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🧵
Are you interested in working with a fresh team on cutting edge protein design?
We're hiring a Technical Assistant for our protein design team at DKFZ Heidelberg: jobs.dkfz.de/en/jobs/1683...
Please share!
🎉🎉 Our paper on temperature-dependent protein control using AsLOV2 variants is now published at @natchembio.nature.com: www.nature.com/articles/s41... including some new insights from extend variant characterizations.
Many congratulations to first author @neuroscinikolai.bsky.social for his amazing work and to all other authors that were involved, including Anna Von Bachmann for her excellent bioinformatics skills!
Together, POGO-PANCE and RAMPhaGE offer a versatile framework for evolving molecular switches and decoding allosteric architecture—by letting evolution sculpt the signal.
Using RAMPhaGE, we diversified the AraC–LOV2 linker and applied POGO-PANCE evolution, uncovering a single-residue deletion/substitution that markedly improved switching likely by stabilizing a continuous α-helix at the sensor-effector junction, yielding near-digital on/off behavior.
To go beyond point mutations, we also built RAMPhaGE: a retron-recombineering platform for targeted phage genome editing. RAMPhaGE enables targeted and cummulative substitutions, insertions, and deletions—supporting tunable library design and high-efficiency in vivo diversification.
Long-read sequencing across POGO-PANCE cycles revealed mutational trajectories and co-evolving networks spanning both AraC and LOV; We observed allosteric communication emerging in real time, structured by alternating selection and mapped as mutational hotspots.
To overcome this, we developed POGO-PANCE—a system that evolves protein switches by mimicking nature. By alternating positive and negative selection aligned with the presence or absence of an inducer, we yielded Optogenetic AraC-AsLOV2 variants with >1000-fold dark/light expression changes.
The principles underlying allostery remain elusive; engineering synthetic allostery is an even greater challenge. While tools like ProDomino can identify permissive allosteric insertion sites in proteins, achieving efficient switching output often still requires further optimization.
Inspired by how nature evolves trigger responsiveness through alternating pressures, we are excited to present POGO-PANCE and RAMPhaGE:
Phage-assisted evolution platforms for engineering allosteric protein switches under dynamic selection.
Preprint: doi.org/10.1101/2025...
Check out the new pre-print from our lab on phage-assisted evolution of light-switchable, allosteric proteins. Congrats to first author @neuroscinikolai.bsky.social, co-corresponding author @jmathony.bsky.social and everyone from the @niopeklab.bsky.social involved!
www.biorxiv.org/content/10.1...
Congratulations to all authors, especially 1st author @pmuench.bsky.social as well as @neuroscinikolai.bsky.social and Matteo Fiumara for their important contributions!
Many thanks to @graeffjohannes.bsky.social for the productive and fun collaboration.
Excited to announce our optogenetic transcriptional deactivation toolbox is now out in its final form at Nucleic Acids research: academic.oup.com/nar/advance-....
Many congratulations to first author @bene837.bsky.social for his amazing work and to all other authors that were involved into the extensive experimental validation, especially @pegish.bsky.social and Sabine Aschenbrenner.
Especially the chemically regulated Cas12a variants showed potent editing and extremely strong response to the inducer. Importantly, all our reported protein switches were generated without any downstream optimization.
Using ProDomino, we created potent optogenetic variants of the puromycin and chloramphenicol antibiotic resistances. Moreover, we engineered blue light-responsive Cas9-VPR transcriptional activators and light- or cortisol-dependent variants of MbCas12a.
This approach, together with ESM-2-based embeddings and a masking strategy enabled us to train a model that showed high success rates in the subsequent wetlab validation.
A main limitation is the absence of sufficiently large experimental datasets that could be used to train ML models. For our new model ProDomino (protein domain insertion optimizer), we leveraged intradomain insertions in natural proteins identified based on CATH/Interpro annotations.
Allosteric protein switches are usually created by inserting a receptor domain into an effector protein. However, the identification of suitable insertion sites remained challenging and nearly impossible to predict.
We are thrilled to share ProDomino a model for the prediction of domain insertion sites in proteins. Our approach enables the simple and rapid engineering of highly potent switchable proteins, as we exemplify by creating novel inducible variants of Cas9 and Cas12a.
www.biorxiv.org/content/10.1...
Deep congrats to co-first authors @lucabrenker.bsky.social , Sabine Aschenbrenner, and Felix Bubeck as well as all other authors.
On the fly, we also created circularly permuted variants of human receptor domains that should be well-suited for allosteric protein control beyond the anti-CRISPR space.
(2/3)
New pre-print from our group reporting engineered, broad-spectrum anti-CRISPR proteins based on AcrIIA5, a type II inhibitor, and AcrVA1, a type V inhibitor, for opto- and chemogenetic control of CRISPR-Cas9 and -Cas12a:
www.biorxiv.org/content/10.1...
(1/3)
Deep congrats to co-first authors Luca Brenker, Sabine Aschenbrenner, and Felix Bubeck as well as all other authors.
On the fly, we also created circularly permuted variants of human receptor domains that should be well-suited for allosteric protein control beyond the anti-CRISPR space.
2/3
Congrats to co-first authors Tobias Stadelmann, Daniel Heid and @mjendrusch.bsky.social as well as everyone involved.
(4/4)
Moreover, we find that AcrIIA5 can actually inhibit Cas9 DNA binding in E. coli to some extent, a property which can be enhanced by certain mutations within or close to its internal IDR.
(3/4)
We here report comprehensive single mutation maps for two anti-CRISPR proteins, AcrIIA4 and AcrIIA5, and identify mutation tolerant regions of interest for Acr engineering.
(2/4)
