Interfering with mitochondrial dynamics sensitizes glioblastoma multiforme to temozolomide chemotherapy

Abstract Glioblastoma multiforme (GBM) is a primary tumour of the central nervous system (CNS) that exhibits the highest degree of malignancy. Radiotherapy and chemotherapy are essential to prolong the survival time of patients. However, clinical work has demonstrated that sensitivity of GBM to chemotherapy decreases with time. The phenomenon of multi‐drug resistance (MDR) reminds us that there may exist some fundamental mechanisms in the process of chemo‐resistance. We tried to explore the mechanism of GBM chemo‐resistance from the perspective of energy metabolism. First, we found that the oxidative phosphorylation (OXPHOS) level of SHG44 and U87 cells increased under TMZ treatment. In further studies, it was found that the expression of PINK1 and mitophagy flux downstream was downregulated in GBM cells, which were secondary to the upregulation of TP53 in tumour cells under TMZ treatment. At the same time, we examined the mitochondrial morphology in tumour cells and found that the size of mitochondria in tumour cells increased under the treatment of TMZ, which originated from the regulation of AMPK on the subcellular localization of Drp1 under the condition of unbalanced energy supply and demand in tumour cells. The accumulation of mitochondrial mass and the optimization of mitochondrial quality accounted for the increased oxidative phosphorylation, and interruption of the mitochondrial fusion process downregulated the efficiency of oxidative phosphorylation and sensitized GBM cells to TMZ, which was also confirmed in the in vivo experiment. What is more, interfering with this process is an innovative strategy to overcome the chemo‐resistance of GBM cells.

radiotherapy and chemotherapy. 5 However, the postoperative recurrence of GBM is inevitable and the survival period of patients with recurrence is usually no more than 12 months. 6,7 Currently, the first-line chemotherapy agent used for GBM is temozolomide (TMZ), an imidazole tetrazine alkylating agent. 8 However, clinical observations have shown that the sensitivity of GBM to TMZ treatment decreases over time and that this may be closely related to multiple inherent or acquired mechanisms that confer resistance to chemotherapy.
In the studies of the chemo-resistance, mitochondria are under spotlight for a long time. Mitochondria are involved in vital processes such as energy synthesis, regulation of calcium ion, ROS generation and apoptosis execution 9 , which all play roles in chemo-resistance.
In addition, the mitochondrial quality control system also affects the sensitivity of tumour cells to chemotherapy. 10 As we all know, TMZ kills tumour cells by inducing DNA damage, and the survival cells are always with upregulated DNA damage response or multiple stress responses. Although these responses vary, energy support is essential. 8 That's why we took the energy supply in glioma cells under TMZ treatment as subject in this study.
Under conditions of chemotherapeutic stress, the metabolic pattern of tumour cells undergoes reprogramming in order to adapt to cellular stress status. 11 In this study, we found that the metabolic reprogramming of tumour cells involved in mitochondrial dynamics is a mechanism underlying the resistance of GBM to TMZ. Mitochondrial dynamics is an important promoter of metabolic reprogramming. Mitochondrial dynamics consists of a range of processes, including mitochondrial fission, fusion and mitophagy. 12 Mitochondrial fission excludes severely damaged mitochondria from the mitochondrial system, while mitochondrial fusion can achieve structural and functional complementation between damaged mitochondria, 13 thereby affecting cell metabolism by controlling the quantity and quality of the mitochondria. Interfering with the process of mitochondrial dynamics and the metabolism in tumour cells can sensitize GBM to TMZ. The study provides us with new ideas for the treatment of GBM.

| Cellular and animal models
Human glioblastoma cell lines (SHG44, U87MG and U251MG) were obtained from the Chinese Academy of Medical Sciences and cultured in Dulbecco's modified Eagle's medium supplemented with 10% foetal calf serum, penicillin (50 U/ml) and streptomycin (50 µg/ml) purchased from Life Technologies and incubated at 37°C in a 5% CO 2 atmosphere.

