Mechanisms for activation and inhibition of inflammasomes

Transcription

1 University of Iowa Iowa Research Online Theses and Dissertations 2014 Mechanisms for activation and inhibition of inflammasomes John Roger Janczy University of Iowa Copyright 2014 John Roger Janczy This dissertation is available at Iowa Research Online: Recommended Citation Janczy, John Roger. "Mechanisms for activation and inhibition of inflammasomes." PhD (Doctor of Philosophy) thesis, University of Iowa, Follow this and additional works at: Part of the Immunology of Infectious Disease Commons

2 MECHANISMS FOR ACTIVATION AND INHIBITION OF INFLAMMASOMES by John Roger Janczy A thesis submitted in partial fulfillment of the requirements for the Doctor of Philosophy degree in Immunology in the Graduate College of The University of Iowa May 2014 Thesis Supervisor: Associate Professor Fayyaz S. Sutterwala

3 Copyright by JOHN ROGER JANCZY 2014 All Rights Reserved

4 Graduate College The University of Iowa Iowa City, Iowa CERTIFICATE OF APPROVAL PH.D. THESIS This is to certify that the Ph.D. thesis of John Roger Janczy has been approved by the Examining Committee for the thesis requirement for the Doctor of Philosophy degree in Immunology at the May 2014 graduation. Thesis Committee: Fayyaz Sutterwala, Thesis Supervisor Jerrold Weiss Lee-Ann Allen Jon Houtman Mary Wilson

5 To my family: Mom, Dad, Grandma, Krysta and Cait; your love, patience, and prayers have seen me through. ii

6 Have no fear of moving into the unknown. Simply step out fearlessly knowing that I am with you, therefore no harm can befall you; all is very, very well. Do this in complete faith and confidence Pope John Paul II Public Address iii

7 ACKNOWLEDGMENTS I extend my deepest gratitude to my advisor and mentor, Dr. Fayyaz Sutterwala. You have done so much for me over the past five years; you have gone above and beyond, thank you. Your time, effort, and patience have been invaluable in my development as a scientist and as a person. Thank you for directing my energy down effective pathways. You have helped me to develop the skills at and away from the bench that I will need to become an effective, and successful independent investigator. To my thesis committee: Dr. Jerry Weiss, Dr. Lee-Ann Allen, Dr. Mary Wilson, and Dr. Jon Houtman, thank you so much for your time, critiques, and help that you have provided over the past five years. I do not believe this day would have come without your help. And finally, I need to extend my gratitude to our program s administrative staff: Paulette Villhauer and Joshua Lobb. To the members of the Sutterwala and Cassel labs, both past and present, thank you for all of your help, mentorship, and friendship. To Dr. Sophie Joly and Dr. Shankar Iyer, thank you for being able to mentor me and help start me down this path. To all of the other students and postdoctoral scholars: Tyler, Stevie, Ceren, Ann, Eric, Anna Paula, and Shruti, thank you for all of the time spend discussing our projects, dreams, and goals. And to our techs Jeffery Sadler and Vickie Knepper-Adrian, without whom this work would not have happened, thank you for making sure that I never took myself too seriously, and for all of the help you have given. All of you have made this one of the best times in my life, thus far. To my family, this would not have been possible without you. Mom and Dad, you are the reason I am here. You set an example I hope that I can follow; your unconditional love and support have seen me through so much, the highs and lows of life. I cannot begin to express how thankful I am. Krysta, you are my best friend, and your support has helped me through this trying time. You have been the best sister I could iv

8 ever ask for. To my entire family that would sit politely while I excitedly described my research, thank you for smiling and nodding in the right places, I really hope I did not put you to sleep too often. You have all kept me sane throughout this process, but more importantly I would never have reached this spot without you. Cait, I cannot fully express how much you have helped me. Thank you for all of your help. You have helped shoot down some of my more off the wall ideas, both in and out of the lab. You have been my rock, and I am forever grateful. You know better than I that this dissertation would not exist if it were not for you, thank you. Now that we are both done with our time in graduate school, onward and upward! v

9 ABSTRACT Activation of the cysteine protease caspase-1 and the subsequent processing and secretion of the pro-inflammatory cytokines IL-1β and IL-18 is central to the inflammatory response as well as the induction of adaptive immune responses. Caspase-1 is activated as a part of a high-molecular weight multi-protein complex termed the inflammasome. The NLRP3 inflammasome is by far the best studied of these complexes, and it is the most promiscuous in terms of activating signals. The diversity of NLRP3 activating signals makes it likely that NLRP3 does not recognize each agonist directly, rather it detects a molecule that is generated, revealed, or altered by cellular stress. Recent studies have indicated that mitochondrial dysfunction is crucial for NLRP3 inflammasome activation, yet the activating ligand has not yet been identified. Appropriate and timely activation of this inflammatory pathway is required for host immunity to a variety of pathogens, however dysregulated activation leads to autoinflammation and potentially autoimmunity. Hence it is important to identify mechanisms for inflammasome activation and regulation. Therefore, this dissertation has focused on investigating the mechanisms for activation and regulation of the NLRP3 inflammasome, and the biological consequences of these changes. We show that the mitochondrial lipid cardiolipin is required for NLRP3 inflammasome activation. We have also identifying a novel mechanism by which inflammasome activation is regulated. Data presented in this dissertation shows that IgG immune complexes effectively suppress inflammasome activation and the subsequent processing and secretion of IL-1α and IL- 1β. Furthermore we show that immunization with IgG immune complexes suppresses both Th2 and Th17 immune responses. Together these data provide novel insights into the activating and regulatory pathways of both the innate and adaptive immune systems. vi

10 TABLE OF CONTENTS LIST OF FIGURES... ix LIST OF ABBREVIATIONS... xi CHAPTER I. INTRODUCTION...1 Pattern Recogniton...1 Pattern Recognition Receptors...2 Inflammasomes...5 Priming the inflammasome for activation...6 Inflammasome activation...7 The NLRP3 inflammasome...8 Mitochondria and Nlrp3 inflammasome activation...11 Cardiolipin...13 Inflammasomes and adaptive immune responses...14 Regulation of inflammasome activation...17 Immunomodulatory functions of IgG Fc receptors...19 Research Focus...21 II. CARDIOLIPIN IS SUFFICIENT AND REQUIRED FOR NLRP3 INFLAMMASOME ACTIVATION...24 Abstract...24 Introduction...24 Materials and Methods...26 Reagents...26 Cell culture...26 Palmitate treatment...27 Immunoblotting...27 Broken cell system...27 Measurement of mitochondrial ROS...28 Measurement of calcium flux...28 Results...29 Cardiolipin depletion inhibits Nlrp3 inflammasome activation...29 Palmitate treatment does not interfere with upstream steps in the NLRP3 activation pathway...30 Cardiolipin is sufficient for caspase-1 activation in a broken cell system...31 Discussion...31 III. IMMUNOGLOBULIN G-ANTIGEN IMMUNE COMPLEXES INHIBITS INFLAMMASOME ACTIVAITON...41 Abstract...41 Introduction...41 Materials and Methods...42 Mice...42 Reagents...43 Immune complexes...43 vii

11 In vitro stimulation of bone-marrow derived macrophages...44 Immunoblotting...44 ASC speck assay...45 Statistical analysis...45 Results...45 Immune complexes inhibit the secretion of IL-1α, IL-1β, and IL-18 in vitro...45 Immune complexes inhibit the activation of caspase Immune complexes do not prevent the synthesis of pro-il-1β, Nlrp3, ASC, or caspse Inflammasome inhibition by IgG immune complexes requires signaling though the FcRγ chain...48 Immune complexes require uptake to inhibit inflammasome activation...48 Immune complexes inhibit activation of NLRP3, NLRC4, and AIM2 inflammasomes...49 Immune complexes block oligomerization of inflammasomes...50 Disucssion...51 IV. IMMUNOGLOBULIN G-ANTIGEN IMMUNE COMPLEXES BLOCK THE DEVELOPMENT OF T H 2 AND T H 17 CD4 + T CELL RESPONSES...65 Abstract...65 Introduction...65 Materials and Methods...67 Mice...67 Reagents...68 Immune Complexes...68 In vitro stimulation of BMDM...68 Induction and evaluation of airway inflammation...68 T cell restimulation and proliferation...69 Statistical analysis...70 Results...70 Antigen-IgG immune complexes suppress the development of alum-driven Th2 and Th17 responses in vivo...70 Immune complexes suppress Th2 and Th17 responses through inhibition of IL-1α and IL-1β secetion rather than enhanced IL-10 secretion...71 Suppression of CD4 + T cell responses by immune complexes is rescued by exogneous IL-1α or IL-1β...73 Disucssion...74 V. SYNTHESIS...85 REFERENCES...97 viii

12 LIST OF FIGURES Figure 1.1 Model of NLRP3 inflammasome activation Cardiolipin depletion interferes with NLRP3 inflammasome activation Palmitate treatment does not interfere with mitochondrial ROS production Palmitate treatment does not interfere with the role of camp in NLRP3 inflammasome activation Palmitate treatment does not interfere with NLRP3 agonist induced calcium flux Cardiolipin is sufficient to activate caspase-1 in a broken cell system Cardiolipin liposomes activate caspase-1 in palmitate treated macrophages IgG immune complexes suppress Nlrp3 inflammasome activation Immune complexes block caspse-1 activation Immune complexes interfere with inflammasome priming Immune complexes do not effect the synthesis of inflammasome components Signaling through the FcRγ chain but not FcγRIIb is required for inflammasome suppression Immune complex uptake is required for inflammasome inhibitions Immune complex mediated inflammasome suppression is not via autophagy Immune complexes suppress Nlrp3 inflammasome activation in response to microbial stimuli in vitro and in vivo Immune complexes inhibit the Nlrc4 and AIM2 inflammasomes Immune complexes block inflammasome oligomerization Schematic representation of our allergic airway disease model Immunization with IgG immune complexes suppresses airway inflammation Immunization with IgG immune complexes suppresses the generation of Th2 and Th17 responses Immune complexes do not interfere with antigen-specific CD4 + T cell proliferations...80 ix

13 4.5 IL-10 is not required for immune complex mediated suppression of Th2 and Th17 responses The NLRP3 inflammasome is important for the generation of alum-induced Th2 and Th17 immune responses IL-1α and IL-1β are both important in alum-driven Th2 and Th17 responses Immune complex mediated downregulation of adaptive immune responses is rescued by exogenous IL-1α and IL-1β...84 x

14 LIST OF ABBREVIATIONS [Ca 2+ ] I ADCC ALS ASC ATP BMDC BMDM camp CAPS CARD CASR CFA ciap Cytoplasmic calcium Antibody dependent cellular cytotoxicity Amyotrophic lateral sclerosis Apoptosis-associated speck-like protein containing a CARD Adenosine triphosphate Bone marrow-derived dendritic cell Bone marrow-derived macrophage Cyclic adenosine monophosphate Cryopyrin associated periodic syndrome Caspase activation and recruitment domain Calcium sensing receptor Complete Fruend's adjuvant Cellular inhibitor of apoptosis protein Chronic infantile neurologic cutaneous and articular CINCA CLR CPPD DAMP DC DC-SIGN DMEM DPBS DPI EAE ER syndrome C-type lectin receptors Calcium pyrophosphate crystals Danger Associated Molecular Patterns Dendritic cell DC-specific ICAM-grabbing non-integrin Dulbecco's modification of Eagle's media Dulbecco's phosphate buffered saline Diphenyleneiodonium Experimental autoimmune encephalomyelitis Endoplasmic reticulum xi

15 FBS Fc FCAS FcRγ FcγR FIIND HBSS Fetal bovine serum Fragment crystallizable Familial cold autoinflammatory syndrome Fc receptor γ chain Immunoglobulin-g Fc receptor Function-to-find domain Hank's balanced salt solution HMGB1 High mobility box group 1 HSA Human serum albumin Hsp60 Heat shock protein 60 IAPP IFN IgG IL- IMM inos ITAM ITIM LDL LPS LRR MAPK MDP NACHT NK cell Islet amyloid polypeptide Interferon Immunoglobulin-g Interleukin- Inner mitochondrial membrane Inducible nitric oxide synthase Immunoreceptor tyrosine-based activation motif Immunoreceptor tyrosine-based inhibitor motif Low-density lipoprotein Lipopolysaccharide Leucine-rich repeat Mitogen activated protein kinase Muramyl dipeptide Nucleotide-binding oligomerization Natural killer cell Nucleotide-binding domain leucine rich repeat containing NLR receptor (NLR) xii

16 NOMID OMM Ova PAMP PKC PLC PRR PVDF RAR RIG-I RLR ROR ROS STAT Th TIR TLR TNF TRAF TXNIP VDAC Neonatal-onset multisystem inflammatory disorder Outer mitochondrial membrane Ovalbumin Pathogen Associated Molecular Patters Protein kinase C Phospholipase C Pattern Recognition Receptors Polyvinylidene difluoride Retinoic acid receptor Retinoic acid-inducible gene-i RIG-I like receptors RAR related orphan receptor Reactive oxygen species Signal transducer and activator of transcription T helper Toll-interleukin 1 receptor Toll-like receptors Tumor necrosis factor TNF receptor associated factor Thioredoxin interacting protein Voltage dependent anion channel xiii

17 1 CHAPTER I INTRODUCTION Pattern Recognition Throughout their lives, mammals are constantly exposed to both beneficial and pathogenic microorganisms. Under some circumstances, these microorganisms can lead to life threatening conditions. The interplay between microorganisms and multicellular organisms has lead to the development of a number of mechanisms by which multicellular organisms deal with the ever-present threat of infection. In mammals, the immune system is divided into two branches, the innate and adaptive immune systems, which are responsible for immediate and long-term immune responses, respectively. The adaptive immune response, which is found only in vertebrates, is highly specific and is responsible for lasting immunity. Cells of the innate immune system coordinate activation of the adaptive response. The innate immune system responds to a number of defined stimuli and is central to early pathogen clearance and tissue repair 1,2. Upon activation, cells of the innate immune system upregulate inflammatory mediators and induce increased expression of co-stimulatory and adhesion molecules. In turn, these molecules drive the activation of the adaptive arm of the immune system. Traditionally, it was thought that the innate immune system recognized harmful conditions, such as the presence of microbes, through its ability to distinguish pathogens as non-self 3,4. However, this paradigm cannot account for the activation that follows sterile insults, wherein the inflammatory response is triggered in the absence of invading pathogens. It is now known that the ability of cells of the innate immune system to execute their inflammatory and tissue repair programs is dependent upon germ-line encoded pattern recognition receptors (PRR), which identify molecular structures associated with cellular stress and death, known as damage associated molecular patterns (DAMPs), as well as

18 2 conserved pathogen derived structures, or pathogen associated molecular patterns (PAMPs) 3,5. PAMPs are highly conserved, microbe-derived structures. Generally these structures are distinct from host self-antigens, and are typically essential for the survival of the microbe. Recognition of highly conserved microbial structures allows for the use of a limited number of PRRs. Further, this helps eliminate the possibility of escape mutants, as these structures are required for microbial survival, and mutations would lead be lethal 1-3,6. DAMPs can be cytosolic or nuclear components, such as the chromatinassociated protein high mobility box group 1, HMGB1, cytosolic host DNA or the chaperone heat shock protein 60 (Hsp60), which under homeostatic conditions are not exposed to PRRs. Alternatively, components of the extracellular matrix, such as hyaluronan and heparan sulfate, can also serve as DAMPs 7. A number of host-derived DAMPs, such as amyloid peptides and uric acid, have been implicated in the generation of a range of inflammatory diseases, including Type 2 diabetes, Alzheimer s disease, and gout The inflammatory response can cause bystander host tissue damage and has a high metabolic cost; thus multiple regulatory mechanisms control the extent, duration, and type of response. To prevent undesirable responses, interactions of intracellular and extracellular molecules with PRRs must be tightly controlled to allow cells of the innate immune system to activate in an appropriate and timely manner. Pattern Recognition Receptors In mammals, four major families of PRRs have been identified. These include the membrane anchored Toll-like Receptors (TLRs) and C-Type Lectin Receptors (CLRs) and the cytoplasmic receptors that include Retinoic acid-inducible Gene (RIG)-I like Receptors (RLRs) and nucleotide-binding domain leucine-rich repeat containing (NLR) family receptors. TLRs are type 1 transmembrane proteins, with ectodomains containing

