Release of pig leukocytes and reduced human NK cell recruitment during ex vivo perfusion of HLA-E/human CD46 double-transgenic pig limbs with human blood
Gisella Puga Yung,Anjan K Bongoni,Amandine Pradier,Natacha Madelon,Maria Papaserafeim,Riccardo Sfriso,David L Ayares,Eckhard Wolf,Nikolai Klymiuk,Andrea Bähr,Mihai A Constantinescu,Esther Voegelin,David Kiermeir,Hansjörg Jenni,Robert Rieben,Jörg D Seebach
DOI: https://doi.org/10.1111/xen.12357
Xenotransplantation
Abstract:Background: In pig-to-human xenotransplantation, interactions between human natural killer (NK) cells and porcine endothelial cells (pEC) are characterized by recruitment and cytotoxicity. Protection from xenogeneic NK cytotoxicity can be achieved in vitro by the expression of the non-classical human leukocyte antigen-E (HLA-E) on pEC. Thus, the aim of this study was to analyze NK cell responses to vascularized xenografts using an ex vivo perfusion system of pig limbs with human blood. Methods: Six pig forelimbs per group, respectively, stemming from either wild-type (wt) or HLA-E/hCD46 double-transgenic (tg) animals, were perfused ex vivo with heparinized human blood for 12 hours. Blood samples were collected at defined time intervals, cell numbers counted, and peripheral blood mononuclear cells analyzed for phenotype by flow cytometry. Muscle biopsies were analyzed for NK cell infiltration. In vitro NK cytotoxicity assays were performed using pEC derived from wt and tg animals as target cells. Results: Ex vivo, a strong reduction in circulating human CD45 leukocytes was observed after 60 minutes of xenoperfusion in both wt and tg limb groups. NK cell numbers dropped significantly. Within the first 10 minutes, the decrease in NK cells was more significant in the wt limb perfusions as compared to tg limbs. Immunohistology of biopsies taken after 12 hours showed less NK cell tissue infiltration in the tg limbs. In vitro, NK cytotoxicity against hCD46 single tg pEC and wt pEC was similar, while lysis of double tg HLA-E/hCD46 pEC was significantly reduced. Finally, circulating cells of pig origin were observed during the ex vivo xenoperfusions. These cells expressed phenotypes mainly of monocytes, B and T lymphocytes, NK cells, as well as some activated endothelial cells. Conclusions: Ex vivo perfusion of pig forelimbs using whole human blood represents a powerful tool to study humoral and early cell-mediated rejection mechanisms of vascularized pig-to-human xenotransplantation, although there are several limitations of the model. Here, we show that (i) transgenic expression of HLA-E/hCD46 in pig limbs provides partial protection from human NK cell-mediated xeno responses and (ii) the emergence of a pig cell population during xenoperfusions with implications for the immunogenicity of xenografts.