Classification of chromosomal aberrations following somatic reprogramming
Synthetic modified KMOS mRNAs (KLF4, cMYC, OCT4 and SOX2) were transfected into HDFs with normal karyotype (Fig. 1a) to initiate the reprogramming process. Following transfections, human iPSC-like colonies were picked and seeded into 24-well plates at one clone per well (Passage 1). Eight days later, cells were processed for karyotyping. Although the vast majority of clones maintained their normal karyotypes, three of them were found to be chromosomally abnormal (6.12%) (Fig. 1b). In one iPS cell clone, the same chromosomal aberration was found in every progeny cell (Fig. 1b, d). This type of aberration is referred to as type-1 clone, and we suspected that the mutation occurred during the reprogramming process (Fig. 1d). The other two karyotypically abnormal iPS clones were found to carry chromosomal aberrations only in partial cells (26.7 % and 53.3%, respectively), and they were referred to as type-2 clones (Fig. 1 b, d). In these two type-2 clones, all karyotypically abnormal cells carried the same chromosomal aberration, suggesting that they were derived from one mutated cell (Fig. 1d). Intriguingly, the proportion of chromosomally abnormal cells in these two clones was either one fourth or one half, implying that chromosomal aberrations occurred during the first or second cell division after the fate conversion from somatic cell to iPSC (Fig. 1d).
Besides the normal typical morphology (Fig. 2a), RT-PCR showed that the karyotypically abnormal iPS cell clones expressed pluripotent marker genes such as NANOG, OCT4, SOX2 and GAL at similar levels to those in normal ones and human embryonic stem cell (hESC) line H9 (Fig. 2b). The immunological features of karyotypically abnormal iPS cell clones were also similar to those of normal ones (Fig. 2c). Furthermore, Teratoma were also formed after transplanting karyotypically abnormal iPSCs into immunodeficient nude mice, and three germ-layer tissue cells were detected, showing the pluripotency of their differentiation (Fig. S1). Thus, the chromosomal aberrations did not appear to compromise the pluripotency and differentiation potential of iPSCs. Conceivably, destructive chromosomal aberrations also inevitably occurred in the process of reprogramming, but they were difficult to detect due to their adverse effects on the survival or self-renewal of iPSCs.
Evidence for mutagenicity during reprogramming process
Karyotyping only involves the examination of chromosomes at the metaphase stage of mitosis. HDFs isolated from skin tissue were a mixed population, and it is conceivable that some exiguous karyotypical abnormalities may not be detected in the mixture. Therefore, it remains plausible that iPS cell clones with abnormal karyotypes, especially type-1 clones, may be reprogrammed from rare chromosomally abnormal parental cells in this study. To rule out this possibility, we obtained single cell-derived karyotypically normal HDF clones to perform clonal reprogramming. When HDFs were plated at clonal density, only a small percentage (1.02 ± 0.15%) of cells grew into clones, and these clones showed poor proliferative capacity with significant aging characteristics (cells became larger with emerging processes) (Fig. 3a). Therefore, it is technically challenging to derive a large number of cells from a single primary clone for somatic cell reprogramming. It was reported that addition of four reprogramming proteins (KLF4, cMYC, OCT4 and SOX2) fused with 9 arginine (9R, a cell-penetrating peptide sequence) to the culture medium can significantly enhance their proliferative capacity [11]. Therefore, HDFs were also tested with the combined cell extracts of four 293T cell lines expressing individual reprogramming factor. After 16 hour of protein transduction, cells were cultured in fibroblast medium for 6 additional days, followed by digesting and passaging (Fig. 3b). After four cycles of repeated protein treatment and subculturing, their cloning ability improved dramatically (Fig. 3d) and many vigorous clones were obtained (Fig. 3c). These clones could be further subcultured for amplification, and EDU labeling showed their significantly higher proliferative capacity compared to primary HDFs (Fig. 3f, g). More importantly, they maintained the typical morphology of fibroblasts (Fig. 3e) and Vimentin expression (Fig. S2), while the immunophenotypes of pluripotent markers such as SSEA-4, OCT4, NANOG, TRA-1-60 and TRA-1-81 were negative (Fig. S2). Furthermore, their karyotypes remained normal and no chromosomal aberration was detected (Fig. 3h).
