Structural and functional characterization of the SLA′ structure at the 3′ terminus of the Zika virus negative-strand intermediate

Maintenance of cell lines

Baby hamster kidney-21 (BHK-21) cells were maintained in vented, tissue-culture treated flasks (Sarstedt) in complete media (Dulbecco's minimal essential medium [DMEM] supplemented with 10% fetal bovine serum [Wisent], 1× MEM nonessential amino acids [Corning], and 2 mM l-glutamine [Gibco]). Cells were cultured at 37°C and 5% CO₂ with regularly screening for mycoplasma contamination.

Cloning

To generate the ZIKV NS5-His expression plasmid for production of full-length ZIKV NS5, the pET22b plasmid was first modified by reverse PCR with Q5 Hot Start High-Fidelity 2X Master Mix (New England BioLabs) using the pET22b GSSS RPCR Fwd and pET22b GSSS RPCR Rev primers (Supplemental Table S1) to add a GSSS linker sequence to the N terminus of the encoded 6X His-tag. Next, the NS5 coding sequence from the ZIKV PRVABC59 strain (GenBank accession: KU501215.1) was codon optimized for BL21 Escherichia coli using the IDT Codon Optimization Tool to generate a geneblock (IDT, Supplemental Table S2). The codon-optimized NS5 sequence was inserted between the start codon and the GSSS linker of the modified pET22b expression plasmid via Gibson assembly using the ZIKV CO NS5 and pET22b CO RPCR primer sets (Supplemental Table S1). The resulting plasmid, pET22b ZIKV NS5-His, encodes the codon-optimized ZIKV full-length NS5 fused to a 6X His-tag via a GSSS linker. The plasmid was prepared using the QIAprep Spin Miniprep kit (QIAGEN), according to the manufacturer's instructions and verified via Sanger sequencing (Sequencing + Bioinformatics Consortium) and Oxford Nanopore sequencing (Plasmidsaurus).

To produce the IVT template plasmids, geneblocks (IDT) were generated for the 5′ end of the positive strand for DENV1 (GenBank accession: NC_001477.1), DENV2 (GenBank accession: NC_001474.2), DENV3 (GenBank accession: NC_001475.2), and DENV4 (GenBank accession: NC_002640.1), and for the 3′ end of the negative strand of ZIKV and DENV2 (Supplemental Table S2). The geneblocks for the 5′ end of the positive strand of DENV1–4 consisted of an in vitro transcription cassette in which the T7 promoter controls transcription of the first 359 nt of the respective reference genome followed by nt 360–383 of the ZIKV PRVABC59 genome (complementary to the ZIKV [+] 5′-UTR PE oligo) and the hepatitis delta virus ribozyme (HDVr). This transcription cassette was flanked by EcoRI and BamHI cut sites to facilitate cloning into the pUC18 backbone. For the 3′ end of the negative strand of ZIKV and DENV2, the region between the T7 promoter and the HDVr in the transcription cassette was replaced with the respective terminal 163 nt of the 3′ end of the negative strand. In all the above cases, the geneblock was amplified with EcoRI-T7 and BamHI-HDVr primers and then inserted into pUC18 using EcoRI and BamHI (Supplemental Table S1). The IVT template plasmid for the ZIKV 5′ end of the positive strand was generated by replacing the region in the ZIKV 3′ negative-strand IVT template plasmid between the T7 promoter and the HDVr ribozyme with the first 383 nt of the ZIKV PRVABC59 genome. This replacement was performed via Gibson assembly using the ZIKV 5′ (+) and T7-HDVr RPCR primer sets (Supplemental Table S1). A similar replacement was performed to generate the 3′ end of the negative-strand IVT template plasmid for DENV1, DENV3, and DENV4, where the respective terminal 163 nt of the 3′ end of the negative strand was inserted between T7 and HDVr via Gibson assembly using the T7-HDVr RPCR primer set and either the DENV1 3(−) Fwd and DENV1 3(−) Rev, DENV3 3(−) Fwd and DENV1 3(−) Rev, or DENV4 3(−) Fwd and DENV4 3(−) Rev primers, respectively (Supplemental Table S1). Plasmids were purified with the QIAprep Spin Miniprep kit (QIAGEN), according to the manufacturer's instructions and verified via Sanger sequencing (Sequencing + Bioinformatics Consortium).

