|Home | About | Journals | Submit | Contact Us | Français|
Deposition of antibody-antigen immune complexes in tissues underlies the pathogenesis of a range of human autoimmune diseases from glomerulonephritis, systemic lupus erythematosus (SLE), arthritis and transplantation rejection to rheumatic fever. There is experimental and clinical evidence to suggest that autoimmunity is a risk factor for cardiovascular disease. Here, we overview the literature that links autoimmune diseases and accelerated atherosclerosis and review our understanding of IC-induced inflammation. In particular, we will discuss the contribution of the location and composition of immune complexes (ICs) to the cellular and molecular mechanisms of disease that develop and detail the steps leading to IC mediated leukocyte recruitment and tissue injury. The focus will be on neutrophil dependent tissue injury. Although these cells have not found a prominent place in discussions of chronic inflammatory conditions such as autoimmunity and atherosclerosis, recent literature as discussed here indicates that neutrophils initiate IC induced inflammation that in turn launches and informs the subsequent immune response. A fuller understanding of the mechanisms involved in the initiation and perpetuation of antigen-IgG antibody inflammation may lead to the design of innovative therapeutic strategies to prevent cardiovascular complications in patients with autoimmune diseases.
Well-controlled studies have demonstrated that autoimmunity is an independent risk factor for atherosclerotic lesions and premature myocardial infarction1, 2. Population-based studies in various countries have shown that cardiovascular disease-associated mortality in patients with rheumatoid arthritis is 1.5 to 2.0 times higher than control individuals3-6. Similarly, increased rates of CVD have been reported in other forms of inflammatory arthritis including giant cell arteritis7 and psoriatic arthritis8. The risk for CVD and myocardial infarction (MI) is also pronounced in patients with systemic lupus erythematosus (SLE), a complex autoimmune disease affecting multiple organs. The Pittsburgh lupus cohort presented a 52-fold increase in cardiovascular events in young women with SLE 9. The Framingham Heart Study risk factors contribute to CVD in patients with SLE and other autoimmune disorders probably in a proportional manner. However, after statistical control for these factors the risk for MI in SLE patients is 10-fold and for overall CVD is 17-fold higher than in controls10. Patients with anti-phospholipid syndrome, typically characterized by recurrent thrombosis and pregnancy loss have increased rates of atherosclerotic lesions. The antibodies in this syndrome are directed against plasma proteins complexed to anionic phospholipids such as β2-glycoprotein I (β2GPI) found in atherosclerotic lesions11, 12, and interfere with endothelial cell function13. Epidemiological studies have identified a number of immune markers to correlate with the development of CVD in patients with autoimmune disorders. Rheumatoid factor (IgM anti-Fc IgG antibody) has been reported to represent an independent risk for the development of CVD in patients with RA14. Erythrocyte sedimentation rate, VCAM and ICAM serum levels, markers of immune cell activation correlate with the development of atherosclerosis in RA patients15. Similarly in patients with SLE, serum levels of cytokines and chemokines like IL-6, MCP-1, ICAM, VEGF, soluble CD40 ligand and TNFα correlate in cross-sectional studies with the development of CVD3, 16.
In addition to clinical evidence, animal studies provide strong arguments that autoimmunity predisposes to accelerated atherosclerosis. Vessel inflammation and fatty lesions are increased in the lupus prone mouse MRL/lpr17 and increased rate of myocardial infarcts have been noted in male NZW × BXSB mice, also prone to developing lupus18. Similarly, B6.Faslpr/lpr (Fas deficient mice) and B6.Sle1.2.3 (triple congenic mice in which 3 Sle loci from the NZM2410 mouse have been transferred to the B6 mouse) display increased rates of CVD19. Evidence that ICs accelerate atherosclerosis in animal models have also been provided. Repeated immunization with protein antigens that leads to IC formation accelerated diet-induced atherosclerosis in rabbits and mice20. A deficiency in FcγRs, receptors for monomeric and complexed IgG, limited development and progression of atherosclerosis in mice21. Moreover, immunoglobulin (IVIG) therapy, used in the treatment of immune-mediated disorders for more than 25 years22, markedly suppressed atherosclerosis through immunomodulation of FcγRs23. Some clues on how ICs may be atherogenic are emerging. Oxidized lipids and antibodies to oxidized low density lipoproteins (LDL) are elevated in SLE11, 12 and macrophages incubated with oxidized LDL-IC demonstrate enhanced uptake, cholesterylester accumulation and foam cell formation in vitro compared to native LDL24. Uptake through FcγRs was associated with macrophage activation and release of cytokines which are characteristics of atherosclerotic lesions25. The independent contribution of SLE and IgG-ICs to the development of atherosclerosis is further testified by the observation that apoE and Faslpr/lpr (lupus prone mice) double deficient B6 mice spontaneously develop lupus-like disease and develop far more atherosclerotic lesions than mice deficient in apoE or Fas alone. Elevated anti-oxidized phospholipid IgG antibody levels in the double deficient was associated with increased IgG deposition in the aorta intima mice and significantly correlated with atherosclerotic lesion area suggesting that autoantibodies play a role in the pathogenesis of atherosclerosis26.
