Characterization and Targeting of the Tumor Microenvironment in Colorectal Cancer

Colorectal cancer (CRC) remains a health burden globally, accounting for 10% of cancer incidence and is the second leading cause of cancer-related deaths. Current standard therapies for CRC patients typically involve surgery, either alone or combined with chemotherapy or targeted treatments, depending on the disease stage. While most patients show an initial response, relapse often occurs within a few months and over 50% of CRC patients present with metastasis at diagnosis. These challenges underscore a critical unmet need for more effective treatment options. Notably, only 5% of drugs that demonstrate efficacy in preclinical studies ultimately achieve clinical approval after phase III trials, highlighting the urgent demand for innovative therapeutic strategies. To address this, current efforts are increasingly focused on three critical areas: in-depth understanding of the tumor microenvironment (TME), innovative drug discovery approaches, and the development of physiologically relevant preclinical models that better recapitulate human disease. To this end, we constructed a comprehensive single-cell transcriptome atlas of over 400,000 cells from 137 patients, covering both normal and CRC tissues. We then integrated these data with Xenium spatial transcriptomics from 15 CRC patients, enabling the spatial profiling of more than 10 million cells. This revealed that three of five cancer-associated fibroblast (CAF) subsets likely originate from normal colonic fibroblasts, while the remaining two are cancer-specific and may serve as therapeutic targets with limited impact on normal fibroblast populations. Concurrently, we developed a high-throughput screening (HTS) assay using 3D CRC models, allowing for the evaluation of over 1,000 drugs across two CRC cell lines and identifying more than 40 hits per line. Six promising compounds were selected for further validation in preclinical CRC models. Finally, we established an orthotopic CRC mouse model via submucosal organoid injection that support in vivo evaluation of tumor growth and therapeutic response within a physiologically relevant context. Together, these approaches represent a multi-faceted strategy to accelerate therapeutic discovery and development in CRC. By integrating deep TME analysis, cutting-edge drug screening, and advanced preclinical modeling, this work lays the groundwork for translating experimental findings into impactful clinical therapies. Further investigation and refinement of these strategies hold great promise for improving the prognosis and treatment outcomes for CRC patients.