OCT4 Spliced Variants Are Differentially Expressed in Human Pluripotent and Nonpluripotent Cells


  • Yaser Atlasi,

    1. Department of Genetics, Faculty of Basic Sciences, Tarbiat Modares University, Tehran, Iran
    Search for more papers by this author
  • Seyed J. Mowla,

    Corresponding author
    1. Department of Genetics, Faculty of Basic Sciences, Tarbiat Modares University, Tehran, Iran
    • Seyed J. Mowla, Seyed J. Mowla, Ph.D., Department of Genetics, Faculty of Basic Sciences, Tarbiat Modares University, Tehran, Iran, P.O. Box: 14115-175. Telephone: 98-2182883464; Fax: 98-2188009730

      Peter W. Andrews, Correspondence: Peter W. Andrews, Ph.D., Centre for Stem Cell Biology, Department of Biomedical Science. University of Sheffield, Western Bank, Sheffield, S10 2TN, U.K. Telephone: 44-1142224173; Fax: 44-1142222399

    Search for more papers by this author
  • Seyed A.M. Ziaee,

    1. Urology and Nephrology Research Center, Labbafi-Nejad Medical Center, Shahid Beheshti University of Medical Sciences, Tehran, Iran
    Search for more papers by this author
  • Paul J. Gokhale,

    1. Centre for Stem Cell Biology, Department of Biomedical Science, University of Sheffield, Sheffield, United Kingdom
    Search for more papers by this author
  • Peter W. Andrews

    Corresponding author
    1. Centre for Stem Cell Biology, Department of Biomedical Science, University of Sheffield, Sheffield, United Kingdom
    • Seyed J. Mowla, Seyed J. Mowla, Ph.D., Department of Genetics, Faculty of Basic Sciences, Tarbiat Modares University, Tehran, Iran, P.O. Box: 14115-175. Telephone: 98-2182883464; Fax: 98-2188009730

      Peter W. Andrews, Correspondence: Peter W. Andrews, Ph.D., Centre for Stem Cell Biology, Department of Biomedical Science. University of Sheffield, Western Bank, Sheffield, S10 2TN, U.K. Telephone: 44-1142224173; Fax: 44-1142222399

    Search for more papers by this author


OCT4 is a master regulator of self-renewal in embryonic stem cells and can potentially encode two spliced variants, designated OCT4A and OCT4B. We have examined the expression pattern of these OCT4 isoforms in various human pluripotent and nonpluripotent cells. Our data revealed that whereas OCT4A expression is restricted to embryonic stem (ES) and embryonal carcinoma (EC) cells, OCT4B can be detected in various nonpluripotent cell types. Furthermore, we detected a novel OCT4 spliced variant, designated OCT4B1, that is expressed primarily in human ES and EC cells and is downregulated following their differentiation. We also found a significantly higher level of OCT4B1 expression in stage-specific embryonic antigen-3 (SSEA3)(+) compared with SSEA3(−) subpopulations of cultured ES cells. Taken together, our data demonstrated a distinctive expression pattern for OCT4 spliced variants in different cell types and highlight the necessity of defining the type of OCT4 when addressing the expression of this gene in different human cells.

Disclosure of potential conflicts of interest is found at the end of this article.


Author contributions: Y.A., performance of experiments, experiment design, data interpretation, manuscript writing and editing; S.J.M., experiment design, data interpretation, manuscript writing and editing; S.A.M.Z. and P.J.G., data interpretation, manuscript writing and editing; P.W.A., experiment design, data interpretation, manuscript writing and editing.

OCT4 belongs to the family of POU genes and is distinguished by its crucial role in regulating the self-renewal and pluripotency networks in embryonic stem (ES) and embryonal carcinoma (EC) cells [1, 2, 34]. In the developing mouse embryo, OCT4 is expressed in early cleavage stage, inner cell mass, primitive ectoderm, and primordial germ cells and downregulated in the trophectoderm [5]. Suppression of OCT4 expression in developing embryo or in cultured human or mouse ES and EC cells leads to loss of pluripotency and differentiation into trophectoderm [4, 5], whereas transgene-mediated overexpression of OCT4 triggers differentiation of embryonic cells into endodermal or mesodermal structures [6, 7].

