Synergistic combination of DT‐13 and Topotecan inhibits aerobic glycolysis in human gastric carcinoma BGC‐823 cells via NM IIA/EGFR/HK II axis

Abstract DT‐13 combined with topotecan (TPT) showed stronger antitumour effects in mice subcutaneous xenograft model compared with their individual effects in our previous research. Here, we further observed the synergistically effect in mice orthotopic xenograft model. Metabolomics analysis showed DT‐13 combined with TPT alleviated metabolic disorders induced by tumour and synergistically inhibited the activity of the aerobic glycolysis‐related enzymes in vivo and in vitro. Mechanistic studies revealed that the combination treatment promoted epidermal growth factor receptor (EGFR) degradation through non‐muscle myosin IIA (NM IIA)‐induced endocytosis of EGFR, further inhibited the activity of hexokinase II (HK II), and eventually promoted the aerobic glycolysis inhibition activity more efficiently compared with TPT or DT‐13 monotherapy. The combination therapy also inhibited the specific binding of HK II to mitochondria. When using the NM II inhibitor (‐)002Dblebbistatin or MYH‐9 shRNA, the synergistic inhibition effect of DT‐13 and TPT on aerobic glycolysis was eliminated in BGC‐823 cells. Immunohistochemical analysis revealed selective up‐regulation of NM IIA while specific down‐regulation of p‐CREB, EGFR, and HK II by the combination therapy. Collectively, these findings suggested that this regimen has significant clinical implications, warranted further investigation.


| INTRODUC TI ON
Gastric cancer (GC) is one of the most common cancers worldwide, especially in developing countries. 1,2 Despite improved surgical resection and efficient adjuvant therapy at the early stages of the disease, GC patients have a poor prognosis and low 5 year survival rate, with the frequent occurrence of ensuing relapse and metastasis. 3 Moreover, advanced gastric cancers are resistant to traditional therapies and modern treatments are therefore failed. 4,5 Therefore, it is of great importance to develop novel therapies for gastric cancer.
Aerobic glycolysis was a process with the involvement of many enzymes, for example hexokinase II (HK II), phosphofructokinase-1 (PFK-1), pyruvate kinase M2 (PKM2) and lactate dehydrogenase (LDHA). 6 Some of them have been reported to be overexpressed in tumours, including GC, and can be regulated by many oncoproteins to promote tumour proliferation, migration and chemoresistance. 7 Accelerated glucose uptake during aerobic glycolysis and loss of regulation between glycolytic metabolism and respiration are the major metabolic changes found in malignant cells. 7,8 Interest in targeting cancer metabolism has been renewed in recent years, and more and more enzymes related to aerobic glycolysis were found to be as potential targets for drug treatment, including HK II, PKM2, glucose transporter 1 (GLUT1) and so on. 9,10 Metabolism and the aerobic glycolysis of cancer cells are seen as specific target of cancer cell, which have provided a new view for cancer treatment. 11,12 Topotecan, a semi-synthetic analogue of the new Topo-inhibitor camptothecin, has been licensed as an anticancer agent for small cell lung cancer (SCLC), 13 ovarian cancer, 14 head-neck tumours, and gastrointestinal carcinoma. The usage of TPT on treating GC and SCLC is limited by its side-effects such as toxicity and suffered from the potential drug resistance. DT-13, a saponin of the dwarf lilyturf tuber Ophiopogon japonicus wall (Family: Convallariaceae), possesses anticancer activities against various types of cancers and antiangiogenesis activity 15 on multiple targets, such as early growth response 1 (Egr-1), VEGF, CCR-5, HIF-1α and MMP2/9. [16][17][18] To increase the efficacy of TPT and avoid the risk of resistance development, we designed combination therapies of DT-13 and TPT and found that DT-13 enhanced the pro-apoptotic effect of TPT on GC by up-regulating NM IIA/EGFR/Cav-1 axis. 19 As a key target of our combination therapies, NM IIA, encoded by MYH-9, is an ATP-driven molecular motor that plays diverse roles in cell physiological functions such as cell migration, adhesion, polarization and cytokinesis. [20][21][22] NM IIA was essential for the endocytosis of EGFR and the modulation of the EGFR-dependent activation of downstream signals, including ERK1/2 and AKT. 23 In our previous research, we further confirmed the relationship of NM IIA and EGFR and that the ability of DT-13 to enhance the pro-apoptotic effect of TPT on GC via myosin IIA-induced endocytosis of EGFR in vitro and in vivo. 19 Meanwhile, it had been reported that EGFR signalling was also associated with increased glycolysis: activated the first step in glycolysis. 24 On the contrary, EGFR inhibitors may reactivate oxidative phosphorylation of cancer cells and provide a mechanistic clue for the rational combination of agents targeting EGFR-dependent proliferation and glucose metabolism in cancer therapy. 25 Based on the finding that NM IIA induced the endocytosis of EGFR and the fact that EGFR could increase the activity of aerobic glycolysis, we developed the hypothesis that DT-13 might be synergistically combined with TPT to inhibit aerobic glycolysis in high EGFR expression GCs through increasing NM IIA modulation of EGFR endocytosis and downstream signalling.

