This was a case-control study involving 85 T2DM Patients and 95 controls of Cameroonian origin, aged 24 and over. Patients with T2DM (with glutamic acid decarboxylase (GAD) antibody negativity) were enrolled in the study at the National Center for Obesity at the Yaoundé Central Hospital and normoglycemic controls (with no personal or family history of diabetes) with fasting plasma glucose level less than 1.1 g/L from the general population after providing detailed information on the objectives and procedures of the study. Physical examination was conducted with data collection on sex, age and clinical parameters were collected for all participants. Height, waist circumference and hip circumference to the nearest 0.5 cm, and weight to the nearest 0.1 kg, were measured. The Body Mass Index (BMI) and the Waist-to-Hip Ratio (WHR) were defined according to the International Diabetes Foundation (IDF) criteria. Blood pressure was the mean of two measurements performed at least three minutes apart on the right arm, with the subject seated after a 15 min rest, using an automatic Omron HEM-705 CP sphygmomanometer recorder (OMRON^® MX2 Basic, OMRON HEALTHCARE, INC. Bannockburn, Illinois 60015.CHINA) with standardized procedures. The percentage of fat mass was measured using bioimpedance (TANITA^® BC 420 MA, TANITA Corporation 1-14-2 Maeno-cho, Tabashi-ku, Tokyo Japan). Diagnosis of T2DM was based on a history of doctor-diagnosis that was conducted according to the American Diabetes Association (ADA) and World Health Organization (WHO) definitions [20] [21]: FPG concentration ≥ 7.0 mmol/L (or ≥126 mg/dL) and/or 2-hour post-Oral Glucose Tolerance Test (OGTT) plasma glucose ≥ 11.1 mmol/L (or ≥ 200 mg/dL). Alternatively, this was accepted if an individual was on pharmacological treatment for T2DM and a review of medical records indicated adequate justification for that therapy. Added to this, the detection of autoantibodies to GAD antibody was used to exclude probable cases of type 1 diabetes. Glycated haemoglobin (HbA1c) was measured by High Performance Liquid Chromatography (HPLC) with an Automated Biochemistry Analyzer Bio-Rad in2it TM (I). The Homeostatic Model Assessment Insulin Resistance (HOMA-IR) and functional β-cells (HOMA-B) were estimated using the formula: HOMA-IR = [fasting insulin concentration (mU/L) × FPG (mmol/L]/22.5] [22] (with Insulin resistance defined by any value of the HOMA-IR index equal to or above 2.1) [23] and 0.27 × fasting C-peptide (μU/ml)/fasting glucose (mmol/ml) − 3.5 respectively [22].

Two venous blood samples were aseptically collected in EDTA and plain tubes in the morning following an overnight fasting for at least 12 hours. They were kept on ice (4˚C) and immediately transported to the laboratory where serum specimen was separated accordingly from the plain tube sample by centrifuging for 5 minutes at 3000 rpm. The extraction of genomic DNA was done from the dried blood stains on the Whatmann filter paper (from EDTA tube sample) by the Chelex method [24] and stored at −20˚C for further analysis.

FPG levels were measured by the glucose oxidase–peroxidase method using an automatic portable Life Scan OneTouch Ultra glucometer (Johnson & Johnson, USA). Total cholesterol (TC) (cholesterol oxidase phenol-4-amino antipyrene peroxidase method, derived from that described by Allain and collaborators) [25], serum triglycerides (TG) (glycerol phosphatase oxidase-phenol4-amino antipyrene peroxidase method, derived from that described by Buccolo and Davids) [26], and high-density lipoprotein (HDL)-cholesterol (cholesterol oxidase phenol-4-amino antipyrene peroxidase method, in the supernatant, derived from that described by Burstein and collaborators) [27] were measured by standardized enzymatic methods on a spectrophotometer (UV Mini 1240) using Chronolab kits. The Friedwald formula was used to calculate Low-density lipoprotein (LDL)-cholesterol concentrations, when TG levels were below 4 mmol/L, and measured when TG values were over 4 mmol/L [28]. Dyslipidemia was considered as an elevated level of TC (>6.2 mmol/L) and/or an elevated level of LDL-cholesterol (>4.1 mmol/L) and/or a low HDL-cholesterol level (<1.04 mmol/L in men and 1.29 mmol/L in women) and/or an elevated level of TG (≥1.7 mmol/L) [29].

