We previously reported a family in which a heterozygous missense mutation in Akt2 led to a dominantly inherited syndrome of insulin-resistant diabetes and partial lipodystrophy. To determine whether genetic variation in AKT2 plays a broader role in human metabolic disease, we sequenced the entire coding region and splice junctions of AKT2 in 94 unrelated patients with severe insulin resistance, 35 of whom had partial lipodystrophy. Two rare missense mutations (R208K and R467W) were identified in single individuals. However, insulin-stimulated kinase activities of these variants were indistinguishable from wild type. In two large case-control studies (total number of participants 2,200), 0 of 11 common single nucleotide polymorphism (SNPs) in AKT2 showed significant association with type 2 diabetes. In a quantitative trait study of 1,721 extensively phenotyped individuals from the U.K., no association was found with any relevant intermediate metabolic trait. In summary, although heterozygous loss-of- function mutations in AKT2 can cause a syndrome of severe insulin resistance and lipodystrophy in humans, such mutations are uncommon causes of these syndromes. Furthermore, genetic variation in and around the AKT2 locus is unlikely to contribute significantly to the risk of type 2 diabetes or related intermediate metabolic traits in U.K. populations.

The serine/threonine-protein kinase Akt/protein kinase B (PKB) plays a critical role in insulin receptor–coupled phosphatidylinositol 3-kinase–mediated signaling (1). There are three Akt mammalian isoforms (Akt1–3), of which Akt2 is the most important in glucose metabolism (1). Mice deficient in Akt2 exhibit fed and fasting hyperglycemia, hyperinsulinemia, glucose intolerance, and impaired muscle glucose uptake (2,3). We identified a missense mutation in the kinase domain of Akt2 (R274H) in a single family with autosomal dominantly inherited severe insulin resistance and diabetes (4). The proband of this family had partial lipodystrophy (4), suggesting that Akt2 may play a role in adipogenesis. To date, detailed genetic association studies of AKT2 have not been reported.

We undertook studies to determine 1) whether other missense/nonsense mutations in AKT2 might result in human syndromes of severe insulin resistance with or without accompanying lipodystrophy and 2) whether common genetic variants in AKT2 might be associated with metabolic phenotypes related to insulin resistance. The entire coding sequence and splice junctions of AKT2 were screened in 94 probands with severe insulin resistance, 35 of which had partial lipodystrophy. The results of this screen are shown in online appendix Table 1 (available at http://dx.doi.org/10.2337/db06-0921). We identified two novel missense mutations (Fig. 1A). R467W was found in a white female patient with type 2 diabetes and partial lipodystrophy. This variant was present in neither 47 ethnically matched control subjects nor in 2 unaffected sons of the carrier. R208K was identified in a white female patient with severe insulin resistance and acanthosis nigricans. This variant was not present in her affected son but was present in 1 of 47 white control subjects. Unfortunately, parental DNA was not available to determine whether these mutations were inherited or spontaneous.

We investigated whether these mutations might cause functional impairment of the Akt2 kinase in vitro. CHO-T cells overexpressing the insulin receptor were transfected with either wild-type HA-Akt2 or mutant HA-Akt2. The ability of Akt2 to phosphorylate an artificial peptide substrate based on glycogen synthase kinase-3 was measured in an in vitro kinase assay using anti-HA immunoprecipitates from the transiently transfected cells, which were serum starved and then stimulated with insulin or left untreated. The known kinase-dead R274H mutant (4) was used as a negative control. Neither R208K nor R467W significantly altered either the basal or the insulin-stimulated kinase activity of Akt2 on the peptide substrate (Fig. 1B). The mutations did not affect expression levels (Fig. 1C). Upon stimulation with insulin, both Akt2 mutants were phosphorylated on T309 and S474 to a similar extent as wild-type Akt2 (Fig. 1C).

Although it is possible that these in vitro assays may fail to detect subtle changes in function of the Akt2 kinase, including, for example, selective impairments of activity at specific substrates, the normality of responses of the mutant kinases in two different assay systems suggests that the their function is likely to be unimpaired. Taken together, these data suggest these mutations are unlikely to be directly implicated in the severe insulin resistance of the probands.

To explore whether common single nucleotide polymorphisms (SNPs) in AKT2 are associated with type 2 diabetes, we studied two U.K. case-control studies: the Cambridgeshire Case-Control (CCC) Study and the European Prospective Investigation into Cancer and Nutrition (EPIC)-Norfolk Study. Based on the criteria described (see research design and methods), we obtained results for 11 polymorphic SNPs (Table 1). Average call rates were 93.3 and 97.7% for the CCC and EPIC-Norfolk studies, respectively. All SNPs were in Hardy-Weinberg equilibrium (HWE) (P > 0.01) in these populations. To increase power, we performed a joint analysis of the two case-control studies, with a term for study, and tested for heterogeneity between the studies (Table 1). When the genotype frequencies were compared between the combined case and control subjects, no statistically significant associations were detected (Table 1). Additional genotyping of other loci in these populations did not show evidence for heterogeneity; thus, the nominally significant associations in the separate case-control studies are likely to be due to chance. However, the heterogeneity could also be due to subtle differences in population substructure between study populations.

