ZAK activation at the collided ribosome

Figures

• 

  Fig. 1. Cryo-EM structure of ZAK bound to a colliding disome. a, Polysome profile (top) and immunoblots of a Phos-tag gel (IBs; bottom) from sucrose gradient fractions of HEK293T cells transfected with Strep-tagged WT and kinase inactive ZAK(T161A/S165A) expressed from a complete CMV promoter plasmid. Polysome profile from WT transfection is shown. b, Cut top view on the ZAK–disome model shown as low-pass-filtered surface, highlighting mRNA, tRNA and EDF1-binding sites. The boxed panels show a zoomed-in view on the ANS (red)-binding site of refined cryo-EM maps (transparent) of the stalled (right) and collided (left) 80S. c,d, Composite cryo-EM map (c) and molecular model (d) of the ZAK–disome complex. Structural features of ZAK are indicated as SAM domain (SAM), pin, RIH, RIM and RIH-p. e, Schematic of the ZAK domain architecture based on AlphaFold prediction: N-terminal kinase domain, leucine zipper (LZ) region, SAM, YLD, sensing region and the CTD. Residues relevant for ribosome interaction as identified in this study are depicted in blue boxes. Vertical lines with P (red) indicate previously published subset of RSR phosphorylation sites (cluster 2)12. The blots represent at least two…

• 

  Fig. 2. The C terminus of ZAK mediates ribosome binding and collision-specific rRNA interaction. a, Top view on the molecular model of the ZAK–disome. The box highlights the site of the pin on the stalled ribosome, comprising the collision interface formed by ES7 and eS27 (both on stalled) and ES6c on the collided ribosome. b, Molecular model of the pin within the ZAK RBR, bound to eS27 on the stalled 80S (see also Extended Data Figs. 5a–e and 6a). c, Zoomed-in view on the collision interface as outlined in panel a. The dashed line indicates regions flanking the pin. d, Schematic of the ZAK domain architecture, with important residues in the RBR highlighted. Residues marked in red were targeted for mutational analysis. S, sensing domain. e, Immunoblots of sucrose gradient fractions collected from HEK293T ZAK-knockout cells transfected with partial CMV promoter plasmids expressing various N-terminal FLAG ZAK constructs at endogenous levels. Immunoblots were probed for FLAG to visualize ZAK. A representative eS24 blot is shown (Extended Data Fig. 3b). f, Immunoblots of total lysate from gradients shown in panel e. g, CLIP-seq of FLAG-tagged ZAK(T161A/S165A) expressed from the full CMV…

• 

  Fig. 3. ZAK interaction with RACK1 via RIH mediates ribosome binding. a, Zoomed-in view of the molecular model of the ZAK–disome focusing on the RACK1 collision interface. The boxes highlight the position of the RIH and the RIH-p. b, Zoomed-in view focusing on the RIH of the stalled 80S (RIH(s)) and the RIH-p extending towards the disome interface to form contacts with uS3 and h37–h38 of the collided 80S. c, Two views highlighting the interactions of RIH Y611 and V615 with RACK1. Additional contacts may be formed by R616 and R617 of ZAK with D220 and E262 of RACK1. d, Schematic of the ZAK domain architecture with RIH highlighted. Residues marked in red were targeted for mutational analysis. e, Immunoblots of sucrose gradient fractions collected from HEK293T ZAK-knockout cells transfected with partial CMV promoter plasmids expressing various N-terminal FLAG ZAK constructs at endogenous levels. Immunoblots were probed for FLAG to visualize ZAK (see Extended Data Fig. 3c for eS24 blots). f, Immunoblots of total lysate from gradients shown in panel e. g, Immunoblots of sucrose gradient fractions collected from HEK293T WT or RACK1-knockout cells. Immunoblots were probed for endogenous ZAK…

• 

  Fig. 4. RIM on ZAK mediates collision sensitivity and activation. a, Zoomed-in view on the molecular model of the ZAK–disome as in Fig. 3a. The boxes highlight the position of the RIM. b, Zoomed-in view highlighting the interactions between RACK1 and the ZAK RIM–F417 that caps the α5 of SAM, which binds into a groove clad by Q119 and N133 of RACK1 blade 3; L420 inserts into a hydrophobic pocket formed by F77, L89 and F113 between RACK1 blades 2 and 3. In addition, K422 forms a salt bridge with T93 and a hydrogen bond to Q76 of RACK1 blade 2 and H416 of SAM α5 contacts K139 in blade 3 of RACK1. Beyond F417, the RIM merges into the last α-helix of the SAM domain (α5) that extends into the solvent space above the 40S head. c, A 180° rotated view of b with RACK1 shown as surface colour coded according to lipophilicity potential (scale bar shown). d, Schematic of the ZAK domain architecture with the RIM highlighted. Residues marked in red were targeted for mutational analysis. Predicted FPxL motif residues are underlined. e, Immunoblots of sucrose gradient fractions collected from HEK293T ZAK-knockout cells transfected with partial CMV promoter plasmids expressing various N-terminal FLAG ZAK…