19 3 leucine rich repeats (LRR) that serve as ligand binding domains and mediate pattern recognition. TLRs signal through cytosolic Toll-interleukin 1 (IL-1) receptor (TIR) domains 11. A variety of PAMPs and DAMPs ranging from microbe derived lipids, lipoproteins, and nucleic acids, to host derived substances such as heat shock proteins and hyaluronan are recognized by TLRs TLRs are subdivided into two groups based upon their cellular localization. TLR1, TLR2, TLR4, TLR5, TLR6, TLR10 and TLR11 are expressed on cell surfaces, whereas TLR3, TLR7, TLR8 and TLR9 are localized to intracellular compartments. Ligand binding by TLRs leads to dimerization and activation of cellular signaling cascades, through the adapter proteins MyD88 or TRIF, culminating in the activation of the transcription factors NF-κB, AP-1, IRF-3, IRF-5 and/or IRF-7 to produce a number of inflammatory mediators 12. A family of cytosolic PRRs that recognizes cytosolic RNA and activates the type- I interferon response and NF-κB has also been identified. The three members of the RLR family, RIG-I, MDA-5, and LGP2, are members of the DExD/H-box containing RNA helicase family RIG-I and MDA-5 both possess caspase activation and recruitment (CARD) domains, and interact with MAVS (IPS-1) which is involved in the activation of the type-i interferon response 18. The recognition of RNA viruses by specific RLRs is apparently based upon RNA length C-Type Lectin Receptors are a part of a superfamily of Metazoan proteins containing C-type lectin-like domains, a compact structural motif that is generally responsible for carbohydrate recognition and substrate specificity 22. This superfamily includes more than a thousand members. CLRs are Ca 2+ -dependent carbohydrate binding lectins and have recently been identified as an important family of PRRs. They are involved in the induction of the inflammatory program either directly or through modulation of TLR signaling While TLRs, as group recognize a wide variety of PAMPs and DAMPs, CLR mediated pattern recognition is limited to carbohydrate structures mainly mannose, fructose, and glucans. CLRs can initiate gene transcription

20 4 either directly via signaling motifs present in their cytoplasmic domains, or through association with immunoreceptor tyrosine-based activation motif (ITAM) containing adapter proteins, such as the fragment crystallizable (Fc) receptor γ chain (FcRγ) 27. Substrate recognition by some CLRs, such as dectin-1 or dectin-2, can initiate inflammatory responses independently of signaling through other PRRs 26,28. Ligand engagement by CLRs such as dectin-1, DC-specific ICAM3-grabbing non-integrin (DC- SIGN), or MICL, can also modulate other PRRs signaling 24,25,29. Upon recognition of their cognate PAMPs CLRs, like TLRs, are able to activate dendritic cells (DCs), and help direct CD4 + activation 23,30. The NLR family of PRRs has been recently identified and has been shown to be involved in diverse immunological functions. NLRs can coordinate responses with both TLRs and CLRs. To date 22 members of the NLR family have been identified in humans. NLRs are characterized by a central nucleotide-binding oligomerization (NACHT) domain that is flanked by a C-terminal LRR domain and an N-terminal effector domain 31. The NLRs can be grouped based upon the particular N-terminal effector domain they contain. In general, those that contain an N-terminal pyrin domain are members of the NLRP group and those with a CARD are part of the NLRC subgroup. The biological functions of many of the NLRs have yet to be elucidated, yet an examination of the NLRs with defined functions makes it clear that the role of NLRs in regulating the innate immune response is both important and varied. For example, NOD1 and NOD2 are cytosolic PRRs that activate NF-κB in response to bacterial peptidoglycans Beyond activation of the NF-κB inflammatory pathway, NLRs modulate both the innate and adaptive immune responses. NLRP6, a pyrin domain containing NLR, has been shown to negatively regulate NF-κB activation and mitogen-activated protein kinase (MAPK) signaling, as mice genetically deficient in NLRP6 are protected from Listeria monocytogenes infection. NLRP6 -/- mice showed reduced bacterial burdens, greater inflammatory cell influx, and macrophages from NLRP6 -/- mice showed greater

21 5 activation of MAPK and NF-κB upon stimulation with either L. monocytogenes or the TLR2 agonist Pam3CSK4 36. NLRP6 was also shown to have a role in controlling intestinal epithelial cell turnover and in the wound healing response 37. NLRP10, the only NLR family member without the putative ligand binding LRR, was first identified as a negative regulator of NF-κB and IL-1β secretion, through overexpression studies 38,39. However, recent studies using NLRP10 knockout mice have shown that NLRP10 plays an integral role in the generation of adaptive immune responses 40,41. Interestingly, NLRP10 -/- DCs are unable to traffic to the draining lymph node following immunization, indicating that NLRP10 plays a role in DC migration 40. NLRP12 also regulates DC migration and plays an integral role in contact hypersensitivity 42. However, the function of NLRP12 is likely not limited to DCs migration. Studies utilizing overexpression systems and NLRP12 deficient mice have shown that NLRP12 serves as a negative regulator of NF-κB and inflammation, in a DSS induced model of colitis Despite growing understanding of NLR function, what determines the function of the individual NLRs remains unclear. Inflammasomes By far, the most well characterized function of the NLR family is regulation of inflammasome activation 31. Three members of the NLR family (NLRP1, NLRP3, NLRC4) and one PYHIN family member (AIM2) have been shown to form highmolecular weight, multi-protein complexes called inflammasomes. Inflammasomes typically contain an NLR, the adaptor protein apoptosis-associated speck-like protein containing a CARD domain (ASC), and the cystine protease caspase-1. The precise order of events leading to inflammasome activation remains elusive for all inflammasomes. By bringing two or more monomers into proximity it is believed these complexes serve to initiate the autocatalytic processing of caspase-1, which then processes the zymogens IL- 1β and IL-18 to their biologically active forms 46, and initiates a proinflammatory form of

22 6 programmed cell death called pyroptosis 47. IL-1β and IL-18 are potent inflammatory mediators, which are important in driving antigen specific adaptive immune responses and in tissue repair 48. However, unchecked IL-1β secretion can lead to fibrosis as well as autoinflammatory disorders 47. Priming the inflammasome for activation Activation of inflammasomes requires two signals. Signal one results in NF-κB activation and causes the production of pro-il-1β and pro-il-18 and upregulates the transcription and translation of NLRP3 49. Numerous inflammatory signals including TLR stimulation and activation by pro-inflammatory cytokines e.g. TNF-α or IL-1β are capable of providing this priming signal Initially, priming was thought only serve to increase the levels of pro-il-1β, thereby increasing the substrate availability for caspase- 1. Recent work on the requirement for priming of the NLRP3 inflammasome has demonstrated that priming regulates the expression of NLRP3, this increased expression is required for activation of the NLRP3 inflammasome 49. However, there is now evidence suggesting acute LPS exposure licenses NLRP3 and AIM2 inflammasome activation even upon co-administration of LPS and activating signal 52, or in the presence of cyclohexamide, which prevents synthesis of new proteins 53. Further, LPS treatment enhanced caspase-1 activation in Nlrp3 deficient bone-marrow derived macrophages that stably and constitutively expressed NLRP3 52. Recent studies have shown that priming also induces various post-translational modifications of inflammasome components that permit inflammasome activation. Under resting conditions NLRP3 is polyubiquinated via K63 linkage. However, upon LPS treatment NLRP3 is rapidly deubiquintated by BRCC3, a member of the JAMM domain-containing Zn 2+ metalloprotease deubiquitinating enzyme family, which is essential for NLRP3 inflammasome activation 53,54. Additionally, phosphorylation of ASC has been linked in NLRP3 and AIM2 inflammasome activation. A recent study by Hara

23 7 et al. elegantly provided evidence that Y144 phosphorylation of ASC by the kinases Syk and Jnk is required for inflammasome activation 55. Combined, these data have shown that priming regulates inflammasome activation via transcriptional and post-translational mechanisms, and serves as a potent check on inflammasome activation. Inflammasome Activation Signal two causes the assembly and activation of the inflammasome and can be provided by a variety of stimuli 46. The four PRRs known to form inflammasomes each respond to specific stimuli. Human NLRP1, the NLR which first defined the inflammasome 56, is unique in its domain structure among inflammasome forming molecules, in that it has an N-terminal pyrin domain (PYD), a C-terminal function-to-find domain (FIIND), and a CARD 31. The NLRP1b inflammasome has been shown to activate in response to anthrax lethal toxin NLRP1b, in conjunction with NOD2, has also been implicated in IL-1β secretion in response to muramyl dipeptide (MDP) 58, however these data have been called into question as others have not been able to replicate these results 59. Genetic analyses have also implicated NLRP1 in the autoimmune disease vitiligo 60. NLRC4 is activated by bacterial type 3 and type 4 secretion systems or cytosolic flagellin through the utilization of the NLR family members NAIP2 or NAIP5, respectively In this respect, NLRC4 is unique among inflammasome forming NLRs, as NLRC4 is the only member identified thus far that does not directly recognize its activating ligand. Phosphorylation of NLRC4 at Ser 533 by protein kinase C (PKC)δ is important for NLRC4 inflammasome activation 65. However, the precise role Ser 533 phosphorylation plays in activation of the NLRC4 inflammasome remains unclear. The mechanism by which NLRC4 and NAIPs interact and activate caspase-1 remains elusive. The PYHIN family member AIM2 also forms an inflammasome. AIM2 is composed of a PYD domain and a HIN-200C. Aim2 serves as a major cytosolic PRR for

24 8 the recognition of cytosolic DNA. Activation of Aim2 by cytosolic DNA causes the formation of an Aim2, Asc, and caspase-1 containing inflammasome Activation of the AIM2 inflammasome is important in host defense against various viruses and the Gram-negative bacterium Francisella tularensis The NLRP3 Inflammasome To date, the NLRP3 inflammasome is the best characterized of the receptors known to form an inflammasomes, and in terms of activating stimuli seems to be the most promiscuous (Figure 1.1). NLRP3 was originally identified as a gain of function mutation associated with the autoinflammatory disorders familial cold autoinflammatory syndrome (FCAS), chronic infantile neurologic cutaneous and articular syndrome (CINCA, neonatal-onset multisystem inflammatory disease/ NOMID) and Muckle-Wells syndrome, together these diseases are now known as cryopyrin-associated periodic syndromes (CAPS) 76,77. The NLRP3 inflammasome is activated by diverse stimuli that have divergent molecular characteristics, making it unlikely that NLRP3 recognizes each directly. Rather, it is more probable the actions of the various agonists converge to generate, modify or expose a common cytosolic product, which serves as the physical and activating ligand of NLRP3. One broad category of NLRP3 activators is crystalline substances, phagocytosis of which has been shown to activate the NLRP3 inflammasome. The crystals can be environmental irritants such as asbestos, silica or alum 78-82, or can be host derived DAMPs, like uric acid or cholesterol crystals 9,83. Uric acid has been shown to activate the NLRP3 inflammasome in the context of gout 9 and bleomycin induced lung injury and fibrosis 84, whereas cholesterol crystals and oxidized low-density lipoprotein (LDL) were shown to activate the NLRP3 inflammasome in atherosclerosis 83. Hornung et al. 82 demonstrated that phagocytosis of crystals leads to lysosomal disruption, supporting a hypothesis that lysosomal damage triggers NLRP3 inflammasome activation. They also

25 9 showed that disruption of the lysosome, even in the absence of particulate DAMPs within the phagosomes, was able to activate the NLRP3 inflammasome 82. Additionally, phagocytosis of both β-amyloid and islet amyloid polypeptide (IAPP), which are associated with Alzheimer s disease and type 2 diabetes respectively, induces lysosomal damage and IL-1β secretion that is dependent upon the NLRP3 inflammasome 8,85. Damage or rupture of the lysosome leads to the release of lysosomal components into the cytosol. It has been proposed that cathepsin B released from damaged lysosomes could be the direct activator of the NLRP3 inflammasome, as inhibition of lysosomal acidification and the cathepsin B inhibitor CA-074-Me were both able to prevent inflammasome activation by silica and alum, as well as influenza A virus challenge 82. However, the role of cathepsin B in inflammasome activation has been called into question because a deficiency in cathepsin B did not prevent the activation of caspase-1 86, suggesting it may be a effect of CA-074-Me other than cathepsin B inhibition that is responsible for NLRP3 inflammasome inhibition. Lysosomal disruption also fails to explain how soluble DAMPs such as ATP activate the NLRP3 inflammasome 82,87. Numerous studies have confirmed potassium efflux, increased intracellular calcium and reactive oxygen species (ROS) are necessary for NLRP3 inflammasome activation 78,79,88,89. The requirement for potassium efflux in vitro has been shown, indirectly, through the use of high concentrations of extracellular potassium that disrupts the concentration gradient of potassium between the intra-and extra-cellular spaces, as well as directly through pharmacological inhibition of potassium channels 78-81,86,90,91. While it is known that potassium efflux is necessary for NLRP3 inflammasome activation, its precise role in NLRP3 inflammasome activation remains unclear. Recent evidence has emerged of a role for calcium and cyclic adenosine monophosphate (camp) in NLRP3 inflammasome activation The requirement for increased cytoplasmic calcium ([Ca 2+ ] I ) in IL-1β release, upon treatment with ATP or nigericin, was recognized prior to the identification of the NLRP3 inflammasome 92. In

26 10 this study, Brough et al. showed that treatment of macrophages with BAPTA-AM, a membrane permeable calcium chelator, was able to significantly impair IL-1β release upon ATP or nigericin challenge. Additionally, they showed that depletion of endoplasmic reticulum (ER) calcium stores via thapsigargin treatment impairs NLRP3 inflammasome activation 92. Work by Murakami et al. 93 has suggested that increased [Ca 2+ ] I, either via ER calcium stores or influx across the plasma membrane, plays a critical role in NLRP3 inflammasome activation. They also provided evidence that Phospholipase C, through the production of the second messenger inositol 1,4,5- trisphosphate (IP 3 ), regulates NLRP3 activation. A recent study by Lee et al. 96 proposed that enhanced extracellular calcium levels can activate the NLRP3 inflammasome via the calcium sensing receptor (CASR). The authors demonstrated that activation of CASR by increased extracellular calcium, in this case CaCl 2, leads to activation of phospholipase C (PLC), and subsequent production of IP 3, decreased camp, and ultimately NLRP3- infalmmasome activation 96. The ability of CASR signaling to serve as an activating signal for NLRP3, must be taken with caution, as the addition CaCl 2 to phosphate containing buffers can lead to the formation of calcium pyrophosphate crystals (CPPD) 96, which can activate the NLRP3 inflammasome 9. A further requirement for NLRP3 inflammasome activation is the production of ROS. The importance of ROS production in NLRP3 inflammasome activation has been shown through the use of the flavoprotien inhibitor diphenyleneiodonium (DPI), or ROS scavengers, such as N-acetyl cysteine (NAC) 78,79,88. Recently, Park et al. 97 solved the crystal structure of the NLRP3 pyrin domain. They identified two conserved cysteine residues at positions 8 and 108, which potentially allow for the formation of a disulfide bridge. The formation of such a bridge might endow NLRP3 with a level of redox sensitivity, and therefore may shed light on the apparent requirement for ROS in NLRP3 inflammasome activation 97. The source of ROS has been an area of interest. It had been postulated that ROS were derived from phagosomal NADPH oxidase as Dostert et al. 79

27 11 showed that knock down of p22 phox, an essential part of a number of phagosomal NADPH oxidases, decreased IL-1β secretion. However, further studies have shown that deletion of various components of the NADPH oxidase in fact does not affect NLRP3 inflammasome activation, suggesting the source of ROS is not phagosomal NADPH oxidase 82,98,99. Mitochondria and NLRP3 inflammasome activation Mitochondria are a primary source of cellular ROS and are known to increase ROS production in response to stress conditions, including plasma membrane destabilization 100. Recently two publications have suggested that mitochondria are the source of ROS involved in NLRP3 inflammasome activation 101,102. Zhou et al. 102 showed that induction of mitochondrial ROS production following inhibition of Complex I and Complex III, through the use of rotenone and Antimycin A respectively, not only induced robust ROS production, but also IL-1β secretion that was NLRP3 dependent. Impairment of mitochondrial ROS production through RNA knockdown of the voltage dependent anion channel VDAC1 or VDAC2 decreased both ROS production and IL-1β secretion upon the addition of NLRP3 agonists 102, thus providing further evidence that mitochondria play a role in NLRP3 inflammasome activation. This connection is supported by the finding that, although both NLRP3 and Asc are associated with the ER under resting conditions, upon activation both relocate to the mitochondria 102. Mitophagy, a specialized form of autophagy that targets dysfunctional mitochondria 103, has also been shown to play a role in NLRP3 inflammasome activation as RNA-mediated knockdown of the autophagy proteins beclin-1 and ATG5 102, as well as genetic disruption of LC3, ATG16L1 and beclin-1 101,104 caused increased IL-1β secretion upon challenge with NLRP3 agonists. Disruption of mitophagy through deletion of LC3 not only increased basal production of mitochondrial ROS, but mitochondria also showed increased damage upon treatment with the NLRP3 agonist ATP 101. Damage to