We next evaluated the degree of linkage disequilibrium among the 11 SNPs genotyped (Fig. 2). As shown, all but two SNPs (rs4273150 and rs3730256) fall into two linkage disequilibrium blocks (5,6). Haplotype analysis in each of these blocks did not provide evidence of statistically significant associations with disease risk (data not shown). Furthermore, to estimate coverage of the genetic variation captured, we compared our selected SNPs with those in Centre d’Etude du Polymorphism Humain samples in HapMap II (release no. 20). There are 21 SNPs with minor allele frequency (MAF) >5%, spanning from 22 kb upstream of the start of the transcript to the 3′ untranslated region of the gene. We have tagged all SNPs present in HapMap (pairwise tagging with r2 = 0.8 and MAF >5%) except for SNPs rs10426842 and rs12460555 (each of these just tag themselves). Taken together, our data suggest that common variants in AKT2 do not significantly contribute to the type 2 diabetes status in the U.K. population.

To examine whether the common variants in AKT2 are associated with metabolic parameters related to type 2 diabetes, we further genotyped 1,721 unrelated U.K. white participants in the Medical Research Council (MRC) Ely Study, a prospective population-based study of the etiology of type 2 diabetes (7). Eleven SNPs were genotyped with an average call rate of 96.6%, although SNP8 and SNP11 were not in HWE (P < 0.01) (Table 2). We tested for association of these SNPs with fasting and 2-h postchallenge plasma glucose levels, fasting plasma insulin levels, and 30-min insulin incremental response, a measure of insulin secretion (8) (Table 2). None of the SNPs showed any association (P < 0.01) with glucose or insulin levels. Therefore, these data suggest that common polymorphisms in AKT2 do not significantly contribute to plasma glucose or insulin levels.

In summary, we screened AKT2 as a candidate gene for human insulin resistance and partial lipodystrophy and detected two novel missense mutations, which did not affect function, at least in the assays used. We conducted the first association study of common variants in the AKT2 gene with human type 2 diabetes and related metabolic phenotypes. No statistically significant associations were found, suggesting that common variants in AKT2 are not associated with type 2 diabetes, in the populations studied. For a SNP with a MAF of 0.2, the power to detect an effect with an odds ratio >1.2 (or <0.82) is 80%, α = 0.05 (960 case and 1,386 control subjects). Therefore, small to moderate effects, if they exist, would be detected by our study. Further studies will be required to completely rule out small effects of AKT2 on type 2 diabetes risk.

Evidence from recent years has suggested that studying rare monogenic forms of diabetes not only contributes to our understanding of the mechanisms of glucose homeostasis but also might constitute an effective strategy to identify genes involved in more common and complex forms of diabetes. Indeed, while rare highly penetrant mutations in HNF4A, PPARG, and KIR6.2 lead to monogenic forms of diabetes, common variants in each of these genes have been shown to increase disease risk in type 2 diabetes (9). This study suggests that this is not the case for AKT2.

Genetic screening.

Genomic DNA was preamplified in a GenomiPhi reaction (Amersham Biosciences). Twelve primer pairs were designed to cover all coding exons and intron-exon boundaries using Primer3 software (online appendix Table 2). PCR was performed using Taq polymerase (Abgene), and PCR products were sequenced using Big Dye Terminator 3.1 DNA sequencing kit (Applied Biosystems).

Plasmid cloning and mutagenesis.

Akt2 cDNA was provided by D. Alessi (10). Mutagenesis was performed by QuickChange Site Directed Mutagenesis Kit (Stratagene).

Transient transfection and protein blotting.

CHO-T cells overexpressing the insulin receptor were transfected with pCDNA3.1-Akt2 or pCDNA3.1-Akt2 mutants using polyethyleneimine (11). Twenty-four hours after transfection, cells were serum starved for 8 h; 100 nmol/l insulin (human Actrapid; Abbott Laboratories) was added for 10 min, and the cells were lysed. Total protein was quantified using Bio-Rad Dc Protein Assay, and equal amounts of protein were run on 10% SDS-PAGE followed by Western blotting with anti–phospho-Akt (T308), anti–phospho-Akt (S473), and total Akt antibodies (Cell Signaling).

Immunoprecipitation and in vitro kinase assay.

Immunoprecipitation was carried out using 10 μg anti-HA (F-7) agarose conjugate (Santa Cruz) for 4 h at 4°C. Immunoprecipitates containing HA-Akt2 and HA-Akt2 mutants were assayed for in vitro kinase activity as previously described (10). Crosstide (Sigma) was used as the substrate in a reaction containing 50 mmol/l Tris, pH 7.5, 0.1 mmol/l EGTA, 0.1% (wt/vol) β-mercaptoethanol, 10 mmol/l magnesium acetate, 100 μmol/l 32PγATP, and 30 μmol/l Crosstide. The reaction was stopped after incubation at 30°C for 30 min, and the incorporated radioactivity was counted.

Severe insulin-resistant cohort.

The inclusion criteria for this cohort were 1) a fasting insulin >100 pmol/l or an insulin requirement >200 units/day, 2) acanthosis nigricans, and 3) BMI <33 kg/m2. From this cohort, 94 individuals who have not been previously screened for mutations in AKT2 were screened for mutations in all coding sequence and splice junctions of AKT2. After confirmation of genetic variants in the severe insulin resistant (SIR) cohort using unamplified genomic DNA, a control population, consisting of 47 white individuals, was screened for the presence of these variants. Lipodystrophic patients were a subset of the SIR patients in whom partial or complete lipodystrophy was diagnosed on the basis of clinical examination and in whom sequencing of the LMNA gene exons 8–12 and PPARG gene were normal.

Case-control studies

CCC Study.

This population-based study has been previously described (12). Briefly, this study consists of 552 type 2 diabetic patients aged 47–75 years and individually age-, sex-, and geographical location–matched control subjects. Case subjects were defined by onset of diabetes after the age of 30 years without insulin treatment in the 1st year following diagnosis. Potential control subjects that had A1C levels >6% were excluded.

