key: cord-103868-iwpiti2h authors: Harrison, Angela R.; Moseley, Gregory W. title: The Ebola virus interferon antagonist VP24 undergoes active nucleocytoplasmic trafficking date: 2020-08-11 journal: bioRxiv DOI: 10.1101/2020.08.10.245563 sha: doc_id: 103868 cord_uid: iwpiti2h Viral interferon (IFN) antagonist proteins mediate evasion of IFN-mediated innate immunity and are often multifunctional, having distinct roles in viral replication processes. Functions of the Ebola virus (EBOV) IFN antagonist VP24 include nucleocapsid assembly during cytoplasmic replication and inhibition of IFN-activated signalling by STAT1. For the latter, VP24 prevents STAT1 nuclear import via competitive binding to nuclear import receptors (karyopherins). Many viral proteins, including proteins from viruses with cytoplasmic replication cycles, interact with the trafficking machinery to undergo nucleocytoplasmic transport, with key roles in pathogenesis. Despite established karyopherin interaction, the nuclear trafficking profile of VP24 has not been investigated. We find that VP24 becomes strongly nuclear following overexpression of karyopherin or inhibition of nuclear export pathways. Molecular mapping indicates that cytoplasmic localisation of VP24 depends on a CRM1-dependent nuclear export sequence at the VP24 C-terminus. Nuclear export is not required for STAT1 antagonism, consistent with competitive karyopherin binding being the principal antagonistic mechanism while export mediates return of nuclear VP24 to the cytoplasm for replication functions. Thus, nuclear export of VP24 might provide novel targets for antiviral approaches. Importance Ebola virus (EBOV) is the causative agent of ongoing outbreaks of severe haemorrhagic fever with case-fatality rates between 40 and 60%. Proteins of many viruses with cytoplasmic replication cycles similar to EBOV interact with the nuclear trafficking machinery, resulting in active nucleocytoplasmic shuttling important to immune evasion and other intranuclear functions. However, exploitation of host trafficking machinery for nucleocytoplasmic transport by EBOV has not been directly examined. We find that the EBOV protein VP24 is actively trafficked between the nucleus and cytoplasm, and identify the specific pathways and sequences involved. The data indicate that nucleocytoplasmic trafficking is important for the multifunctional nature of VP24, which has critical roles in immune evasion and viral replication, identifying a new mechanism in infection by this highly lethal pathogen, and potential target for antivirals. transport, with key roles in pathogenesis. Despite established karyopherin interaction, the 23 nuclear trafficking profile of VP24 has not been investigated. We find that VP24 becomes 24 strongly nuclear following overexpression of karyopherin or inhibition of nuclear export 25 pathways. Molecular mapping indicates that cytoplasmic localisation of VP24 depends on a 26 CRM1-dependent nuclear export sequence at the VP24 C-terminus. Nuclear export is not 27 required for STAT1 antagonism, consistent with competitive karyopherin binding being the 28 principal antagonistic mechanism while export mediates return of nuclear VP24 to the 29 cytoplasm for replication functions. Thus, nuclear export of VP24 might provide novel targets 30 for antiviral approaches. HEK293T cells ( Figure S1 ). Thus, GFP-VP24 can localise into the nucleus in complexes with 149 K1, indicating that cytoplasmic localisation, which is required for roles in nucleocapsid 150 assembly/condensation (4, 5, 7), derives from active nuclear export. can diffuse through the NPC and lacks NLSs or NESs, was diffusely localised between the dependent NESs ( Figure 2A , Figure S2A ). Importantly, the Fn/c for GFP-VP241-251 in LMB-169 treated cells was higher than that for GFP alone ( Figure 2C ), indicative of accumulation. Thus, 170 VP24 localisation appears to be dynamic, involving nuclear entry and rapid nuclear export via 171 CRM1 interaction. 172 To determine which of the predicted NESs is/are responsible for nuclear export, we generated 175 constructs to express truncated VP24 proteins comprising N-terminal (VP241-88), central 176 (VP2489-172) and C-terminal (VP24173-251) portions fused to GFP; each of these contained one 177 or more of the potential NESs ( Figure 2A ). The truncated proteins were designed to be of 178 similar length and to avoid disruption of key structural elements (e.g. alpha helices and beta 179 sheets), based on the VP24 crystal structure (20). All proteins were predominantly cytoplasmic 180 at steady state ( Figure 2B ). Localisation of the N-terminal fragment was largely unaffected by 181 LMB treatment, and LMB produced only a small (≤ 1.4 fold) increase for the Fn/c of the central 182 fragment ( Figure 2B ,C). In contrast, a consistent and substantial increase (> 2 fold) in the Fn/c 183 for the C-terminal fragment was observed following LMB treatment. VP24173-251 also displayed 184 a consistently reduced Fn/c at steady state compared with the other truncated proteins. Thus, it 185 appeared that prominent discrete CRM1-dependent NES activity is located in the C-terminal 186 region of VP24. 