Materials availability

Plasmids and strains generated in this study are available from the corresponding authors upon request. Materials, chemicals and reagents are listed in Supplementary Table 2.

Strains and plasmids

Yeast and bacterial strains are listed in Supplementary Table 2. Growth/culture conditions for the respective experiments are described in the method details.

Drug treatment of GFP/mCherry-tagged strains

The Saccharomyces cerevisiae leptomycin sensitive (crm1T539C) mutant strain chromosomally expressing Bud20-GFP and the mCherry-tagged nuclear membrane protein Nic96 was grown to OD600 0.6 in SD medium supplemented with all amino acids. Thereafter, LmB (135 ng/µl) and/or diazaborine (10 µg/ml) were added, and cells were inspected by fluorescence microscopy after 30 minutes of treatment. As negative control, untreated cells were inspected. A Leica DM6 equipped with a ×100 HC PL APO (1.4) objective was used for fluorescence imaging.

Pre-ribosomal particle purification for cryo-EM and SPR

Pre-60S particles were isolated from a LmB-sensitive S. cerevisiae mutant strain (crm1T539C) using Bud20-TAP as bait. Cells were grown to late log-phase (OD600 1–1.2) and then treated with 135 ng/µl µM LmB for 30 minutes. Cells were collected by centrifugation (4,500g, 1 minute, 25 °C) and stored at −20 °C. Affinity purification via the Tandem affinity purification (TAP) tag was performed as described38,51,52: crude extracts were prepared by breaking the cells in a Merkenschlager bead mill in the presence of 0.6 mm glass beads (4 minutes with CO2 cooling every 30 seconds) in buffer A (20 mM HEPES-NaOH, pH 7.5, 10 mM KCl, 2.5 mM MgCl2, 1 mM EGTA, 1 mM DTT, 0.5 mM PMSF and protease-inhibitor-mix FY (Serva)). The homogenate was centrifuged at 40,000g, 4 °C, 30 minutes. For cryo-EM analysis, export-competent particles were purified using homemade magnetic beads containing covalently linked rabbit IgG and subsequent TEV protease cleavage38. For SPR analysis, particles were purified on calmodulin resin using the calmodulin binding peptide of the TAP tag and subsequently eluted with EGTA. The particle concentration of the eluate was estimated by OD260 measurement of the RNA content in the sample.

Drg1 purification

GST-tagged wild-type Drg1 and the EQ1 (E346Q) variant were overexpressed in yeast as described12,16: the expression strain was inoculated to a starting OD600 of 0.01, incubated at 30 °C at 110 r.p.m. in baffled flasks. Protein expression was induced by immediate addition of 0.025 µM CuSO4, and cells were collected after 24 hours. Affinity purification was performed as described12,16,17,18: frozen cells were thawed in lysis buffer (50 mM Tris-HCl, 150 mM NaCl, pH 7.4, 1 mM DTT, 1× complete protease inhibitor cocktail (Roche) and 0.5 mM PMSF) and disrupted by vigorous shaking with 0.6 mm glass beads in a beadmill (Merckenschlager). Homogenates were centrifuged twice at 40,000g at 4 °C for 15 and 30 minutes, respectively. Crude extracts were incubated for 90 minutes at 4 °C with GSH-agarose beads (Sigma Aldrich) for affinity purification of GST-tagged Drg1. After consecutive washing steps (3× with lysis buffer plus 1 mM EDTA, 1 mM DTT and 1× with elution buffer plus 1 mM DTT), the protein was eluted. For elution, the GST tag was cleaved off using Prescission protease (GE Healthcare/Cytiva) overnight at 4 °C on a rotator in elution buffer suitable for the respective experiment. Protein concentration was measured using the Bradford assay (Biorad) with BSA calibration curve.

