Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Mar;627(8003):437-444.
doi: 10.1038/s41586-024-07093-w. Epub 2024 Feb 21.

The UFM1 E3 ligase recognizes and releases 60S ribosomes from ER translocons

Affiliations

The UFM1 E3 ligase recognizes and releases 60S ribosomes from ER translocons

Linda Makhlouf et al. Nature. 2024 Mar.

Abstract

Stalled ribosomes at the endoplasmic reticulum (ER) are covalently modified with the ubiquitin-like protein UFM1 on the 60S ribosomal subunit protein RPL26 (also known as uL24)1,2. This modification, which is known as UFMylation, is orchestrated by the UFM1 ribosome E3 ligase (UREL) complex, comprising UFL1, UFBP1 and CDK5RAP3 (ref. 3). However, the catalytic mechanism of UREL and the functional consequences of UFMylation are unclear. Here we present cryo-electron microscopy structures of UREL bound to 60S ribosomes, revealing the basis of its substrate specificity. UREL wraps around the 60S subunit to form a C-shaped clamp architecture that blocks the tRNA-binding sites at one end, and the peptide exit tunnel at the other. A UFL1 loop inserts into and remodels the peptidyl transferase centre. These features of UREL suggest a crucial function for UFMylation in the release and recycling of stalled or terminated ribosomes from the ER membrane. In the absence of functional UREL, 60S-SEC61 translocon complexes accumulate at the ER membrane, demonstrating that UFMylation is necessary for releasing SEC61 from 60S subunits. Notably, this release is facilitated by a functional switch of UREL from a 'writer' to a 'reader' module that recognizes its product-UFMylated 60S ribosomes. Collectively, we identify a fundamental role for UREL in dissociating 60S subunits from the SEC61 translocon and the basis for UFMylation in regulating protein homeostasis at the ER.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests

Figures

Fig. 1
Fig. 1. Cryo-EM structure of the 60S–UREL–UFC1–UFM1 complex.
a, ER-associated 60S ribosomal subunits are preferentially UFMylated in cells. The membrane fractions from WT HEK293 cells were solubilized and layered over a 10–30% sucrose gradient to separate individual ribosomal subunits. The 60S and 80S fractions were analysed for RPL26 UFMylation by immunoblotting (IB) using the indicated antibodies. The blot is representative of n = 3 independent experiments. Superscript values 2 and 3 refer to di- and tri-UFM1 RPL26 modifications, respectively. b, Composite cryo-EM density of the UREL ligase complex and UFM1 bound to the 60S ribosome. The 60S ribosome map is coloured in grey, and the UREL and UFM1 density map is coloured by protein as shown in c. c, Schematic of the domain architecture of UREL components (UFL1, UFBP1, CDK5RAP3) and the UFC1–UFM1 mimic. WH, winged-helix domain. NTD, N-terminal domain. d, Superpositions of UREL with published ribosome structures highlighting clashes with 60S ribosome-binding components. Left, the UFL1 C-terminal domain (CTD) occupies A-, P- and E-tRNA-binding sites and clashes with superimposed A/P and P/E tRNA (Protein Data Bank (PDB): 6W6L). Middle, the UFL1 CTD clashes with the 40S subunit (PDB: 6ip8). Right, the UFBP1 helix clashes with the SEC61 translocon (PDB: 6r7q). The arrows indicate clashes. Source Data
Fig. 2
Fig. 2. Summary of key interacting modules of the 60S–UREL–UFC1–UFM1 complex.
UFL1 interacts with the PTC (i). Top, the UFL1 C-terminal domain sits within the tRNA-binding groove and is clamped between 28S rRNA helices 69 (h96) and 38a (h38a). Bottom, the UFL1 loop (residues 436–448; solid line) reaches into the PTC and binds proximal to the P-site. Disordered connecting residues are shown as dashed lines. The substrate-recognition module (ii). UFL1 and CDK5RAP3 bind to RPL10a from the L1 stalk. The ribosome docking module (iii). The UFBP1 helix docks onto the ribosome adjacent to the peptide exit tunnel, near 28S rRNA helices 47 and 24, as well as RPL31. The catalytic module of the UREL complex (iv). UFM1 is bound by CDK5RAP3 and UFBP1, with UFM1 positioned directly above the C terminus of RPL26 near the site of UFMylation.
Fig. 3
Fig. 3. The UFIM motif of UFBP1 binds to UFM1.
a, View of the catalytic region of the UREL complex bound to the 60S ribosome showing the UFBP1 UFIM and CDK5RAP3 N terminus bound to UFM1. The distance between the C termini of UFM1 and RPL26 in our cryo-EM model is shown as a black dashed line. Missing residues of UFM1 and RPL26 are shown as dashed lines in their respective colours. b, 3DFlex was used to generate motion models of the UREL ligase complex. The black arrows indicate the direction of motion observed. c, AlphaFold prediction of the UFBP1 UFIM–UFM1 complex. Inset: an enlarged view of the UFBP1 β1 and UFM1 β2 interface; key interactions are highlighted. d, Cryo-EM density map of UFMylated 60S ribosomes shown in transparent grey with a model of UREL ligase bound to 60S rigid-body fitted into the density.
Fig. 4
Fig. 4. The mechanism of E2 recognition and UFMylation.
a, The crystal structure of E3mUU(ΔUFIM) bound to UFC1 shown as a cartoon representation. Inset: enlarged view of the hinge region that connects UFL1 WH1 to α1. b, Enlarged view of the interaction between UFC1 α2 and UFL1 α1. c, Immunoblot analysis of in vitro 60S ribosome UFMylation in the presence of the indicated E2 mutants that are defective in E3 binding. The blot is representative of n = 3 independent experiments. d, Lysine-discharge assays in the presence of E3mUU bearing truncations in the hinge region. The asterisk indicates UFL1(Δ21–24) and UFL1(Δ25–28). The Coomassie-stained SDS–PAGE gel is representative of n = 3 independent experiments. e, CDK5RAP3 and UFL1 form a composite binding site for the UFC1–UFM1 conjugate. Top, SEC elution profiles of UFL1(Δα1)–UFBP1–UFC1–UFM1, UFL1(Δα1)–UFBP1–CDK5RAP3 and UFL1(Δα1)–UFBP1–CDK5RAP3–UFC1–UFM1 complexes. Bottom, the corresponding peak fractions were separated on a 4–12% SDS–PAGE gel under reducing conditions and Coomassie stained. A280, absorbance at 280 nm. f, The crystal structure of the UFC1–UFM1 conjugate shown as a cartoon representation. UFC1 is coloured in purple, UFM1 is coloured in orange and the isopeptide bond formed between UFM1 Gly83 and UFC1(C116K) is highlighted and shown as a ball and stick representation. Source Data
