Posttranslational mutagenesis: A chemical strategy for exploring protein side-chain diversity

INTRODUCTION Natural posttranslational modifications (PTMs) to proteins expand the chemical groups available to proteins. The ability to expand posttranslational functional group diversity in an unbounded manner could, in principle, allow exploration and understanding of how these groups modulate biological function. Natural PTMs feature bonds to heteroatoms (non-carbon) made at the γ (Cys Sγ, Thr Oγ, Ser Oγ) or ω (Lys Nω, Tyr Oω) positions of side chains. However, one of the central features of biomolecules is C (sp 3 )–C (sp 3 ) bond formation. Because all amino acid side chains contain this C–C bond, mastering its construction on proteins could allow free-ranging structural alteration of residues in proteins (both natural and unnatural). RATIONALE In principle, C (sp 3 )–C (sp 3 ) disconnections at the β,γ C–C bond would allow the chemical installation of a wide range of amino acid functionalities. Traditional two-electron chemistry (using nucleophiles and electrophiles) requires reagents that are often incompatible with biological substrates and/or water. Free radicals can be tolerant of aqueous conditions and unreactive (and thereby compatible) with the majority of functionality present in biomolecules. We therefore reasoned that mild, carbon-centered free radical chemistry would be enabled by matching free-radical reactivity with a suitable, uniquely reactive functional group partner that possesses a chemical affinity for such singly occupied molecular orbitals. The amino acid residue dehydroalanine (Dha) can be readily introduced in a site-selective manner genetically, biosynthetically, or chemically; upon reaction with a suitable radical, Dha would favorably generate a stabilized Cα radical 1 . Suitable “quenching” of the central Cα radical intermediate 1 generated after formation of the critical C–C bond would thus allow “chemical mutation” of the side chain. RESULTS A range of precursor halides ( R -Hal, Hal = I or Br) allowed the creation of radicals R •. These radicals reacted selectively with Dha in peptides and proteins with excellent site selectivity and regioselectivity (>98% β,γ) and typically with a diastereoselectivity of ~1:1. Combined use of R-Hal with NaBH 4 under low-oxygen conditions suppressed competing oxidation and disubstitution side reactions of intermediates 1 . This allowed for rapid reactions (typically 30 min) with improved efficiency across a range of representative protein types and scaffolds (all α, α/β folds, all β, receptor, enzyme, antibody). The reactivity of primary, secondary, and tertiary alkyl halides allowed installation of natural, simple hydrophobic residue side chains. Charged or polar protic (e.g., OH, NH) functionality in amino acid side chains was also possible. Even the use of side-chain reagents in unprotected form proved possible, thus highlighting not only exquisite chemoselectivity but also compatibility with common biological functional groups. These transformations enabled the creation of a wide diversity of natural, unnatural, posttranslationally modified (methylated, glycosylated, phosphorylated, hydroxylated) and labeled (fluorinated, isotopically labeled) side chains, as well as difficult-to-access but important residues in proteins (e.g., methyl-Arg, citrulline, ornithine, methyl-Gln, phospho-Ser). CONCLUSION This approach to chemical editing of amino acid residues, outside of the rigid constraints of the ribosome and enzymatic processing, may prove to be a general technology for accessing diverse, previously unattainable proteins.