Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2024 Jun 17.
Published in final edited form as: Nat Struct Mol Biol. 2024 Mar 11;31(5):747–751. doi: 10.1038/s41594-024-01243-4

Structural basis of human U5 snRNP late biogenesis and recycling

Daria Riabov Bassat 1,#, Supapat Visanpattanasin 1,#, Matthias K Vorländer 1, Laura Fin 1, Alexander W Phillips 1, Clemens Plaschka 1,*
PMCID: PMC7616108  EMSID: EMS196538  PMID: 38467876

Abstract

Pre-mRNA splicing by the spliceosome requires the biogenesis and recycling of its small nuclear ribonucleoprotein complexes (snRNPs), which are consumed in each round of splicing. The human U5 snRNP is the ~1 megadalton “heart” of the spliceosome and is recycled through an unknown mechanism involving major architectural rearrangements and the dedicated chaperones CD2BP2 and TSSC4. Late steps in U5 snRNP biogenesis similarly involve these chaperones. Here, we report cryo-electron microscopy structures of four human U5 snRNP–CD2BP2–TSSC4 complexes, revealing how a series of molecular events primes the U5 snRNP to generate the ~2 megadalton U4/U6.U5 tri-snRNP, the largest building block of the spliceosome.

Pre-mRNA splicing is carried out by the dynamic and multi-megadalton spliceosome, which assembles anew on each intron from pre-formed small nuclear ribonucleoprotein particles (snRNPs; U1, U2, U4, U5, U6) and non-snRNP proteins. The U2-U6 snRNPs undergo major compositional and conformational changes throughout spliceosome assembly, activation, catalysis, and disassembly1–5. Among these, the U5 snRNP undergoes particularly dramatic changes, while serving as the ~1 megadalton “heart” of the spliceosome6 around which pre-mRNA, U1, U2, U4, and U6 snRNPs, and non-snRNP proteins organize for splicing (Fig. 1a). After spliceosome disassembly, the post-splicing U5 snRNP7 is recycled into a ‘20S U5 snRNP’8, which subsequently scaffolds the formation of the ~2 megadalton U4/U6.U5 tri-snRNP for the next round of splicing9. While the molecular basis of pre-mRNA splicing has been extensively studied1–5,10,11 the structural mechanism of U5 snRNP biogenesis and recycling remains unknown.

Figure 1. Structures of the human U5 snRNP.

Figure 1

Open in a new tab

a. Cartoon schematic of pre-mRNA splicing with focus on U5 snRNP recycling by CD2BP2 and TSSC4.

b. Scheme describing the largest differences between the four U5 snRNP structures determined in this study (State I, II, III, and IV; see also Extended Data Fig. 3).

c. U5 snRNP cryo-EM structure in State I shown in front and back views. Transparent models and asterisks on protein labels identify regions not seen in our cryo-EM densities, but for which structural data is available, including the flexibly linked PRP8RH domain (PDB ID 6QW6), the PRP6TPR (PDB ID 6QW6), the PRP8JAB1/MPN–BRR2Helicase–TSSC4153-318 (PDB ID 7PX3), and the CD2BP2 GYF–DIM1 heterodimer (PDB ID 1SYX). Note that DIM1 is sub-stoichiometric in the purified U5 snRNP.

d. U5 snRNP cryo-EM structure in State II. Labels as in panel c.

e. U5 snRNP cryo-EM structure in State III. The PRP8JAB1/MPN–BRR2Helicase–DDX23 (helix α5) are shown as a transparent model (PDB 6QW6). Labels as in panel c, but lacking TSSC4 and containing DDX23.

f. U5 snRNP cryo-EM structure in State IV. Labels as in panel e.

U5 snRNP recycling involves the U5-specific chaperones CD2BP29 (yeast Lin112) and TSSC413, but does not require ATP hydrolysis. While CD2BP2 and TSSC4 associate with the U5 snRNP, they are absent from the U4/U6.U5 tri-snRNP9,14. In the late steps of nuclear U5 snRNP biogenesis15,16, when biogenesis and recycling events become indistinguishable, CD2BP2 and TSSC4 also act on newly made U5 snRNPs through the same mechanisms9,13,14. During these biogenesis and recycling steps, the 11-subunit core of the U5 snRNP (U5 snRNA, PRP8, BRR2, SNU114, SNRNP40, seven-subunit Sm ring)6,11,17,18 incorporates four additional subunits (DIM1, PRP6, DDX23, USP39)9,19–21, and the U4/U6 di-snRNP, which is regenerated independently22. Further, essential domains of the U5 snRNP subunits PRP8 and BRR2 arrange into their U4/U6.U5 tri-snRNP conformations19–21. These combined events yield a new U4/U6.U5 tri-snRNP19 (Fig. 1a). Despite its importance, the structural basis of U5 snRNP biogenesis and recycling is not understood. Here, we report cryo-electron microscopy (cryo-EM) structures of four U5 snRNP complexes, which reveal how CD2BP2 and TSSC4 guide U5 snRNP biogenesis, recycling, and U4/U6.U5 tri-snRNP formation.

To gain structural insights into U5 snRNP biogenesis and recycling, we overexpressed GFP-tagged CD2BP2 in human K562 cells and purified the endogenous 20S U5 snRNP via the GFP-tag from nuclear extract. The purified complexes contained the U5 snRNP core, CD2BP2, TSSC4, PRP6 (Extended Data Fig. 1b, c, Supplementary Table 1), while DIM1, USP39, and subunits of the post-splicing U5 snRNP were sub-stoichiometric, including the NTC and NTR complexes. Purification of GFP-CD2BP2 complexes from cytoplasmic extract yielded U5 snRNPs of the same composition (Extended Data Fig. 1b), suggesting that late steps of U5 snRNP biogenesis and recycling may converge at CD2BP2 and TSSC4. We determined a consensus cryo-EM density of the nuclear 20S U5 snRNP from 256,427 images at an overall resolution of 2.6 Å (Extended Data Figs 1d, e, 2). Subsequent focused image classification yielded additional cryo-EM densities in a range of 2.7 Å to 4.6 Å resolution from which we built the atomic models of four different U5 snRNP complexes, named State I, II, III, and IV (Fig. 1c-f, Table 1, Extended Data Fig. 3, 6, Table 1). While the four states shared the same conformation of the U5 snRNP core, they differed in protein composition and the localization of peripheral U5 snRNP domains.

