Although diabetes-associated SNPs in SHBG are not associated with PCOS status, rs1799941 and rs727428 genotypes are independently associated with SHBG levels in PCOS women.
Abstract
Context:
Single-nucleotide polymorphisms (SNPs) in the SHBG gene are associated with type 2 diabetes mellitus. SHBG has also been proposed as a candidate gene for the polycystic ovary syndrome (PCOS).
Objective:
The study aims were 1) to determine whether any of four SHBG SNPs (rs1779941, rs6297, rs6259, and rs727428) are associated with PCOS and 2) to determine whether SNP genotype influences SHBG levels in PCOS women.
Design:
Using the transmission disequilibrium test, evidence of associations between SHBG SNPs and PCOS were analyzed. Additionally, correlations between SHBG levels and SNP genotype, body mass index, non-SHBG-bound testosterone, and insulin resistance estimated by the homeostasis model were determined.
Setting:
The study was conducted at academic medical centers.
Patients or Other Participants:
A total of 430 families having a proband with PCOS were included in the family-based study. Associations between SNP genotypes, SHBG, and metabolic parameters were determined in 758 women with PCOS including probands from the family cohort.
Main Outcome Measures:
Primary outcome measures included transmission frequency of SNP alleles and correlation coefficients between SHBG and allele frequency/metabolic parameters.
Results:
No evidence of association between SNPs of interest and PCOS was found. However, in multivariate analyses, SHBG levels varied significantly with rs1799941 and rs727428 genotype after controlling for body mass index, non-SHBG-bound testosterone, and homeostasis model for insulin resistance.
Conclusions:
Although SHBG SNPs associated with type 2 diabetes mellitus do not appear to be associated with PCOS status, rs1799941 and rs727428 genotypes are associated with SHBG levels independent of the effects of insulin resistance and obesity.
Polycystic ovary syndrome (PCOS) is characterized by anovulation and hyperandrogenism. The syndrome, which affects an estimated 4–8% of reproductive-age women (1, 2), is characterized by insulin resistance and hyperinsulinemia (3, 4) and is associated with an increased risk of glucose intolerance and type 2 diabetes mellitus (T2DM) (5). Alterations in several different metabolic pathways have been implicated in the pathophysiology of PCOS, including abnormalities in steroid hormone regulation (6–8) and insulin signaling pathways (9, 10). Previous observations support a familial aggregation of PCOS and, thus, suggest a genetic basis for the syndrome (11, 12). Although the specific genetic alterations that contribute to the development of PCOS remain unclear, several candidate genes have been proposed including the SHBG gene (13, 14).
The SHBG gene, located on chromosome 17 (17p13-1p12), encodes a 373-amino-acid polypeptide that regulates the bioavailability of sex steroids by binding androgens, particularly testosterone, and estrogens (15, 16). The majority of SHBG is produced by the liver, and hepatic production of SHBG is regulated by several metabolic factors including insulin and androgens (17–19). Levels of SHBG are decreased in women with PCOS, which has been primarily ascribed to an inhibitory effect of hyperinsulinemia on SHBG production (18). In addition to metabolic and nutritional factors, circulating levels of SHBG are partially determined by genetic variation (20, 21). Alterations in SHBG may contribute to the observed metabolic and, consequently, phenotypic abnormalities in affected women. Given the high binding affinity of SHBG for serum androgens, decreases in serum SHBG among women with PCOS may lead to increases in free androgen levels, magnifying the biological impact of androgens.
In addition to metabolic regulation of SHBG by insulin, studies suggest that genetic variation in SHBG, and subsequently changes in sex steroid bioavailability, may contribute to broader physiological changes including glucose homeostasis and insulin resistance (22–26). Two recent studies have demonstrated that single-nucleotide polymorphisms (SNPs) in the SHBG gene are associated with alterations in circulating SHBG levels and predict the development of T2DM in men and women (23, 25). Specifically, Ding and colleagues (25) investigated the relationships between two SHBG SNPs, rs6257 and rs6259, and T2DM. In their study, carriers of the rs6257 minor allele (C) demonstrated significantly lower levels of SHBG and increased risk of T2DM. Conversely, carriers of a minor allele (A) for rs6259 demonstrated increased serum levels of SHBG and a lower risk of T2DM. Further supporting the biological association between alterations in SHBG and glucose homeostasis, Perry and colleagues (23) reported that carriers of the minor allele (A) of a distinct SNP, rs1799941, had increased levels of SHBG that were associated with reduced risk of T2DM.
In light of the conclusions of these studies, as well as others describing alterations in serum SHBG levels in women affected by PCOS compared with normal women (27), the central role of insulin resistance in the disorder (3, 4), and significant associations between PCOS and glucose intolerance and T2DM (3, 5), we hypothesized that the transmission of polymorphisms in the SHBG gene (rs1799941, rs6257, rs6259, and rs727428) previously associated with risk of T2DM may also be independently associated with the development of PCOS. Specifically, we tested for associations between alleles for each of the SHBG SNPs of interest and PCOS using the transmission disequilibrium test (TDT) (28). Furthermore, we examined the relationships between each SNP genotype and circulating levels of SHBG in women with PCOS.
Subjects and Methods
PCOS subjects and phenotyping
A total of 477 offspring from 387 simplex and 43 multiplex families having at least one daughter with PCOS (group 1) were available for participation in the family-based association analysis of SHBG SNPs (rs179994, rs6257, rs6259, and rs7272428) and PCOS. Data regarding the transmission of other genes of interest in PCOS have been published previously based on this cohort of families (29). In addition to the family-based cohort, a second cohort of 346 unrelated women diagnosed with PCOS (group 2) was studied. The self-identified ethnicities of probands in the families were 87% White, 4% Hispanic, 1% Black, and 7% other or unknown.
