Significance of histochemical expression of Hanganutziu-Deicher antigens in pig, baboon and human tissues

Design and testing of a humanized porcine donor for xenotransplantation

Recent human decedent model studies1,2 and compassionate xenograft use3 have explored the promise of porcine organs for human transplantation. To proceed to human studies, a clinically ready porcine donor must be engineered and its xenograft successfully tested in nonhuman primates. Here we describe the design, creation and long-term life-supporting function of kidney grafts from a genetically engineered porcine donor transplanted into a cynomolgus monkey model. The porcine donor was engineered to carry 69 genomic edits, eliminating glycan antigens, overexpressing human transgenes and inactivating porcine endogenous retroviruses. In vitro functional analyses showed that the edited kidney endothelial cells modulated inflammation to an extent that was indistinguishable from that of human endothelial cells, suggesting that these edited cells acquired a high level of human immune compatibility. When transplanted into cynomolgus monkeys, the kidneys with three glycan antigen knockouts alone experienced poor graft survival, whereas those with glycan antigen knockouts and human transgene expression demonstrated significantly longer survival time, suggesting the benefit of human transgene expression in vivo. These results show that preclinical studies of renal xenotransplantation could be successfully conducted in nonhuman primates and bring us closer to clinical trials of genetically engineered porcine renal grafts.

Xenogeneic cross-circulation for physiological support and recovery of ex vivo human livers

BACKGROUND AND AIMS

The scarcity of suitable donor livers highlights a continuing need for innovation to recover organs with reversible injuries in liver transplantation.

APPROACH AND RESULTS

Explanted human donor livers (n = 5) declined for transplantation were supported using xenogeneic cross-circulation of whole blood between livers and xeno-support swine. Livers and swine were assessed over 24 hours of xeno-support. Livers maintained normal global appearance, uniform perfusion, and preservation of histologic and subcellular architecture. Oxygen consumption increased by 75% ( p = 0.16). Lactate clearance increased from -0.4 ± 15.5% to 31.4 ± 19.0% ( p = 0.02). Blinded histopathologic assessment demonstrated improved injury scores at 24 hours compared with 12 hours. Vascular integrity and vasoconstrictive function were preserved. Bile volume and cholangiocellular viability markers improved for all livers. Biliary structural integrity was maintained.

CONCLUSIONS

Xenogeneic cross-circulation provided multisystem physiological regulation of ex vivo human livers that enabled functional rehabilitation, histopathologic recovery, and improvement of viability markers. We envision xenogeneic cross-circulation as a complementary technique to other organ-preservation technologies in the recovery of marginal donor livers or as a research tool in the development of advanced bioengineering and pharmacologic strategies for organ recovery and rehabilitation.

Immune characterization of a xenogeneic human lung cross-circulation support system

Improved approaches to expanding the pool of donor lungs suitable for transplantation are critically needed for the growing population with end-stage lung disease. Cross-circulation (XC) of whole blood between swine and explanted human lungs has previously been reported to enable the extracorporeal recovery of donor lungs that declined for transplantation due to acute, reversible injuries. However, immunologic interactions of this xenogeneic platform have not been characterized, thus limiting potential translational applications. Using flow cytometry and immunohistochemistry, we demonstrate that porcine immune cell and immunoglobulin infiltration occurs in this xenogeneic XC system, in the context of calcineurin-based immunosuppression and complement depletion. Despite this, xenogeneic XC supported the viability, tissue integrity, and physiologic improvement of human donor lungs over 24 hours of xeno-support. These findings provide targets for future immunomodulatory strategies to minimize immunologic interactions on this organ support biotechnology.

The Structural Complexity and Animal Tissue Distribution of N-Glycolylneuraminic Acid (Neu5Gc)-Terminated Glycans. Implications for Their Immunogenicity in Clinical Xenografting

N-Glycolylneuraminic acid (Neu5Gc)-terminated glycans are present in all animal cells/tissues that are already used in the clinic such as bioprosthetic heart valves (BHV) as well as in those that potentially will be xenografted in the future to overcome end stage cell/organ failure. Humans, as a species lack this antigen determinant and can react with an immune response after exposure to Neu5Gc present in these products/cells/tissues. Genetically engineered source animals lacking Neu5Gc has been generated and so has animals that in addition lack the major αGal xenoantigen. The use of cells/tissues/organs from such animals may improve the long-term performance of BHV and allow future xenografting. This review summarizes the present knowledge regarding structural complexity and tissue distribution of Neu5Gc on glycans of cells/tissue/organs already used in the clinic or intended for treatment of end stage organ failure by xenografting. In addition, we briefly discuss the role of anti-Neu5Gc antibodies in the xenorejection process and how knowledge about Neu5Gc structural complexity can be used to design novel diagnostics for anti-Neu5Gc antibody detection.

Human antibody reactivity against xenogeneic N-glycolylneuraminic acid and galactose-α-1,3-galactose antigen

BACKGROUND

Despite the development of α1,3-galactosyl transferase-knockout (GTKO) pigs, acute humoral xenograft rejection caused by antibodies against non-Gal antigens, along with complement activation, are hurdles that need to be overcome. Among non-Gal antigens, N-glycolylneuraminic acid (Neu5Gc) is considered to play an important role in xenograft rejection in human.

METHODS

We generated human embryonic kidney 293 (HEK293) cells that expressed xenogeneic Neu5Gc (HEK293-pCMAH) or α1,3Gal (HEK293-pGT) antigen and investigated the degree of human antibody binding and complement-dependent cytotoxicity (CDC) against these antigens using 100 individual human sera.