Chronic systemic inflammation, endothelial dysfunction and immune dysregulation observed in SLE and RA are known to also promote atherosclerosis11. Indeed, the cholesterol lowering drug HMG-CoA reductase inhibitor simvastatin, also known for its anti-inflammatory effects, reduced the atherosclerotic lesion area and proinflammatory cytokine production in mice with lupus-like features, independently of effects on cholesterol levels27. LDL and cholesterol may not contribute differently in SLE patients than in normal controls, but disease duration and lack of aggressive treatment with immunosuppressants1 and the high levels of homocysteine2 contribute independently to early CVD. These clinical data point to an independent contribution of SLE immune features to the development of CVD. Consistent with this, studies in animal models show that transfer of bone marrow cells from B6.Sle1.2.3 mice (which readily develop lupus and CVD) to the B6.LDLR-/- mice (an accepted model of atherosclerosis) increased the rate of formation of atherosclerotic lesions which were heavily infiltrated by CD3+ cells28. In a human artery-severe combined immunodeficiency mouse chimera, adoptively transferred human T cells instigated arterial wall inflammation only when wall embedded dendritic cells were conditioned with TLR ligands29. If the transferred T cells produce TRAIL then it may bind to DR5 (a TRAIL receptor) on the surface of smooth muscle cells and cause their apoptotic death30. It appears accordingly that activated T cells, probably displaying adhesion molecules31, present in patients with SLE may enter vessels in which the resident DC have been activated and contribute significantly to the expression of vessel pathology in the form of atherosclerotic plaques. Macrophages may differentiate in the present of lupus serum to DC32 and may account for increased presence of DC in the vessel walls of SLE patients.
Less well appreciated is the contribution of neutrophils to atherosclerotic lesion development in the context of autoimmunity. Neutrophils are abundant in joint lesions of RA patients and play a prominent role in RA development in mouse models33. Neutrophils are also present in renal tissue of lupus nephritis patients34. Although a direct role for neutrophils in murine lupus nephritis models has not been directly shown they do significantly contribute to immune mediated anti-glomerular basement membrane (GBM) nephritis in mice as discussed later in this review. Murine anti-GBM nephritis and SLE nephritis likely share underlying mechanisms as there is concordance in the molecules required for progression of renal disease in these two models35. PMNs have been detected in human atherosclerotic plaques36 albeit they are fewer compared to other immune cell types. PMNs have been found in the plaque erosion or rupture on atherectomy specimens from patients with angina pectosis and myocardiac infarction suggesting that they may contribute to plaque destabilization. Atheromatous lesions contain markers of neutrophil degranulation such as myeloperoxidase (MPO) and neutrophil proteases indicating the presence of active neutrophils. MPO generates reative oxygen species that promote endothelial cell apoptosis and tissue factor expression as well as LDL protein nitration and lipid peroxidative which can advance lesion development. In parallel, the peripheral blood neutrophil number is an independent risk factor and a predictor for future occurrences of cardiovascular events. Indeed high peripheral neutrophil counts correlate with the increase of coronary disease better than any other leukocyte subset including total white blood cell, lymphocyte or monocyte count 37. In animal models neutrophils and/or monocytes, were detected within atherosclerotic lesions38 and colocalized with myeloperoxidase39. Neutrophil depletion decreased neutrophil content and lesion size in ApoE null mice. An inhibition specifically of CXCR2 on neutrophils decreased neutrophil accumulation in the aortic root and arch 40. On the other hand, an antagonist for CXCL12, a chemokine that retains granulocytic precursors within the bone marrow increased the number of neutrophils both in circulation and aortic root plaques and adventitia in ApoE null mice. This treatment was associated with an increase in the proportion of apoptotic neutrophils, particularly on the luminal side of the atherosclerotic plaque40. These studies together suggest a causal relationship between neutrophils and its products and atherosclerotic lesion development but definitive mechanisms by which these cells accelerate lesion formation in the context of autoimmune disorders remains unclear. Migration and activation of Fc-bearing neutrophils within plaques could occur secondary to stimulation with ICs. hFcγRIIA expressing human monocytes adhere to endothelial cells covered with immobilized ICs formed with anti-oxLDL and LDL41. Furthermore, FcγRIIA-dependent immune cell adhesion increases the secretion of inflammatory cytokines or chemokine, which initiates vascular injury and/or recruitment of other immune cells41-43.
TLRs have also been widely implicated in atherogenesis. Toll-like receptors (TLRs), a type of pattern recognition receptor, have emerged as key components of the innate immune system. They activate multiple steps in the inflammatory reaction in response to microbial proteins, DNAs and RNAs, either from invading microbes or host-derived self ligands associated with cell stress or death (DAMPs, damage-associated molecular patterns). Deficiency of TLR4, TLR2 and MyD88, an adapter molecule which is essential for TLR signaling pathway, impaired disease in atherosclerosis prone ApoE or LDLR deficient mouse strains44. TLRs also play important roles in SLE. Both deficiency of TLR7 and MyD88 decreased anti-DNA antibody levels and ameliorated SLE activity while TLR3 deficient animal showed comparable level of anti-DNA production to control animals. The effect of TLR-9 deficiency on the lupus phenotype remains poorly understood45, 46. In dendritic cells, several groups have also recently shown that FcγRIIA facilitates the delivery of nucleic acid containing ICs, which are salient features of SLE, to intracellular vesicles containing TLR7 and TLR9 which initiate signaling46-48. Human neutrophils express mRNA for all the TLRs, except TLR3 49-51. Whether the FcγRIIA/TLR pathway described in dendritic cells operates in neutrophils is not known and is a fruitful area for investigation.
