Research Article: Co-Transplantation of Endothelial Progenitor Cells and Pancreatic Islets to Induce Long-Lasting Normoglycemia in Streptozotocin-Treated Diabetic Rats

Date Published: April 14, 2014

Publisher: Public Library of Science

Author(s): Paola Quaranta, Sara Antonini, Saturnino Spiga, Benedetta Mazzanti, Michele Curcio, Giovanna Mulas, Marco Diana, Pasquina Marzola, Franco Mosca, Biancamaria Longoni, Lucienne Chatenoud.


Graft vascularization is a crucial step to obtain stable normoglycemia in pancreatic islet transplantation. Endothelial progenitor cells (EPCs) contribute to neoangiogenesis and to the revascularization process during ischaemic events and play a key role in the response to pancreatic islet injury. In this work we co-transplanted EPCs and islets in the portal vein of chemically-induced diabetic rats to restore islet vascularization and to improve graft survival. Syngenic islets were transplanted, either alone or with EPCs derived from green fluorescent protein (GFP) transgenic rats, into the portal vein of streptozotocin-induced diabetic rats. Blood glucose levels were monitored and intraperitoneal glucose tolerance tests were performed. Real time-PCR was carried out to evaluate the gene expression of angiogenic factors. Diabetic-induced rats showed long-lasting (6 months) normoglycemia upon co-transplantation of syngenic islets and EPCs. After 3–5 days from transplantation, hyperglycaemic levels dropped to normal values and lasted unmodified as long as they were checked. Further, glucose tolerance tests revealed the animals’ ability to produce insulin on-demand as indexed by a prompt response in blood glucose clearance. Graft neovascularization was evaluated by immunohistochemistry: for the first time the measure of endothelial thickness revealed a donor-EPC-related neovascularization supporting viable islets up to six months after transplant. Our results highlight the importance of a newly formed viable vascular network together with pancreatic islets to provide de novo adequate supply in order to obtain enduring normoglycemia and prevent diabetes-related long-term health hazards.

Partial Text

Pancreatic islet transplantation is a widely accepted therapy for the cure of insulin-dependent diabetes mellitus (IDDM) [1], [2]. Compared to solid organ transplantation, it offers some advantages: low invasive surgery and low incidence of peri-operative risks. Pancreatic islets have a peculiar micro vascular system, known as the insulo-acinar portal system [3]–[5], which is largely destroyed during the isolation procedure, thus requiring rapid revascularization to preserve its performance in the transplant. In the whole pancreas transplantation the anastomosis of blood vessels can lead to rapid revascularization [6], with vessel density and oxygen tension being one and half times that of pancreatic islets [7]–[12], thereby suggesting that reduced oxygen supply may lead to impaired islet function [11]–[13]. After transplantation islets receive nutrients and oxygen only by diffusion mechanisms whatever the implantation site and after one month islets are still not yet fully revascularized [7]. Carlsson et al. compared blood perfusion and oxygenation of transplanted islets in three different sites (kidney capsule, liver and spleen) showing that though the three implantation organs differed markedly in their blood perfusion, the islet graft blood perfusion and oxygen pressure was similar irrespective of the implantation site. This suggests that the intrinsic properties of the transplanted islets are more important than the choice of the implantation site [7].