| Animal models
The experimental protocol was approved by the local ethics committee (20200921-1). Nude mice were adapted to the experimental environment for a week prior to experimentation. Then, 5 × 10 6 U87MG cells were suspended in 100 µl of phosphate buffer saline (PBS) (Beyotime Institute of Biotechnology) and injected subcutaneously into the right armpit of the nude mice. The animals were raised in a specific pathogen-free (SPF) environment, and tumour volume was observed regularly. One week later, when the tumour volume reached 50 mm 3 , the nude mice were randomly divided into four groups, with eight mice in each group. The treatment conditions for each group were as follows: control group: 200 µl of 5% Carboxymethyl Cellulose (CMC)-daily gavage; TMZ group: 40 mg of TMZ suspended in 200 µl of 5% CMC-daily gavage; WY14643 group: 10 mg of WY14643 suspended in 200 µl of 5% CMC-daily gavage; TMZ+WY14641 group: TMZ (40 mg) + WY14643 (10 mg) suspended in 200 µl of 5% CMC-daily gavage. The status of the nude mice was observed daily. Mice were weighed and the size of each tumour was measured every 2 days.
The experiment was terminated when the longest diameter of the tumour reached 2 cm or signs of ulceration appeared. Upon termination of the experiment, the nude mice were sacrificed, and the tumour was stripped, photographed and weighed. The tumour tissue was used for the following experimental tests: (1) evaluating the ATP content of tissue cells; (2) the total protein was extracted from the tumour tissue for Western blotting (WB); (3) part of the tumour tissue was stored in 2.5% dialdehyde and analysed by transmission electron microscopy; and (4) part of the tumour tissue was stored in 4% paraformaldehyde for immunohistochemical analysis.

| Western blot protein analysis
Cells were lysed in Prusiner's buffer (Tris-HCl 1 M; pH 7.5 containing NaCl (150 mM), EDTA (5 mM), Triton ×-100 (0.5%), and deoxycholate and protease inhibitor cocktail). The homogenates obtained were then briefly sonicated. Aliquots of 10µg of total protein were then loaded onto 8%-16% SDS-PAGE gels. After migration, proteins were wet-transferred to PVDF membranes and immunoblotted using certain antibodies. Immunological complexes were detected with either anti-rabbit or anti-mouse IgG-coupled peroxidase antibodies (ProteinTech Group, Inc.) by the electrochemiluminescence detection method (Roche Diagnostics S.A.S). Images were further processed by ImageJ Software version 1.52 s (National Institutes of Health) to assess the intensity of immunofluorescence (IF), as well as the mitochondrial parameters. For the evaluation of mitochondrial fusion, we evaluated three images in each treatment group from three independent experiments using ImageJ Software. Two parameters were used to assess the level of mitochondrial fusion: the mean size of the mitochondrial network (MS) and the mean length of mitochondria (ML).

| Measurement of glucose and lactate concentration
Cells were seeded in 6-well plates at a density of 3 × 10 5 cells/well. Cells were then incubated overnight at 37°C, and the medium was replaced with fresh complete medium. After 24 h, the culture medium was collected and proteins were extracted by sonication. Extracted proteins were then quantified with a Bradford Protein Assay kit (Beyotime Institute of Biotechnology). Then, we measured the concentrations of glucose and lactate with glucose (RsBio) and lactate assay kits (Jiancheng Bio), respectively. The glucose consumption in each experimental group was calculated as follows: Glucose consumption = glucose concentration (fresh complete medium)-glucose concentration (experimental group).

| Determination of ATP concentration
Cells were seeded in 6-well plates at a density of 5 × 10 5 cells/well.
Following overnight incubation at 37°C, the medium was replaced with fresh culture medium. After 6 or 24 h, in accordance with the manufacturer's instructions, cells were washed with PBS, and then, their ATP levels were determined using an ATP Assay Kit (Beyotime Institute of Biotechnology).

| Cell viability assays
Cells were seeded in 96-well plates with 100µl of complete DMEM medium at a density of 1 × 10 4 cells per well. After exposure to TMZ and/or Mdivi1 or WY14643, for 24 h, the cells in each well were incubated with MTT solution (0.5 mg/ml) dissolved in PBS for 4 h at 37°C. Then, 150 µl of DMSO was added to each well. The absorbance was measured at 570 nm using a CLARIOstar microplate reader (BMG Labtech). Cell viability was then calculated as follows: cell viability = absorbance of experimental group/absorbance of control group × 100%.

