POLYCYSTIC OVARY SYNDROME (PCOS), characterized

Transcription

1 X/06/$15.00/0 The Journal of Clinical Endocrinology & Metabolism 91(12): Printed in U.S.A. Copyright 2006 by The Endocrine Society doi: /jc Increased Visfatin Messenger Ribonucleic Acid and Protein Levels in Adipose Tissue and Adipocytes in Women with Polycystic Ovary Syndrome: Parallel Increase in Plasma Visfatin Bee K. Tan,* Jing Chen,* Janet E. Digby, Stephen D. Keay, C. Richard Kennedy, and Harpal S. Randeva Endocrinology and Metabolism Group (B.K.T., J.C., J.E.D., S.D.K., H.S.R.), Clinical Sciences Research Institute, Warwick Medical School, University of Warwick, and Centre for Reproductive Medicine (C.R.K.), University Hospitals, Coventry CV2 2DX, United Kingdom Context: Polycystic ovary syndrome (PCOS) is a multifaceted metabolic disease linked with insulin resistance (IR) and obesity. Recent studies have shown that plasma levels of the insulin-mimetic adipokine visfatin increase with obesity. Currently, no data exist on the relative expression of visfatin in either plasma or adipose tissue of PCOS women. Objectives: We investigated the mrna expression of visfatin from sc and omental (om) adipose tissue and sc adipocytes in women with PCOS compared with matched normal women, as well as visfatin protein in adipose tissue; plasma visfatin was also assessed. Design: Real-time RT-PCR and Western blotting were used to assess the relative mrna and protein expression of visfatin. Biochemical measurements were performed. POLYCYSTIC OVARY SYNDROME (PCOS), characterized by menstrual dysregulation, hyperandrogenism, and hyperinsulinemia (1), is a multifaceted metabolic disease with insulin resistance (IR), dyslipidemia, impaired glucose tolerance, overt type 2 diabetes mellitus (T2DM), and visceral obesity being more prevalent in women with PCOS than in weight-matched controls (1, 2). The metabolic syndrome is closely associated with excessive accumulation of body fat. Besides its role in energy storage, adipose tissue, an endocrine organ, produces several hormones and cytokines that have wide-ranging effects on carbohydrate and lipid metabolism and therefore appear to play an important role in the pathogenesis of diabetes, IR, and atherosclerosis (3). However, it is apparent that accumulation of visceral adipose tissue poses a greater cardio-metabolic risk than sc adipose tissue, even though the latter is the larger adipose tissue depot of the two (4); removal of visceral rather than sc adipose tissue First Published Online September 26, 2006 * B.K.T. and J.C. have contributed equally to the manuscript. Abbreviations: BMI, Body mass index; DHEA-S, dehydroepiandrosterone sulfate; E 2,17 -estradiol; HOMA, homeostasis model assessment; IR, insulin resistance; om, omental; PCOS, polycystic ovary syndrome; PVDF, polyvinylidene difluoride; TBS, Tris-buffered saline; T2DM, type 2 diabetes mellitus; WHR, waist to hip circumference ratio. JCEM is published monthly by The Endocrine Society ( endo-society.org), the foremost professional society serving the endocrine community. Results: There was significant up-regulation of visfatin mrna in both sc (P 0.05) and om (P 0.05) adipose tissue of PCOS women, when compared with normal controls; these findings were also reflected in isolated sc adipocytes (PCOS controls; P 0.05). In addition to elevated plasma visfatin levels in women with PCOS (mean SD, vs ng/ml; P 0.01) when compared with normal controls, visfatin protein levels were significantly greater in both sc and om adipose tissue of PCOS women (P 0.05 and P 0.01, respectively). Conclusions: The precise reason for the up-regulation of visfatin seen in women with PCOS, a proinflammatory state, is unknown. Additional studies are needed to clarify the potential role of visfatin in the pathophysiology of PCOS. (J Clin Endocrinol Metab 91: , 2006) improves insulin sensitivity (5). Additionally, differences in gene expression of adipocyte-secreted molecules