Expression of Galα(1,3)Gal by porcine islet cells and its relevance to xenotransplantation

Abstract: In pig-to-human transplantation, one of the major obstacles is the expression of Galα(1,3)Galactose by the endothelium of vascularized human tissues and the presence in all humans of IgG and IgM antibodies to this epitope: xenotransplantation would be followed by hyperacute rejection (HAR). However, the findings in endothelial cells of all vessels and the parenchyma of tissues such as kidney and liver do not extend to islet cells. Histological studies demonstrate that the adult pig islet (apart from endothelial cells) does not express Galα(1,3)Gal, nor do fresh neonatal or fetal pig islet cells; after tissue culture (under a variety of conditions) large amounts of Galα(1,3)Gal are expressed by the pig islet cells. However, double staining studies show that Gal+ cells do not secrete insulin, glucagon, or somatostatin and these cells may be spared from potential antigen antibody-mediated rejection after transplantation. The relevance of Gal expression in islet cells is discussed–short term studies–preferably in pig to Old World Monkey transplants could provide the answer to the relevance of Galα(1,3)Gal expression in islet cells. Thus far it appears that some islets can be destroyed by antibody; others are spared and it is important to determine the nature of the sparing mechanism.     

Variability of anti-αGal antibodies in human serum and their relation to serum cytotoxicity against pig cells

Abstract: One of the major obstacles in pig-to-human xenografting is hyperacute rejection (HAR) of pig cells caused by preformed anti-pig antibodies and complement. In 1991 we suggested that anti-αGal antibodies play a major role in the HAR of pig cells. Anti-αGal antibodies recognize terminal α-galactose-containing epitopes on glycoproteins and glycolipids. They are present in humans, apes, and Old World monkeys, but not in lower mammals such as pigs. However, pigs, unlike humans, express terminal α-galactose epitopes on vascular endothelium which represent targets for human anti-αGal antibodies. Despite increasing recognition that anti-αGal antibodies are an important factor, many questions related to their precise role in HAR remain to be answered.
In this study, we analyzed 75 human AB sera in terms of (i) cytotoxic activity against cultured pig (PK-15) cells, (ii) anti-αGal titers, (iii) immunoglobulin-binding to pig cells, and (iv) immunoglobulin concentrations. The results demonstrated considerable variability in cytotoxicity, anti-αGal titers, and immunoglobulin binding to pig cells, whereas the serum immunoglobulin concentrations were less variable. Positive correlations were found between cytotoxicity and binding of IgG and IgM to the surface of pig cells. The surface-bound IgG and IgM also correlated with the serum anti-αGal IgG and IgM titers. Anti-αGal IgA, however, did not show any relation with cytotoxicity or cell binding. Concentrations of serum total immunoglobulins correlated with neither I cytotoxic activity nor cell binding.     

A comparison of fetal and adult porcine islets with regard to Gal alpha (1,3)Gal expression and the role of human immunoglobulins and complement in islet cell cytotoxicity

