|Home | About | Journals | Submit | Contact Us | Français|
In the last decade, two advances have shifted attention from cellular rejection to antibody-mediated rejection (AMR) of cardiac transplants. First, more sensitive diagnostic tests for detection of AMR have been developed. Second, improvements in immunosuppression have made severe acute cellular rejection uncommon, but have had less effect on AMR. Antibodies can contribute to graft rejection by activation of complement, by activation of vascular endothelial and smooth muscle cells, and by activation of neutrophils, macrophages or natural killer cells. Because acute rejection is a risk factor for chronic rejection in all types of organ transplants, it is has been proposed that AMR can cause chronic rejection. Small animal models need to be developed to gain further insights into AMR and the role of antibodies in chronic graft arteriopathy. This article reviews the current clinical data and existing mouse models for AMR.
Small animal models have been instrumental in immunological advances of relevance to cardiac transplantation. Two major examples are the discovery of immunological tolerance in mice by Billingham, Brent and Medawar (1) that set the ultimate goal for all organ transplants, and the demonstration of the powerful immunosuppressive effects of cyclosporine on transplants that revived cardiac transplantation as a realistic clinical treatment (2). Continued advances in immunosuppression have made severe acute cellular rejection uncommon. This has revealed another form of rejection, namely antibody-mediated rejection (AMR), that is resistant to current immunosuppressive therapy. Although there is now general agreement that AMR is real and potentially lethal, major gaps remain in our understanding of this form of rejection, including the incidence, risk factors, diagnostic criteria, contributory mediators, tissue responses and possible chronic sequellae. This last question is of greatest concern because chronic arteriopathy remains the major barrier to long-term survival of cardiac transplants. Advanced studies with intravascular ultrasound (IVUS) of coronary arteries have demonstrated increased intimal thickness, one of the features that define chronic arteriopathy, in about half of cardiac transplants within one year after transplantation (3). Unfortunately, current small animal models have been of limited value in examining either AMR or the role of antibodies in chronic graft arteriopathy. In this article, we will review the clinical data and critically evaluate the existing mouse models.
The largest clinical experience with AMR has been in renal allografts. Criteria for acute AMR in renal transplants were published in 2003 (4). These include circulating antibodies to donor MHC antigens, diffuse deposition of the complement split product C4d in peritubular capillaries, and morphologic evidence of acute tissue injury, such as margination of macrophages in capillaries. Using these criteria, AMR was diagnosed in 3–6% of unsensitized patients (5), and most sensitized patients (6, 7). The diagnostic interpretation of C4d deposits and marginated macrophages in cardiac transplants is still debated. However, several large studies indicate that C4d is associated with donor specific antibodies and an increase risk of rejection (8–11). What additional criteria are needed to improve the specificity and sensitivity of this marker is the subject of investigation in many centers.
C4d is a product of the initial steps of the classical and lectin pathways of complement activation (Fig 1). A single C1 molecule bound to a pair of antibodies can cleave many C4 molecules. C4b, the larger split product of C4, has the unusual capacity of forming a covalent bond with nearby proteins or carbohydrates. When C4b binds to endothelial cells, it is quickly cleaved to the smaller biologically inactive C4d. This end product of C4 activation is easy to detect because it is deposited in larger quantities than antibody and it has a longer half-life (12). However, C1 and the C4 split products have limited proinflammatory effects compared to the subsequent complement components, most importantly the split products of C3 and C5. Activation of C3 and C5 produces the soluble chemotactic split products C3a and C5a in addition to the larger C3b and C5b. C3b, like C4b, can bind covalently to tissue where, in the process of regulation, it is cleaved first to iC3b and then to C3d. C5b is the first component of the membrane attack complex (MAC) that is formed by the terminal complement components. Of relevance to AMR, neutrophils and macrophages have complement receptors for C3b (CR1; CD35) and iC3b (CR3; CD11b/CD18). B cells have receptors for C3d (CR2 or CD21).
