This article is mentioned in:

Introduction

A glioma is one of the most common intracranial tumors which is characterized by invasive growth, high risk of relapse and cellular atypia (1). Current treatments which aim to improve the survival of patients with glioma include surgery, radiotherapy and chemotherapy (2). However, all of these treatments have not yet achieved significant outcomes. The prognosis of gliomas is far from satisfactory as it has a median survival time of only 14.6 months and a 5-year survival rate <9.8% (3). With the current developments in immunohistochemistry, molecular biology, genetic testing and treatment, many cytokines involved with the genesis and development of glioma have been identified, which has provided a new direction in the diagnosis and treatment of gliomas (4).

Histone methylation is an important post-translational modification which affects many biochemical processes including chromatin formation, translational regulation and DNA damage repair (5). It has been confirmed that the methylation of multiple histone H3 lysine residues may occur. Among them, the methylation of H3K4, H3K36, H3K79 is associated with transcriptional activation, and the methylation of H3K9 and H3K27 plays a role transcriptional repression (6). To date, dozens of histone demethylases have been identified (6). These known histone demethylases are divided into two categories: one represented by lysine-specific histone demethylase 1 (LSD1) which requires flavin adenine dinucleotide as cofactor in the event of a catalytic reaction; the other comprises histone demethylases containing a Jmjc domain which require Fe(II) and α-ketoglutarate in the demethylation process (6).

Jumonji AT-rich interactive domain 1B (JARID1B) belongs to the second category and it specifically removes the trimethyl modification of H3K4 in order to inhibit the transcription of the corresponding genes (7,8). Previous studies have suggested that the JARID1B protein plays a vital role in the development of breast cancer and prostate cancer, and it is therefore considered to be an important drug target protein (8–11). Although JARID1B expression has been studied in some types of cancer, little is known about the expression of JARID1B in glioma and its function in tumorigenesis.

In the present study, we examined the roles of JARID1B in the genesis of glioma. We demonstrated that the expression of JARID1B was elevated in the majority of the human glioma tissues examined. The overexpression of JARID1B significantly enhanced cell proliferation, migration and sphere formation in vitro. JARID1B overexpression also increased the tumorigenicity of cells in a nude mouse xenograft model of glioma. Moreover, JARID1B increased the protein levels of phosphorylated (p-)Smad2. It also increased the expression of octamer-binding transcription factor 4 (Oct4); nestin; BMI1 proto-oncogene, polycomb ring finger (Bmi-1) and CD133. Conversely, JARID1B knockdown by short hairpin RNA (shRNA) inhibited the proliferation and migration of glioma cells as well as sphere formation. Moreover, the activation of p-Smad2 contributes to JARID1B-induced oncogenic activities. These findings suggest that high JARID1B expression is involved in the pathogenesis of glioma.

Materials and methods

Samples, cells and antibodies

All experiments described in the present study were performed after obtaining informed consent and were in accordance with an Institutional Review Board protocol approved by the Partners Human Research Committee at Dalian Medical University (Dalian, China). The human normal brain tissue samples and glioma tissue samples were provided by the Department of Neurosurgery at the Second Affiliated Hospital of Dalian Medical University (Dalian, China) between 2010 and 2012. The human glioma cell lines, U87, T98G, SW1088 and SW1783, were obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA). The human glioma cell lines, U251 (#09063001) and U373 (#08061901) were obtained from Sigma (St. Louis, MO, USA). The cells were maintained in Minimum Essential Medium (MEM) supplemented with 10% fetal bovine serum (FBS) (both from Invitrogen, Carlsbad, CA, USA). Mouse monoclonal JARID1B (ab198884), p-Smad2 (ab53100), Smad2 (ab119907), transforming growth factor-β1 (TGF-β1; ab92486), Oct4 (ab181557), nestin (ab11306) and Bmi-1 (ab85688) antibodies were purchased from Abcam (Cambridge, MA, USA). Mouse monoclonal β-actin antibody was obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). LY364947, an inhibitor of TGF-βRI, was obtained from Selleckchem (Houston, TX, USA).

