p53 isoform Δ133p53 promotes efficiency of induced pluripotent stem cells and ensures genomic integrity during reprogramming

Human induced pluripotent stem (iPS) cells have great potential in regenerative medicine, but this depends on the integrity of their genomes. iPS cells have been found to contain a large number of de novo genetic alterations due to DNA damage response during reprogramming. Thus, to maintain the genetic stability of iPS cells is an important goal in iPS cell technology. DNA damage response can trigger tumor suppressor p53 activation, which ensures genome integrity of reprogramming cells by inducing apoptosis and senescence. p53 isoform Δ133p53 is a p53 target gene and functions to not only antagonize p53 mediated apoptosis, but also promote DNA double-strand break (DSB) repair. Here we report that Δ133p53 is induced in reprogramming. Knockdown of Δ133p53 results 2-fold decrease in reprogramming efficiency, 4-fold increase in chromosomal aberrations, whereas overexpression of Δ133p53 with 4 Yamanaka factors showes 4-fold increase in reprogamming efficiency and 2-fold decrease in chromosomal aberrations, compared to those in iPS cells induced only with 4 Yamanaka factors. Overexpression of Δ133p53 can inhibit cell apoptosis and promote DNA DSB repair foci formation during reprogramming. Our finding demonstrates that the overexpression of Δ133p53 not only enhances reprogramming efficiency, but also results better genetic quality in iPS cells.


Human induced pluripotent stem (iPS) cells have great potential in regenerative medicine, but this depends on the integrity of their genomes. iPS cells have been found to contain a large number of de novo genetic alterations due to DNA damage response during reprogramming. Thus, to maintain the genetic stability of iPS cells is an important goal in iPS cell technology. DNA damage response can trigger tumor suppressor p53 activation, which ensures genome integrity of reprogramming cells by inducing apoptosis and senescence. p53 isoform
Human induced Pluripotent Stem (iPS) cells can be generated by viral-based ectopic expression of specific transcription factors (e.g., Oct4, Sox2, Klf4, and c-Myc), which provides great potential for use in research and regenerative medicine. However, a number of studies have shown that the reprogramming process can induce genetic abnormalities in iPS cells [1][2][3][4][5][6] . More than 1000 heterozygous single-nucleotide variants were found in human iPS cell lines induced even by non-integrating plasmid expression 3 . These studies raise great concerns on the chromosome aberrations for future application of iPS cells. The most possible reason for generation of genetic variants in iPS cells is that early reprogramming of iPS cells induced by Yamanaka factors triggers the DNA damage response 7,8 . A method for maintaining the genetic stability of iPS cells is very crucial for practical application.
Tumor suppressor p53 is activated by DNA damage and plays a central role in the DNA damage response. The activation of p53 induces cell cycle arrest, DNA damage repair, apoptosis and/or senescence to guard genome stability. Previous studies showed that the p53 signal pathway is activated and DNA double-strand break (DSB) repair foci are formed during cell reprogramming, which suggests that the process of cell reprogramming causes DNA DSBs 7,9,10 . The most toxic lesion in DNA is the DSB. To combat this toxic insult, a number of pathways have evolved to repair DNA DSBs: Homologous Recombination (HR), Non-Homologous End Joining (NHEJ) and Single-Strand Annealing (SSA). In contradict to its tumor suppression role, p53 protein inhibits the HR, NHEJ

Results
Δ133p53 is induced in cell reprogramming and functions to promote reprogramming efficiency. The role of Δ 133p53 in DNA DSB repair prompted us to speculate that Δ 133p53 may have an effect in cell reprogramming. We first checked the expression of Δ 133p53 at 1, 2, 9, 12 and 17 days post infection (dpi) during the reprogramming of human skin fibroblast (CDD-1079sk) cells mediated by the four Yamanaka factors. Interestingly, we found that Δ 133p53 protein and transcript were induced, as were those of full-length p53, from 9 dpi (Fig. 1A,B). Notably, Δ 133p53 was also expressed in the human embryonic stem cells (embryonic cell line 14) 22 , but not in mouse embryonic fibroblast (MEF) cells (Fig. 1A,B). Next, we combined the four Yamanaka factors with specific short hairpin RNAs (shRNA) to knockdown p53 or Δ 133p53, or used ef1a-Δ133p53 to overexpress Δ 133p53 during reprogramming (Fig. 1C). The knockdown and ectopic expression of Δ 133p53 did not have much effect on the level of full-length p53 protein (Fig. 1C). However, the knockdown of full-length p53 also downregulated the expression of Δ 133p53 (Fig. 1C), which is consistent with that Δ 133p53 is a p53 target gene. Similar to the previous studies 7 , the knockdown of full-length p53 promoted the reprogramming efficiency in an approximately 2-fold increase in compared to the control cells co-infected with a nonspecific shRNA (shSTD) (Fig. 1D,E). In contrast, the knockdown of Δ 133p53 resulted in a 2-fold decrease and the overexpression of Δ 133p53 showed a 4-fold increase in reprogramming efficiency (Fig. 1D,E). Combining the knockdown of p53 and the overexpression of Δ 133p53 resulted in a further 4-fold increase compared to the knockdown of p53 alone (Fig. 1D,E). These results demonstrate that Δ 133p53 promotes iPS cell reprogramming.
