Abstract 19055: MicroRNA Gene Expression of Heart Transplant Endomyocardial Biopsy

Introduction Endomyocardial biopsy is the standard surveillance method to detect cardiac allograft rejection. While microRNAs (miRNA) play a major role in regulating mRNA, their nature and role in the biology is not well understood. We hypothesized that specific mRNA-miRNA networks can be identified underlying the clinical phenotypes of different forms of cardiac allograft rejection. Method Twenty one tissue samples from 14 post-HTx patients were subjected to genome wide miRNA sequencing. A non-parametric empirical Bayes framework removed batch effect and filtered genes with low variability. Weighted Gene Correlation Network Analysis (WGCNA) clustered genes into related eigengene modules based on their gene expression. Identified miRNAs were subjected to target prediction and compared with mRNA expression profiles previously identified on the same biopsies. Gene Ontology (GO) was used for biological interpretation of selected genes. Results 1270 miRNAs were used to construct 9 eigengene modules. Module-Trait relationship were then investigated as shown in Figure. The top ten miRNA probe sets filtered by the highest intra-module correlation and statistical significance were hsa-miR-141-3p, hsa-miR-150-5p, hsa-miR-605, hsa-miR-582-5p, hsa-miR-3150b-3p, hsa-miR-508-3p, hsa-miR-652-5p, hsa-miR-26a-1-3p, hsa-miR-3667-3p and hsa-miR-3911. Target prediction analysis resulted in 724 gene targets. GO analysis revealed 184 categories enriched by these genes including regulation of protein kinase activity, cardiac muscle cell differentiation and epithelial cell migration among others. Compared to mRNA previously identified in the same heart biopsies showed 685 overlapping gene targets. Conclusion WGCNA identified miRNA modules correlated with different clinical phenotypes of rejection. MRNA-miRNA pairs were identified to help understand the biology of rejection and as interesting candidates for diagnostic or therapeutic applications.