Transformation of somatic cells into stem cell-like cells under a stromal niche

Abstract

To study the genomic plasticity of somatic cells without ectopic genetic manipulation, we cultured mouse fibroblasts with ovarian cells, embryonic fibroblasts of different strains, and parthenogenetic embryonic stem cells (ESCs). Of 41 trials, cell aggregation resembling nascent ESC colony from inner cell mass was detected in 9 cases (22%), and 6 cases (67%) yielded fibroblast-derived colonies with ESC morphology. Cells used in coculture provided the critical (P=0.0061) inducing factor for the aggregation. These colony-forming fibroblasts (CFFs) showed similar characteristics to those in ESCs and induced pluripotent stem cells (iPSCs), including pluripotency gene expression, in vitro differentiation, and teratoma formation. Furthermore, CFFs produced somatic chimera, although none showed germline chimerism. CFFs had a tetraploid-like karyotype, and their imprinting patterns differed from parthenogenetic ESCs, thereby confirming their nongermline transmissibility. We observed dysregulation of cell cycle-related proteins, as well as both homologous and heterologous recombination of genomic single-nucleotide polymorphisms in CFFs. Our observations provide information on somatic cell plasticity, resulting in stemness or tumorigenesis, regardless of colony-forming cell progenitors in the fibroblast population. The plasticity of somatic genomes under environmental influences, as well as acquisition of pluripotency by cell fusion, is also implicated.—Lee, S. T., Gong, S. P., Yum, K. E., Lee, E. J., Lee, C. H., Choi, J. H., Kim, D. Y., Han, H., Kim, K.-S., Hysolli, E., Ahn, J. Y., Park, I.-H., Han, J. Y., Jeong, J.-W., Lim, J. M. Transformation of somatic cells into stem cell-like cells under a stromal niche.

Success in generating induced pluripotent stem cells (iPSCs) from various tissues (1) has led to excitement among basic and clinical scientists regarding the potential to study cancer pathogenesis as well as develop cell therapies. Somatic cell transformation into iPSCs is similar to cell tumorigenesis, as is the acquisition of cellular pluripotency (2, 3). Recent advances in direct (4–6), small chemical-induced (7–10), or niche-induced (11, 12) cell reprogramming suggest that specific environments may control cell transformation. However, consistency in observing such a phenomenon, which leads to the establishment of standardized protocol, is difficult, because the many factors presumably responsible for the transformation may miscellaneously combine with each other.

We previously observed spontaneous cell transformation of fibroblasts into colony-like cell aggregations in chickens (13). In addition, we reported that the culture of ovarian stromal cells with fibroblasts induces stem cell-like cell transformation (14). Mesoderm-derived cells, such as fibroblasts and stromal cells, may be involved in cell transformation themselves or may have the capacity to stimulate various transforming processes, including tumorigenesis (15–17), epithelial-to-mesenchymal transition (18), and even genetic aberrations (19). Such results were observed in a variety of extracellular niches, where artificial manipulation of gene function was not involved.

A previous report (12) confirmed that a specific microenvironment triggers single-directional differentiation of stem cells. We subsequently addressed in this study whether a specific niche can trigger differentiation of terminally differentiated somatic cells. Our hypothesis is that fibroblasts can transform into stem cell-like cells under specific environmental conditions without artificial manipulation of gene function. We selected F₁ mice of different origins as the model animal of cell transformation, which made it possible to consider several factors simultaneously.

MATERIALS AND METHODS

Experimental design

These experiments consisted of two parts: first, induction of cell transformation in designed microenvironments, and second, cellular and cytogenetic characterization of the transformed cells. For control cell lines, referenced embryonic stem cell (ESC) and iPSC lines were employed. To evaluate niche effects on cell transformation, we used ovarian cells (predominantly consisting of ovarian stromal cells) or parthenogenetic ESCs (pESCs) and fibroblasts of different strains based on their potential to support cell transformation (14, 20–22). Mitomycin C (MC) treatment was simultaneously used to support cell transformation because it can induce cell stress without decreasing cell viability. For the characterization of transformed cells, we primarily applied conventional methods, including morphological evaluation using microscopy, identification of stemness-specific genes and protein expression, and evaluation of in vivo and in vitro differentiation potential. Subsequently, methylation status of imprinting genes was identified, which provided detailed genetic and cellular characteristics, as well as the origin of the transformed cells. miRNA expression and cell properties of ESCs, embryonic germ cells (EGCs), mouse fetal fibroblasts (MFFs), and colony-forming fibroblasts (CFFs) were determined; and cytogenetic analyses, including karyotyping with G-banding, comparative genome hybridization (CGH) array, and selective genomic single-nucleotide polymorphism (SNP) assays, were also conducted.

Animals

B6D2F1 (C57BL/6×DBA2), B6CBAF1 (C57BL/6×CBA/ca), or outbred ICR mice were employed for cell donors. All animal handling and experimentation procedures followed the standard operation protocols of Seoul National University, under the approval of the review board of the Institutional Animal Care and Committee of Seoul National University (approval no. SNU-050331-2).

Fibroblast preparation

For isolation of the MFFs, 13.5-d-old fetuses were retrieved from pregnant female mice, and the visceral organs, head, and extremities were removed. The remaining tissues were incubated for 6 min with agitation in 0.04% (v/v) trypsin-EDTA (Gibco Invitrogen, Grand Island, NY, USA) and subsequently centrifuged once at 110 g. The supernatants were diluted in Dulbecco's modified Eagle's medium (DMEM; Gibco Invitrogen) containing 10% (v/v) fetal bovine serum (FBS; HyClone Laboratories, Logan, UT, USA). After centrifugation for 4 min at 390 g, the pellets containing the fibroblasts were suspended in new medium and cultured on tissue culture dishes. When the fibroblasts formed a confluent monolayer, they were frozen in 10% (v/v) dimethyl sulfoxide (Sigma-Aldrich, St. Louis, MO, USA) until use. To isolate neonatal and adult skin fibroblasts, 2- or 3-d-old neonates and 8-wk-old adult mice were euthanized, and the skin was removed. The dermis was separated from the epidermis by treatment with 0.1% dispase in DMEM overnight at 4°C and was subsequently digested by mincing into small pieces, followed by digestion for 1 h at 37°C with 0.75% collagenase in DMEM. The isolated fibroblasts were cultured in DMEM containing 10% (v/v) FBS and 1% (v/v) lyophilized mixture of penicillin and streptomycin.