The human OCT4 gene is located on chromosome 6 in the region of the major histocompatibility complex and can potentially encode two different spliced variants, described initially as OCT4A and OCT4B [8]. Recent reports have shown that both isoforms share identical POU DNA-binding and C-terminal transactivation domains but differ in their N termini. In OCT4B, the N-terminal domain has an inhibitory effect on the DNA-binding domain, consequently OCT4B cannot stimulate the transcription from OCT4-dependent promoters [9, 10]. In addition, whereas OCT4A orchestrates the transcription of different genes in the nucleus, OCT4B is localized mainly in the cytoplasm and cannot sustain the self-renewal and pluripotency of ES cells [9, 10]. The role of the OCT4B isoform is still not clear, and no mouse ortholog has been found for this isoform. Most of the transcribed OCT4 pseudogenes have high similarity to the OCT4A but not the OCT4B sequence.

In addition to embryonic cells, germ cells, and germ cell tumors, OCT4 expression has already been reported in several adult somatic cells [11] and in various cancer cell lines and primary tumors [12, 13] (supplemental online Table 1). OCT4 expression in somatic cells has been suggested to be restricted to small populations of multipotent cells with high self-renewal capacity, namely the adult stem cells in normal tissues, or cancer stem cells in tumor samples [11, 14]. Recently, OCT4 expression has been also reported in terminally differentiated peripheral blood mononuclear cells (PBMCs) [15].

Because few studies on OCT4 expression discriminate between the two spliced variants [9, 11, 16], little information is available on the expression pattern of each isoform in different cell types. Furthermore, the presence of several transcribed pseudogenes, which have high similarity to OCT4A sequence, can be a potential source of false-positive results and might be misinterpreted in reverse transcription (RT)-polymerase chain reaction (PCR) experiments addressing OCT4 expression. Thus, we have now investigated the expression pattern of OCT4A and OCT4B spliced variants in different human pluripotent and nonpluripotent cell types.

Materials and Methods

Cell Culture

The human ES cell lines H7 [17], HUES-1 [18], and Shef5 (provided by H.D. Moore, Centre for Stem Cell Biology) were cultured in knockout Dulbecco's modified Eagle's medium (DMEM) supplemented with 20% serum replacement (Invitrogen, Paisley, U.K., http://www.invitrogen.com) and 4 ng/ml basic fibroblast growth factor (Peprotech, Rocky Hill, N.J., http://www.peprotech.com) on mouse embryonic fibroblast feeders mitotically inactivated with mitomycin C (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com), as previously described [19]. The human EC cell lines NTERA2 cl.D1 (NT2) [20] and 2102Ep cl. 2A6 (2102Ep) [21] and TERA-1 cells [22] were cultured as previously described in DMEM (Invitrogen) supplemented with 10% fetal calf serum (FCS; Invitrogen) in a humidified atmosphere of 10% CO2 in air.

Other cultured somatic cells that were tested included three cell lines derived from transitional cell carcinoma of bladder (RT4, SD, and HT 1376 [23]), two choriocarcinoma cell lines (BeWo and JAR [24]), SW480 (colon cancer [25]), Hela (cervical carcinoma), K562 (leukemia [26]), U-2 OS (osteosarcoma [27]), human mesenchymal stem cells (hMSCs) (obtained from the Institute for Cancer Studies, University of Sheffield), WI38 (fetal lung fibroblasts [28]), and 1411H (yolk sac carcinoma [29]). These cells were cultured in DMEM or α-minimal essential medium (for hMSCs) supplemented with 10% FCS in a humidified atmosphere of 5% CO2 in air. For passaging, NT2 cells were scraped by glass beads, and other cell types were released with 0.25% trypsin (wt/vol)/1 mM EDTA (Invitrogen).

