Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Loss-of-function mutations in SLC30A8 protect against type 2 diabetes


Loss-of-function mutations protective against human disease provide in vivo validation of therapeutic targets1,2,3, but none have yet been described for type 2 diabetes (T2D). Through sequencing or genotyping of 150,000 individuals across 5 ancestry groups, we identified 12 rare protein-truncating variants in SLC30A8, which encodes an islet zinc transporter (ZnT8)4 and harbors a common variant (p.Trp325Arg) associated with T2D risk and glucose and proinsulin levels5,6,7. Collectively, carriers of protein-truncating variants had 65% reduced T2D risk (P = 1.7 × 10−6), and non-diabetic Icelandic carriers of a frameshift variant (p.Lys34Serfs*50) demonstrated reduced glucose levels (−0.17 s.d., P = 4.6 × 10−4). The two most common protein-truncating variants (p.Arg138* and p.Lys34Serfs*50) individually associate with T2D protection and encode unstable ZnT8 proteins. Previous functional study of SLC30A8 suggested that reduced zinc transport increases T2D risk8,9, and phenotypic heterogeneity was observed in mouse Slc30a8 knockouts10,11,12,13,14,15. In contrast, loss-of-function mutations in humans provide strong evidence that SLC30A8 haploinsufficiency protects against T2D, suggesting ZnT8 inhibition as a therapeutic strategy in T2D prevention.


Genome-wide association studies (GWAS) have identified 65 genomic loci associated with T2D risk7, highlighting previously unidentified pathological pathways. Translation of these loci into novel therapeutic targets16 requires the identification of causal mutations and genes, as well as information on the directional relationship between protein activity and disease risk17. Toward this end, loss-of-function mutations that protect against disease (without adverse phenotypes) are among the most useful findings from human genetics, suggesting targets that, upon inhibition, might prevent disease in the general population.

To identify loss-of-function variants protective against T2D, in 2009, a collaboration of Pfizer, Inc., Massachusetts General Hospital, the Broad Institute and Lund University sequenced the exons of 115 genes near T2D association signals identified by GWAS (Supplementary Fig. 1 and Supplementary Tables 1 and 2) in 758 individuals from Finland or Sweden (modeling previous studies18). To increase power, we selected individuals at the extremes of T2D risk, including 352 young and lean T2D cases and 406 elderly and obese euglycemic controls19 (Supplementary Table 3). In total, we identified 1,768 nonsynonymous variants (1,683 single-nucleotide variants (SNVs) and 85 indels), 1,474 (83%) with minor allele frequency (MAF) of <1% and 1,108 (63%) observed in only one individual. We found no evidence of association with T2D when testing individual variants or a burden of rare variants within genes (Supplementary Fig. 2). Genotyping of 71 select SNVs (showing nominally significant association (P < 0.05) or predicted to affect protein structure) in up to 13,884 additional individuals (6,388 cases and 7,496 controls) also yielded results consistent with the null distribution of no association with T2D (Supplementary Fig. 3).

Among these SNVs, we noted a nonsense variant (c.412C>T, p.Arg138*) in SLC30A8 (transcript accession NM_173851) that trended toward protection from T2D rather than risk (7 case and 21 control observations, odds ratio (OR) = 0.38, P = 0.05). To propagate this variant into larger sample sizes, we added p.Arg138* to the design of the Illumina Human Exome array (along with six other variants from the original sequencing that were predicted to cause protein truncation). Using this array, we genotyped the 13,884 individuals already genotyped for p.Arg138*, as well as 7,212 additional Finnish or Swedish individuals (Supplementary Tables 4 and 5). In the combined analysis of 10,534 cases and 10,562 controls, the protective association for p.Arg138* remained nominally significant (OR = 0.46, P = 0.01; Supplementary Table 6). We genotyped p.Arg138* in an additional 26,566 European individuals (8,210 T2D cases and 18,356 controls; Supplementary Table 7), with 16 heterozygotes observed (2 cases and 14 controls). Combining all available genotype data for the variant encoding p.Arg138*, heterozygosity was estimated to yield a 53% reduction in T2D risk (P = 0.0067, N = 48,115).

SLC30A8 encodes an islet zinc transporter ZnT8 (NP_776250; ref. 20), which is necessary for zinc flux into β cell insulin-secretory granules4 and subsequent insulin crystallization10,12. Upon co-secretion with insulin, zinc also fulfills autocrine and paracrine signaling roles21. A previously identified common missense variant in SLC30A8 (rs13266634; c.973T>A, p.Trp325Arg) associates with T2D risk7,22 and glucose5 and proinsulin6 levels with significance far exceeding genome-wide thresholds (P < 5 × 10−8).

Cellular characterization has suggested that the risk-increasing allele encoding p.Trp325Arg reduces ZnT8 zinc transport activity8,9. In Slc30a8 knockout mice, observations range from no effect on insulin secretion or glucose homeostasis to modest hyperglycemia on a high-fat diet, with the phenotype varying with sex and genetic background8,9,10,11,12,13,14,15. Furthermore, data obtained from a recent β cell–specific Slc30a8 knockout mouse model suggest a multi-organ effect on the phenotype, with circulating zinc also influencing hepatic insulin clearance in these mice21. Thus, the directional relationship between perturbed ZnT8 function and whole-organism phenotype is uncertain, despite much genetic and biological data.