Ovarian cell preparation

Ovarian cells prepared for fibroblast coculture were retrieved from the ovaries of B6CBAF1 female adult mice by our standard protocol (14). After the removal of adherent tissue by chopping of collected tissue with a surgical blade, the specimens were incubated for 30 min at 37°C in dissociation medium that consisted of a 50:50 (v:v) mixture of 0.25% (v/w) trypsin-EDTA (Gibco Invitrogen) and DMEM supplemented with 750 U/ml collagenase type I (Sigma-Aldrich) and 0.03% (v/v) FBS. The dissociated cells were filtered through a 40-μm cell strainer (BD Falcon, Franklin Lakes, NJ, USA) and centrifuged at 390 g for 4 min. The prepared cells were precultured on 60- × 15-mm culture dishes. Fibroblasts that attached quickly to the bottom of the dishes were discarded by collecting only buoyant cells 30 min after seeding. The collected buoyant cells were subsequently used for coculture.

Coculture of fibroblasts and ovarian cells

Three types of fibroblasts (MFFs, neonatal skin fibroblasts, and adult skin fibroblasts) were treated for 3 h with 0 or 10 μg/ml MC (Sigma-Aldrich) in 0.1% (v/v) gelatin-coated 60-mm tissue culture dishes. Cells were subsequently cocultured with prepared cells, including ovarian cells or mixed populations of MFFs and pESCs. The mixed population of MFFs and pESCs was treated with 5 or 10 μg/ml MC at 37°C under 5% CO₂ in a humidified air atmosphere. The culture medium was DMEM supplemented with 0.1 mM β-mercaptoethanol, 1% (v/v) nonessential amino acids (Gibco Invitrogen), 2 mM l-glutamine (Sigma-Aldrich), 1% (v/v) lyophilized mixture of penicillin and streptomycin (Gibco Invitrogen), 5000 U/ml mouse leukemia inhibitory factor (LIF; Chemicon, Temecula, CA, USA), and 15% (v/v) FBS. At the end of primary culture, cultured cells were replated in the same medium except for the LIF concentration, which was reduced from 5000 U/ml in primary culture to 1000 U/ml for the subcultures. Colony-forming cells were mechanically removed using a capillary pipette for subpassaging. The cells were subpassaged at intervals of 3 d, whereas the medium was changed daily.

Establishment of iPSCs

The isolated fibroblasts were cleaned with Ca²⁺- and Mg²⁺-free Dulbecco's phosphate-buffered saline (DPBS; Gibco Invitrogen) and plated on 35-mm culture dishes containing culture medium. The culture medium was DMEM supplemented with heat-inactivated 10% (v/v) FBS. On d 7 of primary culture, the cultured fibroblasts were cryopreserved until use. The procedure to generate iPSCs from tail-tip fibroblasts followed our standard protocol, based on retroviruses expressing 4 reprogramming factors (OCT4, SOX2, KLF4, and MYC; refs. 23, 24). The iPSCs established at the Yale Stem Cell Center were isolated, cultured, and frozen at Seoul National University.

Characterization of CFFs

For characterization using stem cell-specific markers, CFFs and iPSCs were collected at the 20th subpassage, fixed for 10 min at room temperature in 4% (v/v) formaldehyde (Sigma-Aldrich), and immunostained with stem cell-specific antibodies: stemness-specific stage-specific embryonic antigen (SSEA-1; Santa Cruz Biotechnology, Santa Cruz, CA, USA), SSEA-3 (Santa Cruz Biotechnology), SSEA-4 (Santa Cruz Biotechnology), Oct-4 (BD Biosciences, San Jose, CA, USA), integrin-α6 (Santa Cruz Biotechnology), and integrin-β1 (Santa Cruz Biotechnology). Antibody staining was visualized using Alexa Fluor 488-conjugated anti-mouse antibody (Molecular Probes, Eugene, OR, USA), and the DakoCytomation kit (DakoCytomation, Carpinteria, CA, USA). The reactivity of the CFFs and iPSCs to alkaline phosphatase (AP) was assessed using Fast Red TR/naphthol AS-MX phosphate (Sigma-Aldrich).

To confirm spontaneous differentiation in vitro, CFFs and iPSCs were treated with 0.04% (v/v) trypsin-EDTA (Gibco Invitrogen), and the dissociated cells were subsequently transferred to 100-mm plastic Petri dishes that contained LIF-free DMEM (Gibco Invitrogen) supplemented with 15% (v/v) FBS. The cells were grown until embryoid bodies (EBs) formed. The EBs were seeded separately into 96-well culture plates and cultured for 7 d. EBs were immunostained with the following specific markers for the 3 germ layers: nestin (Santa Cruz Biotechnology) and S-100 (Biodesign International, Saco, ME, USA) for ectodermal cells, smooth muscle actin (SMA; Biodesign International) and desmin (Santa Cruz Biotechnology) for mesodermal cells, and α-fetoprotein (Biodesign International) and troma-1 (Developmental Studies Hybridoma Bank, Iowa City, IA, USA) for endodermal cells. Antibody visualization was performed using the DakoCytomation kit.

To confirm in vivo differentiation, 1 × 10⁷ CFFs and iPSCs retrieved at the 20th subpassage were subcutaneously injected into adult NOD-SCID mice. Teratomas that formed in the subcutaneous region were collected 6 wk post-transplantation and fixed with 4% (v/v) paraformaldehyde (Sigma-Aldrich). After embedding in paraffin blocks, the tissues were stained with hematoxylin and eosin for examination using phase-contrast microscopy (BX51TF; Olympus, Kogaku, Japan).

To confirm ultrastructural morphology, ESCs, MFFs, CFFs, and iPSCs were fixed overnight at 4°C in Karnovsky's fixative, consisting of 2% (v/v) paraformaldehyde and 2% (v/v) glutaraldehyde in 0.05 M sodium cacodylate buffer (pH 7.2). Fixed cells were washed 3 times for 10 min each in the cacodylate buffer and were postfixed for 2 h in 1% (w/v) OsO₄. Cells were rinsed twice in distilled water at room temperature and stained overnight at 4°C with 0.5% (w/v) uranyl acetate in distilled water. After en bloc staining, they were dehydrated by placing sequentially in 30, 50, 70, 80, 90, and 100% (v/v) ethanol before embedding in Epon. After polymerization for 24 h at 70°C, trimmed blocks were cut into semithin and ultrathin sections using an ultramicrotome (MT-X; RMC, Tucson, AZ, USA). Ultrathin sections, in which the cells were embedded, were attached to a copper grid and contrasted with 2% uranyl acetate and Reynolds' lead citrate. Cell ultrastructure was examined using a transmission electron microscope.

