In vitro models of xenograft rejection : studies on leukocyte-endothelial cell interactions

Abstract

Xenotransplantation is the transplantation of cells, tissues or organs between different species and is a potential alternative to allotransplantation because of the lack of human donors. The pig is believed to be a suitable donor-species to man for medical, technical, financial and ethical reasons. Porcine vascularized organs transplanted into humans are hyperacutely rejected (HAR) due to xenoreactive natural anti-pig Abs present in man. These Abs arc mainly directed against the Gal(alpha)1,3Gal (alpha-Gal) epitope expressed on most porcine cells, including the endothelium. Following their binding to the endothelium, the complement cascade is activated leading to graft destruction. If HAR is prevented, the vascularized grafts are generally lost in an acute vascular rejection (AVR). AVR is characterized by type 11 (i.e. transcriptional) endothelial cell (EC) activation, thrombosis with fibrin deposition and cellular infiltration dominated by monocytes/macrophages, natural killer (NK) cells and neutrophils. In contrast, cellular xenografts, e.g. isolated islets of Langerhans, are subjected to an instant blood-mediated inflammatory reaction (IBMIR) upon intra-portal transplantation and a subsequent cell-mediated rejection dependent on CD4+ T cells and macrophages.

The aims of the work outlined in this thesis were: (i) to investigate the interaction between human leukocyte subpopulations and porcine endothelium in vitro; and to evaluate the effect of the alpha-Gal epitope and human anti-alpha-Gal Abs in this interaction; (ii) to study the cytokine/chemokine profile of isolated adult porcine islets (APIs) and to examine if these factors could induce activation of human ECs and/or attract human leukocytes in vitro.

In the absence of xenoreactive natural anti-pig Abs and complement, human neutrophils and monocytes did not adhere more to porcine than to human non- and TNF-alpha-stimulated endothelium, as assessed under static and dynamic conditions. The adhesion of human neutrophils increased significantly after antialpha-Gal Ab-mediated activation of porcine endothelium. Porcine EC activation was demonstrated by an increase in E-selectin and VCAM-1 expression, and by the production of soluble factors such as IL-8 and PAF. These factors induced migration of human neutrophils. Aberrant expression of the alpha-Gal epitope on human endothelium using recombinant lentiviruses did not confer susceptibility to human NK cell-mediated cytotoxicity or increased NK cell adhesion. NK cells did not produce IFN-gamma upon interaction with the alpha-Gal-expressing human endothelium. In addition, aberrantly expressed alphaGal epitopes did not increase the adhesion, under static or dynamic conditions, of human monocytes to human endothelium. Neither did it increase monocyte transendothelial migration. Recombinant lentiviruses proved to be effective vehicles for gene transfer into primary human ECs. Cultured APIs expressed mRNAs for MCP-1, IL-1beta and TNF-alpha. Supernatants from cultured APIs induced migration of human monocytes, which could be blocked by an anti-MCP-1 Ab. Isolated islets and islets in situ produced MCP-1 as detected by immunohistochemistry. Recombinant porcine IL- 1beta and TNF-alpha induced the expression of E-selectin and VCAM-1 on human endothelium.

With regard to novel therapeutic strategies to disrupt the infiltration of human leukocytes into organ xenografts, our data suggest that efforts should be focused on adhesion receptor-ligand interactions rather than on carbohydrate remodeling of donor pigs. Induced Abs of alpha-Gal or non-alpha-Gal specificity may increase the adhesion and infiltration of, not only human neutrophils, but also of other human leukocyte subpopulations. Indeed, preventing the induced anti-porcine Ab response will be an important factor to prevent AVR. The work on isolated porcine islets highlighted the importance of proinflammatory and chemotactic factors produced by the graft itself, and that these factors may have profound effects on the rejection process in vivo.