187 9 Notably, only full-length VP24 displayed accumulation into the nucleus following LMB 189 treatment, with all truncated proteins remaining significantly less nuclear than GFP alone. This 190 suggests that the full protein sequence is required for efficient nuclear accumulation, such that 191 truncations remove key sequences or otherwise impact conformation to affect important 192 interactions. The crystal structure of VP24 bound to K5 indicates that three regions contact 193 the K (CL1 and CL2/3, separated by 40-60 residues, Figure 2A induced only a small increase in Fn/c for GFP-UL44NLS-VP2489-172 ( Figure 3A ,B), suggestive 209 of cytoplasmic retention or nuclear export mediated largely via an alternative mechanism to 210 CRM1-dependent export. However, LMB induced substantial nuclear localisation of GFP-211 UL44NLS-VP24173-251 (> 4.6 fold increase in Fn/c; Figure 3B ) that clearly exceeded nuclear 212 localisation of GFP-VP24173-251 ( Figure 2C ), consistent with a classical CRM1-dependent NES 213 counteracting the activity of the heterologous UL44 NLS. The Fn/c for GFP-VP2489-251 was 214 also markedly increased by LMB treatment but did not attain an Fn/c similar to that of full-215 required for efficient nuclear localisation. Nevertheless, these data clearly indicate that VP24 217 contains classical CRM1-dependent NES activity and that the principal NES is within VP24173-218 251. The C-terminal CRM1-dependent NES is the principal sequence mediating nuclear export of 221 VP24 222 To confirm that the C-terminal NES is the major sequence driving CRM1-dependent export of 223 VP24, we used site-directed mutagenesis to disable the NES motif. Analysis of the VP24 C-224 terminal region identified residues 241-251 (comprising the C-terminal 11 residues) as 225 containing a sequence strongly conforming to a NES ( To further examine effects of altered VP24 nuclear trafficking on STAT1 responses, we 263 assessed nuclear import of STAT1 using CLSM analysis of COS7 cells expressing GFP-VP24 264 and immunostained for STAT1 following treatment without or with IFN- and/or LMB. In 265 agreement with results of the luciferase reporter assays, we observed that despite substantial 266 re-localisation of GFP-VP24 to the nucleus in LMB-treated cells, IFN--dependent STAT1 267 nuclear localisation remained clearly inhibited ( Figure S3 ). Together, these data indicate that 268 nuclear export of VP24 is not required for inhibition of STAT1 responses, consistent with K 269 binding representing the major antagonistic mechanism. Thus, it appears that active In this study we have shown that EBOV VP24 undergoes active trafficking between the nucleus 275 and cytoplasm involving CRM1-dependent nuclear export via a NES at the VP24 C-terminus. 276 The acquisition of active nuclear trafficking sequences is consistent with a requirement for 277 highly regulated/dynamic localisation; furthermore, since VP24 is reported to oligomerise 278 (potentially as tetramers) (38), it is likely that active nuclear trafficking is required for transport 279 of VP24 multimers. The identified NES was not resolved in VP24 crystal structures (20, 47, 280 48) but localisation at the C-terminal end would be consistent with exposure and accessibility 281 to CRM1 (20), and the predominantly cytoplasmic localisation of GFP-VP24 in resting cells 282 suggests that the NES is the dominant trafficking signal at steady state. Intriguingly, previous 283 studies indicated that a mutated VP24 protein defective for K-binding was more cytoplasmic 284 than WT protein (49). This would be consistent with karyopherin binding mediating import; 285 one might thus speculate that VP24 would require export mechanisms to enable cytoplasmic 286 localisation/functions. Our findings are the first to confirm this is the case. Notably, the EBOV 287 matrix protein VP40 has also been reported to localise to the nucleus in infected and transfected 288 cells (16, 50); however, a direct role for active trafficking pathways to regulate localisation, 289 distinct from mechanisms such as diffusion or interaction with other host factors, has not been 290 defined. Thus, our data provides, to our knowledge, the first direct demonstration of a filovirus 291 protein exploiting specific host trafficking machinery for nucleocytoplasmic transport, 292 identifying a new mechanism in infection by these highly lethal pathogens. 