Cell Biochemistry and Biophysics

, Volume 48, Issue 2, pp 191–199

Improving islet transplantation by gene delivery of hepatocyte growth factor (HGF) and its downstream target, protein kinase B (PKB)/Akt


  • Nathalie Fiaschi-Taesch
    • Division of EndocrinologyUniversity of Pittsburgh
  • Andrew F. Stewart
    • Division of EndocrinologyUniversity of Pittsburgh
    • Division of EndocrinologyUniversity of Pittsburgh
Original Paper

DOI: 10.1007/s12013-007-0024-7

Cite this article as:
Fiaschi-Taesch, N., Stewart, A.F. & Garcia-Ocaña, A. Cell Biochem Biophys (2007) 48: 191. doi:10.1007/s12013-007-0024-7


Clinical studies have demonstrated that islet transplantation may be a useful procedure to replace beta cell function in patients with Type 1 diabetes. Islet transplantation faces many challenges, including complications associated with the procedure itself, the toxicity of immunosuppression regimens, and to the loss of islet function and insulin-independence with time. Despite the current successes, and residual challenges, these studies have pointed out an enormous scarcity of islet tissue that precludes the use of islet transplantation in a clinical setting on a wider scale. To address this problem, many research groups are trying to identify different islet growth factors and intracellular molecules capable of improving islet graft survival and function, therefore reducing the number of islets needed for successful transplantation. Among these growth factors, hepatocyte growth factor (HGF), a factor known to improve transplantation of a variety of organs/cells, has shown promising results in increasing islet graft survival and reducing the number of islets needed for successful transplantation in four different rodent models of islet transplantation. Protein kinase B (PKB)/Akt, a pro-survival intracellular signaling molecule is known to be activated in the beta cell by several different growth factors, including HGF. PKB/Akt has also shown promising results for improving human islet graft survival and function in a minimal islet mass model of islet transplantation in diabetic SCID mice. Increasing our knowledge on how HGF, PKB/Akt and other emerging molecules work for improving islet transplantation may provide substrate for future therapeutic approaches aimed at increasing the number of patients in which beta cell function can be successfully replaced.


DiabetesGene therapyPancreatic beta cellHepatocyte growth factorIslet transplantationProtein kinase BAkt


The incidence of diabetes is relentlessly increasing, affecting hundreds of millions of people worldwide. For many years, researchers have believed that replacement of beta cells in an immunosuppressed environment could be an ideal alternative to the continuous and intense insulin therapy in patients with diabetes. In theory, this could be achieved by transplanting the whole pancreas or by transplanting isolated islets into an appropriate recipient organ. While whole pancreas transplantation has proven to restore glucose control in patients with diabetes [1], the morbidity associated to this procedure makes difficult to advocate for most of the patients. On the other hand, islet transplantation represents an attractive and safer alternative for replenishing beta cell mass and restoring euglycemia in immunosuppressed diabetic patients [26]. Unfortunately, islet transplantation as initially employed in the 1970s and 1980s did not generate sustained efficacy: insulin independence rates in transplanted patients were disappointing [7]. In 2000, Shapiro and colleagues at the University of Alberta in Edmonton, Canada, developed an islet transplant protocol known as the “Edmonton Protocol” that improves islet allograft survival and, at least temporarily, resulted in insulin independence [36, 8]. Although the Edmonton study constituted an undeniable success, it highlighted several associated novel challenges [6, 9]. These include complications associated with the procedure itself, the toxicity of the immunosuppressants used, the progressive loss of islet graft function and insulin independence over time, and the paucity of islet tissue available for transplantation, since islets from 2 to 4 cadaver donors were needed for each transplanted patient [36, 9]. Although it subsequently has been reported that successful islet transplantation may be achieved using a single cadaver-donor or islets from a single living donor [10, 11], there remains an enormous mismatch between the number of patients with diabetes and the numbers of islets available for transplantation, rendering it currently impossible to use this protocol in a clinical setting on a wider scale. During the last two decades, many research groups have tried to find ways to expand beta cell mass, to protect beta cells against diabetogenic insults and to augment beta cell function with the idea of reducing the number of islets needed for successful transplantation [12, 13]. In this review, we focus on hepatocyte growth factor (HGF), a growth factor that has shown promising results in increasing islet engraftment and islet transplant performance in diabetic rodent models, suggesting that it may have therapeutic value for decreasing the number of islets needed for successful transplantation. Furthermore, we also review recent data indicating that activation of one of the signaling molecules stimulated by HGF and shown to be important for pancreatic beta cell growth and survival, protein kinase B/Akt, can be also a potential therapeutic target for improving islet transplantation.