28 12 mitochondria can cause release of mitochondrial components into the cytosol, and it has been proposed that one or more of these components could be the ligand for NLRP3. It has been noted that thioredoxin interacting protein (TXNIP) dissociates from thioredoxin and associates with NLRP3 under conditions of oxidative stress 105. Further, TXNIP also localizes to the mitochondria upon addition of NLRP3 agonists 102. Nevertheless, another report indicates that mitochondrial DNA is released into the cytosol following mitochondrial damage, and this release is necessary for NLRP3 inflammasome activation 101. Interestingly, the requirement for calcium influx for NLRP3 inflammasome activation provides a link from potassium efflux to mitochondrial stress. It has been shown that potassium efflux precedes the rise in [Ca 2+ ] 93 I. Further, it is well established that calcium mobilizes from the ER rapidly, and these changes in [Ca 2+ ] I can cause mitochondrial damage 93, Together these reports strongly support the notion that mitochondria are critically involved in NLRP3 inflammasome activation, at least in part through serving as the main cellular source for the requisite ROS, and potentially as the source of the activating signal for NLRP3. While ROS may be directly required, it is equally likely ROS may be an earlier common step in the path to NLRP3 activation, triggering events in the mitochondria that lead to NLRP3 activation. Given the integral role of mitochondrial damage and dysfunction in NLRP3 inflammasome activation, it is not surprising to note that there are several parallels between NLRP3 activation and the apoptotic machinery, specifically the apoptosome. At first glance, the most obvious parallel is the central role of mitochondria in both NLRP3 inflammasome induced pyroptosis 110 and apoptosis 111. Activation of the NLRP3 inflammasome shows a general requirement for mitochondrial ROS 101,102, which is known to be important for the induction of apoptosis 112. As stated above one of the central mediators of NLRP3 inflammasome activation is K + efflux. This also appears to be a shared requirement for apoptosome formation and activation Cain et al. 114

29 13 established that physiological intracellular concentrations of K + are able to effectively suppress formation of the apoptosome. Furthermore, the authors showed that physiological concentrations of K + increases the amount of cytochrome C required to induce Apaf-1 oligomerization 114, and therefore apoptosome formation. Recent studies have also demonstrated that apoptotic stimuli, such as UV radiation and staurosporine, are able to activate the NLRP3 inflammasome 117,118. It has even been suggested that cellular inhibitor of apoptosis proteins (ciap1 and ciap2), which serve as negative regulators of apoptosis and promote cell survival 119, are required to efficiently activate caspase Labbe et al. showed that ciap1 and ciap2, in conjunction with Tumor necrosis factor (TNF) receptor associated factor (TRAF) 2, direct nondegradative K-63 polyubiquitination of caspase Given the similarities between NLRP3 inflammasome activation and apoptotic cell death, it is reasonable to hypothesize that the molecules that regulate the initiation apoptosis and apoptosome formation may be involved in the initiating events of NLRP3 inflammasome activation. Cardiolipin Cardiolipin is a tetraacylated diphosphatidylglycerol, non-bilayer forming, mitochondrial membrane lipid that is involved in a number of mitochondrial functions, including apoptosis 121. It is also found in the membranes of gram-positive bacteria 122. Cardiolipin is considered the signature lipid of mitochondria 123 as it is found specifically in mitochondrial membranes. Under homeostatic conditions in the mitochondria, cardiolipin is primarily localized to the inner mitochondrial membrane (IMM), with trace amounts found on the outer mitochondrial membrane (OMM) 124. Under resting conditions, cardiolipin appears to be important in maintaining a number of protein complexes, including the respiratory supercomplexes III and IV 125. Beyond these roles, recent work suggests that cardiolipin plays an integral role in the apoptotic pathways. Cardiolipin tightly binds cytochrome C, and this bond must be broken for

30 14 cytochrome C to be released from the mitochondria during apoptosis Dissociation of cytochrome C from cardiolipin requires peroxidation of cardiolipin, which appears to be mediated by cytochrome C 129,130. Cardiolipin is also integral to the recruitment of the proapoptotic factors truncated BID (tbid) and BAX to the contact sites of the IMM and the OMM Furthermore, it has been shown via the use of 10-N nonyl-acridine orange, a fluorescent dye that preferentially binds cardiolipin, that cardiolipin is available on the OMM following apoptotic stimuli 135. Together, all of these data clearly show that cardiolipin is a critical regulator of the apoptotic pathway. Additionally, these data demonstrate that under resting conditions cardiolipin is effectively sequestered on the IMM, and that apoptotic stimuli can make it available to cytosolic components. This would be consistent with the idea that cardiolipin serves as a signal indicating mitochondrial dysfunction. Given, that mitochondrial dysfunction is integral to the activation pathway of the NLRP3 inflammasome, and that NLRP3 can be activated by apoptotic stimuli 117,118, it seems likely that cardiolipin would be available upon inflammasome activation. Therefore, it is reasonable to hypothesize that cardiolipin, in some fashion, regulates NLRP3 inflammasome activation. Inflammasomes and adaptive immune responses Activation of the adaptive immune response is intimately linked with activation of the innate immune response through the upregulation of inflammatory, co-stimulatory, and adhesion molecules. A great deal of work has been performed that shows that pathogen recognition by PRRs, such as TLRs, is essential in the generation of protective adaptive immune responses The polarizing cytokine environment present during the initiation of antigen-specific T cell responses directs naïve CD4 + T cell responses toward distinct functional lineages. This environment is dependent upon the activation of innate immune cells via PRRs. Polarized CD4 + T cells express defined cytokine profiles and are regulated by specific transcription factors that serve as master regulators for that

31 15 subset. The cytokines IL-1β, IL-1α, IL-18 and IL-33, members of the IL-1 family of cytokines, drive CD4 + T cell responses. IL-1 has long been known to be able to help induce T cell proliferation and activation IL-1α and IL-1β, which both signal through the same receptor, are important in the generation of Th17 responses In a recent study, Chung et al. 142 showed that Il1r1 -/- mice are protected from experimental autoimmune encephalomyelitis (EAE), an experimental model of the autoimmune disease multiple sclerosis, which is driven by a Th17 responses. The authors showed that signaling of IL-1α and IL-1β through IL-1R1 is critical for the generation of Th17 cells, the induction of the transcription factor retinoic acid receptor (RAR) related orphan receptor (ROR)γt as well as the expansion and maintenance of differentiated Th17 cells. IL-18, another cytokine that is processed by caspase-1, is important for and acts synergistically with IL-12 in the generation of Th1 responses 146. Inflammasome activation and the subsequent secretion of IL-1β and IL-18 are important for the generation of adaptive immune responses. However, the role of the NLRP3 inflammasome in the generation of adaptive responses elicited by certain adjuvants is not fully understood. The adjuvant properties of alum, which generate Th2 responses, have been reported to be dependent upon the NLRP3 inflammasome, as Nlrp3 - /- mice showed decreased antibody production, and less T cell proliferation as compared to wild-type mice 80,147. Furthermore, in a model of allergic airway disease Nlrp3 -/-, Asc -/- and Casp1 -/- mice showed significant decreases in pulmonary eosinophil infiltration, and a significant decrease in IL-5 secretion from lymphocytes 80. These data suggest that the NLRP3 inflammasome is important for the generation of adaptive immune responses upon immunization with alum. However, others studies have suggested that NLRP3 and Caspase-1 mice are dispensable in the induction of allergic airway disease as measured by cytokine secretion into the airway or by T cells, as well as airway resistance, or pulmonary inflammation 148,149. These data would indicate that, while alum induces inflammasome activation and the subsequent secretion of inflammatory mediators, this

32 16 inflammatory response is dispensable for the generation of alum-induced Th2 responses. Given these disparate findings in similar experimental settings, it remains to be seen precisely what role the NLRP3 inflammasome and its products play in the generation of Th2 responses in allergy. As discussed above, the generation of a Th17 T cell responses depends upon IL- 1R engagement by IL-1α or IL-1β. IL-17A and IL-17F produced by Th17 cells act on a broad range of cell types to mediate the production of a variety of inflammatory cytokines, chemokines, and matrix mettaloproteinases 150. Th17 responses are highly inflammatory, and are important for protection against a variety of pathogens, such as Candida albicans and Bordetella pertussis 30,143,151,152. IL-1, TGF-β, IL-6 and IL-23 are responsible for driving Th17 responses 142,143, Given, the importance of IL-1 in the development of Th17 responses it is not surprising that various NLRP3 inflammasome agonists are important in driving protective Th17 responses. The fungus C. albicans is an opportunistic pathogen that is a major concern for immunocomprimised populations. Various innate immune pathways play important roles controlling C. albicans infection 158, among them is the activation of the NLRP3 inflammasome 86,159,160. Activation of the NLRP3 inflammasome has been shown to be central in driving Th17 immune responses, which are required for host immunity to C. albicans 161,162. While the Th17 response is important in protective immunity, it can drive certain experimental autoimmune diseases, specifically collagen induced arthritis and EAE 154,156. Additional studies provided evidence that activation of the NLRP3 inflammasome can drive EAE and Th17 responses to low dose Complete Fruend s Adjuvant (CFA), or heat killed Mycobacterium spp. emulsified in mineral oil 163. However, this study also showed that at higher doses of CFA the induction of Th17 responses can be induced independently of NLRP3 activation 163. Uric acid, a prototypical NLRP3 agonist, has also been shown to induce Th17 responses 164. Together, these data all strongly implicate inflammasomes as a critical regulator of the induction of adaptive immune responses.

33 17 Regulation of inflammasome activation Control of the inflammatory response is of paramount importance, as inflammation is metabolically expensive and leads to bystander host tissue damage. This is especially true of inflammasomes, given that the products of its activation are highly inflammatory. Unchecked inflammasome activation and secretion of IL-1β is central to the development of autoinflammatory disorders, specifically the CAPS family of disorders 76,77. Murine models have also implicated dysregulated IL-1β secretion is important in the pathogenesis of Kawasaki s Disease 165,166, a pediatric acute febrile disorder 167. Furthermore, dysregulation of IL-1β is also central to the pathologies associated with ischemia/reperfusion injury and tumor development 51,168. With regard to ischemia/reperfusion injury, Iyer et al. 51 demonstrated that Nlrp3 -/- mice largely survived ischemic acute tubular necrosis, whereas nearly all wild-type mice died following the procedure. Compared to wild-type mice Nlrp3 -/- mice also showed significantly lower production of inflammatory mediators, such as KC (CXCL1) and IL-1β, fewer neutrophils infiltrating the kidney, and better renal function as measured by plasma creatinine and urea levels 51. Additionally, caspase-1 activation and IL-1β but not IL-18 secretion accelerates the development of a mouse model of amyotrophic lateral sclerosis (ALS) 169. These data show that overly vigorous or inappropriate activation of the NLRP3 inflammasome can lead to severe pathological consequences. Given the potential for severe pathological consequences associated with inflammasome activation, it is not surprising that there are a number of regulatory mechanisms targeting inflammasome activation. Type-1 interferon has been known to have anti-inflammatory effects, and has been used as a successful treatment multiple sclerosis 170,171. Recent studies have suggested that type-1 interferon (IFN) can effectively inhibit inflammasome activation 163,172. Guarda et al. 172 recently showed that IFN-α and IFN-β suppress activation of both the NLRP1b and NLRP3 but not the AIM2 and NLRC4 inflammasomes, via two mechanisms. First, the authors showed that type-1

34 18 interferon decreases synthesis of pro-il-1β through the induction of IL-10 and subsequent activation of Signal Transducer and Activator of Transcription (STAT) They further showed that type-1 interferon activation of STAT1 is able to directly block caspase-1 activation, however they were unable to identify by what mechanism this occurs. Finally, the authors showed that type-1 interferon, induced by poly I:C, was able to block IL-1β secretion in vivo 172. Activated T cells have also been shown to serve as potent regulators of inflammasome activation in antigen presenting cells 173. In this study, Guarda et al. 173 showed that activated T effector and memory cells are able to suppress activation of the NLRP1b and NLRP3 but not the NLRC4 inflammasome. This study demonstrated that this suppression requires cell-to-cell contact, via TNF-family ligands, such as RANKL and CD40L expressed by activated T cells 173. However, the precise mechanism by which these TNF-family ligands are able to suppress inflammasome activation remains unclear. Post-translational modifications of NLRP3 have been shown to disrupt NLRP3 inflammasome activation 174. Recently, Mishra et al. 174 demonstrated that production of nitric oxide can inhibit NLRP3, but not AIM2, inflammasome activation. The authors found that the induction of IFN-γ by Mycobacterium tuberculosis lead to thiolnitrosylation of NLRP3, via the production of nitric oxide. Additionally, loss of Ifng or Nos2 lead to enhanced immunopathology, which could be traced to enhanced NLRP3 inflammasome activation and subsequent IL-1β secretion. Furthermore, they suggested IFN-γ produced by activated T cells down-regulated Nlpr3 activity, and that inclusion of IFN-γ during priming lead to inducible nitric oxide synthase (inos) dependent thiolnitrosylation of NLRP3 and decreased IL-1β processing 174. These data strongly suggest that cross-talk between the adaptive and innate immune systems, as well as between different innate immune pathways can serve as critical regulators of the inflammasome activation pathways.

35 19 Immunomodulatory functions of IgG Fc receptors Immunoglobulin G (IgG) Fc receptors (FcγR) were first identified when it was observed that opsonization of erythrocytes with IgG antibodies cause the erythrocytes to become cytophilic for macrophages 175. Since this observation over 45 years ago, FcγRs have been shown to be integral in immune responses, through both their direct activities and various mechanisms of immune modulation. Signals from FcγRs can be either activating or inhibitory in nature, dependent upon the specific FcγR and the cell type being acted upon In mice, there are four FcγRs, of which three are activating (FcγRI, FcγRIII and FcγRIV) and one is inhibitory (FcγRIIb). Orthologs of these four proteins have been identified in primates and humans 176. Ligation of FcγRs leads to a myriad of cellular effects, such as the initiation of phagocytosis, induction of antibody dependent cellular cytotoxicity (ADCC) in natural killer (NK) cells, and repression of antibody production by B cells The activating FcγRs signal through the ITAM containing adapter molecule, FcRγ 185,186. Crosslinking of these receptors leads to calcium mobilization via PLC-γ, activation of the protein tyrosine kinase Syk through the Src family kinases lck, lyn, and hck, as well as the eventual activation of the mitogen activated protein kinase (MAPK), Erk1/2 184, The activating receptors serve many functions within the cells, including the initiation of phagocytosis 179 and inflammatory reactions Ravetch s group has shown that activation of FcγRs was sufficient for the Arthus reaction, as well as the initiation of models of immune thrombocytopenia Initially, it was thought that activation of the complement cascade and signaling through complement receptors mediated much of the inflammation associated with the Arthus reaction. However, Sylvestre et al. showed in two studies that complement is dispensable, while FcγRs are required for the Arthus reaction 196,197. Furthermore, numerous studies have shown that IgG opsonization of a target is able to enhance DC activation, maturation, antigen

36 20 presentation and cross-presentation, and therefore enhance the subsequent generation of adaptive immune responses The inhibitory Fc receptor, FcγRIIb, signals through a immunoreceptor tyrosinebased inhibitory motif (ITIM) on its cytosolic portion 205. Ligation of the FcγRIIb has been shown to have a powerful inhibitory effect on the immune system. It was shown that the Fc portion of IgG is able to suppress the antibody response 181. Ligation of FcγRIIb on B cells blocks B cell activation 182. Furthermore, IgG-ovalbumin (Ova) immune complexes suppress the generation of a novel anti-ova antibody response 180. FcγRIIb -/- mice that were immunized with IgG-Ova showed an enhanced anti-ova antibody response as compared to wild-type animals. These data showed that ligation of this receptor can serve as a potent check on the production of inflammatory antibodies 180. In contrast to the known pro-inflammatory effect of IgG-immune complexes and their interactions with activating FcγRs, there is evidence to suggest that IgG-antigen immune complex ligation of activating FcγRs can have potent immunosuppressive activities on macrophages and DCs Furthermore, IgG immune complexes can strongly influence the generation of novel adaptive immune responses 206,207, Early work by Sutterwala et al. 206,207 established that ligation of FcγRI by IgG-immune complexes on macrophages that had been treated with LPS strongly suppressed secretion of the proinflammatory cytokine IL-12p40, while concurrently upregulating the secretion of the anti-inflammatory cytokine IL-10. The changes in cytokine production were due to changes in the transcription of these genes 206,207. Follow up work by Lucas et al. 213 identified a histone modification of the Il10 promoter as being responsible for enhance IL-10 secretion in macrophages that were primed with LPS in the presence of IgG immune complexes. They demonstrated that pharmacological inhibition and sirna knockdown of both p38 and ERK MAP kinases lead to an abrogation in enhanced IL-10 secretion. Furthermore, they determined via chromatin immunoprecipitation that ERK

