Lymphatic metastasis of lung adenocarcinoma characterized by network with correlations between microbiota and upregulated gene sets involved in tumor migration and development

To the Editor: Lung cancer is the leading cause of cancer death worldwide. The most prevalent pathological subtype of lung cancer is lung adenocarcinoma (LUAD).[1] Genetic and epigenetic variants are crucial, as well as the gut and intratumor microbiota, in affecting lung cancer susceptibility and migration. Therefore, we retrospectively studied the molecular characteristics of patients with LUAD to determine whether the intratumor microbiota affects lymph node metastasis. This study was approved by the Institutional Ethics Review Committee of First People’s Hospital of Yulin (No. YLSY-IRB-SR-2024122). The need for informed consent was waived because of the retrospective nature of this study. We collected genomic, transcriptomic, microbial, and clinical data of patients with LUAD from The Cancer Genome Atlas, cBioPortal and from a publicly available dataset (ftp://ftp.microbio.me/pub/cancer_microbiome_analysis/). Samples without microbial or RNA-seq data were excluded from the analysis. The patients with LUAD were categorized into two groups based on tumor stage: LN+ (T ≥1N >0M0 disease) and LN– (T ≥1N0M0 disease) [Supplementary Figure 1A, https://links.lww.com/CM9/C726]. The methods are detailed in the Supplementary File, https://links.lww.com/CM9/C726. The study workflow is shown in Supplementary Figure 1A, https://links.lww.com/CM9/C726, 458 patients were included, with 152 and 306 in the LN+ and LN– groups, respectively [Supplementary Tables 1 and 2, https://links.lww.com/CM9/C727]. The baseline patient characteristics were similar between the groups [Supplementary Table 1, https://links.lww.com/CM9/C727]. A total of 332 (72%) of the included patients were 60 years of age or older, 250 (55%) were women, and 317 (69%) were smokers [Supplementary Table 1, https://links.lww.com/CM9/C727]. The risk of death was 2.59-fold higher in patients with LN metastasis (hazard ratio [HR]: 2.59, 95% confidence interval [CI]: 1.89–3.55; P <0.001) [Figure 1A].Figure 1: Patients with lymph node metastases were characterized by a correlated network between microbiota and upregulated gene sets involving tumor migration and development. (A) Tumors with lymph node metastases were associated with shorter overall survival. (B) Associations of expression levels of genes driving lung cancer with lymph node metastasis status. (C) Gene set enrichment analysis showing enriched biomarker gene sets. (D) Ranking of bacterial abundance at the genus level. * P-value <0.05. (E) Taxonomic cladogram from linear discriminant analysis of effect size, depicting taxonomic associations between bacterial communities from patients in LN+ and LN– groups. Each node represents a specific taxonomic type. Red and blue nodes denote the taxonomic types with higher abundance in LN+ than in LN– and in LN– than in LN+, respectively. (F) Linear discriminant analysis score computed from features differentially abundant between LN+ and LN– patients. The criterion for feature selection was |linear discriminant analysis score| >3. Correlation network between gene sets and microbiota biomarkers in (G) LN+ and (H) LN– groups. Each line represents a microbe–gene correlation pair, where red and blue lines indicate positive and negative correlations, respectively. The node size is proportional to the degree of correlation; and green, red and blue nodes indicate the gene, LN+ microbe, and LN- microbe biomarkers, respectively. Biomarker gene sets (H1–H7) in gene set enrichment analysis were included in the microbiota–marker network. NE: Not evaluated; LN+: T ≥1N >0M0 disease; LN–: T ≥1N0M0 disease.Mutational analysis identified 166,613 single-nucleotide variants, 4811 insertions/deletions, and 2562 insertions. The most commonly mutated genes were TP53 (48%), TTN (47%), MUC16 (41%), CSMD3 (37%), and RYR2 (36%) [Supplementary Figure 1B, https://links.lww.com/CM9/C726]. The frequencies of these mutations did not differ between the LN+ and LN– groups. No genes known to drive lung cancer were associated with metastasis [Figure 1B]. Signature 4 (smoking-related) was the predominant mutational signature for carcinogenesis [Supplementary Figure 1C, https://links.lww.com/CM9/C726]. Several key gene sets, including Myc targets, E2F targets, the G2M checkpoint, mitotic spindle, epithelial–mesenchymal transition (EMT), glycolysis, and oxidative phosphorylation, were over-represented at the transcriptome level in the LN+ group [Figure 1C]. The five immunological scores did not markedly differ between the groups. Analyzing the composition of immune cells showed that the anticancer status was weaker in the LN+ group, which was characterized by reduced levels of CD8 + T and natural killer (NK) cells, along by an upregulation of M0 macrophages [Supplementary Figure 2A, B, https://links.lww.com/CM9/C726]. This observation implies a role for these gene sets and immune cells in facilitating lymph node metastasis. We did not find any significant differences between the groups in the microbial α-diversity (P >0.05) or β-diversity indices (P = 0.156) based on the Bray–Curtis distances [Supplementary Figure 3A, B, https://links.lww.com/CM9/C726]. The relative abundance of the Listeria clade was high in the LN– group, which includes the Firmicutes phylum, Bacilli, Bacillales, Listeriaceae, and Listeria [Figure 1D, E; Supplementary Figure 3C and D, https://links.lww.com/CM9/C726]. Terrabacter, Bordetella, Listeria, Luteibacter, Proteus, Chlamydia, Klebsiella, and Bacteroides were substantially