Gene Transfer of Constitutively Active Akt Markedly Improves Human Islet Transplant Outcomes in Diabetic Severe Combined Immunodeficient Mice

Adenovirus containing the cDNA of a constitutively active form of Akt (Adv-CA-Akt) or β-galactosidase were prepared according to methods we have previously described in detail (13), originally described by Becker et al. (19). The adenovirus LacZ (Adv-LacZ) was generously provided by Dr. Christopher Newgard (Duke University, Durham, NC). The Adv-CA-Akt was a generous gift from Dr. Thomas Soderling (Vollum Institute, Portland, OR). The constitutively active form of Akt contains the NH₂-terminal c-Src myristolation signal fused to Akt, which directs Akt to the plasma membrane leading to its constitutive activation, and a hemagglutinin tag for monitoring expression. Multiplicity of infection (MOI) was determined by optical density at 260 nm and by plaque assay. MOI calculations assume 1,000 cells per islet equivalent (IEQ). Human islets were kept in culture for 4–7 days before they were transduced with the adenoviruses as previously described in detail (13). Uninfected and adenovirus-transduced human islets were used 24 h after infection for the different experiments described below.

Human islets from pancreatic cadaver donors (aged 39 ± 7 years, n = 11) were provided by the University of Pittsburgh Human Islet Isolation Facility and were prepared according to Ricordi and Rastellini (20) as previously described in detail (21).

Islet extracts were made in freshly prepared ice-cold lysis buffer (PBS with 1% Nonidet P-40, 0.5% sodium deoxyxcholate, 0.1% sodium dodecyl sulfate, 1 mmol/l sodium orthovanadate, 100 μg/ml phenylmethylsulfonyl fluoride, and 57 μg/ml aprotinin) incubated 30 min on ice and sonicated; the supernatant containing the cell lysate was separated; and protein concentrations were measured using the MicroBCA assay (Pierce, Rockford, IL). Twenty-five micrograms of protein from uninfected or adenovirus infected islets was resuspended in loading buffer containing β-mercaptoethanol, boiled for 5 min, and analyzed using a 10% SDS-polyacrylamide gel. Proteins were transferred to Immobilon-P membrane (Millipore, Bedford, MA) using standard techniques. Blots were then probed with antibodies against Akt, phospho-Ser473-Akt, phospho-Thr308-Akt (Cell Signaling Technology, Beverly, MA), and actin (Santa Cruz Biotechnology, Santa Cruz, CA) following previously described methods (13,22). Chemiluminescence was detected using the ECL system (Amersham Pharmacia, Piscataway, NJ).

The Akt immunoprecipitation and kinase assay was carried out using a glycogen synthase kinase (GSK)-3α/β phosphorylation kit (Cell Signaling Technology) following the manufacturer’s procedure and as previously described (13). Briefly, 200 μg of islet protein extracts obtained 24 h after adenovirus transduction or from uninfected islets was incubated with protein G/Akt1 monoclonal antibody complex for 2 h at 4°C. After several washes, immunoprecipitates were resuspended in kinase buffer containing 200 μmol/l ATP and 1 μg of GSK-3 fusion protein and incubated at 30°C for 30 min. The reaction was stopped with SDS gel loading buffer, and samples were analyzed by immunoblotting (as described above) with the phospho-GSK-3α/β (Ser 21/9) antibody.

Insulin release from uninfected and adenovirus-transduced islets was measured in triplicate for each glucose concentration tested, as previously described (12,22,23). Briefly, uninfected and adenovirus-transduced islets were preincubated in Krebs-Ringer bicarbonate buffer supplemented with 10 mmol/l HEPES, 1% BSA, and 5 mmol/l glucose for 1 h at 37°C in a 5% CO₂ incubator. After washing the islets once with the same solution, groups of 10 islets of similar size for each condition were incubated in 1 ml of fresh Krebs-Ringer bicarbonate buffer plus 1% BSA and 5 or 22 mmol/l glucose at 37°C in the 5% CO₂ incubator. After 30 min of incubation, buffer was removed and frozen at −20°C until insulin measurement by radioimmunoassay (Linco Research, St. Louis, MO). Islets were then digested overnight in NaOH at 37°C, and protein was measured by the Bradford method after neutralization with HCl. Results are expressed as a percentage of insulin concentration obtained with uninfected islets incubated at 5 mmol/l glucose.

