Stoichiometry of transcription factors is critical for cardiac reprogramming.

D irect reprogramming of fibroblasts into cardiomyocytes holds great potential for cardiovascular disease research and treatment. This new technology may be used for patient-specific drug screening, cardiac disease modeling, and regen-erative purposes. We reported first that a combination of 3 cardiac-specific transcription factors, Gata4 (G), Mef2c (M), and Tbx5 (T), constituted the minimum requirement to directly reprogram mouse cardiac fibroblasts (CFs) into induced cardiomyocytes (iCMs). 1 Subsequently, multiple groups have achieved and improved direct cardiac reprogramming from mouse and human fibroblasts by overexpressing other combinations of cardiac transcription factors and cardiac-enriched miRNAs. 2–9 Intriguingly, it was reported that gene delivery of reprogramming factors into mouse infarcted hearts converted endogenous CFs into functional iCMs, reduced scar size, and improved cardiac function after myocardial infarction. Although these recent achievements are promising, the low reprogramming efficiency of fully reprogrammed functional iCMs and the reproducibility of cardiac reprogramming continue to be controversial aspects of this technology. Chen et al 12 showed that transduction of pooled lentiviruses expressing G, M, and T was insufficient to reprogram fibroblasts into a cardiac fate. We reported that a polycistronic vector and a mixture of individual vectors expressing G, M, and T could reprogram resident CFs into iCMs in vivo, but the efficiency of cardiac reprogramming in this study was lower than that in other reports. 13 These findings suggest that slight technical or biological differences might result in variable reprogramming efficiency. 14 Therefore, critical factors for cardiac reprogramming need to be identified, and standardized platforms for efficient and reproducible cardiac induction should be established to advance this field. In this issue, Wang et al 15 generated a complete set of polycistronic constructs encoding G, M, and T with all possible splicing orders and analyzed the cardiac reprogramming efficiency using these vectors. They found that each polycistronic vector gave rise to distinct G, M, and T protein expression levels, and that an optimal balance of G, M, and T protein expression greatly improved both the efficiency and the quality of cardiac reprogramming. Previous studies mainly used a mixture of viruses expressing individual factors for cardiac reprogramming. This approach has heterogeneous and uncontrollable ratios of reprogramming factor expression among infected fibroblasts, which may lead to variable and low reprogramming efficiency. To overcome this issue, we previously developed the polycistronic construct TMG to express each factor in a homogeneous ratio and to improve in vivo reprogramming, but the optimal stoichi-ometry of …