Long noncoding RNAs (lncRNAs) have emerged as new regulators of stem cell pluripotency and tumorigenesis. The SOX2 gene, a master regulator of pluripotency, is embedded within the third intron of a lncRNA known as SOX2 overlapping transcript (SOX2OT). SOX2OT has been suspected to participate in regulation of SOX2 expression and/or other related processes; nevertheless, its potential involvement in tumor initiation and/or progression is unclear. Here, we have evaluated a possible correlation between expression patterns of SOX2OT and those of master regulators of pluripotency, SOX2 and OCT4, in esophageal squamous cell carcinoma (ESCC) tissue samples. We have also examined its potential function in the human embryonic carcinoma stem cell line, NTERA2 (NT2), which highly expresses SOX2OT, SOX2, and OCT4. Our data revealed a significant coupregulation of SOX2OT along with SOX2 and OCT4 in tumor samples, compared to the non-tumor tissues obtained from the margin of same tumors. We also identified two novel splice variants of SOX2OT (SOX2OT-S1 and SOX2OT-S2) which coupregulated with SOX2 and OCT4 in ESCCs. Suppressing SOX2OT variants caused a profound alteration in cell cycle distribution, including a 5.9 and 6.9 time increase in sub-G1 phase of cell cycle for SOX2OT-S1 and SOX2OT-S2, respectively. The expression of all variants was significantly diminished, upon the induction of neural differentiation in NT2 cells, suggesting their potential functional links to the undifferentiated state of the cells. Our data suggest a part for SOX2OT spliced variants in tumor initiation and/or progression as well as regulating pluripotent state of stem cells. Stem Cells2014;32:126–134
Long noncoding RNAs (lncRNAs) constitute a large portion of the human transcriptome, and represent a new level of gene expression regulation [1, 2]. LncRNAs are a diverse class of noncoding RNAs with a size of more than 200 nt and an expanding list of biological functions. Among these various functions, lncRNAs are found to be involved in epigenetic modifications of DNA, immune signaling, and regulation of pluripotency and differentiation processes [3-6].
Recent reports of misregulated expression of lncRNAs across many cancer types suggest that their aberrant expression might be a major cause of tumorigenesis, and those different types of cancers can be distinguished according to their altered lncRNAs expression signatures [7-9]. LncRNAs are also introduced as a newly emerging class of oncogene and tumor-suppressor genes .
While little evidence exists on lncRNAs functions, their specific expression pattern, promoter conservation, alternative splicing, and association with particular chromatin signatures all suggest that most of these transcripts are functional [11-14]. Moreover, it is recently claimed that lncRNAs display an enhancer-like function in human cells . They are also implicated in the basal regulation of protein-coding genes, including those central to normal development and tumorigenesis, at both transcriptional and post-transcriptional levels [16, 17]. To date, an increasing number of lncRNAs have been functionally validated to affect different cellular pathways [18-21].
SOX2 is a HMG-box transcription factor, which is crucial for maintaining the pluripotency state and survival of human embryonic stem cells (hESCs). Recent studies have revealed its amplification and activation in several squamous cell carcinomas (SCC) [22-26]. In hESCs, Sox2 regulates the expression of more than 1,000 gene, many of them are coregulated with the other master regulator of pluripotency such as Oct4 . Interestingly, the intronless gene of Sox2 is embedded within the third intron of a lncRNA gene, known as SOX2 overlapping transcript (SOX2OT), which is transcribed in the same orientation as SOX2 . SOX2OT has been suspected to participate in regulation of SOX2 expression and/or related processes, but its potential function has still remained untested .
NTERA2 (NT2), a human embryonal carcinoma (EC) stem cell line, shares many features with hESCs, including similar patterns of gene expression [30-33]. Thus, this cell type is an excellent prototype model of pluripotent stem cells to study gene expression and differentiation associated to human embryonic development . U-87 malignant glioblastoma (U-87 MG) is another cell line that contains cancer stem cell-like properties .
