Overexpressions of RHOA, CSNK1A1, DVL2, FZD8, and LRP5 Genes Enhance Gastric Cancer Development in the Presence of Helicobacter Pylori

Introduction: Intestinal metaplasia (IM), and Helicobacter pylori (HP) infection can be shown as risk factors in the development of gastric cancer (GC). WNT signaling pathway plays a critical role in carcinogenesis. However, the literature studies are limited on the signicance of this pathway for the transition from IM to GC. Methods: We aimed to investigate the importance of the genes of WNT signaling pathways diagnostic and prognostic markers in the presence and absence of HP in conversion from IM to GC. 104 patients, (GC group n=35, IM group n=45, control group n=25) were included in this case-control study. Expression of genes in WNT signalling were searched in study groups with qRT-PCR array and qRT-PCR method. Data were analysed using PCR array data analysis software. Results: Statistically signicant overexpression of RHOA, CSNK1A1, DVL2, FZD8 and LRP5 genes was detected in the GC and IM groups compared to the control group (p <0.05). Statistically signicant overexpression of RHOA, CSNK1A1, DVL2, FZD8 and LRP5 genes was observed in patients with metastatic GC compared to patients with GC without metastasis (p <0.05). It was found that the RHOA, CSNK1A1, DVL2, FZD8 and LRP5 genes were statistically signicantly over-expressed in diffuse GC patients compared to non-diffuse GC patients (p <0.05). Statistically signicant overexpression of RHOA, CSNK1A1, DVL2, FZD8 and LRP5 genes was detected in HP positive IM patients compared to HP negative IM patients (p <0.05). Conclusion: Overexpression of RHOA, CSNK1A1, DVL2, FZD8 and LRP5 genes in IM may suggest that these genes are important markers in the development of IM and inammation with HP. In addition, these genes are linked to tumor burden in the GC group. Consequently, we can conclude that these genes are poor prognosis biomarkers for GC and have the potential to be used as markers for future treatment monitoring.


Introduction
Intestinal metaplasia (IM) and Helicobacter Pylori (HP) are major risk factors for gastric cancer (GC). IM usually develops as a result of atrophic gastritis (AG) infected by HP and may cause gastric dysplasia and/or cancer [1,2]. Throughout this development process, HP triggers the response by CD4 Th1 cells due to in ammation. Thus, macrophages and neutrophils soon reach this site and are stimulated by the release of reactive oxygen and reactive nitrogen products. These reactive substances cause DNA damage with superoxide and hydroxyl ions created by the HP, as well as oxidative damage [3,4]. HP is considered to trigger cause intestinal type GC with two major direct and indirect mechanisms. The indirect mechanism includes in ammation due to increased cell cycle and accumulation mitotic defects within a few decades. In this process called Correa hypothesis, AG occurs rst, followed by IM, colonic metaplasia, dysplasia, and nally GC. The other mechanism is directly associated with DNA damage [3].
WNT signaling pathway which preserved in the evolutionary process have a role in important processes including development, differentiation, cell proliferation, mothology, mortality etc. [5,6]. Therefore, functional genes on the WNT signaling pathway are important for carcinogenesis and have been intensely studied with respect to the pathophysiology of carcinogenesis. The aberrant activation of the WNT signaling pathway is involved in the development and progression of a signi cant proportion of GC [7][8][9]. In a review by Chiurillo (2015), it was suggested that WNT signallingdue to the HP infection and suppression of WNT inhibitors play crucial role in GC. Also, CTNNB1 and APC were listed as pivotal genes in the development of GC [8]. GC aggressiveness is dependent on WNT5A expression level [10]. WNT5A may help building favorable conditions for GC progression (Kazawa 2013) [11]. The RHOA gene in the WNT signaling pathway was described as both a potential therapeutic target and a biomarker for GC (Chang et al. 2016) [12]. However, the literature studies are limited on the signi cance of this pathway for GC development in the presence of HP. In this context, this case-control study investigated the activity of WNT signaling pathway genes as diagnostic and prognostic markers in the presence and absence of HP in development process from IM to GC.

