Maintenance of Multipotency in Human Dermal Fibroblasts Treated with Xenopus laevis Egg Extract Requires Exogenous Fibroblast Growth Factor-2

Abstract

Direct reprogramming of a differentiated somatic cell into a developmentally more plastic cell would offer an alternative to applications in regenerative medicine that currently depend on either embryonic stem cells (ESCs), adult stem cells, or induced pluripotent stem cells (iPSCs). Here we report the potential of select Xenopus laevis egg extract fractions, in combination with exogenous fibroblast growth factor-2 (FGF2), to affect life span, morphology, gene expression, protein translation, and cellular localization of OCT4 and NANOG transcription factors, and the developmental potential of human dermal fibroblasts in vitro. A gradual change in morphology is accompanied by translation of embryonic transcription factors and their nuclear localization and a life span exceeding 60 population doublings. Cells acquire the ability to follow adipogenic, neuronal, and osteogenic differentiation under appropriate induction conditions in vitro. Analysis of active extract fractions reveals that Xenopus egg protein and RNAs as well as exogenously supplemented FGF2 are required and sufficient for induction and maintenance of this phenotypic change. Factors so far identified in the active fractions include FGF2 itself, transforming growth factor-β, maskin, and nucleoplasmin. Identification of critical factors needed for reprogramming may allow for nonviral, chemically defined derivation of human-induced multipotent cells that can be maintained by exogenous FGF2.

Introduction

Mammalian oocytes and eggs have long been recognized as suitable “environments” for studying fundamental questions in cell biology, such as regulation of the cell cycle, DNA replication, chromatin remodeling, and transcription. While both developmental stages, immature oocyte (arrested at prophase of the first meiosis, MI) and mature egg (arrested at metaphase of the second meiosis, MII), may be used as recipient cytoplasm, it is important to remember that oocytes are actively engaged in transcription (Gall and Callan, 1962; Scheer et al., 1976), whereas eggs are transcriptionally inactive but able to support DNA replication after activation (Blow and Laskey, 1986). It has been shown that nuclear components that are released from oocyte nuclei during maturation from MI to MII are required for successful reprogramming of somatic cell nuclei after nuclear transplantation, because removal of oocyte nuclei also removes nuclear components required for reprogramming (Hansis et al., 2004; Kikyo et al., 2000). Several cloning strategies have demonstrated that nuclei from terminally differentiated cells, when introduced into mature eggs, can be successfully reprogrammed, form embryos, and, after transplantation into recipient animals, develop to term (Wilmut et al., 1997).

Developmental plasticity of somatic cell nuclei can also be at least partially reactivated when nuclei are exposed to factors present in cytoplasm of pluripotent cell types, such as embryonic stem cells (ESCs) and embryonic carcinoma (EC) cells. For example, fusion of somatic cells with ESCs or EC cells leads to X chromosome reactivation within the hybrids (Tada et al., 2001), changes in gene expression profile (Cowan et al., 2005; Pereira et al., 2008; Zhou and Melton 2008), and acquisition of stem cell properties, including contribution to all germ layers in teratomas and in aggregation chimeras (Cowan et al., 2005; Pells et al., 2002; Tada et al., 1997; Tada et al., 2001; Terada et al., 2002; Ying et al., 2002).

Treatment of reversibly permeabilized somatic cells with extracts of ESC or EC cells induces expression of genes associated with pluripotency, such as OCT4, NANOG, and SOX2; causes downregulation of somatic cell-specific genes, such as lamin A; and enhances their in vitro differentiation capacity (Taranger et al., 2005). These modifications are observed at least temporarily after the resealing of permeabilized cells incubated in cell extracts (Collas and Taranger, 2006). In addition to mammalian systems, Xenopus oocyte extracts have been used extensively for their ability to affect chromatin structure (Dimitrov and Wolffe, 1996), cell cycle and DNA replication (Lu et al., 1999), and gene expression in cultured mammalian cells (Gurdon and Byrne, 2003; Miyamoto et al., 2007), demonstrating conservation of molecular regulatory mechanisms across species. The cytoplasm of mature Xenopus oocytes alters chromatin structure through a series of DNA and DNA-binding protein changes that lead to expression of early embryonic and developmental genes (Byrne et al., 2003; Freberg et al., 2007; Kimura et al., 2004; Taranger et al., 2005). Specifically, treatment of mammalian somatic cells with complete Xenopus egg cytoplasmic extracts induces expression of OCT4 (Byrne et al., 2003), NANOG (Koziol et al., 2007), and SOX2 (Miyamoto et al., 2007). Although these very significant changes do take place, there are no reports of attempts to maintain the extract-reprogrammed cells for longer periods of time in culture (over 20 days).

