key: cord-0861402-d87j54ef
authors: Miyamoto, Yoichi; Itoh, Yumi; Suzuki, Tatsuya; Tanaka, Tomohisa; Sakai, Yusuke; Koido, Masaru; Hata, Chiaki; Wang, Cai-Xia; Otani, Mayumi; Moriishi, Kohji; Tachibana, Taro; Kamatani, Yoichiro; Yoneda, Yoshihiro; Okamoto, Toru; Oka, Masahiro
title: SARS-CoV-2 ORF6 disturbs nucleocytoplasmic trafficking to advance the viral replication
date: 2021-02-24
journal: bioRxiv
DOI: 10.1101/2021.02.24.432656
sha: 60c6b2bed39dc3e576c4f96e04f7e6ec2a8fe6ff
doc_id: 861402
cord_uid: d87j54ef

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the virus responsible for the coronavirus disease 2019 pandemic. ORF6 is known to antagonize the interferon signaling by inhibiting the nuclear translocation of STAT1. Here we show that ORF6 acts as a virulence factor through two distinct strategies. First, ORF6 directly interacts with STAT1 in an IFN-independent manner to inhibit its nuclear translocation. Second, ORF6 directly binds to importin α1, which is a nuclear transport factor encoded by KPNA2, leading to a significant suppression of importin α1-mediated nuclear transport. Furthermore, we found that KPNA2 knockout enhances the viral replication, suggesting that importin α1 suppresses the viral propagation. Additionally, the analyses of gene expression data revealed that importin α1 levels decreased significantly in the lungs of older individuals. Taken together, SARS-CoV-2 ORF6 disrupts the nucleocytoplasmic trafficking to accelerate the viral replication, resulting in the disease progression, especially in older individuals.

6 α proteins 9 . However, the exact molecular mechanism of the effects of SARS-CoV-2 ORF6 on the 96 nucleocytoplasmic trafficking remains largely unknown.

In this study, therefore, we characterized the function of SARS-CoV-2 ORF6 on the 98 nucleocytoplasmic protein transport. First of all, we examined the subcellular localization of ORF6 protein in 99 cells infected with distinct SARS-CoV-2 strains using originally established antibodies, and also demonstrated 100 that ORF6 functions as a virulence factor for COVID-19 using a hamster model and a newly produced replicon 101 system. Furthermore, we found that ORF6 directly binds to STAT1 to suppress the IFN-induced nuclear 102 localization and the nuclear shuttling in the absence of IFN-stimulation. In addition, the direct binding of ORF6 103 to importin α1 significantly reduces the cNLS-cargo transport, indicating that ORF6 influences importin α1 104 independently of STAT1. We also found that the viral replication of SARS-CoV-2 is enhanced in KPNA2 105 knockout cells, suggesting that importin α1 acts for the suppression of viral propagation. Lastly, the analyses of 8 lower levels in the lung cells infected with SARS-CoV-2/ΔORF6 than in those infected with the WT virus ( Fig.   134 1G). These data suggest that ORF6 is involved in the viral replication and pathogenesis of 135 136 ORF6 inhibits the nuclear localization of STAT1 following the IFN stimulation 137 Several studies have already shown that ORF6 inhibits the nuclear localization of STAT1 in response to type-I 138 IFN (IFN-α or -β) stimulation 7, 8, 9, 10, 13 . Here, we attempted to verify the inhibitory effect of SARS-CoV-2 139 ORF6 on a type-II IFN (IFN-γ)-activated STAT1. The AcGFP fused ORF6 was transfected into HeLa cells, and 

Next, in order to confirm that the PY-STAT1 is excluded from the nucleus by ORF6, we evaluated the 146 expression of PY-STAT1 downstream genes using quantitative RT-PCR (qRT-PCR) . In comparison to the 147 AcGFP-transfected control cells, the AcGFP-ORF6-transfected cells showed significant down-regulation of 148 IFN-γ-inducible protein 10 (IP-10) 30 mRNA 6 h post-transfection or later upon stimulation (Fig. 2D ). To clarify 149 whether the nuclear exclusion of PY-STAT1 by ORF6 affects the expression of the IFN-stimulated response 150 element (ISRE)-containing gene, a luciferase assay was performed in Huh7 cells. The cells were transfected 151 with a luciferase reporter plasmid including ISRE together with AcGFP or AcGFP-ORF6. We observed a 152 significant repression of the relative luciferase values in the AcGFP-ORF6-transfected cells compared to that 153 of AcGFP-transfected cells (Fig. 2E) . Finally, we confirmed that the relative ISRE luciferase value was 154 significantly suppressed when the SARS-CoV-2 infected cells were stimulated by IFN-γ (Fig. 2F) . These results

indicate that SARS-CoV-2 ORF6 suppresses the nuclear translocation of PY-STAT1 to inhibit the activation of 156 STAT1-downstream genes. 173 174 ORF6 directly binds to STAT1 in an IFN-independent manner 175 Next, in order to characterize the interplay between ORF6 and STAT1 in more detail, we analyzed the 176 subcellular distribution of Flag-tagged STAT1 in AcGFP-ORF6-transfected HeLa cells. In addition to the WT 177 ORF6, the in-frame 9 amino acids deletion mutant (loss of amino acids 22 to 30; referred to as ORF6Δ9) which 178 were reported in the previous works 31, 32 was also examined. Under non-stimulated conditions, Flag-STAT1 179 was mainly localized in the cytoplasm in either AcGFP-, AcGFP-ORF6-or AcGFP-ORF6Δ9-transfected cells 180 (Fig. 4A, B) . Upon IFN-γ stimulation, Flag-STAT1 was distributed into the nucleus in the AcGFP-transfected 10 cells. In contrast, it was retained in the cytoplasm in AcGFP-ORF6 WT-and Δ9-transfected cells (Fig. 4C, D) .

