
==== Front
Discov Oncol
Discov Oncol
Discover Oncology
2730-6011
Springer US New York

1285
10.1007/s12672-024-01285-6
Analysis
Impact the impact of gut microbiota on gastric cancer via immune cells: a comprehensive Mendelian randomization study and mediation analysis
Wang Chao 12
Wang Jia 2
Guan Wenxian wenxian_guan820@sina.com

3
Fei Bojian jdfyfbj1971@163.com

12
1 https://ror.org/02ar02c28 grid.459328.1 0000 0004 1758 9149 Department of Gastrointestinal Surgery, Affiliated Hospital of Jiangnan University, Wuxi, 214062 Jiangsu China
2 https://ror.org/04mkzax54 grid.258151.a 0000 0001 0708 1323 Wuxi School of Medicine, Jiangnan University, Wuxi, 214122 Jiangsu China
3 grid.41156.37 0000 0001 2314 964X Division of Gastric Surgery, Department of General Surgery, Nanjing Drum Tower Hospital, Affiliated Hospital of Medical School, Nanjing University, Nanjing, 210008 China
31 8 2024
31 8 2024
12 2024
15 38911 5 2024
27 8 2024
© The Author(s) 2024
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Purpose

Recent observational studies have highlighted the role of altered gut microbiota (GM) in the activation of the host immune system and the resulting of gastric cancer (GC). However, the exact causal relationship and mechanisms of action are still not fully understood.

Materials and methods

Genetic data from published genome-wide association studies (GWASs) were employed to determine the causal effects of 207 taxa and 205 bacterial pathways on GC via two-sample Mendelian randomization (MR) and two-step mediation MR analysis. In this study, 731 immune cell traits served as potential mediators. An inverse variance-weighted (IVW) estimation, augmented by a range of alternative estimators, notably the Bayesian-weighted MR method, was employed as the primary methodological approach.

Results

Four taxa and five bacterial pathways were found to be negatively correlated with GC, whereas one taxon and two bacterial pathways were a positively correlated with GC. Reverse causality was not found in the reverse MR analysis. Additional validation was performed using a sensitivity analysis. Mediation MR analyses revealed that the GM influences GC through various phenotypes of 16 immune cells that act as mediators. For example, s_Alistipes_sp_AP11 was found to inhibit GC through NKT %T cell (total effect: -0.3234, mediation effect: 0.0212). This mediating effect further highlights the complex relationship among GMs, immune cell traits, and their combined effects on GC.

Conclusions

Our findings highlight the genetic connection between specific GMs and GC, emphasizing the potential role of immune cells as mediators, and offering valuable perspectives on potential therapeutic strategies that manipulating the GM to address GC.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-024-01285-6.

Keywords

Gut microbiota
Gastric cancer
Immune cells
Mendelian randomization study
Single nucleotide polymorphisms
Wuxi Health Commission Youth ProjectQ202242 Wang Chao top Talent Support Program for young and middle-aged people of Wuxi Health CommitteeHB2023046 Wang Chao National Natural Science Foundation of China82303117 Wang Chao issue-copyright-statement© Springer Science+Business Media, LLC 2024
==== Body
pmcIntroduction

Gastric cancer (GC) is a major healthcare issue worldwide, with approximately 1,089,103 new cases and 768,793 deaths recorded in 2020. As a result, GC is the fifth most common cancer and the fourth highest cause of cancer-related deaths globally [1], Early detection of GC is challenging as reliable biomarkers for accurate diagnosis and prognosis monitoring during the adjuvant stage are unavailable. As a result, GC is associated with a high mortality rate, with only 20% of patients surviving beyond five years after diagnosis [2, 3]. GC is influenced by genetics, environmental factors, and microbial factors, such as Helicobacter pylori infection [4, 5]. Various studies have indicated variations in stomach microbiota among individuals at various phases of gastric precancerous and malignant lesion progression, specifically a decline in total microbial variety and an increase in gut-friendly bacteria, particularly those with nitrosative capabilities [5]. The connection between the gut microbiota (GM) and GC may be due to the ongoing activation of the host's immune system by the GM, causing a breakdown in communication between host epithelial cells and microbes, ultimately leading to a condition of persistent inflammation [6]. However, the identification of non-H. pylori drivers associated with GC have not been identified. Therefore, more studies are needed to elucidate the exact mechanisms by which the GM affects the immune response during GC development and progression. Moreover, precise and easily accessible biomarkers are needed for the early detection of GC.

The human microbiome comprises a variety of microorganisms, such as viruses, fungi, and bacteria.

And Firmicutes, Bacteroidetes, Actinobacteria and Proteobacteria constitute the most abundant phyla in the human intestinal microbiota [7]. Dysbiosis, characterized by changes in the microbiota's composition and function of the microbiota, can be influenced by factors, such as antibiotic use, external microbial infections, and host genetics [8]. A well-regulated and consistent microbiome is essential to prevent cancer progression, whereas an imbalanced microbiome provides minimal defense and could potentially stimulates carcinogenesis. The gut microbiome is intricately linked to various mechanisms of carcinogenesis, such as inflammation, which facilitates tumor proliferation, modifications in the immune response, and generation of pro-carcinogenic metabolites [9]. Recent developments in metagenomics and high-throughput sequencing have revealed that the stomach, once thought to be devoid of microorganisms, harbors acid-resistant bacteria, including H. pylori [10]. The microbiota of the digestive system, such as Lactobacillus, Streptococcus, Veillonella, Prevosia, Fusobacterium, Lachnospiraceae, Leptotrichia, and Clostridium, are critical for the development of GC [5, 11]. Various immune-mediated inflammatory conditions have been found to alter GM composition and its metabolites. GM-derived metabolites have a wide range effects on various immune cell reactions, including T cells, B cells, dendritic cells, and macrophages [12]. Dysbiosis of the GM may lead to an atypical immune response, which can lead to the onset of GC.

Mendelian randomization (MR) analysis is a unique approach to uncovering causal relationships in observational data using genetic variants as instrumental variables (IVs) [13]. This approach is especially effective at elucidating the impact of the GM on disease pathogenesis via reducing confounding and reverse causation, which are inherent challenges in traditional observational studies. The immune system, with its various of cell types and signaling pathways, plays a crucial role in the development of GC. By analyzing 731 immune cell characteristics [14], specific immune pathways or cells that act as mediators connecting the gut microbiome to the onset or progression of GC can be identified.

This study aimed to explore the association between the GM, immune cells, and GC development. These findings improve our understanding of how the GM influences GC and shed light on the role of immune factors in disease progression.

Materials and methods

Study design

This study comprised two stages of analyses (Fig. 1). In phase 1, a two-sample MR method was used to evaluate the links between 412 GM (207 taxa and 205 bacterial pathways) and GC using single nucleotide polymorphisms (SNPs) as instrumental variables (IVs) for each factor. In phase 2, mediators from 731 immune cell signatures were selected to explore the role of immune characteristics in connecting GM and GC. The mediating effects of these immune traits were calculated using a two-step MR mediation approach. For causal inference to be valid, IVs must meet three key assumptions [15]: (1) high correlation with the exposure variables; (2) no connection with confounding variables; and (3) no direct effect on the outcome, affecting the outcome solely through the exposure.Fig.1 Assumptions and design of the relationship between gut microbiota and gastric cancer with immune cells as mediators in the bidirectional and mediation Mendelian randomization (MR) analyses. a A two-sample bidirectional Mendelian randomization (MR) approach was employed to examine the causal associations between gut microbiota (exposure) and gastric cancer (GC, outcome). A total of 731 immune cell traits were identified as potential mediators for further mediation analyses. Finally, a two-step MR analysis was conducted to determine the potential mediating role of immune cells. In step 1, the impact of gut microbiota on immune cells was determined while in step 2, the influence of immune cells on GC was examined. b Mendelian randomization analysis flow chart

Sources of information on exposure, mediators, and outcome

The MR study used summary-level data from GWASs primarily conducted with individuals of European ancestry to gather information on exposure (207 species and 205 bacterial pathways), mediators (731 immune cell traits), and outcome (GC).

The complete GWAS information for 207 taxa and 205 bacterial pathways in 7,738 participants from the DMP cohort was downloaded directly from the NHGRI-EBI GWAS Catalog (https://www.ebi.ac.uk/gwas/) using the study access codes GCST90027446 to GCST90027857 (specific access codes for species and pathways are listed in Supplementary Table 1) or at https://dutchmicrobiomeproject.molgeniscloud.org [16].

