GKN1 is absent in human gastric tumors and acts as a tumor suppressor, regulating cell proliferation, apoptosis, migration and invasion in gastric cancer cell lines (10). Stimulating the expression of Fas receptor and the activation of caspase-3, this protein modulates apoptotic signals, playing an important role in the repair of tissues during the early stages of neoplastic transformation (7).

In AGS, MKN-1 and MKN-28 gastric cancer cell lines transfected with GKN1, the re-expression of p16 and a reduction in CDK4, cyclin D1 and E2F levels was observed (8) In gastric cancer SGC7901 cells, GKN1 reduces the expression of MMP2, through the deactivation of NF-κB (15), and induces the expression of miRNA (miR)-185. The extreme 3′ untranslated region (UTR) of RhoA mRNA has sequences with affinity to miR-185 and, when this hybrid miRNA with RhoA mRNA reduces its translation, the silencing of RhoA is indirectly mediated by GKN1. c-Myc is a transcription factor that activates RhoA expression and is a target of miR-34a, a miRNA whose expression is promoted by GKN1. Thus, GKN1 also deactivates RhoA via miR-34a. These data suggest that GKN1 inhibits cell motility and invasion by means of the deactivation of RhoA (16).

Transcription factors are able to activate or repress the expression of a gene (63,64). Little is known about the transcriptional regulation of GKN1. Yoon et al (65), using luciferase and chromatin immunoprecipitation assays, confirmed that NKX6.3 is a transcription factor for GKN1, and located the recognition sequence corresponding to NKX6.3 in the promoter region of the GKN1 gene (Fig. 1A). NKX6.3 positively modulates the transcription of GKN1, which is reflected in the increased level of both mRNA and protein (65).

By means of in silico analysis, conducted using the MatInspector (66) (http://www.genomatix.de/matinspector.html), AliBaba2.1 (67) (http://gene-regulation.com/pub/programs/alibaba2/index.html) and TfsiteScan 68) (http://www.ifti.org) programs, transcription factors were identified with affinity to recognition sequences in the GKN1 promoter region (Fig. 2A and Table I). It is likely that one or more of these transcription factors, predicted bioinformatically, are involved in the transcriptional regulation of GKN1. Experimental confirmation of the effect exerted by the proposed transcription factors on the modulation of GKN1 expression will improve understanding of the mechanisms involved in the regulation of the expression of this protein.

In patients with gastric pathology, in murine models or in gastric epithelial cell lines, the expression of trans-acting T-cell-specific transcription factor GATA-3 (GATA-3), STAT-1, STAT-3, transcription factor Sp1 (Sp1), cyclic AMP-responsive element-binding protein 3-like protein 4 (CREB), AP-1 transcription factor (AP-1) and Oct-1 increases, while the levels of CCAAT enhancer binding protein-α (CEBPα) and NKX6.3 decrease (65,69-82). The expression of GATA-3 was found to have increased at different stages of the carcinogenesis associated with H. pylori in patient biopsies, murine models and human gastric epithelial cells (73,74).

CagA and OipA of H. pylori induce the activation of transcription factors such as AP-1, NF-κB, STAT-3, CREB and nuclear factor of activated T cells (NFAT), which favor the expression of IL-6, and cytokines, which promote inflammation (83-89). IL-6 stimulates the activation of the signaling pathway gp130/STAT3 in gastric cancer cell lines (90), while CagA stimulates the expression of the NFAT transcription factor in AGS cells (88).

The protein Tipα, produced by H. pylori, activates the IL-6/STAT3 pathway (89). H. pylori cagA⁺ strains induce signaling through the MAPK pathway, thus increasing proliferation and activating transcription factors such as AP-1 (70). Through toll like receptor (TLR)2 and TLR9, H. pylori activates the MAPK pathway and, downstream, the factors AP-1 and CREB, which positively regulate the transcription of cyclooxygenase 2 (COX-2) (91). CREB and STAT-3 are activated by H. pylori and positively regulate the transcription of COX-2 in gastric epithelial cells (72,83). Increased STAT-3 expression has been found in biopsies of the gastric mucosa infected with H. pylori cagA⁺ (80), as well as in cell lines and murine models (86). CagA promotes the phosphorylation of STAT-3 in gastric epithelial cells (92).

In vitro and in vivo experiments have shown that the protein OipA of H. pylori stimulates the phosphorylation of STAT-1 (93) and that H. pylori alters the STAT-1 signaling induced by IFN-γ in gastric epithelial cells. This event may represent an adaptation of the bacteria in order to modulate the immune response of the host mucosa, allowing the bacteria to survive in the stomach (94).

