Expression of a negative dominant mutant of human p55 tumor necrosis factor-receptor inhibits TNF and monocyte-induced activation in porcine aortic endothelial cells

Use of porcine tumor necrosis factor receptor 1-Ig fusion protein to prolong xenograft survival

BACKGROUND

Delayed rejection of xenografts is a major hurdle that needs to be addressed to achieve long-term engraftment in the pig-to-primate transplant setting. Both vascular and avascular xenografts are susceptible to a delayed rejection process that comprises humoral and cellular responses. Tumor necrosis factor (TNF) is believed to play a role in this process by promoting cell activation, apoptosis and the recruitment of inflammatory cells. To address this problem, we engineered the donor cell in such a way that it could block both human and porcine TNF.

METHODS

We produced a recombinant fusion protein containing the extracellular domain of the porcine TNF-Receptor 1 and an IgG Fc moiety (pTNFR1Ig). We first evaluated by flow cytometry the pTNFR1Ig capacity to prevent TNF alpha-induced expression of SLAI, SLAII, VCAM-1, ICAM-1 and E-selectin on the cell surface of porcine aortic endothelial cells (PAEC). The effect on TNF alpha-mediated cell death was also assessed by propidium iodide staining after incubating PAEC with TNF alpha plus cycloheximide for 24 h. PAEC and porcine fibroblasts were subsequently engineered by retroviral infection to express and secrete pTNFR1Ig and their resistance to the TNF alpha effects was tested in vitro. Finally, we transplanted mock-control and pTNFR1Ig-expressing PAEC under the kidney capsule of BALB/c mice in the absence of immunosuppression and examined the degree of rejection at 2 and 3 weeks post-transplantation.

RESULTS

Treatment with pTNFR1Ig resulted in a very potent blockade of human, porcine and murine TNF alpha activity on porcine cells. It inhibited the upregulation of all cell surface markers of activation tested as well as the TNF alpha-mediated cell death. Moreover, pTNFR1Ig-expressing PAEC showed prolonged engraftment in a pig-to-mouse xenotransplant model.

CONCLUSIONS

Incorporation of strategies that block TNF may prove useful in the development of xenografts resistant to delayed rejection.

Xenogeneic transplantation

This review summarizes the clinical history and rationale for xenotransplantation; recent progress in understanding the physiologic, immunologic, and infectious obstacles to the procedure's success; and some of the strategies being pursued to overcome these obstacles. The problems of xenotransplantation are complex, and a combination of approaches is required. The earliest and most striking immunologic obstacle, that of hyperacute rejection, appears to be the closest to being solved. This phenomenon depends on the binding of natural antibody to the vascular endothelium, fixation of complement by that antibody, and finally, activation of the endothelium and initiation of coagulation. Therefore, these three pathways have been targeted as sites for intervention in the process. The mechanisms responsible for the next immunologic barrier, that of delayed xenograft/acute vascular rejection, remain to be fully elucidated. They probably also involve multiple pathways, including antibody and/or immune cell binding and endothelial cell activation. The final immunologic barrier, that of the cellular immune response, involves mechanisms that are similar to those involved in allograft rejection. However, the strength of the cellular immune response to xenografts is so great that it is unlikely to be controlled by the types of nonspecific immunosuppression used routinely to prevent allograft rejection. For this reason, it may be essential to induce specific immunologic unresponsiveness to at least some of the most antigenic xenogeneic molecules.

Factors in xenograft rejection

Important mechanisms underlying immediate xenograft loss by hyperacute rejection (HAR), in the pig-to-primate combination, have been recently delineated. There are now several proposed therapies that deal with the problem of complement activation and xenoreactive natural antibody (XNA) binding to the vasculature that have been shown to prevent HAR. However, vascularized xenografts are still lost, typically within days, by delayed xenograft rejection (DXR), alternatively known as acute vascular rejection (AVR). This process is characterized by endothelial cell (EC) perturbation, localization of XNA within the graft vasculature, host NK cell and monocyte activation with platelet sequestration and vascular thrombosis. Alternative immunosuppressive strategies, additive anti-complement therapies with the control of any resulting EC activation processes and induction of protective responses have been proposed to ameliorate this pathological process. In addition, several potentially important molecular incompatibilities between activated human coagulation factors and the natural anticoagulants expressed on porcine EC have been noted. Such incompatibilities may be analogous to cross-species alterations in the function of complement regulatory proteins important in HAR. Disordered thromboregulation is potentially relevant to the progression of inflammatory events in DXR and the disseminated intravascular coagulation seen in primate recipients of porcine renal xenografts. We have recently demonstrated the inability of porcine tissue factor pathway inhibitor (TFPI) to adequately neutralize human factor Xa (FXa), the aberrant activation of both human prothrombin and FXa by porcine EC and the failure of the porcine natural anticoagulant, thrombomodulin to bind human thrombin and hence activate human protein C. The enhanced potential of porcine von Willebrand factor to associate with human platelet GPIb has been demonstrated to be dependent upon the isolated A1 domain of von Willebrand factor. In addition, the loss of TFPI and vascular ATPDase/CD39 activity following EC activation responses would potentiate any procoagulant changes within the xenograft. These developments could exacerbate vascular damage from whatever cause and enhance the activation of platelets and coagulation pathways within xenografts resulting in graft infarction and loss. Analysis of these and the other putative factors underlying DXR should lead to the development and testing of genetic approaches that, in conjunction with selected pharmacological means, may further prolong xenograft survival to a clinically relevant extent.

Membrane-associated lymphotoxin on natural killer cells activates endothelial cells via an NF-kappaB-dependent pathway

BACKGROUND

Inhibition of complement in small animal models of xenotransplantation has demonstrated graft infiltration with natural killer (NK) cells and monocytes associated with endothelial cell (EC) activation. We have previously demonstrated that human NK cells activate porcine EC in vitro, which results in adhesion molecule expression and cytokine secretion. In this study, we used the NK cell line NK92 to define the molecular and cellular basis of NK cell-mediated EC activation.

METHODS

EC were transfected with either reporter constructs containing the luciferase gene driven either by E-selectin or interleukin (IL)-8 promoters or a synthetic NF-kappaB-dependent promoter. In addition, a dominant-negative mutant tumor necrosis factor receptor I (TNFRI) expression vector was co-transfected in inhibition studies. Forty-eight hours after transfection, EC were stimulated with NK cells or NK cell membrane extracts for 7 hr and activation was measured by a luciferase assay.

RESULTS

Co-culture of NK cells with transfected EC enhanced E-selectin, IL-8, and NF-kappaB-dependent promoter activity. NK cell membrane extracts retained the capacity to activate EC and induced nuclear translocation of NF-kappaB (p50 and p65). Western blotting of NK cell and membrane extracts detected the presence of Lymphotoxin-alpha (LTalpha) but not tumor necrosis factor-alpha. Furthermore, LTalpha was secreted in NK:EC co-cultures. Co-transfection with dominant-negative mutant TNFRI inhibited EC activation by NK cell membrane extracts and by NK cells by 80% and 47%, respectively. The same pattern of inhibition was observed using anti-human LT sera.

CONCLUSIONS

Human NK cell membrane-bound LT signals across species via TNFRI, leading to NF-kappaB nuclear translocation and transcription of E-selectin and IL-8, which results in EC activation. The discrepancy in the degree of inhibition by membrane extracts and NK cells with mutant TNFRI suggests that additional pathways are utilized by NK cells to activate EC.