RNA affinity pulldown mass spectrometry
RNA antisense purification was performed according to a protocol based on 17. Briefly, 6*10^7 HEK293 cells per condition were lysed in a buffer containing 20 mM Tris/HCl pH 7.5, 100 mM KCl, 5 mM MgCl2, 1 mM DTT, 0.5 % Igepal CA630 (Sigma-Aldrich), 1× cOmplete™ Protease Inhibitor Cocktail (Roche), 40 U/ml RNase inhibitor (Molox). The cleared lysate was incubated with in vitro transcribed RNA corresponding to the SARS-CoV-2 − 1PRF site, which was immobilized on streptavidin hydrophilic magnetic beads (NEB) by biotin-streptavidin interaction. After three washes with binding buffer (50 mM HEPES/KOH pH 7.5, 100 mM NaCl, 10 mM MgCl2) and two washes with wash buffer (50 mM HEPES/KOH pH 7.5, 250 mM NaCl, 10 mM MgCl2), bound proteins were eluted by boiling the sample in 1× NuPAGE LDS sample buffer (Thermo Fisher Scientific) supplemented with 40 mM DTT. For infected as well as uninfected Calu-3 cells the procedure was performed similarly. In order to inactivate the virus, the lysis buffer contained Triton-X100 and inactivation was confirmed by plaque assays.
For LC-MS/MS, the eluted proteins were alkylated using iodoacetamide followed by acetone precipitation. In solution digests were performed in 100 mM ammonium bicarbonate and 6 M urea using Lys-C and after reducing the urea concentration to 4 M with trypsin. Peptides were desalted using C18 stage tips and lyophilized. LC-MS/MS was performed at the RVZ Proteomics Facility (Würzburg) and analyzed as described previously 61. Gene ontology (GO) term analysis was performed with Panther 62. The list of all identified proteins is given in Supplementary Table 1.
Co-immunoprecipitation
Endogenous interaction partners of ZAP were identified by co-immunoprecipitation followed by mass spectrometry as published previously 63. Briefly, uninfected and SARS-CoV-2-infected Calu-3 cells were lysed in lysis buffer (10 mM Tris/HCl pH 7.4, 150 mM NaCl, 1% Igepal CA630 (Sigma-Aldrich) and 1x cOmplete™ Protease Inhibitor Cocktail (Roche)). The lysis buffer was supplemented either with 40 U/ml RNase inhibitor (Molox) or 50 U/ml Benzonase (Roche) to differentiate between RNA- and protein-dependent interactions. 1 mg of cell lysate was cleared with protein A magnetic beads to remove non-specific interactions (S1425S, NEB) and incubated overnight with anti-ZAP antibody (Proteintech 16820-1-AP) or anti-IgG from rabbit as a control (Cell Signaling, a gift from Dr. Mathias Munschauer, HIRI-HZI). Antibodies were captured with protein A magnetic beads, washed with lysis buffer, and eluted by boiling in 1X NuPAGE™ LDS Sample Buffer (Thermo Fisher Scientific) supplemented with 40 mM DTT. LC-MS/MS proceeded as described above. A list of all identified proteins can be found in Supplementary Table 2.
Plasmid construction
To generate dual-fluorescence reporter constructs frameshift sites of SARS-CoV, SARS-CoV-2, MERS-CoV, BtCoV 273, HIV-1, JEV, PEG10, WNV were placed between the coding sequence of EGFP and mCherry (parental construct was a gift from Andrea Musacchio (Addgene plasmid # 87803 64) by site-directed mutagenesis in a way that EGFP would be produced in 0-frame and mCherry in − 1-frame. EGFP and mCherry were separated by StopGo 65 signals as well as an alpha-helical linker 66. A construct with no PRF insert and mCherry in-frame with EGFP served as a 100% translation control and was used to normalize EGFP and mCherry intensities.
To generate screening vectors, protein-coding sequences of DDX17 (NM_001098504.2), DDX36 (NM_020865.3), ELAVL1 (NM_001419.3), GNL2 (NM_013285.3), HNRNPF (NM_001098204.2), HNRNPH1 (NM_001364255.2), HNRNPH2 (NM_001032393.2), IGF2BP1 (NM_006546.4), MATR3 iso 2 (NM_018834.6), MMTAG2 (NM_024319.4), NAF1 (NM_138386.3), NHP2 (NM_017838.4), POP1 (NM_001145860.2), RAP11B (NM_004218.4), RSL1D1 (NM_015659.3), SFL (NM_018381.4), SURF6 (NM_001278942.2), TFRC (NM_003234.4), ZAP (NM_024625.4) and ZNF346 (NM_012279.4) were placed in frame with the coding sequence for ECFP in pFlp-Bac-to-Mam (gift from Dr. Joop van den Heuvel, HZI, Braunschweig, Germany 67) via Gibson Assembly 68.
