The role of TRPV1 ion channels in the suppression of gastric cancer development

Twenty pairs of GC and adjacent tissues from the surgical patients in Xinqiao Hospital of Third Military Medical University (during 2016 and 2017) were used for real-time quantitative PCR, and all resected specimens were confirmed by pathological examination. Informed consent was obtained from all patients. All clinical studies were approved by the Clinical Research Ethics Committee of Third Military Medical University and were performed in accordance with approved guidelines. GC and adjacent tissue microarray for immunostaining were purchased from SHANGHAI OUTDO BIOTECH CO., LTD (Shanghai, China).

MKN45, SGC7901, AGS, MGC803, BGC823 human gastric cancer cell lines and the human gastric normal epithelial mucosa cell line (GES-1) were purchased from Chinese Academy of Sciences (Shanghai, China). All cells were cultured in RPMI-1640 or DMEM-HIGH GLUCOSE medium (HyClone, USA) supplemented with 10% fetal bovine serum (HyClone, USA), 100 IU/mL penicillin and 100 μg/mL streptomycin (Invitrogen, USA). All cells were grown in a 37 °C humidified atmosphere containing 5% CO₂.

Lentiviruses were purchased from Genechem Co., Ltd. (Shanghai, China). A lentivirus containing the full-length coding sequence (CDS) of TRPV1 (NM_080704) was designed to increase its expression in BGC823 cells, and lentiviral-based shRNA was used to silence the expression of TRPV1 in MKN45 cells. Sequences for TRPV1 shRNA and control were as follows: shRNA-1 (5′-GCATCTTCTACTTCAACTTCC-3′), shRNA-2 (5′-GGCCGACAACACGAAGTTTGT-3′) and control (5′-GTTCTCCGAACGTGTCACGT-3′). All shRNA groups that do not have a designed number used shRNA-1. Cells were infected with lentiviruses according to the protocol of the manufacturer. Briefly, cells were plated in 24-well plates at 1 × 10⁵ cells/well, lentiviruses were added into culture medium separately (the volume of lentiviruses was calculated as a MOI of 20), and medium were refreshed after 8 h. Puromycin was used to screen the stable cells after 72 h of lentivirus infection.

The full length CDS of AMPK (NM_006251) was cloned into pcDNA3.1 to prepare overexpression plasmids; AMPK-siRNA sequence 5′-CTGCTTGATGCACACATGAAT-3′; CaMKKβ-siRNA sequence 5′- GTCAAGTTGGCCTACAATG-3′ were cloned into the GV102 vector separately to prepare siRNA-knocked down plasmids. Transfection of plasmids into cells was performed according to the protocol of FuGENE@ HD Transfection Reagent (Cat. No. E2311, Promega, USA). In brief, 1 × 10⁵ cells were cultured in each well of 24-well plates, and 1.5 μL of transfection reagent with 0.5 μg of plasmids were mixed and added to 500 μL culture medium. Cells were incubated with the mixture for 72 h.

Total RNA was extracted from each group using the RNAiso Plus reagent (Cat. No. 9109, Takara, Japan). cDNA was synthesized using PrimeScript® RT-polymerase (Cat. No. R050A, Takara, Japan). Next, 50 ng of each cDNA was amplified as a template, and qPCR was performed using a SteponePlus device (Art. No. 272008342, Life Technologies, USA) with a SYBR® Premix Ex TaqTM II kit (Cat. No. RR820A, Takara, Japan). All samples were run in triplicate, and β-actin was used as an internal control. Data were quantified using the 2^−ΔΔCt relative quantitative method, normalized to β-actin expression, and expressed as the ratio of TRPV1 to β-actin mRNA levels. Primers were designed as follows:

TRPV1: 5′-TGGTATTCTCCCTGGCCTTG-3′ (forward)

5′- CTTCCCGTCTTCAATCAGCG-3′ (reverse)

AMPK: 5′-TGGTAGGAAAAATCCGCAGA-3′ (forward)

5′-CGACTTTCTTTTTCATCCAGC-3′ (reverse)

β-actin: 5′-GGCATCCACGAAACTACCTT-3′ (forward)

5′-CGGACTCGTCATACTCCTGCT-3′ (reverse)

