(W-063) Cell-line Specific Network Modeling to Assess Differential Signal Transduction of Oxaliplatin-based Drug Combinations for Colorectal Cancer
Wednesday, November 13, 2024
7:00 AM – 1:45 PM MST
Angelica Squarzoni, N/A – PhD Candidate, Experimental Neurology Unit, University of Milano-Bicocca; Guido Cavaletti, MD – Professor and Head, Experimental Neurology Unit, University of Milano-Bicocca; Donald Mager, PharmD, PhD – Professor and Chair, Department of Pharmaceutical Sciences, University at Buffalo, Enhanced Pharmacodynamics
PhD Candidate Department of Pharmaceutical Sciences, University at Buffalo, United States
Disclosure(s):
Se Jin Kim, PharmD: No financial relationships to disclose
Objectives: Oxaliplatin is a platinum-based chemotherapy agent for the management of colorectal cancer (CRC) in combination regimens with other chemotherapeutic agents. Oxaliplatin-induced peripheral neuropathy (OIPN) is a dose-limiting toxicity with a prevalence of 85–95% and 10–15% for acute and chronic OIPN1. There are no approved agents for OIPN prevention, and duloxetine is the only available pharmacological treatment, which leads to dose reduction, poor adherence, and termination of therapy. Histone deacetylase (HDAC) inhibitors can prevent and reverse neuropathic pain after exposure to chemotherapy and potentiate cytotoxic effects, but the mechanism of action is unknown2,3. The aim of this study was to utilize a network modeling approach to assess differences in signal transduction for synergistic combinations of oxaliplatin and HDAC inhibitors in CRC cell lines.
Methods: 3 CRC cell lines (Caco2, HCT15, HT29) were incubated with oxaliplatin as a single agent and as a combination with each of three HDAC inhibitors (vorinostat, romidepsin, and SW100) for 24 hours. The extent of cell kill and protein expression of 16 proteins were measured at time 0 and 24 hours after exposure. A prior CRC logic model was modified to include key proteins of interest4. Proteomic data were used construct cell-line specific models through CellNOptR, which estimates parameters through an enhanced scatter search and dynamic hill climbing algorithms5. Estimated parameters were fixed to simulate the time-course of signaling proteins using Odefy, and the area under the curve (AUC) of each protein was calculated6.
Results: Oxaliplatin exhibited the most synergistic cell killing effect across all cell lines with SW100. Model parameters that describe the signaling activity of KRAS were statistically correlated with the fold change (FC) in cell killing for oxaliplatin combinations. Direct upstream regulators of KRAS (EGFR, HGF, and IGF1) and their interactions with KRAS were statistically greater than KRAS and its interaction with direct downstream species (BRAF, M3K1, and PI3K). A smaller AUC for AKT and mTOR were associated with greater FC in cell kill (R > 0.7, p-value < 0.05), which is consistent with literature experiments where a decrease in signaling was observed in the MAPK/ERK pathways in colon cancer after HDAC KO7. AUC of KRAS was better correlated with the FC in cell killing than the raw proteomic data (R=0.66, p-value=0.054 vs. R=0.42, p-value=0.26).
Conclusion: Our cell-line specific CRC logic models identified differential KRAS signaling as a potential mechanism for the enhanced cell kill efficacy of oxaliplatin and SW100. This study highlights the capabilities of generating cell-line specific logic models to compare signaling transduction and potential mechanisms of drug interaction. Such models may inform target selection for additional combination regimens for CRC and inadequate responders to current chemotherapy options.
