Gastric cancer (GC) is a common primary tumor of the digestive system. In 2020, >1 million new cases of GC were reported worldwide and it accounted for 769,000 deaths, rendering it the fifth most common cancer and the fourth leading cause of cancer-associated death globally (1). Incidence and mortality of GC have increased with changes in living habits and environmental factors (2). Numerous studies have shown that the development of GC is associated with Helicobacter pylori infection (2), dietary habits (3), genetic factors (4), and the local and regional environment (5), and these factors are associated with each other. The human gastrointestinal microecosystem is one of the most complex microecosystems in the body, and the relative dynamic balance is closely related to health status. When body functions are disturbed by certain factors such as ulcerative colitis, the dynamic balance of microbiota in the body is disrupted, which can lead to the formation of gastrointestinal microecosystem dysfunction between the host and the flora (6). In addition, tumor progression may occur due to the presence of bacteria that have not yet been detected, whereas the gut microbial community may also shape the microbiota for tumor survival such as induce DNA damage, enhance inflammatory response, and affect the tumor microenvironment to promote tumor growth (7). Therefore, investigation of the association between the changes in intestinal flora and the development of GC is important for the early detection, clinical symptomatic treatment and improvement of survival of patients with GC.

Randomized controlled trials (RCTs) are the gold standard for inferring causality in epidemiology; however, given the ethical constraints and moral limitations such as inappropriate use of placebos, there are difficulties in implementing RCTs (8). Mendelian randomization (MR) studies comprise a statistical method that has been primarily applied to infer causality in epidemiological diseases such as Coronavirus Disease 2019(9). Different genotypes represent different intermediate phenotypes; when the phenotype represents an exposure characteristic such as intestinal flora, the association effect between a genotype and a disease can represent the influence of exposure factors on the disease. Since alleles follow the principle of random distribution, traditional epidemiology does not consider confounding and reverse causality (10). With the public release of large-scale gene-wide association data, a large number of reliable genetic variants are available for MR studies (11). Therefore, the present study analyzed the causal association between gut flora and GC to aid the development of novel strategies for the clinical intervention of GC.

In the present study, the MR method was utilized to explore the causal relationship between the relative abundance of gut microbes and GC. Trillions of symbiotic bacteria colonize the gut and serve a key role in body homeostasis and host defense against pathogenic invasion (16). A healthy microbiota resists colonization and invasion by harmful microorganisms through direct and indirect mechanisms (17,18). For example, short-chain fatty acids (SCFAs), a major metabolite produced by microorganisms, induce the production of antimicrobial peptides by inhibiting the activity of histone deacetylase-3, thereby enhancing the antibacterial activity in infected mouse models (19,20).

Multiple studies have shown decreased diversity of intragastric flora in patients with GC (21-23); however, other studies have suggested a quantitative difference in the composition of the flora between patients with GC and those with dyspepsia (24,25). Aviles-Jimenez et al (22) demonstrated a decrease in the abundance of Porphyromonas, Neisseria and Streptococcus buglossi, and an increase in the abundance of Lactobacillus spp. and Trichosporon spp. during the disease progression of GC. Demiryas et al (25) demonstrated that patients with GC with increased homogeneity and diversity of flora compared with healthy controls. A study of 276 patients with GC demonstrated that the abundance of Streptococcus spp., Clostridium spp., Crescentomonas spp., Propionibacterium spp. and Corynebacterium spp. is increased in cancerous tissues (26). A Korean study concluded that Prevotella and Propionibacterium acnes are causative agents of GC, whereas Lactococcus lactis serves a protective role in the development of GC (27). Furthermore, H. pylori infection is a major risk factor for gastric carcinogenesis; however, the majority of infected individuals do not develop GC and significant genomic diversity of strains is associated with virulence factors (28).

