Harout DerSimonian

Xenogeneic human anti‐pig cytotoxicity mediated by activated natural killer cells

Abstract: Even if hyperacute rejection, which is mediated by human natural antibodies (nAb) and complement, could be prevented, xenoreactive human anti‐pig cellular responses may lead to delayed and/or chronic xenograft rejection. Among the cell populations participating in such rejection, NK cells have been proposed as an important component. In this study we report the in vitro cytotoxic activity of natural killer (NK) cells obtained from healthy human donors against porcine target cells. Freshly isolated peripheral blood mononuclear cells (PBMC) and purified NK cells (CD16+/CD56+, CD3‐, CD20‐, CD33‐) exhibited little or no cytotoxic activity when tested on porcine phytohemagglutinin (PHA)‐stimulated lymphoblasts or bone marrow‐ or aortic‐derived endothelial cell lines in the presence of serum‐free medium. Killing was considerably higher in the presence of human decomplemented plasma, containing xenoreactive nAb, or purified Gal(α1,3)Gal‐reactive antibodies, suggesting that antibody dependent cell‐mediated cytotoxicity (ADCC) mediated by NK cells is an important mechanism involved in xenogeneic cytotoxicity. After incubation of human PBMC for 6 days in the presence of irradiated xenogeneic porcine or allogeneic stimulator cells, or in the presence of exogenous interleukin 2 (IL‐2), the cytotoxic activity of the bulk cultures as well as that of isolated NK cells (separated from stimulated bulk cultures) against xenogeneic targets increased considerably, and corresponded to an increased NK‐specific lysis of K562 target cells. Cell surface staining and flow cytometry showed that CD16+/CD56+, CD3‐ NK cells composed ca. 25% of short‐term (6 days) xenogeneic, allogeneic, or IL‐2 stimulated bulk cultures. In summary, these data suggest that, in contrast to allogeneic cell‐ mediated killing, xenogeneic human anti‐porcine cytotoxicity includes an important contribution from NK cells.

Heterogeneity of human anti-pig natural antibodies cross-reactive with the Gal(alpha1,3)Galactose epitope.

BACKGROUND The cell surface carbohydrate moiety, Gal(alpha1,3)Galactose (alphaGal), has been implicated as the major determinant recognized by more than 80% of human anti-porcine natural antibodies (NAb). An ELISA system was developed for the detection of this subpopulation of porcine cell-reactive NAb using synthetic alphaGal conjugated to bovine serum albumin. METHODS A screen of 95 human serum samples by this method demonstrated marked variability in the alphaGal reactivity of unrelated donors. The percentage of alphaGal-reactive NAb relative to total immunoglobulin was determined for 10 donors. RESULTS alphaGal-reactive NAb comprised 1.0-2.4% of total serum IgG, whereas the range was from 3.9% to 8.0% for IgM. CONCLUSIONS The higher level of alphaGal-reactive IgM suggests that xenoreactive NAbs may be the product of germ-line genes. Two-dimensional gel analysis of affinity-purified alphaGal-reactive NAb from two donors provided evidence suggesting that IgM from this subpopulation of NAb were restricted in protein charge heterogeneity.

Immortalized bone-marrow derived pig endothelial cells.

Abstract: Primary cultures of porcine endothelial cells (EC) can only be maintained for a limited number of passages. To facilitate studies of xenogeneic human anti‐pig immune responses in vitro, pig microvascular bone‐marrow (BM) and macrovascular aortic EC were obtained from our herd of partially inbred miniature swine, homozygous for the major histocompatibility locus, and immortalized with a modified SV40 large T vector. The resulting BM‐derived (2A2) and aortic (PEDSV.15) immortalized EC lines showed unlimited growth and EC phenotype as indicated by expression of von Willebrand Factor (vWF) and low density lipoprotein (LDL) receptors as well as by formation of typical cobblestone monolayers. Ultrastructural studies revealed morphological similarities in primary and immortalized EC. Flow cytometry analysis demonstrated constitutive SLA class I expression by all lines whereas SLA class II was only expressed after stimulation with porcine IFNγ. Furthermore, pig CD34 mRNA was detected by Northern blot analysis in primary and immortalized aortic EC but not in 2A2. Both EC lines expressed a number of myeloid markers, adhesion molecules and xeno‐antigens, the latter being determined by binding of human natural antibodies. Gene transfer into the porcine EC lines was successfully performed by electroporation or calcium‐phosphate transfection, as well as by adenoviral infection. Finally, the functional similarity between primary and immortalized EC was demonstrated in adhesion and cytotoxicity assays. Together, these results suggest that 2A2 and PEDSV.15 represent valuable tools to study both human cellular and humoral immune responses in vitro against pig EC derived from microvascular and large vessels.

