By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
Practical clinical applications for current induced pluripotent stem cell (iPSC) technologies are hindered by very low generation efficiencies. Here, we demonstrate that newborn human (h) and mouse (m) extra-embryonic amnion (AM) and yolk-sac (YS) cells, in which endogenous KLF4/Klf4, c-MYC/c-Myc and RONIN/Ronin are expressed, can be reprogrammed to hiPSCs and miPSCs with efficiencies for AM cells of 0.02% and 0.1%, respectively. Both hiPSC and miPSCs are indistinguishable from embryonic stem cells in colony morphology, expression of pluripotency markers, global gene expression profile, DNA methylation status of OCT4 and NANOG, teratoma formation and, in the case of miPSCs, generation of germline transmissible chimeric mice. As copious amounts of human AM cells can be collected without invasion, and stored long term by conventional means without requirement for in vitro culture, they represent an ideal source for cell banking and subsequent ‘on demand’ generation of hiPSCs for personal regenerative and pharmaceutical applications.
Induced pluripotent stem cells (iPSCs) have been generated through nuclear reprogramming of somatic cells via retrovirus or lentivirus-mediated transduction of exogenous reprogramming factors Oct4, Sox2, Klf4 and C-Myc (Yamanaka 2007). This has led to greatly enhanced promise for exploring the causes of, and potential cures for, many genetic diseases, as well as increased promise for regenerative medicine. Improvements in delivery methodology have further facilitated iPSC generation by minimizing the requirement for genetic modification (Feng et al. 2009). Notably, generation of genetic modification-free iPSCs with reprogramming proteins (Kim et al. 2009; Zhou et al. 2009) suggests regenerative medicine with personal iPSCs could soon be realized. However, the markedly low efficiency of iPSC generation, with all adult somatic cell types tested to date, remains problematic (Wernig et al. 2008). Technological advancements in this field have mainly been achieved using mouse embryonic fibroblasts (MEFs), in which the efficiency of iPSC generation is 10–100 times higher than that with adult somatic cells (Yu et al. 2007; Wernig et al. 2008). Therefore, current methods would appear to be less than ideal for generating iPSCs from adult somatic cells.
Here, to find nuclear reprogramming-sensitive cells collectable with no risk by physical invasion, we generated iPSCs from human and mouse newborn extra-embryonic membranes, amnion (AM) and yolk sac (YS), which consist huge amounts of discarded cells after birth. Interestingly, the efficiency of mouse iPSC (miPSC) generation from the AM was comparable to that of MEFs by retroviral transduction with Oct4, Sox2, Klf4 and c-Myc. Importantly, human iPSC (hiPSC) is also efficiently generated from human AM cells. Expression of the endogenous KLF4/Klf4, c-MYC/c-Myc and RONIN/Ronin in human/mouse AM cells may function in facilitating the generation efficiency of iPSCs. The human AM cell, which is conventionally freeze-storable, could be a useful cell source for the generation of pluripotent stem cells including iPSCs mediated by nuclear reprogramming in the purpose of personal regenerative and pharmaceutical cure in the future of infants.
Generation of iPSCs from mouse AM and YS cells
Extra-embryonic membranes, AM (amniotic ectoderm and mesoderm layers) and YS (visceral yolk sac endoderm and mesoderm layers) express a high level of proto-oncogene (Curran et al. 1984) which function, at least in part, to maintain and protect the fetus in utero. In E18.5 mouse embryos just before birth, AM and YS can be easily recognized microscopically (Fig. 1a). The membranes were dissected from Oct4-GFP (OG)/Neo-LacZ (Rosa26) embryos as approximately 5–10 mm2 sections and digested with collagenase. Isolated cells were cultured for 4–5 days resulting in morphologically heterogeneous populations (Fig. 1a) in which OG expression was undetectable. Approximately 1 × 105 cells were then retrovirally transfected with exogenous Oct4, Sox2, Klf4 and c-Myc (OSKM). After approximately 3 weeks, OG-positive embryonic stem cell (ESC)-like miPSC colonies were picked and expanded without drug selection. All AM (female) and YS (male)-miPSC lines generated here, which closely resembled ESCs in morphology (Fig. 1a), had a 2n = 40 normal karyotype (data not shown).