The Xenopus egg extract system offers an opportunity to identify cytoplasmic factors that have the ability to reprogram somatic cell nuclear memory. We investigated the composition and the ability of different extract fractions to reprogram nuclei of adult human dermal fibroblasts. We monitored removal of somatic histone H1, reactivation of stem cell transcription factors, and their expression and localization to cell nuclei. We investigated the role of fibroblast growth factor-2 (FGF2) in not only induction of these changes, but also its role in their maintenance over prolonged periods of time in vitro. Finally, we examined cells' potency by the ability to cross lineage differentiation boundaries upon select differentiation conditions and the ability to form teratomas in severe combined immunodeficient (SCID) mice.

Materials And Methods

Cell culture

Primary adult human dermal fibroblasts (CRL #2352, American Type Culture Collection) were obtained at passage p3. Cells were expanded in Dulbecco's modified Eagle medium (DMEM)/Ham's F12, supplemented with 10% fetal calf serum [Fetal Clone III (FCIII), Hyclone], and 4 mM l-glutamine (MediaTech). Cultures were passaged at 80% confluence using standard procedures. Ten days prior to electroporation, serum concentration in culture medium was gradually decreased to 0.5% to induce accumulation of cells in G₁/G₀. After electroporation, cells were seeded into 24-well plates at 2000 cells per well and grown in DMEM/F12, supplemented with 10% FCIII and 2 mM l-glutamine, with or without 4 ng/mL FGF2 (PeproTech). The number of population doublings was calculated as log₂ (#final/#initial). All cultures were incubated at 37°C, 5% O₂, 5% CO₂, and high humidity. Teratocarcinoma cells (NCCIT, American Type Culture Collection) were grown as recommended by the supplier.

A sample of control and extract-treated (electroporated) cells was incubated with 10 μM bromodeoxyuridine (BrdU; Invitrogen) for 24 h, washed, and fixed in ice-cold methanol. DNA was denatured with 1 N HCl, and cells were incubated with mouse monoclonal anti-BrdU-AlexaFluor 488–conjugated antibody (Caltag Laboratories). DNA was labeled with 0.2 μg/mL 4′,6-diamidino-2-phenylindole (DAPI).

EGFP–histone H1 vector

Human histone gene 1.2 was amplified from genomic DNA by PCR using the following primer pairs (hHis1.2, forward, 5′-GGATCCATGTCCGAGACTGCTCCTGCC-3′, and hHis1.2, reverse, 5’′-CCCGGGCTATTTCTTCTTGGGCGCCGC-3′). The underlined sequences denote the built-in BamHI and SmaI restriction sites for the forward and reverse primers, respectively. Histone 1.2 was cloned into a pCR2.1 TOPO vector (Invitrogen), and histone 1.2 inserts (654 bp) were isolated by BamHI/SmaI digestion. These fragments were cloned into a BglII/SmaI-digested vector set pAcGFP1-C1 to develop in-frame EGFP–histone 1.2 fusion genes, where the histone genes were fused to the carboxyl terminal of enhanced green fluorescent protein (eGFP) so as to retain full functionality of the histone gene.

Expression of GFP was monitored in real time using an IX81 inverted microscope (Olympus), equipped with fluorescence and a cooled CCD camera (Orca, Hamamatsu). Images were acquired using a GFP filter and processed using SlideBook (Olympus).

Xenopus laevis egg extract preparation

Metaphase-arrested egg extracts were prepared using published protocols (Danilchick et al., 1991). Mature Xenopus laevis females were superovulated with pregnant mare's serum gonadotropin (PMSG; Calbiochem) and 72 h later induced to ovulate with human chorionic gonadotropin (hCG; Calbiochem). Eggs were collected in cold MMR buffer (100 mM NaCl, 2 mM KCl, 1 mM MgCl₂, 2 mM CaCl₂, and 5 mM HEPES) and washed two times with High Salt Barth Solution (NaCl 110 mM, Tris-HCl 15 mM, KCl 2 mM, NaHCO₃ 2 mM, MgSO₄ 1 mM, Na₂HPO₄ 0.5 mM, and EGTA 2 mM). The jelly coats were removed with cold 2% l-cysteine-free base with 2 mM EGTA at pH 7.8 (adjusted with 6 N NaOH). Eggs were washed in inactivating extract buffer (50 mM KCl, 50 mM HEPES, 5 mM MgCl₂, 5 mM EGTA, and 2 mM β-mercaptoethanol). Eggs were crushed by centrifugation at 10,000 rpm for 15 min. The cytoplasmic layer was removed and centrifuged at 20,000 rpm for 15 min at 4°C. The translucent layer was removed and diluted 1:2 with extract dilution buffer at 4°C (50 mM KCl, 50 mM HEPES, 0.4 mM MgCl₂, and 0.4 mM EGTA) supplemented just before use with 2 mM dithiothreitol (DTT), 10 μg/mL aprotinin, leupeptin, and cytochalasin B, each. Extracts were centrifuged again at 100,000×g for 1.5 h at 4°C. High-speed egg extract was aliquoted at 50 μL/vial, snap frozen, and stored in liquid nitrogen. Total RNA was isolated and used for electroporation to determine whether RNA components alone could account for the fraction activity.

