Direct cardiac reprogramming was a pioneering discovery when it was first reported in 2010, but the cardiomyocyte induction efficiency at the time was low and needed improvement. However, since studies on direct cardiac programming were first published, various approaches have been explored to improve its efficiency. These approaches can be broadly divided into (1) improving direct cardiac reprogramming factors, (2) improving cell culture conditions, and (3) regulating epigenetic factors. Previously unknown factors that inhibit reprogramming have been identified, and next-generation sequencing has elucidated the role of these individual factors in the reprogramming process.
Improving direct cardiac reprogramming factors
microRNA (miRNA) has been explored as an alternative approach to improving induction efficiency by including transcription factors with GMT, such as GHMT and GMT + Mesp1 and Myocd. miRNAs are small RNA molecules of approximately 21–25 bases that suppress translation by binding to target messenger RNA (mRNA) and play an important role in determining cell fate. Jayawardena reported that directly inoculating mouse hearts with a lentiviral vector hosting four cardiogenic miRNAs (miR-1, miR-133, miR-208, and miR-499) after myocardial infarction directly reprogrammed cardiac fibroblasts into cardiomyocytes and improved cardiac function [18]. We reported that adding these four miRNAs individually and simultaneously to GMT resulted in faster and more efficient cardiomyocyte induction, especially with miR-133 added to GMT, compared to treatment with GMT alone. In human fibroblasts, adding miR-133 to the human cardiomyocyte induction factor GMTMM improved the reprogramming efficiency by approximately 10-fold. miR-133 represses fibroblastic genes from an early point in iCM induction. miR-133 represses Snai1, a key gene in this pathway and a master regulator of the epithelial-to-mesenchymal transition [19]. This discovery established that iCM induction can be enhanced by suppressing fibroblastic traits and laid the groundwork for numerous later discoveries.
Investigating culture conditions that enhance direct cardiac reprogramming
Direct reprogramming of fibroblasts into iCMs requires the elimination of fibroblast traits. Zhao et al. found that intracellular and extracellular signaling pathways work to maintain the functions of a cell, and in the case of fibroblasts that possess high chemotactic ability and proliferation potential, TGF-β (transforming growth factor-β), Wnt, and ROCK (Rho-associated coiled-coil-containing protein kinase) cytokines are involved in activating their chemotactic ability and proliferation potential. Direct cardiac programming is enhanced by inhibiting the TGF-β and ROCK signaling pathways with low-molecular-weight molecules. Inhibiting the TGF-β and ROCK pathways significantly improves reprogramming efficiency and also allows for the earlier formation of iCMs with a spontaneous beat [20]. Among the 5500 compounds screened for identifying signaling pathways that enhance the efficiency of direct cardiac reprogramming, inhibitors of TGF-β and Wnt signals improved the efficiency of induction with GMT [21].
Signaling pathways that are related to cardiac development and cardiac hypertrophy also enhance direct cardiac reprogramming. Eliciting Akt1 overexpression by adding Akt1 to GHMT resulted in enhanced cardiac reprogramming efficiency and the induction of mature cardiomyocytes [22]. Cardiac reprogramming was also enhanced by repressing Notch signaling. Mef2c transcription is suppressed by Notch signaling, and the addition of DAPT ((S)-tert-butyl 2-((S)-2-(2-(3,5-difluorophenyl) acetamido) propanamido)-2-phenylacetate), a Notch signaling pathway inhibitor, enhanced direct cardiac reprogramming [23]. Our group also attempted to optimize culture media for direct cardiac reprogramming by adding cell growth factors and other cytokines to serum-free media. A combination of fibroblast growth factor (FGF)-2, FGF-10 and vascular endothelial growth factor (VEGF: FFV) enhanced the induction efficiency of beating iCMs approximately 40-fold compared to conventional culture methods. The mechanism behind this effect was the activation of the PI3K/Akt and p38MAPK pathways by cell growth factors, which upregulated the gene clusters involved in cardiac function [24].
Enhancing direct cardiac reprogramming by regulating epigenetic factors
Cells regulate gene transcription and decide the fate of cell differentiation by DNA methylation and histone modification during the differentiation process. The epigenetics of target genes are also presumed to change substantially during the process of direct reprogramming of fibroblasts into cardiomyocytes. In 2010, we reported that introducing GMT into fibroblasts caused the demethylation of Nppa and Myh6, cardiogenic genes that are methylated in fibroblasts. Histone H3 lysine 4 trimethylation (H3K4me3: active modification) and histone H3 lysine 27 trimethylation (H3K27me3: repressive modification) are two of the most widely recognized histone modifications in cardiogenic and fibroblastic genes [25]. Introducing GMT into fibroblasts increased H3K4me3 and reduced H3K27me3 in cardiogenic gene regions, but decreased H3K4me3 and increased H3K27me3 in fibroblastic gene regions. GMT together with short hairpin RNA (shRNA) for genes involved in histone modifications was introduced into fibroblasts and assessed for epigenetic factors that regulate direct cardiac reprogramming. Bmi1, a component of a polycomb protein complex (PRC1) that causes mono-ubiquitination of histone H2A on lysine 119 (H2AK119ub: repressive modification), was found to epigenetically inhibit the expression of cardiogenic genes. Reducing Bmi1 expression altered the histone modification of cardiogenic genes, resulting in a less repressed state. This modification enhanced the reprogramming efficiency, suggesting that Bmi1 is an epigenetic barrier to reprogramming [26].
