Introduction

Gastric cancer is the fifth most common and the third most lethal malignancy in the world with 951,000 newly diagnosed cases and 723,000 deaths per year [1–3]. Due to its asymptomatic nature and late clinical presentation, gastric cancer is often diagnosed at advanced stages [4, 5], when it disseminates into peritoneum or to distant sites via the lymphatic system and blood [6]. Surgical resection for <stage 1A and combined modality approaches for ≥stage 1B diseases remain the standard therapies although targeted therapeutics are emerging [7]. Recurrences are common with 40% of patients relapse within 2 years of surgery [6, 8, 9]. Thus, more effective and precise therapeutic strategies are needed to improve the patient outcome.

Precise medicine has been on the centre stage of cancer research as a result of the rapid development of high throughput genomic analysis in the past decade. Mutations or dysregulations of a number of oncogenes like KRAS, CTNNB1, ERBB2, PIK3CA [10–13] and tumour suppressor genes, such as CDH1, CDKN2A, TP53, and ARID1A [14–17] are associated with gastric cancer, and have been the basis for multiple clinical trials of many targeted agents such as trastuzumab and cetuximab in patients with gastric cancer [7]. Accumulating evidence suggests that annexins, which are a group of Ca2+ and phospholipid binding proteins, are frequently dysregulated in gastric cancer. These include annexins A1 [18], A2 [19, 20], A3 [21], A4 [22], A6 [23], A7 [24, 25], and A10 [26] although their alteration patterns and functions vary. Nevertheless, the potential of these proteins as therapeutic targets in gastric cancer is yet to be characterized.

Annexin A7 is encoded by ANXA7 gene and was the first isolated annexin, mediating the Ca2+-regulated chromaffin granule exocytosis [27]. Monoallelic deletion of ANXA7 causes insulin secretion defect in mice; while biallelic deletion can lead to embryonic lethality [28], implicating its pivotal physiological roles. Dysregulation of annexin A7 is associated with multiple disorders including cancer [29]. In particular, annexin A7 is likely functioning in multifaceted and possibly distinctive ways in different cancer settings. In glioblastoma, melanoma, and prostate cancers, annexin functions as a tumour suppressor [30–35], whereas in liver, colorectal, and breast cancers [36–39], annexin A7 likely promotes tumour progression by regulating metastasis. Moreover, a previous study in glioblastoma demonstrated that annexin A7 may interfere with signalling through epidermal growth factor receptor (EGFR) [30, 35]. Of importance, overexpression of annexin A7 has been linked with gastric cancer metastasis, and its expression correlates with gastric cell differentiation [24, 25, 29] although it has no association with the patient survival [24]. These findings highlight the clinical significance in gastric cancer; however, the functional mechanism of annexin A7 in this malignancy remains poorly understood.

In this study, we characterized the functions of annexin A7 in gastric cancer cell apoptosis, a programmed cell death which represents a key protection mechanism in human but commonly circumvented by cancer cells [40]. Our data showed that silencing ANXA7 expression in vitro and in vivo induced gastric cancer cell apoptosis and inhibited tumour growth in mice. The mechanism underlying the effect of ANXA7 suppression on apoptosis was also investigated.

Materials and methods

Antibodies and reagents

A rabbit anti-annexin A7 polyclonal antibody (10154-2-AP) was purchased from Proteintech (Rosemont, IL). Rabbit polyclonal antibodies against glyceraldehyde-3-phosphate dehydrogenase (GAPDH, YT5052), Bax (YT0459), Bcl-2 (YT0470), caspase-3 (YT0656), and caspase-9 (YT0664) were obtained from ImmunoWay (Plano, TX). Fluorescence-conjugated secondary antibodies against rabbit were obtained from LiCor Biosciences (Lincoln, NE). All chemicals were purchased from Sigma (St. Louis, MO) unless otherwise specified.

Cell culture

A human gastric cell line BGC823, which was originated from poorly differentiated gastric adenocarcinoma, was obtained from the Science and Research Centre, the 4th Affiliated Hospital, Hebei Medical University, China. The cells were maintained at 37 °C in a humidified 5% CO2 atmosphere in RPMI 1640 medium (Thermo Scientific, Waltham, MA) containing 10% (v/v) fetal bovine serum, 100 units/ml penicillin, and 100 units/ml streptomycin.