Tissue samples were paraffin embedded and cut into 5 mm slices. After dewaxing and rehydration, tissue samples were incubated with anti-TRPV1 (Cat. No. ab3487, Abcam, UK) overnight at 4 °C after blocking. TRPV1 was detected using HRP-conjugated anti-rabbit secondary antibody (Cat. No. ZB-2301, ZSGB-BIO, China) and visualized with DAB. The negative control contained secondary antibody only. The gastric cancer tissue microarray we purchased also contains pathological score data for Ki67, VEGFR, E-cadherin and other proteins. The degree of staining in the TRPV1 sections was observed and scored by a pathologist. According to previously defined criteria [31, 32], the percentage of TRPV1 positivity was scored from 0 to 3 as follows: 0, < 10%; 1, 10–30%; 2, 30–50%; 3, > 50%. Staining intensity was scored according to a 4-point scale as follows: 0 (no staining); 1 (weak staining, light yellow); 2 (moderate staining, yellowish brown); and 3 (strong staining, brown). Subsequently, TRPV1 expression was calculated by multiplication of the percent positivity score and staining intensity score, resulting in a final score ranging from 0 to 9. Since TRPV1 was expressed as 0 in a large number of gastric cancer samples, expression of 0 was set as low expression and 1–9 as high expression. Ki67 was positively localized in the nucleus, VEGFR was positively localized in the cytoplasm and E-cadherin was positively localized in the plasma membrane. The scoring standard was the same as that described above.

Cells of each group were plated onto coverslips in 35 mm dishes and fixed in 4% polyformaldehyde for 15 min at room temperature. Coverslips were washed in PBS three times for 5 min. Cells were blocked in goat serum for 1 h at room temperature and then incubated with anti-TRPV1 antibody overnight at 4 °C. After three washes in PBS, cells were incubated with Cy3 labeled anti-rabbit (Cat. No. A0516, Beyotime, China) secondary antibody for 1 h at room temperature. Finally, nuclei were stained with DAPI for 10 min. Images were captured on a confocal microscope (Leica SP5, Germany).

Whole-cell lysates were separated by SDS-PAGE on denaturing 10% or 12% gels and transferred to polyvinylidene fluoride membranes (Cat. No. ISEQ00010, Millipore, USA). Blots were blocked in 5% milk for 1 h at room temperature and then separately incubated at 4 °C overnight with the following specific primary antibodies: anti-TRPV1, anti-Ki67 (Cat. No. ab15580, Abcam, UK), anti-AMPK (Cat. No.5832, Cell Signaling Technology, USA), anti-phospho-AMPK (Cat. No.2535, Cell Signaling Technology, USA), anti-CaMKKβ (Cat. No. ab168818, Abcam, UK), anti-cyclin D1 (Cat. No. ab16663, Abcam, UK), anti-MMP2 (Cat. No.87809S, Cell Signaling Technology, USA), anti-β-catenin (Cat. No.8480, Cell Signaling Technology, USA), anti-phospho-β-catenin (Cat. No.9567, Cell Signaling Technology, USA), anti-AKT (Cat. No.9272, Cell Signaling Technology, USA), anti-phospho-AKT (Cat. No.4060, Cell Signaling Technology, USA), anti-ERK1/2 (Cat. No. ab184699, Abcam, UK), anti-phospho-ERK1/2 (Cat. No. ab214362, Abcam, UK), and anti-GAPDH (Cat. No. TA-08, ZSGB-BIO, China). All primary antibodies were diluted 1:1000. After rinsing, blots were incubated in HRP-conjugated anti-rabbit or anti-mouse (Cat. No. A0239 and A0216, Beyotime, China) secondary antibodies for 1 h at room temperature. Enhanced chemiluminescence (Cat. No. 34094, Thermo, USA) was used to detect immunoreactive bands. A human phospho-kinase array was used to detect changes in tumor-related signaling pathways (Cat. No. ARY003B, R&D Systems, USA), and 2 μM BAPTA-AM (Cas No. 126150–97-8, MedChemExpress, USA) was used to chelate intracellular calcium with cells being treated with BAPTA-AM for 2 h. Each experiment was performed in triplicate and repeated three times. The gray value of the bands was measured by ImageJ software for statistics.

Cell viability and proliferation were measured by CCK8 assay (Cat. No.C0038, Beyotime Biotechnology, China). Cells were plated in 96-well plates (3000 cells/well) in triplicate and cultured for 0, 24, 48, 72 h or 96 h followed by addition of 100 μl of medium/CCK8 mixture (medium: CCK8, 9:1) 1–2 h before the endpoint of incubation. A Multiskan EX plate reader (Thermo Fisher Scientific, Germany) was used to quantify viable cells by measuring the absorbance at 450 nm, which estimates relative cell numbers rather than counting cells.

