The International Xenotransplantation Association consensus statement on conditions for undertaking clinical trials of porcine islet products in type 1 diabetes--chapter 5: Strategies to prevent transmission of porcine endogenous retroviruses.

Xenotransplantation using porcine cells, tissues, or organs may offer a potential solution for the shortage of allogeneic human organs. Prior to the clinical use of porcine xenotransplants, three main hurdles must be overcome: immunologic rejection, physiologic incompatibility, and risk of transmission of porcine pathogens. Designated pathogen-free breeding of pigs can prevent transmission of most porcine microbes. However, this is not possible in the case of porcine endogenous retroviruses (PERV), which are integrated in the pig genome and can infect human cells in vitro. In order to assess the probability of transmission of PERV, a careful screening of the source pig herd is recommended. Two types of PERV are present in pigs, human-tropic PERV-A and PERV-B, which are both present in the genome of all pigs, and PERV-C, which is not ubiquitous and infects only pig cells. In addition to these viruses, recombinant PERV-A/C viruses have recently been described that (i) are able to infect human cells; (ii) are characterized by high titre replication; and (iii) are associated with proviruses that are de novo integrated in the DNA of somatic pig cells, but not yet in the pig germ line. The risks presented by PERV-A/C recombinant viruses could easily be eliminated by using pigs not containing PERV-C in their germ line, thereby effectively preventing recombination with PERV-A. Selection of PERV-C-free animals, if possible, therefore reduces the risk of PERV-A/C transmission to humans. Although it is unclear whether PERV-C may be transmitted in vivo from pig-to-pig, an infection of PERV-C-free animals with this virus may be prevented. To select pigs with low-level expression of PERV-A and PERV-B, it is recommended to apply assays based on real-time reverse transcriptase polymerase chain reaction (RT-PCR), which enables discrimination between pigs with high-level expression and low-level expression. Screening xenotransplant recipients for PERV transmission can be done in a number of ways. Provirus integration and PERV expression could theoretically be detected in peripheral blood mononuclear cells using PCR and RT-PCR. However, as the cells in which PERV replicates are still unknown, it is unclear whether this will be a reliable approach. Applying sufficiently sensitive assays to differentiate between transmission and chimerism is recommended. As can be commonly observed after retrovirus infection, the detection of virus-specific antibodies may indicate infection; however, the possibility of abortive infection, antigen exposure without infection, or cross-reactive response must be considered and explored as alternative explanations. On the other hand, it remains unclear whether the absence of specific antibodies indicates the absence of such infection, in particular if recipients of a xenotransplantation product are under chronic immunosuppression that could prevent antibody formation. Antibodies may be detected by Western blot or ELISA, using purified virus or recombinant viral proteins as antigens. Finally, it may be possible to detect cross-species PERV transmission by evaluating cells from the recipient for their in vitro potential of transmission to specified target cells (the human renal epithelial 293 cell line being the best example). There is no in vivo animal model for cross-species PERV transmission, and therefore it is not possible to validate monitoring assays for PERV transmission in an in vivo situation. Finally, virus safety of xenotransplantation is a fast-developing field, and new experimental findings will change existing strategies and introduce new ones.

No in vivo infection of triple immunosuppressed non-human primates after inoculation with high titers of porcine endogenous retroviruses.