GWAS data on immunity were used to examine the potential mediating functions of immune cells. Summary statistics for different immune characteristics were acquired from the GWAS Catalog, covering the accession numbers, GCST0001391 to GCST0002121 [14]. A total of 731 immunophenotypes were examined, including B cells, CDCs, mature T cells, monocytes, myeloid cells, TBNK (T, B cells, and natural killer cells), and Treg panels.

GC data were acquired from the https://storage.googleapis.com/finngen-public-data-r9/summary_stats/finngen_R9_C3_STOMACH_EXALLC.gz (737 cases and 287,137 controls).

All GWASs included in this study were authorized by the relevant institutional review boards and informed consent was obtained from all participants. Ethical approval was not required as summary-level data were employed in this study.

SNP selection

IVs were selected as SNPs linked to exposures showing a genome-wide significant association (p < 5 × 10−8) with the traits of interest. Subsequently, SNPs showing linkage disequilibrium (LD) were excluded from the study. To meet these requirements, the LD between the selected SNPs must have an r2 value of less than 0.001 and be at least 10,000 kb apart [17, 18]. The explained variance (R2) and F-statistic parameters were computed to assess the strength of the association between the identified IVs and exposure. Generally, SNPs with F-statistic values less than 10 are considered to be ineffective instruments. The genetic instruments for exposures were validated by calculating the F-statistic using the formula: F = R2 × [(N –1 − K)/K] × (1 − R2)[19–21], where R2 represents the cumulative variance accounted for by the selected SNPs, N denotes the size of the sample sizes, and K denotes the number of SNPs examined.

MR analysis

Two-sample MR studies were conducted to explore the possible causal link between the makeup of GM and GC makeup. The inverse-variance-weighted (IVW) [22] and Bayesian weighted MR (BWMR) methods [23] were used to calculate impacts; beta (β) values with standard errors were reported for continuous results and odds ratios (OR) with 95% confidence intervals (CI) were reported for binary results. P < 0.05 was considered to indicate significance.

Mediation MR analysis

Two-step MR analysis was conducted to examine the involvement of immune cells as mediators. Initially, the causal associations between the GM and immune cells were individually examined using MR. Subsequently, the causal link between immune cells and GC was further investigated. To determine the role of immune cells in the connection between GM and GC, the total effect was represented as β1. The direct effect between GM and immune cells was designated as β2, and the direct effect between immune cells and GC was designated as β3. The product of β2 and β3 represented mediating effects.

Sensitivity analysis

Several MR methods, such as MR-Egger, weighted median, IVW, simple mode, weighted mode, BWMR, and MR-PRESSO approaches, were employed. However, the IVW method, which is a frequently used approach in MR research [24, 25], served as the main examination approach. Cochran's Q test was used to evaluate the heterogeneity of each SNP, and scatter plots were created to demonstrate the connections between SNPs, exposures, and outcomes in the MR analysis [25]. A leave-one-out analysis was conducted to assess the impact of each SNP on the results. Thereafter, MR-PRESSO [26] and MR-Egger regression [27] methods were applied to examine potential horizontal pleiotropy effects. MR-PRESSO was used to identify and address significant outliers, thereby correcting for horizontal pleiotropy. Statistical analyses were conducted using R software version 4.2.1 and the R-based package “TwoSampleMR” for MR analysis.

Results

Two sample MR analysis between GM and GC

To explore the causal effect of GM on GC, a two-sample MR analysis was performed using IVW and BWMR methods as the main analytical strategies. Supplementary Table 2 contains the precise information on the 4048 SNPs linked to 207 taxa and 205 bacterial pathway traits. As depicted in Fig. 2A, the IVW model identified 25 GMs associated with GC. The BWMR approach tackles the uncertainty surrounding the estimated minor effects and weak horizontal pleiotropic effects, and identifies outliers resulting from significant horizontal pleiotropic effects [23]. Therefore, the IVW results were validated using the BWMR method, which led to the identification of 18 GC-related GMs (Fig. 2B). Finally, a reverse MR study was performed using 18 sets of GM data that were statistically significant to mitigate the potential influence of reverse causation in the context of GC. Supplementary Table 3 shows the SNPs associated with GC. The results of the reverse MR analysis did not indicate of reverse causation (P > 0.05), as shown in Supplementary Table 4.Fig.2 MR analyses highlight the causal effects between gut microbiota and gastric cancer. a Forest plot of the causal effect of gut microbiota on GC via the IVW method. b Forest plot of the causal effect of gut microbiota on GC via the BWMR method. OR odds ratio, CI confidence interval

As displayed in Fig. 2, MR analysis using the IVW and BWMR methods indicated a genetic prediction of one taxon and three bacterial pathways linked to a higher risk of GC. The presence of g_Haemophilus (IVW OR = 1.4965, 95%CI = 1.0640–2.10480, P = 0.0205; BWMR OR = 1.5121, 95%CI = 1.0537–2.1698, P = 0.0248), “PWY.5659 GDP mannose biosynthesis” (IVW OR = 2.1118, 95%CI = 1.2293–3.6278, P = 0.0068; BWMR OR = 2.1384, 95%CI = 1.2067–3.7897, P = 0.0092), “PWY.724 superpathway of L lysine, L threonine and L methionine biosynthesis II” (IVW OR = 2.3932, 95%CI = 1.5945–1.0623, P = 0.0243; BWMR OR = 1.5768, 95%CI = 1.0472–2.3742, P = 0.0292), and “TRNA CHARGING PWY tRNA charging” (IVW OR = 1.5515, 95%CI = 1.0560–2.2796, P = 0.0253; BWMR OR = 1.5478, 95%CI = 1.0541–2.2729, P = 0.0258) were found to significantly elevate the risk of GC.

Eight taxa and six bacterial pathways were associated with a lower likelihood of developing GC. g_Oxalobacter (IVW OR = 0.7038, 95%CI = 0.5345–0.9269, P = 0.0124; BWMR OR = 0.6964, 95%CI = 0.5204–0.9320, P = 0.0150), s_Alistipes_sp_AP11 (IVW OR = 0.7237, 95%CI = 0.5560–0.9420, P = 0.01620; BWMR OR = 0.7177, 95%CI = 0.5525–0.9322, P = 0.0129), f_Oxalobacteraceae (IVW OR = 0.7040, 95%CI = 0.5255–0.9430, P = 0.0186; BWMR OR = 0.6955, 95%CI = 0.5010–0.9485, P = 0.0218), s_Oxalobacter_formigenes (IVW OR = 0.7039, 95%CI = 0.5253–0.9433, P = 0.0187; BWMR OR = 0.6954, 95%CI = 0.5097–0.948, P = 0.0219), o_Burkholderiales (IVW OR = 0.6640, 95%CI = 0.4615–0.9553, P = 0.0273; BWMR OR = 0.6621, 95%CI = 0.4535–0.9666, P = 0.0327), f_Bacteroidaceae (IVW OR = 0.6483, 95%CI = 0.4380–0.9595, P = 0.0303; BWMR OR = 0.6436, 95%CI = 0.4260–0.9724, P = 0.0364), s_Coprococcus_catus (IVW OR = 0.5983, 95%CI = 0.3690–0.9701, P = 0.0372; BWMR OR = 0.5933, 95%CI = 0.3550–0.9914, P = 0.0463), s_Escherichia_coli (IVW OR = 0.7489, 95%CI = 0.5653–0.9922, P = 0.0439; BWMR OR = 0.7393, 95%CI = 0.5501–0.9937, P = 0.0453), “PWY.6147:6-hydroxymethyl dihydropterin diphosphate biosynthesis I” (IVW OR = 0.5590, 95%CI = 0.3923–0.7966, P = 0.0013; BWMR OR = 0.5719, 95%CI = 0.3973–0.8232, P = 0.0026), “POLYAMSYN.PWY: superpathway of polyamine biosynthesis I” (IVW OR = 0.5003, 95%CI = 0.3118–0.8028, P = 0.0041; BWMR OR = 0.4879, 95%CI = 0.29293 ~ 0.8128, P = 0.0059), “PWY_ARG.POLYAMINE.SYN:superpathway of arginine and polyamine biosynthesis” (IVW OR = 0.4894, 95%CI = 0.2950–0.8120, P = 0.0057; BWMR OR = 0.5000, 95%CI = 0.3025–0.8249, P = 0.0067), “PWY.6897:thiamin salvage II” (IVW OR = 0.5859, 95%CI = 0.3815–0.8996, P = 0.0146; BWMR OR = 0.5770, 95%CI = 0.3661–0.9092, P = 0.0178), “PWY.6284:superpathway of unsaturated fatty acids biosynthesis(E.coli)” (IVW OR = 0.6316, 95%CI = 0.4256–0.9373, P = 0.0225; BWMR OR = 0.6111, 95%CI = 0.4043–0.9237, P = 0.0195), and “ARO.PWY:chorismate biosynthesis I” (IVW OR = 0.6102, 95%CI = 0.3982–0.9352, P = 0.0234; BWMR OR = 0.5961, 95%CI = 0.3757–0.9459, P = 0.0281) were found to significantly decrease in the risk of GC (Fig. 2).