Cell Differentiation

Differentiation of NT2 cells was induced as described previously [30]. Briefly, cells were seeded at a density of 2 × 106 cells per T75 flask and treated with all-trans-retinoic acid (Sigma-Aldrich) (10−5 M) for 2–21 days. Fresh media supplemented with retinoic acid were supplied every 7 days. Cells were harvested at different time points and used for RNA extraction.

For the generation of embryoid bodies, H7 human embryonic stem (hES) cells were dissociated into small clumps by 3-mm glass beads (Philip Harris Scientific, Leicestershire, U.K., http://www.philipharris.co.uk) and cultured in suspension as EBs in the same hES medium but on nonadherent sterile bacterial dishes [31]. EBs were fed with fresh medium every 3 days, harvested at days 6 and 11, and used for RNA extraction.

RNA Extraction and cDNA Synthesis

Total RNA was extracted using TRIzol reagent (Invitrogen) (107 cells per milliliter). To remove any DNA contamination, this RNA was treated with TURBO DNase (Ambion, Austin, TX, http://www.ambion.com) and quantified by spectrophotometry and electrophoresis on a 1% agarose gel. The first strand of cDNA was synthesized using 100 pmol of oligo(dT)18 primer (MWG Biotech, Ebersberg, Germany, http://www.mwg-biotech.com), 200 units of RevertAid H Minus MMuLV Reverse Transcriptase (MBI Fermentas, St. Leon-Rot, Germany, http://www.fermentas.com), and 1 μg of total RNA according to the manufacturer's instructions. For each sample, a no-reverse transcription (No-RT) control was used in parallel from the DNase-treated RNA to detect any potential nonspecific amplification of genomic DNA.

In Silico Analysis

To find specific nucleotides for OCT4A that could be used to design specific primers for the sequence, we performed a multiple alignment approach between OCT4A and other pseudogene sequences using the MegAlign software (DNASTAR Inc., Madison, WI, http://www.dnastar.com). Sequences corresponding to OCT4A, OCT4B, and six pesudogene (PG) sequences located on chromosomes 1, 3, 8, and 12 were used in this approach. The corresponding GenBank numbers were as follows: NM_002701.4 (OCT4A), NM_203289.3 (OCT4B), NG_005793 (PG on chromosome 12), NG_006104 (PG on chromosome 10), NG_006105 (PG on chromosome 3), NG_006106 (PG on chromosome 3), NR_002304.1 (PG on chromosome 8), and HS.632482 (PG on chromosome 1).

Semiquantitative and Real-Time RT-PCR Analysis

The appropriate PCR primers (MWG Biotech) were designed using Genrunner software (version 3.02; Hastings Software) and previously described human OCT4 and β2 microglobulin (β2M) sequences (GenBank accession numbers NM_002701/NM_203289 and NM_004048, respectively). The forward (F) and reverse (R) primers used were as follows: β2M-F, 5′-GGGTTTCATCCATCCGACATTG-3′; β2M-R, 5′-TGGTTCACACGGCAGGCATAC-3′; OCT-AF, 5′-CTTCTCGCCCCCTCCAGGT-3′; OCT-FB, 5′-AGACTATTCCTTGGGGCCACAC-3′; OCT-RB1, 5′-AAATAGAACCCCCAGGGTGAGC-3′; OCT-RB2, 5′-CTCAAAGCGGCAGATGGTCG-3′; OCT-RB4, 5′-CCCCCTGTCCCCCATTCCTA-3′.

PCR was performed using 0.5 μl of cDNA or No-RT sample with 1 U of Taq polymerase (Invitrogen), 1.5 mM MgCl2, 200 μM dNTPs, and 0.4 μM of each primer in 25 μl of PCR. The PCR amplification was performed for 35 (OCT4A), 32 (OCT4B), or 26 (β2M) cycles, and the thermal profile was as follows: 94°C for 30 seconds, 64°C for 30 seconds, and 72°C for 35 seconds, with a final extension at 72°C for 10 minutes. The PCR primers amplified 496-, 267-, 492-, and 167-base pair (bp) fragments from OCT4A, OCT4B, OCT4B1, and β2M cDNA, respectively. PCR products were separated on a 1.5% agarose gel, stained with ethidium bromide, and visualized under the UV light. The identity of the PCR products was confirmed by direct DNA sequencing (Genetics Core Facility, University of Sheffield).