Diagnostic criteria for PCOS have been described in detail elsewhere (6, 12) and were identical for both groups of women. Women were considered affected with PCOS if they had elevated total testosterone (T) (greater than 58 ng/dl) or non-SHBG-bound testosterone (uT) (greater that 15 ng/dl) and six or fewer menses per year. Women with clinical or biochemical evidence of nonclassical 21-hydroxylase deficiency, hyperprolactinemia, or androgen-secreting tumors were excluded (30). Women included in the study were either evaluated on site at one of three study centers (n = 479) or off site at a local hospital or clinical laboratory (n = 279). For subjects studied on site, height and weight were measured to the nearest 0.1 cm and 0.1 kg, respectively. In subjects studied off site, height and weight were self-reported. In both groups, body mass index (BMI) (kg/m2) was calculated. A validation study of self-reported height and weight in a subset of women studied previously concluded that self-reported values and on-site measurements were similar and did not result in a significant differences in BMI values (31).
In all subjects, fasting venous blood samples were collected between 0800 and 1000 h for measurement of plasma glucose, insulin, lipids, SHBG, T, and uT. Plasma glucose levels were determined using glucose oxidase methodology. Insulin, T, and uT levels were determined as previously described (12). Insulin resistance was estimated by the homeostasis model of insulin resistance: HOMA-IR = [fasting glucose (milligrams per deciliter) × fasting insulin (microunits per milliliter)]/405 (32). SHBG was determined using a two-site immunoradiometric assay (Diagnostic Systems Laboratories, Inc., Webster, TX). All assays had interassay coefficients of variation below 10%.
This study was approved by the institutional review boards of the University of Pennsylvania, The Pennsylvania State University College of Medicine, Brigham and Women's Hospital, and Northwestern University. Written informed consent was obtained from all adult subjects and from a parent or guardian for minor subjects.
Fasting glucose status
During the baseline medical evaluation, a self-reported a history of diabetes mellitus (DM) or gestational DM (GDM) was collected. For this study, glucose tolerance status was defined according to a combination of self-reported history of DM and fasting plasma glucose levels. Women with a self-reported history of GDM were excluded from all analyses. In women without a self-reported history of DM or GDM, glucose tolerance status was further defined based on fasting plasma glucose values (33). Specifically, self-reported normal women were classified as having normal fasting glucose (NFG), impaired fasting glucose (IFG), or DM based on fasting plasma glucose values of lower than 100 mg/dl, 100–125 mg/dl, and higher than 125 mg/dl, respectively. Women reporting a history of DM were considered to have DM regardless of fasting glucose values.
SNP genotyping
We genotyped four SNPs in the SHBG gene (rs1799941, rs6257, rs6259, and rs727428; Table 1) using an Applied Biosystems predesigned TaqMan SNP Genotyping Assay (C_8727483_10, C_11955742_10, C_11955739_10 and C_3290005_10, respectively). Allelic PCR products were analyzed using the Applied Biosystems 7900HT Sequence Detection System and SDS version 2.2 software. Genotypes were auto-called with the quality value set at 0.95. Three Centre d'Etude du Polymorphisme Humain individuals were genotyped on each 96-well plate, and no discrepancies were observed; genotypes were all in Hardy-Weinberg equilibrium. Error checking of genotypes in the family material was performed with Merlin software (34), and families with discrepancies were excluded from the analysis.
Table 1.
Characteristics of SHBG SNPs of interest
| Marker ID | Position (chromosome 17) | Location in SHBG | Alleles | MAF |
|---|---|---|---|---|
| rs1799941 | 7474148 | 86 bp 5′ | G:A | A:0.264 |
| rs6297 | 7474442 | Intron | T:C | C:0.096 |
| rs6259 | 7477252 | Exon | G:A | A:0.103 |
| rs727428 | 7478517 | 1.1 kb 3′ | C:T | T:0.422 |
Minor alleles are in bold. MAF, minor allele frequency.
Statistical analysis
Anthropometric and biochemical variables that were not normally distributed were log-transformed for all statistical analyses and reported back-transformed in their original units. All results are reported as means or geometric means for transformed variables with 95% confidence intervals. To account for multiple comparisons involving the four SNPs, Bonferroni correction procedure was applied, and P values <0.01 were considered significant. With the exception of the TDT analysis, all statistical analyses were performed using JMP (version 8; SAS Institute Inc., Cary, NC).
Differences in anthropometric and metabolic characteristics according to glucose tolerance status were assessed in both groups of PCOS women using ANOVA. When ANOVA testing confirmed significant group differences, Tukey's honestly significant different testing was performed to compare mean values between specific groups. Linkage and association between SHBG SNPs and PCOS was analyzed in women (n = 477) in the family-based study (group 1) using the TDT (28). Specifically, TDT analyses were performed for each of the four SNPs separately in families with NFG offspring (n = 400) and those with IFG or DM offspring (n = 77). For the remaining analyses, 412 unrelated probands from group 1 and an additional collection of 346 unrelated women with PCOS from group 2 were combined in a single study population (n = 758). Using the combined population, χ2 analysis was performed to assess differences in SNP allele frequency between PCOS women with NFG (n = 634) and women with IFG/DM (n = 114). ANOVA testing was performed to assess differences in serum SHBG according to the number of variant SNP alleles in the subset of 558 women with PCOS and NFG for whom SHBG levels were known. Univariate correlation analyses were performed in the subset of PCOS-NFG women using Pearson's correlation tests. Lastly, multiple linear regression analyses were performed with SHBG level as the primary dependent variable of interest considering SNP copy number, BMI, uT, and HOMA-IR as independent variables.
Results
Metabolic characteristics of women with PCOS in the two cohorts
The metabolic characteristics of women with PCOS in each of the two groups (probands in family-based study and unrelated women) according to fasting glucose status are summarized in Table 2. PCOS women with IFG or DM had higher BMI values compared with PCOS-NGT women (P < 0.001). As expected, women with IFG/DM also had significantly higher fasting glucose, fasting insulin, and HOMA-IR values than women with NGT (P < 0.001 for each variable). However, mean T, uT, and SHBG values did not differ significantly among women with PCOS according to fasting glucose status categories.
Table 2.
