ELEVATED CIRCULATING LEVELS OF MACROPHAGE MIGRATION INHIBITORY FACTOR IN POLYCYSTIC OVARY SYNDROME

Abstract

Women with Polycystic Ovary Syndrome (PCOS) have chronic low level inflammation which can increase the risk of atherogenesis. We evaluated the status of circulating macrophage migration inhibitory factor (MIF), a proinflammatory cytokine involved in atherogenesis, in women with PCOS and weight-matched controls. Two-way analysis of variance models adjusted for age were fit to evaluate the effect of PCOS status (PCOS vs. controls) and weight-class (obese vs. lean) on MIF and other parameters. MIF levels were significantly (p<0.001) higher in women with PCOS (lean: 37.7±10.6 ng/ml; obese: 54.6±15.2 ng/ml) compared to controls (lean: 4.8±0.6 ng/ml; obese: 17.5±8.0 ng/ml) regardless of weight class. CRP levels were significantly (p<0.001) higher in obese subjects (PCOS: 6.2±1.9 mg/l; controls: 6.7±1.4 mg/l) compared to lean subjects (PCOS: 0.9±0.4 mg/l; controls: 0.2±01 mg/l) after controlling for PCOS status. MIF levels directly correlated with % truncal fat (r=0.41, p<0.05), and plasma levels of CRP (r=0.42, p=0.05), LH (r=0.45, p=0.04), testosterone (r=0.53, p<0.008), androstendione (r=0.58, p<0.005). IS_OGTT inversely correlated with plasma levels of MIF (r= −0.51, p<0.02) and CRP (r= −0.73, p<0.001). Circulating MIF is elevated in PCOS independent of obesity, but both PCOS and obesity contribute to a proatherogenic state. In PCOS, abdominal adiposity and hyperandrogenism may exacerbate the risk of atherosclerosis.

1. INTRODUCTION

The Polycystic Ovary Syndrome (PCOS) is one of the most common female endocrinopathies affecting between 8–12% of reproductive age women [1]. The disorder is characterized by hyperandrogenism, chronic oligo- or anovulation and polycystic ovaries, with 2 out of these 3 findings required to diagnose PCOS [2]. As many as 70% of women with PCOS exhibit insulin resistance, with the compensatory hyperinsulinemia considered to be a promoter of hyperandrogenism in the disorder [3,4]. Insulin resistance is also associated with accelerated atherogenesis.⁵ Indeed, women with PCOS have a higher prevalence of coronary artery calcification, a radiographic marker of atherosclerosis [5]. Women with PCOS are often afflicted by obesity, which is another risk factor for developing atherosclerosis [6].

Macrophage migration inhibitory factor (MIF) is an evolutionary ancient and highly conserved cytokine, and is the first described member of the cytokine family [7,8]. MIF plays pleiotropic roles in acute and chronic inflammatory diseases such as septic shock, rheumatoid arthritis, and colitis [9,10]. Most recently, MIF has emerged as a key player in cardiovascular disease due to its ability to facilitate migration of mononuclear cells (MNC) into the vascular interstitium [11]. In addition, MIF plays a similar role to that of CRP by enhancing lipid uptake into MNC-derived foamy macrophages within atherosclerotic plaques [12,13]. Plasma MIF levels are elevated in obese individuals without PCOS [14]. In contrast, a recent study failed to demonstrate elevations in plasma MIF in women with PCOS [15]. The fact that control subjects in this study were not matched for weight with the women with PCOS may account for this finding.

In the present study, we determined the status of circulating MIF levels in women with PCOS compared to weight-matched controls. We also examined the relationship of MIF with body composition, insulin resistance, and circulating levels of CRP and androgens. We hypothesized that circulating MIF is increased in PCOS, and that MIF is related to measures of adiposity, insulin resistance, CRP and androgens.

2. MATERIALS AND METHODS

2.1. Subjects

Twelve women with PCOS (6 lean and 6 obese) and 12 weight-matched control subjects (6 lean and 6 obese) volunteered to participate in the study. The women with PCOS were diagnosed on the basis of oligo-amenorrhea 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 were ovulatory as evidenced by regular menses and a luteal phase serum progesterone level greater than 5 ng/ml. All control subjects exhibited normal circulating androgen levels and the absence of polycystic ovaries on ultrasound. All of the subjects provided written informed consent in accordance with Institutional Review Board guidelines for the protection of human subjects.