suggest that there are intrinsic adipose tissue depot-specific differences in the endocrine function of adipose tissue. Visfatin (pre-b-cell colony-enhancing factor, PBEF), is a recently described adipokine reported to be preferentially produced by human visceral adipose tissue as compared with sc adipose tissue (6). It reportedly has insulin-mimetic actions, mediated by activation of the insulin receptor in a manner distinct from that of insulin (6). Also, plasma levels increased with obesity and correlated positively with visceral adiposity (6). Interestingly, others have not noted this association (7). Moreover, plasma concentrations of visfatin are increased in subjects with T2DM (8). With the aforementioned in mind and the observation that women with PCOS have an increased prevalence of visceral obesity and the metabolic syndrome (1, 2), we studied the mrna and protein expression of visfatin in both sc and visceral adipose tissue depots in these women and matched controls and their respective plasma visfatin levels. Subjects Subjects and Methods After an overnight fast, blood samples and abdominal (sc) and omental (om) adipose tissue were obtained ( h) from adult female patients undergoing elective surgery for infertility investigation. The sc 5022

2 Tan et al. Visfatin in PCOS J Clin Endocrinol Metab, December 2006, 91(12): biopsies were obtained from the same site, i.e. from a 3-cm horizontal midline incision approximately 3 cm above the symphysis pubis. All samples were obtained during the early follicular phase (d 2 4 from the first day of spontaneous bleeding episode); serum/plasma were immediately aliquoted on ice and stored at 80 C. Adipose tissue biopsies were frozen in liquid nitrogen immediately and stored at 80 C. The sc adipose tissues from subjects were placed into a sterile container containing Medium 199 (Sigma-Aldrich, Gillingham, UK) for primary adipocyte isolation. The control group had no discernible cause for infertility (unexplained infertility). Subjects underwent laparoscopy to rule out fallopian tube(s) obstruction. Furthermore, in most centers in the United Kingdom, and most certainly in our department, either a HysteroContrast Sonography or a laparoscopy is employed to assess fallopian tube(s) patency before commencing treatment (for example, clomiphene citrate) to correct anovulation, as the case may be for PCOS women, and thereby to rule out concurrent fallopian tube(s) obstruction. All patients underwent anthropometric measurements, i.e. weight, height, and waist to hip circumference ratio (WHR). Equal numbers of women (n 8) were recruited in each group, one that consisted of women with PCOS and the other healthy controls (Table 1). Eight women were overweight, i.e. with body mass index (BMI) between 25 and 30 (four PCOS and four controls); eight were obese, i.e. with BMI between 30 and 35 (four PCOS and four controls). The Local Research Ethics Committee approved the study, and all patients involved gave their informed consent, in accordance with the guidelines in The Declaration of Helsinki All PCOS patients met each of the three criteria of the revised 2003 Rotterdam ESHRE/ASRM PCOS Consensus Workshop Group diagnostic criteria (9). Furthermore, all subjects in the control arm had normal findings on pelvic ultrasound scan, regular periods, and no hirsutism/ acne. No women were amenorrheic. Exclusion criteria for the study included age over 40 yr, known cardiovascular disease, thyroid disease, neoplasms, current smoking, diabetes mellitus, hypertension (blood pressure 140/90 mm Hg), renal impairment (serum creatinine 120 mol/liter). None of these women, PCOS and controls, were on any medications for at least 6 months before the study, including oral contraceptives; glucocorticoids; ovulation induction agents; antidiabetic and antiobesity drugs; or estrogenic, antiandrogenic, or antihypertensive medication. Also, the presence of other endocrinopathies were ruled out