Mutagenesis of SLA to disrupt the SLA′ structure via the introduction of G·U wobble base pairs was performed using the ZIKV PRVABC59 subgenomic Renilla luciferase (RLuc) reporter replicon system (Barnard et al. 2023). In brief, the replicon contains the ZIKV 5′ UTR, the first 38 amino acids of the capsid coding region fused to RLuc, the FMDV2A peptide, and the last 30 amino acids of the Envelope coding region through to the 3′ terminus of the ZIKV genome. To introduce the desired mutations into SLA, the region from PacI to ApaLI in the subgenomic replicon was subcloned into pUC18 via restriction digest and ligation with T4 DNA ligase (New England BioLabs). Mutations that introduce G·U wobble base pairs into the base stem, lower, and upper portions of the TSL, were added via reverse PCR with the SLA GU Base RPCR, SLA GU TSL Base RPCR, or SLA GU TSL Top RPCR primer sets, respectively (Supplemental Table S1). The subcloned region was amplified with the 5′ Rep SC GA primer set and reassembled with the PacI to ApaLI replicon via Gibson assembly (Table 1). Resulting replicon-containing bacterial artificial chromosomes were purified using the QIAGEN Large-Construct kit (QIAGEN), and sequences were verified by Oxford Nanopore sequencing (Plasmidsaurus).

BL21-pRARE competent cell preparation

BL21(DE3)-R3-pRARE2 was a gift from Opher Gileadi (Addgene plasmid # 26242) (Savitsky et al. 2010). These bacteria were used to inoculate a 5 mL culture of LB with 10 µg/mL chloramphenicol and incubated overnight at 37°C at 200 rpm. Next, 1 mL of the overnight culture was used to inoculate 99 mL of LB media with 10 µg/mL chloramphenicol and incubated at 37°C at 200 rpm until the culture reached an OD₆₀₀ of 0.3–0.5. Bacteria were pelleted at 4500 rpm for 15 min at 4°C and then gently resuspended in 20 mL of ice-cold 1 m CaCl₂. The resuspended cells were allowed to chill on ice for 15 min prior to centrifugation at 4500 rpm for 15 min at 4°C. Pelleted bacteria were then gently resuspended in 5 mL of ice-cold 0.1 m CaCl_2, supplemented with 15% glycerol, and filter sterilized with a 0.22 µm filter (Millex). Aliquots of the competent cells were transferred to sterile, ice-cold microcentrifuge tubes and stored at −70°C for up to 1 year.

ZIKV NS5 production and purification

Competent BL21(DE3)-R3-pRARE2 cells were transformed with pET22b ZIKV NS5-His via a 30 sec heat shock at 42°C followed by 5 min on ice and a 45 min recovery period at 37°C and 300 rpm. Cells were plated on a LB plate containing 100 µg/mL ampicillin and 10 µg/mL chloramphenicol and incubated overnight at 37°C at 200 rpm. Multiple colonies were then used to inoculate 100 mL of LB medium containing 100 µg/mL ampicillin and 10 µg/mL chloramphenicol, and incubated overnight at 37°C at 200 rpm. Next, 10 mL of the starter culture was used to inoculate 1 L of LB medium containing 100 µg/mL ampicillin and 10 µg/mL chloramphenicol, which was incubated using the above conditions until the culture reached an OD₆₀₀ of 0.6–0.8. At this point, the culture was cold shocked twice in an ice bath for 2 min with swirling before addition of isopropyl β-d-1-galactopyranoside to a final concentration of 0.5 mM. Induced cultures were incubated at 16°C at 230 rpm for 18 h and then pelleted at 4500 rpm for 15 min at 4°C. Pelleted cells were washed once with 20 mL of ice-cold phosphate buffered saline (PBS) and frozen at −70°C.

