Helicobacter pylori regulates ILK signaling pathways to inuence autophagy in gastric epithelial cells

Background The ability of Helicobacter pylori to manipulate host autophagy is an important pathogenic mechanism. Results We found a negative correlation between the expression of ILK and the autophagy marker protein LC3B in H. pylori-positive human samples and in H. pylori-infected GES-1 cell lines. There was a signicant accumulation of autophagosomes in ILK-knockdown GES-1 cells, and the expression levels of both LC3B and p62 were also increased. Here, we showed the activities of Rac1 and RhoA were decreased in H. pylori-infected GES-1 cells and ILK-knockdown GES-1 cells. Inhibition of Rac1 and RhoA increased LC3B levels and autophagosome formation in GES-1 cells after H. pylori infection. Simultaneously, H. pylori infection activated downstream signal molecules of Rac1 (PAK1, LIMK1 and colin) and RhoA (ROCK1, ROCK2 and LIMK1 and colin). Our results demonstrated that H. regulated through and signaling pathways in cells. transfection, western blot and real-time PCR.


Background
Helicobacter pylori resides in more than half of the human population and has co-evolved with humans for more than 58,000 years (Otero et al., 2014). H. pylori can evade the host immune defense system, colonize in human gastric mucosa and maintain long-term chronic infection (Dunn et al., 1997). It has evolved cytotoxin-associated gene A (CagA) and vacuolating cytotoxin (VacA) to achieve this purpose (Radosz-Komoniewska et al., 2005). It employs a multicomponent type IV secretion system (T4SS) to secrete CagA into the cytosol of infected host cells and initiate various host responses (Schuelein et al., 2011). In addition to CagA, VacA is identi ed as necessary and su cient conditions of H. pylori to induce autophagy in gastric epithelial cells (Raju et al., 2012).
Autophagy is an evolutionary conservative process with the degradation of cytoplasmic organelles, proteins, or pathogens in eukaryotic cells (Klionsky and Emr, 2000). Autophagy fuses autophagosomes with lysosomes to form autolysosome, and it is regulated by a variety of autophagy-related genes (Klionsky et al., 2003). Upon bacterial infection, autophagy is induced through host signaling pathways, which results in pathogen being encapsulated within autophagosomes, and then fuses with lysosomes, thereby leading to the degradation of the encapsulated bacteria (Rich et al., 2003). Autophagy as an intracellular defense to eliminate pathogens can be induced by H. pylori in gastric epithelial cells (Terebiznik et al., 2009;Tang et al., 2012). However, previous studies have indicated that H. pylori could prevent its degradation, reside and multiply in autophagosomes, and maintain persistent infection (Sit et al., 2020;Wang et al., 2009). H. pylori T4SS makes a contact with integrin 1 directly that strongly supports the translocation of CagA into the cytosol of infected host cells (Schuelein et al., 2011). Integrin-linked kinase (ILK) plays a central role in the integrin signaling pathway and couples integrin activation to actin cytoskeleton (Ghatak et al., 2013), which transmits extracellular signals to cells and mediates cell growth and differentiation (Grashoff et al., 2012). ILK can activate the family members of Rho GTPases (Graness et al., 2006), which mediate changes in actin polymerization by manipulating the terminal effector protein co lin-1, leading to the formation of stress bers, lopodia or lamellipodia, and ru es (Dedhar, 2000). Previous studies have shown that actin dynamics play a vital role in the whole process of autophagy (Kast and Dominguez, 2017). In this study, we explore the mechanism of H. pylori affecting autophagy by regulating ILK.