Pioneer transcription factors have attracted much attention because they provide important insights into the mechanisms responsible for binding to nucleosomes and subsequent transcriptional regulation [27]. In addition, finding new pioneer transcription factors in cardiac direct reprogramming may help to gain efficiency.
Age- and inflammation-related suppression of direct cardiac reprogramming
Fibroblastic features and epigenetics have both been identified as barriers to direct cardiac reprogramming. Older neonatal or adult fibroblasts also transform into cardiomyocytes less efficiently than immature embryonic fibroblasts [20, 28]. The underlying mechanism is unknown, and clinical applications would favor efficient cardiomyocyte induction from neonatal or adult fibroblasts. We therefore searched for compounds that enhance cardiomyocyte induction from these fibroblasts.
A library of 8400 chemical compounds was screened for compounds that enhance cardiomyocyte induction, and four compounds that enhance cardiomyocyte induction from neonatal fibroblasts were identified. Further screening revealed that cardiomyocyte induction was enhanced significantly by diclofenac, commonly known as a non-steroidal anti-inflammatory drug (NSAID), to GHMT [29]. Adding diclofenac to GHMT upregulated the expression of cardiogenic proteins and created a distinct striated muscle pattern specific to the heart. Adding diclofenac also increased the number of cardiomyocytes that exhibit spontaneous beating, a characteristic feature of more mature cardiomyocytes, by approximately fourfold. This observation indicated that diclofenac not only enhances cardiomyocyte induction, but also promotes the maturation of the induced cardiomyocytes.
The enhancement of cardiomyocyte induction by diclofenac was specific to neonatal and adult fibroblasts and was absent in more immature fetal fibroblasts. Fibroblast gene expression changes with age, and there is an increased signaling of cyclooxygenase (COX)-2, prostaglandin E2 (PGE2), and PGE2 receptor EP4 in the arachidonate cascade with increasing fibroblast age. This signaling pathway also induced inflammatory and fibrosing genes further downstream via COX-2/PGE2/EP4/IL-1β/IL-1R1 signaling, which suppressed reprogramming-based cardiomyocyte induction. Diclofenac enhances cardiomyocyte induction by repressing this pathway, and age and inflammation act as barriers to reprogramming (Fig. 3).
Wang L et al. found that Beclin1 (Becn1), an autophagy factor, suppressed the induction of iCM, in a pathway unrelated to autophagy. Conversely, deletion of Becn1 resulted in the high efficient induction of iCM from mouse and human fibroblasts [30].
The role of transcription factors in direct reprogramming
The quantitative balance among the transcription factors introduced during direct cardiac reprogramming has a major effect on the efficiency of cardiomyocyte induction. Wang et al. reported that high levels of Mef2c expression and low levels of Gata4 and Tbx5 expression were important for enhanced cardiac reprogramming [31]. Next-generation sequencing has enabled the elucidation of the role of each transcription factor in the direct reprogramming process. Hashimoto et al. revealed that fibroblastic genes were repressed, and cardiogenic genes were induced, from early stages of the reprogramming process, using ChIP-seq analysis of the enhancer regions of fibroblastic and cardiogenic genes during reprogramming with GMT [32]. These enhancer regions contain a high frequency of Mef2c binding sites, and the Gata4 and Tbx5 remaining at the Mef2c binding sites probably induce and regulate cardiogenic genes synergistically from early in the reprogramming process. Adding Hand2 and Akt1 results in binding at new enhancer regions, in addition to the previous enhancer regions, and therefore enhances the efficiency of induction [32]. Stone and colleagues reported that Mef2c and Tbx5 caused epigenetic changes between 24 and 48 h after the start of direct cardiac reprogramming by GMT, early in the reprogramming process [33]. Future advances in analytical technology will help to elucidate the molecular mechanisms related to reprogramming, and enable further improvements in direct programming.
The role of mechanobiology in direct reprogramming
Mechanobiology is believed to be involved in cardiac embryology and cardiac diseases. However, its role in direct cardiac reprogramming remains to be elucidated. The extracellular environment is a key factor in the success or failure of reprogramming. The polystyrene culture dishes in which cells are cultured ex vivo have a hardness of 1 GPa (= 1 × 106 kPa), which is harder than the cartilage tissue (~ 100 kPa) in vivo. The hardness of the heart in vivo is around 10 kPa, but fibrotic scar tissue caused by myocardial infarction is 20–50 kPa. It has been reported that the microenvironment may affect the differentiation of mesenchymal stem cells [34].
Cardiomyocytes derived from iPS cells show enhanced cardiomyocyte maturation upon external stretch stimulation [35]. It has been reported that the reprogramming efficiency of iCMs can be enhanced by seeding fibroblasts on a substrate with microgrooves and transducing reprogramming factors [36]. Mechanotransduction via the extracellular matrix may also induce changes in the chromatin status and affect gene expression. It has been reported that the efficiency of miRNA-mediated direct myocardial reprogramming increases when cells are cultured in a three-dimensional hydrogel, compared to culturing in a two-dimensional culture [37].
We developed a new hydrogel culture system which reproduces the stiffness of myocardial tissue (10 kPa) in vitro. The number of beating iCMs (functional iCMs) increased threefold in the soft hydrogel culture dishes compared to regular polystyrene dishes. We found that YAP/TAZ suppression is involved in soft ECM-mediated cardiac reprogramming. In contrast, the hard dishes increased YAP/TAZ signaling and fibroblast programming, which inhibits direct reprogramming into CMs [38] (Fig. 4). This study is the first report to describe mechanotransduction of matrix stiffness during direct cardiac reprogramming [38].