Oxidative Stress in Response to Saturated Fat Ingestion Is Linked to Insulin Resistance and Hyperandrogenism in Polycystic Ovary Syndrome

Abstract

Context

Oxidative stress and insulin resistance are often present in polycystic ovary syndrome (PCOS).

Objective

We determined the effect of saturated fat ingestion on leukocytic reactive oxygen species (ROS) generation, p47^phox expression, and circulating thiobarbituric acid–reactive substances (TBARS) in women with PCOS.

Design

Cross-sectional study.

Setting

Academic medical center.

Patients

Twenty women of reproductive age with PCOS (10 lean, 10 with obesity) and 19 ovulatory control subjects (10 lean, 9 with obesity).

Main Outcome Measures

ROS generation and p47^phox mRNA and protein content were quantified in leukocytes, and TBARS was measured in plasma from blood drawn while the subjects were fasting and 2, 3, and 5 hours after saturated fat ingestion. Insulin sensitivity was derived from an oral glucose tolerance test (IS_OGTT). Androgen secretion was assessed from blood drawn while the subjects were fasting and 24, 48, and 72 hours after human chorionic gonadotropin (HCG) administration.

Results

Regardless of weight class, women with PCOS exhibited lipid-induced increases in leukocytic ROS generation and p47^phox mRNA and protein content as well as plasma TBARS compared with lean control subjects. Both PCOS groups exhibited lower IS_OGTT and greater HCG-stimulated androgen secretion compared with control subjects. The ROS generation, p47^phox, and TBARS responses were negatively correlated with IS_OGTT and positively correlated with HCG-stimulated androgen secretion.

Conclusion

In PCOS, increases in ROS generation, p47^phox gene expression, and circulating TBARS in response to saturated fat ingestion are independent of obesity. Circulating mononuclear cells and excess adipose tissue are separate and distinct contributors to oxidative stress in this disorder.

We studied the effect of lipid on oxidative stress in PCOS. We found increased obesity-independent lipid-stimulated oxidative stress in PCOS that is linked to insulin resistance and hyperandrogenism.

Polycystic ovary syndrome (PCOS) is the most common female endocrinopathy, affecting as many as 15% to 18% of women of reproductive age (1, 2). The disorder is characterized by hyperandrogenism, ovarian dysfunction, and polycystic ovarian morphology (2). Insulin resistance and dyslipidemia are common features of PCOS, and their prevalence is as high as 70% (3, 4). Furthermore, women with PCOS often have concomitant obesity, which is strongly associated with insulin resistance and dyslipidemia (5, 6).

PCOS is a nutrient-triggered proinflammatory state that promotes insulin resistance (7‒9). The initial clue that inflammation is present in the disorder came from our report of elevated circulating levels of TNF-α in lean women with PCOS (10). As a known mediator of insulin resistance, TNF-α upregulates suppressor of cytokine-3 (SOCS-3), which in turn truncates insulin signaling, thereby attenuating facilitative glucose transport (11‒13). We have recently shown that in PCOS, saturated fat ingestion stimulates increases in circulating TNF-α and peripheral leukocytic SOCS-3 expression (14).

Oxidative stress in response to saturated fat ingestion is an intermediate step in stimulating TNF-α secretion from circulating leukocytes. Superoxide is a reactive oxygen species (ROS) produced when NADPH is oxidized by membrane-bound NADPH oxidase (15, 16). Saturated fat ingestion upregulates the gene expression of p47^phox, a key cytosolic component of NADPH oxidase (17–19). Circulating free fatty acids arising from ingested saturated fat are used by leukocytes in the β-oxidation cycle to generate energy (20). Some free fatty acids are diverted for conversion to glycerol-3-phosphate, the precursor of diacylglycerol (21, 22). Diacylglycerol activates protein kinase C, which in turn phosphorylates p47^phox to initiate its translocation to the cell membrane to form a functional enzyme complex (22‒24). Superoxide-induced oxidative stress activates nuclear factor κB (NF-κB), the cardinal signal of inflammation that upregulates TNF-α gene transcription (25, 26). Thus, lipid-induced ROS generation from leukocytes may serve as a pro-oxidant trigger of inflammation to induce insulin resistance in PCOS.

We evaluated the effect of saturated fat ingestion on ROS generation from leukocytes in women with PCOS. We also evaluated this effect on the mRNA and protein content of p47^phox, the key component of NADPH oxidase, and on plasma thiobarbituric acid–reactive substances (TBARS), a commonly used index of oxidative stress–related lipid peroxidation. We hypothesized that, in response to saturated fat ingestion, ROS generation, p47^phox mRNA and protein content, and TBARS are increased in women with PCOS compared with ovulatory control subjects of similar age and body mass index (BMI) and that these markers of inflammation are linked to adiposity, insulin sensitivity, levels of fasting lipids, and ovarian androgen secretion. Lean women with PCOS (representing the authentic syndrome) were evaluated separately from women with obesity and PCOS (representing the superimposed effects of obesity on this disorder).

