Serine and threonine sit at the heart of cellular regulation, serving as the primary attachment points for phosphorylation and O-glycosylation across the proteome. Yet while chemists have turned cysteine, lysine, and tyrosine modification into routine operations, analogous control over serine and threonine in unprotected polypeptides has remained largely inaccessible. Two intrinsic obstacles conspire against progress: the hydroxyl side chains of both residues are weak nucleophiles under aqueous conditions, and multiple indistinguishable sites within a single protein offer no obvious handle for selectivity. Forcing reactivity through highly electrophilic reagents sacrifices selectivity, while bottom-up chemical protein synthesis requires sophisticated, often fragile, pre-installed building blocks for each modification type. Neither path offers a general, programmable solution.
Researchers in the Li Group at The University of Hong Kong, published in the Journal of the American Chemical Society, report a two-stage surrogate strategy that sidesteps the nucleophilicity problem entirely. Rather than coaxing the native hydroxyl into reacting, the team replaces it temporarily with aminooxy serine or aminooxy threonine, residues whose stronger aminooxy nucleophilicity drives a chemoselective N-terminal ligation, termed aminooxy ligation, AOL. The key mechanistic finding is that AOL proceeds via a transient six-membered 1,2,4-oxadiazinane intermediate, confirmed by NMR isolation of a capped dipeptide model: this ring closure positions the acyl group for a proximity-induced, irreversible 1,5-acyl transfer, releasing a salicylaldehyde and delivering the ligated peptide cleanly in aqueous buffer. The aminooxy group remaining after ligation then serves as the entry point for a second transformation, chemoselective ester ligation, CEL, in which reaction with a keto acid proceeds through an oxaziridine-mediated rearrangement to deliver direct O-acylation at the target residue. Isotope-labeling experiments with H218O traced the oxygen origin in the ester carbonyl, supporting the proposed mechanism and distinguishing the productive pathway from competing amide-forming side routes.
Together, AOL and CEL constitute a modular editing platform that converts a single surrogate residue into a native O-glycosylated, O-acylated, or restored Ser/Thr residue under mild aqueous conditions, without protecting-group manipulation of other sites. The authors demonstrate the approach across therapeutic peptide analogues and extend it to convergent synthesis of full-length histone H2B carrying site-specific modifications. The full scope of substrates, mechanistic controls, and biological applications awaits in the original publication.