Suppression of annexin A7

In in vitro assays, expression of annexin A7 in BGC823 cells was silenced by using small interference RNA (siANXA7) from Qiagen (Hilden, Germany). The sequences of siANXA7 are shown in Table 1, targeting nucleotide sequence from 725 to 743 of the ANXA7 coding sequence. A non-specific siRNA (siNS, Qiagen) which shares no sequence identity with any known genes in human was used as a control. The siRNAs were transfected in BGC823 cells, at a final concentration of 80 nM based on a preliminary titration, with a HiPerFect Transfection Reagent (Qiagen) following the manufacturer’s instructions. After incubating the cells with the transfection complex for 6 h, the medium was replaced with fresh medium and the cells were cultured for another 48 h before being used in the in vitro assays. Untransfected parental cells were included as a control (Ctrl).

Table 1 Sequences of siRNAs, shRNAs, and primers
Full size table

In in vivo experiment, expression of annexin A7 in BGC823 cells was suppressed with pGPH1-based small hairpin RNAs (shANXA7) from GenePharma (Shanghai, China) using a non-specific shRNA (shNS) as a control (the sequences of shRNAs are shown in Table 1). The plasmids were purified with an endotoxin-free Maxiprep kit from Tiangen Biotech (Beijing, China) before administration into mice. An Entranster in vivo transfection reagent from Thermo Scientific was used following the manufacturer’s instructions. Twenty microliters of shANXA7 or shNS at 1 μg/μl was mixed with 20 μl of 20% glucose while 40 μl of transfection reagent was combined with 40 μl 10% glucose. The two mixtures were then pooled, mixed, and incubated at room temperature (RT) for 15 min before intratumoural injection. In the control group of mice without shRNA treatment (Ctrl), 20 μl of saline was used instead of shRNAs. The shRNA treatment was repeated every 2 days for a total of seven times.

Mouse xenograft of BGC823 cells

Experiments using mice were approved by Animal Ethics Committee of the 4th Affiliated Hospital, Hebei Medical University, China. BGC823 cells in the exponential stage were detached and dissociated into single cells with 0.25% trypsin in phosphate-buffered saline (PBS). 1 × 106 cells in 0.2 ml of PBS were injected subcutaneously into the right armpits of 4–6-week-old male BALB/C nude mice (HFK Bioscience, Beijing, China). The mice were maintained under a pathogen-free condition at the Animal Centre of the 4th Affiliated Hospital, Hebei Medical University, China at a constant 12:12 h light:dark circadian cycle. Tumour growth was monitored by measuring with a calliper every 3 days. Tumours were allowed to grow to on average 5 mm in diameter before treatment (5 mice/treatment).

At the end of the experiment, mice were euthanized and immersed in 75% ethanol for 5 min and the peritumoural area was disinfected with povidone iodine. The tumours were excised and rinsed three times with RPMI medium before removing connective and adipose tissues. The tumours were weighed and measured with a calliper for the largest and smallest diameters, which were used to estimate the tumour volume as previously described [41]. Tumour tissues were collected for the determination of cell apoptosis and the activities of caspases 3 and 9, pathological analysis, and RNA and protein analyses.

Flow cytometry

Apoptosis of cells freshly detached from the culture surface or dissociated from the mouse xenografts was assessed by flow cytometry analysis using a FITC Annexin V Apoptosis Detection Kit from BD Biosciences (San Jose, CA) following the manufacturer’s instructions. To dissociate cells from mouse xenografts, the tissue blocks were minced and forcefully passed through a 100-μm stainless steel mesh and rinsed with saline; the flowthrough was then filtered with a 50-μm mesh to remove the tissue clumps. The obtained cells were pelleted and washed twice with pre-chilled PBS and stained with Annexin V-FITC in the dark for 15 min at RT and then with propidium iodide (PI) for 5 min before analysis on a FACSCalibur instrument (Becton Dickinson, Franklin Lakes, NJ). Apoptosis rate was calculated as the percentage of both early (Annexin V positive but PI negative) and late apoptotic cells (positive for both Annexin V and PI) in the total population.

Real-time quantitative PCR (qPCR)

Total RNA was isolated from cells or xenograft tissues using a TRIzol reagent (Thermo Scientific) and employed as a template to synthesize the complementary DNA using an M-MLV reverse transcriptase (Promega, Madison, WI). qPCR analysis of the transcripts of interest was then performed with a GoTaq qPCR Master Mix (Promega) and gene-specific primers using GAPDH as a normalization control. The sequences of primers are shown in Table 1, and all primers were synthesized by Dingguo Changsheng Biotechnology (Beijing, China). Abundances of the target transcripts were assessed by using a comparative CT method, and averaged from 3 independent experiments.