Hepatocyte growth factor: an important factor for improving organ/cell transplantation

Hepatocyte growth factor (HGF) is a mesenchyme-derived protein with a wide variety of cellular activities in many different cell types [14]. HGF exhibits mitogenic, motogenic, anti-apoptotic, angiogenic, antifibrotic, and morphogenic activities in cells expressing its tyrosine kinase receptor, c-met [14, 15]. Active HGF is a disulfide-linked heterodimeric protein composed of a 69-kDa alpha-chain and a 34-kDa beta-chain containing an inactive serine-protease-like domain. Activation of HGF from a single chain precursor occurs after proteolytic processing by serine proteases, such as the blood coagulation factor XIIa, urokinase, tissue-type plasminogen activator, and HGF activator [16].

Due to its impressive capacity as an anti-apoptotic, pro-angiogenic and antifibrotic agent, HGF has being tested for improving survival and engraftment of several transplanted organs and/or cells including kidney, liver, bone marrow, heart, hepatocytes, and skeletal muscle cells [1722]. HGF has been shown to protect against acute renal damage induced by ischemia in rodent models [23, 24], suggesting that HGF could improve kidney engraftment after transplantation. In fact, a recent publication has demonstrated that HGF gene transfer into porcine kidneys ex-vivo before transplantation protects the transplanted organ against ischemic injury [17]. Interestingly, no initial tubular damage was observed, and no interstitial fibrosis was detected at 6 months post-transplantation [17]. Similarly, HGF administration has been shown to prevent chronic allograft dysfunction and to suppress fibrosis in liver-transplanted rats immunosuppressed with tacrolimus [21]. Furthermore, it has recently been reported that administration of HGF enhances the survival of cardiac allograft transplants by its cardioprotective and immunomodulative potencies [19]. In this context, HGF also improves skeletal myoblast transplantation in ischemic and infarcted rat hearts, and increases the graft volume and vascularity, at the same time reducing cardiac fibrosis [18]. All these examples, and many others in the literature, make the point that HGF can prevent the initial ischemic damage incurred by an organ/tissue/cell during the transplantation process, and prevents further injury and dysfunction of the graft.

As previously reported by Davalli et al. [25], the first few hours and days after islet transplantation under the kidney capsule in mice are characterized by substantial islet cell dysfunction and beta cell death, presumably due to the hypoxic and nutrient-deprivation conditions in which islets are now immersed. Importantly, these studies suggested that increasing the protection of islet grafts from the initial post-transplantation “damage” and accelerating the vascularization—and therefore the oxygen and nutrient supply—of the islet graft could be important therapeutic approaches for enhancing islet engraftment and performance. Since HGF increases graft survival and performance of different organs/cells, it seems logical to wonder: “Can HGF protect an islet graft against damage induced during transplantation?” “Can the mitogenic and pro-survival actions of HGF help to preserve and expand the mass of islets before and after transplantation?” “Can HGF enhance vascularization of an islet graft?” And, “Can HGF contribute to increase beta cell neogenesis from precursor cells?” Since the 1990s, several groups, including ours, have performed studies attempting to answer these questions and trying to establish whether HGF can be of therapeutic use in islet transplantation. The results are summarized in the following sections. Taken together, all the data suggest that HGF can be a very important therapeutic factor for improving beta cell growth, survival, function, and engraftment, at least in rodent models. Whether this is the case in higher species is under active investigation.