37 21 activation lead to serine phosphorylation on histone H3, thus making the Il10 promoter region more accessible 213. IgG immune complexes can have drastic effect on immune responses to infection. Miles et al. 214 showed that cutaneous infection with Leishmania major in the presence of anti-l. major-antibodies leads to enhanced disease, as measured by lesion size. They further found that immunization of L. major resistant C57Bl/6 mice against Ova, and then infection in the presence of ova lead to increases in both parasite burden and lesion size. Additionally, they demonstrated that inclusion of anti-l. major-antibodies during infection lead to a Th2 biased adaptive immune response, as well as enhanced IL-10 production. Finally, they found the enhancement of IL-10 production lead to increased lesion size, as IL-10 neutralizing antibodies were able to reverse the phenotype 214. These data demonstrate that IgG-immune complexes can be highly immunomodulatory. Thus far, the effect of IgG immune complexes on the generation of Th1 immune responses has been addressed. However, the affect of these immune complexes on IL-1β production and the generation of Th17 responses are currently unknown. Research Focus The studies discussed in this introduction provide robust evidence for the importance of inflammasomes in both the innate and adaptive immune systems. While aberrant inflammasome activation is linked to the pathogenesis of autoinflammatory disorders 76,77, inflammasome activation during infection is important for both immediate and long-term immunity 86, ,215. Given this central role for the NLRP3 inflammasome in both pathology and immunity it is important to determine the precise mechanism by which it is activated, mechanisms for its suppression, as well as the biological consequences of suppression of the NLRP3 inflammasome. The work in this dissertation addresses three primary areas. First, I will show that cardiolipin is critically important in the activation of the NLRP3 inflammasome. In

38 22 particular, I will focus on the effect of disrupting cardiolipin synthesis on NLRP3 inflammasome activation. I will also show the effect of adding cardiolipin to a broken cell system to promote the activation of caspase-1. Given, the similarities discussed between NLRP3 inflammasome activation and apoptosis, and critical role cardiolipin plays in apoptosis I hypothesize that cardiolipin will play a critical role in the activation pathway of the NLRP3 inflammasome. Second, I will investigate the effect of IgG immune complexes on NLRP3 inflammasome activation. Immune complexes are known to have profound immunomodulatory effects, and their interactions with the activating FcγRs are highly inflammatory and depending on the local environment, can also be highly antiinflammatory. Nevertheless, their effect on inflammasome activation has not been investigated. Given the potent immunomodulatory effects of IgG immune complexes, I hypothesize that IgG immune complexes will suppress inflammasome activation. Finally, I will investigate the effect of IgG immune complexes on the generation of adaptive immune responses following immunization against the model antigen ova, using adjuvant alum. For these studies we will focus primarily on the generation of Th17 responses. Given the importance of the IL-1R signaling in the generation of Th17 responses, and the requirement for caspase-1 activation in the processing and activation of IL-1β, I hypothesize that immunization with IgG immune complexes will effectively suppress the generation of Th17 immune responses. The successful completion of these studies will significantly enhance our understanding of how the NLRP3 inflammasome is both activated and regulated. This work will also provide novel insights into the immunoregulatory functions of IgG immune complexes for both the innate and adaptive immune responses. Furthermore, our work may provide a novel mechanism for adaptive immune response control of the innate system. Finally, this work could provide evidence for the ability of humoral immune responses to modulate or suppress the initiation of novel cellular immune responses.

39 Figure 1.1: Model of NLRP3 inflammasome activation. Structurally and biologically diverse agonists activate the Nlrp3 inflammasome. Agonists include soluable DAMPs such as ATP, pathogens like C. albicans, and crystalline irritants like silica. Common upstream steps in the activation pathway include potassium efflux, increased cytosolic calcium, the production of mitochondrial ROS, decreases in intracellular camp, and mitochondrial dysfunction. Ultimately, activation of NLRP3 by its cognate ligand activates caspase-1, which leads to the secretion of the pro-inflammatory cytokines IL-1β and IL

40 24 CHAPTER II CARDIOLIPIN IS SUFFICIENT AND REQUIRED FOR NLRP3 INFLAMMASOME ACTIVAITON Abstract NLRP3 inflammasome activation occurs in response to numerous agonists, but the specific mechanism by which activation occurs remains unclear. Virtually every agonist evaluated induces the generation of mitochondrial ROS, with the notable exception of the oxazolidinone antibiotic linezolid. Several studies have provided evidence that both ROS-dependent and ROS-independent NLRP3 activation converge upon mitochondrial dysfunction. Previous work in our lab has shown that NLRP3 specifically binds the mitochondrial lipid cardiolipin. Here we show that interference with cardiolipin synthesis specifically inhibits NLRP3 inflammasome activation. We further show that cardiolipin is sufficient, in a broken cell system, to activate caspase-1. Together these data suggest that mitochondria play a critical role in the activation of the NLRP3 inflammasome, through the direct binding of NLRP3 to cardiolipin. Introduction The NLRP3 inflammasome is a multi-protein complex consisting of the NLR family member NLRP3, the adaptor protein ASC, and the cysteine protease caspase The NLRP3 inflammasome can activate caspase-1 in response to cellular danger, resulting in the processing and secretion of the proinflammatory cytokines IL-1β and IL- 18 9, A diverse array of stimuli can activate the NLRP3 inflammasome, including both PAMPs and endogenous host-derived molecules indicative of cellular damage 31,110. Because of the divergent qualities of the NLRP3 inflammasome agonists, it has been hypothesized that these agonists converge on a common pathway, with a final endogenous ligand activating NLRP3. NLRP3 agonists share several attributes, including

41 25 the ability to induce potassium efflux, the generation of ROS, and the induction of mitochondrial dysfunction 78-80,88,101,102,118,219. Importantly, each of these events had been shown to be required for NLRP3 inflammasome activation. Importantly, recent studies have revealed that the ROS generated by NLRP3 agonists are of mitochondrial origin, and are generated independently of the phagosomal NADPH oxidases. 98,101,102,220. Combined, these data suggest a model wherein ROS serves as a requisite upstream mediator of NLRP3 inflammasome activation. Despite this, recent data has suggested that ROS production is not essential for all NLRP3 inflammasome agonists. Linezolid, an oxazolidinone antibiotic, was recently shown to be an NLRP3 inflammasome agonist by our lab 118. In this study, we determined that linezolid activated caspase-1 in an NLRP3-dependent manner. Furthermore, we found that, like other studied Nlpr3 agonists, linezolid required potassium efflux for caspase-1 activation. Unlike other NLRP3 agonists, linezolid did not require ROS production, as pharmacological inhibitors of ROS failed to prevent IL-1β secretion. Additionally, we found that treatment of J774A.1 macrophages with linezolid failed to induce mitochondrial ROS production as measured by MitoSOX red. These data strongly suggest that ROS production is not an absolute requirement for NLRP3 inflammasome activation. Our lab and others have reported that mitochondrial dysfunction is an important step in the activation of the NLRP3 inflammasome 101,102,118. Previous studies have also demonstrated that NLRP3 associates with the mitochondria upon activation 102,221. Data generated in our lab has shown that NLRP3 interacts specifically with the mitochondrial lipid cardiolipin, a non-bilayer forming phospholipid 118. We have also reported that cardiolipin interacts with the leucine-rich repeat (LRR) region of NLRP3, suggesting that cardiolipin may regulate NLRP3. The work described in this chapter addresses the role of the mitochondrial lipid cardiolipin in the activation of the NLRP3 inflammasome. Based

42 26 on previous studies by our lab we postulated that cardiolipin may be a critical regulator of NLRP3 inflammasome activation. Materials and Methods Reagents Dulbecco s phosphate-buffered saline (DPBS) (with or without Ca 2+ /Mg 2+ ), Hank s balanced salt solution (HBSS), and Dulbecco s Modification of Eagle s Medium (DMEM) were purchased from Mediatech Inc. Cardiolipin, phosphatidic acid, and phosphatidylcholine were from Avanti Polar Lipids. Phosphatidylserine, palmitate, and ATP were from Sigma. Rotenone was purchased from Calbiochem. E. coli LPS (serotype 0111:B4) was from Invivogen. Silica (Min-U-Sil-5) was purchased from Pennsylvania Glass Sand Corporation. KH7 was from Tocris Biosciences. ELISA antibody pairs for murine IL-1β were from R&D systems. ELISA antibodies pairs for TNF-α were purchased from ebiosciences. Cyclic AMP XP assay kit was purchased from Cell Signaling. Fura-2 AM and MitoSOX were from Invitrogen. Rabbit polyclonal anti-mouse caspase-1 p10 antibody (SC-514) was from Santa Cruz Biotechnology. Mouse monoclonal anti-mouse GAPDH was from Calbiochem. Cell culture J774A.1 murine macrophages (ATCC) were maintained DMEM supplemented with 10% fetal bovine serum (FBS), penicillin (100 U/mL), streptomycin (100 µg/ml), L-glutamine (2 mm), at 37 C and 5% CO 2. For LPS priming, cells were treated with 50 ng/ml LPS for 4-6 hours at 37 C. Where indicated, cells were challenged with silica (50 µg/cm 2 ), ATP (5 mm), or KH7 (100 µm) and then incubated at 37 C and 5% CO 2 for an additional 6-12 hours. For ATP stimulation, the media was changed 20 minutes after the addition of ATP.

43 27 Palmitate treatment J774A.1 murine macrophages were incubated with 0.5 mm palmitate in a 2:1 molar ratio with human serum albumin (HSA) for 8-16 hours in serum-free DMEM. To produce the palmitate/hsa complex, a known mass of sodium palmitate was dissolved chloroform:methanol (2:1). A volume equivalent to 2.56 mg palmitate was aliquoted, and the organic solvent was evaporated using nitrogen. Upon addition of 330 mg of HSA (25%) to the palmitate, the solution was sonicated in a bath sonicator for 10 minutes. Serum-free DMEM was then added to a final volume of 20 ml. Immunoblotting Electrophoresis of lysates was performed using the NuPAGE system (Invitrogen) according to the manufactures protocol. Briefly, lysates were prepared with LDS sample loading buffer (Invitrogen) and sample reducing buffer (Invitrogen). Proteins were resolved on 4-12% Bis-Tris polyacrylimide gels, and transferred to polyvinylidene difluoride (PVDF) membranes by wet electrophoresis. Membranes were blocked with 5% milk and incubated at 4 C in primary antibody overnight, followed by incubation in HRP-conjugated secondary antibody for two hours at 4 C. Proteins were detect by enhanced chemiluminescence, and imaged using the LI-COR Odyssey (LI-COR Biosciences) imaging system. Broken cell system LPS-primed J774A.1 cells were resuspended (1 x 10 7 cells/ml) in DPBS (with Ca 2+ /Mg 2+ ) supplemented with 25 mm KCl. For certain experiments, cells were pretreated with 0.5 mm palmitate for 8 hours as described prior to priming. Cells were lysed by nitrogen cavitation as previously described 51, after which lysates were incubated at 37 C for one hour in the presence of the indicated liposomes. Lysates were then analyzed by SDS-PAGE and immunoblot. Liposomes were produced as previously described 222. In brief, 10 mm stocks of each lipid were prepared as follows: cardiolipin in

44 28 ethanol, phosphatidylcholine in chloroform:methanol (2:1), phosphatidic acid in chloroform:methanol (2:1), and phosphatidylserine in chloroform. Stock lipids were combined with phosphatidylcholine (1:1 molar ratio), which served as the vehicle for the liposomes. Organic solvents were evaporated under nitrogen and lipids were resuspended in DPBS. Liposomes were then sonicated with a probe sonicator for 10 minutes. Cardiolipin liposomes were added at concentrations of 300, 100, or 30 µm; phosphatidylserine, phosphatidic acid, and phosphatidylcholine liposomes were added at a concentration of 300 µm. Measurement of mitochondrial ROS Production of mitochondrial ROS was measured with MitoSOX red according to the manufacturer s protocol. Briefly, cells were pretreated with 0.5 mm palmitate as described above. Cells were resuspended and aliquoted (1 x 10 6 cells) to Falcon tubes. Cells were then LPS primed, washed and resuspended in 1 ml of Hank s balanced salt solution (HBSS). Macrophages were treated with rotenone (40 µm) for one hour, with MitoSOX red (2.5 µm) added for the final 15 minutes of stimulation. Data were acquired with an Accuri C6 flow cytometer (BD Biosciences) and analyzed with CFlow Plus software (BD Biosciences) and FloJo software (Tree Star, Inc.). Measurement of calcium flux Calcium flux was measured in single cells as described previously 95. Briefly, J774A.1 cells were pretreated with 0.5 mm palmitate as described above, plated in 12 well chamber slides, and primed with LPS. Cells were loaded for 30 minutes at 37 C with 2 µm fura-2 AM in the presence Pluronic F127 in Tyrode solution containing 138 mm NaCl, 2.7 mm KCl, 1.06 mm MgCl 2, 1.8 mm CaCl 2, 5.6 mm glucose and 12.4 mm HEPES (ph 7.4). The fluorescence emission at 510 nm of fura-2 excited at 340 nm and 380 nm was acquired and analyzed with the NIS-Elements Advanced Research software

45 29 package (Nikon) using an inverted Nikon To-E microscope equipped with a Xenon lamp (Hamamatsu). Results Cardiolipin depletion inhibits NLRP3 inflammasome activation It has been established that NLRP3 associates with the mitochondria upon activation 102,118,221. Previously, our lab has shown that cardiolipin, a non-bilayer-forming phospholipid found only on the inner mitochondrial membrane in eukaryotic cells, is able to bind directly to NLRP3 via the LRR domain 118. Importantly, binding of cardiolipin to NLRP3 was specific, as NLRP3 only weakly associated with other lipids. However, the significance of the NLRP3-cardioipin interaction remained unclear. To determine if the association between cardiolipin and NLRP3 had biological relevance to NLRP3 inflammasome activation, we interfered with cellular cardiolipin synthesis. Treatment of cells with the saturated long chain fatty acid palmitate (C16:0) has been shown to diminish cardiolipin synthesis in cardiomyocytes 223. We confirmed by thin-layer chromatography that J774A.1 macrophages grown in serum-free media in the presence of palmitate generated less cardiolipin than control cells 118. Strikingly, treatment of LPSprimed macrophages grown in serum-free media with palmitate markedly reduced IL-1β secretion in response to the NLRP3 agonists silica and ATP (Figure 2.1). In contrast, palmitate treatment did not affect TNF-α production in response to LPS priming (Figure 2.1). We also found that palmitate treatment did not affect IL-1β secretion in response to the NLRC4 agonist Pseudomonas aeruginosa 118, indicating that cardiolipin depletion specifically affected activation of the NLRP3 inflammasome. Inhibition of NLRP3 activation was specific to palmitate, as oleate, which does not interfere with cardiolipin synthesis 223, did not suppress IL-1β secretion 118.

46 30 Palmitate treatment does not interfere with upstream steps in the NLRP3 activation pathway The vast majority of NLRP3 agonists require the production of ROS, for inflammasome activation with the notable exception of the antibiotic linezolid 110,118. To determine if palmitate treatment interfered with the ability of mitochondria to produce ROS, we induced mitochondrial ROS production by challenging palmitate treated cells with the mitochondrial complex 1 inhibitor rotenone (Figure 2.2). Mitochondrial ROS production as measured by MitoSOX red, a mitochondria specific probe for ROS, remained intact in palmitate fed cells. Recent studies have also implicated calcium flux and camp as critical regulators of NLRP3 inflammasome activation 93,95,96. We confirmed that the generation of camp was intact in palmitate treated cells (Figure 2.3A). Previous studies have determined that the adenylate cyclase inhibitor KH7 induces IL-1β secretion from macrophages in an NLRP3 inflammasome dependent manner, additionally this study suggested that reductions in intracellular camp is a prerequisite for NLRP3 inflammasome activation 96. Consistent with these data we found that KH7 induced IL-1β secretion in LPS-primed J774A.1 murine macrophages; however, palmitate treatment markedly reduced IL-1β secretion in response to KH7 (Figure 2.3B). These data suggest that the reduction of intracellular camp precedes the interaction of NLRP3 with cardiolipin. We also found that palmitate treatment did not affect calcium flux upon challenge with the NLRP3 agonists silica or ATP (Figure 2.4). These data all suggest that interfering with the production of cardiolipin via palmitate treatment does not interfere with known upstream components of the NLRP3 activation pathway. Thus, these data indicate that cardiolipin is playing a downstream role in the activation of NLRP3.