Survival analysis

In order to determine the association between JARID1B and patient survival, Kaplan-Meier survival analysis was performed. All the carcinoma tissues which were used for immunohistochemical analysis were randomly collected from glioma tissue samples. Follow-up data were summarized at the end of March 2014, with a median observation time of 759 days.

Reverse transcription-polymerase chain reaction (RT-PCR)

Total RNA was extracted from the glioma tissues and the cell lines using TRIzol reagent (Invitrogen). Reverse transcription was performed using the ThermoScript RT-PCR system (Invitrogen). HotStart PCR conditions were set as follows: 45 sec at 94°C, 30 sec at 55°C, and 1 min at 72°C for 28–30 cycles (for JARID1B) or 26 cycles [for glyceraldehyde 3-phosphate dehydrogenase (GAPDH)]. The following primers were used: JARID1B sense, 5′-AGAGTTTCATCCCCAAAGACAA-3′ and antisense, 5′-AGTTCAGGCAGTAGGCAAAGTC-3′; GAPDH sense, 5′-TGCCTCCTGCACCACCAACT-3′ and antisense, 5′-CCC GTTCAGCTCAGGGATGA-3′.

Western blot analysis

The samples and the cells were solubilized in radioimmunoprecipitation assay lysis buffer [50 mmol/l Tris-HCl (pH 7.4), 1% NP-40, 0.25%Na-deoxycholate, 150 mmol/l NaCl, 1 mmol/l EDTA, 1 mmol/l phenylmethylsulfonyl fluoride, 1 mg/ml each of aprotinin, leupeptin and pepstatin, 1 mmol/l Na3VO4, 1 mmol/l NaF]. The supernatants, which contained the whole-cell protein extracts, were obtained following centrifugation of the cell lysates at 10,000 × g for 10 min at 4°C. The protein samples (20 µg) were loaded on a sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel (5% stacking gel and 12% separating gel). The proteins were then transferred to polyvinylidene difluoride membranes (Millipore, Bedford, MA, USA). The membranes were first probed with a primary antibody and then with a secondary antibody. The bound antibody was detected by enhanced chemiluminescence detection reagents (Amersham Bioscience, Piscataway, NJ, USA) according to the manufacturer's instructions. The band intensity was quantified with the use of ImageQuant software (Molecular Dynamics, Sunnyvale, CA, USA).

Plasmid construction

JARID1B human cDNA was cloned as previously reported (12). The full-length cDNAs were subcloned into the multiple cloning sites of the pBabe plasmid, to form the pBabe-JARID1B expression plasmids. shRNA targeting JARID1B (sense, 5′-TCGACGCGAATCTCTCTTTGGCAAGTTCAAGAGACTTGCCAAAGAGAGATTCGTTTTTTGGAAT-3′ and antisense, 5′-CTAGATTCCAAAAAACGAATCTCTCTTTGGCAAGTCTCTTGAACTTGCCAAAGAGAGATTCGCG-3′) was initially inserted into the SalI and XbaI sites of pSuper plasmid, forming the pSuper-shJARID1B plasmids. The pBabe and pSper plasmids were obtained from Addgene (Cambridge, MA, USA).

Generation of stable cell lines

The U251 cells were transfected with the pBabe or pBabe-JARID1B plasmid using the Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. The SW1783 cells were transfected with the pSuper or pSuper-shJARID1B plasmid using Lipofectamine 2000. Stable transfectants were obtained after selection with puromycin (10 µg/ml; Invitrogen) for 2 weeks. The mRNA and protein expression of JARID1B in the stable cell lines were analyzed by RT-PCR and western blot analysis, respectively.

