(C) PLOS One This story was originally published by PLOS One and is unaltered. . . . . . . . . . . TREX (transcription/export)-NP complex exerts a dual effect on regulating polymerase activity and replication of influenza A virus [1] ['Lingcai Zhao', 'Moe Joint International Research Laboratory Of Animal Health', 'Food Safety', 'Engineering Laboratory Of Animal Immunity Of Jiangsu Province', 'College Of Veterinary Medicine', 'Nanjing Agricultural University', 'Nanjing', 'Qingzheng Liu', 'Jingjin Huang', 'Yuanlu Lu'] Date: 2022-11 Influenza A viruses effectively hijack the intracellular "resources" to complete transcription and replication, which involve extensive interactions between the viral and host proteins. Herein, we screened the host factors, which belong to DExD/H-box protein family members, RNA-binding proteins or mitochondrial anchoring proteins, to investigate their effects on polymerase activity. We observed DDX39B and DDX39A, DEAD-box RNA-Helicases, exert a dual effect on regulating polymerase activity and replication of influenza A viruses. We further revealed that DDX39B and DDX39A interact with viral NP and NS1 proteins. Interestingly, the viral NP proteins could reverse the inhibitory effect of excess DDX39B or DDX39A on polymerase activity. Mechanistically, the TREX complex subunits, THOC1, THOC4 and CIP29, were recruited to DDX39B-DDX39A-NP complex in an ATP-dependent manner, via the interaction with DDX39B or DDX39A, followed by excess TREX-NP complexes interfere with the normal oligomerization state of NP depending on the ratio between the viral and host proteins. On the other hand, the TREX complex, an evolutionarily conserved protein complex, is responsible for the integration of several mRNA processing steps to export viral mRNA. Knockdown of TREX complex subunits significantly down-regulated viral titers and protein levels, accompanied by retention of viral mRNA in the nucleus. Taken together, screening the host factors that regulate the replication of influenza virus advances our understanding of viral pathogenesis and our findings point out a previously unclear mechanism of TREX complex function. In this study, we investigated the regulation of polymerase activity by host factors associated with vRNPs (PB2 627 E, PB2 627 K, PB2 627 domain del, PB2 627 CON) and provided novel insights into regulatory mechanisms of DDX39B and DDX39A during viral replication. Our results demonstrated that DDX39B and its paralog DDX39A inhibited polymerase activity via forming TREX-NP complex with concomitant effects on the oligomeric state of NP proteins. Moreover, TREX complex is necessary for expression of viral proteins. Our findings provided potential therapeutic targets for dealing with IAV infection. Funding: This work was supported by National Key Research and Development Program of China [Grant number: 2021YFD1800205 (JP)], National Natural Science Foundation of China [Grant number: 31772775 (JP); 32272992 (JP)] and the State Key Laboratory of Veterinary Biotechnology [Grant number: SKLVBF202103 (JP)]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Copyright: © 2022 Zhao et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Protein-coding genes are transcribed as a pre-mRNA in the nucleus and pre-mRNA undergo several RNA processing steps. The TREX (transcription/export) complex is responsible for the integration of the several mRNA processing steps to export mRNA and is recruited to mRNA in a splicing-dependent manner [ 22 ]. The assembly of the TREX complex is dynamically associated with the ATPase cycle of DDX39B and acts as a binding platform for NXF1, and DDX39B is removed from the TREX complex during NXF1 recruitment [ 23 – 25 ]. Our work revealed DDX39B and DDX39A played a dual role in regulating influenza virus polymerase activity. We found that DDX39B and its paralog DDX39A play similar role in the replication of influenza virus. Further investigation of the molecular mechanism demonstrated that DDX39B and DDX39A rely on ATP binding activity to recruit TREX complex subunits THOC1, THOC4 and CIP29, and then rely on the binding of RNA to NP. Excessive DDX39B or DDX39A was not conducive to polymerase activity of influenza virus. Meanwhile, we demonstrated that DDX39B, DDX39A and other TREX complex subunits are necessary factors for expression of viral proteins, due to their involvement in mRNA transport. In this study, we reconstructed the influenza virus polymerase