Extract fractionation by gel filtration chromatography

High-speed egg extract in extract dilution buffer was supplemented just before use with 2 mM DTT and 10 μg/mL each aprotinin, leupeptin, and cytochalasin B. The material was processed over a Superdex200 gel filtration resin (GE Healthcare, cat. no. 17-1043-02) using an XK26/100 column (GE Healthcare). Superdex200 (S200) resin was equilibrated with extract dilution buffer overnight at 4°C, and the column was calibrated by running gel filtration standards (BioRad Laboratories). Ten percent column volume of high-speed extract was loaded using an AKTÄ Explorer A200 FPLC, and flow rate was maintained constant at 2 mL/min. During the run, absorbance was recorded at A280 nm and A254 nm. The chromatogram generated from the fractionation process was analyzed in comparison with the chromatogram generated from the column calibration. Ninety-two 4-mL fractions were collected in the range between 10 and 600 kDa. Concentration of protein and RNA in each fraction was determined by fluorometry (Qubit, Invitrogen). The proteins in each 4-mL fraction were precipitated with acetone and resuspended to 10 mg/mL final concentration in extract dilution buffer. Samples for denaturing sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) analysis were prepared by mixing 40 μL of each fraction with 10 μL of 5×SDS-PAGE loading buffer. Gels underwent electrophoresis at 160 V for 1 h (BioRad Laboratories) and silver stained. RNA was isolated from fractions using the procedure described below, and DNA was removed by treatment with rDNase (according to the manufacturer's instructions; Ambion) and separated on 4% agarose gels.

Extract electroporation

Approximately 1×10⁶ cells were electroporated in 100 μL of pooled extract fractions (500 μg/mL protein) with three pulses (10 sec apart), at 0.56 kV/cm for 1 msec/pulse (BTX). Electroporated cells were resuspended in serum-free medium. Removal of EGFP–histone H1 was monitored in real time using an Olympus IX81 inverted microscope equipped with an environmental chamber. Images were acquired using Slidebook software (Olympus) during the 2-h period postelectroporation. Cells were subsequently plated in DMEM/F12, 10% FCIII, with or without 4 ng/mL FGF2. All cell cultures were carried out at 37°C, 5% O₂, 5% CO₂, and high humidity.

Total RNA extraction and RT-PCR

Total RNA was isolated using TRIzol (Invitrogen) following the manufacturer's protocol. Four micrograms of total RNA was used to perform first-strand cDNA synthesis using Superscript (Invitrogen). PCR was performed using 0.5 μL of first-strand cDNA in Mg²⁺-free PCR TaKaRa polymerase buffer supplemented with 1.5 mM MgCl₂, 200 μM each of deoxynucleotide triphosphates (dNTPs), 25 pmoles each of forward and reverse primers, and 0.5 U of TaKaRa ExTaq polymerase per reaction. PCR cycling was done as follows: Initial denaturation at 95°C for 2 min followed by 35 cycles of denaturation at 95°C for 15 sec; annealing at primer-specific annealing temperature for 1 min; and extension at 72°C for 1:30 min. Final extension was done at 72°C for 10 min, and the samples were held at 4°C until use. Amplification products were resolved on 2% agarose gels containing 0.5 μg/mL ethidium bromide in 1×TAE buffer and photographed using a Kodak 4000MM Image Station. Primers for amplification of human OCT4 (POU5F1) transcript were designed to detect the ESC-specific transcript that produces the 360-amino-acid variant (NM_002701). Primer sequences used in this study are presented in Table 1.

Table 1.