Several regulators of complement activation can terminate the complement cascade before C3 is cleaved. These include membrane bound Decay Accelerating Factor (DAF; CD 55) and membrane cofactor protein (MCP; CD46), a cofactor for factor I. DAF dissociates the C3 convertase formed by C4b and C2a as well as the subsequent C5 convertase, and MCP catalyzes cleavage of C4b and C3b to C4d and C3d, respectively. Therefore, it has be proposed that effective complement regulation of C4 may result in C4d deposits with little inflammatory response and no impairment of graft function (13, 14). A disconnection of C4d from graft injury is found in some patients, who receive major blood group incompatible renal transplants under treatment with plasmapheresis and intravenous immunoglobulin (7). The term “accommodation” has been coined for this phenomenon (13). Whether accommodation occurs in cardiac transplants remains to be determined, but data have been presented to support the use of C3d in addition to C4d as a pathological marker of AMR (9, 11). This raises the further possibility that upregulation of DAF, MCP or CD59 (a regulator of MAC) may cause certain cardiac transplants to be more resistant to AMR (11). Although it is not known what mechanisms control the expression of complement regulators in clinical cardiac transplants, many mediators that have been implicated in acute and chronic responses to transplants regulate DAF expression by endothelial cells in vitro. Mediators of acute inflammation, such as thrombin, TNFα and IFNγ, increase DAF expression but not CD59 or MCP expression (15, 16). The effects of cytokines on DAF expression are augmented in endothelial cells with MAC deposited on their surface. MAC upregulation of DAF expression may represent a feedback mechanism to protect vascular integrity in sites of complement activation. Mediators associated with chronic inflammation, such as basic FGF and VEGF also upregulate the expression of DAF. The finding that cyclosporine A inhibits the upregulation of DAF expression suggests that some of the mechanisms that protect tissues from complement activation may be inhibited in transplant patients (17).
In addition to activation of complement, antibodies can modulate inflammatory responses in transplants by direct activation of vascular endothelial and smooth muscle cells as a consequence of cross-linking MHC antigens (18–20) and by activation of neutrophils, macrophages or natural killer cells through Fc receptors (21). Extensive studies by Reed and co-workers have demonstrated that even in the absence of leukocytes and complement, binding of antibodies to MHC class I antigens on endothelial cells or smooth muscle cells causes release of growth factors, upregulation of receptors and cell proliferation (18). Importantly, these proliferative responses were blocked by the mTOR inhibitor rapamycin (20). Lowenstein and his colleagues have demonstrated that antibody-mediated cross-linking of MHC class I antigens on endothelial cells can also induce exocytosis of von Willebrand factor (vWf) and P-selectin from Weibel-Palade storage granules (19). Although most of these experiments were performed on isolated endothelial cells in vitro without complement, purified components of MAC can induce endothelial cell proliferation (22) and exocytosis of vWf and P-selectin (23). Undoubtedly, these mechanisms do not occur independently in vivo, but rather interact to modulate inflammation. In vivo, non-complement activating antibodies induce the exocytosis of vWf and P-selectin followed by adhesion of platelets and neutrophils to capillaries in skin grafts, but complement activating antibodies have a more prolonged effect (24). Similarly, antibodies alone can cause endothelial cells to produce cytokines such as IL-6 and MCP-1, but increased production of these cytokines occurs when antibodies also activate macrophages through Fc receptors (21). These cytokine responses are likely enhanced further in transplanted organs by complement because macrophages express an array of receptors for complement split products.
During immune responses, complement and antibody participate in feedback loops to regulate antibody production by B cells. C3d binds to CD21, which acts as a co-receptor for the membrane Ig (mIg) antigen receptor on B cells. C3d bound to antigen has been found to increase the antibody response of B cells by lowering the threshold of concentration or affinity for an antigen to cross-link mIg (25). Antibodies bound to antigens can deliver a counterbalancing signal to B cells through the FcγRIIB for IgG. This receptor delivers an inhibitory signal to B cells and functions as a negative feedback for continued antibody production (26).