HGF/c-met and the pancreatic beta cell

HGF and c-met receptor have a widespread distribution in the body. In the pancreas, their mRNAs are highly expressed during early development and then maintained at a low level during adult life [26, 27]. Interestingly, c-met receptor expression co-localizes with insulin-expressing cells in the islet [27]. Taking advantage of this observation, several groups have studied the effects of HGF actions on the beta cell. Pioneering work by Hayek’s group demonstrated that HGF is a mitogen and an insulinotropic agent for fetal and adult human islets in vitro [2830]. Recently, Hayek’s group has demonstrated that by maintaining the three-dimensional configuration of human islets in culture, it is feasible to expand human beta cells in response to HGF, and to preserve physiologic glucose responsiveness both in vitro and in vivo after transplantation into nude mice [31]. In vivo studies performed by our group have shown that overexpression of HGF in the beta cell of adult transgenic mice directed by the rat insulin type II promoter (RIP) results in increased beta cell mass and beta cell proliferation, enhanced beta cell survival against the diabetogenic effects of streptozotocin (STZ), and improved beta cell differentiation and function [3234]. The increase in beta cell proliferation appears to be at least in part responsible for the increased islet size and concomitant increased beta cell mass observed in RIP–HGF transgenic mice. Initial analysis of the intracellular signaling mechanisms implicated in the HGF-induced mitogenic effect in the beta cell indicate that activation of phosphatidyl-inositol 3-kinase (PI3K) seems to play a central role in this effect [35].

Although an increase in beta cell proliferation could explain the increase in islet mass observed in RIP–HGF transgenic mice, a further inhibitory effect of HGF on the already low apoptotic rate in the normal beta cells in the adult pancreas is also important to consider. Indeed, HGF has been shown to protect pancreatic beta cells against the diabetogenic and cytotoxic effects of STZ in vivo and in vitro [34]. Activation of PI3K–PKB/Akt pathway seems to be one of the intracellular signaling pathways involved in the protective effect of HGF against STZ-induced cell death in pancreatic beta cells [34]. Taken together, these results confirm the cytoprotective effect of HGF in beta cells treated with STZ. However, they do not provide any insight to whether HGF protects against insults characteristic of Type 1 and 2 diabetes, such as cytokines or gluco-lipotoxicity.

Interestingly, the islet hyperplasia induced by HGF in RIP–HGF transgenic mice is associated with an increase in beta cell differentiation and function, as evidenced by increases in insulin, Glut-2 and glucokinase (GK) mRNA and protein expression in the beta cell [33]. Furthermore, RIP–HGF islets display enhanced glucose transport and metabolism, resulting in increased glucose sensitivity and augmented insulin secretion. In this context, it has been shown that HGF is capable of inducing the expression of Glut-2 in exocrine cells in culture [36, 37]. Furthermore, HGF also upregulates Na+/glucose co-transporter, SGLT1, and the facilitative glucose transporter, Glut-5, in rat intestine epithelial cells [38]. Whether upregulation of glucose transporters is a common feature of HGF in different cell types is unknown. Importantly, as a result of this increase in beta cell hyperplasia and function, RIP–HGF transgenic mice display decreased blood glucose levels in fasting and non-fasting conditions and inappropriate hyperinsulinemia in both situations [32].

RIP–HGF mice overexpressing HGF in the pancreatic beta cell display increased number of islets throughout the pancreas [32]. This could potentially be the result of HGF-induced proliferation of single beta cells that give rise to individual islets, or perhaps paracrine effects of HGF, increasing the number of progenitor cells being converted into insulin-producing cells. Regarding this last possibility, exogenous HGF, alone or in combination with activin A, is able to induce the expression of insulin and Glut-2 mRNAs in rat pancreatic AR42J exocrine cells in vitro [36].