RT-PCR Primer Sequences

Target gene	Forward primer	Reverse primer	Amplicon size (bp)	Genomic size (bp)	Accession no.
OCT4	5′-GTTGATCCTCGGACCTGGCTA-3′	5′-GGTTGCCTCTCACTCGGTTCT-3′	646	5339	NM_002701
NANOG	5′-TGTCTTCTGCTGAGATGCCTCACA-3′	5′-CCTTCTGCGTCACACCATTGCTAT-3′	387	4811	NM_024865
hTERT	5′-GCTTCCTCAGGAACACCAAGA-3′	5′-TGCAACTTGCTCCAGACACTC-3′	298	1657	NM_198253
GAPDH	5′-ATCACCATCTTCCAGGAGCGA-3′	5′-TTCTCCATGGTGGTGAAGACG-3′	101	101	NM_002046

Western blotting

High-speed extract was supplemented with complete protease inhibitor cocktail (PIC; Santa Cruz Biotechnology) and 1 mM DTT. Protein concentration was determined with a Quant-iT protein assay kit (Invitrogen). Equal amounts of protein and 2×sample buffer (BioRad Laboratories) were mixed and heated to 95°C for 5 min. Proteins were separated on 4–20% gradient SDS-PAGE gels and transferred to nitrocellulose membranes (BioRad Laboratories) using Towbin's transfer buffer (25 mM Tris, 192 mM glycine, 20% methanol, and 0.037% SDS). The membranes were blocked with Tween Tris-buffered saline (TTBS; 25 mM Tris, 137 mM NaCl, 2.7 mM KCl, and 0.2% Tween), 5% dry milk (Santa Cruz Biotechnology), and 5% fetal calf serum (FCS). The same buffer was used for primary and secondary antibody incubations. Primary antibodies against OCT4 (Chemicon), NANOG (Abcam), FGF2, TGFβ1, and maskin (all from Santa Cruz Biotechnology) and nucleoplasmin (Developmental Studies Hybridoma Bank) were used. Appropriate horseradish peroxidase (HRP)-conjugated secondary antibodies were used for detection. In between antibody incubations, membranes were washed three times with Tween/Tris-buffered saline (TTBS). Membranes were incubated in luminol (Santa Cruz Biotechnology), and luminescence was detected with Kodak 4000MM imager.

Immunocytochemistry

Cells were fixed with ice-cold methanol or 2% methanol-free formaldehyde. Samples were treated with 1 N HCl prior to blocking and labeling. Cells were blocked with 5% FCS in TTBS. Primary antibodies against and fibroblast-specific protein (FSP; Sigma), NANOG (Abcam), OCT4 (Chemicon), nestin (Abcam), and neuronal β-tubulin III (Tuj1; Covance) were diluted in TTBS and 3% FCS. Cells were labeled for 45 min at room temperature with primary antibodies. Plates were washed three times with TTBS and appropriate secondary antibody conjugated to AlexaFluor 488 (Molecular Probes) diluted in TTBS with serum added for 45 min. Antibodies were cross-linked with 4% formaldehyde for 5 min, and washed three times with DAPI added to the last wash. Plates were stored in TBS in the dark at 4°C until imaged. Cells were analyzed using an IX81 inverted microscope (Olympus) equipped with fluorescence and a cooled CCD camera (Orca, Hamamatsu). Images were acquired using appropriate filters and processed using SlideBook (Olympus).

Differentiation protocols

Adipogenic differentiation

Cells were cultured in growth medium supplemented with 10 μM dexamethasone, 100 mg/mL 3-isobutyl-1-methylxanthine, 50 mM indomethacin, and 10 mg/mL insulin (all from Calbiochem) for 3 weeks with a medium change twice a week. Cells were then rinsed twice with phosphate-buffered saline (PBS), fixed with 10% formalin for 10 min, washed with distilled water, rinsed in 60% isopropanol, and covered with a 0.3% Oil Red O solution (HyClone) in 60% isopropanol. After 10 min, cultures were briefly rinsed in 60% isopropanol and thoroughly in distilled water and then dried at room temperature. Samples were counterstained with Hematoxylin.

Osteogenic differentiation

Cells were grown in medium supplemented with 1 μM dexamethasone, 10 mM β-glycerophosphate disodium, and 50 mg/mL ascorbic acid (all from Sigma). After 4 weeks, cells were rinsed twice with PBS, fixed with formalin for 10 min, and washed with distilled water. To stain calcium deposits, cells were covered with a 2% aqueous solution of Alizarin Red S (Sigma) at pH 4.2 for 3 min. Cultures were then washed thoroughly with distilled water and dried at room temperature.

