Nomination And Exploration Of XPO1 Inhibitors For The Treatment Of Triple-Negative Breast Cancer

Abstract

Triple-negative breast cancer (TNBC) is widely recognized as the most aggressive subtype of breast carcinoma, where high rates of relapse, poor overall prognosis, and a lack of effective targeted therapies are commonly present. The considerable inter- and intratumoral heterogeneity observed in TNBC further complicates treatment and emphasizes the urgent need for rationally designed therapeutic strategies. In this dissertation, I nominated the class of exportin 1 (XPO1) inhibitors as a potential therapeutic compound for use in TNBC. XPO1 plays a vital role in the nuclear-cytoplasmic transport of over two-hundred different proteins and RNA species. In addition, I utilized a framework that considers tumor heterogeneity to guide the rational design of drug combinations. Based on this approach, I proposed enhancing XPO1 inhibition by using parthenolide in combination.I applied a computational drug repurposing pipeline that integrates large-scale drug response datasets with transcriptomic profiles from patient tumors. Through this approach, I identified the class of XPO1 inhibitors as compounds with preferential sensitivity in TNBC compared to other breast cancer subtypes. I then validated these findings experimentally using a diverse panel of TNBC cell lines, which demonstrated robust sensitivity to selinexor, an FDA-approved XPO1 inhibitor. Follow-up mechanistic investigations revealed that selinexor induces apoptosis in TNBC cells by suppressing NF-κB signaling, via nuclear retention of NFKBIA. Given the challenges associated with monotherapy in the context of heterogeneous cancers such as TNBC, I next employed computational modeling in conjunction with experimental validation to identify rational drug partners for selinexor. Rather than relying solely on traditional definitions of synergy or additivity, I prioritized drug combinations based on the principle of independent drug action (IDA). To more accurately reflect the physiological diversity of TNBC, I developed a mixed co-culture model by hierarchically clustering transcriptomic profiles from patient tumor samples with those of cancer cell lines, thereby ensuring that the model captured the full spectrum of TNBC heterogeneity. Through this approach, parthenolide, a modulator of the NF-κB pathway, was identified as a promising candidate for combination therapy. In the mixed co-culture system, the combination of selinexor and parthenolide resulted in significantly greater growth inhibition than either drug alone. Further analyses revealed that the degree of synergy was dependent on TNBC subtype, highlighting the importance of accounting for intertumoral diversity when designing combination treatment regimens. In summary, these findings support XPO1 inhibitors as a promising therapeutic strategy for TNBC and establish proof of concept for a computationally guided, heterogeneity-informed approach to drug combination discovery. By integrating transcriptomic profiling, computational modeling, and experimental validation, this dissertation provides both mechanistic insight and a translational framework for the development of more durable therapies in TNBC. Generally applied, this strategy demonstrates how transcriptomic profiles and molecular heterogeneity can be leveraged to inform therapeutic innovation in oncology.