ABSTRACT
SARS-CoV-2 recently emerged as a human pathogen and is the causative agent of the COVID-19 pandemic. A molecular framework of how the virus manipulates host cellular machinery to facilitate infection remains unclear. Here, we focus on SARS-CoV-2 NSP1, which is proposed to be a virulence factor that inhibits protein synthesis by directly binding the human ribosome. Using extract-based and reconstitution experiments, we demonstrate that NSP1 inhibits translation initiation on model human and SARS-CoV-2 mRNAs. NSP1 also specifically binds to the small (40S) ribosomal subunit, which is required for translation inhibition. Using single-molecule fluorescence assays to monitor NSP1–40S subunit binding in real time, we demonstrate that eukaryotic translation initiation factors (eIFs) modulate the interaction: NSP1 rapidly and stably associates with most ribosomal pre-initiation complexes in the absence of mRNA, with particular enhancement and inhibition by eIF1 and eIF3j, respectively. Using model mRNAs and an inter-ribosomal-subunit FRET signal, we elucidate that NSP1 competes with RNA segments downstream of the start codon to bind the 40S subunit and that the protein is unable to associate rapidly with 80S ribosomes assembled on an mRNA. Collectively, our findings support a model where NSP1 associates with the open head conformation of the 40S subunit to inhibit an early step of translation, by preventing accommodation of mRNA within the entry channel.
SIGNIFICANCE STATEMENT SARS-CoV-2 is the causative agent of the COVID-19 pandemic. A molecular framework for how SARS-CoV-2 manipulates host cellular machinery to facilitate infection is needed. Here, we integrate biochemical and single-molecule strategies to reveal molecular insight into how NSP1 from SARS-CoV-2 inhibits translation initiation. NSP1 directly binds to the small (40S) subunit of the human ribosome, which is modulated by human initiation factors. Further, NSP1 and mRNA compete with each other to bind the ribosome. Our findings suggest that the presence of NSP1 on the small ribosomal subunit prevents proper accommodation of the mRNA. How this competition disrupts the many steps of translation initiation is an important target for future studies.
INTRODUCTION
Beta-coronaviruses (CoVs) are a family of RNA viruses that include human pathogens1. In the last two decades, two CoVs have emerged from animal hosts to cause epidemic diseases of the human respiratory tract: Severe Acute Respiratory Syndrome (SARS-CoV, in 2002)2,3 and Middle East Respiratory Syndrome (MERS-CoV, in 2012)4. A third CoV emerged in late 2019 — SARS-CoV-2 — that is responsible for the ongoing COVID-19 pandemic5. Given the lack of effective therapies against SARS-CoV-2, there is an urgent need for a molecular understanding of how the virus manipulates the machineries present in human cells.
SARS-CoV-2 and the closely-related SARS-CoV have single-stranded, positive-sense RNA genomes nearly 30 kb in length6,7. Upon entry of a virion into human cells, the genomic RNA is released into the cytoplasm where it must hijack human translation machinery to synthesize viral proteins8. As the genomic RNA has a 7-methylguanosine (m7G) cap on the 5’-terminus, viral protein synthesis likely proceeds via a process reminiscent of that which occurs on typical human messenger RNAs (mRNAs)9. However, as viral proteins accumulate, human translation is inhibited and host mRNAs are destabilized, which facilitates suppression of the host immune response10–13.
Studies on SARS-CoV have implicated non-structural protein 1 (NSP1), the first encoded viral protein, as a virulence factor with a key role in the shutdown of host translation10,11,14. In infected cells or upon its ectopic expression, NSP1 inhibits human translation, which is dependent on its association with the small (40S) subunit of the human ribosome12–17. In a connected but separable activity, NSP1 destabilizes at least a subset of human mRNAs, likely via recruitment of an unidentified human endonuclease12,13,15,16,18. NSP1 from SARS-CoV-2 is expected to employ similar mechanisms, given its approximately 85% sequence identity with the SARS-CoV protein. Thus, NSP1 has a near singular ability to dramatically disrupt host gene expression; yet, the mechanism by which this inhibition occurs is not clear.
The 40S subunit is the nexus for translation initiation, recruiting an m7G-capped mRNA through a multistep, eukaryotic initiation factor(eIF)-mediated process. Prior to recruitment of an mRNA, the 40S subunit is bound by numerous eIFs, including eIF1, eIF1A, eIF3, eIF5, and the eIF2–Met-tRNAMeti–GTP ternary complex (TC)19. The eIFs make extensive contacts with the 40S subunit, including the ribosomal A and P sites20,21. They also manipulate the dynamics of the 40S head region to facilitate mRNA recruitment, which has structural consequences at both the mRNA entry (3’ side of mRNA) and exit (5’ end of mRNA) channels. Following mRNA recruitment and scanning of the 5’ untranslated region (UTR), a series of compositional and conformational changes occur22,23. This ultimately repositions the 40S subunit head into the closed conformation and accommodates the anticodon-stem-loop of the initiator tRNA at the start codon20–23. Taken together, the intrinsic dynamics of translation initiation present many opportunities and obstacles for NSP1 association with the ribosome, and its subsequent inhibition of translation.
Here, we merge biochemical and single-molecule approaches to probe the molecular function of SARS-CoV-2 NSP1 and its interaction with the human ribosome. We illustrated that NSP1 inhibited translation initiation on human and SARS-CoV-2 model mRNAs, and then determined how the NSP1– 40S subunit interaction was modulated by eIFs and mRNA. Our results reveal allosteric control of NSP1 association by eIF1 and likely eIF3j, and competition between NSP1 and mRNA to bind the ribosome. Combined with recent structural studies, our study suggests a mechanism for how NSP1 inhibits translation initiation.
RESULTS
NSP1 inhibited translation initiation of host and SARS-CoV-2 model mRNAs
The sequence conservation (≈85% identity) of SARS-CoV-2 NSP1 with the homologous protein from SARS-CoV suggested that it would inhibit human translation (Sup. Fig. 1A). To test this hypothesis, we employed a cell-free in vitro translation (IVT) assay using HeLa cellular extract. As a model of a host mRNA, the 5’ and 3’ UTRs from human GAPDH mRNA were fused to a nanoLuciferase (nLuc) open-reading frame (ORF) (Fig. 1A). This GAPDH reporter mRNA was added to IVT reactions that contained increasing concentrations of purified SARS-CoV-2 NSP1 (Sup. Fig. 1B). Following incubation for 45 minutes, we observed a concentration-dependent reduction in nLuc activity by wild-type NSP1 (Fig. 1B), with an IC50 of 510 ± 20 nM (95% C.I.; R2 =0. 83). Given its multi-functional potential, we also tested two mutant NSP1 proteins with alanine substitutions that replaced both the conserved RK124-125 residues (RK/AA) implicated in mRNA destabilization or the conserved KH164-165 residues (KH/AA) implicated in ribosome binding (Sup. Fig. 1). As predicted based on the SARS-CoV protein16, the NSP1(RK/AA) mutant inhibited translation similar to the wild-type protein (IC50 ≈ 420 ± 11 nM, 95% C.I., R2 = 0.89), while the NSP1(KH/AA) mutant failed to inhibit translation (Fig. 1B). The inhibitory effect of NSP1 on protein synthesis therefore may be meditated by an NSP1-ribosome interaction, independent of mRNA degradation.
In the context of infection, full-length SARS-CoV-2 genomic RNA and its sub-genomic mRNAs must be translated in the presence of NSP1 protein. To examine whether RNA elements within the viral UTRs facilitate evasion of NSP1-mediated inhibition, we constructed model SARS-CoV-2 mRNAs in which the nLuc ORF was flanked on the 5’ end by either the full-length viral 5’UTR or the sub-genomic 5’ leader sequence (LDR) (Fig. 1C). At the 3’ end, we fused two different versions of the 3’UTR, beginning after the stop codon for N protein (L) or ORF10 (S), to account for ambiguity in ORF10 coding potential24. Translation of all four model viral mRNAs was reduced significantly by approximately 50% upon addition of 400 nM NSP1 relative to reactions that lacked NSP1 (p ≤ 0.0008, unpaired t-test) (Fig. 1D). The degree to which NSP1 inhibited translation of the viral reporters was similar to the inhibition observed for GAPDH reporter mRNA (p ≥ 0.2, one-way ANOVA). However, translation of the 5’UTR-3’UTR(S) model viral mRNA was 36% higher than that of the host and other viral reporters (p ≤ 0.0006, one-way ANOVA) in our IVT system (Fig. 1D). This may suggest that enhanced translational activity of viral RNAs relative to host mRNAs may play a role in SARS-CoV-2 infection.
