Sex hormone-binding globulin gene expression and insulin resistance.

J C E M

O N L I N E

A d v a n c e s

i n

G e n e t i c s — E n d o c r i n e

R e s e a r c h

Sex Hormone-Binding Globulin Gene Expression and Insulin Resistance Stephen J. Winters, Jyothi Gogineni, Marjan Karegar, Charles Scoggins, Chris A. Wunderlich, Richard Baumgartner, and Dushan T. Ghooray Division of Endocrinology, Metabolism and Diabetes (S.J.W., J.G., M.K., D.T.G.), Division of Surgical Oncology (C.S.), Clinical Pathology Associates, Norton Healthcare (C.A.W.), and Department of Epidemiology and Population Health (R.B.), University of Louisville, Louisville, Kentucky 40202

Context: The plasma level of sex hormone binding globulin (SHBG), a glycoprotein produced by hepatocytes, is subject to genetic, hormonal, metabolic, and nutritional regulation, and is a marker for the development of the metabolic syndrome and diabetes. Objective: Because the mechanism for these associations is unclear, and no studies of SHBG gene expression in humans have been published, SHBG mRNA was measured in human liver samples and related to anthropometric data. Setting: Inpatients at a private, nonprofit, university-associated hospital were studied. Participants: Subjects were fifty five adult men and women undergoing hepatic resection as treatment for cancer. Main Outcome Measures: Main outcome measures were SHBG mRNA and serum SHBG levels. Results: SHBG mRNA was a strong predictor of serum SHBG with higher levels of the mRNA and protein in women than in men. The relationship between SHBG mRNA and circulating SHBG differed in males and females consistent with a sex difference in post-transcriptional regulation. A strong positive correlation was found between the level of the mRNA for the transcription factor HNF4␣ and SHBG mRNA. Insulin resistance (IR), assessed by homeostatis model assessment, was related inversely to SHBG mRNA and to HNF4␣ mRNA as well as to circulating SHBG levels. These mRNAs, as well as serum SHBG, were higher when the hepatic triglyceride concentration was low, and decreased with increasing body mass index but were unrelated to age. Conclusions: Fat accumulation in liver and IR are important determinants of SHBG gene expression and thereby circulating SHBG levels that are perhaps mediated through effects on the transcription factor HNF4␣. These findings provide a potential mechanism to explain why low SHBG predicts the development of type 2 diabetes. (J Clin Endocrinol Metab 99: E2780 –E2788, 2014)

S

ex hormone– binding globulin (SHBG) is a 90 –100 KDa homodimeric glycoprotein that is encoded by a single gene on the short arm of chromosome 17, and is produced primarily by hepatocytes (1). SHBG transports testosterone (T) and other steroids in the blood plasma, reduces their metabolic clearance rate, and regulates their access to target tissues (2). In addition, there is

some evidence that ligand-bound SHBG binds to membrane receptors, and stimulates cAMP production (3), and/or enters cells by binding to the membrane protein megalin (4) to initiate a biological effect. Although the level of SHBG in a given individual is relatively constant (5), and is unrelated to meals or time of day (6), there is a 10-fold variation among individuals that is influenced

ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright © 2014 by the Endocrine Society Received June 12, 2014. Accepted September 10, 2014. First Published Online September 16, 2014

Abbreviations: BMI, body mass index; HNF4␣, hepatocyte nuclear factor 4-␣; HOMA, homeostatis model assessment; IR, insulin resistance; MetS, metabolic syndrome; n.s., not significant; SHBG, sex hormone binding globulin; T, testosterone; T2DM, type 2 diabetes.

E2780

jcem.endojournals.org

J Clin Endocrinol Metab, December 2014, 99(12):E2780 –E2788

doi: 10.1210/jc.2014-2640

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 December 2014. at 11:00 For personal use only. No other uses without permission. . All rights reserved.

doi: 10.1210/jc.2014-2640

by genetic, developmental, hormonal, and nutritional factors (7, 8). SHBG levels are lower with increasing obesity (9), and increase following weight loss, either by dieting (10) or after bariatric surgery (11). SHBG levels are reduced in type 2 diabetes (T2DM) (12, 13) and in patients with metabolic syndrome (MetS) (14). Moreover, a low level of SHBG is associated with an increased risk for developing MetS (15), gestational diabetes (16), and T2DM (17–19), as well as the cardiovascular disease that accompanies these conditions (20). Furthermore, SHBG genetic variation has been proposed to contribute to the development of T2DM (21, 22). Accordingly, SHBG has emerged as a biomarker for insulin resistance (IR) (23). However, the mechanism for this association remains controversial. Low SHBG in IR has been attributed to hyperinsulinemia (24). Studies have found an inverse correlation between SHBG and fasting (21), glucose-stimulated (25) or 24-hour mean insulin or C-peptide levels (26, 27), and SHBG levels increase when IR improves and insulin levels decline with weight loss (10, 28) or following treatment with insulin-sensitizing drugs (29). SHBG levels also increased when insulin levels were reduced by administering diazoxide to women with polycystic ovary syndrome (30) or to obese men (31). Moreover, adding insulin to HepG2 hepatocarcinoma cells reduced the production of SHBG protein (32, 33) and decreased SHBG mRNA levels (33). More recent studies, however, also using HepG2 cells, found no effect of insulin on SHBG secretion or mRNA, which instead were reduced by glucose and fructose (34) and by rosiglitazone (35). SHBG levels are not low in patients with type 1 diabetes (36) who are also hyperglycemic, however; and SHBG levels increase rather than decrease when patients with T2DM are treated with rosiglitazone (37). Several recent studies found that obese IR patients with fatty liver disease have low SHBG levels (38, 39), and one study of subjects at risk for T2DM found that the amount of liver fat was the strongest predictor of SHBG, and adjustment for liver fat abolished the relationship between SHBG and insulin sensitivity as determined by euglycemic hyperinsulinemic clamp (40). The nuclear receptor hepatocyte nuclear factor 4-␣ (HNF4␣) activates the promoters of multiple genes expressed in liver that function in lipid metabolism (41). The proximal promoter of the SHBG gene contains an HNF4␣ binding site, and overexpression by transient transfection of HNF4␣ in HepG2 cells increased transcription of a SHBG-luciferase reporter (42). Although the ligand activation of HNF4␣ is not well understood, TNF␣ (43) and IL1␤ (44) were shown to reduce SHBG expression in


E2781

HepG2 cells by decreasing HNF4␣ through a mechanism involving nuclear factor ␬B (45). Partly because SHBG mRNA is not naturally expressed in rodent liver, there has been limited research on SHBG gene expression using in vivo models, and to our knowledge no studies of SHBG expression in humans have been reported. Accordingly, this research was initiated to better understand the factors that regulate SHBG gene expression in human liver and the level of SHBG in plasma, with the long-term goal of understanding why SHBG predicts the development of the MetS and T2DM, and whether novel therapeutic strategies might evolve from this knowledge.