Characterization of AM and YS-miPSCs
As with ESCs, all AM- and YS-miPSC colonies were positive for alkaline phosphatase (ALP) (Fig. 1b). Immunohistochemical analyses also demonstrated that the cells were positive for pluripotent cell-specific nuclear proteins Oct4 and Nanog, and the surface glycoprotein SSEA1 (Fig. 1b). Thus, the expression profile of all marker proteins tested in AM and YS-miPSCs was similar to that observed in ESCs.
To examine the global transcription profile of these cells, comparative Affymetrix gene expression microarray analyses were performed between AM cells, YS cells, YS-miPSCs and R1 ESCs (Fig 1c). The global gene expression profile of YS-miPSCs was significantly different from that of YS cells. We detected a similar behavior between AM-miPSCs and AM cells (data not shown). Notably, the profile was similar to that of ESCs (Fig. 1c). Together, the data indicate that significant global nuclear reprogramming had occurred in these cells in response to OSKM transfection. We next applied RT-PCR analysis to gain a more focused transcriptional profile of pluripotent cell-specific marker genes in the induced cells. We found that Nanog, Rex1, ERas, Gdf3, Zfp296 and Ronin were expressed in both AM and YS-miPSCs, whereas the AM and YS genes, Igf1 and Ccl6 were silenced (Fig. 1c). Notably, Ronin was expressed not only in AM and YS-miPSCs but also in the precursor AM and YS cells. To investigate whether the exogenous Oct4, Sox2, Klf4 and c-Myc genes were silenced by DNA methylation as reported for other iPSCs (Jaenisch & Young 2008) in the AM and YS-miPSCs, we examined expression using gene-specific primer sets designed to distinguish endogenous and exogenous transcripts. In all miPSC lines, the expression of endogenous Oct4, Sox2, Klf4 and c-Myc was similar to that in R1 ESCs, whereas the exogenous c-Myc and Klf4 were fully silenced in some YS-miPSC clones but not in others (Fig. 1c). Notably, high-level expression of endogenous Klf4 and c-Myc was detected even in AM and YS cells, consistent with the expression of proto-oncogene (Curran et al. 1984). Endogenous expression of Klf4, c-Myc and Ronin genes that are involved in maintaining pluripotency may play a key function in enhancing the generation efficiency of miPSCs from AM and YS cells.
Timing and efficiency of miPSC generation
The molecular mechanisms that govern OKSM-induced nuclear reprogramming of somatic cells to iPSCs are poorly understood. It has been demonstrated that activation of endogenous Oct4 may be a landmark for irreversible epigenetic transition toward fully reprogrammed iPSCs (Sridharan & Plath 2008). Thus, the timing of reactivation of OG is closely linked with the efficiency of reprogramming. Activation of exogenous OG was detected in some cell populations in every colony around 10 days after OSKM transfection of AM and YS cells, similar to control MEFs examined here and those reported previously (Fig. 1d) (Brambrink et al. 2008). The reprogramming efficiency of AM and YS cells was estimated by ALP-staining 21 days after OSKM transfection with reseeding at day 4. Notably, the number of ALP-positive colonies was similar between AM cells (4373 ± 983; mean ± SEM, n = 3) and MEFs (4997 ± 1049, n = 3), and ∼50% in YS cells (2293 ± 487, n = 3). Thus, the efficiency of AM reprogramming by OSKM is comparable to that of MEFs, and far exceeds that of adult somatic cells (Fig. 1d).
