Cytosolic phosphoenolpyruvate carboxykinase (PEPCK-C), encoded by the PCK1 gene in humans, is an enzyme centrally involved in gluconeogenesis, glyceroneogenesis and cataplerosis. Normal expression of PCK1 is under hormonal control, regulated at the transcriptional level by both activators, such as glucagon, and inhibitors, such as insulin. The metabolic functions of PEPCK-C advocate the PCK1 gene as a strong candidate for conferring susceptibility to type 2 diabetes (T2D), a theory that is supported by the effects of gene modulation in mouse and rat models. These include over-expression of PEPCK-C resulting in insulin resistance and T2D [1].

Genetic studies have previously implicated the region of chromosome 20q in which PCK1 lies in T2D susceptibility [2, 3]. Cao et al. [4] reported on the discovery of a single nucleotide polymorphism (SNP) in the promoter region of PCK1 (-232C/G) that was associated with T2D in Canadian Caucasian and Oji-Cree cohorts (odds ratio (OR) = 2.8, 95% CI 1.7 - 4.7, p = 4 × 10^-5 and OR = 1.9, 95% CI 1.2 - 3.0, p = 9 × 10^-3 respectively). In addition, luciferase reporter assays demonstrated that the -232G allele increased expression of PCK1 compared to the -232C allele in multiple cell lines, with no down regulation by insulin. As with many genetic association studies investigating polygenic diseases, attempts to replicate the association between PCK1 and T2D have produced mixed results. A haplotype of PCK1 variants was shown to confer risk of developing the disease in a Korean population (OR not given, p = 6 × 10^-3) [5], and a screen of 134 candidate susceptibility SNPs showed that the -232C/G SNP was a risk factor for T2D in a Finnish cohort (OR = 1.27, 95% CI 1.02 - 1.57, p = 0.031) [6]. A recent study reported that multiple PCK1 variants are associated with T2D in a Chinese population [7]. The authors of this paper reported -232C as a risk allele, although its association with the disease fell short of statistical significance (OR = 1.24, 95% CI 1.00 - 1.55, p = 0.057). Studies investigating other populations, however, have found no evidence for an association between PCK1 variants and T2D [8, 9].

Within the UK, T2D is 4 to 6 times more common in the South Asian population compared to the indigenous white Caucasian population [10]. Over 10% of South Asian adults will develop the disease and yet only a small number of investigations have reported on the genetic susceptibility to T2D in this ethnic group. None of these studies has looked into the possible effects of the PCK1 gene. We therefore aimed to investigate the association between common variants of the PCK1 gene and T2D in a UK-resident South Asian population of Punjabi ancestry.

Type 2 diabetic subjects (N = 903) were recruited to the United Kingdom Asian Diabetes Study (UKADS), a multiple risk factor intervention trial investigating the impact of a culturally-sensitive, enhanced diabetes care package on the risk of cardiovascular disease in South Asian type 2 diabetes patients living in Birmingham and Coventry, UK [11]. All subjects were of Punjabi ancestry, confirmed over three generations, and originated predominantly from the Mirpur area of Azad Kashmir, Pakistan. Ethnically-matched normoglycaemic control subjects (N = 471) were recruited from the same geographical areas through community screening. Normal glucose tolerance was defined as fasting plasma glucose <6 mmol/l and 2 hr plasma glucose <7.8 mmol/l on a 75 g OGTT. Where OGTT was not feasible, normal glucose tolerance was defined as random blood glucose <7 mmol/l. Venous blood was collected from each subject after obtaining informed consent and genomic DNA extracted using an adaptation of the Nucleon^® protocol (Nucleon Biosciences, Coatbridge, UK). The study was approved by the Birmingham East, North and Solihull Research Ethics Committee.

LD patterns (Figure 1) and allele frequencies for all variants were generally similar to those seen in the CEPH (CEU) HapMap samples (CEU allele frequencies: rs2071023 (G) - 0.52, rs6070157 (T) - 0.21, rs2070756 (T) - 0.25, rs2179706 (T) - 0.52, rs1042531 (G) - 0.31). Genotyping success rate was ≥ 97.5% for all SNPs studied. Approximately 15% of all individuals were re-genotyped for the estimation of error rate, which was <1% for all variants. All SNPs conformed to Hardy-Weinberg equilibrium with the exception of rs6070157 (p = 0.01). As the error rate for this SNP was zero and it displayed no significant association with any variable measured, no further action was taken to investigate this anomaly.

