Differentiated embryonic chondrocyte-expressed gene 1 promotes temozolomide resistance by modulating the SP1-MGMT axis in glioblastoma

Abstract

Glioblastoma multiforme (GBM) is a malignant brain tumor with a high mortality rate and poor prognosis. Temozolomide (TMZ) is a first-line drug against GBM, but resistance limits its use. We previously reported that differentiated embryonic chondrocyte (DEC1) expression is associated with TMZ resistance and poor prognosis in GBM; however, the underlying mechanism remains unclear. By using glioma cell lines with stably overexpressed or silenced DEC1, we examined the effects of DEC1 on TMZ sensitivity using proliferation assays, Western blotting, and flow cytometry. We demonstrated that DEC1 overexpression suppressed, whereas DEC1 knockdown enhanced, TMZ-induced cell apoptosis in methylguanine methyltransferase (MGMT)-positive T98G and LN18 cells but not in MGMT-negative U251 cells. Mechanistically, DEC1 positively regulated MGMT through specificity protein 1 (SP1). MGMT silencing in DEC1-overexpressing cells or overexpression in DEC1-silenced cells abrogated DEC1’s effects on TMZ sensitivity, and siRNA-mediated SP1 knockdown phenocopied TMZ sensitivity, which was rescued by MGMT overexpression. Thus, DEC1 may control TMZ resistance via the SP1-MGMT axis. Immunohistochemical staining of the human glioma tissue microarray revealed that the expression levels of DEC1 and MGMT were correlated. Therefore, DEC1 expression has a predictive value for TMZ resistance and poor outcome in glioma patients, and is a novel therapeutic target in TMZ-resistant glioma.

Introduction

Glioma is a common primary tumor in the central nervous system and classified into four pathological grades (World Health Organization [WHO] classification, grades I to IV). Glioblastoma multiforme (GBM), which belongs to WHO grade IV, constitutes more than 70% of primary intracranial tumors and has a high recurrence and mortality [1]. GBM is a highly malignant, actively proliferating, and intensively invading cancer with a very poor prognosis [2]. At present, the standard strategy for GBM is a comprehensive treatment combining advanced surgery, radiotherapy, and chemotherapy such as temozolomide (TMZ). Nonetheless, the median survival time of patients with GBM is only approximately 12-15 months [3]. Drug resistance represents a key challenge in clinical practice and limits the therapeutic application of TMZ. Therefore, understanding the mechanisms underlying this drug resistance and identification of novel therapeutic targets and strategies are of great importance.

Temozolomide, as the first-line chemotherapeutic agent in glioma, exerts its anticancer effects by methylating the O⁶ position of guanine (O⁶-G) in the deoxyribonucleic acid (DNA) molecule. O⁶-methylguanine (O⁶-MeG) mispairs with thymine during DNA replication, which evokes the DNA mismatch repair system to recognize this mismatch region and excise the thymine. Persistent O⁶-MeG caused by TMZ in the template strand and the following mispair with thymine triggers futile repair cycles, ultimately leading to DNA strand breaks and, thus, cell death [4]. O⁶-methylguanine-DNA methyltransferase (MGMT), a suicide DNA repair enzyme, is able to directly reverse O⁶-MeG lesions by transferring these groups to its active center and thus becoming inactivated [5]. Therefore, MGMT is a key player to mediate the resistance to TMZ, and MGMT levels are regarded as an important biomarker to evaluate the responsiveness to TMZ [5]. It has been well known that the epigenetic regulation of MGMT plays an important role in glioma, with approximately 20% of primary GBMs lacking MGMT activity and approximately 40% of the gliomas displaying hypermethylation in the MGMT promoter region [6,7]. Moreover, several transcription factors, such as secreted protein 1 (SP1), activator protein 1 (AP1), CCAAT-enhancer-binding proteins (CEBP), hypoxia inducible factor-1α (HIF-1α), and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), bind to the MGMT promoter to induce or suppress MGMT expression [4]. Previous studies have reported that the MGMT level is primarily determined by the transcription factor SP1 and upregulated by glucocorticoids [8], whereas p53 can negatively regulate MGMT expression via interaction with and sequestration of SP1 to the cognate elements in the MGMT promoter [9]. However, the regulation of MGMT and its relevance are far from being clarified.

