Wright Lecture
| Wednesday, June 24, 2026 |
| 4:30 PM - 5:15 PM |
| LT1 |
Speaker
Dr Vinaya Gunasri
NIHR Clinical Lecturer
University College London
Colorectal carcinogenesis: immune microenvironment, tumour evolution and the microbiome
Abstract
Colorectal cancer (CRC) develops through a well-characterised adenoma-carcinoma sequence, yet changes in the tumour microenvironment and microbiome across this transition remain incompletely understood. This lecture presents findings from two interconnected research programmes.
The first examined immune escape mechanisms in CRC using a multiregion, multi-omic dataset of matched genome, transcriptome, and chromatin accessibility profiles from 495 single glands (29 CRCs), supplemented with high-resolution spatially resolved neoantigen sequencing and multiplexed imaging of the tumour microenvironment from 82 microbiopsies within 11 CRCs. Cytotoxic T cell depletion and regulatory T cell enrichment were identified as global features of the tumour microenvironment (TME), with increased spatial segregation between epithelial and cytotoxic T cells in cancer versus normal tissue. Somatic chromatin accessibility loss at antigen-presenting genes and selective silencing of clonal neoantigens were identified as mechanisms underpinning immune evasion. These findings indicate that immune evasion in CRC is an early, spatially uniform event consistent with a 'Big Bang' model of tumour evolution (Lakatos* & Gunasri* et al., Nature Genetics, 2025).
The second programme presents preliminary data from 6 patients characterising the intra-tumoural microbiome across the normal-polyp-cancer transition using multi-platform sequencing (16S full-length, 16S V3-V4, and shotgun metagenomics). Following contaminant removal (170 genera excluded), strong cross-platform concordance validated the analytical approach. Lesion type significantly predicted microbial community composition after controlling for patient and anatomical site, with cancer versus polyp representing the most significant pairwise difference. Cancer samples showed reduced alpha diversity, with enrichment of known pathobionts including Fusobacterium nucleatum compared to normal mucosa.
Together, these findings support a model in which immune remodelling and microbial dysbiosis are early, convergent features of colorectal carcinogenesis. Future work will integrate spatial transcriptomics with microbiome data to interrogate the mechanistic links between specific bacteria and immune exclusion.
The first examined immune escape mechanisms in CRC using a multiregion, multi-omic dataset of matched genome, transcriptome, and chromatin accessibility profiles from 495 single glands (29 CRCs), supplemented with high-resolution spatially resolved neoantigen sequencing and multiplexed imaging of the tumour microenvironment from 82 microbiopsies within 11 CRCs. Cytotoxic T cell depletion and regulatory T cell enrichment were identified as global features of the tumour microenvironment (TME), with increased spatial segregation between epithelial and cytotoxic T cells in cancer versus normal tissue. Somatic chromatin accessibility loss at antigen-presenting genes and selective silencing of clonal neoantigens were identified as mechanisms underpinning immune evasion. These findings indicate that immune evasion in CRC is an early, spatially uniform event consistent with a 'Big Bang' model of tumour evolution (Lakatos* & Gunasri* et al., Nature Genetics, 2025).
The second programme presents preliminary data from 6 patients characterising the intra-tumoural microbiome across the normal-polyp-cancer transition using multi-platform sequencing (16S full-length, 16S V3-V4, and shotgun metagenomics). Following contaminant removal (170 genera excluded), strong cross-platform concordance validated the analytical approach. Lesion type significantly predicted microbial community composition after controlling for patient and anatomical site, with cancer versus polyp representing the most significant pairwise difference. Cancer samples showed reduced alpha diversity, with enrichment of known pathobionts including Fusobacterium nucleatum compared to normal mucosa.
Together, these findings support a model in which immune remodelling and microbial dysbiosis are early, convergent features of colorectal carcinogenesis. Future work will integrate spatial transcriptomics with microbiome data to interrogate the mechanistic links between specific bacteria and immune exclusion.
Chair
Mark Arends
Head Of Pathology
University Of Edinburgh