Overexpression of Δ133p53 inhibits apoptosis during reprogramming. To investigate whether the increase of iPS cell reprogramming efficiency is correlated with Δ 133p53's anti-apoptotic activity, we performed a fluorescence-activated cell sorting (FACS) analysis with anti-Annexin V antibody staining at 9 and 12 dpi. The results showed that the percentage of reprogramming cells undergoing apoptosis at 9 dpi was significantly increased more than 2 folds in the treatment with the knockdown of Δ 133p53, whereas the percentage of apoptotic cells was slightly decreased in the treatments with either the overexpression of Δ 133p53 or the knockdown of p53, compared to that in the control reprogramming cells co-infected with shSTD ( Fig. 2A,B). The analysis from 12 dpi showed that the percentage of apoptotic cells was 5.46% four-fold lower in the treatment with the overexpression of Δ 133p53 and was 10.86% two-fold lower in the treatment with the knockdown of p53, whereas the percentage of apoptotic cells was increased about 8% in the treatment with the knockdown of Δ 133p53, compared to that (20.23%) in the control treatment ( Fig. 2A,B). The percentage of apoptotic cells (9.01%) in the treatment with combining the knockdown of p53 and the overexpression of Δ 133p53 was higher than that in the treatment with the overexpression of Δ 133p53 alone, but still two-fold lower than that in the shSTD co-infected control group. Nevertheless, the results suggest that one of reasons for Δ 133p53 to promote reprogramming efficiency is inhibition of apoptosis.
Δ133p53 promotes DNA DSB repair in cell reprogramming. Previous reports showed that DNA DSB repair foci are formed during cell reprogramming 7,8 , which suggests cell reprogramming can induce DNA DSBs. Our recent finding demonstrated that Δ 133p53 promotes DNA DSB repair by upregulating the expression of RAD51, LIG4 and RAD52 21 . Therefore, we checked the protein accumulation of these three genes in reprogramming at 12 dpi using Western blot (Fig. 3A). The results showed that three DNA DSB repair genes, RAD51, LIG4 and RAD52 were all up-regulated at 12 dpi after reprogramming (Fig. 3A). The expression of these genes after reprogramming was down-regulated by the knockdown of Δ 133p53 and enhanced by overexpression of Δ 133p53 (Fig. 3A). The results suggested that cell reprogramming triggers DNA DSB response and Δ 133p53 may promote DNA DSB repair during cell reprogramming.
Next, we investigated the function of Δ 133p53 in the formation of the DNA DSB repair foci of phosphorylated H2AX (γ H2AX) and RAD51 at 9 and 12 dpi during reprogramming. γ H2AX is one of the early DNA DSB repair markers. RAD51 is a recombinase and required for HR repairs which executes high fidelity DNA repair by using the undamaged sister chromatid or homologous DNA as a template to faithfully repair the damage. Similar effects of Δ 133p53 on DNA damage repair were observed at both 9 and 12 dpi. The proportion of cells with RAD51 positive staining (including foci and pan-nuclear signals) increased approximately 2 to 3-fold with the overexpression of Δ133p53 and decreased almost 5 to 7-fold with the knockdown of Δ133p53, whereas the percentage of RAD51 positive cells was not significantly changed by the knockdown of p53, compared to that in the control cells co-infected with shSTD ( Fig. 3B,C). However, the proportion of cells with γ H2AX positive staining (including foci and pan-nuclear signals) was significantly increased by the knockdown of either p53 (about 2-fold) or Δ 133p53 (about 3-fold), but significantly decreased by overexpression of Δ 133p53 (about 3-fold), compared to the control (Fig. 3B,C). From these results, we speculated that Δ 133p53 protects iPS cell genomic stability by promoting DNA DSB repair.

