Clinical Xenotransplantation: Pigs Might Fly?

Authors


Abstract

Xenotransplantation has the potential to deliver an unlimited supply of organs for transplantation. However, this promise has yet to translate into clinical application, despite substantial research efforts in the last decade. Although increasing numbers of studies are being performed in relevant pre-clinical (pig-to-primate) transplantation models, so far these have highlighted the apparent elusiveness of long-term xenograft survival. Humoral rejection remains the main obstacle to success, but control of T cell-mediated rejection will be a problem in the future and there are major concerns about the possible transmission of porcine endogenous retroviruses (PERV) and other infectious agents. This article reviews recent advances in the understanding of acute vascular rejection (AVR), acute T cell-mediated rejection and PERV transmission and highlights some of the strategies that may prove successful in overcoming these problems. Although progress has been slow, the promise of an inexhaustible supply of organs is sufficient reason to continue research in these areas. Assuming the specific problem of AVR can be ameliorated by one of a number of strategies currently under investigation, there are grounds to believe that xenotransplantation will become a clinical reality. Pig xenografts, currently grounded, might eventually fly!

Introduction

This is a time of relative uncertainty in the xenotransplantation field, following slow progress in overcoming xenograft rejection and a revision of strategy by some of the pharmaceutical companies providing significant research funds. As such, it is an opportune moment to review recent progress and ask whether clinical xenotransplantation will ever deliver on the promise of an unlimited supply of organs or tissues for transplantation. Paraphrasing a well-known line by Lewis Carroll, ‘The time has come to talk of whether pig xenografts will fly’.

Recent Advances and Current Uncertainties

Although development has been slow in the last few years, much is now known of the separate elements that mediate early xenograft rejection. A wealth of data from small animal transplantation models continues to provide valuable insights into xenograft rejection, but a significant number of experiments in the last few years have been performed using pig-to-primate combinations, allowing assessment of preclinical strategies to improve xenograft survival. Figure 1 illustrates the sequential immunological hurdles confronting solid organ xenografts and lists potential clinical strategies to overcome these.

Figure 1.

Sequential stages of xenograft rejection.

A. Acute vascular rejection (AVR)

Acute vascular rejection (AVR) (also called delayed xenograft rejection [DXR] or acute humoral xenograft rejection [AHXR]) has been the major focus of research for several years since the development of transgenic pigs expressing human regulators of complement activity (RCA), organs from which were shown to be resistant to hyperacute rejection (HAR). Two advances in our understanding of AVR have occurred in recent years, each with therapeutic implications. The first concerns the role that complement plays in the rejection of organs from RCA-transgenic pigs.

The original descriptions of AVR in the pig-to-primate model contained descriptions of two separate pathophysiological mechanisms. Pruitt, for example, defined complement-dependent AVR after using soluble complement receptor 1 to inhibit the HAR of pig hearts transplanted into Cynomolgus monkeyseys(1). Rejection was associated with increasing complement activity in the serum of recipient monkeys and with deposition of complement, alongside xenoreactive natural antibodies (XNA), in the graft. In contrast, Leventhal defined a complement-independent AVR after using cobra venom factor (CVF) to prevent the HAR of porcine hearts transplanted into primates. These grafts were rejected by a type of vascular rejection with deposition of XNA but no complement and no evidence of systemic complement activity (2).

With the advent of pigs transgenic for human CD55 (3,4), it was assumed by many working in the field that these organs would suffer AVR that was complement independent. However, new studies have shown this not to be the case. For instance, AVR of these organs after transplantation into primates is associated with endothelial cell deposition of terminal complement components (5,6) and systemic complement inhibitors have been shown to enhance survival (7,8). These data show that organs from these pigs are not completely resistant to primate complement, thereby highlighting the conditions under which they can be used to study the pathophysiology of complement-independent AVR in primates.

The second major advance has been understanding the role of antigraft (in particular anti-gal) antibodies in AVR. Xenoreactive natural antibodies are capable of mediating, in vitro, changes in endothelial cells that are consistent with a role in AVR (9–11). In vivo, there is no evidence from experiments using pig-to-primate grafts that complement alone, in the absence of antigraft antibody, can mediate AVR. Instead, AVR is always associated with antibody deposition in the graft, even if (following depletion of pre-existing antigraft XNA) rejection occurs when antigraft antibodies are circulating at very low or undetectable levels. In addition, depleting preformed XNA prior to transplantation usually protects against AVR (12). Taken together, these data strongly suggest a pathogenic role for antigraft antibody in AVR.

However, adequate protection against vascular rejection can only be guaranteed if elicited antigraft antibody production (stimulated by the immune response to the graft) is significantly suppressed (13), something which is currently possible only by deploying severely toxic immunosuppressive regimens. In practice, this means that inhibition of AVR using conventional immunosuppressive protocols is still associated with an extremely high morbidity and mortality.

A significant step forward was achieved with the recent finding that the antibodies mediating AVR are predominantly specific for the galα(1–3)gal epitope, as in HAR (14). This finding was surprising as a substantial proportion of the IgG XNA in primates is specific for epitopes other than galα(1–3)gal (15,16).