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay

An MTT assay was performed in order to determine the effect of overexpression or knockdown of JARID1B on the proliferation of glioma cells. Five thousand cells were seeded into a 96-well plate. The cells were cultured in 100 µl growth medium. At various time-points, 20 µl of sterile MTT dye (5 mg/ml; Sigma) was added, followed by incubation at 37°C for 4 h. The MTT solution was then replaced with dimethyl sulfoxide (DMSO) (200 µl) and thoroughly mixed for 30 min. The absorbance at 570 nm was measured using a microplate reader (Spectra Max 340; Molecular Devices, Sunnyvale, CA, USA) with background subtraction at 660 nm.

Wound healing assay

Equal numbers of different cells were seeded into 6-well tissue culture plates. When 90% confluence was reached, a single wound was created by gently removing the attached cells using a sterile plastic pipette tip. Debris was removed by washing the cells with serum-free medium. The migration of the cells into the wounded area was observed and noted at different time-points. Migration and extended protrusion of cells from the border of the wound were visualized and images were captured using an inverted microscope (Olympus DP27).

Assays of cell invasion and migration

The invasion of cells was measured in Transwell inserts (6.5 mm; Costar, Manassas, VA, USA) coated with Matrigel (BD Biosciences, Franklin Lakes, NJ, USA) containing polycarbonate filters with 8-µm pores as described previously (13–15). The inserts were coated with 50 µl of 1 mg/ml Matrigel matrix according to the manufacturer's instructions. The cells (2×105) in 200 µl of serum-free medium were plated in the upper chamber whereas 600 µl of medium with 10% FBS was added to the lower well. After a 24 h incubation, the top cells were removed and the number of cells on the bottom surface of the Transwell insert was counted. The cells that had migrated to the lower surface of the membrane were fixed in 4% paraformaldehyde and stained with 0.5% crystal violet (Sigma). For each membrane, five random fields were counted at x10 magnification. The mean was calculated and the data are presented as the means ± SD from three independent experiments performed in triplicate. The migration assay was similar to the Matrigel invasion assay except that the Transwell insert was not coated with Matrigel.

Sphere formation assay

Mammosphere culture was performed as previously described by Sun et al with slight modifications (14). Briefly, single-cell suspensions were plated in ultralow attachment 96-well plates (Costar) at different densities of viable cells. The cells were grown in a serum-free mammary epithelial growth medium (MEGM), supplemented with 1:50 B27 (Invitrogen), 20 ng/ml epidermal growth factor (EGF), 20 ng/ml basic fibroblast growth factor (bFGF) (BD Biosciences) and 10 µg/ml heparin (Sigma). The numbers of spheroids were counted after 7–10 days.

Immunohistochemical staining

Paraffin-embedded sections were deparaffinized, blocked, and incubated with 1:200 anti-JARID1B antibody at 4°C overnight. Horseradish peroxidase-conjugated secondary antibody (1:500) was then added and further incubated for 1 h at room temperature. The sections were developed using a 3,3′-diaminobenzidine tetrahydrochloride (DAB) substrate kit (Thermo Fisher Scientific, Waltham, MA, USA) at room temperature for 1–5 min and then counterstained with hematoxylin (Sigma).

In vivo tumor formation assay

U251-pBabe and U251-pBabe-JARID1B cells at their exponential growth phase were harvested and washed twice in phosphate-buffered saline (PBS). The cells were resuspended in PBS at a density of 5×107 cells/ml. The cell suspension (0.1 ml; 5×106 cells) was subcutaneously injected into the right flank of 5- to 6-week-old female BALB/c-nu/nu mice (n=3/group). The mice were purchased from the Shanghai Slac Laboratory Animal Co. Ltd and maintained in microisolator cages. Tumor volume (mm3) was measured every 7 days in two dimensions using a caliper and was calculated as 0.4 × (short length)2 × long length. The mice were humanely euthanized by carbon dioxide when the subcutaneous tumors reached 15 mm in diameter. After euthanasia, the tumors were removed from the mice and weighed. This experiment was conducted according to international Guidelines for the Care and Use of Laboratory Animals, and was approved by the Institutional Animal Care and Use Committee of Dalian Medical University (Dalian, China) (approval no. DMU201507).