47 31 Cardiolipin is sufficient for caspase-1 activation in a broken cell system Activation of an inflammasome leads to the autocatalytic cleavage of the 45 kd pro-caspase-1 to generate two subunits, p20 and p10. To assess whether the interaction between NLRP3 and cardiolipin could result in inflammasome activation, we utilized a broken cell system. Cells were primed with LPS and then disrupted by nitrogen cavitation. We found that cardiolipin-containing liposomes activated caspase-1, as determined by the appearance of the p10 cleavage produce, in a dose dependent manner (Figure 2.5A). Additionally, we found that non-cardiolipin containing liposomes had no effect on caspase-1 activation (Figure 2.5B). To confirm that the inhibition NLRP3 inflammasome activation associated with palmitate treatment was due to interference with cardiolipin synthesis, palmitate-treated and control-treated cells were LPS primed and subjected to the broken cell system. We found that addition of cardiolipin liposomes to broken palmitate-treated cells restored caspase-1 activation (Figure 2.6). Together, these data strongly suggest that cardiolipin may be sufficient for caspase-1 activation. As such, the ability of cardiolipin to interact with NLRP3 and directly activate NLRP3 may be biologically relevant. Discussion The data described in this chapter provide strong evidence that cardiolipin plays an integral role in the activation of the NLRP3 inflammasome. Additionally, these data support a hypothesis that cardiolipin serves as the direct activating signal for the NLRP3 inflammasome. The diverse molecules capable of activating the NLRP3 inflammasome are structurally and biologically unrelated, making it unlikely that they directly interact with NLRP3. Rather, it is more likely that these dissimilar agonists converge on a shared pathway leading to NLRP3 inflammasome activation. Steps in NLRP3 inflammasome

48 32 activation that are shared by all known activators have, until now, included potassium efflux, mitochondrial dysfunction, and the production of mitochondrial ROS 89,101,102,110. However, we have found evidence that the interaction between cardiolipin and NLRP3 is downstream of these events, and common to all tested NLRP3 inflammasome agonists. Given that cardiolipin is exclusively found in bacteria and on the inner mitochondrial membrane in eukaryotes, and that mitochondria are phylogenetically bacterial symbionts of early eukaryotic cells, it is attractive to postulate that cardiolipin functions as an endogenous PAMP that is revealed upon mitochondrial dysfunction and detected by NLRP3. Disruption of macrophage cardiolipin synthesis by palmitate exposure resulted in specific defects in NLRP3 (but not NLRC4 118 ) activation, suggesting that cardiolipin plays a critical and selective role in NLRP3 inflammasome activation. Depletion of cardiolipin via palmitate feeding can affect other aspects of cellular function, such as increasing cytochrome c release 223. However, our findings are consistent with the conclusion that it is cardiolipin depletion that is responsible for the defects observed in NLRP3 inflammasome activation. We found no alterations in a number of events known to be upstream of NLRP3 inflammasome activation, such as mitochondrial ROS production and calcium flux. Furthermore, we observed no difference in TNF-α production upon LPS stimulation, suggesting that cells fed palmitate were competent to produce cytokines. Additionally, palmitate treatment specifically impaired IL-1β release upon stimulation with NLRP3 agonists, but not the NLRC4 agonist P. aeruginosa, indicating that caspase-1 activation was not globally impaired 118. Finally, the addition of exogenous cardiolipin to a broken cell system was sufficient to induce caspase-1 activation. These data suggest that the observed defect in NLPR3 inflammasome activation in palmitate treated cells was likely due to perturbations in cardiolipin levels, rather than another cellular process. These data also suggest that cardiolipin is necessary for NLRP3 inflammasome activation and sufficient to drive caspase-1 activation in a broken cell system.

49 33 We have provided data that suggests cardiolipin is sufficient to activate caspase-1 in a broken cell system. We have determined that cardiolipin activates caspase-1 in a dose dependent manner, and that cardiolipin but not other phospholipids can activate caspase-1 in our broken cell system. We have additionally demonstrated that in our broken cells system the addition of cardiolipin to palmitate fed cells induced caspase-1 activation. All together these data suggest that cardiolipin is sufficient for caspase-1 activation in our broken cell system. However, it must be noted that this system does not necessarily reflect perfectly the state of the cell following challenge with an NLRP3 agonist. Cellular disruption by nitrogen cavitation may allow interactions to occur between proteins, and subcellular compartments, that would not come into contact even when the cell is stressed. Additionally, our cardiolipin containing liposomes do not recapitulate the full lipid profile of mitochondria, which may have enhanced the ability of NLRP3 to recognized cardiolipin. It would be useful for future work to determine what effect alterations in the lipid content of the cardiolipin containing liposomes have on caspase-1 activity. Additionally, it should be determined which cellular fractions are required for caspase-1 activation in this broken cell system. If the model presented in this chapter is accurate, and cardiolipin is sufficient for caspase-1 activation, one could predict that the addition of cardiolipin would render mitochondria dispensable. Ultimately, these observations should be tested in a true cell free system, where the inflammasome activation by cardiolipin could be recapitulated using purified proteins, because such an experiment would largely eliminate artifacts that may be induced by using cellular lysates. Cardiolipin is known to play an important role in the initiation of apoptosis. Apaf- 1 can form a large multiprotein complex called the apoptosome, which is responsible for the activation of caspase-9 in apoptosis. Parallels between the NLRP3 inflammasome and the apoptosome have been highlighted previously, when it was observed that low intracellular potassium is required for activation of the apoptosome, as it is for the

50 34 NLRP3 inflammasome 89,114. Additionally, interference with the mitochondrial membrane channel VDAC not only inhibits apoptosis, but also interferes with NLRP3 inflammasome-mediated IL-1β release 102. Cardiolipin is important in the activation pathways of both caspase-8 and ,224. Thus, it appears that inflammatory and apoptotic pathways converge on cardiolipin. It remains unclear what factors are required to induce NLRP3 inflammasome activation without additionally activating apoptosis. Nevertheless, it is recognized that apoptotic stimuli such as staurosporine can activate caspase-1, if cells are first primed with LPS 118. It is possible that a preceding inflammatory stimulus may be the point of divergence. Alternatively mitochondrial DNA has been suggested to play a role in NLRP3 inflammasome activation 101,219, through acting in concert with cardiolipin to determine if mitochondrial dysfunction leads to inflammation or cell death. Further study is required to identify this point of divergence. In summary, we have provided strong evidence that cardiolipin plays a central role in the activation pathway of the NLRP3 inflammasome. Data here suggests that cardiolipin is both necessary and sufficient to drive NLRP3 inflammasome activation. However, further study is required to determine the factors that regulate whether mitochondrial dysfunction leads to inflammation or apoptosis.

51 Figure 2.1: Cardiolipin depletion interferes with NLRP3 inflammasome activation. J774A.1 macrophages grown in the presence or absence of 0.5 mm palmitate were LPS primed, followed by stimulation with silica or ATP (5 mm). At 6 hours after stimulation, culture supernatants were collected and IL-1β and TNF-α release was measured by ELISA. Determinations were performed in triplicate and are expressed as the mean±sd Data shown are representative of three independent experiments. Reprinted from Immunity, 39(2), Iyer, S., et al. Mitochondrial cardiolipin is required for Nlrp3 inflammasome activation , Copyright (2013), with permission from Elsevier. 35

52 Figure 2.2: Palmitate treatment does not interfere with mitochondrial ROS production. J774A.1 macrophages were cultured in 0.5 mm palmitate for 16 hours. Macrophages were then LPS-primed, followed by stimulation with rotenone for one hour and staining with MitoSOX red. MitoSOX fluorescence was then measure by flow cytometry. Data shown are representative of three independent experiments. Reprinted from Immunity, 39(2), Iyer, S., et al. Mitochondrial cardiolipin is required for Nlrp3 inflammasome activation , Copyright (2013), with permission from Elsevier. 36

53 Figure 2.3: Palmitate treatment does not interfere with the role of camp in NLRP3 inflammasome activation. A. J774A.1 macrophages grown in the presence or absence of 0.5 mm palmitate were LPS-primed then treated with silica or ATP for 6 hours. Intracellular camp levels were assessed using an ELISA-based competitive immunoassay. B. J774A.1 macrophages grown in the presence or absence of 0.5 mm palmitate were LPS-primed and challenged with KH7 for 12 hours. Culture supernatants were collected and IL-1β release was measured by ELISA. Data are expressed as mean ± SD. Results are representative of three independent experiments. Reprinted from Immunity, 39(2), Iyer, S., et al. Mitochondrial cardiolipin is required for Nlrp3 inflammasome activation , Copyright (2013), with permission from Elsevier. 37

54 Figure 2.4: Palmitate treatment does not interfere with NLRP3 agonists induced calcium flux. J774A.1 macrophages grown in the presense or abscnece of 0.5 mm palmitate were LPS-primed, loaded with fura-2 AM, and then challenged with silica or ATP. The maximux elavations of intracellular free calcium ([Ca 2+ ] i ), represented by ΔF340/F380 are shown. Silica or ATP were added 1 minute after the start of recording calcium imaging. Imaging was collected over 10 minute periods for each condition. Stimulation in the absence (+ Ca 2+ ) or presence of 0.5 mm EGTA (- Ca 2+ ) are shown. Data are expressed as mean ± SD, and are representative of two independent experiments. Reprinted from Immunity, 39(2), Iyer, S., et al. Mitochondrial cardiolipin is required for Nlrp3 inflammasome activation , Copyright (2013), with permission from Elsevier. 38

55 Figure 2.5: Cardiolipin is sufficient to activate caspase-1 in a broken cell system. A. LPS-primed J774A.1 cells were lysed by nitrogen cavitation in the presence of liposomes containing cardiolipin (300, 100, or 30 µm) or phosphatidylcholine (300 µm), incubated for 1 hour, and analyzed for caspase-1 and GAPDH by immunoblot. B. LPS-primed J774A.1 cells were lysed by nitrogen cavitation in the presence of liposomes containing cardiolipin, phosphatidylserine, phosphatidic acid, or phosphatidylcholine (300 µm), incubated for 1 hour, and analyzed for caspase-1and GAPDH by immunoblot. Data shown are representative of at least three independent experiments. Reprinted from Immunity, 39(2), Iyer, S., et al. Mitochondrial cardiolipin is required for Nlrp3 inflammasome activation , Copyright (2013), with permission from Elsevier. 39

56 Figure 2.6: Cardiolipin liposomes activate caspase-1 in palmitate treated macrophages. J774A.1 macrophages grown in the presence or absence of 0.5 mm palmitate were LPS-primed and lysed by nitrogen cavitation in presense or absence of cardiolipin-containing liposomes (300 µm). Lysates were incubated for one hour. Caspase-1 and GAPDH were detected by immunoblot. Data shown are representative of three independent experiments. Reprinted from Immunity, 39(2), Iyer, S., et al. Mitochondrial cardiolipin is required for Nlrp3 inflammasome activation , Copyright (2013), with permission from Elsevier. 40

57 41 CHAPTER III IMMUNOGLOBULIN G-ANTIGEN IMMUNE COMPLEXES INHIBIT INFLAMMASOME ACTIVATION Abstract Immunoglobulin G (IgG) immune complexes can modify immune responses driven by antigen presenting cells in either a pro- or anti-inflammatory direction, depending upon the context of stimulation. However, the ability of immune complexes to modulate the inflammasome-dependent innate immune response is unknown. Here we show that IgG immune complexes suppress IL-1α and IL-1β secretion through inhibition of inflammasome activation. The mechanism by which this inhibition occurs is via immune complex ligation of activating Fcγ receptors (FcγR), resulting in prevention of both activation and assembly of the inflammasome complex in response to NLRP3, NLRC4, or AIM2 agonists. Our data suggest an unexpected mechanism where the generation of antigen-specific humoral immune responses may suppress the early inflammatory response. Introduction IgG-antigen immune complexes have numerous effects on the host immune system driven by their ability to act on Fcγ receptors (FcγR). Signals from FcγR can be either activating or inhibitory in nature. These signals regulate the activation of the innate immune system, leading to enhanced phagocytosis, the induction of antibody dependent cellular cytotoxicity, degranulation, and profound modifications in cytokine production by macrophages, dendritic cells, and monocytes Immune complexes have been shown to upregulate IL-10 production while concurrently suppressing IL-12 p40 generation 206,207,209,211,225. These changes to innate immune cytokine production have

58 42 a profound influence on the generation of an antigen specific adaptive response and have been shown to repress Th1 responses while enhancing Th2 responses 209,210. IL-1α and IL-1β are closely related cytokines that signal through the IL-1R1. As they are potent pro-inflammatory cytokines, the secretion of each is tightly regulated. IL- 1β secretion is dependent upon the activation of caspase-1, which is activated as part of a multi-protein complex called the inflammasome 76,226,227. In contrast, IL-1α processing and secretion can proceed in either an inflammasome-dependent or independent manner, depending on the particular agonist 228. Upon activation, the NLR family members (NLRP1, NLRP3, NLRC4) and one PYHIN family member, AIM2, forms an inflammasome complex. These complexes typically contain an NLR (or AIM2), the adapter protein ASC, and the cysteine protease caspase-1. Activation of the inflammasome follows a two-step process. The first signal, priming, occurs in response to either microbial or endogenous danger signals 50,51. Priming results in the generation of pro-il-1α, pro-il-1β and pro-il-18. In addition, it also readies the inflammasome for activation through an unknown mechanism. Numerous stimuli can serve as signal two, which leads to the activation of specific inflammasome complexes, and ultimately caspase-1 activation. In this study we demonstrate that the ligation of FcγR by IgG immune complexes during priming inhibits the assembly and activation of the inflammasome complex. This in turn markedly diminishes production of IL-1α and IL-1β. Further, we demonstrate that IgG immune complexes suppress in vivo production of IL-1β. Materials and Methods Mice C57BL/6 mice were obtained from the NCI mouse repository (Frederick, MD). Fcer1g -/- and Fcgr2b -/- mice have been described previously 186,229. Map1lc3b -/- mice were

59 43 obtained from The Jackson Laboratory (Bar Harbor, ME). The Institutional Animal Use and Care Committees at the University of Iowa approved all protocols. Reagents DPBS and DMEM were purchased from Mediatech Inc. Rabbit anti-sheep red blood cell antibodies were purchased from Rockland. Rabbit anti-ova IgG was from MP Cappel. Rabbit anti-c. albicans polyclonal IgG, and Imject alum were from Thermo Scientific. Chicken egg ovalbumin (Grade V) and ATP were from Sigma. E. coli LPS (serotype 0111:B4) was from Invivogen. Staphylococcus aureus lipoteichoic acid (LTA) was purchased from Sigma-Aldrich. Silica (Min-U-Sil-5) was purchased from Pennsylvania Sand and Glass Corporation. ELISA antibody pairs for IL-1β were from R&D Systems. ELISA antibody pairs for IL-1α, IL-10 and IL-12p40 were from ebiosciences. IL-18 ELISA antibody pairs were from MBL. For serum IL-1β quantification an R&D systems Quantikine ELISA kit was used per the manufacturers instructions. Antibodies for immunoblot detection of caspase-1, IL-1β, ASC, NLRP3, and GAPDH, were performed with rabbit polyclonal anti-mouse casapse-1 p10 antibodies (Santa Cruz), rabbit polyclonal anti-mouse IL-1β antibodies (Millipore), mouse monoclonal anti-mouse NLRP3 antibodies (Enzo Life Sciences), and mouse monoclonal anti-mouse GAPDH antibodies (Calbiochem), respectively. Rhodamine phalloidin was from Molecular Probes, and VectaSheild with DAPI was purchased from Vector Laboratories. FITC conjugated donkey anti-rabbit IgG was from BioLegend. Immune complexes IgG opsonized sheep erythrocytes were produced as described previously 207. Briefly, stock sheep erythrocytes were diluted 1:10 with DPBS and anti-sheep red blood cell antibodies were added (final concentration: 400 µg/ml, unless otherwise noted), incubated at 4 C for 40 minutes with agitation, and washed 3x with DPBS. For IgG-Ova immune complexes, chicken egg ovalbumin (Ova) and rabbit anti-ova IgG were mixed at

60 44 a 1:32 (µg Ova: µg IgG) ratio, and incubated for 30 minutes at room temperature with agitation. To produce IgG-opsonized C. albicans, log growth phase C. albicans (3-5 x 10 7 yeast/ml) were incubated with 0.5 mg/ml rabbit anti-c. albicans polyclonal IgG for 40 minutes at 4 C. The IgG-opsonized C. albicans was washed and resuspended in DPBS. In vitro stimulation of bone marrow-derived macrophages Bone marrow-derived macrophages (BMDM) were generated as described previously 206. Briefly, bone marrow cells were grown for 6-10 days in DMEM supplemented with 10% FBS, penicillin (100 U/mL), streptomycin (100 µg/ml), L- glutamine (2 mm), and 20% L292 cell conditioned media, at 37 C and 5% CO 2. BMDM were either left unstimulated, primed with LPS (50 ng/ml), LPS and immune complexes, or LPS and particle control for 3-4 hours. BMDM were then challenged with ATP (5 mm), silica (50 µg/cm 2 ), C. albicans FC20 strain (MOI 10:1), P. aeruginosa PAK strain (MOI 1:1) for 6 hours, or F. tularensis LVS strain (MOI 50:1) for 9 hours. For ATP stimulation the media was changed at 20 minutes post-stimulation. For experiments where LTA was used for priming, the macrophages were treated as described above, but were primed for 4 hours with 50 µg/ml LTA. Immunoblotting Electrophoresis of lysates was performed using the NuPAGE system (Invitrogen) according to the manufactures protocol. Briefly, cells were lysed with 1% (v/v) Triton X- 100 in Tris-buffered saline with EDTA (5 mm). Lysates were prepared with LDS sample loading buffer (Invitrogen) and sample reducing buffer (Invitrogen). Proteins were resolved on 4-12% Bis-Tris polyacrylimide gels (Invitrogen), and transferred to polyvinylidene difluoride (PVDF) membranes by wet electrophoresis. Membranes were blocked with 5% milk, and incubated in primary antibody overnight at 4 C. This was followed by treatment with an HRP-conjugated secondary antibody for two hours at 4 C.