2.1
The ILK expression decreased and the LC3B expression increased duringH. pyloriinfection in gastric epithelial cells ILK and LC3B, as represented by green uorescence, were distributed in H. pylori-negative or -positive human gastric epithelial cells ( Fig. 1a and 1b). The average uorescence intensity of ILK was markedly decreased (P < 0.001) in H. pylori-positive human gastric epithelial cells compared with that in H. pylorinegative human gastric epithelial cells (Fig. 1c). And the average uorescence intensity of autophagy marker protein LC3B was signi cantly higher (P < 0.001) in H. pylori-positive human gastric epithelial cells than that in H. pylori-negative human gastric epithelial cells (Fig. 1d). We speculated that the reduced ILK in human samples could affect the LC3B. Therefore, Pearson's correlation analysis was used to compare the relative levels of ILK and LC3B in these samples. There was statistically a signi cant negative correlation (P < 0.001) between ILK and LC3B only in H. pylori-positive human gastric epithelial cells (R = 0.653) (Fig. 1e); this nding indicated that ILK was inversely correlated with LC3B in H. pylori-positive human samples.
We also proved that ILK was downregulated and autophagosomes were increaed in GES-1 cell lines after H. pylori infection. As shown in Fig. 2a, infection with H. pylori with or without CagA decreased ILK phosphorylation in GES-1 cells, and CagA knockout H. pylori (CagA − ) led lower level of ILK phosphorylation (P < 0.01) compared to that caused by H. pylori wild type. The total amount and phosphorylated ILK were down-regulated (P < 0.05) at 2, 4, 6, and 8 h in GES-1 cells after H. pylori challenge (Fig. 2b). The tandem mRFP-GFP-LC3B adenovirus assay is based on the pH difference between the acidic autolysosome and the neutral autophagosome. The progress of autophagosome to autolysosome was manipulated by the differences in pH sensitivity of green uorescent protein (GFP) and red uorescent protein (RFP). Once autophagosome and lysosome fused to form autolysosomes, the GFP moiety degraded from the tandem protein, but mRFP-LC3B still maintains the punta. Induction of autophagosome (merge of GFP-LC3B and mRFP-LC3B) and autolysosome (mRFP-LC3B) formation in GES-1 cells after H. pylori infection occurred in a time-dependent manner (Fig. 2c). The numbers of autophagosome increased at 4, 6 and 8 h in GES-1 cells after H. pylori infection, and the numbers were the most at 6 h after bacterial infection, and the numbers of autolysosome gradually increased from 2 to 8 h and reached the peak at 6 h in GES-1 cells after H. pylori infection (Fig. 2d). The p62 levels decreased at 2 h and increased at 4, 6 and 8 h after H. pylori infection (P < 0.05) (Fig. 2e). These results further indicated that H. pylori disrupted autophagy in GES-1 cells.

Downregulation of ILK disrupted autophagy in GES-1 cells
RNAi was used to downregulate the expression of ILK in GES-1 cells. Western blot analysis demonstrated that ILK was reduced compared to that in the negative group at 48 h after transfection with sequencespeci c siRNA (HSS140843, HSS179923, or HSS179924) targeting ILK (Fig. 3a). Ultimately, we chose HSS140843 as the target siRNA to reduce the expression of ILK for subsequent experiments. To further study the in uence of ILK on autophagy in H. pylori-infected GES-1 cells, the tandem mRFP-GFP-LC3B adenovirus was used to evaluate the formation of autophagosome and autolysosome. As shown in Fig. 3b and 3c, we found increased formation of yellow uorescent autophagosomes and red uorescent autolysomes in ILK-knockdown GES-1 cells compared with those in control siRNA-treated GES-1 cells, regardless of whether or not GES-1 cells were infected by H. pylori. The number of autophagosomes were 104.3 ± 7.57 per cell, which was 1.57-fold higher (P = 0.017) in ILK knockdown GES-cells after H. pylori infection than that in control siRNA-treated GES-1 cells after H. pylori infection. To further corroborate the in uence of ILK on autophagy, the appearance of double-membraned autophagosomes was investigated by transmission electron microscopy (Fig. 3d). There was a signi cant accumulation of autophagosomes in ILK-knockdown GES-1 cells compared to that in control siRNA-treated GES-1 cells. The LC3B levels were then assessed by western blot analysis. As shown in Fig. 3e, LC3B was signi cantly up-regulated (P < 0.05) and p62 was signi cantly down-regulated (P < 0.05) in GES-1 cells after H. pylori infection. The LC3B and p62 levels were both higher (P < 0.05) in ILK-knockdown GES-1 cells than those in control siRNA-treated GES-1 cells after H. pylori infection. These results indicated that downregulation of ILK disrupted autophagy in H. pylori-infection GES-1 cells.