Because the observed protective association between p.Arg138* and T2D risk was statistically modest, we sought additional evidence. Unfortunately, the near absence of the p.Arg138* variant outside of western Finland limited our ability to further characterize its effect in other populations (Supplementary Figs. 4 and 5). We thus sought to identify a wider spectrum of protein-truncating variants in SLC30A8 by investigating the catalog of 35 million variants collected by deCODE Genetics through whole-genome sequencing23. The variant encoding p.Arg138* was not observed in this data set. However, an independent protein-truncating variant was observed at a frequency of 0.17%—a deletion (c.100_106delAAAGATC, p.Lys34Serfs*50; Supplementary Figs. 6 and 7) predicted to cause a frameshift and loss of all six transmembrane domains encoded by the islet-specific transcript (NM_173851) of SLC30A8 (ref. 4).

Heterozygosity for the variant encoding p.Lys34Serfs*50 was associated with 80% reduced T2D risk, with 2 observations of this variant in 2,953 T2D cases (0.03%) versus 248 observations in 67,919 controls (0.18%; OR = 0.18, P = 0.004; Supplementary Tables 8 and 9). Taking into consideration the ancestral relationship between Norwegian and Icelandic populations, we genotyped the variant in 5,714 Norwegians (Supplementary Table 8) and observed no carriers in 1,645 cases and 3 carriers in 4,069 controls. Combining the evidence for the p.Lys34Serfs*50 and p.Arg138* variants strengthened the association between SLC30A8 protein-truncating variants and reduced T2D risk (combined OR = 0.32, P = 2.4 × 10−4).

Both rare SLC30A8 variants are bioinformatically predicted to cause ZnT8 truncation and, consequently, to affect activity. To test this prediction, we assessed overexpressed V5-tagged ZnT8 variants (Trp325, Arg325, *138 (as well as Arg138*) and Ser34fs*50) in HeLa cells20 (Fig. 1a). Despite similar RNA transcript levels for all variants (Supplementary Fig. 8), only Trp325 and Arg325 ZnT8 proteins were easily detectable in cells8, with Arg138*, *138 and Ser34fs*50 ZnT8 present at low to undetectable levels (Fig. 1b,c). Similar results were obtained using antibodies against the native protein or the V5 tag in protein blot analysis (Fig. 1b) and immunofluorescence (Fig. 1c) and in HeLa as well as in Ins1e rat insulinoma cells (Fig. 1d). Coexpression of *138 or Ser34fs*50 ZnT8 with Trp325 ZnT8 did not decrease expression of the full-length allele or rescue expression of either truncating variant (Supplementary Fig. 9).

Figure 1: Overexpression of Arg138* and Ser34fs*50 ZnT8 in HeLa cells.

We sought to experimentally evaluate whether the Arg138* or Ser34fs*50 ZnT8 variants resulted in decreased ZnT8 expression and/or activity. (a) Depiction of the C-terminally V5-tagged ZnT8 proteins encoded by SLC30A8 ORFs (tag highlighted in green). (b) Protein blot of HeLa lysates after transient overexpression of V5-tagged ZnT8 variants (using antibody to the V5 tag). Antibody against tubulin was used as a loading control for each sample, and untransfected cell lysate was used to demonstrate the specificity of the antibody to V5. (c,d) Immunofluorescent staining of ZnT8 variants in HeLa (c) and Ins1e (d) cells. ZnT8 was detected using antibodies against the C-terminal V5 tag or the N terminus of the endogenous protein (both green), as indicated. BFP-V5 and untransfected HeLa cells serve as controls. Cells were also stained with Hoechst 33342 to mark nuclei (blue). Within each row, identical exposure times were used for all images. (e) ZnT8 variant expression, as detected by immunostaining with antibody to V5, following 4-h treatment with inhibitors of the lysosome (chloroquine, 100 μM) or proteasome (MG132, 10 μM). Images were acquired using either a 63× (top row) or 10× (all other rows) objective and identical exposure times. Scale bars, 100 μm.

We hypothesized that decreased expression of these two mutants might be due to protein instability and/or enhanced degradation. After treatment of cells with chloroquine or MG132 (lysosomal and proteasomal inhibitors, respectively24), higher *138 and Ser34fs*50 ZnT8 expression was detected via immunofluorescence (but remained undetectable via protein blot analysis; Fig. 1e and Supplementary Fig. 10). These results are consistent with but do not prove instability and subsequent degradation of these truncated proteins25. In additional experiments (data not shown), we observed zinc transport in cells expressing Arg325 and Trp325 ZnT8 but not in cells expressing *138 or Ser34fs*50 ZnT8 (an expected result given the low levels of mutant protein). Further experiments are needed to assess the in vivo impact of these variants, including their susceptibility to nonsense-mediated decay and potential dominant-negative effects on protein oligomerization.

