Identificación y caracterización funcional de microarns

Resumen

Background and aim Chronic inflammation has an important role in the pathogenesis of obesity and type 2 diabetes mellitus (T2DM). Obese adipose tissue is characterized by an increase in macrophage infiltration and by an increment in releasing of adipokines and inflammatory cytokines, such as MCP-1, TNFα and IL-6, not only by the adipocyte but also by adipose tissue macrophages. The inflammatory environment in adipose tissue impairs insulin signaling, which leads to local and systemic insulin resistance. Tumor Necrosis Factor alpha (TNFα) has been identified as one of the responsible agents for the development of insulin resistance during obesity. In fact, chronic exposure of adipocytes to low concentrations of TNFα dramatically reduces glucose uptake upon insulin stimulation. It has already been described that TNFα directly interferes with insulin signaling by inhibiting the auto-phosphorylation of its receptor and the phosphorylation of its substrate (IRS-1), and consequently blocking the biological actions of insulin. However, the mechanisms involved have only been partially described. MicroRNAs (miRNAs) are highly-conserved non-coding RNA molecules (19-21 base pairs) of endogenous origin. miRNAs negatively regulate gene expression, either by degrading mRNA transcripts or by repressing protein translation. In previous studies, altered expression of miRNAs in insulin-sensitive tissues (liver, muscle, endothelium or adipose tissue, among others)of obese and/or diabetic patients has been described, suggesting that miRNAs play an important role in these pathologies. Moreover, miRNAs have recently emerged as main regulators of metabolism. Therefore, the deregulation of miRNA expression may contribute to metabolic abnormalities. All these data indicate that miRNAs could become potential therapeutic targets to alleviate or improve complications from metabolic disorders such as obesity and T2DM. In addition, of the presence of circulating miRNAs in biological fluids, such as serum, has revealed their potential role as signaling molecules and biomarkers of endocrine diseases. Changes in miRNA expression in response to abnormal levels of cytokines can cause alterations in cellular homeostasis. In this regard, TNFα is directly involved in the modulation of several miRNAs expression in different cell types, for example immune cells or endothelial cells. While the modulation of miRNAs expression by TNFα in adipose tissue has been widely described, very little is known about their role in regulation of the insulin pathway in mature adipocytes. Therefore, the underlying mechanisms that connect miRNAs and TNFα-induced insulin resistance (IR) in the mature adipocyte of obese adipose tissue remain to be deciphered. Methodology The expression of 25 selected miRNAs was determined in human adipocytes after an inflammatory stimulus induced by TNFα, as an IR adipocyte model. The expression of miR-181a-5p, miR-23a-3p and miR-155-5p was analyzed by qRT-PCR from RNA extracted from paired biopsies of subcutaneous (SAT) and visceral (VAT) adipose tissue. Adipose tissue biopsies correspond to an age- and sex-matched cohort of patients stratified according to their body mass index (BMI) <30 (N = 30) and BMI ≥30 (N = 28), and different glucose tolerance status. To evaluate the function of both miR-181a-5p and miR-23a-3p in the insulin pathway, we transiently overexpressed them using mimic-miRNAs Mimics in mature SGBS (Simpson-Golabi-Behmel Syndrome) adipocytes, which we then stimulated with insulin or insulin plus TNFα. We then measured phosphorylated (p)AKT and AKT substrate ok 160 kDa (pAS160) levels as a surrogate measure of insulin signaling. Using target prediction algorithms for miRNAs (miRanda-mirSVR and microT-CDS) we searched for insulin pathway regulators potentially modulated by miR-181a-5p and miR-23a-3p. Predicted targets were validated by luciferase reporter assay in HEK293 cells; and by measuring the endogenous protein levels of PTEN and S6K in SGBS adipocytes transfected with mimics or miRNA inhibitors. In addition, to assess the potential value of miR-181a-5p and miR-23a-3p as prognostic markers, we measured their levels in serum in a cohort of 101 normoglycemic subjects that were followed for 4 years, with 48 diagnosed preadiabetic and 53 remaining normoglycemic. We also selected 100 age-, sex- and BMI-matched patients, from the Di@bet.es study (50 controls and 50 subjects with T2DM), in order to check circulating levels of miR-181a-5p and miR-23a -3p in serum of diabetic patients. Results We have observed that TNFα significantly affects expression of 3 from 25 miRNAs selected in our study (miR-181a-5p, miR-23a-3p and