UC San Diego

Proteasome inhibitors (PIs) have recently emerged as a therapeutic strategy in cancer chemotherapy with the FDA approval of bortezomib. The marine actinobacterium Salinispora tropica, discovered in sediments off the Bahamas, produces a potent natural product PI, salinosporamide A (NPI-0052 or marizomib), which is now in clinical trials for the treatment of multiple myeloma. A chlorine atom, incorporated via the novel polyketide synthase extender unit chloroethylmalonyl-CoA, confers highly potent and irreversible inhibition of the eukaryotic 20S proteasome. Herein I report the in vitro characterization of one enzyme, the short-chain dehydrogenase/reductase SalM, responsible for the oxidation of 5-chloro-5-deoxy-D-ribose to 5-chloro-5-deoxy -D-ribono-gamma-lactone en route to chloroethylmalonyl- CoA. Using heterologously produced SalM, a sensitive, real -time ¹³C NMR assay was developed to monitor transient product formation followed by spontaneous lactone hydrolysis. SalM was determined to have an atypical divalent cation dependence (Mg²⁺, Mn²⁺ or Ca²⁺) and to oxidize tetrose or pentose furanoses with hydroxy stereochemistry equivalent to that of D-ribose, making it the first reported stereospecific non-phosphorylated ribose-1-dehydrogenase. Additionally, I explored the question of PI self-resistance in S. tropica as actinobacteria possess 20S proteasome machinery. A secondary catalytic [beta]-subunit (SalI) encoded adjacent to the salinosporamide biosynthetic gene cluster was characterized by heterologous expression and in vitro assaying of the [alpha]/SalI complex. An altered proteolytic specificity and 30-fold resistance toward salinosporamide A inhibition was demonstrated for the [alpha]/SalI complex relative to the housekeeping [alpha]/ [beta]₁ complex. Sequence comparison of these two [beta]- subunits revealed two mutations, M45F and A49V, which likely conferred resistance. Mutational analysis demonstrated that the A49V mutation of SalI is partially responsible for resistance which correlates to identical mutations observed in bortezomib resistant human cancer cell lines. The [alpha]/SalI complex was also cross- resistant to bortezomib and to salinosporamide analogs, suggesting that S1 binding pocket mutation leads to resistance against all proteasome [beta]-subunit inhibitors. As bortezomib therapy is plagued by intrinsic and acquired resistance, it is critical to determine if salinosporamide A will suffer the same fate. My analysis suggests that bortezomib resistant cancer cell lines are likely cross-resistant to salinosporamide A. Moreover, these results suggest that self-resistance to natural PIs may predict clinical outcomes