complex in vitro, and detected the interacting proteins through immune-precipitation (IP) and mass spectrometry. We verified associated DExD/H-box protein family members, RNA-binding proteins, and mitochondrial anchoring proteins via viral microgenome polymerase activity experiments. And we first discovered that DDX39B, DDX39A and other RNA binding proteins regulate polymerase activity of influenza virus, thereby regulating replication of influenza virus. DExD/H-box helicases are further distinguished based on the amino acid sequence of the eponymous conserved helicase motif II (DEAD, DEAH, DExH and DExD helicases), which are involved in various RNA metabolic processes, including transcription, translation, RNA splicing, RNA transport, and RNA degradation [ 15 ]. There is evidence that lots of DExD/H-box helicases have also been identified as essential host factors for the replication of different viruses, suggesting that viruses "hijack" RNA helicase activities for their benefit [ 16 , 17 ]. Influenza virus recruits cellular RNA helicase eIF4A3 to promote viral mRNA splicing and spliced mRNA nuclear export [ 18 ]. In an interesting twist, DExD/H-box helicases also engage in anti-viral immunity, such as RIG-like helicases (RIG-I), MDA5, act as important cytosolic pattern recognition receptors for viral RNA. It is well established that viral single-stranded or short double stranded RNA bearing a 5’-triphosphate is recognized by the cytosolic RNA helicase RIG-I [ 19 – 21 ]. Multiple studies have sought to identify host factors related to influenza virus replication [ 6 – 9 ], in particular those that may account for the species-specificity of PB2-627K [ 10 , 11 ]. The human and avian versions of the host protein ANP32A are now known to play an important role in the host-range restriction of influenza viruses: Briefly, ANP32A from avian origin could remarkably promote the replication of avian influenza virus (PB2 627 E) [ 12 – 14 ]. In addition, multiple other host factors are believed to interact with the viral polymerase complex to regulate its (host-specific) activity. The influenza viral polymerase complex is composed of three subunits (PA, PB1, PB2) which, together with the NP nucleoprotein, transcribes viral RNA (vRNA) into mRNA and replicates it through complementary RNA (cRNA) replication intermediates. The so-called 627 domain of PB2 forms a unique structure that protrudes from the polymerase core [ 1 ]. At amino acid position 627 of PB2, most human influenza viruses encode lysine (K), while most avian influenza viruses encode glutamic acid (E) [ 2 ]. The E-to-K mutation at PB2-627 considerably enhances the polymerase activity of avian influenza viruses at low temperature in mammalian cells, resulting in more effective virus replication in the upper respiratory tract of mammals (33–35°C) [ 3 – 5 ]. Collectively, the PB2-E627K mutation is now recognized as a strong determinant of host range restriction that enables avian influenza viruses to replicate efficiently in mammalian cells. Viral pathogens including influenza, corona, and Ebola viruses can cause high morbidity and mortality with tremendous consequences for human health and the global economy. Among them, influenza virus with its wide host range is one of the most destructive pathogens that threaten the livestock industry, and the general public through annual epidemics and sporadic pandemics, such as the pandemics of 1918, 1957, 1968, and 2009. Results Effect of human and avian host factors on vRNP activity Next, we asked if selected DExD/H-box protein family members, RNA-binding proteins, and mitochondrial anchor proteins affect influenza vRNP activity. Briefly, human 293T cells were transfected with plasmids over-expressing the indicated host protein (Fig 2). The transfected cells were co-transfected with plasmids encoding PB2 627 E- or PB2 627 K-vRNPs (Fig 2A). In the minireplicon assay (Fig 2A), overexpression of chicken ANP32A protein (which is known to increase the replicative ability of avian-type polymerase complexes in mammalian cells [14]) indeed strongly upregulated reporter protein expression by PB2 627 E-vRNP, but not by PB2 627 K-vRNP. Overexpression of several human host factors (most prominently, DDX39B, a member of the DExD/H box family of ATP-dependent RNA-helicases which is involved in mRNA splicing and nuclear export processes) reduced reporter protein expression from the influenza vRNP in the minireplicon