An additional investigation into the causal connection between the 18 GMs and GC is outlined in Fig. 3A–R. Sensitivity analysis provided detailed information confirming the strength of the observed causal relationships. The MR-Egger regression intercept method showed no bias from genetic pleiotropy in the results (Supplementary Table 5), and MR-PRESSO analysis confirmed the lack of horizontal pleiotropy in the MR study (P > 0.05, Supplementary Table 6). The Cochran's Q tests revealed no significant heterogeneity based on the results (P > 0.05; Supplementary Table 7). Furthermore, scatter plots (Fig. S1), forest plots (Fig. S2), and funnel plots (Fig. S3) further supported the stability of the results, and analyses using the “leave-one-out” method indicated that none of the individual variables had a significant impact on the causal inferences (Fig. S4).Fig.3 Forest plots display the causal associations between gut microbiota and GC using different methods. a Forest plot of the causal effect of the “6-hydroxymethyl dihydropterin diphosphate biosynthesis I” bacterial pathway on GC. b Forest plot of the causal effect of the “superpathway of polyamine biosynthesis I” bacterial pathway on GC. c Forest plot of the causal effect of the “superpathway of arginine and polyamine biosynthesis” bacterial pathway on GC. d Forest plot of the causal effect of the “GDP mannose biosynthesis” bacterial pathway on GC. e Forest plot of the causal effect of the g_Oxalobacter on GC. f Forest plot of the causal effect of the “thiamin salvage II” bacterial pathway on GC. g Forest plot of the causal effect of the s_Alistipes_sp_AP11 on GC. h Forest plot of the causal effect of the f_Oxalobacteraceae on GC. i Forest plot of the causal effect of the s_Oxalobacter_formigenes on GC. j Forest plot of the causal effect of the g_Haemophilus on GC. k Forest plot of the causal effect of the “superpathway of unsaturated fatty acids biosynthesis (E. coli)” bacterial pathway on GC. l Forest plot of the causal effect of the “chorismate biosynthesis I” bacterial pathway on GC. m Forest plot of the causal effect of the “superpathway of L lysine, L threonine and L methionine biosynthesis II” bacterial pathway on GC. n Forest plot of the causal effect of the “TRNA.CHARGING.PWY:tRNA.charging” bacterial pathway on GC. o Forest plot of the causal effect of the o_Burkholderiales on GC. p Forest plot of the causal effect of the f_Bacteroidaceae on GC. q Forest plot of the causal effect of the s_Coprococcus_catus on GC. (r) Forest plot of the causal effect of the s_Escherichia_coli on GC. IVW inverse variance weighting, OR odds ratio, CI confidence interval

Effect of 18 GMs on 731 immune cell traits

The crucial role of the immune system in GC is widely recognized. Thus, we proceeded to clarify how the GM affects immune cell characteristics and their potential influence on the risk of GC. Mediation MR analyses were performed with 731 immune cell characteristics as mediators to explore the relationships between the GM and GC. Significant relationships were found between the 18 bacterial traits and various immune cell traits. A total of 467 immune cell characteristics from a pool of 731 potential mediators met the screening criteria (P_IVW < 0.05) and were used in the mediation MR analyses (Supplementary Table 8).

Effect of each immune cell trait on GC

These results underscore the complex interplay between immune cell characteristics and the GM, laying the groundwork for further mediation analyses. Subsequent investigations revealed the potential mediating effects of exposure to these noteworthy mediators (467 immune cell traits from 18 gut microbiota) in GC. Based on IVW (Fig.S5), 12 GMs affected GC through 16 immune cell traits.

Following the elucidation of how immune cell features affect GC, we examined the direct influence of 12 GM on these crucial mediators (16 immune cell traits). Our analysis revealed several important discoveries (Fig. S6). The causal association between 12 GMs and 16 immune cell traits was investigated using a two-sample MR analysis. The MR-Egger regression intercept method revealed that genetic pleiotropy did not affect these findings (Supplementary Table 9). Moreover, Cochran’s Q tests indicated no significant heterogeneity (P > 0.05, Supplementary Table 10), and MR-PRESSO analysis confirmed the absence of horizontal pleiotropy in the MR study (P > 0.05 Supplementary Table 11). The scatter plots in Fig. S7, forest plots in Fig. S8, funnel plots in Fig. S9, and "leave-one-out" analyses (Fig. S10) highlighted the consistency of the results.

MR analysis was conducted to determine the protective effects of the eight immunophenotypes on GC using the IVW method. These immunophenotypes include (6-hydroxymethyl dihydropterin diphosphate biosynthesis I) CD39 + CD4 + %CD4 + , (f_Oxalobacteraceae) CD4 Treg %T cell, (f_Oxalobacteraceae) CD8br and CD8dim %leukocyte, (s_Oxalobacter_formigenes) CD4 Treg %T cell, (GCST90027810) ebi-a-GCST90001614, (g_Haemophilus) HVEM on naive CD4 + , (chorismate biosynthesis I) CD28- CD8dim %CD8dim, and (tRNA.charging) CD39 + CD4 + %CD4 + . Comparable results were observed when four additional approaches were employed: MR eager, weighted median, simple mode, and weighted mode (Fig. S11).

The IVW approach was employed to examine the enhancing effect of four immunophenotypes on GC, specifically (superpathway of polyamine biosynthesis I) CD8br %T cell, (superpathway of arginine and polyamine biosynthesis) SSC-A on NKT, (thiamin salvage II) CD8br %T cell, (s_Alistipes_sp_AP11) NKT %T cell, (g_Haemophilus) CCR2 on CD14- CD16 + monocyte, (chorismate biosynthesis I) CD28- DN (CD4-CD8-) %T cell, (superpathway of L lysine, L threonine and L methionine biosynthesis II) HLA DR on CD33br HLA DR + CD14dim, and (s_Coprococcus_catus) HLA DR on CD33br HLA DR + CD14dim. Four additional techniques, namely MR eager, weighted median, simple mode, and weighted mode, had similar outcomes (Fig. S11) to the above approaches.

The MR-Egger intercept (Supplementary Table 12) and MR-PRESSO analyses indicated no horizontal pleiotropy in the studied associations (Supplementary Table 13). Cochran's Q tests did not show any significant heterogeneity based on the results shown in Supplementary Table 14 (P > 0.05). The reliability of the findings was reinforced by scatter diagrams (Fig. S12), forest charts (Fig. S13), funnel charts (Fig. S14), and “leave-one-out” examinations (Fig. S15).

Mediation MR analysis

A two-step MR study revealed a causal relationship between the GM and GC, mediated by immune cell traits. The key factors affecting GC were determined, and the resulting impact of exposure on mediation was measured by calculating the mediation effect (indirect effect, Table 1).Table 1 Mediation Mendelian randomization analyses of the causal effects among gut microbiota, immune cells and GC

