Association Study of AMP-Activated Protein Kinase Subunit Genes in Polycystic Ovary Syndrome
Total Page:16
File Type:pdf, Size:1020Kb
European Journal of Endocrinology (2009) 161 405–409 ISSN 0804-4643 CLINICAL STUDY Association study of AMP-activated protein kinase subunit genes in polycystic ovary syndrome Kari Sproul1,2, Michelle R Jones3, Ricardo Azziz1,2,4 and Mark O Goodarzi1,3,4,5 1Department of Obstetrics and Gynecology, Cedars-Sinai Medical Center, Los Angeles, California 90048, USA, 2Department of Obstetrics and Gynecology, the David Geffen School of Medicine at UCLA, Los Angeles, California 90095, USA, 3Division of Endocrinology, Diabetes and Metabolism, Department of Medicine, Cedars-Sinai Medical Center, 8700 Beverly Boulevard, Room B-131, Los Angeles, California 90048, USA, 4Department of Medicine, the David Geffen School of Medicine at UCLA, Los Angeles, California 90095, USA and 5Medical Genetics Institute, Cedars-Sinai Medical Center, Los Angeles, California 90048, USA (Correspondence should be addressed to M O Goodarzi at Division of Endocrinology, Diabetes and Metabolism, Department of Medicine, Cedars-Sinai Medical Center; Email: [email protected]) Abstract Objective: To examine the genes for AMP-activated protein kinase (AMPK) subunits a2(PRKAA2) and g3(PRKAG3) as candidates for polycystic ovary syndrome (PCOS) and its component traits. Design and methods: A total of 287 white PCOS women were recruited from the reproductive endocrinology clinic at the University of Alabama at Birmingham and 187 white control subjects were recruited from the surrounding community. Seven PRKAA2 single nucleotide polymorphisms (SNPs) and four PRKAG3 SNPs were genotyped in PCOS cases and controls. Genotyping and association analysis were performed at Cedars-Sinai Medical Center. Results: Nominal associations of PRKAA2 variants with insulin-related traits and the PRKAG3 Pro71Ala variant with PCOS were not statistically significant after multiple testing correction. Among PCOS patients, there were no associations between variants in AMPK subunit genes and androgenic or reproductive traits. Conclusions: Variants in genes for AMPKa2 and AMPKg3 were not associated with PCOS or its component traits. Our evidence does not demonstrate that AMPK is a major genetic risk factor for PCOS. European Journal of Endocrinology 161 405–409 Introduction PCOS ovarian theca cells, suggesting that metformin stimulation of AMPK in the ovary itself may promote Polycystic ovary syndrome (PCOS) is a common complex ovulation (7). We therefore examined the genes for genetic disorder, affecting w6–8% of reproductive age AMPKa2(PRKAA2) and AMPKg3(PRKAG3)as women (1, 2). PCOS is characterized by hyperandro- candidate genes for PCOS and its component traits. genism, ovulatory dysfunction, polycystic ovarian morphology, and frequently insulin resistance (IR). AMP-activated protein kinase (AMPK) is a cellular energy sensor, acting to maintain ATP levels in the cell Subjects and methods by sensing the AMP/ATP ratio and then phosphorylat- ing metabolic enzymes and regulating gene expression Subjects and phenotyping with the net result of activating processes that generate A total of 287 consecutive unrelated white patients with ATP and inhibiting processes that consume ATP. AMPK PCOS, aged 13–47 years, and 187 healthy unrelated influences food intake, fatty acid and glucose uptake in white control women, aged 14–60, were recruited from skeletal and cardiac muscle, insulin secretion, and the Birmingham, AL, USA area. PCOS cases were hepatic fatty acid synthesis and gluconeogenesis (3). recruited from the reproductive endocrine practice of Consequently, AMPK subunit genes have been one of the investigators (RA) at the University of examined as candidates for type 2 diabetes suscep- Alabama at Birmingham (UAB). All subjects were tibility, with conflicting results (4, 5). non-Hispanic Caucasians. Participation in research As PCOS is a metabolic disorder associated with an studies was offered to patients meeting inclusion increased risk of diabetes, AMPK may play a role in this criteria (premenopausal, not pregnant, on no hormonal association. Metformin, commonly prescribed for PCOS therapy for at least 3 months, and meeting diagnostic patients, activates AMPK activity in skeletal muscle (6). criteria for PCOS). PCOS was diagnosed following the In addition, AMPK activation appears to be decreased in 1990 NIH consensus criteria (8). q 2009 European Society of Endocrinology DOI: 10.1530/EJE-09-0245 Online version via www.eje-online.org Downloaded from Bioscientifica.com at 09/27/2021 12:40:30AM via free access 406 K Sproul and others EUROPEAN JOURNAL OF ENDOCRINOLOGY (2009) 161 Controls were healthy women with a history of Statistical analysis regular menstrual cycles and no family history of 2 hirsutism. These women had no