Gastrin activates autophagy and increases migration and survival of gastric adenocarcinoma cells

Autophagy is an evolutionarily conserved process wherein the cytoplasmic components are degraded to provide cells with energy during starvation. Basal autophagy is necessary to maintain homeostasis and can be induced in response to cellular stress [1, 2]. The process of macroautophagy (herein referred to as autophagy) involves the engulfment of cytoplasmic material into de novo generated double membrane vesicle called autophagosomes. The isolated material is degraded after the fusion with the lysosomes [3]. The process of autophagy is orchestrated by a set of AuTophaGy-related genes (ATGs) that were first identified in yeast, but later shown to have orthologs in mammals [4]. Microtubule-associated protein 1 light chain 3 beta (MAP1LC3B-I/II/ LC3B) is lipidated when autophagy is induced and plays an essential role in the autophagosome formation [5]. Sequestosome 1 (SQSTM1/p62) facilitates the degradation of polyubiquitinated substrates by autophagy via the direct interaction with ubiquitinated proteins and MAP1LC3B located on the autophagosomal membrane [6]. MAP1LC3B and SQSTM1 are both produced and degraded in a coordinated manner during autophagy and therefore, are used as markers to study this process [7, 8].

The initiation of autophagy is orchestrated by the activity of the ULK1 (ATG1) kinase complex [9]. The activity of the ULK1 complex is positively regulated by the adenosine monophosphate-activated protein kinase (PRKAA2/AMPK) and inhibited by mammalian target to rapamycin (mTOR). This leads to balancing of cellular catabolic routes according to the innate needs of the cell. The activity of the ULK1 complex can be monitored by using specific antibodies,that recognize the phosphorylation of ULK1 on Ser 555 or Ser 317 (stimulate the activity) or on ULK1 Ser 757 (inhibit the activity) [10‐12].

The peptide hormone gastrin (G-17) is the central regulator in the maintenance and organization of the gastric mucosa and plays a pivotal role in gastric acid secretion in the stomach [13]. In addition, gastrin exerts growth-promoting effects in both normal and malignant gastrointestinal tissues in the oxyntic mucosa and the gastric epithelial cells [14]. Gastrin has been found to stimulate proliferation of cancer cell lines isolated from the stomach, pancreas and colon [15‐17]. It has been reported to promote cellular responses such as migration, invasion and survival [18‐20]. However, the role of gastrin in the progression of gastric adenocarcinoma is not completely understood. Nonetheless, hypergastrinemia in combination with H. pylori infections are considered to be a risk factor for the development of gastric adenocarcinomas [21].

We have previously reported that gastrin treatment of the pancreatic adenocarcinoma cell line AR42J resulted in differentially expressed genes which were annotated to cellular responses such as unfolded protein response (UPR)/ER stress and survival [22]. It is well established that UPR/ER stress is counteracted by increased autophagy [23]. Thus, we hypothesized that gastrin may be involved in the activation of autophagy in human gastric cancer cells. In this study, we find that gastrin treatment induces autophagy in the gastric adenocarcinoma cell lines AGS-Gr and MNK45, concomitant with the activation of the STK11-PRKAA2-ULK1 signaling cascade. Further, we demonstrate that gastrin treatment reduces the cytotoxic effect exerted by cisplatin. We propose that gastrin induced autophagy is in part responsible for the increased migration and chemoresistance of the AGS-Gr cells.

Induced autophagy represents a survival mechanism in tumour cells that may enable them to survive during stressful conditions including exposure to cytostatic drugs [27]. We examined whether gastrin induced autophagy was essential for the increased cell survival. Consistent with previous reports, gastrin caused a slight, but consistent and significant increase in the number of viable cells under serum-free conditions using annexin-PI staining (Fig. 4a). The lysosomal inhibitor BafA1 did not affect the viability of AGS-Gr cells by itself, but interestingly, adding BafA1 together with gastrin reduced the pro-survival effect of gastrin by 30% (Fig. 4a & Additional file 1: Figure S4a). The data suggests that induced autophagy contributes to enhanced cell survival in response to gastrin.