Human islet cell proliferation was assayed as reported previously (22). Briefly, uninfected or adenovirus-transduced islets were incubated in RPMI medium containing 5 mmol/l glucose and 1% penicillin and streptomycin for 24 h at 37°C. One microcurie of [³H]thymidine (Amersham Pharmacia) was then added to the medium for an additional 24 h. To quantitate [³H]thymidine incorporation, islets were washed with ice-cold PBS and precipitated in ice-cold 10% trichloroacetic acid twice for 10 min on ice. The resulting precipitant was resuspended in 0.1 N NaOH at 60°C for 30 min and neutralized with 0.1 N HCl. Aliquots were counted in a β-scintillation counter and assayed for protein using the Bradford assay.

For immunohistological confirmation of [³H]thymidine incorporation, human islets were labeled with bromodeoxyuridine (BrdU) using the same protocol as described above except that a 1:1,000 dilution of BrdU (Amersham Cell Proliferation kit) was added to the culture medium in place of [³H]thymidine for 18–20 h. The isolated islets were fixed in Bouin’s solution for 1 h at room temperature and stored in neutral-buffered formalin before embedding in paraffin, as previously described (22). Briefly, serial 5-μm sections of the isolated islets were deparaffinized, rehydrated, microwaved in citrate buffer for 5 min, quenched with 3% H₂O₂ in methanol for 10 min at room temperature, rinsed, and blocked with 2% goat serum. Sections were then incubated with an anti-insulin antibody (Zymed, South San Francisco, CA) at 1:50 dilution and an anti-BrdU antibody (Amersham Pharmacia) at 1:5 dilution for 1 h at room temperature. Visualization of the staining was achieved by using fluorescein isothiocyanate–conjugated anti-guinea pig IgG and tetramethylrhodamine isothiocyanate (TRITC)–conjugated rabbit anti-mouse IgG secondary antibodies (Sigma), respectively. Hoestch 33258 (Sigma) was used to detect all the nuclei in the sections.

Twenty-four hours after infection with the adenovirus, adenovirus-transduced and uninfected islets were incubated in RPMI medium without glucose and fetal bovine serum (FBS) for 24 h or in RPMI medium containing FBS and either 5 or 22 mmol/l glucose for 5 days (chronic hyperglycemia). At the end of the incubation period, islets were fixed in Bouin’s solution, embedded in paraffin, and sectioned. β-Cell death was detected by enzymatic in situ labeling for DNA strand breaks using the terminal deoxynucleotidyltransferase-mediated dUTP nick end-labeling (TUNEL) method using the In Situ Cell Death Detection kit (Roche Molecular Biochemicals, Mannheim, Germany) according to the manufacturer’s protocol. Subsequent to TUNEL, sections were stained with the anti-insulin antibody as described above. Visualization was performed by using a TRITC-conjugated rabbit anti-guinea pig IgG secondary antibody. Hoestch 33258 was used to detect all of the nuclei in the sections.

Human islets, uninfected or transduced with Adv-LacZ or Adv-CA-Akt as described above, were transplanted under the kidney capsule of STZ-induced diabetic SCID mice (BALB/cByJ) (Jackson Laboratory, Bar Harbor, ME) as previously described (13,23). Briefly, SCID mice were rendered diabetic by injecting 250 mg/kg body wt STZ i.p. Diabetes was confirmed by the presence of hyperglycemia (>300 mg/dl), polyuria, and weight loss. Random nonfasted blood glucose was measured from the snipped tail by a Precision Q.I.D. portable glucometer (Medisense, Bedford, MA). After 3 consecutive days of hyperglycemia, SCID diabetic mice were transplanted with either 1,500 or 4,000 uninfected, 1,500 Adv-LacZ–or 1,500 Adv-CA-Akt–transduced IEQs (1 IEQ = 125-μm-diameter islet) beneath the kidney capsule, as previously described (13,23). Blood glucose levels were measured at days 1 and 7 and then weekly until the transplanted mice underwent unilateral nephrectomy. At day 54, retroorbital bleeds were performed to measure plasma human insulin levels with a human insulin-specific radioimmunoassay (Linco). All studies were performed with the approval of and in accordance with guidelines established by the University of Pittsburgh Institutional Animal Care and Use Committee.