Assembly of the Drg1–pre-60S particle complex for cryo-EM

For reconstruction of the Drg1–pre-60S complex, purified Drg1 was eluted in cryo-elution buffer (20 mM HEPES-KOH, 150 mM potassium acetate, 5 mM magnesium acetate, 0.005% Tween-20, 1 mM DTT, pH 7.6), adjusted to 1.75 mg/ml (12 µM) and mixed in a 1:1 ratio with 200 nM export-competent pre-60S particles in the presence of 2 mM ATPγS and incubated for 10 minutes prior to plunge freezing. Quantifoil (R1.2/1.3) copper grids were glow-discharged for 60 seconds using an ELMO glow discharge system. Grids were loaded with 4 µl of the Drg1–pre-60S mixture and plunge frozen in liquid ethane using an FEI Vitrobot Mark IV at 4 °C and an environment with 100% humidity. The blotting force was set to 6, along with a blotting time of 7 seconds.

Cryo-EM imaging settings

Cryo-EM data for the Drg1–pre-60S complex were collected on a FEI Titan Krios G3i in conjunction with a Gatan K3 BioQuantum direct electron detector using a slit width of 20 eV. The camera was operated in counting mode using hardware binning (pixel size of 1.07 Å pixel−1) and dose fractionation. One movie contains 54 images resulting in a total dose of 60 e–/Å2 with a total exposure time of 4.84 seconds. The microscope was operated at 300 kV in nanoProbe energy-filtered transmission electron microscopy (EFTEM) mode at a nominal magnification of ×81,000. The dataset was acquired using SerialEM53 with an active beam tilt and astigmatism compensation. The acquisition scheme was designed to collect nine holes once per stage movement. Since the resulting maps showed preferred orientation, an additional, 34°-tilted dataset was collected with the same settings. The acquisition scheme was adapted to three holes per stage movement along the tilting axes. The required tilt angle was calculated with Relion v3.0.

Image processing

Image processing was mostly performed in Cryosparc v3.0 (ref. 54). Micrograph preprocessing included motion correction (patch motion correction) and CTF determination (patch CTF estimation). Micrographs and power spectra were individually inspected using the manually curate exposures function to exclude low quality micrographs due to ice contamination, devitrification or strong drift.

For the Drg1–pre-60S complex, 6,214 high-quality micrographs from the untilted dataset as well as 4,241 micrographs from the tilted dataset were used for particle picking. Initial 2D class averages generated from manually picked particles of the Drg1–pre-60S complex were used for template picking in both datasets. Selected 2D classes from both datasets were combined, resulting in an initial set of 3,645,306 particles. After multiple rounds of 2D classification, 1,782,014 particles were used to generate an initial ab initio 3D model. Subsequently, multiple rounds of heterogeneous refinement, as well as 3DVA followed by clustering, were performed to separate pre-ribosomal particle populations. On the basis of the conformation of the L1 stalk, export-competent Bud20-TAP particles (closed L1) were differentiated from earlier Bud20-TAP populations or mature 60S subunits (both open L1). Heterogeneity analysis was performed with CryoDRGN v.3.0.2 (ref. 55).

In a second picking round on the same dataset, exclusively particles representing Drg1 alone (not associated with pre-ribosomes) were picked. Initial 2D class averages generated from manually picked particles were used for automated template picking generating an initial set of 3,148,330 particles. Multiple rounds of 2D classification were used to clean the dataset and remove pre-ribosomes. 3D classification by heterogeneous refinement resulted in a final set of 114,728 particles that was used for refinement and 3DVA56.

3DVA (principal component analysis) allows sorting of structural variants from complex particle mixtures. It was performed on the refined particles in Cryosparc v3.0 with a filter resolution of 7 Å, followed by a 3D Var display job with eight frames. Visualization was performed as a volume series in UCFS ChimeraX v1.3. Conformational heterogeneity in the final particle populations for both the free Drg1 hexamers as well as the Drg1–pre-60S complex was detected and visualized using 3DVA.

After homogeneous refinement in Cryosparc v3.0, particles were transferred to RELION v3.0 using pyEM v0.5 (ref. 57) for further processing. Finally, maps were post-processed using DeepEMhancer58.

Model building and refinement

As an initial model the symmetric Drg1 hexamer was used (PDB: 7NKU)17. Model building was performed in Coot v0.9.6 (ref. 59), followed by refinement using Phenix v1.18.2-3874 (ref. 60) and ISOLDE v1.2.2 (ref. 61).