These genetic and functional data suggest that SLC30A8 haploinsufficiency reduces T2D risk. However, confidence in this conclusion would be further increased through the observation of multiple additional putative loss-of-function variants demonstrating protective effects. As part of the Genetics of Type 2 Diabetes (Go-T2D) and Type 2 Diabetes Genetic Exploration by Next-Generation Sequencing in Multi-Ethnic Samples (T2D-GENES) consortia, we sequenced SLC30A8 exons in 12,294 individuals spanning multiple ancestry groups (Supplementary Table 10). Nine additional protein-truncating variants were identified—2 frameshift indels and 2 nonsense, 4 splice-site and 1 initiator codon SNV—in 23 heterozygous individuals of African-American, East Asian and South Asian ancestry (Supplementary Data Set 1). The variant encoding p.Arg138* was seen in three additional carriers (one case and two controls); the variant encoding p.Lys34Serfs*50 was not observed.

In aggregate, carriers of these additional variants exhibited 60% reduced T2D risk (4 observations of the variants in cases versus 18 observations in controls; OR = 0.38, P = 0.0025), with similar effects and statistical significance observed upon analysis of only carriers of frameshift or nonsense variants (2 observations of variants in cases versus 13 observations in controls; OR = 0.37, P = 0.0027). Combining all data from sequencing and genotyping in 149,134 subjects, heterozygosity for any of the 12 protein-truncating variants was associated with 65% reduced T2D risk (OR = 0.34, P = 1.7 × 10−6), which represents a statistically significant association even after correction for the 20,000 genes in the human genome (Table 1).

Table 1 Association of SLC30A8 variants with T2D

We investigated potential confounding factors for the observed protective association. First, we assessed whether the haplotypic background encoding the p.Trp325Arg variant might influence results. Whereas the variants encoding p.Lys34Serfs*50 and p.Met50Ile were observed on the protective common variant haplotype, the remaining variants (including the one encoding p.Arg138*) were observed on the risk-associated common variant haplotype. Thus, independent protective protein-truncating variants were observed on opposite haplotypic backgrounds encoding p.Trp325Arg. Second, we tested for a survivor effect, where carriers of rare variants with diabetes would die at a younger age than non-carriers. However, (i) carrier ages did not differ significantly from non-carrier ages for either p.Arg138* (69.6 ± 8.4 years versus 65.5 ± 11.0 years for cases (P > 0.1), 46.4 ± 15.7 years versus 50.3 ± 15.5 years for controls (P > 0.1)) or p.Lys34Serfs*50 (70.5 ± 4.5 years versus 65.6 ± 13.8 years for cases (P > 0.1), 48.5 ± 20.1 years versus 50.0 ± 23.2 years for controls (P > 0.1)), and (ii) p.Lys34Serfs*50 association attained equivalent significance even when the analysis was restricted to age-matched controls. Finally, although we employed well-established methods using linear mixed models or principal components to control for population structure26, we acknowledge the noted challenges to address this potential confounder in rare variant association studies27. We had insufficient data to control for stratification via a family-based transmission disequilibrium test (pedigree information was only available for Icelanders, with three carrier parents all transmitting the risk-associated allele to affected children). However, the consistent association of multiple independent protein-truncating variants across multiple cohorts and ancestry groups argues against population stratification being responsible for protective association.

These data thus provide compelling evidence that mutations inactivating one copy of SLC30A8 reduce T2D risk in humans. In addition to T2D risk, the common SLC30A8 variant encoding p.Trp325Arg is associated with proinsulin and fasting plasma glucose levels at genome-wide significance5,6, as well as with 2-h glucose levels after an oral glucose tolerance test (OGTT) at nominal significance (Supplementary Table 11)27. We asked whether rare protein-truncating SLC30A8 variants also affected T2D-related phenotypes, particularly glycemic traits that might be indicative of altered islet function.

Among the traits analyzed (Supplementary Table 12), the strongest statistical evidence for association was observed in Icelanders between p.Lys34Serfs*50 and random (non-fasting) glucose levels: non-diabetic carriers of the protective allele had lower glucose levels (β = −0.17 s.d.; N = 182 carriers; P = 4.6 × 10−4), with a consistent effect seen in three Norwegian carriers (β = −0.3 s.d.; P > 0.1). Glucose levels were lower at 1 h in the small number of p.Lys34Serfs*50 carriers characterized by OGTT (β = −0.73 s.d.; N = 4 carriers; P = 0.05). We did not observe a significant difference in fasting glucose or insulin levels, although the directions of effect were consistent with those described above—these levels were lower for fasting glucose (average β = −0.10 s.d.; N = 146 carriers; P > 0.1) and higher for fasting insulin (β = 0.24 s.d.; N = 52 carriers; P = 0.09). The parallel directionality of glucose levels and T2D risk is similar to the pattern observed for p.Trp325Arg, where the T2D-protective allele also associates with lower glucose levels (Supplementary Table 9), providing further evidence against a survivor effect or population stratification driving the protective association.

In summary, we identified 12 rare predicted protein-truncating SLC30A8 variants (Fig. 2). Carriers of these variants had 65% reduced T2D risk at a level of significance adequate to correct for the 20,000 genes in the human genome (P = 1.7 × 10−6). Non-diabetic Icelandic carriers of p.Lys34Serfs*50 also demonstrated lower glucose levels (β = −0.17 s.d., P = 4.6 × 10−4). Notably, initial sequencing of 115 genes in 758 individuals with extreme T2D phenotypes resulted in only 2 observations of the variant encoding p.Arg138*, without significant evidence of association of low-frequency or rare variants, individually or in aggregate, for any of the sequenced genes. Rather, establishing the association of SLC30A8 protein-truncating variants with T2D protection at levels of exome-wide significance (with correction for 20,000 genes) required genotyping of 150,000 individuals spanning multiple ancestry groups. Detecting similar effects in genes without previous evidence of association may require analysis on a similar or larger scale, for not only T2D but also for other complex traits.