Δ133p53 reduces chromosomal aberrations in iPS cells.
To confirm this speculation, we selected five independent iPS cell clones from each treatment and performed a chromosomal damage analysis at passage four using a karyotype assay. The characteristics of the selected iPS cell clones were confirmed by different iPS markers (Fig. S1A,B). Pluripotency of iPS cell clones from both of the control group and the treatment with co-expression of Δ 133p53 was identified by the analysis of teratoma formation (Fig. S2). Chromosomal aberration events, including chromosome breakages and end-to-end fusions, indeed increased 2-fold in the iPS cells with a p53 knockdown and 3-fold in cells with a Δ 133p53 knockdown, compared to that in shSTD infected controls (Fig. 4A,B). Strikingly, there were only half as many aberration events in the iPS cells overexpressing Δ 133p53 as in the shSTD infected controls, even though the reprogramming efficiency in the iPS cells overexpressing Δ 133p53 was increased 4-fold. The overexpression of Δ 133p53 significantly decreased the chromosomal aberration events caused by the knockdown of p53 in the iPS cells (Fig. 4A,B), which is consistent with that Δ 133p53 promotes DNA DSB repair independent of p53. These data demonstrate that the genetic quality of iPS cells can be improved by the overexpression of Δ 133p53.

Discussion
De novo genetic variants in iPS cells have been observed in many studies [1][2][3][4][5][6] . To minimize the genomic instabilities of iPS cells, strategies of generating integration-free iPS cells have been developed. However, iPS cells generated either with episomal vector or protein-base method were still found to carry a large number of de novo genetic variants 3,23 . One of the most important reasons for the de novo genetic variants in iPS cells is that reprogramming process can trigger DNA damage response. Therefore, faithful repairing DNA damages during reprogramming is very crucial for maintenace of genomic integrity. Tumour repressor p53, often known as the "guardian of the genome", is a key regulator in DNA damage response. It has demonstrated that p53 inhibits cell reprogramming by promoting reprogramming cells to undergo apoptosis and senescense 7,9,10 . When p53 is absent, reprogramming efficiency is significantly increased. However, the genetic quality of generated iPS cells is getting worse 7,13 .
In the last decade, p53 has been found to encode a large number of isoforms 24,25 . It has demonstrated that p53 isoforms can modulate p53 functions either synergistically or antagonistically 26 . Our recent studies showed that Δ 133p53, an N-terminal truncated p53 isoform, not only antagonizes p53-mediated apoptosis, but also promotes DNA DSB repair 21 . Here, we report that Δ 133p53 is induced during reprogramming. Δ 133p53 not only promotes efficiency of cell reprogramming by its anti-apoptotic function, but also ensures genetic stability by promoting DNA DSB repair. Our results imply that overexpression of Δ 133p53 during reprogramming may provide a solution for improving iPS genetic quality due to its ability to increase RAD51 foci formation and decrease γ H2AX foci formation and chromosome aberrations in iPS cells.

RNA Analysis and qRT-PCR.
For quantitative real-time reverse transcriptional PCR (qRT-PCR), total RNA was treated with DNaseI prior to reverse transcription and purified with RNeasy mini kit (QIAGEN). First strand cDNA was synthesized using M-MLV Reverse Transcriptase (Invitrogen). Reaction was performed in CFX96 TM Real-Time System (Bio-Rad) using SsoFast EvaGreen Supermix (Bio-Rad) according to the manufacturer's instructions. Total RNA was normalized with β-actin. Statistics was obtained from three repeat experiments. Primers sequences used are listed in Supplemental Table S1.
Immuno-blotting. For Western blotting, total protein was extracted using standard SDS sample buffer.
Western blotting was performed as described 21 .
