Discordant xenografts surviving the initial hyperacute rejection phase may be subject to cellular rejection processes mediated by infiltrating leukocytes including T cells, NK cells and monocytes. The stable adhesion of these cell types to endothelial cells is due to the molecular interaction of the integrins VLA-4 and LFA-1 with their ligands vascular cell adhesion molecule (VCAM) and ICAM-1 present on the endothelial cells. Human VLA-4 binds to porcine VCAM, and blocking mAbs specific for porcine VCAM have been developed. We have localized the epitope of the antiporcine VCAM blocking mAbs 2A2 and 3F4 to domains 1 and 2, respectively. Humanized antibodies (IgG4 isotype) were constructed from these anti-porcine VCAM antibodies and demonstrated to inhibit adhesion of Ramos, Jurkat and YT cells, as well as purified resting and activated human T cells, to porcine aortic endothelial cells (PAEC). These cell types express both LFA-1 as well as VLA-4, suggesting blockade of human VLA-4 interaction with porcine VCAM may alone be sufficient to significantly impair adhesion of human leukocytes to porcine endothelial cells. The chimeric antiporcine VCAM (pVCAM) HuG4 antibodies promoted increased adhesion of Fc receptor (FcR) positive cells such as U937 monocytic cells to PAEC. In contrast, chimeric anti-porcine VCAM antibodies created using the CH1 and hinge region from human IgG2 and the CH2 and CH3 regions from human IgG4 (HuG2/G4 antibodies) inhibited binding of FcR positive cells to PAEC. These chimeric anti-pVCAM antibodies should allow delineation of the in vivo role of VLA-4/VCAM interaction in porcine-to-primate xenotransplants. Further, the design of the HuG2/G4 antibodies should render them efficacious in multiple settings requiring elimination of FcR binding.

The serious shortage of available donor organs for patients with end stage organ failure who are in need of solid organ transplantation has led to a heightened interest in xenotransplantation. The major barrier to successful discordant xenotransplantation is hyperacute rejection. Hyperacute rejection results from the deposition of preformed antibodies that activate complement on the luminal surface of the vascular endothelium, leading to vessel occlusion and graft failure within minutes to hours. Endogenous membrane-associated complement inhibitors normally protect endothelial cells from autologous complement -- however, these molecules are species-restricted and therefore are ineffective at inhibiting activated xenogeneic complement. To address the pathogenesis of hyperacute rejection in the pig-to-human combination, F1 offspring were generated from a transgenic founder animal that was engineered to express the human terminal complement inhibitor hCD59. High-level cell surface expression of hCD59 was detected in the hearts and kidneys of these transgenic F1 animals, similar to expression levels in human kidney tissue. The hCD59 was expressed on both large vessel and capillary endothelium. Ex vivo perfusion experiments, using human blood as the perfusate, were performed with transgenic porcine hearts and kidneys to evaluate the ability of hCD59 to inhibit hyperacute rejection. These experiments demonstrated that transgenic organs expressing hCD69 resisted hyperacute rejection, as measured by increased organ function for both the hearts and the kidneys, as compared with control pig organs. Hearts from hCD59-expressing animals demonstrated a five-fold prolongation in function compared with controls, 109.8 +/- 20.7 min versus 21.2 +/- 2.9 min (P = 0.164). The hCD59-expressing kidneys also demonstrated significantly prolonged function at 157.8 +/- 27.0 min compared with 60.0 +/- 6.1 min for controls (P = 0.0174). Deposition of C9 neoantigen In the vasculature of porcine organs perfused with human blood was markedly reduced in organs expressing hCD59. These studies demonstrate that C5b-9 plays an important role in hyperacute rejection of a porcine organ perfused with human blood and suggest that donor pigs transgenic for hCD59 may be an integral component of successful clinical xenotransplantation.

