As of today, lipofibroblasts remain as one of the most ill-defined cell populations is the lung. They are part of the lung mesenchyme and contain lipid droplets. Since they are located near alveolar epithelial type II cells, they shuttle the accumulated lipid droplets into AECII cells and support them with surfactant production. Our data revealed that lipofibroblasts start to emerge at E16.5 which is around the time that AECIIs start to appear. FGF10 acts in an autocrine and paracrine fashion on mesenchymal and epithelial cells which are crucial for the formation of LIFs and AECIIs. In the next step we have provided evidence regarding the contribution of lipofibroblasts to the pool of activated myofibroblasts during progression of fibrosis. In brief LIFs transdifferentiate to activated myofibroblasts due to hyperactivity of TGFbeta1 signaling. Following the fate of these cells, our lineage tracing using AdrpCre-ERT2; tdTomatoflox mouse line, confirmed that during resolution phase, activated myofibroblasts, which were resulted from transdifferentiation of lipofibroblasts, revert back to their original status. Activation of PPARg signaling via treatment of cells with rosiglitazone negated the effects of TGFbeta1 treatment and enforced the lipogenic phenotype in the pulmonary fibroblasts. Following this concept, we hypothesized that treatment of the fibrotic lungs with reagents which activate PPARg signaling and enforce lipogenic phenotype in the activated myofibroblasts would yield beneficial effects for the patients. However, since rosiglitazone might not be the best choice due to increased risk of ischemic heart failure in the patients, we set the journey to find other antidiabetic drugs which would manipulate the cellular metabolism of activated myofibroblasts in the favor of lipofibroblasts. Thus, we decided to investigate the potential of metformin in conducting such effects. Treatment of pulmonary fibroblasts with TGFbeta1 (forcefully transdifferentiating them to activated myofibroblasts) followed by metformin resulted in efficient suppression of TGFbeta1 signaling and transdifferentiation to lipofibroblasts. To better mimic the in vivo situation, we cultured fresh PCLSs from human IPF patients and treated them with metformin. The results showed obvious improvement of tissue architecture, reduction of COL1A1 and increasing number of lipofibroblasts. The same readouts were also confirmed via treatment of bleomycin injured ACTA2-CreERT2, tdTomatoflox mice starting at 14 d.p.i. for the next 14 days. Metformin is a well-known agonist of AMPK. While, our data showed that metformin reduces expression of COL1A1 in an AMPK-dependent manner, induction of lipogenic markers such as PLIN2 and PPARg via metformin treatment were independent of AMPK. To explore the molecular mechanism behind this event, the transcriptome of fibroblasts treated with metformin for 72 h was analyzed. KEGG analysis revealed that in addition to alteration of many of the metabolic pathways in favor of inducing lipid droplets, BMP2 was expressed in much higher levels compared to control group. Treatment of fibroblasts with BMP2 induced expression of lipogenic markers and accumulation of lipid droplets. However, BMP2 treatment was not enough to reduce expression of COL1A1. Post translational modification analysis of PPARg following BMP2 or metformin treatment, illustrated a significant increase in PPARg phosphorylation at Ser-112. Hence, metformin suppresses expression of COL1A1 through activation of AMPK, in parallel to induction of lipogenic markers via BMP2-PPARg axis.
Verknüpfung zu Publikationen oder weiteren Datensätzen
