In recent years, stem cell therapy has emerged as a promising approach to promote tissue repair after myocardial infarction. Despite encouraging initial results, clinical translation of cellular therapy has been impeded by several challenges, including poor engraftment 32
, malignant transformation 33
, or acute death of the transplanted cells 34
. Innovative approaches to overcome these challenges have relied on genetic modification of the delivered stem cells to enable expression of transgenes. Stable expression of such transgenes can enable a variety of valuable enhancements to cellular therapy, including enhanced vasculogenesis 35
, improved cell survival 15, 36
, cellular tracking using clinical imaging modalities 37
, and the availability of a “safety switch” to induce apoptosis in the event of cellular misbehavior 17
. However, commonly employed gene therapy vectors such as lentivirus and retrovirus can produce random integrations that are subject to the unpredictable effects of insertional mutagenesis. Hence site-specific integration techniques are needed to safely translate gene therapy into the clinical setting.
In recent years, stem and progenitor cells have emerged as excited new tools for ameliorating left ventricular dysfunction following myocardial infarction. hCPCs hold considerable potential in this regard, as they are a clonogenic, self-renewing population that is pre-programmed for differentiation into cells of all three cardiac lineages in vitro
: myocytes, smooth muscle cells, and endothelial cells. Several parameters for hCPC identification are currently found in the literature, including isolation of c-kit+
, cardiosphere-derived cells 38
, and Isl1+
. These various hCPC formulations have all demonstrated a degree of efficacy in animal studies, although evaluation of the degree of overlap between these various populations and their comparative efficacy is still needed. We employed Sca-1+
hCPCs in this study, as we and others have found that hCPCs isolated using the Sca-1 antibody consistently provide high yields of homogeneous cells with identical morphology, differentiation, and surface marker expression 39, 41, 42
. However, the human equivalent for murine Sca-1 has not yet been identified. Presumably, the Sca-1 antibody cross-reacts with a human protein to yield a homogeneous population, but further work is needed to elucidate the identity of this epitope.
Here, we employed the phiC31 integrase to achieve site-specific integration of a triple fusion reporter gene into a known chromosomal context in hCPCs and hECs. This proof-of-principle study demonstrates the utility of the phiC31 integration technique in producing both progenitor cells (hCPCs) and differentiated cells (hECs) that have been genetically modified to stably express a transgene of interest. Moreover, the techniques described herein are easily applied in any context with just basic molecular biology expertise and plasmid manipulation tools 43
. Genomic modification of hCPCs did not appreciably alter the progenitor cell phenotype, as assessed by global transcriptome microarray analysis and immunohistochemical analysis of differentiating progenitor cells. Indeed, phiC31-modified hCPCs retained their ability to improve myocardial contractility by a statistically significant margin compared to controls when assessed two weeks after intramyocardial injection. Likewise, phiC31-modified hECs maintained their ability to improve ischemic hindlimb perfusion after genomic modification.
We have the found the phiC31 integrase technique a facile and efficient approach for generating site-specific integrants in progenitor and differentiated cells. Alternate site-specific integration techniques are significantly less efficient than phiC31-integrase mediated recombination because they often rely on successful activation of mammalian cells’ homologous recombination machinery. One such technique involves the introduction of a double-strand break at the desired genomic location by a tandem pair of zinc-finger nucleases (ZFNs) 44
. Although ZFNs allow targeting of a wider variety of genomic loci, their design and validation require extensive time and expertise. In addition to verifying proper targeting, cells’ exposure to ZFNs must be minimized in order to prevent any off-target cleavage events. The phiC31 integrase system is easily implemented using common plasmid manipulation and molecular biology techniques.
In summary, our report establishes for the first time the feasibility of site-specific genetic integration in human cardiac progenitor cells and human endothelial cells using non-viral vectors. This type of genetic engineering of human stem cells enables a multitude of improvements to overcome many of cellular therapy’s current limitations. The ease, efficiency, and safety of the phiC31 integrase system make it an appealing approach to such attempts at stable genetic modification of therapeutic cell populations.