Association of Obesity, but not Diabetes or ... - Wiley Online Library

Association of Obesity, but not Diabetes or Hypertension, with Glucocorticoid Receptor N363S Variant Ruby C.Y. Lin,* Xing Li Wang,† Bronwen Dalziel,‡ Ian D. Caterson,‡ and Brian J. Morris*

Abstract LIN, RUBY C.Y., XING LI WANG, BRONWEN DALZIEL, IAN D. CATERSON, AND BRIAN J. MORRIS. Association of obesity, but not diabetes or hypertension, with glucocorticoid receptor N363S variant. Obes Res. 2003;11:802-808. Objective: To determine whether the N363S variant in the glucocorticoid receptor (encoded by nuclear receptor subfamily 3, group C, member 1: NR3C1) is associated with obesity, type 2 diabetes, or hypertension. Research Methods and Procedures: This was a crosssectional case-control study involving 951 Anglo-Celtic/ Northern European subjects from Sydney. This study consisted of the following: 1) an obesity clinic group, most of whom had “morbid obesity” (mean BMI for group ⫽ 43 ⫾ 8 kg/m2; n ⫽ 152); 2) a type 2 diabetes clinic group (n ⫽ 356); 3) patients with essential hypertension who had a strong family history (n ⫽ 141); and 4) normal healthy controls (n ⫽ 302). N363S genotype, BMI, and a range of other parameters relevant to each group were measured. Results: Compared with the frequency of 0.04 in nonobese healthy subjects, the S363 allele was significantly higher in obesity clinic patients (0.17; p ⫽ 5.6 ⫻ 10– 8), subjects with diabetes who were also obese (0.09; p ⫽ 0.0045), subjects with hypertension who were also overweight (0.08; p ⫽ 0.0016), and overweight healthy subjects (0.12; p ⫽ 0.0004).

Received for review November 29, 2002. Accepted in final form April 14, 2003. *Basic & Clinical Genomics Laboratory, Department of Physiology, School of Medical Sciences and Institute for Biomedical Research and ‡Human Nutrition Unit, Department of Biochemistry, The University of Sydney, Sydney, Australia; †Cardiovascular Genetics Laboratory, Prince of Wales Hospital, Centre for Thrombosis and Vascular Research, University of New South Wales, Sydney, Australia; and Southwest Foundation for Biomedical Research, San Antonio, Texas. Address correspondence to Professor Brian J. Morris, DSc, Basic & Clinical Genomics Laboratory, Department of Physiology, School of Medical Sciences and Institute for Biomedical Research, Building F13, The University of Sydney, NSW 2006, Australia. E-mail: [email protected] Copyright © 2003 NAASO

802

OBESITY RESEARCH Vol. 11 No. 6 June 2003

Discussion: The NR3C1 N363S variant is associated with obesity and overweight in a range of patient settings but is not associated with hypertension or type 2 diabetes. Key words: glucocorticoid receptor, morbid obesity, type 2 diabetes, essential hypertension, molecular genetics

Introduction The metabolic syndrome is defined as having two or more of a range of conditions that include obesity, type 2 diabetes, and hypertension (1,2) and is caused by an effect of environmental factors on individuals who are genetically predisposed. One of the most important regulators of metabolic and cardiovascular function is the glucocorticoid class of hormones, whose effects are mediated by the glucocorticoid receptor (GR).1 The GR gene (NR3C1) exhibits a number of mutations and polymorphisms (3–12). Point mutations or deletions can reduce intracellular concentration or biological activity of GR in glucocorticoid target tissues (3,5,13). Patients with such NR3C1 defects can present with conditions that include Cushingoid symptoms and signs, hypertension (HT), hypokalemic alkalosis, and/or familial glucocorticoid resistance. It is also known that tissue sensitivity to glucocorticoids is associated with insulin resistance, glucose intolerance, and hypertriglyceridemia (14). In early studies, an association of obesity, abdominal visceral fat, and hyperinsulinemia with an intronic BclI restriction fragment length polymorphism of NR3C1 was made (4,15–17). Since then, other work has shown an association between an N363S variant (18) and elevated BMI in a Dutch cohort (7) and overweight in two different groups of Anglo-Celtic whites (10,19). Studies of Danish subjects (20) and Swedish men (11) proved negative, how-

1 Nonstandard abbreviations: GR, glucocorticoid receptor; HT, hypertension; BP, blood pressure; PCR, polymerase chain reaction; OR, odds ratio.