To determine the phase of protein synthesis inhibited by NSP1, we employed a real-time translation assay in which nLuc activity was continuously monitored in situ. From the resulting time course data, we extracted the protein synthesis rate, mean synthesis time, and translational productivity25 from samples where NSP1 (400 nM) was either omitted from the IVT reaction, added simultaneously with mRNA (GAPDH), added after the reaction was pre-incubated with mRNA, or pre-incubated in the extract prior to mRNA addition (Fig. 1E). Pre-incubation of the extract with mRNA allows translation to initiate in the absence of NSP1, while pre-incubation with NSP1 primes the extract for inhibition. If NSP1 disrupts translation initiation, the inhibition would be dependent on the temporal availability of NSP1; otherwise, if NSP1 inhibits translation elongation or termination, the inhibition would be insensitive to the timing of NSP1 addition.
The impact of NSP1 on translation was dependent on its time of addition to the IVT reaction. Consistent with the end-point assays, we observed an approximate 2-fold reduction in the protein synthesis rate when NSP1 and GAPDH reporter mRNA were added simultaneously (p < 0.02, one-way ANOVA) (Fig 1F,G). While protein synthesis rates for the other samples were similar to the reaction that lacked NSP1, we observed a dramatic delay in the appearance of nLuc signal when cell extracts were pre-incubated with NSP1, but not vice versa (Fig 1F). This lag between the addition of mRNA (t = 0) and the initial appearance of nLuc signal is the time needed for a full round of translation (‘synthesis time’; sum of initiation, elongation, and termination). To compare synthesis times quantitatively, we fit the second derivative of the nLuc time course data to a Gaussian distribution (Fig 1H). In this analysis, the mean of the distribution represents the mean synthesis time and its amplitude is a gauge of translational productivity. As suggested by the raw data, the mean synthesis time when extracts were pre-incubated with NSP1 (637 ± 41 s) increased by 54% (p<0.0001, one-way ANOVA) compared to the reaction without NSP1 (413 ± 5 s), while mean times of the two other conditions (411 ± 6 s and 435 ± 8 s) were similar to the control (Fig. 1I). Pre-incubation with NSP1 also reduced translational productivity approximately 2-fold (p = 0.04, one-way ANOVA) (Fig. 1J). In contrast, pre-incubation with mRNA yielded mean synthesis times and translation productivity similar to reactions that lacked NSP1 (Fig. 1I,J). Thus, ribosomes pre-loaded with mRNA evaded NSP1-mediated inhibition, which strongly suggests that NSP1 is a potent inhibitor of translation initiation, perhaps linked to mRNA recruitment or accommodation.
NSP1 stably associated with ribosomal pre-initiation complexes
To test whether NSP1 binds directly to the human 40S ribosomal subunit, we employed native gel shift assays with purified components. Using an 11-amino acid ybbR tag, single cyanine dye fluorophores were conjugated site-specifically onto purified NSP1 (Supp.Fig. 2A,B)26,27. When incubated with increasing concentrations of purified 40S subunits, the amount of fluorescently-labeled NSP1 that co-migrated with 40S subunits increased (Supp. Fig. 2C). In contrast, NSP1 did not co-migrate with human 60S or yeast 40S subunits (Supp. Fig. 2D). Whereas NSP1(KH/AA) was unable to block the NSP1–40S subunit interaction, inclusion of either wild-type NSP1 or NSP1(RK/AA) at 150-fold molar excess prevented co-migration of labeled NSP1 with human 40S subunits (Supp. Fig. 2E). Thus, NSP1 specifically binds to the human 40S ribosomal subunit, dependent on the presence of an intact KH motif in the C-terminus of the protein. Together with our extract-based assays, these data indicate that NSP1 binds the human 40S subunit to inhibit translation initiation.
Throughout translation initiation, there is a complex choreography between eIFs and the 40S ribosomal subunit. Yet, it was unknown how the NSP1–40S subunit interaction is affected by eIFs, either through direct interactions or induced changes in ribosome conformation. We therefore initially examined NSP1 binding to the 40S subunit in the presence of either 6 µM eIF1, eIF1A, or eIF3j. Each of these canonical eIFs bind with high affinity to the 40S subunit28 and may alter its conformation20,21. Inclusion of eIF1 increased the intensity of the NSP1–40S subunit band approximately 2-fold (mean ≈ 2 ± 0.4, 95% C.I.), while eIF3j eliminated the band (Supp. Fig. 3A-C). Unlike eIF1 and eIF3j, eIF1A had little impact on the NSP1–40S subunit complex (Supp. Fig. 3A,D). The NSP1–40S subunit interaction therefore was modulated inversely by two eIFs.
To define the kinetics of NSP1 binding to 40S subunits, we established a single-molecule assay to monitor NSP1 association with ribosomal pre-initiation complexes directly in real time. First, biotin was attached to purified 40S subunits that contained the ybbR tag on the ribosomal protein RACK1 (Supp. Fig. 4A), using the same strategy as with fluorescent dyes29. We then tethered preassembled eIF1–biotinylated-40S subunit complexes to thousands of zero-mode waveguide (ZMW) surfaces coated with neutravidin (Supp. Fig. 4B)30. Upon start of data acquisition, Cy3-NSP1 was added to the ribosomal complex (Fig. 2A), which had translation-inhibition and ribosome-binding activities similar to the wild-type protein (Supp. Fig. 4C,D). Association of NSP1 with the 40S subunit was manifested by a burst of Cy3 fluorescence (Fig. 2B). When NSP1 was added at 75 nM, the majority of ZMWs (56 ± 7 %) with tethered eIF1–40S subunit complexes contained at least one NSP1 binding event (≥ ≈ 5 s in length) (Fig. 2C). In contrast, the number of ZMWs with binding events was reduced dramatically in the absence of the tethered complex (9 ± 4 %). Similarly, at two different NSP1 concentrations, pre-incubation with 2.5 µM eIF3j reduced NSP1 binding to baseline levels (from 48 ± 7 % and 60 ± 7 %, to 6 ± 3 and 7 ± 3 %). Results consistent with specific binding also were obtained using total internal reflection fluorescence microscopy (TIRFM) at equilibrium (Supp. Fig. 4E,F). Thus, our assay directly monitored real-time association of NSP1 with tethered 40S ribosomal complexes and demonstrated competition by eIF3j for NSP1– 40S complex formation.
NSP1 bound the eIF1–40S subunit complex with high affinity. As predicted for a bimolecular interaction, NSP1 association times (the time elapsed from its addition until appearance of Cy3 signal) decreased with increasing concentration of NSP1 at 20 °C (Fig. 2D and Supp. Fig. 4G). Linear-regression analysis of the observed rates at various NSP1 concentrations yielded a bimolecular association rate of 0.3 ± 0.1 µM-1 s-1 (95% C.I.) (Fig. 2D and Supp. Table 2). The observed lifetime of the NSP1–40S subunit interaction (the duration of the Cy3 signal) was dependent on the power of the excitation laser (Supp. Fig. 4H,I), which indicated our measurements were limited by dye photostability. Nevertheless, with our longest measured lifetime as a lower bound (238 ± 6 s, 95% C.I.), we estimated that the equilibrium dissociation constant (KD) of the NSP1 interaction with eIF1–40S subunit complexes was ≤ 12.5 nM at 20 °C, similar to that of eIFs28.
NSP1 rapidly and stably associated with different ribosomal pre-initiation complexes. When added at 25 nM to eIF1–40S subunit complexes, median NSP1 association times (see, Material and Methods) were similar at 20 and 25 °C (67–101 s and 66–97 s, 95% C.I.), but they decreased nearly two-fold at 30 and 35 °C (38–54 s and 25–42 s, 95% C.I.) (Fig. 2E and Supp. Fig. 4J). We therefore measured NSP1 association times and lifetimes with 40S subunits in complex with various canonical eIFs at 30 °C. Consistent with our gel-based assays, the median NSP1 association time (at 25 nM) was decreased about 2-fold in the presence of eIF1 relative to 40S subunits alone (38–54 s versus 91–137 s, 95% C.I.) (Fig. 2F,G and Supp. Fig. 4K). Further inclusion of eIF1A, eIF3 that lacked the 3j subunit (eIF3), eIF5, and/or an eIF2–Met-tRNAMeti –GMPPNP ternary complex (TC-GMPPNP) also yielded modest reductions in NSP1 association times. NSP1 lifetimes on the various eIF–40S subunit complexes were similar (Fig. 2G and Supp. Fig. 4L) and likely limited by dye photostability. Together with the temperature-dependence, the eIF-mediated modulation of NSP1 association with the 40S subunit, particularly by eIF1 and eIF3j, suggested that NSP1 may associate with a particular conformation of the 40S ribosomal subunit.
mRNA within the entry channel of the 40S subunit inhibited NSP1 association
While this work was in progress, multiple groups reported structures of NSP1 bound to the human ribosome31,32. Despite its apparent flexibility, the N-terminal globular domain of NSP1 was localized to the solvent-exposed surface of the 40S subunit, near the entrance to the mRNA entry channel (Supp.Fig. 5A). This domain appears anchored by the two most C-terminal α-helices of NSP1, which were dynamic and unstructured in the free SARS-CoV NSP1 structure solved by NMR33; in the NSP1–40S subunit complex, these helices were well-resolved, docked within the mRNA entry channel where they contact ribosomal proteins uS3 and uS5, and helix 18 of the 18S rRNA. Further guided by our findings above, we therefore hypothesized that association of NSP1 with the 40S subunit would be sensitive to the conformation of the mRNA entry channel and to the presence of mRNA within it.
