Cloning, plasmid construction and strains
All constructs were cloned using Gibson Assembly. For overexpression of the ligase complex in P. pastoris, DNA sequences coding for T. thermophilus, S. cerevisiae or C. thermophilum components were synthesized and codon optimized by GeneArt of Life Technology. Pex12 contained an N-terminal SBP tag followed by a HRV3C protease cleavage site, Pex10 contained a C-terminal haemagglutinin (HA) tag, and Pex2 contained a C-terminal FLAG tag. All genes were cloned between the EcoRI and NotI sites of the vector pPlCZA-LINK, which was modified from pPlCZA to allow the expression of multiple genes36. For expression of ring fingers in Escherichia coli, DNA sequences of S. cerevisiae genes were synthesized and cloned between the BamHI and NotI sites of the vector pGEX6P-1 containing an N-terminal glutathione S-transferase (GST) tag followed by a HRV3C protease cleavage site. Variants of all constructs were generated using Quikchange (Agilent).
The S. cerevisiae wild-type strain UTL7A (MATa, ura3-52, trp1, leu2-3/112) was kindly provided by R. Erdmann (Ruhr-University Bochum). All single and double knockout strains were constructed using standard transformation techniques and the pFA6A-NatMX or KanMX plasmids. The FLAG-tagged versions of Pex2, Pex10 and Pex12 were introduced at their endogenous locus by homologous recombination using the pFA6A-HygMX vector. For Pex5 overexpression, a HA tag followed by a 6×His tag was fused to the C terminus of full-length Pex5. The gene contained approximately 500 bp upstream of the ATG start codon, so that Pex5 was expressed from its endogenous promoter. The sequence was inserted into the multiple-cloning site of the CEN plasmid pRS416 (ref. 32). All transformations of S. cerevisiae cells were done with the LiAc-PEG method37.
Protein expression, purification and nanodisc reconstitution
The P. pastoris wild-type strain SMD1168 was obtained from Life Technology. Transformations were performed by electroporation, following the manufacturer’s instructions. Transformed yeast cells were grown on Zeocin-containing YPDS (YPD medium plus sorbitol) plates at 30 °C for 3 days. A single colony was picked to inoculate a starting culture, which was incubated at 30 °C overnight. A large culture was then inoculated by diluting the starter culture 1:100 into buffered minimal glycerol (BMG) medium. The culture was incubated at 30 °C for about 24 h and protein expression was induced by switching to the same volume of buffered minimal methanol (BMM) medium. After incubation for about another 20 h at 28 °C, the cells were harvested by centrifugation at 4,500g for 10 min. The pellet was stored at −80 °C until further use.
Cell pellets of about 100 g were resuspended in 100–120 ml buffer A (25 mM HEPES pH 7.4, and 400 mM NaCl) supplemented with 2 mM phenylmethane sulfonyl fluoride (PMSF) and 2 μM pepstatin A. The cells were lysed in a BioSpec Beadbeater for 40 min with 20 s/60 s on/off cycles in a water-ice bath. The homogenate was centrifuged at 10,000g for 20 min to remove cell debris. The supernatant was subjected to centrifugation in a Ti45 rotor (Beckman) at 44,000 r.p.m. for 1 h at 4 °C. The pelleted membranes were resuspended with a Dounce homogenizer in buffer A and pelleted again by centrifugation. The membranes were resuspended in 200–250 ml of buffer A containing 1% laurylmaltose neopentylglycol (LMNG) and a protease inhibitor cocktail and incubated for 60 min at 4 °C. Insoluble material was removed by centrifugation in a Beckman Ti45 rotor at 44,000 r.p.m. for 30 min. The supernatant was transferred to a new tube and incubated with 2 ml high-capacity streptavidin resin (Thermo Scientific) for 1 h. The resin was washed with 20–30 ml buffer A containing 0.1% digitonin (EMD Millipore), and bound protein was eluted with 10–15 ml of buffer B (25 mM HEPES pH 7.4, 150 mM NaCl, 10% glycerol and 0.1% digitonin) supplemented with 2 mM biotin (Sigma). The complex was concentrated with a 100-kDa cut-off Amicon filter (Sigma-Millipore) and further purified by size-exclusion chromatography on a Superose 6 3.2/300 Increase column (GE Healthcare), equilibrated with buffer C (25 mM HEPES pH 7.4, 150 mM NaCl and 0.05% digitonin).