Cisplatin is used in the treatment of gastric adenocarcinomas [34, 35], and we tested whether gastrin induced autophagy could modify the cellular response to cisplatin in vitro. Initially, AGS-Gr cells were treated with increasing concentrations of cisplatin (1–90 μM) in the presence and absence of gastrin for 24 h, 48 h and 72 h. The numbers of surviving cells were determined by their metabolic activity (XTT assay). We found that gastrin treatment reduced the sensitivity towards cisplatin (Fig. 4b & Additional file 1: Figure S4 (c) & (d)). To explore if this gastrin induced survival was due to the induced autophagy, we performed the experiments in the presence of the autophagy inhibitor HCQ. Interestingly, addition of HCQ reduced the survival effect of gastrin (Fig. 4c) and the viability of cells treated with gastrin + cisplatin (Fig. 4d). HCQ diminished the survival effect induced by gastrin with increasing concentrations (1 μM, 4 μM and 7.5 μM) of cisplatin (Fig. 4d). This indicates that the survival effect of gastrin involves the induction of autophagy. Notably, HCQ did not affect cell viability by itself (Fig. 4c). As anticipated, HCQ treated cells showed an accumulation of both SQSTM1 and MAP1LC3B-II (Additional file 1: Figure S4 (e)). Next, we treated AGS-Gr cells with the ULK1 inhibitor SBI in the presence of BafA1 + gastrin. This resulted in reduced accumulation of SQSTM1 and MAP1LC3B-II (Additional file 1: Fig. S4 (f)). Further, AGS-Gr cells were treated with gastrin +/− SBI and cell viability determined by XTT assay. Consistently, gastrin alone increased the cell viability, but in the presence of SBI, cell viability was reduced by approx. 15% compared to cells treated with gastrin alone, at both 24 and 48 h (Fig. 4e). As previously reported the inhibitor on its own reduced cell viability [36].

To establish the link between gastrin induced autophagy and apoptosis, we examined the activation of caspase 3/7 in the AGS-Gr cells. Initially, the cells were treated with gastrin +/− BafA1. Gastrin treatment alone reduced the induction of caspase 3/7 activity (Fig. 4f), this coincides with a study in the gastrin responsive AR42J cells [37]. The reduced activation of caspases in the presence of gastrin was counter acted by BafA1 treatment (Fig. 4f). When, the cells were treated with gastrin in the presence or absence of cisplatin (autophagy blocked 16 h), we find that the inhibition of autophagy increased the activation of caspase 3/7 (Cisplatin + gastrin + BafA1 versus Cisplatin + gastrin) (Fig.4f). However, activation of caspases in the presence of cisplatin alone or in combination with BafA1 was found not to be significant. Further, when gastrin was added to cells treated with cisplatin, we observed a reduced caspase activity, suggesting that gastrin exerts a cytoprotective effect on these cells (Fig. 4f). Collectively, these results suggest that gastrin induced autophagy is linked to the anti-apoptotic effect exerted by gastrin.