To test our hypothesis, we used the internal ribosome entry site (IRES) from hepatitis C virus (HCV), a structured RNA that directly binds to the human 40S subunit with high affinity (2–4 nM)34 (Supp.Fig. 5B). A flexible segment of the IRES (domain II) swivels the head of the 40S subunit to open the mRNA entry channel35–37 and allows accommodation of the mRNA coding region downstream from the start codon. We generated five HCV IRES model RNAs that were 5’-biotinylatyed and contained 0, 6, 12, 24, or 48 nucleotides downstream (3’) of the start codon (Fig. 3A). Based on the above structural models, mRNAs with more than 6 nucleotides after the start codon are predicted to at least partially occlude the NSP1 binding site within the mRNA entry channel. Following incubation of the biotinylated RNAs with fluorescently-labeled (Cy5 dye) ribosomal subunits, IRES–40S subunit complexes were tethered to a ZMW surface, and Cy3-NSP1 was added at 25 nM upon start of data acquisition (Fig. 3B,C).
Short and long segments of RNA downstream of the start codon had opposite effects on NSP1 association. With HCV+0 and HCV+6, NSP1 efficiently (80 ± 3% and 81 ± 3%, 95% C.I.) and rapidly associated (apparent kon-1 ≈ 11 ± 0.6 s and 8 ± 0.7 s, 95% C.I.) with tethered IRES-40S subunit complexes when added at 25 nM (Fig. 3D-F and Supp. Table 3). NSP1 association to this complex was approximately 2.5-fold faster than to the eIF1– 40S subunit complex (apparent kon-1 ≈ 29 ± 0.2 s, 95% C.I.) (Fig. 3F and Supp. Fig. 5C). In contrast, NSP1 associated less efficiently (52 ± 4% and 31 ± 4%) and much more slowly (apparent kon-1 ≈ ≥ 520 s) with HCV+24 and HCV+48 complexes relative to HCV+0 (Fig. 3D-F). We reasoned that the relative lack of inhibition we observed on HCV+12 could be due to inefficient accommodation of the mRNA into the entry channel. Indeed, inclusion of eIFs utilized for initiation by the HCV IRES (eIF1, eIF1A, eIF5, eIF3, and TC-GMPPNP) exacerbated the increase in NSP1 association times on HCV+12 to nearly 8-fold slower (apparent kon-1 ≈ 77 ± 2 s, 95% C.I.) relative to HCV+0 (Fig. 3D-F). In parallel to inhibited association, we also observed at least 3-fold decreased median NSP1 lifetimes on HCV+24 and HCV+48 complexes (11–16 s and 9–15 s, 95% C.I.) relative to the dye-limited measurements on HCV+0 and HCV+6 (29–50 s and 38–58 s) (Fig. 3F and Supp. Fig. 5D). We therefore used the same strategy as above to estimate that the lifetime of NSP1 on the 40S– HCV+0 complex was ≥ ≈230 s (Supp. Fig. 5E). Consequently, the KD of the NSP1 interaction with the IRES–40S subunit complex was increased at least 500-fold (from ≤ ≈4 nM to ≥ ≈2–3 µM) by long segments of RNA downstream of the start codon.
To test whether our findings were generalizable to other RNAs, we preformed analogous experiments as above in two formats using an unstructured model mRNA (M+41) that contained 41 nucleotides downstream of the start codon (Fig. 3A). In the first, 3’-biotinylated M+41 RNA bound to 40S-Cy5 subunits were tethered to the imaging surface (Supp.Fig. 5F,G). In the second, 40S-biotin subunits bound to fluorescently-labeled M+41 were tethered (Supp.Fig. 5H,I). In both scenarios, we observed inefficient association of NSP1 with the mRNA–40S subunit complexes (16 ± 3% with 40S-biotin, 95% C.I.) and at least 48-fold increases in NSP1 association times (apparent kon-1 ≈ ≥ 500 s for both) relative to HCV+0 or tethered ribosomal complexes that lacked the model mRNA (apparent kon-1 ≈ 27 ± 3 s, 95% C.I.) (Fig. 3D-F, Supp.Fig. 5J,K, and Supp. Table 3). NSP1 association with mRNA–40S subunit complexes therefore was inhibited markedly by RNA segments downstream of the start codon that at least partially occlude its binding site in the mRNA entry channel.
Intriguingly, NSP1 also has been visualized bound to 80S ribosomal complexes31. To examine whether NSP1 could associate with 80S ribosomes, we used CRISPR-Cas9 and homology-directed repair to establish a FRET signal between the 40S and 60S subunits of the ribosome, analogous to our signal in yeast38. The ybbR tag was appended to all endogenous copies of ribosomal proteins uS19 (40S subunit) or uL18 (60S subunit) (Supp.Fig. 6A-D), which are within predicted FRET distance (≈50 Å) in structural models of 80S ribosomes (Fig. 4A). The tagged ribosomes were functional in cells (Supp.Fig. 6E), and purified 40S-ybbR and 60S-ybbR subunits were labeled efficiently (50– 80%) with Cy3 (FRET donor) and Cy5 (FRET acceptor) fluorescent dyes, respectively (Supp.Fig. 6F). After incubation with the IRES from the intergenic region of cricket paralysis virus (CrPV IRES), which assembles ribosomal subunits into 80S ribosomes independent of eIFs39, we observed a FRET efficiency distribution (mean ≈ 0.5 ± 0.01, 95% C.I.) between the labeled 40S and 60S subunits, consistent with structural predictions (Fig. 4B,C).
By leveraging the FRET signal and the CrPV IRES, we examined whether NSP1 associated with 80S ribosomes assembled on an mRNA (Fig. 4D). We generated RNAs as above with 1, 6, and 48 nucleotides downstream of the CCU codon present in the ribosomal A site (Supp. Fig. 7A). With these models, the RNA is shifted three nucleotides further into the entry channel relative to the HCV IRES40. Therefore, CrPV+6 and CrPV+48 will have mRNA that at least partially occludes the NSP1 binding site, while CrPV+1 will not. When added at 25 nM to 40S–
CrPV+1 complexes, we observed slower (median association time ≈ 119–189 s, 95% C.I.) and less efficient (45 ± 4%, 95% C.I.) Cy5.5-NSP1 association relative to that of 40S–HCV+0 complexes (median ≈ 46–82 s and 72 ± 4%, 95% C.I.) (Supp. Fig. 7B,C and Supp. Table 4). This finding likely reflects heterogeneity of the 40S subunit head conformation when bound to the CrPV IRES41, unlike the near-homogenous open conformation induced by the HCV IRES. Further inclusion of 60S subunits to yield 80S– CrPV+1 complexes inhibited NSP1 association (median association time ≈ 235–285 s, 95% C.I.), similar to the inhibition observed on both 80S–CrPV+6 and 80S–CrPV+48 complexes (median association times ≈ 243–292 s and 288–354 s, 95% C.I.) (Fig. 4E,F and Supp. Fig. 7C). Thus, even when mRNA was absent from it, the conformation of the mRNA entry channel on 80S–CrPV IRES complexes was incompatible with rapid NSP1 association.
Our findings support a model where NSP1 associates with the open conformation of the 40S ribosomal subunit, in the absence of mRNA within the entry channel. They also suggest that NSP1 is unable to rapidly bind 80S ribosomes in either the canonical or rotated states prior to peptide bond formation and translocation. Future studies will be needed to examine whether NSP1 accesses other states of the 80S ribosome.