Materials and Methods Subjects Adult men and women undergoing hepatic resection as treatment for cancer were recruited for this study approved by the Institutional Board of the University of Louisville. Subjects were Eastern Cooperative Oncology Group performance status 0: fully active, and able to carry on all predisease performance without restriction. Subjects with other liver diseases, such as hepatitis C, were excluded. The time from diagnosis to surgery was 1–5 weeks. During this time, there was a median change in weight of ⫺5 lbs (range, ⫺20 to ⫹10 lbs). Patients were not instructed to take nutritional supplements before surgery, and no patients were treated with chemotherapy or x-irradiation. Following informed consent, the patient’s medical history was reviewed and anthropometric data were collected, and a fasting blood sample was obtained in which glucose was measured in a biomedical panel and an aliquot was frozen at ⫺70°C for the measurement of SHBG, insulin, and T in males. After resection of the liver specimen at surgery, tumor was removed for clinical analysis by the pathologist, and adjacent normal liver was stored in RNAlater (Life Technologies) for subsequent analysis. Insulin resistance (IR) was calculated using the homeostatis model assessment (HOMA) method in which HOMA-IR ⫽ fasting insulin (mU/L) ⫻ glucose (mg/dl)/405. Metabolic syndrome (MetS) was defined using the definition advocated by the National Cholesterol Education Program’s Adult Treatment Panel III (NCEP:ATPIII).

RNA isolation and real-time RT-PCR analysis Total RNA was extracted from liver tissue using RNAeasy columns (QIAGEN). The concentration of total RNA was determined using a spectrophotometer. Sample purity was determined by calculating the ratio of sample absorbance at 260: 280 nm, and samples were rejected if below 1.8. RNA (1 ␮g) was reverse transcribed using an oligo dT (12–24) as the primer. qRT-PCR was performed on a Stratagene MX4000 Multiplex Quantitative PCR System using gene-specific primers: SHBG (NM_001040.3) Forward GACCTCACCAAGA-TCACA, Reverse TGCCTGAGTCCCTGGA. HNF4␣ (NM_000457.3) Forward GGACAGATGTGTGAGTGGCCCCGAC, Reverse CCAGAGCAGGGCGTCAAGGGT. These primers amplified a single band of 516 and 374 bp following gel electrophoresis of


E2782

Winters et al

SHBG mRNA and Insulin Resistance

the PCR product. Accumulation of PCR products was monitored in real time, and the crossover threshold was determined using Mx4000 software (Stratagene). The specificity of real-time PCR result was confirmed by analysis of the melting curve and PCR product separated by agarose gel electrophoresis. Standard curves of known amounts of the target mRNAs were prepared from human liver as previously described (46).

Immunoassays Serum T and SHBG levels were measured with ELISA kits from ALPCO Diagnostics. Insulin was measured using a specific ELISA kit from Mercodia. Hepatic triglyceride levels were determined using a Triglyceride Colorimetric Assay (Cayman Chemical Co). Tissue was homogenized in diluent (100 mg/0.4 ml) containing 10 ␮l/ml protease inhibitors (RIPA Lysis buffer; Santa Cruz Biotechnology) and samples were diluted 1:5 for assay.

Hepatic steatosis The degree of steatosis in sections of normal liver was assessed in a blinded fashion by a hepatopathologist (C.W.) using a morphological semiquantitative approach with a four-graded scale: 0 –3, corresponding to none, slight, moderate, and severe, and scores 1–3 corresponding roughly to ⬍ 1%, 1–15%, and ⬎ 15% fat deposition.

Statistical analysis All study variables with the exception of sex and T2DM diagnosis were continuously distributed. Preliminary analyses indicated several variables (HOMA, serum SHBG, SHBG mRNA, HNF4␣ mRNA) were positively skewed. As a result some statistical tests were conducted using raw variables as well as variables transformed to approximate normal distribution. Differences in means between groups (sexes, T2DM) were tested using the Student t test on both raw and transformed variables. Because there were no meaningful differences for the statistical significance of group differences regardless of transformation, only results for means of raw variables are provided to simplify interpretation. Associations among variables were analyzed using Pearson correlations and least squares linear regression. Spearman correlations among variables were calculated also, but not found to differ meaningfully from Pearson. Least squares regression analysis is generally robust to departures from normal distributions and was therefore conducted using only the raw, untransformed variables to simplify interpretations. Bivariate

Table 1.


scatterplots were created and inspected visually for departures from linearity. Where a nonlinear relationship seemed possible, we fitted an appropriate polynomial regression and evaluated the improvement in fit using partial R2s. Multiple linear regression analysis was used to identify the best combination of independent variables that predicted a dependent variable of interest. The nominal level for statistical significance for all analyses was set at ␣ ⫽ 0.05; no corrections were made for multiple comparisons because the sample size was small. All analyses were conducted using STATA version 11.2.

