Skip to main content
NCBI home page
Neuro-Oncology logoLink to Neuro-Oncology
. 2024 Nov 8;27(3):711–726. doi: 10.1093/neuonc/noae237

Glioma–astrocyte connexin43 confers temozolomide resistance through activation of the E2F1/ERCC1 axis

Yanping Gui 1,#, Hongkun Qin 2,#, Xinyu Zhang 3, Qianqian Chen 4, Fangyu Ye 5, Geng Tian 6, Shihe Yang 7, Yuting Ye 8, Di Pan 9, Jieying Zhou 10, Xiangshan Fan 11,✉, Yajing Wang 12,✉, Li Zhao 13,14,✉
PMCID: PMC11889727  PMID: 39514365

Abstract

Background

Glioma is the most prevalent and lethal tumor of the central nervous system. Routine treatment with temozolomide (TMZ) would unfortunately result in inevitable recurrence and therapy resistance, severely limiting therapeutic efficacy. Tumor-associated astrocytes (TAAs) are key components of the tumor microenvironment and increasing evidence has demonstrated that aberrant expression of connexin43 (Cx43) was closely associated with glioma progression and TMZ resistance. However, the specific role of Cx43 in mediating TMZ resistance through glioma and astrocyte interactions has not been fully explored.

Methods

The expression and prognostic value of Cx43 were evaluated in tumor samples and clinical databases. ShRNA-medicated knockdown and Gfap-Cre Cx43flox/flox gene mouse were used to assess the role and functional significance of Cx43 in vitro and in vivo. Moreover, we performed mass spectrometry analysis, chromatin immunoprecipitation, and other biochemical assays to define the molecular mechanisms by which Cx43 promotes TMZ resistance.

Results

We confirmed that the upregulation of Cx43 expression between TAAs and glioma cells contributed to TMZ resistance and tumor recurrence. Genetic knockdown or pharmacological inhibition of Cx43 enhanced TMZ-induced cytotoxicity. Mechanistically, elevated Cx43 expression induced β-catenin accumulation at the cell surface of glioma cells, suppressing T-cell factor/lymphoid enhancer-binding factor transcription. This led to impaired miR-205-5p expression and subsequent activation of the E2F1/ERCC1 axis, which eventually led to chemoresistance.

Conclusions

Our study reveals a novel regulatory mechanism in which the Cx43/miR-205-5p/E2F1/ERCC1 axis contributes to TMZ resistance in glioma. These findings further highlight the potential of targeting Cx43 as a therapeutic strategy in glioma.

Keywords: Connexin 43, drug resistance, ERCC1, glioma, temozolomide

Graphical Abstract

Graphical Abstract.

Graphical Abstract

Open in a new tab

Key Points.

  • Astrocytes promote the expression of Cx43 in glioma cells.

  • Impairing Cx43 in astrocytes and glioma cells enhances temozolomide (TMZ)-induced cytotoxicity.

  • Cx43 contributes to TMZ resistance via β-catenin/miR-205-5p/E2F1/ERCC1 axis.

Importance of the Study.

This study addresses that in the glioma microenvironment, astrocytes interact with glioma and upregulate Cx43 expression in both cell types. Here, we further clarify that genetic knockdown or application of inhibitors effectively resensitizes glioma to temozolomide (TMZ). A key breakthrough of our study is the discovery that dual knockout Gja1 (the gene encoding Cx43) in astrocytes and glioma cells significantly extends the efficacy of TMZ in the murine orthotopic glioma model, thereby prolonging survival time. Mechanistically, elevated expression of Cx43 leads to the accumulation of β-catenin at the glioma cell membrane, which suppresses the transcription of miR-205-5p. This suppression results in increased levels of the transcription factor E2F1 and upregulation of the DNA damage repair protein ERCC1. Ultimately, this molecular pathway impairs the sensitivity of glioma cells to TMZ both in vitro and in vivo.

Glioma is the most common and lethal intracranial tumor in adults, characterized by highly aggressive growth and diffuse infiltration.1 Temozolomide (TMZ) is among the first-line chemotherapy agents routinely used for glioma patients, inducing cytotoxicity by causing DNA double-strand breaks through alkylation.2 Unfortunately, the emergence of TMZ resistance has become an urgent clinical problem in glioma. The most common mechanism of TMZ resistance involves the direct repair of cytotoxic lesions by the DNA repair enzyme O6-methylguanine DNA methyltransferase (MGMT).3 However, patients with low MGMT expression also develop resistance to TMZ.4,5 Deeper insights into MGMT-independent TMZ resistance are, therefore, urgently needed.

A major mechanism underlying the chemotherapeutic resistance in cancer cells is the nucleotide excision repair (NER) pathway, which is responsible for removing alkylating damage.6,7 Increasing evidence implicates that enhanced NER is accountable for acquired TMZ resistance in glioma,8–10 and NER enhancement is driven not only by alterations in endogenous gene expression patterns within tumors but also by exogenous signals from the tumor microenvironment (TME).11–13 Among the various cells in the glioma TME, astrocytes are particularly abundant in the brain and play a crucial role in glioma development, progression, invasion, and therapy resistance.14,15 Studies have revealed that chronic inflammation and hypoxia during the progression transform peripheral astrocytes into reactive astrocytes, known as tumor-associated astrocytes (TAAs). TAAs reciprocally interact with glioma cells, providing essential environmental cues to support glioma propagation and invasion via secretion or direct contact.16,17 However, whether astrocytes directly interact with glioma cells to potentiate TMZ resistance through the NER pathway remains elusive.

Cx43 is widely expressed in both glioma and astrocytes and exhibits increased expression in TAAs.18 As a channel-forming protein crucial for intercellular communication, Cx43 forms gliomas–astrocyte gap junction intercellular communication at the tumor border and regulates the response of glioma cells to TMZ. Overexpression of Cx43 in glioma has been reported to attenuate the efficacy of TMZ in a channel-dependent manner.16,19-21 Additionally, accumulating evidence demonstrates that Cx43-mediated carcinoma–astrocyte gap junction intercellular communication promotes brain metastasis and chemotherapy resistance.18,22–24 Cx43 is also known to confer TMZ resistance in glioma through channel-independent mechanisms and exosomes.25,26 However, it remains unclear how Cx43-mediated interactions between glioma and astrocytes modify the NER pathway-induced TMZ resistance in glioma.

In this study, we investigated the role of Cx43 in mediating the protective effects of astrocytes on glioma TMZ resistance. Glioma cells could activate astrocytes through direct contact, leading to the upregulation of Cx43. Elevated Cx43 expression results in the accumulation of β-catenin at the cell surface, reducing its nuclear translocation and suppressing the transcriptional activity of miR-205-5p by the T-cell factor/lymphoid enhancer-binding factor (TCF/LEF) family, in turn elevating the expression of miR-205-5p target gene E2F1, ultimately leading to TMZ resistance through upregulating the NER protein ERCC1 in glioma cells. Our study suggests that targeting Cx43-mediated interactions between glioma and astrocytes could be a promising therapeutic strategy to enhance the chemotherapeutic efficacy of TMZ against glioma.

Materials and Methods

Primary Glioma Specimens

Human glioma specimens were obtained from the Affiliated Drum Tower Hospital of Nanjing University Medical School between 2016 and 2019. All human glioma specimens used in this study were approved by the ethics committees of the Affiliated Drum Tower Hospital of Nanjing University Medical School (Approval No.2022-129-01), with informed consent from patients or their guardians. Histopathological diagnoses of glioma specimens were performed by pathologists according to the 2016 World Health Organization (WHO) classification.

