Overexpression of Glutathione Peroxidase With Two ... - Diabetes

Overexpression of Glutathione Peroxidase With Two Isoforms of Superoxide Dismutase Protects Mouse Islets From Oxidative Injury and Improves Islet Graft Function Tharun B. Mysore,1 Trixie A. Shinkel,1 James Collins,2 Evelyn J. Salvaris,1 Nella Fisicaro,1 Lisa J. Murray-Segal,1 Lucinda E.A. Johnson,1 Diana A. Lepore,1 Stacey N. Walters,2 Rebecca Stokes,2 Abhilash P. Chandra,2 Philip J. O’Connell,2 Anthony J.F. d’Apice,1,3 and Peter J. Cowan1,3

Primary nonfunction of transplanted islets results in part from their sensitivity to reactive oxygen species (ROS) generated during the isolation and transplantation process. Our aim was to examine whether coexpression of antioxidant enzymes to detoxify multiple ROS increased the resistance of mouse islets to oxidative stress and improved the initial function of islet grafts. Islets from transgenic mice expressing combinations of human copper/zinc superoxide dismutase (SOD), extracellular SOD, and cellular glutathione peroxidase (Gpx-1) were subjected to oxidative stress in vitro. Relative viability after hypoxanthine/xanthine oxidase treatment was as follows: extracellular SOD ⴙ Gpx-1 ⴙ Cu/Zn SOD > extracellular SOD ⴙ Gpx-1 > extracellular SOD > wild type. Expression of all three enzymes was the only combination protective against hypoxia/ reoxygenation. Islets from transgenic or control wildtype mice were then transplanted into streptozotocininduced diabetic recipients in a syngeneic marginal islet mass model, and blood glucose levels were monitored for 7 days. In contrast to single- and double-transgenic grafts, triple-transgenic grafts significantly improved control of blood glucose compared with wild type. Our results indicate that coexpression of antioxidant enzymes has a complementary beneficial effect and may be a useful approach to reduce primary nonfunction of islet grafts. Diabetes 54:2109 –2116, 2005

From the 1Immunology Research Centre, St. Vincent’s Hospital, Melbourne, Australia; 2the Centre for Transplant and Renal Research, Westmead Millennium Institute, University of Sydney, Sydney, Australia; and 3the Department of Medicine, University of Melbourne, St. Vincent’s Hospital, Melbourne, Australia. Address correspondence and reprint requests to Peter J. Cowan, PhD, Immunology Research Centre, St. Vincent’s Hospital Melbourne, P.O. Box 2900, Fitzroy 3065, Victoria, Australia. E-mail: [email protected]. Received for publication 26 July 2004 and accepted in revised form 11 April 2005. T.B.M. and T.A.S. contributed equally to this work. P.J.O. is a member of the Rapimmune Advisory Panel for Wyeth Australia and of the Australian National Advisory Panel for Novartis. D.A.L. is currently affiliated with the Pituitary Research Unit, Murdoch Children’s Research Institute, Royal Children’s Hospital, Melbourne, Australia. BGL, blood glucose level; FOX, ferric-xylenol orange; GPI, glycophosphatidyl inositol; Gpx-1, glutathione peroxidase; HX/XO, hypoxanthine/xanthine oxidase; IFN-␥, interferon-␥; MTT, 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide; ROS, reactive oxygen species; SOD, superoxide dismutase. © 2005 by the American Diabetes Association. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

DIABETES, VOL. 54, JULY 2005

I

slet transplantation has the potential to replace pancreatic endocrine function in patients with type 1 diabetes (1), but the requirement for two or more cadaveric donors per recipient limits wider application of the technique. This problem has led to investigation of the causes of islet primary nonfunction, so that protective strategies may be developed. A significant number of transplanted islets are lost because of apoptosis and necrosis (2), mediated at least in part by oxidative injury. Native islets are richly supplied with blood but lose their vasculature during isolation (3), and full revascularization takes weeks (4). Hypoxia/reoxygenation generates reactive oxygen species (ROS), including superoxide radicals (O2䡠⫺), hydrogen peroxide (H2O2), and hydroxyl radical (䡠OH) (5). In addition, nitric oxide (NO) is generated by upregulation of inducible NO synthase and from nitrated residues of necrotic cells (6). Simultaneous generation of superoxide radicals and NO results in the production of peroxynitrite (ONOO⫺) (7), which is a potent mediator of human ␤-cell destruction (8). ROS directly injure cells by DNA strand breakage and peroxidation of proteins and lipids but also activate a number of signaling pathways involved in inflammation and apoptosis (9). Islets appear to be more vulnerable to oxidative injury than other tissues because they express low levels of antioxidant enzymes, including superoxide dismutase (SOD), which converts superoxide radicals to H2O2, and catalase and cellular glutathione peroxidase (Gpx-1), which detoxify H2O2 (10). Treatment of human islets with an SOD mimic promoted islet cell survival (11). The protective effect of overexpressing antioxidant enzymes has also been examined in vitro. SOD, catalase, and Gpx-1 provided rodent ␤-cell lines with various degrees of protection from exogenous oxidants and inflammatory cytokines (10,12–17). However, it is difficult to extrapolate from these relatively simple models to the complex process of ROS generation in a transplant setting, and in vivo results have been mixed (18,19). Overexpression of a combination of antioxidant enzymes in islets may provide the most effective protection from ROS-mediated injury (20). SOD is required to deal with superoxide radicals, but Gpx-1 or catalase may also 2109

