A Low-Frequency Inactivating AKT2 Variant Enriched in the Finnish Population Is Associated With Fasting Insulin Levels and Type 2 Diabetes Risk

Abstract

To identify novel coding association signals and facilitate characterization of mechanisms influencing glycemic traits and type 2 diabetes risk, we analyzed 109,215 variants derived from exome array genotyping together with an additional 390,225 variants from exome sequence in up to 39,339 normoglycemic individuals from five ancestry groups. We identified a novel association between the coding variant (p.Pro50Thr) in AKT2 and fasting plasma insulin (FI), a gene in which rare fully penetrant mutations are causal for monogenic glycemic disorders. The low-frequency allele is associated with a 12% increase in FI levels. This variant is present at 1.1% frequency in Finns but virtually absent in individuals from other ancestries. Carriers of the FI-increasing allele had increased 2-h insulin values, decreased insulin sensitivity, and increased risk of type 2 diabetes (odds ratio 1.05). In cellular studies, the AKT2-Thr50 protein exhibited a partial loss of function. We extend the allelic spectrum for coding variants in AKT2 associated with disorders of glucose homeostasis and demonstrate bidirectional effects of variants within the pleckstrin homology domain of AKT2.

Figure 1

AKT2 Pro50Thr association with FI levels. A: For each study, the square represents the estimate of the additive genetic effect for the association of the AKT2 Pro50Thr allele with log-transformed FI levels and the horizontal line gives the corresponding 95% CI of the estimate. Inverse-variance meta-analyses were performed for all discovery studies, all replication studies, and all studies combined. The vertical dashed lines indicate the 95% CI for the estimate obtained in the meta-analysis of all studies combined. DPS, The Finnish Diabetes Prevention Study; DR’s EXTRA, Dose-Responses to Exercise Training study; FIN-D2D, National Diabetes Prevention Programme in Finland; PPP, Prevalence, Prediction and Prevention of Diabetes (PPP)-Botnia study. B: MAF for each available region and ancestry. Across countries of the world, the MAF ranges from 0 to 1.1%. The relative sample sizes (N) for each region/ancestry are displayed with the blue circles and the relative MAFs of AKT2 Pro50Thr are displayed with the purple circles, with the size of the circles showing comparative differences. Within Finland (inset), where the MAF ranges from 0.9 to 1.7%, birthplace and study center data were used to show the allele distribution across the country. ^aFINRISK 2007, ^bFIN-D2D 2007, ^cFINRISK 1997 and 2002.

Figure 2

Expression and conservation properties. A: Amino acid alignment and conservation of the three AKT proteins in vertebrates. The x-axis gives the amino acid position and the height of the lines shows the conservation score across 100 vertebrate genome alignments. The functional domains are the PH domain (blue) and the kinase domain (green). The position of AKT2 Pro50Thr is shown in red and the locations of the other AKT2 disease-causing mutations (–40) are shown in orange: Glu17Lys, Arg208Lys, Arg274His, and Arg467Trp. B: WebLogo plots of amino acids 35–60 are shown for AKT2, AKT1, and AKT3, contrasting the homology of the three isoforms. The height of letters gives the relative frequency of different amino acids across the 100 vertebrate species, with the colors showing amino acids with similar charge. C: Expression of AKT1, AKT2, and AKT3 in eight insulin-sensitive tissues using RNA sequencing data from the GTEx Consortium. subcut., subcutaneous.

Figure 3

Functional properties of AKT2-Thr50. A: Predicted protein structure of AKT2. Domain and variants are highlighted as in Fig. 2A. The relative spatial positioning of the AKT2-Pro50 residue is magnified within the inset. B: HeLa cells were infected with lentiviral V5-AKT2, V5-AKT2-Lys17, V5-AKT2-Thr50, V5-AKT2-Lys208, V5-AKT2-His274, or V5-AKT2-Trp467; starved for 18 h (white bar); and stimulated for 20 min with 100 nmol/L insulin (gray bar). V5-tagged AKT2 was isolated from cell lysates with anti-V5 agarose beads and incubated with GSK3β-GST peptide in an in vitro kinase assay. Quantification of phosphorylated substrate peptide (pGSK3β) relative to total peptide (GST-GSK3β) is shown at the inset. Immunoblots and quantification shown are representative of three independent replicates. Linear model statistical analyses across all three independent replicates are available in Supplementary Fig. 9. The in vitro kinase was immunoblotted (IB) with the indicated antibodies. C: HuH7 cells were infected with lentiviral V5-AKT2, V5-AKT2-Thr50, or control pLX304. At 72 h, relative cellular proliferation was determined with WST-1 assay of HuH7 cells. Error bars represent SD. ***P = 4.5 × 10⁻⁵.

Figure 4

Genetic architecture of rare, low-frequency, and common variants associated with FI levels. In this plot, the absolute values of the percent change in FI level due to rare monogenic mutations (diamonds) and common genetic variants (circles) are plotted against the MAF of the variant. The extremely rare monogenic mutations (above the dashed line to the left of the x-axis) were observed in 2–18 individuals (,–40,48,53,54), with the height of the point indicating the percent change in FI levels of mutation carriers from 40 pmol/L, an estimate of population mean FI level. Mutations in INSR and AKT2 p.Arg274His cause compensatory hyperinsulinemia, individuals with TBC1D4 p.Arg363Ter show normal FI levels but postprandial hyperinsulinemia, and mutations in PTEN cause enhanced insulin sensitivity providing protection against type 2 diabetes. For common variants, the percent change in FI levels per insulin-increasing allele is plotted above the solid horizontal axis. These observations are from sequencing (6) and array-based genome-wide association studies (3). For several genes, the effects from rare mutations can be compared with the effects of common variants in or near the gene: PPARG (blue), TBC1D4 (green), PTEN (orange), and AKT2 (red). ^aDonohue syndrome: biallelic LoF mutations in INSR (54). ^bRabson-Mendenhall syndrome: biallelic LoF mutations in INSR (54). ^cPostpubertal severe insulin resistance: heterozygous or homozygous LoF mutations in INSR (54). ^dLoF PTEN mutations cause Cowden syndrome in which carriers exhibit a lowered FI level (mean 29 pmol/L) compared with matched control subjects (3). ^eCarriers with the AKT2 p.Glu17Lys mutation were described with hypoinsulinemic hypoketotic hypoglycemia and hemihypertrophy with undetectable serum insulin (37,38).