In this study, we have examined a potential correlation between the expression patterns of SOX2OT and its novel splice variants (SOX2OT-S1 and SOX2OT-S2), and those of master regulators of pluripotency (SOX2 and OCT4) in esophageal squamous cell carcinoma (ESCC). We also investigated the function of SOX2OT novel splice variants in NT2 cells, as a prototype model of pluripotent cells.
Materials and Methods
Patients and Clinical Samples
Tissue biopsies were obtained from Iran national tumor bank of Imam-Khomeini cancer institute (Tehran University of Medical Sciences). The samples had been immediately snap-frozen in liquid nitrogen and had been stored in −70°C, until being collected for RNA extraction. The tissues were categorized in two groups: 36 tumor samples from 36 patients with ESCC and 36 paired non-tumor tissues (as control), which were taken from the apparently normal tissues from the margin of same tumors. Histopathological parameters were evaluated according to WHO criteria for grade and TNM system for stage classification (Supporting Information Table S1). The experimental design was approved by the Ethics Committees of Tarbiat Modares University.
Cell Culture and Differentiation
The human NT2 (NTera2; kindly provided by Dr. Peter Andrews at Sheffield University) cell line was propagated in Dulbecco's modified Eagle's medium (DMEM)/F-12 (Invitrogen, Gaithersburg, MD) supplemented with 10% fetal bovine serum (FBS; Invitrogen) and 1% penicillin/streptomycin (Sigma, St. Louis, MO) at 37°C in 5% CO2. NT2 cells were treated with all-trans retinoic acid (ATRA, Sigma-Aldrich) to differentiate into the neuron-like cells, according to a modified protocol provided by Atlasi et al. . Briefly, 2 days before the first ATRA application, cells were seeded at a density of 4 × 104 cells in 2 ml growth medium per well in six-well plates. ATRA was added to the growth medium at a final concentration of 10 µM, and the differentiation medium was renewed twice a week for 4 weeks at regular time intervals. Cells were then harvested for RNA extraction, from three biological replicates at 0, 1, 2, 3, and 4 weeks after ATRA treatment. The control time point (t = 0) expression level was set to 100% and the expression of the treated samples represented as a percentage of the control. The human malignant glioma cell line U-87 MG (derived from a malignant glioma tumor) was obtained from Pasteur Institute (Tehran, Iran) and cultured in RPMI medium (Sigma-Aldrich, Ltd., St. Louis, Mo.) supplemented with 10% FBS and 1% penicillin/streptomycin, and incubated at 37°C under humidified 5% CO2.
Isolation of Nuclear and Cytoplasmic RNAs and cDNA Synthesis
Nuclear and cytoplasmic RNAs were isolated from the human NT2 cell line. Cells were maintained at 37°C in DMEM high glucose containing 10% FBS. After the addition of 1 ml trypsin, cells were collected by centrifugation and washed by phosphate buffered salin (PBS). The cells were then resuspended in 120 µl of LB buffer (10 mM NaCl, 2 mM MgCl, 10 mM Tris-HCl [pH 7.8], 5 mM dithiothreitol (DTT), and 1% Triton X-100). After centrifugation, the supernatant and pellet were partitioned, where supernatant suspected to comprise cytoplasmic RNA and pellet to contain nuclear RNA. The following steps of RNA isolation from both cell lines and patients specimens were carried out using Trizol Reagent (Invitrogen, Carlsbad, CA), according to the manufacturer's instructions. To remove any potential DNA contamination, the extracted RNA was treated with RNase-free DNase (Takara, Japan). The first strand of cDNA synthesized by Reverse Transcriptase (Takara, Japan), using both oligo dT and random hexamer primers (Takara, Japan), according to the manufacturer's instructions. For each sample, a no-reverse transcription (No-RT) control was used in parallel to the DNase-treated RNA, to detect nonspecific amplification of genomic DNA.
