Glycolysis is essential for chemoresistance induced by transient receptor potential channel C5 in colorectal cancer

The wild human CRC cell line HCT-8 (KG028) and LoVo (SCSP-514) were purchased from Keygen Biotech Co. Ltd. (Nanjing, Jiangsu Province, China) and the Cell Resource Center of Shanghai Institutes for Biological Sciences, Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China) respectively. Fluorouracil (5-Fu)-resistant HCT-8 cells (HCT-8/5-Fu) (KG333) was purchased from Keygen Biotech Co. Ltd. 5-Fu-resistant LoVo cells (LoVo/5-Fu) were derived by treating LoVo cells with stepwise increasing concentrations of 5-Fu (Jinyao Amino Acid Co. Ltd., Tianjin, China) over 6 months. The wild human CRC cells and 5-Fu-resistant CRC cells were cultured as we reported previously [2].

HCT-8/5-Fu cells on 50–70% confluence were treated with TRPC5-shRNA (sc-42,670, Santa Cruz Biotechnology, Dallas, TX, USA) (HCT-8/5-Fu/RNAi) (scrambled siRNA as control, HCT-8/5-Fu/Scrambled). 3-bromopyruvate (3-BP) (SML2000, Sigma Chemical Co., St. Louis, MO, USA) (40 μM, 24 h) or 2-Deoxy-D-glucose (2DG) (D8375, Sigma Chemical Co.) (20 mM, 24 h) was used to inhibit the glycolysis. If needed, cells were treated with BAPTA-AM (A1076, Sigma Chemical Co.) (20 μM, 1 h) before glycolysis inhibition. Expression of TRPC5 and GLUT1 were deternmined by Real-time PCR and western blot.

Whole-cell protein was obtained using RIPA containing 1 mM PMSF. An equal quantity of total proteins was electrophoresed on 8% polyacrylamide gel containing 0.1% SDS and then transferred to PVDF membrane. After blocked with phosphate-buffered saline tween containing 5% non-fat milk, the PVDF membranes were incubated with the primary antibodies anti-TRPC5 (ACC-020, Alomone labs, Jerusalem, State of Israel) (1:500), anti-caspase-3 (ab32351, Abcam Biotechnology, Cambridge, MA, USA) (1:500), anti-glucose transporter 1 (GLUT1) (ab115730, Abcam Biotechnology) (1:1000), β-actin (AA128, Beyotime Biotechnology) (1:1000) and subsequently with the corresponding secondary antibodies [goat anti-rabbit IgG (A0208, Beyotime Biotechnology) and goat anti-mouse IgG (A0216, Beyotime Biotechnology)]. The bands were quantified using ImageJ software (NIH, Bethesda, MD). β-actin was used as the internal control for normalization.

TRIzol (10296–010, Camarillo, CA, USA) was used to extract total RNA from cells. Real-time PCR and the comparision of the mRNA levels were performed according to the reported study [2]. Table 1 listed the primer pairs used in this study.

Twelve hours after 10⁴ CRC cells (200 μl) seeded in 96-well plates, the cells were treated with 5-Fu of different concentrations. After 48 h, the cells in each well were incubated with resh RMPI1640 (200 μl) containing 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyl tetrazolium bromide (MTT) (M2128, Sigma Chemical Co.) (5 mg/ml) for 4 h. Dimethyl sulfoxide (DMSO) (D8418, Sigma Chemical Co.) (150 μl) was added to each well and then the absorbance was detected at 490 nm.

We used GECO1.2 (a calcium indicator) to measure [Ca²⁺] _i level. The fluorescence signals of GECO1.2 reflected the [Ca²⁺] _i levels. The detailed procedure was in accordance with the previously reported study [2].

About 1 × 10⁶ cells were seeded in 6-well cell culture microplates. The medium was replaced with 3 ml RMPI-1640 without fetal calf serum the next day. Twenty four hours later, the medium was collected and the glucose concentration in the medium was determined according to Glucose (HK) kit (GAHK-20, Sigma Chemical Co.). Glucose consumption rate was defined as the ratio of the glucose concentration after twenty four hours divided by the glucose concentration before twenty four hours.

On reaching 50–70% confluence, ells seeded in 6-well cell culture microplates were treated with lysis reagent to release ATP. The supernatant was obtained to measure ATP according to the manufacturer’s protocol (S0026, Beyotime Biotechnology).

Quantitation of apoptotic cells was obtained using the Annexin V-FITC/PI Apoptosis Detection Kit (C1062, Beyotime Biotechnology) according to the manufacturer’s protocol. Cells in logarithmic phase were detached to obtain a single cell suspension. After a total of 1 × 10⁶ cells were washed in PBS for 2 times, 195 μL of binding buffer solution was added for cell resuspension. Then 5 μL of annexin V-FITC and 10 μL of PI were added into culture solution for mixing, with incubation 30 min at 4 °C. Flow cytometry was used to make a comparison of the apoptotic cells ratio.

Ethical permission was obtained from the Ethics Committee at the Affiliated Hospital of Jiangnan University and conformed to the provisions of the Declaration of Helsinki (as revised in Fortaleza, Brazil, October 2013). The advanced CRC patients who received a biopsy and/or surgery for a primary lesion and postoperatively 5-Fu based first-line systematic chemotherapy at the Affiliated Hospital of Jiangnan University from January 2010 to December 2016 were enrolled in this study. The exclusion criteria was according with our previous study [12]. Treatment response was evaluated according to the Response Evaluation Criteria in Solid Tumors 1.1 (RECIST 1.1) guideline [13] after 2 cycles of chemotherapy. Patients achieved progressive disease (PD) or stable disease (SD) were considered as non-responders, and patients achieved partial response (PR) or complete response (CR) were considered as responders. Immunohistochemistry staining was performed to detect TRPC5 and GLUT1 protein expression in CRC tissue slides with the procedure we reported previously [12]. The results were judged according to German semi-quantitative scoring system [14] (no staining = 0; weak staining = 1, moderate staining = 2, strong staining = 3) and the extent of stained cells (0% = 0, 1–24% = 1, 25–49% = 2, 50–74% = 3, 75–100% = 4). The final score was determined by multiplying the intensity score with the extent score, ranging from 0 to 12. Each grade of TRPC5 and GLUT1 were from the same sample.

