Microbial and Metabolomic Variations Correlated With Gastric Cancer Subtypes and Prognosis

Yan Yang, Liping Wen, Lin Wu, Yiran Chen, Rui Yang, Chao He, Yingzi Zhang, Jing Zhang, Haohao Wang, Haiyong Wang, Lisong TengMicrobiologyOpen 2025 · DOI

TL;DR

This study delves into the complex interplay between the gut microbiome, metabolic profiles, and gastric cancer (GC) subtypes. By analyzing tumor and normal tissues from GC patients, researchers identified distinct microbial and metabolic signatures associated with different GC classifications (Lauren and ZJU), revealing that the ZJU classification appears more closely linked to patient prognosis.

What Problem Does This Paper Address?

Gastric cancer (GC) is a significant global health challenge, particularly in China. While it's known that GC exhibits diverse features based on its subtypes (like the Lauren classification), and that the gut microbiome and host metabolism are intertwined with cancer development, the specific microbial and metabolic differences across these subtypes and their impact on prognosis have remained largely unexplored. This paper aims to bridge this knowledge gap by comprehensively investigating these variations, seeking to refine our understanding of GC heterogeneity and its clinical implications.

Methods

The researchers employed a multi-omics approach, combining two key techniques:

  • 16S rRNA Gene Sequencing: This standard technique was used to profile the microbial communities present in tumor and normal tissue samples. By amplifying and sequencing specific regions of the 16S rRNA gene, they could identify and quantify different bacterial genera.
  • Untargeted Metabolomics (LC-MS): This method was used to identify and quantify a broad range of small molecules (metabolites) within the tissue samples. Liquid chromatography coupled with mass spectrometry (LC-MS) allows for the detection of thousands of metabolites, providing a snapshot of the tissue's metabolic state.

Bioinformatics analyses, including LefSe (Linear discriminant analysis Effect Size) for identifying differential taxa, PCoA (Principal Coordinate Analysis) for visualizing community structure, PLS-DA (Partial Least Squares Discriminant Analysis) for metabolite classification, and KEGG pathway analysis for understanding metabolic functions, were used to interpret the data. Correlation analyses (Spearman and Pearson) were performed to link microbial and metabolic findings.

Datasets

The study utilized paired tumor and adjacent normal tissue samples from 50 gastric cancer patients. Of these, 16S rRNA gene sequencing was performed on all 50 patients, while untargeted metabolomics was conducted on 40 pairs. The clinicopathological data for these patients, including Lauren classification, ZJU classification, TNM stage, and survival status, were also collected and analyzed.

Key Insights and Ideas

  • Subtype-Specific Signatures: Both the Lauren and ZJU classifications revealed distinct microbial and metabolic profiles in tumor tissues compared to normal tissues. This highlights that GC isn't a monolithic disease but rather a spectrum of conditions with unique biological underpinnings.
  • Shared and Unique Features: While both classification systems identified some common differential microbes (e.g., Fusobacterium, Haemophilus) and metabolites, each classification also possessed unique signatures. This suggests that while there's overlap, each system captures different facets of GC biology.
  • ZJU Classification's Prognostic Link: The ZJU classification, a molecularly-derived system, showed a stronger association with prognosis compared to the Lauren classification. This is a crucial finding, suggesting that molecular subtyping might be more informative for predicting patient outcomes.
  • Prognostic Biomarkers Identified: The study pinpointed Eubacterium_ventriosum_group as a microbe positively correlated with poor prognosis, and N6-Succinyl Adenosine as a metabolite negatively correlated with poor prognosis. These could potentially serve as novel biomarkers for risk stratification.
  • Microbe-Metabolite Interplay: The research explored how specific microbes and metabolites interact within different GC subtypes, revealing complex relationships that could influence tumor development and progression.

Key Results

  • Variations in microbial composition and metabolite landscapes were observed across different subtypes, irrespective of whether the Lauren or ZJU classification was used.
  • Both Lauren and ZJU classifications shared differential microbiota, including Fusobacterium and Haemophilus.
  • 38 out of the top 50 differential metabolites were common to both classifications.
  • The ZJU classification showed distinct features, with Eubacterium_ventriosum_group and N6-Succinyl Adenosine being characteristic differences.
  • Multivariate survival analysis identified Eubacterium_ventriosum_group as a positive predictor of poor prognosis and N6-Succinyl Adenosine as a negative predictor.
  • Diffuse-type GC patients had significantly worse outcomes compared to intestinal-type patients.
  • Among ZJU subtypes, subtype 3 showed the worst prognosis, followed by subtype 2.

Limitations and Future Work

The authors acknowledge several limitations:

  • Small Sample Size: The study involved 50 GC patients, which they note is a small sample size that could bias results and limit generalizability.
  • Drug Influence: Metabolite profiling may have been influenced by patients' comorbidities and medication use, potentially confounding results.
  • Nontargeted Metabolomics Variability: Nontargeted metabolomics can be variable, and the absence of a validation set to confirm findings is a constraint.

Future work could focus on validating these findings in larger, independent cohorts. Further research into the specific mechanisms by which the identified microbial and metabolic factors influence GC development and progression is warranted. Exploring the therapeutic potential of targeting these factors could also be a promising avenue.

Summary updated

Abstract

Gastric cancer (GC) persists as the third most prevalent malignancy in China. GC exhibits distinct features when stratified by Lauren/ZJU subtypes. The interdependence of microbes, metabolites, and tumor evolution is recognized. Nevertheless, the specific microbial and metabolite disparities related to the Lauren and ZJU subtypes of GC have yet to be thoroughly investigated. In this study, we employed 16S sequencing of microbial communities and conducted untargeted metabolomic assessments on tumor tissues and their matched normal controls from 50 GC patients. We observed variations in microbial composition and metabolite landscapes across subtypes, irrespective of the Lauren or ZJU classification. We explored the associations and differences between the Lauren and the ZJU classification. It was found that both classifications share differential microbiota, including Fusobacterium and Haemophilus. Additionally, 38 of the top 50 differential metabolites are common to both classifications. However, distinct classifications also exhibit unique differential microbiota and metabolite characteristics. Among them, Eubacterium_ventriosum_group and N6-Succinyl Adenosine are both characteristic differences of the ZJU classification. Multivariate survival analysis disclosed that Eubacterium_ventriosum_group positively correlates with poor prognosis, whereas N6-Succinyl Adenosine negatively correlates with poor prognosis. Our research delineates the microbiota and metabolites specific to different subtypes of GC and investigates the interplay between these differential elements, as well as their prognostic significance. We have identified two distinct features that are both associated with the ZJU classification, suggesting that the ZJU classification is more closely related to prognosis.
BibTeX
@inproceedings{yang2025microbial,
  author = {Yan Yang and Liping Wen and Lin Wu and Yiran Chen and Rui Yang and Chao He and Yingzi Zhang and Jing Zhang and Haohao Wang and Haiyong Wang and Lisong Teng and Haiyong Wang and Lisong Teng},
  title = {{Microbial and Metabolomic Variations Correlated With Gastric Cancer Subtypes and Prognosis}},
  booktitle = {MicrobiologyOpen},
  year = {2025},
  doi = {10.1002/mbo3.70139},
}
Figure