Reverse transcription-polymerase chain reaction (RT-PCR) and quantitative RT-PCR analyses

Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Valencia, CA, USA). cDNA was synthesized from ∼1 μg RNA using the SuperScript III First-Strand Synthesis System for RT-PCR (Invitrogen, Carlsbad, CA, USA) and subjected to PCR amplification with specific primers. The PCR products were size fractionated by 1.2% agarose gel electrophoresis and visualized by ethidium bromide staining. Primary3 software (Whitehead Institute/Massachusetts Institute of Technology Center for Genome Research, Cambridge, MA, USA) was used to design the specific primers used in these experiments. All PCR primers were designed based on mouse cDNA and genomic DNA sequences obtained from GenBank (U.S. National Center for Biotechnology Information, Bethesda, MD, USA). The specificities of the designed primers were tested by conducting 40 PCR cycles of 95°C for 30 s, an annealing temperature for 45 s, and 72°C for 30 s. The primer sequences are listed in Supplemental Table S3. To compare relative mRNA level of samples, the cDNA was quantified by real-time PCR using the DyNAmo SYBRGreen qPCR Kit (Finnzymes, Espoo, Finland). The mRNA level of each gene was normalized to that of β-actin and defined as 2^−ΔCt, where C_t is the threshold cycle for target amplification, and ΔC_t = C_{t tg} − C_{t ir}, where tg is the target gene, and ir is the internal reference (β-actin).

Sex determination using genomic DNA-PCR analyses

Total genomic DNA from each sample was extracted using the G-spin Genomic DNA Extraction Kit (iNtRON Biotechnology, Seoul, Korea) according to the manufacturer's instructions. The extracted genomic DNA was PCR amplified using primers for the Zfy1 (Y chromosome-specific) and Xist (X chromosome-specific) genes. The PCR products were size fractionated by 1.2% (v/w) agarose gel electrophoresis and visualized by ethidium bromide staining.

Telomerase activity assays

Telomerase activity was determined using the TRAP_EZE Telomerase Detection Kit (Chemicon) with slight modifications for easy reading of data. The established lines were analyzed at the 20th subpassage and, PCR amplification was performed for 27 cycles. The protein concentration was adjusted (50 pg in this set of experiments) before assay, and the PCR product was separated by nondenaturing polyacrylamide gel electrophoresis.

DNA microsatellite analyses

DNA microsatellite analyses were performed with genomic DNA samples from fibroblast donors (B6D2F1 MFFs), ovary donors (B6CBAF1 ovaries), and newly established CFFs. Five specific mouse microsatellite primers (D3Mit200, D11Mit4, D15Mit159, D19Mit33, and D4Mit251) were used in this study, which can be referenced online (http://www.cidr.jhmi.edu/mouse/mmset.html). The genomic DNA from each sample was PCR amplified for the 5 microsatellite loci. Forward primers were synthesized with a fluorescent tag (FAM, TET, or HEX) at the 5′ end, and fluorescent PCR amplification was performed using the PC808 program. The PCR products were subsequently analyzed in an ABI Prism 310 DNA automated sequencer (Applied Biosystems, Foster City, CA, USA). Digital images were obtained using Genescan Data Collection 2.5 software (Applied Biosystems). Each fluorescent peak was quantified for base pair size, peak height, and peak area.

Chimera production

To confirm pluripotency on a larger scale, 10–15 CFFs maintained for different numbers of subpassages (7–13 or 32–39 subpassages) were aggregated with diploid 8-cell embryos (25), injected into diploid blastocysts, or injected into tetraploid blastocysts that had developed after electrofusion at the 2-cell stage. The blastocysts reconstituted by each method were transferred to the uteri of 2.5-dpc pseudopregnant ICR females.

Hormone assays

To confirm the ability to derive oocyte-like cells, CFFs, iPSCs, and control ESCs were seeded at a density of ∼1–2.5 × 10⁴ cells/cm² and grown in 0.1% gelatinized tissue culture plates in ESC medium without feeder cells and LIF (26). Cultured cells were maintained without a medium change for the first 3 d, and the medium was changed daily until d 9 of culture after nonadherent cells were removed on d 3–4 of culture. The culture medium was then changed every other day, and the formation of cell aggregates and follicle-like structures was monitored. The estradiol and progesterone levels in culture medium were measured by enzyme immunoassay using the Estradiol EIA Kit and Progesterone EIA Kit (Oxford Biomedical Research, Oxford, MI, USA), respectively. The measurement was recorded every 2 d.

Pyrosequencing for imprinted gene analyses

Genomic DNA was extracted from cultured cells using the DNeasy Blood and Tissue Kit (Qiagen). The target point primers and sequencing primers were designed using the sequence of each differentially methylated region (DMR) using Methyl Primer software and ADS software. The methylation patterns of DMRs in NANOG, Oct-4, H19, Gtl2, Snrpn, and Peg3 genes were analyzed using specific primer sets (Supplemental Table S3). DNA methylation analyses were performed with bisulfite-treated DNA using the MethylEasy Xceed-Rapid DNA bisulfite modification kit (Human Genetic Signatures Pty Ltd., Sydney, NSW, Australia). DNA was amplified using bisulfite-PCR with a biotin-labeled primer that was used to purify the final PCR product with streptavidin beads (Streptavidin Sepharose HP; Amersham Biosciences AB, Uppsala, Sweden) and the pyrosequencing vacuum prep workstation (Pyrosequencing, Inc., Westborough, MA, USA). Pyrosequencing was performed using the PyroMark Q96 MD Pyrosequencing System (Pyrosequencing, Inc.). The resulting pyrograms and associated sequences were generated and analyzed automatically using PSQ 96 SQA software (Qiagen).

SNP analyses of cell lines

SNP genotyping to distinguish genetic polymorphism between C57BL/6 and DBA2 was performed using iPLEXTM for use with the MassARRAY platform (Sequenom, San Diego, CA, USA; ref. 27). Genomic DNA was extracted from each sample using the AccuPrep Genomic DNA extraction kit (Bioneer, Seoul, South Korea). Primers specific to each SNP and assays were designed using the MassARRAY assay design software (Sequenom).