Exposure of nonprimate cells, tissues, and organs to human blood typically results in a rapid form of rejection that is mediated by high-titer naturally occurring antibodies and complement. This hyperacute rejection presents a formidable barrier to the use of nonprimate animals as transplant donors. The major xenoantigen responsible for such rejection has recently been identified as a single carbohydrate structure, the α-galactosyl epitope. This epitope is expressed on the cells of most mammals, with the exception of humans and other Old World primates. Conversely, the anti-α-galactosyl antibody (anti-Gal) is unique to Old World primates. Complement-mediated inactivation of certain type C retroviruses is also restricted to Old World primates, an observation that has recently prompted the question of whether complement activation by these viruses involves anti-Gal. Two decades ago, this species-restricted viral inactivation was attributed to an antibody-independent mechanism mediated by direct activation of complement by the virus. However, recent studies now challenge this hypothesis and suggest that viral inactivation occurs through an antibody-dependent mechanism that is initiated by anti-Gal. Understanding the mechanisms responsible for serum inactivation of retroviruses has become increasingly important in the field of retroviral-mediated gene transfer, as in vivo applications have now entered the clinic. In this review we discuss the significance of anti-Gal and its carbohydrate epitope in retroviral-mediated gene transfer and xenotransplantation and the possible contribution of this antibody–antigen interaction to natural immunity in humans. The α-galactosyl epitope (Galα1–3Galβ1–4GlcNAc-R) is a glycosidic moiety that is expressed on the surface of cells from most mammalian species excluding humans and other Old World primates (catarrhine primates). Addition of this epitope to lipids and proteins occurs in the Golgi apparatus via transfer of galactose from the sugar donor nucleotide, uridine diphosphate galactose (UDP-galactose), to the acceptor N-acetyllactosaminyl group (Figure 1, reaction A). The enzyme that mediates this transfer is α1–3galactosyl transferase, a protein that is absent in Old World primates. Transcription from the α1–3galactosyl transferase gene in Old World primates is undetectable, and nonsense mutations are present within the coding region of some species. The presence of a functional α1–3galactosyl transferase in New World monkeys indicates that this gene was inactivated after the divergence of Old and New World primates approximately 40 million years ago. It has been speculated that α1–3galactosyl transferase gene null mutations in Old World primates were selected following an epidemic potentiated by an infectious agent that either carried the α-galactosyl epitope or used it as a receptor. Indeed, epitopes reactive with anti-Gal have been detected on several potential pathogens including enveloped viruses, bacteria, and protozoa (reviewed in7Galili U Springer Semin. Immunopathol. 1993; 15: 155-171Crossref PubMed Scopus (177) Google Scholar). A reciprocal relationship exists between α-galactosyl epitope expression and production of anti-Gal; the absence of this epitope in Old World primates correlates with production of anti-Gal in these species. In humans, as much as 1% of the total serum IgG is specific for the α-galactosyl moiety. Significant levels of anti-Gal IgM and IgA have also been reported. It has been hypothesized that elevated levels of anti-Gal persist due to constant antigenic stimulation by α-galactosyl-like epitopes found on the surface of bacteria constituting the normal flora, as has been proposed for the production of anti-blood group antibodies (reviewed in7Galili U Springer Semin. Immunopathol. 1993; 15: 155-171Crossref PubMed Scopus (177) Google Scholar). Type C retroviruses from various species often show a wide in vitro host range, and sequence analysis indicates that interspecies transmission of these viruses has occurred during evolution. However, type C retroviruses are rapidly inactivated in human serum in vitro (18Welsh Jr., R.M Cooper N.R Jensen F.C Oldstone M.B.A Nature. 1975; 257: 612-614Crossref PubMed Scopus (185) Google Scholar). Over 15 years ago, such retroviral inactivation was attributed to an innate immune mechanism mediated by direct binding of C1q (which initiates the classical complement pathway) to the retroviral envelope protein, p15E (3Cooper N.R Jensen F.C Welsh Jr., M.W Oldstone M.B.A J. Exp. Med. 1976; 144: 970-984Crossref PubMed Scopus (178) Google Scholar, 2Bartholomew R.M Esser A.F Muller-Eberhard H.J J. Exp. Med. 1978; 147: 844-853Crossref PubMed Scopus (99) Google Scholar). However, two recent studies have challenged this long-standing hypothesis by identifying an antibody-dependent mechanism as the predominant mediator of viral inactivation. Natural antibody specific for the envelope-associated α-galactosyl epitope was shown to be primarily responsible for complement activation and virolysis of murine-derived retroviruses in Old World primate sera (12Rother R.P Fodor W.L Springhorn J.P Birks C.W Setter E Sandrin M.S Squinto S.P Rollins S.A J. Exp. Med. 1995; 182: 1345-1355Crossref PubMed Scopus (157) Google Scholar). This inactivation was effectively inhibited by blockade of anti-Gal or down-regulation of the α-galactosyl epitope. These results were confirmed by 16Takeuchi Y Porter C.D Strahan K.M Preece A.F Gustafsson K Cosset J.