SOD/Gpx EXPRESSION PROTECTS ISLET GRAFTS

be necessary to eliminate the resulting H2O2, which damages mitochondria in islets and blunts insulin secretion (21). The ultimate failure of manganese SOD– expressing islet grafts in a NOD scid mouse model may have been due to low expression of Gpx-1 and catalase (18). Furthermore, it is important to consider the cellular location of increased antioxidant activity because of the likely diversity of sources of ROS during islet transplantation. Here, we report the first demonstration that coexpression of enzymes to catabolize superoxide radicals and H2O2 protects islet grafts. We produced transgenic mice overexpressing the intracellular enzymes copper/zinc (Cu/Zn) SOD and Gpx-1 and the extracellular enzyme extracellular SOD. Because extracellular SOD is not tightly bound to the cell surface and can be lost during cellular activation, it was modified to incorporate a glycophosphatidyl inositol (GPI) tail. Gpx-1 was chosen for its broad cellular distribution and pleiotropic actions. Gpx-1 has a high affinity for H2O2 and catalyzes the reduction of organic and lipid hydroperoxides, which impair insulin secretion by islets (22). Gpx-1 overexpression protects against Fas-mediated apoptosis (23), increases the resistance of cultured ␤-cells to hypoxia/reoxygenation (24), and neutralizes peroxynitrite (25). Peroxynitrite is thought to be responsible for much of the oxidative damage previously ascribed to other radicals (26,27). We found that triple-transgenic islets showed the broadest resistance to oxidative stress in vitro and when grafted provided significantly improved early control of blood glucose compared with double-transgenic, single-transgenic, and wild-type grafts. RESEARCH DESIGN AND METHODS Preparation of transgenic constructs and design of screening primers. Constructs were prepared by cloning cDNAs for human antioxidant enzymes into a vector containing the mouse H-2Kb promoter (28) (Fig. 1A). The cDNAs for hemagglutinin-tagged Cu/Zn SOD and His-tagged Gpx-1 were as described previously (24) except that the latter lacked its endogenous intron. Construction of a modified extracellular SOD cDNA required several steps. First, the extracellular SOD coding region was amplified from human genomic DNA. The downstream primer incorporated a site for the restriction enzyme BspHI to allow the construction of a hybrid extracellular SOD/CD55 cDNA using a previously described strategy (29). The hybrid cDNA encoded an extracellular SOD protein in which the COOH terminus including the heparin-binding domain was replaced by the GPI anchor of human CD55. Finally, a splice-overlap PCR technique (30) was used to incorporate a FLAG epitope immediately downstream of the signal peptide. The following primers (Fig. 1A, arrowheads) were used for screening: F1, 5⬘-CTCAGAAGTCGAT CAAGCT-3⬘; F2, 5⬘-GCATGACTTCTGCGCTAAGATTGTC-3⬘; F3, 5⬘-TTGAAT TCGGCGTGGCCTAGCGAGTTATG-3⬘; R1, 5⬘-TTTTATGTTTCAGGTTCAGGG3⬘; R2, 5⬘-CCGAATTCAGGCATAATCTGGCAGGTCGTAAGGGTATTGGG-3⬘; R3, 5⬘-CATGGCGCAATTGTCCAAGAAGCCA-3⬘; and R4, 5⬘-TCATCGTCGTCCTTG TAGTCGGCGTCCGAGGCACCGGT-3⬘. Generation of transgenic mice. Constructs were microinjected separately at 10 ␮g/ml into CBA/C57BL6 F1 oocytes as described previously (31). An equimolar mixture at the same total concentration was microinjected to obtain triple-transgenic mice. Transgenic animals were identified by PCR of tail-tip genomic DNA prepared using the QIAamp system (Qiagen, Clifton Hill, Australia). The primers used and products obtained were as follows: Cu/Zn SOD: F3/R1, 915 bp; Gpx-1: F2/R3, 534 bp; and extracellular SOD: F2/R4, 298 bp. All procedures involving mice were performed with the approval of the St. Vincent’s Hospital Melbourne Animal Ethics Committee. Islet isolation and culture. Islets were isolated from pancreas by collagenase digestion and density centrifugation (32). Islets were maintained in RPMI-1640 with 10% FCS, 20 mmol/l HEPES, 1% L-glutamine, and 1% penicillin/ streptomycin, in 5% CO2 in air at 37°C unless otherwise indicated. Culture reagents were from Life Technologies (Grand Island, NY), collagenase type V and Histopaque from Sigma (St. Louis, MO), and deoxyribonuclease one from Worthington Biochemical (Lakewood, NJ). 2110