Germline-transmissible chimeras with AM and YS-miPSCs
To address in vivo differentiation potential of the AM and YS-miPSCs, approximately 10 agouti miPSCs were microinjected into C57BL/6J × BDF1 blastocysts (black), and transferred into white ICR foster mothers to generate chimeras. Three male YS-miPSC and two female AM-miPSC lines were tested for chimera formation. X-gal staining analysis on sections of E15.5 embryos demonstrated successful generation of normally developing chimeric embryos with OG/Neo-LacZ miPSC contribution to the majority of tissues in all miPSC lines examined (data not shown). We next examined the miPSC potential for normal growth to sexual maturity and germline transmission. Two high-degree chimeric mice with a YS-miPSC line and three high-degree chimeric mice with two AM-miPSC lines, characterized by the >50% contribution of agouti coat color (Fig. 2a), developed normally into adulthood. However, an adult YS-miPSC chimera developed a neck tumor around 8–10 weeks after birth, which may be due to reactivation of the exogenous c-Myc as reported previously (Nakagawa et al. 2008). Testes isolated from affected males were bisected and one-half was X-gal-stained for LacZ activity whereas the other half was cryosectioned. Blue staining in the seminiferous tubule indicated that YS-miPSCs could contribute to germ cell development. To confirm this, testis cryosections immunohistochemically stained with antibodies against LacZ (iPSC-derived cell marker) and TRA98 (spermatogonia and spermatocyte marker) (Fig. 2b). Germ cells in all tubules were positive for TRA98, whereas germ cells in only some seminiferous tubules were positive for LacZ, clearly demonstrating that YS-miPSCs are capable of contributing to the differentiating germ line in chimeras. Finally, to examine whether the genetic information of YS-miPSCs was transmissible to the next generation, DNA isolated from progeny of the remaining YS-miPSC chimera was analyzed by genomic PCR with a primer set specific to Neo. Seven of the thirty-five pups examined were positive, demonstrating that YS-miPSCs are able to differentiate into fully functional germ cells (Fig. 2c). In one of three female AM-miPSC chimeric mice, competence for contribution to germ cells was detected by X-gal staining analysis of ovaries (data not shown).
Teratoma formation with AM and YS-miPSCs
The differentiation competence of AM and YS-miPSCs was further tested by teratoma formation induced by injection of cells into the inguinal region of immunodeficient SCID mice. Teratomas were isolated 5–8 weeks after for histological analysis and for gene expression analysis. Hematoxylin–eosin (HE) staining of paraffin sections demonstrated that the three primary layers were generated as morphologically shown by ectodermal glia and neuroepithelium, mesodermal muscle and endodermal ciliated epithelium and cartilage (Fig. 2d). Multi-lineage differentiation of miPSCs was verified by transcription of endodermal, mesodermal and ectodermal genes in the majority of teratomas (Fig. 2e).
Generation of iPSCs from human AM cells
To examine whether hiPSCs could be efficiently generated from primary AM cells isolated from the amniotic membrane (∼100 cm2) of the placenta of newborn human (Fig. 3a), the reprogramming factors OCT4, SOX2, KLF4 and c-MYC were introduced by vesicular stomatitis virus G glycoprotein (VSV-G) retroviral transduction. About 20 AM-hiPSC lines were established from 1.0 × 105 AM cells infected (0.02%). The efficiency of AM-hiPSC generation is markedly high relative to that with cells from human adult tissues (Yu et al. 2007). AM-hiPSCs were morphologically similar to human ESCs (hESCs) (Fig. 3a). Immunohistochemical analyses demonstrated expression of the pluripotent cell-specific nuclear proteins OCT4, SOX2 and NANOG, and the keratan sulfate proteoglycan TRA-1-60 (Fig. 3b) consistent with the profile observed in hESCs. To extend this analysis, we examined the expression profile of genes by RT-PCR. The endogenous reprogramming factor genes OCT4, SOX2, KLF4 and c-MYC were all activated in AM-hiPSCs, whereas the transgenes were fully silenced (Fig. 3c). Expression of pluripotent cell-specific genes NANOG, REX1, GDF3, ESG1, FGF4, TERT and RONIN were also activated in all AM-hiPSC clones consistent with the profile of control hESCs (Fig. 3c). Notably, transcription of KLF4, c-MYC, and RONIN was detected not only in AM-hiPSCs but also AM cells. Similar to mouse AM and YS cells, endogenous expression of KLF4, c-MYC and RONIN in human AM cells may facilitate acquisition of reprogramming competency for efficient generation of hiPSCs.
