Federal University of Goiás, Health Science Pos Graduation Program, Goiânia, GO, Brazil; Children´s Hospital of Goiás, Goiânia, GO, Brazil; Pontifical Catholic University of Goiás, Replicon Research Center, Goiânia, GO, BrazilBrazil
Pontifical Catholic University of Goiás, Replicon Research Center, Goiânia, GO, BrazilBrazil
Federal University of Goiás, Health Science Pos Graduation Program, Goiânia, GO, BrazilBrazil
Pontifical Catholic University of Goiás, Replicon Research Center, Goiânia, GO, Brazil; Department of the State of Goiás, Laboratory of Human Cytogenetics and Molecular Genetics, Secretary of State for Health of Goiás, Goiânia, GO, BrazilBrazil
Federal University of Goiás, Goiânia, GO, BrazilBrazil
Also available as PDF (figures and tables appear in the PDF).
Background Genetic polymorphisms that affect the availability and secretion of dopamine can affect the risk of obesity. Objectives To investigate the relationship between pediatric obesity and cardiovascular risk factors (CRF) with the polymorphisms of “Fat Mass and Obesity Associated” (FTO) rs9939609, “Dopamine Receptor type 2” (DRD2) rs6277 and “Ankyrin Repeat and Kinase Domain Containing 1” (ANKK1) rs18000497 genes. Methods Case-Control study conducted with 226 pediatric patients from 5 to 16-years of age. The two main groups, Obese (O) and Eutrophic (E), were subdivided according to the value of HOMA-IR into obese with insulin resistance (ORI) or insulin sensitivity (OSI) and eutrophic resistant (ERI) or sensitive (ESI) to insulin. According to the presence of two or more CRF, they were subdivided into metabolically unhealthy or metabolically healthy groups: Obese Metabolically Unhealthy (OMU), Obese Metabolically Healthy (OMH), Eutrophic Metabolically Unhealthy (EMU) and Eutrophic Metabolically Healthy (EMH). Polymorphisms were determined by real-time Polymerase Chain Reaction (PCR) or Restriction Fragment Length Polymorphisms (PCR-RFLP). Results In the obese group, the higher the number of risk alleles of FTO and ANKK1 genes isolated and the three genes combined, the higher the mean BMI (p<0.0001). Regarding the FTO gene: the frequency of the risk allele was: 57.7%-ERI, 37.4%-ESI (p=0.048), and the homozygous wild genotype was: 29.5%-OMU, 37.5%-OMH (p=0.02). Regarding the DRD2 gene: the genotypes with the risk allele were present in 84.6%-OMU and 67.5%-OMH (p=0.031). Regarding the ANKK1 gene: the frequency of the homozygous risk genotype was current in 15.4%-ERI and 13.5%-ESI (p<0.0001) and 62.5%-EMU and 41.5%-OMH (p=0.031). Conclusion Risk alleles of FTO, DRD2 and ANKK1 genes had an additive effect on the outcome of pediatric obesity in Brazilian children and conferred a higher risk of insulin resistance (FTO and ANKK1) and CRF.
Keywords: Childhood obesity; Genetic polymorphism; Insulin resistance; Metabolic syndrome; Dopamine
Genetic polymorphisms that affect the availability and secretion of dopamine can affect the risk of obesity.
To investigate the relationship between pediatric obesity and cardiovascular risk factors (CRF) with the polymorphisms of “Fat Mass and Obesity Associated” (FTO) rs9939609, “Dopamine Receptor type 2” 1” (ANKK1) rs18000497 genes.
Case-Control study conducted with 226 pediatric patients from 5 to 16-years of age. The two main groups, Obese (O) and Eutrophic (E), were subdivided according to the value of HOMA-IR into obese with insulin resistance (ORI) or insulin sensitivity (OSI) and eutrophic resistant (ERI) or sensitive (ESI) to insulin. According to the presence of two or more CRF, they were subdivided into metabolically unhealthy or metabolically healthy groups: Obese Metabolically Unhealthy (OMU), Obese Metabolically Healthy (OMH), Eutrophic Metabolically Unhealthy (EMU) and Eutrophic Metabolically Healthy (EMH). Polymorphisms were determined by real-time Polymerase Chain Reaction (PCR) or Restriction Fragment Length Polymorphisms (PCR-RFLP).
In the obese group, the higher the number of risk alleles of FTO and ANKK1 genes isolated and the three genes combined, the higher the mean BMI (p<0.0001). Regarding the FTO gene: the frequency of the risk allele was: 57.7%-ERI, 37.4%-ESI (p=0.048), and the homozygous wild genotype was: 29.5%-OMU, 37.5%-OMH (p=0.02). Regarding the DRD2 gene: the genotypes with the risk allele were present in 84.6%-OMU and 67.5%-OMH (p=0.031). Regarding the ANKK1 gene: the frequency of the homozygous risk genotype was current in 15.4%-ERI and 13.5%-ESI (p<0.0001) and 62.5%-EMU and 41.5%-OMH (p=0.031).
Risk alleles of FTO, DRD2 and ANKK1 genes had an additive effect on the outcome of pediatric obesity in Brazilian children and conferred a higher risk of insulin resistance (FTO and ANKK1) and CRF.
Childhood obesity; Genetic polymorphism; Insulin resistance; Metabolic syndrome; Dopamine.
cc Copyright 2021 by Pinto RM. This is an open-access article distributed under allows to copy, redistribute, remix, transform, and reproduce in any medium or format, even commercially, provided the original work is properly cited. (DRD2) rs6277 and “Ankyrin Repeat and Kinase Domain Containing Creative Commons Attribution 4.0 International License (CC BY 4.0), which
besity is recognized by the World Health Organization (WHO) as one of the ten leading health problems in most O 1 diverse societies, and in 2019 chronic non-communicable diseases: obesity, diabetes, cancer, and cardiovascular diseases reached the second position in the ranking world of health challenges. The prevalence of obesity has increased worldwide, with epidemic proportions in many developed and developing countries. Currently, obesity is considered a neurobehavioral disease in which hypothalamic changes occur in the control of hunger, satiety, and energy expenditure. It is a condition of multifactorial etiology influenced by genetic, epigenetic, endocrine-metabolic, and behavioral fac- 6,7 tors.