Inhibition of ILK in uenced Rac1 and RhoA activation in GES-1 cells
Total Rac1 and GTP-Rac1, and total RhoA and GTP-RhoA were detected by western blot. GTP/Total Rac1 and GTP/Total RhoA were decreased in GES-1 cells after H. pylori infection and in ILK-knockdown GES-1 cells regardless whether or not GES-1 cells were infected with H. pylori (Fig. 4). GTP/Total Rac1 and GTP/Total RhoA were were 3/4-fold (P < 0.05) and 2/5-fold (P < 0.05) lower in H. pylori-infected GES-1 cells than in the blank GES-1 cells, respectively.
2.4 Inhibition of Rac1 and RhoA in uenced autophagy in GES-1 cells The cells survival rates were to detect concentration-dependent and time-dependent toxicity of inhibitors against GES-1 cells. No signi cant difference was observed in cell survival rate between inhibitor-treated groups (10 and 50 µM of NSC23766, and 2 and 5 µM of CCG-1423-treated GES-1 cells) and blank group.
The GES-1 cells treated with 50 µM of NSC23766 and 5 µM of CCG-1423 at 1, 2 and 4 h showed no change in cell survival rates (Fig. 5a). The LC3B levels were assessed by western blot to clarify whether inhibition of Rac1 and RhoA in uenced autophagy. As shown in Fig. 5b, LC3B was signi cantly increased (P < 0.05) in Rac1-or RhoA-inhibited GES-1 cells compared to that in control GES-1 cells, regardless of whether the cells were infected with H. pylori. To further study the in uence of Rac1 and RhoA on autophagy in GES-1 cells, the tandem mRFP-GFP-LC3B adenovirus was used to evaluate the formation of autophagosome and autolysosome. As shown in Fig. 5c and 5d, we found increased formation of yellow uorescent autophagosomes in H. pylori-infected, Rac1-and RhoA-inhibited GES-1 cells with or without H. pylori compared to those in control GES-1 cells. The red uorescent autolysosomes in H. pylori aloneinfected, Rac1-and RhoA-inhibited GES-1 cells were more (P < 0.01) than that in untreated GES-1 cell, but the autolysosomes in Rac1-and RhoA-inhibited GES-1 cells with H. pylori infection were remarkably decreased compared to that in GES-1 cells infected by H. pylori alone. These results indicated that the inhibition of Rac1 or RhoA only increased autophagosome formation in GES-1 cells after H. pylori infection.

Inhibition of Rac1 and RhoA in uenced downstream signaling molecules
To determine whether PAK1, ROCK1, ROCK2, LIMK1, and co lin were regulated by Rac1 or RhoA, we investigated PAK1, ROCK1, ROCK2, LIMK1, and co lin mRNA and protein phosphorylation levels in Rac1or RhoA-inhibited GES-1 cells through qRT-PCR and western blot. The mRNA levels of PAK1, ROCK1, ROCK2 and LIMK1 were signi cantly higher (P < 0.05), and co lin was remarkably lower (P < 0.01) in GES-1 cells after H. pylori infection than those in untreated GES-1 cells. The mRNA levels of PAK1 and LIMK1 were signi cantly decreased (P < 0.001) and co lin was remarkably increased (P < 0.05) in NSC23766treatment GES-1 cells compared to those in untreated GES-1 cells, and the same trends were observed in NSC23766-treatment GES-1 cells after H. pylori infection compared to those in H. pylori-infected GES-1 cells. The mRNA levels of ROCK1, ROCK2 and LIMK1 were signi cantly decreased (P < 0.01) and co lin was remarkably increased (P < 0.05) in CCG-1423-treatment GES-1 cells compared to those in untreated GES-1 cells, and the same trends were observed in NSC23766-treatment GES-1 cells after H. pylori infection compared to those in H. pylori-infected GES-1 cells (Fig. 6a). The protein phosphorylation levels of PAK1, ROCK2 and LIMK1 were signi cantly higher (P < 0.05), and co lin was remarkably lower (P < 0.05) in GES-1 cells after H. pylori infection than those in untreated GES-1 cells. The protein phosphorylation levels of PAK1 and LIMK was signi cantly decreased, co lin protein phosphorylation level was signi cantly increased in GES-1 cells treated with Rac1 inhibitor NSC 23766. The protein phosphorylation levels of ROCK1, ROCK2 and LIMK was signi cantly reduced, co lin protein phosphorylation level was signi cantly increased in GES-1 cells treated with the RhoA inhibitor CCG-1423 ( Fig. 6b).