Frozen pellets were thawed on ice and resuspended in lysis buffer (20 mM Tris-HCl pH 7.5, 500 mM NaCl, 1.4 mM β-mercaptoethanol, 10% glycerol, 30 mM imidazole) with protease inhibitors (1 mM PMSF, 1 mM benzamidine, 0.7 µg/mL pepstatin A, 5 µg/mL leupeptin). Resuspended pellets were lysed via 4 × 1 min sonication pulses at 18% amplitude, and the resulting lysate was centrifuged at 18,000 rpm for 50 min at 4°C. Supernatants were then loaded onto a 5 mL HisTrap FF column (Cytiva) for initial purification via nickel affinity chromatography. The column bound ZIKV NS5-His was washed with 10 column volumes of lysis buffer and eluted with nickel affinity elution buffer (20 mM Tris-HCl pH 7.5, 500 mM NaCl, 1.4 mM β-mercaptoethanol, 10% glycerol, 300 mM imidazole). Fractions containing ZIKV NS5-His were then pooled, diluted in cation affinity loading buffer (20 mM Tris-HCl pH 7.5, 1 mM DTT, 10% glycerol) to reduce the salt concentration to 100 mM, and filtered through a 0.45 µM filter (Millex) to remove any resulting precipitate. The desalted filtrate was then loaded onto a 5 mL SP HiTrap column (Cytiva) for cation affinity chromatography and washed with 10 column volumes of cation affinity washing buffer (20 mM Tris-HCl pH 7.5, 100 mM NaCl, 1 mM DTT, 10% glycerol). ZIKV NS5-His was eluted with a linear gradient of NaCl to a final concentration of 500 mM NaCl. Fractions containing ZIKV NS5-His were then pooled and concentrated before a final size exclusion chromatography purification on a Superdex 200 Increase 10/300 column (Cytiva), where buffer was exchanged for S200 running buffer (50 mM Tris-HCl pH 7.5, 500 mM NaCl, 5 mM DTT, 10% glycerol) under RNase-free conditions. The resulting fractions containing the purified ZIKV NS5-His were pooled and concentrated to ∼2 mg/mL using 50 kDa MWCO centrifugal filters (Amicon), and then flash frozen and stored at −70°C.

RNA secondary and tertiary structure prediction

Prediction of RNA secondary structure and Gibbs free energy were performed using the RNAstructure software package (available at https://rna.urmc.rochester.edu/index.html) (Reuter and Mathews 2010). In general, RNA sequences of interest were loaded into the RNAstructure software using the “Fold RNA Single Strand” command. To generate RNA structure predictions using in vitro SHAPE data as a constraint, corresponding RNA sequences were loaded as a “Single Sequence” using the “Structure Prediction” command. Normalized SHAPE data were then input as a constraint on secondary structure prediction as a .shape file. Outputs were saved as dot bracket files and visualized via VARNA (Darty et al. 2009).

Prediction of RNA tertiary structure was performed using AlphaFold Server (available at https://alphafoldserver.com/) (Abramson et al. 2024). The first 70 nucleotides (nt) of the positive strand and the last 70 nt of the negative strand were used as templates for predictions of SLA and SLA′, respectively. The five predicted structures with the highest AlphaFold ranking score were aligned and visualized in PyMol (Schrodinger and DeLano 2020, http://www.pymol.org/pymol.).

In vitro transcription of genome termini RNA

For in vitro SHAPE and EMSA experiments, DNA templates were generated by PCR and in vitro transcribed into RNA with the T7 RiboMAX Kit (Promega), according to the manufacturer's instructions. DNA templates corresponding to the first 383 nt of the ZIKV and DENV1–4 genomes were generated from their respective IVT template plasmids using the EcoRI-T7 and BamHI-HDVr primers (Supplemental Table S1). To generate DNA templates corresponding to the first 163 nt of the ZIKV PRVABC59 genome, EcoRI-T7 and ZIKV + IVT 163 nt Rev (5′-CGTTTTAGCATATTGACAATCCGGAATC-3′) primers were used to amplify the region from the IVT template plasmids (Supplemental Table S1). Finally, amplification of DNA templates arising from the 3′ end of the negative strand of either ZIKV or DENV was performed using the EcoRI-T7 and BamHI-HDVr primers (Supplemental Table S1). All DNA templates were size-verified on a 2% agarose gel and purified using the QIAQuick PCR Purification Kit (QIAGEN), according to the manufacturer's instructions. In vitro transcribed RNA generated using the T7 RiboMAX Kit was treated with DNase I (NEB) for 20 min at 37°C, and then ethanol precipitated and purified. RNA was then gel purified on an 8% Urea-PAGE gel cast using the SequaGel UreaGel System (National Diagnostics) and run at 30 W for 2.5 h. Excised bands were incubated in 400 μL elution buffer (500 mM NH4OAc, 10 mM EDTA, and 0.1% sodium dodecylsulfate) overnight at 4°C with gentle agitation to extract the RNA. Gel-purified RNA was ethanol precipitated and stored at −20°C until use.