61 45 Proteins were detected by enhanced chemiluminescence, and imaged using the LI-COR Odyssey (LI-COR Biosciences) imaging system. ASC Speck Assay Immunofluorescence studies were carried out as previously described 230. Briefly, macrophages were seeded on glass coverslips and challenged as described above. The cells were washed and fixed in 4% (v/v) paraformaldehyde in PBS. The cells were then blocked and permeabilized in 5% FBS in PBS with 0.5% (w/v) saponin. Cells were immunostained with rabbit anti-asc (Enzo Life Sciences) as the primary antibody and FITC conjugated donkey anti-rabbit IgG. Actin was stained with rhodamine phalloidin according to the manufacturer s protocol. Coverslips were mounted on slides using VectaSheild with DAPI and imaged using a confocal microscope (Zeiss 710, Carl Zeiss, Inc.). Statistical Analysis Statistical analysis was done using Prism 5.0a (GraphPad Software). Unless otherwise noted, statistical significance for single comparisons was determined by Student s t-test; ANOVA with a Bonferroni post-test was used for multiple comparisons. Results Immune complexes inhibit the secretion of IL-1α, IL-1β, and IL-18 in vitro. Cytokines secreted by antigen presenting cells are required to instruct the differentiation of CD + T cells and can help direct the tissue healing response. To assess if these cytokines were affected by immune complex uptake, bone marrow-derived macrophages (BMDM) were primed with LPS in the presence of either IgG-opsonized sheep red blood cells (SRBC; eigg) or unopsonized SRBC (e). The cells were then challenged with the NLRP3 inflammasome agonist silica. As expected, stimulation of

62 46 LPS-primed BMDM with silica resulted in the secretion of IL-1β and IL-18. However, the presence of eigg immune complexes during LPS priming strongly inhibited both IL- 1β and IL-18 secretion (Figure 3.1A). Consistent with previous studies, immune complexes suppressed LPS-induced IL-12p40 production while concurrently elevating IL-10 production (Figure 3.1A). This inhibition was not specific to the type of immune complex or the NLRP3 agonist, as soluble IgG-Ova immune complexes were also capable of inhibiting IL-1β secretion induced by the soluble NLRP3 agonist ATP, or the crystalline NLRP3 agonist alum (Figure 3.1B and C). Immune complexes inhibit the activation of caspase-1 Caspase-1 activation involves the autocatalytic processing of the 45 kda procaspase-1 to generate two subunits, p20 and p10. Caspase-1 activation in silica and ATP stimulated LPS-primed BMDM was detected in immunoblots by the appearance of the p10 cleavage product (Figure 3.2A). However, if BMDM were LPS-primed in the presence of eigg immune complexes, caspase-1 activation was not observed in response to silica or ATP challenge (Figure 3.2A). Unlike IL-1β, cleavage and secretion of IL-1α can be either inflammasome-dependent or independent, depending on the NLR agonist 228. In general, crystalline activators of the NLRP3 inflammasome, like alum, induce caspase-1-independent IL-1α secretion, whereas soluble activators, such as ATP, induce caspase-1-dependent IL-1α secretion. Consistent with the inhibition of caspase-1 by IgG immune complexes, we observed a suppression of ATP-induced, but not aluminduced, IL-1α secretion (Figure 3.2B and C). Taken together, these data suggest that immune complex mediated blockage of IL-1β secretion and ATP-induced IL-1α secretion occurs via the inhibition caspase-1 activation.

63 47 Immune complexes do not prevent the synthesis of pro-il- 1β, NLRP3, ASC or caspase-1 We next asked at what step immune complexes acted to interfere with inflammasome activation. Inhibition of IL-1β secretion occurred when immune complexes were added up to three hours following the addition of LPS, but not if the immune complexes were added concurrently with the NLRP3 agonists. This suggested that immune complexes interfered with the priming signal required for inflammasome activation (Figure 3.3A). Additionally we found that immune complex-mediated inhibition of IL-1β secretion was dependent upon the number of immune complexes used as well as the concentration of antibody used to opsonize the target (Figure 3.3B and C). We next asked if the TLR agonist used played a role in the observed suppression. We found that IgG-immune complexes suppressed IL-1β secretion from lipoteichoic acid (LTA), which activates the TLR2/6 heterodimer 231, primed BMDMs (Figure 3.3D). These data strongly suggest that immune complexes suppress inflammasome activation by interfering with the required priming signal. The precise mechanism by which priming readies the inflammasome for activation is unknown; however it has been suggested that upregulation of NLRP3 expression is one factor in the priming process 49. To determine if immune complexes blocked the expression of NLRP3, or other inflammasome components, BMDM were LPS-primed in the presence or absence of IgG immune complexes for 4 hours and subjected to immunoblot. Similar amounts of pro-caspase-1, pro-il-1β, NLRP3 and ASC were detected in the absence or presence of immune complexes (Figure 3.4), thus suggesting that inhibition of inflammasome complex component expression is not responsible for immune complex-mediated caspase-1 inhibition.

64 48 Inflammasome inhibition by IgG immune complexes requires signaling through the FcRγ chain FcγRs can be either activating (FcγRI, FcγRIII, and FcγRIV) or inhibitory (FcγRIIb). The activating receptors require the ITAM containing FcRγ chain (encoded by the Fcer1g gene) for signal propagation 186, whereas the inhibitory FcγRIIb signals through a cytosolic ITIM motif 205. To determine the contribution of activating and inhibitory FcγR in immune complex-mediated inflammasome inhibition, we utilized BMDM derived from FcRγ chain-deficient mice (Fcer1g -/- ) or FcγRIIb-deficient (Fcgr2b -/- ) mice. Immune complexes failed to suppress the secretion of IL-1β secretion in response to silica or ATP in BMDM from Fcer1g -/- mice (Figure 3.5A). In contrast, immune complexes suppressed IL-1β secretion from BMDM derived from Fcrg2b -/- mice (Figure 3.5B). These data suggest that inhibition of inflammasome activation by immune complexes is not mediated through the inhibitory receptor FcγRIIb. Rather, this indicates that signaling through the FcRγ chain is required, implicating an activating FcγR in the process. Immune complexes require uptake to inhibit inflammasome activation To elucidate the mechanism behind immune complex-mediated inhibition of inflammasome activation, we next asked if uptake of the immune complexes was required for inflammasome inhibition, or if signaling through the FcRγ is sufficient to inhibit inflammasome activation. To address this question we LPS-primed BMDM that had been plated in tissue culture dishes coated with either IgG or Ova as a control. Strikingly, the BMDM that had been plated on IgG coated tissue culture plates showed no difference in IL-1β secretion upon silica challenge when compared to BMDM alone (Figure 3.6). Enhancement in IL-10 secretion and suppressed IL-12p40 production

65 49 confirmed that FcγR had been ligated by the plate bound IgG (Figure 3.6). Together these data suggest that immune complex uptake is required for inflammasome inhibition. Autophagy has been proposed to be a critical regulator of inflammasome activation 101,104. Additionally, FcγR recruits autophagic proteins to the nascent phagosomes 232. Given this information, we hypothesized that uptake of immune complexes would induce autophagy, thus suppressing inflammasome activation. To test this hypothesis we utilized BMDM derived from mice deficient in the autophagy protein LC3 (Map1lc3b -/- ). We found that immune complexes suppressed IL-1β in Map1lc3b -/- BMDM that had been challenged with ATP or silica (Figure 3.7A and B). These data demonstrate that autophagy is not involved in inflammasome inhibition by immune complexes. Immune complexes inhibit activation of NLRP3, NLRC4, and AIM2 inflammasomes To determine if inflammasome inhibition by immune complexes is specific to the NLRP3 inflammasome or affects other inflammasomes as well, we utilized three pathogens, Candida albicans, Pseudomonas aeruginosa, and Francisella tularensis LVS strain, which activate the NLRP3, NLRC4, and AIM2 inflammasomes, respectively. Consistent with our observations with other NLRP3 agonists, we found a marked inhibition of IL-1β in response to C. albicans in BMDM that had been LPS-primed in the presence of immune complexes (Figure 3.8A). Direct IgG opsonization of C. albicans also resulted in significantly reduced IL-1β secretion from bone-marrow derived dendritic cells (BMDC) as compared to challenge with unopsonized C. albicans (Figure 3.8B). Systemic infection of mice with C. albicans in the presence of IgG-Ova immune complexes also resulted in diminished serum IL-1β levels compared to mice challenged with C. albicans in the presence of Ova alone (Figure 3.8C).

66 50 IgG immune complexes were also capable of effectively suppressing IL-1β secretion from BMDM challenged with P. aeruginosa or F. tularensis LVS, which activate the NLRC4 and AIM2 inflammasomes respectively (Figure 3.9 A and B). Together these results demonstrate that immune complexes inhibit caspase-1 activation induced by multiple different inflammasomes. These data additionally indicate that this inhibition occurs at a point discrete from the individual NLR or AIM2 receptors, yet common to each inflammasome. Immune complexes block oligomerization of inflammasomes To elucidate the mechanism by which immune complexes inhibit inflammasome activation, we next asked what effect they had on inflammasome oligomerization upon receiving an activating signal. Upon activation of the inflammasome, the adapter protein ASC forms cytosolic aggregates, or specks, that co-localize with caspase-1 and their initiating PRR. These foci can be visualized using confocal microscopy 230,233. We therefore tested if IgG immune complexes were able to block the formation of ASC specks. BMDM were challenged with ATP following LPS priming in the presence or absence of immune complexes. We observed the formation of ASC specks upon ATP challenge in BMDM that had been LPS primed in the absence of immune complexes (Figure 3.10A and B). However, ATP challenged BMDM that had been primed in the presence of immune complexes had a significant reduction in ASC specks (Figure 3.10A and B). Similar to the NLRP3 inflammasome agonist ATP, we found that BMDM challenged with the NLRC4 agonist P. aeruginosa showed a significant reduction in ASC speck formation upon priming in the presence of immune complexes (Figure 3.10C). These data indicate that immune complexes block caspase-1 activation and the subsequent secretion of IL-1β by preventing inflammasome assembly.

67 51 Discussion The data described in this chapter demonstrate that immune complexes are able to potently suppress caspase-1 activation and the subsequent processing and secretion of IL- 1β, both in vitro and in vivo. Furthermore, we found that this inhibition was not specific, but rather affected all inflammasomes tested. Finally, our data suggest that immune complexes prevent inflammasome oligomerization. Swift and robust innate immune responses are required for the control of microbial pathogens, but a continued or disproportionate innate response can cause collateral tissue damage and lead to autoimmunity. Considering the strong proinflammatory activity of IL-1α and IL-1β, their processing and secretion must be tightly controlled. In this chapter, we have described a novel pathway by which inflammasome activation and assembly in macrophages can be modulated, namely via the internalization of IgG immune complexes. While previous studies have shown that immune complexes can modify innate immune responses through regulating the release of IL-10 and IL-12, their impact on inflammasome activation has not been determined. Herein, we present a novel function of immune complexes, wherein their presence inhibits inflammasome activation. Previous studies suggest that the modulation of IL-10 and IL-12p40 by IgG immune complexes occurs at the level of transcription 206. Interestingly, immune complexes do not appear to suppress inflammasome activation in a similar manner. Rather, they appear to act a post-translational level by preventing inflammasome assembly and caspase-1 activation. Upon exposure to activating stimuli ASC will form cytosolic aggregates, or specks, that colocalize with caspase-1, pro-il-1β and NLRP3 230,234,235. These aggregates likely represent the active inflammasome complex, which is thought to be multimer of the NLR, ASC, and caspase-1 heterotrimeric complex 235 The frequency of ASC specks we observed is similar to previously published results 230, however it is unlikely that the ASC speck positive cells we observed are the

68 52 entirety of the population of cells that are speck positive upon inflammasome activation. Active caspse-1 p10 can be detected within 30 minutes of ATP stimulation by immunoblot 236, therefore it is possible that a larger proportion of macrophages are speck positive that what we observed at our chosen time point. A recent study by Misawa et al. has provided evidence that assembly of the NLRP3 inflammasome complex and localization to the mitochondria depends upon dynein-mediated transport, which is acetylated α-tubulin dependent 221. Perhaps FcγR signaling interferes with the production of acetylated α-tubulin, thus preventing inflammasome assembly and activation. However, such a mechanism would likely only apply to NLRP3 as these dynein-mediated rearrangements do not appear to be involved in NLRC4 or AIM2 inflammasome activation 221. Nevertheless, it would be interesting to fully characterize the dynamics of inflammasome assembly following inflammasome activation, perhaps using fluorescently labeled ASC or NLRP3 to follow ASC speck formation using live-cell time-lapse fluorescent microscopy. Such studies may help identify the mechanism by which immune complexes block inflammasome activation. Data presented in this chapter suggests that the inhibition of inflammasome activation by immune complexes is not specific to the NLRP3 inflammasome, but also applies to both the NLRC4 and AIM2 inflammasomes, as the presence of immune complexes inhibited IL-1β release induced by both P. aeruginosa and F. tularensis. Therefore, this novel inhibition of inflammasome activation by immune complexes appears to apply to multiple types of inflammasomes. Given the differences between the activating signals of the inflammasomes tested, it seems unlikely that immune complexes would be acting via distinct mechanisms on each. Rather it seems more likely that immune complexes would inhibit inflammasome activation through a step common to all inflammasomes, such as priming. Importantly, we show that the initial priming steps occur in cells treated with immune complexes, as determined by the upregulation of inflammasome components. However, the degree of inflammasome inhibition is

69 53 dependent upon the timing of the addition of immune complexes, as immune complexes inhibit inflammasome activation only if added prior to the activating signal. Furthermore, we have provided data that suggests immune complexes will suppress inflammasome activation in response to non-lps priming, as IgG immune complexes suppressed IL-1β secretion from BMDM primed with LTA. Thus the inhibition of inflammasome activation occurs following the initial priming signal, but prior to the assembly of the inflammasome. Recent studies have provided evidence that inflammasome priming requires more that simple enhanced synthesis of inflammasome components and pro-il- 1β 52-54,236. A recent study by Ghonime et al. 236 suggested that TLR4 activation of ERK1 is required for non-transcriptional priming of the NLRP3 inflammasome. This requirement for ERK in inflammasome priming suggests a tantalizing link between priming and FcγR signaling, as ERK is downstream of FcγR 237. Perhaps FcγR ligation leads to sequestration ERK, or an upstream component of this cascade, thus suppressing its activity during priming. Sequestration of an upstream component of the ERK activating cascade, such as Syk, could help explain that apparent requirement of IgG immune complexes to be internalized to inhibit inflammasome activation. Additional work is required to determine what if any role ERK plays in immune complex mediated inflammasome suppression. Future studies should focus on determining the precise mechanism by which immune complexes inhibit inflammasome activation. Elucidating this mechanism could provide novel insight in to other steps necessary to prime the inflammasome for activation. FcγRs can be either activating or inhibitory. All of the activating FcγRs signal through the common FcRγ. We show a mechanism by which immune complexes abrogate inflammasome activation via activating FcγRs, as FcRγ-deficient BMDMs show intact IL-1β secretion when treated with immune complexes. NLRP3 inflammasome activation by Schistosoma mansoni and C. albicans requires the CLR dectin-2 86,91,152, which couples with the FcRγ chain for signal propogation 28. Hence, signals from the

70 54 FcRγ chain can be either activating or inhibitory for the NLRP3 inflammasome depending on the upstream receptor initiating the signal. It is possible that this divergence is due to the inflammasome agonist (i.e. C. albicans) providing the signal that initiates signaling through the FcRγ chain. However this seems unlikely considering that IgG opsonized C. albicans induced less IL-1β secretion from BMDCs than C. albicans not complexed with IgG. It would be interesting to determine what causes signaling through dectin-2 to induce inflammasome activation, while signaling through FcγRs to inhibit inflammasome activity. The presence of immune complexes is a marker for the successful development of an adaptive immune response, thus the ability of immune complexes to shut off inflammasomes represents a negative feedback loop. Another example of adaptive immune downregulation of innate responses is seen by the ablation by CD4 + T effector and memory cells of inflammasome activation, which is in a CD40/CD40L dependent manner 173. Together, our data suggest that the presence of an effective adaptive immune response through antigen-antibody immune complexes acts as a negative feedback loop, shutting off inflammasome activation and providing the signal to terminate the inflammatory response.

