The rs 4430796 SNP of the HNF 1 β gene associates with type 2 diabetes in older adults

INTRODUCTION: The impact of type 2 diabetes mellitus raises interest in understanding its evolutionary-genetic basis, to unveil yet unknown pathways that may have immediate medical relevance. The HNF1β gene (hepatocyte nuclear factor-1 beta) is a transcription factor expressed in tissues such as liver, kidney, genital tract and pancreas that is known to be essential for insulin secretion and glucose balance. We tested the association of allelic variants produced by the HNF1β gene (rs4430796) variation with the clinical and biochemical profile of elderly Brazilian outpatients with metabolic disorders. MATERIAL AND METHODS: Anthropometry, blood pressure, glycaemia, lipemia and other parameters were assessed in 184 Brazilians aged 60 or older in clinical care settings. Alleles were determined by amplification of the polymorphic site by real time polymerase chain reaction. RESULTS: Analysing variables across the genotypes, a statistically significant difference was noticed in the allele frequencies among diabetic patients, with 30.8% of the A homozygous bearing the condition compared to a prevalence of 12.2% between G homozygotes. CONCLUSION: Our results corroborate the possible protective property of the GG genotype from the rs4430796 variation (already presented in the literature) against occurrence of diabetes mellitus, which appears applicable to elderly individuals as well, even in the context of multiple metabolic disorders so typical in older Brazilians.

Diabetes is a group of metabolic diseases characterized by hyperglycaemia resulting from defects in insulin secretion, insulin action, or both 1 .In 2012, an estimated 1.5 million deaths were directly caused by diabetes and another 2.2 million deaths were attributable to high blood glucose 2 .Type 2 diabetes mellitus (DM2) is projected to be the seventh leading cause of death worldwide by 2030.This impact of DM2 has raised concerns and interest in under-REV ASSOC MED BRAS 2018; 64 (7):586-589 tial for normal glucose-stimulated insulin secretion from the analysis of β-cell-conditional HNF1β knockout mice 5 .Besides leading to an impaired glucose metabolism, selective deletion of HNF1β in β-cells leads to a variety of islet gene expression dysregulation.Some GWAS revealed that several tag SNPs in the HNF1β gene were associated with the susceptibility of DM2 and such associations were well replicated in many countries 6 Thus, HNF1β is prominent in the regulation of human insulin gene transcription 7 .
The purpose of this study was to investigate the association of allelic variants produced by the rs4430796 variation in the HNF1β gene with the glycaemic profile of Brazilian elderly patients as well as with their clinical, biochemical and inflammatory features.
For that, a sample of consecutive non-institutionalized patients aged 60 or older were enrolled between 2011 and 2013 at two general geriatrics outpatient clinics located at the metropolitan area of the Federal District, Brazil.The clinics were the Geriatrics Service of the Catholic University of Brasília (HUCB) and the Geriatrics Center of the University of Brasília (UnB).The main inclusion criterion was to spontaneously seek primary or secondary care for circulatory events.Exclusion to integrate these analyses were active inflammatory and/or infectious conditions, malignancies of any kind (current or past) or important kidney impairment (creatinine clearance < 25 ml/min/1.73m 2 ) coupled or not with abnormal titers of liver function markers.Patient enrolment was done consecutively, with no active search for particular conditions or events.The study was approved by the institutional research ethics committee and procedures were in accordance with the ethical standards of the Helsinki Declaration, with all participants having signed informed consent before assessments.
Diagnosis of type 2 diabetes included self-report of the condition confirmed by clinical aspect in anamnesis and/or by fasting glycosylated haemoglobin (Hb1Ac) ≥ 6.5% 1 .Practitioners of physical exercises were those exhibiting 30 minutes or over of exercises at any bout for at least four days a week 8 , while the smoking habit was defined as report of usual, active consumption of cigarettes 9 .Metabolic syndrome (MS) was identified according to NCEP-ATPIII criteria 10 .For biochemical analysis, the following measures were obtained according to routine clinical biochemistry and expressed in standard units: blood glucose (GLU), glycated haemoglobin (HbA1c), total cholesterol (CHL), and high density lipoprotein (HDL), triglycerides (TGL), C-reactive protein high sensitivity (CRP), thyroid stimulating hormone (TSH), Cockcroft-Gault creatinine clearance (CGault), gamaglutamiltranspeptidase (γGT), aspartate aminotransferase (AST) and alanine aminotransferase (ALT).
All subjects were submitted to assessments of total body mass (kg), body height (m), and blood pressure (mmHg).Body mass index (BMI; kg/m 2 ) was defined as usual whereas waist circumference (WC; cm) was measured 2 cm above the umbilicus scar.
Whole blood was obtained during sampling for  biochemical analysis and stored at -20°C until use.
Genomic DNA was purified according to standard extraction kits (QIAamp DNA Mini Kit, Qiagen, Brazil).