-L Weiss R.A Collins M.K.L Nature. 1996; 379: 85-88Crossref PubMed Scopus (233) Google Scholar, who further demonstrated that type C retroviruses propagated through human cells expressing an exogenous α1–3galactosyl transferase gene acquire sensitivity to human serum killing. Based on these findings and an earlier report indicating that retroviruses produced in human cells display reduced complement sensitivity (15Takeuchi Y Cosset F.-L Lachmann P.J Okada H Weiss R.A Collins M.K.L J. Virol. 1994; 68: 8001-8007Crossref PubMed Google Scholar), it has been proposed that propagation of retrovirus in cells that do not express the α-galactosyl epitope will generate complement-resistant virus. Complement-mediated inactivation of retroviruses in human serum triggered by anti-Gal represents a unique type of natural immunity. Antibodies are normally considered to be an element of acquired immunity elicited in response to a specific foreign antigen. In contrast, anti-Gal is preformed and recognizes a carbohydrate moiety that is present on a variety of viruses. Prior to identification of anti-Gal, several studies reported naturally occurring antibody that recognized envelope glycoproteins from various type C retroviruses. It is likely that the natural antibody described in these early studies recognized the α-galactosyl epitope. For example, envelope glycoproteins were only targeted by antibody when the retroviruses were grown in non–Old World primate cells. In addition, serum proteins from New World primates or other mammalian species, but not Old World primates, shared the envelope glycoprotein determinants recognized by the natural antibody. These data suggest that the α-galactosyl epitope may be commonly associated with the envelopes of many different retroviruses when passaged through cells that express a functional α1–3galactosyl transferase (references can be found in1Barbacid M Bolognesi D Aaronson S.A Proc. Natl. Acad. Sci. USA. 1980; 77: 1617-1621Crossref PubMed Scopus (57) Google Scholar). In addition to retroviruses, other enveloped viruses contain proteins and/or lipids that bear the α-galactosyl epitope. The envelope of the eastern equine encephalitis virus has recently been shown to incorporate α-galactosyl modified glycoproteins when propagated through murine but not Old World primate cells (10Repik P.M Strizki J.M Galili U J. Gen. Virol. 1994; 75: 1177-1181Crossref PubMed Scopus (37) Google Scholar). Other enveloped viruses, including the lymphocytic choriomeningitis virus (LCMV) and Newcastle disease virus, acquire sensitivity to human serum when passaged through cells that express α1–3galactosyl transferase (19Welsh Jr., R.M J. Immunol. 1977; 118: 348-354PubMed Google Scholar). Conversely, LCMV grown in human cells or α-galactosyl epitope-deficient baby hamster kidney cells is completely resistant to human serum killing. Although the α-galactosyl epitope has not been identified on the surface of these viruses, the parallel between serum sensitivity and cell passage history strongly suggests its presence. The potential importance of human anti-Gal in providing natural immunity against enveloped viruses originating from animal species that express the α-galactosyl epitope is not known. Conservation of high-titer anti-Gal expression in Old World primates, including humans, suggests a critical function for this immunoglobulin. However, the occurrence of human zoonotic diseases such as influenza, rabies, and vesicular stomatitis suggests that this antibody is not a strong impediment to cross-species transfer of all viruses. Furthermore, agammaglobulinemia patients, who should lack this type of immunological barrier, do not appear to show a higher incidence of such zoonotic infections. The occurrence of zoonotic diseases does not rule out a role for anti-Gal in natural immunity against some viruses. Effective elimination of viruses expressing the α-galactosyl epitope is likely to be dependent on several factors including route of transmission and viral dose. For example, viruses introduced through aerosol transmission will not initially encounter high-titer anti-Gal IgG. Although gamma A class secretory anti-Gal may be encountered during this mode of transmission, IgA is a poor activator of complement and its binding has been shown to actually protect some bacteria from complement lysis (8Hamadeh R.M Estabrook M.M Zhou P Jarvis G.A Griffiss J.M Infect. Immun. 