Glucocorticoid Receptor in Obesity Genetics, Lin et al.

ever, possibly because of genetic differences, which are almost impossible to match between populations (21), differences in ascertainment, or in the interaction between genotype and the very different environmental factors in each geographic region. Other variants in or near NR3C1 have also been tested but showed no association with overweight (22). The N363S variant represents a GR with an extra serine residue. Because phosphorylation of serine and threonine residues is important for DNA binding by GR dimers (23,24), this has the potential to offer an additional phosphorylation site, which might then have biological consequences. The N363S polymorphism is located close to the transactivation domain (␶1), amino acids 77–262 (25), that interacts with co-activator and co-repressor proteins in a transcription complex formation. Moreover, deletion of amino acids 328 –382 from the human mineralocorticoid receptor decreases transcriptional activation markedly (26). This offers potential support for a role for this region in GR. The aim of the present study was to perform an association analysis of the GR N363S variant in obesity clinic patients to see if an association stronger than seen with overweight might exist. Furthermore, in view of the many effects of glucocorticoids, we performed association analyses of the N363S variant in type 2 diabetes and essential HT.

Research Methods and Procedures Study Cohorts All subjects were unrelated Australian whites of AngloCeltic or Northern European ancestry who were residents of Sydney. Obesity and overweight were as defined by the International Obesity Task Force: normal weight, 18.5–24.9 kg/m2; overweight, 25.0 –29.9 kg/m2; class I obesity, 30.0 – 34.9 kg/m2; class II obesity, 35.0 –39.9 kg/m2; and class III obesity, ⱖ40.0 kg/m2 (27). Obesity clinic individuals were interviewed by a dietitian, and diabetic or hypertensive patients were interviewed by a physician for medical history and current medication. Height and weight, as well as blood pressure (BP), taken after a 10-minute rest, were recorded, and a 10-mL fasting blood sample was collected for DNA, plasma lipids, as well as, in the diabetes group, glucose and insulin. The study received approval from the Ethics Review Board, and all subjects gave informed consent. Specific details of each group follow. Controls. These comprised 302 volunteers recruited from the Sydney Red Cross Blood Bank. They were nondiabetic and had no heart or kidney disease or HT (BP ⬍ 140/90 mm Hg). Their parents were also free of these conditions past the age of 50 years. This was determined by completion of a questionnaire during interview by the investigators. Based on BMI and health, the entire group was suitable as controls for each patient group. This cohort has, moreover, been used as controls for a large number of genetic studies over the

past decade, and, in all cases, allele frequencies for polymorphisms tested were similar to those reported by others for healthy white populations elsewhere. Healthy weight and overweight subgroups were selected from these subjects. Obesity Clinic Subjects. There were 157 subjects with BMI ⱖ30 kg/m2 recruited by the Metabolism and Obesity Services Unit, Royal Prince Alfred Hospital, Sydney, while attending the clinic for initial assessment. Most had class III (“morbid”) obesity (mean BMI, 43 ⫾ 8 kg/m2). Patients with diabetes, congestive heart failure, kidney disease, and weight loss over the previous 3 months were excluded from the study. Although we eliminated subjects with diabetes, some were glucose intolerant. Moreover, for those who were included, fasting glucose and insulin levels between the genotypes did not differ (data not shown). Percentage of body weight attributable to fat was determined by bioelectrical impedance analysis using a low-frequency current (50 kHz) analyzer (BIA-101; RJL Systems, Detroit, MI). Type 2 Diabetes Patients. There were 359 diabetes patients who had been diagnosed by National Diabetes Data Group criteria and enrolled consecutively at the Diabetes Center at the Prince of Wales Hospital, Sydney, from May 1996 to May 1997. No patients were excluded unless they declined participation themselves. Among 333 patients with proper documentation, 130 (40.8%) had a positive family history, 61 (18.3%) had vascular disease, 68 (22.9%) had neuropathy, 43 (18.9%) had retinopathy, 23 (7.0%) had nephropathy, and 84 (29.5%) had microalbuminuria. Genotype data were obtained for 356 of the diabetes subjects after exclusion of 3 whose samples did not yield suitable polymerase chain reaction (PCR) products. Hypertensive Subjects. These consisted of 141 recruits from community advertising in which one selection criterion was that both of their parents had to have had HT. All lacked renal disease, coronary artery diseases, heart failure, diabetes, and a diagnosis of secondary HT. The requirement for strong family history not only improved the chance of demonstrating an existing genetic association with HT, but also meant that they had moderate-to-severe (Table 1) earlyonset (35 ⫾ 9 years) HT. Moreover, it meant selection was from a hypertensive population that was 10 times larger (28). Genotyping DNA was isolated from whole blood using a DNA extraction kit (Qiagen, Hilden, Germany) or, for the obesity clinic cohort, a Nucleon BACC DNA extraction kit (Amersham Life Sciences, Buckinghamshire, UK). Genotypes for the A1218G (N363S) NR3C1 variant were determined by PCR-restriction fragment length polymorphism analysis, as described previously (22), except that PCR involved 94 °C for 2 minutes, and then 35 cycles of 94 °C, 64 °C, and 72 °C for 1 minute each, finishing with a step at 72 °C for OBESITY RESEARCH Vol. 11 No. 6 June 2003