| qPCR assay
RNA was extracted using a RNeasy kit (Qiagen) in accordance with the manufacturer's protocol. Quantitative real-time reverse transcription-PCR (qRT-PCR) was then performed using a two-step method. Data analysis was based on the delt-delt-Ct method using β-Actin as a normalization control. PCR was conducted in a QuantStudio5 real-time PCR system (Thermo Fisher) and analysed using QuantStudio Design & Analysis software v1.3.1 (Thermo Fisher).
In order to evaluate the mtDNA copy number, we extracted DNA from cells in each group and determined the copy number of the mtDNA gene ND1, as normalized to the 18S gene. In order to evaluate the extent of DNA damage, we selected two fragments of mtDNA (79bp and 230bp in length) as target templates for a qPCR array using the 18S gene as an internal reference. The extent of mtDNA damage was parallel to the value of 79bp/230bp.
The primers used for the real-time PCR are listed in Table 1.

| The extraction of nuclear and mitochondrial proteins
The Mitochondria and Nuclear Extraction Kit for Cells (Beyotime Biotechnology) was used to isolate nuclear and mitochondrial fractions from each experimental sample. In each case, we followed the manufacturer's guidelines.  T TCGAGATGT TCCGAGAGC TGA ATG AGGC C T TGG A   AC TCA AGGATGCCCAGGC TGGGA AGGAGCCAGGGGGGAG   CAGGGCTCACTCCAGCCACCTGAAGTCCAAAAAGGGTCAGTC   TACC TCCCGCCATA A A A A AC TCATGT TCA AGACAGA AGGG CCTGACTCAGACTGA.

| Immunohistochemistry
Tumour tissues were fixed in 4% paraformaldehyde for 24 h and em- Images were captured on a microscope (Discover ECHO Company).

| TEM observation
Tumour tissues from the different treatment groups were fixed in 2.5% glutaraldehyde and then embedded in an embedding agent.
Next, specimens were sliced into sections (50-70 nm in thickness). The sections were then dyed with uranyl acetate and citrate dye solution.
The TEM system (Thermo Fisher Scientific Company) operated at an acceleration voltage of 80 kV and an electron tomography voltage of 120 kV. Specimens were magnified by ×150,000, and images were acquired. Three tumour specimens in each treatment group were randomly selected for TEM examination. And each specimen outputted a clear picture of which the mean length of mitochondria was measured directly by the software of TEM. Then, the data were analysed.

| Flow cytometry
Apoptosis assays were performed using an Apoptosis Detection Kit (BD Biosciences). Cells from different treatment groups were trypsinized with

Genes Primers
Reverse-GCTGTCACCTTCACCGTTCC 0.25% trypsin, and assays were conducted in accordance with the manufacturer's protocol. In total, 10000 cells were counted in each treatment group using a Guava ® easyCyte flow cytometer (Merck KGaA).
For the analysis of mitochondrial mass, cells were grown in 6-well plates and treated with the different agents. Cells were trypsinized with 0.25% trypsin and suspended cells were stained with MitoTracker™ Red FM (Invitrogen) (1:10000) for 30 min at 37℃. Cells were then rinsed three times with PBS, and 10,000 cells were counted for each treatment group using a Guava ® easyCyte flow cytometer (Merck KGaA).
The production of ROS was measured by a ROS Assay Kit (Beyotime Biotechnology). In brief, cells were collected and resuspended in DMEM medium after treatment and then incubated with DCFH-DA for 30 min at 37°C. Cells were then rinsed three times in PBS, and 10,000 cells were for each treatment group using a Guava ® easyCyte flow cytometer (Merck KGaA).

| Statistical analysis
Data from three independent experiments were collected and ana- presented as means ± standard deviation, and p < 0.05 was considered to indicate a statistically significant difference.

| The level of oxidative phosphorylation and mitochondrial mass in GBM cells increased under TMZ treatment
As an alkylating agent, TMZ can induce damage to nuclear DNA and mitochondrial DNA. 14 It is generally believed that DNA damage, especially mitochondrial DNA damage, will affect the expression of subunits within the respiratory chain complex, 15 thus influencing the metabolic activity of tumour cells. Using SHG44 and U87 cell lines, we first evaluated the metabolic activity of cells under TMZ treatment. When treated with TMZ, both cell lines showed increased oxygen consumption rate ( Figure 1A), increased levels of intracellular ATP and glucose consumption ( Figure 1B), while the lactic acid secretion ( Figure 1B) and the levels of intracellular ROS remained unchanged ( Figure 1C). These findings suggested that TMZ treatment increased the level of oxidative phosphorylation in SHG44 and U87 cells but without increasing the levels of ROS.
In order to identify the mechanisms underlying the increased level of oxidative phosphorylation, we first detected the mass of mitochondria in the two cell lines, using three methods: (1)