BACKGROUND: It is still debated whether fetal or adult porcine islets should be the preferred choice for future clinical islet xenotransplantation. Each type of islet preparation has advantages and disadvantages compared with the other. Here we present a direct comparison between fetal and adult porcine islets with regard to Gal alpha(1,3)Gal expression, immunoglobulin and complement binding, and cytotoxicity after exposure to fresh human serum. METHOD: Islet single cell suspensions were prepared from adult and fetal islets by trypsin digestion. Fluorescein isothiocyanate-conjugated Bandeiraea simplicifolia isolectin B4 (BS-IB4) and affinity-purified chicken anti-Gal alpha(1,3)Gal antibody was used to detect Gal alpha(1,3)Gal expression. Immunoglobulin and complement binding to the islet cells and cytotoxicity for islet cells was compared after incubation with fresh and heat-inactivated human sera and with an immune serum from a diabetic patient who received a fetal porcine islet transplant. Furthermore, two pools of human AB sera were depleted of porcine endothelial cell cytotoxic human anti-Gal alpha(1,3)Gal antibodies by absorption and were used to analyze the effect of Gal alpha(1,3)Gal antibody removal on islet cell cytotoxicity. RESULTS: Fetal islet cells readily bound both BS-IB4 and the chicken anti-Gal alpha(1,3)Gal antibody. None of 10 adult porcine islet preparations were stained by BS-IB4. In comparison, IgY anti-Gal Ab binding was detected in two of eight adult islet isolations, whereas the other six preparations showed marginal/no binding. After incubation of fetal islet cells with fresh human serum, C3c binding was strongly positive and IgM binding variable, with occasional binding of IgG and no detectable binding of IgA. Adult islet cells were also strongly positive for C3c but did not bind detectable amounts of IgM, IgG, or IgA. Immune sera from a patient who had received fetal porcine islets showed the presence of induced antibodies that bound to fetal islet cells and to porcine peripheral blood lymphocytes, whereas binding to adult islet cells was barely detectable. Fresh human sera showed a high and similar level of complement-mediated lytic activity for both adult islet cells (78+/-22%) and fetal islet cells (75+/-16%). Cytotoxicity for fetal islet cells and peripheral blood lymphocytes was significantly reduced when the corresponding sera were depleted of anti-Gal antibodies before use (P=0.002 and P=0.003, respectively). In contrast, no difference in cytotoxicity for adult islet cells was detected when anti-Gal-depleted human sera were used. CONCLUSION: Gal alpha(1,3)Gal expression is occasionally detectable on adult porcine islet cells, but not as readily and at a lower level, compared with fetal islet cells. Thus, as porcine fetal islets mature to adult islets, the expression of the Gal alpha(1,3)Gal epitope gradually diminishes. Consequently, cytotoxic anti-Gal alpha(1,3)Gal antibodies in human serum play an important role in the lysis of fetal but not adult porcine islet cells.     

Reduction of alpha-galactosyl xenoantigen by expression of endo-beta-galactosidase C in pig endothelial cells

Background: Elimination of the Galα1–3Galβ1–4GlcNAc (αGal) epitope has been considered to be essential for successful pig-to-human xenotransplantation but, unfortunately, has not been achieved. Endo-β-galactosidase C (EndoGalC) is an endoglycosidase which cleaves the Galβ1–4GlcNAc linkage in the αGal epitope and digests out the Galα1–3Gal disaccharide. Because of its potent activity in physiological pH conditions, EndoGalC can remove αGal epitopes expressed on the cell surface of pig erythrocytes and vascular endothelial cells almost completely. In vivo or ex vivo administration of EndoGalC successfully reduced αGal expression in pig kidneys to an undetectable level, but αGal epitopes soon reappeared. Gene expression of EndoGalC in pig cells was attempted to solve this problem. As the terminal αGal is transferred in the trans-Golgi network by α-1,3-galactosyltransferase (α1,3GT), colocalization of the EndoGalC gene with the α1,3GT gene was expected to be one of the most reliable ways to eliminate the αGal epitope. Methods and results : The sequence of pig α1,3GT, including the cytoplasmic tail, transmembrane domain and stem region, was ligated upstream of EndoGalC, and the conjugated gene was expressed in pig aortic endothelial cells and COS7 cells. Following the introduction of the gene, the αGal epitope on pig aortic endothelial cells was effectively reduced. Transfection studies in COS7 cells using EndoGalC combined with α1,3GT showed that the expressed EndoGalC was localized not only inside, but also outside, the cells. The expression of EndoGalC conjugated with a murine immunoglobulin (Igκ)-chain signal sequence also showed a similar effect. Conclusions : These results suggest the effectiveness of gene transfer with EndoGalC into pig endothelial cells, and strongly encourage us to produce transgenic animals with the expressed enzyme.     

Recognition of an octapeptide sequence by multiple Galα(1,3)Gal-binding proteins

Abstract: Strategies are now being designed to overcome the hyperacute rejection of vascularized pig xenografts by human natural anti-Galα(1,3)Gal antibodies. We now demonstrate that a synthetic octapeptide, Gal pep 1 (DAHWESWL), isolated from a peptide epitope library using the α-galactosyl specific lectin IB4, "mimics" the carbohydrate epitope Galα(1,3)Gal. In vitro studies demonstrated that the binding of the IB4 lectin or human natural antibodies to pig cells could be specifically blocked by Gal pep 1, in several different serological assays: a) the hemagglutnation of pig erythrocytes; b) cytofluorographic analysis of pig lymphocytes and endothelial cells; c) cytotoxicity of human serum against pig endothelial cells; d) an ELISA assay. The relative affinity of the peptide for the IB4 lectin was similar to that of α-galactosyl sugars, although it was bound at a lower affinity by human antibodies. The implications of these observations to xenotransplantation are that peptides could theoretically be used to block hyperacute rejection.     

