Back

A Two-Step Synthesis of Covalent Genetically-Encoded Libraries of Peptide-Derived Macrocycles (cGELs) enables use of electrophiles with diverse reactivity

Walker, J. H. B.; Ekanayake, A. I.; Pedowitz, N.; Qiu, R.; Girnt, P.; Babin, B. M.; Atrazhev, A.; Vukovic, L.; Bogyo, M.; Derda, R.

2025-08-28 biochemistry
10.1101/2025.08.25.672157 bioRxiv
Show abstract

Genetically-encoded libraries of peptide-derived macrocycles containing electrophile warheads (cGELs) can be used to identify potent and selective covalent ligands for protein targets. Such cGELs are synthesized either by incorporation of unnatural amino acids that display mild electrophiles on their side chains or by chemical post-translational modification (cPTM) of mRNA or phage-displayed peptide libraries. Here we investigate fundamental barriers to the synthesis of cGELs. We observe that a previously reported cPTM that proceeds in neutral-to-basic conditions creates mixtures of regioisomers. The complexity of the resulting mixture scales with the electrophilicity of the warhead used in the linker, with some electrophiles being not suitable for use under basic conditions. In contrast, use of a Knorr-pyrazole cPTM enables attachment of electrophiles in acidic pH, thus preventing unwanted reactions with nucleophilic sidechains. The Electrophile is activated only upon mixture with the desired protein target in neutral pH. We use this approach to generate a cGEL with alkyne-bearing macrocycles and use it to identify covalent macrocyclic ligands for pyruvate kinase 2 (PKM2). Our results suggest that construction of cGELs should be performed in conditions that silence the electrophiles (e.g., acidic environment) to prevent unwanted side reactions. In addition to the Knorr-pyrazole method, many other biocompatible bond-forming processes that proceed in mildly acidic pH are likely to be equally effective in constructing cGELs.

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.