Thread by @WoutOosterheert

Very proud that our manuscript on the assembly and aging of actin filaments (and my first publication as a postdoc) is finally online @Nature (open access!)

nature.com/articles/s4158…

This was a great team effort from the @Intein lab @mpimoph.

A thread (1/17)

Actin is a highly abundant protein and a major component of the eukaryotic cytoskeleton. It exists in two main forms, as monomer (G-actin) and as filament (F-actin). The dynamic turnover of actin filaments is crucial for controlling cell shape, polarity and movement.

(2/17)

After polymerization, F-actin rapidly hydrolyzes ATP and undergoes transitions dependent on its bound nucleotide. 'Aged' ADP-F-actin can be depolymerized back to G-actin.

We aimed to visualize and understand F-actin assembly and aging at the highest molecular detail.

(3/17)

Through an optimized cryo-EM workflow using @SPHIRE_Package and #relion, we determined structures of Mg-F-actin in three nucleotide states (ATP-mimic, ADP-Pi, ADP) at ~2.2 Å resolution.

The high structural detail allowed for the modeling of hundreds of water molecules.

(4/17)

Mg is bound to actin in vivo. However, actin is standardly purified in buffer containing Ca. To assess the effect of the bound cation on the F-actin structure, we also determined structures of Ca-F-actin at similarly high resolutions.

(5/17)

When comparing the six nucleotide-state structures, we globally observed no major structural rearrangements.

Thus, the differences were in the details. But luckily, at these resolutions, the structures displayed a lot of detail :)

(6/17)

First, by comparing our cryo-EM structures with x-ray structures of G-actin, we could show that the G- to F-actin transition triggers the relocation of water molecules near the ATP binding site. This explains why F-actin hydrolyzes ATP within seconds but G-actin does not.

(7/17)

To recapitulate the video: In Mg-actin, water molecules are relocated during the G- to F-actin transition but the Mg-ion coordinating waters (in magenta) are not altered, retaining an octahedral coordination.

But why is the polymerization rate of Ca-actin so much slower?
(8/17)

As it turns out, in Ca-actin, water relocation necessitates the movement of a water (Wx) within the coordination sphere of Ca.

We propose that this required rearrangement of the Ca-coordination sphere could pose a kinetic barrier for the G- to F-actin transition.

(9/17)

Zooming in on ATP hydrolysis in F-actin: We observed a water hydrogen-bonded to Q137 within 3.6 Å of the Pγ-analog, which likely is the nucleophilic water (Wnuc) that hydrolyzes ATP.

The position of Wnuc is further stabilized by another water (Wbridge), D154 and H161
(10/17)

We next compared the ADP-Pi and ADP states of F-actin to investigate the mechanism of Pi release.

Pi release is predicted to involve a 'back door' of residues R177 and N111. Interestingly, the back door was previously proposed to be open in ADP-F-actin.

(11/17)

To our surprise, the back door is closed in structures of both ADP-Pi and ADP F-actin, indicating that the F-actin conformation that allows for the exit of Pi is a transient state.

We envision that the Pi release mechanism can be further explored in future experiments
(12/17)

Finally, a dogma in the field is that changes in the F-actin nucleotide are coupled to the filament periphery. Our structures now suggest a path for how the nucleotide-binding site and the F-actin C-terminus are coupled.
Please check the paper for many more details!
(13/17)

We hope that our work will be of valuable interest to all cytoskeleton enthousiasts, and hope that it will now also pave the way for high-resolution structures of F-actin bound to ABPs.
In general, these are extremely exciting time in the actin structural biology field.

(14/17)

Complementary to our work, a beautiful study by the @Alushin_Lab lab, also out in @Nature today, investigated the bending of F-actin, see: nature.com/articles/s4158…

See also the associated News and Views article discussing both our papers: nature.com/articles/d4158…

(15/17)

In addition, the Maeda and Takeda labs determined high-resolution x-ray structures of monomeric actin in an F-like conformation.
doi.org/10.1073/pnas.2…

(16/17)

I would like to thank @Intein for the invaluable support and @alexanderbelyy1, Björn & Sabrina for their great contributions.
Finally, I gratefully acknowledge the @AvHStiftung for funding my postdoctoral fellowship. Much more exciting science to come, watch this spot!

(17/17)