Ultrasonography to detect and measure intima-media thickness and electron beam computed tomography are readily used to detect preclinical atherosclerosis in patients with rheumatic and other immune disorders3. Although the lesions are suspected to contain a larger number of immune cells, at this point there are no pathology reports on possible differences between the atherosclerotic lesions in patients with autoimmune and without autoimmune disorders. The available epidemiologic evidence urges clinicians to monitor classical CVD risk factors and advise patients appropriately. Early, non-invasive detection of atherosclerotic lesions would suggest aggressive intervention with cholesterol lowering agents which are known in addition to tame the activated immune system in various autoimmune disease animal models52. In addition, the preceding discussion on the involvement of cellular components of the immune system in the expression of CVD in patients with autoimmune diseases, calls for sharp and aggressive control of the aberrantly activated immune system.
In summary, the clinical and experimental data together signify an association between autoimmunity and atherosclerosis. There is also emerging evidence that ICs and neutrophils, prevalent features of autoimmune disease, may contribute to atherosclerotic lesion development. Thus a better understanding of how these aspects of the innate immune response modulate inflammation may provide insights into mechanisms of accelerated atherosclerosis in autoimmune disorders.
Advanced atherosclerosis leading to MI or unstable angina leads to periods of ischemia, and therapeutic intervention can result in reperfusion injury. Although there is no clinical data to suggest that rheumatoid arthritis or SLE predispose to ischemia/reperfusion (I/R) injury in patients, recent work in animal models suggests that autoantibodies present in SLE may contribute to I/R-induced organ injury as has been described for natural antibodies. B6.MRL/lpr mice subjected to mesenteric I/R exhibited enhanced intestinal injury charaterized by profound neutrophil infiltration and remote lung injury compared to non-autoimmune mice. Moreover, anti-DNA and anti-histone mAbs often found in sera of lupus mice reconstituted I/R injury in I/R-induced injury resistant Rag-1 deficient mice which lack IgG. Anti-DNA and anti-histone mAbs may bind to apoptotic intestinal cells which leads to in situ IC formation, neutrophil infiltration and neutrophil mediated tissue damage in response to I/R. Thus SLE autoantibodies may serve as a second hit to tissue damage following reperfusion 53. Further work is needed to assess whether this mechanism operates in the context of autoimmunity in patients.
ICs are produced continuously in response to infection, tissue injury and foreign antigens. The Fc portion of complexed IgG or IgM can activate complement, a system of serum and cell surface proteins that interact in a cascade to generate important effectors of immunity. Antibody and complement components assemble on microbes and their toxins to target them for destruction by the innate immune system. However when the same immune reactants interact with antigens associated with host tissue, the immune response inflicts tissue injury that extends beyond the collateral tissue damage observed during the course of an infectious challenge. An immune response to exogenous antigens can lead to discomforts such as skin irritation to potentially fatal diseases such as bronchial asthma. A response to endogenous tissue antigens can result in transfusion reactions and graft rejection when non-self antigens are encountered, and autoimmune disease when self antigens are recognized on host cells due to a break in immunologic tolerance.
Injury in antibody mediated diseases may be categorized as Type I, II or III hypersensitivity reactions based largely on the nature and location of the antigen that is the target of the response and the effector mechanisms of tissue injury (Table 1). In the clinical setting the situation is more complex as diseases are triggered by a combination of humoral and cell-mediated factors stimulated by a single or multiple antigen, but the classification is useful because it may predict the pathologic features of a given disease. Type I hypersensitivity, results from the cross-linking of IgE Fc receptors on mast cells of individuals previously sensitized to the antigen. As the focus of this review is on IgG-ICs, Type I hypersensitivity will not be further discussed. In Type II hypersensitivity, pathogenic IgG or IgM antibody against cell surface or extracellular matrix antigens, or exogenous “planted” antigens such as bacteria directly induce inflammation or interfere with cellular functions without tissue injury. In some cases ICs fail to produce an inflammatory response because they are inaccessible to leukocytes (e.g. Heymann nephritis), the antibody bound receptor is rapidly internalized (e.g. Myasthenia Gravis) or is against blood elements that are efficiently removed from the circulation by the phagocyte system in the liver and spleen (e.g. autoimmune hemolytic anemia). In Type III hypersensitivity, ICs formed by IgG or IgM antibody against circulating antigens form and lodge in the capillaries, the surrounding basement membrane and tissues, thus triggering an inflammatory response (Table 1). Within Type II and III responses the sequence of events in inflammation depends on the relative accessibility of the IC deposits to circulating inflammatory cells. Intravascular deposited ICs may directly engage circulating leukocytes. Perivascular and/or extravascular ICs may first engage resident cells that then release inflammatory mediators capable of activating the endothelium and their ability to recruit cells. Prototypic animal models of Type II and Type III hypersensitivity responses are anti-glomerular basement membrane (GBM) nephritis, also referred to as nephrotoxic serum nephritis and the Reverse Passive Arthus (RPA) reaction respectively. Anti-GBM nephritis is induced by the intravenous injection of nephrotoxic serum in animals pre-immunized with IgG. The antibody recognizes components of the GBM through open endothelial fenestrae, leading to renal leukocyte infiltration, extensive glomerular injury, and proteinuria that mimics features of Goodpasture's syndrome in humans 54. The RPA reaction is induced in animals by the systemic delivery of antigen and the local tissue injection of antibody. This results in IC formation in the intravascular, perivascular and tissue compartment that triggers intense PMN infiltration, edema and hemorrhage. Histologically, RPA lesions exhibit features of vasculitis in SLE 55.