Cells were digested, centrifuged and washed twice in cold PBS. The supernatant was subsequently discarded, and pre-cooled 70% ethanol was slowly added to the cell pellets that were kept at 4 °C overnight. After centrifugation the next day, the ethanol was removed, cells were washed once with PBS and were then incubated in a solution containing 0.2% Tween 20, 100 U/mL RNase, and 50 μg/mL propidium iodide for 20 min at 37 °C. Cell cycle analysis was performed using flow cytometry in which samples were gated on live cells with an excitation wavelength of 488 nm and an emission wavelength of 620 nm. LMD files were further analyzed using ModFit LT (Verity Software House, Topsham, ME).

Long-term survival of cells was assessed by their ability to form colonies. Cells were plated in 6-well plates with 3 mL culture medium with 500 cells/well. After 10–12 days, the cell culture medium was removed, and cell clones were washed in PBS and fixed in 4% polyformaldehyde. Clone numbers were quantified after staining with crystal violet (Relative clonogenicity = Clone numbers/Average clone numbers of control group).

Twenty-four-well transwell chambers (Corning, USA) were used for this assay. 5 × 10⁴ cells were plated into each upper chamber with 8-μm pores and cultured in 200 μL serum-free RPMI-1640. The lower chambers were filled with 500 μL complete RPMI-1640 medium. After 24 h of incubation, cells that had migrated onto the lower surface were stained with crystal violet, and counted under a microscope (Olympus Corporation, Japan). The average value of three randomly selected fields was recorded as the number of migrated cells. Then, the upper surface of the polycarbonate filter was coated with 10% Matrigel (Collaborative Biomedical, USA), and 1 × 10⁵ cells were added to detect cell invasion. The other conditions were the same as those in the migration assay.

Plating 1 × 10⁴ cells on coverslips, cells were loaded with 5 μM Fura-2 AM (Cat. No. F1221, Invitrogen, USA) in physiological salt solution (PSS) at 37 °C for 60 min and then washed in PSS or PSS with and 50 μM of the TRPV1 inhibitor SB-705498 (Cas No.501951–42-4, MCE, USA). Next, cells on coverslips were mounted in a standard perfusion chamber on a Nikon microscope stage. The ratio of Fura-2 fluorescence at 340 or 380 nm excitation (F340/380) was followed over time and captured with an intensified CCD camera (ICCD200) and a MetaFluor Imaging System (Universal Imaging, Downingtown, PA). PSS used in Ca²⁺ measurement contained the following: 140 mM Na⁺, 5 mM K⁺, 2 mM Ca²⁺, 147 mM Cl⁻, 10 mM HEPES, and 10 mM glucose at pH 7.4.

The animal use protocol was approved by the Third Military Medical University Committee on Investigations Involving Animal Subjects. All animal care and experimental studies were conducted in accordance with the guidelines of the Animal Ethical Committee of Third Military Medical University and the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No.8023, revised 1978). Animal studies were reported in compliance with the ARRIVE guidelines. First, 1 × 10⁶ TRPV1-overexpressing BGC823 cells and negative control (NC) cells were injected into the armpits of seven 4-week old male nude mice for tumor xenograft assay. TRPV1-overexpressing BGC823 cells were injected into the right armpit, and NC cells were injected into the left side. After 30 days of implantation, mice were sacrificed and tumor volumes in mm³ were calculated using the formula 1/2(length× width²). For the peritoneal dissemination assay, 1 × 10⁶ TRPV1-overexpression BGC823 cells and the NC cells were injected into the abdominal cavity of nude mice. Five weeks later, mice were sacrificed, and nodules were observed and quantified.

The aberrant expression of ion channels is usually associated with the development of several human cancers. Since TRPV1 expression in human GC is lacking, we first predicted its expression from Oncomine tumor database, which indicated low expression of TRPV1 in GC (data not shown). Second, we collected human primary GC tissues and corresponding adjacent tissues to compare TRPV1 expression. As shown in Fig. 1a, TRPV1 expression was lower at transcriptional levels in human GC tissues than in adjacent tissues. Therefore, TRPV1 expression was decreased in human primary GC, consistently with the prediction from Oncomine tumor database.

