Proefschrift lieshout by Nicole Nijhuis

Cell-specific Cell-specific recognition recognition receptor receptor signaling signaling inin antibacterial antibacterial defense defense Miriam Miriam H.P. H.P. vanvan Lieshout Lieshout

Cell-specific pattern recognition receptor signaling in antibacterial defense

Miriam H.P. van Lieshout

Colofon Cell-specific pattern recognition receptor signaling in antibacterial defense Academic Thesis, University of Amsterdam, The Netherlands Copyright ÂŠ 2015 Miriam H.P. van Lieshout, Amsterdam, the Netherlands. All rights reserved. No part of this thesis may be reproduced, stored, or transmitted in any form or by any means without prior permission of the author Printing of this thesis was financially supported by: Psychiatrische Praktijk J.J. van Lieshout; http://www.vvpao.nl. Printed by Gildeprint - Enschede Cover design: Cherry blossom resembles lung tissue sections. Original photograph taken in Philadelphia by Miriam van Lieshout Editing by Joost van Lieshout

Cell-specific pattern recognition receptor signaling in antibacterial defense ACADEMISCH PROEFSCHRIFT ter verkrijging van de graad van doctor aan de Universiteit van Amsterdam op gezag van de Rector Magnificus prof. dr. D.C. van den Boom ten overstaan van een door het College voor Promoties ingestelde commissie, in het openbaar te verdedigen in de Agnietenkapel op donderdag 24 september 2015, te 10.00 uur door Miriam Hanneke Petra van Lieshout geboren te Amsterdam

Promotiecommissie Promotor: prof. dr. T. van der Poll Copromotores: dr. C. van â&#x20AC;&#x2DC;t Veer dr. A.F. de Vos

Universiteit van Amsterdam

Overige leden: prof. dr. J.M. Prins prof. dr. K. Brinkman prof. dr. M.G. Netea prof. dr. R.J.M. ten Berge dr. W.J. Wiersinga dr. J.C. Leemans Faculteit der Geneeskunde

Universiteit van Amsterdam OLVG Radboud Universiteit Nijmegen Universiteit van Amsterdam Universiteit van Amsterdam Universiteit van Amsterdam

Universiteit van Amsterdam Universiteit van Amsterdam

Chapter 1

Table of contents 7

General introduction and outline of this thesis

Chapter 2

Host defense during Klebsiella pneumonia relies on hematopoietic expressed TLR4 and TLR2 European Respiratory Journal, 2011

Chapter 3

Differential roles of MyD88 and TRIF in hematopoietic and resident cells during murine gram-negative pneumonia Journal of Infectious Diseases, 2012

Chapter 4

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis PLoS Pathogens, 2014

Chapter 5

TIR-domain-containing adaptor-inducing interferon-Î˛ (TRIF) mediates antibacterial defense during gram-negative pneumonia by inducing Interferon-Îł Journal of Innate Immunity, 2015

Chapter 6

103

Lung epithelial MyD88 drives early pulmonary clearance of Pseudomonas aeruginosa by a flagellin dependent mechanism Submitted for publication

Chapter 7

123

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia American Journal of Respiratory Cell and Molecular Biology, 2014

Chapter 8

165

ASC and NLRP3 impair host defense during lethal pneumonia caused by highly virulent serotype 3 Streptococcus pneumoniae Submitted for publication

Chapter 9

189

Single immunoglobulin interleukin-1 receptor related molecule impairs host defense during pneumonia and sepsis caused by Streptococcus pneumoniae Journal of Innate Immunity, 2014

Chapter 10

209

TLR4 inhibition impairs bacterial clearance in a therapeutic setting in murine abdominal sepsis Inflammation Research,2014

Chapter 11

223

Summary and general discussion of this thesis “Cell-specific pattern recognition receptor signaling in antibacterial defense”

Chapter 12

235

Nederlandse samenvatting en discussie van dit proefschrift “Cell-specific pattern recognition receptor signaling in antibacterial defense”

Addendum

255

Curriculum vitae

256

PhD portfolio

257

Dankwoord

261

Chapter 1 General introduction and outline of this thesis

Miriam H.P. van Lieshout

Chapter 1

General introduction and outline of this thesis

General introduction Infection and sepsis Sepsis is the syndrome of infection complicated by acute organ dysfunction. It is a leading cause of morbidity and mortality worldwide, both in developing countries and the developed world (1, 2). The most common cause of sepsis is pneumonia, followed by abdominal and genitourinary infections (1). Despite all efforts in the past decades to improve the outcome of sepsis, the mortality rate remains as high as 20-40 % (3-5). Moreover, antimicrobial resistance rates of common pathogens are increasing, including Streptococcus (S.) pneumoniae. Especially alarming is the emergence of extended-spectrum Î˛-lactamases producing strains of Escherichia (E.) coli and Klebsiella (K.) pneumoniae, as well as multi-drug resistant and extreme drug resistant strains of other gram-negative bacteria including Pseudomonas (P.) aeruginosa (5-10). These resistant strains are often confined to health-care institutions, but some become increasingly prevalent in the community. Infections with these resistant pathogens are associated with increased morbidity, mortality and economic costs. These major health concerns call for the development of new therapies against infection and sepsis. Host defense against infection In pneumonia, there are several lines of host defense against the infective pathogen. The respiratory tract is a large surface within the body that mediates gas exchange with the environment but therefore also is a large area of potential contact with pathogens. The first lines of defense consist of the pseudostratified mucosal barrier of the tracheobronchial tree where ciliated and secretory cells (goblet and Clara cells) work together to protect the airways. Mucus produced by secretory cells contains antimicrobial peptides, enzymes and surfactant proteins and entraps pollutants and pathogens, after which it is transported by the cilia in ascending direction. The lower airway surface is covered by type I alveolar cells that have gas exchange as primary function and type II alveolar cells that maintain the alveolar space by the secretion of several surfactant proteins that also serve an antimicrobial function by opsonizing pathogens (11). When these lines of defense fail, the innate and adaptive immune systems are the next defense mechanisms to induce an antibacterial response to eliminate pathogens (12, 13). Alveolar macrophages and dendritic cells reside in the lungs and therefore function in addition to airway epithelial cells as sentinel cells of the innate immunes system (12, 13). When pathogens are detected, sentinel cells attract larger numbers of phagocytes such as neutrophils from the bloodstream to the site of infection in interplay with respiratory epithelial cells, via the secretion of various chemokines and cytokines, thereby also enhancing their effector functions of phagocytosis and killing (12-15). Moreover, alveolar macrophages themselves can also engulf pathogens and apoptotic neutrophils and in this way eliminate pathogens and contribute to the resolution of pneumonia (16). The secretion of interferon (IFN)-Îł by both cells of the innate and adaptive system is known to 8

General introduction and outline of this thesis

Figure 1: Cell types involved in the innate immune response in the alveolar space.

powerfully enhance these macrophage effector functions (17). The abdominal cavity is normally sterile and as such, has fewer defense mechanisms than the respiratory tract. In the case of infectious peritonitis, reticuloendothelial cells, mesothelial cells, and peritoneal macrophages detect pathogens and contribute to host defense. Role of pattern recognition receptors Innate immune cells detect pathogens by recognition of conserved microbial molecules (pathogen associated molecular patterns or PAMPS) with sensors called pattern recognition receptors (PRRs) (13, 18, 19). Toll-like receptors (TLRs) prominently feature herein, detecting a variety of conserved microbial patterns as well as â&#x20AC;&#x153;danger signalsâ&#x20AC;? released from host cells as a consequence of injurious inflammation. As such, TLRs play an important role in the initiation and amplification of the host response (14, 18, 19). To date, 10 human TLRs have been identified, located on the cell surface (TLR1,2,4-6,10) or in membranes of endosomes and lysosomes (TLR3, TLR7-9) (14, 18, 19). Different microbial and endogenous ligands have been identified for most TLRs and for each one pathogen there are a variety of molecular patterns detectable by different TLRs (14). 9

Chapter 1

Once TLRs are activated, they propagate their signal via intracellular adapters, activating nuclear factor kappa-light-chain-enhancer of activated B cells (NFκB) and Mitogen-activated protein kinases (MAP kinases). The universal adapter for all TLRs - except TLR3 - is myeloid differentiation primary response gene (MyD)88 (18, 19). In addition, MyD88 mediates IL-1β and IL-18 receptor signaling (20). TIRdomain-containing adapter-inducing interferon-β (TRIF) is the sole adapter for TLR3 and contributes to TLR4 signaling (18, 19). Unrestrained activation of TLRs may result in excessive inflammation and collateral tissue damage. Several negative regulators of TLR signaling are known to balance the inflammatory response. Single immunoglobulin IL-1 receptor-related molecule (SIGIRR) has been shown to inhibit TLR-dependent and IL-1R like receptor induced NFκB activation (21). Another family of PRRs are the nucleotide-binding and oligomerization-domain proteins, with NLR family, pyrin domain containing 3 (NLRP3) as the best characterized member. NLRP3, together with the adaptor protein apoptosisassociated speck-like protein containing a CARD (ASC), takes part in the formation of large multi-protein complexes called inflammasomes. These are crucial to the antimicrobial response because they activate caspase-1 dependent IL-1β and IL-18 maturation. Caspase-1 activation also leads to inflammatory cell death or pyroptosis (22, 23). Activation of caspase-1 by the NLRP3 inflammasome is a multi-signal process, requiring at least two signals (22, 23). Although a brisk initiation and amplification of the host response is indisputably important, the inflammatory response may, also dependent on the virulence of the pathogen and host characteristics, induce local tissue injury and lead to a systemic inflammatory response that may contribute to organ injury (1). Little is known about the contributions of non-hematopoietic cells (i.e., lung epithelial cells, the vascular endothelium) to both the antibacterial response and potentially harmful side effects of the inflammatory response. Some evidence indicates that tissue and organ injury during sepsis may be negatively impacted by endothelial induced inflammation (24-27). With the aim of attenuating excessive inflammation during sepsis, several anti-TLR therapies have been developed in recent years (28, 29). Until now, results have only been very modest and not proven of additional value in the clinical setting. Infection models used in this thesis and the innate immune receptors involved Infections of the respiratory tract are in the top ten causes of death both nationally and globally; mortality affects mainly children and the elderly (30-32). In pneumonia, community-acquired infection is distinguished from pneumonia that is associated to hospital admission, mechanical ventilation or out-of hospital health care settings since they are very different with regard to microbial etiology and prognosis. Klebsiella (K.) pneumoniae is a gram-negative pathogen of the Enterobacteriaceae family that frequently causes pneumonia and blood-stream infections, especially in hospitals and health-care related settings (33-35). In experimental K. pneumoniae respiratory infection particularly TLR4, that detects the gram-negative cell wall 10

General introduction and outline of this thesis

constituent lipopolysaccharide (LPS) as well as various endogenous danger signals, and TLR9, that detects bacterial DNA, were found to be protective (3638). Universal TLR-adapter MyD88 is of crucial importance for host defense and survival in K. pneumonia respiratory infection (39). The adapter protein TRIF mediates TLR3 signaling in response to double-strand RNA and contributes to TLR4 dependent signaling. TRIF has been shown to be required for optimal host resistance in Klebsiella pneumonia (39). Streptococcus (S.) pneumoniae is the most frequent cause of communityacquired pneumonia and responsible for a considerable part of the health burden that pneumosepsis places on society, especially among young children and the elderly (30, 34). More than 90 serotypes have been identified, several of which cause invasive and severe disease and mortality (40). In this thesis, two different pneumococcal strains are used: a serotype 3 strain (ATCC 6303) that causes lethal disease in mice after low-dose infection and a serotype 2 (D39) strain that is used at a high dose in experimental pneumonia thereby inducing about 25% mortality in immunocompetent mice (41-49). In humans, infections caused by serotype 3 pneumococci are common and associated with a complicated course and an increased risk of death (40, 50, 51), whereas infections with the serotype 2 are uncommon in the western world. Several TLRs contribute to the host response during pneumococcal infection. TLR2 that detects lipoteichoic acid (LTA), a constituent of the pneumococcal cell wall (46, 52), has a modest role in the cytokine response during pneumococcal pneumonia after infection with a serotype 3 S. pneumoniae (47). TLR4 contributes to host defense during S. pneumoniae pneumonia by recognition of pneumolysin and serves a protective role during pneumococcal infection of the lower airways (37, 53). In the absence of pneumolysin, TLR2 limited bacterial growth during infection with the serotype 2 D39 pneumococcus (45). Finally, TLR9 deficient (Tlr9/) mice showed enhanced bacterial growth and dissemination after induction of pneumococcal pneumonia (54). Myd88-/- mice had a profoundly enhanced growth of pneumococci and a strongly reduced survival after intranasal infection with a serotype 4 S.pneumoniae strain (55). In recent years, the importance of the inflammasome components NLRP3, a member of the NOD like receptor family, and/ or the adapter protein ASC for the host response during pneumococcal pneumonia was demonstrated in studies using S. pneumoniae strains with a relatively low virulence (42, 43, 56). This protective effect is hypothesized to be dependent on the activation of NLRP3 by pneumolysin, a crucial virulence factor expressed by S. pneumoniae (43, 57, 58). Respiratory tract infection with Pseudomonas (P.) aeruginosa, a flagellated gramnegative opportunistic pathogen, often occurs in hospitalized and/or mechanically ventilated patients and frequently results in severe disease (59, 60). Moreover this pathogen tends to induce chronic lung inflammation after colonization of the airways of patients suffering from chronic lung diseases thereby causing further decline in pulmonary function (61, 62). In experimental models of Pseudomonas infection the importance of TLR dependent clearance of this pathogen was clearly 11

Chapter 1

illustrated. MyD88 deficient (Myd88-/-) mice were hypersusceptible to Pseudomonas pneumonia (63-65). TLR2, TLR4 and TLR5 (that detects flagellin) had redundant functions, but control of the bacterium required detection of either LPS or flagellin in a way that also depended on the infectious inoculum (66, 67). Abdominal infection, together with urinary tract infection is the second most common cause of sepsis and Escherichia (E.) coli is among the most frequently cultured gram-negative bacteria in sepsis and peritonitis patients (68-70). Typically, peritonitis results from perforation of a hollow abdominal organ, spilling gut content into the normally sterile cavity leading to polymicrobial infection (secondary peritonitis). In certain susceptible patients however primary peritonitis may result from bacterial translocation (70-73). Since bacteria can quickly spread via the bloodstream from the peritoneal cavity, this can lead to the rapid onset of systemic inflammation and sepsis, resulting in a very high mortality rate of up to 60% (71, 72). Complicating treatment, E. coli has increasing extended antimicrobial resistance rates, both in hospitals and in the community (9). The experimental model of E. coli peritonitis used in this thesis has a low infectious inoculum of the virulent O18:K1 strain in contrast to frequently used models with high infectious doses of less virulent bacterial strains. In the model here used, TLR4 is important for the initial host defense, while during later stage infection the role of TLR2 becomes significant (74).

Aim and outline of this thesis The general aim of this thesis is to advance our understanding of TLR-dependent, especially MyD88-dependent signaling in experimental models of sepsis, with a focus on pneumonia, its most common cause. We also explored the role of the “inflammasome”’ during pneumococcal pneumonia. Secondly, since innate immune sensors are widely distributed among different cell types in the airways and body, we aimed to gain insight in the contribution of different cell types and body compartments to TLR- and MyD88-dependent signaling during infection and sepsis. In chapter 2 we dissected the role of TLR2 and 4 and in chapter 3 of MyD88 and TRIF dependent signaling in hematopoietic and non-hematopoietic cells during K. pneumoniae airway infection by the use of bone-marrow chimeras. In chapter 4 we further dissected the role of MyD88 dependent signaling during Klebsiella pneumosepsis in myeloid and endothelial cells by the use of tissue specific conditional knockouts for MyD88. In chapter 5 we studied the importance of TRIF in the secretion of interferon (IFN)-γ in response to K. pneumoniae and the potential of recombinant IFN-γ to restore the impaired antibacterial defense in TRIF-deficient mice. In chapter 6 we studied the role of MyD88 dependent signaling in lung epithelial cells versus myeloid cells during respiratory tract infection with P. aeruginosa, and in addition used bone marrow chimeric mice to establish the different contribution of TLR5 on hematopoietic versus non-hematopoietic cells to the innate host response to this flagellated bacterium. In chapter 7 we investigated the differential role of NLRP3 and ASC during 12

General introduction and outline of this thesis

pneumococcal pneumonia with the serotype 2 D 39 strain. In more depth we analyzed the early in vivo immune response by whole-genome transcriptional profiling and the differential contribution of ASC and NLRP3. In chapter 8 we demonstrate an opposite role of NLRP3 and ASC compared to the previous chapter during infection with a serotype 3 pneumococcal strain. In addition, we investigated the role of TLR-dependent signaling during infection with this strain by the use of MyD88 deficient mice. Furthermore, in chapter 9 we investigated the role of SIGIRR, a negative regulator of TLR and IL-1 receptor dependent signaling, during lethal pneumococcal pneumonia. In chapter 10 we studied the effects of anti-TLR4 therapy on anti-bacterial defense, inflammatory response and organ injury in a delayed treatment model of E. coli peritonitis.

Chapter 1

References

1. Angus DC, van der Poll T. Severe Sepsis and Septic Shock. N Engl J Med 2013;369:840-851.

2. van der Poll T, Opal SM. Host-Pathogen Interactions in Sepsis. Lancet Infect Dis 2008;8:3243.

3. van der Poll T. Immunotherapy of Sepsis. Lancet Infect Dis 2001;1:165-174.

4. Webster NR, Galley HF. Immunomodulation in the Critically Ill. Br J Anaesth 2009;103:70-81.

5. World Health Organisation. Antimicrobial Resistance: Global Report on Surveillance.; 2014.

6. Coque TM, Baquero F, Canton R. Increasing Enterobacteriaceae in Europe. Euro Surveill 2008;13.

7. Giamarellou H. Multidrug Resistance in Gram-Negative Bacteria That Produce ExtendedSpectrum Beta-Lactamases (ESBLs). Clin Microbiol Infect 2005;11 Suppl 4:1-16.

8. Schwaber MJ, Carmeli Y. Mortality and Delay in Effective Therapy Associated With ExtendedSpectrum Beta-Lactamase Production in Enterobacteriaceae Bacteraemia: a Systematic Review and Meta-Analysis. J Antimicrob Chemother 2007;60:913-920.

9. Rossolini GM, Arena F, Pecile P, Pollini S. Update on the Antibiotic Resistance Crisis. Curr Opin Pharmacol 2014;18C:56-60.

10. Mesaros N, Nordmann P, Plesiat P, Roussel-Delvallez M, Van EJ, Glupczynski Y, Van LY, Jacobs F, Lebecque P, Malfroot A, et al. Pseudomonas Aeruginosa: Resistance and Therapeutic Options at the Turn of the New Millennium. Clin Microbiol Infect 2007;13:560578.

11. Camelo A, Dunmore R, Sleeman MA, Clarke DL. The Epithelium in Idiopathic Pulmonary Fibrosis: Breaking the Barrier. Front Pharmacol 2014;4:173.

12. Dickson RP, Erb-Downward JR, Huffnagle GB. The Role of the Bacterial Microbiome in Lung Disease. Expert Rev Respir Med 2013;7:245-257.

13. Opitz B, van Laak V, Eitel J, Suttorp N. Innate Immune Recognition in Infectious and Noninfectious Diseases of the Lung. Am J Respir Crit Care Med 2010;181:1294-1309.

14. Mizgerd JP. Acute Lower Respiratory Tract Infection. N Engl J Med 2008;358:716-727.

15. Parker D, Prince A. Innate Immunity in the Respiratory Epithelium. Am J Respir Cell Mol Biol 2011;45:189-201.

16. Aggarwal NR, King LS, Dâ&#x20AC;&#x2122;Alessio FR. Diverse Macrophage Populations Mediate Acute Lung Inflammation and Resolution. Am J Physiol Lung Cell Mol Physiol 2014;306:L709-L725.

17. Schroder K, Hertzog PJ, Ravasi T, Hume DA. Interferon-Gamma: an Overview of Signals, Mechanisms and Functions. J Leukoc Biol 2004;75:163-189.

18. Beutler BA. TLRs and Innate Immunity. Blood 2009;113:1399-1407.

Prevalence

ESBL-Producing

General introduction and outline of this thesis

19. Kawai T, Akira S. Toll-Like Receptors and Their Crosstalk With Other Innate Receptors in Infection and Immunity. Immunity 2011;34:637-650.

20. Adachi O, Kawai T, Takeda K, Matsumoto M, Tsutsui H, Sakagami M, Nakanishi K, Akira S. Targeted Disruption of the MyD88 Gene Results in Loss of IL-1- and IL-18-Mediated Function. Immunity 1998;9:143-150.

21. Riva F, Bonavita E, Barbati E, Muzio M, Mantovani A, Garlanda C. TIR8/SIGIRR Is an Interleukin-1 Receptor/Toll Like Receptor Family Member With Regulatory Functions in Inflammation and Immunity. Front Immunol 2012;3:322.

22. Strowig T, Henao-Mejia J, Elinav E, Flavell R. Inflammasomes in Health and Disease. Nature 2012;481:278-286.

23. Franchi L, Munoz-Planillo R, Nunez G. Sensing and Reacting to Microbes Through the Inflammasomes. Nat Immunol 2012;13:325-332.

24. Ding J, Song D, Ye X, Liu SF. A Pivotal Role of Endothelial-Specific NF-KappaB Signaling in the Pathogenesis of Septic Shock and Septic Vascular Dysfunction. J Immunol 2009;183:40314038.

25. Song D, Ye X, Xu H, Liu SF. Activation of Endothelial Intrinsic NF-{Kappa}B Pathway Impairs Protein C Anticoagulation Mechanism and Promotes Coagulation in Endotoxemic Mice. Blood 2009;114:2521-2529.

26. Xu H, Ye X, Steinberg H, Liu SF. Selective Blockade of Endothelial NF-KappaB Pathway Differentially Affects Systemic Inflammation and Multiple Organ Dysfunction and Injury in Septic Mice. J Pathol 2010;220:490-498.

27. Ye X, Ding J, Zhou X, Chen G, Liu SF. Divergent Roles of Endothelial NF-KappaB in Multiple Organ Injury and Bacterial Clearance in Mouse Models of Sepsis. J Exp Med 2008;205:13031315.

28. Hennessy EJ, Parker AE, Oâ&#x20AC;&#x2122;Neill LA. Targeting Toll-Like Receptors: Emerging Therapeutics? Nat Rev Drug Discov 2010;9:293-307.

29. Savva A, Roger T. Targeting Toll-Like Receptors: Promising Therapeutic Strategies for the Management of Sepsis-Associated Pathology and Infectious Diseases. Front Immunol 2013;4:387.

30. World Health Organisation. World Health Observatory, Top Ten Causes of Death 2012.; 2012.

31. Centraal Bureau voor de Statistiek. Statline:Overledenen; Belangrijke Doodsoorzaken.; 2014.

32. Center for Disease Control and prevention. National Vital Statitistics Reports: Deaths: Final Data for 2010.; 2013.

33. Kollef MH, Shorr A, Tabak YP, Gupta V, Liu LZ, Johannes RS. Epidemiology and Outcomes of Health-Care-Associated Pneumonia: Results From a Large US Database of Culture-Positive Pneumonia. Chest 2005;128:3854-3862.

34. Welte T, Torres A, Nathwani D. Clinical and Economic Burden of Community-Acquired Pneumonia Among Adults in Europe. Thorax 2012;67:71-79.

Chapter 1

35. Zahar JR, Timsit JF, Garrouste-Org, Francais A, Vesin A, Descorps-Declere A, Dubois Y, Souweine B, Haouache H, Goldgran-Toledano D, et al. Outcomes in Severe Sepsis and Patients With Septic Shock: Pathogen Species and Infection Sites Are Not Associated With Mortality. Crit Care Med 2011;39:1886-1895.

36. Bhan U, Lukacs NW, Osterholzer JJ, Newstead MW, Zeng X, Moore TA, McMillan TR, Krieg AM, Akira S, Standiford TJ. TLR9 Is Required for Protective Innate Immunity in Gram-Negative Bacterial Pneumonia: Role of Dendritic Cells. J Immunol 2007;179:3937-3946.

37. Branger J, Knapp S, Weijer S, Leemans JC, Pater JM, Speelman P, Florquin S, van der Poll T. Role of Toll-Like Receptor 4 in Gram-Positive and Gram-Negative Pneumonia in Mice. Infect Immun 2004;72:788-794.

38. Schurr JR, Young E, Byrne P, Steele C, Shellito JE, Kolls JK. Central Role of Toll-Like Receptor 4 Signaling and Host Defense in Experimental Pneumonia Caused by GramNegative Bacteria. Infect Immun 2005;73:532-545.

39. Cai S, Batra S, Shen L, Wakamatsu N, Jeyaseelan S. Both TRIF- and MyD88-Dependent Signaling Contribute to Host Defense Against Pulmonary Klebsiella Infection. J Immunol 2009;183:6629-6638.

40. Weinberger DM, Harboe ZB, Sanders EA, Ndiritu M, Klugman KP, Ruckinger S, Dagan R, Adegbola R, Cutts F, Johnson HL, et al. Association of Serotype With Risk of Death Due to Pneumococcal Pneumonia: a Meta-Analysis. Clin Infect Dis 2010;51:692-699.

41. Ogunniyi AD, LeMessurier KS, Graham RM, Watt JM, Briles DE, Stroeher UH, Paton JC. Contributions of Pneumolysin, Pneumococcal Surface Protein A (PspA), and PspC to Pathogenicity of Streptococcus Pneumoniae D39 in a Mouse Model. Infect Immun 2007;75:1843-1851.

42. Fang R, Tsuchiya K, Kawamura I, Shen Y, Hara H, Sakai S, Yamamoto T, Fernandes-Alnemri T, Yang R, Hernandez-Cuellar E, et al. Critical Roles of ASC Inflammasomes in Caspase-1 Activation and Host Innate Resistance to Streptococcus Pneumoniae Infection. J Immunol 2011;187:4890-4899.

43. McNeela EA, Burke A, Neill DR, Baxter C, Fernandes VE, Ferreira D, Smeaton S, El-Rachkidy R, McLoughlin RM, Mori A, et al. Pneumolysin Activates the NLRP3 Inflammasome and Promotes Proinflammatory Cytokines Independently of TLR4. PLoS Pathog 2010;6:e1001191.

44. Dessing MC, Knapp S, Florquin S, de Vos AF, van der Poll T. CD14 Facilitates Invasive Respiratory Tract Infection by Streptococcus Pneumoniae. Am J Respir Crit Care Med 2007;175:604-611.

45. Dessing MC, Florquin S, Paton JC, van der Poll T. Toll-Like Receptor 2 Contributes to Antibacterial Defence Against Pneumolysin-Deficient Pneumococci. Cell Microbiol 2008;10:237-246.

46. Dessing MC, Schouten M, Draing C, Levi M, von AS, van der Poll T. Role Played by Toll-Like Receptors 2 and 4 in Lipoteichoic Acid-Induced Lung Inflammation and Coagulation. J Infect Dis 2008;197:245-252.

47. Knapp S, Wieland CW, van â&#x20AC;&#x2DC;, V, Takeuchi O, Akira S, Florquin S, van der Poll T. Toll-Like Receptor 2 Plays a Role in the Early Inflammatory Response to Murine Pneumococcal Pneumonia but Does Not Contribute to Antibacterial Defense. J Immunol 2004;172:31323138.

General introduction and outline of this thesis

48. van der Windt GJ, Hoogendijk AJ, Schouten M, Hommes TJ, de Vos AF, Florquin S, van der Poll T. Osteopontin Impairs Host Defense During Pneumococcal Pneumonia. J Infect Dis 2011;203:1850-1858.

49. Rijneveld AW, Florquin S, Branger J, Speelman P, Van Deventer SJ, van der PT. TNF-Alpha Compensates for the Impaired Host Defense of IL-1 Type I Receptor-Deficient Mice During Pneumococcal Pneumonia. J Immunol 2001;167:5240-5246.

50. Benfield T, Skovgaard M, Schonheyder HC, Knudsen JD, Bangsborg J, Ostergaard C, Slotved HC, Konradsen HB, Thomsen RW, Lambertsen L. Serotype Distribution in NonBacteremic Pneumococcal Pneumonia: Association With Disease Severity and Implications for Pneumococcal Conjugate Vaccines. PLoS ONE 2013;8:e72743.

51. Jansen AG, Rodenburg GD, van der Ende A, van AL, Veenhoven RH, Spanjaard L, Sanders EA, Hak E. Invasive Pneumococcal Disease Among Adults: Associations Among Serotypes, Disease Characteristics, and Outcome. Clin Infect Dis 2009;49:e23-e29.

52. Schroder NW, Morath S, Alexander C, Hamann L, Hartung T, Zahringer U, Gobel UB, Weber JR, Schumann RR. Lipoteichoic Acid (LTA) of Streptococcus Pneumoniae and Staphylococcus Aureus Activates Immune Cells Via Toll-Like Receptor (TLR)-2, Lipopolysaccharide-Binding Protein (LBP), and CD14, Whereas TLR-4 and MD-2 Are Not Involved. J Biol Chem 2003;278:15587-15594.

53. Malley R, Henneke P, Morse SC, Cieslewicz MJ, Lipsitch M, Thompson CM, Kurt-Jones E, Paton JC, Wessels MR, Golenbock DT. Recognition of Pneumolysin by Toll-Like Receptor 4 Confers Resistance to Pneumococcal Infection. Proc Natl Acad Sci U S A 2003;100:19661971.

54. Albiger B, Dahlberg S, Sandgren A, Wartha F, Beiter K, Katsuragi H, Akira S, Normark S, Henriques-Normark B. Toll-Like Receptor 9 Acts at an Early Stage in Host Defence Against Pneumococcal Infection. Cell Microbiol 2007;9:633-644.

55. Albiger B, Sandgren A, Katsuragi H, Meyer-Hoffert U, Beiter K, Wartha F, Hornef M, Normark S, Normark BH. Myeloid Differentiation Factor 88-Dependent Signalling Controls Bacterial Growth During Colonization and Systemic Pneumococcal Disease in Mice. Cell Microbiol 2005;7:1603-1615.

56. Witzenrath M, Pache F, Lorenz D, Koppe U, Gutbier B, Tabeling C, Reppe K, Meixenberger K, Dorhoi A, Ma J, et al. The NLRP3 Inflammasome Is Differentially Activated by Pneumolysin Variants and Contributes to Host Defense in Pneumococcal Pneumonia. J Immunol 2011;187:434-440.

57. Kadioglu A, Weiser JN, Paton JC, Andrew PW. The Role of Streptococcus Pneumoniae Virulence Factors in Host Respiratory Colonization and Disease. Nat Rev Microbiol 2008;6:288-301.

58. Hoegen T, Tremel N, Klein M, Angele B, Wagner H, Kirschning C, Pfister HW, Fontana A, Hammerschmidt S, Koedel U. The NLRP3 Inflammasome Contributes to Brain Injury in Pneumococcal Meningitis and Is Activated Through ATP-Dependent Lysosomal Cathepsin B Release. J Immunol 2011;187:5440-5451.

59. Chastre J, Fagon JY. Ventilator-Associated Pneumonia. Am J Respir Crit Care Med 2002;165:867-903.

Chapter 1

60. Jones RN. Microbial Etiologies of Hospital-Acquired Bacterial Pneumonia and VentilatorAssociated Bacterial Pneumonia. Clin Infect Dis 2010;51 Suppl 1:S81-S87.

61. King PT, Holdsworth SR, Freezer NJ, Villanueva E, Holmes PW. Microbiologic Follow-Up Study in Adult Bronchiectasis. Respir Med 2007;101:1633-1638.

62. Lambiase A, Raia V, Del PM, Sepe A, Carnovale V, Rossano F. Microbiology of Airway Disease in a Cohort of Patients With Cystic Fibrosis. BMC Infect Dis 2006;6:4.

63. Power MR, Peng Y, Maydanski E, Marshall JS, Lin TJ. The Development of Early Host Response to Pseudomonas Aeruginosa Lung Infection Is Critically Dependent on Myeloid Differentiation Factor 88 in Mice. J Biol Chem 2004;279:49315-49322.

64. Skerrett SJ, Liggitt HD, Hajjar AM, Wilson CB. Cutting Edge: Myeloid Differentiation Factor 88 Is Essential for Pulmonary Host Defense Against Pseudomonas Aeruginosa but Not Staphylococcus Aureus. J Immunol 2004;172:3377-3381.

65. Skerrett SJ, Wilson CB, Liggitt HD, Hajjar AM. Redundant Toll-Like Receptor Signaling in the Pulmonary Host Response to Pseudomonas Aeruginosa. Am J Physiol Lung Cell Mol Physiol 2007;292:L312-L322.

66. Morris AE, Liggitt HD, Hawn TR, Skerrett SJ. Role of Toll-Like Receptor 5 in the Innate Immune Response to Acute P. Aeruginosa Pneumonia. Am J Physiol Lung Cell Mol Physiol 2009;297:L1112-L1119.

67. Ramphal R, Balloy V, Jyot J, Verma A, Si-Tahar M, Chignard M. Control of Pseudomonas Aeruginosa in the Lung Requires the Recognition of Either Lipopolysaccharide or Flagellin. J Immunol 2008;181:586-592.

68. Angus DC. The Search for Effective Therapy for Sepsis: Back to the Drawing Board? JAMA 2011;306:2614-2615.

69. Ranieri VM, Thompson BT, Barie PS, Dhainaut JF, Douglas IS, Finfer S, Gardlund B, Marshall JC, Rhodes A, Artigas A, et al. Drotrecogin Alfa (Activated) in Adults With Septic Shock. N Engl J Med 2012;366:2055-2064.

70. Brook I. Microbiology and Management of Abdominal Infections. Dig Dis Sci 2008;53:25852591.

71. McClean KL, Sheehan GJ, Harding GK. Intraabdominal Infection: a Review. Clin Infect Dis 1994;19:100-116.

72. Wiest R, Krag A, Gerbes A. Spontaneous Bacterial Peritonitis: Recent Guidelines and Beyond. Gut 2012;61:297-310.

73. Cheong HS, Kang CI, Lee JA, Moon SY, Joung MK, Chung DR, Koh KC, Lee NY, Song JH, Peck KR. Clinical Significance and Outcome of Nosocomial Acquisition of Spontaneous Bacterial Peritonitis in Patients With Liver Cirrhosis. Clin Infect Dis 2009;48:1230-1236.

74. van â&#x20AC;&#x2DC;t Veer C, van den Pangaart PS, Kruijswijk D, Florquin S, de Vos AF, van der Poll T. Delineation of the Role of Toll-Like Receptor Signaling During Peritonitis by a Gradually Growing Pathogenic Escherichia Coli. J Biol Chem 2011;286:36603-36618.

Chapter 2 Host defense during Klebsiella pneumonia relies on hematopoietic expressed TLR4 and TLR2 European Respiratory Journal 2011 Apr;37(4):848-57 DOI: 10.1183/09031936.00076510 Catharina W. Wieland 1, 2, 3, 4 Miriam H.P. van Lieshout 1, 2 Arie J. Hoogendijk 1, 2 Tom van der Poll 1, 2 Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands: 1 Center of Experimental and Molecular Medicine 2 Center of Infection and Immunity 3 Laboratory of Experimental Intensive Care and Anesthesiology 4 Department of Intensive Care

Chapter 2

Abstract In this study the relative roles of Toll-like receptor (TLR)2 and TLR4 were investigated independently and together. Moreover, we studied the role of hematopoietic compartment in anti-Klebsiella host defense. We infected TLR2, TLR4 single- and TLR2x4 double knock-out (KO) animals with different doses of Klebsiella pneumoniae. In addition, bone marrow chimeric mice were created and infected. TLR4 played a more prominent role in antibacterial defense than TLR2, considering that only TLR4 KO mice demonstrated enhanced bacterial growth in lungs and spleen 24 h after infection with 3x103 colony-forming units of Klebsiella compared with wild-type (WT) mice. In late stage infection or after exposure to a higher infectious dose, bacterial counts in lungs of TLR2 KO animals were elevated compared to WT mice and TLR2x4 KO animals were more susceptible to infection than TLR4 KO mice. TLR signaling on cells of hematopoietic origin is of primary importance in host defense against K. pneumoniae. These data suggest that 1) TLR4 drives the antibacterial host response after induction of pneumonia with relatively low Klebsiella doses; 2) TLR2 becomes involved at a later phase of the infection and/or upon exposure to higher bacterial burdens and (3) hematopoietic TLR2 and TLR 4 are important for an adequate host response during Klebsiella pneumonia.

Host defense during Klebsiella pneumonia relies on hematopoietic expressed TLR4 and TLR2

Introduction Gram-negative pneumonia is a common and serious illness that is a major cause of morbidity and mortality in humans. Klebsiella (K.) pneumoniae is a frequently isolated causative pathogen in nosocomial lower respiratory tract infection (1, 2). The increasing microbial resistance to antibiotics, resulting in therapy failure and higher mortality rates, is an issue of major concern (1). Therefore, it is important to gain more insight into the pathogenesis of pneumonia. Toll-like receptors (TLRs) recognize pathogens, resulting in onset of the inflammatory response (3). TLRs are expressed in both cells of hematopoietic origin and stromal cells (e.g. lung epithelium). When K. pneumoniae enters the lung, bacteriumspecific TLRs are activated, triggering the release of cytokines and chemokines that attract and activate neutrophils. In the best case scenario, these neutrophils kill all bacteria after ingestion. TLR4 has been implicated as the most important TLR for the recognition of K. pneumoniae by virtue of its capacity to sense lipopolysaccharide (LPS) present in the outer membrane of this Gram-negative pathogen (3). Indeed, in previous research we found that TLR4 mutant mice were highly susceptible to pulmonary infection with K. pneumoniae regardless of the infectious dose (4). The indispensable role of TLR4 for antibacterial defense against Klebsiella has subsequently been confirmed by other studies using different serotypes and different infection models (5-7). Notably, evidence indicates that other TLRs also contribute to host defense against Klebsiella pneumonia. Mice deficient for TLR9 (which is expressed within endosomes and recognizes bacterial DNA (3) had an impaired host defense after infection with K. pneumoniae via their airways, due to reduced dendritic cell accumulation and dendritic cell and macrophage activation in their lungs (8). Moreover, mice deficient for MyD88 (myeloid differentiation primary response gene 88; which mediates signaling of all TLRs except TLR3) or TIRAP (Toll-IL-1R domain-containing adaptor protein; an essential adapter for TLR1, TLR2, TLR4, and TLR6 signaling) displayed a diminished antibacterial defense during Klebsiella pneumonia (9, 10). In the present study, we tested the hypothesis that TLR2, in conjunction with TLR4, is an important player in the protective immune response during respiratory tract infection by K. pneumoniae. In theory, TLR2 can contribute to the recognition of Klebsiella through an interaction with bacterial lipoproteins (3). Moreover, macrophages and dendritic cells can be activated by the K. pneumoniae pathogen associated molecular pattern (PAMP) outer membrane protein A (OmpA) through TLR2 (11). In addition, a recent study demonstrated that both TLR4 and TLR2 mRNA and protein are upregulated after stimulation of A549 cells (human lung epithelial cells) with K. pneumoniae (12). That study suggested that capsular polysaccharides are Klebsiella PAMPs responsible for TLR upregulation in lung epithelium. In the present study, we infected mice deficient for TLR2, TLR4 or both TLR2 and TLR4 via the airways with K. pneumoniae and studied antibacterial host defense and immune responses. Interestingly, TLR2x4 double KO mice were found to be more susceptible to Klebsiella pneumonia than animals deficient for TLR4 only. This study shows that TLR2 helps antibacterial host defense at a late stage of the 21

Chapter 2

infection and/or upon exposure of the host to high bacterial numbers. Moreover, after creating bone marrow chimeric mice, we found that TLR2 and TLR4 expressed within the hematopoietic compartment are crucial for host defense against this nosocomial pathogen.

Materials and Methods Animals TLR4 (13) and TLR2 KO mice (14) were generously provided by Dr. S. Akira (Research Institute for Microbial Disease, Osaka, Japan). TLR2x4 KO mice were generated by intercrossing TLR2 KO and TLR4 KO mice. All genetically modified mice were back-crossed at least 6 times on a C57Bl/6 genetic background and bred in the animal facility of the Academic Medical Center (University of Amsterdam, Amsterdam, the Netherlands). Age- and sex matched wild-type (WT) C57Bl/6 control mice were obtained from Harlan Nederland (Horst, the Netherlands). Mice were infected at 10-12 weeks of age. The Animal Care and Use Committee of the University of Amsterdam approved all experiments. Induction of pneumonia Pneumonia was induced as described previously (15). Briefly, K. pneumoniae serotype 2 (ATCC 43816; American Type Culture Collection, Manassas, VA) was grown for 3 h to mid-logarithmic phase at 37°C using Tryptic Soy broth (Difco, Detroit, MI). Bacteria were harvested by centrifugation at 1500 x g for 15 min, and washed twice in sterile isotonic saline. Bacteria were then resuspended in sterile isotonic saline at a concentration of 3x103 or 1x104 colony-forming units (CFUs) /50µl, as determined by plating serial 10-fold dilutions on sheep-blood agar plates. Mice were lightly anesthetized by inhalation of isoflurane (Upjohn, Ede, the Netherlands) and bacteria were inoculated intranasally. Determination of bacterial outgrowth Five, 24 or 48 h after infection, mice were anesthetized with medetomidine (Domitor, Pfizer Animal Health Care, Capelle aan der IJssel, the Netherlands) and ketamine (Nimatek, Eurovet Animal Health, Bladel, the Netherlands) and sacrificed by heart puncture. Blood was collected in EDTA containing tubes. Lungs, liver and spleen were harvested and homogenized in sterile saline (weight: volume 1:5) using a tissue homogenizer (Biospec Products, Bartlesville, Oklahoma). CFUs in organ homogenates and blood were determined from serial dilutions plated on blood agar plates, incubated at 37°C for 16 h before colonies were counted. Preparation of lung homogenates for cytokine measurements For cytokine measurements, lungs were excised, weighed and homogenized in saline (weight: volume 1:5). Lung homogenates were diluted 1:2 in lysis buffer containing 300 mM NaCl, 30 mM Tris, 2 mM MgCl2, 2 mM CaCl2, 2 % Triton X-100, and AEBSF (4-(2-aminoethyl)benzeensulfonyl fluoride, Na2EDTA, pepstatin and leupeptin (all 8 µg/ml; pH 7.4) and incubated on ice for 30 min. Homogenates were centrifuged at 1500 × g at 4 °C for 15 min and stored at −20 °C until assays were 22

Host defense during Klebsiella pneumonia relies on hematopoietic expressed TLR4 and TLR2

performed. TNF-α, IL-1β, IL-6, keratinocyte-derived chemokine (KC, also known as CXCL1) and macrophage inflammatory protein 2α (MIP-2, also known as CXCL2) were measured by ELISA using matched antibody pairs according to the manufacturer’s instructions (R&D Systems Inc., Minneapolis, MN, USA). Detection limits were 63 pg/mL for TNF-α, IL-1β, IL-6 and MIP-2, and 15 pg/mL for KC. Histologic examination Lungs were removed and fixed in 10 % formalin in PBS for 24 h and embedded in paraffin. Hematoxylin- and eosin- stained slides were coded and semi-quantitatively scored for inflammatory parameters by a pathologist who was not aware of the origin of the tissue samples. To score lung inflammation and damage, the entire lung surface was analyzed with respect to the following parameters: interstitial inflammation, edema, endothelialitis, bronchitis and pleuritis. Each parameter was graded on a scale of 0 to 4 (0: absent; 1: mild; 2: moderate; 3: severe; 4: very severe). The percentage pneumonia was scored and graded according on a scale of 0 to 4 (0: absent; 1: 5-20% confluent pneumonia; 2: 21-40%; 3: 41-60%; 4: 6180%; 5: 81-100%). The total “lung inflammation score” was expressed as the sum of the scores for each parameter, the maximum being 25 (15,16). Granulocyte staining was performed as described (17). In brief, slides were deparaffinized and rehydrated. Endogenous peroxidase activity was quenched with a solution of 0.1% NaN3 and 0.03% H2O2 (Merck, Whitehouse Station, NJ). Slides were then digested with a solution of 0.25% pepsin (Sigma, St. Louis, MO) in 0.01 M HCl. After being rinsed, the sections were incubated in 10% normal goat serum (Dako, Glostrup, Denmark) and then exposed to fluorescein isothiocyanate (FITC)-labeled antimouse Ly-6-G monoclonal antibody (Pharmingen, San Diego, CA). After washes, slides were incubated with a rabbit anti-FITC antibody (Dako) followed by further incubation with a biotinylated pig anti-rabbit antibody (Dako), rinsed again, incubated in a streptavidine-horseradish peroxidase solution (Dako) and developed using 1% H2O2 and 3,3-diaminobenzidine tetrahydrochoride (Sigma) in Tris-HCl. The sections were counterstained with methyl green and mounted in glycerin gelatin. The numbers of Ly-6G-positive cells were counted in 10 non-overlapping fields (at 400x magnification) (18). Bone marrow transplantation To examine the relative roles of TLR2 plus TLR4 expression in hematopoietic (H) and structural (S) cells in the response to K. pneumoniae, we created bone marrow chimeric mice in essence as described previously (19, 20). Briefly, bone marrow cells were harvested from 7-9 week old WT (CD45.1+ or CD45.2+) and TLR2x4 KO (CD45.2+) mice (all age- and sexmatched). Cells were isolated by flushing tibia and femurs with PBS containing 10% fetal calf serum (BioWitthaker, Heidelberg, Germany), 100 U/ml penicillin (BioWitthaker), and 100 µg/ml streptomycin (BioWitthaker), and single cells were prepared by pulling the tissue clumps three times through a 25-gauge needle. Next, the cells were centrifuged at 250 x g for 10 minutes, aspirated, washed, and resuspended in PBS. At the start of the experiment recipient mice were six weeks old. The recipient groups received a lethal total body irradiation of two times 4.5 Gy with three hours between the two doses, using a 137Cs 23

Chapter 2

irradiator (CIS Bio International, Gif, France) at a dose rate of 0.5 Gy/min, followed by intravenous injection of 5x106 bone marrow cells from donor animals. To protect the irradiated recipient mice from immediate infections, the mice were also injected with 2x105 splenocytes from donor animals that were crushed through 40 µm filter, washed and resuspended in PBS. Moreover, mice were provided with autoclaved, acidified drinking water containing 0.16% neomycin sulfate (Sigma Chemical Co. St.Louis, MO) from one week before until five weeks after transplantation, and they were housed in sterile filter top cages in a laminar flow chamber. Mice entered the infection experiment six weeks after bone marrow transplantation, one week after stopping the antibiotics. Engraftment was confirmed by flow cytometry in the peripheral blood just before starting the infection experiment. As a control for the transplantation procedure, we not only administered TLR2x4 KO bone marrow cells (H-) into WT recipient mice (S+) and WT bone marrow cells (H+) into TLR2x4 KO recipient mice (S-), but also WT bone marrow (H+) to WT mice (S+) and TLR2x4 KO bone marrow (H-) to TLR2x4 KO mice (S-). Thus, four groups of mice were generated (H+/S-, H-/S+ and as controls H+/S+ and H-/S-). Flow cytometry Blood was drawn by heart puncture and erytrocytes were lysed with ice-cold isotonic NH4Cl solution (155 mM NH4Cl, 10 mM KHCO3, 0.1 mM EDTA, pH 7.4); the remaining cells were washed twice with RPMI 1640 (Bio Whittaker, Verviers, Belgium), and counted by using a hemocytometer. The percentages of monocytes and neutrophils were determined using a FACSCalibur (BD, San Jose, CA). Cells were brought to a concentration of 1x107 cells/ml in FACS buffer (PBS supplemented with 0.5% PBS, 0.01% NaN3 and 0.35 mM EDTA). Immunostaining for cell surface molecules was performed for 30 minutes at 4°C using directly labeled antibodies (abs) against GR-1 (GR-1 FITC, BD Pharmingen, San Diego, California), CD45.1phycoerythrin (CD45.1- PE; BD Pharmingen), CD45.2 Peridinin-chlorophyll-protein complex (CD45.2-PerCP, BD Pharmingen) and an allophycocyanine labeled antibody against F4/80 (Serotec, Oxford, United Kingdom). All abs were used in concentrations recommended by the manufacturer. Neutrophils were counted using the scatter pattern and GR-1 high gate, monocytes in the sidescatter low and F4/80 positive gate. Statistical analysis Data are expressed as mean ± SEM. Survival curves were compared using the long rank test. Comparisons between multiple groups were performed using the Kruskall-Wallis test with the Mann Whitney U test as a post test, using GraphPad Prism version 4.00, GraphPad Software (San Diego, CA). p < 0.05 was considered to be statistically significant.

Host defense during Klebsiella pneumonia relies on hematopoietic expressed TLR4 and TLR2

Results Delayed early inflammatory response in the absence of TLR4, but enhanced response in TLR2 KO mice The success of combating pulmonary infections strongly depends on the efficacy of the local inflammatory response elicited. Early recognition by TLRs with subsequent release of chemokines (to attract immune cells) and cytokines (activation) has been proven to be crucial for successful host defense (21). We therefore studied bacterial growth, influx of neutrophils and levels of cytokines and chemokines after 5h of infection with 3x103 CFU Klebsiella. At this time-point pulmonary bacterial loads were equal in the 4 mouse strains studied (figure 1A), whereas cultures from distant organs remained sterile. In order to study the influx of neutrophils at this early time point, we performed immunohistochemical staining of lung tissue slides and counted the number of Ly6+ neutrophils in 10 high-power fields (400x magnification). Despite no differences in pulmonary bacterial loads between WT, TLR2 KO, TLR4 KO and TLR2x4 KO mice, we found more neutrophils in lungs of TLR2 KO animals when compared to the other mouse strains (figure 1 B-D). In order to further dissect the early inflammatory response, we measured several important cytokines and chemokines in lung homogenates (figure 1E-H). When compared to WT mice, TLR2x4 KO mice demonstrated reduced lung levels of the chemokines KC and MIP-2 as well as the cytokines IL-1β and IL-6. In TLR4 KO mice KC and IL-1β were reduced 5 hours after infection. Interestingly, lungs of TLR2 KO animals contained significantly more IL-1β; moreover, KC and IL-6 demonstrated a trend towards higher levels in this mouse strain. No differences in TNF-α levels were detected in lung homogenates at this early time point of infection (data not shown). TLR2x4 double KO mice display a more profoundly disturbed antibacterial defense than TLR4 KO mice As a next step, we studied antibacterial host defense after 24h of infection, using two different bacterial doses: 3x103 and 1x104 CFU. After inoculation with the lower bacterial dose, TLR4 and TLR2x4 double KO mice but not TLR2 KO mice had higher bacterial burdens in lung (figure 2A) and spleen (figure 2B) at this time-point of infection. In the liver, only the mice that lacked both TLR2 and TLR4 had higher bacterial loads (figure 2C). After infection with the higher inoculum, no differences were found in lung, spleen and liver of TLR2 and TLR4 KO mice at 24h (figure 2DE). Interestingly, TLR2x4 double KO mice had higher bacterial burdens in all organs examined (figure 2D-E). These data suggest that antibacterial defense after low dose infection primarily is driven by TLR4, with a modest additional role for TLR2; however, after infection with a higher bacterial inoculum, apparently TLR2 can apparently, at least partially, compensate for TLR4, considering that only TLR2x4 double KO mice, but not TLR4 KO mice, displayed enhanced bacterial outgrowth. 25

Chapter 2

Figure 1: Impact of TLR4 and TLR2 on the early inflammatory response during Klebsiella pneumonia. WT, TLR2, TLR4 and TLR2x4 KO mice were inoculated with K. pneumoniae (3x103 CFU) and lung bacterial counts were determined after 5 h of infection (A). Each symbol represents an individual mouse. Horizontal lines indicate medians. The number of Ly6+ neutrophils (B, t=5h) was significantly higher in TLR2 KO mice (representative picture shown in panel D, original magnification 40x) compared to WT animals (representative picture shown in panel C) after Klebsiella pneumonia as counted in 10 randomly selected high-power fields (B). Chemokines KC (E) and MIP-2 (F) and cytokines IL-1Î˛ (G) and IL-6 (H) were measured in lung homogenates. Data are shown as means Âą SEM of 8 mice per group. *p < 0.05; **p < 0.01, ***p < 0.001 vs WT mice.

Host defense during Klebsiella pneumonia relies on hematopoietic expressed TLR4 and TLR2

We did not find consistent differences between mouse strains in pulmonary cytokine and chemokine levels (Table 1). After infection with 3x103 CFU TLR2x4 double KO mice displayed the lowest cytokine and chemokine levels, but the differences compared to WT mice were only significant for IL-1β and TNF-α. Overall, TLR2 KO mice had similar mediator levels in their lungs as compared to WT mice, with the exception of IL-1β and MIP-2 which were higher in TLR2 KO mice after infection with the lower and higher bacterial inoculums, respectively. Histopathology scores were consistent with cytokine and chemokine levels: only in the low dose infection did TLR2x4 double KO mice display significantly reduced histopathology scores. TLR2 is important during late stage of infection with K. pneumonia Next we wished to study the late host response during Klebsiella pneumonia. For this, we infected mice with 3x103 CFU K. pneumoniae, seeking to examine bacterial numbers and inflammatory responses 48h after infection. However, in the first experiment, in which we compared TLR2x4 double KO with WT mice, the former mouse strain proved to be hypersusceptible: all TLR2x4 double KO mice died 24 - 48h after infection, whereas the first deaths amongst WT mice occurred beyond the 48h time point (figure 3A). In the next experiment we infected WT, TLR2 KO and TLR4 KO mice with the same bacterial dose, again seeking to determine bacterial loads 48h after infection. Like TLR2x4 double KO mice, TLR4 KO mice demonstrated lethality beyond the 24h time-point. Whereas at 48h after infection all WT and TLR2 KO mice were alive, 5/8 TLR4 KO mice had died; the remaining 3 TLR4 KO mice showed high bacterial loads in all organs examined, especially in spleen and liver (figure 3B-D). Most interestingly, bacterial numbers in lungs of TLR2 KO mice were higher than bacterial counts in lungs of WT mice. These data again suggest that, although TLR4 is pivotal for adequate host defense against K. pneumoniae, TLR2 plays a role in local antibacterial host defense in a late stage of the infection. TLR2 and TLR4 on hematopoietic cells are pivotal for host defense against K. pneumoniae TLRs are expressed on cells of hematopoietic and non-hemapoietic origin, and both are potentially important in host defense against Klebsiella. To address whether TLR2 and TLR4 expression in either (radiosensitive) hematopoietic or (radioresistant) non-hematopoietic cells is sufficient for the innate immune response to Klebsiella, we generated bone marrow chimeric mice using WT and TLR2x4 double KO (the most susceptible strain) mice. In brief, either WT (CD45.1+) or TLR2x4 double KO mice (CD45.2+) were lethally irradiated and reconstituted with bone marrow from TLR2x4 double KO (CD45.2+) or WT mice (CD45.1+) respectively, creating WT mice reconstituted with TLR2x4 KO bone marrow (H-/S+) and TLR2x4 KO mice reconstituted with WT bone marrow (H+/S-). Control groups were also generated by transferring bone marrow from WT to WT mice 27

Chapter 2

Figure 2: TLR4 KO mice demonstrate enhanced bacterial outgrowth after low dose infection, while after high dose infection bacterial loads are only increased in TLR2x4 double KO mice. WT, TLR2, TLR4 and TLR2x4 KO mice were inoculated with two doses K. pneumoniae: 3x103 CFU (A-C) or 1x104 CFU (D-F). After 24 h of infection, bacterial burdens were determined in lung (A, D), spleen (B, E) and liver (C, F) homogenates. Each symbol represents an individual mouse. Horizontal lines indicate medians. *p < 0.05; **p < 0.01 vs WT mice; # p < 0.05 vs TLR4 KO mice.

Host defense during Klebsiella pneumonia relies on hematopoietic expressed TLR4 and TLR2 Table 1: Lung cytokine and chemokine levels and histopathology scores 24h after infection

CFU

3x10

TLR2KO

TLR4KO

TLR2x4KO

4150 ± 477

3688 ± 346

3261 ± 698

2035 ± 471

MIP-2

4290 ± 1308

6132 ± 848

3137 ± 862

1717 ± 526

IL-1β

2401 ± 309

3842 ± 620*

1549 ± 436

781 ± 580***

IL-6

1602 ± 325

2934 ± 580

1964 ± 445

1452 ± 319

TNF-α

188 ± 18

256 ± 29

146 ± 13*

129 ± 2**

8.5 ± 0.6

6.2 ± 0.9

Score

1x104

7.3 ± 0.6

4.6 ± 0.7**

6360 ± 667

5490 ± 432

3982 ± 1023

4459 ± 1056

MIP-2

12782 ± 3566

32579 ± 1671**

6466 ± 2831

17088 ± 5739

IL-1β

1936 ± 391

2504 ± 302

799 ± 301

1763 ± 580

IL-6

1528 ± 480

3355 ± 776

1637 ± 667

3693 ± 1473

TNF-α

965 ± 103

1551 ± 341

650 ± 130

719 ± 147

Score

10.1 ± 1.2

10.5 ± 0.4

7.9 ± 1.0

9.4 ± 1.1

WT, TLR2, TLR4 and TLR2x4 KO mice were infected with the indicated inoculum of K. pneumoniae. After 24h, mice were sacrificed, right lungs were removed and KC, MIP-2, IL-1β, IL-6 and TNF-α were determined using ELISA (pg/mL). Left lungs were used for determining histopathology scores (score) as described in the Methods section. Data are means ± SEM, n = 7 or 8 mice per group. *p < 0.05, **p < 0.01, ***p < 0.001 vs WT levels

(H+/S+) and from TLR2x4 KO to TLR2x4 KO mice (H-/S-). Engraftment was confirmed by flow cytometry on peripheral blood directly prior to induction of pneumonia (6 weeks after transplantation) and 24h after infection, revealing that, in accordance with our earlier data (19, 20), the mean percentage of neutrophils and monocytes in blood of uninfected chimeras derived from the donor mouse was >90% in all groups transplanted; 24h after infection, the percentages of donor neutrophils and monocytes in blood remained > 90% in all groups (data not shown). Six weeks after transplantation, we infected all groups with 104 CFU K. pneumoniae and studied bacterial growth in lungs and spleen 24h later. The procedure of irradiation and bone marrow transfer did not affect host defense, because the difference observed between irradiated WT mice reconstituted with WT bone marrow (H+/S+) and irradiated TLR2x4 KO mice reconstituted with TLR2x4 KO bone marrow (H-/S-) confirmed our earlier findings in non-irradiated animals: H-/Smice displayed higher bacterial loads than H+/S+ mice in both lungs and spleen (figure 4). Our main finding was that TLR2x4 KO mice reconstituted with WT bone

Chapter 2

Figure 3: TLR2 is important during the late phase of infection. WT (n = 8; closed symbols) and TLR2x4 KO mice (n=8; open symbols) were inoculated with 3x103 CFU K. pneumoniae and followed for 14 days (A), p< 0.001. WT, TLR2 and TLR4 mice (n=8 mice per group) were inoculated with 3x103 K. pneumoniae. After 48 h of infection, bacterial burdens were determined in lung (B), spleen (C) and liver (D) homogenates of the remaining animals. Each symbol represents an individual mouse. Horizontal lines indicate medians. *p < 0.05 vs WT mice.

marrow (H+/S-) displayed equal amounts of bacteria in lungs and spleen when compared to H+/S+ animals. Moreover, irradiated WT reconstituted with TLR2x4 KO bone marrow were more susceptible to infection: H-/S+ mice displayed increased bacterial outgrowth in lung and spleen when compared to H+/S+ mice. Finally, bacterial loads in H-/S+ and H-/S- mice were similar. Together, these findings demonstrate the importance of TLR2 and TLR4 expression on radiosensitive hematopoietic cells for an adequate antibacterial defense during Klebsiella pneumonia. Consistent with earlier experiments (table 1), pulmonary levels of KC and IL-1Î˛ were reduced in H-/S- mice in comparison to H+/S+ mice despite higher bacterial loads (table 2). Mice with TLR2- and TLR4-deficient stroma (S-) demonstrated reduced KC levels regardless of the origin of reconstituted hematopoietic cells (H+ or H-), implying KC is mainly produced by stromal cells, such as epithelia.

Host defense during Klebsiella pneumonia relies on hematopoietic expressed TLR4 and TLR2

Figure 4: TLR2/4 expressed by hematopoietic cells are important for antibacterial defense during K. pneumoniae pneumonia. WT (S+) and TLR2x4 KO (S-) mice were irradiated and injected with WT (H+) or TLR2x4 KO (H-) bone marrow cells. Six weeks after transplantation, mice were infected with 104 CFU K. pneumoniae and sacrificed 24h later. Pulmonary (A) and splenic (B) outgrowth of K. pneumonia was determined in organ homogenates. Each symbol represents an individual mouse. Horizontal lines indicate medians. *p<0.05; **p<0.01 vs H+/S+. ns = not significant.

Table 2: Lung cytokine and chemokine levels in chimeric mice 24h after infection.

TLR2/4

H+/S+

H+/S-

H-/S+

H-/S-

11224 ± 778

6115 ± 1605*

11198 ± 682

6954 ± 1321*

MIP-2

22822 ± 11200

32113 ± 12403

34712 ± 11590

4245 ± 1035

IL-1β

1949 ± 472

3411 ± 1122

2674 ± 490

855 ± 154*

IL-6

859 ± 156

1271 ± 236

1328 ± 176

1106 ± 129

TNF-α

1368 ± 302

1944 ± 648

1755 ± 395

1978 ± 425

Chimeric mice were infected with 4x104 CFU K. pneumoniae 6 weeks after bone marrow transplantation. Twenty-four h later, mice were sacrificed, lungs were removed and KC, MIP-2, IL-1β and IL-6 were determined using ELISA (pg/mL). Data are means ± SEM, n = 7 or 8 mice per group. *p<0.05, **p<0.01, ***p<0.001 vs WT levels. H: hematopoietic; S: structural

Chapter 2

Discussion K. pneumoniae is a clinically important Gram-negative pathogen in hospitalacquired pneumonia. Previous research demonstrated that TLR4 is important for an adequate host defense in K. pneumoniae pneumonia (4, 5, 7). Given that Klebsiella expresses several TLR2 ligands (3, 11), we here sought to determine the role of TLR2, in the presence or absence of functional TLR4, in the innate immune response to respiratory tract infection by this bacterium in vivo. In addition, we aimed to assess the relative contribution of TLR2 and TLR4 on hematopoietic and stromal cells herein. Our main findings were: 1) TLR4 drives the antibacterial host response after infection with relatively low Klebsiella doses; 2) TLR2 becomes involved at a later phase of the infection and/or upon exposure of the host to higher bacterial burdens; and 3) TLR2 and TLR4 expressed by radiosensitive hematopoietic cells and not by radioresistant stromal cells are important for an adequate host response. Klebsiella pneumonia and pneumosepsis are common in and outside the hospital environment (1, 2). Nosocomial pneumonia mainly affects patients with pre-existing diseases that may impact on host defense in the lung. Our model uses previously healthy mice and, as such more closely resembles community-acquired Klebsiella pneumonia. Other laboratories have used the same method to obtain insight into the innate immune response during respiratory tract infection by this pathogen (22). Of note, the ATCC strain of K. pneumoniae serotype 2 used in this study is a common laboratory strain that was used in many previous studies investigating the host response against K. pneumoniae. Capsular serotypes 1 and 2 are the most common and the most virulent Klebsiella serotypes (2, 23, 24). Upon infection of TLR2 KO mice with K. pneumoniae, we discovered a dual role for TLR2. In the early initial recognition phase, TLR2-related pathways delayed IL-1Î˛ release and neutrophil influx. We previously observed a similar dampening function of TLR2 in the early host response during another Gram-negative pneumonia, caused by Acinetobacter baumannii (25). Along the same line, TLR2 KO mice were reported to be less susceptible to lethal infections with Yersinia enterocolitica or Candida albicans through a mechanism that involved a stronger type 1 cytokine response (26, 27). Although the exact mechanisms behind these possible anti-inflammatory properties of TLR2 remain unclear, it is possible that lack of TLR2 signaling was associated with upregulation of other receptors with mainly pro-inflammatory properties, such as has been described for TLR4 in TLR2 KO mice infected with Pseudomonas aeruginosa (28). Interestingly, the role of TLR2 changed during the course of infection. While TLR2 deficiency did not impact on bacterial growth early after infection with a low bacterial dose, at a later stage or after infection with a higher inoculum, TLR2 did contribute to antibacterial defense. Indeed, relative to WT mice, TLR2 KO mice had higher bacterial counts in lungs 48 h (but not 24 h) after infection with 3x103 CFU Klebsiella, whereas TLR2x4 double KO mice displayed higher bacterial burdens than TLR4 KO mice in lungs and distant organs 24 h after infection with 104 CFU Klebsiella, which was associated with an accelerated lethality beyond this time-point. TLR4 clearly played a more prominent role in antibacterial defense than TLR2, especially in the initial phase of the infection, considering that TLR4 KO (but not TLR2 KO) mice 32

Host defense during Klebsiella pneumonia relies on hematopoietic expressed TLR4 and TLR2

demonstrated enhanced bacterial growth in lungs and spleen 24 h after infection with 3x103 CFU Klebsiella when compared with WT mice, and TLR4 KO mice and TLR2x4 double KO mice harbored equal bacterial loads in lungs and spleen at this time point. Together, these data suggest that the interaction between LPS and TLR4 drives the early host response during Klebsiella pneumonia, whereas the interaction between TLR2 ligands expressed by this bacterium and TLR2 becomes a factor upon exposure to higher bacterial numbers. In clinical practice, patients would have been treated with antibiotics before lethality occurred. It is very well possible that the outcome in the TLR deficient animals would be different in the context of antibiotic therapy. Hypothetically, it might be beneficial to lack TLR mediated hyperinflammation when antibiotics are taking care of bacterial elimination. It is therefore important to realize that the primary aim of our study was to determine the role of TLRs in the innate immune response during Klebsiella pneumonia, rather than to investigate the therapeutic potential of TLR inhibition (which clearly should be studied in animals concurrently treated with antibiotics). Neutrophil recruitment to the lungs is an important first line of defense against bacterial infections (21). However, in addition to neutrophils and lung macrophages, non-hematopoietic cells, such as lung epithelium and endothelium, contribute to the initial recognition of bacteria and production of inflammatory mediators and thus, host defense. To dissect the role of hematopoietic cells and non-hematopoietic cells, we generated bone marrow chimeric mice using WT and TLR2x4 double KO mice (the most susceptible mouse strain). We found that injecting WT bone marrow into irradiated TLR2x4 KO animals resulted in similar bacterial growth as in syngeneic transplanted WT mice and reduced outgrowth when compared to syngeneic TLR2x4 KO animals, meaning that hematopoietic cells are of utmost importance in host defense. In line, irradiated WT mice that received TLR2x4 KO bone marrow did worse than syngeneic transplanted WT animals. Since differences in antibacterial host defense between syngeneic transplanted (H+/S+ and H-/S-) mice on the one hand and non-transplanted WT and TLR2x4 KO animals on the other hand were similar, we believe that our results are not an artifact introduced by the bone marrow transplantation procedure. Several earlier studies investigated the role of TLRs in host defense against Klebsiella pneumonia. In accordance with the current data, C3H/HeJ mice (which harbor a mutation in TLR4 that renders this receptor dysfunctional) were reported to have an enhanced bacterial growth and dissemination and a reduced survival (4, 5). TLR4 can signal via two intracellular routes, relying on the adaptors TRIF (TIR domain-containing adaptor-inducing IFN-Î˛) and MyD88 respectively (3), and both TRIF KO and MyD88 KO displayed an impaired host defense during Klebsiella pneumonia (9). Notably, MyD88 mediates signaling of all TLRs excluding TLR3 and of the signaling receptors for IL-1 and IL-18 (29); the possibility that multiple TLRs are involved in protective immunity during respiratory tract infection by K. pneumoniae is further supported by investigations revealing an enhanced susceptibility of mice deficient for either TLR9 (8) or TIRAP (which mediates signaling of TLR1, TLR2, TLR4, and TLR6) (10). We here expand these previous data showing that during 33

Chapter 2

pneumonia caused by K. pneumoniae TLR4 and TLR2 expressed by hematopoietic cells interact in mediating an effective antibacterial defense in a manner that is dependent on the stage of the infection and the bacterial load to which the host is exposed. The present study further suggests that TLR2 plays a dual role in the host response to Klebsiella pneumonia: while TLR2 signaling dampens the initial inflammatory response after relatively low dose infection without influencing bacterial expansion, at later stages TLR2 is important in limiting bacterial growth irrespective of the presence of TLR4.

Acknowledgements The authors thank Anita de Boer for expert technical assistance.