The expression level of the Sp1 transcription factor increases in gastric adenocarcinoma and is related to the cancer stage, the depth of infiltration and an unfavorable prognosis for patients (82). The expression of Sp1 differs between intestinal-type and diffuse-type cancer, while low-level expression of Sp1 is related to the progression and metastasis of intestinal-type cancer, in contrast to diffuse-type cancer (95). Sp1 is essential in the regulation of genes that determine the characteristics of cancer (96). In AGS cells, the ERK1/2 signaling pathway, activated in response to H. pylori infection, in turn activates Sp1, which modulates the transcription of vascular endothelial growth factor-A (69). It is probable that the factors AP-1, Oct-1, STAT-1, STAT-3, GATA-3, Sp1, CREB and NFAT, with recognition sequences in the GKN1 promotor and being activated by H. pylori, repress the transcription of GKN1 in infected mucosa or mucosa with gastric cancer.

In normal gastric mucosa, CEBPα is expressed in the foveolar epithelium and is reduced in the tumor tissue of patients with gastric cancer (79), and in the cell lines MKN45 and MKN74 (97). The ectopic expression of CEBPα in gastric cancer cell lines reduces cell viability (97). The level of CEBPα expression gradually decreases in line with the advancing carcinogenesis associated with H. pylori infection (73). Given the function of CEBPα in the regulation of the viability of cancerous cells and the fact that GKN1 and CEBPα levels gradually decrease in line with the progress of the lesion, it is probable that this factor is an activator of GKN1 transcription. The reduced expression of CEBPα and NKX6.3 in gastric cancer may be due to the negative regulation mediated by microRNAs.

Changes in the methylation of DNA lead to changes in gene expression. The hypermethylation of CpG islands located in the promoter region of a gene results in the decrease or silencing of the expression of the gene.

The methylation of the GKN1 promoter was previously studied, finding hypermethylation in the CpG islands of the promoter region in only two of 25 gastric tumors. This evidence indicates that the low or null GKN1 expression in the inflamed tissue, either tumoral or infected by H. pylori, is not due to the methylation of its promoter in all cases (Fig. 1B) (20).

The protein EBNA1 of EBV is able to bind to the promoter region of various genes of the host (28). It has been found to have an affinity with the sequences contained in the GKN1 promoter and, binding at these sites, contributes to the deregulation of GKN1 in gastric cancer associated with Epstein-Barr infection (Fig. 1C) (29). Only a small proportion of gastric tumors contain EBV.

The modification of histones is an epigenetic mechanism influencing gene expression. Altieri et al (26) analyzed six gastric tumors in order to determine whether histone modification contributes to GKN1 regulation. Chromatin immunoprecipitation assays were conducted on a fragment of 600 pb of the GKN1 gene promoter, including the 5′UTR, finding trimethylation in lysine 9 of histone 3 (H3K9triMe), among bases -148 and -310 of the GKN1 gene promoter in the six gastric tumors. H3K9triMe is a gene repression marker that generates binding sites for histone deacetylase I (HDAC1) (98) (Fig. 1D). The inhibition of HDAC1 activity with trichostatin A, a hypomethylating agent, is related to increased GKN1 mRNA levels but not to the protein itself. These findings suggest that the regulation of GKN1 may occur at the post-transcriptional level via miRNAs (26).

In the gastric mucosa, miRNAs can be expressed by epithelial cells, infiltrating inflammatory cells, transformed cells or cancerous cells (31). In the regulation of gene expression, miRNAs inhibit the translation or induce the degradation of target transcripts. It is likely that some miRNAs impede the translation of GKN1 mRNA and, consequently, are responsible for the reduction in the protein level, although there are no reports indicating whether a miRNA is involved in the regulation of GKN1 expression to the best of our knowledge.

In order to explore whether GKN1 mRNA has binding sites for one or more miRNAs, an in silico analysis was conducted using programs for predicting miRNA targets: TargetScan (99) (2015, http://www.targetscan.org), miRanda (100) (http://www.microrna.org), miRDB (101) (http://mirdb.org), miRSystem (102) (http://mirsystem.cgm.ntu.edu.tw/index.php) and DianaTools (103) (http://www.microrna.gr/microT-CDS), based on thermodynamic and base complementarity analysis. miRNAs were found with sequences complementary to sites located in the 3′UTR region of GKN1 mRNA (Fig. 2B), four of which are able to hybridize with canonical sites of GKN1 mRNA and possess two or three guanine or cytosine residues in the seed region of the miRNA, conferring them greater binding stability. An adenine in position 1 of the 3′UTR region of GKN1 mRNA ensures the recognition of the transcript by the RNA-induced silencing complex. These characteristics increase the probability that a miRNA will interact with the 3′UTR of GKN1 mRNA (Table II).

Multiple prediction programs may be used to locate binding sites for miRNAs in gene transcripts. The results of the analysis facilitated the selection of miRNAs with a higher probability of binding to their target, miR-544a was predicted by five programs, according to the aforementioned criteria. It is highly probable that miR-544a is a regulator of GKN1 (104,105). This proposal is strengthened by experimental data on the expression of miR-544a, which is found at increased levels in gastric cancer cell lines (106). To the best of our knowledge, research has not been conducted on the expression of hsa-miR-1245b-3p, hsa-miR-892c-5p and hsa-miR-548d-3p in the gastric mucosa, be that infected, inflamed, atrophic or with gastric cancer. However, the results of the in silico analysis suggested a high probability that these miRNAs regulate the expression of GKN1, either inhibiting the translation of the transcript or promoting its degradation (Fig. 1E and F).