Materials and Methods

Participants

Twenty women with PCOS (10 lean and 10 with obesity), 18 to 35 years of age, and 19 control subjects (10 lean and 9 with obesity), 21 to 40 years of age with a similar BMI, volunteered to participate in the study. Some subjects in the current study were involved in our previous work on lipopolysaccharide-mediated inflammation in PCOS (14). Lean subjects had a BMI between 18 and 25 kg/m². Obesity was defined as a BMI between 30 and 40 kg/m². The diagnosis of PCOS was based on the presence of oligo-amenorrhea and hyperandrogenemia after excluding nonclassic congenital adrenal hyperplasia, Cushing syndrome, hyperprolactinemia, and thyroid disease. All subjects with PCOS also exhibited polycystic ovaries on ultrasound. All control subjects had regular menses lasting 25 to 35 days and a luteal range serum progesterone level consistent with ovulation (>5 ng/mL). Serum androgen levels were normal in all control subjects, and none of these subjects had skin manifestations of androgen excess or polycystic ovaries on ultrasound.

None of the subjects had diabetes or inflammatory illnesses, although two women with obesity and PCOS had impaired glucose tolerance and metabolic syndrome based on World Health Organization and Adult Treatment Panel III criteria, respectively (27, 28). Thirteen women with PCOS (five lean, eight with obesity) and 10 control subjects (five lean, five with obesity) had a family history of type 2 diabetes. None of the subjects smoked tobacco, and none used medications that would affect carbohydrate metabolism or immune function for a minimum of 6 weeks before study participation. During the 6 months before study entry, all subjects were weight stable within 5 pounds and were either sedentary or lightly active, the latter of which was defined as physical exertion beyond routine lasting no more than 30 minutes no more than once or twice a week. The level of physical activity was similar among study groups. All subjects provided written informed consent in accordance with the Institutional Review Board guidelines for the protection of human subjects.

Study design

A cream challenge test (CCT) was performed on all study subjects between days 5 and 8 after the onset of menses. In four amenorrheic subjects with PCOS (two lean, two with obesity), menses was induced with a 5-day course of micronized progesterone. An oral glucose tolerance test (OGTT) was performed the next day. All subjects were required to fast overnight for ∼12 hours before undergoing both tests. A healthy diet consisting of 50% carbohydrate, 35% fat, and 15% protein was provided to all subjects for three consecutive days before the CCT and after completing the CCT on the day preceding the OGTT. Body composition was assessed on the same day as the CCT. All subjects underwent a human chorionic gonadotropin (HCG) stimulation test over 4 days beginning on the day of the OGTT.

CCT

All subjects ingested 100 mL of dairy cream (gourmet heavy whipping cream; Land O Lakes Inc., Arden Hills, MN) as adapted from Deopurkar et al. (29). The dairy cream preparation had a saturated fat content of 70% (28% unsaturated fat), a protein content of <2%, and a glucose content of 0%. Blood samples were drawn while subjects were fasting and 2, 3, and 5 hours after cream ingestion to quantify molecular markers of oxidative stress from leukocytes isolated as previously described (30). Plasma was isolated from these same blood samples and stored at –80°C until assayed for TBARS and fasting lipids.

OGTT

All subjects ingested a beverage containing 75 g of glucose. Blood samples were drawn while the subjects were fasting and at 30, 60, 90, 120, and 180 minutes after glucose ingestion. For each blood sample, plasma glucose was measured immediately and insulin was measured later from plasma stored at –80°C. Insulin sensitivity was derived from the OGTT (IS_OGTT) using the Matsuda index formula: 10,000 divided by the square root of the fasting glucose level multiplied by the fasting insulin level, and that result multiplied by the product of the mean glucose level and mean insulin level (31).

HCG stimulation test

A baseline blood sample was drawn at 8 am in all subjects after an overnight fast of ∼12 hours and was immediately followed by an IM injection of 5000 IU HCG (Pregnyl; Merck & Co., Whitehouse Station, NJ). Fasting blood samples were drawn at 24, 48, and 96 hours after the HCG injection. Serum was isolated from these samples and stored at –80°C until assayed for testosterone (T), androstenedione (A), and 17-hydroxyprogesterone (17-OHP). The trapezoidal rule was used to calculate the area under the curve (AUC) for androgens and 17-OHP (32).

Body composition assessment

Height without shoes was measured to the nearest 1.0 cm. Body weight was measured to the nearest 0.1 kg. All subjects also underwent dual-energy X-ray absorptiometry to assess the percentage of total body fat, percentage of truncal fat, and R1 central abdominal fat using the QDR 4500 Elite model scanner (Hologic Inc., Waltham, MA) as previously described (33, 34).

ROS generation assay

Respiratory burst activity of mononuclear cells (MNCs) and polymorphonuclear cells (PMNs) was measured by detection of superoxide radical via chemiluminescence as previously described (35). As previously validated, the variation of ROS generation by leukocytes in humans using this method is <8% over a 2-week period (36).

