Effects of polymorphisms in APOB, APOE, HSD11β1, PLIN4, and ADIPOQ genes on lipid profile and anthropometric variables related to obesity in children and adolescents

Abstract Genes can influence lipid profile and anthropometric variables related to obesity. The present study aimed to verify if variants of the APOE, APOB, ADIPOQ, HSD11β1, and PLIN4 genes are associated with lipid levels or anthropometric variables in a sample comprised of 393 Euro-Brazilian children and adolescents. DNA was genotyped by TaqMan allelic discrimination assay. The ε4 and ε2 alleles of the APOE gene were associated respectively with lower high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C) levels (p=0.015 and p=0.012, respectively), while the ε3 allele was associated with higher abdominal circumference (p=0.0416) and excess weight (p=0.0001). The G allele (rs846910) of the HSD11β1 gene was also associated with excess weight (p=0.039). No other association was found. Our results indicate that the ε4 and ε2 alleles could contribute to lower HDL-C and LDL-C levels, respectively, furthermore, the ε3 allele and the G allele (rs846910) of HSD11β1 gene may be risk factors for excess of weight.These findings are very important because we observed that some genetic variants influence the lipid profile and anthropometric variables early in life.


Introduction
Dyslipidemia is closely related to the development of cardiovascular and cerebrovascular diseases, such as atherosclerosis, acute myocardial infarction, ischemic heart disease, and cerebrovascular accident, and therefore of great relevance for public health (ANVISA, 2011;Maria et al., 2011). It is estimated that 53% of American adults have lipid abnormalities (Tóth et al., 2012). In Brazil, according to Alcântara Neto et al. (2012), the prevalence of dyslipidemia among children and adolescents enrolled in the public school system was 25.5%. They also found a positive association between dyslipidemia and overweight (Alcântara Neto et al., 2012). Worldwide, in 2015, the number of overweight children under five years old had been estimated at more than 42 million (WHO, 2016).
The human gene encoding adiponectin, ADIPOQ gene (3q27), is the most expressed gene in adipose tissue (Maeda et al., 1996). Obesity, and in particular the accumulation of abdominal visceral fat, as well as type 2 diabetes mellitus, coronary disease, and arterial hypertension are accompanied by a reduction of serum adiponectin (Arita et al., 1999;Lara-Castro et al., 2007). The SNP of the ADIPOQ gene was rs1501299: NM_001177800.1:c.214+62G > T.
Hence, the aim of the present study was to investigate possible influences of the PLIN4 (rs8887), APOB (rs5742904), ADIPOQ (rs1501299), HSD11b1 (rs848910 and rs12086634), and APOE (rs7412 and rs429358; alleles e2, e3, and e4) genes on lipid and glucose levels, abdominal circumference, and obesity in a sample of children and adolescents from a population in southern Brazil.

Subjects
The sample was comprised of 393 Euro-Brazilians (13.54 ± 0.095 years old) living in Curitiba, PR, of which 143 were eutrophic and 250 overweight. Of these 393 individuals, 128 were girls (21.09% eutrophic and 78.91% overweight) and 265 were boys (43.94% eutrophic and 56.06% overweight). This study was approved by the Institutional Ethics Committee and informed consent was signed by participants and their parents or legal guardians.
Body mass index (BMI) was calculated as weight (kg) divided by the square of height (m). Age-and sex-specific BMI z-score and percentiles were calculated using CDC 2000 growth charts (Kuczmarski et al., 2002). Eutrophic was defined as a < 85 percentile, overweight as a ³85 percentile, and obesity as ³95 percentile. The abdominal circumference (AC) was measured in centimeters (cm) at the level of the iliac crest. Thus, subjects were classified as eutrophic (percentile < 85) and overweight/obese (percentile ³ 85) (Kuczmarski et al., 2002).
Blood samples were collected in the morning after 12 hours of fasting to perform measurements of glucose (Glu), triglycerides (TG), total cholesterol (TC), and high density lipoprotein cholesterol (HDL-C) by standard automated methods. Low density lipoprotein cholesterol (LDL-C) levels were calculated using the Friedewald equation (Friedewald et al., 1972), for TG levels below 200 mg/dL.

Genotyping assays
DNA was extracted from peripheral blood by a salting-out method (Lahiri and Numberger, 1991) and was diluted to 20 ng/mL. All SNPs were genotyped by TaqMan allelic discrimination assay on StepOnePlus real time PCR systems (Applied Biosystems, USA). Each reaction contained 3.0 mL of Master Mix (2X), 1.7 mL of ultrapure water, 0.3 mL of primer and 3.0 mL of DNA. The reactions were performed according to the following protocol: 50°C for 2 min, 95°C for 10 min, and 50 cycles of 95°C for 15 s and 62 ºC for 1 min.

Statistical analysis
Samples were classified into two groups, eutrophic and overweight (overweight + obese), categorized into above and below the median for age, AC, Glu, TC, LDL-C, HDL-C and TG levels. Chi-square tests were performed using Clump (Jakobsson and Rosenberg, 2007) to test for Hardy-Weinberg equilibrium and to compare allele proportions between groups above and below the median and also between eutrophic and overweight. Logistic regression analyses were performed to identify variables influencing serum glucose, lipid concentrations, and AC. False discovery rate (FDR) corrections (Benjamini and Hochberg, 1995) were performed for multiple testing. The significance level adopted was 0.05 (5%).