Differentiated embryonic chondrocyte-expressed gene 1 (DEC1), also known as split- and hairy-related protein-2 (SHARP-2), basic helix-loop-helix binding protein 2 (BHLHB2), or stimulated with retinoic acid 13 (Stra13), belongs to a subfamily of the basic helix-loop-helix family, which also contains DEC2 (BHLHE41/SHARP-1) [10]. The DEC1 gene is located at chromosome 9q33.1. The human DEC1 protein consists of 412 amino acid residues with basic helix-loop-helix (bHLH) and Orange domains. As a transcription factor, DEC1 either represses gene transcription by directly binding to E-box motifs of its target genes such as DEC2 [11], CCND1 [12], and LKB1 [13], or activates gene transcription by binding to the SP1 site of its target genes, such as SURVIVIN [14]. DEC1 can also act as a cofactor to inhibit the SP1-mediated claudin-1 (CLDN1) transactivation [15]. In function, DEC1 is an essential regulator of circadian rhythms [16]. In addition, DEC1 has an important role in various cellular processes such as cell growth [10], cell cycles [17], cell metabolism [18], cell differentiation, and apoptosis [19]. The aberrant expression of DEC1 is involved in the pathologies of various disorders [20]. Particularly, DEC1 is closely associated with cancer but its functions are controversial in different types of cancer. It was reported that DEC1 suppresses cell proliferation, migration, and invasion in breast and oral cancer [21], and its expression is correlated with clinicopathological parameters [22]. However, in thyroid cancer, DEC1 is significantly overexpressed in all major histologic types and promotes cell growth and invasiveness via an interplay with NOTCH1 [17]. We previously reported that DEC1 expression is increased in glioma in comparison to normal tissues, and its expression is correlated with malignancy grade, prognosis, as well as TMZ resistance [3]. Furthermore, DEC1 was reported to promote GBM tumor cell proliferation, migration, and invasion [23]. However, to the best of our knowledge, far less is known about whether, and how, DEC1 is involved in TMZ resistance.

Since MGMT is the most important factor in the response of glioma cells to TMZ as mentioned above, we speculated that DEC1 may regulate cellular sensitivity to TMZ via MGMT. Therefore, the present study intended to explore the role of DEC1 in the response to TMZ and the interactive relationships between DEC1 and MGMT. Our results showed that DEC1 can positively regulate the expression of MGMT through SP1, and this axis can promote therapeutic resistance in response to TMZ.

Material and methods

Patients’ samples

The tissue microarrays were preloaded with brain tissue samples from five healthy controls and 35 patients with GBM, containing 80 points (10 for normal brains and 70 for GBM); they were commercially obtained from Xi’an Alenabio Technology Co. Ltd. (Gl805; Xi’an, China). The experiments were approved by the Research Ethics Committee and the investigations were carried out following the rules of the Declaration of Helsinki of 1975 (https://www.wma.net/what-we-do/medical-ethics/declaration-of-helsinki/), revised in 2013.

The Cancer Genome Atlas (TCGA)

DEC1 gene expression profile for GBM patients was obtained from the The Cancer Genome Atlas (TCGA) data portal (https://tcga-data.nci.nih.gov/tcga/), which contains 249, 265 and 153 samples of Gll, Glll and GlV, respectively.

Reagents and plasmids

Dimethyl sulfoxide (DMSO; cat no. DH105-2) was purchased from Xi’an Kehao Bioengineering Co., Ltd. (Xi’an, China). Temozolomide (cat no. 85622-93-1) was purchased from Sigma-Aldrich (St. Louis, MO, USA) and dissolved in DMSO to obtain a 100-mM stock concentration. Puromycin (cat no. HY-CL13900) was purchased from MCE Bioengineering Co., Ltd. (Shanghai, China) and dissolved in phosphate-buffered saline (PBS) at a 1 mg/ml concentration. Hygromycin B (cat no. HY-B0490) was purchased from MCE Bioengineering Co., Ltd. (Shanghai, China) and dissolved in phosphate-buffered saline (PBS) at a 500 mg/ml concentration. Geneticin (cat no. DH138-2) was purchased from Xi’an Kehao Bioengineering Co., Ltd. and was dissolved in PBS to obtain a stock solution of 20 mg/ml.

Lentivirus plasmids containing full-length DEC1 (EX-DEC1-pEZ-Lv105) or MGMT (EX-MGMT-pEZ-Lv151) or their corresponding empty cherry plasmids (i.e., EX-pEZ-Lv105 and EX-pEZ-Lv151) were commercially purchased (Guangzhou Fulengen Co., Ltd., Guangdong, China). The following lentivirus plasmids were purchased from Guangzhou FulenGen Co., Ltd.: short hairpin DEC1 and its empty scrambled plasmid (CSHCTR001-LVRU6P), as well as short hairpin MGMT and its empty scrambled plasmid (CSHCTR001-LVRU6GH). The full-length MGMT promoter-reporter construct (pGL3.0-pMGMT-Luc vector, forward: 5’-AATATCAGTCGGCACTCTGATTCCA-3’. reverse: 5’-CTCTTGCTTTTCTCAGGTCC-3’) and its empty cherry vector were purchased from BokeGen Co., Ltd. (Xi’an, China).