EPIC-Norfolk participants.

This is a nested case-control study within the EPIC-Norfolk prospective cohort study; both the case-control and full cohort study (13,14) have been previously described in detail. Briefly, the case-control study consists of 417 incident type 2 diabetic case and control subjects and two sets of 417 control subjects, matched by age, sex, time in study, and family physician with the second set additionally matched for BMI. A case was defined by a physician’s diagnosis of type 2 diabetes, with no insulin prescribed within the 1st year after diagnosis and/or A1C >7% at baseline or the follow-up health check. Control subjects were selected from those in the cohort who had not reported diabetes, cancer, stroke, or myocardial infarction at baseline and who had not developed diabetes by the time of selection. Potential control subjects with measured A1C levels >6% were excluded. DNA was available for this analysis from 354 case and 741 control subjects.

MRC Ely Study.

This is a population-based cohort study of the etiology and pathogenesis of type 2 diabetes and related metabolic disorders in the U.K. (7). It is an ethnically homogeneous Europid population in which phenotypic data have been recorded at the outset and after 4.5 years. This analysis included 1,721 men and women aged 35–79 years without diagnosed diabetes who attended the study clinic for a health check between 2000 and 2004. Of these, 1,005 individuals were attending a follow-up health check, while the remaining 716 were newly recruited in 2000 from the original population sampling frame. Participants attending the health check underwent standard anthropometric measurements and a 75-g oral glucose tolerance test.

Ethical permission.

Ethical permission for the three studies was granted by their respective local research ethics committee, and study participants provided informed consent.

SNP selection.

SNP selection for this study was undertaken before HapMap I (15) data were available. SNPs were selected from the National Center for Biotechnology Information dbSNP database and direct sequencing of SIR samples. SNPs with a MAF ≥5% were selected. For dbSNPs with no frequency information, SNPs that had been validated by cluster and by submitter were preferentially selected. Where possible, it was attempted not to have gaps >2.5–3 kb between any consecutive SNPs selected. Nineteen different SNPs were chosen from the dbSNP database for genotyping. One additional SNP (SNP8) was identified through direct sequencing. Of these 20 SNPs, 5 failed assay design (rs4803320, rs892120, rs4803322, rs7409393, and rs6508935), 5 were monomorphic (rs2288917, rs3730260, rs1804324, rs1142298, and rs7247518), and 1 failed genotyping (rs892119) due to low call rate (<80%). After HapMap I data were available, two additional SNPs (rs3730051 and rs8100018) were selected to increase coverage of genetic variation in this gene. In total, data from 11 SNPs were available for analysis.

Methods for genotyping.

For the case-control populations, case and control samples were randomly distributed across each 96-well plate, with approximately the same number of case and control subjects per plate. Between 3 and 8.5% internal replicate samples were included in each population in all genotyping tests to assess genotyping accuracy. Genotyping of samples was performed in 384-well plates at the Wellcome Trust Sanger Institute, Cambridge, U.K., using an adaptation of the homogenous MassExtend protocol supplied by Sequenom for the MassArray system (Sequenom) (16). Call rates were ≥90%, and concordance rates between duplicate samples were ≥99.8% for all assays included in the analysis.

Statistical analyses.

All analyses used Stata/SE 9.2 for Windows (Stata Corporation, College Station, TX). Genotype frequencies were tested for HWE using a χ2 goodness-of-fit test in all samples in the Ely Study and in the control subjects of the case-control studies. For each SNP, an additive model (linear trend, which assumes an additive effect for the presence of zero, one, or two rare alleles) on 1 degree of freedom and a general model (compares the three genotypes as a categorical variable) on 2 degrees of freedom were performed to assess the association with diabetes and quantitative traits. Unconditional logistic regression was applied to the two combined case-control studies adjusted for age, sex, BMI, and study, and the matching variables were included as covariates. Heterogeneity between studies was tested, and, if significant, we used a random-effect meta-analysis, which incorporates an estimate of the between-study variation. Quantitative trait analysis was undertaken in the Ely Study population. Association between quantitative traits and genotype were tested in a linear regression model and adjusted for age, sex, and BMI. The likelihood-ratio test comparing statistical models did not show the general model to be better than the additive model in any of the SNPs; hence, we presented only the additive model.

FIG. 1.

Rare mutations in Akt2 in human individuals. A: Location of the identified mutations R208K and R467W in relation to functional domains and known phosphorylation sites. B: In vitro kinase assay of Akt2 mutants. HA-Akt2 and HA-Akt2 mutants were immunoprecipitated from lysates of appropriately transfected CHO-T cells treated with (▪) or without (□) 100 nmol/l insulin 10 min before lysis. Kinase assays were performed as described in research design and methods. Kinase activities are adjusted relative to that of unstimulated wild-type control subjects (100%). Data measured are means ± SD of five independent experiments. C: Equal amounts of lysates were immunoblotted with anti–phospho-Thr308 Akt (upper panel [pT309]) and anti–phospho-Ser473 Akt (middle panel [pS474]) antibodies to demonstrate increased phosphorylation in response to insulin. Immunoprecipitates were also immunoblotted with anti-HA antibody (lower panel [HA]) to demonstrate similar levels of immunoprecipitates used in the kinase assays.

FIG. 1.