Host defense during Klebsiella pneumonia relies on hematopoietic expressed TLR4 and TLR2

References 1. Burwen DR, Banerjee SN, Gaynes RP. Ceftazidime Resistance Among Selected osocomial Gram-Negative Bacilli in the United States. National Nosocomial Infections Surveillance System. J Infect Dis 1994;170:1622-1625. 2. Podschun R, Ullmann U. Klebsiella Spp. As Nosocomial Pathogens: Epidemiology, Taxonomy, Typing Methods, and Pathogenicity Factors Clin Microbiol Rev 1998;11:589-603. 3. Ishii KJ, Koyama S, Nakagawa A, Coban C, Akira S. Host Innate Immune Receptors and Beyond: Making Sense of Microbial Infections. Cell Host Microbe 2008;3:352-363. 4. Branger J, Knapp S, Weijer S, Leemans JC, Pater JM, Speelman P, Florquin S, van der Poll T. Role of Toll-Like Receptor 4 in Gram-Positive and Gram-Negative Pneumonia in Mice. Infect Immun 2004;72:788-794. 5. Schurr JR, Young E, Byrne P, Steele C, Shellito JE, Kolls JK. Central Role of Toll- Like Receptor 4 Signaling and Host Defense in Experimental Pneumonia Caused by Gram-Negative Bacteria. Infect Immun 2005;73:532-545.

6. Song J, Bishop BL, Li G, Duncan MJ, Abraham SN. TLR4-Initiated and CAMP-Mediated Abrogation of Bacterial Invasion of the Bladder. Cell Host Microbe 2007;1:287-298.

7. Chan YR, Liu JS, Pociask DA, Zheng M, Mietzner TA, Berger T, Mak TW, Clifton MC, Strong RK, Ray P, et al. Lipocalin 2 Is Required for Pulmonary Host Defense Against Klebsiella Infection. J Immunol 2009;182:4947-4956.

8. Bhan U, Lukacs NW, Osterholzer JJ, Newstead MW, Zeng X, Moore TA, McMillan TR, Krieg AM, Akira S, Standiford TJ. TLR9 Is Required for Protective Innate Immunity in Gram Negative Bacterial Pneumonia: Role of Dendritic Cells. J Immunol 2007;179:3937-3946.

9. Cai S, Batra S, Shen L, Wakamatsu N, Jeyaseelan S. Both TRIF- and MyD88- Dependent Signaling Contribute to Host Defense Against Pulmonary Klebsiella Infection. J Immunol 2009;183:6629-6638. 10. Jeyaseelan S, Young SK, Yamamoto M, Arndt PG, Akira S, Kolls JK, Worthen GS. Toll/IL-1R Domain-Containing Adaptor Protein (TIRAP) Is a Critical Mediator of Antibacterial Defense in the Lung Against Klebsiella Pneumoniae but Not Pseudomonas Aeruginosa. J Immunol 2006;177:538-547. 11. Jeannin P, Bottazzi B, Sironi M, Doni A, Rusnati M, Presta M, Maina V, Magistrelli G, Haeuw JF, Hoeffel G, et al. Complexity and Complementarity of Outer Membrane Protein A Recognition by Cellular and Humoral Innate Immunity Receptors. Immunity 2005;22:551 560. 12. Regueiro V, Moranta D, Campos MA, Margareto J, Garmendia J, Bengoechea JA. Klebsiella Pneumoniae Increases the Levels of Toll-Like Receptors 2 and 4 in Human Airway Epithelial Cells. Infect Immun 2009;77:714-724. 13. Hoshino K, Takeuchi O, Kawai T, Sanjo H, Ogawa T, Takeda Y, Takeda K, Akira S. Cutting Edge: Toll-Like Receptor 4 (TLR4)-Deficient Mice Are Hyporesponsive to Lipopolysaccharide: Evidence for TLR4 As the Lps Gene Product. J Immunol 1999;162:3749-3752.

Chapter 2

14. Takeuchi O, Hoshino K, Kawai T, Sanjo H, Takada H, Ogawa T, Takeda K, Akira S. Differential Roles of TLR2 and TLR4 in Recognition of Gram-Negative and Gram-Positive Bacterial Cell Wall Components. Immunity 1999;11:443-451.

15. Renckens R, Roelofs JJTH, Bonta PI, Florquin S, de Vries CJM, Levi M, Carmeliet P, vanâ&#x20AC;&#x2122;t Veer C, van der Poll T. Plasminogen Activator Inhibitor Type 1 Is Protective During Severe Gram-Negative Pneumonia. Blood 2007;109:1593-1601.

16. Wieland CW, Stegenga ME, Florquin S, Fantuzzi G, van der Poll T. Leptin and Host Defense Against Gram-Positive and Gram-Negative Pneumonia in Mice. [Miscellaneous Article]. Shock 2006;25:414-419.

17. Knapp S, Leemans JC, Florquin S, Branger J, Maris NA, Pater J, van RN, van der Poll T. Alveolar Macrophages Have a Protective Antiinflammatory Role During Murine Pneumococcal Pneumonia. Am J Respir Crit Care Med 2003;167:171-179. 18. Leemans JC, Butter LM, Pulskens WP, Teske GJ, Claessen N, van der Poll T, Florquin S. The Role of Toll-Like Receptor 2 in Inflammation and Fibrosis During Progressive Renal Injury. PLoS One 2009;4:e5704.

19. Leemans JC, Stokman G, Claessen N, Rouschop KM, Teske GJ, Kirschning CJ, Akira S, van der Poll T, Weening JJ, Florquin S. Renal-Associated TLR2 Mediates Ischemia/ Reperfusion Injury in the Kidney. J Clin Invest 2005;115:2894-2903.

20. Pulskens WP, Teske GJ, Butter LM, Roelofs JJ, van der Poll T, Florquin S, Leemans JC. Toll-Like Receptor-4 Coordinates the Innate Immune Response of the Kidney to Renal Ischemia/Reperfusion Injury. PLoS One 2008;3:e3596.

21. Mizgerd JP. Acute Lower Respiratory Tract Infection. N Engl J Med 2008;358:716-727.

22. Karaolis DKR, Newstead MW, Zeng X, Hyodo M, Hayakawa Y, Bhan U, Liang H, Standiford TJ. Cyclic Di-GMP Stimulates Protective Innate Immunity in Bacterial Pneumonia. Infect Immun 2007;75:4942-4950.

23. Fung CP, Hu BS, Chang FY, Lee SC, Kuo BI, Ho M, Siu LK, Liu CY. A 5-Year Study of the Seroepidemiology of Klebsiella Pneumoniae: High Prevalence of Capsular Serotype K1 in Taiwan and Implication for Vaccine Efficacy. J Infect Dis 2000;181:2075-2079.

24. Yu WL, Ko WC, Cheng KC, Lee CC, Lai CC, Chuang YC. Comparison of Prevalence of Virulence Factors for Klebsiella Pneumoniae Liver Abscesses Between Isolates With Capsular K1/K2 and Non-K1/K2 Serotypes. Diagnostic Microbiology and Infectious Disease 2008;62:1-6.

25. Knapp S, Wieland CW, Florquin S, Pantophlet R, Dijkshoorn L, Tshimbalanga N, Akira S, van der Poll T. Differential Roles of CD14 and Toll-Like Receptors 4 and 2 in Murine Acinetobacter Pneumonia. Am J Respir Crit Care Med 2006;173:122-129.

26. Sing A, Rost D, Tvardovskaia N, Roggenkamp A, Wiedemann A, Kirschning CJ, Aepfelbacher M, Heesemann J. Yersinia V-Antigen Exploits Toll-Like Receptor 2 and CD14 for Interleukin 10-Mediated Immunosuppression. J Exp Med 2002;196:1017-1024.

27. Netea MG, Sutmuller R, Hermann C, Van der Graaf CA, Van der Meer JW, van Krieken JH, Hartung T, Adema G, Kullberg BJ. Toll-Like Receptor 2 Suppresses Immunity Against Candida Albicans Through Induction of IL-10 and Regulatory T Cells. J Immunol 2004;172:3712-3718.

Host defense during Klebsiella pneumonia relies on hematopoietic expressed TLR4 and TLR2

28. Lorenz E, Chemotti DC, Vandal K, Tessier PA. Toll-Like Receptor 2 Represses Nonpilus Adhesin-Induced Signaling in Acute Infections With the Pseudomonas Aeruginosa PilA Mutant. Infect Immun 2004;72:4561-4569.

29. Dinarello CA. Immunological and Inflammatory Functions of the Interleukin-1 Family. Annu Rev Immunol 2009;27:519-550.

Chapter 3 Differential roles of MyD88 and TRIF in hematopoietic and resident cells during murine gram-negative pneumonia Journal of Infectious Diseases 2012 Nov;206(9):1415-23 DOI: 10.1093/infdis/jis505 Miriam H.P. van Lieshout 1,2 Dana C. Blok 1,2 Catharina W. Wieland 1,3 Alex F. de Vos 1,2 Cornelis van â&#x20AC;&#x2122;t Veer 1,2 Tom van der Poll 1,2,4 Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands: 1 Center of Infection and Immunity Amsterdam 2 Center of Experimental and Molecular Medicine 3 Laboratory of Experimental Intensive Care and Anesthesiology 4 Division of Infectious Diseases

Chapter 3

Differential role of MyD88 and TRIF signaling in pneumonia

Abstract Background: Pneumonia is frequently caused by gram-negative pathogens, among which Klebsiella pneumoniae prominently features. Recognition of pathogen-associated molecular patterns by Toll-like receptors (TLRs) is important for an appropriate immune response during infection. TLR signaling can proceed via two distinct routes which are dependent on the adaptor proteins Myeloid differentiation primary response gene (88) (MyD88) and TIR-domain-containing adaptor-inducing interferon-Î˛ (TRIF) respectively. Aim of the study was to determine the relative contribution of MyD88 and TRIF signaling in resident and hematopoietic cells to host defense during pneumonia. Methods: Bone marrow chimeras of MyD88 deficient/wild type and TRIF mutant/wild type mice were created and infected with K. pneumoniae via the airways. Main results: MyD88 in both resident and hematopoietic cells contributed to survival and antibacterial defense in late stage infection, whereas only TRIF in hematopoietic cells was protective. On the other hand, resident MyD88 and hematopoietic TRIF contributed to distant cellular injury. Resident MyD88 was pivotal for early chemokine release and neutrophil recruitment in the bronchoalveolar space. Conclusion: MyD88 and TRIF dependent signaling have a differential contribution to host defense in different cell types that changes from early to late stage gram-negative pneumonia.

Differential role of MyD88 and TRIF signaling in pneumonia

Introduction Pneumonia is the most common cause of sepsis. Lower respiratory tract infections are frequently caused by gram-negative pathogens, including Klebsiella pneumoniae (1,2). Emerging microbial resistance amongst Enterobacteriaceae is an issue of major concern, limiting therapeutic options and increasing mortality rates (3). Tolllike receptors (TLRs) occupy a prominent position in the innate immune system by virtue of their capacity to recognize bacterial components (4, 5). TLR signaling can proceed via two distinct routes which are dependent on myeloid differentiation primary response gene 88 (MyD88) and TIR-domain-containing adaptor-inducing interferon-Î˛ (TRIF) (4, 5). MyD88 is the universal adaptor for all TLRs except TLR3; TRIF is the sole adaptor for TLR3, and in addition contributes to TLR4 signaling. Very recently, mice deficient for either MyD88 or TRIF were found to be more susceptible to death after infection with Klebsiella via the airways, which was accompanied by enhanced bacterial growth in both mouse strains (6). Several MyD88-dependent TLRs likely contribute to the susceptible phenotype of MyD88 deficient mice during Klebsiella pneumonia, in particular TLR4 and TLR9, whereas TLR2 may contribute to host defense during late stage infection (7-10). Other MyD88 dependent TLRâ&#x20AC;&#x2122;s have not been studied in models of Klebsiella infection, but are less likely to be involved considering their specificity for pathogen ligands that are not expressed by this bacterium. TLRs are expressed by both hematopoietic and resident cells and both cell types contribute to an effective host defense during respiratory tract infections (1113). We here aimed to investigate the cell-type specific role of MyD88 and TRIF during early and late stage Klebsiella infection, by creating bone marrow chimeras expressing MyD88 or TRIF only in radioresistant (resident, R) cells or radiosensitive (hematopoietic, H) cells.

Methods Animals MyD88 gene deficient (Myd88-/-) mice were provided by Dr. S. Akira (Research Institute for Microbial Diseases, Osaka, Japan) and backcrossed > 6 times to a C57Bl/6 genetic background (14). TRIF mutant mice, generated on a C57Bl/6 genetic background (15), were provided by Dr B. Beutler (the Scripps Research Institute, La Jolla, CA). Age- and sex matched wild-type (WT) C57Bl/6 control mice were obtained from Harlan Nederland (Horst, the Netherlands). Mice were infected at 10-12 weeks of age. The Animal Care and Use Committee of the University of Amsterdam approved all experiments. Induction of pneumonia and sampling of organs Pneumonia was induced by intranasal inoculation with 7x103 colony forming units (CFU) of K. pneumoniae serotype 2 (ATCC 43816; American Type Culture Collection, Manassas, VA) (9, 16). Mice were followed for a maximum of 14 days or in separate experiments euthanized at 6, 24 or 48 hours after infection mice after which organs were harvested and processed for the determination of 41

Chapter 3

bacterial outgrowth and cytokine levels as described (9). In some experiments bronchoalveolar lavage (BAL) was performed and cell counts determined in BAL fluid (BALF) (17). Assays TNF-α, IL-6, IL-10 and MCP-1 (monocyte chemotactic protein 1, also known as CCL2) were measured by using a cytometric bead array multiplex assay (BD Biosciences, San Jose, CA). Cytokine-induced neutrophil chemoattractant (KC, also known as CXCL1), MIP-2 (Macrophage inflammatory protein 2 alpha, also known as CXCL2), LPS-induced CXC chemokine (LIX) and E-selectin were measured by ELISA’s (R&D Systems, Minneapolis, MN). Lactate dehydrogenase (LDH), aspartate aminotransferase (AST) and alanine transaminase (ALT) were measured with kits from Sigma (St. Louis, MO), using a Hittachi analyzer (Boehringer Mannheim, Mannheim, Germany). Bone marrow transplantation Bone marrow chimeric mice were generated as described (9, 18, 19). Briefly, recipient groups received a lethal total body irradiation of two times 4.5 Gy with three hours between the two doses, using a 137Cs irradiator (CIS Bio International, Gif, France) at a dose rate of 0.5 Gy/min, followed by intravenous injection of 5x106 bone marrow cells that were isolated from donor animals as decribed before (9). To protect the irradiated recipient mice from immediate infections, the mice were also injected with 2x105 splenocytes from donor animals that were crushed through 40 μm filter, washed and resuspended in PBS. Moreover, mice were provided with autoclaved, acidified drinking water containing 0.16% neomycin sulfate (Sigma Chemical Co. St.Louis, MO) from one week before until four weeks after transplantation. Pneumonia was induced 6 weeks after transplantation. Engraftment was checked by flow cytometry based on differential expression of CD45.1 and CD45.2 by donor and recipient cells exactly as described (9,19). As a control for the transplantation procedure, we not only administered TRIF mutant or Myd88-/- bone marrow cells (H-) into WT recipient mice (R+) and WT bone marrow cells (H+) into TRIF mutant or Myd88-/- recipient mice (R-), but also WT bone marrow (H+) to WT mice (R+) and TRIF mutant or Myd88-/- bone marrow (H-) to TRIF mutant or Myd88/mice (R-) respectively. Thus, in each experiment with chimeric mice four groups of mice were generated (R-/H+, R+/H- and as controls R+/H+ and R-/H-). Statistical analysis Data are expressed as means ± standard error of the mean; as medians with individual data points (for bacterial loads); or as Kaplan-Meier plots. For experiments with two groups Mann Whitney U test was used to determine significance. For experiments with more than two groups Kruskall-Wallis test was used as a pretest, in order to reduce the chance of committing a type 1 error. When appropriate, Mann Whitney U tests were used as follow-up tests to compare individual genetically modified groups to the R+H+ control group. Survival curves were compared using log-rank test. All analyses were done using GraphPad Prism (San Diego, CA). p < 0.05 was considered statistically significant. 42

Differential role of MyD88 and TRIF signaling in pneumonia

Results Both MyD88 and TRIF are crucial for survival and the antibacterial response in gram-negative pneumonia We first infected Myd88-/-, TRIF mutant and WT mice with K. pneumoniae via the airways and followed them for 10 days in two independent survival experiments (Figure 1A and D). All Myd88-/- mice died before 48 hours, while 50% of WT mice remained alive until the end of the experiment (p < 0.001). TRIF mutant mice also showed a higher and accelerated mortality compared to WT mice (p < 0.01), although lethality did not occur as rapidly as in Myd88-/- mice. Next, we euthanized Myd88-/-, TRIF mutant and WT mice at predefined time points for quantitative cultures of lungs, blood and spleen. In experiments comparing Myd88/and WT mice, these analyses were confined to the first 24 hours considering the large number of deaths amongst Myd88-/- mice thereafter. At 6 hours, the bacterial loads in the lungs of both Myd88-/- and TRIF mutant mice were similar to those in WT mice and cultures of blood and spleen remained sterile in all but one Myd88-/mouse. After 24 hours, Myd88-/- mice had about 2-log more bacteria in their lungs when compared to WT mice (Figure 1B, p < 0.01), while dissemination to blood (not shown) and spleen (Figure 1C) was not different. At 48 (but not 24 hours), TRIF mutant mice had remarkably higher bacterial burden in their lungs (Figure 1E), as well as in blood (not shown) and spleen (Figure 1F, p < 0.001 versus WT mice). The fact that TRIF mutant mice showed enhanced bacterial loads in their spleens while Myd88-/- mice did not, is likely explained by the different durations of the infection at the time of euthanasia (48 versus 24 hours respectively). These data confirmed the essential role of MyD88 and TRIF in host defense during Klebsiella pneumonia (6) and further show that the role of TRIF in protective immunity becomes apparent later in the course of the infection in comparison to MyD88. MyD88 expression in both hematopoietic and resident cells contributes to survival, while hematopoietic TRIF expression is most important for survival To dissect the role of MyD88 and TRIF dependent TLR signaling in hematopoietic (H) versus resident (R) cells, we created bone marrow chimeric mice for MyD88 and TRIF. In accordance with our earlier reports (9, 18, 19), the mean percentage of donor derived neutrophils and monocytes in blood from all chimeric mice was >90% (data not shown). MyD88 R-/H- mice displayed a strongly accelerated mortality when compared with MyD88 H+/R+ mice (p < 0.0001; Figure 2A). Of more interest, MyD88 R+/H- mice and MyD88 R-/H+ mice also demonstrated an accelerated mortality when compared with MyD88 R+/H+ mice (both p < 0.001), whereas the mortalities amongst both chimeric MyD88 strains (R+/H- and R-/H+) were significantly delayed when compared with MyD88 R-/H- mice (p < 0.001 and p <0.05 respectively). Mortality curves of MyD88 R+/H- and R-/H+ mice were not different. TRIF R-/H- mice displayed a strongly accelerated mortality when compared with 43

Chapter 3

Figure 1: MyD88 and TRIF protect against lethality and restrict bacterial growth in gram negative pneumonia. WT, Myd88-/- and TRIF mutant mice were inoculated with 7x103 CFU K. pneumoniae and monitored for survival or sacrificed at designated time-points. Survival of WT (closed squares) and Myd88-/- (open rounds) mice (n=8 per group) (A). Bacterial loads in lung (B) and spleen (C) 6 and 24 hours after infection in WT (closed squares) and Myd88-/- mice (open rounds). Survival of WT (closed squares) and TRIF mutant mice (open rounds) (n=17 per group) (D). Bacterial loads in lung (E) and spleen (F) 6, 24 or 48 hours after infection in WT (closed squares) or TRIF mutant mice (open rounds) (n=7-8 per group). Each symbol represents an individual mouse, with horizontal lines showing medians. * p < 0.05, ** p < 0.01, *** p < 0.001 compared to WT mice determined with Mann-Whitney U test.

TRIF R+/H+ mice (p < 0.0001, Figure 2B). TRIF R+/H- mice and TRIF R-/H+ mice both showed an accelerated mortality when compared with TRIF R+/H+ mice (both p < 0.01), but there was no significant difference between the mortality curves of TRIF R-/H+ mice and TRIF R+/H- mice. Importantly, the mortalities amongst TRIF R-/H+ mice were delayed when compared with TRIF R-/H- mice (p < 0.05) while the survival of TRIF R+/H- mice was not significantly different from TRIF R-/H- mice. MyD88 expression in both hematopoietic and resident cells limits bacterial growth, whereas hematopoietic but not resident TRIF expression reduces bacterial multiplication We next infected MyD88 and TRIF chimeric mice with Klebsiella via the airways and euthanized them at 24 (MyD88 chimeras) or 48 hours (TRIF chimeras), i.e. at time points that had revealed the importance of MyD88 and TRIF in reducing bacterial growth in mice with a general deficiency for these adaptor proteins. As expected, MyD88 R-/H- and TRIF R-/H- mice demonstrated enhanced bacterial outgrowth in their lungs at 24 and 48 hours after infection respectively when compared to their respective R+/H+ controls (Figure 3A and D). Of considerable 44

Differential role of MyD88 and TRIF signaling in pneumonia

Figure 2: MyD88 expression in both resident and hematopoietic cells contributes to survival from gram negative pneumonia, while hematopoietic TRIF expression is most important for survival. WT (R+) and Myd88-/- or TRIF mutant (R-) mice were irradiated and injected with WT (H+), Myd88-/- or TRIF mutant (H-) bone marrow cells. Six weeks after transplantation mice were infected with 7x103 CFU K. pneumoniae and survival was monitored for 14 days. Survival of MyD88 chimeras (n=7-19 per group) (A) and TRIF chimeras (n=18-20 per group) (B). p-values for the comparison between different recipient groups are summarized in tables: * p < 0.05, ** p < 0.01, *** p < 0.001, ns = non-significant determined with Log-Rank Test.

interest, 24 hours after infection not only MyD88 R-/H- mice showed enhanced bacterial growth in lungs and distant body sites (blood and spleen), but also mice deficient for MyD88 in either hematopoietic or resident cells had increased bacterial loads in all body compartments when compared with MyD88 H+/R+ mice (Figure 3AC). The difference between MyD88 R+/H- and R+/H+ mice was larger than between MyD88 R-/H+ and R+/H+ mice, indicating that MyD88 signaling in resident cells does contribute to host defense but that the expression of MyD88 in hematopoietic cells is even more important. In contrast, the experiments with TRIF chimeras only revealed a role for TRIF expressed in hematopoietic cells in reducing bacterial growth and dissemination. Indeed, at 48 hours after infection TRIF R+/H- mice, but not TRIF R-/H+ mice, displayed higher bacterial burdens in lungs, blood and spleen when compared with TRIF R+/H+ mice; in addition, bacterial loads in TRIF R+/H- mice were similar to those in TRIF R-/H- mice in all body compartments (Figure 3D-F).

Chapter 3

Limited role of MyD88 or TRIF in lung cytokine and chemokine production during late stage infection We measured the lung concentrations of the proinflammatory cytokines TNF-Îą, IL-6, the anti-inflammatory cytokine IL-10 and the chemokines KC, MIP-2 and MCP-1 at 24 hours (MyD88 chimeric mice) or 48 hours (TRIF chimeric mice) after infection. The levels of all mediators showed a large variation amongst groups and differences between groups were modest at best (Supplementary tables 1 and 2 ). These data suggest that neither MyD88 nor TRIF signaling is essential for the production of these mediators in the lungs during late stage pneumonia.

Figure 3: MyD88 expression in both resident and hematopoietic cells limits bacterial growth, whereas hematopoietic but not resident TRIF expression reduces bacterial multiplication. WT (R+) and Myd88 -/- or TRIF mutant (R-) mice were irradiated and injected with WT (H+), Myd88 -/- or TRIF mutant (H-) bone marrow cells. Six weeks after transplantation, mice were infected with 7x10 3 CFU K. pneumoniae. Bacterial loads in lung (A), blood (B) and spleen (C) of MyD88 chimeras 24 hours after infection (n = 8-14 per group). Bacterial loads in lung (D), blood (E) and spleen (F) of TRIF mutant chimeras 48 hours after infection (n = 9-11 per group). Each symbol represents an individual mouse, with horizontal lines showing medians.* p < 0.05, ** p < 0.01, *** p < 0.001 vs R+/ H+ mice determined with Mann-Whitney U test as a follow-up test on Kruskall-Wallis test.

Differential role of MyD88 and TRIF signaling in pneumonia

Resident MyD88 and hematopoietic TRIF expression contribute to distant organ injury in late stage infection The model of gram-negative pneumonia and sepsis used here is associated with rises in the plasma concentrations of LDH (indicative for cellular injury in general) and AST/ALT (reflecting hepatocellular injury) in the late stage of infection (16). Remarkably, none of the partially or fully MyD88 deficient mice demonstrated evidence for enhanced cell injury when compared with MyD88 R+/H+ mice (Figure 4): plasma LDH and AST concentrations were even slightly lower in MyD88 R-/H+ mice (p < 0.05 versus MyD88 R+/H+ mice), suggesting that MyD88 in resident cells contributes to cell injury during Klebsiella infection (Figure 4A-C). In TRIF chimeras there was a distinct contribution of TRIF in hematopoietic cells to cellular injury : TRIF R+/H- mice and TRIF R-/H- mice had significantly lower plasma LDH, AST and ALT values when compared to TRIF R+/H+ or TRIF R-/H+ mice (p < 0.05 to p <0.001, Figure 4D-F). On the other hand, the plasma levels of cellular injury markers were not different between TRIF R+/H+ and TRIF R-/H+ mice. Differential contribution of MyD88 and TRIF in hematopoietic and resident cells in early stage infection We repeated the experiments in MyD88 and TRIF chimeras, this time sacrificing the mice after 6 hours of infection (Figure 5). This early time point was not chosen to determine the impact of cell-specific MyD88 or TRIF deficiency on bacterial growth, but rather to establish their role in induction of an early innate immune response. MyD88 R-/H+ mice had a slightly higher bacterial burden in their lungs when compared with MyD88 R+/H+ mice (p < 0.05), whereas both TRIF R-/H+ and TRIF R-/H- mice displayed modestly elevated bacterial counts when compared with TRIF R+/H+ mice (p < 0.05). Blood cultures were sterile in all chimeric animals. Both hematopoietic and resident MyD88 and TRIF contributed to recruitment of neutrophils, although with different relative importance. Indeed, MyD88 R-/H+ and R-/H- mice showed a similar dramatic reduction in neutrophil influx into BALF when compared with MyD88 R+/H+ mice (both p < 0.001), whereas MyD88 R+/H- demonstrated an intermediate phenotype (p < 0.01 versus MyD88 R+/H+ mice and p < 0.001 versus both MyD88 R-/H- and R-/H+ mice; Figure 6A), indicating that especially MyD88 in resident cells drives the early neutrophil migration into the alveolar space during Klebsiella pneumonia. On the other hand, hematopoietic and resident TRIF appeared to be equally important: relative to TRIF R+/H+ mice, TRIF R+/H-, R-/H+ and R-/H- mice all had similarly reduced neutrophil numbers in BALF (all p < 0.05 versus TRIF R+/H+ mice; Figure 6E). Consistent with the more prominent role for resident MyD88 in neutrophil recruitment, the BALF levels of neutrophil attracting chemokines KC, MIP-2 and LIX were especially reduced in MyD88 R-/H+ mice (Figure 6B-D). With regard to KC and LIX, MyD88 R-/ H+ and R-/H- mice demonstrated similarly reduced BALF levels when compared with MyD88 R+/H+ mice; MIP-2 levels were even only diminished in MyD88 R-/H+ mice. In contrast, BALF MIP-2 and LIX concentrations were similar in all TRIF chimeras, 47

Chapter 3

Figure 4: Resident MyD88 and hematopoietic TRIF contribute to distant organ injury. WT (R+) and Myd88-/- or TRIF mutant (R-) mice were irradiated and injected with WT (H+), Myd88-/- or TRIF mutant (H-) bone marrow cells. Six weeks after transplantation, mice were infected with 7x103 CFU K. pneumoniae. Plasma levels of LDH (A), AST (B) and ALT (C) in MyD88 chimeras 24 hours after infection (n = 8-14 per group). Plasma levels of LDH (D), AST (E) and ALT (F) in TRIF mutant chimeras 48 hours after infection (n = 9-11 per group). Bars represent mean Âą standard error of the mean. * p < 0.05, ** p < 0.01, *** p < 0.001 vs R+/H+ mice determined with Mann-Whitney U test as a follow-up test on Kruskall-Wallis test.

Figure 5: Bacterial loads in early stage infection in mice chimeric for MyD88 or TRIF. WT (R+) and Myd88-/- or TRIF mutant (R-) mice were irradiated and injected with WT (H+) or Myd88-/-/ TRIF mutant (H-) bone marrow cells. Six weeks after transplantation, mice were infected with 7x103 CFU K. pneumoniae and sacrificed 6 hours later (n = 8-12 per group). Bacterial loads in lung homogenates of MyD88 chimeras (A) and of TRIF mutant chimeras (B). Each symbol represents an individual mouse, with horizontal lines showing medians. * p < 0.05, ** p < 0.01 vs R+/H+ mice determined with Mann-Whitney U test as a followup test on Kruskall-Wallis test.

Differential role of MyD88 and TRIF signaling in pneumonia

Figure 6: Differential contribution of MyD88 and TRIF in resident and hematopoietic cells to the host response in early stage infection. WT (R+) and Myd88-/- or TRIF mutant (R-) mice were irradiated and injected with WT (H+) or Myd88-/-/ TRIF mutant (H-) bone marrow cells. Six weeks after transplantation, mice were infected with 7x103 CFU K. pneumoniae and sacrificed 6 hours later (n = 8-12 per group). Number of neutrophils (A) and levels of KC (B), MIP-2 (C) and LIX (D) in BALF of MyD88 chimeras. Number of neutrophils (E) and levels of KC (F), MIP-2 (G) and LIX (H) in BALF of TRIF mutant chimeras. Bars represent mean Âą standard error of the mean.* p < 0.05, ** p < 0.01, *** p < 0.001 vs R+/ H+ mice determined with Mann-Whitney U test as a follow-up test on Kruskall-Wallis test.

Chapter 3

Figure 7: Levels of cytokines in BALF and levels of E-selectin in lung homogenates in early stage infection in mice chimeric for MyD88 or TRIF. WT (R+) and Myd88 -/- or TRIF mutant (R-) mice were irradiated and injected with WT (H+) or Myd88 -/-/ TRIF mutant (H-) bone marrow cells. Six weeks after transplantation, mice were infected with 7x10 3 CFU K. pneumoniae and sacrificed 6 hours later (n = 8-12 per group). Levels of TNF-α (A) and IL-6 (B) in BALF and levels of E-selectin in lung homogenate (C) of MyD88 chimeras. Levels of TNF-α (D) and IL-6 (E) in BALF and levels of E-selectin in lung homogenate (F) of TRIF mutant chimeras. Bars represent mean ± standard error of the mean.* p < 0.05, ** p < 0.01, *** p < 0.001 vs R+/H+ mice determined with Mann-Whitney U test as a follow-up test on Kruskall-Wallis test.

whereas KC levels were reduced in mice lacking TRIF in either hematopoietic or resident cells or both (Figures 6F-H). Resident MyD88 clearly also was most important for the early release of proinflammatory cytokines into BALF: TNF-α and IL-6 levels were equally reduced in MyD88 R-/H+ and R-/H- mice (Figure 7A and B). In contrast, hematopoietic TRIF determined TNF-α and IL-6 release: TRIF R+/H- and R-/H- mice displayed equally reduced BALF TNF-α and IL-6 concentrations when compared with either TRIF R+/ H+ or TRIF R-/H+ mice (Figure 7D and E). In addition, resident MyD88 was most important for the upregulation of E selectin (a marker for endothelial cell activation) in the lungs: MyD88 R-/H+ and R-/H mice had significantly lower lung E-selectin levels when compared to MyD88 R+/H+ mice (p < 0.05 and p < 0.001; Figure 7C). The absence of TRIF in neither resident nor hematopoietic cells influenced lung E-selectin levels (Figure 7F).

Differential role of MyD88 and TRIF signaling in pneumonia

Discussion MyD88 is the common adaptor for all TLRs (except TLR3) and deficiency of this proximal protein in TLR signaling has recently been demonstrated to result in a strongly impaired host defense during respiratory tract infection by K. pneumoniae (6). Deficiency of TRIF, which in addition to MyD88 is responsible for cellular activation by TLR4, also was associated with a hypersusceptible phenotype during Klebsiella pneumonia (6). We here show that MyD88 induced protection during Klebsiella pneumonia and sepsis is mediated by both hematopoietic and resident cells, while TRIF mediated protection is primarily driven by hematopoietic cells. These cell-specific protective functions of MyD88 and TRIF corresponded with their role in limiting bacterial growth, but not with the extent of cellular injury in distant organs, as measured by the plasma concentrations of AST, ALT and LDH. Indeed, our results indicate that MyD88 in resident cells and TRIF in hematopoietic cells contributed to cell injury during late stage infection. Hence, these findings suggest that in the present model mortality likely occurs as a consequence of excessive bacterial growth and nicely illustrate the â&#x20AC;&#x153;double edged swordâ&#x20AC;? character of innate immune activation via MyD88 and TRIF dependent signaling. The role of hematopoietic and resident MyD88 has been studied previously in murine Pseudomonas aeruginosa pneumonia (20). Herein, mice expressing MyD88 only in resident cells cleared the pathogen equally well as their controls with intact MyD88 expression (20). Accordingly, selective expression of MyD88 in lung epithelial cells was sufficient for clearance of Pseudomonas from the lungs (21). Notably, the Pseudomonas pneumonia model in mice differs considerably from the Klebsiella model used here; the current model more closely resembles the clinical scenario of a gradually growing bacterial load (22). We here demonstrate that MyD88 expressed by hematopoietic and resident cells is involved in early (< 6 hours) influx of neutrophils during Klebsiella pneumonia, but that clearly resident MyD88 plays the more prominent part. The importance of MyD88 dependent signaling in resident cells for the attraction of neutrophils in mice chimeric for MyD88 was reported previously in a model of Pseudomonas pneumonia (20). Accordingly, selective expression of MyD88 in lung epithelial cells was sufficient for neutrophil attraction during Pseudomonas pneumonia (21). TRIF signaling also contributed to neutrophil attraction into BALF, wherein hematopoietic and resident TRIF seemed to be of similar importance. The reduced BALF CXC chemokine levels in MyD88 and TRIF chimeras likely contributed to this attenuated neutrophil migration (23-25). Potential radioresistant cell populations that contribute to protective TLR signaling during Klebsiella pneumonia include epithelial, endothelial and stromal cells. Amongst these, in particular respiratory epithelial cells have been implicated to play an important role in the early phase of infection by virtue of their capacity to release an array of antimicrobial peptides and to secrete chemokines that orchestrate the recruitment of neutrophils to the alveolar space (13). Airway epithelial cells 51

Chapter 3

especially express TLR2-6 (13), strategically positioned to enable immediate recognition of organisms entering the airways. Several MyD88 dependent receptors can contribute to the hypersusceptible phenotype of MyD88 deficient mice: MyD88 is not only the adaptor for multiple TLRs but also for the IL-1 and IL-18 receptors. MyD88 dependent receptors contributing to host defense in Klebsiella pneumonia include TLR2, TLR4 and TLR9 (7-10); the potential roles of other MyD88 dependent TLRs and the IL-18 receptor have not been studied thus far, whereas IL-1 did not play a role of significance (26). The phenotype of (bone marrow chimeric) TLR2/4 double KO mice was remarkably similar to that of (bone marrow chimeric) Myd88-/- mice, suggesting that TLR2 and TLR4 are the most important MyD88 dependent receptors involved during Klebsiella infection (9). The protective role of hematopoietic TRIF reported here most likely is mediated via TLR4, considering that mice deficient for TLR3 (which relies exclusively on TRIF for signaling) demonstrate similar bacterial loads during Klebsiella pneumonia when compared with WT mice (our own unpublished data). Of note, a subset of macrophages harvested from TRIF mutant mice were reported to still respond to LPS, most likely via the TRAM adapter (15). Considering the strong phenotype of TRIF mutant mice shown here, it is likely that this pathway does not contribute significantly to protective immunity during Klebsiella infection. Our observations are in contrast with the only other report that studied the relative role of TRIF in a mouse bone marrow chimera model in a lung infection model: there the TLR3-TRIF dependent axis in resident cells was shown to be crucial for an effective host response against Aspergillus, by balancing the Th1 and Th17 response (27). Clearly, the pathogenesis and defense mechanisms that lead to a beneficial outcome during aspergillosis are different from these processes during Klebsiella infection. In conclusion, we here document for the first time to our knowledge the relative importance of the essential TLR adaptors MyD88 and TRIF in different cell types and how their contribution changes during early and late stage infection. Our results provide new insights in the pathophysiology of Klebsiella pneumonia and the potential of therapeutic targeting of TLR dependent pathways.

Acknowledgements The authors thank Dr. S. Akira (Research Institute for Microbial Diseases, Osaka, Japan) and Dr. B. Beutler (The Scripps Research Institute, La Jolla, CA) for providing the Myd88-/- and TRIF mutant mice respectively. The authors thank Joost Daalhuisen, Marieke ten Brink and Hans Roodermond for excellent technical assistance.

Differential role of MyD88 and TRIF signaling in pneumonia

References

1. Kollef MH, Shorr A, Tabak YP, Gupta V, Liu LZ, Johannes RS. Epidemiology and Outcomes of Health-Care-Associated Pneumonia: Results From a Large US Database of Culture-Positive Pneumonia. Chest 2005;128:3854-3862.

2. Opal SM, Garber GE, Larosa SP, Maki DG, Freebairn RC, Kinasewitz GT, Dhainaut JF, Yan SB, Williams MD, Graham DE, et al. Systemic Host Responses in Severe Sepsis Analyzed by Causative Microorganism and Treatment Effects of Drotrecogin Alfa (Activated). Clin Infect Dis 2003;37:50-58.

3. Giamarellou H. Multidrug Resistance in Gram-Negative Bacteria That Produce ExtendedSpectrum Beta-Lactamases (ESBLs). Clin Microbiol Infect 2005;11 Suppl 4:1-16.

4. Beutler BA. TLRs and Innate Immunity. Blood 2009;113:1399-1407.

5. Kawai T, Akira S. Toll-Like Receptors and Their Crosstalk With Other Innate Receptors in Infection and Immunity. Immunity 2011;34:637-650.

6. Cai S, Batra S, Shen L, Wakamatsu N, Jeyaseelan S. Both TRIF- and MyD88-Dependent Signaling Contribute to Host Defense Against Pulmonary Klebsiella Infection. J Immunol 2009;183:6629-6638.

7. Branger J, Knapp S, Weijer S, Leemans JC, Pater JM, Speelman P, Florquin S, van der Poll T. Role of Toll-Like Receptor 4 in Gram-Positive and Gram-Negative Pneumonia in Mice. Infect Immun 2004;72:788-794.

8. Schurr JR, Young E, Byrne P, Steele C, Shellito JE, Kolls JK. Central Role of Toll-Like Receptor 4 Signaling and Host Defense in Experimental Pneumonia Caused by GramNegative Bacteria. Infect Immun 2005;73:532-545.

9. Wieland CW, van Lieshout MH, Hoogendijk AJ, van der Poll T. Host Defence During Klebsiella Pneumonia Relies on Haematopoietic-Expressed Toll-Like Receptors 4 and 2. Eur Respir J 2011;37:848-857.

10. Bhan U, Lukacs NW, Osterholzer JJ, Newstead MW, Zeng X, Moore TA, McMillan TR, Krieg AM, Akira S, Standiford TJ. TLR9 Is Required for Protective Innate Immunity in Gram-Negative Bacterial Pneumonia: Role of Dendritic Cells. J Immunol 2007;179:3937-3946.

11. Mizgerd JP. Acute Lower Respiratory Tract Infection. N Engl J Med 2008;358:716-727.

12. Opitz B, van Laak V, Eitel J, Suttorp N. Innate Immune Recognition in Infectious and Noninfectious Diseases of the Lung. Am J Respir Crit Care Med 2010;181:1294-1309.

13. Parker D, Prince A. Innate Immunity in the Respiratory Epithelium. Am J Respir Cell Mol Biol 2011;45:189-201.

14. Adachi O, Kawai T, Takeda K, Matsumoto M, Tsutsui H, Sakagami M, Nakanishi K, Akira S. Targeted Disruption of the MyD88 Gene Results in Loss of IL-1- and IL-18-Mediated Function. Immunity 1998;9:143-150.

15. Hoebe K, Du X, Georgel P, Janssen E, Tabeta K, Kim SO, Goode J, Lin P, Mann N, Mudd S, et al. Identification of Lps2 As a Key Transducer of MyD88-Independent TIR Signalling. Nature 2003;424:743-748.

Chapter 3

16. Renckens R, Roelofs JJ, Bonta PI, Florquin S, de Vries CJ, Levi M, Carmeliet P, Van’t Veer C., van der Poll T. Plasminogen Activator Inhibitor Type 1 Is Protective During Severe GramNegative Pneumonia. Blood 2007;109:1593-1601.

17. van der Windt GJ, Schouten M, Zeerleder S, Florquin S, van der Poll T. CD44 Is Protective During Hyperoxia-Induced Lung Injury. Am J Respir Cell Mol Biol 2011;44:377-383.

18. Leemans JC, Stokman G, Claessen N, Rouschop KM, Teske GJ, Kirschning CJ, Akira S, van der Poll T, Weening JJ, Florquin S. Renal-Associated TLR2 Mediates Ischemia/Reperfusion Injury in the Kidney. J Clin Invest 2005;115:2894-2903.

19. Pulskens WP, Teske GJ, Butter LM, Roelofs JJ, Van Der Poll T, Florquin S, Leemans JC. TollLike Receptor-4 Coordinates the Innate Immune Response of the Kidney to Renal Ischemia/ Reperfusion Injury. PLoS ONE 2008;3:e3596.

20. Hajjar AM, Harowicz H, Liggitt HD, Fink PJ, Wilson CB, Skerrett SJ. An Essential Role for Non-Bone Marrow-Derived Cells in Control of Pseudomonas Aeruginosa Pneumonia. Am J Respir Cell Mol Biol 2005;33:470-475.

21. Mijares LA, Wangdi T, Sokol C, Homer R, Medzhitov R, Kazmierczak BI. Airway Epithelial MyD88 Restores Control of Pseudomonas Aeruginosa Murine Infection Via an IL-1-Dependent Pathway. J Immunol 2011;186:7080-7088.

22. Knapp S, Schultz MJ, van der Poll T. Pneumonia Models and Innate Immunity to Respiratory Bacterial Pathogens. Shock 2005;24 Suppl 1:12-18.

23. Frevert CW, Huang S, Danaee H, Paulauskis JD, Kobzik L. Functional Characterization of the Rat Chemokine KC and Its Importance in Neutrophil Recruitment in a Rat Model of Pulmonary Inflammation. J Immunol 1995;154:335-344.

24. Greenberger MJ, Strieter RM, Kunkel SL, Danforth JM, Laichalk LL, McGillicuddy DC, Standiford TJ. Neutralization of Macrophage Inflammatory Protein-2 Attenuates Neutrophil Recruitment and Bacterial Clearance in Murine Klebsiella Pneumonia. J Infect Dis 1996;173:159-165.

25. Jeyaseelan S, Manzer R, Young SK, Yamamoto M, Akira S, Mason RJ, Worthen GS. Induction of CXCL5 During Inflammation in the Rodent Lung Involves Activation of Alveolar Epithelium. Am J Respir Cell Mol Biol 2005;32:531-539.

26. Tanabe M, Matsumoto T, Shibuya K, Tateda K, Miyazaki S, Nakane A, Iwakura Y, Yamaguchi K. Compensatory Response of IL-1 Gene Knockout Mice After Pulmonary Infection With Klebsiella Pneumoniae. J Med Microbiol 2005;54:7-13.

27. de Luca A., Bozza S, Zelante T, Zagarella S, D’Angelo C, Perruccio K, Vacca C, Carvalho A, Cunha C, Aversa F, et al. Non-Hematopoietic Cells Contribute to Protective Tolerance to Aspergillus Fumigatus Via a TRIF Pathway Converging on IDO. Cell Mol Immunol 2010;7:459470.

Differential role of MyD88 and TRIF signaling in pneumonia

Supplementary appendix chapter 3 Supplementary Table 1: Lung cytokine levels in MyD88 chimeric mice.

MyD88

R+/H+

R-/H+

R+/H-

R-/H-

TNF-α

380 ± 70

444 ± 72

246 ± 40

171± 62 *

IL-6

557 ± 94

545 ± 65

662 ±132

734 ± 305

IL-10

35593 ± 4279

25895 ± 3876

41537 ± 6774

13685 ± 4239 **

MCP-1

3916 ± 440

4044 ± 414

2679 ± 237

2450 ± 770

MIP-2

8868 ± 1680

15018 ± 2886

12555 ± 2469

10999 ± 2497

Lung

Mice chimeric for MyD88 were infected with 7x103 CFU K. pneumonia 6 weeks after bone marrow transplantation. Twenty-four hours after infection, mice were sacrificed, lungs were removed and cytokine levels determined in lung homogenates. KC and MIP-2 were determined using ELISA. TNF-α, IL-6, IL-10 and MCP-1 levels were determined by Cytometric Bead Assay. Data are presented in pg/ ml lung homogenate as mean ± SEM. N=8-14 mice per group. Bd= below detection level. * p < 0.05, ** p < 0.01, *** p < 0.001 vs R+/H+ levels determined with Mann Whitney U test. R recipient and H hematopoietic.

Chapter 3 Supplementary Table 2: Lung cytokine levels in TRIF chimeric mice.

TRIF

R+/H+

R-/H+

R+/H-

R-/H-

TNF-α

196 ± 71

156 ± 39

241 ± 68

328 ± 240

IL-6

1195 ± 760

1701 ± 509

5713 ± 2035

2625 ± 1753s

IL-10

12355 ± 1829

8213 ± 1038

13561 ± 2698

11673 ± 1971 **

MCP-1

1501 ± 1047

901 ± 299

1006 ± 265

1488 ± 1157

MIP-2

6329 ± 1802

2327 ± 720

9435 ± 2262

10359 ± 2590

Lung

Mice chimeric for TRIF were infected with 7x103 CFU K. pneumoniae 6 weeks after bone marrow transplantation. Forty-eight hours after infection, mice were sacrificed, lungs were removed and cytokine levels determined in lung homogenates. KC and MIP-2 were determined using ELISA. TNF-α, IL-6, IL10 and MCP-1 levels were determined by Cytometric Bead Assay. Data are presented in pg/ml lung homogenate as mean ± SEM. N=9-11 mice per group. * p < 0.05, ** p < 0.01, *** p < 0.001 vs R+/H+ levels determined with Mann Whitney U test. R recipient and H hematopoietic.

Chapter 4 Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gramnegative Pneumonia Derived Sepsis PLoS Pathogens 2014 Sep 25;10(9):e1004368 DOI: 10.1371/journal.ppat.1004368 Miriam H.P. van Lieshout 1,2 Adam A. Anas 1,2 Sandrine Florquin 3 Baidong Hou 4 Cornelis van â&#x20AC;&#x2122;t Veer 1,2 Alex F. de Vos 1,2 Tom van der Poll 1,2,5 Academic Medical Center, Amsterdam, the Netherlands: 1 Center of Infection and Immunity Amsterdam 2 Center of Experimental and Molecular Medicine 3 Department of Pathology 5 Division of Infectious Diseases Institute of Biophysics, Chaoyang District, Beijing, China: Key Laboratory of Infection and Immunity,

Chapter 4

Abstract Klebsiella pneumoniae is an important cause of sepsis. The common Toll-like receptor adapter myeloid differentiation primary response gene (MyD)88 is crucial for host defense against Klebsiella. Here we investigated the role of MyD88 in myeloid and endothelial cells during Klebsiella pneumosepsis. Mice deficient for MyD88 in myeloid (LysM-Myd88-/-) and myeloid plus endothelial (Tie2-Myd88-/-) cells showed enhanced lethality and bacterial growth. Tie2-Myd88-/mice reconstituted with control bone marrow, representing mice with a selective MyD88 deficiency in endothelial cells, showed an unremarkable antibacterial defense. Myeloid or endothelial cell MyD88 deficiency did not impact on lung pathology or distant organ injury during late stage sepsis, while LysM-Myd88-/mice demonstrated a strongly attenuated inflammatory response in the airways early after infection. These data suggest that myeloid but not endothelial MyD88 is important for host defense during gram-negative pneumonia derived sepsis.

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis

Introduction Globally, lower respiratory tract infections are in the top ten causes of death, both in high- and low-income countries (1). Pneumonia is the most common cause of sepsis and frequently caused by gram-negative pathogens from the family of Enterobacteriaceae, including Klebsiella (K.) pneumoniae (2-4). Increasing rates of extended-spectrum β-lactamases producing Enterobacteriaceae are a major health concern and make the development of new therapies urgent, since infection with such pathogens is associated with increased mortality (5-7). Infection is detected by sensors of the innate immune system collectively called pattern recognition receptors (8, 9). Toll-like receptors (TLRs) prominently feature herein, able to detect a variety of conserved microbial patterns as well as “danger signals” released from host cells as a consequence of injurious inflammation. As such, TLRs play an important role in the initiation and amplification of the host response (8, 9). The universal adaptor for all TLRs except TLR3 is myeloid differentiation primary response gene (MyD)88, that propagates the signal of activated TLRs intracellularly, leading to NFκB and MAP kinase activation. In addition, MyD88 mediates IL-1β and IL-18 receptor signaling (10). We and others recently demonstrated the importance of MyD88 dependent signaling for survival and antibacterial defense during K. pneumoniae infection (3, 11, 12, 12). During respiratory tract infection different MyD88 expressing cells may contribute to host defense, including innate immune cells, such as alveolar macrophages, intraepithelial dendritic cells and migrated leukocytes, and parenchymal cells, such as lung epithelium and endothelium (13-15). By creating chimeric mice using bone marrow (BM) transplantation, we reported the importance of MyD88 in both radiosensitive (hematopoietic) cells and radioresistant (parenchymal) cells for antibacterial defense and survival during Klebsiella pneumonia derived sepsis (12). Whereas the role of hematopoietic cells in host defense against bacteria is undisputed, there are only few reports about the specific contribution of the vascular endothelium to the pathophysiology of infection and sepsis. Some evidence points to an attenuation of tissue and organ injury during polymicrobial sepsis when endothelial NFκB signaling was specifically targeted, without an effect on bacterial clearance (16-19). However on the other hand the specific expression of endothelial TLR4 was reported to be sufficient for adequate bacterial clearance in a model of gram-negative infection (20). Therefore, we here aimed to study the role of MyD88 dependent signaling in myeloid and endothelial cells during K. pneumoniae pneumosepsis by using mice with cell-specific targeted deletion of Myd88 and BM transfer. We demonstrate that myeloid, but nor endothelial cell MyD88 is important for host defense during pneumonia derived sepsis caused by Klebsiella.

Chapter 4

Results Genetic and functional characterization of primary cells from LysM-Myd88-/- and Tie2-Myd88-/- mice To investigate the relative contribution of MyD88 dependent signaling in myeloid and endothelial cells to protective immunity during gram-negative pneumosepsis we crossed mice homozygous for the conditional Myd88 flox allele (Myd88fl/fl mice) (21) with mice expressing Cre under control of the myeloid cell LysM promoter (to generate LysM-Myd88-/- mice) (22) or the myeloid plus endothelial cell Tie2 promoter (to generate Tie2-Myd88-/- mice) (23). To determine the efficiency of Cre-induced Myd88 deletion in specific cell types, we performed qPCR to quantify the remaining Myd88fl/fl in blood total leukocytes, granulocytes, monocytes and lymphocytes, in alveolar and peritoneal macrophages, in splenocytes and in lung endothelial and epithelial cells (figure 1A). As expected, the deletion efficiency of Cre in LysMMyd88-/- was very high in the myeloid compartment, especially in macrophages, granulocytes and to a lesser extent monocytes; lymphocytes and endothelial cells were unaffected. As anticipated, the Myd88fl/fl allele was almost completely absent in endothelial cells of Tie2-Myd88-/- mice. In addition, excision of the Myd88fl/fl allele was also virtually complete in all hematopoietic cell types of Tie2-Myd88-/- mice, as well as in lymphocytes and (accordingly) in splenocytes. Next, to determine the functional consequences of these Cre-mediated cell-specific Myd88 deletions, we incubated whole blood leukocytes, alveolar and peritoneal macrophages and splenocytes obtained from LysM-Myd88-/-, Tie2-Myd88-/- and control mice with K. pneumoniae LPS or heat-killed K. pneumoniae, using TNFÎą release as readout; we focused on these cell types since they confer protective functions during infection and sepsis (24-26). In agreement with the genetic characterization of cells from LysM-Myd88-/- and Tie2-Myd88-/- mice, whole blood leukocytes from both genotypes showed a clearly reduced responsiveness to Klebsiella and Klebsiella LPS, with Tie2-Myd88-/- leukocytes showing the largest defect (figure 1B). In addition, LysMMyd88-/- and Tie2-Myd88-/- alveolar and peritoneal macrophages displayed strongly reduced TNF-Îą release upon stimulation (figure 1C,D), while the strongest defect in splenocyte responsiveness was seen in Tie2-Myd88-/- cell cultures (figure 1E). Together these results indicate that Tie2-Myd88-/- mice are MyD88 deficient in hematopoietic, lymphoid and endothelial cells, while in LysM-Myd88-/- mice MyD88 deficiency is restricted to hematopoietic cells.

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis

Figure 1: Genetic and functional characterization of primary cells from LysM-Myd88-/- and Tie2Myd88-/- mice. The residual amount of the MyD88fl/fl allel in blood and primary cells LysM-Myd88-/- and Tie-Myd88-/- mice was quantified via qRT-PCR relative to the unaffected Socs-3 gene. The amount of remaining “floxed” MyD88 region in LysM-MyD88-/- and Tek-MyD88-/- mice was calculated using the 2-deltaCt (ΔΔCt) method using the amount of genomic DNA from Myd88fl/fl mice for the no-deletion control. The deletion efficiency was calculated as (1 - residual Myd88fl) x100% (A). Whole blood (B), alveolar and peritoneal macrophages (C,D) and splenocytes (E) derived from control, LysM-Myd88-/- and Tie2Myd88-/- mice (n=3 per group) were in vitro stimulated with LPS derived from Klebsiella pneumoniae (1μg/ml ) or heat killed K. pneumoniae in two concentrations (2x 105 CFU/ml or 2 x107/ml) for 20 hours. Data are expressed as mean (SE). * p < 0.05, ** p < 0.01, *** p < 0.001.

Chapter 4

LysM-Myd88-/- and Tie2-Myd88-/- mice demonstrate a strongly impaired host defense during gram-negative pneumosepsis Next, we infected LysM-Myd88-/- , Tie2-Myd88-/- and Myd88fl/fl Cre negative control mice with K. pneumoniae via the airways and monitored mortality during a 5-day follow up (figure 2A). LysM-Myd88-/- and Tie2-Myd88-/- mice displayed massive mortality within the first 2 days after infection with median survival times of 1.8 and 1.5 days respectively, while control mice had a median survival time of 2.9 days (both p < 0.001 versus control mice). Notably, Tie2-Myd88-/- mice showed an accelerated mortality relative to LysM-Myd88-/- mice (p < 0.01 for the difference between groups). To obtain insight in the cause of early lethality of LysM-Myd88-/- and Tie2-Myd88-/- mice we next infected mice with Klebsiella in a separate experiment and harvested lungs, blood, spleen and liver for quantitative cultures 24 hours post infection (i.e. shortly before the first deaths were expected to occur), seeking to collect data representative for host defense at the primary site of infection and bacterial dissemination. At this time point, both LysM-Myd88-/and Tie2-Myd88-/- mice had ≥ 2-log more bacteria in their lungs relative to control mice (p < 0.01 and 0.001 respectively compared to controls, figure 2B). Moreover, bacterial counts were significantly higher in blood and spleen of LysM-Myd88-/and Tie2-Myd88-/- mice (both p < 0.05 compared to control mice, figure 2C and D). In addition, Tie2-Myd88-/- mice had significantly higher amounts of bacteria in their livers (p < 0.01 compared to control mice, figure 2E). Tie2-Myd88-/- mice had higher bacterial counts when compared with LysM-Myd88-/- mice in all body sites, although these differences did not reach statistical significance. Together these data indicate that LysM-Myd88-/- and Tie2-Myd88-/- mice demonstrate a strongly enhanced bacterial growth and dissemination during gram-negative pneumonia derived sepsis, resulting in accelerated mortality. LysM-Myd88-/- and Tie2-Myd88-/- mice show modest alterations in the inflammatory and injurious response during gram-negative pneumosepsis To obtain insight in local inflammation at the primary site of infection we harvested lungs from LysM-Myd88-/-, Tie2-Myd88-/- and control mice 24 hours post infection for semi-quantitative histopathology, focusing on key histological features characteristic for severe pneumonia (figure 3). The extent of lung pathology did not differ between groups. LysM-Myd88-/- mice had lower myeloperoxidase (MPO) concentrations in whole lung homogenates, indicative of a reduced neutrophil content. In accordance, the number of Ly6+ cells was lower in LysM-Myd88-/- mice relative to controls. To obtain further insight in the role of MyD88 in cells targeted by LysM- and Tie2-driven Cre recombinase in lung inflammation during Klebsiella pneumonia, we measured the levels of the proinflammatory cytokines IL-1β, TNF-α, IL-6, the anti-inflammatory cytokine IL-10 and the neutrophil attracting chemokines CXCL-1 and CXCL-2 in lung homogenates (table 1). The pulmonary concentrations of all mediators were similar in LysM-Myd88-/-, Tie2-Myd88-/- and control mice, with the exception of TNF-α levels which were significantly lower in Tie2-Myd88-/- mice 62

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis

Figure 2: Impaired survival and bacterial defense in LysM-Myd88-/- and Tie2-Myd88-/- mice. Control, LysM-Myd88-/- and Tie2-Myd88-/- mice were intranasally infected with ~6 x103 CFU K. pneumoniae. Survival of control (dark grey symbols, n=37), LysM-Myd88-/- (light grey symbols, n=9) and Tie2-Myd88-/mice (white symbols, n=13) expressed as Kaplan-Meier plot (A), bacterial loads in lung (B), blood (C), spleen (D) and liver (E), of control (dark grey bars, n=8), LysM-Myd88-/- (light grey bars, n=8) and Tie2Myd88-/- mice (white bars, n=5 mice). Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. BC+ = number of positive blood cultures. Survival curves were compared with Log-Rank test Bacterial loads were compared to control mice determined with Mann-Whitney U test: * p < 0.05, ** p < 0.01, *** p < 0.001.

Chapter 4

Figure 3: Lung inflammatory response. Mice were intranasally infected with ~ 6 x103 CFU K. pneumoniae; Histological scores 24 hours after infection determined as described in the Methods section, in control (dark grey, n=8), LysM-Myd88-/- (light grey, n=8) and Tie2-Myd88-/- mice (white, n=5) (A). Panel (B) shows representative lung histology of control, LysM-Myd88-/- and Tie2-Myd88-/- mice H&E staining, original magnification 20x. Neutrophil influx compared between mouse groups as reflected by Ly6 lung surface positivity (C) and whole lung MPO levels (D). Panel E shows representative images of Ly-6 staining on lung slides from control, LysM-Myd88-/- and Tie2-Myd88-/- mice; Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. * p < 0.05, ** p < 0.01,

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis

(p < 0.01 compared to controls). Plasma IL-6 levels were significantly increased in LysM-Myd88-/- and Tie2-Myd88-/- mice (p < 0.05 to 0.01 respectively compared to controls, table 1) likely as a result of higher bacterial loads. In addition, we determined E-selectin levels in both lung homogenates and plasma as a reflection of endothelial cell activation (27) and observed that lung levels of E-selectin were significantly increased in Tie2-Myd88-/- mice, probably as a result of the higher bacterial burden (p < 0.05 compared to control mice) (figure S1). The model of Klebsiella pneumonia and sepsis used here is associated with rises in the plasma concentrations of LDH (indicative for cellular injury in general) and AST (reflecting hepatocellular injury) in the late stage of infection (28). To study if the absence of MyD88 in myeloid and/or endothelial cells affected the degree of liver and cellular injury we determined the plasma levels of these parameters but observed no differences (figure S2). Together these data suggest that the increased mortality in LysM-Myd88-/- and Tie2-Myd88-/- mice occurred as a result of overwhelming bacterial growth rather than as a result of pulmonary or distant organ injury. MyD88 dependent signaling in the hematopoietic compartment is crucial for antibacterial defense while MyD88 in endothelial cells is not important Considering that Tie2-Myd88-/- mice have strongly impaired MyD88 signaling in hematopoietic and endothelial cells, we decided to restore the hematopoietic compartment of Tie2-Myd88-/- mice with BM of Myd88fl/fl control mice after lethal irradiation, thereby creating mice with a more exclusive MyD88 deficiency in endothelial cells. In order to adequately estimate the effect size, we created two control groups: Tie2-Myd88-/- mice transplanted with Tie2-Myd88-/- BM and control mice transplanted with control BM. After 6 weeks of recovery, we infected mice with K. pneumoniae intranasally and sacrificed them 24 hours later. In addition, to check the efficiency of the BM transplantation to restore the responsiveness of relevant cell types from Tie2-Myd88-/- mice to Klebsiella, we euthanized 2-3 uninfected mice of each recipient group and repeated cell stimulation experiments as described above. These experiments revealed that transfer of control BM in Tie2-Myd88-/- mice fully restored the capacity of blood leukocytes, and alveolar and peritoneal macrophages, and partially that of splenocytes, to produce TNFÎą upon exposure to Klebsiella in vitro (figure 4A-D). The response of cells obtained from the two control groups transplanted with isogenic BM (control mice + control BM and Tie2-Myd88-/- mice + Tie2-Myd88-/- BM) replicated the impaired response of untransplanted Tie2-Myd88-/- mice relative to control mice. Importantly, after 24 hours of infection, lung bacterial loads of Tie2-Myd88-/- + control BM mice were indistinguishable from control + control BM mice, while the difference between Tie2-Myd88-/- + Tie2-Myd88-/- BM mice and control + control BM mice phenocopied the difference between Tie2-Myd88-/- and control mice observed in untransplanted mice (p < 0.001, figure 5A). In line, lung bacterial levels were significantly lower in Tie2-Myd88-/- + control BM mice compared to Tie2-Myd88-/- + Tie2-Myd88-/- BM mice (p < 0.01). Bacterial numbers in blood and spleen confirmed the protective 65

Chapter 4

Figure 4: Bone marrow transfer restores responsiveness of hematopoietic cells from Tie2-Myd88-/mice to Klebsiella. Whole blood (A), alveolar and peritoneal macrophages(B,C) and splenocytes (D) derived from control mice transplanted with control bone marrow (Co+ Co BM, grey bars) and Tie2Myd88-/- mice transplanted with control bone marrow (Tie2-Myd88-/- + Co BM, white dotted bars) or Tie2-Myd88-/- bone marrow (Tie2-Myd88-/- + Tie2-Myd88-/- BM, white bars). (n=2-3 per group) were in vitro stimulated with LPS derived from Klebsiella pneumoniae (1Îźg/ml ) or heat killed K. pneumoniae in two concentrations (2x 105 CFU/ml or 2 x107/ml) for 20 hours. Data are expressed as mean (SE). * p < 0.05, ** p < 0.01. ND = not determined.

effect of reconstitution of the hematopoietic compartment of Tie2-Myd88-/- mice with MyD88 sufficient BM (p < 0.05 versus Tie2-Myd88-/- + Tie2-Myd88-/- BM mice, figure 5B,C). The extent of lung pathology, lung MPO levels and the number of Ly6+ cells in lung tissue were not different between groups (figure S3). Moreover, lung and plasma cytokine/chemokine and E-selectin concentrations were not affected by the selective absence of endothelial MyD88 in Tie2-Myd88-/- + control BM mice, except for slightly lower lung levels of IL-10 compared to control + control BM mice (table S1; figure S4). Also, the plasma levels of AST and LDH did not differ between groups (figure S4). Together, these data indicate that endothelial cell MyD88 has no role in antibacterial defense or in lung or distant organ injury after infection with Klebsiella via the airways. LysM-Myd88-/- mice demonstrate an attenuated early inflammatory response Mice with a complete MyD88 deficiency show a strongly impaired antibacterial defense after infection with Klebsiella via the airways caused by a mitigated neutrophil recruitment into the airways associated with strongly reduced local levels of neutrophil attracting mediators (11). We wished to determine whether a similar mechanism is at play in LysM-Myd88-/- mice. Thus, LysM-Myd88-/-

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis

Figure 5: The absence of MyD88 in the hematopoietic compartment determines the impaired antibacterial defense of Tie2-Myd88-/- mice. Control and Tie2-Myd88-/- mice were irradiated and injected with control or Tie2-Myd88-/- bone marrow cells. Six weeks after transplantation, mice were infected with 6 x103 CFU K. pneumoniae and sacrificed after 24 hours. Bacterial loads in lung (A), blood (B), spleen (C) of control mice transplanted with control bone marrow (Co+ Co BM, grey bars, n=8) and Tie2-Myd88-/- mice transplanted with control bone marrow (Tie2-Myd88-/- + Co BM, white dotted bars) or Tie2-Myd88-/- bone marrow (Tie2-Myd88-/- + Tie2-Myd88-/- BM, white bars). Data are expressed as boxand-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. * p < 0.05, ** p < 0.01.

and control mice were infected with K. pneumoniae intranasally and lungs and bronchoalveolar lavage (BAL) fluid was harvested 6 hours later. LysM-Myd88-/mice showed higher bacterial loads in whole lung homogenates, but not in BAL fluid (figure 6A). Importantly, LysM-Myd88-/- mice displayed a strongly attenuated influx of neutrophils into the bronchoalveolar compartment (figure 6B), which was associated with markedly reduced levels of TNFÎą, CXCL-1 and CXCL-2 in BAL fluid; IL-6 concentrations in BAL fluid did not differ between groups (figure 6D). Hence, these data suggest that LysM-Myd88-/- mice replicate the phenotype of Myd88-/- mice with regard to impaired neutrophil influx in the airways during early Klebsiella pneumonia at least in part caused by a reduced chemotactic gradient due to impaired chemoattractant production.

Chapter 4

Figure 6: LysM-Myd88-/- mice demonstrate an attenuated early inflammatory response. Control and LysM-Myd88-/-- mice were intranasally infected with ~6 x103 CFU K. pneumoniae. Bacterial loads in lung (A) and BALF (B), number of neutrophils (C) and levels of TNF-Îą, CXCL-1, CXCL-2 and IL-6 (D) in BALF of control (dark grey symbols, n=8) and LysM-Myd88-/- mice (light grey symbols, n=8). Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. * p < 0.05, ** p < 0.01.

