The paradoxical functions of B cells and antibodies in transplantation¹

Abstract

Scarcely anyone would dispute that donor-specific B cells and the antibodies they produce can cause rejection of transplants. Less clear and more controversial however is the possibility that donor-specific B cells and the antibodies they produce might by one or more means protect transplants from injury. Here we review and discuss this later possibility and consider how less well-known functions of B cells and antibodies might impact on the design of therapeutics and the management of transplant recipients.

Introduction

B cells nearly always respond to transplantation. The antibodies produced by responding B cells provided the first clear evidence that transplantation provokes immunity (1) and that individuals inherit the capacity to respond or not to respond (1, 2). So reliable are B cell responses to transplantation that allo-specific antibodies produced during this response identified and mapped the genetic loci encoding the antigens the response targets (3). While antibodies produced by allo-specific B cells responding to transplantation provide the best markers of allo-immunity and histocompatibility (1), and while the potential of those antibodies for inflicting injury on transplants is now beyond dispute (4–6), that potential was once the subject of heated debate (7). This debate was fuelled in part by incomplete understanding of the molecular basis of immune recognition but it was also fuelled by observations, now nearly forgotten, that antibodies like regulatory T cells and certain cytokines, can inhibit some types of immunity and protect immunological targets. Here we consider this subject in a critical light, discussing how antibodies and B cells can modify immune responses and the susceptibility of tissues to cellular and humoral immunity, and the implications for therapeutics and clinical management of the recipients of transplants.

Impact of Antibodies on Transplants

The impact of antibodies on transplants depends to the greatest extent on the origin of the blood vessels feeding those transplants (8, 9). Organ transplants have donor-derived blood vessels that can be targeted directly by donor-specific antibodies. Binding of antibodies to foreign blood vessels generates hyperacute, antibody-mediated acute and chronic rejection, and accommodation (9). Cell and tissue grafts, however, are fed mainly by blood vessels of the recipient that grow into the grafts. These blood vessels are not targeted by donor-specific antibodies and indeed, in the absence of inflammation, the blood vessels block ready access of antibodies to the graft (9). Accordingly, cell and tissue grafts are not generally subject to hyperacute, antibody-mediated acute or chronic rejection (9). But, because activated T cells migrate efficiently through vessels, cell and tissue grafts are susceptible, and some believe more susceptible, to cellular rejection. Before the importance of anti-donor antibodies in the outcome of grafts was understood, donor specific antibodies were sometimes considered of little or no import in transplantation immunity. In classic reports, Mitchison (10), showed that transfer of donor-specific antibodies had little impact on allogeneic tumor grafts and Snell (11) found that these antibodies can actually “enhance” growth of these grafts. One happy by-product of this incomplete understanding however was the emergence (some would say re-emergence) of the appreciation of cells, as effectors of immunity (10).

If antibodies are now appreciated as effectors of organ transplant rejection (6), the availability of these antibodies and hence the susceptibility of a graft to antibody-mediated rejection is sometimes underestimated. Large organs, such as kidney and hearts absorb huge amounts of anti-donor antibodies (12–15) and in this way can clear the blood of all or nearly all of the antibodies as long as perfusion of the graft is unimpaired. Thus, the level of these antibodies in the blood increases dramatically following removal of an organ transplant (16) or severe graft injury (17, 18). Hence, while new technologies facilitate assay of donor specific antibodies (19) and while these assays are gaining wide application for analysis of pre- and post-transplant risk, one might exert some caution, if not skepticism, since after organ transplantation the antibodies of highest affinity and perhaps highest biological import will be absorbed in preference to antibodies of lower affinity (20). Also, because MHC class I is expressed in grafts more widely than MHC class II, and hence antibodies against MHC class I might be more fully absorbed than antibodies against MHC class II, one should exercise caution when interpreting reports that associate anti-MHC class II antibodies with graft rejection.