Immunoblot analyses for cell cycle proteins

The expression of cell cycle proteins in CFFs and the control cell lines was monitored using the Bradford method (28). Cells were harvested, washed twice with PBS, and lysed for 30 min on ice in lysis buffer [20 mM Tris (pH 7.5), 1 mM EDTA, 1 mM EGTA, 1% (v/v) Triton X-100, 1 mg/ml aprotinin, 1 mM phenylmethylsulfonylfluoride, and 0.5 mM sodium orthovanadate]. The lysates were clarified by centrifugation (10 min at 15,000 rpm, 4°C) and the protein concentration was determined. Equal amounts of protein (20 μg) were resolved by 10% (w/v) SDS-PAGE and transferred to nitrocellulose. After the blots were washed with TBST [10 mM Tris-HCl (pH 7.6), 150 mM NaCl, and 0.05% (v/v) Tween 20], membranes were blocked for 1 h with 5% (w/v) skim milk and incubated with the appropriate cyclin D1, cyclin E, CDK2, CDK4, p53, and pRb antibodies obtained from Santa Cruz Biotechnology, Inc. at dilutions recommended by the supplier. The membranes were washed, and primary antibodies were detected using goat anti-rabbit IgG conjugated to horseradish peroxidase (Jackson Immunoresearch, West Grove, PA, USA). The bands were visualized using enhanced chemiluminescence (Amersham Pharmacia Biotech, Little Chalfont, UK).

Fluorescence-activated cell sorter (FACS) analyses

FACS analyses were used to measure DNA content. Harvested cells were washed in DPBS (Gibco Invitrogen) and suspended in 70% (v/v) ethanol (Sigma-Aldrich) for 1 h at 4°C. The cells were centrifuged for 4 min at 390 g and resuspended in 0.5 ml of Ca²⁺- and Mg²⁺-free DPBS (Gibco Invitrogen) containing 0.1 mg/ml ribonuclease (Sigma-Aldrich) and 0.1 mg/ml propidium iodide (Sigma-Aldrich). After incubation for 30 min at room temperature in the dark, the cell suspension was analyzed using a Becton Dickinson FACS-Vantage SE (Becton Dickinson, San Jose, CA, USA) equipped with a water-cooled laser. The data were analyzed using Cell Quest 3.3 software (Becton Dickinson). For identification of stem cell-specific marker expression in MFFs, the prepared B6D2F1 MFFs were reacted with primary antibodies, biotin-conjugated CD34 (BD Biosciences), PE-conjugated CD44 (BD Biosciences), PerCP-Cy5-conjugated Oct-4 (BD Biosciences), and PE-conjugated Nanog (BD Biosciences) for 1 h at room temperature. CD34 was additionally reacted with streptavidin-FITC (BD Biosciences). After a subsequent washing step, the antigen-antibody complexes were analyzed by flow cytometry (FACSCalibur; Becton Dickinson).

Statistical analyses

Statistical analyses were conducted to compare numerical data obtained from >3 replications in each experiment. A generalized linear model (PROC-GLM) from a statistical analysis system (SAS) program (SAS Institute, Cary, NC, USA) was used to evaluate the effect of each treatment. When analysis of variance (ANOVA) in the SAS package detected a significant main effect, the least-square method was conducted. A value of P < 0.05 indicates statistical significance.

RESULTS

Cell transformation in a specific extracellular environment

Forty-one attempts were made to trigger fibroblast transformation (Supplemental Table S1). Of these, cell aggregates during the primary passage were observed in 9 cases (22%). Cell aggregation in 6 of 9 cases transformed into ESC-like colonies (67%). All established colonies were maintained >20 subpassages. All data from each replicate were reallotted to consider the effects of the experimental system factors on the cell transformation. As shown in Table 1, a significant model effect on cell aggregation was detected in cocultured cell types (P=0.0061). The origin, strain, MC treatment, gender of fibroblasts, and strain of cocultured cells did not influence the aggregation. Short tandem-repeat microsatellite analyses (Table 2) showed that the microsatellite peaks of the CFFs accurately matched those of B6D2F1 fibroblasts for all 5 markers tested. Contribution of ovary-derived cells to form ESC-like colonies was tested, but the culture of ovarian cells without MFFs yielded neither aggregated cells nor cell colonies from the observation of total 20 trials (Supplemental Table S2). Oocytes were not visible in the ovarian cell pellets prepared by our standard preparation method. Culture of primordial germ cells with MFFs did not yield colonized cells.

Table 1.

Factors affecting transformation of fibroblasts into colony-forming cells in different culture conditions and treatments

Factors	Groups	Trials [n (%)]			Colonies [n (%)]c
		Attempted	Yielded cell aggregatesa	Yielded coloniesb	Subpassaged >20	Derived from fibroblasts
Whole replicates	Merged	41	9 (22)	6 (67)	6 (100)	6 (100)
Fibroblast origin	13.5-dpc fetuses	32	9 (28)	6 (67)	6 (100)	6 (100)
	Neonates	6	0 (0)	0 (0)	0 (0)	0 (0)
	Adults	3	0 (0)	0 (0)	0 (0)	0 (0)
	P		0.2089
Fibroblast strain	B6D2F1	39	9 (23)	6 (67)	6 (100)	6 (100)
	ICR	2	0 (0)	0 (0)	0 (0)	0 (0)
	P		0.4545
Fibroblast treatment	None	2	0 (0)	0 (0)	0 (0)	0 (0)
	Mitomycin	39	9 (23)	6 (67)	6 (100)	6 (100)
	P		0.4545
Fibroblast gender	Male	6	0 (0)	0 (0)	0 (0)	0 (0)
	Female	3	0 (0)	0 (0)	0 (0)	0 (0)
	Mixed	32	9 (28)	6 (67)	6 (100)	6 (100)
	P		0.2089
Cocultured cell typed	Ovarian cells	37	6 (16)	6 (100)	6 (100)	6 (100)
	Mixed populatione	4	3 (75)	0 (0)	0 (0)	0 (0)
	P		0.0061
Cocultured cell straind	B6D2F1	6	3 (50)	0 (0)	0 (0)	0 (0)
	B6CBAF1	33	6 (18)	6 (100)	6 (100)	6 (100)
	C57BL6	2	0 (0)	0 (0)	0 (0)	0 (0)
	P		0.1752

Data were extracted from Supplemental Table S1. All factors were extracted from the whole replications. P values indicate model effect of treatment.

^aPercentage of attempted trials.