Ethics
Permission for the study was granted by the Local Council of Ethics in 2013 (#147). A signed informed consent form was obtained from each subject prior to any study related procedures.

Study Design and Patients
Between September 2014 and September 2015, 34 GC cases with upper gastrointestinal endoscopy and 45 IM cases who applied to Manisa Celal Bayar University Gastroenterology Department were included in the study. In addition, 25 patients who with normal biopsy-endoscopy underwent endoscopic biopsy due to upper gastrointestinal system/functional dyspepsia complaints, with normal biopsy-endoscopy were selected as the control group. Age, gender, and family history of all subjects were questioned and recorded.

Inclusion Criteria
(1) Adult patients older than 18 years who have not received chemotherapy or radiotherapy treatment with new GC diagnosis, (2) Adult patients over 18 years of age who were diagnosed as IM and underwent control endoscopy for dysplasia, (3) Adult healthy volunteers over 18 years of age who had endoscopic biopsy of the cause of functional dyspepsia and whose endoscopy was normal and biopsy pathology was normal.

Exclusion Criteria
(1) Taking chemotherapy or radiotherapy at any time with the diagnosis of GC or another malignancy, (2)Having a malignancy other than GC,(3)Patients who did not accept biopsy during endoscopy or did not receive biopsy because they were using antibiotics (4)Cases who underwent subtotal or total gastrectomies for any reasonPET or PET/CT images were taken for staging purposes in all GC patients. CT or PET/CT evaluation of opere and non-opere patients and whose postoperative pathological evaluation were performed according to TNM classi cation of AJCC.

Endoscopy and Tissue Sampling
Endoscopic examinations of all cases were carried out at Manisa Celal Bayar University, Department of Gastroenterology. Endoscopy was performed with Olympus Lucera CLV260SL device. At least 2 biopsy specimens of antrum and corpustum were taken from tumor tissue in GC patient group, IM, and control patient group using olympus biopsy forceps. Pathology evaluations of biopsy specimens were performed and the data in the results were recorded.
Mucosal biopsy specimens taken for analysis of WNT signaling gene expression from all cases were immediately stored in RNAlater RNA stabilization reagent (Qiagen, Hilden, Germany) solution.

RNA Isolation
RNA from tissue was isolated using RNeasy Mini Kit (Qiagen, Hilden, Germany) according to Kasap  Amount and Purity of RNA.
The amount of RNA was measured at absorbance at 260 nm (A260) and RNA purity was determined by the ratio A260 / A280 using spectrophotometer.  Data analysis PCR array and qPCR primary assay data were analyzed in the online system (http://www.sabiosciences.com/pcrarraydataanalysis.php). Fold change and fold regulation values above 2, indicative of overexpressed genes; fold change values below 0.5 and fold regulation values below − 2 were considered as indicative of underexpressed genes.

Statistics
In the results, mean values are shown as ± standard deviation, and the 2 -ΔCt values obtained for each gene in the study group were analyzed based on Student's t-test. A p value less than or equal to 0.05 was considered statistically signi cant.

Patient Groups
When the three groups in the study were compared, there was no difference between the age groups and gender distribution ( Table 2). In patients with GC, the tumor was largely localized in the proximal stomach. Localization-based distribution was as follows: Proximal Stomach: 47% (n = 16), Distal Stomach: 35% (n = 12), and Diffuse Stomach Involvement: 18% (n = 16). Thirty of 34 GC cases had full-body computed tomography and distant metastasis was detected in 52% (n = 16) of these cases. Fifty-six percent (n = 9) of those patients had a peritoneal metastasis, 32% (n = 5) had a liver metastasis, 6% (n = 1) had a pancreatic metastasis, and 6% (n = 1) had a lung metastasis.
Regarding survival, 29% (n = 10) of the patients with GC survived less than 3 months, and 24% (n = 8) survived for a period between 3 months and 1 year.
When GC cases were compared with IM cases, the GC group displayed no signi cantly overexpressed gene with regard to the latter (p > 0.05).

