Here we present a robust method to reprogram primary embryonic fibroblasts into functional cardiomyocytes through overexpression of GATA4, Hand2, Mef2c, Tbx5, miR-1, and miR-133 (GHMT2m) alongside inhibition of TGF-β signaling. Our protocol generates beating cardiomyocytes as early as 7 days post-transduction with up to 60% efficiency.
Trans-differentiation of one somatic cell type into another has enormous potential to model and treat human diseases. Previous studies have shown that mouse embryonic, dermal, and cardiac fibroblasts can be reprogrammed into functional induced-cardiomyocyte-like cells (iCMs) through overexpression of cardiogenic transcription factors including GATA4, Hand2, Mef2c, and Tbx5 both in vitro and in vivo. However, these previous studies have shown relatively low efficiency. In order to restore heart function following injury, mechanisms governing cardiac reprogramming must be elucidated to increase efficiency and maturation of iCMs.
We previously demonstrated that inhibition of pro-fibrotic signaling dramatically increases reprogramming efficiency. Here, we detail methods to achieve a reprogramming efficiency of up to 60%. Furthermore, we describe several methods including flow cytometry, immunofluorescent imaging, and calcium imaging to quantify reprogramming efficiency and maturation of reprogrammed fibroblasts. Using the protocol detailed here, mechanistic studies can be undertaken to determine positive and negative regulators of cardiac reprogramming. These studies may identify signaling pathways that can be targeted to promote reprogramming efficiency and maturation, which could lead to novel cell therapies to treat human heart disease.
Ischemic heart disease is a leading cause of death in the United States1. Approximately 800,000 Americans experience a first or recurrent myocardial infarction (MI) per year1. Following MI, the death of cardiomyocytes (CMs) and cardiac fibrosis, deposited by activated cardiac fibroblasts, impair heart function2,3. Progression of heart failure following MI is largely irreversible due to the poor regenerative capacity of adult CMs4,5. While current clinical therapies slow disease progression and decrease risk of future cardiac events6,7,8,9, no therapies reverse disease progression due to the inability to regenerate CMs post-infarction10. Novel cell therapies are emerging to treat patients following MI. Disappointingly, clinical trials delivering stem cells to the heart following MI thus far have shown inconclusive regenerative potential11,12,13,14,15,16,17,18.
The generation of human-derived induced pluripotent stem cells (hiPSCs) from fibroblasts by overexpression of four transcription factors, first demonstrated by Takahashi & Yamanaka, opened the door to new breakthroughs in cell therapy19. These cells can differentiate into all three germ layers19, and several highly efficient methods for generating large numbers of CMs have been previously shown20,21. HiPSC-derived CMs (hiPS-CMs) offers a powerful platform to study cardiomyogenesis and may have important implications for repairing the heart following injury. However, hiPS-CMs currently face translational hurdles due to concerns of teratoma formation22, and their immature nature may be pro-arrhythmogenic23. Reprogramming fibroblasts into hiPSCs sparked interest in directly reprogramming fibroblasts into other cell types. Ieda et al. demonstrated that overexpression of GATA4, Mef2c, and Tbx5 (GMT) in fibroblasts results in direct reprogramming to cardiac lineage, albeit at low efficiency24. Reprogramming efficiency was improved with the addition of Hand2 (GHMT)25. Since these early studies, many publications have demonstrated that altering the reprogramming factor cocktail with additional transcription factors26,27,28,29, chromatin modifiers30,31, microRNAs32,33, or small molecules34 leads to improved reprogramming efficiency and/or maturation of induced cardiomyocyte-like cells (iCMs).
Here we provide a detailed protocol to generate iCMs from mouse embryonic fibroblasts (MEFs) with high efficiency. We previously showed that the GHMT cocktail is significantly improved with the addition of miR-1 and miR-133 (GHMT2m) and is further improved when pro-fibrotic signaling pathways including transforming growth factor β (TGF-β) signaling or Rho-associated protein kinase (ROCK) signaling pathways are inhibited35. Using this protocol, we show that approximately 60% of cells express cardiac Troponin T (cTnT), approximately 50% express α-actinin, and a high number of beating cells can be observed as early as Day 11 following transduction of reprogramming factors and treatment with the TGF-β type I receptor inhibitor A-83-01. Furthermore, these iCMs express gap junction proteins including connexin 43 and exhibit spontaneous contraction and calcium transients. This marked improvement in reprogramming efficiency compared to earlier studies demonstrates the potential to regenerate CMs from endogenous cell populations that remain in the heart post-infarction.
