We demonstrate that a HAC vector containing expression cassettes for four reprogramming factors and a p53-knockdown construct efficiently reprogrammed somatic cells to pluripotency. In addition, we established integration-free iPS cells derived from these reprogrammed cells. Inserting all expression constructs into a defined cloning site on the HAC vector, which was maintained stably and independently of host chromosomes in cells, resulted in homogenous expression levels of the transgenes in HAC donor and recipient cells. Once configuration of an expression cassette in the HAC vector is optimized for the generation of iPS cells, the resulting uniformity of transgene expression in target cells is an advantage in promoting reprogramming efficiency and reducing clonal variation in the resulting iPS cells. Thus, the HAC-based reprogramming strategies are expected to be more effective in establishing homogenous iPS clones than other methods, including DNA transfection and viral transduction, which are both unable to regulate the quantity of xeno-products in modified cells. Nonetheless, the transfer rate of HAC vectors via MMCT is relatively low, i.e., 10-5
. To overcome this drawback, our protocol was enhanced with two procedures. First, to sustain high expression levels of individual reprogramming factors, each factor was surrounded with insulators. Second, to potentiate the reprogrammed state, miR294 cluster mimics, which promote induced pluripotency 
, were added after MMCT. Indeed, although the overall efficiency of reprogramming by our iHAC strategy was approximately 0.001%, more than half of iHAC-bearing cells developed an ES-like phenotype (iHAC1, 57%; and iHAC2, 62%). Furthermore, vector-free, transgene-free iPS cells were established from a third of the iHAC2-iPS lines. Notably, the effect of iHAC2 on the generation of iPS cells was no longer dependent on the addition of miR294 cluster mimics, because the pluripotent state induced by iHAC2 alone was sufficiently high. Therefore, the HAC vector system did facilitate somatic cell reprogramming by homogenous expression of the transgenic reprogramming factors and established vector-free, transgene-free iPS cells, which are suitable for clinical applications. Nevertheless, the MMCT frequency needs to improve. An improved MMCT technology may enhance the frequency of reprogramming by iHAC (50-100 times) 
All iHAC1-iPS cells satisfied some criteria of pluripotency (e.g., alkaline phosphatase staining, EB formation, ability of EB cells to differentiate into three germ layers); however, the iHAC1-iPS cells expressed only low levels of various pluripotent markers, indicating only partial reprogramming. In contrast, expression of pluripotent markers in most of the iHAC2-iPS cells was significantly upregulated and close to that in ES cells. These results were consistent with previous studies, which demonstrated that increasing Oct4 expression relative to the other three reprogramming factors 
and suppression of the p53 pathway 
resulted in enhanced reprogramming efficiency. Notably, in both iHAC1-iPS and iHAC2-iPS cells, gene silencing of the transgenic reprogramming factors was incomplete or nonexistent because each reprogramming factor was surrounded by multiple copies of the insulator. Moreover, expression levels of anti-proliferative genes in the iHAC1-iPS and the iHAC2-iPS cells were similar to those in ES cells, and were not upregulated (Fig. S6
). Therefore, the major cause of the partial reprogramming in the iHAC1-iPS cells, but not the iHAC2-iPS cells, may be inadequate activation of endogenous pluripotent genes rather than sustained expression of transgenes or induction of anti-proliferative genes. Recently, it has been demonstrated that Nanog drives partially reprogrammed cells into ground state pluripotency 
. Therefore, we can assume that p53 shRNA and the additional copies of Oct4 encoded by iHAC2 may have contributed to consolidating connections between core transcription factor networks and the enhanced expression of genes like Nanog 
. This hypothesis is supported by the evidence that supplementation of miR294/295 was no longer required for the iHAC2 reprogramming protocol. These results indicate that transgene integration into a defined cloning site on a HAC vector may also be useful for screening other reprogramming factors that improve the quality of iPS cells and increase overall efficiency.
Removal of potential obstacles, such as persistent exogenous genes or chemicals, the maintenance of normal cellular functions and the preservation of genome integrity are fundamental to the application of iPS cells in regenerative medicine. Here, we demonstrated that a HAC vector can mediate somatic cell reprogramming and that transgene-free, vector-free iPS cells can be obtained from iHAC2-iPS cells using simple FACS sorting; in contrast, other systems require prolonged culture, vector excision and drug selection to obtain integration-free iPS cells. Furthermore, the iHAC-free iPS cells generated using iHAC2 did not exhibit chromosomal aberrations, although the present version of iHAC2 contained a p53-knockdown construct. Thus, the strategy of using HAC vectors to generate integration-free iPS cells might be valuable. Moreover, the HAC vector itself was safe because it was maintained independently of host chromosomes, and importantly, chimeric mice were produced from mouse ES/iPS cells harboring a HAC vector 
. We have demonstrated that HAC vectors are able to carry therapeutic genes, especially genomic loci larger than 1 Mb, and that they are able to correct multiple cellular defects in target cells without transgene integration 
. As illustrated in , our overall strategy for regenerative medicine using HAC vectors equipped with one or more suicide genes as a safeguard system may facilitate the generation of patient-specific iPS cells that complement genetic traits causing diseases without risking genomic alteration and other undesirable outcomes.
Regenerative medicine strategy using HAC vectors.