• 

  Fig. 5. SAM dimerization of ZAK regulates its kinase activity. a, Zoomed-in view on the molecular model of the ZAK–disome as in Fig. 3a. The boxes highlight the position of the hyperactive phenotype mutants (red dashed box) and SAM dimer interface (green dashed box) as shown in b. b, Location of hyperactive phenotype mutants within the SAM five-helix bundle (left) and zoomed-in view on amino acids involved in a network of salt bridges (highlighted by dashed lines) within the asymmetric head-to-tail interface of the SAM dimer (right). c, Schematic of ZAK SAM bridge interaction with characterized interface residues highlighted. d, Immunoblots of total lysate from HEK293T ZAK-knockout cells transfected with partial CMV promoter plasmids expressing various SAM mutant N-terminal FLAG ZAK constructs at endogenous levels. e, Zoomed-in view on low-pass-filtered cryo-EM maps (transparent) of the ZAK–disome (left) and ZAK(K394D)–disome (right), highlighting the RACK1 collision interface with molecular models for a RACK1-bound SAM dimer or SAM monomer (on RACK1(c)) and SERBP1 (on RACK1(s)) docked. f, Immunoblots of total lysate from HEK293T ZAK-knockout cells transfected with partial CMV promoter…

• 

  Extended Data Fig. 1. Conservation of ZAK and characterization of various in vivo and in vitro systems. a, Immunoblots from wildtype (WT) HEK293Ts and ZAK knockout HEK293Ts cells without or after addition of anisomycin (−/+ ANS) and with different ZAK (mutant) constructs expressed under a partial or full CMV promoter. Plasmid DNA concentrations for transfection indicated. b, Coomassie stained SDS gels of elution fractions from native pull-down of tagged ZAK mutants expressed in Expi293 cells and used for cryo-EM. See Extended Data Fig. 2 for joined data from the ZAK-K45M pull-down (left panel, ANS) and the ZAK-T161/165 A pull-down (both after ANS addition), Extended Data Fig. 3a for the pull-down from untreated cells (left panel, (−)) and Extended Data Fig. 9 for the ZAK-K45M-K394D pull-down (right panel) c, Consurf evolutionary conservation profile of certain ZAK regions of interest with RIM, RIH, eS27-pin and ES7-patch highlighted. Residues with red * were targeted for mutational analysis. d, Coomassie stained SDS gel of purified GST-tagged ribosome binding region (RBR) of ZAK. e, Coomassie stained SDS gels of in vitro binding assays using purified subunits (40S and 60S) or 80S ribosomes and…

• 

  Extended Data Fig. 2. Cryo-EM data analysis, classification and resolution of the ZAK-disome complex. The sample was obtained from a native pull-down using kinase inactive (K45M or T161/165 A) ZAK mutants after ANS addition (see Extended Data Fig. 1b, left and middle panels). a, Processing scheme: from a total of 95,813 micrographs 1,373,448 80S ribosomal particles were selected after 2D classification, 3D refined and classified. 80S classes with strong density for neighbors, indicative of stable disomes, displayed additional density at RACK1 (indicated with a black arrow), accounting for the ZAK SAM dimer. The two most abundant classes were classified with a soft mask focusing on the RACK1 region (shown in pink), revealing sub-classes with the ZAK SAM dimer varying in flexibility. Classes displaying strong SAM density were joined and refined to an overall resolution of 2.3 Å for both the stalled 80S and the collided 80S, and then subjected to local refinement. Local resolution was determined for the RACK1-SAM regions ranging from below 3 Å (for the RACK1-RIM interaction) to 5-8 Å for the globular SAM domain, revealing the position of the ZAK RIM and ZAK RIH. From a sub-dataset, maps for the entire…

• 

  Extended Data Fig. 3. Cryo-EM data analysis, classification and resolution of the ZAK-bound ribosomal complexes obtained without ANS addition and eS24 control blots. a, The sample was obtained from a native pull-down using the kinase inactive (K45M) ZAK mutant without ANS addition (see Extended Data Fig. 1b, left panel). 3D classification of 325,370 particles picked from a total of 13,154 micrographs showed 80S classes representing translating 80S with tRNAs in hybrid state (A/P, P/E) or with tRNAs in POST state (P/P, E/E) state, as well as 80S with a tRNA in the E site and bound to eEF2 and SERBP1, indicative of hibernating ribosomes. No classes indicative for stable disomes were found. The hybrid state translating and the hibernating 80S classes were first globally refined to 2.2 Å and 2.6 Å, respectively, followed by local refinement focusing on either the entire 40S subunit (2.3 Å for the translating and 2.7 Å for the hibernating 40S) or the RACK1 region (2.7 Å for the translating and 3.0 Å for the hibernating 40S head). Both maps revealed the ZAK eS27-pin, the ZAK RIH bound to RACK1 and the RACK1-bound SERBP1 C-terminal region. b, eS24 blots corresponding to Fig. 2e. c, eS24 blots corresponding…