2.2. Study Design

All study subjects underwent an oral glucose tolerance test (OGTT) between days 5 and 8 following the onset of menses, and an overnight fast of ~12 hours. The women were provided with a healthy diet consisting of 50% carbohydrate, 35% fat and 15% protein for 3 consecutive days before the test. All subjects also underwent body composition assessment on the same day the OGTT was performed.

2.3. Oral Glucose Tolerance Test (OGTT)

All subjects ingested a 75 gram glucose beverage. Blood samples were drawn while fasting and 2 hours after ingestion of the glucose beverage to measure glucose and insulin. Plasma glucose concentrations were assayed immediately from the blood samples collected while plasma to measure insulin was stored at −80°C and assayed later. Additional plasma was isolated from the fasting blood samples and stored at −80°C until assayed for MIF, CRP and lipids. Insulin sensitivity was derived by IS_OGTT using the following formula: 10,000 divided by the square root of (fasting glucose × fasting insulin) × (mean glucose × mean insulin) [16].

2.4. 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 [17]. In addition, all subjects underwent dual energy absorptiometry (DEXA) to determine % total body fat and % truncal fat using the QDR 4500 Elite model scanner (Hologic Inc., Waltham, MA) as previously described [18,19].

2.5. Plasma Measurements

Plasma glucose concentrations were measured by the glucose oxidase method (YSI, Yellow Springs, OH) while plasma insulin concentrations were measured by a double antibody RIA (Linco Research, St. Charles, MO). Luteinizing hormone (LH), testosterone, androstenedione and dehydroepiandrosterone-sulfate (DHEA-S) levels were measured by RIA (Diagnostic Products Corporation, Los Angeles, CA). Plasma MIF concentrations were measured by ELISA (R&D Systems, Minneapolis, MN). Plasma CRP concentrations were measured by a high sensitivity ELISA (Alpha Diagnostics International, San Antonio, TX). Levels of total cholesterol, triglyceride and high-density lipoprotein (HDL) cholesterol were measured by enzymatic methods (SYNCHRON LX20 PRO automatic analyzer, Beckman Coulter, Inc., Fullerton, CA). Low-density lipoprotein (LDL) cholesterol was calculated using Friedewald’s formula [20]. All samples from each subject were measured in duplicate in the same assay at the end of the study. The interassay and intraassay coefficients of variation for all assays were 7% and 12% respectively.

2.6. Statistics

All values were initially examined graphically for departures from normality, and the natural logarithm transformation was applied as needed. A subsequent two-way analysis of variance (ANOVA) model was fit to evaluate the effect of PCOS status (PCOS vs. control) and weight class (obese vs. lean) on each variable after adjusting for age since the mean age of women with PCOS was significantly lower compared to controls. The results of the main effect tests are reported because none of the study variables exhibited a statistically significant interaction effect between PCOS status and weight class. The Spearman rank correlation coefficient was used to estimate the correlation between parameters. All values are expressed as means ± SE. All calculated p-values are two-sided, and an α-level of 0.05 was used to determine statistical significance.

3. RESULTS

3.1. Body Composition, Blood Pressure and Lipids

As expected, body mass index, % total body fat, % truncal fat and waist circumference were significantly (p<0.04) greater in obese subjects compared to lean subjects (Table 1). Systolic blood pressure, diastolic blood pressure, and triglyceride levels were also significantly (p<0.04) higher in obese subjects compared to the lean subjects after controlling for the effect of PCOS status. There were no significant differences in these parameters between the women with PCOS and control subjects after controlling for the effect of weight class.

Table 1.

Age, body composition, blood pressure, endocrine and metabolic parameters of subjects.