Clinical and metabolic characteristics of women with PCOS according to study cohort and glucose tolerance status
| Characteristic | Proband women with PCOS (group 1) |
Unrelated women with PCOS (group 2) |
||||
|---|---|---|---|---|---|---|
| NFG | IFG | DM | NFG | IFG | DM | |
| n | 342 | 45 | 25 | 292 | 39 | 15 |
| Age (yr) | 27.1 (26.5–27.7) | 29.4 (27.4–31.3)a | 32.4 (29.7–35.1)a | 28.6 (27.9–29.3) | 29.4 (27.8–31.0) | 32.2 (29.6–34.8)c |
| BMI (kg/m2) | 34.6 (33.7–35.5) | 39.1 (36.0–42.1)a | 40.2 (37.0–43.4)a | 34.5 (33.5–35.5) | 38.6 (35.6–41.7)c | 40.0 (35.2–44.9)c |
| T (ng/dl)e | 73.1 (70.3–76.0) | 80.0 (71.9–88.9) | 72.7 (63.0–83.8) | 76.9 (73.9–80.0) | 70.6 (63.7–78.3) | 83.9 (71.0–99.2) |
| uT (ng/dl)e | 23.3 (22.1–24.5) | 28.6 (24.7–33.0) | 24.8 (20.5–30.1) | 23.7 (22.4–25.0) | 23.2 (20.1–26.8) | 27.3 (21.5–34.6) |
| SHBG (nmol/liter)e | 56.0 (52.4–59.8) | 45.6 (38.1–54.6) | 48.3 (38.1–61.2) | 58.0 (54.0–62.3) | 47.6 (39.7–57.2) | 46.4 (34.6–62.3) |
| Fasting glucose (mg/dl) | 86 (85.4–86.9) | 107 (105.3–108.6)a | 141 (124.1–158.6)a,b | 87 (85.6–87.4) | 106 (104.5–107.9)c | 126 (105.8–146.4)c,d |
| Fasting insulin (μU/ml)e | 21.0 (19.7–22.3) | 34.2 (29.0–40.3)a | 33.4 (25.1–39.3)a | 22.2 (20.7–23.8) | 27.6 (23.1–33.1) | 39.7 (29.1–54.3)c |
| HOMA-IRe | 4.4 (4.16–4.73) | 9.0 (7.56–10.72)a | 10.6 (8.33–13.38)a | 4.7 (4.38–5.09) | 7.2 (5.98–8.77)c | 12.0 (8.61–16.75)c,d |
Women with PCOS, including the proband cohort (group 1), were recruited as part of a family-based association study investigating genetic causes of PCOS; a second cohort of unrelated women diagnosed with PCOS (group 2) using the same criteria as probands in the family cohort was also included in this study. Data are presented as means (95% confidence intervals) unless otherwise noted. Glucose tolerance status was defined according to self-reported history of diabetes and fasting plasma glucose values: DM (fasting plasma glucose >125 mg/dl and/or history of DM); IFG (fasting plasma glucose 100–125 mg/dl without a history of DM); NFG (fasting plasma glucose <100 mg/dl without a history of DM). To convert values for T and uT to millimoles per liter, multiply by 0.03467; to convert values for glucose to millimoles per liter, multiply by 0.05551; to convert values for insulin to picomoles per liter, multiple by 7.175. Differences in mean values according to glucose tolerance groups were compared using ANOVA. When ANOVA testing was significant (P < 0.001), specific group means were compared using Tukey's honestly significant different testing.
P < 0.05 compared with proband NFG group.
P < 0.05 compared with proband IFG group.
P < 0.05 compared with unrelated NFG group.
P < 0.05 compared with unrelated IFG group.
Geometric means.
TDT analyses for association between SHBG SNPs and PCOS
The results of the TDT analyses for family-based associations between four SNPs in SHBG (rs 1799941, rs6257, rs6259, and rs727428) and PCOS among offspring with NFG (n = 400) and those with IFG or DM (n = 77) are shown in Table 3. The primary outcome of interest was a significant association between PCOS and alleles of each of the four SHBG SNPs among women with NFG. In these analyses, no significant overtransmission of alleles at any of the four SNPs was observed, indicating that none of these SNPs are associated with PCOS. A separate TDT analysis in PCOS offspring with IFG/DM also failed to find any significant associations.
Table 3.
TDT analyses for the association between four SNPs in SHBG in affected offspring with PCOS according to glucose tolerance status
| SNP | Transmitted allele | NFG (n = 400) |
IFG/DM (n = 77) |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Ta | Not Ta | Total TDT | Transmission frequency | TDT χ2 | P | Ta | Not Ta | Total TDT | Transmission frequency | TDT χ2 | P | ||
| rs1799941 | A | 147 | 127 | 274 | 0.536 | 1.46 | >0.05 | 16 | 28 | 44 | 0.364 | 3.27 | >0.05 |
| rs6257 | C | 72 | 55 | 127 | 0.567 | 2.28 | >0.05 | 13 | 13 | 26 | 0.500 | 0.00 | >0.05 |
| rs6259 | G | 76 | 55 | 131 | 0.580 | 3.37 | >0.05 | 14 | 16 | 30 | 0.467 | 0.133 | >0.05 |
| rs727428 | C | 178 | 163 | 341 | 0.522 | 0.66 | >0.05 | 36 | 33 | 69 | 0.522 | 0.13 | >0.05 |
Glucose tolerance status was defined according to self-reported history of diabetes and fasting plasma glucose values: DM (fasting plasma glucose >125 mg/dl and/or history of DM); IFG (fasting plasma glucose 100–125 mg/dl without a history of DM); NFG (fasting plasma glucose <100 mg/dl without a history of DM).
T, Number of transmissions in TDT analysis.
Differences in SHBG SNP genotypes in PCOS women with NGT compared with IFG/DM
After combining the PCOS probands in the family-based study and the cohort of unrelated women with PCOS (n = 758), genotype frequency for each of the four SNPs according to fasting glucose status was investigated (Table 4). No significant differences in the frequency of SHBG SNP genotypes were observed between PCOS women with NFG and IFG/DM. Given the potential confounding association between IFG or DM and SHBG, only women without a self-reported history of DM and normal fasting glucose levels were considered in the remaining analyses.