Exposure	Mediator	Total effect	β1	β2	Indirect effect(β1*β2)	
6-hydroxymethyl dihydropterin diphosphate biosynthesis I	CD39 + CD4 + %CD4 + 	− 0.5815	− 0.1877	− 0.1051	0.0197	
superpathway of polyamine biosynthesis I	CD8br %T cell	− 0.6925	0.3163	0.1323	0.0418	
superpathway of arginine and polyamine biosynthesis	SSC-A on NKT	− 0.7145	− 0.3302	0.0783	− 0.0258	
thiamin salvage II	CD8br %T cell	− 0.5347	0.2454	0.1323	0.0325	
s_Alistipes_sp_AP11	NKT %T cell	− 0.3234	0.1378	0.1538	0.0212	
f_Oxalobacteraceae	CD4 Treg %T cell	− 0.3511	− 0.1684	− 0.1782	0.0300	
f_Oxalobacteraceae	CD8br and CD8dim %leukocyte	− 0.3511	0.1353	− 0.3758	− 0.0508	
s_Oxalobacter_formigenes	CD4 Treg %T cell	− 0.3511	− 0.1685	− 0.1782	0.0300	
s_Oxalobacter_formigenes	CD8br and CD8dim %leukocyte	− 0.3511	0.1354	− 0.3758	− 0.0509	
g_Haemophilus	HVEM on naive CD4 + 	0.4031	0.3610	− 0.1297	− 0.0468	
g_Haemophilus	CCR2 on CD14− CD16 + monocyte	0.4031	0.1763	0.0908	0.0160	
ARO.PWY:chorismate biosynthesis I	CD28− CD8dim %CD8dim	− 0.4940	0.2558	− 0.0621	− 0.0159	
ARO.PWY:chorismate biosynthesis I	CD28− DN (CD4− CD8−) %T cell	− 0.4940	0.2642	0.1580	0.0418	
superpathway of L lysine, L threonine and L methionine biosynthesis II	HLA DR on CD33br HLA DR + CD14dim	0.4665	− 0.3427	0.1193	− 0.0409	
tRNA.charging	CD39 + CD4 + %CD4 + 	0.4392	0.2298	− 0.1051	− 0.0241	
s_Coprococcus_catus	HLA DR on CD33br HLA DR + CD14dim	− 0.5137	0.2997	0.1193	0.0357	

Some data from Table 1 were briefly explained, particularly “PWY.6147:6-hydroxymethyl dihydropterin diphosphate biosynthesis I” hinders the advancement of GC by interacting with “CD39 + CD4 + %CD4 + ” cells, resulting in a total effect size of -0.5815, of which the mediated effect size of “CD39 + CD4 + %CD4 + ” was 0.0197. Moreover, “POLYAMSYN.PWY: superpathway of polyamine biosynthesis I” was demonstrated to impede the advancement of GC through mediation by “CD8br %T cell”, with a total effect size of -0.6925; the mediated effect size of “CD8br %T cell” was 0.0418. The promotion of GC progression was facilitated by the interaction of “PWY.724: superpathway of L lysine, L threonine and L methionine biosynthesis II” and “HLA DR on CD33br HLA DR + CD14dim”, resulting in a total effect size of 0.4665, with “HLA DR on CD33br HLA DR + CD14dim” cells contributing a mediated effect size of -0.0409. These findings emphasize the intricate relationship between distinct GM exposures, their mediators, and their collective influence on GC.

Discussion

The relationship between the GM and immune-related illnesses is a growing area of research. Imbalances in the GM can lead to immune system disorders; however, their impact on GC is complex and not fully understood. To our knowledge, this MR analysis is the first to elucidate the possible causal connection between immune cells in the GM and GC. We identified 12 GMs linked to GC with fluctuations in their abundance affecting immune cells and potentially increasing the risk of GC. We also identified 16 immune cells associated with these bacteria and GC, providing insights into the complex relationship between the GM and GC.

The GM coexists with the human host and has many advantages, such as regulation of immune homeostasis. An imbalance in the composition of the GM can lead to increased susceptibility to colonization of potentially harmful microorganisms and hinder the synthesis of key microbiota-derived metabolites necessary for immune cell maturation and maintenance. GM dysbiosis is associated with various diseases, including tumorigenesis and cancer progression. The treatment of many cancers directly affects gut bacteria. In fact, prophylactic antibiotics are needed to reduce the risk of infection. During radiotherapy and chemotherapy, gastrointestinal toxicity can occur in patients with cancer. Thus, treatment can subsequently alter the composition of the microbiota and may lead to persistent dysbiosis in cancer survivors. Exploring strategies to alleviate dysbiosis and restore microbiota balance plays a role in the management of inflammation-related neoplastic diseases. In the present study, GC was found to be positively correlated with g_Haemophilus and two bacterial pathways (Table 1). Such finding suggests that an increased presence of these specific types of taxonomic flora or alterations in these bacterial processes may lead to a higher likelihood of GC. Owing to its ability to reduce nitrates, the Haemophilus may play a role in the ongoing inflammatory process [28]. A negative relationship has also been found between the levels of Haemophilus and IL-1B mRNA [29]. Previous studies revealed a decrease in Haemophilus in the oral microbiome of individuals with GC and a reduction in the presence of Haemophilus parainfluenzae in GC [30]. Nevertheless, another study found indicated an increase in the prevalence of Haemophilus in GC [31], potentially linked to its connection with functional dyspepsia in the control group rather than intestinal metaplasia in the control group [29]. These results indicate that Haemophilus has different effects on GC at different stages. The present study shows that Haemophilus promotes effect on GC. Higher levels of tRNA and amino acid pairing typically lead to increased tumor formation [32]; however, specific metabolic processes can also contribute to the development of GC. The metabolic process of the “superpathway involving L lysine, L threonine, and L methionine biosynthesis II” remains largely unexplored in GC. Based on our findings, this metabolic pathway plays a role in promoting GC. Thus, how this pathway affects the GC progression through immune cells is further discussed.

The bacteria s_Alistipes_sp_AP11, f_Oxalobacteraceae, s_Oxalobacter_formigenes, and s_Coprococcus_catus, and the five bacterial pathways had an inverse relationship with GC. This finding suggests that these groups of organisms or bacterial processes could potentially provide defense mechanisms against the illnesses. Alistipes, a recently discovered group of bacteria, has mainly been found mainly in medical samples, although not as frequently as other bacteria in the Bacteroidetes phyla, which play significant roles in dysbiosis and disease. Alistipes can offer protection against certain illnesses, such as via cancer immunotherapy [33]. Similarly, we found that s_Alistipes_sp_AP11 exerted a protective effect against occurrence of GC. Regarding f_Oxalobacteraceae, prior studies showed that following the successful elimination of H._pylori in patients with early-stage GC, the average of unidentified Oxalobacteraceae in the range of stomach mucosa-related microbiota was reduced in biopsy samples compared to that in patients without early-stage GC [34]. Reducing the levels of this bacterium may lower the risk of GC; however, further studies are needed to fully understand this connection after the successful elimination of H. pylori. S_Oxalobacter_formigenes, an exclusively anaerobic bacterium that depends entirely on oxalate for its development, plays a crucial role in the breakdown of oxalate in the digestive system of mammals [35]. Therefore, we speculate that reducing the abundance of this bacterium can regulate immune cells and thus reduce the risk of GC, which may be related to oxalate metabolism. Fermentation by various intestinal bacteria can result in the production of lactic acid, which can disturb the gut microbiome by reducing the pH of the lumen, leading to various negative health effects associated with its buildup. Coprococcus catus is a crucial species in the human gut and is essential for lactic acid metabolism. C. catus is important for the growth of lactic acid, sugar, or a combination of both, and influences how other intestinal bacterial species use lactic acid [36]. This role may explain the increased abundance of C. catus, which reduces the occurrence of GC.

The microbiota in the digestive system is important in the development of GC and may have an impact on the response to immunotherapy, enhancing the immune response to tumors in a various ways, such as by activating T cells in response to bacterial antigens that may also target tumor antigens and producing metabolites that have systemic effects on the host [37]. The current findings suggest that the gut microbiome increases the likelihood of GC by influencing 12 types of immune cells (Table 1). Furthermore, four GMs were found to decrease the likelihood of GC by controlling four immune cell characteristics (Table 1). CD39, also called ectonucleoside triphosphate diphosphohydrolase 1 (ENTPD1), is crucial for producing immunosuppressive adenosine and plays a significant role in cancer advancement [38]. In fact, high levels of CD39 in GC patients with GC post-surgery are linked to a negative outcome [39]. Our findings indicate that the “PWY.6147:6-hydroxymethyl dihydropterin diphosphate biosynthesis I” and “TRNA.CHARGING.PWY:tRNA.charging” are involved in promoting and inhibiting the risk of GC through “CD39 + CD4 + %CD4 + ” cell, respectively. Currently, how these bacterial pathways interact with the “CD39 + CD4 + %CD4 + ” cells remain unknown.

CD8 + T cells are the main cells in tumors that fight cancer; however, studies on their relationship with GC prognosis have yielded conflicting results, and their role in outcomes is debated [40]. Spermidine can boost fatty acid oxidation by activating mitochondrial trifunctional proteins, leading to increased mitochondrial activity and cytotoxic function in CD8 + T cells, ultimately improving anti-tumor immunity [41]. The breakdown of glutamine is a key feature of T cell activation and changes in metabolism. Studies using isotopic tracers on CD8 + T cells activated by antigens have indicated that glutamine is a primary source of carbon for the production of polyamines, such as putrescine, spermidine, and spermine. These polyamines are crucial for the proliferation of T cells triggered by activation and production of hypusine. Hypusine is a product of spermidine and is attached to the translational elongation factor, eukaryotic initiation factor 5A (eIF5A) [42]. Our findings indicate that “POLYAMSYN.PWY:superpathway of polyamine biosynthesis I” and “PWY.6897:thiamin salvage I” are involved in mediating “CD8br %T” cell activity. However, the relationship between “PWY.6897: thiamin salvage II” and “CD8 + T” cells must to be further experimentally verified.