at the time of recruitment (more than 7 d before surgery) by measuring basal serum 17-hydroxyprogesterone and prolactin and by measuring h cortisol after 1.0 mg (2300 h) overnight dexamethasone suppression (a value of 30 nmol/liter was considered to rule out Cushing s syndrome). All subjects suppressed cortisol less than 30 nmol/liter. Biochemical and hormonal analysis Assays for glucose, insulin, LH, FSH, 17 -estradiol (E 2 ), progesterone, testosterone, androstenedione, dehydroepiandrostenedione sulfate (DHEA-S), and SHBG were performed using an automated analyzer (Abbott Architect; Abbott Laboratories, Abbott Park, IL). The estimate of IR by homeostasis model assessment (HOMA) score was calculated as I o G o /22.5, where I o is the fasting insulin and G o is the fasting glucose, as described by Matthews et al. (10). Plasma visfatin levels were measured using a commercially available ELISA (EIA kit; Phoenix Pharmaceuticals, Belmont, CA) according to the manufacturer s protocol, with a coefficient of variation of less than 6%. Isolation of primary adipocytes Primary adipocytes were isolated by a method modified from that of Rodbell (11) and Gliemann et al. (12). Fat samples were washed with a large volume of PBS, meticulously cleaned of obvious blood vessels, connective tissue, and skin, and then chopped with sterile scissors. The cleaned samples were washed again with saline, and the floating, clean fat recovered by centrifugation. Adipocytes were isolated by collagenase digest (Hanks balanced salt solution containing 3 mg/ml collagenase type II and 1.5% BSA) in a shaking water bath at 37 C for up to 60 min. Mature adipocytes were separated from the stromal vascular cells through an inert oil, bis-3,5,5-trimethylhexyl phthalate (Fluka Chemicals, Gillingham, UK). Total RNA extraction and cdna synthesis from isolated adipocytes were performed immediately as described below. To exclude potential contamination of the adipocytes, von Willebrand factor was examined, revealing its absence in adipocytes, employing human umbilical vascular endothelial cells as the positive control; moreover, there was negligible CD45 and CD14 gene expression when compared with adipose tissue as the positive control (data not shown). Total RNA extraction and cdna synthesis Total RNA was extracted from whole adipose tissue samples and isolated adipocytes using QIAGEN RNeasy Lipid Tissue Mini Kit according to the manufacturer s guidelines (QIAGEN, Crawley, UK). Firststrand cdna synthesis was carried out using Moloney Murine Leukemia Virus reverse transcriptase and random hexamers as primers. RT-PCR Quantitative PCR of visfatin, CD45, and CD14 were performed on a Roche Light Cycler system (Roche Molecular Biochemicals, Mannheim, Germany). PCR were carried out in a reaction mixture consisting of 5.0 l reaction buffer and 2.0 mm MgCl 2 (Biogene, Kimbolton, UK), 1.0 l of each primer (10 ng/ l), 2.5 l cdna and 0.5 l Light Cycler DNA Master SYBR Green I (Roche). Protocol conditions consisted of denaturation of 95 C for 15 sec, followed by 40 cycles of 94 C for 1 sec, 58 C for 10 sec, and 72 C for 12 sec, followed by melting-curve analysis. For analysis, quantitative amounts of genes of interest were standardized against the housekeeping gene -actin. The RNA levels were expressed as a ratio, using the delta-delta method for comparing relative expres- TABLE 1. Clinical, hormonal, and metabolic features of women with PCOS and controls Variable PCOS Controls Significance Age (yr) NS BMI (kg/m 2 ) NS WHR NS Glucose (mmol/liter) NS Insulin (pmol/liter) P 0.05 HOMA P 0.05 LH (IU/liter) NS FSH (IU/liter) NS Prolactin (mu/liter) NS E 2 (pmol/liter) NS Testosterone (nmol/liter) P 0.05 Androstenedione (nmol/liter) P 0.05 DHEA-S ( mol/liter) P 0.05 SHBG (nmol/liter) P 0.05 FAI P 0.05 Visfatin (ng/ml) P 0.01 Values are given as mean SD. Free androgen index (FAI) testosterone (nmol/liter)/shbg (nmol/liter) 100. n 8 in each group. NS, Not significant.