Forty-nine days after islets were transplanted in STZ-induced SCID diabetic mice (see above), glucose tolerance was analyzed in 16-h–fasted mice by injection of 2 g glucose/kg body wt i.p. as previously reported (13,23). Blood samples were obtained from the snipped tail and analyzed for glucose levels using the portable glucometer.

Kidneys containing islet grafts were fixed in Bouin’s solution, paraffin embedded, sectioned, and stained with the anti-insulin antibody (see above) and an anti-cytokeratin-19 (anti-CK-19) antibody (Novocastra/Vector Labs, Burlington, Ontario, Canada) at 1:75 dilution. Similarly, sections were also stained with anti-glucagon (Biogenex, San Ramon, CA) at 1:50 dilution and anti-somatostatin antibodies (Novacastra, U.K.) at 1:200 dilution. Visualization of the staining was achieved by using the avidin-biotin immunoperoxidase complex system (Biogenex). Sections were counterstained using hematoxylin.

To detect the presence of cells expressing CA-Akt in adenovirus-transduced islets and islet grafts, sections were stained with an anti-hemagglutinin tag monoclonal antibody (Cell Signaling Technology) at 1:300 dilution and with the anti-insulin antibody described above. Subsequently, fluorescein isothiocyanate–or TRITC-conjugated rabbit anti-mouse and rabbit anti-guinea pig IgG secondary antibodies (Zymed) were used. Slides were then cover-slipped using gelmount antifade (Biomeda, Foster City, CA).

Results are expressed as means ± SE. Statistical differences were determined by a two-tailed unpaired Student’s t test. P < 0.05 was considered statistically significant.

We first examined the ability of Adv-CA-Akt to transduce human islets and whether this transduction resulted in overexpression of phosphorylated Akt. As shown in Fig. 1A, 24 h after the transduction of human islets with 500 MOI of Adv-CA-Akt, ∼40–50% of islet cells were transduced with Adv-CA-Akt as determined by immunofluorescent staining with an antibody against the hemagglutinin epitope tag. Furthermore, the expression of Akt and its phosphorylation status was examined by Western blot analysis. As shown in Fig. 1B, exposure of human islets to increasing doses of Adv-CA-Akt (250–1,000 MOI) for 1 h at 37°C resulted in an increase in Akt expression at 24 h after transduction. Activation of Akt1 requires phosphorylation of residues Ser473 and Thr308 (5–8). Western blot analysis of extracts from human islets exposed to Adv-CA-Akt revealed an increase in the expression of Akt phosphorylated on Ser473 and Thr308 (Fig. 1B). This indicates that exposure of human islets to Adv-CA-Akt results in overexpression and increased phosphorylation of Akt.

Next, we examined whether phosphorylation of Akt in extracts from Adv-CA-Akt–transduced human islets resulted in increased Akt kinase activity compared with uninfected and Adv-LacZ–transduced human islets. As shown in a representative immunoblot in Fig. 1C, Akt activity is significantly four- to sevenfold increased in human islets exposed to 500 MOI of Adv-CA-Akt (7.6 ± 1.7-fold, n = 5, P < 0.05) compared with human islets exposed to the same MOI of Adv-LacZ (1.9 ± 0.9-fold, n = 5) or uninfected islets (1.0 ± 0.3-fold, n = 5). Taken together, these results demonstrate that adenovirus containing the cDNA of a constitutively active form of Akt is able to: 1) efficiently transduce human islet cells; and 2) increase Akt expression, phosphorylation, and activation in human islets.