Discussion Several MyD88 dependent TLRs are known to be important for the innate immune response to respiratory tract infection with K. pneumoniae, particularly TLR4 and TLR9, and during late stage infection or in the presence of high bacterial numbers, TLR2 (29-32). Since TLRs and other innate immune sensors are widely distributed among different cell types in the airways, comprising both hematopoietic and nonhematopoietic cells, our laboratory engaged in several studies seeking to dissect the cell-specific contribution of TLR and MyD88 signaling in host defense during Klebsiella pneumonia derived sepsis (12, 32). Using BM chimeras we reported that TLR2 and TLR4 expression in hematopoietic cells are crucial for antibacterial defense, while MyD88 in hematopoietic and parenchymal cells is equally important (12, 32). BM transplantation can introduce artefacts caused by the irradiation and/ or incomplete replacement of recipient hematopoietic cells, and cannot provide detailed information about the specific cell type that is affected (33). In the present study we used the Cre-lox system combined with BM transfer to study the role of myeloid and endothelial cell specific MyD88 signalling in the host response during Klebsiella induced pneumosepsis. We demonstrate that while myeloid MyD88 contributes significantly to host defense, endothelial cell MyD88 has no role herein. 68

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis

Endothelial cells are resident cells implicated in sepsis pathogenesis and the induction of organ injury (34). Earlier investigations examined the contribution of TLR and NFκB signaling within the vascular endothelium to the host response during experimental sepsis. Inhibition of endothelial NFκB signaling by overexpression of a degradation-resistant form of the NF-κB inhibitor I-κBα under the control of the endothelial cell specific VE-cadherin-5 promoter attenuated tissue inflammation and organ injury during endotoxemia and abdominal sepsis (16-19). In addition, these mice displayed strongly reduced coagulation activation upon administration of endotoxin (19). Endothelial cell specific NFκB inhibition did not influence the clearance of Listeria monocytogenesis, Streptococcus pneumoniae or Salmonella enterica after intravenous infection (16-19). However, transgenic Tie2 driven expression of TLR4 in Tlr4-/- mice, resulting in mice with TLR4 expression restricted to endothelial cells, was sufficient for adequate bacterial clearance after intraperitoneal infection with Escherichia coli (20). We used mice with Tie2 driven expression of Cre recombinase to delete hematopoietic and endothelial MyD88 in Myd88fl/fl mice and observed a strongly impaired host defense as reflected by very high bacterial loads and increased mortality. Previous studies support Tie2 expression in hematopoietic cells and the lack of specificity for endothelial cells (23, 35). Similarly, the VE-cadherin-5 promoter is reported to drive Cre recombinase gene expression not only in endothelial cells but also in a subset of hematopoietic cells (36). As such, the Cre-lox system seems less suitable to specifically study the function of genes in endothelial cells. Therefore, to generate mice with endothelial cell specific MyD88 deficiency, we reconstituted Tie2-Myd88-/- mice with BM of control mice and confirmed functional recovery of their hematopoietic cells with regard to responsiveness to Klebsiella. These mice were indistinguishable from control mice with regard to antibacterial defense, inflammation and distant organ injury, strongly suggesting that endothelial cell MyD88 does not play an important role in the host response during Klebsiella induced pneumosepsis. Although this “negative” finding may seem to contrast with previous studies on the role of endothelial cells in severe infection (16-20), our approach clearly differs from these earlier investigations, both with regard to the target of genetic manipulation (deletion of MyD88 versus inhibition of NFκB (16-19) and endothelial cell TLR4 expression on an otherwise TLR4 deficient background (20) and the sepsis model used (pneumonia versus abdominal or intravenous infection (16-20). Importantly, mice with TLR4 exclusively on endothelial cells were unable to recruit neutrophils into the lungs upon intratracheal LPS administration (20) and, similarly, studies in TLR4 BM chimeras have indicated that neutrophil influx after airway exposure to LPS occurs by mechanisms that do not rely on TLR4 expression by radioresistant (including endothelial) cells (37), which is completely consistent with our present data. Of note, findings in TLR4 BM chimeras have suggested that neutrophil accumulation in lungs upon intravenous LPS challenge does largely dependent on TLR4 in radioresistant cells (38), indicating that the role of cell-specific TLR signaling in neutrophil recruitment likely depends on the route by which the bacterial stimulus is administered.

Chapter 4

LysM-Cre mediated deletion of the floxed Myd88 allele resulted in MyD88 deficiency especially in macrophages and neutrophils, and to a lesser extent monocytes (22, 39). Clearly, these myeloid cell MyD88 deficient mice showed a strongly compromised host defense after infection with Klebsiella, as reflected by enhanced mortality, increased bacterial numbers at the primary site of infection and an impaired early neutrophil influx and cytokine/chemokine release in the airways. Thus, MyD88 expressed by alveolar macrophages and neutrophils is essential for initiation of an adequate early innate immune response in the lung after infection with Klebsiella via the airways and the absence thereof results in uncontrolled bacterial growth and death. The phenotype of LysM-Myd88-/- mice was very similar to the previously documented phenotype of Myd88-/- mice during Klebsiella pneumonia (3, 11, 12), underlining the importance of myeloid cell MyD88 during respiratory tract infection. The Klebsiella strain used here cannot be killed by macrophages or neutrophils in vitro, illustrating its high virulence and precluding analysis of a possible direct role of MyD88 in killing. Previous studies have reported a role for MyD88 in killing of commensal and attenuated pathogenic Gram-negative bacteria (40), but not in killing of Listeria by macrophages (41). While innate immunity is important for antibacterial defense, it can also cause harm by hyperinflammation induced organ injury (42, 43). Deficiency of MyD88 has been shown to be protective in polymicrobial sepsis, in which especially liver injury was found to be associated with MyD88 dependent signaling (43, 44). A recent study demonstrated that mice with selective expression of MyD88 in myeloid cells displayed enhanced hepatocellular injury during abdominal sepsis induced by cecal ligation and puncture (45). Here we found no evidence for a role of either myeloid or endothelial cell MyD88 in hepatocellular damage during pneumonia derived sepsis caused by K. pneumoniae. Hence, although MyD88 may contribute to organ injury during sepsis, its role likely depends on the type and primary source of the infection. Using MyD88 BM chimeras, we recently reported a role for both hematopoietic and parenchymal MyD88 in host defense in this model (12). Since we could not demonstrate a role for endothelial cell MyD88 in the present investigation, MyD88 expressed in the respiratory epithelium may be involved. Indeed, lung epithelial cells have been implicated in host defense during respiratory tract infection (15). The importance of MyD88 dependent signaling in lung epithelial cells was recently elegantly demonstrated in a model of Pseudomonas pneumonia in epithelial specific MyD88 knock-in mice (46, 47). Selective expression of MyD88 in the airway epithelium was sufficient for neutrophil recruitment to the site of infection and bacterial clearance (47). In addition, transgenic overexpression of IÎşB-Îą in alveolar and bronchial epithelium in mice resulted in a reduced neutrophil influx into BAL fluid upon intrapulmonary delivery of LPS (48, 49) and an increased growth of the gram-positive pathogen Streptococcus pneumoniae upon intratracheal infection (50). Studies using mice in which Myd88 is deleted specifically in respiratory epithelium are warranted to establish the role of epithelial MyD88 in host defense against Klebsiella pneumonia derived sepsis. However, our first preliminary 70

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis

results with mice generated from intercrossings of Myd88fl/fl mice and mice with Cre recombinase controlled by the surfactant protein C promoter (51), resulting in mice with a targeted deletion of Myd88 in distal airway epithelium, suggest that epithelial cell MyD88 does not contribute to protective immunity during Klebsiella pneumonia. Therefore, our earlier data using MyD88 BM chimeras (12) may have been confounded by incomplete replacement of recipient (MyD88 sufficient) hematopoietic cells. LysM-Myd88-/- mice showed a strongly impaired neutrophil influx into the bronchoalveolar space 6 hours after infection with Klebsiella, together with markedly reduced local concentrations of neutrophil attracting mediators such as TNF-Îą, CXCL1 and CXCL2. Notably, global MyD88 deficiency similarly results in an early impairment of neutrophil chemoattractant release and neutrophil migration into the airways in mouse models of pneumonia caused by a variety of bacterial and viral species (11, 52-57), as well as during sterile lung inflammation (58, 59). These data suggest that hematopoietic and global MyD88 deficiency impair host defense during pneumonia by a largely similar mechanism that involves an inability to produce a chemotactic gradient that would normally attract neutrophils to the site of the infection. Notably, MyD88 deficient mice showed extensive lung inflammation, including high E-selectin levels, at 24 hours after infection, suggesting that these late responses can be induced by Klebsiella via MyD88-independent mechanisms (e.g., via the TRIF pathway) in the presence of the (by then) very high bacterial loads. Similarly, global Myd88-/- mice were previously reported to show profound lung inflammation during late stage bacterial pneumonia in the presence of high bacterial loads (54, 56, 60). E-selectin, while implicated in the rolling of neutrophils along the vascular endothelium (61), does not seem to play a role in neutrophil recruitment to the lungs elicited by bacterial stimuli (62, 63). In conclusion, to our knowledge, we here report for the first time on the role of MyD88 in myeloid and endothelial cells in severe bacterial infection, using a clinically relevant model of gram-negative pneumonia derived sepsis characterized by gradual growth of bacteria at the primary site of infection followed by dissemination, tissue injury and death. While myeloid MyD88 was crucial for protective immunity, endothelial MyD88 played no role herein. Our results suggest that myeloid MyD88 deficiency results in enhanced lethality during Klebsiella pneumonia by a mechanism that involves a strongly attenuated early inflammatory response at the primary site of infection and as a consequence thereof uncontrolled bacterial growth. These data provide new insights in the pathophysiology of gram-negative sepsis and may be helpful for the development of therapeutics aimed at specific cell types.

Chapter 4

Materials and methods Ethics statement Experiments were carried out in accordance with the Dutch Experiment on Animals Act and approved by the Animal Care and Use Committee of the University of Amsterdam (Permit number: DIX 100121, sub-protocols DIX102300 and DIX101613). Animals Homozygous Myd88fl/fl mice (21) were crossed with LysM-Cre (22) or Tie2-Cre mice (23), both obtained from the Jackson Laboratory (Bar Harbor, Maine), to generate myeloid (LysM-Myd88-/-) and myeloid plus endothelial cell (Tie2-Myd88-/-) specific MyD88 deficient mice. Myd88fl/fl Cre negative littermates were used as controls. All mice were backcrossed at least 8 times to a C57Bl/6 background and age- and sex matched when used in experiments. Harvest of primary cells for genetic and functional characterization of LysM-Myd88-/and Tie2-Myd88-/- mice Peritoneal lavage was performed with 5 ml sterile PBS under isoflurane anesthesia and lavage fluid was collected in PBS containing a final concentration of 10% FBS, 1% antibiotics (penicillin- streptomycin- amphotericin B (Gibco, Paisley, United Kingdom); heart puncture was performed and blood was collected in EDTA or heparin containing tubes; BAL was performed with 10 ml PBS in portions to obtain alveolar macrophages and spleens were harvested. For whole blood stimulation, 100 μl of heparinized blood was pipetted in a 96 wells U-bottom cell culture plate (Greiner bio-one, Alphen a/d Rijn, Netherlands). Spleens were crushed through a 40μm mesh and after lysis of erythrocytes with an ammoniumchloride containing lysis buffer, splenocytes were seeded in RPMI complete (containing 10% FBS, 1% antibiotics, 10mM L-glutamine, Gibco) at a density of 500.000 cells per well in 96 wells U-bottom culture plate (Greiner bio-one). Peritoneal and alveolar macrophages were seeded in flat bottom 96 wells cell culture plates (Greiner Bio-one) at a density of approximately 50.000 and 30.000 respectively per well in RPMI complete and left to adhere overnight. Cells were stimulated for 20 hours with the indicated concentrations of heat-killed K. pneumoniae or LPS derived from Klebsiella pneumoniae (Sigma) diluted in RPMI complete medium in a final volume of 200 microliter. Whole blood leukocyte genomic DNA was isolated from fresh EDTA blood and primary cells using the Nucleospin Blood Kit (Machery Nagel, Düren, Germany) and in addition, from FACS purified monocytes, neutrophils and lymphocytes. For this, erythrocyte lysis of EDTA blood with ammoniumchloride containing lysis buffer was performed and cells were stained for cell surface molecules using FITCconjugated anti-mouse Ly6-C &Ly6-G (Gr-1), PE-conjugated anti-mouse CD11b (BD Biosciences) and biotinylated anti-mouse CD115 (eBioscience), secondary staining was performed with streptavidin-APC (BD Biosciences). Monocytes were identified as Gr-1dim/CD-115+ and neutrophils as Gr-1high/ Cd115- within the fraction of CD11b+ cells, the fraction of Cd11b- cells with a low Side Scatter and Forward 72

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis

Scatter pattern were identified as lymphoid cells (64). Real-time PCR Total RNA was reverse transcribed using oligo (dT) primer and Moloney murine leukemia virus reverse transcriptase (Invitrogen, Breda, the Netherlands). We quantified the residual amount of the “floxed” region of MyD88 in LysM-Myd88-/- and Tie2-Myd88-/- mice in blood and particular cell types using the primer sequences 5’-ACGCCGGAACTTTTCGAT-3’ (forward); 5’-TTTTCTCAATTAGCTCGCTGG-3’ relative to the unaffected Socs-3 gene with primer sequences 5’- ACCTTTCTTATCCGCGACAG- 3’ (forward) and 5’TGCACCAGCTTGAGTACACAG-3’ (reverse) in a SybrGreen reaction on an LightCycler system (LC480, Roche Applied Science, Mannheim, Germany). The amount of remaining “floxed” MyD88 region in LysM-MyD88-/- and Tie2-MyD88-/mice was calculated using the 2-deltaCt (ΔΔCt) method using the amount of genomic DNA from Myd88fl/fl mice for the no-deletion control (21). The deletion efficiency was calculated as (1 - residual Myd88fl) x100. Induction of pneumonia and sampling of organs Pneumonia was induced by intranasal inoculation with ~ 6x103 colony forming units (CFU) of K. pneumoniae serotype 2 (ATCC 43816; American Type Culture Collection, Manassas, VA) and survival was monitored or in separate experiments mice were euthanized after 6 or 24 hours of infection when organs were harvested and processed exactly as described (12, 32). Measurements of inflammatory proteins and clinical chemistry Lung (and cell supernatant) levels of IL-1β, TNF-α, IL-6, IL-10, CXCL-1 and CXCL-2 were measured by ELISA (R&D Systems, Minneapolis, MN). Plasma levels of TNF-α, IL-6, and IL-10 were measured by using a cytometric bead array multiplex assay (BD Biosciences). MPO was measured by ELISA from HyCult Biotechnology (Uden, the Netherlands).Lactate dehydrogenase (LDH) and aspartate aminotransferase (AST were measured using kits from Sigma and a Hittachi analyzer (Boehringer Mannheim). Histopathology Histologic examination of lungs was performed exactly as described (32). For granulocyte immunohistochemic stainings lung tissue slides were deparaffinized and rehydrated. Endogenous peroxidase activity was quenched by a solution of 0.3% H2O2 (Merck). Slides were then digested by a solution of pepsin 0.025% (Sigma, St. Louis, MO, USA) in 0.1 M HCl. After being rinsed, the sections were incubated in Ultra V Block (Thermo Scientific, Fremont, CA) and then exposed to a FITC-labeled anti-mouse Ly6-G and Ly6-C monoclonal antibody (BD Pharmingen, San Diego, CA). After washes, slides were incubated with a rabbit anti-FITC antibody (Nuclilab, Ede, the Netherlands) followed by further incubation with Brightvision poly-horseradish peroxidase anti Rabbit IgG (Immunologic, Duiven, the Netherlands), rinsed again and developed using Bright DAB (Immunologic, Duiven, the Netherlands). The sections were counterstained with methyl green and 73

Chapter 4

mounted in Pertex mounting medium (Histolab, Gothenburg, Sweden). The Ly-6G and Ly-6C+ percentage of total lung surface was determined with imageJ software (Rasband, W.S., ImageJ, U. S. National Institutes of Health, Bethesda, Maryland, USA, http://rsb.info.nih.gov/ij/, 1997-2011). Bone marrow transplantation BM transplantation was done as described previously (12). Three groups were generated: Tie2-Myd88-/- (recipient) + control BM (donor), Tie2-Myd88-/- + Tie2Myd88-/- BM and control + control BM mice. Myd88fl/fl mice and BM were used as control. Statistical analysis Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation (in vivo experiments) or as means Âą standard error of the mean (tables, cell stimulation experiments); Comparison of these data was done by Mann Whitney U test. Differences in the proportion of positive cultures were analyzed by Fisherâ&#x20AC;&#x2122;s exact test. Survival curves are depicted as Kaplan-Meier plots and compared using logrank test. These analyses were done using GraphPad Prism (San Diego, CA). p < 0.05 was considered statistically significant.

Acknowledgments Anthony L. DeFranco (Department of Microbiology & Immunology, University of California, San Francisco) kindly supplied us with the Myd88 fl/fl mice. The authors thank Regina de Beer, Marieke ten Brink and Joost Daalhuisen (Center of Experimental and Molecular Medicine), Onno de Boer (Department of pathology) and Berend Hooijbrink, (Flow Cytometry Facility, Department of Cell Biology) from the Academic Medical Center, Amsterdam, the Netherlands for expert technical assistance.

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis

References

1. World Health Organisation WHO Causes of Death 2008, Global Health Observatory, World Health Organisation.; 2012.

2. Kollef MH, Shorr A, Tabak YP, Gupta V, Liu LZ, Johannes RS. Epidemiology and Outcomes of Health-Care-Associated Pneumonia: Results From a Large US Database of Culture-Positive Pneumonia. Chest 2005;128:3854-3862.

3. Welte T, Torres A, Nathwani D. Clinical and Economic Burden of Community-Acquired Pneumonia Among Adults in Europe. Thorax 2012;67:71-79.

4. Zahar JR, Timsit JF, Garrouste-Org, Francais A, Vesin A, Descorps-Declere A, Dubois Y, Souweine B, Haouache H, Goldgran-Toledano D, et al. Outcomes in Severe Sepsis and Patients With Septic Shock: Pathogen Species and Infection Sites Are Not Associated With Mortality. Crit Care Med 2011;39:1886-1895.

5. Coque TM, Baquero F, Canton R. Increasing Enterobacteriaceae in Europe. Euro Surveill 2008;13.

6. Giamarellou H. Multidrug Resistance in Gram-Negative Bacteria That Produce ExtendedSpectrum Beta-Lactamases (ESBLs). Clin Microbiol Infect 2005;11 Suppl 4:1-16.

7. Schwaber MJ, Carmeli Y. Mortality and Delay in Effective Therapy Associated With ExtendedSpectrum Beta-Lactamase Production in Enterobacteriaceae Bacteraemia: a Systematic Review and Meta-Analysis. J Antimicrob Chemother 2007;60:913-920.

8. Beutler BA. TLRs and Innate Immunity. Blood 2009;113:1399-1407.

9. Kawai T, Akira S. Toll-Like Receptors and Their Crosstalk With Other Innate Receptors in Infection and Immunity. Immunity 2011;34:637-650.

10. Adachi O, Kawai T, Takeda K, Matsumoto M, Tsutsui H, Sakagami M, Nakanishi K, Akira S. Targeted Disruption of the MyD88 Gene Results in Loss of IL-1- and IL-18-Mediated Function. Immunity 1998;9:143-150.

11. Cai S, Batra S, Shen L, Wakamatsu N, Jeyaseelan S. Both TRIF- and MyD88-Dependent Signaling Contribute to Host Defense Against Pulmonary Klebsiella Infection. J Immunol 2009;183:6629-6638.

12. van Lieshout MH, Blok DC, Wieland CW, de Vos AF, van â&#x20AC;&#x2DC;t Veer C, van der Poll T. Differential Roles of MyD88 and TRIF in Hematopoietic and Resident Cells During Murine Gram-Negative Pneumonia. J Infect Dis 2012;206:1415-1423.

13. Kawai T, Akira S. The Role of Pattern-Recognition Receptors in Innate Immunity: Update on Toll-Like Receptors. Nat Immunol 2010;11:373-384.

14. Opitz B, van L, V, Eitel J, Suttorp N. Innate Immune Recognition in Infectious and Noninfectious Diseases of the Lung. Am J Respir Crit Care Med 2010;181:1294-1309.

15. Parker D, Prince A. Innate Immunity in the Respiratory Epithelium. Am J Respir Cell Mol Biol 2011;45:189-201.

Prevalence

ESBL-Producing

Chapter 4

16. Ding J, Song D, Ye X, Liu SF. A Pivotal Role of Endothelial-Specific NF-KappaB Signaling in the Pathogenesis of Septic Shock and Septic Vascular Dysfunction. J Immunol 2009;183:40314038.

17. Xu H, Ye X, Steinberg H, Liu SF. Selective Blockade of Endothelial NF-KappaB Pathway Differentially Affects Systemic Inflammation and Multiple Organ Dysfunction and Injury in Septic Mice. J Pathol 2010;220:490-498.

18. Ye X, Ding J, Zhou X, Chen G, Liu SF. Divergent Roles of Endothelial NF-KappaB in Multiple Organ Injury and Bacterial Clearance in Mouse Models of Sepsis. J Exp Med 2008;205:13031315.

19. Song D, Ye X, Xu H, Liu SF. Activation of Endothelial Intrinsic NF-{Kappa}B Pathway Impairs Protein C Anticoagulation Mechanism and Promotes Coagulation in Endotoxemic Mice. Blood 2009;114:2521-2529.

20. Andonegui G, Zhou H, Bullard D, Kelly MM, Mullaly SC, McDonald B, Long EM, Robbins SM, Kubes P. Mice That Exclusively Express TLR4 on Endothelial Cells Can Efficiently Clear a Lethal Systemic Gram-Negative Bacterial Infection. J Clin Invest 2009;119:1921-1930.

21. Hou B, Reizis B, DeFranco AL. Toll-Like Receptors Activate Innate and Adaptive Immunity by Using Dendritic Cell-Intrinsic and -Extrinsic Mechanisms. Immunity 2008;29:272-282.

22. Clausen BE, Burkhardt C, Reith W, Renkawitz R, Forster I. Conditional Gene Targeting in Macrophages and Granulocytes Using LysMcre Mice. Transgenic Res 1999;8:265-277.

23. Koni PA, Joshi SK, Temann UA, Olson D, Burkly L, Flavell RA. Conditional Vascular Cell Adhesion Molecule 1 Deletion in Mice: Impaired Lymphocyte Migration to Bone Marrow. J Exp Med 2001;193:741-754.

24. Eddens T, Kolls JK. Host Defenses Against Bacterial Lower Respiratory Tract Infection. Curr Opin Immunol 2012;24:424-430.

25. Takabayshi K, Corr M, Hayashi T, Redecke V, Beck L, Guiney D, Sheppard D, Raz E. Induction of a Homeostatic Circuit in Lung Tissue by Microbial Compounds. Immunity 2006;24:475-487.

26. Mebius RE, Kraal G. Structure and Function of the Spleen. Nat Rev Immunol 2005;5:606-616.

27. Achouiti A, de Vos AF, de Beer R., Florquin S, Vanâ&#x20AC;&#x2122;t Veer C., van der Poll T. Limited Role of the Receptor for Advanced Glycation End Products During Streptococcus Pneumoniae Bacteremia. J Innate Immun 2013.

28. Renckens R, Roelofs JJ, Bonta PI, Florquin S, de Vries CJ, Levi M, Carmeliet P, Vanâ&#x20AC;&#x2122;t Veer C., van der Poll T. Plasminogen Activator Inhibitor Type 1 Is Protective During Severe GramNegative Pneumonia. Blood 2007;109:1593-1601.

29. Bhan U, Lukacs NW, Osterholzer JJ, Newstead MW, Zeng X, Moore TA, McMillan TR, Krieg AM, Akira S, Standiford TJ. TLR9 Is Required for Protective Innate Immunity in Gram-Negative Bacterial Pneumonia: Role of Dendritic Cells. J Immunol 2007;179:3937-3946.

30. Branger J, Knapp S, Weijer S, Leemans JC, Pater JM, Speelman P, Florquin S, van der Poll T. Role of Toll-Like Receptor 4 in Gram-Positive and Gram-Negative Pneumonia in Mice. Infect Immun 2004;72:788-794.

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis

31. Schurr JR, Young E, Byrne P, Steele C, Shellito JE, Kolls JK. Central Role of Toll-Like Receptor 4 Signaling and Host Defense in Experimental Pneumonia Caused by GramNegative Bacteria. Infect Immun 2005;73:532-545.

32. Wieland CW, van Lieshout MH, Hoogendijk AJ, van der Poll T. Host Defence During Klebsiella Pneumonia Relies on Haematopoietic-Expressed Toll-Like Receptors 4 and 2. Eur Respir J 2011;37:848-857.

33. Duran-Struuck R, Dysko RC. Principles of Bone Marrow Transplantation (BMT): Providing Optimal Veterinary and Husbandry Care to Irradiated Mice in BMT Studies. J Am Assoc Lab Anim Sci 2009;48:11-22.

34. Schouten M, Wiersinga WJ, Levi M, van der Poll T. Inflammation, Endothelium, and Coagulation in Sepsis. J Leukoc Biol 2008;83:536-545.

35. Tang Y, Harrington A, Yang X, Friesel RE, Liaw L. The Contribution of the Tie2+ Lineage to Primitive and Definitive Hematopoietic Cells. Genesis 2010;48:563-567.

36. Alva JA, Zovein AC, Monvoisin A, Murphy T, Salazar A, Harvey NL, Carmeliet P, Iruela-Arispe ML. VE-Cadherin-Cre-Recombinase Transgenic Mouse: a Tool for Lineage Analysis and Gene Deletion in Endothelial Cells. Dev Dyn 2006;235:759-767.

37. Hollingsworth JW, Chen BJ, Brass DM, Berman K, Gunn MD, Cook DN, Schwartz DA. The Critical Role of Hematopoietic Cells in Lipopolysaccharide-Induced Airway Inflammation. Am J Respir Crit Care Med 2005;171:806-813.

38. Andonegui G, Bonder CS, Green F, Mullaly SC, Zbytnuik L, Raharjo E, Kubes P. EndotheliumDerived Toll-Like Receptor-4 Is the Key Molecule in LPS-Induced Neutrophil Sequestration into Lungs. J Clin Invest 2003;111:1011-1020.

39. Kong X, Thimmulappa R, Craciun F, Harvey C, Singh A, Kombairaju P, Reddy SP, Remick D, Biswal S. Enhancing Nrf2 Pathway by Disruption of Keap1 in Myeloid Leukocytes Protects Against Sepsis. Am J Respir Crit Care Med 2011;184:928-938.

40. Laroux FS, Romero X, Wetzler L, Engel P, Terhorst C. Cutting Edge: MyD88 Controls Phagocyte NADPH Oxidase Function and Killing of Gram-Negative Bacteria. J Immunol 2005;175:5596-5600.

41. Edelson BT, Unanue ER. MyD88-Dependent but Toll-Like Receptor 2-Independent Innate Immunity to Listeria: No Role for Either in Macrophage Listericidal Activity. J Immunol 2002;169:3869-3875.

42. Mancuso G, Midiri A, Beninati C, Biondo C, Galbo R, Akira S, Henneke P, Golenbock D, Teti G. Dual Role of TLR2 and Myeloid Differentiation Factor 88 in a Mouse Model of Invasive Group B Streptococcal Disease. J Immunol 2004;172:6324-6329.

43. Weighardt H, Kaiser-Moore S, Vabulas RM, Kirschning CJ, Wagner H, Holzmann B. Cutting Edge: Myeloid Differentiation Factor 88 Deficiency Improves Resistance Against Sepsis Caused by Polymicrobial Infection. J Immunol 2002;169:2823-2827.

44. Weighardt H, Mages J, Jusek G, Kaiser-Moore S, Lang R, Holzmann B. Organ-Specific Role of MyD88 for Gene Regulation During Polymicrobial Peritonitis. Infect Immun 2006;74:36183632.

Chapter 4

45. Gais P, Reim D, Jusek G, Rossmann-Bloeck T, Weighardt H, Pfeffer K, Altmayr F, Janssen KP, Holzmann B. Cutting Edge: Divergent Cell-Specific Functions of MyD88 for Inflammatory Responses and Organ Injury in Septic Peritonitis. J Immunol 2012;188:5833-5837.

46. Hajjar AM, Harowicz H, Liggitt HD, Fink PJ, Wilson CB, Skerrett SJ. An Essential Role for Non-Bone Marrow-Derived Cells in Control of Pseudomonas Aeruginosa Pneumonia. Am J Respir Cell Mol Biol 2005;33:470-475.

47. Mijares LA, Wangdi T, Sokol C, Homer R, Medzhitov R, Kazmierczak BI. Airway Epithelial MyD88 Restores Control of Pseudomonas Aeruginosa Murine Infection Via an IL-1-Dependent Pathway. J Immunol 2011;186:7080-7088.

48. Poynter ME, Cloots R, van WT, Butnor KJ, Vacek P, Taatjes DJ, Irvin CG, JanssenHeininger YM. NF-Kappa B Activation in Airways Modulates Allergic Inflammation but Not Hyperresponsiveness. J Immunol 2004;173:7003-7009.

49. Skerrett SJ, Liggitt HD, Hajjar AM, Ernst RK, Miller SI, Wilson CB. Respiratory Epithelial Cells Regulate Lung Inflammation in Response to Inhaled Endotoxin. Am J Physiol Lung Cell Mol Physiol 2004;287:L143-L152.

50. Quinton LJ, Jones MR, Simms BT, Kogan MS, Robson BE, Skerrett SJ, Mizgerd JP. Functions and Regulation of NF-KappaB RelA During Pneumococcal Pneumonia. J Immunol 2007;178:1896-1903.

51. Korfhagen TR, Glasser SW, Wert SE, Bruno MD, Daugherty CC, McNeish JD, Stock JL, Potter SS, Whitsett JA. Cis-Acting Sequences From a Human Surfactant Protein Gene Confer Pulmonary-Specific Gene Expression in Transgenic Mice. Proc Natl Acad Sci U S A 1990;87:6122-6126.

52. Albiger B, Sandgren A, Katsuragi H, Meyer-Hoffert U, Beiter K, Wartha F, Hornef M, Normark S, Normark BH. Myeloid Differentiation Factor 88-Dependent Signalling Controls Bacterial Growth During Colonization and Systemic Pneumococcal Disease in Mice. Cell Microbiol 2005;7:1603-1615.

53. Bradley LM, Douglass MF, Chatterjee D, Akira S, Baaten BJ. Matrix Metalloprotease 9 Mediates Neutrophil Migration into the Airways in Response to Influenza Virus-Induced TollLike Receptor Signaling. PLoS Pathog 2012;8:e1002641.

54. Hawn TR, Smith KD, Aderem A, Skerrett SJ. Myeloid Differentiation Primary Response Gene (88)- and Toll-Like Receptor 2-Deficient Mice Are Susceptible to Infection With Aerosolized Legionella Pneumophila. J Infect Dis 2006;193:1693-1702.

55. Naiki Y, Michelsen KS, Schroder NW, Alsabeh R, Slepenkin A, Zhang W, Chen S, Wei B, Bulut Y, Wong MH, et al. MyD88 Is Pivotal for the Early Inflammatory Response and Subsequent Bacterial Clearance and Survival in a Mouse Model of Chlamydia Pneumoniae Pneumonia. J Biol Chem 2005;280:29242-29249.

56. Skerrett SJ, Liggitt HD, Hajjar AM, Wilson CB. Cutting Edge: Myeloid Differentiation Factor 88 Is Essential for Pulmonary Host Defense Against Pseudomonas Aeruginosa but Not Staphylococcus Aureus. J Immunol 2004;172:3377-3381.

57. Wiersinga WJ, Wieland CW, Roelofs JJ, van der Poll T. MyD88 Dependent Signaling Contributes to Protective Host Defense Against Burkholderia Pseudomallei. PLoS ONE 2008;3:e3494.

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis

58. Doz E, Noulin N, Boichot E, Guenon I, Fick L, Le BM, Lagente V, Ryffel B, Schnyder B, Quesniaux VF, et al. Cigarette Smoke-Induced Pulmonary Inflammation Is TLR4/MyD88 and IL-1R1/MyD88 Signaling Dependent. J Immunol 2008;180:1169-1178.

59. Gasse P, Mary C, Guenon I, Noulin N, Charron S, Schnyder-Candrian S, Schnyder B, Akira S, Quesniaux VF, Lagente V, et al. IL-1R1/MyD88 Signaling and the Inflammasome Are Essential in Pulmonary Inflammation and Fibrosis in Mice. J Clin Invest 2007;117:3786-3799.

60. Wieland CW, Florquin S, Maris NA, Hoebe K, Beutler B, Takeda K, Akira S, van der Poll T. The MyD88-Dependent, but Not the MyD88-Independent, Pathway of TLR4 Signaling Is Important in Clearing Nontypeable Haemophilus Influenzae From the Mouse Lung. J Immunol 2005;175:6042-6049.

61. Kolaczkowska E, Kubes P. Neutrophil Recruitment and Function in Health and Inflammation. Nat Rev Immunol 2013;13:159-175.

62. Burns JA, Issekutz TB, Yagita H, Issekutz AC. The Alpha 4 Beta 1 (Very Late Antigen (VLA)4, CD49d/CD29) and Alpha 5 Beta 1 (VLA-5, CD49e/CD29) Integrins Mediate Beta 2 (CD11/ CD18) Integrin-Independent Neutrophil Recruitment to Endotoxin-Induced Lung Inflammation. J Immunol 2001;166:4644-4649.

63. Mizgerd JP, Meek BB, Kutkoski GJ, Bullard DC, Beaudet AL, Doerschuk CM. Selectins and Neutrophil Traffic: Margination and Streptococcus Pneumoniae-Induced Emigration in Murine Lungs. J Exp Med 1996;184:639-645.

64. de Bruin AM, Libregts SF, Valkhof M, Boon L, Touw IP, Nolte MA. IFNgamma Induces Monopoiesis and Inhibits Neutrophil Development During Inflammation. Blood 2012;119:15431554.

Chapter 4

Supplementary appendix Chapter 4

Figure S1: Lung endothelial cell activation as reflected by e-selectin is higher in Tie2-MyD88-/mice. Control, LysM-MyD88-/- and Tie2-MyD88-/- mice were inoculated with Ě´ 6x103 CFU K. pneumoniae and sacrificed 24 hours later. Homogenates were prepared from right lungs. E-selectin levels are presented in pg/ml lung homogenate (A) or plasma (B). Data are mean (SE) of 5-8 mice per group. *p < 0.05 vs control mice.

Figure S2: Absence of hematopoietic or endothelial Myd88 does not impact on organ injury. Control, LysM-MyD88-/- and Tie2-MyD88-/- mice were inoculated with Ě´ 6x103 CFU K. pneumoniae and sacrificed 24 hours later. Plasma levels of LDH (A) and AST (B) after 24 hours. Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation.

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis

Figure S3: Local inflammatory response is not affected by the absence of MyD88 expression in the endothelial compartment Control and Tie2-MyD88-/- mice were irradiated and injected with control or Tie2-MyD88-/- bone marrow cells. Six weeks after transplantation, mice were infected with 6 x103 CFU K. pneumoniae. Histological scores 24 hours after infection were determined of control mice transplanted with control bone marrow (Co+ Co BM, grey bars, n=8) and Tie2-MyD88-/- mice transplanted with control bone marrow (Tie2MyD88-/- + Co BM, white dotted bars) or Tie2-MyD88-/- bone marrow (Tie2-MyD88-/- + Tie2-MyD88-/- BM, white bars) (A). Panel (B) show representative lung histology of Co+ Co BM mice, Tie2-MyD88-/- + Co BM mice and Tie2-MyD88-/- + Tie2-MyD88-/- BM, H&E staining, original magnification 20x. Neutrophil influx as reflected by Ly6-G and Ly6-C lung surface positivity (C) and whole lung MPO levels (D). Panel E shows representative images of Ly-6G and Ly-6C staining on lung slides Co+ Co BM mice, Tie2-MyD88-/- + Co BM mice and Tie2-MyD88-/- + Tie2-MyD88-/- BM. Data are expressed as box-andwhisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. * p < 0.05, ** p < 0.01, *** p < 0.001 compared to control mice determined with MannWhitney U test.

Chapter 4

Figure S4: Systemic inflammation is not affected by the absence of MyD88 expression in the endothelial compartment, but lung E-selectin is higher. Control and Tie2-MyD88-/- mice were irradiated and injected with control or Tie2-MyD88-/- bone marrow cells. Six weeks after transplantation, mice were infected with 6 x103 CFU K. pneumoniae and sacrificed 24 hours later. Homogenates were prepared from right lungs. E-selectin levels of (Co+ Co BM, grey bars, n=8) and Tie2-MyD88-/- mice transplanted with control bone marrow (Tie2-MyD88-/- + Co BM, white dotted bars) or Tie2-MyD88-/- bone marrow (Tie2-MyD88-/- + Tie2-MyD88-/- BM, white bars) are presented in pg/ml lung homogenate (A) or plasma (B). Plasma levels of LDH (C) and AST (D) after 24 hours. Data are expressed as box-andwhisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. *p < 0.05 vs control mice.

Hematopoietic but not Endothelial Cell MyD88 Contributes to Host Defense during Gram-negative Pneumonia Derived Sepsis Table S1: Inflammatory response in LysM-MyD88-/- and Tie2-MyD88-/- during K. pneumonia pulmonary tract infection.

Recipient

Tie2-MyD88-/-

Bonemarrow

Tie2-MyD88-/-

TNF-α

1122 (37)

1138 (58)

958 (55)

IL-1β

1912 (691)

3991 (1021)

2091 (435)

IL-6

3697 (1336)

5142 (1403)

6222 (694)

IL-10

47 (3)

30 (4)*

35 (5)

CXCL-1

4358 (1194)

7580 (2371)

11340 (2560)*

CXCL-2

18487 (2479)

26592 (5357)

22539 (1449)

TNF-α

10 (3)

102 (30)

136 (85)**

IL-6

814 (565)

2110 (1248)

3380 (1278)**

IL-10

IL-12

CCL-2

311 (149)

3085 (1793)

3771 (1061)**

Lung

Plasma

Control and Tie2-MyD88-/- mice were irradiated and injected with control or Tie2-MyD88-/- bone marrow cells. Six weeks after transplantation, mice were infected with 6 x103 CFU K. pneumoniae and sacrificed after 24 hours. Homogenates were prepared from right lungs. Cytokine and chemokine levels are presented in pg/ml lung homogenate or plasma. Data are mean (SE) of 5-8 mice per group. *p < 0.05, ** p < 0.01 vs control mice transplanted with control bone marrow.

Chapter 5 TIR-domain-containing adaptor-inducing interferon-β (TRIF) mediates antibacterial defense during gram-negative pneumonia by inducing Interferon-γ Journal of Innate Immunity, 2015 June 9, e-pub ahead of print Miriam H.P. van Lieshout 1,2 Sandrine Florquin 3 Cornelis van’t Veer 1,2 Alex F. de Vos 1,2 Tom van der Poll 1,2,4 Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands: 1 Center of Infection and Immunity Amsterdam 2 Center of Experimental and Molecular Medicine 3 Department of Pathology 4 Division of Infectious Diseases

Chapter 5

Abstract Klebsiella pneumoniae is an important cause of gram-negative pneumonia and sepsis. Mice deficient for TIR-domain-containing adaptor-inducing interferon-β (TRIF) demonstrate enhanced bacterial growth and dissemination during Klebsiella pneumonia. We here show that the impaired antibacterial defense of TRIF mutant mice is associated with absent interferon (IFN)-γ production in the lungs. IFN-γ production by splenocytes in response to K. pneumoniae in vitro was critically dependent on Toll-like receptor 4 (TLR4), the common TLR adapter myeloid differentiation primary response gene (MyD88) and TRIF. Reconstitution of TRIF mutant mice with recombinant IFN-γ via the airways reduced bacterial loads in lungs and distant body sites to levels measured in wild-type mice, and partially restored pulmonary cytokine levels. The IFN-γ induced improved enhanced antibacterial response in TRIF mutant mice occurred at the expense of increased hepatocellular injury. These data indicate that TRIF mediates antibacterial defense during gram-negative pneumonia at least in part by inducing IFN-γ at the primary site of infection.

TIR-domain-containing adaptor-inducing interferon-β (TRIF) mediates antibacterial defense during gram-negative pneumonia by inducing Interferon-γ

Introduction Globally, pneumonia is a common cause of morbidity and mortality and the most common cause of sepsis (1-3). The emerging antibiotic resistance among gram-negative pathogens, including Enterobacteriaceae such as Klebsiella (K.) pneumoniae, is an issue of major concern, since therapeutic options are limited and infections with these pathogens are associated with an unfavorable outcome (3, 4). K. pneumoniae is a common sepsis pathogen in humans, in particular in the context of lower respiratory tract infection (2). Pathogens entering the lower airways are detected by innate immune cells via pattern recognition receptors, among which the family of Toll-like receptors (TLRs) features prominently; this interaction initiates the early immune response (5). TLR signaling can proceed via two different routes that are dependent on myeloid differentiation primary response gene 88 (MyD88) and TIR-domain-containing adaptor-inducing interferon-β (TRIF) respectively (6). MyD88 is the universal adaptor for all TLRs except TLR3 and leads to NF-kB and MAP kinase activation and the induction of inflammatory cytokines. TRIF is the sole adaptor for TLR3 and in addition contributes to TLR4 signaling, leading to the activation of NF-kB and Interferon regulatory factor 3 (IRF3) and the induction of type I interferon (IFN) and inflammatory cytokine production (6). Notably, TLR4, that recognizes lipopolysaccharide (LPS), first activates the MyD88-dependent pathway before it initiates downstream signaling via the TRIF-dependent pathway once TLR4 complex is transported to the endosome for degradation (7). However, activation of both pathways is necessary for the induction of inflammatory cytokines via TLR4 (7). We previously reported about the crucial role of the TLR adaptors MyD88 and TRIF during K. pneumoniae infection and their differential contribution to the host response in different body compartments (8, 9). In these studies we noted that mice deficient for TRIF were incapable of IFN-γ production at the primary site of infection (unpublished data). IFN-γ is an important cytokine for innate and adaptive immunity that influences a wide array of immunologically relevant cellular programs, such as the enhancement of leukocyte attraction, up-regulation of pathogen recognition, antigen processing and presentation, and microbicidal effector cell functions (10). A previous report demonstrated the importance of IFN-γ for antibacterial defense and survival during K. pneumoniae by the use of IFN-γ gene deficient mice (11, 12). Moreover, IFN-γ deficient mice were more susceptible to airway infection with Legionella pneumophila and Burkholderia pseudomallei (13, 14), and therapeutic administration of recombinant IFN-γ was beneficial in several models of experimental respiratory tract infection (15, 16). Also, rIFN-γ demonstrated a beneficial effect in several human studies when used as an adjunctive therapy for opportunistic pathogens (17-21). We here report the impact of TRIF deficiency on pulmonary IFN-γ production during Klebsiella pneumonia. Furthermore, we explored to which extent the absence of local IFN-γ production during K. pneumoniae pneumonia in TRIF deficient mice contributes to their susceptible phenotype. We demonstrate that TRIF dependent signaling is crucial for IFN-γ production in vivo and in vitro and that reconstitution 87

Chapter 5

of IFN-γ levels in the airways improves antibacterial defense in TRIF deficient but not in wild-type (WT) mice.

Materials and methods Animals TRIF mutant mice, generated on a C57Bl/6 genetic background (22), were provided by Dr B. Beutler (Center for the Genetics of Host Defense, University of Texas Southwestern Medical Center, Texas). MyD88 deficient (Myd88-/-) (23) and Tlr4−/− mice (24) were provided by Dr. S. Akira (Research Institute for Microbial Diseases, Osaka, Japan) and backcrossed > 8 times to a C57Bl/6 genetic background. All gene deficient mice were bred in the animal facility of the Academic Medical Center (Amsterdam, the Netherlands). Age- and sex matched WT C57Bl/6 control mice were obtained from Harlan Nederland (Horst, the Netherlands). Mice were infected at 10-12 weeks of age. The Animal Care and Use Committee of the University of Amsterdam approved all experiments. Induction of pneumonia and sampling of organs Pneumonia was induced by intranasal inoculation with ~ 1 x 104 colony forming units (CFU) of K. pneumoniae serotype 2 (ATCC 43816; American Type Culture Collection, Manassas, VA) (8,9). Mice were sacrificed at the indicated time points after infection and organs were harvested and processed exactly as described (8, 25). In the reconstitution experiment, mice were administered 50 ng of recombinant mouse IFN-γ (rIFN-γ) (R&D systems, Abbington, United Kingdom) or vehicle (0.1% human serum albumin in sterile saline) intranasally 30 minutes before and 24 hours after inoculation; mice were euthanized after 48 hours of infection. Quantitative RT-PCR RNA was isolated from lung homogenates using the Nucleospin RNA II kit (Machery-Nagel, Duren, Germany). Total RNA was reverse transcribed using oligo (dT) primer and Moloney murine leukemia virus reverse transcriptase (Invitrogen, Breda, The Netherlands). Quantitative PCR of Ifng gene product was performed as described (26). Data was analyzed using the LinRegPCR program. Results were normalized to β2m transcript. In vitro studies Splenocytes were obtained, seeded at a density of 500.000 cells per well and cultured exactly as decribed (27). Cells were stimulated for 48 hours in at least quadruplicate with the indicated concentrations of mitomycin C-treated (0.05 mg/ml) (Sigma-Aldrich) growth-arrested K. pneumoniae diluted in RPMI medium without antibiotics, LPS derived from Klebsiella pneumoniae (Sigma) (100 ng/ml) or ultrapure Escherichia coli O111 B4 LPS (Invivogen) (100 ng/ml) diluted in RPMI medium with antibiotics in a final volume of 200 microliter. Supernatants were stored and analyzed for cytokine concentrations by ELISA.

TIR-domain-containing adaptor-inducing interferon-β (TRIF) mediates antibacterial defense during gram-negative pneumonia by inducing Interferon-γ

Assays IFN-γ levels in cell supernatants and lung levels of IL-1β, CXCL1, CXCL2 and CCL2 were measured by ELISA (R&D Systems, Minneapolis, MN and Invitrogen, Breda, the Netherlands). Lung levels of IFN-γ, TNF-α, IL-6 and IL-10 were measured by using a cytometric bead array multiplex assay (BD Biosciences). Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were measured using kits from Sigma and a Hittachi analyzer (Boehringer Mannheim). Histopathology Histologic examination of lungs and liver was performed exactly as described (25, 28). Granulocyte immunohistochemic stainings were prepared using a FITClabeled anti-mouse Ly6-C/G mAb (BD Biosciences, San Jose, CA) exactly as described before (27). Statistical analysis Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation (in vivo experiments) or as means ± standard error of the mean (tables, cell stimulation experiments); Bacterial loads are expressed as scatter plots, each symbol representing an individual mouse, with horizontal lines indicating medians. For experiments with 2 groups, the Mann–Whitney U test was used to determine statistical significance. For experiments with > 2 groups, the Kruskall-Wallis test was used, followed by Mann–Whitney U tests to compare individual genetically modified groups with the WT or TRIF mutant control group when appropriate. Fisher’s exact test was used to determine if the proportion of positive test results was different. These analyses were done using GraphPad Prism (San Diego, CA). p < 0.05 was considered statistically significant.

Results IFN-γ production is impaired in TRIF mutant mice during Klebsiella pneumonia In our previous studies on the role of TRIF during K. pneumoniae airway infection we demonstrated that TRIF mutant mice have an impaired antibacterial defense as illustrated by significantly higher bacterial loads in lungs, blood and spleen (8); in these investigations we also observed higher bacterial loads in livers of TRIF mutant mice and TRIF bone marrow chimeras lacking TRIF in hematopoietic cells (supplementary Figure 1A,B). We noticed in a multiplex cytokine assay performed on whole lung homogenates that IFN-γ levels remained undetectable in TRIF mutant mice throughout the infection (<5 pg/ml), while in WT mice lung IFN-γ concentrations increased after Klebsiella inoculation, peaking after 24 hours (p < 0.05 to 0.001 for the difference between groups, figure 1A). TRIF mutant mice also showed strongly reduced IFN-γ mRNA expression in lungs during Klebsiella pneumonia (p < 0.01 versus WT mice, figure 1B).

Chapter 5

Figure 1: TRIF mediates IFN-γ production during K. pneumoniae airway infection. WT and TRIF mutant mice (n=7-8 per group) were infected with ~1 x 104 CFU K. pneumoniae and sacrificed at designated time points. IFN-γ levels in lungs of mice were determined by cytometric bead assay (A) and quantitative real-time RT-PCR (qRT-PCR) (B). Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation (A,B) or as mean (SE) (C). p < 0.05, ** p < 0.01 determined with Mann–Whitney U test. ### p < 0.001 determined with Fisher’s exact test.

IFN-γ production in response to Klebsiella is TLR4 dependent via both Myd88 and TRIF Next, we stimulated splenocytes, as a source of IFN-γ producing cells, with growtharrested K. pneumoniae in vitro. In a pilot-study, we observed significantly impaired IFN-γ secretion by TRIF mutant cells stimulated with either 2 x 105 or 2 x 106 bacteria (data not shown). We repeated this experiment, this time including splenocytes of Tlr4-/- and Myd88-/- mice in addition to splenocytes from TRIF mutant and WT mice. IFN-γ production in response to growth-arrested K. pneumoniae was most severely impaired in Myd88-/- cells, followed by Tlr4-/- and then TRIF mutant cells (figure 2A, p < 0.05 to 0.01 compared to WT cells). In addition, we stimulated cells with LPS derived from K. pneumoniae or ultra-purified LPS derived from E. coli, and found virtually absent IFN-γ release by Tlr4-/-, Myd88-/- and TRIF mutant cells (figure 2B, p < 0.01 versus WT cells). Antibacterial defense of TRIF mutant mice can be restored by local treatment with IFN-γ To test if the strongly reduced pulmonary IFN-γ levels contribute functionally to the impaired antibacterial defense of TRIF mutant mice we treated WT and TRIF mutant mice with IFN-γ intranasally 30 minutes before and 24 hours after infection with Klebsiella; we used 48 hours of infection as pre-defined endpoint since this was the time point at which the enhanced growth of Klebsiella in TRIF mutant relative to WT mice was most clear (8). While TRIF mutant mice treated with vehicle displayed undetectable pulmonary IFN-γ concentrations, confirming the results presented in figure 1A, TRIF mutant mice administered with rIFN-γ had lung IFN-γ levels that were similar to those measured in WT mice (figure 3A); WT mice that received rIFN-γ had significantly higher lung levels than WT mice treated with vehicle (p < 90

TIR-domain-containing adaptor-inducing interferon-β (TRIF) mediates antibacterial defense during gram-negative pneumonia by inducing Interferon-γ

Figure 2: IFN-γ secretion by splenocytes is dependent on TLR4, MyD88 and TRIF. Splenocytes derived from WT, Tlr4-/-, Myd88-/- and TRIF mutant mice were stimulated with different concentrations growth arrested K. pneumoniae, and LPS derived from E. coli or K. pneumoniae (n=4-6 for each condition), and IFN-γ levels were determined after 48 hours. Data are expressed as mean (SE). * p < 0.05, ** p < 0.01 determined with Mann–Whitney U test (performed as post hoc following Kruskal-Wallis test).

Chapter 5

0.05). We reproduced the previously described phenotype in TRIF mutant mice (8), showing 100-1000 fold higher bacterial loads in their lungs relative to WT mice, together with increased bacterial dissemination to blood and spleen (figure 3B-D, p < 0.001). Importantly, we observed a spectacular improvement of antibacterial defense in rIFN-γ treated TRIF mutant mice compared to vehicle treated TRIF mutant mice (p < 0.01 to 0.001), as reflected by bacterial loads similar to WT mice in all organs. Of note, we observed no effect on bacterial burdens in WT mice treated with rIFN-γ compared to vehicle treated WT mice (figure 3B-D).

Figure 3: Administration of rIFN-γ via the airways restores antibacterial defense in TRIF mutant mice. WT and TRIF mutant mice were infected with ~1 x 104 CFU K. pneumonia; 50 ng recombinant IFN-γ or vehicle was administered intranasally 30 minutes before infection and 24 hours thereafter (n=8 mice each group). Mice were sacrificed after 48 hours of infection. IFN-γ levels in lung homogenates 48 hours after infection (A) depicted as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. Bacterial loads in lung (B), blood (C) and spleen (D) 48 hours after infection. Each symbol represents an individual mouse, horizontal lines represent medians. ** p < 0.01, *** p < 0.001 vs WT mice treated with vehicle, ## p < 0.01, ### p < 0.001 vs TRIF mutant mice treated with vehicle determined with Mann-Whitney U test, Fisher’s exact test was used in panel 1A for comparison between TRIF mutant groups (performed as post hoc after following Kruskal-Wallis test).

TIR-domain-containing adaptor-inducing interferon-β (TRIF) mediates antibacterial defense during gram-negative pneumonia by inducing Interferon-γ

Impact of IFN-γ treatment on the inflammatory response to pneumonia To obtain insight in the extent of local inflammation at the primary site of infection in TRIF mutant and WT mice, and the effect of rIFN-γ treatment hereon, we semiquantitatively scored lung histopathology of tissue samples harvested 48 hours after infection, focusing on key histological features characteristic for severe pneumonia. While total lung histopathology scores were not different between groups (table 1), rIFN-γ treated TRIF mutant and WT mice had more signs of bronchitis and less signs of pleuritis when compared to their respective vehicle treated controls (table 1, p < 0.05 compared to the respective controls and figure 4A-F). Neutrophil recruitment to the lungs, measured as the percentage of Ly-6+ positive lung cell surface, was significantly higher in vehicle treated TRIF mutant mice compared to vehicle treated WT mice at this late stage of infection (table 1, p < 0.01). Administration of rIFN-γ reduced total neutrophil numbers in lung tissue of TRIF mutant mice similar to those measured in WT mice (p < 0.05 compared to vehicle treated TRIF mutant mice); rIFN-γ treatment did not influence lung neutrophil counts in WT mice (table 1). Moreover, when the Ly-6 stainings were studied in detail, the number of intrabronchial neutrophils appeared to be larger after rIFN-γ treatment (Figure 4E-H). We next determined the effect of rIFN-γ treatment on the induction of proinflammatory cytokines (TNF-α, IL-1β, IL-6), the anti-inflammatory cytokine IL10 and chemokines CXCL1, CXCL2 and CCL2 in whole lung homogenates. TRIF mutant mice demonstrated reduced levels of TNF-α, IL-1β, CXCL1, CXCL2 and CCL2 relative to WT mice (table 2, p < 0.05 to 0.001). Treatment of TRIF mutant mice with rIFN-γ partially restored the inflammatory profile with the exception of IL-1β: TNF-α and CXCL2 were not significantly different from vehicle treated WT mice, levels of CXCL1 and CCL2 were still significantly lower although differences were smaller (p < 0.05 to 0.01 compared to vehicle treated WT mice). The change in levels of inflammatory cytokines and chemokines after treatment with rIFN-γ of TRIF mutant mice was significant for TNF-α, IL-6, CXCL2 and CCL2 compared to vehicle treated TRIF mutant mice (table 2, p < 0.05 to 0.001 between groups). IFN-γ deficiency protects TRIF mutant mice from liver injury Klebsiella induced pneumonia derived sepsis is associated with hepatocellular injury, as reflected by increased plasma concentrations of AST and ALT (8, 29). TRIF mutant mice had lower AST and ALT plasma levels 48 hours after infection when compared with WT mice (p < 0.01, figure 5A,B) as well as fewer signs of liver inflammation as determined by liver histopathology scores (p < 0.01, figure 5C, Supplemental figure 2). Remarkably, rIFN-γ treatment significantly increased AST and ALT levels in TRIF mutant mice compared to vehicle treated TRIF mutant mice (p < 0.01 to 0.001) to levels similar to those measured in WT mice. In WT mice, rIFN-γ treatment reduced transaminase levels, significantly so for AST (p < 0.05, figure 5A).

Chapter 5 Table 1: Histological scores

Mice

TRIF mutant

Treatment

vehicle

rIFN-ɣ

vehicle

rIFN-ɣ

Total pathology score lung

14.5 (0.6)

13.8 (0.8)

14.5 (1.2)

13.1 (0.6)

Pneumonia % of lung surface

15 (3)

6 (4)

22 (6)

7 (3)

Interstitial inflammation

3.1 (0.1)

3.0 (0.5)

2.8 (0.7)

2.4 (0.3)

Oedema

2.8 (0.2)

2.5 (0.3)

3.4 (0.5)

3.0 (0.2)

Endothelialitis

2.5 (0.2)

2.9 (0.1)

3 (0.2)

2.6 (0.2)

Bronchitis

2.9 (0.1)

3.5 (0.2)*

2.6 (0.3)

3.8 (0.2)#

Pleuritis

1.8 (0.3)

1.3 (0.2)*

1.5 (0.3)

0.8 (0.2)#

Ly6+ % of total lung surface

2.3 (0.4)

2.1 (0.5)

8.6 (1.2)**

3.9 (0.8)#

WT and TRIF mutant mice were infected with 1x104 CFU K. pneumoniae and 50 ng recombinant IFNγ was administered intranasally upon infection and after 48 hours. Histological scores determined 48 hours after infection. Total pathology score is the sum of the histological subscores determined as decribed in the methods. Data are mean (SE) of 7–8 mice per group. * p < 0.05, ** p < 0.01 compared to vehicle treated WT mice. # p < 0.05 rIFN-γ treated TRIF mutant mice compared to vehicle treated TRIF mutant mice. Table 2: Inflammatory response

Mice

TRIF mutant

Treatment

vehicle

rIFN-ɣ

vehicle

rIFN-ɣ

TNF-α

892 (245)

760 (57)

191 (48)**

501 (115)#

IL-1β

7434 (642)

4950 (753)*

4168 (731)**

4525 (557)**

IL-6

1914 (451)

2030 (624)

2446 (398)

1507 (216)#

IL-10

14 (2)

11 (1)

14 (2)

CXCL1

12586 (1899)

9453 (1645)

3625 (871)**

4255 (828)*

CXCL2

20553 (6546)

38048 (7157)

6432 (1532)*

28943 (5785)###

CCL2

4619 (541)

3718 (366)

1841 (210)***

2126 (240)**#

WT and TRIF mutant mice were infected with 1x104 CFU K. pneumoniae and 50 ng recombinant IFNγ was administered intranasally upon infection and after 48 hours. Homogenates were prepared from right lungs. Cytokine and chemokine levels are presented in pg/ml lung homogenate. Data are mean (SE) of 7–8 mice per group. Bd= below detection.* p < 0.05, ** p < 0.01, *** p < 0.001 compared to vehicle treated WT mice. # p < 0.05. ## p < 0.01, ### p < 0.001 rIFN-γ treated TRIF mutant mice compared to vehicle treated TRIF mutant mice.

TIR-domain-containing adaptor-inducing interferon-β (TRIF) mediates antibacterial defense during gram-negative pneumonia by inducing Interferon-γ Figure 4: Effect of IFN-γ treatment on lung pathology. WT and TRIF mutant mice were infected with ~ 1x 104 CFU K. pneumonia; 50 ng recombinant IFN-γ or vehicle was administered intranasally 30 minutes before infection and 24 hours thereafter. Mice were sacrificed after 48 hours of infection. Representative lung histology (H&E staining) of WT mice treated with vehicle (A), WT mice treated with rIFN-γ (B), TRIF mutant mice treated with vehicle (C) and TRIF mutant mice treated with rIFN-γ (D). In each upper panel arrows indicate signs of bronchitis (original magnification 10x) and in the lower panel asterisks indicate pleuritis (original magnification 20x). Representative lung histology (Ly6 staining, indicating neutrophils) of lungs of WT mice treated with vehicle (E), WT mice treated with rIFN-γ (F), TRIF mutant mice treated with vehicle (G) and TRIF mutant mice treated with rIFN-γ (H), original magnification 10x.

Chapter 5

Figure 5: TRIF mutant mice have attenuated liver injury that increases after rIFN-γ treatment. WT and TRIF mutant mice were infected with ~ 1 x 104 CFU K. pneumonia; 50 ng rIFN-γ or vehicle was administered intranasally 30 minutes before infection and 24 hours thereafter. Mice were sacrificed after 48 hours of infection. AST (A) and ALT (B) plasma levels and liver histopathology scored as decribed in the methods (C) expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. * p < 0.05, ** p < 0.01, *** p < 0.001 determined with Mann-Whitney U (performed as post hoc following Kruskal-Wallis test).

Discussion K. pneumoniae is a clinically important gram-negative bacterium in pneumonia and one of the pathogens that causes major concern because of increasing antimicrobial resistance rates, limiting therapeutic options (2-4, 30). Previous research has documented the importance of TLR signaling for host defense during K. pneumoniae pneumonia, notably of TLR4, TLR2 and TLR9 (25, 31, 32), and we and others previously described the pivotal role for the TLR-adapters MyD88 and TRIF herein (8, 33). Given our discovery that in the absence of TRIF lung levels of IFN-γ were undetectable during the course of K. pneumoniae airway infection we here explored the functional importance thereof. Our main findings were that indeed TRIF is crucial for IFN-γ production in response to K. pneumonia, together with TLR4 and MyD88, and that reconstitution of TRIF mutant mice with rIFN-γ improves antibacterial defense to the level of WT mice, but at the expense of enhanced liver injury. Earlier, we and others described the susceptible phenotype of TRIF deficient mice in Klebsiella pneumonia, marked by a clearly impaired antibacterial defense with a 100-1000 fold increase in bacterial loads 48 hours after infection, a finding that we reproduced in the current report (8, 25, 33). The early inflammatory response of mice partially or fully deficient for TRIF is characterized by impaired neutrophil 96

TIR-domain-containing adaptor-inducing interferon-β (TRIF) mediates antibacterial defense during gram-negative pneumonia by inducing Interferon-γ

influx probably as a result of impaired CXCL1 secretion and lower levels of TNF-α and IL-6. However, during the course of the infection and in response to higher bacterial loads all of these cytokines gradually increased in spite of (partial) TRIF deficiency (8). Notably, in the current study CXCL1, CXCL2 and TNF-α levels were still reduced in TRIF mutant mice 48 hours post infection, while lung neutrophil numbers as determined by immunohistochemistry were significantly higher. This is probably due to the very high bacterial numbers present in TRIF mutant mice at this moment, leading to tissue injury and neutrophil attraction via mechanisms other than provided by the chemoattractant gradient by the afore mentioned mediators. Remarkably, however, IFN-γ levels remained virtually undetectable in TRIF mutant mice throughout, which formed the rationale for the current study. We hypothesized that deficient IFN-γ production could at least in part be responsible for the impaired antibacterial defense of TRIF mutant mice, considering that IFN-γ is a powerful pleiotropic cytokine that during bacterial infection can enhance leukocyte attraction, pathogen recognition, antigen processing and presentation, and microbicidal effector cell functions (10). We extended our in vivo observation of decreased IFN-γ levels in TRIF mutant mice by demonstrating that also under controlled conditions with equal amounts of growth-arrested bacteria the capacity of TRIF mutant splenocytes to secrete IFN-γ is impaired. Moreover, IFN-γ production was critically dependent on MyD88 and TLR4. This is not surprising, since it is well known that these innate immune sensors are highly important for the induction of the inflammatory response to K. pneumoniae and the phenotype of Myd88-/- and Tlr4-/- mice is more severe than that of TRIF mutant mice during in vivo infection (8, 25). However, the role of these receptors specifically in the induction IFN-γ in response to pathogens is less well known. In accordance with the present report, TRIF deficient mice were demonstrated to produce lower IFN-γ levels during Aspergillus airway infection in vivo (34). In the current study, our results suggest that TLR2 dependent signals play a role in response to K. pneumoniae in addition to TRIF, MyD88 and TLR4, since IFN-γ levels secreted by TRIF mutant and Tlr4-/cells gradually increased with increasing bacterial concentrations, which is in line with the role of TLR2 during infection with Klebsiella in vivo (25). We observed a spectacular effect on bacterial loads after reconstitution of TRIF mutant mice with rIFN-γ, which coincided with a partial recovery of the inflammatory cytokine profile. The importance of IFN-γ during K. pneumoniae infection was demonstrated before since Ifn-γ-/- mice displayed an impaired antibacterial defense and increased mortality (11, 12, 25). The other way around, in a rat model of ethanol intoxication followed by Klebsiella airway infection, adenoviral expression of IFN-γ improved antibacterial defense (35). Likewise, conditional adenoviral expression of IFN-γ improved clearance of Klebsiella from the lungs in mice (36). Strikingly, in our study there was no effect of rIFN-γ on bacterial loads in WT mice, suggesting that local rIFN-γ administration is only beneficial when it compensates for a clearly deficient production. Also, in WT mice rIFN-γ treatment, with the exception of IL-1β, did not affect lung cytokine concentrations, while in TRIF mutant mice it increased the levels of TNF-α, IL-6, CXCL1, CXCL2 and CCL2. The mechanism by which rIFN-γ improves bacterial defense in TRIF mutant mice might be by enhancing the 97

Chapter 5

bacterial killing capacity of alveolar macrophages (37). Unfortunately, the Klebsiella strain used here cannot be killed by macrophages or neutrophils in vitro (our own observations), illustrating its high virulence and precluding further in vitro analyses. Improved monocyte and macrophage function was also presumed to play a role in human clinical trials wherein treatment with rIFN-γ demonstrated beneficial effects in Mycobacterium (M.) tuberculosis and M. avium infections, Leishmaniasis and fungal sepsis, although the exact mechanisms are currently unknown (17-21). Recently, however, it was demonstrated in fungal sepsis patients that the ex vivo cytokine response was enhanced in patients treated with rIFN-γ (17). In our study, TRIF mutant mice treated with rIFN-γ also had higher plasma IFN-γ levels when compared with TRIF mutant mice treated with vehicle, even though rIFN-γ was instilled locally in the airways. Hence, although it is likely that the reduced bacterial loads at distant body sites in rIFN-γ treated TRIF mutant mice at least in part are the consequence of lower bacterial burdens at the primary site of infection, we cannot exclude an additional systemic effect of local rIFN-γ treatment. Another aspect of the inflammatory response that we observed in our study is that while total lung histopathology scores were not different between groups, rIFN-γ treated Trif-/- and WT mice had more signs of bronchitis and lower scores on pleuritis when compared to their respective vehicle treated controls, possibly indicating a redistribution in the pattern of inflammatory cell migration. This might be secondary to a higher intrabronchial rIFN-γ concentration after intranasal administration, resulting in increased attraction of inflammatory cells to the intrabronchial and intraalveolar compartment (see also figure 4). Possibly, this contributed to a better containment of the infection. In this and in our previous study we demonstrated significantly lower levels of AST and ALT in mice (partially) deficient for TRIF, despite higher levels of bacterial loads in the blood and liver (8). Although liver bacterial loads were not determined in rIFN-γ treated mice, it is unlikely that the increased hepatocellular injury in these animals was caused by higher bacterial burdens in livers considering the reduced Klebsiella numbers in blood and spleen. This illustrates the double edged sword character of the innate immune response that is on the one hand essential for early antibacterial defense but on the other hand contributes to collateral tissue damage in sepsis as was illustrated in previous work (38-40). Strikingly, the reconstitution of TRIF mutant mice with rIFN-γ deteriorated liver injury. This suggests that IFN-γ is involved in inflammation driven liver injury, as was proposed before in an intravenous model of K. pneumoniae sepsis in Ifn-γ-/- mice (12). However, the lower AST levels in WT mice treated with rIFN-γ are more difficult to explain and require further investigation. Possibly, the increased plasma levels of the anti-inflammatory cytokine IL-10 in rIFN-γ treated WT mice (albeit not significant) played a role herein. In conclusion, we demonstrate a crucial role for TRIF in IFN-γ production during K. pneumoniae pneumonia. TRIF mediated IFN-γ release is essential for an adequate innate immune response as reflected by the fact that the strongly impaired antibacterial defense of TRIF mutant mice can be restored by reconstitution of IFN-γ levels in the lungs by local treatment. These data provide new insight into 98

TIR-domain-containing adaptor-inducing interferon-Î˛ (TRIF) mediates antibacterial defense during gram-negative pneumonia by inducing Interferon-Îł

how TRIF mediates protective immunity during gram-negative infection.

Acknowledgments We thank Regina de Beer, Joost Daalhuisen and Marieke ten Brink for expert technical assistance. This work was supported by the AMC Graduate School of Medical Science (to M. H. P. v. L).

Chapter 5

References

1. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR. Epidemiology of Severe Sepsis in the United States: Analysis of Incidence, Outcome, and Associated Costs of Care. Crit Care Med 2001;29:1303-1310.

3. World Health Organisation. World Health Observatory, Top Ten Causes of Death 2012.; 2012.

4. Schwaber MJ, Carmeli Y. Mortality and Delay in Effective Therapy Associated With ExtendedSpectrum Beta-Lactamase Production in Enterobacteriaceae Bacteraemia: a Systematic Review and Meta-Analysis. J Antimicrob Chemother 2007;60:913-920.

5. Mizgerd JP. Acute Lower Respiratory Tract Infection. N Engl J Med 2008;358:716-727.

6. Kawai T, Akira S. Toll-Like Receptors and Their Crosstalk With Other Innate Receptors in Infection and Immunity. Immunity 2011;34:637-650.

8. van Lieshout MH, Blok DC, Wieland CW, de Vos AF, van â&#x20AC;&#x2DC;t Veer C, van der Poll T. Differential Roles of MyD88 and TRIF in Hematopoietic and Resident Cells During Murine Gram-Negative Pneumonia. J Infect Dis 2012;206:1415-1423.

9. van Lieshout MH, Anas AA, Florquin S, Hou B, Vanâ&#x20AC;&#x2122;t Veer C, de Vos AF, van der Poll T. Hematopoietic but Not Endothelial Cell MyD88 Contributes to Host Defense During GramNegative Pneumonia Derived Sepsis. PLoS Pathog 2014;10:e1004368.

10. Schroder K, Hertzog PJ, Ravasi T, Hume DA. Interferon-Gamma: an Overview of Signals, Mechanisms and Functions. J Leukoc Biol 2004;75:163-189.

11. Yoshida K, Matsumoto T, Tateda K, Uchida K, Tsujimoto S, Iwakurai Y, Yamaguchi K. Protection Against Pulmonary Infection With Klebsiella Pneumoniae in Mice by Interferon-Gamma Through Activation of Phagocytic Cells and Stimulation of Production of Other Cytokines. J Med Microbiol 2001;50:959-964.

