Solid-Phase Methods for the Preparation of Epoxysuccinate-Based Inhibitors of Cysteine Proteases

Abstract

ReceiVed July 21, 2006 The CA clan papain family is one of the largest and best studied subfamilies of the cysteine proteases. In addition to their association with physiological processes such as antigen presentation,1 it has become clear that these enzymes are involved in several pathological processes including rheumatoid arthiritis,2 Alzheimer’s disease,3 cancer invasion and metastasis,4 and several parasitic diseases.5 Selective inhibition of papain cysteine proteases will aid in our understanding of their role in these human diseases. More importantly, recent studies using papain family selective epoxysuccinyl inhibitors in mouse models of cancer suggest they may represent a viable new class of anticancer chemotherapeutic agents.6 Currently, one such epoxysuccinate inhibitor is in preclinical trials at the National Cancer Institute. Thus, methods allowing the rapid synthesis of libraries of diverse epoxysuccinates that contain reduced peptide character will be critical for the identification of optimal lead compounds for use in human clinical trials. The natural product E-647 (1, Figure 1) and its closely related analog JPM-5658 (2) are epoxysuccinyl-based covalent inhibitors of papain-like cysteine proteases. Their broadspectrum activity against this family of enzymes is derived from the leucine residue that mimics the P2 position of a substrate and interacts with the hydrophobic S2 pocket of most papain-like proteases. The ethyl ester of 2, JPM-OEt (3), has been used in in vivo studies using a mouse model for pancreatic cancer.6 Treatment with JPM-OEt resulted in a significant reduction in angiogenesis, tumor volume, and tumor invasiveness. Combined with overall low toxicity, these results support the development of this class of inhibitors as anticancer chemotherapeutics. However, the very short half-life of JPMOEt in vivo and its overall limited bioavailability makes it less suitable as a drug candidate. We recently reported a solid-phase synthesis route that allows the incorporation of variable peptide elements on both sides of the epoxide function.9 Synthesis of a number of “double-headed” epoxides using this method confirmed the importance of the interactions of inhibitors with regions of the active site on both sides of the reactive cysteine nucleophile. However, the current synthetic methods are only able to produce compounds that are highly peptidic in nature and therefore unlikely to have optimal pharmacological properties. Furthermore, the incorporation of diversity elements is limited by the need for suitably protected amino acid building blocks for attachment to the resin. We reasoned that the incorporation of non-natural elements would be an advantageous strategy for the design of a cathepsin inhibitor because it would allow for the introduction of a larger diversity of R groups within the inhibitor scaffold (4, Figure 1). In addition, the reduction of the peptide character would significantly increase the druglike properties of the resulting compounds. The use of diverse amines to generate the R1 element on the inhibitor scaffold (see Figure 1) precludes the use of the standard Rink linkage to the resin. To address this issue, we investigated resins that have been applied to make Cterminal-modified peptides.10 In our first approach, a safety-catch resin was loaded with the critical P2 amino acid using reported procedures (Scheme 1, method A).11 Next, we introduced the epoxysuccinate warhead using the activated nitrophenyl ester 5, followed by hydrolysis of the ethyl ester to introduce a R3 group at the prime side of the molecule. However, treatment of the resin under saponifying conditions, as previously described,9 resulted in premature cleavage from the solid support and therefore proved incompatible with the safety-catch resin. To overcome this problem, we coupled epoxysuccinate 6,12 having two free acid groups, with PyBOP/DIEA. The resulting immobilized acid could then be directly extended by coupling of the R3 amine using a standard PyBOP/DIEAmediated coupling.9 Activation of the sulfonamide function by iodoacetonitrile and the subsequent release from the resin using an excess of the R-NH2 yielded the desired compounds in a 12-20% yield after HPLC purification (based on original resin load reported by vendor; see Table 1). Although the safety-catch resin provides the desired products in reasonably good yields, the presence of an excess of the amine during cleavage from the resin results in a * To whom correspondence should be addressed. E-mail: mbogyo@ stanford.edu. Figure 1. Structures of natural product E-64, its synthesized analogs, JPM-565 and JPM-OEt, and the general scaffold 4. 802 J. Comb. Chem. 2006, 8, 802-804