Highly efficient and chemoselective peptide ubiquitylation.

Post-translational modifications play an important role in regulating protein structure and function in health and disease. Ubiquitylation is one example for such a modification wherein both the extent (polyversus mono-ubiquitylation) and the sequence position of this modification dictates the function and fate of the ubiquitylated protein. In the ubiquitylation process, three distinct enzymes, known as the E1–E3 system, collaborate to achieve a site-specific tagging of the lysine residue(s) in target protein. This condensation step generates an isopeptide linkage between the e-NH2 of the lysine residue and the activated C-terminal glycine of ubiquitin (Ub). The overwhelming majority of the studies in the field rely on the in vitro enzymatic reconstitution of this complex posttranslational modification for the protein of interest. This process is often challenged by the heterogeneity of the modified protein, the isolation of the specific ligase (E3), and obtaining reasonable quantities of the ubiquitylated protein. In principle, adopting chemical and semisynthetic strategies for the preparation of posttranslationally modified protein could overcome these challenges and allow for sitespecific ubiquitylation of the modified protein in sufficient quantities for biochemical and structural studies. With respect to Ub, early efforts in these directions include the total synthesis of Ub using stepwise solid-phase peptide synthesis (SPPS) and native chemical ligation (NCL), semisynthesis of Ub analogues with C-terminal electrophile, and chemical manipulation of the polyubiquitin chain. Recently, Muir and co-workers devised an elegant semisynthetic approach that allows site-specific peptide and protein mono-ubiquitylation.This strategy relies on attaching a photoremovable auxiliary to a lysine residue to mediate isopeptide formation with the Ub-a-thioester. This concept was then applied to synthesize a homogeneous ubiquitylated histone H2B, which was used for the assembly of the nucleosome in vitro. This study paved the way to unravel the effect of mono-ubiquitylation on the mechanism of stimulation of the Dot1-catalyzed histone H3 methylation. Despite the success in obtaining a purely modified ubiquitylated histone H2B, the crucial step in the synthesis, that is, the isopeptide formation mediated by the photoremovable auxiliary, proceeded slowly, and required five days for ubiquitylation and seven days for SUMOylation (small ubiqutin-like modification) to achieve a reasonable conversion to the mono-ubiquitylated product. The slow ligation rate was attributed to the involvement of the secondary amine in the S-N acyl transfer, despite proceeding through a fivemembered ring transition state. Indeed, in sugarand sidechain-assisted ligation, the rates are reasonable (6–48 h) despite the reaction proceeding through 14–15-membered ring intermediates for S-N acyl transfer. The involvement of a primary amine in the acyl transfer step, along with to the proximity effect, is a crucial factor. Inspired by NCL, the ligation method developed by Kent and co-workers, we reasoned that the formation of fivemembered-ring transition state involving the primary e-NH2 should have a major influence on the ligation rate. This method would require the installation of a thiol group at the lysine side-chain specifically on the d-carbon generating a cysteine-like system (Scheme 1). With this in mind, we devised a new thiol-modified lysine analogue, that is, d-mercaptolysine, to allow a thiol capture step with the Uba-thioester, followed by S-N acyl transfer, to form the isopeptide linkage. Moreover, this modification should be removable by applying the desulfurization reaction developed by Dawson and Yan to furnish the unmodified lysine (Scheme 1). Initially, we focused on the synthesis of the d-mercaptolysine. We realized that the chiral center on the d-carbon could be installed in its diastereomeric form, as it will be removed after the desulfurization, which simplifies the synthesis and purification steps. The synthesis of modified lysine started from commercially available l-glutamic acid, which was converted into aldehyde 1 in three steps according to a previously reported procedure (Scheme 2). Subsequently, the Henry reaction was applied to 1 with nitromethane, in the presence of TBAF as a base, to give a diastereomeric mixture of nitro alcohol 2 in a 1:1 ratio (based on H NMR spectroscopy; see the Supporting Information). The one-pot acetylation, followed by elimination of the alcohol functionality on the nitro alcohol 2 using acetic anhydride and 4-dimethylaminopyridine, afforded the E/Z mixture of the Michael acceptor 3 in 71% yield. To avoid any racemization at the a carbon, the reaction of lithium tert-butylsulfide with conjugated nitro olefin 3was performed at 78 8C to furnish a diasteromeric mixture of tert-butylmercapto nitro compound 4 in a ratio of 52:48 (see the Supporting Information). Next, the reduction of nitro group to the amine using NaBH4 and NiCl2 6H2O, [19] followed by protection with allyloxycarbonyl chloride, afforded the alloc-protected tert-butyl [*] Dr. K. S. Ajish Kumar, M. Haj-Yahya, Dr. A. Brik Department of Chemistry, Ben-Gurion University of the Negev Beer Sheva 84105 (Israel) Fax: (+972)8-647-2943 E-mail: [email protected] Homepage: http://www.bgu.ac.il/~abrik