To determine an optimal dose of Adv-CA-Akt that induces overexpression and activation of Akt in human islets without adversely affecting insulin secretion, glucose-stimulated insulin secretion (GSIS) was examined 24 h after adenoviral transduction. The insulin secretory response from human islets transduced with 500 MOI of either Adv-LacZ or Adv-CA-Akt was similar to nontransduced islets, indicating that islet function is not adversely attenuated by this dose of adenovirus (Fig. 1D). In contrast, GSIS was negatively affected in islets transduced with 1,000 MOI of either adenovirus. Taken together, the results shown in Fig. 1 suggest that 500 MOI is an optimal and perhaps maximal MOI that can be used for subsequent experiments.

To determine whether constitutively active Akt could also induce an increase in the proliferation rates of human islet cells as has been reported to occur in rodent islets (10), we performed [³H]thymidine incorporation experiments in Adv-CA-Akt–transduced and control human islets. To eliminate variability in islet size, we selected islets of similar size and also expressed [³H]thymidine incorporation (counts per minute) per microgram of protein. As shown in Fig. 2A, human islets transduced with 500 MOI of Adv-CA-Akt displayed a twofold increase (P < 0.05) in [³H]thymidine incorporation compared with uninfected or Adv-LacZ–transduced human islets. These studies indicate that Adv-CA-Akt delivery to human islets increases islet cell proliferation. However, they do not provide any information on whether the proliferation is occurring in β-cells. To determine whether this is the case, sections from adenovirus-transduced and uninfected human islets were stained for insulin and BrdU. The percentage of insulin-positive cells, expressed as a function of total number of cells, in the different human islet preparations used in the current studies was 55 ± 10% (n = 11). As can be observed in the representative panels included in Fig. 2B, the number of BrdU-positive nuclei in sections from islets transduced with 500 MOI of Adv-CA-Akt appears to be clearly increased. Quantitation of BrdU incorporation rates indicated that constitutive activation of Akt in human islets significantly increase (1.8-fold, P < 0.05) islet cell replication compared with uninfected or Adv-LacZ–transduced human islets (Fig. 2C). Furthermore, quantitation of insulin- and BrdU-positive cells in those sections indicated that constitutive activation of Akt resulted in a significant increase in proliferation of β-cells (Fig. 2D). Taken together, these results indicate that adenoviral delivery of constitutively active Akt stimulates human β-cell proliferation in vitro.

Akt is a survival factor in rodent systems (9,10,12,18). To examine whether constitutive activation of Akt increases the survival of human islet β-cells in vitro, we exposed Adv-CA-Akt–transduced and control islets to two cell death inducers: combined glucose and serum deprivation and chronic high glucose levels (24–26).

First, the ability of constitutively active Akt to protect against glucose and serum deprivation-induced β-cell apoptosis was evaluated. As shown in the representative photomicrograph in Fig. 3A, exposure of uninfected or Adv-LacZ–transduced human islets to RPMI medium without glucose and serum for 24 h resulted in an increased number of nuclei displaying DNA fragmentation (TUNEL-posi-tive nuclei) compared with islets grown in regular conditioned medium (Fig. 3A). Interestingly, and in striking contrast to uninfected or Adv-LacZ–transduced islets, Adv-CA-Akt–transduced human islets cultured for 24 h in the absence of glucose and serum show a striking reduction in the number of TUNEL-positive nuclei (Fig. 3A). Quantitation of the number of insulin- and TUNEL-positive nuclei in those sections indicate that survival in Adv-CA-Akt–transduced human islets from apoptosis induced by glucose and serum deprivation is increased by ∼80% (Fig. 3B).