HGF improves islet transplantation in several diabetic rodent models

In this section, we describe the studies performed with HGF to improve islet transplant outcomes and to reduce the number of islets needed for successful transplantation. These studies are summarized in Table 1. The study performed by Nakano et al. in 2000 was the first one to describe the use of HGF as a potential therapeutic agent for improving islet transplantation outcomes [39]. In this study, intraperitoneal administration of HGF and dextran sulfate for 8 days induced sustained euglycemia for 90 days in diabetic mice transplanted intraportally with a marginal mass of syngeneic islets. HGF has a very short half-life, being cleared from the circulation within minutes after injection, and sulfated oligosaccharides such as dextran sulfate have been shown to stabilize and activate HGF and facilitate c-met receptor dimerization [40]. Thus, in the Nakano study, administration of dextran sulfate was essential for HGF to induce its beneficial effects on islet transplantation, since administration of HGF alone was ineffective [39]. These results suggest either that continuous administration of HGF may be required to achieve therapeutic plasma concentrations, or that HGF potentiators may need to be administered along with HGF to improve islet transplantation outcomes. Unfortunately, transgenic mice with ubiquitous overexpression of HGF using the metallothionein promoter display markedly elevated levels of HGF, but also display diverse tumorigenesis in a broad array of tissues [41]. Taken together, these studies suggest that prolonged, high circulating levels of HGF, although therapeutically relevant, may be worrisome. This suggests that temporary local expression of HGF in the islet may be a preferable therapeutic option. Here, it is important to note that overexpression of HGF in the beta cell in transgenic mice does not result in pancreatic tumor formation during their 2-year life span.
Table 1

Summary of the different studies performed to determine the therapeutic potential of HGF for improving islet transplantation in STZ-induced diabetic recipients



Site of transplant





Syngeneic islets; marginal islet mass


HGF (100 μg) + DS ip injection/8 days




Syngeneic islets; marginal islet mass


HGF (100 μg) ip injection/8 days

No effect



SCID mouse; marginal islet mass

Renal subcapsular

RIP–HGF TG islets




SCID mouse; marginal islet mass

Renal subcapsular

Ex vivo AdvHGF-transduced islets




Allogeneic islets; marginal islet mass; immunosuppressants


Ex vivo AdvHGF-transduced islets

Reduced blood glucose levels


Abbreviations: Dextran sulfate, DS; intraperitoneal, ip; severe combined immunodeficient, SCID; transgenic, TG; and adenovirus, Adv

Based on these observations, and since overexpression of HGF in the beta cell of transgenic mice results in increased beta cell growth, survival and function, we wondered whether RIP–HGF islets might function more effectively than normal islets in a transplant setting [33]. For these experiments, we used a marginal islet mass model transplanted under the kidney capsule in STZ-induced diabetic SCID mice. In dramatic contrast to the performance of a marginal mass of normal islets, the same dose of RIP–HGF islets was able to immediately and chronically (8 weeks) normalize blood glucose levels in diabetic SCID mice. Furthermore, blood glucose levels in mice transplanted with RIP–HGF islets were comparable to those of mice transplanted with twice as many normal islets, and to normal mice, suggesting that only half as many RIP–HGF transgenic islets as normal islets are required to produce sustained normalization in blood glucose in diabetic mice. The increase in insulin content, the enhanced insulin secretion and the improved beta cell survival observed in RIP–HGF islets may have participated in the initial decrease in blood glucose levels observed in diabetic mice transplanted with RIP–HGF islets. Increased beta cell proliferation might be another mechanism implicated in the beneficial effects of RIP–HGF transgenic islet grafts.

As described above, RIP–HGF mouse islets function superiorly to normal islets in a transplant setting [33]. However, these islets have chronic overexpression of HGF and display important beneficial features already before transplantation. Therefore, we next determined whether acute overexpression of HGF into normal islets by adenovirus (Adv)-mediated gene transfer of HGF gene ex-vivo could confer the advantages that RIP–HGF islets display in the marginal islet mass transplant model in STZ-induced diabetic SCID mice [34]. In marked contrast to uninfected or Adv-β-galactosidase (LacZ)-transduced islets, AdvHGF-transduced islets were able to immediately and chronically (8 weeks) reduce blood glucose levels in the diabetic SCID renal transplant model. Marked enhancement of the size of the islet grafts was detected in kidneys carrying Adv-HGF-transduced islets, as compared with minuscule or absent islet grafts observed in the kidneys of mice transplanted with Adv-LacZ-transduced or uninfected islets [34].

