Since its establishment in 2015, the transcriptomics-based consensus molecular subtype (CMS) classification has unified our understanding of colorectal cancer (CRC). Each of the four CMS exhibited distinctive high-level molecular signatures that correlated well with prognosis and treatment response. Nonetheless, many key aspects of CRC progression and intra-subtype heterogeneity remain unresolved. This is partly because the bulk transcriptomic data used to define CMS contain substantial interference from non-tumor cells. Here, we propose a concise panel of 62 genes that not only accurately recapitulates all key characteristics of the four original CMS but also identifies three additional subpopulations with unique molecular signatures. Validation on independent cohorts confirm that the new CMS4 intra-subtypes coincide with single-cell-derived intrinsic subtypes and that the panel consists of many immune cell type markers that can capture the status of tumor microenvironment. Furthermore, a 2D embedding of CMS structure based on the proposed gene panel provides a high-resolution view of the functional pathways and cell type markers that underlie each CMS intra-subtype and the continuous progression from CMS2 to CMS4 subtypes. Our gene panel and 2D visualization refined the delineation of CRC subtypes and could aid further discovery of molecular mechanisms in CRC. Implications: Well-selected gene panel and representation can capture both the continuum of cancer cell states and tumor microenvironment status.

Background: LIN28B is functionally driving malignant transformation and relevance to the worse disease outcomes by promoting cancer aggressiveness. However, a typical role of LIN28B in cholangiocarcinoma (CCA) is primarily unknown. In this study, the tumorigenic potential of LIN28B in the cholangiocyte context was investigated. Methods: Stable LIN28B expression in MMNK-1 cells was generated by infecting with retrovirus-containing LIN28B gene. LIN28B-overexpressing cells were further validated the amount of released cytokines by using human cytokine arrays. After treatment of chemo-drugs, cell viability was subsequently measured using MTT assay. Aldehyde dehydrogenase (ALDH) activity was determined using ALDEFLUOR assay Kit and analyzed by flow cytometry. The mRNA and protein expression levels were respectively assayed by RT-qPCR and western blot. Results: Cytokine release results showed that numerous inflammatory cytokines-chemokines related to cancer initiation and development, such as IL-8, IL-6, VEGF, MCP1, TNF-α were significantly increased in LIN28B-overexpressing MMNK-1 cells. Drug sensitivity test showed that LIN28B-overexpressing MMNK-1 treated cells had a high percentage of cell viability compared to MMNK-1-control treated cells. Activity and expression level of a cancer stem cell marker, ALDH was significantly elevated in LIN28B-overexpressing MMNK-1 cells. Moreover, the activation of an oncogenic signaling pathway, signal transducer and activator of transcription 3 (STAT3) was enhanced in LIN28B-overexpressing MMNK-1 cells. Whereas, growth capacity of LIN28B-overexpressing MMNK-1 cells was found to be reduced in STAT3 inhibition. Conclusion: LIN28B can regulate the inflammatory response and resistance to chemotherapy of cholangiocytes through modulation of STAT3 signaling pathway.A recent study suggests that activated cholangiocytes can be induced by regulation of LIN28B/STAT3 pathway and this may partially contribute to the initiating CCA. Here, LIN28B and its downstream signaling could be considered as an attractive therapeutic target in patients with CCA.

<p>S1. Expression profiles of cell type-specific markers. S2. Example of a subnetwork consisting of 1000 samples from the original CMS dataset. S3. Robustness analysis of the 2-round network clustering method. S4. Hierarchical clustering of CMS2 samples. S5. Hierarchical clustering of CMS4 samples. S6. Agreement between original CMS and new intra-CMS subtypes. S7. Comparison of our gene panel with previously published panels for CRC. S8. Expression of 62-gene panel on CMS2/iCMS2 samples in the SG-Bulk cohort. S9. Expression of 62-gene panel on every epithelial cell from four selected patients from the iCMS cohort. S10. Progressive changes in inflammatory, immune, and proliferative pathway activities from CMS2 to CMS4 intra-subtypes. S11. Expression profiles of cellular senescence markers. S12. Mutation profiles and rates for KRAS, BRAF and PIK3CA. S13. Progressive changes in inflammatory, immune, and proliferative pathway activities between CMS1, CMS3, and CMS4 intra-subtypes. S14. Progressive changes in WNT, EMT, MAPK, and VEGF sub-pathway modules. S15. Expression profiles for JAK-STAT and amino sugar and nucleotide sugar metabolism pathways. S16. Confusion matrix for the classification of 7 intra-subtypes with an artificial neural network model.</p>

<p>Mutation rate and MSI status statistics for each subtype</p>

<p>List of 62 genes of our proposed panel</p>