Golden Gate compatible vectors for heterologous overexpression in E. coli, in vitro translation in RRL, and lentivirus production, were generated by Golden Gate or Gibson Assembly. A dropout cassette was included to facilitate the screening of positive colonies. Protein-coding sequences were introduced by Golden Gate Assembly using AarI cut sites 69. pET-SUMO-GFP (gift from Prof. Utz Fischer, Julius-Maximilians-University, Würzburg, Germany) was used as the parental vectors for protein overexpression in E. coli. The lentivirus plasmid was a gift from Prof. Chase Beisel (HIRI-HZI, Würzburg, Germany). An ALFA-tag was included to facilitate the detection of the expressed protein 70. The frameshift reporter vector for the in vitro translation contained ß-globin 5' and 3' UTRs as well as a 30 nt long poly(A) tail. The insert was derived from nucleotides 12686–14190 of SARS-CoV-2 (NC_045512.2); a 3×FLAG-tag was introduced at the N-terminus to facilitate detection. To generate 0% and 100% − 1PRF controls, the − 1PRF site was mutated by disrupting the pseudoknot structure as well as the slippery sequence.
Optical tweezers constructs were based on the wild type SARS-CoV-2 frameshifting site (nucleotides 13475–13541) cloned into the plasmid pMZ_lambda_OT, which encodes for the optical tweezer handle sequences (2Kb each) flanking the RNA structure (130 nt). Constructs were generated using Gibson Assembly. Sequences of all plasmids and oligos used in this study are given in Supplementary Table 4.
Cell culture, transfections, generation of polyclonal stable cell lines
HEK293 cells (gift from Prof. Jörg Vogel, HIRI-HZI) and Huh7 cells (gift from Dr. Mathias Munschauer, HIRI-HZI), were maintained in DMEM (Gibco) supplemented with 10% FBS (Gibco) and 100 U/ml streptomycin and 100 mg/ml penicillin. Calu-3 cells (ATCC HTB-55) were cultured in MEM (Sigma) supplemented with 10% FBS. Cell lines were kept at 37°C with 5% CO2. Transfections were performed using PEI (Polysciences) according to manufacturer's instructions. For co-transfections, plasmids were mixed at a 1:1 ratio.
VSV-G envelope pseudo-typed lentivirus for the generation of stable cell lines was produced by co-transfection of each transfer plasmid with pCMVdR 8.91 71 and pCMV-VSV-G (gift from Prof. Weinberg, Addgene plasmid # 8454 72). 72 h post-transfection, the supernatant was cleared by centrifugation and filtration. The supernatant was used to transduce naïve Huh7 cells in the presence of 10 µg/ml polybrene (Merck Millipore). After 72 h, the cells were selected with 10 µg/ml blasticidin (Cayman Chemical) for 10 days to generate polyclonal cell lines.
SARS-CoV-2 infection
For infection with SARS-CoV-2, we used the strain hCoV-19/Croatia/ZG-297-20/2020, a kind gift of Prof. Alemka Markotic (University Hospital for Infectious Diseases, Zagreb, Croatia). The virus was raised for two passages on Caco-2 cells (HZI Braunschweig). Calu-3 cells (ATCC HTB-55) were infected with 2000 PFU/ml corresponding to an MOI of 0.03 at 24h post-infection, cells were collected and lysed for proteomic and ribosome-interaction experiments. To study the effect of ZAP-S on SARS-CoV-2 infection, Huh-7 cells were employed. One hour before infection, Huh-7 cells both naïve or ZAP-S-overexpressing cells were either pre-stimulated with IFN-β (500 U/ml), IFN-ɣ (500 U/ml), IFN- ƛ1 (5 ng/ml), or left untreated. Cells were infected with 200000 pfu/ml, corresponding to an MOI of 0.03 at 24h post-infection, cell culture supernatants were collected and titrated by plaque assay on Vero E6 cells (ATCC CRL-1586). Briefly, confluent Vero E6 cells in 96-well plates were inoculated with dilutions of the virus-containing supernatants for one hour at 37°C, the inoculum was removed and cells were overlaid with MEM containing 1.75% methyl-cellulose. At three days post-infection, whole wells of the plates were imaged using an IncuCyte S3 (Sartorius) at 4x magnification, and plaques were counted visually.