DNA methylation of OCT4 and NANOG in AM-hiPSCs
To further characterize the pluripotent nature of AM-hiPSCs, the promoter CpG methylation status of key pluripotency genes was examined by bisulfite-modified DNA sequencing. Promoters of both OCT4 and NANOG were found to highly methylated in hAM cells, consistent with transcriptional silencing in these cells. Conversely, both promoter regions were hypo-methylated in AM-hiPSCs consistent with the observed reactivation (Fig. 3d). These data demonstrate that human AM cells are capable of being epigenetically reprogrammed to AM-hiPSCs through forced expression of reprogramming factors.
Teratoma formation with AM-hiPSCs
To address whether the AM-hiPSCs have competence to differentiate into specific tissues, teratoma formation was induced by implantation under the kidney capsule of immunodeficient nude mice. Twenty-one out of twenty-four AM-hiPS independent clones induced teratoma formation within 6–10 weeks of implantation (1.0 × 107 cells/site). Histological analysis by HE staining of paraffin-embedded sections demonstrated that the three primary layers were generated as shown by ectodermal glia and neuroepithelium, mesodermal muscle and endodermal ciliated epithelium and cartilage morphologically (Fig. 3e). Thus, the majority of AM-hiPSC clones have potential for multi-lineage differentiation in vivo.
We here demonstrated that hiPSCs and miPSCs were efficiently generated from newborn AM cells, in which endogenous Klf4, c-Myc and Ronin were highly expressed. The generation efficiency of miPSCs from AM cells was comparable to that from MEFs in mice and was notably high to that from adult somatic cells in humans. The properties of AM-hiPSCs and AM or YS-miPSCs resemble those of fully reprogrammed iPSCs from other tissues and ESCs.
iPSCs are generated through epigenetic reprogramming of somatic cells. Information on the base sequence of DNA in nuclei is unchanged through the reprogramming, although the gene expression profile is altered through the reprogramming from the somatic cell to the iPSC type. Developmentally rewound iPSCs retain aged DNA base sequence information inherited from somatic cells. The base sequence of DNA accumulates mutations through aging with cell division and mis-repair. Young somatic cells are suitable for iPSC generation rather than aged somatic cells. Therefore, it is suggested that the AM cells accumulating less genetic mutation are safer than the adult somatic cells as a cell source for iPSC generation.
The generation efficiency of OG-positive colonies was approximately four times lower than that of ALP-positive colonies and it is likely that miPSC generation will be further reduced (Wernig et al. 2008). Furthermore, when pre-iPSCs are reseeded, the generation efficiency of iPSC outcome could be roughly estimated as 1/2X (X = reseeded day after infection or transfection; doubling time of pre-iPSC is estimated as 24 h). Recently, iPSC generation technology has been developed and improved with MEFs and human embryonic or newborn fibroblasts (HNFs) as representative somatic cells. Even with these types of cells, application of the current technology resulted in a marked decrease in iPSC generation efficiency. The retroviral transduction-mediated miPSC generation efficiency is 0.05–0.1% with MEFs (Takahashi et al. 2007; Wernig et al. 2007). The generation efficiency of hiPSCs (∼0.01% in ALP-positive colony and 0.0025% in hiPSC outcome) (Yu et al. 2007; Wernig et al. 2008) is ∼10 times lower than that of miPSCs. The generation efficiency of genetic modification-free hiPSCs from HNFs by direct delivery of reprogramming proteins is estimated at about 0.001% in outcome (Kim et al. 2009). Notably, it is evident that the generation of hiPSCs from adult somatic cells is much harder than that from MEFs. In fact, analysis with a secondary dox-inducible transgene system shows that the efficiency varies between different somatic cell types (Wernig et al. 2008). Thus, for practical application of iPSC technology to medical care, identification of reprogramming-sensitive cell types is a key issue. Human primary keratinocytes are one candidate cell type for efficient generation of hiPSCs from adult patients (the efficiency of ALP-positive colony = 1.0%) (Aasen et al. 2008). Here, we have shown that human and mouse AM cells, in which the endogenous KLF4/Klf4, c-MYC/c-Myc and RONIN/Ronin are naturally expressed, are highly reprogramming-sensitive (hiPSC generation efficiency was approximately 0.02% in outcome). An important point is that relatively huge amounts of human AM cells can be collected from discarded AM membranes at birth with no risk to the individual. Furthermore, these cells can be kept in long-term storage without requirement for amplification by in vitro cell culture.
