Introduction

Colorectal cancer (CRC) is the second leading cause of cancer-related deaths, with approximately one in four patients presenting with metastatic disease at the time of their diagnosis. Therapeutic management of metastatic CRC (mCRC) continues to be a major challenge due to therapy resistance and disease heterogeneity. Given the central role of the MAPK signaling pathway in CRC, multiple MAPK pathway inhibitors (MAPKi) have been developed and have shown promising results in patients. However, clinical trials using MAPKi therapies consistently demonstrate that patients obtain short-lived benefits, usually lasting less than six months before their disease progresses. As such, further research is needed to understand the biology limiting MAPKi therapeutic responses in patients with mCRC.

Transcriptomic data from patients with mCRC treated with dabrafenib (BRAFi), trametinib (MEKi) and sparatlizumab (anti-PD1) was analyzed using gene signatures (n=53) to identify potential mechanisms associated with poor treatment response. Results showed that the SCS signature, which we previously demonstrated to denote stem-like CRC cells, was upregulated in nonresponders (NES=2.71, FDR<0.001), suggesting that tumor propagating cells (TPC) may influence MAPKi therapy response. TPCs are unique stem-like, self-renewing, chemoresistant cells that readily adapt to new conditions and have been linked to poor outcomes in CRC. Therefore, we postulate that the limited response durability of MAPKi-therapy is driven by the enrichment of innately resistant TPCs in mCRC tumors.

Methods

ASCL2 was recently reported to denote stemness activity in CRC cells, and we previously showed that the SCS signature correlated with ASCL2 expression. To validate using ASCL2 as a TPC marker, a metacohort of single cell RNA sequencing (scRNAseq) data from patients with CRC (n=72) was compiled, and ASCL2⁺ cells were found to have higher expression of stemness signatures (p<0.001), confirming that ASCL2 does denote a TPC population in CRC. In order to facilitate transcriptomic identification of ASCL2⁺ TPCs, an ASCL2 gene signature was generated and the resulting expression score was termed ‘ASCL2 TPC Index.’ To further explore the impact of ASCL2⁺ TPCs in MAPKi therapy, we utilize cellular barcoding, scRNAseq, and engineered translational CRC models.

Results

Patient-derived organoid and cell line CRC models revealed that ASCL2 mRNA increased 2-6x when treated with MAPKi therapy. Additionally, MAPKi-treated patient-derived xenografts (PDX) showed a significant upregulation of the ASCL2 TPC Index (NES=1.3, FDR=0.02), indicating that MAPKi therapy alters the TPC compartment. An ASCL2^GFP reporter system was generated and multiple CRC lines were established, all of which displayed significant increases in ASCL2⁺ TPC prevalence when treated with MAPKi therapy (p<0.0001), confirming that MAPKi enriches intratumoral TPCs. Lastly, combining molecularly barcoded PDX models with single cell transcriptomics, to enable comparison of individual cells across condition, we found that MAPKi treated cells had higher ASCL2 TPC Index expression (p=0.01) compared to pre-treatment, thus revealing that MAPKi therapy causes induction of the TPC phenotype.

Conclusion

Overall, we identify TPCs as a mechanism involved in determining MAPKi therapy response, and demonstrate that MAPKi treatment results in the enrichment of ASCL2⁺ TPCs through induction of the TPC phenotype in CRC cells. Further studies are underway to expand on our observations and provide a more comprehensive understanding of the underlying tumor biology impacting TPCs and MAPKi responses. Ultimately, these observations will help guide optimization of future MAPKi therapeutic regimens to prolong response durability and improve overall outcomes for patients with mCRC.

(OV is supported by CPRIT Research Training Grant RP210028.)