Re-emergence of undifferentiated cells from transplants of human induced pluripotent stem cells as a possible risk factor of tumourigenesis

Summary

Abstract:
Although induced pluripotent stem cells (iPSCs) are a potential source for transplantation therapy, malignant transformation (tumourigenesis) remains a major concern in their safe clinical application. iPSCs are considered more tumourigenic than embryonic stem cells (ESCs) because of genetic and epigenetic manipulations. We generated 22 human iPSC lines from normal human fibroblasts and injected three of these cell lines into SCID mice, and produced three tumours, all of which were identified as teratomas with at least two germ layers. Using cells cultured from them, RT-PCR showed that the cells expressed undifferentiated cell markers, including OCT4 and NANOG. This suggests that some undifferentiated cells remain in the teratoma during its formation. We also found emergence of cells expressing undifferentiated cell markers from teratoma-derived cells during culturing with the ESC medium. Immunocytochemical analyses showed that NANOG-, OCT4- and SSEA4-positive cells appeared and increased with time in culture. These data indicate that iPSC-like undifferentiated cells can emerge from differentiated cells under certain condition and they may present a potential risk of tumourigenesis, as do residual iPSCs.

Keywords:
differentiated cell; iPS cell; teratoma; transplantation; undifferentiated cell markers

Introduction

Human somatic cells were reprogrammed to pluripotent stem cells by ectopic expression of four reprogramming factors (OCT4, SOX2, KLF4 and MYC), their introduction being done by retrovirus-mediated genomic integration (Takahashi and Yamanaka, 2006; Takahashi et al., 2007). The generated induced pluripotent stem cells (iPSCs) are very similar to embryonic stem cells (ESCs) in their expression profiles of many genes, epigenetic status and the ability of teratoma formation, which shows the capacity to differentiate into the all three germ layers (Takahashi and Yamanaka, 2006; Maherali et al., 2007; Okita et al., 2007; Takahashi et al., 2007; Wernig et al., 2007; Yu et al., 2007). Therefore, many types of differentiated cells from somatic cells can be produced through iPSCs in vitro and used for patient-specific cell therapy.

However, there are serious safety concerns in clinical application of iPSCs, because integrated DNA has a risk of insertional mutagenesis. Two of reprogramming genes, MYC and KLF4, are themselves potent oncogenes. Therefore, the possible development of a tumour in transplantation of iPSC-derivatives exists, which must be addressed (Miura et al., 2009; Fong et al., 2010). Reactivation of integrated reprogramming genes is also concerned, such as Myc reactivation in tumour formation of chimeric mice (Okita et al., 2007). Reactivation of the other reprogramming factors may also cause tumours (Okita et al., 2007; Nakagawa et al., 2008; Miura et al., 2009; Walia et al., 2012), which makes them unsafe for clinical use (Miura et al., 2009; Walia et al., 2012).

If undifferentiated cells are present after in vitro or in vivo differentiation, it could become tumourigenic during transplantation (Erdö et al., 2003; Fujikawa et al., 2005; Teramoto et al., 2005; Brederlau et al., 2006; Miura et al., 2009). Fu et al. (2012) showed that a small number of mouse iPSCs could maintain their pluripotency during in vitro and in vivo differentiation, and that these contaminated undifferentiated iPSCs were still tumourigenic during transplantation. They have also observed the phenomena in mouse ESCs (Fu et al., 2012). Miura et al. (2009) also found that a small number of Nanog-positive cells exist during induced differentiation of mouse iPSCs. Therefore, it is important to evaluate this phenomenon in human iPSCs.

We have previously generated iPSC lines from the normal human fibroblast TIG-1 (Kumazaki et al., 2011) by using pCX-OKS-2A and pCX-c-Myc plasmids (Okita et al., 2008). And we have also established normal immortal cell lines from the same TIG-1 (Kamada et al., 2012). These cell lines with the same genetic background are useful for the study of both aging and differentiation, and the former (iPSCs) are especially useful for the study of regenerative medicine. Using these iPSCs, we obtained teratomas by injecting them into severe combined immune-deficiency (SCID) mice and isolated cells from the teratomas for culture, and now show evidence of their existence of undifferentiated cells in the culture. Teratoma-derived differentiated cells were also seen to convert to undifferentiated cells after culturing them with the medium for ESCs. Thus iPSCs or iPSC-like cells probably emerge from differentiated cells derived from iPSC teratoma.