The regulation of food intake by the Central Nervous System (CNS) depends on the interaction of a homeostatic component that aims to balance energy and nutrients and a hedonic component, which seeks pleasure associated with food. The ingestion of highly palatable foods (rich in sugar and fats) can deregulate the homeostatic control of appetite, perpetuating the stimulus to eat, which makes the intake primarily mediated by hedonic and non-homeostatic needs.10
Cortico-limbic-striated structures and circuits form the Brain Reward System (BRS). BRS is activated by various stimuli that trigger a cascade of neurotransmitter secretion that ultimately releases dopamine (DA), bringing a sense of pleasure and leading the individual to seek positive reinforcements. The brain reward cascade starts in the hypothalamus, where serotonin acts as a neurotransmitter stimulating the enkephalin release, which in turn inhibits GABAergic neurons in the substantia nigra. These GABAergic neurons act in the fine adjustment of DA amount that will be released in the NAc, the brain reward site (Figure 1).
Figure 1. Interaction of Various Neurotransmitters that forms the “Brain Reward Cascade Source: Pinto et al Considering that DA plays a crucial role in BRS, it is in- volved in eating behavior, and that genetic factors are responsible for up to 80% of the inter-individual variation of nutritional sta- tus,13 the study of genetic polymorphisms that affect the availabil-
ity and secretion of DA have gained space in the literature.12
The dopamine receptor type 2 (DRD2) is the main self-regulator of DA release in the brain reward dopaminergic system, with pre and 2 postsynaptic effects.14
3,4 The DRD2 gene, located at 11q22-q23, has several polymorphisms capable of influencing the quantity and functionality of receptors in the membrane of the postsynaptic cell, which directly implies in the predisposition of individuals to seek any sub stance or behavior that stimulates BRS15 with a consequent higher propensity to gain weight in the future.
The single nucleotide polymorphism (SNP) rs6277 of the DRD2 gene consists of the change of an Arginine (A) by Guanine (G) in the position 957pb in exon 7. There are only a few 8,9 researches that studied this polymorphism from a metabolic point of view, but it is known that the presence of the risk allele (G) leads to a reduction in the affinity of the receptors and is associated with: psychiatric disorders related to the deficiency of dopaminergic function and processing of emotions,17 a greater sense of reward even in individuals of normal weight,18 higher consumption of sugar and increased body mass index (BMI) from normal to overweight/obesity in adults.19
10,11 Polymorphism
The Ankyrin repeat and Kinase domain containing 1 gene (ANKK1) is also located in the 11q22-q23 region. It’s rs1800497 SNP interferes with the availability of DRD2 receptors and was considered part 12 of the DRD2 gene itself, under the name of Taq1A polymorphism of the DRD2 gene. This SNP influences the signaling of DRD2 receptors: healthy individuals who have the risk allele (T) have a reduction of between 30-40% in the density of DRD2 in the striatum when compared to those without T-allele, and reduced glucose metabolism in the regions of BRS that contain dopaminergic neurons.20 This reduction in the BRS’s dopaminergic action reflects 16 16 clinically in a higher risk of obesity in adults,16 greater weight gain,16 more intense reinforcement (feeling of pleasure) after a meal, and increased energy intake, especially in obese.21 There is a wide variation in genotype and allele frequencies in different countries, and even within the same country, there is inconsistency regarding the association or not with metabolic outcomes. Chart 1 shows the main results of these scientific researches.
The Fat Mass and Obesity Associated (FTO) gene, located at 16q12.2, encodes a 2-oxoglutarate-dependent demethylase involved in posttranslational modifications, deoxyribonucleic acid (DNA) repair, and fatty acid acidification.22 The presence of at least one risk allele (A) of this SNP is associated with higher caloric intake, less satiety, and a higher frequency of episodes of compulsive eating when 23-26 compared with homozygotes for the wild allele (T).
Figure 1. Summary of Studies Carried Out in the Pediatric Population that Evaluated the rs1800497 Polymorphism of the ANKK1 gene
26 Children from 8 to 12-years old with 84 overweight and parents with obesity Presence of theT allele associated with higher BMI and also weight modification Epstein et al Country: USA Children There 46 obese proportion was observed: 85 50 eutrophic Ergum et al Country:Turkey Eutrophic:T- 52% C- 48% 44 female teenagers 16 Observation of MRI in a situation of food anticipation showed attenuated Stice et al Country: USA
279 Adolescents 86 Allele Strien et al Country: Netherlands T -18.3%; C- 81.7%
448 Adolescents TT-15.2; CT- 46.5; CC- 38.3; 87 T-38.5; C- 61.5 Duran-Gonzalez et al Country: USA
200 pediatric patients aged 2 to 17 years. T-20;C - 80 42 50% obese 50% eutrophic Araz et al Country:Turkey TT-4; CT-31; CC- 65
583 eutrophic adults TT-2.4%; CT- 29.3%; CC- 68.3% 28 overweight children T- 17%; C- 83% 88 423 children with obesity Roth et al Country: USA allele; however, the TT genotype showed a worse response after a lifestyle
Pediatric patients from 5 to 16 years of age 76 Genotype and allele frequency (%) in controls: Pinto et al 55 obese 50 eutrophic Country Brazil
84 teenagers of Asian origin 77 TT + CT 63%; CC 37% Yeh et al Country: USA 258 Pediatric patients aged 8 to 14 years 115 obese, Genotype frequency (%) 78 42 overweight, TT-9.7; CT- 38.4; CC- 51.9 Yobregon et al 101 eutrophic No difference in BMI,T allele associated with greater pleasure in eating in obese Country: Chile 286 Pediatric patients aged 7 to 12 years Genotype frequency (%) 79 TT– 12.9; CT– 39.5; CC- 47.6 Cartel et al Country: USA TT homozygotes reported 20% higher energy intake of carbohydrates, and
Currently, FTO is the gene that shows the most significant association with human polygenic obesity. After evaluating 3337 obese and 3159 controls, a recent meta-analysis study definitively confirms its correlation with obesity in European populations.27 The association of FTO with obesity has been confirmed in other 28 29,30 31-33 ethnicities such as Asians and Hispanics. Few studies have evaluated this SNP in the Brazilian population, and only 7 have
after a weight loss program lasting 0.5 to 1-year.
