Genes and Immunity (2011) 12, 341–351 & 2011 Macmillan Publishers Limited All rights reserved 1466-4879/11 www.nature.com/gene ORIGINAL ARTICLE Peripheral blood gene expression profiles in metabolic syndrome, coronary artery disease and type 2 diabetes BL Grayson1, L Wang2 and TM Aune1,3 1Department of Microbiology and Immunology, Vanderbilt University School of Medicine, Nashville, TN, USA; 2Department of Biostatistics, Vanderbilt University School of Medicine, Nashville, TN, USA and 3Department of Medicine, Vanderbilt University School of Medicine, Nashville, TN, USA To determine if individuals with metabolic disorders possess unique gene expression profiles, we compared transcript levels in peripheral blood from patients with coronary artery disease (CAD), type 2 diabetes (T2D) and their precursor state, metabolic syndrome to those of control (CTRL) subjects and subjects with rheumatoid arthritis (RA). The gene expression profile of each metabolic state was distinguishable from CTRLs and correlated with other metabolic states more than with RA. Of note, subjects in the metabolic cohorts overexpressed gene sets that participate in the innate immune response. Genes involved in activation of the pro-inflammatory transcription factor, NF-kB, were overexpressed in CAD whereas genes differentially expressed in T2D have key roles in T-cell activation and signaling. Reverse transcriptase PCR validation confirmed microarray results. Furthermore, several genes differentially expressed in human metabolic disorders have been previously shown to participate in inflammatory responses in murine models of obesity and T2D. Taken together, these data demonstrate that peripheral blood from individuals with metabolic disorders display overlapping and non-overlapping patterns of gene expression indicative of unique, underlying immune processes. Genes and Immunity (2011) 12, 341–351; doi:10.1038/gene.2011.13; published online 3 March 2011 Keywords: coronary artery disease; gene expression profiles; gene set; metabolic syndrome; rheumatoid arthritis; type 2 diabetes Introduction tension or raised fasting plasma glucose. The prevalence of MetS in the United States is as high as 39% using the Type 2 diabetes (T2D) is a metabolic disorder of International Diabetes Federation criteria. Diagnosis of peripheral insulin resistance resulting in hyperglycemia MetS confers a 1.5–2.6 relative risk of developing CAD and ultimately decreased insulin secretion from the and a 3.5–7.5 relative risk of developing T2D. Addition- pancreas. Risk factors for T2D include obesity, physical ally, the Framingham study determined that a portion of inactivity and family history of metabolic disorders or these relative risks persist even in the absence of obesity. atherosclerosis.1 Diabetes currently affects 6.3% of the This trio of disorders poses a significant threat to public United States population and B90% of these cases are health in the United States. non-insulin dependent or T2D.2 Coronary artery disease Inflammatory processes are involved in the pathogen- (CAD) results from atherosclerotic plaque development esis of T2D and CAD. Visceral adipose tissue, present in coronary arteries. These fibrous, fatty deposits can in abundance in many patients with T2D, produces ultimately block the flow of blood resulting in angina inflammatory cytokines like interleukin (IL)-6 and tumor and/or myocardial infarction. Hyperlipidemia predis- necrosis factor (TNF)-a´ that are known to aid in the poses to the development of these plaques, making impairment of insulin signaling in adipocytes. These obesity and physical inactivity also risk factors for CAD.3 cytokines can activate a systemic immune response and The prevalence of CAD in the United States is 4.1%.4 recruit inflammatory cells, like lymphocytes, to visceral Metabolic syndrome (MetS 0) is a precursor state to both adipose tissue.14 In the case of CAD, the lesion is not T2D and CAD.5–13 The International Diabetes Federation visceral adipose tissue, but rather fatty deposits in the defines this pre-disease state as central obesity plus any vasculature. These deposits contain fat-laden macro- two of the following four characteristics: hypertriglycer- phages and immunoreactive T cells.3 idemia, low high-density lipoprotein cholesterol, hyper- Gene expression profiling of blood or tissue samples is one way to assess cellular changes due to cell differ- entiation and aging,15,16 disease pathogenesis17–19 or Correspondence: Dr TM Aune, Department of Medicine, Vanderbilt pharmacological responses.20,21 One example of this is University School of Medicine, MCN T-3219, 1161 21st Avenue tumor typing; gene expression signatures are presently South, Nashville, TN 37232, USA. E-mail: [email protected] used to classify tumor types in breast cancer biopsies. Received 7 December 2010; revised and accepted 20 January 2011; This method can also be used to assess changes in published