NK cells, also known as crucial lymphocytes in the innate immune system, are essential for inhibiting the onset, progression, and spread of GC. According to various clinical studies, enhancing NK cell quantity or NK cell anti-tumor function is a promising approach for patients with GC [43]. Arginase 2 (ARG2)-induced arginine depletion creates an immunosuppressive environment for traditional T cells. The “PWY_ARG.POLYAMINE.SYN: superpathway of arginine and polyamine biosynthesis”, and s_Alistipes_sp_AP11 were discovered to reduce the likelihood of GC by affecting “SSC-A on NKT” cells. Alistipes, a genus in the phylum Bacteroidetes, is a symbiotic bacterium in the intestine that is not motile, produces indole, does not reduce nitrate, and does not hydrolyze arginine or urea. The bacterium has been demonstrated to play a positive role in cancer treatment by influencing the environment around the tumor, primarily by stimulating the production of tumor necrosis factor (TNF) by immune cells near the tumor and using a combination of intratumoral CpG-oligodeoxynucleotides (ODN) to activate TLR9 and inhibitory IL-10R antibodies, ultimately leading to tumor eradication [33]. Therefore, a possible connection exists between this taxon and GC involving the NKT cells.

Recent studies have shown that the recruitment of regulatory T cells (Tregs) to tumors is a strategy for immune evasion. Elevated levels of Tregs have been observed in the tumor mucosa of various types of cancers, such as GC, and these Tregs have been linked to unfavorable outcomes in GC [44]. The f_Oxalobacteraceae and s_Oxalobacter_formigenes are modulated via “CD4 Treg %T” cells, which is essential for immune system evasion. Therefore, the oxalate metabolism pathway may play a role in Tregs; however, further experiments are needed to verify this notion. Interestingly, f_Oxalobacteraceae and s_Oxalobacter_formigenes have been shown to reduce GC risk through “CD8br and CD8dim %leukocyte”. Our findings also indicates that g_Haemophilus could potentially elevate the likelihood of GC by regulating of “HVEM on naive CD4 + ” and “CCR2 on CD14- CD16 + monocyte”. HVEM, a member of the tumor necrosis factor receptor family (TNFRSF14), is commonly altered in cancer and is believed to have a tumor suppressor function in certain cancer situations [45]. The activation of costimulatory signals through the binding of the TNF ligand to its corresponding TNF receptor (TNFR) superfamily is crucial for the proliferation, maturation, and viability of antigen-experienced CD4 + and CD8 + T cells, which play key roles in adaptive immunity and various diseases [46]. Monocytes travel through the bloodstream and move to areas of inflammation where they can have either negative or positive effects according to their characteristics [47]. Monocytes/macrophages that express CCR2, especially in the tumor microenvironments, can strongly suppress the immune system. Recently, studies have used CCR2 antagonism to attract immunosuppressive monocytes/macrophages to tumors to change the tumor environment and improve the anti-tumor immune response [48]. Currently, studies elucidating how g_Haemophilus interacts with the HVEM/CCR2 pathway are lacking.

CD28 plays key roles in activating T cells and regulating immune tolerance. Reduced CD28 expression in senescent T cells can lead to an increase in CD8 + CD28-senescent populations in tumors [49]. The “ARO.PWY:chorismate biosynthesis I” was found to be mediated through “CD28-CD8dim %CD8dim” and “CD28-DN (CD4-CD8-) %T” cells. However, further studies are required to elucidate regulation of immune cells through this metabolic pathway. HLA-DR, a part of the major histocompatibility complex II that exhibits abnormal expression in specific types of tumors reasons, might be a noteworthy indicator [50]. “PWY.724 superpathway of L lysine, L threonine and L methionine biosynthesis II” was discovered to elevate the likelihood of GC via “HLA DR on CD33br HLA DR + CD14dim” cells, whereas s_Coprococcus_catus was found to lower the likelihood of GC through “HLA DR on CD33br HLA DR + CD14dim” cells. These metabolites highlight the significant and diverse effects of microbial metabolism on the host immune response and overall well-being.

This study distinguishes itself from other related studies through its thorough methodology, incorporating several rigorous analyses to examine the connections between GM and GC. Our conclusions are reinforced by the uniformity of our findings using various analytical methods, including weighted median, MR-Egger, and primary IVW. We sought to thoroughly examine specific GM genera and their connections to GC, which is a significant aspect of our study. These findings offer compelling insights into potential biological mechanisms underlying GC development.

Limitations

This study used genetic data from the a GWAS database, which only included European populations; therefore, the findings may not apply to all ethnic groups and populations. Furthermore, our analysis primarily focused on adenocarcinoma as a singular category within the pathological classification of GC, neglecting other pathological classifications within GC. This omission may obscure the potential impact of the GM on the susceptibility to GC across various pathological classifications. In addition, owing to the intricate nature of the immune system, additional mediating factors may not have been considered. Therefore, caution should be exercised when interpreting these results in a clinical context, particularly in relation to the individual health status of patients. The findings should be regarded as provisional and further verifications through subsequent investigations are warranted.

Conclusions

Our study emphasized the important role of the GM in affecting immune reactions and their possible effects on GC. The identified relationships and intermediary impacts established a basis for additional research, underscoring the importance of the gut-immune connection in well-being and illness.

Supplementary Information

Supplementary Material 1. The following supporting information has been uploaded, Fig.S1 Scatter plots for the causal association between 18 bacterial traits and GC; Fig.S2 Forest plot of the causal effect of 18 bacterial traits on GC; Fig.S3 Funnel plot of the causal effect of 18 bacterial traits on GC; Fig.S4 Leave-one-out plots for the causal association between 18 bacterial traits and GC; Fig.S5 MR analyses show causal effects between 16 immune cell traits and gastric cancer; Fig.S6 Forest plots showed the causal associations between 12 bacterial traits and 16 immune cell traits by using different methods; Fig.S7 Scatter plots for the causal association between 12 bacterial traits and 16 immune cell traits; Fig.S8 Forest plot of the causal effect of 12 bacterial traits on 16 immune cell traits; Fig.S9 Funnel plot of the causal effect of 12 bacterial traits on 16 immune cell traits; Fig.S10 Leave-one-out plots for the causal association between 12 bacterial traits and 16 immune cell traits; Fig.S11 Forest plots showed the causal associations between 16 immune cell traits and GC by using different methods; Fig.S12 Scatter plots for the causal association between 16 immune cell traits and GC; Fig.S13 Forest plot of the causal effect of 16 immune cell traits on GC; Fig.S14 Funnel plot of the causal effect of 16 immune cell traits on GC; Fig.S15 Leave-one-out plots for the causal association between 16 immune cell traits and GC; Supplementary Figure Legends; Supplementary Table 1: NHGRI-EBI GWAS Catalog accession numbers for each taxa and pathway; Supplementary Table 2: The detailed information of the 4048 SNPs for 207 taxa and 205 bacterial pathways trait; Supplementary Table 3: SNP for MR analysis between GC and 18 GC-related gut microbiotas; Supplementary Table 4: Sensitivity analysis in reverse MR analysis between GC and 18 GC-related gut microbiotas; Supplementary Table 5: Effect estimates of the associations of 18 bacterial traits with GC in the pleiotropy analysis; Supplementary Table 6: Effect estimates of the associations of 18 bacterial traits with GC in the MR PRESSO analysis; Supplementary Table 7: Effect estimates of the associations of 18 bacterial traits with GC in the heterogeneity analyses; Supplementary Table 8: 467 immune cell traits were included in the mediated MR analysis; Supplementary Table 9: Effect estimates of the associations of 16 immune cell traits with 12 bacterial traits in the pleiotropy analysis; Supplementary Table 10: Effect estimates of the associations of 16 immune cell traits with 12 bacterial traits in the heterogeneity analyses; Supplementary Table 11: Effect estimates of the associations of 12 immune cell traits with 16 bacterial traits in the MR PRESSO analysis; Supplementary Table 12: Effect estimates of the associations of GC with 16 immune cell traits in the pleiotropy analysis; Supplementary Table 13: Effect estimates of the associations of GC with 16 immune cell traits in the MR PRESSO analysis; Supplementary Table 14: Effect estimates of the associations of GC with 16 immune cell traits in the heterogeneity analyses.

