Modeling and Therapeutic Development for the Tuberous Sclerosis Related Neoplasm Lymphangioleiomyomatosis

Abstract

The multisystemic tumors characteristic of the monogenic neoplastic diseases, tuberous sclerosis complex (TSC) and lymphangioleiomyomatosis (LAM), share common signaling aberrations upon the loss of heterozygosity in either the TSC1 or TSC2 genes. However, their physical manifestations are vastly different and can generally be classified as being either neurological (TSC) or mesenchymal (TSC & LAM; referred to herein as LAM for simplicity) in origin. In this study, I present a comprehensive stem cell model of LAM utilizing multiple TSC2 knockout (TSC2-/-) pluripotent stem cell lines differentiated to the putative cell of origin for mesenchymal tumors, neural crest cells (NCCs). TSC2-/- NCCs faithfully recapitulate LAM phenotypes and temporal RNA-seq analysis of neural and neural crest differentiation was performed to model disease pathogenesis. Analysis revealed immediate activation of stress response signaling resulting in protein aggregation and lysosome and autophagosome accumulation upon neuralization in TSC2-/- cells. This resulted in acute and lasting effects specific to neural progenitor cells (NPCs), that are transient and ameliorated in NCCs. These lineage-specific effects resulted in selective sensitization of NPCs to cell death via proteasome inhibition, suggesting a potential therapeutic avenue for neurological TSC, but not LAM. Thus, a genome-wide CRISPR knockout screen was performed in TSC2-/- NCCs. Analysis of synthetic lethal genes reveals pathways previously targeted for LAM, but provides gene-level resolution to the vulnerable nodes within these pathways. Importantly, 18 novel gene targets were identified that display synthetic lethality to TSC2-/- cells with high specificity. 3 genes within this list were targetable using commercially available small molecule inhibitors, one of which, FGFR1, shows highly selective lethal targeting of TSC2-/- NCCs. Importantly, this model system, paired with the expansive resource of transcriptomic and synthetic lethal data, serves as a foundation for the development of next generation treatment strategies for LAM, and potentially the entire spectrum of TSC manifestations.