Fig. 5
Fig. 5. UFMylation of RPL26 mediates dissociation of the 60S ribosomal subunit from the ER translocon.
a, UREL–UFM1 superimposed onto the structure of 80S–SEC61 (PDB: 6R7Q). Modelling UREL binding to SEC61 translocon-bound ribosomes suggests that the ligase complex will destabilize the 60S–SEC61 interaction. The approximate position of the ER membrane and ER lumen relative to the SEC61 translocon are shown. b, Loss of UFMylation leads to an accumulation of 60S ribosomes with the translocon. SEC61β association with 60S ribosomes was analysed in membrane fractions from parental WT and CDK5RAP3-KO cells. Lysates were normalized to uniform RNA concentration (using absorbance at 254 nm) and fractionated on 10–30% sucrose density gradients. The fractions corresponding to 60S from three independent replicates were analysed by immunoblotting using the indicated antibodies. c, Model for 60S–SEC61 translocon recognition by UREL, showing the multistep catalytic mechanism leading to RPL26 UFMylation and the subsequent conversion of UREL to a reader that splits the 60S ribosomal subunit from the translocon (see Discussion). The diagram in c was created using BioRender.
Extended Data Fig. 1
Extended Data Fig. 1. Preparation of complexes for cryo-EM.
a, UREL selectively UFMylates 60S in vitro. Ribosome UFMylation by the addition of UBA5, UFC1, UFM1 and UREL (UFL1, UFBP1 and CDK5RAP3) to a mixture of 60S ribosomes and 2-fold excess of 80S ribosomes. The reaction mixture was separated on a sucrose density gradient (left) and the fractions analysed for RPL26 UFMylation by immunoblotting (right). Data are representative of n = 3 independent experiments. b, UREL binds to 60S ribosomes. UREL was incubated with a mixture of 60S and 2.5-fold excess 80S ribosomes, separated on a sucrose density gradient and analysed by immunoblotting with the indicated antibodies. c & d, UREL associates with UFMylated 60S ribosomes in cells. Membrane fractions from HEK 293 WT cells left untreated (top) or treated with anisomycin (bottom) were separated on a sucrose density gradient and the different fractions were immunoblotted using the indicated antibodies. e, In vitro UFMylation assay to generate UFMylated 60S ribosomes in the presence of UFL1/UFBP1 or UREL for LC-MS/MS analysis. f, The reaction products from (e) were analysed by LC-MS/MS to identify the UFMylation site, linkage type and to quantify UFMylation of RPL26 under different conditions. Abundance of K134-GG remnants in the presence of E1 and E2 alone (A), UFL1/UFBP1 (B) and UREL (C). (Un: Unmodified RPL26, Mono: MonoUFMylated RPL26, Di: DiUFMylated RPL26) (n > 2 technical replicates). g, Immunoblot showing UFMylation of RPL26 in the presence UFM1 WT, K69R, K69A or lysine-less UFM1(K0). h, Mode of UFMylation of RPL26 as inferred from LC-MS/MS and biochemical experiments from e to g. i, Coomassie stained SDS-PAGE gel of purified UREL, UFC1-UFM1 and 60S ribosomes used in the preparation of samples for visualization by cryo-EM. j, Coomassie stained SDS-PAGE gel showing in vitro reaction for the generation of UFC1-UFM1. k, Coomassie stained SDS-PAGE gel analysing stability of UFC1-UFM1. Purified UFC1-UFM1 was incubated with UFL1/UFBP1 at 37 °C to monitor hydrolysis of UFC1-UFM1 conjugate. The reaction was stopped at indicated time points and separated on a 4-12% SDS-PAGE gel followed by Coomassie staining. l, Schematic showing reconstitution of stable UREL:60S complexes for cryo-EM analysis. m, Preparation of stable UREL:60S complex as outlined in l. Source Data
