SLC30A8 polymorphism and type 2 diabetes risk: Evidence from 27 study groups
Y.L. Jing,Q.M. Sun,Y. Bi,S.M. Shen,D.L. Zhu
DOI: https://doi.org/10.1016/j.numecd.2009.11.004
2011-01-01
Abstract:Methods and results We searched PubMed through October 2009 to identify all relevant papers. Odds ratios (ORs) and 95% confidence intervals (CIs) were extracted under an additive genetic model. In the current meta-analysis, we identified a total of 27 groups including 42,609 cases and 69,564 controls. In analyses of the case–control studies by ethnicity, the results indicated that SLC30A8 polymorphism was related to elevate risks of type 2 diabetes both in Europeans (OR = 1.15, 95% CI 1.11–1.18, P < 0.001) and Asians (OR = 1.15, 95% CI 1.11–1.19, P < 0.001). Next, we separated hospital-based case–control studies from population-based case–control studies, however, there was no apparent difference between population-based case–control study groups (OR = 1.15, 95% CI 1.12–1.17, P < 0.001) and hospital-based case–control study groups (OR = 1.16, 95% CI 1.07–1.25, P < 0.001). Conclusion Our present meta-analysis provided evidence that SLC30A8 (rs13266634) C allele carriers could elevate the risk of type 2 diabetes, especially in Europeans and Asians. Keywords Type 2 diabetes SLC30A8 (rs13266634) polymorphism Meta-analysis Introduction Type 2 diabetes is a complex disease characterized by insulin resistance and pancreatic β-cell dysfunction, and the prevalence of type 2 diabetes is increasing rapidly worldwide. Many environmental factors combined with multiple genetic variants are involved in the onset and development of type 2 diabetes. To date, intense research has been performed to identify the genetic risk factors in type 2 diabetes, and 18 genetic loci have been reported to be associated with the risk of type 2 diabetes [1,2] . Zinc transporter protein member 8 ( ZnT-8 ), which is a member of the zinc transporter family, is only expressed in pancreatic beta cells and provides zinc to insulin maturation and/or storage processes in insulin-secreting pancreatic beta cells. ZnT-8 ( SLC30A8 ) is a 369-amino acid protein coded by the gene SLC30A8 (chromosome 8q24.11). The non-synonymous single-nucleotide polymorphism (SNP) in SLC30A8 (rs13266634) causes an arginine to tryptophan change (Arg325Trp) in the last exon of the solute carrier family 30 [3,4] . There is accruing evidence that SLC30A8 polymorphism is associated with β-cell dysfunction. Kirchhoff et al. indicated that the risk allele of rs13266634 in SLC30A8 plays an important role in impaired proinsulin conversion and processing [5] . In addition, the polymorphism of SLC30A8 was also associated with the secretion of insulin. In the study of 846 non-diabetic but diabetes-prone European offspring of type 2 diabetic patients, they found that the function and/or production of ZnT-8 were reduced when carrying the risk C allele of SLC30A8 . The risk C allele might also modulate the function of neighboring cells via paracrine/autocrine interactions, and variation in ZnT-8 is a cause of impaired pancreatic beta cell function. The SLC30A8 loci were also related to altered glucose-stimulated insulin secretion [6,7] . A recent genome-wide association study of the French population by Sladek et al. identified four novel risk loci for type 2 diabetes, and rs13266634 in the SLC30A8 gene locus has been found to be associated with type 2 diabetes [8] . Subsequently, replication studies on the association between SLC30A8 (rs13266634) polymorphism and type 2 diabetes susceptibility have been performed in various ethnic populations, but the results were inconclusive, which might be due to the relatively small sample size in the included studies. Therefore, it is critical to summarize the available evidence from case–control studies and prospective studies to further estimate the association between SLC30A8 polymorphism and susceptibility to type 2 diabetes. In the present study, we conducted a meta-analysis on SLC30A8 (rs13266634) polymorphism and type 2 diabetes susceptibility based on various ethnic groups from Europe, Asia and Africa [8–29] . Methods Search strategy We searched PubMed through October 2009 to identify all relevant papers, using the key words Zinc transporter protein member 8, ZnT-8 , SLC30A8 , SLC30A8 gene