SUMMARY. Salivary gland cancer (SGC) is an orphan disease for which no targeted therapies are approved.
SGCs are divided into histotypes, the most common being adenoid cystic carcinoma (ACC) and mucoepidermoid
carcinoma (MEC). We have generated one of the largest reported SGC PDX banks including major histotypes
like ACC and the first reported PDX models of MEC. Our SGC PDX bank includes several cases from the same
patient, collected after subsequent relapses, which has allowed an exploration of the acquisition of oncogenic
gene events. Similar to other reports, our SGC models had low mutation burden by whole exome sequencing,
and RNA-seq analysis identified known (MYB-NFIB), novel (NFIB-MTFR2), and even dual gene fusions in ACC
cases that gave rise to oncogenic fusion protein products. Other genetic events in SGC clustered in the PI3K
and mTOR pathways, which impinge upon protein synthesis. Both MEC and ACC overexpress key transcription
factors such as Myc or SOX2 that dictated tumorigenicity and growth in SGC. Overall, we propose that protein
synthesis is an unexplored target in SGC. SVC112 is a small molecule that inhibits protein synthesis at the
elongation step by inhibiting eEF2. In our seminal studies where SVC112 was first reported, protein synthesis
and growth inhibition were associated; SVC112 had greater effect on cancer over non-cancer cells; SVC112
depleted SOX2, Myc, and Cyclin D1, and arrested growth in vivo. In both ACC and MEC cell lines we found that
SVC112 inhibits both native Myb and the protein products of MYB fusions and key proteins like Myc. SVC112
inhibited proliferation and sphere formation in both ACC and MEC in vitro assays, and had notable single agent
activity (including actual tumor shrinkage) in 2 ACC PDX, one with a MYB fusion and another one showing non-
fusion mediated Myb upregulation. We propose to study the role of oncogenic fusions in SGC, the mechanism
of action of SVC112 in SGC, and the efficacy of SVC112 in complex SGC models. First, we will catalogue and
prioritize MYB fusions from large patient and PDX cohorts. Then, to understand key genetic events we will insert
fusions in non-cancer cell lines, and deploy CRISPR and lentiviral vectors in fusion-and non-fusion-bearing SGC
cell lines, respectively; we will also test SVC112 inhibition to add an additional layer of testing of the impact of
protein modulation. The hypothesis that SVC112’s selective effect is due to selective depletion of key proteins
will be tested by ribosome profiling (to identify mRNA targets) and proteomics analysis (to identify proteins
targets). Lastly, we will test the in vivo efficacy of SVC112 using both native SGC PDX models bearing MYB and
other fusions and wild-tpy, as well as in vivo models with the CRISPR-depleted MYB fusions strains. Overall our
goal is to examine dependence on fusion events vs protein overexpression by other molecular mechanisms for
SVC112 susceptibility. This project will propel the translation of SVC112, a new drug for SGC discovered in
Colorado, by dissecting the basis for its effect and therapeutic window, and identifying its efficacy in advanced
SGC models with a plan that enables identification of which subset of SGC patients may derive more benefit.