Reduction of the major porcine xenoantigen Galoc(1,3)Gal by expression of α(1,2)fucosyltransferase

Abstract: Although removal of the Galα(1,3)Gal antigen from pigs would prevent hyperacute graft rejection, the technique of homologous recombination to knock out the α 1,3 galactosyltransferase gene is not available for pigs, and an alternative strategy is presented. As both α 1,3 galactosyltransferase and α 1,2 fucosyltransferase use the same substrate (N-acetyl lacto-samine), competition between the transferases in vitro and in vivo was examined. The data show that there is indeed a hierarchy of these gly-cosyltransferases competing for the same substrate, and that α 1,2 fuco-syltransferase takes precedence over α 1,3 galactosyltransferase: a) COS cells simultaneously transfected with cDNA clones encoding α, 2 fuco-syltransferase and α 1,3 galactosyltransferase show preferential expression of the H substance (synthesised by α 1,2fucosyltransferase) rather than Galα(1,3)Gal (synthesised by α 1,3galactosyltransferase), even though α 1,3galactosyltransferase mRNA and functional enzyme was present, b) In a pig kidney cell line that expressed both the Galα(1,3)Gal and H, the increased expression of H induced by the transfection and stable expression of α 1,2fucosyltransferase resulted in decreased expression of Galα(1,3)Gal. c) Coexpression of α 1,2fucosyltransferase and α 1,3galactosyltransferase in either COS cells or the pig cell line resulted in decreased human antibody binding and complement-mediated cell lysis, d) Transgenic mice, ubiquitously expressing α 1,2fucosyltransferase show a major decrease in Galα-(1,3)Gal expression and a decrease in natural human antibody binding. These findings have important implications for xenotransplantation in that α,2fucosyltransferase transgenic pigs could be a source of donors for xenotransplantation to humans.     

Continued production of xenoimmune antibodies 6–8 years after clinical transplantation of fetal pig islet-like cell-clusters

Abstract: We have monitored the humoral immune responses of 10 type I diabetic patients, xenotransplanted with fetal porcine islet-like cell clusters for up to 8 years after xenotransplantation. We investigated the immunoglobulin subclass distribution as well as specificity differences of xenoreactive antibodies. Hemagglutintion tests, using pig erythrocytes, showed that some patients maintained higher titers of xenoreactive IgM antibodies during the entire follow up period, compared with pretransplant levels. In microcytotoxicity tests all but one patient tested showed higher than pretransplant levels of cytotoxic antibodies against pig peripheral blood mononuclear cells (PBMC) 6–8 years after transplantation. Levels of Galα1,3Gal specific antibodies, were also high. Antibody dependent cellular cytotoxicity (ADCC) activity against a Galα1,3Gal expressing human B cell line was detected in four patients while ADCC reactivity against adult pig islet cells was detected in only two patients, 6–8 years after transplantation. Immune sera collected 30 days and 1 year after transplantation showed positive staining of adult pig islet cells in fluoromicroscopy whereas sera from later time points did not.
Western blot experiments showed that some patients had IgG1 antibodies reactive against epitopes on pig cells other than Galα1,3Gal, while xenoreactive IgM and IgG2 antibodies mainly reacted with Galα1,3Gal-containing epitopes as shown by absorption experiments.
These results show that patients continue to produce higher than pretransplant levels of IgM and IgG2 xenospecific antibodies against Galα1,3Gal for extended time periods following xenotransplantation. Some patients also produce xenoreactive IgG1 antibodies directed against non-Galα1,3Gal epitopes.     