3 5024 J Clin Endocrinol Metab, December 2006, 91(12): Tan et al. Visfatin in PCOS sion results between treatments in real-time PCR (13). The following primers were used for our study: visfatin (228 bp), 5 -AAGAGACT- GCTGGCATAGGA-3 and 5 -ACCACAGATACAG GCACTGA-3 ; CD45 (194 bp), 5 -AGGAGAGTGAATGCCTTCAG-3 and 5 -GCC TC- TACTTGAACCATCAG-3 ; and CD14 (173 bp), 5 -CGCAACACAG- GAATGGAGAC-3 and 5 -CCAGCGAACGACAGATTGAG-3. Ten microliters of the reaction mixture was subsequently electrophoresed on a 1% agarose gel and visualized by ethidium bromide, using a 1-kb DNA ladder (Life Technologies, Inc.-BRL, Paisley, UK) to estimate the band sizes. As a negative control for all the reactions, preparations lacking RNA or reverse transcriptase were used in place of the cdna. RNAs were assayed from three independent biological replicates. Sequence analysis The PCR products from the adipose tissue and adipocyte samples were purified from the 1% agarose gel using the QIAquick Gel Extraction Kit (QIAGEN). PCR products were then sequenced in an automated DNA sequencer, and the sequence data were analyzed using Blast Nucleic Acid Database Searches from the National Centre for Biotechnology Information, confirming the identity of our products. Western blotting Protein lysates were prepared by homogenizing adipose tissue in radioimmunoprecipitation lysis buffer (Upstate Biotechnology, Lake Placid, NY) according to the manufacturer s instructions. Equal amounts of Laemmli buffer (5 m urea, 0.17 m SDS, 0.4 m dithiothreitol, and 50 mm Tris-HCl, ph 8.0) were added, mixed, and placed in a boiling water bath for 5 min. Lysates were allowed to cool at room temperature. The proteins (38.7 g/lane) were separated by SDS-PAGE (10% resolving gel) and transferred to polyvinylidene difluoride (PVDF) membranes at 100 V for 1 h in atransfer buffer containing 20 mm Tris, 150 mm glycine, and 20% methanol. PVDF membranes were then blocked in Tris-buffered saline (TBS) containing 0.1% Tween 20 and 5% BSA for 2 h. After three washes with TBS/0.1% Tween, the PVDF membranes were incubated with primary rabbit-antihuman antibody for visfatin (Bethyl Laboratories Inc., Montgomery, TX) at a 1:10,000 dilution or primary rabbitantihuman antibody for -actin (Cell Signaling Technology Inc., Beverly, MA) at a 1:1000 dilution overnight at 4 C. The membranes were washed thoroughly for 60 min with TBS/0.1% Tween before incubation with the secondary antirabbit horseradish-peroxidase-conjugated Ig (1:2000) for 1 h at room temperature. Antibody complexes were visualized using chemiluminescence (ECL; Amersham Pharmacia, Little Chalfont, UK). Visfatin peptide was used as the positive control and water as the negative control (data not shown). Statistical analyses For the real-time PCR measurements, Western immunoblotting, and blood analysis, results were evaluated between groups by using unpaired Student s t test, with significance determined at the level of P The Wilcoxon matched-pairs test and the Mann-Whitney U test were used for comparisons within and between groups, respectively, and P 0.05 was considered significant. For Western immunoblotting experiments, the densities were measured using a scanning densitometer coupled to scanning software ImageQuant (Molecular Dynamics, Amersham Pharmacia). Spearman rank correlation was used for calculation of associations between variables, and multiple regression analysis contained insulin, HOMA, testosterone, and SHBG; P 0.05 was considered significant. Results Demographic data Table 1 shows the anthropometric, biochemical, and hormonal data in both PCOS and control women, and the data are shown as the mean sd. As expected, insulin, testosterone, and androstenedione levels, HOMA, and free androgen index were significantly higher in women with PCOS. ELISA analysis of plasma visfatin levels revealed that PCOS patients have significantly elevated levels when compared with controls ( vs ng/ml; P 0.01; Table 1). Serum progesterone levels in all women confirmed follicular phase of the menstrual cycle ( nmol/liter; follicular phase, 6 nmol/liter; luteal phase, 6 64 nmol/liter). mrna expression of visfatin in normal and PCOS women Using RT-PCR analysis, we detected visfatin mrna in adipose tissue and in adipocytes, and subsequent sequencing of the PCR products confirmed the gene identity. Furthermore, real-time RT-PCR analysis corrected over -actin showed a significant up-regulation of visfatin in both sc and om adipose tissue of PCOS women, when compared with normal controls (Fig. 1, A and B; P 0.05). Similar findings were noted in isolated