Association of VDR gene ApaI polymorphism with obesity in Iranian population.

INTRODUCTION: Identifying obesity risk factors as a health problem facing communities is crucial given its complexity. The vitamin D receptor gene has been reported as a possible cause of this disease.
OBJECTIVE: To study the association of the VDR gene ApaI, BsmI, and TaqI polymorphisms with obesity in an Iranian population.
MATERIAL AND METHODS: We analyzed the genotypes of 348 obese (BMI≥30 kg/m2) and 320 non-obese people (BMI: 18.5-24.9 kg/m2) using PCR-RFLP. We measured FBS, TG, total cholesterol, and HDL and LDL cholesterol levels in an automatic biochemical analyzer.
RESULTS: We found significantly higher BMI, FBS, and TG levels in the obese group compared to the control. In the obese individuals, the frequency of genotype AA was 47.1% and that of the combined Aa+aa genotype, 52.9% while in the control group they were 30% and 70%, respectively (p=0.024, 95% confidence interval (CI)=1.100-3.933, odds ratio (OR)=2.08). A and a alleles frequencies for the ApaI polymorphism were statistically significant in the two groups (allele A vs. a; p=0.017). No significant relationship was observed between TaqI genotypes and alleles in the control and obese subjects.
CONCLUSION: We found that VDR ApaI (rs7975232 C/A) polymorphism appeared to be a risk factor for obesity. Especially, the A allele and the AA genotype in ApaI were associated with the obesity phenotypes.

Obesity is a complex disease influenced by environmental and genetic factors, which consists of the excessive accumulation of fats in adipose tissues as a result of an imbalance between energy consumption and intake. In the past 50 years, unhealthy habits in lifestyle have resulted in the so-called “obesogenic” environment, i.e., the intake of easily accessible energy-dense foods coupled with decreased physical activities. Researchers have suggested that genetic factors play a key role in regulating body weight . Obesity can cause many disorders, including cancer, cardiovascular disease, impaired glucose tolerance, hypertension, type 2 diabetes, sleep apnea, osteoarthritis, and gallbladder and liver disorders The body mass index (BMI) has been commonly used as a surrogate marker of excessive body fats in the absence of accurate yet simple techniques for measuring them . Identifying the genetic factors contributing to the risk of obesity may help broaden the basic biological knowledge about the energy imbalance and determine the pathways and molecules that may be targeted for therapeutic purposes in humans.

Although vitamin D deficiency is a proven risk factor for obesity -, the exact relationship between vitamin D status and obesity remains unclear. Vitamin D function is induced by the attachment of 1,25-dihydroxyvitamin D3, one of vitamin D active forms, to VDR as a nuclear receptor and a product of the VDR gene locus on chr12q13.1. A nuclear receptor acts as a ligandinducible transcription factor. Some polymorphisms reported in the VDR gene are associated with certain diseases and phenotypes Several VDR gene polymorphisms, including TaqI, BsmI, and ApaI, are located near the 3’ un-translated region (UTR). In the human VDR, the FokI translation-start site polymorphism was found in the 5’ UTR.

In Iran, the prevalence of overweight/obesity is 63.6% in adults while that of abdominal obesity is 75.2%, and it is higher in women (32.2%) than in men (15.1%) . Mirzazadeh, et al., have reported an obesity prevalence of 13.7% in men and 27.3% in women . Frequently, VDR gene polymorphisms have been found to be related to obesity, although these findings differ depending on the population -. Considering them as risk factors for obesity, we studied three of these polymorphisms: TaqI, ApaI, and BsmI, and their relationships with serum factors in an obese Iranian population.

Materials and methods

Data collection and participants

We conducted a case-control study in the private nutrition clinic of Abolabbas, a charity institution in Khorasgan, Isfahan, Iran, from October, 2014, to March, 2015. Multiple clinical evaluations were performed and a questionnaire was completed by participants. The parameters recorded included weight, height, and BMI calculated by the accurate measurement of height (meters) and weight (kg) under the supervision of a nutritionist. According to the BMI classification proposed by the World Health Organization (WHO), 320 subjects were categorized as healthy controls with normal weight (BMI: 18.5-24.9 kg/m2) and no chronic diseases while 348 participants were considered obese individuals (BMI≥30 kg/m2) ; their ages ranged from 25 to 80 years. Pregnant or breastfeeding women were excluded, as well as those with a family history of obesity, severe psychological disorders, or other serious diseases, and those consuming metformin, vitamin D, calcium supplements, or cholesterol-lowering medications.