As mentioned above, the first few hours and days after islet transplant are characterized by substantial islet dysfunction and death [25]. Interestingly, beta cell death rates in islet grafts obtained 24 h after the transplant were strikingly reduced in AdvHGF grafts compared with AdvLacZ grafts. In parallel, the insulin content in the AdvHGF grafts was significantly higher than in AdvLacZ grafts obtained both 24 h after the transplant. These studies collectively indicate that AdvHGF-transduced islets, when transplanted under the kidney capsule into SCID diabetic mice, clearly improve blood glucose control and graft survival, as compared to mice transplanted with uninfected or AdvLacZ-transduced islets. Whether accelerated and increased vascularization might contribute to the increased islet engraftment and performance induced by HGF is being explored.

These studies document that HGF markedly improves graft function in murine islets. We next wondered whether adenoviral gene transfer of HGF gene could also result in improved islet transplant outcomes in a marginal mass rat islet transplant model that more closely resembles human islet transplantation protocols: allogeneic islet transplant; intraportal delivery of the islets; and, islet transplant recipients immunosuppressed with the same doses and the same circulating concentrations of the immunosuppressants employed by the Edmonton group [3, 42]. The results of this study clearly demonstrated that ex vivo islet adenoviral gene therapy with HGF markedly improves islet transplant outcomes in immunosuppressed rats. This HGF-mediated improvement in islet graft function persisted for at least 4 months. More importantly, HGF was highly effective in this model of marked insulin resistance and beta cell toxicity induced by the Edmonton immunosuppression regimen. Surprisingly, normal rats exposed to the same circulating concentrations of immunosuppressants used in the Edmonton protocol displayed insulin resistance, a progressive decrease in islet function, and ultimately frank diabetes [42]. If translated to humans, these results may partially explain the decrease in islet graft survival and function, and the return to insulin dependence, observed in most patients 5 years after islet transplantation, as recently reported by the Edmonton group [6]. Taken together, these studies indicate that ex vivo adenoviral gene transfer of HGF to normal rodent islets improves islet graft function and survival and reduces the number of islets needed for transplantation.

Activation of PKB/Akt for improving islet transplantation

PKB/Akt has emerged as a focal point for multiple intracellular signal transduction pathways, regulating many cellular processes, including gene transcription, protein synthesis, cell survival and proliferation, and glucose metabolism [43, 44]. Activation of PI3K generates phosphorylated phosphoinositides that bind Akt through a pleckstrin homology (PH) domain and induce the translocation of Akt to the plasma membrane, where phosphoinositide-dependent kinase-1 (PDK-1) and probably another unknown kinase activate Akt by phosphorylation on residues Thr308 and Ser473 (in the Akt1 isoform), respectively.

Glucose and growth factors such as insulin-like growth factor I (IGF-1), glucagon-like peptide (GLP-1), glucose-dependent insulinotropic polypeptide (GIP) or HGF have been shown to increase the phosphorylation/activation of Akt in rodent beta cells [34, 4549]. Importantly, activation of the PI3K–Akt pathway by these growth factors seems to play a key role in enhancing beta cell proliferation and survival in vitro under normal as well as stress-induced conditions [34, 4549]. Furthermore, recent studies by Aikin and colleagues in human islets have shown that inhibition of PI3K/Akt pathway enhances basal and cytokine-mediated cell death [50].

Constitutively active forms of Akt have been generated by fusion of N-terminal c-Src myristoylation residues to Akt [5153]. The resultant myristoylation targets Akt directly to the membrane, where it is phosphorylated and constitutively activated inside the cell. Transgenic expression of constitutively active Akt1 in the mouse pancreatic beta cell results in increased beta cell mass (8- to 10-fold), decreased blood glucose levels, and hyperinsulinemia [51, 52]. This increase in beta cell mass is at least in part mediated by an increase in beta cell size (hypertrophy) and probably increased islet neogenesis, although this is a difficult concept to quantitate. In Bernal-Mizrachi’s studies, the authors also observed an increase in beta cell proliferation in adult mice [52], a change that was not detected in the studies of Tuttle et al. [51]. Importantly, in both studies, constitutively active Akt induces an increase in beta cell survival in vivo after treatment of transgenic mice with STZ. These studies strongly indicate that constitutively active Akt induces hyperinsulinemia, and beta cell hyperplasia and survival in vivo. Interestingly, a recent report has demonstrated that transplantation of rat adult bone marrow-derived mesenchymal stem cells genetically engineered to overexpress Akt1 can repair infarcted myocardium, prevent deleterious myocardial remodeling and restore cardiac performance [54]. Collectively, these studies suggest that activation of PKB/Akt in human islets could be an important therapeutic strategy for enhancing human islet graft survival and performance after transplantation in diabetic rodents.