Detection of transgene mRNA in islets. Total RNA was extracted from hand-picked islets using TRIzol (Life Technologies). First-strand cDNA was synthesized from total RNA using SuperScript III (Life Technologies) and oligo-dT. PCR was performed using the primer sets F1/R2 for Cu/Zn SOD, F1/R3 for Gpx-1, and F1/R4 for extracellular SOD. The common forward primer F1 is specific for transcripts from the H-2Kb promoter (Fig. 1A), and the primer sets were designed to span the intron to distinguish mRNA-derived products (350, 358, and 122 bp, respectively) from genomic DNA-derived products (230 bp larger). As a control, murine GAPDH was amplified with primers 5⬘-GGTGAAGGTCGGTGTGAAC-3⬘ and 5⬘-TAGTAGAGGCGGGGAA GACG-3⬘. Flow cytometric analysis. Mouse splenocytes were cultured for 18 h in RPMI-1640 with 200 units/ml mouse interferon-␥ (IFN-␥; Chemicon, Temecula, CA), washed, incubated sequentially with anti-FLAG monoclonal antibody M2 (Sigma) and fluorescein-conjugated sheep anti-mouse IgG1 (Chemicon), and analyzed on a FACSCalibur flow cytometer (Becton Dickinson, Franklin Lakes, NJ). Measurement of antioxidant enzyme activity. Splenocytes cultured as above or hand-picked islets cultured for 48 h in CMRL medium-1066 with or without mouse IFN-␥ were homogenized using a PRO200 homogenizer (PRO Scientific, Oxford, CT) in 10 volumes of 50 mmol/l phosphate buffer, pH 7.4 (SOD assay), or 100 mmol/l phosphate buffer, 5 mmol/l EDTA, and 1 mmol/l dithiothreitol (Gpx assay). Homogenates for the SOD assay were incubated with phosphatidylinositol-specific phospholipase C (BIOMOL, Plymouth Meeting, PA) at 2.5 units/ml for 10 min at 30°C to release GPI-anchored extracellular SOD. All homogenates were then centrifuged at 8,000 ⫻ g for 20 min at 4°C, and supernatants were collected and assayed for protein content using the Bio-Rad Protein Assay (Bio-Rad, Hercules, CA). Total SOD activity in islet lysates was measured using the SOD assay kit WST (Dojindo Laboratories, Kumamoto, Japan). Gpx activity in splenocyte lysates was measured according to Paglia and Valentine (33). Glucose-stimulated insulin secretion assay. Overnight-cultured islets were washed three times with Krebs-Ringer bicarbonate buffer containing 5.5 mmol/l glucose and 0.1% BSA. Twenty islets were counted into glass tubes containing 200 ␮l of the same medium and preincubated for 30 min at 37°C with shaking. The medium was replaced with 200 ␮l of the same medium and incubated for 30 min. The medium was removed and replaced by 200 ␮l Krebs-Ringer bicarbonate buffer containing 22 mmol/l glucose and incubated for a further 30 min. Insulin content in the medium after the 30-min incubations in 5.5 and 22 mmol/l glucose was measured using a rat insulin radioimmunoassay kit (Linco Research, St. Charles, MO). In each experiment, at least 4 groups of 20 islets isolated from two mice were assayed. Exposure of islets to oxidative stress in vitro. One-hundred overnightcultured islets in CMRL medium-1066 were exposed to 10 mU/ml xanthine oxidase (ICN Biomedicals, Costa Mesa, CA) and 1 mmol/l hypoxanthine (Sigma) for 1 h at 37°C. Treated and control untreated islets were then cultured for 24 h in complete medium (CMRL-1066 containing 10% FCS and 1% penicillin/streptomycin) before determining viability. In addition, islets prepared as above were subjected to 18 h of hypoxia in CMRL medium-1066 in a Perspex hypoxia chamber flushed with 95% nitrogen/5% CO2, followed by 48 h of reoxygenation in complete medium at 37°C. In vitro viability assay. Islet cell viability was assessed using the 3-[4,5dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay. Islets were incubated with 0.5 mg/ml MTT in culture medium for 2 h at 37°C. Formazan produced by viable cells was dissolved in DMSO in a 96-well microtiter plate in a volume of 100 ␮l and measured by absorbance at 550 nm. Results were expressed as percent viability compared with parallel untreated samples of the same islets. There was no difference in MTT activity between untreated transgenic and wild-type islets (data not shown). Islet transplantation. Diabetes was induced in recipient CBA/C57BL6 F1 mice by a single intraperitoneal injection of streptozotocin in 0.9% NaCl at 0.3 mg/g body wt. Blood glucose was measured daily by tail vein prick using Accu-Chek Advantage II strips (Roche) and a Reflux II blood glucose monitor (Roche). The mice were considered diabetic if blood glucose was ⬎15 mmol/l for at least 2 consecutive days. Islets were not picked or cultured before transplantation. Islets from each donor group were pooled, quantified by calculating islet equivalents (IEQ) (34), and transplanted under the left renal capsule. Blood glucose was measured for 7 days posttransplantation. Grafts were removed by unilateral nephrectomy at day 8, and recipients were monitored for a further 2–3 days to confirm return to hyperglycemia. Ferric-xylenol orange assay. Protein hydroperoxide formed in islet grafts were measured using the ferric-xylenol orange (FOX) assay (35). Grafts were removed from beneath the kidney capsule, washed in 0.9% NaCl, and resuspended in 1 ml of 1:1 (vol:vol) 0.9% NaCl and CHCl3. The aqueous layer was removed after centrifugation and incubated in 2.8 mmol/l sulfuric acid, 75 ␮mol/l xylenol orange (Sigma), and 75 ␮mol/l ferrous ammonium sulfate for DIABETES, VOL. 54, JULY 2005

T.B. MYSORE AND ASSOCIATES

FIG. 1. A: Constructs used to generate transgenic mice. The cDNAs (below) were cloned separately into the EcoRI site of the vector (top). Arrowheads indicate primer binding sites. Gray shading indicates open reading frames, and black shading indicates the location of epitope tags. Elements are drawn to scale except the H-2Kb promoter. TCR, T-cell receptor. B: Detection of transgene mRNA by RT-PCR of islet total RNA. Primer sets are indicated at the top; the presence of individual transgenes is indicated below (ⴙ). Arrows indicate products of the expected size. Each set of reactions is separated by a 100-bp molecular weight ladder. C: Detection of cell surface expression of extracellular (EC) SOD-GPI by flow cytometric analysis. IFN-␥–treated wild-type and extracellular SOD-GPI transgenic splenocytes stained using an antibody to the FLAG epitope tag are indicated by the thin and thick lines, respectively. D: Total SOD and Gpx activity in islet and splenocyte lysates. 䊐, wild type; 2, extracellular SOD-GPI; f, Cu/Zn SOD; o, triple-transgenic line C; gray bars, Gpx-1. DIABETES, VOL. 54, JULY 2005