Next, we exposed adenovirus-transduced and uninfected human islets to increasing glucose concentrations (from 5 to 22 mmol/l) for 5 days. As shown in the representative photomicrograph of Fig. 3C, chronic high glu-cose levels increased the number of TUNEL-positive nuclei in uninfected or Adv-LacZ–transduced human islets compared with islets grown in regular conditioned medium containing 5 mmol/l glucose. However, islet sections from Adv-CA-Akt–transduced islets treated with high glucose for 5 days displayed reduced numbers of apoptotic nuclei compared with control islets (Fig. 3C). Quantitation of β-cell apoptotic rates in these sections revealed that transduction with Adv-CA-Akt increased survival of human β-cells treated with high glucose concentrations for 5 days by ∼80% (Fig. 3D). In addition, transduction with Adv-CA-Akt did not significantly change baseline apoptotic levels at 5 mmol/l glucose (Figs. 3B and D). Taken together, these results indicate that constitutive activation of Akt confers a survival advantage to human islet β-cells exposed to two different cell death inducers in vitro: glucose and serum deprivation and chronic high glucose levels.

The results presented above suggest that human islets transduced with Adv-CA-Akt might function more efficiently than normal islets in a transplant setting. To address this hypothesis, we compared the ability of a marginal mass, 1,500 IEQs of uninfected, Adv-LacZ–and Adv-CA-Akt–transduced human islets to normalize blood glucose in STZ-induced diabetic SCID mice, as previously performed using mouse islets (13,23). As shown in Fig. 4A, transplantation of 1,500 uninfected or Adv-LacZ–transduced human IEQs under the kidney capsule of diabetic SCID mice did not induce euglycemia in these mice, indicating that this amount of islets was clearly insufficient to achieve sustained normal blood glucose levels. In marked contrast, when 1,500 Adv-CA-Akt IEQs were transplanted under the kidney capsule of diabetic SCID mice, they rapidly normalized blood glucose concentrations, and the mice remained euglycemic throughout the 56-day period of the study (Fig. 4A). The kidney bearing the graft was removed in three of six mice transplanted with Adv-CA-Akt–transduced islets at day 56 after transplantation. Blood glucose levels promptly returned to the original pretransplant values, indicating that the human islets engrafted under the kidney capsule were responsible for the euglycemia observed in these mice. Importantly, random, nonfasting total (human plus rodent) circulating insulin levels were significantly higher in mice transplanted with Adv-CA-Akt–transduced islets compared with mice transplanted with Adv-LacZ–transduced islets (2.90 ± 0.7 ng/ml [n = 6] vs. 0.80 ± 0.2 ng/ml [n = 5], P < 0.05). Furthermore, human insulin levels at day 54 after transplantation were 10-fold higher (P < 0.05) in SCID mice transplanted with Adv-CA-Akt–transduced islets (2.4 ± 0.4 ng/ml, n = 6) compared with mice transplanted with Adv-LacZ–transduced islets (0.25 ± 0.08 ng/ml, n = 5). In addition, blood glucose levels in 1,500-IEQ Adv-CA-Akt–transplanted mice were similar to the blood glucose levels reached with 4,000 uninfected IEQs, an amount that was sufficient to induce euglycemia in SCID diabetic mice.

To further define the function of the transplanted Adv-CA-Akt–transduced islets, we performed intraperitoneal glucose tolerance tests (IPGTTs) on day 49 in mice transplanted with adenovirus-transduced islets and in normal nondiabetic littermates. After an overnight fast, blood glucose levels were significantly (P < 0.05) lower in SCID mice transplanted with Adv-CA-Akt islets (81 ± 6 mg/dl, n = 6) as compared with SCID mice transplanted with Adv-LacZ islets (181 ± 27 mg/dl, n = 5) (Fig. 4B). Importantly, after glucose injection, glucose tolerance in mice transplanted with Adv-CA-Akt–transduced islets was similar to that in normal nondiabetic mice and was remarkably and significantly (P < 0.025) superior to that in the mice transplanted with Adv-LacZ–transduced islets (Fig. 4B). These results indicate that Adv-CA-Akt–transduced islets confer improved fasting and nonfasting blood glucose control and superior glucose tolerance when transplanted in diabetic SCID mice.