Genomic DNA was extracted from filter paper stained with whole blood by the Chelex method. One hundred and eighty (180) participants (85 type 2 diabetic patients and 95 healthy controls) were genotyped for KLF14 rs4731702 (C/T) by PCR-RFLP. The KLF14 rs4731702 (C/T) polymorphism was genotyped using its corresponding number and size of appearing bands. PCR was performed with these Forward primer 5’-AATCCCAAGGCATCTATC-3’ and Reverse primer 5’-CTTGGATTTTGATTACGG-3’ (SIGMA-ALDRICH, United States). For the amplification, each 25 μL PCR reaction mixture consisted of 3 μL of genomic DNA, 0.25 μL of each primer (10 pmol/L), 12.5 μL of Go Taq Green PCR Master Mix (20 mM Tris-HCl, pH 8.3; 100 mM KCl; 3 mM MgCl2; 0.1 U Taq polymerase/μL; 500 μM dNTP each QIAGEN) and 9 μL of ddH2O (DNase/RNase-free). The PCR was carried out under the following conditions: initial denaturation at 94˚C for 5 minutes, followed by 35 cycles of denaturation at 94˚C for 45 seconds, primer annealing at 53˚C for 45 seconds, primer extension at 72˚C for 45 seconds and a final extension at 72˚C for 7 minutes. The amplicons were analyzed by agarose gel electrophoresis using a 2% agarose gel and 6 μL of positive amplification products (347 bp) digested with 5U of Bacillus stearothermophilus I (BsrI) restriction enzyme at 65˚C overnight. Agarose gels (3%) stained with ethidium bromide were used to separate the digested amplicons which were then visualized under a UV lamp transilluminator. This restriction enzyme digestion produced three (3) bands of different sizes: a 347 bp corresponding to the TT genotype (restriction site absent); a set of 133 bp and 214 bp corresponding to the CC genotype (restriction site present) and a set of 133 bp, 214 bp and 347 bp corresponding to CT genotype (restriction site present and absent). To ensure that the genotyping was of good quality, all gels were reread blindly by another person without any change and random duplicates were performed in about 10% of the samples in each genotyping assay. No genotype errors were detected in the random duplicates.

This was performed prior to the initiation of the study. A sample size of 80 T2DM patients and 80 healthy controls was required to detect the polymorphism with an effect size of 1.5 for significance threshold of 0.05 and an expected allele frequency of 0.10 in the general population, with statistical power of 80% (α = 0.05 and β = 0.20) at this locus. The Quanto statistical software was used to calculate the power and Sample size [30].

Participant data obtained were entered into a log book, and later keyed into a computer using the 2016 version of Microsoft excel sheets and verified for the possibility of entering errors. Genotype and allele distributions in patients with T2DM and healthy controls were obtained by direct counting. Qualitative variables were analyzed by the Chi square (χ²) test or Fisher’s exact test. Quantitative variables were compared using non-parametric tests (Mann Whitney or Kruskall Wallis with post hoc multiple comparison by Dunn-Sidak test) and described using median and inter quartile domain (IQR, 25th-75th percentiles) or mean ± standard deviation (M ± SD), and categorical variables using their frequency and percentage. The genotype frequencies were tested for the Hardy-Weinberg equilibrium using a χ² test. Odd Ratios (OR) with 95% confidence intervals (CI), were calculated using unconditional logistic regression. Data were coded, entered and analyzed using the Statistical Package for Social Sciences (SPSS) version 20.0 for Windows (Chicago, Illinois, USA) and were considered as statistically significant when p < 0.05.

The characteristics of the studied population are presented in Table 1. There were no significant difference distributions between diabetic patients and normoglycemic controls for gender (male/female ratio 0.89 vs. 0.73; p = 0.5494), age

Notes: Data are medians (interquartile range) unless otherwise stated. Abbreviations: BMI, Body Mass Index; WHR, Waist to Hip Ratio; SBP, Systolic Blood Pressure; DBP, Diastolic Blood Pressure; FPG, Fasting Plasma Glucose; TC, Total Cholesterol; TG, Triglycerides; HDL-C, High Density Lipoprotein cholesterol; LDL-C, Low Density Lipoprotein Cholesterol; AI, Atherogenic Index = LDL cholesterol/HDL-cholesterol.