Abbreviations

GM Gut microbiota

GC Gastric cancer

MR Mendelian randomization

GWAS Genome-Wide Association Studies

IVW Inverse variance-weighted

SNP Single nucleotide polymorphisms

IV Instrumental variable

LD Linkage disequilibrium

BWMR Bayesian weighted MR

OR Odds ratios

CI Confidence intervals

H. pylori Helicobacter pylori

ODN Oligodeoxynucleotides

TNF Tumor necrosis factor

Tregs Regulatory T cells

TNFRSF Tumor necrosis factor receptor family

TNFR TNF receptor

ENTPD1 Triphosphate diphosphohydrolase 1

eIF5A Eukaryotic initiation factor 5A

Acknowledgements

We thank all the GWASs for making the summary data publicly available, and we are grateful for all the investigators and participants who contributed to those studies.

Author contributions

JW analyzed the data and prepared the original draft. BJF and WXG validated the data and revised the original draft. CW conceived the entire study, collated the data, prepared the original draft, and completed the final revision of the manuscript. All authors reviewed the manuscript.

Funding

This work was supported by grants from the Wuxi Health Commission Youth Project (Q202242), Top Talent Support Program for Young and Middle-Aged People of the Wuxi Health Committee (HB2023046), and National Natural Science Foundation of China (82303117).

Data availability

Data is provided within the manuscript or supplementary information files. Publicly available data were analyzed in this study, which can be found here: https://www.ebi.ac.uk/gwas/; https://storage.googleapis.com/finngen-public-data-r9/summary_stats/finngen_R9_C3_STOMACH_EXALLC.gz.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Chao Wang and Jia Wang contributed equally to this work and shared first authorship.
==== Refs
References