Including young control subjects within a case-control analysis can artificially reduce effect size and statistical significance, as it can increase the chance of including subjects who will develop T2D later in life. As we have done previously with TCF7L2 [15], we re-analyzed our data using subsets of the control group defined by different minimum age cut-offs. For SNP rs2071023 there was a significant relationship between control-group minimum age cut-off and both OR (r² = 0.912, p = 3.86 × 10^-7) and statistical significance of the logistic regression test (r² = 0.838, p = 1.12 × 10^-5), up until maximum statistical significance (minimum p-value) was reached at an age cut-off of 47 years. At this age cut-off the effect size of the association had greatly increased (OR = 1.31, 95% CI 1.10 - 1.56, p = 2 × 10^-3), remaining significant even after correcting for multiple testing (Bonferroni correction for testing 5 SNPs, p = 0.015). After this age cut-off both relationships began to deteriorate as the number of individuals within the control group was further reduced. It is interesting to note that when an age cut-off of 47 years was applied to the control group, statistical power actually increased to 90% due to the increase in OR, despite a drop in subject numbers.

As previously reported in a number of studies[4, 6] our results suggest that the -232C/G SNP (rs2071023) located in the promoter region of the PEPCK-C-encoding PCK1 gene is associated with T2D.

It has been suggested that T2D could be caused by either excessive PEPCK-C production in the liver or reduced levels of PEPCK-C in adipose tissue [16]. In addition, from expression analysis of luciferase reporter constructs in multiple cell lines, Cao et al. [4] demonstrated that the -232G risk allele resulted in increased basal gene expression when compared to the -232C allele. It is possible, therefore, that the -232C/G polymorphism may confer increased risk of T2D development by increasing PCK1 expression in the liver. This would result in an upregulation of gluconeogenesis and increased blood glucose levels. Unfortunately we were not able to investigate the relationship between PCK1 genotype and blood glucose levels in this study, as fasting blood glucose data were only available for a small subset of our control subjects. Interestingly, a recent study has shown that liver-specific silencing of Pck1 can improve glycaemic control and insulin sensitivity in a T2D mouse model [17], supporting the role of PEPCK-C in T2D pathology and providing a potential therapeutic target for treatment of the disease.

Although the use of our entire cohort resulted in an OR of 1.21, the removal of young control subjects should have increased the validity of the control group. The OR of 1.31 resulting from our reduced dataset, similar to that seen in a Finnish cohort (OR = 1.27) [6], may therefore be a more representative estimate of the true effect size of this SNP, one that is not inconsiderable compared to recently discovered T2D susceptibility variants.

There are a number of limiting factors to be considered when interpreting our results. Firstly, our cohort is limited in size. Secondly, there is the possibility of cryptic relatedness within our cohort as the study subjects were recruited from a relatively small migrant population. In addition to this consanguinity is relatively common in the Mirpuri population. These limitations increase the risk of our result being a false positive. Unfortunately we cannot control for relatedness in our analyses as we do not have the necessary data. The impact of cryptic relatedness on association studies, however, increases with sample size [18], so it may be that our relatively small cohort is to some degree protected from this effect. Furthermore, we have no reason to believe that the degree of relatedness would differ significantly between the diabetic and control groups and so the effect of any cryptic relatedness may be negated.

This study is the first to investigate the association between variants of the PCK1 gene and T2D in South Asians. In agreement with studies in other ethnic groups [4, 6], our analyses suggest that the -232C/G promoter SNP (rs2071023) confers susceptibility to T2D. Due to the limitations discussed within this manuscript, however, we cannot exclude the possibility that our findings are a false positive result. We strongly advocate that replication of this association in larger independent South Asian cohorts is needed to confirm the -232C/G polymorphism of PCK1 as a true T2D susceptibility variant in this ethnic group.