2111


FIG. 2. Glucose-stimulated insulin secretion by hand-picked islets from transgenic mice and wild-type (WT) littermate controls. Triple, triple-transgenic line C. Insulin secretion from 20 islets was measured after incubation for 30 min in 5.5 mmol/l (䊐) or 22 mmol/l (f) glucose. Results are expressed as means ⴞ SE from four experiments. There was no statistically significant difference between the WT islets and any of the transgenic islets. EC, extracellular. 30 min at 25°C. The FOX complex formed was measured by absorbance at 560 nm, and the concentration of protein hydroperoxide was calculated using the extinction coefficient of 48,000 (35). The concentration of protein was measured using the Bio-Rad Protein Assay (Bio-Rad). Statistical analysis. Statistical analyses were performed using SigmaStat statistical software (SPSS, Chicago, IL). Data were calculated as means ⫾ SE. Differences between means of in vitro treatment groups were analyzed by Student’s unpaired t test. Differences in islet graft function between groups in transplantation studies were assessed by Kruskal-Wallis one-way ANOVA on ranks. P ⬍ 0.05 was considered significant.

RESULTS

Transgenic mice. Five extracellular SOD-GPI transgenics were generated from 49 mice (10%), 10 Cu/Zn SOD transgenics from 75 mice (13%), and 7 Gpx-1 transgenics from 65 mice (11%). RT-PCR screening of RNA from peripheral blood leukocytes identified transgene expression in 5 of 5 extracellular SOD-GPI, 7 of 10 Cu/Zn SOD, and 5 of 7 Gpx-1 transgenic mice (data not shown). The highest expressing lines were maintained by breeding with CBA/ C57BL6 F1 mice and crossbred to generate double- and triple-transgenic mice. Ten triple-transgenic mice were also generated from 67 born (15%) by coinjection. Of these, three mice expressed all three transgenes (data not shown). The highest expressor was bred as above to establish a line, designated C (for coinjected), to distinguish it from triple-transgenic mice generated from singletransgenic lines, which were designated B (for bred). The transgenes were inherited en bloc. RT-PCR was used to demonstrate transgene expression at the mRNA level in islets (Fig. 1B). Triple-transgenic line C islets showed similar expression of transgene mRNA to islets from the individual transgenic lines. To confirm that extracellular SOD-GPI was expressed on the cell surface, splenocytes were isolated and treated with interferon-␥ (which upregulates the H-2Kb promoter used to drive transgene expression), stained using an antibody for the FLAG epitope tag, and analyzed by flow cytometry. Cell surface expression was evident on extracellular SOD-GPI transgenic splenocytes (Fig. 1C). Treatment with phosphatidylinositol-specific phospholipase C caused a reduction in FLAG expression (data not shown), confirming that the protein was attached by a GPI linkage. 2112

Total SOD activity was modestly increased in extracellular SOD-GPI, Cu/Zn SOD, and triple-transgenic islets compared with wild-type islets (Fig. 1D). A more significant increase (2.3-, 3.8-, and 4.6-fold, respectively) was observed after IFN-␥ treatment (Fig. 1D). Because it proved impractical to obtain sufficient protein from handpicked islets to measure Gpx activity, we assayed Gpx activity in splenocytes. In the IFN-␥ treatment group, Gpx-1 and triple-transgenic splenocytes showed 2.9- and 2.0-fold higher Gpx activity than wild-type splenocytes (Fig. 1D). Overexpression of antioxidant enzymes in islets does not affect glucose-stimulated insulin secretion. Islet function was assessed in vitro by measuring insulin secretion after incubation of islets in medium containing low (5.5 mmol/l) followed by high (22 mmol/l) glucose. Wildtype islets showed an 11-fold increase in secreted insulin in response to high glucose (Fig. 2). None of the transgenic islets showed a significantly different response to that of wild-type islets, indicating that transgene expression was not detrimental to this aspect of islet function. Gpx-1 enhances extracellular SOD-mediated protection of islets from extracellular superoxide radicals. Treatment with hypoxanthine/xanthine oxidase (HX/XO), which generates constant production of superoxide radicals close to the cell membrane (36), caused a 68% reduction in viability of wild-type islets (Fig. 3). Resistance to HX/XO toxicity was significantly increased by expression of extracellular SOD-GPI and further enhanced by coexpression of Gpx-1, despite the latter having no effect on its own (Fig. 3). Interestingly, Cu/Zn SOD islets were more sensitive than wild-type islets (P ⬍ 0.02), although this sensitivity was abolished by coexpression of Gpx-1 (Fig. 3). Maximum viability (88%) was observed in tripletransgenic islets, which were significantly more resistant than extracellular SOD-GPI/Gpx-1 islets (P ⫽ 0.005). Coexpression of Cu/Zn SOD, Gpx-1, and extracellular SOD-GPI protects islets from hypoxia/reoxygenation injury. The resistance of islets to hypoxia/reoxygenation injury was tested in vitro by incubation in nutrient-poor medium under hypoxic conditions for 18 h, followed by DIABETES, VOL. 54, JULY 2005