^bPercentage of trials that yielded cell aggregates.

^cPercentage of trials that yielded colony formation.

^dCocultured cells with fibroblasts of different origins.

^eMixed population of 13.5-dpc fetal fibroblasts and parthenogenetic embryonic stem cells.

Table 2.

Confirmation of colony-forming cell origins by microsatellite analysis

Marker and size	CFF-1	CFF-2	CFF-3	CFF-4	CFF-5	CFF-6	Ovary cell	Fibroblast
D3Mit200
1	101.2	101.1	101.1	101.0	101.2	101.2	124.4	101.1
2	124.3	124.3	124.3	124.2	124.4	124.3	—	124.3
D11Mit4
1	249.0	248.9	248.9	248.7	248.9	248.7	248.7	248.8
2	285.1	285.1	285.3	285.1	285.1	285.1	295.1	285.1
D15Mit159
1	112.0	110.8	110.8	111.9	111.8	111.8	137.4	111.8
2	137.5	137.4	139.5	137.5	137.4	137.4	139.4	137.4
D19Mit33
1	251.6	251.6	251.5	251.5	251.7	251.6	251.6	251.6
2	253.6	253.6	253.6	253.6	253.5	253.6	253.6	253.7
D4Mit251
1	107.2	107.2	107.2	107.2	107.2	107.3	107.3	107.3
2	120.6	120.6	120.6	120.7	120.6	120.8	116.3	120.7
F1 origin	B6D2	B6D2	B6D2	B6D2	B6D2	B6D2	B6CBA	B6D2

Microsatellite markers used were selected from MIT database for discerning the strains of ovary cells (B6CBAF1) and fibroblasts (B6D2F1).

Comparison of CFFs with ESCs and iPSCs

To presume the progenitor colony-forming cells, stem cell-like activity of MFFs or other stem cell-like cells in MFF population were monitored. Intact MFFs were negative for female germ cell-specific Stella, germ cell-specific Vasa, and follicular cell-specific AMH (Fig. 1A). FACS-based cell sorting did not detect hematopoietic stem cell-specific CD34⁺ and pluripotency-specific Oct-4⁺ cells and Nanog⁺ cells in MFF population. However, the presence of mesenchymal stem cell-specific CD44⁺ cells was detected. In addition, the intact MFF population expressed skin stem cell-specific Snail, Slug, Pax3, Dermo-1, and Sox9, while CFFs expressed only Slug and Dermo-1 (Supplemental Fig. S1). Expression of skin stem cell-specific Twist was not detected in either MFFs or CFFs.

Figure 1.

Morphology of fibroblasts, CFFs, and iPSCs. A) Identification of marker expression in B6D2F1 MFFs. mRNA expression of female germ cell-specific Stella, germ cell-specific Vasa, and follicular cell-specific AMH was not detected in MFFs, while adult testicular cells expressed Vasa. Cells positive for hematopoietic stem cell-specific CD34 or pluripotency-specific Oct-4 and Nanog were not detected, while mesenchymal stem cell-specific CD44⁺ cells were identified by flow cytometry. B) Morphology of CFFs and iPSCs. CFFs have a similar morphology with referenced ESCs, and both CFFs and iPSCs were stably maintained during long-term culture. Scale bars = 50 μm. C) Ultrastructure of CFFs, iPSCs, referenced ESCs, and MFFs. In the ESCs, a high nuclear-to-cytoplasmic volume ratio was detected, and multiple nucleoli with a well-developed nuclear membrane were visible. In addition, the microvilli of the cytoplasmic membrane and the cytoplasmic microorganelles were underdeveloped in ESCs. Similar morphology compared with E14 ESCs was detected in both CFFs and iPSCs. In contrast, low nuclear-to-cytoplasmic volume ratio and a well-developed cytoplasm containing a large number of microorganelles were detected in MFFs. Scale bars = 2.5 μm (left panels); 1 μm (right panels).

On phenotype analyses, the morphology of CFFs was similar to that of ESCs or iPSCs. CFFs were stably maintained during long-term culture, similar to other cell types (Fig. 1B). Transmission electron micrographs showed a clear difference in fibroblast morphology before and after colony formation (Fig. 1C). The ultrastructure of CFFs showed similar cellular morphology compared to ESCs and iPSCs, which had a high nuclear-to-cytoplasmic volume ratio. The microvilli of the cytoplasmic membrane and the cytoplasmic microorganelles were undeveloped. In contrast, in the CFFs, a relatively low nuclear-to-cytoplasmic volume ratio and a well-developed cytoplasm containing a large number of microorganelles were detected.

The CFFs were characterized by comparing the control cell lines, and as shown in Fig. 2A, CFF, ESC, and iPSC lines were all positive for SSEA-1, Oct-4, integrin-α6, integrin-β1, and AP, and negative for SSEA-3 and SSEA-4, while difference in expression strength was detected. Similar mRNA expression levels in NANOG, Rex-1, Dnmt3b, Tert, Lif Rc, Fgf4, Foxd3, CD9, and Gdf3 were detected among CFFs, E14 ESCs, and iPSCs, whereas the expression of Oct-4, Cripto, Stat3, Bmp4, and Sox2 was significantly different (Fig. 2B). In addition, all 6 CFF lines showed strong telomerase activity, which was not detected in MFFs (Fig. 2C). All CFF lines established turned out to have XX chromosomes, because a male-specific Zfy-1 gene was not detected in any of them (Fig. 2D). miRNA microarray analysis showed that expression of stemness-specific miRNAs was higher in the CFF and ESC lines compared to MFFs and ovary cells (Supplemental Fig. S2A, B). Real-time PCR analyses showed that the expression of stemness-specific miR-290, miR-293, and miR-295 was higher in CFF-1 and CFF-3 cells than in ESCs (Supplemental Fig. S2C).

Figure 2.

Characterization of CFFs. A, B) Stemness-specific marker reactivity (A) and relative mRNA expression (B) of CFFs, reference ESCs, and iPSCs. Similar to the ESCs, CFF and iPSC lines were positive for SSEA-1, Oct-4, integrin-α6, integrin-β1, and AP, but negative for SSEA-3 and SSEA-4. Scale bars = 50 μm. Expression of ESC-specific genes was detected in CFFs. C, CFF; i, iPSC; E, E14 ESC. Bars with different letters differ significantly (P<0.05). C) Telomerase activity of CFFs was detected by the telomeric repeat amplification protocol assay. The ladder of telomerase products amplified by PCR is shown, with 6-base increments starting at 50 nt. All 6 lines, like ESCs, exhibited high levels of telomerase activity. NC, negative control without template addition. D) PCR analysis was conducted to determine the sex of cells using primers for X (Xist) and Y (Zfy1) chromosome-specific genes. Late-passage CFFs were used for PCR analysis, and all 6 lines expressed Xist but not Zfy1.

