Title: Hypoxia Enhances the Generation of Induced Pluripotent Stem Cells
Abstract: Mouse and human somatic cells can be reprogrammed to induced pluripotent stem cells (iPSCs) by the transduction of four transcription factors, Oct 3/4, Sox2, Klf4, and c-Myc (Maherali et al., 2007Maherali N. Sridharan R. Xie W. Utikal J. Eminli S. Arnold K. Stadtfeld M. Yachechko R. Tchieu J. Jaenisch R. et al.Cell Stem Cell. 2007; 1: 55-70Abstract Full Text Full Text PDF PubMed Scopus (1412) Google Scholar, Meissner et al., 2007Meissner A. Wernig M. Jaenisch R. Nat. Biotechnol. 2007; 25: 1177-1181Crossref PubMed Scopus (657) Google Scholar, Okita et al., 2007Okita K. Ichisaka T. Yamanaka S. Nature. 2007; 448: 313-317Crossref PubMed Scopus (3532) Google Scholar, Takahashi et al., 2007Takahashi K. Tanabe K. Ohnuki M. Narita M. Ichisaka T. Tomoda K. Yamanaka S. Cell. 2007; 131: 861-872Abstract Full Text Full Text PDF PubMed Scopus (14962) Google Scholar, Takahashi and Yamanaka, 2006Takahashi K. Yamanaka S. Cell. 2006; 126: 663-676Abstract Full Text Full Text PDF PubMed Scopus (18861) Google Scholar, Wernig et al., 2007Wernig M. Meissner A. Foreman R. Brambrink T. Ku M. Hochedlinger K. Bernstein B.E. Jaenisch R. Nature. 2007; 448: 318-324Crossref PubMed Scopus (2217) Google Scholar). Patient or disease-specific human iPSCs could be used for studying pathogenesis, or potentially also to treat patients suffering from incurable diseases by transplanting the regenerated grafts derived from their own cells. However, the low induction efficiency and high tumorigenesis rate due to the use of proto-oncogenes, such as c-Myc, continue to hinder the clinical application of iPS technology. Many efforts have been made to find other factors or small molecules that facilitate the reprogramming process (Huangfu et al., 2008Huangfu D. Maehr R. Guo W. Eijkelenboom A. Snitow M. Chen A.E. Melton D.A. Nat. Biotechnol. 2008; 26: 795-797Crossref PubMed Scopus (1309) Google Scholar, Shi et al., 2008bShi Y. Do J.T. Desponts C. Hahm H.S. Scholer H.R. Ding S. Cell Stem Cell. 2008; 2: 525-528Abstract Full Text Full Text PDF PubMed Scopus (590) Google Scholar). In this study, we show that conducting reprogramming in hypoxic conditions results in improved efficiency for both mouse and human cells. Somatic stem cells reside in specific microenvironments, called niches, and environmental changes, such as stromal cell contacts, extracellular matrix proteins, temperature, and O2 tension, have a great influence on stem cell function and differentiation. Notably, low O2 tension promotes the survival of neural crest cells and hematopoietic stem cells and prevents differentiation of human ESCs (Danet et al., 2003Danet G.H. Pan Y. Luongo J.L. Bonnet D.A. Simon M.C. J. Clin. Invest. 2003; 112: 126-135Crossref PubMed Scopus (288) Google Scholar, Ezashi et al., 2005Ezashi T. Das P. Roberts R.M. Proc. Natl. Acad. Sci. USA. 2005; 102: 4783-4788Crossref PubMed Scopus (682) Google Scholar, Morrison et al., 2000Morrison S.J. Csete M. Groves A.K. Melega W. Wold B. Anderson D.J. J. Neurosci. 2000; 20: 7370-7376Crossref PubMed Google Scholar). Moreover, mammalian embryonic epiblasts reside in a physiologically hypoxic environment. These observations led us to the hypothesis that hypoxic conditions might promote the reprogramming process and thus iPS cell generation. To quantify the effect of hypoxia on iPS cell generation, we performed comparison experiments on mouse embryonic fibroblasts (MEFs) carrying the Nanog-GFP-Ires-Puror cassette (Okita et al., 2007Okita K. Ichisaka T. Yamanaka S. Nature. 2007; 448: 313-317Crossref PubMed Scopus (3532) Google Scholar). Four or three transcription factors (Oct3/4, Sox2, Klf4, +/− c-Myc) were introduced into MEFs with retroviral vectors. Four days after transduction, the cells were trypsinized and seeded onto the feeder layer of mitomycin C-treated STO cells. The cells were cultivated under 21%, 5%, or 1% O2 from day 5 to day 14 after transduction. GFP+ iPS cell colonies could be first detected on days 10–14 after viral transduction, and we counted the number of GFP-positive colonies on days 21 and 28 after transduction. Under 5% O2, the GFP-positive colonies derived from four-factor transduced MEFs increased 7.4-fold on day 21 and 3.1-fold on day 28, and those from three-factor transduced MEFs increased 20-fold on day 21 and 7.6-fold on day 28 under 5% O2 (Figures 1Aa and 1Ab). Moreover, hypoxic treatment increased the percentage of GFP-positive colonies in total colonies from four- or three-factor transduced MEFs (Figures 