Multi-Omics-Based Anticancer Microbiome Research Using Urine and Tissue

Abstract

The role of the microbiome in human health, especially concerning various cancers, has gained increasing attention. Recent studies have revealed that genitourinary organs previously considered sterile, such as the urinary bladder and uterus, possess unique microbiomes. However, limited research has focused on the urine and tissue microbiomes in genitourinary malignancies. In this study, we demonstrate that tissue and urine microbiomes in bladder cancer (BCa) are closely associated in terms of microbial composition and metabolic pathways. Furthermore, we reveal that urine and tissue microbiomes in BCa are significantly linked to antitumor activity, treatment response, and disease prognosis. We analyzed urine and tissue microbiomes in BCa patients using 16S rRNA gene sequencing, RNA sequencing, and shotgun metagenomics sequencing. First, we conducted 16S rRNA gene sequencing on urine samples from patients with non-muscle-invasive bladder cancer (NMIBC) and benign prostatic hyperplasia (BPH) as controls. Our analysis revealed distinct bacterial compositions between benign and malignant samples, suggesting a potential role for the toluene degradation pathway in reducing BCa risk. Responders to BCG therapy exhibited unique microbial compositions with increased quinolone synthesis and a higher abundance of two Bifidobacterium species, BB_003 and B. longum, both of which were associated with prolonged recurrence-free survival (RFS). Additionally, we developed highly accurate models to predict malignancy and BCG response. We next performed RNA sequencing on tissue samples from NMIBC patients. Non-recurrence was associated with higher levels of Lactococcus lactis and BB_003, with BB_003 linked to extended RFS. In silico analyses indicated that BB_003 plays a role in a pro-inflammatory cascade involving macrophages and interferons. In vitro, BB_003-β strain demonstrated antitumor potential by upregulating pro-inflammatory genes and increasing levels of effector CD8 T cells and M2 macrophages. In vivo, direct administration of live BB_003-β into tumors reduced tumor growth, increased CD8 T cell populations in the tumor, and enhanced the presence of cDC1 and M1 macrophages in tumor-draining lymph nodes. Intravesical administration of live bacteria notably reduced bladder weight and increased interferon gene expression. We further analyzed urine, tissue, and stool microbiomes using shotgun metagenomics sequencing to examine the relationships between these sample types. Our findings reveal a close association between urine and tissue microbiomes in terms of microbial composition and metabolic pathways. Spearman’s correlation analysis showed that the tissue and urine (TU) microbiomes had a stronger correlation and lower variability than either the tissue and stool (TS) or urine and stool (US) pairs. In addition, we developed a predictive model for tissue microbial composition based on the urine microbiome. In conclusion, we successfully inferred several beneficial microbial species and their mechanisms of action through comprehensive bioinformatics analyses across multiple omics datasets. These findings were validated in experimental studies using cell lines and mouse models, leading to the development of a prediction model for malignancy and BCG treatment response, as well as a novel, non-invasive model for tissue microbiome composition. Our research on the microbiomes of BCa underscores a close association between these microbiomes and key clinical outcomes, including antitumor activity, treatment response, and disease prognosis.