We describe in this report an efficient method for generating human iPSCs with an intact genome by using a single episomal vector to transfect briefly-cultured adult PB or CB MNCs. Transient expression of a second vector (pEB-Tg) further increased reprogramming efficiencies of adult PB CD34+ and MNCs by 4-fold, although its stimulatory activity is much less on CB cells that have a much higher reprogramming efficiency. Tg stimulation on reprogramming was observed by several groups using different vectors and somatic cell types 17, 27
. Transient expression of Tg by episomal vectors did not necessarily alter karyotypes of the derived iPSCs (17
and this study), or global epigenetic signatures as shown in . Transient p53 inhibition by shRNA expression from an episomal vector also enhanced reprogramming efficiency as previously observed in studies using genome-integrating viral vectors 28, 29
, although its stimulatory effect on PB cell reprogramming is less than Tg over-expression. Interestingly, we noticed that all 5 of the iPSC lines derived from blood cells by non-integrating vectors we analyzed so far clustered together (and close to a group of hESC lines), based on promoter DNA methylation signatures. They appear distinct from other iPSCs derived by integrating vectors from either blood CD34+ cells or MSCs/fibroblasts (). A recent study found that mouse blood cell-derived iPSCs more closely resemble ESCs epigenetically and functionally than those derived from fibroblasts 30
. Whether the virus-free and integration-free human iPSCs we derived from blood cells and by the non-integrating vectors are superior in quality (closer to the best human ESCs) than previous iPSC lines derived by integrating vectors remains to be determined.
Our EBNA1/OriP episomal vector system offers several important advantages over the previous episomal vector combinations for reprogramming postnatal human cell types 17
. First, the pEB-C5 plasmid (+/− pEB-Tg) generated a higher percentage of TRA-1-60+ pre-iPSC colonies, especially in the presence of NaB, facilitating the efficient isolation of the desired TRA-1-60+ colonies. Second, the platform is more flexible if one needs to omit or replace Tg with other factors for mechanistic investigations. More importantly, our vector system, together with optimized cell culture conditions, allows us to efficiently generate integration-free human iPSCs from blood MNCs isolated from a few milliliters of blood after suspension cultures for 8–9 days. When further enhancement is needed for other cell types, or patient's blood MNCs that are more refractory to reprogramming by the current episomal vectors, we can easily add additional reprogramming genes over the 5–6 transgenes we used. Since we can deliver multiple (at least 8) genes from a single EBNA1/OriP plasmid for simultaneous and transient expression, it is no longer critical to reduce the numbers of reprogramming factors used. Obviously, small numbers of vectors (1–2) are preferred to ensure the lack of episomal vector DNA in the reprogrammed and expanded iPSC lines. Using PCR (3 sets of specific primers) and Southern blot, we did not detect vector DNA either as episomes or anywhere in the genome of the iPSCs analyzed. More sensitive methods are needed to detect residual vector DNA or to eliminate (episomal) vector DNA more efficiently than the passive expansion and dilution approach we used here.
Increasing evidence suggests that both DNA replication-dependent and -independent mechanisms are involved in promoting reprogramming, which is fundamentally an epigenetic process 31
. Our success in generating iPSCs from un-fractionated human blood MNCs depends on a culture condition to obtain a proliferating cell population that is neither T nor B cells, before plasmid-mediated reprogramming. We also detected that the proliferating cells derived from human blood MNCs or purified CD34+ cells display a global profile of promoter DNA methylation that groups closer to human ESCs/iPSCs than age-matched non-hematopoietic cells such as fibroblastic and endothelial cells. The unique blood cell DNA methylation and gene expression signatures may contribute to the high efficiency of iPSC derivation from blood cells we achieved by using a single episomal vector after one-time DNA transfection. Our data further substantiate the hypothesis that differences in epigenetic profiles among different human adult cell types contribute to differences in reprogramming efficiencies as well as the properties of derived iPSCs, as previously shown in the mouse system 30
. It is of interest to determine if the serial delivery of synthetic mRNAs encoding 4–5 reprogramming factors or purified proteins could reprogram human blood MNCs more efficiently than human fibroblasts as reported recently.
The facile method described here that reprograms blood cells after minimal culture and using plasmids provides several important advantages over other virus-free and integration-free methods. First, one-time delivery of 1–2 plasmids that are much easier to produce and more stable than proteins or mRNAs makes our method more attractive or feasible for generation of clinical-grade iPSCs under current good manufacture practice (cGMP) conditions. Second, the capability to reprogram PB MNCs not only provides a readily available source of human cells for iPSC derivation 1
, but also shortens the cell culture time required to prepare and prime the target cells for reprogramming (now < 10 days from blood versus
≥ 4 weeks from skin biopsies). Our combined method of using MNCs from a few milliliters of blood and non-integrating plasmids brings the iPSC derivation technology to a new level, and offers greater promise for the use of patient iPSCs for disease modeling and future clinical applications in regenerative medicine.