Cell culture, cell transfection, and lentivirus infection

Human GBM cell lines (T98G, LN18, and U251), normal human glial cells (HEB) and human embryonic kidney (HEK) 293T cells (American Type Culture Collection, Manassas, VA, USA) were maintained in Dulbecco’s Modified Eagle’s Medium (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% fetal bovine serum (Sigma-Aldrich; Merck Millipore, Darmstadt, Germany) in a humidified incubator containing 5% carbon dioxide (CO₂) at 37°C. In addition, 100 μg/mL streptomycin and 100 U/mL penicillin were routinely added to the medium. For lentiviral packaging, HEK293T cells were seeded into 10-cm plates, and when the density reached 80%, the Lenti-Pac HIV Expression Packaging Kit (Genecoponia, Inc., Rockville, MD, USA) was applied. The recombinant expression plasmid or empty cherry plasmid were transfected into HEK293T cells together with the packing plasmid mix (Genecoponia, Inc., Rockville, MD, USA) using lipofectamine 2000 (Thermo Fisher Scientific, Inc.). The supernatants which contain virus were harvested after 48 hours. Glioma cells were then infected with the individual recombinant virus. Stable T98G, LN18, and U251 cells with overexpressed/silenced DEC1 and their corresponding control cells expressing Cherry were generated using selection with 0.8 mM (for T98G and LN18) or 0.4 mM (for U251) puromycin (Invitrogen; Thermo Fisher Scientific, Inc.). Stable T98G and LN18 cells overexpressing MGMT or Cherry were generated by lentivirus infection and selection with 400 or 600 mM G418. Stable T98G and LN18 cells with suppressed MGMT or negative control were generated by lentivirus infection and selection with 600 mM hygromycin B in T98G or LN18 cells, respectively. For transient transfection, cells were seeded into six-well plates and transfected with individual plasmids or siRNAs using lipofectamine 2000 (Thermo Fisher Scientific, Inc.), based on the manufacturer’s instructions.

Quantitative real-time PCR analysis

Total RNA was extracted from U251, T98G, or LN18 cells using RNAiso Plus (Takara Bio, Inc., Kusatsu, Japan). The One-Step SYBR PrimeScript RT-PCR Kit (Takara Bio, Inc., Kusatsu, Japan) was used for qRT-PCR detection of DEC1, MGMT, or β-actin. The primers were as follows: DEC1 forward: 5’-ACTTACCTTGAAGCATGTGAAAGCA-3’, reverse: 5’-CATGTCTGGAAACCTGAGCAGAA-3’; β-actin forward: 5’-TGACGTGGACATCCGCAAAG-3’, reverse: 5’-CTGGAAGGTGGACAGCGAGG-3’; MGMT forward: 5’-GCAATTAGCAGCCCTGGCA-3’, reverse: 5’-CACTCTGTGGCACGGGAT-3’. The qRT-PCR was conducted using a Bio-Rad C1000 thermal cycler with the following conditions: 95°C for 3 minutes, followed by 40 cycles of 95°C for 10 seconds and 55°C for 30 seconds. The expression levels were calculated relative to the β-actin levels as the endogenous control. Relative expression was calculated as 2^{(Ct test gene - Ct ACTB)}.

Luciferase reporter assay

HEK293T cells, at 60% confluence, were seeded into six-well plates and then transfected with DEC1-encoding plasmid (EX-DEC1-pEZ-Lv105) or its corresponding empty cherry plasmid (EX-pEZ-Lv105), together with the control luciferase vector pGL3.0-Luc or the luciferase reporter pGL3.0-pMGMT-Luc which were commercially purchased (Guangzhou Fulengen Co., Ltd., Guangdong, China). The cells were harvested 48 hours after transfection. The Luc activity was determined, based on the manufacturer’s instructions (Promega, Madison, WI, USA).

RNA interference

Chemically synthesized siRNA duplexes were obtained from Shanghai GenePharma Co., Ltd. (Shanghai, China). Negative control siRNA sequence: 5’-UUCUCCGAACGUGUCACGUTT-3’; siSP1#1: 5’-GCCGUUGGCUAUAGCAAAUTT-3’; siSP1#2: 5’-CCAGCAACAUGGGAAUUAUTT-3’; siSP1#3: 5’-GUGCAAACCAACAGAUUAUTT-3’. siRNA transfections were conducted with lipofectamine 2000, based on the manufacturer’s instructions.

Cell proliferation assay

T98G, LN18, and U251 cells were seeded at a density of 1,000 cells/well in 96-well plates in septuplets. Cell proliferation was estimated using the CCK8 (Apexbio, Co., Ltd., Boston, MA, USA; cat no. k1018). To each well, 10 μL of CCK-8 solution were added, and the cells were incubated for 3 hours at 37°C. The absorbance at 490 nm was measured using a microplate reader (SH-9000; Hitachi, Tokyo, Japan).