Immunofluorescence-staining. To analyze RAD51 and γ H2AX foci formation in reprogamming cells, CCD-1079sk cells were reprogramed with Yamanaka 4 factors or combined with other factors as described in the section of Lentiviral transduction and reprogramming culture. At 9 and 12 dpi, cells were collected and washed with hES culture medium and then plated on a Coverglass For Growth (Fisher Scientific, FIS12-545-82) which were covered with gelatin. After 6 h of culture, cells were rinsed with PBS and incubated with Permeate Buffer (0.2% Triton X-100 in PBS) at room temperature for 3 min. Cells on coverslips were rinsed twice with ice-cold PBS and then fixed with 4% PFA (Sigma) on ice for 15 mins. Cells were washed twice with PBS, and permeabilized with PBST (0.2% triton X-100 in PBS) at room temperature (RT) for 15 mins. After blocking in FDB (0.2% Triton X-100, 2% donkey serum, 3% bovine serum albumin, 1 × PBS) for 30 mins at RT, the coverslips were incubated with primary antibody for 1 hour (h) at RT, followed by 3 × 3 mins washes with PBST. A secondary antibody (Invitrogen) (1:400 diluted in blocking solution) was added and incubated for a further 1 h at RT. After a total 3 rounds of washing with PBST quickly, the coverslips were mounted on slides with a mount medium containing DAPI (VectaShield). RAD51 polyclonal antibody (ct-1201, Cell Application) and γ H2AX S139 monoclonal antibody (#05-636, Millpore) were used for immunostaining. Total number of γ -H2AX/Rad51 positive cells were counted from randomly picked up 150 cells in each sample.
iPS colony immunostaining was performed as described above. The antibodies were used as follow: Anti-SOX2 (rabbit IgG, 1:1000 Lentiviral transduction and reprogramming culture. CCD-1079sk cells at passage 6 were cultured in DMEM medium supplemented with 10% FBS. 2.0 × 10 5 cells were transduced with a cocktail of lentivirus carrying 4 Yamanaka factors, or combined with a lentivirus shSTD, shp53, shΔ 133p53 and shp53 plus Δ 133p53 separately. Transduction medium were supplemented with 0.1% polybrene, and the day was defined as "0 Day post infection (dpi)". The infected cells were plated to a 6-well plate. At 24 hour post infection (hpi), the medium was changed to fresh DMEM medium (with 10% FBS). At 2 dpi, cells were transfered to a new 6-well plate covered with mouse embryonic fibroblast (MEF) feeder cells and cultured for another 3 days. The medium was replaced with human stem cell medium (hES medium; Invitrogen) for each of 2-days. After 12 dpi, the medium was substituted with a mixed medium consisting of hES medium, Condition Medium (CM; Invitrogen) and basic fibroblast growth factor (bFGF; Invitrogen) ( hES: CM = 1:1, bFGF 2 ng/ml) in each of 2-days. At 20 dpi, the formed iPS colonies were subjected to AP Staining. Around 25-30 dpi, the colonies were picked out for expansion growth with hES medium in a 48-well plate covered with MEF feeder cells. Finally, iPS colonies were cultured in a 25 cm 2 flask with feeder cells for other experiments or storage. The cryopreservation media for iPS colonies consisted of 20% qualified embryonic stem cell FBS (GIBCO), 70% hES medium and 10% DMSO. AP staining. At day 20 dpi, reprogramming cell colonies was stained with Alkaline Phosphotase (AP) Staining Kit (Sidansai) as manufacturer's instruction. AP positive colonies in each well were photographed with Sony W570 camera and the number of colonies in each well was counted for statistical analysis.
Karyotype analysis. At 25-30 dpi, more than 5 reprogramming cell colonies from each treatment were separately picked into a new 12-well plate for further expansion. At passage 4, part of cells of each colony were subjected to AP staining and immunostaining with different iPS marker genes. Five AP and iPS marker positive colonies from each treatment were selected for continuing culture. About 2 × 10 7 cells from each colony were sent to ADICON Clinical Lab INC (Hangzhou) for karyotype analysis. In each iPS clone, 25 metaphases and about 1000 chromosomes were observed. Average abnormal chromosome events came from 5 independent iPS clones in each treatment.
Teratoma formation. IPS cells (four factors, or four factors plus Δ 133p53) (10 6 cells) were subcutaneously injected into irradiated (4 Gy) nude mice (injections were performed 1 day after irradiation). Teratomas were surgically removed or after 9 weeks of injection. Tissue was fixed in formalin at 4°C, embedded in paraffin wax, and sectioned at a thickness of 5 mm. Sections were stained with haematoxylin and eosin for pathological examination.