Structure of the Dicer-2–R2D2 heterodimer bound to a small RNA duplex – Nature

Expression and purification of the Dicer-2–R2D2 heterodimer

Dicer-2 and R2D2 were co-expressed in Sf9 insect cells using the Bac-to-Bac system (Invitrogen). The gene encoding Dicer-2 (residues 1–1722) was cloned into a modified pFastBac vector (Invitrogen), in which the N-terminal 6×His tag was replaced with an 8×His–GFP tag. Dicer-2 was also cloned into the modified pFastBac vector, in which the N-terminal 6×His tag and the following TEV protease cleavage site were replaced with an 8×His tag, to improve the yield of the purified protein. The gene encoding R2D2 (residues 1–311) was cloned into a modified pFastBac vector, in which the N-terminal 6×His tag was replaced with a 3×Flag tag. The sequences of the DNA oligonucleotides used for the vector construction are listed in Supplementary Table 2.

For the preparation of the Dicer-2–R2D2 heterodimer, Sf9 cells at a cell density of 3.0 × 10⁶ cells per ml in Sf900 II medium (Thermo Fisher Scientific) were co-infected with the baculoviruses expressing 8×His–GFP-tagged Dicer-2 and 3×Flag-tagged R2D2 and incubated for 72 h at 27 °C. The cells were then collected and disrupted in lysis buffer (30 mM Hepes-KOH, pH 7.4, 100 mM potassium acetate, 2 mM magnesium acetate, 0.5% NP-40, and 5% glycerol). The lysate was centrifuged at 40,000g for 30 min, and the supernatant was incubated with anti-Flag M2 affinity resin (Sigma-Aldrich) for 1 h. The resin was washed with wash buffer (30 mM Hepes-KOH, pH 7.4, 800 mM NaCl, 2 mM magnesium acetate, 1% Triton-X 100, and 5% glycerol), and the protein was then eluted with elution buffer (30 mM Hepes-KOH, pH 7.4, 300 mM NaCl, 2 mM MgCl₂, 0.1 mg ml⁻¹ 3×Flag peptide, 20 mM imidazole, and 5% glycerol). The eluted protein was incubated with Ni-NTA Agarose resin (Qiagen), and the protein was eluted with elution buffer (30 mM Hepes-KOH, pH 7.4, 300 mM NaCl, 2 mM MgCl₂, 300 mM imidazole and 5% glycerol). The eluted protein was treated with TEV protease and dialysed against dialysis buffer (30 mM Hepes-KOH, pH 7.4, 300 mM NaCl, 2 mM MgCl₂, 40 mM imidazole and 5% glycerol). The protein sample was passed through a Ni-NTA Agarose column, to remove the 8×His–GFP and TEV protease. The Dicer-2–R2D2 protein was further purified on a Superdex 200 10/300 Increase column (GE Healthcare), equilibrated in 30 mM Hepes-KOH, pH 7.4, 100 mM KCl, 2 mM MgCl₂, 1 mM DTT and 0.02% glycerol.

For the preparation of the Dicer-2–R2D2–siRNA complex, Sf9 cells were co-infected with the baculoviruses expressing 8×His-tagged Dicer-2 and 3×Flag-tagged R2D2, and the Dicer-2–R2D2 protein was purified using anti-Flag M2 affinity resin and Ni-NTA Agarose, as described above. The Dicer-2–R2D2 protein was further purified by chromatography on a HiLoad Superdex 200 16/600 column (GE Healthcare), equilibrated with 30 mM Hepes-KOH, pH 7.4, 100 mM potassium acetate, 2 mM magnesium acetate, 1 mM DTT, and 0.02% glycerol. The purified Dicer-2–R2D2 and the let-7 siRNA duplex (Ajinomoto Bio-Pharma) were mixed at a 1:2 ratio at room temperature, and the Dicer-2–R2D2–siRNA complex was purified by chromatography on a Superdex 200 10/300 Increase column, equilibrated in 30 mM Hepes-KOH, pH 7.4, 100 mM potassium acetate, 2 mM magnesium acetate, 1 mM DTT and 0.02% glycerol.