was no difference in the allelic distribution of the groups.The following
Obese:T- 51%, C- 49%
response of the CNS of patients with the presence of theT allele. Genotype frequency: TT-2.9%; CT - 30.8%; CC - 66.3% frequency:
The presence of theT allele increased the intake of emotional background Genotype and allele frequency (%):
T allele associated with central obesity Genotypic and allele frequency (%) in Obese: TT-7; CT-26; CC- 67
Genotype and allele frequency (%) in Controls:
No association ofT allele and BMI n children, the following genotypic and allelic frequencies were found:
There was no association between elevated BMI with the presence of theT change program aimed at weight loss.
Genotype and allele frequency (%) in the obese: TT-14.5; CT-40; CC-45.5 T-34.5;C 65.5 TT-10; CT-26; CC-64 T-23 allele; C-77 T allele associated with obesity and alteration of glycemic homeostasis Genotype frequency (%) No difference in BMI, the presence of theT allele associated with increased consumption of carbohydrates and fast food
boys.
greater deposition of visceral fat.
been carried out in the pediatric population of our country. Chart 2 shows the main findings of these studies.
27 34 Recently Sun et al34 reviewed evidence that suggest a role of two common gene variants, FTO and ANKK1, in driving geneenvironment interactions leading to obesity, metabolic dysfunc-
Figure 2. Summary of Studies Conducted with Brazilian Children and Adolescents to Evaluate the rs9939609 Polymorphism of the FTO gene Silva et al Lourenco et al AA- 15.6; AT- 44.8;TT- 39.6 Pereira et al Reuter et al
Nascimento et al
Rodrigues et al
Todendi et al AlleleA associated with increase in BMI
Legend: AC: Abdominal circumference; AM: Amazonas; BMI: Body mass index; MA: Maranhão; MG: Minas Gerais; PR: Paraná; RS: Rio Grande do Sul.
tion, and cognitive change via their influence on DRD2 signaling. It was also demonstrated that the FTO protein regulates dopamine signaling via DRD2, and the polymorphisms rs9939609 of the FTO gene and rs18000497 of the ANKK1 gene interact.35 A study conducted in humans found that FTO risk variants affect the brain dopaminergic response and modify a pleasurable task’s learning. The FTO gene modulates the BRS connectivity, suggesting that the increased risk for obesity is related to the processing of the sensation of pleasure.35
The main objective of this study is to investigate the relationship between pediatric obesity and the polymorphisms of the ANKK1 (rs18000497), DRD2 (rs6277) and FTO (rs9939609) genes, in a population of Brazilians. We also intend to correlate the studied polymorphisms with the degree of insulin sensitivity and cardiovascular risk factors (CRF) components of the metabolic syndrome (MS): Blood pressure (BP), Abdominal circumference (AC), Triglycerides (TG), High-density cholesterol (HDL) and blood glucose.
This research is a Case-Control study carried out jointly by LaGene - Laboratory of Human Cytogenetics and Molecular Genetics of the Health Department of the State of Goiás (LaGene/Lacen/ SES-GO), NPR- Replicon Research Center (PUC-GO), and the Children’s Hospital of Goiânia/Child Endocrinology Office and
N = 348 Genotype and allele frequency (%) 59 Age: 0-8-years AA- 16.4; AT- 46.3;TT- 37.4 A - 40;T- 60 State: RS Association with higher BMI and adiposity N = 1088 Genotype frequency (%) 60 Age: 0-10-years Association with weight gain State: AM 195 obese Allele frequency (%) 64 153 eutrophic Obese: A- 49.5;T-50.5 Eutrophic: A- 46.7;T-53.27 State: MG There was no difference between groups N = 406 Genotype and allele frequency (%) 61 Age: 7-17-years AA-13.3; AT-44.3;TT- 42.4 A - 35.5;T- 64.5 State: RS AA genotype associated with obesity/overweight 136 obese Genotype and allele frequency (%) 172 eutrophic Obese: AA- 14.7; AT- 46.3;TT- 38.9 62 Age: 8-17-years Eutrophic: AA-9.9; AT-52.3;TT- 37.8 AlleleA associated with lower AC reduction State: PR after intervention with diet and exercise 378 obesos Genotype and allele frequency (%) 378 eutrophic Obese: AA -14.5; AT-44.9,TT-40.5 / A-36.8;T-63.2 65 Age:18-9-years Eutrophic: AA-13.5; AT- 47.9;TT-38.6 / A- 37.4; T-62.6 State: MA There was no difference between groups N = 871 63 Age: 7-17-years State: RS
Federal University of Goiás (LabMut/FG). The project was approved by the Research Ethics Committee of PUC Goiás, with the number: 16303313.4.0000.0037. All parents or guardians were interviewed and signed the informed consent form consenting with their child’s participation in the study and data for research. The researchers clarified the methodological research procedures before and during the study to all participants.