Immunohistochemical analysis of sections from kidneys containing the grafts obtained at 56 days after the transplant of 4,000 uninfected or 1,500 Adv-CA-Akt–transduced IEQs revealed the presence of insulin-positive cells and ductal structures staining for the ductal cell marker CK-19 (Figs. 4C–F). Furthermore, insulin-positive cells could be seen within the ductal structures, suggesting the presence of β-cell neogenesis from the ductal formations in both types of grafts (Figs. 4G–H).

To assess the long-term function of Adv-CA-Akt–transduced islet grafts implanted in diabetic SCID mice, we allowed one-half of the Adv-CA-Akt mice in Fig. 4 to retain their graft for 6 months. As shown in Fig. 5, normalization of blood glucose levels in diabetic mice transplanted with Adv-CA-Akt was sustained for the complete 6-month period of the study compared with the diabetic blood levels observed in mice transplanted with Adv-LacZ–transduced islets. Importantly, when uninephrectomy was performed in mice transplanted with Adv-CA-Akt–transduced islets at 6 months after the transplant, blood glucose levels increased to pretransplant levels, indicating that the graft was responsible for the normal blood glucose observed in these mice. These results indicate that the improvement in blood glucose induced by Adv-CA-Akt islet grafts in diabetic SCID mice is persisting, lasting at least one-fourth of the murine life span

Next, we immunohistochemically analyzed sections from kidneys containing the grafts obtained at 6 months after the transplant. Grafts were easily and abundantly visualized in the kidneys containing Adv-CA-Akt–transduced islets. This was in marked contrast with kidneys from mice transplanted with Adv-LacZ islets in which islet grafts were undetectable. As shown in the representative photomicrographs included in Fig. 6, adenovirus-CA-Akt islet grafts contained insulin-, glucagon-, and somatostatin-positive cells, as well as ductal structures staining for CK-19 (Figs. 6A–D). Furthermore, insulin-positive cells can also be seen within the ductal structures (Figs. 6E–F). To determine whether Adv-CA-Akt–transduced islet cells were still present in the grafts 6 months after the transplant and taking advantage of the presence of an hemagglutinin tag in CA-Akt (Fig. 1A), we stained kidney sections containing adenovirus-CA-Akt islet grafts with an antibody against the hemagglutinin epitope tag. As shown in the representative photomicrographs in Fig. 6 (G–H), cells positive for hemagglutinin can be detected in the grafts in insulin-positive and insulin-negative cells. As expected, no hemagglutinin staining was detected in 4,000 uninfected IEQ grafts (Figs. 6I–J).

Collectively, these results indicate that Adv-CA-Akt–transduced islets, when transplanted under the kidney capsule into SCID diabetic mice, clearly improve graft survival compared with the mice transplanted with Adv-LacZ–transduced islets.

The recent demonstration that human pancreatic islet transplantation is both technically and immunologically feasible has revealed the need for large numbers of human pancreatic islets to treat an increasing number of patients with diabetes (1–3). This scarcity of human islets might potentially be reduced by increasing proliferation, differentiation, function, and survival of the donated human pancreatic islets. β-Cell–specific expression of constitutively active Akt1 in mice results in increased β-cell proliferation, differentiation, and survival (9,10). However, the effects of constitutive activation of Akt in human β-cell proliferation, survival, and transplant performance are entirely unknown. In the studies described herein, we document three novel observations: adenovirus-mediated expression of constitutively active Akt1 (Adv-CA-Akt) in human islets results 1) in increased human β-cell proliferation, 2) in enhanced survival in vitro, and 3) more importantly, in prolonged improvement in human islet transplant outcomes in diabetic SCID mice.