Selective 2′ hydroxyl acylation analyzed by prime extension (SHAPE)

In vitro SHAPE analysis was performed in quadruplicate as previously described (Chahal et al. 2019). In brief, 5 pmol of in vitro transcribed RNA was refolded and incubated in SHAPE buffer (333 mM HEPES pH 8.0, 20 mM MgCl₂, 333 mM NaCl) at 37°C for 10 min. RNA was then treated with 0.01 m NAI-N₃ or dimethyl sulfoxide (DMSO; treatment control) at 37°C for 5 min or 15 min for 383 and 163 nt IVT RNA, respectively. Treated RNA was then extracted using TRIzol reagent (Thermo Fisher), according to the manufacturer's instructions. Primer extensions (PEs) were performed in a 20 µL reaction volume containing 5 pmol of NAI-N₃-labeled ZIKV RNA and 1 pmol of the 6-FAM-labeled PE oligo. ZIKV (+) 5′-UTR 6-FAM (5′-CAG CAT GGC AGC CAG ATC TTT CTT-3′) was used as the PE oligo for the 383 nt IVT RNAs corresponding to the 5′ ends of both the ZIKV and DENV positive strands, while the ZIKV (−) 3′ PE 6FAM (5′-AGT TGT TGA TCT GTG-3′) and DENV (−) 3′ PE 6FAM (5′-AGT TGT TAG TCT ACG-3′) oligos were used for PE of the 163 nt IVT RNA corresponding to the 3′ end of the negative strands of ZIKV and DENV, respectively. Matched sequencing ladders were generated for all reactions by performing primer extension reactions using 5 pmol of the respective untreated IVT RNA, 1 pmol of the respective NED-labeled PE oligo (ZIKV [+] 5′-UTR PE NED, ZIKV [−] 3′ PE NED, or DENV [−] 3′ PE NED), and either 0.5 mM ddCTP or ddGTP. Primer extensions were performed for 30 min at 52°C for the ZIKV (+) 5′-UTR PE reactions, and 45 min at 52°C for the ZIKV (−) 3′ PE and DENV (−) 3′ PE reactions. Following extension, 2 m NaOH was added and incubated at 95°C for 5 min to degrade the input RNA. The remaining cDNA was then diluted to a total volume of 100 µL with water and ethanol precipitated with 2.5 volumes 95% ethanol and 0.1 volumes 3 m sodium acetate pH 5.2 for storage at −70°C. Pelleted cDNA was washed three times with cold 85% ethanol and then dried at 37°C for 10 min. Dried pellets were sent for capillary electrophoresis on an ABI 3100 Genetic Analyzer at Plate-forme d'Analyses Génomiques de l'Université Laval. Raw fluorescence data were analyzed using the QuShape software package to generate normalized SHAPE reactivities for each nucleotide (Karabiber et al. 2013).