The study included pediatric patients from 5 to16-years of age who sought the pediatric endocrinology office of the Children’s Hospital of Goiânia (Goiânia-Goiás-Brazil) years 2017 and 2018. The responsible researcher was personally assessed in complete pediatric consultation with thorough anamnesis and physical examination. Exclusion criteria were: overweight, malnutrition, severe chronic diseases, presence of genetic syndromes, use of medications known to alter weight.
The nutritional status diagnosis was based on the BMI with interpretation according to the reference values proposed by WHO. According to the BMI, children were divided into two main groups: Obese (whose BMI was above +2 SD) and Eutrophic (whose BMI was between -2 SD and +1 SD) that constituted the control group.
The BP was evaluated according to the Hypertension Guidelines for children and Adolescents,37 which follows the values of the Consensus of the American Academy of Pediatrics.38 Arterial hypertension was considered if the systolic and/or dia-
stolic blood pressure values were equal to or higher than the 95th percentile for sex, age, and height percentile on three or more oc- 37,38 casions.
The AC was measured at the midpoint between the last th fixed rib (10th) and the iliac crest’s upper border. P90 was considered for age, sex, and ethnicity as the maximum value of normality, according to the values described by Freedman and collaborators.39
After 8 to 12-hours of fasting, blood was collected to determine the lipid profile, blood glucose, and insulin. Through the equip- ® ment Architect c8000® Plasma total cholesterol, HDL-cholesterol, and TG were determined by enzymatic-colorimetric assay, and Glucose values were determined by the enzymatic method (Glucose Oxidase - Latest, SP, Brazil). Insulin was measured by the chemiluminescence technique using an immunometric immunoassay in a piece of automated equipment Architect i2000®.
The Brazilian Society of Pediatrics’ recommendation for the Brazilian pediatric population was used as the reference for serum lipids.40 The reference values for fasting blood glucose do not differ between children and adults, with typical values between 60 and 100 mg/dL, pre-diabetes between 101 and 125 mg/dL, and diabetes ≥ 126 mg/dL.41 Fasting insulinemia is considered normal from 2.5 to 25 mIU/mL in adults and up to 15 mIU/mL in children.42 The determination of fasting blood glucose and insulin in the same sample allows the calculation of the homeostatic model assessment insulin resistance (HOMA-IR) index, a method used to quantify insulin resistance (IR).43 The interpretation of HOMA-IR values followed the parameters established by de Almeida et al44 for the Brazilian pediatric population. HOMA was considered altered when the value was 2 standard deviations (SD) above the mean for age and sex.
According to the value of HOMA-IR, the main groups were subdivided into obese with insulin resistance (ORI) or insulin sensitivity (OSI) and eutrophic resistant (ERI) or sensitive (ESI) to insulin. According to the presence of two or more CRF, they were subdivided into metabolically unhealthy (MU) or metabolically healthy (MH) groups: OMU, OMH, EMU, EMH.
We choose to use the individualized evaluation of CRF and not the presence or absence of MS as there are no reference criteria for MS before ten-years of age, and even after that age, the criteria for MS remain controversial. More than 40 different cutoff points have already been used,45 and even after the consensus attempt by the International Diabetes Federation, there is still no consensus on which criteria should be used for the diagnosis of MS in adolescence.46 Thus, this research followed the American Academy of Pediatrics’s recommendation to focus on the CRF individually instead of focusing on the diagnosis of MS.47
The polymorphisms of DRD2 and FTO genes were determined th by the real-time polymerase chain reaction (PCR) method. The rs9939609 and rs6277 SNPs were genotyped using the TaqMan ® Real-Time PCR kit (SNP Genotyping kit, from AppliedBiosystems, USA), following the manufacturer’s concentration guidelines. This protocol included the primer oligonucleotide sequences and ® two fluorophore-labeled TaqMan® minor groove-binding (MGB) ® ™ probes, named VIC® and FAM™.
To evaluate the polymorphism of the ANKK1 gene, the polymerase chain reaction-restriction fragment length polymorphisms (PCR-RFLP) strategy was used, using the restriction enzyme TaqIA whose cutting site contains an SNP C/T (C32806T), following the methodology proposed by Behravan et al.48
The Primers used were: GGTTCCTTGCGACTGCTGT- GAATTT [T/A] GTGATGCACTTGGATAGTCTCTGTT (Sequence 5´→3´) for FTO; TCTTCTCTGGTTTGGCGGGGCT- GTC [G/A] GGAGTGCTGTGGAGACCATGGTGGG ® (Sequence 5´→3´) for DRD2, and ACCCTTCCTGAGTGTCAT- CA (Forward) and ACGGCTGGCCAAGTTGTCTA (reverse) for ANKK1.
Two hundred twenty-six children and adolescents were evaluated, 124 (54.9%) had obesity, and 102 (45.1%) were eutrophic. The two groups were similar in age and sex distribution (49.2% were female in the obese group and 50.9% in the eutrophic group). Table 1 shows the clinical and biochemical parameters and the proportion of obese and eutrophic patients with altered parameters used for the main groups’ subdivision.
All patients were genotyped for the ANKK1 gene; however, due to an insufficient amount of DNA, two eutrophic and six children with obesity were not genotyped for the FTO and DRD2 genes. For the combined analyzes of the three genes, these eight patients were excluded.
There was no statistical difference between the obese and eutrophic groups in the allelic and genotypic distribution of the three genes studied (Table 2). The risk allele of the three genes was more frequent in the obese group concerning the eutrophic group, and the homozygous genotypes for the wild allele of the FTO and ANKK1 genes were more frequent in the eutrophic group, but none of these differences reached statistical significance.