Activation of Akt is associated with increased proliferation in several cell types (8,27). However, whether activation of Akt is also associated with increased pancreatic β-cell replication remains controversial. Several pancreatic β-cell mitogens such as GLP-1, GIP, HGF, and glucose have been shown to stimulate Akt activity in these cells (15–18). In all of these cases, inhibition of the PI3K-Akt pathway reduced or blunted β-cell proliferation induced by these factors. In addition, studies performed by Dickson et al. (11) have demonstrated that adenoviral delivery of CA-Akt into INS-1 cells enhanced two- to threefold β-cell mitogenesis at 3 mmol/l glucose but did not significantly alter β-cell replication at 15 mmol/l glucose. Furthermore, Bernal-Mizrachi et al. (10) have recently reported that constitutive activation of Akt in the pancreatic β-cell of transgenic mice is associated with an increase in adult β-cell replication. On the other hand, Tuttle et al. (9) have observed that transgenic mice overexpressing constitutively active Akt in the pancreatic β-cell have β-cell proliferation rates unaffected. These, in some cases, conflicting studies have been performed in rodent pancreatic β-cells in vitro or in transgenic mice in vivo. However, the role of Akt activation on human β-cell proliferation has not been studied. To begin to evaluate this question, we have analyzed proliferation rates in control human islets and in islets transduced with Adv-CA-Akt by two different methods: [³H]thymidine and BrdU incorporation. In these proliferation studies performed at 5 mmol/l glucose, we observed that gene delivery of CA-Akt induced an increase in human islet cell proliferation, and, more importantly, constitutive activation of Akt resulted also in specifically increased β-cell mitogenesis. These results indicate that activation of Akt could expand human pancreatic β-cells in culture, a result that has potential clinical value given the urgent need to increase the availability of human β-cells for transplantation.

Therapeutic interventions aimed to inhibit β-cell death will help to increase cell viability and function during and after islet isolation and islet transplantation. It has been extensively reported that Akt activation plays a pivotal role in cell survival in many cell systems (5,6,8). Among these studies, it has been shown that blockade of the PI3K/Akt pathway in rodent β-cells blunts the pro-survival effects induced by glucose, HGF, GIP, GLP-1, or IGF-I in these cells (12,13,16–18). In addition, transfection of MIN-6 cells with a constitutively active form of Akt enhances the survival effect of glucose in these cells (18). Furthermore, expression of constitutively active Akt in the β-cell of transgenic mice confers striking resistance to cell death induced by the diabetogenic and cytotxic agent STZ (9,10). Collectively, these rodent studies suggest that Akt activation could be an important pro-survival event for improving human β-cell resistance to cell death. To assess this hypothesis, we studied the effect of constitutive activation of Akt on human β-cell survival in vitro by using two different cell death inducers implicated in the massive β-cell loss that occurs immediately after islet transplantation (28) and in β-cell failure in type 2 diabetes (29): glucose and serum deprivation and chronic hyperglycemia. We found that gene transfer of CA-Akt protects human β-cells against cell death induced by both types of insults. Glucose and serum deprivation increased apoptotic rates more than eightfold in human β-cells in uninfected or Adv-LacZ–transduced islets. In marked contrast, Adv-AC-Akt–transduced human islets displayed markedly reduced number of apoptotic β-cells, similar to the apoptotic rates found in islets incubated in normal conditioned medium. In addition, incubation of control (uninfected or Adv-LacZ–transduced) human islets in 22 mmol/l glucose for 5 days resulted in a four- to sixfold increase in β-cell apoptosis. Again, CA-Akt transfer in human islets resulted in a striking decrease in the number of apoptotic β-cells. Contreras et al. (30) have recently demonstrated that activation of Akt is essential for the cytoprotective effect induced by the cholesterol-lowering drug simvastatin on isolated human islets in vitro. Furthermore, recent studies by Aikin et al. (31) have shown that inhibition of PI3K/Akt pathway enhances basal and cytokine-mediated cell death in human islets. These studies, together with the present observations, suggest that activation of Akt in human islets could be an important therapeutic strategy for enhancing human β-cell survival under normal and stress-induced conditions.