FIG. 3. Viability of transgenic and wild-type (WT) control islets after treatment with 1 mmol/l hypoxanthine and 10 mU/ml xanthine oxidase for 1 h. Triple, triple-transgenic line C. Results are expressed as means ⴞ SE from at least three experiments. *P < 0.05 vs. WT and **P < 0.01 vs. WT. EC, extracellular.

incubation in complete medium under normoxia for 48 h. This caused a 51% reduction in viability of wild-type islets (Fig. 4). Triple-transgenic islets demonstrated a small but significant (P ⬍ 0.01) improvement in viability compared with wild-type islets, whereas single- and double-transgenic islets showed no protection (Fig. 4). Coexpression of Cu/Zn SOD, Gpx-1, and extracellular SOD-GPI improves early function of marginal islet grafts. Islets from single transgenic, triple-transgenic (line C), and negative littermate controls were transplanted under the renal capsule of streptozotocin-induced diabetic recipients. Line C mice were used rather than the crossbred triple-transgenic line B because islets from line C and B mice showed equivalent protection from in vitro oxidative challenge (data not shown). To eliminate the possibility of variable rejection responses due to the mixed genetic background (CBA/C57BL6) of the donors, CBA/ C57BL6 F1 recipients were used. This allowed the analysis

of protection from primary nonfunction in the absence of rejection. Recipients were transplanted with a “marginal” islet mass of 1,500 IEQ. Preliminary experiments showed that 1,500 IEQ wild-type grafts conferred only partial glycemic control (blood glucose level [BGL], 10 –15 mmol/l) in the 1st week posttransplantation. Overexpression of Cu/Zn SOD or Gpx-1 did not improve marginal islet graft function compared with wild type (Fig. 5A). In contrast, grafts from triple-transgenic donors performed significantly better than wild type (Fig. 5B). Surprisingly, recipients of extracellular SOD-GPI marginal grafts fared worse than those of wild-type grafts, with mean BGL consistently higher than 20 mmol/l (Fig. 5B). Given the apparently detrimental action of extracellular SOD-GPI when expressed in isolation, its effect when combined with Gpx-1 and Cu/Zn SOD expression was determined by performing a second set of transplants

FIG. 4. Viability of transgenic and wild-type (WT) control islets after hypoxia in nutrientpoor medium for 18 h followed by reoxygenation in complete medium for 48 h. Triple, triple-transgenic line C. Results are expressed as means ⴞ SE from at least three experiments. **P < 0.01 vs. WT. EC, extracellular. DIABETES, VOL. 54, JULY 2005

2113


FIG. 6. Function of marginal islet grafts (1,000 IEQ) in diabetic recipients (mean BGL ⴞ SE). Œ, wild-type (WT) (n ⴝ 8); ‚, Cu/Zn SOD/Gpx-1 double-transgenic (n ⴝ 8); E, triple-transgenic line C (n ⴝ 8). Triple-transgenic grafts performed better than double-transgenic (P < 0.05) and WT (P < 0.001) grafts. There was no statistically significant difference between double-transgenic and WT grafts (P ⴝ 0.08).

FIG. 5. Function of marginal islet grafts (1,500 IEQ) in diabetic recipients, expressed as mean BGL ⴞ SE. A: Œ, wild type (WT) (n ⴝ 4); 〫, Cu/Zn SOD transgenic (n ⴝ 6); 䡺, Gpx-1 transgenic (n ⴝ 6). There was no statistically significant difference between these groups. B: Œ, WT (n ⴝ 4); E, triple-transgenic line C (n ⴝ 6); F, extracellular SOD-GPI transgenic (n ⴝ 6). Compared with WT grafts, triple-transgenic grafts performed better (P < 0.01) and extracellular SOD-GPI grafts performed worse (P < 0.001).

comparing Cu/Zn SOD/Gpx-1 grafts with triple-transgenic grafts. Graft size was reduced to 1,000 IEQ to increase the likelihood of detecting subtle differences in protection. Although the mean BGL in Cu/Zn SOD/Gpx-1 graft recipients was lower than that of wild-type graft recipients in the 3 days after transplantation, the difference over 7 days was not significant (P ⫽ 0.08) (Fig. 6). Triple-transgenic grafts performed significantly better than both wild-type (P ⬍ 0.001) and Cu/Zn SOD/Gpx-1 (P ⬍ 0.05) grafts (Fig. 6). This indicated that expression of extracellular SOD-GPI was beneficial in vivo but only in the context of coexpression of Gpx-1 and Cu/Zn SOD. In addition to the experiments shown, control grafts were performed from each donor group using an “excess” islet mass of 4,000 IEQ. Recipients receiving the control grafts had a mean BGL of between 3.9 and 4.5 mmol/l in the 7 days posttransplantation (data not shown), confirming that each islet preparation was capable of stably reversing hyperglycemia if sufficient islets were transplanted. This indicates that it is unlikely that the differences in graft performance were due to variation in the quality of the islet preparations. 2114

Coexpression of antioxidants protects islet grafts from oxidative injury. To determine whether the improved function of triple-transgenic islets was associated with a reduction in oxidative injury, total protein was extracted from two wild-type and two triple-transgenic 4,000 IEQ day 7 grafts and assayed for hydroperoxide content using the FOX assay. Wild-type grafts contained 2.4-fold more protein hydroperoxide (11.5 ⫾ 2.6 ␮mol/mg protein) than triple-transgenic grafts (4.9 ⫾ 0.5 ␮mol/mg protein). Analysis of lipid peroxidation by the FOX assay is complicated by the reaction to cyclic peroxides (37) and was not performed in this study. DISCUSSION

