Right ventricular myocardial biomarkers in human heart failure

Abstract

Background: Right ventricular (RV) dysfunction contributes to mortality in chronic heart failure (HF). However, the molecular mechanisms of RV failure remain poorly understood, and RV myocardial biomarkers have yet to be developed.

Methods and results: We performed RNA sequencing (RNA-seq) on 22 explanted human HF RVs and 5 unused donor human heart RVs (DON RV) and compared results to those recently reported from 16 explanted human LVs We used Bowtie-Tophat for transcript alignment and transcriptome assembly, DESeq for identification of differentially expressed genes (DEGs) and Ingenuity for exploration of gene ontologies. In the HF RV, RNA-seq identified 130,790 total RNA transcripts including 13,272 protein coding genes, 10,831 long non-coding RNA genes and 8,605 pseudogenes. There were 800-1000 DEGs between DON and HF RV comparison groups with differences concentrated in cytoskeletal, basement membrane, extracellular matrix (ECM), inflammatory mediator, hemostasis, membrane transport and transcription factor genes, lncRNAs and pseudogenes. In an unbiased approach, the top 10 DEGs SERPINA3, SERPINA5, LCN6, LCN10, STEAP4, AKR1C1, STAC2, SPARCL1, VSIG4 and F8 exhibited no overlap in read counts between DON and HF RVs, high sensitivities, specificities, predictive values and areas under the receiver operating characteristic curves. STEAP4, SPARCL1 and VSIG4 were differentially expressed between RVs and LVs, supporting their roles as RV-specific myocardial biomarkers.

Conclusions: Unbiased, comprehensive profiling of the RV transcriptome by RNA-seq suggests structural changes and abnormalities in inflammatory processes and yields specific, novel HF RV vs HF LV myocardial biomarkers not previously identified by more limited transcriptome profiling approaches.

Keywords: Human heart failure; RNA-seq; biomarkers; myocardial; right ventricle; transcriptome.

Fig. 1.

Read count ranges for selected biomarkers. The lowest and highest read counts in the donor (DON) right ventricles (RVs; n = 5) vs heart failure (HF) RVs (n = 22) are presented for SERPINA3, SERPINA6, LCN6, LCN10, STEAP4, ANP, BNP, Gal3, ST2, and Cystatin C. Top row: for these 5 established biomarkers, there was no overlap in RV myocardial read ranges. Bottom row: for these 5 established biomarkers, there was overlap in RV myocardial read ranges.

Fig. 2.

Read counts for NPPA, NNPB, LGALS3, ST2, and CST3. The histograms show adjusted counts for NPPA, NNPB, GAL3, ST2, and CST3 for all 22 heart failure patients. There were substantial differences in counts for all 5 different transcripts between heart failure right ventricles. As discussed in the text, neither clinical characteristics nor NPPA promoter–associated transcription factors accounted for the variation in NPPA read counts between subjects.

Fig. 3.

Panel A displays the sensitivity and specificity for established heart failure biomarkers. Panel B displays the area under the received operating characteristic (ROC) curve (AUC) performance for established heart failure biomarkers compared to SERPINA3 AUC. Panel C displays the sensitivity, specificity, positive and negative predictive values (PPV and NPV, respectively) and ROC areas for established heart failure biomarkers with the cut-off values as shown. *P < .05 testing for difference from SERPINA3 AUC. PPV, positive predictive value; NPV, negative predictive value.

Fig. 4.

Ingenuity “Cardiovascular Disease, Organismal Injury, and Abnormalities” network results and predictions for potential novel right ventricular biomarkers. Given the observed down-regulation of SERPINA3, SERPINA5, F8, VSIG4, SPARCL1, STEAP4, and AKRIC1 (green; see inset Prediction Legend), Ingenuity predicts inhibition of TNF, TGF-b1, APP, CTSG, SYP, KLK6, chymotrypsin, F2, F11, NADPH oxidase, and AChR (blue) and activation of CTF1, serine protease, allopregnanolone, and FIRSP12 (brown). (Color version of the figure is available online.)

Fig. 5.

Ingenuity “Coagulation System” and “Acute Phase Response Signaling” canonical pathway effects identified from novel right ventricular biomarkers as inputs. (A) Ingenuity “Coagulation System” canonical pathway. Given the observed down-regulation of SERPINA5 and F8, Ingenuity predicts activation of the F3/F5a/F7a/F10a complex, F2a, Protein C, Fibrin, F13a, and Fibrin cross-link polymer (brown). (B) Ingenuity “Acute Phase Response Signaling” canonical pathway. Given the observed down-regulation of SERPINA5 and F8, Ingenuity predicts widespread inhibition of a variety of acute phase signaling effectors, including STAT3, NF-κB, c-JUN, NF-IL6, glucocorticoid receptors, and corticosteroid receptors (blue). (Color version of the figure is available online.)