assay. Overall, overexpression of several DExD/H-box protein family members (i.e., DDX39, DDX24), mitochondrial anchoring proteins (i.e., DNAJA3, ATP5B), and RNA-binding proteins (i.e., SRSF3, RBMX, RBM10, RBMS1, NH2L1) affected reporter protein expression from influenza vRNPs. PPT PowerPoint slide PNG larger image TIFF original image Download: Fig 2. Screening host factors that regulate influenza virus polymerase activity. (A) Host proteins were selected to analyze their roles in polymerase activity (PB2 627 E or PB2 627 K) of influenza virus H7N9 (A/Anhui/1/2013). The log2 logarithmic values of the ratio of the luciferase activity transfected with the corresponding host factors and the empty pCAGGS plasmids were used as the PB2 627 E or PB2 627 K polymerase activity. (B) Host proteins from avian were selected to analyze their roles in polymerase activity (PB2 627 E) in HEK293T cells. RBM10-ck v1 and RBM10-ck v2, and RBMX-ck v2 and RBMX-ck v3 are two different transcripts of avian-derived RBM10 and RBMX, respectively. The detailed base sequences are in the attached S3 Table. Statistical differences between groups are labeled according to a one-way ANOVA followed by a Dunnett’s test. Each treatment was repeated three times in parallel. The results are presented as means ± standard deviations. *, P < 0.05; ***, P < 0.001; ****, P < 0.0001; ns, no significance. ck, chicken; hu, human. https://doi.org/10.1371/journal.ppat.1010835.g002 Previous studies established that host proteins of human or avian origin can differently affect the activity of viruses encoding PB2 627 K or PB2 627 E, respectively [26]. We, therefore, compared the effect of selected human (-hu) and chicken (-ck) RNA-binding host proteins (including splice variants of avian host proteins) on reporter protein expression in minireplicon assays (Fig 2B). Overexpression of SRSF3-hu or SRSF3-ck (serine- and arginine-rich splicing factor 3) increased the reporter protein expression levels in human cells, which is consistent with the results of our previous study [27]. On the other hand, overexpression of RBMS1-hu and RBMS1-ck reduced the reporter protein levels expressed from PB2 627 E-vRNPs (Fig 2A and 2B). This pattern indicated species-specific differences in host factor interaction with the influenza virus replication complex. Effect of human and avian DDX39B or DDX39A on vRNP activity DDX39B (also known as UAP56) is known to interact with the influenza virus NP protein [28]. However, the molecular mechanism by which DDX39B regulates viral replication remains unclear. We found that in human 293T and chicken DF-1 cells, overexpression of DDX39B-hu that was only detected in immunoprecipitates of PB2 627 E-vRNPs, DDX39A-hu (a mammalian paralog of DDX39B [29,30] which was not detected in immunoprecipitates of vRNPs (see S1 Table)), or DDX39B-ck reduced the reporter protein expression levels from vRNPs (Fig 3A–3C), consistent with the data shown in Fig 2A. Increasing amounts of DDX39B/A tended to have a stronger inhibitory on the levels of reporter protein expression from vRNPs (Fig 3D), while overexpression of NP reversed the inhibitory effect of DDX39B/A (Fig 3E). Based on these data, we reasoned that siRNA-mediated knock-down of DDX39B and DDX39A should upregulate reporter protein expression from vRNPs in minireplicon assays, but detected the opposite effect (Fig 3F). Thus, both overexpression and down-regulation of DDX39B and DDX39A reduced reporter protein expression from influenza vRNPs in vitro, suggesting that the level of DDX39B may be important for efficient reporter protein expression in minireplicons. PPT PowerPoint slide PNG larger image TIFF original image Download: Fig 3. Effect of human and avian DDX39B or DDX39A on vRNP activity. (A, B and C) Polymerase genes (PB2 627 E) from either influenza virus H1N1 (A/WSN/1933) (A) or H9N2 (A/Chicken/Anhui/LH99/2017) (B), poll-Luc or pollck-Luc, RL-TK and corresponding plasmids encoding HA-DDX39B-hu, HA-DDX39A-hu, HA-DDX39B-ck or empty pCAGGS vectors were co-transfected into HEK293T cells (A and B) or DF-1 cells (C), 48h after transfection, the polymerase activity was detected. (D) Above polymerase genes (PB2 627 E), poll-Luc, RL-TK and gradually increased dose of HA-DDX39B or HA-DDX39A were co-transfected into HEK293T cells, 48h after transfection, the polymerase activity was detected. (E) Above polymerase genes (PB2 627 E), poll-Luc, RL-TK and gradually increased dose of NP