polymorphism or SLC30A8 variant in combination with diabetes or type 2 diabetes or T2DM. The titles and abstracts were scanned to exclude studies that were clearly irrelevant to the current topic. The full text of the remaining articles was read to determine whether they contained information of interest. Additionally, the reference lists of the retrieved articles were examined to identify other relevant publications. We performed a systematic literature search for all relevant studies in the English language. All searches were conducted independently by two investigators. Study selection To be eligible, studies had to fulfill the following criteria: 1) had a case–control study design or prospective study design; 2) reported the number of subjects with each allele or genotype in cases and controls; 3) reported the risk estimates [odds ratio (OR)] and 95% confidence intervals (CI) or provided raw data that allowed us to calculate them. Studies were excluded if there were no available data reported. When the same data were included in more than one publication, the complete or most recent study was included. We also conducted a meta-regression to identify characteristics (body mass index (BMI), age) that significantly contributed to heterogeneity. Data extraction The following data were extracted independently by two authors following the guidelines of the Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group: first author's name, publication year, ethnicity of the study population, study design (population-based controls or hospital-based controls for case–control studies and prospective studies), numbers of cases and controls or prospective cohort size, the number of cases and controls with each of the three genotypes, allele frequencies and the most completely adjusted estimate (OR and 95% CI) [30] . Discrepancies were reconciled through discussion and resolved by a third investigator until consensus was achieved on every item. Statistical analysis ORs with 95% CIs were used to assess the strength of the association between SLC30A8 polymorphism and the risk of type 2 diabetes in both case–control studies and prospective studies. Study-specific ORs and corresponding 95% CIs were extracted from every included study and then log OR values were weighted by the inverse of their variances to obtain a pooled OR and its 95% CI. The Z -test was used to determine the statistical significance of the pooled OR, and P < 0.05 was considered as statistically significant. We evaluated the association between C/T allele and type 2 diabetes risk under an additive genetic model. Overall OR and 95% CI were directly extracted from 17 studies [9–15,17,18,20,21,23–27,29] . In order to improve comparability with other studies, we calculated the total OR and 95% CI with the information provided under the additive genetic model in the remaining five studies [8,16,19,22,28] . We performed analysis on ethnicity (Europeans, Asians and Africans) separately. Stratified analysis was conducted on the source of controls (hospital-based or population-based) in case–control studies. The Q test ( P < 0.10 was considered representative of statistically significant heterogeneity) and I 2 statistic were used to evaluate possible heterogeneity of study results [31] . If significant heterogeneity existed, the pooled OR estimate of each study was calculated by the random-effects model, otherwise the fixed-effects model was used [32] . We also performed sensitivity analyses to assess the stability of the results. The Funnel plot and Egger regression test were used to assess potential publication bias, and P < 0.10 was considered as a statistically significant publication bias [33] . All of the statistical analyses were conducted with STATA 9.2 software (STATA Corporation, College Station, TX, USA). Results We identified 44 potentially relevant studies after the primary literature search. Nineteen studies were excluded after screening abstracts and full texts, because they were either review articles, or irrelevant to our current study. Three studies [34–36] were excluded because they focused on SLC30A8 polymorphism and type 1 diabetes risk. One study was not included because it defined hyperglycemia (HG) as IFG or T2D. [37] We also performed a stepwise process sensitivity analysis, and