sc adipocytes from these women, where PCOS women had significantly higher expression of visfatin compared with controls (Fig. 1C; P 0.05). Interestingly, visfatin was seen to be significantly up-regulated in om adipose tissue of PCOS patients when compared with corresponding sc adipose tissue (Fig. 1D; P 0.01), whereas there was no significant difference in the expression of visfatin in the om and sc adipose tissue depots of normal controls (Fig. 1D; P 0.05). mrna expression of CD45 and CD14 in normal and PCOS women Once again, RT-PCR analysis detected CD45 and CD14 mrna in adipose tissue of both groups of women, and subsequent sequencing of the PCR products confirmed the identity of both genes (Fig. 2A-B). The levels of CD45 and CD14 mrna, a marker for macrophage index, were assessed in normal and PCOS women, both in sc and om adipose tissues. Real-time RT-PCR analysis revealed that there was no statistical difference in CD45 and CD14 between both groups of women in either adipose tissue depot (P 0.05). Protein expression of visfatin in adipose tissue in normal and PCOS women The changes noted at the mrna level were also reflected at the protein level in PCOS women, with significantly greater visfatin expression levels in om vs. sc adipose tissue (Fig. 3A; P 0.05), but no significant difference in -actin protein was noted as per immunoblotting (Fig. 3B; P 0.05). Interestingly, unlike the mrna data, visfatin protein was significantly higher in om vs. sc adipose tissue in normal controls (Fig. 3A; P 0.05). The detected proteins for visfatin and -actin have apparent molecular masses of 55 and 45 kda, respectively (Fig. 3B, inset). Furthermore, when compared with normal controls, visfatin protein levels were significantly greater in both sc and om adipose tissue depots of PCOS women (Fig. 3B; P 0.05 and P 0.01, respectively). Association of visfatin with other covariates When both groups were analyzed collectively, no significant correlation was noted between plasma visfatin and BMI (P 0.12) or WHR (P 0.15). Plasma visfatin was positively associated with insulin (P 0.01), HOMA (P 0.01), tes-

4 Tan et al. Visfatin in PCOS J Clin Endocrinol Metab, December 2006, 91(12): FIG. 1. A and B, Expression of visfatin (mean SEM) is significantly increased in both human sc and om adipose tissue depots when comparing all PCOS women to all normal controls, using real-time RT- PCR. C, Also, expression of visfatin (mean SEM) is significantly greater in human sc adipocytes when comparing PCOS women with normal controls, using real-time RT-PCR. D, Expression of visfatin (mean SEM) is significantly up-regulated in the om adipose tissue of PCOS patients when compared with the sc adipose tissue, using realtime RT-PCR. However, no significant (P 0.05) difference in the expression of visfatin was observed when comparing om with sc adipose tissue in normal controls collectively, using real-time RT- PCR. *, P 0.05; **, P tosterone (P 0.03), and E 2 (P 0.046) and negatively with SHBG (P 0.03). Multiple regression analysis contained insulin, HOMA, testosterone, and SHBG (Table 2). Interestingly, HOMA was the only predictor of plasma visfatin ( 1.23; P 0.02), with insulin failing to reach significance ( 0.54; P 0.09). Similar findings were noted when the groups were analyzed individually. Furthermore, plasma visfatin levels correlated positively with visfatin mrna from sc (P 0.01) and om (P 0.037) and visfatin protein levels from sc (P 0.002) and om adipose tissue (P 0.008). Unlike plasma visfatin, both sc and om adipose tissue visfatin mrna expression were positively correlated with BMI (sc, P 0.05; om, P 0.01) and WHR (sc, P 0.04; om, P 0.015). However, similar to plasma visfatin, we did not find any significant correlation between sc and om adipose tissue visfatin protein levels and BMI (sc, P 0.418; om, P 0.381) and WHR (sc, P 0.452; om, P 0.555). Moreover, a significant positive correlation was observed between sc and om visfatin gene expression (P 0.007) and between sc and om visfatin protein levels (P 0.008), but no significant correlation was noted between visfatin mrna and visfatin protein levels in either fat depot (sc, P 0.073; om, P 0.157). Furthermore, although sc and om visfatin protein levels were significantly positively correlated with insulin, HOMA, and testosterone and negatively correlated with SHBG (Tables 3 and 4), in multiple linear regression analysis, only insulin ( 1.15l; P 0.05) and HOMA ( 2.35; P 0.014) were predictors of visfatin protein in om adipose tissue; HOMA just failed to reach significance in sc adipose tissue (P 0.07). Discussion In our present study, we show the expression of visfatin, a recently described adipokine (6), in corresponding sc and om human adipose tissues at both mrna and protein levels. Furthermore, to our knowledge, this is the first report to demonstrate the presence and significant up-regulation of visfatin gene expression and protein in both these adipose tissue depots in women with PCOS, including significantly higher plasma visfatin levels, compared with matched controls. The reasons for our observations remain to be determined. The up-regulation of visfatin in PCOS women, a prodiabetogenic state, is of interest given that subjects with T2DM have higher plasma and adipose tissue mrna expression of visfatin levels, compared with controls (8, 14),

5 5026 J Clin Endocrinol Metab, December 2006, 91(12): Tan et al. Visfatin in PCOS FIG. 2. Expression of CD45 and CD14 (mean SEM) in both human sc and om adipose tissue depots comparing all PCOS women with all normal controls, using real-time RT-PCR. No significant differences in either CD45 or CD14 mrna expression levels were detected (P 0.05). after accounting for BMI and WHR. In our study, it is unlikely that either BMI or WHR is responsible for higher visfatin levels in PCOS women, given that both groups were matched for these variables. Besides, Hammarstedt et al. (14), like us, did not find any significant correlation between WHR, a surrogate marker for visceral fat, and circulating levels of visfatin. Moreover, like others (7), although we detected a significant positive relationship between visfatin gene expression and BMI and WHR in both adipose tissue depots, we report novel data of no significant correlation between these anthropometric measures and visfatin protein in adipose tissue, in keeping with our plasma visfatin data. Our observations and those of others (7, 14) are in contrast to the original findings by Fukuhara et al. (6) and the more recent findings by Chen et al. (8), where a positive correlation was noted between plasma visfatin and visceral fat mass and also WHR. Moreover, unlike us, they did not study visfatin protein expression in adipose tissue and its relation to BMI/ WHR. The precise reason for these differences is not clear at present, but it may be related to ethnic differences, where some studies are from Far Eastern populations (6, 8) and others (7, 14), like ours, from Caucasians. The insulin-mimetic action of visfatin has been proposed to contribute to the development of the metabolic syndrome (6). Women with PCOS, as in our study, have a higher incidence of IR. We noted a positive correlation between plasma visfatin levels and both insulin and HOMA, as previously reported by some authors (8, 15) but not others (7); however, in multiple linear regression analysis, only HOMA was significantly associated with plasma visfatin levels. We went on further to determine the relationship between visfatin protein in adipose tissue and metabolic parameters and FIG. 3. Western blot analysis of protein extracts from adipose tissues of all normal controls and PCOS women. A, Densitometric analysis of visfatin immunocomplexes (mean SEM) having normalized to actin revealed that there is a significant (*, P 0.05) increase in protein expression of visfatin in om vs. sc adipose tissue of both normal and PCOS women. B, Western blot analysis of protein extracts from adipose tissues of normal and PCOS women demonstrate that the antibody against visfatin and the antibody against -actin recognized bands with apparent molecular masses of 55 and 45 kda, respectively, in both sc and om adipose tissue depots (inset). Densitometric analysis of visfatin immunocomplexes (mean SEM), having normalized to -actin, revealed that there is a significant increase in protein expression of visfatin in both sc (*, P 0.05) and om (**, P 0.01) adipose tissues of all PCOS women when compared with all normal controls. report new findings that insulin and HOMA are significantly correlated to visfatin protein in om but not sc adipose tissue. In addition, we noted a significant correlation between plasma visfatin and protein visfatin levels in both om and sc adipose tissue, a finding not described before. TABLE 2. Linear regression analysis of variables associated with plasma visfatin in subjects studied Variable Simple Multiple Estimate P Estimate P BMI (kg/m 2 ) WHR Insulin (pmol/liter) HOMA E 2 (pmol/liter) Testosterone (nmol/liter) DHEA-S ( mol/liter) SHBG (nmol/liter) In multiple linear regression analysis, values included were insulin, HOMA, testosterone, and SHBG.