Quantitative polymerase chain reactions (Q-PCRs) were performed using SYBR Green chemistry with the SYBR-Green JumpStart Taq ReadyMix kit (Sigma-Aldrich) on an iCycler Real Time Quantitative PCR System (Bio-Rad, Hercules, CA, http://www.bio-rad.com). cDNA corresponding to 20 ng of RNA was added to the SYBR-Green JumpStart Taq Ready Mix (0.2 μM of each specific primer, 10 μl of SYBR-Green JumpStart Taq ReadyMix) in a total reaction volume of 20 μl. Reactions were run in a 96-well format (thus permitting parallel analysis of genes in different cell lines) and always included H2O and No-RT negative controls, and β2M for positive control and normalization purposes. Reactions were repeated in triplicate, and the resultant mean threshold cycles were used for further analysis. The thermal profile included 95°C for 2 minutes, 40 cycles of denaturation at 95°C for 30 seconds, annealing at 64°C for 1 minute, and elongation with optics on for fluorescence monitoring at 72°C for 1 minute.

The threshold cycle (Ct) for individual reactions was identified using iCycler IQ sequence analysis software (Bio-Rad). OCT4B and OCT4B1 expression data were normalized to β2M as an internal control, and relative gene expressions were presented with the 2−ΔΔCt method [32]. For each individual Q-PCR, a final melting curve analysis from 72°C to 95°C was performed to ensure the homogeneity of PCR products.

Specific primers amplifying OCT4B and OCT4B1 isoforms individually were used in Q-PCRs and were as follows: for OCT4B1, OCT-FB and OCT-RB4 (5′-CTTAGAGGGGAGATGCGGTCA-3′); for OCT4B, FB and OCT-RB5 (5′-GGCTGAATACCTTCCCAAATAGA-3′); for β2M amplification, B2M-F and B2M-R.


Cells were grown in a four-well tissue culture plates and fixed in 4% paraformaldehyde for 10 minutes at room temperature. Cells were then permeabilized in 0.2% Triton X-100 (in 0.1% Tween 20-phosphate-buffered saline [PBS]) for 10 minutes and blocked in 3% bovine serum albumin (Sigma-Aldrich) before being incubated in antibody.

Two different antibodies were used: a mouse monoclonal antibody, sc-5279 (Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com), which recognizes amino acids 1–134 of OCT4A protein and therefore recognizes only OCT4A isoform, and an affinity-purified goat polyclonal antibody, sc-8629 (Santa Cruz Biotechnology), which reacts with the C terminus and therefore recognizes both OCT4A and OCT4B isoforms.

Cells were incubated with anti-OCT4 antibodies overnight at 4°C, and dilutions of 1:100 and 1:25 were used for sc-8629 and sc-5279 antibodies, respectively, as described previously [9, 33]. Cells were washed three times in PBS and incubated with fluorescein isothiocyanate (FITC)-conjugated anti-mouse IgG (Invitrogen) or anti-goat IgG (Sigma-Aldrich) at dilutions of 1:100 for 2 hours at 4°C. The cells were then washed three times in PBS, counterstained with Hoechst 33258, and visualized under a UV microscope. For negative control samples, all the conditions were kept the same, except that the primary antibody was omitted.

Fluorescence-Activated Cell Sorting

ES cells were harvested and sorted by fluorescence-activated cell sorting (FACS) as described previously [34]. Briefly, Shef5 cells were dissociated into single cells with trypsin-EDTA solution and resuspended in FACS buffer with 10% fetal calf serum at 5 × 106 cells per milliliter. The cells were incubated for 15 minutes on ice to block nonspecific binding sites and stained with monoclonal antibody MC631, which recognizes anti-stage-specific embryonic antigen-3 (anti-SSEA3), for 30 minutes on ice. Cells then were stained with FITC-conjugated antimouse IgG (Invitrogen) and used for FACS analysis using a MoFlo fluorescent cell sorter (DakoCytomation, Glostrup, Denmark, http://www.dakocytomation.com). The MC631 antibody used was concentrated hybridoma supernatant, pretitered to ensure maximal binding to target cells. The immunofluorescence with SSEA3 antibody was compared with that from the negative control antibody obtained from the parent myeloma cell line P3X63Ag8.