In both Type II and III responses, neutrophils are the first immune cells to be recruited. Neutrophils account for 50-70% of all circulating leukocytes. They are terminally differentiated cells with a relatively short half life of 4-10hr in circulation that can be extended up to 48hrs during inflammation. They circulate in a quiescent state. However, following an insult to tissues, pro-inflammatory stimuli and adhesion molecules presented by the inflamed endothelium rapidly induce neutrophil rolling and adhesion to the vessel wall and subsequent transmigration into tissue56. Resting neutrophils can also acquire a state of preactivation that amplifies their responsiveness to subsequent external stimuli in a process referred to as priming. This both localizes and augments the neutrophil response during inflammation57-59. Fundamental to an appreciation of the role of neutrophils in IC induced injury is the principle that the arsenal of weapons that they have evolved to destroy microorganisms also play a significant role in damaging tissue. Engagement of FcγRs and complement receptors triggers phagocytosis, release of enzymes and vasoactive amines and generation of a vigorous oxidative burst. Enzymes stored in specialized intracellular granules promote hydrolytic substrate degradation as in classical lysosomes but also kill ingested bacteria and regulate various physiological and pathological processes when secreted. Among these, the antimicrobial protein myeloperoxidase (MPO) and proteinases are of particularly interest as they are both effectors and targets of autoimmune responses. Neutrophil proteases degrade the extracellular matrix, serve as ligands and/or modulate the function of adhesion molecules on the surface of neutrophils60-62, and inactivate anti-inflammatory mediators63. Neutrophil derived reactive oxygen species (ROS) are directly injurious to host cells and participate in autocrine and paracrine cell signaling64 that promotes inflammation. Finally, the transcriptional upregulation of cytokines and chemokines by activated neutrophils65 66 enhances neutrophil influx and the recruitment and activation of monocytes, lymphocytes and dendritic cells. Products from mononuclear cells in turn activate neutrophils and regulate their release from the bone marrow65, 67. Thus alarm signals from the injured host, and a number of pro-inflammatory mediators released by neutrophils together serve to marshall and coordinate the adaptive immune response. This potentially places neutrophils at the forefront of IC mediated pathogenesis.
Three clinical examples of autoantibodies produced in disease, antineutrophil cytoplasmic autoantibodies, anti-endothelial cell antibodies and anti-glomerular basement membrane nephritis will be discussed. These were chosen as they target different components of the vasculature, are featured in several types of unrelated autoimmune disorders and studies have elegantly illustrated the point that autoantibodies generated in these Type II responses are pathogenic in vivo and trigger neutrophil activation and vascular damage.
Anti-neutrophil cytoplasmic autoantibodies (ANCA) are associated with small vessel necrotizing vasculitis with a paucity of immunoglobulin deposits in the vessel wall. The major antigen specificities of ANCA in vasculitis are myeloperoxidase (MPO-ANCA) or serine proteinase 3 (PR3-ANCA) which are components of neutrophil granules. The ANCA antigens are translocated to the cell surface upon stimulation of cells with pro-inflammatory cytokines. Their subsequent ligation by the F(ab2)′ component of the ANCA-IgG enhances neutrophil activation with the additional release of granule contents and ROS. Cell bound ANCA IgG may also be recognized by FcγR on neighboring neutrophils, and trigger PMN activation. MPO-ANCAs have been shown to activate MPO and generate hypochlorous acid that causes cytolysis of endothelial cells, thus identifying another mechanism of ANCA pathogenicity. Circulating ANCA are reported in 90% of patients with Wegener's granulomatosis, microscopic polyangiitis and renal limited pauci-immune necrotizing and crescentic glomerulonephritis. This as well as additional circumstantial clinical evidence argues for a pathogenic role for ANCA in patients68 69. There is also compelling in vivo data in mice in support of this argument: Anti-MPO IgG leads to the development of nephritis and vasculitis70 and augments the adhesion and transmigration of leukocytes across the endothelium71.
AECA, a distinct entity from ANCA, represent a heterogenous group of autoantibodies to endothelial antigens and endothelial cell-planted antigens that are unified in their potential for vessel wall damage72. AECA lesions are induced by the adoptive transfer of IgG purified from the blood of individuals with the disease. Vasculitis-like lesions were observed with murine AECA generated in response to immunization with AECA IgG fractions of patients with Wegener's granulomatosis (WG). Albeit human AECA itself was not shown to be pathogenic, mice exhibited IgG deposits in blood vessel walls as a result of the anti-anti-AECA, which had binding properties similar to the original human AECA 73. AECAs planted on endothelial cells can support interactions with neutrophils and can also directly activate and damage the endothelium74.