| In vivo tumorigenicity
Female athymic BALB/c nude mice (5-6 weeks old) with body masses ranging from 18 to 20 g were supplied by the Shanghai Institute of Materia Medica, Chinese Academy of Sciences.
BGC-823 cells were collected in serum-free medium (10 6 cells/100 μL). Then, the cell suspension was injected subcutaneously into four mice in one flank as host. After 2 weeks, the resulting subcutaneous tumour (≈2 cm in diameter) were surgically removed under strict aseptic conditions following removal of necrotic tissue from the central tumour areas, and cut into small cubic fragments of approximately 1 mm 3 . Nude mice were randomly divided into two groups: 54 as tumour bearing mice were explanted with tumour fragments from the BGC-823 cell lines, while the other six were assigned as sham operation group (SHAM). In brief, mice were anaesthetized by i.p. injection of sterile pentobarbital solution (50 mg/kg of body weight, China Academy of Military Medical Science). For implantation, the mouse stomach was gently exteriorized via a left-side upper abdominal incision, and one small tissue pocket was prepared in the middle wall of the greater curvature using microscissors, and then, one tumour piece was placed into the pocket. The stomach was then returned to the peritoneal cavity, and the abdominal wall was closed with 4-0 absorbable sutures. The mice were given special care and fed in cages as usual after surgery. After a week, 54 tumour-bearing mice were divided into nine groups according to their weight (n = 6): BGC-823 tumour-bearing mice (BGC); BGC mice injected intravenously with the combination therapy. Collectively, these findings suggested that this regimen has significant clinical implications, warranted further investigation.

K E Y W O R D S
aerobic glycolysis, DT-13, EGFR, gastric cancer, NM IIA, TPT 8.0 mg/Kg (HTPT) or 0.5 mg/Kg (LTPT) dissolved in saline twice a week; BGC mice administrated intragastrically with 2.5 mg/Kg (HDT), 1.25 mg/Kg (MDT) or 0.625 mg/Kg DT-13 (LDT) suspended in 0.5% sodium carboxymethylcellulose (CMC-Na) once a day; BGC mice combined administration group with 0.5 mg/Kg TPT and 2.5 mg/Kg (HD-LT), 1.25 mg/Kg (MD-LT) or 0.625 mg/Kg DT-13 (LD-LT). However, the mice in SHAM group were treated with normal saline. All nude mice were observed the growth and physical condition every day, weighed three times a week for 3 weeks.
After 3 weeks of administration, the nude mice were killed, and the tumours were completely removed. Animal care and surgery protocols were approved by the Animal Care Committee of China Pharmaceutical University. All animals were treated appropriately and used in a scientifically valid and ethical manner.