• 

  Extended Data Fig. 4. AlphaFold models and density fits for ZAKα. a, AlphaFold model of ZAKα, colored according to a per-model confidence score (pLDDT; from 0 to 100). Blue regions display a very high confidence (pLDDT > 90), red/orange region low confidence (pLDDT <50). b, AlphaFold model of ZAKα with individual domains color coded. Regions not visible in our cryo-EM maps are shown in transparent grey except for KD (red), LZ (yellow), RBR (pale blue) and CTD (plum). KD = kinase domain; LZ = leucine zipper; RBR = ribosome binding region; CTD = C-terminal domain; RIM = RACK1-interacting motif; RIH = RACK1-interacting helix; RIH-p = RIH-peptide; pin = eS27-pin; YLD = YEATS-like domain; SAM = sterile alpha motif; helix = α-helix formed by residues 568–583 of ZAK. c, AlphaFold model of a SAM dimer with asymmetric head-to-tail interface. α-helices α1-5 are indicated. d, Three views on the AlphaFold model for SAM(s) (top) and SAM(c) fitted into respective isolated densities. The densities (mesh) were extracted from the local refined stalled and collided 80S (see also Extended Data Fig. 2) and were Gaussian low-pass filtered with a standard deviation of 0.75 in ChimeraX 1.9. ED indicates low-resolution…

• 

  Extended Data Fig. 5. Fitting of ZAK eS27 pin, RIH, RIM and SERBP1-C and model of EDF1 bound to the colliding disome. a, Top rows show cryo-EM maps of the individually refined stalled and collided 80S (Extended Data Fig. 2) from the dataset obtained from the kinase inactive ZAK pull-down after ANS treatment (+ANS). Below a view focusing on the eS27-pin, two views on RACK1-bound RIH and one view on the RACK1-bound RIM (with SAM α5) are shown. b, c, same as (a) for the untreated (-ANS) kinase inactive ZAK pull-down. Shown are cryo-EM maps for the refined hibernating (b) and hybrid state translating (c) 80S (see also Extended Data Fig. 3a). Views focusing on the ZAK eS27-pin, RIH and SERBP1-C (at the same RACK1 binding site as the ZAK RIM) and are shown as in (a). d, Cryo-EM map and view focusing on the eS27-pin for the reconstituted ZAK-RBR-40S complex. e, same as (a) with cryo-EM maps obtained from the kinase inactive ZAK-K394D pull-down after ANS treatment dataset (Extended Data Fig. 9). For all density snapshots focusing on RACK1 the cryo-EM maps are derived from local refinements on the RACK1-SAM region (see also Extended Data Figs. 2, 3a and 9). For (a) and (e) composite disome maps are shown with…

• 

  Extended Data Fig. 6. C-terminus structural validation and CLIP-seq analysis. a, Model of the eS27-pin fitted into the 2.3 Å cryo-EM map (mesh) of reconstituted ZAK-RBR-40S complex (see also Extended Data Fig. 5d). The main interaction with eS27 is mediated by W768, which intercalates between R80 and K36 of eS27. The following residues (T769, K770 and V771) are packed on top of the 3-stranded β-sheet of eS27 with T769 and V771 forming additional interactions with eS27. b, View focusing on the collision interface of the ZAK-disome complex surrounding the eS27-pin. The model for the ZAK-disome was fitted into the cryo-EM map of the entire stable ZAK-disome (Extended Data Fig. 2, right side). Dashed lines indicate regions flanking the eS27-pin. c, CLIP-seq of Flag tagged WT and T161A/S165A ZAK expressed from partial CMV promoter in HEK293T ZAK-KOs and its associated ribosomal RNA (black lines) in untreated and ANS treated conditions. Sequence reads were mapped to 18S rRNA and normalized to non-ribosome binding 1–649 aa ZAK truncation sample. The horizontal line represents 2-fold enrichment. Shading indicates standard error of the mean from two biological replicates. Reads from CLIP-seq were mapped to the…

• 

  Extended Data Fig. 7. Characterization of ZAK interactions in vivo and in vitro. a, Immunoblots of co-immunoprecipitation of FLAG tagged or HA tagged ZAK either in untreated or ANS treated conditions. b, Immunoblots of total lysate from HEK293T ZAK knockout cells transfected with partial CMV promoter plasmids expressing various SAM mutant N-terminal FLAG ZAK constructs at endogenous levels. Blots represent at least two independent replicates (see “Statistics and Reproducibility” section). See Supplementary Fig. 11 for source data.