Abstract
Preclinical studies have suggested that transplanted human pluripotent stem cell-derived cardiomyocyte (hPSC-CM) grafts expand due to proliferation. This knowledge came from cell cycle activity measurements that cannot discriminate between cytokinesis or DNA synthesis associated with hypertrophy. To refine our understanding of hPSC-CM cell therapy, we genetically engineered a cardiomyocyte-specific fluorescent barcoding system into an hPSC line. Since cellular progeny have the same color as parental hPSC-CMs, we could identify subsets of engrafted hPSC-CMs that clonally expanded, with the remainder being non-proliferative and hypertrophic.
Main Text
Preclinical studies have suggested that transplanted human pluripotent stem cell-derived cardiomyocyte (hPSC-CM) grafts expand due to proliferation.1 This knowledge came from cell cycle activity measurements that cannot discriminate between cytokinesis or DNA synthesis associated with hypertrophy. To refine our understanding of hPSC-CM cell therapy, we genetically engineered a cardiomyocyte-specific fluorescent barcoding system into an hPSC line. Since cellular progeny have the same color as parental hPSC-CMs, we could identify subsets of engrafted hPSC-CMs that clonally expanded, with the remainder being non-proliferative and hypertrophic.
We generated our hPSC line by knocking four copies of the Cre-dependent Brainbow 3.2 lineage reporter2 into WTC11 cells (Figure A). These rainbow hPSCs were transduced with cardiac troponin T (cTnT)-driven Cre, which restricts expression of the rainbow barcoding system to committed cardiomyocytes (Figure A). Rainbow-labeling was observed after 7 days of differentiation and immunostaining confirmed the labeled cells expressed cTnT (Figure B). We aimed for sparse labeling so we could track individual cells over time since neighboring cells would not have the same color (Figure C). Cre-mediated recombination elicited all eighteen of the possible hues, but monocolor-labeling dominated (Figure D).
By day 14, we observed that hPSC-CMs had clonally expanded (Figure E). On average, the number of cardiomyocytes per clone went from 1.03 to 1.71 (day 7 versus 14, p<0.001, Figure F). While most rainbow hPSC-CMs did not proliferate over time, some were highly proliferative and a subset of hPSC-CMs continued to proliferate after replating at day 14 (Figure F). Repeat imaging of the replated hPSC-CMs over days 15-28 confirmed that (1) neighboring cells had unique hues (Figures F and G) and (2) all daughter cardiomyocytes inherited the parental fluorescent-barcode (Figure G), definitively demonstrating these clusters arise from clonal expansion (Figure H). We noted that hPSC-CM underwent hypertrophic growth as measured by cell area (Figure I), and by day 28, clonally expanded hPSC-CMs were 4.07-fold smaller than non-dividing hPSC-CMs (p<0.0001, Figure I). Staining reconfirmed cTnT-driven labeling was specifically induced in cardiomyocytes and showed multinucleation in non-dividing, hypertrophied hPSC-CMs (Figures H and J).
For the transplantation studies, sparsely-labeled day 14 hPSC-CMs were dissociated, resuspended in Matrigel with prosurvival cocktail, and transplanted into the hearts of immune-compromised athymic rats.1 Hearts were harvested two weeks after hPSC-CM engraftment, which equates to the 28-day timepoint in vitro. Most engrafted hPSC-CMs hypertrophied (1.75-fold increase versus injectate, p<0.002) while some subsets clonally expanded with the average number of cells per clone increasing to 1.77 (p<0.03, Figures K-O). Twenty-two clusters of engrafted hPSC-CM clones were identified, of those 36.4% had undergone at least one cell division and 18.2% had divided multiple times. To ensure that the results were not driven by false positives, the cell injectate was imaged to confirm that neighboring hPSC-CMs did not express the same barcode at the time of transplantation (Figures K and M). Notably, the heterogeneous proliferative potential among engrafted hPSC-CMs, which would not have been observed without the rainbow single-cell reporter, was not due to differences in sarcomere content, as the proportion of α-actinin+ area was the same in both clonally expanding and non-expanding groups (p>0.77, Figures O-P).
As hPSC-CM therapy is rapidly approaching clinical use, it is critical to understand how these cells behave in vivo. Indeed single cell transcriptomics assays3 and DNA content analysis4 have revealed profound molecular heterogeneity among these hPSC-CMs. It remains unclear how these variances translate to functional heterogeneity, but by generating a cTnT lineage rainbow reporter we could longitudinally track the hypertrophic and proliferative growth of individual hPSC-CMs. This approach demonstrated for the first time that hPSC-CMs have heterogeneous levels of proliferation in vitro and after engraftment in host myocardium, revealing a dichotomy between non-dividing and clonally expanding hPSC-CMs in their capacity to hypertrophy. By examining the generation of newly formed cardiomyocytes, rather than utilizing proxies for cell proliferation, this study distinguished bona fide cardiomyocyte division versus incomplete cell cycle activation, which has been a source of controversy in the field. Moreover, this analysis found instances of multinucleation that could have led to false identification of cell proliferation with gold standard DNA synthesis or mitosis markers that measure newly formed genomes or nuclei rather than cell division. While the heterogenous proliferative capacity among hPSC-CMs was initially surprising to us, our results are consistent with findings that demonstrated only a few clonally dominant cardiomyocytes give rise to most of the adult zebrafish heart5, suggesting a similar mechanism may underlie hPSC-CM graft expansion. The finding that proliferative hPSC-CM express normal levels of sarcomere contractile elements suggests increasing hPSC-CM clonal expansion will more efficiently repopulate myocardium lost to injury. This is in line with cardiac cell therapy optimizations, such as co-transplantation of hPSC-CMs with epicardial cells1, demonstrating significantly improved outcomes from stimulating grafted hPSC-CM proliferation. Thus, controlling engrafted hPSC-CM clonal expansion holds great promise for improving cardiac regenerative therapies.
Article Information
Data sharing: All data and materials will be available from the corresponding author by request.
Sources of Funding
This work was supported by NIH HL141187 & HL142624 (J.D.), NSF CMMI-1661730 (N.J.S.), NIH F32HL143851 (D.E.), Gree Family Gift (J.D./N.J.S./C.E.M.). C.E.M. was also supported by NIH grants R01HL128362, U54DK107979, R01HL128368, R01HL141570, R01HL146868, and a grant from the Foundation Leducq Transatlantic Network of Excellence.
Disclosures
C.E.M. is a founder and equity holder in Cytocardia. The other authors report no conflicts of interest.