Table 1. Cryo-EM data collection and refinement statistics of human U5 snRNP complexes.

a
Map 1 Map 2 Map 3 Map 4 Map 5 Map 6
Data collection
   Particles 15,174 19,393 5,636 6,879 256,427 256,427
   Pixel Size (Å) 1.24 1.24 1.24 1.24 1.24 1.24
   Defocus range (μm) –0.5 to –2.5 –0.5 to –2.5 –0.5 to –2.5 –0.5 to –2.5 –0.5 to –2.5 –0.5 to –2.5
   Voltage (kV) 300 300 300 300 300 300
   Electron dose (e-/Å2) 40 40 40 40 40 40
Reconstruction (cryoSPARC)
   Resolution (Å) 3.8 3.2 3.9 4.6 2.7 2.6
   Map sharpening B-factor (Å2) –30 –30 –30 –30 –65 –60
Model composition
   Non-hydrogen atoms 28,086 28,612 31,351 41,435
   Protein residues 3,493 3,562 4,072 6,108
   Nucleotide residues 104 104 104 104
   Ligands 1 1 1 1
Refinement (PHENIX)
   Map CC (around atoms) 0.70 0.79 0.62 0.53
B-factor (overall) 219 145 148 786
Rms deviations
   Bond lengths (Å) 0.010 0.010 0.01 0.010
   Bond angles (°) 1.13 1.13 1.12 1.19
Validation
   Molprobity score 1.40 1.40 1.38 1.38
   All -atom clashscore 3.29 3.31 3.43 3.09
   Rotamer outliers (%) 0.08 0.08 0.04 0.04
   C-beta deviations 0 0 0 0
Ramachandran plot
   Outliers (%) 0.1 0.1 0.1 0.1
   Allowed (%) 4.0 3.9 3.6 3.9
   Favoured (%) 95.9 96.0 96.3 96.0
Data Deposition
   EMDB ID EMD-18234 EMD-18229 EMD-18235 EMD-18237 EMD-18239 EMD-18238
   PDB ID 8Q7V 8Q7Q 8Q7W 8Q7X
b
graphic file with name EMS196538-f009.jpg
Open in a new tab

The structures of U5 snRNP State I and State II contain CD2BP2 and TSSC4, differing only in a part of CD2BP2 (residues 153-231; Figs 1c, d, 2a, b, Extended Data Fig. 3). The chaperone CD2BP2 comprises three domains, which we name the ‘Hook’, the ‘Bundle’, and the previously identified GYF domain23. The CD2BP2 ‘Hook’ (residues 64-123) bridges the PRP8 N-terminal (PRP8N) and Large (PRP8L) domains and binds a bent conformation of the regulatory PRP8 α-finger11 (Fig. 2b, c, Extended Data Fig. 4d). The ‘Hook’ connects to the helical ‘Bundle’ (residues 153-231), which is likely to bind PRP8N reversibly: it is resolved in State II (Figs 1d, 2b), but is mobile in State I (Fig. 1c). The ‘Bundle’ extends to the CD2BP2 ‘GYF’ domain and its interactor DIM1 (ref.23), which are not resolved in our structures (Extended Data Fig. 2d, e). The chaperone TSSC4 (residues 77-98) binds the PRP8 reverse transcriptase (PRP8RT) domain (Figs 1c, d, 2a, b), consistent with prior biochemical data13. Previous data also show that this TSSC4 region is conserved in vertebrates and that deletion of TSSC4 residues 76-101 leads to defects in TSCC4–U5 snRNP association and in U4/U6.U5 tri-snRNP assembly13. The connected TSSC4 C-terminal region (residues 153-318) binds interfaces of the PRP8JAB1/MPN–BRR2 helicase (BRR2Helicase) domains, as shown in a previously determined cryo-EM structure of these isolated regions14. TSSC4 thereby prevents stable docking of BRR2Helicase on the U5 snRNP but may assist with anchoring the BRR2Helicase to the U5 snRNP, acting jointly with the PRP8-bound BRR2 N-terminus and the PRP8JAB1/MPN domains (Fig. 1b-f, Extended Data Fig. 3). These combined observations suggest functions for CD2BP2 and TSSC4: CD2BP2 locks the PRP8N and PRP8L domains through its ‘Hook’, arranging these domains near their eventual U4/U6.U5 tri-snRNP locations (Extended Data Fig. 4a), while TSSC4 tethers the mobile BRR2Helicase near its U5 snRNP binding site (Fig. 1c, d). Thus, both TSSC4 and CD2BP2 organize the U5 snRNP architecture for the integration of additional subunits.

Figure 2. Details of U5 snRNP–CD2BP2 and –TSSC4 interfaces.