the result indicated that the second group in the study by Cauchi et al. [19] was the main source of the heterogeneity in the European groups, so it was excluded in the meta-analysis. Therefore, there were a total of 24 groups from 20 case–control studies [8–27] and three groups from two prospective studies [28,29] including 42,609 cases and 69,564 controls in the current meta-analysis on SLC30A8 polymorphism in relation to susceptibility to type 2 diabetes risk ( Table 1 ). There were duplicated numbers of cases and/or controls in five studies [8,13,17,19,22] , so the large numbers were extracted when summarizing the total number of cases and controls. Among the case–control studies, three groups were hospital-based case–control samples and 21 were population-based case–control samples. In the case–control studies, 12 groups were Asian descendents, 10 were European descendents and the remaining two were African descendents. Subjects in the two prospective studies were European descendents. Age, sex and BMI were considered as the potential confounders in most of the individual studies. In several studies, BMI and age have been demonstrated to be involved in type 2 diabetes and its complications [38–41] , so we attempted to take these covariates into account by meta-regression; however, we failed to find any significant results. The genotypic distributions in the controls of all included studies were in agreement with Hardy–Weinberg equilibrium except for two studies by Ng et al. and Horikawa et al., but the summarized estimates were not significantly changed with or without the inclusion of these two studies [17,20] . In the analyses of the case–control studies by ethnicity (European, Asian and African), the results indicated that SLC30A8 polymorphism was related to elevated risks of type 2 diabetes both in European groups (OR = 1.15, 95% CI, 1.11–1.18, Z = 8.31, P < 0.001) and Asian groups (OR = 1.15, 95% CI, 1.11–1.19, Z = 7.54, P < 0.001) ( Fig. 1 ), however, the results from the African groups showed no statistically significant association between SLC30A8 polymorphism and type 2 diabetes risk (OR = 1.09, 95% CI, 0.89–1.33, Z = 0.83, P = 0.404) ( Table 2 ). When we separated hospital-based case–control studies from population-based case–control studies, there was a statistically significant association in the population-based case–control study groups (OR = 1.15, 95% CI, 1.12–1.17, Z = 11.42, P < 0.001) ( Fig. 2 ), and a similar result was observed in the hospital-based case–control study groups (OR = 1.16, 95% CI, 1.07–1.25, Z = 3.69, P < 0.001) ( Fig. 2 ). Of all the included studies, no evidence of obvious asymmetry was observed which indicated that no publication bias was revealed by Egger's test ( t = −0.65, P = 0.519). Discussion The current meta-analysis including 24 case–control studies and three prospective studies was conducted to explore the association between SLC30A8 polymorphism and susceptibility to type 2 diabetes. The results indicated that SLC30A8 polymorphism carried an elevated risk of type 2 diabetes in both hospital-based and population-based case–control studies, which indicated that individuals with the variant genotype had a higher risk for the development of type 2 diabetes than participants with the wild genotype. The pooled OR on the analysis by ethnicity indicated that the elevated risk existed both in Europeans and Asians, but not in Africans. Since rs13266634 in the SLC30A8 gene locus was identified as one of the risk loci for type 2 diabetes, the function of ZnT-8 and the SNP of this gene have been widely studied in the etiology of type 2 diabetes. In two cohort studies from the Insulin Resistance Atherosclerosis Family Study, the non-synonymous variant at the SLC30A8 locus was associated with variation in disposition index (an index of beta-cell function) in an African-American cohort and more modestly in a Hispanic-American cohort [42] . Recently, a three-stage genome-wide association study was conducted to identify novel type 2 diabetes gene variants and confirm previously identified variants in the Japanese population; the authors confirmed the association between the SLC30A8 variant and type 2 diabetes risk [43] . Evidence from animal studies confirmed that ZnT-8 modulated glucose homeostasis in mice, and