RT-PCR and Quantitative Real-Time PCR
The appropriate SOX2OT PCR primers were designed using Genrunner software (version 3.02; Hastings Software, Inc.) and manufactured by MICROGENE Company (Supporting Information Table S2). PCR reactions were performed using 2 µl of cDNA or No-RT samples with 1 U of Taq polymerase (Cinnagen, Iran), 1.5 mM MgCl2, 200 µM dNTPs, and 0.5 µM of each primer in a 20 µl PCR reaction. The PCR amplification was carried out using the following cycling conditions: initiation at 94°C for 5 minutes, amplification for 35 cycles with denaturation at 94°C for 1 minute, annealing at 60°C for 20 seconds, and extending at 72°C for 1 minute and a final extension at 72°C for 10 minutes. Direct DNA sequencing (MICROGENE, South Korea) was performed to confirm the authenticity of the PCR products. Specific real-time PCR primers (Supporting Information Table S2) were designed for SOX2, SOX2OT, SOX2OT-S1, SOX2OT-S2, OCT4, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (GenBank accession numbers: NM_003106.2, NR_004053.2, JN711430.1, JN882275.1, NM_002701.4, NM_002046.3, respectively), using AlleleID 6.0 software (Primer BioSoft, Palo Alto, CA; www.premierbiosoft.com) and Gene Runner software (version 3.02; Hastings Software, Inc.). TaKaRa SYBR Premix Ex TaqII master mix (2×), supplemented with ROX reference Dye II, were used for all real-time PCR reactions. For compensating variations in the amount of input RNA and the efficacy of reverse transcriptase, GAPDH mRNA was also quantified as an internal control, and the expression of other genes was normalized to its expression value. All real-time PCR reactions were carried out with the ABI 7500 real-time PCR systems (Applied Biosystems, Foster City, CA) using the following cycling conditions: initiation at 94°C for 1 minute, amplification for 40 cycles with denaturation at 94°C for 10 seconds, annealing at 62°C for 10 seconds, and extending at 72°C for 40 seconds. The authenticity of the PCR products was further confirmed by melt curve analysis and direct sequencing of the products.
Two pairs of siRNAs targeting exon 1 and exon 5 regions of human SOX2OT were used to suppress SOX2OT expression in NT2 and U-87 MG cells. One pair of scrambled siRNAs was also used as a control. All siRNAs with symmetric 3′-TT overhangs were designed using Invitrogen's Block-It RNAi Designer (https://rnaidesigner.invitrogen.com/rnaiexpress). The sequences of the SOX2OT-specific siRNAs were as follows: 5′-GGAUAGGCCUCACUUACAA-3′ (SOX2OT-siRNA1) and 5′-GGAGAUUGUGACCUGGCUU-3′ (SOX2OT-siRNA2) and sequence of the scrambled siRNA was 5′-GGUAUUCCUCUCGUUUACA-3′. For transient siRNA transfection, NT2 cells were seeded at a concentration of 4 × 104 cells per well in 12-well plates, and U-87 MG cells were seeded at a density of 2.5 × 105 cells per well into six-well culture plates. Both cell lines then incubated for 24 hours in culture medium. The NT2 cells were then transfected with 100 nM and U-87 MG cells with 50 nM SOX2OT-siRNAs or scrambled-siRNA and Lipofectamine RNAiMax (Invitrogen, Carlsbad, CA), according to the manufacturer's instructions. Forty-eight hours after transfection, the cells were harvested for RNA extractions and flow cytometry analysis. shRNAs sequences for silencing SOX2OT-S1 and SOX2OT-S2 were searched and blasted using Invitrogen's Block-It RNAi Designer from the Invitrogen website. The shRNAs targeting SOX2OT variants were designed and chemically synthesized as SOX2OT-S1 shRNA (5′-CCGGCATGGACATATCCAACTTACTCGAGTAAGTTGGATATGTCCATGTTTTTG 3′) and SOX2OT-S2 shRNA (5′-CCGGCTATTCCAATCCAACTTAGCTCGAGCTAAGTTGGATTGGAATAGTTTTTG 3′), and a scrambled/nontarget sequence (CCTAAGGTTAAGTCGCCCTCGCTCGAGCGAGGGCGACTTAACCTTAGG) was also used as a control for knockdown study. For shRNAs transfection, NT2 cells were seeded at a concentration of 4 × 104 cells per well in 12-well plates and incubated for 24 hours in culture medium. The NT2 cells were then transfected with 1 µg pLKO.1 vector (Addgene; Cambridge, MA) expressing SOX2OT-S1, SOX2OT-S2, or scrambled/nontarget shRNA, and FuGENE HD (Roche, UK) transfection reagent, according to the manufacturer's instructions. Forty-eight hours after transfection, the cells were harvested for RNA extractions and flow cytometry analysis.