12. Moore TA, Perry ML, Getsoian AG, Newstead MW, Standiford TJ. Divergent Role of Gamma Interferon in a Murine Model of Pulmonary Versus Systemic Klebsiella Pneumoniae Infection. Infect Immun 2002;70:6310-6318.

13. Shinozawa Y, Matsumoto T, Uchida K, Tsujimoto S, Iwakura Y, Yamaguchi K. Role of Interferon-Gamma in Inflammatory Responses in Murine Respiratory Infection With Legionella Pneumophila. J Med Microbiol 2002;51:225-230.

14. Easton A, Haque A, Chu K, Lukaszewski R, Bancroft GJ. A Critical Role for Neutrophils in Resistance to Experimental Infection With Burkholderia Pseudomallei. J Infect Dis 2007;195:99-107.

100

The Role of Pattern-Recognition Receptors in Innate Immunity: Update on Toll-Like Receptors. Nat Immunol 2010;11:373-384.

TIR-domain-containing adaptor-inducing interferon-β (TRIF) mediates antibacterial defense during gram-negative pneumonia by inducing Interferon-γ

15. Beck JM, Liggitt HD, Brunette EN, Fuchs HJ, Shellito JE, Debs RJ. Reduction in Intensity of Pneumocystis Carinii Pneumonia in Mice by Aerosol Administration of Gamma Interferon. Infect Immun 1991;59:3859-3862.

16. Buccheri S, Reljic R, Caccamo N, Meraviglia S, Ivanyi J, Salerno A, Dieli F. Prevention of the Post-Chemotherapy Relapse of Tuberculous Infection by Combined Immunotherapy. Tuberculosis (Edinb ) 2009;89:91-94.

17. Delsing CE, Gresnigt MS, Leentjens J, Preijers F, Frager FA, Kox M, Monneret G, Venet F, Bleeker-Rovers CP, van de Veerdonk FL, et al. Interferon-Gamma As Adjunctive Immunotherapy for Invasive Fungal Infections: a Case Series. BMC Infect Dis 2014;14:166.

18. Gao XF, Yang ZW, Li J. Adjunctive Therapy With Interferon-Gamma for the Treatment of Pulmonary Tuberculosis: a Systematic Review. Int J Infect Dis 2011;15:e594-e600.

19. Milanes-Virelles MT, Garcia-Garcia I, Santos-Herrera Y, Valdes-Quintana M, ValenzuelaSilva CM, Jimenez-Madrigal G, Ramos-Gomez TI, Bello-Rivero I, Fernandez-Olivera N, Sanchez-de la Osa RB, et al. Adjuvant Interferon Gamma in Patients With Pulmonary Atypical Mycobacteriosis: a Randomized, Double-Blind, Placebo-Controlled Study. BMC Infect Dis 2008;8:17.

20. Murray HW. Interferon-Gamma in Infection and Immunoparalysis. Intensive Care Med 1996;22 Suppl 4:S455.

21. Wunderink RG. Adjunctive Therapy in Community-Acquired Pneumonia. Semin Respir Crit Care Med 2009;30:146-153.

22. Hoebe K, Du X, Georgel P, Janssen E, Tabeta K, Kim SO, Goode J, Lin P, Mann N, Mudd S, et al. Identification of Lps2 As a Key Transducer of MyD88-Independent TIR Signalling. Nature 2003;424:743-748.

23. Adachi O, Kawai T, Takeda K, Matsumoto M, Tsutsui H, Sakagami M, Nakanishi K, Akira S. Targeted Disruption of the MyD88 Gene Results in Loss of IL-1- and IL-18-Mediated Function. Immunity 1998;9:143-150.

24. Hoshino K, Takeuchi O, Kawai T, Sanjo H, Ogawa T, Takeda Y, Takeda K, Akira S. Cutting Edge: Toll-Like Receptor 4 (TLR4)-Deficient Mice Are Hyporesponsive to Lipopolysaccharide: Evidence for TLR4 As the Lps Gene Product. J Immunol 1999;162:3749-3752.

25. Wieland CW, van Lieshout MH, Hoogendijk AJ, van der Poll T. Host Defence During Klebsiella Pneumonia Relies on Haematopoietic-Expressed Toll-Like Receptors 4 and 2. Eur Respir J 2011;37:848-857.

26. van ‘t Veer C, van den Pangaart PS, Kruijswijk D, Florquin S, de Vos AF, van der Poll T. Delineation of the Role of Toll-Like Receptor Signaling During Peritonitis by a Gradually Growing Pathogenic Escherichia Coli. J Biol Chem 2011;286:36603-36618.

27. van Lieshout MH, Anas AA, Florquin S, Hou B, Van’t Veer C, de Vos AF, van der Poll T. Hematopoietic but Not Endothelial Cell MyD88 Contributes to Host Defense During GramNegative Pneumonia Derived Sepsis. PLoS Pathog 2014;10:e1004368.

28. Achouiti A, Vogl T, Urban CF, Rohm M, Hommes TJ, van Zoelen MA, Florquin S, Roth J, Van’t Veer C., de Vos AF, et al. Myeloid-Related Protein-14 Contributes to Protective Immunity in Gram-Negative Pneumonia Derived Sepsis. PLoS Pathog 2012;8:e1002987.

101

Chapter 5

29. Renckens R, Roelofs JJ, Bonta PI, Florquin S, de Vries CJ, Levi M, Carmeliet P, Vanâ&#x20AC;&#x2122;t Veer C., van der Poll T. Plasminogen Activator Inhibitor Type 1 Is Protective During Severe GramNegative Pneumonia. Blood 2007;109:1593-1601.

30. Vardakas KZ, Tansarli GS, Rafailidis PI, Falagas ME. Carbapenems Versus Alternative Antibiotics for the Treatment of Bacteraemia Due to Enterobacteriaceae Producing ExtendedSpectrum Beta-Lactamases: a Systematic Review and Meta-Analysis. J Antimicrob Chemother 2012;67:2793-2803.

31. Bhan U, Lukacs NW, Osterholzer JJ, Newstead MW, Zeng X, Moore TA, McMillan TR, Krieg AM, Akira S, Standiford TJ. TLR9 Is Required for Protective Innate Immunity in Gram-Negative Bacterial Pneumonia: Role of Dendritic Cells. J Immunol 2007;179:3937-3946.

32. Branger J, Knapp S, Weijer S, Leemans JC, Pater JM, Speelman P, Florquin S, van der Poll T. Role of Toll-Like Receptor 4 in Gram-Positive and Gram-Negative Pneumonia in Mice. Infect Immun 2004;72:788-794.

33. Cai S, Batra S, Shen L, Wakamatsu N, Jeyaseelan S. Both TRIF- and MyD88-Dependent Signaling Contribute to Host Defense Against Pulmonary Klebsiella Infection. J Immunol 2009;183:6629-6638.

34. de Luca A., Bozza S, Zelante T, Zagarella S, Dâ&#x20AC;&#x2122;Angelo C, Perruccio K, Vacca C, Carvalho A, Cunha C, Aversa F, et al. Non-Hematopoietic Cells Contribute to Protective Tolerance to Aspergillus Fumigatus Via a TRIF Pathway Converging on IDO. Cell Mol Immunol 2010;7:459470.

35. Kolls JK, Lei D, Stoltz D, Zhang P, Schwarzenberger PO, Ye P, Bagby G, Summer WR, Shellito JE, Nelson S. Adenoviral-Mediated Interferon-Gamma Gene Therapy Augments Pulmonary Host Defense of Ethanol-Treated Rats. Alcohol Clin Exp Res 1998;22:157-162.

36. Ruan S, Young E, Luce MJ, Reiser J, Kolls JK, Shellito JE. Conditional Expression of Interferon-Gamma to Enhance Host Responses to Pulmonary Bacterial Infection. Pulm Pharmacol Ther 2006;19:251-257.

37. Serezani CH, Chung J, Ballinger MN, Moore BB, Aronoff DM, Peters-Golden M. Prostaglandin E2 Suppresses Bacterial Killing in Alveolar Macrophages by Inhibiting NADPH Oxidase. Am J Respir Cell Mol Biol 2007;37:562-570.

38. Hochhauser E, Avlas O, Fallach R, Bachmetov L, Zemel R, Pappo O, Shainberg A, Ben AZ. Bone Marrow and Nonbone Marrow Toll Like Receptor 4 Regulate Acute Hepatic Injury Induced by Endotoxemia. PLoS ONE 2013;8:e73041.

39. Kitazawa T, Tsujimoto T, Kawaratani H, Fukui H. Salvage Effect of E5564, Toll-Like Receptor 4 Antagonist on D-Galactosamine and Lipopolysaccharide-Induced Acute Liver Failure in Rats. J Gastroenterol Hepatol 2010;25:1009-1012.

40. Moore TA, Lau HY, Cogen AL, Monteleon CL, Standiford TJ. Anti-Tumor Necrosis FactorAlpha Therapy During Murine Klebsiella Pneumoniae Bacteremia: Increased Mortality in the Absence of Liver Injury. Shock 2003;20:309-315.

102

Chapter 6 Lung epithelial MyD88 drives early pulmonary clearance of Pseudomonas aeruginosa by a flagellin dependent mechanism Submitted

Adam A. Anas 1,2 * Miriam H.P. van Lieshout 1,2 * Theodora A.M. Claushuis 1,2 Alex F. de Vos 1,2 Sandrine Florquin 3 Onno J. de Boer 3 Baidong Hou 4 Cornelis van â&#x20AC;&#x2122;t Veer 1,2 Tom van der Poll 1,2,5 *These authors contributed equally. Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands: 1 Center of Infection and Immunity Amsterdam 2 Center of Experimental and Molecular Medicine 3 Department of Pathology 5 Division of Infectious Diseases Institute of Biophysics, Chaoyang District, Beijing, China: Key Laboratory of Infection and Immunity

Chapter 6

Abstract Pseudomonas (P.) aeruginosa is a flagellated pathogen frequently causing pneumonia in hospitalized patients and sufferers of chronic lung disease. Tolllike receptors (TLRs) comprise a family of pattern recognition receptors crucial for induction of innate immunity, including during Pseudomonas infections of the lower respiratory tract. Here we investigated the role of the common TLR adaptor myeloid-differentiation factor (MyD)88 in myeloid versus lung epithelial cells in clearance of P. aeruginosa from the airways. Mice deficient for MyD88 in lung epithelial cells (Sftpccre-Myd88-lox mice) demonstrated a reduced influx of neutrophils into the bronchoalveolar space and an impaired early antibacterial defense after infection with P. aeruginosa, while the response of mice deficient for MyD88 in myeloid cells (LysMcre-Myd88-lox mice) was unremarkable. The immune enhancing role of epithelial MyD88 was dependent on a recognition of pathogenderived flagellin by epithelial TLR5, as demonstrated by an unaltered clearance of mutant P. aeruginosa lacking flagellin from the lungs of Sftpccre-Myd88-lox mice, and an impaired bacterial clearance in bone marrow chimeric mice lacking TLR5 in parenchymal cells. Together these data indicate that clearance of P. aeruginosa from the airways is dependent on flagellin-TLR5-MyD88 dependent signaling in respiratory epithelial cells.

104

Lung epithelial MyD88 drives early pulmonary clearance of Pseudomonas aeruginosa by a flagellin dependent mechanism

Introduction Pseudomonas (P.) aeruginosa pneumonia frequently occurs in hospitalized patients and is associated with high mortality rates and substantial financial costs (1, 2). In addition, Pseudomonas often colonizes the airways of patients suffering from chronic lung diseases such as cystic fibrosis, chronic obstructive pulmonary disease and bronchiectasis. Colonization by Pseudomonas induces chronic inflammation and contributes to a further decline in lung function (3, 4). Moreover, antibiotic multiresistance of Pseudomonas is an increasing problem (5, 6). Hence, studies on induction of host defense during airway infection by Pseudomonas and mechanisms by which this pathogen initiates inflammation are of great importance. Toll-like receptors (TLRs) occupy a prominent position in the innate immune system by virtue of their capacity to recognize bacterial components (7, 8). Pseudomonas possesses ligands for several TLRs, including TLR2 (lipoprotein), TLR4 (lipopolysaccharide, LPS), TLR5 (flagellin) and TLR9 (cytosinephosphateguanosine DNA) (9), all of which rely on the common adapter myeloid-differentiation factor (MyD)88 for intracellular signaling (7, 8). The importance of TLR dependent signaling for clearance of this pathogen was illustrated by the strongly impaired defense of MyD88 deficient (Myd88-/-) mice during Pseudomonas pneumonia (1012). The interplay between TLR2, TLR4 and TLR5 and the redundancy of these receptors during Pseudomonas infection have been elegantly demonstrated by experiments in which Tlr2-/-, Tlr4-/- and Tlr2-/-/Tlr4-/- mice were infected with wild-type (WT) or a flagellin deficient strain of P. aeruginosa (13, 14). In addition, a recent study in Tlr5-/- mice demonstrated that TLR5 contributes to the early antibacterial response and the recruitment of neutrophils during Pseudomonas pneumonia (15). Several cell types express TLRs in human and murine lung tissue, most notably airway epithelial cells, neutrophils, and alveolar macrophages (9). Respiratory epithelial cells are assumed to play an important role in the initiation of the host response and the attraction of inflammatory cells when they first encounter a pathogen (16). The importance of MyD88 dependent signaling in non-hematopoietic cells for the induction of an effective innate host response against Pseudomonas was demonstrated in a mouse bone marrow (BM) chimera model (17). In accordance, selective expression of MyD88 in lung epithelial cells in otherwise MyD88 deficient mice was sufficient to control bacterial growth, although this effect was largely dependent on MyD88 mediated in IL-1Î˛ receptor signaling (18). Additional studies making use of TLR5 BM chimeras revealed that the expression of TLR5 on residential cells is crucial for the induction of a proinflammatory response to purified flagellin in the lungs (19, 20). However, at present the relative contribution of TLR5 dependent signaling in resident and hematopoietic cells to the innate immune response during infection with a flagellated pathogen is unknown. Therefore, in the present study we aimed to investigate the cell-type specific role of MyD88 in myeloid versus lung epithelial cells and the role of the interaction between TLR5 and flagellin herein. To this end we performed experiments in myeloid and epithelial cell-specific MyD88 deficient mice using WT and flagellin-deficient P. aeruginosa, as well as TLR5 BM chimeras. 105

Chapter 6

Results Sftpccre-Myd88-lox mice have impaired early bacterial clearance after infection with Pseudomonas via the airways To investigate the relative contribution of MyD88 dependent signaling in myeloid and respiratory epithelial cells in host defense during Pseudomonas pneumonia, we crossed mice homozygous for the conditional Myd88 flox allele (Myd88fl/fl mice) (21) with mice expressing Cre under control of the myeloid cell lysozyme M (LysM) promoter (22) (to generate LysMcre-Myd88-lox mice) (23) or the surfactant protein C (Sftpc) promoter (24, 25) (to generate Stfpccre-Myd88-lox mice). In two separate experiments Stfpccre-Myd88-lox mice and LysMcre-Myd88-lox were infected with P. aeruginosa (strain PA01, 5 x 106 colony forming units (CFU) via the airways, and lung bacterial loads were compared with those measured in Cre negative Myd88fl/fl littermates at 6 or 24 hours thereafter. At 6 hours after infection StfpccreMyd88-lox mice had 10-100 fold higher bacterial burdens in lungs (Figure 1A) and bronchoalveolar lavage fluid (BALF) (Figure 1C) when compared with littermate controls (p < 0.001); the impaired antibacterial defense in Stfpccre-Myd88-lox mice was further illustrated by the fact that 50% (4/8) of these animals had a positive blood culture for Pseudomonas, versus none of 8 control mice (p < 0.05). In contrast, bacterial loads in lung and BALF of LysMcre-Myd88-lox and control mice were similar at this early time point (Figure 1B,D) and neither group had positive blood cultures. At 24 hours post infection, lung bacterial loads in Stfpccre-Myd88lox and LysMcre-Myd88-lox mice were similar to those in their respective littermate control mice (Figure 1A,B). These data suggest that epithelial MyD88, but not myeloid MyD88, contributes to an effective early clearance of Pseudomonas from the airways. Sftpccre-Myd88-lox mice have an impaired early pulmonary inflammatory response during Pseudomonas infection The impaired early bacterial clearance in Stfpccre-Myd88-lox mice at 6 hours post infection coincided with a markedly diminished influx of neutrophils into the bronchoalveolar space of these animals at this early time point, as demonstrated by reduced neutrophil counts in BALF (Figure 2A, p < 0.01 versus control mice). The number of neutrophils in lung tissue did not differ between Stfpccre-Myd88-lox and control mice, as reflected by similar myeloperoxidase (MPO) concentrations in whole lung homogenates (Figure 2C) and equal numbers of Ly6+ neutrophils in lung tissue slides, quantified by digital image analysis (Figure 2E; representative pictures in Figure 2G). In LysMcre-Myd88-lox mice BALF neutrophil counts, lung MPO levels and the number of Ly6+ neutrophils in lung tissue were not altered relative to control animals (Figure 2B,D,F,H). These results indicate that epithelial MyD88 is important for a swift influx of neutrophils into the alveolar space during Pseudomonas pneumonia, while myeloid MyD88 has a limited role herein. 106

Lung epithelial MyD88 drives early pulmonary clearance of Pseudomonas aeruginosa by a flagellin dependent mechanism

The extent of lung pathology, quantified at 6 and 24 hours after infection according to the scoring system described previously (26) and in the Methods section, was similar in Stfpcre-Myd88-lox and LysMcre-Myd88-lox mice when compared with their respective controls (Supplementary Figure 1). To obtain insight into the contribution of Myd88 dependent signalling in myeloid and respiratory epithelial cells to the early release of inflammatory mediators in the lungs, we measured the concentrations of proinflammatory cytokines and chemokines in whole lung homogenates harvested from Stfpc-Myd88-lox, LysMcre-Myd88-lox and control mice 6 hours after infection with Pseudomonas (Table 1). Stfpccre-Myd88-lox mice displayed higher (CXCL1, IL-6) or unaltered (CXCL2, TNF-α, IL-1β, G-CSF) lung levels of neutrophil attracting mediators; of all mediators measured, only CCL20 levels were significantly lower in lungs of Stfpccre-Myd88-lox mice (p < 0.001 compared to controls, table 1). Remarkably, while mediator levels in lungs of LysMcre-Myd88-lox and control mice were largely similar, the former mouse strain showed reduced TNF-α and IL-1β concentrations (p < 0.05 relative to controls). Together these data suggest that epithelial and myeloid MyD88 differentially contribute to proinflammatory mediator release in the lungs, wherein epithelial MyD88 in particular mediates CCL20 release, whereas myeloid MyD88 is important for TNF-α and IL-1β production. Considering the attenuated neutrophil influx into the bronchoalveolar space of Stfpccre-Myd88lox mice we also measured neutrophil attracting chemokines in BALF of these animals (supplementary table 1); CXC chemokines were either higher (CXCL1) in Stfpccre-Myd88-lox mice or similar (CXCL2, CXCL5) when compared with controls.

Figure 1: Sftpccre-Myd88-lox mice have impaired early bacterial clearance after infection with Pseudomonas. Control, Sftpccre-Myd88- lox and LysMcre-Myd88-lox mice were intranasally infected with 5x106 CFU P. aeruginosa and sacrificed after 6 or 24 hours. Bacterial loads in lung (A, B) 6 and 24 hours after infection and in BALF 6 hours after infection (C,D) of control (grey bars, n =4-8), SftpccreMyd88- lox (white bars, n = 8) and LysMcre-Myd88-lox mice (striped bars, n = 4-7 mice). Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. Values were compared to control mice determined with MannWhitney U test: *** p < 0.001.

107

Chapter 6

108

2939 (494)

8200 (994)

3572 (689)

16211 (3316)

7280 (716)

4452 (833)

18306 (1431)

12496 (1870)

TNF-Îą

IL-1Î˛

IL-6

CXCL1

CXCL2

CCL2

CCL20

G-CSF

9224 (1836)

6853 (581)***

7199 (1006)

5391 (560)

49598 (7699)**

11566 (1043)**

6978 (1447)

2499 (623)

Stfpccre-Myd88-lox

3818 (438)

12376 (2915)

3166 (499)

6134 (753)

16029 (5692)

4070 (1131)

6902 (1523)

1951 (294)

Control

1749 (581)

8873 (1555)

4134 (1208)

3839 (1380)

9377 (3386)

4277 (1764)

1513 (351)*

489 (192)*

LysMcre-Myd88-lox

Lung cytokine and chemokine levels in Stfpccre-Myd88-lox and LysMcre-Myd88-lox mice after P. aeruginosa airway infection. Mice were infected with 5x106 CFU P. aeruginosa and sacrificed after 6 hours. Homogenates were prepared from right lungs. Cytokine and chemokine levels are presented in pg/ml lung homogenate. Data are mean (SE) of 4-8 mice per group. *p < 0.05, ** p < 0.01, *** p < 0.001 vs control mice.

Control

Lung

Table 1: Lung cytokine and chemokine levels after infection with P. aeruginosa via the airways of mice deficient for MyD88 in epithelial or myeloid cells

Lung epithelial MyD88 drives early pulmonary clearance of Pseudomonas aeruginosa by a flagellin dependent mechanism

Figure 2 (see page 108): Sftpccre-Myd88-lox mice have impaired neutrophil influx into the alveolar space. Control, Sftpccre-Myd88- lox and LysMcre-Myd88-lox mice were intranasally infected with 5x106 CFU P. aeruginosa. Neutrophil numbers in BALF (A, B), myeloperoxidase (MPO) levels in lung (C,D) and number of Ly6+ cells in lung tissue (E-F) 6 hours after infection of control (grey bars, n = 4-8), Sftpccre-Myd88- lox (white bars, n = 8) and LysMcre-Myd88-lox mice (striped bars, n = 4 mice). Panels (G,H) show representative images of Ly-6 staining on lung slides from control, SftpccreMyd88- lox and LysMcre-Myd88-lox mice (original magnification 20x). Data are expressed as box-andwhisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. ** p < 0.01 as compared to control mice determined with Mann-Whitney U test

109

Chapter 6

Sftpc-Myd88-lox mice have unremarkable antibacterial response and neutrophil recruitment after infection with flagellin-deficient Pseudomonas We argued that epithelial cell MyD88 might initiate a protective immune response during Pseudomonas airway infection by recognition of bacterial flagellin via TLR5. To test this possibility we infected Stfpccre-Myd88-lox and control mice with flagellindeficient Pseudomonas (PAO1ΔfliC) (14). We expected that if flagellin drives TLR dependent MyD88 activation in respiratory epithelial cells, the impaired host defense of Stfpccre-Myd88-lox mice seen after infection with WT Pseudomonas PAO1 would not be demonstrable after infection with PAO1ΔfliC. Indeed, bacterial loads were similar in lungs and BALF of Stfpccre-Myd88-lox and control mice at 6 hours after infection with PAO1ΔfliC (Figure 3A,B). In addition, neutrophil numbers in BALF (which were reduced in Stfpccre-Myd88-lox mice after infection with WT PAO1; Figure 2A) did not differ between mouse strains after infection with PAO1ΔfliC (Figure 3C). Lung MPO concentrations (Figure 3D) did not differ between StfpccreMyd88-lox and control mice. Similar to our findings in Stfpccre-Myd88-lox mice after infection with WT PAO1, lung cytokine and chemokine levels were similar in the two mouse strains after infection with PAO1ΔfliC, with the exception of CCL20 (Table 2). Table 2: Lung cytokine and chemokine levels after infection with flagellin deficient P. aeruginosa via the airways of epithelial cell MyD88 deficient mice

Lung

Control

Stfpccre-Myd88-lox

TNF-α

2691 (475)

2583 (581)

IL-1β

9894 (1454)

7752 (1539)

IL-6

21990 (4709)

27528 (6660)

CXCL1

44694 (7782)

108085 (35964)

CXCL2

8349 (1141)

6729 (581)

CCL2

8137 (1710)

8902 (1271)

CCL20

9625 (827)

5349 (238)***

G-CSF

23557 (2480)

19552 (2045)

Lung cytokine and chemokine levels in Stfpccre-Myd88-lox mice after airway infection with a flagellin deficient P. aeruginosa strain. Stfpccre-Myd88-lox mice and control mice were infected with 5x106 CFU P. aeruginosa PAO1ΔfliC and sacrificed after 6 hours. Homogenates were prepared from right lungs. Cytokine and chemokine levels are presented in pg/ml lung homogenate. Data are mean (SE) of 4-8 mice per group. *** p < 0.001 vs control mice.

110

Lung epithelial MyD88 drives early pulmonary clearance of Pseudomonas aeruginosa by a flagellin dependent mechanism

Figure 3: Sftpc-Myd88-lox mice have unremarkable antibacterial response and neutrophil recruitment after infection with flagellin-deficient Pseudomonas. Control and Sftpccre-Myd88- lox mice were intranasally infected with 5x106 CFU of the unflagellated P. aeruginosa strain PAO1ﾎ認liC. Bacterial loads in lung (A) and BALF (B), neutrophil numbers in BALF (C) and myeloperoxidase (MPO) levels in lung (D) 6 hours after infection of control (grey bars, n = 8) and Sftpccre-Myd88- lox mice (white bars, n = 8). Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. Differences between groups were not significant.

TLR5 expressed by parenchymal cells drives early clearance of Pseudomonas from the lungs Previous studies have indicated that TLR5 facilitates clearance of Pseudomonas from the airways (15), most likely through recognition of flagellin (27). To further establish that an interaction between flagellin and epithelial TLR5 drives the clearance of Pseudomonas from the airways, we continued with experiments using Tlr5-/- mice. First, we confirmed a beneficial role for TLR5 in antibacterial defense during Pseudomonas pneumonia by showing higher bacterial loads in lungs and BALF of Tlr5-/- mice 6 hours after infection, when compared with WT mice (p < 0.01, Figure 4A, B). To dissect the contribution of parenchymal (P) and hematopoietic (H) TLR5 in the TLR5-mediated clearance of Pseudomonas from the airways, we created BM chimeras for Tlr5 according to previously described methods (23, 28). To this end, irradiated WT recipient mice were infused with Tlr5-/- BM and vice versa, thereby creating WT mice reconstituted with Tlr5-/- BM (P+/H-) and Tlr5-/mice reconstituted with WT BM (P-/H+), as well as two groups transplanted with autologous BM as controls for the BM transfer procedure: WT mice transplanted with WT BM (P+/H+) and Tlr5-/- mice transplanted with Tlr5-/- BM (P-/H-). All mice were infected with P. aeruginosa PAO1 and euthanized 6 hours later for analyses.

111

Chapter 6

The impaired antibacterial defense found in Tlr5-/- mice was reproduced: Tlr5 P-/Hmice had significantly higher bacterial loads in their lungs and BALF compared to Tlr5 P+/H+ mice (p < 0.05 to < 0.01)(Figure 4C,D). Clearly, TLR5 expression on parenchymal cells was more important for clearance of Pseudomonas from the respiratory tract than TLR5 expression on hematopoietic cells: Tlr5 P-/H+ mice had significantly higher bacterial loads in lungs and BALF than Tlr5 P+/H+ mice (p < 0.01 and p < 0.05), while median lung and BALF CFU counts did not differ between Tlr5 P+/H- and Tlr5 P+/H+ mice, hinting to an insignificant role for hematopoietic TLR5 in antibacterial defense during P. aeruginosa pneumonia.

Figure 4: Parenchymal TLR5 mediates bacterial clearance after infection with Pseudomonas via the airways. Tlr5-/- and WT mice were intranasally infected with 5x106 CFU P. aeruginosa. Bacterial loads in lung (A) and BALF (B) 6 hours after infection of WT (grey bars) and Tlr5-/- mice (white bars). WT (P+) and Tlr5-/- (P−) mice were irradiated and injected with WT (H+) or Tlr5-/- (H−) bone marrow cells. Six weeks after transplantation, mice were infected with 5x106 CFU P. aeruginosa. Bacterial loads in lung (C) and BALF (D) of TLR5 chimeras 6 hours after infection (n = 11–12 per group). Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. ** p < 0.01 compared to WT mice with Mann-Whitney U test in panels A,B. *p < 0.05, **p < 0.01 vs P+/H+ mice determined with Mann–Whitney U test as a follow-up test on Kruskall– Wallis test in panels C,D.

112

Lung epithelial MyD88 drives early pulmonary clearance of Pseudomonas aeruginosa by a flagellin dependent mechanism

Discussion P. aeruginosa is an important cause of pneumonia in hospitalized patients and sufferers from chronic lung disease (29, 30). Previous studies have documented the importance of MyD88 and TLR dependent signaling for the early induction of bacterial clearance during Pseudomonas airway infection (10-13). In the present study we aimed to identify the role of MyD88 dependent signaling in myeloid cells versus type II alveolar lung epithelial cells using the Cre-lox system in a model of acute P. aeruginosa pneumonia. We demonstrate that mice with a selective deficiency of MyD88 in lung epithelial cells (but not mice with myeloid specific MyD88 deficiency) have an impaired clearance of P. aeruginosa from the airways. Additional studies provided evidence that epithelial MyD88 drives pulmonary host defense during Pseudomonas pneumonia by TLR5 mediated recognition of flagellin. Indeed, MyD88 expression in lung epithelial cells was dispensable for an adequate immune response during infection with a mutant Pseudomonas strain lacking flagellin, and BM chimeric mice deficient for TLR5 in parenchymal (including epithelial) cells showed a similarly impaired bacterial clearance as epithelial cell MyD88 deficient mice after infection with WT Pseudomonas. Together these data suggest that early MyD88 dependent signaling in lung epithelial cells mediates clearance of Pseudomonas from the airways by a mechanism that depends on the presence of flagellin (expressed by the pathogen) and TLR5 (expressed by the host). Previously, several reports pointed to a prominent role for parenchymal cells in host defense during acute P. aeruginosa pneumonia, first illustrated in a model of MyD88 BM chimeras (17). Myd88-/- mice transplanted with WT BM showed impaired neutrophil attraction and a delayed bacterial clearance during early stage infection, reproducing the phenotype of Stfpccre-Myd88-lox mice described here, while WT mice transplanted with MyD88 deficient BM were as capable to reduce bacterial loads as control WT chimeras (17). Our results are also in accordance with an earlier report in which MyD88 was over-expressed in CC10 positive (Clara) epithelial cells in otherwise MyD88 deficient mice; these Clara epithelial cell selective MyD88 transgenic mice showed enhanced bacterial clearance and an increased number of migrating neutrophils into the lung after infection with P. aeruginosa when compared with complete Myd88-/- mice (18). This protective effect of selective MyD88 overexpression in lung epithelium was due to its role in IL-1 receptor signaling since it could be partially blocked by an IL-1 receptor antagonist (18). Of note, however, in this study an unflagellated P. aeruginosa strain was used (18), which is of importance since mice deficient for the IL-1 receptor had an improved bacterial clearance of a flagellated Pseudomonas strain (31, 32). These data signify our current data on the importance of MyD88 dependent signaling in lung epithelial cells in early neutrophil recruitment and bacterial clearance during infection with P. aeruginosa by a TLR5flagellin dependent mechanism. In accordance, airway instillation of P. aeruginosa resulted in rapid NF-ÎşB activation in the lungs that was primarily localized to the bronchial epithelium, and NF-ÎşB inhibition reduced neutrophil influx and impaired bacterial clearance (33). Interestingly, while neutrophil numbers in BALF were 113

Chapter 6

reduced in Stfpccre-Myd88-lox mice upon infection with Pseudomonas, neutrophil counts in lung tissue, determined by MPO levels in whole lung homogenates and quantitative Ly6 staining of lung tissue slides, were not altered, suggesting that epithelial MyD88 contributes to transmigration of neutrophils from the interstitium into the bronchoalveolar space. We observed a selective impairment in CCL20 production in Stfpccre-Myd88-lox mice early after infection with either WT or flagellin deficient P. aeruginosa. CCL20 was shown to be specifically produced by type II alveolar cells in response to LPS (a TLR4 ligand) (34), which may explain that Stfpccre-Myd88-lox mice still produced less CCL20 in response to a flagellin deficient mutant strain. Of note, CCL20 exerts bactericidal activity towards Pseudomonas (35), suggesting that impaired CCL20 production may contribute to the impaired bacterial clearance in Stfpccre-Myd88lox mice. In contrast, levels of CXCL1 and IL-6 were higher in Stfpccre-Myd88-lox mice; although the production of these cytokines was found to be MyD88 dependent in both alveolar macrophages and epithelial cells in vitro (36) and in MyD88-/- mice in vivo (37), the presence of high bacterial loads most likely caused cells other than alveolar type II cells to produce high levels of these inflammatory mediators. Lung IL-1β and TNF-α levels were markedly reduced in LysMcre-Myd88-lox mice early after infection. In accordance, a previous study localized IL-1β production primarily to alveolar macrophages, not epithelial cells, after infection with Pseudomonas (38), and data from BM chimeras indicated that the production of IL-1β and TNF-α is induced in a MyD88 dependent way in radiosensitive cells (17). Earlier reports have suggested that TLR2, TLR4 and TLR5 have redundant roles in the detection of P. aeruginosa by demonstrating that either the presence of flagellin or LPS is sufficient for efficient bacterial clearance (13, 37). However, in these studies the use of a flagellin deficient Pseudomonas strain probably concealed the role of TLR5 since this strain is also less motile and therefore less virulent (13, 27). Later, the importance of TLR5 for bacterial clearance was shown in Tlr5-/- mice infected with the PAK strain of P. aeruginosa (15). Our data confirmed this beneficial role of TLR5 in airway infection by P. aeruginosa PA01, and further revealed for the first time, using TLR5 BM chimeras, that parenchymal (most likely epithelial) cells mediate this effect. In accordance, lung inflammation induced by purified flagellin relied on TLR5 expression by parenchymal cells (19, 39). Notably, Pseudomonas flagellin not only is sensed by TLR5, but can also activate the NLRC4 inflammasome (40). Unlike Tlr5-/- and Tlr5 P-/H+ mice, Nlrc4-/- mice demonstrated enhanced clearance of Pseudomonas from the airways (38). Together these data suggest that flagellin triggered induction of NLRC4 inflammasome signaling (ascribed primarily to the alveolar macrophage) (38) and TLR5 signaling (in epithelial cells) have opposite effects on P. aeruginosa clearance during pneumonia. In conclusion, we demonstrate here in vivo that clearance of P. aeruginosa from the airways is dependent on flagellin-TLR5-MyD88 dependent signaling in respiratory epithelial cells. The current results further elaborate insight in the pathophysiology of 114

Lung epithelial MyD88 drives early pulmonary clearance of Pseudomonas aeruginosa by a flagellin dependent mechanism

Pseudomonas pneumonia and may be helpful for the development of therapeutics aimed at specific cell types as an adjunctive therapy to antibiotics for which this pathogen is increasingly resistant

Methods Animals Homozygous Myd88fl/fl mice (21) were crossed with LysMcre (22) (Jackson Laboratory, Bar Harbor, Maine), or Stfpccre mice (24, 25), to generate myeloid (LysMcre-Myd88-lox) and type II lung alveolar epithelial (Stfpccre-Myd88-lox) specific MyD88 deficient mice respectively. Myd88fl/fl Cre negative littermates were used as controls in all experiments. In studies using Tlr5-/- mice, generated as described (19), WT C57Bl/6 mice were obtained from Harlan (Horst, the Netherlands) as controls. All genetically modified mice were backcrossed at least 8 times to a C57Bl/6 background and age- and sex matched when used in experiments. Mice were infected at 9-12 weeks of age. The Animal Care and Use Committee of the University of Amsterdam approved all experiments. Induction of pneumonia and sampling of organs Pneumonia was induced by intranasal inoculation with 5x106 CFU of P. aeruginosa PAO1 or PAO1ΔfliC (14), as described (41, 42). After 6 or 24 hours of infection mice were euthanized after injection anesthesia with ketamine/medotomidine and heart puncture as described before (28). For BAL the trachea was exposed through a midline incision; after cannulation of the trachea and occlusion of the left main bronchus with suture tread lavage of the right lung was performed by instilling 2 × 0.3 mL of sterile phosphate-buffered saline; the left lung was preserved for histopathology after fixation in 10% formalin. Lung was homogenized in sterile saline (1:5, weight/vol) using a tissue homogenizer (Biospec Products, Bartlesville, Oklahoma), and CFUs were determined in lung homogenates, BALF and blood from serial dilutions plated on blood agar plates, incubated at 37°C for 16 hours before colonies were counted. Cell counts were determined for each BALF sample in a hemocytometer (Beckman Coulter, Fullerton, CA, USA) and differential cell counts were performed on cytospin preparations stained with Giemsa stain (DiffQuick; Dade Behring AG, Düdingen, Switzerland). For cytokine and chemokine measurements lung homogenates were lysed in an equal volume of lysis buffer (150mM NaCl,15mM Tris, 1mM MgCl, 1mM CaCl2, 1% Triton, pH 7.4) with protease inhibitors (Roche Complete Protease Inhibitor cocktail) on ice for 30 minutes and spun down. BALF and lung homogenate supernatants were stored at -20 oC until further analysis. Assays TNF-α, IL-6 and CCL2 were measured by using a cytometric bead array multiplex assay (BD Biosciences, San Jose, CA) or ELISA (R&D Systems, Minneapolis, MN). IL-1β, CXCL1, CXCL2, CXCL5, CCL20 and G-CSF were measured by ELISA’s (R&D Systems, Minneapolis, MN); MPO was measured by ELISA from Hycult Biotechnology (Uden, the Netherlands). 115

Chapter 6

Bone marrow transplantation BM chimeric mice were generated exactly as described (26, 28). Briefly, recipient groups (6 weeks of age) received a lethal total body irradiation of two times 4.5 Gy with three hours between the two doses, using a 137Cs irradiator (CIS Bio International, Gif, France) at a dose rate of 0.5 Gy/min, followed by intravenous injection of 5x106 BM cells and 2x105 splenocytes (to protect the irradiated recipient mice from immediate infections) isolated from donor animals as described before (26). Engraftment was checked by flow cytometry based on differential expression of CD45.1 and CD45.2 by donor and recipient cells exactly as described (26). BM chimeras were infected at 12 weeks of age and euthanized 6 hours after infection for analyses. Histologic examination For histologic examination left lungs were harvested and instantly fixed in formalin. After paraffin embedding 5 μm sections were made and stained with hemotoxylin and eosin. These lung tissues were scored by a pathologist blinded for experimental groups at a scale of 0 (absent) to 4 (very severe) with respect to the following parameters: interstitial damage, endothelialitis, peri-bronchitis, oedema, thrombus formation and pleuritis and the surface of the lung that was affected by confluent pneumonia, as described (26). Granulocyte immunohistochemic stainings were prepared using a FITC-labeled anti-mouse Ly6-C/G mAb (BD Biosciences, San Jose, CA) exactly as described before (23). Statistical analysis Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation or as means ± standard error of the mean (tables). For experiments with 2 groups, the Mann–Whitney U test was used to determine statistical significance. For comparisons between more than two groups Kruskall-Wallis test was used as a pretest, followed by Mann Whitney U tests where appropriate. All analyses were done using GraphPad Prism (San Diego, CA). p < 0.05 was considered statistically significant.

Acknowledgments We express our gratitude to Anthony L. DeFranco (Department of Microbiology & Immunology, University of California, San Francisco) for providing us with the Myd88fl/fl mice, to Dr. Richard A. Flavell (Yale University School of Medicine, New Haven, CT) for the Tlr5-/- mice and Dr. Brigid Hogan, (Duke University School of Medicine, Durham, NC) for the Stfpccre mice. We thank Dr. Reuben Ramphal (University of Florida, Gainesville, FL) for the P. aeruginosa PAO1ΔfliC and the PAO1 parent strain and Marieke ten Brink and Joost Daalhuisen for expert technical assistance.

116

Lung epithelial MyD88 drives early pulmonary clearance of Pseudomonas aeruginosa by a flagellin dependent mechanism

Disclosure The authors declare to have no commercial or other associations that might pose a conflict of interest. Funding: M.H.P.v.L. and A.P.N.A.d.P. were supported by a grant from the AMC Graduate School of Medical Sciences (no grant number available).

117

Chapter 6

References

1. Chastre J, Fagon JY. Ventilator-Associated Pneumonia. Am J Respir Crit Care Med 2002;165:867-903.

2. Jones RN. Microbial Etiologies of Hospital-Acquired Bacterial Pneumonia and VentilatorAssociated Bacterial Pneumonia. Clin Infect Dis 2010;51 Suppl 1:S81-S87.

3. Lambiase A, Raia V, Del PM, Sepe A, Carnovale V, Rossano F. Microbiology of Airway Disease in a Cohort of Patients With Cystic Fibrosis. BMC Infect Dis 2006;6:4.

4. Oliver A, Canton R, Campo P, Baquero F, Blazquez J. High Frequency of Hypermutable Pseudomonas Aeruginosa in Cystic Fibrosis Lung Infection. Science 2000;288:1251-1254.

5. Rossolini GM, Arena F, Pecile P, Pollini S. Update on the Antibiotic Resistance Crisis. Curr Opin Pharmacol 2014;18C:56-60.

6. Mesaros N, Nordmann P, Plesiat P, Roussel-Delvallez M, Van EJ, Glupczynski Y, Van LY, Jacobs F, Lebecque P, Malfroot A, et al. Pseudomonas Aeruginosa: Resistance and Therapeutic Options at the Turn of the New Millennium. Clin Microbiol Infect 2007;13:560578.

7. Beutler BA. TLRs and Innate Immunity. Blood 2009;113:1399-1407.

8. Kawai T, Akira S. Toll-Like Receptors and Their Crosstalk With Other Innate Receptors in Infection and Immunity. Immunity 2011;34:637-650.

9. McIsaac SM, Stadnyk AW, Lin TJ. Toll-Like Receptors in the Host Defense Against Pseudomonas Aeruginosa Respiratory Infection and Cystic Fibrosis. J Leukoc Biol 2012;92:977-985.

10. Power MR, Peng Y, Maydanski E, Marshall JS, Lin TJ. The Development of Early Host Response to Pseudomonas Aeruginosa Lung Infection Is Critically Dependent on Myeloid Differentiation Factor 88 in Mice. J Biol Chem 2004;279:49315-49322.

11. Skerrett SJ, Liggitt HD, Hajjar AM, Wilson CB. Cutting Edge: Myeloid Differentiation Factor 88 Is Essential for Pulmonary Host Defense Against Pseudomonas Aeruginosa but Not Staphylococcus Aureus. J Immunol 2004;172:3377-3381.

12. Skerrett SJ, Wilson CB, Liggitt HD, Hajjar AM. Redundant Toll-Like Receptor Signaling in the Pulmonary Host Response to Pseudomonas Aeruginosa. Am J Physiol Lung Cell Mol Physiol 2007;292:L312-L322.

13. Ramphal R, Balloy V, Jyot J, Verma A, Si-Tahar M, Chignard M. Control of Pseudomonas Aeruginosa in the Lung Requires the Recognition of Either Lipopolysaccharide or Flagellin. J Immunol 2008;181:586-592.

14. Fleiszig SM, Arora SK, Van R, Ramphal R. FlhA, a Component of the Flagellum Assembly Apparatus of Pseudomonas Aeruginosa, Plays a Role in Internalization by Corneal Epithelial Cells. Infect Immun 2001;69:4931-4937.

15. Morris AE, Liggitt HD, Hawn TR, Skerrett SJ. Role of Toll-Like Receptor 5 in the Innate Immune Response to Acute P. Aeruginosa Pneumonia. Am J Physiol Lung Cell Mol Physiol 2009;297:L1112-L1119.

118

Lung epithelial MyD88 drives early pulmonary clearance of Pseudomonas aeruginosa by a flagellin dependent mechanism

16. Parker D, Prince A. Innate Immunity in the Respiratory Epithelium. Am J Respir Cell Mol Biol 2011;45:189-201.

17. Hajjar AM, Harowicz H, Liggitt HD, Fink PJ, Wilson CB, Skerrett SJ. An Essential Role for Non-Bone Marrow-Derived Cells in Control of Pseudomonas Aeruginosa Pneumonia. Am J Respir Cell Mol Biol 2005;33:470-475.

18. Mijares LA, Wangdi T, Sokol C, Homer R, Medzhitov R, Kazmierczak BI. Airway Epithelial MyD88 Restores Control of Pseudomonas Aeruginosa Murine Infection Via an IL-1-Dependent Pathway. J Immunol 2011;186:7080-7088.

19. Feuillet V, Medjane S, Mondor I, Demaria O, Pagni PP, Galan JE, Flavell RA, Alexopoulou L. Involvement of Toll-Like Receptor 5 in the Recognition of Flagellated Bacteria. Proc Natl Acad Sci U S A 2006;103:12487-12492.

20. Janot L, Sirard JC, Secher T, Noulin N, Fick L, Akira S, Uematsu S, Didierlaurent A, Hussell T, Ryffel B, et al. Radioresistant Cells Expressing TLR5 Control the Respiratory Epithelium’s Innate Immune Responses to Flagellin. Eur J Immunol 2009;39:1587-1596.

21. Hou B, Reizis B, DeFranco AL. Toll-Like Receptors Activate Innate and Adaptive Immunity by Using Dendritic Cell-Intrinsic and -Extrinsic Mechanisms. Immunity 2008;29:272-282.

22. Clausen BE, Burkhardt C, Reith W, Renkawitz R, Forster I. Conditional Gene Targeting in Macrophages and Granulocytes Using LysMcre Mice. Transgenic Res 1999;8:265-277.

23. van Lieshout MH, Anas AA, Florquin S, Hou B, Van’t Veer C, de Vos AF, van der Poll T. Hematopoietic but Not Endothelial Cell MyD88 Contributes to Host Defense During GramNegative Pneumonia Derived Sepsis. PLoS Pathog 2014;10:e1004368.

24. Okubo T, Hogan BL. Hyperactive Wnt Signaling Changes the Developmental Potential of Embryonic Lung Endoderm. J Biol 2004;3:11.

25. Okubo T, Knoepfler PS, Eisenman RN, Hogan BL. Nmyc Plays an Essential Role During Lung Development As a Dosage-Sensitive Regulator of Progenitor Cell Proliferation and Differentiation. Development 2005;132:1363-1374.

26. Wieland CW, van Lieshout MH, Hoogendijk AJ, van der Poll T. Host Defence During Klebsiella Pneumonia Relies on Haematopoietic-Expressed Toll-Like Receptors 4 and 2. Eur Respir J 2011;37:848-857.

27. Balloy V, Verma A, Kuravi S, Si-Tahar M, Chignard M, Ramphal R. The Role of Flagellin Versus Motility in Acute Lung Disease Caused by Pseudomonas Aeruginosa. J Infect Dis 2007;196:289-296.

28. van Lieshout MH, Blok DC, Wieland CW, de Vos AF, van ‘t Veer C, van der Poll T. Differential Roles of MyD88 and TRIF in Hematopoietic and Resident Cells During Murine Gram-Negative Pneumonia. J Infect Dis 2012;206:1415-1423.

29. Chastre J, Fagon JY. Ventilator-Associated Pneumonia. Am J Respir Crit Care Med 2002;165:867-903.

30. Jones RN. Microbial Etiologies of Hospital-Acquired Bacterial Pneumonia and VentilatorAssociated Bacterial Pneumonia. Clin Infect Dis 2010;51 Suppl 1:S81-S87.

119

Chapter 6

31. Cohen TS, Prince AS. Activation of Inflammasome Signaling Mediates Pathology of Acute P. Aeruginosa Pneumonia. J Clin Invest 2013;123:1630-1637.

32. Schultz MJ, Rijneveld AW, Florquin S, Edwards CK, Dinarello CA, Van Der Poll T. Role of Interleukin-1 in the Pulmonary Immune Response During Pseudomonas Aeruginosa Pneumonia. Am J Physiol Lung Cell Mol Physiol 2002;282:L285-L290.

33. Sadikot RT, Zeng H, Joo M, Everhart MB, Sherrill TP, Li B, Cheng DS, Yull FE, Christman JW, Blackwell TS. Targeted Immunomodulation of the NF-KappaB Pathway in Airway Epithelium Impacts Host Defense Against Pseudomonas Aeruginosa. J Immunol 2006;176:4923-4930.

34. Yamamoto K, Ferrari JD, Cao Y, Ramirez MI, Jones MR, Quinton LJ, Mizgerd JP. Type I Alveolar Epithelial Cells Mount Innate Immune Responses During Pneumococcal Pneumonia. J Immunol 2012;189:2450-2459.

35. Yang D, Chen Q, Hoover DM, Staley P, Tucker KD, Lubkowski J, Oppenheim JJ. Many Chemokines Including CCL20/MIP-3alpha Display Antimicrobial Activity. J Leukoc Biol 2003;74:448-455.

36. Raoust E, Balloy V, Garcia-Verdugo I, Touqui L, Ramphal R, Chignard M. Pseudomonas Aeruginosa LPS or Flagellin Are Sufficient to Activate TLR-Dependent Signaling in Murine Alveolar Macrophages and Airway Epithelial Cells. PLoS ONE 2009;4:e7259.

37. Ramphal R, Balloy V, Huerre M, Si-Tahar M, Chignard M. TLRs 2 and 4 Are Not Involved in Hypersusceptibility to Acute Pseudomonas Aeruginosa Lung Infections. J Immunol 2005;175:3927-3934.

38. Cohen JM, Khandavilli S, Camberlein E, Hyams C, Baxendale HE, Brown JS. Protective Contributions Against Invasive Streptococcus Pneumoniae Pneumonia of Antibody and Th17Cell Responses to Nasopharyngeal Colonisation. PLoS One 2011;6:e25558.

39. Van Maele L, Fougeron D, Janot L, Didierlaurent A, Cayet D, Tabareau J, Rumbo M, CorvoChamaillard S, Boulenouar S, Jeffs S, et al. Airway Structural Cells Regulate TLR5-Mediated Mucosal Adjuvant Activity. Mucosal Immunol 2014;7:489-500.

40. Lavoie EG, Wangdi T, Kazmierczak BI. Innate Immune Responses to Pseudomonas Aeruginosa Infection. Microbes Infect 2011;13:1133-1145.

41. Renckens R, van Westerloo DJ, Roelofs JJ, Pater JM, Schultz MJ, Florquin S, van der PT. Acute Phase Response Impairs Host Defense Against Pseudomonas Aeruginosa Pneumonia in Mice. Crit Care Med 2008;36:580-587.

42. van Zoelen MA, Florquin S, Meijers JC, de BR, De Vos AF, de Boer OJ, Van Der Poll T. PlateletActivating Factor Receptor Contributes to Host Defense Against Pseudomonas Aeruginosa Pneumonia but Is Not Essential for the Accompanying Inflammatory and Procoagulant Response. J Immunol 2008;180:3357-3365.

120

Lung epithelial MyD88 drives early pulmonary clearance of Pseudomonas aeruginosa by a flagellin dependent mechanism

Supplementary appendix chapter 6 Supplementary Table 1: CXC chemokine levels in bronchoalveolar lavage fluid of mice deficient for MyD88 in epithelial cells after infection with P. aeruginosa via the airways

BALF

Control

Stfpccre-Myd88-lox

CXCL1

2626 (256)

4448 (509)**

CXCL2

1969 (257)

1905 (184)

CXCL5

9158 (1128)

7791 (623)

Control and Sftpccre-Myd88- lox mice were intranasally infected with 5x106 CFU P. aeruginosa. Six hours after infection, mice were sacrificed, the right lung was lavaged and cytokine levels were determined in BALF supernatant. Data are presented in pg/ml BALF as mean Âą SEM. N=8 mice per group. ** p < 0.01, Legend Supplementary figure 1 (see page 122): Lung inflammation in Sftpccre-Myd88-lox and LysMcre-Myd88-lox mice during P. aeruginosa infection. Control, Sftpccre-Myd88- lox and LysMcreMyd88-lox mice were intranasally infected with 5x106 CFU P. aeruginosa and sacrificed after 6 or 24 hours. Total lung inflammation score as described in the methods (A, B) 6 and 24 hours after infection of control (grey bars, n =4- 8), Sftpccre-Myd88- lox (white bars, n = 8) and LysMcre-Myd88-lox mice (striped bars, n = 4-7 mice). Panels C and D show representative representative images of H&E staining on lung slides from control, Sftpccre-Myd88- lox and LysMcre-Myd88-lox mice 6 and 24 hours after infection, original magnification 20x.Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation.

121

Chapter 6

Supplementary figure 1: Lung inflammation in Sftpccre-Myd88-lox and LysMcre-Myd88-lox mice during P. aeruginosa infection. See legend on page 121.

122

Chapter 7 NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia American Journal of Respiratory Cell and Molecular Biology 2014 Apr;50(4):699-712 DOI: 10.1165/rcmb.2013-0015OC Miriam H.P. van Lieshout 1,2 Brendon P. Scicluna 1,2 Sandrine Florquin 3 Tom van der Poll 1,2,4 Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands: 1 Center of Infection and Immunity Amsterdam 2 Center of Experimental and Molecular Medicine 3 Department of Pathology 4 Division of Infectious Diseases

Chapter 7

Abstract StreptococcusÂ (S.) pneumoniaeÂ is the most frequently isolated causative pathogen of community-acquired pneumonia, a leading cause of mortality worldwide. Inflammasomes are multiprotein complexes that play crucial roles in the regulation of inflammation. NLRP3 (Nod-like receptor family, pyrin domain containing 3) is a sensor that functions in a single inflammasome, whereas ASC (the adaptor apoptosisassociated speck-like protein containing a caspase activation and recruitment domain) is a common adaptor of several inflammasomes. We investigated the role of NLRP3 and ASC during S. pneumoniae pneumonia by comparing bacterial growth and spreading, and host innate immune responses in wild-type mice and mice deficient for either NLRP3 (Nlrp3-/-) or ASC (Asc-/-). Asc-/ - mice had increased bacterial dissemination and lethality compared to Nlrp3-/- mice, although the cytokine response was impaired in both mouse strains. By detailed analysis of the early inflammatory response in the lung by whole genome transcriptional profiling, we identified several mediators that were differentially expressed between Nlrp3-/- and Asc-/ - mice. Of these, interleukin-17, granulocyte-macrophage colony-stimulating factor and integrin alpha M were significantly attenuated in Asc-/- relative to Nlrp3/mice as well as a number of genes involved in the adaptive immune response. These differences may explain the increased susceptibility of Asc-/ - mice during S. pneumoniae infection and suggest that either ASC-dependent NLRP3-independent inflammasomes or inflammasome independent ASC functions may be involved.

124

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

Introduction Streptococcus (S.) pneumoniae is the most frequently isolated causative pathogen in patients with community-acquired pneumonia (1, 2) and a common cause of sepsis, especially in the context of pneumonia (3). Worldwide, the mortality rate associated with pneumococcal pneumonia ranges from 6 to >40% depending on the setting of outpatients or patients in general hospital wards or intensive cares (2). This, together with the increasing incidence of antibiotic resistance in S. pneumoniae (4), emphasizes the importance of expanding our knowledge of host defense mechanisms that influence the outcome of pneumococcal pneumonia. In recent years, the importance of Nod-like receptors (NLRs) for the antimicrobial response has become apparent (5, 6). NLRs are cytosolic receptors that can be part of large multi-protein complexes called inflammasomes. Interleukin-1 beta (IL-1β), one of the most potent proinflammatory cytokines (5, 6), is at a post-transcriptional level tightly regulated by these inflammasomes, together with the proinflammatory cytokine IL-18 (5, 6). Inflammasomes recognize a diverse set of inflammationinducing stimuli including pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs) and consist of a cytosolic sensor that (together with an adaptor protein) recruits pro-caspase-1, resulting in the generation of active caspase-1, the enzyme responsible for activation of pro-IL1β to mature IL-1β. The sensor protein can be either a member of the NLR or the pyrin and HIN domain-containging(PYHIN) family and, if the NLR sensor does not have a caspase activation and recruitment domain (CARD), ASC (the adaptor apoptosis-associated speck-like protein containing a CARD) is necessary to recruit and bind caspase-1 (6). NLRP3 (NLR family, pyrin domain containing 3) is one of the best studied members of the NLR family that can be activated by a large variety of signals, including pneumolysin, a major virulence factor of S. pneumoniae (7). Recently, the role of the NLRP3 and ASC inflammasome complexes in S. pneumoniae induced cell activation in vitro and infection in vivo was investigated (7-9). We here expanded these previous studies using a model of pneumococcal pneumonia and observed a remarkable susceptibility of especially Asc deficient (Asc-/-) mice when compared with Nlrp3-/- mice. Considering that the afore mentioned investigations on the role of the inflammasome during pneumococcal pneumonia predominantly focused on mechanistic studies in purified macrophages (7-9), we here aimed to characterize the initiation of the host response to pulmonary infection in mice deficient for inflammasome components in vivo in more detail by performing a genome-wide scan of transcriptional responses in lung tissue at an early time point after lower respiratory tract infection with S. pneumoniae. We identified a strong influence of the inflammasome components ASC and NLRP3 on the transcriptional response during pneumococcal pneumonia and in addition a differential gene expression pattern between Asc-/- and Nlrp3-/- mice.

125

Chapter 7

Methods Mice Nlrp3-/- and Asc-/- mice (10) were backcrossed 9 times to a C57Bl/6 background. Age- and sex-matched wild type (WT) C57Bl/6 mice were from Harlan (Horst, the Netherlands). Mice were infected at 9-12 weeks. All experiments were approved by the Institutional Animal Care and Use Committee of the Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands. Induction of pneumonia and tissue harvest Pneumonia was induced by intranasal inoculation with 1-2x107 CFU of S. pneumoniae D39 (serotype 2) (11). For two experiments isogenic pneumolysindeficient D39 was used (11). Mice were followed for 14 days or euthanized at 6 or 48 hours, and bacterial outgrowth and cytokine levels were determined (11). Cytokine assays Tumor necrosis factor-α (TNF-α), Interleukin (IL)-6, IL-10 and chemokine (C-C motif) ligand 2 (CCL2) were measured by a cytometric bead assay (BD Biosciences, San Jose, CA). IL-1β, Chemokine (C-X-C motif) ligand 1 and 2 (CXCL1 and 2), granulocyte-macrophage colony-stimulating factor (GM-CSF), granulocyte colony-stimulating factor (G-CSF) and IL-17A (cross-reactivity with IL-17A/F) were measured by ELISA (R&D Systems, Minneapolis, MN). Myeloperoxidase (MPO) ELISA was from HyCult Biotechnology, Uden, the Netherlands. Histology Lungs were fixed in formalin and embedded in paraffin. 5 μm sections were stained with hemotoxylin and eosin. The following parameters were scored on a scale of 0 (absent) to 4 (very severe) by a pathologist blinded for experimental groups: interstitial damage, vasculitis, peri-bronchitis, oedema, thrombus formation and pleuritis. Granulocyte staining was performed (12). RNA preparation and genome-wide transcriptional profiling RNA was isolated from lung homogenates using the Nucleospin RNA II kit (MacheryNagel, Duren, Germany). 750ng of biotinylated cRNA was hybridized onto the Illumina MouseRef-8v2 Expression BeadChip (Eindhoven, the Netherlands). The samples were scanned using the Illumina iScan array scanner(Eindhoven, the Netherlands. Preparation of cRNA, chip hybridization, washing, staining and scanning were carried out at ServiceXS (Leiden, the Netherlands). The raw scan data were read using the beadarray package (version 1.12.1) (13), available through Bioconductor (14) in the R statistical environment (version 2.13.2; R Foundation for Statistical Computing, Vienna, Austria). All non-normalized and normalized data are available at the gene expression omnibus of NCBI (GEO) with accession number GSE42464. Details on bioinformatics are available in the online supplement.

126

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

Quantitative real time PCR Total RNA was reverse transcribed using oligo (dT) primer and Moloney murine leukemia virus reverse transcriptase (Invitrogen, Breda, the Netherlands). Quantitative PCR of 4933427D14Rik, Nlrp3, Asc, Gdpd3, Csf2, Csf3, Trim 16, Zfp39, Unc 45b and Mid-1genes was performed. Data was analyzed using the LinRegPCR program (15). Results were normalized to Gapdh transcript. Primer pairs were designed to avoid overlap with array probes (Table S5). Statistical analysis and bioinformatics Data are expressed as means Âą standard error of mean or as box-and-whisker diagrams. Comparison between two groups was by Mann Whitney U test. For experiments with > 2 groups, the Kruskall-Wallis test was used as a pretest, followed by Mann Whitney U tests where appropriate. Proportions of positive cultures were compared by Fisherâ&#x20AC;&#x2122;s exact test. Kaplan-Meier survival plots were compared using log-rank test. For these analyses, GraphPad Prism (San Diego, CA) was used. p< 0.05 was considered statistically significant. Differential probe intensities on microarrays were identified using the R package limma (version 3.8.3) (16). Enrichment of functional annotations was performed by the DAVID Database (17). We report Benjamini and Hochberg (BH) multiplecomparison corrected p-values.

Results Both NLRP3 and ASC contribute to the antibacterial response, but only ASC is crucial for survival during pneumococcal pneumonia We first infected WT, Nlrp3-/-, and Asc-/- mice with S. pneumoniae via the airways and euthanized the mice at 6 or 48 hours for quantitative cultures of lungs, blood and spleen. For the predefined 6-hour time point, Nlrp3-/-, Asc-/- and WT mice were infected concurrently; the 48-hour time point was investigated in two separate experiments for Nlrp3-/- and Asc-/- mice (each with a matched WT group) because of difficulties in synchronizing their breeding. At 6 hours, the bacterial loads in the lungs of both Nlrp3-/- and Asc-/- mice were similar to those in WT mice (Figure 1A). However, whereas after 48 hours WT mice showed a reduction of bacterial loads, pulmonary bacterial counts in both Nlrp3-/- and Asc-/- mice remained significantly higher (both p< 0.05 versus WT mice). Both Nlrp3-/- and Asc-/- mice showed an earlier dissemination of pneumococci from the primary site of infection, as reflected by a significantly higher proportion of positive blood cultures 6 hours post infection when compared with WT mice: 50% of Nlrp3-/- and Asc-/- mice had a positive blood culture while blood cultures of all WT mice remained sterile (Figure 1B, both p <0.05 versus WT mice). Enhanced infection of distant organs was also indicated by higher bacterial loads in blood and spleens of Asc-/- mice 48 hours after infection (all p < 0.05 versus WT mice); notably, Nlrp3-/- mice did not show higher bacterial burdens in blood and spleen at this late time point. Previous studies have documented an important role for the S. pneumoniae virulence factor pneumolysin in activation of 127

Chapter 7

Figure 1: NLRP3 and ASC contribute to bacterial clearance of S. pneumoniae, but only ASC is important for survival. WT, Nlrp3 -/- and Asc-/- mice were inoculated with 1-2x107 CFU S. pneumoniae and sacrificed at designated time-points or monitored for survival. Bacterial loads in lung (A), blood (B) and spleen (C) 6 and 48 hours after infection in WT (white), Nlrp3-/- (dark grey) and Asc-/- mice (light grey) (n=6-9 per group). Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. Survival of WT (white symbols), Nlrp3/(light grey symbols) and Asc-/- mice (dark grey symbols) (n=19-28 per group) expressed as KaplanMeier plot (D). * p < 0.05 compared to WT mice determined with Mann-Whitney U test. # p < 0.05 compared to WT mice determined with Fishers exact test, *** p < 0.001 for the comparison between Asc/and WT mice, and between Asc-/- and Nlrp3-/- mice as determined by Log-Rank test. ND = not done.

the NLRP3 inflammasome in macrophages in vitro (7-9). In accordance, 48 hours after infection with an isogenic pneumolysin deficient S. pneumoniae D39 strain bacterial loads were similar in lungs and distant organs of Nlrp3-/- and Asc-/- mice when compared with WT mice (Figure S1). Thus, further experiments were done with WT S. pneumoniae exclusively. Next, we infected Nlrp3-/-, Asc-/- and WT mice with S. pneumoniae and followed them for 14 days in a survival experiment (Figure 1D). After 68 hours, mice in all groups started dying; however, in Asc-/- mice mortality was significantly higher and accelerated compared to WT mice: 50% of Asc-/- mice died within 4 days and <10% remained alive until the end of the experiment, while > 60% of WT mice survived (p < 0.001 for the difference between Asc-/- and WT mice). The mortality curve of Nlrp3-/- mice was similar to that of WT mice and their survival was significantly better than that of Asc-/- mice (p< 0.001 Nlrp3-/- compared to Asc-/- mice).

128

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

Impact of NLRP3 and ASC deficiency on the early inflammatory response in the lungs after infection with S. pneumoniae The early immune response to S. pneumoniae in the lower airways is of utmost importance for an adequate defense against uncontrolled bacterial multiplication (2, 18). The studies described here documented that Nlrp3-/- and Asc-/- mice displayed enhanced bacterial growth relative to WT mice during late stage pneumococcal pneumonia, which in Asc-/- mice was accompanied by enhanced bacterial dissemination and an increased lethality. Early after infection, at a time point (6 hours) when normal WT mice show a brisk inflammatory response in their lungs in this model (11), bacterial loads were still similar in lungs of Nlrp3-/-, Asc-/- and WT mice, allowing a detailed analysis of the impact of NLRP3 and ASC deficiency on the induction of innate immunity at the primary site of infection without bias caused by differences in bacterial loads (which have been shown to be a major denominator of the extent of lung inflammation during murine pneumococcal pneumonia) (19). Therefore, we measured several components of the inflammatory response in lung tissue harvested from Nlrp3-/-, Asc-/- and WT mice 6 hours post infection. Overall, Nlrp3-/- and Asc-/- mice showed reduced cytokine and chemokine levels in whole lung homogenates when compared with WT mice at this early time point (Table 1). Lung levels of IL-6, CXCL1, CXCL2 and CCL2 were significantly lower in both knockout strains; in addition, Nlrp3-/- mice showed significantly diminished IL-1Î˛ levels (p < 0.05 compared to WT mice) and Asc-/- mice had significantly lower TNF-Îą and IL-17 levels (p < 0.05 and 0.001 compared to WT mice respectively). Cytokine and chemokine levels were not significantly different between Nlrp3-/- and Asc-/- mice except for IL-17 that was significantly lower in lungs of Asc-/- mice (p < 0.05). In plasma, IL-6 and CCL2 levels were lower in Nlrp3-/- and Asc-/- mice at this early time point (Table S1). To obtain further insight into the role of NLRP3 and ASC in induction of lung inflammation during S. pneumoniae pneumonia, we prepared hemotoxylin and eosin stained slides and scored key histological features according to the scoring system described in the Methods section. All three mouse strains showed extensive signs of pneumonia, characterized by interstitial inflammation, pleuritis, vasculitis, bronchitis and edema 6 hours after infection (Figure 2A-C). However, neither total inflammation scores nor individual components of the inflammation scores were different between genotypes (Figure 2D). Similarly, the extent of neutrophil influx, as measured by the number of Lymphocyte antigen6G (Ly-6G) and Lymphocyte antigen-6C (Ly-6C) positive cells in lung tissue slides (Figure 2E-G) and the concentrations of MPO in whole lung homogenates (Figure 2H), did not differ between Nlrp3-/-, Asc-/- and WT mice.

129

Chapter 7 Table 1: Nlrp3-/- and Asc-/- mice have an impaired inflammatory response early after induction of pneumonia by S. pneumoniae.

Lung

Nlrp3-/-

Asc-/-

IL-1β

617 (35)

436 (42)**

558 (73)

TNF-α

2763 (322)

2115 (733)

1770 (479)*

IL-6

5259 (474)

1919 (471)**

1379 (252)***

IL-10

12 (1)

14 (2)

IL-17

73 (5)

85 (10)

53 (4)**^

CXCL1

50590 (4810)

13553 (2728)***

19101 (3878)**

CXCL2

48971 (6592)

19857 (2493)**

23153 (4686)*

CCL2

5026 (280)

2309 (355)**

2363 (260)***

WT, Nlrp3-/- and Asc-/- mice were inoculated with 1-2x107 CFU S. pneumoniae and sacrificed 6 hours later. Homogenates were prepared from right lungs. Cytokine and chemokine levels are presented in pg/ml lung homogenate. Data are mean (SE) of 6-9 mice per group.* p < 0.05, ** p < 0.01, *** p < 0.001 vs WT mice. ^ p < 0.01 Asc-/- versus Nlrp3-/- mice.