Western blotting

Total protein samples were prepared by lysing the adherently cultured cells or mouse xenografts in a lysis buffer containing 50 mM Tris(hydroxymethyl) aminomethane hydrochloride (pH 7.5), 150 mM NaCl, 1 mM dithiothreitol, 1% Nonidet P-40, and 10% glycerol, in the presence of 1 mM phenylmethane sulfonyl fluoride and 1 mM sodium orthovanadate. After the removal of insoluble fraction by centrifugation at 8228×g, the supernatant was collected and the total protein concentration in the lysates was measured by a bicinchoninic acid (BCA) assay (Thermo Scientific). Equal amount of total protein was separated by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene fluoride membranes. After blocking in 5% non-fat milk in a Tris-buffered saline containing 0.1% Tween-20 (TBST) for 2 h at RT, membranes were incubated overnight with primary antibodies in the blocking buffer at 4 °C and washed three times with TBST, each for 10 min. The membranes were then incubated with appropriate secondary antibodies in the blocking buffer at RT for 2 h followed by three washes with TBST. The fluorescence signals were scanned with an Odyssey imager (LiCor), and the intensities were assessed by a densitometry analysis to measure the relative expression of the target proteins using GAPDH as a control.

Caspase activity assays

Activities of caspases 3 and 9 were determined by using colorimetric assay kits purchased from Biobox Biotech (Nanjing, China) following instructions from the manufacturer. Briefly, cells (3–5 × 106) from the adherent culture or freshly dissociated from mouse xenografts were washed with PBS and lysed on ice. Equal amount of total protein (100–200 μg) was added with appropriate amount of Ac-DEVD-pNA or Ac-LEHD-pNA, which is the colorimetric substrate for caspases 3 or 9, respectively, and incubated in the supplied reaction buffer for 4 h at 37 °C. The activities of caspases 3 and 9 in each sample were determined by measuring the absorbance at 405 nm. Final values are normalized to the negative control.

Statistical analysis

All data were processed with a SPSS 19.0 software package (IBM, Armonk, NY), and p < 0.05 was considered statistically significant. Counting data are presented as mean ± standard deviation. Significance of variances between groups was assessed by one-way analysis of variance. A Fisher’s least significant difference test or Dunnett’s T3 test (α = 0.05) was performed depending on the homogeneity or heterogeneity of variances.

Results

Silencing annexin A7 expression promotes apoptosis of BGC823 cells

Expression of annexin A7 in BGC823 cells was suppressed by transient transfection with an ANXA7-specific siRNA, siANXA7. Cells treated with a non-specific siRNA, siNS, and cells with no siRNA treatment were used as controls. After 48 h, cells were collected and the reduction in ANXA7 expression was confirmed by qPCR and western blotting using GAPDH mRNA and protein as controls. The results indicated an 89.9% depletion of ANXA7 mRNA (Fig. 1a) and a 55.1% reduction of annexin A7 protein (Fig. 1b), while shNS-treated cells showed no significant changes in the expression of ANXA7, compared to the untransfected control (Fig. 1).

Fig. 1
figure 1

Silencing ANXA7 expression induces apoptosis in BGC823 gastric cancer cells. a The suppression of ANXA7 expression was confirmed by qPCR and (b) western blotting-densitometry analysis. Ctrl, untransfected cells; siNS, a non-specific control siRNA; and siANXA7, a specific siRNA targeting ANXA7 mRNA. Relative mRNA and protein expressions were normalized against the averages of the Ctrl group. c Cell apoptosis was determined by flow cytometry analysis of cells stained with Annexin V-FITC and PI and (d) the percentage of apoptotic cells were compared. Quantification data were averaged from 3 independent experiments and presented as mean ± standard deviation (SD). *p < 0.001

Full size image

Significant silencing of ANXA7 expression in BGC823 cells in vitro with siRNA allowed us to examine the effects of annexin A7 inhibition on the cellular function. Flow cytometry analysis of BGC823 cells with 48-h suppression of ANXA7 demonstrated a 6.4-fold increase in apoptosis (Fig. 1c, d). The apoptotic population of cells, including both the early and late apoptotic cells, increased from 3.89 ± 0.28% in the untreated control cells to 24.87 ± 4.80% in the siANXA7-treated cells (p < 0.001, Fig. 1c, d). The non-specific siRNA, siNS, which targets no transcripts in human, led to no obvious change in the percentage of apoptotic cells compared to the untransfected cells (Fig. 1c, d).