Figure 2: Protein-truncating variants identified in SLC30A8.

Through sequencing and genotyping of nearly 150,000 individuals across 5 ancestry groups, we identified 12 SLC30A8 variants—each rare and predicted to cause premature protein truncation. (a) Shown is the position of each variant on the islet-specific SLC30A8 transcript (NM_173851; ref. 20). p.Met50Ile is predicted to alter the initiator codon in other transcripts of SLC30A8. Lines are drawn from each variant to the ancestry groups in which carriers were observed, with greater width corresponding to more observations. Lines are further drawn from each ancestry group to the populations (cohorts) in which carriers were identified. From left to right, the cohorts are JHS, WFS, Botnia, Danish, deCODE, Finnish, HUNT2, KORA, Malmö, PIVUS/ULSAM, WTCCC, LOLIPOP, Singapore Indians and KARE (cohort information appears in the Supplementary Note). Ancestry groups or cohorts with no observations are not shown. (b) Graphical representation of the case (red) and control (blue) frequencies for each observed variant. Wider bars correspond to more observations. A quantitative and complete representation of these data is given in Table 1.

Previous modeling of the relationship between ZnT8 activity and T2D risk centered on p.Trp325Arg, which results in mildly attenuated zinc transport that is concomitant with increased T2D risk8, and Slc30a8 knockout mice, where phenotypic heterogeneity is observed13,15. We find a clear and consistent association between putative SLC30A8 loss-of-function variants and T2D risk across multiple ancestry groups, demonstrating convincingly that a 50% reduction in gene dosage protects against T2D in humans. These data reject the model in which loss of ZnT8 function is associated with as little as a 1.2-fold increase in T2D risk (similar to the increase in risk observed for the common p.Trp325Arg variant) at significance of P ≈ 1 × 10−9. Phenotypic interrogation of human mutation carriers is needed to determine the physiological mechanism behind this protective association and to establish the effects of SLC30A8 haploinsufficiency in the pancreas and other tissues21.

The observed human genetics data present several implications for SLC30A8 function in T2D pathophysiology. The identification of multiple disease-associated protein-altering variants in SLC30A8 unambiguously (albeit unsurprisingly) documents SLC30A8 as the causal gene behind GWAS association signals. The observation that protein-truncating variants protect against T2D defines the directional relationship between ZnT8 activity and T2D risk in humans. The expanded SLC30A8 allelic series offers a more functionally informative catalog of variation compared to the p.Trp325Arg variant alone, enabling future experiments investigating potential mechanisms. Although substantial work is required to understand how reduced ZnT8 activity lowers T2D risk, the current observations motivate experiments to test ZnT8 inhibition in T2D treatment in human populations.


Sequencing and genotyping.

Individuals were selected for initial sequencing from several population-based cohorts from Finland and Sweden. A custom hybrid selection array was used to target genes, which were sequenced on an Illumina HiSeq 2000. Additional individuals from these same cohorts, as well as from other cohorts drawn from different European populations, were genotyped for the SLC30A8 nonsense SNV encoding p.Arg138* using the Illumina HumanExome v1.1 array. All sequenced individuals were also genotyped, with data showing 100% concordance.

Icelandic individuals were genotyped for the frameshift variant encoding p.Lys34Serfs*50 using a combination of whole-genome sequencing and imputation (either direct imputation based on chip genotyping or family-based imputation). Sanger sequencing was used to confirm carriers. Norwegian individuals were genotyped with a fragment length–based method using differentially labeled fluorescent primers, with Sanger sequencing again used to confirm carriers. Further SLC30A8 sequencing (aimed at identifying additional carriers of rare variants) was performed as part of a whole-exome sequencing experiment, with the Agilent SureSelect Human All Exon platform used to capture exons and an Illumina HiSeq 2000 used for sequencing.

These studies were performed using protocols approved by the ethics committees of Helsinki University Hospital, Finland, and Lund University, the Data Protection Commission of Iceland and the National Bioethics Committee of Iceland, the Regional Committee for Research Ethics and the Norwegian Data Inspectorate, and the Massachusetts Institute of Technology Institutional Review Board, as well as with informed consent from all participants.

Association analysis.

Association analysis was performed separately for three groups of variants: p.Arg138*, p.Lys34Serfs*50 and the remaining variants. For p.Arg138*, association analysis was separate for each analyzed cohort and used a linear mixed model to account for sample structure, including population stratification and genetic relatedness. Results were combined via a fixed-effects meta-analysis. For p.Lys34Serfs*50, association analysis was performed in Iceland using logistic regression, with controls matched to cases on the basis of how informative the imputed genotypes were, and in Norway using a simple logistic regression with significance calculated via the score statistic. For the remaining variants, all individuals were analyzed jointly via a collapsing method, regressing phenotype on the presence of any variant, with a linear mixed model used to account for sample structure. The resulting three association statistics were combined via a random-effects meta-analysis to obtain combined estimates of effect size and statistical significance.