Flow cytometry
HEK293 cells were transiently transfected with either the control construct or the − 1PRF construct encoding for the dual-fluorescence EGFP-mCherry translation reporter as outlined in Fig. 2A. Cells were harvested at 24 h post-transfection and fixed with 0.4% formaldehyde in PBS. After washing with PBS, flow cytometry was performed on a FACSAria III (BD Biosciences) or a NovoCyte Quanteon (ACEA) instrument. Flow cytometry data were analyzed with FlowJo software (BD Biosciences). ECFP-positive cells were analyzed for the ratio between mCherry and EGFP. FE was calculated according to the following formula:
where mCherry represents the mean mCherry intensity, EGFP the mean EGFP intensity, test represent the tested sample and control represents the in-frame control where mCherry and EGFP are produced in an equimolar ratio 73. Data represent the results of at least three independent experiments.
Purification of recombinant proteins
Recombinant ZAP-S N-terminally tagged with 6×His-SUMO was purified from E. coli Rosetta 2 cells (Merck) by induction with 0.2 mM isopropyl β-d-1-thiogalactopyranoside for 18 h at 18°C. Cells were collected, resuspended in lysis buffer (50 mM HEPES/KOH pH 7.6, 1 M NaCl, 1 mM DTT, 1 mM PMSF) and lysed in a pressure cell. The lysate was cleared by centrifugation and ZAP-S was captured using Ni-NTA resin (Macherey-Nagel). After elution with 500 mM imidazole, ZAP-S was further purified and the bound nucleic acids removed by size exclusion chromatography (HiLoad® 16/600 Superdex® 200) in 20 mM HEPES/KOH pH 7.6, 1 M KCl, 1 mM DTT, 20% glycerol. Protein identity was verified by SDS-PAGE as well as western blotting (Supplementary Fig. 2D). Purified ZAP-S was rapidly frozen and stored in aliquots at -80°C.
His-SUMO-IGF2BP3 as well as His-SUMO were kind gifts from Dr. Andreas Schlundt (Goethe University, Frankfurt, Germany)
Western blots
Protein samples were denatured at 95°C and resolved by 12 % SDS-PAGE at 30 mA for 2 h. After transfer using Trans-Blot (Bio-Rad), nitrocellulose membranes were developed using the following primary antibodies: anti-His-tag (ab18184), anti-DDDDK (ab49763), anti-ALFA (FluoTag®-X2 anti-ALFA AlexaFluor 647), anti-ZC3HAV1 (Proteintech 16820-1-AP). The following secondary antibodies were used: IRDye® 800CW Goat anti-rabbit and IRDye® 680RD Donkey anti-Mouse (both LI-COR). Bands were visualized using an Odyssey Clx infrared imager system (LI-COR) or a Typhoon7000 (GE Healthcare).
In vitro translation assays
mRNAs were in vitro transcribed using T7 polymerase purified in-house using linearized plasmid DNA as the template. These mRNAs were capped (Vaccinia Capping System, NEB) and translated using the nuclease-treated rabbit reticulocyte lysate (RRL; Promega). Typical reactions were comprised of 75% v/v RRL, 20 µM amino acids, and were programmed with ∼50 µg/ml template mRNA. ZAP-S was titrated in the range of 0–3 µM. Reactions were incubated for 1 h at 30°C. Samples were mixed with 3X volumes of 1X NuPAGE™ LDS Sample Buffer (Invitrogen), boiled for 3 min, and resolved on a NuPAGE™ 4 to 12% Bis-Tris polyacrylamide gel (Invitrogen). The products were detected using western blot (method as described above). The nitrocellulose membranes were developed using anti-DDDDK primary (Abcam ab49763) and IRDye® 680RD donkey anti-mouse secondary antibody (LI-COR). Bands were visualized using an Odyssey Clx infrared imager system (LI-COR). Bands corresponding to the − 1 or 0-frame products, 58 kDa and 33 kDa respectively, on western blots of in vitro translations were quantified densitometrically using ImageJ software 74. FE was calculated as previously described, by the formula intensity (–1-frame)/ (intensity (–1-frame) + intensity (0-frame)) 11. The change in FE was calculated as a ratio of FE of each condition to the FE of no-protein control in each measurement. Experiments were repeated at least 3 independent times.