Extended Data Fig. 2
Extended Data Fig. 2. Cryo-EM data processing pipeline.
Cryo-EM data processing steps to obtain cryo-EM maps for the 80S ribosome, 60S ribosome subunit, UREL ligase-bound 60S ribosome and UREL bound to RPL10a. ~2.2 million picked particles were extracted using a box size of 588 pixels (pix), rescaled to 128 pix. After several rounds of 2D classification ~1.6 million ribosome-like particles were selected. All ribosome-like particles were pooled to generate an initial 3D model, followed by 3D refinement. 3D variability analysis (3DVA) separated three major classes: 80S ribosome, 60S subunit and UREL-bound 60S (60S+ligase). These underwent further rounds of 3DVA and cryoDRGN particle sorting to obtain homogenous particles, which were then re-extracted using the original box size, followed by a final 3D refinement. To generate the ligase+RPL10a map, signal corresponding to the 60S ribosome was subtracted and the region corresponding to the UREL ligase and RPL10a was locally refined. This was then further refined using 3DFlex training and reconstruction. Final maps are coloured by local resolution. 3D angular distribution representation and FSC curves are shown, calculated using the gold standard FSC cutoff of 0.143.
Extended Data Fig. 3
Extended Data Fig. 3. UREL:60S subunit interactions.
Main interactions between UREL and the 60S ribosome. Throughout, side chains are displayed as ball and stick and hydrogen bonds shown as black dashed lines. a, UFL1 N-terminus and UFBP1 C-terminus form a composite winged helix domain (pWH/pWH’). b, 60S ribosomal proteins RPL10a, RPL11, RPL36a, RPL36 and RPL13 interact with UFL1. c, Hydrogen bond network between UFL1 winged helix domains WH1 and WH2 and RPL13. d, UFL1 and CDK5RAP3 bind to RPL10a of the L1 stalk. Atomic model cartoon is coloured by protein and cryo-EM density shown in transparent grey. RBD is CDK5RAP3 ribosome binding domain. e, Hydrophobic residues at the RPL10a:UREL interface. f, Hydrogen bonding residues at the RPL10a:UREL interface. g, Overview of CDK5RAP3 domains. RBD is RPL10a binding domain. CCD is coiled-coil domain. UUBD is UFM1/UFBP1 binding domain. h, Main electrostatic interactions between CDK5RAP3 RBD and UFL1. I, Main electrostatic interactions between CDK5RAP3 CCD and UFBP1. j, Immunoblotting of membrane fractions from HEK293 WT, CDK5RAP3 KO or UFSP2, ODR4 double KO cells untreated or treated with 200 nM anisomycin for 60 min. Asterisk indicates empty lane. k, Superposition of CCDC47 (PDB ID 7tm3) with cryoEM structure of UREL:60S complex shown in cartoon representation reveals similar mode of ribosome docking. Source Data
Extended Data Fig. 4
Extended Data Fig. 4. UFL1 loop binds near the P-site of the peptidyl transferase centre (PTC).
a, Comparison between P-site regions of the 60S only map versus the UREL-bound 60S map, viewed at similar thresholds. Additional density was observed in the ligase-bound 60S map which corresponds to UFL1 (green). b, UFL1 loop positioning within PTC with surrounding rRNA bases shown. Cryo-EM density for UFL1 loop shown in transparent green. Arrows indicate loop N- and C-termini. c, 28S A4548 moves towards P-site to stack with UFL1 Y443. Transparent grey 60S model represents non-ligase bound 60S (PDB ID 6r7q). Opaque grey 60S model is UREL bound 60S. d, 28S U4452 moves towards A-site to sit proximal to G437. Transparent grey 60S model represents non-ligase bound 60S (PDB ID 6r7q). Opaque grey 60S model is UREL bound 60S. e, UFL1 N439 stacks with 28S rRNA A3908 and hydrogen bonds with G3807 (dashed line). f, UFL1 R441 stacks with 28S rRNA A4385 and hydrogen bonds with surrounding phosphates of 28S rRNA (dashed lines).
Extended Data Fig. 5
