Cells and plasmids
Human embryonic kidney 293T cells (HEK-293T) were maintained in Dulbecco’s modified Eagle medium (DMEM) with 10% fetal bovine serum (FBS; Sigma-Aldrich) at 37°C and 5% CO2. BSR-T7/5 cells, cells stably expressing the T7 polymerase that are derived from baby hamster kidney cells (BHK-21) [43] were maintained in DMEM with 10% FBS and 1 mg/ml G418/Geneticin (Roche) every third cell culture passage. The Spodoptera frugiperda Sf9 cells were maintained in Sf-900™ II SFM medium (Thermo Scientific) at 27°C.
pCAGGs plasmids expressing NiV L, P and N proteins (Bangladesh, 2004) as well as the pUC57 plasmid expressing T7-dependent NiV bicistronic minigenome (BMG) [10] were kindly provided by Micheal Lo (CDC). Plasmids expressing truncated versions of L proteins were generated by Gibson assembly (NEB).
Protein expression and purification of NiV L-P complex
The coding sequences for Malaysian strain NiV L (NCBI: NP_112028.1) with an N terminal twin-Strep tag and P (NP_112022.1) with the C-terminal octa-His tag were codon optimized for insect cell expression and cloned into a pFL plasmid downstream of polyhedrin and p10 promoters, respectively. The preparation of Baculovirus stocks and protein expression were performed following the Bac to Bac manual (Invitrogen). Two litres of Sf9 cells expressing the L-P complex were harvested 72 hours post infection. The cells were resuspended in Buffer A (50 mM HEPES pH 7.4, 500 mM NaCl, 10% vol/vol glycerol, 2 mM tris(2-carboxyethyl) phosphine (TCEP), 5 mM MgCl2) which was further supplemented with 0.05% wt/vol n-octyl beta-d-thioglucopyranoside, one protease inhibitor cocktail tablet (Roche, cOmplete Mini, EDTA-free), 2 mM phenylmethylsulfonyl fluoride (PMSF), Bensonase, and RNase. Cells were lysed using a Dounce homogeniser and clarified with centrifugation. The lysate was incubated with the pre-equilibrated Strep-Tactin® XT Sepharose resin (IBA Lifesciences) for 3 h and resin was washed with the BufferA. The NiV L-P complex was eluted with 50 mM Biotin in Buffer A and further purified with Superose 6 increase 10/300 size exclusion column (GE Healthcare) in Buffer A. The fractions eluting after the void volume (between 10-12ml) was collected and concentrated to 0.5 mg/ml. The L-D832A mutant, in complex with P, was expressed and purified using the same methods as the wild-type L-P complex.
Protein expression and purification of CD of NiV L
Residues 1480–1742, encoding the CD of L, were cloned into a pET28a vector with an N-terminal hexa-His and Sumo tag. The construct was transformed into BL21(DE3) cells and protein expression was induced with 0.5 mM IPTG at OD600 = 0.6 and expression was carried out at 18°C for 18 hours. Six litres of bacterial cell pellet were lysed by sonication in a Lysis Buffer containing 50 mM HEPES pH 7.6, 500 mM NaCl, 10% vol/vol glycerol supplemented with protease inhibitors (Roche, cOmplete Mini, EDTA-free), RNase A, and lysozyme (Sigma). Supernatant cleared after ultracentrifugation was filtered and loaded onto HisTrap 5 ml HP columns. Protein was washed and eluted with a gradient increase in imidazole concentration. Elution fractions were combined, desalted into the Final Buffer containing 20 mM HEPES pH 7.6, 500 mM NaCl, 10% vol/vol glycerol and then incubated with Ulp1 overnight at 4°C. Cleaved protein was re-injected into HisTrap 5 ml HP to remove Ulp1 enzyme as well as the cleaved His-Sumo tag. The flowthrough was collected, concentrated, and injected into Superdex 75 16/600 (GE Healthcare) that was pre-equilibrated with the Final Buffer. The protein was concentrated to 10 mg/ml and stored at -80°C.
In vitro RNA synthesis assay
For the NiV L-P complex in vitro assay, 3 µl reactions were set up containing reaction buffer (20 mM Tris pH 7.5, 10 mM KCl, 2 mM DTT, 0.5% triton, 10% DMSO, 1 U Rnasin (Promega), 5 mM MgCl2), 0.25 µM RNA-template derived from the NiV leader sequence (UGGUUUGUUCCC or UGGUCUGUUCCC), 0.5 µM recombinant L-P complex, 0.5 mM ATP, 0.5 mM CTP, 0.1 µM GTP and 200 µM primer (pACCA). ATP was substituted with remdesivir triphosphate (APExBIO), where indicated. The radioisotope tracer in these reactions was [α32P] GTP (Revvity). The reactions were incubated at 30 ̊C for 1 h, stopped with the addition of 3 µl formamide loading buffer and denatured at 95 ̊C for 3 minutes. A 32P-5’end labelled 20 nucleotide-long DNA served as a marker. RNA products were separated on a 22% polyacrylamide urea gel for 2.5 hours at 35 W and the level of [α32P] GMP incorporation was imaged by phosphorimaging on a FLA-5000 scanner (Fuji).