Neuronal differentiation

Cell aggregates were transferred to cell culture dishes and cultured in Neurobasal medium with 1×N2 (Invitrogen), 4 ng/mL FGF2, and 1×insulin, selenium, and transferrin (ITS; Invitrogen) for 10 days. Neuroepithelium-like cultures were immunostained with antibodies against nestin and neuronal β-tubulin III and detected with AlexaFluor 488 and AlexaFluor 568-conjugated secondary antibodies, respectively. DNA was stained with DAPI.

Teratoma formation assay

Animal studies were done with Institutional Animal Care and Use Committees (IACUC)- approved protocols and in accordance with animal care and use procedures at Worcester Polytechnic Institute, Worcester, MA. One million of control and one million of extract-treated fibroblasts were mixed with 8- to 12-μm-diameter carbon beads in sterile Dulbecco's phosphate-buffered saline (DPBS) and injected into the hind leg muscle of SCID mice (Charles River Laboratories, Wilmington, MA). Animals were euthanized 6 weeks after injection, and the muscle was excised and processed for histology. Tissues were fixed in 4% formaldehyde in DPBS and embedded in paraffin. Sections were stained with Hematoxylin & Eosin (H&E), and the injection site was located by visualization of the carbon beads with microscopy.

This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocols were approved by the Institutional Animal Care and Use Committee of the Worcester Polytechnic Institute (protocol numbers #07-05 and 08-010). All efforts were made to minimize animal pain and suffering.

Results

Cell morphology and their in vitro life span

To eliminate the variability of cell response due to the stage of the cell cycle, the cells were synchronized in G₁ by gradually reducing serum concentration in the culture medium over 10 days from 10% to 0.5% (Fig. 1Aa). The BrdU assay was performed to show that all the cells had stopped cycling, thereby synchronizing them in G₀. By the end of the serum reduction period, cells stopped proliferating, as indicated by the absence of BrdU incorporation over 24 h of culture (Fig. 1Ab). Upon electroporation and addition of serum to the culture medium, both extract-treated cells as well as control cells re-entered the cell cycle. (Fig. 1Ac). To verify that the electroporation protocol indeed enabled transfer of large protein molecules into cells, we supplemented the extract with goat 9-amino-6-chloro-2-methoxyacridine (AMCA)-conjugated immunoglobulin G (IgG; Abcam) and monitored fluorescence in cells over 2 h postelectroporation (Fig. 1 Ba). Morphology of electroporated cells gradually changed during the following 6 weeks in culture, and cells displayed a higher nucleus-to-cytoplasm ratio. Cells were grown as adherent monolayers (Fig. 1Bb). For differentiation experiments, cells were transferred to low-adhesion tissue culture plastic and allowed to form aggregates resembling embryonic bodies (Fig. 1Bc).

FIG. 1.

Morphology, cell cycle synchronization, uptake of Xenopus egg extract, and long-term culture of electroporated human fibroblasts. Human dermal fibroblasts were grown to subconfluence (A, column a). Serum concentration was gradually reduced over 10 days to drive cells into G₁ as shown by lack of BrdU incorporation (A, column b). After electroporation (A, column c). (B) Uptake of extract was verified by 9-amino-6-chloro-2-methoxyacridine (ACMA)-labeled IgG fluorescence (a) and cells were grown either in adherent monolayers (b) or embryonic bodies (c). Bars, 20 μm (A) and 100 μm (B).

To examine whether or not the new cell phenotype could be maintained for prolonged periods of time, we cultured extract-treated and untreated cells in DMEM/F12, 10% FCIII with or without exogenously supplemented FGF2 (4 ng/mL, PeproTech). The number of population doublings between the two was significantly different depending on the presence or absence of FGF2. FGF2-supplemented cultures underwent over 60 population doublings as compared to 25–35 population doublings for non-FGF2 cultures (Fig. 2).

FIG. 2.

Cumulative population doubling data. Population doubling data of human dermal fibroblasts electroporated with Xenopus egg extract (XOE treated) and supplemented with FGF2 (diamond), electroporated with Xenopus egg extract (XOE treated) and no FGF2 added (squares), buffer electroporated control treated with added FGF2 (triangles), and non-electroporated human dermal fibroblast control grown with no added FGF2 in the culture media (circles).