RESULTS

Both IgM and IgG bound to α1,3Gal, while only IgG bound to Neu5Gc. Of the ABO blood groups, the degree of IgG binding to α1,3Gal was highest for blood group A. The degree of CDC against HEK293-pCMAH cells was significantly lower than that against HEK293-pGT cells. However, CDC against HEK293-pCMAH cells was significantly higher than that against control HEK293 cells. In addition, the severity of CDC against HEK293-pCMAH cells positively correlated with that against GTKO pig aortic endothelial cells (PAECs), suggesting that Neu5Gc is the main antigen in GTKO PAECs. Similar to antibody-binding activity, only IgG binding correlated with CDC against HEK293-pCMAH cells. The most common subclass of IgGs against Neu5Gc was IgG1, which typically induces strong complement activation.

CONCLUSIONS

We showed that IgG-mediated CDC was detected in Neu5Gc-overexpressed HEK293 cells incubated with human sera; however, this antibody reactivity to Neu5Gc was highly variable among individuals. Our results suggest that additional modifications to the CMAH gene should be considered for widespread use of pig organs for human transplants.

A pig-to-mouse coronary artery transplantation model for investigating the pathogenicity of anti-pig antibody

BACKGROUND

Rejection of Gal-free (GTKO) donor pig cardiac xenografts is strongly associated with vascular non-Gal antibody binding, endothelial cell (EC) injury, and activation and microvascular thrombosis. We adopted a pig-to-SCID/beige small animal transplant model to compare the pathogenicity of baboon and human anti-pig antibody.

METHODS

Wild-type (GT(+) ) or GTKO porcine coronary arteries (PCAs) were transplanted into the infrarenal aorta of SCID/beige mice. Three days after transplant, recipients were infused with anti-pig antibody (anti-SLA class I, an isotype control, naive or sensitized baboon serum, or naive human serum). PCAs were recovered 24 h after antibody infusion and examined using histology, immunohistochemistry, and in situ hybridization.

RESULTS

Dose-dependent intragraft thrombosis occurred after infusion of anti-SLA I antibody (but not isotype control) in GT(+) and GTKO PCA recipients. Naive baboon serum induced thrombosis in GT(+) grafts. Thrombosis was significantly reduced by pre-treating naive baboon serum with Gal polymer and not observed when this serum was infused to GTKO PCA recipients. Naive human serum caused dose-dependent intragraft thrombosis of GTKO PCAs. In all cases, thrombosis involved graft-specific vascular antibody and complement deposition, macrophage adherence, EC delamination, and subendothelial thrombus formation.

CONCLUSIONS

This study provides the first direct in vivo comparison of the pathogenicity of naive human and baboon serum. The results suggest that human preformed non-Gal antibody may have increased pathogenicity compared to baboon. This model, which showed a rejected graft histopathology similar to antibody-mediated rejection in cardiac xenotransplantation, may be useful to assess the pathogenicity of individual protein or carbohydrate specific non-Gal reactive antibodies.

Comparative N-linked glycan analysis of wild-type and α1,3-galactosyltransferase gene knock-out pig fibroblasts using mass spectrometry approaches

Carbohydrate antigens expressed on pig cells are considered to be major barriers in pig-to-human xenotransplantation. Even after α1,3-galactosyltransferase gene knock-out (GalT-KO) pigs are generated, potential non-Gal antigens are still existed. However, to the best of our knowledge there is no extensive study analyzing N-glycans expressed on the GalT-KO pig tissues or cells. Here, we identified and quantified totally 47 N-glycans from wild-type (WT) and GalT-KO pig fibroblasts using mass spectrometry. First, our results confirmed the absence of galactose-alpha-1,3-galactose (α-Gal) residue in the GalT-KO pig cells. Interestingly, we showed that the level of overall fucosylated N-glycans from GalT-KO pig fibroblasts is much higher than from WT pig fibroblasts. Moreover, the relative quantity of the N-glycolylneuraminic acid (NeuGc) antigen is slightly higher in the GalT-KO pigs. Thus, this study will contribute to a better understanding of cellular glycan alterations on GalT-KO pigs for successful xenotransplantation.

Production of biallelic CMP-Neu5Ac hydroxylase knock-out pigs

After the knock-out (KO) of α1,3 galactosyltransfease (Gal-T), the Hanganutziu-Deicher antigen became a major antigen of the "non-Gal antigen" that is implicated in subsequent xenograft rejection. For deletion of non-Gal antigen, we successfully produced zinc finger nuclease (ZFN)-mediated monoallelic/biallelic male and female CMP-N-acetylneuraminic acid hydroxylase (CMAH) KO miniature pigs: the efficiency of the gene targeting (41.7%) was higher when donor DNA was used with the ZFN than those of ZFN alone (9.1%). Monoallelic KO pigs had no integration of exogenous DNA into their genome, indicating that this technique would provide a new avenue to reduce the risk of antibiotics resistance when organs from genetically modified pigs are transplanted into patients. Until now, both monoallelic and biallelic CMAH KO pigs are healthy and show no sign of abnormality and off-target mutations. Therefore, these CMAH null pigs on the Gal-T KO background could serve as an important model for the xenotransplantation.

The alpha1,3GalT knockout/alpha1,2FucT transgenic pig does not appear to have an advantage over the alpha1,3GalT knockout pig with respect to glycolipid reactivity with human serum antibodies

BACKGROUND

The human H-transferase (α2FucT) was introduced in Gal-negative pigs to produce pig organs not only free from Gal-antigens, but also in which the uncapped N-acetyllactosamine precursor had been transformed into non-xenogenic blood group H type 2 compounds. This work is the first descriptive analysis of glycolipids from the GalT-KO/FucT-TG pig. The aim was to investigate the cell membrane antigens in GalT-KO/FucT-TG tissues to explore its efficacy as an organ donor. Also, detailed knowledge on the correlation between the cellular glycosyltransferase configuration and the resulting carbohydrate phenotype expression is valuable from a basic glycobiological perspective.

METHODS

Neutral and acidic glycolipids from GalT-KO/FucT-TG small intestine were compared with glycolipids from two wildtype and two GalT-KO pig intestines. Glycolipid reactivity was tested on thin layer chromatography plates using chemical reagents, antibodies, lectins, and human serum. Structural characterization of neutral glycolipids was performed by LC-ESI/MS and proton NMR spectroscopy.