Interestingly, statistical analysis revealed that even under non-stimulated conditions, the cytoplasmic intensities 183 of Flag-STAT1 were significantly higher in the AcGFP-ORF6 WT-or Δ9-transfected cells compared to those 184 observed in the AcGFP-transfected control cells (Fig. 4B ), suggesting that STAT1 may shuttle between the 185 nucleus and the cytoplasm in the absence of IFN stimulation and be trapped in the cytoplasm by ORF6.

Therefore, to address the possibility of the interaction between ORF6 and non-activated STAT1, we 187 first performed an immunoprecipitation assay using AcGFP-ORF6 (HA-ORF6)-and Flag-STAT1-transfected 188 cells. As a result, we found that HA-ORF6 was precipitated with Flag-STAT1 in the absence of IFN stimulation 189 ( Fig. 4E ). Next, to know whether ORF6 directly binds to STAT1, bacterially purified recombinant STAT1 190 protein was incubated with either recombinant GST-GFP or recombinant GST-GFP fused ORF6 full-length 191 protein (GST-GFP-ORF6). Fig. 4F showed that STAT1 was directly bound to GST-GFP-ORF6. To address 192 whether the binding of ORF6 to STAT1 is mediated by the amino acids 49-61 in the C-terminal region (referred 193 to as M0 in this study, see Fig. 3A ), we produced a recombinant protein consisting of the 13 amino acids (M0) 194 which was fused to GST and GFP (GST-M0-GFP). As shown in Fig. 4G 

Previous studies showed that the SARS-CoV-and SARS-CoV-2-associated inhibition of nuclear translocation 201 of STAT1 is accomplished through the interaction between ORF6 and importin α1/KPNA2 7, 10 . However, as 202 described above, ORF6 binds directly to STAT1 in the absence of IFN stimulation in the cytoplasm, which 203 raises a possibility that the binding of ORF6 to importin α proteins might not be required to inhibit the nuclear 204 accumulation of STAT1. Therefore, we attempted to examine the interplay between the ORF6 and importin α 205 11 proteins. Since ORF6 has been shown to alter the distribution of importin α1/KPNA2 and/or importin 206 α5/KPNA1 7, 9 , we first validated these findings concerning all human importin α subtypes. Consistent with the 207 previous reports, in AcGFP-transfected cells (control) , the overexpressed Flag-importin α proteins were mainly 208 localized in the nucleus (Fig. 5A, B, Fig. S2 ). In contrast, in AcGFP-ORF6-transfected cells, the localization of 209 Flag-importin α1, α3, α4, α6, and α8 remarkably changed to the cytoplasm, while Flag-importin α5 and α7 were 210 still mostly localized in the nucleus (Fig. 5A, B, Fig. S2 ). The analysis of nuclear fluorescence intensities ratio 211 clearly showed that in AcGFP-ORF6-transfected cells, the nuclear distribution of Flag-importin α1 was 212 drastically shifted to the cytoplasm, while the Flag-importin α5 was mainly retained in the nucleus (Fig. 5C ).

These data indicate that ORF6 has distinct effects on each importin α subtype.

Importin α1 shuttles between the nucleus and the cytoplasm in the presence of ORF6 216 Next, we examined whether ORF6 directly binds to importin α proteins. Purified recombinant GST-GFP-ORF6 217 was incubated with Flag-importin α1 or Flag-importin α5, respectively, and then the GST-proteins were pulled-218 down using glutathione Sepharose beads to know whether the Flag-importin α proteins were co-precipitated or 219 not. Using western blotting, we identified that Flag-importin α1, but not Flag-importin α5, mainly binds to 220 ORF6 (Fig. 5D ).

Direct binding of ORF6 to importin α1 suggest a possibility that ORF6 may inhibit the mobility of 222 importin α1 to tether it in the cytoplasm, as reported previously with SARS-CoV ORF6 7 . Therefore, we next 223 tried to know whether importin α1 moves from the cytoplasm to the nucleus even in the presence of ORF6. For 224 this, we focused on the characteristic feature of importin α1 that it accumulates in the nucleus in response to 225 cellular stresses such as oxidative stress 33, 34 . That is, it has been clearly demonstrated that while in unstressed 226 cells importin α1 shuttles between the nucleus and the cytoplasm, in stressed cells importin α1 accumulates in 227 the nucleus due to the inhibition of RanGTP-dependent nuclear export of importin α by a collapse in the RanGTP 228 gradient 33, 34 . Therefore, we speculated that if importin α1 cannot move from the cytoplasm to the nucleus by 12 ORF6, we cannot observe its nuclear accumulation under stress conditions. Hence, HeLa cells were transfected 230 with AcGFP-ORF6 and Flag-importin α1 and treated with 200 μM of H2O2 for 1 h. Unexpectedly, we found 231 that under the oxidative stress conditions, Flag-importin α1 was localized into the nucleus in the ORF6-232 transfected cells (Fig. 5E , F). The same phenomenon was also observed for endogenous importin α1 ( Fig. S3A- 

C). Thus, these results mean that although importin α1 seems to be tethered in the cytoplasm in the ORF6-234 transfected cells, it retains its original capacity to move between the nucleus and the cytoplasm even in the 235 ORF6-transfected cells. Taken together with the results that the subcellular localization of neither importin β1 236 nor CAS, the export factor for importin α, was dramatically altered in the ORF6-transfected cells (Fig. S3D , E),

it is most likely that the nuclear exclusion of STAT1 by ORF6 is not due to the cytoplasmic tethering of importin 238 α1.