Gene Expression Analysis based on site of Gastric Involvement
In patients with GC, a comparison between those with a diffuse gastric involvement (n = 6, 18%) and the control group showed a signi cantly higher overexpression of RHOA, CCND2, DVL2, FZD8, and LRP5 genes (p < 0.05).
A comparison of metastatic GC cases with non-metastatic ones showed that RHOA, CCND2, DVL2, FZD8, and LRP5 genes were overexpressed on a higher level, which was signi cant (p < 0.05).

Gene Expression Analysis Based on Survival of Patients with GC
Regarding survival, 53% (n = 16) of the patients with gastric cancer became exitus. Fifty percent of them (n = 8) survived for 3 months, 19% (n = 3) 6 months, and 31% (n = 5) 12 months. The patients who became exitus had a signi cantly higher overexpression of RHOA, CCND2, DVL2, FZD8, and LRP5 genes than the control group (p < 0.05). There was no signi cant difference between exitus and surviving patients in terms of gene overexpression (p > 0.05).
A gene analysis comparison between HP-positive and HP-negative IM cases revealed that overexpressed RHOA, CCND2, DVL2, FZD8, and LRP5 genes were signi cantly higher in HP-positive patients (p < 0.05).

Discussion
The present study focuses on the overexpression of RHOA, CXADR, CSNK1A1, CCND2, DVL2, FZD8, and NFACT1 genes belonging to the WNT signaling pathway. Expression rates were compared with GC, IM patients and control group. In addition, effects of HP on the expression of these genes in IM patients have been evaluated.
RHOA (Ras homolog gene family member A), a key step along the WNT signaling pathway, is a member of RHO family, which is considered to play a role in the development of various cancers and active in cellular signal transmission over GTPase (Maeda et al. 2015) [17]. It plays a role in events such as cell cycle, polarization, and migration. RHOA GTPase activation may be altered in the tumor tissue due to mutation. Consequently, GTPase-in uenced G1-S phase of the cell cycle may gain pace or may slow down. A RHOA mutation is obtained in GC by cell culture assays. In these mutations, RhoA-speci c siRNA, which is active in the G1-S domain of cell division, is detected ( [24]. In another analysis by Chang et al (2016), the role of RHOA pathway dysregulation in GC and the feasibility of clinically targeting RHOA for GC therapy were systematically assessed. As a result, they demonstrated that RHOA gene in the WNT signaling pathway was described as both a potential therapeutic target and a biomarker for the Asian GC [12].
The WNT signaling pathway gets activated via canonical and non-canonical ways. In the non-canonical   [32][33][34]. A study by Metcalfe et al (2010) found overexpressed DVL2 in colorectal adenocarcinoma (CRC) cells. Based on the idea that the inhibition of DVL2 by siRNA reduces the activity of Beta-catenin signaling pathway, it was thought that DVL2 activation in patients with CRC led to a hyper-activated Beta-catenin signaling pathway. An analysis conducted on 24 patients with CRC detected that the DVL2 level in 1/3 of CRC cells was higher than that of the normal tissue on the tumor surgical border. The study proceeded by increasing the number of subjects to 393. Patients were divided into the groups of normal mucosa, hyperplastic, adenomatous polyp, and CRC. When Beta-catenin-positive nuclei were ordered from normal tissue to the carcinomatous tissue, the expression rate was observed to increase within the cell groups. This increased expression level was also the case for Axin and DVL2, as well. The preservation of this rate across different tumor stages suggests the upregulation of DVL2 and Axin genes in patients with CRC in response to APC loss. With a WNT simulation, the