Results Serum and liver tissues were received from 55 subjects. Of these, 49 (23 women and 26 men) agreed to a medical interview and a review of their medical record. Twenty seven patients had metastatic colorectal cancer, seven had hepatocellular cancer, three had renal cell cancer, and 12 had miscellaneous cancers that were treated by partial hepatectomy. Nineteen patients were diagnosed with T2DM, of whom eight were treated with antidiabetic drugs and four with insulin, and 18 patients were diagnosed with dyslipidemia and were treated with a lipid-lowering agent. Descriptive statistics for the study variables are shown in Table 1. The patients ranged in age from 39 – 82 years (61.8 ⫾ 11.2 y; mean ⫾ SEM). They weighed 43.1–130 kg (84.5 ⫾ 20.2 kg), and their body mass index (BMI) ranged from 16 – 45 kg/m2 (28.9 ⫾ 6.5 kg/m2). There were no significant differences between men and women for age, BMI, serum insulin, HOMA-IR, HNF4␣ mRNA, levels or hepatic triglyceride content. Serum SHBG levels ranged from 40 –242 nmol/L in women, and from 27–182 nmol/L in men; mean levels in women were 2-fold higher (P ⬍ .01) than in men. There was a 40-fold variation in hepatic SHBG mRNA concentrations, and mean levels in women were 42% higher (P ⫽ .07) than in men. Figure 1 illustrates the strong positive relationship between SHBG mRNA in liver and serum SHBG levels

Descriptive Statistics for Study Participants

Characteristic Age, years BMI, kg/m2 Insulin, mU/La HOMA-IRa Serum SHBG, nmol/L SHBG mRNA, copies/␮g RNA HNF4␣ mRNA, copies/␮g RNA Hepatic Triglyceride, mg/dl

Men (n ⴝ 26) Mean ⴞ SE

Women (n ⴝ 23) Mean ⴞ SE

P Value

63.15 ⫾ 2.11 29.59 ⫾ 1.14 10.27 ⫾ 1.27 2.94 ⫾ 0.44 63.66 ⫾ 7.62 0.89 ⫾ 0.09 ⫻ 106 1.17 ⫾ 0.13 ⫻ 107

60.17 ⫾ 2.44 28.11 ⫾ 1.49 8.57 ⫾ 1.96 1.98 ⫾ 0.46 116.75 ⫾ 13.30 1.24 ⫾ 0.17 ⫻ 106 1.22 ⫾ 0.23 ⫻ 107 538 ⫾ 107

ns ns ns ns ⬍.001 .07 ns ns

Abbreviation: ns, not significant. a

Due to missing data, n ⫽ 24 in men and n ⫽ 22 in women.


doi: 10.1210/jc.2014-2640


Figure 1. Association of serum SHBG with hepatic expression of SHBG mRNA in women and men undergoing partial hepatectomy. Subject A was treated with tamoxifen, and two women were treated with oral estradiol.B, The straight lines represent the linear regression of serum SHBG on SHBG mRNA for men (R2 ⫽ 0.45, P ⫽ .005) and women (R2 ⫽ 0.32, P ⫽ .003). Of the additional three women with high serum SHBG levels, one was treated with spironolactone, and one was thin with a BMI of 16 kg/m2. Of the three men with high values, one had a BMI of 16.1 kg/m2, and one was treated with dutasteride. The manufacturer’s reference range for serum SHBG was 7–70 nmol/L for men and 15–120 nmol/L for women.

(R2 ⫽ 0.40, P ⬍ .001). Interestingly, SHBG protein and mRNA levels were high in two estradiol-treated women, but only serum SHBG was elevated in the one woman treated with the selective estrogen receptor modulator tamoxifen. As a result, these three women were excluded from subsequent regression analyses. For both men (b ⫽ 5.4 ⫾ 1.2 ⫻ 10⫺5, R2 ⫽ 0.45, P ⫽ .005) and women (b ⫽ 5.5 ⫾ 1.6 ⫻ 10⫺5, R2 ⫽ 0.32, P ⫽ .003), there was a statistically significant positive association between these variables of closely similar magnitude and slope, offset by the sex difference in mean serum SHBG and SHBG mRNA (intercept ⫽ 15.13 in men vs 39.62 in women) and taken together, sex and SHBG mRNA explained 51% of the variance in serum SHBG. Bivariate (Pearson) correlations between the variables are shown in Table 2. Values for men are below the dotted Table 2.

E2783

lines that form a diagonal in the table, and those for women are above the diagonal. Hence, the correlation for BMI with insulin in men (r ⫽ 0.361) can be found below the diagonal, and the corresponding correlation in women (r ⫽ 0.514) is above the diagonal. Statistically significant correlations (P ⬍ ⫽ .01) are shown in bold. None of the variables correlated significantly with age in women or men. As expected, BMI was positively correlated with circulating insulin levels, HOMA-IR, and hepatic triglyceride concentrations, but was negatively correlated with serum SHBG in both genders. In regression analysis, SHBG mRNA was inversely related to BMI (R2 ⫽ 0.087: P ⬍ .025) more so among men (R2 ⫽ 0.13) than among women (R2 ⫽ 0.039). SHBG mRNA increased with age in women (R2 ⫽ 0.16; P ⫽ .05) but not men (R2 ⫽ 0.0008), and there was a nonsignificant positive association (R2 ⫽ 0.06) between plasma T and SHBG mRNA among men. Given that overexpression of the transcription factor HNF4␣ stimulates the SHBG promoter in HepG2 hepatocarcinoma cells (42), we measured HNF4␣ mR⌵〈 in human liver. As shown in Figure 2, there was a 150-fold variation in HNF4␣ mRNA, and the level of HNF4␣ mRNA was a strong predictor of SHBG mRNA (overall R2 ⫽ 0.30, P ⬍ .001). This association was somewhat stronger in men (b ⫽ 0.05 ⫾ 0.01, R2 ⫽ 0.46, P ⬍ .001) than in women (b ⫽ 0.03 ⫾ 0.01, R2 ⫽ 0.24, P ⫽ .03); however, there was no statistically significant interaction between sex and HNF4␣ on SHBG mRNA. Taken together, sex and HNF4␣ explained 35% of the variance in liver SHBG mRNA expression. HNF4␣ mRNA and SHBG mRNA expression levels were inversely associated with BMI in men (r ⫽ ⫺0.34 and ⫺0.37, respectively) and women (r ⫽ ⫺0.27 and ⫺0.170, respectively), but these associations were not statistically significant. SHBG and HNF4␣ expression were significantly inversely associated with HOMA in men (r’s ⬎ ⫺0.52; P ⬍ .05) but not in women. HNF4␣ was also inversely associated with insulin

