Serum Soluble Glycoprotein 130 Concentration Is Inversely Related to Insulin Sensitivity in Women With Polycystic Ovary Syndrome

The study group consisted of 78 women with PCOS (35 lean, BMI <25 kg/m², and 43 overweight or obese, BMI >25 kg/m²) and 34 healthy normally menstruating women (18 lean and 16 overweight or obese). The subjects' recruitment and diagnosis of PCOS according to Rotterdam criteria (15) was described in detail previously (16). Before entering the study, a physical examination and appropriate laboratory test were performed. A standard 75-g oral glucose tolerance test was performed, and all participants had normal glucose tolerance. None of the women had morbid obesity (BMI >40 kg/m²), cardiovascular disease, hypertension, infections, or other serious medical problems; all were nonsmokers, and they were not taking any anti-inflammatory drugs and drugs known to affect glucose and lipid metabolism. Studies were performed in the PCOS group 3–5 days after a spontaneous menses or independent of cycle phase in the presence of amenorrhea and in regularly cycling women during the early follicular phase (3–5 days) of their menstrual cycle. All analyses were performed after an overnight fast. The study protocol was approved by the ethics committee of the Medical University of Białystok, Poland. All participants gave written informed consent before entering the study. Anthropometric measurements were performed as previously described (16).

Insulin sensitivity was evaluated by the euglycemic-hyperinsulinemic clamp technique as previously described (16). The rate of whole-body glucose uptake (M value) was calculated as the mean glucose infusion rate from 80 to 120 min, corrected for glucose space and normalized per kilogram of fat-free mass (M/FFM).

Plasma glucose was measured immediately by the enzymatic method using a glucose analyzer (YSI 2300 STAT Plus; Yellow Springs Instruments, Yellow Springs, OH). The serum insulin was measured with immunoradiometric assay with the sensitivity of 1 μIU/ml and intra- and interassay coefficients of variation (CVs) <2.2 and 6.5%, respectively (BioSource Europe, Nivelles, Belgium). Serum lipids were measured as previously described (16).

Before estimation of serum insulin, sgp130, IL-6, and sIL6-R, samples were kept frozen at −80°C. Serum sgp130 was measured with a quantitative sandwich enzyme immunoassay kit (ELISA) (R&D Systems, Minneapolis, MN) with the lowest detectable limit of 0.08 ng/ml and with intra- and interassay CVs below 5.5 and 5.2%, respectively. Serum high-sensitivity IL-6 (hsIL-6) and sIL-6R concentrations were determined with ELISA kits (R&D Systems). The minimum detectable concentration was 0.039 pg/ml for hsIL-6 and 6.5 pg/ml for sIL-6R. Intra- and interassay CVs were, respectively, <7.8 and 9.6% for hsIL-6 and <8.6 and 9.6% for sIL-6R. Serum high-sensitive C-reactive protein (hsCRP) was measured by means of particle enhanced immunonephelometry using a CardioPhase kit (Dade Behring, Marburg, Germany) with a sensitivity of 0.175 mg/l and intra- and interassay CVs <4.5 and 5.8%, respectively.

Serum luteinizing hormone, follicle-stimulating hormone, testosterone, and estradiol were measured by the chemiluminescence method (ACS Chirone 180), and serum sex hormone–binding globulin (SHBG) was measured by the immunoradiometric assay (ZenTech, Angleur, Belgium). The sensitivity and CVs for all these assays were identical, as reported previously (16). Free androgen index was calculated as serum testosterone (nmol/l) × 100/SHBG (nmol/l) ratio (17).

The statistics were performed with STATISTICA 8.0 software. The variables, which did not have normal distribution (fasting and postload insulin, triglycerides, IL-6, sIL-6R, sgp130, SHBG, testosterone, free androgen index, estradiol, and hsCRP) were log-transformed before the analyses. For the purpose of the data presentation, these variables were transformed to absolute values in the results. The differences between the groups were estimated with factorial ANOVA, with PCOS status and obesity as categorical factors and the studied parameters as dependent variables. The relationships between variables were studied with the Pearson product-moment correlation analysis and with multiple regression analysis. The level of significance was accepted at P < 0.05.

The clinical characteristics of the studied groups is shown in Table 1. In the factorial ANOVA, we observed that the PCOS women had higher serum sgp130 concentrations than the control group (P = 0.0002). We also found that the obese women had higher serum sgp130 concentrations than the lean women (P < 0.0001). The interaction between PCOS status and obesity was not significant (P = 0.29), which suggests that an increase in sgp130 in PCOS was independent of obesity and an increase in sgp130 in obesity was independent of PCOS status. Additionally, PCOS women had significantly lower serum sIL-6R (P = 0.0036). In contrast, serum sIL-6R did not differ between lean and obese women (P = 0.93). Serum IL-6 and hsCRP did not differ between PCOS and control subjects, whereas they were increased in the obese group compared with the lean group (P = 0.007 and P < 0.001, respectively).

Both the presence of PCOS and obesity were independently characterized by lower insulin sensitivity (P = 0.0021 and P = 0.0033, respectively; P = 0.62 for the interaction), lower serum SHBG (P = 0.011 and P < 0.0001, respectively; P = 0.22 for the interaction), and higher free androgen index (both P < 0.0001, P = 0.34 for the interaction). PCOS women also had higher serum luteinizing hormone and testosterone concentrations (both P < 0.0001) (Table 2).