Statistical analysis

Experimental data are presented as the means ± standard deviation (SD). The results from different treatment groups were compared using a two-tailed Student's t-test. P<0.05 was considered to indicate a statistically significant difference. Statistical analysis was performed using SPSSn Win11.0 software (SPSS, Inc., Chicago, IL, USA).

Results

JARID1B is overexpressed in human glioma samples

Firstly, we detected the mRNA and protein levels of JARID1B in the seven normal brain samples and 13 glioma samples. The expression of JARID1B in normal brain samples, seven low-grade glioma samples (grades I–III) and six high-grade glioma tissue samples (grade IV) was analyzed by RT-PCR (Fig. 1) and western blot analysis (Fig. 2), respectively. Compared with the normal brain tissues, we found that the mRNA and protein expression of JARID1B was upregulated in seven out of seven (100%) low-grade samples and six out of six (100%) high-grade glioma samples. Furthermore, we also found that the expression level of JARID1B was significantly higher in high-grade primary gliomas than in low-grade gliomas (Figs. 1 and 2).

Figure 1 Elevated mRNA expression of jumonji AT-rich interactive domain 1B (JARID1B) in gliomas tissues. (A) JARID1B expression were measured by RT-PCR in thirteen gliomas tissues (T1-T13) and seven normal tissues (N1–N7). (B) JARID1B mRNA expression is significantly elevated in gliomas tissues in comparison with that in normal tissues, and JARID1B mRNA expression was significantly higher in high-grade gliomas than in low-grade gliomas. **P<0.01 vs. normal brain tissue and ##P<0.01 vs. low-grade glioma samples. (based on the Student's t-test). All results are representative of three independent experiments.

Figure 2 Elevated protein expression of jumonji AT-rich interactive domain 1B (JARID1B) in gliomas tissues. (A) JARID1B protein levels were measured by western blot analysis in gliomas tissues and normal tissues. (B) JARID1B is significantly elevated in gliomas tissues in comparison with that in normal tissues, and the protein expression level of JARID1B was significantly higher in high-grade gliomas than in low-grade gliomas. **P<0.01 vs. normal brain tissue and ##P<0.01 vs. low-grade glioma samples. (based on the Student's t-test). All results are representative of three independent experiments.

Secondly, we performed immunohistochemical staining in order to analyze the expression of JARID1B in more samples (low-grade glioma, 10; high-grade glioma, 30). The results showed that JARID1B protein was expressed in all human glioma samples. The protein was localized in the nuclei of the tumor cells (Fig. 3A). The nuclei of the majority of the high-grade glioma cells (Fig. 3A) were more intensely stained compared with those of the low-grade tumors (Fig. 3A). The ratio of JARID1B-positive cells in the high-grade glioma samples was higher than that in the low-grade glioma samples (Fig. 3B). Survival analysis revealed significantly better survival for glioma patients with low expression of JARID1B (<50%) than for glioma patients with high expression of JARID1B (>50%) (Fig. 3C). These results suggest that JARID1B expression in gliomas is not simply a reflection of the expression pattern of the tumor cells of origin nor a consequence of therapeutic agents but rather represents an aberrant gene activation event associated with tumor progression.

Figure 3 Immunohistochemical analysis reveals that Jumonji AT-rich interactive domain 1B (JARID1B) expression is elevated in high grade glioma tissues. (A) The expression of JARID1B was measured by immunohistochemistry in glioma tissues. (B) JARID1B expression is significantly elevated in high grade gliomas tissues in comparison with that in low grade gliomas tissues. (C) Elevated expression of JARID1B is associated with poor patient survival in individuals with gliomas. **P<0.01 is based on the Student's t-test; **P<0.01 vs. low-grade gliomas. All results are representative of three independent experiments. Scale bar, 200 nm.