Alterations in the gut microbiota may increase the susceptibility to GC through several mechanisms. Gastrointestinal flora produce a number of metabolites and enhance inflammatory responses, antagonize the development of tumors and activate alternative mechanisms via PI3K/AKT, MAPK, JAK-STAT and other signaling pathways, which can lead to disruption of the intestinal flora and further contribute to development of GC (29). The gastric microbiome and metabolome profiles of 37 cases of GC and matched non-tumor tissue were previously characterized by 16S ribosomal RNA technology. The relative abundance of amino acids, carbohydrates, carbohydrate conjugation, glycerophospholipids and nucleosides in GC tissue was revealed to be higher than that in non-tumor tissues (30). Furthermore, the combination of 1-methylnicotinamide and n-acetyl-D-glucosamine 6-phosphate is a reliable biomarker to distinguish GC from normal tissue (31). SCFAs mainly comprise acetic, butyric and propionic acids; these metabolites are important energy sources for gut microbes and epithelial cells, in addition to their different immunomodulatory functions (32). Previous studies have demonstrated that SCFAs serve a tumorigenic role by blocking activation of the NF-κB signaling pathway and inducing the differentiation of regulatory T cells (33,34). Among them, butyrate not only promotes energy metabolism and maintains a low-oxygen environment in the intestinal lumen, but also activates peroxisome proliferator-activated receptor γ in intestinal cells, inhibits expression of the nitric oxide synthase 2 gene and the synthesis of inducible nitric oxide synthase, decreases nitrate production and restricts proliferation of pathogenic anaerobic bacteria, thus inhibiting gastrointestinal tract inflammation and carcinogenesis (35). Bile reflux-generated bile acids are high-risk factors for GC and secondary bile acids promote GC cell proliferation (36). Therefore, the incidence of GC may be decreased by regulating specific types of bile acid.

To the best of our knowledge, the present study is the first to identify a causal association between gut microorganisms and GC, in which elevated abundance of the genera Prevotella 7, Roseburia and Ruminococcaceae UCG014 may increase the risk of GC. In addition, the family Enterobacteriaceae, the genera Allisonella, Lachnospiraceae FCS020, Ruminococcaceae UCG004 and Ruminococcaceae UCG009, and the order Enterobacteriales decreased the risk of GC. Ruminococcus is one of the earliest discovered gastric bacteria and serves a crucial role in metabolism. Cellulose is broken down by rumen bacteria to obtain nutrients. Ruminococcus is also capable of fermenting glucose and xylose. In addition to this function, it is able to stabilize the intestinal barrier, prevent diarrhea, reduce the risk of colorectal cancer, reduce kidney stone formation and increase energy (37). Ruminalococcus spp. has decreased abundance in ulcerative colitis, allergic disease and cerebral palsy, indicating its function as a beneficial bacterium (38). Notably, the results of the present reverse MR study did not support a causal association between GC and altered intestinal flora.

The causal relationship identified in the present study may provide candidate gut microbiota for subsequent functional studies. However, there are limitations. First, the threshold for screening the gut microbiome IVs was P<1x10^-5 and although measures were taken to ensure validity by calculating the F-statistic for each SNP, there is the possibility of false-negative errors due to insufficient statistical validity. Second, while the majority of patients in the GWAS pooled data were European, only a small number of gut microbiome data came from other ethnicities, which could lead to biased estimates and could affect generalizability. Third, due to the strict threshold, a number of genetic locus of the gut microbiota were excluded at the IV selection stage, which may have led to some results being missed.

In conclusion, the causal association between intestinal microorganisms and GC was investigated in the present study using MR analysis. The genera Prevotella 7, Roseburia and Ruminococcaceae UCG014 were associated with increased risk of GC, whereas the family Enterobacteriaceae, the genera Allisonella, Lachnospiraceae FCS020, Ruminococcaceae UCG004 and Ruminococcaceae UCG009, and the order Enterobacteriales reduced the risk of GC development, suggesting that intestinal microorganisms serve a role in the process of GC development and may have potential for the treatment of GC.