NSP1 remained bound to 40S subunits upon association with model mRNAs
Our data above indicated that mRNA inhibited NSP1 association with the 40S ribosomal subunit. However, it remained unclear whether RNA segments downstream of a start codon would destabilize the NSP1–40S subunit complex upon mRNA recruitment. To focus solely on NSP1, the 40S subunit, and mRNA, we again leveraged the HCV IRES. Using Cy5.5-NSP1 and 40S-Cy3 subunits, we pre-formed NSP1–40S complexes and added the complex at 15 nM to ZMWs with surface-immobilized HCV+0 or HCV+48 model mRNAs (Supp. Fig. 8A-C). On HCV+0 and HCV+48, NSP1 co-associated with 57 ± 6% and 60 ± 6% (99% C.I.) of 40S subunits (Fig. 5A,B and Supp. Table 5), which indicated near saturation of 40S subunits with NSP1. Association of the NSP1–40S subunit complex with the tethered RNAs had kinetics similar to 40S subunits alone (Fig. 5C,D and Supp. Fig. 8D,E). After association, NSP1 remained bound to the 40S subunit for approximately 60 s on average (median lifetimes 39–74 s and 52–82 s, 95% C.I.) (Fig. 5D and Supp. Fig. 8F). Consistent with our findings using the HCV IRES, NSP1 lifetimes on the ribosomal subunit were similar for both CrPV+1 and CrPV+48 model RNAs (Supp. Fig. 8G-N). Thus, the stability of NSP1 on the 40S subunit was unaffected upon direct association with the structured mRNAs we examined.
To delineate competition between NSP1 and mRNA for the 40S cleft, we next asked whether NSP1 could re-associate stably with single IRES–40S subunit complexes and how re-association was impacted by long segments of RNA downstream of the start codon. Indeed, following loss of the initial NSP1 signal (due to dye photobleaching or NSP1 departure), 80 ± 6% (99% C.I.) of 40S–HCV+0 complexes had at least one additional stable (≥ 20 s) NSP1 binding event (Fig. 5A,B and Supp. Fig. 8B). In contrast, only 32 ± 7% (99% C.I.) of 40S– HCV+48 complexes had a second, stable NSP1 event (Fig. 5A,B and Supp. Fig. 8C). When multiple NSP1 association events were observed on a single 40S–IRES complex, the time that separated them was at least 55-fold longer on HCV+48 (apparent kon-1 ≈ ≥ 700 s) relative to HCV+0 (apparent kon-1 ≈ 13 ± 1 s, 95% C.I.) (Fig. 5C and Supp. Fig. 8O). The lifetimes of initial and re-associated NSP1 binding events were similar (Fig. 5D and Supp. Fig. 8P). Together, these findings indicated that once NSP1 dissociated from the 40S–HCV+48 complex, mRNA was accommodated more rapidly into the mRNA entry channel, thereby inhibiting re-association of NSP1.
DISCUSSION
Shutdown of host protein synthesis is a common feature of viral infection. Most characterized mechanisms involve the covalent inactivation of key eIFs or their regulators (e.g., eIF2 and eIF4F42). Here, we provide insight into a distinct form of translation inhibition employed by SARS-CoV-2 and closely-related CoVs. The first protein encoded in the viral genomic RNA, NSP1, directly targets the small subunit of the human ribosome to inhibit protein synthesis. Based on our findings and recent structural studies31,32, we suggest that NSP1 preferentially associates with the open conformation of the 40S subunit to prevent proper accommodation of mRNA during translation initiation (Fig. 6).
NSP1 is a potent inhibitor of human translation initiation. When we added purified NSP1 to human HeLa cell extract, we observed a dramatic reduction in translation of our model for human GAPDH mRNA. Inhibition was specific, as mutations in two NSP1 amino acids (KH164-165) necessary for 40S subunit binding abrogated the NSP1 inhibition. The apparent IC50 value for wild-type NSP1-mediated inhibition suggests near stochiometric association of NSP1 with 40S subunits in the cell extract, which agrees well with our best estimate for the KD of the interaction (≤ 4 nM). Based on our kinetic analysis of protein synthesis, we demonstrate that such NSP1 association with the 40S subunit inhibits the initiation phase of translation. Consistently, a recent study31 illustrated that ectopic expression of SARS-CoV-2 NSP1 reduced the abundance of actively-translating polysomes and increased the abundance of 80S monosomes, a hallmark of translation initiation defects. Similar findings also have been observed with the SARS-CoV protein16,17.
NSP1 inhibited translation of SARS-CoV-2 model mRNAs at levels comparable to that for a model human mRNA. This finding suggests that NSP1 is a general inhibitor of protein synthesis. The high affinity of NSP1 for the 40S subunit likely demands buildup of NSP1 protein levels before translation is inhibited broadly, which may enable viral protein synthesis to proceed unimpeded during early stages of infection. Once NSP1 has accumulated, the increased translation efficiency of the viral mRNAs relative to human mRNAs we and others32 have observed may enable the virus to synthesize sufficient amounts of viral proteins, even when translation is largely shutdown. However, it is possible that another viral protein, a different segment of the viral genome, or another mechanism (e.g., sequestration) allows the virus to evade translation inhibition. Future studies in the context of infected human cells are needed to deconvolute these possibilities.
To inhibit translation initiation, multiple lines of evidence suggest that NSP1 preferentially associates with the open conformation of the 40S subunit. First, of all eIFs we examined, NSP1 association was enhanced the most (≈ 2-fold) by eIF1. This protein binds to the 40S subunit at the ribosomal P site43–45, where it has a critical role during start codon recognition46–50. Upon its association, eIF1 repositions the 40S subunit head, which concomitantly opens the mRNA entry channel51,52. Given that eIF1 and NSP1 binding sites are non-overlapping, our findings indicate that eIF1 allosterically enhances NSP1 association, likely by altering the conformation of the mRNA entry channel to allow NSP1 access. Second, the most rapid NSP1 association with the 40S subunit we observed was in the presence of the HCV IRES. This structured RNA directly manipulates the ribosomal subunit to bypass eIFs and initiate translation53. One of its flexible segments, domain II, makes extensive contacts with the 40S subunit head, which swivels, opening the mRNA entry channel35–37. In our assays, the estimated rate of NSP1 association with the HCV+0 model mRNA was 3–4 µM-1 s-1, nearly an order of magnitude faster than with 40S subunits alone. Enhanced NSP1 association with this complex relative to the eIF1–40S subunit complex likely reflects inherent differences in their stability. Whereas the IRES–40S subunit complex is quite stable (KD ≈ 2 nM, koff ≈ 0.002 s-1)34,54, the eIF1–40S subunit interaction is more closed conformations. Consistently, NSP1 associated more slowly with the 40S–CrPV+1 complex, which likely also contains a heterogenous mix of entry channel conformations41.
In striking contrast, NSP1 competes with mRNA to bind the ribosome. When 40S subunits were pre-incubated with an mRNA that had at least 12 nucleotides downstream of the start codon, we observed marked inhibition of NSP1 association with the ribosomal subunit. On such mRNAs, NSP1 contacts with helix 18 of the 18S rRNA and ribosomal proteins uS3 and uS5 in the mRNA entry channel are at least partially occluded by the accommodated mRNA. Consistently, when we pre-incubated mRNA in extracts prior to NSP1 addition, NSP1 failed to inhibit translation in our real-time protein synthesis assays. And in the reciprocal experiment, pre-incubation with NSP1 yielded a delay in protein synthesis (≈ 200 s) similar in length to our best estimate for the lifetime of the NSP1–40S subunit interaction (at least ≈ 240 s). In parallel, NSP1 association with the 40S subunit also was inhibited by pre-incubation with eIF3j, which binds near the mRNA entry channel55,56. The C-terminus of eIF3j may sterically block NSP1 association55 or eIF3j may limit movement of the 40S subunit head to promote a conformation of the entry channel inaccessible to NSP1, which may reflect the likely preference of eIF3j for the closed conformation of the 40S subunit57.