1995; 63: 4900-4906Crossref PubMed Google Scholar). In addition, if inactivation of a particular virus by anti-Gal is complement dependent, concentration of complement at the site of viral entry may play a critical role. Finally, the size of the initial dose of virus could determine the difference between pathogenesis or viral clearance. If a single virion escapes anti-Gal-mediated inactivation and is amplified in a human cell, it will no longer express the α-galactosyl epitope and could more readily spread to other cells. By the same token, once a virus successfully infects one individual, transmission to other humans will not be subject to inactivation through this mechanism. The role of the α-galactosyl epitope in serum killing of retroviruses has important implications in the field of retroviral-mediated gene therapy. Although the transfer of genes via retroviral particles has proven successful ex vivo, gene transfer in vivo has been primarily restricted to nonprimate species. One of the major obstacles to in vivo retroviral gene transfer in Old World primates, including humans, is the rapid complement-mediated inactivation of murine-derived retroviruses in primate blood. The observation that retroviral inactivation is mediated through recognition of the α-galactosyl epitope has thus prompted attempts to prevent incorporation of this carbohydrate structure into viral envelopes. One strategy that has recently been employed to generate serum-resistant retroviruses involves the manipulation of currently available high-titer retroviral producer cells. This strategy is based on previous observations showing that competition between glycosyl transferases for a common acceptor can determine which terminal glycosidic residues are added. The N-acetyllactosaminyl group (shown in Figure 1) is the acceptor for sugars from at least five glycosyl transferases in mammalian cells, including α1–3galactosyl transferase and α1–2fucosyl transferase (H-transferase). H-transferase generates the terminal glycosidic structure H-antigen (corresponding to the O blood group phenotype), which is universally compatible in human blood transfusions. A recent study has shown that expression of recombinant H-transferase in cells containing a functional α1–3galactosyl transferase gene results in elevated expression of H-antigen modified membrane proteins (Figure 1, reaction B) with a concomitant reduction in α-galactosyl epitope expression (14Sandrin M.S Fodor W.L Mouhtouris E Osman N Cohney S Rollins S.A Guilmette E.R Setter E Squinto S.P McKenzie I.F.C Nature Med. 1995; 1: 1261-1267Crossref PubMed Scopus (276) Google Scholar). Using this same strategy, it has been demonstrated that expression of H-transferase in retroviral producer cells effectively reduces the level of α-galactosyl residues on the cell surface and renders both cells and virus resistant to complement-mediated damage (12Rother R.P Fodor W.L Springhorn J.P Birks C.W Setter E Sandrin M.S Squinto S.P Rollins S.A J. Exp. Med. 1995; 182: 1345-1355Crossref PubMed Scopus (157) Google Scholar). Cells that are deficient in α-galactosyl epitope expression are also good candidates for the generation of cell lines that produce serum-resistant retrovirus. Such cells include those of humans and other Old World primates. Human producer cells have already been reported to produce retrovirus that is resistant to inactivation in human serum (5Cosset F.-C Takeuchi Y Battini J.-L Weiss R.A Collins M.K.L J. Virol. 1995; 69: 7430-7436Crossref PubMed Google Scholar). In addition, producer lines can be generated from nonprimate cells that do not express the α-galactosyl epitope. These include naturally deficient cells such as baby hamster kidney or Chinese hamster ovary, as well as cells obtained from α1–3galactosyl transferase gene knockout mice. Upon establishing circulation to a nonprimate xenogeneic organ transplanted into an Old World primate, complement is activated and blood flow to the organ ceases within minutes. This hyperacute form of graft rejection is primarily mediated by recognition of the α-galactosyl epitope on the surface of xenogeneic endothelial cells by anti-Gal. Efforts are now underway to reduce or eliminate the expression of this epitope in transgenic animals. These include down-regulation of the α-galactosyl epitope through the expression of H-transferase and the targeted knockout of the α1–3galactosyl transferase gene. Other strategies to abrogate this antibody–antigen interaction include transient depletion of anti-Gal in the circulation through extracorporeal absorption and antibody blockade using synthetic carbohydrates (references can be found in14Sandrin M.S Fodor W.L Mouhtouris E Osman N Cohney S Rollins S.A Guilmette E.R Setter E Squinto S.P McKenzie I.F.C Nature Med. 1995; 1: 1261-1267Crossref PubMed Scopus (276) Google Scholar). Cultured cells from nonprimate species are also being utilized in transplantation. For example, murine-derived retrovirus producer cells have recently been introduced into humans as a means of increasing transduction efficiency. However, recent evidence shows that murine producer cells are rapidly killed in human serum, primarily through anti-Gal-mediated complement lysis (13Russell D.W Berger M.S Miller A.D Hum. Gene Ther. 1995; 6: 635-641Crossref PubMed Scopus (55) Google Scholar, 11Rollins S.A Birks C.W Setter E Squinto S.P Rother R.P Hum. Gene Ther. 1996; 7: 619-626Crossref PubMed Scopus (24) Google