Direct Reprogramming for Cardiac Regeneration
From Dream to Reality
Direct reprogramming of endogenous fibroblasts into functional cardiomyocytes to regenerate the heart after myocardial infarction (MI) has long been a dream. A few years ago, this would have been considered by most to be science fiction. Certainly, the most optimistic scientists would probably categorize this as “highly unlikely to ever work.” And yet, 2 papers,1,2 including one in this issue of Circulation Research,2 have achieved this, using two distinct approaches. These landmark studies will hopefully pave the way for effective approaches to restoring cardiac function after cardiac injury.
Article, see p 1465
Following MI, a large portion of the heart dies, and muscle is replaced by a fibrotic scar. This scar contributes to impaired cardiac function, which eventually results in heart failure in patients who survive an infarct. Much emphasis in cardiac research has been toward developing strategies to enhance cardiac function post-MI.3,4 The majority of approaches have been cell-based, involving the introduction of cells back into the injured heart. Strategies in this category have included bone marrow-derived mesenchymal cells, endogenous cells from the heart, or cardiomyocytes derived from embryonic stem or induced pluripotent stem (iPS) cells. To date these approaches have not resulted in long-term restoration of new myocardium, although clinical trials have shown some promising results in improving cardiac function.5,6 As an alternative to cell replacement strategies, others have turned to direct reprogramming.
Direct reprogramming involves transforming one somatic cell type into another, by introducing regulators of the desired cell type and attempting to coax the genome of the recipient cell into resetting itself. The most dramatic example of reprogramming is the reinduction of pluripotency, by introduction of defined factors, or other means.7 Others have coaxed fibroblasts to become neurons, pancreatic exocrine cells to adopt a beta cell phenotype, or driven cells toward other induced fates.8,9 Two different approaches have been used to reprogram fibroblasts into cardiomyocyte-like cells in cell culture.10,11 One approach exploited the partial reprogramming inherent to the iPS cell reprogramming strategy, to divert incompletely reprogrammed cells toward a cardiac phenotype.10 The other was modeled after the iPS cell reprogramming and by a systematic selection process identified a combination of 3 transcription factors—Gata4, Mef2c, and Tbx5 (referred to as GMT)—that when introduced into cultured fibroblasts could reprogram them into cells that closely resembled cardiomyocytes.11 The dream based on these promising results has been to adopt this approach in vivo. This has now been achieved.1,2
Writing in the current issue of Circulation Research, Jayawardena and colleagues demonstrate the potential for microRNAs, short RNAs that regulate gene expression via translational and transcriptional modulation, to reprogram in vitro cultured and endogenous resident mouse cardiac fibroblasts toward a cardiomyocyte phenotype. A candidate approach for selecting microRNAs was used, and a combination of miR-1, miR-133, miR-208, and miR-499 was found to be effective. These microRNAs were selected based on their known roles in regulating broad aspects of cardiac gene expression.12,13 Jayawardena et al used a Myh6 promoter-driven CFP reporter transgene (Myh6::CFP) to monitor cardiac gene expression. Transfecting cardiac fibroblasts with mimics to several microRNAs (miRNAs), 1.5% to 5% of transfected cells expressed Myh6::CFP. Interestingly, this was increased to 13% to 27% with a JAK inhibitor that had been previously shown to drive cardiac differentiation from partially reprogrammed cells.10 However, the authors did not quantitate endogenous cardiac markers to assess the efficiency of reprogramming; a previous study showed that the Myh6 promoter is more sensitive to reprogramming factors than many endogenous genes.11 Increased mRNA levels for cardiac genes were measured in pooled cells, but it is not known in how many cells these were active. Functional assessments showed that only 1% to 2% of transfected cells were beating, similar to transcription factor-based methods.11 Ca2+ oscillations were also observed in some transfected cells, indicative of functional excitation-contraction coupling. Thus in vitro microRNAs can at least partially reprogram cardiac fibroblasts toward a cardiac phenotype.
Jayawardena et al followed these promising in vitro results in vivo, using lentiviral delivery of miR-1 or miR-1, miR-133, miR-208, and miR-499 in the border zone of postinfarction mouse heart. Either manipulation resulted in the appearance of cardiac Troponin T+ cells that by genetic lineage tracing were likely to be derived from fibroblasts. As a transfection marker was not included, it was not possible to evaluate the efficiency of reprogramming using this approach. Cardiac function was not assessed, so the benefits of the intervention are not known. The authors also did not address the possibility that the in vivo reprogramming was due to cell fusion of fibroblasts with existing cardiomyocytes. Despite these caveats, this study shows that a combination of miRNAs can induce endogenous fibroblasts to undergo a conversion toward a cardiac phenotype. Because miRNAs can be delivered systemically via nonviral methods, this study raises the exciting possibility that therapeutic cardiac reprogramming could be achieved without a need for viral delivery methods.
