Correlation between Soluble Endoglin, Vascular Endothelial Growth Factor Receptor-1, and Adipocytokines in Preeclampsia Hisashi Masuyama, Hideki Nakatsukasa, Norio Takamoto, and Yuji Hiramatsu Department of Obstetrics and Gynecology, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Shikata, Okayama 700-8558, Japan Context: Recent reports have demonstrated that soluble endoglin (sEng), an antiangiogenic protein thought to impair TGF- binding to receptors, and soluble vascular endothelial growth factor receptor (sVEGFR)-1 play important roles in the pathophysiology of pre- eclampsia (PE). Moreover, insulin resistance, which is greatly influ- enced by adipocytokines, characterizes PE. Objectives: We examined possible links between sEng, VEGF, sVEGFR, and adipocytokines in the pathophysiology of PE. Study Design: We performed a cross-sectional study in 30 PE pa- tients and controls matched for gestational age and body mass index. Blood samples were collected soon after disease onset. We measured serum concentrations of leptin, adiponectin, sEng, VEGF, placental growth factor (PlGF), and sVEGFR [soluble fms-like tyrosine kinase 1 (sFlt-1) and soluble fetal liver kinase 1 (sFlk-1)], and examined the placental protein content of sEng and sFlt-1. Results: sEng concentrations in PE patients (60.9 28.8 ng/ml) were significantly higher than those in controls (11.2 4.4 ng/ml). There was a significant correlation between sEng and sFlt-1 or PlGF. More- over, there were significant differences in mean blood pressure be- tween the high and low sEng groups, and in proteinuria between the high and low sFlt-1 groups, and significant differences in placental sEng and sFlt-1 contents between patients with and without severe hypertension or proteinuria. sEng was also correlated positively with adiponectin levels and negatively with the leptin to adiponectin ratio. Conclusions: Along with sFlt-1 and PlGF, sEng might play a role in the pathophysiology of PE, especially in elevating blood pressure, while the association with hypoadiponectinemia and the high leptin to adiponectin ratio in pregnancy seem to be risk factors for PE. (J Clin Endocrinol Metab 92: 2672���2679, 2007) PREECLAMPSIA (PE) IS characterized by the onset of high blood pressure and proteinuria. It occurs in about 5���10% of pregnancies and results in substantial maternal and neonatal morbidity and mortality (1, 2). The pathogenesis of PE is thought to involve three steps: abnormal remodeling of the placental bed vasculature, placental ischemia, and endothelial cell dysfunction (3, 4). Angiogenic factors such as vascular en- dothelial growth factor (VEGF) and placental growth factor (PlGF) are thought to be potentially important regulators in the human placenta (5), while decreased concentrations of circu- lating free VEGF and PlGF have been noted in PE (6���8). More- over, recent reports have indicated that the soluble form of VEGF receptor-1 [sVEGFR-1 or soluble fms-like tyrosine kinase 1 (sFlt-1)] is increased in the placenta and serum of women with PE (8���11) although we have demonstrated that there is no relationship between PE and another soluble form of the VEGF receptor, VEGFR-2 [or soluble fetal liver kinase 1 (sFlk-1)] (8). Thus, sFlt-1 may act by sequestering free PlGF and VEGF, thereby preventing interaction between endothelial receptors and free PlGF or VEGF on the cell surface, and inducing en- dothelial cell dysfunction. Endoglin (Eng) is a component of the receptor complex for TGF- and interacts efficiently with TGF- (12). It modulates the TGF- signaling pathway by interacting with TGF- re- ceptors I and II, and a proliferation-associated, hypoxia-in- ducible protein abundantly expressed in angiogenic endo- thelial cells and the cytotrophoblast. Several studies have suggested that Eng is a pro-angiogenic component that pro- tects endothelial cells under hypoxia-induced apoptosis and regulates nitric oxide-dependent vasodilatation (13���15). On the other hand, soluble endoglin (sEng) is an antiangiogenic protein thought to impair TGF- 1 binding to cell surface receptors and to decrease endothelial nitric oxide signaling (16, 17). Recent reports have demonstrated that sEng, which is placental in origin, is elevated in the sera of preeclamptic individuals, increasing with disease severity and falling after delivery, and that administration of both sFlt1 and sEng produces a severe PE-like animal model with hypertension, proteinuria, glomerular endotheliosis, and features of HELLP (hemolysis, elevated liver enzymes, low platelets) syndrome (16, 18). These data suggest