Culturing CFFs in LIF-free medium allowed spontaneous in vitro differentiation into EBs as early as 4 d after culture (Fig. 3A). The EBs were positive for the germ layer-specific markers S-100, nestin, SMA, desmin, α-fetoprotein, and troma-1. Transplantation of the CFFs of all 6 lines into NOD-SCID mice induced the teratoma formation (Fig. 3B for CFF-1 line, not shown for other lines) consisting of cells derived from 3 germ layers, including neuroepithelial rosettes (ectodermal), keratinized stratified squamous epithelial cells (ectodermal), osteoid islands showing bony differentiation (mesodermal), muscle (mesodermal), pancreatic tissue (endodermal), and ciliated columnar epithelial cells (endodermal). The same in vitro and in vivo differentiation patterns were demonstrated in iPSCs and E14 ESCs (data not shown).

Figure 3.

Differentiation of CFFs and iPSCs. A) In vitro differentiation into EBs of CFFs and iPSCs. a1) EBs formed at d 4 after culture in leukemia inhibitory factor (LIF)-free culture medium. a2–a7) Differentiation was detected by immunocytochemistry using the 3 germ-layer-specific markers: S-100 (a2) and nestin (a3) for ectoderm, SMA (a4) and desmin (a5) for mesoderm, α-fetoprotein (a6), and troma-1 (a7) for endoderm. a8) EBs of iPSCs formed at d 4 after culture in LIF-free culture medium. a9–a14). Differentiation was detected by immunocytochemistry using the 3 germ-layer-specific markers: S-100 (a9) and Nestin (a10) for ectoderm, SMA (a11) and Desmin (a12) for mesoderm, α-fetoprotein (a13) and Troma-1 (a14) for endoderm. Scale bars = 100 μm. B) Teratoma formation after subcutaneous injection of CFFs and iPSCs into NOD-SCID mice. b1–b6) The cell mass formed by the CFF injection contained 3 germ-layer-specific cell types: neuroepithelial rosettes (b1) and keratinized stratified squamous epithelial cells (b2) for ectoderm; osteoid islands showing bony differentiation (b3) and muscle (b4) for mesoderm; pancreatic tissue (b5) and ciliated columnar epithelial cells (b6) for endoderm. b7–b12) The cell mass formed by the iPSC injection contained 3 germ layer-specific cell types: neuronal epithelial cells (b7) and keratinized stratified squamous epithelial cells (b8) for ectoderm; chondrogenic differentiation (b9) and muscle (b10) for mesoderm; cuboidal to columnar epithelium (b11) and ciliated cuboidal to columnar epithelium (b12). Scale bars = 100 μm (b1–b3, b7–b12); 50 μm (b4–b6). C) Chimerism induced by CFFs. Ten to 15 CFFs were aggregated with 8-cell embryos or injected into diploid and tetraploid blastocysts, and each resulting blastocyst was transferred into the uterine horn of a surrogate mother. See Table 3 for summary of outcomes. Live offspring with different coat colors were determined to be somatic chimeras. Images show the somatic chimeras derived from the transfer of CFF-aggregated blastocysts (c1, c2) or CFF-injected blastocysts (c3, c4). Asterisks and arrows indicate the surrogate mothers and 14- and 10-d-old chimeric progenies, respectively.

Different characterization of CFFs with ESCs and iPSCs

As shown in Fig. 3C and Table 3, in the first trial employing 960 blastocysts that were aggregated with CFFs of different subpassages as 8-cell-stage embryos, 74 offspring were delivered and 9 (19%) were somatic chimeras. Four of the chimeras survived (3 females and 1 male), but progeny testing did not yield a germline transmission. In the second set of experiments using diploid blastocysts, 298 injections of CFF-1 or CFF-4 were made, and 2 somatic chimeras were derived. In the final set using 100 tetraploid embryos, no offspring was derived after the CFF injection, followed by transfer to surrogate mothers. However, placental and fetal formation was detected in 36 and 9 cases, respectively. Higher placental formation (80 vs. 36%) and lower fetal formation (13 vs. 25%) were observed in the development of tetraploid embryos that were transferred without CFFs in 20 cases. No germline transmission was detected in any trial.

Table 3.

Chimerism induced by CFFs

Trial	Embryos transferred (n)	Recipients (n)	Offspring delivered (n)b	Total	Somatic chimeras [n (%)]a
					Live male	Live female	Dead
Diploid aggregation with CFF-1
7–13 subpassages	573	46	48	9 (19)	1	3	5
32–39 subpassages	387	27	26	0 (0)	0	0	0
Diploid injection
CFF-1, passage 11	200	13	127	2 (2)	1	1	0
CFF-4, passage 7	98	5	46	0 (0)	0	0	0

	Embryos transferred (n)	Recipients (n)	Placentas formed [n (%)]d	Fetuses [n (%)]c
				Formed	Normal morphology
Tetraploid injection
CFF-4, passage 7	100	6	36 (36)	9 (25)	1 (3)
Without CFF-4	20	1	16 (80)	2 (13)	0 (0)

Outcome of chimera production after aggregation of CFFs with diploid embryos, injection of CFFs into diploid blastocysts, or injection of CFFs into tetraploid blastocysts (see Fig. 3).

^aPercentage of offspring delivered.

^bIn some cases, fetuses were delivered by cesarean section at 19.5 dpc.

^cPercentage of placentas formed.

^dPercentage of embryos transferred.

