Senescence impairs successful reprogramming to pluripotent stem cells

Abstract

Somatic cells can be reprogrammed into induced pluripotent stem (iPS) cells by overexpressing combinations of factors such as Oct4, Sox2, Klf4, and c-Myc. Reprogramming is slow and stochastic, suggesting the existence of barriers limiting its efficiency. Here we identify senescence as one such barrier. Expression of the four reprogramming factors triggers senescence by up-regulating p53, p16(INK4a), and p21(CIP1). Induction of DNA damage response and chromatin remodeling of the INK4a/ARF locus are two of the mechanisms behind senescence induction. Crucially, ablation of different senescence effectors improves the efficiency of reprogramming, suggesting novel strategies for maximizing the generation of iPS cells.

Figure 1.

Induction of senescence upon expression of the four reprogramming factors. (A) Timeline of the experiments in this figure and Figure 2. IMR90 human fibroblasts were infected with a polycistronic vector expressing the four factors (Oct4, Sox2, Klf4, and c-Myc). Cells were selected for 7 d. (B) Effect of the expression of the four factors on the growth of IMR90 cells. (C) Crystal violet-stained plates of IMR90 cells infected with the indicated vectors. (D) Expression of the four reprogramming factors causes a decrease in the percentage of IMR90 cells incorporating BrdU. (E) Effect of the vector expressing the four reprogramming factors over SA β-Gal activity and SAHF formation in IMR90 cells.

Figure 2.

Molecular analysis of senescence induced by expression of reprogramming factors. (A–D) Expression of the four reprogramming factors results in an increase in the percentage of cells positive for DNA damage (as measured by pST/Q IF), p53, p21^CIP1, and p16^INK4a. An explanation of the procedures used to calculate the percentage of positive cells is detailed in the Supplemental Material and in Supplemental Figure 4. (E,F) The ability of the four reprogramming factors to block proliferation of IMR90 cells is alleviated by coexpression of HPV E6 and/or HPV E7 or knockdown of p53, p16^INK4a, or p21^CIP1 using shRNAs as measured by crystal violet staining (E) or BrdU incorporation (F).

Figure 3.

The histone demethylase JMJD3 contributes to the regulation of the INK4a/ARF locus during RIS. (A) Cartoon showing the organization of the human INK4b/ARF/INK4a locus and primer sets used. (B,C) Expression of the four reprogramming factors results in a loss of H3K27me3 marks (B) and an increase in H3K4me3 marks (C) in the INK4b/ARF/INK4a locus. (D) The H3K27me3 histone demethylase JMJD3 is induced in response to the expression of the four reprogramming factors. (E) Up-regulation of JMJD3 transcript in response to the expression of the four reprogramming factors. (F) ChIP analysis showing an enrichment of JMJD3 and a loss of EZH2 in the INK4a promoter in response to expression of the four reprogramming factors.

Figure 4.

Role of p21^CIP1 during RIS. (A) Individual effect of reprogramming factors on the induction of DNA damage and the expression of p53 and p21^CIP1 as measured by immunofluorescence. The expression of c-Myc and Klf4 induces p53 up-regulation. The expression of c-Myc, Klf4, or Sox2 induces the up-regulation of p21^CIP1. Similar results were obtained in two independent experiments. (O) Oct4; (S) Sox2; (K) Klf4; (M) c-Myc. (B) Expression of miRNAs of the miR-302 cluster analyzed by TaqMan in IMR90 cells infected with vector or four factors (Oct4, Sox2, Klf4, and c-Myc), three iPS cells (iPS B6, B7, and B8), and H1 hES cells (hES). (C) Expression of the miR-302 cluster (302) alleviates OSKM-induced senescence. (Left) IMR90 cells were infected with the indicated vectors and, 10 d post-infection, BrdU incorporation was measured after a 16-h pulse was given. (Right) Crystal violet-stained plates are shown. (Ctrl.) Control vector. (D,E) miR-302a–d prevents the up-regulation of p21^CIP1 (D) and p130 (E) upon expression of reprogramming factors as analyzed by immunofluorescence.

Figure 5.

Inhibition of senescence improves reprogramming efficiency. (A) BJ human fibroblasts were infected with shRNAs targeting senescence effectors. Quantitative RT–PCR showing the levels of the transcripts for p53, CDKN1A (encoding for p21^CIP1), and INK4a (encoding for p16^INK4a). (B) The BJ fibroblasts described above were transduced with lentiviruses expressing Oct-4, Sox2, Klf4, and c-Myc and grown in culture conditions compatible with pluripotent stem cell growth. Colonies were analyzed for NANOG and TRA-1-60. The number of positive colonies is shown as the mean ± SD of three representative experiments. (C) Table summarizing the results of quantitative RT–PCR analysis of hES cell markers in iPS cells. (Green) Positive; (yellow) negative; (n.d.) nondetermined. The graphs of selected quantitative RT–PCR of this experiment are shown in Supplemental Figure 15. (D) hiPS cell lines generated upon knockdown of p53 or p16 express pluripotency markers. (E) hIPS cells generated upon knockdown of p53 or p16 can differentiate into extraembryonic tissues and into derivatives of the three germ layers.