| Preparation of samples and acquisition of 1 H NMR spectra
The deep-frozen serum samples were thawed at 4°C overnight and then were vortexed to remove any precipitates. To 300 μL of each serum sample, 150 μL phosphate buffer was added, which was dissolved in D 2 O (0.2 M, Na 2 HPO 4 /NaH 2 PO 4 and pH 7.0) containing 0.05% TSP-d 4 as chemical shift reference. Samples were vortexed and then centrifuged at 12 000 g for 10 minutes at 4°C to afford 500 μL of supernatant.
Frozen tissue sections were weighed (ca. 250 mg), homogenized in precooled acetonitrile/water (vol/vol = 1:1, 5 mL/g tissue) kept in an ice/water bath and centrifuged (12 000 g, 10 minutes, 4°C). 38 The supernatant was lyophilized and then reconstituted in 600 mL phosphate buffer dissolved in D 2 O. After vortexing and centrifugation (12 000 g, 10 minutes, 4°C), a total of 550 mL of the supernatants was pipetted into 5 mm NMR tubes for analysis.
All the 1 H NMR spectra were recorded at 298 K on a Bruker AV-500 MHz spectrometer. The water-suppressed Carr-Purcell-Meiboom-Gill (CPMG) spin-echo pulse sequence (RD-90°-(τ-180°-τ) n-ACQ) with a total spin-echo delay (2nτ) of 40 ms was used to attenuate broad signals from proteins and lipoproteins. Typically, 128 transients were acquired with 32 K data points for each spectrum with a spectral width of 10 kHz.

| NMR data processing and multivariate data analysis
With TopSpin software (version 3.0, Bruker Biospin), all the 1 H NMR spectra were automatically corrected for phase and baseline distortions, and calibrated to TSP at 0.00 ppm. The spectral regions between δ 0.8 and 9.0 ppm for each tissue sample were automatically data reduced to integral segments using an adaptive binning approach based on the code 39 implemented in MATLAB (version 7.3, MathWorks). The area under the spectrum was then calculated for each segmented region and expressed as an integral value. The regions of water resonance (4.59-5.15 ppm) were removed to avoid the effects of imperfect water suppression. To account for variations in concentration of metabolites due to dilute, the binned data were probabilistic quotient normalized using a median calculated spectrum, 40 mean-centred and pareto scaled in R, a freely available, open-source software (R Development Core Team, http://cran.r-proje ct.org/) prior to multivariate statistical analysis. A supervised orthogonal partial least-squares discriminant analysis (OPLS-DA) was performed using scripts written in R language.

| Identification of metabolites
The statistical total correlation spectroscopy (STOCSY) technique was used to identify multiple NMR peaks from the same molecule in a complex mixture, which took advantage of the multicollinearity of the intensity variables in a set of spectra. Aided by STOCSY, metabolites were identified by comparing with those reported in the literature and/or registered in Human Metabolome Database (HMDB) (www.hmdb.ca). The assignments were further confirmed by 2D NMR techniques such as total correlation spectroscopy (TOCSY) and heteronuclear singular quantum correlation (HSQC).

| Cell culture
Human gastric cancer BGC-823 cells were purchased from The Shanghai Institute of Life Science, Chinese Academy of Sciences.

| Cell transfection
The MYH-9 lentivirus shRNA was purchased from Shanghai GenePharma Co., Ltd. The cells were seeded at a density of 5 × 10 4 / well in a 24-well plate 24 hour before transfection to achieve more than 70% confluence. 20 μL MYH-9 lentivirus shRNA and 20 μL scrambled sequence lentivirus shRNA were added into 2 mL fresh medium individually and then added 2 μL polybrene (Santa Cruz) after 24 hour treatment, and lentivirus medium was replaced by fresh medium. The plate was incubated at 37°C for 48-72 hours until the transfection efficiency was more than 80% and was then used in the experiments described below. The sequence for MYH-9 shRNA was forward, 5-GAGGCAAUGAUCACUGACUdTdT-3; reverse, 5-AGUCAGUGAUCAUUGCCUCdTdT-3.

| Co-immunoprecipitation assays
HK II (Santa Cruz) was immunocaptured from cells extracts using polyclonal antibodies to HK II cross-linked to protein G-agarose beads (Thermo Fisher Scientific). The immune complexes were analysed by Western blotting and probed with antibody against VDAC (Cell Signaling Technology).