Depleting annexin A7 affects the expression of apoptosis-mediating genes

To characterize the function of annexin A7 in the apoptosis in gastric cancer cells, we examined the expression of known apoptosis regulatory genes, such as B-cell lymphoma 2 (Bcl-2), Bcl-2-associated X (Bax), Caspases 3 (Casp3) and 9 (Casp9), in BGC823 cells after ANXA7 suppression. qPCR analysis showed that downregulation of ANXA7 in BGC823 resulted in a 37.2% decrease of Bcl-2 mRNA (p < 0.05) and concurrent 2.7-, 1.9-, and 5.7-fold increases of Bax, Casp3, and Casp9 transcripts, respectively, compared to the untreated cells. The non-specific siRNA-treated cells showed no changes in the expression of these genes compared to the untreated cells (Fig. 2a). These RNA alterations were consistent with a more than 75.8% reduction in Bcl-2 protein expression and the 7.0-, 5.0-, and 3.3-fold increases in the expression of Bax, Caspases 3 and 9 proteins, respectively (Fig. 2b).

Fig. 2
figure 2

Inhibiting ANXA7 expression in BGC823 cells affects the expression of Bcl-2, Bax, Casp3, and Casp9. a Levels of Bcl-2, Bax, Casp3, and Casp9 mRNAs were quantified by qPCR using GAPDH as a control. b Expressions of Bcl-2, Bax, Caspase 3, and Caspase 9 proteins were determined by western blotting analysis (top), and a densitometry analysis indicates the fold changes in the protein levels (bottom). Relative mRNA and protein expressions were normalized against the averages of the Ctrl group. Quantification data were averaged from 3 independent experiments and presented as mean ± (SD). *p < 0.05

Full size image

To further confirm the effect of ANXA7 silencing on the apoptosis mediated by caspases, the activities of caspases 3 and 9 were determined by in vitro enzymatic assays using substrates specific for these two enzymes. Our results showed 2.95- and 3.70-fold increases in the activities of caspases 3 and 9 (Fig. 3a, b), respectively, in BGC823 cells with ANXA7 silencing, compared to the untransfected cells. The non-specific siRNA, on the contrast, showed no significant influence on the activities of two caspases (Fig. 3a, b).

Fig. 3
figure 3

Annexin A7 regulates caspase activity. a Suppressing ANXA7 expression in BGC823 cells induces the enzymatic activities of caspase 3 and (b) caspase 9. Values were normalized against the averages of the Ctrl group. Data were averaged from three independent experiments and presented as mean ± (SD). *p < 0.05

Full size image

Suppressing ANXA7 expression in vivo inhibits tumour growth

Results from the in vitro assays indicated significant increases in cell apoptosis caused by the depletion of annexin A7, we next examined in vivo whether targeting ANXA7 could enhance the programmed death of BGC823 cells and reduce the growth of tumours. In BALB/c nude mice, tumours of BGC823 cells were established, and in vivo transfection was carried out through intratumoural injection. Vectors harbouring DNA sequences expressing shRNAs specifically targeting ANXA7, shANXA7, or no known transcript in human, shNS, were introduced into tumour cells, using the untreated xenografts as controls. Our results showed that in vivo transfection of shANXA7 significantly inhibited the tumour growth in BGC823 mouse xenografts compared to shNS-treated mice (Fig. 4a). The volume of tumours from these mice at sacrifice was 614.72 ± 121.72 mm3, significantly lower than that in the control mice with no shRNA treatment (1072.65 ± 238.47 mm3) (p < 0.05, Fig. 4b). Noteworthy, the non-specific shRNA control, shNS, showed no obvious impact on the tumour growth of tumours, which averaged 955.82 ± 152.29 mm3 at the sacrifice of the mice. Consistently, the weights of tumours at harvest in the shANXA7-treated group (1.36 ± 0.39 g) were significantly lower than those from the shNS-treated and untreated groups 2.13 ± 0.32 g and 2.20 ± 0.34 g, respectively (Fig. 4c, p < 0.05).