We used biochemical kits to evaluate fasting blood sugar (FBS) (Biorexfars, Iran), triglycerides (TG) (ParsAzmun, Iran), total cholesterol (ParsAzmun, Iran), high-density lipoprotein cholesterol (HDL cholesterol) (ParsAzmun, Iran), and low-density lipoprotein cholesterol (LDL cholesterol) (calculated parameter) after eight hours of nocturnal fasting. The results of tests related to blood biochemical parameters were evaluated by an internist and a nutritionist.

All the participants signed written informed consent forms before sampling and data collection. The Research Committee of the Isfahan (Khorasgan) Branch, Islamic Azad University, Isfahan, Iran approved the present study.

Genotyping

Blood samples were collected in EDTA tubes and DNA was extracted from peripheral blood leukocytes through standard salting-out. We used the PCR- RFLP technique with forward and reverse primers to amplify and analyze the genomic DNA of VDR genotype TaqI, ApaI, and BsmI polymorphisms following the protocol of a previous study (14). Allelic nomenclatures of dominant (ABT) alleles are based on endonuclease success over its TaqI, ApaI, and BsmI restriction sites. Recessive (abt) alleles were therefore utilized when these endonucleases failed to cut their corresponding DNA molecules.

Data analyses

We analyzed the data using the SPSS-20.0™ software. We examined the Hardy-Weinberg genotype equilibrium in the control group using the chi- squared test. The descriptive data were expressed in means and standard deviations, as well as allele and genotype frequency. The Kolmogorov- Smirnov test was used to analyze the normality of distribution for each variable. To compare groups we used the chi-square test for qualitative data and for quantitative data, independent samples T-test. Odds ratios (OR) were calculated using logistic regression with 95% confidence intervals (CI). A p-value less than 0.05 was considered statistically significant.

Results

The characteristics of the obese and control subjects and the details of the control group (mean age: 50.02 ± 12.47 years) and the obese participants (mean age: 36.30 ± 10.19 years) are shown in table 1. There were significantly more women in the obese group than in the control group. FBS, TG, and BMI were also significantly higher among the obese subjects compared to the controls. There were no significant differences between the two groups in terms of smoking status, alcohol consumption, HDL cholesterol, LDL cholesterol, total cholesterol, and heart disease; however, there were significantly more subjects who practiced physical activity in the control group than in the obese group (p=0.010).

Table 1

Anthropometric and biochemical characteristics of study subjects

Variable		Normal weight (Control group n=80)	Obese (Case group n=87)	p-value
Sex; n (%)	Male	Sex; n (%)	Sex; n (%)	0.002a
	Female
Smoking status; n (%)	Yes	Smoking status; n (%)	Smoking status; n (%)	0.342a
	No
Physical activity; n (%)	Yes	Physical activity; n (%)	Physical activity; n (%)	0.010a
	No
Alcohol intake; n (%)	Yes	Alcohol intake; n (%)	Alcohol intake; n (%)	0.428b
	No
Heart disease status; n (%)	Yes	Heart disease status; n (%)	Heart disease status; n (%)	1.00b
	No
Age (year)	Mean ± SD	Age (year)	Age (year)	<0.001c
BMI (kg/m2)	Mean ± SD	BMI (kg/m2)	BMI (kg/m2)	<0.001d
FBS (mg/dl)	Mean ± SD	FBS (mg/dl)	FBS (mg/dl)	<0.001c
LDL cholesterol (mg/dl)	Mean ± SD	LDL cholesterol (mg/dl)	LDL cholesterol (mg/dl)	0.244c
HDL cholesterol (mg/dl)	Mean ± SD	HDL cholesterol (mg/dl)	HDL cholesterol (mg/dl)	0.132c
Triglycerides (mg/dl)	Mean ± SD	Triglycerides (mg/dl)	Triglycerides (mg/dl)	<0.001d
Total cholesterol (mg/dl)	Mean ± SD	Total cholesterol (mg/dl)	Total cholesterol (mg/dl)	0.071d

p-values from a: Chi square test; b: Fisher exact test; c: Mann-Whitney test; d: independent t study

BMI: body mass index; FBS: fasting blood sugar; LDL: low-density lipoprotein; HDL: high-density lipoprotein

The Hardy-Weinberg equilibrium was observed in the TaqI and ApaI genotypes distributions in the control group (p=0.086 and 0.262, respectively) (table 2). A significant deviation was, however, observed in the Hardy- Weinberg equilibrium of the BsmI polymorphism in the controls (p<0.001).