In a second paper, published in Nature, Qian et al1 extended the 3 transcription factors (GMT) approach previously used in cell culture,11 to assess its in vivo potency. Using a retroviral system to overexpress GMT, Qian et al found that fibroblasts could only be infected after MI, consistent with their active proliferative state. Addition of GMT post-MI resulted in an efficient conversion of fibroblasts to a cardiac phenotype, as evidenced by the activation of several endogenous markers (α-actinin, cardiac troponin T, alphaMHC, α-Topomyosin) and a transgenic Myh6::EGFP reporter, overlapping with a red fluorescent transfection marker. Here the fibroblast origin of the reprogrammed cells was also assessed by lineage tracing, using 2 different fibroblast-specific Cre lines. Neither line was active in cardiomyocytes following injury and exclusively marked noncardiomyocytes. The potential conversion of endothelial cells was excluded using lineage tracing with Tie2::Cre. The possibility of cell fusion was also excluded using an inducible Myh6::MerCreMer transgene, which when activated prior to reprogramming did not label newly reprogrammed cardiomyocytes. The collective lineage tracing experiments indicate that proliferating fibroblasts are directly reprogrammed to cardiomyocytes by GMT. GMT factors were somewhat more efficient in vivo than in vitro: A 12% reprogramming efficiency was achieved in vivo, compared to 5% to 10% in vitro. Most importantly, the in vivo reprogrammed cells more closely resembled endogenous cardiomyocytes than their in vitro reprogrammed counterparts. Many of the in vivo reprogrammed cells were rod-shaped and binucleated, expressed many cardiac markers, and had Cx43+ gap junctions at their ends. Most importantly, these cells also had important functional hallmarks of a mature cardiomyocyte, including mature action potentials and the ability to contract. Another key factor was that the reprogrammed cells were electrically coupled to their neighbors, which is an essential feature of any approach that seeks to restore new cardiomyocytes to the injured heart. Therefore, the combination of GMT could convert endogenous fibroblasts to cells closely resembling cardiomyocytes.
Comparing the efficiency of reprogramming in vitro, versus in vivo, it is clear that there is a significant contribution from the cellular environment to the efficiency and extent of reprogramming. It is known that signaling cues can influence cellular reprogramming14; identifying which of these in the adult heart are the key potentiators of cardiac reprogramming will be of considerable interest to understand the process and to enhance it.
An important aspect to the Qian et al study is that reprogramming fibroblasts to cardiomyocytes led to clear and long-lasting improvements in cardiac function post-MI. Although not restored to pre-MI levels, function was improved, as measured by echo and MRI, 8 to 12 weeks after delivery. Additionally, a significant reduction in scar size was measured, and there was clear evidence of new myocytes within the scar. An interesting added twist was that the addition of Thymosin β4, a cytokine that had been previously shown to improve cardiac function post-MI,15 further enhanced the improvements in cardiac function and scar size reduction by GMT, presumably by increasing fibroblast mobilization.
Some questions and issues arise from these studies. One is that the reprogramming efficiency in both studies is still low. In addition to cell numbers, a strict definition of a functional cardiomyocyte is required to measure the degree of efficiency. At the very least, reprogrammed cardiomyocytes should have structural and physiological characteristics, including the ability to electrically couple, that are functionally indistinguishable from their endogenous counterparts to be therapeutically useful. Many studies in iPS cell reprogramming have identified several factors, eg, chromatin remodelers16 or growth factors,14,17 that can improve reprogramming. Another challenge will be delivery in large animals. This is a 2-part challenge. The first is to ensure that the much thicker myocardium of the human heart can be efficiently and broadly targeted. Trials in large animals will be needed to address this. The second is to ensure safe delivery. Currently both the miRNA and transcription factor-based approaches use viral delivery. The safety of gene therapy approaches using these types of vectors have been contentious, and thus avoiding these would be ideal. Again in the iPS cell field, there has been some success replacing certain reprogramming transcription factors with small molecules;14,18 it is likely that similar small molecule approaches will be effective in the context of cardiac reprogramming. Alternatively, adeno-associated virus vectors have been shown to be safe and effective in delivering to the heart, and both transcription factors and miRNA could be delivered using this technology. A final important question is the mechanism underlying the reprogramming. Transcription factors have defined genomic targets, and miRNAs have specific mRNA targets, so it is certainly a tractable problem. It will be fascinating to uncover the pathways that are activated or repressed during the process of reprogramming fibroblasts to cardiomyocytes.
The papers by Jayawardena et al and Qian et al provide an exciting hope for direct reprogramming as a viable strategy for cardiac repair. Beyond post-MI myocardium, several other cardiac diseases could benefit from this approach. It is clear that we have departed from the realm of science fiction and can now consider the very real future of cardiac reprogramming.
Sources of Funding
My laboratory is funded by grants from the NIH, CIRM, and the Lawrence J. and Florence A. DeGeorge Charitable Trust/American Heart Association Established Investigator Award.
The opinions expressed in this article are not necessarily those of the editors or of the American Heart Association.
- © 2012 American Heart Association, Inc.
- Qian L,
- Huang Y,
- Spencer CI,
- Foley A,
- Vedantham V,
- Liu L,
- Conway SJ,
- Fu JD,
- Srivastava D
- Jayawardena TM,
- Egemnazarov B,
- Finch EA,
- Zhang L,
- Payne JA,
- Pandya K,
- Zhang Z,
- Rosenberg P,
- Mirotsou M,
- Dzau VJ
- Williams AR,
- Hare JM