Objective
We evaluated mononuclear cell (MNC) preactivation in women with polycystic ovary syndrome (PCOS) by examining the effect of in vitro lipopolysaccharide (LPS) exposure on cytokine release in the fasting state.
Study Design
Twenty women with PCOS (10 lean, 10 obese) and 20 weight-matched controls (10 lean, 10 obese) volunteered for study participation. Tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6) release was measured from mononuclear cells isolated from fasting blood samples and cultured in the presence and absence of LPS. Plasma IL-6 was measured from the same fasting blood samples. Insulin sensitivity was derived from an oral glucose tolerance test using the Matsuda index, and truncal fat was measured by dual-energy x-ray absorptiometry.
Results
The percent change from baseline in TNF-α and IL-6 release from MNC following LPS exposure was increased ( P < .04) in lean and obese women with PCOS and obese controls compared with lean controls. Plasma IL-6 was increased ( P < .02) in obese women with PCOS compared with lean women with PCOS, which in turn was increased ( P < .02) compared with lean controls. The MNC-derived TNF-α and IL-6 responses from MNCs were negatively correlated with insulin sensitivity ( P < .03) and positively correlated with testosterone ( P < .03) and androstenedione ( P < .006) for the combined groups. Plasma IL-6 was positively correlated with percentage truncal fat ( P < .008).
Conclusion
In PCOS, increased cytokine release from MNCs following LPS exposure in the fasting state reveals the presence of MNC preactivation. Importantly, this phenomenon is independent of obesity and may contribute to the development of insulin resistance and hyperandrogenism in PCOS. In contrast, the source of plasma IL-6 elevations in PCOS may be excess adiposity.
The polycystic ovary syndrome (PCOS) affects as many as 15% of reproductive-aged women and is characterized by hyperandrogenism, chronic oligo- or anovulation, and polycystic ovaries. Obesity and insulin resistance are often present in PCOS and the compensatory hyperinsulinemia is thought to promote the hyperandrogenism. Many young women with PCOS also possess risk factors for cardiovascular disease such as metabolic syndrome, type 2 diabetes, dyslipidemia, and hypertension, which are associated with atherosclerosis.
In PCOS, oxidative stress and chronic low-grade inflammation have been implicated in the development of insulin resistance and accelerated atherogenesis. Women with PCOS exhibit increased circulating levels of protein carbonyls, tumor necrosis factor-α (TNF-α), and interleukin-6 (IL-6), decreased insulin signaling, and a higher prevalence of coronary artery calcification.
Our previous studies have highlighted the ability of nutrient ingestion to trigger a prooxidant proinflammatory response from peripheral blood mononuclear cells (MNCs) that is independent of obesity. Indeed, MNCs of lean women with PCOS exhibit increases in reactive oxygen species generation and nuclear factor-κB (NFκB) activation following the ingestion of glucose and lipid. NFκB regulates the transcription of a variety of inflammatory mediators including that of TNF-α and IL-6. These cytokines exert positive feedback to up-regulate the preceding molecular events that mediate insulin resistance and atherogenesis. Thus, the MNCs of women with PCOS have an increased sensitivity to nutrient ingestion.
MNC preactivation in the fasting state may account for increased MNC sensitivity in PCOS. In this instance, the NFκB inflammation pathway that culminates in cytokine release from MNCs may already be up-regulated prior to nutrient ingestion, similar to what has been reported in obesity. As such, we evaluated MNCs obtained in the fasting state for the amount of TNF-α and IL-6 release following in vitro exposure to lipopolysaccharide (LPS) and also evaluated the status of fasting plasma IL-6 in women with PCOS. We anticipated that plasma IL-6 would be elevated in these women based on previous studies. However, we hypothesized that MNC-derived cytokine release is increased in response to in vitro LPS exposure in women with PCOS compared with weight-matched controls and that this cytokine response to LPS along with plasma IL-6 levels are related to abdominal adiposity, insulin sensitivity, and circulating androgens.
Materials and Methods
Subjects
Twenty women with PCOS (10 lean and 10 obese) 20-34 years of age and 20 weight-matched control subjects (10 lean and 10 obese) 20-39 years of age volunteered for study participation. Some subjects in the current study were involved in our previous work on PCOS and insulin resistance. Additional subjects were recruited via newspaper advertisements and flyers within the Indiana University community between the years 2011 and 2013.