803


Table 1. Characteristics of subjects in each group Control

Obese

Diabetes

HT

n Male:female Age (y)

302 173:129 47 ⫾ 11

BMI (kg/m2)

26 ⫾ 4

152 31:121 44 ⫾ 13 (0.16) 43 ⫾ 8 (⬍0.0001) 127 ⫾ 17 (⬍0.0001) 81 ⫾ 13 (⬍0.0001) 5.3 ⫾ 0.1 (0.004) 1.6 ⫾ 0.05 (0.015) 1.3 ⫾ 0.02 (0.99) 3.4 ⫾ 0.1 (0.02)

356 186:170 63 ⫾ 12 (⬍0.0001) 30 ⫾ 7 (⬍0.0001) 142 ⫾ 18 (⬍0.0001) 82 ⫾ 10 (⬍0.0001) 5.6 ⫾ 0.07 (⬍0.0001) 2.1 ⫾ 0.07 (⬍0.0001) 1.2 ⫾ 0.02 (⬍0.0001) 3.5 ⫾ 0.05 (0.64)

141 67:74 54 ⫾ 13 (⬍0.0001) 26 ⫾ 5 (0.95) 173 ⫾ 26* (⬍0.0001) 110 ⫾ 17* (⬍0.0001) 5.8 ⫾ 0.1 (⬍0.0001) 2.5 ⫾ 0.2 (⬍0.0001) 1.1 ⫾ 0.05 (⬍0.0001) 3.6 ⫾ 0.1 (0.99)

Systolic BP (mm Hg)

120 ⫾ 11

Diastolic BP (mm Hg)

73 ⫾ 8

Total cholesterol (mM)

5.0 ⫾ 0.1

Triglycerides (mM)

1.3 ⫾ 0.05

High-density lipoprotein-cholesterol (mM)

1.3 ⫾ 0.03

Low-density lipoprotein-cholesterol (mM)

3.7 ⫾ 0.1

* Pretreatment BPs. Values are mean ⫾ SD or, for lipid measurements, mean ⫾ SE. Shown in parentheses, p values from t tests that compared values for study cohort with control, with Bonferroni corrections for multiple comparisons. ND, not determined.

20 minutes. PCR products were digested at 65 °C for 1 hour with 1 U of Tsp509I (New England BioLabs, Beverley, MA). Random samples were phenol-chloroform purified and cloned using a pGEM-T Easy Vector System (Promega, Madison, WI) and then sequenced (Australian Genome Research Facility, Brisbane, Australia) to confirm accuracy of genotyping. In addition, 106 were subjected to melt curve genotyping analysis using real-time PCR (Rotorgene; Corbett Research, Sydney, Australia) with FAM and Cy5 labeled probes and the following primers: forward, 5⬘TTCTCAACAGCAGGATCAGAAGC-3⬘; reverse, 5⬘-TGTTCGACCAGGGAAGTTCAGAG-3⬘. Genotypes were assigned according to the appearance of peaks in the melt curves. Statistical Analysis Data were tested by ␹2, one-way ANOVA, and logistic regression analyses using StatView (Abacus Concepts, Berkeley, CA) and SPSS v9.0 for Windows (SPSS Inc., Chicago, IL). In some subgroup analyses, S363 carriers were grouped together because of low S/S homozygote prevalence. Power (at 0.005 endpoint) indicated adequate 804


sample size of each cohort: obesity clinic group 72%, type 2 diabetes 88%, hypertensives 90%, and healthy controls 99%, where each is influenced by relative risk of the condition, a parameter that differs for each group. Odds ratio (OR) was calculated as described (29).