A key question that emerges from these initial in vitro studies is whether constitutive activation of Akt can improve human islet engraftment and performance when transplanted in diabetic mice. The results obtained in the studies presented herein are very clear: gene transfer of constitutively active Akt to a marginal mass of human islets dramatically improved human islet graft function and survival in a transplant setting in diabetic SCID mice. Importantly, this euglycemic effect in the mice transplanted with CA-Akt–transduced islets was long lasting. Contreras et al. (30) have recently reported that treatment of human islets with simvastatin improved islet transplant outcomes in 58% of the transplanted diabetic SCID mice. Furthermore, this improvement in the islet transplant outcomes was completely abolished when islets were pretreated with the PI3K/Akt pathway inhibitor, LY294002. Taken together, these studies suggest that activation of Akt in human islets could decrease the number of islets needed for successful transplantation.

In the present studies, we have used an adenovirus system to deliver a constitutively active form of Akt1 into human islets. The rate of infection achieved with this system is similar to that achieved in other studies (30–70%) in which rodent or human islets have been transduced with adenoviruses (13,32,33). Because Akt is an intracellular protein, one can argue that with the use of the adenoviral system only 30–70% of the cells will acquire the potential positive benefits that constitutively active Akt may promote in human islet cells. Regarding the importance of these findings for human islet transplantation, three points are relevant. First, the present studies serve as a “proof of principle” studies designed to determine whether activation of Akt in human islets is beneficial for these islets. The answer to this question is surely affirmative. Second, unlike rodent islets, human islets seem to display a more peripheral distribution of β-cells (Figs. 1–3) (22), which could result in increased efficiency of adenoviral transduction in β-cells. Third, despite transducing only a portion of the cells in the islet, human islets infected with Adv-CA-Akt display increased proliferation, enhanced survival, and improved human islet transplant outcomes, as shown in Figs. 2–5. This is very significant, because one can readily envision future vector approaches that will result in enhanced efficiency of transduction, improved β-cell specificity, an adjustable level of expression, and a regulatable temporal expression pattern.

In the current studies, we detect the presence of CA-Akt in graft cells at 6 months after the transplant. Because overexpression of CA-Akt increases human islet cell proliferation and survival in vitro, one might worry that CA-Akt gene delivery into human islets might cause abnormal proliferation or tumorigenesis of graft cells. Although, we are not aware of any report indicating that continuous Akt activation is implicated in generation of human insulinomas, amplification and overexpression of Akt2 in a subset of human pancreatic ductal adenocarcinomas has been reported (34). Several points are relevant here. First, as mentioned above, the current study was intended as a “proof of principle” to analyze whether Akt activation in human islets is beneficial for improving islet transplant outcomes. Second, diabetic SCID mice transplanted with Adv-CA-Akt–transduced human islets do not apparently display hypoglycemia, a crude measure of islets tumors. Third, although we observed the presence of a large number of ductular structures in the CA-Akt islet graft at 6 months after the transplant, we did not find any invasion of the graft into the kidney. In fact, similar observations have been reported by others in transplants performed with human islets not treated with CA-Akt (35).

In summary, our studies demonstrate that adenoviral delivery of CA-Akt to human islets ex vivo improves islet transplant survival and performance in STZ-induced diabetic SCID mice. These studies demonstrating that CA-Akt expression can increase human β-cell proliferation and survival reveal two critical components required for expanding human β-cells in vitro and in vivo. These experiments also suggest that CA-Akt could play an important role in preventing early failure and promoting expansion of the islet graft. These beneficial effects could be mediated by constitutively active Akt or by other factors secreted by Adv-CA-Akt–transduced cells that could also affect nontransduced cells in a paracrine way. Controlled activation of Akt may be of therapeutic use for expanding and protecting isolated human islets, improving islet transplant outcomes, and reducing the number of islets required for successful islet transplantation.

This work was supported by grants from the National Institutes of Health (DK-067351 and DK-068836), the Juvenile Diabetes Foundation International, and the American Diabetes Association (Junior Faculty grant).

We are grateful to Dr. Andrew F. Stewart and Dr. Rupangi C. Vasavada for the frequent and thoughtful discussions of the ideas in this article. We thank Vasumathi Reddy and Darinka Sipula for superb technical assistance.