Big congratulations and a huge thank you to everyone involved. I am especially greatful for the great collaboration with the shared first author Ying Zhang who was doing the biochemical experiments and handled the yeast.

These findings provide key insight into cotranslational protein folding and proteostasis in eukaryotic cells, resolve the cotranslational cycle of Ssb, and support a unified model integrating prior structural and biochemical data.

Further, we show that initial recruitment of Ssb to the ribosome is independent of direct interaction with RAC but is stabilized by engagement with the NC accompanied by ATP hydrolysis.
This is a criminally short summary of many experiments, please check the paper if you want to know more 😜

Point mutations not only verified our model, but also established this as the interface which is engaged by Ssb in the precatalytic ATP bound state and where Ssb in the ADP state remains bound in the post catalytic ADP state.

We identified Rpl25 as the major interaction site, which was resolved to ~4Å. Being able to assign side chains, we further observed several charged interactions governing this interaction. This was our key finding, which allowed us to decipher the early events of the Ssb cotranslational cycle.

We modeled Met-elongator tRNA base pairing to the final AUG codon and the nascent chain in the exit tunnel. Owing to limited Ssb–SBD-β resolution due to flexibility, the bound NC was only apparent in low-pass filtered maps.

Ending up with just ~14K and ~33K particles we solved the structure of RNC-Ssb in the ADP state in two conformations called S1 and S2. Global resolution was estimated at 2.8Å and 3.0Å for S1 and S2 state respectively. Local resolution of the 80S-Ssb interface, however, was limited to 4-10Å.

Therefore, we started to iteratively classify our particles using increasingly tighter masks. Due to our particle numbers dropping fast, we needed a lot of data. Backfilling our Krios queue, we finally collected >54.000 micrographs.

Focusing on this stalled ribosome population, masked classifications began to converge, and we were able to identify the density as Ssb’s C-terminal substrate binding domain (SBD). With <25 kDa the SBD was too small for local refinements severely limiting the local resolution of the SBD.

The silver bullet was to assess global heterogeneity first. We used cryoDRGN and analyzed the landscape analysis. We found that non-rotated ribosomes with a P-site tRNA and a density corresponding to a trailing ribosome (Figure, Cluster 1), were ever so slightly enriched in Ssb occupancy.

In the initial consensus reconstructions, we noted a small, very weak density close to the exit tunnel in low-pass filtered, weakly thresholded maps. Owing to its low occupancy, masked classification in CS or RELION did not permit further subdivision of the particle stack.
- Figure not published -

We expressed a truncated Flag-Pgk(1-70) mRNA without a stop codon in a yeast in-vitro system with added recombinant Ssb. Thus, we were able to enrich stalled RNC using the nascent chain's (NC) N-terminal Flag tag and prepare cryoEM grids.

To close this knowledge gap, we turned to the yeast Hsp70 Ssb. First, we (@haselbachlab.bsky.social, @impvienna.bsky.social in collaboration with the Rospert group @uni-freiburg.de ) started working on cryoEM structures of ribosome-nascent-chain-complexes (RNC) cotranslational engaged with Ssb.

Coupling ribosomal translation to cotranslational folding is essential for cellular homeostasis. In eukaryotes, this process relies on Hsp70 and its J-domain cochaperone RAC, but how Hsp70 activity is coordinated with translation remains unclear.

I didn't realize. They nonetheless ended up in my stomach - so tasty :)

Congratulations Juan. Really amazing 🤩