Removal of histone H1

Removal of over 90% of linker histones during initial stages of reprogramming has been reported in somatic nuclei incubated with Xenopus egg extract (Dimitrov and Wolffe, 1996). To ascertain that the extract was having an effect on the initial stages of chromatin remodeling in fibroblast nuclei, we transfected human fibroblasts with EGFP–histone H1 fusion protein vector. Stable transfectants resulted in high proportions of cells expressing nuclear localized EGFP–histone H1 protein. Removal of nuclear fluorescence was monitored in real-time during 2 h postelectroporation with Xenopus egg extract. Strength of fluorescence signal was compared to non-electroporated cells and cells electroporated with buffer alone. Complete extract electroporation induced rapid removal of histone H1 (Fig. 3A), and the majority of the protein was undetectable 60 min after electroporation. Cells were subsequently electroporated with pools of 10 adjacent extract fractions obtained from the Superdex200 column. Only pools of extract fractions that demonstrated histone H1 removal were used for the electroporation experiments on non-transformed fibroblasts.

FIG. 3.

Removal of histone H1 and expression of stem cell genes in electroporated human fibroblasts. (A) Time-lapse of EGFP–histone H1 fusion protein removal from adult fibroblast nuclei at 15, 45, and 100 min after extract delivery. (B) RT-PCR analysis. Cells were not electroporated (Non Elect.) or were electroporated with extract dilution buffer (Elect. Buffer) or with Xenopus egg extract (Elect. Extract). Teratocarcinoma cells (NCCIT) were used as a positive control, and no template was a negative control. FGF2 was supplemented in the culture media as indicated by (+) or (−). (C) Western blot analysis for presence of OCT4 and NANOG in cells electroporated with extract dilution buffer (Elect. Buffer) and extract electroporated cells (Elect. Extract) grown with or without FGF2 for 7 days. (D) Immunocytochemistry of extract electroporated and buffer electroporated fibroblasts for FSP, SOX2, OCT4, and NANOG. Green, Alexafluor 488–conjugated secondary antibody; red, Alexafluor 568–conjugated secondary antibody; blue, DAPI. Bars, 10 μm (A and D, buffer and extract-treated cells); 200 μm (D, hESCs).

Expression of early developmental genes

To evaluate stem cell transcription factor activation, we investigated the presence of transcripts and their protein products in extract electroporated cells. No effect of extract electroporation on transcription of OCT4 or NANOG was observed by traditional RT-PCR on day 7 after electroporation (Fig. 3B). Non-electroporated fibroblasts supplemented with FGF2 in the culture medium contained significant amounts of mRNA for all of the transcription factors examined, an observation that we have recently extended for several stem cell–specific genes in human dermal fibroblasts treated with FGF2 under low oxygen conditions (Page et al., 2009). There was no detectable telomerase reverse transcriptase (hTERT) transcript in control fibroblasts, and electroporation alone or with extract fractions did not appear to induce hTERT expression. The increase in population doublings was accompanied by maintenance of their stem cell transcription factor expression and absence of hTERT (Fig. 3B).

Examination of cells using western blotting and immunocytochemistry (ICC), however, demonstrated that translation of OCT4 and NANOG occurred only in the extract electroporated group, indicating translational activation of these genes (Fig. 3C, D). In parallel with induction of OCT4 and NANOG mRNA translation, there was a concomitant reduction in the levels of FSP expression. OCT4 and NANOG proteins localized to the nuclei of electroporated cells, which would correspond to their site of activity as transcription factors. Overall, approximately 30% of extract electroporated cells displayed the above-described protein expression and localization properties. When tested alone for its potential to contribute to expression of stem cell genes, FGF2 had the ability to induce expression of OCT4 and NANOG proteins, as observed previously (Page et al., 2009; Page et al., 2011), an effect that could be achieved by the extract in the absence of exogenous FGF2 (Fig. 3C).

Components of Xenopus laevis egg extract

Our strategies in identification of active “dedifferentiation activity” components employed extract fractionation approach, reconstitution of activity with combinations of pools of adjacent fractions, and evaluation of the activity by a two-step bioassay. Step 1 involved real-time monitoring of removal of nuclear histone H1, and step 2 involved detection of OCT4 and NANOG proteins. On the basis of this two-step bioassay, two distinct pools of fractions were identified after Superdex200 fractionation. The high-molecular-weight pool (∼150–200 kDa) contained 10 fractions and the low-molecular-weight group (∼15–40 kDa) contained 20 fractions. Combination of the high- and low-molecular-weight pools together with addition of exogenous FGF2 reconstituted the activity of the total extract (Fig. 4A). Interestingly, electroporation of extract RNA alone did not induce removal of histone H1 or induction of OCT4/NANOG protein expression but rather caused significant cell death (data not shown).

FIG. 4.