In further support of a competition model, NSP1 remained bound to the 40S ribosomal subunit upon recruitment of an mRNA— regardless of its length. This finding suggests that mRNA itself is insufficient to dislodge or destabilize the NSP1–40S subunit interaction. Based on our observed lifetime of that interaction (at least ≈ 240 s), NSP1 likely will remain associated with the ribosomal subunit for longer than the time frame of translation initiation on transient (KD ≈ 50 nM, koff ≈ 0.36 s-1)28,50, which many mRNAs (< 60 s)58,59, thereby blocking full would lead to a mixed population of open and accommodation of the mRNA. Whether NSP1 can be dislodged by other host proteins or eIFs, such as helicases eIF4A, DDX3X, or DHX29, remains unclear. Nevertheless, once NSP1 dissociated, mRNA was accommodated into the entry channel, which prevented re-association of NSP1. Thus, our findings suggest that the presence of NSP1 or mRNA within the mRNA entry channel of the 40S subunit are mutually exclusive. It is feasible that such competition blocks eIF4F-mediated recruitment of an m7G-capped mRNA to the 40S subunit or a subsequent step in initiation, which awaits elucidation by future studies. Collectively, our work provides a biophysical foundation for deeper understanding of NSP1-mediated shutdown of host translation and its impact on SARS-CoV-2 viral pathogenesis.
MATERIALS AND METHODS
Molecular cloning
See Supplementary Table 6 for all relevant sequences.
NSP1
Codon-optimized SARS-CoV-2 NSP1 and relevant mutants were cloned into a vector purchased from the UC Berkeley QB3 MacroLab (vector 1B) using their standard protocol. Synthetic DNA that encoded wild-type, RK124-125AA mutant, KH164-165AA mutant, and ybbR-tagged NSP1 sequences were purchased from Integrated DNA Technologies (IDT). All sequences were verified by Sanger sequencing. The resulting plasmids encoded NSP1 proteins tagged on the N-terminus with a 6-histidine tag followed by a TEV protease cleavage site (NH2-6His–TEV–NSP1-COOH). Wild-type NSP1 reference sequence (nts 266-805) was obtained from NCBI GenBank accession MN997409.1. When noted, a ybbR tag was included either on the N-terminus (ybbR-NSP1) or the C-terminus (NSP1-ybbR).
HCV IRES
A synthetic DNA was purchased from IDT that contained a 5’ flanking sequence (pUC19 backbone sequence), a T7 promoter (TAATACGACTCACTATAG), and the HCV IRES (nts 1-344, including the AUG codon). Downstream of the AUG included the rest of domain IV, the core sequence, and a 3’ extension: AGCACGAATCCTAAACCTCAAAGAAAAACCGCCAGAACCATGGAAGAC. DNA templates that encoded HCV+0, +6, +12, +24, and +48 (relative to the ‘G’ of the start codon, 3’-end indicated by underlined nts) were generated via standard PCRs using NEB Phusion polymerase (25 cycles), a common 5’ primer (upstream of the T7 promoter), and specific 3’ primers.
CrPV IRES
A plasmid that encodes the IRES of the intergenic region of CrPV with the first codon (Ala) replaced with a Phe codon (TTC) was described previously60. The sequence from CCU (bolded) of the IRES to the 3’ end was: CCTTTCACATTTCAAGATACCGGCGCCATGGAAGACGCCAAAAACATAAAG. DNA templates were generated as for the HCV IRES, with the underlined nts representing the 3’-terminus. We selected +1 as the shortest segment to promote proper folding of the IRES.
GAPDH nanoLuciferase
The reporter was designed to contain a 5’ T7 promoter, the nLuc coding sequence (Promega) flanked by the 5’ and 3’UTRs from human GAPDH (NCBI GenBank accession: AF261085), a poly(A) tail, and PmeI and SpeI restriction enzyme consensus sites. The reporter construct was purchased from IDT as a plasmid with an pUCIDT backbone and propagated in the DH5α strain of E. coli. DNA sequence identity was confirmed by Sanger sequencing. The plasmid was linearized by restriction digest with SpeI (NEB, #R0133) for templated in vitro transcription.
SARS-CoV-2 nanoLuciferase
Viral DNA (NCBI GenBank accession MN997409.1) constructs were designed such that the nLuc coding sequence was flanked by either the full-length 5’UTR or the subgenomic 5’ leader sequence and one of two 3’’UTR sequences, and a poly A tail followed by a SpeI consensus sequence. The full-length 5’UTR incuded the first 27 nt of the viral ORF in order to maintain the predicted stem-loop structure within the 5’UTR. The 3’UTR (L) began after the N protein stop codon and the 3’UTR (S) began after ORF10. Synthetic DNAs were purchased as GeneBlocks from IDT (5’ leader constructs) or as Gene Parts from GenScript (5’UTR constructs). All synthetic DNAs had a 5’-terminal XmaI consensus sequence and 3’-terminal HindIII consensus sequence. Restriction digest cloning was used to insert viral reporter DNA into a pUC19 vector (NEB, #R10180S, #R3104S, #M2200S).
NSP1 expression, purification, & labeling
NSP1 expression plasmids were transformed into OneShot BL21(DE3) cells (Invitrogen) and grown overnight at 37 °C on LB agar plates supplemented with 50 µg/mL kanamycin. Liquid cultures of single colonies were grown to OD600 ≈ 0.5 at 37 °C in LB supplemented with kanamycin. Cultures were shifted to 18 °C for 30 minutes, 0.5 mM IPTG was added, and cultures were grown for 16-20 h at 18 °C. Cells were harvested by centrifugation at 5,000 x g for 15 min at 4 °C in a Fiberlite F9 rotor (ThermoFisher, cat. # 13456093). Cells were lysed by sonication in lysis buffer (20 mM Tris-HCl pH 8.0, 300 mM NaCl, 10% (v/v) glycerol, 40 mM imidazole, and 5 mM β-mercaptoethanol), and lysates were cleared by centrifugation at 38,000 x g for 30 min at 4 °C in a Fiberlite F21 rotor followed by filtration through a 0.22 µm syringe filter. Clarified lysate was loaded onto a Ni-NTA gravity flow column equilibrated in lysis buffer, washed with 20 column volumes (CV) of lysis buffer, 20 CV of wash buffer (20 mM Tris-HCl pH 8.0, 1 M NaCl, 10% (v/v) glycerol, 40 mM imidazole, and 5 mM β-mercaptoethanol), and 10 CV of lysis buffer. Recombinant proteins were eluted with five sequential CV of elution buffer (20 mM Tris-HCl pH 8.0, 300 mM NaCl, 10% (v/v) glycerol, 300 mM imidazole, and 5 mM β-mercaptoethanol). Fractions with recombinant protein were identified by SDS-PAGE analysis. The relevant fractions were dialyzed overnight at 4 °C into ybbR-labeling buffer (50 mM HEPES-KOH pH 7.5, 250 mM NaCl, 10 mM MgCl2, 10% (v/v) glycerol, and 1 mM DTT) or TEV Buffer (20 mM Tris-HCl pH 8.0, 250 mM NaCl, 10% (v/v) glycerol, 10 mM imidazole, and 5 mM β-mercaptoethanol), as appropriate. Fluorescent labeling via the ybbR tag was performed essentially as described26,29. Briefly, 14.5 µM ybbR-NSP1 or NSP1-ybbR were incubated at 37 °C for 90 min with 4 µM Sfp synthase enzyme and 20 µM of either Cy3-CoA, Cy5-CoA, or Cy5.5-CoA substrates, respectively. Free dye was removed via purification over 10DG-desalting columns (Bio-Rad, cat.# 7322010) equilibrated in TEV buffer. Dye-labeled and non-labeled NSP1 proteins were incubated with TEV protease at 22 °C for 1.5 hrs followed by 30 min at 37 °C. TEV protease, Sfp synthase, and the cleaved 6His tag were removed via a subtractive Ni-NTA gravity column equilibrated in TEV buffer, with the flow-through collected. NSP1 proteins were subjected to a final purification step using size exclusion chromatography (SEC) on a Superdex 75 column (23 mL) equilibrated in SEC buffer (20 mM HEPES-KOH pH 7.5, 250 mM KOAc, 10% (v/v) glycerol, and 1 mM DTT). Fractions containing NSP1 were concentrated using a 10 kD MWCO Amicon Ultra centrifugal filter, aliquoted, flash frozen on liquid N2, and stored at −80 °C. Protein concentrations were determined via absorption at 280 nm using a nanodrop for total protein, and at 548 nM or 646 nM for Cy3 or Cy5/5.5 labeled proteins, respectively. For ybbR-NSP1, labeling efficiencies were 50-70%. For NSP1-ybbR, the labeling efficiency was much lower (<20%), and the protein had reduced translation inhibition activity, which is why it was excluded from single-molecule analyses.
nLuc in vitro translation assays
HeLa cell-free translation (ThermoFisher, #88884) reactions setup according to manufacturer’s protocol were programed with a final mRNA concentration of 200 nM (endpoint) or 80 nM (real-time). For reactions containing NSP1, an equal volume of NSP1 buffer was added to the paired no NSP1 control reaction. Prism8 was used for dose-response analysis ([inhibitor] vs. response with variable slope (four parameters) nonlinear fit) and statistical analysis (one-way ANVA Turkey’s multiple comparisons test, unpaired t-test).