Accession codes.

SLC30A8 transcript, NM_173851 (ref. 20); ZnT8 protein, NP_776250 (ref. 20).

Accession codes

Primary accessions

NCBI Reference Sequence


  1. 1

    Nassar, M.A. et al. Nociceptor-specific gene deletion reveals a major role for Nav1.7 (PN1) in acute and inflammatory pain. Proc. Natl. Acad. Sci. USA 101, 12706–12711 (2004).

    CAS  Article  Google Scholar 

  2. 2

    Cohen, J. et al. Low LDL cholesterol in individuals of African descent resulting from frequent nonsense mutations in PCSK9. Nat. Genet. 37, 161–165 (2005).

    CAS  Article  Google Scholar 

  3. 3

    Sullivan, D. et al. Effect of a monoclonal antibody to PCSK9 on low-density lipoprotein cholesterol levels in statin-intolerant patients: the GAUSS randomized trial. J. Am. Med. Assoc. 308, 2497–2506 (2012).

    CAS  Article  Google Scholar 

  4. 4

    Chimienti, F. et al. In vivo expression and functional characterization of the zinc transporter ZnT8 in glucose-induced insulin secretion. J. Cell Sci. 119, 4199–4206 (2006).

    CAS  Article  Google Scholar 

  5. 5

    Dupuis, J. et al. New genetic loci implicated in fasting glucose homeostasis and their impact on type 2 diabetes risk. Nat. Genet. 42, 105–116 (2010).

    CAS  Article  Google Scholar 

  6. 6

    Strawbridge, R.J. et al. Genome-wide association identifies nine common variants associated with fasting proinsulin levels and provides new insights into the pathophysiology of type 2 diabetes. Diabetes 60, 2624–2634 (2011).

    CAS  Article  Google Scholar 

  7. 7

    Morris, A.P. et al. Large-scale association analysis provides insights into the genetic architecture and pathophysiology of type 2 diabetes. Nat. Genet. 44, 981–990 (2012).

    CAS  Article  Google Scholar 

  8. 8

    Nicolson, T.J. et al. 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, 2070–2083 (2009).

    CAS  Article  Google Scholar 

  9. 9

    Rutter, G.A. Think zinc: new roles for zinc in the control of insulin secretion. Islets 2, 49–50 (2010).

    Article  Google Scholar 

  10. 10

    Lemaire, K. et al. Insulin crystallization depends on zinc transporter ZnT8 expression, but is not required for normal glucose homeostasis in mice. Proc. Natl. Acad. Sci. USA 106, 14872–14877 (2009).

    CAS  Article  Google Scholar 

  11. 11

    Pound, L.D. et al. Deletion of the mouse Slc30a8 gene encoding zinc transporter-8 results in impaired insulin secretion. Biochem. J. 421, 371–376 (2009).

    CAS  Article  Google Scholar 

  12. 12

    Wijesekara, N. et al. Beta cell–specific Znt8 deletion in mice causes marked defects in insulin processing, crystallisation and secretion. Diabetologia 53, 1656–1668 (2010).

    CAS  Article  Google Scholar 

  13. 13

    Pound, L.D. et al. The physiological effects of deleting the mouse Slc30a8 gene encoding zinc transporter-8 are influenced by gender and genetic background. PLoS ONE 7, e40972 (2012).

    CAS  Article  Google Scholar 

  14. 14

    Hardy, A.B. et al. Effects of high-fat diet feeding on Znt8-null mice: differences between β-cell and global knockout of Znt8. Am. J. Physiol. Endocrinol. Metab. 302, E1084–E1096 (2012).

    CAS  Article  Google Scholar 

  15. 15

    da Silva Xavier, G., Bellomo, E.A., McGinty, J.A., French, P.M. & Rutter, G.A. Animal models of GWAS-identified type 2 diabetes genes. J. Diabetes Res. 2013, 906590 (2013).

    Article  Google Scholar 

  16. 16

    van de Bunt, M. & Gloyn, A.L. From genetic association to molecular mechanism. Curr. Diab. Rep. 10, 452–466 (2010).

    CAS  Article  Google Scholar 

  17. 17

    Plenge, R.M., Scolnick, E.M. & Altshuler, D. Validating therapeutic targets through human genetics. Nat. Rev. Drug Discov. 12, 581–594 (2013).

    CAS  Article  Google Scholar 

  18. 18

    Nejentsev, S., Walker, N., Riches, D., Egholm, M. & Todd, J.A. Rare variants of IFIH1, a gene implicated in antiviral responses, protect against type 1 diabetes. Science 324, 387–389 (2009).

    CAS  Article  Google Scholar 

  19. 19

    Guey, L.T. et al. Power in the phenotypic extremes: a simulation study of power in discovery and replication of rare variants. Genet. Epidemiol. doi:10.1002/gepi.20572 (9 February 2011).

  20. 20

    Chimienti, F., Devergnas, S., Favier, A. & Seve, M. Identification and cloning of a β-cell-specific zinc transporter, ZnT-8, localized into insulin secretory granules. Diabetes 53, 2330–2337 (2004).

    CAS  Article  Google Scholar 

  21. 21

    Tamaki, M. et al. The diabetes-susceptible gene SLC30A8/ZnT8 regulates hepatic insulin clearance. J. Clin. Invest. 123, 4513–4524 (2013).