Cell Lines and Cell Culture

U251, U87, and HEK-293T cell lines were obtained from the Cell Bank of the Chinese Academy of Sciences. Normal human astrocytes (NHAs) were obtained from the American Type Culture Collection. GL261 cell line was kindly gifted from Huashan Hospital Affiliated to Fudan University. Cell lines were authenticated using short tandem repeat profiling. The U251, U87, NHA, and HEK-293T cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Cat# 31800-014) with 10% fetal bovine serum (FBS; Wisent, Cat#086-150). GL261 cells were cultured in DMEM/F-12 medium (Gibco, Cat#12500062) with l-glutamine (Gibco, Cat#2323163) and 10% FBS. All the cells were cultured at 37 °C in a humidified atmosphere with 5% CO2 and were verified to be free of mycoplasma contamination by PCR analysis.

Mice

mGFAP-cre (Stock No. 004600) and Cx43flox (Stock No. 008039) transgenic mice were purchased from the Jackson Laboratory. The Cx43flox line was crossed with the mGFAP-cre line to generate conditional knockout homozygous (Gfapcre-Cx43f/f, Cx43 cKO) mice and littermate controls (Cx43f/f). The genotypes of all mice were determined by PCR analysis of tail genomic DNA with appropriate primers (Supplementary Table 2). Male and female mice were used for all experiments without bias; 6- to 8-week-old female athymic BALB/c nude mice and C57BL/6 mice were purchased from Beijing Vital River Laboratory Animal Technology Co. Ltd. Mice were housed under standard specific pathogen-free conditions, and all animal experiments were performed in accordance with protocols approved by the Animal Welfare and Ethics Committee of China Pharmaceutical University.

Antibodies and Reagents

The source of antibodies and reagents is listed in Supplementary Table 3.

Primary Cells Isolation and Culture

See Supplementary Methods for details.

Cell Coculture Model

Glioma or astrocyte cells were stably transfected with a red fluorescent protein to distinguish them from the direct coculture cells. In the direct coculture experiment, glioma and astrocyte cells were mixed at a 1:1 ratio. For the transwell coculture experiment, glioma and astrocyte cells were cultured separately in the lower and upper chambers of a 0.4 μM transwell apparatus (Millipore, Cat#PIHT30R48). To determine the chemoprotective roles of astrocytes, the overnight cocultured cells were treated with TMZ for 48 h. To investigate the TMZ-resistance reversing effects of carbenoxolone (CBX) and meclofenamate (Meclo), overnight coculture cells were treated with TMZ combined CBX or Meclo for 48 h. Cocultured cells were then subjected to further analysis.

Subcutaneous Xenograft and Intracranial Murine Model

Details regarding subcutaneous xenograft model and intracranial murine model are provided in Supplementary Methods.

Apoptosis Analysis

Apoptotic analysis was performed using an Annexin V-DAPI apoptosis detection kit (Miltenyi, Cat#130-093-060) according to the manufacturer’s instructions. Cells were analyzed using MACSQuant Analyzer 10 (Miltenyi) or CytoFLEX S flow cytometer (Beckman Coulter). Flow cytometry data processing was performed with FlowJo software. TUNEL staining was performed using the TUNEL apoptosis kit (Yeasen, Cat#40306ES60) according to the manufacturer’s instructions. The images were acquired with a fluorescent (DM5200, Leica).

Flow Cytometry Cx43 Detection Assay

Coculture cells were collected and fixed with 4% paraformaldehyde for 20 min. After fixation, the cells were washed twice with phosphate-buffered saline (PBS) and incubated with 5% bovine serum albumin (BSA) for 2 h followed by labeled with Cx43 primary antibody for 2 h at room temperature. Next, cells were washed twice with PBS and labeled with fluorescent conjugate secondary antibody for 1 h at room temperature. Labeled cells were then washed, and analyzed on a MACSQuant Analyzer 10 (Miltenyi) or CytoFLEX S flow cytometer (Beckman Coulter), and the data were analyzed with FlowJo software.

Vector and Vector Construction

pLKO.1-puro, pLenti PGK V5-LUC Neo, lentiCRISPRv2, human beta-catenin pcDNA3, psPAX2, pMD2.G, M50 Super 8x TOPFlash, and M51 Super 8x FOPFlash vectors were purchased from Addgene. pRL-TK, pGL3-basic, and pmirGLO vectors were purchased from Promega. p3xFLag-CMV-14, pEGFP-C1, and pLenti-CMV-BSD vectors were gifted by Prof. Lei Qiang from China Pharmaceutical University. Primer sequences used for vector construction are listed in Supplementary Table 4.

Cell Transfection

The miR-205-5p mimic (miR-205-5p) and the human miRNA negative control (miR-NC) were purchased from GenePharma. Cells were transfected with shRNAs, miRNA mimics, or vectors using Lipofectamine 2000 (Invitrogen, Cat#11668019). For establishing stable transfected cell lines, lentiviral vectors, psPAX2 and pMD2.G (4:3:1), were transfected into HEK293T cells to generate lentiviruses. These lentiviruses were then used to infect target cells. Stable transfected cell lines were selected by culturing the infected cells in media containing either puromycin (2.5 μg/mL), G418 (200 μg/mL), or blasticidin S (10 μg/mL) for 7 days.

Subcellular Protein Extraction

Subcellular proteins were extracted using a subcellular protein extraction kit (Keygen, Cat#KGBSP002), and the fractions were subjected to SDS-PAGE and western blotting.

Immunohistochemistry

For immunohistochemical analysis, sections were deparaffinized and rehydrated through a descending alcohol series, followed by antigens retrieval, and endogenous peroxidase activity blocking. The sections were then incubated with primary antibodies followed by visualization with a two-step process DAB staining kit (ZSGB-BIO, Cat#ZLI-9019). Finally, slides were counterstained with hematoxylin, dehydrated, and mounted. Each specimen was assigned a score based on staining intensity (no staining = 0, weak staining = 1, moderate staining = 2, strong staining = 3) and the proportion of stained cells (0% = 0, 1%–24% = 1, 25%–49% = 2, 50%–74% = 3, 75%–100% = 4). For homogeneous staining, the sum IHC score was calculated by multiplying the intensity score with the extent score. For heterogeneous staining, each component was scored independently, and the scores were summed to obtain the final results. All staining assessments were performed by a pathologist who was blinded to the origin of the samples and the outcomes of the subjects. Positive and negative controls of IHC staining are shown in Supplementary Figure 9.

Immunofluorescence Staining

Details regarding immunofluorescence staining are provided in Supplementary Methods.

RNA Extraction and qRT-PCR

The total RNA of cells and clinical tissues was extracted using TRIzol (Invitrogen, Cat#15596018). cDNA was synthesized with HiScript III 1st Strand cDNA Synthesis Kit (Yeasen, Cat#11121ES60). SYBR Green (Yeasen, Cat#11202ES08) was used for qRT-PCR. Quantification of the miRNA was performed with a stem-loop real-time PCR miRNA kit (Vazyme, Cat#MR101-01). GAPDH and U6 were utilized as reference genes. Sequences of the primers are listed in Supplementary Table 5.

miRNA Sequencing

Total RNA was extracted from control and Cx43 knockdown cells using TRIzol. Each group was prepared with 3 parallel replicates. Further miRNA sequencing detection and analysis were conducted by Sangon Biotech.

Immunoprecipitation

Detailed information regarding immunoprecipitation is provided in Supplementary Methods.

Protein LC-MS/MS Analysis

Flag-Cx43 and empty vector-transfected HEK-293T cells were lysed, and Cx43 interaction proteins were immunoprecipitated using anti-Flag affinity gel (GNI, Cat#GNI4510-FG). The samples were then analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS) (OE Biotech) for the identification of proteins bound with Cx43.

Chromatin Immunoprecipitation

Details regarding chromatin immunoprecipitation (ChIP) assays are provided in Supplementary Methods. Primer sequences are shown in Supplementary Table 4.