All experiments requiring animals were approved by the Institutional Animal Care and Use Committee at the UC Denver Anschutz Medical Campus.
1. Isolation of MEFs
2. Production of Retrovirus and Transduction of MEFs
NOTE: All steps in this section must be carried out in a Biosafety Level 2 Cabinet. Since this protocol utilizes retroviral transduction, ensure that safety precautions are taken including treating all waste containing viral media with 10% bleach for at least 20 min. Refer to Figure 1B for a timeline for reprogramming.
3. Flow Cytometry for Cardiac Markers
4. Immunostaining of Reprogrammed MEFs
5. Calcium Imaging
NOTE: If imaging at 10X magnification, standard cell culture dishes and plates are suitable for calcium imaging. Imaging at higher magnification requires that cells be plated on glass coverslips or glass bottom dishes.
Using the reprogramming strategy outlined above and in Figure 1B, we generated iCMs with approximately 70% of cells expressing cardiac Troponin T and approximately 55% of cells expressing cardiac α-actinin, quantified by flow cytometry at Day 9 following transduction of GHMT2m (Figure 2A and B). Additionally, the majority of cells express cardiac Troponin T, Troponin I, and cardiac α-actinin as well as the gap junction marker Connexin 43 at Day 14 following transduction (Figure 2C and D). Imaging with higher magnification reveals well-defined sarcomere structure formation and gap junctions forming between neighboring cells (white arrowheads, Figure 2D). Furthermore, spontaneous contraction and calcium transients indicate functionality of iCMs (Figure 3A-C and Movie 1).
Figure 1: Timeline of Reprogramming and Optimal Plating of Cells. (A) Scatterplot of log-transformed RNA-seq data for MEFs generated in our lab versus ENCODE MEFs. The R2 value and linear regression line are shown. (B) Schematic of all critical steps in the GHMT2m-mediated reprogramming of MEFs. (C) PE cells at 70-80% confluence prior to transfection (top row, left panel) and GFP signal 48 hours post-transfection to indicate transfection efficiency (top row, middle and right panels). MEFs seeded sparsely prior to infection to prevent overcrowding over the time period of reprogramming (bottom row, left panel) and GFP signal 48 hours post-infection to indicate infection efficiency (bottom row, middle and right panels). Scale bar = 400 µM. Please click here to view a larger version of this figure.
Figure 2: Quantification and Characterization of Cardiac Protein Expression in iCMs. (A and B). Representative flow cytometry for cardiac Troponin T (A) and α-actinin (B) at Day 9 following GHMT2m transduction, n = 3. (C) ICC staining for cardiac markers at Day 14 following GHMT2m transduction. Green: cardiac Troponin I, Red: cardiac Troponin T (middle panel) and α-actinin (right panel), and Blue: Hoechst staining for nuclei. Representative images (n = 3). Scale bar = 200 µM (D) High magnification imagines of sarcomere structure (Red: cardiac α-actinin (top row) and cardiac Troponin T (bottom row)) and expression of gap junction protein Connexin 43 (green). White arrowheads indicate gap junctions formed between neighboring cells. Representative images (n = 3). Scale bar = 100 µM. Please click here to view a larger version of this figure.
Figure 3: Functional Quantification of iCMs. (A) Time course of beating cell counts following GHMT2m transduction. Beating cells were counted by eye and cell counts from 10 fields per dish over 3 dishes per experiment were averaged and included in panel A. (B and C) Recorded calcium transients of iCMs. Calcium transient frequency is altered in iCMs following treatment with 1 µM Isoproterenol or 10 µM nifedipine. F340/F380, the ratio of fluorescence intensity at 340 and 380 nm. Please click here to view a larger version of this figure.