Rare mutations in Akt2 in human individuals. A: Location of the identified mutations R208K and R467W in relation to functional domains and known phosphorylation sites. B: In vitro kinase assay of Akt2 mutants. HA-Akt2 and HA-Akt2 mutants were immunoprecipitated from lysates of appropriately transfected CHO-T cells treated with (▪) or without (□) 100 nmol/l insulin 10 min before lysis. Kinase assays were performed as described in research design and methods. Kinase activities are adjusted relative to that of unstimulated wild-type control subjects (100%). Data measured are means ± SD of five independent experiments. C: Equal amounts of lysates were immunoblotted with anti–phospho-Thr308 Akt (upper panel [pT309]) and anti–phospho-Ser473 Akt (middle panel [pS474]) antibodies to demonstrate increased phosphorylation in response to insulin. Immunoprecipitates were also immunoblotted with anti-HA antibody (lower panel [HA]) to demonstrate similar levels of immunoprecipitates used in the kinase assays.

FIG. 2.

Location and linkage disequilibrium map of AKT2 SNPs genotyped. Thirteen exons of AKT2 are represented by solid bars (numbered 1–13); intronic regions and 5′ and 3′ regions are represented by solid lines. The positions of SNPs 1–11 are indicated. The dbSNP reference numbers are indicated below each SNP. The pairwise linkage disequilibrium coefficient r2 (top) and D′ (bottom) for the control subjects in the case-control studies were calculated for genotyped SNPs using Haploview. Haplotype blocks were identified using Haploview.

FIG. 2.

Location and linkage disequilibrium map of AKT2 SNPs genotyped. Thirteen exons of AKT2 are represented by solid bars (numbered 1–13); intronic regions and 5′ and 3′ regions are represented by solid lines. The positions of SNPs 1–11 are indicated. The dbSNP reference numbers are indicated below each SNP. The pairwise linkage disequilibrium coefficient r2 (top) and D′ (bottom) for the control subjects in the case-control studies were calculated for genotyped SNPs using Haploview. Haplotype blocks were identified using Haploview.

TABLE 1

Relationship between genotype in AKT2 and type 2 diabetes status in a U.K. population

StudyCases*
Controls*
OR (95% CI)P (trend)P (heterogeneity)
111222111222
EPIC-Norfolk          
    SNP1 325 28 676 60 0.88 (0.55–1.43) 0.627  
    SNP2 265 78 535 176 23 0.87 (0.67–1.13) 0.312  
    SNP3 161 143 40 382 279 57 1.25 (1.02–1.53) 0.031  
    SNP4 270 73 595 130 10 1.21 (0.91–1.61) 0.190  
    SNP5 297 54 622 110 0.95 (0.68–1.33) 0.783  
    SNP6 145 142 36 332 256 50 1.24 (1.01–1.53) 0.045  
    SNP7 183 138 30 433 260 43 1.25 (1.01–1.54) 0.037  
    SNP8 159 146 37 370 284 54 1.22 (0.99–1.49) 0.061  
    SNP9 165 151 37 386 294 57 1.19 (0.97–1.45) 0.091  
    SNP10 265 81 540 174 23 0.91 (0.70–1.18) 0.462  
    SNP11 102 167 80 182 370 176 0.92 (0.76–1.11) 0.390  
CCC          
    SNP1 439 45 458 37 1.19 (0.77–1.83) 0.433  
    SNP2 338 123 12 344 118 1.09 (0.84–1.42) 0.505  
    SNP3 242 173 33 226 199 44 0.81 (0.65–1.00) 0.047  
    SNP4 405 75 391 102 0.69 (0.49–0.96) 0.028  