    CAS  Article  Google Scholar 

  22. 22

    Sladek, R. et al. A genome-wide association study identifies novel risk loci for type 2 diabetes. Nature 445, 881–885 (2007).

    CAS  Article  Google Scholar 

  23. 23

    Gudmundsson, J. et al. A study based on whole-genome sequencing yields a rare variant at 8q24 associated with prostate cancer. Nat. Genet. 44, 1326–1329 (2012).

    CAS  Article  Google Scholar 

  24. 24

    Bloom, J. & Pagano, M. Experimental tests to definitively determine ubiquitylation of a substrate. Methods Enzymol. 399, 249–266 (2005).

    CAS  Article  Google Scholar 

  25. 25

    Waters, P.J. Degradation of mutant proteins, underlying “loss of function” phenotypes, plays a major role in genetic disease. Curr. Issues Mol. Biol. 3, 57–65 (2001).

    CAS  PubMed  Google Scholar 

  26. 26

    Mathieson, I. & McVean, G. Differential confounding of rare and common variants in spatially structured populations. Nat. Genet. 44, 243–246 (2012).

    CAS  Article  Google Scholar 

  27. 27

    Price, A. L., Zaitlen, N. A., Relch, D. & Patterson, N. New approaches to population stratification in genome-wide association studies. Nat. Rev. Genet. 11, 459–463 (2010).

    CAS  Article  Google Scholar 

  28. 28

    Saxena, R. et al. Genetic variation in GIPR influences the glucose and insulin responses to an oral glucose challenge. Nat. Genet. 42, 142–148 (2010).

    CAS  Article  Google Scholar 

Download references


This manuscript is dedicated to the memory of David R. Cox, our dear friend and colleague, who was relentlessly supportive of this work—and more generally, of the use of human genetics to improve human health. He is missed, but his legacy goes on. We gratefully acknowledge the contribution of all 150,000 participants from the various population studies that contributed to this work. J.F. was supported in part by US National Institutes of Health (NIH) Training Grant 5-T32-GM007748-33. D.A. was supported by funding from the Doris Duke Charitable Foundation (2006087). N.L.B. was supported by a Fulbright Diabetes UK Fellowship (BDA 11/0004348). This work was supported in part by funding to the Broad Institute (principal investigator D.A.) from Pfizer, Inc. Funding for the GoT2D and T2D-GENES studies was provided by grants 5U01DK085526 (NIH/NIDDK; Multiethnic Study of Type 2 Diabetes Genes), DK088389 (NIH/NIDDK; Low-Pass Sequencing and High-Density SNP Genotyping for Type 2 Diabetes) and U54HG003067 (National Human Genome Research Institute (NHGRI); Large-Scale Sequencing and Analysis of Genomes), as well as by NIH grants U01 DK085501, U01 DK085524, U01 DK085545 and U01 DK085584. The Malmö Preventive Project and the Scania Diabetes Registry were supported by grants from the Swedish Research Council (Dnr 521-2010-3490 to L.G. and Dnr 349-2006-237 to the Lund University Diabetes Centre), as well as by a European Research Council (ERC) grant (GENETARGET T2D, GA269045) and two European Union grants (ENGAGE (2007-201413) and CEED3 (2008-223211)) to L.G. The Botnia study was supported by funding from the Sigrid Juselius Foundation and the Folkhälsan Research Foundation. P.R.N. was funded by the ERC (AdG 293574), the Research Council of Norway (197064/V50), the KG Jebsen Foundation, the University of Bergen, the Western Norway Health Authority, the European Association for the Study of Diabetes Sabbatical Leave Programme and Innovest. The Danish studies were supported by the Lundbeck Foundation (Lundbeck Foundation Centre for Applied Medical Genomics in Personalised Disease Prediction, Prevention and Care (LuCamp); and the Danish Council for Independent Research. The Novo Nordisk Foundation Center for Basic Metabolic Research is an independent Research Center at the University of Copenhagen, partially funded by an unrestricted donation from the Novo Nordisk Foundation ( The PIVUS/ULSAM cohort was supported by Wellcome Trust grants WT098017, WT064890 and WT090532, Uppsala University, the Uppsala University Hospital, the Swedish Research Council and the Swedish Heart-Lung Foundation. The METSIM study was supported by the Academy of Finland (contract 124243), the Finnish Heart Foundation, the Finnish Diabetes Foundation, Tekes (contract 1510/31/06), the Commission of the European Community (HEALTH-F2-2007-201681) and grants R01DK062370 and R01DK072193 (NIH/NIDDK) and grant Z01HG000024 (NHGRI). The FUSION study was supported by grants R01DK062370 and R01DK072193 (NIH/NIDDK) and grant Z01HG000024 (NHGRI). The DR's EXTRA Study was supported by the Ministry of Education and Culture of Finland (627; 2004-2011), the Academy of Finland (102318; 123885), Kuopio University Hospital, the Finnish Diabetes Association, the Finnish Heart Association and the Päivikki and Sakari Sohlberg Foundation and by grants from the European Commission Framework Programme 6 Integrated Project (EXGENESIS); LSHM-CT-2004-005272, City of Kuopio and Social Insurance Institution of Finland (4/26/2010). V. Salomaa is funded by the Academy of Finland, grant 139635, and by the Finnish Foundation for Cardiovascular Disease. Sequencing and genotyping of British individuals was supported by Wellcome Trust grants WT090367, WT090532 and WT098381 and by NIH/NIDDK grant U01-DK085545. Funding for the Jackson Heart Study (JHS) was provided by the National Heart, Lung, and Blood Institute (NHLBI) and by the National Institute on Minority Health and Health Disparities (N01 HC-95170, N01 HC-95171 and N01 HC-95172). A.P.M. acknowledges support from Wellcome Trust grants WT098017, WT090532 and WT064890. F.V.-S. and H.S. were supported by the European Union Seventh Framework Programme, DIAPREPP (Diabetes Type 1 Prediction, Early Pathogenesis and Prevention, grant agreement 202013), and by the Swedish Child Diabetes Foundation (Barndiabetesfonden).