more abundant in the LN– group than in the LN+ group and are potential microbial biomarkers. Conversely, Aeromonas and Neisseria were more abundant in the LN+ group than in the LN– group [Figure 1F]. A strong correlation between Listeria and M2 macrophage abundances was detected in the LN+ group [Supplementary Figure 4A, B, https://links.lww.com/CM9/C726], suggesting an enhanced carcinogenic inflammatory status. This combination of enrichment and microbial variety provides insights into the relationship between the bacteria and gene expression profiles in LUAD. Comparing the bacterial biomarker networks between the LN+ and LN– groups revealed differences in topology: the network was closely interconnected in the LN+ group [Figure 1G, Supplementary Table 3, https://links.lww.com/CM9/C727], whereas the network interconnections were weaker in the LN– group [Figure 1H, Supplementary Table S4, https://links.lww.com/CM9/C727]. The subnetworks focused on Aeromonas and Listeria markedly differed in connectivity. Aeromonas was positively associated with oxidative phosphorylation and Myc target gene sets in the LN+ network. The LN+ network, not the LN– network, was characterized by these links to Listeria [Supplementary Figure 5A, B, https://links.lww.com/CM9/C726]. Listeria significantly and negatively correlated with gene sets related to energy metabolism (glycolysis and oxidative phosphorylation) and tumor migration (EMT) in the LN+ network. However, Listeria abundance positively correlated with cell-cycle-related (G2M checkpoint and mitotic spindle) gene sets. The abundance of Listeria in the LN+ group was substantially lower than that in the LN– group, indicating that decreased Listeria abundance activates energy metabolism and tumor cell migration, contributing to lymph node metastasis. The Listeria abundance was higher in the lymph nodes than in tumor lesions and adjacent tissues of those in the Chinese LUAD cohort [Supplementary Table 5, https://links.lww.com/CM9/C727]. Gene set enrichment analysis confirmed the marked upregulation of the biomarkers of LUAD such as Myc targets, E2F targets, G2M checkpoints, mitotic spindles, and glycolysis in tumors and lymph nodes. Our findings show that Listeria acts as a core node in tumor and metastatic lymph node networks. Moreover, Listeria abundance negatively correlated with the Myc and E2F target gene sets in metastatic lymph nodes [Supplementary Figure 6A–F, https://links.lww.com/CM9/C726]. Thus, the diverse correlations derived of Listeria and the biomarkers of different tumor niches further increase our understanding of the relevant mechanisms through which LUAD occurs. The interactions between commensal bacteria in the lower airways, such as Veillonella and Streptococcus, and the upregulation of specific pathways such as IL17 and ERK/PI3K, may contribute to LUAD progression.[2,3] However, evidence is limited on the mechanism through which microbiota promote lymph node metastasis. Our study suggests that lymph node metastatic LUAD is characterized by interactions involving the microbiota and upregulated biomarkers of tumor migration. We identified distinct relationships between the microbiota in primary tumors, adjacent tissues, and lymph node tissues in a LUAD cohort of patients in China. Listeria may act as modulators of the tumor ecosystem, similar to the intratumor microbiota, through promoting breast cancer metastasis.[4] Certain beneficial bacteria can trigger the production of interleukin (IL)-1β and IL-23 by myeloid cells, leading to the growth and activation of specific immune cells, Vγ6/Vδ1/γδT cells, which produce IL-17, promoting inflammation.[5] Microbiota biomarkers, such as Aeromonas and Listeria, can reshape the immune environment of the host. The strong connection between Aeromonas and CD8+T cells also supports the idea that microbes induce tumor progression. The connection between Listeria and M2 macrophages suggests that the microbiota reshapes innate immunity and induces lymph node metastasis. Furthermore, Listeria could be engineered to act as a delivery system to trigger specific immune responses against tumors. CD8+T-cell tolerance resembles stem-like cells from tumor-draining lymph nodes and was disrupted by Myc-mediated protein translation induced using Listeria monocytogenes injection.[6] Thus, the intratumor microbiota could be a target for modifying the tumor environment. Lymph node metastasis is a crucial factor for distant metastasis in LUAD. In general, lymph node metastasis involves four main steps: vascular endothelial growth factor (VEGF)-C/VEGF-D/VEGFR-3 axis-regulated tumor-associated lymph angiogenesis in the vicinity of tumors, chemokine-assisted tumor cell migration to lymphatic vessels, tumor-induced lymph angiogenesis in lymph nodes, and morphology and function changes in the blood and lymphatic vessels in lymph nodes. Additionally, the intestinal microbiota stimulates the secretion of Paneth cells to regulate the lymphatic vessel proliferation in mesenteric lymph nodes, which is a prerequisite for lymph node metastasis. The regulatory effect of the microbiota on tumor migration, which is achieved via affecting energy metabolism and EMT pathways, was further investigated. Our results emphasize the importance of microbial homeostasis in lymph node metastasis. We were unable to determine the intratumoral-microbiota-specific mechanisms underlying metastasis and immune environment remodeling given the retrospective nature of our study. Preclinical models and prospective studies should be conducted to identify the role of microbiota in activating antitumor immunity. Conflicts of interest None.