JARID1B facilitates the proliferation of glioma cells

We detected the mRNA and protein expression levels of JARID1B in the six glioma cell lines by RT-PCR and western blot analysis, respectively (Fig. 4A–C). The results revealed high levels of JARID1B expression in four grade IV glioblastoma cell lines (T98G, SW1783, U87 and SW1088). By contrast, the other two low-grade glioma cell lines (U373 and U251) expressed low levels of JARID1B.

Figure 4 Expression profile of jumonji AT-rich interactive domain 1B (JARID1B) in glioma cell lines. (A) JARID1B mRNA levels were determined by RT-PCR in the glioma cell lines. (B) JARID1B protein levels were measured by western blot analysis in gliomas cell lines. (C) Relative JARID1B protein expression was measured relative to β-actin in the glioma cell lines. (D) Establishment of JARID1B overexpression in U251 cells. The results were analyzed by western blot analysis. (E) Proliferation of U251-pBabe-JARID1B cells is significantly accelerated compared to normal U251 control cells (mock) measured by MTT assay. **P<0.01 is based on Student's t-test; **P<0.01 vs. mock cells. All results are representative of three independent experiments.

To demonstrate the functional role of JARID1B in glioma, we established a glioma cell clone that stably overexpressed JARID1B protein (pBabe JARID1B) and transfected U251 cells with pBabe JARID1B in order to examine the effect on cell proliferation. The U251 cell line was selected for use in this experiment as a result of having the lowest expression of JARID1B in the six glioma cell lines. Following selection with puromycin, JARID1B expression was evaluated by western blot analysis (Fig. 4D). A high level of JARID1B was expressed in the U251-pBabe-JARID1B cells, whereas there was lower expression of JARID1B in the control cells transfected with the empty pBabe plasmid (U251-pBabe) and mock U251 cells. An MTT assay was performed to detect proliferation and the results revealed that U251-pBabe-JARID1B displayed higher proliferation rates than the control cells (Fig. 4E).

JARID1B enhances sphere formation as well as the migratory and invasive capacities of glioma cells

To explore the role of JARID1B in cellular transformation, the ability of the JARID1B-expressing U251 cells to form spheres was analyzed. The parent U251 cells and U251-pBabe cells formed only a few spheres (Fig. 5A). By comparison, the U251 cells expressing pBabe-JARID1B showed a significant increase in the number of spheres formed. To determine whether the overexpression of JARID1B affects cell migratory and invasive capacities, the wound healing assay, Transwell assay and Matrigel assay were performed using the established U251 stable transfectants and its parent cells. The results showed that U251-pBabe-JARID1B cells exhibited marked increases in cell migratory activity compared to the control U251-pBabe cells and the parent cells, suggesting that JARID1B may also enhance the rate of migration of glioma cells (Fig. 5B). This result was confirmed by the findings of the Transwell assay (Fig. 5C). Furthermore, U251-pBabe-JARID1B cells exhibited a greater degree of invasion through Matrigel (Fig. 5C).

Figure 5 Overexpression of jumonji AT-rich interactive domain 1B (JARID1B) enhances the migration and invasiveness of glioma cells as well as sphere formation. (A) Sphere formation of glioma cells was evaluated by the sphere formation assay. Scale bars, 100 µm. The number of spheres in each culture well was scored, and the results are expressed as means ± SD; **P<0.01 vs. pBABE group (based on Student's t-test). (B) Wound healing assay results showed that U251-pBabe-JARID1B cells exhibited significantly faster closure of the wound area. (C) U251-pBabe-JARID1B and its control cells were subjected to Transwell migration and Matrigel invasion assays. Quantification of migrated cells through the membrane and invaded cells through Matrigel of each cell line is presented as the proportions of their vector controls. **P<0.01 vs. pBABE group (based on Student's t-test). All results are representative of three independent experiments. Scale bar, 100 nm (A), 500 nm (B), 100 nm (C).