The data presented above are consistent with a gastrin induced autophagy that stimulates migration and potentiates cell survival. The ULK1 kinase is the master regulator of autophagy by coordinating the initial steps of autophagosome formation. Since the data suggests that gastrin induces autophagy, we asked if this involves the activation of ULK1. We have recently demonstrated that gastrin induces STK11 Ser 428 phosphorylation in AGS-Gr and MKN45 cells [25], indicating that the STK11-PRKAA2-ULK1 signalling pathway might be involved in gastrin mediated induction of autophagy. Thus, we treated cells with gastrin and examined the phosphorylation of PRKAA2 by immunoblotting. In line with a gastrin induced activation of STK11, the phosphorylation of the STK11 targeted site in PRKAA2 (Thr 172) was transiently elevated in both AGS-Gr and MKN45 cells (Fig. 5a & b). Concurrent with an elevated activity of PRKAA2, we found that gastrin treatment also increased the phosphorylation of the autophagy activating sites of the ULK1 complex (Ser 317 and Ser 555) (Fig 5a & c). These data indicate a direct signalling pathway mediated by gastrin/CCKBR to the activation of ULK1 via increased activity of STK11 and PRKAA2. However, ULK1 may in addition be regulated indirectly by the same pathway, ie. if the gastrin induced PRKAA2 activity results in reduced mTOR activity. To further unravel signalling pathways involved in autophagy, we examined gastrin mediated phosphorylation of Regulatory-associated protein of mTOR (Raptor) Ser 792, which is known to inhibit mTOR activity. As shown in (Fig 5a & b), gastrin induced the phosphorylation of Raptor Ser 792. In the AGS-Gr cells, the phosphorylation of Raptor Ser 792 appeared as early as 5 min, and in the MKN45 cells at 15 min. Consistent with the activation of ULK1 on the autophagy activating sites (Ser 555 and Ser 317) we also found that the mTOR substrate 4EBP1 was less phosphorylated after gastrin treatment of the AGS-Gr cells (Fig. 5d). Collectively, these results suggest that the elevated autophagy in response to gastrin treatment is both due to a direct effect on PRKAA2, which induces the ULK1 activity, and the indirect effect via reduced mTOR activity. Additionally, AGS-Gr cells were transfected with siRNA towards STK11 and subsequently treated with gastrin before the assessment of the autophagy markers. The protein level of STK11 was reduced by ~ 60% compared to cells transfected with non-targeting siRNA (Fig. 6a). As shown in Fig. 6b, we observed a reduction in gastrin induced expression of MAP1LC3B-II and SQSTM1 when STK11 was knocked down. Similarly, targeting PRKAA2 with siRNA or a chemical inhibitor (Comp C) resulted in the downregulation of SQSTM1 (Fig. 6c, d & Additional file 1: Figure S5). Taken together, our results are congruent with a gastrin induced autophagy involving the activation of STK11–PRKAA2-ULK1 signaling pathway.

To elucidate a functional relationship between the gastrin-induced signalling cascades detailed above and migration, we treated AGS-Gr cells with the PRKAA2 inhibitor Compound C (Comp C). We found that Comp C decreased the gastrin-induced migration by approx. 40% (18 h) (Fig. 6e), indicating the involvement of the autophagy regulated signalling pathway. Together, these results suggest that the STK11 - PRKAA2 pathway controls autophagy and that this is important for the enhanced cell survival and migration in response to gastrin.

Gastrin exerts a growth promoting effect on several gastrointestinal cancer cells and a variety of neoplasms that express CCKBR, including neuroendocrine, pancreatic, medulla thyroid and lung cancer [38‐40]. Interestingly, Hur et al. demonstrated that gastrin and CCKBR are expressed in approx. 50% of gastric carcinoma tissues, and patients with diffuse type of gastric carcinoma expressing both gastrin and CCKBR had poorer prognosis compared to those who were negative for both [41]. However, the role of gastrin in adenocarcinoma is still not completely understood, and whether gastrin acts as an autocrine/paracrine growth factor in gastric carcinoma is unclear. In the current study, we report that gastrin induces autophagy and increases cell migration and survival in vitro, and suggest that these molecular mechanisms may contribute to tumor progression of gastric cancer cells.