Results Group Characteristics Table 1 shows group characteristics. Hardy-Weinberg equilibrium was seen for each group: control, ␹2 ⫽ 6.0 (p ⫽ 0.051); obesity clinic, ␹2 ⫽ 3.3 (p ⫽ 0.19); diabetes, ␹2 ⫽ 3.2 (p ⫽ 0.21); and hypertensive, ␹2 ⫽ 1.9 (p ⫽ 0.39). The frequency (0.04) noted for S363 in normal-weight controls was similar to values reported by others of 0.03 (7,18), 0.04 (20), and 0.05 (11) for healthy white Anglo-Celtic or Northern European subjects. Association Analyses in Diabetes and Hypertension Table 2 presents genotype data for each group and comparisons with the controls. No association was seen for type 2 diabetes (OR ⫽ 1.0; 95% CI, 0.6 to 1.2) or hypertension


Table 2. Association analyses of NR3C1 N363S variant in type 2 diabetes and hypertension Total alleles on all chromosomes (frequency)

Genotypes (frequency) Group

N/N

Control

263 (0.87) N/N

31 8 (0.10) (0.03) N/S ⫹ S/S

263 303 (0.85) 303 129 (0.91) 129

39

Diabetes

Hypertension

N/S

S/S

46 (0.13)

7 (0.02) 52

10 (0.07)

2 (0.02) 12

(OR ⫽ 1.0; 95% CI, 0.9 to 1.2), the latter being consistent with our previous study that involved fewer subjects (21). S363 allele frequency was higher in male (0.08) than in female (0.02) hypertensives (␹2 ⫽ 5.5, p ⫽ 0.020). Failure to detect association with hypertension also applied in the obese and type 2 diabetes cohorts. In the diabetes group, logistic regression analysis, in which other variables including age, gender, BMI, waist-to-hip ratio, and cigarette smoking were controlled for as confounding factors, was also unable to reveal an association with diabetes. S363 in males (0.07) and females (0.10) was similar, and there was no association of microalbuminuria, myocardial infarction, angina, hypertension, neuropathy, nephropathy, or retinopathy with genotype. After Bonferroni corrections, there was no association of BP, plasma lipids, and age of onset or duration of diabetes with the S363 allele. Presence of diabetes in the obese cohort also showed no association with S363. Further analyses relevant to body weight for the diabetes and hypertensive groups are shown below. Association with Obesity We found that S363 frequency in the obesity clinic group (0.17) was significantly higher than in normal-weight controls (0.04) but was not higher than in overweight controls (0.12; Table 3; OR for S363 carriers to confer obesity ⫽ 1.7; 95% CI, 1.1 to 2.7). Male and female values did not differ for obesity clinic (0.23 vs. 0.15, p ⫽ 0.12), overweight, or normal-weight groups. Lipid-lowering treatment, common in subjects with BMI ⱖ35 kg/m2, did not differ between genotypes. Genotype data in obese subjects with normal BP (AA ⫽ 0.75, AG ⫽ 0.19, GG ⫽ 0.06) were similar to obese with hypertension (AA ⫽ 0.66, AG ⫽ 0.27, GG ⫽ 0.07; within the group, p ⫽ 0.54). After Bonferroni

␹2

p

1.4

0.50

0.41 1.9

0.52 0.39

1.8

0.18

␹2

p

60 (0.08)

0.18

0.67

14 (0.05)

2.4

0.12

N

S

557 (0.92)

47 (0.08)

652 (0.92) 268 (0.95)

corrections, we saw no association between either BP or plasma lipids and genotype, but we did observe significantly higher waist circumference in S363 carriers vs. N363 homozygotes (118.6 vs. 117.3 cm, respectively; Bonferronicorrected p ⫽ 0.012). In addition, total body fat in S363 carriers was lower (40.5%) than in N363 homozygotes (46.2%; Bonferroni-corrected p ⫽ 0.030). Furthermore, subgroup analyses controlled for gender did not show any significant sex-specific difference across genotypes. Analysis of Obese Subgroups of Diabetes and Hypertensive Cohorts Overweight or obesity also demonstrated association with the S363 variant in the diabetes and hypertensive patients (Table 3). Furthermore, we observed significant tracking of BMI with genotype in the hypertensives (N/N ⫽ 26 ⫾ 4 kg/m2, N/S ⫽ 29 ⫾ 2 kg/m2, S/S ⫽ 38 ⫾ 11 kg/m2; p ⬍ 0.001). After correction for multiple comparisons, no difference was seen for BP or lipid variables across genotypes in these groups.