it was required for normal insulin crystallization and insulin release mainly in vivo . Additionally, down-regulation of ZnT-8 expression was associated with a marked decrease in intracellular insulin content and a reduction in the secretion of insulin in cultured rat beta-cell line cells. Moreover, SLC30A8 expression-level polymorphism could affect insulin secretion and the glycemic response in vivo by reducing the endogenous expression of the ZnT-8 transporter [44,45] . The results in our present study should be interpreted with caution when heterogeneity is present. Obvious heterogeneity was observed in European groups, however, after exclusion of one study group that caused the heterogeneity, no evidence of heterogeneity was observed ( P = 0.300). This implies that different genetic backgrounds and various environmental factors were present in different ethnicities, and a small study sample was also a source of heterogeneity. Some limitations of this meta-analysis must be considered when interpreting the results. First, our search was limited to English-language studies and papers published in indexed journals, and we were not able to obtain the original data. Second, all studies which we included were observational studies, which presented particular challenges because of different study designs and inherent biases, however, no publication bias was observed in our study, suggesting that the overall result is likely to be unbiased. Third, study samples were relatively small when investigating the relationship between the two factors in Africans, so a more precise association needs to be explored further with sufficient data. Fourth, we were not able to conduct subgroup analyses on obesity status because of insufficient data and different definitions of obesity status among different ethnicities; only one study separated obese from non-obese participants [22] , so whether obesity status affects the genetic susceptibility to type 2 diabetes needs further research. Fifth, there was statistically significant heterogeneity among the European group, which may be attributed to the multiple ethnic groups in the world, the complicated interactions between gene–gene and gene–environment, the limited study numbers and relatively small sample sizes. Variable definition of cases and controls might also be a source of heterogeneity. In addition, the numbers of prospective studies were too limited to perform analysis on ethnicity and the source of controls so as to certify the results of the case–control studies. Sixth, not all of the included studies have adjusted for potential confounders, such as age, sex, BMI (body mass index), family history, environmental factors and lifestyle, all of which could influence the relationship between SLC30A8 polymorphism and the risk of type 2 diabetes. In conclusion, our data indicated an elevated risk of type 2 diabetes in carriers of an SLC30A8 (rs13266634) C allele, especially in Europeans and Asians. However, further large sample prospective studies on different ethnic groups are warranted to confirm our results. In addition, comprehensive interactions on gene–gene and gene–environment should also be evaluated in future analyses. Acknowledgements This work was supported by grants from the National Natural Science Foundation of China (No. 30671004 and No. 30800539), the development program of Nanjing Health Bureau (No. ZKX05022) and the Natural Science Foundation of Jiangsu Province (No. BK2008061). Conflict of interest None disclosed. References [1] M.A. Permutt J. Wasson N. Cox Genetic epidemiology of diabetes J Clin Invest 115 2005 1431 1439 [2] J.R. Perry T.M. Frayling New gene variants alter type 2 diabetes risk predominantly through reduced beta-cell function Curr Opin Clin Nutr Metab Care 11 2008 371 377 [3] T.W. Boesgaard J. Zilinskaite M. Vänttinen M. Laakso P.A. Jansson A. Hammarstedt The common SLC30A8 Arg325Trp variant is associated with reduced first-phase insulin release in 846 non-diabetic offspring of type 2 diabetes patients – the EUGENE2 study Diabetologia 51 2008 816 820 [4] H. Staiger F. Machicao N. Stefan O. Tschritter C. Thamer K. Kantartzis Polymorphisms within novel risk loci for type 2 diabetes determine beta-cell function PLoS One 2 2007 e832 [5] K. Kirchhoff F. Machicao A. Haupt S.A. Schäfer