Cryo-EM sample preparation

The Dicer-2–R2D2 complex was concentrated to A₂₈₀ = 0.6, using a Vivaspin centrifugal filter device (100 kDa MW cut-off, Sartorius). The sample (3 μl) was applied to a freshly glow-discharged Cu 300 mesh R1/1 grid (Quantifoil), in a Vitrobot Mark IV (FEI) at 4 °C, with a waiting time of 30 s and a blotting time of 4 s under 100% humidity conditions. The Dicer-2–R2D2–siRNA complex was concentrated to A₂₈₀ = 0.9, using the Vivaspin centrifugal filter device. The sample (3 μl) was applied to a freshly glow-discharged Au 300 mesh R1/1 grid (Quantifoil), in a Vitrobot Mark IV at 4 °C, with a waiting time of 30 s and a blotting time of 4 s under 100% humidity conditions. The grids were plunge-frozen in liquid ethane cooled at liquid nitrogen temperature.

Cryo-EM data collection and processing

Cryo-EM data were collected using a Titan Krios G3i microscope (Thermo Fisher Scientific), running at 300 kV and equipped with a Gatan Quantum-LS Energy Filter (GIF) and a Gatan K3 Summit direct electron detector in the electron counting mode.

Micrographs for Dicer-2–R2D2 were recorded at a nominal magnification of ×105,000, corresponding to a calibrated pixel size of 0.83 Å at the electron exposure of 15.8 e⁻ per pixel per s for 2.30 s, resulting in an accumulated exposure of 53 e⁻ Å⁻². The data were automatically collected by the image shift method using the SerialEM software⁴¹, with a defocus range of −1.6 to −0.8 μm, and 2,745 movies were obtained and processed using RELION-3.1. From the 2,745 motion-corrected and dose-weighted micrographs, 1,688,210 particles were initially picked, and extracted at a pixel size of 3.66 Å. These particles were subjected to several rounds of 2D and 3D classifications. The selected 324,630 particles were re-extracted at a pixel size of 1.25 Å, and then subjected to 3D refinement, per-particle defocus refinement, beam-tilt refinement, Bayesian polishing⁴² and 3D classification with the mask focusing on Dicer-2 CRBD and R2D2. The selected 144,979 particles were subjected to 3D refinement, and subsequent postprocessing of the map improved its global resolution to 3.3 Å, according to the Fourier shell correlation (FSC) = 0.143 criterion⁴³. The local resolution was estimated by RELION-3.1.

Micrographs for Dicer-2–R2D2–siRNA were recorded at a nominal magnification of ×105,000, corresponding to a calibrated pixel size of 0.83 Å at the electron exposure of 15 e⁻ per pixel per s for 2.30 s, resulting in an accumulated exposure of 48 e⁻ Å⁻². The data were automatically collected by the image shift method using the SerialEM software, with a defocus range of −1.6 to −0.8 μm. In total, 3,663 movies were obtained, and the beam-induced motion correction, dose-weighting and CTF estimation were conducted similarly to those for Dicer-2–R2D2. From the 3,663 motion-corrected and dose-weighted micrographs, 2,181,396 particles were initially picked, and extracted at a pixel size of 4.15 Å. These particles were subjected to several rounds of 2D and 3D classifications. The selected 179,826 particles were then re-extracted at a pixel size of 0.99 Å, and subjected to 3D refinement, per-particle defocus refinement, beam-tilt refinement and Bayesian polishing. The particles were again subjected to 3D refinement, and subsequent postprocessing of the map improved its global resolution to 3.3 Å, according to the FSC = 0.143 criterion.