Superdex200 fractionation of Xenopus laevis egg extract. Peak alignment from extract fractionation; the red boxes outline groups of fractions with reprogramming activity (A). Agarose gel separation of RNAs (B) and silver-stained SDS-PAGE analysis of some of the fractions collected during the fractionation step (C). Western blotting detected nucleoplasmin, maskin, and FGF2 and TGF-β1 in high- and low-molecular-weight fractions, respectively (D). EX1 and EX2 represent protein from two independently isolated egg extracts.

RNA was isolated from approximately every second fraction between fractions 28 and 92. Four percent agarose failed to resolve any RNA up to fraction 58. Similar RNA profiles were observed in fractions 63, 64, and 72, and a different pattern was seen in fractions 74 and up, with very little trace of RNA remaining in fraction 90. The lower number fractions exhibited the presence of distinct bands of small RNAs, with their levels gradually decreasing in the higher fraction. These small RNAs were above 100 bases or more, suggesting that they may include either primary miRNA or pre-miRNA (Ambady et al., 2012). The pools of fractions containing activity to induce OCT4/NANOG protein expression (Fig. 4B, fractions 74 and up) also contained significant amounts of large RNAs that were not detectable in fractions without this activity (Fig. 4B, lanes 59–72). Silver staining of SDS-PAGE–separated protein indicates significant enrichment of proteins in the expected molecular weight range (Fig. 4C). Western blot analyses identified the presence of FGF2 and TGF-β1 in the low-molecular-weight fractions and nucleoplasmin and maskin in the high-molecular-weight fractions (Fig. 4D).

Differentiation potential of extract-electroporated cells

Extract-treated cells were grown in aggregates resembling embryonic bodies (Fig. 5A). In three independent replicates from fibroblasts used for dedifferentiation at 10, 15, and 20 population doublings after treatment, cells repeatedly followed the expected differentiation pathway. Potential of cells to follow ectodermal tissue lineage development was indicated by acquisition of a neuroepithelial phenotype (Fig. 5B) and expression of nestin and Tuj1 (Fig. 5C, D), both markers of neuronal differentiation. Deposition of calcium in electroporated cells exposed to osteogenic conditions was detected with Alizarin Red (Fig. 5E). Under conditions that promote adipogenic differentiation, cells began accumulating lipids, detected with Oil Red O stain (Fig. 5F). Control (non-electroporated) cells remained fibroblast-like and showed no presence of adipogenic cells, Ca²⁺ deposits, or expression of neuronal markers. When injected into SCID mice, the extract-treated, FGF2-grown cells did not form teratomas.

FIG. 5.

In vitro differentiation of extract-treated fibroblasts. Embryonic-like bodies (A) formed neuroepithelial-like cells (B), expressed nestin (C), and supported maturation of cells into neurons (D, neuronal-specific β-tubulin Tuj1). Osteogenic induction produced cells expressing osteocalcin, detected with Alizarin Red (E). Adipogenic induction produced lipid-accumulating cells detected with Oil Red O stain and counterstained with Hematoxylin (F). Bar, 200 μm (A), and 50 μm (B–F).

Discussion

To reprogram a differentiated cell into a multi- or pluripotent phenotype, a number of molecular changes need to take place. It has been proposed that a sequence of events, including remodeling of nuclear chromatin leading to selective activation or repression of gene transcription, is required for transient or stable alteration of cell fate. Removal of DNA-binding proteins, such as linker histone H1 proteins, is important for initial chromatin decondensation (Kikyo et al., 2000) and for acquisition of transcriptional pluripotency, a process that is efficiently accomplished by nucleoplasmin (Dimitrov and Wolffe, 1996). Use of EGFP–histone H1 fusion protein allowed monitoring of its removal from the nucleus upon exposure to extract fractions. Our observation of rapid disappearance of EGFP–histone H1 is indicative of at least a partial erasure of the cells' epigenetic status, and is similar in its kinetics to the one observed by Dimitrov and Wolffe (1996). Two pools of extract fractions were identified that in combination and with addition of exogenous FGF2 demonstrated reprogramming activity and contained nucleoplasmin, a known H1 chaperone.