Western blot analysis

Cells were lysed in radioimmunoprecipitation assay (RIPA) buffer (50 nM Tris-HCl, 150 nM NaCl, 1% NP40, 0.1% sodium dodecyl sulfate [SDS], 0.5% deoxycholate, and 1 mM phenylmethanesulfonyl fluoride) for 30 minutes on ice. A total of 30 μg protein was separated by gradient SDS-polyacrylamide gel electrophoresis (SDS-PAGE) (4%-12% polyacrylamide). Proteins were then transferred to nitrocellulose membranes. The membranes were blocked, immunoblotted with the indicated primary antibodies, and subsequently incubated with the corresponding horseradish peroxidase-conjugated secondary antibodies. The signals were detected with an enhanced chemiluminescence system (Tanon 5500; Tanon Science and Technology, Co., Ltd., Shanghai, China). The following commercial antibodies were purchased: DEC1 (dilution 1:5,000; rabbit monoclonal; cat no. ab70723; Abcam, Cambridge, UK), MGMT (dilution 1:1,000; mouse monoclonal; cat no. ab69629; Abcam, Cambridge, UK), Bcl-2 (dilution 1:1,000; rabbit monoclonal; cat no. 2876; CST, Boston, MA, USA), SP1 (dilution 1:1,000; rabbit monoclonal; cat no. 5931; CST, Boston, MA, USA), GAPDH (dilution 1:5,000; rabbit monoclonal; Zhuangzhibio; Xi’an, China). Horseradish-peroxidase-conjugated secondary antibodies (cat no. abs921; absin, Shanghai, China).

Apoptosis assay

V-fluorescein isothiocyanate (FITC)/PI apoptosis detection kit (Cat 55647; BD, Franklin Lakes, NJ, USA) was used. Cells were treated with TMZ or DMSO. A total of 10⁵ cells were collected and resuspended in binding buffer. The FITC-Annexin V and PI (each 5 μL) were then added to the cell suspension and incubated for 15 minutes at room temperature in the dark. Stained cells were analyzed using a flow cytometer (FACSCalibur, Becton Dickinson, Franklin Lakes, NJ, USA). Annexin V+/PI- (i.e., early apoptotic cells) and Annexin V+/PI+ (i.e., late apoptotic cells) were calculated as total apoptotic cells (from triplicates and using Student’s t-test).

Immunohistochemistry

Immunohistochemistry was conducted, based on standard procedures. Primary antibodies directed against DEC1 (1:250 dilution; cat no. ab90594; Abcam, Cambridge, UK), or MGMT (1:100 dilution; cat no. ab69629; Abcam, Cambridge, UK) were used with the ABC Kit and DAB (both obtained from Vector Resources, Inc., Torrance, CA USA) for staining. The results were obtained using a digital slide scanner (3DHISTECH, Ltd., Budapest, Hungary). The immunoreactive score (IRS) was applied. The deliberate score standard was as follows (presented as the percentage of cells with positive staining and score number): < 5%, 0; 5%-25%, 1; 26%-50%, 2; 51%-75%, 3; >75%, 4. The color score standard used the staining intensity, as follow: colorless, 0; mild, 1; moderate, 2; and strong, 3.

Colony formation assay

T98G, LN18, and U251 cells were seeded into six-well plates at a density of 300 cells/well in triplicate. The plates were incubated at 37°C with 5% CO₂ for 3 weeks. The plates were then washed with PBS, the developed cell colonies were fixed with 4% formaldehyde, and the cells were stained with Giemsa for 20-30 minutes. The number of colonies was directly reported.

Statistical analysis

All statistical analyses were conducted using SPSS 13.0 package (SPSS Inc., Chicago, IL, USA). Quantitative data are presented as the mean ± the standard deviation of at least three independent experiments. For comparison of two-groups datasets, 2-sided, unpaired t-test was applied. For comparison of the clinical relevance of DEC1 and MGMT expression, 2-sided, unpaired t-test and chi-square test was applied. Statistical significance was defined as P < 0.05. *; P < 0.05, **; P < 0.01, ***; P < 0.001, and ****; P < 0.0001.

Results

DEC1 promotes TMZ resistance in MGMT-positive GBM cells

Our previous study showed that DEC1 expression is positively correlated with the malignancy grade and the TMZ resistance of GBM [3]. Therefore, we speculated that DEC1 is involved in TMZ resistance by regulating key molecules that determine TMZ sensitivity. To test this hypothesis, we chose three GBM cell lines: T98G, LN18, and U251. Among these cell lines, U251 is MGMT-deficient and TMZ-sensitive, whereas T98G and LN18 express high levels of MGMT, rendering them resistant to TMZ [24]. Consistent with the literature, our Western blot analysis showed that T98G and LN18, but not U251, expressed high levels of MGMT (Figure S1A). Moreover, the cell counting kit-8 (CCK-8) assay, used to test the sensitivity of cells to TMZ, demonstrated that TMZ significantly inhibited cell proliferation in U251, but T98G and LN18 exhibited much more resistance to TMZ (Figure S1B). Therefore, a TMZ concentration of 600 μM in LN18 and T98G and 200 μM in U251 were applied in the subsequent experiments, according to our result and the recent study [43]. We then established stable T98G, LN18, and U251 cell lines by individually infecting cells with DEC1 or shDEC1 recombinant lentivirus to up- or downregulate the expression of DEC1. The success of stable infection in these cell lines was confirmed by Western blotting analysis and by real-time polymerase chain reaction (PCR) assay (Figure S1C and S1D). The CCK-8 and the colony formation assays demonstrated that upregulation of DEC1 significantly promoted cell growth, colony-forming ability and cell migration, while downregulation of DEC1 had the opposite effect (Figure S1E-J). These results indicated that DEC1 has an important role in the proliferation and migration of the chosen GBM cell lines.