Movie 1: Beating iCMs. Movie showing beating iCMs at Day 12 in vitro. Please click here to view this video. (Right-click to download.)
iCM Media | ||
Name of Reagent | Volume (mL) | Final Concentration |
DMEM High Glucose | 320 | |
Medium 199 | 80 | |
Fetal Bovine Serum | 50 | 10% |
Donor Horse Serum | 25 | 5% |
MEM Essential Amino Acids, 50x | 10 | 1x |
Sodium Pyruvate Solution, 100x | 5 | 1x |
MEM Non-Essential Amino Acids, 100x | 5 | 1x |
MEM Vitamin Solution, 100x | 5 | 1x |
Insulin-Transferrin-Selenium, 100x | 5 | 1x |
B27, 50x | 10 | 1x |
Penicilin-Streptomycin | 5.5 | 1.1% |
L-Glutamine supplement | 5.5 | 1.1% |
PE Media | ||
Name of Reagent | Volume (mL) | Final Concentration |
DMEM High Glucose | 450 | |
Fetal Bovine Serum | 50 | 10% |
Penicilin-Streptomycin | 5.5 | 1.1% |
L-Glutamine supplement | 5.5 | 1.1% |
Blasticidin-HCl – 10mg/mL | 0.5 | 10 µg/mL |
Puromycin dihydrochloride | 0.05 | 1 µg/mL |
Growth Media | ||
Name of Reagent/ Equipment | Volume (mL) | Final Concentration |
DMEM High Glucose | 450 | |
Fetal Bovine Serum | 50 | 10% |
Penicilin-Streptomycin | 5.5 | 1.1% |
L-Glutamine supplement | 5.5 | 1.1% |
Table 1: Culture Media. Recipes for culture medium used in this protocol.
PE Cells | ||||||
Cell culture dish | Surface Area (cm2) | Seeding Density (cells) | Media Volume (mL) | Total DNA per transfection (µg) | Transfection reagent (µL) | Opti-MEM (µL) |
15 cm | 176.7 | 11.25 x 106 | 20 | 36 | 108 | 1080 |
10 cm | 78.5 | 5 x 106 | 10 | 12 | 36 | 360 |
60 mm | 28.2 | 1.7 x 106 | 4 | 4 | 12 | 120 |
6 well plate | 9 | 0.54 x 106 | 2 | 1.3 | 3.9 | 39 |
12 well plate | 4 | 0.24 x 106 | 1 | 0.6 | 1.8 | 18 |
24 well plate | 2 | 0.12 x 106 | 0.5 | 0.3 | 0.9 | 9 |
MEFs | ||||||
Cell culture dish | Surface Area (cm2) | Seeding Density (cells) | Media Volume (mL) | Hexadimethrine bromide (µL) | ||
15 cm | 176.7 | 1.35 x 106 | 20 | 12 | ||
10 cm | 78.5 | 0.6 x 106 | 10 | 6 | ||
60 mm | 28.2 | 0.2 x 106 | 4 | 2.4 | ||
6 well plate | 9 | 0.1 x 106 | 2 | 1.2 | ||
12 well plate | 4 | 0.4 x 105 | 1 | 0.6 | ||
24 well plate | 2 | 0.2 x 105 | 0.5 | 0.3 |
Table 2: Seeding Densities for Reprogramming. Table of seeding densities for PE cells and MEFs for common cell culture dish formats.
The present study outlines a high-efficiency strategy to directly reprogram fibroblasts into functional iCMs via delivery of GHMT2m reprogramming factors combined with suppression of pro-fibrotic signaling pathways. Using flow cytometry, immunofluorescent imaging, calcium imaging, and beating cell counts, we show the majority of cells in this protocol undergo successful reprogramming and adopt CM lineage fate. We have previously shown that the addition of anti-fibrotic compounds including the TGF-β type I receptor inhibitor A-83-01 yields an approximately 6-fold increase in the number of beating iCMs compared to transduction with GHMT2m alone, indicating that pro-fibrotic signaling is a strong endogenous barrier to cardiac reprogramming. This system can, therefore, be harnessed for drug screening to investigate mechanisms governing cardiomyogenesis; the addition of a small molecule alongside A-83-01 could uncover negative regulators of cardiomyogenesis. Conversely, the addition of a small molecule to GHMT2m alone could elucidate positive regulators of cardiomyogenesis. Unraveling the mechanisms through which these positive and negative regulators govern CM differentiation will result in a better understanding of cardiomyogenesis and lead to even more efficient reprogramming and maturation protocols.