Table 4.
Frequency of SHBG SNP genotypes in women with PCOS according to glucose tolerance status
| SNP | Glucose tolerance status | Total (n) | Genotype [n (%)] | Genotype [n (%)] | Genotype [n (%)] | χ2 | P |
|---|---|---|---|---|---|---|---|
| rs1799941 | AA | AG | GG | ||||
| NFG | 600 | 45 (7.5) | 195 (32.5) | 360 (60.0) | |||
| IFG/DM | 117 | 8 (6.8) | 38 (32.5) | 71 (60.7) | 0.07 | 0.97 | |
| rs6257 | CC | CT | TT | ||||
| NFG | 620 | 8 (1.3) | 119 (19.2) | 493 (79.5) | |||
| IFG/DM | 121 | 1 (0.8) | 20 (16.5) | 100 (82.7) | 0.69 | 0.71 | |
| rs6259 | AA | AG | GG | ||||
| NFG | 619 | 6 (1.0) | 111 (17.9) | 502 (81.1) | |||
| IFG/DM | 118 | 2 (1.7) | 21 (17.8) | 95 (80.5) | 0.49 | 0.78 | |
| rs727428 | CC | CT | TT | ||||
| NFG | 613 | 197 (32.1) | 288 (47.0) | 128 (20.9) | |||
| IFG/DM | 118 | 37 (31.3) | 60 (50.9) | 21 (17.8) | 0.79 | 0.67 |
Minor alleles are in bold. Glucose tolerance status was defined according to self-reported history of diabetes and fasting plasma glucose values: DM (fasting plasma glucose >125 mg/dl and/or history of DM); IFG (fasting plasma glucose 100–125 mg/dl without a history of DM); NFG (fasting plasma glucose <100 mg/dl without a history of DM). The complete study population consisted of 758 women with PCOS; however, genotype data for rs1799941, rs6257, rs6259, and rs727428 were unavailable in 5.4, 2.2, 2.8, and 3.6% of women, respectively.
Relationship between SNP allele copy number and serum SHBG levels
Table 5 outlines the differences in mean serum SHBG levels according to genotype for each of the four SNPs among PCOS-NFG women (n = 558). For the SNP rs727428, mean levels of SBHG differed significantly according to SNP genotype (P=0.001). Specifically, levels of SHBG were decreased approximately 13% in CT carriers and 27% in TT homozygotes compared with CC homozygotes. Significant differences in mean SHBG levels were not observed according to rs1799941, rs6257, or rs6259 genotypes in unadjusted analyses.
Table 5.
Mean SHBG values (nanomoles per liter) for women with PCOS and NFG according to SHBG SNP genotype
| rs1799941 | rs6257 | rs6259 | rs727428 | |
|---|---|---|---|---|
| n | 530 | 545 | 544 | 540 |
| Genotype: SHBG | AA: 66.6 (55.9–79.3) | CC: 44.5 (29.0–68.2) | AA: 50.6 (30.4–84.1) | CC: 64.0 (58.8–69.7) |
| AG: 58.7 (53.9–64.0) | CT: 53.9 (48.3–60.2) | AG: 55.2 (49.1–61.9) | CT: 55.8 (52.0–59.9) | |
| GG: 54.4 (51.0–58.0) | TT: 57.2 (54.2–60.4) | GG: 57.3 (54.3–60.5) | TT: 46.8 (44.8–55.4) | |
| P value | 0.06 | 0.35 | 0.75 | 0.001 |
Results are shown as geometric means (95% confidence intervals) and were log-transformed before analysis. Minor alleles are in bold. P values are based on ANOVA testing. In the combined study population, 558 women with PCOS and NFG had data available on SHBG. Of these women, genotype data for rs1799941, rs6257, rs6259, and rs727428 were unavailable in 5.3, 2.3, 2.5, and 3.2% of women, respectively.
Determinants of serum SHBG level in women with PCOS
Despite the significant association between rs727428 SNP genotype and serum SHBG levels in this cohort of women with PCOS, several other metabolic factors may confound the observed variations including obesity, hyperandrogenemia, and insulin resistance. In fact, highly significant inverse linear relationships were observed between SHBG levels and BMI (r = −0.39; P < 0.0001), uT (r = −0.52; P < 0.0001), fasting insulin levels (r = −0.44; P < 0.0001), and HOMA-IR (r = −0.45; P < 0.0001). A strong positive correlation was observed between SHBG and T (r = 0.30; P < 0.0001).
As a result, multilinear regression analysis was performed to determine whether the association between SNP genotype and serum levels of SHBG remained significant after controlling for the influence of BMI, hyperandrogenemia, and insulin resistance. The multivariate models included SNP allele copy number, BMI, uT, and HOMA-IR as independent variables with SHBG levels as the dependent variable; separate multivariate analyses were performed for each of the four SNPs. All models were highly statistically significant [P < 0.0001; adjusted R2 = 0.378 for model with rs1799941 (adjusted R2 reflects the proportion of variability in SHBG that is accounted for by each statistical model); adjusted R2 = 0.368 for model with rs6257; R2 = 0.360 for model with rs6259; adjusted R2 = 0.382 for model with rs727428]. The parameter estimates for the independent variables in each model are outlined in Table 6. In all four models, BMI, uT, and HOMA-IR continued to demonstrate significant negative associations with SHBG levels. In the model including rs727428, SNP allele copy number remained an independent predictor of SHBG. Furthermore, although a significant relationship was not observed in the unadjusted analysis, rs1799941 genotype demonstrated a significant, independent association with SHBG levels after controlling for the influence of BMI, uT, and HOMA-IR. Specifically, levels of SHBG increased significantly in women with increasing copies of the rs1799941 minor (A) allele. However, SNP allele copy number was not a significant predictor of SHBG levels in the models including rs6257 and rs6259.
Table 6.