RESULTS

Characterization of the glycolipid expression in GalT-KO/FucT-TG intestine showed absence of Gal antigens and decreased/unchanged levels of the N-acetyllactosamine precursor and the blood group H type 2 expression, when compared with the wildtype. The reactivity of human serum antibodies to GalT-KO/FucT-TG derived glycolipids was similar or slightly elevated when compared with GalT-KO glycolipids. Results from LC-ESI/MS and proton NMR spectroscopy revealed no established neutral xenogenic antigens in the GalT-KO/FucT-TG pig, and could thus not explain the immunologic reactivity to human serum antibodies. The antibody binding to acidic glycolipids is most likely to be explained by the abundance of N-glycolylneuraminic acid epitopes in pig tissues. Six neutral complex biantennary glycolipids with blood group H type 1, 2, Lewis(x) and Lewis(y) determinants were found, of which three were identified in this work for the first time. One of these was a nonaglycosylceramide with blood group H type 2 and lactosyl determinants linked to a lactotetraosyl core, and the other two were decaglycosylceramides with blood group H type 1 and H type 2 determinants linked to a neolactotetraosyl core, and Lewis(x) and blood group H type 1 determinants on a lactotetraosyl core, respectively.

CONCLUSIONS

Lipid-linked carbohydrate antigens in the GalT-KO/FucT-TG pig intestine showed no or minor qualitative difference when compared with GalT-KO pigs. The GalT-KO/FucT-TG pig did not appear to have an advantage over the GalT-KO pig with respect to reactivity with human antibodies from a xenotransplantation perspective.

The sweets standing at the borderline between allo- and xenotransplantation

Animal cells are densely covered with glycoconjugates, such as N-glycan, O-glycan, and glycosphingolipids, which are important for various biological and immunological events at the cell surface and in the extracellular matrix. Endothelial α-Gal carbohydrate epitopes (Galα3Gal-R) expressed on porcine tissue or cell surfaces are such glycoconjugates and directly mediate hyperacute immunological rejection in pig-to-human xenotransplantation. Although researchers have been able to develop α1,3-galactosyltransferase (GalT) gene knockout (KO) pigs, there remain unclarified non-Gal antigens that prevent xenotransplantation. Based on our expertise in the structural analysis of xenoantigenic carbohydrates, we describe the immunologically significant non-human carbohydrate antigens, including α-Gal antigens, analyzed as part of efforts to assess the antigens responsible for hyperacute immunological rejection in pig-to-human xenotransplantation. The importance of studying human, pig, and GalT-KO pig glycoprofiles, and of developing adequate pig-to-human glycan databases, is also discussed.

Deficiency of N-glycolylneuraminic acid and Galα1-3Galβ1-4GlcNAc epitopes in xenogeneic cells attenuates cytotoxicity of human natural antibodies

BACKGROUND

A number of carbohydrate structures apart from Galα1-3Galβ1-4GlcNAc (Gal) have been implicated as potential xenoantigens. Epitopes of another carbohydrate structure, namely N-glycolylneuraminic acid (NeuGc), are widely expressed on the surfaces of endothelial cells of all mammals, except humans, and are likely targets of anti-non-Gal antibodies (Abs). The purpose of this study is to assess whether deficiency of NeuGc and Gal epitopes in xenogeneic cells attenuates the cytotoxicity of naturally occurring antibodies in human sera.

METHODS

We generated mice deficient in both α1,3-galactosyltransferase (GalT) and cytidine monophospho-N-acetylneuraminic acid hydroxylase (CMAH). Flow cytometric and immunohistochemical studies confirmed the complete absence of Gal and NeuGc expression in hematopoietic cells and tissue sections of the heart, lungs, liver, kidney, and pancreas in GalT(-/-) CMAH(-/-) double knockout (DKO) mice. The thymocytes obtained from wild-type (WT), GalT(-/-), CMAH(-/), and DKO mice were used to capture xenoreactive Abs present in human sera from healthy volunteers of each blood group. The titers and complement-dependent cytotoxicity (CDC) of these Abs were determined by flow cytometry and the (51) Cr release assay, respectively.

RESULTS

In all sera, IgM and IgG Abs bound to the thymocytes of WT, GalT(-/-), CMAH(-/-), and DKO mice. The average values of IgM Abs against thymocytes of WT and CMAH(-/-) mice were similar, but were statistically higher than those against thymocytes of GalT(-/-) and DKO mice. The average value of IgG Abs against WT mouse thymocytes was similar to that against GalT(-/-) mouse thymocytes, but was significantly higher than that against thymocytes of CMAH(-/-) and DKO mice. Remarkable CDC of human sera was observed against the thymocytes of WT and CMAH(-/-) mice, whereas thymocytes of GalT(-/-) and DKO mice were more resistant to lysis. CDC of human sera against CMAH(-/-) mouse thymocytes was significantly lower than that against WT mouse thymocytes. In addition, CDC against DKO mouse thymocytes, which appeared undetectable even at lesser serum dilutions, was significantly lower than that against GalT(-/-) mouse thymocytes.

CONCLUSIONS

A DKO mice strain lacking both Gal and NeuGc antigens in several tissues has been generated. Comparing CDC of human sera against mouse thymocytes, the DKO strain shows a statistically significant but a small additional reduction in CDC compared to that already achieved in GalT(-/-) mice.

Investigation of potential carbohydrate antigen targets for human and baboon antibodies

BACKGROUND

The continued presence of a primate antibody-mediated response to cells and organs from alpha1,3-galactosyltransferase gene-knockout (GTKO) pigs indicates that there may be antigens other than Gal alpha 1,3Gal (alpha Gal) against which primates have xenoreactive antibodies. Human and baboon sera were tested for reactivity against a panel of saccharides that might be potential antigen targets for natural anti-non-alpha Gal antibodies.