Fluorescence-Activated Cell Sorting

Cocultured NHA-RFP and U251/U87 cells were collected and washed by PBS twice and incubated with 5% BSA. Cells were then analyzed on CytoFLEX SRT flow cytometer (Beckman Coulter), and the data were analyzed with FlowJo software.

Public Datasets Analysis

Details regarding data analysis are provided in Supplementary Methods.

Statistical Analysis

Each experiment was repeated at least three times to ensure the reliability of the results. All data were represented by mean ± SD. Significant differences between the groups were estimated by Student’s t-test (2-sided) or 1-way analysis of variance. The Kaplan–Meier curves were used to describe the survival, and the log-rank test was applied to assess the statistical significance between groups. A value of P < .05 was considered statistically significant. All statistical analyses were performed using GraphPad 8.0 (GraphPad Software). The details of the statistical analyses are presented in the figure legends.

Results

Upregulation of Cx43 Is Associated With TMZ Resistance and Glioma–Astrocyte Interactions in the Tumor Microenvironment

Previous studies have shown that Cx43 expression is associated with TMZ resistance in glioma.19,20 To further corroborate the expression profile of Cx43 in glioma, we performed immunohistochemistry (IHC) staining of Cx43 in glioma cases. Cx43 expression was significantly higher in glioma tissues compared with nonglioma tissues and increased with ascending WHO pathological grades of glioma (Figure 1A). Western blotting analysis also showed significant upregulation of Cx43 expression in glioma samples (Figure 1B). In the Chinese Glioma Genome Atlas (CGGA) database, GJA1, the coding gene of Cx43 protein, was found to be expressed at the highest level among all the connexins genes in recurrent glioma with TMZ treatment (Figure 1C). Survival analysis revealed that a higher level of GJA1 was associated with a worse outcome in patients with TMZ treatment (Figure 1D). Similarly, data from the Cancer Genome Atlas Program (TCGA) database showed that high GJA1 expression was positively associated with recurrence and poor prognosis in glioma patients treated with TMZ (Figure 1E and F). Taken together, these results suggested that glioma patients with high Cx43 levels benefited less therapeutic efficacy from TMZ.

Figure 1.

Figure 1.

Open in a new tab

Cx43 is elevated in glioma tissues and associated with TMZ resistance. (A) Representative images of IHC staining and analysis of Cx43 in nonglioma brain and glioma tissues. Normal (n = 20 patients), WHO 2 (n = 20 patients), WHO 3 (n = 20 patients), and WHO 4 (n = 20 patients). (B) Western blotting analysis of Cx43 expression in 7 glioma and corresponding parenchyma adjacent tissues. (C) Bioinformatic analysis of gap junction gene expression in recurrent glioma patients with TMZ treatment in CGGA mRANseq_693 dataset. (D) Kaplan–Meier survival analysis of the overall survival with log-rank test stratified by GJA1 level in glioma patients with TMZ treatment in CGGA mRANseq_693 dataset. The cutoff is the median. (E) Bioinformatic analysis of GJA1 expression in recurrent and nonrecurrent glioma patients with TMZ treatment in TCGA GBM combined LGG database. (F) Kaplan–Meier survival analysis of the overall survival with log-rank test stratified by GJA1 level in glioma patients with TMZ treatment in TCGA GBM combined LGG database. The cutoff is the median. (G) Representative immunofluorescence assays of GFAP expression in NHA cells in a contact coculture model. (H) Representative immunofluorescence assays of Cx43 expression in the U251 + NHA-RFP and U87 + NHA-RFP cell contact coculture models. (I) Flow cytometry assays of Cx43 expression on U251 and U87 cell membrane in a contact coculture model. (J) Representative images of H&E and IHC staining of GFAP and Cx43 in murine orthotopic glioma model. (K) Representative images of Cx43 IHC staining in glioma and paraneoplastic tissue from patient samples. Data were presented as mean ± SD. ***P < .001, **P < .01, and *P < .05 by one-way ANOVA.

Glioma cells have been demonstrated to interact with surrounding astrocytes, transforming them into a reactive status, known as TAAs.27 Consistent with this, our study demonstrated that when red fluorescence-labeled NHAs were cocultured with U251, U87, and CG1 glioma cells, there was a subsequent upregulation of GFAP, a marker of astrocyte reactivity (Figure 1G). Meanwhile, the expression of Cx43, a key gap junction protein in glioma cell membrane, was upregulated when cocultured with astrocytes, especially at the direct-contact sites between the two cell types (Figure 1H and I and Supplementary Figure 1A). To further characterize glioma–astrocyte interactions in vivo, a murine orthotopic glioma model was established. Syngeneic murine glioma GL261 cells infiltrated into the surrounding brain tissues, resulting in increased GFAP and Cx43 protein levels in the glioma-normal tissue contacted region (Figure 1J, Supplementary Figure 1B). Notably, in clinical glioma samples, enhanced expression of Cx43 in glioma tissue was detected in the tumor-normal tissue contact zone (Figure 1K). Moreover, comparative analysis of the single-cell RNA sequencing from the Darmanis dataset28 confirmed that GJA1 was preferentially expressed in the invasive region of gliomas (Supplementary Figure 1C). All these findings suggested a crucial role for increased Cx43 in the glioma microenvironment.

Astrocytes Confer Glioma Cell Resistance to TMZ Related with Cx43

Astrocytes have been implicated as a potentiating factor in sustaining favorable habitat for cancer cell survival in the brain microenvironment.29 To clarify how astrocytes enabled glioma cells’ escape from TMZ cytotoxicity, we established both no-contact and direct-contact coculture models using murine primary astrocytes and glioma cells at a ratio of 1:1, and the results indicated that the apoptotic glioma cells were reduced in both coculture models compared with the mono-culture group (Supplementary Figure 1D), and astrocytes conferred greater protection against glioma apoptosis induced by TMZ in direct-contact model. Similar reduced apoptotic cell proportions of human glioma cells were also observed in flow cytometry apoptosis assay and TUNEL staining (Figure 2A and Supplementary Figure 1E).

Figure 2.

Figure 2.

Open in a new tab

Astrocytes promote TMZ resistance of gliomas via Cx43 in vivo and in vitro. (A) Apoptosis assays of apoptotic U251 and GG1 primary glioma cells in contact coculture model with TMZ treatment for 48 h. (B) Apoptosis assays of U251, U87, and GL261 cells treated with TMZ (1000 μM) combined with CBX or Meclo in a contact coculture model for 48 h. (C–E) Schema of experiment design for the orthotopic model, Kaplan–Meier survival analysis with log-rank test, and bioluminescence imaging of tumor in GL261 murine orthotopic glioma model treated with TMZ and/or Meclo (n = 5 per group). (F) Apoptosis assays of apoptotic glioma cells in Cx43-knockdown U251, U87, and NHA cell contact coculture model with TMZ treatment (1000 μM, 48 h). (G and H) Schema of experiment design for orthotopic model and Kaplan–Meier survival analysis with log-rank test of different groups of syngeneic mice orthotopic glioma model (n = 5 per group). Data were presented as mean ± SD. ***P < .001, **P < .01, and *P < .05 by one-way ANOVA or two-sided Student’s t-test.

Carcinoma–astrocyte gap junctions have been reported previously to facilitate glioma invasion, breast cancer brain metastasis, and melanoma cell chemoresistance.22,30 To expand previous findings, gap junction activity inhibitors CBX and Meclo were used to identify whether gap junction is involved in astrocyte-induced chemoprotection. We demonstrated that treatment with 100 μM of CBX or 10 μM of Meclo alone inhibited gap junction activity without significantly altering the apoptotic ratio of glioma cells and astrocytes (Supplementary Figure 1F and G). Treatment with either CBX or Meclo significantly reversed the chemoprotection effect of astrocytes in the contact coculture model when exposed to TMZ (Figure 2B). Building on the findings of Schneider et al., which demonstrated that Meclo treatment enhanced TMZ sensitivity in primary glioma cells,31 we confirmed that Meclo significantly prolonged the median survival of TMZ-treated mice and rendered the glioma cells more sensitive to TMZ-induced apoptosis (Figure 2C-E and Supplementary Figure 1H). At the same time, no significant toxicity to major organs was observed (Supplementary Figure 1I).