Mulivariate analyses for association between serum SHBG and SHBG SNP genotype, BMI, uT, and insulin resistance in women with PCOS
| Variable | rs1799941 |
rs6257 |
rs6259 |
rs727428 |
||||
|---|---|---|---|---|---|---|---|---|
| β-Estimate (95% CI) | P value | β-Estimate (95% CI) | P value | β-Estimate (95% CI) | P value | β-Estimate (95% CI) | P value | |
| SNP genotype | −0.119 | <0.001 | 0.044 | 0.33 | 0.023 | 0.64 | −0.098 | <0.001 |
| (−0.182 to −0.057) | (−0.044 to 0.132) | (−0.072 to 0.118) | (−0.154 to −0.042) | |||||
| BMI | −0.012 | <0.0001 | −0.012 | <0.0001 | −0.012 | <0.0001 | −0.012 | <0.0001 |
| (−0.018 to −0.006) | (−0.018 to −0.006) | (−0.018 to −0.006) | (−0.018 to −0.007) | |||||
| uT | −0.517 | <0.0001 | −0.497 | <0.0001 | −0.504 | <0.0001 | −0.486 | <0.0001 |
| (−0.614 to −0.420) | (−0.590 to −0.404) | (−0.598 to −0.411) | (−0.578 to −0.394) | |||||
| HOMA-IR | −0.205 | <0.0001 | −0.219 | <0.0001 | −0.204 | <0.0001 | −0.212 | <0.0001 |
| (−0.291 to −0.120) | (−0.305 to −0.134) | (−0.291 to −0.118) | (−0.297 to −0.127) | |||||
SHBG, uT, and HOMA-IR values were log-transformed before analysis. P values reported are for individual parameter estimates within each model. Each of the four models for SHBG were also highly significant overall (P < 0.0001; adjusted R2 = 0.378 for model with rs1799941; adjusted R2 = 0.368 for model with rs6257; R2 = 0.360 for model with rs6259; adjusted R2 = 0.382 for model with rs727428). CI, Confidence interval.
Discussion
The aim of this study was to determine whether an independent association exists between PCOS and SHBG SNPs previously reported to be associated with T2DM. Using data collected from 430 families with female offspring affected by PCOS, we did not find any evidence of linkage or association between the four SHBG SNPs of interest (rs1799941, rs6257, rs6259, and rs727428) and PCOS. We did observe that serum SHBG levels decreased significantly with increasing copies of the minor (T) allele for rs727429, and this relationship remained significant after controlling for the influence of other potential confounding variables including BMI, unbound testosterone, and insulin resistance. Furthermore, in multivariate analysis, rs1799941 genotype also emerged as an independent predictor of SHBG levels. In contrast to the relationship observed with rs727428, SHBG levels increased with increasing copies of the minor (A) allele for rs1799941. We did not observe statistically significant relationships between SNP genotype and serum SHBG levels for the remaining two SNPs studied (rs6257 and rs6259) in either the unadjusted or multivariate analyses.
Despite evidence suggesting that SBHG may be a candidate gene for PCOS (13, 14), our finding of no association between the four SHBG SNPs studied and PCOS are consistent with reported results from other groups investigating associations between SHBG gene polymorphisms and PCOS (27, 35). Bendlová and colleagues (35) compared the frequency of rs6259 mutations in the SHBG gene (referred to as D327N in their paper) in PCOS and healthy control women and found that genotype distribution was not associated with PCOS status. Similarly, Ferk and colleagues (27) investigated the association between a microsatellite polymorphism located in the SHBG gene promoter region (TAAAA)n and PCOS. Although serum SHBG levels appeared to be strongly influenced by presence of the (TAAAA)n, the polymorphism was not present in significantly higher rates in women with PCOS compared with normal control women.
Three of the four SHBG SNPs evaluated in this study (rs1799941, rs6257, and rs6259) have previously been associated with T2DM (23, 25). Therefore, women with overt T2DM or IFG were excluded from our primary analyses to limit the potential confounding effects of glucose intolerance on the associations between SHBG polymorphisms and PCOS status. However, as part of a secondary analysis that did include PCOS women with IFG and DM, we failed to observe significant differences in SHBG SNP genotype frequency between women with NFG and IFG/DM. In light of the relatively small number of women with IFG/DM (n = 77) genotyped, our study was likely underpowered to detect significant associations between the various SHBG SNP genotypes and DM.
The observed discrepancies between studies examining the associations between SHBG SNP transmission and risk of T2DM and PCOS may be explained by the significant genetic heterogeneity responsible for both disorders. Biyasheva and colleagues (29) recently examined the association between polymorphisms in another T2DM susceptibility locus, transcription factor 7-like 2 (TCF7L2), and PCOS. They confirmed that two SNPS in TCF7L2 previously associated with T2DM were not associated with PCOS. However, two PCOS-specific loci were identified after analyzing the entire genomic region encoding for TCF7L2. Although our study did not find an association between PCOS and specific SHBG SNPs previously associated with T2DM, an exhaustive interrogation of the SHBG gene was not conducted. Thus, the possibility remains that other distinct loci in or near the SHBG gene are associated with PCOS and warrants further investigation.
Among PCOS women in our study, serum SHBG levels were found to correlate inversely with BMI, unbound testosterone, and estimates of insulin resistance. The observed relationship between insulin resistance and serum SHBG is consistent with previous studies demonstrating that insulin suppresses hepatic SHBG production in vitro (19), and plasma levels of SHBG increase after the inhibition of insulin release in vivo (18). SHBG levels have also been shown to correlate with both fat mass (18) and BMI (36). Furthermore, SHBG levels increase after significant weight loss (37). Visceral adiposity, in particular, appears to be related to circulating levels of SHBG (38). Although waist circumference or waist-to-hip ratio may be used as a surrogate for central adiposity, such data were not available for the majority of women in our study and could not be accounted for in our analysis. Moreover, the observed relationship between BMI and SHBG levels must be considered in light of the potential confounding effect of obesity-associated hyperinsulinemia. Interestingly, in this study, we found that BMI remained a significant predictor of SHBG levels after controlling for HOMA-IR in multivariate analysis, suggesting that obesity may contribute to alterations in SHBG independent of its effects on insulin resistance.