To examine whether DEC1 expression affects the sensitivity of glioma cells to TMZ, we conducted a CCK-8 assay in these cells upon treatment with TMZ. As shown in Figure 1A and 1B, T98G and LN18 cells stably overexpressing DEC1 were more resistant to TMZ, compared to the vector cells, whereas DEC1 knockdown significantly reverted TMZ resistance in these cells. Next, we conducted propidium iodide (PI)-annexin V double staining and flow cytometry analysis to examine apoptosis in these cells after treatment with TMZ. Our results demonstrated that DEC1 overexpression was able to decrease apoptosis in T98G cells (4.60% ± 0.45% vs 16.89% ± 1.29%, P < 0.01) and LN18 cells (4.06% ± 0.19% vs 5.58% ± 0.09%, P < 0.01; Figure 1C). By contrast, DEC1 knockdown increased apoptosis in T98G cells (14.49% ± 0.78% and 11.40% ± 0.94% vs 7.96 ± 0.88%, P < 0.05, and P = 0.15, respectively) and in LN18 cells (16.12% ± 2.72% vs 6.41% ± 0.81%, P < 0.05, and 11.10% ± 0.36% vs 6.41% ± 0.81%, P < 0.01, respectively; Figure 1D). In a similar manner, no significant difference existed in TMZ-induced cell apoptosis between U251-DEC1 cells and U251-vector cells (12.16% ± 2.51% vs 9.55% ± 2.23%, P = 0.40), as well as between U251-scramble cells and U251-shDEC1#1 and U251-shDEC1#2 cells (11.26% ± 1.53% and 12.08% ± 2.38% vs 9.30% ± 1.63%, P = 0.40 and P = 0.40, respectively; Figure 1C and 1D). These data suggested that DEC1 regulated TMZ resistance in an MGMT-dependent manner. To further confirm these results, Bcl-2 level, which has been widely used as an anti-apoptosis indicator, was assessed by using Western blotting and exhibited consistent alterations (Figure 1E).

Figure 1.

DEC1 promotes TMZ resistance in MGMT-positive GBM cells. Stable GBM cell lines (T98G, LN18, and U251) were established by an infection with DEC1- or shDEC1-recombinant lentivirus and selection with puromycin. (A and B) LN18 and T98G were treated with 600 μM TMZ at different time points. cell survival was evaluated with CCK-8 assays (n = 3). (C and D) Cells were treated with TMZ (600 μM for LN18 and T98G, and 200 μM for U251) for 48 hours; cell apoptosis was evaluated using flow cytometry analysis. Representative images are shown (n = 3). (E) Cells were treated with 600 μM TMZ for 48 hours; the expression level of DEC1 and Bcl-2 were evaluated using Western blot (n = 3). In (A-D), data are presented as the mean ± SD. Unpaired t test for two-groups datasets. *; P < 0.05, **; P < 0.01, ***; P < 0.001. CCK-8, cell counting kit 8; DEC1, differentiated embryonic chondrocyte; GBM, glioblastoma multiforme; MGMT, methylguanine methyltransferase; RT-qPCR, quantitative reverse transcription polymerase chain reaction; SD, standard deviation; shDEC1, short hairpin differentiated embryonic chondrocyte; specificity protein 1; TMZ, temozolomide.

DEC1 positively regulates SP1 and MGMT

It is a consensus that MGMT is a critical molecule in the development of resistance to TMZ [26]. High levels of MGMT are associated with TMZ resistance and poor prognosis of patients with GBM [27]. On account that DEC1 modulates TMZ sensitivity in MGMT-positive but not MGMT-negative cell lines, we tested whether DEC1 is able to regulate MGMT expression. As shown in Figure 2A, Western blot results demonstrated that MGMT could be upregulated by DEC1 overexpression and downregulated by DEC1 knockdown. Moreover, real-time PCR assays revealed similar results (Figure 2B and 2C), which indicated that DEC1 enhances MGMT at the transcription level. To further confirm this finding, pGL3.0-pMGMT-Luc, a luciferase reporter plasmid containing the promoter of MGMT, or the control luciferase vector pGL3.0-Luc was transfected into 293T cells together with the DEC1-overexpression vector or an empty vector to evaluate luciferase activity. As shown in Figure 2D, DEC1 significantly enhanced the fluorescence signal, compared to the control group. This indicated that DEC1 positively regulates MGMT at the transcriptional level.

Figure 2.