Real-time PCR

Total RNA was isolated from MNCs that were previously stabilized in RNAlater® (Sigma-Aldrich, St. Louis, MO). Real-time PCR was used to quantify the mRNA content of p47^phox as previously described (37). However, an ABI Prism 7300 Sequence Detection System (Applied Biosystems, Foster City, CA) was used for the current study. PRIMER EXPRESS software (PE Biosystems, Foster City, CA) was used to select the primer sequences for p47^phox (GenBank AF330627; forward primer: 5′-AGAGCGGTTGGTGGTTCTGT-3′; reverse primer: 5′-GGAAGGATGCTGGGATCCA-3′). The rRNA signal for the housekeeping gene ribosomal protein L13a was used to normalize against differences in RNA isolation and degradation and in reverse transcription and PCR efficiencies using the comparative cycle threshold method.

Western blotting

The protein content of p47^phox and actin from MNCs was quantified by Western blotting as previously described using a monoclonal antibody against p47^phox subunit (BD Transduction Laboratories, San Diego, CA) or actin (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) at a dilution of 1:500 (38). Densitometry was performed on the scanned films of Western blots using Molecular Imaging software version 5.0.2.30 (Carestream Health, Rochester, NY), and all values for p47^phox were corrected for loading using those obtained for actin.

Plasma and serum measurements

Plasma TBARS was measured by fluorescence (OXItek; ZeptoMetric Corporation, Buffalo, NY) (sensitivity, 1.0 nmol/mL; intra-assay coefficient of variation [CV], 6.5%; interassay CV, 5.5%). Plasma glucose was measured by the glucose oxidase method (YSI, Yellow Springs, OH). Plasma insulin was measured by radioimmunoassay (Millipore, Billerica, CA) (sensitivity, 2.72 µU/mL; intra-assay CV, 3.8%; interassay CV, 3.2%). Plasma total cholesterol, triglycerides, high-density lipoprotein cholesterol, and low-density lipoprotein (LDL) cholesterol were measured by enzymatic methods (SYNCHRON LX20 PRO automatic analyzer; Beckman Coulter, Inc., Fullerton, CA) (sensitivities, 25, 5, 3, and 8 mg/dL; intra-assay CV, 0.5% to 4.0%; interassay CV, 1.1% to 4.5%). Serum LH, T, A, and 17-OHP were measured by radioimmunoassay (Siemens, Los Angeles, CA) (sensitivities, 0.1 IU/L, 5 ng/dL, 0.1 ng/mL, and 0.1 ng/mL; intra-assay CV, 4.1% to 6.8%; interassay CV, 4.0% to 11.2%). The T assay demonstrates good correlation with commercial liquid chromatography–tandem mass spectrometry (39). Serum dehydroepiandrosterone-sulfate (DHEA-S) was measured by chemiluminescence on a UniCel^® DxI 800 Immunoassay System (Beckman Coulter, Inc., Chaska, MN) (sensitivity, 1 µg/dL; intra-assay CV, 4.7%; interassay CV, 6.1%). All samples from each subject were measured in duplicate in the same assay at the end of the study.

Statistics

The StatView software package (SAS Institute, Cary, NC) was used to perform the statistical analysis. All values were initially examined graphically for departure from normality. The presence or absence of normality was confirmed using the Shapiro-Wilk test. The natural logarithm transformation was applied to total cholesterol and LH before the analysis because these values were not normal. Treatment effects on markers of oxidative stress were determined by calculating percent change from baseline for each participant to account for intersubject variability. The trapezoidal rule was used to calculate the incremental AUC (iAUC) for each oxidative stress marker (29). Because prior work by our group suggests that obesity increases oxidative stress and reduces insulin sensitivity in PCOS (7‒9, 14, 30, 40, 41), ANOVA was used to compare data from this study across groups (lean PCOS vs lean control vs obese PCOS vs obese control) followed by post hoc analyses using the Tukey honestly significant difference test to identify the source of significance. Differences across groups in the response of oxidative stress markers over time during the CCT were analyzed using repeated measures ANOVA followed by post hoc analyses. Pearson product moment correlation coefficients were calculated initially for correlation analyses. Partial Pearson correlations with markers of oxidative stress were calculated that separately adjusted for each measure of adiposity due to collinearity. Data are presented as mean ± SE, and results with a two-tailed α level of 0.05 were considered to be significant.

Results

Age, body composition, and blood pressure

All four groups were similar in age, height, and systolic and diastolic blood pressures (Table 1). Subjects who were obese had significantly (P < 0.05) higher weight, BMI, percent total body fat, percent truncal fat, and R1 fat compared with subjects who were lean, regardless of whether they had PCOS. However, these measures of body composition were similar when women with PCOS were compared with control subjects of similar weight class.

Table 1.