Extended Data Fig. 5. Analysis of UFBP1 UFIM:UFM1 and UREL:UFC1~UFM1 interactions.
a, ITC titration curve and the corresponding fitting curve for UFM1 and UFBP1 UFIM (178-204). Data are representative of n = 2 independent experiments. b, (Left) In vitro UFMylation assays with UFBP1 UFIM mutants and immunoblotted with the indicated antibodies. Data are representative of n = 3 independent experiments. (Right) Quantitative representation showing percentage of di-UFMylation (mean ± SD; n = 3) in the in vitro UFMylation assays. c, Overlay of the cryo-EM structure of the UREL complex and the crystal structure of E3mUU in complex with UFC1. d, Close-up view showing the conformational change of the DDRGK motif of UFBP1, highlighted in red, in the cryo-EM structure. e, Gel filtration chromatograms of E3mUU-ΔUFIM:UFC1 complex. Approximately 40 µM E3mUU were incubated with 15 µM UFC1 for 15 min at 4 °C and loaded on a Superdex 200 3.2/300 gl column. The corresponding peak fractions were collected and analysed on a 4-12% SDS-PAGE gel followed by Coomassie staining. f, to k, Gel filtration chromatograms of E3mUU-ΔUFIM:UFC1 complexes with mutations at the UFL1 α1/UFC1 α2 interface. l, Lysine discharge assays in the presence of E3mUU mutants. (Top) Schematic describing the assay workflow. (Bottom) Coomassie stained SDS-PAGE gel showing aminolysis of UFM1 from UFC1~UFM1 in the presence of E3mUU-ΔUFIM mutants. m, and n, SEC elution profiles of E3mUU-ΔUFIM:UFC1 complexes with mutations disrupting the UFBP1 R265:UFC1 D50 interaction. Source Data
Extended Data Fig. 6
Extended Data Fig. 6. Binding curves for ITC experiments.
a – l, Representative ITC binding curves analysing interactions between indicated E3mUU proteins and UFC1, UFM1 or UFC1-UFM1. All experiments were performed in duplicates, the dissociation constants and stoichiometries were calculated based on both experiments. (Bottom) Summary of disassociation constants of the different E3mUU constructs with UFC1-UFM1, UFC1 and UFM1 measured by ITC. nd indicates that no binding was detected.
Extended Data Fig. 7
Extended Data Fig. 7. Existence of a composite binding site for UFC1~UFM1 on UREL.
a, Mapping crosslinked residues on the UREL:60S cryo-EM structure. Residues on CDK5RAP3 and UFL1 crosslinked with UFC1 are highlighted in yellow. b, Residues on UFBP1 and RPL26 crosslinked with UFC1 may constitute two distinct interfaces for charged E2 interaction. Cryo-EM model of UREL:60S is shown in cartoon representation and crosslinked residues shown as ball and sticks are highlighted in yellow. (c-f) Models depicting different intermediate stages of UFC1 prior to conjugation of UFM1 on RPL26. c, Model to show that UFC1 is potentially engaged closer to the interface formed by CDK5RAP3 and UFL1 as observed in XL-MS data shown in (a). Model was generated by superposition of crystal structures of E3mUU bound to UFC1 and cryo-EM structure of UREL:60S ribosome. d, Model generated by superposition of crystal structure of UFC1-UFM1 conjugate onto cryo-EM structure of UREL:60S ribosome to suggest UFC1’s proximity to RPL26 and UFBP1 as observed in the XL-MS data shown in (b). e, (Top) Individual SEC elution profiles for UFL1-Δα1/UFBP1, E3mUU, CDK5RAP3 and UFC1-UFM1. (Bottom) The corresponding peak fractions were separated on a 4-12% SDS-PAGE gel under reducing conditions and Coomassie stained. f, (Top) Gel filtration chromatograms of E3mUU:UFC1-UFM1, E3mUU:CDK5RAP3 and E3mUU:CDK5RAP3/UFC1-UFM1. (Bottom) The corresponding peak fractions were separated on a 4-12% SDS-PAGE gel under reducing conditions and Coomassie stained. g, ITC binding curves for preformed E3mUU/UFC1-UFM1 complex and CDK5RAP3. Data are representative of n = 2 independent experiments. The dissociation constant and stoichiometry were calculated based on both experiments. h, Control experiment for (g), where no UFC1-UFM1 was added to E3mUU. nd indicates that no binding was detected. Data are representative of n = 2 independent experiments.