Our findings illustrate that human AM cells are a strong candidate cell source for collection and banking that could be retrieved on demand and used for generating personalized genetic modification-free iPSCs applicable for clinical treatment and drug screening.
Amnion and yolk sac cells
In mice, AM and YS membranes collected from E18.5 embryos from GOF-18/delta PE/GFP (Oct4-GFP) transgenic females (Yoshimizu et al. 1999) mated with 129/Rosa26 transgenic males (Friedrich & Soriano 1991) were digested with 0.1% collagenase (Wako, Osaka, Japan) and 20% fetal bovine serum (FBS) at 37 °C for 1 h, and then repeatedly passed through a 26-gauge needle. The cell suspension was cultured with mES medium (DMEM/F12 (Dulbecco’s modified Eagle’s medium/Ham’s F12) (Wako) supplemented with 15% FBS, 10−4m 2-mercaptoethanol (Sigma) and 1000 U/mL of recombinant leukemia inhibitory factor (Chemicon, Temecula, CA, USA) containing 5 ng/mL basic fibroblast growth factor (bFGF) (Peprotech, Rocky Hill, NJ, USA). Following culture for 2–3 days, the adherent AM and YS cells growing to near-confluence were applied for iPSC experiments.
In humans, the AM membrane was cut into tiny pieces with dissection scissors. The AM membrane pieces were cultured in DMEM with 10% FBS for 7–10 days. The adherent AM cells growing to near-confluence were applied for iPSC experiments. Primary AM cells were provided from the cell bank of RIKEN Bioresource Center, Japan.
Generation of iPSCs
In mouse, each of pMXs-Oct4, Sox2, Klf4, c-Myc and DsRed (an indicator of retroviral silencing) was transfected into the Plat-E cells using the FuGENE6 Transfection Reagent (Roche Diagnostics, Indianapolis, IN, USA). A 1 : 1 : 1 : 1 : 4 mixture of Oct4, Sox2, Klf4, c-Myc and DsRed retroviruses in supernatants with 4 μg/mL polybrene (Nacalai Tesque, Kyoto, Japan) was added to AM and YS cells at 1.0 × 105 cells per 3 cm well. At day 4 after infection, the cells were reseeded into a 10 cm culture dish on feeder cells with mES medium. Colonies were picked around day 20.
In humans, pMXs-OCT4, SOX2, KLF4 or c-MYC, pCL-GagPol, and pHCMV-VSV-G vectors were transfected into 293FT cells (Invitrogen, Carlsbad, CA, USA) using the TransIT-293 reagent (Mirus). A 1 : 1 : 1 : 1 mixture of OCT4, SOX2, KLF4 and c-MYC viruses in supernatant with 4 μg/mL polybrene were added to AM cells at 1.0 × 105 cells per 3 cm well. The cells were subcultured on feeder cells into a 10 cm dish with the iPSellon medium (Cardio) supplemented with 10 ng/mL bFGF (Wako) (hES medium). Colonies were picked up around day 28.
Human and mouse cells were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) for 10 min at 4 °C. After washing with 0.1% Triton X-100 in PBS (PBST), the cells were prehybridized with blocking buffer for 1–12 h at 4 °C and then incubated with primary antibodies; anti-SSEA4 (1 : 300) (Chemicon), anti-TRA-1–60 (1 : 300) (Chemicon), anti-Oct4 (1 : 50) (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-Nanog (1 : 300) (ReproCELL, Tokyo, Japan), anti-Sox2 (1 : 300) (Abcam, Cambridge, UK) and/or anti-SSEA1 (1 : 1000) (DSHB) antibodies for 6–12 h at 4 °C. They were incubated with secondary antibodies; anti-rabbit IgG, anti-mouse IgG or anti-mouse IgM conjugated with Alexa 488 or 546 (1 : 500) (Molecular Probes, Eugene, OR, USA) in blocking buffer for 1 h at room temperature. The cells were counterstained with 4,6-diamidino-2-phenylindole (DAPI) and then mounted with a SlowFade light antifade kit (Molecular Probes). To examine germline competence, cryosections of a half of a testis of 4- to 5-week-old chimeric mice were fixed with 4% paraformaldehyde in PBS for overnight at 4 °C, and then prehybridized with blocking buffer. The sections were double-stained with primary antibodies; anti-LacZ antibody (1 : 500) (Promega, Madison, WI, USA) specific to miPSC-derived cells and with anti-TRA98 antibody (1 : 500) specific to spermatogonia and spermatocytes. The remaining testis and ovaries were stained with X-gal.