Discussion We have found that frequency of the S363 allele of GR is markedly elevated in obesity. This finding adds to previous data for overweight (19). The data are consistent with obesity being part of a continuum from overweight, having the same genetic cause, at least in the case of the variant we tested. Moreover, although others have noted increased waist-to-hip ratio for the S363 variant in males but not females (10), we saw no sex difference. No association was apparent for the other disorders examined. Interestingly, we noted that S363 carriers in the obese cohort showed significantly higher waist measurements but OBESITY RESEARCH Vol. 11 No. 6 June 2003

805


Table 3. Association of obesity and overweight with NR3C1 N363S variant Total alleles on all chromosomes (frequency)

Genotypes (frequency) Group Control with BMI ⬍ 25 (22 ⫾ 1 kg/m2)

Control with BMI 25 to 29.9 (26 ⫾ 1 kg/m2) Obesity clinic BMI ⱖ 30 (43 ⫾ 8 kg/m2)

Diabetes with BMI ⱖ 30 (35 ⫾ 6 kg/m2) HT with BMI ⬍ 25 (22 ⫾ 2 kg/m2) HT with BMI ⱖ 25 (29 ⫾ 4 kg/m2)

N/N

N/S

141 (0.93) N/N

9 1 (0.06) (0.01) N/S ⫹ S/S

141 94 (0.81) 94 110 (0.72) 110

10

141 (0.84) 141 57 (1.0) 72 (0.86)

17 (0.15)

5 (0.04) 22

32 (0.21)

10 (0.07) 42

23 (0.14)

␹2

S/S

4 (0.02)

10

p

0.0064

9.5 24 (2.7)† 23 (2.7)† 7.0

0.0021 5.9 ⫻ 10⫺6 (0.26)† 1.3 ⫻ 10⫺6 (0.10)† 0.030

6.9

0.0085

27 0 (0.0) 12 (0.14)

8.9

0.0029

␹2

N

S

291 (0.96)

11 (0.04)

205 (0.88)

27 (0.12)

252 (0.83)

52 (0.17)

29 (3.1)†

5.6 ⫻ 10⫺8 (0.077)†

305 (0.91)

31 (0.09)

8.1

0.0045

114 (1.0) 154 (0.92)

0 (0.0) 14 (0.08)

13

10

p

0.0004

0.0016

* All comparisons are with values for “Control with BMI ⬍ 25 kg/m2” group, except where indicated †, in which the comparison was “Control with BMI 25–29.9 kg/m2” vs. “Obesity Clinic BMI ⱖ 30 kg/m2” group. Low n values for certain HT and diabetes subgroups meant the BMI categories shown were the most appropriate. In column 1, values in parentheses are mean ⫾ SD BMI for group.

lower percentages of total body fat, which suggests a difference in body fat distribution in S363 carriers to one of greater abdominal obesity. This is in accordance with a study that found an association between abdominal visceral fat and a BclI variant in GR (16). The biological mechanism leading to the association noted was not immediately apparent from our data. Although no association was seen with lipid variables in the obese group, this could have been obscured by lipid-lowering medication in these patients, the types of which did not differ between genotypes. We have, however, observed an association of S363 with total cholesterol, triglycerides and total cholesterol/high-density lipoprotein-cholesterol ratio in coronary artery disease subjects (30). In the case of type 2 diabetes, our inability to observe an association with this condition does not support a primary effect of the S363 variant on insulin resistance. Glucocorticoids impair insulin-mediated glucose metabolism and induce insulin resistance. However, in subjects with the S363 allele, dexamethasone induces a larger increase in serum 806