Endpoint assays
IVT reactions were incubated at 37°C for 45 min and then immediately transferred to an ice water bath and diluted 1:1 with cold Glo Lysis Buffer (Promega, #E2661). All samples were brought to room temperature and mixed with a 1:1 volume of nGlow solution (Promega, #N1110). Samples (90% of total volume) were loaded into non-adjacent wells of a 384-well plate. Sample luminescence was measured 7 min post nGlow solution addition using a BioTek Neo2 multi-mode plate reader (25°C, 114LUM1537 filter, gain of 135). Luminescence signal was monitored for an additional 30 min at 3 min intervals to verify luminesce signal of all samples decayed at the same rate.
Real-time assays
HeLa IVT reactions were prepared by addition nGlow substrate to the cell-free translation mix with a 1:10 v/v ratio. Before the addition of mRNA and/or NSP1, the IVT reactions were transferred to non-adjacent wells in a 384-well plate and equilibrated to 30°C in the plate reader. All other reagents were maintained at 30°C and then added to the IVT reactions according to the order-of-addition assay schematic outlined in Fig. 1D. The preincubation (30°C, 2 min) was performed in the plate reader. Kinetic monitoring of the samples (36 min, 15 s intervals) was initiated during the equilibration step. Reagent additions and plate transfer times were noted during the experiment and used to post-synchronize t0 to mRNA addition. Data was analyzed in MatLab using the approach developed by Vassilenko et al.25. The raw data was smoothed using Savitzky-Golay filtering (frame length ≤ 5, 2nd degree polynomial) to maintain the shape and magnitude of the raw data. The plateau value of the 1st derivative of the smoothed curve is equal to the synthesis rate25,61. Mean synthesis time, which is lag time between mRNA addition and the initial detection of nLuc signal, is the sum of initiation, elongation, and termination time. Synthesis times were extracted from the data by fitting the 2nd derivative of the smoothed data to a Gaussian distribution using the equation where a = amplitude, b = centroid (median), c = variance25,62. Statistical analysis of the real-time nLuc activity was done as described above.
Native gel assays
Native composite agarose-acrylamide gels were prepared as described63. Briefly, 2.75% acrylamide (37.5:1), 0.5% Nusieve GTG agarose composite gels were prepared in the following buffer: 25 mM Tris-OAc pH 7.5, 4 mM KOAc, 2 mM Mg(OAc)2, 0.5 mM DTT, 2.5% glycerol, 0.1% (v/v) TEMED, and 0.1% (v/v) fresh ammonium persulfate. The gels were cooled at 4 °C for 20 minutes and allowed to further polymerize at room temperature for 90 min. Prior to removal of combs, gels were placed at 4 °C for 15 min. Gels were run in ice-cold running buffer (25 mM Tris-OAc pH 7.5, 4 mM KOAc, and 2 mM Mg(OAC)2) for 30-60 min at 4 °C. All assays were repeated at least three times. For complex formation, the indicated components were incubated in ribosome assay buffer (30 mM HEPES-KOH pH 7.4, 100 mM KOAc, 2 mM Mg(OAc)2) at 37 °C for 15 minutes. Unless noted, NSP1 with a C-terminal ybbR tag conjugated to a Cy5 dye was used in all gel shift experiments, as C-terminally tagged NSP1 from SARS-CoV was functional in cellular assays12 and Cy5 provides cleaner signal upon image acquisition with a Typhoon imager. For competition experiments, the competitor protein was pre-incubated with ribosomal subunits at 37 °C for 15 minutes, prior to addition of the labeled protein.
Purification of human eIFs
eIF1 and eIF1A
Plasmids for the expression of recombinant human eIF1 and eIF1A were gifted by Christopher S. Fraser (UC Davis). Detailed cloning and recombinant protein expression and purification methods have been described previously55. Changes made to the eIF1 and eIF1A purification scheme include use of a HiTrap SP-HP (Cytiva Lifesciences, #17115201) column for the IEX chromatography step and the addition of a SEC step in which fractions eluted from the IEX column containing either eIF1 or eIF1A, as determined by SDS-PAGE analysis, were pooled and passed over a Superdex 75 10/300 (Cytiva, #17-5174-01) column to remove protein oligomers/aggregates and exchange purified factors into a storage buffer (20 mM Hepes, pH 7.5, 250 mM KOAc, 1 mM dithiothreitol (DTT), and 10% glycerol (v/v)). Purified factors were concentrated to ∼ 200 μM using a Millipore centrifugal filter and stored at −80°C.
eIF2 and eIF3
Endogenous initiation factor complexes were purified from 400 mL of HeLa post-nuclear extract that was kindly donated to our group by Robert Tijan (UC Berkeley) using previously detailed methods55,64. Additional guidance was received from Christopher S. Fraser (UC Davis).
eIF5
The expression and purification of recombinant eIF5 was carried out using a using a modified protocol based on previously established methods65. Alterations to the protocol included substitution of the MonoQ column with a HiTrap Q HP column (Cytiva Lifesciences, #29051325) for the anion exchange chromatography step, and adding a SEC step using a Superdex 200 10/300 (Cytiva Lifesciences, 28-9909-44) column to exchange purified eIF5 in to a storage buffer containing 20 mM Hepes, pH 7.5, 250 mM KOAc, 1 mM DTT, and 10% glycerol (v/v).
eIF3j
Human eIF3j was expressed and purified as done for NSP1, with the following changes. Human eIF3j was codon optimized for expression in E. coli and cloned with an N-terminal 6xHis tag and TEV protease cleavage site into a pET28b vector at the NcoI site. Protein was expressed by induction with 0.5 mM IPTG overnight at 17 °C. Following cleavage of the 6xHis tag and the subtractive Ni-NTA step, eIF3j was subjected to a final purification step using SEC on a 23 mL Superdex 75 10/300 (Cytiva, #17-5174-01) column equilibrated in storage buffer (20 mM HEPES-KOH pH 7.5, 150 mM KOAc, 10% (v/v) glycerol, and 1 mM DTT). Fractions with purified eIF3j were identified via SDS-PAGE, concentrated, flash frozen, and stored at −80 °C.
tRNAi
Human tRNAi was transcribed from a DNA template with a 5’-end T7 promoter and hammer head ribozyme55. The plasmid was linearized via digestion with BstNI for transcriptional termination at the 3’-CCA of tRNAi. Purified and linearized plasmid was used as the template for in vitro transcription with T7 polymerase at 37 °C for 4 hrs at 16 mM MgCl2, during which the ribozyme self-cleaved (>80% efficiency). Mature tRNAi was separated from pre-cursor RNA and cleaved ribozyme via 10% acrylamide gel electrophoresis in the presence of 8M urea. After excision of the tRNAi band, the RNA was extracted 3x at room temperature for 12 hrs with 300 M ammonium acetate, and ethanol precipitated. tRNAi was resuspended in 10 mM NaCl,10 mM Bis-tris, pH 7.0 and stored at −80°C. To charge tRNAi with methionine, 100 µL of 60μM human tRNAi transcript was mixed with 618.6 µL ddH2O, and heated at 95 °C for 2 min, then immediately chilled on ice for 5 min. The resulting solution was mixed with 200 µL of 5x charging buffer (200 mM Tris-HCl pH7.5, 50 mM Mg(OAc)2, 5 mM DTT), 20 µL 100 mM ATP, 30 µL 10 mM L-methionine, 2 µL 1 M Mg(OAc)2 and 29.4 µL 34 µM yeast MetRS66, and incubated at 30 °C for 45 min. The charging reaction was stopped by addition of 100 µL 3M NaOAc pH5.2. The resulting tRNA was purified by phenol-chloroform-isoamyl alcohol (25:24:1, pH 5.2) extraction and ethanol precipitation. The pellet was resuspended with tRNA storage buffer (10 mM NaOAc pH 5.2, 50 mM Mg(OAc)2) and further purified by passing through BioRad P-6 columns that were equilibrated with tRNA storage buffer. The charging efficiency was ≈70% based on acid urea PAGE analyses67 of the final tRNA product.