| Reduced mitophagic flux was an important factor underlying the increased mitochondrial mass
The mass of mitochondria within a cell is precisely controlled by mitophagy. 17 In order to investigate whether mitochondrial dynamics is involved in metabolic reprogramming under chemotherapy stress, we evaluated the flux in mitophagy. There are many pathways that mediate mitophagy; the most dominant pathway is the PINK1-Parkin pathway. 18,19 We first evaluated the level of the fusion between mitochondria and lysosomes in the two cell lines under TMZ treatment and found that the co-localization of the two organelles was reduced ( Figure 3A). We also studied the co-localization of Parkin and mitochondria with IF assay ( Figure 3B). And the further WB assay with mitochondrial protein revealed that when cells were treated with TMZ, the accumulation of Parkin and PINK1 in mitochondria decreased ( Figure 3C). In the IF assay and WB assay with whole-cell proteins, the expression levels of Parkin were unchanged while the expression of PINK1 and PINK1-specific phosphorylation product (ser65) phosphorylated ubiquitin decreased when treated with TMZ ( Figure 3D, E). In order to further confirm the change of (ser65) phosphorylated ubiquitin, we introduced the agent MG132 and combined it with TMZ in the treatment of SHG44 and U87 cells ( Figure S1A). We also evaluated the expression of PINK1 in transcriptional level, in which, the downregula-  Figure 4A, B). 20 We also found that the copy number of the 79bp/230bp fragments of mtDNA increased under TMZ treatment in SHG44 cells, thus indicating increased levels of mtDNA damage ( Figure 4C). 21 Secondly, we observed the elevated expression of the stress response protein P53 in the two cell lines ( Figure 4D, E), along with the increased accumulation of P53 protein in the nucleus ( Figure 4F). As an important stress response protein in the cell, P53 can act as a transcriptional regulator in the nucleus, thus regulating the expression of a variety of downstream cell cycle monitoring genes and apoptosis-related genes. 22 In order to determine whether there is a correlation between the expression of P53 and PINK1, we performed correlation analysis of the expression levels of these two molecules using tumour databases CGGA, TGCA and Rembrandt. We found that the expression of these two proteins was negatively correlated ( Figure 4G). These results were consistent to our immunohistochemistry analysis of tumour tissues ( Figure 4H).
To further verify the regulatory effect of P53 on the expression of to act as a nuclear transcriptional regulator. 24 In this model, we observed that TMZ treatment increased the expression of mut-P53 protein in U251 cells, and also increased the expression of PINK1 protein in U251 cells ( Figure 5D, E). What is more, the localization of Parkin to mitochondria reduced ( Figure 3F). These findings confirmed that wt-TP53 participates in the downregulation of mitophagy flux by reducing the expression levels of PINK1; this mechanism is consistent with another excellent study. 25

| The subcellular distribution of Drp1 impacted the mitochondrial morphology which may play role in metabolic reprogramming
It is generally believed that the downregulation of mitophagy flux is accompanied by the accumulation of damaged mitochondria, which is the main source of excessive ROS production in cells. 26,27 However, in the present study, the downregulation of mitophagy flux was not

| The activated AMPK regulated the translocation of Drp1 to mitochondria
We were also interested in the mechanisms responsible for the reduced translocation of DRP1 to the mitochondria. We hypothesize F I G U R E 3 Mitophagy flux in GBM cells decreased under TMZ treatment. (A) Mitotracker (red) and Lysotracker (green) were used to label mitochondria and lysosomes, respectively. This analysis revealed a reduction in the co-localization of mitochondria and lysosomes in the two cell lines when treated with TMZ; (B) IF experiments further showed that the co-localization of Parkin (green) and mitochondria (red) in the TMZ group was reduced when compared to the Control; (C) Analysis of mitochondrial protein suggested that the accumulation of PINK1 and Parkin was decreased in the mitochondria of cells treated with TMZ. (D) IF assays also showed that the expression of PINK1 (green) was lower in cells that were treated with TMZ; (E) WB assays further confirmed that TMZ treatment downregulated the expression of PINK1 in both SHG44 and U87 cell lines. The specific phosphorylated product of PINK1 (phosphorylated (Ser65) ubiquitin) was also downregulated. The expression of Parkin in these cells did not change significantly. ( Figure 7B). AMPK, as an 'energy sensor' in cells, is an important molecule that regulates metabolism. 29 We next consid-  Figure S1C). We also found that activation of the AMPK pathway was positively correlated with the expression of P53 in the nucleus. This may be related to the phosphorylation of P53 protein by AMPK; as this leads to a reduction in the degradation of P53. 32 We also confirmed the negative correlation between TP53 and PINK1 expression (Figure 7 C, Figure S1B). These findings sug- What is more, to confirm that the AMPK regulated the function of mitochondria, we performed the mito stress test by using the seahorse analyzer and revealed that: 1.TMZ elevated the OXPHOS potential of mitochondria while repressing the activity of AMPK impaired the effect ( Figure S1D).