Third, immunohistochemstry study was applied to detect protein expression of TRPV1 in human specimens from 100 patients with GC. Among these patients, their average age was 64 years old, 64% was male, 57% was diagnosed with advanced-stage (III/IV), and 73% had lymphatic metastasis (Table 1). As shown in Fig. 1b and c, the protein expression of TRPV1 was decreased in GC tissues compared to their adjacent tissues, but no staining was detected in the negative control without primary antibody, indicating its specific staining to TRPV1 proteins.

To examine the involvement of TRPV1 in GC progression and development, the association between TRPV1 expression and clinicopathologic parameters of GC progression was subsequently analyzed. As shown in Table 1, the down-regulation of TRPV1 expression was correlated with the large tumor size (P < 0.001), high histological grade (P = 0.006), lymphatic metastasis (P = 0.001), and advanced clinical stage (P < 0.001), but not associated with patients’ age (P = 0.476), gender (P = 0.223). Furthermore, Kaplan-Meier analysis showed that the median survival times of GC patients with high and low TRPV1 expression were 63 and 22 months, respectively, indicating that the GC patients with high TRPV1 expression had a better prognosis, but those with low expression had a poor prognosis (P = 0.008, Fig. 1d). Therefore, the close association between TRPV1 expression and clinicopathologic parameters strongly suggests a role for TRPV1 in the progression and development of GC.

Since Ki67, VEGFR, and E-cadherin are well-recognized tumor markers for proliferation and metastasis of cancer cells, we further analyzed the association between these tumor markers and TRPV1 expression in GC. As shown in Fig. 1e–g, TRPV1 expression was negatively correlated with Ki67 (r = − 0.2001, P = 0.0471) and VEGFR (r = − 0.3291, P = 0.0009), but positively correlated with E-cadherin (r = 0.4279, P < 0.0001). Togetether, these data suggest that TRPV1 might be a tumor suppressor in human GC.

After analyzing the association between TRPV1 expression in human primary GC tissues and clinicopathologic parameters, we compared TRPV1 expression level among 5 GC cell lines (MKN45, SGC7901, AGS, MGC803 and BGC823) and a normal gastric mucosal cell line (GES-1). QPCR study showed that TRPV1 was expressed at varying degrees at the levels of transcripts in aforementioned all cell lines (Fig. 1i), which was further confirmed at the levels of proteins by immunoblotting analysis (Fig. 1h). The expression of TRPV1 was consistent at the levels of transcripts and proteins among these cell lines (Fig. 1h and i). Compared with normal GES-1 cells, MKN45 cells had the highest expression of TRPV1 but BGC823 cells had the lowest expression, indicating an obvious heterogeneity of GC cells [33].

We also performed immunofluorescence analysis to study the expression and localization of TRPV1 proteins. As shown in Fig. 1j, it was further confirmed for the highest TRPV1 expression in MKN45 cells but the lowest expression in BGC823 cells. Moreover, TRPV1 proteins were predominately expressed on the plasma membrane, but also in the cytoplasm at a small amount, consistently with the previous report that TRPV1 also existed in the endoplasmic reticulum [34]. Therefore, MKN45 and BGC823 cells were selected for further molecular and cell biological studies in the following experiments.

Our clinicopathologic study suggested that TRPV1 might be a human GC suppressor, and therefore its inhibition on GC cell proliferation was tested since high proliferation is a major characteristic of cancer cells. To this end, we applied lentiviruses infection to make TRPV1-overexpressed BGC823 cells and TRPV1-knockeddown MKN45 cells. As shown in Fig. 2a as representative image and Fig. 2b and c as summary data (n = 3 for each column), after successful overexpression or knockdown of TRPV1 in BGC823 or MKN45 cells at mRNA level (Fig. 2d) and protein level (Fig. 2a and b), respectively; proliferation marker Ki67 was decreased in TRPV1-overexpressed BGC823 cells but increased in TRPV1-knockeddown MKN45 cells compared with their corresponding negative controls (NC) (Fig. 2a and c). Therefore, our results obtained from GC cells are consistent with those obtained from primary GC tissues (Fig. 1e). Parallelly, TRPV1 overexpression decreased cell viability while TRPV1 knockdown increased it (Fig. 2e). Moreover, plate cloning experiments confirmed the inhibitory role of TRPV1 on clonogenicity in GC cells (Fig. 2i and j) but not in normal GES-1 cells with lower TRPV1 expression (Fig. 2h).