Model building and validation

The initial model of Dicer-2–R2D2 was built using Buccaneer⁴⁴, and the model was then manually built using COOT⁴⁵. The model of the Dicer-2–R2D2–siRNA complex was built based on the Dicer-2–R2D2 model. The density maps were improved with the DeepEMhancer program⁴⁶. The models were refined using Servalcat Refmac5 (ref. ⁴⁷), with external restraints prepared by ProSMART⁴⁸ and LIBG⁴⁹. The structures were validated using MolProbity⁵⁰ from the PHENIX package. In the Dicer-2–R2D2 complex, residues 1–7, 90–94, 254–272, 346–353, 426–435, 539–553, 693–700, 955–968, 1052–1063, 1120–1129, 1146–1169, 1414–1421, 1564–1605, 1674–168, and 1721–1722 of Dicer-2, residues 1–2, 70–89, 164–187, 212–214 and 310 of R2D2 are not included in the final model, since these regions are not well resolved in the density map. In the Dicer-2–R2D2–siRNA complex, residues 1–7, 255–271, 345–351, 427–435, 539–553, 606–610, 693–700, 837–840, 955–968, 1042–1082, 1120–1129, 1146–1170, 1414–1421, 1564–1605 and 1673–1681 of Dicer-2, residues 1–94, 164–187 and 211–215 of R2D2, nucleotides g6–g21 and p1–p14 of the dsRNA, and nucleotides g21 and p21 of the siRNA are not included in the final model. The curves representing model versus full, half1 and half2 were calculated using Servalcat⁴⁷. In brief, the final models were ‘shaken’ by introducing random shifts to the atomic coordinates with a root mean squared deviation of 0.3 Å, and were refined against the first half map. The statistics of the 3D reconstruction and model refinement are summarized in Supplementary Table 1. The cryo-EM density maps were calculated with UCSF ChimeraX⁵¹, and the molecular graphics were prepared with CueMol (http://www.cuemol.org).

Pull-down experiments

The Dicer-2–R2D2 heterodimer (wild-type or mutant 8×His–Dicer-2 and 3×Flag–R2D2) was co-expressed in Sf9 cells (ATCC, catalogue no. CRL-1711), using the Bac-to-Bac system. The cells were collected by centrifugation, and then solubilized in lysis buffer (30 mM Hepes-KOH, pH 7.4, 100 mM potassium acetate, 2 mM magnesium acetate, 0.5% NP-40, and 5% glycerol). The lysate was centrifuged at 10,000g for 15 min, and then half of the supernatant was incubated with anti-Flag M2 affinity magnetic beads for 1 h. The beads were washed with wash buffer (30 mM Hepes-KOH, pH 7.4, 800 mM NaCl, 2 mM magnesium acetate, 1% Triton-X 100, and 5% glycerol), and then with lysis buffer without NP-40. The beads were treated with SDS–PAGE sample buffer. The lysate and the bound proteins were analysed by 5–20% SDS–PAGE. Western blotting was performed at 20 V (constant voltage) for 30 min, using an Immobilon-P Transfer Membrane (Merck), an EzFastBlot HMW (ATTO), and a Trans-Blot Turbo Transfer System (Bio-Rad). After the transfer, the membranes were incubated on a tilting shaker with blocking buffer (TBST containing 1% skim milk) for 10 min, primary antibodies: anti-Flag antibody (1:2,000) and anti-Dicer-2 antibody (1:1,000) for 60 min, and horseradish peroxidase-conjugated secondary antibodies: goat anti-rabbit IgG (H+L) antibody (1:4,000) and goat anti-mouse IgG (H+L) antibody (1:4,000) for 30 min. The chemiluminescence was then detected using Luminata Forte Western HRP substrate (Merck) and an Amersham Imager 600 (GE Healthcare). R2D2 and Dicer-2 were detected with an anti-Flag-antibody (Sigma) and an anti-Dicer-2 antibody (Abcam), respectively.

Photocrosslinking experiments

For photocrosslinking experiments, siRNA-1 and siRNA-4 were 5′-radiolabelled using T4 polynucleotide kinase, and then annealed with a 1.5-fold excess of siRNA-2 or siRNA-3 and siRNA-2 (1-nt overhang) or siRNA-5 (0-nt overhang), respectively (Supplementary Table 2). The 5-iodouracil-containing siRNA duplex (20 nM) and Dicer-2–R2D2 (40 nM) were incubated for 30 min. The samples (7 μl per well) were transferred to a Terasaki plate (Greiner BIO-ONE), and then exposed to > 300 nm UV radiation for 60 s, using a UV crosslinker (SP-11 Spot Cure, Ushio) with a uniform radiation lens (USHIO) and a long-path filter (300 nm, ASAHI SPECTRA), at 15 cm from the light. Aliquots of the reaction mixture were transferred into a new tube and mixed with SDS–PAGE sample buffer. The samples were analysed by 5–20% SDS–PAGE, and crosslinked proteins were detected by phosphorimaging.

Reporting summary

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