Obesity was defined as a body mass index (BMI) between 30 and 40 kg/m 2 . Lean subjects had a BMI between 18 and 25 kg/m 2 . The women with PCOS were diagnosed on the basis of oligoamenorrhea and hyperandrogenemia after excluding nonclassic congenital adrenal hyperplasia, Cushing’s syndrome, hyperprolactinemia, and thyroid disease. Polycystic ovaries were present on ultrasound in all subjects with PCOS. All control subjects had regular menses lasting 25-35 days and a luteal range serum progesterone level consistent with ovulation (>5 ng/mL). All control subjects exhibited normal circulating androgen levels and did not have any skin manifestations of androgen excess or polycystic ovaries on ultrasound.
Diabetes and inflammatory illnesses were excluded in all subjects. None of these suffered from depression, smoked tobacco, ingested more than 2 alcoholic beverages per month, or used medications that could have an impact on carbohydrate metabolism or immune function for a minimum of 6 weeks before beginning the study. All subjects were sedentary, defined as exercise less than once a month during the 6 months before the study participation. Written informed consent was obtained from all subjects according to institutional review board guidelines.
Study design
All study subjects were provided with a healthy diet consisting of 50% carbohydrate, 35% fat, and 15% protein for 3 consecutive days before testing between days 5 and 8 after the onset of menstruation. All but 1 of the subjects was compliant with the diet based on completion of a checklist of consumed food and inspection of returned empty food containers. The less compliant individual was a lean control who consumed 85% of the food provided. Testing began with an assessment of body composition. After an overnight fast of approximately 12 hours, a blood sample was obtained for MNC isolation and culture and to isolate plasma that was stored at –80°C until assayed for IL-6. All study subjects then underwent an oral glucose tolerance test (OGTT) to assess insulin sensitivity.
Body composition assessment
Height without shoes was measured to the nearest 1.0 cm. Body weight was measured to the nearest 0.1 kg. Waist circumference was measured at the level of the umbilicus and used to estimate abdominal adiposity. In addition, all subjects underwent dual-energy x-ray absorptiometry to determine the percentage total body fat and percentage truncal fat using the QDR 4500 Elite model scanner (Hologic Inc, Waltham, MA) as previously described.
MNC culture
MNCs were isolated by density gradient centrifugation in Histopaque-1077 (Sigma-Aldrich Co, St. Louis, MO). The cells were washed twice in pyrogen-free saline, resuspended in RPMI 1640 (0.3 mg/mL L-glutamine, 100 U/mL penicillin, 100 μg/mL streptomycin) with TCH serum replacement (MD Biomedicals, Inc, Irvine, CA) and seeded in coated culture plates (VWR International, West Chester, PA) at a concentration of 2.5 × 10 6 cells/mL. The cells were then incubated (95% humidity, 5% CO 2 , 37° C) for 24 hours in the presence (exposed) and absence (unexposed) of 1 ng/mL of LPS endotoxin from Escherichia coli 0127:B8 (Sigma-Aldrich Co). The culture supernatants were collected and stored at –80°C until assayed for TNF-α and IL-6.
Oral glucose tolerance test
Blood samples were drawn at 0, 30, 60, 90, 120, and 180 minutes after ingestion of a 75 g glucose beverage to measure glucose and insulin. Plasma glucose concentrations were assayed immediately, and insulin measurements were performed later from plasma stored at –80°C. Insulin sensitivity was derived using the Matsuda index formula (IS OGTT ): 10,000 divided by the square root of (fasting glucose × fasting insulin) × (mean glucose × mean insulin). This particular formula is highly correlated with insulin sensitivity measurements using the gold standard hyperinsulinemic-euglycemic clamp procedure.