evidence of hirsutism, Unpaired t-tests and c tests were used to compare acne or alopecia, or endocrine dysfunction and had not clinical characteristics between women with and taken hormonal therapy (including oral contraceptives) without PCOS; quantitative trait values were log- or for at least 3 months prior to testing. Controls were square root-transformed as appropriate to reduce non- recruited by word of mouth and advertisements in normality. the Birmingham, Alabama area, through a call for Association with the presence/absence of PCOS and ‘healthy women’ without detailing further the nature qualitative traits was evaluated using logistic of the studies. regression, adjusting for age and body mass index Subjects underwent a physical examination and (BMI) by including both as independent variables. serum androgen measurement, as per a previously Association with quantitative traits was evaluated described protocol (1). Fasting glucose and insulin were using analysis of covariance, again adjusting for age also obtained in a subset of the cohort (w70%). This and BMI. Recognizing that insulin secretion responds to subset did not differ demographically or hormonally ambient IR, HOMA-%B associations were also calcu- from the study subjects overall. The computer-based lated with HOMA-IR as a covariate in addition to age homeostasis model assessment (HOMA, www.dtu.ox.ac. and BMI. Quantitative trait associations were conducted in the PCOS and control groups separately. To handle uk/homa) utilizes fasting glucose and insulin to ! calculate indices of IR (HOMA-IR) and insulin secretion multiple testing, significance was taken as P 0.006, (HOMA-%B) (9). For the insulin-related traits only, six considering that we analyzed two genes against four PCOS subjects with diabetes were excluded because the families of traits (PCOS diagnosis, androgens, metabolic hyperglycemia of diabetes may induce secondary traits and reproductive endpoints), yielding a Bonfer- changes in insulin-related traits that reduce their utility roni correction factor of 8 (i.e. eight independent for genetic analysis. In addition, in PCOS cases only, we comparisons). Association analyses excluding 26 evaluated the following reproductive traits: severe post-menopausal subjects from the control group oligomenorrhea (defined as O3monthsbetween yielded the same results as analyses conducted in the menses), infertility (yes/no), and parity (yes/no). entire cohort; therefore, the latter results are reported. Clinical characteristics of the subjects are given in supplementary Table 1, which can be viewed online at http://www.eje-online.org/supplemental/. All subjects gave written informed consent, and the Results study was performed according to the guidelines of the Institutional Review Boards of UAB and Cedars- PRKAA2 Sinai Medical Center, Los Angeles, CA, USA. We genotyped seven SNPs spanning the PRKAA2 gene (Fig. 1A). SNP frequencies are shown in Table 1. All Genetic analysis markers were in Hardy–Weinberg equilibrium. Linkage disequilibrium among these markers (D0) ranged from We selected six PRKAA2 single nucleotide polymorph- 0.45 to 1.0 (average pairwise D0 of 0.90). The r2 values isms (SNPs; rs11206887, rs2143749, rs2746349, ranged from 0 to 0.98 (average 0.23). rs12749128, rs857155, and rs3738568) using geno- Among the women with PCOS, carriers of the minor type data of the Caucasian population of the HapMap alleles of SNPs rs2143749, rs2746349, or rs857155 database (release 24, phase II) (10). These SNPs capture had higher fasting insulin levels than noncarriers 28 out of 35 (80%) of the Caucasian HapMap SNP (PZ0.044, 0.021, and 0.027 respectively). Carriers of alleles at r2O0.8. An additional PRKAA2 SNP the minor allele of SNP rs2143749 also had increased (rs2051040) was selected because it was associated HOMA-%B (PZ0.030), which was not significant when with type 2 diabetes in a Japanese cohort (4). Four HOMA-IR was also included as a covariate (PZ0.26). validated PRKAG3 SNPs (rs6436094, rs16859382, Lastly, PCOS women who were carriers of the minor rs692243, and rs650898) were genotyped, selected allele of SNP rs2746349 had increased HOMA-IR based on frequency O5% because only two SNPs compared with noncarriers (PZ0.039). Trait values in HapMap (rs6436094, rs692243) are polymorphic for PCOS women are shown in the electronic supple- in Caucasians. SNP rs692243 is a nonsynonymous mental Table 3, which can be viewed online at http:// variant (proline to alanine at codon 71). The eleven www.eje-online.org/supplemental/.Noneofthese SNPs were genotyped in the 474 subjects using PRKAA2 associations were significant after considering the 50-exonuclease assay (TaqMan MGB, Applied multiple testing. We did not find significant associations Biosystems, Foster City, CA, USA). Supplementary between PRKAA2 variantsandPCOSitself,BMI,