Materials and methods

Cell culture

Feeder cell line SNL76/7 provided from Welcome Trust Sanger Institute was cultured with Dulbecco's modified eagle medium (DMEM, Invitrogen) containing 7 or 10% fetal bovine serum (FBS, Gibco). Human iPSCs lines (Kumazaki et al., 2011), which were generated from TIG-1 fibroblasts (Ohashi et al., 1980) by introduction of pCX-OKS-2A and pCX-c-Myc plasmids (Okita et al., 2008), were cultured with the medium for human ESCs (Knockout DMEM (Invitrogen) containing 15% Knockout serum replacement (Invitrogen), 2 mM glutamine (Gibco), 1 mM 2-mercaptoethanol (Gibco), 0.1 mM non-essential amino acids (Invitrogen) and 4 ng/ml recombinant human FGF-2 (Miltenyi Biotech)).

Teratoma formation and culture of teratoma-derived cells

The iPSCs were harvested by centrifugation after treatment with dissociation solution (ReproCELL, Japan), and the cell pellet was resuspended in human ESC medium. One half or all of the cells from 9 cm culture dish at ∼50% confluence was injected subcutaneously in the lateral flank of SCID mice (CREA, Japan). Five to 13 weeks after injection, the tumours were resected, fixed in 4% paraformaldehyde (Sigma), and prepared for hematoxylin and eosin (HE) at Sapporo General Pathology Laboratory Co., Ltd. (Japan).

This animal study was carried out in accordance with the recommendation in the Guide for the Care and Use of Laboratory Animals of National Institutes of Health. The protocol was approved by the Committee on the Ethics of Animal Experiments of Tokushima Bunri University (Permit Number: KP10-61-004). Surgery was done under ether anesthesia, and all efforts were made to minimise suffering. In parallel with fixation of tissues, a part of fresh tumour was minced with knives and treated with 0.5% collagenase and 0.125% trypsin for 20–30 min at 37°C. The digested tissue was filtered through 100 μm cell strainer (Falcon), the isolated cells collected by centrifugation and seeded in appropriate size of dishes with MCDB 131 (Sigma) containing 10% FBS. The cells derived from K9, K12 and K17 teratomas were named as K9te, K12te and K17te, respectively. After 2 weeks or more, cell colonies with different morphology formed in the culture dishes, and selected colonies were expanded in a separate culture dishes as clones, of which 7, 6 and 5 came, respectively, from K9, K12 and K17 teratomas. Some of the cultures were grown in medium for human ESCs.

RNA isolation, reverse transcription and polymerase chain reaction

Total RNA was purified by using RNeasy mini kit (Qiagen, Germany) and treated with DNase using Turbo DNA-free kit (Ambion). One microgram of total RNA was used for reverse transcription reaction with Transcriptor high fidelity cDNA synthesis kit (Roche, Switzerland). PCR was performed with appropriate primer set and Primestar DNA polymerase (Takara Bio, Japan). A semi-quantitative analysis of PCR bands was performed using Multi Gauge ver. 3.0 (Fujifilm). Primer sequences used are given in Table 1.

Immunocytochemistry

Cultured cells were fixed with 4% paraformaldehyde for 15 min at room temperature. After washing with PBS, the cells were treated with 0.1% Triton X-100 for 15 min at room temperature, and blocked with 3% bovine serum albumin (Nacalai tesque, Japan) for 3 h at room temperature. The cells were treated with primary antibodies overnight at room temperature. Primary antibodies used were anti-OCT4 (1:500, rabbit polyclonal, Abcam) and anti-human NANOG (1:500, mouse monoclonal, Abnova, Taiwan). Secondary antibodies (1:3,000, Molecular Probes) used for these primary antibodies were Alexa488-conjugated goat anti-mouse IgM, Alexa488-conjugated goat anti-rabbit IgG, Alexa555-conjugated goat anti-mouse IgM, Alexa555-conjugated goat anti-rabbit IgG or Alexa594-conjugated goat anti-rabbit IgG. Primary antibodies for SSEA4 (1:100–1:300, BD Pharmingen) was conjugated with Alexa Flour 555. The fluorescence microscope used for all experiments is an Axiovert 200M with DAPI, FITC, Rhodamine and Cy5 filter sets (Carl Zeiss). The fluorescence of Alexa488, Alexa555 and Alexa594 were detected by FITC, Rhodamine and Cy5 filters, respectively. Fluorescent area was quantified with Multi Gauge ver. 3.0. Three photographs for each sample were analysed and the mean percentage of fluorescent area was determined. Statistical analyses used one-way analysis of variance (ANOVA) followed by Tukey's test to determine significant differences at P < 0.05.