Table 3 and Figure 2 show the evolution of the mean Z Score of the BMI (Z-BMI) within the obese group, according to the number of risk alleles. In the combined evaluation of the three genes, the presence of 0, 1, and 6 risk alleles was disregarded due to the small number of patients. The highest number of risk alleles correlated with a higher Z-BMI for the FTO and ANKK1 genes individually and in the sum of the three genes studied, showing an upward curve of the lowest number of alleles and a lower Z-BMI for the highest number of alleles and higher Z-BMI (Figure 2), with a statistically significant difference for all ranges of the number of risk alleles considered.
Table 1. Clinical and Biochemical Parameters Observed in the Obese and Eutrophic Groups Age (years) 9.7 2.4 Weight (kg) 55.6 19.8 Height (cm) 142.7 13.6 Z-BMI 3.05 0.97 Mother´s BMI (kg/m ) 29.7 6.3
Father´s BMI (kg/m ) 32.4 6.9 Glucose (mg/dL) 87.08 6.4 Insulin (µui/mL) 12.68 6.7 HOMA 2.77 1.53 TC (mg/dL) 170.34 29.9 HDL (mg/dL) 43.25 8.67 LDL (mg/dL) 107.19 29.02 TG (mg/dL) 94.18 44.57
TG 64 (51.6) 60 (48.4) 31 (30.4) 71 (69.6) 0.0012* HDL 83 (66.9) 41 (33.1) 40 (39.2) 62 (60.8) 0,0003* LDL 50 (40.3) 74 (59.7) 25 (24.5) 77 (75.5) 0.0119* Glucose 2 (1.6) 122 (98.4) 1 (0.9) 101 (99.1) 0.679 BP 4 (3.2) 120 (96.8) 0 (0) 102 (100) NA AC 107 (86.3) 17 (13.7) 0 (0) 102 (100) NA MS# 83 (66.9) 41 (33.1) 24 (23.5) 78 (76.5) <0.0001* Subgroup OMU OMH EMU EMH HOMA 78 (62.9) 46 (37.1) 13 (12.7) 89 (87.3) <0.0001* Subgroup ORI OSI Total:N (%) 124 (100%) Legend: AC: Abdominal circumference; BMI: Body mass index; EIS: Eutrophic with insulin sensitivity; ERI: Eutrophic resistant to insulin resistance, EMH: phic metabolically unhealthy; HDL: High density cholesterol; LDL: Low density cholesterol; HOMA: Homeostatic Model Assessment Insulin Resistance; ORI: Obese resistant to insulin; OSI: Obese sensitive to insulin; OMH: Obese metabolically healthy; OMU: Obese metabolically unhealthy; BP: Blood pressure; MS: Metabolic syndrome;TC:Total cholesterol;TG:Triglycerides;; SD: Standard deviation; Z-BMI:Z score of the BMI;# considering 2 or more parameters; *p level with statistical significance.
Table 4 shows the genotypic and allelic distribution of the FTO, DRD2 and ANKK1 genes in the subgroups that were divided according to insulin sensitivity criteria: ORI, OSI, ERI, and ESI.
Regarding the FTO gene, in the eutrophic subgroups, the wild homozygous genotype (TT) was more prevalent in the insulin-sensitive group-ESI than in the insulin-resistant group-ERI (36.8×23%, p=0.039). The risk allele (A) was more prevalent in the ERI group (57.7%) than in the ESI group (37.4%), with an odds ratio (OR) of 2.29 for IR when the A allele was present (p=0.048).
Regarding the DRD2 gene, we did not observe any statistically significant difference in the evaluation of the groups together or separately.
Regarding the ANKK1 gene: the ERI group showed a
10.1 2.6 0.114 30.4 10.1 < 0.0001* 135.3 15.7 0.121 - 0.45 0.81 < 0.0001* 24,2 3.3 < 0.0001* 26.9 4.7 0.003* 86.4 6.7 0.337 6.28 2.7 < 0.0001* 1.32 0.6 < 0.0001* 162.05 26.5 0.13 48.23 10.01 0.1 97.37 24.08 0.186 73.65 27.12 < 0.0001*
ERI ESI 102 (100%)
Eutrophic metabolically healthy; EMU: Eutro-
higher proportion of the homozygous genotype for the risk allele (TT) when compared to ESI (15.4%×13.5%, p<0.00001).
The evaluation of the genotypic and allelic distribution of the FTO, DRD2 and ANKK1 genes in the groups subdivided according to the CRF number is shown in Table 5.
Regarding the FTO gene, in the evaluation of the obese subgroups, we observed that the wild homozygous (TT) genotype is present in 37.5% of the metabolically healthy group (OMH) and 29.5% of the metabolically sick group (OMU) (p=0.02).
Regarding the DRD2 gene, in the evaluation of the obese subgroups, the homozygous genotype for the non-risk allele (AA) is more prevalent in the healthy subgroup – OMH than in the subgroup with two or more CRF-OMU (32.5%×15.4%, p=0.053). In
Table 2. Genotypic and Allelic Distribution of FTO, DRD2 and ANKK1 genes in the Obese and Eutrophic Groups. (Risk alleles: FTO=A, DRD2=G, ANKK1= T)
Genotypes AA 25 (21.2) 15 (15) AT 55 (46.6) 50 (50) TT 38 (32.2) 35 (35) Total 118 (100) 100 (100) Alleles A 105 (44.5) 80 (40) T 131 (55.5) 120 (60) 0.89 0.344 Total 236 (100) 200 (100)
Genotypes GG 43 (36.4) 29 (29) AG 50 (42.4) 50 (50) AA 25 (21.2) 21 (21) Total 118 (100) 100 (100) Alleles G 136 (57.6) 108 (54) A 100 (42.4) 92 (46) 0.577 0.4471 Total 236 (100) 200 (100)
Genotypes TT 13 (10.5) 14 (13.7) CT 45 (36.3) 29 (28.5) CC 66 (53.2) 59 (57.8) Total 124 (100) 102 (100) Alleles T 71 (28.6) 57 (28) C 177 (71.4) 147 (72) 0.026 0.871 Total 248 (100) 204 (100) Legend: A: Adenine; C: Cytosine; G: Guanine;T:Thymine.