(median age 50.00 years vs. 49.00; p = 0.1599) and BMI (median age 27.56 years vs. 27.51; p = 0.8042). On the contrary, for WHR (median value 0.90 vs. 0.85; p < 0.001), FPG (median level 3.44 vs. 0.92; p < 0.0001), TC (median level 4.13 vs. 1.83; p < 0.0001), HDL-C (median level 0.83 vs. 0.49; p < 0.0001) and LDL-C (median level 2.42 vs. 1.05; p < 0.0001), the same significant difference distributions were observed between diabetic patients and normoglycemic controls respectively. HbA1c (median value 8.6%), HOMA (median value 9.04) and β-cell functions (median value 0.01) were assessed only for diabetic patients.

All T2DM patients and normoglycemic controls were positively genotyped for the KLF14 rs4731702(C/T) gene polymorphism. After amplification characterized on agarose gel by one band of 347 bp (Figure 1), the genotyping was characterized on agarose gel by two bands of 214 bp and 133 bp for the wild type homozygote CC, one band of 347 bp for the mutant homozygote TT, and three bands of 347 pb, 214 pb and 133 bp for the mutant heterozygote CT (Figure 2).

In the general population, the frequency of the CC genotype was 46.11% (83/180), vs. 20.00% (36/180) and 33.89% (61/180) for the CT and the TT genotypes respectively.

This matching population was not in Hardy-Weinberg equilibrium (χ² = 34.76; p < 0.0001). The C allele was major with a frequency of 56.11%, as compared to the minor T allele which showed a frequency of 43.89%.

The KLF14 rs4731702 (C/T) gene polymorphism was strongly associated with T2DM, with OR = 4.877 (95% CI [3.112 - 7.644] and p < 0.0001) for the T allele and OR = 2.721 (95% CI [1.263 - 5.865] and p = 0.0145), OR = 3.907 (95% CI [2.022 - 7.549] and p < 0.0001) for the CT and TT genotypes respectively. The T allele (minor) was demonstrated to be the risk allele (63.53% diabetic patients vs. 26.32% healthy controls, OR = 4.877 and p < 0.0001) while the C allele was demonstrated to be the protective allele (36.47% diabetic patients vs. 73.68% healthy controls, OR = 0.205 and p < 0.0001). The CC genotype was protective (22.35% diabetic patients vs. 67.37% healthy controls, OR = 0.139 and p < 0.0001) against the disease (Table 2).

According to the rs4731702 (C/T) genotypes, the genotype-phenotype correlations were performed. The clinical (BMI, WHR, SBP, DBP and Fat Mass percentage) and biochemical (FPG, TC, LDL-C, HDL-C, TG, AI (LDL-C/HDL-C), HbA1c (%), HOMA-β (Pep-C) and HOMA-IR) characteristics of the total population were stratified (Table 3). The median FPG was significantly different (p < 0.0001) in the three (CC, CT and TT) genotype groups with the higher values in the individuals presenting the CT genotype. The medians concentrations of TC (p < 0.0001), HDL-C (p = 0.0004) and LDL-C (p < 0.0001) were significantly

Notes: The results are given as n (%). Abbreviations: KLF14, Kruppel Like Factor 14; OR, Odds Ratio; CI, Confidence Interval.

Notes: Data are medians (interquartile range) unless otherwise stated. Abbreviations: BMI, Body Mass Index; WHR, Waist to Hip Ratio; SBP, Systolic Blood Pressure; DBP, Diastolic Blood Pressure; FPG, Fasting Plasma Glucose; TC, Total Cholesterol; TG, Triglycerides; HDL-C, High Density Lipoprotein cholesterol; LDL-C, Low Density Lipoprotein Cholesterol; AI, Atherogenic Index = LDL Cholesterol/HDL-cholesterol; HbA1C, glycated haemoglobin; HOMA-IR, Homeostasis Model Assessment of Insulin Resistance; HOMA-β, β-cell functions; (a), only for the 85 type 2 diabetic patients.

different with the higher values in the individuals presenting the TT genotype. As for the KLF14 rs4731702 (C/T) gene polymorphism, no significant genotype-phenotype characteristics were observed in the control group and also in a specific phenotype of T2DM (data not shown).