Long Non-coding RNA Expression Profiling in Biopsy to Identify Renal Allograft at Risk of Chronic Damage and Future Graft Loss

Jing Xu,Jinglei Hu,Heng Xu,Honghao Zhou,Zhaoqian Liu,Yong Zhou,Rong Liu,Wei Zhang
DOI: https://doi.org/10.1007/s12010-019-03082-2
2019-01-01
Applied Biochemistry and Biotechnology
Abstract:The loss of allograft from chronic damage is still the major risk that renal transplant recipients face today. Biomarkers for early detection of chronic damage are needed to improve the long-term graft survival. This study aimed to identify long non-coding RNA (lncRNA) biomarkers associated with chronic damage and graft loss after renal transplantation. Gene Expression Omnibus (GEO) datasets including GSE57387 ( n = 101), GSE21374 ( n = 282), and GSE25902 ( n = 24) from three high-quality studies were analyzed. By repurposing the publicly available array-based data coupled with Affymetrix Human Exon 1.0 ST and Human U133 Plus 2.0 arrays, we obtained expression profiles of 11323 and 3383 lncRNAs in biopsies after renal transplantation, respectively. The logistic regression model and Cox regression model were applied to identify lncRNAs associated with chronic damage and graft survival. High AC093673.5 expression was identified as significantly associated with the three endpoints including chronic damage, progressive chronic histological damage, and graft failure across these three datasets. A six-lncRNA signature was created to predict renal allograft at risk of chronic damage with a high predictive ability (AUC = 0.94). Gene set enrichment analysis (GSEA) indicated that our lncRNA signature was related with allograft rejection and immunity. Our study highlights the importance of lncRNAs in chronic graft damage and allograft loss, supporting their potential role as prognosis biomarkers.
What problem does this paper attempt to address?