Pearson Correlations Among Variables Women

Men

Age BMI Insulin HOMA-IR Serum SHBG SHBG mRNA HNF4␣ mR⌵〈 Hepatic Triglyceride

Age

BMI

Insulina

HOMA-IRa

Serum SHBG

SHBG mRNA

HNF4␣ mRNA

Trig

— ⫺0.119 0.040 0.097 ⫺0.032 ⫺0.029 ⫺0.005 ⫺0.050

⫺0.051 — 0.361 0.502 ⫺0.342 ⫺0.366 ⫺0.349 0.334

⫺0.194 0.514 — 0.909 ⫺0.236 ⴚ0.522 ⴚ0.528 0.1297

⫺0.284 0.463 0.830 — ⫺0.290 ⴚ0.570 ⴚ0.517 0.205

0.080 ⫺0.129 0.061 0.088 — 0.668 0.137 ⫺0.337

0.178 ⫺0.272 ⫺0.109 ⫺0.070 0.566 — 0.676 ⫺0.324

⫺0.031 ⫺0.170 ⫺0.253 ⫺0.227 ⫺0.074 0.502 — ⫺0.216

⫺0.003 0.434 0.412 0.244 ⫺0.305 ⫺0.207 ⫺0.071 —

Correlations for men (n ⫽ 26) are below the dashed diagonal; women (n ⫽ 23) are above the dashed diagonal. Statistically significant correlations (P ⬍ .05) are shown in bold. a

Due to missing data, n ⫽ 24 in men and n ⫽ 22 in women.


E2784

Winters et al


Figure 2. Association between the expression levels in liver of SHBG mRNA with HNF4␣ mRNA in women and men undergoing partial hepatectomy. The straight line represents the linear regression of SHBG mRNA expression on HNF4␣ mRNA, excluding women treated with tamoxifen or oral estradiol. R2 ⫽ 0.30, P ⬍ .001; values were R2 ⫽ 0.46, P ⬍ .001 for men, and R2 ⫽ 0.24, P ⫽ .03 for women.

in men (r ⫽ ⫺0.39; P ⬍ .05). Although serum SHBG was significantly (P ⬍ .05) positively correlated with SHBG mRNA expression in both men (r ⫽ 0.67) and women (r ⫽ 0.57), it was unaffected by HNF4␣ despite significant positive associations between SHBG mRNA and HNF4␣ in both sexes (0.68 in men; 0.50 in women). Low SHBG levels have been linked to insulin resistance (12) and to hyperinsulinemia (47). Figure 3 shows the regressions of SHBG mRNA and HNF4␣ mRNA on HOMA-IR by sex, which was b⫽ ⫺11.4 ⫾ 3.50 ⫻ 104 (R2 ⫽ 0.33, P ⫽ .004) in men and b ⫽ ⫺6.22 ⫾ 7.53 ⫻ 104 (R2 ⫽ 0.04, n.s.) in women. The corresponding regressions for HNF4␣ on HOMA-IR were b ⫽ ⫺16.34 ⫾ 5.77 ⫻ 105 (R2 ⫽ 0.27, P ⫽ .01) in men and b ⫽ ⫺13.74 ⫾

Figure 3. Association of liver expression of SHBG mRNA,A; and HNF4␣ B, with HOMA-IR in women and men undergoing partial hepatectomy. The straight lines represent the linear regressions of SHBG mRNA and HNF4␣ on HOMA-IR in men and women, respectively. R2 ⫽ 0.33, P ⫽ .004 for SHBG mRNA on HOMA-IR in men, and R2 ⫽ 0.04, n.s. in women. The regressions for HNF4␣ on HOMA-IR were R2 ⫽ 0.27, P ⫽ .01 in men and R2 ⫽ 0.06, n.s. in women.


13.49 ⫻ 105 (R2 ⫽ 0.06, n.s.) in women. Although these differences in regression lines suggest modification of the associations of HOMA-IR with SHBG mRNA and HNF4␣ by sex, interaction terms were not statistically significant, possibly due to small sample sizes and low statistical power. Because serum SHBG levels are low in individuals with nonalcoholic fatty liver disease even after adjusting for BMI (38), fat accumulation was evaluated by measuring hepatic triglyceride content, and the relationship to SHBG was examined. Hepatic triglyceride levels varied 30-fold from 134 –2679 mg/dl. Although the linear associations were nonsignificant (Table 2), there were significant inverse curvilinear relationships between hepatic triglyceride concentration and the levels of both serum SHBG (R2 ⫽ 0.16, P ⫽ .02) and SHBG mRNA (R2 ⫽ 0.23, P ⫽ .003) when fitted using second degree polynomials (Figure 4). These curvilinear associations did not differ by sex and were independent of HNF4␣, age, BMI, HOMA, and diagnosis of T2DM. In contrast, there was a statistically significant curvilinear association between HNF4␣ and hepatic triglycerides in men (R2 ⫽ 0.23, P ⫽ .04) but not in women (R2 ⫽ 0.09, n.s.), and the overall R2 was 0.1 (P ⫽ .12). There was good agreement between the level of hepatic triglyceride and the degree of steatosis estimated histologically (Supplemental Table 1). We also calculated the ratio of SHBG/HNF4 mRNAs, and found that it was unrelated to hepatic triglyceride levels (R2 ⫽ 0.01) Subjects with self-reported T2DM had lower levels of circulating SHBG (P ⫽ .03) than nondiabetics whereas HNF4␣ mRNAs and hepatic triglycerides were not significantly different (Supplemental Table 2). When subjects were divided into two groups based on low (134 – 412 mg/dl) or high (415–2679 mg/dl) hepatic triglyceridelevels, subjects with high hepatic triglyceride levels were more insulin resistant (P ⬍ .01) based on HOMA-IR (3.31 ⫾ 0.51 vs 1.60 ⫾ 0.28) and had a higher fasting blood glucose level (122 ⫾ 9.5 vs 99 ⫾ 2.8 mg/dl; P ⫽ .02). Among subjects for whom the level of HbA1C was known (n ⫽ 28), there was no relationship between A1c and hepatic triglyceride content (R2 ⫽ 0.003). Because serum SHBG levels represent a biomarker for the development of MetS and T2DM, we sought to determine the optimal combination of study variables that predicted variation in serum SHBG using multiple regression methods. The best equation for predicting serum SHBG consisted of SHBG mRNA, HNF4␣ mRNA, and sex and is shown in Supplemental Table 3. It is interesting that SHBG mRNA and HNF4␣ mRNA have statistically significant, independent associations with serum SHBG despite their strong positive correlation with each other.


doi: 10.1210/jc.2014-2640


E2785

Discussion

Figure 4. Association between serum SHBG,A; SHBG mRNA, B; and HNF4␣ mRNA, C with hepatic triglyceride levels in women and men undergoing partial hepatectomy. The curvilinear lines were fitted using second-degree polynomials as described in the Methods section. There was a statistically significant curvilinear association between HNF4␣ and hepatic triglycerides in men (R2 ⫽ 0.23, P ⫽ .04) but not women (R2 ⫽ 0.09, n.s.), and the overall R2 was 0.1 (P ⫽ .12).