Minireplicon assay
BSR-T7/5 cells were seeded in 24-well plates at 5 x 104 cells/well. 24 hours later, cells were transfected in duplicates with the BMG plasmid (0.4 µg), as well as pCAGGs plasmids expressing for NiV L (0.4 µg), P (0.4 µg), N (0.25 µg) using LT-1 transfection reagent (MirusBio). For negative controls pCAGGs-NiV-L was substituted with a pCAGGs-empty vector. After 48 hours, Gaussia luciferase activity was determined using the Renilla Luciferase Assay System (Promega). For this purpose, cells were lysed in 50 µl Renilla lysis buffer on a microplate shaker at room temperature. After 45 min, 20 µl of the cell lysates were mixed with 50 µl of the Renilla luciferase reagent and RLUs were analysed using a GLOMAX 20/20 luminometer (Promega).
Cryo-EM sample preparation, data acquisition and processing
3.5 µl of the NiV L-P complex at 0.5 mg/ml was applied to a freshly glow-discharged UltrAufoil Au 1.2/1.3 300 mesh grid. The sample was blotted for 6 s and plunge frozen in liquid ethane. All grids were prepared using a Vitrobot Mark VI (FEI) under conditions of 100% humidity and 20°C. Cryo-EM data was collected at the Oxford Particle Imaging Centre (OPIC) using a 300 kV G3i Titan Krios microscope (Thermo Fisher Scientific) equipped with a SelectrisX energy filter and Falcon 4i direct electron detector. Automated data collection was setup in EPU 3.4 and a total of 22,859 movies were recorded in EER format, of which 14,319 were collected with a tilt angle of 30°. Data was collected using AFIS with a total dose of ~ 50 e-/Å2, a calibrated pixel size of 0.7303 Å/pix, defocus range of -1.4 to -2.4, and with 10 eV slit. Cryo-EM data collection parameters and refinement statistics are summarized in Table S1.
Data processing was performed in CryoSparc v4.4.1 [44] by following the workflow outlined in Supplementary Fig. 2. Briefly, motion correction and patch CTF estimations were performed for movie frames, and low-quality images were eliminated by manual inspection and excluded from further analysis. A template search was prepared using the PIV5 L-P structure (PDB 6V85) as the template. Following template picking, 2D classes were obtained, and the best-resolved classes were selected for ab-initio model generation. The ab-initio models were refined using Heterogenous refinement. The resulting optimal map was refined using NU-refinement with C1 symmetry. This refined map was then used to train Topaz picking. The picked particles were directly subjected to heterogeneous refinement, and the best resulting map was selected for further refinement using NU-refinement. This entire process, from particle picking to Topaz training, was repeated three times to increase the number of picked particles.
The map, containing 490,675 particles, was corrected for reference motion and subjected to 3D classification to identify classes that represented clear densities for the L and P-XD domain. A total of 299,261 particles were selected and refined for the L domain using local NU refinements, resulting in a 2.52 Å resolution map. Meanwhile, 206,718 particles were refined for the P tetramer using local NU refinement, yielding a 2.75 Å resolution map. The local maps were further processed with DeepEMhancer.
Cryo EM model building and refinement
Initial models were generated for L and P tetramer separately using ModelAngelo [45]. To improve model geometry, multiple cycles of manual building were performed using COOT [46] followed by real space refinement against the corresponding local maps in PHENIX [47]. The model geometry was validated using MolProbity [48]. Comprehensive statistics for the map and model are presented in the Supplementary Table 1. Structural analysis and figure preparation were conducted using UCSF Chimera [49] and UCSF ChimeraX [50].
Crystallization of CD, X-ray data collection and structure solving
Freshly purified CD, at a concentration of 10 mg/mL, was mixed with a set of sitting-drop crystallization screens in ratios of 1:1, 2:1, and 1:2, and incubated at 20°C. The crystals were obtained in 25 days from a condition consisting of 35% tert-Butanol and 0.1M tri-Sodium Citrate pH 5.6. The crystals were harvested, cryo-preserved in the crystallization condition containing 20% v/v glycerol and flash frozen in liquid nitrogen.
X-ray diffraction data were collected on the I04 beamline at Diamond Light Source (Harwell, U.K.). The data processing was performed using the program DIALS in combination with XIA2 [51]. The structure was solved via PHASER molecular replacement [52], employing a search model generated by AlphaFold [53]. Subsequent manual model building was performed in COOT [46], followed by refinement in PHENIX [47]. X-ray data collection and refinement statistics are summarized in Table S2.
AlphaFold modelling
The RNA-bound NiV L structure was predicted in AlphaFold 3 by providing the leader and product RNA sequences as the input along with the corresponding protein sequences [54]. For the X-ray crystallography analysis, a reference model for molecular replacement of the CD was prepared using AlphaFold 2 [53].