#IDIBELLfellows: Blanca Rius & Irene Peñuelas

Integrative Metabolomic and Genomic Analysis Identifies the Polyamine Metabolite Acisoga as Candidate Metabolic Marker for Colorectal Neoplasia in a Colorectal Cancer Screening Cohort
Blanca Rius Sansalvador – Colorectal cancer
Colorectal cancer (CRC) is a leading cancer in terms of incidence and mortality worldwide. Early detection and accurate risk stratification of precursor lesions remain critical challenges. Metabolomics, combined with genetic analysis, provides a potential avenue for diagnostic biomarker identification and better understanding of CRC development. Untargeted metabolomics was performed on serum from 513 individuals spanning 185 controls, 74 low-, 98 intermediate-, and 100 high-risk lesions groups, as well as 56 CRC cases. A total of 1,562 metabolic features were analysed. Linear trend models identified features with differential intensities across diagnostic groups, controlling for multiple testing with false discovery rate (FDR) correction. Genome-wide association studies (GWAS) were performed for significant metabolites, with subsequent SNP annotation and locus visualization. Associations between genetically-predicted metabolite levels and CRC risk were evaluated using summary statistics from a large CRC GWAS meta-analysis. Three metabolic features, identified as the polyamine Acisoga, showed statistically significant trends across diagnostic groups after FDR correction, with Acisoga levels increasing with lesion progression. Metabolite GWAS of the significant hits revealed genome-wide associations mapping to genes such as EWAST1, RSAD2, and CMKLR1 among others. Genetically predicted Acisoga levels were positively associated with CRC risk (p = 0.003), consistent with the observed trend across diagnostic groups. Importantly, the association of Acisoga with CRC was externally validated in the independent cohort EPIC, supporting its robustness as a potential early biomarker. This integrative analysis identified novel metabolites and genetic loci associated with CRC and its precursor lesions. These findings contribute to the understanding of molecular changes during CRC progression and provide a foundation for future biomarker development and risk prediction.

Targeting ROBO and SLIT guidance cues in the immunosuppressive stromal context of pancreatic càncer
Irene Peñuelas Haro – Pancreatic Cancer
Pancreatic ductal adenocarcinoma (PDAC) is highly lethal, partly due to late diagnosis, high recurrence, and a dense, immunosuppressive tumor microenvironment (TME). The SLIT-ROBO signaling pathway, originally described in neural axon guidance, is disrupted in up to 30% of PDAC patients. Apart from operating in the nervous system, recent evidences showed that the SLIT-ROBO pathway also influences key processes in tumor cells and the TME, although little is known of its role in the context of PDAC. Underpinned by bioinformatics analyses, including compartment-specific scRNA-seq and spatial transcriptomics data, as well as primary in vitro cultures, we have deciphered the importance of cell type expression of the different members of the pathway and their interactions in human PDAC samples. We found that SLIT2/3 are mainly expressed by stromal cells, while ROBO1 is variably expressed by tumor cells. Specific stromal SLIT2/3 expression patterns associate with increased infiltration of exhausted CD8⁺ T cells. Functional co-culture experiments support a role for SLIT-ROBO in modulating T-cell behavior. Overall, our results suggest that that SLIT-ROBO pathway play a role in defining the immunosuppressive tumor microenvironment of PDAC tumors. Given the prevalence of alterations in SLIT-ROBO, unveiling the role of the pathway may translate into innovative tailored therapies for a significant proportion of PDAC patients.