Moreover, HNF4␣ mRNA seems to be inversely associated with serum SHBG when controlling for SHBG mRNA and sex. Taken together, SHBG mRNA, HNF␣ mRNA, and sex explained 67% of the variance in serum SHBG. Because no other variables, including age, BMI, insulin, HOMA, or hepatic triglycerides were significantly (P ⬍ .05) associated with serum SHBG over and above SHBG mRNA, HNF␣ mRNA, and sex, this analysis suggests that these are the primary, proximal determinants of circulating SHBG levels.

Although SHBG has emerged as an early marker for the development of the metabolic syndrome and T2DM, the lack of an animal model to study metabolic control of SHBG expression has limited understanding of the mechanisms that link SHBG to these conditions. To the best of our knowledge, this is the first study to examine SHBG gene expression in humans. SHBG mRNA expression levels in surgical specimens of human liver were found to be an important determinant of the plasma level of SHBG, but for a given level of SHBG mRNA, the plasma SHBG was higher in women than in men. Circulating SHBG levels are known to increase with estrogen treatment and during pregnancy. Each SHBG monomer has one O-linked glycosylation site on Thr7 and two N-linked glycosylation sites on Asn351 and Asn367, and is modified by estrogens (48) which may prolong clearance (49) to increase the SHBG level in plasma. Nearly all of our female subjects were postmenopausal, however, whose estradiol levels are similar to, if not lower than, those of men (50). Thus, the sex difference may be partly through T because T treatment lowers circulating SHBG slightly (51). SHBG mRNA levels were also higher in women than in men, and were especially high in two women who were treated with estradiol. Although the molecular mechanism for this effect is not known, ␮M concentrations of estradiol increased SHBG mRNA in Hep-G2 cell cultures that were stably transfected with a human estrogen receptor-␣ expression vector (52), and increased estradiol secretion by these cells (53). Circulating SHBG levels are known to increase when postmenopausal women are treated with tamoxifen (54). An interesting finding in this study was that circulating SHBG was elevated whereas SHBG mRNA was low in the one woman treated with tamoxifen, implying a post-transcriptional effect although tamoxifen-like estradiol increased SHBG mRNA in Hep G2/ H5A cells (55). Tamoxifen might also increase serum SHBG through a reduction in GH and IGF-I (56). Androgens may also contribute to the sex difference in SHBG mRNA because one study in two castrated male monkeys found that the level of SHBG mRNA in liver increased but plasma SHBG level declined following treatment with T (57) although the inverse relationship between SHBG mRNA and serum T levels in men in our study was not statistically significant. Prospective studies have shown that low SHBG levels are associated with IR and with an increased risk for developing T2DM (17), and in this study serum SHBG as well as SHBG mRNA levels were lower with increasing IR as assessed by HOMA-IR. The association between SHBG


E2786

Winters et al


genotype, circulating SHBG, and a higher risk for T2DM among those with the SHBG-lowering allele has led to the idea that a change in SHBG function, perhaps related to the bioavailability of sex hormones, may be a primary defect in the pathogenesis of the disorder (22). Our results suggest that the low circulating SHBG in IR can be explained by a lower level of SHBG mRNA. Patients with IR often have high serum insulin concentrations, and there is a substantial body of literature implicating insulin as a negative regulator of SHBG in plasma (47) whereas other results have refuted this hypothesis (34, 40). In our study, there was a weak association between fasting insulin and SHBG mRNA. Fasting insulin levels provide only a limited analysis of 24-hour insulin secretion, however, so that a definitive conclusion on the importance of insulin in the control of SHBG gene expression is not possible from our study. Our results reveal that a low level of SHBG mRNA is largely due to a low level of HNF4␣ mRNA expression in liver. This result was predicted by the increase in SHBG promoter activity when HepG2 cells were transfected with an HNF4␣ expression vector (42). HNF4␣ is a member of the nuclear receptor superfamily, and is a master regulatory protein with many of its target genes involved in metabolic processes such as lipid, organic acid, and carbohydrate metabolism (58). In adult hepatocyte-specific conditional HNF-4␣ knockout mice (H4LivKO), expression of a large number of genes whose gene products are essential for adult liver function is disrupted (59). This protein also influences the insulin secretory pathway and is linked to the rare monogenic disorder, maturity-onset diabetes of the young (MODY-1). Moreover, HNF-4␣ is reduced in liver of obese/insulin resistant db/db mice (60). Results from several human populations suggest that HNF4␣ single-nucleotide polymorphisms and haplotypes are associated with IR and the risk for T2DM (61– 62). Serum SHBG levels as well as HNF4␣ and SHBG mRNAs were lower in subjects with high concentrations of hepatic triglycerides. Hepatic steatosis is associated with IR, and is influenced by the delivery of free fatty acids principally from adipose tissue, de novo lipogenesis in hepatocytes, and by dietary fat intake; however, the mechanism linking IR and hepatic steatosis is incompletely understood (63). In clinical studies, hepatic fat content is generally determined semiquantitatively by ultrasound or by magnetic resonance spectroscopy. Using the latter method, Peter et al (40) reported a strong inverse correlation between liver fat and SHBG in persons at risk for T2DM, and others have confirmed this association (39, 64, 65). Moreover, SHBG levels increase as liver fat decreases with weight loss (66). In one study of men at-risk for metabolic syndrome (67), in contrast, circulating trig-


lycerides and fasting blood glucose, but not hepatic triglyceride content, were inversely related to SHBG. Serum SHBG levels, as well as SHBG and HNF4␣ mRNAs, were higher in those subjects with lower levels of hepatic triglyceride; however, the curvilinear associations suggest there may be a threshold above which the accumulation of liver fat has no further effect on SHBG. This conclusion is tentative, however, given the relatively small sample size. In this study the scatterplots and Pearson correlation coefficients reveal statistically significant associations; however, the strength of these associations, attenuated by uneven tissue distribution of analytes and measurement error, imply that multiple factors regulate SHBG. Serum hormone levels were measured only once on the day of surgery which may have weakened the associations found. Moreover, the samples were from patients with a variety of cancers that might have influenced the results. Although this was a cross-sectional study that cannot establish causality, our findings support the idea that HNF4␣ expression is an important determinant of SHBG perhaps because of IR and signaling pathways initiated by hepatic triglyceride. These results provide a plausible mechanism for the notion that low circulating SHBG levels represent an early biomarker for the inheritance of IR and the predisposition to T2DM (68, 69).