same study also demonstrated that the upregulation of DVL2 might also occur post-transcriptionally. This led the authors to evaluate DVL2 in APC-mutant rats. A low level of DVL2 was found proportionated to reduced tumor burden; and shortened small intestine, as well as reduced number and diameter of crypts were observed in DVL2 reduction [35]. A study by Chong et al (2014) found that DVL2 and ETS1 proteins were more speci c to diffuse and mixed type GC than the intestinal type. Both genes display a methylation similar to the normal tissue and intestinal type GC, whereas the methylation rate was high in the mixed and diffuse type GC. These ndings suggest that abnormal methylation starts from the non-carcinomatous normal tissue in the mixed and diffuse type GC [36].
CyclinD2 (CCND2) is a member of the cyclin family. CCND2 regulates CDK kinase and forms a complex with CDK4 and CDK6. CCND2 is the regulator of this complex. It enables cell cycle to proceed from G1 to S phase. CCND2 also participates in the phosphorylation of RB (Retinoblastoma-associated protein), which is a tumor suppressor protein (Truong et al. 2015) [37]. Therefore, the activation of CDK/Cyclin complex may gain importance in carcinogenesis (Myant et al. 2011) [38]. In a study conducted on rats by Cole et al (2010), authors suggested that the inhibition of WNT signaling pathway could prove to be a treatment alternative for patients with CRC. The study demonstrated that Cyclin D/CDK4-6 complex increased in the intestinal tissues of APC-negative rats. In APC-positive rats, however, CyclinD2 level was low in the intestinal tissue. The most important ndings were as follows: CyclinD2 did not affect normal enterocyte development and CDK4/6 inhibition suppressed the adenomatous tissue in knockout rats [39]. A retrospective study conducted by Shi et al (2015) on patients with GC demonstrated signi cantly increased CyclinD2 (CCND2) in the tumorous tissue compared to the normal tissue (P < 0.001). The same analysis also detected that increased CCND2 expression was associated with the depth of invasion, lymph node positivity and distant metastasis, and advanced TNM staging. Authors suggested that increased CCND2 expression could be a prognostic factor [40].
Two casein proteins were puri ed: Casein kinase (CK) proteins CD1 and CD2. Casein kinase 1 (CK1) is a monomeric protein of a serine and threonine structure. It takes part in tasks such as DNA repair, cell division, nuclear settlement, membrane transport, etc. CK1-alpha is thought to phosphorylate beta-catenin and played a role in the degradation process when it is not bound to the WNT receptor (FZD  [42]. Gamma-delta T-cells in the epithelial tissue play a critical role for defense against environmental factors such as infection, trauma and malignancy. The CXADR gene, which is included in the present study, has been previously isolated in pancreatic and small intestine tissues and demonstrated to be active in cell proliferation through cytokines, which suggests that this gene could be active in GIS adenocarcinomas. Studies on NFACT1 gene, which is a part of this study, are largely conducted osteocytes. There is no study available on its relationship with adenocarcinomas; however, it has been demonstrated that HP causes the inhibition of NFACT, increases VacA secretion, and consequently leads to the vacuolization of epithelial cells. This intensi es the process whereby HP becomes chronic (Gebert et al. 2003) [43].
Although there is no relevant study in the literature and our ndings contradict it, there are some data available to claim that HP becomes chronic by inhibiting NFACT, which suggests that this gene may be associated with adenocarcinoma.