Age, Body Composition, and Endocrine and Metabolic Parameters of Subjects

	Control Subjects		PCOS
	Lean	Obese	Lean	Obese
Age, y	30 ± 2	30 ± 2	28 ± 1	29 ± 2
Height, cm	165.3 ± 1.6	163.9 ± 2.6	163.3 ± 1.9	161.7 ± 3.3
Body weight, kg	61.0 ± 2.0	94.0 ± 4.0a,b	61.1 ± 1.7	89.2 ± 3.9c,d
BMI, kg/m2	23.0 ± 0.8	34.9 ± 0.7a,b	22.9 ± 0.4	34.0 ± 0.8c,d
Total body fat, %	29.5 ± 1.8	42.3 ± 0.9a,b	32.1 ± 1.9	43.8 ± 1.2c,d
Truncal fat, %	24.7 ± 2.4	41.9 ± 1.4a,b	28.1 ± 2.3	43.2 ± 1.4c,d
Central fat (R1), g	759 ± 87	2137 ± 135a,b	913 ± 93	2117 ± 114c,d
Systolic blood pressure, mm Hg	114 ± 4	127 ± 3	116 ± 5	122 ± 4
Diastolic blood pressure, mm Hg	72 ± 2	76 ± 3	70 ± 4	76 ± 2
TBARS, nmol/mL	2.3 ± 0.3	2.5 ± 0.3	2.5 ± 0.2	2.8 ± 0.2
Fasting glucose, mg/dL	89 ± 2	89 ± 2	86 ± 3	92 ± 2
2-Hour glucose, mg/dL	97 ± 5	86 ± 9	95 ± 8	134 ± 8c,d,f
Fasting insulin, µU/mL	4.1 ± 1.1	15.5 ± 3.0a,b	7.6 ± 1.3	19.3 ± 4.1c,d
ISOGTT	14.1 ± 2.2	5.4 ± 1.0a	7.7 ± 1.1e	3.0 ± 0.4c,d
Total cholesterol, mg/dL	143 ± 6	141 ± 6b	173 ± 6e	183 ± 9c,f
Triglycerides, mg/dL	58 ± 6	81 ± 11	62 ± 5	143 ± 22c,d,f
HDL cholesterol, mg/dL	53 ± 3	48 ± 3	55 ± 2	49 ± 4c,d
LDL cholesterol, mg/dL	78 ± 7	76 ± 6b	106 ± 5e	107 ± 7c,f
LH, mIU/mL	6.6 ± 0.5	4.7 ± 0.8b	14.4 ± 1.7e	13.4 ± 1.8c,f
Testosterone, ng/dL	32.3 ± 5.2	25.5 ± 4.3b	62.8 ± 4.5e	72.3 ± 6.8c,f
Androstenedione, ng/mL	1.7 ± 0.2	2.1 ± 0.2b	4.0 ± 0.3e	3.9 ± 0.2c,f
DHEA-S, µg/dL	211 ± 22	157 ± 22	256 ± 34	205 ± 27
Testosterone, AUC	3401 ± 453	3700 ± 201b	6468 ± 786e	7130 ± 1137c,f
Androstenedione, AUC	576 ± 51	329 ± 37b	506 ± 31e	576 ± 51c,f
17OH-progesterone, AUC	10,112 ± 1226	9848 ± 594b	21,387 ± 2859e	23,057 ± 2782c,f

Values are expressed as means ± SE. Conversion factors to SI units: testosterone, ×3.467 (nmol/L); androstenedione, ×3.492 (nmol/L); DHEA-S, ×0.002714 (µmol/L); glucose, ×0.0551 (mmol/L); insulin, ×7.175 (pmol/L).

Abbreviation: HDL, high-density lipoprotein.

^aObese control vs lean control, P < 0.006.

^bObese control vs lean PCOS, P < 0.05.

^cObese PCOS vs lean control, P < 0.002.

^dObese PCOS vs lean PCOS, P < 0.02.

^eLean PCOS vs lean control, P < 0.007.

^fObese control vs obese PCOS, P < 0.001.

Markers of oxidative stress in leukocytes and plasma

In response to saturated fat ingestion, the change from baseline in ROS generation from MNCs and PMN, p47^phox mRNA and protein content, and plasma TBARS decreased in lean control subjects and was significantly (P < 0.0001) different compared with the increase observed in lean women and women with obesity and PCOS and control subjects with obesity after 2 and 3 hours (Figs. 1–3). In all four groups, the maximum response was reached at 2 hours for ROS generation from MNCs and PMN and at 3 hours for p47^phox mRNA and protein content and plasma TBARS. Compared with control subjects with obesity, women with obesity and PCOS exhibited significantly (P < 0.02) greater ROS generation responses from MNCs and PMN after 2 hours and greater increases in p47^phox protein content after 2 and 3 hours. All five markers of oxidative stress returned to baseline in both lean groups and in control subjects with obesity after 5 hours. In contrast, women with obesity and PCOS exhibited significantly (P < 0.02) greater residual responses in all five markers of oxidative stress compared with the other three groups after 5 hours.

Figure 1.

Comparison of the four study groups of the change from baseline (%) in leukocytic ROS generation from (A) MNCs and (B) PMNs from blood samples collected while fasting and 2, 3, and 5 h after saturated fat ingestion. Lean women and women with obesity and PCOS and control subjects with obesity exhibited a greater ROS generation response from MNCs (*P < 0.0001) and PMNs (*P < 0.02) compared with lean control subjects at 2 and 3 h after saturated fat ingestion. Women with obesity and PCOS exhibited a greater ROS generation response from MNCs (^†P < 0.02) and PMNs (^†P < 0.03) compared with control subjects with obesity at 2 h after saturated fat ingestion and a greater residual ROS generation response from MNCs (^‡P < 0.005) and PMNs (^‡P < 0.02) compared with lean women with PCOS and control subjects who were lean or had obesity at 5 h after saturated fat ingestion.