Circulating IgG against GBM can clinically produce anti-GBM disease including glomerulonephritis and when accompanied by pulmonary hemorrhage, is known as the Goodpasture syndrome 75, 76. Renal histology typically reveals focal or segmental glomerular necrosis with destruction of the GBM, and cellular proliferation leading to crescent formation. Human anti-GBM antibody binds the non-collagenous domain of type IV collagen which is critical for the maintenance of the GBM superstructure77. Injection of experimentally induced, human anti-GBM antibody or antibody generated in GBM immunized animals, promotes glomerulonephritis in mice77 78, 79. Inflammation and damage to the vessel wall is a primary concern as the anti-GBM is accessible to the circulating leukocytes. In animals models of anti-GBM disease, recruited PMNs stay in the capillary to induce injury likely through oxidative injury to endothelial cells80, 81 and don't appear to infiltrate into the subepithelial area perhaps as a result of downregulation of chemokine receptors or reduced PMN deformability as illlustrated in ANCA disease69. Neutrophil induced injury to the glomerulus, the primary site of IC formation, promotes interstitial accumulation of macrophages and T cells. Studies of human biopsies revealed prominent accumulation of neutrophils within glomerular capillaries in membranous proliferative GN, crescentic GN and lupus nephritis 34, 82, 83 that was associated with significant interstitial macrophage influx. Although anti-GBM nephritis was not included in the panel of biopsies, these reports illustrate the potential pathogenic role of neutrophils in orchestrating glomerular damage in prevalent chronic immune complex mediated glomerular diseases.
Although the location of ICs is relatively well-defined when antibodies are generated in situ against specific tissue antigens as illustrated in examples of Type II reactions, the fate of circulating ICs in Type III responses is less predictable. While large aggregates, or insoluble ICs are primarily cleared by the mononuclear phagocyte system in the liver and spleen, small soluble ICs have a propensity for tissue deposition that is influenced by systemic factors, physiochemical properties of the ICs and tissue specific hemodynamics. Lines of evidence in humans and animal models suggest that circulating ICs first localize within the vasculature and then translocate into extravascular tissue 84-89. Changes in vascular permeability84-87 induced by cytokines and/or lipid mediators secreted by hematopoietic cells and mast cells 90-92 is likely the initial trigger for IC deposition. Also, ICs themselves promote vascular leakage in “permeability” susceptible tissues such as the joint tissue which may provide an explanation for why circulating ICs can promote tissue specific disease such as rheumatoid arthritis (RA)93 94. However, permeability is likely an amplifier rather than a prerequisite for localization of arthritogenic antibody and subsequent development of RA 95. A change in the physicochemical parameters of ICs, reflected by a change in the properties of the antigen (e.g. charge, valence, size etc) or the antibody (e.g. subclass, size, charge, precipitability complement fixing abilities), can modulate the extent and location of IC deposition96. For example IC formed with an alteration in the ratio of antigen and antibody, or composed of cation antigen have a higher potential to activate and bind complement C1q 97, 98 an early component of the classical complement pathway, that promotes IC binding to endothelial cells 99, 100 and extracellular matrix molecules 88, 101, 102 and modifies the lattice structure of ICs to facilitate deposition 103. C1q also activates C3 to C3b. C3b covalently linked to ICs binds to cell surfaces 103, 104. It also solubilizes ICs by interfering with Fc-Fc interactions96, which reduces their size and thus their clearance. Moreover, C3b-based convertases activate C5 that can in turn promote IC deposition 105.
Studies in the past several years suggested that FcγRs play primary roles in diseases initiated by antibodies 106, 107. FcγRs are members of the immunoglobulin gene superfamily that bind the Fc binding domain of IgG and are widely expressed in the hematopoietic system. Currently, two groups of FcγRs are recognized on cells of the immune system. The high affinity FcγRI which binds monomeric IgG and the low affinity receptors, FcγRII and FcγRIII which preferentially bind complexed IgG. In humans, the low affinity FcγRs are present in multiple isoforms, FcγRII (CD32)A and B and FcγRIII (CD16) A and B. The FcγRs are further classified as activating or inhibitory. Signals from these receptors are transmitted via immunoreceptor tyrosine-based activation (ITAM) or inhibitory (ITIM) motifs respectively. FcγRI, FcγRIIA and FcγRIIIA are activating receptors with ITAMs present in the cytoplasmic domain of the α-subunit (FcγRIIA) or in an accessory signaling γ-chain with which the α-subunit associates (FcγRI and FcγRIII). Upon cross-linking of the receptors by ICs, the tyrosine residues in the ITAM motifs are phosphorylated by src family tyrosine kinases, which initiate a cascade of signaling events that trigger neutrophil effector responses. The exception is the uniquely human FcγRIIIB, which anchors to the outer leaflet of the neutrophil plasma membrane through a glycosylphosphatidylinositol (GPI) linkage and does not contain or associate with ITAMs. It may signal by associating with FcγRIIA and the integrin complement receptor 3 (CR3, Mac-1) which serve as signaling partners, and/or localizing to membrane rafts enriched in signaling molecules like Src protein kinases. The ITIM motif is present in the cytoplasmic domain of the single chain, inhibitory FcγRIIB. Coligation of activating and inhibitory receptors on the same cell by ICs inhibits ITAM-triggered activation thus providing a higher threshold for activation of cells107. Human neutrophils express FcγRI, which is upregulated on the surface by cytokines such as IFNγ, and contain low levels of FcγRIIB but this is not consistently reported108-110. They also express the two low affinity receptors, FcγRIIIB and FcγRIIA for which there are no genetic equivalents in mice or other mammals111. FcγRIIIB's surface expression is 4-5 fold higher than FcγRIIA and it is present only on granulocytes. Neutrophil activation by inflammatory mediators leads to rapid FcγRIIIB shedding and modulates FcγRIIA expression and function112. Thus, the repertoire or activity of hFcγRs is regulated at an inflammatory site, which in turn may determine the magnitude of neutrophil effector responses.