Results and discussion

Formation of teratoma from human iPSC lines

We generated 22 iPSC lines from human fibroblast strain, TIG-1 (Kumazaki et al., 2011), from which cells of TIG-1 iPSC lines K9, K12 and K17 were injected into SCID mice and developed tumours over 5–13 weeks. Sections of K9 had glandular epithelium (endoderm), blood vessels and striated muscle (mesoderm; Figure 1A). Those of K12 tumour had glandular epithelium (endoderm) and connective tissue (mesoderm; Figure 1B); and those of K17 tumour had glandular epithelium hyperplasia (endoderm), blood vessel hyperplasia and cartilage-like tissue (mesoderm; Figure 1C). Histopathologically the tumours were classified as teratomas.

Expression of undifferentiated-cell specific genes in teratoma-derived cells

To investigate whether the suppression of undifferentiated-cell markers, such as OCT4 and NANOG, occurred in the teratoma clones, the expression of these markers was examined by RT-PCR. Among 7 clones from K12 teratoma, 4 clones (L2, L4, L11 and L12) expressed human NANOG and 3 clones (L4, L11 and L12) expressed endogenous human OCT4 when they were cultured in MCDB medium (Figure 2A). L4, L11 and L12 also expressed hTERT, suggesting that they are residual undifferentiated cells in the teratomas. The difference in expression of undifferentiated cell markers among these clones might be due to differences in their degree of differentiation.

Expression of undifferentiated-cell markers in K9te and K17te clones that were cultured with MCDB medium were also examined. NANOG was expressed in five clones (LR2, LR10, LR21, LR42 and LW22) in K9te clones, though expression of endogenous OCT4 occurred only in two clones (LR10 and LW22; Figure 2B, lanes M). In K17te clones, expression of NANOG occurred in three clones (#38, #18 and #20), though that of endogenous OCT4 was detected only in one clone (#38) (Figure 2C, lanes M). Therefore, all the teratomas appeared to contain residual undifferentiated cells, as shown in mouse iPSCs (Fu et al., 2012).

Emergence of undifferentiated cells from teratoma-derived differentiated cells

All the clones showed different morphologies from those of iPSCs in MCDB medium (Figure 3), though some clones expressed undifferentiated cell markers, as mentioned above. Some clones, such as R4, R7 from K12te, R16 from K9 and #25 from K17te, did not express NANOG, h-OCT4 andh-TERT (Figure 2). When MCDB medium was switched to ESC medium, most cells died, probably because of lack of serum and/or growth factors, except for FGF-2. However, some growing cells appeared later. Morphology of these cells differed from those of iPSCs and the original differentiated cells (Figure 3). The teratoma-derived cells cultured with MCDB 131 medium were spindle-shaped and grew as a flat monolayer. In contrast, re-emerging cells in ESC medium had a small rounded shape and piled up like colonies. The colony of re-emerging cells had unclear boundaries, although those of iPSCs were a flat monolayer mass with striking boundaries. These re-emerging cells were considered to be developing into iPSCs.

Regarding the expression of undifferentiated-cell markers in these cultures as detected by RT-PCR and semi-quantitative analyses (measuring their PCR band densities), RNAs were prepared from K9te clones on 11–25 days after switching to ESC medium. Expression of endogenous human OCT4 was detected in three clones (LR2, LR21 and R16) cultured with ESC medium (Figure 2B, lanes E), even though OCT4 expression had been absent from them when cultured in MCDB medium (Figure 2B, lanes M). This indicates that OCT4 expression could be induced by ESC medium in these cells. OCT4expression was increased in 2 clones (LR42, ∼7.8-fold and LW22, ∼2.9-fold) that expressed OCT4 at low level when they were cultured with MCDB medium (Figure 2B). Similarly, expression of NANOGwas also induced (R16) or increased (LR21, ∼4.2-fold and LW22, ∼1.8-fold) by ESC medium. In testing whether induction of undifferentiated cell markers occurred in K17 clones, expression of endogenous OCT4 was found to be induced only in one clone (#25) by switching to ESC medium. In contrast, NANOG expression was induced (#25 and #1) or increased (#38; ∼2.6-fold (3d), #18; ∼3-fold (3d) and #20; ∼2.7-fold (3d)) in all the five clones (Figure 2C, lanes E). The data suggest that undifferentiated cells re-emerged and increased when cultured in ESC medium. Two possible reasons are, first, residual undifferentiated cells grow selectively in ESC medium. Since serum-free ESC medium contains only FGF-2 as a growth factor, normal differentiated cells probably cannot grow. ESC medium is suitable for the growth of undifferentiated cells, such as ESCs and iPSCs. Second, teratoma-derived differentiated cells revert back to an undifferentiated state. FGF-2 is involved in dedifferentiation of keratinocytes and oligodendroglia (Grinspan et al., 1996; Sun et al., 2011); these cells could be converted into their precursor cells in the presence of FGF-2. FGF-2 has a direct role in modulating downstream in the Wnt pathway to maintain undifferentiated human ES cell (Ding et al., 2010). Therefore, FGF-2 in ESC medium may be a main factor causing dedifferentiation of teratoma-derived cells.