Blocking of human anti-pig xenoantibodies by soluble GALα1–3GAL and Galα1–2GAL disaccharides; studies in a pig kidney in vitro perfusion model

Depletion of anti-pig xenoantibodies reduces cell cytotoxicity of human serum to pig endothelial cells and lymphocytes. The aim of this study was to test, in a pig kidney xenoperfusion model, the ability of soluble αGal terminated disaccharides to prevent the hyperacute rejection process in an organ. Porcine kidneys were perfused with whole human blood lacking saccharide and blood supplemented with Galα1–3GAL, Galα1–2Gal and lactose. Parameters evaluated were, urine production, renal blood flow, vascular resistance, renal clearance, blood cell counts, xenoantibody titers, complement activation and histopathology. The blood flow was higher in the Galα1–3Gal (155 ± 31 ml/min × 100 g^–1 kidney tissue) group compared to Galα1–2Gal (138 ± 16), lactose (92 ± 78) and controls (69 ± 16). When calculated as percent of the blood flow value at 1 min, the blood flow at 30 min was 157 % for the Galα1–3Gal and for 187 % the Galα1–2Gal. The corresponding values for the lactose and control groups were 102 % and 74 %, respectively. Urine production in the lactose/control groups was lower (0.7 ml/min × 100 g^–1 kidney tissue) compared to Galα1–3Gal (3.0) and Galα1–2Gal (3.7). Urine sodium excretion was reduced in the lactose/control groups, compared to the Galαα1- groups during the perfusions. An increase in urine potassium excretion was found in the Galαα1-groups while a reduction occurred in the lactose/control experiments. An initial 40–50 % reduction in platelet count was observed in all groups while the leukocyte count showed a continuous decrease. Immunohistochemistry revealed less deposition of IgM, IgG, C3 and C1 q in the Galαα1-saccharide groups compared to the lactose/control groups. Soluble Galαa1-disaccharides improved both functional and histological parameters. However, significant pathological changes were still present indicating that this approach to inhibit HAR must be used in combination with additional therapeutic approaches such as solid phase xenoantibody immunoadsorption and blocking of complement activation. Received: 19 August 1999 Revised: 3 May 2000 Accepted: 10 August 2000     

Expression of human soluble complement receptor 1 by a pig endothelial cell line inhibits lysis by human serum

The importance of complement activation and naturally occurring anti-pig antibodies in the hyperacute rejection (HAR) observed in models of pig-to-human xenotransplantation is well established. To overcome this, much effort has been dedicated to preparing transgenic pigs by knocking out Galalpha(1-3)Gal expression in these animals, or knocking in the expression of human complement regulatory proteins (CRPs), such as CD59 or decay accelerating factor. A soluble form of another membrane CRP, complement receptor type 1 (CR1), has also been shown to inhibit complement activation. Here, we show that transfection of a pig endothelial cell line with a truncated form of human soluble complement receptor 1 (sCR1) almost completely protected these cells from complement-mediated lysis by human AB serum. Pigs genetically manipulated to express human sCR1 may represent an additional strategy to inhibit HAR of pig-to-human transplanted organs.     

Detection, immunoabsorption, and inhibition of cytotoxic activity of anti-αGal antibodies using newly developed substances with synthetic Gal α1–3Gal disaccharide epitopes

Abstract: The presence of naturally occurring anti-Galα1–3Gal (anti-αGal) antibodies in human serum is believed to be a major factor in the hyperacute rejection of discordant organ xenografts such as the pig-to-human combination. Galα1–3Gal epitopes are expressed on pig tissues and the binding of anti-αGal leads to endothelial cell activation and complement-mediated, hyperacute graft rejection. One possible method to overcome this problem is to absorb anti-αGal antibodies from the plasma of the xenograft recipient using a suitable immunoabsorbent or to interfere with their binding to tissues and thus prevent their cytotoxic activity by the intravenous injection of soluble antigen. We describe here the use of new synthetic antigens containing the Galα1–3Gal disaccharide (Bdi) epitope. Soluble conjugates of the Bdi with polyacrylamide (PAA-Bdi) were used as coating antigens for an anti-αGal ELISA as well as for in vitro inhibition of the cytotoxicity of anti-αGal. An immunoabsorbent consisting of PAA-Bdi coupled to macroporous glass (Sorbent Bdi) was tested for absorption of anti-αGal from human serum. Anti-αGal IgM, IgG and IgA could be detected by the anti-αGal ELISA and were specifically absorbed by Sorbent Bdi with absorption efficiencies ranging from 70 to 50% for anti-αGal IgG and 60 to 25% for anti-αGal IgM. A comparison of the anti-αGal absorption by Sorbent Bdi and rabbit red blood cells revealed a qualitatively (isotype distribution) and quantitatively similar pattern. Nonspecific absorption by Sorbent Bdi was low, as detected by the reduction of anti-A trisaccharide antibodies. The cytotoxic effect of human serum on pig kidney (PK15) cells was almost totally inhibited by the addition of synthetic B disaccharide or by adsorption of the serum through immunoaffinity columns of PAA-Bdi. We conclude that the newly developed, synthetic αGal1–3Gal antigens are suitable for the detection and immunoabsorption or inhibition of anti-αGal antibodies.     