Human anti‐pig T‐cell mediated cytotoxicity

Abstract: Miniature swine have a variety of advantages as potential donors for human xenotransplantation, including size, physiological similarities, and breeding characteristics. To investigate the nature of the human anti‐pig xenogeneic cellular response, we performed standard 51Cr‐release cell‐mediated lympholysis (CML) experiments. The major histocompatibility complex (MHC) allele specificity of the xenogeneic cellular response was tested on porcine target cells of three distinct homozygous MHC haplotypes (SLAaa, SLACC, SLAdd) and three intra‐MHC recombinant haplotypes (SLAjj, SLAgg, SLAkk), obtained from our herd of partially inbred miniature swine. After stimulation with irradiated porcine peripheral blood mononuclear cells (PBMC) of SLAaa haplotype, a strong nonspecific cytotoxic response of the bulk culture against xenogeneic targets of all three haplotypes was observed. However, SLA allele specificity could be demonstrated after T cell enrichment, and mapping experiments revealed predominantly SLA class I restriction of xenoreactive cytotoxic T lymphocytes (CTLs), although some class II restriction was also observed. The experiments were repeated in the presence of anti‐T cell monoclonal antibodies, anti‐CD3 (OKT3), anti‐CD2 (35.1), anti‐CD4 (OKT4), or anti‐CD 8 (OKT8). The bulk xenogeneic CML was not inhibited by any of the anti‐T cell antibodies tested. However, after T cell‐enrichment, lysis of porcine targets was significantly inhibited by anti‐CD3 or anti‐CD8 antibody and partially inhibited by anti‐CD2 antibody. In comparable assays, the human allogeneic CML was blocked by anti‐CD3 and anti‐CD8, but not by anti‐CD2 or anti‐CD4 antibodies. Finally, the cytotoxic activity of A3b3, a human CD4+ T‐cell clone, was tested. A3b3 lysed xenogeneic targets of SLAaa haplotype, but not SLACC or allogeneic targets, and was inhibited by anti‐CD4, anti‐CD2, and anti‐CD3 antibodies, but not by anti‐CD8. With the aid of intra‐MHC recombinant haplotypes, the xenogeneic CML reactivity of A3b3 was mapped to SLA class II, suggesting direct xenogeneic recognition of porcine MHC class II antigens by human T cells. Thus, the human anti‐pig cell‐mediated cytotoxic response is similar in magnitude to comparable allogeneic responses, and involves both SLA class I and class II restricted T‐cell mediated cytotoxicity, as well as additional nonspecific killing, possibly by NK cells.

Relationship between ABO blood group and levels of Gal alpha,3Galactose-reactive human immunoglobulin G.

BACKGROUND The terminal Gal alpha1,3Galactose (alphaGal) determinant is present on all porcine glycoproteins and glycolipids, but is not expressed by human cells. Consequently human sera contain anti-alphaGal natural antibodies. The human blood group B antigen [Gal alpha1,3(Fuc1,2)Galactose] is differentiated from the alphaGal epitope by the presence of a fucosyl group. METHODS To determine whether the expression of the B antigen has any effect on the level of alphaGal-reactive natural antibodies, equal numbers (n=12) of A, B, AB, and O serum samples were evaluated by ELISA and flow cytometry. RESULTS A significant reduction in IgG alphaGal reactivity was observed with serum samples from B antigen-expressing donors (B, AB) relative to non-B antigen-expressing donors (A, O). CONCLUSIONS These results are consistent with the possibility that anti-alphaGal antibodies in non-B antigen-expressing individuals include a subset that is reactive with the structurally related B antigen and that this subset is absent in B and AB individuals.

Modulation of the in vivo primate anti-Gal response through administration of anti-idiotypic antibodies.

Polyclonal anti‐idiotypic antibodies (AIA) were generated against human Galα1,3Gal antibodies (anti‐Gal) isolated from a single donor. Specificity of the AIA was demonstrated by selective binding to anti‐Gal antibodies (Ab) and absence of reactivity to non‐Gal Ab. The idiotopes identified by AIA were present on anti‐Gal Ab from all of the human samples evaluated (n=59) as well as on pooled samples, demonstrating that a restricted number of dominant idiotopes characterized the human anti‐Gal Ab response. Furthermore, the AIA had cross‐species reactivity with baboon serum samples (n=19), suggesting that the overall shape of the anti‐Gal Ab combining site is conserved throughout the Old World primates and providing additional evidence of the limited heterogeneity of the anti‐Gal Ab repertoire. In order to evaluate the potential effect of AIA in the modulation of the anti‐Gal response in vivo, a baboon was injected with repeated doses of the purified AIA. Following AIA treatment, new Ab were generated that reduced Ab‐mediated cytotoxicity to porcine cells. Furthermore, administration of the AIA to a baboon prolonged the survival of intravenously infused pig hematopoietic cells when compared with their survival in a control baboon that did not receive prior AIA treatment but underwent a similar conditioning regimen.