Accommodation and Enhancement

Antibodies against a transplant can protect the transplant from immunological injury by inducing accommodation, acquired resistance to immune and inflammatory injury (21–23) or enhancement, humoral suppression of immunity (11). Accommodation, although still a subject of controversy, has been the subject of recent critical reviews (23). Enhancement, although reviewed critically (7), and possibly less controversial, has been ignored in recent decades. Both merit consideration here because the processes challenge the notion that B cell responses invariably harm transplants.

Accommodation was first observed in the 1980s when some ABO-incompatible kidney transplants were found to function normally and suffer little or no injury in the face of high levels of circulating antibodies against donor blood group (21). Accommodation is associated with heightened expression of complement regulatory proteins and cytoprotective genes in the transplant (24–29). Whether either or both of these mediate accommodation is yet unclear, although absence of either dramatically heightens susceptibility of tissues to injury of every type (30, 31). The original reports of accommodation use donor-specific antibodies as markers (8); however, because organ transplants can absorb vast amounts of antibody, this marker lacks sensitivity. Expression of cytoprotective genes does appear to mark accommodation, but the high levels of expression found in rejecting organs undermines the specificity of these markers (31). The presence of C4d in normally functioning organ grafts might help identify accommodation in recipients having little or no donor-specific antibodies in the blood. Still more sensitive might be donor specific B cell responses, which we recently detected in many transplants (Lynch, Submitted 2012). The high prevalence of donor-specific B cell responses, if confirmed, would suggest accommodation could in fact be the most common outcome of organ transplantation.

The possibility that donor specific antibodies might suppress transplant immunity was first enunciated by Snell (11) who observed that administration of tumor vaccines sometimes “enhanced” rather than retarded cancer growth. Kaliss et al. (32), who studied the phenomenon most extensively, showed that antibodies mediate enhancement, since they can be used in lieu of vaccines and that antibodies do so by “blocking“ immunological recognition (7). One might expect that any suppressive properties of donor-specific antibodies would be eclipsed by humoral injury they cause to organ transplants. However, the survival of experimental and clinical organ transplants is sometimes improved and allo-immunity sometimes suppressed by evoking or administering anti-donor antibodies (7, 33, 34).

How exactly anti-donor B cell responses and anti-donor antibodies can suppress immunity is far from clear. Early investigation seemed to implicate blockade of antigen recognition and this mechanism has not been excluded. However, apart from the antibodies they produce, B cells can suppress immunity by secreting IL-10 and this function can control allo-immunity in mice (discussed below). Enhancement might also reflect antibody-dependent suppression of cellular immunity, which De Groot et al. (35) ascribe to activation of FoxP3-positive T regulatory cells by IgG derived peptides associated with HLA class II. Regardless of the mechanism, enhancement might account for the paradoxical intensification of cellular immunity associated with depletion of B cells in transplant recipients (36). That these mechanisms have not been implicated in enhancement probably reflects neglect of that subject more than contrary evidence.

Antibody-dependent T cell responses

Antibodies enhance antigen presentation to T cells. Binding of antibodies can concentrate antigen and direct to sites of antigen presentation. IgM captures antigen in blood, transports it to the spleen, and retains it in marginal zones (37). The antigen receptor of marginal zone B cells capture antigen and transports it to follicular dendritic cells in lymphoid follicles (38). Besides facilitating antigen presentation, Ig can directly promote cellular immunity. Peptides originating from Ig variable regions can associate with MHC and stimulate cellular immunity directed against IgG idiotypes originating from VH germline and mutated variants (39). T cellsalso can recognize peptides derived from mutated λ2 light chains (40) and recognition cangenerate delayed type hypersensitivity (41). Since transplants adsorb and process large amounts of donor specific antibody, this mechanism may explain why recipients producing such antibodies have a high incidence and greater severity of cellular rejection.