Discussion
The ability of H. pylori to regulate host autophagy is one of main pathogenic mechanism (Rubio et al., 2012). Autophagy can be employed as means of host defense and elimination of invasive bacteria at the initial stage; however, with the prolongation of exposure time, H. pylori can disrupt autophagy by preventing maturation of the autolysosome, leading to a chronic sense of bacterial persistent infections (Rubio et al., 2012). Although novel insights into the mechanism of autophagy in bacterial infection have emerged, many aspects remain largely unknown.
ILK, a serine/threonine kinase, is the key mediator of integrin signaling pathway and regulates cell migration, cell proliferation, and apoptosis (Dedhar, 2000). We found a negative correlation between ILK and LC3B in H. pylori-positive human samples (Fig. 1). This conclusion was further con rmed by an in vitro experiment, which demonstrated that total and phosphorylated ILK were downregulated, and the numbers of autophagosome and autolysome were increased in GES-1 cell lines after H. pylori infection ( Fig. 2a-2d). As shown in Fig. 2a, CagA knockout H. pylori led lower level of ILK phosphorylation compared to that caused by H. pylori wild type, which indicated that ILK phosphorylation might be partly CagA-dependent pattern. The p62 levels decreased at 2 h and increased at 4, 6 and 8 h after H. pylori infection (Fig. 2e). The p62 is an adaptor protein that directs polyubiquitinated proteins to nascent autophagosomes, and disruption of autophagy results in the accumulation of p62 (Rubio et al., 2012).
Tang et al. found that p62 was autophagic degraded in the initial stage of H. pylori infection. With the prolongation of H. pylori exposure time, the autophagy process was interrupted and p62 accumulated in cells (Kamm and Stull, 2001). These results together con rmed that ILK might be a critical regulator for H. pylori-induced autophagy.
To further study the in uence of ILK on autophagy in H. pylori-infected GES-1 cells, RNAi was used to downregulate ILK level, and autophagy in GES-1 cells after H. pylori infection was then assessed. There was a signi cant accumulation of autophagosomes and increased LC3B levels both in H. pylori-infected and ILK-knockdown GES-1 cells (Fig. 3b-3e). The LC3B and p62 levels were both higher (P < 0.05) in ILKknockdown GES-1 cells than those in control siRNA-treated GES-1 cells after H. pylori infection. These results point to ILK as a novel regulator of autophagy, related to H. pylori infection strategy. ILK had been known to associated with the cytoplasmic tail of the β1 integrin, which was rst identi ed as a receptor of the H. pylori T4SS (Grashoff et al., 2012;Schuelein et al., 2011). The cytotoxin-associated genes pathogenicity island (cagPAI) encoded proteins to form a functional T4SS, and T4SS then delivered the virulence factors such as cagA into host target cells. Previous study had demonstrated that autophagy induction is independent on the cagPAI and T4SS (Deen et al., 2015;Terebiznik et al., 2009), thus further research on which component of H. pylori affected ILK and autophagy would be needed in the future. ILK can activate the family members of Rho GTPases (Rac1 and RhoA), which mediate changes in actin polymerization, thus appearing to be speci cally implicated in autophagy (Jang et al., 2012;Ivanovska et al., 2013;Xu et al., 2016). As shown in Fig. 4, Rac1 and RhoA activities were inhibited both in H. pyloriinfected GES-1 cells and ILK-knockdown GES-1 cells regardless of whether or not GES-1 cells were infected with H. pylori ( Fig. 5a and 5b). The results demonstrated that H. pylori reduced Rac1 and RhoA activities by affecting ILK levels. As shown in Fig. 5, inhibition of Rac1 or RhoA activities increased autophagosome and autolysosome formation in GES-1 cells. This indicated that the ILK/Rac1 and ILK/RhoA signaling pathways should affect autophagy in GES-1 cells.
Small GTP-binding proteins of the Rho family, such as Rac-1 and RhoA, can regulate the stability of newly formed actin laments and actin dynamics by regulating downstream signal molecules (Ivanovska et al., 2013;Xu et al., 2016;Yuan et al., 2017). The Rho GTPase Rac1 can bind to and modulate the activation of PAK1 (Xu et al., 2014), which can also interact with LIMK1 to stabilize actin lament dynamics and inhibit downstream co lin, an actin depolymerization agent (Hou et al., 2016). Co lin-1 acts as a terminal effector of the Rho GTPases signaling cascade, which regulates actin dynamics by increasing the depolymerization of F-actin laments (Wang et al., 2015). As shown in Fig. 6, the mRNA and protein phosphorylation levels of PAK1 and LIMK were signi cantly increased and mRNA and protein phosphorylation level of co lin was signi cantly decreased in GES-1 cells after H. pylori infection. The results demonstrated that H. pylori activated Rac1/PAK1/LIMK1/co lin signaling pathway in GES-1 cells. RhoA is the best-characterized Rho-associated coiled-coil containing kinases (ROCKs) regulators (Watanabe et al., 2015). ROCKs activation is accomplished by phosphorylation of several downstream target proteins to promote the formation of stress bers (Watanabe et al., 2015). LIMK1 is also downstream of ROCKs, which phosphorylate and inactivate co lin, thereby leading to a dynamic regulation of actin cytoskeleton (Ivanovska et al., 2013). There are two mammalian ROCK homologs, namely ROCK1 and ROCK2. Activated ROCK1 binds and phosphorylates to Beclin1 at Thr119 site to promote autophagy. Gurkar et al. showed that mice knocked out of ROCK1 gene resulted in impaired autophagy and reduced autophagosome formation (Zhang et al., 2018). In this study, ROCK1 and ROCK2 mRNA levels were signi cantly up-regulated, and ROCK1 phosphorylation levels were no obvious change in GES-1 cells after H. pylori infection. the mRNA and protein phosphorylation levels of ROCK1, ROCK2 and LIMK were signi cantly reduced and mRNA and protein phosphorylation level of co lin was signi cantly increased in GES-1 cells treated with the RhoA inhibitor CCG-1423 (Fig. 6); this indicated that RhoA could control ROCK1/ROCK2/LIMK1/co lin signaling pathway in H. pylori-infection GES-1 cells.