Knowledge of the mechanisms of antibody-mediated immune responses has provided insights for potential therapeutic interventions. For example, the finding that MHC cross-linking by antibody involves the mTOR pathway has obvious implications for the use of rapamycin for immunosuppression (20). Understanding feedback regulation through FcgRIIB on B cells may lead to improvements in formulations of intravenous immunoglobulin (IVIg) to inhibit alloantibody production. Ravetch and colleagues have demonstrated that the affinity of Ig binding to FcgRIIB is dependent on the carbohydrate side chain on the IgG (26). Therefore, calculated modifications of the carbohydrate side chain may increase the inhibitory effect of IVIg on B cells. The complement cascade is also a therapeutic target. Monoclonal antibodies to C5 have been used to inhibit antibody-mediated rejection (27, 28). Preventing the cleavage of C5 into C5a and C5b blocks two proinflammatory links. C5a chemoattracts neutrophils and macrophages. It also activates endothelial cells as well as neutrophils and macrophages. C5b initiates the formation of MAC. Therefore, monoclonal antibodies to C5 can inhibit leukocyte and endothelial cell activation.
Chronic rejection is the tissue response to the cumulative injury to a transplant over time. This includes the innate and adaptive immune responses initiated before transplantation by brain death to the donor, at the time of transplantation by ischemia-reperfusion, and after transplantation by subclinical as well as clinically apparent rejection episodes. Arteries are the most universally affected structure in all types of organ transplants. In cardiac transplants, arteriopathy of the conductance arteries is the major finding. Careful studies at the level of capillaries demonstrate a decrease in numbers (“capillary dropout”). This suggests that the capacitance vessels are also a target of the rejection process. These changes are eventually accompanied by parenchymal fibrosis.
Chronic graft arteriopathy, also known as graft vasculopathy, is characterized in cardiac transplants by a diffuse and concentric narrowing of the coronary arteries as a result of neointimal expansion that is often accompanied by adventitial fibrosis. The intimal expansion initially includes an inflammatory cell component of macrophages and T cells that then is replaced by mesenchymal cells many of which develop characteristics of smooth muscle cells. Initially, all of the smooth muscle cells were thought to originate from the media, but more recently two additional sources of these cells have been identified: circulating recipient stem cells and donor or recipient endothelial cells. In small animal models that use suboptimal levels of immunosuppression, vascular injury is extensive and recipient cells may be the major source of both endothelial and smooth muscle cells in the arterial lesions (29). However, careful studies in human cardiac transplants indicate that only about 5–10% of the endothelial cells and 2–4% of the smooth muscle cells in larger arteries are of recipient origin (30). Endothelial cell to mesenchymal cell transition also may occur under the influence of TGFβ (31). Outward expansion of the artery to compensate for the thickened intima is prevented by inflammation leading to fibrosis of the adventitia. This results in decreased blood flow and ischemic changes.
Graft arteriopathy is detected clinically by radiological techniques. The International Society for Heart and Lung Transplantation (ISHLT) Registry reported that the incidence of graft arteriopathy is 7% at one year, 32% at five years, and 53% at ten years after cardiac transplantation for patients followed between 2002–2006 (32). These data are based on standard coronary angiography. Studies using more sensitive intravascular ultrasound (IVUS) detect new arteriopathy in about half of cardiac transplants within one year after transplantation (3).
Because acute rejection is a risk factor for chronic rejection in all types of organ transplants, it is has been proposed that AMR can cause chronic rejection of renal transplants (33). The infrequent demonstration of C4d in renal transplant biopsies with evidence of chronic rejection has been ascribed to low levels of antibody or repeated cycles of antibody production. This is an attractive concept for cardiac transplants because antibodies and complement cause vasculocentric pathology. Moreover, many of the cells (macrophages, platelets, endothelial cells, smooth muscle cells, fibroblasts) and growth factors (basic fibroblast growth factor, platelet derived growth factor) associated with vasculopathy in transplants are stimulated by antibodies and complement (reviewed in (34). However, direct clinical evidence for C4d in arteries of cardiac transplants is limited because large arteries are not contained in diagnostic endomyocardial biopsies. In addition, few patients have circulating antibodies to donor MHC antigens or C4d deposition on the graft at the time of explantation or death.
DNA microarray analysis provided unexpected evidence for local production of antibodies in graft arteriopathy. Microarray studies comparing the gene expression of coronaries with dilated cardiomyopathy, atherosclerosis and chronic graft arteriopathy revealed that coronaries with chronic arteriopathy had an upregulation of immunoglobulin genes including immunoglobulin heavy chain as well as kappa and lambda light chains (34). These findings correlated with histological analyses of cardiac allografts explanted as a result of chronic arteriopathy. B cells and plasma cells were frequently found in clusters within and surrounding coronary arteries with chronic arteriopathy (Fig 2). Infiltrating plasma cells contained kappa and lambda light chains as well as all IgG subclasses.
