@sagarkhare.bsky.social
Professor, Chemistry and Chemical Biology, Rutgers University. Interested in protein design, enzymes, molecular modeling, biophysics, AI/ML for bioX. Here for science, society and lack of toxicity
Nature magazine asked me to write a "News& Views" article on a recent paper on computational enzyme design. We enjoyed writing it (especially for non-specialists), hope you enjoy reading it (and the excellent paper by Fleishman et al. it accompanies)!
Shared PDF link: rdcu.be/evlwJ
lnkd.in/eKxXm_b6
📢 New preprint, lead by Alex Bronstein, @sankethvedula.bsky.social and colleagues:
Guiding AlphaFold with experimental data.
This approach generates conformational ensembles guided by NMR, X-ray (or whatever) data. I am thrilled to tackle lots of exciting problems with it!
arxiv.org/abs/2502.09372
So, how does the chaperone HtrA1 do it? Well, it uses cool tricks: it uses two (important) places to bind not one, it not just binds but also cuts its target, and it binds to various species (seeds, monomers) along the aggregation pathway. See how we figured it out! (4/n)
In the classical inhibition paradigm, a protein is inhibited by (think antibodies) by binding tightly at one site. This is quite challenging to do if you have, literally, a very moving target in the form of a disordered protein. Moreover, aggregation pathways involve species with >1 chains (3/n)
With this broader motivation and in collaboration with Jean Baum's group, we brought to bear a number of computational and experimental biophysical tools (including cool Atomic Force Microscopy in Liquids!) to figure out how HtrA1 inhibits the aggregation and prion-like spread of a-synuclein. (2/n)
New pre-print alert! How does a chaperone prevent protein aggregation, especially of intrinsically disordered proteins? If we learn how, perhaps we could design inhibitory proteins based on the same mechanistic blueprint as much-needed therapeutic leads against neurodegenerative diseases. (1/n)
Hello world! Feels good to be here :)