Statistical Analysis

Differences in gene expression between the two groups were evaluated using the Mann-Whitney U test, which was performed by the SPSS 11.0 software package for Windows (SPSS, Chicago, http://www.spss.com). A p value less than .05 was considered statistically significant.


OCT4A and OCT4B Isoforms Have Different Expression Patterns in Different Cell Types

To study the expression pattern of each spliced variant in a panel of cultured cells, we used primers that specifically amplify each isoform. Since most of the transcribed OCT4 pseudogenes have high similarity to the OCT4A sequence, we first used a multiple alignment approach to find nucleotide differences between OCT4A and other pseudogenes, to design specific primers for OCT4A. Accordingly, we found that nucleotides T102 and C288 on exon 1a are unique to the OCT4A transcript, and we used this feature in designing suitable primers. Using these designed primers (AF and RB1) to amplify OCT4 in a panel of different cultured cells, we found that OCT4A expression is confined to undifferentiated ES and EC cells, with no detectable expression in the other nonpluripotent cell types examined, including the gestational choriocarcinoma lines BeWo and JAR and the teratocarcinoma-derived yolk sac carcinoma line 1411H (Fig. 1A). To amplify the OCT4B isoform, we used a primer set located on exons 1b and 2 of OCT4 transcript (FB and RB1). In this case, the expected 228-bp band was detected in RT-PCR amplification of almost all the cells examined (Fig. 1A). DNA sequencing and BLAST analysis confirmed the specificity and accuracy of PCR products.

Figure Figure 1..

Expression of OCT4A and OCT4B in different human pluripotent and nonpluripotent cultured cells. (A): Polymerase chain reaction analysis for OCT4B was performed using primer set FB/RB1 as described in Materials and Methods. β2M was used as an internal control. Note that OCT4A expression was restricted to ES and EC cells, but OCT4B could be detected in almost all cell types tested. (B): Immunostaining analysis of OCT4A and OCT4B isoforms. Immunocytochemistry was performed using a specific monoclonal antibody raised against the OCT4A peptide (sc-5279) and a polyclonal antibody that can recognize both OCT4A and OCT4B isoforms (sc-8629), as described before. Different ES, EC, and somatic cells were used, and the results for NT2 (embryonic carcinoma) and RT4 (bladder cancer) cells are shown. Both antibodies gave a nuclear signal in NT2 cells, which colocalized with the Hoechst 33258 staining (blue signals). In contrast, although no nuclear signal could be detected in RT4 cells, a cytoplasmic signal was observed using the sc-8629 antibody, which recognizes both OCT4A and OCT4B isoforms. No signals were detected in NC. Abbreviations: EC, embryonal carcinoma; ES, embryonic stem; NC, negative controls; NT2, NTERA2 cl.D1.

To further examine the expression and intracellular localization of OCT4A and OCT4B proteins, we used immunocytochemistry by using a specific monoclonal antibody against OCT4A, as well as a polyclonal antibody that could detect both isoforms. Using the anti-OCTA antibody, a nuclear signal was easily detectable in NT2 (ES/EC) cells but not in other nonpluripotent somatic cells, such as RT4 (Fig. 1B). We also confirmed the absence of OCT4A protein by Western blotting in three non-ES lines, RT4, SD, and HT-1376 (data not shown). In contrast, using a polyclonal antibody that can recognize both OCT4 isoforms, we could detect a strong nuclear signal, as well as a very faint cytoplasmic signal, in ES/EC cells, whereas the expression of OCT4 in other cell types was primarily localized to the cytosol (Fig. 1B).