ROS have been implicated as important mediators of ␤-cell dysfunction in islet transplantation. Superoxide radicals and NO, produced in the extracellular compartment by activated recipient and intragraft macrophages, can in turn generate H2O2, hydroxyl radical, and peroxynitrite. Each of these molecules can damage the ␤-cell membrane and (with the exception of superoxide) readily cross into the cytosol and mitochondria to cause further damage to key enzymes, lipids, and DNA. This situation is worsened by intracellular generation of superoxide radicals by xanthine oxidase and NO by inducible NO synthase. Thus, the transplanted ␤-cell requires protection from a cocktail of ROS within and without. The complexity of ROS metabolism suggests that a multi-antioxidant enzyme strategy may be most effective (12,13,20). In this study, we used transgenic mice to demonstrate that this is the case in a transplant setting. We first used in vitro assays to show that the level of expression of individual enzymes was functionally significant. Extracellular SOD-GPI protected islets from the action of extracellular superoxide radicals generated by HX/XO, but Cu/Zn SOD and Gpx-1 did not. Coexpression of Gpx-1 with extracellular SOD-GPI further increased resistance, suggesting that although the primary cause of cell death by HX/XO is extracellular injury by superoxide DIABETES, VOL. 54, JULY 2005


radicals, secondary damage can result from entry of its by-products (including H2O2) into the cell. The highest level of protection from HX/XO was observed for tripletransgenic islets. Expression of all three transgenes was necessary to increase the resistance of islets to hypoxia/ reoxygenation injury. However, the degree of protection from hypoxia was modest, suggesting that the activity of one or more of the enzymes may be insufficient to fully cope with the volume of ROS generated. It is conceivable that substitution of catalase for Gpx-1 would increase protection in this in vitro model, because although catalase has a lower affinity for H2O2 it has a higher capacity for inactivation (12). Although the in vitro models provided an indication of additive or synergistic protection, they are unlikely to accurately reflect the complex in vivo transplantation setting. We therefore used a marginal islet graft model to determine whether coexpression was also protective in vivo. Although others have reported that 500 hand-picked mouse islets were sufficient to cure diabetes and grafts of 250 –300 islets were suboptimal (38,39), we found that wild-type grafts of 1,500 IEQ were marginal. This apparent discrepancy is probably due to differences in islet graft quantitation rather than the quality of our islet preparations, because in our hands, grafts of 400 hand-picked islets are sufficient to reverse hyperglycemia (data not shown). In the current study, we did not hand-pick islets before transplantation and accordingly measured IEQ, which is the standard procedure for human islet allografts. The marginal grafts of 1,500 IEQ corresponded to approximately one donor per recipient, which is also suboptimal in human islet transplantation. Triple-transgenic islet grafts, but not Cu/Zn SOD/Gpx-1 or single-transgenic grafts, improved control of blood glucose compared with wild-type grafts. This suggests that extracellular superoxide radicals play a significant role in graft injury even when intracellular antioxidant activity is elevated. However, expression of extracellular SOD-GPI alone significantly worsened graft function, indicating that intracellular ROS also contribute to ␤-cell loss. Peroxynitrite is an obvious candidate because it can be produced from superoxide radicals and nitric oxide even in the presence of SOD (26). Furthermore, the activity of peroxynitrite is enhanced by a positive feedback loop: it inhibits mitochondrial enzymes and increases generation of H2O2, which in turn prolongs the half-life of peroxynitrite (40). The detrimental effect of extracellular SOD-GPI in isolation may be due to accumulation of H2O2 and its entry into cells, compounded by production of intracellular superoxide radicals and nitric oxide. The opposite effects of extracellular SOD-GPI in vitro (HX/XO assay) and in vivo highlight the difficulty of accurately modeling the injury associated with islet transplantation. Even the hypoxia model, which may approximate some of the conditions of transplantation, does not allow for the contribution of surrounding recipient tissue and immune cells, and the kinetics of ROS generation may be quite different. We would argue that in vitro assays should be used only to demonstrate functional antioxidant gene expression and that graft function is the only reliable measure of efficacy. To our knowledge, this is the first study of islets from DIABETES, VOL. 54, JULY 2005

mice coexpressing three transgenes, without apparent detrimental effects on islet function. Whether this strategy can be used in human islet allotransplantation is not clear. Safe and efficient ex vivo gene delivery into human islets has not yet been demonstrated, and the requirement for expression of multiple genes presents an additional challenge. It may be preferable to express a single enzyme capable of scavenging a range of ROS, such as metallothionein (41). The use of different models precludes comparison of the relative efficacy of metallothionein and the enzymes used here. The potential value of antioxidant expression may lie in islet xenotransplantation. The use of nonhuman donor animals such as pigs circumvents the problems of ex vivo gene delivery because it allows direct genetic modification of the donor. We have shown that it is possible to coexpress at least three transgenes in pigs, with the genes coinherited as a single unit (28). The use of transgenic donors offers the additional advantage over postisolation gene transfer that protective genes will be expressed in islets during the damaging isolation procedure. Expression of antioxidant enzymes, perhaps in combination with other graft-protective molecules (38,42– 44), is thus a promising approach to reducing primary nonfunction and rejection of islet allografts and xenografts alike. ACKNOWLEDGMENTS