Figure 3.

Comparison of the four study groups of the change from baseline (%) in plasma TBARS levels from blood samples collected while fasting and 2, 3, and 5 h after saturated fat ingestion. Lean women and women with obesity and PCOS and control subjects with obesity exhibited a greater plasma TBARS response (*P < 0.0001) compared with lean control subjects at 2 and 3 h after saturated fat ingestion. Women with obesity and PCOS exhibited a greater residual plasma TBARS response (^†P < 0.02) compared with lean women with PCOS and control subjects who were lean or had obesity at 5 h after saturated fat ingestion.

Figure 2.

Comparison of the four study groups of the change from baseline (%) in MNC-derived p47^phox (A) mRNA content and (B) protein content from blood samples collected while fasting and 2, 3, and 5 h after saturated fat ingestion. Representative Western blots show the change in quantity of p47^phox and actin in MNC homogenates in samples collected before and after saturated fat ingestion. The samples used to quantify proteins by densitometry were run on the same gel. Lean women and women with obesity and PCOS and control subjects with obesity exhibited greater responses (*P < 0.0001) in p47^phox protein content (*P < 0.0001) compared with lean control subjects at 2 and 3 h after saturated fat ingestion. Women with obesity and PCOS exhibited a greater p47^phox mRNA content response (^†P < 0.02) compared with control subjects with obesity at 2 and 3 h after saturated fat ingestion and a greater residual response in p47^phox mRNA content (^†P < 0.0001) and protein content (^‡P < 0.0001) compared with lean women with PCOS and control subjects who were lean or had obesity at 5 h after saturated fat ingestion.

The iAUC for ROS generation from MNCs and PMN, p47^phox mRNA, and protein content and plasma TBARS decreased in lean control subjects and was significantly (P < 0.0001) different compared with the increase observed in lean women and women with obesity and PCOS and in control subjects with obesity (Fig. 4). Women with obesity and PCOS exhibited a significantly (P < 0.05) greater iAUC for ROS generation from MNCs and PMNs and for p47^phox mRNA content compared with lean women with PCOS and control subjects with obesity and for p47^phox protein content compared with control subjects with obesity.

Figure 4.

Comparison of the four study groups of the iAUC in response to saturated fat ingestion for leukocytic ROS generation from (A) MNCs and (B) PMNs, MNC-derived p47^phox (C) mRNA content and (D) protein content, and (E) plasma TBARS levels. Lean women and women with obesity and PCOS and control subjects with obesity exhibited a greater iAUC for ROS generation from MNCs (*P < 0.0001) and PMN (*P < 0.0001), MNC-derived p47^phox mRNA content (*P < 0.0001) and protein content (*P > 0.0001), and plasma TBARS (*P < 0.0001) compared with lean control subjects. Women with obesity and PCOS exhibited a greater iAUC for ROS generation from MNCs (^†P < 0.007) and PMN (^†P < 0.007) and MNC-derived p47^phox mRNA content (†P < 0.05) and protein content (†P < 0.0006) compared with control subjects with obesity. Women with obesity and PCOS also exhibited a greater iAUC for ROS generation from MNCs (^‡P < 0.05) and p47^phox protein content (^‡P < 0.03) compared with lean women with PCOS.

Insulin sensitivity and fasting lipids

IS_OGTT was significantly lower (P < 0.02) in subjects with obesity whether or not they had PCOS compared with lean control subjects, in women with obesity and PCOS compared with lean women with PCOS, and in lean women with PCOS compared with lean control subjects (Table 1). Plasma cholesterol and LDL were significantly (P < 0.05) higher in women with PCOS compared with control subjects regardless of body composition status. Plasma triglycerides were significantly (P < 0.05) higher in women with obesity and PCOS compared with those of the other three groups. Plasma high-density lipoprotein was similar in all four groups. Nevertheless, the only clinically elevated mean plasma lipid was the LDL level (>100 mg/dL) in women with obesity and PCOS.

Basal hormone levels and HCG-stimulated androgen and 17-OHP responses

Serum levels of LH, T, and A were significantly (P < 0.05) higher in women with PCOS compared with control subjects regardless of body composition status. However, DHEA-S levels were similar in all four groups (Table 1).

The AUCs for the HCG-stimulated responses of T, A, and 17-OHP were significantly (P < 0.05) higher in women with PCOS compared with control subjects regardless of weight class.

Correlations

Adiposity vs oxidative stress

BMI, percent total body fat, percent truncal fat, and R1 fat were positively correlated with the iAUC for ROS generation from MNCs and PMNs, p47^phox mRNA, and protein content and plasma TBARS for the combined groups and separately in control subjects (Table 2). In women with PCOS, BMI and R1 fat were positively correlated with the iAUC for p47^phox protein content (r = 0.46, P < 0.05; r = 0.52, P < 0.03), and percent total body fat was positively correlated with the iAUC for p47^phox mRNA content (r = 0.46, P < 0.05).

Table 2.