But, humoral immunity does not always enhance cellular immunity, it can also eradicate or regulate it. Thus, expression of a transgenic Ig κ light-chain caused deletion of κ light-chain peptide specific CD4-positive T cells (42), T cell tolerance to an Ig idiotype (Id+) prevented disease otherwise caused by Id+ T cells (43) and, induction of tolerance to an Ig idiotype prevented lupus in the NZB/NZW F1 mice (44). Besides causing deletion of idiotype specific T cells, Ig can induce regulation of those T cells (35). Whether and to which extent peptides from donor specific antibodies underlie regulation of T cell responses to transplants is not known but the emergence of B cell therapeutics heightens the importance of this subject.

B cell “dependent” T cell responses

Following the seminal observations of Mitchison establishing that immune cells rather than antibodies reject transplants (10), Szenberg (45) and others (46) showed that offending cells originate in the thymus and not in the bursa. Cell and tissue transplants were thus clearly subject to cellular and not antibody-mediated rejection, as transplants in B cell deficient mice confirmed (47).

However, the relationship between B cell functions and the outcome of transplantation would prove more complex than these results would suggest. First, as discussed above, it became apparent that rejection of organ transplants including cellular-mediated rejection, could in some circumstances, be promoted or in others be suppressed by B cells and/or antibodies. Organ transplant recipients with donor-specific antibodies more often experience cellular than antibody-mediated rejection (48). And, while no one disputes that the incidence of cellular mediated rejection in recipients with donor-specific antibodies might simply reflect prior sensitization, it is also possible that antibodies (and/or the B cells that produce the antibodies) facilitate cellular immune responses. Consistent with that concept, Sarwal et al. (49), Hippen et al. (50), Tsai et al. (51), and Zarkhin et al. (52) showed CD20-positive B cell clusters in C4d-negative biopsies of acutely rejecting kidney allografts suggesting that B cells may promote rejection perhaps by facilitating cellular immune responses. However, whether B cells in rejecting kidneys contribute to or merely mark rejection, is not known.

Depletion of B cells should offer clues to their function in transplantation. Antibodies against CD20, which is expressed on B cells but not on plasma cells, decrease the levels of donor-specific antibodies (53), improves outcome of transplants in clinical transplant recipients (54, 55), improves survival of islet allografts (56) and attenuates allograft vasculopathy in non-human primates (57). These findings might suggest that pathogenic antibodies against MHC might originate with partly differentiated plasma cells or memory B cells. Still, the failure of anti-CD20 to reduce or eliminate anti-HLA antibodies (58), suggests that other functions of B cells (59) and/or other sources of anti-HLA (plasma cells) might have a greater impact on transplant outcome.

While the impact of B cell depletion on antibody-mediated and chronic rejection has been studied in detail (60), the impact on cellular immunity and cellular rejection has not. Depletion of B cells at the time of transplantation in sensitized individuals decreased the frequency and severity of cellular rejection (53). Work in mice suggests at least some of this benefit might reflect hindered antigen presentation (61), although impact on anti-graft antibodies was not fully excluded. On the other hand, B cell depletion can also heighten risk and severity of cellular rejection. Five of six renal allograft recipients treated with anti-CD20 as an induction therapy were reported to develop acute cellular rejection (36). Memory B cells promoted unresponsiveness to co-stimulation blockade enhancing cellular rejection of heart allografts in mice (62) and, depletion of B cells in mice accelerated rejection of minor histocompatibility discordant skin grafts and major histocompatibility discordant kidney allografts (63).

How B cells control cellular immunity

B cells establish the microenvironments of lymphoid tissues that enable the mounting of cellularimmunity and function of adoptively transferred T cells. By providing lymphotoxin, B cells promote lymphoid organogenesis (64), differentiation of follicular dendritic cells (64), development of B cell follicles (65) and T cell zones in spleen (66), and differentiation of M cells in the gut associated lymphoid organs (67). The architecture of lymphoid tissues not only supports effector responses, it also supports regulation of cellular immune responses as established from the study of transplants in mice (68). As only one example, antibodies against CD62L, which impair T cell migration into lymph nodes, prevent establishment of transplant tolerance in mice with cardiac allografts (69). Whether depletion of B cells by agents such as anti-CD20 disrupts lymphoid architecture and in this way cellular immunity or regulation in the context of transplantation is unknown.