Impact of NLRP3 and ASC deficiency on the late inflammatory response during pneumococcal pneumonia We determined the extent of local and systemic inflammation at 48 hours post infection, i.e., shortly before the first deaths occurred. At this late time point, lung pathology scores were slightly lower relative to those documented at 6 hours (Figure 3), most likely reflective of the lower bacterial loads. Whereas, pathology scores were similar in Nlrp3-/- and WT mice, Asc-/- mice showed significantly lower scores. The initially impaired pulmonary cytokine/chemokine response of Nlrp3/and Asc-/- mice had disappeared. In fact, at this late time point the lung levels of cytokines and chemokines were higher in inflammasome deficient mice (likely as a consequence of their higher bacterial burdens), although only reaching statistical significance for IL-6 in Nlrp3-/- mice and CXCL2 for Asc-/- mice (Table 2). Similarly, plasma levels of cytokines and chemokines were higher in inflammasome deficient mice, reaching significance for TNFα and IL-6 in Nlrp3-/- mice (Table S2). Finally, we measured markers of cellular injury in plasma (urea, aspartate aminotransferase, alanine aminotransferase and lactate dehydrogenase) as a readout for distant organ damage (Table 3). Although the plasma concentrations of these markers tended to be higher in Nlrp3-/- and Asc-/- mice, differences between groups were not statistically significant.

130

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

Figure 2: Inflammatory response early after infection with S. pneumoniae. For legend, see page 132.

131

Chapter 7 Legend Figure 2 (page 131): Inflammatory response early after infection with S. pneumoniae. WT, Nlrp3 -/- and Asc-/- mice were inoculated with 1-2x107 CFU S. pneumoniae (n=6-9 per group); 6 hours after infection histological scores, determined as described in the Methods section, were similar in WT, Nlrp3-/- and Asc-/- mice (D). Panels A-C show representative lung histology of WT (A), Nlrp3-/- (B) and Asc-/- mice (C) H&E staining, original magnification 20x. Neutrophil influx did not differ between mouse groups as reflected by similar Ly6-G and Ly6-C lung surface positivity (H) and whole lung MPO levels (I). Panels E-G show representative images of Ly-6G and Ly-6C staining on lung slides from WT (E), Nlrp3-/- (F) and Asc-/- mice (G); Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. Table 2: Lung levels of chemokines and cytokines 48 hours after induction of pneumonia by S. pneumoniae.

Lung

Nlrp3-/-

Asc-/-

IL-1Î˛

TNF-Îą

228 (49)

411 (81)

715 (105)

1538 (467)

IL-6

228 (72)

810 (205)*

1086 (361)

1590 (477)

IL-10

10 (1)

12 (2)

18 (3)

262 (125)

CXCL1

4720 (503)

6381 (499)

5145 (936)

5969 (807)

CXCL2

2067 (141)

4820 (1360)

5297 (1397)

20032 (7193)*

CCL2

5910 (949)

9338 (834)*

7217 (942)

8278 (924)

WT, Nlrp3-/- and Asc-/- mice were inoculated with 1-2x107 CFU S. pneumoniae and sacrificed 48 hours later. Homogenates were prepared from right lungs. Cytokine and chemokine levels are presented in pg/ml lung homogenate. Data are mean (SE) of 6-9 mice per group. Below detection (bd), * p < 0.05 vs WT mice. Table 3: Parameters of organ and cellular injury 48 hours after induction of pneumonia by S. pneumoniae.

Plasma

Nlrp3-/-

Asc-/-

Urea (mmol/L)

9 (1)

24 (8)

9 (1)

ASAT (U/L)

112 (14)

242 (97)

119 (15)

470 (210)

ALAT (U/L)

14 (1)

11 (4)

13 (4)

12 (6)

LDH (U/L)

505 (128)

546 (185)

425 (73)

710 (187)

WT, Nlrp3-/- and Asc-/- mice (n=6-9 per group) were inoculated with 1-2x107 CFU S. pneumoniae and sacrificed 48 hours later. Plasma levels are presented as mean (SE).

132

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

Figure 3: Inflammatory response during late-stage infection with S. pneumoniae. WT, Nlrp3 -/and Asc-/- mice were inoculated with 1-2x107 CFU S. pneumoniae (n=6-9 per group); 48 hours after infection histological scores, determined as described in the Methods section, were similar in Nlrp3-/- and WT mice but Asc-/- mice showed significantly lower scores (E). Panels A-D show representative lung histology of WT (A), Nlrp3-/- (B), WT (C) and Asc-/- mice (D) H&E staining, original magnification 20x. Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation.

133

Chapter 7

Genome-wide pulmonary transcriptional response in WT, Asc-/- and Nlrp3-/- mice in pneumococcal pneumonia To further investigate the molecular basis of the observed differences in susceptibility between Asc-/- and Nlrp3-/- mice in pneumococcal pneumonia, we performed whole genome transcriptional profiling of lung tissues obtained 6 hours post infection, e.g. at a time that bacterial loads were still similar, from 3-4 female WT, Asc-/- and Nlrp3-/- mice and from matched uninfected controls using pan-genomic expression arrays. After pre-processing and quality control, 23742 transcripts were available for differential abundance analysis. First, we analyzed for gene expression differences between non-infected WT, Asc-/- and Nlrp3-/- mice. Pair-wise analysis (moderated t test) of WT and Nlrp3-/- mice revealed four differentially abundant transcripts (p < 0.05), namely, 4933427D14Rik (probe ID: ILMN_1245850), Zfp39 (probe ID: ILMN_1241611), Trim16 (probe ID: ILMN_2482494) and Pttg1 (probe ID: ILMN_2809167). These transcripts physically map to within ~20Mb of the Nlrp3 locus on chromosome 11 (Figure S2). No differences were detected for the (uninfected) WT and Asc-/- comparison. The genome-wide response to early S.pneumoniae infection was then assessed in WT mice. Pair-wise analysis of the gene expression profiles between uninfected and infected WT mice yielded 5346 differentially abundant transcripts (BH p < 0.05). Figure 4A illustrates the volcano plot (integrating nominal p-values and log2 fold change) of the transcriptional response to S. pneumoniae infection in WT mice. Considering a fold change â&#x2030;Ľ 2 we detected 360 transcripts, while 20 transcripts were detected with a fold change â&#x2030;¤ 2. Notably, Nlrp3 transcript abundance was significantly increased post-infection (p = 4.5x10-7, log2 fold change = 2.1), whereas Asc transcript abundance was not different. Functional annotation enrichment of the gene transcripts that increased in abundance after S. pneumoniae infection is shown in Figure 4B. These annotation clusters include defense response (p = 4.4x10-34), regulation of cytokine production (p = 2.2x10-12), chemotaxis (p= 5.6x1012 ), cell death (p = 4.8x10-8), protein kinase cascade (p = 9.3x10-9) and regulation of adaptive immune response (p = 3.7x10-10) terms. Figure 4C shows the functional annotation enrichment results of those gene transcripts that decreased in abundance after S. pneumoniae infection. These clusters include the mitochondrial part (p = 8.2x10-10), mitochondrial envelope (p= 3.0x10-12), respiratory system development (p = 0.0098) and ion binding (p = 0.035) annotations. Next, we analyzed the pulmonary transcriptional response to early S. pneumoniae infection in Asc-/- and Nlrp3-/- mice. Using the f test approach (ANOVA) and considering a BH corrected p-value < 0.05 we detected 56 gene transcripts that differed in abundance between WT, Asc-/- and Nlrp3-/- mice (Table 4, see pages 138-141). By using an unsupervised hierarchical clustering approach we show that these transcripts clearly discriminate WT, Nlrp3-/- and Asc-/- mouse samples (Figure 5, see page 137). These genes significantly enrich (BH p < 0.05) functional annotation clusters that included regulation of cytokine secretion, extracellular space, and defense response. Prominent genes within these pathways include Csf2, Csf3, Il17f and Foxp3. By univariate analysis we delineated 20 transcripts (BH p < 0.05) that discriminate Nlrp3-/- and Asc-/- mouse genome-wide responses 134

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

Figure 4: Genome-wide pulmonary transcriptional response in WT mice in pneumococcal pneumonia. WT mice were inoculated with 1-2x107 CFU S. pneumoniae and genome-wide lung transcriptional responses were assessed 6 hours after infection (n=4) and compared with the lung transcriptome obtained in uninfected WT mice (n= 3). Volcano plot analysis (integrating nominal p-values and fold changes) of the transcriptional response after infection in WT mice, 5346 transcripts were differentially expressed (BH p-value<0.05; red line) (A). Pie chart representations (pie slices denote enrichment, E, scores for the particular significant cluster) of the transcripts in pneumonia p, BenjaminiHochberg multiple comparison corrected p-values (B,C see page 136).

135

Chapter 7

Figure 4 (continued): Genome-wide pulmonary transcriptional response in WT mice in pneumococcal pneumonia.

136

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

Figure 5: Unsupervised hierarchical clustering heat map of top 56 differentially expressed genes among WT, Nlrp3-/- and Asc-/- mice. WT, Nlrp3 -/- and Asc-/- mice were inoculated with 1-2x107 CFU S. pneumoniae and genome-wide lung transcriptional responses were assessed 6 hours after infection and compared with the lung transcriptome obtained in uninfected mice (n=3-4 mice per strain for each condition). F statistics (ANOVA) of the expression profiles in the lungs of WT, Nlrp3-/- and Asc-/- mice yielded 56 differentially expressed genes (Benjamini-Hochberg p-value < 0.05) represented by an unsupervised hierarchical clustering heatmap plot. Rows correspond to genes while columns correspond to mouse samples. Red denotes high expression, blue denotes low expression. MGI official symbols are illustrated.

137

138

EntrezID

74477

211623

216799

12985

68616

12981

50501

94092

66824

22698

78354

16409

18054

217012

ProbeID

ILMN_1245850

ILMN_2710159

ILMN_1237471

ILMN_1217948

ILMN_2893879

ILMN_2749412

ILMN_3147259

ILMN_2482494

ILMN_29364764

ILMN_1241611

ILMN_1226767

ILMN_2696017

ILMN_1228832

ILMN_3153940

Unc45b

Ngp

Itgam

2210407C18Rik

Zfp39

Asc

Trim16

Prok2

Csf2

Gdpd3

Csf3

Nlrp3

Plac9

4933427D14Rik

Gene Symbol

-1.23656

-1.11445

-0.40243

1.59297

-1.13306

0.249558

-1.51257

-1.42428

-2.15991

-0.14058

-2.03124

-2.18135

1.787369

0.275962

-2.67972

-1.82175

0.088175

0.313151

-1.56012

-0.59139

-2.37021

-0.84308

4.624858

-1.76356

-0.67436

1.789564

0.165437

-1.51252

1.565271

1.419325

1.504794

-1.44621

1.809683

-0.92119

0.945934

-1.31683

-4.76544

-0.26768

-1.50699

-0.0022

-2.80188

v Asc-/-

v WT -2.63645

Nlrp3-/-

log2 FC

Asc-/- v

log2 FC

Nlrp3-/-

log2 FC

28.76223

29.227

31.68581

32.38318

32.898

38.90442

39.62573

40.29537

41.34397

42.89634

49.31849

51.82589

54.25563

81.10563

2.21E-07

1.91E-07

9.01E-08

7.34E-08

6.33E-08

1.24E-08

1.03E-08

8.75E-09

6.76E-09

4.65E-09

1.10E-09

6.52E-10

4.00E-10

4.64E-12

Table 4: Differential transcriptional response in WT, Nlrp3-/- and Asc-/- mice after induction of pneumococcal pneumonia.

0.000341

0.000315

0.000161

0.000142

0.000133

3.20E-05

3.00E-05

2.90E-05

2.61E-05

2.16E-05

6.38E-06

5.04E-06

4.64E-06

1.07E-07

BH p

Chapter 7

EntrezID

66039

66628

101490

257630

16407

30939

16878

98365

16415

21940

103712

26904

26410

216850

27078

ProbeID

ILMN_2693858

ILMN_2979430

ILMN_2785512

ILMN_2741201

ILMN_2699898

ILMN_2809167

ILMN_3137291

ILMN_2663249

ILMN_1258735

ILMN_1233589

ILMN_2594139

ILMN_2978617

ILMN_1245924

ILMN_3095624

ILMN_2713969

B9d1

Kdm6b

Map3k8

Sh2d1b1

6330403K07Rik

Cd27

Itgb2l

Slamf9

Lif

Pttg1

Itgae

Il17f

Inpp5f

Thg1l

D14Ertd449e

Gene Symbol

0.786093

-1.48959

-0.51505

1.096595

-1.16149

0.88401

-0.89513

1.33698

-0.42806

-1.39936

1.454311

0.716829

0.213559

1.591472

0.181671

-0.41878

-0.87112

1.06378

0.191912

0.665586

-1.82136

0.365524

-1.30856

-0.21191

0.218711

-0.76781

0.830906

-0.11472

1.763997

0.604422

-1.07081

0.356068

0.032814

-1.3534

0.218423

0.926226

0.971456

0.880503

-1.18745

1.2356

1.484635

-0.61735

1.706197

-0.33081

v Asc-/-

v WT 1.433187

Nlrp3-/-

log2 FC

Asc-/- v

log2 FC

Nlrp3-/-

log2 FC

15.97045

16.31111

16.74478

17.58096

17.58899

17.88613

18.57409

18.67304

19.51646

19.80135

20.70172

20.99229

26.41031

26.71657

27.77398

2.76E-05

2.36E-05

1.95E-05

1.35E-05

1.19E-05

8.87E-06

8.51E-06

6.02E-06

5.36E-06

3.75E-06

3.35E-06

4.75E-07

4.29E-07

3.03E-07

0.018898

0.017125

0.014566

0.010443

0.009816

0.007612

0.007585

0.005577

0.005178

0.003951

0.003696

0.000611

0.000584

0.000439

BH p

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

139

140

EntrezID

20371

56369

21983

71520

21869

27055

94091

68458

24050

14778

16319

83397

12796

434760

15978

ProbeID

ILMN_2635132

ILMN_1234781

ILMN_1250696

ILMN_2617656

ILMN_1260212

ILMN_1249550

ILMN_2435814

ILMN_2659824

ILMN_1218471

ILMN_2715546

ILMN_1216213

ILMN_2624622

ILMN_2766604

ILMN_2917647

ILMN_2685712

Ifng

Rhox2d

Camp

Akap12

Incenp

Gpx3

Sept3

Ppp1r14a

Trim11

Fkbp9

Nkx2-1

Grap

Tpbg

Apip

Foxp3

Gene Symbol

0.275066

1.040516

-0.88631

-0.66209

0.866477

-0.75945

1.507093

-0.12876

-0.97242

0.142067

-0.1227

-0.99604

-0.20086

0.139456

-1.18156

0.148384

-2.26747

-0.69725

0.272439

-0.05641

0.050678

0.750609

0.184316

-0.56994

0.612782

0.377567

-0.83329

0.879188

-0.42954

1.456626

0.892131

1.381158

0.035168

0.594038

-0.70304

1.456414

-0.87937

-1.15674

0.712003

-0.73548

-1.37361

0.632426

-0.73973

1.275045

v Asc-/-

v WT 0.845506

Nlrp3-/-

log2 FC

Asc-/- v

log2 FC

Nlrp3-/-

log2 FC

13.93113

13.98752

14.06886

14.3575

14.40149

14.48708

14.61403

14.73513

14.756

15.31883

15.44467

15.5374

15.63157

15.89562

15.9093

7.22E-05

7.02E-05

6.75E-05

5.87E-05

5.74E-05

5.51E-05

5.19E-05

4.90E-05

4.85E-05

3.72E-05

3.51E-05

3.36E-05

3.22E-05

2.85E-05

2.84E-05

0.034838

0.034613

0.033991

0.030217

0.029714

0.02863

0.027686

0.022122

0.021415

0.021071

0.020736

0.018898

BH p

Chapter 7

380924

105428

238871

26569

19730

17318

240263

210146

269016

26382

19288

1199138

ILMN_3161834

ILMN_2641270

ILMN_1236700

ILMN_2709355

ILMN_2689119

ILMN_3159435

ILMN_2724409

ILMN_2715466

ILMN_2942551

ILMN_2737903

ILMN_2662802

ILMN_2676066 Stard7

Ptx3

Fgd2

Sh3rf2

Irgq

Fem1c

Mid1

Ralgds

Slc27a4

Pde4d

Fam149b

Olfm4

Gene Symbol

-0.2594

-0.79681

0.916668

-0.23723

1.054852

0.260737

-1.8982

-0.58143

-0.51691

0.07914

0.11043

-0.51749

-1.08702

0.412132

0.563574

0.125884

-0.66049

0.405619

-1.2147

-1.02555

-0.79903

-0.94065

-1.16156

0.258091

0.290217

0.504536

-0.8008

0.928968

0.921224

-2.30382

0.633271

0.508633

0.878169

1.051076

0.395484

v Asc-/-

v WT -0.76608

Nlrp3-/-

Asc-/- v

log2 FC

Nlrp3-/-

log2 FC

12.79716

13.01888

13.11564

13.20866

13.21971

13.3038

13.32926

13.36526

13.41779

13.56195

13.58135

13.82476

0.000127

0.000114

0.000108

0.000103

9.84E-05

9.72E-05

9.54E-05

9.30E-05

8.65E-05

8.57E-05

7.60E-05

0.049101

0.044614

0.043219

0.041967

0.041467

0.041433

0.039324

0.035953

BH p

WT, Nlrp3-/- and Asc-/- mice were inoculated with 1-2x107 CFU S. pneumoniae and genome-wide lung transcriptional responses were assessed 6 hours after infection. Differential pulmonary transcriptional response to early S. pneumoniae infection in WT, Asc-/- and Nlrp3-/- mice was evaluated by means of the f-test. Shown are the 56 genes that were different among the three genotypes with a Benjamini-Hochberg corrected p-value < 0.05. log2 FC, Log 2 transformed foldchange. F, F statistic. p, nominal p-value. BH p, Benjamini-Hochberg multiple comparison corrected p-value. Mouse Genome Informatics (MGI) official symbols are illustrated.

EntrezID

ProbeID

log2 FC

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

141

Chapter 7

to S.pneumoniae pulmonary infection (Table 5). As expected, Asc and Nlrp3 were amongst the discriminative transcripts. Quantitative real time PCR (qRT-PCR) analysis of Gdpd3, Unc45b, Mid1, (Figure S3) and Csf2 (Figure 6A) lend further weight to our genomics data, suggesting that components of the inflammasome macromolecular complex influence, at least in part, distinct transcriptional responses in pneumococcal pneumonia. Tables S4 and S5 show the univariate analyses for the comparison between WT and Nlrp3-/- mice, and WT and Asc-/mice. In silico mapping of these differentially expressed genes (Tables S4 and S5) for cell-specific gene co-expression characteristics derived from the Toppgene suite (http://toppgene.cchmc.org) unmasked a significant association of the Nlrp3/mouse transcriptome with lung dendritic cells (online supplement, Table S6). No significant association was uncovered for the Asc-/- mice. The emerging role of Csf2 and Csf3, respectively encoding for the growth factors GM-CSF and G-CSF, in regulating lung-protective immunity against S. pneumoniae (20, 21), coupled with our differential gene expression analysis led us to further evaluate their protein abundance. G-CSF protein levels (Figure 6D) reflected both the array and qRT-PCR results for Csf3 transcript abundance (Table 4 and Figure 6C); G-CSF presented statistically significant lower abundance in lungs of Asc-/when compared to WT mice (Figure 6D). No differences in G-CSF protein abundance were observed between Nlrp3-/- and Asc-/- mouse lung samples. Interestingly, GMCSF was significantly lower in Asc-/- mice, but not in Nlrp3-/- mice (Figure 6B), which did not reflect the results obtained from either microarray analysis or qRT-PCR for Csf2 transcript abundance (Table 4 and Figure 6A): mRNA expression was only slightly decreased in Asc-/- compared to WT mice, but strongly decreased in Nlrp3/mice (Figure 6A).

142

74477

22698

66824

217012

68616

78354

216799

66628

257630

103712

20371

16407

71520

12981

ILMN_1245850

ILMN_1241611

ILMN_2936476

ILMN_3153940

ILMN_2893879

ILMN_1226767

ILMN_1237471

ILMN_2979430

ILMN_2741201

ILMN_2594139

ILMN_2635132

ILMN_1217629

ILMN_2617656

ILMN_2749412

Probe ID

Entrez ID

Csf2

Grap

Itgae

Foxp3

6330403K07Rik

Il17f

Thg1l

Nlrp3

2210407C18Rik

Gdpd3

Unc45b

Pycard

Zfp39

4933427D14Rik

Symbol

-1.31683

-1.37361

1.375726

1.275045

-1.3534

1.484635

1.706197

-1.50699

1.504794

-4.76544

-1.51252

1.809683

-1.44621

-2.80188

Asc-/-

Nlrp3-/- v

log2 FC

-5.3602

-5.52507

5.58263

5.633688

-5.65899

6.361767

6.752563

-6.81456

6.947761

-7.21234

-7.34241

7.364298

-7.93374

-11.7189

1.21E-05

7.78E-06

6.68E-06

5.84E-06

5.46E-06

8.70E-07

3.20E-07

2.73E-07

1.95E-07

1.01E-07

7.29E-08

6.90E-08

1.73E-08

5.06E-12

0.01862

0.012878

0.011908

0.01127

0.002239

0.000926

0.000904

0.000753

0.000466

0.000422

0.0002

1.17E-07

BH p

Table 5: Differential transcriptional response between Nlrp3-/- and Asc-/- mice after induction of pneumococcal pneumonia.

3.055198

3.42206

3.54952

3.662289

3.718067

5.235958

6.049902

6.176843

6.447499

6.976465

7.232193

7.274957

8.358301

14.17256

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

143

144

94091

30939

16409

17318

327957

ILMN_2435814

ILMN_2809167

ILMN_2696017

ILMN_3159435

ILMN_2847437 A430084P05Rik

Mid1

Itgam

Pttg1

Trim11

Trim16

Symbol

1.299602

-2.30382

1.419325

-1.18745

-1.15674

-0.92119

Asc-/-

Nlrp3-/- v

4.916517

-4.98778

5.024978

-5.04292

-5.21736

-5.24174

3.93E-05

3.25E-05

2.95E-05

2.81E-05

1.76E-05

1.65E-05

0.041447

0.035893

0.034125

0.02403

0.023927

BH p

2.057287

2.218402

2.30239

2.342876

2.735426

2.790114

WT, Nlrp3-/- and Asc-/- mice were inoculated with 1-2x107 CFU S. pneumoniae and genome-wide lung transcriptional responses were assessed 6 hours after infection. Univariate analysis (moderated t statistics) of the Nlrp3-/- and Asc-/- mouse samples yielded 20 gene expression profiles (Benjamini-Hochberg p < 0.05) that discriminate Nlrp3-/- and Asc-/- transcriptional responses in pneumonia. log2 FC, Log 2 transformed fold-change. t, t statistic. p, nominal p-value. BH p, Benjamini-Hochberg multiple comparison corrected p-value. B, Beta probability (distribution). Mouse Genome Informatics (MGI) official symbols are illustrated.

94092

Entrez ID

ILMN_2482494

Probe ID

log2 FC

Chapter 7

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

Figure 6: Validation of Csf2, Csf3 gene expression and lung protein levels of their respective products GM-CSF and G-CSF. WT, Nlrp3-/- and Asc -/- mice (n=6-8 per group) were inoculated with 1-2x107 CFU S. pneumoniae and sacrificed 6 hours later. qRT-PCR results for normalized lung expression of Csf2 (A) and Csf3 (C) genes (n=3-4 per group). GM-CSF (B) and G-CSF (D) were determined in lung homogenates using ELISA (n=6-9 per group). Data are expressed as mean Âą standard error of the mean; * p < 0.05, *** p < 0.001 vs WT mice determined with Mann-Whitney U test.

Discussion Inflammasomes are multiprotein complexes activated upon cellular infection or stress that play essential roles in the regulation of inflammation (5, 6). Many different inflammasomes have been discovered in recent years, each containing either a member of the NLR or PYHIN family. Whereas the function of NLRP3 is restricted to one NLR inflammasome, ASC is a common adaptor of several NLR inflammasomes. We here compared the roles of NLRP3 and ASC in innate immunity during respiratory tract infection caused by S.pneumoniae, the responsible pathogen in the majority of cases of community-acquired pneumonia (1, 2). Both NLRP3 and ASC were shown to contribute to antibacterial defense, as reflected by higher bacterial burdens at the primary site of infection in Nlrp3-/- and Asc-/mice, the role of ASC clearly was more prominent, indicated by enhanced bacterial dissemination to distant body sites and increased lethality in mice lacking ASC but not in mice deficient for NLRP3. Focusing on the early immune response, we identified 20 transcripts in the lungs of infected animals that discriminated between Nlrp3-/- and Asc-/- mice. Of these, secreted growth factors and cytokines encoded by Csf2 (GM-CSF), Csf3 (G-CSF) and Il17f, integrin molecules Itgae and Itgam (CD11b), the transcriptional regulator Foxp3 and the protein folding chaperone Unc45b were prominent features. Moreover, our data provide an attractive role for transcripts of as yet unknown function, namely, 4933427D14Rik, 2210407C18Rik, 145

Chapter 7

6330403K07Rik and A430084P05Rik, in the distinct pulmonary transcriptional responses to pneumococci linked to specific components of inflammasome macromolecular complexes. Three recent investigations studied the role of NLRP3 and/or ASC in pneumococcal pneumonia (7-9). These studies reported enhanced bacterial growth in the lungs of Nlrp3-/- mice after infection with a serotype 2 pneumococcus (D39) (7, 8), which is consistent with our current results using the same strain, but not after infection with a serotype 3 S. pneumoniae (9); dissemination of pneumococci to distant body sites was not described (7-9). One study directly compared the roles of NLRP3 and ASC after intranasal infection with S. pneumoniae D39 given at a slightly higher dose (5 x 107 CFU) than used here (1-2 x 107 CFU), revealing a reduced resistance of both Nlrp3-/- and Asc-/- mice relative to WT mice but no differences between the two inflammasome deficient mouse strains with regard to pulmonary bacterial growth and mortality, although lethality tended to be higher amongst Asc-/- mice (8). The in vivo role of NLRP3 and ASC in the inflammatory response to S. pneumoniae was only studied to a limited extent; these previous reports rather focused on mechanisms involved in the activation of the NLRP3 inflammasome and caspase-1 by S. pneumoniae in macrophages in vitro, all demonstrating an important role for the pneumococcal virulence factor pneumolysin (7-9). Our investigation partially confirms and extends these earlier data: we show that neither ASC nor NLRP3 impact on bacterial growth after infection with pneumolysin deficient S. pneumoniae; that after infection with WT S. pneumoniae ASC but not NLRP3 reduces bacterial dissemination and lethality after induction of pneumococcal pneumonia; and in addition, we describe the early innate immune response in the lungs of Nlrp3-/-, Asc-/- and WT mice in great detail, making use of whole genome mRNA profiling. The higher susceptibility of Asc-/- mice is not unexpected, since ASC can form inflammasome complexes with other NLRs, as well as with Absent in melanoma 2 (AIM2), a cytosolic DNA sensor, suggesting a broader role for ASC, relative to NLRP3, in the host response to invading pathogens. Indeed, AIM2 was shown to be involved in IL-1Î˛ secretion and pyroptosis in macrophages in response to S. pneumoniae (8). However, to our knowledge there are no in vivo studies with Aim2/mice infected with S. pneumoniae. Notably, at 48 hours after infection Nlrp3-/- and Asc-/- mice demonstrated higher lung cytokine levels than WT mice, which most likely was caused by the higher bacterial loads in inflammasome deficient mice and indicating that neither NLRP3 nor ASC are required for the induction of cytokines by S. pneumoniae in the respiratory tract. In accordance, Nlrp3-/- mice were reported to have equal or elevated lung cytokine levels at 48 hours after induction of pneumonia by a serotype 3 pneumococcus (9). The most evident difference between Asc-/- and Nrlp3-/- mice at 48 hours was the fact that the former mouse strain displayed less lung inflammation at this late stage of infection, as determined by an established quantitative histology score, in spite of similar bacterial loads. At present it is unclear whether this distinction between strains contributed to the clear difference in survival. Biochemical markers for distant organ injury did not differ between inflammasome deficient mice. Our analysis of the pulmonary transcriptome induced by S. pneumoniae in WT mice (Figure 4), encompassing 5346 transcripts, can be subdivided into functional 146

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

units of distinct biological processes. We observed an increase in abundance of transcripts involved in defense responses, cytokine regulation, chemotaxis and apoptosis (Figure 4B). A previous study reported a comparable transcriptional response in an aerosol-induced pneumococcal pneumonia model, although not extensively characterized through bioinformatics (22). A substantial proportion of the transcripts within the induced biological themes are involved in neutrophil-specific functions. Indeed, our data show that highly induced transcripts after pneumococcal infection serve as neutrophil recruiters, such as the chemoattractants Cxcl1 and Cxcl2 as well as Ptx3 (Pentraxin-3). Collectively, our comprehensive analysis of the pulmonary transcriptome at six hours post-pneumococcal infection suggests that while immune cell (innate, humoral and/or adaptive) related transcripts are maximal, epithelial cell transcripts as well as genes involved in mitochondrial function are negatively affected. In our attempt to dissect differences between Nlrp3-/- and Asc-/- mice during pneumococcal pneumonia we focused on the early immune response considering that the immediate inflammatory reaction in the airways is decisive for an adequate defense (2) and since bacterial loads were still similar at the 6-hour time point allowing an unbiased comparison between mouse strains not hampered by differences in the bacterial stimulus. We first determined cytokine and chemokine levels in whole lungs of infected mice and found a markedly impaired response in both Nlrp3-/- and Asc-/- mice. Importantly, relative to WT mice, levels of IL-1β were only significantly reduced in lungs of Nlrp3-/- mice, whereas TNFα was only lower in Asc-/- mice; the lung levels of CXCL1, CXCL2, CCL2 and IL-6 were significantly lower in both knockout strains. Of these differentially expressed mediators, especially TNFα plays an eminent role in host defense during pneumococcal pneumonia (23), but also IL-1β (24), IL-6 (25) and CCL2 (26) have been implicated in protective immunity. Thus, these differentially expressed mediators likely contributed to the impaired host defense of Asc-/- and Nlrp3 -/- mice. IL-17 was the only cytokine that was differentially expressed between Nlrp3-/- and Asc-/- mice, with lowest levels in Asc-/- mice, suggesting that reduced levels of IL-17 coincide with increased susceptibility of especially these mice in our model of pneumococcal pneumonia. Indeed, previous reports have documented that IL-17 is important for mucosal immunity and bacterial clearance, as well as colonization induced immunity against S. pneumonia (27, 28). The comparison of the lung transcriptional response between WT, Asc-/- and Nlrp3-/- mice during early pulmonary infection with S. pneumoniae yielded 56 gene transcripts that differed in abundance between the three genotypes (table 4). Among the genes that were differentially expressed between genotypes were Csf2 (encoding GM-CSF) and Csf3 (encoding G-CSF). Considering that both GMCSF and G-CSF have been implicated in protective immunity during pneumococcal pneumonia (20, 21), we expanded our analyses by determining protein levels of these growth factors in whole lung homogenates. We found significantly lower GMCSF concentrations in lungs of Asc-/ - mice relative to both WT and Nlrp3-/- mice, while mRNA levels (on microarrays and confirmed by RT-PCR) were only slightly 147

Chapter 7

lower in Asc-/- but nearly absent in Nlrp3-/- mice. This might point to an inability of Asc-/- mice to secrete the protein GM-CSF. Alternatively, the discrepancy between GM-CSF mRNA and protein data in Asc-/- and Nlrp3-/- mice may have been caused by altered mRNA expression before the 6-hour time point. Nonetheless, reduced GM-CSF levels in lungs of Asc-/- mice might explain their increased susceptibility at least in part, since prophylactic as well as therapeutic treatment with GM-CSF directly into the airways was demonstrated to improve survival and bacterial clearance in murine pneumococcal pneumonia (21). The protein levels of G-CSF were in line with mRNA expression data and lower in both Asc-/- and Nlrp3-/- mice. It would be interesting to evaluate if reconstitution of Asc-/- mice with GM-CSF or Asc-/- and Nlrp3-/- mice with G-CSF improves bacterial clearance and mortality rates and how this relates to their phenotype. Another gene of functional interest is integrin alpha M (CD11b). In the current study, integrin alpha M mRNA was lower in both knockout strains compared to WT mice, but significantly lower in Asc-/- than in Nlrp3-/- mice. Considering that CD11b deficient mice have impaired antibacterial defense against pneumococci (29), these data suggest a role for reduced integrin alpha M expression in the enhanced susceptibility of Asc-/- mice relative to Nlrp3-/mice during respiratory tract infection by S. pneumoniae. Another difference in gene expression in infected Nlrp3-/- and Asc-/- mice related to the adaptive immune response (in particular Il17F, Foxp3 and Itgae) and all of these genes were higher expressed in Nlrp3-/- mice. The adaptive immune response is intimately tied to inflammasome activation and several reports point to a role for the adaptive immune response in pneumococcal pneumonia (27, 30). For example, regulatory T cells, for which Foxp3 is a marker, are probably important to limit excessive inflammation during pneumococcal pneumonia (30). The product of Itgae, Integrin alpha E (CD103), for which epithelial E-cadherin is the ligand, is expressed on certain populations of intraepithelial T cells and some regulatory T cells as well as on dendritic cells (31, 32). A recent study demonstrated that the activation of Integrin alpha E on dendritic cells can protect from lethal pneumonia with S. pneumoniae (33). Interestingly, the expression of Integrin alpha E on dendritic cells was reported to be induced by GM-CSF in an infection model (34), which is in line with the impaired levels of this growth factor detected here in Asc-/- mice. In addition, it was very recently reported that Asc-/- dendritic cells have impaired antigen presenting capacity and that Asc-/- mice have impaired lymphocyte migration independent of caspase-1 activation, pointing to an inflammasome independent role for ASC in the induction of the adaptive host response (35). As a preliminary analysis of differences in inflammatory cell populations we evaluated co-clustering of genes that were significantly different between genotypes by use of the Immgen.org database. Dendritic cells were predicted to be significantly different between genotypes Nlrp3-/- and WT mice (p= 7,80E-03). This might suggest that dendritic cells compensate the lack of NLRP3 inflammasome dependent functions to some extent in Nlrp3-/- mice but not in Asc-/- mice. It remains to be determined if migration of dendritic cells to the lungs of Asc-/- mice is indeed impaired and if this contributes to their susceptible phenotype, for instance in adoptive transfer 148

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

experiments. In conclusion, we here demonstrate different susceptibilities of Nlrp3-/- and Asc-/ - mice after infection with S. pneumoniae via the airways reflected by enhanced bacterial dissemination and lethality in the latter mouse strain. While the vast majority of studies focused on the role of inflammasomes in caspase-1 activation, the release of IL-1β and IL-18 and pyroptosis, we here performed a detailed analysis of the early inflammatory response in the lungs by whole genome expression arrays. We identified several genes influenced by NLRP3 and ASC not classically linked to inflammasome activation and in addition found mediators that were differentially expressed in Nlrp3-/- and Asc-/ - mice. Of these, IL-17, GM-CSF and integrin alpha M were of particular interest considering that they were significantly attenuated in Asc/relative to Nlrp3-/- mice, which in light of their protective role during pneumococcal pneumonia may at least in part explain the relative hypersusceptibility of Asc-/ mice. In addition, Asc-/ - mice displayed a reduced expression of a number of genes involved in the adaptive immune response. Taken together with previous studies from our laboratory that found only limited roles for inflammasome dependent cytokines IL-1β and IL-18 in host defense against pneumococcal pneumonia (24, 36), the current data suggest that especially ASC impacts on the immune response to pneumococci by mechanisms not directly linked to the release of IL-1β or IL-18. It remains to be determined to which extent inflammasome independent functions by ASC, such as recently described (35), contribute to its role in host defense against pneumonia caused by S. pneumoniae.

Acknowledgments Fayaz S. Sutterwala and Richard A. Flavell kindly provided us with the Nlrp3and Asc-/ - mice. We thank Joost Daalhuisen and Marieke ten Brink for expert technical assistance and James C. Paton for providing the pneumolysin deficient S. pneumoniae D39.

149

Chapter 7

References

1. Mandell LA, Wunderink RG, Anzueto A, Bartlett JG, Campbell GD, Dean NC, Dowell SF, File TM, Jr., Musher DM, Niederman MS, et al. Infectious Diseases Society of America/American Thoracic Society Consensus Guidelines on the Management of Community-Acquired Pneumonia in Adults. Clin Infect Dis 2007;44 Suppl 2:S27-S72.

2. van der Poll T, Opal SM. Pathogenesis, Treatment, and Prevention of Pneumococcal Pneumonia. Lancet 2009;374:1543-1556.

3. Zahar JR, Timsit JF, Garrouste-Org, Francais A, Vesin A, Descorps-Declere A, Dubois Y, Souweine B, Haouache H, Goldgran-Toledano D, et al. Outcomes in Severe Sepsis and Patients With Septic Shock: Pathogen Species and Infection Sites Are Not Associated With Mortality. Crit Care Med 2011;39:1886-1895.

4. Ramirez JA, Anzueto AR. Changing Needs of Community-Acquired Pneumonia. J Antimicrob Chemother 2011;66 Suppl 3:iii3-iii9.

5. Schroder K, Tschopp J. The Inflammasomes. Cell 2010;140:821-832.

6. Strowig T, Henao-Mejia J, Elinav E, Flavell R. Inflammasomes in Health and Disease. Nature 2012;481:278-286.

7. McNeela EA, Burke A, Neill DR, Baxter C, Fernandes VE, Ferreira D, Smeaton S, El-Rachkidy R, McLoughlin RM, Mori A, et al. Pneumolysin Activates the NLRP3 Inflammasome and Promotes Proinflammatory Cytokines Independently of TLR4. PLoS Pathog 2010;6:e1001191.

8. Fang R, Tsuchiya K, Kawamura I, Shen Y, Hara H, Sakai S, Yamamoto T, Fernandes-Alnemri T, Yang R, Hernandez-Cuellar E, et al. Critical Roles of ASC Inflammasomes in Caspase-1 Activation and Host Innate Resistance to Streptococcus Pneumoniae Infection. J Immunol 2011;187:4890-4899.

9. Witzenrath M, Pache F, Lorenz D, Koppe U, Gutbier B, Tabeling C, Reppe K, Meixenberger K, Dorhoi A, Ma J, et al. The NLRP3 Inflammasome Is Differentially Activated by Pneumolysin Variants and Contributes to Host Defense in Pneumococcal Pneumonia. J Immunol 2011;187:434-440.

10. Sutterwala FS, Ogura Y, Szczepanik M, Lara-Tejero M, Lichtenberger GS, Grant EP, Bertin J, Coyle AJ, Galan JE, Askenase PW, et al. Critical Role for NALP3/CIAS1/Cryopyrin in Innate and Adaptive Immunity Through Its Regulation of Caspase-1. Immunity 2006;24:317-327.

11. Dessing MC, Florquin S, Paton JC, van der Poll T. Toll-Like Receptor 2 Contributes to Antibacterial Defence Against Pneumolysin-Deficient Pneumococci. Cell Microbiol 2008;10:237-246.

12. Hoogendijk AJ, Roelofs JJ, Duitman J, van Lieshout MH, Blok DC, van der Poll T, Wieland CW. R-Roscovitine Reduces Lung Inflammation Induced by Lipoteichoic Acid and Streptococcus Pneumoniae. Mol Med 2012;18:1086-1095.

13. Dunning MJ, Smith ML, Ritchie ME, Tavare S. Beadarray: R Classes and Methods for Illumina Bead-Based Data. Bioinformatics 2007;23:2183-2184.

14. Reimers M, Carey VJ. Bioconductor: an Open Source Framework for Bioinformatics and Computational Biology. Methods Enzymol 2006;411:119-134.

150

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

15. Ruijter JM, Pfaffl MW, Zhao S, Spiess AN, Boggy G, Blom J, Rutledge RG, Sisti D, Lievens A, De PK, et al. Evaluation of QPCR Curve Analysis Methods for Reliable Biomarker Discovery: Bias, Resolution, Precision, and Implications. Methods 2012.

16. Smyth GK. Linear Models and Empirical Bayes Methods for Assessing Differential Expression in Microarray Experiments. Stat Appl Genet Mol Biol 2004;3:Article3.

17. Huang dW, Sherman BT, Lempicki RA. Systematic and Integrative Analysis of Large Gene Lists Using DAVID Bioinformatics Resources. Nat Protoc 2009;4:44-57.

18. Koppe U, Suttorp N, Opitz B. Recognition of Streptococcus Pneumoniae by the Innate Immune System. Cell Microbiol 2012;14:460-466.

19. Giebelen IA, Leendertse M, Florquin S, van der Poll T. Stimulation of Acetylcholine Receptors Impairs Host Defence During Pneumococcal Pneumonia. Eur Respir J 2009;33:375-381.

20. Knapp S, Hareng L, Rijneveld AW, Bresser P, van der Zee JS, Florquin S, Hartung T, van der Poll T. Activation of Neutrophils and Inhibition of the Proinflammatory Cytokine Response by Endogenous Granulocyte Colony-Stimulating Factor in Murine Pneumococcal Pneumonia. J Infect Dis 2004;189:1506-1515.

21. Steinwede K, Tempelhof O, Bolte K, Maus R, Bohling J, Ueberberg B, Langer F, Christman JW, Paton JC, Ask K, et al. Local Delivery of GM-CSF Protects Mice From Lethal Pneumococcal Pneumonia. J Immunol 2011;187:5346-5356.

22. Evans SE, Tuvim MJ, Zhang J, Larson DT, Garcia CD, Martinez-Pro S, Coombes KR, Dickey BF. Host Lung Gene Expression Patterns Predict Infectious Etiology in a Mouse Model of Pneumonia. Respir Res 2010;11:101.

23. van der Poll T, Keogh CV, Buurman WA, Lowry SF. Passive Immunization Against Tumor Necrosis Factor-Alpha Impairs Host Defense During Pneumococcal Pneumonia in Mice. Am J Respir Crit Care Med 1997;155:603-608.

24. Rijneveld AW, Florquin S, Branger J, Speelman P, Van Deventer SJ, van der Poll T. TNFAlpha Compensates for the Impaired Host Defense of IL-1 Type I Receptor-Deficient Mice During Pneumococcal Pneumonia. J Immunol 2001;167:5240-5246.

25. van der Poll T, Keogh CV, Guirao X, Buurman WA, Kopf M, Lowry SF. Interleukin-6 GeneDeficient Mice Show Impaired Defense Against Pneumococcal Pneumonia. J Infect Dis 1997;176:439-444.

26. Winter C, Taut K, Srivastava M, Langer F, Mack M, Briles DE, Paton JC, Maus R, Welte T, Gunn MD, et al. Lung-Specific Overexpression of CC Chemokine Ligand (CCL) 2 Enhances the Host Defense to Streptococcus Pneumoniae Infection in Mice: Role of the CCL2-CCR2 Axis. J Immunol 2007;178:5828-5838.

27. Zhang Z, Clarke TB, Weiser JN. Cellular Effectors Mediating Th17-Dependent Clearance of Pneumococcal Colonization in Mice. J Clin Invest 2009;119:1899-1909.

28. Graeber KE, Olsen NJ. Th17 Cell Cytokine Secretion Profile in Host Defense and Autoimmunity. Inflamm Res 2012;61:87-96.

151

Chapter 7

29. Rijneveld AW, de Vos AF, Florquin S, Verbeek JS, van der Poll T. CD11b Limits Bacterial Outgrowth and Dissemination During Murine Pneumococcal Pneumonia. J Infect Dis 2005;191:1755-1760.

30. Neill DR, Fernandes VE, Wisby L, Haynes AR, Ferreira DM, Laher A, Strickland N, Gordon SB, Denny P, Kadioglu A, et al. T Regulatory Cells Control Susceptibility to Invasive Pneumococcal Pneumonia in Mice. PLoS Pathog 2012;8:e1002660.

31. Van den Bossche J, Malissen B, Mantovani A, De BP, Van Ginderachter JA. Regulation and Function of the E-Cadherin/Catenin Complex in Cells of the Monocyte-Macrophage Lineage and DCs. Blood 2012;119:1623-1633.

32. Leepiyasakulchai C, Ignatowicz L, Pawlowski A, Kallenius G, Skold M. Failure to Recruit Anti-Inflammatory CD103+ Dendritic Cells and a Diminished CD4+ Foxp3+ Regulatory T Cell Pool in Mice That Display Excessive Lung Inflammation and Increased Susceptibility to Mycobacterium Tuberculosis. Infect Immun 2012;80:1128-1139.

33. Ivanov S, Fontaine J, Paget C, Macho FE, Van ML, Renneson J, Maillet I, Wolf NM, Rial A, Leger H, et al. Key Role for Respiratory CD103(+) Dendritic Cells, IFN-Gamma, and IL-17 in Protection Against Streptococcus Pneumoniae Infection in Response to AlphaGalactosylceramide. J Infect Dis 2012;206:723-734.

34. Zhan Y, Carrington EM, van NA, Bedoui S, Seah S, Xu Y, Wang N, Mintern JD, Villadangos JA, Wicks IP, et al. GM-CSF Increases Cross-Presentation and CD103 Expression by Mouse CD8(+) Spleen Dendritic Cells. Eur J Immunol 2011;41:2585-2595.

35. Ippagunta SK, Malireddi RK, Shaw PJ, Neale GA, Walle LV, Green DR, Fukui Y, Lamkanfi M, Kanneganti TD. The Inflammasome Adaptor ASC Regulates the Function of Adaptive Immune Cells by Controlling Dock2-Mediated Rac Activation and Actin Polymerization. Nat Immunol 2011;12:1010-1016.

36. Lauw FN, Branger J, Florquin S, Speelman P, Van Deventer SJ, Akira S, van der Poll T. IL-18 Improves the Early Antimicrobial Host Response to Pneumococcal Pneumonia. J Immunol 2002;168:372-378.

152

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

Supplemental methods Assays for clinical chemistry Plasma levels of urea, lactate dehydrogenase (LDH), aspartate aminotransferase (AST) and alanine transaminase (ALT) were measured with kits from Sigma (St. Louis, MO), using a Hittachi analyzer (Boehringer Mannheim, Mannheim, Germany). Processing of RNA & Bioinformatics Total RNA was isolated from lung homogenates using the Nucleospin RNA II kit (Machery-Nagel, Duren, Germany). Yield and purity (260nm:280nm) were determined by Nanodrop ND-1000. The integrity (RIN>7.0) of the re-suspended total RNA was determined by using the RNA Nano Chip Kit on the Bioanalyzer 2100 and the 2100 Expert software (Agilent). The Illumina TotalPrep-96 RNA Amplification Kit was used to generate biotin labeled (biotin-16-UTP) amplified cRNA starting from 200ng total RNA. A total of 750ng biotinylated cRNA was hybridized onto the Illumina MouseRef-8v2 Expression BeadChip. The samples were scanned using the Illumina iScan array scanner. Preparation of cRNA, chip hybridization, washing, staining and scanning were carried out at ServiceXS (Leiden, the Netherlands). The raw scan data were read using the beadarray package (version 1.12.1) (1), available through Bioconductor (2) , using the R statistical package (version 2.13.2; R Foundation for Statistical Computing, Vienna, Austria). Estimated background was subtracted from the foreground for each bead. For replicate beads, outliers greater than 3 median absolute deviations from the median were removed and the average signal was calculated for the remaining intensities. For each probe a detection score was calculated by comparing its average signal with the summarized values for the negative control probes. Resulting data were neqc normalized (3). Quality control was performed both on bead level and on bead summary data. The arrayQualityMetrics package v2.6.0 (4) was used for further quality assessment of the normalized bead summary data. Probes were re-annotated using the package illuminaMousev2.db from Bioconductor (5). Annotation quality was assessed by means of the illuminaMousev2PROBEQUALITY call of the illuminaMousev2. db package as described in (5). Probes that showed â&#x20AC;&#x153;no matchâ&#x20AC;? to transcript or genomic regions were filtered out of subsequent differential gene expression analysis. Differential gene expression analysis was performed by means of the limma package (version 3.8.3), which implements linear models for microarray data (Smyth GK, 2004). A factorial design matrix was constructed to indicate the RNA sample identity of each array, that is, uninfected wildtype (WT), Nlrp3-/-, Asc-/- and D39 infected wildtype, Nlrp3-/-, Asc-/- mouse lung RNA samples. Next, a contrast matrix was defined to specify the comparisons to be made between the RNA samples: a. genotype effect (uninfected samples only) Nlrp3_geno=(Nlrp3KO-WT), Asc_geno=(AscKO-WT) b. infection (infected samples only) Asc_infection=(AscKO-WT), Nlrp3_infection=(Nlrp3KO-WT), Asc_Nlrp3_ difference=((Nlrp3KO-WT)-(AscKO-WT)) The lmFit and eBayes functions coupled with array weights (implemented in 153

Chapter 7

limma) were used to statistically assess differential expression. Empirical Bayesian methods moderate the standard errors of the estimated log-fold changes resulting in more stable inferences and improved power (Smyth GK, 2004). Thus, p-values were obtained from moderated t statistics or f statistics, which are then adjusted for multiple comparisons with Benjamini and Hochbergâ&#x20AC;&#x2122;s (BH) method to control the false discovery rate. The toptable function was used to generate annotated outputs for which the most significant probe ID was presented. All non-normalized and normalized data are available at the gene expression omnibus of NCBI (GEO) with accession number GSE42464. Differentially expressed genes as defined by multiple-test corrected p < 0.05 were analyzed for enrichment of functional annotations using the Database for Annotation, Visualization and Integrated Discovery (DAVID) v6.7 (reference 17 in main manuscript). We analyzed up- and down-regulated gene expression profiles defined by positive and negative log2 foldchanges, respectively. All analyses were performed using default parameters. In order to assess enrichment for cellspecific gene co-expression signatures we used the coexpression atlas of the Immunological Genome database (Immgen.org) available through the Toppgene suite (http://toppgene.cchmc.org). Genes listed in table S6 were analyzed using default parameters. Significance for both DAVID and Toppgene outputs were defined by bonferroni p < 0.05.

References

1. Dunning MJ, Smith ML, Ritchie ME, Tavare S. Beadarray: R Classes and Methods for Illumina Bead-Based Data. Bioinformatics 2007;23:2183-2184.

2. Reimers M, Carey VJ. Bioconductor: an Open Source Framework for Bioinformatics and Computational Biology. Methods Enzymol 2006;411:119-134.

3. Shi W, Oshlack A, Smyth GK. Optimizing the Noise Versus Bias Trade-Off for Illumina Whole Genome Expression BeadChips. Nucleic Acids Res 2010;38:e204.

4. Kauffmann A, Rayner TF, Parkinson H, Kapushesky M, Lukk M, Brazma A, Huber W. Importing ArrayExpress Datasets into R/Bioconductor. Bioinformatics 2009;25:2092-2094.

5. Barbosa-Morais NL, Dunning MJ, Samarajiwa SA, Darot JF, Ritchie ME, Lynch AG, Tavare S. A Re-Annotation Pipeline for Illumina BeadArrays: Improving the Interpretation of Gene Expression Data. Nucleic Acids Res 2010;38:e17.

154

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

Supplemental tables Table S1: Plasma levels of chemokines and cytokines 6 hours after induction of pneumonia by S. pneumoniae.

Plasma

Nlrp3-/-

Asc-/-

TNF-Îą

27 (4)

16 (3)

19 (4)

IL-6

1038 (140)

416 (149)*

138 (35)**

IL-10

7 (1)

6 (1)

13 (5)

IL-12

12 (4)

9 (2)

22 (12)

CCL2

682 (118)

198 (48)**

131 (35)**

WT, Nlrp3 -/- and Asc-/- mice were inoculated with 1-2x107 CFU S. pneumoniae and sacrificed at 48 hours after infection in WT, Nlrp3-/- and Asc-/- mice. Plasma cytokine and chemokine levels are presented in pg/ml. Data are expressed as mean (SE). * p < 0.05, **p <0.01 compared to WT mice determined with Mann-Whitney U test. Table S2: Plasma levels of chemokines and cytokines 48 hours after induction of pneumonia by S. pneumoniae.

Plasma

Nlrp3-/-

Asc-/-

TNF-Îą

4 (1)

70 (47)*

168 (95)

1020 (589)

IL-6

22 (12)

955 (632)**

1947 (1265)

3241 (1556)

IL-10

38 (28)

12 (2)

14 (19)

322 (227)

IL-12

104 (94)

14 (4)

27 (17)

15 (4)

CCL2

42 (10)

255 (138)

2093 (1214)

3805 (1733)

WT, Nlrp3-/- and Asc-/- mice were inoculated with 1-2x107 CFU S. pneumoniae and sacrificed 48 hours later. Plasma cytokine and chemokine levels are presented in pg/ml. Data are mean (SE) of 6-9 mice per group.* p < 0.05, ** p < 0.01 vs WT mice.

155

Chapter 7 Table S3: Primer sequences used for qRT-PCR validation of transcripts

Primer ID

Sequence 5’→3’

Gdpd3_Forward

ATGGCGATCGTTCTGGATAC

Gdpd3_Reverse

AATGGTAGCCGACAGTTGGT

Csf2_Forward

CTTGAACATGACAGCCAGCTA

Csf2_Reverse

TCATTTTTGGCCTGGTTTTT

Csf3_Forward

GTCTCCTGCAGGCTCTATCG

Csf3_Reverse

CTGGAAGGCAGAAGTGAAGG

Nlrp3_Forward

AGGACTCAGGCTCCTCTGTG

Nlrp3_Reverse

GATCATTGTTGCCCAGGTTC

Pycard_Forward

CCAGTGTCCCTGCTCAGAGT

Pycard_Reverse

TGTCTTGGCTGGTGGTCTCT

Unc45b_Forward

AGAACCATGACCAGCTACGG

Unc45b_Reverse

GCGTTCTGCAGTTTGTGGT

Mid1_Forward

CCGTCACTACTGGGAAGTGG

Mid1_Reverse

ATGGAGCCGTTGTCGTAGTC

156

EntrezID

74477

216799

12981

94092

12985

211623

78354

66628

217012

22698

16407

30939

98365

16407

216850

Probe ID

ILMN_1245850

ILMN_1237471

ILMN_2749412

ILMN_2482494

ILMN_1217948

ILMN_2710159

ILMN_1226767

ILMN_2979430

ILMN_2732747

ILMN_1241611

ILMN_2699898

ILMN_2809167

ILMN_2663249

ILMN_1217629

ILMN_3095624

Nlrp3-/- v WT

log2 FC

1.787

-2.031

-1.513

-2.160

-2.181

Kdm6b

Itgae

Slamf9

Pttg1

Itgae

Zfp39

Unc45b

Thg1l

-1.490

1.421

1.337

-1.399

1.454

-1.133

-1.159

1.591

2210407C18Rik 1.593

Plac9

Csf3

Trim16

Csf2

Nlrp3

4933427D14Rik -2.636

Symbol

-5.699

5.949

6.093

-6.131

6.267

-6.412

-6.456

6.498

7.587

7.948

-8.219

-8.879

-9.070

-10.176

-11.376

4.91E-06

2.54E-06

1.75E-06

1.58E-06

1.11E-06

7.63E-07

6.82E-07

6.13E-07

3.99E-08

1.67E-08

8.78E-09

1.91E-09

1.24E-09

1.12E-10

9.84E-12

P.Value

Table S4: Differential transcriptional response between WT and Nlrp3-/- mice after pulmonary infection.

0.00670

0.00368

0.00270

0.00262

0.00215

0.00161

0.00158

0.00012

5.53E-05

3.39E-05

1.11E-05

9.58E-06

1.30E-06

2.28E-07

adj.P.Val

3.880

4.437

4.754

4.836

5.132

5.446

5.539

5.629

7.861

8.556

9.064

10.247

10.576

12.358

14.073

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

157

EntrezID

27078

16319

21940

434760

14778

26382

24050

103712

210146

16992

66039

23845

103268

18476

18729

73914

Probe ID

ILMN_2713969

158

ILMN_1216213

ILMN_1233589

ILMN_2917647

ILMN_2715546

ILMN_2737903

ILMN_1218471

ILMN_2594139

ILMN_2715466

ILMN_1229804

ILMN_2693858

ILMN_3121522

ILMN_2807750

ILMN_2640971

ILMN_2691049

ILMN_1253972

1.507

0.917

-0.759

1.041

0.884

0.866

0.786

Nlrp3-/- v WT

log2 FC

Irak3

Pira6

Pafah1b3

Cep57l1

Clec5a

D14Ertd449e

Lta

Irgq

-0.761

-1.294

0.809

0.754

-0.998

1.433

1.097

1.055

6330403K07Rik -1.161

Sept3

Fgd2

Gpx3

Rhox2d

Cd27

Incenp

B9d1

Symbol

-4.725

-4.819

4.842

-4.903

4.918

4.941

4.961

-5.010

5.052

5.067

-5.112

5.140

5.280

5.365

5.604

6.56E-05

5.11E-05

4.80E-05

4.08E-05

3.92E-05

3.68E-05

3.49E-05

3.06E-05

2.74E-05

2.63E-05

2.33E-05

2.17E-05

1.49E-05

1.19E-05

6.31E-06

P.Value

0.04607

0.03698

0.03587

0.03261

0.03243

0.03161

0.03112

0.02840

0.02647

0.02457

0.02392

0.01727

0.01451

0.00812

adj.P.Val

1.662

1.878

1.932

2.071

2.106

2.159

2.205

2.318

2.413

2.448

2.552

2.615

2.935

3.128

3.668

Chapter 7

238037

14999

319172

13388

ILMN_2760057

ILMN_1244977

ILMN_2836654

ILMN_2721188

Dll1

Hist1h2ab

H2-DMb1

BC068281

Symbol

-1.013

1.243

0.896

-0.879

Nlrp3-/- v WT

log2 FC

-4.660

4.672

4.678

-4.711

7.82E-05

7.57E-05

7.44E-05

6.81E-05

P.Value

0.04897

0.04870

0.04642

adj.P.Val

1.511

1.540

1.554

1.630

WT, Nlrp3 -/- and Asc -/- (Pycard KO) mice were inoculated with 2x107 CFU S. pneumoniae and genome-wide lung transcriptional responses were assessed 6 hours after infection. Univariate analysis (moderated t statistics) of the WT and Nlrp3-/- mouse samples yielded 35 gene expression profiles (BenjaminiHochberg p < 0.05) that discriminate Nlrp3 -/- transcriptional response from WT in pneumonia. log2 FC, Log 2 transformed fold-change. t, t statistic. p, nominal p-value. BH p, Benjamini-Hochberg multiple comparison corrected p-value. B, Beta probability.

EntrezID

Probe ID

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

159

EntrezID

211623

50501

68616

12985

16409

66824

18054

101490

66039

16878

16415

26410

21983

56369

12796

26904

Probe ID

ILMN_2710159

160

ILMN_3147259

ILMN_2893879

ILMN_1217948

ILMN_2696017

ILMN_2936476

ILMN_1228832

ILMN_2785512

ILMN_2693858

ILMN_3137291

ILMN_1258735

ILMN_1245924

ILMN_1250696

ILMN_1234781

ILMN_2766604

ILMN_2978617

Sh2d1b1

Camp

Apip

Tpbg

Map3k8

Itgb2l

Lif

D14Ertd449e

Inpp5f

Ngp

Asc

Itgam

Csf3

Gdpd3

Prok2

Plac9

Symbol

1.064

-2.267

0.879

-0.833

-0.871

-1.821

-1.309

1.764

0.831

-2.680

-1.560

-1.822

-1.764

4.625

-2.370

1.790

logFC

5.231

-5.295

5.435

-5.483

-5.742

-6.092

-6.203

7.075

7.147

-7.640

-7.655

-7.777

-8.341

8.440

-8.897

9.301

1.70E-05

1.43E-05

9.89E-06

8.70E-06

4.39E-06

1.75E-06

1.31E-06

1.42E-07

1.19E-07

3.51E-08

3.39E-08

2.52E-08

6.60E-09

5.23E-09

1.83E-09

7.40E-10

P.Value

Table S5: Differential transcriptional response between WT and Asc-/- mice after pulmonary infection.

0.0232

0.0208

0.0153

0.0144

0.0078

0.0034

0.0028

0.0004

0.0003

0.0001

3.82E-05

2.12E-05

1.72E-05

adj.P.Val

2.858

3.006

3.327

3.438

4.026

4.810

5.057

6.924

7.072

8.070

8.100

8.340

9.417

9.601

10.426

11.126

Chapter 7

19730

380924

99138

19288

83397

11504

67676

18033

240327

ILMN_2689119

ILMN_3161834

ILMN_2676066

ILMN_2662802

ILMN_2624622

ILMN_2761082

ILMN_2726308

ILMN_2592476

ILMN_2904001

Gm4951

Nfkb1

Rpp21

Adamts1

Akap12

Ptx3

Stard7

Olfm4

Ralgds

Slc27a4

0.906

-0.495

0.515

-1.213

-0.697

-1.087

-0.517

-1.162

-1.215

-1.026

-4.852 4.792

4.853

-4.860

-4.871

-4.942

-5.055

-5.167

-5.169

-5.175

5.48E-05

4.67E-05

4.57E-05

4.44E-05

3.68E-05

2.72E-05

2.02E-05

2.01E-05

1.97E-05

0.0471

0.0416

0.0387

0.0300

0.0234

1.842

1.982

1.983

2.000

2.025

2.190

2.451

2.711

2.716

2.730

WT, Nlrp3-/- and Asc-/- mice were inoculated with 2x107 CFU S. pneumoniae and genome-wide lung transcriptional responses were assessed 6 hours after infection. Univariate analysis (moderated t statistics) of the WT and Asc-/- mouse samples yielded 26 gene expression profiles (Benjamini-Hochberg p < 0.05) that discriminate Asc-/- transcriptional response from WT in pneumonia. log2 FC, Log 2 transformed fold-change. t, t statistic. p, nominal p-value. BH p, Benjamini-Hochberg multiple comparison corrected p-value. B, Beta probability.

26569

ILMN_2709355

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

161

Chapter 7 Table S6 Nlrp3 -/- vs WT mice, enrichment for cell-specific gene co-expression.

Coexpression atlas ID

Name

bonferroni p

Genes

GSM538231_500

Myeloid Cells, DC.103+11b-.Lu, CD11chi CD11b low CD103+ MHCII+ SiglecF-, Lung, avg-3

7,80E-03

Fgd2, Slamf9, Sept3, Clec5a, Itgae, Gpx3, Nlrp3

Toppgene suite (http://toppgene.cchmc.org) derived enrichment for cell-specific gene co-expression characteristics available through the Immunological Genome database (Immgen.org). Differentially expressed genes between WT and Nlrp3-/- mouse samples suggest lung dendritic cells (DC) sorted by CD11c hi CD11b low CD103+ MHCII+ SiglecF- markers as putatively prominent cell-type. Coexpressed genes are Fgd2, Slamf9, Sept3, Clec5a, Itgae, Gpx3, Nlrp3.

Supplemental figures

Figure S1: Pulmonary clearance of pneumolysin deficient S. pneumoniae D39 is not impaired in Nlrp3-/- and Asc-/- mice. T, Nlrp3 -/- and Asc-/- mice were inoculated with 2x107 CFU isogenic pneumolysin deficient S. pneumoniae D39 (Ply-) and sacrificed 48 hours after infection. Bacterial loads in lungs of WT (white), Nlrp3-/- (light grey) and Asc-/- mice (dark grey) (n=6-9 per group). Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation.

Figure S2: Position of differentially expressed genes in uninfected Nlrp3-/- mice on chromosome 11 in relation to the Nlrp3 gene.

162

NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia

Figure S3: Validation of transcripts differentially expressed on arrays by quantitative reversetranscriptase PCR (qRT-PCR). WT, Nlrp3-/- and Asc-/- mice were inoculated with 2x107 CFU S. pneumoniae and genome-wide lung transcriptional responses were assessed 6 hours after infection. qRT-PCR results for normalized expression of Nlrp3, Asc, Gdpd3 and Unc45b. Data are expressed as mean Âą standard error of the mean, * p < 0.05 vs WT mice determined with Mann-Whitney U test.

163

164

Chapter 8 ASC and NLRP3 impair host defense during lethal pneumonia caused by highly virulent serotype 3 Streptococcus pneumoniae Submitted for publication

Miriam H.P. van Lieshout 1,2 * Alex F. de Vos 1,2 * Mark C. Dessing 3 Alexander P.N.A. de Porto 1,2 Onno J. de Boer 3 Regina de Beer 1,2 Sanne Terpstra 1,2 Sandrine Florquin 3 Cornelis van â&#x20AC;&#x2122;t Veer 1,2 Tom van der Poll 1,2,4 * These authors contributed equally. Academic Medical Center, University of Amsterdam, the Netherlands: 1 Center of Infection and Immunity Amsterdam 2 Center for Experimental and Molecular Medicine 3 Department of Pathology 4 Division of Infectious Diseases

Chapter 8

Abstract Streptococcus (S.) pneumoniae is the most common cause of community-acquired pneumonia. The Nod-like receptor family pyrin domain containing 3 (NLRP3) inflammasome, consisting of NLRP3, ASC (the adaptor apoptosis-associated speck-like protein containing a CARD) and caspase-1, has been implicated in protective immunity in bacterial infections, amongst other during pneumonia induced by high doses of a moderately virulent strain of S. pneumoniae. Here we investigated the role of the NLRP3 inflammasome in the host response during airway infection with a low dose of a highly virulent gradually growing serotype 3 S. pneumoniae in a model that more closely mimics human disease. Mice were euthanized at predefined endpoints for analysis or observed in survival studies. In additional studies Tlr2-/-/Tlr4-/- mice and Myd88-/- mice, incapable of Toll-like receptor signaling, were studied. Surprisingly, and in stark contrast with existing literature, both Nlrp3-/- and Asc-/- mice showed a strongly improved host defense, as reflected by a markedly reduced mortality accompanied by diminished bacterial growth and dissemination. Lung inflammation and pathology were attenuated in Nlrp3-/- and Asc-/- mice during the late stages of the infection. Host defense was unaltered in Tlr2-/-/Tlr4-/- mice and Myd88-/- mice, although Myd88-/- mice demonstrated attenuated lung inflammation in the presence of high pneumococcal burdens. These results show that the NLRP3 inflammasome impairs host defense during lethal pneumonia caused by highly virulent serotype 3 S. pneumoniae. Our findings in this clinically relevant model of a highly prevalent human infection challenge the current paradigm that proximal innate detection systems, including the NLRP3 inflammasome and Toll-like receptors (TLRs), are indispensable for an adequate host immune response against bacteria.

166

ASC and NLRP3 impair host defense during lethal pneumonia caused by highly virulent serotype 3 Streptococcus pneumoniae

Introduction Community-acquired pneumonia is an important cause of mortality world-wide. The most frequently isolated causative pathogen is Streptococcus (S.) pneumoniae, a Gram-positive diplococcus of which over 90 serotypes have been identified (13). Depending on health care settings, the mortality rate associated with pneumococcal pneumonia ranges from 6 to > 40% (2, 3). The health burden caused by pneumococcal pneumonia is further aggravated by the increasing incidence of antibiotic resistance in S. pneumoniae (4). In recent years, the importance of large multi-protein complexes called “inflammasomes” for the antimicrobial response has become apparent (5, 6). Inflammasomes consist of a cytosolic sensor protein and caspase-1. The NLRP3 (NLR family pyrin domain containing 3) inflammasome uses the adaptor protein ASC (apoptosis-associated speck-like protein containing a CARD) to recruit and bind caspase-1, which upon activation cleaves immature pro-IL-1β and pro-IL-18 to the mature secreted forms (5, 6). NLRP3-mediated secretion of IL-1β and IL-18 is a tightly regulated process that requires at least two signals (5, 6). In response to bacterial infection, the first signal occurs when pathogen-associated molecular patterns (PAMPs) derived from the invading pathogen stimulate innate immune cells through activation of pattern recognition receptors, most notably members of the Toll-like receptor (TLR) family, resulting in the transcription and translation of the pro-forms of IL-1β and IL-18. The second signal requires the assembly and activation of the NLRP3 inflammasome resulting in the formation of active caspase-1 by an ASC dependent mechanism (5, 6). Activation of the NLRP3 inflammasome can also induce pyroptosis, which, in contrast to caspase-3-mediated apoptotic cell death, is associated with the release of pro-inflammatory mediators (7). Recent investigations have revealed an important protective role for the NLRP3 inflammasome during pneumococcal pneumonia (8-11) as indicated by enhanced bacterial growth in the lungs of NLRP3 deficient (Nlrp3-/-) and Asc-/- mice after infection with a serotype 2 pneumococcus (D39) (8-10). One study examined the contribution of NLRP3 to host defense during serotype 3 S. pneumoniae pneumonia, reporting a more modest protective role (11). S. pneumoniae can activate the NLRP3 inflammasome thorugh its crucial virulence factor pneumolysin (10, 12). Besides NLRP3 and ASC, several TLRs have been reported to contribute to the protective immunity during pneumococcal infection. TLR2, that detects lipoteichoic acid, a constituent of the pneumococcal cell wall (13, 14), played a modest role in the cytokine response during pneumococcal pneumonia after infection with a serotype 3 pneumococcus (15), while TLR4 limited the growth of pneumococci during nasopharyngeal colonization (16) and lower respiratory infection (17), at least in part by recognition of pneumolysin (16, 18). Likewise, Tlr9-/- mice showed enhanced bacterial growth and dissemination after induction of pneumococcal pneumonia (19). The intracellular adaptor protein myeloid differentiation primary response gene 88 (MyD88) is a critical component in TLR2, TLR4 and TLR9 signaling, and in accordance Myd88-/- mice showed a profoundly enhanced bacterial outgrowth and a strongly reduced survival after intranasal infection with serotype 2, 4 or 19F S. pneumoniae strains (20, 21). 167

Chapter 8

The serotype of the infecting S. pneumoniae is an important risk factor for the occurrence of invasive disease and mortality (22). In humans, infections caused by serotype 3 pneumococci are associated with a complicated course and an increased risk of death (22-25). In conjunction with the implementation of the 7-valent pneumococcal conjugate vaccine (26, 27), the prevalence of invasive pneumococcal disease caused by non-vaccine serotypes, including serotype 3, has increased. Our laboratory has used a highly virulent serotype 3 S. pneumoniae strain (6303) in a mouse model of pneumococcal pneumonia to mimic severe infection associated with this serotype (15, 28-30). In contrast to models with the commonly used D39 serotype 2 strain, low dose infection with S. pneumoniae 6303 results in a high mortality in immune competent mice caused by a gradual growth and subsequent dissemination of bacteria, thereby more closely mimicking human disease (8, 10, 14, 15, 18, 28-31). Here we infected Nlrp3-/- and Asc-/- mice with low dose S. pneumoniae 6303 via the airways to investigate the role of the NLRP3 inflammasome in the host response. Much to our surprise we found that both Nlrp3-/- and Asc-/- mouse strains were strongly protected against lethality. We also infected Myd88-/- mice with S. pneumoniae 6303 and found that MyD88-dependent signaling does not contribute to protective immunity in this model. Our data report for the first time a detrimental role of the NLRP3 inflammasome during respiratory tract infection in a clinically relevant model of a common human disease.