Conclusion
In summary, ILK could be a novel regulator of autophagy related to H. pylori infection strategy. We found a negative correlation between ILK and LC3B in H. pylori-positive human samples and H. pylori-infected GES-1 cell lines. Low expression of ILK targeted Rac1 and RhoA to induce autophagosome formation, and inhibited autophagy ux in GES-1 cells. Thus, H. pylori regulated ILK to disrupt autophagy through Rac1 and RhoA signaling pathways in gastric epithelial cells.

Clinical specimens
Eighteen H. pylori-positive and Nighteen H. pylori-negative gastric tissues were collected form patients at the Yantai A liated Hospital of Binzhou Medical University (Yantai, China). The diagnoses were based on clinical and histological laboratory examination.

Immuno uorescence for LC3B and ILK
Clinical samples of human stomach sections stored in paraformaldehyde were dehydrated, embedded in para n, sliced (Green eld, Jones, 2013). Slides were treated with 3% H 2 O 2 for 30 minutes and blocked with sheep serum for 1 h. The slides were incubated with primary antibody (anti-LC3B or anti-ILK, respectively) and a secondary uorescent antibody, and then the nuclei were stained by DAPI. Sections were evaluated using laser scanning confocal microscopy (Zeiss MIC-SYSTEM). On randomly selected images of gastric epithelial cells (n ≥ 180), the average uorescence intensity of LC3B and ILK signals in human biopsies was measured by the software ZEN 2.5 lite of confocal microscopy.

Human gastric epithelial cells
The GES-1 cell lines maintained in Dulbecco Modi ed Essential Medium (DMEM) with 10% fetal bovine serum (FBS) under a humidi ed atmosphere of 5% CO 2 at 37˚C. When cells reached approximately 70%-90% con uence in 6-well plates, cells were treated with 0.25% trypsin and passaged .