Microscale thermophoresis
Short frameshifting RNA constructs were in vitro transcribed using T7 polymerase as described above. RNAs were labeled at the 3' end using pCp-Cy5 (Cytidine-5'-phosphate-3'-(6-aminohexyl) phosphate) (Jena Biosciences). For each binding experiment, RNA was diluted to 10 nM in Buffer A (50 mM Tris-HCl pH 7.6, 250 mM KCl, 5 mM MgCl2, 1 mM DTT, 5% glycerol supplemented with 0.05% Tween 20 and 0.2 mg/ml E. coli tRNA). A series of 16 tubes with ZAP-S dilutions were prepared in Buffer A on ice, producing ZAP-S ligand concentrations ranging from 40 pM to 2 µM. For measurements, each ligand dilution was mixed with one volume of labeled RNA, which led to a final concentration of 5.0 nM labeled RNA. The reaction was mixed by pipetting, incubated for 10 min at room temperature, followed by centrifugation at 10,000 × g for 5 min. Capillary forces were used to load the samples into Monolith NT.115 Premium Capillaries (NanoTemper Technologies). Measurements were performed using a Monolith Pico instrument (NanoTemper Technologies) at an ambient temperature of 25°C. Instrument parameters were adjusted to 5% LED power, medium MST power, and MST on-time of 5 seconds. An initial fluorescence scan was performed across the capillaries to determine the sample quality and afterward, 16 subsequent thermophoresis measurements were performed. Data of three independently pipetted measurements were analyzed for the ΔFnorm values, and binding affinities were determined by the MO. Affinity Analysis software (NanoTemper Technologies). Graphs were plotted using GraphPad Prism 8.4.3 software.
Microscopy
HEK293 cells were cultured on glass slides and transfected as described above. The cells were fixed with 4% paraformaldehyde in 1x PBS for 15 min at room temperature. After washing with 1x PBS, cells were mounted in ProLong Antifade Diamond without DAPI (Invitrogen). Microscopy was performed using a Thunder Imaging System (Leica) using 40% LED power and the 40x objective. EGFP was excited at 460–500 nm and detected at 512–542 nm. mCherry was excited at 540–580 nm and detected at 592–668 nm. The images were processed with the LasX software (Leica).
Polysome profiling analysis
A plasmid expressing ZAP-S N-terminally tagged with a His-tag was transfected into HEK293 cells using PEI, as described above. To check endogenous ZAP-S expression, HEK cells were transfected with a plasmid containing the same backbone and His-tag. At 24 h post-transfection, cycloheximide (VWR) was added to the medium at a final concentration of 100 µg/ml to stop translation. Approximately 107 HEK cells were lysed with 500 µl lysis buffer (20 mM Tris-HCl pH 7.4, 150 mM NaCl, 5 mM MgCl2, 1 mM DTT, 100 µg/ml Cycloheximide, 1% Triton X), and the lysate was clarified by centrifugation at 17,0000 × g for 10 min at 4°C. Polysome buffer (20 mM Tris-HCl pH 7.4, 150 mM NaCl, 5 mM MgCl2, 1 mM DTT, 100 µg/ml Cycloheximide) was used to prepare all sucrose solutions. Sucrose density gradients (5–45% w/v) were freshly made in SW41 ultracentrifuge tubes (Beckman) using a Gradient Master (BioComp Instruments) according to manufacturer's instructions. The lysate was then applied to a 4–45% sucrose continuous gradient and centrifuged at 35,000 rpm (Beckmann Coulter Optima XPN) for 3 h, at 4°C. The absorbance at 254 nm was monitored and recorded and 500 µl fractions were collected using a gradient collector (BioComp instruments). The protein in each fraction was pelleted with trichloroacetic acid, washed with acetone, and subjected to western blotting, as described above.
Ribosome pelleting assay
SARS-CoV-2 Calu-3 infected lysates were prepared as described above. 300 µl of the lysate was loaded onto a 900 µl 1M sucrose cushion in polysome buffer (described above) in Beckman centrifugation tubes. Ribosomes were pelleted by centrifugation at 75,000 rpm for 2 h, at 4°C, using a Beckmann MLA-130 rotor (Beckman Coulter Optima MAX-XP). After removing the supernatant, ribosome pellets were resuspended in polysome buffer and was used for western blotting, as described above.