A study of the xenoantigenicity of adult pig islets cells

Abstract: Background: The pig pancreas is considered to be the most suitable source of islets for xenotransplantation into patients with type I diabetes. The purpose of this study was to assess the antigenicity of pig islets, including the Galα1‐3Galβ1‐4GlcNAc‐R (the α‐Gal) and Hanganutziu‐Deicher (H‐D) antigens, and the pathway involved in human complement activation. Methods: The expression of α‐Gal on islets from adult pigs was investigated by immunohistochemical staining and flowcytometric analysis. The α1,3 galactosyltransferase (α1,3GT) activity of islets was determined by high‐performance liquid chromatography. Antigenicity to human natural antibodies, including the H‐D antigen of pig islets was next examined by treatment of pig islets with tunicamycin, D‐threo‐1‐phenyl‐2‐decanoylamino‐3‐morpholino‐1‐propanol (PDMP) and/or neuraminidase. In addition, complement‐mediated islets lysis was examined using factor D‐deficient and C1‐deficient sera. Results: Adult pig islets expressed negligible amounts of α‐Gal epitope, and α1,3GT activity was also undetectable. However, human natural antibodies, immunoglobulin G and M, and the anti H‐D antibody react to the adult islet. Treatment of pig islets with tunicamycin, but not PDMP, led to a drastic reduction in antigenicity to human serum, indicating the importance of N‐linked sugars on the islets. Neuraminidase treatment indicated the presence of, not only the H‐D antigen, but also other sialic acid antigens that reacted with the human natural antibody. The complement deposition of C4, C3 and factor B on islets was demonstrated. The alternative pathway‐mediated pig islet killing accounted for approximately 30% of that by the total complement pathway. Conclusion: The origin of antigenicity of pig islets is mainly N‐linked sugars including sialic acid antigens, but not the α‐Gal, and pig islets can be injured by both the classical and the alternative complement pathway in human serum.     

Production of human monoclonal antibodies against Galalpha(1-3)Gal

The hyperacute rejection observed in models of pig-to-human xenotransplantation is mainly because of the presence of natural antibodies in human blood with specificity for the Galα(1–3)Gal (Gal) carbohydrate moiety present on the surface of porcine endothelial cells. Human monoclonal anti-Gal antibodies could be of use both in the study of the basic mechanisms of hyperacute rejection as well as in its clinical prevention. In the present study we prepared 42 heterohybridomas (human–mouse) secreting antibodies with specificity for the Gal epitope. All of the antibodies produced were of the IgM isotype, according to a dot-blot assay. Twenty-seven antibodies were further characterized, and shown to be specific for Gal by different methods, including an enzyme-linked immunosorbent assay, in which the plates were sensitized with mouse laminin as a source of Gal. Specificity was also confirmed using purified Gal carbohydrate in a hemagglutination inhibition assay. The antibodies were shown to mediate lysis of Gal-expressing rabbit erythrocytes in the presence of complement. However, the heterohybridomas themselves were shown to express Gal, a result of the mouse P3x63Ag8.653 hybridoma cells used during hybridoma generation. The presence of this epitope on the surface of anti-Gal-producing cells, and on the antibody itself, represents a limitation to the production of high affinity anti-Gal antibodies.     