Suppression of JARID1B inhibits the proliferation, migration and invasiveness of glioma cells as well as sphere formation

To examine the effect of JARID1B knockdown on the proliferation, migration and invasiveness of glioma cells as well as sphere formation, the SW1783 human glioma cell line, a highly tumorigenic cell line commonly used in glioma research, was transfected with pSuper-shJARID1B or the control (pSuper). As shown by western blot analysis (Fig. 6A), JARID1B protein expression levels were significantly suppressed in the cells transfected with pSuper-shJARID1B. The results of the MTT assay also revealed that the suppression of JARID1B was associated with a decrease in cell proliferation (Fig. 6B). The sphere formation assay showed that the suppression of JARID1B significantly decreased the number of spheres formed (Fig. 6C). As indicated by the Transwell assay and Matrigel assay, knockdown of JARID1B expression significantly inhibited the migration and invasiveness of glioma cell (Fig. 6D). These results provide further evidence to indicate that JARID1B is involved in the proliferation and migration of glioma cells.

Figure 6 Downregulation of jumonji AT-rich interactive domain 1B (JARID1B) reduces tumorigenesis in vitro. (A) Establishment of downregulation of JARID1B in SW1783 cells. The results were evaluated by western blot analysis. (B) Proliferation of SW1783-pSuper-shJARID1B cells was significantly reduced compared with that in normal SW1783 control cells as measured by the MTT assay. (C) Sphere formation of glioma cells was evaluated by the sphere formation assay. Scale bars, 100 µm. The number of spheres in each culture well was counted, and the results are expressed as the means ± SD. (D) SW1783-pSuper-shJARID1B and its control cells were subjected to Transwell migration and Matrigel invasion assays. Quantification of migrated cells through the membrane and invaded cells through Matrigel of each cell line is shown as proportions of their vector controls. **P<0.01 vs. pSUPER group (based on the Student's t-test). All results are representative of three independent experiments. Scale bar, 100 nm (D).

JARID1B promotes tumorigenesis in glioma cells in vivo

To examine the function of JARID1B in vivo in glioma carcinogenesis, a xenograft model of glioma was established by implanting U251-pBabe and U251-pBabe-JARID1B cells subcutaneously into the right flanks of nude mice. The tumor size was continuously monitored weekly. The tumors of the mice injected with U251-pBabe-JARID1B cells were significantly larger than those of the control mice injected with U251-pBabe cells 2 weeks after tumor cell injection (Fig. 7A). The average weight of tumors from the mice injected with U251-pBabe-JARID1B cells was 5 g while that of the control mice injected with U251-pBabe cells was 2 g (Fig. 7B). Apparently, the tumor growth rate in the mice injected with U251-pBabe-JARID1B cells was greater than that in the control mice and this was confirmed by measuring the mean tumor volume 42 days after injection (Fig. 7C). Furthermore, JARID1B expression in the xenograft was confirmed by immunohistochemical analysis (Fig. 7D).

Figure 7 Jumonji AT-rich interactive domain 1B (JARID1B) promotes tumorigenesis in glioma cells in vivo. (A) Stable U251 transfectants were inoculated subcutaneously into the right flank of BALB/c-nu/nu mice. (B Tumor weight was measured 42 days after tumor cell injection, and the results are expressed as the means ± SD (n=3; **P<0.01). (C) Tumor volumes were measured every week after tumor cell injection, and the results are expressed as the means ± SD (n=3; **P<0.01). (D) Immunohistochemical analysis confirmed JARID1B expression in xenograft. **P<0.01 is based on Student's t-test. All results are representative of three independent experiments. Scale bar, 200 nm.