Fig. 4
figure 4

Inhibiting ANXA7 affects tumour growth in vivo. a Silencing ANXA7 expression in BGC823 cell xenografts in mice with a shRNA targeting ANXA7, shANXA7, affects tumour growth, in comparison with the controls without treatment (Ctrl) or treated with a non-specific shRNA, shNS. b, c Tumour volumes and weights are compared. Data are presented as mean ± (SD). *p < 0.05

Full size image

Silencing ANXA7 induces apoptosis in vivo and alters expression of apoptosis-mediating genes

Analysis of ANXA7 mRNA unveiled a 34.5% decrease in the tumour tissues from the shANXA7-treated mice compared to those from the untreated control group (p < 0.05, Fig. 5a). On the contrast, the non-specific shRNA caused no apparent changes in the levels of ANXA7 mRNA. Concordantly, western blotting analysis of the tumour tissues demonstrated about 54% decrease in the expression of annexin A7 protein in tumours transfected with shANXA7 (p < 0.05), compared with the untreated controls, whereas the shNS transfection showed no significant effect (Fig. 5b).

Fig. 5
figure 5

Silencing ANXA7 expression in BGC823 cell xenografts in mice leads to an increase of apoptosis. a The suppression of ANXA7 expression was confirmed by qPCR and (b) western blotting-densitometry analysis. Relative mRNA and protein levels were normalized against the averages of the Ctrl group. c Tumour cells dissociated from the xenografts were stained with annexin V-FITC and PI and analysed by flow cytometry. d The percentages of apoptotic cells are summarized. Quantification data were averaged from the results of all tumours in each group and presented as mean ± (SD). *p < 0.05

Full size image

Cells were dissociated from fresh tumour tissues collected from BGC823 xenografts and stained with annexin V-FITC and PI to quantify the population of cells in apoptotic cells. As shown in Fig. 5c, d, silencing ANXA7 in tumour xenografts increased cell apoptosis by 3.35-fold, while the non-specific shRNA showed no change, in comparison with the untreated tumours. These data suggested promotive effect on apoptosis in vivo by inhibiting annexin A7 expression.

Further qPCR analysis of the tumour tissues demonstrated that the in vivo suppression of ANXA7 inhibited the transcription of Bcl-2 while enhancing the expression of Bax, Casp3, and Casp9 by 2.4-, 4.3-, and 4.6-fold, respectively, compared to the untreated controls (Fig. 6a). Consistently, the level of Bcl-2 protein was reduced by 67.0%, whereas the expression of Bax, caspase 3, and caspase 9 proteins were significantly upregulated by 3.3-, 1.8-, and 3.9-fold, respectively (Fig. 6b). Of note, the non-specific shRNA treatment led to no statistically significant changes in the expression of these genes.

Fig. 6
figure 6

Inhibiting ANXA7 expression in BGC823 cell xenografts in mice alters the expression of Bcl-2, Bax, Casp3, and Casp9 genes. a Levels of Bcl-2, Bax, Casp3, and Casp9 mRNAs in xenografts treated with shANXA7, shNS, or vehicle (Ctrl) were quantified by qPCR using GAPDH as a control. b Expressions of Bcl-2, Bax, Caspase 3, and Caspase 9 proteins were determined by western blotting-densitometry analysis. Relative mRNA and protein expressions were normalized against the averages of the Ctrl group. Quantification data were averaged from the results of all tumours in each group and presented as mean ± (SD). *p < 0.05

Full size image

We next examined the enzymatic activities of caspases 3 and 9 in the protein extracts from the tumour xenografts. The results showed elevated activities of both caspase 3 (1.78-fold) and caspase 9 (2.71-fold) in tumours with ANXA7 silencing, compared to those from the shNS-treated and untreated tumour samples (Fig. 7a, b). These data, which were consistent with the results of qPCR and western blot analysis, suggested the activation of tumour caspases-mediated apoptosis pathway induced by ANXA7 suppression in vivo.