    SNP5 421 58 419 76 0.78 (0.54–1.12) 0.182  
    SNP6 288 205 45 243 227 51 0.78 (0.64–0.95) 0.012  
    SNP7 286 170 19 263 194 35 0.74 (0.59–0.92) 0.007  
    SNP8 253 166 36 220 181 45 0.80 (0.65–0.99) 0.042  
    SNP9 254 189 38 235 215 47 0.82 (0.67–1.00) 0.050  
    SNP10 338 132 10 358 128 10 1.09 (0.84–1.41) 0.518  
    SNP11 123 229 124 149 231 114 1.17 (0.98–1.41) 0.082  
EPIC-Norfolk and CCC          
    SNP1 764 73 1,134 97 1.04 (0.76–1.43) 0.789 0.373 
    SNP2 603 201 20 879 294 32 0.98 (0.81–1.17) 0.794 0.237 
    SNP3 403 316 73 608 478 101 1.01 (0.66–1.54) 0.981 0.003 
    SNP4 675 148 986 232 12 0.92 (0.53–1.60) 0.764 0.012 
    SNP5 718 112 1,041 186 0.87 (0.68–1.11) 0.269 0.415 
    SNP6 433 347 81 575 483 101 0.98 (0.62–1.55) 0.938 0.001 
    SNP7 469 308 49 696 454 78 0.96 (0.58–1.61) 0.882 0.001 
    SNP8 412 312 73 590 465 99 0.99 (0.66–1.48) 0.959 0.006 
    SNP9 419 340 75 621 509 104 0.99 (0.68–1.43) 0.938 0.010 
    SNP10 603 213 18 898 302 33 0.99 (0.83–1.19) 0.941 0.336 
    SNP11 225 396 204 331 601 290 1.04 (0.92–1.19) 0.528 0.068 
StudyCases*
Controls*
OR (95% CI)P (trend)P (heterogeneity)
111222111222
EPIC-Norfolk          
    SNP1 325 28 676 60 0.88 (0.55–1.43) 0.627  
    SNP2 265 78 535 176 23 0.87 (0.67–1.13) 0.312  
    SNP3 161 143 40 382 279 57 1.25 (1.02–1.53) 0.031  
    SNP4 270 73 595 130 10 1.21 (0.91–1.61) 0.190  
    SNP5 297 54 622 110 0.95 (0.68–1.33) 0.783  
    SNP6 145 142 36 332 256 50 1.24 (1.01–1.53) 0.045  
    SNP7 183 138 30 433 260 43 1.25 (1.01–1.54) 0.037  
    SNP8 159 146 37 370 284 54 1.22 (0.99–1.49) 0.061  
    SNP9 165 151 37 386 294 57 1.19 (0.97–1.45) 0.091  
    SNP10 265 81 540 174 23 0.91 (0.70–1.18) 0.462  
    SNP11 102 167 80 182 370 176 0.92 (0.76–1.11) 0.390  
CCC          
    SNP1 439 45 458 37 1.19 (0.77–1.83) 0.433  
    SNP2 338 123 12 344 118 1.09 (0.84–1.42) 0.505  
    SNP3 242 173 33 226 199 44 0.81 (0.65–1.00) 0.047  
    SNP4 405 75 391 102 0.69 (0.49–0.96) 0.028  
    SNP5 421 58 419 76 0.78 (0.54–1.12) 0.182  
    SNP6 288 205 45 243 227 51 0.78 (0.64–0.95) 0.012  
    SNP7 286 170 19 263 194 35 0.74 (0.59–0.92) 0.007  
    SNP8 253 166 36 220 181 45 0.80 (0.65–0.99) 0.042  
    SNP9 254 189 38 235 215 47 0.82 (0.67–1.00) 0.050  
    SNP10 338 132 10 358 128 10 1.09 (0.84–1.41) 0.518  
    SNP11 123 229 124 149 231 114 1.17 (0.98–1.41) 0.082  
EPIC-Norfolk and CCC          
    SNP1 764 73 1,134 97 1.04 (0.76–1.43) 0.789 0.373 
    SNP2 603 201 20 879 294 32 0.98 (0.81–1.17) 0.794 0.237 
    SNP3 403 316 73 608 478 101 1.01 (0.66–1.54) 0.981 0.003 
    SNP4 675 148 986 232 12 0.92 (0.53–1.60) 0.764 0.012 
    SNP5 718 112 1,041 186 0.87 (0.68–1.11) 0.269 0.415 
    SNP6 433 347 81 575 483 101 0.98 (0.62–1.55) 0.938 0.001 
    SNP7 469 308 49 696 454 78 0.96 (0.58–1.61) 0.882 0.001 
    SNP8 412 312 73 590 465 99 0.99 (0.66–1.48) 0.959 0.006 
    SNP9 419 340 75 621 509 104 0.99 (0.68–1.43) 0.938 0.010 
    SNP10 603 213 18 898 302 33 0.99 (0.83–1.19) 0.941 0.336 
    SNP11 225 396 204 331 601 290 1.04 (0.92–1.19) 0.528 0.068 
*