Activation of p-Smad2 contributes to JARID1B-induced oncogenic activities

The overexpression of JARID1B in the U251 cells significantly increased the expression of p-Smad2 (Fig. 8A). In comparison, the expression of TGF-β and Smad2 were unchanged. Significantly, we also found that the expression of cancer stem cell-related markers, namely CD133, Oct4, nestin and Bmi-1 were induced by JARID1B in the U251-pBabe-JARID1B cells, compared with that in the parental cells and U251-pBabe cells (Fig. 8A). Inversely, the knockdown of JARID1B in the SW1783 cells significantly decreased the expression of p-Smad2, CD133, Oct4, nestin and Bmi-1 (Fig. 8B) whereas it did not affect the expression of TGF-β and Smad2. As for p-Smad2, a well-known regulator of cancer stem cells, we examined whether the activation of p-Smad2 contributes to JARID1B-induced oncogenic activities. The suppression of p-Smad2 receptor signaling using the inhibitor LY364947 suppressed Smad2 phosphorylation, without affecting the expression of Smad2 in the U251-pBabe-JARID1B cells (Fig. 9). Treatment with LY364947, an inhibitor of p-Smad2 receptor signaling, suppressed the expression of CD133, Oct4, nestin and Bmi-1 in the U251-pBabe-JARID1B cells (Fig. 9). The suppression of p-Smad2 signaling also reduced the migration, invasiveness, proliferation of U251-pBabe-JARID1B cells as well as sphere formation (Fig. 10). This indicates that the induction of p-Smad2 expression by JARID1B activates a critical signaling pathway implicated in the formation of gliomas.

Figure 8 Jumonji AT-rich interactive domain 1B (JARID1B) upregulates the expression of phosphorylated (p-)SMAD family member 2 (Smad2) and cancer stem cell-related markers in glioma cells. (A) Western blot analysis of transforming growth factor-β1 (TGF-β1), p-Smad2, Smad2, CD133, octamer-binding transcription factor 4 (Oct4), nestin and BMI1 proto-oncogene, polycomb ring finger (Bmi-1) protein levels in U251-pBabe-JARID1B and its control cells. (B) Western blot analysis of TGF-β1, p-Smad2, Smad2, CD133, Oct4, nestin and Bmi-1 protein levels in SW1783-pSuper-shJARID1B and its control cells. β-actin was used as the internal control for western blot analysis.

Figure 9 Suppression of phosphorylated (p-)SMAD family member 2 (Smad2) signaling by LY364947 reverses the inductive effect of jumonji AT-rich interactive domain 1B (JARID1B) on the expression of glioma cancer stem cell-related markers. The expression of transforming growth factor-β1 (TGF-β1), p-Smad2, Smad2, CD133, octamer-binding transcription factor 4 (Oct4), nestin and BMI1 proto-oncogene, polycomb ring finger (Bmi-1) protein levels in LY364947-exposed U251-pBabe-JARID1B and its control cells were measured by western blot analysis. β-actin was used as the internal control for western blot analysis.

Figure 10 Treat with transforming growth factor-β (TGF-β) signaling pathway inhibitor LY364947 reduces the proliferation, migration, and invasiveness properties of U251-pBabe-jumonji AT-rich interactive domain 1B (JARID1B) cells as well as sphere formation. (A) MTT assay measured the growth rates in LY364947-treated U251-pBabe-JARID1B cells and the controls. (B) Sphere formation in the LY364947-treated U251-pBabe-JARID1B cells and the controls was evaluated by sphere formation assay. The number of spheres in each culture well was counted, and the results are expressed as the means ± SD. (C) LY364947-treated U251-pBabe-JARID1B cells and the controls were subjected to Transwell migration and Matrigel invasion assays. Quantification of migrated cells through the membrane and invaded cells through Matrigel of each cell line is presented as proportions of their vector controls. **P<0.01 is based on Student's t-test; **P<0.01 vs. pBabe JARID1B. All results are representative of three or four independent experiments. Scale bar, 100 nm (B and C).

