Comprehensive Structure-Activity Relationship Studies of Cepafungin Enabled by Biocatalytic C-H Oxidations

Abstract

The cepafungins are a class of highly potent and selective eukaryotic proteasome inhibitor natural products with potential to treat refractory multiple myeloma and other cancers. The structure-activity relationship of the cepafungins is not fully understood. This Article chronicles the development of a chemoenzymatic approach to cepafungin I. A failed initial route involving derivatization of pipecolic acid prompted us to examine the biosynthetic pathway for the production of 4-hydroxylysine, which culminated in the development of a 9-step synthesis of cepafungin I. An alkyne-tagged analogue enabled chemoproteomic studies of cepafungin and comparison of its effects on global protein expression in human multiple myeloma cells to the clinical drug bortezomib. A preliminary series of analogues elucidated critical determinants of potency in proteasome inhibition. Herein we report the chemoenzymatic syntheses of 13 additional analogues of cepafungin I guided by a proteasome-bound crystal structure, 5 of which are more potent than the natural product. The lead analogue was found to have 7-fold greater proteasome β5 subunit inhibitory activity and has been evaluated against several multiple myeloma and mantle cell lymphoma cell lines in comparison to the clinical drug bortezomib.

Figure 1

Representative structures of syrbactin natural products and clinical proteasome inhibitors.

Scheme 1. (A) Prior Approaches to Diastereomers of 4-Hydroxylysine and (B) First-Generation Synthetic Strategy toward Cepafungin

See the Supporting Information for experimental details.

Scheme 2. (A) Annotation of Glidobactin Biosynthetic Gene Cluster and Product Scope of GlbB and (B) Optimized Chemoenzymatic Route to 1

Figure 2

Preliminary series of cepafungin analogues and their cytotoxicities and comparison to prior syringolin analogues.

Figure 3

Bound crystal structure of cepafungin I at β5 subunit of yeast 20S proteasome. PDB ID: 4FZC. Dashed lines indicate polar contacts. The catalytic β5 N-terminal threonine covalently bonded to the cepafungin macrocycle is labeled “T1”. β5 residues involved in polar contacts are Gly47, Ala49, Thr21, and Asp126. A hydrophobic channel for the tail fragment involves β6 subunit residues Pro127, Val128, Pro104, Tyr106, and Tyr5.

Scheme 3. Synthesis of Cepafungin Analogues 40–51

Figure 4

Biological evaluation of cepafungin (1) and analogues 40–51 in RPMI 8226 cells. (A) Proteasome inhibition screening at 30 nM compound concentration with subunit-specific fluorogenic proteasome substrates Suc-LLVY-AMC (β5) and Ac-RLR-AMC (β2) (n = 3). (B) Cytotoxicity screening at 30 and 100 nM compound concentration (n = 3). (C) EC₅₀ measurement of 50 in comparison to 1 (n = 3). (D) Proteasome subunit IC₅₀ measurements of 50 in comparison to 1 (n = 3). (E) Cytotoxicity comparison of 50 and bortezomib in human MM (RPMI8226, U266, H929) and MCL (JeKo-1, HBL2, Mino) cell lines (n = 3). All error bars represent standard deviation.

Figure 5

Comparative global proteomics study of the mode of action of BTZ and 50. (A) Volcano plots representing the global proteome profile of RPMI 8226 cells treated with BTZ or 50 versus DMSO for 14 h. Data are represented as log2 fold change; dotted lines represent a false discovery rate of 5% and an S₀ of 0.1. Quantification was performed using the LFQ method (n = 6). Red dots indicate proteins with a significant increase in expression levels in response to treatment with both BTZ and 50. (B) Venn diagram representing the overlap in significantly upregulated proteins from panel A. (C) Table of common upregulated proteins between BTZ and 50 compared to DMSO. Shown are log2 fold change ratios and −log(P-values).