of influenza virus H7N9 (A/Anhui/1/2013) or H1N1 (A/WSN/1933), 48h after transfection, the polymerase activity was detected. (F) Detecting the effect of siRNA-mediated knockdown of DDX39B or DDX39A on polymerase activity. (G) HA-DDX39B, HA-DDX39A or empty pCAGGS vectors were transfected into HEK293T cells, 24 hours after transfection, cells were infected with the recombinant influenza viruses WSN-H7 (627K) or WSN-H7 (627E) at MOI = 0.05. The supernatants were sampled at 24h post infection. (H) HEK293T cell lines with shRNA-mediated knockdown of DDX39B or DDX39A were infected with recombinant influenza viruses WSN-H7 (627K) or WSN-H7 (627E) at MOI = 0.05. The supernatants were sampled at 24h post infection. (I) HEK293T cells with siRNA-mediated knockdown of DDX39B and DDX39A, cells were infected with the recombinant influenza viruses WSN-H7 (627E) at MOI = 0.05. The supernatants were sampled at 24h post infection. (J) HA-DDX39B or HA-DDX39A was transfected into HEK293T cells with siRNA-mediated knockdown of DDX39A and DDX39B, respectively. 24 hours after transfection, cells were infected with the recombinant influenza viruses WSN-H7 (627E) at MOI = 0.05. The supernatants were sampled at 24h post infection. The protein levels were measured by western blotting and viral titers were measured by plaque assay. Statistical differences between groups are labeled according to a one-way or two-way ANOVA. The results are presented as means ± standard deviations. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, no significance. ck, chicken; hu, human. https://doi.org/10.1371/journal.ppat.1010835.g003 DDX39B and DDX39A expression levels affect influenza virus replication Next, we tested the effect of DDX39B and DDX39A overexpression on influenza virus titers in infected cells. Over-expression of either DDX39B or DDX39A hardly influenced the titers of viruses encoding PB2 627 K or PB2 627 E (Fig 3G). These data suggested that DDX39B and DDX39A may exert effects in multiple processes of virus replication, or the virus adopts certain strategies to suppress their inhibitory effects. shRNA-mediated down-regulation of DDX39B or DDX39A reduced virus titers in infected cells (Fig 3H); this inhibitory effect was further enhanced when both DDX39B and DDX39A were down-regulated (Fig 3I). Moreover, the siRNA-mediated reduction of virus titers was offset by overexpression of the DDX39 paralog (Fig 3J). Collectively, our data indicate that DDX39B and DDX39A are essential host factors for efficient viral replication, and their inhibitory effects on polymerase activity may depend on the ratio between viral and host proteins. DDX39B and DDX39A expression levels and intracellular localization are not affected by influenza virus infection To identify the molecular mechanisms through which DDX39B and DDX39A affect the influenza virus live cycle, we first assessed their expression levels and intracellular localization. In four different mammalian and avian cell lines, the expression levels of DDX39B and DDX39A vary widely (Fig 4A), in line with RNA-seq tissue data from public databases [31]. GFP- and mCherry-tagged DDX39B and DDX39A fusion proteins, respectively, are primarily expressed in the nucleus (Figs 4B, 4C, and S2A). Infection of human A549 cells with four different influenza viruses expressing PB2 627 K or PB2 627 E had no significant effect on the expression levels (Fig 4D) and intracellular localization (S2B and S2C Fig) of DDX39B and DDX39A. PPT PowerPoint slide PNG larger image TIFF original image Download: Fig 4. DDX39B and DDX39A inhibit polymerase activity depending on ATP binding and hydrolytic activity. (A) The protein levels of DDX39B and DDX39A in HEK293T, A549, Vero and DF-1 cells were measured by western blotting. (B) DDX39B-GFP or DDX39A-mCherry was transfected into HEK293T cells, 24 hours after transfection, the cells were then fixed, permeabilized. Cells were stained with the 4, 6-diamidino-2-phenylindole (DAPI) and were examined via fluorescence microscopy. (C) DDX39B-GFP and DDX39A-mCherry were co-transfected into HEK293T cells, 24 hours after transfection, the cells were treated the same as in (B). (D) A549 cells were infected with influenza virus WSN/1933 627K, WSN/1933 627E or recombinant influenza virus WSN-H7 (627K), WSN-H7 (627E) at MOI = 0.5. At 12h, 24h after infection, the cells were lysed and protein levels were measured by western blotting. (E) Polymerase genes (PB2 627 