Discussion

In the present study, we examined the role of JARID1B in glioma. Compared with the normal brain tissues, we detected raised levels of JARID1B in the majority of the glioma tissues. It was found that the overexpression of JARID1B promoted the proliferation, migration and invasiveness of glioma cells as well as sphere formation, and accelerated tumor growth in nude mice in vivo. Conversely, the knockdown of JARID1B suppressed the proliferation, migration and invasiveness of glioma cells as well as sphere formation. We also demonstrated that the function of JARID1B was associated with p-Smad2. Taken together, these findings suggest that JARID1B is involved in gliomagenesis.

JARID1B belongs to the family of histone demethylases, and it is capable of specifically removing the trimethyl modification of H3K4 in order to inhibit the transcription of the corresponding genes (16–18). Previous studies have suggested that JARID1B protein plays a vital role in the genesis of breast cancer and prostate cancer, and it is therefore considered to be an important drug target protein (8–10). Although the expression of JARID1B has been studied in some types of cancer, little is known about the expression of JARID1B in gliomas and its function in tumorigenesis. Thus, we examined the expression of JARID1B in glioma. We found that compared with the normal brain tissues, JARID1B was overexpressed in the majority of the high-grade glioma tissues examined.

Furthermore, we found that the overexpression of JARID1B increased the proliferation of glioblastoma cells and promoted tumorigenesis in vitro and in vivo, whereas the knockdown of JARID1B expression by shRNA produced an inhibitory effect on tumorigenesis. Notably, we found that the overexpression of JARID1B enhanced cell migration and invasion, whereas the knockdown of JARID1B produced an opposite effect. To the best of our knowledge, this is the first study to demonstrate that JARID1B regulates the proliferation and migratory abilities of glioma cells.

p-Smad2 is a member of the TGF-β superfamily which is involved in signal transduction (19). Signal pathways regulated by TGF-β include the Smad-dependent and Smad-independent pathways (20). TGF-β1/Smad is a classic signal transduction pathway. Its basic signal transduction process is: TGF-β extracellularly conjoins TβRII to activate TβR, then TβRI phosphorylates Smad2/3, and finally p-Smad2/3 in combination with Smad4 enters the cell nucleus and binds to DNA in order to regulate the transcription of many genes (20). p-Smad2 signaling is involved in regulating the proliferation, differentiation and survival/or apoptosis of many types of cells, including glioma cells (21,22). The increased expression of p-Smad2 correlates with the degree of malignancy of human gliomas (23). p-Smad2 may contribute to tumor pathogenesis by directly supporting tumor growth, the self-renewal of glioma initiating stem cells and inhibiting antitumor immunity (21). Therefore, we hypothesized that JARID1B may also modulate p-Smad2 expression in glioma cells. In this study, we found that the overexpression of JARID1B significantly increased the expression of p-Smad2 whereas the knockdown of JARID1B produced the opposite effect. We also found that JARID1B regulated the expression of cancer stem cell-related markers, namely CD133, nestin, Oct4 and Bmi-1. Suppression of the p-Smad2 signaling pathway by LY364947 inhibited the inductive effect of JARID1B on the expression of the cancer stem cell-related markers. Moreover, the suppression of p-Smad2 signaling also reduced the migration, invasiveness and proliferation of U251-pBabe-JARID1B cells as well as sphere formation. Thus, aberrant JARID1B expression in gliomas may enhance the oncogenic effects of the activated p-Smad2 signaling pathway. However, further investigation of the mechanisms involved in the expression profile of other genes is warranted.

In conclusion, we have demonstrated for the first time, to the best of our knowledge, that JARID1B is overexpressed in glioma tissues, and the overexpression of JARID1B increased the growth and migration of glioma cells in vitro and enhanced glioma tumorigenesis in vivo. Taken together, these findings suggest that JARID1B is potentially an important molecular target for the design of novel anti-glioma therapies.

Acknowledgments

The present study was supported by a grant from the National Natural Science Foundation of China (no. 81301168).

References