Co-culture of GES-1 cells with H. pylori
H. pylori strain 26695 was from the H. pylori Research Laboratory of the Chinese Center for Disease Control and Prevention (Beijing, China). Bacteria were grown on chocolate agar plate supplemented with 5% sheep's blood at 37 ˚C under microaerophilic conditions (85% N 2 , 10% CO 2 and 5% O 2 ) . GES-1 cells were plated onto 6-well plates in Dulbecco Modi ed Essential Medium (DMEM) with 10% fetal bovine serum (FBS) overnight, and then co-cultured with H. pylori at a multiplicity of infection (MOI) of 1:100 for 0, 2, 4, 6 and 8 h. GES-1 cells were immediately used to detect total ILK, phosphorylated ILK and p62 expression by western blot analysis and LC3B expression by mRFP-GFP-LC3B-Adenovirus transfection.

Western blot analysis
The GES-1 cells were lysed with 50 µL cell lysis buffer [RIPA (50 mM Tris pH 7.4, 0.1% SDS, 1% TritonX-100, 150 mM NaCl, 1 mM EDTA):PMSF = 16:1]. BCA Protein Assay Kit was used to determine the protein content. The 40 mg total protein was carried out by 12% SDS polyacrylamide gel electrophoresis and transferred to the polyvinylidene di uoride (PVDF) membrane. The membranes were incubated overnight with the primary antibody at 4 ˚C, and then incubated with secondary antibody at room temperature for 2 h. According to the manufacturer's instruction, the blots were detected with an enhanced chemiluminescence detection kit. GAPDH was used as control. ImageJ soft-ware was used to quantify the bands intensities (Zhang et al., 2017).

Adenovirus transfection
For transfection of mRFP-GFP-LC3B adenovirus, GES-1 cells were grown on coverslips at a density of 1.0 × 10 5 cells/well in Opti-MEM. The adenovirus was transfected into GES-1 cells (MOI = 100:1) for 12 h at 37 ˚C, and then maintained in DMEM for 48 h. Cells were washed three times with PBS and xed with 4% of paraformaldehyde. Samples were analyzed for merge of GFP-LC3B and mRFP-LC3B (yellow), and mRFP-LC3B (red) puncta on confocal microscope at 40× (Tang et al., 2012). The numbers of yellow spots (autophagosome) and red spots (autolysosome) were calculated. During this experiment, adenovirus was rst transfected into cells for 12 h, and ILK siRNA was used to interfere with the expression of ILK for 48 h, and then co-cultured with H. pylori for 6 h to observe the distribution of uorescent spots in the cells.

Knockdown of ILK in GES-1 cells co-cultured withH. pylori
RNA interference (RNAi) was performed as previously described (Zhang et al., 2017). Brie y, knockdown of ILK was performed using siRNAs of HSS140843, HSS179923 and HSS179924 (Table 1). GES-1 were transfected with 100 nmol L − 1 ILK siRNA using a Lipofectamine RNAiMAX Transfection kit according to the manufacturer's protocol. Brie y, 1.0 × 10 5 GES-1 cells/well were mixed with 2.5 µL ILK siRNA and 7.5 µL Lipofectamine RNAiMAX into 500 µL serum-free DMEM were co-cultured at 37 ˚C for 48 h in a humidi ed atmosphere of 5% CO 2 . The GES-1 cells with control siRNA were used as the negative group.
Western blot was employed to detect the effect of ILK knockdown according to the previous methods.
H. pylori were harvested and quanti ed using a spectrophotometer; quantities reached 1.0 × 10 8 CFU mL − 1 . The GES-1 cells, GES-1 cells with control siRNA, and GES-1 cells transfected with ILK siRNA (HSS140843) were respectively seeded at a density of 1.0 × 10 5 cells/well, and then were incubated with the H. pylori at a multiplicity of infection of 1:100. The untreated GES-1 cells were as the blank group. The control siRNA transfected GES-1 cells were considered to be the negative group. The ILK siRNA transfected GES-1 cells were used as the siRNA group. The cells were then harvested after 6 h and assessed for adenovirus transfection, transmission electron microscopy, western blot. In order to determine whether ILK reduction was CagA-dependent after H. pylori infection, the GES-1 cells were cocultured with CagA-knockout H. pylori (CagA − ) and wild-type H. pylori (CagA + ). The cells were then harvested after 6 h for western blot.