Real-time single-molecule assays using ZMWs
ZMW-based imaging
All real-time imaging was conducted using a modified Pacific Biosciences RSII DNA sequencer, which was described previously30. Unless otherwise noted, Cy3 dyes were excited using the 532 nm excitation laser at 0.6 µW/µm2. Cy5 and Cy5.5 dyes were excited with the 642 nm laser at 0.1 µW/µm2. In nearly all experiments, data were collected at 10 frames per second. The exception was when the 532 nm laser was used at 0.16 µW/µm2; in this case, data were collected at 3 fps to increase signal to noise ratios. ZMW chips were purchased from Pacific Biosciences. Prior to imaging, all ZMW chips were washed with 0.2% Tween-20 and TP50 buffer (50 mM Tris-OAc pH 7.5, 100 mM KCl). Washed chips were coated with neutravidin by a 5 min incubation with 1 µM neutravidin diluted in TP50 buffer supplemented with 1.3 µM of DNA blocking oligos and 0.7 mg mL-1 UltraPure BSA. Chips were washed with TP50 and ribosome assay buffer (30 mM HEPES-KOH pH 7.4, 100 mM KOAc, 2 mM Mg(OAc)2). For imaging, the ribosome assay buffer was supplemented with casein (62.5 µg mL-1) and an oxygen scavenging system68: 2 mM TSY, 2 mM protocatechuic acid (PCA), and 0.06 U/µL protocatechuate-3,4-dioxygenase (PCD). In all real-time single-molecule experiments, fluorescently-labeled NSP1 with an N-terminal ybbR tag was used, since it had translation inhibition and 40S-binding activities similar to the wild-type protein.
Tethered eIF1–40S and eIF1–eIF3j–40S complexes (Figure 2)
Prior to tethering, 200 nM biotin-40S subunits were incubated in ribosome assay buffer with 6 µM eIF1 at 37 °C for 20 min. For tethering, 1 nM of pre-formed complexes (by biotin-40S) were incubated on the neutravidin-coated ZMW surface for 15 min at room temperature. Non-tethered complexes were washed away with ribosome assay buffer prior to imaging. Upon start of data acquisition and in the presence of 1 µM eIF1, Cy3-NSP1 was added at the indicated concentrations and temperatures. In experiments with eIF3j, 2.5 µM eIF3j was pre-incubated for ∼5 min with tethered ribosomal complexes prior to addition of ybbR-NSP1 labeled with Cy3 dye (‘Cy3-NSP1’). Temperatures and laser powers were as indicated.
Tethered ribosomal pre-initiation complexes (Figure 2)
Complexes were formed by incubating 40 nM biotin-40S subunits in ribosome assay buffer with the indicated eIFs at 37 °C for 15 min. eIF concentrations were: 1 µM for eIF1, eIF1A, & eIF5; and 200 nM for TC-GMPPNP & eIF3. Tethering was conducted as above. As indicated, imaging was conducted in the presence of: 1 µM eIF1, eIF1A, & eIF5; 100 nM TC-GMPPNP; and 50 nM eIF3. Upon start of data acquisition at 30 °C, Cy3-NSP1 was added at 25 nM final concentration.
Tethered 40S–HCV IRES complexes (Figure 3)
The indicated HCV IRES RNAs were diluted to 200 nM in refolding buffer (20 mM cacodylate-NaOH pH 7.0, 100 mM KCl, and 1 mM EDTA), heated to 95 °C for 2 min, and slow cooled to room temperature (∼45 min). Once cooled, 4 mM Mg(OAc)2 was added to quench the EDTA, and RNAs were stored on ice until use. To form complexes, 75 nM of 40S-RACK1-Cy5 subunits were incubated in ribosome assay buffer with 20 nM of the indicated IRES RNAs at 37 °C for 20 min. As indicated (‘+eIFs’), 1 µM eIF1, 1 µM eIF1A, 1 µM eIF5, 300 nM TC-GMPPNP, and 240 nM eIF3 were included. Tethering was performed as above, except 1 nM of complex was incubated for 5 min at room temperature. Upon start of data acquisition at 30 °C, Cy3-NSP1 was added at 25 nM final concentration. As indicated (+eIFs), data acquisition was performed in the presence of 1 µM eIF1, 1 µM eIF1A, 1 µM eIF5, 200 nM TC-GMPPNP, and 150 nM eIF3.
Tethered M+41–40S subunit complexes (Figure 3)
The biotinylated model mRNA (M+41) and Cy5-labeled model mRNA (M+41-Cy5) were described previously38. To form complexes with M+41, 40 nM 40S-RACK1-Cy5 subunits were incubated in ribosome assay buffer with 500 nM M+41, 1 µM eIF1, 1 µM eIF1A, 1µM eIF5, and 400 nM TC-GMPPNP at 37 °C for 20 min. Tethering was performed as above, except 1 nM of complex was incubated for 10 min at room temperature. Upon start of data acquisition at 30 °C, Cy3-NSP1 was added at 25 nM final concentration in the presence of 1 µM eIF1, 1 µM eIF1A, 1 µM eIF5, and 200 nM TC-GMPPNP. For M+41-Cy5, conditions were identical except TC-GMPPNP was present at 200 nM and 100 nM during complex formation and imaging, respectively; and, eIF3 was included at 200 nM and 50 nM. eIF3 was included in these experiments to promote complex formation, which was inefficient.
Tethered 40S–CrPV and 80S–CrPV complexes (Figure 4)
CrPV IRES RNAs were refolded as with the HCV IRES. To form 40S–CrPV IRES complexes, 40S-uS19-Cy3 subunits at 75 nM were incubated in ribosome assay buffer with 20 nM of the indicated RNA at 37 °C for 15 min. To form 80S–CrPV IRES complexes, 75 nM 40S-uS19-Cy3, 150 nM 60S-uL18-Cy5, and 20 nM RNA were incubated. Tethering was performed as with the 40S– HCV complexes. Upon start of data acquisition at 30 °C, Cy5.5-NSP1 was added at 25 nM final concentration.
Addition of NSP1–40S subunit complexes (Figure 5)
HCV and CrPV IRES RNAs were refolded as above. To form NSP1–40S subunit complexes, 550 nM Cy5.5-NSP1 was incubated in ribosome assay buffer with 275 nM 40S-uS19-Cy3 subunits at 37 °C for 15 min. To tether biotinylated RNAs, 0.16 nM of refolded IRES was incubated on the neutravidin-coated ZMW surface for 5 min. Upon start of data acquisition at 30 °C, either 15 nM of 40S-Cy3 subunits alone or 15 nM of Cy5.5-NSP1–40S-Cy3 pre-formed complexes (final concentrations) were added.
Data analysis
Experimental movies that captured fluorescence intensities over time were processed using custom MATLAB scripts, as described previously29,30. Briefly, ZMWs with the desired fluorescence signals were identified by filtering for the signals at appropriate time points. Individual binding events were assigned manually based on the appearance and disappearance of the respective fluorescence signals. Unless noted, only the first NSP1 binding event longer than 4-5 s that occurred within the first 500 s of imaging was analyzed. Unless intractable due to complex stability or inhibition, approximately 200-300 single molecules were analyzed, indicated by single-step photobleach events. Association times were defined as the time elapsed from the addition of the labeled component until a burst of fluorescence for that component. The time of addition is controlled by the instrument and varies from experiment to experiment, but typically occurs within the first 10 s of data acquisition. Lifetimes were defined as the duration of the corresponding fluorescence signal.
Kinetic parameters were extracted by fitting observed data to single- or double-exponential functions as described30. On some complexes, the presence of a large, slow association phase made it difficult to derive reliable rates, as amplitudes for the association rates are assigned semi-arbitrarily during the fits. When this occurred, comparisons of median association times were used instead, which better reflected the raw data. In the results section, this is indicated by ‘median association times’, which only pertains to the indicated final concentration of NSP1. When possible, observed association rates derived from fits to exponential functions were reported, which are indicated by ‘apparent kon-1’ and were converted to times to facilitate comparisons across experiments where medians are reported. Regardless, all derived association rates, median association times, lifetimes, and the number of molecules examined are reported in Supplementary Tables 2-5. As indicated in the tables, fits to linear functions were used to estimate very slow association rates observed when NSP1 association was inhibited.
Statistical analyses
To calculate errors for NSP1 binding efficiency (e.g., Figure 3E), bootstrap analyses (n = 10,000) were performed to calculate 99% confidence intervals (C.I.) for the observed proportions using R. To calculate errors for median association times and lifetimes, bootstrap analyses (n = 10,000) were performed to calculate 95% C.I. of the observed median using MATLAB. Reported errors for derived rates represent 95% C.I. yielded from fits to linear, single-exponential, or double-exponential functions, as indicated.