Cross-sectionally, serum sgp130 was related to BMI (r = 0.35, P < 0.001), waist (r = 0.36, P < 0.001), body fat (r = 0.28, P = 0.005), postload glucose and insulin (r = 0.29, P = 0.002, and r = 0.29, P = 0.002, respectively), triglycerides (r = 0.20, P = 0.032), hsCRP (r = 0.24, P = 0.012), and free androgen index (r = 0.28, P = 0.002). We also observed the significant inverse correlation of sgp130 with insulin sensitivity (r = −0.36, P < 0.0001) (Fig. 1) and SHBG (r = −0.23, P = 0.015). In the subgroup analysis, the correlation between serum sgp130 and insulin sensitivity was present in the PCOS subjects (r = −0.34, P = 0.002) but not in the control group (r = −0.16, P = 0.35).

Using M value normalized for body weight (M/BW) instead of M/FFM increased the effect of obesity (Table 1) and the correlation with sgp130 was slightly stronger (r = −0.40, P < 0.001).

Serum sIL-6R was related to follicle-stimulating hormone in the entire study group (r = 0.23, P = 0.017) and in the subgroup of PCOS women (r = 0.24, P = 0.041).

In multiple regression analysis, we observed that the relationship between sgp130 and insulin sensitivity was independent of BMI, waist, percent of body fat, postload glucose and insulin, triglycerides, and hsCRP (all adjusted β values between −0.23 and −0.33, all P < 0.05), whereas correlations of sgp130 with SHBG and free androgen index disappeared after controlling for insulin sensitivity.

Exclusion of subjects with the highest values of postload insulin did not change any of the results.

The main finding of the present study is an increased serum sgp130 concentration in obesity and PCOS and an inverse relationship between sgp130 and insulin sensitivity in PCOS women.

So far, serum sgp130 has not been determined in obesity and PCOS. Escobar-Morreale et al. (18) did not find the difference in circulating sgp130 between control and hyperandrogenic women; however, almost 40% of subjects in the hyperandrogenic group did not have PCOS. The frontier between PCOS and hyperandrogenism is diffuse; however, because of the large heterogeneity of PCOS itself, differences in characteristics of the studied groups are likely to influence the results. There are studies showing the association between polymorphisms in IL-6 or IL-6R and gp130 genes with clinical characteristics of PCOS and hyperandrogenism (18,19). Regarding the studied polymorphism of the gp130 gene, it did not affect the circulating level of this factor. In another study, polymorphism at the gp130 locus was associated with traits of the metabolic syndrome; however, circulating sgp130 was not reported (20).

Serum sgp130 was inversely related to insulin sensitivity in PCOS women. As mentioned, it was hypothesized that the gp130 cytokine family might exert beneficial effects and might be useful in the treatment of obesity and insulin resistance (4). Most gp130 cytokines occur in circulation in small, usually undetectable, concentrations. It is possible that they might exert auto- and paracrine effects in insulin-sensitive tissues, as in experimental studies, where the ciliary neurotrophic factor was demonstrated to regulate metabolic processes in adipocytes and skeletal muscle cells (4,10). Given the fact that circulating sgp130 acts as an inhibitor of intracellular gp130 signaling (11), the correlation observed in our study might reflect this inhibitory effect.

The question arises as to whether an association between sgp130 and insulin sensitivity is specific only for PCOS. We did not observe a significant correlation in the control group. Other populations with different clinical characteristics should probably be investigated. However, sgp130 was also increased in obese insulin-resistant women without PCOS.

Serum sgp130 was also related to SHBG and free androgen index; however, these associations were no longer significant after controlling for insulin sensitivity. IL-6, sIL-6R, and sgp130 might also regulate other reproductive functions, such as granulosa cell survival, follicular growth, development of human preovulatory processes, and blastocyst implantation (21). We cannot exclude the possibility that sgp130 influences other features of PCOS phenotype, not assessed in the present study. Nevertheless, these are mainly local actions of sgp130. In our study, estimation of sgp130 serum concentration was performed during the 3–5 days of the menstrual cycle, and the influence of sgp130 on endometrium in the proliferative phase of the menstrual cycle seemed to be relatively weak (21).

There are studies that demonstrate an increase in circulating IL-6 and hsCRP in PCOS (22). However, our results on the lack of an increase in these parameters in PCOS are in agreement with other data (23,24).

To our knowledge, decrease in sIL-6R in PCOS was not reported previously. Polymorphisms in the IL-6R gene were associated with BMI and characteristics of the metabolic syndrome (18,25), but not with hyperandrogenism (18). The balance between sIL-6R and sgp130 in PCOS seems to be directed toward an inhibition of IL-6 action; however, the significance of this finding remains unclear.

In conclusion, our data indicate that serum sgp130 is inversely and independently associated with insulin sensitivity in women with PCOS. An increased serum sp130 in obesity and PCOS suggests an inhibition of intracellular gp130 signaling in insulin-resistant conditions.

This study was supported by Grants 3-50815L and 3-50816L from the Medical University of Bialystok.

No potential conflicts of interest relevant to this article were reported.