Nuclear MxA proteins form a complex with in ̄uenza virus NP and inhibit the transcription of the engineered in ̄uenza virus genome

Mx proteins belong to the dynamin superfamily of high molecular weight GTPases and interfere with multiplication of a wide variety of viruses. Earlier studies show that nuclear mouse Mx1 and human MxA designed to be localized in the nucleus inhibit the transcription step of the in ̄uenza virus genome. Here we set a transient in ̄uenza virus transcription system using luciferase as a reporter gene and cells expressing the three RNA polymerase subunits, PB1, PB2 and PA, and NP. We used this reporter assay system and nuclear-localized MxA proteins to get clues for elucidating the anti-in ̄uenza virus activity of MxA. Nuclear-localized VP16-MxA and MxA-TAg NLS strongly interfered with the in ̄uenza virus transcription. Over-expression of PB2 led to a slight resumption of the transcription inhibition by nuclear MxA, whereas over-expression of PB1 and PA did not affect the MxA activity. Of interest is that the inhibitory activity of the nuclear MxA was markedly neutralized by over-expression of NP. An NP devoid of its C-terminal region, but containing the N-terminal RNA binding domain, also neutralized the VP16-MxA activity in a dose-dependent manner, whereas an NP lacking the N-terminal region did not affect the VP16-MxA activity. Further, not only VP16-MxA but also the wild-type MxA was found to interact with NP in in ̄uenza virus-infected cells. This indicates that the nuclear MxA suppresses the in ̄uenza virus transcription by interacting with not only PB2 but also NP. INTRODUCTION Mx proteins are interferon-inducible and have antiviral activity against a wide variety of viruses (1±7). Studies on Mx proteins started with identi®cation of murine Mx1 in a mouse strain, A2G, resistant to in ̄uenza virus infection (8). Mx1 accumulates in the nucleus and inhibits the in ̄uenza virus transcription (9), thereby inhibiting in ̄uenza virus multiplication. Thereafter, a number of Mx proteins were identi®ed in higher eukaryotes including ®sh, birds and mammals as a homolog of murine Mx1 (10±16). MxA, a human homolog of Mx1, accumulates in the cytoplasm and interferes with multiplication of orthomyxoviruses (2,17,18), paramyxoviruses (19), bunyaviruses (3) and togavirus (4). When cells expressing MxA are infected with in ̄uenza virus, mRNA synthesis by primary transcription with virion-associated RNA polymerases is undergone at the same level as in MxA-negative cells (17). In contrast, viral protein synthesis and genome ampli®cation are strongly inhibited (17). However, the molecular function and exact target(s) of MxA in in ̄uenza virus-infected cells remain unknown. In Thogoto virus-infected cells, MxA interacts with viral nucleocapsids and inhibits transport of viral nucleocapsids into the nucleus (20,21). Recently it has been revealed that MxA inhibits multiplication of hepatitis B virus, a DNA virus (6). In HBV-infected cells, export of viral mRNA from the nucleus is blocked by MxA. These two cases suggest that MxA may interfere with translocation of viral components between the nucleus and the cytoplasm. To approach the molecular mechanism of the anti-in ̄uenza virus activity of MxA, two lines of evidence by earlier studies could be useful clues: MxA designed to be present in the nucleus inhibits the transcription of the in ̄uenza virus genome as Mx1 (22), and the anti-in ̄uenza virus activity of Mx1 is suppressed by over-expression of PB2 (23,24). In order to get more clues for the anti-in ̄uenza virus action of MxA and to look for a candidate(s) for a viral target molecule(s) of *To whom correspondence should be addressed. Tel: +81 298 53 3233; Fax: +81 298 53 3134; Email: [email protected] The authors wish it to be known that, in their opinion, the ®rst two authors should be regarded as joint First Authors Nucleic Acids Research, 2004, Vol. 32, No. 2 643±652 DOI: 10.1093/nar/gkh192 Nucleic Acids Research, Vol. 32 No. 2 ã Oxford University Press 2004; all rights reserved Published online January 29, 2004 at Penylvania State U niersity on Feruary 3, 2013 http://narrdjournals.org/ D ow nladed from MxA, we tried to design experiments based on the above two facts. We set a reporter assay system, which utilized plasmids used for the DNA transfection-mediated virus-like particle generation system (25). We constructed a host cell RNA polymerase I (Pol I)-based plasmid, from which an engineered in ̄uenza virus genome containing a reporter gene of negative sense is generated. When the plasmid is introduced into cells expressing the viral RNA polymerase and NP, the engineered viral RNA generated by Pol I is transcribed. With this system, we found that nuclear MxA interferes with expression of the reporter gene. This inhibition was neutralized by overexpression of PB2 but not by that of PB1 or PA. Interestingly, over-expression of NP led to a signi®cant suppression of the inhibitory activity of nuclear MxA. Immunoprecipitation assays revealed that nuclear MxA and NP form a complex under certain conditions. These results suggest that nuclear MxA proteins interfere with the viral transcription directly or indirectly through NP. Further, we found that NP is also immunoprecipitated with the wild-type