It is generally accepted that B cells present antigens. However, whether or not the type of T cell responses initiated by B cell antigen presentation is distinct from those initiated by conventional antigen presenting cells is not known. B cells present antigens like conventional APC following antigen phagocytosis or pinocytosis and may do so more efficiently than previously thought (70, 71). B cells also present antigen by a novel “cognate” pathway that might have particular importance in transplantation. Lanzavecchia (72) showed that B cells can, via their antigen receptor, take up antigen which is then processed and presented to MHC class II restricted T cells. B cell cognate antigen presentation differs from conventional antigen presentation owing to B cells’ unique ability to take up antigen specifically through binding clonotypic B cell receptors which, in turn, concentrates presentation of peptides from one antigen (73). Also, in contrast to conventional antigen presenting cells which do not divide, B cells proliferate forming clones, amplifying presentation of the antigen uptake by the B cell receptor. B cell cognate antigen presentation might be important in transplantation because the nearly universal B cell response to MHC infers that every individual has some, perhaps many, B cells capable of cognate presentation of MHC and perhaps some minor antigens; it might also explain how cellular immunity and cellular rejection frequently occur in spite of potent immunosuppression. The importance of cognate presentation in transplantation would seem ripe for study, given therecent attention devoted to depletion of B cells.

Besides establishing lymphoid microenvironments that nurture immune activation and regulation, B cells directly control immunity, including transplantation immunity. B cell control immunity in part by secreting IL-2, IL-4, IL-10, IL-13, IFNγ, IL-12 and TNFα (74, 75). B cells also promote differentiation of regulatory T cells (76–78). Excellent reviews of the properties of B regulatory B cells have been recently published (75, 79). Consistent with a regulatory function of B cells in transplantation, B cell genes are expressed more frequently in tolerant than in non-tolerant renal transplant recipients of renal allografts (80, 81), tolerant kidney transplant recipients have large numbers of B cells producing IL-10 (82) and tolerance to experimental cardiac allografts in rats can be transferred by a purified population of B cells (83). As mentioned above, an increased risk of acute cellular rejection following depletion of B cells prior to transplantation, could be due to loss of regulatory B cells (36).

In addition to regulating the response of T cells to transplantation and other stimuli, B cells, and particularly diverse Ig, expand the repertoire of T cells (84). This property of B cells may be particularly relevant in transplantation since transplantation introduces powerful immunogens (like MHC) which cannot be cleared, inducing chronic immune activation, biasing the lymphocyte repertoire (85). And, while rejection of transplants does not require T cell diversity (86), regulation of allo-immunity might (86).

Conclusion

Recent years have brought heightened appreciation that donor-specific B cells pose a key barrier to successful transplantation, especially of organs. This appreciation has naturally sparked interest in development of agents for depletion and suppression of B cells. Yet, this interest also heightens the importance if not appreciation of paradoxical functions of B cells in transplantation. No less important than effector B cell functions, might be those B cell functions that prevent or suppress immunity against the graft or protect the graft against injury. Enhancement, accommodation and tolerance all reflect functions of B cells and all are potentially compromised by therapeutics aimed at B cells. That is not to say that B cell therapeutics should be put aside but rather that a broader view of the outcomes might be needed. We consider especially urgent the testing of whether anti-CD20 or other such agents modify lymphoid architecture or the numbers and functions of regulatory T cells. For example B cell depletion therapies may originate structural and/or functional defects in lymphoid tissue which in turn, may limit their efficacy or decrease immune responses to vaccines. Equally important would be probing of whether these agents impair accommodation, enhancement and/or spontaneous tolerance. New, more sensitive assays for donor-specific antibodies should allow more attention to be devoted to learning when and these antibodies prevent graft injury.