Given the discovery of highly expressed Cx43 in glioma and inhibition of Cx43 had a potential role in TMZ resistance as illustrated above, we intended to investigate whether TMZ resistance induced by glioma–astrocyte gap junctions is related to Cx43. Knockdown of Cx43 in either glioma cells or astrocytes abolished the chemoprotective effects of astrocytes in the coculture model, enhancing the TMZ sensitivity of glioma cells (Supplementary Figure 2A and Figure 2F). To clarify these findings in vivo, we utilized a Gfap-Cre mouse model crossed with Cx43 floxed (Cx43f/f) mice to selectively eliminate Cx43 expression in astrocytes (Supplementary Figure 2B and C). Notably, implantation of Cx43 knockout GL261-luc cells in Gfap-Cre Cx43f/f mice significantly extended survival time and reduced bioluminescence signal, indicating a robust tumor-suppressive effect of TMZ (Figure 2G and H and Supplementary Figure 2A and D). TUNEL staining further showed that knocking out Cx43 in both glioma cells and astrocytes exhibited remarkable enhancement of the therapeutic efficacy of TMZ in vivo (Supplementary Figure 2E). Taken together, these results underscore the crucial role of Cx43 in astrocyte-induced TMZ resistance in glioma cells.

Elevated Cx43 Increases the Membrane Localization of β-Catenin, Resulting in the Suppression of TCF/LEF Transcription

To further specify the role of Cx43 in TMZ resistance, a Flag-Cx43 vector was transfected into HEK-293T cells, followed by immunoprecipitation assays and LC-MS/MS to identify related proteins interacting with Cx43. Notably, β-catenin was identified as one of the Cx43 interaction partners (Figure 3A and Supplementary Figure 3A). Coincident with the finding in human prostate cancer cells and murine mammary glands,32 we further verified the interaction between Cx43 and β-catenin in glioma cells by coimmunoprecipitation and immunofluorescence assays (Figure 3B and C). Since the cytosolic C-terminal tail of Cx43 serves as a platform for diverse protein-protein interactions,33 we constructed GFP-Cx43-CT and GFP-Cx43-ΔCT vectors. To narrow down the interaction domains of β-catenin, 3 nonoverlapping fragment vectors were constructed and subjected to coimmunoprecipitation analysis (Figure 3D). As indicated, the C-terminal domain of Cx43 was identified as the dominant one to which β-catenin was directly bound, and Cx43 interacted primarily with the C-terminal domain of β-catenin (Figure 3E). Immunofluorescence staining showed strong colocalization of ectopically expressed Cx43-CT with β-catenin-CT (Supplementary Figure 3B).

Figure 3.

Figure 3.

Open in a new tab

Cx43 interacts with β-catenin and regulates its localization and downstream TCF/LEF transcriptional activation. (A) Representative proteins determined by LC-MS/MS and β-catenin peptide mass spectrum by product ion scan. (B and C) Co-IP and IF assays of Cx43 and β-catenin in U251 and GG1 primary glioma cells. (D) Schematic illustration of GFP-tagged full-length Cx43 (FL), Flag-tagged full-length β-catenin (FL), and corresponding functional domain truncations. (E) Co-IP interaction assays of β-catenin and Cx43 functional domain truncations in HEK-293T cells. (F) TOP-/FOP-flash luciferase reporter assays in Cx43 transfected U251 and U87 cells. (G) TOP-/FOP-flash luciferase reporter assays in Cx43-knockdown and Cx43 truncations transfected HEK-293T cells. (H) Western blotting analysis of β-catenin distribution in Cx43 transfected HEK-293T cells. (I) Representative immunofluorescence images of β-catenin distribution in Cx43 transfected U251 and U87 cells. Data were presented as mean ± SD. ***P < .001, **P < .01, and *P < .05 by one-way ANOVA or two-sided Student’s t-test.

Interestingly, there were no significant changes in the mRNA and protein levels of β-catenin in Cx43-overexpressing glioma cells or HEK-293T cells (Supplementary Figure 3C and D). However, the TOP/FOP ratios were significantly reduced in Cx43-overexpressing glioma cells, reflecting decreased TCF/LEF transcriptional activity (Figure 3F). This activity was significantly increased when Cx43 was knocked down and was restored by re-expression of Cx43 (Figure 3G). Furthermore, immunofluorescence and western blotting analyses revealed that Cx43 mainly colocalized with β-catenin at the cell membrane, and the distribution of β-catenin in the nucleus was reduced (Figure 3H and I). These results suggested that Cx43 could trap β-catenin at the cell membrane, inhibiting its nuclear translocation and consequently preventing its downstream TCF/LEF transcriptional activity.

miR-205-5p/E2F1 Is Transcriptional Regulated by Cx43/β-Catenin Axis

Several studies have demonstrated that β-catenin, the primary downstream effector of the canonical Wnt pathway, can regulate chemosensitivity to TMZ in glioma cells.34 Our study found that increased Cx43 can trap β-catenin, preventing its nuclear translocation and promoting TMZ resistance. This indicated that inhibition of the β-catenin pathway caused by glioma–astrocyte Cx43 gap junctions may be involved in regulating TMZ sensitivity, but the underlying mechanism remains unclear. miRNAs have been proven to be capable of transferring from glioma cells to astrocytes in a Cx43-dependent manner.22 As posttranscriptional repressors, miRNAs exhibit profound regulatory effects by potently repressing their target genes.

To determine whether miRNAs were involved in the Cx43/β-catenin axis, we performed a miRNA high-throughput sequencing on Cx43 knockdown U251 cells and corresponding control cells; 19 miRNAs were significantly upregulated, whereas 35 miRNAs were significantly downregulated in Cx43 knockdown U251 cells (Figure 4A and Supplementary Table 1). Among them, miR-205-5p was significantly upregulated and has been reported to have a critical role in a variety of tumors. RT-qPCR results validated that miR-205-5p was upregulated in Cx43 knockdown U251 and CG1 cells, and vice versa (Figure 4B). Also, miR-205-5p levels in parenchyma tissues were significantly higher than those in glioma specimens and were significantly lower in recurrent patients than that in primary glioma patients in the CGGA database (Supplementary Figure 4A and B), suggesting that miR-205-5p may involve in chemotherapy resistance of glioma. Further investigation showed that Cx43 significantly reduced both pri-MIR205 and pre-MIR205 (Supplementary Figure 4C and D) and Cx43 overexpression significantly attenuated the promoter transcriptional activity of MIR205HG. β-Catenin significantly enhanced promoter transcriptional activity, which could be rescued by restoration of Cx43 (Figure 4C). Additionally, analyses of the potential binding regions of TCF/LEF transcriptional factors like TCF-1, TCF-3, TCF-4, and LEF-1 on this promoter in the JASPAR and PROMO databases with construction of serial luciferase reporter vectors containing MIR205HG promoter sequences of various length demonstrated that −220~+10 was the exact region where the TCF/LEF transcriptional factors bind to it mainly (Figure 4D).

Figure 4.

Figure 4.