71 Figure 3.1: IgG immune complexes suppress NLRP3 inflammasome activation. A. BMDM were LPS primed in the presence of unopsonized (e) or IgG-opsonized erythrocytes (eigg) for 4 hours and then challenged with silica for 6 hours. Cytokine secretion into the culture supernatant was measured by ELISA. B and C. BMDM were LPS primed in the presence or absence of Ova or IgG-Ova, and then challenged with either B. ATP or C. alum for 6 hours. Culture supernatants were collected and analyzed for IL-1β by ELISA. Determinations were performed in triplicate and are expressed as mean±sd. Results shown are representative of at least three independent experiments. 55

72 Figure 3.2: Immune complexes block caspase-1 activation. A. BMDM were LPS primed in the presence or absence of eigg. Macrophages were then challenged with either silica or ATP. 6 hours later cell lysates were collected and analyzed for caspase-1 activation by immunoblot. GAPDH serves as the loading control. B and C. BMDM were LPS primed with or without Ova or IgG-Ova. Cells were then challenged with B. ATP or C. alum. Culture supernatants were harvested and analyzed for IL-1α by ELISA. Determinations were performed in triplicate and are expressed as the mean±sd. Results shown are representative of at least three independent experiments. 56

73 Figure 3.3: Immune complexes interfere with inflammasome priming. A. BMDM were LPS primed and eigg was added at the indicated time points after LPS treatment. Four hours after the addition of LPS, BMDM were then challenged with silica for 6 hours. Supernatants were harvested and IL-1β was measured by ELISA. B. BMDM were LPS primed in the presence of increasing numbers of eigg (eigg:bmdm ratio of 10:1, 5:1, 2.5:1, and 1:1). BMDM were challenged with silica for 6 hours and culture supernatants were analyzed for IL-1β by ELISA. C. BMDM were LPS primed in the presence of erythrocytes that were IgG opsonized with decreasing concentrations of anti- SRBC IgG (400, 200, 100, and 50 µg/ml). BMDM were then challenged with silica for 6 hours and culture supernatants were analyzed for IL-1β by ELISA. D. BMDM were primed with LTA (50 µg/ml) in the presence or absence of Ova or IgG-Ova. Macrophages were then challenged with silica for 6 hours, and culture supernatants were analyzed for IL-1β by ELISA. Determinations were performed in triplicate and are expressed as mean±sd. Data shown are representative of at least three independent experiments. 57

74 Figure 3.4: IgG immune complexes do not affect the synthesis of inflammasome components. BMDM were LPS primed with or without Ova or IgG-Ova. After 4 hours lysates were collected and analyzed by immunoblot for pro-caspase-1, pro-il-1β, NLRP3, ASC, and GAPDH. GAPDH served as the loading control. Data shown is representative of three independent experiments. 58

75 Figure 3.5: Signaling through the FcRγ chain but not FcγRIIb is required for inflammasome suppression. BMDM from A. Fcer1g -/- or B.Fcgr2b -/- were LPS primed with or without eigg and challenged with silica or ATP for 6 hours. Culture supernatants were harvested and analyzed for IL-1β by ELISA. *p 0.05, ***p by Student s T- test. Determinations were performed in triplicate and are expressed as the mean±sd. Results shown are representative of at least three independent experiments. 59

76 Figure 3.6: Immune complex uptake is required for inflammasome inhibition. BMDM were plated onto tissue culture plates that had been coated with IgG (2 µg/ml) or Ova (2 µg/ml), and then LPS primed. Macrophages were then challenged with silica for 6 hours. Culture supernatants were collected and analyzed for IL-1β, IL-10, and IL-12p40 by ELISA. ***p<0.001 by Student s T-test. Determinations were performed in triplicate and data is expressed as mean±sd. Data shown are representative of three independent experiments. 60

77 Figure 3.7: Immune complex mediated inflammasome suppression is not via autophagy. LC3-deficient (Map1lc3b -/- ) BMDM were LPS primed in the presence or absence of eigg, and then challenged with either A. ATP or B. silica for 6 hours. Culture supernatants were collected and analyzed by ELISA for IL-1β. ***p by Student s t-test. Determinations were performed in triplicate, and are shown as mean±sd. Data shown are representative of three independent experiments. 61

78 Figure 3.8: Immune complexes suppress NLRP3 inflammasome activation in response to microbial stimuli in vitro and in vivo. A. BMDM were LPS primed with or without e or eigg and then challenged with C. albicans (MOI 10:1) for 6 hours. Culture supernatnats were collected and IL-1β was measured by ELISA. B. BMDC were challenged with either unopsonized or IgG-opsonized C. albicans (MOI 10:1) for 6 hours. Culture supernatants were collected and anlyzed for IL-1β secretion by ELISA. **p 0.01 by Student s t-test. C. WT mice were infected i.v. with 2 x 10 6 C. albicans yeast, resuspended in the presence of Ova or IgG-Ova. After 6 hours serum IL-1β levels were determined by ELISA. *p 0.05 by Mann-Whitney test (N=14 mice per group). Determinations were performed in triplicate and are expressed as the mean±sd (A and B) or mean±sem (C). Data shown are representative of at least three independent experiments (A and B) or are pooled from four independent experiments (C). 62

79 Figure 3.9: Immune complexes inhibit the NLRC4 and AIM2 inflammasomes. BMDM were LPS primed with or without e or eigg and then challenged with either A. P.aeruginosa PAK strain (MOI 1:1) or B. F. tularensis LVS strain (MOI 50:1) for 6 and 9 hours respectively. IL-1β in supernatants was measured by ELISA. Determinations were performed in triplicate and are expressed as the mean±sd. Results shown are representative of at least three independent experiments. 63

80 Figure 3.10: Immune complexes block inflammasome oligomerization. BMDM were LPS primed in the presence or absence of IgG-Ova, and were then challenged with ATP for 1 hour (A and B) or P. aeruginosa (MOI 1:1) for 6 hours; cells were fixed, permiablized and stained for ASC. A. Representative confocal images are shown. Blue is DAPI (DNA), red is rhodamine phalloidin (F-actin), and green is ASC. Number of speck positive cells was quantified (B and C). ***p by 1-way ANOVA with Bonferroni post-test. Determinations were performed in triplicate and are expressed as the mean±sd. Results shown are representative of at least three independent experiments. 64

81 65 CHAPTER IV IMMUNOGLOBULIN G-ANTIGEN IMMUNE COMPLEXES BLOCK THE DEVELOPMENT OF T H 2 AND T H 17 CD4 + T CELL RESPONSES Abstract The precise context in which the innate immune system is activated, and the cytokines it subsequently produces, play a central role in determining the type of CD4 + T helper (Th) cell response. Antigen-IgG immune complexes are known to be highly immunomodulatory. Th1 responses are downregulated when antigen is encountered in the context of IgG immune complexes. However, the effect of immune complexes on the generation of Th17 responses is currently unknown. To assess if Th17 responses to antigen are similarly influenced by the presence of immune complexes, we utilized an allergic airway disease model. In this model, mice immunized with the adjuvant alum (aluminum hydroxide salts) and Ova induce a potent Ova-specific Th2 and Th17 response. Here we show that substituting IgG-Ova for Ova results in the suppression of both Th2 and Th17 responses. In agreement with previous studies, we found that IL-1R signaling is required for Th2 and Th17 responses. Further, we determined that immune complex mediated suppression of Th2 and Th17 responses was due to suppression of IL- 1α and IL-1β secretion. The data presented in this chapter suggest a novel mechanism whereby the generation of an antigen specific humoral immune response may alter subsequent cellular immune responses. Introduction Antigen-specific CD4 + T cell responses are shaped by instruction from the innate immune system. The cytokine signals that will ultimately shape the CD4 + T cell response reflect the environment in which the antigen presenting cell (APC) initially encounters antigen. This includes whether the antigen is associated with specific adjuvants (e.g. in a

82 66 vaccine), PAMPs or DAMPs (e.g. during infection), and if the antigen is complexed with antibodies. Differentiation of naïve CD4 + T cells down specific effector pathways, including T helper (Th) 1, Th2, or Th17 lineages, requires the presentation of processed antigen in the presence of costimulatory molecules, with precise combinations of cytokines. Circulating immune complexes are associated with both the initiation and the progression of many diseases, in particular autoimmune disorders. Immune complexes are highly immunomodulatory, and can regulate both the innate and adaptive immune responses. Additionally, the immune complexes can either enhance or repress inflammatory responses 179,195,197,206,207. Ligation of activating FcγRs on macrophages by IgG-immune complexes results in suppressed IL-12p40 production 206, which plays a critical role in Th1 differentiation. Previous studies have demonstrated that antigen-igg immune complexes are capable of effectively suppressing Th1 responses 209,210, partially through changes in the production of IL-12p40 and IL-10. However, the effect of IgGimmune complexes on the generation of Th17 responses is currently unknown. Th17 cells are defined by their production and secretion of the proinflammatory cytokines IL-17A, IL-17F, and IL , and are reliant on the master transcription factor RORγt 239. Functionally, Th17 responses play an integral role in host defense against numerous microbes 30,143,151,240. Nevertheless, they also drive pathologic responses that can underlie autoimmune disorders 154,156. The cytokines IL-1, IL-6, IL-23, and TGF-β are important in the generation of these pathological Th17 responses 142,238,241. Signaling through the IL-1R on T cells is critical for the induction of Th17 responses, as mice deficient in IL-1R1 had defective Th17 responses in an experimental autoimmune encephalomyelitis (EAE) model. This defect was accompanied by reduced disease severity 142,241. The closely related cytokines IL-1α and IL-1β both signal through the IL-1R1, yet their individual contributions to the generation of Th17 responses remain unclear. Even

83 67 though IL-1α and IL-1β share a common downstream receptor, their processing and secretion proceeds via distinct mechanisms. IL-1β secretion is dependent upon processing by caspase-1, while IL-1α processing and secretion may be inflammasome-dependent or independent, depending on the agonist 110,228. While IL-1β requires processing to be biologically active, IL-1α can be released passively and does not require cleavage to interact with and activate signaling through IL-1R Based on the results generated in the preceding chapter, which demonstrated that IgG-immune complexes effectively limit IL-1α and IL-1β secretion, we hypothesized that immunization with immune complexes would impair the development of Th17 responses. To address this hypothesis we utilized a model of allergic airway disease. Here we demonstrate that immunization with adjuvant alum and immune complexes strongly impairs the development of both Th2 and Th17 responses. Additionally, we show that in vivo alum driven adaptive immune responses depend on the presence of both IL-1α and IL-1β. Finally, we show that immune complex mediated suppression of IL-1α and Il-1β results in the inhibition of effector CD4 + T cell responses. Materials and Methods Mice C57BL/6 and CD45.1 (B6Ly5.2Cr) mice were obtained from the NCI mouse repository (Frederick, MD). The generation of Nlrp3 -/-, Asc -/-, Casp1 -/-, Il10 -/-, Il1r1 -/-, Il1a -/-, and Il1b -/- mice have been described previously 218, OT-II (B6.Cg- Tg(TcraTcrb)425Cbn/J) transgenic mice 247 were purchased from Jackson Laboratories (Bar Harbor, ME). The Institutional Animal Use and Care Committees at the University of Iowa approved all protocols.

84 68 Reagents DMEM, RPMI 1640, and DBPS were purchased from Mediatech Inc. Chicken egg ovalbumin (Ova) (Grade V) and ATP was purchased from Sigma. Rabbit anti-ova IgG was from MP Cappel. Rabbit anti-sheep red blood cell antibodies were from Rockland. Imject Alum was purchased from Thermo Scientific. Silica (Min-U-Sil-5) was purchased from Pennsylvania Sand and Glass Corp. Recombinant IL-1α and IL-1β were purchased from R&D Systems. ELISA antibody pairs for IL-17A, IL-13, and IL-4 were purchased from ebiosciences, and antibody pairs for IL-1β were from R&D Systems. For flow cytometry brefeldin A, monensin, FACS Fixation/Permiabilization buffer, and anti- CD3, -CD4, and IL-13 were from ebiosciences, and anti-il-17a was from BD Biosciences. CD4 + T cell Positive Selection MACS kit was purchased from Miltenyi. CFSE was purchased from Invitrogen. E. coli LPS (serotype 0111:B4) was purchased from Invivogen. Carboxyfluorescein succinimidyl ester (CFSE) was from Invitrogen. HEMA3 stain was purchased from Fisher Scientific. Immune complexes IgG opsonized sheep erythrocytes and IgG-Ova immune complexes were produced as described in Chapter 2. In vitro stimulation of BMDM BMDM were produced, and stimulated in vitro, as described in Chapter 2. Induction and evaluation of airway inflammation Mice were sensitized on day 0 by intraperitoneal injection with either 2 mg alum and 20 µg Ova or 2 mg alum and IgG-Ova (20 µg Ova). On days 15, 16, and 17 mice were intranasally challenged with 20 µg Ova in 50 µl PBS. Lymph nodes, lungs, blood, and BAL fluid were harvested on day 18. Bronchoalveolar lavage (BAL) was performed by delivering 1 ml cold DPBS into the airway via tracheal cannula and gently aspirating

85 69 the fluid. The lavage was repeated three times. The cells were stained with trypan blue to determine viability, and total nucleated cell counts were obtained using a hemocytometer. Cytospin slides were prepared, and percentage of neutrophils, eosinophils, lymphocytes, and DC/Macs was determined after HEMA3 staining. Ova specific IgG1 and IgG2c and total IgE in serum was determined by ELISA at day 19 as previously described 248,249. Lungs were fixed, embedded in paraffin and 5 µm sections were stained with hematoxylin and eosin (H&E). T cell restimulation and proliferation Cells from the draining lymph nodes (LNs) were cultured in RPMI 1640 supplemented with 10% FBS, penicillin (100 U/mL), streptomycin (100 µg/ml), L- glutamine (2 mm), and β-mercaptoethanol with 10 µm Ova for 72 hours. Supernatants were collected and analyzed via ELISA for IL-17A, IL-13, and IL-4. For flow cytometric analysis of intracellular cytokines, Ova-restimulated LN cells were incubated for 4 hours in the presence of brefeldin A (3 µg/ml) and monensin (2 µm). Cells were then fixed and permeabilized using Fixation/Permiabilization buffer from ebiosciences, and then stained with anti-cd3, -CD4, IL-13, and IL-17A antibodies. Flow cytometric analysis was performed on a Becton Dickinson LSR II and data was analyzed with FloJo software (Tree Star Inc.). For proliferation analysis, splenic CD4 + T cells from OT-II transgenic mice were prepared by positive selection using CD4 Miltenyi beads per the manufacturer s instructions. CD4 + T cells were labeled with CFSE (2.5 µm) for 5-7 minutes at 37 C. Labeled CD4 + T cells (3 x 10 6 cells) were transferred intravenously into CD45.1 congenic mice. Mice were immunized as above with alum/ova or alum/igg-ova. T cell proliferation was assessed, as CFSE dilution, at day 3 post-immunization by flow cytometry on an Accuri C6 flow cytometer (BD Biosciences). Data were analyzed with FlowJo software.