E), poll-Luc, RL-TK and corresponding HA-tagged mutants of DDX39B or DDX39A or empty pCAGGS vectors were co-transfected into HEK293T cells, 48h after transfection, the polymerase activity is detected. Statistical differences between groups are labeled according to two-way ANOVA. The results are presented as means ± standard deviations. **, P < 0.01; ****, P < 0.0001; ns, no significance. https://doi.org/10.1371/journal.ppat.1010835.g004 DDX39B and DDX39A mutants no longer reduce influenza virus polymerase activity in minireplicon assays DDX39B is an essential splicing factor which is required for the U2 snRNP-branchpoint interaction during pre-spliceosome assembly, and its ATP-binding, ATPase, and double-stranded (ds) RNA unwinding/helicase activities are essential for in vitro pre-mRNA splicing [32]. To determine if these functions are important for the regulation of influenza virus polymerase activity, we constructed DDX39B mutants lacking ATP-binding (K95A), ATPase activity (E197A), or dsRNA unwinding/helicase activity (D199A) (S3 Fig) [32–34]. Compared with wild-type DDX39B and DDX39A, the K95A and E197A mutants lost the ability to reduce reporter protein expression from influenza vRNPs in minireplicon assays (Fig 4E), demonstrating that the ATP-binding and ATPase activities of DDX39B and DDX39A are needed to suppress reporter protein expression in minireplicon assays. DDX39B and DDX39A interact with the influenza virus NP and NS1 proteins DDX39B co-precipitated with vRNP complexes (S1 Table). DDX39B is known to interact with influenza virus NP, but has not been tested for interaction with the other components of the viral replication complex. We, therefore, performed co-immunoprecipitation studies with the three viral polymerase proteins and NP, but only the latter interacted with the host factors (Fig 5A–5E). The interaction of DDX39B and DDX39A with NP was also detected in the context of virus-infected cells (Fig 5F), in line with the co-localization of these proteins in virus-infected cells (Fig 5G). PPT PowerPoint slide PNG larger image TIFF original image Download: Fig 5. DDX39B and DDX39A interact with NP and NS1 protein. (A, B, C, D and E) HEK293T cells were co-transfected with HA-DDX39B, HA-DDX39A or HA-DDX39B-ck and FlAG-PB2 (A), FlAG-PA (B), FlAG-PB1 (C), FlAG-NP (D) of influenza virus A/Anhui/1/2013 (H7N9) or FlAG-NP of influenza virus A/WSN/1933 (H1N1) (E), at 48h after transfection, the cells were lysed, followed by co-IP with anti-HA mouse Mab and western blotting using anti-FlAG mouse Mab and anti-HA mouse MAb. (F) HEK293T cells were transfected with HA-DDX39B or HA-DDX39A, at 24h after transfection, the cells were infected with recombinant influenza virus WSN-H7 627E at MOI = 1. At 24h after infection, the cells were lysed, followed by co-IP with anti-HA mouse Mab and western blotting using anti-NP mouse Mab and anti-HA mouse MAb. (G) DDX39B-mCherry or DDX39A-mCherry together with NP were transfected into HEK293T cells, 24 hours after transfection, the cells were fixed and examined via fluorescence microscopy. (H) HEK293T cells were co-transfected with HA-DDX39B, HA-DDX39A or HA-DDX39B-ck and FLAG-NS1 of influenza virus A/WSN/1933 (H1N1), at 48h after transfection, the cells were lysed, followed by co-IP with anti-HA mouse Mab and western blotting using anti-FlAG mouse Mab and anti-HA mouse MAb. https://doi.org/10.1371/journal.ppat.1010835.g005 We also tested the remaining influenza viral proteins for an interaction with DDX39B or DDX39A and found that the viral interferon-antagonist NS1 protein co-precipitated with both of these host factors (Fig 5H), consistent with a report that these proteins co-localize [35]. However, overexpression of NS1 had no significant effect on DDX39B’s ability to down-regulate reporter protein expression from vRNPs in minireplicon assays (S4A Fig). The other viral proteins tested (i.e., the M1 matrix protein, the M2 ion channel protein, and the NS2/NEP nuclear export protein) did not interact with DDX39B and DDX39A (S4B–S4D Fig) and their overexpression did not affect the DDX39A-mediated reduction of reporter protein expression from vRNPs in minireplicon assays (S4A Fig). [END] --- [1] Url: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1010835 Published and (C) by PLOS One Content appears here under this condition or license: Creative Commons - Attribution BY 4.0. via Magical.Fish Gopher News Feeds: gopher://magical.fish/1/feeds/news/plosone/