α-Amino-γ-lactam, Agl, residues are prized peptidomimetic building blocks: they lock backbone geometry in a way that can improve membrane permeability, yet incorporating them into genetically encoded peptide libraries has remained an open problem. Existing routes, including on-resin lactam annulation, reductive amination, and selenomethionine-mediated substitution, all require conditions incompatible with the aqueous, RNA-containing milieu of mRNA display. An earlier attempt to load presynthesized Agl-containing dipeptides directly into ribosomes achieved only roughly 5% translation efficiency, too low for practical screening. Meanwhile, chemical post-translational modification, PTM, strategies in display systems have focused almost exclusively on macrocyclization, leaving RNA-compatible backbone-modifying chemistries largely unexplored. The gap matters because backbone topology, not just side-chain identity, governs the metabolic stability and permeability that candidate peptides need to become drug leads.
Researchers in the Suga Group at The University of Tokyo, published in J. Am. Chem. Soc., recognized that L-aspartyl aldehyde residues spontaneously form hemiaminals with the adjacent backbone amide and asked whether that intermediate could be coaxed, under mild acidic conditions, into a full Pictet–Spengler-type cascade cyclization with a neighboring nucleophilic side chain. Systematic SPPS model studies mapped the pH dependence, nucleophile scope, and stereochemical outcome of the reaction, identifying tryptophan and cysteine as the only proteinogenic partners that react efficiently at biocompatible pH values; tyrosine, histidine, and lysine were unreactive under the conditions tested. A semicarbazone-protected aspartyl aldehyde building block was charged onto tRNA via flexizyme catalysis and incorporated into peptides by the FIT cell-free translation system. After translation, mild acidification unmasked the aldehyde and drove cyclization to yield polycyclic Agl, pcAgl, motifs with near-quantitative conversion overnight, while acetone precipitation cleanly removed the acid and preserved sufficient RNA integrity for reverse transcription and PCR amplification.
Integrating this PTM step into the RaPID selection platform, the team ran a proof-of-concept campaign against MAT2A, a SAM-synthesizing enzyme relevant to MTAP-deficient cancers, and discovered macrocyclic peptide inhibitors whose pcAgl motifs proved structurally critical by saturation mutagenesis. Iterative affinity maturation then guided the search toward lariat topologies with further improved potency. The work establishes a blueprint for purely chemical, recognition-sequence-independent backbone modification in mRNA display, and the full stereochemical analysis, selection data, and inhibitor characterization are reported in the original publication.