Removal of histone H1 can result in transcriptional activation of specific genes, such as 5S rRNA (Bouvet et al., 1994), and is followed by chromatin decondensation and establishment of transcriptionally permissive chromatin. RT-PCR analysis of transcripts for several early developmental and stem cell–associated genes, however, did not reveal a significant change in levels of any of the transcripts examined (Fig. 3B). This confirms our recent data where a number of differentiated cell lines contained detectable amounts of both OCT4 and NANOG transcripts (Ambady et al., 2010). Low levels of OCT4 expression have previously been reported in leukocytes (Hansis et al., 2004), peripheral blood mononuclear cells (Zangrossi et al., 2007), and recently in adult human fibroblasts (Page et al., 2009). The levels of NANOG and OCT4 transcripts, however, have no direct predictive value for the levels of respective proteins. It has been shown that the levels of transcripts represent both the embryonic gene and its pseudogenes, with which they share significant homology (Ambady et al., 2010; Atlasi et al., 2008; Zangrossi et al., 2007). Although little or no effect of egg extract in our study can be ascribed to transcription of the select stem cell genes, translation of their mRNAs was significantly affected by the extract in combination with exogenously added FGF2, and this potential mechanism needs further investigation. Immunocytochemistry and western blotting detected the presence of OCT4 and NANOG proteins in treated cells and the transcription factors localized to the cells' nuclei. Previously we described a very similar posttranscriptional activation of these genes using defined culture conditions with FGF2 supplementation of cell culture medium (Page et al., 2009); however, exogenously supplemented FGF2 alone does not have the potential to dedifferentiate human dermal fibroblasts into a multipotent state. Our active extract fraction pools contain significant amounts of FGF2 based on western blot analysis, similar to previously reported amounts in complete Xenopus egg extract (Kimelman et al., 1988).

Our findings show that full reprogramming of human dermal fibroblasts through permeabilization in vitro requires both the combination of active fraction pools from the extract as well as exogenously added FGF2. Whereas FGF2 has not been implicated previously in transcriptional activation of OCT4 or NANOG, it has been determined that the maintenance of expression of these genes and cell pluripotency is dependent on FGF2 (Levenstein et al., 2006). The proposed action of FGF2 involves induction of members of the transforming growth factor-β (TGF-β); TGF-β ligands maintain expression of OCT4, SOX2, and NANOG, which in turn activate expression of endogenous FGF2 to complete this regulatory loop (Greber et al., 2007). It is reasonable to hypothesize that a similar FGF2-dependent core regulatory circuit maintaining the self-renewal ability and maintenance of pluripotent state in embryonic stem cells may at least in part be needed for reprogramming of differentiated somatic cells. However, lack of hTERT reactivation and lack of the potential of the cells to form tumors in vivo suggest that additional events may need to take place within the cells before true pluripotency is achieved. One of these missing events could be regulated by extract RNAs, because only fractions containing both protein and RNAs demonstrated reprogramming activity.

The regularity of protein/small RNA co-elution may indicate a novel mechanism of transcriptional, translational, and/or post-translational regulation. Recently, Awe and Byrne identified eight candidate oocyte reprogramming factors (CORFs) (Awe and Byrne 2013) that closely correlate with the “chromatin loosening/enhanced reprogramming” hypothesis previously described by Gurdon and Wilmut (2011). Five of those eight CORFs (ARID2, ASF1A, ASF1B, H1FOO, and KDM6B) not only are expressed in Xenopus laevis oocyte (Awe and Byrne 2013), but their molecular weights also fit the S200 elution profile in our study, suggesting that these factors may be present in the pools of fractions that can reconstitute reprogramming activity of the total extract. Small RNAs, microRNAs in particular, are becoming increasingly important in our understanding of posttranscriptional regulation. While hundreds of targets will likely emerge from this research, it may be possible to identify a much smaller number of the “absolutely required” targets through development of an integrated map of components, their relationships, and their established cellular activities. This approach would enable selection of the most upstream acting molecules, their testing in the bioassay, and progressive elimination of unlikely targets. It is also expected that a number of novel proteins will be identified that may play an important role in the process.

The natural reprogramming activity of egg cytoplasm without the need for induced gene overexpression may lead to identification of the molecules and mechanisms of nuclear reprogramming, and consequently the use these molecules to enhance the efficiency of reprogramming could be applied for derivation of therapeutically relevant, patient-specific cell types. This would be of interest especially if the reprogrammed cells of one type could yield cells of another unrelated cell-type, rather than being pluripotent and potentially tumorigenic.

Acknowledgments

The anti-maskin antibody and anti-nucleoplasmin antibody were obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the National Institute of Child Health and Human Development (NICHD) and maintained by The University of Iowa, Department of Biological Sciences, Iowa City, IA 52242. This work was funded by the National Institutes of Health (grant no. R01GM085456) and DARPA and US Army Research Office award no. W911NF-09-0004 to Tanja Dominko.

Author Disclosure Statement

The authors declare that no conflicting financial interests exist.