Figure 2

Open in a new tab

a. Domain organization of CD2BP2 and TSSC4. Solid black lines show regions included in the atomic model, grey lines show regions previously determined in the CD2BP2 GYF–DIM1 crystal structure (PDB ID 1SYX) and the PRP8JAB1/MPN–BRR2Helicase–TSSC4 C-terminus cryo-EM structure (PDB ID 7PX3).

b. U5 snRNP–CD2BP2 and U5 snRNP–TSSC4 interfaces in the U5 snRNP. The CD2BP2 ‘Hook’ bridges the PRP8N and PRP8L domains (States I-IV), suggesting a mechanism for priming DIM1-binding (see Extended Data Fig. 4a). The TSSC4 ‘Tether’ (States I-II) binds the PRP8RT domain, indicating how it may assist U5 snRNP recycling and prevent premature U4/U6.U5 tri-snRNP assembly (see Extended Data Fig. 4a, 4f and main text for details).

c. CD2BP2 binds a conformation of the PRP8 α-finger bent at PRP8 Pro1530, which may be adopted during U5 snRNP biogenesis and relieved during U4/U6.U5 tri-snRNP assembly.

d. U4/U6.U5 tri-snRNP–DIM1 and U4/U6–U5 snRNP interfaces (PDB ID 6QW6) reveal interactions that are mutually exclusive with CD2BP2 and TSSC4 in the isolated U5 snRNP structures in States I-IV and I-II, respectively.

e. In the U4/U6.U5 tri-snRNP, the PRP8 α-finger adopts a helical conformation stabilized by interactions with U4/U6.U5 tri-snRNP subunits DIM1, PRP31, the PRP8Th/X, and U4 and U6 snRNAs (see also Extended Data Fig. 4 for details).

f. Gallery of human U5 snRNP structures visualized in the intron-lariat spliceosome (U5 snRNP core in the ILS; PDB ID 6ID1), in States II and IV of biogenesis and recycling, and in the U4/U6.U5 tri-snRNP (PDB ID 6QW6). The structures reveal a molecular cascade of conformational and compositional rearrangements that generate the U4/U6.U5 tri-snRNP for the next round of splicing. Colored arrows indicate large conformational rearrangements of U5 snRNP subunits between steps. The insets outline the U5 snRNP core in ILS and the complete U5 snRNP in the U4/U6.U5 tri-snRNP structures, respectively.

In U5 snRNP State III, TSSC4 is dissociated, and the ATPase domain of the U5 snRNP subunit DDX23 binds at the PRP8N domain. The DDX23 α-helix ‘-5’ is known to bind the PRP8JAB1/MPN–BRR2Helicase domains19,20, which would sterically clash with the TSSC4 C-terminus14. This suggests that competitive displacement of TSSC4 by DDX23 on the PRP8JAB1/MPN–BRR2Helicase domains dissociates TSSC4 from the U5 snRNP (Fig. 1d, e, Extended Data Fig. 3). In the subsequent U5 snRNP State IV, the DDX23 α-helix ‘-5’ and the associated PRP8JAB1/MPN–BRR2Helicase domains are docked at the PRP8RT domain, near their eventual U4/U6.U5 tri-snRNP locations (Fig. 1f). This docking is likely reversible, as it relies on small protein-protein interfaces, which become enlarged only in the U4/U6.U5 tri-snRNP through interfaces with additional proteins and the U4/U6 di-snRNP. The tethering and docking of the BRR2Helicase domain, via the stepwise actions of TSSC4 and DDX23, may thus prime the U5 snRNP for further assembly steps.

To gain insights into U5 snRNP recognition by CD2BP2 during biogenesis, we compared U5 snRNP State I to the crystal structure of an early yeast Prp8-Aar2 biogenesis intermediate24 (Extended Data Fig. 5a). This revealed that the cytoplasmic biogenesis factor Aar2 (ref.24) binds PRP8N and PRP8L surfaces that overlap with the conserved CD2BP2 ‘Hook’–PRP8 interfaces. Aar2 also binds the Prp8 α-finger, which is bent at a conserved Proline (yeast Pro1602, human Pro1530; Extended Data Fig. 5b). A highly similar conformation of the human PRP8 α-finger is bound by CD2BP2 (Fig. 2c), suggesting that the PRP8 α-finger may be a sensor of U5 snRNP biogenesis and recycling states, reminiscent of its regulatory roles during pre-mRNA splicing1–3,11. Future work is needed to understand the molecular exchange of AAR2 for CD2BP2, which may involve the nuclear phosphorylation of AAR216,25,26.

To investigate the mechanism of U4/U6.U5 tri-snRNP assembly, we compared U5 snRNP State IV to the human U4/U6.U5 tri-snRNP cryo-EM structure19–21. The essential U5 snRNP proteins DIM1 and PRP6 interact with each other, with CD2BP29,27,28 and scaffold the U4/U6.U5 tri-snRNP19–21 (Fig. 2d, Extended Data Fig. 1c). In our U5 snRNP State IV structure, we observe the N-terminus of PRP6 (residues 99-121) bound to PRP8N (Figs 1f, 2b), but its DIM1-interacting regions, DIM1 itself, and the PRP6 linker towards and including the PRP6 TPR repeat (PRP6TPR) were not resolved (Extended Data Figs 3, 4e, 4f). In contrast, these elements are stably bound in the U4/U6.U5 tri-snRNP (Fig. 2d, Extended Data Fig. 4f). In the U4/U6.U5 tri-snRNP, DIM1 binds similar PRP8N and PRP8L interfaces as the CD2BP2 ‘Hook’ in the U5 snRNP (Fig. 2d). However, DIM1 binds an altered conformation of the PRP8 α-finger in the U4/U6.U5 tri-snRNP, which forms an extended α-helix (Fig. 2e). This suggests an unexpected mechanism by which CD2BP2 could promote U4/U6.U5 tri-snRNP assembly and its own release (Fig. 2f). First, the CD2BP2 ‘Hook’ would chaperone PRP8N and PRP8L domains of the post-splicing U5 snRNP into defined locations (State I). The CD2BP2 ‘Bundle’ would bind PRP8N and deliver the CD2BP2 GYF–DIM123 complex to the nearby PRP6 N-terminus, tethering DIM1 near its PRP8-binding site, which is occluded by the CD2BP2 ‘Hook’ (States II-IV) (Extended Data Fig. 4e). Next, DDX23 would bind the U5 snRNP (State III), displacing TSSC4. The subsequent association of the U4/U6 di-snRNP proteins and USP39 could promote the stable association of the PRP6 N-terminal regions along PRP8N, PRP8 α-helical bundle (PRP8HB), and PRP8RT domains (Extended Data Fig. 4f). The PRP8HB would thus shift in position and straighten the bent PRP8 α-finger into an α-helix (Fig. 2d, e). The PRP8 α-finger would clash with CD2BP2, ultimately displacing CD2BP2 Hook-C from PRP8 (Fig. 2, Extended Data Fig. 4a, d) and in the same process generating a high-affinity binding site for DIM1. Once PRP6 and DIM1 dock in their U4/U6.U5 tri-snRNP binding sites, all CD2BP2–U5 snRNP interfaces would have been severed, releasing CD2BP2 from the complex. Thus, CD2BP2 controls its own release through the delivery of its cargo, DIM1, to its final binding site (Fig. 2f). After the CD2BP2–DIM1 exchange, PRP6 and DIM1 reside at the interface of U5 snRNP and U4/U6 di-snRNP, stabilizing the newly formed U4/U6.U5 tri-snRNP. In the future, the complete in vitro reconstitution of the human U4/U6.U5 tri-snRNP will be needed to probe these events in further detail.