Antisense inhibition of pig α1,3galactosyl-transferase leads to a reduction in expression of the major target for human natural antibodies on pig vascular endothelial cells

Abstract: The expression of epitopes on pig cells for human natural antibodies (NAb) currently constitutes a major barrier to the use of pig organs and tissues in clinical transplantation. This is of particular significance to vascularized organs where the presence of carbohydrate antigens invariably leads to a hyperacute rejection when exposed to human blood or serum. It has been strongly suggested that the major antigen in this context consists of a terminal Galα1,3Galβ1,4GlcNAc trisaccharide. We have previously proposed that decreased expression of this epitope for human NAb may lead to elimination of hyperacute rejection. We have now used mRNA targeted antisense oligonucleotides to decrease the expression of the α1,3galactosyltransferase directing the terminal synthesis of this epitope on pig vascular endothelial cells. This mRNA antisense targeting leads to a decreased expression of the Galα1,3Galβ1,4GlcNAc structure as detected by epitope specific lectin. Moreover, a very similar reduction in mean fluorescence was seen when staining the same cells with human Galα1,3Galα1,4GlcNAc-specific IgM NAb. We feel these findings constitute an important step in the process of providing conclusive evidence that reduction of or complete elimination of this epitope will overcome the hyperacute response in this species combination. In addition, such pig tissues may be less susceptible to further antibody rejection once the hyperacute phase has been overcome.     

Gal alpha 1,3Gal expression on porcine pancreatic islets, testis, spleen, and thymus

Abstract:  Gal α 1,3Gal (Gal) is the first target in antibody-mediated rejection of pig-to-non-human primate xenograft. Its expression may vary between organs and constituents of organs. Gal expression was studied in pancreas, testis, spleen and thymus of 22 pigs, with ages ranging from 1 to 22 months. The immunoperoxidase technique using the biotinylated lectin, Griffonia simplicifolia (IB4), was used. In the pancreas, neither endocrine (islet cells) nor exocrine cells expressed Gal. The Sertoli cells in the testis were negative. The spleen capsule and trabeculae did not stain for Gal, although both splenic T and B lymphocytes expressed Gal (B > T). Thymocytes were weakly positive, whereas thymic epithelial cells were negative for Gal. No age-related differences were seen in any tissues. Porcine islets of Langerhans, Sertoli cells, and the splenic and thymic structural frameworks did not express Gal, and therefore, should be relatively resistant to anti-Gal antibody-mediated rejection. The availability of pigs deficient in Gal as a source of islets may therefore not be beneficial in extending islet graft survival in non-human primate models.     

Absence of humoral and cellular alloreactivity in baboons sensitized to pig antigens

AIM: to study whether sensitization to pig antigens results in humoral and/or cellular sensitization to alloantigens in baboons, and thus increases the risks of organ allotransplantation after xenotransplantation. Serum from baboons that were naive (n = 4), sensitized to Gal alpha 1,3Gal (Gal) antigens (n = 2), or sensitized to Gal + non-Gal pig antigens (n = 2) were tested by flow cytometry for the presence of immunoglobulin G (IgG) and IgM antibodies that bind to pig or baboon peripheral blood mononuclear cells (PBMC). Two allosensitized baboons were used as positive controls. The same 10 sera were tested in a complement-mediated cytotoxicity assay to detect cytotoxic antibodies against pig, allo and self-PBMC. The T-cell responses of the same baboons to allogeneic and pig PBMC stimulators in mixed lymphocyte reaction (MLR) were studied. All baboon sera contained cytotoxic antibodies that bound to pig PBMC. Binding and cytotoxicity were higher in xenosensitized baboons, particularly in those sensitized to Gal + non-Gal antigens (P < 0.001). None of the naive or xenosensitized baboon sera bound to baboon PBMC. Serum from allosensitized baboons showed anti-baboon IgG and IgM binding, but there was no increase in binding to pig PBMC or in cytotoxicity to pig cells. The MLR response to pig stimulators in baboons sensitized to non-Gal pig antigens was greater than that of naive or Gal-sensitized baboons (P < 0.001), but there was no increase in the response to baboon cells. In baboons, no in vitro evidence that a previous pig xenograft might endanger the outcome of a subsequent allograft was documented.     

