5-HT2A Receptor Gene Promoter Polymorphism in Relation to Abdominal Obesity and Cortisol
Department of Clinical Chemistry, Sahlgrenska University Hospital, S-41345 Göteborg, Sweden. E-mail: firstname.lastname@example.org
Objective: There is considerable evidence that cortisol secretion is associated with obesity. The regulation of the 5-hydroxytryptamine receptor 2A (5-HT2A) gene might play an essential role because it is involved in the control of cortisol secretion. Therefore, we examined the potential impact of the 5-HT2A −1438G/A promoter polymorphism on obesity and estimates of insulin, glucose, and lipid metabolism as well as circulating hormones, including salivary cortisol, in 284 unrelated Swedish men born in 1944.
Research Methods and Procedures: The subjects were genotyped by using polymerase chain reaction amplification of the promoter region of the gene for 5-HT2A followed by digestion of the reaction product with the restriction enzyme MspI.
Results: The frequencies were 0.39 for allele −1438A and 0.61 for allele −1438G. Homozygotes for the −1438G allele had, in comparison with −1438A/A subjects, higher body mass index, waist-to-hip ratio, and abdominal sagittal diameter. Moreover, cortisol escape from 0.25-mg dexamethasone suppression was found in subjects with the −1438A/G genotype. Serum leptin, fasting insulin, and glucose, as well as serum lipids, were not different across the −1438G/A genotype groups.
Discussion: From these results, we suggest the possibility that an abnormal production rate of the 5-HT2A gene product might lead to the development of abdominal obesity. The pathophysiology could involve stress factors that destabilize the serotonin-hypothalamic-pituitary-adrenal system in those with genetic vulnerability in the serotonin receptor gene.
Serotonin (5-hydroxytryptamine; 5-HT) is a neurotransmitter that contributes to the regulation of many physiological processes such as sleep, appetite, and hormone secretion (1,2,3). Abnormalities of the serotonergic system have been implicated in a number of human diseases including obesity (4,5). Like other neurotransmitters, 5-HT is released into the synaptic junction and exerts its effect on specific receptors on postsynaptic membranes (6). Over the years, a multitude of such 5-HT receptors have been identified, including the 5-HT2A receptor (6).
Excessive fat accumulation (obesity) affects many aspects of human health, and extensive research has disclosed that the abdominal distribution of body fat carries the greatest risk for type 2 diabetes, stroke, and coronary artery disease (7). Our knowledge of human obesity has progressed beyond the simple generalization that abdominal obesity is fully explained by inappropriate eating and physical inactivity. Over the last decade, a considerable amount of evidence has emerged regarding the pathogenic role of cortisol in abdominal obesity (8). That 5-HT is involved in regulating cortisol secretion has long been recognized (3), and evidence suggests that cortisol secretion is regulated by central 5-HT2A/2C receptors (9).
Lately, an MspI restriction fragment length polymorphism in the promoter region of the 5-HT2A gene (−1438G/A) has been described (10). The human 5-HT2A gene is located on 13q14-q21 (11) and consists of three exons separated by two introns and spans over 20 kb (12). More recent data suggest that the MspI polymorphism of the 5-HT2A gene may influence food and alcohol intake in obese subjects (13). In addition, some, but not all studies, have indicated a role for the −1438G/A variant in the pathogenesis of anorexia nervosa (10,14,15,16,17,18,19,20).
In this study, we addressed the hypothesis that the −1438G/A promoter variant of the of the 5-HT2A gene is involved in the pathogenesis of abdominal obesity and related perturbations in insulin, glucose, and lipid metabolism, as well as circulating hormones including salivary cortisol.
Research Methods and Procedures
In this study, we recruited subjects from an ongoing cohort study of men born in 1944 (21,22). The study was initiated in 1992. Based on self-measured waist-to-hip ratio (WHR) the following three subgroups, each of 150 men, were selected for further studies: the lowest values (≤0.885), the highest values (≥1.01), and the arithmetic mean (0.94 to 0.96). In 1995, we examined these men (age 51 years), and 284 (63%) volunteered to participate (23,24). Nine subjects refused to be involved in the genetic studies, and in 11 subjects, the amount of blood samples was not sufficient to conduct genetic analysis. All men gave written informed consent before participating in the study, which was approved by the Göteborg University Ethics Committee.
Body weight, height, circumferences, and abdominal sagittal diameter were measured as described previously (23,24).
Venous blood was obtained after overnight fasting. Commercial radioimmunoassay kits was used for the determination of serum testosterone, insulin-like growth factor I, insulin, and leptin. Glucose and serum lipids were determined enzymatically as detailed previously (23,24).
Diurnal cortisol secretion was measured by a series of saliva sampling over an ordinary working day, in which cortisol levels were measured. Additionally, a dexamethasone suppression test was done at home, using 0.25 or 0.5 mg dexamethasone. The details of the procedures have been described previously (23,24).