Taken together, our data reveal how the U5 snRNP-specific chaperones CD2BP2 and TSSC4 facilitate the ATP-independent biogenesis and recycling of the human U5 snRNP and prime the assembly of the U4/U6.U5 tri-snRNP for a new round of splicing. The modes of CD2BP2 and TSSC4 action suggest a framework to understand how snRNP-specific chaperones use protein-protein interactions to position key, mobile snRNP domains for the efficient binding of snRNP-specific subunits or snRNAs.

Methods

Vectors and sequences

For endogenous purification of U5 snRNP–CD2BP2 complexes from human K562 cells (DSMZ), the CD2BP2 cDNA and an N-terminal GFP-3C tag were cloned into a lentiviral vector backbone (Addgene #31485) using overhang primers for CD2BP2 and eGFP (tcggcgcgccagtcctccgagccaccatggtgagcaagggcgagga, cgggccctgaaacagaacttccagcttgtacagctcgtccatgccga, ctggaagttctgtttcagggcccgatgccaaagaggaaagtgacct, aggttgattgttccagacgcggatctcaggtgtagaggtcaaagt), yielding a plasmid containing pRRL-SFFV-GFP-3C-CD2BP2.

Human GFP-3C-CD2BP2 K562 cell line

For endogenous purification of U5 snRNPs–CD2BP2, lentiviral particles carrying the GFP-3C-CD2BP2 construct were generated in Lenti-X 293T cells (Takara) via polyethyleneimine transfection (Polysciences) of the viral carrier plasmid pCMVR8.74 (Addgene #22036) and helper plasmid pCMV-VSV-G (Addgene #8454), according to standard procedures. K562 (DSMZ) cells were infected at limiting dilutions and GFP-positive cells were isolated using a BD FACSAria III cell sorter (BD Biosciences). Viral integration was confirmed by immunoblotting for CD2BP2 and GFP. Lenti-X and K562 cells tested negative for mycoplasma.

Preparation of nuclear and cytoplasmic extracts

To prepare nuclear (NE) and cytoplasmic (CE) extracts, 60 L of human K562 cells overexpressing GFP-3C-CD2BP2 were grown in RPMI media (20% FBS, Sigma Aldrich; 10% L-Glutamine, Gibco; 10% Ampicillin streptomycin; 10% Na-pyruvate, Sigma) to a 1.5x106 density at 37°C, 5% CO2, stirred at 70 rpm. The NE and CE fractions were prepared as previously described29 and dialysed against buffer E (20 mM HEPES pH 7.9, 100 mM KCl, 20% (w/v) glycerol, 0.2 mM EDTA, 2 mM DTT).

Purification of endogenous U5 snRNP complexes

10 ml of GFP-3C-CD2BP2 K562 NE (or CE) was incubated with GFP-Trap Agarose resin (Chromotek) pre-equilibrated with binding buffer F (20 mM HEPES pH 7.9, 75 mM KCl, 2 mM MgCl2, 0.05% (w/v) NP-40, 8% (v/v) Glycerol, 1 mM DTT, cOmplete EDTA-free protease inhibitor cocktail (Roche)) for 2 hrs at 4°C under constant rotation. After five washes, the U5 snRNPs were eluted by cleavage using 3C PreScission Protease diluted in binding buffer for 1h. The eluate was loaded onto a GraFix30 10–30% w/v sucrose density gradient containing 0.05% glutaraldehyde in buffer G (20 mM HEPES pH 7.9, 100 mM KCl, 2 mM MgCl2, 2 mM TCEP) and centrifuged at 91,000 g for 16 hrs in 4°C in a SW60 Ti rotor (Beckman Coulter). The sedimentation coefficients were simulated using the CowSuite software (https://www.cow-em.de). Fractions containing the U5 snRNP were quenched for 15 min using a final concentration of 50 mM lysine, pooled, concentrated in a 0.5 mL 100kDa MWCO Amicon Ultra concentrator (Sigma) against buffer G and immediately used for EM grid preparation.

Western Blots

To assess the quality of NE and CE fractionation, samples were applied to SDS–PAGE in a ratio of 1:3.45 (NE:CE; according to extract dilution ratio), transferred onto a PVDF membrane (ThermoScientific) and probed with anti-PRP8 (ab190347, Abcam, dilution 1:1000), anti-CD2BP2 (ab241947, Abcam, dilution 1:1000), anti-HSP90 (HRP-conjugated, 79641S, Cell signalling, dilution 1:1000) and anti-HNRNPA1 (HRP-conjugated, ab198535, Abcam, dilution 1:5000) primary antibodies. Goat anti-Rabbit IgG-HRP (31466, Invitrogen, dilution 1:10000) was used as a secondary antibody. Antibody detection was performed with Amersham ECL Select Western Blotting Detection Reagent (GE Healthcare) and a ChemiDoc MP imaging system (Bio-Rad Laboratories).