The vigorous HDF clones derived from three donors were randomly selected, and each clone was subjected to two treatments i.e. KMOS mRNAs transfection and GFP mRNA transfection according to the protocol described in Fig. 4a. In KMOS mRNAs transfection group, cellular division rapidly accelerated, and the EDU positive rate on d5 increased to the highest value, which was about twice that of d1(Fig. 4b). While the KMOS mRNAs transfection group obtained a 0.5-0.8% iPSC induction efficiency, no iPSC-like colony was found in the GFP mRNA transfection group. As a result, human iPSC-like colonies were only picked from the KMOS mRNAs group for karyotyping. While the vast majority of clones kept the primary normal karyotype, chromosomal aberrations were found in three donor HDF-derived iPS cell clones (Fig. 4c, d). These results confirmed the ‘chromosomal mutagenicity’ of reprogramming process. The chromosomal aberrations appeared to occur randomly, and no apparent locations or patterns were noticed (Fig. 4c). To exclude the effects of mRNA transfection on karyotypic stability, we examined the karyotypes of cells of GFP mRNA transfection group and found no chromosomal aberration (data not shown). Together, these findings demonstrated that reprogramming process mediated by KMOS mRNA transfection could trigger chromosomal aberration at a low frequency.
Restoration of karyotypic stability in established iPS cell lines
The emergence of karyotypically abnormal iPS cell clones also prompted us to investigate whether it was caused by the deterioration of chromosomal stability after reprogramming. Therefore, we tracked the karyotypic stability of iPS cell clones in the subsequent subculturing. In this study, 30 karyotypically normal iPS cell clones and 10 abnormal clones were selected and passaged every 6 days by collagenase IV digestion, and karyotyping was performed every 5 generations. For karyotypically normal iPS cell clones, their chromosomes kept unchanged (Fig. 5a). Of the karyotypically abnormal clones, the primary aberrations were retained, and no new aberrations were observed throughout the subculturing (Fig. 5b). These results indicate that iPSCs possess a reliable mechanism for karyotypic stability after the re-establishment of stemness.
Antioxidants reduced chromosomal aberrations during reprogramming
As somatic reprogramming to iPSCs undergoes a rapid increase in the rate of cell division (Fig. 4b), it leads to a sharp increase in energy demand and the transformation of oxidative respiration to oxidative glycolysis [14]. The metabolic transformation causes a significant increase in the level of reactive oxygen species (ROS) [15,16], and high ROS levels can result in oxidative DNA damage [17-22]. Consistent with this principle, we detected a significant increase in the proportion of KMOS mRNAs transfected fibroblasts with γH2AX foci, a widely used marker for monitoring the levels of DNA double-strand breaks (DSBs), while treatment with the antioxidant N-acetyl-cysteine (NAC) reduced significantly γH2AX-positive cells (Fig. 6a, b). We suspect that the DSBs induced by reprogramming may contribute to the chromosomal aberration, and antioxidant may have the potential to reduce the chromosomal aberrations associated with the reprogramming process. To test this idea, we included the antioxidant NAC in the medium during iPSC induction. Random selection of iPSC-like colonies for karyotyping (Fig. 6c) showed that the percentage of karyotypically abnormal iPS cell clones in the NAC treatment group was significantly lower than that of control (Fig. 6d), indicating that NAC treatment effectively reduced the occurrence of chromosomal aberration.
DSBs are the most common form of DNA damage which can be repaired by two different pathways: error-prone nonhomologous end joining (NHEJ) and error-free homologous recombination repair (HRR) [23]. When DSBs are introduced, embryonic stem cells (ESCs) predominantly adopt high fidelity HRR to repair the lesions rather than NHEJ [24-28].However, when ES cells differentiate into somatic cells, the expression of HRR-related enzymes is down-regulated, whereas the expression of NHEJ-related enzymes, such as DNA Ligase IV, is up-regulated. As a result, DSB repair pathway shifts from HRR to NHEJ [26,29]. Consistently, the expression level of DNA Ligase IV during reprogramming displayed a significant downregulation after KMOS mRNAs transfection (Fig. 6e, f). In contrast, the expression levels of the HRR pathway-related proteins Rad51 and Rad52 increased gradually [29], reaching a peak on day 7 (Fig. 6g-j). These data suggest that the NHEJ pathway slowly attenuates during reprogramming while the high-fidelity HRR pathway gradually increases, in an opposite direction of iPSC differentiation [25,26].
At the early stage of reprogramming, the DSBs induced by the rapid increase of ROS are repaired mainly by NHEJ pathway, which could contribute to the chromosomal aberration in the process of reprogramming. Consistently, our studies revealed that the percentage of karyotypically abnormal clones was reduced significantly by NAC treatment in the first seven days of reprogramming, and the effect was similar to that of the group treated with NAC throughout the reprogramming (Fig. 6k, l). Moreover, adding NAC to the culture medium after the seventh day had no significant effect (Fig. 6k, l). Collectively, these findings strongly suggested the ‘mutagenicity’ of reprogramming process itself.