Nicholas Southern

1/28 How do you optimize a dynamic protein property that emerges from multiple states? Our finally published paper in @NatureComms takes on one of the hardest problems in protein engineering with phage assisted evolution: evolving allosteric switches 🧵

GROQ-seq Datasets Across Transcription Factors (LacI, RamR, VanR), T7 RNA Polymerase and TEV Protease biorxiv.org/content/10.648… #biorxiv_bioeng

Diversity-generating retroelements for programmable targeted hypermutagenesis go.nature.com/3OG2nKN

Nice work on de novo switches by Guo et al.

@Jtuscher @NatureComms Thanks Jen!! 😄

@NeurosciNikolai @NatureComms Congrats Nick🎉

1/28 How do you optimize a dynamic protein property that emerges from multiple states? Our finally published paper in @NatureComms takes on one of the hardest problems in protein engineering with phage assisted evolution: evolving allosteric switches 🧵

28/28 Code, scripts, and sequencing data are available on GitHub, Zenodo, and NCBI. All plasmids including our POGO-PANCE circuits, phages, and RAMPhaGE editing plasmid RcP9 are available on Addgene. Full paper here: rdcu.be/fdwKu Questions welcome!

27/28 That said, we are excited to apply this to more effector-receptor pairs to uncover more generalizable principles of allostery. The more systems we evolve, the closer we get to understanding the rules that govern it!

26/28 Paired with ProDomino for site identification, there is now a clear end-to-end pipeline for allosteric switch engineering. One key caveat: protein activity must be linkable to phage fitness, which requires upfront circuit engineering and is not trivial for every target.

25/28 We applied RAMPhaGE to our LOV-inserted starting point, diversifying linker junctions enriched under prior evolution. Combined with POGO-PANCE cycling, this uncovered a different solution class that resulted in near-digital switches inaccessible to point mutagenesis alone.

24/28 Editing came at a cost: unlike DP6, which can expand phage pools from low phage titers, RAMPhaGE output titers closely mirrors input. Sequential passaging strategies can address this, and perhaps combinations with selection could be a promising direction.