Smad2 Phosphorylation in Diabetic Kidney Tubule Epithelial Cells Is Associated with Modulation of Several Transforming Growth Factor-beta Family Members

摘要

Background: The role of transforming growth factor-beta (TGF-beta) has recently gained much attention in diabetic nephropathy and kidney fibrosis. In this study, we extend this to an assessment of transcriptional regulation of the entire TGF-beta superfamily in kidneys from diabetic vs. healthy mice. In order to study the translation between mouse model and patients, we evaluated the signature of phosphorylated Sma- and Mad-related protein 2 (pSmad2), as molecular marker of TGF-beta/activin activity, in the kidneys of streptozotocin (STZ)-treated mice compared to that of type 1 diabetes (T1D) patients. Methods: Patterns of pSmad2 were determined in kidneys from T1D patients with progressed diabetic nephropathy (DN), defined by hyperglycemia, microalbuminuria, and increased levels of serum creatinine. They were compared to changes seen in the STZ-induced DN mouse model. This was studied by immunohistochemistry (IHC) with an antibody specific for pSmad2. Diabetic mice were also characterized by pSmad1/5/8 (IHC), pSmad2/3 (flow cytometry), and TGF-beta family members including bone morphogenetic protein (BMP)-like proteins (quantitative real-time polymerase chain reaction [qPCR]). Results: Renal tubules in DN patients and in STZ mice showed upregulation of pSmad2 concomitant with significantly enlarged distal tubule lumens (p 0.0001). Renal-derived CD11b+ cells from STZ mice showed elevated pSmad2/3, while endothelial cells had reduced pSmad2/3 levels. No pSmad1/5/8 was observed in the tubule compartment of STZ-treated mice. On total kidney mRNA level, a signature favoring activation of the TGF-beta/activin pathway and inhibition of the BMP pathway was demonstrated by qPCR. Conclusion: Although the pre-clinical DN model lacks the features of fibrosis present in human DN, both species show induction of a local milieu favoring pSmad2 signaling, which may be useful as a disease biomarker in pre-clinical models. (C) 2017 S. Karger AG, Basel