This work was supported by grants from the Juvenile Diabetes Research Foundation and the Roche Organ Transplantation Research Foundation. We thank Helen Barlow and the staff of the Immunology Research Centre mouse facility for animal maintenance. REFERENCES 1. Shapiro AM, Lakey JR, Ryan EA, Korbutt GS, Toth E, Warnock GL, Kneteman NM, Rajotte RV: Islet transplantation in seven patients with type 1 diabetes mellitus using a glucocorticoid-free immunosuppressive regimen. N Engl J Med 343:230 –238, 2000 2. Biarnes M, Montolio M, Nacher V, Raurell M, Soler J, Montanya E: ␤-Cell death and mass in syngeneically transplanted islets exposed to short- and long-term hyperglycemia. Diabetes 51:66 –72, 2002 3. Parr EL, Bowen KM, Lafferty KJ: Cellular changes in cultured mouse thyroid glands and islets of Langerhans. Transplantation 30:135–141, 1980 4. Griffith RC, Scharp DW, Hartman BK, Ballinger WF, Lacy PE: A morphologic study of intrahepatic portal-vein islet isografts. Diabetes 26:201–214, 1977 5. Li C, Jackson RM: Reactive species mechanisms of cellular hypoxiareoxygenation injury. Am J Physiol Cell Physiol 282:C227–C241, 2002 6. Lepore DA: Nitric oxide synthase-independent generation of nitric oxide in muscle ischemia-reperfusion injury. Nitric Oxide 4:541–545, 2000 7. Beckman JS, Beckman TW, Chen J, Marshall PA, Freeman BA: Apparent hydroxyl radical production by peroxynitrite: implications for endothelial injury from nitric oxide and superoxide. Proc Natl Acad Sci U S A 87:1620 –1624, 1990 8. Lakey JR, Suarez-Pinzon WL, Strynadka K, Korbutt GS, Rajotte RV, Mabley JG, Szabo C, Rabinovitch A: Peroxynitrite is a mediator of cytokineinduced destruction of human pancreatic islet beta cells. Lab Invest 81:1683–1692, 2001 9. Evans JL, Goldfine ID, Maddux BA, Grodsky GM: Are oxidative stressactivated signaling pathways mediators of insulin resistance and ␤-cell dysfunction? Diabetes 52:1– 8, 2003 10. Tiedge M, Lortz S, Drinkgern J, Lenzen S: Relation between antioxidant enzyme gene expression and antioxidative defense status of insulinproducing cells. Diabetes 46:1733–1742, 1997 11. Bottino R, Balamurugan AN, Bertera S, Pietropaolo M, Trucco M, Piganelli JD: Preservation of human islet cell functional mass by anti-oxidative action of a novel SOD mimic compound. Diabetes 51:2561–2567, 2002 12. Tiedge M, Lortz S, Munday R, Lenzen S: Complementary action of 2115


antioxidant enzymes in the protection of bioengineered insulin-producing RINm5F cells against the toxicity of reactive oxygen species. Diabetes 47:1578 –1585, 1998 13. Tiedge M, Lortz S, Munday R, Lenzen S: Protection against the co-operative toxicity of nitric oxide and oxygen free radicals by overexpression of antioxidant enzymes in bioengineered insulin-producing RINm5F cells. Diabetologia 42:849 – 855, 1999 14. Azevedo-Martins AK, Lortz S, Lenzen S, Curi R, Eizirik DL, Tiedge M: Improvement of the mitochondrial antioxidant defense status prevents cytokine-induced nuclear factor-␬B activation in insulin-producing cells. Diabetes 52:93–101, 2003 15. Karsten V, Sigrist S, Moriscot C, Sorg T, Lemarchand P, Belcourt A, Benhamou PY, Pinget M, Kessler L: How can adenovirus-mediated catalase and superoxide dismutase gene transfer improve the outcome of pancreatic cells for transplantation? Transplant Proc 33:575–576, 2001 16. Lortz S, Tiedge M, Nachtwey T, Karlsen AE, Nerup J, Lenzen S: Protection of insulin-producing RINm5F cells against cytokine-mediated toxicity through overexpression of antioxidant enzymes. Diabetes 49:1123–1130, 2000 17. Moriscot C, Pattou F, Kerr-Conte J, Richard MJ, Lemarchand P, Benhamou PY: Contribution of adenoviral-mediated superoxide dismutase gene transfer to the reduction in nitric oxide-induced cytotoxicity on human islets and INS-1 insulin-secreting cells. Diabetologia 43:625– 631, 2000 18. Bertera S, Crawford ML, Alexander AM, Papworth GD, Watkins SC, Robbins PD, Trucco M: Gene transfer of manganese superoxide dismutase extends islet graft function in a mouse model of autoimmune diabetes. Diabetes 52:387–393, 2003 19. Xu B, Moritz JT, Epstein PN: Overexpression of catalase provides partial protection to transgenic mouse beta cells. Free Radic Biol Med 27:830 – 837, 1999 20. Lortz S, Tiedge M: Sequential inactivation of reactive oxygen species by combined overexpression of SOD isoforms and catalase in insulin-producing cells. Free Radic Biol Med 34:683– 688, 2003 21. Maechler P, Jornot L, Wollheim CB: Hydrogen peroxide alters mitochondrial activation and insulin secretion in pancreatic beta cells. J Biol Chem 274:27905–27913, 1999 22. Miwa I, Ichimura N, Sugiura M, Hamada Y, Taniguchi S: Inhibition of glucose-induced insulin secretion by 4-hydroxy-2-nonenal and other lipid peroxidation products. Endocrinology 141:2767–2772, 2000 23. Gouaze V, Andrieu-Abadie N, Cuvillier O, Malagarie-Cazenave S, Frisach MF, Mirault ME, Levade T: Glutathione peroxidase-1 protects from CD95induced apoptosis. J Biol Chem 277:42867– 42874, 2002 24. Lepore DA, Shinkel TA, Fisicaro N, Mysore TB, Johnson LE, d’Apice AJ, Cowan PJ: Enhanced expression of glutathione peroxidase protects islet beta cells from hypoxia-reoxygenation. Xenotransplantation 11:53–59, 2004 25. Sies H, Sharov VS, Klotz LO, Briviba K: Glutathione peroxidase protects against peroxynitrite-mediated oxidations: a new function for selenoproteins as peroxynitrite reductase. J Biol Chem 272:27812–27817, 1997 26. Beckman JS: Oxidative damage and tyrosine nitration from peroxynitrite. Chem Res Toxicol 9:836 – 844, 1996 27. Murphy MP, Packer MA, Scarlett JL, Martin SW: Peroxynitrite: a biologically significant oxidant. Gen Pharmacol 31:179 –186, 1998 28. Cowan PJ, Aminian A, Barlow H, Brown AA, Chen CG, Fisicaro N, Francis DM, Goodman DJ, Han W, Kurek M, Nottle MB, Pearse MJ, Salvaris E, Shinkel TA, Stainsby GV, Stewart AB, d’Apice AJ: Renal xenografts from triple-transgenic pigs are not hyperacutely rejected but cause coagulopa-