This recent discovery raises a hope that organs from cloned pigs lacking the galactosyltransferase enzyme may prove relatively resistant to AVR. Such pigs have been made possible by advances in nuclear transfer technology (17) and, recently, two independent groups have reported the generation of founder lines of gal-knockout (KO) pigs (See http://www.ppl-therapeutics.com/html/cfml/index_fullstory.cfm?StoryID=50 and Lai et al., http://www.sciencemag.org/cgi/content/abstract/1068228v1).

Transplantation studies using organs from homozygous gal-KO pigs should be possible in the near future, to test the hypothesis that these are resistant to AVR. An assessment of the vigor of the humoral immune response to these grafts will be important, for it is highly likely that they will stimulate an antigraft antibody response against epitopes other than galα(1–3)gal. Whether these antibodies will precipitate AVR or whether their production will be easier to suppress than anti-gal antibodies will be of particular interest.

Two other areas of research have proved fruitful in recent years. The first concerns the phenomenon of accommodation. This is defined as allo- or xenograft survival despite detectable antigraft antibody, either circulating or deposited in the graft, a situation that would normally precipitate vascular rejection (18). Accommodation is now known to involve the acquisition of a ‘resistance to injury’, by the cells making up the vasculature of the xenograft, involving the expression of ‘survival-gene’ products such as Bcl-2, A20 and crucially, the antioxidant protein hemoxygenase-1 (HO-1), which mediates the production of carbon monoxide (CO) (19). The association between organ survival and expression of HO-1 and CO is being increasingly appreciated [for instance, see (20)], and an understanding of the molecular mechanisms underpinning the protection provided by these two molecules may lead eventually to the development of therapies to prevent vascular rejection.

The second area of research concerns the role that coagulation and coagulation factors play in AVR (21). In vivo, intragraft thrombosis is now understood to play a central role in the early pathophysiology of AVR, rather than being a terminal event as had been assumed (22). Also, several groups have recently documented the systemic coagulation abnormalities that occur as part of AVR (23,24), the predominant feature being a consumptive coagulopathy initiated by the graft itself. The implication of this work is that adjunctive therapy directed at preventing clotting after xenotransplantation may have a potential to improve graft survival and reduce the immunosuppressive burden.

An important area for future research concerns the question of whether the clotting abnormalities in AVR are reliant or independent of deposition of antigraft antibody. This is relevant as porcine cells in vitro have an inherent tendency to spontaneously clot human plasma (25), an effect that appears dependent on certain molecular incompatibilities between human and porcine regulators of coagulation (26). If clotting abnormalities occur independently of antibody, this may be one reason why AVR in pig-to-primate models has been so difficult to prevent or treat.

An indication that clotting abnormalities may arise independently of antigraft antibody has come from experiments with porcine islets. When injected intraportally, they suffer an immediate humoral rejection response which severely limits functional integrity (27). This appears to be a different process to that mediating classical AVR of solid organs, at least when adult pig islets are used (islets from fetal pigs bind anti-gal antibody whereas those from adults do not (28), indicating that anti-gal antibodies play no role in the humoral rejection of adult islets after IV injection), as it involves complement and clotting factors but not anti-pig XNA. Further work is needed to clarify the pathophysiology of this problem, particularly since it also appears to affect allogeneic islets after exposure to whole blood (27).

B. T-cell mediated rejection

Some in vivo studies, using closely related species [for instance after transplantation of islets from one species of primate into another (29)] have shown prolonged inhibition of the T-cell response using near-conventional immunosuppressive regimens. However, these studies give no insight into the efficacy of these regimens after transplantation of tissue from more distantly related species, such as the pig. Instead, the evidence suggests that conventional immunosuppressive regimes will be relatively impotent at preventing T-cell-mediated rejection of porcine xenografts (30,31), primarily because the human T-cell response to porcine cells is very vigorous compared to allogeneic cells (32). Alternative therapies will therefore be needed for successful clinical application, if the side-effects of high-dose systemic immunosuppression are to be avoided in xenograft recipients.

Although few such alternatives have evolved in clinical allotransplantation, despite the description of multiple experimental strategies to promote graft survival, they might be easier to develop in xenotransplantation, for the following reasons.

First, with adequate planning, there should be time to manipulate the recipient prior to transplantation to influence antidonor T-cell responsiveness and, with an unlimited supply of donor tissue [with major histocompatibility complex (MHC) in-bred donors], to test the efficacy of this manipulation. Moreover, there is the opportunity to manipulate the xenogeneic donor to reduce organ immunogenicity or promote tolerance over sensitization. Finally, the fundamental differences between species may be advantageous, allowing generation of tolerogenic reagents with true graft specificity. To illustrate the final point, my group has recently described porcine CD152-Ig, which preferentially binds pig rather than human (or murine) CD80/86 over a wide range of concentrations (33). This reagent is currently being evaluated in a small animal model to determine if selective binding translates into graft-specific suppression of T-cell responses after xenotransplantation.