online 3 March 2011 peripheral whole blood of patients with common, Peripheral blood gene expression profiles in MetS, CAD and T2D BL Grayson et al 342 complex diseases.22–24 Individuals with autoimmune Results diseases (type 1 diabetes, multiple sclerosis, systemic lupus erythematosus and rheumatoid arthritis (RA)) Peripheral blood gene expression profiling using micro- display unique gene expression signatures in peripheral arrays has been shown sufficient to distinguish between whole blood. Portions of these signatures are expressed phenotypically distinct cohorts of patients.22–24 We in first degree unaffected relatives,25 however, disease- sought to determine if subjects with MetS, CAD or T2D specific signatures are also found in peripheral blood also possessed a gene expression signature in blood and are sufficient to distinguish individuals with disease sufficient to distinguish these subjects from CTRL from control (CTRL) individuals.26 Moreover, peripheral subjects and, if so, did this signature bear any resem- blood gene expression profiling can give insight into blance to the signature of an autoimmune disease, RA. To disease processes and suggest specific functional defects do so, we recruited subjects with MetS, CAD and T2D in cells. For example, peripheral blood gene expression (n ¼ 6, n ¼ 6, n ¼ 8, respectively), six subjects with RA, in patients with RA contains low transcript levels of the and nine subjects who had never been diagnosed with a tumor suppressor protein, p53. Consequently, T cells chronic illness, and were not presently taking medica- from patients with RA are resistant to g-radiation tions for any diagnosed state, to serve as the CTRL induced apoptosis, a p53 dependent pathway.27 Gene cohort. expression profiling may also aid in diagnosing patients We analyzed all 35 peripheral blood samples for gene who have these often difficult to diagnose diseases; expression using the human exonic evidence-based therefore, analysis of peripheral blood gene expression oligonucleotide array. Next, we normalized the data to represents one approach to assessing immune system a sum total intensity of 10 000, giving an average changes, predicting cellular defects and diagnosing intensity per oligonucleotide probe of 0.2. Genes, with patients with immune-related disease in a minimally an average intensity of greater than 0.2, were used as invasive way. data points for clustering analysis. The intensity values T2D, CAD and their precursor, and MetS are not of the filtered set of genes for each array were inputted autoimmune diseases but feature inflammation as a into The Institute for Genomic Research’s multi-experi- possible pathogenic component. The purpose of our ment viewer. studies was to assess if these diseases also possess unique Initially the samples were imported in sets of CTRL peripheral blood gene expression profiles and if so, what plus one disease or pre-disease state (RA, MetS, CAD or do the profiles indicate about the relationships among T2D, respectively) and then all samples were imported MetS, CAD and T2D. To address this question, we together. Using the Support Tree function, we created a compared profiles of each disease state to CTRL subjects, dendrogram based on unsupervised clustering of each to an autoimmune disease, RA, and to each other. group of samples by similarity (Figure 1). In other words, 100% support 90-100% support 80-90% support RA 01 RA 02 RA 03 RA 04 RA 05 RA 06 70-80% support MetS 03 MetS 01 MetS 02 MetS 04 MetS 05 MetS 06 CTRL 05 CTRL 07 CTRL 03 CTRL 05 CTRL 07 CTRL 03 CTRL 06 CTRL 09 CTRL 08 CTRL 06 CTRL 09 CTRL 01 CTRL 02 CTRL 04 CTRL 01 CTRL 02 CTRL 04 CTRL 08 60-70% support 50-60% support 0-50% support 0% support unrecovered node T2D 06 T2D 04 T2D 03 T2D 05 T2D 01 T2D 02 T2D 07 CAD 03 CAD 05 CAD 01 CAD 02 CAD 04 CAD 06 CTRL 05 CTRL 07 CTRL 03 CTRL 08 CTRL 06 CTRL 09 CTRL 04 CTRL 01 CTRL 02 CTRL 04 CTRL 08 CTRL 02 CTRL 03 CTRL 05 CTRL 07 CTRL 01 CTRL 06 CTRL 09 RA 01 RA 02 RA 03 RA 04 RA 05 RA 06 T2D 06 T2D 05 T2D 04 T2D 01 T2D 02 T2D 08 T2D 03 T2D 07 CAD 03 CAD 05 CAD 01 CAD 02 CAD 04 CAD 06 MetS 02 MetS 04 MetS 03 MetS 05 MetS 01 MetS 06 CTRL 04 CTRL 01 CTRL 06 CTRL 09 CTRL 08 CTRL 02 CTRL 03 CTRL 05 CTRL 07 Figure 1 Unsupervised hierarchical clustering of individual disease cohorts with CTRL. To determine if differential patterns of gene expression could be found among combinations of samples, normalized intensity data points from oligos with an average intensity of X0.20 (average array intensity) were inputted into The Institute for Genomic Research’s Multi-Experiment Viewer. For each comparison, gene intensity averages were calulcated and those X0.20 were selected as input in each comparison. The CTRL vs RA input was 4969 gene and gene splice data points; CTRL vs MetS input was 4225 data points; CTRL vs CAD input contained 4271 data points and the CTRL vs T2D comparison featured an input of 4983 data points.