METHODS

Human sera (n = 16) and baboon sera (n = 15) of all ABO blood types were tested using an enzyme-linked immunoadsorbent assay for binding of IgM and IgG to a panel of synthetic polyacrylamide-linked saccharides (n = 15). Human sera were also tested after adsorption on alpha Gal immunoaffinity beads. Sera from healthy wild-type (WT, n = 6) and GTKO (n = 6) pigs and from baboons (n = 4) sensitized to GTKO pig organ or artery transplants (of blood type O) were also tested. Forssman antigen expression on baboon and pig tissues was investigated by immunohistochemistry.

RESULTS

Both human and baboon sera showed high IgM and IgG binding to alpha Gal saccharides, alpha-lactosamine, and Forssman disaccharide. Human sera also demonstrated modest binding to N-glycolylneuraminic acid (Neu5Gc). When human sera were adsorbed on alpha Gal oligosaccharides, there was a reduction in binding to alpha Gal and alpha-lactosamine, but not to Forssman. WT and GTKO pig sera showed high binding to Forssman, and GTKO pig sera showed high binding to alpha Gal saccharides. Baboon sera sensitized to GTKO pigs showed no significant increased binding to any specific saccharide. Staining for Forssman was negative on baboon and pig tissues.

CONCLUSIONS

We were unable to identify definitively any saccharides from the selected panel that may be targets for primate anti-non-alpha Gal antibodies. The high level of anti-Forssman antibodies in humans, baboons, and pigs, and the absence of Forssman expression on pig tissues, suggest that the Forssman antigen does not play a role in the primate immune response to pigs.

Brief report: a new profile of terminal N-acetyllactosamines glycans on pig red blood cells and different expression of alpha-galactose on Sika deer red blood cells and nucleated cells

It has been reported that: (1) large variations were found in the number of sialic acid (SA) capped with N-acetyllactosamines (SA-Galbeta1-4GlcNAc-R) and alpha-Gal epitopes (Galalpha1-3Galbeta1-4GlcNAc-R) or uncapped N-acetyllactosamines (Galbeta1-4GlcNAc-R) on different mammalian red blood cells, and on nucleated cells originating from a given tissue in various species; (2) goat, sheep, horse and mouse red blood cells lack alpha-Gal epitopes, despite the expression of this epitope on a variety of nucleated cells in these species, including lymphocytes differentiated from the same hematopoietic origin. In this study, flow cytometry and Western blot analyses of pig red blood cells showed that alpha-Gal epitopes on pig red cells developed concomitantly after treatment with neuraminidase, suggesting that the terminal N-acetyllactosaminide glycans were capped with SA-alpha-Gal epitopes. Whereas, the expression of the alpha-Gal epitopes on red blood cells from Sika deer (Cevus nippon hortulorum) were found to be absent even though the epitopes were present on their white blood cells. Thus, these results add new data not only for the terminal carbohydrate structures on cell surface glycans of various mammalian cells, but also for wide variety of epitope expression on the cells from different tissues, which might be useful for understanding their unique states resulting from differentiation and evolution.

Structural characterization of alpha1,3-galactosyltransferase knockout pig heart and kidney glycolipids and their reactivity with human and baboon antibodies

BACKGROUND

alpha1,3-galactosyltranferase knockout (GalT-KO) pigs have been established to avoid hyperacute rejection in GalT-KO pig-to-human xenotransplantation. GalT-KO pig heart and kidney glycolipids were studied focusing on elimination of Gal-antigens and whether novel antigens would appear. Non-human primates are used as pre-clinical transplantation experimental models. Therefore, sera from baboons transplanted with GalT-KO hearts were compared with human serum regarding reactivity with pig glycolipids.

METHODS

Neutral and acidic glycolipids were isolated from GalT-KO and WT pig hearts and kidneys. Glycolipid immune reactivity was tested on TLC plates using human affinity-purified anti-Gal Ig, anti-blood group monoclonal antibodies, lectins, and human serum as well as baboon serum collected before and after GalT-KO pig heart transplantations. Selected glycolipid fractions, isolated by HPLC, were structurally characterized by mass spectrometry and proton NMR spectroscopy.

RESULTS

GalT-KO heart and kidney lacked alpha3Gal-terminated glycolipids completely. Levels of uncapped N-acetyllactosamine precursor compounds, blood group H type 2 core chain compounds, the P1 antigen and the x(2) antigen were increased. Human serum antibodies reacted with Gal-antigens and N-glycolylneuraminic acid (NeuGc) in WT organs of which only the NeuGc reactivity remained in the GalT-KO tissues. A clear difference in reactivity between baboon and human antibodies with pig glycolipids was found. This was most pronounced for acidic, not yet identified, compounds in GalT-KO organs which were less abundant or lacking in the corresponding WT tissues.

CONCLUSIONS

GalT-KO pig heart and kidney completely lacked Gal glycolipid antigens whilst glycolipids synthesized by competing pathways were increased. Baboon and human serum antibodies showed a different reactivity pattern to pig glycolipid antigens indicating that non-human primates have limitations as a human pre-clinical model for immune rejection studies.

Cloning and functional characterization of pig CMP-N-acetylneuraminic acid hydroxylase for the synthesis of N-glycolylneuraminic acid as the xenoantigenic determinant in pig–human xenotransplantation

In the present study, the pig CMP-N-acetylneuraminic acid hydroxylase gene (pcmah), a key enzyme for the synthesis of NeuGc (N-glycolylneuraminic acid), was cloned from pig small intestine and characterized. The ORF (open reading frame) of pcmah was 1734 bp, encoding 577 amino acids and consisting of 14 exons. Organ expression pattern analysis reveals that pcmah mRNA is mainly expressed in pig rectum, tongue, spleen and colon tissues, being the most highly expressed in small intestine. In the ectopic expression of pcmah, when pig kidney PK15 cells and human vascular endothelial ECV304 cells were transfected with the cloned pcmah, the NeuGc contents of these transfectants were greater in comparison with vector transfectants used as controls. In addition, in the functional analysis of NeuGc, HSMC (human-serum-mediated cytotoxicity) was elevated in the ectopic NeuGc-expressing pcmah-transfected cells compared with controls. Moreover, binding of human IgM to the pcmah-transfected cells was significantly increased, whereas binding of IgG was slightly increased, indicating that the human IgM type was a major anti-NeuGc antibody. Furthermore, pcmah silencing by shRNA (short hairpin RNA) resulted in a decrease in NeuGc content and xenoantigenicity in PK15. From the results, it was concluded that the pcmah gene was capable of synthesizing the NeuGc acting as a xenoantigen in humans, confirming the NeuGc-mediated rejection response in pig–human xenotransplantation.

