iiinova — where your digital self keeps evolving

Carl von Hallerstein @jchvh.bsky.social · 28/09/2026

I'm hugely grateful to Angie Russell and @philbiggin.bsky.social for being amazingly supportive throughout this project! I also want to give a shoutout to Johnny Colburn and to all the other first-class colleagues in Oxford SBCB and the @sykhalid.bsky.social group who helped me at various stages!

230

Carl von Hallerstein @jchvh.bsky.social · 28/09/2026

We then performed an AF2-based screen on other class A GPCRs and found that 13 shared an identical residue at position 6.52 and 187 at position 5.47, so this kind of hydrophobic levering mechanism might have applicability beyond GPR84...

110

Carl von Hallerstein @jchvh.bsky.social · 28/09/2026

Validation work using cell-based assays and more sims followed. My personal highlight was the fact that the Russell group were able to design a ligand library that used substituents with different van der Waals radii to exploit the L336 lever as a way to modify bias. Check out the level of control!

110

Carl von Hallerstein @jchvh.bsky.social · 28/09/2026

It turns out that in the absence of a G-protein binding partner, the -CF3 group on the ligand wedges between TM5 and TM6 and forces them apart. The hydrophobic residues can also act as a lever that twists TM6 outwards and "unplugs" the residue Y332 (6.48) from its H-bond network…

110

Carl von Hallerstein @jchvh.bsky.social · 28/09/2026

Since the structure provided no direct insight, we thought the answer might lie in how the ligand affects receptor dynamics, and so turned to Phil Biggin, my eventual master's supervisor. With Phil, I performed large-scale MD simulations (10 × 2 µs) that confirmed our suspicions…

110

Carl von Hallerstein @jchvh.bsky.social · 28/09/2026

She also showed me the GPR84–Gi–ligand cryo-EM structure from a collab with the Zhang lab. By itself, the structure was bafflingly uninformative: the only real difference compared to the structure with a reference ligand was a minor outward displacement of two hydrophobic residues, L336 and F187.

110

Carl von Hallerstein @jchvh.bsky.social · 28/09/2026

Back in 2023, when I was looking for master's projects, Angie Russell showed me some structure–activity data from her group's latest library. The results were striking: with a simple -CH3 → -CF3 substitution, beta-arrestin recruitment was completely switched off, without impacting Gi-signalling.

110

Carl von Hallerstein @jchvh.bsky.social · 28/09/2026

Really chuffed that this has finally made it out after so many years in the making. TLDR: we chased down the mechanism behind an extremely steep and unusual structure-activity cliff at a class A GPCR. It’s a bit different from my current work on Mass Spec+EM, but I thought I’d do a little 🧵 anyway:

120

Reposted by Carl von Hallerstein

Anika Baier @anikabaier.bsky.social · 17/09/2025

Big thanks to @arnemoeller.bsky.social , @djanuliene.bsky.social, all speakers, and everyone who made the #WEHeraeusSeminar - Cryo-EM on the Move such an amazing event! It was a blast! #cryoEM #StructuralBiology #Microscopy

093