A review of our ndings
We classi ed the ndings of genes in two groups. The rst group included the genes RHOA, LRP5, FZD8, DVL2, and CCND2. Similar to the reports in the literature, RHOA, LRP5, FZD8, DVL2, and CCND2 were found signi cantly overexpressed compared to the patients in the control group. Again, similar to the ndings reported in the literature, we detected that these genes were signi cantly overexpressed in diffuse type GC patients compared with both the control group and patients without a diffuse involvement. Patients with a distant metastasis had a signi cantly higher overexpression of RHOA, LRP5, FZD8, DVL2, and CCND2 than non-metastatic patients. This nding contradicted previous reports for RHOA. Compared with survival, exitus patients presented an overexpression of RHOA, LRP5, FZD8, DVL2, and CCND2, although the difference was not signi cant. As part of the present study, we checked RHOA, LRP5, FZD8, DVL2, and CCND2 in patients with IM and found that they were signi cantly overexpressed with respect to the control group. HP-positive and/or HP-negative patients had a signi cantly higher overexpression of RHOA, LRP5, FZD8, DVL2, and CCND2. Patients with HP-positive IM showed signi cantly overexpressed RHOA, LRP5, FZD8, DVL2, and CCND2 than the HP-negative ones. A comparison with GC cases revealed no signi cant difference from the patients with IM in terms of gene expression (p > 0.05). RHOA, LRP5, FZD8, DVL2, and CCND2 were overexpressed in HP-positive and HP-negative patients. The overexpression was higher in the former group than the latter. The second group included the genes CSNK1A1, CXADR, and NFACT. CSNK1A1 showed no signi cant overexpression in GC or IM patient groups compared with the control group. Contrary to previous reports on patients with adenocarcinoma, the present study did not detect a signi cantly overexpressed CXADR in GC or IM patient groups than the control group. Although our ndings do not support it, previous reports claiming that HP becomes chronic by causing NFACT inhibition suggest that this gene could be associated with adenocarcinoma. Conclusion 1. We are of the opinion that the signi cantly higher overexpression of RHOA, LRP5, FZD8, DVL2, and CCND2 in the patient group than the control group and in the metastatic group than the nonmetastatic group as well as the fact that all of them were overexpressed in patients with diffuse involvement and metastatic patients are associated with tumor burden. The absence of a signi cant difference between exitus and surviving patients in the GC group suggest that it may not be related to prognosis. Analyses with a larger sample size and to include other factors may provide us with deeper knowledge about its impact on survival.
2. Given that RHOA, LRP5, FZD8, DVL2, and CCND2 were signi cantly overexpressed in IM cases compared with the control group, one may think that carcinogenesis starts at the IM phase.
3. The signi cant overexpression of RHOA, LRP5, FZD8, DVL2, and CCND2 in HP-positive patients than the HP-negative ones suggest that HP might in uence gene expression by intensifying in ammation. These ndings need to be supported by further studies. 4. In accordance with the HP-B/catenin relationship reported in the literature, the present study also found that HP-positive and/or HP-negative IM cases presented a signi cant overexpression of RHOA, LRP5, FZD8, DVL2, and CCND2 compared with the control group. HP-positive IM cases had a signi cantly higher overexpression of RHOA, LRP5, FZD8, DVL2 and CCND2 than the HP-negative cases.
5. It is likely that RHOA, LRP5, FZD8, DVL2, and CCND2, which play a key role in the gastric carcinogenesis, are triggered by in ammation and in uenced by environmental factors such as HP.
. In the light of the data we obtained and the literature, we are of the opinion that CSNK1A1 does not have an impact on the pathogenesis of adenocarcinoma.
7. Contrary to the data in the literature, we did not detect a signi cant overexpression of CXADR in GC or IM cases compared with the control group. Therefore, further studies are required to shed light on its role in the pathogenesis of GC.
. Although it is not supported by our data, previous studies reporting that HP becomes chronic by causing the inhibition of NFACT suggest that this gene might be associated with adenocarcinoma.
More studies with larger study groups are needed to elucidate the role of the detected genes in the pathogenesis of GC.

Limitations
The sample size in our study is limited. A total of 11 cases were excluded from the study due to RNA degradation. In this respect, this study was categorized as a case-control study. Further and comprehensive studies are needed for these genes about this topic.

Declarations
Funding: This study was supported by Manisa Celal Bayar University Scienti c Research Projects (#2014-094). WNT Gene Expressions in HP Negative IM