All the patients have been genotyped for the HNF1β
A/G transition (rs4430796).For the identification of the polymorphisms, we have used the The Kolmogorov-Smirnov test was used to test normal distribution of continuous variables through within-genotype approach.The Hardy-Weinberg equilibrium was tested using chi-square test.Data is expressed as means ± standard deviation (SD) or frequency (%).Analysis of variance (ANOVA) was used to compare means of the continuous variables (e.g.: age) across genotypes.Frequencies of categorical variables (e.g.: gender) across genotypes were compared using the chi-square test.All analyses were performed employing the Statistical Package for Social Sciences (SPSS) for Windows (version 13.0).A p value < 0.05 was considered significant.
Data of the 184 patients was analysed for anthropometric, clinical and metabolic features, with mean values and proportions in Table 1.The overall profile of the patients is compatible with an important frequency of metabolic disorders, with a high prevalence of diabetic (> 30%) dyslipidemic (> 50%) and hypertensive (> 75%) cases as well as of smokers (> 30%) and sedentary individuals (> 50%).
The frequencies of the HNF1β rs4430796 alleles in the sample were consistent with the Hardy-Weinberg equilibrium (p value < 0.05).A statistically significant difference was noticed in the frequency of diabetic patients across genotypes, with 30.8% of the A homozygous bearing the condition compared to a prevalence of 12.2% between G homozygotes (c 2 = 6.04; df = 2; p < 0.05).Genotypes of the HNF1β gene associated with no other features of the sample.
Given that the HNF1β gene, located on 17q21.3,encodes a transcription factor involved in tissue-specific regulation of gene expression and embryonic development of numerous organs 4, 11 , and that a total of 30 heterozygous mutations in the HNF1β gene have been described so far (including missense, nonsense, frameshift, insertion/deletions, and splice site mutations) 12 , we reiterate that our investigation focused on the association of a particular A/G allelic variation (rs4430796) with a range of clinical and biochemical variables in elderly patients, having found evidence for an specific association of the corresponding GG genotype with the occurrence of T2DM.Such an association has already been described elsewhere, as follows.
The major G allele of the rs4430796 SNP of the HNF1β gene was associated with decreased risk of T2DM in physically active in contrast to the increased risk in sedentary, pointing out the complexity of the relationship between the gene and diabetes 13 .The evidence for an association is reinforced by the genome-wide association study (GWAS) with a Chinese Hans sample which confirm the association of the HNF1β gene rs4430796 with T2DM (OR = 1.19; p = 1.52 x 10 -11 ) 14 .It should be considered that the Brazilian population was formed by an admixture of races 15 .In ancestry-specific GWAS meta-analyses, the rs4430796 polymorphism was also confirmed in association with T2DM (Cochran's Q p-value 3,6 x 10 -1 ; p = 8.9 x 10 -9 ) 16 , as well as in another GWAS that validated this association in European, African and Asian backgrounds and found protective properties of the rs4430796 polymorphism against T2DM (OR= 0.91; p = 2.7 x 10 -7 ) 17 .
To our knowledge, this study is the first report to demonstrate that a polymorphism in the HNF1β gene is associated with T2DM in Brazilian elderly patients.In clinical terms, the sample characterization revealed significant prevalence of metabolic disorders compatible with a profile eligible for primary or secondary prevention of vascular conditions.Even so, the analyses showed no association of genotypes with other basic variables apart from the glycaemic traits, rendering this relationship as plausible even in the context of multiple metabolic disorders, typical of older Brazilians 18 .Unfortunately, sample size severely limited statistical power for subgroup analysis.Further thorough investigations with larger sample sizes could confirm our findings.
In summary, we tested the association of allelic variants from the rs4430796 SNP with the clinical, biochemical and inflammatory profile in elderly patients.Our findings suggest a protective character for the GG genotype in the development of T2DM, which appears applicable to elderly individuals as well.Whether this information will prove useful for targeted prevention or treatment of type 2 diabetes remains to be determined.

TABLE 1 .
AVERAGE Data shown as mean ± SD or frequency within genotype.* p < 0.05 when compared to any other genotypic group.BMI = body mass