Determination of Genotypes
Genotyping of the −1438G/A polymorphism of the 5-HT2A gene was carried out by polymerase chain reaction (PCR) and restriction digestion as described previously (10). Genomic leukocyte DNA (100 ng in a final volume of 10 μL) was amplified by PCR using the following primers: 5′-AAGCTGCAAGGTAGCAACAGC-3′ and 5′-AACCAACTTATTTCCTACCAC-3′ (10). The primers amplified a product of 468 bp. The PCR conditions were as follows: an initial denaturation step at 94 °C for 3 minutes, followed by 35 cycles of denaturation at 94 °C for 15 seconds, annealing at 55 °C for 15 seconds, and extension at 72 °C for 30 seconds, with a final extension of 10 minutes at 72 °C. The PCR reaction product was digested at 37 °C overnight with 5 U of the restriction enzyme MspI (New England Biolabs, Beverly, MA). The −1438G allele was cut into 244-bp and 224-bp fragments, whereas the −1438A allele remained undigested. The fragments were separated on a 2% agarose gel.
Statistical differences between genotypes were tested using a one-way analysis of covariance model including genotype as independent factors and body mass index and WHR as covariates. To determine which mean values differed (p < 0.05), we used the Duncan multiple-range test. All statistical analyses were performed using the SAS System for Windows, release 8.0 (SAS Institute Inc., Cary, NC).
The frequency of allele −1438A was 0.39 and 0.61 for allele −1438G. The observed genotype frequencies were 35.6%, 51.6%, and 12.9% for −1438A/A, −1438A/G, and −1438G/G, respectively. Genotype frequencies were in a Hardy-Weinberg equilibrium.
Table 1 presents the differences in anthropometric, endocrine, and metabolic measurements among the 5-HT2A −1438G/A genotypes. Homozygotes for the −1438G allele had, in comparison with −1438A allele homozygotes, higher body mass index, WHR, and abdominal sagittal diameter. Serum leptin, fasting insulin, and glucose, as well as serum lipids, were not different across the −1438G/A genotype groups.
Table 1. Differences in anthropometric, endocrine, and metabolic measurements by 5-HT2A −1438G/A genotypes
|Body mass index (kg/m2)||25.1 (3.5)a||26.7 (4.0)a||26.9 (4.3)b||0.017|
|Waist-to-hip ratio||0.92 (0.06)a||0.94 (0.07)b||0.96 (0.06)b||0.015|
|Abdominal sagittal diameter (cm)||21.9 (3.4)a||22.9 (3.8)b||23.6 (3.2)b||0.039|
|Testosterone (nM)||20.2 (5.0)||19.6 (5.9)||19.0 (5.1)||>0.20|
|Insulin-like growth factor I (μg/L)||208.4 (56.4)||204.0 (71.0)||203.3 (62.1)||>0.20|
|Leptin (μg/L)||5.4 (4.1)||6.4 (4.3)||7.3 (4.4)||>0.20|
|Fasting insulin (mU/L)||10.4 (6.1)||13.5 (13.0)||13.8 (10.1)||>0.20|
|Fasting glucose (mM)||4.4 (0.7)||4.6 (1.1)||4.4 (0.6)||>0.20|
|Triglycerides (mM)||1.8 (0.9)||1.8 (1.2)||1.8 (0.7)||>0.20|
|Total cholesterol (mM)||6.1 (1.0)||6.2 (1.1)||6.2 (1.2)||>0.20|
|HDL cholesterol (mM)||1.2 (0.3)||1.3 (0.4)||1.2 (0.3)||0.113|
|LDL cholesterol (mM)||4.0 (1.0)||4.2 (1.0)||4.2 (1.1)||>0.20|
Table 2 shows the results of the differences in salivary cortisol measurements by genotype of the MspI polymorphism of the 5-HT2A promoter. Differences in diurnal salivary cortisol variables between the three genotype groups were not significant. However, subjects with the −1438A/G genotype showed significantly less suppression of cortisol with 0.25-mg dexamethasone compared with −1438A/A homozygotes (p = 0.021).
