Introduction
Bone morphogenetic protein 10 (BMP10) is gaining attention as a blood biomarker for predicting rhythm in atrial fibrillation (AF), AF recurrence, and AF-associated cardiovascular complications such as stroke. BMP10 is a ligand of the TGF-beta superfamily, primarily secreted by cardiomyocytes; however, the regulatory mechanisms of BMP10 secretion and its downstream effects in cardiomyocytes remain largely unclear. This study aimed to investigate the secretion dynamics and effects of BMP10 in cardiomyocytes using a human in vitro 3D engineered heart tissue (EHT) model.
Methods
BMP10 serum levels were analyzed across plasma samples from patients presenting different atrial rhythms at the time of blood draw (n=254–814/group). For in vitro human studies, cardiomyocytes differentiated from human induced pluripotent stem cells (hiPSC, both atrial and standard ventricular-like) were embedded in a fibrin matrix to produce EHT. Atrial EHT were either optogenetically paced (3–5 Hz) or kept at their intrinsic beating rate (~2 Hz) for periods ranging from 24 hours to 15 days (n=3–5/group). Sequential EHT media samples were analyzed via ELISA for BMP10 and other cardiac biomarkers. Ventricular EHT (vEHT) and atrial EHT (aEHT) were treated with recombinant human BMP10 (rhBMP10) for 10 days and subjected to RNA sequencing (n=5/group). Additionally, the electrophysiological effects of rhBMP10 were tested in ventricular hiPSC-CMs using the voltage-sensitive dye FluoVolt (n=6/group).
Results
BMP10 plasma levels showed a significant increase across the three patient categories: patients in sinus rhythm (2.07±0.02 ng/mL), patients diagnosed with AF but in sinus rhythm during blood draw (2.20±0.04 ng/mL), and patients with AF at the time of blood draw (2.77±0.05 ng/mL). In the human in vitro model, BMP10 was secreted by atrial but not vEHT within 48 hours of culture. Continuous optogenetic pacing at high rates (3–5 Hz) led to approximately a 3-fold increase in BMP10 release from atrial EHT, with a delay of at least 24 hours after pacing began; in contrast, troponin-I and NT-proBNP levels rose within the first 24 hours. RhBMP10 exposure resulted in upregulation of transcripts involved in the TGF-beta pathway and gene ontology (GO) networks such as “Atrial fibrillation,” “ECG morphology,” and “PR interval” in vEHT, along with increased expression of atrial genes including PITX2 and NPPB. Furthermore, rhBMP10 influenced ventricular CM function and electrophysiology by prolonging contraction time-to-peak and relaxation, and shortening the action potential duration. RhBMP10 also enhanced contraction force of aEHT.
Conclusion
Atrial EHT composed of hiPSC-derived cardiomyocytes express and secrete BMP10. Elevated atrial rates promote BMP10 gene expression and active release both in vitro and in vivo. High concentrations of rhBMP10 induce AF-associated transcriptomic networks in vEHT and affect cardiomyocyte function. These results support the potential of BMP10 as a biomarker for atrial fibrillation and reveal a functional role for BMP10 in AF-related cardiomyopathy.