Murine neutrophils express FcγRI and the inhibitory FcγRII and unlike humans, have two activating receptors FcγRIII and IV with low and intermediate affinity for IgG respectively, which rely on the common γ-chain for expression and signaling113, 114. Based on sequence homology of the extracellular domains, mouse FcγRIII resembles human FcγRIIA and FcγRIV is the orthologue of human FcγRIIIA107. Thus in humans, the low affinity human receptors are single polypeptide molecules with FcγRIIA containing its own signaling domain while the murine counterparts function as multi-protein complexes, with ligand binding and signaling functions present on separate polypeptides (Figure 1). Mice deficient in the α-subunit of specific activating receptors or the common γ-chain are protected in acute and progressive glomerulonephritis, autoimmune skin diseases, arthritis, SLE nephritis, and the RPA reaction107. However, conclusions in γ-chain deficient mice should take into account the increasingly apparent role for the γ-chain in the function of several receptors besides FcγRs. These include integrins on platelets and leukocytes, lymphocyte antigen receptors, and cytolytic receptors on NK cells and receptors on osteoclasts 115. Mice lacking the inhibitory FcγRII exhibit exaggerated responses in the aforementioned autoimmune models and develop spontaneous autoimmune kidney disease 106 113 indicating that FcγRIIB is an important regulator of its activating counterparts. Together these data infer a central role for FcγRs in modulating a spectrum of antibody mediated diseases in the skin, kidney, lung and joints of mice.
In addition to the strict requirement for FcγRs, complement is required for IC-induced inflammation and subsequent end organ damage. The three major pathways of complement activation are the classical pathway, activated by ICs composed of antibody of certain isotypes, the alternative pathway triggered by microbial cell surfaces and the lectin pathway activated by plasma lectin that binds mannose residues on microbes. Complement activated by all three pathways serves the same function. The central event is the generation of complement protein C3, which is then cleaved by C3 convertase to give rise to C3a and C3b. Further binding of C3b to C3 convertase leads to the formation of C5 convertase that cleaves C5 to generate C5a and initiates the terminal steps of complement activation. C3b covalently attaches to surfaces of cells in which complement is activated to target them for phagocytosis by macrophages. Complement products C3a and C5a are powerful anaphylotoxins that were suspected for a long time to be the primary mechanisms by which IgG-ICs trigger inflammation. Unexpectedly, mice deficient in complement C3 mounted normal IC induced inflammatory responses in several immunologic models shown to be Fcγ-chain dependent, leading to the conclusion that FcγRs are the primary initiators of IgG-IC mediated inflammation. However, the importance of complement was reasserted in subsequent studies which showed that the receptor for complement component C5a is essential for IC-mediated inflammatory responses in the lung, peritoneum 116, 117 and kidney 118, 119 of mice. In models that require C5a receptor but not complement C3, C5a may be generated by a proteolytic pathway independent of C3b. In support of this, C5a receptor blockade significantly attenuated RPA induced neutrophil accumulation and edema in C3 deficient mice 117 and in the genetic absence of C3, thrombin substituted for the C3 dependent C5 convertase to generate C5a 120. C5a receptor engagement on mast cells and macrophages increases expression of activating FcγRs and suppresses inhibitory FcγRIIB expression 106. Therefore C5a, which is a potent chemoattractant, may have a broader role in regulating the relative levels of inhibitory and activating FcγRs on tissue resident cells121. A proximal event in the classical pathway of complement activation is the binding of the C1q component of complement C1 to the Fc region of complexed IgG or IgM. This results in the cleavage of C4 and C2 to form an enzyme complex that acts as a C3 convertase and cleaves C3. The role of early components of the classical complement pathway in autoimmune disorders such as SLE is paradoxical. Unlike the later complement activation products such as C5, which are proinflammatory, the early complement components C1q and C4 facilitate IC disposal and the normal clearance of apoptotic cells that presumably bear SLE autoantigens. Thus a deficiency in C1q or C4 leads to a greater susceptibility to SLE122, 123. On the other hand, complement C1q itself is a target of autoimmune antibodies. Anti-C1q antibodies are strongly associated with lupus nephritis and have been detected in several mouse models of SLE 124, 125.
Neutrophil accumulation is an early and consistent feature in IC mediated diseases. The importance of FcγRs in this process is shown by the reliable reduction in neutrophil recruitment observed in Fcγ-chain deficient mice in various models of inflammation106, 113. Similarly, complement deficiency or blockade of complement components, C5a, C5aR or C3 leads to a reduction in neutrophil accumulation in select models of IC mediated disease117, 126. The prevailing paradigm is that tissue resident cells (mast cells and macrophages) sense ICs through FcγRs and complement receptors. This results in the elaboration of secondary mediators such as TNF and chemotactic chemokines106, which activate endothelial cells. Adhesion receptors together with surface bound chemokines upregulated on the surface of activated endothelial promote neutrophil recruitment through the well-described multistep process requiring selectin mediated rolling, and integrin mediated adhesion and transmigration 56, 89, 127. Recent work suggests that this paradigm of IC-induced neutrophil accumulation may need to be revisited in some cases. That is, when ICs are accessible to circulating neutrophils as in the case of in situ AECA, anti-GBM or soluble ICs initially deposited within the vasculature, ICs themselves may directly tether neutrophils. In vitro, FcγRIIIB played a primary role in neutrophil tethering to immobilized ICs under flow but in the framework of activated endothelial cells and monomeric AECA antibodies, FcγRIIA cooperated with chemokines to enhance PMN adhesion 128, 129. In vivo, the contribution of FcγRs in specific steps of neutrophil recruitment was explored by intravital microscopy (IVM) which allows the real time analysis of the complex behavior of neutrophils within the vessel wall. Deposition of IC within the vessel wall led to FcγR dependent slowing of the velocity of cells tethered to the endothelial selectin P-selectin, and increased subsequent adhesion and transmigration 88. Complement C3 played no role in these steps. TNF priming of neutrophils enhanced IC induced leukocyte recruitment 130, which may aid in localizing neutrophil influx to sites of IC inflammation. FcγRs also played an important role in neutrophil slow rolling and adhesion when neutrophils themselves were engaged by ANCA71.