Pearson Correlations of Oxidative Stress Markers iAUC During the Cream Challenge Test With Measures of Adiposity and Insulin Sensitivity

	Significance	ROS MNC iAUC	ROS PMN iAUC	p47phox mRNA Content iAUC	p47phox Protein Content iAUC	Plasma TBARS iAUC
Combined groups
BMI, kg/m2	r	0.482	0.413	0.506	0.464	0.447
	P	0.002a	0.009a	0.002a	0.003a	0.005a
Total body fat, %	r	0.496	0.412	0.543	0.464	0.492
	P	0.002a	0.012a	0.0005a	0.003a	0.002a
Truncal fat, %	r	0.505	0.347	0.500	0.461	0.498
	P	0.002a	0.043a	0.002a	0.004a	0.002a
Central fat (R1), g	r	0.503	0.364	0.462	0.504	0.459
	P	0.002a	0.032a	0.004a	0.002a	0.004a
ISOGTT	r	−0.649	−0.481	−0.549	−0.661	−0.578
	P	0.0001a	0.002a	0.0004a	0.0001a	0.0002a
Control subjects
BMI, kg/m2	r	0.801	0.750	0.774	0.846	0.721
	P	0.0001a	0.0002a	0.0001a	0.0001a	0.0005a
Total body fat, %	r	0.657	0.574	0.678	0.816	0.668
	P	0.002a	0.010a	0.001a	0.0001a	0.002a
Truncal fat, %	r	0.701	0.545	0.647	0.798	0.670
	P	0.0008a	0.016a	0.003a	0.0001a	0.002a
Central fat (R1), g	r	0.762	0.650	0.687	0.869	0.711
	P	0.0001a	0.003a	0.001a	0.0001a	0.0006a
ISOGTT	r	−0.674	−0.534	−0.531	−0.670	−0.600
	P	0.002a	0.019a	0.019a	0.002a	0.007a

Insulin sensitivity index derived from an oral glucose tolerance test.

^a P < 0.05.

Insulin sensitivity vs oxidative stress

IS_OGTT was negatively correlated with BMI (r = −0.58, P < 0.0001), percent total body fat (r = 0.60, P < 0.0001), percent truncal fat (r = −0.64, P < 0.0001), and R1 fat (r = −0.62, P < 0.0001) for the combined groups. IS_OGTT was also negatively correlated with the iAUC for ROS generation from MNCs and PMNs, p47^phox mRNA and protein content, and plasma TBARS for the combined groups and in control subjects (Table 2) and with the iAUC for p47^phox protein content (r = 0.47, P < 0.04) in women with PCOS.

Fasting lipids vs oxidative stress

Plasma total cholesterol and triglycerides were positively correlated with the iAUC for ROS generation from MNCs and PMNs, p47^phox mRNA and protein content, and plasma TBARS for the combined groups (Table 3). Plasma triglycerides were positively correlated with the iAUC for ROS generation from MNCs (r = 0.48, P < 0.04) and PMNs (r = 0.53, P < 0.02) and p47^phox protein content (r = 0.49, P < 0.03) in women with PCOS and with the iAUC for ROS generation from MNCs (r = 0.48, P < 0.05) and PMNs (r = 0.48, P < 0.04) and plasma TBARS (r = 0.49, P < 0.05) in control subjects.

Table 3.

Pearson Correlations of Oxidative Stress Markers iAUC During the Cream Challenge Test With Circulating Lipids For the Combine Groups

	Significance	ROS MNC iAUC	ROS PMN iAUC	p47phox mRNA Content iAUC	p47phox Protein Content iAUC	Plasma TBARS iAUC
Total cholesterol, mg/dL	r	0.525	0.440	0.393	0.477	0.393
	P	0.0006a	0.005a	0.013a	0.002a	0.013a
Triglycerides, mg/dL	r	0.458	0.498	0.354	0.478	0.450
	P	0.003a	0.001a	0.027a	0.002a	0.004a
HDL cholesterol, mg/dL	r	−0.075	−0.118	−0.050	−0.055	−0.018
	P	0.648	0.474	0.762	0.741	0.916
LDL cholesterol, mg/dL	r	0.435	0.326	0.285	0.311	0.271
	P	0.006a	0.043a	0.079	0.054	0.096

Abbreviation: HDL, high-density lipoprotein.

^a P < 0.05.

LH and androgens vs oxidative stress

For the combined groups, there were positive correlations between basal LH and ROS generation from PMNs, p47^phox mRNA and protein content and plasma TBARS, basal and HCG-stimulated AUC for testosterone and androstenedione and ROS generation from MNCs and PMNs, p47^phox mRNA and protein content and plasma TBARS, along with HCG-stimulated 17-OHP AUC and ROS generation from MNCs and PMNs and p47^phox mRNA and protein content (Table 4). In women with PCOS, there were positive correlations between p47^phox mRNA content and basal testosterone (r = 0.59, P < 0.007) and HCG-stimulated AUC for testosterone (r = 0.71, P < 0.0006), HCG-stimulated AUC for androstenedione and ROS generation from MNCs (r = 0.47, P < 0.04) and PMN (r = 0.44, P < 0.05) as well as DHEA-S and ROS generation from MNCs (r = 0.50, P < 0.03). In control subjects, basal androstenedione was positively correlated with ROS generation from MNCs (r = 0.51, P < 0.03).