Methods Mice Nlrp3-/-, Asc-/- and Myd88-/- mice were generated as described (32-34). Tlr2-/-/Tlr4/mice were generated by intercrossing Tlr2-/- and Tlr4-/- mice (35, 36). Mice were backcrossed 6-9 times to a C57Bl/6 genetic background. Age and sex matched wild-type (WT) C57Bl/6 mice were obtained from Harlan Nederland (Horst, the Netherlands). Mice were infected at 9-12 weeks of age. The Animal Care and Use Committee of the University of Amsterdam approved all experiments. Induction of pneumonia and tissue harvest Mice were lightly anesthetized by inhalation of isoflurane (Upjohn, Ede, the Netherlands) and inoculated intranasally with ~ 5x104 or ~ 1 x 107 CFU (in 50 µl) of S. pneumoniae 6303 (serotype 3; American Type Culture Collection, Manassas, VA) as described (15, 28-30). Mice were followed for a maximum of 7 days for survival studies or euthanized at 6, 24 or 48 hours after infection by heart puncture under ketamine / medotomidine anesthesia. Subsequently lungs and spleen were harvested and processed for the determination of bacterial outgrowth and cytokine levels exactly as described (37, 38). Assays TNF-α and IL-6 were measured using a cytometric bead array (BD Biosciences, San Jose, CA). IL-1β, CXCL1 and CXCL2 were measured by ELISA (R&D Systems, Minneapolis, MN). Myeloperoxidase (MPO) was measured by ELISA (HyCult Biotechnology, Uden, the Netherlands). 168

ASC and NLRP3 impair host defense during lethal pneumonia caused by highly virulent serotype 3 Streptococcus pneumoniae

Histologic examination For histologic examination, lung sections were stained with hematoxylin and eosin, and lung inflammation was scored exactly as described (37). For analysis of neutrophil influx in the lung, immunohistochemical stainings were prepared using a FITC-labeled anti-mouse Ly-6C/G mAb (BD Biosciences, San Jose, CA) and Ly6C/G was quantified as described (39). Analysis of γ-H2AX expression in lung tissue was performed in a similar manner using an antibody specific for phospho-H2AX (Ser139; Cell Signaling Technology, Danvers, MA) and peroxidase-conjugated antirabbit IgG (Immunologic, Duiven, the Netherlands). Statistical analysis Data are expressed as medians with individual data points (bacterial loads), KaplanMeier plots (survival) or box and whisker plots showing the smallest observation, lower quartile, median, upper quartile and largest observation. Data in tables are mean ± SEM. Comparisons between mouse strains were done by log-rank test for survival curves, Fisher’s exact test for proportions of positive blood cultures and Mann Whitney U test for other outcomes. These analyses were done using GraphPad Prism (San Diego, CA), p < 0.05 was considered statistically significant.

Results Nlrp3-/- and Asc-/- mice display a strongly improved host defense during lethal pneumonia caused by highly virulent S. pneumoniae We studied mortality and bacterial growth and dissemination in lethal pneumonia caused by low dose infection (5 x 104 CFU) with a virulent serotype 3 S. pneumoniae strain (6303). For this we performed two separate survival studies, comparing Nlrp3-/- or Asc-/- mice with normal WT mice. As expected, in both experiments WT mice showed massive mortality between 48 and 72 hours; the last WT mice died just after 3 days of infection (Figure 1A-B). Surprisingly, both Nlrp3-/- and Asc-/- mice displayed a strongly reduced mortality; at the end of the 7-day observation period 9/15 Nlrp3-/- and 7/16 Asc-/- mice were still alive (both p < 0.001 versus WT mice). To study antibacterial defense, we infected mice with the same bacterial dose in separate experiments and subsequently euthanized them after 6, 24 or 48 hours for quantitative cultures of lungs (the primary site of infection), blood and spleen (to obtain insight into dissemination to distant organs). After 6 hours, bacterial loads in the lungs of Nlrp3-/- and Asc-/- mice were similar to those in lungs of WT mice (Figure 1C-D). At later time points, however, Nlrp3-/- or Asc-/- mice showed strongly reduced bacterial burdens in their lungs (p < 0.01 - < 0.001 versus WT mice). Strikingly, while in all WT mice the infection disseminated to blood and spleen from 24 hours after infection onward (Figure 1E-H), blood and spleen cultures remained sterile in the vast majority of Nlrp3-/- and Asc-/- mice (p < 0.01 - < 0.001 versus WT mice). Together these data suggest that Nlrp3-/- and Asc-/- mice are strongly protected against lethality during pneumonia caused by highly virulent S. pneumoniae by a robust reduction of bacterial growth and dissemination. 169

Chapter 8

Figure 1: NLRP3 and ASC contribute to mortality and enhance bacterial growth of S. pneumoniae. WT, Nlrp3-/- and Asc-/- mice were inoculated with ~ 5 x 104 CFU S. pneumoniae and monitored for survival or sacrificed at designated time-points. Survival of WT (closed squares), Nlrp3-/- (white triangles) and Asc-/- mice (open circles) (n=15-16 per group) expressed as Kaplan-Meier plot (A, B). Bacterial loads in lung (C,D), blood (E,F) and spleen (G,H) 6, 24 and 48 hours after infection in WT (closed squares), Nlrp3-/- (white triangles) and Asc-/- mice (open circles) (n=7-8 per group). Each symbol represents an individual mouse, with horizontal lines showing medians. ### p < 0.001 for the comparison between Asc-/- and WT mice, and between Nlrp3-/- and WT mice as determined by Log-Rank test. * p < 0.05, ** p < 0.01, *** p < 0.001 compared to WT mice determined with Mann-Whitney U test or Fisherâ&#x20AC;&#x2122;s exact test for the proportion of positive blood and spleen cultures. ND= not done.

170

ASC and NLRP3 impair host defense during lethal pneumonia caused by highly virulent serotype 3 Streptococcus pneumoniae

Nlrp3-/- and Asc-/- mice demonstrate increased pulmonary cytokine and chemokine concentrations early after infection For an adequate defense against uncontrolled bacterial multiplication early induction of the immune response to S. pneumoniae in the lower airways is of utmost importance (2, 40). The virulent S. pneumoniae strain used here logarithmically proliferates in the lungs during the early stages of the infection without causing a clear inflammatory response (15, 15, 28, 29). Indeed, WT mice showed a relatively modest increase in levels of proinflammatory cytokines (TNF-α, IL-1β, IL-6) and chemokines (CXCL1, CXCL2) in whole lung homogenates harvested 6 hours post infection compared to values of uninfected mice (Table 1A, B, C). Remarkably, at this early time point the lung levels of all proinflammatory mediators were higher in Nlrp3-/- and Asc-/- mice than in their respective WT groups, significantly so for all mediators in Asc-/- mice and for IL-1β in Nlrp3-/- mice (Table 1B,C). The early enhanced cytokine and chemokine response in lungs of Nlrp3-/- and Asc-/- mice did not result in an altered recruitment of neutrophils into lungs, as reflected by similarly low MPO concentrations in whole lung homogenates (Figure 2A-B) and similarly low number of Ly-6C/G+ cells in lung tissue slides when compared with WT mice (Figure 2C-F). Likewise, the extent of lung pathology was similar in all mouse strains at this early time point (Figure 2G-J). Nlrp3-/- and Asc-/- mice show strongly attenuated lung inflammation during late stages of the infection In accordance with previous reports (15, 28-30), WT mice displayed profound lung pathology during late stages of the infection (Figure 2G-J) associated with accumulation of neutrophils in lung tissue (Figure 2A-F). These changes were strongly reduced in Nlrp3-/- and Asc-/- mice: histopathology scores were much lower in inflammasome deficient mice, especially at 48 hours after infection, i.e., shortly before the first WT mice started dying (Figure 2G-J, both p < 0.01 versus WT mice), as were pulmonary MPO levels (Figure 2A,B, both p < 0.05 versus WT mice) and the number of Ly-6C/G+ cells in lung tissue (Figure 2C-F, both p < 0.05 versus WT mice). In addition, the lung levels of TNFα, IL-1β, IL-6, CXCL1 and CXCL2 were all significantly lower in Nlrp3-/- or Asc-/- mice than in WT mice at 24 and 48 hours after infection (Table 1).

171

Chapter 8

172

ASC and NLRP3 impair host defense during lethal pneumonia caused by highly virulent serotype 3 Streptococcus pneumoniae

3208 (825)** 9977 (23) 4518 (841) 5642 (562)

561 (81)

T=6

140 (15)

7(2)

15 (9)

276 (53)

2223 (266)

CXCL-2

Lung

IL-1β

TNF-α

IL-6

CXCL-1

CXCL-2

2447 (231)

3565 (1488)* 13328 (4347) 1106 (215)*** 9169 (2007)

917 (104) CXCL-1

411 (54)

6 (2) 82 (70) 23 (8)*** 1721 (623)

27 (12) IL-6

25 (10)

5 (1)* 17 (4) 14 (4)*** 247 (56)

bd TNF-α

13 (4)

4530 (1683) 8811 (1139) 40 (3)*** 327 (100)

bd IL-1β

229 (29) *

Nlrp3 -/WT WT

WT Lung

Nlrp3 -/-

474 (100)

944 (132)

10 (5)

Nlrp3 -/-

T=48 T=24

542 (135)

1064 (80)

Asc -/Nlrp3 -/Uninfected A

Table 1 Lung cytokine and chemokine levels in Nlrp3-/- and Asc-/- mice before and after pulmonary infection with S. pneumonia

Figure 2 (page 172): Neutrophil infiltration and lung inflammation after S. pneumoniae infection. WT, Nlrp3-/- and Asc-/- mice were inoculated with ~ 5 x 104 CFU S. pneumoniae and sacrificed at designated time-points. MPO levels in whole lung homogenate (A,B) and neutrophil influx in the lung as reflected by Ly-6C/G lung surface positivity (C,D) at 6, 24 and 48 hours after infection in WT (dark grey), Nlrp3-/- (light grey) and Asc-/- mice (white) (n=7-8 per group).Representative images of Ly-6C/G staining on lung slides from WT, Nlrp3-/- and Asc-/- mice, original magnification 10x (E,F). Histological scores were determined as described in the Methods section (G,H). Panels I,J show representative lung histology of WT, Nlrp3-/- and Asc-/- mice, H&E staining, original magnification 10x. Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. * p < 0.05, ** p < 0.01, *** p < 0.001 compared to WT mice determined with Mann-Whitney U test.

173

174

157 (12)

24 (2)

32 (5)

798 (46)

2744 (106)

Lung

IL-1Î˛

TNF-Îą

IL-6

CXCL-1

CXCL-2

3351 (47)**

991 (68)*

81 (7)***

35 (4)*

200 (9)*

Asc -/-

4583 (905)

4993 (1269)

1048 (264)

399 (93)

347 (115)

5720 (475)

1370 (159)***

87 (69)**

40 (21)**

176 (30)

17421 (3831)

8809 (1821)

487 (272)

54 (20)

5460 (1346)

Asc-/-

T=48

T=24

4212 (1048)**

1227 (378)***

25 (10)**

18 (4)

841 (116)***

Asc-/-

Uninfected mice (A) were sacrificed. WT, Nlrp3-/- (B) and Asc-/- (C) mice were inoculated with ~5 x 104 CFU S. pneumoniae and sacrificed after 6, 24 or 48 hours. Cytokine and chemokine levels are presented in pg/ml lung homogenate. Data are mean (SEM) of 7-8 mice per group.* p < 0.05, ** p < 0.01, *** p < 0.001 vs WT mice determined with Mann-Whitney U test. bd = below detection.

T=6

Chapter 8

ASC and NLRP3 impair host defense during lethal pneumonia caused by highly virulent serotype 3 Streptococcus pneumoniae

Improved host defense in Nlrp3-/- and Asc-/- mice is associated with diminished cell death during late stages of the infection Besides processing of IL-1β and IL-18, activation of the NLRP3 inflammasome may trigger pyroptotic cell death (7, 41, 42). Since pneumococcal pneumonia in Nlrp3-/and Asc-/- mice was characterized by reduced bacterial dissemination and reduced lung pathology, we wanted to determine whether the improved host defense response in these inflammasome-deficient mice resulted from diminished cell death early during infection. Since specific markers for pyroptosis detection in situ have not been described yet, we analyzed the presence of phosphorylated histone H2AX (γ-H2AX) in the lung. H2AX is modified to γ-H2AX immediately following DNA double-strand breaks by various means including apoptosis and necroptosis (43, 44). Immunohistological staining of γ-H2AX in lung sections showed only few strongly positive cells 6 and 24 h after infection and revealed no differences between WT and Nlrp3-/- or Asc-/- mice (Fig 3A-D). During late stages of the infection, strongly γ-H2AX-positive cells were detected in areas with neutrophil accumulation and cells along the pleura in the lung tissue of WT mice, but not in Nlrp3-/- and Asc-/- mice (Fig 3A-D). These data suggest that NLRP3-inflammasome activation does not evoke cell death in the lung early during S. pneumonia 6303-induced pneumonia, but is associated with increased cell death during late stages of the infection.

175

Chapter 8

Figure 3: Reduced numbers of γ-H2AX-positive cells in the lung of Nlrp3-/- and Asc-/- mice during late stages of pneumococcal pneumonia. WT, Nlrp3-/- and Asc-/- mice were inoculated with ~ 5 x 104 CFU S. pneumoniae and sacrificed at designated time-points. Expression of γ-H2AX in lung sections was quantified by digital image analysis as described in the Methods; the amount of γ-H2AX positivity was expressed as a percentage of the total lung surface area (A, B). Representative images of γ-H2AX staining on lung slides from WT, Nlrp3-/- and Asc-/- mice (C,D), original magnification 20x. Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. * p < 0.05, ** p < 0.01 compared to WT mice determined with Mann-Whitney U test.

MyD88 is not involved in protective immunity against lethal pneumococcal pneumonia In a previous study, we found that protective immunity during respiratory tract infection caused by S. pneumoniae D39 is strongly dependent on MyD88 signaling, as reflected by increased bacterial growth and lethality in Myd88-/- mice (21), which is in full accordance with a previous investigation using serotype 4 or 19F S. pneumoniae (20). Considering the unexpected role of NLRP3/ASC in airway infection caused by S. pneumoniae 6303 reported here, we also evaluated the role of MyD88 in host defense against this virulent strain. Myd88-/- and WT mice were infected with 5 x 104 CFU S. pneumoniae 6303 and followed for 5 days. Remarkably, lethality progressed in Myd88-/- mice at the same pace as in WT mice and all animals had succumbed at the end of the observation period (Figure 4A). In separate experiments we euthanized Myd88-/- and WT mice at 6 or 48 hours post infection to obtain insight in the role of MyD88 in controlling bacterial growth and 176

ASC and NLRP3 impair host defense during lethal pneumonia caused by highly virulent serotype 3 Streptococcus pneumoniae

Figure 4: MyD88-dependent signaling is not important for survival and does not restrict bacterial defense, nor do TLR2 and TLR4 contribute to antibacterial defense. WT, Myd88-/- and Tlr2-/-/Tlr4/mice were inoculated with ~ 5 x 104 CFU S. pneumoniae and monitored for survival or sacrificed 6 or 48 hours after infection. Survival of WT (closed squares) and Myd88-/- mice (open squares) (n=12 per group) expressed as Kaplan-Meier plot (A). Bacterial loads in lung (B, D) and blood (C) 6 and 48 hours after infection in WT (closed squares) and Myd88-/- (open squares) and Tlr2-/-/Tlr4-/- mice (diamonds) (n=7-8 per group). Each symbol represents an individual mouse, with horizontal lines showing medians. BC+= ratio of positive blood cultures in Tlr2-/-/Tlr4-/- mice.

dissemination. In accordance with the unaltered lethality, Myd88-/- mice displayed similar pneumococcal loads in the lung as WT mice at both time points (Figure 4B). In addition, analysis of dissemination of pneumococci to the circulation revealed no differences between the number of WT and Myd88-/- mice with positive blood cultures or bacterial loads in blood (Figure 4C). To obtain further proof for an absent role for MyD88 in host defense against pneumonia caused by S. pneumoniae 6303, we also infected Tlr2-/-/Tlr4-/- and WT mice with 5 x 104 CFU of this strain and determined bacterial loads in lungs at 6 and 48 hours after infection; no differences were found between mouse strains (Figure 4D). Together these data suggest that MyD88-dependent signaling does not contribute to protective immunity during lethal pneumonia caused by S. pneumoniae 6303.

177

Chapter 8

MyD88 is not involved in the early inflammatory response in the lung during pneumonia caused by S. pneumoniae 6303 Myd88-/- and WT mice had similar (relatively low) lung levels of proinflammatory cytokines (TNFα, IL-1β, IL-6) and chemokines (CXCL1, CXCL2) at 6 hours after infection (Table 2A, B). Likewise, early neutrophil recruitment was comparable in both mouse strains, as indicated by lung MPO concentrations (Figure 5A) and the number of Ly-6C/G+ cells in lung tissue slides (Figure 5B,C), as was the extent of lung pathology (Figure 5D,E). Hence, these results indicate that MyD88 is not involved in the induction of the inflammatory response at the primary site of infection after intrapulmonary delivery of S. pneumoniae 6303. Myd88-/- mice show reduced pulmonary cytokine levels during late stages of the infection Cytokine and chemokine levels measured in whole lung homogenates harvested 48 hours after infection, i.e. just before the first mice started dying, were lower in Myd88-/- than in WT mice, significantly so for IL-1β, CXCL1 and CXCL2 (Table 2). Similarly, lung MPO levels were lower in Myd88-/- mice at this late time point (Figure 5A). The number of Ly-6C/G+ cells in lung tissue (Figure 5B,C) and the extent of lung histopathology (Figure 5D,E) were similar in both mouse strains at 48 hours post infection. MyD88 mediates the early inflammatory response during pneumonia caused by high dose S. pneumoniae 6303 An important difference in pneumonia models evoked by different pneumococcal serotypes is the number of bacteria used to induce disease with approximately similar severity: typically the infectious dose of S. pneumoniae 6303 is 100-1000fold lower (~104 CFU) than the dose used in investigations using S. pneumoniae D39 (~106 – 107) (8, 10, 15, 18, 28-31, 45). The finding that MyD88 did not contribute to lung inflammation early after infection with S. pneumoniae 6303, but that differences in cytokine production became evident when bacterial loads had become much higher, prompted us to infect Myd88-/- and WT mice with 1 x 107 CFU S. pneumoniae 6303 and evaluate bacterial loads and the extent of lung inflammation at 6 and 24 hours post infection. In contrast to pneumonia evoked by 5 x 104 CFU S. pneumoniae 6303 (see above), 1 x 107 CFU S. pneumoniae 6303 caused an early severe inflammatory response in the lungs of WT mice with significant cytokine and chemokine release and neutrophil influx at 6 hours after inoculation (Table 3). Importantly, Myd88-/- and WT mice displayed similar pulmonary bacterial loads at both 6 and 24 hours in this high dose infection model; in addition, the number of positive blood cultures did not differ between strains (Figure 6A,B). Moreover, the extent of lung injury, that was remarkably more severe 6 hours after infection than after infection with the lower infectious dose, was not different between Myd88-/178

bd bd bd 506 (60) 1492 (105)

MyD88--/137 (4) bd bd 371 (71) 443 (63)

489 (37)

1319 (146)

T=6

143 (9)

1056 (358)

665 (169)

Lung

IL-1β

TNF-α

IL-6

CXCL-1

CXCL-2

Lung

IL-1β

TNF-α

IL-6

CXCL-1

CXCL-2

5236 (1520)

9394 (2319)

178 (49)

97 (43)

1728 (778)

T=48

1074 (319)

808 (306)*

38 (13)*

4 (1)*

168 (22)*

MyD88-/-

WT and MyD88-/- mice were inoculated with ~ 5 x 104 CFU S. pneumoniae and sacrificed after 6 or 48 hours. Cytokine and chemokine levels are presented in pg/ml lung homogenate. Data are mean (SEM) of 7-8 mice per group.* p < 0.05 vs WT mice determined with Mann-Whitney U test. bd = below detection.

MyD88-/-

Uninfected mice

Table 2: Lung cytokine and chemokine levels in MyD88-/- mice after pulmonary infection with S. pneumoniae

ASC and NLRP3 impair host defense during lethal pneumonia caused by highly virulent serotype 3 Streptococcus pneumoniae

179

Chapter 8

Figure 5: Reduced neutrophil activation in Myd88-/- mice during late stage infection, while lung inflammation is not determined by MyD88-dependent signaling. MPO levels (A) and Ly-6C/G lung surface positivity (B) as a reflection of neutrophil influx in the lung at 6 and 48 hours after infection in WT(dark grey) and Myd88-/- mice (dotted white) (n=7-8 per group). Representative images of Ly-6C/G staining on lung slides from WT and Myd88-/- mice, original magnification 10 x (C). Histological scores were determined as described in the Methods section (D).In panel E representative lung histology of WT, and Myd88-/- mice are shown, H&E staining, original magnification 10x. Data are expressed as box and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation.* p < 0.05 compared to WT mice determined with Mann-Whitney U test.

180

503 (111) 86 (54)** 291 (178)* 14720 (10648)** 1489 (688)*

554 (47)

481 (53)

1087 (117)

111234 (17635)

4560 (197)

IL-1Î˛

TNF-Îą

IL-6

CXCL-1

CXCL-2

3719 (278)

23082 (7596)

436 (52)

120 (23)

1943 (295)

T=24

2407 (299)**

3761 (271)**

336 (59)

33 (7)**

488 (42)***

MyD88-/-

WT and MyD88-/- mice were inoculated with ~1 x 107 CFU S. pneumoniae and sacrificed after 6 or 24 hours. Cytokine and chemokine levels are presented in pg/ml lung homogenate. Data are mean (SEM) of 7-8 mice per group.* p < 0.05, ** p < 0.01, *** p < 0.001 vs WT mice determined with Mann-Whitney U test.

MyD88-/-

Lung

T=6

Table 3: Lung cytokine and chemokine levels in MyD88-/- mice after high dose pulmonary infection with S. pneumoniae

ASC and NLRP3 impair host defense during lethal pneumonia caused by highly virulent serotype 3 Streptococcus pneumoniae

181

Chapter 8

and WT mice (Figure 6C). However, neutrophil influx, as reflected by MPO levels in whole lung homogenates (Figure 6D), and lung cytokine (TNFα, IL-1β, IL-6) and chemokine (CXCL1 and CXCL2) concentrations were lower in Myd88-/- mice at 6 and 24 hours after challenge (Table 3; not significant for IL-1β at 6 hours and IL-6 at 24 hours). These data further indicate that S. pneumoniae 6303 is recognized by TLRs when present in high numbers and that the virulence of this strain in part might be related to the fact that it can multiply in the airways without being noticed by TLRs during the early phase of eventually lethal infection.

Figure 6: MyD88-dependent signaling is involved in the inflammatory response after infection with a supralethal dose of S. pneumoniae. WT and Myd88-/- mice were inoculated with ~ 1 x 107 CFU S. pneumoniae and sacrificed 6 and 24 hours after infection. Bacterial loads in lung (A), blood (B) 6 and 24 hours after infection in WT (closed squares), Myd88-/- (open squares) (n=7-8 per group); each symbol represents an individual mouse, with horizontal lines showing medians. MPO levels as a reflection of neutrophil influx in whole lung homogenate 6 and 48 hours after infection in WT(dark grey) and Myd88/mice (dotted white) (n=7-8 per group) (C). Histological scores (D) were determined as described in the Methods section. Data are expressed as box-and-whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile, and largest observation. * p < 0.05, ** p < 0.01, *** p < 0.001 compared to WT mice determined with Mann-Whitney U test.

182

ASC and NLRP3 impair host defense during lethal pneumonia caused by highly virulent serotype 3 Streptococcus pneumoniae

Discussion S. pneumoniae is the most frequent causative pathogen in community-acquired pneumonia, a leading infectious disease globally (1-3). In recent years, it was concluded that the inflammasome components NLRP3 and/or ASC are crucial for an effective host defense response during pneumococcal pneumonia from studies using S. pneumoniae strains with a relatively low virulence, in particular the commonly used serotype 2 D39 strain (8-11). In the current study, we set out to evaluate their roles in a respiratory tract infection model evoked by a low dose of a gradually growing, highly virulent serotype 3 pneumococcal strain and found in stark contrast with these earlier studies that both NLRP3 and ASC strongly impair host defense. In addition, we show that MyD88, the common TLR adapter previously revealed as a major protective mediator during pneumonia caused by relatively modestly virulent S. pneumoniae strains (20, 21), does not contribute to an effective innate immune response in this model of severe pneumococcal pneumonia. Inflammasome activation results in the processing of pro-IL-1β and pro-IL-18 into the mature secreted forms of these proinflammatory cytokines by caspase-1 (5, 6). In accordance, the NLRP3 inflammasome contributed to IL-1β and IL-18 release induced by pneumococci in vitro and during pneumonia in vivo (8, 10, 11, 46). Importantly, however, previous work from our laboratory suggests that the protective phenotype of Asc-/- and Nlrp3-/- mice in this model of pneumonia caused by S. pneumoniae 6303 is not likely explained by the role of the NLRP3 inflammasome in IL-1β and IL-18 maturation, since both IL-1R1-/- and IL-18-/- mice demonstrated a transient increase in bacterial lung counts with an unaltered survival (30, 47). Accordingly, treatment with recombinant IL-1 receptor antagonist resulted in transiently increased bacterial loads in the lung without influencing survival in pneumonia caused by S. pneumoniae 6303 (48). Another group reported that IL-1β was important for limitation of bacterial growth and survival after airway infection with the serotype 3 strain WU2 (49). Hence, these data strongly argue against a role for attenuated IL-1β and/or IL-18 maturation in the improved outcome of Nrlp3-/- and Asc-/- mice. The finding that Nrlp3-/- and Asc-/- mice had attenuated lung inflammation during the late phase of the infection most likely was the consequence of the much lower bacterial loads in these mice. Another hall-mark feature of inflammasome activation is the occurrence of caspase1-dependent pyroptosis, a form of programmed cell death considered to be part of a protective host response to intracellular bacteria (50). In the present study we did not find proof for inflammasome-mediated cell death in the lung early during pneumococcal pneumonia according to analysis of γ-H2AX expression. Previous work has indicated that pyroptosis does occur upon exposure to S. pneumoniae in vitro in an ASC- and caspase-1-dependent way, triggered by pneumolysin (8). Although the role of pyroptosis during infections in vivo is not well established, the general assumption is that pyroptosis serves an antibacterial function (50, 51). However, a recent publication provided evidence for an opposite role of caspase-1dependent pyroptosis during pneumonia caused by the same S. pneumoniae strain used here (52). This supports our observation that deficiency of inflammasome 183

Chapter 8

components can be beneficial during pneumococcal infection and we hypothesize that S. pneumoniae 6303 benefits from preemptive inflammasome-induced celldeath by facilitating bacterial growth and dissemination and possibly downregulation of early cytokine production in an inflammasome-dependent manner, as suggested by our data. Further studies are required to provide more insight in the process of pyroptotic cell death in this model of severe pneumococcal pneumonia. The differential contribution of innate immune sensors to control the growth and dissemination of S. pneumoniae 6303 and D39 is further illustrated by the fact that Myd88-/- mice displayed an unaltered capacity to control bacterial proliferation during pneumonia caused by S. pneumoniae 6303, while these mice have a significantly reduced capacity to limit the growth of S. pneumoniae D39 (21). In our current experiments, MyD88 deficiency only influenced the induction of proinflammatory mediators in the lung in the presence of high burdens of S. pneumoniae 6303, i.e., either during late phase infection after inoculation with a low infectious dose or early after inoculation with a supralethal dose, suggesting that S. pneumoniae 6303 is not sensed by the TLR system when present in low yet eventually lethal quantities. Less virulent strains, such as S. pneumoniae D39, only cause significant disease when administered in relatively high infectious doses; these strains are then readily sensed by multiple innate recognition systems, including the NLRP3 inflammasome and TLRs, which in this case are important for protective immunity. It is tempting to speculate that the different outcomes of Asc-/- and Nlrp3-/- mice after infection with S. pneumoniae 6303 and D39 at least in part depend on the net balance between the seemingly opposite roles of ASC and NLRP3 in key antibacterial responses, i.e. their ability to induce bacteria-induced pyroptosis (inhibiting bacterial killing) versus their ability to produce proinflammatory cytokines (favoring bacterial clearance), and that these roles are influenced by the bacterial load and the consequent capacity of innate sensing systems to detect S. pneumoniae. Multiple reports have pointed to a key role for ASC and/or NLRP3 in protective innate immunity during infectious diseases (5, 6), including pneumococcal pneumonia (8, 10, 11). We here demonstrate that ASC and NLRP3 strongly impair host defense during lower airway infection with a highly virulent serotype 3 S. pneumoniae strain in a model characterized during the early phase after infection by bacterial multiplication without noticeable TLR sensing. These data are of clinical relevance, considering the capacity of serotype 3 pneumococcal strains to cause severe disease in humans (22-25, 53) and considering that in real life lower respiratory tract infections are unlikely to be initiated by the high infectious doses used in most murine models of pneumococcal pneumonia. The current results are the first to expose a detrimental role of ASC and NLRP3 in antibacterial defense during a clinically relevant model of community-acquired pneumonia.

Acknowledgments We are indebted to Fayaz S. Sutterwala and Richard A. Flavell for kindly providing us with Nlrp3-/- and Asc-/ - mice, and to Shizuo Akira for kindly providing us with Myd88-/-, Tlr2-/- and Tlr4-/- mice.

184

ASC and NLRP3 impair host defense during lethal pneumonia caused by highly virulent serotype 3 Streptococcus pneumoniae

References

2. van der Poll T, Opal SM. Pathogenesis, Treatment, and Prevention of Pneumococcal Pneumonia. Lancet 2009;374:1543-1556.

3. Welte T, Torres A, Nathwani D. Clinical and Economic Burden of Community-Acquired Pneumonia Among Adults in Europe. Thorax 2012;67:71-79.

4. Ramirez JA, Anzueto AR. Changing Needs of Community-Acquired Pneumonia. J Antimicrob Chemother 2011;66 Suppl 3:iii3-iii9.

5. Strowig T, Henao-Mejia J, Elinav E, Flavell R. Inflammasomes in Health and Disease. Nature 2012;481:278-286.

6. Franchi L, Munoz-Planillo R, Nunez G. Sensing and Reacting to Microbes Through the Inflammasomes. Nat Immunol 2012;13:325-332.

7. Bergsbaken T, Fink SL, Cookson BT. Pyroptosis: Host Cell Death and Inflammation. Nat Rev Microbiol 2009;7:99-109.

9. van Lieshout MH, Scicluna BP, Florquin S, van der Poll T. NLRP3 and ASC Differentially Affect the Lung Transcriptome During Pneumococcal Pneumonia. Am J Respir Cell Mol Biol 2014;50:699-712.

10. McNeela EA, Burke A, Neill DR, Baxter C, Fernandes VE, Ferreira D, Smeaton S, El-Rachkidy R, McLoughlin RM, Mori A, et al. Pneumolysin Activates the NLRP3 Inflammasome and Promotes Proinflammatory Cytokines Independently of TLR4. PLoS Pathog 2010;6:e1001191.

11. Witzenrath M, Pache F, Lorenz D, Koppe U, Gutbier B, Tabeling C, Reppe K, Meixenberger K, Dorhoi A, Ma J, et al. The NLRP3 Inflammasome Is Differentially Activated by Pneumolysin Variants and Contributes to Host Defense in Pneumococcal Pneumonia. J Immunol 2011;187:434-440.

12. Hoegen T, Tremel N, Klein M, Angele B, Wagner H, Kirschning C, Pfister HW, Fontana A, Hammerschmidt S, Koedel U. The NLRP3 Inflammasome Contributes to Brain Injury in Pneumococcal Meningitis and Is Activated Through ATP-Dependent Lysosomal Cathepsin B Release. J Immunol 2011;187:5440-5451.

13. Schroder NW, Morath S, Alexander C, Hamann L, Hartung T, Zahringer U, Gobel UB, Weber JR, Schumann RR. Lipoteichoic Acid (LTA) of Streptococcus Pneumoniae and Staphylococcus Aureus Activates Immune Cells Via Toll-Like Receptor (TLR)-2, Lipopolysaccharide-Binding Protein (LBP), and CD14, Whereas TLR-4 and MD-2 Are Not Involved. J Biol Chem 2003;278:15587-15594.

185

Chapter 8

14. Dessing MC, Schouten M, Draing C, Levi M, von AS, van der Poll T. Role Played by Toll-Like Receptors 2 and 4 in Lipoteichoic Acid-Induced Lung Inflammation and Coagulation. J Infect Dis 2008;197:245-252.

15. Knapp S, Wieland CW, van â&#x20AC;&#x2DC;t Veer C, Takeuchi O, Akira S, Florquin S, van der Poll T. TollLike Receptor 2 Plays a Role in the Early Inflammatory Response to Murine Pneumococcal Pneumonia but Does Not Contribute to Antibacterial Defense. J Immunol 2004;172:31323138.

16. Malley R, Henneke P, Morse SC, Cieslewicz MJ, Lipsitch M, Thompson CM, Kurt-Jones E, Paton JC, Wessels MR, Golenbock DT. Recognition of Pneumolysin by Toll-Like Receptor 4 Confers Resistance to Pneumococcal Infection. Proc Natl Acad Sci U S A 2003;100:19661971.

17. Branger J, Knapp S, Weijer S, Leemans JC, Pater JM, Speelman P, Florquin S, van der Poll T. Role of Toll-Like Receptor 4 in Gram-Positive and Gram-Negative Pneumonia in Mice. Infect Immun 2004;72:788-794.

18. Dessing MC, Florquin S, Paton JC, van der Poll T. Toll-Like Receptor 2 Contributes to Antibacterial Defence Against Pneumolysin-Deficient Pneumococci. Cell Microbiol 2008;10:237-246.

19. Albiger B, Dahlberg S, Sandgren A, Wartha F, Beiter K, Katsuragi H, Akira S, Normark S, Henriques-Normark B. Toll-Like Receptor 9 Acts at an Early Stage in Host Defence Against Pneumococcal Infection. Cell Microbiol 2007;9:633-644.

20. Albiger B, Sandgren A, Katsuragi H, Meyer-Hoffert U, Beiter K, Wartha F, Hornef M, Normark S, Normark BH. Myeloid Differentiation Factor 88-Dependent Signalling Controls Bacterial Growth During Colonization and Systemic Pneumococcal Disease in Mice. Cell Microbiol 2005;7:1603-1615.

21. de Vos AF, Dessing MC, Lammers AJ, de Porto AP, Florquin S, de Boer OJ, de BR, Terpstra S, Bootsma HJ, Hermans PW, et al. The Polysaccharide Capsule of Streptococcus Pneumonia Partially Impedes MyD88-Mediated Immunity During Pneumonia in Mice. PLoS ONE 2015;10:e0118181.

22. Weinberger DM, Harboe ZB, Sanders EA, Ndiritu M, Klugman KP, Ruckinger S, Dagan R, Adegbola R, Cutts F, Johnson HL, et al. Association of Serotype With Risk of Death Due to Pneumococcal Pneumonia: a Meta-Analysis. Clin Infect Dis 2010;51:692-699.

23. Bender JM, Ampofo K, Byington CL, Grinsell M, Korgenski K, Daly JA, Mason EO, Pavia AT. Epidemiology of Streptococcus Pneumoniae-Induced Hemolytic Uremic Syndrome in Utah Children. Pediatr Infect Dis J 2010;29:712-716.

24. Burgos J, Lujan M, Falco V, Sanchez A, Puig M, Borrego A, Fontanals D, Planes AM, Pahissa A, Rello J. The Spectrum of Pneumococcal Empyema in Adults in the Early 21st Century. Clin Infect Dis 2011;53:254-261.

25. Yu J, Salamon D, Marcon M, Nahm MH. Pneumococcal Serotypes Causing Pneumonia With Pleural Effusion in Pediatric Patients. J Clin Microbiol 2011;49:534-538.

26. Tan TQ. Pediatric Invasive Pneumococcal Disease in the United States in the Era of Pneumococcal Conjugate Vaccines. Clin Microbiol Rev 2012;25:409-419.

186

ASC and NLRP3 impair host defense during lethal pneumonia caused by highly virulent serotype 3 Streptococcus pneumoniae

27. Weil-Olivier C, van der Linden M, de S, I, Dagan R, Mantovani L. Prevention of Pneumococcal Diseases in the Post-Seven Valent Vaccine Era: a European Perspective. BMC Infect Dis 2012;12:207.

28. Dessing MC, Knapp S, Florquin S, de Vos AF, van der Poll T. CD14 Facilitates Invasive Respiratory Tract Infection by Streptococcus Pneumoniae. Am J Respir Crit Care Med 2007;175:604-611.

29. van der Windt GJ, Hoogendijk AJ, Schouten M, Hommes TJ, de Vos AF, Florquin S, van der Poll T. Osteopontin Impairs Host Defense During Pneumococcal Pneumonia. J Infect Dis 2011;203:1850-1858.

30. Rijneveld AW, Florquin S, Branger J, Speelman P, Van Deventer SJ, van der PT. TNF-Alpha Compensates for the Impaired Host Defense of IL-1 Type I Receptor-Deficient Mice During Pneumococcal Pneumonia. J Immunol 2001;167:5240-5246.

31. Ogunniyi AD, LeMessurier KS, Graham RM, Watt JM, Briles DE, Stroeher UH, Paton JC. Contributions of Pneumolysin, Pneumococcal Surface Protein A (PspA), and PspC to Pathogenicity of Streptococcus Pneumoniae D39 in a Mouse Model. Infect Immun 2007;75:1843-1851.

32. Kuida K, Lippke JA, Ku G, Harding MW, Livingston DJ, Su MS, Flavell RA. Altered Cytokine Export and Apoptosis in Mice Deficient in Interleukin-1 Beta Converting Enzyme. Science 1995;267:2000-2003.

33. Sutterwala FS, Ogura Y, Szczepanik M, Lara-Tejero M, Lichtenberger GS, Grant EP, Bertin J, Coyle AJ, Galan JE, Askenase PW, et al. Critical Role for NALP3/CIAS1/Cryopyrin in Innate and Adaptive Immunity Through Its Regulation of Caspase-1. Immunity 2006;24:317-327.

34. Adachi O, Kawai T, Takeda K, Matsumoto M, Tsutsui H, Sakagami M, Nakanishi K, Akira S. Targeted Disruption of the MyD88 Gene Results in Loss of IL-1- and IL-18-Mediated Function. Immunity 1998;9:143-150.

35. Hoshino K, Takeuchi O, Kawai T, Sanjo H, Ogawa T, Takeda Y, Takeda K, Akira S. Cutting Edge: Toll-Like Receptor 4 (TLR4)-Deficient Mice Are Hyporesponsive to Lipopolysaccharide: Evidence for TLR4 As the Lps Gene Product. J Immunol 1999;162:3749-3752.

36. Takeuchi O, Hoshino K, Kawai T, Sanjo H, Takada H, Ogawa T, Takeda K, Akira S. Differential Roles of TLR2 and TLR4 in Recognition of Gram-Negative and Gram-Positive Bacterial Cell Wall Components. Immunity 1999;11:443-451.

37. Wieland CW, van Lieshout MH, Hoogendijk AJ, van der Poll T. Host Defence During Klebsiella Pneumonia Relies on Haematopoietic-Expressed Toll-Like Receptors 4 and 2. Eur Respir J 2011;37:848-857.

38. van Lieshout MH, Blok DC, Wieland CW, de Vos AF, vanâ&#x20AC;&#x2122;t Veer V, van der Poll T. Differential Roles of MyD88 and TRIF in Hematopoietic and Resident Cells During Murine Gram-Negative Pneumonia. J Infect Dis 2012;206:1415-1423.

39. van Lieshout MH, Anas AA, Florquin S, Hou B, Vanâ&#x20AC;&#x2122;t Veer C, de Vos AF, van der Poll T. Hematopoietic but Not Endothelial Cell MyD88 Contributes to Host Defense During GramNegative Pneumonia Derived Sepsis. PLoS Pathog 2014;10:e1004368.

40. Koppe U, Suttorp N, Opitz B. Recognition of Streptococcus Pneumoniae by the Innate Immune System. Cell Microbiol 2012;14:460-466.

187

Chapter 8

41. Creagh EM. Caspase Crosstalk: Integration of Apoptotic and Innate Immune Signalling Pathways. Trends Immunol 2014;35:631-640.

42. Leissinger M, Kulkarni R, Zemans RL, Downey GP, Jeyaseelan S. Investigating the Role of Nucleotide-Binding Oligomerization Domain-Like Receptors in Bacterial Lung Infection. Am J Respir Crit Care Med 2014;189:1461-1468.

43. Baritaud M, Cabon L, Delavallee L, Galan-Malo P, Gilles ME, Brunelle-Navas MN, Susin SA. AIF-Mediated Caspase-Independent Necroptosis Requires ATM and DNA-PK-Induced Histone H2AX Ser139 Phosphorylation. Cell Death Dis 2012;3:e390.

44. Rogakou EP, Nieves-Neira W, Boon C, Pommier Y, Bonner WM. Initiation of DNA Fragmentation During Apoptosis Induces Phosphorylation of H2AX Histone at Serine 139. J Biol Chem 2000;275:9390-9395.

45. Berry AM, Paton JC. Additive Attenuation of Virulence of Streptococcus Pneumoniae by Mutation of the Genes Encoding Pneumolysin and Other Putative Pneumococcal Virulence Proteins. Infect Immun 2000;68:133-140.

46. Shoma S, Tsuchiya K, Kawamura I, Nomura T, Hara H, Uchiyama R, Daim S, Mitsuyama M. Critical Involvement of Pneumolysin in Production of Interleukin-1alpha and Caspase-1Dependent Cytokines in Infection With Streptococcus Pneumoniae in Vitro: a Novel Function of Pneumolysin in Caspase-1 Activation. Infect Immun 2008;76:1547-1557.

47. Lauw FN, Branger J, Florquin S, Speelman P, Van Deventer SJ, Akira S, van der Poll T. IL-18 Improves the Early Antimicrobial Host Response to Pneumococcal Pneumonia. J Immunol 2002;168:372-378.

48. Rijneveld AW, Florquin S, Speelman P, Edwards CK, Dinarello CA, van der Poll T. Interleukin-1 Receptor Antagonist Transiently Impairs Antibacterial Defense but Not Survival in Murine Pneumococcal Pneumonia. Eur Cytokine Netw 2003;14:242-245.

49. Kafka D, Ling E, Feldman G, Benharroch D, Voronov E, Givon-Lavi N, Iwakura Y, Dagan R, Apte RN, Mizrachi-Nebenzahl Y. Contribution of IL-1 to Resistance to Streptococcus Pneumoniae Infection. Int Immunol 2008;20:1139-1146.

50. Miao EA, Rajan JV, Aderem A. Caspase-1-Induced Pyroptotic Cell Death. Immunol Rev 2011;243:206-214.

51. Lamkanfi M, Dixit VM. Modulation of Inflammasome Pathways by Bacterial and Viral Pathogens. J Immunol 2011;187:597-602.

52. Dela Cruz CS, Liu W, He CH, Jacoby A, Gornitzky A, Ma B, Flavell R, Lee CG, Elias JA. Chitinase 3-Like-1 Promotes Streptococcus Pneumoniae Killing and Augments Host Tolerance to Lung Antibacterial Responses. Cell Host Microbe 2012;12:34-46.

53. Hausdorff WP, Feikin DR, Klugman KP. Epidemiological Differences Among Pneumococcal Serotypes. Lancet Infect Dis 2005;5:83-93.

188

Chapter 9 Single immunoglobulin interleukin-1 receptor related molecule impairs host defense during pneumonia and sepsis caused by Streptococcus pneumoniae Journal of Innate Immunity 2014;6(4):542-52. doi: 10.1159/000358239 Dana C. Blok 1 Miriam H.P. van Lieshout 1 Arie J. Hoogendijk 1 Sandrine Florquin 2 Onno J. de Boer 2 Cecilia Garlanda 4 Alberto Mantovani 4,5 Cornelis van â&#x20AC;&#x2122;t Veer 1 Alex F. de Vos 1 Tom van der Poll 1,3 Academic Medical Center, University of Amsterdam, the Netherlands: 1 Center of Experimental and Molecular Medicine, Center of Infection and Immunity Amsterdam 2 Department of Pathology 3 Division of Infectious Diseases University of Milan, Rozzano, Italy: Humanitas Clinical and Research Center, Department of Inflammation and Immunology 5 Department of Translational Medicine 4

Chapter 9

Abstract Streptococcus (S.) pneumoniae is a common cause of pneumonia and sepsis. Toll-like receptors (TLRs) play a pivotal role in host defense against infection. We here sought to determine the role of Single immunoglobulin interleukin-1 receptorrelated molecule (SIGIRR a.k.a. TIR8), a negative regulator of TLR signaling, in pneumococcal pneumonia and sepsis. Wild type and SIGIRR deficient (Sigirr -/-) mice were infected intranasally (to induce pneumonia) or intravenously (to induce primary sepsis) with S. pneumoniae and euthanized after 6, 24 or 48 hours for analyses. Additionally, survival studies were performed. Sigirr -/- mice showed a delayed mortality during lethal pneumococcal pneumonia. In accordance, Sigirr -/- mice displayed lower bacterial loads in lung and less dissemination of the infection at 24 hours after induction of pneumonia. SIGIRR deficiency was associated with increased interstitial and perivascular inflammation in the lung tissue early after infection without impacting on neutrophil recruitment or cytokine production. Sigirr -/- mice also demonstrated reduced bacterial burdens at multiple body sites during S. pneumoniae sepsis. Sigirr -/- alveolar macrophages and neutrophils exhibited an increased capacity to phagocytose viable pneumococci. These results suggest that SIGIRR impairs antibacterial host defense during pneumonia and sepsis caused by S. pneumoniae.

190

Single immunoglobulin interleukin-1 receptor related molecule impairs host defense during pneumonia and sepsis caused by Streptococcus pneumoniae

Introduction The gram-positive diplococcus Streptococcus (S.) pneumoniae is the leading cause of community-acquired pneumonia and the most common cause of death from infection in developed countries today (1, 2). In the United States alone, S. pneumoniae is responsible for more than half a million pneumonia cases and 50,000 episodes of bacteremia each year, with case fatality rates of 7 and 20% respectively (3); similar figures have been reported for Europe (4). Globally, the annual pneumococcal related death toll has been estimated at approximately 2 million (2). As such, S. pneumoniae represents a major health burden despite vaccination programs and effective antibiotic treatments. Toll-like receptors (TLRs) form an important part of innate defense against infection (5, 6). TLRs recognize conserved motifs expressed by microbes (pathogen/ microbe-associated molecular patterns or PAMPs/MAMPs) and host derived damage-associated molecular patterns (DAMPs), resulting in the recruitment of intracellular adaptor molecules, the activation of nuclear factor (NF)-ÎşB and other signaling pathways, and the production of pro-inflammatory cytokines. Multiple TLRs are involved in the detection of pneumococci. TLR2 is mainly responsible for recognition of S. pneumoniae cell wall components (7-9), while TLR4 induces cytokine release in response to pneumolysin, a toxin expressed by all virulent pneumococcal strains (10, 11). During experimental pneumococcal pneumonia protective roles have been reported for TLR4 (10-12) and TLR9 (13), the receptor that recognizes bacterial DNA (5, 6), while TLR2 contributes to the induction of inflammation in the airways (7, 14). All TLRs involved in sensing S. pneumoniae signal via a common adapter: myeloid differentiation primary response gene 88 (MyD88), which also mediates the intracellular effects of the interleukin (IL)-1 receptor (R) and IL-18R (15). Not unexpectedly, mice with a genetic deletion of the Myd88 gene (Myd88 -/- mice) showed a strongly impaired host defense during pneumococcal pneumonia, as reflected by enhanced bacterial growth and an increased mortality (16). Unrestrained activation of TLRs can cause disproportionate inflammation and collateral tissue damage. Therefore, TLR signaling is securely regulated in order to avoid such injurious inflammatory responses (17). Single immunoglobulin interleukin-1 receptor related molecule (SIGIRR or TIR8) has been shown to inhibit NF-ÎşB activation dependent on TLRs and IL-1R like receptors (ILRs) (18). SIGIRR is ubiquitously expressed in different tissues, including the lung, where the main SIGIRR positive cell types are bronchial epithelium, leukocytes and blood endothelial cells (19). Recent research has implicated SIGIRR as an important regulator of inflammation in the respiratory tract. In a model of acute pneumonia caused by Pseudomonas (P.) aeruginosa, a gram-negative pathogen primarily affecting immune compromised hosts, Sigirr -/- mice showed increased lethality and higher bacterial burdens together with exaggerated local and systemic inflammation (19). Likewise, in chronic lung infection caused by Mycobacterium tuberculosis SIGIRR deficiency was associated with excessive lung and systemic inflammation, and as a consequence thereof increased lethality (20). In accordance, overexpression of SIGIRR in lung epithelial cells attenuated acute lung injury elicited by airway 191

Chapter 9

exposure to LPS, the toxic component of gram-negative bacteria (21). Thus far, the contribution of SIGIRR to the host response during gram-positive infection has not been studied. We here sought to determine the role of SIGIRR in pneumonia and sepsis caused by S. pneumoniae.

Materials and methods Animals Specific pathogen free 9-11 week old C57BL/6 wild-type (WT) mice were from Charles River (Maastricht, the Netherlands). Sigirr -/- mice (22), backcrossed six times to a C57BL/6 background, were bred in the animal facility of the Academic Medical Center in Amsterdam. Age- and sex-matched animals were used in all experiments. The Animal Care and Use Committee of the University of Amsterdam approved all experiments. Experimental infections The models of pneumococcal pneumonia and pneumococcal sepsis have previously been described (23, 24). In short, mice were inoculated intranasally (to induce pneumonia) with 5x104 CFU of S. pneumoniae (serotype 3; American Type Culture Collection, ATCC 6303, Rockville, MD) or intravenously (to induce primary sepsis) with 5x105 CFU S. pneumoniae. Lung, blood, spleen and liver were harvested 6, 24, or 48 post infection for quantitative bacterial cultures as described (n = 8 or 16 per group at each time point) (23, 24). Neutrophil counts in bronchoalveolar lavage (BAL) fluid were determined as described (25). In separate studies, mice were followed for 4 days and survival was monitored at least every 12 hours (n = 20 per group). Histopathological analysis Lung histopathology was semi-quantitatively analyzed as described (7, 26). In short: the “lung inflammation score” was expressed as the sum of six, on a scale of 0 (‘absent’) to 4 (‘severe’) graded, parameters: pleuritis, bronchitis, edema, interstitial inflammation, percentage of pneumonia, and endothelialitis. Granulocyte staining was done with a Ly-6G monoclonal antibody (BD Pharmingen, San Diego, CA, USA) as described previously (23, 27). The entire Ly-6G stained lung sections were digitized with a slide scanner (Olympus, Tokyo, Japan). Immunopositive (Ly-6G+) areas were analyzed with ImageJ (version 2006.02.01, US National Institutes of Health, Bethesda, MD) and expressed as the percentage of the total lung surface area (26, 27). Analyses were performed in a blinded way, i.e., without knowledge of genotype (n=8 per group at each time point). Assays Lung homogenates were prepared as described (7). In lung homogenates, tumor necrosis factor (TNF)-α, IL-1β, IL-6, macrophage inflammatory protein (MIP-2, also known as CXCL2) and cytokine-induced neutrophil chemo attractant (KC, also known as CXCL1) were measured using specific enzyme-linked immunosorbent assays (R&D systems, Abingdon, UK) in accordance with the manufacturer’s 192

Single immunoglobulin interleukin-1 receptor related molecule impairs host defense during pneumonia and sepsis caused by Streptococcus pneumoniae

recommendations. In plasma TNF-Îą, IL-6 and monocyte chemoattractant protein 1 (MCP1, also known as CCL2) were measured by cytometric bead array multiplex assay (BD Biosciences, San Jose, CA). Phagocytosis Heparinized whole blood from WT and Sigirr -/- mice was collected and murine alveolar macrophages (AMs) were obtained by bronchoalveolar lavage and cultured to adhere overnight. Phagocytosis of UV irradiated (254 nm; 30 minutes at 0.12J/ cm2; in a BLX-254; Vilber Lourmat, France) CFSE labeled (Invitrogen) opsonized (10% autologous normal mouse serum) S. pneumoniae by alveolar macrophages (MOI 100) and in whole blood by GR-1 identified neutrophils (8x107 CFU/ml blood) was determined with the help of flow cytometry as described previously (23). The percentage phagocytosing cells at 37 oC was corrected for the percentage phagocytosis at 4 oC. Statistical analysis Data are expressed as box and whisker diagrams depicting the smallest observation, lower quartile, median, upper quartile and largest observation or as bar graphs depicting means Âą SEM. Differences were analyzed by Mann Whitney U test. Survival was compared by Kaplan-Meier analysis followed by a log rank test. A value of p < 0.05 was considered statistically significant.

Results Sigirr -/- mice show delayed mortality and diminished bacterial outgrowth during S. pneumoniae pneumonia To obtain insight into the potential role of SIGIRR in the outcome of pneumococcal pneumonia WT and Sigirr -/- mice were infected intranasally with S. pneumoniae and followed for 4 days (Figure 1A). Sigirr -/- mice showed a prolonged survival (median survival time 63 hours) when compared to WT mice (median survival time 54 hours, p < 0.01). In order to determine whether the survival advantage of Sigirr -/mice corresponded with an improved antibacterial response, we next harvested lungs, blood, livers and spleens from both mouse strains at predefined time points following induction of pneumonia for quantitative cultures (Figure 1B-E). At 6 hours post infection, pneumococci were cultured from lungs only (with the exception of one positive blood culture in each group) and bacterial loads were similar in Sigirr -/and WT mice. In contrast, at 24 hours after infection, Sigirr -/- mice showed lower bacterial counts in all body sites examined when compared with WT mice (lungs p < 0.05; blood, spleen, liver all p < 0.01). At 48 hours, differences in bacterial loads between mouse strains had subsided in distant organs, whereas at this late time point Sigirr -/- mice still had lower pneumococcal burdens in their lungs (p < 0.01 versus WT mice). Together these data suggest that Sigirr -/- mice show a prolonged survival in this lethal model of pneumococcal pneumonia resulting from a transient limitation of bacterial growth and dissemination. 193

Chapter 9

Figure 1: Survival and bacterial loads in WT and Sigirr -/- mice during pneumococcal pneumonia. Wild-type (grey dots & boxes) and Sigirr -/- mice (white dots & boxes) were inoculated with S. pneumoniae intranasally. Lack of SIGIRR improved survival following intranasal infection (A). Sigirr -/- mice displayed decreased bacterial loads after 24 and 48 hours in lung (B), and after 24 hours in blood (C), liver (D) and spleen (E). CFUs are expressed as box and whisker plots showing the smallest observation, lower quartile, median, upper quartile and largest observation. Survival: N= 20 mice per group, t=6 & 24 hours n=16 mice per group, t=48 hours n=8 mice per group. * p < 0.05, ** p < 0.01

194

Single immunoglobulin interleukin-1 receptor related molecule impairs host defense during pneumonia and sepsis caused by Streptococcus pneumoniae

Sigirr -/- mice demonstrate an increased pulmonary inflammatory response early after induction of pneumonia To obtain insight in the role of SIGIRR in the regulation of lung inflammation induced by S. pneumoniae, lung tissue slides prepared 6, 24 and 48 hours after infection were semi-quantitatively analyzed according to the scoring system described in Methods section (Figure 2A-E). As reported earlier, this model of pneumococcal pneumonia is characterized by a gradually developing inflammatory response in lung tissue with typical features of lower respiratory tract infection, including bronchitis, perivascular and interstitial inflammation, edema and, especially during the progressed phase, accumulation of neutrophils (12, 23, 24). At the earliest time point (6 hours) Sigirr -/- mice showed significantly increased lung inflammation (p < 0.05 relative to WT mice), which was caused by enhanced interstitial and perivascular inflammation. At later time points, when the extent of lung pathology had clearly increased, pathology scores did not differ between Sigirr -/- and WT mice. The recruitment of neutrophils to the primary site of infection is a prominent part of the innate immune response to invading respiratory pathogens. We determined the number of neutrophils in lung tissue by quantifying the amount of Ly-6+ cells in lung slides by digital imaging (Figure 3A-E). In both mouse strains, the number of Ly6+ cells in lung tissue gradually increased during the course of the infection; no differences between groups were observed at any time point (6h p = 0.5; 24h p = 0.9; 48h p = 0.13). Considering the importance of early neutrophil influx into the bronchoalveolar space, we also analyzed the number of neutrophils in BAL fluid harvested 6 hours after infection; no difference between Sigirr -/- and WT mice was found (p = 0.13) (Figure 3F). BAL fluid macrophage,and lymphocytes numbers were also similar between groups at this early time point (data not shown). Together these data argue against an important role for SIGIRR in neutrophil recruitment during pneumococcal pneumonia; SIGIRR deficiency did impact on early interstitial and perivascular inflammation. Sigirr -/- mice demonstrate unaltered lung cytokine and chemokine levels in lungs following S. pneumoniae infection To obtain further insight in the potential role of SIGIRR in the regulation of lung inflammation during pneumococcal pneumonia, we measured proinflammatory cytokines (TNF-α, IL-1β, IL-6) and chemokines (MIP-2, KC) in whole lung homogenates harvested 6, 24 or 48 hours after intranasal inoculation with S. pneumoniae (Table 1). Lung cytokine and chemokine levels were similar in Sigirr -/and WT mice at all time points. Moreover, as a readout for systemic inflammation, we measured the plasma concentrations of TNFα, IL-6 and MCP-1; no differences were found between groups (Table 1).

195

Chapter 9

Figure 2: Histopathology of lungs from WT and Sigirr -/- mice during pneumococcal pneumonia. Semi-quantitative histology scores of lung slides as determined by the scoring system described in the Methods section from WT (grey bars) and Sigirr -/- mice (white bars) (A) and a haematoxylin-eosin staining of the lung 6 (B-C) and 48 hours (D-E) following pneumococcal infection. Representative lung slides of WT (B,D) and Sigirr -/- mice (C,E); original magnification x10. Histology scores are mean Âą SEM of 8 mice per group at each time point. Arrows in C indicate several foci of increased perivascular and interstitial inflammation.

196

Single immunoglobulin interleukin-1 receptor related molecule impairs host defense during pneumonia and sepsis caused by Streptococcus pneumoniae

Figure 3: Neutrophil influx into lungs of WT and Sigirr -/- mice during pneumococcal pneumonia. Neutrophil numbers in WT (grey bars) and Sigirr -/- lung tissue (white bars) were evaluated by Ly-6 staining of lung slides (A-E) during pneumonia. Neutrophil numbers were further counted in bronchoalveolar lavage fluid 6 hours following inoculation (F). Representative Ly-6 stained lung sections of WT (B,D) and Sigirr -/- mice (C,E) are depicted 6 (B-C) and 48 hours (D-E) post induction of pneumonia: original magnification x 10. Data in panels A and F are means Âą SEM of 8 mice per group at each time point.

197

Chapter 9 Table 1: Cytokines measured in lung homogenates and plasma during S. pneumoniae pneumonia. T=6

T=24

T=48

Sigirr-/-

TNF-α

45 ± 17

109 ± 58

73 ± 34

124 ± 60

138 ± 39

199 ± 43

IL-1β

90 ± 45

164 ± 79

491 ± 260

677 ± 504

488 ± 240

1828 ± 768

IL-6

48 ± 12

102 ± 30

616 ± 194

574 ± 249

1603 ± 354

2384 ± 658

MIP-2

477 ± 78

894 ± 255

1560 ± 694

2050 ± 1025

33555 ± 5024

96068 ± 36498

1398 ± 290

942 ± 255

8882 ± 3474

4318 ± 1563

28174 ± 4591

21275 ± 7723

TNF-α

5±1

11 ± 2

15 ± 3

60 ± 5

45 ± 7

IL-6

3 ± 0.4

7±2

133 ± 38

124 ± 46

1087 ± 262

1197 ± 339

MCP-1

23 ± 2

19 ± 4

157 ± 42

161 ± 52

343 ± 57

269 ± 54

Lung

Plasma

Mice were infected with S. pneumoniae at t=0. Data are means ± SEM of 8 mice per group.

198

Single immunoglobulin interleukin-1 receptor related molecule impairs host defense during pneumonia and sepsis caused by Streptococcus pneumoniae

Enhanced phagocytosis of S. pneumoniae by Sigirr neutrophils ex vivo

-/-

alveolar macrophages and

Next, we investigated the ability of WT and Sigirr -/- alveolar macrophages and neutrophils to phagocytose S. pneumoniae. For this alveolar macrophages and whole blood were exposed to growth arrested CFSE labeled bacteria for 60 minutes at 4 or 37 oC and internalization was analyzed by flow cytometry (Figure 4). Alveolar macrophages showed a relatively low ability to phagocytose viable S. pneumoniae; nonetheless, Sigirr -/- alveolar macrophages demonstrated an enhanced capacity to internalize S. pneumoniae compared to WT macrophages (Figure 4A, p < 0.001). Neutrophils of both genotypes readily phagocytosed pneumococci; clearly Sigirr -/neutrophils showed an increased ability to phagocytose S. pneumoniae relative to WT neutrophils (Figure 4B, p < 0.001).

Figure 4: Phagocytosis of S. pneumoniae by alveolar macrophages and neutrophils. Alveolar macrophages (AM) and whole blood were exposed to fluorescently labelled S. pneumoniae for 60 minutes at 4 or 37 oC. Depicted are percentages of phagocytosing AM (A) and neutrophils (B) at 37 oC when corrected for their 4 oC controls. Data are means Âą SEM (AM: n = 6 WT vs 8 Sigirr -/-, PMN n= 8 vs 8). *** p < 0.001

199

Chapter 9

Sigirr -/- mice show diminished bacterial outgrowth during primary S. pneumoniae sepsis We wondered whether the improved antibacterial defense in Sigirr -/- mice after induction of pneumonia was primarily caused by limitation of bacterial growth in the lungs and as a consequence thereof impeded dissemination and/or by an additional reduction of pneumococcal multiplication in body sites distant from the lungs. To address this question, Sigirr -/- and WT mice were infected with S. pneumoniae by intravenous injection via the tail vein, thereby by passing the initial interaction between host and pathogen in the respiratory tract, and euthanized 6, 24 or 48 hours later for quantitative cultures of multiple body sites. Once more lower bacterial counts were observed in Sigirr -/- mice, especially at 48 hours after infection when SIGIRR deficiency was associated with reduced bacterial loads in blood (p < 0.05 versus WT mice), spleen (p < 0.05), liver (p < 0.01) and lungs (p < 0.01, Figure 5). Notably, in these experiments SIGIRR deficiency did not significantly influence the plasma concentrations of TNFÎą, IL-6 or MCP-1 (Table 1).