UNLABELLED: Porcine endogenous retroviruses (PERVs) released from pig tissue can infect selected human cells in vitro and therefore represent a safety risk for xenotransplantation using pig cells, tissues, or organs. Although PERVs infect cells of numerous species in vitro, attempts to establish reliable animal models failed until now. Absence of PERV transmission has been shown in first experimental and clinical xenotransplantations; however, these trials suffered from the absence of long-term exposure (transplant survival) and profound immunosuppression. METHODS: We conducted infectivity studies in rhesus monkeys, pig-tailed monkeys, and baboons under chronic immunosuppression with cyclosporine A, methylprednisolone, and the rapamycin derivative. These species were selected because they are close to the human species and PERVs can be transmitted in vitro to cells of these species. In addition, the animals received twice, a C1 esterase inhibitor to block complement activation before inoculation of PERV. In order to overcome the complications of microchimerism, animals were inoculated with high titers of cell-free PERV. In addition, to enable transmission via cell-cell contact, some animals also received virus-producing cells. For inoculation the primate cell-adapted strain PERV/5 degrees was used which is characterized by a high infectious titer. Produced on human cells, this virus does not express alpha 1,3 Gal epitopes, does not contain porcine antigens on the viral surface and is therefore less immunogenic in non-human primates compared with pig cell-derived virus. Finally, we present evidence that PERV/5 degrees productively infects cells from baboons and rhesus monkeys. RESULTS: In a follow-up period of 11 months, no antibody production against PERV and no integration of proviral DNA in blood cells was observed. Furthermore, no PERV sequences were detected in the DNA of different organs taken after necropsy. CONCLUSION: These results indicate that in a primate model, in the presence of chronic immunosuppression, neither the inoculation of cell-free nor cell-associated PERV using a virus already adapted to primate cells results in an infection; this is despite the fact that peripheral blood mononuclear cells of the same animals are infectible in vitro.

Analysis of pig-to-human porcine endogenous retrovirus transmission in a triple-species kidney xenotransplantation model.

Clinical pig-to-human xenotransplantation might be associated with the risk of transmission of xenozoonoses, especially porcine endogenous retroviruses (PERVs). We have established a pig-to-humanised-cynomolgus monkey xenotransplantation model allowing the analysis of potential PERV-transmission from normal or transgenic porcine organs to human vascular tissue. Pig-to-human kidney xenotransplantation was performed in cynomolgus monkeys. An interposition graft constructed from a human saphena vein replaced the porcine kidney vein. After graft rejection and/or death of the recipient (survival 2, 4, 6, 13, 16, 19 days), the human interposition grafts were removed. Human endothelial cells (huECs) were isolated from the interposition grafts and cultivated in vitro. Explanted human vascular tissue, isolated huECs, plasma and serum samples of the graft recipients were characterised by flow cytometry and immunohistochemistry and screened for indications of PERV transmission by quantitative polymerase chain reaction (PCR), reverse transcriptase-polymerase chain reaction (RT-PCR) and RT assay. PERV-specific immune response of recipients was analysed by Western blot. No evidence of PERV infection or PERV-specific immune response was detected.

Marked prolongation of porcine renal xenograft survival in baboons through the use of alpha1,3-galactosyltransferase gene-knockout donors and the cotransplantation of vascularized thymic tissue.

The use of animal organs could potentially alleviate the critical worldwide shortage of donor organs for clinical transplantation. Because of the strong immune response to xenografts, success will probably depend upon new strategies of immune suppression and induction of tolerance. Here we report our initial results using alpha-1,3-galactosyltransferase knockout (GalT-KO) donors and a tolerance induction approach. We have achieved life-supporting pig-to-baboon renal xenograft survivals of up to 83 d with normal creatinine levels.

Heart transplantation in baboons using alpha1,3-galactosyltransferase gene-knockout pigs as donors: initial experience.

Hearts from alpha1,3-galactosyltransferase knockout pigs (GalT-KO, n = 8) were transplanted heterotopically into baboons using an anti-CD154 monoclonal antibody-based regimen. The elimination of the galactose-alpha1,3-galactose epitope prevented hyperacute rejection and extended survival of pig hearts in baboons for 2-6 months (median, 78 d); the predominant lesion associated with graft failure was a thrombotic microangiopathy, with resulting ischemic injury. There were no infectious complications directly related to the immunosuppressive regimen. The transplantation of hearts from GalT-KO pigs increased graft survival over previous studies.

Evidence and consequence of porcine endogenous retrovirus recombination.