Flow Cytometry and Cell Cycle Analysis
For cell cycle analysis, NT2 cells were collected 48 hours after transfection with SOX2OT variants or scrambled siRNAs. Cells then washed in PBS and trypsinized with 0.025% trypsin-EDTA to yield single-cell suspensions. Then cells were fixed in ice-cold 70% ethanol and stained with 50 µg/ml propidium iodide solution containing 0.1% Triton X-100 and 10 µg/ml RNaseA (Takara, Japan). The single-cell suspensions were then used for flow cytometric analysis, using a FACSCanto II system (Becton Dickinson Bioscience). Experiments were repeated at least twice and cell cycle profiles were analyzed using Flowing software.
RNA Stability Assay
NT2 cells were grown to ∼50% of confluence, before addition of 10 mg/ml actinomycin D (Sigma, St. Louis, Missouri) to block RNA polymerase activity. RNAs were extracted from two biological replicates at 0, 0.5, 1.5, 3, 6, and 12 hours after treatment. SOX2OT, SOX2OT-S1, and SOX2OT-S2 expression levels were determined from RNA extracted at different time points after treatment, and quantified relative to GAPDH expression level.
All real-time PCR experiments were carried out in duplicates. For calculation of expression fold change of the examined genes, the expression level of them in each sample was normalized to that of GAPDH, as an internal control. Then, the expression of candidate genes in tumor samples was normalized to their matched non-tumor samples (2−ΔΔCT). All experimental tests were repeated at least three times. The data were expressed as means ± SD. Statistical analysis was performed using Student's t test (two-tailed). The criterion for statistical significance was p < .01. Receiver operating characteristic (ROC) curve analysis was plotted to determine whether the expression of the examined genes have the sensitivity and specificity to discriminate between tumor and non-tumor specimens as well as between different grades of tumor samples. The correlation coefficients of all genes were also analyzed by using the Spearman's rho test.
The LncRNA SOX2OT is Coupregulated with Master Regulators of Pluripotency, SOX2 and OCT4, in ESCC
We used a real-time reverse transcriptase PCR (qRT-PCR) approach to examine the possibility of SOX2OT, SOX2, and OCT4 expression in tumor and non-tumor samples of ESCC. Accordingly, specific PCR primers were designed to amplify each transcript in biopsy samples of ESCCs. Our data revealed a high level of SOX2OT expression in tumor samples of ESCC, compared to that of apparently normal marginal tissues obtained from the same patients (p < .01, Fig. 1A). The consistent and high expression level of SOX2OT in tumor samples of ESCC suggested a potential role for this transcript in tumorigenesis of esophagus. The authenticity of the PCR products was further confirmed by direct sequencing of the products.