Table 2. Differences in salivary cortisol measurements by 5-HT2A −1438G/A genotypes
|Cortisol level (nM) in the morning||15.6 (8.4)||14.5 (6.8)||14.1 (6.1)||>0.20|
|Cortisol level (nM) at 11:45 am||7.0 (3.3)||7.6 (6.8)||6.2 (2.8)||>0.20|
|Cortisol level (nM) at 30 minutes after lunch||8.4 (6.2)||8.6 (9.9)||6.6 (2.4)||>0.20|
|Cortisol level (nM) at 45 minutes after lunch||7.8 (5.9)||7.5 (5.5)||6.4 (3.0)||>0.20|
|Cortisol level (nM) at 60 minutes after lunch||6.9 (4.3)||7.0 (6.5)||5.7 (1.9)||>0.20|
|Cortisol level (nM) at 5:00 pm||4.7 (2.4)||4.8 (2.4)||4.4 (2.4)||>0.20|
|Cortisol level (nM) before bedtime||3.3 (5.8)||3.3 (3.3)||3.4 (3.4)||>0.20|
|Dexamethasone (0.25 mg) suppression test (nM)||16.6 (18.3)a||5.0 (5.7)b||7.9 (2.3)||0.021|
|Dexamethasone (0.5 mg) suppression test (nM)||11.9 (5.8)||11.9 (5.1)||12.9 (5.3)||>0.20|
The examined men were selected from an ongoing cohort study, and 80% volunteered to participate in the first part of the study. The second part, which was laboratory-based, attracted fewer participants (63%), but there was no difference between nonresponders and responders concerning the prevalence of hypertension, diabetes mellitus, myocardial infarction, stroke and angina pectoris, education level, housing conditions, smoking, and alcohol habits (23,24). We therefore believe that the participating men are representative of men at this age in the city of Göteborg, Sweden.
The assessment of the hypothalamic-pituitary-adrenal (HPA)-axis function in this study included measurements of salivary cortisol levels under basal conditions and after dexamethasone suppression. The assessment of cortisol in saliva is specific for the detection of unbound, free cortisol, and the concentrations of cortisol in saliva is independent of the saliva flow (25). The test is sufficiently sensitive to measure cortisol levels in normal subjects and to distinguish normal secretory pattern from hypo- and hypercortisolism (26,27). Moreover, cortisol in saliva reflects accurately the free fraction of cortisol in plasma (25), which has also been confirmed in our laboratory (r > 0.90, data not shown). From a practical point of view, the assessment of cortisol from saliva samples represents less of a burden for the subjects.
The major findings of this study are that homozygotes for the −1438G allele had increased body mass and abdominal distribution of body fat (WHR and abdominal sagittal diameter) along with less suppression of cortisol with 0.25-mg dexamethasone compared with other MspI 5-HT2A genotypes. In addition, the lack of association between the 0.5-mg dexamethasone test and the MspI polymorphism (Table 2) suggests that subtle physiological alterations in the HPA-axis activity are involved in the development of obesity. These results might be important in understanding the pathophysiology of abdominal obesity. Several features of abdominal obesity suggest that a relative deficiency of serotoninergic effects is involved. These include traits of depression and anxiety (21,28), carbohydrate craving (29), alcohol use, and smoking (7).
We have previously found that a number of psychosocial and socioeconomic handicaps that are known to activate the HPA axis are also associated with abdominal obesity (22,30). The most prominent factors are divorce, solitude, poor economy and low education, unemployment, and problems at work when employed. Interestingly, socioeconomic inequality and low educational level have recently been shown to be associated with elevated stress-related cortisol secretion as well as abdominal obesity (31). One might speculate that various environmental factors in the context of stress might destabilize the serotonin-hypothalamic-pituitary-adrenal system in those with genetic vulnerability. Distressing events or situations evoke the HPA system predominantly, which habituates as exposure to the stressor is repeated (32,33). However, after long-term exposure, the HPA axis eventually becomes dishabituated, resulting in a disruption of central regulatory systems, which, in genetically susceptible individuals, is commonly accompanied by a central arousal syndrome and abdominal obesity, insulin resistance, dyslipidemia, and hypertension (34,35,36,37).
During the course of antidepressant pharmacotherapy of depression, the relative resistance to serum cortisol suppression by exogenous glucocorticoids seems to retreat (4,5). Furthermore, long-term treatment with antidepressant drugs in rats results in decreased basal and stress-related plasma levels of corticotropin and cortisol (4,5). Beside normalization of the HPA-axis activity, serotoninergic drugs increase energy expenditure in humans (4,5). To a considerable extent, serotonin agonists, such as fluoxetine, produce a resumption of the normal eating patterns by diminishing meal size and disrupting night-eating behavior (4,5). Recent data indicate that serotoninergic agents decrease abdominal fat mass and improve glucose and insulin metabolism (4,5).
In summary, the results of our study suggest that more emphasis is needed on the serotonergic system in treatment. Moreover, there is evidence that some selective serotonin reuptake inhibitors are effective as antiobesity drugs. Because abdominal obesity is an unsolved therapeutic problem associated with a high mortality and morbidity, new approaches are clearly warranted. This is particularly important for the individuals who are at risk as a result of a genetic predisposition. The example of the 5-HT2A gene-promoter polymorphism provided here seems to represent one way through which such a predisposition may occur.
This study was supported by grants from the Swedish Medical Research Council (K97–19X-00, 251–35A) and the Pennington Biomedical Research Center. R. Rosmond would also like to acknowledge the Henning and Johan Throne-Holst Foundation for the support of a postdoctoral fellowship at the Pennington Biomedical Research Center. C. Bouchard is partially supported by the George A. Bray Chair in Nutrition.