Table 4.

Pearson Correlations of Oxidative Stress Markers iAUC During the Cream Challenge Test With Circulating LH and Androgens for the Combined Groups

	Significance	ROS MNC iAUC	ROS PMN iAUC	p47phox mRNA Content iAUC	p47phox Protein Content iAUC	Plasma TBARS iAUC
LH, iU/mL	r	0.287	0.327	0.341	0.367	0.442
	P	0.076	0.049a	0.043a	0.032a	0.005a
Testosterone, ng/dL	r	0.422	0.488	0.580	0.494	0.365
	P	0.008a	0.002a	0.0002a	0.002a	0.034a
Androstenedione, ng/mL	r	0.680	0.651	0.606	0.652	0.436
	P	0.0001a	0.0001a	0.0001a	0.0001a	0.007a
DHEA-S, μg/dL	r	0.026	0.244	0.095	0.036	0.040
	P	0.874	0.135	0.566	0.830	0.807
Testosterone AUC	r	0.436	0.512	0.642	0.525	0.357
	P	0.006a	0.001a	0.0001a	0.0007a	0.034a
Androstenedione AUC	r	0.565	0.570	0.509	0.633	0.382
	P	0.0003a	0.0003a	0.001a	0.0001a	0.023a
17OH-progesterone AUC	r	0.489	0.397	0.310	0.573	0.238
	P	0.002a	0.023a	0.048a	0.0002a	0.145

^a P < 0.05.

The relationships of markers of oxidative stress with insulin sensitivity, fasting lipids, and hormones were maintained after adjusting for adiposity (data not shown).

Discussion

Our data clearly show that, in PCOS, oxidative stress originating from leukocytes in response to saturated fat ingestion is increased independent of obesity. Lean women with PCOS exhibit lipid-induced increases in ROS generation, p47^phox gene expression, and plasma TBARS compared with lean control subjects. Furthermore, women with obesity and PCOS exhibit lipid-induced increases in ROS generation and p47^phox gene expression compared with control subjects with obesity. ROS generation, p47^phox gene expression, and plasma TBARS in response to saturated fat ingestion are negatively associated with insulin sensitivity and positively associated with basal and HCG-stimulated androgen secretion, further bolstering the idea that lipid-stimulated oxidative stress may be a key driver of insulin resistance and hyperandrogenism in PCOS. The positive association between molecular and circulating markers of oxidative stress and measures of adiposity also suggests that, in PCOS, excess adipose tissue is a contributor to the pro-oxidant burden and an additional regulator of insulin action.

Suppression of leukocytic oxidative stress after saturated fat ingestion may be the normal in vivo response in lean, healthy women of reproductive age. The responses of ROS generation from MNCs and PMNs, p47^phox mRNA, and protein content and plasma TBARS are suppressed in lean control subjects. We have previously shown that glucose ingestion elicits a similar response in normal women of reproductive age for leukocytic markers of oxidative stress and inflammation, including that of leukocytic TNF-α secretion (9, 40–43). This is key because macrophages derived from MNCs present in excess adipose tissue and muscle exert a paracrine effect on insulin action (44, 45). Moreover, ROS-induced oxidative stress from MNC-derived macrophages triggers NF-κB activation and subsequent TNF-α secretion, which can decrease insulin signaling to impair glucose uptake (11). Conversely, ablation of MNC-derived macrophages in muscle of insulin-resistant animals improves insulin sensitivity (46). Thus, suppression of NADPH oxidase activity to limit ROS generation after saturated fat ingestion may serve as a protective effect to maximize insulin signaling for glucose disposal in lean, healthy women of reproductive age.

Leukocytes of women with PCOS exhibit a pro-oxidant response to saturated fat ingestion. In the current study, leukocytic ROS generation and p47^phox protein content increased after saturated fat ingestion in lean women and women with obesity and PCOS compared with control subjects of their respective weight class, as did p47^phox mRNA content and plasma TBARS in lean women with PCOS compared with lean control subjects. These results corroborate our previous reports of leukocyte preactivation in the fasting state and glucose-stimulated leukocytic ROS generation in women with PCOS (40–43, 47). Most notably, they highlight the discrete role of circulating leukocytes in establishing a pro-oxidant state that may be the underpinning of lipid-induced increases in circulating TNF-α and leukocytic SOCS-3 expression in women with PCOS (14). Protein ingestion also provokes a pro-oxidant, proinflammatory response in MNCs (48). Thus, dietary intake alone can stimulate oxidative stress in PCOS and culminates in an acute inflammatory response that promotes insulin resistance in the absence of excess adiposity. Further corroboration is provided by the reductions in oxidative stress and inflammation observed in normal individuals after a 2-day fast, along with the inverse relationship between insulin sensitivity and markers of oxidative stress in the current study (49).