Serum, plasma and culture supernatant measurements
Serum luteinizing hormone (LH), testosterone, androstenedione, and dehydroepiandrosterone-sulfate (DHEA-S) levels were measured by a radioimmunoassay (Siemens Medical Solutions Diagnostics, Los Angeles, CA). Plasma glucose was measured by the glucose oxidase method (YSI 2300 STAT Plus, Yellow Springs, OH), whereas plasma insulin was measured by a double-antibody radioimmunoassay (Millipore, St. Charles, MO). Plasma IL-6 concentrations along with those of TNF-α and IL-6 in MNC culture supernatants were measured by an enzyme-linked immunosorbent assay (eBioscience, San Diego, CA). The interassay and intraassay coefficients of variation for all assays did not exceed 7.4% and 12%, respectively.
Statistics
Data were analyzed using StatView (SAS Institute, Cary, NC). All values were initially examined graphically for departure from normality, and the natural logarithm transformation was applied as needed. Descriptive data and change from baseline of variables were compared between groups using an analysis of variance for multiple group comparisons. The source of significance was subsequently identified by Tukey’s post hoc test. Furthermore, analysis of covariance was performed to confirm significance of inflammation variables using significantly altered metabolic parameters as covariates that might influence inflammation when comparing lean PCOS vs lean controls or obese PCOS vs obese controls.
The treatment effect was determined by calculating the percentage change in LPS-exposed cytokine release from the unexposed baseline for each subject in view of intersubject variability. Pearson linear regression was used for correlation analyses using the method of least squares. Data are presented as mean ± SE, and results with a 2-tailed α-level of 0.05 were considered significant.
Results
Age, body composition, blood pressure, and lipids
All 4 groups were similar in age and height ( Table 1 ). Obese subjects had significantly ( P < .05) higher weight, BMI, percentage total body fat, percentage truncal fat, and waist circumference compared with lean subjects whether or not they had PCOS. However, these measures of body composition were similar when women with PCOS were compared with weight-matched controls.
| Variable | Controls | PCOS | ||
|---|---|---|---|---|
| Lean | Obese | Lean | Obese | |
| Age, y | 30 ± 2 | 30 ± 2.0 | 27 ± 2 | 26 ± 2 |
| Height, cm | 165.9 ± 1.3 | 164.0 ± 2.1 | 162.4 ± 3.4 | 165.5 ± 1.7 |
| Body weight, kg | 62.8 ± 2.0 | 93.4 ± 3.4 a,b | 61.1 ± 1.9 | 96.6 ± 3.5 c,d |
| Body mass index, kg/m 2 | 22.8 ± 0.5 | 34.7 ± 0.9 a,b | 23.3 ± 0.7 | 35.2 ± 1.0 c,d |
| Total body fat, % | 32.9 ± 1.7 | 42.3 ± 0.8 a,b | 30.3 ± 1.4 | 44.5 ± 1.2 c,d |
| Truncal fat, % | 30.2 ± 1.7 | 41.7 ± 0.8 a,b | 29.3 ± 2.2 | 46.1 ± 1.1 c,d |
| Waist circumference, cm | 74.2 ± 1.4 | 100.8 ± 2.9 a,b | 77.1 ± 2.3 | 99.3 ± 5.1 c,d |
| Systolic blood pressure, mm Hg | 109 ± 2.0 | 118 ± 4.0 | 112 ± 2.0 | 118 ± 4.0 |
| Diastolic blood pressure, mm Hg | 64 ± 1.0 | 75 ± 3 a,b | 65 ± 3.0 | 72 ± 4.0 |
| Total cholesterol, mg/dL | 177 ± 12 | 188 ± 20 | 177 ± 12 | 178 ± 12 |
| Triglycerides, mg/dL | 63 ± 18 | 113 ± 35 | 122 ± 34 | 94 ± 14 |
| HDL cholesterol, mg/dL | 54 ± 5.0 | 48 ± 4.0 | 52 ± 5.0 | 47 ± 4.0 |
| LDL cholesterol, mg/dL | 115 ± 13 | 115 ± 18 | 108 ± 10 | 118 ± 11 |
| LH, mIU/mL | 4.9 ± 0.4 | 2.8 ± 0.4 b,e | 10.7 ± 1.2 f | 8.3 ± 1.1 c,d |
| Testosterone, ng/dL | 44.5 ± 3.3 | 31.4 ± 3.9 b,e | 66.1 ± 6.5 f | 72.6 ± 4.7 c |
| Androstendione, ng/mL | 1.4 ± 0.1 | 1.8 ± 0.1 b,e | 3.3 ± 0.3 f | 3.5 ± 0.2 c |
| DHEA-S, μg/dL | 118 ± 13 | 161 ± 26 b,e | 343 ± 41 f | 291 ± 46 c |
| Fasting glucose, mg/dL | 87 ± 2.0 | 87 ± 4.0 | 84 ± 2.0 | 88 ± 2.0 |