We further examined the effect of capsaicin (CAP), a well-known selective TRPV1 activator, on GC cell proliferation. As shown in Fig. 2f, although either CAP (50 μM) or TRPV1 overexpression alone inhibited GC cell proliferation with similar potency, a combination of them could not result in any superimposed inhibition. Moreover, CAP still inhibited proliferation in TRPV1-knockeddown MKN45 cells (Fig. 2g). Therefore, CAP suppression of GC cell proliferation was TRPV1-independent, as reported previously [35, 36]. Since capsaicin works on GC cells in a TRPV1-independent manner, we focused on the role of TRPV1 instead of CAP in the following experiments.

To further verify the role of TRPV1 in GC growth in vivo, xenografted GC model of nude mice was applied. In this model, the overexpression of TRPV1 in BGC823 cells markedly suppressed growth ability of GC cells after their implantation, leading to significant decreases in both tumor weights by about 52% and tumor volume by about 60% compared to their NC groups (Fig. 3a and b). Therefore, TRPV1 could not only inhibit GC cell proliferation in vitro but also suppress tumor growth in vivo, confirming our early notion that TRPV1 acts as a tumor suppressor in GC.

We further elucidated the underlying mechanisms of how TRPV1 inhibits GC cell proliferation in vitro and tumor growth in vivo. Since the cell cycle is an essential biological process that controls cell proliferation, we investigated how TRPV1 controls the GC cell cycle through genetic manipulation of TRPV1 expression and flow cytometry analysis [37]. As shown in Fig. 3c and d, compared to NC cells, the number of TRPV1-overexpressed BGC823 cells in G1 phase was increased by around 24%, and in S phase was decreased by around 31%. In contrast, the number of TRPV1-knockeddown MKN45 cells in G1 phase was decreased by around 30–40%, and in S phase was increased by around 50–60% (Fig. 3e and f). However, the G2 phase of the cell cycle were not significantly altered in response to manipulation of TRPV1 expression in GC cells. Therefore, mechanistically, TRPV1 blocked cell cycle at G1 phase to inhibit GC cell proliferation in vitro and tumor growth in vivo.

Since high migration and invasion of GC cells are not only their important characteristics but also major reasons to cause high mortality [38], we examined the role of TRPV1 in regulating GC cell migration and invasion. Our transwell assays showed that cell migration was reduced in TRPV1-overexpressed BGC823 cells, but enhanced in TRPV1-knockeddown MKN45 cells (Fig. 4a and b). Furthermore, changes in cell migration in both groups were over 50% compared to their NC groups (Fig. 4a and b). The trend of GC cell invasion was similar to that of migration (Fig. 4c and d), and the invasion in both groups was significantly altered at least 50% (Fig. 4c and d).

We applied abdominal transplantation tumor model of nude mice to verify the role of TRPV1 in suppressing GC invasion in vivo, and found that the overexpression of TRPV1 in BGC823 cells markedly suppressed GC cell metastasis after their peritoneal implantation (Fig. 4e), leading to a decrease in tumor numbers by about 40% compared to NC group (Fig. 4f). Therefore, TRPV1 suppressed GC cell migration and invasion both in vitro and in vivo.

Since various cell signaling pathways are involved in GC tumorigenesis [39, 40], we initially screened TRPV1-mediated downstream signaling in GC cells. After screening the phosphorylation of 40 key proteins which are known to be closely related to tumorigenesis, we found that the phosphorylation of AMPK was significantly increased in TRPV1-overexpressed BGC823 cells compared to their corresponding negative controls (Fig. 5a), suggesting that AMPK was most likely involved. In addition, we assessed whether other key signaling proteins were also likely involved, including ERK1/2, β-catenin and AKT. As shown in Fig. 5b, TRPV1 overexpression attenuated but TRPV1 knockdown enhanced the phosphorylation of ERK1/2, suggesting TRPV1 suppresses ERK1/2 signaling in GC cells. However, due to the well-documented oncogenic role of ERK1/2 signaling in many cancers, including GC, we did not focus on it in further study. The other two signaling proteins, β-catenin and AKT were not significantly altered in response to genetic manipulation of TRPV1, excluding their involvements in TRPV1-mediated downstream signaling in GC cells (Fig. 5c and d).