As shown in Fig. 4, cytogenetic analyses demonstrated that all CFF lines were aneuploid with a tetraploid-like chromosome number (71 to 78 chromosomes, with an average of 74.3±4.7, Supplemental Fig. S3A). iPSCs and MFFs had normal diploid karyotypes. Deletions and translocations were detected at multiple chromosomal sites in CFFs, and more mutated chromatids were represented in CFFs compared to iPSCs or MFFs (Supplemental Fig. S3A). The deleted chromatids appeared sporadic, and no typical chromosome abnormality patterns were detected by CGH array (Supplemental Fig. S3B). Cell cycle protein analyses in the established CFFs and control ESCs showed less phosphorylation of p53 at serine 15 and 20 sites in the CFFs compared to E14 or R1 ESCs or MFFs (Fig. 4B). Concomitantly, increased cyclin E and cyclin D1 were detected after colony formation (CFFs), at similar levels to the expression level in ESCs (Fig. 4C). Statistical significance showing as a model effect was detected in the comparison of p53 at serine 15 and cyclin D1 translation. The ability of CFFs to differentiate into follicle structures in vitro was confirmed by recapitulating their ability to form oocyte-containing follicle-like structures (Fig. 5). CFF, iPSC, and ESC lines formed follicle-like structures after treatment. Estradiol and progesterone production were different among the cell lines of iPSCs and ESCs, with no significant changes in secretion activity throughout culture in CFFs.

Figure 4.

DNA contents and karyotype analysis of CFFs, iPSCs, and their progenitor MFFs and expression of the cell cycle proteins (CDK2, CDK4, cyclin E, and cyclin D) and p53 in different cell lines. A) DNA contents of each cell line were measured by flow cytometric analysis; karyotype data are shown as GTG-banding karyograms for 6 CFF, iPSC, and MFF lines. Arrows in each image indicate structural changes, including deletional, additional or translocational chromatid configurations on each chromosome. All 6 CFF lines show tetraploid-like chromosome status, while iPSCs and MFFs show normal karyotype. B, C) MFFs of different batches and 2 lines of CFFs were provided for Western blot analysis. Referenced ESCs of E14 and/or R1 strains were used as controls. B) Less phosphorylation of p53 at serine 15 and 20 was detected in CFFs than in the reference E14 or R1 ESC or MFF lines. Bars with different letters differ significantly (P<0.05). C) Similar protein expression was detected between CFF and ESC lines, while weak expression was detected in 2 MFF lines.

Figure 5.

Differences in differentiation activity and stemness-related or imprint gene methylation among stemness-acquired cell lines. Differentiation of CFFs into follicle-like structure and steroid production of CFFs, iPSCs, and referenced ESCs after the formation of follicle-like structure in vitro. All cell lines evaluated formed the follicle-like structure after treatment. Scale bars = 200 μm. Estradiol and progesterone production were different among cell lines. There was no significant change in secretion activity throughout culture in CFFs, in contrast to both iPSCs and ESCs. See Table 4. Error bars = sem. *P < 0.05.

Pyrosequencing analyses to assess the methylation status of paternally imprinted H19 and Gtl2 and maternally imprinted Peg3 and Snrpn showed lower methylation in CFFs compared with iPSCs, ESCs, and pESCs (Table 4). In the case of embryonic germ cells, complete demethylation in H19, Igf2r, Nanog, and Oct4 was detected (Supplemental Fig. S4), which differed from ESCs and MFFs. Only little difference in methylation pattern was detected between CFFs and embryonic germ cells. SNP genotyping results demonstrated that CFFs had both homologous and heterologous recombination of genomic SNPs (Fig. 6), and complete heterozygous SNPs were detected in EGCs, ESCs, and MFFs of the B6D2F1 strain. pESC lines showed both homozygous and heterozygous chromosome recombination. As expected, only homozygotic SNP loci were detected in the fibroblasts of maternal or paternal origin.

Table 4.

Methylation of stemness genes Oct4 and Nanog, maternally imprinted genes Snrpn and Peg3, and paternally imprinted genes Gtl2 and H19 in mouse MFFs, CFFs, iPSCs, pESCs, and ESCs

Gene	Cell line
	MFF	CFF	iPSC	pESC	ESC
Stemness related (%)
Oct4	67	8	6	14	8
	68	3	2	12	3
	44	12	9	13	11
Nanog	47	8	6	11	8
	35	7	7	10	8
Maternally imprinted (%)
Snrpn	30	5	21	23	37
	43	4	33	40	42
	42	8	32	33	38
Peg3	39	19	24	40	36
	50	5	30	52	43
	46	7	28	46	41
Paternally imprinted (%)
Gtl2	48	4	48	18	44
	45	3	56	11	46
	42	4	51	7	42
H19	53	13	58	37	86
	48	8	58	27	81

Genomic DNAs isolated from each cell line were subjected to pyrosequencing analysis after bisulfite treatment. Different methylation status in both imprinted genes was detected among cell lines, while stemness-related genes in all cell lines except for MFF were highly demethylated.

Figure 6.

SNP genotyping analysis of CFFs. Heterozygosity or homozygosity of SNP loci were analyzed, and CFFs of B6D2F1 were compared with those of EGCs, ESCs, MFFs, and pESCs of B6D2F1 and somatic fibroblasts of DBA2 and C57B6. Both homozygosity and heterozygosity were concomitantly detected in the CFF line. EGCs, ESCs, and MFFs of F1 strain showed heterozygosity alone, while only homozygotic SNP loci were detected in the fibroblasts of inbred strain. pESC line possessed both homozygotic and heterozygotic chromosomes.

DISCUSSION

In this study, we confirmed that a specific niche can induce somatic cell transformation into stem cell-like cells, and fibroblasts or the cells mixed in the fibroblast population can participate in niche-induced transformation. Stemness acquisition under the niche designed in this study, however, might not occur completely, and several different cell characteristics were observed in CFFs compared with the control ESC or iPSC lines. Based on our observation that several CFF characteristics were similar to tumor cells, we further hypothesized that niche-induced, incomplete cell transformation can also lead to tumorigenesis. Terminally differentiated, somatic cells may have both cellular and genetic plasticity in response to specific extracellular environments (29, 30).

The morphology of the fibroblasts before and after colony formation was typical of cells undergoing slow (i.e., well-developed cytoplasm and microvillar structure with a condensed nucleolus, frequently observed in cells entering a long G₁ phase) and rapid (i.e., large nucleus to cytoplasm ratio with less-developed cytoplasm) cell cycle progression, respectively. The data from microsatellite analyses and imprinted gene methylation results confirmed the occurrence of cell transformation in fibroblasts, not in contaminating cells of other strain, such as germ cells. As a matter of fact, our standard protocol for collecting ovarian cells extremely minimizes contamination of the cells > 40 μm in diameter, and most growing oocytes, antral follicles, and preantral follicles, at least beyond the late primary stage, can be eliminated from the list of CFF progenitor cells. Cell aggregates and colonies were not derived from the culture of primary follicles, immature oocytes, or granulosa cells (14). Results of immunocytochemistry (data not shown) demonstrated the absence of developing oocytes and PGCs in ovarian cell pellets.