The genetic nature and biological effects of recombination between porcine endogenous retroviruses (PERV) were studied. An infectious molecular clone was generated from a high-titer, human-tropic PERV isolate, PERV-A 14/220 (B. A. Oldmixon, et al. J. Virol. 76:3045-3048, 2002; T. A. Ericsson et al. Proc. Natl. Acad. Sci. USA 100:6759-6764, 2003). To analyze this sequence and 15 available full-length PERV nucleotide sequences, we developed a sequence comparison program, LOHA(TM) to calculate local sequence homology between two sequences. This analysis determined that PERV-A 14/220 arose by homologous recombination of a PERV-C genome replacing an 850-bp region around the pol-env junction with that of a PERV-A sequence. This 850-bp PERV-A sequence encompasses the env receptor binding domain, thereby conferring a wide host range including human cells. In addition, we determined that multiple regions derived from PERV-C are responsible for the increased infectious titer of PERV-A 14/220. Thus, a single recombination event may be a fast and effective way to generate high-titer, potentially harmful PERV. Further, local homology and phylogenetic analyses between 16 full-length sequences revealed evidence for other recombination events in the past that give rise to other PERV genomes that possess the PERV-A, but not the PERV-B, env gene. These results indicate that PERV-A env is more prone to recombination with heterogeneous backbone genomes than PERV-B env. Such recombination events that generate more active PERV-A appear to occur in pigs rather frequently, which increases the potential risk of zoonotic PERV transmission. In this context, pigs lacking non-human-tropic PERV-C would be more suitable as donor animals for clinical xenotransplantation.

Mouse retrovirus mediates porcine endogenous retrovirus transmission into human cells in long-term human-porcine chimeric mice.

Porcine endogenous retrovirus (PERV) is a potential pathogen in clinical xenotransplantation; transmission of PERV in vivo has been suggested in murine xenotransplantation models. We analyzed the transmission of PERV to human cells in vivo using a model in which immunodeficient NOD/SCID transgenic mice were transplanted with porcine and human lymphohematopoietic tissues. Our results demonstrate, we believe for the first time, that human and pig cells can coexist long-term (up to 25 weeks) without direct PERV infection of human cells. Despite the transplantation of porcine cells that did not produce human-tropic PERV, human cells from the chimeric mice were frequently found to contain PERV sequences. However, this transmission was due to the pseudotyping of PERV-C (a virus without human tropism) by xenotropic murine leukemia virus, rather than to de novo generation of human-tropic PERV. Thus, pseudotyping might account for the PERV transmission previously observed in mice. The absence of direct human cell infection following long-term in vivo coexistence with large numbers of porcine cells provides encouragement regarding the potential safety of using pigs that do not produce human-tropic PERV as source animals for transplantation to humans.

Xenotransplantation: infectious risk revisited.

Xenotransplantation is a possible solution for the shortage of tissues for human transplantation. Multiple hurdles exist to clinical xenotransplantation, including immunologic barriers, metabolic differences between pigs--the source species most commonly considered--and humans, and ethical concerns. Since clinical trials were first proposed almost 10 years ago, the degree of risk for infection transmitted from the xenograft donor to the recipient has been extensively investigated. A number of potential viral pathogens have been identified including porcine endogenous retrovirus (PERV), porcine cytomegalovirus (PCMV), and porcine lymphotropic herpesvirus (PLHV). Sensitive diagnostic assays have been developed for each virus. Human-tropic PERV are exogenous recombinants between PERV-A and PERV-C sequences and are present in only a subset of swine. Porcine cytomegalovirus can be excluded from herds of source animals by early weaning of piglets. In contrast, the risks associated with PLHV remain undefined. Microbiologic studies and assays for potential xenogeneic pathogens have furthered understanding of risks associated with xenotransplantation. Thus far, clinical xenotransplantation of pig tissues has not resulted in transmission of viral infection to humans; significant risks for disease transmission from swine to humans have not been confirmed. If immunologic hurdles can be overcome, it is reasonable to initiate carefully monitored clinical trials.

