Expression profiling of human adipose tissue in obese subjects

The molecular mechanisms by which obesity increases the risk of cardiovascular diseases are poorly understood. The purpose of this study was to identify candidate biomarkers overexpressed in adipose tissue of obese subjects that could link expanded fat mass to atherosclerosis. We compared gene expression profile in subcutaneous adipose tissue (scWAT) of 28 obese and 11 lean subjects using microarray technology. This analysis identified 240 genes significantly overexpressed in scWAT of obese subjects. The genes were then ranked according to the correlation between gene expression and body mass index (BMI). In this list, the elastolytic cysteine protease cathepsin S was among the highly correlated genes. RT-PCR and Western blotting confirmed the increase in cathepsin S mRNA (P=0.006) and protein (P<0.05) in obese scWAT. The circulating concentrations of cathepsin S were also significantly higher in obese than in nonobese subjects (P<0.0001). Both cathepsin S mRNA in scWAT and circulating levels were positively correlated with BMI, body fat, and plasma triglyceride levels. In addition, we show that the proinflammatory factors, lipopolysaccharide, interleukin-1β, and tumor necrosis factor-α increase cathepsin S secretion in human scWAT explants. This study identifies cathepsin S as a novel marker of adiposity. Since this enzyme has been implicated in the development of atherosclerotic lesions, we propose that cathepsin S represents a molecular link between obesity and atherosclerosis. Keywords: disease state analysis scWAT samples were obtained, after an overnight fast, by needle biopsy from the peri-umbilical area while the female subjects were under local anesthesia (1% xylocaïne). In a first exploratory phase, differential gene expression of scWAT was compared between 11 nonobese [mean body mass index (BMI) 23.3±0.56 range 19-26 kg/m2] and 28 obese women (mean BMI 40.3±1.41 range 32-60 kg/m2). To analyze adipose tissue gene expression in the first set of lean and obese subjects, we used a common reference pool that was generated by mixing equal amounts of total RNA extracted from adipose tissue samples of patients undergoing plastic surgery. Amplified RNA (aRNA) from the reference pool was labeled with Cy3, and aRNA from the 39 subjects was labeled with Cy5. A total of 39 pangenomic microarray hybridizations were performed for this differential comparison. After scanning the 39 arrays, the images were analyzed and data were normalized using the Stanford Microarray Database Online resources (http://genomewww5. stanford.edu/MicroArray/SMD/). The log2 Cy5/Cy3 ratios of the 39 subjects were extracted. Spots with an average intensity <2.5-fold above the background were eliminated. We filtered genes with signal recovered in at least 8/11 lean and 20/28 obese subject ratios.