Open in a new tab

Cx43/β-catenin axis regulates miR205-5p and its target gene E2F1. (A) Heatmap and relative level of differentially expressed miRNAs from 3 independent samples in Cx43-knockdown U251 cells and control cells by HGS. (B) RT-qPCR analysis of miR-205-5p expression in Cx43-knockdown and Cx43-overexpressed U251 cells. (C) Relative MIR205HG promoter luciferase activity in Cx43, β-catenin, and Cx43 + β-catenin transfected HEK-293T and CG1 cells. (D) Predicted TCF/LEF binding sites and transcription activity of different regions of MIR205HG promoter in β-catenin transfected HEK-293T cells. (E) Venn diagram of predicted miR-205-5p target genes in TarBase, miRWalk, ENCORI, TargetScan, and MiRDB databases. (F) Relative luciferase activity of wt-E2F1-3’-UTR and mut-E2F1-3’-UTR in response to miR-205-5p or miR-NC in HEK-293T cells. (G) Western blotting analysis of E2F1 protein expression in Cx43 and miR-205-5p transfected U251 and U87 cells. (H) Representative images of IHC staining and analysis of E2F1 in nonglioma brain and glioma tissues. Normal (n=20 patients), WHO 2 (n=20 patients), WHO 3 (n=20 patients), and WHO 4 (n=20 patients). (I) Kaplan–Meier survival analysis of the overall survival with log-rank test stratified by E2F1 level in glioma patients with TMZ treatment in CGGA mRANseq_693 dataset. The cutoff is the median. (J) Apoptosis assays of apoptotic glioma cells in miR-205-5p and E2F1 transfected U251 and U87 cells treatment with TMZ (1000 μM, 48 h) in contact coculture model. (K) Tumor volume of U251 subcutaneous xenografts with different miR-205-5p and E2F1 levels and combined with or without TMZ (25 mg/kg, ig, qod) treatment. Data were presented as mean ± SD. ***P < .001, **P < .01, and *P < .05 by one-way ANOVA or two-sided Student’s t-test.

To uncover the molecular mechanisms of miR-205-5p, we intersected several miRNA target gene prediction databases and noted that E2F1 could be one of the potential target genes of miR-205-5p (Figure 4E). As indicated in Figure 4F, miR-205-5p significantly reduced the luciferase activity of wild-type E2F1 3’-UTR, while having no significant effect on mutant E2F1 3’-UTR activity. Moreover, overexpression of miR-205-5p significantly decreased E2F1 expression (Supplementary Figure 4E). E2F1 expression was robustly increased in Cx43 overexpression cells, which could be reversed by transfection with miR-205-5p mimic (Figure 4G). These results indicated that E2F1 expression might be regulated by the Cx43/miR-205-5p axis.

Our previous study in 2021 demonstrated that E2F1 is highly expressed in both clinical glioma samples and glioma cell lines.35 In the present study, using a large cohort of clinical samples, we further confirmed that E2F1 expression was increased with the ascending WHO pathological grade of glioma (Figure 4H). Meanwhile, the E2F1 gene expression profiles and clinical glioma data derived from the CGGA database also suggested that it was significantly higher in recurrent patients than in primary patients (Supplementary Figure 4F), and the Kaplan–Meier survival analysis also showed that patients with low E2F1 expression significantly benefit from TMZ chemotherapy (Figure 4I), which demonstrated that E2F1 might play a crucial role in glioma TMZ resistance.

To identify the roles of miR-205-5p and its target gene E2F1 in astrocytes-mediated TMZ resistance in glioma, glioma cells cocultured with NHA cells were detected. As shown in Figure 4J, glioma cells with overexpressed E2F1 were more resistant upon TMZ treatment, whereas the increased apoptosis was significantly reversed in miR-205-5p overexpressed cells. Consistent with the in vitro results, E2F1 significantly attenuated the chemotherapy sensitivity enhanced by miR-205-5p in vivo (Figure 4K and Supplementary Figure 4G and H). Moreover, IHC staining of xenograft tumors also suggested that E2F1 expression could be regulated by miR-205-5p (Supplementary Figure 4I). The data above supported that E2F1 was a direct and functional target of miR-205-5p, conferring astrocytes-mediated TMZ resistance in glioma.

E2F1 Confer TMZ Resistance Through Upregulate ERCC1 Expression

In addition to O6-MGMT activity, enhanced NER capacity and other DNA repair pathways have been reported to contribute to TMZ resistance in glioma.8,36 Considering that ERCC1, the key protein of the NER pathway, forms a heterodimer with XPF to cleave damaged DNA strands during DNA repair, we were curious whether ERCC1 is implicated in Cx43-induced TMZ resistance. Notably, in recurrent glioma treated with TMZ in the CGGA database, ERCC1 had the highest expression level among other ERCC family genes and positively correlated with poor prognosis (Figure 5A and B). Furthermore, the mRNA and protein level of ERCC1 was significantly downregulated compared with other ERCC family genes upon transfected with ShGJA1-2 in U251 and U87 cells (Figure 5C and Supplementary Figure 5A). To this end, IHC staining in glioma samples also demonstrated that ERCC1 expression increased with ascending WHO pathological grade (Figure 5D). To evaluate the repair capacity of ERCC1, glioma cells were treated with TMZ at a temporal gradient. DNA damage marker γ-H2AX and ERCC1 both increased in a time-dependent manner, indicating that the repair capacity of ERCC1 gradually increased during TMZ-induced DNA damage in glioma (Figure 5E). All the results indicated that ERCC1 might contribute to TMZ resistance.

Figure 5.

Figure 5.

Open in a new tab

E2F1 mediated TMZ resistance in glioma through transcription of ERCC1. (A) Bioinformatic analysis of ERCC family genes expression in recurrent glioma patients with TMZ treatment in CGGA mRANseq_325 dataset. (B) Kaplan–Meier survival analysis of the overall survival with log-rank test stratified by ERCC1 level in glioma patients with TMZ treatment in CGGA mRANseq_325 dataset. The cutoff is the median. (C) RT-qPCR analysis of ERCC family expression in Cx43-knockdown U251 and U87 cells. (D) Representative images of IHC staining and analysis of ERCC1 in nonglioma brain and glioma tissues. Normal (n = 20 patients), WHO 2 (n = 20 patients), WHO 3 (n = 20 patients), and WHO 4 (n = 20 patients). (E) Western blotting analysis of γ-H2AX and ERCC1 expression in U251 and U87 cells treatment with TMZ (500 μM) for 0, 0.5, 1, and 3 h. (F) Schematic representation of E2F1 motif and construction of ERCC1 promoter dual-luciferase reporter vector. (G) Relative luciferase activity of wt-ERCC1 promoter and mut-ERCC1 promoter transfected U251 and U87 cells upon transfection of E2F1 expression vector or empty vector. (H) ChIP assays of E2F1 enrichment on the ERCC1 promoter region (−1516 nt~ −1526 nt) in U251 and U87 cells. (I) Apoptosis assays of glioma cells in E2F1 and shERCC1 transfected U251 and U87 cells treated with TMZ (1000 μM, 48 h) in contact coculture model. (J-L) Tumor volume (J), photographs (K), and tumor weight (L) of U251 subcutaneous xenografts with different ERCC1 and E2F1 levels treated with or without TMZ (25 mg/kg, ig, qod). Data were presented as mean ± SD. ***P < .001, **P < .01, and *P < .05 by 1-way ANOVA or two-sided Student’s t-test.

According to the TCGA GBM (glioblastoma) combined LGG (low grade glioma) database, the transcription factor E2F1 showed a strong correlation with ERCC1 (Supplementary Figure 5B). Moreover, we performed bioinformatics analysis of the promoter region of the ERCC1 gene and predicted the most putative DNA binding sites for E2F1 in the JASPAR and PROMO database (Figure 5F). Data in Figure 5G revealed that transfection of E2F1 significantly upregulated ERCC1 promoter activities. Consistent with this finding, ChIP assays demonstrated that E2F1 recruited to the −1526~−1516 region of ERCC1 promoter (Figure 5H). Together, these results suggested that ERCC1 expression was regulated by the E2F1.