DEC1 positively regulates SP1 and MGMT. (A) The expression level of DEC1, MGMT and GAPDH in stable GBM cells was detected with Western blot analysis (n = 3). (B and C) The transcriptional levels of MGMT was evaluated using RT-qPCR (n = 3). (D) The 293T cells were transiently co-transfected with the DEC1-encoding construct or empty vector, and the control luciferase vector pGL3.0-Luc or the luciferase reporter pGL3.0-pMGMT-Luc. Luciferase activity was evaluated 48 hours after transfection and shown as the ratio of promoter-reporter activity to control vector luciferase activity (n = 3). (E) The LN18 cells were transfected with siNC, siSP1#1, siSP1#2 or siSP1#3. The expression levels of SP1, DEC1, and MGMT were evaluated with Western blot (n = 3). (F) In stable DEC1 GBM cell lines (i.e., T98G, LN18, and U251), the expression level of SP1 along with DEC1 and MGMT were evaluated with Western blot (n = 3). (G) Stable DEC1-overexpressing LN18 was transiently transfected with siSP1#1 or siNC. The expression levels of MGMT along with SP1 and DEC1 were evaluated with Western blot. In (B-D), the data are presented as the mean ± SD. Unpaired t test for two-groups datasets. *; P < 0.05, **; P < 0.01, ***; P < 0.001.

We were also interested in determining whether DEC1 acted directly on the MGMT promoter or whether its activity relied on other transcription factors. It has been reported that the MGMT promoter contains binding sites for a series of transcription factors such as SP1 and AP1 [9]. Interestingly, Zheng et al. demonstrated that DEC1 suppresses the transcription of CLDN1 by interaction with SP1 but not by binding to the E-box motifs in the CLDN1 promoters, which also contain SP1 binding sites [15]. Similar to CLDN1, the MGMT promoter contains SP1 binding sites and E-box motifs; therefore, we wondered whether DEC1 regulates MGMT transcription via SP1.

We first examined SP1 levels in DEC1-silencing cells. Strikingly, we found that SP1 was positively regulated by DEC1 in LN18, T98G, and U251 cells (Figure 2F). To address whether DEC1 regulates MGMT transcription via SP1, we silenced SP1 expression using siRNAs in T98G and LN18 (Figure 2E) to examine the expression of MGMT and DEC1. As shown in Figure 2E, the silencing of SP1 led to a dramatic decrease in MGMT expression but had no effect on DEC1 levels. Thus, SP1 is a downstream target of DEC1 and an upstream regulator of MGMT.

We then knocked down SP1 with siRNAs in DEC1-overexpressing LN18 cells to observe its effects on DEC1-dependent MGMT alterations. The results showed that the regulation of MGMT by DEC1 could be abrogated by SP1 (Figure 2G). Together, these data indicated that SP1 is an indispensable factor for MGMT regulation by DEC1.

The SP1-MGMT axis is essential for DEC1-mediated TMZ resistance

We have already provided evidence that DEC1 can promote TMZ resistance and upregulate MGMT expression in GBM cell lines. We next examined whether DEC1 exerts its role in the cellular response to TMZ via MGMT. For this purpose, we further knocked down MGMT in DEC1-overexpressing cells or overexpressed MGMT in stable DEC1-silenced T98G and LN18 cells. The expression levels of MGMT were confirmed by Western blot analysis (Figure 3A). We then treated these cells with TMZ and determined cell survival via CCK-8 assay and cell apoptosis via flow cytometry assay. As shown in Figure 3B-D, cell survival was significantly increased while apoptosis decreased after DEC1 overexpression, and this was reversed when MGMT was simultaneously downregulated. Similarly, DEC1-silenced T98G and LN18 cells exhibited significantly lower viability but higher apoptosis rates than their scrambled control group, but this effect could be rescued by forced MGMT upregulation. Western blot analyses of Bcl-2 also corroborated these results (Figure 3A). These data indicated that MGMT is an indispensable factor that mediates the effect of DEC1 on TMZ resistance.

Figure 3.

The SP1-MGMT axis is essential for DEC1-mediated TMZ resistance. After being individually infected with DEC1 or shDEC1 recombinant lentivirus, the T98G and LN18 cells were further infected with MGMT, shMGMT or their individual empty control recombinant lentivirus and selected with hygromycin and G418 to establish stable cell lines. (A) Cells were treated with TMZ (600 μM) for 48 hours. The expression level of MGMT, DEC1, and Bcl-2 were determined by Western blot. (B) Cells indicated were treated with different concentration of TMZ for 48 hours, and cell survival was evaluated by CCK-8 assay (n = 3). (C and D) Cells were treated with 600 μM TMZ for 48 hours and cell apoptosis was evaluated by flow cytometry analysis (n = 3). Representative images are shown in (C). In (B and D), data are presented as the mean ± the SD. Unpaired t test for two-groups datasets. *; P < 0.05, **; P < 0.01, ***; P < 0.001, ****; P < 0.0001.

Secreted protein 1 (SP1) is indispensable for MGMT regulation by DEC1 (see Figure 2G); therefore, we then investigated whether SP1 has an important role in MGMT expression and resistance to TMZ. To test this, we knocked down SP1 via siRNA in T98G and LN18 cells and observed the effects on TMZ sensitivity by performing CCK-8 assays, apoptosis analyses, and Western blots. When SP1 was repressed, MGMT expression was also markedly downregulated (Figure 4A), and more cells underwent apoptosis (Figure 4A-C). More importantly, the effects of SP1 knockdown on cell survival and apoptosis was entirely abrogated if MGMT was forcedly overexpressed in the cells at the same time (Figure 4A-C). These results collectively indicated that the SP1-MGMT axis is essential for DEC1-mediated TMZ resistance.