ORF6 negatively regulates the importin α/β1 pathway 241 As described above, since we found that ORF6 directly binds to importin α1, we next tried to know whether 242 ORF6 directly affects the importin α1-mediated nuclear transport pathway or not. For this, HeLa cells were 243 transfected with mCherry fused SV40T-NLS (mCherry-NLS) together with AcGFP or AcGFP-ORF6, and then 244 the nuclear intensities of the fluorescent substrate were measured. Although the majority of mCherry-NLS was 245 observed in the nucleus of AcGFP-ORF6-transfected cells, its cytoplasmic proportion was significantly 246 increased compared to that observed in the AcGFP-transfected control cells (Fig. 6A, B ).

Next, we compared the binding ability of importin α1 to the cNLS-containing cargo (GST-NLS-GFP) 248 and ORF6 (GST-GFP-ORF6). The GST pull-down assay showed that importin α1 more efficiently interacted 249 with cNLS than ORF6 (Fig. 6C) . To assess the inhibitory effects of ORF6 on the importin α/β1-mediated nuclear 250 transport of GST-NLS-mRFP, a digitonin-permeabilized semi-intact nuclear transport assay was performed. As 251 a result, we observed that the addition of excess amounts of ORF6 significantly inhibited the nuclear 252 13 translocation of the cNLS-cargo (Fig. 6D , E), suggesting that ORF6 affects the importin α/β1 pathway, when it 253 exists in large quantities in cells.

To further investigate the effects of ORF6 on other important signaling pathways mediated by importin 255 α/β1 other than STAT1, we focused on the following signaling molecules, hypoxia inducible factor 1α (HIF-256 1α) and Nuclear factor-kappa B (NF-κB) component p65/RelA, since they have been shown to be transported symptoms. Using the GTEx dataset, we analyzed whether the importin α1 expression level in lung tissues and 281 whole blood cells, with sexes as well as across age is correlated with symptoms using the European ancestry 282 (EUR) samples. We found that the importin α1 levels significantly decreased with ages in lung tissues (Fig. 7A) 283 and that the tendency was associated with males rather than female (Fig. 7B ). On the other hand, the expression 284 levels in whole blood increased in an age-dependent manner in both sexes (Fig. 7C , D). We also found that the 285 KPNA2 expression levels in the lungs were significantly lower in Asian ancestry than those in European, African, 286 and other ancestries (Fig. S5 ). These data suggest that the low expression level of importin α1 in the lungs might 287 represent a COVID-19-associated concern observed in older patients, particularly for males, and be one of the 288 risk factors for infection of SARS-CoV-2.

In this study, we demonstrated that SARS-CoV-2 ORF6 plays an important role in viral replication and 291 pathogenesis of COVID-19 in vivo. In addition, we discovered that ORF6 acts on the nucleocytoplasmic 292 signaling via two distinct ways. That is, first, ORF6 inhibits the nuclear translocation of one of the key signaling 293 molecules for COVID-19, STAT1, through its direct binding to antagonize the IFN signaling. Second, ORF6 294 directly affects the function of importin α to impair the nuclear transportation of cNLS-containing cargos 295 including signaling molecules such as HIF-1α and NF-κB p65.

A previous study reported that SARS-CoV ORF6 tethered importin α1/KPNA2, but not importin 297 α5/KPNA1, to the ER and, as a result, sequestered importin β1 into the ER/Golgi segment through the 298 interaction with importin α1, resulting in the inhibition of PY-STAT1 nuclear import 7 . Since the nuclear 299 transport of PY-STAT1 is known to be mediated by importin α5/KPNA1 24, 25 and we showed here that importin 300 α1 can enter the nucleus even in the presence of ORF6 under oxidative stress conditions, it is unlikely that 301 SARS-CoV-2 ORF6 tethers importin α1 to the ER to cause the nuclear exclusion of STAT1.

In contrast, we found that the Flag-STAT1 was more abundantly localized in the cytoplasm in the 303 absence of IFN stimulation in the ORF6-transfected cells than in the control WT cells (Fig. 4A, B) . It was 304 previously demonstrated that unphosphorylated STAT1 shuttles between the nucleus and the cytoplasm, and 305 this shuttling might play an important role in regulating the expression of IFN stimulated genes 12, 38, 39, 40 . 306 Moreover, in this study, we found that the bacterially purified STAT1 proteins, which are not phosphorylated, 307 binds to ORF6. Since the phosphorylation of STAT1 has been shown to be unaffected by ORF6 upon the IFN 308 stimulation 8, 10 , the interaction between ORF6 and STAT1 might occur in a phosphorylation-independent 309 manner. Collectively, we propose a scenario that the nuclear exclusion of STAT1 is caused by the direct binding 310 with ORF6 independently of importin α proteins.

On the other hands, we found that the subcellular localizations of all importin α subtypes were altered 312 in ORF6-transfected cells, suggesting a possibility that ORF6 directly or indirectly influences the importin α/β1-313 16 mediated nuclear transport pathways. Consistent with this possibility, we further found that the nuclear 314 accumulation of the mCherry-NLS substrate was significantly reduced in the AcGFP-ORF6-transfected cells.

In addition, using the semi-intact nuclear transport assay, we demonstrated that the addition of ORF6 inhibits 316 the nuclear transportation of GST-NLS-mRFP. Furthermore, we found that ORF6 negatively regulates the 317 nuclear import of HIF-1α and NF-κB p65, which have been already shown to be mediated by importin α proteins Recently, it has been shown that the specific interaction of ORF6 with the NPC components, Nup98 321 and Rae1, might disrupt the nuclear transport 9, 28, 41 . Consistently, it has been already known that the hijacking 322 of the NPC components suppresses the host mRNA export system 41 . Moreover, Miorin et al. provided a 323 supporting evidence that the arginine substitution at residue 58 of ORF6 abolishes its IFN antagonistic function 324 9 . In this study, we demonstrated that STAT1 binds to ORF6 via the residues 49-61 from the C-terminus (M0) 325 which contains a methionine amino acid at the 58 position. Furthermore, we found that this C-terminal region 326 of ORF6 plays a critical role in altering the subcellular localization of importin α1. Taken together, we propose 327 that the ORF6 negatively regulates the nucleocytoplasmic trafficking through the binding with importin α1 and 328 some NPC components. Considering that various mutations and deletions have been found in ORF6 31, 42 , further 329 studies will be required to understand how ORF6, especially its C-terminus region functions for the viral 330 infection in order to utilize it as a therapeutic target.

Consistently, we found that the viral propagation of SARS-CoV-2 is enhanced in KPNA2 KO cells. In 332 addition, the GTEx dataset revealed that the KPNA2 levels in lung tissue significantly decrease with the 333 advancement in age, and that this decrease is more remarkable in males than in females. Moreover, the another 334 GTEx analysis indicated that the levels of KPNA2 in the lungs tend to be lower in Asians than in other genetic 335 ancestries, although further data need to be required to make a definite conclusion, since the sample size for the 336 Asian group is smaller than that of the other groups. Since hypoxia has been recognized as a pathogenic factor 337 17 in COVID-19 patients 43, 44 , we suspect that ORF6 might impair the nuclear import of HIF-1α, which is required 338 for the response to hypoxia, to affect lung cell functions in the COVID-19 patients, so that the lung oxygen 339 levels cannot be maintained appropriately. Thus, the downregulation of importin α1 may accelerate the 340 replication of the virus in the lungs, in particular in older individuals. In conclusion, we propose that in the lungs 341 of older individuals, SARS-CoV-2 ORF6 exhibits dual effects on the viral proliferation by inhibiting the STAT1 342 signaling and negatively regulating the importin α/β1-mediated nuclear transport pathways to avoid the 343 interferon immune responses. Understanding the effects of SARS-CoV-2 proteins on the nucleocytoplasmic 344 trafficking system might provide a novel approach for COVID-19 therapeutics. 

CoV-2 ORF6-specific antibody was screened using ELISA, western blotting, and immunostaining of hybridoma 366 supernatants. Finally, hybridoma clone producing the monoclonal antibody later named 8B10, was selected.

Using a rat isotyping kit the MAb 8B10 was found to be an IgG 1 (k) antibody subtype. The monoclonal antibody 368 against SARS-CoV-2 NP (3A9 clone) was generated by Cell Engineering Corporation (Osaka, Japan). 

Full-length STAT1 was amplified from a previously subcloned plasmid 46 using the primers described 392 in Table S1 . The PCR products were cloned into a pcDNA5/FRT/3xFLAG expression vector 47 . Human 393 20 importin αs including importin α1/KPNA2 and importin α5/KPNA1 were cloned into a pcDNA5/FRT/FLAG 394 expression vector, as previously described 47 . For constructs encoding the SV40T antigen NLS (NLS; 395 PKKKRKVED), the relevant oligonucleotides (Table S1) were ligated into the pmCherry-C1 vector (Clontech).

Full-length ORF6 cDNA was amplified using the specific primers described in Table S1 . The PCR products 399 were cloned into a pGEX6P2 vector (Clontech) which was subcloned the GFP gene at the N-terminus 48 .

Construct integrity was confirmed by DNA sequencing. For constructs encoding the C-terminus of ORF6 (M0), 401 the relevant oligonucleotides (Table S1) 

For IP-10, total RNA was isolated using ReliaPrep™ RNA Tissue Miniprep System (Promega, Madison, WI, 461 USA) according to the manufacturer's instructions. One microgram of total RNA and the PrimeScript RT 462 reagent kit (Takara Bio.) were used to perform the first-strand cDNA synthesis. The PCR reaction was 463 performed as previously described 50 . The PCR primers including those of β-actin are described in Table S2 .

For detection of N2 in SARS-CoV-2, total RNA of Huh7-ACE2 or lung homogenates were isolated 465 using ISOGENE II (Nippon Gene, Toyama, Japan). Real-time RT-PCR was performed with the Power SYBR

Green RNA-to-CT 1-Step Kit (Applied Biosystems, Foster City, CA, USA) using an AriaMx Real-Time PCR 467 system (Agilent, Santa Clara, CA, USA). The relative quantification of the target mRNA levels was performed 468 using the 2 -ΔΔCT method. β-actin was used as the housekeeping gene. The primers used are described in Table   469 S2.

The amount of RNA copies in the culture medium was determined using a qRT-PCR assay as previously Vero-TMPRSS2 were seeded into 24-well plates (80,000 cells/well) at 37 °C in 5% CO2 for overnight. The 482 supernatants were serially diluted using inoculated medium and incubated for 2 h. Next, the culture medium 483 was removed, fresh medium containing 1% methylcellulose (1.5 mL) was added, and the cells were cultured 484 for 3 more days. Lastly, the cells were fixed with 4% paraformaldehyde in PBS (Nacalai Tesque, Kyoto, Japan) For the single-guide RNA (sgRNA) targeting KPNA2, the targeting sequences were designed using three 504 different sequences for each gene as previously described 52 . The targeting sequences were synthesized using 505 DNA oligos (Eurofins Genomics, Tokyo, Japan), and cloned into the lentiCRISPR v2 (Addgene, #52961) 506 digested by BmsBI (New England Biolab, MA, USA). The target sequences for KPNA2 were described in Table   507 S3. Lentiviruses expressing three types of target sequences per gene were mixed, introduced into the Huh7 cells 508 expressing the ACE2 receptor (Huh7-ACE2), and maintained in a culture medium supplemented with 1 µg/mL 509 puromycin for 3 weeks. For viral infection, sgControl (sgCtl) or sgKPNA2 Huh7-ACE2 cells were seeded into 510 24-well plates and incubated at 37 °C for 24 h. The different SARS-CoV-2 strains were used to infect the cells 511 25 (MOI 0.1) and supernatants were collected at 0, 6, 12, and 24 h. The intracellular viral RNA was quantified 512 using qRT-PCR while the viral titers were quantified using the plaque forming assay.

SARS-CoV-2 replicon vector, pBAC-SCoV2-Rep, was generated using the CPER reaction as previously 516 described 53 , with some modifications. Briefly, seven DNA fragments covering the SARS-CoV-2 genome 517 (excluding the region spanning from S gene to ORF8 gene) were amplified using PCR, and subcloned into a 518 pCR-Blunt vector (Invitrogen). The DNA fragments containing cytomegalovirus (CMV) promoter, a 25 519 nucleotide synthetic poly(A), a hepatitis delta ribozyme as well as a bovine growth hormone (BGH) termination, 520 and a polyadenylation sequences (the lightly shaded region in Fig. 3D ) were amplified using a conventional 521 overlap extension PCR, and subcloned into the NotI sites of pSMART BAC vector (Lucigen, Middelton, WI, 522 USA). The luciferase reporter vector pGL4 was used as the template for PCR amplification of Renilla luciferase 523 gene. For CPER reaction, nine DNA fragments that contain approximately 40-bp overlapping ends for two 524 neighboring fragments were amplified by PCR using the aforementioned plasmids. Next, the PCR fragments 525 were mixed equimolarly (0.1 pmoL each) and subjected to CPER reaction using the PrimeSTAR GXL DNA 526 polymerase (Takara Bio.). The CPER product was extracted using phenol-chloroform, followed by ethanol 527 precipitation, resolved in TE buffer, and transformed into the BAC-Optimized Replicator v2.0 Electrocompetent 528 Cells (Lucigen). The replicon vector was maxipreped using a NucleoBond Xtra BAC kit (Takara Bio.).

SARS-CoV-2 recombinants were generated by CPER reaction as previously described 54 with some CoV-2 and treated with IFN-γ (50 ng/mL) for 12 h. The luciferase activity was detected using the Dual-

Luciferase Reporter Assay System (Promega) according to the manufacturer's instructions.

Semi-intact nuclear transport assay 562 A digitonin-permeabilized in vitro nuclear transport assay was performed as previously described 47 . The NLS 563 substrate GST-NLS-GFP was used 4 pmoL in 10 L of reaction mixture, and the competitive substrate AcGFP-

ORF6 was added to the assay with 20 pmoL, 40 pmoL, and 80 pmoL which represented 5×, 10×, or 20× the 565 NLS-substrate dosage, respectively.

We obtained publicly available data regarding the expression levels of KPNA genes, sex, age category (20-29, 

Viral RNA in the supernatants was quantified using qRT-PCR. ***P < 0.001, two-tailed Student's t-test. E.

Percent body weight changes were calculated for all hamsters infected with SARS-CoV-2 WT or ΔORF6. Data 598 are mean ± SD from four independent animals. F. Viral RNA in lung homogenates from hamsters was quantified 599 using qRT-PCR. *P < 0.05, two-tailed Student's t-test. G. Immunohistochemistry of SARS-CoV-2 antigen (NP 600 protein) in lung lobes of hamster infected with SARS-CoV-2 WT or ΔORF6, respectively. Scale bars: 100 μm. 

The graph represents the relative fluorescence values of Flag-STAT1 in the nucleus compared to those of the 632 31 whole cells in A. Signal intensities from total 45 nuclei from two independent experiments. ***P < 0.001, one-633 way ANOVA. C. Immunofluorescence of Flag-STAT1 in HeLa cells transfected with AcGFP, AcGFP-ORF6

WT or AcGFP-ORF6Δ9 following IFN-γ stimulation. Anti-GFP or anti-Flag antibodies were used for detection.

DAPI was used to stain the DNA. Scale bars: 30 μm. D. The graph represents the relative fluorescence values 636 of Flag-STAT1 in the nucleus compared to those of the whole cells in C. Signal intensities from total 45 nuclei 637 from two independent experiments were measured. ***P < 0.001, one-way ANOVA. E. Immunoprecipitation 

The bottom panel represents the proteins bound to the beads and stained with CBB. Inputs are 1/10th of the 672 amount of each importin α that was used for the reaction. D. An in vitro semi-intact nuclear transport assay was 673 performed to measure the nuclear import of GST-NLS-mRFP in the presence of AcGFP-ORF6. Digitonin-674 permeabilized HeLa cells were incubated with GST-NLS-mRFP, importin α1, importin β1, RanGDP, p10/NTF2, 675 GTP, and ATP regeneration system. The reaction mixture was added 5×, 10×, or 20× concentration of AcGFP-676 ORF6 compared to that of the NLS-substrate. After incubation for 30 min, the mRFP signals were detected 677 using a fluorescence microscope. DAPI was used to stain the DNA. Scale bars: 30 μm. E. The graph represents 678 the nuclear fluorescence values of GST-NLS-mRFP in D. Signal intensities of total 100 nuclei were measured 679 and the statistically analyzed using a one-way ANOVA (***P < 0.001). F. Immunofluorescence of HIF-1α in 33 HeLa cells transfected with AcGFP or AcGFP-ORF6 following CoCl2 treatment. Anti-GFP or anti-Flag 681 antibodies were used for detection. DAPI was used to stain the DNA. Scale bars: 30 μm. G. The graph represents 682 the relative fluorescence values of the nucleus compared to those of the entire cells in F. Signal intensities of 683 total 50 nuclei from two independent experiments were measured. ***P < 0.001, two-tailed Student's t-test. H.

Immunofluorescence of NF-κB p65 in HeLa cells transfected with AcGFP or AcGFP-ORF6 following TNF-α 685 stimulation. Anti-GFP or anti-Flag antibodies were used for detection. DAPI was used to stain the DNA. Scale 686 bars: 30 μm. I. The graph represents the relative fluorescence values of the nucleus compared to whole cells in 687 H. Signal intensities of total 50 nuclei from two independent experiments were measured. ***P < 0.001, two-688 tailed Student's t-test. J-K. Huh7 cells expressing the ACE2 receptor (Huh7-ACE2) introduced with sgControl 689 (sgCtl) or sgKPNA2 were infected with SARS-CoV-2 and supernatants were collected at 0, 6, 12, and 24 h.

Intracellular viral RNA was quantified using qRT-PCR (J) while the viral titers (K) were quantified using plaque 691 forming assay. Statistical significance was determined using a two-way ANOVA (***P < 0.001). 

GTEx donors whose estimated ancestry was EUR (n = 436 for lung tissues in A and B, and n = 558 for whole 695 blood in C and D) were used. P-values for the trends between KPNA2 expression levels and age categories were 696 obtained using the two-sided Jonckheere-Terpstra test. The box represented the first and third quartiles and the 697 center line represented the median. The upper whisker extended from the hinge to the highest value that is within 698 the 1.5 × IQR of the hinge, the lower whisker extended from the hinge to the lowest value within the 1.5 × IQR 699 of the hinge, and the data beyond the end of the whiskers were plotted as points. F, female; M, male; IQR, 700 interquartile range. Viral RNA in the supernatants was quantified using qRT-PCR. ***P < 0.001, two-tailed Student's t-test. E.

Percent body weight changes were calculated for all hamsters infected with SARS-CoV-2 WT or ΔORF6. Data are mean ± SD from four independent animals. F. Viral RNA in lung homogenates from hamsters was quantified using qRT-PCR. *P < 0.05, two-tailed Student's t-test. G. Immunohistochemistry of SARS-CoV-2 antigen (NP protein) in lung lobes of hamster infected with SARS-CoV-2 WT or ΔORF6, respectively. Scale bars: 100 μm.

This work was funded by the Japan Agency for Medical Research and Development (AMED)

AUTHOR CONTRIBUTION: 706 Conceptualization: Y.M. and

708 Investigation

A pneumonia outbreak associated with a new coronavirus of probable bat origin

Genomic characterization of the 2019 novel human-pathogenic coronavirus isolated 724 from a patient with atypical pneumonia after visiting Wuhan. Emerging microbes & infections 9

Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for 728 virus origins and receptor binding

The Architecture of SARS-CoV-2 Transcriptome

Molecular immune pathogenesis and diagnosis of COVID-19

SARS-CoV-2 nsp13, nsp14, nsp15 and orf6 function as potent interferon antagonists

Emerging microbes & infections 9

Severe acute respiratory 740 syndrome coronavirus ORF6 antagonizes STAT1 function by sequestering nuclear import factors on the 741 rough endoplasmic reticulum/Golgi membrane

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

SARS-CoV-2 Orf6 hijacks Nup98 to block STAT nuclear import and antagonize 747 interferon signaling

Evasion of Type I Interferon by SARS-CoV-2

Mechanisms of type-I-and type-II-interferon-mediated signalling

A Positive Feedback Amplifier Circuit That 755

Stimulated Gene Expression and Controls Type I and Type II IFN Responses. 756 Frontiers in immunology

The ORF6, ORF8 and nucleocapsid proteins of SARS-CoV-2 inhibit type I interferon 759 signaling pathway

The Part and the Whole: functions of nucleoporins in nucleocytoplasmic 762 transport

Structure, dynamics and function of nuclear pore complexes

Nuclear import by karyopherin-betas: recognition and inhibition

Diversification of importin-α isoforms in cellular trafficking and disease 771 states

Nuclear transport is becoming crystal clear

Molecular basis for specificity of nuclear import and prediction of nuclear localization

Regulated nucleocytoplasmic transport during 779 gametogenesis

Importin α: a key molecule in nuclear transport and non-transport 782 functions

Importin α: a multipurpose nuclear-785 transport receptor

Evolution of the metazoan-specific importin α gene family

Extracellular signal-dependent nuclear

Regulated nuclear import of the STAT1 795 transcription factor by direct binding of importin-α

Importin α nuclear localization 798 signal binding sites for STAT1, STAT2, and influenza A virus nucleoprotein

Molecular basis for the recognition of 802 phosphorylated STAT1 by importin α5

A SARS-CoV-2 protein interaction map reveals targets for drug repurposing

The nonstructural protein 8 (nsp8) of the SARS coronavirus interacts with its ORF6 808 accessory protein

CD44 participates in IP-10 induction in cells in which hepatitis C virus RNA is replicating, 811 through an interaction with Toll-like receptor 2 and hyaluronan

A Rare Deletion in SARS-CoV-2 ORF6 Dramatically Alters the Predicted Three-815 Dimensional Structure of the Resultant Protein. bioRxiv : the preprint server for biology

SARS-CoV-2 ORF6 disrupts nucleocytoplasmic transport through interactions with 819 Rae1 and Nup98. bioRxiv : the preprint server for biology

Cellular stresses induce the nuclear accumulation of importin α and cause a 822 conventional nuclear import block

Multiple mechanisms promote the inhibition of classical 825 nuclear import upon exposure to severe oxidative stress

NF-κB is transported into the nucleus by 828 38 importin α3 and importin α4

Nuclear translocation of hypoxia-inducible factors (HIFs): involvement of the 831 classical importin alpha/beta pathway

XPO7 and IPO8 mediate the translocation ofNF-κB/p65 into the nucleus

Nucleocytoplasmic shuttling of STAT transcription factors

Unphosphorylated STAT1 prolongs the expression of interferon-induced immune 840 regulatory genes

IFNβ-dependent increases in STAT1, STAT2, and IRF9 mediate resistance to viruses 843 and DNA damage

Overexpression of SARS-CoV-2 protein ORF6 dislocates RAE1 and NUP98 from the 846 nuclear pore complex

Different mutations in SARS-CoV-2 associate with severe and mild 849 outcome

HIF-1α, and COVID-852 19: from pathogenic factors to potential therapeutic targets

Oxygen Sensing and Viral Replication: Implications for 855 Tropism and Pathogenesis

A novel method of preparing rat-monoclonal antibody-858 producing hybridomas by using rat medial iliac lymph node cells

Interferon-gamma-dependent nuclear 861 import of Stat1 is mediated by the GTPase activity of Ran/TC4

Functional characterization of importin α8 as a classical nuclear localization signal 864 receptor

Importin α can migrate into the nucleus in an importin -and Ran-independent 867 manner

Data on dimer formation between importin α subtypes

Genetic loss of importin α4 causes abnormal sperm morphology and impacts on 873 male fertility in mouse

A Simplified Quantitative Real-Time PCR Assay for Monitoring SARS-CoV-876 2

Improved vectors and genome-wide libraries for CRISPR screening

A novel bacterium-free method for generation of flavivirus infectious DNA by circular 882 polymerase extension reaction allows accurate recapitulation of viral heterogeneity

Establishment of a reverse genetics system for SARS-CoV-2 using circular polymerase 886 extension reaction. bioRxiv : the preprint server for biology

The GTEx Consortium atlas of genetic regulatory effects across human tissues

The GTEx Consortium atlas of genetic regulatory effects across human tissues

Figure 2. Inhibition of nuclear localization of PY-STAT1 by ORF6

Signal intensities of total 100 different nuclei from two independent experiments. ***P < 0.001, two-tailed Student's t-test. D. qRT-PCR analysis of IP-10 mRNA in AcGFP and AcGFP-ORF6

ORF6-M0: wild type ORF6; ORF6-M1: ORF6 with amino acids 49-52 substituted for alanine; ORF6-M2: ORF6 with amino acids 53-55 substituted for alanine; ORF6-M3: ORF6 with amino acids 56-61 substituted for alanine. B. Immunofluorescence of PY-STAT1 in HeLa cells transfected with the AcGFP-ORF6 mutants following IFN-γ stimulation. DAPI was used to stain the DNA. Scale bars: 30 μm. C. The graph represents the relative fluorescence values in the nucleus compared to those of the whole cells in B. Signal intensities of total 50 nuclei from two independent experiments were measured. ***P < 0.001, one-way ANOVA. D. Schematic representation of SARS-CoV-2 replicon DNA, pBAC-SCoV2-Rep. The genetic structure of the SARS-CoV-2 replicon is shown at the top of the panel. The dark shaded box indicates the core sequence of transcription regulating sequence

Rz, hepatitis delta virus ribozyme

Relative luciferase values for each ORF6 mutant in replicon (n=3). **P < 0.01, ***P < 0.001, one-way ANOVA