O. Tschritter H. Staiger Polymorphisms in the TCF7L2, CDKAL1 and SLC30A8 genes are associated with impaired proinsulin conversion Diabetologia 51 2008 597 601 [6] E. Mocchegiani R. Giacconi M. Malavolta Zinc signalling and subcellular distribution: emerging targets in type 2 diabetes Trends Mol Med 14 2008 419 428 [7] F. Chimienti A. Favier M. Seve ZnT-8, a pancreatic beta-cell-specific zinc transporter Biometals 18 2005 313 317 [8] R. Sladek G. Rocheleau J. Rung C. Dina L. Shen D. Serre A genome-wide association study identifies novel risk loci for type 2 diabetes Nature 445 2007 881 885 [9] L.J. Scott K.L. Mohlke L.L. Bonnycastle C.J. Willer Y. Li W.L. Duren A genome-wide association study of type 2 diabetes in Finns detects multiple susceptibility variants Science 316 2007 1341 1345 [10] R. Saxena B.F. Voight V. Lyssenko N.P. Burtt P.I. de Bakker H. Chen Genome-wide association analysis identifies loci for type 2 diabetes and triglyceride levels Science 316 2007 1331 1336 [11] E. Zeggini M.N. Weedon C.M. Lindgren T.M. Frayling K.S. Elliott H. Lango Replication of genome-wide association signals in UK samples reveals risk loci for type 2 diabetes Science 316 2007 1336 1341 [12] M. Horikoshi K. Hara C. Ito N. Shojima R. Nagai K. Ueki Variations in the HHEX gene are associated with increased risk of type 2 diabetes in the Japanese population Diabetologia 50 2007 2461 2466 [13] V. Steinthorsdottir G. Thorleifsson I. Reynisdottir R. Benediktsson T. Jonsdottir G.B. Walters A variant in CDKAL1 influences insulin response and risk of type 2 diabetes Nat Genet 39 2007 770 775 [14] J.K. Hertel S. Johansson H. Raeder K. Midthjell V. Lyssenko L. Groop Genetic analysis of recently identified type 2 diabetes loci in 1638 unselected patients with type 2 diabetes and 1858 control participants from a Norwegian population-based cohort (the HUNT study) Diabetologia 51 2008 971 977 [15] Y.H. Lee E.S. Kang S.H. Kim S.J. Han C.H. Kim H.J. Kim Association between polymorphisms in SLC30A8, HHEX, CDKN2A/B, IGF2BP2, FTO, WFS1, CDKAL1, KCNQ1 and type 2 diabetes in the Korean population J Hum Genet 53 2008 991 998 [16] D.K. Sanghera L. Ortega S. Han J. Singh S.K. Ralhan G.S. Wander Impact of nine common type 2 diabetes risk polymorphisms in Asian Indian Sikhs: PPARG2 (Pro12Ala), IGF2BP2, TCF7L2 and FTO variants confer a significant risk BMC Med Genet 9 2008 59 [17] M.C. Ng K.S. Park B. Oh C.H. Tam Y.M. Cho H.D. Shin Implication of genetic variants near TCF7L2, SLC30A8, HHEX, CDKAL1, CDKN2A/B, IGF2BP2, and FTO in type 2 diabetes and obesity in 6719 Asians Diabetes 57 2008 2226 2233 [18] Y. Furukawa T. Shimada H. Furuta S. Matsuno A. Kusuyama A. Doi Polymorphisms in the IDE-KIF11-HHEX gene locus are reproducibly associated with type 2 diabetes in a Japanese population J Clin Endocrinol Metab 93 2008 310 314 [19] S. Cauchi D. Meyre E. Durand C. Proença M. Marre S. Hadjadj Post genome-wide association studies of novel genes associated with type 2 diabetes show gene–gene interaction and high predictive value PLoS One 3 2008 e2031 [20] Y. Horikawa K. Miyake K. Yasuda M. Enya Y. Hirota K. Yamagata Replication of genome-wide association studies of type 2 diabetes susceptibility in Japan J Clin Endocrinol Metab 93 2008 3136 3141 [21] J.P. Lewis N.D. Palmer P.J. Hicks M.M. Sale C.D. Langefeld B.I. Freedman Association analysis in African Americans of European-derived type 2 diabetes single nucleotide polymorphisms from whole-genome association studies Diabetes 57 2008 2220 2225 [22] S. Cauchi K.T. Nead H. Choquet F. Horber N. Potoczna B. Balkau The genetic susceptibility to type 2 diabetes may be modulated by obesity status: implications for association studies BMC Med Genet 9 2008 45 [23] Y. Wu H. Li R.J. Loos Z. Yu X. Ye L. Chen Common variants in CDKAL1, CDKN2A/B, IGF2BP2, SLC30A8, and HHEX / IDE genes are associated with type 2 diabetes and impaired fasting glucose in a Chinese Han population Diabetes 57 2008 2834 2842 [24] J. Xiang X.Y. Li M. Xu J. Hong Y. Huang J.R. Tan Zinc transporter-8 gene (SLC30A8) is associated with type 2 diabetes in Chinese J Clin Endocrinol Metab 93 2008 4107 4112 [25] S. Omori Y. Tanaka A. Takahashi H. Hirose A. Kashiwagi K. Kaku Association of CDKAL1, IGF2BP2, CDKN2A/B, HHEX, SLC30A8, and KCNJ11 with susceptibility to type 2 diabetes in a