Figure 4.

SP1 regulates TMZ sensitivity via MGMT. The T98G and LN18 cells infected with MGMT or empty Cherry recombinant lentivirus were further transiently transfected with siSP1#NC or siSP1#1. A. Cells were treated with TMZ at 600 μM for 48 hours. The expression level of SP1, MGMT, DEC1, and Bcl-2 were analyzed using Western blot. B. Cells were treated with different concentrations of TMZ for 48 hours and cell survival were evaluated using CCK-8 assays (n = 3). C. Cell apoptosis was evaluated by using flow cytometry analysis (n = 3). Data are presented as the mean ± the SD. Unpaired t test for two-groups datasets. *; P < 0.05, **; P < 0.01, ***; P < 0.001, ****; P < 0.0001.

DEC1 is overexpressed in GBM and is correlated with the expression level of MGMT

Based on the findings of our previous study [3], high expression levels of DEC1 are associated with poor prognosis and chemotherapeutic resistance to TMZ. In this study, we examined the discrepancy between endogenous DEC1 expression levels in GBM (T98G, LN18, and U251) and normal cerebral (i.e., HEB) cells by Western blot, and found that DEC1 was overexpressed in GBM cell lines (Figure 5A). We also analyzed the glioma database provided by the Cancer Genome Atlas (Gll: 249 samples, Glll: 265 samples, GIV: 153 samples). Student’s t-test analysis confirmed that the expression level of DEC1 in glioblastoma was significantly higher than those in Gll and Glll giloma (Figure 5B, GlV vs Gll, G vs Glll, P < 0.0001). To further explore the clinical relevance of DEC1 and MGMT expression, immunohistochemical (IHC) staining of glioblastoma TMA, containing 35 patients and 5 normal tissue samples, was used to determine the expression levels of DEC1 and MGMT. The result showed that the expression levels of both DEC1 and MGMT were significantly higher in glioblastoma tissues than in normal brain tissues (Figure 5C and 5D, P < 0.01); moreover, DEC1 expression levels were evaluated by IHC, and patients were stratified as low DEC1 expression (IRS < 6) or high DEC1 expression (IRS ≥ 6). According to the IRS of MGMT, we found that percentage of high MGMT expression (IRS ≥ 6) in high DEC1 expression group is much higher than those in low DEC1 expression group (Figure 5E and 5F, P < 0.05). These findings are in accordance with our cytological results, indicating that the expression of DEC1 is positively associated with that of MGMT. In summary, DEC1 is upregulated in GBM tissue, compared to normal human tissue, and its expression is correlated with the GBM malignancy grade, and the expression of MGMT.

Figure 5.

The expression of DEC1 is upregulated in GBM and correlated with the expression of MGMT. A. The expression level of DEC1 was analyzed using Western blot in normal human brain cell line HEB and GBM cell lines (T98G, LN18, and U251). B. The correlation between the expression level of DEC1 and the glioma grade was analyzed, based on the database information provided by the Cancer Genome Atlas (TCGA). C and D. The expression level of DEC1 and MGMT were analyzed by immunohistochemical staining of glioma tissue microarray, which contains 35 GBM tissues and 5 normal brain tissues. E. Representative immunohistochemistry images with high and low levels of DEC1 and MGMT are shown. F. The IRS scores of DEC1 and that of MGMT (low expression group: IRS < 6 or high expression group: IRS ≥ 6) were counted and the percentage of high/low MGMT in DEC1 high/low were shown. P < 0.05, chi-square test was applied.

Discussion

Our previous study showed that DEC1 is positively associated with grade, prognosis, and TMZ resistance of glioma [3], but the molecular mechanisms for these effects remained unclear. Our study demonstrated that forced expression of DEC1 suppressed TMZ-induced cell apoptosis, whereas the knockdown of DEC1 had the opposite effect in MGMT-positive glioma cell lines, which indicated that DEC1 promoted TMZ resistance in an MGMT-dependent manner. Regarding the mechanism, we confirmed that DEC1 positively regulated MGMT at the transcriptional level. We further showed that SP1 was an essential factor for the regulation of MGMT by DEC1. Moreover, DEC1-dependent TMZ resistance can be blocked by MGMT manipulation. These data illustrated that DEC1 is involved in glioma cell sensitivity to TMZ by regulating the SP1-MGMT axis. In the samples of TCGA, DEC1’s expression level increased in GBM compared to gradell or gradelll samples, which indicated that DEC1 was positively associated with the glioma grades. Finally, immunohistochemistry confirmed that the expression of DEC1 and MGMT was increased in GBM tissues and the expression level of DEC1 was positively associated with those of MGMT. Our data demonstrated for the first time that DEC1, as an important upstream regulator of MGMT, controlled TMZ sensitivity. This finding suggested that it is a novel potential target for glioma treatment.