Characterization of germline porcine endogenous retroviruses from Large White pig.

Porcine endogenous retroviruses (PERV) are of concern when the microbiological safety aspects of xenotransplantation are considered. Four unique isolates of PERV B have been identified previously from a lambda library constructed from genomic DNA from a Large White pig. This study shows that none of these isolates are replication competent when transfected into permissive human or pig cells in vitro, and the removal of flanking genomic sequences does not confer a human tropic replication competent (HTRC) phenotype on these PERV proviruses. Analysis of the envelope sequences revealed that PERV B demonstrated high similarity to the envelope sequences derived from replication-competent PERV, indicating that lack of replication competence does not appear to be attributable to this region of the provirus. These data complement recent findings that HTRC PERV are recombinants between the PERV A and PERV C subgroups, and that these recombinants are not present in the germline of miniature swine. Together, these results indicate that these individual PERV B proviruses are unlikely to give rise to HTRC PERV.

Risk factors for the development of post-transplant lymphoproliferative disorder in a large animal model.

A high incidence of a post-transplant lymphoproliferative disorder (PTLD) is observed in miniature swine conditioned for allogeneic hematopoietic cell transplantation using a protocol involving T-cell depletion and cyclosporine therapy. This study was designed to assess contributing factors to disease development. Forty-six animals were studied including 12 (26%) that developed PTLD. A number of risk factors for PTLD were examined, including degree of immunosuppression, degree of MHC mismatch and infection by a porcine lymphotrophic herpesvirus (PLHV-1). Flow cytometry was used to measure host and donor T- and B-cell levels in the peripheral blood. Porcine lymphotrophic herpesvirus viral load was determined by quantitative PCR. Animals developing PTLD had significantly lower levels of T cells on the day of transplant. Cyclosporine levels did not differ significantly between animals with and without PTLD. Animals receiving transplants across a two-haplotype mismatch barrier showed an increased incidence of PTLD. All animals with PTLD had significant increases in PLHV-1 viral loads. Porcine lymphotrophic herpesvirus viral copy numbers remained at low levels in the absence of disease. The availability of a preclinical large-animal model with similarities to PTLD of humans may allow studies of the pathogenesis and treatment of that disorder.

No transmission of porcine endogenous retrovirus after transplantation of adult porcine islets into diabetic nude mice and immunosuppressed rats.

BACKGROUND: The aim of this study was to investigate whether transmission of porcine endogenous retrovirus (PERV) occurs in a model of diabetes reversal by the xenotransplantation of adult porcine islets (APIs) into immunoincompetent diabetic rodents. METHODS: Black-6 nu/nu mice and Lewis rats were immunosuppressed with cyclosporin A (CsA) and FTY 720, and rendered diabetic with streptozotocin. Purified APIs were transplanted into the renal subcapsular space; 5,000 islet equivalents (IEQs) were used in the nude mice (n = 4) and 40,000 IEQs in the rats (n = 4). The nude mice were sacrificed at 75 days after transplantation. In order to confirm chronic xenograft function, the graft-bearing kidney was removed prior to sacrifice. The rats were followed until xenograft rejection, at which time they were sacrificed. Immediately after sacrifice, tissue samples (liver, spleen, and small intestine) were taken for analysis. Quantitative polymerase chain reaction (PCR) was used to assess evidence of PERV transmission, and porcine cell chimerism. RESULTS: All animals became normoglycemic within 48 h of transplantation. The nude mice remained normoglycemic during the 75-day study period, with removal of the graft-bearing kidney resulting in prompt hyperglycemia. The rats remained normoglycemic until xenograft rejection, which occurred at 66 +/- 28 days. Despite the evidence of porcine cell microchimerism in recipients, real-time PCR detected no evidence of PERV transmission in any of the tissue specimens tested. CONCLUSIONS: There was no evidence of PERV transmission following transplantation of pig islets into diabetic nude mice and immunosuppressed rats.