Currently, >5,000 miRNAs are registered on miRBase, while in silico predictions estimate that more than one-third of the human transcriptome can be regulated by miRNAs. In gastric cancer, inflammatory processes and H. pylori infection, miRNAs fulfil an important function in the deregulation of gene expression (31,33). GKN1 is absent in cancer, both with and without H. pylori, and is reduced in patients with gastritis, in gastric mucosa infected by H. pylori and in atrophic gastritis (30). The reduction in or absence of GKN1 transcripts or protein is not completely explained by the studied mechanisms of transcriptional and post-translational regulation. It is likely that some miRNAs regulate GKN1 expression directly or indirectly at the post-transcriptional level.

It has been reported that ROS deregulate the expression of miRNAs in tissue infected with H. pylori or gastric cancer tissue (107,108). It is also known that ROS induce a decrease in the number of copies of GKN1 mRNA in tissue infected by H. pylori (48). These data support the hypothesis that, in H. pylori infection, ROS alter the expression of miRNAs, among which are those with the GKN1 transcript as a target. The cytotoxins VacA and CagA, lipopolysaccharide and peptidoglycan, among other components of H. pylori, are able to induce the increased expression of miRNAs that inhibit translation or induce the degradation of the GKN1 transcript, thus modulating the decrease in the levels of this protein.

From the first stages of H. pylori infection, the inflammation associated with it causes changes in the expression of proteins and miRNAs, alterations in cell signaling, and unbalanced cell proliferation and apoptosis in gastric epithelial cells, promoting the progression of gastritis to pre-neoplastic and neoplastic lesions (109). The abnormal expression of miRNAs is common in different types of cancer (110), with the evidence indicating changes in the expression profiles of miRNAs in gastric cancer and in mucosa infected by H. pylori.

Alterations in the expression of miRNAs can manifest either as increases or decreases (111). In the gastric mucosa, both with and without H. pylori infection, it has been found that miRNAs with changes in their expression levels in response to H. pylori can be similar to or different from those observed in gastric cancer with a negative result for bacteria (31) (Tables III and IV). Chang et al (33) found that hsa-miR-99b-3p, hsa-miR-564 and hsa-miR-658 expression increased in cancerous tissue infected with H. pylori, while hsa-miR-204-5p, hsa-miR-338-5p, hsa-miR-375 and hsa-miR-548c-3p were found to be overexpressed in cancer tissue without H. pylori (33).

In infected mucosa, hsa-miR-223 expression was found to be increased, while in mucosa without H. pylori, hsa-miR-203, hsa-miR-204, hsa-miR-455, hsa-miR-141 and hsa-let-7f were found to be overexpressed (31). The levels of let-7, miR-125a and miR-500 were found to be significantly reduced in cells infected with cagA⁺ strains, although not in those infected with cagA⁻ strains. These results indicate that miRNAs participate in gastric pathogenesis, whether associated and not associated with H. pylori, and suggest that the CagA oncoprotein of H. pylori regulates the differential expression of miRNAs in epithelial gastric cells (31). Increased hsa-miR-127-5p, hsa-miR195, hsa-miR-196a, hsa-miR-206, hsa-miR-216 and miR-488 expression has been found, while decreased hsa-miR-103, hsa-miR-141, hsa-miR-17-3p, hsa-miR34a and let-7i expression has been found in gastric epithelial cells infected with different H. pylori-cagA⁺ strains (36).

H. pylori is able to modify the expression of miRNAs by means of inflammatory effectors (112). In gastric epithelial cells, the pro-inflammatory cytokines IL-8, tumor necrosis factor-α and IL-1β induce the expression of miR-146a (113), while the oncoprotein CagA positively regulates c-myc, which is related to the decreased expression of miR-26a and miR-101. The decrease in the expression of these miRNAs contributes to increased levels of the histone methyltransferase EZH2 and methyltransferase DNMT3B, which promote the methylation of the let-7 promoter (114).

E3 ubiquitin-protein ligase UBR5 (UBR5) is an E3 ubiquitin ligase that participates in the ubiquitin-proteasome system, regulating protein concentration via ubiquitination and degradation, and is deregulated in different types of cancer (115). UBR5 increases in the cancerous tissues of gastric cancer patients, while an interaction between UBR5 and GKN1 has been observed through immunoprecipitation assays. These results suggest that UBR5 participates in the ubiquitination of GKN1, and that at least part of this protein is sent to be degraded by the proteasome (27) (Fig. 1G). Thus, UBR5 contributes to the regulation of gastric carcinogenesis, inducing the degradation of tumor suppressing proteins, such as GKN1 (27).

Therefore, promoter methylation, trimethylation of histones and ubiquitination are mechanisms that contribute to the regulation of the GKN1 expression in gastric cancer; however, they do not explain the absence of the protein in cell lines and cancerous human tissue.