Soluble saccharides block the inhibition of agonist-induced human platelet aggregation observed after in vitro incubation of human platelet-rich plasma with porcine aortic endothelial cells

Platelet aggregation is a prominent feature in the hyperacute process of vascularized allografts and xenografts. In a study of extracorporeal connection of pig kidneys to the blood circulation of human volunteers, we observed in one case considerable destruction of human platelets in the pig kidney without signs of hyperacute rejection or microthrombi formation. In the present study, we have investigated the agonist-induced aggregation of human platelets in mixtures with porcine aortic endothelial cells (PAEC). In vitro incubation of human platelet-rich plasma (PRP) with PAEC inhibited platelet aggregation induced by ADP, collagen and arachidonic acid in a time-dependent manner and partially inhibited adrenalin-induced aggregation. Aggregation of the human platelets could not be induced by high concentrations of ADP (20 μM) to overcome the inhibition capacity of the PAEC. The PAEC inhibiting effect could be transferred by the supernatants of PAEC/PRP and PAEC/PPP incubation mixtures. Preincubation of the PAEC with aspirin, but not with N^G-methyl-L-Arg, reduced the aggregation inhibitory effect. Control experiments mixing human umbilical vein endothelial cells (HUVEC) and human PRP or mixing porcine PRP and PAEC did not elicit any inhibition of ADP-induced platelet aggregation. The aggregation inhibition effect could partially be blocked by preincubation of PRP with soluble Galα1–3Gal, Galα1–3β1–4GlcNAc, lactose, galactose, and glucose, but not by lactosamine, galactosamine, or glucosamine. The Galα1–3Gal disaccharide was most effective in blocking aggregation inhibition, and to a similar extent as its ability to block the human anti-pig lymphocytotoxicity reaction. In conclusion, the data indicate that PAEC, upon stimulation by human anti-pig xenoantibodies in a nondynamic system, inhibits agonist-induced human platelet aggregation, and that this effect is probably at least partially caused by prostacyclin released from the PAEC. Received: 12 August 1997 Received after revision: 19 March 1998 Accepted: 15 April 1998     

Reduction in Gal-α1,3-Gal epitope expression in transgenic mice expressing human H-transferase

Abstract: There is now considerable evidence implicating anti-Gal xenoantibodies as a key instigator in the hyperacute rejection of discordant xenografts. As a consequence it is generally held that elimination or reduction of the Gal/anti-Gal component is critical to overcoming hyperacute rejection. We have recently shown that in mice inactivation of the GalT gene by homologous recombination completely eliminates the expression of the Gal-epitope and that hearts from these mice demonstrate prolonged survival when perfused ex vivo with human plasma. Unfortunately this strategy is currently not feasible in pigs because the technology to isolate porcine embryonic stem cells, which are critical for homologous recombination, is not yet available. This study investigates an alternative competition-based transgenic strategy to suppress the level of the Gal epitope by expression of H-transferase (α1,2-fucosyltransferase) an enzyme which has the same substrate specificity (lactobiose) as α1,3-galactosyltransferase. In vitro transfection of murine cells with H-transferase reduced Gal-epitope expression by 80–90%. A similar reduction in Gal expression was observed on PBL and thymocytes from H-transferase transgenic mice. This reduction in Gal epitope expression resulted in a marked reduction in the reactivity of these cells with human serum. In tissues from these mice the reduction in Gal expression was inversely proportional to the endogenous level of Gal. The results of this study support pursuing this strategy as a means to reduce the xenoantigenicity of porcine tissues.     

Human DAF on pig cells protects against human and non-human primate sera cytotoxicity mediated by exogenous or endogenous complement, as determined by flow cytometry

Expression of human complement regulatory proteins (CRP) in pig cells through transgenesis was proposed to prevent complement activation and the ensuing rejection of pig tissues and organs following pig-to-primate transplantation. Transplantation in non-human primates of organs from transgenic pigs for human decay accelerating factor (hDAF) did not undergo hyperacute rejection, but hDAF could not prevent humoral xenograft rejection (AHXR). A possible explanation for the lack of efficacy of the expression of human complement regulatory proteins in pig cells to prevent AHXR may be interspecies differences between human and non-human complement regulatory system. We assayed the efficacy of transgenic hDAF expressed on porcine cells to inhibit the in vitro complement activity of primate sera. The individual cytotoxicity of sera from seven untreated baboons and of pools of normal human and baboon sera was assayed with endogenous and exogenous complement using a flow-cytometry complement-mediated cytotoxicity assay (FCCA) against peripheral blood lymphocytes (PBL) from hDAF and non-transgenic pigs. We also analyzed the anti-Galalpha1-3Gal (alphaGal) antibody titre of the baboon sera by ELISA and the expression of hDAF on the PBL surface by immunofluorescence. Transgenic hDAF expression was capable of protecting pig cells against injury produced by both baboon and human serum. Cellular expression of hDAF reduced cytotoxicity mediated by endogenous and exogenous complement, although the former was slightly higher. Humoral cytotoxicity was not related to a particular antibody but was inversely related to hDAF expression. The presence of hDAF protected pig cells against lysis by NHS more effectively than against NBS. These results confirm in vitro the protective role of hDAF in pig cells to heterologous complement mediated damage, but they also suggest that the extent of hDAF protection decreases, however, if cells express low levels of hDAF.     