As a result of the SHBG-regulating effects of these anthropometric and metabolic factors, we used multivariate analysis to determine the independent effect of SHBG SNP genotype on circulating SHBG levels. After multivariate analysis, we observed an independent association between rs1799941 minor allele (A) copy number and serum SHBG levels among women with PCOS. This observed association between rs1799941 genotype and SHBG levels is consistent with the results of other groups of investigators, including the previously described study by Perry et al. (23). Likewise, Riancho and colleagues (39) also demonstrated an association between rs1799941 genotype and SHBG levels that persisted after adjusting for weight in a population of 753 postmenopausal women. We also observed a significant association between a second SHBG SNP, rs727428, and serum SHBG levels.
However, in contrast to the study conducted by Ding and colleagues (25), we did not find significant associations between SHBG levels and genotypes for rs6257 or rs6259, even after adjusting for the potential influence of other confounding variables. Similar to our findings, Bendlová and colleagues (35) did not demonstrate an association between rs6259 genotype and SHBG levels among women in their study. Riancho et al. (39) also did not observe a significant relationship between rs6259 allele frequency and SHBG levels. However, Riancho and colleagues (39) did report a significant association between rs6257 genotype and levels of circulating SHBG, although the rs6257 SHBG relationship was not as robust as the one observed with rs1799941. The discrepancies between SHBG levels and rs6257 and rs6259 allele frequency between the various studies may be partially explained by differences in the study populations and range of serum SHBG values. In their study, Riancho et al. (39) studied postmenopausal women exclusively. In light of evidence suggesting that sex steroids influence SHBG levels, hormonal alterations during the perimenopausal transition may obscure the associations between SHBG SNP genotype and circulating levels of the protein. The Ding study included an even more heterogeneous population of both men and women (25), and there was greater clinical variability in SHBG levels over the entire study population compared with our study involving only women with PCOS. In fact, in the study by Ding et al. (25), subset analysis involving only female subjects no longer revealed a significant difference in SHBG levels according to rs6259 genotype.
The SHBG gene is complex with multiple first exons and promoters (40). One of the SNPs significantly associated with SHBG levels, rs179994, is located in exon 1L within the 5′ untranslated region of the gene. The observed relationship between rs1799941 genotype and serum SHBG suggests a dominant role for this splice variant, which is abundantly expressed in liver, in the control of circulating SHBG levels. In preliminary studies, we have ruled out a role for rs1799941 in transcriptional regulation of the SHBG gene (Chen, C., J. C. Smothers, and J. F. Strauss, unpublished observations), making it likely that its impact is on mRNA stability or translational control. These mechanisms warrant further investigation. The rs727428 SNP is 1.1 kb 3′ to the gene sequence, and the mechanism of its influence on SHBG levels remains to be determined.
In summary, we did not find evidence of an association between PCOS status and any of four candidate SNPs in the SHBG gene previously associated with T2DM. We did, however, observe that circulating levels of SHBG were independently associated with minor allele frequency for two of the SHBG SNPs, rs1799941 and rs727428, after controlling for the potential impact of BMI, hyperandrogenemia, and insulin resistance on SHBG. Such findings underscore the influence of genetic variation on circulating levels of SHBG among women with PCOS. Additional studies are necessary to determine whether other novel polymorphisms in SHBG contribute to the pathophysiology of PCOS.
Acknowledgments
We thank and acknowledge the women and their families who participated in the study. We also thank the study coordinators (B. Sheetz, S. Ward, and J. Schindler) and the General Clinical Research Center nursing staff at Brigham and Women's Hospital, Northwestern University, and Pennsylvania State University.
This research was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, through cooperative agreement [U54 HD034449 (to J.E.N., J.F.S., and R.S.L.)] as part of the Specialized Cooperative Centers Program in Reproduction and Infertility Research. Additional support was provided through the following National Institutes of Health Grants: K23 HD053742 (to E.P.W.), K24 HD40327 (to J.E.N.), P50 HD044405 (to A.D.), RR10732 and C06 RR016499 [to The Pennsylvania State University General Clinical Research Center (GCRC)], M01 RR00048 (to Northwestern University GCRC), and M01 RR10732 and M01 RR02635 (to Brigham and Women's Hospital GCRC). This project was also funded, in part, under a grant with the Pennsylvania Department of Health using Tobacco Settlement Funds (SAP 41-000-26343 to R.S.L). The Department specifically disclaims responsibility for any analyses, interpretations, or conclusions.
Disclosure Summary: The authors have no potential conflicts of interest to disclose.
Footnotes
- adjusted R2
- Reflects the proportion of variability in SHBG that is accounted for by each statistical model
- BMI
- body mass index
- DM
- diabetes mellitus
- GDM
- gestational DM
- HOMA-IR
- homeostasis model of insulin resistance
- IFG
- impaired fasting glucose
- NFG
- normal fasting glucose
- PCOS
- polycystic ovary syndrome
- SNP
- single-nucleotide polymorphism
- T
- total testosterone
- T2DM
- type 2 diabetes mellitus
- TDT
- transmission disequilibrium test
- uT
- non-SHBG-bound testosterone.