	PCOS		CONTROL
	Lean (n=6)	Obese (n=6)	Lean (n=6)	Obese (n=6)
Age, yr †	27±2	25±2	32±2	32±3
Body mass index, kg/m2 ‡	22.9±1.1	35.6±0.9	22.1±1.0	35.8±1.1
Total body fat, % ‡	31.7±2.7	44.2±2.0	30.3±1.6	41.9±1.3
Truncal Fat, % ‡	29.5±3.6	45.6±1.8	27.8±2.4	41.7±1.0
Waist circumference, cm ‡	75.2±2.1	102.3±2.9	78.0±2.8	99.2±3.4
Systolic blood pressure, mmHg ‡	112±4	126±6	108±2	123±6
Diastolic blood pressure, mmHg ‡	73±4	78±5	62±4	78±4
Total cholesterol, mg/dl	164±11	161±11	174±8	196±30
Triglycerides, mg/dl ‡	53±4	132±31	56±8	122±53
HDL – cholesterol, mg/dl	50±8	44±5	53±5	50±4
LDL – cholesterol, mg/dl	106±9	94±8	114±8	113±23
LH, mIU/ml †,‡	14.2±1.8	8.1±1.0	3.2±0.5	2.8±0.5
Testosterone, ng/dl †	84.0±10.7	79.2±11.1	43.0±5.1	34.3±5.3
Androstendione, ng/ml †	3.8±0.4	3.4±0.2	1.8±0.2	1.7±0.2
DHEA-S, µg/dl	301±48	269±46	239±18	162±33
Fasting Glucose, mg/dl †	87.2±3.0	90.2±1.5	88.2±1.3	81.7±5.1
2 Hour Glucose, mg/dl	99.5±6.0	118.3±7.4	109.2±7.1	118.8±4.7
Fasting Insulin, µiU/ml ‡	9.3±1.5	18.8±2.9	7.2±1.3	12.9±2.8
ISOGTT†,‡	7.2±0.7	3.3±0.6	10.9±1.4	5.9±1.4

Values are expressed as means ± SE. Conversion factors to SI units: Testosterone ×3.467 (nmol/liter), Androstenedione ×3.492 (nmol/liter), DHEA-S ×0.002714 (umol/liter), Glucose ×0.0551 (mmol/liter), Insulin ×7.175 (pmol/liter). Age is a covariate in the two-way ANOVA models used for statistical analysis. The natural logarithm transformation was applied to total cholesterol, triglyceride and LH values before the analysis. There was no significant interaction effect between PCOS status and weight class, P ≥ 0.22.

^†Main effect of PCOS status (PCOS vs. controls) controlling for the effect of weight class, P < 0.05.

^‡Main effect of weight class (lean vs. obese) controlling for the effect of PCOS status, P < 0.04.

3.2. Plasma Hormone Levels

Circulating levels of LH, testosterone and androstenedione were significantly (p<0.05) elevated in women with PCOS compared to control subjects regardless of weight class (Table 1). Circulating DHEA-S levels were similar in women with PCOS compared to controls after controlling for weight status.

3.3. Glycemic Status and Insulin Resistance

Glucose levels while fasting were significantly (p<0.05) higher in women with PCOS compared to controls, but were similar 2 hours post glucose ingestion after controlling for the effect of weight class. Fasting insulin levels were significantly (p<0.04) higher in obese subjects compared to lean subjects, regardless of PCOS status. IS_OGTT was significantly (p<0.04) lower in obese individuals compared to lean individuals, and significantly (p<0.05) lower in women with PCOS compared to the control subjects.

3.4. Plasma MIF and CRP Levels

Plasma MIF concentrations were significantly (p<0.001) higher in women with PCOS compared to control subjects regardless of weight class (Figure 1A). Plasma CRP concentrations were significantly (p<0.001) higher in obese individuals compared to lean individuals after controlling for PCOS status (Figure 1B).

Figure 1.

(A) Plasma macrophage migration inhibitory factor (MIF) measured in fasting blood samples. In a two-way ANOVA model that included age as a covariate, there was no significant interaction effect between PCOS status and weight class, P = 0.58. ^† Main effect of PCOS status (PCOS vs. controls) controlling for the effect of weight class, P < 0.001. ^‡ Main effect of weight class (lean vs. obese) controlling for the effect of PCOS status, P = 0.15. (A) Plasma C-reactive protein (CRP) measured in fasting blood samples. In a two-way ANOVA model that included age as a covariate, there was no significant interaction effect between PCOS status and weight class, P = 0.06. The natural logarithm transformation was applied to CRP values before the analysis. ^† Main effect of PCOS status (PCOS vs. controls) controlling for the effect of weight class, P = 0.59. ^‡ Main effect of weight class (lean vs. obese) controlling for the effect of PCOS status, P < 0.001.