Once neutrophils are recruited they can signal cytotoxic functions through FcγR and/or complement receptors that promote tissue damage. In vitro, engagement of FcγRIIA promotes phagocytosis, degranulation and ROS generation. FcγRIIIB cross-linking induces calcium mobilization and triggers degranulation and leukotriene release131, 132, but it has not been consistently shown to induce other neutrophil cytotoxic responses. Physiological evidence for FcγR's contribution to neutrophil dependent tissue damage has been challenging as protection from immune mediated disease in mice lacking activating FcγRs is associated with a reduction in neutrophil recruitment thus precluding conclusions about their role in neutrophil cytotoxicity. Recent work demonstrate that a deficiency in neutrophil Vav proteins, that signal downstream of activating FcγRs, confers resistance to the RPA in the lung and skin despite normal neutrophil accumulation in these tissues. The uncoupling of neutrophil accumulation and tissue injury in these models implicates FcγRs in neutrophil cytotoxicity in vivo133.
Given the structural variances in FcγRs between mice and humans and differences in their distribution in some cases, it is not clear how well studies mediated by the murine receptors alone accurately reflect human inflammation. For example, FcγRIIA is present on human macrophages and platelets while murine macrophages express FcγRIII and FcγRI and murine platelets are devoid of FcγRs. Recent studies have begun to address this gap in knowledge. A human FcγRIIA trangene under the control of its own promotor, was expressed in platelets and macrophages of mice 134. These mice developed enhanced autoimmune thrombocytopenia and collagen-induced arthritis (CIA) 135, 136 and spontaneous autoimmune disease likely as a result of an unexplained exuberant adaptive immune response 136. The retention of endogenous murine FcγRs, and thus an overall increase in the burden of FcγRs in these transgenic strains could contribute to the observed phenotype.
The expression of the uniquely human FcγRIIA and FcγRIIIB selectively on neutrophils of γ-chain deficient mice was undertaken to elucidate the physiological role of these two receptors in vivo. Neutrophil expression of both FcγRIIA and RIIIB was sufficient to restore susceptibility to the RPA and anti-GBM nephritis in Fcγ-chain deficient mice. This, in the absence of FcγRs in mast cells, platelets and macrophages, provided physiological evidence that the human FcγRs on neutrophils are sufficient for inducing antibody mediated tissue injury. Analysis of mice expressing one or the other human FcγR revealed that both FcγRIIIB and FcγRIIA promoted neutrophil recruitment, while only FcγRIIA induced macrophage accumulation and tissue injury 137. To further explore the role of these two receptors in IC induced neutrophil recruitment in vivo, and to determine whether they have evolved differential roles in this process, the interactions of neutrophils with the vessel wall were analyzed by intravital microscopy (IVM). In the context of strictly intravascular ICs in the cremaster muscle that was not associated with significant inflammation88, FcγRIIIB predominated in neutrophil slow rolling, adhesion and transmigration137. In contrast, FcγRIIA supported recruitment following induction of the RPA reaction in the cremaster137 which induces a complex inflammatory environment of intravascular and tissue ICs, activated endothelial cells and platelets, cytokines/chemokines and complement activation 89, 127, 138. These results suggest that human FcγRIIA and FcγRIIIB play context-dependent roles in IC-induced neutrophil recruitment.
What is the physiological role of FcγRIIIB mediated neutrophil recruitment? FcγRIIIB may be required for tissue clearance of ICs under homeostatic conditions (Figure 2). FcγRIIIB's presence on finger like projections on the neutrophil called microvilli 128, its fast mobility at the membrane 112 coupled with its weak capacity to signal tissue damage may suit it for tethering to and subsequently clearing intravascular ICs. When the FcγRIIIB mediated removal of IC is overwhelmed or defective, ICs may translocate and accumulate in extravascular tissue. Here they may persist and activate resident cells to release inflammatory mediators that lead to FcγRIIIB shedding, enhance FcγRIIA dependent recruitment and activity and thus instigate a cycle of vessel damage that predisposes to IC mediated disease (Figure 2). Indirect clinical evidence in support of FcγRIIIB's putative role in IC clearance comes from recent studies demonstrating that a low expression of FcγRIIIB, is associated with increased susceptibility to lupus nephritis and glomerulonephritis139-141.