insulin (7), consistent with increased sensitivity to glucocorticoids. Insulin, by increasing another factor that affects GR phosphorylation state, can boost the transactivation potency of the GR on a glucocorticoid-responsive element-containing promoter in dexamethasone-treated cells (31,32) and cause concomitant modification of the GR at the posttranslational level (32). The response to insulin in S363 carriers is not known, although stimulation of promoter activity by each GR allele using dexamethasone alone is similar (7). Abnormal GR activity had earlier been suggested to be the cause of an increase in the effect of glucocorticoids leading to mild HT, insulin resistance, and hyperglycemia in men, particularly in young men with higher BP (33). The patients exhibited increased tissue sensitivity to cortisol, as well as enhanced ligand-binding affinities for dexamethasone in leukocytes. However, we were unable to detect an association of hypertension with the N363S variant and have previously obtained negative data in association studies involving three NR3C1 markers as well as in


sib-pair linkage analysis (22). Others also failed to see linkage of the NR3C1 locus to HT (34). In addition, others also found no association of BP with the N363S allele (7). In conclusion, the present study has demonstrated that obesity and overweight are associated with the S363 allele of the N363S variant. Early testing for the S363 allele might have merit in overall evaluation of a person’s risk of overweight/obesity later in life.

Acknowledgments This study was supported by the National Health and Medical Research Council of Australia. We thank Stephen Colagiuri, Judith O’Neill, and Alison K. Gosby for help in collection of patient samples and Adam V. Benjafield for assistance in some of the statistical analyses.

11.

12.

13.

14.

15. References 1. Connell JM, Kenyon CJ, Ingram M, et al. Corticosteroids in essential hypertension: multiple candidate loci and phenotypic variation. Clin Exp Pharmacol Physiol. 1996;23:369 –74. 2. Bjo¨rntorp P, Rosmond R. Obesity and cortisol. Nutrition. 2000;16:924 –36. 3. Hurley DM, Accili D, Stratakis CA, et al. Point mutation causing a single amino acid substitution in the hormone binding domain of the glucocorticoid receptor in familial glucocorticoid resistance. J Clin Invest. 1991;87:680 – 6. 4. Weaver JU, Hitman GA, Kopelman PG. An association between a BclI restriction fragment length polymorphism of the glucocorticoid receptor locus and hyperinsulinaemia in obese women. J Mol Endocrinol. 1992;9:295–300. 5. Karl M, Lamberts SW, Detera-Wadleigh SD, et al. Familial glucocorticoid resistance caused by a splice site deletion in the human glucocorticoid receptor gene. J Clin Endocrin Metab. 1993;76:683–9. 6. Dahia PLM, Honegger J, Reincke M, et al. Expression of glucocorticoid receptor gene isoforms in corticotropin-secreting tumors. J Clin Endocrinol Metab. 1997;82:1088 –93. 7. Huizenga NATM, Koper JW, deLange P, et al. A polymorphism in the glucocorticoid receptor gene may be associated with increased sensitivity to glucocorticoids in vivo. J Clin Endocrinol Metab. 1998;83:144 –51. 8. de Lange P, Koper JW, Brinkmann AO, de Jong FH, Lamberts SW. Natural variants of the beta isoform of the human glucocorticoid receptor do not alter sensitivity to glucocorticoids. Mol Cell Endocrinol. 1999;153:163– 8. 9. Lind U, Greenidge P, Gillner M, Koehler KF, Wright A, Carlstedt-Duke J. Functional probing of the human glucocorticoid receptor steroid-interacting surface by site-directed mutagenesis. Gln-642 plays an important role in steroid recognition and binding. J Biol Chem. 2000;275:19041–9. 10. Dobson MG, Redfern CP, Unwin N, Weaver JU. The N363S polymorphism of the glucocorticoid receptor: po-

16.

17.

18.

19.

20.

21. 22.

23.

24.