Odds ratio (OR) calculated per allele 2 from an additive logistic model on the genotypes, adjusted for age, sex, and BMI.

P <0.05 for linear trend model.

P <0.01 for heterogeneity between studies. Results were adjusted for heterogeneity whenever significant heterogeneity between studies was present. Because there were six SNPs with significant heterogeneity (P < 0.05), we reported the results for each case-control study and the combined data. 1, major allele, 2, minor allele.

TABLE 2

Relationship between genotype in AKT2 and glucose and insulin levels in a U.K. population

Genotype frequencies111222P (HWE)
    SNP1 1,479 124 0.719 
    SNP2 1,203 430 38 0.954 
    SNP3 839 656 165 0.029 
    SNP4 1,362 297 15 0.788 
    SNP5 1,452 227 0.554 
    SNP6 834 664 166 0.047 
    SNP7 944 605 129 0.021 
    SNP8 847 637 166 0.005* 
    SNP9 849 664 161 0.063 
    SNP10 1,205 438 41 0.873 
    SNP11 472 768 409 0.006* 
Genotype frequencies111222P (HWE)
    SNP1 1,479 124 0.719 
    SNP2 1,203 430 38 0.954 
    SNP3 839 656 165 0.029 
    SNP4 1,362 297 15 0.788 
    SNP5 1,452 227 0.554 
    SNP6 834 664 166 0.047 
    SNP7 944 605 129 0.021 
    SNP8 847 637 166 0.005* 
    SNP9 849 664 161 0.063 
    SNP10 1,205 438 41 0.873 
    SNP11 472 768 409 0.006* 
Parameter111222P (trend)
PG0     
    SNP1 5.04 (5.01–5.07) 5.00 (4.89–5.12) 5.20 (4.35–6.22) 0.668 
    SNP2 5.00 (4.97–5.04) 5.10 (5.04–5.16) 5.11 (4.90–5.32) 0.034 
    SNP3 5.04 (5.00–5.09) 5.02 (4.97–5.08) 5.04 (4.94–5.14) 0.541 
    SNP4 5.04 (5.00–5.07) 5.04 (4.96–5.11) 5.01 (4.68–5.35) 0.459 
    SNP5 5.03 (5.00–5.07) 5.03 (4.94–5.12) 5.28 (4.79–5.82) 0.549 
    SNP6 5.04 (4.99–5.08) 5.02 (4.97–5.07) 5.05 (4.95–5.16) 0.701 
    SNP7 5.03 (4.98–5.07) 5.04 (4.99–5.09) 5.04 (4.93–5.15) 0.934 
    SNP8 5.05 (5.00–5.09) 5.02 (4.97–5.07) 5.04 (4.95–5.14) 0.366 
    SNP9 5.04 (5.00–5.09) 5.02 (4.97–5.07) 5.03 (4.93–5.13) 0.396 
    SNP10 5.01 (4.97–5.04) 5.11 (5.04–5.17) 5.11 (4.91–5.32) 0.034 
    SNP11 5.08 (5.02–5.14) 5.02 (4.97–5.06) 5.02 (4.95–5.08) 0.318 
2-h plasma glucose     
    SNP1 6.03 (5.93–6.14) 6.06 (5.71–6.44) 6.61 (4.18–10.45) 0.923 
    SNP2 6.02 (5.90–6.14) 6.02 (5.83–6.22) 6.28 (5.63–7.01) 0.745 
    SNP3 6.07 (5.93–6.21) 5.90 (5.75–6.06) 6.31 (5.99–6.65) 0.790 
    SNP4 5.99 (5.88–6.10) 6.19 (5.95–6.43) 6.32 (5.28–7.56) 0.070 
    SNP5 6.00 (5.90–6.11) 6.10 (5.83–6.38) 6.82 (5.11–9.10) 0.497 
    SNP6 6.06 (5.92–6.21) 5.91 (5.76–6.07) 6.26 (5.94–6.59) 0.804 
    SNP7 6.04 (5.90–6.17) 5.95 (5.79–6.11) 6.24 (5.89–6.61) 0.704 
    SNP8 6.07 (5.93–6.21) 5.87 (5.72–6.03) 6.29 (5.97–6.63) 0.908 
    SNP9 6.07 (5.93–6.21) 5.92 (5.77–6.08) 6.28 (5.95–6.62) 0.829 
    SNP10 6.01 (5.90–6.13) 6.03 (5.83–6.22) 6.36 (5.72–7.07) 0.839 
    SNP11 6.21 (6.02–6.41) 5.90 (5.76–6.04) 6.10 (5.90–6.30) 0.633 
INS0§     
    SNP1 49.3 (48.0–50.6) 50.9 (46.5–55.8) 51.9 (25.5–105.5) 0.246 
    SNP2 48.5 (47.1–49.9) 49.1 (46.8–51.6) 54.1 (46.0–63.6) 0.658 
    SNP3 49.6 (47.9–51.4) 47.8 (46.0–49.8) 51.3 (47.5–55.5) 0.830 
    SNP4 49.2 (47.8–50.5) 48.3 (45.6–51.2) 59.0 (45.5–76.4) 0.307 
    SNP5 48.6 (47.3–49.9) 50.8 (47.5–54.4) 71.1 (48.5–104.3) 0.025 
    SNP6 49.2 (47.5–51.0) 47.3 (45.5–49.3) 52.2 (48.3–56.5) 0.816 
    SNP7 49.5 (47.9–51.2) 47.9 (46.0–49.9) 50.8 (46.4–55.5) 0.482 
    SNP8 49.6 (47.9–51.4) 47.9 (46.0–49.9) 51.3 (47.5–55.5) 0.875 
    SNP9 49.5 (47.8–51.3) 47.5 (45.7–49.4) 51.4 (47.4–55.6) 0.770 
    SNP10 48.6 (47.2–50.0) 49.7 (47.4–52.2) 51.0 (43.5–59.8) 0.672 
    SNP11 50.3 (48.0–52.7) 48.3 (46.6–50.2) 48.5 (46.1–51.0) 0.348 
INS incremental     
    SNP1 30.4 (29.4–31.5) 30.5 (27.0–34.5) 22.9 (9.2–56.9) 0.874 
    SNP2 30.1 (29.0–31.1) 31.7 (29.7–33.9) 27.7 (22.2–34.7) 0.447 