6 Tan et al. Visfatin in PCOS J Clin Endocrinol Metab, December 2006, 91(12): TABLE 3. Linear regression analysis of variables associated with sc adipose tissue visfatin protein in subjects studied Variable In our control group, similar to others (7, 14), we found no difference in visfatin mrna levels between om and sc adipose tissue, but we did observe significantly higher visfatin protein expression in om when compared with sc tissue. In our PCOS group, om visfatin mrna expression and protein level was significantly higher than sc adipose tissue. The precise reason for higher omental visfatin in both control and PCOS women is unclear. Recently, macrophages in human visceral adipose tissue have been described to be a source of visfatin (16). In our study, in both controls and PCOS women, as expected, there was higher CD14 and CD45 gene expression, hallmarks of macrophages, in om vs. sc adipose tissue depots. More importantly, PCOS women did not have significantly higher CD14 and CD45 levels in either adipose tissue depot compared with controls. Therefore, although they may contribute to the visfatin pool, it is unlikely that macrophages are the key reason for our observations. Furthermore, in isolated sc adipocytes, visfatin expression was significantly higher in PCOS women. Unfortunately, because of technical limitations in om adipose tissue procurement, we were unable to obtain sufficient amounts of sample per patient to perform adipocyte separation in om adipose tissue depots. These limitations notwithstanding, it is clear that both adipose tissue and adipocytes in PCOS women express more visfatin, with a parallel increase in plasma visfatin. Studies in adipocytes from women with PCOS reveal adipocyte insensitivity to inhibition of lipolysis by insulin as well as a decrease in maximal rates of adipocyte glucose uptake (17). We hypothesize that the elevated visfatin levels in both plasma and adipose tissue in PCOS women in our study may suggest a role of visfatin signaling in target tissues TABLE 4. Linear regression analysis of variables associated with om adipose tissue visfatin protein in subjects studied Variable Simple Multiple Estimate P Estimate P BMI (kg/m 2 ) WHR Insulin (pmol/liter) HOMA E 2 (pmol/liter) Testosterone (nmol/liter) DHEA-S ( mol/liter) SHBG (nmol/liter) In multiple linear regression analysis, values included were insulin, HOMA, testosterone, and SHBG. Simple Multiple Estimate P Estimate P BMI (kg/m 2 ) WHR Insulin (pmol/liter) HOMA E 2 (pmol/liter) Testosterone (nmol/liter) DHEA-S ( mol/liter) SHBG (nmol/liter) In multiple linear regression analysis, values included were insulin, HOMA, testosterone, and SHBG. or the dysregulation in biosynthesis and thereby the increased need for insulin receptor signaling in PCOS adipose tissue. In addition, we speculate that the increased visfatin in PCOS women may be a compensatory response to IR in specific tissues or a marker of tissue-specific inflammatory cytokine action and that this may account, in particular, for the higher om visfatin expression. Finally, visfatin may not only have an endocrine role in modulating insulin sensitivity (6), but also have autocrine/paracrine effects, given that it is expressed in pure adipocytes, as shown here and by others (16). A limitation of our study may relate to the number of subjects studied. However, obtaining BMI/WHR-matched and menstrual cycle-synchronized blood and tissue samples impeded subject recruitment. This may in part explain why in bivariate analysis, androgens (testosterone and DHEA-S) significantly correlated with only insulin and HOMA, and testosterone with visfatin (data not shown); however, in multivariate analysis, insulin and HOMA were the only significant factors correlating with visfatin. It is also important to recognize that potentially several of these correlations may be without causative significance and result from the fact that hyperandrogenism and IR characterizes PCOS. Nonetheless, our observations are consistent and significant and raise interesting questions on the mechanisms regulating visfatin expression. Furthermore, it will be interesting to see whether these findings are also present