Transmission electron microscopy
For the transmission electron microscopy (TEM) to analyze autophagosomes and autolysosome, GES-1 cells were xed in 2.5% glutaraldehyde at 4 ˚C for overnight. The GES-1 cells were washed three times with PBS (0.1mol•L − 1 , pH 7.2) and then xed again in 1% osmium acid solution for 1.5 h, which were dehydrated in gradient ethanol, embedded in Epon. The Epon was cut into ultrathin sections with a thickness of 70 nm, and then it was dyed with 2% uranium dioxide acetate solution in dark for 20 min, and then dyed again with lead citrate solution for 7 min. Observation and photography of cells were used by transmission electron microscopy (Hitachi HT-7800) at 80 kV (Dejaeger et al., 2017).

Cell survival rate assays
Cell survival rates were assessed using Meilun Cell Counting Kit-8 (Meilunbio, China) according to the manufacturer's instructions. Brie y, the GES-1 cells were seeded on a 96-well plate at a concentration of 5 × 10 3 cells per well. The GES-1 cells were treated with Rac1 inhibitor NSC23766 (10, 50, 100 and 200 µM) or RhoA inhibitor CCG-1423 (2, 5, 10 and 25 µM) for 2 h in an incubator at 37 °C in 5% CO 2 to detect concentration-dependent toxicity of inhibitors. The GES-1 cells were treated with 50 µM of NSC23766 or 5 µM of CCG-1423 at 1, 2, 4, 8 and 12 h in an incubator at 37 °C in 5% CO 2 to detect time-dependent toxicity of inhibitors. After treatment duration, the CCK-8 assay reagent was added to culture media and incubated for 2 h. Absorbance was read at 450 nm on a multifunctional microplate reader (Wang et al., 2019).

RhoA and Rac1 activation assay
RhoA and Rac1 activation assays were measured using RhoA/Rac1/Cdc42 Combo Activation Assay Kit (Abcam, ab211168) according to the manufacturer's instructions. Brie y, the GES-1 cells were lysed in cold lysis buffer containing protease inhibitor for 20 min. Thereafter, they were centrifuged at 14,000 × g for 10 min at 4 °C, and the supernatant was then for measurement of RhoA and Rac1 kinase activity. 40 µL of Rhotekein RBD beads or PAK1 PBD beads were separately added to the supernatant for 1 h at 4 °C to bind to GTP-RhoA or GTP-Rac1, and then centrifuged at 14,000 × g for 10 s at 4 °C. Mixture of the supernant and 5 × SDS-PAGE sample buffer was boiled at 100 °C for 10 min as total RhoA or Rac1 (T-RhoA or T-Rac1). The pellet was washed three times with 0.5 mL of 1 × Assay Buffer, resuspended in 40 µL of 2 × SDS-PAGE sample buffer and then boiled at 100 °C for 10 min. Thereafter, they were centrifuged at 14,000 × g for 10 s at 4 °C, and the supernatant was used as active RhoA or Rac1 (GTP-RhoA or GTP-Rac1). The active and total RhoA, or active and total Rac1 were analyzed by western blot (Barbati et al., 2015).