Immune responses to alpha1,3 galactosyltransferase knockout pigs

PURPOSE OF REVIEW

To summarize the current knowledge of the immune response generated against xenografts stemming from alpha1,3-galactosyltransferase knockout (GalT-KO) pigs. In particular, we will address the nature of potentially remaining Gal epitopes, the role of non-Gal xenoantigens, and the cellular response directed against GalT-KO tissues.

RECENT FINDINGS

New findings support the view that porcine cells do not express isoglobotrihexosylceramide 3, and GalT-KO pigs, if at all, express negligible levels of Gal. The anti-non-Gal antibody response to GalT-KO cells allowed the identification of several potentially relevant porcine xenoantigens, mainly carbohydrates. Coculture of wildtype pig aortic endothelial cells but not of GalT-KO pig aortic endothelial cells with whole human blood induces the secretion of porcine and human cytokines and the upregulation of E-selectin; in contrast, the transmigration of human leukocytes across porcine endothelium is not regulated by Gal.

SUMMARY

New immunological problems are arising after the elimination of Gal by the generation of GalT-KO pigs; these include non-Gal antibodies and the identification of their elusive antigens, as well as cellular components of the immune system, including neutrophils, macrophages, natural killer cells, and T cells.

Qualitative and quantitative comparison of N-glycans between pig endothelial and islet cells by high-performance liquid chromatography and mass spectrometry-based strategy

N-glycan structures released from miniature pig endothelial and islet cells were determined by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF), negative ion electrospray ionization (ESI) MS/MS and normal-phase high performance liquid chromatography (NP-HPLC) combined with exoglycosidase digestion. Totally, the identified structures were 181 N-glycans including 129 sialylated and 18 alpha-galactosylated glycans from pig endothelial cells and 80 N-glycans including 41 sialylated and one alpha-galactosylated glycans from pig islet cells. The quantity of the alpha-galactosylated glycans from pig islet cells was certainly neglectable compared to pig endothelial cells. A number of NeuGc-terminated N-glycans (80 from pig endothelial cells and 13 from pig islet cells) are newly detected by our mass spectrometric strategies. The detailed structural information will be a matter of great interest in organ or cell xenotransplantation using alpha 1,3-galactosyltransferase gene-knockout (GalT-KO) pig.

Structural analysis of alpha-Gal and new non-Gal carbohydrate epitopes from specific pathogen-free miniature pig kidney

The major barrier in transplantation of pig organs into humans is the presence of surface carbohydrate antigens (e.g., the Gal alpha 1-3 Gal beta 1-4GlcNAc-R (alpha-Gal) epitope) expressed on pig endothelial cells. In this study, total N-glycans from membrane glycoproteins derived from specific pathogen-free miniature pig kidney are identified by MALDI-TOF, negative ion ESI MS/MS and normal-phase HPLC (NP-HPLC) combined with exoglycosidase digestion. Over 100 N-glycans, including sialylated and neutral types, were identified. As well as the known alpha-Gal antigens, some of these glycans contained novel non-Gal carbohydrate antigens such as (Neu5Gc-Gal-GlcNAc) and Gal alpha 1-3 Lewis(x) (Gal-Gal-(Fuc)GlcNAc) which have not been reported before in N-glycans from pig organs. The ability of MALDI, ESI, and HPLC to measure the relative proportions of the glycans was evaluated. The HPLC resolution was insufficient for accurate work and some minor differences were noted in the ionization efficiencies of different glycan groups when measured by the two mass spectrometric techniques. However, the results indicated that the relative quantity of alpha-Gal epitope was in the region of 50% of the complex glycans. High-mannose type glycans were also abundant (35-43%) but appeared to be ionized more efficiently than the complex glycans by ESI than by MALDI.

Studies on glycolipid antigens in small intestine and pancreas from alpha1,3-galactosyltransferase knockout miniature swine

BACKGROUND

To avoid hyperacute rejection of xeno-organs, alpha1,3-galactosyltransferase knockout (GalT-KO) pigs have been produced. Galalpha1,3Gal determinant elimination may expose cryptic carbohydrate antigens and/or generate new antigens. This is the first biochemical study of carbohydrate antigens in GalT-KO pig organs.

METHODS

Neutral and acidic glycolipids were isolated from small intestine and pancreas of two GalT-KO and one wild-type (WT) pig. Glycolipid immune reactivity was tested on thin-layer chromatograms. Small intestine neutral glycolipids were separated by high-performance liquid chromatography and selected fractions were analyzed by proton nuclear magnetic resonance spectroscopy. Total gangliosides were quantified on thin-layer chromatograms and in microtiter wells.

RESULTS

Using Galalpha1,3nLc4 glycolipid reference, total Galalpha1,3Gal glycolipid antigens in the WT animal was estimated at about 30 microg (small intestine) and 3 microg (pancreas) per gram of dry tissue. Galalpha1,3Gal determinants were not detected in GalT-KO tissues at a detection limit of less than 0.25% (small intestine) and 0.5% (pancreas) of the WT tissues. Isoglobotriaosylceramide (iGb3) was absent but trace amounts of Fuc-iGb3 was found in both GalT-KO and WT pig small intestine. Blood group H type 2 core saccharide compounds were increased in GalT-KO pancreas. Total amount of gangliosides was decreased in GalT-KO tissues. The alpha1,3-galactosyltransferase acceptor, N-acetyllactosamine determinant, was not increased in GalT-KO tissues. Human serum antibodies reacted with WT organ Galalpha1,3Gal antigens and gangliosides, of which the ganglioside reactivity remained in GalT-KO tissues.