Cryo-electron microscopy

4 μL of concentrated and crosslinked U5 snRNPs were applied to glow discharged R2/1 holey carbon grids coated with a home-made 2 nm continuous carbon layer. Grids were blotted at 8 °C and 80% humidity and plunged into liquid ethane using a Leica EM GP2. Data was collected on a ThermoFischer Titan Krios G4i, equipped with a cold FEG (field emission gun) operated at 300 keV in EFTEM, and equipped with a Thermo Scientific Selectris energy filter set to a slit width of 10 eV and a Falcon 4i direct electron detector operated in counting mode. Both datasets were collected at a pixel size of 0.74 Å pixel-1, a total dose of 40 e– Å-2 and a defocus range of –0.5 to –2 μm using Thermo Scientific EPU. The dose rate was 6 e– pixel-1 sec-1. The dataset comprised 21,574 micrographs.

Data processing

Pre-processing

Data was pre-processed using Warp v1.0931. CTF parameters were estimated with a spatial resolution of 6 by 6 and a fitting range from 25 Å to 3 Å. Motion-correction was performed with a spatial resolution of 6 by 6. We picked 788,114 particles in total in Warp using a custom BoxNet model and extracted these particles in RELION 4.032 using a box size of 3602 pixels. For initial classification, particles were binned to 1.24 Å pixel-1. The gold-standard Fourier Shell Correlation (FSC; 0.143 criterion) was used to determine resolution in cryoSPARC33.

Particle classification and refinement

The cryo-EM data was processed in cryoSPARC33 as described in Extended Data Fig. 2. Briefly, the initial reference map for processing the data set was obtained from the first 57,138 particles. These particles were subjected to 2D classification, and 7,224 particles were used to obtain an ab-initio reconstruction. This map was filtered to 30 Å and used as the reference map for processing the full data set. After two rounds of heterogenous classification with three classes and without a mask, we selected highest quality class 3 particles in each round and refined these using a global mask, yielding Map 6 at 2.6 Å resolution from 256,427 particles. We subsequently applied several masks for focused classifications and refinement to improve the local density quality in various U5 snRNP regions. A mask encompassing the PRP8L domain was used for focused 3D refinement yielding Map 5 at 2.7 Å resolution. 3D variability analysis of the Map 5 particles with 10 clusters, yielded cluster 1 from which we determined Map 2 at 3.2 Å resolution (State II). A mask around the complete U5 snRNP was used for 3D classification of the Map 6 particles with 50 classes. 3D refinement of the class 1 particles yielded Map 4 at 4.6 Å resolution, comprising State IV. 3D refinement of the class 11 particles yielded Map 3 at 3.9 Å resolution, comprising state III. 3D variability analysis of the PRP8N domain using 10 clusters and the Map 6 particles, revealed cluster 4 that refined to 3.8 Å resolution Map 1, and comprised State I.

Model building

To prepare the complete U5 snRNP model, we employed ModelAngelo34, to build PRP8, SNU114, CD2BP2 Bundle (residues 153-231), Sm ring, and U5 snRNA into Map 6. The CD2BP2 Hook domain (residues 64-112) was rigid-body fitted in Chimera X35 based on an Alphafold236 prediction, followed by model building in COOT37 and refinement in Namdinator38. Subsequently, we modeled CD2BP2 Hook-N (residues 64-112), BRR2 N-terminus (residues 36-63), and PRP8 endonuclease and Linker domains (residues 679-1755) in Map 5 using both COOT37 and ISOLDE39 in Chimera X35. The CD2BP2 Hook-C (residues 116-123) was modeled in Map 1; PRP8N, SNU114, PRP6 (residues 99-121) and TSSC4 (residues 77-98) were modeled in Map 6. DDX23 α-helix ‘-5’ was modeled in Map 4; DDX23, BRR2, and SNRNP40 were rigid-body fitted into Map 4, providing States III and IV of the recycling process. Model regions with low-resolution density were trimmed to poly-alanine. The model coordinates were refined into the respective sharpened maps in PHENIX40 using the phenix.real_space_refine routine and applying secondary structure and rotamer restraints. Figures were made with UCSF Chimera X35.

Extended Data

Extended Data Figure 1. Purification and cryo-EM imaging of the human U5 snRNP complexes.

Extended Data Figure 1

Open in a new tab

a. Endogenous CD2BP2 is detected by western blot in nuclear and cytoplasmic extracts (NE and CE) from K562 wild-type or K562 GFP-CD2BP2 overexpressing cells (representative image, n=4). Note that the western blot band for CD2BP2 migrates at a higher molecular weight in CE compared to NE fractions, indicating that post-translational modifications could regulate CD2BP2 activity. Consistent with this, CD2BP2 has been reported to be phosphorylated, which may influence its interactions with the U5 snRNP41.

b. GFP-CD2BP2 purifications from nuclear or cytoplasmic extracts (NE and CE) yield U5 snRNPs of similar protein composition (n=2).

c. Purification scheme (top) and SDS-PAGE analysis (bottom) of the human U5 snRNP complexes from K562 cells. The most abundant proteins in the samples are annotated according to mass spectrometry analysis of SDS-PAGE gel slices. The asterisk annotates contaminants (n=9).

d. Denoised cryo-EM micrographs of U5 snRNP complexes. The dataset contained 21,572 micrographs. Scale bar, 300 Å (n=1).

e. Cryo-EM 2D class averages of U5 snRNP complexes in State I, II, III, IV. The particles subsets for the shown States I-IV 2D class averages were obtained from U5 snRNP image processing shown in Extended Data Fig. 2. Scale bar, 100 Å.