For the Bud20-TAP particle, initial models of the following components were taken from published early cytoplasmic pre-60S particles: Arx1 was taken from PDB 6RZZ28. Ribosomal proteins, 25S rRNA, Mrt4, Nog1, Bud20, Rlp24 and YBl028C) were taken from PDB 6N8K27. An initial model for L12 was taken from the mature 80S ribosome (PDB 4V6I). As an initial model for the ES27 rRNA segment, PDB 3IZD was used. Adjustment and real-space refinement of these models was performed in Coot v0.9.2 and phenix refine v1.18.2-3874, respectively. For molecular visualization, UCSF chimera v1.14 (refs. 62,63) and ChimeraX v1.3 were used64,65.

Drg1–pre-60S complex crosslinking MS (Arx1-TAP particle)

Pre-60S particles were purified via Arx1-TAP from a thermosensitive drg1-18 mutant after 1 hour of incubation at 37 °C as described previously12. After TEV cleavage, the eluate from 2 L of culture was incubated with purified Drg1 (150 µg) in the presence of 1 mM ATPγS for 30 minutes on a rotator at room temperature. Afterwards, one-third of the sample was supplemented with 1.5 mM isotopically labeled crosslinking reagent A (disuccinimidyl suberate, DSS-d0/DSS-d12, Creative molecules) and one-third with 1.5 mM isotopically labeled Crosslinker reagent B (bis(sulfosuccinimidyl)suberate, BS3-d0/BS3-d12, Sigma). One-third was further purified via Calmodulin sepharose beads without crosslinking. Crosslinking reactions were incubated at 30 °C in a thermomixer at 300 r.p.m. After 30 minutes, the reactions were quenched with 5 µl of a 1 M NH4HCO3 stock solution for further 10 minutes at 30 °C. Subsequently, samples were concentrated in a speedvac and stored at −20 °C.

Fractionation and enrichment of crosslinked peptides by SEC

Crosslinked samples were processed essentially as described28. In short, samples were dried (Eppendorf, Concentrator plus), resuspended in 8 M urea, reduced, alkylated and digested with trypsin (Promega). Digested peptides were separated from the solution and retained by a solid-phase extraction system (SepPak, Waters). Crosslinked peptides were enriched by size exclusion chromatography (SEC) using an ÄKTAmicro chromatography system (GE Healthcare) equipped with a Superdex Peptide 3.2/30 column (column volume, 2.4 ml). For each crosslinked sample four fractions were measured in technical duplicates. Therefore, the elution fractions 0.9–1.0 and 1.0–1.1 ml were pooled and the three elution fractions 1.1–1.2, 1.2–1.3 and 1.3–1.4 ml were analyzed separately by liquid chromatography with tandem MS (LC–MS/MS). Absorption levels at 215 nm of each fraction were used to normalize peptide amounts prior to LC–MS/MS analysis.

LC–MS/MS analysis

LC–MS/MS analysis was carried out on an Orbitrap Fusion Tribrid mass spectrometer (Thermo Electron). Peptides were separated on an EASY-nLC 1200 system (Thermo Scientific) at a flow rate of 300 nl/minute over an 80-minute gradient (5% acetonitrile in 0.1% formic acid for 4 minutes, 5%–35% acetonitrile in 0.1% formic acid in 75 minutes, 35%–80% acetonitrile in 1 minute). Full-scan mass spectra were acquired in the Orbitrap at a resolution of 120,000, a scan range of 400–1,500 m/z, and a maximum injection time of 50 ms. The most intense precursor ions (intensity ≥ 5.0 × 103) with charge states 3–8 and monoisotopic peak determination set to ‘peptide’ were selected for MS/MS fragmentation by CID at 35% collision energy in a data-dependent mode. The duration for dynamic exclusion was set to 60 seconds. MS/MS spectra were analyzed in the ion trap at a rapid scan rate.

Identification of crosslinked peptides

MS raw files were converted to centroid files and searched using xQuest/xProphet66 in ion-tag mode, with a precursor mass tolerance of 10 ppm against a database containing ribosomal proteins and known assembly factors (total of 184 proteins). For each experiment, only unique crosslinks were considered and only high-confidence crosslinked peptides that were identified with a delta score (deltaS) below 0.95 and an ld-Score above 32, translating to an FDR ≤ 0.5, were selected for this study. Crosslink networks were visualized with xiNet67.