Scholar). This is in contrast to data showing that murine-derived producer cells survive exposure to human cerebrospinal fluid; a finding that is probably attributable to low levels of antibody and/or complement in the central nervous system (13Russell D.W Berger M.S Miller A.D Hum. Gene Ther. 1995; 6: 635-641Crossref PubMed Scopus (55) Google Scholar). It is likely that survival of transplanted xenogeneic cells in tissues that are exposed to higher levels of antibody and complement can also be potentiated by manipulation of the α-galactosyl epitope. It has been observed that intravenous administration of replication-competent murine amphotropic retrovirus does not cause disease in Old World primates (4Cornetta K Moen R.C Culver K Morgan R.A McLachlin J.R Sturm S Selegue J London W Blaese R.M Anderson W.F Hum. Gene Ther. 1990; 1: 15-30Crossref PubMed Scopus (146) Google Scholar). In this study, retrovirus was generated from murine cells and would be expected to express α-galactosyl residues. By contrast, replication-competent Moloney murine leukemia virus introduced into primate stem cells ex vivo has been shown to induce T cell lymphoma in rhesus monkeys following bone marrow reconstitution (6Donahue R.E Kessler S.W Bodine D Goodman S Agricola B Byrne E Raffeld M Moen R Bacher J Zsebo K.M Nienhuis A.W J. Exp. Med. 1992; 176: 1125-1135Crossref PubMed Scopus (466) Google Scholar). The pathogenicity of this retrovirus may reflect the absence of the α-galactosyl epitope on the viral envelope following passage through primate cells, although duplication of a core enhancer element in recombinant retrovirus isolated from the lymphoma cells has recently been implicated as the pathogenic factor (9Purcell D.F.J Broscius C.M Vanin E.F Buckler C.E Nienhuis A.W Martin M.A J. Virol. 1996; 70: 887-897Crossref PubMed Google Scholar). The generation of α1–3galactosyl transferase gene knockout mice, which have been shown to produce anti-Gal (17Thall A.D Maly P Lowe J.B J. Biol. Chem. 1995; 270: 21437-21440Crossref PubMed Scopus (313) Google Scholar), should provide a model for assessing the importance of this antibody in immunity against replication-competent retrovirus. Taken together, these data underscore the need for stringent assays to detect replication-competent retrovirus in viral populations produced for administration to humans. Although the introduction of various deletions and mutations in the helper virus genome and the separation of helper virus coding regions (gag-pol and env) in packaging cell lines have decreased the generation of replication-competent retroviruses, evidence that endogenous viral sequences found in eukaryotic cells may also participate in recombination events has been demonstrated (9Purcell D.F.J Broscius C.M Vanin E.F Buckler C.E Nienhuis A.W Martin M.A J. Virol. 1996; 70: 887-897Crossref PubMed Google Scholar). This issue has recently been addressed through the generation of packaging cell lines in which the retroviral vector and packaging cell are derived from different species in hope of eliminating the occurrence of endogenous sequences that can recombine with the vector (5Cosset F.-C Takeuchi Y Battini J.-L Weiss R.A Collins M.K.L J. Virol. 1995; 69: 7430-7436Crossref PubMed Google Scholar). Concerns that human pathogens may be generated through the elimination of the α-galactosyl epitope are also relevant to the field of xenotransplantation. Based on anatomical, physiological, and ethical considerations, the primary donor candidate for xenotransplantation into humans is the pig. Questions have been raised regarding the possibility of infecting recipients with potentially pathogenic retroviruses from the porcine donor. Removal of the α-galactosyl epitope in porcine organs or removal of anti-Gal in the recipient may increase the possibility of interspecies retroviral transmission by affecting natural immunity against the virus. Efforts are currently underway to isolate and characterize porcine-derived retroviruses and assess their potential danger to humans. A better understanding of the α-galactosyl epitope and its potential role in restricting the interspecies transmission of viruses and other pathogens to humans is needed for the development of safe and efficacious strategies of xenotransplantation, as well as retroviral-mediated gene therapy. The absence of the α-galactosyl epitope in humans and other Old World primates poses many fascinating questions. Why do members of this lineage carry null mutations in the α1–3galactosyl transferase gene, while evolutionary pressures have conserved this enzymatic activity in all other mammalian species? What is the function of anti-Gal, and is this antibody an important component of natural immunity against potential human pathogens that carry the α-galactosyl epitope? Will the risk of infection associated with the introduction of retroviruses and animal donor organs into patients be increased by the elimination of the α-galactosyl epitope? Recent findings have provided some tantalizing hints, but there is much yet to learn concerning the significance of anti-Gal and its carbohydrate epitope.