We also investigated the expression of transgenes in these clones. The ectopic expression of Oct4,Klf4 and Sox2 were detected in all K9te clones when cultured in MCDB medium (Figure 2B). After replacement with ESC medium, transgene expression increased in some clones. For example, the expression of Oct4-Klf4 in LR2 and LR21 were increased ∼6-fold. In the case of K17te clones, the ectopic expression of Oct4 and Klf4 were not detected in #25 clone cultured in MCDB medium. However, the transgene expression was induced by culturing in ESC medium (Figure 2C). In #38 clone, the transgene expression was weak, even when the clone was cultured in ESC medium. The expression of Oct4-Klf4 in #38 (lane E-17d) was ∼6% of that in #25 (lane E-28d). Interestingly, inductions of endogenous NANOG and OCT4 expression by culturing in ESC medium were weak in this clone (Figure 2C). The expression of NANOG in #38 (lane E-17d) was 5% of that in #25 (lane E-28d). Transgene expression might be involved in the induction of endogenous undifferentiated-cell markers.

Expression of undifferentiated cell specific proteins

We also examined the expression of undifferentiated cell markers by immunocytochemical staining using two clones of K17te. An undifferentiated cell surface protein, SSEA4, was barely detectable in the cells of #25 clone cultured in MCDB medium (Figure 4A, panels MCDB). Other undifferentiated-cell markers, OCT4 and NANOG were also weakly detected. After 7 days in ESC medium, SSEA4, OCT4 and NANOG were detected, although very few cells were expressing these markers (Figure 4A, panels ESCM 7d). With continued culture in ESC medium, undifferentiated cell marker-positive cells increased (Figure 4A, panels ESCM 3d–21d), as also with other clones, for example K17te #1 (Figure4B). Figure 4C shows the fluorescent area getting more extensive with culture time. These results are consistent with the RT-PCR findings. We considered that these cells were iPSC-like cells reprogrammed again by culturing in ESC medium, because reprogramming genes were integrated in their genomes and ectopic expression of Oct4, Sox2, Klf4 and c-Myc was detected. NANOG is involved in self-renewal of undifferentiated ES cell and essential for attaining the ground state of authentic pluripotency in the final phase of somatic cell reprogramming (Chambers et al., 2003; Chambers et al., 2007). Therefore, re-emerging iPSC-like cells expressing NANOG seemed to be pluripotent undifferentiated cells. In fact, partially reprogrammed pre-iPSCs do not express NANOG (Silva et al., 2008, 2009; Papp and Plath, 2011).

Our in vitro study suggests that iPSC-like cells might re-emerge from the population of differentiated cells under certain conditions. Since substances such as FGF-2 included in ESC medium are also present in the matrix of body, we believe that this phenomenon can also occur in vivo. The re-emerging iPSC-like cells may be tumourigenic, like residual iPSCs are when transplanted in vivo. However, further in vivo study will be needed to demonstrate actual tumourigenic differentiation or tumour development of the dedifferentiated cells after their re-transplantation. We suggest a new view point on the risk of tumourigenesis, which is the emergence of iPSC-like cells from differentiated cells in some conditions.

Acknowledgments and funding

Financial support for this work was given by a regular research fund from Tokusima Bunri University.

Author contribution

Tsutomu Kumazaki and Youji Mitsui conceived, designed and performed experiments. Tsutomu Kumazaki and Tomoko Takahashi analysed and Taira Matsuo and Mizuna Kamada interpreted the data. Taira Matsuo, Youji Mitsui and Tsutomu Kumazaki wrote the paper.

Corresponding author: e-mail: y-mitsui@kph.bunri-u.ac.jp