Porcine endogenous retrovirus transmission characteristics of galactose alpha1-3 galactose-deficient pig cells.

Galactose alpha1-3 galactose (Gal) trisaccharides are present on the surface of wild-type pig cells, as well as on viruses particles produced from such cells. The recognition of Gal sugars by natural anti-Gal antibodies (NAb) in human and Old World primate serum can cause the lysis of the particles via complement-dependent mechanisms and has therefore been proposed as an important antiviral mechanism. Recently, pigs have been generated that possess disrupted galactosyl-transferase (GGTA1) genes. The cells of these pigs do not express Gal sugars on their surface, i.e., are Gal null. Concerns have been raised that the risk of virus transmission from such pigs may be increased due to the absence of the Gal sugars. We investigated the sensitivity of porcine endogenous retrovirus (PERV) produced from Gal-null and Gal-positive pig cells to inactivation by purified NAb and human serum. PERV produced in Gal-null pig cells was resistant to inactivation by either NAb or human serum. In contrast, although Gal-positive PERV particles were sensitive to inactivation by NAb and human serum, they required markedly higher concentrations of NAb for inactivation compared to the Gal-positive cells from which they were produced. Complete inactivation of Gal-positive PERV particles was not achievable despite the use of high levels of NAb, indicating that NAb-mediated inactivation of cell-free PERV particles is an inefficient process.

Reduced sensitivity to human serum inactivation of enveloped viruses produced by pig cells transgenic for human CD55 or deficient for the galactosyl-alpha(1-3) galactosyl epitope.

Complement activation mediated by the major xenogeneic epitope in the pig, galactosyl-alpha(1-3) galactosyl sugar structure (alpha-Gal), and human natural antibodies could cause hyperacute rejection (HAR) in pig-to-human xenotransplantation. The same reaction on viruses bearing alpha-Gal may serve as a barrier to zoonotic infection. Expressing human complement regulatory proteins or knocking out alpha-Gal epitopes in pig in order to overcome HAR may therefore pose an increased risk in xenotransplantation with regard to zoonosis. We investigated whether amphotropic murine leukemia virus, porcine endogenous retrovirus, and vesicular stomatitis virus (VSV) budding from primary transgenic pig aortic endothelial (TgPAE) cells expressing human CD55 (hCD55 or hDAF) was protected from human-complement-mediated inactivation. VSV propagated through the ST-IOWA pig cell line, in which alpha-galactosyl-transferase genes were disrupted (Gal null), was also tested for sensitivity to human complement. The TgPAE cells were positive for hCD55, and all pig cells except the Gal-null ST-IOWA expressed alpha-Gal epitopes. Through antibody binding, we were able to demonstrate the incorporation of hCD55 onto VSV particles. Viruses harvested from TgPAE cells were relatively resistant to complement-mediated inactivation by the three sources of human sera tested. Additionally, VSV from Gal-null pig cells was resistant to human complement inactivation. Such protection of enveloped viruses may increase the risk of zoonosis from pigs genetically modified for pig-to-human xenotransplantation.

Absence of replication-competent human-tropic porcine endogenous retroviruses in the germ line DNA of inbred miniature Swine.