It has been shown that peripheral T cell tolerance can be induced by systemic antigen administration. We have been interested in using this phenomenon to develop antigen-specific immunotherapies for T cell-mediated autoimmune diseases. In patients with the demyelinating disease multiple sclerosis (MS), multiple potentially autoantigenic epitopes have been identified on the two major proteins of the myelin sheath, myelin basic protein (MBP) and proteolipid protein (PLP). To generate a tolerogenic protein for the therapy of patients with MS, we have produced a protein fusion between the 21.5-kD isoform of MBP (MBP21.5) and a genetically engineered form of PLP (deltaPLP4). In this report, we describe the effects of treatment with this agent (MP4) on clinical disease in a murine model of demyelinating disease, experimental autoimmune encephalomyelitis (EAE). Treatment of SJL/J mice with MP4 after induction of EAE either by active immunization or by adoptive transfer of activated T cells completely prevented subsequent clinical paralysis. Importantly, the administration of MP4 completely suppressed the development of EAE initiated by the cotransfer of both MBP- and PLP-activated T cells. Prevention of clinical disease after the intravenous injection of MP4 was paralleled by the formation of long-lived functional peptide-MHC complexes in vivo, as well as by a significant reduction in both MBP- and PLP-specific T cell proliferative responses. Mice treated with MP4 were resistant to disease when rechallenged with an encephalitogenic PLP peptide emulsified in CFA, indicating that MP4 administration had a prolonged effect in vivo. Administration of MP4 was also found to markedly ameliorate the course of established clinical disease. Finally, MP4 therapy was equally efficacious in mice defective in Fas expression. These results support the conclusion that MP4 protein is highly effective in suppressing disease caused by multiple neuroantigen epitopes in experimentally induced demyelinating disease.

The introduction of retroviral vector producer cells (VPC) into tumors as a means of increasing transduction efficiency has recently been employed in human gene therapy trials. However, the fate of these xenogeneic cells in humans is not well understood. In the present study, we used an in vitro model to examine the survival of commonly used VPC lines in serum from humans and various other species. VPC derived from the murine NIH-3T3 cell line, including PA317, Psi CRIP, and GP + E-86, were effectively killed in sera from Old World primates, including human and baboon. Conversely, the same murine cell lines survived exposure to sera from dog, rabbit, rat, and mouse. This pattern of serum killing parallels the occurrence of the anti-alpha-galactosyl natural antibody (Ab) found exclusively in Old World primates. The anti-alpha-galactosyl Ab targets the terminal glycosidic structure Gal alpha 1-3Gal beta 1-4GlcNAc-R (alpha-galactosyl epitope) found on the surface of mammalian cells, excluding Old World primates. All murine-derived VPC tested expressed high levels of the alpha-galactosyl epitope as determined by FACS analysis. VPC killing was complement-mediated, because preincubation of human serum with a functionally blocking anti-C5 mAb completely abolished cell lysis. Furthermore, addition of soluble galactose(alpha 1-3)galactose (Gal alpha 1-3Gal) to human serum or down-regulation of the alpha-galactosyl epitope on the surface of VPC effectively reduced VPC killing, indicating that complement activation by these cells is primarily initiated by natural antibody recognition of the alpha-galactosyl epitope. Finally, VPC incubated with human serum for 8 hr in the presence of complement inhibition continued to produce viable retroviral particles, thus demonstrating a correlation between VPC and particle survival. Taken together, these data suggest that elimination of the alpha-galactosyl epitope or complement blockade may provide a strategy to prolong the survival of VPC and the particles that they produce in vivo.

Over the past few years, several major advances have occurred in the understanding of how the humoral and cellular immune system of humans recognizes and destroys transplanted cells, tissues and organs derived from animal sources. Consequently, armed with this new knowledge, several laboratories have now developed novel immunoprotective technologies that may allow xenotransplantation to be clinically feasible.

Activation of the complement system contributes significantly to the pathogenesis of numerous acute and chronic diseases. Recently, a monoclonal antibody (5G1.1) that recognizes the human complement protein C5, has been shown to effectively block C5 cleavage, thereby preventing the generation of the pro-inflammatory complement components C5a and C5b-9. Humanized 5G1.1 antibody, Fab and scFv molecules have been produced by grafting the complementarity determining regions of 5G1.1 on to human framework regions. Competitive ELISA analysis indicated that no framework changes were required in the humanized variable regions for retention of high affinity binding to C5, even at framework positions predicted by computer modeling to influence CDR canonical structure. The humanized Fab and scFv molecules blocked complement-mediated lysis of chicken erythrocytes and porcine aortic endothelial cells in a dose-dependent fashion, with complete complement inhibition occurring at a three-fold molar excess, relative to the human C5 concentration. In contrast to a previously characterized anti-C5 scFv molecule, the humanized h5G1.1 scFv also effectively blocked C5a generation. Finally, an intact humanized h5G1.1 antibody blocked human complement lytic activity at concentrations identical to the original murine monoclonal antibody. These results demonstrate that humanized h5G1.1 and its recombinant derivatives retain both the affinity and blocking functions of the murine 5G1.1 antibody, and suggest that these molecules may serve as potent inhibitors of complement-mediated pathology in human inflammatory diseases.