Extended Data Figure 2. Cryo-EM image processing of the human U5 snRNP complexes.

Extended Data Figure 2

Open in a new tab

a. Three-dimensional image classification tree of the U5 snRNP cryo-EM data. The dataset contained 21,574 micrographs from which 788,114 particles were picked and extracted. An initial volume was generated from 57,138 particles in cryoSPARC33 using the ab-initio reconstruction algorithm, which served as a reference volume to classify the entire dataset using two rounds of heterogenous classification (see Methods). The cleaned U5 snRNP particle stack contained 256,427 particles and was refined to an overall resolution of 2.6 Å. Focused three-dimensional classifications and subsequent refinements, yielded six cryo-EM densities, labeled 1-6, in resolutions varying from 2.6 to 4.6 Å.

b. Gold-standard Fourier Shell Correlation (FSC = 0.143) of U5 snRNP cryo-EM maps 1-6.

c. Orientation distribution plots for all particles contributing to the respective U5 snRNP cryo-EM maps 1-6.

d. Composite U5 snRNP cryo-EM maps 1-6 shown from the front view, colored by local resolution as determined by cryoSPARC33.

e. As panel d. but the maps are colored by U5 snRNP subunits as in Fig. 1.

f. Gallery of cryo-EM densities from the U5 snRNP subunits CD2BP2 (map 5), TSSC4 (map 5), SNU114 (map 6) and U5 snRNA (map 6), superimposed on the final coordinate models.

Extended Data Figure 3. Structures of the human U5 snRNP State II and IV.

Extended Data Figure 3

Open in a new tab

a, b. Back views of the final U5 snRNP coordinate models of States II (panel a) and IV (panel b), highlighting the differences PRP8–DDX23 interactions in State IV (and III) but absent from States II (and I). Colors as in Fig. 1. Transparent models indicate regions not observed in our densities.

Extended Data Figure 4. Structural comparisons of U5 snRNP to U4/U6.U5 tri-snRNP and CD2BP2–DIM1 structures.

Extended Data Figure 4

Open in a new tab

a. Comparison of U5 snRNP State II (this study) and U4/U6.U5 tri-snRNP structures (PDB ID 6QW6). For the U5 snRNP (left), PRP8N and PRP8L, U5 snRNA, TSSC4 and PRP6 are shown alone for clarity. For the U4/U6.U5 tri-snRNP (center) the same regions are shown, in addition to the PRP8-interaction segments of U4/U6 snRNAs, of PRP31, and complete DIM1. The structures were aligned on their PRP8L domains (right), with the U4/U6.U5 tri-snRNP rendered transparent. Black arrows indicate movements within U5 snRNP compared to the equivalent U4/U6.U5 tri-snRNP regions.

b. As in panel a, the structures were aligned on the PRP8L domain (right), to visualize the clash of the CD2BP2 ‘Hook-N’ (U5 snRNP, left) with PRP8α-finger–PRP31 interfaces and with DIM1 in the U4/U6.U5 tri-snRNP (center).

c. The CD2BP2 ‘GYF’ domain from the CD2BP2 GYF–DIM1 crystal structure (PDB ID 1SYX, left) clashes with PRP8L–PRP31 interfaces in the U4/U6.U5 tri-snRNP structure (PDB ID 6QW6, center). The CD2BP2 GYF–DIM1 crystal structure was aligned with the U4/U6.U5 tri-snRNP on DIM1 (right).

d. As in panel a, the structures were aligned on the PRP8L domain (right), to visualize the clash between the CD2BP2 ‘Hook’ (U5 snRNP, left) and the extended PRP8 α-finger (U4/U6.U5 tri-snRNP, center).

e. As in panel a, but the structures were aligned on the PRP8N domain (right) to visualize the clash between the CD2BP2 ‘Hook-C’ (U5 snRNP, left) and the PRP6 N-terminus (residues 72-99; U4/U6.U5 tri-snRNP, center).

f. As in panel a, but the structures were aligned on the PRP8RT domain (right) to visualize the clash between TSSC4 (U5 snRNP, left) and the PRP6 N-terminus (residues 236-281; U4/U6.U5 tri-snRNP, center).

Extended Data Figure 5. Structural comparison of the human U5 snRNP to the yeast Prp8-Aar2 structure.

Extended Data Figure 5

Open in a new tab

a. Comparison of PRP8L domains in the human U5 snRNP (State I, left) and the yeast Saccharomyces cerevisiae Prp8-Aar2 crystal structures24 (PDB ID 4I43, right). The PRP8 α-finger is highlighted with a black outline. Colors as in Fig. 1; Aar2 (light pink).

b. Comparison of PRP8 α-finger conformations in the human U5 snRNP (State I, left) and the yeast Prp8-Aar2 crystal structures24 (PDB ID 4I43, right).

Extended Data Figure 6. Fourier shell correlations between U5 snRNP cryo-EM maps and respective coordinate models.

Extended Data Figure 6

Open in a new tab

Fourier shell correlations between the cryo-EM maps 1, 2, 3, 4 with the respective refined coordinate models of U5 snRNP State I, II, III, and IV using phenix.mtriage.