Acknowledgments We thank the staff of the Louisville Tissue Repository Program at Norton Hospital for their help obtaining samples for this research. Address all correspondence and requests for reprints to: Stephen J. Winters, MD, Division of Endocrinology, Metabolism and Diabetes, University of Louisville, ACB-A3G11, 550 Jackson Street, Louisville, KY, 40202. E-mail [email protected]. A portion of this research was presented at 94th Annual Meeting of the Endocrine Society, Houston, TX, June 23–26, 2012. This work was supported in part by a gift from the Walter and Avis Jacobs Foundation. Disclosure Summary: The authors have nothing to disclose.

References 1. Hammond GL, Bocchinfuso WP. Sex hormone-binding globulin: Gene organization and structure/function analyses. Horm Res. 1996;45–197–201. 2. Joseph DR. Structure, function, and regulation of androgen-binding protein/sex hormone-binding globulin. Vitam Horm. 1994;49 – 197–280. 3. Rosner W, Hryb DJ, Khan MS, Nakhla AM, Romas NA. Sex hormone-binding globulin mediates steroid hormone signal transduction at the plasma membrane. J Steroid Biochem Mol Biol. 1999; 69:481– 485.


doi: 10.1210/jc.2014-2640

4. Hammes A, Andreassen TK, Spoelgen R, et al. Role of endocytosis in cellular uptake of sex steroids. Cell. 2005;122:751–762. 5. Jayagopal V, Kilpatrick ES, Jennings PE, Hepburn DA, Atkin SL. The biological variation of testosterone and sex hormone-binding globulin (SHBG) in polycystic ovarian syndrome: Implications for SHBG as a surrogate marker of insulin resistance. J Clin Endocrin Metab. 2003;88:1528 –1533. 6. Sørensen K, Andersson AM, Skakkebaek NE, Juul A. Serum Sex Hormone-Binding Globulin Levels in Healthy Children and Girls with Precocious Puberty before and during Gonadotropin-Releasing Hormone Agonist Treatment. J Clin Endocrinol Metab. 2007;92: 3189 –3196. 7. Anderson DC. Sex-hormone-binding globulin. Clin Endocrinol. 1974;3:69 –96. 8. Vanbillemont G, Lapauw B, De Naeyer H, Roef G, Kaufman JM, Taes YE. Sex hormone-binding globulin at the crossroad of body composition, somatotropic axis and insulin/glucose homeostasis in young healthy men. Clin Endocrinol. 2012;76(1):111–118. 9. Glass AR, Swerdloff RS, Bray GA, Dahms WT, Atkinson RL. Low serum testosterone and sex-hormone-binding-globulin in massively obese men. J Clin Endocrinol Metab. 1977;45:1211–1219. 10. Guzick DS, Wing R, Smith D, Berga SL, Winters SJ. Endocrine consequences of weight loss in obese, hyperandrogenic, anovulatory women. Fertil Steril. 1994;61:598 – 604. 11. Hammoud A, Gibson M, Hunt SC, et al. Effect of Roux-en-Y gastric bypass surgery on the sex steroids and quality of life in obese men. J Clin Endocrinol Metab. 2009;94:1329 –1332. 12. Birkeland KI, Hanssen KF, Torjesen PA, Vaaler S. Level of sex hormone-binding globulin is positively correlated with insulin sensitivity in men with type 2 diabetes. J Clin Endocrinol Metab. 1993;76: 275–278. 13. Goodman-Gruen D, Barrett-Connor E. Sex hormone-binding globulin and glucose tolerance in postmenopausal women. The Rancho Bernardo Study. Diabetes Care. 1997;20:645– 649. 14. Laaksonen DE, Niskanen L, Punnonen K, et al. Sex hormones, inflammation and the metabolic syndrome: A population-based study. Eur J Endocrinol. 2003;149:601– 608. 15. Brand JS, van der Tweel I, Grobbee DE, Emmelot-Vonk MH, van der Schouw YT. Testosterone, sex hormone-binding globulin and the metabolic syndrome: A systematic review and meta-analysis of observational studies. Intl J Epidemiol. 2011;40:189 –207. 16. Veltman-Verhulst SM, van Haeften TW, Eijkemans MJ, de Valk HW, Fauser BC, Goverde AJ. Sex hormone-binding globulin concentrations before conception as a predictor for gestational diabetes in women with polycystic ovary syndrome. Hum Reprod. 2010;25: 3123–3128. 17. Ding EL, Song Y, Manson JE, et al. Sex hormone-binding globulin and risk of type 2 diabetes in women and men. N Engl J Med. 2009;361:1152–1163. 18. Haffner SM, Laakso M, Miettinen H, Mykkänen L, Karhapää P, Rainwater DL. Low levels of sex hormone-binding globulin and testosterone are associated with smaller, denser low density lipoprotein in normoglycemic men. J Clin Endocrinol Metab. 1996;81: 3697–3701. 19. Lindstedt G, Lundberg PA, Lapidus L, Lundgren H, Bengtsson C, Björntorp P. 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. 1991;40:123–128. 20. Calderon-Margalit R, Schwartz SM, Wellons MF, et al. Prospective association of serum androgens and sex hormone-binding globulin with subclinical cardiovascular disease in young adult women: The “Coronary Artery Risk Development in Young Adults” women’s study. J Clin Endocrinol Metab. 2010;95:4424 – 4431. 21. Gascón F, Valle M, Martos R, Ruz FJ, Ríos R, Montilla P, Cañete R. Sex hormone-binding globulin as a marker for hyperinsulinemia and/or insulin resistance in obese children. Eur J Endocrinol. 2000; 143:85– 89.


