Colorectal cancer (CRC) shows a dramatic drop in the five-year survival rate from 90% for localized tumors to 14% for metastatic CRC (mCRC). Our lab has identified LIN28B, an RNA-binding protein overexpressed in 30% of CRCs, as an oncogene that promotes CRC metastasis. However, the mechanisms through which LIN28B promotes these effects are unknown. The phosphatidylinositol 3-kinase (PI3K)/AKT signaling pathway, altered in 50-70% of CRCs and often harboring PIK3CA mutations (20-30% of cases), is linked to poor prognosis. Within this signaling cascade, the ribosomal protein S6 kinase (S6K) plays a key role in protein synthesis and cell growth. Despite the prevalence of PIK3CA mutations, targeted therapies for PIK3CA-mutant CRC are lacking. Alpelisib, a PI3Kα-specific inhibitor approved for a subset of PIK3CA-mutant breast cancers, remains underexplored in CRC. We hypothesize that LIN28B-mediated CRC metastasis is facilitated by the activation of the PI3K/AKT signaling pathway, presenting a potential target for Alpelisib and other pathway inhibitors.
Induction of LIN28B expression in CRC cell lines activated the PI3K/AKT pathway as measured by AKT phosphorylation array, and injection of these cell lines into the portal vein of immunocompromised mice promoted liver metastasis. Treatment with SC79, a pan-AKT activator, enhanced CRC cell migration and invasiveness. Consistent with in vitro activation, VillinCre;R26Pik3ca transgenic mice, which express a constitutively active mutant form of PI3Kα in the intestinal epithelium, exhibited crypt hyperplasia and an amplified proclivity for tumorigenesis. Similarly, VillinCreERT;R26Pik3ca mice, in which the mutant form of PI3Kα is inducible, developed invasive colonic tumors and liver metastases when treated with the carcinogen azoxymethane. Notably, Alpelisib inhibited LIN28B-driven CRC progression in cell lines, transgenic mice-derived organoids, mCRC mouse models, and CRC patient-derived organoids (PDOs) from liver metastases. The inhibitory effects with Alpelisib were mirrored by S6K inhibition using LY2584702. Combining low-dose Alpelisib with LY2584702 synergistically restricted the growth of both primary tumor- and matched liver metastasis-derived PIK3CA-mutant CRC PDOs. Immunohistochemical analysis of tissue microarrays, comprising samples from normal colon, primary tumors, and liver metastases of 60 CRC patients, confirmed the correlation between PI3K-S6K signaling and disease progression.
Our data show that the PI3K-S6K signaling is crucial in LIN28B-driven CRC metastasis. Moreover, we introduce the first genetically engineered mouse model with colonic tumors that progress to form liver metastases within an intact immune system. This work opens up promising treatment avenues for PIK3CA-mutant CRC patients and emphasizes the effectiveness of combination therapy in improving anti-metastatic outcomes.

Schematic diagram illustrating the mouse models and translational methods used to study the role of PI3K/AKT-S6K signaling on CRC metastasis. VillinCreERT;R26Pik3ca transgenic mice were used to examine the effects of PI3Kα activation on metastasis of colonic tumors. A portal vein injection model and patient-derived organoids (PDOs) were used to assess the efficacy of Alpelisib and LY2584702 treatments. Tissue microarrays confirmed the association between PI3K-S6K signaling and CRC progression. Image created with BioRender.com.