For nanodisc reconstitution, a lipid stock was first prepared. The stock contained 10 mM each of 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE) and 1,2-dioleoyl-sn-glycero-3-phospho-l-serine (DOPS) and was prepared as in ref. 38. Protein purified in digitonin was incorporated into nanodisc using a 1:240:5 molar ratio of protein:lipid:membrane-scaffold protein 1D1 (MSP1D1). This mixture was incubated at 4 °C for 2 h with gentle agitation. Then, Bio-Beads (Bio-Rad) were added at 4 °C overnight with continuous rotation. The Bio-Beads were removed and the reconstitution mixture was centrifuged at 12,000g for 20 min to remove aggregated protein. The supernatant was loaded onto a Superose 6 3.2/300 Increase size-exclusion column (GE Healthcare) in gel-filtration buffer (25 mM HEPES pH 7.4 and 150 mM NaCl). Fractions containing nanodisc-reconstituted ligase complex were pooled and concentrated to about 10 μM. This material was used to generate Fabs.
To assemble a Fab–ligase complex, purified ligase complex in digitonin was incubated with Fab at a 1:1.5 molar ratio on ice for 1 h. The Fab–ligase complex was concentrated and loaded on a Superose 6 3.2/300 Increase size-exclusion column (GE Healthcare) in buffer C (25 mM HEPES pH 7.4, 150 mM NaCl and 0.05% digitonin). Peak fractions were pooled and concentrated to approximately 7 mg ml–1 for cryo-EM analysis.
For the purification of GST-tagged ring finger domains, the proteins were expressed in the E. coli strain BL21(DE3). A cell lysate was first subjected to centrifugation at 10,000g for 30 min at 4 °C, and the filtered supernatant was applied to a glutathione resin (GE Healthcare). After elution with reduced glutathione, proteins were concentrated and loaded onto a Superdex 200 3.2/300 Increase size-exclusion column (GE Healthcare) equilibrated in gel-filtration buffer (25 mM HEPES pH 7.4, 150 mM NaCl, 1 mM Tris (2-carboxyethyl) phosphine hydrochloride (TCEP) and 5% glycerol). Peak fractions were pooled and stored at −80 °C until use.
For purification of the Pex12 ring finger domain used for crystallization, GST-3C–Pex12 RF bound to the glutathione resin was incubated with PreScission protease at 4 °C overnight to cleave off the GST tag. The flow-through fraction was collected and loaded onto a Mono Q ion-exchange column (GE Healthcare). The protein was eluted with a salt gradient (0–500 mM NaCl and 25 mM HEPES pH 7.4) and peak fractions were pooled. The fractions corresponding to the Pex12 ring finger domain were concentrated and applied to a HiLoad 16/60 Superdex 75 gel-filtration column (GE Healthcare) pre-equilibrated in gel-filtration buffer (25 mM HEPES, pH 7.4, 150 mM NaCl, 1 mM TCEP and 5% glycerol). The purified protein was concentrated to 6 mg ml–1, flash-frozen in liquid nitrogen and stored at −80 °C.
Crystallization of the S. cerevisiae Pex12 ring finger domain
The Pex12 ring finger domain was crystallized by the hanging-drop vapour diffusion method by mixing 0.2 μl of the protein at 6 mg ml–1 with 0.2 μl of the 100 μl reservoir solution containing 0.1 M bis-Tris propane, pH 8.5, 0.2 M sodium fluoride and 20% PEG3350. Crystals were cryo-protected by gradually increasing the glycerol concentration in the drop by repeated additions of well solution supplemented with 30% glycerol. The crystal was removed from the drop and swiped through another drop of well solution supplemented with 30% glycerol and then flash-frozen in liquid nitrogen. Data were collected at 100 K at the 24-ID-C beamline at the Advanced Photon Source (APS). Data were indexed, integrated and scaled with HKL2000/3000 packages to 1.5 Å resolution.
Identification of ligase complex-specific Fabs using phage display
The purified T. thermophilus ligase complex was reconstituted into nanodiscs as described above, except that chemically biotinylated MSP1D1 was used. The efficiency of biotinylation was evaluated by capturing the nanodiscs with streptavidin-coated paramagnetic beads (Promega). For Fab selection, a previously described Fab phage library was used39. The nanodiscs containing the ligase complex and the library were diluted into selection buffer (20 mM HEPES pH 7.4, 150 mM NaCl and 1% BSA). Five rounds of sorting were performed as previously described40,41. In the first round, biopanning was performed manually using 400 nM of reconstituted ligase complex. To increase the stringency of selection pressure, four additional rounds of sorting were performed semi-automatically by stepwise reduction of the target concentration: 200 nM in the second round, 100 nM in the third round, 50 nM in the fourth round and 25 nM in the fifth round. All rounds of phage display were performed using a previously described protocol39. For each round except the first, the amplified phage population from each preceding round was used as the input pool. Of empty MSP1D1 nanodiscs, 2 μM was used throughout the selection as competitors.