Equilibrium single-molecule analyses using total internal reflection fluorescence microscopy (TIRFM)
Our home-built, prism-based TIRFM instrument has been described previously68–70. All data were collected at room temperature at 10 frames per second with the EM gain set to 650. To form NSP1–40S subunit complexes, 40 nM biotin-40S subunits were incubated in ribosome assay buffer with 500 nM NSP1-ybbR-Cy5 and 6 µM eIF1 at 37 °C for 15 min. Pre-formed complexes were tethered to the surface of a neutravidin-coated quartz slide as with ZMWs above. Buffer conditions during TIRFM imaging were identical to those with ZMWs. During imaging via excitation with the 647 nm laser, 1 µM eIF1 and 4 nM NSP1-Cy5 (C-terminal ybbR tag) were present. To form 80S–IRES complexes, 100 nM of refolded CrPV IRES60 biotinylated on the 3’-terminus was incubated in ribosome assay buffer with 250 nM 40S-uS19-Cy3 and 750 nM 60S-uL18-Cy5 subunits. The complex was tethered as above at 1 nM. Emission data were collected in both the Cy3 (donor) and Cy5 (acceptor) channels following excitation of the Cy3 donor dye with the 532 nm laser. Co-localized molecules were identified and analyzed using custom MATLAB scripts. FRET events were assigned manually to a single state, with single molecules indicated by single-step photobleach events of the donor fluorophore. Background fluorescence intensities were corrected and normalized, and the gamma-corrected FRET efficiency distribution was calculated as described71.
Ribosome purifications and labeling
40S and 60S ribosomal subunits were purified from the indicated cell lines and labeled with biotin or dyes as described29. To specifically install biotin on the ribosomal subunit, purified 40S-RACK1-ybbR subunits were incubated with 4-fold molar excess PEG11-biotin (ThermoFisher #21911) conjugated to co-enzyme A (CoA) and 2-fold molar excess Sfp synthase enzyme. To install fluorophores, purified 40S-RACK1-ybbR, 40S-uS19-ybbR, and 60S-uL18-ybbR subunits were incubated with 4-fold molar excess Cy3 or Cy5 dyes conjugated to CoA and 2-fold molar excess of Sfp synthase. All reactions proceeded for 90 min at 37 °C. To remove free biotin, dyes, and/or Sfp synthase, reactions were layered onto a low-salt sucrose cushion (30 mM HEPES-KOH pH 7.5, 100 mM KOAc, 0.5 M sucrose, 5 mM Mg(OAc)2, and 2 mM DTT), and centrifuged at 287,582 x g (90,000 rpm) for 1 hr at 4 °C in a TLA100.2 rotor (Beckman, ref. 362046) in 11×34 mm thick-wall polycarbonate ultracentrifuge tubes (Beckman, ref. 343788). Ribosome pellets were washed once and subsequently resuspended with ribosome storage buffer (30 mM HEPES-KOH pH 7.4, 100 mM KOAc, 5 mM Mg(OAc)2, 6% (v/v) sucrose, and 2 mM DTT), aliquoted, flash frozen in liquid N2, and stored at −80 °C.
RNA in vitro transcriptions
HCV & CrPV IRES RNAs
Purified PCR products were used as the templates for in vitro transcription with the T7 MEGAScript T7 Transcription kit (ThermoFisher #AMB13345). Standard reaction conditions were used, except each nucleotide was present at a final concentration of 4 mM and the reaction was supplemented with 6 mM of 5’-biotin-G-Monophosphate (Trilink, #N-6003). Reactions were incubated at 37 °C for 2 hours. Transcribed RNAs were purified using the GeneJET RNA Purification Kit (ThermoFisher, #K0732).
Reporter mRNAs for IVT assays
Linearized plasmids were used as templates for in vitro transcription with the MessageMAX T7 ARCA-Capped Message Transcription Kit (Cell Script, # CMMA60710). Transcription reactions were setup using standard conditions and then incubated at 37°C for 30 min. Transcripts were purified using the MEGAclear Transcription Clean-up Kit (ThermoFisher, AM1908). The integrity and homogeneity of the reporter mRNAs were verified via native- and denaturing-PAGE.
CRISPR-Cas9 & Homology-directed repair
Sequences
To generate the 40S-uS19-ybbR and 60S-uL18-ybbR cell lines, the guide sequences CGCAACACTCACCATCTTGC (uS19/RPS15) and AAAAATCATAGAAAATTGCT (uL18/RPL5) were cloned into the pX458 vector backbone using the published approach72. To insert the tandem ybbR and flag tags onto the endogenous copies of the genes, single-stranded DNA ultramer repair templates were purchased from Integrated DNA Technologies that contained about 40-60 nts of flanking sequence on either side of the desired insertion. See Supplementary Table 6 for all guide oligo, repair template, and PCR screening oligo sequences.
Transfections, sorting, & screening overview
Approximately 24 h post seeding in a well of a 6-well plate, low-confluency (∼30%) wild-type HEK293T cells were transiently transfected (Liopfectamine 3000, ThermoFisher) with 1 µg of the relevant pX458 plasmid and 2 µg of ssDNA repair template. Cells were allowed to recover for 48 hrs. Single, eGFP-positive cells were sorted at the Stanford Shared FACS Facility into a well of a 96-well plate that contained 50% conditioned DMEM(high glucose) medium. Individual colonies were allowed to recover until they were visible by eye (approximately 1.5-2 weeks), upon which colonies were transferred to a well of a 24-well plate. Once confluent, colonies were screened via PCR, Sanger sequencing, and Western blot analyses. Successfully edited cell lines were expanded and ultimately stored as stocks in 5% DMSO-FBS solution.
PCR screening & Sanger sequencing
Genomic DNA was isolated from candidate cell lines using the QuickExtract DNA Extraction Solution (Lucigen, #QE09050) essentially as described, except 10 µL of resuspended cells were added to 50 µL of QuickExtract solution. Following extraction, 1 µL of gDNA was added to a standard 24 µL 2X GoTaq Green PCR (Promega, #M7123), typically ∼35 cycles. Sequences for uS19 and uL18 screening oligos are available in Supplementary Table 6. Following amplification, PCRs were analyzed using 2% agarose gel electrophoresis. PCR products with the desired insertion size were submitted for Sanger sequencing, following their purification.
Western blots
Whole cell lysates were analyzed via 4-20% SDS-PAGE followed by transfer to PVDF membranes. All antibodies were diluted in 5% Blotting-Grade Blocker (BioRad, #1706404) in TBST buffer. The following primary antibodies were used: anti-RPS15/uS19 at 1:500 from Abcam (ab168361); and, anti-RPL5 at 1:500 from GeneTex (GTX101821). Incubations with primary antibodies were for > 12 h at 4 °C. The secondary antibody was GeneTex Rabbit IgG antibody (HRP) (GTX213110-01) and was used at 1:10,000. Blots were imaged via chemiluminescence.
Polysome profiling assays
Polysome profiling assays were performed as described29 and in duplicate for the indicated cell lines.
Cell lines and growth
HAP1 cells that express RACK1-yybR were described previously29. HAP1 cells were grown at 37 °C with 5% CO2 in Iscove’s Modified Dulbecco’s medium (IMDM) (Gibco, #sh30228) supplemented with 10% (v/v) fetal bovine serum, 2 mM L-glutamine (Gibco), and 1X penicillin/streptomycin (Gibco). Wild-type HEK293T cells were purchased from ATCC (CRL-3216) and engineered HEK293T cells are described here. All HEK293T cell lines were grown as HAP1 cells except Dulbecco’s Modified Eagle’s Medium (DMEM) High Glucose medium (Gibco, #11965092) was used instead. Cell lines were not tested for mycobacterium.
Data sharing plan
All plasmids, data, and custom analysis scripts are available upon reasonable request.
SUPPLEMENTARY FIGURES
ACKNOWLEDGEMENTS
We are grateful to Michael Lawson, Jan Carette, and other members of the Puglisi and Carette labs for helpful guidance, discussions, and feedback. We also appreciate discussions with and advice from Chris Fraser and Masa Sokabe. We thank Peter Sarnow and the Sarnow lab for sharing cell culture equipment, and the Stanford Shared FACS Facility for completion of all single-cell sorting. C.P.L. is a Damon Runyon Fellow supported by the Damon Runyon Cancer Research Foundation (DRG-#2321-18); A.G.J. was supported by a National Science Foundation Graduate Research Fellowship (DGE-114747); and J.W. was supported by a postdoctoral scholarship from the Knut and Alice Wallenberg Foundation (KAW 2015.0406). Research on eukaryotic translation in the laboratory of J.D.P. is funded by the National Institutes of Health (GM011378, AI047365, and AG064690).