Identification of a Novel OCT4 Spliced Variant, OCT4B1, Retaining Intron 2 as a Cryptic Exon

Using another primer set for OCT4B isoform, in which the forward and reverse primers were located on exons 1b and 3, respectively (FB and RB2), we found that ES and EC cells consistently expressed an ∼500-bp band in RT-PCRs instead of the predicted 267-bp PCR product (Fig. 2). In other nonpluripotent cells, besides the primarily expressed 267-bp band, a faint expression of the new ∼500-bp band could also be detected in a few samples. The fact that prior to RT-PCR, all of the RNA samples used had been DNase-treated to eliminate the risk of DNA contamination, indicated the authenticity of the amplified band as derived from mRNA. This interpretation was confirmed by the absence of the band in No-RT controls, used in parallel to all RT-PCR repeats (Fig. 2A).

Figure Figure 2..

Reverse transcription (RT)-polymerase chain reaction (PCR) analysis and exon organization of OCT4B and OCT4B1 transcripts. (A): RT-PCR was performed using primer set FB and RB2 as described in Materials and Methods. An expected 267-base pair (bp) band, corresponding to OCT4B, was detected in nonpluripotent cells (RT-4 and SD cells). However, embryonic stem and embryonal carcinoma (H7 and NT2) cells primarily expressed a 500-bp PCR product, which corresponds to the novel OCT4B1 spliced variant. No band was detected in No-RT control samples. No-RT controls 1 and 2 were prepared from NT2 and RT-4 RNA samples, respectively. (B): Schematic representation of exon organization and protein domains in OCT4 splice variants. OCT4B1 consists of all five exons of OCT4B, plus a novel exon, exon 2b, flanked by exons 2 and 3. Arrows show PCR primer positions. Primer set FB/RB3 was used for sequence analysis of OCT4B1 spliced variant. Abbreviations: CTD, C-terminal domain; No-RT, no reverse transcription; NT-2, NTERA2 cl.D1; NTD, N-terminal domain; POUH, POU homeodomain; POUS, POU-specific domain.

The 500-bp PCR bands were extracted from the gel, sequenced, and aligned against human genome and transcript sequences. The results showed the existence of a novel OCT4 transcript retaining the whole 225-bp intron 2 sequence (corresponding to nucleotides 4,292–4,518 of the published sequence of OCT4 gene; GenBank Accession Z11900) as a putative novel exon, termed exon 2b (Fig. 2B). We designated this novel variant OCT4B1 (GenBank accession no. EU518650). Acquisition of a new in-frame TGA stop codon within the novel exon 2b results in an open reading frame of 348 nucleotides, predicting a potential truncated 115-amino acid peptide.

The amino acid sequence of OCT4B1 was aligned and compared with that of OCT4B spliced variant. Amino acids 1–80 were identical between OCT4B and OCT4B1 proteins, including a similar N-terminal domain and a part of the POU-specific (POUS) domain, comprising amino acids 43–80. However, OCT4B1 lacks the rest of the POUS domain, as well as the POU homeodomain and the C-terminal transactivation domain (Fig. 2B).

OCT4B1 Is Highly Expressed in Human ES and EC Cells and Downregulated Following Differentiation

To study the expression pattern of OCT4B1 in different cell types, we performed a Q-PCR analysis with primers amplifying OCT4B and OCT4B1 individually. Among the different pluripotent and nonpluripotent cells examined, we found a high level of OCT4B1 expression in undifferentiated ES/EC cells with no or low detectable level in most other cell types. However, occasional non-ES cells (notably SD and HT-1376) did express significant levels of OCT4B1, whereas the undifferentiated ES/EC cells highly expressed OCT4B1 transcript, the expression of OCT4B, if any, was very low. (Fig. 3).

Figure Figure 3..

Quantitative reverse transcription (Q-RT)-polymerase chain reaction (PCR) analysis of OCT4B and OCT4B1 expression in different cell types. Q-RT-PCR was performed using specific primers for OCT4B (white bars) and OCT4B1 (gray bars) in a panel of pluripotent and nonpluripotent cell types. Experiments were performed in triplicate, and values are expressed as means ± SE (normalized to β2M). Note that in contrast to somatic cells, ES and EC cells highly expressed OCT4B1 transcript, with no detectable level of the OCT4B spliced variant. Abbreviations: EC, embryonal carcinoma; ES, embryonic stem; h-MSCs, human mesenchymal stem cells; NT-2, NTERA2 cl.D1.