Many variables affect reprogramming efficiency. We find that cell plating density greatly impacts both the production of retroviral particles in the case of PE cells, and the efficient uptake of all reprogramming factors in the case of MEFs. Furthermore, the passage number can affect these parameters as well. To ensure optimal production of viral particles, transfect PE cells before passage 20. To ensure efficient uptake of retroviral particles, do not passage MEFs after freezing.
Several recent studies have also demonstrated high-efficiency reprogramming of MEFs by knocking down PRC1 member Bmi along with expression of GMT30 or by overexpressing Akt in addition to GHMT37. Addition of ZFN281 to GHMT + Akt led to a 4-fold increase in troponin T positive cells in adult tail tip fibroblasts, attributed in part to suppression of inflammatory gene signatures38. Furthermore, the combination of a TGF-β signaling inhibitor and WNT signaling inhibitor increased GMT-mediated cardiac reprogramming39. Importantly, the signaling pathways involved in these studies do not appear to overlap. Therefore, it is likely that cross-talk between multiple signaling pathways could further enhance cardiac reprogramming.
A number of studies demonstrated reprogramming in vivo through the delivery of reprogramming factors in mice following left anterior descending (LAD) artery ligation25,32,39,40,41. These studies did show modest improvement in ventricular function following MI, underscoring the therapeutic potential of cardiac reprogramming on heart regeneration following injury. Our protocol reprograms MEFs with the highest efficiency in vitro to date. Therefore, it will be interesting to see if high-efficiency reprogramming is capable of fully restoring heart function post-injury in vivo. These studies could serve as a basis for translating this method into novel cell therapies for patients' post-infarction.
The authors have nothing to disclose.
This research was supported by funds from the Boettcher Foundation's Webb-Waring Biomedical Research Program, American Heart Association Scientist Development Grant (13SDG17400031), University of Colorado Department of Medicine Outstanding Early Career Scholar Program, University of Colorado Division of Cardiology Barlow Nyle endowment, and NIH R01HL133230 (to K.S). A.S.R was supported by NIH/NCATS Colorado CTSA Grant Number TL1TR001081 and a pre-doctoral fellowship from the University of Colorado Consortium for Fibrosis Research & Translation (CFReT). This research was also supported by the Cancer Center Support Grant (P30CA046934), the Skin Diseases Research Cores Grant (P30AR057212), and the Flow Cytometry Core at the University of Colorado Anschutz Medical Campus.
C57BL/6 Mice | Charles River's Laboratory | 027 | For MEF isolation |
Platinum E (PE) Cells | Cell Biolabs, INC | RV-101 | For retrovirus production |
DMEM High Glucose | Gibco | SH30022.FS | Component of iCM, PE, and Growth media |
Medium 199 | Life Technologies | 11150-059 | Component of iCM media |
Fetal Bovine Serum | Gemini | 100106 | Component of iCM, PE, and Growth media |
Donor Horse Serum | Gemini | 100508 500 | Component of iCM media |
MEM Essential Amino Acids, 50X | Life Technologies | 11130051 | Component of iCM media |
Sodium Pyruvate Solution, 100X | Life Technologies | 11360070 | Component of iCM media and for calcium imaging |
MEM Non-Essential Amino Acids, 100X | Life Technologies | 11140050 | Component of iCM media |
MEM Vitamin Solution, 100X | Life Technologies | 11120-052 | Component of iCM media |
Insulin-Transferrin-Selenium | Gibco | 41400045 | Component of iCM media |
B27 | Gibco | 17504-044 | Component of iCM media |
Penicilin-Streptomycin | Gibco | 15140-122 | Component of iCM, PE, and Growth media |
GlutaMAX (L-Glutamine Supplement) | Gibco | 35050-061 | Component of iCM, PE, and Growth media |