CONCLUSIONS

Knockout of porcine alpha1,3-galactosyltransferase gene results in elimination of Galalpha1,3Gal-terminated glycolipid compounds. GalT-KO genetic modification did not produce new compensatory glycolipid compounds reactive with human serum antibodies.

Characterization of natural human anti-non-gal antibodies and their effect on activation of porcine gal-deficient endothelial cells

BACKGROUND

The generation of Galalpha1-3Gal (Gal) transferase deficient pigs has increased the interest in non-Gal antigens potentially representing important targets for xenoreactive antibody binding leading to xenograft rejection. The present study addressed the levels and immunoglobulin isotypes of preformed human anti-non-Gal antibodies and their potential to activate porcine endothelial cells.

METHODS

Porcine endothelial cells lacking the Gal epitope (Gal-/-) were used to measure immunoglobulin (Ig) M and IgG subclass anti-non-Gal antibodies, using sera from 80 blood donors and pooled human AB serum. Antibodies specific for the non-Gal Hanganutziu-Deicher (HD) xenoantigen were measured by enzyme-linked immunosorbent assay. Activation of Gal-/- and Gal+/+ endothelial cells by human serum was measured, in the presence or absence of complement inhibitors, by E-selectin cell-surface expression using flow cytometry.

RESULTS

Anti-non-Gal antibody levels varied considerably among individual sera and comprised approximately 10% of total anti-porcine antibodies without sex or age differences. Among the IgG subclasses only IgG1 and IgG2 were detected. Human serum-induced E-selectin expression on Gal-/- cells was less than 20% compared with Gal+/+ cells, correlated with anti-HD IgM and IgG antibody levels (P=0.027 and 0.032, respectively), and was largely complement-independent in accordance with the lack of IgG3 anti-non-Gal antibodies. In contrast, E-selectin upregulation on Gal+/+ cells was reduced in complement blocking experiments.

CONCLUSION

Preformed anti-non-Gal antibodies, in particular anti-HD antibodies, were present in all human sera samples, activated porcine endothelial cells, and may therefore play a role in xenograft rejection using organs from GalT-/- pigs.

A study of the xenoantigenicity of neonatal porcine islet-like cell clusters (NPCC) and the efficiency of adenovirus-mediated DAF (CD55) expression

BACKGROUND

The pig pancreas is considered to be the most suitable source of islets for xenotransplantation in patients with type I diabetes. The objective of this study was to assess the antigenicity of neonatal porcine islet-like cell clusters (NPCC), including the Galalpha1-3Galbeta1-4GlcNAc-R (alpha-Gal) and Hanganutziu-Deicher (H-D) antigens, and the pathway involved in human complement activation. The efficiency of expression of human decay-accelerating factor (DAF: CD55) on NPCC by adenoviral transduction was also examined, and the functional capacity of DAF was also estimated.

METHODS

The deposition of human natural antibodies, immunoglobulin (Ig)G and IgM, and the expression of alpha-Gal and H-D antigens on NPCC were investigated by FACS analysis. The downregulation in the antigenicity to human natural antibodies, including the alpha-Gal and H-D antigens on NPCC by treatment with tunicamycin, PDMP and neuraminidase were also examined. In addition, complement-mediated islet lysis was examined using factor D-deficient and C1-deficient sera. An adenovirus encoding DAF under the control of the cytomegalovirus promoter, Ad.pCMV-DAF, was then constructed, and used for transducing NPCC. The amelioration of complement-dependent cytotoxicity of the NPCC by the transduced DAF was assessed as an in vitro hyperacute rejection model of a pig to human xenograft.

RESULTS

The NPCC clearly expressed the alpha-Gal epitope, and the human natural antibodies, IgG and IgM, and the anti-H-D antibody also reacted with the NPCC. Treatment of NPCC with tunicamycin led to a drastic reduction in the extent of deposition of IgG, indicating the importance of N-linked sugars on the islets, presumably related to alpha-Gal expression on N-linked sugars. Neuraminidase treatment indicated the presence of, not only the H-D antigen, but also other sialic acid antigens which reacted with the human natural antibody, especially IgG. The complement deposition of factor B on NPCC was clear, and the alternative pathway-mediated NPCC killing accounted for approximately 30% of that by the total complement pathway. On the other hand, approximately 90% of the NPCC could be transduced to express DAF by the adenovector, Ad.pCMV-DAF. The expressed DAF showed an approximately 50-62% suppression in complement-dependent NPCC lysis.

CONCLUSION

The origin of the antigenicity of NPCC is mainly N-linked sugars including alpha-Gal and sialic acid antigens, and NPCC expressed the transduced molecule in high efficiency by the adenovector.

Carbohydrates in xenotransplantation

The success of allotransplantation has led to an increasing shortage of human organs from deceased donors. This crisis could be resolved by the use of organs from an anatomically suitable animal, such as the pig. The pig and human have, however, been evolving differently for approximately 80 million years, and numerous immunological and physiological barriers have developed that need to be overcome. Differences in carbohydrate epitopes on pig and human cells have been found to play a major role in some of the immunological barriers that have been identified to date. The rejection caused by the presence of galactose-alpha1,3-galactose (Gal) on the pig vascular endothelium and of natural anti-Gal antibodies in humans has recently been prevented by the breeding of pigs that do not express Gal, achieved by knocking out the gene for the enzyme alpha1,3-galactosyltransferase, which was made possible by the introduction of nuclear transfer/embryo transfer techniques. N-glycolylneuraminic acid (the so-called Hanganutziu-Deicher antigen) has been identified as another carbohydrate antigen present in pigs that may need to be deleted if xenotransplantation is to be successful, although some doubt remains regarding its importance. There remain other antipig antibodies against hitherto unidentified antigenic targets that may well be involved in graft destruction; their possible carbohydrate target epitopes are discussed.