Figure 2. Mean BMI Z-Score in the Obese group According to the Number of Risk Alleles of the FTO, DRD2 and ANKK1 Genes A) FTO, DRD2 and ANKK1 genes Separately
Table 3. Mean BMI Z score in the Obese Group, According to the Number of Risk Alleles of the FTO, DRD2 and ANKK1 Genes Individually and of the 3 Genes Combined 1.38 0.500
* 0 (TT) 2.92 0.68 0.000 * 1 (AT) 3.06 1.09 0.000 * 2 (AA) 3.21 1.05 0.000
0 (AA) 3.21 1.12 0.453 1 (AG) 3.00 1.05 0.574 2 (GG) 3.01 0.76 0.404 1.594 0.450 ANKK1
* 0 (CC) 2.95 0.84 0.000 * 1 (CT) 3.15 1.17 0.000 * 2 (TT) 3.16 0.76 0.000
* 2 2.89 0.93 0.000 * 3 3.09 1.08 0.000 * 4 3.13 1.09 0.000 * 5 3.45 0.64 0.000 Legend: A: Adenine; C: Cytosine; G: Guanine; SD: Standard deviation;T:Thymine; 1.763 0.414 * Z-BMI:Z score of the BMI; p level with statistical significance.
B) FTO, DRD2 and ANKK1 genes Combined
Table 4. Genotypic and Allelic Distribution of FTO, DRD2 and ANKK1 genes in the Groups: ORI (Obese Resistant to Insulin), OSI (Obese Sensitive to Insulin), ERI (Eutrophic Resistant to Insulin Resistance) and ESI (Eutrophic Sensitive to Insulin)
Genotypes AA 15 (20) 10 (23.3) 5 (38.5) 10 (11.5) AT 36 (48) 19 (44.2) 5 (38.5) 45 (51.7) 7.15 a0.039 TT 24 (32) 14 (32.5) 3 (23) 32 (36.8) AA+AT 51 (68) 29 (67.5) 10 (77) 55 (63.2) TT 24 (32) 14 (32.5) 3 (23) 32 (36.8) Total 75 (100) 43 (100) 13 (100) 87 (100) Alleles A 66 (44) 39 (45.3) 15 (57.7) 65 (37.4) T 84 (56) 47 (54.7) 11 (42.3) 109 (62.6) 4.76 a0.048 Total 150 (100) 86 (100) 26 (100) 174 (100)
Genotypes GG 27 (36) 16 (37.2) 5 (38.5) 24 (27.6) AG 28 (37.3) 22 (51.2) 5 (38.5) 45 (51.7) 6.54 0.3649 AA 20 (26.7) 5 (11.6) 3 (23) 18 (20.7) GG+AG 55 (73.3) 38 (88.4) 10 (77) 69 (79.3) AA 20 (26.7) 5 (11.6) 3 (23) 18 (20.7) Total 75 (100) 43 (100) 13 (100) 87 (100) Alleles G 82 (54.7) 54 (62.8) 15 (57.7) 93 (53.4) A 68 (45.3) 32 (37.2) 11 (42.3) 81 (46.6) 2.20 0.530 Total 150 (100) 86 (100) 26 (100) 174 (100)
Genotypes TT 9 (11.5) 4 (8.7) 2 (15.4) 12 (13.5) CT 28 (35.9) 17 (37) 3 (23.1) 26 (29.2) 2.18 0.414 CC 41 (52.5) 25 (54.3) 8 (61.5) 51 (57.3) TT+CT 37 (47.5) 21 (45.7) 5 (38.5) 38 (42.7) CC 41 (52.5) 25 (54.3) 8 (61.5) 51 (57.3) Total 78 (100) 46 (100) 13 (100) 89 (100) Alleles T 46 (29,5) 25 (27,2) 7 (26,9) 50 (28,1) C 110 (70,5) 67 (72,8) 19 (73,1) 128 (71,9) 0.19 0.978 Total 156 (100) 92 (100) 26 (100) 178 (100) Legend: A: Adenine; C: Cytosine; EIS: Eutrophic with insulin sensitivity; ERI: Eutrophic resistant to insulin; G: Guanine; ORI: Obese resistant to insulin; OSI: Obese sensitive to insulin;T:Thymine;a – p value in the comparison between ERI and ESI.
patients with obesity, the genotypes with the presence of the risk allele of the DRD2 gene (GG+AG) increased the risk for metabolic disease by 2.65 times (p=0.031).
