Comparative Analysis of Circulating Noncoding RNAs Versus Protein Biomarkers in the Detection of Myocardial Injury

TY - JOUR

T1 - Comparative Analysis of Circulating Noncoding RNAs Versus Protein Biomarkers in the Detection of Myocardial Injury

AU - Schulte, Christian

AU - Barwari, Temo

AU - Joshi, Abhishek

AU - Theofilatos, Konstantinos

AU - Zampetaki, Anna

AU - Barallobre-Barreiro, Javier

AU - Singh, Bhawana

AU - Sörensen, Nils A

AU - Neumann, Johannes T

AU - Zeller, Tanja

AU - Westermann, Dirk

AU - Blankenberg, Stefan

AU - Marber, Michael

AU - Liebetrau, Christoph

AU - Mayr, Manuel

PY - 2019/7/19

Y1 - 2019/7/19

N2 - RATIONALE: Noncoding RNAs (ncRNAs), including microRNAs (miRNAs), circular RNAs (circRNAs), and long noncoding RNAs (lncRNAs), are proposed novel biomarkers of myocardial injury. Their release kinetics have not been explored without confounding by heparin nor has their relationship to myocardial protein biomarkers.OBJECTIVE: To compare ncRNA types in heparinase-treated samples with established and emerging protein biomarkers for myocardial injury.METHODS AND RESULTS: Screening of 158 circRNAs and 21 lncRNAs in human cardiac tissue identified 12 circRNAs and 11 lncRNAs as potential biomarkers with cardiac origin. Eleven miRNAs were included. At low spike-in concentrations of myocardial tissue, significantly higher regression coefficients were observed across ncRNA types compared with cardiac troponins and cMyBP-C (cardiac myosin-binding protein C). Heparinase treatment of serial plasma and serum samples of patients undergoing transcoronary ablation of septal hypertrophy removed spurious correlations between miRNAs in non-heparinase-treated samples. After transcoronary ablation of septal hypertrophy, muscle-enriched miRNAs (miR-1 and miR-133a) showed a steeper and earlier increase than cardiac-enriched miRNAs (miR-499 and miR-208b). Putative cardiac lncRNAs, including LIPCAR (long intergenic noncoding RNA predicting cardiac remodeling and survival), did not rise, refuting a predominant cardiac origin. Cardiac circRNAs remained largely undetectable. In a validation cohort of acute myocardial infarction, receiver operating characteristic curve analysis revealed noninferiority of cardiac-enriched miRNAs, but miRNAs failed to identify cases presenting with low troponin values. cMyBP-C was validated as a biomarker with highly sensitive properties, and the combination of muscle-enriched miRNAs with high-sensitive cardiac troponin T and cMyBP-C returned the highest area under the curve values.CONCLUSIONS: In a comparative assessment of ncRNAs and protein biomarkers for myocardial injury, cMyBP-C showed properties as the most sensitive cardiac biomarker while miRNAs emerged as promising candidates to integrate ncRNAs with protein biomarkers. Sensitivity of current miRNA detection is inferior to cardiac proteins but a multibiomarker combination of muscle-enriched miRNAs with cMyBP-C and cardiac troponins could open a new path of integrating complementary characteristics of different biomarker types.

AB - RATIONALE: Noncoding RNAs (ncRNAs), including microRNAs (miRNAs), circular RNAs (circRNAs), and long noncoding RNAs (lncRNAs), are proposed novel biomarkers of myocardial injury. Their release kinetics have not been explored without confounding by heparin nor has their relationship to myocardial protein biomarkers.OBJECTIVE: To compare ncRNA types in heparinase-treated samples with established and emerging protein biomarkers for myocardial injury.METHODS AND RESULTS: Screening of 158 circRNAs and 21 lncRNAs in human cardiac tissue identified 12 circRNAs and 11 lncRNAs as potential biomarkers with cardiac origin. Eleven miRNAs were included. At low spike-in concentrations of myocardial tissue, significantly higher regression coefficients were observed across ncRNA types compared with cardiac troponins and cMyBP-C (cardiac myosin-binding protein C). Heparinase treatment of serial plasma and serum samples of patients undergoing transcoronary ablation of septal hypertrophy removed spurious correlations between miRNAs in non-heparinase-treated samples. After transcoronary ablation of septal hypertrophy, muscle-enriched miRNAs (miR-1 and miR-133a) showed a steeper and earlier increase than cardiac-enriched miRNAs (miR-499 and miR-208b). Putative cardiac lncRNAs, including LIPCAR (long intergenic noncoding RNA predicting cardiac remodeling and survival), did not rise, refuting a predominant cardiac origin. Cardiac circRNAs remained largely undetectable. In a validation cohort of acute myocardial infarction, receiver operating characteristic curve analysis revealed noninferiority of cardiac-enriched miRNAs, but miRNAs failed to identify cases presenting with low troponin values. cMyBP-C was validated as a biomarker with highly sensitive properties, and the combination of muscle-enriched miRNAs with high-sensitive cardiac troponin T and cMyBP-C returned the highest area under the curve values.CONCLUSIONS: In a comparative assessment of ncRNAs and protein biomarkers for myocardial injury, cMyBP-C showed properties as the most sensitive cardiac biomarker while miRNAs emerged as promising candidates to integrate ncRNAs with protein biomarkers. Sensitivity of current miRNA detection is inferior to cardiac proteins but a multibiomarker combination of muscle-enriched miRNAs with cMyBP-C and cardiac troponins could open a new path of integrating complementary characteristics of different biomarker types.

KW - Artifacts

KW - Biomarkers/blood

KW - Cardiomyopathies/blood

KW - Carrier Proteins/blood

KW - Heparin

KW - Heparin Lyase/pharmacology

KW - Humans

KW - MicroRNAs/blood

KW - Myocardium/chemistry

KW - Plasma/drug effects

KW - RNA, Untranslated/blood

KW - Real-Time Polymerase Chain Reaction

KW - Troponin T/blood

U2 - 10.1161/CIRCRESAHA.119.314937

DO - 10.1161/CIRCRESAHA.119.314937

M3 - SCORING: Journal article

C2 - 31159652

VL - 125

SP - 328

EP - 340

JO - CIRC RES

JF - CIRC RES

SN - 0009-7330

IS - 3

ER -