In the contact coculture model, disruption of ERCC1 expression significantly compromised the protective effect of astrocytes followed by TMZ treatment, while E2F1 restoration mitigated the process (Figure 5I and Supplementary Figure 5C). Consistent with our in vitro findings, analysis of the tumor size (Figure 5J and K), tumor weight (Figure 5L), and apoptotic xenograft cells (Supplementary Figure 5D) in a subcutaneous xenograft model demonstrated that disrupting ERCC1 expression in E2F1-overexpressing cells markedly restored sensitivity to chemotherapy. Moreover, IHC staining also suggested that ERCC1 expression was regulated by E2F1 (Supplementary Figure 5E). These data suggested E2F1 leading to TMZ resistance in glioma cells through ERCC1.

Cx43 Potentiates TMZ Resistance Through miR-205-5p/E2F1/ERCC1 Axis

To verify whether the miR-205-5p/E2F1/ERCC1 axis is involved in the maintenance of the TMZ resistance mediated by Cx43, we employed a knockdown approach targeting GJA1. Subsequently, downregulation of both E2F1 and ERCC1 was observed (Figure 6A and B and Supplementary Figure 6A). Disruption of Cx43 expression significantly enhanced the apoptosis and DNA damage of glioma cells by TMZ treatment, whereas ERCC1 restoration mitigated the apoptosis and DNA damage (Figure 6C and D and Supplementary Figure 6B-E). Consistent with the results in vitro, the tumor size (Figure 6E and F), tumor weight (Figure 6G), and apoptotic xenograft cells (Figure 6H) showed that overexpressing ERCC1 in Cx43 disrupting cells significantly restored chemotherapy sensibility. Moreover, IHC staining also suggested that E2F1 and ERCC1 expression was regulated by Cx43 (Supplementary Figure 7A). Meanwhile, miR-205-5p dramatically inhibited E2F1 and ERCC1 protein expression in both cytoplasm and nucleus (Supplementary Figure 7B and C). The restoration of E2F1 effectively rescued the inhibitory effect of miR-205-5p on ERCC1 in vivo and in vitro (Supplementary Figure 7D-F). Moreover, western blotting analysis of fluorescence-activated cell sorting-sorted glioma cells in the direct-contact coculture model with astrocytes revealed significantly induced Cx43, E2F1, and ERCC1 protein levels compared with noncocultured glioma cells (Supplementary Figure 8A and B). Correlation analysis of Cx43/E2F1/ERCC1 expression and clinical parameters in glioma samples revealed that high Cx43/E2F1/ERCC1 was associated with ki67 positivity and WHO grade (Supplementary Table 6-8). These data suggested a novel regulatory mechanism of Cx43/miR-205-5p/E2F1 leading to TMZ resistance in glioma cells through ERCC1.

Figure 6.

Figure 6.

Open in a new tab

The Cx43/E2F1/ERCC1 axis promotes TMZ resistance in glioma cells. (A) Representative immunofluorescence assays of E2F1 and ERCC1 expression in shGJA1-2 transfected U251 and U87 cells. (B) Western blotting analysis of the protein expression of Cx43, E2F1, and ERCC1 in U251 and U87 cells with stable knockdown of GJA1. (C) Western blotting analysis of the protein expression of Cx43, E2F1, and ERCC1 in U251 and U87 cells with stable knockdown of GJA1 and overexpression of ERCC1. (D) Apoptosis assays of glioma cells in ShGJA1-2 and ERCC1 transfected U251 and U87 cells treated with TMZ (1000 μM, 48 h). (E-H) Tumor volume (E), photographs (F), tumor weight (G), and representative images of TUNEL assays (H) of U251 subcutaneous xenografts with different Cx43 and ERCC1 levels and combined with or without TMZ (25 mg/kg, ig, qod) treatment. Data were presented as mean ± SD. ***P < .001, **P < .01, *P < .05, and n.s. no significance by 1-way ANOVA or 2-sided Student’s t-test.

Discussion

Astrocytes represent a major cell type in the glioma ecology, and glioma-activated astrocytes are pivotal for tumor proliferation, invasion, and therapeutic resistance. In this study, we revealed that astrocytes interact with glioma cells through direct contact to enhance DNA repair and induce TMZ resistance. As illustrated in a model proposed in the graphical abstract, Cx43 is upregulated in glioma and activated astrocytes, largely contributing to the tumor-protective effects of astrocytes and attenuating the chemotherapeutic efficacy of TMZ. The C-terminus of Cx43 interacts with the C-terminus of β-catenin, sequestering β-catenin at the cell surface and preventing its nuclear localization. This interaction inhibits TCF/LEF-mediated transactivation of miR-205-5p, which in turn activates the expression of the target gene E2F1. Activated E2F1 eventually confers TMZ resistance to glioma cells through the translational activation of ERCC1, which mediates NER activity. Our work provides new insights into the role of astrocytes in supporting the malignant phenotype of glioma and highlights Cx43 as a critical molecular link mediating glioma–astrocyte interactions, with therapeutic potential as a TMZ-sensitizing target.

β-Catenin functions not only as a key transducer of the canonical Wnt/β-catenin pathway but also as a mediator of adherent junction protein complexes. A hallmark of the Wnt/β-catenin pathway is the dramatic nuclear translocation of β-catenin and the formation of complexes with TCF/LEF family transcription factors to activate downstream genes. Emerging evidence suggests that the primary mechanism regulating β-catenin localization involves protein interactions rather than shuttling. This is supported by the findings that E-cadherin, N-cadherin, as well as AKT1 sequester β-catenin at the cell surface, thereby abrogating TCF/LEF transcription.37,38 It has been reported that knockdown of Cx43 disrupts the Cx43/β-catenin junctional complex in the membrane of neural progenitor cells, leading to an increase in β-catenin nuclear translocation, transcription of pro-neuronal genes, and a higher number of neurons.39 This is supported by our findings that elevated Cx43 in glioma cells traps β-catenin at the cell surface, subsequently reducing the nuclear translocation of β-catenin and suppressing the TCF/LEF-mediated transcriptional activity of miR-205-5p. Furthermore, our results verified that the interaction of Cx43 and β-catenin is mediated by their respective C-terminus.

Accumulating evidence indicates that abnormal expression of miRNAs is closely associated with the tumorigenesis of glioma through the regulation of target mRNAs, making them valuable molecular biomarkers for diagnosis, prognosis, and therapy response. In previous studies, miR-205-5p has been identified as an inhibitory miRNA in various cancer types.40 Here, our functional experiments in vitro and in vivo demonstrated that miR-205-5p acts as a tumor-suppressive miRNA against astrocyte-induced TMZ resistance in glioma. Studies have identified the important functions of miRNAs and showed that aberrant miRNA expression patterns in glioma can be rescued using target delivery strategies.41 However, it is necessary to identify the transcription factors that regulate miRNA expression to restore these miRNAs to their normal expression levels. The transcriptional regulation of miR-205-5p remains poorly understood either. Some studies have reported that miR-205-5p can be transcriptionally regulated by p53, p63, and YB1,42,43 yet how β-catenin and TCF/LEF family transcription factors regulate miR-205-5p transcription have not been adequately explored. Herein, our data confirmed that β-catenin is essential for Cx43-regulated miR-205-5p transcription and identified the specific binding region for TCF/LEF family transcription factors that promote miR-205-5p transcription.

The primary mechanism for TMZ resistance involves MGMT-mediated direct DNA repair, which removes the genotoxic O6-methylguanine adduct, and MGMT expression has consistently been inversely correlated with tumor chemosensitivity. Recent research devoted to the mechanisms of TMZ chemoresistance has illustrated that multiple DNA damage repair mechanisms operate independently of MGMT expression, affecting cellular sensitivity to TMZ. This indicates that DNA damage repair should not be viewed merely as an intrinsic property of tumor cells but also as a biological process that can be modulated by the microenvironment.44,45 Emerging studies imply that enhanced ERCC1, a key player in the NER pathway and part of the ERCC1/XPF endonuclease complex, is a significant barrier to effective chemo-treatment, including treatments with platinum, paclitaxel, and fluorouracil.46,47 Moreover, ERCC1 has been widely used as a predictive marker for the efficacy of cisplatin-based regimens in nonsmall cell lung cancer patients. Our study reveals that ERCC1 is highly expressed in glioma and underscores its critical role in glioma therapy.