Blasticidin-HCl | Life Technologies | A11139-03 | Component of PE media |
Puromycin dihydrochloride | Life Technologies | A11138-03 | Component of PE media |
0.25% Trypsin/EDTA | Gibco | 25200-056 | For detaching cells from culture dishes |
A-83-01 | R&D Systems – Tocris | 2939/10 | Treat cells to inhibit TGF-β signaling – promotes high efficiecy reprogramming. Use at 0.5 µM |
DMSO | Thermo Scientific | 85190 | For dilution and storage of A-83-01 and component of Freeze Medium |
SureCoat | Cellutron | SC-9035 | For coating dishes to plate MEFs |
FuGENE 6 Transfection Reagent | Promega | E2692 | Transfection Reagent |
Opti-MEM Reduced Serum Media | Gibco | 11058-021 | Transfection Reagent |
pBabe-X Myc-GATA4 | Plasmid containing reprogramming factor | ||
pBabe-X Myc-Hand2 | Plasmid containing reprogramming factor | ||
pBabe-X Myc-Mef2c | Plasmid containing reprogramming factor | ||
pBabe-X Myc-Tbx5 | Plasmid containing reprogramming factor | ||
pBabe-X miR-1 | Plasmid containing reprogramming factor | ||
pBabe-X miR-133 | Plasmid containing reprogramming factor | ||
pBabe-X GFP | Plasmid containing reprogramming factor | ||
Polybrene (Hexadimethrine bromide) | Sigma | H9268-5G | For viral induction. Use at a concentration of 6 µg/mL |
Vacuum Filter + bottles (0.22 µm pores) | Nalgene | 569-0020 | For filtering media |
Syringes | Bd Vacutainer Labware | 309654 | For viral filtration |
0.45 µm Filters | Celltreat | 229749 | For viral filtration |
70 µm cell strainers | Falcon | 352350 | For MEF isolation and Flow Cytometry |
Cytofix/Cytoperm Solution | BD | 554722 | For fixation and permeabilization of cells for flow cytometry |
perm/wash buffer | BD | 554723 | For washing cells for flow cytometry |
DPBS 1X | Gibco | 14190-250 | For washing cells |
Bovine Serum Albumin | VWR | 0332-100g | For flow cytometry and calcium imaging |
Goat Serum | Sigma | G9023 | For blocking cells for Flow Cytometry |
Donkey Serum | Sigma | D9663-10mg | For blocking cells for Flow Cytometry |
Mouse Troponin T | Thermo Scientific | ms-295-p | 1:400 IF, 1:200 Flow Cytometry |
Mouse α-actinin | Sigma | A7811L | 1:400 IF, 1:200 Flow Cytometry |
Rabbit Connexin 43 | Sigma | C6219 | 1:400 IF |
Rabbit Troponin I | PhosphoSolutions | 2010-TNI | 1:400 IF |
Hoechst | Life Technologies | 62249 | 1:10000 IF |
Alexa 488, rabbit | Life Technologies | A-11034 | 1:800 IF |
Alexa 555, mouse | Life Technologies | A-21422 | 1:800 IF |
Alexa 647, mouse | Life Technologies | A-31571 | 1:200 Flow Cytometry |
27-color ZE5 Flow Cytometer | Bio-RAD | For FACS | |
Paraformaldehyde | sigma | P6148-500mg | For fixing cells for IF |
Triton X-100 | Promega | H5142 | For permeabilization of cells for IF |
EVOS™ FL Color Imaging System | Thermo Scientific | AMEFC4300 | For IF |
NaCl | RPI | S23020-5000 | For calcium imaging |
KCl | VWR | 395 | For calcium imaging |
CaCl2 | Fisher | C614-500 | For calcium imaging |
MgCl2 | VWR | 97061-352 | For calcium imaging |
glucose | sigma | G7528-250g | For calcium imaging |
HEPES | sigma | H4034-500g | For calcium imaging |
Fura-2 AM | Life Technologies | F1221 | For calcium imaging |
Fluronic F-127 | Sigma | P2443-250g | For calcium imaging |
Nifedipine | Sigma | N7634-1G | For disruption calcium transients in iCMs – use at 10 µM |
Isoproterenol | sigma | I6504-1g | For increasing number of calcium transients in iCMs – use at 1-2 µM |
Marianas Spinning Disk Confocal microscope | 3i | For calcium imaging | |
ethanol | Decon Laboratories | 2801 | |
bleach | Clorox | ||
50 mL conical tubes | GREINER BIO-ONE | 227261 | |
15 mL conical tubes | GREINER BIO-ONE | 188271 | |
15 cm cell culture dishes | Falcon | 353025 | |
10 cm cell culture dishes | Falcon | 353003 | |
60 mm cell culture dishes | GREINER BIO-ONE | 628160 | |
6 well cell culture plates | GREINER BIO-ONE | 657160 | |
12 well cell culture plates | GREINER BIO-ONE | 665180 | |
24 well cell culture plates | GREINER BIO-ONE | 662160 |