miR-155-5p) in human adipocytes. miR-155 is a well known miRNA induced by inflammatory stimuli, but reduction of miR-181a-5p and miR-23a-3p by TNFα in mature human adipocytes has been revealed for the first time in our study. When we analyzed miRNA expression in human adipose tissue, we observed that miR-181a-5p, miR-23a-3p and miR-155-5p were differentially overexpressed in VAT, with respect to subcutaneous adipose tissue (SAT), indicating that those miRNAs could be actively involved in VAT function, which is closely linked to obesity-related comorbidities. The expression of both miR-181a-5p and miR-23a-3p were significantly reduced in VAT from BMI≥30 individuals compared to controls. Interestingly, we observed a lower expression of both miRNAs in subjects who, in addition to being obese, had also developed T2DM compare to diabetic patients with BMI<30; pointing a combined effect of comorbidity (obesity + T2DM). In addition, we found that expression of miR-181a-5p and miR-23a-3p in the VAT was inversely related to HOMA- IR, blood glucose levels and adiposity. Moreover, we observed that TNFα expression levels, which were increased in VAT, correlated inversely with expression levels of both miRNAs. These data suggested a possible role for miR-181a-5p and miR-23a-3p in IR in the obese adipose tissue. Regarding miR-155-5p, we did not detect differences of its expression neither in SAT nor in VAT according to BMI and glucose tolerance status. We also determined the functional role of miR-181-5p and miR-23a-3p in the insulin pathway in adipocytes. We observed that overexpression of miR-181a-5p and miR-23a-3p in human adipocytes improves insulin sensitivity by increasing pAKT and pAS160 protein levels. In addition, we observed a significant protective effect in TNFα-induced IR when both miRNAs miR-181a-5p and miR-23a-3p were co-transfected. PTEN and S6K are negative regulators of insulin signaling and they were identified as the best shared predicted targets for miR-181a-5p and miR-23a-3p. Target validation by luciferase reporter assays in HEK293 and transfection with mimics and inhibitors in SGBS adipocytes confirmed the presence of functional targets in the 3'UTR regions of PTEN and S6K for both miR-181a-5p and miR-23a-3p. Circulating miRNAs are attractive candidates as biomarkers for non-invasive diagnosis and monitoring due to their great stability and easy detection in liquid biopsies. Given the functional role of these miRNAs in IR, we check potential value of miR-181a-5p and miR-23a-3p as prediabetes prognostic biomarkers in the cohort of the prospective PIZARRA study. We observed that the levels of both miRNAs were reduced in serum of prediabetic subjects compared to controls, but only miR-181a-5p was already reduced at baseline, pointing its value. To evaluate the usefulness of circulating miR-181a-5p as a potential prognostic biomarker of prediabetes status, we performed a Receiver operating characteristic (ROC) curve analysis. This analysis did not reveal a robust predictive power for miR-181a-5p alone, with an AUC of 0.633. However, when we performed a multivariate logistic regression analysis including miR-181a-5p along with other serological signatures, which also changed between cases and controls at baseline (HDL cholesterol, C reactive protein and Adiponectin), we got a better ROC curve (AUC = 0.793). This model was able to correctly classify 72.3% of the patients with significantly higher specificity (72.9%). We then applied a Partial Least Squares Discriminant Analysis (PLS-DA) to evaluate the potential of miR-181a-5p in stratification of patients. The model was constructed using anthropometric, clinical and biochemical variables. Cross-validation analysis showed that at the first component had a low accuracy of 54.9%. With regards to the importance in this model, miR-181a-5p ranked fourth in importance, behind HDL Cholesterol, C-reactive protein (CRP) and adiponectin. Thus, although the diagnostic accuracy is moderate for all the indexes we examined, the four serum signatures (HDL cholesterol, CRP, adiponectin and miR-181a-5p) could represent a potential biomarker panel that allows an early diagnostic of prediabetic patients. Conclusions The expression of miR-181a-5p and miR-23a-3p is decreased in VAT of obese patients (BMI≥30); this reduction is more accentuated with the presence of T2DM. The expression of miR-181a-5p and miR-23a-3p are inversely related to adiposity, independently of the fat depot, and the presence of IR. miR-181a-5p and miR-23a-3p have a possible regulatory role in obesity and IR induced by TNFα in human adipocytes. Circulating levels of miR-181a-5p are deregulated before the onset of prediabetes indicating a possible prognostic role in this disease.