MxA when lysates prepared from in ̄uenza virus-infected cells were used. A possible function of MxA in the context of its interaction with NP is also discussed. MATERIALS AND METHODS Construction of plasmid vectors A plasmid vector, from which an arti®cial in ̄uenza virus genome containing luciferase gene of negative polarity as a reporter is synthesized in cells by the mouse DNA-dependent Pol I, was constructed as follows. The mouse Pol I promoter region was ampli®ed by PCR with primers, 5¢-TAATACGACTCACTATA-3¢ and 5¢-GTCGGTACCTATCTCCAGGTCCA-3¢ using pMrBKSP11 (a gift from Dr K. Yamamoto) (26) as a template. PCR products were phosphorylated with T4 polynucleotide kinase (TOYOBO) and digested with KpnI (TaKaRa). This fragment was cloned into pHH21 containing the promoter region of human ribosomal RNA gene (a gift from Dr Y. Kawaoka) (25,27), which had been digested with BssHII followed by treatment with Klenow fragment (TaKaRa) and subsequent digestion with KpnI. Thus, the treatment described above removed the human promoter in pHH21 and then replaced it with the mouse promoter. The resultant plasmid was designated pHMP1. Next, a fragment containing luciferase gene sandwiched by 5¢and 3¢-terminal sequences of in ̄uenza A virus (WSN/33) segment 8 encoding non-structural protein (NS) was ampli®ed by PCR with primers, 5¢-GTAGTAGAAACAAGGGTGTTTTTTACTCGAGATCTTACAATTTGGACTTTCCGCCCTT-3¢ and 5¢-GATCCGTCTCCGGGAGCAAAAGCAGGGTGACAAAGACATAATGCATATGGAAGACGCCAAAAACATAAAGAAAGG-3¢ using pGV-B (TOYO Ink) as a template. pHMP1 was digested with KpnI, blunted with T4 DNA polymerase (TOYOBO) and de-phosphorylated with calf intestine alkaline phosphatase (TaKaRa). The ampli®ed PCR product was phosphorylated with T4 polynucleotide kinase and cloned into pHMP1 treated as above, resulting in construction of pHMP1-pre-vNS-Luc. Then, pHMP1-prevNS-Luc was digested with BsmB1 and re-ligated to produce pHMP1-vNS-Luc, in which the luciferase gene of reverse orientation sandwiched with 23 and 26 nucleotide-long 5¢and 3¢-terminal sequences of the in ̄uenza virus segment 8 is placed under the control of mouse Pol I promoter (Fig. 1A). A plasmid encoding the wild-type MxA protein was constructed by subcloning of the full-length MxA fragment into a eukaryotic expression vector, pCAGGS (28). The MxA fragment prepared by digestion of pET3a-MxA (a gift from Dr Staeheli) with NdeI (BioLabs) and BamHI (TOYOBO) was once cloned into NdeIand BamHI-digested pET14b (Novagen). The MxA fragment derived from pET14b-MxA by digestion with XbaI (TOYOBO) and EcoRV (TOYOBO) and blunted with Klenow fragment. Resultant DNA fragment was cloned into pCAGGS digested by XhoI (TOYOBO) and blunted with Klenow fragment. To construct a plasmid for expression of MxA fused to the trans-activation domain of herpes simplex virus VP16, pVP16 (Clontech), a plasmid for mammalian two-hybrid system was used. A fragment containing MxA gene was prepared from pET14b-MxA by digestion with NdeI followed by treatment with Klenow fragment and digestion with BamHI. The puri®ed MxA fragment was cloned into pVP16 that had been digested with EcoRI (TOYOBO) and blunted with Klenow fragment followed by BamHI digestion. For expression of MxA tagged with the nuclear localization signal (NLS) sequence of SV40 T antigen, pHMG-TMxA was used (a gift from Dr Staeheli). In order to construct pCHA-Mx1 encoding HA-tagged mouse Mx1, Mx1 cDNA was cloned into pCHA (29). The Mx1 cDNA portion was ampli®ed by using KOD-plus (TOYOBO) with a previously constructed plasmid containing Mx1 cDNA (30) as a template and speci®c primers, 5¢-TAGGCTAGCATGGATTCTGTGAATAATCTGTGC-3¢ and 5¢-TGAGCTAGCTTAATCGGAGAATTTGGCAAGCTT-3¢. The PCR-ampli®ed Mx1 fragment was digested with NheI (TOYOBO) and cloned into pCHA digested with NheI. pCHA-Mx1DC plasmid expressing Mx1 lacking its C-terminal (between amino acid positions 563 and 631) was spontaneously generated during the construction of pCHAMx1. Plasmids encoding NP (pCAGGS-NP) and three subunits of the in ̄uenza virus RNA polymerase, PB1, PB2 and PA (pcDNA-PB1, pcDNA-PB2 and pcDNA-PA) were gifts from Dr Y. Kawaoka (25). In order to construct pCAGGSNP-Myc for expression of Myc epitope-tagged NP, DNA fragments were ampli®ed by PCR with primers, 5¢CCGAATTCCATGGCGTCTCAAGGCACCAAA-3¢ and 5¢CGCGCCATGGCATTATCGTATTCCTCTGCA-3¢ and a plasmid containing the NP gene as a template. The ampli®ed fragment was digested with NcoI and ligated into NcoIdigested pBluescript II vector (Stratagene) containing a Myc tag sequence. Subsequently, a DNA fragment containing Myctagged NP sequence was excised from the pBluescript II and inserted into pCAGGS that had been digested with EcoRI and blunted with Klenow fragment. pSEAP2-control (Clontech) containing a secreted alkaline phosphatase (SEAP) gene under the control of the SV40 early promoter was used as a control for normalization of transfection. The nucleotide sequence of each plasmid was con®rmed by DNA sequencing. Cells and transfection Swiss mouse 3T3 cell lines, Swiss3T3-Neo (Swiss3T3) and Swiss3T3-MxA cells (31), were kindly provided by Dr Haller. Either mouse Swiss3T3 cells or Clone 76 cells [a gift from 644 Nucleic Acids Research, 2004, Vol. 32, No. 2 at Penylvania State U niersity on Feruary 3, 2013 http://narrdjournals.org/ D ow nladed from