What is the contribution of other well-described FcγR bearing cells such as macrophages and mast cells to IC induced disease progression? In the RPA and arthritis models, the role of mast cells is still being defined as studies on different mast cell deficient strains, namely the KitW/KitW-v and W/Wsh mice have reported contrasting results. KitW/KitW-v mice exhibit a partial to severe reduction in edema in the RPA 117, 142, 143 and fail to develop disease in the serum induced K/BxN model of arthritis 144. On the other hand, W/Wsh mice which are also profoundly mast cell deficient have a normal response in the arthritis model 145 and the RPA (N.T. and T.N.M, unpublished data). It is possible that the observed basal neutropenia in the KitW/KitW-v mice allows phenotypic complementation by mast cells or the recently described splenomegaly and expansion of the hematopoietic compartment (neutrophilia and thrombocytosis) in W/Wsh mice146 overshadows the need for mast cells in these models. One explanation for these results is that in the context of robust neutrophil responses/function, mast cells may not be required while in the face of neutrophil insufficiency, mast cells may enhance neutrophil IC induced functional responses by releasing neutrophil priming and activating agents90, 147.
Macrophages are required for the development of IC-mediated diseases. This is best illustrated by studies in trangenic mice with diphtheria toxin inducible depletion of macrophages. These mice exhibit reduced number of glomerular crescents and renal CD4+ T cells, markedly attenuated tubular injury, improved renal function, and reduced proteinuia following induction of anti-GBM nephritis 148. FcγRs on macrophages may only partially contribute to disease in this model as transgenic re-expression of the Fcγ-chain selectively in macrophages and monocytes of Fcγ-chain deficient mice only partially restored nephritis to wild-type levels following injection of anti-GBM antisera 149. In the human neutrophil FcγRIIA transgenic mice, anti-GBM induced neutrophil accumulation preceded interstitial macrophage accumulation and tissue injury suggesting that FcγR-bearing neutrophils signal macrophage recruitment into the kidney 137. Macrophages have the capacity to induce apoptotic cell death of renal parenchymal cells, and promote epithelial cell proliferation and interstitial fibrosis by modulating the balance of enzymes regulating matrix turnover and deposition, and the repertoire of inflammatory cytokines in the lesion148.
Patients with autoimmune diseases can exhibit accelerated atherosclerosis as well as advanced subclinical atherosclerosis that cannot be explained by traditional risk factors indicating that autoimmunity is an independent risk factor for development of cardiovascular disease (CVD). Studies in mouse models of autoimmune disease and atherosclerosis support this view. The inititators of autoimmunity are diverse but autoimmune diseases have in common pathways of immune-mediated injury. Neutrophils are present in atherosclerotic lesions, their peripheral counts are a predictor of future cardiovascular disease and the literature discussed herein and other recent reviews90, 106 strengthen the view that neutrophils and their opsonic receptors, FcγR and complement play fundamental roles in initiating and then perpetuating IC mediated immune responses. Studies of murine receptors have identified the FcγR and complement family of immunoreceptors as playing dominant roles in immune mediated inflammation in the context of a number of autoimmune diseases. Studies of mice engineered with human neutrophil FcγRs suggest the idea that FcγRs on these cells are the first sensors of ICs. Their apparent role in neutrophil recruitment, a proximal event in IC-induced inflammation could help answer one of the central questions in the field as to what triggers IC induced inflammation. It also infers that neutrophil FcγRs may be attractive therapeutic targets in IgG mediated inflammatory diseases. Many issues remain to be addressed before experimental findings can be translated to clinical medicine. First further clarification of the roles of the human FcγRs on neutrophils and other hematopoietic cell types in IC induced immunological responses is needed given the significant differences in structure and distribution of FcγRs between humans and other species. Furthermore, a more cogent understanding of how IgG-ICs and complement, and their respective receptors intersect to coordinate IC induced inflammation is needed.
If human neutrophil FcγRs prove to be an attractive therapeutic target then identification of molecular targets that negate FcγR function is required. Recent progress has been made in targeting a proximal signaling tyrosine kinase Syk that signals via ITAM containing receptors. The Syk inhibitor protects against experimentally induced arthritis 150, 151 and SLE 152 in mice, and its therapeutic benefit has been shown in clinical trials for human arthritis 153, 154. The inhibitor likely has its effects by targeting FcγR as well as a number of other Syk dependent receptors in hematopoietic and intrinsic tissue cells 115. Nonetheless, this type of approach is more specific than the immunosuppression regimens currently available for treatment of autoimmune disease which are largely non-selective with significant side effects. Cell-specific neutralization of human FcγRs or its downstream signaling partners would offer further specificity for potential therapeutics. For example, specific targeting of human neutrophil FcγRIIA would prevent autoimmune induced tissue damage while preserving the ability of FcγRIIA expressing macrophages to clear circulating ICs, which are constantly produced in autoimmune diseases. Moreover, specific targeting of FcγRIIA that spared FcγRIIIB would be desirable if FcγRIIIB proves to be important in clearance of intravascular ICs during homeostasis. These may be realistic goals given the recent advent of cell type specific silencing of proteins in leukocytes in vivo. Antibody to the hematopoietic restricted CD18 integrin, LFA-1 was used to deliver siRNA that silenced specific proteins in lymphocytes155, 156. The challenge lies in translating these approaches to neutrophils, which are short lived cells with a high turnover rate and a propensity for activation upon engagement of receptors/molecules on their surface. The opportunity is that if this obstacle is surmounted, therapeutics can be designed to eliminate specific neutrophil functions deleterious to tissue integrity in autoimmune disease while largely preserving inflammatory responses required for host defense.
Funding Sources Arthritis Foundation (NT) and NIH RO1 HL065095, AR050800 and DK077111(TNM).