Expression and purification of GST-Arx1

S. cerevisiae Arx1 was expressed as an N-terminal GST fusion from pGEX-6P-1 in Escherichia coli Rosetta (DE3) pLysS cells. Five hundred milliliters of LB medium (+100 µg/ml ampicillin and 40 µg/ml chloramphenicol) were inoculated to a starting OD600 0.04 and grown at 37 °C (170 r.p.m.). At an OD600 of 0.4, the culture was shifted to a shaking water bath at 16 °C, and after 30 minutes, heterologous protein expression was induced with 0.4 mM IPTG for 18 hours. Collected cells were washed once with aqua bidest and stored at −80 °C. For cell lysis, cells were resuspended in lysis buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM DTT, 0.5 mM PMSF, 1× HP protease inhibitor cocktail (Serva)) and incubated with 1 mg/ml lysozyme for 45 minutes prior to sonification. After removal of cell debris by centrifugation (40,000g, 25 min, 4 °C), the supernatant was incubated with GSH-agarose beads (Sigma Aldrich) at 4 °C for 60 minutes. Afterwards, beads were washed twice with lysis buffer, once with washing buffer (50 mM Tris-HCl, pH 7.4, 1 M NaCl, 1 mM DTT) and twice with storage buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 10% vol/vol glycerol, 1 mM DTT). The beads with the bound GST-Arx1 were resuspended in storage buffer, portioned, shock frozen in liquid nitrogen and stored at −80 °C.

GST pulldown

GST-Arx1 or GST-Rlp24 beads were thawed and washed once with binding buffer (20 mM HEPES-KOH, 150 mM potassium acetate, 5 mM magnesium acetate, 0.1% Tween-20, 1 mM DTT, pH 6.8). Eighty micrograms Drg1 were incubated with the GST-tagged bait proteins on the beads in binding buffer (120 µl) in the presence of either 1 mM ATP or ATPγS or no nucleotide for 60 minutes at room temperature under constant rotation. Empty GST was included as control for unspecific binding. For the GST-Arx1 pulldown, samples using GST-Rlp24C12 as bait protein were included as additional reference. After five binding-buffer washing steps, followed by centrifugation (500g, 1 minute, 4 °C), Arx1 and co-purifying Drg1 were separated from the beads by Prescission protease treatment (GE Healthcare) overnight, which cleaves off the GST tag. Samples using GST or GST-Rlp24C as baits were eluted by addition of 20 mM free GSH (Sigma Aldrich). The eluates were analyzed on a NuPAGE 4–12% Bis-Tris gel (Invitrogen).

ATPase activity assay (Malachite Green Phosphate Assay)

Drg1 ATPase activity was measured using the Malachite green phosphate assay68 as reported previously12,16. Essentially, purified Drg1 was eluted in 20 mM HEPES-KOH, 150 mM potassium acetate, 5 mM magnesium acetate, 0.1% Tween-20, 1 mM DTT, pH 6.8). HIS6-tagged Rlp24C or the indicated variants thereof were heterologously expressed in E. coli and purified as described in ref. 16. The activity of 5 µg/100 µl Drg1 was measured either alone (basal activity) or in the presence of 2 µg (0.8 µM) HIS6-Rlp24C and/or 2.7 µg (0.4 µM) Arx1. All samples contained 1 mM ATP (Sigma Aldrich). The released phosphate was quantified using the Malachite Green Phosphate Assay kit (Bioassay Systems). The absorbance of the samples at 600 nm was measured at a GeniusPro TECAN plate reader using a Microsoft Excel data collection plugin (XFluor4 v4.51). The specific activity (µmol ATP/hour/mg Drg1) of all samples was normalized to the Drg1 basal activity to display relative activities. Three biological replicates were measured with three technical replicates to determine mean and s.d.

Spot assays

Shuffle strains (drg1Δ, drg1Δ/arx1Δ or rlp24Δ) carrying the wild-type genes with their endogenous promoter and terminator on URA3 (pRS316) plasmids were transformed with plasmids expressing the indicated variants of Drg1, Arx1 or Rlp24. For plasmid shuffling, the transformed strains were grown in SD-leu (pRS315 plasmids) or in SD-his-leu (pRS315 and pRS313 plasmids) and spotted in a serial dilution on selective media and 5-FOA agar plates. For overexpression of Drg1 variants from the CUP1 promoter, the cells were spotted on selective medium containing different concentrations of CuSO4.