In PCOS, there is a relationship between oxidative stress induced by saturated fat ingestion and adiposity. The iAUC for all five lipid-stimulated markers of oxidative stress are positively associated with measures of adiposity for the combined groups. Lipid-induced p47^phox gene expression in particular is positively associated with BMI, total body fat, and abdominal fat in women with PCOS. Phagocytic activity of resident macrophages derived from migrant MNCs within the stromal-vascular compartment of excess adipose tissue in response to hypoxia-related adipocyte death increases ROS generation to promote oxidative stress (44, 50). Subsequent macrophage TNF-α production stimulates adipocyte TNF-α production through a paracrine effect that promotes insulin resistance (51). In fact, measures of adiposity including abdominal adiposity are negatively associated with insulin sensitivity. Thus, circulating leukocytes and excess adipose tissue jointly contribute to oxidative stress–related systemic inflammation and insulin resistance when obesity is present in PCOS.

In PCOS, there is also a relationship between oxidative stress induced by saturated fat ingestion and dyslipidemia. In the current study, total cholesterol and triglycerides are positively associated with the iAUC for ROS generation and p47^phox gene expression for the combined groups and in women with PCOS. Triglycerides are also positively associated with plasma TBARS for the combined groups. Hypercholesterolemia increases the cholesterol content within plasma membrane microdomains known as lipid rafts, which leads to increases in membrane-bound NADPH oxidase activity, and enhances inhibitory binding of a lipid raft scaffolding protein known as caveolin to endothelial nitric oxide synthase, thereby decreasing production of nitric oxide, the deactivator of ROS (52, 53). These key enzyme alterations lead to increased ROS generation (52, 53). ROS promotes production of TNF-α, which in turn stimulates hepatic fatty acid synthesis, adipose tissue lipolysis, and fatty acid transport to the liver, culminating in increased production and secretion of hepatic triglyceride and triglyceride-rich very-low-density lipoprotein (VLDL) into the bloodstream (54). The increase in VLDL in particular enhances the transfer of triglyceride from VLDL to LDL. Hydrolysis of LDL by hepatic lipase yields small, dense LDL, a highly atherogenic molecule that easily enters the vascular subendothelium, where it undergoes oxidation for optimal intake by foamy macrophages within atherosclerotic plaques (55). In the current study, cholesterol and LDL are higher in women with PCOS regardless of weight class, whereas triglycerides are higher in women with obesity and PCOS in particular. Thus, dyslipidemia may be a formidable instigator of lipid-induced oxidative stress in PCOS especially when obesity is present, with the abnormal lipid profile and oxidative stress working through a positive feedback loop that can accelerate atherogenesis at an early age.

In PCOS, oxidative stress induced by saturated fat ingestion may directly stimulate hyperandrogenism. Basal levels of LH and androgens along with the HCG-stimulated androgen secretion are positively associated with various lipid-stimulated markers of oxidative stress for the combined groups and in women with PCOS. This reaffirms similar results from our past reports (7–9, 30, 40, 41). Although the association with LH suggests a central impact of oxidative stress on androgen production, local effects are well described. Macrophages derived from MNCs increase in number within the ovary after saturated fat ingestion (56). Pro-oxidants upregulate the gene expression of the androgen-producing steroidogenic enzyme CYP17 in ovarian theca cells (57). In fact, ROS-induced TNF-α secretion from macrophages promotes serine phosphorylation, which may increase the activity of the 17,20-lyase arm of CYP17 (58), and stimulates theca cell proliferation (59). Furthermore, chronic suppression of lipid-induced ROS generation and p47^phox gene expression with salicylate therapy reduces basal and HCG-stimulated androgen secretion in lean insulin-sensitive women with PCOS (60). Thus, excess ovarian androgen production in women with PCOS may be the result of a local inflammatory response to lipid-induced oxidative stress from leukocytes trafficking into the polycystic ovary culminating in increased theca cell proliferation and steroidogenic activity.

In summary, women with PCOS exhibit increases in leukocytic ROS generation, p47^phox gene expression, and plasma TBARS in response to saturated fat ingestion that are independent of obesity. The resultant oxidative stress induces a proinflammatory state that can promote insulin resistance, dyslipidemia, and hyperandrogenism in PCOS. The association between lipid-induced markers of oxidative stress and measures of adiposity points to excess adipose tissue as an additional perpetuator of inflammation in this disorder. These findings showcase the separate and distinct contribution of circulating MNCs and excess adipose tissue in the evolution of metabolic and endocrine dysfunction in PCOS.

Glossary

Abbreviations:

17-OHP

17-hydroxyprogesterone

A

androstenedione

AUC

area under the curve

BMI

body mass index

CCT

cream challenge test

CV

coefficient of variation

DHEA-S

dehydroepiandrosterone-sulfate

HCG

human chorionic gonadotropin

iAUC

incremental area under the curve

IS_OGTT

insulin sensitivity derived from the oral glucose tolerance test

LDL

low-density lipoprotein

MNC

mononuclear cell

NF-κB

nuclear factor κB

OGTT

oral glucose tolerance test

PCOS

polycystic ovary syndrome

PMN

polymorphonuclear cell

ROS

reactive oxygen species

SOCS-3

suppressor of cytokine-3

T

testosterone

TBARS

thiobarbituric acid–reactive substances

VLDL

very-low-density lipoprotein