IDPSeminars is back TOMORROW! Come see @tanjamittag.bsky.social and @xsalvatella1.bsky.social ! Not to be missed!!
04.03.2026 15:20
👍 19
🔁 8
💬 0
📌 0
IDPSeminars is back TOMORROW! Come see @tanjamittag.bsky.social and @xsalvatella1.bsky.social ! Not to be missed!!
Interesting meeting for those interested in the interface between machine learning and non-globular proteins such a IDPs.
ml4ngp.eu/conference-w...
The microelectrode used is AFAIK a redox electrode so I think is modified in a similar manner.
This project was driven from start-to-finish by postdoc extra-ordinaire Ankush Garg. (I think she might be looking for a job in the New Delhi area!)
Also, great collaboration with @csbrasnett.bsky.social, @cg-martini.bsky.social and Klaus Koren).
We don't think there oxygen is special in this context - rather this represents the baseline which the partitioning of metabolites default to in the absence of interactions with the condensate.
We think this modifies our under standing of small molecule partiotioning into condensates. ⬇️
Instead, we found a striking anti-correlation between protein and oxygen concentrations in the condensates.
More dense condensates = less oxygen inside.
We think this represents an accessible volume effect, where the protein excludes oxygen - and by extension other solutes.
We measured partitioning across a range of synthetic condensates composed of intrinsically disordered proteins (RLPs).
We expected to see a correlation with hydrophobicity as observed for other apolar metabolites. We got no correlation polarity assessed either from sequence of fluorescent sensors.
Firstly, we had to establish ways to measure oxygen partitioning into condensate. We ended up with 2 experimental + 1 simulation (@cg-martini.bsky.social).
Microelectrodes are straight-forward, but requires pellting of condensates. PLIM is more elaborate, but works for dispersed condensates.
🚨New paper: We found - contrary to expectations - that oxygen is partially excluded from biomolecular condensates.
We find a strong anti-correlation between oxygen and protein concentrations in the condensate showing that accessible volume dominates over polarity.
www.nature.com/articles/s41...
This work is the bulk of Alex Harvey's PhD thesis @molbiolau.bsky.social defended a few weeks ago with great success.
Rumor has it that Alex is currently looking for a job!
www.linkedin.com/in/alexharve...
Finally, the mice show impaired learning in the Morris Maze and the Barnes maze - behavioural assays of spatial learning.
Basically, chopping of the 3'UTR of Grin2b makes the mice stupid!
The 3'UTR knock-out mutant has reduced synaptic plasticity (both LTP and LTD) - suggesting defects on the physiological function of the NMDA receptor at the cellular level.
The Grin2B mRNA is normally targeted to synapses - presumably for local translation. This is also observed for many other synaptic proteins due to the highly branched structure of the neuron.
The mutant removing the 3'UTR removes the synaptic targeting.
Alex Harvey - PhD since a few weeks - CRISPRed most of the 3'UTR out of the murine GRIN2B gene which codes for the GluN2B sub-unit of the NMDA receptor.
The knock-out does not affect mRNA levels, but protein levels as ~50% due to reduced translation.
Why does a synaptic protein need a 24kb 3'UTR? We asked ourselves that question when writing a review on the IDRs of the NMDA receptor.
Alex answered this using a knock-out mouse ⬇️: Synaptic targeting of Grin2b mRNA is crucial for synaptic plasticity and learning.
www.pnas.org/doi/10.1073/...
🚨Postdoc opening: Integrative structural biology and interactions of disordered regions in plant transporters.
Skills sought: Protein experience (ideally IDPs/MPs), NMR, SAXS, smFRET, biophysics and similar.
Come find me at #BPS26 or reach out to learn more!
mbg.au.dk/aktuelt/ledi...
Thx, we should cite your paper during revisions. 😏
I guess for sub-stoichiometric clients, high concentrations does not necessarily lead to slow diffusion in itself. However, it is hard to envision a case that results in high partitioning and fast diffusion.
Or by the Arosio lab due to opening of the active site:
www.nature.com/articles/s41...
Diffusion limits have been discussed by many others, so not surprising in itself.
A unique aspectof, is the modulation of diffusion via chain length, which is typically not possible Without changing intrinsic rates.
I also suspect there is a publication bias studies showing an enhancement. Retardation of reactions sounds less interesting .
There are a couple of papers showing HDX in condensates (Cited within). Water should diffuse anymore or less the same rates inthe dense and dilute phas - it is mainly macromolecules that are slowed down. pH is an open question, but it is a charge neutral condensate so shouldn’t shift much.
Thanks!
The project was driven over three years by postdoc Nicolas Gonzalez-Foutel - now Novonesis - with help from Ankush and Evi and collaborators at iNANO and in Brussels to get it over the (first) finishing line 🤞.
We try to distinguish between two types of diffusion limitations: Encounter limited and mass transport limited.
We find that mass transport limitations explain the data best!
We derive a rate equation as a function of volume fraction and fit it the kinetics above - and correlate dense phase reactions rates to diffusion.
The correlation is striking: Slow diffusion = slow catalysis.
Due to the modular design, we can change the partitioning and diffusion of our substrates without changing the intrinsic kinetics.
These experiments show enhancement of reaction rates in the dilute phase (likely nanoclusters) and a length dependent inhibition in the condensate.
We use numerical simulations combining mass action and an inhibitory condensate to show the possible kinetic behaviors.
Especially, the variation of volume fraction of condensate is instructive in underlying effects:
In a elaborate control experiment, we use hydrogen-deuterium exchange MS to probe the structure of the kinase inside the condensate...
... and find no changes at all. The inhibition is not due to changes of the structure and dynamics of the condensate.
We establish a modular and tunable model system for investigating how synthetic condensates affect phosphorylation by protein kinase A.
We find that despite strong co-localisation of enzyme and substrate, the condensates decrease the rate of the reaction.
Graphical abstract of pre-print
Does targeting enzymes and substrates in a condensate lead to rate enhancement? No. Here, we investigate how the condensate environment can inhibit an enzyme reaction.
Spoiler: Mass-transport limitations. We find a strong correlation between diffusion and reaction rates.
doi.org/10.64898/202...