in matched lean (BMI 25) PCOS women vs. normal controls. In conclusion, we present novel data with regard to the increased expression both at the mrna and protein levels as well as plasma levels of visfatin in women with PCOS. Although the up-regulation of visfatin seen in PCOS women may act in an endocrine manner, we speculate that visfatin may have important autocrine/paracrine effects. The physiological and pathological significance of our findings remain to be elucidated. Acknowledgments We thank Dr. E. Karteris for proofreading this manuscript. H.S.R. acknowledges S. Waheguru, University of Warwick, for his continual support. Received May 3, Accepted September 20, Address all correspondence and requests for reprints to: Dr. Harpal S. Randeva, MBCHB, FRCP, Ph.D., Endocrinology and Metabolism Group, Department of Biological Sciences, Warwick Medical School, The University of Warwick, Coventry CV4 7AL, United Kingdom. hrandeva@bio.warwick.ac.uk. This study was funded by The General Charities of the City of Coventry. Disclosure statement: The authors have nothing to declare. References 1. Dunaif A 1997 Insulin resistance and the polycystic ovary syndrome: mechanism and implications for pathogenesis. Endocr Rev 18: Wild RA, Painter RD, Coulson PB, Carruth KB, Ranney RB 1985 Lipoprotein lipid concentrations and cardiovascular risk in women with polycystic ovary syndrome. J Clin Endocrinol Metab 61: Kershaw EE, Flier JS 2004 Adipose tissue as an endocrine organ. J Clin Endocrinol Metab 89: Wajchenberg BL 2000 Subcutaneous and visceral adipose tissue: their relation to the metabolic syndrome. Endocr Rev 21: Thorne A, Lonnqvist F, Apelman J, Hellers G, Arner P 2002 A pilot study of

7 5028 J Clin Endocrinol Metab, December 2006, 91(12): Tan et al. Visfatin in PCOS long-term effects of a novel obesity treatment: omentectomy in connection with adjustable gastric banding. Int J Obes Relat Metab Disord 26: Fukuhara A, Matsuda M, Nishizawa M, Segawa K, Tanaka M, Kishimoto K, Matsuki Y, Murakami M, Watanabe E, Takagi T, Akiyoshi M, Ohtsubo T, Kihara S, Yamashita S, Makishima M, Funahashi T, Yamanaka S, Hiramatsu R, Matsuzawa Y, Shimomura I 2005 Visfatin: a protein secreted by visceral fat that mimics the effects of insulin. Science 307: Berndt J, Kloting N, Kralisch S, Kovacs P, Fasshauer M, Schon MR, Stumvoll M, Bluher M 2005 Plasma visfatin concentrations and fat depot-specific mrna expression in humans. Diabetes 54: Chen MP, Chung FM, Chang DM, Tsai JC, Huang HF, Shin SJ, Lee YJ 2006 Elevated plasma level of visfatin/pre-b cell colony-enhancing factor in patients with type 2 diabetes mellitus. J Clin Endocrinol Metab 91: Fauser B 2004 Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS). Hum Reprod 19: Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC 1985 Homeostasis model assessment: insulin resistance and -cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 28: Rodbell M 1964 Metabolism of isolated fat cells. Effects of hormones on glucose metabolism and lipolysis. J Biol Chem 239: Gliemann J, Osterlind K, Vinten J, Gammeltoft S 1972 A procedure for measurement of distribution s in isolated fat cells. Biochim Biophys Acta 286: Pfaffl MW 2001 A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 29:e Hammarstedt A, Pihlajamaki J, Rotter Sopasakis V, Gogg S, Jansson PA, Laakso M, Smith U 2006 Visfatin is an adipokine, but it is not regulated by thiazolidinediones. J Clin Endocrinol Metab 91: Haider DG, Schindler K, Schaller G, Prager G, Wolzt M, Ludvik B 2006 Increased plasma visfatin concentrations in morbidly obese subjects are reduced after gastric banding. J Clin Endocrinol Metab 91: Curat CA, Wegner V, Sengenes C, Miranville A, Tonus C, Busse R, Bouloumie A 2006 Macrophages in human visceral adipose tissue: increased accumulation in obesity and a source of resistin and visfatin. Diabetologia 49: Ek I, Arner P, Bergqvist A, Carlstrom K, Wahrenberg H 1997 Impaired adipocyte lipolysis in nonobese women with the polycystic ovary syndrome: a possible link to insulin resistance? J Clin Endocrinol Metab 82: JCEM is published monthly by The Endocrine Society ( the foremost professional society serving the endocrine community.