| Interfering with mitochondrial fission and fusion affected the metabolic changes induced by TMZ
In order to verify that the mobilization of mitochondrial dynamics was involved in the metabolic reprogramming of tumour cells under chemotherapy stress and to explore whether interfering with mitochondrial dynamics could sensitize GBM to TMZ treatment, we used two different agents: WY14643 to promote mitochondrial fission 33,34 and Mdivi1 to promote mitochondrial fusion. 35 In the TMZ+WY14643 treatment group, we observed that the level of mitochondrial fusion was lower than that of the TMZ treatment group; the level of mitochondrial fusion in the TMZ+Mdivi1 treatment group was higher than that of the TMZ treatment group ( Figure 8A). Furthermore, cell viability assays suggested that certain concentrations of WY14643 shifted the TMZ concentration-cell viability curve of SHG44 and U87 cells downwards. Certain concentrations of Mdivi1 shifted the TMZ concentration-cell viability curve upwards ( Figure 8B). Evaluating the metabolism of different treatment groups revealed that the TMZ+WY14643 treatment group had lower intracellular levels of ATP, while the TMZ+Mdivi1 treatment group had more intracellular ATP than cells in the TMZ treatment group. However, the secretion of lactic acid did not increase significantly ( Figure 8C). These results suggested that the promotion of F I G U R E 4 TMZ treatment-induced DNA damage in GBM cells and the resultant TP53 overexpression affected the expression of PINK1. (A) WB assays showed that TMZ treatment increased the expression of γ-H2A.X, a marker of cellular DNA damage. (B) IF assays also confirmed that TMZ increased the expression of γ-H2A.X (green) in the nucleus; (C) qPCR was used to evaluate the copy number of two fragments of mitochondrial DNA (230bp fragment and 79bp fragment) and showed that in SHG44 cells, the ratio of the 79bp/230bp fragments in the TMZ treatment group was higher than that in the control group; (D) TMZ treatment led to the upregulated transcription of TP53; (E) The expression of P53 protein was also upregulated in the TMZ treatment group; (F) The combination of cell nucleus extraction and WB assays revealed an increased accumulation of P53 in the nucleus of cells undergoing TMZ treatment; (G) Analysis of the correlation between TP53 and PINK1 expression in the tumour database CGGA (R = −0.13, p < 0.001), TGCA (R = −0.50, p < 0.001) and Rembrandt (R = −0.56, p < 0.001) indicated that the expression levels of TP53 and PINK1 were negatively correlated in glioma; (H) In U87 cell xenograft tumour specimens, the expression levels of P53 protein and PINK1 were negatively correlated. (*p < 0.05, **p < 0.01, ***p < 0.001) mitochondrial fission by WY14643 partially reversed the metabolic changes induced by TMZ treatment in GBM cells. This was also supported by the detection of cellular ROS levels: the WY14643+TMZ treatment group exhibited higher levels of ROS ( Figure 8D).
Finally, we found that WY14643 promoted mitochondrial fission and increased the efficacy of TMZ chemotherapy. This treatment also enhanced the efficacy of TMZ to induce apoptosis of SHG44 and U87 cells (Figure 8 E).

| In vivo experiments confirmed that the induction of mitochondrial fission can sensitize GBM to TMZ treatment
In order to further verify that the disruption of TMZ-induced mitochondrial fusion could sensitize GBM to TMZ chemotherapy, we designed and established a nude mouse subcutaneous xenograft tumour model, and administered WY14643 and/or TMZ. We found that the combination of WY14643 and TMZ demonstrated the most powerful inhibitory effect over tumour growth across all four groups ( Figure 9A, B, C, D). Evaluation of the expression levels of apoptosis-related proteins within tumour tissue confirmed that the highest expression levels of apoptotic proteins occurred in the group featuring the combination of two agents ( Figure 9E). The