Human cells express cell surface complement regulatory molecules that inhibit the activity of the C3/C5 convertases (DAF, MCP, CR1) or inhibit the membrane attack complex (CD59). A single molecule that inhibits both the convertase activity and formation of the membrane attack complex has never been characterized. To this end, we have developed two reciprocal chimeric complement inhibitors (CD, NH2-CD59-DAF-GPI; and DC, NH2-DAF-CD59-GPI) that contain the functional domains of decay accelerating factor (DAF; CD55) and CD59. Cell surface expression of the CD and DC chimeric proteins was detected with DAF- and CD59-specific antisera. Cell surface C3d deposition was inhibited on cells expressing the chimeric molecules, thereby indicating that the DAF moiety was functional in both molecules. Conversely, Ab-blocking experiments demonstrated that only the DC molecule retained CD59 function. Therefore, the DC molecule represents a novel potent chimeric bifunctional complement inhibitor that retains the functional domains of two distinct complement regulatory molecules.

Prevention of hyperacute xenograft rejection in the pig-to-primate combination has been accomplished by removal of natural antibodies, complement depletion with cobra venom factor, or prevention of C3 activation with the soluble complement inhibitor sCR1. Although these strategies effectively prevent hyperacute rejection, they do not address the relative contribution of early (C3a, C3b) versus late (C5a, C5b-9) activated complement components to xenogeneic organ damage. To better understand the role of the terminal complement components (C5a, C5b-9) in hyperacute rejection, an anti-human C5 mAb was developed and tested in an ex vivo model of cardiac xenograft rejection. In vitro studies demonstrated that the anti-C5 mAb effectively blocked C5 cleavage in a dose-dependent manner that resulted in complete inhibition of both C5a and C5b-9 generation. Addition of anti-C5 mAb to human blood used to perfuse a porcine heart prolonged normal sinus cardiac rhythm from a mean time of 25.2 min in hearts perfused with unmodified blood to 79,296, or > 360 min when anti-C5 mAb was added to the blood at 50 micrograms/ml, 100 micrograms/ml, or 200 micrograms/ml, respectively. In these experiments, activation of the classical complement pathway was completely inhibited. Hearts perfused with blood containing the highest concentration of anti-C5 mAb had no histologic evidence of hyperacute rejection and no deposition of C5b-9. These experiments suggest that the activated terminal complement components C5a and C5b-9, but not C3a or C3b, play a major role in tissue damage in this porcine-to-human model of hyperacute rejection. They also suggest that targeted inhibition of terminal complement activation by anti-C5 mAbs may be useful in clinical xenotransplantation.

The use of retroviral-mediated gene transfer as an in vivo approach for the treatment of human disease has been limited by the inability of retroviral vectors to survive the host humoral immune system. The rapid inactivation of retroviruses that ensues following exposure to human or primate serum has been largely attributed to the activition of the complement cascade mediated through the classical pathway. We have extensively investigated and experimentally defined the relationship between the human complement cascade and retroviral vector particle inactivation. These research efforts have allowed us to develop several related solutions to the problems of retroviral inactivation by human complement. Furthermore, the approaches we describe here establish commercially viable methods. to prevent complement-mediated inactivation of retroviral vector particles and producer cells in human blood which should enable and expedite the direct in vivo retroviral-mediated gene transfer in humans.

The rapid inactivation of murine-derived retroviral vectors in human or nonhuman primate sera is largely attributed to the activity of complement mediated through the classical pathway. In this study, we have further investigated the relationship between the human complement cascade and retrovirus inactivation. Preincubation in normal human serum effectively inactivated LXSN retroviral vector particles, whereas the vector maintained the ability to transduce cells following incubation in sera deficient in either the C1, C2, C3, C5, C6, C8, or C9 human complement proteins. Preincubation of serum with monoclonal antibodies (mAbs) that functionally block specific complement components, including C5, C6, C8, and C9, successfully protected the LXSN vector from complement-mediated inactivation. Treatment of serum with cobra venom factor, which consumes terminal complement, also effectively protected the vector from inactivation. LXSN vector survival in serum corresponded inversely to the level of complement activity following treatment of serum with anti-C5 mAb as assessed in an erythrocyte hemolytic assay. Additionally, pretreatment of human whole blood with anti-C5 mAb effectively inhibited inactivation of the LXSN vector. Taken together, these data demonstrate that formation of the membrane attack complex (MAC, C5b-9) is required for the inactivation of the murine-based LXSN retroviral vector in human blood and that this process can be abrogated with the use of soluble complement inhibitors.