Table 2

Hardy-Weinberg equilibrium in the control group

Genotype		Observed	Expected	χ2	p-value
Apal	aa	12	14.5	1.256	0.262
	Aa	44	39.1
	AA	24	26.5
Taql	tt	8	11.6	2.957	0.086
	Tt	45	377
	TT	27	30.6
Bsml	bb	4	171	34.778	<0.001
	Bb	66	39.8
	BB	10	23.1

VDR polymorphisms and obesity risk

Table 3 shows the allelic and genotypic frequencies of VDR ApaI and TaqI polymorphisms in both groups. ApaI and TaqI genotypes were expressed as recessive homozygous (aa, tt), dominant homozygous (AA, TT), and heterozygous (Aa, Tt). The frequency of genotype AA was 47.1% and that of the combined genotype Aa+aa, 52.9% in the obese group, figures that corresponded to 30% and 70% in the controls, were respectively (p=0.024, OR=2.08, 95%CI=1.100-3.933). Compared to genotype aa, Aa was found as a potential risk factor for obesity (p=0.029, OR=3.417, 95%CI=1.135-10.283). The frequencies of A and a alleles in the ApaI polymorphism were statistically significant in the two groups (allele A vs. a, p=0.017). As shown in table 3, there was no significant association between TaqI genotypes and alleles in neither group.

Table 3

Apal and Taql polymorphisms and obesity risk

Genotype	Obese (%)	Control (%)	p-value	OR	95%CI
Apal
aa	6 (6.9)	12 (15.0)		1
Aa	40 (46.0)	44 (55.0)	0.029	3.417	1.135-10.283
AA	41 (47.1)	24 (30.0)	0.273	1.818	0.624-5.298
Aa+aa	46 (52.9)	56 (70.0)		1
AA	41 (47.1)	24 (30.0)	0.024	2.08	1.100-3.933
A	52 (29.9)	68 (42.5)		1
A	122 (70.1)	92 (575)	0.017	1.734	1.104-2.723
Taql
tt	8(9.2)	8(10.0)		1
Tt	42(48.3)	45(56.3)	0.574	1.370	0.457-4.110
TT	37(42.5)	27(33.8)	0.899	0.933	0.321-2.711
Tt+tt	50 (57.5)	53 (66.3)		1
TT	37(42.5)	27(33.8)	0.245	1.453	0.775-2.724
t	58 (33.3)	61 (38.1)		1
T	116 (66.7)	99 (61.9)	0.361	1.232	0.787-1.930

P-value, OR and 95%CI based on logistic regression

Apal genotypes associations with biochemical parameters

In table 4 we present the distribution of clinical variables by genotypes in the obese and control groups. We combined genotypes Aa and aa given the limited frequency of aa genotype in the VDR ApaI polymorphism. Moreover, 86.2% of the AA genotype and 70.6% of the Aa + aa group corresponded to women. According to the Chi-squared test, the two groups were significantly different in terms of gender distribution and the vast majority of the participants with genotype AA were female (p= 0.020).

Table 4

Apal polymorphism and obesity components

Variable		Apal AA	Apal Aa+aa	p-value
Gender, n (%)	Male	9 (13.8)	30 (29.4)	0.020a
	Female	56 (86.2)	72 (70.6)
Heart disease, n (%)	Yes	3 (4.6)	1 (1.0)	0.300b
	No	62 (95.4)	101 (99.0)
BMI (kg/m2)	Mean ± SD	31.56 ± 7.72	28.55 ± 6.62	0.022c
FBS (mg/dl)	Mean ± SD	114.61 ± 32.43	103.69 ± 29.41	0.003c
LDL cholesterol (mg/dl)	Mean ± SD	10771 ± 37.16	109.74 ± 31.34	0.718d
HDL cholesterol (mg/dl)	Mean ± SD	46.25 ± 12.59	45.33 ± 10.65	0.627d
Triglycerides (mg/dl)	Mean ± SD	162.98 ± 86.38	149.97 ± 82.56	0.242c
Total cholesterol (mg/dl)	Mean ± SD	184.29 ± 45.25	185.82 ± 43.65	0.832d

P-value from a: Chi square test; b: Fisher exact test; c: Mann-Whitney test; d: independent t study

BMI: body mass index; FBS: fasting blood sugar; LDL: low-density lipoprotein; HDL: high- density lipoprotein

As for the ApaI variant, BMI (p=0.022) and FBS (p=0.003) were higher in the subjects carrying the AA genotype compared to those with the Aa+aa genotype. Other parameters such as heart disease, LDL cholesterol, HDL cholesterol, TG, and total cholesterol were not associated with the AA and Aa+aa genotypes.

Discussion

A complex feedback mechanism mediated by receptors, enzymes, and hormones, regulates 1,25(OH)2D. With a key function in bone and calcium homeostasis, 1,25(OH)2D/VDR signaling can regulate proliferation, differentiation, and various cellular responses in the cardiovascular and immune systems ,. The evidence suggests that VDR polymorphisms and vitamin D deficiency can increase the risk of osteoporosis, calcium stones, diabetes, and prostate cancer and the first may affect development and growth processes.