To evaluate the correlation between OCT4B1 expression and the undifferentiated state of human ES and EC cells, we treated NT2 human EC cells with all-trans-retinoic acid, and we also induced the differentiation of H7 human ES cells in embryoid bodies by culture in suspension in nonadherent dishes. In both experiments, OCT4B1 expression was sharply downregulated during the course of differentiation. The expression of OCT4A was also downregulated, although much more rapidly than OCT4B1 was downregulated (Fig. 4A).

Figure Figure 4..

Expression patterns of OCT4B1 during differentiation and in different subsets of embryonic stem (ES)/embryonal carcinoma (EC) cells. (A): RA treatment and embryoid body formation were used to induce cellular differentiation in NT2 EC and H7 ES cells, respectively. RNA samples were collected in D0–D21 of RA treatment, or D0–D11 for embryoid bodies, and used in semiquantitative reverse transcription (RT)-polymerase chain reaction (PCR) analysis. OCT4B1 expression was readily detected in undifferentiated cells, but it was downregulated in both cell types after differentiation. OCT4A expression was also downregulated, albeit more rapidly, in both experiments. (B): Fluorescence-activated cell sorting was used to isolate the SSEA3(+) (undifferentiated) and SSEA3(−) (spontaneously differentiated) subpopulations of Shef5 ES cells, as described in Materials and Methods. Relative OCT4B1 (normalized to B2M) expression was significantly much higher in SSEA3(+) cells compared with SSEA3(−) cells, as examined by quantitative-RT-PCR analysis (p < .001). Note that only a basal expression level could be detected for OCT4B isoform in the examined cells. Abbreviations: D, day; NT2, NTERA2 cl.D1; RA, retinoic acid; SSEA, stage-specific embryonic antigen.

We also studied OCT4B1 expression in spontaneously differentiating subsets of ES cells defined by expression of the human ES cell marker antigen SSEA3. Two cultures of Shef5 human ES cells were separately labeled and sorted by flow cytometry for expression of SSEA3. Fractions corresponding to the brightest and dimmest cells were collected and processed for quantitative RT-PCR analysis for OCT4B1 expression. A significantly much higher level of OCT4B1 expression was detected in the SSEA3(+) cells compared with the SSEA3(−) cells (p < .05; Fig. 4B).


The expression of OCT4 has been shown to be pivotal for embryonic development and maintaining the pluripotent identity of embryonic stem cells. Several groups, however, have reported the expression of OCT4 in normal and/or tumor somatic cells, suggesting a possible role for OCT4 in regulating the self-renewal of normal/cancer stem cells, thus extending the role of the gene from the embryo to the adult [11, 13] (supplemental online Table 1). In contrast, some recent reports have challenged the expression of OCT4 in somatic cells and have suggested that the reported observations might simply be a misinterpretation caused by the presence of the OCT4 pseudogenes or its other splice variant(s) [13, 33, 35].

Accordingly, there exist several expressed pseudogenes in human genome, with high similarity to the OCT4A sequence [36, 37], which raises the possibility of false-positive results in RT-PCR experiments [35, 37]. Using a primer set specifically designed to avoid the pseudogenes, as well as a specific antibody against OCT4A isoform, our results suggest that OCT4A expression is restricted to ES/EC cells with no detectable expression in other examined somatic cells. This is in accordance with other recent reports of the absence of OCT4A expression in HeLa cells, MCF7 cells, and PBMCs [16, 33]. More experiments, however, are required to address the issue of OCT4A expression in human adult stem cells and somatic tumors. In line with this, Oct4 expression has been reported in murine adult stem cells [38, 39, 40, 41, 42, 4344], and its oncogenic potential has also been shown by ectopic expression in nontumorigenic somatic cells [45, 46]. These findings suggest that Oct4 can potentially be functional in mouse somatic cells. However, the only reported Oct4 isoform in mouse cells is orthologous to the human OCT4A isoform.