Extended Data Fig. 8
Extended Data Fig. 8. Active conformation of UFC1~UFM1 and UFMylation-dependent SEC61 dissociation from 60S.
a, Comparison of apo and E3 bound states of E2~Ubiquitin conjugate (PDB IDs 3ugb, 4auq) with apo UFC1-UFM1 conjugate. Ubiquitin (dark orange) and UFM1 (light orange) are shown as surfaces overlaid on cartoons. UBE2D2/3 (green) and UFC1 (purple) are shown in cartoon representation. b, Enlarged view of the interface between UFC1 α0 and UFM1 with the interacting residues highlighted in ball and stick representation. c, Model depicting open and closed states of UFC1~UFM1 conjugate. Inset highlights the clash between UFC1 α0 and an incoming UFM1 suggesting a requirement to remodel UFC1 α0 to accommodate UFM1 in the closed-active state. d, Crystal structure of UFC1-UFM1 conjugate in an intermediate conformation superimposed onto cryo-EM structure of UREL:UFM1 bound 60S ribosome. C-terminal glycine of UFM1 (G83), catalytic cysteine to lysine mutation of UFC1 (C116K) and most C-terminal residue of RPL26 for which density is present in the cryo-EM map (V128) are depicted as circles. e, Crystal structure of Ubiquitin-E2 conjugate (PDB ID 7r71) in a closed conformation superimposed onto the cryo-EM structure of UREL:UFM1-bound 60S ribosome in the same view as (d). C-terminal glycine of ubiquitin (G76), catalytic cysteine to lysine mutation C85K of ubiquitin E2 (UBE2D2) and most C-terminal residue of RPL26 for which density is present in the cryo-EM map (V128) are depicted as circles. f, Quantification of immunoblots from (Figure 5b). UFMylated RPL26 band intensity normalized to intensity of total RPL26 (left) and SEC61β band intensities normalized to RPL10a. Data show mean ± SD. ***p < 0.0001 (Student t-test). Data are representative of n = 2 independent experiments. g, UFMylation mediates dissociation of 60S from the translocon. In vitro UFMylation reactions were performed on membrane associated 60S ribosomal subunit-SEC61 translocon complexes isolated from CDK5RAP3 KO cells. 60S-SEC61 complexes were incubated with UBA5, UFC1, UREL and UFM1 either in the presence or absence of ATP and the reaction products were separated on a sucrose gradient and analysed by immunoblotting with the indicated antibodies. Blot is representative of n = 2 independent experiments. h, Dissociation of 60S from the ER translocon requires the UFIM motif. In vitro 60S-SEC61 UFMylation and translocon dissociation assay performed as in (d) using UREL complex with the UFBP1 UFIM mutant F196A. Blot is representative of n = 2 independent experiments. i, Comparison of the UFL1 PTC loop (P-site) with the eRF1 catalytic centre (A-site; PDB ID 6ip8). eRF1 catalytic residues GGQ and UFL1 loop residues GGN coloured in yellow. j, Superimposition of UREL complex and NEMF:Listerin complex (PDB ID 3j92) bound to the 60S ribosome. Missing Listerin model is depicted as dashed line. Source Data

References

    1. Walczak CP, et al. Ribosomal protein RPL26 is the principal target of UFMylation. Proc. Natl Acad. Sci. USA. 2019;116:1299–1308. doi: 10.1073/pnas.1816202116. - DOI - PMC - PubMed
    1. Wang L, et al. UFMylation of RPL26 links translocation-associated quality control to endoplasmic reticulum protein homeostasis. Cell Res. 2020;30:5–20. doi: 10.1038/s41422-019-0236-6. - DOI - PMC - PubMed
    1. Peter JJ, et al. A non-canonical scaffold-type E3 ligase complex mediates protein UFMylation. EMBO J. 2022;41:e111015. doi: 10.15252/embj.2022111015. - DOI - PMC - PubMed
    1. Banerjee S, Kumar M, Wiener R. Decrypting UFMylation: How Proteins Are Modified with UFM1. Biomolecules. 2020;10:1442. doi: 10.3390/biom10101442. - DOI - PMC - PubMed
    1. Millrine, D., Peter, J. & Kulathu, Y. 1 A guide to UFMylation, an emerging post-translational modification. FEBS J.290, 5040–5056 (2023). - PMC - PubMed

MeSH terms