| Immunofluorescence assays
The cells were fixed with 4% paraformaldehyde in PBS at 30 minutes intervals, permeabilized with 0.5% Triton X-100 and blocked with 3%

| Immunohistochemical analysis for NM IIA, EGFR, p-CREB and HK II in tumour tissues
Paraffin-embedded tissue sections were deparaffinized in xylene followed by treatment with a graded series of alcohol [100%, 95% and 80% (pH 7.5). Antigen retrieval for ethanol/double-distilled

| Western blot analysis
Cellular proteins were extracted, and Western blot analysis was performed as previously described after BGC-823 cells were treated with DT-13 and TPT for 48 hour. All primary antibodies were purchased from Cell Signaling Technology, Inc (Cell Signaling Technology). Horseradish peroxidase (HRP)-conjugated antimouse immunoglobulin G (Sigma-Aldrich) and anti-rabbit immunoglobulin G (Cell Signaling Technology) were used as the secondary antibodies. Protein bands were visualized using enhanced chemiluminescence reagents (Millipore).

| Glucose, lactate, ROS, ATP detection
The levels of lactic acid, glucose, ROS, ATP in cells supernatant were collected after the combination treatment for 48 hours and then assayed following the manufacturer's instructions of the related kit.

| Statistical analysis
All of the results excluding metabolomics analysis in vivo were presented as the mean ± SD from triplicate experiments performed in a parallel manner unless otherwise indicated. Statistically significant differences (one-way ANOVAs followed by Bonferroni's multiple comparison test) were determined using GraphPad Prism 6 software. A value of P < 0.05 was considered significant, and values of P < 0.01 and P < 0.001 were considered highly significant.
The P-values of metabolomics analysis in vivo were corrected by BH (Benjamini-Hochberg) methods and were calculated based on a parametric Student's t test or a nonparametric Mann-Whitney test (dependent on the conformity to normal distribution), * P < 0.05, ** P < 0.01, *** P < 0.001.

| DT-13 combined with TPT had synergistically antitumour effect and aerobic glycolysis inhibition effect in BGC-823 orthotopic xenograft nude mice
In our previous study, we had found that DT-13 combined with TPT showed stronger antitumour effect in mice subcutaneous xenograft model in vivo compared with their individual effects. 19 Here, we fur-  Figure 1D). It could be concluded that the combination treatment group could more effectively reverse the abnormal metabolic status towards a normal condition.

| The combination treatment synergistically inhibited the expression of the aerobic glycolysisrelated enzymes in vivo and in vitro
Metabolomics analysis in vivo revealed the combination treatment had stronger inhibition effect on glucose uptake and lactate production than the individual treatment effects, which indicate that the combination treatment might synergistically inhibit the aerobic glycolysis. In

