Disulfide-bond assisted peptide ligation for chemical protein synthesis

Abstract

Chemical protein synthesis serves as a powerful approach for protein preparation and modifications. It can provide atom-by-atom control during the whole synthesis process. Hence, it can provide not only natural proteins but also modified proteins and D-proteins. Solid-phase peptide synthesis has been developed to prepare peptide fragments with limited length. Chemical ligations realize chemoselective and stereoselective reactions between unprotected peptide fragments to reach larger homogeneous peptides/proteins. Over the years, many ligation methods have been developed, such as native chemical ligation, serine/threonine ligation, cysteine/penicillamine ligation, and -ketoacid-hydroxylamine ligation. With these advances, various soluble small proteins have been successfully synthesized. However, it remains challenging to synthesize difficult proteins and huge proteins. The difficulties mainly come from the poor solubility and low ligation efficiency of peptides. Peptides with aggregation tendency show poor solubility during the purification and ligation step, which greatly challenges chemical protein synthesis. Many strategies have been introduced to solve the solubility issues. However, all these methods have their limitations, such as limited improving effects, sequence limitation due to specific amino acids required, and tedious linker synthesis. Herein, we proposed an alternative and simple solubilizing tags method, the reducible solubilizing tags (RST) strategy. The solubilizing tags are easily installed into peptides via a traceless disulfide linker and removed in due course. We have incorporated diverse solubilizing tags (i.e., poly-Arg tag, poly-Lys tag, polyethylene glycol tag) into peptides at the internal cysteine site or the salicylaldehyde ester moiety. With the RST strategy, we successfully synthesized the protein 2B4 tail and membrane protein FCER1G. Notably, our method is also well compatible with the semi-synthesis of proteins. We effectively semi-synthesized protein HMGB1. In addition, another major problem of chemical protein synthesis is inefficient chemical ligations. As peptides become long, the difficulty in dissolving peptides in the ligation buffer to achieve high peptide concentration increases, leading to inferior ligation performance. Furthermore, some ligation sites present low reactivity even with good solubility. To solve the problem, we developed the disulfide-bond assisted serine/threonine ligation. We utilized the disulfide bridge to bring two peptides together, turning the ligation into an intramolecular reaction. As a result, disulfide-bond assisted ligation accomplished high-efficient ligations even at low concentrations, enabling the high-yield peptide ligation in a shorter period of time. Based on our model study, the reaction goes well with various C-terminal amino acids, including the sites with high steric hindrance (e.g., Thr, Ile, Pro, Val). Caveolin-1 (CAV-1) is a transmembrane protein that plays vital roles in the membrane lipid composition and signal transduction. It is involved in many disease processes, and thus it is an important research target in the study and treatment of many diseases. Its post-translational modifications have a close relationship with its functions. However, its chemical synthesis remains challenging. By using disulfide-bond assisted serine/threonine ligation, we fulfilled the total synthesis of CAV-1 cytoplasmic domain (2-104) and its two phosphorylated forms. From these successful examples, we envisioned that our disulfide-bond assisted serine/threonine ligation could be utilized to synthesize extremely large proteins, reaching a new milestone in chemical protein synthesis.