Natural protection from zoonosis by alpha-gal epitopes on virus particles in xenotransmission

Clinical transplantation has become one of the preferred treatments for end-stage organ failure, and one of the novel approaches being pursued to overcome the limited supply of human organs involves the use of organs from other species. The pig appears to be a near ideal animal due to proximity to humans, domestication, and ability to procreate. The presence of Gal-alpha1,3-Gal residues on the surfaces of pig cells is a major immunological obstacle to xenotransplantation. Alpha1,3galactosyltransferase (alpha1,3GT) catalyzes the synthesis of Gal alpha 1-3Gal beta 1-4GlcNAc-R (alpha-gal epitope) on the glycoproteins and glycolipids of non-primate mammals, but this does not occur in humans. Moreover, the alpha-gal epitope causes hyperacute rejection of pig organs in humans, and thus, the elimination of this antigen from pig tissues is highly desirable. Recently, concerns have been raised that the risk of virus transmission from such pigs may be increased due to the absence of alpha-gal on their viral particles. In this study, transgenic cells expressing alpha1,3GT were selected using 1.25 mg/ml neomycin. The development of HeLa cells expressing alpha1,3GT now allows accurate studies to be conducted on the function of the alpha-gal epitope in xenotransmission. The expressions of alpha-gal epitopes on HeLa/alpha-gal cells were demonstrated by flow cytometry and confocal microscopy using cells stained with IB4-fluorescein isothiocyanate lectin. Vaccinia viruses propagated in HeLa/alpha-gal cells also expressed alpha-gal on their viral envelopes and were more sensitive to inactivation by human sera than vaccinia virus propagated in HeLa cells. Moreover, neutralization of vaccinia virus was inhibited in human serum by 10 mm ethylene glycol bis(beta-aminoethylether)tetraacetic acid (EDTA) treatment. Our data indicated that alpha-gal epitopes are one of the major barriers to zoonosis via xenotransmission.     

Transplantation of mouse pancreatic islets into primates--in vivo and in vitro evaluation.

BACKGROUND: Islets transplanted from other species to man has the potential to cure diabetes but whether islets are subject to hyperacute rejection after xenotransplantation is contentious. We transplanted mouse pancreatic islets of mouse beneath the primate renal capsule and assessed natural xenoantibody binding, complement activation and cell lysis in vitro. METHODS: Freshly isolated mouse islets were transplanted in a blood clot under the renal capsule of cynolmogus monkeys. The graft was removed after 24 hr for histological and ultrastructural analysis. Freshly isolated mouse pancreatic islets were analyzed in vitro by immunohistochemistry for Gal(alpha1,3)Gal and Von Willebrand factor expression and for IgG, IgM, C3, C4, and C5b-9 binding after incubation in 100% human serum. Complement mediated cell lysis was evaluated by 51Cr release assays after incubation of islets for 4 hr in human serum, plasma, and lymph with and without added neutrophils. RESULTS: Mouse islets transplanted under the renal capsule of cynomolgus monkeys were destroyed within 24 hr by a process involving necrosis with neutrophil and mononuclear cell infiltration. Gal(alpha1,3)Gal was strongly positive on only 10% of islet cells. After islet incubation in 100% human serum before frozen section, human IgG and IgM, C3, C4, and C5b-9 was deposited on islets with increased intensity in the periphery. Measurement of 51Cr release from labeled fresh islets after four hours incubation in 100% human serum showed 17% lysis and was not changed by addition of neutrophils. CONCLUSION: These results indicate that mouse islets in a primate recipient undergo rapid destruction by a process that has features similar to hyperacute rejection in vascularized organs and we propose the same term be used.