The suitability of expression level of examined genes to discriminate between tumor and non-tumor states of the ESCC samples was further analyzed by ROC analysis. According to ROC curve analysis, SOX2OT (AUC = 0.84, p = .008; Supporting Information Fig. S1A), SOX2 (AUC = 0.75, p = .015; Supporting Information Fig. S1B), and OCT4A (AUC = 0.70, p = .05 Fig. 1C) seem to be good candidates for discriminating tumor from non-tumor samples. An AUC>70 indicates a good ability of a marker to discriminate two groups of samples. On contrary, ROC curve analysis failed to discriminate high-grade samples from low-grade ones. We further examined the possibility of any correlation between the expression patterns of main promoters of reprogramming (OCT4 and SOX2) and SOX2OT in ESCC samples. According to Spearman's rho test, SOX2OT expression level had a strong correlation to that of SOX2 (83%, p = .025) and a relatively high correlation to that of OCT4A (61%, p = .041).
SOX2OT Revealed a Close Expression Pattern with SOX2 and OCT4 in NT2 Cells
We used a qRT-PCR approach to examine the relationship between expression patterns of human SOX2OT and two master regulators of pluripotency, OCT4 and SOX2, in a well-characterized human EC cell line, NT2. The expression level of studied gene was quantified relative to that of GAPDH, as a house-keeping internal control. SOX2OT demonstrated a very close expression pattern with its intronic coding gene SOX2, as well as the other important master pluripotency gene, OCT4, in these cells (Fig. 1B). It seems that the expression level of SOX2OT was higher than that of SOX2 in NT2 cells.
Suppressing SOX2OT Diminished SOX2 and OCT4 Expression in NT2 Cells
Considering the close expression pattern of SOX2OT and those of master regulators of pluripotency, SOX2 and OCT4, we hypothesized that SOX2OT might have a role in regulation of pluripotency pathway in human stem cells. To examine the potential function of SOX2OT in human pluripotent cells, we designed two pairs of specific siRNAs that target exon 1 (SOX2OT-siRNA1) and exon 5 (SOX2OT-siRNA2) regions of SOX2OT. As expected, transfection of NT2 and U-87 MG cells with SOX2OT specific siRNAs efficiently reduced RNA level of SOX2OT. Moreover, qRT-PCR revealed that silencing of SOX2OT resulted in a significant reduction of SOX2 and OCT4 RNA levels in NT2 cells. Compared to the scrambled siRNA, SOX2OT-siRNA2 caused more than 90% reduction in the expression level of SOX2OT, SOX2, and OCT4A in NT2 cells (p < .01; Fig. 1C). Moreover, suppression of SOX2OT caused a significant decrease in SOX2 expression in U-87 MG cells (p < .01; Fig. 1D).
We next examined the intracellular localization of SOX2OT, by means of RT-PCR analysis of cytoplasmic and nuclear RNAs extracted from NT2 cells. The result revealed a nuclear localization for SOX2OT, similar to that of U6 RNA which was used as a control for nuclear RNAs.
Identification of Two Novel Splice Variants of Human SOX2OT (SOX2OT-S1 and SOX2OT-S2) that are Upregulated in ESCC
We used specific primer set 1 (Supporting Information Table S2) to amplify SOX2OT in ESCC samples, where the forward and reverse primers were located on exons 3 and 5 of the gene, respectively (Fig. 2A). Unexpectedly, the melt curve analysis of the real-time PCR products demonstrated the existence of two distinct melt curves for the amplified products, suggesting the amplification of two different fragments. Running the PCR products on an agarose gel, we found that tumor samples of ESCC and some paired non-tumor tissues consistently produced an additional 93 bp fragment in addition to the predicted 201 bp PCR product (Fig. 2B). Direct sequencing of the amplified products further confirmed the identity of the 93 bp band as an alternatively spliced form of SOX2OT, lacking exon 4 of the main transcript (SOX2OT-S1; Accession no: JN711430, GI:379031002).
Finding the aforementioned variant encouraged us to search for other potential variants of SOX2OT. Using another specific set of primers (set 2), in which the forward and reverse primers were located on the first and last exons (exon 5) of the gene (Fig. 3A, 3B), another splice variant of SOX2OT lacking exons 3 and 4 was identified by direct sequencing of the PCR products (SOX2OT-S2; Accession no: JN882275, GI:379031003).