Despite intense research carried out on EPC biology in the last 10 years a consensus on the definitive appearance and function of EPCs have not been yet reached [37]. Very little data are available on rat EPC characterization and at present a clear definition of EPC surface markers still remains elusive. Therefore there are controversial results obtained in EPC therapy derived from different EPC populations [39], [40]. In vitro, two different populations of EPCs can be derived: early and late EPCs. Even if these cells are characterized by different morphology, proliferation potential and phenotype, in vivo these seem to improve vascularization [23]. In this work we studied the effect of bone marrow-derived rat late-EPCs in a marginal mass model of pancreatic islet transplantation in chemically-induced diabetic rats (STZ-treated), in an attempt to re-establish islet microvasculature destroyed during the islet isolation procedure. In a previous work we showed that transplantation of 700 IE in diabetic rat induced a significant decrease in blood glucose but failed to attain normoglycemia therefore this is considered a marginal mass model of islet transplantation [30]. We decided to transplant only a marginal mass of islets to overcome one of the main problems in islet transplantation: the limited donor-islet availability [41], [42]. On the basis of our results, we suppose that in the same model of full grafts (i.e. a transplantation of 1400 IE, [30]) the EPCs would be able to support islet revascularization better than in the marginal mass model. It is well known in fact that long-lasting hyperglicemia impairs the vascular network [14] preventing the revascularization process. In the full graft model normoglycemia is reached faster and for a long period than in the marginal mass, thus the EPC efficacy in the islet revascularization could be increased. Our results highlight that neo-vascularization is a crucial process in promoting a viable and enduring islet pancreatic transplant in experimental diabetes. GFP+ EPCs that lasted for 180 days after co-transplant promoted a newly formed vascular network. Accordingly, histological analysis revealed insulin-positive clusters of islets surrounded by GFP-expressing newly formed endothelial cells only in livers of recipients that received a co-transplant. This was combined with a long-lasting normoglycemia and a modulation of the expression of angiogenesis-related genes, which accompanied a new blood vessel formation. Since we did not detect any residual beta cell mass in pancreatic tissues of STZ-treated animals by H&E analysis, we speculated that the role of EPCs was to sustain the function of donor grafted islets more than residual islets of the recipients. Our data is in accordance with recently published results by Oh et al. They showed that not only donor EPC co-transplanted with islets are able to improve intra-islet microvasculature but also contribute to maintain islet organization and morphology. Unlike us they used early EPCs co-transplanted with a marginal mass of syngenic islets under kidney capsule of diabetic-induced mice [43]. Impaired revascularization is one of the main issues of graft failure [44] and attempts have been made to overcome this obstacle through administration of angiogenic factors such as VEGF [45] and/or mesenchymal stem cells [46], or stimulation with GM-CSF to mobilize bone marrow-derived EPCs [47]. Nevertheless angiogeneic factors had a short half-life as consequence their use is not free of safety concerns [48]. On the other side the migration of mesenchymal stem cells towards the site of inflammation and their dispersion in several organs of the recipient [49]–[53] reduced their efficacy on the grafted islets and could increase the risk of tumour development [54], even though recently a pooled analysis showed no correlation between MSCs and malignancies [55]. GM-CSF treatment could be appealing for its immediate clinical translational potential, on the other hand this approach might have significant effects on the immune system of the recipient [47]. In this context the interest towards the effect of EPCs in islet transplantation arised; in 2004, Brissova et al. reported that in vitro EPCs co-cultured with pancreatic islets were able to improve beta cell survival and insulin secretion [56] and recently many works on EPC and islet co-transplantation have been published [27], [43], [57]. The EPC-cotransplant method described here could be a more physiological way of inducing neovascularization in islet transplant. EPCs in fact did not disperse into the recipient but were confined in the implantation site around the transplanted islets thus making them safe for use in clinical settings. Overall EPCs were found to support beta cell proliferation [56], [58], cause a threefold improvement of beta cell volume and double functional blood vessels [57]. In addition, it has to be considered that patients with diabetes are characterized by low levels of circulating EPCs correlated to the impaired endothelium, so that the revascularization process is delayed upon an ischaemic insult (one week later compared to the healthy patient) [16], [59]. A re-establishment of a complete vascular network was successfully observed in the present study. Most importantly, the overlapping of blood glucose levels, blood glucose control and revascularization lasting 180 days suggests that donor EPCs may play a key role in enhancing and maintaining revascularization over long periods of time. Many works have also investigated the structure and function of intra-islet endothelial cells to clarify their role in blood vessel regeneration and in revascularization of islet graft [56], [60], [61]. Nyqvist et al. observed that the transplantation of freshly isolated islets with a relevant number of endothelial cells, in contrast to cultured islets, markedly improved their vascularization, thus a preservation of intra-islet endothelial cell mass was able to improve the long-term graft function [62]. Later the same authors further observed that donor islet endothelial cells contributed to the revascularization of freshly isolated islets by participating in early processes of vessel formation; nevertheless, these cells did not increase the vascular density or improve the endocrine function of the grafts [63]. In our work the absence of residual islets of the recipient in co-transplanted animals after STZ treatment supports the hypothesis that normoglycemia is due to donor islets and EPCs. Donor EPCs could effectively contribute to the intra-islet EPCs to support islet function and maintain morphological organization. Furthermore, the induced neovascularization in the co-transplanted group by EPCs is explained by the modulation of specific gene expression involved in the angiogenic process, whereas this is not observed in the control group. VEGF is the most important gene involved in the regulation of blood vessel sprouting during development, growth and disease; in particular, VEGF-A member is positively regulated by hypoxia [64]. Our data show a marked increase in the VEGF-A level in liver tissues of animals in the first 15 days after transplantation for groups receiving both 700 syn-IE and 700 syn-IE plus EPCs. It has to be considered that islets secrete VEGF-A after isolation, as the result of the ischemic insult derived from it and from the culture condition and also that in the first days after transplantation there is an incomplete vascular network formation and only a partial