E2787

22. Perry JR, Weedon MN, Langenberg C, et al. Genetic evidence that raised sex hormone binding globulin (SHBG) levels reduce the risk of type 2 diabetes. Hum Mol Genet. 2010;19:535–544. 23. Le TN, Nestler JE, Strauss JF, 3rd, Wickham EP. 3rd 2012 Sex hormone-binding globulin and type 2 diabetes mellitus. Trends Endocrinol Metab. 2012;23:32– 40. 24. Pugeat M, Crave JC, Elmidani M, et al. Pathophysiology of sex hormone binding globulin (SHBG): Relation to insulin. J Steroid Biochem Molec Biol. 1991;40:841– 849. 25. Galloway PJ, Donaldson MD, Wallace AM. Sex hormone binding globulin concentration as a prepubertal marker for hyperinsulinaemia in obesity. Arch Dis Child. 2001;85:489 – 491. 26. Toscano V, Balducci R, Bianchi P, Guglielmi R, Mangiantini A, Sciarra F. Steroidal and non-steroidal factors in plasma sex hormone binding globulin regulation. J Steroid Biochem Molec Biol. 1992; 43:431– 437. 27. Katsuki A, Sumida Y, Murashima S, et al. Acute and chronic regulation of serum sex hormone-binding globulin levels by plasma insulin concentrations in male noninsulin-dependent diabetes mellitus patients. J Clin Endocrinol Metab. 1996;81:2515–2519. 28. Pasquali R, Vicennati V, Scopinaro N, et al. Achievement of nearnormal body weight as the prerequisite to normalize sex hormonebinding globulin concentrations in massively obese men. Int J Obes Relat Metab Disord. 1997;21:1–5. 29. Nestler JE, Powers LP, Matt DW, et al. A direct effect of hyperinsulinemia on serum sex hormone-binding globulin levels in obese women with the polycystic ovary syndrome. J Clin Endocrinoland Metab. 1991;72:83– 89. 30. Nestler JE, Barlascini CO, Matt DW, et al. Suppression of serum insulin by diazoxide reduces serum testosterone levels in obese women with polycystic ovary syndrome. J Clin Endocrinol Metab. 1989;68:1027–1032. 31. Pasquali R, Casimirri F, De Iasio R, et al. Insulin regulates testosterone and sex hormone-binding globulin concentrations in adult normal weight and obese men. J Clin Endocrinol Metab. 1995;80: 654 – 658. 32. Plymate SR, Matej LA, Jones RE, Friedl KE. Inhibition of sex hormone-binding globulin production in the human hepatoma (Hep G2) cell line by insulin and prolactin. J Clin Endocrinol Metab. 1988;67:460 – 464. 33. Loukovaara M, Carson M, Adlercreutz H. Regulation of production and secretion of sex hormone-binding globulin in HepG2 cell cultures by hormones and growth factors. J Ciin Endocrinol Metab. 1995;80:160 –164. 34. Selva DM, Hogeveen KN, Innis SM, Hammond GL. Monosaccharide-induced lipogenesis regulates the human hepatic sex hormonebinding globulin gene. J Clin Invest. 2007;117:3979 –3987. 35. Selva DM, Hammond GL. Peroxisome-proliferator receptor gamma represses hepatic sex hormone-binding globulin expression. Endocrinology. 2009;150(5):2183–189. 36. Danielson KK, Drum ML, Lipton RB. Sex hormone-binding globulin and testosterone in individuals with childhood diabetes. Diabetes Care. 2008;31:1207–1213. 37. Kapoor D, Channer KS, Jones TH. Rosiglitazone increases bioactive testosterone and reduces waist circumference in hypogonadal men with type 2 diabetes. Diab Vasc Dis Res. 2008;5:135–137. 38. Shin JY, Kim SK, Lee MY, et al. Serum sex hormone-binding globulin levels are independently associated with nonalcoholic fatty liver disease in people with type 2 diabetes. Diabetes Res Clin Pract. 2011;94:156 –162. 39. Flechtner-Mors M, Schick A, Oeztuerk S, et al. 2013 Associations of fatty liver disease and other factors affecting serum SHBG concentrations: A population based study on 1657 subjects. Horm Metabolic Res. 2014;46:287–293. 40. Peter A, Kantartzis K, Machann J, et al. Relationships of circulating sex hormone-binding globulin with metabolic traits in humans. Diabetes. 2010;59:3167–3173. 41. Rhee J, Ge H, Yang W, et al. Partnership of PGC-1alpha and


E2788

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54. 55.

Winters et al


HNF4alpha in the regulation of lipoprotein metabolism. J Biol Chem. 2006;281:14683–14690. Jänne M, Hammond GL. Hepatocyte nuclear factor-4 controls transcription from a TATA-less human sex hormone-binding globulin gene promoter. J Biol Chem. 1998;273:34105–34114. Simó R, Barbosa-Desongles A, Lecube A, Hernandez C, Selva DM. Potential role of tumor necrosis factor-␣ in downregulating sex hormone-binding globulin. Diabetes. 2012;61:372–382. Simó R, Barbosa-Desongles A, Hernandez C, Selva DM. IL1␤ down-regulation of sex hormone-binding globulin production by decreasing HNF-4␣ via MEK-1/2 and JNK MAPK pathways. Mol Endocrinol. 2012;26:1917–1927. Simó R, Barbosa-Desongles A, Sáez-Lopez C, Lecube A, Hernandez C, Selva DM. Molecular mechanism of TNFalpha-induced downregulation of SHBG expression. Mol Endocrinol. 2012;26:438 – 446. Moore JP, Jr, Winters SJ. Weaning and the developmental changes in follicle-stimulating hormone, pituitary adenylate cyclase-activating polypeptide, and inhibin B in the male rat. Biol Reprod. 2008; 78:752–760. Pugeat M, Nader N, Hogeveen K, Raverot G, Déchaud H, Grenot C. Sex hormone-binding globulin gene expression in the liver: Drugs and the metabolic syndrome. Mol Cell Endocrinol. 2010;316: 53–59. Tardivel-Lacombe J, Degrelle H. Hormone-associated variation of the glycan microheterogeneity pattern of human sex steroid-binding protein (hSBP). J steroid Biochem Molec Biol. 1991;39:449 – 453. Cousin P, Déchaud H, Grenot C, Lejeune H, Hammond GL, Pugeat M. Influence of glycosylation on the clearance of recombinant human sex hormone-binding globulin from rabbit blood. J Steroid Biochem Molec Biol. 1999;70:115–121. Jasuja GK, Travison TG, Davda M, et al. Age trends in estradiol and estrone levels measured using liquid chromatography tandem mass spectrometry in community-dwelling men of the Framingham Heart Study. J Gerontol A Biol Sci Med Sci. 2013;68:733–740. Winters SJ, Atkinson L. Serum LH concentrations in hypogonadal men during transdermal testosterone replacement through scrotal skin: Further evidence that ageing enhances testosterone negative feedback. The Testoderm Study Group. Clin Endocrinol. 1997;47: 317–322. Nader N, Raverot G, Emptoz-Bonneton A, et al. Mitotane has an estrogenic effect on sex hormone-binding globulin and corticosteroid-binding globulin in humans. J Clin Endocrinol Metab. 2006; 91:2165–2170. Edmunds SE, Stubbs AP, Santos AA, Wilkinson ML. Estrogen and androgen regulation of sex hormone binding globulin secretion by a human liver cell line. J Steroid Biochem Molec Biol. 1990;37:733– 739. Bruning PF, Bonfrer JM, Hart AA, et al. Tamoxifen, serum lipoproteins and cardiovascular risk. Br J Cancer. 1988;58:497– 499. Mercier-Bodard Christine, Nivet V, Baulier EE. Effects of hormones