Optical tweezers constructs
5' and 3' DNA handles, and the template for in vitro transcription of the SARS-CoV-2 putative pseudoknot RNA were generated by PCR using the pMZ_lambda_OT vector. The 3′ handle was labeled during the PCR using a 5′ digoxigenin-labeled reverse primer. The 5′ handle was labeled with Biotin-16-dUTP at the 3′ end following PCR using T4 DNA polymerase. The RNA was in vitro transcribed using T7 RNA polymerase. Next, DNA handles (5′ and 3′) and in vitro transcribed RNA were annealed in a mass ratio 1:1:1 (5 µg each) by incubation at 95°C for 10 min, 62°C for 2 h, 52°C for 2 h and slow cooling to 4°C in annealing buffer (80% formamide, 400 mM NaCl, 40 mM HEPES, pH 7.5, and 1 mM EDTA, pH 8) to yield the optical tweezer suitable construct (Fig. 4E). Following the annealing, samples were concentrated by ethanol precipitation, pellets were resuspended in 40 µl RNase-free water, and 4 µl aliquots were stored at − 20°C until use.
Optical tweezers data collection and analysis
Optical tweezers measurements were performed using a commercial dual-trap platform coupled with a microfluidics system (C-trap, Lumicks). For the experiments, optical tweezers (OT) constructs were mixed with 4 µl of polystyrene beads coated with antibodies against digoxigenin (AD beads, 0.1% v/v suspension, Ø 1.76 µm, Spherotech), 10 µl of assay buffer (20 mM HEPES, pH 7.6, 300 mM KCl, 5 mM MgCl2, 5 mM DTT and 0.05% Tween) and 1 µl of RNase inhibitor. The mixture was incubated for 20 min at room temperature in a final volume of 19 µl and subsequently diluted by the addition of 0.5 ml assay buffer. Separately, 0.8 µl of streptavidin-coated polystyrene beads (SA beads, 1% v/v suspension, Ø 2 µm, Spherotech) were mixed with 1 ml of assay buffer. The flow cell was washed with the assay buffer, and suspensions of both streptavidin beads and the complex of OT construct with anti-digoxigenin beads were introduced into the flow cell. During the experiment, an anti-digoxigenin (AD) bead and a streptavidin (SA) bead were trapped and brought into proximity to allow the formation of a tether. The beads were moved apart (unfolding) and back together (refolding) at a constant speed (0.05 µm/s) to yield the force-distance (FD) curves. The stiffness was maintained at 0.31 and 0.24 pN/nm for trap 1 (AD bead) and trap 2 (SA bead), respectively. For experiments with ZAP-S protein, recombinantly expressed ZAP-S was diluted to 400 nM in assay buffer and introduced to the flow cell. FD data were recorded at a rate of 78125 Hz.
Raw data files were processed using our custom-written python algorithm called Practical Optical Tweezers Analysis TOol (POTATO, https://github.com/lpekarek/POTATO.git, manuscript in preparation). In brief, raw data were first down sampled by a factor of 20 to speed up subsequent processing, and the noise was filtered using Butterworth filter (0.05 filtering frequency, filter order 2). Numerical time derivation was calculated separately for force and distance data. These derivations were statistically analyzed to identify the folding events and their coordinates. For data fitting, we employed a combination of two worm-like chain models (WLC1 for the fully folded double-stranded parts and WLC2 for the unfolded single-stranded parts) as described previously 53. Firstly, the initial contour length of the folded RNA was set to 1231 nm, and the persistence length of the double-stranded part was fitted 53. Then, the persistence length of the unfolded RNA was set to 1 nm, and the contour length of the single-stranded part was fitted. Data were statistically analyzed, and the results were plotted using Prism 8.0.2 (GraphPad).
qRT-PCR
Total RNA was isolated as described previously 75, and the reverse transcription using RevertAid (Invitrogen) was primed by oligo(dT). Reactions of quantitative real-time PCR (qRT-PCR) were set up using POWER SYBR green Master-mix (Invitrogen) according to manufacturer's instructions and analyzed on the CFX96 Touch Real-Time PCR Detection System (Bio-Rad) under the following cycling condition: 50°C for 2 min, 95°C for 2 min, followed by 40 cycles of 95°C for 15 s and 60°C for 30 s, and ending with a melt profile analysis. The fold change in mRNA expression was determined using the 2-ΔΔCt method relative to the values in uninfected samples, after normalization to the housekeeping gene (geometric mean) GAPDH. Statistical analysis was conducted using an unpaired two-tailed t-test with Welch's correction comparing delta Ct values of the respective RNA in uninfected and infected cells. The results were plotted using Prism 8.0.2 (GraphPad).
Quantification and statistical analysis
All statistical analyses and software used have been mentioned in the Figure Legends and Materials & Methods. Measurements from the in vitro western blot assay and in vivo dual fluorescence assay resulted from 3 technical replicates. Measurements from single-molecule experiments resulted from a specified number (n) of traces from a single experiment. For the ensemble MST analysis, all analysis from 3 individual replicates was performed in Nanotemper MO. Affinity software.