Figure 5: Bacterial loads in WT and Sigirr -/- mice during pneumococcal sepsis. WT (grey boxes) and Sigirr -/- mice (white boxes) were inoculated with S. pneumoniae intravenously. Sigirr -/- mice displayed decreased bacterial loads after 24 and 48 hours in the lung (A), and after 48 hours in the blood (B), liver (C) and spleen (D). Data are expressed as box and whisker plots showing the smallest observation, lower quartile, median, upper quartile and largest observation; n=8 mice per group at each time point. * p < 0.05, ** p < 0.01

200

Single immunoglobulin interleukin-1 receptor related molecule impairs host defense during pneumonia and sepsis caused by Streptococcus pneumoniae

Discussion S. pneumoniae is a common human pathogen that can reside as a commensal in the nasopharynx, from where it is able to enter the lower respiratory tract, causing pneumonia and sepsis (1, 2). The family of TLRs forms an important part of innate defense against invading pneumococci. SIGIRR is a negative regulator of TLR and ILR signaling that is abundantly expressed in the lungs. We here investigated the potential role of SIGIRR in pneumococcal pneumonia and sepsis. SIGIRR was found to impair antibacterial host defense during both pneumonia and sepsis caused by S. pneumoniae, as reflected by a reduced survival accompanied by increased bacterial growth and dissemination in WT mice when compared with Sigirr -/- animals. Upon recognition of S. pneumoniae by in particular TLR2 (7, 8), TLR4 (10, 11) and TLR9 (13) NF-κB is activated and pro-inflammatory cytokines are produced with the distinct purpose of eradicating the pathogen (5). Uncontrolled stimulation of TLRs leads to disproportionate inflammation and tissue injury. Excessive TLR activation is prevented by negative regulators of TLR signaling of which several have been identified (17), including SIGIRR (18, 28). Specifically, inhibitory activity of SIGIRR has been demonstrated on signaling by TLR4, TLR7, TLR9, IL-1R type I (IL-1RI), IL18R and ST2 (18). The current study does not elucidate via which receptor SIGIRR exerts its detrimental effect during pneumococcal pneumonia. Potential candidates are TLR4, TLR9, IL-1RI and IL-18R, considering that mice deficient for either one of these signaling pathways were reported to have an impaired antibacterial defense response during infection with S. pneumoniae. Indeed, mice with a mutant nonactive form of TLR4 showed enhanced pneumococcal growth and dissemination in models of nasopharyngeal colonization and lower respiratory tract infection, accompanied by increased lethality (10-12). Similarly, TLR9 -/- mice displayed accelerated growth of pneumococci upon infection of the lower airways together with enhanced mortality (13), while IL-1rI -/- (29) and IL-18 -/- (30) mice showed a more modestly impaired immune response only reflected by higher bacterial loads. To our knowledge, the role of TLR7 has not been studied in the context of pneumococcal infections, while our laboratory recently showed that ST2 does not contribute to host defense during S. pneumoniae pneumonia (26). Together these data suggest that SIGIRR may impair host defense during pneumococcal pneumonia by inhibition of TLR4, TLR9, IL-1RI and/or IL-18R. In contrast to the results presented here, previous studies revealed a protective role of SIGIRR in host defense against pulmonary infections. In a high dose lung infection model with P. aeruginosa associated with acute inflammation, Sigirr -/mice showed a reduced survival and diminished bacterial clearance relative to WT mice (19). Sigirr -/- mice had markedly elevated concentrations of proinflammatory cytokines in whole lung homogenates, including IL-1β, and elimination of IL-1RI signaling in Sigirr -/- mice partially reversed their worse outcome (19). Similarly, Sigirr -/- mice displayed exaggerated pulmonary inflammation and strongly elevated plasma levels of TNFα and IL-1β during experimental lung tuberculosis, and combined treatment with anti-TNFα and anti-IL-1β antibodies improved their 201

Chapter 9

survival in this model (20). Hence, the protective role of SIGIRR in these previous investigations on lung infection most likely was related to its inhibitory effect on the production of proinflammatory cytokines (19, 20). Our current data, indicating a detrimental role for SIGIRR, suggest another underlying mechanism, considering that SIGIRR deficiency did not alter the local or systemic cytokine response during S. pneumoniae pneumonia or sepsis. Nonetheless, studies in which IL-1 signaling is inhibited are warranted to address this issue. Of note, our model of gram-positive pneumonia differs considerably from the previously reported model of gramnegative pneumonia caused by P. aeruginosa (19). Indeed, while S. pneumoniae (infectious dose 5 x 104 CFU) gradually grows in lungs of normal mice resulting in a slowly building inflammatory response, airway infection by P. aeruginosa (infectious inoculum 106 CFU) is associated with a brisk inflammatory reaction while bacteria are cleared from the airways. The difference between S. pneumoniae and P. aeruginosa pneumonia models is further illustrated by the different roles of proinflammatory cytokines, which play a protective role in pneumococcal pneumonia (29-31), while they hamper bacterial clearance during Pseudomonas pneumonia (32-34). Earlier studies have indicated that the role for SIGIRR in the immune response to infection varies depending on the infecting organism and the infected site. In accordance with the enhanced susceptibility of Sigirr -/- mice during P. aeruginosa pneumonia (19), an anti-SIGIRR antibody caused increased bacterial loads in corneas during experimental Pseudomonas keratitis (35). In addition, Sigirr -/- mice were more susceptible to infection by mucosal and systemic infection by Candida albicans and to lung infection by Aspergillus fumigatus (36). In contrast, however, Sigirr -/- mice demonstrated transiently reduced bacterial burdens in kidneys during Escherichia coli pyelonephritis, possibly caused by a faster recruitment of neutrophils to the primary site of infection (37). Although clearly host defense against gram-positive and gram-negative bacteria is regulated in partially different ways (5), the present results, obtained after infection with a gram-positive bacterium, taken together with previous reports using gram-negative organisms (19, 35, 37), cannot be used to establish a differential role of SIGIRR in infections caused by either one of these very broad groups of pathogens. Indeed, the receptors known to be influenced by SIGIRR (see above) are not exclusively activated by either grampositive or gram-negative bacteria. We consider it likely that the primary site of infection, the initial bacterial load and whether the pathogen multiplies or is cleared, together with differential expression of PAMPs, determines the eventual role of SIGIRR in host defense. We provided evidence that SIGIRR attenuates S. pneumoniae phagocytosis by neutrophils and alveolar macrophages, which at least in part may explain the lower bacterial loads in Sigirr -/- mice. TLR stimulation induced expression of phagocytic genes (Fc and complement receptor genes, scavenger receptor and scavenger receptor pathway genes) in bone marrow derived macrophages (38), and resulted in enhanced uptake of both gram-negative (Escherichia coli) and gram-positive (Staphylococcus aureus) bacteria by RAW cells in a MyD88 and p38 mitogen activated kinase (MAPK) dependent manner (38). p38 MAPK signaling dependency 202

Single immunoglobulin interleukin-1 receptor related molecule impairs host defense during pneumonia and sepsis caused by Streptococcus pneumoniae

was established in LPS enhanced microglial Fc receptor mediated phagocytosis as well (39). In addition, IL-1Î˛ and IL-18 have been shown to enhance phagocytosis (39, 40). Thus, in theory, SIGIRR deficiency could impact phagocytosis by either inhibiting TLR effects on phagocytic gene expression and/or by attenuating IL-1Î˛ and/or IL-18 production and/or signaling. To the best of our knowledge this is the first report implicating SIGIRR in phagocytosis. The fact that this highly virulent S. pneumoniae strain cannot be killed by macrophages or neutrophils ex vivo ((24) and data not shown) precludes studies on the effect of SIGIRR on bacterial killing. Contrary to E. coli pyelonephritis (37), SIGIRR deficiency did not influence neutrophil influx into the lungs during pneumonia caused by either P. aeruginosa (19) or S. pneumoniae (reported here). SIGIRR is expressed by both hematopoietic and parenchymal cells (19). Bone marrow transfers, creating chimeric mice expressing SIGIRR only in the hematopoietic or parenchymal compartment, can provide insight into which cells drive the phenotype of Sigirr -/- mice in pneumococcal infection. The pneumococcus is a highly relevant human pathogen, especially in the context of community-acquired pneumonia. The immune system rapidly responds to pneumococci that try to invade the lower airways. TLRs are of paramount importance for the recognition of S. pneumoniae and for the activation of inflammatory pathways among which ILRs. We here report on the role of SIGIRR, a negative regulator of TLRs and ILRs, in S. pneumoniae pneumonia and sepsis. In contrast to earlier investigations that studied the function of SIGIRR during lung infections caused by Aspergillus fumigatus (36), Mycobacterium tuberculosis (20) or P. aeruginosa (19), in which SIGIRR improved outcome by inhibition of excessive inflammation, our results indicate that SIGIRR impairs host defense during pneumonia and sepsis caused by S. pneumoniae.

Acknowledgements We thank Joost Daalhuisen and Marieke ten Brink for expert technical assistance.

203

Chapter 9

References 1.

van der Poll T, Opal SM. Pathogenesis, treatment, and prevention of pneumococcal pneumonia. Lancet 2009;374(9700):1543-1556.

Dockrell DH, Whyte MK, Mitchell TJ. Pneumococcal pneumonia: Mechanisms of infection and resolution. Chest 2012;142(2):482-491.

Prevention of pneumococcal disease: Recommendations of the advisory committee on immunization practices (acip). MMWR Recommendations and reports : Morbidity and mortality weekly report Recommendations and reports / Centers for Disease Control 1997;46(RR-8):124.

Welte T, Torres A, Nathwani D. Clinical and economic burden of community-acquired pneumonia among adults in europe. Thorax 2012;67(1):71-79.

Ishii KJ, Koyama S, Nakagawa A, Coban C, Akira S. Host innate immune receptors and beyond: Making sense of microbial infections. Cell host & microbe 2008;3(6):352-363.

Kawai T, Akira S. Toll-like receptors and their crosstalk with other innate receptors in infection and immunity. Immunity 2011;34(5):637-650.

Knapp S, Wieland CW, van â&#x20AC;&#x2DC;t Veer C, Takeuchi O, Akira S, Florquin S, van der Poll T. Toll-like receptor 2 plays a role in the early inflammatory response to murine pneumococcal pneumonia but does not contribute to antibacterial defense. J Immunol 2004;172(5):3132-3138.

Mogensen TH, Paludan SR, Kilian M, Ostergaard L. Live streptococcus pneumoniae, haemophilus influenzae, and neisseria meningitidis activate the inflammatory response through toll-like receptors 2, 4, and 9 in species-specific patterns. Journal of leukocyte biology 2006;80(2):267-277.

Dessing MC, Schouten M, Draing C, Levi M, von Aulock S, van der Poll T. Role played by toll-like receptors 2 and 4 in lipoteichoic acid-induced lung inflammation and coagulation. The Journal of infectious diseases 2008;197(2):245-252.

10.

Malley R, Henneke P, Morse SC, Cieslewicz MJ, Lipsitch M, Thompson CM, Kurt-Jones E, Paton JC, Wessels MR, Golenbock DT. Recognition of pneumolysin by toll-like receptor 4 confers resistance to pneumococcal infection. Proceedings of the National Academy of Sciences of the United States of America 2003;100(4):1966-1971.

11.

Srivastava A, Henneke P, Visintin A, Morse SC, Martin V, Watkins C, Paton JC, Wessels MR, Golenbock DT, Malley R. The apoptotic response to pneumolysin is toll-like receptor 4 dependent and protects against pneumococcal disease. Infection and immunity 2005;73(10):64796487.

12.

Branger J, Knapp S, Weijer S, Leemans JC, Pater JM, Speelman P, Florquin S, van der Poll T. Role of toll-like receptor 4 in gram-positive and gram-negative pneumonia in mice. Infection and immunity 2004;72(2):788-794.

13.

Albiger B, Dahlberg S, Sandgren A, Wartha F, Beiter K, Katsuragi H, Akira S, Normark S, Henriques-Normark B. Toll-like receptor 9 acts at an early stage in host defence against pneumococcal infection. Cellular microbiology 2007;9(3):633-644.

204

Single immunoglobulin interleukin-1 receptor related molecule impairs host defense during pneumonia and sepsis caused by Streptococcus pneumoniae 14.

Dessing MC, Florquin S, Paton JC, van der Poll T. Toll-like receptor 2 contributes to antibacterial defence against pneumolysin-deficient pneumococci. Cellular microbiology 2008;10(1):237-246.

15.

Dinarello CA. Immunological and inflammatory functions of the interleukin-1 family. Annual review of immunology 2009;27:519-550.

16.

Albiger B, Sandgren A, Katsuragi H, Meyer-Hoffert U, Beiter K, Wartha F, Hornef M, Normark S, Normark BH. Myeloid differentiation factor 88-dependent signalling controls bacterial growth during colonization and systemic pneumococcal disease in mice. Cellular microbiology 2005;7(11):1603-1615.

17.

Kondo T, Kawai T, Akira S. Dissecting negative regulation of toll-like receptor signaling. Trends in immunology 2012;33(9):449-458.

18.

Riva F, Bonavita E, Barbati E, Muzio M, Mantovani A, Garlanda C. Tir8/sigirr is an interleukin-1 receptor/toll like receptor family member with regulatory functions in inflammation and immunity. Frontiers in immunology 2012;3:322.

19.

Veliz Rodriguez T, Moalli F, Polentarutti N, Paroni M, Bonavita E, Anselmo A, Nebuloni M, Mantero S, Jaillon S, Bragonzi A, et al. Role of toll interleukin-1 receptor (il-1r) 8, a negative regulator of il-1r/toll-like receptor signaling, in resistance to acute pseudomonas aeruginosa lung infection. Infection and immunity 2012;80(1):100-109.

20.

Garlanda C, Di Liberto D, Vecchi A, La Manna MP, Buracchi C, Caccamo N, Salerno A, Dieli F, Mantovani A. Damping excessive inflammation and tissue damage in mycobacterium tuberculosis infection by toll il-1 receptor 8/single ig il-1-related receptor, a negative regulator of il-1/tlr signaling. J Immunol 2007;179(5):3119-3125.

21.

Chen X, Zhao Y, Wu X, Qian G. Enhanced expression of single immunoglobulin il-1 receptor-related molecule ameliorates lps-induced acute lung injury in mice. Shock 2011;35(2):198204.

22.

Garlanda C, Riva F, Polentarutti N, Buracchi C, Sironi M, De Bortoli M, Muzio M, Bergottini R, Scanziani E, Vecchi A, et al. Intestinal inflammation in mice deficient in tir8, an inhibitory member of the il-1 receptor family. Proceedings of the National Academy of Sciences of the United States of America 2004;101(10):3522-3526.

23.

van der Windt GJ, Blok DC, Hoogerwerf JJ, Lammers AJ, de Vos AF, Vanâ&#x20AC;&#x2122;t Veer C, Florquin S, Kobayashi KS, Flavell RA, van der Poll T. Interleukin 1 receptor-associated kinase m impairs host defense during pneumococcal pneumonia. The Journal of infectious diseases 2012;205(12):1849-1857.

24.

van der Windt GJ, Hoogendijk AJ, Schouten M, Hommes TJ, de Vos AF, Florquin S, van der Poll T. Osteopontin impairs host defense during pneumococcal pneumonia. The Journal of infectious diseases 2011;203(12):1850-1858.

25.

Hoogendijk AJ, Roelofs JJ, Duitman J, van Lieshout MH, Blok DC, van der Poll T, Wieland CW. R-roscovitine reduces lung inflammation induced by lipoteichoic acid and streptococcus pneumoniae. Molecular medicine 2012;18:1086-1095.

26.

Blok DC, van der Sluijs KF, Florquin S, de Boer OJ, van â&#x20AC;&#x2DC;t Veer C, de Vos AF, van der Poll T. Limited anti-inflammatory role for interleukin-1 receptor like 1 (st2) in the host response to murine postinfluenza pneumococcal pneumonia. PloS one 2013;8(3):e58191.

205

Chapter 9 27.

Kager LM, Wiersinga WJ, Roelofs JJ, Meijers JC, Levi M, Van’t Veer C, van der Poll T. Plasminogen activator inhibitor type i contributes to protective immunity during experimental gram-negative sepsis (melioidosis). Journal of thrombosis and haemostasis : JTH 2011;9(10):2020-2028.

28.

Wald D, Qin J, Zhao Z, Qian Y, Naramura M, Tian L, Towne J, Sims JE, Stark GR, Li X. Sigirr, a negative regulator of toll-like receptor-interleukin 1 receptor signaling. Nature immunology 2003;4(9):920-927.

29.

Rijneveld AW, Florquin S, Branger J, Speelman P, Van Deventer SJ, van der Poll T. Tnf-alpha compensates for the impaired host defense of il-1 type i receptor-deficient mice during pneumococcal pneumonia. J Immunol 2001;167(9):5240-5246.

30.

Lauw FN, Branger J, Florquin S, Speelman P, van Deventer SJ, Akira S, van der Poll T. Il18 improves the early antimicrobial host response to pneumococcal pneumonia. J Immunol 2002;168(1):372-378.

31.

van der Poll T, Keogh CV, Buurman WA, Lowry SF. Passive immunization against tumor necrosis factor-alpha impairs host defense during pneumococcal pneumonia in mice. American journal of respiratory and critical care medicine 1997;155(2):603-608.

32.

Skerrett SJ, Martin TR, Chi EY, Peschon JJ, Mohler KM, Wilson CB. Role of the type 1 tnf receptor in lung inflammation after inhalation of endotoxin or pseudomonas aeruginosa. The American journal of physiology 1999;276(5 Pt 1):L715-727.

33.

Schultz MJ, Knapp S, Florquin S, Pater J, Takeda K, Akira S, van der Poll T. Interleukin-18 impairs the pulmonary host response to pseudomonas aeruginosa. Infection and immunity 2003;71(4):1630-1634.

34.

Schultz MJ, Rijneveld AW, Florquin S, Edwards CK, Dinarello CA, van der Poll T. Role of interleukin-1 in the pulmonary immune response during pseudomonas aeruginosa pneumonia. American journal of physiology Lung cellular and molecular physiology 2002;282(2):L285290.

35.

Huang X, Hazlett LD, Du W, Barrett RP. Sigirr promotes resistance against pseudomonas aeruginosa keratitis by down-regulating type-1 immunity and il-1r1 and tlr4 signaling. J Immunol 2006;177(1):548-556.

36.

Bozza S, Zelante T, Moretti S, Bonifazi P, DeLuca A, D’Angelo C, Giovannini G, Garlanda C, Boon L, Bistoni F, et al. Lack of toll il-1r8 exacerbates th17 cell responses in fungal infection. J Immunol 2008;180(6):4022-4031.

37.

Leemans JC, Butter LM, Teske GJ, Stroo I, Pulskens WP, Florquin S. The toll interleukin-1 receptor (il-1r) 8/single ig domain il-1r-related molecule modulates the renal response to bacterial infection. Infection and immunity 2012;80(11):3812-3820.

38.

Doyle SE, O’Connell RM, Miranda GA, Vaidya SA, Chow EK, Liu PT, Suzuki S, Suzuki N, Modlin RL, Yeh WC, et al. Toll-like receptors induce a phagocytic gene program through p38. The Journal of experimental medicine 2004;199(1):81-90.

39.

Ferreira R, Santos T, Viegas M, Cortes L, Bernardino L, Vieira OV, Malva JO. Neuropeptide y inhibits interleukin-1beta-induced phagocytosis by microglial cells. Journal of neuroinflammation 2011;8:169.

206

Single immunoglobulin interleukin-1 receptor related molecule impairs host defense during pneumonia and sepsis caused by Streptococcus pneumoniae 40.

Henan X, Toyota N, Yanjiang X, Fujita Y, Zhijun H, Touma M, Qiong W, Sugimoto K. Enhancement of phagocytosis and cytotoxicity in macrophages by tumor-derived il-18 stimulation. BMB reports 2013.

207

208

Chapter 10 TLR4 inhibition impairs bacterial clearance in a therapeutic setting in murine abdominal sepsis Inflammation Research 2014 Nov;63(11):927-33. doi: 10.1007/s00011-014-0766-9 Miriam H.P. van Lieshout 1,2 Tom van der Poll 1,2,3 Cornelis van â&#x20AC;&#x2122;t Veer 1,2 Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands: 1 Center of Infection and Immunity 2 Center of Experimental and Molecular Medicine 3 Division of Infectious Diseases

Chapter 10 TLR4 inhibition impairs bacterial clearance in a therapeutic setting in murine abdominal sepsis

Abstract Objective and design: to investigate the therapeutic effect of E5564 (a clinically used TLR4 inhibitor) in murine abdominal sepsis elicited by intraperitoneal infection with a highly virulent Escherichia coli in the context of concurrent antibiotic therapy. Methods: Mice were infected with different doses (~2 x 104 - 2 x 106 CFU) of E. coli O18:K1 and treated after 8 hours with ceftriaxone 20 mg/kg i.p. combined with either E5564 10mg/kg i.v. or vehicle. For survival studies this treatment was repeated every 12 hours. Bacterial loads and inflammatory parameters were determined after 20 hours in peritoneal lavage fluid, blood, liver and lung tissue. Plasma creatinin, AST, ALT and LDH were determined to assess organ injury. Results: E5564 impaired bacterial clearance under the antibiotic regime after infection with a low dose E. coli (1.7 x 104 CFU) while renal function was slightly preserved. No differences were observed in bacterial load and organ damage after infection with a tenfold higher (1.7 × 105 E. coli) bacterial dose. While treatment with E5564 slightly attenuated inflammatory markers provoked by the sublethal doses of 104–105 E. coli under the antibiotic regime, it did not affect lethality evoked by infection with 1.7 × 106 E. coli. Conclusions: The impact of TLR4 inhibition during abdominal sepsis by virulent E. coli bacteria is only beneficial at low infection grade at cost of bactericidal activity.

210

TLR4 inhibition impairs bacterial clearance in a therapeutic setting in murine abdominal sepsis

Introduction Sepsis is a leading and increasing cause of death in non-coronary intensive care units (1). The abdomen is the second most common site of primary infection in sepsis (2). Sepsis is frequently caused by gram-negative pathogens, amongst which Escherichia (E.) coli is still one of the most frequently isolated organisms (2). Despite all efforts to improve therapeutic outcome, the mortality rate of sepsis remains as high as 20-40%. The inflammatory response to pathogens is initiated by the recognition of bacterial structures by Pattern Recognition Receptors (PRRs) such as Toll-like receptors (TLRs). The proinflammatory host response is on the one hand essential to combat the infection adequately, but on the other hand can contribute to systemic inflammation and tissue injury, such as occurs during severe sepsis. TLR4 is central in the pathway that initiates and amplifies the inflammatory response since it not only recognizes exogenous microbial ligands but also endogenous ligands released from damaged tissues and dying cells collectively called alarmins (1, 3). Therefore, this pathway is regarded as a potential therapeutic target for sepsis. A highly potent synthetic inhibitor of TLR4, known as E5564 (Eritoran®) that binds to the essential co-receptor of LPS-signaling MD2 in a competitive way, was developed and shown to block LPS activity in vitro and in vivo (4). In 2011 the phase III multicenter “ACCESS” trial study, designed to study E5564 in patients with severe sepsis and a high mortality risk, was completed; although the results have not been published yet, E5564 was reported not to be effective in reducing 28-day mortality (5). We here determine the effect of therapeutic administration of E5564 in a model of gram-negative sepsis in a clinically relevant setting, i.e. in the context of postponed treatment together with concomitant antibiotic therapy and observed that TLR4 is involved in bacterial clearance under these conditions.

Methods Mice Specific pathogen-free female C57BL/6 mice were obtained from Harlan (Horst, the Netherlands) (6-8). Mice were 10 weeks of age at the start of experiments. All experiments were approved by the Animal Care and Use Committee of the University of Amsterdam. Experimental design Peritonitis was induced as described (9) by intraperitoneal injection of different dosages of viable E. coli O18:K1 in 200 μl pyrogen free 0.9% NaCl (Baxter) as described. In the first experiment survival of mice was monitored after infection with increasing doses of E. coli (2 x 104 to 2x 106 CFU per mouse) with or without i.p. antibiotic treatment with 20 mg/kg ceftriaxone (Bipharma, Almere, the Netherlands) or vehicle (200 μl sterile 0.9% NaCl) starting from 8 hours after infection, this was repeated every 12 hours (10-12). Ceftriaxone is used against Gram-positive and Gram-negative bacteria and commonly used for treatment of pneumonia, bacterial 211

Chapter 10

meningitis, sepsis and other bacterial infections (13). In subsequent experiments mice received ceftriaxone i.p. and 10 mg/kg E5564 or vehicle (200 μl 0.9% NaCl) intravenously via the tail vein 8 hours after infection and survival was monitored or mice were sacrificed 20 hours after infection, and peritoneal lavage fluid (PLF), blood, liver and lung were harvested, processed and analyzed for CFU levels as described (9). We were advised on the treatment dose of E5564 by the supplier. In the survival studies treatments were repeated every 12 hours. Assays Tumor necrosis factor (TNF)-α, interleukin (IL)-6, IL-10 and monocyte chemoattractant protein (MCP)-1 were measured by cytometric bead array (BD Biosciences, San Jose, CA). IL-1β, cytokine-induced neutrophil chemoattractant (KC), macrophage inflammatory protein (MIP)-2 and soluble E-selectin were measured by ELISA’s (R&D Systems, Minneapolis, MN). Creatinin, AST, ALT and LDH were measured by kits from Sigma (St. Louis, MO), using a Hittachi analyzer (Boehringer Mannheim, Mannheim, Germany). Statistics Data are expressed as Kaplan Meier plots (survival curves), medians with individual data points (bacterial loads) or means ± standard error of the mean. Comparison between groups was done by Mann Whitney U tests. p < 0.05 was considered statistically significant.

Results TLR4 inhibition does not affect survival, but impairs bacterial clearance during abdominal sepsis in the context of antibiotic treatment First, we aimed to study the effect of delayed combined treatment with antibiotics and TLR4 inhibition in a lethal model of abdominal sepsis caused by intraperitoneal infection with E. coli. To determine the infectious dose that was lethal despite antibiotic treatment, we first performed a pilot study in which mice were infected with increasing doses of E. coli and were treated with ceftriaxone or vehicle (n=5 per treatment group) after 8 hours and every subsequent 12 hours. Mice were followed until 48 hours (Figure 1). Ceftriaxone prevented mortality of mice after infection with 2 x 104 and 2 x 105 E. coli, but not the lethality provoked by 2 x 106 E. coli. Based on these pilot data, we infected two groups of 20 mice with 2 x 106 CFU and subjected them to antibiotic treatment with or without E5564 and scored survival rates. From 15 hours onward mortality occurred in both treatment groups and survival curves were not different after treatment with E5564 (median survival 26 hours in both groups; Figure 2A). To study the effect of E5564 on antibacterial defense and the inflammatory response in this setting of delayed antibiotic treatment, we performed separate experiments in which we infected mice with sublethal doses of 1.7 x 104 or 1.7 x 105 CFU and sacrificed them after 20 hours. After infection with 1.7 x 104 E. coli, cultures of PLF and blood showed no growth of bacteria in the vast 212

TLR4 inhibition impairs bacterial clearance in a therapeutic setting in murine abdominal sepsis

Figure 1: Pilot survival study to establish lethality with different infectious inocula in the context of antibiotic treatment. Mice were inoculated with three different 10-fold increasing doses of E. coli intraperitoneally (i.p.) and treated after 8 hours with 20mg/kg ceftriaxone (open symbols) or vehicle i.p. (closed symbols) (n=5 per treatment group); treatment was repeated twice daily and mice were followed until 48 hours.

majority of mice irrespective of E5564 treatment (Figure 2B, C). In liver and lungs however, median bacterial loads were almost 10-fold higher in E5564 treated mice (P < 0.05 versus vehicle for liver and P < 0.01 versus vehicle for lung) (Figure 2D, E) indicating that E5564 decreased host defense in organ parenchyma. In contrast, in mice infected with a 10-fold higher dose (Figure 2B-E) of E.coli (1.7 x 105) E5564 did not influence bacterial loads after infection with the higher dose (Figure 2B-E). Strikingly, bacterial loads in lungs of mice that were inoculated with 1.7 x 104 CFU E. coli and treated with E5564 were higher than in the mice that received 1.7 x 105 CFU E. coli treated with E5564. The latter indicates that the number of bacteria found in the organs is not just a function of the inoculated dose, but also depends on host defense. Indeed killing of the E. coli type O18:K1 bacterium, is dependent on cytokine activated macrophages (14) and thus we evaluated the host response. Effect of TLR4 inhibition on inflammatory response To evaluate the effect of delayed TLR4 inhibition on the inflammatory response we determined levels of proinflammatory cytokines (TNF-Îą, IL-6, IL-1Î˛), an antiinflammatory cytokine (IL-10) and chemokines (KC, MIP-2, MCP-1, LIX) at the primary site of infection and in the circulation. Although trends were observed, E5564 did not significantly downregulate cytokine or chemokine levels in PLF (relative to treatment with ceftriaxone plus vehicle; Table 1). Moreover, there were no differences in total cell numbers and composition of the recruited cells in PLF between treatment groups (data not shown). Plasma levels of cytokines and chemokines tended to be lower in both E5564 treated groups, which was significant for MCP-1 (p < 0.05 and p < 0.01 versus vehicle treated groups). As a read-out for endothelial cell activation we determined the plasma levels of soluble E-selectin. TLR4 inhibition was associated with lower plasma soluble E-selectin concentrations after infection with the higher E. coli dose (p < 0.01 versus vehicle).

213

Chapter 10

Figure 2: TLR4 inhibition impairs bacterial clearance during abdominal sepsis in the therapeutic setting with concurrent antibiotic treatment. Mice were inoculated with E. coli intraperitoneally (i.p.) and treated after 8 hours with 20mg/kg ceftriaxone i.p. and 10 mg/kg E5564 (open rounds) or vehicle (closed rounds) intravenously; treatment was repeated twice daily during survival study. Survival from mice infected with 2 x 106 CFU E. coli (n=20 mice per treatment group) (A). Bacterial loads in peritoneal lavage fluid (PLF) (B), blood (C), liver (D) and lung (E) 20 hours after infection with 1.7 x 104 (n=8 per treatment group) or 1.7 x 105 CFU E coli ( n=8 per treatment group). Each symbol represents an individual mouse, with horizontal lines showing medians. ** p < 0.01 versus vehicle treated group that was infected with the same dose of bacteria, determined with Mann-Whitney U test.

Effect of TLR4 inhibition on tissue injury E5564 preserved renal function in mice infected with the lower E. coli dose as reflected by lower plasma levels of creatinin (Figure 3A, p < 0.05 versus vehicle). E5564 did not impact on hepatocellular injury as evaluated by the plasma concentrations of ALT and AST, nor did it attenuate cellular injury in general as indicated by plasma LDH (Figure 3B-D).

214

TLR4 inhibition impairs bacterial clearance in a therapeutic setting in murine abdominal sepsis Table 1: Effect of TLR4 inhibition on peritoneal and plasma levels of cytokines and chemokines.

1.7 x 104 CFU Ceftriaxone vehicle

1.7 x 105 CFU + Ceftriaxone + E5564

Ceftriaxone vehicle

+ Ceftriaxone + E5564

PLF IL-1β

174 ± 26

228 ± 25

303 ± 72

164 ± 22

TNF-α

11 ± 3

9±2

IL-6

4 ± 0.5

177 ± 126

48 ± 33

IL-10

53 ± 3

633 ± 355

200 ± 108

MCP-1

90 ± 15

66 ± 18

669 ± 298

397 ± 180

MIP-2

424 ± 36

393 ± 45

1051 ± 391

635 ± 107

LIX

177 ± 64

78 ± 35

TNF-α

121 ± 75

36± 15

IL-6

4449 ± 1753

115 ± 72

13 ± 7

IL-10

127 ± 88

88 ± 20

MCP-1

3940 ±47

52 ±11**

1568 ± 713

189 ± 47*

E-selectin

275 ± 57

240 ± 29

338 ±54

144 ±31**

Plasma

Mice were inoculated with 1.7 x 104 (n = 8 per treatment group) or 1.7 x 105 CFU (n = 8 per treatment group) E. coli intraperitoneally (i.p.) and treated after 8 hours with 20mg/kg ceftriaxone i.p. and 10 mg/ kg E5564 or vehicle intravenously. Data are means ± standard error of the mean. Cytokines and chemokines levels are in pg/mL; soluble E-selectin levels in ng/mL. PLF = peritoneal lavage fluid.* p < 0.05, ** p < 0.01 versus vehicle treated group that was infected with the same dose of bacteria, determined with Mann-Whitney U test. Bd= below detection level.

215

Chapter 10

Figure 3: TLR4 inhibition preserves renal function during abdominal sepsis without influencing (hepato)cellular injury. Mice were inoculated with 1.7 x 104 (n = 8 per treatment group) or 1.7 x 105 CFU (n= 8 per treatment group) E. coli intraperitoneally (i.p.) and treated after 8 hours with 20mg/kg ceftriaxone i.p. and 10 mg/kg E5564 or vehicle intravenously. Plasma levels (20 hours post infection) of creatinin (A), ALT (B), AST (C) and LDH (D). Bars represent mean Âą standard error of the mean. * p < 0.05 versus vehicle treated group that was infected with the same dose of bacteria, determined with Mann-Whitney U test.

Discussion Several adjunctive therapies to improve sepsis outcome have been evaluated in clinical trials during the last decades, almost invariably yielding negative results (15). Anti-TLR4 therapy was designed as an additional therapy to attenuate excessive inflammation caused by high bacterial loads and the release of endogenous â&#x20AC;&#x153;danger moleculesâ&#x20AC;?. On the other hand, TLR4 is important for bacterial clearance of several gram-negative pathogens (9, 16, 17). We here studied the effects of TLR4 inhibition in a murine model of abdominal sepsis caused by E. coli and revealed part of the dual function of TLR4: after infection with a relatively low bacterial inoculum E5564 treatment impaired bacterial clearance in the presence of concurrent antibiotic therapy but at the same time attenuated the cytokine response and preserved renal function. Previous studies have shown that anti-TLR4 directed therapy can be protective during rodent peritonitis (5, 17-21). In the model of colon ascendens stent peritonitis anti-TLR4 was protective if initiated immediately upon the intervention (18). In two separate reports, antibodies directed against TLR4 reduced mortality from E. coli peritonitis if administered prophylactically in infection models with high doses of less virulent bacteria and antibiotics (19, 20) or therapeutically in a postponed treatment 216

TLR4 inhibition impairs bacterial clearance in a therapeutic setting in murine abdominal sepsis

setting in combination with TLR2 inhibition (19, 20). This suggests that excessive TLR4 mediated hyperinflammation is important in causing lethality during bacterial peritonitis. We recently reported an impaired antibacterial defense in TLR4 deficient mice, especially in early stage E. coli peritonitis by infection with a relatively low amount of highly virulent bacteria (9). In the present study we established by use of the MD2/TLR4 inhibitor E5564 (22, 23) that TLR4 is important for antibacterial defense even during progressive infection after infection with a relatively low dose of bacteria (i.e. when inhibited 8 hours after infection) and in the presence of antibiotic treatment. TLR4 inhibition did not impact on bacterial loads after high dose infection, suggesting that in this condition after 8 hours the time window during which TLR4 contributes to antibacterial defense has passed or becomes redundant. The differential effect of anti-TLR4 treatment on antibacterial defense after infection with increasing doses might be explained by the different amounts of LPS that induce the production of cytokines, which in turn, initiate macrophages to kill E. coli (14). Indeed mice infected with the high dose E. coli displayed high cytokine/chemokine levels at the primary (peritoneal) site of infection irrespective of TLR4 inhibition (table 1). It is noteworthy that the irritants from the peritoneal cavity are drained by the thoracic duct via the subclavial vein in the lung circulation. Potentially peritoneal originated mediators aid in bacterial control in the lung and are involved in our observation that the bacterial load is actually lower in the lungs of 1.7 x 105 infected E5564 treated mice compared to E5564 treated mice infected with 1.7 x 104 E. coli (Figure 2E). Moreover, this could be due to the low peritoneal as well as low systemic cytokine/chemokine levels in the 1.7 x 104 infected E5564 treated group (table 1). Alternatively, E5564 treatment could impair phagocytosis, since the phagocytic capacity of macrophages has been reported to be enhanced by TLR4 stimulation and MD-2 was reported to function as an opsonin (24, 25). Overall the results indicate that antibiotic control of established E. coli infection in the organs falls short when TLR4 is inhibited. Postponed TLR4 inhibition attenuated the systemic levels of cytokines to some extent and partially preserved renal function. This is in line with recent reports that TLR4 is involved in the pathophysiology of sepsis induced acute kidney injury (26, 27). In accordance, E5564 was reported to be protective in a model of renal ischemia-reperfusion injury (28). Moreover, E5564 treated mice showed lower circulating levels of soluble E-selectin, indicating that TLR4 is involved in endothelial cell activation during progressive abdominal sepsis. Discrepancies between the previous reports on the role of TLR4 during murine peritonitis (1820) and our earlier (9) and current reports may be largely due to differences in the models: we administered relatively low doses of a highly pathogenic E.coli O18:K1 strain that induces 80-100% lethality if not treated (9), while in the previous studies 1000-100,000x higher doses of less virulent E. coli were administered (18-20). Also, we used a MD2/TLR4 inhibitor which binds only to MD2, and it is under debate whether alarmin signalling by TLR4 is MD2 dependent (29). Furthermore, in here we used the antibiotic ceftriaxone that is reported to have an adequate penetration in lung tissue and abdominal organs (30, 31) and showed that although blood and peritoneal cavity indeed became virtually sterile, E. coli may persist in organs such 217

Chapter 10

as the liver and the lungs during the initial phase of TLR4 treatment. It is clear that at high bacterial loads in addition to TLR4 other TLRs such as TLR2 are important in the recognition of E. coli (9, 20, 32), while for the initial detection of a lower dose of (pathogenic) bacteria TLR4 is most important as our group demonstrated in vivo and in vitro (9). Specifically, TLR2 can be activated by Peptidoglycan-associated lipoprotein (PAL), a constituent of the outer membrane of E. coli (33).

Conclusions This study illustrates that therapeutic targeting of TLR4 signalling can potentially interfere with bacterial clearance even in the context of antibiotic therapy, but on the other hand can help limit some aspects of tissue injury depending on the infectious dose and kinetics of the infection model.

Acknowledgments We acknowledge Eisai Inc., Woodcliff Lake, NJ for the supply of E5564 and Joost Daalhuisen and Marieke ten Brink for expert technical assistance. This work was supported by the AMC Graduate School of Medical Science (to M. H. P. v. L.)

218

TLR4 inhibition impairs bacterial clearance in a therapeutic setting in murine abdominal sepsis

References

1. van der Poll T, Opal SM. Host-Pathogen Interactions in Sepsis. Lancet Infect Dis 2008;8:3243.

2. Ranieri VM, Thompson BT, Barie PS, Dhainaut JF, Douglas IS, Finfer S, Gardlund B, Marshall JC, Rhodes A, Artigas A, et al. Drotrecogin Alfa (Activated) in Adults With Septic Shock. N Engl J Med 2012;366:2055-2064.

3. Rittirsch D, Flierl MA, Ward PA. Harmful Molecular Mechanisms in Sepsis. Nat Rev Immunol 2008;8:776-787.

4. Barochia AV, Cui X, Natanson C, Eichacker PQ. Risk of Death and the Efficacy of Eritoran Tetrasodium (E5564): Design Considerations for Clinical Trials of Anti-Inflammatory Agents in Sepsis. Crit Care Med 2010;38:306-308.

5. Angus DC. The Search for Effective Therapy for Sepsis: Back to the Drawing Board? JAMA 2011;306:2614-2615.

6. Renckens R, Roelofs JJ, Florquin S, van der PT. Urokinase-Type Plasminogen Activator Receptor Plays a Role in Neutrophil Migration During Lipopolysaccharide-Induced Peritoneal Inflammation but Not During Escherichia Coli-Induced Peritonitis. J Infect Dis 2006;193:522530.

7. Renckens R, Roelofs JJ, Florquin S, de Vos AF, Pater JM, Lijnen HR, Carmeliet P, van ‘, V, van der PT. Endogenous Tissue-Type Plasminogen Activator Is Protective During Escherichia Coli-Induced Abdominal Sepsis in Mice. J Immunol 2006;177:1189-1196.

8. Schouten M, Van’t Veer C, Levi M, Esmon CT, van der Poll T. Endogenous Protein C Inhibits Activation of Coagulation and Transiently Lowers Bacterial Outgrowth in Murine Escherichia Coli Peritonitis. J Thromb Haemost 2011;9:1072-1075.

9. van ‘t Veer C, van den Pangaart PS, Kruijswijk D, Florquin S, de Vos AF, van der Poll T. Delineation of the Role of Toll-Like Receptor Signaling During Peritonitis by a Gradually Growing Pathogenic Escherichia Coli. J Biol Chem 2011;286:36603-36618.

10. Turnbull IR, Wlzorek JJ, Osborne D, Hotchkiss RS, Coopersmith CM, Buchman TG. Effects of Age on Mortality and Antibiotic Efficacy in Cecal Ligation and Puncture. Shock 2003;19:310313.

11. Rijneveld AW, Florquin S, Hartung T, Speelman P, van der PT. Anti-Tumor Necrosis Factor Antibody Impairs the Therapeutic Effect of Ceftriaxone in Murine Pneumococcal Pneumonia. J Infect Dis 2003;188:282-285.

12. Wang E, Bergeron Y, Bergeron MG. Ceftriaxone Pharmacokinetics in Interleukin-10-Treated Murine Pneumococcal Pneumonia. J Antimicrob Chemother 2005;55:721-726.

13. Gladwin M. Clinical Microbiology Made Ridiculously Simple, 4th ed. Miami Med Master, Inc.; 2007.

14. Cross A, Asher L, Seguin M, Yuan L, Kelly N, Hammack C, Sadoff J, Gemski P, Jr. The Importance of a Lipopolysaccharide-Initiated, Cytokine-Mediated Host Defense Mechanism in Mice Against Extraintestinally Invasive Escherichia Coli. J Clin Invest 1995;96:676-686.

219

Chapter 10

15. Webster NR, Galley HF. Immunomodulation in the Critically Ill. Br J Anaesth 2009;103:70-81.

16. Branger J, Knapp S, Weijer S, Leemans JC, Pater JM, Speelman P, Florquin S, van der Poll T. Role of Toll-Like Receptor 4 in Gram-Positive and Gram-Negative Pneumonia in Mice. Infect Immun 2004;72:788-794.

17. Wieland CW, van Lieshout MH, Hoogendijk AJ, van der Poll T. Host Defence During Klebsiella Pneumonia Relies on Haematopoietic-Expressed Toll-Like Receptors 4 and 2. Eur Respir J 2011;37:848-857.

18. Daubeuf B, Mathison J, Spiller S, Hugues S, Herren S, Ferlin W, Kosco-Vilbois M, Wagner H, Kirschning CJ, Ulevitch R, et al. TLR4/MD-2 Monoclonal Antibody Therapy Affords Protection in Experimental Models of Septic Shock. J Immunol 2007;179:6107-6114.

19. Roger T, Froidevaux C, Le RD, Reymond MK, Chanson AL, Mauri D, Burns K, Riederer BM, Akira S, Calandra T. Protection From Lethal Gram-Negative Bacterial Sepsis by Targeting Toll-Like Receptor 4. Proc Natl Acad Sci U S A 2009;106:2348-2352.

20. Spiller S, Elson G, Ferstl R, Dreher S, Mueller T, Freudenberg M, Daubeuf B, Wagner H, Kirschning CJ. TLR4-Induced IFN-Gamma Production Increases TLR2 Sensitivity and Drives Gram-Negative Sepsis in Mice. J Exp Med 2008;205:1747-1754.

21. Solomon SB, Cui X, Gerstenberger E, Danner RL, Fitz Y, Banks SM, Natanson C, Eichacker PQ. Effective Dosing of Lipid A Analogue E5564 in Rats Depends on the Timing of Treatment and the Route of Escherichia Coli Infection. J Infect Dis 2006;193:634-644.

22. Visintin A, Halmen KA, Latz E, Monks BG, Golenbock DT. Pharmacological Inhibition of Endotoxin Responses Is Achieved by Targeting the TLR4 Coreceptor, MD-2. J Immunol 2005;175:6465-6472.

23. Wasan KM, Sivak O, Cote RA, MacInnes AI, Boulanger KD, Lynn M, Christ WJ, Hawkins LD, Rossignol DP. Association of the Endotoxin Antagonist E5564 With High-Density Lipoproteins in Vitro: Dependence on Low-Density and Triglyceride-Rich Lipoprotein Concentrations. Antimicrob Agents Chemother 2003;47:2796-2803.

24. Deng T, Feng X, Liu P, Yan K, Chen Y, Han D. Toll-Like Receptor 3 Activation Differentially Regulates Phagocytosis of Bacteria and Apoptotic Neutrophils by Mouse Peritoneal Macrophages. Immunol Cell Biol 2012.

25. Jain V, Halle A, Halmen KA, Lien E, Charrel-Dennis M, Ram S, Golenbock DT, Visintin A. Phagocytosis and Intracellular Killing of MD-2 Opsonized Gram-Negative Bacteria Depend on TLR4 Signaling. Blood 2008;111:4637-4645.

26. Castoldi A, Braga TT, Correa-Costa M, Aguiar CF, Bassi EJ, Correa-Silva R, Elias RM, Salvador F, Moraes-Vieira PM, Cenedeze MA, et al. TLR2, TLR4 and the MYD88 Signaling Pathway Are Crucial for Neutrophil Migration in Acute Kidney Injury Induced by Sepsis. PLoS ONE 2012;7:e37584.

27. Watts BA, III, George T, Sherwood ER, Good DW. A Two-Hit Mechanism for Sepsis-Induced Impairment of Renal Tubule Function. Am J Physiol Renal Physiol 2013.

28. Liu M, Gu M, Xu D, Lv Q, Zhang W, Wu Y. Protective Effects of Toll-Like Receptor 4 Inhibitor Eritoran on Renal Ischemia-Reperfusion Injury. Transplant Proc 2010;42:1539-1544.

220

TLR4 inhibition impairs bacterial clearance in a therapeutic setting in murine abdominal sepsis

29. Chun KH, Seong SY. CD14 but Not MD2 Transmit Signals From DAMP. Int Immunopharmacol 2010;10:98-106.

30. Martin C, Ragni J, Lokiec F, Guillen JC, Auge A, Pecking M, Gouin F. Pharmacokinetics and Tissue Penetration of a Single Dose of Ceftriaxone (1,000 Milligrams Intravenously) for Antibiotic Prophylaxis in Thoracic Surgery. Antimicrob Agents Chemother 1992;36:2804-2807.

32. Elson G, Dunn-Siegrist I, Daubeuf B, Pugin J. Contribution of Toll-Like Receptors to the Innate Immune Response to Gram-Negative and Gram-Positive Bacteria. Blood 2007;109:15741583.

33. Liang MD, Bagchi A, Warren HS, Tehan MM, Trigilio JA, Beasley-Topliffe LK, Tesini BL, Lazzaroni JC, Fenton MJ, Hellman J. Bacterial Peptidoglycan-Associated Lipoprotein: a Naturally Occurring Toll-Like Receptor 2 Agonist That Is Shed into Serum and Has Synergy With Lipopolysaccharide. J Infect Dis 2005;191:939-948.

221

222

Chapter 11 Summary and general discussion of this thesis “Cell-specific pattern recognition receptor signaling in antibacterial defense”

Miriam H.P. van Lieshout

Chapter 11

Summary Sepsis, the syndrome that describes infection complicated by acute organ failure, is most frequently caused by bacterial pneumonia and infection originating from the abdominal cavity and is a major cause of morbidity and mortality globally. Mortality rates remain high, despite the improvement of supportive therapies in the past decades. The global rise of antimicrobial resistance rates together with the fact that there currently is no prospect of the availability of new antimicrobial agents is alarming. A brisk and firm initial host response is needed for clearance of the pathogen, but on the other hand can induce tissue and organ injury. More insight in the initial immune response during pneumonia and sepsis potentially offers new targets for the development of therapeutic agents. Chapter 1 is a general introduction describing the mechanisms of the initial host response and the role of pattern recognition receptors during infection and the relevant disease models that were used in this thesis. In the first part of the thesis we focused on the role of Toll-like receptors (TLRs) and their intracellular adapter proteins myeloid differentiation primary response gene (MyD88) and TIR-domain-containing adapter-inducing interferon-Î˛ (TRIF) during pneumonia with the gram-negative pathogens Klebsiella (K.) pneumoniae and Pseudomonas (P.) aeruginosa. Specifically, we investigated the role of these proteins in different body compartments and cell-types. Chapter 2 describes the role of TLR2 and TLR4 during Klebsiella pneumonia. Tlr4-/-, but not Tlr2-/- mice demonstrated enhanced bacterial growth 24 hours after infection, indicating that TLR4 initially was more important for antibacterial defense than TLR2. However during later stage infection or after high dose infection, Tlr2-/- mice had higher bacterial loads than normal wild-type mice, and Tlr2/Tlr4 double deficient mice were more susceptible than Tlr4-/- mice. Moreover, using bone marrow chimeras, we demonstrated that hematopoietic TLR2 and TLR4 determined antibacterial capacity, while parenchymal TLRs did not contribute to a significant extent. This chapter shows that TLR4 is important for the initiation of the early host defense against K. pneumoniae, while TLR2 contributes to host defense against a higher infectious inoculum or during later stage infection. Moreover, TLR2 and TLR4 dependent signaling in hematopoietic cells is of primary importance to limit bacterial growth. In chapter 3 we report on the role of MyD88 and TRIF during K. pneumoniae infection and their respective roles in hematopoietic and non-hematopoietic cells. Both TLR adapters were crucial for antibacterial defense and survival, while Myd88-/- mice, that all died within 48 hours after infection, were even more susceptible than TRIF mutant mice. MyD88 in both hematopoietic and non-hematopoietic cells contributed to antibacterial defense during late stage infection and survival, while only TRIF in hematopoietic cells was protective. On the other hand, MyD88 in resident cells and TRIF in hematopoietic cells contributed to distant tissue injury. Early after infection, MyD88 in non-hematopoietic cells was crucial for the local production of chemokines and neutrophil attraction, while TRIF in both hematopoietic and non-hematopoietic cells was equally important. We here conclude that MyD88and TRIF-dependent signaling have a different contribution to the host defense in different cell types that changes during the course of gram-negative infection. 224

Summary and general discussion of this thesis “Cell-specific pattern recognition receptor signaling in antibacterial defense”

In chapter 4 we further studied the role of specific cell MyD88 signalling during K. pneumoniae infection. Mice deficient for MyD88 in myeloid (LysM-Myd88-/-) and myeloid plus endothelial (Tie2-Myd88-/-) cells showed enhanced mortality and higher bacterial loads. Tie2-Myd88-/- mice reconstituted with control bone marrow, representing mice with a selective MyD88 deficiency in endothelial cells, showed an unremarkable antibacterial defense. Myeloid or endothelial cell MyD88 deficiency did not impact on lung pathology or distant organ injury during late stage sepsis, while LysM-Myd88-/- mice demonstrated a strongly attenuated inflammatory response in the airways early after infection. These data suggest that myeloid but not endothelial MyD88 is important for host defense during gram-negative pneumonia derived sepsis. In chapter 5 we further focused on the role of TRIF during gram-negative pneumonia and especially its role in the induction of interferon (IFN)-γ. The impaired antibacterial defense of TRIF mutant mice was associated with absent IFN-γ production in the lungs. Furthermore, in vitro IFN-γ production by splenocytes in response to K. pneumoniae was critically dependent on TLR4, MyD88 and TRIF. When TRIF mutant mice were reconstituted with recombinant IFN-γ via the airways, bacterial loads were reduced in lungs and distant body sites similar to levels measured in wild-type mice, while pulmonary cytokine levels were partially restored. The IFN-γ induced improved enhanced antibacterial response in TRIF mutant mice occurred at the expense of increased hepatocellular injury. These data indicate that TRIF mediates antibacterial defense during gram-negative pneumonia at least in part by inducing IFN-γ at the primary site of infection. In chapter 6 we investigated the role of MyD88 dependent signaling during airway infection with P. aeruginosa. Sftpccre-Myd88-/- mice, that are deficient for MyD88 in lung epithelial cells, demonstrated an impaired early bacterial clearance of P. aeruginosa, as well as impaired neutrophil recruitment and a selective impairment of CCL20 secretion. Tlr5-/- mice also demonstrated an impaired early antibacterial and inflammatory response that was dependent on non-hematopoietic cells in bone marrow chimeras, findings pointing to an interaction between TLR5 expressed by epithelial cells with flagellin (TLR5 ligand) expressed by Pseudomonas. Indeed, by the use of an unflagellated Pseudomonas mutant we could demonstrate that the detection of flagellin by MyD88 in lung epithelial cells is crucial for the initiation of the host response. Together, these data indicate that recognition of Pseudomonas flagellin by epithelial cell TLR5-MyD88 initiates host defense to induce clearance of P. aeruginosa from the airways. In the next part of the thesis we focused on another component of the innate immune system demonstrated to be important for antimicrobial defense, the “NLRP3-inflammasome” (consisting of a protein complex formed by NLR family, pyrin domain containing 3 (NLRP3) and the adaptor apoptosis-associated specklike protein containing a CARD (ASC)), during pneumonia caused by Streptococcus (S.) pneumoniae, the most common causative agent in community-acquired pneumonia. Chapter 7 describes the role of NLRP3 and ASC in the host response during pneumococcal pneumonia with the serotype 2 D39 strain. Both Nlrp3-/and Asc-/- mice demonstrated impaired bacterial clearance from the lung, but only Asc-/- mice had increased mortality rates compared to wild-type mice. The early inflammatory response was disturbed in both genetically modified mouse strains as 225

Chapter 11

reflected by impaired cytokine secretion. Detailed analysis of the early inflammatory response in the lung by whole-genome transcriptional profiling identified several mediators that were differentially expressed between Nlrp3-/- and Asc-/- mice, possibly explaining the differences in lethality of these mice. These data confirmed a prominent role for the NLRP3-inflammasome during pneumococcal pneumonia and suggest that either ASC-dependent NLRP3-independent inflammasomes or inflammasome-independent ASC functions may be involved. In chapter 8, however, we demonstrate an opposite role of NLRP3 and ASC during infection with a serotype 3 pneumococcal strain. Notably, both Nlrp3-/- and Asc-/- mice showed a strongly improved host defense, as reflected by markedly improved survival rates and accompanied by diminished bacterial growth and dissemination. The early inflammatory response was slightly enhanced in Nlrp3-/- and Asc-/- mice, while lung inflammation and pathology were attenuated in Nlrp3-/- and Asc-/- mice during the late stages of the infection. Moreover, we investigated the contribution of MyD88 dependent signaling during infection with this specific virulent pneumococcal strain. Bacterial growth and survival were unaltered in Myd88-/- mice, although these mice demonstrated attenuated lung inflammation in the presence of high pneumococcal burdens. These data demonstrate that the contribution of proximal innate detection systems, such as the NLRP3-inflammasome and TLRs, can be dispensable depending on the pathogen and can vary between strains within the same bacterial species. In chapter 9 the role of Single immunoglobulin IL-1 receptor-related molecule (SIGIRR), a negative regulator of TLR- and IL-1 receptor dependent inflammation during pneumococcal airway infection and sepsis is described. Sigirr-/- mice demonstrated delayed mortality and had significantly lower bacterial loads both during pneumococcal airway and bloodstream infection, while the inflammatory response was not different. However, in vitro, Sigirr-/- alveolar macrophages and neutrophils demonstrated higher phagocytic rates, possibly explaining enhanced bacterial defense. Finally, we examined the effect of anti-TLR4 therapy, designed as an additional treatment to modulate excessive inflammation in human sepsis, in a model of murine abdominal sepsis. Chapter 10 shows that mice infected with a low infectious inoculum demonstrated higher bacterial loads after receiving anti-TLR4 therapy in a delayed treatment model of E. coli peritonitis, while organ injury was slightly preserved and survival was similar to the placebo treatment group. This illustrates that therapeutic targeting of TLR4 signaling can potentially interfere with bacterial clearance even in the context of antibiotic therapy, but on the other hand can help limit some aspects of tissue injury depending on the infectious dose and kinetics of the infection model.

226

Summary and general discussion of this thesis â&#x20AC;&#x153;Cell-specific pattern recognition receptor signaling in antibacterial defenseâ&#x20AC;?

General discussion With the experimental studies presented in this thesis we intended to gain more insight in the role of innate immune receptor signaling, including TLR dependent and especially MyD88-dependent pathways as well as inflammasome receptors, in experimental models of infection and sepsis. The focus was on pneumonia since this is the most common cause of sepsis. Secondly, the aim was to gain more insight in the contribution of different cell types and body compartments to TLR and MyD88 dependent signaling during infection and sepsis, considering that innate immune sensors are widely distributed among different cell types. In sepsis, inflammation is a key element of host defense, but on the other hand can result in tissue damage, a severe systemic inflammatory response and organ failure. Different cell types may contribute differentially to these inflammatory processes; this is potentially interesting for the development of new therapies. In the first part of the thesis, we focused on the role of TLR-dependent signaling during K. pneumoniae airway infection and sepsis. We observed a dynamic pattern in the contribution of different TLRs during the course of infection (chapter 2): TLR4 was essential for the early initiation of the host defense while TLR2 became important when bacterial loads were higher possibly (in part) by amplifying TLR4 signaling. These observations are in line with previous research on experimental E. coli peritonitis (1), despite the profound differences between these models of infection, underlining the importance of TLR4 and TLR2 during various infections. These findings are also in accordance with the fact that the phenotype of MyD88 deficient mice during K. pneumoniae infection was more severe than that of TRIF mutant mice (chapter 3), since MyD88 mediates the signals of both TLR2 and TLR4, while TRIF only participates in TLR4 dependent signaling. The observation that the expression of TLR2 and TLR4 on hematopoietic cells is sufficient to limit bacterial growth was corroborated by the finding that TRIF in hematopoietic cells determined bacterial growth during later stage infection and not the expression of TRIF in resident cells (chapter 3), reflecting the importance of TLR4 mediated signaling. However, the expression of MyD88 in both hematopoietic and resident cells was important for the antibacterial response during late stage Klebsiella infection (chapter 3), pointing to a role for MyD88 in the lung epithelium or other resident cells, similar as was demonstrated before in a model of P. aeruginosa airway infection (2). Indeed, during early stage infection MyD88 in resident cells was especially important for neutrophil attraction and cytokine production. We confirmed the importance of hematopoietic and especially myeloid MyD88 during late stage infection in chapter 4 and ruled out a contribution of endothelial MyD88 to antibacterial defense. However so far, we were not able to identify the resident cell type in which MyD88 dependent signaling confers protection in this model of gram-negative pneumonia since preliminary data show that MyD88 dependent signaling in type II lung alveolar cells or in Clara epithelial cells is not protective in this K. pneumoniae model. Possibly, our bone marrow chimera experiments were confounded by incomplete repopulation, or alternatively another (lung) cell type not targeted so far is more important. The finding that MyD88 in lung epithelial cells does 227

Chapter 11

not contribute to host defense during K. pneumoniae infection is in contrast with our studies on P. aeruginosa airway infection (chapter 6). Here, we demonstrate that MyD88 in lung epithelial cells is crucial for the early initiation of the host response and the recruitment of neutrophils, and that the detection of flagellin is involved in this process. In line with this, we demonstrate that TLR5 (that detects flagellin) is involved in the initiation of early antibacterial defense and that its expression on resident cells is most important. This was the first time that the role of TLR5 on parenchymal cells in Pseudomonas airway infection was directly demonstrated. The observations above illustrate that host defense during airway infection is a dynamic process and that the way that different TLRs and their adapters in various cell types contribute during the course of the infection vary depending on the causative pathogen and the stage of the infection. Our findings in chapter 7 and 8 further illustrate the complicated nature of the innate immune response, since we here demonstrate opposite roles for the inflammasome proteins ASC and NLRP3 during pneumococcal infection with two different bacterial strains: these proteins conferred protection in a model using one specific pneumococcal strain that is frequently used in experimental models, but played a detrimental role during infection with a virulent, clinically relevant S. pneumoniae strain. It is tempting to speculate that these different outcomes may depend on the net balance between the seemingly opposite roles of ASC and NLRP3 in key antibacterial responses, i.e., their ability to mediate bacteria-induced pyroptosis (an inflammatory form of cell death that may inhibit bacterial killing) and autophagy (a non-inflammatory form of cell death that is thought to attenuate pyroptosis) versus their ability to produce proinflammatory cytokines (favoring bacterial clearance); however the exact nature and effects of these processes have not been unraveled so far (3, 4). Our studies illustrate that the balance of benefit and harm that can result from these innate immune defense pathways can vary even after infection with pathogens within the same species. This is also illustrated by our finding that MyD88 dependent signaling during infection with a serotype 3 pneumococcus was not protective (chapter 8), while this was previously considered to be part of a protective host response after infection with moderately virulent pneumococcal strains (5); our group observed the same protective role of MyD88 during infection with the serotype 2 D39 strain (6). Our observations in chapter 9 further illustrate the complicated interactions of innate immune receptor signaling. In this chapter we found a negative effect of the negative TLR and IL-1R regulator SIGIRR on survival and bacterial loads during pneumonia and blood-stream infection caused by the serotype 3 S. pneumoniae also used in chapter 8. SIGIRR is known to exert inhibiting activity on TLR4, TLR7, TLR9, IL-1R type I (IL-1RI), IL-18R, and ST2 (7). In earlier research mice deficient for TLR4, TLR9, IL-1R or IL-18R were reported to have an impaired bacterial response against S. pneumoniae (8-11) and therefore these receptors might be involved as targets for SIGIRR in this case. However when considering the observations in chapter 8 of this thesis, the here observed role of SIGIRR is still difficult to interpret, since all of its target proteins are (partially) dependent on MyD88 mediated signaling, except for ST2 that was demonstrated to have a limited role during infection with 228

Summary and general discussion of this thesis â&#x20AC;&#x153;Cell-specific pattern recognition receptor signaling in antibacterial defenseâ&#x20AC;?

the same pathogen (12, 13). Possibly, the observed effect of SIGIRR results from the combined inhibition of multiple receptors or unknown interactions or feed-back mechanisms between them. Together, this part of the thesis illustrates that the effect of the various innate immune receptors on antibacterial host defense results of a complicated interplay between various (anti)inflammatory processes that moreover may be specific for each pathogen and stage of the infection. During infection, survival of the host is not only determined by the effectivity of the antimicrobial response and the clearance of the pathogen, but also by the severity of organ injury that is thought to occur as a result of a strong inflammatory response. This double edged sword character of the immune response during sepsis was clearly illustrated in some of our experimental models where the combat against bacteria occurred at the cost of organ injury. For example, during Klebsiella induced pneumonia derived sepsis organ injury was attenuated in mice deficient for MyD88 in resident cells as well as in mice deficient in TRIF in hematopoietic cells, while they had higher bacterial loads (chapter 3). Furthermore, improvement in antibacterial defense in TRIF deficient mice by the administration of IFN-Îł (chapter 5) occurred at the expense of organ injury. Eventually, survival was determined by uncontrolled bacterial growth rather than the degree of organ injury in this experimental model of gram-negative pneumosepsis without antibiotic treatment and was therefore impaired in mice deficient for TLR2/4 pathways or their adapters MyD88 and TRIF. However, we observed a similar effect of attenuated TLR4 signaling on the balance between antibacterial defense and organ injury in the delayed treatment model of peritonitis that was used in chapter 11. Here, the selective targeting of TLR4 slightly improved organ injury but still resulted in higher peripheral bacterial loads despite antibiotic therapy. Previously, there was some evidence pointing to a contribution of endothelial TLRdependent signaling in the development of organ injury during sepsis as well as in the antibacterial defense (14-18); however, we could not confirm this in our model of gram-negative pneumosepsis in chapter 4. Then again, there were major differences between the infection models that were used and the way inflammatory pathways in endothelial cells were targeted, including the use of Cre-promoters that are not completely cell-specific. In our study we corrected for the leakage of the endothelial promoter Tie2 to the hematopoietic compartment via bone marrow transplantation, which is why we believe our observations truly reflect that in this model there is no role for TLR dependent signaling in the endothelium in antibacterial defense or in organ injury. In conclusion in this thesis several key points on experimental sepsis were identified: 1) the crucial role of TLR2 and TLR4 and their intracellular adapters MyD88 and TRIF during gram-negative pneumosepsis and their respective contributions in different body compartments as well as during different stages of the infection. 2) the observation that the NLRP3 inflammasome can be both beneficial and detrimental for host defense in pneumococcal pneumonia, depending on the particular pathogen strain. 3) the inflammatory response contributes to organ injury, but on the other hand attenuation of this response by use of adjunctive anti-TLR4 229

Chapter 11

directed therapy may result in less effective bacterial clearance. Of course, the final question is how these findings in experimental murine sepsis relate to the pathophysiology and treatment of human sepsis. First of all, there are strong limitations in the translation of experimental sepsis studies to the human situation, but at the same time there is no other way to gather new insights. The available evidence is largely obtained from gene deficient mice, while the limitations of genetically modified mice as a model are well known. In humans, true genetic deficiencies for innate immune pathways are relatively rare, likely because they are under evolutionary pressure, but if they do occur the phenotype can be very different from that observed in mice (19). A strong example is provided by human MyD88 deficiency, which is associated with more severe and in a vast percentage even lethal infections in infancy and childhood albeit with only a small group of pathogens (S. pneumoniae, S. aureus and P. aeruginosa) (20). MyD88 deficient subjects are at a later age otherwise healthy, while murine MyD88 deficiency results in a higher susceptibility to virtually all pathogens. Furthermore, while several known human TLR polymorphisms may be associated with a somewhat higher susceptibility to some infectious diseases, this remains an issue of debate and in general do not explain a large part of the burden of sepsis (21, 22). The large body of evidence that is available from in vitro, animal and pre-clinical studies, as well as genetic studies on the role of TLRs in sepsis, has resulted in the design of TLR-directed therapies. Two different TLR4 inhibiting agents were tested in phase III studies in severe sepsis patients; both failed to improve 28 day mortality rates and further studies were suspended (13, 23). These failures illustrate the problems that were encountered with many other adjunctive sepsis therapies as well; in pre-clinical studies many demonstrated promising results but once studied in a clinical setting therapeutic advantages could not be established. The most likely obstacle in studying adjunctive therapies is that sepsis describes a syndrome, and represents a heterogeneous spectrum of different diseases and pathophysiologies. As described in this thesis, the contribution of different innate immune receptors varies between different experimental sepsis models; between different pathogens and even between pathogens of the same species as well as during the course of the infection. In human sepsis, little is known about the actual activity of different innate immune receptors, let alone about their kinetics or contribution in specific conditions. Importantly, in clinical practice the causative pathogen is initially unknown since traditional cultures at least need 24 hours to yield a positive result and are known to lack sensitivity. Moreover, the stage of the inflammatory response may differ between patients, i.e., the course of the infection may not be similar in all sepsis patients, unlike in experimental sepsis studies. Additionally, the host may or may not have substantial co-morbidity that alters the host response and outcome. Finally, unlike in experimental sepsis models, the quality of standard supportive care has improved substantially and is in most cases sufficient to prevent mortality in the first days; however, severe sepsis patients that do survive initially remain at high risk for mortality due to organ failure and complications, even beyond 28 days (20, 24). For all these reasons, the concept of early intervention in the host innate immune response via targeting of just one receptor or pathway in the whole 230

Summary and general discussion of this thesis â&#x20AC;&#x153;Cell-specific pattern recognition receptor signaling in antibacterial defenseâ&#x20AC;?

group of severe sepsis patients may be too simplistic and therefore not result in measurable treatment effects. The first step in the improvement of additional sepsis treatments may very well be earlier identification of the pathogen, preferable including resistance characteristics, to improve antimicrobial therapy. In addition, enhanced insight in the specifics of the derailed host response in individual patients is warranted using (sets of) biomarkers to better direct and monitor targeted immune modulatory therapies. Future research should seek to integrate detailed observational studies in patients with sepsis in different stages of the disease course with data derived from experimental sepsis models, and link the knowledge obtained with the development of novel targeted interventions and rapid beside tests to identify patients who might benefit from these interventions and to monitor their effect.

231

Chapter 11

References

1. van â&#x20AC;&#x2DC;t Veer C, van den Pangaart PS, Kruijswijk D, Florquin S, de Vos AF, van der Poll T. Delineation of the Role of Toll-Like Receptor Signaling During Peritonitis by a Gradually Growing Pathogenic Escherichia Coli. J Biol Chem 2011;286:36603-36618.

2. Hajjar AM, Harowicz H, Liggitt HD, Fink PJ, Wilson CB, Skerrett SJ. An Essential Role for Non-Bone Marrow-Derived Cells in Control of Pseudomonas Aeruginosa Pneumonia. Am J Respir Cell Mol Biol 2005;33:470-475.

3. Creagh EM. Caspase Crosstalk: Integration of Apoptotic and Innate Immune Signalling Pathways. Trends Immunol 2014;35:631-640.

4. Leissinger M, Kulkarni R, Zemans RL, Downey GP, Jeyaseelan S. Investigating the Role of Nucleotide-Binding Oligomerization Domain-Like Receptors in Bacterial Lung Infection. Am J Respir Crit Care Med 2014;189:1461-1468.

5. Albiger B, Sandgren A, Katsuragi H, Meyer-Hoffert U, Beiter K, Wartha F, Hornef M, Normark S, Normark BH. Myeloid Differentiation Factor 88-Dependent Signalling Controls Bacterial Growth During Colonization and Systemic Pneumococcal Disease in Mice. Cell Microbiol 2005;7:1603-1615.

6. de Vos AF, Dessing MC, Lammers AJ, de Porto AP, Florquin S, de Boer OJ, de BR, Terpstra S, Bootsma HJ, Hermans PW, et al. The Polysaccharide Capsule of Streptococcus Pneumonia Partially Impedes MyD88-Mediated Immunity During Pneumonia in Mice. PLoS ONE 2015;10:e0118181.

7. Riva F, Bonavita E, Barbati E, Muzio M, Mantovani A, Garlanda C. TIR8/SIGIRR Is an Interleukin-1 Receptor/Toll Like Receptor Family Member With Regulatory Functions in Inflammation and Immunity. Front Immunol 2012;3:322.

8. Albiger B, Dahlberg S, Sandgren A, Wartha F, Beiter K, Katsuragi H, Akira S, Normark S, Henriques-Normark B. Toll-Like Receptor 9 Acts at an Early Stage in Host Defence Against Pneumococcal Infection. Cell Microbiol 2007;9:633-644.

9. Branger J, Knapp S, Weijer S, Leemans JC, Pater JM, Speelman P, Florquin S, van der Poll T. Role of Toll-Like Receptor 4 in Gram-Positive and Gram-Negative Pneumonia in Mice. Infect Immun 2004;72:788-794.

10. Lauw FN, Branger J, Florquin S, Speelman P, Van Deventer SJ, Akira S, van der Poll T. IL-18 Improves the Early Antimicrobial Host Response to Pneumococcal Pneumonia. J Immunol 2002;168:372-378.

11. Rijneveld AW, Florquin S, Branger J, Speelman P, Van Deventer SJ, van der PT. TNF-Alpha Compensates for the Impaired Host Defense of IL-1 Type I Receptor-Deficient Mice During Pneumococcal Pneumonia. J Immunol 2001;167:5240-5246.

12. Blok DC, de Vos AF, Florquin S, van der Poll T. Role of Interleukin 1 Receptor Like 1 (ST2) in Gram-Negative and Gram-Positive Sepsis in Mice. Shock 2013;40:290-296.

13. Rice TW, Wheeler AP, Bernard GR, Vincent JL, Angus DC, Aikawa N, Demeyer I, Sainati S, Amlot N, Cao C, et al. A Randomized, Double-Blind, Placebo-Controlled Trial of TAK-242 for the Treatment of Severe Sepsis. Crit Care Med 2010;38:1685-1694.

232

Summary and general discussion of this thesis “Cell-specific pattern recognition receptor signaling in antibacterial defense”

14. Ding J, Song D, Ye X, Liu SF. A Pivotal Role of Endothelial-Specific NF-KappaB Signaling in the Pathogenesis of Septic Shock and Septic Vascular Dysfunction. J Immunol 2009;183:40314038.

15. Song D, Ye X, Xu H, Liu SF. Activation of Endothelial Intrinsic NF-{Kappa}B Pathway Impairs Protein C Anticoagulation Mechanism and Promotes Coagulation in Endotoxemic Mice. Blood 2009;114:2521-2529.