that sEng plays an important role in the pathophysiology of PE. In addition to alterations in angiogenesis, there is an as- sociation between markers of insulin resistance and PE, and insulin resistance syndrome has been observed before, dur- ing, and after this pregnancy complication (19). Moreover, recent reports have suggested that insulin signaling and an- giogenesis are intimately related (20, 21), and that insulin regulates the expression of genes involved in angiogenesis, First Published Online April 10, 2007 Abbreviations: BMI, Body mass index Eng, endoglin PE, preeclampsia PlGF, placental growth factor sEng, soluble endoglin sFlk-1, soluble fetal liver kinase 1 sFlt-1, soluble fms-like tyrosine kinase 1 sVEGFR, soluble VEGFR VEGF,vascularendothelialgrowthfactor VEGFR,VEGFreceptor. JCEM is published monthly by The Endocrine Society (http://www. endo-society.org), the foremost professional society serving the endo- crine community. 0021-972X/07/$15.00/0 The Journal of Clinical Endocrinology & Metabolism 92(7):2672���2679 Printed in U.S.A. Copyright �� 2007 by The Endocrine Society doi: 10.1210/jc.2006-2349 2672

including the expression of VEGF mRNA, and VEGF sig- naling also activates Akt/protein kinase B phosphorylation in the insulin signaling pathway (22, 23). These data suggest that alternations in angiogenesis and insulin resistance may have an additive effect that leads to alterations in critical cellular functions, endothelial cell injury, and, subsequently, increased risk of developing PE. Adipose tissue functions as a highly specialized endocrine and paracrine tissue, produc- ing an array of adipocytokines such as leptin, TNF- , and adiponectin. Such factors have local and systemic biological effects, and play important roles in insulin signaling and the development of metabolic diseases (24, 25). In this study, to determine possible links among sEng, angiogenic factors, and adipocytokines in the pathophysiol- ogy of PE, we first examined serum concentrations of sEng in pregnant women with PE and healthy pregnant women. Second, to determine any relationships between sEng and angiogenic factors, we measured levels of the circulating angiogenic factors VEGF and PlGF, and the soluble VEGF receptors sFlt-1 and sFlk-1 in women with PE and in healthy pregnant women. Third, we examined the association of sEng and sFlt-1 with severity of PE, and finally, whether there are significant correlations between sEng and adipo- cytokines in PE and healthy pregnant women. Patients and Methods Patients A total of 60 pregnant Japanese women who visited the Department of Obstetrics and Gynecology, Okayama University Hospital, Japan, were included in the present study. Of these 60 women, 30 had severe PE, and 30 were controls matched for age, gestational age, parity, and body mass index (BMI), with normotensive pregnancies. There were 10 age-matched nonpregnant women, 20 healthy pregnant women in the first trimester, and those in the second trimester also included as vol- unteers. According to the definition of the Japan Society of Obstetrics and Gynecology, PE was defined as a persistent elevation of systolic blood pressure to 140 mm Hg and diastolic blood pressure to 90 mmHg on two occasions several hours apart, with proteinuria of more than 300 mg in a 24-h urine collection. Severe PE was defined as either severe hypertension (systolic blood pressure 160 mm Hg or diastolic blood pressure 110 mm Hg) or severe proteinuria ( 2.0 g protein in a 24-h urine collection). None of the patients with PE had any prior history of renal disorder or essential hypertension. The healthy pregnant women had no history of illness and no form of hypertension or renal disorder. PE patients were also subdivided using the median values of sEng, sFlt-1, or sFlt-1/PlGF as cutoff points: those with a high-circulating sEng level [high sEng group, 55 ng/ml (n 15)] low-circulating sEng level [low sEng group, 55 ng/ml (n 15)] high-circulating sFlt-1 level [high sFlt-1 group, 5300 pg/ml (n 15)] low-circulating sFlt-1 level [low sFlt-1 group, 5300 pg/ml (n 15)] high-circulating sFlt-1/PlGF ratio [high sFlt-1/PlGF group, 33 (n 15)] and low-circulating sFlt-1/PlGF ratio [low sFlt-1/PlGF group, 33 (n 15)]. Clinical records were carefully reviewed, and some patients were interviewed further at the time of sample collection, with those not meeting the aforementioned criteria being eliminated from the study. The study was approved by the Institutional Ethical Review Board of Okayama University Hospital, and all subjects gave informed consent. Blood samples were