The MTT assay was performed to determine the half maximal inhibitory concentration of 5-Fu (5-Fu IC50) of the CRC cells. HCT-8/5-Fu (5-Fu IC50: 122.3 mg/L) and LoVo/5-Fu (5-Fu IC50: 44.76 mg/L) showed more resistance to cytotoxicity of 5-Fu than in HCT-8 (5-Fu IC50: 13.8 mg/L) and LoVo (5-Fu IC50: 2.611 mg/L) cells (Fig. 1a). Further real time PCR and western blot showed a much higher expression of TRPC5 in HCT-8/5-Fu and LoVo/5-Fu cells than their parental lines (Fig. 1b, c). Overproduced GLUT1 was reported to be essential for the increased glucose import in aerobic glycolysis in cancer [7, 15, 16]. In this study, glycolytic activity was determined by examination of GLUT1 expression, glucose consumption and celluar ATP production. Real-time PCR and western blot showed a much higher expression of GLUT1 in 5-Fu-resistant CRC cells than in their parental lines (Fig. 1b, c). Additionally, 5-Fu-resistant CRC cells showed higher glucose consumption rates and more ATP production than the wild type cells (Fig. 1d, e).

The roles of trp channels in cancer include changes in [Ca²⁺] _i level [17]. TRPC5 is a nonselective cation channel with Ca²⁺ permeability [18]. In our previous study [2], TRPC5 was proven to be required for the increase of [Ca²⁺] _i in HCT-8/5-Fu cells. In present study, according with the up-regulated expression of TRPC5, the level of [Ca²⁺] _i in HCT-8/5-Fu cells was higher than in HCT-8 cells. Further inhibition of TRPC5 by shRNA resulted in decreased TRPC5 protein expression (Fig. 2a) along with dramatically decreased [Ca²⁺] _i level (Fig. 2b).

Several studies showed up-regulated expression of trp proteins [3] could lead to [Ca²⁺] _i overload, which was demonstrated to induce apoptosis [3, 19, 20]. With regard to the augmentation of Ca²⁺ influx through up-regulated expression of TRPC5, ATP-dependent Ca²⁺ efflux should be crucail to prevent [Ca²⁺] _i overload related apoptosis. Since reprogramed energy metabolism to glycolysis was demonstrated to be the major mechanism of generating ATP [4, 7, 21] and the major ATP source for Ca²⁺ efflux in cancer [4], we explored the potential mechanism of [Ca²⁺] _i homeostasis in chemoresistance induction by TRPC5. 3-BP and 2DG, inhibitors of glycolysis [10, 22], were used to inhibit glycolysis in human CRC cells. Administration of 3-BPor 2DG caused a remarkable ATP production decrease and increasement of [Ca²⁺] _i level in HCT-8/5-Fu cells, while caused no obvious change in ATP production and [Ca²⁺] _i level in HCT-8 cells (Fig. 3a, c). In addition, western blot showed administration of 3-BP or 2DG increased cleaved Caspase-3 in HCT-8/5-Fu cells significant, while showed little change of cleaved Caspase-3 level in HCT-8 cells (Fig. 3b). FCM showed administration of 2DG apoptotic cells rate obviously increased in HCT-8/5-Fu cells (Fig. 3d). Further study showed the elevated cleaved Caspase-3 and apoptotic cells rate induced by 2DG dramatically decreased with the administration of BAPTA-AM (Fig. 3b, d). MTT assay showed the 5-Fu IC50 of HCT-8/5-Fu cells treated with 3-BP (HCT-8/5-Fu/3-BP) or 2DG (HCT-8/5-Fu/2DG) decreased to 44.7 ng/ml (95%CI: 36.8–48.6 mg/L) and 38.48 ng/ml (95%CI: 30.64 to 48.33 mg/L), while no significant change of 5-Fu IC50 of HCT-8 cells treated with 3-BP (HCT-8/3-BP) was observed (Fig. 1a). Similar results were obtained in experiments in LoVo/5-Fu cells. The 5-Fu IC50 of LoVo/5-Fu cells dramatically decreased to 27.77 mg/L (95%CI: 23.92–32.23 mg/L) in LoVo/5-Fu/2DG cells (Fig. 1a).

As was shown in Table 2, among the 147 advanced CRC patients enrolled in this study, 53 patients achieved CR/PR (responders) and 94 patients achieved SD/PD (non-responders) after chemotherapy. Different levels of TRPC5 and GLUT1 protein were observed in tumor tissues from different CRC patients (Fig. 4). ROC analysis identified 5 and 6.3 as the optimal cutoff value of TRPC5 score and GLUT1 score respectively to discriminate responders from non-responders (Fig. 5). Pearson’s chi-squared test showed the positive correlation between TRPC5 and GLUT1 protein levels and a high TRPC5/GLUT1 expression was closed correlated with chemoresistance (Table 3), which was consistent with our previous findings [12]. Interesting, high TRPC5 expression was found to be significantly associated with chemoresistance only in case of high GLUT1 expression, while no association was observed between TRPC5 expression and chemotherapy outcome in the case of low GLUT1 expression (Table 4).