In summary, our results unveiled that the astrocytes in glioma microenvironment upregulated Cx43 on glioma cells, triggering the activation of E2F1 through the β-catenin/miR-205-5p axis. This activation subsequently directed the transcription of the downstream protein ERCC1 to promote DNA repair, thereby mitigating TMZ-induced cytotoxicity both in vitro and in vivo. However, it is important to note that other mechanisms involving Cx43, aside from the miR-205-5p/E2F1/ERCC1 axis, may also contribute to the development of TMZ resistance. Additionally, the precise signaling pathways or mediators responsible for the upregulation of Cx43 remain unclear, warranting further investigation. Moreover, the glioma samples utilized in this study, sourced from clinical and bioinformatics databases, were not specifically categorized, such as GBM IDH-wild type. Additionally, the Kaplan–Meier analyses conducted within the CGGA database were not strictly confined to a consistent dataset, which may limit the clinical relevance of our findings. Nevertheless, our study highlights an unexpected role of Cx43 in glioma TMZ resistance and sheds light on the possibility of designing novel small molecular inhibitors of Cx43 as a potential strategy to combat TMZ resistance.

Supplementary material

Supplementary material is available online at Neuro-Oncology (https://academic.oup.com/neuro-oncology).

noae237_suppl_Supplementary_Materials
noae237_suppl_supplementary_materials.docx (10.1MB, docx)

Acknowledgments

We thank the help given by Jia Li from the Pathology and PDX Efficacy Evaluation Center, China Pharmaceutical University for technical guidance. Also, we thank Prof. Lei Qiang from China Pharmaceutical University for gifting plasmids. Further, we acknowledge pathologist Dr. Zhiwen Li from the Affiliated Drum Tower Hospital of Nanjing University Medical School for the glioma samples collection and immunoreactive score.

Contributor Information

Yanping Gui, School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University, Nanjing, P.R. China.

Hongkun Qin, School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University, Nanjing, P.R. China.

Xinyu Zhang, School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University, Nanjing, P.R. China.

Qianqian Chen, School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University, Nanjing, P.R. China.

Fangyu Ye, School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University, Nanjing, P.R. China.

Geng Tian, School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University, Nanjing, P.R. China.

Shihe Yang, Public Experimental Platform, China Pharmaceutical University, Nanjing, P.R. China.

Yuting Ye, Public Experimental Platform, China Pharmaceutical University, Nanjing, P.R. China.

Di Pan, The High Efficacy Application of Natural Medicinal Resources Engineering Center of Guizhou Province, School of Pharmaceutical Sciences, Guizhou Medical University, Guiyang, P.R. China.

Jieying Zhou, Department of Chemistry and Biochemistry, Florida International University, Miami, Florida.

Xiangshan Fan, Department of Pathology, Affiliated Nanjing Drum Tower Hospital of Nanjing University School of Medicine, Nanjing, P.R. China.

Yajing Wang, School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University, Nanjing, P.R. China.

Li Zhao, Public Experimental Platform, China Pharmaceutical University, Nanjing, P.R. China; School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University, Nanjing, P.R. China.

Conflict of interest statement. None declared.

Funding

This work was supported by the program of the National Natural Science Foundation of China [grant numbers 81773774, 81872903, and 82273962].

Author contributions

L.Z., Y.J.W., H.K.Q, and Y.P.G. designed the research; Y.P.G., H.K.Q, X.Y.Z, Q.Q.C., F.Y.Y, and G.T. performed experimental work; S.H.Y., Y.T.Y. D.P., and X.S.F. provided methodology instructions and formal analysis. X.S.F. provided specimen resources. H.K.Q, Y.P.G., X.Y.Z, and Q.Q.C. analyzed the data. Y.P.G. and H.K.Q. wrote the manuscript. L.Z., Y.J.W., and J.Y.Z. reviewed and edited the manuscript.

Data availability

All relevant data supporting the findings of this study are available within the paper and supplementary information files, and other data that support this study are available from the corresponding author upon reasonable request.