F I G U R E 9
Interfering with the mitochondrial fusion enhanced the TMZ efficacy to inhibit GBM tumour growth in nude mice. (A) Images of tumours acquired from different treatment groups. The TMZ+WY14643 treatment group showed the strongest tumour suppressive effect. (B) Volume curve of xenograft tumours in different treatment groups. The TMZ+WY14643 treatment group showed the strongest levels of tumour suppression. (C) Tumour weight statistics from different treatment groups; the TMZ+WY14643 treatment group showed the strongest tumour suppressive effect; (D) Weight curve for nude mice in different treatment groups; (E) The expression of apoptosisrelated proteins in tumour tissues from different treatment groups; the expression levels of apoptosis-related protein were the highest in the TMZ+WY14643 treatment group; (F) The evaluation of ATP levels in tumour tissues revealed that the ATP levels of the TMZ group were higher than that in the control group, while also higher than that in the TMZ+WY14643 treatment group; (G) Transmission electron microscopy examination revealed that the mean length of mitochondria in the TMZ treatment group was greater than that of the control, while the mean length of mitochondria in the TMZ+WY14643 group was less than that in the TMZ group. (*p < 0.05, **p < 0.01, ***p < 0.001) evaluation of ATP levels in tumour tissues within each treatment group revealed that the ATP levels in tissues from the TMZ treatment group were higher than that in the Control group; the levels of ATP in the WY14643+TMZ group were lower than those in the TMZ treatment group ( Figure 9F). The morphology of mitochondria in tumour tissue was evaluated by transmission electron microscopy (TEM) and revealed that mitochondria in the TMZ+WY14643 group had a shorter mean length than those in the TMZ treatment group  36 And it has been found that when tumour cells become resistant to certain chemotherapeutic agents, they usually show resistance to a variety of chemotherapeutics. This phenomenon is known as multi-drug resistance (MDR). 37 The exact mechanism underlying MDR has yet to be mice from the WY14643+TMZ treatment group had a shorter mean mitochondrial length, a lower tissue ATP level, a higher expression level of apoptosis-related proteins, and more significant growth inhibition, than those in the TMZ treatment group. This indicated that by disrupting mitochondrial dynamics, we may be able to overcome chemo-resistance in GBM. However, the precise effect of mitochondrial fission on cell fate remains inconclusive. Some studies have found that inducing mitochondrial fission promotes the occurrence of apoptosis, 45,46 while other studies have found that mitochondrial fission exhibited anti-apoptotic properties. 47 These differences may be related to differences in cell type and experimental conditions. In support of the fact that mitochondrial fission may promote apoptosis, it is believed that Drp1 promotes the release of cytochrome c and the activation of Caspases. 48 Studies have also suggested that Drp1 may be associated with the Bcl2 family, by promoting the pro-apoptotic protein Bax to translocate to the mitochondria, 49 or because the interaction of Drp1 with Bcl2 promotes cell apoptosis. 50 In the present study, we found that under TMZ treatment, mitochondrial fission influenced the response of GBM cells to chemotherapy by affecting energy metabolism.
We should also highlight the fact that genetic polymorphisms exist in GBM cells. The frequency of TP53 mutations is known to be significantly different when compared between primary and recurrent glioblastomas, 51 and that the frequency of TP53 mutations also differs across different subtypes of GBM, accounting for 54%, 32%, 21% and 0% of Proneural, Interstitial, Neuro and Classical GBM, respectively. 52 However, wt-TP53 and mut-TP53 play different roles in the regulation of mitochondrial dynamics. This is reflected in our study of the TP53 mutant cell line, U251. The regulatory function of mut-TP53 on mitochondrial dynamics needs to be investigated further. In addition, the specific molecular mechanisms underlying the regulation of AMPK signalling and translocation of Drp1 between subcellular structures, also needs to be further clarified.

| CON CLUS ION
The study of the mechanisms underlying chemo-resistance has always been popular in the field of chemotherapy. Recent research studies have revealed a diverse and complex array of drug resistance mechanisms in tumour cells. In the present study, we identified a new mechanism of chemo-resistance, the reprogramming of metabolism. By disrupting mitochondrial dynamics, it was possible to reduce the sensitivity of GBM cells to TMZ; these findings were confirmed by both in vitro and in vivo studies. This study provides us with a new concept for basic and clinical research into the chemotherapeutic strategies used to treat GBM.

ACK N OWLED G EM ENTS
This study was supported by the Department of Science and

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available on request from the corresponding author.