KLF15 SNP rs9838915 A allele is associated with increased LV mass in patients with 2 diabetes.
KLF15 SNP rs9838915 predicts incident heart failure hospitalization.
Genotyping the KLF15 SNP rs9838915 A allele allowed more precise stratification of the risk of heart failure hospitalization in patients already at high risk due to the presence of echocardiographic LVH.
Left ventricular hypertrophy (LVH) is a heritable trait that is common in patients with diabetes. The Kruppel like factor 15 (KLF15) is expressed in the heart and acts as a repressor of cardiac hypertrophy and fibrosis. Our study provides evidence that genetic variation in KLF15 is associated with LVH in patients with type 2 diabetes and these findings were then replicated in an independent cohort of patients with type 2 diabetes. The KLF15 genetic variant was also associated with first heart failure hospitalization. These findings add to our understanding of the molecular mechanisms that contribute to increased LV mass.
Left ventricular (LV) hypertrophy (LVH) is a heritable trait that is common in type 2 diabetes and is associated with the development of heart failure. The transcriptional factor Kruppel like factor 15 (KLF15) is expressed in the heart and acts as a repressor of cardiac hypertrophy in experimental models. This study investigated if KLF15 gene variants were associated with LVH in type 2 diabetes. In stage 1 of a 2-stage approach, patients with type 2 diabetes and no known cardiac disease were prospectively recruited for a transthoracic echocardiographic assessment (Melbourne Diabetes Heart Cohort) (n = 318) and genotyping of two KLF15 single nucleotide polymorphisms (SNPs) (rs9838915, rs6796325). In stage 2, the association of KLF15 SNPs with LVH was investigated in the Genetics of Diabetes Audit and Research in Tayside Scotland (Go-DARTS) type 2 diabetes cohort (n = 5631). The KLF15 SNP rs9838915 A allele was associated in a dominant manner with LV mass before (P = 0.003) and after (P = 0.001) adjustment for age, gender, BMI and hypertension, and with adjusted septal (P < 0.0001) and posterior (P = 0.004) wall thickness. LVH was present in 35% of patients. Over a median follow up of 5.6 years, there were 22 (7%) first heart failure hospitalizations. The adjusted risk of heart failure hospitalization was 5.5-fold greater in those with LVH and the rs9838915 A allele compared to those without LVH and the GG genotype (HR 5.5 (1.6–18.6), P = 0.006). The association of rs9838915 A allele with LVH was replicated in the Go-DARTS cohort. We have identified the KLF15 SNP rs9838915 A allele as a marker of LVH in patients with type 2 diabetes, and replicated these findings in a large independent cohort. Studies are needed to characterize the functional importance of these results, and to determine if the SNP rs9838915 A allele is associated with LVH in other high risk patient cohorts.
This clinical study investigated the association between the KLF15 gene and LVH in patients with type 2 diabetes. A key finding in the discovery cohort was that the A allele at rs9838915 SNP in the KLF15 gene was associated with increased LV mass in patients with type 2 diabetes. Carriers of the AA genotype had a 14 g/m2 increase in LV mass compared to those with the GG genotype that was independent of age, gender, BMI and hypertension. Furthermore, we replicated the association of the KLF15 SNP rs9838915 A allele with LVH in a large, independent cohort of patients with type 2 diabetes, the Go-DARTS cohort (n = 5631). We found no association of the KLF15 rs6796325 T/C SNP with any echocardiographic parameter in either the discovery or the replication cohorts.
In a preliminary analysis, we explored the association between LVH, KLF15 genotype and heart failure outcomes in the discovery cohort. Echocardiographic LVH was present in 35% of patients and was associated with a 3-fold increased risk of first heart failure hospitalization that was independent of age, gender, systolic blood pressure, BMI and hypertension. The adjusted risk of heart failure hospitalization was 5.5-fold greater in those with LVH and the KLF15 rs9838915 A allele compared to the group with no LVH and the GG genotype. These results, albeit in a small cohort, suggest that patients with type 2 diabetes and LVH may be able to be more precisely stratified for risk of the development of heart failure according to their KLF15 rs9838915 genotype.
Diabetes is an independent predictor of LVH, but not all patients with diabetes develop LVH, suggesting that there is a genetic component to LV mass and the development of LVH. In this study, patients had relatively well controlled diabetes (HbA1c ~ 7–7.5%) with ~ 90% on glucose lowering therapy, but despite this a significant number had LVH. The genetic basis of LVH has been mainly studied in the general population, whereas we specifically chose to study patients at high risk of LVH due to the presence of diabetes. GWAS has identified some genetic loci associated with electrocardiographic or echocardiographic LVH (van der Harst et al., 2016) but none of the SNPs reached genome wide significance on chromosome 3, which is the location of the KLF15 gene ( Vasan et al., 2009). The EchoGen study (n = 12,612) investigated associations with echocardiographic LV mass in 5 community-based cohorts of European ancestry (Vasan et al., 2009) and Shah et al. investigated associations with ECG-LVH in 10,258 individuals in 3 population based cohorts (Shah et al., 2011). Neither study was enriched for risk factors for LVH such as hypertension, obesity and diabetes. A recent meta-analysis of GWAS in a much larger cohort of 73,518 European ancestry individuals identified 52 genomic loci associating with 4 ECG measured QRS complex phenotypes (van der Harst et al., 2016).
It is unknown how rs9838915 SNP located in intron 2 influences LV mass. KLF15 is expressed in cardiomyocytes and acts as a repressor of pathological cardiac hypertrophy (Wang et al., 2008; Fisch et al., 2007; Haldar et al., 2010 ; Leenders et al., 2012) through inhibition of the cardiac transcriptional factors GATA4 and MEF2 (Fisch et al., 2007). One possibility is that genetic variation disrupts the ability of KLF15 to repress hypertrophic transcriptional factors, leading to increased expression of these genes and thus cardiac hypertrophy. KLF15 is also a transcriptional inhibitor of cardiac fibrosis (Wang et al., 2008), which is a key feature of diabetic heart disease (Rubler et al., 1972). It is also possible that the observed genetic variation in KLF15 could lead to increased cardiac fibrosis which would in turn contribute to increased LV mass and myocardial stiffness. Further studies are required to determine whether the association we observe is due to SNP rs9838915 or a neighboring SNP that may be in strong linkage disequilibrium with rs9838915. We performed a bioinformatics analysis to identify putative causal variants underlying the association of the KLF15 SNP rs9838915 with LVH. We explored the co-localisation of the association signal with features indicative of functional genomic elements, including evidence of transcription factor binding, DNase hypersensitivity and histone modification marks. An analysis of the RegulomeDB suggests that SNP rs9838915 is located in a transcription factor binding site and an enhancer element. These functional elements were identified in human LV tissue. We used the SNAP tool (http://www.broadinstitute.org/mpg/snap) by the Broad Institute to identify SNPs that were in LD (r2 > 0.8) with SNP rs9838915. We identified a SNP located 635 bp upstream of rs9838915 that was highly correlated (r2 = 1.0) and located within an enhancer element identified in human LV, right atrium and ventricle tissue.
We identified a gene variant, the rs9838915 SNP in the KLF15 gene that is relevant to increased LV mass in 318 patients with type 2 diabetes, and validated these findings in a large independent cohort of > 5000 individuals with type 2 diabetes. Studies are now needed to characterize the functional importance of these results, to understand the biological mechanisms involved, and to determine if the KLF15 SNP rs9838915 A allele is associated with LVH in patients without diabetes.