Author information





This manuscript describes an analysis spanning four initially distinct sequencing studies—a collaborative project between Pfizer, Massachusetts General Hospital, the Broad Institute and Lund University entitled “Towards Therapeutic Targets for Type 2 Diabetes and Myocardial Infarction in the Background of Type 2 Diabetes” (PMBL), an effort by deCODE Genetics to use whole-genome sequencing and imputation to identify and genotype over 35 million variants in up to 370,000 Icelanders23, the Genetics of Type 2 Diabetes (GoT2D) project and the Type 2 Diabetes Genetic Exploration by Next-Generation Sequencing in Multi-Ethnic Samples (T2D-GENES) project—as well as four additional genotyping efforts. The overall study bringing together data from these efforts was coordinated by J.F. and D.A., with final analysis combining data from all variants collected by J.F. The manuscript was written by J.F., D.A. and K.S., and all authors reviewed, edited and approved the manuscript. Author contributions specific to the sequencing or genotyping studies are as follows.

Pfizer, Massachusetts General Hospital, Broad Institute and Lund University. Study design. B.F.V., F.B., S.P., N.P.B., D.R.C., T.R., L.G., S.K. and D.A. Clinical investigation and sample management. T. Forsén, B.I., T. Tuomi, L.G. and J. Kravic. Sequencing, genotyping and data processing. N.P.B., T. Fennell and S.G.; performed at the Broad Institute. Analysis. J.F. (p.Arg138*) and A.V.S. (p.Trp325Arg). Functional studies. N.L.B., S.B.R.J., Z.D. (cellular models), F.V.-S., H.S. and D.A.D. (screen for carrier autoantibodies). Leadership and management. N.P.B., A.-M.R., J. Brosnan, J.K.T., S.K., T.R., L.G., D.R.C. and D.A.

deCODE Genetics. Clinical investigation and sample management. A.B.H. and R.B. Sequencing, genotyping and data processing. G.T., V. Steinthorsdottir, P.S., G.M., D.F.G. and A.K.; performed at deCODE Genetics. Analysis. G.T., V. Steinthorsdottir, P.S., G.M., D.F.G. and A.K. Leadership and management. A.K., U.T. and K.S.

T2D-GENES and Go-T2D. Clinical investigation and sample management. G.A., J. Blangero, D.W.B., J.C., Y.S.C., R.D., B.G., C.H., J. Kooner, M.L., T.M., J.-Y.L., E.S.T., Y.Y.T. and J.G.W. Sequencing and data processing. N.P.B., Y.F., T. Fennell and S.G.; performed at the Broad Institute. Analysis. J.F. (SLC30A8), P.F., A.P.M. and T. Teslovich (exome wide). Leadership and management. M.I.M., M.B. and D.A. HUNT2 sample management, genotyping, analysis and leadership involved K.H., A. Molven, S.J. and P.R.N. Danish sample management, genotyping, analysis and leadership involved N.G., R.R.-M., M.E.J., C.C., I.B., A.L., T.J., T.H. and O.P. PIVUS/ULSAM sample management, genotyping, analysis and leadership involved A. Mahajan, C.M.L., L.L., E.I. and A.P.M. Finnish sample management, genotyping, analysis and leadership involved C.F., H.M.S., M.L., K.L.M., R.R., V. Salomaa, J.T. and M.B.

Corresponding authors

Correspondence to Kari Stefansson or David Altshuler.

Ethics declarations

Competing interests

G.T., V. Steinthorsdottir, P.S., G.M., D.F.G., A.K., U.T. and K.S. are employed by deCODE Genetics/Amgen, Inc. S.P., A.-M.R., J. Brosnan, J.K.T., T.R. and D.R.C. are employees of Pfizer, Inc. F.B. is a former employee of Pfizer, Inc., and retains shares in the company. All other authors declare no competing financial interests.

Additional information

Full lists of members and affiliations appear in the Supplementary Note.

Full lists of members and affiliations appear in the Supplementary Note.

Integrated supplementary information

Supplementary Figure 1 Technical quality control metrics of targeted sequencing.