| Combinational aerobic glycolysis inhibition effect was promoted by NM IIA
As we found the inhibitory of combination treatment on aerobic glycolysis in vivo and in vitro, we desired to find out the potential F I G U R E 1 Combination treatment and inhibition of tumour growth in BGC-823 xenograft nude mice, and metabolomics analysis in vivo. A, The weight of the tumours in nude mice was examined. B, Representative photographs of tumours, respectively, in each group at the end of experiment. C, Scores plot and colour-coded loadings plot according to the OPLS-DA analysis of NMR data from different tissue extracts of mice: serum; kidney; liver. Significantly changed metabolites were assigned in the loadings plots. Negative signals represent increased and positive signals represent decreased concentrations in model group. D, Potential marker metabolites in mice serum, kidney, liver identified by 1H-NMR and their fold changes among groups and the associated P-values. Colour-coded according to the fold change value, red represents increased and blue represents decreased concentrations of metabolites. P-values corrected by BH (Benjamini-Hochberg) methods were calculated based on a parametric Student's t test or a nonparametric Mann-Whitney test (dependent on the conformity to normal distribution). * P < 0.05, ** P < 0.01, *** P < 0.001 mechanism. According to the finding that DT-13 enhanced the proapoptotic effect of TPT on GC via myosin IIA-induced endocytosis of EGFR in vitro and in vivo, 19 we wondered whether the combinational aerobic glycolysis inhibition effect was correlated to NM IIA.
We first examined the relationship of NM IIA and aerobic glycolysis.
In NM IIA knock-downed BGC-823 cells, the mRNA levels of HK II, PKM2, PFK-1 and LDHA were all increased, and only HK II protein level was increased markedly (Figure S2A,B). Meanwhile, NM IIA knock-down could increase the lactate production and glucose uptake ( Figure S2C).
To further test whether the combinational aerobic glycolysis inhibition effect was correlated to NM IIA, we next detected the change of the activities of the enzymes in NM IIA knock-downed BGC-823 cells treated with DT-13-TPT combination. We found that in NM IIA knock-downed cell, the combinational treatment inhibition effect was reversed totally on HK II mRNA level; reversed a bit on PKM2 mRNA level; but could not reverse little on PFK-1 and LDHA mRNA levels ( Figure 3A). What's more, knocking-down NM IIA could only reverse the combination treatment inhibition effect on HK II protein level in BGC-823 cells ( Figure 3B). The combination treatment promotion effect on ROS level was reversed in NM IIA knocking-down cell ( Figure S3A). We also found that knockingdown NM IIA and (-)-blebbistatin could both reverse the activation of AMPK induced by the combination treatment ( Figure 3C). When pre-treated with APMK inhibitor dorsomorphin or ROS scavenger NAC for 2 hours and then treated with the combinational treatment F I G U R E 2 DT-13 combined with TPT promoted aerobic glycolysis inhibition effects in BGC-823 cells in vivo and in vitro by up-regulating the activity of AMPKα. (A-B) Effects of DT-13 combined with TPT on the mRNA levels A, and protein levels B, of the aerobic glycolysisrelated enzymes in BGC-823 tumour tissue. C, The AMPK-mTOR pathway was detected using Western blot analysis when treated with DT-13 combined with TPT in BGC-823 tumour tissue. D, The combination inhibitory effects of DT-13 combined with TPT on the levels of lactate, ATP, glucose and ROS in BGC-823 cells were measured after the combined treatment for 48 hours. The aerobic glycolysis-related enzymes were assessed by PCR and Western blot analysis in BGC-823 cells (E-F). G, After the cells were treated with the combined treatment for 48 hours, the AMPK-mTOR pathway was assessed by Western blotting analysis in BGC-823 cells. Statistical analysis was performed using one-way ANOVA followed by Bonferroni's multiple comparison test, *P < 0.05; **P < 0.01; ***P < 0.001; for A, D and E, statistical analysis was performed using at least three independent replicates for 48 hours, the decrease of HK II protein in BGC-823 cells induced by the combinational treatment could not be reversed by dorsomorphin or NAC ( Figure S3B,C). These findings showed that the change of HK II happened before the AMPK activation and ROS level increase; we speculated that DT-13 combined with TPT increased the aerobic glycolysis inhibition activity through myosin IIA-induced inactivation of HK II.

| Effects of EGFR pathway on HK II in BGC-823 cells
To verify our conjecture that DT-13 combined with TPT increased the aerobic glycolysis inhibition activity through myosin IIA-induced inactivation of HK II, we examined the effect of DT-13 and TPT on HK II in BGC-823 cells, respectively. We found that the expression of HK II was suppressed by DT-13, whereas TPT increased the expression of HK II in BGC-823 ( Figure 4A,B). DT-13 combined with TPT induced the endocytosis of EGFR by up-regulation of NM IIA ( Figure S3D), and EGFR pathway was correlated to aerobic glycolysis. 25 We next examined the effects of EGFR pathway on the activity of HK II and PKM2. As shown in Figure 4C  (20 μmol/L) or U0126 (10 μmol/L) separately for 2 hour before DT-13-TPT combination treatment, the HK II level was determined by Western blot and PCR analysis, and the total/phospho AKT, total/phospho S6K and total/phospho ERK1/2 were detected by Western blot analysis. Statistical analysis was performed using one-way ANOVA followed by Bonferroni's multiple comparison test, *P < 0.05; **P < 0.01; ***P < 0.001; for b, d and g, statistical analysis was performed using at least three independent replicates