Regarding the ANKK1 gene, when we compared the “healthy obese” group with the “unhealthy eutrophic” group, we observed that the homozygous genotype for the risk allele (TT) was twice as frequent in the EMU group as in the OMH group
Table 5. Genotypic and Allelic Distribution of the FTO, DRD2 and ANKK1 Genes in the Groups: OMU (Obese MetaBolically Unhealthy), OMH (Obese Metabolically Healthy), EMU (Eutrophic Metabolically Unhealthy) and EMH (Eutrophic Metabolically Healthy)
Genotypes AA 12 (15.4) 13 (32.5) 4 (17.4) 11 (14.3) AT 43 (55.1) 12 (30) 12 (52.2) 38 (49.3) 9.88 a0.020 TT 23 (29.5) 15 (37.5) 7 (30.4) 28 (36.4) AA + AT 55 (70.5) 25 (62.5) 16 (69.6) 49 (63.6) 1.23 0.745 1.14 0.765 TT 23 (29.5) 15 (37.5) 7 (30.4) 28 (36.4) Total 78 (100) 40 (100) 23 (100) 77 (100) Alleles A 67 (42.9) 38 (47.5) 20 (43.5) 67 (42.9) T 89 (57.1) 42 (52.5) 26 (56.5) 89 (57.1) 1.63 0.650 Total 156 (100) 80 (100) 46 (100) 154 (100)
Genotypes GG 28 (35.9) 15 (37.5) 7 (30.4) 22 (28.6) AG 38 (48.7) 12 (30) 12 (52.2) 38 (49.4) 7.53 a0.053 AA 12 (15.4) 13 (32.5) 4 (17.4) 17 (22) GG + AG 66 (84.6) 27 (67.5) 19 (82.6) 60 (88) 4.88 a0.031 3.75 0.289 AA 12 (15.4) 13 (32.5) 4 (17.4) 17 (22) Total 78 (100) 40 (100) 23 (100) 77 (100) Alleles G 94 (60.3) 42 (52.5) 26 (56.5) 82 (53.2) A 62 (39.7) 38 (47.5) 20 (43.5) 72 (46.8) 2.02 0.567 Total 156 (100) 80 (100) 46 (100) 154 (100)
Genotypes TT 49 (59) 4 (9.8) 5 (20.8) 44 (56.4) CT 25 (30.1) 20 (48.8) 4 (16.7) 25 (32.1) 9.01 b0.031 CC 9 (10.9) 17 (41.4) 15 (62.5) 9 (11.5) TT + CT 34 (41) 24 (58.5) 9 (37.5) 34 (53.6) 0.60 0.895 4.18 0.242 CC 49 (59) 17 (41.5) 15 (62.5) 44 (46.4) Total 83 (100) 41 (100) 24 (100) 78 (100) Alleles T 43 (25.9) 28 (34.1) 14 (29.2) 43 (27.6) C 123 (74.1) 54 (65.9) 34 (70.8) 113 (72.4) 0.9 0.591 Total 166 (100) 82 (100) 48 (100) 156 (100) Legend: A: Adenine; C: Cytosine; EMH: Eutrophic metabolically healthy; EMU: Eutrophic metabolically unhealthy; G: Guanine; OMH: Obese metabolically healthy; OMU: Obese metabolically unhealthy;T:Thymine; a – p value in the comparison between OMU and OMH,b – p value in the comparison between OMH e EMU.
(20.8%×9.8%, p=0.031).
The polygenic nature of common obesity discovers risky genes and their variants a challenging task. Genome-Wide Association Study (GWAS) studies have brought new insights into the understanding of polygenic obesity, but even today, specific genes’ contribution
to BMI variability is still poorly understood.6 No GWAS study has included the Brazilian population, and studies in the pediatric age group are scarce. Although carried out with a small number of patients and with groups composed following the simple random sampling strategy without replacement, the present research contributed to the knowledge of genetic factors that interfere in the phenotype of pediatric obesity in Brazilians.
The presence of higher-levels of parent’s BMI of the obese children corroborates studies that indicate a vital genetic 49-51 component in human obesity. In addition to the genetic factor, these children live in the same environment as their family, therefore, they are exposed to their lifestyle. It is known that children whose both parents are obese have an 80% possibility of being obese, 50% when one parent is obese, and 9% when the parents are not obese. It is worth mentioning that the father’s average BMI is above normal in both groups. The group of eutrophic children has overweight fathers, a fact consistent with the advancement of overweight and obesity in Brazilians, which affects 56.9% of the adult population.
Both obese and eutrophic groups had similar blood glucose levels. The obese group, however, showed higher-levels of insulin and HOMA-IR. These insulin sensitivity changes are attributed to obesity per se, which is a significant risk factor for diabetes mellitus (DM) type 2. The worsening of the situation of the pancreatic malfunction is related to obesity evolves to DM.
We observed that TG levels were higher in the obese group, an expected result since these children also had elevated insulin and HOMA. Hypertriglyceridemia is a pervasive alteration in patients with obesity and is related to insulin resistance: obese patients with IR, even without DM2 installed, have selective hepatic IR: insulin cannot suppress hepatic glucose production while continuing to perpetuate lipogenesis, resulting in hypertriglyceride- 57,58 mia.
The other lipid profile parameters, the mean TC, highdensity cholesterol (HDL), and low-density cholesterol (LDL), did not show a statistical difference. The mean levels of TC and LDL were higher in the obese group but still within the normal range, while the mean value of HDL was lower in the obese group and was in the sub-normal range. When assessing the number of individuals in each group with LDL and HDL outside the normal range (Table 1), a significantly higher proportion of obese people was in this condition. Indeed, a similar means between the obese and eutrophic groups might be due to the presence of 33.1% of “Healthy Obese” (OMH group) and 23.5% of “Sick Eutrophic,” also called “false thin” or “obese eutrophic,” that composed the EMU group.
It was observed that the risk allele of the three genes was more frequent in the obese group and that the homozygous genotype for the non-risk allele of the FTO and ANKK1 genes was more frequent in the eutrophic, but none of these differences reached statistical significance. The small number of participants evaluated may justify the lack of statistical significance. When evaluating the effect of the presence of the risk alleles within the obese group, we observed that the higher the number of risk alleles, the greater the Z-BMI for the FTO and ANKK1 genes alone, and in the sum of the 3 genes studied. The upward curve of the lowest number of alleles and lowest Z-BMI for the highest number of alleles and the highest Z-BMI, with a statistically significant difference for all ranges of the number of risk alleles considered, indicate that the studied polymorphisms have an additive effect on the outcome of childhood-onset obesity. This result is unprecedented in the literature.