1. Sung H Ferlay J Siegel RL Laversanne M Soerjomataram I Jemal A Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries CA Cancer J Clin 2021 71 3 209 249 10.3322/caac.21660 33538338
Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209–49.33538338 10.3322/caac.21660
2. Smyth EC Nilsson M Grabsch HI van Grieken NC Lordick F Gastric cancer Lancet 2020 396 10251 635 648 10.1016/S0140-6736(20)31288-5 32861308
Smyth EC, Nilsson M, Grabsch HI, van Grieken NC, Lordick F. Gastric cancer. Lancet. 2020;396(10251):635–48.32861308 10.1016/S0140-6736(20)31288-5
3. Machlowska J Baj J Sitarz M Maciejewski R Sitarz R Gastric cancer: epidemiology, risk factors, classification, genomic characteristics and treatment strategies IJMS. 2020 21 11 4012 10.3390/ijms21114012 32512697
Machlowska J, Baj J, Sitarz M, Maciejewski R, Sitarz R. Gastric cancer: epidemiology, risk factors, classification, genomic characteristics and treatment strategies. IJMS. 2020;21(11):4012.32512697 10.3390/ijms21114012
4. Morgan E Arnold M Camargo MC Gini A Kunzmann AT Matsuda T The current and future incidence and mortality of gastric cancer in 185 countries, 2020–40: a population-based modelling study EClinicalMedicine 2022 47 101404 10.1016/j.eclinm.2022.101404 35497064
Morgan E, Arnold M, Camargo MC, Gini A, Kunzmann AT, Matsuda T, et al. The current and future incidence and mortality of gastric cancer in 185 countries, 2020–40: a population-based modelling study. EClinicalMedicine. 2022;47: 101404.35497064 10.1016/j.eclinm.2022.101404
5. Bessède E Mégraud F Microbiota and gastric cancer Semin Cancer Biol 2022 86 Pt 3 11 17 10.1016/j.semcancer.2022.05.001 35533800
Bessède E, Mégraud F. Microbiota and gastric cancer. Semin Cancer Biol. 2022;86(Pt 3):11–7.35533800 10.1016/j.semcancer.2022.05.001
6. Wang M Yang G Tian Y Zhang Q Liu Z Xin Y The role of the gut microbiota in gastric cancer: the immunoregulation and immunotherapy Front Immunol 2023 14 1183331 10.3389/fimmu.2023.1183331 37457738
Wang M, Yang G, Tian Y, Zhang Q, Liu Z, Xin Y. The role of the gut microbiota in gastric cancer: the immunoregulation and immunotherapy. Front Immunol. 2023;14:1183331.37457738 10.3389/fimmu.2023.1183331
7. Hidalgo-Cantabrana C Delgado S Ruiz L Ruas-Madiedo P Sánchez B Margolles A Bifidobacteria and their health-promoting effects Microbiol Spectr. 2017 10.1128/microbiolspec.BAD-0010-2016 28643627
Hidalgo-Cantabrana C, Delgado S, Ruiz L, Ruas-Madiedo P, Sánchez B, Margolles A. Bifidobacteria and their health-promoting effects. Microbiol Spectr. 2017. 10.1128/microbiolspec.BAD-0010-2016.28643627 10.1128/microbiolspec.BAD-0010-2016
8. Hou K Wu Z-X Chen X-Y Wang J-Q Zhang D Xiao C Microbiota in health and diseases Signal Transduct Target Ther 2022 7 1 135 10.1038/s41392-022-00974-4 35461318
Hou K, Wu Z-X, Chen X-Y, Wang J-Q, Zhang D, Xiao C, et al. Microbiota in health and diseases. Signal Transduct Target Ther. 2022;7(1):135.35461318 10.1038/s41392-022-00974-4
9. González-Sánchez P DeNicola GM The microbiome(s) and cancer: know thy neighbor(s) J Pathol 2021 254 4 332 343 10.1002/path.5661 33723873
González-Sánchez P, DeNicola GM. The microbiome(s) and cancer: know thy neighbor(s). J Pathol. 2021;254(4):332–43.33723873 10.1002/path.5661
10. Wang L-L Yu X-J Zhan S-H Jia S-J Tian Z-B Dong Q-J Participation of microbiota in the development of gastric cancer World J Gastroenterol 2014 20 17 4948 4952 10.3748/wjg.v20.i17.4948 24803806
Wang L-L, Yu X-J, Zhan S-H, Jia S-J, Tian Z-B, Dong Q-J. Participation of microbiota in the development of gastric cancer. World J Gastroenterol. 2014;20(17):4948–52.24803806 10.3748/wjg.v20.i17.4948
11. Nasr R Shamseddine A Mukherji D Nassar F Temraz S The crosstalk between microbiome and immune response in gastric cancer IJMS. 2020 21 18 6586 10.3390/ijms21186586 32916853
Nasr R, Shamseddine A, Mukherji D, Nassar F, Temraz S. The crosstalk between microbiome and immune response in gastric cancer. IJMS. 2020;21(18):6586.32916853 10.3390/ijms21186586
12. Yang W Cong Y Gut microbiota-derived metabolites in the regulation of host immune responses and immune-related inflammatory diseases Cell Mol Immunol 2021 18 4 866 877 10.1038/s41423-021-00661-4 33707689
Yang W, Cong Y. Gut microbiota-derived metabolites in the regulation of host immune responses and immune-related inflammatory diseases. Cell Mol Immunol. 2021;18(4):866–77.33707689 10.1038/s41423-021-00661-4
13. Sekula P Del Greco MF Pattaro C Köttgen A Mendelian randomization as an approach to assess causality using observational data J Am Soc Nephrol 2016 27 11 3253 3265 10.1681/ASN.2016010098 27486138
Sekula P, Del Greco MF, Pattaro C, Köttgen A. Mendelian randomization as an approach to assess causality using observational data. J Am Soc Nephrol. 2016;27(11):3253–65.27486138 10.1681/ASN.2016010098
14. Orrù V Steri M Sidore C Marongiu M Serra V Olla S Complex genetic signatures in immune cells underlie autoimmunity and inform therapy Nat Genet 2020 52 10 1036 1045 10.1038/s41588-020-0684-4 32929287
Orrù V, Steri M, Sidore C, Marongiu M, Serra V, Olla S, et al. Complex genetic signatures in immune cells underlie autoimmunity and inform therapy. Nat Genet. 2020;52(10):1036–45.32929287 10.1038/s41588-020-0684-4
15. Bowden J Holmes MV Meta-analysis and Mendelian randomization: a review Res Synth Methods 2019 10 4 486 496 10.1002/jrsm.1346 30861319
Bowden J, Holmes MV. Meta-analysis and Mendelian randomization: a review. Res Synth Methods. 2019;10(4):486–96.30861319 10.1002/jrsm.1346
16. Lopera-Maya EA Kurilshikov A van der Graaf A Hu S Andreu-Sánchez S Chen L Effect of host genetics on the gut microbiome in 7738 participants of the Dutch Microbiome Project Nat Genet 2022 54 2 143 151 10.1038/s41588-021-00992-y 35115690
Lopera-Maya EA, Kurilshikov A, van der Graaf A, Hu S, Andreu-Sánchez S, Chen L, et al. Effect of host genetics on the gut microbiome in 7738 participants of the Dutch Microbiome Project. Nat Genet. 2022;54(2):143–51.35115690 10.1038/s41588-021-00992-y
17. Myers TA Chanock SJ Machiela MJ LDlinkR: an r package for rapidly calculating linkage disequilibrium statistics in diverse populations Front Genet 2020 11 157 10.3389/fgene.2020.00157 32180801
Myers TA, Chanock SJ, Machiela MJ. LDlinkR: an r package for rapidly calculating linkage disequilibrium statistics in diverse populations. Front Genet. 2020;11:157.32180801 10.3389/fgene.2020.00157
18. Slatkin M Linkage disequilibrium–understanding the evolutionary past and mapping the medical future Nat Rev Genet 2008 9 6 477 485 10.1038/nrg2361 18427557
Slatkin M. Linkage disequilibrium–understanding the evolutionary past and mapping the medical future. Nat Rev Genet. 2008;9(6):477–85.18427557 10.1038/nrg2361
19. Palmer TM Lawlor DA Harbord RM Sheehan NA Tobias JH Timpson NJ Using multiple genetic variants as instrumental variables for modifiable risk factors Stat Methods Med Res 2012 21 3 223 242 10.1177/0962280210394459 21216802
Palmer TM, Lawlor DA, Harbord RM, Sheehan NA, Tobias JH, Timpson NJ, et al. Using multiple genetic variants as instrumental variables for modifiable risk factors. Stat Methods Med Res. 2012;21(3):223–42.21216802 10.1177/0962280210394459
20. Levin MG Judy R Gill D Vujkovic M Verma SS Bradford Y Genetics of height and risk of atrial fibrillation: a Mendelian randomization study PLoS Med 2020 17 10 e1003288 10.1371/journal.pmed.1003288 33031386
Levin MG, Judy R, Gill D, Vujkovic M, Verma SS, Bradford Y, et al. Genetics of height and risk of atrial fibrillation: a Mendelian randomization study. PLoS Med. 2020;17(10): e1003288.33031386 10.1371/journal.pmed.1003288
21. Gill D Efstathiadou A Cawood K Tzoulaki I Dehghan A Education protects against coronary heart disease and stroke independently of cognitive function: evidence from Mendelian randomization Int J Epidemiol 2019 48 5 1468 1477 10.1093/ije/dyz200 31562522
Gill D, Efstathiadou A, Cawood K, Tzoulaki I, Dehghan A. Education protects against coronary heart disease and stroke independently of cognitive function: evidence from Mendelian randomization. Int J Epidemiol. 2019;48(5):1468–77.31562522 10.1093/ije/dyz200
22. Burgess S Small DS Thompson SG A review of instrumental variable estimators for Mendelian randomization Stat Methods Med Res 2017 26 5 2333 2355 10.1177/0962280215597579 26282889
Burgess S, Small DS, Thompson SG. A review of instrumental variable estimators for Mendelian randomization. Stat Methods Med Res. 2017;26(5):2333–55.26282889 10.1177/0962280215597579
23. Zhao J Ming J Hu X Chen G Liu J Yang C Bayesian weighted Mendelian randomization for causal inference based on summary statistics Bioinformatics 2020 36 5 1501 1508 10.1093/bioinformatics/btz749 31593215
Zhao J, Ming J, Hu X, Chen G, Liu J, Yang C. Bayesian weighted Mendelian randomization for causal inference based on summary statistics. Bioinformatics. 2020;36(5):1501–8.31593215 10.1093/bioinformatics/btz749
24. Burgess S Scott RA Timpson NJ Davey Smith G Thompson SG Using published data in Mendelian randomization: a blueprint for efficient identification of causal risk factors Eur J Epidemiol 2015 30 7 543 552 10.1007/s10654-015-0011-z 25773750
Burgess S, Scott RA, Timpson NJ, Davey Smith G, Thompson SG. Using published data in Mendelian randomization: a blueprint for efficient identification of causal risk factors. Eur J Epidemiol. 2015;30(7):543–52.25773750 10.1007/s10654-015-0011-z
25. Burgess S Butterworth A Thompson SG Mendelian randomization analysis with multiple genetic variants using summarized data Genet Epidemiol 2013 37 7 658 665 10.1002/gepi.21758 24114802
Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. 2013;37(7):658–65.24114802 10.1002/gepi.21758
26. Verbanck M Chen C-Y Neale B Do R Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases Nat Genet 2018 50 5 693 698 10.1038/s41588-018-0099-7 29686387