Drug treatment in GES-1 cells co-cultured withH. pylori
The Drugs NSC 23766 or CCG-1423 were used as inhibitors of Rac1 or RhoA, which was dissolved in 10% DMSO and diluted with Opti-MEM. Brie y, GES-1 cells were grown at a density of 1.0 × 10 5 cells/well in Opti-MEM and treated with either NSC 23766 (50 µM) (Maekawa et al., 1999) or CCG-1423 (5 µM) (Montani et al., 2009) for 2 h. The untreated GES-1 cells were employed the blank group. H. pylori were added to the cells at a multiplicity of infection of 100:1. The cells were then harvested after 6 h and assessed for adenovirus transfection, western blot and real-time PCR. 5.13 Quanti cation of ROCK1, ROCK2, PAK1, LIMK1 and co lin1 by quantitative real-time RT-PCR The ROCK1, ROCK2, PAK1, LIMK1 and co lin1 mRNA expression in GES-1 were detected by quantitative real-time polymerase chain reaction (RT-PCR). The SYBR Green RT-PCR assay was performed in an ABI PRISM 7500 Sequence Detection System (ThermoFisher Scienti c, USA). The ampli cations were conducted in 20 µL using the following thermal pro le: 95 °C for 3 min, followed by 40 cycles at 95 °C for 15 s and 58 °C for 33 s. The ROCK1-speci c primers ROCK1-F and ROCK1-R, ROCK2-speci c primers ROCK2-F and ROCK2-R, PAK1-speci c primers PAK1-F and PAK1-R, LIMK1-speci c primers LIMK1-F and LIMK1-R, co lin1-speci c primers CFL1-F and CFL1-R and GAPDH-speci c primers GAPDH-F and GAPDH-R (Table 1), were used to amplify the corresponding products. The comparative average-cycle threshold method was used to analyze ROCK1, ROCK2, PAK1, LIMK1 and co lin1 mRNA levels, which could be calculated with an n-fold difference relative to the GAPDH (Zhu et al., 2012).

Statistical analysis
Data are presented as mean ± standard deviations (SD) from at least ve independent experiments. Pearson correlation coe cient was used to study the relationship between LC3B and ILK in clinical samples. All data were subjected to one-way analysis of variance. P ≤ 0.05 was considered statistically signi cant. 6. Declarations 6.1 Ethics approval and consent to participate Informed consent for study has been obtained from all patients. The present study was sanctioned by the ethics committees of Binzhou Medical University.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This work was supported by grants from the National Natural Science Foundation of China (grant numbers: 81771709 and 81471561); Department of Education of Shandong Province (2019KJK012).
6.6 Authors's contributions YZ and BL were major contributors in project administration, funding acquisition and writing-review and editing. ZX was a major contributor in methodology, data curation and writing-original draft preparation. YD, RZ and XT performed formal analysis and investigation. YW and XJ performed western blot assays.

Acknowledgements
Not applicable.

Figure 3
Downregulation of ILK in uenced autophagy in GES-1 cells. (a) Western blot analysis to detect ILK expression at 48 h in GES-1 cells after transfection with sequence-speci c siRNA (HSS140843, HSS179923, or HSS179924) targeting ILK. Data are presented as mean ± SD of ve independent assays.
Two asterisks indicate signi cant difference compared to that treated with control siRNA (siControl). (b) GES-1 cells, which had been transfected with mRFP-GFP-LC3B adenovirus for 12 h, were transfected with control siRNA (siControl) or ILK siRNA (siILK) for 48 h and then infected with or without H. pylori for 6 h.
Scale bar: 10 µm. (c) The number of autophagosome and autolysosome was calculated. Data are presented as mean ± SD of 30 cells. (d) Ultrastructural features of GES-1 cells were analyzed by transmission electron microscopy. Ultrastructural features of H. pylori-uninfected and H. pylori-infected GES-1 cells are shown in the rst and third lines, respectively. Scale bar: 5 μm. The typical images of autophagosomes (red arrow) in the second and fourth lines are shown at higher magni cation according to ultrastructural features of the rst and third lines, respectively. Scale bar: 1 μm. (e) Western blot analysis to detect LC3B and p62 in blank, siControl-or siILK-treatment GES-1 cells infected with or without H. pylori for 6 h. Data are presented as mean ± SD of ve independent assays (*P < 0.05, **P < 0.01).

Figure 4
The Rac1 and Rho activation assay in the GES-1 cells. The GTP-Rac1/T-Rac1 (a) and GTP-RhoA/T-RhoA (b) levels in ILK-knockdown GES-1 cells after H. pylori infection through western blot analysis. Blank: The untreated GES-1 cells; siILK: The GES-1 cells were transfected with ILK siRNA; siControl: The GES-1 cells were transfected with control siRNA. The quantitative data are presented as means ± SD of 5 independent assays (*P < 0.05).