To examine the expression pattern of the newly discovered variants in ESCC samples, specific primer pairs for SOX2OT splice variants on regions that are distinct for each transcript were used. The results of qRT-PCR revealed that SOX2OT-S1 and SOX2OT-S2 were overexpressed in tumor samples of ESCC, compared to their marginal non-tumor tissue counterparts. Furthermore, overexpression of SOX2OT-S2 in tumor samples, similar to SOX2 and OCT4, was statistically significant (p < .01; Fig. 3C). According to the ROC curve analysis, SOX2OT-S1 (AUC = 0.71, p = .0095; Supporting Information Fig. S1D) and SOX2OT-S2 (AUC = 0.735, p = .005; Supporting Information Fig. S1E) seem to be good candidates for discriminating tumor from non-tumor samples. According to the Spearman's rho test, SOX2OT-S1 showed a high correlation to SOX2 (72%, p = .01) and a weak correlation to OCT4A (37%, p = .05). Similarly, SOX2OT-S2 showed a relatively high correlation to SOX2 (65%, p = .015) and OCT4A (59%, p = .022).
We also examined the expression patterns of the novel variants of SOX2OT (normalized to that of internal control, GAPDH) in NT2 cells by qRT-PCR. SOX2OT-S1 had the least expression level, compared to SOX2OT and SOX2OT-S2 splice variants in the cells.
SOX2OT-S2 Is the Most Stable Transcript Among SOX2OT Splice Variants
As mentioned above, the SOX2OT-S1 and SOX2OT-S2 variants are expressed in NT2 cells, albeit at lower levels compared to those of OCT4 and SOX2. To examine whether splicing influence the stability of SOX2OT transcripts, we treated NT2 cells with a general transcription inhibitor agent, actinomycin D, and quantified the expression levels of the variants in the treated cells after 0.5, 1.5, 3, 6, and 12 hours. The obtained data revealed that SOX2OT-S2 is the most stable variant, compared to SOX2OT and SOX2-S1 variants (p < .01; Fig. 4A).
Suppressing SOX2OT-S1 and SOX2OT-S2 Caused a Profound Alteration in Cell Cycle Distribution of NT2 Cells
Following the identification of novel splice variants of SOX2OT, and the fact that these variants are stable and overexpressed in ESCC, we explored their function in NT2 cells. For this purpose, we knocked down SOX2OT-S1 and SOX2OT-S2 expression using specific shRNAs and then analyzed the gene expression and cell cycle alterations in NT2 cells. Using distinct shRNAs specifically targeting SOX2OT-S1 and SOX2OT-S2, along with a scrambled/nontarget shRNA as a negative control, we demonstrated a noticeable decrease in SOX2OT-S1 and SOX2OT-S2 expression after 48 hours in NT2 cells. Following SOX2OT-S1 suppression, we observed a significant decrease in SOX2 expression in NT2 cells (p < .01), but we failed to observe any significant difference in OCT4A expression (Fig. 4B). Conversely, SOX2OT-S2 knockdown revealed no significant difference in SOX2 and OCT4A expression.
We further tested whether suppression of SOX2OT-S1 and SOX2OT-S2 can cause any alteration on cell cycle distribution. Compared to the control cells transfected with scrambled/nontarget shRNA, flow cytometry analysis revealed dramatic changes in distribution of cells in G1 and sub-G1 phases 48 hours after SOX2OT-S1 and SOX2OT-S2 knockdown, respectively (p < .01; Fig. 4C).
SOX2OT Variants Are Downregulated Upon the Induction of Neural Differentiation in NT2 Cell Line
NT2 cells were treated with ATRA to induce their differentiation toward neural-like phenotype, and then the cells were collected at different time points. As shown in Figure 5B, SOX2OT variants demonstrated distinct expression pattern during the course of neural differentiation of NT2 cells. While the expression pattern of SOX2OT variants was similar to those of OCT4 and SOX2 during early time points, low expression of SOX2OT variants persisted in later time points (p < .01; Fig. 5).