56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

on SBP mRNA levels in human cancer cells. J steroid Biochem. 1991;40:777–785. Birzniece V, Sutanto S, Ho KK. Gender difference in the neuroendocrine regulation of growth hormone axis by selective estrogen receptor modulators. J Clin Endocrinol Metabol. 2012;97:E521– E527. Kottler ML, Dang CD, Salmon R, Counis R, Degrelle H. Effect of testosterone on regulation of the level of sex steroid-binding protein mRNA in monkey (Macaca fascicularis) liver. J Mol Endocrinol. 1990;5:253–257. Watt AJ, Garrison WD, Duncan SA. HNF4: A central regulator of hepatocyte differentiation and function. Hepatol. 2003;37:1249 – 1253. Hayhurst GP, Lee YH, Lambert G, Ward JM, Gonzalez FJ. Hepatocyte nuclear factor 4alpha (nuclear receptor 2A1) is essential for maintenance of hepatic gene expression and lipid homeostasis. Mol Cell Biol. 2001;21:1393–1403. Weissglas-Volkov D, Huertas-Vazquez A, Suviolahti E, et al. Common hepatic nuclear factor-4alpha variants are associated with high serum lipid levels and the metabolic syndrome. Diabetes. 2006;55: 1970 –1977. Xie X, Liao H, Dang H, et al. Down-regulation of hepatic HNF4alpha gene expression during hyperinsulinemia via SREBPs. Mol Endocrinol. 2009;23:434 – 443. Saif-Ali R, Harun R, Kamaruddin NA, Al-Jassabi S, Ngah WZ. Association of hepatocyte nuclear factor 4 alpha polymorphisms with type 2 diabetes with or without metabolic syndrome in Malaysia. Biochem Genet. 2012;50:298 –308. Ahmed MH BC. 2011. Non-alcoholic fatty liver disease. In: Byrne CD WS, ed. The Metabolic Syndrome 2nd ed. Chichester, UK: Wiley-Blackwell; 245–277. Polyzos SA, Kountouras J, Tsatsoulis A, et al. Sex steroids and sex hormone-binding globulin in postmenopausal women with nonalcoholic fatty liver disease. Hormones (Athens). 2013;12:405– 416. Jones H, Sprung VS, Pugh CJ, et al. Polycystic ovary syndrome with hyperandrogenism is characterized by an increased risk of hepatic steatosis compared to nonhyperandrogenic PCOS phenotypes and healthy controls, independent of obesity and insulin resistance. J Clin Endocrin Metab. 2012;97:3709 –3716. Stefan N, Schick F, Haring HU. Sex hormone-binding globulin and risk of type 2 diabetes. New Engl J Med. 2009;361:2675–2676; author reply 2677–2678. Bonnet F, Velayoudom Cephise FL, Gautier A, et al. Role of sex steroids, intrahepatic fat and liver enzymes in the association between SHBG and metabolic features. Clin Endocrinol. 2013;79: 517–522. Krishnasamy SS, Chang C, Wang C, Chandiramani R, Winters SJ. Sex Hormone-Binding Globulin and the risk for Metabolic Syndrome in Children of South Asian Origin. Endocr Pract. 2012;1–23. Agirbasli M, Agaoglu NB, Orak N, et al. Sex hormones and metabolic syndrome in children and adolescents. Metabolism. 2009;58: 1256 –1262.


Pathophysiology of sex hormone binding globulin (SHBG): relation to insulin.

Effects of thyroid hormone on sex hormone-binding globulin gene expression in human cells.

Age-dependent sex hormone-binding globulin expression in male rat.

Free estradiol and sex hormone-binding globulin.

Genetic insulin resistance is a potent regulator of gene expression and proliferation in human iPS cells.

Vanadate and insulin stimulate gene 33 expression.

Sex hormone-binding globulin (SHBG) expression in ovarian carcinomas and its clinicopathological associations.

Expression and differential glycosylation of human sex hormone-binding globulin by mammalian cell lines.

VDR Gene variation and insulin resistance related diseases.

Feedback inhibition of insulin gene expression by insulin.

Genetic variation of human sex hormone-binding globulin: evidence for a worldwide bi-allelic gene.

The functions of corticosteroid-binding globulin and sex hormone-binding globulin: recent advances.

Decreased sex hormone binding globulin (SHBG) and insulin-like growth factor binding protein (IGFBP-1) in extreme obesity.

Regulation of gene expression by insulin.

Control of insulin gene expression by glucose.

Factors affecting insulin-regulated hepatic gene expression.

Sex Hormone-Binding Globulin in Children and Adolescents.

Hormonal contraceptives and plasma sex hormone binding globulin.

Sex hormone binding globulin: origin, function and clinical significance.

Plasma sex hormone binding globulin and oestrogen therapy [proceedings].

Sex-biased gene expression and sexual conflict throughout development.

Promoter Architecture and Sex-Specific Gene Expression in Daphnia pulex.

Reduced mRNA and a nonsense mutation in the insulin-receptor gene produce heritable severe insulin resistance.

Gene expression of chicken gonads is sex- and side-specific.