Transcriptional trajectories of human kidney disease progression

Abstract

Our molecular understanding of clinical conditions progressing from acute organ injury to irreversible dysfunction is limited. We used renal transplantation as a model to characterize the transcriptional response along the transition from acute kidney injury to allograft fibrosis in humans. The integrated analysis of 163 transcriptomes with machine learning techniques identified shared and divergent transcriptional trajectories determining distinct clinical outcomes in a heterogeneous population. The molecular map of renal responses to injury was validated in a mouse ischemia-reperfusion injury model and highlighted early markers of disease progression. This generally applicable approach opens the way for an unbiased analysis of progressive diseases.