Characterization of 3T3-L1 adipocytes de-differentiation induced by a mitochondrial dysfunction

(eng) Excessive accumulation of adipose tissue (obesity) implies a health risk worldwide. Even though obesity is not a disease by itself, it leads to various metabolic and cardiovascular disorders. Several evidences have shown that a loss, as small as 5-10% of initial body weight, can improve these health complications. In the field of searching for new therapeutic strategies, targeting adipose tissue by inducing mitochondrial energy dissipation through the uncoupling of mitochondrial oxidative phosphorylation (OXPHOS) is an approach that attracts great attention. However, while it is known that mitochondrial uncoupling triggers triglyceride loss in adipocytes, molecular mechanisms pushing mature adipocyte to lose their capacity to accumulate lipids in a partial de-differentiation process towards a more fibroblast-like phenotype, are still poorly understood. In the first part of this study, we characterized the 3T3-L1-adipocyte de-differentiation exposed to a mild and chronic mitochondrial uncoupling induced by carbonyl cyanide (p-trifluoromethoxy)-phenylhydrazone (FCCP). In the model we set up, we showed that adipocytes incubated for several days with FCCP decrease their triglyceride content to a similar extent than the lipid reduction observed in adipocytes treated with TNFα, a pro-inflammatory cytokine known to act as negative regulator of adipose tissue mass and to increase mitochondrial ROS production in cells. Using a low-density cDNA microarray, we analyzed the expression of genes encoding either adipogenic markers and/or effectors and compared the differentially expressed genes in adipocytes treated with FCCP or TNFα. In addition, we demonstrated that the transcriptional activity of the key transcription factors for adipogenesis: PPARγ and CEBP/α, involved in terminal differentiation and maintenance of the adipocyte phenotype, were significantly decreased in TNFα- and FCCP-treated adipocytes. However, the presence of rosiglitazone and 9-cis retinoic acid (PPARγ/RXR-ligands, respectively) does not prevent triglyceride loss in FCCP-treated cells, suggesting that the TG content reduction observed in adipocytes exposed to a mitochondrial uncoupling is independent of alteration in PPARγ activity. Indeed, we studied several metabolic pathways that form the ‘fatty acid partitioning’ in adipocytes treated with TNFα or FCCP. In addition, metabolic assays also revealed that TG reduction in FCCP-treated adipocytes could be mediated by a down-regulation of lipid synthesis rather than an up-regulation of fatty acid β-oxidation. Finally, lipolysis stimulated by the uncoupler also seems to contribute to the TG reduction, a process associated with the perilipin A down-regulation. These results highlight some new mechanisms that might potentially be involved in adipocyte de-differentiation initiated by a mitochondrial uncoupling. Moreover, adipocytes incubated with FCCP adapt their metabolism in order to maintain constant ATP content, by both reducing energy-consuming processes, such as lipid synthesis but increasing glucose consumption via the stimulation of glycolysis and lactate production. As TNFα or the mitochondrial uncoupling seems to largely affect the expression of many adipogenic genes, in the second part of this work, we screened for modifications in the miRNA expression profile of adipocytes responding to a mitochondrial dysfunction, by analyzing the transcript abundance of several miRNAs known or suspected to play a role in adipogenesis and glucose/lipid metabolism. Indeed, miRNA is now considered as an enlarging class of non-coding mRNAs that control gene expression in many biological processes. However, their contribution to metabolic disorders, lipid metabolism and mitochondrial dysfunction is still poorly studied, and their role is thus puzzled. We found that let-7b/c, miR-103, miR-107, miR-182, miR-422b and miR-150 are differentially expressed in adipocytes incubated with either FCCP or TNFα. In addition, some miRNA seem to be specifically regulated, as miR-155 and miR-222 are more abundant in adipocytes challenged with TNFα, while no modification in their abundance could be found in FCCP-treated cells. Unfortunately, we were unable to address the functional role of these miRNAs differentially expressed in adipocytes exposed to a mild mitochondrial uncoupling. Indeed, we could not show that miRNA interferences (either ectopic over-expression or inhibition of endogenous miRNAs) of miR-103 and miR-107, two miRNA paralogs, predicted to target several mRNAs involved in pathways that affect lipid metabolism, modify the triglyceride content in adipocytes or adipocytes incubated either with FCCP or TNFα. In conclusion, our results suggest that the mechanisms and cell signaling pathways leading to the reduction of TG content are different in adipocytes incubated with either TNFα or FCCP. Moreover, 3T3-L1 cells incubated for one or more days in the presence of TNFα or FCCP must adapt their metabolism and change the expression or activity of several genes encoding effectors or markers of differentiation. In addition, these adaptations of metabolism, found in FCCP-treated adipocytes, are reflected by modifications in the energy homeostasis of the cells. We also demonstrated that mitochondrial uncoupling in white adipocytes might represent a putative and interesting target to induce TG content reduction, and to potentially reduce adipose mass. Hopefully, these findings may contribute to a better understanding of the molecular mechanisms involved/associated or responsible for the adipocyte cell response to a mild but chronic mitochondrial uncoupling.