Verbanck M, Chen C-Y, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018;50(5):693–8.29686387 10.1038/s41588-018-0099-7
27. Bowden J Davey Smith G Burgess S Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression Int J Epidemiol 2015 44 2 512 525 10.1093/ije/dyv080 26050253
Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44(2):512–25.26050253 10.1093/ije/dyv080
28. Raubenheimer K Bondonno C Blekkenhorst L Wagner K-H Peake JM Neubauer O Effects of dietary nitrate on inflammation and immune function, and implications for cardiovascular health Nutr Rev 2019 77 8 584 599 10.1093/nutrit/nuz025 31150091
Raubenheimer K, Bondonno C, Blekkenhorst L, Wagner K-H, Peake JM, Neubauer O. Effects of dietary nitrate on inflammation and immune function, and implications for cardiovascular health. Nutr Rev. 2019;77(8):584–99.31150091 10.1093/nutrit/nuz025
29. Kim H-N Kim M-J Jacobs JP Yang H-J Altered gastric microbiota and inflammatory cytokine responses in patients with helicobacter pylori-negative gastric cancer Nutrients 2022 14 23 4981 10.3390/nu14234981 36501012
Kim H-N, Kim M-J, Jacobs JP, Yang H-J. Altered gastric microbiota and inflammatory cytokine responses in patients with helicobacter pylori-negative gastric cancer. Nutrients. 2022;14(23):4981.36501012 10.3390/nu14234981
30. Huang K Gao X Wu L Yan B Wang Z Zhang X Salivary microbiota for gastric cancer prediction: an exploratory study Front Cell Infect Microbiol 2021 11 640309 10.3389/fcimb.2021.640309 33777850
Huang K, Gao X, Wu L, Yan B, Wang Z, Zhang X, et al. Salivary microbiota for gastric cancer prediction: an exploratory study. Front Cell Infect Microbiol. 2021;11: 640309.33777850 10.3389/fcimb.2021.640309
31. Castaño-Rodríguez N Goh K-L Fock KM Mitchell HM Kaakoush NO Dysbiosis of the microbiome in gastric carcinogenesis Sci Rep 2017 7 1 15957 10.1038/s41598-017-16289-2 29162924
Castaño-Rodríguez N, Goh K-L, Fock KM, Mitchell HM, Kaakoush NO. Dysbiosis of the microbiome in gastric carcinogenesis. Sci Rep. 2017;7(1):15957.29162924 10.1038/s41598-017-16289-2
32. Vincent CT Schneider RJ Selective tRNA charging in breast cancer Nat Cell Biol 2022 24 3 287 289 10.1038/s41556-022-00863-6 35288657
Vincent CT, Schneider RJ. Selective tRNA charging in breast cancer. Nat Cell Biol. 2022;24(3):287–9.35288657 10.1038/s41556-022-00863-6
33. Parker BJ Wearsch PA Veloo ACM Rodriguez-Palacios A The genus alistipes: gut bacteria with emerging implications to inflammation, cancer, and mental health Front Immunol 2020 11 906 10.3389/fimmu.2020.00906 32582143
Parker BJ, Wearsch PA, Veloo ACM, Rodriguez-Palacios A. The genus alistipes: gut bacteria with emerging implications to inflammation, cancer, and mental health. Front Immunol. 2020;11:906.32582143 10.3389/fimmu.2020.00906
34. Nakano T Dohi O Takagi T Naito Y Fukui H Miyazaki H Characteristics of gastric mucosa-associated microbiota in patients with early gastric cancer after successful helicobacter pylori eradication Dig Dis Sci 2023 68 12 4398 4406 10.1007/s10620-023-08154-2 37875607
Nakano T, Dohi O, Takagi T, Naito Y, Fukui H, Miyazaki H, et al. Characteristics of gastric mucosa-associated microbiota in patients with early gastric cancer after successful helicobacter pylori eradication. Dig Dis Sci. 2023;68(12):4398–406.37875607 10.1007/s10620-023-08154-2
35. Daniel SL Moradi L Paiste H Wood KD Assimos DG Holmes RP Forty years of Oxalobacter formigenes, a gutsy oxalate-degrading specialist Appl Environ Microbiol 2021 87 18 e0054421 10.1128/AEM.00544-21 34190610
Daniel SL, Moradi L, Paiste H, Wood KD, Assimos DG, Holmes RP, et al. Forty years of Oxalobacter formigenes, a gutsy oxalate-degrading specialist. Appl Environ Microbiol. 2021;87(18): e0054421.34190610 10.1128/AEM.00544-21
36. Sheridan PO Louis P Tsompanidou E Shaw S Harmsen HJ Duncan SH Distribution, organization and expression of genes concerned with anaerobic lactate utilization in human intestinal bacteria Microb Genom. 2022 10.1099/mgen.0.000739 35077342
Sheridan PO, Louis P, Tsompanidou E, Shaw S, Harmsen HJ, Duncan SH, et al. Distribution, organization and expression of genes concerned with anaerobic lactate utilization in human intestinal bacteria. Microb Genom. 2022. 10.1099/mgen.0.000739.35077342 10.1099/mgen.0.000739
37. Zitvogel L Ma Y Raoult D Kroemer G Gajewski TF The microbiome in cancer immunotherapy: Diagnostic tools and therapeutic strategies Science (New York, NY) 2018 359 6382 1366 1370 10.1126/science.aar6918
Zitvogel L, Ma Y, Raoult D, Kroemer G, Gajewski TF. The microbiome in cancer immunotherapy: Diagnostic tools and therapeutic strategies. Science (New York, NY). 2018;359(6382):1366–70.10.1126/science.aar6918
38. Sundström P Stenstad H Langenes V Ahlmanner F Theander L Ndah TG Regulatory T cells from colon cancer patients inhibit effector T-cell migration through an adenosine-dependent mechanism Cancer Immunol Res 2016 4 3 183 193 10.1158/2326-6066.CIR-15-0050 26787824
Sundström P, Stenstad H, Langenes V, Ahlmanner F, Theander L, Ndah TG, et al. Regulatory T cells from colon cancer patients inhibit effector T-cell migration through an adenosine-dependent mechanism. Cancer Immunol Res. 2016;4(3):183–93.26787824 10.1158/2326-6066.CIR-15-0050
39. Cai X-Y Wang X-F Li J Dong J-N Liu J-Q Li N-P High expression of CD39 in gastric cancer reduces patient outcome following radical resection Oncol Lett 2016 12 5 4080 4086 10.3892/ol.2016.5189 27895775
Cai X-Y, Wang X-F, Li J, Dong J-N, Liu J-Q, Li N-P, et al. High expression of CD39 in gastric cancer reduces patient outcome following radical resection. Oncol Lett. 2016;12(5):4080–6.27895775 10.3892/ol.2016.5189
40. Wang J Li R Cao Y Gu Y Fang H Fei Y Intratumoral CXCR5+CD8+T associates with favorable clinical outcomes and immunogenic contexture in gastric cancer Nat Commun 2021 12 1 3080 10.1038/s41467-021-23356-w 34035252
Wang J, Li R, Cao Y, Gu Y, Fang H, Fei Y, et al. Intratumoral CXCR5+CD8+T associates with favorable clinical outcomes and immunogenic contexture in gastric cancer. Nat Commun. 2021;12(1):3080.34035252 10.1038/s41467-021-23356-w
41. Al-Habsi M Chamoto K Matsumoto K Nomura N Zhang B Sugiura Y Spermidine activates mitochondrial trifunctional protein and improves antitumor immunity in mice Science 2022 378 6618 eabj3510 10.1126/science.abj3510 36302005
Al-Habsi M, Chamoto K, Matsumoto K, Nomura N, Zhang B, Sugiura Y, et al. Spermidine activates mitochondrial trifunctional protein and improves antitumor immunity in mice. Science. 2022;378(6618):eabj3510.36302005 10.1126/science.abj3510
42. Elmarsafawi AG Hesterberg RS Fernandez MR Yang C Darville LN Liu M Modulating the polyamine/hypusine axis controls generation of CD8+ tissue-resident memory T cells JCI Insight. 2023 10.1172/jci.insight.169308 37581943
Elmarsafawi AG, Hesterberg RS, Fernandez MR, Yang C, Darville LN, Liu M, et al. Modulating the polyamine/hypusine axis controls generation of CD8+ tissue-resident memory T cells. JCI Insight. 2023. 10.1172/jci.insight.169308.37581943 10.1172/jci.insight.169308
43. Du Y Wei Y Therapeutic potential of natural killer cells in gastric cancer Front Immunol 2018 9 3095 10.3389/fimmu.2018.03095 30719024
Du Y, Wei Y. Therapeutic potential of natural killer cells in gastric cancer. Front Immunol. 2018;9:3095.30719024 10.3389/fimmu.2018.03095
44. Kindlund B Sjöling Å Yakkala C Adamsson J Janzon A Hansson L-E CD4+ regulatory T cells in gastric cancer mucosa are proliferating and express high levels of IL-10 but little TGF-β Gastric Cancer 2017 20 1 116 125 10.1007/s10120-015-0591-z 26782287
Kindlund B, Sjöling Å, Yakkala C, Adamsson J, Janzon A, Hansson L-E, et al. CD4+ regulatory T cells in gastric cancer mucosa are proliferating and express high levels of IL-10 but little TGF-β. Gastric Cancer. 2017;20(1):116–25.26782287 10.1007/s10120-015-0591-z
45. Šedý JR Ramezani-Rad P HVEM network signaling in cancer Adv Cancer Res 2019 142 145 186 10.1016/bs.acr.2019.01.004 30885361
Šedý JR, Ramezani-Rad P. HVEM network signaling in cancer. Adv Cancer Res. 2019;142:145–86.30885361 10.1016/bs.acr.2019.01.004
46. So T Ishii N The TNF-TNFR family of co-signal molecules Adv Exp Med Biol 2019 1189 53 84 10.1007/978-981-32-9717-3_3 31758531
So T, Ishii N. The TNF-TNFR family of co-signal molecules. Adv Exp Med Biol. 2019;1189:53–84.31758531 10.1007/978-981-32-9717-3_3
47. França CN, Izar MCO, Hortêncio MNS, Amaral JB do, Ferreira CES, Tuleta ID, et al. Monocyte subtypes and the CCR2 chemokine receptor in cardiovascular disease. Clin Sci 2017;131(12):1215–24.
48. Fei L Ren X Yu H Zhan Y Targeting the CCL2/CCR2 Axis in cancer immunotherapy: one stone, three birds? Front Immunol 2021 12 771210 10.3389/fimmu.2021.771210 34804061
Fei L, Ren X, Yu H, Zhan Y. Targeting the CCL2/CCR2 Axis in cancer immunotherapy: one stone, three birds? Front Immunol. 2021;12: 771210.34804061 10.3389/fimmu.2021.771210
49. Huff WX Kwon JH Henriquez M Fetcko K Dey M The evolving role of CD8+CD28− immunosenescent T cells in cancer immunology IJMS. 2019 20 11 2810 10.3390/ijms20112810 31181772
Huff WX, Kwon JH, Henriquez M, Fetcko K, Dey M. The evolving role of CD8+CD28− immunosenescent T cells in cancer immunology. IJMS. 2019;20(11):2810.31181772 10.3390/ijms20112810
50. Amrane K Le Meur C Besse B Hemon P Le Noac'h P Pradier O HLA-DR expression in melanoma: from misleading therapeutic target to potential immunotherapy biomarker Front Immunol 2023 14 1285895 10.3389/fimmu.2023.1285895 38299143
Amrane K, Le Meur C, Besse B, Hemon P, Le Noac’h P, Pradier O, et al. HLA-DR expression in melanoma: from misleading therapeutic target to potential immunotherapy biomarker. Front Immunol. 2023;14:1285895.38299143 10.3389/fimmu.2023.1285895