25.

tential contribution to central obesity in men and lack of association with other risk factors for coronary heart disease and diabetes mellitus. J Clin Endocrinol Metab. 2001; 86:2270 – 4. Rosmond R, Bouchard C, Bjo¨rntorp P. Tsp509I polymorphism in exon 2 of the glucocorticoid receptor gene in relation to obesity and cortisol secretion: cohort study. BMJ. 2001; 322:652–3. Ruiz M, Lind U, Gafvels M, et al. Characterization of two novel mutations in the glucocorticoid receptor gene in patients with primary cortisol resistance. Clin Endocrinol. 2001;55: 363–71. Malchoff DM, Brufsky A, Reardon G, et al. A mutation of the glucocorticoid receptor in primary cortisol resistance. J Clin Invest. 1993;91:1918 –25. Lambillotte C, Gilon P, Henquin JC. Direct glucocorticoid inhibition of insulin secretion. An in vitro study of dexamethasone effects in mouse islets. J Clin Invest. 1997; 99:414 –23. Watt GCM, Harrap SB, Foy CJW, et al. Abnormalities of glucocorticoid metabolism and renin-angiotensin system: a four corners approach to the identification of genetic determinants of blood pressure. J Hypertens. 1992;10:473– 82. Buemann B, Vohl M-C, Chagnon M, et al. Abdominal visceral fat is associated with a BclI restriction fragment length polymorphism at the glucocorticoid receptor gene locus. Obes Res. 1997;71:186 –92. Rosmond R, Chagnon YC, Holm G, et al. A glucocorticoid receptor gene marker is associated with abdominal obesity, leptin, and dysregulation of the hypothalamic-pituitary-adrenal axis. Obes Res. 2000;8:211– 8. Koper JW, Stolk RP, de Lange P, et al. Lack of association between five polymorphisms in the human glucocorticoid receptor gene and glucocorticoid resistance. Hum Genet. 1997;99:663– 8. Lin RCY, Wang WYS, Morris BJ. High penetrance, overweight, and glucocorticoid receptor variant: case-control study. BMJ. 1999;319:1337– 8. Echwald SM, Sorensen TI, Andersen T, Pedersen O. The Asn363Ser variant of the glucocorticoid receptor gene is not associated with obesity or weight gain in Danish men. Int J Obes Relat Metab Disord. 2001;25:1563–5. Cardon LR, Palmer LJ. Population stratification and spurious allelic association. Lancet. 2003;361:598 – 604. Lin RCY, Wang WY, Morris BJ. Association and linkage analyses of glucocorticoid receptor gene markers in essential hypertension. Hypertension. 1999;34:1186 –92. Hoeck W, Groner B. Hormone-dependent phosphorylation of the glucocorticoid receptor occurs mainly in the aminoterminal transactivation domain. J Biol Chem. 1990;265: 5403– 8. Bodwell JE, Orti E, Coull JM, Pappin DJ, Smith LI, Swift F. Identification of phosphorylated sites in the mouse glucocorticoid receptor. J Biol Chem. 1991;266:7549 –55. Hollenberg SM, Evans RM. Multiple and cooperative transactivation domains of the human glucocorticoid receptor. Cell. 1988;55:899 –906. OBESITY RESEARCH Vol. 11 No. 6 June 2003

807


26. Govindan MV, Warriar N. Reconstitution of the N-terminal transcription activation function of human mineralocorticoid receptor in a defective human glucocorticoid receptor. J Biol Chem. 1998;273:24439 – 47. 27. Willett WC, Dietz WH, Colditz GA. Guidelines for healthy weight. N Engl J Med. 1999;341:427–34. 28. Morris BJ, Zee RYL, Schrader AP. Different frequencies of angiotensin-converting enzyme genotypes in older hypertensive individuals. J Clin Invest. 1994;94:1085–9. 29. Ott J. Analysis of Human Genetic Linkage. Baltimore, MD: The Johns Hopkins University Press; 1991. 30. Lin RCY, Wang XL, Morris BJ. Association of coronary artery disease with glucocorticoid receptor N363S variant. Hypertension. 2003;41:404 –7.

808


31. Tsawdaroglou NH, Georgakopoulos A, Nikodimou E. Insulin enhances glucocorticoid induction of gene expression in a sequence specific manner. J Steroid Biochem. 1989;34: 333–7. 32. Georgakopoulos A, Tsawdaroglou N. Insulin potentiates the transactivation potency of the glucocorticoid receptor. FEBS Lett. 1996;381:177– 82. 33. Walker BR, Phillips DI, Noon JP, et al. Increased glucocorticoid activity in men with cardiovascular risk factors. Hypertension. 1998;31:891–5. 34. Rice T, Rankinen T, Province MA, et al. Genome-wide linkage analysis of systolic and diastolic blood pressure: the Quebec Family Study. Circulation. 2000;102:1956 – 63.