Key driver genes as potential therapeutic targets in renal allograft rejection
Zhengzi Yi, Karen L. Keung, Li Li, Min Hu, Bo Lu, Leigh Nicholson, Elvira Jimenez-Vera, Madhav C. Menon, Chengguo Wei, Stephen Alexander, Barbara Murphy, Philip J. O’Connell, Weijia Zhang
Zhengzi Yi, Karen L. Keung, Li Li, Min Hu, Bo Lu, Leigh Nicholson, Elvira Jimenez-Vera, Madhav C. Menon, Chengguo Wei, Stephen Alexander, Barbara Murphy, Philip J. O’Connell, Weijia Zhang
View: Text | PDF
Research Article Transplantation

Key driver genes as potential therapeutic targets in renal allograft rejection

Abstract

Acute rejection (AR) in renal transplantation is an established risk factor for reduced allograft survival. Molecules with regulatory control among immune pathways of AR that are inadequately suppressed, despite standard-of-care immunosuppression, could serve as important targets for therapeutic manipulation to prevent rejection. Here, an integrative, network-based computational strategy incorporating gene expression and genotype data of human renal allograft biopsy tissue was applied, to identify the master regulators — the key driver genes (KDGs) — within dysregulated AR pathways. A 982–meta-gene signature with differential expression in AR versus non-AR was identified from a meta-analysis of microarray data from 735 human kidney allograft biopsy samples across 7 data sets. Fourteen KDGs were derived from this signature. Interrogation of 2 publicly available databases identified compounds with predicted efficacy against individual KDGs or a key driver–based gene set, respectively, which could be repurposed for AR prevention. Minocycline, a tetracycline antibiotic, was chosen for experimental validation in a murine cardiac allograft model of AR. Minocycline attenuated the inflammatory profile of AR compared with controls and when coadministered with immunosuppression prolonged graft survival. This study demonstrates that a network-based strategy, using expression and genotype data to predict KDGs, assists target prioritization for therapeutics in renal allograft rejection.

Authors

Zhengzi Yi, Karen L. Keung, Li Li, Min Hu, Bo Lu, Leigh Nicholson, Elvira Jimenez-Vera, Madhav C. Menon, Chengguo Wei, Stephen Alexander, Barbara Murphy, Philip J. O’Connell, Weijia Zhang

×

Figure 4

KDGs in AR differential modules.

Options: View larger image (or click on image) Download as PowerPoint
KDGs in AR differential modules. Each colored node represents a meta-gen...
Each colored node represents a meta-gene, and the KDGs are represented with yellow text. The color of each node represents its effect size, reflecting its relative expression level in AR as depicted by the color scale bar at the bottom. The arrows signify the direction of the relationship between 2 genes as predicted by Bayesian network, such that the arrow tip points toward the gene that is downstream of the other. Width of lines indicates strength of correlation between 2 nodes. Due to the size of the modules, only the first layer of downstream genes that were directly connected to key drivers in M1, M3, and M4 are illustrated, although all module members are shown for M8 and M10. In M1, the largest of the modules, since CASP1 was not downstream to any other KDGs in its module, it was identified as the global driver here. Furthermore, CASP1 also had the largest number of out-degrees and downstream gene count of all the KDGs. M7, although regarded as a differential module from connectivity permutation test between AR and non-AR, did not yield a KDG because none of its members satisfied the downstream counts and out-degrees threshold criteria.