The expression of DEC1 is differentially upregulated or downregulated among various cancers and is associated with tumor grade, malignant characteristics, and patient prognosis in various cancers [10,21,30-32]. However, the evidence is inconsistent and even contradictory regarding whether DEC1 functions as a tumor promoter or suppressor. For example, DEC1 was shown to inhibit cell proliferation of breast cancer [33], colon cancer [34], and oral cancer cells, probably by regulating cyclin D1 [22] or cyclin E [33]. By contrast, DEC1 promotes cell growth in thyroid cancer [17] and gastric cancer [19] by controlling cell cycle-related genes. Moreover, DEC1 facilitates epithelial-mesenchymal transition processes and invasiveness in breast [21], thyroid [17], pancreatic [37], and liver cancers but inhibits migration and invasion in oral cancer cells [10]. Thus, DEC1 may act as a tumor promoter or suppressor, depending on different tumor tissue contexts. Our previous study showed for the first time that in glioma, the expression of DEC1 is upregulated and positively associated with the tumor grade. Our current study further demonstrated that DEC1 can antagonize TMZ-induced cell apoptosis, thereby promoting TMZ resistance. Of note, DEC1 was reported to be upregulated in several types of cancers and to protect against serum-starvation-induced apoptosis by transcriptionally activating SURVIVIN, an anti-apoptotic protein [14], or antagonize DNA damage-induced cell death by interacting with p53 and weakening its induction of MIC-1 [38]. Similarly, DEC1 inhibited apoptotic effects in mouse mammary carcinoma cell lines induced by a TGF-βR kinase inhibitor [39]. Interestingly, there is also contradictory evidence showing that DEC1 exhibits pro-apoptotic effects in human esophageal [40] and breast [41] cancer cells after treatment with chemotherapeutic agents such as cisplatin or paclitaxel [40,41]. As such, the exact role and mechanism of DEC1 may be highly context-dependent in different types of cancer in response to various stimuli, deserving further investigations.

The novel mechanistic finding of our study is that DEC1 regulates the SP1-MGMT axis in glioma cells. It is well-established that MGMT is a key regulator of TMZ sensitivity in glioma cells, and epigenetic silencing, via methylation of its promoter, contributes so much to its expression level that the MGMT methylation status is used as a predictive marker for therapy [13]. Recently, accumulating studies have revealed that MGMT is subjected to regulation by a variety of transcription factors, such as AP1 [42], SP1 [8], and NF-κB, which play an important role in the resistance of glioma to TMZ [24,43].

In particular, many SP1 sites exist within the MGMT promoter so the basal level of MGMT expression is primarily determined by the SP1 level [8]. Consistent with this finding, we showed that SP1 knockdown significantly reduced the MGMT expression level and enhanced TMZ-induced apoptosis in glioma cells. Of note, the study by Kitange et al. [28] also demonstrated that increased SP1 recruitment within the MGMT promoter is involved in MGMT upregulation and acquired TMZ resistance in GBM. Considering that SP1 is upregulated in GBM and promotes glioma cell proliferation [35], whereas its inhibition leads to decreased cell migration and invasion in glioma and other cancers [35,36], our study is in support of other publications and strengthens the notion that SP1 might be a potential therapeutic target in glioma.

In addition, we showed that DEC1 regulates TMZ sensitivity via the SP1-MGMT axis because the effects of DEC1 on TMZ-induced cell apoptosis rely on the presence of MGMT and can be completely abolished by the manipulation of MGMT expression. As a transcription factor, DEC1 can either activate or repress its target genes [25]. Interestingly, DEC1 can not only directly bind to the SP1 sites of its target genes to activate gene transcription such as SURVIVIN [14] but also interact with SP1 and prevent its binding to the promoter to suppress CLDN1 transcription [15]. Our current study showed that DEC1 can also regulate SP1 levels, albeit the detailed mechanism still needs to be clarified. Together, it seems an intimate interplay exists between SP1 and DEC1 in the regulation of gene transcription, wherein DEC1 can function by occupying SP1 sites, sequestering SP1 binding, or regulating SP1 content.

We provided robust evidence that DEC1 regulates TMZ sensitivity via the SP1-MGMT axis and that its expression is upregulated in GBM and correlated with the expression of MGMT; however, we have not clarified how DEC1 regulates SP1. It is still needed to address whether DEC1 regulates SP1 transcriptionally, post-transcriptionally, or epigenetically, and to evaluate its clinical relevance. In addition, a major limitation of this study was that glioma cell lines, but not patient-derived primary glioma cells or an orthotopic glioblastoma mouse model, were utilized to identify the role of DEC1 in glioma, which is worth an in-depth investigation in the future.

In conclusion, and to the best of our knowledge, our study, for the first time provided evidence that DEC1 promotes the chemotherapeutic resistance of glioma cells to TMZ via positively regulating the SP1-MGMT axis. Our study also explained the mechanism of TMZ resistance in GBM treatment. We anticipate that our findings may provide new therapeutic and prophylactic target molecules to improve prognosis for patients.

Disclosure of conflict of interest

None.