Fibroblasts differentiate from embryonic mesenchymal cells, which are the progenitors of various cells and tissues in this study. Considering other data recently reported on direct cell reprogramming (4–6), fibroblasts seem to be an ideal bioresource for cell transformation. The presence of CD44⁺ cells and the expression of skin stem cell-specific genes in fibroblast population in this study support the feasibility of fibroblast populations as stem cell donors. As a matter of fact, there are many mesenchymal lineage cells being potentially mixed in fibroblast populations, which include skin stem cells (31) and circulatory bone marrow mesenchymal (32) or hematopoietic (33) stem cells. Isolation of CD44⁺ cells and overexpression of specific genes in the isolated cells may be one of alternative choices to establish somatic cell-derived stem cells or to differentiate specific functional cells. Furthermore, a fibroblast-employed, niche model system for cell transformation may be an excellent model for elucidating the pathogenesis of mesenchymal cell-derived tumors, such as sarcoma. Nevertheless, there is not still clear evidence whether fibroblasts themselves or fibroblast-mixed cells are involved in the colony formation.

From our observations, cell-to-cell interactions are the most important factor for fibroblast transformation. The number of replicates is not sufficient for full interpretation of the influence of all factors on cell transformation in this study, but our data clearly demonstrate the importance of the cellular niche for fibroblast transformation. According to niche theory, one cell changes lots of cellular events, and mixing of extremely small numbers may change the fate and function of adjacent cells; thus, care must be taken to analyze cell-to-cell interactions and niche signaling for cell transformation. Development of a defined niche for monitoring cell transformation is absolutely required for characterizing rate-limiting factors of cell transformation. We have developed a cellular niche containing stemness-specific peptide motifs, which can maintain ESCs without coculture cells (34, 35).

The cell transformation observed in this study may occur via changes in the cell cycle and its checkpoints, because the fibroblasts transformed had a concomitant acceleration of the cell cycle and breakdown of cell cycle checkpoints. p53 phosphorylation at serines 15 and 20, which is essential for p53-dependent apoptosis (36, 37), was lower in the CFFs compared to the ESCs or nontransformed fibroblasts. This would allow the aneuploid CFFs to survive. Increased cyclin E and cyclin D1 expression was detected in the CFFs compared to intact fibroblasts, which became similar to the expression levels of the control ESCs. Halting or changing the cell cycle before or during transformation is likely an important step. Change of cell cycle checkpoint, whether related to aneuploidy or not, can result in better cell survival in a suboptimal environment. It may also trigger cell transformation as the normal adaptation process against a specific environment.

Based on novel results of stem cell transformation by exposure to ESC extract (38, 39), secretary substances from ovarian cells, ESCs, oocytes, or even fibroblasts may also induce cell transformation. It is impossible to uptake large molecules such as RNA and cytoplasmic proteins into cytoplasm, which can induce cell transformation, by simple diffusion. Instead, stemness-related signaling or its small triggering molecules can be stimulated by either extracellular protein motifs (34, 35) or the active transport system, respectively. However, we have not observed positive results on fibroblast transformation by using medium conditioning with ovarian cells and ESCs or by coculture with ESCs to date (preliminary data not shown).

Tetraploidy found in CFFs may be one of main reasons for the failure to induce germline transmission. Different levels of stemness-specific gene expression in CFFs compared with those in ESCs or iPSCs may another cause for the failure. In addition, both homozygous and heterozygous recombinations were detected in the evaluated CFF lines. Complete heterozygous SNP was detected in ESCs and fibroblasts of the F₁ strain, and, as expected, only homozygotic SNP loci were detected in the fibroblasts of maternal or paternal origin (Fig. 6). These data suggest that genetic instability is the primary cause of impaired development, whereas it neither inhibits stemness acquisition nor critically determines stem cell characteristics. There are a number of aneuploid or polyploid cells exist even in the normal physiological state (40, 41), some of which have multipotent or pluripotent activity. pESC lines also showed both homozygous and heterozygous chromosome recombination, while their patterns of imprinted gene methylation were different from the CFFs. If de novo cell transformation of fibroblast occurs, unique mitotic recombination may occur without a cell division, which contributes to tetraploidy and the mixed homozygous and heterozygous composition of the chromosome. Is such a mixed recombination state one of the stimulants for acquiring stemness or oncogenesis in terminally differentiated cells?

In addition to de novo transformation of the fibroblast itself, cell fusion would be another cause of cell transformation because of tetraploidy of CFFs. We have not observed spontaneous fusion of fibroblasts, and artificial induction of fibroblast fusion by different methods (chemical, viral, and electrical) does not yield colony-forming cells. Results of microsatellite analysis confirm the fusion of fibroblasts with ovary-derived cells, and no evidence of fibroblast fusion for deriving CFFs except for tetraploidy has been recognized to date. Use of a defined niche can provide critical evidence of cell fusion resulting in cell transformation.

Our reported study is a pilot-style experiment for finding novel phenomena, and first of all, the results showed niche-induced, somatic cell plasticity in ESC-like cells without genetic manipulation. Although these data lack genetic or molecular evidence, we suggest a strategy to understand cell transformation and somatic cell plasticity. It seems that niche-based reprogramming is an incomplete process for acquiring stemness. Evaluations of the epigenetic reprogramming of stemness-related genes (to the permissive chromatin states and promoter modifications), as well as of specific gene expression triggering genetic and epigenetic silencing or activation, are necessary for elucidating the mechanism of the fibroblast transformation. We are now primarily conducting cytogenetic analysis using microarrays and SNP genotyping tools, which will be followed by evaluating histone acetylation and methylation. Pursuing a series of hierarchal studies on somatic cell transformation from bioinformatics to phenotype via epigenetics, signal transduction, and protein engineering directly contributes to clinical-friendly regulation of cell transformation, as well as to confirming the feasibility of direct reprogramming of terminally differentiated cells. From a different viewpoint, lots of information can be obtained from this niche-based cell transformation, and the niche model is useful for understanding somatic cell tumorigenesis. A specific (i.e., extreme or hazardous) cellular environment may stimulate stem cell or cancer cell transformation of somatic cells. The mechanism of cell transformation and elucidation of the factors regulating cellular transformation contribute to understanding oncogenic changes in normal tissues and organs, as well as to establishing advanced strategies for preventing cancer-related changes in somatic tissues.