Results
A descriptive analysis of the sample, displaying the variables considered in this study, is shown in Table 1. Significantly higher frequencies were found for the e4 allele in the group below the HDL-C median (p=0.0001), and for the e2 allele in the group below the LDL-C median (p=0.0001). Furthermore the e3 allele was associated with higher AC and excess weight (p=0.0001). The allele frequencies are shown in Table 2. Logistic regression analysis was done using stratified TC as below and above the median as the de-pendent variable, and for the polymorphisms analyzed (dominant model for APOE gene, in which e4 is dominant over e2; for the other polymorphisms, dominant, recessive, and additive models were tested), gender, AG, and anthropometric classification as independent variables. The same logistic regression analysis design was performed using LDL-C, HDL-C, TG, glucose, and AC as the dependent variable and maintaining the same independent variables. We identified the APOE gene e4 allele as a contributing factor in reducing HDL-C levels (b = -0.29 ± 0.08, p=0.015) and the e3 allele as a risk factor for higher AC measures (b = -0.24 ± 0.08, p=0.041). We also found that obesity and overweight are independent risk factors for higher triglyceride levels (b = 0.30 ± 0.08, p=0.021).
Furthermore, we observed that the A allele (rs846910) of the HSD11b1 gene was associated with excessive weight (p=0.039, Chi-square test). It is known that there is variation in metabolic processes inherent to gender, so we conducted the same analyses separately for each gender. We observed that in girls the alleles e2 and e4 of the APOE gene were associated with LDL-C below the median (p=0.0001 by Chi-square test) and HDL-C below the median, independently of the other analyzed variables (b = -0.34 ± 0.08, p=0.0039) (Table 3). Furthermore, eutrophic girls had lower mean TG levels than obese or overweight girls (b = 0.30 ± 0.08, p=0.0039). Regarding boys, we observed that the e2 allele is associated to lower LDL-C levels (p=0.019 by Chi-square test) (Table 3).
Gene polymorphisms and obesity 737

Discussion
Blood lipid levels are influenced by environmental and genetic factors (Crook, 2012), and it is known that LDL-C is the primary target for reducing cardiovascular risk (Catapano et al., 2016). In our study, as shown in Figure 1, it was observed that the APOE e2 allele was associated with lower LDL-C levels in the total sample, as well as in girls and boys, which is consistent with the known protective effect of this allele (Frikke-Schmidt et al., 2000;Bennet et al., 2007;Fuzikawa et al., 2008;Ward et al., 2009;Nascimento et al., 2009;Bazzaz et al., 2010;Ferreira et al., 2010). Our finding is particularly relevant considering that the protective effect of the e2 allele is usually observed in adults, but in our study we observed that it is also present in children and adolescents, and therefore can contribute to lower LDL-C levels early in life.
The APOE-e4 allele seems to be associated with lower HDL-C levels, which support the notion that the e4 allele is an atherogenic risk factor (Frikke-Schmidt et al., 2000;Bennet et al., 2007;Fuzikawa et al., 2008;Ward et al., 2009). Being related to lower HDL-C levels, responsible for cholesterol reverse transport, this allele could contribute to higher cholesterol levels, and this is especially worrisome in children, considering all possible and severe comorbidity (ANVISA, 2011;Maria et al., 2011;Crook, 2012).
Besides its association with the lipid profile, some studies have demonstrated that the APOE gene influences characteristics of obesity (Volcik et al., 2006;Tabatabaei-Malazy et al., 2012). According to the Atherosclerosis Risk in Communities (ARIC) study, the apo E genotypes were associated with BMI following the order: apo E4 > apo E3 > apo E2 (Volcik et al., 2006). Srinivasan et al. (1994), who analyzed a sample of children and adolescents, similar to this study, found that the apo E3 group showed significant associations with obesity measures and lipoprotein variables.
Our work is in agreement with Sun et al. (2016) who also found some increased variables, such as BMI and LDL-C, in e3 allele carriers when compared to e2 allele carriers in the non-metabolic syndrome group (Sun et al., 2016). Some studies also have associated the e4 allele with features related to obesity in different populations (Tabatabaei-Malazy et al., 2012;Alharbi et al., 2017). Therefore, these polymorphisms in the APOE gene influence both lipid profile and traits related to obesity. It is important to highlight the relevance of studies involving this gene, especially in the young. We found a relationship between the G allele of the HSD11b1 gene rs846910 polymorphism and higher AC measurements. Some studies have demonstrated different effects of this polymorphism on serum lipid levels and other associated characteristics (Nair et al., 2004;Duran-Gonzalez et al., 2011;Dujic et al., 2012). Different from this study, Durán-Gonzalez et al. (2011) observed an association between the A allele and higher triglyceride 738 Gasparin et al.  levels, and according to some studies this polymorphism could be associated to metabolic syndrome (Nair et al., 2004;Duran-Gonzalez et al., 2011;Dujic et al., 2012). However, Turek et al. (2014) found that the A allele ws associated with higher HDL-C levels only in women. Furthermore, it is relevant to consider that a possible linkage disequilibrium might exist with another polymorphism in the HSD11b1 gene, and another allele could be the cause of an altered lipid profile or features related to obesity (Malavasi et al., 2010). Although our study had relevant findings, we recognize that the small sample size is a limitation, thus generalizability should be done with caution, and studies with larger samples should be done. In summary, we found that in children and adolescents, as in adults, the e4 and e3 alleles could be considered a contributing factor for dyslipidemia and traits related to obesity, respectively, while the e2 allele seems to be a protective factor, contributing to lower LDL-C and higher HDL-C levels. Furthermore, the HSD11b1 gene G allele seems to be related to obesity. Considering that effects may start early in life, a precocious intervention could be planned, therefore preventing many complications resulting from altered lipid profile and obesity.