| 2 hour glucose, mg/dL | 113 ± 6.0 | 117 ± 5.0 | 106 ± 8.0 | 110 ± 6.0 |
| Fasting insulin, μIU/mL | 5.4 ± 0.9 | 13.9 ± 1.7 a,e | 11.4 ± 1.3 | 19.7 ± 3.1 c,d |
| IS OGTT | 9.4 ± 1.0 | 4.6 ± 0.9 a | 4.3 ± 0.4 f | 3.0 ± 0.5 c |
a Obese control vs lean control, P < .03
b Obese control vs lean PCOS, P < .0007
c Obese PCOS vs lean control, P < .001
d Obese PCOS vs lean PCOS, P < .05
e Obese control vs obese PCOS, P < .05
Systolic blood pressure was similar among groups. Diastolic blood pressure was significantly ( P < .05) higher in obese controls compared with lean women with PCOS and lean controls, but mean values were in the normotensive range. The levels of total cholesterol, triglycerides, and high- and low-density lipoprotein cholesterol were similar among groups.
Plasma hormone levels, glycemic status, and insulin sensitivity
Women with PCOS exhibited significantly ( P < .05) higher serum levels of LH, testosterone, androstenedione, and DHEA-S compared with control subjects, regardless of weight class ( Table 1 ). Women with PCOS who were lean had significantly ( P < .05) higher serum LH levels compared with those who were obese.
Women with PCOS and controls exhibited similar glucose levels while fasting and 2 hours after glucose ingestion, regardless of weight status. All subjects had a normal glucose response during the OGTT, with fasting glucose levels less than 100 mg/dL and 2-hour glucose levels ranging between 75 and 137 mg/dL. Fasting insulin levels were significantly higher ( P < .05) in obese women with PCOS compared with lean women with PCOS and both control groups and in obese controls compared with lean controls. The IS OGTT was significantly higher ( P < .05) in obese subjects, regardless of PCOS status compared with lean controls and in lean women with PCOS compared with lean controls.
MNC-derived TNF-α and IL-6 release and plasma IL-6
Baseline TNF-α and IL-6 release from MNCs in the fasting state was similar in all 4 groups (data not shown). The change from baseline in TNF-α and IL-6 release from MNCs following LPS exposure was significantly ( P < .04) greater in both PCOS groups and obese controls compared with lean controls ( Figure 1 , A and B).
Fasting plasma IL-6 was significantly ( P < .02) higher in obese subjects whether or not they had PCOS and lean women with PCOS compared with lean controls ( Figure 2 ). Obese women with PCOS also exhibited significantly ( P < .02) higher fasting plasma IL-6 levels compared with lean women with PCOS.
The MNC-derived TNF-α and IL-6 responses and fasting plasma IL-6 remained significantly increased in lean women with PCOS when comparing the lean groups after controlling for insulin sensitivity (IS OGTT ) and remained similar when comparing the obese groups after controlling for fasting insulin.
Correlations
Waist circumference was positively correlated with BMI (r = 0.76, P < .0001), percentage body fat (r = 0.61, P < .0001), and percentage truncal fat (r = 0.68, P < .0001), and IS OGTT was negatively correlated with BMI (r = –0.49, P < .002), waist circumference (r = –0.45, P < .005), and percentage truncal fat (r = –0.42, P < .008) for the combined groups.
The change from baseline in IL-6 release from MNC following LPS exposure was positively correlated with that of TNF-α and plasma IL-6 for the combined groups ( Table 2 ). Fasting plasma IL-6 was positively correlated with BMI, waist circumference, percentage body fat and percentage truncal fat. The MNC-derived TNF-α and IL-6 responses and plasma IL-6 were negatively correlated with IS OGTT and were positively correlated with serum levels of testosterone and androstenedione. There was also a positive correlation between MNC-derived TNF-α response and serum LH (r = 0.50, P < .002).