References
- 1. Azziz R, Woods KS, Reyna R, Key TJ, Knochenhauer ES, Yildiz BO. 2004. The prevalence and features of the polycystic ovary syndrome in an unselected population. J Clin Endocrinol Metab 89:2745–2749 [DOI] [PubMed] [Google Scholar]
- 2. Knochenhauer ES, Key TJ, Kahsar-Miller M, Waggoner W, Boots LR, Azziz R. 1998. Prevalence of the polycystic ovary syndrome in unselected Black and White women of the southeastern United States: a prospective study. J Clin Endocrinol Metab 83:3078–3082 [DOI] [PubMed] [Google Scholar]
- 3. DeUgarte CM, Bartolucci AA, Azziz R. 2005. Prevalence of insulin resistance in the polycystic ovary syndrome using the homeostasis model assessment. Fertil Steril 83:1454–1460 [DOI] [PubMed] [Google Scholar]
- 4. Dunaif A, Segal KR, Futterweit W, Dobrjansky A. 1989. Profound peripheral insulin resistance, independent of obesity, in polycystic ovary syndrome. Diabetes 38:1165–1174 [DOI] [PubMed] [Google Scholar]
- 5. Moran LJ, Misso ML, Wild RA, Norman RJ. 2010. Impaired glucose tolerance, type 2 diabetes and metabolic syndrome in polycystic ovary syndrome: a systematic review and meta-analysis. Hum Reprod Update 16:347–363 [DOI] [PubMed] [Google Scholar]
- 6. Urbanek M, Legro RS, Driscoll DA, Azziz R, Ehrmann DA, Norman RJ, Strauss JF, 3rd, Spielman RS, Dunaif A. 1999. Thirty-seven candidate genes for polycystic ovary syndrome: strongest evidence for linkage is with follistatin. Proc Natl Acad Sci USA 96:8573–8578 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Carey AH, Waterworth D, Patel K, White D, Little J, Novelli P, Franks S, Williamson R. 1994. Polycystic ovaries and premature male pattern baldness are associated with one allele of the steroid metabolism gene CYP17. Hum Mol Genet 3:1873–1876 [DOI] [PubMed] [Google Scholar]
- 8. Gharani N, Waterworth DM, Batty S, White D, Gilling-Smith C, Conway GS, McCarthy M, Franks S, Williamson R. 1997. Association of the steroid synthesis gene CYP11a with polycystic ovary syndrome and hyperandrogenism. Hum Mol Genet 6:397–402 [DOI] [PubMed] [Google Scholar]
- 9. Dunaif A, Segal KR, Shelley DR, Green G, Dobrjansky A, Licholai T. 1992. Evidence for distinctive and intrinsic defects in insulin action in polycystic ovary syndrome. Diabetes 41:1257–1266 [DOI] [PubMed] [Google Scholar]
- 10. Baillargeon JP, Diamanti-Kandarakis E, Ostlund RE, Jr, Apridonidze T, Iuorno MJ, Nestler JE. 2006. Altered d-chiro-inositol urinary clearance in women with polycystic ovary syndrome. Diabetes Care 29:300–305 [DOI] [PubMed] [Google Scholar]
- 11. Menke MN, Strauss JF., 3rd 2007. Genetic approaches to polycystic ovarian syndrome. Curr Opin Obstet Gynecol 19:355–359 [DOI] [PubMed] [Google Scholar]
- 12. Legro RS, Driscoll D, Strauss JF, 3rd, Fox J, Dunaif A. 1998. Evidence for a genetic basis for hyperandrogenemia in polycystic ovary syndrome. Proc Natl Acad Sci USA 95:14956–14960 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Xita N, Tsatsoulis A, Chatzikyriakidou A, Georgiou I. 2003. Association of the (TAAAA)n repeat polymorphism in the sex hormone-binding globulin (SHBG) gene with polycystic ovary syndrome and relation to SHBG serum levels. J Clin Endocrinol Metab 88:5976–5980 [DOI] [PubMed] [Google Scholar]
- 14. Cousin P, Calemard-Michel L, Lejeune H, Raverot G, Yessaad N, Emptoz-Bonneton A, Morel Y, Pugeat M. 2004. Influence of SHBG gene pentanucleotide TAAAA repeat and D327N polymorphism on serum sex hormone-binding globulin concentration in hirsute women. J Clin Endocrinol Metab 89:917–924 [DOI] [PubMed] [Google Scholar]
- 15. Bérubé D, Séralini GE, Gagné R, Hammond GL. 1990. Localization of the human sex hormone-binding globulin gene (SHBG) to the short arm of chromosome 17 (17p12–p13). Cytogenet Cell Genet 54:65–67 [DOI] [PubMed] [Google Scholar]
- 16. Hammond GL. 1990. Molecular properties of corticosteroid binding globulin and the sex-steroid binding proteins. Endocr Rev 11:65–79 [DOI] [PubMed] [Google Scholar]
- 17. Edmunds SE, Stubbs AP, Santos AA, Wilkinson ML. 1990. Estrogen and androgen regulation of sex hormone binding globulin secretion by a human liver cell line. J Steroid Biochem Mol Biol 37:733–739 [DOI] [PubMed] [Google Scholar]
- 18. Nestler JE, Powers LP, Matt DW, Steingold KA, Plymate SR, Rittmaster RS, Clore JN, Blackard WG. 1991. A direct effect of hyperinsulinemia on serum sex hormone-binding globulin levels in obese women with the polycystic ovary syndrome. J Clin Endocrinol Metab 72:83–89 [DOI] [PubMed] [Google Scholar]
- 19. Plymate SR, Matej LA, Jones RE, Friedl KE. 1988. Inhibition of sex hormone-binding globulin production in the human hepatoma (Hep G2) cell line by insulin and prolactin. J Clin Endocrinol Metab 67:460–464 [DOI] [PubMed] [Google Scholar]
- 20. Ring HZ, Lessov CN, Reed T, Marcus R, Holloway L, Swan GE, Carmelli D. 2005. Heritability of plasma sex hormones and hormone binding globulin in adult male twins. J Clin Endocrinol Metab 90:3653–3658 [DOI] [PubMed] [Google Scholar]
- 21. Jaquish CE, Blangero J, Haffner SM, Stern MP, MacCluer JW. 1997. Quantitative genetics of serum sex hormone-binding globulin levels in participants in the San Antonio Family Heart Study. Metabolism 46:988–991 [DOI] [PubMed] [Google Scholar]