collected from PE patients soon after disease onset. All PE patients underwent termi- nation within 1 wk from the time of sample collection, due to severe PE. None of the subjects received any medication before blood sampling. Samples from healthy pregnant women in their second or third trimester were matched with PE patients by age, gestational week, parity, and BMI to avoid possible bias. All patients, including healthy controls, visited Okayama University, and all blood samples (5 ml from each patient) were collected between 2000 and 2006. Immediately after sample col- lection, the serum was separated by centrifugation and stored at 80 C until use. The average time of freezer storage in PE patients (2.4 1.6 yr) was not significantly different from that of the controls (2.1 1.3 yr). ELISA for angiogenic factors and adipocytokines Serum levels of VEGF, PlGF, sFlt-1, sFlk-1, adiponectin, leptin, and sEng were determined by ELISA (R&D Systems, Minneapolis, MN), following the manufacturer���s instructions. All samples were examined in duplicate, and the mean values of individual sera were used for statistical analysis. The minimum detectable concentrations in the assays for VEGF, PlGF, sFlt-1, sFlk-1, adiponectin, leptin, and sEng were 5.0, 7.0, 5.0, 5.0, 3.0, 0.5, and 0.1 ng/ml, respectively. The intraassay and inter- assay coefficients of variation for VEGF were less than 4.5% and 7.0%, respectively PlGF, 3.6% and 11.0% sFlt-1, 3.8% and 7.0% sFlk-1, 3.6% and 6.9% adiponectin, less than 3.5% and 7.0% leptin, less than 3.0% and 7.5% and sEng, 3.0% and 6.5%. Western blot analysis Protein extracts were obtained from placental tissues of patients with PE and healthy controls using tissue protein extraction reagent (Pierce, Rock- ford,IL)accordingtothemanufacturer���sprotocol,andstoredat 80Cuntil analysis. Equivalent amounts of protein (50 g/sample) from each extract were determined by bicinchoninic acid protein assay (Pierce), then solu- bilized in sodium dodecyl sulfate buffer [0.05 m Tris/HCl, 2% sodium dodecyl sulfate, 6% mercaptoethanol, and 10% glycerol (pH 6.8)]. Tissue extracts were separated on 10% SDS-PAGE, along with molecular weight markers, and subsequently transferred to polyvinylidene fluoride mem- branes (Amersham Biosciences, Piscataway, NJ). Nonspecific protein bind- ing was blocked with 3% nonfat dry milk in Tris-buffered saline for 1 h at room temperature. The blots were incubated with either rabbit polyclonal antibody raised against sFlt-1 (1:200 dilution Zymed Laboratories, San Francisco, CA) or mouse monoclonal antibody raised against sEng (1:200 dilution Santa Cruz Biotechnology, Santa Cruz, CA) at 4 C overnight. Rabbit polyclonal antibody raised against -actin (1:1000 dilution Santa Cruz Biotechnology) was used to confirm equal loading. The blots were then washed before incubation with horseradish peroxidase-conjugated antirabbit or mouse secondary antibody (1:2000 dilution Santa Cruz Bio- technology) for 1 h at room temperature. The blots were then washed, and specific signals were detected using the ECL chemiluminescence system (Amersham Biosciences) according to the manufacturer���s instructions. The amount of each protein was quantified densitometrically using an Image Scanner (CanoScan D 2400U Canon, Tokyo, Japan) and Bio Image BQ 2.0 software (Bio Image, Ann Arbor, MI). Statistical analysis All values are expressed as the mean sd. The Kruskal-Wallis and Scheffe �����s tests were used for intergroup comparisons of clinical param- eters and serum levels of PlGF, sFlt-1, sFlk-1, adiponectin, leptin, and sEng. The correlations between sEng and several parameters were an- alyzed using Spearman���s rank correlation. Statistical analysis was per- formed with StatView software (Abacus Concepts, Berkeley, CA). P 0.05 was considered statistically significant. Results Characteristics of PE patients and healthy controls When we compared variables (e.g. gestational age, mater- nal age, smoking habit, BMI before pregnancy) between PE and the controls, we found no significant differences. How- ever, mean blood pressure [diastolic blood pressure (sys- tolic blood pressure diastolic blood pressure)/3 mm Hg] in PE was significantly higher than that in healthy pregnant women (Table 1). Serum concentration of sEng and sFlt-1/PlGF ratio To evaluate the gestational pattern of sEng level and sFlt- 1/PlGF ratio, we first examined the serum levels of sEng and the sFlt-1/PlGF ratio in healthy pregnant and nonpregnant Masuyama et al. ��� sEng, sFlt-1, and Adipokines in PE J Clin Endocrinol Metab, July 2007, 92(7):2672���2679 2673