Fig. 7
figure 7

Suppressing ANXA7 in the xenograft tumours in mice increases the enzymatic activities of (a) caspase 3 and (b) caspase 9. Values were normalized against the averages of the Ctrl group. Data were averaged from all tumour samples in individual groups and presented as mean ± (SD). *p < 0.05

Full size image

Discussion

Here we demonstrate that annexin A7 protein has an important and potentially targetable role in gastric cancer. Silencing the expression of ANXA7 in vitro and in vivo induces the apoptosis of BGC823 gastric cancer cells and, importantly, inhibits the growth of tumour xenografts in mice. Annexin A7-mediated apoptosis was found to be associated with alterations of multiple apoptosis-related genes, such as Bcl-2, Bax, Casp3, and Casp9. These findings suggest that inhibiting annexin A7 may potentially be beneficial to the gastric cancer patients.

Data from our in vitro assays suggest that annexin A7 is functionally important in protecting gastric cancer cells from apoptosis. Inhibiting its expression increases BGC823 cell apoptosis by more than five times. This result is consistent with the previous findings in mouse and human hepatocarcinoma cells, in which downregulating ANXA7 expression induces apoptosis and inhibits their metastases [42, 43]. Of potentially clinical importance, our in vivo data in mouse xenografts of BGC823 cells emphasized our in vitro finding. Intratumoural injection of shANXA7 significantly repressed the growth of tumours, suggesting that targeting annexin A7 in the gastric cancer patients may be beneficial in regard to tumour growth control. However, further studies are required to examine the potential side effects of this approach considering the critical physiological roles of this annexin in mice [27, 28].

Apoptosis can be triggered by (1) extrinsic signalling through death receptors, including Fas cell surface death receptor (FasR), tumour necrosis factor receptor 1 (TNFR1), death receptor (DR)3, DR4, and DR5, and activation of perforin/granzyme pathway [44]; (2) endoplasmic reticulum (ER) stress [45]; and (3) intrinsic non-receptor-mediated pathways that disrupt the integrity of mitochondrial membrane [44]. The intrinsic apoptotic pathway is coordinated by the members of the Bcl-2 family of proteins [46], which may also respond to signals from the death-receptor pathway [47]. In this study, both in vitro and in vivo data demonstrate a significant decrease in the expression of Bcl-2, an anti-apoptotic protein of the Bcl-2 family, and a concurrent increase of Bax protein, a pro-apoptotic member of the Bcl-2 family, in BGC823 cells or tumours treated with ANXA7 silencing. As a result, the Bax–to-Bcl-2 ratio, which is believed to be a major checkpoint for the intrinsic apoptotic pathway [48], was significantly increased. The accumulation of Bax allowed the formation of homodimers and heterodimers with other Bcl-2 family proteins, such as Bcl-2 and Bcl-xL, which collaboratively promotes cell apoptosis [49, 50]. Downstream of these, caspases 9 expression and activity were upregulated in BGC823 cells with ANXA7 suppression. This is consistent with the decrease of Bcl-2 and increase of Bax, which are known to initiate the release of cytochrome C from the mitochondria [49]. Cytochrome C associates with apoptotic protease activating factor 1 (APAF-1) in the cytoplasm to form apoptosome which recruits procaspase 9 and converts it into active caspase 9 [51]. Caspase 9 as a caspase initiator further activates caspase 3, the most important executioner caspase in the apoptosis execution pathway [44]. Indeed, both the expression and activity of caspase 9 and 3 were increased by ANXA7 suppression in this study. Taken together, our data suggest that annexin A7 likely mediates gastric cancer cell apoptosis through an intrinsic signalling cascade.

How silencing ANXA7 expression triggered alterations of Bcl-2 and Bax in BGC823 cells is uncertain. In mouse hepatocarcinoma cells, annexin A7 binds to BAG4, which plays a negative role in mitochondrial apoptosis by associating with Bcl-2, and protects cells from apoptosis [52]. In breast cancer cells, annexin A7 interacts with galectin-3 and facilitates the translocation of the latter to perinuclear membranes, which is important for the prevention of mitochondrial damage and cytochrome C release [53]. Whether annexin A7 functions through these mechanisms in gastric cancer cells requires further characterization.

In summary, our data demonstrate that annexin A7 has an important role in gastric cancer cell apoptosis. Silencing the expression of annexin A7 in vitro and in vivo induces dramatic increases in cell apoptosis, and of clinical value, represses tumour growth in mice. Biochemical analysis suggests that annexin A7 likely mediates apoptosis through an intrinsic signalling pathway. Our results implicate potential therapeutic value of targeting annexin A7 for the treatment of gastric cancer although preclinical studies are required to assess the adverse effect induced by the inhibition of annexin A7.