- 22. Kalyani RR, Franco M, Dobs AS, Ouyang P, Vaidya D, Bertoni A, Gapstur SM, Golden SH. 2009. The association of endogenous sex hormones, adiposity, and insulin resistance with incident diabetes in postmenopausal women. J Clin Endocrinol Metab 94:4127–4135 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Perry JR, Weedon MN, Langenberg C, Jackson AU, Lyssenko V, Sparsø T, Thorleifsson G, Grallert H, Ferrucci L, Maggio M, Paolisso G, Walker M, Palmer CN, Payne F, Young E, Herder C, Narisu N, Morken MA, Bonnycastle LL, Owen KR, Shields B, Knight B, Bennett A, Groves CJ, Ruokonen A, Jarvelin MR, Pearson E, Pascoe L, Ferrannini E, Bornstein SR, Stringham HM, Scott LJ, Kuusisto J, Nilsson P, Neptin M, Gjesing AP, Pisinger C, Lauritzen T, Sandbaek A, Sampson M, Zeggini E, Lindgren CM, Steinthorsdottir V, Thorsteinsdottir U, Hansen T, Schwarz P, Illig T, Laakso M, Stefansson K, Morris AD, Groop L, Pedersen O, Boehnke M, Barroso I, Wareham NJ, Hattersley AT, McCarthy MI, Frayling TM. 2010. Genetic evidence that raised sex hormone binding globulin (SHBG) levels reduce the risk of type 2 diabetes. Hum Mol Genet 19:535–544 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Ding EL, Song Y, Malik VS, Liu S. 2006. Sex differences of endogenous sex hormones and risk of type 2 diabetes: a systematic review and meta-analysis. JAMA 295:1288–1299 [DOI] [PubMed] [Google Scholar]
- 25. Ding EL, Song Y, Manson JE, Hunter DJ, Lee CC, Rifai N, Buring JE, Gaziano JM, Liu S. 2009. Sex hormone-binding globulin and risk of type 2 diabetes in women and men. N Engl J Med 361:1152–1163 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Lindstedt G, Lundberg PA, Lapidus L, Lundgren H, Bengtsson C, Björntorp P. 1991. Low sex-hormone-binding globulin concentration as independent risk factor for development of NIDDM. 12-yr follow-up of population study of women in Gothenburg, Sweden. Diabetes 40:123–128 [DOI] [PubMed] [Google Scholar]
- 27. Ferk P, Teran N, Gersak K. 2007. The (TAAAA)n microsatellite polymorphism in the SHBG gene influences serum SHBG levels in women with polycystic ovary syndrome. Hum Reprod 22:1031–1036 [DOI] [PubMed] [Google Scholar]
- 28. Spielman RS, McGinnis RE, Ewens WJ. 1993. Transmission test for linkage disequilibrium: the insulin gene region and insulin-dependent diabetes mellitus (IDDM). Am J Hum Genet 52:506–516 [PMC free article] [PubMed] [Google Scholar]
- 29. Biyasheva A, Legro RS, Dunaif A, Urbanek M. 2009. Evidence for association between polycystic ovary syndrome (PCOS) and TCF7L2 and glucose intolerance in women with PCOS and TCF7L2. J Clin Endocrinol Metab 94:2617–2625 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Dunaif A, Scott D, Finegood D, Quintana B, Whitcomb R. 1996. The insulin-sensitizing agent troglitazone improves metabolic and reproductive abnormalities in the polycystic ovary syndrome. J Clin Endocrinol Metab 81:3299–3306 [DOI] [PubMed] [Google Scholar]
- 31. Legro RS, Bentley-Lewis R, Driscoll D, Wang SC, Dunaif A. 2002. Insulin resistance in the sisters of women with polycystic ovary syndrome: association with hyperandrogenemia rather than menstrual irregularity. J Clin Endocrinol Metab 87:2128–2133 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. 1985. Homeostasis model assessment: insulin resistance and β-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 28:412–419 [DOI] [PubMed] [Google Scholar]
- 33. Genuth S, Alberti KG, Bennett P, Buse J, Defronzo R, Kahn R, Kitzmiller J, Knowler WC, Lebovitz H, Lernmark A, Nathan D, Palmer J, Rizza R, Saudek C, Shaw J, Steffes M, Stern M, Tuomilehto J, Zimmet P. 2003. Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care 26:3160–3167 [DOI] [PubMed] [Google Scholar]
- 34. Abecasis GR, Cherny SS, Cookson WO, Cardon LR. 2002. Merlin: rapid analysis of dense genetic maps using sparse gene flow trees. Nat Genet 30:97–101 [DOI] [PubMed] [Google Scholar]
- 35. Bendlová B, Zavadilová J, Vanková M, Vejrazková D, Lukásová P, Vcelák J, Hill M, Cibula D, Vondra K, Stárka L, Vrbíková J. 2007. Role of D327N sex hormone-binding globulin gene polymorphism in the pathogenesis of polycystic ovary syndrome. J Steroid Biochem Mol Biol 104:68–74 [DOI] [PubMed] [Google Scholar]
- 36. Maggio M, Lauretani F, Basaria S, Ceda GP, Bandinelli S, Metter EJ, Bos AJ, Ruggiero C, Ceresini G, Paolisso G, Artoni A, Valenti G, Guralnik JM, Ferrucci L. 2008. Sex hormone binding globulin levels across the adult lifespan in women: the role of body mass index and fasting insulin. J Endocrinol Invest 31:597–601 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Morisset AS, Blouin K, Tchernof A. 2008. Impact of diet and adiposity on circulating levels of sex hormone-binding globulin and androgens. Nutr Rev 66:506–516 [DOI] [PubMed] [Google Scholar]
- 38. De Simone M, Verrotti A, Iughetti L, Palumbo M, Farello G, Di Cesare E, Bernabei R, Rosato T, Lozzi S, Criscione S. 2001. Increased visceral adipose tissue is associated with increased circulating insulin and decreased sex hormone binding globulin levels in massively obese adolescent girls. J Endocrinol Invest 24:438–444 [DOI] [PubMed] [Google Scholar]
- 39. Riancho JA, Valero C, Zarrabeitia MT, García-Unzueta MT, Amado JA, González-Macías J. 2008. Genetic polymorphisms are associated with serum levels of sex hormone binding globulin in postmenopausal women. BMC Med Genet 9:112. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Nakhla AM, Hryb DJ, Rosner W, Romas NA, Xiang Z, Kahn SM. 2009. Human sex hormone-binding globulin gene expression: multiple promoters and complex alternative splicing. BMC Mol Biol 10:37. [DOI] [PMC free article] [PubMed] [Google Scholar]