Several recent studies have indicated a role of autophagy and autophagy related proteins in the progression of gastric cancer [42, 43]. Immunohistochemistry analysis has identified an elevated staining intensity for autophagy related proteins such as Beclin1, Atg5, Atg8/ MAP1LC3B, and SQSTM1/p62 in gastric cancer tissue [26, 42]. Upregulation of MAP1LC3A was shown to correlate with an increase in Ki67 positive cells in gastrointestinal tumors [44]. Ge et al. demonstrated that ATG5 was highly expressed in gastric cancer patients compared to healthy individuals, which was speculated to contribute to increased chemoresistance [45]. Autophagy is an incessant process, where numerous proteins including MAP1LC3B and SQSTM1 are constantly degraded. Thus, elevated levels of the proteins in a biopsy could result from both activation and inactivation of the process. Additionally, how autophagy could possibly be activated and contribute to gastric cancer development is currently not fully understood. We find that the peptide hormone gastrin induces autophagy in two gastric adenocarcinoma cell lines via the STK11-PRKAA2-ULK1 pathway. This signaling pathway may contribute to cancer development by both increasing cell survival and cell migration.

Loss of STK11 has been shown to correlate with increased cancer development. Nguyen and Liu et al. [46, 47] demonstrated that STK11 deficient melanoma cells have increased invasive properties compared to cells with normal amounts of STK11 protein. Contrarily, STK11 has been reported to be necessary for the survival of colorectal cancer cells, hepatocyte proliferation, liver regeneration and cell survival of liver tumors with constitutive activation of AKT [48, 49]. Our result establishes a functional role of STK11 in gastrin induced autophagy; the inhibition of downstream kinases PRKAA2 and ULK1 reduced gastrin-induced migration and survival. The role of autophagy in the migration of cancer cells is currently being investigated. Kenefic et al. found that autophagy is involved in the turnover of focal adhesion (FA), promoting FA disassembly and that this is dependent upon the autophagic receptor NBR1 [33]. Further, activation of autophagy in glioblastoma cells impairs the migration and invasion capacities via the downregulation of epithelial mesenchymal transition proteins such as SNAIL and SLUG [50]. Autophagy deficiency was found to increase SQSTM1/p62 levels, resulting in reduced E-cadherin expression, stabilization of the oncogenic protein Twist1 and promotion of cell migration, invasion and proliferation of human squamous cell carcinomas [51].

Gastrin is known to induce cell survival via several mechanisms [18, 52]. We find that gastrin partially counteracts the apoptotic effect of cisplatin, and blocking autophagy reduces the survival effect induced by gastrin. Xu et al. reported targeting autophagy sensitizes gastric cancer cells to oxaliplatin induced apoptosis [32]. Inhibiting lysosomal degradation by using BafA1 in combination with 5-Flurouracil (5-FU) resulted in decreased viability and colony forming capacity of SGC-7901 cells [27]. Kim et al. showed that cisplatin treatment induced phosphorylation of AMPK in the AGS cells, and that pharmacological inhibition/siRNA towards AMPK sensitized the cells to cisplatin induced apoptosis [53]. These results suggest that AMPK and autophagy contribute to a general chemo resistance and act in a cytoprotective manner in response to anti-cancer therapy.

Recently Egan et al., found that inhibition of ULK1 reduced the survival of A549 cells significantly, when combined with a mTOR inhibitior [36]. In our study, we report that the ULK1 inhibitor SBI, downregulates gastrin induced autophagy as well as reduces the associated migration and survival. We suggest that increased autophagic flux may be one of the several mechanisms by which gastrin induces migration, cell survival and chemoresistance. Pharmacological interference of ULK1 and PRKAA2 as therapeutic targets might provide a novel therapeutic strategy to target gastric cancer cells, which are resistant to cisplatin/5-FU treatment. In fact, the combination of HCQ with several chemotherapeutic agents is currently being investigated in clinical trials [54, 55]. These results lay the foundation for further investigation of the role of gastrin induced autophagy in survival and metastasis.

Our data demonstrates that gastrin regulates autophagy via the STK11-PRKAA2-ULK1 pathway in vitro. We show that the gastrin induced migration and survival of gastric adenocarcinoma cells is partially dependent on induced autophagy. These results may contribute to a better understanding of the role of gastrin in gastric cancer. Blocking of autophagy may be a therapeutic approach for sensitizing gastric cancer cells to chemotherapy.