The potential transmission of porcine endogenous retroviruses (PERVs) has raised concern in the development of porcine xenotransplantation products. Our previous studies have resulted in the identification of animals within a research herd of inbred miniature swine that lack the capacity to transmit PERV to human cells in vitro. In contrast, other animals were capable of PERV transmission. The PERVs that were transmitted to human cells are recombinants between PERV-A and PERV-C in the post-VRA region of the envelope (B. A. Oldmixon, J. C. Wood, T. A. Ericsson, C. A. Wilson, M. E. White-Scharf, G. Andersson, J. L. Greenstein, H. J. Schuurman, and C. Patience, J. Virol. 76:3045-3048, 2002); these viruses we term PERV-A/C. This observation prompted us to determine whether these human-tropic replication-competent (HTRC) PERV-A/C recombinants were present in the genomic DNA of these miniature swine. Genomic DNA libraries were generated from one miniature swine that transmitted HTRC PERV as well as from one miniature swine that did not transmit HTRC PERV. HTRC PERV-A/C proviruses were not identified in the germ line DNAs of these pigs by using genomic mapping. Similarly, although PERV-A loci were identified in both libraries that possessed long env open reading frames, the Env proteins encoded by these loci were nonfunctional according to pseudotype assays. In the absence of a germ line source for HTRC PERV, further studies are warranted to assess the mechanisms by which HTRC PERV can be generated. Once identified, it may prove possible to generate animals with further reduced potential to produce HTRC PERV.

Identification of exogenous forms of human-tropic porcine endogenous retrovirus in miniature Swine.

The replication of porcine endogenous retrovirus subgroup A (PERV-A) and PERV-B in certain human cell lines indicates that PERV may pose an infectious risk in clinical xenotransplantation. We have previously reported that human-tropic PERVs isolated from infected human cells following cocultivation with miniature swine peripheral blood mononuclear cells (PBMC) are recombinants of PERV-A with PERV-C. Here, we report that these recombinants are exogenous viruses in miniature swine; i.e., they are not present in the germ line DNA. These viruses were invariably present in miniature swine that transmitted PERV to human cells and were also identified in some miniature swine that lacked this ability. These data, together with the demonstration of the absence of both replication-competent PERV-A and recombinant PERV-A/C loci in the genome of miniature swine (L. Scobie, S. Taylor, J. C. Wood, K. M. Suling, G. Quinn, C. Patience, H.-J. Schuurman, and D. E. Onions, J. Virol. 78:2502-2509, 2004), indicate that exogenous PERV is the principal source of human-tropic virus in these animals. Interestingly, strong expression of PERV-C in PBMC correlated with an ability of the PBMC to transmit PERV-A/C recombinants in vitro, indicating that PERV-C may be an important factor affecting the production of human-tropic PERV. In light of these observations, the safety of clinical xenotransplantation from miniature swine will be most enhanced by the utilization of source animals that do not transmit PERV to either human or porcine cells. Such animals were identified within the miniature swine herd and may further enhance the safety of clinical xenotransplantation.

Genotyping of porcine endogenous retroviruses from a family of miniature swine.

The identification of animals in an inbred miniature swine herd that consistently fail to produce replication- competent humantropic porcine endogenous retrovirus (PERV) has prompted studies on the biology of PERV in transmitter and nontransmitter animals. We analyzed PERV RNA transcript profiles in a family of inbred miniature swine (SLA(d/d) haplotype) in which individual members differed in their capacity to generate humantropic and ecotropic (i.e., pigtropic) virus. We identified unique HaeIII and HpaII gag restriction fragment length polymorphism (RFLP) profiles resulting from single nucleotide polymorphisms in blood cells; these were found only in animals that produced humantropic PERV. These HaeIII and HpaII gag RFLP profiles proved to be components of humantropic PERV as they were transmitted to 293 human target cells in vitro. The humantropic HaeIII and HpaII gag RFLP genotypes in the family of study were not present in other miniature swine in the herd that produced humantropic PERV, indicating that these RFLP profiles relate specifically to this family's lineage.

Reduced efficacy of ganciclovir against porcine and baboon cytomegalovirus in pig-to-baboon xenotransplantation.