293 294 Although the nucleus is not directly involved in the replication processes of most RNA viruses, 295 proteins of a number of these viruses are reported to encode nuclear trafficking sequences, 296 indicative of a requirement for dynamic regulation or specific accumulation in particular 297 compartments. For example, the RABV IFN antagonist P protein encodes several NLSs and 298 NESs (32, 43, 51-53), with regulatory mechanisms including co-localisation or overlap of the 299 sequences, enabling co-regulation by mechanisms including phosphorylation (51-53). 300 Although our data identify the C-terminal NES as a principal determinant of nucleocytoplasmic 301 localisation of full-length VP24, the differential localisation and LMB sensitivity of VP241-88 302 and VP2489-172, and the finding that VP2489-251 does not recapitulate nuclear accumulation of 303 full-length VP24, suggest the presence of alternative regulatory sequences/mechanisms, 304 potentially exposed by truncation. For example, VP24 is reported to associate with membranes 305 (38), which might result in tethering within the cytoplasm under certain conditions. 306 Interestingly, a recent study reported that sumoylation of residue K14 of VP24 enhances K 307 binding and IFN antagonistic function (54). In contrast, ubiquitination, including at residue 308 K206 within CL3 (Figure 2A ), appears to negatively regulate IFN antagonist activity (54). 309 Intriguingly, K14 is distal to CL1-3 but is within a predicted NES motif (Figure 2A) . Whether mechanisms will be of interest in defining the processes controlling immune evasion and 312 replication by EBOV. 313 While some viral IFN antagonists use NESs to facilitate immune evasion, including through 315 mislocalisation of associated STATs (33, 34), VP24 uses a mechanism of competitive binding 316 to Ks. Our finding that VP24 nuclear export is not required for STAT antagonism is consistent 317 with this, and indicates that export relates to cytoplasmic roles including in nucleocapsid 318 assembly and transport (4-14). The requirement for efficient translocation out of the nucleus is 319 consistent with interaction of VP24 with K (see above), that underpins distinct functions in 320 immune evasion. This is further supported by our finding that the C-terminal NES motif is 321 conserved among VP24 of several filovirus species that have been shown to bind to Ks or 322 have conserved CL sequences (20, 26, 45), but not in MABV VP24 ( Figure 4E , Figure S2C Thus, targeting VP24 regulatory mechanisms, including its nuclear export, may provide novel 353 targets for anti-EBOV drug design. 354 355 The construct to express the minimal NLS from human cytomegalovirus UL44 protein 358 (residues 425-433) fused to GFP was generated by subcloning from pEPI-GFP-UL44425-433 (42, 359 43) into the pEGFP-C1 vector C-terminal to GFP (Clontech). Constructs to express full-length 360 or truncated EBOV-VP24 protein fused to GFP or GFP-UL44NLS were generated by PCR to express FLAG-tagged K1 was a kind gift from C. Basler (Georgia State University). Other 366 constructs have been described elsewhere (40). Discovery of an Ebolavirus-Like Filovirus 446 in Europe Filoviruses: Ecology, Molecular Biology, and 448 Evolution Descriptive Analysis of Ebola Virus Proteins Infection of naive target cells with virus-like 453 particles: implications for the function of ebola virus VP24 Knockdown of Ebola virus VP24 impairs viral nucleocapsid assembly and 456 prevents virus replication The assembly of Ebola virus nucleocapsid 458 requires virion-associated proteins 35 and 24 and posttranslational modification of 54 Contribution of ebola virus 608 glycoprotein, nucleoprotein, and VP24 to budding of VP40 virus-like particles Both matrix proteins of Ebola 611 virus contribute to the regulation of viral genome replication and transcription Ebola virus 614 (EBOV) VP24 inhibits transcription and replication of the EBOV genome VP24 Is a Molecular Determinant of Ebola Virus Virulence 618 in Guinea Pigs A single phosphorodiamidate morpholino 621 oligomer targeting VP24 protects rhesus monkeys against lethal Ebola virus infection Fn/c) for GFP (mean ± SEM, n  52 637 cells for each condition; results are from a single assay representative of three independent 638 assays). Statistical analysis (Student's t-test) was performed using GraphPad Prism software A) Schematic of full-642 length VP24 and truncated VP24 proteins generated. Location of potential NESs are shown in 643 yellow. Location of clusters (CL1-3) of residues that interact with Ks in the VP24:Kα5 644 complex crystal structure (20) are shown in red. Numbering indicates residue positions in full-645 length VP24; sequences of potential NESs are shown above. (B) COS7 cells Representative images are shown. (C) Images such 648 as those shown in B were analysed to calculate the Fn/c for GFP (C; mean ± SEM; n  31 cells 649 for each condition Statistical analysis used Student's t-test. **** from a single assay representative of two independent assays). Statistical analysis used 657 Student's t-test The authors declare that they have no conflicts of interest with the contents of this article. 441