We computed various metrics to evaluate the quality of the initial sequencing experiment. Shown from the left are the distributions (over all sequenced samples) of the number of variant sites with 10x sequence coverage and hence nonmissing genotypes (Call Rate), the number of minor alleles in genotypes called across all variant sites (Minor Alleles), the number of minor alleles at sites where no other samples have minor alleles (Singletons), the fraction of genotypes called heterozygous (Heterozygosity), the ratio of the number of heterozygous sites to the number of minor allele homozygous sites (Het to Hom ratio), the fraction of sequence reads with the minor allele (averaged over all heterozygous sites; Allele Balance) and the fraction of minor allele genotypes identical to those called from Exome Chip or Metabochip genotyping (Concordance). Absent differential technical artifacts or population structure, distributions are expected to be similar between cases and controls. Statistical comparison between case and control distributions was performed via a Kruskal-Wallis one-way analysis of variance test.

Supplementary Figure 2 Association with T2D of nonsynonymous variants from the initial sequencing experiment.

Based on data from the initial sequencing experiment, we performed two types of as- sociation tests for all low-frequency (below 1%) nonsynonymous variants. First, variants were tested individually using a linear mixed model. Second, variants were collapsed within each gene and tested for aggregate association using the same mixed model approach. Shown are QQ plots of association for each approach: the x-axis plots the expected distribution of (−log10) P-values under the null model whereby no association exists across the entire experiment (e.g. the uniform distribution); the y-axis plots the observed distribution of (−log10) p-values. The blue lines show estimated 95% confidence intervals for observed distributions under the null model. For each plot, only variants (or, analogously, genes) for which ten or more carriers are observed are plotted: for variants with small numbers of counts, the uniform distribution is not a good approximation to the P -value distribution expected under the null model.

Supplementary Figure 3 Association with T2D of 71 variants from small-scale genotyping.

We genotyped select variants from the initial sequencing experiment in up to 13,884 additional Finnish and Swedish individuals. Shown is a QQ plot of associations (with the same format as in Supplementary Figure 2), tested via a logistic regression.

Supplementary Figure 4 Frequency of the p.Arg138* variant in Europe.

The p.Arg138* variant was genotyped in individuals from Finland, Sweden, Denmark, the UK, and Poland (through the Illumina Human Exome Array or custom Sequenom genotyping), as well as Iceland (through whole genome sequencing). Shown are the observed frequencies of the variant in each country; frequency estimates from Finland are stratified by the Botnia region (individuals from the Botnia cohort) and other regions of Finland. N indicates the number of individuals genotyped.

Supplementary Figure 5 Frequency of the p.Arg138* variant in Botnia.

Frequency estimates for Botnia (as computed in Supplementary Figure 4) were further stratified by town in Botnia. Information on the center of sample selection was available for all samples from the Botnia region of Finland. Individuals were grouped by study center, and frequency estimates were computed for each center separately. N indicates the number of individuals genotyped.

Supplementary Figure 6 Position of the p.Lys34Serfs*50 frameshift mutation.

Shown is the genomic position of the frameshift mutation p.Lys34Serfs*50, as visualized in the UCSC genome browser (16).

Supplementary Figure 7 Partial sequence chromatogram for the p.Lys34Serfs*50 frameshift variant.

Sanger sequencing was used to confirm carriers of the frameshift mutation. Shown is a chromatogram for one individual carrier (from Norway) heterozygous for the variant.

Supplementary Figure 8 Transcript levels of SLC30A8 variants.

nCounter analysis of SLC30A8 transcript levels in HeLa cells, following transient over-expression of C-terminal, V5-tagged ZnT8 variants and control ORFs. Data shown are mean, normalized mRNA counts ± s.e.m. of three independent plasmid transfections from two experiments. Non-specific binding is indicated by the red line. Gene expression detected by probes directed against the SLC30A8 sequence encoding the N terminus, the sequence encoding the V5 tag, and TUBB are shown.

Supplementary Figure 9 Coexpression of ZnT8 variants.

HeLa cells overexpressed either (a) singly transfected C-terminal, V5-tagged ZnT8 variants or (b) cotransfections that also included C-terminally HA-tagged Trp325-ZnT8 for 24h, after which ZnT8 protein levels were observed. Cells were immunostained for ZnT8 expression using anti-HA and anti-V5 antibodies and costained with Hoechst 33342 to mark nuclei. Scale bars, 100μm (c) Protein blot analysis of HeLa cell lysates following cotransfection of Trp325-HA ZnT8 and V5-tagged ZnT8 variants. ZnT8 expression was detected using anti-V5 and anti-HA antibodies.

Supplementary Figure 10 Inhibition of protein degradation.

HeLa cells transiently overexpressing V5-tagged ZnT8 variants were treated (+) with chloroquine (100 μM) or MG132 (10 μM) or left untreated in medium absent these chemicals, over a 4-h incubation period, to inhibit lysosomal and proteasomal degradation, respectively. Protein blot analysis was performed using anti-V5 and anti-tubulin antibodies.

Supplementary information

Supplementary Text and Figures

Supplementary Tables 1–12, Supplementary Figures 1–10 and Supplementary Note (PDF 9205 kb)

Supplementary Data Set 1

Sequence read data for variants called from additional sequencing. (ZIP 205 kb)

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Flannick, J., Thorleifsson, G., Beer, N. et al. Loss-of-function mutations in SLC30A8 protect against type 2 diabetes. Nat Genet 46, 357–363 (2014).

Download citation

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing