key: cord-0848415-zq5cdd5z
authors: Yoo, Tae Yeon; Mitchison, Timothy
title: Quantification of nuclear transport inhibition by SARS-CoV-2 ORF6 using a broadly applicable live-cell dose-response pipeline
date: 2021-12-13
journal: bioRxiv
DOI: 10.1101/2021.12.10.472151
sha: 37ed75f3860a5c2117ba4257b192faf16fbb7ac9
doc_id: 848415
cord_uid: zq5cdd5z

SARS coronavirus ORF6 inhibits the classical nuclear import pathway to antagonize host antiviral responses. Several models were proposed to explain its inhibitory function, but quantitative measurement is needed for model evaluation and refinement. We report a broadly applicable live-cell method for calibrated dose-response characterization of the nuclear transport alteration by a protein of interest. Using this method, we found that SARS-CoV-2 ORF6 is ∼5 times more potent than SARS-CoV-1 ORF6 in inhibiting bidirectional nuclear transport, due to differences in the NUP98-binding C-terminal region that is required for the inhibition. The N-terminal region was also required, but its membrane binding function was dispensable, since loss of the inhibitory function due to N-terminal truncation was rescued by forced oligomerization using a soluble construct. Based on these data, we propose that the hydrophobic N-terminal region drives oligomerization of ORF6 to multivalently cross-link the FG domains of NUP98 at the nuclear pore complex.

1 Macromolecular transport across the nuclear envelope occurs through nuclear pore complexes (NPCs) and 2 is tightly regulated in eukaryotic systems. The transport channel of the NPC is densely filled with 3 intrinsically disordered regions of the FG-nucleoporins (FG-NUPs) enriched in phenylalanine-glycine 4 repeats 1-3 . The FG-repeat regions constitute a permeability barrier that specifically interacts with soluble 5 transport carrier proteins, including importin-beta family proteins, to selectively facilitate their diffusion 6 through the pore along with cargoes bound to them 4-7 . The classical nuclear import pathway involves 7 importin beta-1 (KPNB1) as a transport carrier and a class of adaptor proteins called importin alphas 8 (KPNAs) which bind classical nuclear localization signals (NLSs) in complex with KPNB1 8 . The classical 9

nuclear export pathway is mediated by exportin-1 (CRM1/XPO1), another transport carrier in the 10 importin-beta family, which directly binds to leucine-rich classical nuclear export signals 9 . Directionality 11 of cargo transport is provided by compartment-specific assembly and disassembly of cargo-carrier 12 complexes, which is mediated by the small GTPase RAN [10] [11] [12] . 13 14 Nucleocytoplasmic transport is required for several innate immune anti-viral pathways; for example, the 15 interferon (IFN) response requires nuclear import of phosphorylated STAT1 13 . SARS coronaviruses, like 16 several other viruses, were reported to inhibit the nucleocytoplasmic transport system of host cells during 17 infection, resulting in reduced innate immune signaling [14] [15] [16] [17] [18] [19] . A small protein that is unique to these viruses, 18

Open Reading Frame 6 (ORF6), was shown to inhibit the KPNB1-mediated classical nuclear import 19 pathway 20 . Several inhibition mechanisms were proposed. Early studies on SARS-CoV-1 proposed that 20 its ORF6 tethers KPNA2 to membranes of the endoplasmic reticulum (ER) and Golgi apparatus to 21 sequester KPNB1 21 . Interaction of SARS-CoV-2 ORF6 with KPNA2 was also demonstated 16 Full understanding of how ORF6 inhibits nucleocytoplasmic transport requires quantitative measurement 6 of the dose response function, i.e., the relationship between the intracellular expression level of ORF6 and 7 the resulting reduction in the nuclear transport efficiency. Previous studies measured neither of the 8 quantities but instead assessed the average effect of ORF6 overexpression on the steady-state localization 9

of IFN-related transcription factors or artificial cargoes or on the downstream IFN signaling activities. 10

Accurate dose-response characterization would also enable objective comparison of ORF6 activity 11 between viral species and measurement of perturbation effects, for example the influence of variations in 12 ORF6 sequence on its inhibitory function. Such comparison may contribute to understanding of different 13 characteristics between SARS-CoV-1 and SARS-CoV-2 and the prediction of the innate immune response 14

to new SARS-CoV-2 variants which may occur in the future. 15

One reason that inhibition of nuclear transport by ORF6 was not accurately quantified in previous studies 17 was the lack of suitable assays. We previously developed optogenetic assays for quantification of nuclear 18 import and export in living cells 25 . Here, we apply them to measure inhibition by ORF6. We developed a 19 simple measurement pipeline for calibrated quantification of the expression level of an untagged protein 20 in single cells and applied it to measure ORF6 concentrations. In typical studies where results are averaged 21 across many cells, variation in protein expression levels between cells is a technical problem. Conclusions 22 based on bulk averages may not predict the behavior of individual cells whose expression level differs 23 U2OS cells stably expressing the optogenetic nuclear import or export probe were transfected with a 1 plasmid encoding GFP-2A-POI. Due to the self-cleaving activity of the 2A sequence, which occurs 2 during translation 29 , GFP and POI are produced at the same rate as separate polypeptides. This enables 3 the estimation of the POI concentration based on the GFP fluorescence intensity without the use of 4 direct tagging which might disrupt the function of the POI. The GFP intensity and the transport kinetics 5 were repeatedly measured in hundreds of individual cells at 3-6-hour intervals after transfection. The 6 GFP intensity was converted to absolute GFP concentration via a novel calibration procedure, described 7

below. The resulting data shows a wide range of GFP concentration, typically ranging from 0 to 30 µM, 8 due to the stochasticity in the transient transfection as well as the time-dependent increase in the 9 expression level. Plotting the nuclear transport rate (import or export) against the GFP concentration 10 revealed the dose-response curve. To generate parameters from these curves, we fit them to a Hill 11 function. As in pharmacokinetic-pharmacodynamic modeling 30 , this fit should be considered descriptive 12 rather than mechanistic, but it allowed quantitative comparisons across all the data. 13

The GFP calibration is performed as follows (Fig. 1b, Supplementary Fig. 1 ): (Step 1) GFP expressing 15 cells are imaged using the same microscope and image acquisition setting as in the dose-response 16 characterization. The GFP intensity in each cell is quantified in the same way as in the dose-response 17 characterization. ( Step 2) Recombinant GFP is added to the culture at a certain concentration, and the 18 same cells in the Step 1 are imaged again in the presence of GFP surrounding them. Cells would be 19 brighter or darker than the background, depending on the relative concentration of the intracellular to the 20 extracellular GFP concentration. ( Step 3) Plotting the contrast between the intracellular and the 21 background GFP intensity against the GFP intensity measured in Step 1 shows a linear relationship. A 22

linear model is fitted to the relationship to determine the x-intercept, which corresponds to the GFP 23 intensity of a cell whose intracellular GFP concentration is the same as the concentration of the 1 recombinant GFP added to the culture in Step 2. (Step 4) Steps 1 to 3 are repeated with different 2 concentrations of added GFP to obtain the relationship between the GFP intensity and concentration. 3

The slope of the linear model fitted to the relationship serves as the intensity-to-concentration 4 conversion factor. Importantly, this method reports GFP concentration values independent of the optical 5 parameters of the microscope such as the depth of field or illumination brightness (Supplementary Text). 6 7

We tested the dose-response pipeline with negative controls ( Supplementary Fig. 2 ). When only GFP 8 was expressed, the nuclear import rate was constant across the GFP concentration, indicating the 9 absence of artifacts due to GFP overexpression. ORF10 and Candidate ORF15 are gene products of 10 SARS-CoV-2 that have been reported to not affect the IFN signaling and therefore are not likely to alter 11 the nuclear import 16,31 . As expected, these proteins showed flat dose-response curves, confirming that 12 co-expression of GFP and an inactive POI does not perturb transport. 13 14

We applied the dose-response pipeline to compare the inhibitory effects of SARS-CoV-1 and SARS-16

CoV-2 ORF6s (denoted by ORF6 CoV1 and ORF6 CoV2 , respectively) on nuclear transport. The amino acid 17 sequences of ORF6 CoV1 and ORF6 CoV2 are 69% identical; most of the variations lie in the C-terminal 18 half, including the 2 amino acid C-terminal extension of ORF6 CoV1 (Fig. 2a) . We acquired dose-response 19 curves of ORF6 CoV1 and ORF6 CoV2 on the nuclear import and export rates. The Hill function fit well to 20 all the dose-response curves, providing parameters that reflect the amplitude of inhibition (A), half 21 maximal inhibitory concentration (IC50), and the steepness of the dose-response curve (nH or Hill 22 coefficient) (Fig. 2b) . A and IC50 are measures of the efficacy and potency of ORF6 as a nuclear 23 COVID-19 patients in Italy 34 (Fig. 4a) . Together with the comparative mutational analysis of ORF6 CoV1 1 and ORF6 CoV2 (Fig. 3) , these results suggest that the C-terminal end is crucial to the inhibitory function 2 of ORF6. We also tested a point mutation, M58R, which has been shown to abolish the ability of 3 ORF6 CoV2 to bind NUP98-RAE1 while maintaining its KPNA binding 18 . This point mutation also 4 resulted in the complete loss of the inhibitory function of ORF6 CoV2 , suggesting that NUP98-RAE1 5

binding of the C-terminal region is essential while KPNA binding is not (Fig. 4a) . 6 7

We also evaluated the influence of mutations in the N-terminal and middle regions on the inhibitory 8 function of ORF6 CoV2 (Fig. 4b) . The peptide of C-terminal residues 43-61 has been previously shown to 9

interact with NUP98-RAE1 complex 18 . We found that the peptide of residues 38-61 did not show a 10 significant inhibitory effect on the nuclear import (Fig. 4b) , indicating that NUP98-RAE1 binding alone 11

is not sufficient for the transport inhibition. Alternative translation of ORF6 CoV2 gene results in the lack 12 of the first 18 residues 35 , which we found to reduce the potency of ORF6 CoV2 by a factor of 3.4 (IC50 = 13 4.57 ± 0.76 µM) (Fig. 4b) . A previous study discovered ORF6 having a 9-residue deletion from the 14 central part (residues 22-30) in a SARS-CoV-2 strain passaged in vitro in the IFN-deficient Vero E6 15 cell 36 . Although the authors predicted that the deletion would dramatically alter the structure of ORF6 16 binding to the membrane, the deletion did not affect the dose-response characteristics of the nuclear 17 import inhibition (Fig. 4b) . This is consistent with K23R/V24I/S25A and Y31V mutations (denoted by 18 #2 and #3) not affecting the dose-response characteristics ( Fig. 3 and Supplementary Fig. 4) . Taken 19 together, these mutational analyses suggest that the N-terminal and C-terminal regions are 20 simultaneously required for the ORF6 activity in the nuclear transport inhibition, while the central 21

We next sought to examine the localization of ORF6 using small epitope tags and immunofluorescence. 2 Direct tagging of ORF6 with an epitope tag or fluorescent protein has been performed in previous 3 studies to examine the subcellular localization or interactions with other proteins 16,18,20-24 , but the effect 4 of the tagging on the inhibitory function has not been evaluated. We characterized the dose responses of 5 ORF6 CoV2 N-or C-terminally tagged with three different epitope tags: ALFA 37 , Flag, and HA. All three 6 epitope tags attenuated the potency of ORF6 CoV2 when positioned at the C-terminus (Fig. 5a , 7

Supplementary Fig. 5 ). This is consistent with the observation that the ORF6 CoV2 potency was sensitive 8

to the C-terminal modifications ( Fig. 3 and Fig. 4a ). On the other hand, at the N-terminus of ORF6, 9

ALFA and HA tags had negligible influence on the dose-response characteristics, while Flag tag rather 10 increased the potency (Fig. 5a, Supplementary Fig. 5 ). These results suggest that the tagging location 11 and the type of epitope tag should be carefully selected when studying small proteins and must be 12 explicitly reported in publications. 13

We chose to use ALFA tag for subsequent immunofluorescence analyses since there is a well-15 characterized fluorophore-conjugated anti-ALFA single-domain antibody ("nanobody") commercially 16

for 24 hours. Prior to paraformaldehyde fixation, GFP concentration was measured to estimate the co-18 expressed ORF6 level as in the nuclear transport dose-response analysis (Fig 1b) . The fixed cells were 19 permeabilized with digitonin to preserve the intracellular membrane structure and then immunostained 20 using the anti-ALFA nanobody and anti-NUP98 antibody. Interestingly, the N-and C-terminally tagged 21 ORF6 CoV2 showed distinct staining patterns (Fig. 5b) . ALFA-ORF6 CoV2 showed strong nuclear pore 22

staining with negligible staining elsewhere throughout different expression levels. On the other hand, 23 ORF6 CoV2 -ALFA was stained at the cytoplasmic membranes and the nuclear envelope so strongly that 1 staining of individual nuclear pores was barely discernable. At high ORF6 concentrations, we found that 2 NUP98 was dislocated from the NPC, consistent with a previous study 24 (Fig. 5b) . However, the NUP98 3 dislocation was not noticeable at ORF6 concentrations that are low yet high enough to inhibit the 4 nuclear transport, so it is unlikely to be directly involved in the nuclear transport inhibition. 5 6

The difference in the ALFA immunostaining patterns (Fig. 5b) could result from the membrane binding 7 of the N-terminal region being disrupted by the proximal N-terminal ALFA tag but not by the distal C-8 terminal tag. If this is the case, the lack of effect of the N-terminal ALFA tag on the dose-response 9 characteristics ( Fig. 5a ) would suggest that membrane binding is not required for the inhibitory function 10 of ORF6 CoV2 . However, the difference in the ALFA immunostaining patterns could also result from the 11 membrane binding of the N-terminal region restricting the N-terminal ALFA tag from binding the anti-12 ALFA nanobody. To test this possibility, we examined the immunofluorescence localization of ORF6 13

having ALFA tags at both the N-and C-termini (ALFA-ORF6 CoV2 -ALFA) (Fig. 5c) . We found that this 14 construct showed strong ALFA immunostaining at the nuclear pores, similar to ALFA-ORF6 CoV2 . If the 15 absence of the immunostained ALFA-ORF6 CoV2 at the membranes was due to epitope inaccessibility 16 rather than hindrance of membrane binding, ALFA-ORF6 CoV2 -ALFA would have shown strong ALFA 17 immunostaining at the membranes like ORF6 CoV2 -ALFA did. Thus, the N-terminal ALFA tag inhibits 18 the membrane binding of ORF6 CoV2 but does not alter its inhibitory effect on the nuclear transport. This 19

indicates that the role of the N-terminal region in the nuclear transport inhibition is not the membrane 20 binding but something else. 21

Forced homo-oligomerization rescues the inhibitory function of N-terminally truncated ORF6 23 We next sought to investigate alternative roles of the N-terminal region of ORF6 in the transport 1 inhibition. Several other proteins have also been reported to bind FG-NUPs and inhibit the carrier-2 mediated nuclear transport. Such proteins include wheat germ agglutinin 38 (WGA), hyperphosphorylated 3 tau 39 , C9orf72 dipeptide repeats 40 , and vesicular stomatitis virus matrix (VSV M) protein 41,42 . A 4 common feature of these proteins is the ability to form protein aggregates or homo-oligomers which are 5 presumably able to multivalently cross-link FG domains 43-45 . For example, WGA forms a dimer with 6 eight binding sites for O-GlcNAc, a glycosylation abundant in the FG domains 46,47 . A recent 7 computational analysis predicted that ORF6 is also aggregate-prone due to the hydrophobic N-terminal 8 region 48 . Therefore, we hypothesized that the mechanistic role of the N-terminal region of ORF6 in the 9 nuclear transport inhibition is to mediate aggregation or homo-oligomerization, while its membrane 10 binding is rather coincidental. 11

To test this hypothesis, we investigated whether the N-terminal region can be functionally replaced with 13 an oligomerization domain derived from other proteins that do not bind membranes. We first used a 14 chemically inducible oligomerization system based on the F36V mutant of FK506-binding protein 12 15 (FKBP) and the B/B Homodimerizer ligand (equivalent to AP20187) 49 (Fig. 6a) . Addition of the 16 homodimerizer did not affect the dose-response characteristics of the full-length ORF6 CoV2 or the N-17 terminal truncation of ORF6 CoV2 (residues 38-61; CT), confirming the absence of unspecific effects of 18 the homodimerizer on the dose response (Fig. 6a) . When the CT was fused to FKBP, addition of the 19 homodimerizer partially recovered the inhibitory function of the CT (Fig. 6a) . The homodimerizer-20 dependent recovery further increased when the CT was fused to two FKBPs in tandem, which 21

presumably induce higher-order oligomerization (Fig. 6a) . We also tested the tetramerization domain of 22 p53 (residues 326-356; p53TD) 50,51 , which is far smaller than the tandem FKBP (MW ~4 kDa vs ~24 23 kDa) and comparable to the truncated N-terminal region in size. We found that fusing p53TD also 1 rescues the inhibitory function of the CT (Fig. 6b) . We confirmed that the CT and the fusion proteins 2 were localized at the nuclear pores without showing noticeable membrane localization (Fig. 6c, d, e) . 3

The recovery of the inhibitory function via forced homo-oligomerization suggests that the mechanistic 4 role of the N-terminal region in the nuclear transport inhibition may be to drive oligomerization of 5 ORF6 and that membrane binding is dispensable (Fig. 6f) . 6 7 Having multiple NUP98-binding C-terminal regions, the ORF6 oligomer would have a high avidity for 8 the multiple copies of NUP98 clustered at the NPC. The interaction with KPNAs, on the other hand, 9

would not be strengthened by homo-oligomerization because they diffuse freely as monomers. 10

Consistent with this prediction, we found that forced homo-oligomerization increased the nuclear pore 11 localization of the CT relative to the nucleoplasmic localization, which presumably results from the 12 KPNA binding (Fig. 6c, d, e) . Thus, we propose that ORF6 homo-oligomerizes to multivalently cross-13 link the FG domains of NUP98 at the NPC (Fig. 6f) . the widely used Hill function to the dose-response data to obtain quantitative metrics, noting that more 1 mechanistic models may also be used to extract biophysical parameters in future studies. 2 3

The dose-response pipeline quantitatively revealed key mechanistic features of the inhibitory function of 4 ORF6. First, ORF6 inhibits multiple carrier-mediated bidirectional transport pathways. ORF6 showed 5 the same potency and Hill coefficient for inhibitions of KPNB1-mediated nuclear import and 6 CRM1/XPO1-mediated nuclear export (Fig. 2a) . Moreover, a recent study showed that ORF6 disrupts 7 the nuclear export of mRNA 23 , which is mediated by a variety of other export carriers 52 . Therefore, the 8 nuclear transport inhibition likely arises from a carrier unspecific perturbation of the nuclear transport 9 machinery, e.g., the NPC impairment, rather than from specific interactions with KPNAs. Second, 10 NUP98 binding of the C-terminal region is necessary but not sufficient for the inhibitory action of 11 ORF6. C-terminal modifications dramatically altered the dose-response characteristics of ORF6 (Fig. 3,  12 4 and 5). Most importantly, the NUP98 binding deficient mutation, M58R 18 , abolished the inhibitory 13 function of ORF6 (Fig 4a) . Compared to ORF6 CoV1 , ORF6 CoV2 showed a higher potency in inhibiting the 14 nuclear transport in our study (Fig. 2) , primarily due to differences in the C-terminal end. In another 15 study, ORF6 CoV2 showed a stronger NUP98 binding than ORF6 CoV1 23 . This correlation suggests that 16 NUP98 binding of the C-terminal region determines the potency of ORF6. The NUP98 binding C-17 terminal region alone without the N-terminal region was not able to inhibit the nuclear transport, 18

indicating that the ORF6 function requires the joint action of N-and C-terminal regions (Fig. 5) . Finally, 19 membrane binding of the N-terminal region is not required for the inhibitory function of ORF6. 20

Disrupting the membrane binding by N-terminally tagging the ORF6 CoV2 with ALFA did not alter the 21 inhibitory dose-response characteristics (Fig. 5) . The N-terminal region was able to be functionally 22

replaced with oligomerization domains that do not bind membranes (Fig. 6) . Therefore, we argue that 23 the major role of the N-terminal region in the nuclear transport inhibition is to drive oligomerization, 1 while the membrane binding is dispensable. Taken together, we propose that ORF6 forms oligomer via 2 the N-terminal region to multivalently cross-link the FG domains of NUP98 via the C-terminal region 3 (Fig. 6f) . The long FG domain of NUP98 was found to interact in vivo with many other FG-NUPs of 4 native human NPCs 53 , so the cross-linking of the FG domains of NUP98 could have strong influence on 5 the overall permeability of the FG barrier to cargo-carrier complexes. Gly-Ser linker. Oligonucleotides for PCRs were purchased from IDT or Genewiz. All the plasmids 22 generated in this study were verified by Sanger sequencing (Genewiz). The recombinant His6-EGFP 23 used in the GFP intensity calibration (Fig. 1b) was expressed in Rosetta 2 (DE3) competent cells 1 (Millipore Sigma #71400) using pDual-EGFP plasmid (Addgene, #63215), purified using HisPur™ Ni-2 NTA Spin Columns (Thermo Fisher, #88226) via a standard protocol, and dialyzed in 50 mM Tris-HCl 3 pH 7.5, 150 mM NaCl, 10% glycerol, 2 mM DTT. The concentration was determined based on the 488 4 nm absorbance and the previously reported value 54 of the extinction coefficient of EGFP at 488 nm, 5 which is 53,300 M -1 cm -1 . The purified His6-EGFP was aliquoted, snap-frozen in liquid nitrogen, and 6 stored in -80˚C. 7 8

Live-cell imaging 10 U2OS stable cell lines stably expressing H2A-Halo and NES-mCherry-LINuS (import probe) or NLS-11 mCherry-LEXY (export probe) were generated in our previous study 25 and were used in this study for 12 measuring the dose-dependent effect of ORF6 on the nuclear transport kinetics. Cells were seeded at 13 based forced oligomerization experiments (Fig. 6a) , 500 nM B/B Homodimerizer (Takara, #635058) 22

was also added. After 2.5-6 hours of the transfection, cells were continually imaged for the 23 simultaneous measurement of the nuclear transport kinetics and GFP intensity for ~24 hours. Cage 1 microscope incubator (OkoLab) was used to maintain the cells at 37°C in 5% CO2 with high humidity 2 during imaging. Applied Research), CFI Plan Apo 20x/0.75NA objective lens (Nikon), and ZT445/514/561/640tpc 10 (Chroma) polychroic mirror. mCherry-labeled import/export probes were imaged using 561-nm laser 11 and ET605/70m emission filter (Chroma), while H2A-Halo:JF646 was imaged using 642-nm laser and 12 ET700/75m emission filter (Chroma). Live-cell time-lapse images were acquired through a ~11-min 13 imaging cycle that consisted of two acquisition phases: pre-activation and activation. Throughout the 14 cycle, the mCherry-labeled transport probes and H2A-Halo:JF646 were imaged every 10 s with 100-ms 15 exposure time. Three frames (30 s) were acquired in the pre-activation phase to determine the baseline 16 nuclear localization of the probe, followed by a 10-min activation phase (61 frames) in which the 17 optogenetic transport probes were activated by 200-ms exposure to the activation laser (447 nm) every 18 10 s. The activation laser also excites GFP, of which images were collected using ET480/40m emission 19 filter (Chroma). To increase the throughput, the imaging cycle was executed at three different fields 20 simultaneously. The imaging cycle was repeated at 3-6-hour interval at different wells, which was fully 21 10.5281/zenodo.1226458). Then, for each nucleus, the time-trajectory of the mean nuclear mCherry 10 intensity was measured and normalized such that the average intensity during pre-activation phase was 1 11 and the background intensity was 0. The following monoexponential decay model was fitted to the 12 normalized nuclear mCherry intensity trajectory in the activation phase to determine the nuclear 13 transport rate k: 14

When the import probe is used, k corresponds to the import rate and b is > 1 (i.e. y increases over time), 16 while when the export probe is used, k corresponds to the export rate and b is < 1 (i.e. y decreases over 17 time). Trust region reflective least-squares algorithm was used for the nonlinear regression. The fitting 18 result was excluded from further analyses if it met one or more of the following criteria: (1) the fitting 19 algorithm did not converge; (2) there were too many missing time points in the nuclear intensity 20 trajectory (i.e. incomplete nucleus tracking); (3) reduced chi-squared statistics was too large; (4) |1-b| 21 was too small; and (5) k was abnormally large. 22

The GFP images were corrected for camera dark noise and uneven illumination. Then, for each nucleus, 1 mean GFP intensity was measured and converted to a concentration using the conversion factor obtained 2 from the intensity calibration procedure described in Fig 1b and below. Dose-response curve was 3 generated by plotting the nuclear transport rate (k) against the GFP concentration (C). The raw data 4 points were logarithmically binned by the GFP concentration within the range between 0.1 µM and 40 5 µM. Median and interquartile range (IQR) were calculated in each bin and plotted as an error bar in the 6 dose-response curve. Using the Levenberg-Marquardt least squares algorithm, the following Hill 7 function was fitted to the binned median values: 8

A is the amplitude of the inhibition (i.e., efficacy), IC50 the half-maximal inhibitory concentration (i.e., 10 potency), and nH the Hill coefficient. 11

The GFP intensity calibration (Fig. 1b) was performed every 2-3 weeks or whenever there was a major 14 change in the microscope (e.g., laser realignment). For the calibration, the U2OS stable cell line 15 after adding the recombinant GFP (I and Iin, respectively) were quantified in the same way as in the 1 dose-response characterization. The extracellular background GFP intensity (Iex) in the presence the 2 added recombinant GFP was determined by blurring the image with a gaussian filter with sigma = 2 3 pixels and finding the most frequent intensity value. For each recombinant GFP concentration (C), Iin -4

Iex was plotted against I. A linear model was fitted to determine the x-intercept (I0). The slope of the 5 linear model fitted to the plot of I0 against C was used as the intensity-to-concentration conversion 6 factor. Levenberg-Marquardt least squares algorithm was used for the linear regressions. 7 8 Immunofluorescence 9

Sample preparation 10 U2OS cells were seeded at ~10,000 cells per well in a glass-bottom eight-well chambered coverslip 11 (ibidi, #80827) with 250 µl complete DMEM. On the next day, cells in each well were transfected with a 12 plasmid encoding GFP-2A-POI using TransIT-2020 transfection reagent as described above, where POI 13 is an epitope tagged protein (e.g., ALFA-ORF6 CoV2 and ORF6-ALFA CoV2 ). After ~24 hours of 14 transfection, GFP images were acquired to determine the concentration of POI as described above. 15

Then, cells were fixed using 2% PFA/PBS for 20 min at RT, permeabilized using 25 µg/ml digitonin 16 (Millipore Sigma, #300410) in PBS for 20 min at RT, and blocked using 1% BSA/PBS for 1 hour at RT. produced at the same rate as separate polypeptides. The rate of the light-induced nuclear/cytoplasmic 6 translocation of the transport probe and GFP intensity were simultaneously measured in individual cells. 7

The measurement was repeated at 3-6-hr interval for ~24 hrs after the transfection. The GFP intensity 8 was converted to GFP concentration via calibration. The Hill function was fitted to the relationship 9 between the nuclear transport rate and GFP concentration ("dose-response curve") to obtain the dose-10 response parameters. Scale bar 50 µm. b, GFP intensity calibration procedure. (Step 1) Measure GFP 11 intensity (I) of individual cells in the same way as in the dose-response characterization. ( Step 2) Add a 12

known concentration (C µM) of recombinant GFP to the culture and measure the intracellular and 13 extracellular GFP intensities (Iin and Iex). ( Step 3) Fit a linear model to the plot of Iin -Iex vs I to 14

determine the x-intercept (I0), which is the GFP intensity equivalent to the concentration C. ( Step 4) 15

Repeat Steps 1 to 3 with different C's and fit a linear model to the plot of I0 vs C to determine the slope, 16

which serves as the intensity-to-concentration conversion factor. Example data is shown in 17

Supplementary Fig. 1 . and shaded area represent the best Hill function fit and corresponding 95% confidence interval, 7

respectively. The parameter estimates and standard errors are shown in the plots. variant. c, Dose-response curves for the nuclear import inhibition by (left) #13r (deletion of YP from the 5 C-terminal end of ORF6 CoV1 ) and (right) #13 (addition of YP to the C-terminus of ORF6 CoV2 ). Green 6 circles are the raw data points. Black error bar represents median and interquartile range in each bin. 7

Black line and shaded area represent the best Hill function fit and the corresponding 95% confidence 8

interval, respectively. The parameter estimates and standard errors are shown in the plots. Dotted line is 9

the best Hill function fit for the (left) ORF6 CoV1 and (right) ORF6 CoV2 wild types shown for comparison. 10

The dose-response curves for the other variants are shown in Supplementary Figure 4 . Dose-response curves for the nuclear import inhibition by a, ORF6 CoV2 WT (black circles), residues 1-3 37 (yellow upward-pointing triangles), residues 1-55 (green squares), and M58R mutant (red downward-4 pointing triangles); and b, WT (black circles), residues 38-61 (yellow upward-pointing triangles), 5

residues 19-61 (green squares), and D22-30 mutant (red downward-pointing triangles). n>1100 cells for 6 each variant. Error bars show the median and interquartile range in each bin. Shaded area represents 7 95% confidence interval of the best Hill function fit. Fits are not shown for the dose-response curves that 8 failed to reject the null hypothesis that the inhibition amplitude (A) is zero (a= 0.05). Representative NUP98 and ALFA immunofluorescence images of (left) ALFA-ORF6 CoV2 and (right) 6 ORF6 CoV2 -ALFA at low and high GFP concentrations when focused on the basal plane ("bottom") and 1 midplane ("mid") of the nucleus. c, Representative NUP98 and ALFA immunofluorescence image of 2 ALFA-ORF6 CoV2 -ALFA when focused on the midplane of the nucleus. Scale bar 10 µm. Insets: 10x 3 magnification. Green and red colors correspond to the absence and presence of 500 nM B/B homodimerizer, 6

respectively. b, The dose-response curve for ALFA-p53TD-CT, where p53TD is the tetramerization 7 domain (residues 326-356) of p53. In a and b, n>1500 cells for each condition. Error bar shows median 8 and interquartile range in each bin, and the solid line and shaded area represent the best Hill function fit 9

and the corresponding 95% confidence interval, respectively. NUP98 and ALFA immunofluorescence presence of 500 nM B/B homodimerizer, and e, ALFA-p53TD-CT. Samples were focused on the 12

The Structure of the Nuclear Pore Complex (An Update)

Integrative structure and functional anatomy of a nuclear pore complex

Disorder in the nuclear pore 17 complex: the FG repeat regions of nucleoporins are natively unfolded

The location of the transport gate in the nuclear pore complex

Permeability of single nuclear pores

Kinetic analysis of translocation through nuclear pore complexes

Nucleocytoplasmic transport: taking an inventory

Classical nuclear localization signals: definition, function, and interaction with 5 importin alpha

Exportin 1 (Crm1p) is an essential nuclear export 7 factor

Structural Biology and Regulation of Protein Import into the Nucleus

Mechanistic Insights from Structural Analyses of Ran-GTPase-Driven Nuclear 11

Export of Proteins and RNAs

Transport into and out of the nucleus

Regulation of type I interferon responses

Viral Appropriation: Laying 18

Activation and evasion of type I interferon responses by SARS-CoV-2

Evasion of Type I Interferon by SARS-CoV-2

A Rare Deletion in SARS-CoV-2 ORF6 Dramatically Alters the Predicted 1

Three-Dimensional Structure of the Resultant Protein. bioRxiv

The ALFA-tag is a highly versatile tool for nanobody-based bioscience 4 applications

Inhibition of in vitro nuclear 6 transport by a lectin that binds to nuclear pores

Tau Protein Disrupts Nucleocytoplasmic Transport in Alzheimer's

Toxic PRn poly-dipeptides encoded by the C9orf72 repeat expansion block 11 nuclear import and export

The matrix protein of vesicular 14 stomatitis virus inhibits nucleocytoplasmic transport when it is in the nucleus and associated with 15 nuclear pore complexes

VSV disrupts the Rae1/mrnp41 mRNA nuclear export pathway

Aggregation of VSV M protein is reversible 20 and mediated by nucleation sites: implications for viral assembly

Molecular Pathogenesis of the Tauopathies

The C9orf72 GGGGCC repeat is translated into aggregating dipeptide-repeat 3 proteins in FTLD/ALS

Structural basis of multivalent binding to wheat germ agglutinin

Glycosylation of the nuclear pore

Redesigning an FKBP-ligand interface to generate chemical dimerizers with 13 novel specificity

Crystal structure of the tetramerization domain of the 16 p53 tumor suppressor at 1.7 angstroms

High-resolution structure of the oligomerization domain of p53 by 19 multidimensional NMR

mRNA nuclear export at a glance

Insets: 10x magnification. f, Proposed model. The N-terminal 1 region (NT) of ORF6 drives the oligomerization