Expression of carbohydrate xenoantigens on porcine peripheral nerve

BACKGROUND

The use of thin easily revascularized cutaneous nerve autografts, which has been the gold standard, or the alternative use of nerve allografts or artificial grafts for nerve reconstructing have all their pros and cons. Nerve xenotransplantation may offer a potential alternative. In a potential pig to human nerve xenograft transplantation set-up several porcine antigen barriers have to be considered such as carbohydrate antigens system like the blood group A/O, the Galalpha1-3Gal (alphaGal) and the Hanganutziu-Deicher (HD) antigens. The swine leukocyte protein antigens system may also have to bee considered. The knowledge of the antigen expression on pig peripheral nerves is today limited. The present study describes the distribution of glycolipid based carbohydrate xenoantigens in ischiadicus nerve from blood group A and O pigs.

METHODS

Glycolipid fractions were separated on thin layer chromatography plates and immunostained with human AB sera, biotinylated Griffonia simplicifolia isolectin B4, monoclonal antibodies reacting with the HD antigen and with blood group A antigens based on different core saccharide structures. In addition, the subcellular distribution of alphaGal and HD antigens were studied by light- and electron-microscopical immunohistochemistry. The total amount of neutral glycolipids was 15 mg/g tissue for both blood group A and O nerves with mono-glycosylceramides as the dominating component.

RESULTS AND CONCLUSIONS

The total amount of acidic glycolipids (gangliosides and sulpholipids) was 9 mg/g tissue for both the blood group O and A nerves with sulphatides as the dominating components. Analyses of the glycolipid fractions showed strong expression of both the alphaGal and the HD antigens in nerves from both blood group A and O pigs. In addition, small amounts of blood group A antigens were expressed in nerves from blood group A pigs. Staining of neutral glycolipids from blood group A pigs using monoclonal antibodies reacting with A antigen having different core structures suggested that the A epitope expressed on pig ischiadicus nerves is based on the type 1 core chain structure. Light and electron microscopical studies on the alphaGal and HD-antigen distribution revealed that the neural cells were alphaGal antigen negative. Endothelial cells of blood vessels, and lymphatic and perineural cells expressed alphaGal antigen. Both endothelial cells and myelinized axons revealed positively labelled for the HD antigen.

A study of the xenoantigenicity of adult pig islets cells

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 Galalpha1-3Galbeta1-4GlcNAc-R (the alpha-Gal) and Hanganutziu-Deicher (H-D) antigens, and the pathway involved in human complement activation.

METHODS

The expression of alpha-Gal on islets from adult pigs was investigated by immunohistochemical staining and flowcytometric analysis. The alpha1,3 galactosyltransferase (alpha1,3GT) 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 alpha-Gal epitope, and alpha1,3GT 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 alpha-Gal, and pig islets can be injured by both the classical and the alternative complement pathway in human serum.

Are N-glycolylneuraminic acid (Hanganutziu-Deicher) antigens important in pig-to-human xenotransplantation?

BACKGROUND

N-glycolylneuraminic acid (NeuGc) epitopes, so called Hanganutziu-Deicher (HD) antigens, which are widely expressed on endothelial cells of all mammals except humans, are considered to be potential targets for natural and elicited anti-nonGalalpha1-3 Gal (Gal) antibodies in humans. We previously reported that anti-NeuGc antibodies were not detected in healthy humans by enzyme-linked immunosorbent assay (ELISA) using NeuGc-GM3-coated plates, and no antibody production was observed in patients with a history of exposure to pig cells. However, a recent paper has revealed that (i) anti-NeuGc antibodies to porcine red blood cells (PRBC) are detectable in most healthy humans, and (ii) the majority of anti-nonGal antibodies are specific for NeuGc epitopes. The purpose of this study was to reassess whether NeuGc is important as an immunogenic nonGal epitope.

METHODS

The binding of antibodies to PRBC and porcine aortic endothelial cells (PAEC) was compared. Cells were treated with (i) alpha-galactosidase, and then (ii) neuraminidase, which digests sialic acids, including NeuGc epitopes. Cells were incubated with human pooled sera, and applied to flow cytometric analysis. After enzyme digestion, almost complete reduction of Gal and NeuGc expression was confirmed by GS-IB4 and HU/Ch2-7 (a chicken monoclonal antibody against HD antigens), respectively. Trypsin, which removes membrane glycoproteins, and endoglycoceramidase which cleaves glycolipids, were used for differentiating between NeuGc-containing glycoproteins and glycolipids.

RESULTS

Neuraminidase-treatment reduced the binding of immunoglobulin G (IgG) antibodies to PRBC; about half of the anti-nonGal IgG antibodies to PRBC were directed to NeuGc. In contrast, anti-nonGal antibodies to PAEC were not directed to NeuGc. Trypsin-treatment markedly reduced the expression of NeuGc only on PRBC. Endoglycoceramidase-treatment was followed by a greater reduction in NeuGc epitopes on PAEC than on PRBC. Most NeuGc on PRBC appeared to be linked to proteins, but most NeuGc on PAEC was expressed on glycolipids.

CONCLUSIONS

Carbohydrate structures on PRBC are different from those on PAEC. Healthy human sera contain anti-nonGal IgG antibodies to NeuGc expressed on PRBC, but not on PAEC. We speculate that anti-nonGal IgG antibodies to PRBC can recognize both NeuGc and protein, and this may be the reason why such antibodies have not been detected by ELISA. A definite conclusion about the immunogenicity of NeuGc has not been obtained. More sera from patients (not from non-human primates) sensitized with porcine cells or organs need to be studied.