Contreras and colleagues have recently reported that treatment of human islets with simvastatin markedly (58%) improved islet transplant outcomes in diabetic SCID mice [55]. This improvement was completely abolished when islets were pre-treated with the PI3K/Akt pathway inhibitor, LY294002. Based on the several studies described above, we examined whether constitutive activation of Akt in human islets might have any impact in human beta cell proliferation and survival in vitro, and in human islet transplant outcomes in a marginal mass model of islet transplant in diabetic SCID mice [56]. For this purpose, we delivered a constitutively active form of Akt1 using the adenoviral approach previously used for our HGF studies [34, 42]. As anticipated, human islets transduced with adenoviruses encoding constitutively active Akt1 (Adv-CA-Akt) displayed significantly increased beta cell replication as assessed by BrdU incorporation and a significant decrease in beta cell death induced by serum and glucose starvation or chronic hyperglycemia [56]. More importantly, constitutive activation of Akt in human islets resulted in enhanced islet graft survival and function compared with control islets transplanted into STZ-induced diabetic mice (Fig. 1). Interestingly, this superior islet graft function and survival was maintained for a prolonged period of time, 6 months. These results indicate that Akt activation in human islets may be an important therapeutic approach for enhancing human islet transplant efficiency and duration, while reducing the number of islets required for transplantation.
Fig. 1

Representative photomicrographs of kidney sections from STZ-induced diabetic mice transplanted with 1,500 Adv-CA-Akt IE, obtained 6 months after the transplant and stained for insulin (A, E, F), CK-19 (B), glucagon (C), and somatostatin (D). Insulin-positive cells inside of ductal structures can be detected in the grafts (arrows in E and F), suggesting beta-cell neogenesis. (G) Blood glucose levels in diabetic mice transplanted with 1,500 Adv-LacZ- or 1,500 Adv-CA-Akt-transduced IE. The post-transplant blood glucose values in mice transplanted with Adv-CA-Akt-transduced IE were significantly (P < 0.05) lower vs. mice transplanted with Adv-LacZ for the entire 6-month period until UNX was performed and the blood glucose returned immediately to pre-transplant diabetic levels. Results are means ± SE. Adapted from Ref. [56], with permission, copyright by the American Diabetes Association

Summary and conclusion

In summary, the studies described in this brief review clearly indicate that HGF, and its downstream target, PKB/Akt, are effective in improving islet transplant outcomes in models of islet transplantation in diabetic rodents. These are mainly “proof of principle” studies and there remains a long list of questions and issues that need to be addressed before therapeutic testing in humans. For example, “since the studies are performed in rodents, will HGF gene therapy improve islet transplantation in higher species?” “Are there any synergistic factors that may help HGF to further decrease the number of islets needed for successful transplantation?” “Will HGF protein administration be as effective as HGF gene therapy for islet transplantation?” “Is it possible to find less tumorigenic and more stable HGF peptides that may be used to improve islet transplantation?” “Since adenoviruses used in the gene therapy studies described above do not integrate, are easy to make and widely available, and since they disappear in a relative short time [57], is adenoviral-mediated delivery of HGF or Akt a reasonable means of gene delivery for islet transplantation in immunosuppressed recipients?” “Since islets are normally transplanted into the liver, will HGF have any local adverse effects in liver cells?” “What are the different molecular targets downstream of Akt that are important for its beneficial effects in islet survival and function after transplantation?” “Is it possible to find small molecules that can be administered in islet-transplanted patients to activate Akt, enhancing insulin sensitivity and improving islet graft survival?” Answering these questions, and many others, will provide the knowledge needed to determine whether HGF or PKB/Akt will prove to be useful tools for enhancing beta cell replacement in patients with diabetes.


This work was supported by National Institutes of Health grants DK068836 and DK067351 to A.G.-O. and DK55023 and DK R33066127 to AFS.

Copyright information

© Humana Press Inc. 2007