16. Andonegui G, Zhou H, Bullard D, Kelly MM, Mullaly SC, McDonald B, Long EM, Robbins SM, Kubes P. Mice That Exclusively Express TLR4 on Endothelial Cells Can Efficiently Clear a Lethal Systemic Gram-Negative Bacterial Infection. J Clin Invest 2009;119:1921-1930.

18. Ye X, Ding J, Zhou X, Chen G, Liu SF. Divergent Roles of Endothelial NF-KappaB in Multiple Organ Injury and Bacterial Clearance in Mouse Models of Sepsis. J Exp Med 2008;205:13031315.

19. Netea MG, Wijmenga C, O’Neill LA. Genetic Variation in Toll-Like Receptors and Disease Susceptibility. Nat Immunol 2012;13:535-542.

20. von Bernuth H., Picard C, Jin Z, Pankla R, Xiao H, Ku CL, Chrabieh M, Mustapha IB, Ghandil P, Camcioglu Y, et al. Pyogenic Bacterial Infections in Humans With MyD88 Deficiency. Science 2008;321:691-696.

21. Brouwer MC, de GJ, Heckenberg SG, Zwinderman AH, van der Poll T, van de Beek D. Host Genetic Susceptibility to Pneumococcal and Meningococcal Disease: a Systematic Review and Meta-Analysis. Lancet Infect Dis 2009;9:31-44.

22. Schroder NW, Schumann RR. Single Nucleotide Polymorphisms of Toll-Like Receptors and Susceptibility to Infectious Disease. Lancet Infect Dis 2005;5:156-164.

23. Opal SM, Laterre PF, Francois B, Larosa SP, Angus DC, Mira JP, Wittebole X, Dugernier T, Perrotin D, Tidswell M, et al. Effect of Eritoran, an Antagonist of MD2-TLR4, on Mortality in Patients With Severe Sepsis: the ACCESS Randomized Trial. JAMA 2013;309:1154-1162.

24. Cawcutt KA, Peters SG. Severe Sepsis and Septic Shock: Clinical Overview and Update on Management. Mayo Clin Proc 2014;89:1572-1578.

233

234

Chapter 12 Nederlandse samenvatting en discussie van dit proefschrift â&#x20AC;&#x153;Cell-specific pattern recognition receptor signaling in antibacterial defenseâ&#x20AC;? Miriam H.P. van Lieshout

Chapter 12

Algemene inleiding Dit proefschrift gaat over afweerreacties tijdens sepsis en de rol van bepaalde eiwitten hierin. Sepsis beschrijft de ontstekingsreactie die zich voltrekt in het gehele lichaam als reactie op een ernstige infectie gecombineerd met orgaanfalen dat daardoor kan ontstaan. Vaak is de ziekteverwekker (pathogeen) een bacterie. Wereldwijd is sepsis een belangrijke oorzaak van ziekte en sterfte (1, 2). Longontsteking (pneumonie) is de belangrijkste oorzaak van sepsis, gevolgd door infecties in de buikholte waaronder buikvliesontsteking, en infecties van de urinewegen (1). In dit proefschrift wordt gebruik gemaakt van muismodellen om deze ziektebeelden te onderzoeken, vaak in muizen die genetisch gemodificeerd zijn waardoor zij een bepaald gen missen (gen deficiëntie) en daardoor het onderzochte eiwit niet kunnen aanmaken. Ondanks de verbeterde behandelstrategieën en de ontwikkeling van de technische mogelijkheden in de Intensive Care geneeskunde van de afgelopen decennia blijft de sterfte onder sepsispatiënten hoog, ca 20-40%(3-5). Bovendien is er sprake van een sterke toename van resistentie bij veel voorkomende ziekteverwekkers, waardoor de mogelijkheden om infecties met antibiotica te behandelen in sommige gevallen sterk afnemen (5-10). Hierdoor is er een behoefte aan nieuwe en verbeterde behandelingen voor infecties en sepsis. Verdedigingsmechanismen van de gastheer Er zijn verschillende verdedigingsmechanismen die de gastheer moeten beschermen tegen infectie (host defense). De longblaasjes in de long vormen samen een grote oppervlakte die ter behoeve van de gaswisseling in contact staat met de buitenwereld, maar dit maakt de longen ook kwetsbaar voor infecties. Het eerste verdedigingsmechanisme zijn de slijmvliescellen van de luchtwegen (epitheel) die pathogenen afvoeren door het afscheiden van antibacteriële stoffen bevattend slijm wat vervolgens opwaarts wordt afgevoerd door de beweging van trilharen. De longblaasjes zijn bedekt met alveolair type I epitheelcellen (tbv de gaswisseling) en alveolair type II epitheelcellen, die bepaalde eiwitten (surfactant) produceren die de oppervlaktespanning beïnvloeden zodat de longblaasjes goed open blijven en die tevens ook een antibacteriële werking hebben (11). Als pathogenen aan deze verdedigingsmechanismen ontsnappen treden vervolgens het aangeboren (innate) afweersysteem en verworven (adaptive) afweersysteem in werking (12, 13). Er zijn naast de longepitheelcellen verschillende soorten cellen die altijd aanwezig zijn in de longen en de wacht houden (zie Figuur 1), zoals alveolaire macrofagen en dendritische cellen (12, 13). Als deze cellen pathogenen waarnemen leidt dat tot de migratie vanuit de bloedvaten naar de plaats van de infectie van grotere aantallen fagocyten (cellen die pathogenen kunnen opnemen) zoals neutrofielen. Dit proces wordt aangestuurd door de productie van stoffen die immuuncellen aantrekken (chemokines) en activeren tot migratie en het opnemen en doden van pathogenen (cytokines) (12-15). Alveolaire macrofagen kunnen zelf ook pathogenen en dode neutrofielen opnemen en op deze manier bijdragen aan het bestrijden van de ontsteking (16). Interferon (IFN)-γ is een belangrijke 236

Nederlandse samenvatting en discussie (Dutch summary)

cytokine die zowel door cellen van het aangeboren als verworven afweersysteem geproduceerd wordt en de activiteit van macrofagen krachtig stimuleert (17).

Figuur 1: Celtypes betrokken bij de aangeboren afweerrespons in de alveolus. In de buikholte (peritoneaalholte) bevinden zich normaal gesproken geen bacteriën, daarom zijn er minder lokale verdedigingsmechanismen dan in de long. Als er toch een ontsteking van het buikvlies optreedt dan wordt de immuunrespons geinduceerd door cellen van het reticuloendotheliale systeem (fagocyten uit o.a. de lever, milt en lymfeklieren), mesotheelcellen (de cellen die het buikvlies bekleden) en peritoneaal macrofagen ter bestrijding van de pathogenen. De rol van receptoren van het aangeboren immuunsysteem tijdens infectie Cellen van het aangeboren immuunsysteem herkennen specifieke moleculaire patronen van ziekteverwekkers (pathogen associated molecular patterns or PAMPs) met behulp van receptoren (pattern recognition receptors (PRRs)) (13, 18, 19). Toll-like receptoren (TLR’s) zijn een belangrijk onderdeel van deze groep van receptoren, die een grote variëteit aan PAMPs waarnemen, maar ook signalen die vrijkomen uit gastheercellen bij celschade (“danger signals”). Hierdoor spelen TLR’s een belangrijke rol bij het ontstaan en de versterking van de afweerrespons (14, 18, 19). Er zijn 10 verschillende menselijke TLR’s bekend. Van de meeste TLR’s is bekend welke moleculaire patronen van pathogenen of gastheer zij herkennen. 237

Chapter 12

Pathogenen bevatten doorgaans meerdere PAMPs die door verschillende TLR’s herkend kunnen worden (14). Geactiveerde TLR’s versterken het signaal vanaf celmembranen via intracellulaire adaptereiwitten naar de celkern, zodat er uiteindelijk (via veranderingen in genexpressie) eiwitten worden geproduceerd die de onstekingsrespons verder versterken. Het adaptereiwit voor alle TLR’s behalve TLR3 is MyD88 (myeloid differentiation primary response gene) (18, 19). Daarnaast speelt MyD88 een rol in de signalering via de receptoren van IL-1β en IL-18 , twee krachtige cytokines (20). TIR-domain-containing adapter-inducing interferon-β (TRIF) is het enige adaptereiwit voor TLR3 en draagt ook bij aan signalering via TLR4 (18, 19). Excessieve ontstekingsreacties kunnen leiden tot weefselschade, echter er bestaan ook factoren die de activiteit van TLR’s remmen waaronder Single immunoglobulin IL-1 receptor-related molecule (SIGIRR) (21). Een andere groep van receptoren binnen de PRRs zijn de Nod-like receptoren (nucleotide-binding and oligomerization-domain proteins) waarvan NLR family, pyrin domain containing 3 (NLRP3) het meest bekend is. NLRP3, kan samen met het adapter eiwit apoptosis-associated speck-like protein containing a CARD (ASC), een groot eiwitcomplex vormen ,een zogeheten “inflammasome”. Deze eiwitcomplexen zijn erg belangrijk voor de antimicrobiële respons aangezien zij via caspase-1 activatie leiden tot de maturatie van IL-1β and IL-18, twee belangrijke cytokines. Daarnaast speelt caspase-1 activatie een rol in pyroptosis, een vorm van inflammatoire celdood (22, 23). Hoewel een krachtige verdedigingsreactie van de gastheer van groot belang is om de infectie te controleren, kan deze anderzijds ook resulteren in lokale weefselschade en een systemische ontstekingsreactie die waarschijnlijk een rol speelt bij het ontstaan van orgaanfalen. Hierbij spelen ook virulentie (de kenmerken die de schadelijkheid bepalen) van de pathogeen en gastheereigenschappen een rol (1). Er is weinig bekend over de rol van andere cellen dan witte bloedcellen (non-hematopoietische cellen), zoals bijvoorbeeld longepitheelcellen en endotheelcellen (cellen die de bloedvaten bekleden), in de antibacteriële afweerreactie en het ontstaan van weefsel- en orgaanschade ten gevolge van de ontstekingsreactie. Er zijn aanwijzingen dat een ontstekingsreactie van het endotheel bijdraagt aan het ontstaan van weefsel- en orgaanschade tijdens sepsis (24-27). Pathogenen, receptors en experimentele modellen die worden gebruikt in dit proefschrift Er bestaan verschillende verwekkers en vormen van longontsteking. De meest voorkomende verwekker is Streptococcus (S.) pneumoniae (“pneumokok”), die het vaakst “in de omgeving opgelopen” longontsteking (community acquired pneumonie) veroorzaakt (28, 29). Deze vorm van longontsteking veroorzaakt met name ziekte en sterfte onder ouderen en jonge kinderen. Er bestaan meer dan 90 verschillende serotypen van deze bacterie waarvan sommigen virulenter zijn dan anderen (30). In dit proefschrift worden twee verschillende S. pneumoniae stammen in muismodellen gebruikt. De eerste is een serotype 3 bacterie (ATCC 6303) die 238

Nederlandse samenvatting en discussie (Dutch summary)

vaak longontsteking bij mensen veroorzaakt en is geasscocieerd met een ernstiger ziekteverloop en meer complicaties. Deze bacterie veroorzaakt in een muismodel circa 80% sterfte als deze in een relatief lage concentratie van 5 x 104 CFU (colony forming units) in de neus wordt gedruppeld. Daarnaast wordt in dit proefschrift een serotype 2 bacterie (D39) gebruikt, deze bacterie is minder aggressief en wordt vaak gebruikt in experimentele muismodellen waarbij na toediening van een hoge dosis van ca 1x107 CFU circa 20% van de muizen sterft. Deze bacterie veroorzaakt nauwelijks gevallen van longontsteking in de westerse wereld (31-39). Verschillende TLR’s dragen bij aan de afweerreactie tegen de pneumokok. TLR2 detecteert een bestand van de celwand van de bacterie (lipoteichoic acid, LTA) (36, 40). TLR4 herkent pneumolysine, een virulentiefactor van de pneumokok en speelt een beschermende rol tijdens longontsteking met deze bacterie (41, 42). Tenslotte hadden TLR9 deficiënte muizen een grotere hoeveelheid bacteriën in de long en in andere organen tijdens pneumokokkenpneumonie (43). In MyD88 deficiënte muizen was sprake van een sterke toename van de groei van een serotype 4 S.pneumoniae en toegenomen sterfte (44). De afgelopen jaren is de beschermende rol van de NLRP3 inflammasome tijdens pneumokokkenpneumonie in verschillende studies aangetoond (32, 33, 45). Waarschijnlijk werkt de herkenning van pneumolysine door NLRP3 beschermend (33, 46, 47). Klebsiella pneumoniae is een gramnegatieve bacterie die ernstige longontsteking en ook infecties van de bloedstroom kan veroorzaken en vaker infecties veroorzaakt bij patiënten die zijn opgenomen in het ziekenhuis (hospital acquired infectie) of in instellingen voor de gezondheidszorg worden behandeld (health care associated) (28, 48, 49). In experimentele Klebsiella pneumonie spelen met name TLR4, dat lipopolysacharide (LPS) uit de bacteriële celwand van gramnegatieve bacteriën herkent en tevens verschillende “danger signals” afkomstig uit gastheercellen en TLR9 dat bacterieel DNA detecteert, een beschermende rol (42, 50, 51). Ook de universele TLR-adapter MyD88 is van cruciaal belang voor de afweerrespons en overleving tijdens experimentele luchtweginfectie met Klebsiella (52). Het adapter eiwit TRIF, dat signalering via TLR3 voortgeleid en tevens bijdraagt aan TLR4 afhankelijke signalering, is nodig voor een optimale gastheerrespons tijdens Klebsiella pneumonie (52). Pseudomonas aeruginosa is ook een gramnegatieve bacterie die bovendien is uitgerust met een zweepstaart (flagel) waarmee hij zich kan voortbewegen. Deze bacterie veroorzaakt vaak infecties waaronder longontsteking, bij patiënten die in het ziekenhuis zijn opgenomen en/of beademd worden en bij patiënten met een chronische longziekte (53-56). In experimentele modellen van Pseudomonas infectie werd het belang van TLR’s voor de bacteriele klaring aangetoond in MyD88 deficiënte muizen, aangezien deze zeer kwetsbaar bleken voor luchtweginfecties met Pseudomonas en dit resulteerde in een hoge mate van sterfte (57-59). Bij de antibacteriële respons spelen TLR2, TLR4 en TLR5 (detecteert flagelline, afkomstig uit de zweepstaart) allen een rol (60, 61). Infecties afkomstig uit de buik zijn samen met infecties van de urinewegen de 239

Chapter 12

tweede oorzaak van sepsis. Escherichia (E.) coli, een darmbacterie, is een van de gramnegatieve bacteriën die het vaakst worden gekweekt bij patiënten met sepsis en buikvliesontsteking (62-64). Meestal ontstaat buikvliesontsteking (peritonitis) door de perforatie van een hol orgaan in de buik (bv maag, darm of galblaas) waardoor darmbacteriën in de normaal gesproken steriele buikholte komen. Er is dan sprake van een infectie met verschillende soorten bacteriën (6467). Soms ontstaat peritonitis ook doordat bacteriën zich door de darmwand heen verplaatsten en in de buikholte terecht komen in daarvoor bevattelijke patiënten met bijvoorbeeld nier- of leverfalen. Aangezien bacteriën zich vanuit de buikholte snel naar de bloedbaan kunnen verspreiden treedt er een snelle en hevige systemische ontstekingsreactie op en ontwikkeld sepsis zich snel, met een hoge sterfte tot 60% tot gevolg (65, 66). Wat de behandeling van peritonitis verder bemoeilijkt is dat er in toenemende mate sprake is van resistentie voor de meest gangbare antibiotica bij E. coli stammen die zowel in het ziekenhuis als in de samenleving vorkomen (9). Het experimentele model voor E. coli peritonitis dat in dit proefschrift wordt gebruikt bestaat uit toediening van een lage hoeveelheid virulente bacteriën van de O18:K1 stam in de buikholte, in tegenstelling tot veel gebruikte modellen waarbij infectie wordt geinduceerd door toediening van een grote hoeveelheid van minder virulente bacteriën. In het hier gebruikte model is TLR4 belangrijk voor de initiatie van de afweerrespons en tijdens de latere fase van de infectie levert ook TLR2 een bijdrage aan de afweer (68).

Doel van het onderzoek dat wordt beschreven in dit proefschrift Het hoofddoel van dit proefschrift is om meer inzicht te krijgen in de rol van TLR- en met name van MyD88- afhankelijke signalering in experimentele sepsis modellen, met de nadruk op pneumoniemodellen. Daarnaast onderzochten we de rol van de “inflammasome” tijdens pneumokokkenpneumonie. Een ander onderzoeksdoel was om meer inzicht te krijgen in de bijdrage van verschillende celtypen en lichaamscompartimenten aan TLR- en MyD88- afhankelijke signalering tijdens infectie en sepsis, aangezien deze receptoren en adaptereiwitten voorkomen op veel verschillende celtypen maar hier weinig over bekend is.

240

Nederlandse samenvatting en discussie (Dutch summary)

Samenvatting van dit proefschrift Hoofdstuk 1 is een algemene inleiding die de mechanismen van de afweerrespons, de rol van pattern recognition receptors hierin en de ziektemodellen beschrijft, zoals ook hier bovenstaand. In het eerste deel van dit proefschrift onderzochten we met name de rol van Tolllike receptors (TLR’s) en hun intracellulaire adaptereiwitten myeloid differentiation primary response gene (MyD88) en TIR-domain-containing adapter-inducing interferon-β (TRIF) tijdens pneumonie met de gramnegatieve pathogenen Klebsiella (K.) pneumoniae en Pseudomonas (P.) aeruginosa. Bovendien onderzochten we specifiek de rol van deze eiwitten in verschillende lichaamscompartiementen en celtypes. Hoofdstuk 2 beschrijft de rol van TLR2 en TLR4 tijdens Klebsiella pneumonie. In dit hoofdstuk concluderen we dat TLR4 belangrijker is voor de initiële antibacteriële respons dan TLR2, aangezien Tlr4-/- (TLR4 deficiënte) muizen een toegenomen groei van bacteriën in de organen hadden maar Tlr2-/- muizen niet. Echter tijdens de latere stadia van de infectie of na toediening van een hoger aantal bacteriën bleek TLR2 wel een rol te spelen in de afweer. Tlr2/Tlr4 dubbel deficiënte muizen waren nog veel kwetsbaarder dan Tlr4-/- muizen. Bovendien konden we door gebruik te maken van chimere muizen (zie Figuur 2) aantonen dat TLR2 en TLR4 op cellen afkomstig uit het beenmerg (hematopoietische cellen

Figuur 2: Chimere muismodel 241

Chapter 12

zoals witte bloedcellen) de kwaliteit van de antibacteriële reactie bepaalden, terwijl TLR’s op lichaamscellen hier niet in belangrijke mate aan bijdroegen. In hoofdstuk 3 rapporteren we over de rol van MyD88 and TRIF tijdens K. pneumoniae infectie en de rol van deze eiwitten in hematopoietische- en lichaamscellen. Beide TLRadapters bleken cruciaal voor de antibacteriële afweerrespons en overleving van de muizen, hoewel Myd88-/- muizen, die allen stierven binnen 48 uur na infectie, nog kwetsbaarder bleken dan TRIF mutant muizen. MyD88 in zowel hematopoietische als lichaamscellen cellen was belangrijk voor de antibacteriële respons terwijl TRIF in alleen hematopoietische cellen voldoende bescherming boodt. Anderzijds, droeg MyD88 in lichaamscellen en TRIF in hematopoietische cellen bij aan orgaanschade. Kort na infectie bleek MyD88 in lichaamscellen cruciaal voor de productie van chemokines en cytokines en de migratie van neutrofielen, terwijl TRIF zowel in hematopoietische- als lichaamscellen even belangrijk was. In conclusie leveren MyD88- en TRIF afhankelijke signalering een verschillende bijdrage aan de afweerrespons van de gastheer in verschillende celtypen en verandert hun rol tijdens het beloop van Klebsiella infectie. In hoofdstuk 4 verdiepten we ons verder in de rol van cell-specifieke MyD88 afhankelijk signalering tijdens K. pneumoniae infectie. Muizen die deficiënt waren voor MyD88 in bepaalde witte bloedcellen (myeloide cellen, LysM-Myd88-/- muizen) en in myeloide cellen plus endotheelcellen (Tie2-Myd88-/- muizen) vertoonden een toegenomen sterfte en hogere aantallen bacteriën in de organen. Tie2-Myd88-/muizen die beenmerg van controlemuizen hadden toegediend gekregen om de rol van selectieve afwezigheid van MyD88 in endotheelcellen te onderzoeken, vertoonden geen afwijkende afweerrespons. De afwezigheid van MyD88 in myeloide of endotheelcellen had geen invloed op long- of orgaanschade later tijdens de infectie. Wel hadden LysM-Myd88-/- muizen een sterk verminderde ontstekingsrespons kort na de infectie. Concluderend is MyD88 in myeloide maar niet in endotheliale cellen belangrijk voor de afweerrespons tijdens sepsis veroorzaakt door Klebsiella pneumonie. In hoofdstuk 5 onderzochten we de rol van TRIF tijdens Klebsiella pneumonie en met name zijn rol in de productie van interferon (IFN)-γ, een belangrijke cytokine die onder andere de activiteit van macrofagen versterkt. TRIF mutant muizen hadden een verminderde antibacteriële respons die samenging met de afwezigheid van IFN-γ productie in de longen. De productie van IFN-γ door miltcellen (splenocyten) na stimulatie met K. pneumoniae was afhankelijk van TLR4, MyD88 and TRIF. Anderzijds verbeterde de afweer van TRIF mutant muizen na toediening van IFN-γ: de aantallen bacteriën in de longen en andere organen waren gelijk aan die van onbehandelde wild-type muizen terwijl ook de productie van cytokines in de long deels herstelden. Wel trad er na behandeling met IFN-γ meer leverschade op in TRIF mutant muizen. Deze bevindingen wijzen erop dat de bijdrage van TRIF aan de antibacteriële respons in ieder geval deels verklaard kan worden door de productie van IFN-γ in de longen. In hoofdstuk 6 onderzochten we de rol van MyD88 afhankelijke signalering tijdens luchtweginfectie met P. aeruginosa. Sftpccre-Myd88-/- muizen, die geen MyD88 in long epitheel cellen hebben, vertoonden een beperking in de bacteriële klaring kort na de infectie en hadden een afgenomen aantal neutrofielen dat zich in de longen verzamelde. Tlr5-/- muizen hadden ook een verstoorde ontstekingsrespons 242

Nederlandse samenvatting en discussie (Dutch summary)

en bacteriële klaring kort na de infectie die bepaald werd door de afwezigheid van TLR5 in lichaamscellen zoals bleek in experimenten met chimere muizen. Onze bevindingen wezen op een interactie tussen TLR5 op epitheelcellen en flagelline (TLR5 ligand) van de Pseudomonas bacterie. Door gebruik te maken van een Pseudomonas bacterie (mutant) zonder flagel konden we aantonen dat MyD88 afhankelijke TLR signalering in het epitheel belangrijk is voor de ontstekingsrespons. In het volgende deel van dit proefschrift richtten we ons op de rol van een ander onderdeel van het aangeboren afweersysteem, de “NLRP3-inflammasome” (dat bestaat uit een eiwitcomplex gevormd door het sensoreiwit NLRP3 en het adaptereiwit ASC) tijdens pneumokokkenpneumonie. Hoofdstuk 7 beschrijft de de rol van NLRP3 en ASC in de afweerrespons tijdens pneumokokkenpneumonie met een serotype 2 D39 bacterie stam. Beide Nlrp3-/- en Asc-/- muizen vertoonden een afgenomen bacteriële klaring in de long, maar alleen Asc-/- muizen hadden een toegenomen sterfte vergeleken met wild type muizen. In zowel Nlrp3-/- als Asc-/- muizen was de vroege inflammatoire respons verstoord zoals bleek uit een afgenomen cytokineproductie. Door middel van een gedetailleerde analyse van de vroege inflammatoire respons in de longen met behulp van een analyse van genexpressie van alle genen (whole-genome transcriptional profiling) identificeerden we een aantal mediatoren die verschillend waren tussen Nlrp3-/and Asc-/- muizen. Mogelijk verklaarden deze de verschillen in sterfte tussen deze soorten muizen. De bevindingen in dit hoofdstuk bevestigen een belangrijke rol voor het NLRP3-inflammasome tijdens pneumokokkenpneumonie en suggereren daarnaast een belangrijke rol voor ASC naast zijn rol als adaptereiwit voor NLRP3. Anderzijds ontdekten we in hoofdstuk 8 een tegengestelde rol voor NLRP3 and ASC tijdens infectie met een serotype 3 pneumokokken stam. Opvallend genoeg vertoonden beide Nlrp3-/- en Asc-/- muizen een sterk verbeterde afweerreactie, zoals bleek uit een toegenomen overleving en verminderde groei van bacteriën in de long en afgenomen verspreiding van bacteriën naar andere organen. Bovendien onderzochten we de rol van MyD88 afhankelijke signalering tijdens infectie met deze specifieke virulente pneumokokkenstam. De groei van bacteriën en overleving waren onveranderd in Myd88-/- muizen ten opzichte van wild type muizen, hoewel er sprake was van een afgenomen inflammatieresponse in de longen na infectie met hoge aantallen bacteriën. Dit hoofdstuk illustreert dat receptoren van het aangeboren afweersysteem waaronder het NLRP3-inflammasome en TLR’s, afhankelijk van de pathogeen onder sommige omstandigheden niet bijdragen aan bestrijding van de infectie en dat hun rol heel anders kan zijn tijdens infecties met verschillende bacteriestammen, ook al zijn zij nauw verwant. In hoofdstuk 9 wordt de rol van Single immunoglobulin IL-1 receptor-related molecule (SIGIRR), een remmende factor van TLR- and IL-1 receptor afhankelijke inflammatie tijdens pneumokokkenpneumonie en -sepsis beschreven. Sigirr-/muizen vertoonden een afgenomen sterfte en lagere bacterieaantallen zowel tijdens luchtweginfectie als na toediening van bacteriën in de bloedbaan, terwijl de inflammatoire reactie niet anders was. Wel konden alveolaire macrofagen en neutrofielen van Sigirr-/- muizen sneller S. Pneumoniae opnemen in het laboratorium, als mogelijke verklaring voor de verbeterde antibacteriële respons. Tenslotte onderzochten we het effect van behandeling met anti-TLR4 therapie, 243

Chapter 12

die werd ontworpen om overmatige inflammatie in sepsis patiënten af te remmen in een muismodel van E. coli peritonitis. Hoofdstuk 10 laat zien dat muizen die een lage hoeveelheid bacteriën kregen toegediend na behandeling met anti-TLR4 therapie hogere aantallen bacteriën in de organen hadden vergeleken met muizen die het controlemiddel kregen, ondanks gelijktijdige behandeling met antibiotica. Wel was de nierfunctie in lichte mate beter dan in de controlegroep terwijl de sterfte onveranderd was. Dit illustreert dat het afremmen van TLR4 activiteit er onder bepaalde omstandigheden toe kan leiden dat het bestrijden van bacteriën minder effectief verloopt, zelfs als er tegelijk behandeld word met antibiotica. Anderzijds kan het remmen van TLR4 ook orgaanschade in enige mate tegengaan, afhankelijk van de infectieuze dosis en het verloop van de infectie.

Conclusies De bevindingen in de hierboven besproken hoofdstukken illustereren dat de gastheerrespons tijdens pneumonie uit dynamische processen bestaat, waarbij verschillende TLR’s en hun adaptereiwitten in verschillende celtypen (zoals hematopoietisch- versus lichaamscellen en witte bloedcellen, longepitheel- en endotheelcellen) betrokken zijn en hun bijdrage kan veranderen tijdens het verloop van de infectie en ook afhankelijk is van de pathogeen. Zo spelen TLR2 en TLR4 en hun intracellulaire adaptereiwitten MyD88 en TRIF een cruciale rol tijdens Klebsiella pneumonie en sepsis, waarbij de bijdrage van MyD88 in longepitheelcellen en in myeloide cellen anders is tijdens Klebsiella pneumoniae en Pseudomonas aeruginosa infectie. Ons onderzoek laat bovendien zien dat de balans tussen gunstige en schadelijke effecten ten gevolge van de rol van de NLRP3 inflammasome tijdens pneumokokkenpneumonie sterk kan verschillen tijdens infectie met verschillende pathogenen, zelfs als deze nauw verwant zijn. Ook kan de bijdrage van TLR- en MyD88 afhankelijke signalering heel verschillend zijn in deze infectiemodellen. De negatieve rol van de TLR- en IL-1 like receptor remmer SIGIRR tijdens pneumokokkenpneumonie valt niet direct valt af te leidien uit de bevindingen in de voorgaande hoofdstukken en zijn waarschijnlijk de resultante van nog onbekende interacties tussen deze recptoren. Of de gastheer een infectie overleeft hangt niet alleen af van de effecitiviteit van de antimicrobiële respons en of de pathogeen geklaard kan worden, maar ook van de ernst van weefsel en orgaanfalen die als bij-effect van een krachtige ontstekingsreactie kan optreden. Daarom wordt ook wel gesteld dat de immuunrespons tijdens sepsis een mes is dat aan twee kanten snijdt. Dit blijkt ook uit sommige van de bevindingen die we in dit proefschrift deden waarbij het gevecht tegen bacteriën bijdroeg aan orgaanschade. Zo bleek de orgaanschade tijdens Klebsiella infectie en sepsis minder ernstig in muizen die een verminderde antibacteriële afweer hadden doordat zij (deels) deficiënt waren voor MyD88 of TRIF. Ook ging het verbeteren van de antibacteriële respons door het toedienen van IFN-γ aan TRIF deficiënte muizen ten koste van meer orgaanschade. Behandeling met een TLR4-remmer tijdens peritonitis resulteerde in een lichte afname van orgaanfalen maar ook in een vermindere capaciteit om bacteriën te klaren 244

Nederlandse samenvatting en discussie (Dutch summary)

zoals bleek uit hogere aantallen bacteriën in de organen, ondanks gelijktijdige behandeling met antibiotica. Uiteraard is het uiteindelijk de vraag of en hoe deze bevindingen over experimentele sepsis in muismodellen te vertalen zijn naar het ziektebeeld sepsis bij mensen en de ontwikkeling van betere therapieën. Het is algemeen bekend dat het gebruik van genetisch gemodificeerde muizen zijn beperkingen kent. Ook zijn er de afgelopen jaren, naar aanleiding van de kennis die werd opgedaan in onder andere experimentele studies met muismodellen, verschillende therapieën ontworpen die gericht zijn op de functie van TLR’s. Zo werden er twee verschillende TLR4 remmers onderzocht bij sepsispatiënten in klinische fase III studies, maar beiden hadden geen effect op de overleving na 28 dagen en verdere studies werden gestaakt (69, 70). Hetzelfde gebeurde met andere veelbelovende aanvullende therapieën voor sepsis: ondanks gunstige resultaten in pre-klinische studies kon de aanvullende therapeutische waarde in sepsispatiënten niet worden vastgesteld. Een belangrijk obstakel bij het bestuderen van het effect van nieuwe aanvullende therapiën in sepsispatiënten is dat sepsis een syndroom beschrijft en een vergaarbak is van een groot aantal verschillende (infectie)ziekten en ziekteprocessen. Zoals werd beschreven in dit proefschrift, verschilt de rol van de verschillende receptoren die betrokken zijn bij de aangeboren afweerrespons tussen verschillende sepsismodellen en pathogenen (zelfs nauw verwante bacteriestammen) en verandert hun rol tijdens het verloop van de infectie. Er is weinig bekend over de exacte activiteit van TLR’s in sepsispatiënten, laat staan over veranderingen hierin tijdens het verloop van het ziektebeeld of bij infectie met verschillende verwekkers. Het is belangrijk om zich te realiseren dat in de klinische praktijk bij aanvang van de behandeling de ziekteverwekker meestal onbekend is, aangezien bacteriekweken vaak pas na 24 uur bekend worden en bovendien niet altijd resultaat opleveren. Ook is er weinig bekend over het verloop van de inflammatoire respons en het verloop van de infectie bij sepsispatiënten, dat waarschijnlijk ook wordt beïnvloed door allerlei patiëntfactoren (genen, bijkomende ziekten, leeftijd etcetera), terwijl in experimentele modellen het moment van infectie exact bekend is en de ziekte bij alle dieren ongeveer hetzelfde verloopt. Een ander verschil tussen experimentele sepsis en de humane situatie is dat de kwaliteit van ondersteunende therapieën (inclusief orgaanvervangende therapieën) de afgelopen decennia sterk is verbeterd, zodat sterfte in de eerste dagen vaak wordt voorkomen. Patiënten die sepsis aanvankelijk overleven houden echter een sterk verhoogde kans op sterfte, ook nog na 28 dagen, ten gevolge van orgaanfalen en andere complicaties (70, 71). Om al deze redenen is in klinische studies het beïnvloeden van de aangeboren immuunrespons door het moduleren van één receptor waarschijnlijk niet effectief gebleken om in studieverband de sterfte in de hele groep van sepsis patiënten aantoonbaar te veranderen. Een eerste stap in het verbeteren van de behandeling van sepsis zou het versnellen van het identificeren van de pathogeen zijn, bij voorkeur inclusief informatie over het resistentiepatroon zodat de antibiotische therapie beter kan worden afgestemd. Bovendien zou meer inzicht in het verloop van de afweerrespons in de individuele patiënt kunnen worden verkregen door het gebruik van biomarkers. Toekomstig onderzoek 245

Chapter 12

zou zich vervolgens moeten richten op het integreren van deze gedetailleerde observaties in patiënten met sepsis in verschillende fases van de ziekte en data van experimentele sepsismodellen. Deze kennis zou dan weer kunnen worden gebruikt bij het onderzoeken van nieuwe therapieën en het ontwerpen van snelle testen die “aan het bed” kunnen worden uitgevoerd om die patiënten te identificeren die baat zullen hebben van deze nieuwe therapieën en om het effect hiervan te monitoren.

246

Nederlandse samenvatting en discussie (Dutch summary)

Reference List

1. Angus DC, van der Poll T. Severe Sepsis and Septic Shock. N Engl J Med 2013;369:840-851.

2. van der Poll T, Opal SM. Host-Pathogen Interactions in Sepsis. Lancet Infect Dis 2008;8:32-43.

3. van der Poll T. Immunotherapy of Sepsis. Lancet Infect Dis 2001;1:165174.

4. Webster NR, Galley HF. Immunomodulation in the Critically Ill. Br J Anaesth 2009;103:70-81.

5. World Health Organisation. Antimicrobial Resistance: Global Report on Surveillance.; 2014.

6. Coque TM, Baquero F, Canton R. Increasing Prevalence of ESBL-Producing Enterobacteriaceae in Europe. Euro Surveill 2008;13.

7. Giamarellou H. Multidrug Resistance in Gram-Negative Bacteria That Produce Extended-Spectrum Beta-Lactamases (ESBLs). Clin Microbiol Infect 2005;11 Suppl 4:1-16.

8. Schwaber MJ, Carmeli Y. Mortality and Delay in Effective Therapy Associated With Extended-Spectrum Beta-Lactamase Production in Enterobacteriaceae Bacteraemia: a Systematic Review and Meta-Analysis. J Antimicrob Chemother 2007;60:913-920.

9. Rossolini GM, Arena F, Pecile P, Pollini S. Update on the Antibiotic Resistance Crisis. Curr Opin Pharmacol 2014;18C:56-60.

11. Camelo A, Dunmore R, Sleeman MA, Clarke DL. The Epithelium in Idiopathic Pulmonary Fibrosis: Breaking the Barrier. Front Pharmacol 2014;4:173.

12. Dickson RP, Erb-Downward JR, Huffnagle GB. The Role of the Bacterial Microbiome in Lung Disease. Expert Rev Respir Med 2013;7:245-257.

13. Opitz B, van Laak V, Eitel J, Suttorp N. Innate Immune Recognition in Infectious and Noninfectious Diseases of the Lung. Am J Respir Crit Care Med 2010;181:1294-1309. 247

Chapter 12

14. Mizgerd JP. Acute Lower Respiratory Tract Infection. N Engl J Med 2008;358:716-727.

15. Parker D, Prince A. Innate Immunity in the Respiratory Epithelium. Am J Respir Cell Mol Biol 2011;45:189-201.

16. Aggarwal NR, King LS, Dâ&#x20AC;&#x2122;Alessio FR. Diverse Macrophage Populations Mediate Acute Lung Inflammation and Resolution. Am J Physiol Lung Cell Mol Physiol 2014;306:L709-L725.

17. Schroder K, Hertzog PJ, Ravasi T, Hume DA. Interferon-Gamma: an Overview of Signals, Mechanisms and Functions. J Leukoc Biol 2004;75:163189.

18. Beutler BA. TLRs and Innate Immunity. Blood 2009;113:1399-1407.

19. Kawai T, Akira S. Toll-Like Receptors and Their Crosstalk With Other Innate Receptors in Infection and Immunity. Immunity 2011;34:637-650.

21. Riva F, Bonavita E, Barbati E, Muzio M, Mantovani A, Garlanda C. TIR8/ SIGIRR Is an Interleukin-1 Receptor/Toll Like Receptor Family Member With Regulatory Functions in Inflammation and Immunity. Front Immunol 2012;3:322.

22. Strowig T, Henao-Mejia J, Elinav E, Flavell R. Inflammasomes in Health and Disease. Nature 2012;481:278-286.

23. Franchi L, Munoz-Planillo R, Nunez G. Sensing and Reacting to Microbes Through the Inflammasomes. Nat Immunol 2012;13:325-332.

24. Ding J, Song D, Ye X, Liu SF. A Pivotal Role of Endothelial-Specific NF-KappaB Signaling in the Pathogenesis of Septic Shock and Septic Vascular Dysfunction. J Immunol 2009;183:4031-4038.

25. Song D, Ye X, Xu H, Liu SF. Activation of Endothelial Intrinsic NF-{Kappa} B Pathway Impairs Protein C Anticoagulation Mechanism and Promotes Coagulation in Endotoxemic Mice. Blood 2009;114:2521-2529.

248

Nederlandse samenvatting en discussie (Dutch summary)

27. Ye X, Ding J, Zhou X, Chen G, Liu SF. Divergent Roles of Endothelial NF-KappaB in Multiple Organ Injury and Bacterial Clearance in Mouse Models of Sepsis. J Exp Med 2008;205:1303-1315.

28. Welte T, Torres A, Nathwani D. Clinical and Economic Burden of Community-Acquired Pneumonia Among Adults in Europe. Thorax 2012;67:71-79.

29. World Health Organisation. World Health Observatory, Top Ten Causes of Death 2012.; 2012.

30. Weinberger DM, Harboe ZB, Sanders EA, Ndiritu M, Klugman KP, Ruckinger S, Dagan R, Adegbola R, Cutts F, Johnson HL, et al. Association of Serotype With Risk of Death Due to Pneumococcal Pneumonia: a Meta-Analysis. Clin Infect Dis 2010;51:692-699.

32. Fang R, Tsuchiya K, Kawamura I, Shen Y, Hara H, Sakai S, Yamamoto T, Fernandes-Alnemri T, Yang R, Hernandez-Cuellar E, et al. Critical Roles of ASC Inflammasomes in Caspase-1 Activation and Host Innate Resistance to Streptococcus Pneumoniae Infection. J Immunol 2011;187:4890-4899.

33. McNeela EA, Burke A, Neill DR, Baxter C, Fernandes VE, Ferreira D, Smeaton S, El-Rachkidy R, McLoughlin RM, Mori A, et al. Pneumolysin Activates the NLRP3 Inflammasome and Promotes Proinflammatory Cytokines Independently of TLR4. PLoS Pathog 2010;6:e1001191.

34. Dessing MC, Knapp S, Florquin S, de Vos AF, van der Poll T. CD14 Facilitates Invasive Respiratory Tract Infection by Streptococcus Pneumoniae. Am J Respir Crit Care Med 2007;175:604-611.

35. Dessing MC, Florquin S, Paton JC, van der Poll T. Toll-Like Receptor 2 Contributes to Antibacterial Defence Against Pneumolysin-Deficient Pneumococci. Cell Microbiol 2008;10:237-246.

36. Dessing MC, Schouten M, Draing C, Levi M, von AS, van der Poll T. Role Played by Toll-Like Receptors 2 and 4 in Lipoteichoic Acid-Induced Lung Inflammation and Coagulation. J Infect Dis 2008;197:245-252.

37. Knapp S, Wieland CW, van â&#x20AC;&#x2DC;t Veer C, Takeuchi O, Akira S, Florquin S, van der Poll T. Toll-Like Receptor 2 Plays a Role in the Early Inflammatory Response to Murine Pneumococcal Pneumonia but Does Not Contribute to Antibacterial Defense. J Immunol 2004;172:3132-3138. 249

Chapter 12

38. van der Windt GJ, Hoogendijk AJ, Schouten M, Hommes TJ, de Vos AF, Florquin S, van der Poll T. Osteopontin Impairs Host Defense During Pneumococcal Pneumonia. J Infect Dis 2011;203:1850-1858.

39. Rijneveld AW, Florquin S, Branger J, Speelman P, Van Deventer SJ, van der PT. TNF-Alpha Compensates for the Impaired Host Defense of IL-1 Type I Receptor-Deficient Mice During Pneumococcal Pneumonia. J Immunol 2001;167:5240-5246.

40. Schroder NW, Morath S, Alexander C, Hamann L, Hartung T, Zahringer U, Gobel UB, Weber JR, Schumann RR. Lipoteichoic Acid (LTA) of Streptococcus Pneumoniae and Staphylococcus Aureus Activates Immune Cells Via Toll-Like Receptor (TLR)-2, Lipopolysaccharide-Binding Protein (LBP), and CD14, Whereas TLR-4 and MD-2 Are Not Involved. J Biol Chem 2003;278:15587-15594.

41. Malley R, Henneke P, Morse SC, Cieslewicz MJ, Lipsitch M, Thompson CM, Kurt-Jones E, Paton JC, Wessels MR, Golenbock DT. Recognition of Pneumolysin by Toll-Like Receptor 4 Confers Resistance to Pneumococcal Infection. Proc Natl Acad Sci U S A 2003;100:1966-1971.

42. Branger J, Knapp S, Weijer S, Leemans JC, Pater JM, Speelman P, Florquin S, van der Poll T. Role of Toll-Like Receptor 4 in Gram-Positive and Gram-Negative Pneumonia in Mice. Infect Immun 2004;72:788-794.

43. Albiger B, Dahlberg S, Sandgren A, Wartha F, Beiter K, Katsuragi H, Akira S, Normark S, Henriques-Normark B. Toll-Like Receptor 9 Acts at an Early Stage in Host Defence Against Pneumococcal Infection. Cell Microbiol 2007;9:633-644.

44. Albiger B, Sandgren A, Katsuragi H, Meyer-Hoffert U, Beiter K, Wartha F, Hornef M, Normark S, Normark BH. Myeloid Differentiation Factor 88-Dependent Signalling Controls Bacterial Growth During Colonization and Systemic Pneumococcal Disease in Mice. Cell Microbiol 2005;7:1603-1615.

45. Witzenrath M, Pache F, Lorenz D, Koppe U, Gutbier B, Tabeling C, Reppe K, Meixenberger K, Dorhoi A, Ma J, et al. The NLRP3 Inflammasome Is Differentially Activated by Pneumolysin Variants and Contributes to Host Defense in Pneumococcal Pneumonia. J Immunol 2011;187:434-440.

46. Kadioglu A, Weiser JN, Paton JC, Andrew PW. The Role of Streptococcus Pneumoniae Virulence Factors in Host Respiratory Colonization and Disease. Nat Rev Microbiol 2008;6:288-301.

250

Nederlandse samenvatting en discussie (Dutch summary)

47. Hoegen T, Tremel N, Klein M, Angele B, Wagner H, Kirschning C, Pfister HW, Fontana A, Hammerschmidt S, Koedel U. The NLRP3 Inflammasome Contributes to Brain Injury in Pneumococcal Meningitis and Is Activated Through ATP-Dependent Lysosomal Cathepsin B Release. J Immunol 2011;187:5440-5451.

48. Kollef MH, Shorr A, Tabak YP, Gupta V, Liu LZ, Johannes RS. Epidemiology and Outcomes of Health-Care-Associated Pneumonia: Results From a Large US Database of Culture-Positive Pneumonia. Chest 2005;128:38543862.

49. Zahar JR, Timsit JF, Garrouste-Org, Francais A, Vesin A, Descorps-Declere A, Dubois Y, Souweine B, Haouache H, Goldgran-Toledano D, et al. Outcomes in Severe Sepsis and Patients With Septic Shock: Pathogen Species and Infection Sites Are Not Associated With Mortality. Crit Care Med 2011;39:1886-1895.

50. Bhan U, Lukacs NW, Osterholzer JJ, Newstead MW, Zeng X, Moore TA, McMillan TR, Krieg AM, Akira S, Standiford TJ. TLR9 Is Required for Protective Innate Immunity in Gram-Negative Bacterial Pneumonia: Role of Dendritic Cells. J Immunol 2007;179:3937-3946.

51. Schurr JR, Young E, Byrne P, Steele C, Shellito JE, Kolls JK. Central Role of Toll-Like Receptor 4 Signaling and Host Defense in Experimental Pneumonia Caused by Gram-Negative Bacteria. Infect Immun 2005;73:532545.

52. Cai S, Batra S, Shen L, Wakamatsu N, Jeyaseelan S. Both TRIF- and MyD88-Dependent Signaling Contribute to Host Defense Against Pulmonary Klebsiella Infection. J Immunol 2009;183:6629-6638.

53. Chastre J, Fagon JY. Ventilator-Associated Pneumonia. Am J Respir Crit Care Med 2002;165:867-903.

54. Jones RN. Microbial Etiologies of Hospital-Acquired Bacterial Pneumonia and Ventilator-Associated Bacterial Pneumonia. Clin Infect Dis 2010;51 Suppl 1:S81-S87.

55. King PT, Holdsworth SR, Freezer NJ, Villanueva E, Holmes PW. Microbiologic Follow-Up Study in Adult Bronchiectasis. Respir Med 2007;101:16331638.

56. Lambiase A, Raia V, Del PM, Sepe A, Carnovale V, Rossano F. Microbiology of Airway Disease in a Cohort of Patients With Cystic Fibrosis. BMC Infect Dis 2006;6:4.

251

Chapter 12

57. Power MR, Peng Y, Maydanski E, Marshall JS, Lin TJ. The Development of Early Host Response to Pseudomonas Aeruginosa Lung Infection Is Critically Dependent on Myeloid Differentiation Factor 88 in Mice. J Biol Chem 2004;279:49315-49322.

58. Skerrett SJ, Liggitt HD, Hajjar AM, Wilson CB. Cutting Edge: Myeloid Differentiation Factor 88 Is Essential for Pulmonary Host Defense Against Pseudomonas Aeruginosa but Not Staphylococcus Aureus. J Immunol 2004;172:3377-3381.

59. Skerrett SJ, Wilson CB, Liggitt HD, Hajjar AM. Redundant Toll-Like Receptor Signaling in the Pulmonary Host Response to Pseudomonas Aeruginosa. Am J Physiol Lung Cell Mol Physiol 2007;292:L312-L322.

60. Morris AE, Liggitt HD, Hawn TR, Skerrett SJ. Role of Toll-Like Receptor 5 in the Innate Immune Response to Acute P. Aeruginosa Pneumonia. Am J Physiol Lung Cell Mol Physiol 2009;297:L1112-L1119.

61. Ramphal R, Balloy V, Jyot J, Verma A, Si-Tahar M, Chignard M. Control of Pseudomonas Aeruginosa in the Lung Requires the Recognition of Either Lipopolysaccharide or Flagellin. J Immunol 2008;181:586-592.

62. Angus DC. The Search for Effective Therapy for Sepsis: Back to the Drawing Board? JAMA 2011;306:2614-2615.

63. Ranieri VM, Thompson BT, Barie PS, Dhainaut JF, Douglas IS, Finfer S, Gardlund B, Marshall JC, Rhodes A, Artigas A, et al. Drotrecogin Alfa (Activated) in Adults With Septic Shock. N Engl J Med 2012;366:2055-2064.

64. Brook I. Microbiology and Management of Abdominal Infections. Dig Dis Sci 2008;53:2585-2591.

65. McClean KL, Sheehan GJ, Harding GK. Intraabdominal Infection: a Review. Clin Infect Dis 1994;19:100-116.

66. Wiest R, Krag A, Gerbes A. Spontaneous Bacterial Peritonitis: Recent Guidelines and Beyond. Gut 2012;61:297-310.

67. Cheong HS, Kang CI, Lee JA, Moon SY, Joung MK, Chung DR, Koh KC, Lee NY, Song JH, Peck KR. Clinical Significance and Outcome of Nosocomial Acquisition of Spontaneous Bacterial Peritonitis in Patients With Liver Cirrhosis. Clin Infect Dis 2009;48:1230-1236.

252

Nederlandse samenvatting en discussie (Dutch summary)

68. van â&#x20AC;&#x2DC;t Veer C, van den Pangaart PS, Kruijswijk D, Florquin S, de Vos AF, van der Poll T. Delineation of the Role of Toll-Like Receptor Signaling During Peritonitis by a Gradually Growing Pathogenic Escherichia Coli. J Biol Chem 2011;286:36603-36618.

69. Opal SM, Laterre PF, Francois B, Larosa SP, Angus DC, Mira JP, Wittebole X, Dugernier T, Perrotin D, Tidswell M, et al. Effect of Eritoran, an Antagonist of MD2-TLR4, on Mortality in Patients With Severe Sepsis: the ACCESS Randomized Trial. JAMA 2013;309:1154-1162.

70. Rice TW, Wheeler AP, Bernard GR, Vincent JL, Angus DC, Aikawa N, Demeyer I, Sainati S, Amlot N, Cao C, et al. A Randomized, Double-Blind, Placebo-Controlled Trial of TAK-242 for the Treatment of Severe Sepsis. Crit Care Med 2010;38:1685-1694.

71. Cawcutt KA, Peters SG. Severe Sepsis and Septic Shock: Clinical Overview and Update on Management. Mayo Clin Proc 2014;89:1572-1578.

72. von Bernuth H., Picard C, Jin Z, Pankla R, Xiao H, Ku CL, Chrabieh M, Mustapha IB, Ghandil P, Camcioglu Y, et al. Pyogenic Bacterial Infections in Humans With MyD88 Deficiency. Science 2008;321:691-696.

253

254

Addendum

Miriam H.P. van Lieshout

Addendum

Curriculum vitae Miriam Hanneke Petra van Lieshout was born in 1982 in Amsterdam, The Netherlands and grew up in Hem in a family of 6. She graduated cum laude from the OSG West-Friesland in Hoorn in 2000 and started her medical studies at the University of Amsterdam. After a research internship in vascular medicine at the Academic Medical Center and a cum laude doctorate in medicine, she was part of the committee that designed Curius, the present medical curriculum of the AMC. In Gondar, Ethiopia, she studied the prevalence of brucellosis. The plethora of infectious diseases observed in Gondar intensified her interest in infectiology. Upon completion of her clinical internships in Amsterdam and reception of her medical degree in 2008, she was awarded an AMC PhD Scholarship to perform the studies presented in this thesis, at the Center for Experimental and Molecular Medicine under supervision of Tom van der Poll. During these years, she and her partner Onno Holleboom were blessed with the birth of their two sons Imme and Fer. In 2013, Miriam started her residency in internal medicine in the Onze Lieve Vrouwengasthuis in Amsterdam under supervision of dr. Y. Smets. In 2014 she and Onno decided to dedicate yet another year to research and took their family to Philadelphia, US, where Miriam completed this thesis and their daughter Maud was born.

256

Addendum

PhD portfolio Name PhD student:

Miriam van Lieshout

PhD period:

2008-2013

Name PhD supervisor: Prof. dr. T. van der Poll PhD training Year

Workload (Hours/ECTS)

General courses -

AMC Graduate School, Practical biostatistics

2010

1.1

AMC Graduate School, Radiation protection

2009

1.7

AMC Graduate School, DNA technology

2009

2.1

AMC Graduate School, Basic Laboratory safety

2008

0.4

AMC Graduate School, Laboratory Animal course

2008

3.9

AMC Graduate School, Mass Spectrometry, Proteomics and Protein Research

2008

2.1

AMC Graduate School, Crash course

2008

0.7

2008

1.3

Specific courses -

AMC Graduate School, Infectious diseases course

Seminars, workshops and master classes -

Masterclass by Jos van der Meer

2012

0.2

Masterclass by Charles Dinarello

2011

0.2

Weekly infectious diseases journal club

2008-2013

Monthly CEMM posium

2008-2013

1.25

257

Addendum

Year

Workload (Hours/ECTS)

Presentations -

Oral presentation “Opposite roles of ASC and NLRP3 inflammasomes in host defense during infection with serotype 2 and serotype 3 streptococcus pneumonia”. 7th International Federation of Shock Societies and the thity-fifth annual conference on Shock, Miami Beach, USA Poster presentation “Differential roles of Myd88 and TRIF in hematopoietic and resident cells during murine gram-negative pneumonia”

Poster presentation “Host defense during Klebsiella pneumonia mediated by TRIF and MyD88: differential contribution of hematopoietic and somatic cells”, Toll 2011 meeting-decoding innate immunity, Riva del Garda, Italy

Poster presentation “Differential roles of MyD88 and TRIF in hematopoietic and resident cells during murine gram-negative pneumonia” 8th World Congress on Trauma, Shock, Inflammation and sepsis (TSIS), München, Germany

(Inter)national conferences -

Summer Frontiers Symposium on Innate Immunity “Training the immunological memory”, Nijmegen

7th International Federation of Shock Societies and the thity-fifth annual conference on Shock, Miami Beach, USA

Toll 2011 meeting-decoding innate immunity, Riva del Garda, Italy

8th World Congress on Trauma, Shock, Inflammation and sepsis (TSIS), München, Germany

Toll 2008 meeting: recent advances in pattern recognition, Cascais, Portugal

2012

0.5

2012

0.5

2011

0.5

2010

0.5

2012

0.6

2012

2011

1.2

2010

2008

Parameters of Esteem Grants -

AMC PhD Scholarship, Graduate school, Academic Medical Center

2008

Awards and Prizes -

258

Travel award Shock Society

2012

Addendum

List of publications Miriam H.P. van Lieshout, Sandrine Florquin, Cornelis van’t Veer, Alex F. de Vos, Tom van der Poll TIR-domain-containing adaptor-inducing interferon-β (TRIF) mediates antibacterial defense during gram-negative pneumonia by inducing Interferon-γ. J Innate Immun, accepted april 2015 van Lieshout M.H.P, Anas A.A, Florquin S, Hou B, van’t Veer C, de Vos A.F, van der Poll T. Hematopoietic but not endothelial cell MyD88 contributes to host defense during gramnegative pneumonia derived sepsis. PLoS Pathog. 2014 Sep 25;10(9):e1004368. van Lieshout M.H, van der Poll T, van’t Veer C. TLR4 inhibition impairs bacterial clearance in a therapeutic setting in murine abdominal sepsis. Inflamm Res. 2014 Nov;63(11):927-33. De Jong H.K, Koh G.C, van Lieshout M.H, Roelofs J.J, van Dissel J.T, van der Poll T, Wiersinga W.J Limited role for ASC and NLRP3 during in vivo Salmonella Typhimurium infection. BMC Immunol. 2014 Aug 13;15:30. Blok D.C, van Lieshout M.H, Hoogendijk A.J, Florquin S, de Boer O.J, Garlanda C, Mantovani A, van’t Veer C, de Vos AF, van der Poll T. Single immunoglobulin interleukin-1 receptor-related molecule impairs host defense during pneumonia and sepsis caused by Streptococcus pneumoniae. J Innate Immun. 2014;6(4):542-52 van Lieshout M.H, Scicluna B.P, Florquin S, van der Poll T. NLRP3 and ASC differentially affect the lung transcriptome during pneumococcal pneumonia. Am J Respir Cell Mol Biol. 2014 Apr;50(4):699-712. doi: 10.1165/rcmb.2013-0015OC. Miriam H.P. van Lieshout, Dana C. Blok, Catharina W. Wieland, Alex F. de Vos , Cornelis van ’t Veer, Tom van der Poll Differential roles of MyD88 and TRIF in hematopoietic and resident cells during murine gram-negative pneumonia J Infect Dis. 2012 Nov;206(9):1415-23 Arie J. Hoogendijk, Joris J. T. H. Roelofs, JanWillem Duitman, Miriam H.P. van Lieshout, Dana C. Blok, Tom van der Poll, Catharina W. Wieland r-Roscovitine reduces lung inflammation induced by lipoteichoic acid and Streptococcus pneumoniae. Mol Med. 2012 Sep 25;18:1086-95. Wieland C.W, van Lieshout M.H, Hoogendijk A.J, van der Poll T. Host defence during Klebsiella pneumonia relies on haematopoietic-expressed Toll like receptors 4 and 2. Eur Respir J. 2011 Apr;37(4):848-57

259

Addendum Nieuwdorp M, Meuwese M.C, Mooij H.L, van Lieshout M.H, Hayden A, Levi M, Meijers J.C, Ince C, Kastelein J.J, Vink H, Stroes E.S. Tumor necrosis factor-alpha inhibition protects against endotoxin-induced endothelial glycocalyx perturbation. Atherosclerosis. 2009 Jan;202(1):296-303. Nieuwdorp M, van Haeften T.W, Gouverneur M.C, Mooij HL, van Lieshout M.H, Levi M, Meijers J.C, Holleman F, Hoekstra J.B, Vink H, Kastelein J.J, Stroes E.S. Loss of Endothelial Glycocalyx During Acute Hyperglycemia Coincides With Endothelial Dysfunction and Coagulation Activation In Vivo Diabetes. 2006 Feb;55(2):480-6

260

Addendum

Dankwoord Na zoveel jaren zijn er ontzettend veel mensen wie ik dank verschuldigd ben voor hun hulp, helemaal aangezien ik het laboratorium betrad als “domme dokter”. Voorafgaand aan deze periode heeft het onderzoek waaraan ik als student bij Max Nieuwdorp deelnam en het enthousiasme van Erik Stroes er zeker aan bijgedragen om die stap te zetten. Ik hoop dat alle direct wetenschappelijke ervaring maar zeker ook de levenservaring van de afgelopen 7 jaren mij tot een veelzijdigere en betere dokter en mens hebben gemaakt, vooral ook dankzij de velen die hun inzichten en ervaring met mij deelden. Ten eerste gaat op deze dag mijn dank uit naar de leden van de promotiecommissie die het eindresultaat van mijn onderzoek kritisch hebben beoordeeld. Dan natuurlijk mijn promotor Tom van der Poll, jouw kalme persoonlijkheid en overtuiging dat er een proefschrift en promotie zouden komen hebben mij door de jaren heen geholpen. Af en toe hadden we kleine maar nooit ernstige misverstanden, te beginnen met dat jij dacht dat je me al had aangenomen en ik dat nog niet doorhad. Verder heb je me altijd aangemoedigd (“schrijven kreng!”) en iedere keer geïnformeerd naar de gesteldheid van mijn steeds maar groeiende gezin, geen vanzelfsprekendheid. Ik bewonder hoe jij al je taken combineert maar dat ook weet te relativeren, terwijl je niet vergeet één ochtend per week “levens te redden”. Daarnaast roem ik maar weer eens jouw talent om leuke collega’s aan te nemen. Kees, Alex co-promotoren. Jullie hebben ieder op geheel eigen wijze jullie aandeel in de totstandkoming van dit proefschrift. Kees, ik bewonder jouw encyclopedische kennis en jouw vermogen om data te analyseren en hypotheses te genereren over hoe nieuwe bevindingen in het grotere geheel van jouw bevindingen maar ook alle beschikbare wetenschappelijke literatuur zouden kunnen passen. Gelukkig heb je ook een oneindige hoeveelheid geduld om dit vervolgens aan simpelere zielen zoals ik uit te leggen. Dat het vervolgens bevestigen van die hypotheses dan vaak weer een stukje ingewikkelder is en meer tijd kost dan je denkt was voor mij een les in geduld en nauwkeurigheid. En de beste primers worden natuurlijk door jou ontworpen! Daarnaast ben je altijd te porren voor feestelijke activiteiten. Alex, jouw bereidheid om jouw uitgebreide ervaring met muismodellen te delen hebben mij zeker in het begin goed op weg geholpen. Daarnaast heb ik veel geleerd van jouw grondige aanpak bij moeilijke FACS proeven en de nodige bescheidenheid bij het interpreteren van onderzoeksresultaten. Jouw aanwezigheid in het laboratorium houdt iedereen op een positieve manier gemotiveerd om “netjes” onderzoek te doen en de sfeer goed te houden. Verder speciaal veel dank bij het op afstand afronden van de laatste stukjes en voor het aanleveren van foto’s de laatste maanden! Verder veel dank aan Sandrine Florquin en Onno de Boer van de afdeling pathologie bij de beoordeling en het maken van foto’s van vele, vele weefselcoupes. Dan de (ex-) analisten van het CEMM, ontelbare vragen van mij hebben jullie 261

Addendum

onvermoeibaar beantwoord. Regina, dank voor je goede humeur en het kleuren van eindeloze hoeveelheden coupes. Daniëlle, samen begonnen en nu een heel eind van ons werkzame leven verder, dank voor het lyseren van eindeloze hoeveelheden teentjes, Sanne, Miranda, Jennie en Anita ook veel dank. Monique, dank voor alle hulp en bestellingen die je hebt verzorgd. Joost en Marieke, dank voor jullie hulp en gezelschap bij de eindeloze aantallen muizenproeven die ik deed en die dankzij jullie (meestal) in elk geval in technisch opzicht slaagden, hoewel ik de proefopzet volgens jullie altijd veel te ingewikkeld maakte... Dames van het secretariaat MoniQue en Heleen, dank voor jullie hulp en ondersteuning, speciaal als het écht belangrijk of urgent is, is jouw hulp fantastich Heleen. De Tommies, een zich steeds vernieuwende en veranderende groep met altijd fijne collega’s, die goed zijn voor veel gedeelde onderzoeksfrustratie maar vooral veel lol zowel tijdens als buiten werktijd, allen dank voor de tijd die ik met jullie doorbracht. Jacobien, Masha, Joost en Joppe, jullie waren mijn ervaren voorbeelden. Cathrien, dank voor de les “hoe produceer ik efficiënt een artikel” (helaas lag dat bij veel van de volgende hoofdstukken wat ingewikkelder). Marcel, dank voor vele interessante gesprekken en het delen van jouw liefde voor muziek. Jolanda, jij was mijn rolmodel hoe ambitie, werk en moederschap efficiënt te combineren zonder teveel schuldgevoel te hebben op één van die fronten. Rianne (Gerritje), dank voor zowel je schaterlach als serieusheid. Arjan, stille kracht met rake observaties, absurde humor en een grote hulp bij schijnbaar onoplosbare software/hardware en andere technische problemen, je hebt me vaak gered. Floor, bijna tegelijk begonnen en ook klaar, veel gedeeld leed en chokotoffs, je bent oprecht, rechtvaardig en de meest lieve en attente collega die ik ken. Liesbeth, recht op je doel af en ambitieus, waar haal je de energie vandaan? Dana, eigenzinnige motormuis, Achmed(T) leuk dat we bijna tegelijk promoveren, Tijmen, drummer, OLVG-collega en tevens leverancier van de nodige grappen en cynische opmerkingen, ik hoop nog vaak koffie met je te drinken, Daan jouw energie en optimisme lijken twee keer teveel voor je lengte, Anne Jan, nooit voorspelbare onderzoeker&ondernemer, ik ben benieuwd waar het leven jou brengt. Tassili, zet ‘m op het gaat je lukken in alle opzichten. Adam, para-nimf, vele gezamenlijke fietstochten Oost-AMC en vice versa en gedeelde genotyperingservaring, fijn dat jij vandaag naast mij staat. Sacha en Katja, goed gedaan dames jullie zijn mij ruimschoots voorgegaan. Tim, Misha, Lonneke, Maryse en Ingrid, dank. Brendon family- en whole genome arrayman zonder jou geen hoofdstuk 7. Adoptie-Tommies Jan-Willem, dank voor de vele gezamenlijke automaten-koffie en Wytske, we begonnen bijna tegelijk aan het onderzoek en in het OLVG en overwonnen samen de eerste tijd daar, je bent een uitzonderlijke bikkel.

262

Addendum

Collega’s in het OLVG, opleiders en supervisoren, dank voor het vertrouwen toen ik in de kliniek begon en het zelf best een beetje zwaar vond en voor de mogelijkheid om er even tussen uit te gaan, ik heb er weer zin in! Collega arts-assistenten waarvan intussen een heel aantal ex-CEMMers, dank voor jullie humor, hulp en collegialiteit. Vrienden, door de afgelopen jaren van promotieonderzoek, kliniek en gezin heb ik sommigen van jullie te weinig gezien, ik hoop dat er nu weer meer tijd is. Speciaal Marieke, van brugklas tot paranimf vandaag, wat fijn dat wij altijd vriendinnen zijn gebleven. Suthesh, ontmoet tijdens het eerste college en tot nu gezamenlijk onderweg tot internist, alleen ben jij nu geëmigreerd, help daar moet ik nog aan wennen! Marc & Antje, een door jullie bereide maaltijd met goede wijn, thuis of ergens voor een tent met ons en nu ook jullie gezin, wat een geluk! Verder veel dank aan mijn buren uit de Fraunhoferstraat, ik had nooit gedacht zo veel met mijn buren te delen en dank voor alle babyfoonoppas- en speelmomenten die zeker hebben geholpen bij de uitvoering van schrijfactiviteiten en dienst-schema’s. Mijn uitgebreide familie, teveel om op te noemen, maar dank voor al jullie uiteenlopende voorbeelden en de gedeelde familie-band. Speciale dank aan mijn nicht Vera, voor alle aanmoedigingen, het is leuk dat wij ondanks vele verschillen toch ook gelijk op gaan. Mijn allerliefste niet-officiële schoonfamilie: Jurgen en Ineke, dank voor al jullie lieve onvoorwaardelijke steun, hulp en oppassen, zonder jullie was het leven carrièretechnisch een stuk ingewikkelder en daarnaast zijn jullie een super-opa en oma. Jasper, mijn enige zwager dank voor je luchtige en vrolijke maar als het nodig is serieuze jezelf zijn. Michiel, Saskia en Joost, wat fijn dat wij met zijn vieren brusjes zijn, ook al dachten we daar vroeger wel eens anders over. Ieder van jullie heeft op zijn eigen wijze in concrete zin aan dit proefschrift bijgedragen, van incidentele oppas bij noodgevallen, hulp bij de lay-out van dit boek tot het verzorgen van de post tijdens ons overzeese verbijf. Daarnaast is het fijn dat wij vieren, in wisselende samenstelling veel samen doen en delen, ik hoop in de toekomst (weer) meer voor jullie terug te doen! Charlotte en Pauline, fijn dat jullie ook bij de club horen. Papa en mama, jullie zijn uiteraard mijn basis en oorsprong, dank voor het vertrouwen dat jullie altijd in mij hebben gehad en jullie opvoeding die ons op bescheiden wijze toch alle mogelijkheden van onszelf en de wereld liet zien. Het is extra speciaal dat wij “collega’s” zijn. Daarnaast zijn jullie een fantastische opa en oma voor mijn kinderen, ik had nooit vermoed dat dat onze band nog zoveel meer zou versterken. Onno, liefste voordeurdeler en niet-echtgenoot, tot nu toe is het ons aardig gelukt om “gewoon te leven” en daarnaast twee proefschriften en drie kinderen voort te brengen, een huis te verbouwen en aan twee opleidingen tot internist te beginnen in de afgelopen zeven jaar. Jouw vastberadenheid en gedrevenheid zijn daarbij de factoren die ervoor zorgen dat volgens mij onmogelijke projecten toch tot een goed 263

Addendum

einde komen. Je hebt me ontelbare keren aangemoedigd vol te houden en er zelfs voor gezorgd dat ik mijn proefschrift in alle rust in Philly af kon schrijven. Je bent mijn steunpilaar en criticus, dat laatste is zeker wederzijds. Ik hoop oprecht dat we nu dit proefschrift is afgerond ook weer meer tijd met zijn tweeĂŤn hebben en wie weet, om toch ook nog eens te trouwen..? Lieve Imme en Fer, waar hebben jullie toch zulke ouders aan verdiend? Dank voor alles wat jullie zijn, jullie onvoorwaardelijke liefde en het mij afleiden van werk en proefschrift. Het is jullie goed recht. Ik hoop dat ik slaag in mijn streven jullie uit te rusten met dat wat jullie nodig hebben om in het leven je eigen weg te gaan. Lieve Maud, ons kleine Amerikaantje, jij werd geboren daags na het afronden van dit manuscript. Je bent nu al een onmisbaar lid van ons gezin. Miriam

264

Cell-specific recognition receptor signaling in antibacterial defense

Miriam H.P. van Lieshout