Modifying xenogeneic immune recognition and engraftment by genetic engineering

Sammanfattning: Transplantation using xenogenic organs, tissues and cells (i.e. xenotransplantation) is a potential solution to the shortage of those from human sources. Vascular endothelial cells (ECs) are the most immediate barrier between the xenogeneic donor organ and the host defense systems. In order to accomplish gene expression in ECs specifically, EC-specific promoters are preferable to be used. If human EC-specific promoters can be used in porcine ECs, time and efforts will be saved. In acute vascular rejection (AVR), the interaction between porcine endothelium and human NK cells/monocytes has been suggested to depend on the direct recognition of Galalpha1,3Gal (alpha-Gal) epitopes on porcine ECs. Genetic engineering of pancreatic islets prior to transplantation has the potential to improve islets survival through the expression of genes encoding factors supporting islet acceptance by the host. The lentiviral vector system, e.g. an HIV-1 based vector system, has been shown to efficiently and stably transduce primary and post-mitotic cells. The aims of this thesis were: (i) to investigate the activity and cell-specificity of the human EC specific promoters of Flk-1, Flt-1, ICAM-2, thrombomodulin and vWf in porcine cells; (ii) to evaluate the importance of lentiviral-mediated expression of alpha-Gal on ECs for its interaction with human NK cells and monocytes, and to evaluate the transduction efficiency in primary ECs; (iii) to investigate the ability of the lentiviral vector to transduce isolated rat pancreatic islets and its effect on islet function. The promoters for human Flk-1, Flt-1, ICAM-2, thrombomodulin, and vWf supported similar levels of luciferase expression in human (HAECs) and porcine (HAECs) aortic ECs, with the Flk-1 promoter being the strongest. The human EC-specific promoters all showed less activity in porcine kidney microvascular ECs than in liver or brain microvascular ECs. The thrombomodulin and Flk-1 promoters exhibited similar activity in porcine liver and kidney microvascular ECs, whereas the Flk-1 promoter was stronger in aortic and brain microvascular ECs. No difference was detected between non-alpha-Gal and alpha-Gal expressing HAECs in terms of their susceptibility to NK cell-mediated lysis, ability to stimulate IFN-gamma production by NK cells, or ability to support NK cell-adhesion under static and dynamic conditions. In addition, the alpha-Gal epitope did not appear to contribute to increased monocyte adhesion to, or migration across, endothelium. Human monocytes adhered less to PAECs than to HAECs under flow following human, but not porcine, TNF-alpha stimulation. Lentiviral transduction did not result in activation of HAECs, and transduced HAECs responded as expected to TNF-alpha and IFN-gamma stimulation. Lentivirus transduction did not affect rat pancreatic islet s viability and insulin secretion in vitro and its ability to restore euglycemia in diabetic nude mouse in vivo. Furthermore, this vector sustained long-term expression of the transgene in islets. Amongst the EC-specific promoters tested, the human Flk-1 and thrombomodulin promoters are good candidate promoters for strong EC-specific gene expression in genetically modified pigs. Our work on the interaction between alpha-Gal epitopes and NK cells/monocytes suggest that efforts on preventing infiltration of these leukocytes in organ xenograft need to be focused on adhesion receptor-ligand interactions rather than on carbohydrate remodelling of donor pigs. In addition, the lentivirus vector can be used as a gene carrier to modify primary ECs as well as cells in pancreatic islets in order to improve engraftment and prevent rejection.