1Ac and 1Ad). The GFP-positive colonies derived after hypoxic treatment was comparable in morphology and size to those derived under normoxic conditions (Figure S1 available online). Alkaline phosphatase staining showed that cultivation under 5% O2 increased the number of colonies with a positive alkaline phosphatase activity (Figure S2). We also examined whether GFP-positive cells were detected more quickly under hypoxic conditions. The four-factor transduced MEFs were cultivated under 21% O2 or under 5% O2 with or without VPA from day 5 to day 9 after transduction and were subjected to flow cytometric analysis on day 9. Retroviral expression of four factors induced 0.01% of the cells to become GFP-positive on day 9 after transduction. Treating the four-factor transduced MEFs for 4 days with hypoxia or with VPA increased the percentage of GFP-positive cells to 0.40% and 0.48%, respectively. Moreover, cotreatment with hypoxia and VPA increased the percentage of GFP-positive cells to 2.28%. These data suggest that GFP-positive cells can be detected earlier and that the effect of hypoxic culture synergizes with VPA (Figures 1Ba–1Bd). Although neural stem cells that express SOX2 endogenously can be reprogrammed to iPS cells with the transduction of Oct3/4 and Klf4 (Kim et al., 2008Kim J.B. Zaehres H. Wu G. Gentile L. Ko K. Sebastiano V. Arauzo-Bravo M.J. Ruau D. Han D.W. Zenke M. et al.Nature. 2008; 454: 646-650Crossref PubMed Scopus (796) Google Scholar), the reprogramming of MEFs to iPS cells rarely occurs with two transcription factors of Oct3/4 and Klf4. Recently, small-molecule compounds have been reported to enable the reprogramming of Oct3/4 and Klf4-transduced MEFs to iPS cells (Shi et al., 2008aShi Y. Desponts C. Do J.T. Hahm H.S. Scholer H.R. Ding S. Cell Stem Cell. 2008; 3: 568-574Abstract Full Text Full Text PDF PubMed Scopus (721) Google Scholar). We examined whether hypoxic conditions could enhance MEFs to be reprogrammed to iPS cells with Oct3/4 and Klf4 transduction. Figure 1Ae shows an increased efficiency of the iPS cell generation derived from MEFs with Oct3/4 and Klf4 (MEF-2F-iPS) under 5% O2 in comparison to 21% O2. To further evaluate the pluripotency of the iPS cells generated with hypoxic treatment, we randomly picked up and established multiple iPS cell lines from two-, three-, and four-factor-infected MEFs. These iPS cells exhibited typical ES cell morphology. We examined the karyotype of iPS cell lines derived after hypoxic treatment (521AH5-1 and 527CH5-1), and these cell lines showed normal karyotypes (Figure S3). We investigated the mRNA expression of pluripotency-related genes in the iPS cells generated after hypoxic treatments. The mRNA expression patterns of these iPS cells were comparable to those of ESCs (Figure S4). When injected into nude mice subcutaneously, all of the established iPS cell lines gave rise to teratomas with histologic evidence of cells differentiating into all three germ layers (Figure S5). Moreover, MEF-2F-iPS cells derived under hypoxic conditions contributed to the formation of somatic tissue in chimeric mice (Figures 1Ca and 1Cb), but we have not yet obtained germline transmission with these mice, so the extent of reprogramming is not entirely clear. Previous studies have shown that iPSCs generated with the same three or four factors are capable of germline transmission (Nakagawa et al., 2008Nakagawa M. Koyanagi M. Tanabe K. Takahashi K. Ichisaka T. Aoi T. Okita K. Mochiduki Y. Takizawa N. Yamanaka S. Nat. Biotechnol. 2008; 26: 101-106Crossref PubMed Scopus (2200) Google Scholar, Okita et al., 2007Okita K. Ichisaka T. Yamanaka S. Nature. 2007; 448: 313-317Crossref PubMed Scopus (3532) Google Scholar). Recently, it was reported that iPS cells could be established by other methods than retroviruses or lentiviruses. We reported that transient transfection of expression plasmid vectors of four reprogramming factors could reprogram MEFs to iPS cells (Okita et al., 2008Okita K. Nakagawa M. Hyenjong H. Ichisaka T. Yamanaka S. Science. 2008; 322: 949-953Crossref PubMed Scopus (1599) Google Scholar). It was also reported that MEFs could be reprogrammed by transcription factors delivered by piggyBac (PB) transposition system (Kaji et al., 2009Kaji K. Norrby K. Paca A. Mileikovsky M. Mohseni P. Woltjen K. Nature. 2009; 458: 771-775Crossref PubMed Scopus (1039) Google Scholar, Woltjen et al., 2009Woltjen K. Michael I.P. Mohseni P. Desai R. Mileikovsky M. Hamalainen R. Cowling R. Wang W. Liu P. Gertsenstein M. et al.Nature. 