| DT-13 combined with TPT inhibited p-CREB by down-regulating the activity of EGFR downstream pathways
The above data showed that DT-13 combined with TPT inhibited both protein and mRNA level of HK II ( Figure 2E,F). Thus, we speculated that the combination treatment inhibited the transcription process of HK II. cAMP response element binding protein (CREB) was transcription factor, which could bind to the promoter regions of HK II and induced HK II activation. [27][28][29] What's more, CREB was correlated to NM IIB. 30 Thus, we next examined the effect of the combination treatment on the activity of CREB in cell nucleus. We found that the combination treatment exhibited a stronger inhibitory effect on p-CREB level than TPT or DT-13 alone in BGC-823 cells ( Figure 5A). CREB phosphorylation is mainly controlled by protein kinase A (PKA) and calcium-dependent protein kinase (PKC), casein kinase (CK II), MAPK, AKT pathway. 31,32 Next, we further confirmed the effect of EGFR downstream MEK, PI3K pathway on p-CREB level in cell nucleus by using the EGFR and its downstream inhibitors. Except for gefitinib, six inhibitors could inhibit the p-CREB level in cell nucleus ( Figure 5B). Then, we examined the effect of the combination treatment on the p-AKT and p-ERK levels in cell nucleus by using Western blot analysis ( Figure 5C) and cell fluorescence analysis ( Figure 5D,E). We observed that the combination treatment exhibited a stronger inhibitory effect on p-AKT and p-ERK protein level and fluorescence intensity than TPT or DT-13 alone in cell nucleus.
These results confirmed that the combination of DT-13 with TPT inhibited the entrance of p-ERK, p-AKT into the cell nucleus, thereby inhibiting the activity of CREB.
F I G U R E 5 DT-13 combined with TPT could inhibit p-CREB by down-regulating the activity of EGFR downstream pathways in vivo and in vitro. A, The CREB and p-CREB levels in cell nucleus were detected using Western blot analysis when treated with DT-13 combined with TPT for 48 hours in BGC-823. B, The CREB and p-CREB levels in cell nucleus were detected using Western blot analysis when treated with the EGFR inhibitors and the EGFR downstream inhibitors in BGC-823 cells. C, The total/phospho AKT and total/phospho ERK1/2 levels in cell nucleus were detected using Western blot analysis when treated with DT-13 combined with TPT in BGC-823 cells and NM IIA knockdown BGC-823 cells. (D-E) The p-ERK and p-AKT levels in cell nucleus were detected using cell fluorescence analysis when treated with DT-13 combined with TPT in BGC-823 cells. F, The NM IIA, EGFR, p-CREB, HK II were detected using immunohistochemical analysis when treated with DT-13 combined with TPT in BGC-823 tumour tissue. Statistical analysis was performed using one-way ANOVA followed by Bonferroni's multiple comparison test, **P < 0.01; ***P < 0.001; for F, statistical analysis was performed using at least three independent replicates

| Effect of NM IIA on the activity of CREB in BGC-823 cells
We have confirmed that aerobic glycolysis was correlated to NM IIA ( Figure S2A,B,C) and that DT-13 combined with TPT suppressed HK II by up-regulating NM IIA ( Figure 3A,B). To further investigate the effect of NM IIA on HK II promoter activity, we examined the p-CREB, p-AKT, p-ERK levels in the cell nucleus of NM IIA knock-downed BGC-823 cells, where the p-CREB, p-AKT and p-ERK levels were all increased ( Figure S4A). By using cell fluorescence analysis, we also found the fluorescence intensity of p-AKT and p-ERK was increased in the cell nucleus of BGC-823 cells with NM IIA knock-downed or treated with (-)-blebbistatin ( Figure S4B,C). We also examined the p-AKT and p-ERK levels in the nucleus of NM IIA knock-down cells treated with DT-13-TPT combination. We found that NM IIA knock-down reversed the inhibition effect of the combinational treatment on p-AKT and p-ERK protein level in the cell nucleus ( Figure 5F). In tumour tissues, the positive areas for NM IIA were increased, while the positive areas for p-CREB, EGFR and HK II were reduced ( Figure 5G). These results demonstrated that DT-13 combined with TPT inhibited the entrance of p-ERK, p-AKT into the nucleus by up-regulating NM IIA, thereby inhibiting CREB activity, and ultimately inhibiting HK II promoter activity.