Multiple efforts are being made to explore strategies for induction of peripheral T-cell tolerance. Our group has had recent success with a novel graft-specific strategy targeting the interaction between porcine CD80/86 and murine CD28, to interfere with T-cell activation and promote islet xenograft survival (34), a strategy which in theory could be applied directly into clinical practice, but which, practically speaking is many years away from clinical application.

A strategy with more promise for clinical application is tolerance induction through mixed hematopoietic chimerism. This is an effective way of promoting tolerance to transplanted grafts (35) and has already been used clinically to promote immunosuppression-free renal allograft survival (36). This strategy has particular promise in xenotransplantation as B-cell tolerance is also induced, meaning that AVR may be effectively abolished alongside T-cell antigraft responses (37,38).

However, despite progress in small animal xenogeneic models using nonmyeloablative conditioning regimens (39), there are significant problems hindering progress in models involving phylogenetically distant species (40), especially with xenogeneic marrow engraftment, and with a humoral response against the injected xenogeneic material (41,42). Three approaches are currently being taken in an attempt to overcome these difficulties. The first involves characterizing the various growth factors needed to promote marrow engraftment, in an effort to develop pharmaceutical strategies to enhance chimerism (43,44). The second is to use recipient marrow or stem cells transduced with donor elements (MHC molecules) to induce a particular form of chimerism (with ‘transduced-self’) that may be easier to sustain. Preliminary studies have demonstrated the promise of such an approach (45,46), although definitive evidence of resulting transplantation tolerance is awaited. The third has proven very efficacious in experimental models and involves transplanting xenogeneic thymus alongside hematopoietic material to promote tolerance (47,48), although it remains to be seen if this strategy could be developed for clinical use.

C. Porcine endogenous retroviruses (PERV)

Infection of human cells by porcine viruses encoded by endogenous (genomic) DNA sequences was first demonstrated in 1997 (49). Since then, much has been published about the various PERV and attempts have been made to define the risks of transmission in more than 150 individuals exposed to porcine tissue (50–53). Though no infected humans have been identified, these retrospective studies are far from ideal at defining the precise risks of infection. The recent demonstration that human cells used to reconstitute severe combined immunodeficiency (SCID) mice were infected after transplantation of porcine tissue (54) has illustrated the potential for cross-species transmission of PERV after transplantation into immunosuppressed recipients, although the real relevance of this model to a clinical setting is not clear. In preparation for the real risk of PERV transmission, most Western countries have established regulatory mechanisms to ensure that the first xenograft recipients are monitored thoroughly and comprehensively (for information on the USA regulatory framework, go to http://www.fda.gov/cber/xap/xap.htm).

The most important development in recent months has been the identification of a line of in-bred mini-pigs that appear not to transmit retrovirus to human cells, despite secreting intact virus (55). If this translates into a failure to transmit in vivo, this will represent the single most significant advance in this area since PERV were identified.

There is still (and, by definition, always will be) a problem of the ‘unknown’; that is viruses or other infectious agents, not yet identified, that may be transmitted by xenografts after transplantation. Of great concern is the risk that cross-species transmission of any infection might be accompanied by a change in the behavior of the organism, perhaps through mutation, that brings with it the risk of transmission into the general population and the risk of human disease or death (56). It is difficult to know how to address such a concern. Even if quantitating the risk of the emergence of such a pathogen was possible, such an exercise may or may not reassure the public; persuading some to accept any uncertainty, however, small, will be a difficult challenge after recent public-health debacles in various countries.

The infectious issues surrounding clinical xenotransplantation have yet to be fully resolved, and there is an ongoing need for informed public debate.

Summary

The future of clinical xenotransplantation has become uncertain in recent years with the discovery of PERV transmission to human cells and slow progress in overcoming the rejection of discordant grafts. It is correct at such a time to ask whether further research is justified. The answer has to be yes, for no other reason than xenotransplantation promises an unlimited supply of organs for transplantation. But how realistic is the prospect of clinical xenotransplantation?

Recent years have seen real and significant advances in our understanding of the problems of vascular rejection and potential ways to overcome AVR. Simultaneously, some forward progress has been made in developing strategies for tolerance. Of course, all these advances still need to be translated into improved survival of porcine organs after transplantation into primates, and this may take some years. Nevertheless, a scrutiny of recent research in this field, as summarized here, should give grounds for cautious optimism that xenograft rejection will eventually come under control. At the same time there is hope that some pig organs may not transmit PERV, although this has yet to be formally tested.

At the beginning of this article, the question posed was whether pig xenografts would ever fly. The answer is there is every chance that they will, but much more work is required in the ‘hangar’, and clearance from the ‘tower’ (the public and the regulatory authorities) is then needed. Continuing the aviation metaphor, it is worth remembering, in this age when rapid technological advance is sometimes taken for granted, the words of Orville Wright, who described the first flight of his machine as ‘…exceedingly erratic, partly due to the irregularity of the air, and partly to lack of experience in handling…’ This succinct description of xenotransplantation research to date also prophesies what we should expect in the near future.

Acknowledgments

The author currently receives grants from the Medical Research Council, Juvenile Diabetes Research Foundation and the British Heart Foundation.

Ancillary