We next focused on AMPK signaling since it is most likely involved in TRPV1-mediated downstream pathways in GC cells. This notion was further supported by investigation of AMPK phosphorylation. As shown in Fig. 5f and g, the phosphorylation levels at Thr-172 site of AMPK were markedly increased or decreased after TRPV1 was overexpressed in BGC823 cells or knocked down in MKN45 cells, respectively, compared to their corresponding NC; however, mRNA levels of AMPK were unchanged in both cell types (Fig. 5e). Therefore, TRPV1 likely stimulates AMPK phosphorylation in GC cells.

We further examined the critical role of AMPK in TRPV1 activation of GC cells after constructing AMPK overexpression and interference plasmids. As shown in Fig. 5f and g, compared to their corresponding NC, AMPK phosphorylation was markedly decreased by AMPK interference in TRPV1-overexpressed BGC823 cells, or increased by AMPK overexpression in TRPV1-knockeddown MKN45 cells. Therefore, TRPV1 indeed stimulates AMPK phosphorylation, which is critical in TRPV1-mediated signaling in GC cells.

We further verified the role of AMPK in proliferation, migration and invasion of GC cells. As shown in Fig. 6a, reduced proliferation in TRPV1-overexpressed BGC823 cells could be recovered by siAMPK, but enhanced proliferation in TRPV1-knockeddown MKN45 cells could be reduced by AMPK overexpression. Similarly, reduced migration and invasion in TRPV1-overexpressed BGC823 cells were recovered by siAMPK, but the enhanced migration and invasion of TRPV1-knockeddown MKN45 cells were reduced by AMPK overexpression (Fig. 6e and f). Therefore, TRPV1 inhibited GC cell proliferation, migration and invasion through AMPK activation.

We elucidated the underlying mechanisms of how TRPV1-activated AMPK suppresses proliferation, migration and invasion in GC cells. Since both cyclin D1 and matrix metalloproteinase-2 (MMP2) play important roles in controlling the cell cycle and metastasis [41, 42], we investigated whether they are involved in TRPV1/AMPK-mediated suppression of GC cells. Western blotting analysis exhibited that compared to their corresponding NC, the expression of both cyclin D1 and MMP2 was decreased in TRPV1-overexpressed BGC823 cells, which were recovered by AMPK-siRNA (Fig. 6b–d). In contrast, the expression of both cyclin D1 and MMP2 was enhanced in TRPV1-knockeddown MKN45 cells, which was attenuated by AMPK overexpression (Fig. 6b–d). Therefore, TRPV1-activated AMPK suppressed GC development by inhibiting both cyclin D1 and MMP2.

TRPV1 is a Ca²⁺–permeable channel [10] and CaMKKβ is a downstream kinase of calmodulin (CaM), a well-known intracellular Ca²⁺ binding protein [43]. Since CaMKKβ is an upstream AMPK activator [25], we hypothesized that AMPK activation requires the Ca²⁺/CaMKKβ in GC cells. To test this hypothesis, we first measured [Ca²⁺]_i in GC cells. Because CAP (50 μM) did not affect basal [Ca²⁺]_i in GC cells (data not shown), it is not a useful tool for the study on TRPV1 channel in these cells, as reported previously [35, 36]. Since TRPV channels are usually stimulated by GPCR activation, we stimulated Ca-sensing receptor (CaSR) with calcium (5 mM) for this purpose in GC cells [7]. Indeed, CaSR-mediated Ca²⁺ signaling was significantly increased in TRPV1-overexpressed BGC823 cells, which could be inhibited by SB-705498, a specific inhibitor of TRPV1 (Fig. 7a). In contrast, CaSR-mediated Ca²⁺ signaling was significantly decreased in TRPV1-knockeddown MKN45 cells (Fig. 7b). Thus, TRPV1 is a functional Ca²⁺–permeable channel to causes Ca²⁺ entry in GC cells.

We then examined the effect of TRPV1/Ca²⁺ on CaMKKβ, and found that the protein expression of CaMKKβ was markedly increased in TRPV1-overexpressed BGC823 cells, but decreased in TRPV1-knockeddown MKN45 cells compared with their NC (Fig. 7c). Furthermore, the increased CaMKKβ expression could be attenuated either by CaMKKβ-siRNA (Fig. 7d) or by pretreatment with [Ca²⁺]_i chelator, BAPTA-AM (2 μM) (Fig. 7e) in TRPV1-overexpressed BGC823 cells, in which the increased AMPK phosphorylation could be parallelly attenuated by CaMKKβ-siRNA (Fig. 7f). Therefore, TRPV1/Ca²⁺/CaMKKβ pathway is essential for AMPK phosphorylation in GC cells.