SHAPE analysis of the ZIKV IVT RNAs in the presence of recombinant NS5 protein was performed as above with the following exceptions for the NAI-N₃-labeling step. Following refolding, the 5 pmol of RNA was added to a binding reaction with a total reaction volume of 20 µL containing either 1 ng of recombinant ZIKV NS5 (2 n_NS5:1 n_RNA) in NS5 binding buffer (20 mM Tris pH 7.4, 150 mM NaCl, 2 mM MgCl₂, 10% glycerol) or NS5 binding buffer alone (0 n_NS5:1 n_RNA). Binding reactions proceeded at 25°C for 30 min, before treatment with 0.01 m NAI-N₃ or DMSO (treatment control) at 25°C. The Wilcoxon rank test was done between the SHAPE data of RNA with NS5 and that of RNA alone to determine if there was a statistical difference between the SHAPE reactivities. The Wilcoxon test determined there are highly significant differences in SHAPE reactivity between groups for both SLA and SLA′ (P < 0.0001 for SLA; P < 0.05 for SLA). The baseline value for ΔSHAPE (NS5:RNA reactivity−RNA reactivity) was determined by finding the average of the absolute ΔSHAPE values. ΔSHAPE values above the positive baseline were considered significant increases and below the negative baseline value were considered significant decreases in SHAPE reactivity.

Electrophoretic mobility shift assay (EMSA)

First, 2.5 pmol (150 ng) of IVT RNA corresponding to the 163 nt at the 5′ terminus of the positive strand or the 3′ end of the negative strand was heated at 95°C and then snap-cooled on ice. Once cooled, RNA was added to a binding reaction with a total reaction volume of 20 µL containing one dilution of a series of ZIKV NS5-His ranging from a molar ratio (NS5:RNA) of 0–8:1 (0–2 µg of NS5) in NS5 binding buffer. Binding reactions proceeded at 25°C for 30 min before separation on a 1% agarose 0.5× TBE gel run for 1 h at 100 V on ice at 4°C. Gels were stained with 1× SYBR Gold nucleic acid gel stain (Thermo Fisher) on an orbital shaker at 120 rpm for 20 min and visualized on a ChemiDoc MP Imaging System (BioRad).

In vitro transcription of ZIKV replicon RNA

To generate full-length replicon RNAs, templates were linearized with PsrI (Creative Enzymes) and in vitro transcribed with SP6 RNA polymerase. In vitro transcription was performed in a total reaction volume of 60 µL containing 1 mM of each rNTP (New England BioLabs), 0.8 mM m7G(5′)ppp(5′)A Cap Structure Analog (New England BioLabs), 1 U/µL RiboLock RNase inhibitor (Thermo Fisher), 50 mM dithiothreitol (Thermo Fisher), and 1 U/µL SP6 RNA polymerase (New England BioLabs). Reactions proceeded at 37°C for 90 min, followed by a 20 min DNase I (New England BioLabs) treatment at 37°C. Capped firefly luciferase (FLuc) mRNAs were generated as previously described (Rheault et al. 2023). IVTs were then ethanol precipitated as described above and stored at −70°C until use.

Electroporations

BHK-21 cells were trypsinized, washed two times in 25 mL cold PBS, and then resuspended in cold PBS to a concentration of 1 × 10⁷ cells/mL. In an ice-cold 4 mm cuvette, 400 µL of the above cell suspension (4 × 10⁶ cells) was mixed with 10 µg of the respective ZIKV replicon RNA and 1 µg capped Fluc mRNA. Cells were then electroporated with one pulse of 270 V, 975 µF, and infinite resistance on a Bio-Rad Gene Pulser XCell (BioRad). Electroporated cells were then resuspended to a total volume of 8 mL in complete media. Two milliliters of cells were either placed in a 15 mL conical tube (2 h time point) or were plated in 6-well plates (24–72 h time points) and incubated at 37°C and 5% CO₂ until harvest.

Luciferase assay

To collect samples for luciferase assays, electroporated cells were washed with 1 mL PBS, harvested in 100 µL 1× passive lysis buffer (Promega), and stored at 70°C until use. Luciferase assay was performed using the Dual Luciferase Assay kit (Promega), according to the manufacturer's instructions, except that 50 µL of each reagent was used for 10 µL of harvested sample in opaque, white 96-well plates (Greiner Bio-One). Luminescence was measured on a GloMax Explorer plate reader (Promega) with a 10 sec integration time. Luciferase activity of each sample was analyzed in duplicate, and all data are representative of three independent biological replicates.

Data and statistical analyses

Statistical analyses were performed in GraphPad 10 software. For analysis of luciferase assays, multiple unpaired Student's t-tests were performed for each time point in comparison with WT.