Acknowledgments

We thank members of the Plaschka Group for their help and discussions; staff at the Protein Technologies Facility at the Vienna BioCenter Core Facilities (VBCF), a member of the Vienna BioCenter (VBC), for assistance with protein production; staff at the VBCF Electron Microscopy Facility, in particular T. Heuser and H. Kotisch, for support, data collection and maintaining facilities; staff at the IMP-IMBA-GMI BioOptics facility for performing FACS; R. Zimmermann and his team for computational support; K. Mechtler and his team for MS; staff at the in-house Molecular Biology Service for reagents; C. Bernecky for discussions; C. Bernecky, E. Bassat and members of the Plaschka group for critical reading of the manuscript. D.R.B. was supported by a Marie Sklodowska-Curie fellowship (101028744). M.K.V. was supported by an EMBO Postdoctoral Fellowship. C.P. was supported by Boehringer Ingelheim and the European Research Council (ERC-2020-STG 949081 RNApaxport). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript. For the purpose of open access, the author has applied for a CC BY public copyright licence to any Author Accepted Manuscript version arising from this submission.

Footnotes

Author contributions

D.R.B. generated the GFP-3C-CD2BP2 K562 cell line, and with A.W.P. grew cells and prepared nuclear extract. S.V., D.R.B., L.F., purified endogenous complexes and performed biochemical experiments. S.V. and M.K.V. collected cryo-EM data. D.R.B, S.V, M.K.V., and C.P. analysed the cryo-EM data and modelled the U5 snRNP structure. D.R.B. and C.P. analysed the data and prepared the manuscript with input from all authors. C.P. designed and supervised the project.

Competing interests

The authors declare no competing interests.

Supplementary Data

Supplementary Table 1 | Mass spectrometry of human U5 snRNP complexes.

Data availability

The 3D cryo-EM density maps 1, 2, 3, 4, 5, and 6 of the U5 snRNP have been deposited into the Electron Microscopy Data Bank under the accession numbers EMD-18234, EMD-18229, EMD-18235, EMD-18237, EMD-18239, EMD-18238. The coordinate file of the U5 snRNP in States I, II, III, and IV have been deposited into PDB under the accession numbers 8Q7V, 8Q7Q, 8Q7W, and 8Q7X.