2009; 458: 766-770Crossref PubMed Scopus (1444) Google Scholar). The PB insertions can be removed from established iPS cells. These methods minimize the potential for insertional mutagenesis. We examined whether hypoxia could improve the efficiency of iPS cell generation with plasmid vectors and with PB transposition system. Figure 1D shows that hypoxic cultivation significantly increased the number of GFP-positive colonies with transient transfection of plasmid vectors, and Figure 1E shows that hypoxic treatment for 5 and 10 days increased the number of GFP-positive colonies with the piggyBac transposition system by 2.9-fold and 4.0-fold, respectively. These data suggest that hypoxia can increase the efficiency of iPS cell generation by nonviral vectors such as plasmid expression vectors or the piggyBac transposition system. We next examined the effect of hypoxic culture on proliferation, survival, and gene expression. Flow cytometric analysis with annexin V demonstrated that hypoxic culture had no protective effect on mouse ESCs or on four-factor transduced MEFs (Figure S6). Furthermore, hypoxic cultivation showed no effect on proliferation of mouse ESCs (Figure S7). Although hypoxic incubation from day 1 to day 4 after transduction had no significant effect on proliferation of mock-transduced MEFs, it had significant effect on four-factor-transduced MEFs (Figure S8). To investigate the expression profile of cells in reprogramming process, we performed microarray analysis and quantitative real-time RT-PCR. We examined the expression profile of MEFs cultivated under hypoxic and normoxic conditions for 10 days. Microarray analysis shows that 57.2% of ESC-specific genes were upregulated and 67.5% of MEF-specific genes were downregulated in the MEFs cultivated under 5%O2 (Figures S9A and S9B). In Figures S9C and S9D, microarray analysis of four-factor-transduced MEFs cultivated under hypoxic and normoxic conditions from day 1 to day 4 showed that 73.2% of ESC-specific genes (765 genes out of 1045 total genes) were upregulated and 85.8% of MEF-specific genes (980 genes out of 1142 total genes) were downregulated in the cells treated with hypoxia. Moreover, quantitative real-time RT-PCR analysis demonstrated that expression of endogenous Oct3/4 and Nanog increased 3.4-fold and 2.1-fold, respectively, in four-factor-transduced MEFs after 3 days of hypoxic treatment (Figures S9E and S9F). To rule out the possibility that hypoxia enhances iPS cell generation by stimulating STO cells, we examined growth situation of iPS cells under hypoxic cultivation without the feeder layer of STO cells. Figure S10 shows that cultivation under 5%O2 increased the number of GFP-positive colonies, suggesting that hypoxic enhancement of reprogramming was not mediated by STO cells. We next examined whether the exposure to hypoxia increases the efficiency of iPS cell generation from human somatic cells. The four transcription factors were introduced into adult human dermal fibroblasts (HDFs) by retroviral vectors. At six days after transduction, the cells were trypsinized and seeded onto the feeder layer of mitomycin C-treated STO cells. The cells were cultivated under 5% O2 for 7 (1w), 14 (2w), 21 (3w), or 33 days (Long), and the number of human ESC-like colonies was counted on day 24, 32, and 40 after transduction (Figure 2A). Figure 2B shows that 14 day and 21 day hypoxic cultivation increased the efficiency of iPS cell generation by 4.2-fold and 3.6-fold on day 24, by 2.8-fold and 3.0-fold on day 32, and by 2.6-fold and 2.5-fold on day 40, respectively. We randomly selected and established several clones of human iPS cells derived under hypoxic conditions. All of the human iPS cell lines had a typical ESC morphology and were strongly positive for alkaline phosphatase while also expressing pluripotent-related gene markers (Figures 2Ca, 2Cb, and 2D). Moreover, immunocytological staining showed that all of the cell lines expressed SSEA3, SSEA4, and Nanog (Figures 2Cc–2Ce). To investigate the differentiation ability of the human iPS cells derived under hypoxic conditions, we used floating cultivation to form embryoid bodies (EBs). After 8 days, the iPS cells formed round embryoid bodies and we then transferred the EBs to gelatin-coated plates and cultivated them for another 8 days. Immunocytochemical analysis showed that for all the iPS cell lines, attached cells derived from the EBs were positive for alpha-fetoprotein (endoderm), alpha-smooth muscle actin (mesoderm), glial fibrillary acidic protein (ectoderm), and beta-3 tubulin (ectoderm) (Figures 