    SNP3 30.7 (29.3–32.2) 30.3 (28.7–31.9) 28.6 (25.8–31.8) 0.329 
    SNP4 30.8 (29.7–31.9) 29.0 (26.8–31.4) 30.1 (20.8–43.6) 0.245 
    SNP5 30.6 (29.5–31.7) 30.0 (27.4–32.9) 26.3 (14.8–46.8) 0.715 
    SNP6 31.0 (29.5–32.5) 30.3 (28.7–31.9) 28.8 (26.0–32.0) 0.290 
    SNP7 30.9 (29.6–32.3) 30.4 (28.8–32.2) 27.9 (24.8–31.4) 0.183 
    SNP8 30.8 (29.4–32.3) 31.0 (29.3–32.7) 28.5 (25.6–31.6) 0.431 
    SNP9 30.8 (29.4–32.3) 30.4 (28.9–32.1) 28.6 (25.7–31.8) 0.305 
    SNP10 30.1 (29.0–31.3) 31.6 (29.6–33.7) 27.7 (22.3–34.4) 0.525 
    SNP11 29.1 (27.3–31.0) 31.2 (29.8–32.8) 30.5 (28.5–32.6) 0.404 
Parameter111222P (trend)
PG0     
    SNP1 5.04 (5.01–5.07) 5.00 (4.89–5.12) 5.20 (4.35–6.22) 0.668 
    SNP2 5.00 (4.97–5.04) 5.10 (5.04–5.16) 5.11 (4.90–5.32) 0.034 
    SNP3 5.04 (5.00–5.09) 5.02 (4.97–5.08) 5.04 (4.94–5.14) 0.541 
    SNP4 5.04 (5.00–5.07) 5.04 (4.96–5.11) 5.01 (4.68–5.35) 0.459 
    SNP5 5.03 (5.00–5.07) 5.03 (4.94–5.12) 5.28 (4.79–5.82) 0.549 
    SNP6 5.04 (4.99–5.08) 5.02 (4.97–5.07) 5.05 (4.95–5.16) 0.701 
    SNP7 5.03 (4.98–5.07) 5.04 (4.99–5.09) 5.04 (4.93–5.15) 0.934 
    SNP8 5.05 (5.00–5.09) 5.02 (4.97–5.07) 5.04 (4.95–5.14) 0.366 
    SNP9 5.04 (5.00–5.09) 5.02 (4.97–5.07) 5.03 (4.93–5.13) 0.396 
    SNP10 5.01 (4.97–5.04) 5.11 (5.04–5.17) 5.11 (4.91–5.32) 0.034 
    SNP11 5.08 (5.02–5.14) 5.02 (4.97–5.06) 5.02 (4.95–5.08) 0.318 
2-h plasma glucose     
    SNP1 6.03 (5.93–6.14) 6.06 (5.71–6.44) 6.61 (4.18–10.45) 0.923 
    SNP2 6.02 (5.90–6.14) 6.02 (5.83–6.22) 6.28 (5.63–7.01) 0.745 
    SNP3 6.07 (5.93–6.21) 5.90 (5.75–6.06) 6.31 (5.99–6.65) 0.790 
    SNP4 5.99 (5.88–6.10) 6.19 (5.95–6.43) 6.32 (5.28–7.56) 0.070 
    SNP5 6.00 (5.90–6.11) 6.10 (5.83–6.38) 6.82 (5.11–9.10) 0.497 
    SNP6 6.06 (5.92–6.21) 5.91 (5.76–6.07) 6.26 (5.94–6.59) 0.804 
    SNP7 6.04 (5.90–6.17) 5.95 (5.79–6.11) 6.24 (5.89–6.61) 0.704 
    SNP8 6.07 (5.93–6.21) 5.87 (5.72–6.03) 6.29 (5.97–6.63) 0.908 
    SNP9 6.07 (5.93–6.21) 5.92 (5.77–6.08) 6.28 (5.95–6.62) 0.829 
    SNP10 6.01 (5.90–6.13) 6.03 (5.83–6.22) 6.36 (5.72–7.07) 0.839 
    SNP11 6.21 (6.02–6.41) 5.90 (5.76–6.04) 6.10 (5.90–6.30) 0.633 
INS0§     
    SNP1 49.3 (48.0–50.6) 50.9 (46.5–55.8) 51.9 (25.5–105.5) 0.246 
    SNP2 48.5 (47.1–49.9) 49.1 (46.8–51.6) 54.1 (46.0–63.6) 0.658 
    SNP3 49.6 (47.9–51.4) 47.8 (46.0–49.8) 51.3 (47.5–55.5) 0.830 
    SNP4 49.2 (47.8–50.5) 48.3 (45.6–51.2) 59.0 (45.5–76.4) 0.307 
    SNP5 48.6 (47.3–49.9) 50.8 (47.5–54.4) 71.1 (48.5–104.3) 0.025 
    SNP6 49.2 (47.5–51.0) 47.3 (45.5–49.3) 52.2 (48.3–56.5) 0.816 
    SNP7 49.5 (47.9–51.2) 47.9 (46.0–49.9) 50.8 (46.4–55.5) 0.482 
    SNP8 49.6 (47.9–51.4) 47.9 (46.0–49.9) 51.3 (47.5–55.5) 0.875 
    SNP9 49.5 (47.8–51.3) 47.5 (45.7–49.4) 51.4 (47.4–55.6) 0.770 
    SNP10 48.6 (47.2–50.0) 49.7 (47.4–52.2) 51.0 (43.5–59.8) 0.672 
    SNP11 50.3 (48.0–52.7) 48.3 (46.6–50.2) 48.5 (46.1–51.0) 0.348 
INS incremental     
    SNP1 30.4 (29.4–31.5) 30.5 (27.0–34.5) 22.9 (9.2–56.9) 0.874 
    SNP2 30.1 (29.0–31.1) 31.7 (29.7–33.9) 27.7 (22.2–34.7) 0.447 
    SNP3 30.7 (29.3–32.2) 30.3 (28.7–31.9) 28.6 (25.8–31.8) 0.329 
    SNP4 30.8 (29.7–31.9) 29.0 (26.8–31.4) 30.1 (20.8–43.6) 0.245 
    SNP5 30.6 (29.5–31.7) 30.0 (27.4–32.9) 26.3 (14.8–46.8) 0.715 
    SNP6 31.0 (29.5–32.5) 30.3 (28.7–31.9) 28.8 (26.0–32.0) 0.290 
    SNP7 30.9 (29.6–32.3) 30.4 (28.8–32.2) 27.9 (24.8–31.4) 0.183 
    SNP8 30.8 (29.4–32.3) 31.0 (29.3–32.7) 28.5 (25.6–31.6) 0.431 
    SNP9 30.8 (29.4–32.3) 30.4 (28.9–32.1) 28.6 (25.7–31.8) 0.305 
    SNP10 30.1 (29.0–31.3) 31.6 (29.6–33.7) 27.7 (22.3–34.4) 0.525 
    SNP11 29.1 (27.3–31.0) 31.2 (29.8–32.8) 30.5 (28.5–32.6) 0.404 