Release of pig leukocytes during pig kidney perfusion and characterization of pig lymphocyte carbohydrate xenoantigens

The Galalpha1-3Gal (alphaGal) antigen is considered the main xenoantigen in the pig to human species combination but other porcine antigens have to be considered such as the swine lymphocyte antigen (SLA), the blood group A/O and the Hanganutziu-Deicher (H-D) antigens. The H-D antigens are N-glycolyl-neuraminic acid (NeuGc) terminated gangliosides that are widely distributed in mammalian species but absent in humans. Upon exposure to a vascularized pig organ, the human recipient can be immunized by direct interaction with the pig tissue or/and by transfer of tissue/cells from the organ into the recipient. In the present work, we describe the release of cells from porcine kidneys upon perfusion and the expression of glycolipid based alphaGal, blood group A/O and H-D antigens in pig lymphocytes. Pig kidneys were flushed with 20 ml of NaCl or Lidocain containing 5000 U heparin, and thereafter perfused with 3000-ml perfusion solution and the cells released were counted and examined microscopically. Neutral glycolipid and ganglioside fractions were extracted from purified pig lymphocytes. The extracted components were characterized by thin layer chromatography, degradation and mass spectrometry. The expression of alphaGal and H-D epitopes on cells released from pig kidneys and purified pig lymphocytes were studied by immune electron microscopy. A total amount of about 300 x 106 leukocytes, mainly lymphocytes were released in the perfusate from the kidneys, of which about 100 x 106 cells were eluated in the 600 to 2400 ml perfusate fraction. Immunelectron microscopical analysis with Griffonia simplicifolia isolectin B4 showed staining of pig leukocytes and other cells, morphologically similar to endothelial cells, released in the perfusate. The purified porcine lymphocytes contained 930 microg neutral glycolipid (4.2 microg/mg cell protein) of which 95% was glycolipids with one to four sugar residues. Immunostaining of the neutral glycolipid fractions revealed alphaGal terminated compounds migrating in the five and 10 to 12 sugar regions and blood group A compounds in the six and eight sugar regions. Two major gangliosides NeuGc-GM3 and NeuGc-GD3 were found in the pig lymphocytes. In a patient extracorporeally xenoperfused with a pig kidney, an increased staining of both alphaGal terminated structures as well as the H-D reactive gangliosides were found in the post-perfusion serum samples. In summary, leukocytes, mainly lymphocytes are released from pig kidneys during perfusion which may contribute to immunization of human xenograft recipients.

Remodeling of the major pig xenoantigen by N-acetylglucosaminyltransferase III in transgenic pig

We have been successful in generating several lines of transgenic mice and pigs that contain the human beta-d-mannoside beta-1,4-N-acetylglucosaminyltransferase III (GnT-III) gene. The overexpression of the GnT-III gene in mice and pigs reduced their antigenicity to human natural antibodies, especially the Galalpha1-3Galbeta1-4GlcNAc-R, as evidenced by immunohistochemical analysis. Endothelial cell studies from the GnT-III transgenic pigs also revealed a significant down-regulation in antigenicity, including Hanganutziu-Deicher antigen, and dramatic reductions in both the complement- and natural killer cell-mediated pig cell lyses. Changes in the enzymatic activities of other glycosyltransferases, such as alpha1,3-galactosyltransferase, GnT-IV, and GnT-V, did not support cross-talk between GnT-III and these enzymes in the transgenic animals. In addition, we demonstrated the effect of GnT-III in down-regulating the xenoantigen of pig heart grafts, using a pig to cynomolgus monkey transplantation model, suggesting that this approach may be useful in clinical xenotransplantation in the future.

Lack of antibody production against Hanganutziu-Deicher (H-D) antigens with N-glycolylneuraminic acid in patients with porcine exposure history

The significance of non-alphagalactosyl antigens remains unclear in pig-to-primate xenotransplantation. Hanganutziu-Deicher (H-D) antigens with terminal N-glycolylneuraminic acid (NeuGc) are widely expressed on endothelial cells of mammalian species, with the exception of humans. As baboons and monkeys also express H-D antigens, a pig-to-non-human primate experimental model cannot resolve the question of whether H-D antigens can elicit a potent humoral response in human recipients. The purpose of this study was to elucidate the clinical significance of H-D antigens by examining the sera from patients who have been previously exposed to porcine tissue. After the digestion of porcine aortic endothelial cells (PAEC) by neuraminidase, NeuGc and N-acetylneuraminic acid (NeuAc) were quantitated by HPLC. IgG and IgM antibody levels against H-D antigens were measured by NeuGc-GM3-coated ELISA plates in the sera of patients who had undergone ex vivo kidney perfusion 1 to 3 weeks and 2 years previously (n=2) or had been injected with fetal porcine islets 2 months previously (n= 10). HPLC determined that 9.7x 10(7) NeuAc and 6.3x 10(7) NeuGc residues per cell were released from PAEC by neuraminidase, while 25.7x 10(7) NeuAc and an undetectable level of NeuGc were released from human aortic endothelial cells (HAEC). No significant elevation of IgG or IgM antibody levels against NeuGc-GM3 was observed in sera from patients with a history of porcine exposure. Considering the active production of antibody against the foreign galactosyl antigens after pig-to-human xenotransplantation, some production of antibodies against the equally foreign H-D antigens would be expected, because large amounts of NeuGc terminated saccharides are present in the pig endothelial cell surface. However, no production of antibodies directed to H-D antigens could be found in patients exposed to porcine tissue. Further studies are warranted to explain why H-D antigens do not elicit a significant antibody production.

Carbohydrates in transplantation

Carbohydrate materials have become increasingly utilized in transplantation and cell/tissue engineering within the past year. This has been well documented in recent applications of immobilized or soluble alpha-galactosyl epitopes (i.e. oligosaccharides with a terminal Galalpha1-3Gal sequence) in preventing hyperacute rejection in pig-to-primate xenotransplantation. In addition, alpha-galactosyl polymers have been shown to exhibit much greater activity (up to 10(4) times) than alpha-galactosyl monomers in inhibiting the binding of anti-galactosyl antibodies to pig kidney epithelial cells and assisting in the prevention of cytotoxicity in human serum.