A large body of literature has been devoted to four common singlenucleotide polymorphisms in the VDR gene: TaqI C>T (rs731236), ApaI C>A (rs7975232), FokI T>C (rs10735810), and BsmI A>G (rs1544410), whose relationships with different human diseases and traits have been explored . The single-nucleotide polymorphisms near the 3’ UTR include ApaI, TaqI, and BsmI. Despite being non-functional, they are linked with a poly (A) microsatellite repeat in the 3’ UTR, which can affect the stability of the VDR mRNA, the efficiency of protein translation, and the modulation of gene expression while the protein expression can be affected by altered intronic sequences ,.

Our findings suggested an association between the VDR ApaI polymorphism (rs7975232 C/A) and the susceptibility to obesity while the A allele and the AA genotype in ApaI were associated with obesity phenotypes. The relationships of the VDR gene polymorphisms with the anthropometric and biochemical features of obesity were evident in the higher serum levels of FBS and BMI in genotype AA carriers.

Identifying the risk factors and genetic causes of obesity is crucial given its significant frequency in developing and developed countries. Risk factors are different in many populations and several VDR gene variants appear to affect obesity differently. A few studies suggest relationships between VDR polymorphisms and obesity ,,. For example, Ferrarezi, et al., found that BsmI polymorphism was related to the height in a cohort of obese adolescents and children but TaqI and ApaI were not significantly associated with obese adolescents and children . In 2018, Correa-Rodríguez, et al., found that VDR genetic variants did not contribute to obesity phenotypes in a population of Caucasian young adults .

A study in Saudi Arabia reported vitamin D deficiency, VDR BsmI, and Taq1 genotypes as risk factors for obesity . In 2014, another study in Saudi Arabia by Al-Daghri, et al., revealed that polymorphisms affecting the vitamin D/VDR axis played a role in obesity in terms of the inflammation possibly caused by changes in microbial translocation and gut permeability. All the polymorphisms, namely TaqI, ApaI, and BsmI, were found to relate to obesity . In 2015, Ostadsharif, et al., obtained the frequency of alleles F and f in obese and healthy groups. The difference between the two alleles in the control and obese groups was significant (p=0.005) and the individuals with the FF genotype in the control group had lower fasting blood sugar levels compared to the other genotypes . In contrast, a study in Poland reported no statistically significant differences in BMI, nor in the weight or height, for four VDR genotypes, i.e. TaqI, ApaI, BsmI, and FokI .

The study of ApaI and TaqI to evaluate VDR polymorphisms in nephropathic and non-nephropathic patients with type 2 diabetes conducted by Iranian researchers did not show relationships between the ApaI polymorphism and the nephropathic and non-nephropathic diabetic patients, although it showed significant differences in VDR gene TaqI genotypes of diabetic patients with and without nephropathy .

The contribution of VDR gene polymorphisms to susceptibility to neurodegenerative diseases and conditions associated with calcium metabolism has been frequently reported in the literature. In Japan, the lumbar spine BMD of the common genotype AA was lower by 9.3% than in genotype aa in premenopausal women . In Greece, no heritability patterns were reported for the relation of genotype AA and low BMD .

In Iran, Meamar, et al., evidenced that FokI f and ApaI a alleles significantly increase the risk of Parkinson’s disease as was the case of ApaI heterozygous genotype Aa compared to the AA homozygous .

In the present study, we found higher BMI and FBS levels in those subjects with genotype AA than in those with genotype Aa+aa. In a metaanalysis, the authors found significant relations between the VDR gene TaqI, BsmI, ApaI, and FokI polymorphisms and the risk of type 2 diabetes; FokI polymorphism was a risk factor especially in Asians . On the other hand, studies in an Egyptian population showed differences in VDR genotypes Apa-I and Taq-I allele distribution and frequency between diabetic individuals and controls . There are, of course, a few studies demonstrating that the ApaI polymorphism was related to obesity and blood glucose. Most articles focused on type 2 diabetes or polycystic ovary syndrome and other VDR gene polymorphisms ,-.

Given that the outcomes depend on numerous environmental, cultural, geographical, and socioeconomic factors, the results obtained differ by population. Like other studies, ours was not without limitations. The size of the study population and the fact that not all variants in the VDR gene were evaluated are the most important ones. Vitamin D serum levels were not assessed either. On the other hand, the most important strengths of the study were the ethnic homogeneity of participants and the association of genotype and blood biochemical factors with BMI.

We found relationships between the VDR gene ApaI polymorphism and obesity in an Iranian population. The relation of ApaI AA genotype and A allele to obesity phenotypes can make them obesity potential predictors. These findings suggest that the ApaI polymorphism can predict the increased risk of obesity and help to identify novel treatment strategies for this metabolic disorder. We recommend future studies with larger samples to analyze the relationship between environmental factors and single nucleotide polymorphisms involved in obesity in different Iranian populations.