Surface plasmon resonance

SPR measurements were performed on a Biacore X100 (GE Healthcare/Cytiva). To analyze binding of Drg1 to the Rlp24C domain, purified GST-Rlp24C was immobilized as ligand on a CM5 sensor chip using the amine coupling kit (both Cytiva). Analogously, GST alone was immobilized in the reference flow cell. Drg1 was purified as described and eluted in elution buffer (20 mM HEPES, 10 mM KCl, 2.5 mM MgCl2, 100 mM NaCl, 0.05% Tween-20, 1 mM DTT, pH 7.5), which was also used as running buffer additionally supplemented with 1 mM ATPγS (or 1 mM ATP for the release assay). Then, 75–10,000 nM Drg1 supplemented with 1 mM ATPγS was injected with each cycle, composed of 180 seconds association, 120 seconds dissociation and 60 seconds regeneration of the chip surface with 1 M NaCl. Two biological replicates were measured on two separate CM5 chips with two technical replications (n = 4). For binding-affinity determination, measured RUs were normalized to the maximal response of each series (= 100%) and plotted over the concentration in Graphpad prism v3.03. The KD was determined by non-linear regression (one site binding hyperbola).

To measure binding of Drg1 to the export-competent particles, IgG antibodies were immobilized in both flow cells of CM3 chips (Cytiva) by amine coupling. Subsequently, purified export-competent particles still carrying the Protein A moiety of the TAP tag were captured (120-second contact time, resulting in 1,500–1,700 RU followed by a 30-second stabilization period) on the IgGs prior to injection of increasing concentrations of Drg1. All samples contained 1 mM ATPγS. Each injection cycle was composed of a fresh particle capturing step, followed by injection of Drg1 (180 seconds association, 120 seconds dissociation), and finally two 30-second pulses of 10 mM glycine-HCl pH 2.2 to regenerate the binding surface. Three biological replicates were measured on three separate chips (n = 3). For binding-affinity determination, measured RUs were normalized to the maximal response of each series (=100%) and plotted over the concentration in GraphPad prism v3.03. EC50 and Hill slope were determined by non-linear regression (sigmoidal dose–response with variable slope).

For the release assay, particles were analogously captured via IgGs, followed by injection of different Drg1 concentrations either containing 1 mM ATPγS (binding control), 1 mM ATP or 1 mM ATP + 100 µg/ml diazaborine. Linear segments (20 seconds) of the 37.5 nM Drg1 injections in the presence of ATP were used to quantify ΔRU/s. To determine the change in the number of bound particles, we used the manufacturer’s estimate (GE Healthcare Sensor Surface Handbook) that an SPR signal of 1 RU approximately corresponds to a surface concentration of 1 pg/mm2, which is based on an empirical determination using radiolabeled proteins69 and can further be converted to the volume concentration, taking into account the volume of the dextran matrix (see equation 1 in ref. 70). Given that the specific responses produced by biomolecules are largely independent of size69 and that the CM3 sensor chip used for the experiment provides only half the surface volume with the thickness of the dextran matrix being 50 nm, we calculated that a response of ~1,500 RU obtained by the injection of the 2.1 MDa pre-ribosomal particle corresponds to a surface concentration of 0.14 µM. This would give a concentration of ~9.5 nM for 1 RU of the 2.1-MDa pre-ribosomal particle. Finally, we related ΔRU/s to the Drg1 concentration (37.5 nM monomeric Drg1 corresponding to 6.25 nM hexamer) to estimate the rate of particle release per Drg1 hexamer. Four 37.5-nM injections from four biological replicates were used for the quantification (n = 4).

Statistics and reproducibility

The experiments described in Figures 1a,b and 3c,d,g,h and Extended Data Figs. 3b,d,e and 4 were performed twice, and representative results are shown. All attempts at replication were successful. Reproduction and sample numbers of the experiments described in Figs. 3e,i and 6a,d and Extended Data Fig. 3c are described in the respective figure legends and methods details. No statistical methods were used to predetermine sample size.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.