References

  • 1.Kastner B, Will CL, Stark H, Lührmann R. Structural Insights into Nuclear pre-mRNA Splicing in Higher Eukaryotes. Cold Spring Harb Perspect Biol. 2019;11 doi: 10.1101/cshperspect.a032417. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Wilkinson ME, Charenton C, Nagai K. RNA Splicing by the Spliceosome. Annu Rev Biochem. 2020;89:359–388. doi: 10.1146/annurev-biochem-091719-064225. [DOI] [PubMed] [Google Scholar]
  • 3.Wan R, Bai R, Zhan X, Shi Y. How Is Precursor Messenger RNA Spliced by the Spliceosome? Annu Rev Biochem. 2020;89:333–358. doi: 10.1146/annurev-biochem-013118-111024. [DOI] [PubMed] [Google Scholar]
  • 4.Will CL, Lührmann R. Spliceosome structure and function. Cold Spring Harb Perspect Biol. 2011;3 doi: 10.1101/cshperspect.a003707. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Vorländer MK, Pacheco-Fiallos B, Plaschka C. Structural basis of mRNA maturation: Time to put it together. Current Opinion in Structural Biology. 2022;75:102431. doi: 10.1016/j.sbi.2022.102431. [DOI] [PubMed] [Google Scholar]
  • 6.Grainger RJ, Beggs JD. Prp8 protein: at the heart of the spliceosome. Rna. 2005;11:533–57. doi: 10.1261/rna.2220705. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Golas MM, et al. 3D Cryo-EM Structure of an Active Step I Spliceosome and Localization of Its Catalytic Core. Molecular Cell. 2010;40:927–938. doi: 10.1016/j.molcel.2010.11.023. [DOI] [PubMed] [Google Scholar]
  • 8.Bach M, Lührmann R. Protein-RNA interactions in 20S U5 snRNPs. Biochim Biophys Acta. 1991;1088:139–43. doi: 10.1016/0167-4781(91)90164-h. [DOI] [PubMed] [Google Scholar]
  • 9.Laggerbauer B, et al. The human U5 snRNP 52K protein (CD2BP2) interacts with U5-102K (hPrp6), a U4/U6.U5 tri-snRNP bridging protein, but dissociates upon tri-snRNP formation. Rna. 2005;11:598–608. doi: 10.1261/rna.2300805. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Liu Y-C, Cheng S-C. Functional roles of DExD/H-box RNA helicases in Pre-mRNA splicing. Journal of Biomedical Science. 2015;22:54. doi: 10.1186/s12929-015-0161-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Plaschka C, Newman AJ, Nagai K. Structural Basis of Nuclear pre-mRNA Splicing: Lessons from Yeast. Cold Spring Harb Perspect Biol. 2019;11 doi: 10.1101/cshperspect.a032391. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Bialkowska A, Kurlandzka A. Proteins interacting with Lin 1p, a putative link between chromosome segregation, mRNA splicing and DNA replication in Saccharomyces cerevisiae. Yeast. 2002;19:1323–33. doi: 10.1002/yea.919. [DOI] [PubMed] [Google Scholar]
  • 13.Klimešová K, et al. TSSC4 is a component of U5 snRNP that promotes tri-snRNP formation. Nature Communications. 2021;12:3646. doi: 10.1038/s41467-021-23934-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Bergfort A, et al. The intrinsically disordered TSSC4 protein acts as a helicase inhibitor, placeholder and multi-interaction coordinator during snRNP assembly and recycling. Nucleic Acids Res. 2022;50:2938–2958. doi: 10.1093/nar/gkac087. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Boon KL, et al. prp8 mutations that cause human retinitis pigmentosa lead to a U5 snRNP maturation defect in yeast. Nat Struct Mol Biol. 2007;14:1077–83. doi: 10.1038/nsmb1303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Weber G, et al. Structural basis for dual roles of Aar2p in U5 snRNP assembly. Genes Dev. 2013;27:525–40. doi: 10.1101/gad.213207.113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Laggerbauer B, Achsel T, Lührmann R. The human U5-200kD DEXH-box protein unwinds U4/U6 RNA duplices in vitro. Proc Natl Acad Sci U S A. 1998;95:4188–92. doi: 10.1073/pnas.95.8.4188. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Plaschka C, Lin P-C, Charenton C, Nagai K. Prespliceosome structure provides insights into spliceosome assembly and regulation. Nature. 2018;559:419–422. doi: 10.1038/s41586-018-0323-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Charenton C, Wilkinson ME, Nagai K. Mechanism of 5’ splice site transfer for human spliceosome activation. Science. 2019;364:362–367. doi: 10.1126/science.aax3289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Zhan X, Yan C, Zhang X, Lei J, Shi Y. Structures of the human pre-catalytic spliceosome and its precursor spliceosome. Cell Research. 2018;28:1129–1140. doi: 10.1038/s41422-018-0094-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Agafonov DE, et al. Molecular architecture of the human U4/U6.U5 tri-snRNP. Science. 2016;351:1416–20. doi: 10.1126/science.aad2085. [DOI] [PubMed] [Google Scholar]
  • 22.Bell M, Schreiner S, Damianov A, Reddy R, Bindereif A. p110, a novel human U6 snRNP protein and U4/U6 snRNP recycling factor. Embo j. 2002;21:2724–35. doi: 10.1093/emboj/21.11.2724. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Nielsen TK, Liu S, Lührmann R, Ficner R. Structural basis for the bifunctionality of the U5 snRNP 52K protein (CD2BP2) J Mol Biol. 2007;369:902–8. doi: 10.1016/j.jmb.2007.03.077. [DOI] [PubMed] [Google Scholar]
  • 24.Galej WP, Oubridge C, Newman AJ, Nagai K. Crystal structure of Prp8 reveals active site cavity of the spliceosome. Nature. 2013;493:638–43. doi: 10.1038/nature11843. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Jehle S, et al. A human kinase yeast array for the identification of kinases modulating phosphorylation-dependent protein-protein interactions. Mol Syst Biol. 2022;18:e10820. doi: 10.15252/msb.202110820. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Weber G, et al. Mechanism for Aar2p function as a U5 snRNP assembly factor. Genes Dev. 2011;25:1601–12. doi: 10.1101/gad.635911. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Liu S, Rauhut R, Vornlocher HP, Lührmann R. The network of protein-protein interactions within the human U4/U6.U5 tri-snRNP. Rna. 2006;12:1418–30. doi: 10.1261/rna.55406. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Schreib CC, et al. Functional and Biochemical Characterization of Dib1’s Role in Pre-Messenger RNA Splicing. J Mol Biol. 2018;430:1640–1651. doi: 10.1016/j.jmb.2018.04.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Mayeda A, Krainer AR. Preparation of HeLa cell nuclear and cytosolic S100 extracts for in vitro splicing. Methods Mol Biol. 1999;118:309–14. doi: 10.1385/1-59259-676-2:309. [DOI] [PubMed] [Google Scholar]
  • 30.Kastner B, et al. GraFix: sample preparation for single-particle electron cryomicroscopy. Nature Methods. 2008;5:53–55. doi: 10.1038/nmeth1139. [DOI] [PubMed] [Google Scholar]
  • 31.Tegunov D, Cramer P. Real-time cryo-electron microscopy data preprocessing with Warp. Nature Methods. 2019;16:1146–1152. doi: 10.1038/s41592-019-0580-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Scheres SHW. RELION: Implementation of a Bayesian approach to cryo-EM structure determination. Journal of Structural Biology. 2012;180:519–530. doi: 10.1016/j.jsb.2012.09.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Punjani A, Rubinstein JL, Fleet DJ, Brubaker MA. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat Methods. 2017;14:290–296. doi: 10.1038/nmeth.4169. [DOI] [PubMed] [Google Scholar]
  • 34.Jamali K, Kimanius D, Scheres S. ModelAngelo: Automated Model Building in Cryo-EM Maps. 2022 doi: 10.1038/s41586-024-07215-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Goddard TD, et al. UCSF ChimeraX: Meeting modern challenges in visualization and analysis. Protein Science. 2018;27:14–25. doi: 10.1002/pro.3235. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Evans R, et al. Protein complex prediction with AlphaFold-Multimer. bioRxiv. 2022:2021.10.04.463034 [Google Scholar]
  • 37.Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr. 2004;60:2126–32. doi: 10.1107/S0907444904019158. [DOI] [PubMed] [Google Scholar]
  • 38.Kidmose RT, et al. Namdinator - automatic molecular dynamics flexible fitting of structural models into cryo-EM and crystallography experimental maps. IUCrJ. 2019;6:526–531. doi: 10.1107/S2052252519007619. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Croll T. ISOLDE: a physically realistic environment for model building into low-resolution electron-density maps. Acta Crystallographica Section D. 2018;74:519–530. doi: 10.1107/S2059798318002425. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Afonine PV, et al. Real-space refinement in PHENIX for cryo-EM and crystallography. Acta Crystallogr D Struct Biol. 2018;74:531–544. doi: 10.1107/S2059798318006551. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Behrens S-E, Lührmann R. Immunoaffinity purification of a [U4/U6.U5] tri-snRNP from human cells. Genes Dev. 1991;5:1439–1452. doi: 10.1101/gad.5.8.1439. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The 3D cryo-EM density maps 1, 2, 3, 4, 5, and 6 of the U5 snRNP have been deposited into the Electron Microscopy Data Bank under the accession numbers EMD-18234, EMD-18229, EMD-18235, EMD-18237, EMD-18239, EMD-18238. The coordinate file of the U5 snRNP in States I, II, III, and IV have been deposited into PDB under the accession numbers 8Q7V, 8Q7Q, 8Q7W, and 8Q7X.

RESOURCES