Colorectal cancer (CRC) is among the most prevalent cancer types and remains a leading cause of cancer-related mortality. CRC develops through a multistep process in which the accumulation of oncogenic mutations drives the expansion of malignant cells and their progression to invasive carcinoma that can metastasize to distant organs. Recent genomic and gene expression profiling studies have linked cellular heterogeneity to aggressive phenotypes and have highlighted how interactions between cancer cells and the tumor microenvironment (TME) influence metastatic initiation and seeding. However, how cellular interactions contribute to the emergence of distinct cellular states and the initiation of metastasis remains unclear. To address this, we developed an orthotopic mouse perturbation screening platform to systematically assess the interplay between cancer cells and the TME.

We used Pro-Code, a platform that couples CRISPR-Cas9 guide RNAs (gRNAs) to protein epitope tags, to perform a small-scale (75×) in vivo CRISPR screen with optical readouts. Cas9-expressing mouse AKP (Apc-/-KrasG12D/+p53-/-) CRC organoids were transduced with lentiviruses carrying 75 Pro-Code constructs targeting 20 commonly mutated driver genes in CRC metastases. Edited cells were injected orthotopically into mice using a colonoscopy-guided injection method and harvested after 4 or 6 weeks. Colons containing primary tumors were fixed, cryosectioned, and subjected to cyclic immunofluorescence using the Comet™ platform. Nine epitope tags, together with 16 additional markers, were imaged to decode sgRNA identities and to identify distinct cell types within the TME. Multiplexed images were then processed for single-cell segmentation and spatial analysis to quantify significant cellular interactions.

Using multiplexed imaging and downstream analyses across multiple tumors, we observed significant differences in tumor morphology (eg. Differentiated vs tumor budding), clonality (eg, mono vs cooperative), and TME composition (eg. Hot vs cold) associated with specific mutations and metastatic outcome. These findings suggest that distinct mechanisms of tumor–microenvironment crosstalk may promote metastasis initiation. We are currently investigating these mechanisms using spatial and single-cell transcriptomic profiling, as well as targeted perturbations of specific interactions enriched in metastatic mutants.

We established an orthotopic in vivo CRISPR screening platform with a multiplexed imaging-based readout. This approach enables systematic assessment of cancer cell–TME interactions during metastasis initiation. Characterizing and perturbing key interactions may reveal novel therapeutic targets for preventing metastasis.