References

  • 1. Low JT, Ostrom QT, Cioffi G, et al. Primary brain and other central nervous system tumors in the United States (2014-2018): a summary of the CBTRUS statistical report for clinicians. Neurooncol Pract. 2022;9(3):165–182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Mannas JP, Lightner DD, Defrates SR, Pittman T, Villano JL.. Long-term treatment with temozolomide in malignant glioma. J Clin Neurosci. 2014;21(1):121–123. [DOI] [PubMed] [Google Scholar]
  • 3. Zhang J, Stevens MF, Bradshaw TD.. Temozolomide: mechanisms of action, repair and resistance. Curr Mol Pharmacol. 2012;5(1):102–114. [DOI] [PubMed] [Google Scholar]
  • 4. Yi GZ, Huang G, Guo M, et al. Acquired temozolomide resistance in MGMT-deficient glioblastoma cells is associated with regulation of DNA repair by DHC2. Brain. 2019;142(8):2352–2366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Zhang X, Li T, Yang M, et al. Acquired temozolomide resistance in MGMT(low) gliomas is associated with regulation of homologous recombination repair by ROCK2. Cell Death Dis. 2022;13(2):138. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Kara A, Özgür A, Nalbantoğlu S, Karadağ A.. DNA repair pathways and their roles in drug resistance for lung adenocarcinoma. Mol Biol Rep. 2021;48(4):3813–3825. [DOI] [PubMed] [Google Scholar]
  • 7. Duan M, Ulibarri J, Liu KJ, Mao P.. Role of nucleotide excision repair in cisplatin resistance. Int J Mol Sci. 2020;21(23):9248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Boccard SG, Marand SV, Geraci S, et al. Inhibition of DNA-repair genes Ercc1 and Mgmt enhances temozolomide efficacy in gliomas treatment: a pre-clinical study. Oncotarget. 2015;6(30):29456–29468. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Nagel ZD, Kitange GJ, Gupta SK, et al. DNA repair capacity in multiple pathways predicts chemoresistance in glioblastoma multiforme. Cancer Res. 2017;77(1):198–206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Montaldi AP, Lima SCG, Godoy P, Xavier DJ, Sakamoto-Hojo ET.. PARP1 inhibition sensitizes temozolomide-treated glioblastoma cell lines and decreases drug resistance independent of MGMT activity and PTEN proficiency. Oncol Rep. 2020;44(5):2275–2287. [DOI] [PubMed] [Google Scholar]
  • 11. McLaughlin M, Patin EC, Pedersen M, et al. Inflammatory microenvironment remodelling by tumour cells after radiotherapy. Nat Rev Cancer. 2020;20(4):203–217. [DOI] [PubMed] [Google Scholar]
  • 12. Mondal A, Kumari Singh D, Panda S, Shiras A.. Extracellular vesicles as modulators of tumor microenvironment and disease progression in glioma. Front Oncol. 2017;7:144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Scanlon SE, Glazer PM.. Multifaceted control of DNA repair pathways by the hypoxic tumor microenvironment. DNA Repair (Amst). 2015;32:180–189. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Henrik Heiland D, Ravi VM, Behringer SP, et al. Tumor-associated reactive astrocytes aid the evolution of immunosuppressive environment in glioblastoma. Nat Commun. 2019;10(1):2541. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Zhang H, Zhou Y, Cui B, Liu Z, Shen H.. Novel insights into astrocyte-mediated signaling of proliferation, invasion and tumor immune microenvironment in glioblastoma. Biomed Pharmacother. 2020;126:110086. [DOI] [PubMed] [Google Scholar]
  • 16. Chen W, Wang D, Du X, et al. Glioma cells escaped from cytotoxicity of temozolomide and vincristine by communicating with human astrocytes. Med Oncol. 2015;32(3):43. [DOI] [PubMed] [Google Scholar]
  • 17. Yu T, Wang X, Zhi T, et al. Delivery of MGMT mRNA to glioma cells by reactive astrocyte-derived exosomes confers a temozolomide resistance phenotype. Cancer Lett. 2018;433:210–220. [DOI] [PubMed] [Google Scholar]
  • 18. Sin WC, Aftab Q, Bechberger JF, et al. Astrocytes promote glioma invasion via the gap junction protein connexin43. Oncogene. 2016;35(12):1504–1516. [DOI] [PubMed] [Google Scholar]
  • 19. Gielen PR, Aftab Q, Ma N, et al. Connexin43 confers temozolomide resistance in human glioma cells by modulating the mitochondrial apoptosis pathway. Neuropharmacology. 2013;75(1):539–548. [DOI] [PubMed] [Google Scholar]
  • 20. Munoz JL, Rodriguez-Cruz V, Greco SJ, et al. Temozolomide resistance in glioblastoma cells occurs partly through epidermal growth factor receptor-mediated induction of connexin 43. Cell Death Dis. 2014;5(3):e1145–e1145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Pridham KJ, Shah F, Hutchings KR, et al. Connexin 43 confers chemoresistance through activating PI3K. Oncogenesis. 2022;11(1):2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. McCutcheon S, Spray DC.. Glioblastoma-astrocyte connexin 43 gap junctions promote tumor invasion. Mol Cancer Res. 2022;20(2):319–331. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Brandao M, Simon T, Critchley G, Giamas G.. Astrocytes, the rising stars of the glioblastoma microenvironment. Glia. 2019;67(5):779–790. [DOI] [PubMed] [Google Scholar]
  • 24. Jaraíz-Rodríguez M, Talaverón R, García-Vicente L, et al. Connexin43 peptide, TAT-Cx43266-283, selectively targets glioma cells, impairs malignant growth, and enhances survival in mouse models in vivo. Neuro Oncol. 2020;22(4):493–504. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Zhu S, Yang N, Guan Y, et al. GDF15 promotes glioma stem cell-like phenotype via regulation of ERK1/2–c-Fos–LIF signaling. Cell Death Discov. 2021;7(1):1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Murphy SF, Varghese RT, Lamouille S, et al. Connexin 43 inhibition sensitizes chemoresistant glioblastoma cells to temozolomide. Cancer Res. 2016;76(1):139–149. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Guan X, Hasan MN, Maniar S, Jia W, Sun D.. Reactive astrocytes in glioblastoma multiforme. Mol Neurobiol. 2018;55(8):6927–6938. [DOI] [PubMed] [Google Scholar]
  • 28. Darmanis S, Sloan SA, Croote D, et al. Single-cell RNA-Seq analysis of infiltrating neoplastic cells at the migrating front of human glioblastoma. Cell Rep. 2017;21(5):1399–1410. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29. Mega A, Hartmark Nilsen M, Leiss LW, et al. Astrocytes enhance glioblastoma growth. Glia. 2020;68(2):316–327. [DOI] [PubMed] [Google Scholar]
  • 30. Chen Q, Boire A, Jin X, et al. Carcinoma-astrocyte gap junctions promote brain metastasis by cGAMP transfer. Nature. 2016;533(7604):493–498. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Schneider M, Vollmer L, Potthoff AL, et al. Meclofenamate causes loss of cellular tethering and decoupling of functional networks in glioblastoma. Neuro Oncol. 2021;23(11):1885–1897. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Sorgen PL, Trease AJ, Spagnol G, Delmar M, Nielsen MS.. Protein-protein interactions with connexin 43: regulation and function. Int J Mol Sci. 2018;19(5):1428. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Leithe E, Mesnil M, Aasen T.. The connexin 43 C-terminus: a tail of many tales. Biochim Biophys Acta Biomembr. 2018;1860(1):48–64. [DOI] [PubMed] [Google Scholar]
  • 34. Wickstrom M, Dyberg C, Milosevic J, et al. Wnt/beta-catenin pathway regulates MGMT gene expression in cancer and inhibition of Wnt signalling prevents chemoresistance. Nat Commun. 2015;6(1):8904. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Qin H, Gui Y, Ma R, et al. miR-1258 attenuates tumorigenesis through targeting E2F1 to inhibit PCNA and MMP2 transcription in glioblastoma. Front Oncol. 2021;11:671144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Garnier D, Meehan B, Kislinger T, et al. Divergent evolution of temozolomide resistance in glioblastoma stem cells is reflected in extracellular vesicles and coupled with radiosensitization. Neuro Oncol. 2017;20(2):236–248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Osuka S, Zhu D, Zhang Z, et al. N-cadherin upregulation mediates adaptive radioresistance in glioblastoma. J Clin Invest. 2021;131(6):e136098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Gao SP, Kiliti AJ, Zhang K, et al. AKT1 E17K inhibits cancer cell migration by abrogating beta-catenin signaling. Mol Cancer Res. 2021;19(4):573–584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Rinaldi F, Hartfield EM, Crompton LA, et al. Cross-regulation of connexin43 and beta-catenin influences differentiation of human neural progenitor cells. Cell Death Dis. 2014;5(1):e1017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Xiao Y, Humphries B, Yang C, Wang Z.. MiR-205 dysregulations in breast cancer: the complexity and opportunities. Noncoding RNA. 2019;5(4):53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Pottoo FH, Javed MN, Rahman JU, Abu-Izneid T, Khan FA.. Targeted delivery of miRNA based therapeuticals in the clinical management of glioblastoma multiforme. Semin Cancer Biol. 2021;69:391–398. [DOI] [PubMed] [Google Scholar]
  • 42. Piovan C, Palmieri D, Di Leva G, et al. Oncosuppressive role of p53-induced miR-205 in triple negative breast cancer. Mol Oncol. 2012;6(4):458–472. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Liu X, Chen D, Chen H, et al. YB1 regulates miR-205/200b-ZEB1 axis by inhibiting microRNA maturation in hepatocellular carcinoma. Cancer Commun (Lond). 2021;41(7):576–595. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44. Meng X, Duan C, Pang H, et al. DNA damage repair alterations modulate M2 polarization of microglia to remodel the tumor microenvironment via the p53-mediated MDK expression in glioma. EBioMedicine. 2019;41:185–199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45. Saitoh T, Oda T.. DNA damage response in multiple myeloma: the role of the tumor microenvironment. Cancers (Basel). 2021;13(3):504. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Kwon H-C, Roh M, Oh S, et al. Prognostic value of expression of ERCC1, thymidylate synthase, and glutathione S-transferase P1 for 5-fluorouracil/oxaliplatin chemotherapy in advanced gastric cancer. Ann Oncol. 2007;18(3):504–509. [DOI] [PubMed] [Google Scholar]
  • 47. Smith S, Su D, Rigault de la Longrais IA, et al. ERCC1 genotype and phenotype in epithelial ovarian cancer identify patients likely to benefit from paclitaxel treatment in addition to platinum-based therapy. J Clin Oncol. 2007;25(33):5172–5179. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

noae237_suppl_Supplementary_Materials
noae237_suppl_supplementary_materials.docx (10.1MB, docx)

Data Availability Statement

All relevant data supporting the findings of this study are available within the paper and supplementary information files, and other data that support this study are available from the corresponding author upon reasonable request.

Articles from Neuro-Oncology are provided here courtesy of Society for Neuro-Oncology and Oxford University Press

RESOURCES