The Krüppel-like factor (KLF) proteins belong to a family of transcription factors that bind to GC-GT rich sites and CACC boxes of a large range of gene promoters 1234. Binding of these factors is mediated by a highly conserved DNA-binding motif of C2H2 zinc fingers localised at their C-terminus 5, which suggests that these proteins could have a redundant activity. However, accumulated evidence shows that each KLF has its own function and regulation 67. KLF transcription factors act as transcriptional activators or repressors depending on the specificity of promoters to which they bind and the cellular context. Moreover, they interact with various co-activators or co-repressors, regulating via different mechanisms, the transcription of their target genes 4.

We recently identified both rare and frequent genetic variants in KLF11 co-segregating with early onset familial diabetes or associated with late-onset T2D 8. These results prompted us to assess the putative genetic contribution to T2D risk of additional KLF family members, especially those presumably involved in energy and glucose homeostasis, through their expression in the pancreas, adipose tissue, liver and muscle 91011. Indeed, KLF2, KLF5, and KLF6 have been suggested to be involved in adipocyte differentiation 121314; KLF2 and KLF15 may contribute to adipogenesis transcriptional regulation via PPARγ signalling 1516. In particular, this implication in the differentiation process of adipocytes, suggests they may play an important role in insulin resistance, a status prone to T2D development.

Thus, the possible implication of the KLF family in glucose homeostasis and adipogenesis makes its members relevant gene candidates for T2D. Therefore, we assessed if frequent genetic variants in several members of the KLF family are implicated in genetic susceptibility to T2D development in a French population. To assess the putative contribution of genetic variability of KLFs gene family members to type 2 diabetes (T2D) we performed a two-stage association study.

In the first case-control set, we genotyped 28 frequent variants of 9 different members of the KLF gene family with a minor allele frequency (MAF) above 0.10 (table 2). All SNPs were in Hardy-Weinberg equilibrium according to the FINETTI program (data not shown). In table 2, we present the NCBI reference number (dbSNP ID) and details for each SNP genotyped, as well as the MAF and the P-value under an allelic model. From this first study set (D1/C1), three SNPs showed an association with T2D; KLF2 rs12459387 (P = 0.09, MAF = 0.12, OR = 0.75 with 95% CI = 0.53–1.05), KLF4 rs10759240 (P = 0.08, MAF = 0.52, OR = 0.83 with 95% CI = 0.67–1.02) and KLF5 rs11841945 (P = 0.04, MAF = 0.40, OR = 1.27 with 95% CI = 1.02–1.59).

In the second case-control set, we genotyped KLF2 rs12459387, KLF4 rs10759240 and KLF5 rs11841945, but the observed difference in MAF between cases and controls in D1/C1 was not confirmed (Table 2). In fact, the trend for allelic association with T2D observed in the first set was not replicated for these SNPs in the second set (D2/C2), nor were they significant after the combined analysis of both groups (D/C). As D2 had a higher mean body mass index (BMI) compared to D1, we also analyzed the data of individuals with BMI < 27 kg/m² but still no significant difference was observed. In addition, we did not observe associations between these three SNPs and quantitative traits such as fasting insulin, fasting glucose, OGTT.

Since our study is probably limited by the fact that it does not cover all the genetic variability of the KLF genes, we analysed additional SNPs included in our recently published genome wide association study 20. The SNPs located in regions spanning the KLF genes, and analyzed in this genome wide association study (148) showed no significant association with T2D in this French population (data not shown).

We have analysed SNPs within nine KLF genes and we were not able to show a genetic T2D association in our population. A similar study in a Japanese population also showed no evidence for an association with other variants of the same KLF genes analysed in our study (KLF2, -3, -5, -12, -13 and KLF16 21). However, it is worth noting that some KLF genes are not covered extensively with HapMap genotyped SNPs. In fact, this was the case for KLF15 and KLF16 gene regions, where at the start of this study no Taqman^® assays were available. Therefore, we decided to screen these genes for common SNPs. Since we analysed in our study all frequent SNPs without any positive results, we conclude that the KLF15 and KLF16 genes are not likely to be associated with T2D in a French population.

A limitation of our study is the two-staged design, through which we may have missed a weak but true association with T2D. However, the use of T2D patients with affected first degree relatives and controls sharing the same environment as the cases in set 1, probably increased our power to detect an association with T2D. A second limitation of this study is the fact that it does not cover all the genetic variability of the KLF genes. However, our subsequent genome wide association analysis with 1,363 individuals also did not show significant T2D associations for variants of these genes, making it unlikely the nine KLFs play a major role in genetic T2D susceptibility.

We hypothesised that KLF genes could be associated with genetic susceptibility for T2D, based on their role in adipogenesis. In particularly, KLF2 is implicated in the regulation of PPARγ 15, which is a target gene implicated in the genetic susceptibility for T2D. Therefore, we studied this gene more extensively and genotyped 9 SNPs. Also for KLF2 variants, we did not observe an association with T2D in this French population. In addition, a previous study of KLF2 variants and body weight control in another French population did not observe an association with adipose tissue-related traits 22. From these results, it seems unlikely that the genetic variability of the KLF2 gene plays a mayor role in susceptibility to T2D or obesity.

In our study of variants from nine different members of the KLF gene family, we observed no association with T2D in our large French population. Moreover, the genotyped SNPs covered the overall genetic variability of the KLF2, -15 and -16 genes so we may conclude that these genes do not contribute to the genetic susceptibility of T2D.

The author(s) declare that they have no competing interests.

RGA wrote the paper. RGA and BN performed the statistical analyses. YB and EV genotyped the variants of this study. RS provided data and statistical analyses of the genome wide association study. BB, MM, GC and PF established the diverse DNA collection used. PF and BN directed the study and contributed to the redaction of the paper.