The rs9930506 polymorphism of the FTO gene is the SNP that shows the most significant association with human polygenic obesity: a recent meta-analysis study after evaluating 3337 obese and 3159 controls confirms its correlation with obesity in European populations. Few studies have evaluated this SNP in the Brazilian population, and only seven were carried out in the pediatric population of our country, five have found an association with 59-63 adiposity and 2 showed no difference between patients with or 64,65 without obesity.
Our study showed that in the Eutrophic group, the presence of wild homozygous genotype (TT) associates with insulin sensitivity, and the presence of the risk allele (A) associates with 54-56 an increase of 2, 29 times in the risk for IR. This SNP has already been associated with insulin sensitivity changes in insulin sensitivity.66 but this is the first time this result is presented in Brazilian children. Within the obese group, the TT genotype was more frequent in the OMH group (37.5%) than in the OMU group (29.5%); this difference is statistically significant, so we can say that the T allele is associated with the absence of CRF, which can be considered a protective factor within the obese group, a result also unprecedented in the literature for the Brazilian population.
The rs6277 is a poorly studied SNP: only six scientific studies have evaluated the rs6277 variant of the DRD2 gene in children or ado- 67-71 lescents to assess cognitive control and behavioral nuances and choke risk.72
From the metabolic point of view, the rs6277 polymorphism of the DRD2 gene was evaluated in only four articles, all performed in adults. This SNP was associated with compulsive eat- 17,18 19 ing disorder, weight gain, and higher sugar consumption.73,74 The prevalence of the GG genotype varied between 12 and 21% in these studies. Our series found a higher frequency of genotypes with the risk allele (G), GG-36.4 and 29%, AG-42.4%, and 50% in obese and eutrophic individuals respectively, and AA - 21% for both groups. The highest frequency of the G allele found in our study is probably due to the great miscegenation of the Brazilian population, data corroborated by the 1000 Genome Project,75 which shows a frequency of the G allele of 12.5% in Europeans, 8% in Asians, and 93% in Africans.
Ramos-Lopez et al associated the GG genotype with
significantly higher-levels of TG in adults. The present study associated the G allele of the SNP rs 6277 of the DRD2 gene with CRF in obese children and adolescents, which was an unprecedented result in the literature.
We observed that the frequency of the homozygous genotype for the wild allele (CC) was slightly higher in the eutrophic group (57.8%) than in the obese group (53.2%). However, there was no statistically significant difference in the allelic or genotypic distribution. This result disagrees with a previous study carried out by our team that found an association of the T-allele with the presence of 74,75 obesity in children. However, when assessing the evolution of the mean Z-BMI in the obese group, we observed that the higher the number of risk alleles, the higher the Z-BMI, corroborating the previous finding that the T allele is associated with increased BMI.
Few studies have been carried out to verify the association of the rs1800497 polymorphism of the ANKK1 gene in children and adolescents. A considerable variation in allele frequencies was observed in the literature, even within populations of the same country, and inconsistency regarding the association with metabol- 12,76-79 ic outcomes.
In evaluating the subgroups using the HOMA values criterion, the risk allele’s homozygous genotype (TT) was associated with insulin resistance in eutrophic individuals. This result corroborates the finding of previous research carried out by our group, which for the first time in the literature, associated the T-allele of the ANKK1 gene with alteration of glycemic homeostasis.
Clinical and animal studies evidence the relationship between DRD2 receptors and glucose metabolism. Clinical research 80 81 conducted in diabetic patients and animal study demonstrated improved glycemic control using Bromocriptine, a dopaminergic agonist that acts via DRD2 receptors. In 2005, RUBI et al demonstrated for the first time that DRD2 receptors are expressed in the pancreatic beta-cell and modulate insulin secretion. A study carried out in knockout rats for the DRD2 gene revealed that DRD2 receptors play a crucial role in insulin secretion and glycemic homeostasis; rats with no DRD2 receptor showed a failure in insulin response in the face of glucose overload, higher fasting glucose, glucose intolerance, and reduced beta-cell mass. Such results show that DRD2 is essential for the proliferation of beta cells and insulin secretion and can be considered a growth factor essential for glycemic homeostasis.83
Regarding the presence of CRF, the risk allele of the ANKK1 gene was associated with the presence of two or more CRF, as shown by the presence of the homozygous risk genotype (TT) in 20.8% of EMU and only 9.8% of OMH, a result also found to be unprecedented in the literature.
In conclusion, our results found that the risk alleles of the FTO, DRD2 and ANKK1 genes interfered with the outcome of pediatric obesity in Brazilian children:
• The higher number of the risk alleles of FTO and ANKK1 genes and the 3 genes combined significantly increased the mean Z-BMI of the obese group.
• The risk alleles of the FTO and ANKK1 genes were positively associated with IR in eutrophic children.
• The T allele of the FTO gene has a protective effect against CRF in children with obesity • The risk alleles of the DRD2 and ANKK1 genes conferred a higher risk of CRF in children with obesity and normal weight, respectively.
For being a case-control, our findings cannot ascribe causality, only association. Further studies are needed, with a more significant number of patients and longitudinal follow-up, to confirm the possible causality between the risk alleles and our findings.
It is hoped that in the future, more excellent knowledge of the specific contribution of each genetic variant will be a useful tool in clinical practice, being able to guide preventive measures and specific treatment according to genetic risk scores.
This research was supported by grants of CAPES and FAPEG/ DECID/MS
• RMP and ADC designed the research (project conception, devel- 76 opment of overall research plan, and study oversight). • RMP conducted the clinical evaluation JSF, RVM, and LBM conducted the genetic protocols (hands-on conduct of the experiments).
• RMP, LBM, and ADC analyzed data or performed statistical analysis.
82 • RMP, MPC, and ADC interpreted data findings.
• RMP, JSF, MPC, and ADC wrote the paper.
• RMP and ADC had primary responsibility for final content. All authors have read and approved the final manuscript.
83 The authors declare that they have no conflicts of interest.