In addition, nodules of mononuclear cells with compartmentalized T and B cell regions consistent with tertiary lymphoid neogenesis were present in approximately half of the transplants studied. Tertiary lymphoid nodules are recognized components of chronic inflammatory diseases where they are a source of local and long-lived responses to autoantigens in chronic thyroiditis and rheumatoid arthritis (35, 36). B cell nodules have been frequently described in renal transplants, but their contribution to acute and chronic rejection is controversial (37).
In addition to nodules, diffuse infiltrates of B cells and plasma cells were found in the adventitia of coronaries with graft arteriosclerosis. These adventitial infiltrates were often embedded in fibrosis. Mengel and colleagues (38) have reported that transcripts associated with B cells and plasma cells are a signature of scarring in renal transplants. Recent data from experimental models indicate that B cells can promote fibrosis. For example, Sato and Tedder and co-workers (39, 40) have demonstrated that bleomycin induces less fibrosis in skin and lung of CD19 deficient mice. In these experiments, CD19 deficiency also resulted in recruitment of fewer T cells, macrophages and mast cells to sites of bleomycin administration. This is of relevance because Mengel and colleagues reported that transcripts associated with mast cells were typical of fibrotic areas in renal biopsies (38). These observations suggest a potential new role for B cells in fibrogenesis that would be fruitful to explore in animal models.
Experimental models that were often too reductionist to be clinically relevant convinced many basic immunologists and clinicians that mice and rats were not appropriate animals for studying AMR. Most influential were early experimental models, in which either serum or leukocytes were transferred from sensitized mice to naïve transplant recipients. These experiments demonstrated that lymphocytes could cause rejection in the absence of antibodies. Appreciation of the potential proinflammatory effects of antibodies in transplants was eroded further by experiments demonstrating that passive transfer of antibodies often prolonged graft survival in rats and mice. The opinion evolved that rats and mice had “weak” complement systems. Although some strains of mice have complement deficiencies (eg, B10.D2 old, DBA/2, AKR and A/J are C5 deficient), most mouse strains are not complement deficient. Reagents have now been developed to detect C4d and C3d in rats and mice (Fig 3). These reagents have demonstrated that rejection of cardiac transplants is frequently associated with alloantibody production and C4d and C3d deposition (12, 41–44). C4d and C3d deposition is absent in cardiac allografts in immunoglobulin knock out mice and early injury to vascular endothelial cells is minimized. In some strain combinations acute rejection of MHC incompatible is also delayed in immunoglobulin knock out recipients. In these models, passive transfer of complement activating monoclonal antibodies to the donor MHC antigens restores complement deposition and accelerates graft rejection (45, 46).
Induction of AMR in T and B cell immunodeficient mice (RAG−/−) by passive transfer of antibodies to donor MHC antigens has been difficult to achieve. Nozaki et al (44) concluded that titer of antibody is critical to successful passive transfer of AMR to RAG−/− recipients. These investigators found that wild type recipients of MHC incompatible cardiac transplants do not produce high enough titers of alloantibodies to cause AMR in passive transfer studies with RAG−/− recipients. In contrast, CCR5−/− mice produce titers of alloantibodies to MHC mismatched cardiac allografts that are 15- to 25- fold higher than their wild type counterparts. Undiluted sera from CCR5−/− mice containing these higher titers of alloantibodies do cause AMR in RAG−/− recipients, but AMR does not result when the sera are diluted to wild type levels.
Attempts to cause AMR in RAG−/− recipients with monoclonal antibodies to MHC antigens on the transplanted heart have generally produced little pathology even though they cause C4d deposition on capillaries (47, 48). The use of monoclonal antibodies limits the number of epitopes recognized by the antibodies and results in lower densities of antibodies arrayed on the vessels. Whether passively transferred antibodies that cause C4d deposition in the absence of graft rejection induce graft accommodation in mice has not been explored.