Some lncRNAs have been reported to be key regulators of stem cell pluripotency and tumorigenesis. Accordingly, expression patterns of several lncRNAs correlate with those of pluripotency regulators such as OCT4, Nanog, and SOX2 [37, 38]. SOX2 is an intronless gene that lies within the third intron of another gene known as SOX2OT . SOX2OT has been proposed to function in the transcriptional regulation of SOX2 . However, its exact biological function has yet remained elusive.
A common strategy that might provide insight into specific lncRNA functions is to analyze its expression in comparison to its adjacent protein-coding genes. Furthermore, little is known about the involvement of lncRNAs in the tumorigenesis of ESCCs. Therefore, in this work, we examined the expression patterns of SOX2OT and its intronic protein-coding gene, SOX2, along with the expression pattern of SOX2 partner, OCT4, in ESCC samples. Our data revealed that SOX2OT, SOX2, and OCT4 are coupregulated in tumor samples of esophagus, in comparison to their apparently normal marginal tissues. Interestingly, the upregulation of SOX2OT was more than those of SOX2 and OCT4 genes, suggesting its better suitability as a potential tumor marker for ESCC. Accordingly, the data obtained from ROC analysis suggested that the expression levels of SOX2OT, SOX2, and OCT4A have a high sensitivity and specificity in discriminating between tumor and non-tumor states of esophagus samples. However, further works are needed to evaluate the expression of SOX2OT, as a potential general tumor marker, in other types of cancers. The upregulation of these markers also needs to be examined in a larger sample size of tumor and non-tumor specimens, to calculate its better sensitivity and specificity to discriminate tumors versus non-tumors as well as to differentiate different grades of malignancy from each other.
Recently, in order to identify the biological roles of sox2ot, Amaral et al.  analyzed the expression pattern of this lncRNA in several developmental systems, in comparison to the expression of sox2. These analyses revealed that sox2ot, similar to sox2, is expressed in mouse ESCs. They also claimed that sox2ot is a stable transcript which is dynamically regulated, and has conserve functions during the vertebrate development. Based on our data on close expression patterns of SOX2OT transcripts with SOX2 and OCT4, we proposed that these genes could be regulated in the same manner in pluripotent cells such as NT2 cells. Our finding on coexpression of SOX2OT and SOX2 in NT2 cells is in agreement with the hypothesis proposed by Amaral et al. on the possible role of SOX2OT in regulating SOX2 expression. Interestingly, specific siRNAs directed against SOX2OT efficiently caused a diminution in the expression of SOX2 and OCT4 as well as a subG1/G1 cell cycle arrest and prohibition of S-phase entry in NT2 cells. These findings are in agreement with two independent studies which reported that suppression of SOX2 using siRNA could attenuate S-phase entry and accumulation of human glioma cells in the G0/G1 cell cycle phase [39, 40]. Considering our finding on nuclear localization of SOX2OT, it seems that this lncRNA is acting at transcriptional rather than post-transcriptional level to regulate SOX2 expression. Moreover, arrest in subG1/G1 phase of cell cycle and prohibition of S-phase entry by specific siRNA against SOX2OT in NT2 cells suggest a part for SOX2OT in self-renewal of pluripotent cells. SOX2OT knockdown also caused a significant reduction in OCT4 expression. This finding is in agreement with a previous report by Masui et al.  who indicated that Sox2 has essential role in maintaining the pluripotency state of the mouse ESCs by regulating the level of Oct4 expression. Accordingly, the reduction of OCT4 expression might be a consequence of SOX2 downregulation, thus SOX2OT might play an important role in regulation of master regulators of stem cell pluripotency and self-renewal.