Japanese population Diabetes 57 2008 791 795 [26] R. Rong R.L. Hanson D. Ortiz C. Wiedrich S. Kobes W.C. Knowler Association analysis of variation in/near FTO, CDKAL1, SLC30A8, HHEX, EXT2, IGF2BP2, LOC387761, and CDKN2B with type 2 diabetes and related quantitative traits in Pima Indians Diabetes 58 2009 478 488 [27] Y. Tabara H. Osawa R. Kawamoto H. Onuma I. Shimizu T. Miki Replication study of candidate genes associated with type 2 diabetes based on genome-wide screening Diabetes 58 2009 493 498 [28] M. van Hoek A. Dehghan J.C. Witteman C.M. van Duijn A.G. Uitterlinden B.A. Oostra Predicting type 2 diabetes based on polymorphisms from genome-wide association studies: a population-based study Diabetes 57 2008 3122 3128 [29] V. Lyssenko A. Jonsson P. Almgren N. Pulizzi B. Isomaa T. Tuomi Clinical risk factors, DNA variants, and the development of type 2 diabetes N Engl J Med 359 2008 2220 2232 [30] D.F. Stroup J.A. Berlin S.C. Morton I. Olkin G.D. Williamson D. Rennie Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of observational Studies in Epidemiology (MOOSE) group JAMA 283 2000 2008 2012 [31] J.P. Higgins S.G. Thompson Quantifying heterogeneity in a meta-analysis Stat Med 21 2002 1539 1558 [32] D. Petitti Meta-analysis, decision analysis, and cost-effectiveness analysis 1994 Oxford University Press New York [33] M. Egger G. Davey Smith M. Schneider C. Minder Bias in meta-analysis detected by a simple, graphical test BMJ 315 1997 629 634 [34] H. Gohlke U. Ferrari K. Koczwara E. Bonifacio T. Illig A.G. Ziegler SLC30A8 (ZnT8) Polymorphism is associated with young age at type 1 diabetes onset Rev Diabet Stud 5 2008 25 27 [35] C. Brorsson R. Bergholdt M. Sjögren S. Eising K.M. Sørensen D.M. Hougaard A non-synonymous variant in SLC30A8 is not associated with type 1 diabetes in the Danish population Mol Genet Metab 94 2008 386 388 [36] H.Q. Qu S.F. Grant J.P. Bradfield C. Kim E. Frackelton H. Hakonarson Association analysis of type 2 diabetes Loci in type 1 diabetes Diabetes 57 2008 1983 1986 [37] S. Cauchi C. Proença H. Choquet S. Gaget F. De Graeve M. Marre Analysis of novel risk loci for type 2 diabetes in a general French population: the D.E.S.I.R. study J Mol Med 86 2008 341 348 [38] Lin CY, Chen PC, Kuo HK, Lin LY, Lin JW, Hwang JJ. Effects of obesity, physical activity, and cardiorespiratory fitness on blood pressure, inflammation, and insulin resistance in the National Health and Nutrition Survey 1999–2002. Nutr Metab Cardiovasc Dis in press, doi:10.1016/j.numecd.2009.06.005 . [39] E. Morini S. Prudente E. Succurro M. Chandalia Y.Y. Zhang S. Mammarella IRS1 G972R polymorphism and type 2 diabetes: a paradigm for the difficult ascertainment of the contribution to disease susceptibility of ‘low-frequency-low-risk’ variants Diabetologia 52 2009 1852 1857 [40] J.B. McAteer S. Prudente S. Bacci H.N. Lyon J.N. Hirschhorn V. Trischitta The ENPP1 K121Q polymorphism is associated with type 2 diabetes in European populations: evidence from an updated meta-analysis in 42,042 subjects Diabetes 57 2008 1125 1130 [41] Rossi MC, Nicolucci A, Pellegrini F, Comaschi M, Ceriello A, Cucinotta D, et al. Obesity and changes in urine albumin/creatinine ratio in patients with type 2 diabetes: the DEMAND Study. Nutr Metab Cardiovasc Dis in press, doi:10.1016/j.numecd.2009.02.013 . [42] N.D. Palmer M.O. Goodarzi C.D. Langefeld J. Ziegler J.M. Norris S.M. Haffner Quantitative trait analysis of type 2 diabetes susceptibility loci identified from whole genome association studies in the Insulin Resistance Atherosclerosis Family Study Diabetes 57 2008 1093 1100 [43] F. Takeuchi M. Serizawa K. Yamamoto T. Fujisawa E. Nakashima K. Ohnaka Confirmation of multiple risk Loci and genetic impacts by a genome-wide association study of type 2 diabetes in the Japanese population Diabetes 58 2009 1690 1699 [44] T.J. Nicolson E.A. Bellomo N. Wijesekara M.K. Loder J.M. Baldwin A.V. Gyulkhandanyan Insulin storage and glucose homeostasis in mice null for the granule zinc transporter ZnT8 and studies of the type 2 diabetes-associated variants Diabetes 58 2009 2070 2083 [45] Y. Fu W. Tian E.B. Pratt L.B. Dirling S.L. Shyng C.K. Meshul Down-regulation of ZnT8 expression in INS-1 rat pancreatic beta cells reduces insulin content and glucose-inducible insulin secretion PLoS One 4 2009 e5679