Single-point phage ELISA to validate Fab binding to the reconstituted ligase complex
Single-point phage ELISA was performed to validate unique binders obtained from the fourth and fifth rounds of phage display. Sequencing of individual colonies harbouring phagemids was performed at the University of Chicago Comprehensive Cancer Center DNA Sequencing facility, and unique clones were selected and phages were amplified before ELISA as previously described40,41. ELISA was performed using a previously described protocol42.
Fab cloning, expression, purification and validation
Specific binders based on phage ELISA results were sequenced at the University of Chicago Comprehensive Cancer Center DNA Sequencing facility and unique clones were then subcloned into the Fab expression vector RH2.2 (gift of S. Sidhu) using the In-Fusion Cloning kit (Takara). Successful cloning was verified by DNA sequencing. Fabs were then expressed and purified as previously described35. Following purification, Fab samples were verified for purity by 4–20% SDS–PAGE and subsequently dialysed overnight in 25 mM HEPES pH 7.4 and 150 mM NaCl. Purified Fab affinities were estimated by multi-point ELISA42 using the ligase complex in biotinylated nanodiscs.
The ligase complex reconstituted in nanodiscs or in digitonin at a concentration of 0.02 mg ml–1 was applied to glow-discharged continuous carbon grids (Electron Microscopy Sciences, Inc.). After 1 min of adsorption, the grids were blotted with filter paper to remove excess sample, immediately washed twice with 4 μl of 1.5% freshly made uranyl formate solution and incubated with 4 μl of 1.5% uranyl formate solution for an additional 30 s. The grids were then further blotted with filter paper to remove the uranyl formate solution, air dried at room temperature and examined with a Tecnai T12 electron microscope (Thermo Fisher Scientific) equipped with an LaB6 filament and operated at 120 kV acceleration voltage, using a nominal magnification of ×52,000 at a pixel size of 2.13 Å.
Single-particle cryo-EM sample preparation and data acquisition
The concentrated sample was incubated with MS(PEG)12 methyl-PEG-NHS-ester (Thermo Fisher) at a 1:40 molar ratio for 2 h on ice to reduce the preferred orientation of particles on the grids43. Then, 3.0 µl PEGylated sample was applied to a glow-discharged quantifoil grid (1.2/1.3, 400 mesh). The grids were blotted for 7.0 s with a blot force of 12 at approximately 100% humidity and plunge-frozen in liquid ethane using a Vitrobot Mark IV instrument (Thermo Fisher Scientific).
Cryo-EM data were collected on a Titan Krios electron microscope (FEI) operated at 300 kV and equipped with a K2 Summit direct electron detector (Gatan) at the HHMI Janelia Farm Cryo-EM facility. An energy filter slit width of 20 eV was used to remove inelastically scattered electrons. All cryo-EM movies were recorded in super-resolution counting mode using SerialEM. The nominal magnification of ×81,000 corresponds to a calibrated physical pixel size of 1.06 Å and 0.53 Å in the super-resolution mode. The dose rate was 5.28 electrons Å−2 s−1. The total exposure time was 10 s, resulting in a total dose of 52.8 electrons Å−2 fractionated into 50 frames (200 ms per frame). The defocus range for the sample was between −1.0 and −2.5 µm.
A total of 9,019 dose-fractionated super-resolution movies were subjected to motion correction using the program MotionCor2 (ref. 44) with 2× binning, yielding a pixel size of 1.06 Å. A sum of all frames of each image stack (50 in total) was calculated following a dose-weighting scheme and used for all image-processing steps except for defocus determination. The program CtfFind4 (ref. 45) was used to estimate defocus values of the summed images from all movie frames. Particles were autopicked in Relion 3.1 (ref. 46). After manual inspection and sorting to discard poor images, classifications were done in Relion 3.1. A total of 3,224,513 particles was extracted and subjected to one round of reference-free 2D classification to remove false picks and obvious junk classes. To speed up 3D classification during data processing, the entire dataset was divided into four batches in the order of their collection, and each batch was subjected to 3D classification individually. Only one class of each batch showed protein features and particles from this class were combined for further classification (1,007,363 particles in total). Auto-refinement was done on this particle set using the reconstruction from previous 3D classification as initial model and a soft mask surrounding the protein and detergent micelle. After this round of refinement, particles were subjected to Bayesian polishing, followed by another round of auto-refinement and focused refinement using a mask encompassing the ligase complex and Fab. The refinement at this step yielded a 3.3 Å map. Using the angle assignments obtained after the focused refinement, a 1.8° local 3D classification (2 sigma and T = 20) with an adaptive mask was used to further classify the particles. A total of 795,444 particles from one class was selected and subjected to another round of auto-refinement. 3D classification (T30) without alignment, but with a mask, was used to further improve the quality of the map. After selection of 121,644 particles, a final round of auto-refinement followed by focused refinement using the adaptive mask yielded a map at 3.1 Å. Local resolutions were calculated with Resmap v1.1.5 (ref. 47) and map sharpening was performed in Relion 3.1. All reported resolutions are based on gold-standard refinement procedures and the Fourier shell correlation = 0.143 criterion. Histograms of directional Fourier shell correlation curves and sphericity values were calculated with the 3DFSC server48. All software is supported by SBGrid49.