| Effects of the combination treatment on the binding of HK II to the mitochondria
The binding of HK II to mitochondria can combine aerobic glycolysis and oxidative phosphorylation together, not only make the hexokinase greater use of ATP, but also accelerate the glucose metabolism, providing more ADP for mitochondria and accelerate the tricarboxylic acid (TCA) cycle. 33,34 We found that the combinational treatment decreased HK II expression in mitochondria and increased HK II expression in cytosol ( Figure 6A). By using ImageJ software analysis, we also found the combinational treatment decreased the co-localization of HK II and mitochondria ( Figure 6B). We found that both (-)-blebbistatin treatment and NM IIA knockdown reversed the inhibition effect of combinational treatment on the binding of HK II to mitochondria ( Figure 6C,D).
HK II is the only kind of glycolysis enzyme that can bind to mitochondria, whose binding sites are on the mitochondrial outer membrane protein (VDAC) channel. Once bound, it will inhibit mitochondrial release of cytochrome C (Cyt C) and tumour cell apoptosis. 33 We used the co-immunoprecipitation assay to determine the effect of the combinational treatment on the binding of HK II to VDAC. The results showed that the combinational treatment inhibited the binding of HK II to VDAC ( Figure 6E). Meanwhile, the combinational treatment could promote mitochondrial release of Cyt C, which would be reversed in NM IIA knock-downed BGC-823 cells ( Figure 6F). These results demonstrated that DT-13 combined with TPT inhibited the binding of HK II to VDAC by up-regulating NM IIA, thereby inhibiting aerobic glycolysis activity. In recent years, many cancer-related pathways have found to be of profound effects on metabolism of cancer and oncogenic metabolic lesions may be selective targets for new anticancer therapeutics. 8,12,35 Renewed interest in the fact that cancer cells have to reprogramme their metabolism in order to proliferate or resist treatment. It must take into consideration that the ability of tumour cells to adapt their metabolism to the local microenvironment (low oxygen, low nutrients). 36 This variety of metabolic sources might be either a strength, resulting in infinite possibilities for adaptation and increased ability to resist chemotherapyinduced death, or a weakness that could be targeted to kill cancer cells. 7 In this study, we found that DT-13-TPT combination could inhibit the aerobic glycolysis-related enzymes' activity and inhibit F I G U R E 7 A schematic representation of a hypothesized mechanism of the synergistic combination of DT-13 and TPT inhibits human gastric cancer aerobic glycolysis via NM IIA/EGFR/HK II axis the binding of HK II to mitochondria, which have more effectively reverse the abnormal metabolic status towards a normal condition.

| D ISCUSS I ON
Moreover, DT-13 inhibited HIF-1α expression in our previous study, 15 and the HIF-1α expression was strongly correlated with glycolysis. 37 What's more, DT-13 combined with TPT inhibited the binding of HK II to VDAC, inducing the increase of Cyt C release. Hence, whether the aerobic glycolysis inhibition effect of the combination treatment was also induced by HIF-1α inhibition, and the fully suppression of aerobic glycolysis was the cause which induced of the mitochondrial apoptosis may all need to be explored in depth in the future.
The requirement of NM IIA for EGFR endocytosis was first reported by Kim JH.et al 23 and not further explored in depth and the binding of NM IIA and EGFR had not been applied to drug research until now. In our previous studies, DT-13 had the ability to increase NM IIA expression and might promote NM IIA binding to EGFR, thereby expediting the endocytosis of EGFR. 19 Our present study verified DT-13 combined with TPT could inhibit the activity of HK II based on the interaction between NM IIA and EGFR, which efficiently, indicating the potential of this novel mechanism as a target for drug development.
In our previous study, we have reported NM IIA was related to cell apoptosis 19 ; however, the correlation between NM IIA and cell aerobic glycolysis has not been reported previously. Our findings highlight for the first time this correlation of NM IIA with cancer cell aerobic glycolysis by using MYH-9 shRNA. These findings showed that NM IIA knocking down could increase lactate production and glucose uptake, and increase the activity of HK II and PKM2 ( Figure S2). According to the structure of NM IIA and our present data ( Figure S4), we also postulated that NM IIA might directly inhibit the activity of HK II in an EGFR independently way. Therefore, our research group will be investigating the whole picture of the contextual pathways between the NM IIA-mediated HK II inactivation process.

ACK N OWLED G EM ENTS
This work was financially supported by The "National Natural Science Foundation of China" (81573456, 81773766, and 81872892), "Double First-Class" University project (Cpu2018gy38) and The National High Technology Research and Development Program of China (2014AA022208). We are highly thankful to both organizations for this sponsorship.

CO N FLI C T O F I NTE R E S T
No potential conflicts of interests were disclosed.

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 from the corresponding author upon reasonable request.