PDF, English Restricted access: Repository staff only until 7 January 2026. Login+Download (123MB) | Terms of use

Abstract

The tumor microenvironment (TME) of colorectal cancer (CRC) represents a complex ecosystem, encompassing tumor, immune, stromal, and bacterial components. Prior research highlighted the significant impact of these interactions on CRC clinical manifestation and treatment outcome. However, a substantial gap remains in omics datasets capable of capturing these interactive networks comprehensively, a challenge often attributed to the technical difficulties in achieving balanced recovery of all cellular entities within CRC tissues.

This dissertation introduces innovative experimental methodologies and frameworks designed to generate detailed data from primary CRC resections, encompassing both host tissue and its local bacterial populations. By developing and implementing Host-Microbe RNA-sequencing (HMR-seq) technique in a newly established cohort of 47 CRC patients, I identified transcriptional alterations within CRC tissues, mirrored by compositional shifts in tumor-resident bacteria. Joint host-microbiome analyses identified significant associations between intratumoral bacteria and oncogenic and immune-related gene expression signatures, with distinct configuration in MSI and MSS tumor subtypes.

Furthermore, spatial and single-cell transcriptomics analyses uncovered distinct gene expression landscapes across CRC tissues, especially within immune cell populations. These spatial gene expression variances were shaped by genetic backgrounds and were further delineated by regional differences between patient-matched tumor core and invading front sections. This emphasizes spatial organization as a key factor in intratumoral heterogeneity. Additionally, spatial assessments utilizing m-HybISS for microbiome in situ profiling revealed that intratumoral bacteria also displayed varied localization patterns and densities, influenced by the histopathological characteristics of assayed MSI and MSS tumor sections.

Overall, the methodologies and multi-omics dataset presented in this work offer valuable toolkits for dissecting host-microbe interactions within primary CRC resections. The insights reported herein hold promise for broadening our understanding of host-microbe interactions in CRC and may inform therapeutic strategies for this and potentially other solid tumors influenced by microbiome interactions.