AcGFP-ORF6 WT or AcGFP-ORF6Δ9 (anti-GFP or anti-Flag antibodies were used). DAPI was used to stain the DNA. Scale bars: 30 μm. B. The graph represents the relative fluorescence values of Flag-STAT1 in the nucleus compared to those of the whole cells in A. Signal intensities from total 45 nuclei from two independent experiments. ***P < 0.001, oneway ANOVA. C. Immunofluorescence of Flag-STAT1 in

WT or AcGFP-ORF6Δ9 following IFN-γ stimulation. Anti-GFP or anti-Flag antibodies were used for detection

Input is 1/10th of the amount of STAT1 used for the reaction. G. GST-GFP or GST-GFP fused with the C-terminal peptide of ORF6 wild type (49-61 amino acids; GST-M0-GFP) immobilized on glutathione Sepharose beads was incubated with the STAT1 recombinant protein for 1 h. The bottom panel represents the proteins bound to the beads and stained with CBB. Input is 1/10th of the amount of STAT1 used for the reaction. Figure 5. ORF6 affects the subcellular localization of importin α proteins A-B. Immunofluorescence of Flag-importin α1 (Flag-Impα1) and Flag-importin α5 (Flag-Impα5) in HeLa cells transfected with AcGFP or AcGFP-ORF6. Anti-GFP or anti-Flag antibodies were used for detection. DAPI was used to stain the DNA. Scale bars: 30 μm. C. The graph represents the relative fluorescence values of the nucleus compared to those of the whole cells in A and B. Signal intensities of total 50 nuclei from two independent experiments were measured. ***P < 0.001, two-tailed Student's t-test. D. GST-GFP and GST-GFP-ORF6 were immobilized on glutathione Sepharose beads and incubated with bacterially purified Flag-importin α1 (Flag-Impα1) or Flag-importin α5 (Flag-Impα5) for 1 h. The importin α proteins were detected using an anti-Flag antibody. The bottom panel represents the proteins bound to the beads and stained with CBB. Inputs are 1/30th of the amount of each importin α that was used for the reaction. E. Immunofluorescence of Flag-importin α1 (Flag-Impα1) in HeLa cells transfected with AcGFP-ORF6 with or without hydrogen peroxide (200 μM H2O2) for 30 min. Anti-GFP or anti-Flag antibodies were used for detection

The authors declare no competing interests. Signal intensities of total 50 nuclei from two independent experiments were measured. ***P < 0.001, two-tailed Student's t-test. C. GST-GFP, GST-GFP-ORF6 or GST-NLS-GFP were immobilized on glutathione Sepharose beads and incubated with importin α1 (Impα1) for 1 h.The bottom panel represents the proteins bound to the beads and stained with CBB. Inputs are 1/10th of the amount of each importin α that was used for the reaction. D. An in vitro semi-intact nuclear transport assay was performed to measure the nuclear import of GST-NLS-mRFP in the presence of AcGFP-ORF6. Digitoninpermeabilized HeLa cells were incubated with GST-NLS-mRFP, importin α1, importin β1, RanGDP, p10/NTF2, GTP, and ATP regeneration system. The reaction mixture was added 5×, 10×, or 20× concentration of AcGFP-ORF6 compared to that of the NLS-substrate. After incubation for 30 min, the mRFP signals were detected using a fluorescence microscope. DAPI was used to stain the DNA. Scale bars: 30 μm. E. The graph represents the nuclear fluorescence values of GST-NLS-mRFP in D. Signal intensities of total 100 nuclei were measured and the statistically analyzed using a one-way ANOVA (***P < 0.001). F. Immunofluorescence of HIF-1α in HeLa cells transfected with AcGFP or AcGFP-ORF6 following CoCl2 treatment. Anti-GFP or anti-Flag antibodies were used for detection. DAPI was used to stain the DNA. Scale bars: 30 μm. G. The graph represents the relative fluorescence values of the nucleus compared to those of the entire cells in F. Signal intensities of total 50 nuclei from two independent experiments were measured. ***P < 0.001, two-tailed Student's t-test. H.Immunofluorescence of NF-κB p65 in HeLa cells transfected with AcGFP or AcGFP-ORF6 following TNF-α stimulation. Anti-GFP or anti-Flag antibodies were used for detection. DAPI was used to stain the DNA. Scale bars: 30 μm. I. The graph represents the relative fluorescence values of the nucleus compared to whole cells in H. Signal intensities of total 50 nuclei from two independent experiments were measured. ***P < 0.001, twotailed Student's t-test. J-K. Huh7 cells expressing the ACE2 receptor (Huh7-ACE2) introduced with sgControl (sgCtl) or sgKPNA2 were infected with SARS-CoV-2 and supernatants were collected at 0, 6, 12, and 24 h. Intracellular viral RNA was quantified using qRT-PCR (J) while the viral titers (K) were quantified using plaque forming assay. Statistical significance was determined using a two-way ANOVA (***P < 0.001).

GTEx donors whose estimated ancestry was EUR (n = 436 for lung tissues in A and B, and n = 558 for whole blood in C and D) were used. P-values for the trends between KPNA2 expression levels and age categories were obtained using the two-sided Jonckheere-Terpstra test. The box represented the first and third quartiles and the center line represented the median. The upper whisker extended from the hinge to the highest value that is within the 1.5 × IQR of the hinge, the lower whisker extended from the hinge to the lowest value within the 1.5 × IQR of the hinge, and the data beyond the end of the whiskers were plotted as points. F, female; M, male; IQR, interquartile range.