Structural model building, refinement and analysis
All models were built in Coot50 and refined in PHENIX51 using the 3.1 Å sharpened density map. For Pex2, Pex10 and Pex12, transmembrane helices of each protein could easily be identified and were initially built as poly-Ala. We then manually fitted transmembrane helices into the density and traced the backbone of transmembrane segments and the linkers between them in Coot using secondary structure predictions and bulky residues (such as Phe, Trp and Arg) as sign posts. The density map was of sufficient quality to assign rotamers for key residues. In cases in which the rotamer could not be assigned, the side chain was stubbed at the Cβ atom. Models were refined in real space without secondary structure restraints using PHENIX real_space_refine. Strong non-crystallographic symmetry constraints in PHENIX real_space_refine were used to immobilize the domain that was not being refined. Several iterations of manual refinement and global refinement using Phenix and Coot were performed after visual inspection. For the ring finger domains, homology models of each RING domain were generated using RaptorX52 or Alphafold 2 (ref. 53). Then, these models were fit into the density map in UCSF Chimera54 and transferred to Coot for manual model building using secondary structure predictions from the XtalPred server as an additional guide. For Fab model building, the Fab portion of the deposited GgMFSD2A Fab complex (PDB ID: 7MJS) was used as a starting template and manually docked into the cryo-EM density with Chimera. The model was refined by iterative rounds of automated refinement. The structure of the Fab constant domain was removed due to its weak density. For models of lipids, PDB files of ergosterol or POPC were imported into Coot and fit into the density as ligands.
The Pex12 RF crystal dataset was collected at a wavelength of 0.967 Å, at which Zn2+ has a strong anomalous signal. This signal gave good-quality anomalous data that allowed SHELXD to locate zinc atoms in a straightforward manner. Phase probability distributions using this dataset and heavy atom sites were calculated with the SHARP program55. The quality of the derived phases allowed most of the Pex12 RF model to be automatically or manually completed in Coot. Refinement was carried out with REFMAC56.
Visualizations of the atomic models were made using UCSF Chimera54, ChimeraX57 and PyMOL (The PyMOL Molecular Graphics System, version 2.0, Schrödinger, LLC.).
Peroxisome protein import assay
The violacein pathway (VioA, VioB and VioE-SKL)19 was integrated into S. cerevisiae cells at the Leu2 locus. In brief, the plasmids pWCD1401 or pWCD1402 (ref. 58) were digested with NotI-HF and the gel-extracted 11.5-kB fragments were used for transformation. Clones were selected on SD-Leu plates. The strain containing the violacein pathway was used to generate knockouts of Pex2, Pex10 and Pex12, as described above.
To complement strains lacking Pex2, Pex10 or Pex12, FLAG-tagged versions of the wild-type proteins or of mutants were expressed from the endogenous locus of each gene. The genes were introduced by homologous recombination using the vector pFA6A-HygMX, as described above. The plasmids were transformed into cells containing the violacein pathway, but lacking a PEX gene, and selected on SD-Leu medium. Three colonies from each transformation were streaked on a SD-Leu plate, and a single colony from each was picked for overnight growth in YPD medium at 30 °C with shaking at 250 r.p.m. Saturated cultures were then diluted 50-fold into 3 ml of fresh SD-Leu medium and grown for about 60 h.
Extraction of the green pigment PDV was done as follows. The cell pellet was resuspended in 300 μl of glacial acetic acid and transferred to thin-walled Eppendorf tubes. The tubes were then incubated at 95 °C for 15 min, mixed by inversion and incubated for another 15 min. Cell debris were removed first by centrifugation for 5 min at 4,700 r.p.m. and then by filtration of the supernatant with an Acroprep Advance 0.2-μm filter plate (Pall Corporation). The filtrate was transferred to a 96-well non-transparent plate (Greiner Bio-one). Fluorescence was determined with a microplate reader (Bio-Tek Synergy Neo2), using excitation and emission wavelengths of 535 nm and 585 nm, respectively19.