2Cf–2Ci). To evaluate pluripotency in vivo, we transplanted the human iPS cells into the testes of SCID mice. All of the established human iPS cell lines derived after hypoxic treatment developed teratomas and the histological study showed the cells in the teratomas to differentiate into tissues representing all three germ layers (Figure 2E). Although hypoxic conditions promote reprogramming, hypoxia also induces cytotoxicity. There are significant differences between cell types in terms of their susceptibility to hypoxia. In our experiments, HDFs were more susceptible to hypoxic cytotoxicity than MEFs. Cultivation under 1% O2 inhibited the proliferation of HDFs and even led cell death within several days, whereas cultivation under 1% O2 had little effect on the proliferation of MEFs. In our experiments, hypoxic cultivation showed no significant effects on the survival of mouse ESCs and four-factor-transduced cells or on the proliferation of mouse ESCs and mock-transduced MEFs. However, in four-factor-transduced MEFs, hypoxia showed a significant proliferative effect and increased the expression level of Oct3/4 and Nanog. In addition, exposure of MEFs to hypoxic conditions shifted the overall gene expression pattern toward that of ESCs. Although there may be several explanations for the positive effect of hypoxia on the efficiency of reprogramming, these results suggest that hypoxic conditions may contribute to the reprogramming process itself. In this study, we created hypoxic conditions by flushing hypoxic gas mixture, by using Forma Series II Universal CO2 incubators (Thermo Scientific), in which mild hypoxia (5%–6% O2) in a gas phase can be achieved within 10 min after opening and closing of the door. However, because we changed the medium in a laminar flow hood under normoxic atmosphere, there must have been some fluctuation in O2 content after medium change. More strict control of hypoxia, with a closed hypoxia workstation or medium pre-equilibrated under hypoxic conditions, might further increase the efficiency of iPS cell generation. In summary, by comparing the efficiency of iPS cell induction under normoxic and hypoxic conditions, we showed that hypoxic conditions can improve the efficiency of iPS cell generation from mouse and human somatic cells. We have found that cultivation under 5% O2 favors more efficient iPSC generation, but further characterization is needed to determine the optimal O2 concentration and duration of hypoxic culture for promoting reprogramming process. Ultimately, we hope that understanding the basis of this effect of hypoxia will contribute to ongoing efforts to devise a method for efficient iPSC that does not require genetic modification. We thank Drs. Masato Nakagawa, Takashi Aoi, Michiyo Koyanagi, and Koji Tanabe and other members of our laboratory for their valuable scientific comments and fruitful discussions; Tetsuya Ishii and Kanon Takeda for their critical reading of the manuscript; and Drs. Jun K. Yamashita and Masataka Fujiwara for assistance in the cultivation of human ESCs. We are also grateful to Aki Okada, Nanako Takizawa, Misato Nishikawa, and Megumi Kumazaki for technical support and Rie Kato, Ryoko Iyama, Noriyo Maruhashi, and Eri Nishikawa for administrative support. We also thank Dr. Robert Farese, Jr for RF8 ESCs, Dr. Toshio Kitamura for the Plat-E cells and pMX retroviral vectors, Dr. Andras Nagy for piggyBac vectors, and Dr. Malcolm J. Fraser for providing pBSII-IFP2-orf. This study was supported in part by a grant from the Program for Promotion of Fundamental Studies in Health Sciences of NIBIO, a grant from the Leading Project of MEXT, and Grants-in-Aid for Scientific Research of JSPS and MEXT (to S.Y. and Y.Y.). Download .pdf (.49 MB) Help with pdf files Document S1. Ten Figures and Supplemental Experimental Procedures PO2 Matters in Stem Cell CultureWion et al.Cell Stem CellSeptember 04, 2009In BriefAbout a century ago, conditions were worked out for maintaining growing tissue and cells outside the body. From the beginning, care was taken to maintain cultures at a physiological temperature, and to include precise concentrations of salts and other compounds, but the oxygen concentration in the culture medium was simply the result of letting the medium equilibrate with air. This approach was a reasonable first approximation, given that values of partial pressure of oxygen (PO2) in animal tissues were not measured until over a decade later, and all that mattered seemed to be to provide cells with “enough” oxygen. Full-Text PDF Open Archive