recovery in functionality. Nevertheless in the co-transplanted group the increase in VEGF-A expression was lower than in the group of animals receiving islet alone, due to an exogenous administration of EPCs. Our findings are in line with data demonstrating that the vascularization of transplanted islets is delayed by the presence of hyperglycemia, derived from an increase in local oxygen consumption [65]. ANG-1 is responsible for vessel stabilization and promotion of pericytes adhesion by tightening endothelial junctions [66]. Jeansson et al. demonstrated that ANG-1 is not only necessary in the quiescent mature vasculature, but it also exerts a role in the regulation of the response to tissue injury and microvascular disease in diabetes [67]. High levels of ANG-1 gene expression, observed in the 700 syn-IE group, suggest that islets were unable to tighten endothelial junctions and maintain blood vessels in host diabetic environment. On the contrary, the down-regulation of ANG-1 observed in the co-transplanted group, is probably related to the EPC-supported vascularization. At variance with the other genes, PECAM-1 is involved in transendothelial migration of neutrophils, monocytes and natural killer cells both in vitro and in vivo. Indeed transmigration and inflammation can be significantly reduced when antibodies directed against PECAM-1 are used [68]. The strong down-regulation of PECAM-1 in co-transplanted group indicates that administration of exogenous EPCs is effective also in reducing the recruitment of immune system cells thus hampering an inflammatory response as shown by the absence of infiltrating mononuclear cells into the liver parenchyma. This result was supported also by Cantaluppi et al. who demonstrated that microvesicles, released by EPCs on human pancreatic islets, significantly inhibited spontaneous and cytokine-induced peripheral blood mononuclear cell adhesion to islet endothelial cells [69]. Microvescicles expressed CD154 marker as able to bind CD40 expressed by islet endothelial cells, thus interfering with leukocyte adhesion to endothelium. Overall our obtained results by gene expression suggest that day 15 is the crucial time point in graft revascularization. Recently Kang et al. reported that human cord blood-derived EPCs co-transplanted with porcine islets into renal capsules of diabetic nude mice were able to induce a rapid revascularization, a better graft perfusion and a recovery from hypoxia [27]. Although this experimental design was similar in some respects to our transplantation model, there were some differences in the implantation site, the origin of both islets and EPCs, timing of analysis and the islet dose used. To investigate the effect of EPCs on islet vascularization either early or a long time after transplantation, we chose a syngenic transplantation model to avoid any other interferences due to allo- or xenograft rejection. Moreover, the kidney subcapsular site is most commonly used in rodent models and has good results in that diabetes is reversed within a few days. Nevertheless, though the development of an instant blood-mediated inflammatory reaction (IBMIR) upon intraportal islet infusion, the progressive loss of islet function even in recipients of autologous grafts (in humans, but also in the canine model [70]), the portal vein implantation site still remains the gold standard for islet transplantation in the clinical setting. Until now the intraportal site was considered to have a reduced islet survival, not suitable for long-term function [70], while the long-term normoglycemia that we observed highlights a fully revascularization and a complete functionality of engrafted islets even when a marginal mass was used. A limit of our study is the assessment of pancreatic islet transplant only in a syngenic rat model to avoid an immunologic response due to donor-recipient immunologic systems. This needs certainly to be further investigated in the next future in an allogeneic co-transplant model, also by exploiting other cell lines like Sertoli cells which appeared to maintain the immunosuppressive effect during vascularization process [71] or either protocols of immunosuppression [72]. This could overcome obstacles deriving from the host’s immune rejection and impairment of vascularization network in the long term. An interesting issue on the mechanism of islet revascularization involves the emerging measurements of vessel parameters within the endothelial lumen. As reported by Alberts et al., the endothelial cells not only repair and renew the lining of established blood vessels, but also create new blood vessels and have a remarkable capacity to adjust their number and arrangement to suit local requirements [73]. By signaling to the surrounding cells, endothelial cells enable the blood vessel to adapt its diameter and wall thickness to suit the blood flow needed. They can be roused to proliferate with a doubling time of just a few days [73], including variations of thickness as it occurs in the allogeneic portal vein transplants, that developed a significant increase in wall thickness [74]. Recently several studies of co-transplant of pancreatic islets and EPCs reported conflicting results about the effect of EPCs on vascular density and their role in the mechanism of improvement in vascularization process [27], [57]. Our results demonstrated that EPCs were able to increase the vascular density and the endothelial thickness in co-transplant model during the first 30 days after transplantation. Subsequently a decrease of both parameters was observed up to levels comparable with healthy control values. This effect can be attributed to the endothelium transition from a juvenile to a mature status. It is worthy of note that few data is present in the literature about endothelial thickness which for the first time was analyzed in a study about the revascularization process. In conclusion, we provide evidence that co-transplantation of EPCs and a marginal mass of pancreatic islets in portal vein induce a stable rescue of glycemic control lasting for a significant fraction of the animal life span. We show that the glycemic control recovery is associated with EPC-induced neovascularization, which is followed by a stabilization of islet vascular network within a few weeks after transplantation. This is paralleled by a down-regulation of specific genes, such as ANG-1, involved in the vascularization process, and PECAM-1 related to the inflammation process. This provides the experimental evidence for the previously hypothesized revascularization process as a key factor for successful islet transplantation. The present results pave the way to translational experimental testing in humans as a new therapeutic approach to overcome some problems encountered in the search for a successful and long-lasting surgical approach for the cure of IDDM.