2116

thy in non-immunosuppressed baboons. Transplantation 69:2504 –2515, 2000 29. Shinkel TA, Cowan PJ, Barlow H, Aminian A, Romanella M, Lublin DM, Pearse MJ, d’Apice AJ: Expression and functional analysis of glycosylphosphatidyl inositol-linked CD46 in transgenic mice. Transplantation 66:1401–1406, 1998 30. Zaidi SH, Hui CC, Cheah AY, You XM, Husain M, Rabinovitch M: Targeted overexpression of elafin protects mice against cardiac dysfunction and mortality following viral myocarditis. J Clin Invest 103:1211–1219, 1999 31. Cowan PJ, Shinkel TA, Aminian A, Romanella M, Wigley PL, Lonie AJ, Nottle MB, Pearse MJ, d’Apice AJ: High-level co-expression of complement regulators on vascular endothelium in transgenic mice: CD55 and CD59 provide greater protection from human complement-mediated injury than CD59 alone. Xenotransplantation 5:184 –190, 1998 32. Gotoh M, Maki T, Kiyoizumi T, Satomi S, Monaco AP: An improved method for isolation of mouse pancreatic islets. Transplantation 40:437– 438, 1985 33. Paglia DE, Valentine WN: Studies on the quantitative and qualitative characterization of erythrocyte glutathione peroxidase. J Lab Clin Med 70:158 –169, 1967 34. Bretzel RG, Alejandro R, Hering BJ, van Suylichem PT, Ricordi C: Clinical islet transplantation: guidelines for islet quality control. Transplant Proc 26:388 –392, 1994 35. Gay C, Collins J, Gebicki JM: Determination of iron in solutions with the ferric-xylenol orange complex. Anal Biochem 273:143–148, 1999 36. Amstad P, Moret R, Cerutti P: Glutathione peroxidase compensates for the hypersensitivity of Cu,Zn-superoxide dismutase overproducers to oxidant stress. J Biol Chem 269:1606 –1609, 1994 37. Yin H, Porter NA: Specificity of the ferrous oxidation of xylenol orange assay: analysis of autoxidation products of cholesteryl arachidonate. Anal Biochem 313:319 –326, 2003 38. Grey ST, Longo C, Shukri T, Patel VI, Csizmadia E, Daniel S, Arvelo MB, Tchipashvili V, Ferran C: Genetic engineering of a suboptimal islet graft with A20 preserves beta cell mass and function. J Immunol 170:6250 – 6256, 2003 39. Garcia-Ocana A, Takane KK, Reddy VT, Lopez-Talavera JC, Vasavada RC, Stewart AF: Adenovirus-mediated hepatocyte growth factor expression in mouse islets improves pancreatic islet transplant performance and reduces beta cell death. J Biol Chem 278:343–351, 2003 40. Cuzzocrea S, Riley DP, Caputi AP, Salvemini D: Antioxidant therapy: a new pharmacological approach in shock, inflammation, and ischemia/reperfusion injury. Pharmacol Rev 53:135–159, 2001 41. Li X, Chen H, Epstein PN: Metallothionein protects islets from hypoxia and extends islet graft survival by scavenging most kinds of reactive oxygen species. J Biol Chem 279:765–771, 2004 42. Wu Y, Han B, Luo H, Roduit R, Salcedo TW, Moore PA, Zhang J, Wu J: DcR3/TR6 effectively prevents islet primary nonfunction after transplantation. Diabetes 52:2279 –2286, 2003 43. Contreras JL, Smyth CA, Bilbao G, Young CJ, Thompson JA, Eckhoff DE: 17␤b-Estradiol protects isolated human pancreatic islets against proinflammatory cytokine-induced cell death: molecular mechanisms and islet functionality. Transplantation 74:1252–1259, 2002 44. Lopez-Talavera JC, Garcia-Ocana A, Sipula I, Takane KK, Cozar-Castellano I, Stewart AF: Hepatocyte growth factor gene therapy for pancreatic islets in diabetes: reducing the minimal islet transplant mass required in a glucocorticoid-free rat model of allogeneic portal vein islet transplantation. Endocrinology 145:467– 474, 2004

DIABETES, VOL. 54, JULY 2005