The complete absence of T cells in RAG−/− recipients may curtail the effects of antibodies because it is likely that antibodies and T cells have synergistic interactions. Platelets are one potential link between antibodies and T cells. Morrell et al (24) have demonstrated that antibodies to MHC antigens on endothelial cells of skin grafts cause the release of vWf and adhesion of platelets to blood vessels. Activated platelets express and secrete many factors that can localize T cells to an allograft (49, 50). Release of small vasoactive molecules like adenine nucleotides (ADP and ATP), ionized calcium, histamine, serotonin and epinephrine from dense granules of activated platelets can promote leukocyte infiltration and stimulate vasoconstriction and leukocyte diapedesis. The alpha granules of platelets contain several chemokines that attract and activate macrophages and T cells, such as MIP-1α (CCL3), RANTES (CCL5), MCP-3 (CCL7) and PF4 (CXCL4). After contact, macrophages and T cells interact with platelets through receptor/ligand pairs, including P-selectin/PSGL-1 and CD40/CD154. In the converse direction, upregulation of MHC antigens on endothelial cells by IFNγ and other cytokines secreted by T cells likely makes antibodies more effective proinflammatory mediators. These mechanisms may account for the observation by Sis et al (51) that some biopsies with genomic profiles characteristic of activated endothelial cells associated with AMR also have upregulated genes for various T cell profiles.
The use of mice to determine whether alloantibodies can contribute to chronic arteriopathy has also had limited success. Again, the use of polyclonal antibodies apparently is more effective than monoclonal antibodies. Russell and colleagues reported that passive transfer of polyclonal antibodies to SCID mice resulted in extensive arteriopathy (52). However, the same group reported that passive transfer of monoclonal antibodies to donor MHC antigens only causes small lesions at the root of the coronary arteries of cardiac transplants in RAG−/− recipients (47). More widespread neointimal formation with a prominent smooth muscle cell component can be produced in cardiac allografts to rats. Both acute and chronic rejection of cardiac allografts is delayed in rats by a deficiency in C6, one of the terminal complement components that forms MAC (41, 42, 53).
Significant anatomical differences exist between rodent and human coronary arteries that may contribute to the apparent resistance to arteriopathy in cardiac transplants in rodents. First, the large coronary arteries in humans are located on the surface of the heart surrounded by epicardial fat. In contrast only the first few millimeters of the coronary arteries adjacent to the aortic root are on the surface of the heart in rodents. At that point the coronary arteries penetrate the myocardium. This is relevant to the development of arteriopathy because epicardial fat produces proinflammatory cytokines including IL-6, IL-8 and MCP-1 (54). The second difference is that the adventitia of the larger human coronary arteries contains vasa vasorum (55). These vessels have been implicated in the development of atherosclerosis and it is possible that they provide an “outside-in” route for the migration of inflammatory cells and cytokines into coronaries of cardiac transplants in humans but not in rodents. In addition, adult human hearts often have superimposed abnormalities before transplantation. This is most evident by the incidence of subclinical amounts of atherosclerosis found in IVUS studies (3). Pre-existing atheroma is an indicator of more global underlying aging alterations. Moreover, these organs are often transplanted into patients who have had previous surgery (e.g., repairs of congenital abnormalities, replacement of valves, application of coronary bypass grafts, and insertion of ventricular assist devices) with ensuing inflammation, angiogenesis and scarring. In many patients, this local inflammatory environment is compounded by systemic sensitization as the result of transfusions received for the surgical procedure. In contrast to humans, the animals used as donors and recipients are usually young and healthy.
In summary, cell-mediated rejection of cardiac transplants has decreased as immunosuppressive treatment has improved. AMR and chronic arteriopathy are important causes of mortality in cardiac transplant recipients. Although diagnostic criteria have been established for acute AMR in renal transplants, much remains to be learned about AMR in cardiac transplants. Likewise, the pathogenesis of chronic arteriopathy in cardiac allografts is largely unknown. Some small animal models for acute AMR and chronic arteriopathy have been developed, but many models present perplexing idiosyncrasies.
The authors thank Karen Fox-Talbot and Nina Volokh for excellent immunohistological stains. RLF is supported by grants from the National Institutes of Allergy and Infectious Diseases (AI40459 and AI51620), and from the Roche Organ Transplant Research Foundation (#60495086). CNM is supported by NIH grants R01HL093179 and HL94547.