Interrogating the molecular underpinnings of metastatic colonization

Division Of Basic Sciences - Nci

Investigators

Linked publications & trials

Abstract

Sub-Project 1. I used 318 patient samples to demonstrate specific cytoplasmic accumulation of XPO7 within tumor cells in a subset of patients with CCA and further verified this localization pattern in a panel of CCA cell lines. I then used the cells lines to show that XPO7 binds to and stabilizes Ste20-like kinase (SLK), which in turn activates downstream oncogenic signaling pathways, including the PI 3-Kinase-AKT pathway, and subsequently reverted back to tumor samples to verify this relationship. Given the absence of specific SLK inhibitors, we next profiled the kinome-specificity of 250 inhibitors to identify those with activity against SLK. Based on that data, I selected one of those agents, tivozanib, for further study. Using murine xenografts and CCA tumor from a patient in the SMART system (see Project #3), I demonstrated that tivozanib treatment resulted in tumor shrinkage/tumor cell killing and substantial morphologic alterations similar to those observed with XPO7 or SLK knockdown. Based upon these results, I have successful secured tivozanib from Aveo Oncology through a CRADA mechanism and have opened a Phase I/II trial (NCT 04645160) for patients with CCA. Sub-Project 2. Genetic screening identified the amino acid sensor GCN1 as a driver of metastatic outgrowth in the liver by activating a dual translational and a transcriptional gene expression program. While the translational program activates genes involved in nutrient acquisition and intrinsic amino acid biosynthesis in an ATF4-dependent manner, the transcriptional function sustains oncogenic drive by interacting with DNA binding transcription factor HNRNPK, which activates the expression of KRASG12D through DNA quadruplex (G4) unwinding in its promoter. Intriguingly, the GCN1-dependent translational and transcriptional program is negatively regulated by IMPACT, which binds and inhibits GCN1 function. The expression of IMPACT is lost in highly aggressive and metastatic cell lines and inversely correlates with GCN1 pathway activation. Our data suggest that GCN1 signaling is indispensable for in vivo liver metastasis, and agents targeting GCN1 and downstream signaling pathway may have clinical efficacy in the adjuvant setting. Sub-Project 3. I utilized tumor-bearing mesothelium and created an oxygenated perfusion circuit, termed the SMART System (Surgically-resected Mesothelium ContAining UnalteRed Tumor Microenvironment), which I demonstrated maintains tumor with minimal alterations in structure, cellular composition, and transcriptomics. Importantly, cells remain functional and respond to cytokines and drugs. The SMART System is the subject of an EIR. I developed hardware to perform live imaging of the tumor, repurposing our advanced confocal and multiphoton scopes. This technique gives unprecedented insight into tumor biology dynamics. To compliment the SMART system, I conceived and implemented ex vivo perfusion of tumor-bearing liver segments. We initially sought to repurpose commercially available transplant machines, but these lacked real-time adjustment capabilities required for prolonged ex-vivo animation. To circumvent the limitations, I have assembled a team and built a custom liver perfusion system.