In pig-to-baboon xenotransplantation, porcine cytomegalovirus (PCMV) causes viremia, consumptive coagulopathy, and tissue-invasive disease. Baboon cytomegalovirus (BCMV) is associated with invasive disease in xenograft recipients. The efficacy of prophylaxis with intravenous ganciclovir (GCV) was studied for prevention of PCMV and BCMV infections in pig-to baboon xenotransplantation. GCV prophylaxis did not alter the incidence of BCMV activation in recipients, but reduced the amount of virus in tissues (mean 8.38 x 10(2) vs. 3.24 x 10(5) copies/ micro g DNA without treatment) and prevented tissue-invasive infections. PCMV viral loads were unaltered by GCV prophylaxis (8.36 x 10(8) copies/ micro g DNA without prophylaxis vs. 1.20 x 10(9) copies/ micro g DNA with prophylaxis). In vitro, PCMV was relatively resistant to GCV [90% inhibitory concentration (IC90) of 10 micro m, IC50 = 3 micro m], acyclovir (100 micro m), and leflunomide (not achievable). Only cidofovir (IC90 1 micro m) and foscarnet (IC90 100 micro m) might have therapeutic efficacy for PCMV in vivo in achievable concentrations, although these agents often carry significant toxicity in transplant recipients. GCV has limited activity against BCMV and no therapeutic efficacy against PCMV at standard doses in vivo. GCV and other antiviral agents have limited activities against PCMV in vitro. Breeding of PCMV-free xenograft donors may be necessary to prevent PCMV infections in clinical trials.

Packaging of human endogenous retrovirus sequences is undetectable in porcine endogenous retrovirus particles produced from human cells.

The chronic shortage of human donor organs and tissues for allotransplantation could be relieved if clinical xenotransplantation were to become a viable clinical therapy. Balanced against the benefits of xenotransplantation are the possible consequences of zoonotic infections, and in particular, infection by porcine endogenous retrovirus (PERV). An often-proclaimed risk of PERV infection is the possible recombination of PERV with human endogenous retroviruses (HERV). To address this issue, we examined the potential for HERV sequences to be cross-packaged into PERV particles produced from infected human 293 cells. Although HERV-K, W, E, R, and ERV-9 RNA transcripts are expressed in 293 cells, we did not detect cross-packaging of any of these HERV groups. Quantitative analysis indicated that less than approximately 1 in 10(4)-10(7) PERV particles might contain HERV sequences. In comparison, we found that murine leukemia virus (MLV)-based vector transcripts were cross-packaged at a rate of approximately one copy in 10(4) PERV particles. Our results indicate that the potential for recombination of PERV and HERV sequences is low and that novel viruses generated by this mechanism are unlikely to represent a significant risk for xenotransplantation.

Engraftment of retroviral EGFP-transduced bone marrow in mice prevents rejection of EGFP-transgenic skin grafts.

We have investigated whether a state of tolerance toward EGFP-expressing skin tissue can be induced by prior establishment of EGFP molecular chimerism by transplant of gene-transduced bone marrow in mice. Irradiated (10 Gy) C57BL/6J mice were transplanted with bone marrow cells transduced with two different retroviral vectors encoding EGFP. EGFP-transduced, mock-transduced, and age-matched control mice received skin grafts from both C57BL/6 EGFP-transgenic (B6-EGFP. Tg) and MHC-mismatched B10.A donor mice at 8, 29, or 39 weeks after bone marrow transplantation. Although 14 of 17 control mice rejected EGFP.Tg skin grafts within 100 days, 24 of 25 mice receiving EGFP-expressing bone marrow cells accepted their B6-EGFP.Tg grafts out to 200 days after skin grafting, including animals with undetectable levels of EGFP expression in blood cells. The EGFP-transduced animals rejected third-party grafts from MHC-mismatched mice within 20 days, indicating that acceptance of the EGFP-expressing skin grafts was the result of the induction of specific and operational immune tolerance. Thus, our data indicate that (a) EGFP-expressing tissue elicits an immunological rejection in C57BL/6 mice and (b) tolerance can be induced by engrafting relatively small numbers of EGFP-transduced hematopoietic cells. These experiments utilizing EGFP as an immunogen point to the wider therapeutic potential of employing transplantation of gene-transduced hematopoietic cells for establishing immunological tolerance and thereby preventing rejection of gene-corrected cells and tissues.