In vitro polyubiquitylation assay
In vitro polyubiquitylation assays were performed in reaction buffer (25 mM HEPES pH 7.4, 150 mM NaCl, 10 mM MgCl2 and 50 μM TCEP) at 30 °C. The concentrations of the protein components were: 0.1 μM Uba1, 4 μM Ubc4, 0.5 μM GST–RFs and 100 μM ubiquitin. The mixture also contained 1 μM Dylight-Maleimide-800-labelled Cys-ubiquitin. The reaction was started by addition of 5 mM ATP and terminated after 60 min by addition of 4× SDS sample buffer. The samples were analysed by 4–20% SDS–PAGE and fluorescence scanning at 800 nm with an Odyssey scanner (Li-Cor).
Quantitative isobaric tag-based proteomics
The samples were prepared and analysed by liquid chromatography–tandem mass spectrometry, as previously described59. In brief, yeast cells were cultured in 50 ml YNBG medium (0.3% yeast extract, 0.5% peptone, 0.67% yeast nitrogen base with amino acid and 0.5% glucose, pH 6.0) overnight until log phase and then switched into YNBO medium (0.3% yeast extract, 0.5% peptone, 0.67% yeast nitrogen base without amino acid, 0.5% glucose, 0.05% Tween40 and 0.1% oleic acid, pH 6.0) for another 18 h to induce peroxisome proliferation. Cells pellets were resuspended in lysis buffer (8 M urea, 200 mM EPPS pH 8.5 and protease inhibitors (Pierce)) and then lysed using a BioSpec Beadbeater for five cycles, 30 s on followed by 60 s off per cycle. The homogenate was centrifuged at 2,000g for 10 min to remove the cell debris and the supernatant was transferred to a new tube. The sample was reduced with 5 mM TCEP for 30 min, alkylated with 10 mM iodoacetamide for 30 min and then quenched with 10 mM DTT for 15 min. Streamlined tandem mass tag labelling and liquid chromatography–mass spectrometry were all done following the protocol described in ref. 59. Quantitative isobaric tag-based proteomics data processing was done with MSconvert 3.0 (https://proteowizard.sourceforge.io/tools/msconvert.html) and Comet 2021.02 rev. 0 (http://comet-ms.sourceforge.net/).
Yeast cells lacking Pex2 and Pex12 and expressing FLAG-tagged Pex2 and Pex12 wild-type or mutant proteins from the native locus under the endogenous promoter were cultured in 50 ml YNBG medium overnight until log phase. The cells were transferred into YNBO medium for another 18 h to induce peroxisome proliferation. Cells were lysed using glass beads and cell debris were removed by centrifugation, as described above. Membrane fractions were isolated by ultracentrifugation at 45,000 r.p.m. for 60 min. Membranes were homogenized and solubilized in lysis buffer containing 1% Triton X-100 for 1 h. The extract was then incubated with 10 μl of anti-FLAG M2 resin for 2 h at 4 °C. The beads were washed three times with lysis buffer containing 0.1% Triton X-100, and bound proteins were eluted with buffer containing 0.4 mg ml–1 of 3×FLAG peptide (Sigma). Eluted proteins were subjected to SDS–PAGE and immunoblotting. FLAG-tagged Pex2 and Pex12 were detected using anti-FLAG (F7425, Sigma) antibodies at 1:3,000 dilution. As a loading control, total cell lysates were immunoblotted with anti-Sec61α antibody (homemade rabbit serum) at 1:3,000 dilution.
Statistics and reproducibility
All biochemical experiments were independently performed at least three times with similar results. A one-way analysis of variance with multiple comparisons was performed using GraphPad Prism 9.3.0 to evaluate the statistical significance of peroxisomal protein import efficiency in wild-type cells compared to import in mutants of Pex2, Pex10, Pex12 or Pex5. The bar graphs shown in Figs. 3d and 4c–g and Extended Data Figs. 5h and 8c,d show the individual data points, the mean and s.e.m. from three biological repeats. NS, not significant; *P < 0.1, **P < 0.05, ****P < 0.001. The quantitative isobaric tag-based proteomics experiment was performed independently twice with similar results. The statistics of results (Extended Data Fig. 9) were performed by multiple unpaired t-test followed by the method of two-stage step-up (Benjamini, Krieger and Yekutieli, desired false discovery rate (Q) = 1%) based on three biological replicates.
Further information on research design is available in the Nature Research Reporting Summary linked to this paper.