The herpesvirus saimiri genome encodes a complement control protein homolog (CCPH). Stable mammalian cell transfectants expressing a recombinant transmembrane form of CCPH (mCCPH) or a 5'FLAG epitope-tagged mCCPH (5'FLAGmCCPH) conferred resistance to complement-mediated cell damage by inhibiting the lytic activity of human serum complement. The function of CCPH was further defined by showing that the mCCPH and the 5'FLAGmCCPH transfectants inhibited C3 convertase activity and effectively reduced cell surface deposition of the activated complement component, C3d.

Type C retroviruses endogenous to various nonprimate species can infect human cells in vitro, yet the transmission of these viruses to humans is restricted. This has been attributed to direct binding of the complement component C1q to the viral envelope protein p15E, which leads to classical pathway-mediated virolysis in human serum. Here we report a novel mechanism of complement-mediated type C retrovirus inactivation that is initiated by the binding of "natural antibody" [Ab] (anti-alpha-galactosyl Ab) to the carbohydrate epitope Gal alpha 1-3Gal beta 1-4GlcNAc-R expressed on the retroviral envelope. Complement-mediated inactivation of amphotropic retroviral particles was found to be restricted to human and other Old World primate sera, which parallels the presence of anti-alpha-galactosyl natural Ab. Blockade or depletion of anti-alpha-galactosyl Ab in human serum prevented inactivation of both amphotropic and ecotropic murine retroviruses. Similarly, retrovirus was not killed by New World primate serum except in the presence of exogenous anti-alpha-galactosyl Ab. Enzyme-linked immunosorbent assays revealed that the alpha-galactosyl epitope was expressed on the surface of amphotropic and ecotropic retroviruses, and Western blot analysis further localized this epitope to the retroviral envelope glycoprotein gp70. Finally, down-regulation of this epitope on the surface of murine retroviral particle producer cells rendered them, as well as the particles liberated from these cells, resistant to inactivation by human serum complement. Our data suggest that anti-alpha-galactosyl Ab may provide a barrier for the horizontal transmission of retrovirus from species that express the alpha-galactosyl epitope to humans and to other Old World primates. Further, these data provide a mechanism for the generation of complement-resistant retroviral vectors for in vivo gene therapy applications where exposure to human complement is unavoidable.

The major obstacle to successful discordant xenotransplantation is the phenomenon of hyperacute rejection (HAR). In the pig-to-primate discordant transplant setting, HAR results from the deposition of high-titre anti-alpha-galactosyl antibodies and complement activation leading to endothelial cell destruction and rapid graft failure. To overcome HAR, we developed an enzymatic carbohydrate remodelling strategy designed to replace expression of the Gal alpha-1,3-Gal xenoepitope on the surface of porcine cells with the nonantigenic universal donor human blood group O antigen, the alpha-1,2-fucosyl lactosamine moiety (H-epitope). Xenogenic cells expressing the human alpha-1,2-fucosyltransferase expressed high levels of the H-epitope and significantly reduced Gal alpha-1,3-Gal expression. As a result, these cells were shown to be resistant to human natural antibody binding and complement-mediated cytolysis.

Complement activation contributes to the systemic inflammatory response induced by cardiopulmonary bypass. At the cellular level, cardiopulmonary bypass activates leukocytes and platelets; however the contribution of early (3a) versus late (C5a, soluble C5b-9) complement components to this activation is unclear. We used a model of simulated extracorporeal circulation that activates complement (C3a, C5a, and C5b-9 formation), platelets (increased percentages of P-selectin-positive platelets and leukocyte-platelet conjugates), and neutrophils (upregulated CD11b expression). to specifically target complement activation in this model, we added a blocking mAb directed at the human C5 complement component and assessed its effect on complement and cellular activation. Compared with a control mAB, the anti-human C5 mAb profoundly inhibited C5a and soluble C5b-9 generation and serum complement hemolytic activity but had no effect on C3a generation. Additionally, the anti-human C5 mAb significantly inhibited neutrophil CD11b upregulation and abolished the increase in P-selectin-positive platelets and leukocyte-platelet conjugate formation compared to experiments performed with the control mAb. This suggests that the terminal components C5a and C5b-9, but not C3a, directly contribute to platelet and neutrophil activation during extracorporeal circulation. Furthermore, these data identify the C5 component as a site for therapeutic intervention in cardiopulmonary bypass.