Abstract
The reaction between C-terminal α-thioester and N-terminal cysteinyl peptides is known as native chemical ligation (NCL). Alkyl α-thioesters are traditionally prepared in NCL due to their higher thermodynamic stability, which endows resistance to hydrolysis and easier peptide handling. However, the ligation kinetics of these species are slow, and the reaction times exceed the practical limits for chemical protein synthesis. Therefore, conversion to a more reactive phenyl α-thioesters through thiol-thioester exchange is usually employed to enhance the NCL reaction rate. In addition, phenyl thiols can reverse the formation of the less reactive branched thioesters, i.e., thioesters formed with internal Cys residues, and thiolactones. Interestingly, the fastest NCL rates are achieved with phenyl α-selenoesters, though phenylselenol is a poor catalyst for the selenol-(α-alkyl thioester) exchange, particularly with β-branched residues. Hence, it is usually necessary to preform the phenyl α-selenoester and protect internal cysteine residues to preserve the kinetic advantage. Moreover, an ∼2-5-fold excess of the N-terminal cysteine acceptor peptide is typically required to prevent the competition from the cysteine of the ligation product for the phenyl α-selenoester donor that leads to the branched thioester formation. Based on these precedents, we have designed sodium 2-selenoethanesulfonate (SeESNa) as a new selenol catalyst that can overcome these limitations. SeESNa reacts with alkyl α-thioester, N-acyl benzotriazole, and N-acylurea peptides, giving α-SeESNa species. We have determined the rate constants for the ligation with preformed α-SeESNa peptides and show that it is a superior catalyst compared to the known 4-mercaptophenylacetic and 4-mercaptobenzoic acids. The utility of SeESNA was proved through the synthesis of the cardiotoxin A5, a snake venom peptide that contains eight cysteines, without orthogonal cysteine protection. Importantly, it has enabled expressed protein ligation under folding conditions with extraordinary speed, as shown with the Sonic Hedgehog and SUMO2 proteins. Thus, SeESNa is envisaged to have broad applicability in synthetic and semisynthetic protein chemistry.