Compared with our designed OCT4A primers, recently two other primer sets have also been suggested for specific amplification of this isoform. In one multialignment approach, nucleotides 102, 288, and 407 were described as sites that uniquely define the OCT4A spliced variant and primers incorporating nucleotide 288 were used to amplify this isoform specifically [16]. Using these primers, the absence of OCT4A expression has been reported in mononuclear cells and PBMCs. In another study a primer set corresponding to the beginning of the OCT4A sequence was used for Q-PCR analysis in different germ cell tumors (GCTs) and EC cell lines [47]. In the current study we used nucleotide 102 for designing OCT4A-specific primers and validated these primers in different cell types.

Another potential source of controversy is the presence of OCT4B spliced variant, the involvement of which in stem cell self-renewal still remains unclear [10]. However, in our data, most OCT4 detectable in somatic cells is contributed by OCT4B expression, which showed a clear cytoplasmic staining in immunocytochemistry. This accords with the recent report suggesting that the observed OCT4 expression in peripheral blood mononuclear cells, described by Zangrossi et al. [15], is probably contributed by OCT4B but not OCT4A isoform [48]. Furthermore, we previously found that the presence of OCT4 in some bladder tumor sections is confined to the cytoplasm of cells adjacent to the basal lamina [49], again suggesting a potential involvement of OCT4B isoform, rather than OCT4A variant, in some tumor cells.

Previously, a level of OCT4B expression has been reported in HeLa and MCF-7 cells that is almost similar to that in NT2 embryonal carcinoma cells [33]. Our results, however, demonstrated that the nature of OCT4B transcript in ES and EC cells is different from that of somatic cells, due to the presence of a novel OCT4 spliced variant designated OCT4B1. Notably, whereas ES/EC cells express a very low level of OCT4B isoform, the novel OCT4B1 spliced variant is highly expressed in these cells. This OCT4B1 isoform is expressed predominantly in the SSEA3(+) stem cells and is rapidly downregulated upon differentiation. Taken together, these findings suggest a potential correlation between OCT4B1 and the pluripotent/undifferentiated state of human ES and EC cells.

In immunofluorescence staining, we also observed a low cytoplasmic signal for OCT4 in ES/EC cells that express low levels of OCT4B isoform, whereas a high cytoplasmic signal could also be detected in somatic cells that express a high level of OCT4B variant. Since a cytoplasmic localization has been described previously for OCT4B isoform [9, 10] we concluded that the observed cytoplasmic signal in ES/EC cells is contributed mainly to the OCT4B. However, OCT4B1 spliced variant could potentially encode a short truncated protein, but antibodies that specifically detect this isoform are not currently available to address its subcellular localization.


When these data are taken together, our study demonstrates the importance of specifying the exact variant of OCT4 when addressing the issue of OCT4 expression in different human cells. This would also be pertinent to studies addressing OCT4 function, using experimental approaches such as microarray analysis, RNA interference experiments, or somatic cell reprogramming. The results may also have an impact on the diagnosis of GCTs, in which OCT4 has been proposed as an informative marker for GCTs that exhibit features of pluripotentiality [50, 5152]. We believe that this study provides further insight into the transcriptional regulation of OCT4 expression in different cell states and suggests a potential interconnected role for the spliced variants in regulating the self-renewal of stem cells in embryonic or adult tissues.

Disclosure of Potential Conflicts of Interest

The authors indicate no potential conflicts of interest.


We thank Dr. Mark Jones for help in FACS analysis, Margaret Knowles (Leeds Institute of Molecular Medicine, St. James's University Hospital) for kindly providing the bladder carcinoma cell lines, Harry Moore for providing the Shef5 human ES cells, and Drs. Abbas Basiri and Maryam Matin for encouragement and support. This work was supported by research grants from the Iran National Science Foundation (Grant 85093/27) and the Medical Research Council, and by ESTOOLS, an EU FP6 Project.