Data are means (95% CI) adjusted for age, sex, and BMI.

*

P < 0.01 for HWE.

PG0 refers to fasting plasma glucose level (mmol/l).

2-h plasma glucose refers to plasma glucose level 2 h after glucose bolus (mmol/l).

§

INS0 refers to fasting plasma insulin level (pmol/l).

INS incremental refers to a 30-min insulin incremental response (the difference between 30-min and fasting insulin concentrations divided by the 30-min glucose concentration in an oral glucose tolerance test) (pmol/mmol). 1, major allele, 2, minor allele.

Additional information for this article can be found in an online appendix at http://dx.doi.org/10.2337/db06-0921.

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.

I.B. is funded by the Wellcome Trust. S.O. and I.B. acknowledge support from EU FP6 funding (contract no. LSHM-CT-2003-503041). K.T. is supported by A*STAR (Singapore).

The CCC Study is funded by the Wellcome Trust (to N.J.W. and S.O.). The EPIC-Norfolk Study is funded by MRC UK and Cancer Research UK. The MRC Ely Study is funded by the MRC and Wellcome Trust (to N.J.W.).

We thank S. Bumpstead for genotyping technical support and F. Payne for additional technical support.

1.
Taniguchi CM, Emanuelli B, Kahn CR: Critical nodes in signalling pathways: insights into insulin action.
Nat Rev Mol Cell Biol
7
:
85
–96,
2006
2.
Cho H, Mu J, Kim JK, Thorvaldsen JL, Chu Q, Crenshaw EB 3rd, Kaestner KH, Bartolomei MS, Shulman GI, Birnbaum MJ: Insulin resistance and a diabetes mellitus-like syndrome in mice lacking the protein kinase Akt2 (PKB beta).
Science
292
:
1728
–1731,
2001
3.
Garofalo RS, Orena SJ, Rafidi K, Torchia AJ, Stock JL, Hildebrandt AL, Coskran T, Black SC, Brees DJ, Wicks JR, McNeish JD, Coleman KG: Severe diabetes, age-dependent loss of adipose tissue, and mild growth deficiency in mice lacking Akt2/PKB beta.
J Clin Invest
112
:
197
–208,
2003
4.
George S, Rochford JJ, Wolfrum C, Gray SL, Schinner S, Wilson JC, Soos MA, Murgatroyd PR, Williams RM, Acerini CL, Dunger DB, Barford D, Umpleby AM, Wareham NJ, Davies HA, Schafer AJ, Stoffel M, O’Rahilly S, Barroso I: A family with severe insulin resistance and diabetes due to a mutation in AKT2.
Science
304
:
1325
–1328,
2004
5.
Barrett JC, Fry B, Maller J, Daly MJ: Haploview: analysis and visualization of LD and haplotype maps.
Bioinformatics
21
:
263
–265,
2005
6.
Gabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, Higgins J, DeFelice M, Lochner A, Faggart M, Liu-Cordero SN, Rotimi C, Adeyemo A, Cooper R, Ward R, Lander ES, Daly MJ, Altshuler D: The structure of haplotype blocks in the human genome.
Science
296
:
2225
–2229,
2002
7.
Wareham NJ, Byrne CD, Williams R, Day NE, Hales CN: Fasting proinsulin concentrations predict the development of type 2 diabetes.
Diabetes Care
22
:
262
–270,
1999
8.
Wareham NJ, Phillips DIW, Byrne CD, Hales CN: The 30 minute insulin incremental response in an oral glucose tolerance test as a measure of insulin secretion.
Diabet Med
12
:
931
,
1995
9.
O’Rahilly S, Barroso I, Wareham NJ: Genetic factors in type 2 diabetes: the end of the beginning?
Science
307
:
370
–373,
2005
10.
Walker KS, Deak M, Paterson A, Hudson K, Cohen P, Alessi DR: Activation of protein kinase B beta and gamma isoforms by insulin in vivo and by 3-phosphoinositide dependent protein kinase-1 in vitro: comparison with protein kinase B alpha.
Biochem J
331
:
299
–308,
1998
11.
Durocher Y, Perret S, Kamen A: High-level and high-throughput recombinant protein production by transient transfection of suspension-growing human 293-EBNA1 cells.
Nucleic Acids Res
30
:
E9
,
2002
12.
Barroso I, Luan J, Middelberg RP, Harding AH, Franks PW, Jakes RW, Clayton D, Schafer AJ, O’Rahilly S, Wareham NJ: Candidate gene association study in type 2 diabetes indicates a role for genes involved in beta-cell function as well as insulin action.
PLoS Biol
1
:
E20
,
2003
13.
Harding AH, Day NE, Khaw KT, Bingham S, Luben R, Welsh A, Wareham NJ: Dietary fat and the risk of clinical type 2 diabetes: the European Prospective Investigation of Cancer-Norfolk Study.
Am J Epidemiol
159
:
73
–82,
2004
14.
Day N, Oakes S, Luben R, Khaw KT, Bingham S, Welch A, Wareham N: EPIC-Norfolk: study design and characteristics of the cohort: European Prospective Investigation of Cancer.
Br J Cancer
80 (Suppl. 1)
:
95
–103,
1999
15.
The International HapMap Consortium: The International HapMap Project.
Nature
426
:
789
–796,
2003
16.
Whittaker P, Bumpstead S, Downes K, Ghori J, Deloukas P: SNP analysis by MALDI-TOF mass spectrometry. In
Cell Biology: A Laboratory Handbook.
3rd ed. Celis J, Carter N, Simons K, Smoll JV, Hunter T, Eds. San Diego, CA, Academic Press,
2005

Supplementary data