3.4. Correlations

There was a positive correlation between plasma levels of MIF and CRP (r=0.42, p<0.05) for the combined groups. Each of these levels were directly correlated with % body fat (MIF: r=0.44, p<0.04; CRP: r=0.79, p<0.001) and % truncal fat (MIF: r=0.41, p<0.05; CRP: r=0.83, p<0.001). Plasma MIF was also positively correlated with plasma levels of LH (r=0.45, p<0.04), testosterone (r=0.53, p<0.009), androstendione (r=0.58, p<0.005), and fasting insulin (r=0.42, p<0.05). Plasma CRP in turn, was also directly correlated with BMI (r=0.78, p<0.001), waist circumference (r=0.67, p<0.001) and fasting insulin (r=0.73, p<0.001). IS_OGTT was negatively correlated with MIF (r= −0. 51, p<0.02) and CRP (r= −0.73, p<0.001). There were no correlations between plasma CRP and LH or androgens.

4. DISCUSSION

Our data clearly show for the first time that circulating MIF levels are elevated in PCOS independent of obesity. This is evident by the elevated MIF levels in women with PCOS compared to weight-matched controls. In addition, systolic and diastolic blood pressures and triglyceride levels are increased in the obese whether or not they have PCOS, and insulin sensitivity is lower in women with PCOS compared to weight-matched controls with the lowest insulin sensitivity evident in obese women with PCOS. Collectively, these findings support the concept that both PCOS and obesity contribute to inflammation, elevated blood pressure and insulin resistance, which in turn underlie the development of atherosclerosis. MIF is directly related to androgen levels suggesting that proatherogenic inflammation may be promoted by hyperandrogenism in PCOS. Furthermore, the independent associations of MIF and CRP with abdominal adiposity suggest that the accumulation of abdominal fat is an important contributing factor in promoting atherogenesis and subsequent cardiovascular events in obese women with PCOS.

In contrast to plasma MIF, the elevation in plasma CRP appears to be more a function of obesity than PCOS per se. Moreover, CRP in the current study is markedly elevated in obese subjects compared to those who are lean regardless of PCOS status. This latter finding is consistent with previous reports [21,22]. Elevations in MIF and CRP may work in consort in PCOS to promote lipid uptake by MNC-derived foamy macrophages within atherosclerotic plaques. This concept is further supported by the direct relationship between MIF and CRP. Nevertheless, the elevated MIF levels in lean women with PCOS compared to lean controls suggest that accelerated atherogenesis can occur in PCOS independent of obesity. The insulin resistance evident in women with PCOS based on a lower IS_OGTT is a feature highly associated with atherosclerosis [3]. This is confirmed in the present study by the inverse relationship between IS_OGTT and plasma levels of MIF and CRP. Thus, the presence of PCOS in combination with obesity may result in greater risk of inflammation-related atherogenesis compared to having PCOS or obesity alone.

In PCOS, there may be a link between adiposity and circulating MIF levels. While not evident in the present study, our group and other investigators have previously shown that aside from obese women with PCOS, abdominal adiposity can be increased in lean women with the disorder [23–25]. Moreover, plasma MIF is directly related to measures of adiposity, particularly abdominal adiposity. The same is true for plasma CRP, and is corroborated by our previous report [22]. MNC are known to migrate into adipose tissue to stimulate adipocyte production of inflammatory mediators [26]. Inflamed adipose tissue, especially in the abdominal region, may be a perpetuator of elevated circulating MIF levels in obese women with PCOS. Thus, these data are striking because they suggest that adiposity related inflammation may initiate a proatherogenic milieu in women with PCOS at an early age.

In PCOS, hyperandrogenism may be capable of promoting inflammation that can lead to atherosclerosis. This is suggested by the direct relationship between MIF and plasma testosterone and androstenedione. The positive association between circulating androgens and inflammatory mediators in PCOS is a consistent finding corroborated by our previous reports [18,21,27]. In vitro studies have shown that adhesion of MNC to vascular endothelium and oxidation of LDL by MNC-derived macrophages are increased following androgen exposure [28,29]. Furthermore, experimentally induced hyperandrogenism favors the development of atherosclerosis in cholesterol-fed female cynomolgus monkeys [28]. We have previously shown that in PCOS, hyperglycemia causes an increase in ROS generation from MNC [30]. Thus, hyperandrogenism in PCOS may perpetuate NFκB activation following ROS-induced oxidative stress from glucose-activated MNC to upregulate the transcription of inflammatory mediators that are involved in atherogenesis.

In conclusion, circulating MIF levels are elevated in PCOS independent of obesity. This proinflammatory phenomenon may place women with PCOS at an increased risk of developing atherosclerosis. The association of MIF with abdominal fat and circulating androgens suggests that increased abdominal adiposity and hyperandrogenism can contribute significantly to the promotion of atherogenesis in PCOS.