86 70 Statistical analysis Statistical analysis was done using Prism 5.0a (GraphPad Software). Unless otherwise noted, statistical significance for single comparisons was determined by Student s t-test; ANOVA with a Bonferroni post-test was used for multiple comparisons. Results Antigen-IgG immune complexes suppress the development of alum-driven Th2 and Th17 responses in vivo Data presented in the previous chapter demonstrated that IgG-antigen immune complexes effectively suppressed secretion of IL-1α and IL-1β, by blocking inflammasome oligomerization and caspase-1 activation. Given the critical role of IL- 1R1 signaling in the development of adaptive immune responses, we examined the effect of immune complexes on the generation of adaptive immune responses. To accomplish this, we utilized a murine model of allergic airway disease (Figure 4.1). Mice were immunized intraperitoneally with either ovalbumin (Ova) or IgG-Ova immune complexes, along with the adjuvant alum. Fifteen days later, mice were challenged intranasally with Ova for three consecutive days. After 48 hours, the extent of airway inflammation was assessed by histology and inflammatory cell infiltration was determined in the bronchoalveolar lavage (BAL) fluid. As expected, mice immunized with alum/ova showed eosinophilic infiltration in response to intranasal challenge with Ova (Figure 4.2 A and B) 80, Strikingly, those mice immunized with alum/igg-ova showed significantly decreased eosinophilic infiltration following challenge with Ova (Figure 4.2B and C). To determine if subsequent adaptive immune responses to the antigen were affected by immunization with IgG-immune complexes, we looked at serum antibody levels. Specifically we analyzed the levels of the Th2 related antibodies, IgG1 and IgE, and the levels of the Th1 related antibody IgG2c. We found that Ova-specific IgG1 and

87 71 total IgE levels were significantly decreased in mice immunized with alum/igg-ova as compared to mice immunized with alum/ova (Figure 4.2C). However, Ova-specific IgG2c levels were unchanged. These data suggested that Th2 response was impaired upon immunization with IgG-immune complexes. To further characterize the effect of presenting antigen as IgG-immune complexes on the generation of adaptive immune responses, we restimulated lymph node (LN) cells with Ova ex vivo. LN cells from mice immunized with alum/ova produced IL-13, IL-4, and IL-17A upon restimulation with Ova, consistent with an alum-drive Th2 and Th17 response (Figure 4.3A). Surprisingly, LN cells from alum/igg-ova immunized mice secreted significantly less IL-4, IL-13, and IL-17A upon restimulation with Ova (Figure 4.3A). We confirmed that the defect in IL- 13 and IL-17A secretion was due to decreased production by CD4 + T cells by intracellular cytokine staining (ICS) (Figure 4.3B). To determine if the impaired adaptive immune response found in alum/igg-ova immunized mice was due to a failure of APCs to properly process and present the IgG- Ova we adoptively transferred CFSE labeled T-cell receptor (TCR) transgenic OT-II CD4 + T cells to congenically mismatched mice. We then immunized the mice with either alum/ova or alum/igg-ova. We observed that OT-II CD4 + T cells proliferated similarly regardless of whether the mice were immunized with alum/ova or alum/igg-ova (Figure 4.4). These data suggested that processing and presentation of antigen remained intact. Together these data demonstrate that antigen-igg immune complexes effectively suppress the generation of alum driven Th2 and Th17 responses. Immune complexes suppress Th2 and Th17 responses through inhibition of IL-1α and Il-1β secretion rather than enhanced IL-10 production. IL-10 is a potent anti-inflammatory cytokine which affects numerous cell types and is important in limiting inflammation 250. It is known that immune complexes enhance

88 72 IL-10 secretion 207. To determine if enhanced IL-10 secretion by BMDM following immunization with immune complexes was required for the inhibition of Th2 and Th17 responses in vivo, we utilized IL-10-deficient (Il10 -/- ) mice. LN cells from Il10 -/- mice immunized with alum/igg-ova showed suppressed IL-13 and IL-17A production upon ex vivo restimulation, as compared to alum/ova immunized mice (Figure 4.5A). As data presented in the preceding chapter showed that immune complexes suppressed IL-1β secretion, we next asked if immune complex-mediated enhanced IL-10 secretion was responsible for decreased IL-1β secretion in vitro. Similar to our findings in vivo for the Th2 and Th17 responses, we found immune complexes suppressed Il-1β in response to the Nlrp3 agonists silica and ATP in Il10 -/- BMDM (Figure 4.5B). This data suggested that immune complex-mediated IL-1β suppression does not require enhanced IL-10 production. The NLRP3 inflammasome has been implicated in the generation of alum-driven Th2 responses in vivo 80,251. Furthermore, signaling through the IL-1R1 on CD4 + T cells enhances their expansion and differentiation into Th2 and Th17 effector cells 139,142. However, the role of NLRP3 inflammasome activation in alum-driven Th17 responses is unclear. To evaluate if the NLRP3 inflammasome was required for Th17 responses in the Alum/Ova model of allergic airway disease, we immunized mice deficient in NLRP3 (Nlrp3 -/- ), caspase-1 (casp1 -/- ), or IL-1R1 (Il1r1 -/- ) with alum/ova. Consistent with the previous data that had demonstrated a role for IL-1R1 in Th17 differntiation 142,241, we found that LN cells from Il1r1 -/- mice immunized with alum/ova had significantly decreased IL-17A production upon Ova restimulation ex vivo, as compared to WT mice (Figure 4.6A). In agreement with the data from Il1r1 -/- mice, LN cells from Nlrp3 -/- and Casp1 -/- mice secreted significantly less IL-17A than WT mice upon Ova restimulation ex vivo (Figure 4.6A and B). As expected, mice lacking NLRP3, caspase-1, and IL-1R1 also displayed impaired IL-13 production (Figure 4.6A and B). While these data suggest a role for caspase-1 and the NLRP3 inflammasome in the development of alum induced

89 73 Th2 and Th17 immune responses, the further decrease in IL-17A secretion in Il1r1 -/- mice indicates that NLRP3 inflammasome and caspase-1 independent signaling through the IL-1R1 may contribute to alum-driven Th17 responses in vivo. Both IL-1α and IL-1β signal through the IL-1R1, yet their individual contributions to the induction of alum-drive Th17 responses is unknown. To investigate their individual contributions, mice deficient in either IL-1α (Il1a -/- ) or IL-1β (Il1b -/- ) were immunized with alum/ova and then challenged intranasally with Ova. We observed decreased IL-13 and IL-17A secretion in LN cells restimulated ex vivo with Ova, in both Il1a -/- and Il1b -/- mice, as compared to WT mice (Figure 4.7A and B). These suggest that both IL-1α and IL-1β play important roles in the alum-driven Th2 and Th17 immune responses. Suppression of CD4 + T cell responses by immune complexes is rescued by exogenous IL-1α or IL-1β To determine if the suppression of macrophage IL-1α and IL-1β secretion by immune complexes observed in vitro contributed to the ability of immune complexes to suppress the generation of Th2 and Th17 responses in vivo, we immunized mice with alum/igg-ova in the presence or absence of recombinant IL-1α or IL-1β. The presence of either IL-1α or IL-1β during the sensitization phase with alum/igg-ova was sufficient to overcome the effects of IgG immune complexes and restore generation of Th17 responses (Figure 4.8 A and B). However, only exogenous IL-1α, but not IL-1β, was able to rescue the development of alum-drive Th2 responses (Figure 4.8A and B). Taken together these data suggest that both IL-1α and IL-1β are important for the generation of alum-driven Th2 and Th17 responses. Additionally, these data suggest that immune complexes suppress the generation of Th2 and Th17 responses through dampening the secretion of IL-1α and IL-1β. Furthermore, these data provide evidence of differential functions for IL-1α and IL-1β in the generation of the Th2 response.

90 74 Discussion The data presented in this chapter demonstrate that immunization with immune complexes in the presence of adjuvant alum suppressed the generation of Th2 and Th17 responses. Furthermore, we have shown that the NLRP3 inflammasome plays a critical role in the generation of both alum driven Th2 and Th17 immune responses. Finally, we have provided data that suggests that immune complexes suppress the generation of the Th2 and Th17 lineages by interference with the secretion of IL-1α and IL-1β. The induction of CD4 + T effector cell responses is critical for the host immunity. However, a disproportionate or otherwise inappropriate response leads to bystander tissue damage, and potentially autoimmunity. Therefore the induction of these responses must be strictly regulated. Signaling through the IL-1R1 is important in the generation of both Th2 and Th17 responses. First, IL-1R1 engagement in T cells enhanced expansion and differentiation Additionally, IL-1R1 signaling is required for the induction of Th17 responses 142,241. Previous studies have suggested a requirement for the NLRP3 inflammasome in the generation of alum-induced antigen specific Th2 responses 80,147. In this chapter we have expanded on these data, and shown that the NLRP3 inflammasome plays an important role in the generation of alum-driven Th17 responses as well. In agreement with, and expanding upon, previous studies we have provided evidence that IL-1R1 is critically important for the generation of Th17 responses induced by adjuvant alum. We have also presented data that loss of either IL-1α or IL-1β impairs the development of alum-directed Th17 cells in our allergic airway disease model, indicating that both are important for the full development of the CD4 + T effector lineage. It is interesting to note that while our data indicate that IL-1R1 is required for the development of Th17 effector cells, loss of NLRP3 and caspase-1 leads to a partial, yet significant, defect in Th17 development. This discrepancy can be explained by the requirement for both IL-1α and IL-1β in our model of allergic airway inflammation, as IL-1α secretion and activity is not entirely dependent upon processing through caspase-

91 Other data generated in our lab demonstrates that the development of Th17 cells in response to immunization with CFA is independent of the inflammasome and IL-1β, but is dependent upon IL-1α (unpublished data). The reason for this divergence in the use of IL-1α and IL-1β, and if this difference leads to any biologically significant alterations in T cell responses, remains unclear. We have found that presenting antigen in the form of antigen-igg immune complexes during the priming phase of our model of allergic airway disease effectively blocks the generation of Th2 and Th17 T effector responses. The secretion of IL-10, a potent anti-inflammatory mediator, is enhanced by immune complexes 207. However, this enhancement is not required for immune complexes to suppress alum-driven Th2 and Th17 responses, as Th2 and Th17 responses were suppressed in Il10 -/- mice. We found that the addition of exogenous IL-1α or IL-1β during the priming phase could restore Th17 responses in immune complex immunized mice. However, only exogenous IL-1α added during the priming phage restored Th2 responses, which agrees with previous studies that show a crucial role for IL-1α and IL-1R1 in Th2 sensitization in a house dust mite allergic airway disease model 252. These data could suggest that IL-1α and IL-1β have differential effects on the polarization of Th2 and Th17 cells, where IL-1α and IL- 1β both are important for Th17 development but IL-1α is more important for Th2 responses, perhaps due to the presence of an as of yet unidentified accessory protein that sensitizes Th2 cells to IL-1α but not IL-1β. These data could also be due to differences in clearance between IL-1α and IL-1β. Additional work is required to determine why there appears to be this difference in the role of IL-1α and IL-1β in the development of Th2 cells. The presence of immune complexes indicates the presence of effective humoral adaptive immune responses. The data presented in this chapter suggests that successful humoral immune responses may serve as negative feedback loop, preventing the development of novel cellular immune responses upon re-exposure to infection. As our

92 76 data indicates that immunization with IgG-immune complexes in alum blocks the development of Th2 and Th17 responses, it may be possible to use such an immunization therapeutically in the treatment of allergy or autoimmunity.

93 Figure 4.1: Schematic representation of our allergic airway disease model. Mice were immunized with alum/ova or alum/igg-ova on day zero. After two weeks the mice were challenged intranasally with Ova protein on three consecutive days. 48 hours following the final challenge, mediastinal LN, blood, lungs, and BAL fluid were harvested. 77

94 Figure 4.2: Immunization with IgG immune complexes suppresses airway inflammation. Mice were immunized as described in Figure 4.1. A. H&E stained lung sections were imaged. Representative sections of four mice per group are shown; upper panel scale bar = 50 µm; lower panel scale bar = 20 µm. B. Differential cell counts of HEMA3 stained BAL fluid. **p 0.01 by 2-way ANOVA with a Bonferroni post-test. C. Total IgE, and Ova specific IgG1 and IgG2c were measured from the serum by ELISA. ** 0.01, ***p by Mann-Whitney test. Determinations were performed in triplicate and are expressed as the mean±sd (B.) or the mean±sem (C.). Data shown are representative of three independent experiments each with a minimum of three mice per group. 78

95 Figure 4.3: Immunization with IgG immune complexes suppresses the generation of Th2 and Th17 responses. Mice were immunized as described in Figure 4.1. A. Mediastinal lymph nodes were collected and restimulated in vitro with Ova protein. After 72 hours, supernatants were collected and analyzed for IL-17A, IL-13, and IL-4 by ELISA. *p 0.05, ***p by Student s t-test. B. Intracellular cytokine analysis of LN cells stimulated in vitro with Ova for 72 hours. Determinations were performed in triplicate and are expressed as the mean±sd. Data shown are representative of three independent experiments, each with a minimum of three mice per group. 79

96 Figure 4.4: CD4 + T cells proliferation is unaffected by immunization with IgG immune complexes. Congenically mismatched CFSE labeled OTII T cells were transferred i.v. into WT mice. 24 hours later mice were immunized with either alum/ova or alum/igg-ova. After 72 hours the draining lymph nodes were collected and CFSE dilution was analyzed by flow cytometry. Data shown are representative of two independent experiments. 80

97 Figure 4.5: IL-10 is not required for immune complex mediated suppression of Th2 and Th17 responses. A. Il10 -/- mice were immunized as described in Figure hours following the final intranasal Ova challenge, mediastinal lymph nodes were collected and restimulated in vitro with Ova protein. After 72 hours supernatants were collected and analyzed for IL-17A and IL-13 by ELISA. *p 0.05, **p 0.01 by Student s t-test. B. Il10 - /- BMDM were LPS primed with or without eigg and then challenged for six hours with either silica or ATP. Supernatants were harvested and analyzed for IL-1β by ELISA. ***p by Student s t-test. Determinations were performed in triplicate and are expressed as the mean±sd. Data shown are representative of three independent experiments, (A.) with a minimum of three mice per group. 81

98 Figure 4.6: The NLRP3 inflammasome is important for the generation of aluminduced Th2 and Th17 immune responses. A. WT, Nlrp3 -/-, or Il1r1 -/- mice were immunized with alum/ova as described in Figure 4.1. Mediastinal lymph nodes were collected 48 hours after the final intranasal challenge, and were restimulated in vitro with Ova protein. After 72 hours culture supernatants were collected and analyzed for IL-17A and IL-13 by ELISA. **p 0.01, ***p by 1-way ANOVA with Bonferroni posttest. B. WT and Casp1 -/- mice were immunized as described in Figure 4.1. Mediastinal lymph nodes were collected 48 hours after the final intranasal challenge, and were restimulated in vitro with Ova protein. After 72 hours culture supernatants were collected and analyzed for IL-17A and IL-13 by ELISA. *p 0.05, **p 0.01 by Student s t-test. Determinations were performed in triplicate, and are expressed as the mean±sd. Data shown are representative of three independent experiments each with a minimum of three mice per group. 82

99 Figure 4.7: IL-1α and IL-1β are both important in alum-driven Th2 and Th17 responses. WT, Il1a -/- (A.), or Il1b -/- (B.) mice were immunized as described in Figure 4.1. Mediastinal lymph nodes were collected 48 hours after the final intranasal challenge, and were restimulated in vitro with Ova protein. After 72 hours culture supernatants were collected and analyzed for IL-17A and IL-13 by ELISA. *p 0.05, **p 0.01, ***p by Student s t-test. Determinations were performed in triplicate and are expressed as mean±sd. Data shown are representative of three independent experiments each with a minimum of three mice per group. 83