Total RNAs were isolated from mouse and human cells using the TRIzol (Invitrogen) and the RNeasy Plus Mini Kit (Qiagen, Valencia, CA, USA), respectively. cDNAs were synthesized from 1 μg total RNAs using Superscript III reverse transcriptase (Invitrogen) with random hexamers according to the manufacturer’s instructions. Template cDNA was PCR-amplified with gene-specific primer sets (Table 1).
Table 1. Primers for RT-PCR and PCR
Gene expression microarray
Total RNA was extracted from mouse cells using the TRIzol Reagent. Double-stranded cDNA synthesized from the total RNA was amplified and labeled using the One-Cycle Target Labeling and Control Regents (Affymetrix, Santa Clara, CA, USA). Global gene expression was examined with the GeneChip Mouse Genome 430 2.0 Array (Affymetrix). The fluorescence intensity of each probe was quantified by using the GeneChip Analysis Suite 5.0 computer program (Affymetrix). The level of gene expression was determined as the average difference (AD). Specific AD levels were then calculated as percentages of the mean AD level of probe sets for housekeeping genes Actin and Gapdh. To eliminate changes within the range of background noise and to select the most differentially expressed genes, data were used only if the raw data values were less than 50 AD. Further data were analyzed with GeneSpring GX 7.3.1 (Agilent Technologies, Santa Clara, CA, USA).
The reprogramming efficiency of mouse YS and AM cells was estimated by counting the number of ALP-positive colonies 21 days after retroviral infection. The cells in 10 cm culture dish were fixed with 4% paraformaldehyde in PBS for 15 min at room temperature and washed with PBS. After treating with ALP stain (pH 9.0) for 30 min at room temperature, the number of ALP-positive cells was counted.
AM-miPSCs (2n = 40, XX) and YS-miPSCs (2n = 40, XY) were microinjected into blastocysts (C57BL/6J × BDF1). The blastocysts were transferred into the uterus of pseudopregnant ICR female mice. Chimeric mice were mated with C57BL/6J for examining germline transmission. The genotype of the progeny was determined with tail tip DNA by genomic PCR with a Neo-specific primer set (Table 1). All animal experiments were performed according to the guidelines of animal experiments of Kyoto University, Japan.
In mice, cell suspension of 1.0 × 106 AM or YS-miPSCs/100 μL DMEM/F12 was subcutaneously injected into the inguinal region of immunodeficient SCID mice (CLEA). In humans, the 1 : 1 mixture of the AM-hiPSC suspension and Basement Membrane Matrix (BD Biosciences, San Jose, CA, USA) were implanted at 1.0 × 107 cells/site under the kidney capsule of immunodeficient nude mice (CLEA). Teratomas surgically dissected out 5–8 weeks in mice and 6–10 weeks in human after implantation, were fixed with 4% paraformaldehyde in PBS, and embedded in paraffin. Sections at 10 μm in thickness were stained with HE.
Bisulfite-modified DNA sequencing
Genomic DNAs (1 μg) extracted from AM-hiPSCs and hAM cells were bisulfite-treated with EZ DNA methylation-Gold Kit (ZYMO Research, Orange, CA, USA) according to the manufacturer’s instruction. The promoter regions of the human NANOG and OCT4 genes were PCR-amplified with specific primer sets (Table 1). Ten clones of each PCR product were gel-purified, sub-cloned and sequenced with the SP6 universal primer.
We thank Dr Gen Kondoh and Miss Hitomi Watanabe for generating chimeras, and Dr Justin Ainscough for critical comments on the manuscript.