Unexpectedly, during amplification of SOX2OT with different set of primers, we identified two alternative splice variants of the main transcript which were both upregulated in ESCC samples. A close correlation between the expression patterns of SOX2OT transcripts and those of master regulators of pluripotency (SOX2 and OCT4) in ESCC samples suggests a similar function for them. Compared to SOX2OT-S1, the upregulation of SOX2OT-S2 was more significant in tumor samples. ROC analysis also confirmed a high sensitivity and specificity for SOX2OT variants as potential tumor markers in detecting tumor states of esophagus samples. Furthermore, the fact that SOX2OT-S2 is more stable than the other variants suggests a potential causative role for this variant in tumorigenesis of esophagus. A homology search using the BLASTN algorithm on the National Center for Biotechnology Information web server against the human EST database (dbEST) showed multiple partially sequenced EST clones, with GenBank accession numbers BX423294.2, BX442540.2, BX459910.2, DA268964.1, and DA282731.1 which correspond to the sequence of exon 3-exon 5 junction in SOX2OT-S1. Another partially sequenced EST clone, with GenBank accession number DA308672.1, corresponds to the sequence of exon 2-exon 5 junction in SOX2OT-S2. These findings further confirmed our data on alternative splicing of SOX2OT to yield SOX2OT-S1 and SOX2OT-S2 variants. By specific knockdown of SOX2OT variants, we observed a significant decrease in SOX2 expression in NT2 cells treated with SOX2OT-S1 shRNA expression. While there was also a downregulation in SOX2 expression in the cells treated with SOX2OT-S2 shRNA, this alteration was not statistically significant. The latter finding could be a result of innate differences between the function of these variants or simply could be due to the observed weaker ability of SOX2OT-S2 shRNA to suppress SOX2OT-S2.
LncRNAs have also been proposed to be important regulators of neurogenesis. However, little is known about their functional involvement in human neural differentiation [37, 42]. We also examined the expression patterns of SOX2OT and its two novel variants during the course of neural differentiation of NT2 cells. SOX2OT transcripts showed a distinct expression pattern upon the induction of neural differentiation of NT2 cells, similar to that of SOX2 and OCT4. However, in contrast to OCT4 and SOX2, a low expression of SOX2OT transcripts persists in later time points, suggesting its weaker correlation to the state of pluripotency. Nevertheless, the close correlation between the expression patterns of SOX2OT transcripts and the key pluripotency genes suggests that they may be regulated coordinately.
In conclusion, our data provide for the first time an insight into the potential function of the human SOX2OT in regulating SOX2 expression in NT2 pluripotent cells. It also highlights a coupregulation of SOX2OT and its novel splice variants, SOX2OT-S1 and SOX2OT-S2, with key stem cell pluripotency genes, SOX2 and OCT4, in ESCCs. High expression and close correlation in expression patterns with SOX2 and OCT4 advocate that SOX2OT transcripts might play a part in tumorigenesis of ESCC. All together, our data provide a novel regulatory mechanism governing the key stem cell pluripotency genes, SOX2 and OCT4, mediated by the lncRNA SOX2OT.
We are grateful to Dr. Fereydouni, the head, and Drs. Kamali and Hosseini the members of National Tumor Bank of Imam-Khomeini Cancer Institute (Tehran University of Medical Sciences) for their valuable help in providing frozen clinical samples. We also thank MeysamYousefi and Mahboobeh Mohseni for their kind help in experiment data and flow cytometry data analysis. This work was supported partly by a research grant from the research deputies of Tarbiat Modares University and Golestan University of Medical Sciences.
A.S.: conception and design, collection and/or assembly of data, data analysis and interpretation, and manuscript writing; M.R.R.: conception and design, assembly of data, data analysis, and manuscript writing; Y.F.: conception and design and manuscript writing; N.A.O.: conception and design; N.M.S.: provision of study material or patients and manuscript writing; M.S.: manuscript writing; S.S.: manuscript writing and provision of study material or patients; M.V.: provision of study material or patients; S.J.M.: conception and design, administrative support, data analysis and interpretation, manuscript writing, and final approval of manuscript.
Disclosure of Potential Conflicts of Interest
The authors indicate no potential conflicts of interest.