Search

It is generally thought that pathogen-associated molecular patterns (PAMPs) responsible for triggering interferon (IFN) induction are produced during virus replication and, to limit the activation of the IFN response by these PAMPs, viruses encode antagonists of IFN induction. Here we have studied the induction of IFN by parainfluenza virus type 5 (PIV5) at the single-cell level, using a cell line expressing GFP under the control of the IFN-β promoter. We demonstrate that a recombinant PIV5 (termed PIV5-VΔC) that lacks a functional V protein (the viral IFN antagonist) does not activate the IFN-β promoter in the majority of infected cells. We conclude that viral PAMPs capable of activating the IFN induction cascade are not produced or exposed during the normal replication cycle of PIV5, and suggest instead that defective viruses are primarily responsible for inducing IFN during PIV5 infection in this system., (Copyright © 2011 Elsevier Inc. All rights reserved.)

The infection of cells by RNA viruses is associated with the recognition of virus PAMPs (pathogen-associated molecular patterns) and the production of type I interferon (IFN). To counter this, most, if not all, RNA viruses encode antagonists of the IFN system. Here we present data on the dynamics of IFN production and response during developing infections by paramyxoviruses, influenza A virus and bunyamwera virus. We show that only a limited number of infected cells are responsible for the production of IFN, and that this heterocellular production is a feature of the infecting virus as opposed to an intrinsic property of the cells., (Copyright © 2010 Elsevier Inc. All rights reserved.)

A panel of influenza A viruses encoding mutant NS1 proteins was created in which a number of NS1 functions, including interactions with dsRNA, PI3K, CPSF30 and PKR, were inhibited. Surprisingly, given previous reports that NS1 activates PI3K to prevent apoptosis, the mutant viruses rUd-Y89F and rUd-P164/7A that fail to activate PI3K did not induce any more apoptosis than wild-type virus in MRC-5 and A549 cells, even though these cells are highly sensitive to inducers of apoptosis. Induction of cell death by the apoptogenic rUd-184-8(P) virus could not be prevented by serum-mediated activation of PI3K/Akt. Neither infection of MRC-5 or A549 cells with wild-type virus nor constitutive expression of NS1 prevented cell death caused by apoptosis inducers, suggesting that NS1 is not directly anti-apoptotic. Our data suggest that the loss of a functionally intact NS1 protein promotes apoptosis, but this is not due to an inability to activate PI3K.

Posttranslational modification of viral proteins by cellular enzymes is a feature of many virus replication strategies. Here, we report that during infection the multifunctional human influenza A virus NS1 protein is phosphorylated at threonine-215. Substitution of alanine for threonine at this position reduced early viral propagation, an effect apparently unrelated to NS1 antagonizing host interferon responses or activating phosphoinositide 3-kinase signaling. In vitro, a subset of cellular proline-directed kinases, including cyclin dependent kinases (CDKs) and extracellular signal-regulated kinases (ERKs), potently phosphorylated NS1 protein at threonine-215. Our data suggest that CDK/ERK-mediated phosphorylation of NS1 at threonine-215 is important for efficient virus replication.

Influenza A virus NS1 protein is a multifunctional virulence factor. Here, we report a crystal structure for the NS1 effector domain of avian influenza virus A/Duck/Albany/76. Comparison of this structure with that reported for a human strain shows both proteins share a common monomer conformation, albeit with subtle differences. Strikingly, our data reveal a novel helix-helix dimeric interface between monomers of the avian NS1 protein, which is also found in the human NS1 crystal lattice. We re-evaluate the current model of NS1 dimeric assembly, and provide biochemical evidence to show tryptophan-187 (a residue located at the helix-helix interface) is essential for dimerization of this effector domain.

A dynamic model of STAT1 degradation by the V protein of parainfluenza virus 5 (PIV5; formerly SV5) has been proposed. In it, the V protein functions as a bipartite adaptor linking DDB1, a component of a cellular SCF-like ubiquitin E3 ligase complex, to STAT2, which in turn binds STAT1 and presents STAT1 to the E3 ligase complex for ubiquitination and subsequent degradation. Furthermore, it appears that loss of STAT1 from the complex results in decreased affinity of V for STAT2 such that STAT2 either dissociates from V or is displaced by STAT1/STAT2 complexes, facilitating the cycling of the DDB1/PIV5 V containing E3 complex for further rounds of STAT1 ubiquitination and degradation. By determining the approximate number of molecules of V, DDB1, STAT1 and STAT2 present in IFN-treated 2fTGH cells, we provide additional evidence for this dynamic model of STAT1 degradation. These results show that (i) in IFN-treated cells there is approximately 4-fold less STAT2 and 15-fold less DDB1 than STAT1 per cell and thus DDB1 and STAT2 must repeatedly acquire more STAT1 for degradation to go to completion, and (ii) approximately 600 molecules of V protein per cell can target as many as 120,000 molecules of STAT1 for degradation in the absence of either viral or cellular protein synthesis. The importance of this mechanism in terms of the ability of the virus to dismantle the IFN-induced anti-viral state of cells is discussed.

Whilst screening various cell lines for their ability to respond to interferon (IFN), we noted that in comparison to other tissue culture cells AGS tumour cells, which are widely used in biomedical research, had very low levels of STAT1. Subsequent analysis showed that the reason for this is that AGS cells are persistently infected with parainfluenza virus type 5 (PIV5; formally known as SV5), a virus that blocks the interferon (IFN) response by targeting STAT1 for proteasome-mediated degradation. Virus protein expression in AGS is altered in comparison to the normal pattern of virus protein synthesis observed in acutely infected cells, suggesting that the AGS virus is defective. We discuss the relevance of these results in terms of the need to screen cell lines for persistent virus infections that can alter cellular functions.

We have previously reported that the addition of interferon (IFN) to the culture medium of Vero cells (which cannot produce IFN) that were infected with the CPI- strain of parainfluenza virus 5 (PIV5, formally known as SV5), that fails to block IFN signaling, rapidly induces alterations in the relative levels of virus mRNA and protein synthesis. In addition, IFN treatment also caused a rapid redistribution of virus proteins and enhanced the formation of cytoplasmic viral inclusion bodies. The most studied IFN-induced genes with known anti-viral activity are MxA, PKR and the Oligo A synthetase/RNase L system. We therefore examined the effects of these proteins on the replication cycle of PIV5. These studies revealed that while these proteins had some anti-viral activity against PIV5 they were not primarily responsible for the very rapid alteration in virus protein synthesis observed following IFN treatment, nor for the IFN-induced formation of virus inclusion bodies, in CPI- infected cells.

The V protein of the Paramyxovirus simian virus 5 (SV5) is a multifunctional protein containing an N-terminal 164 residue domain that is shared with the P protein and a distinct C-terminal domain that is cysteine-rich and which is highly conserved among Paramyxoviruses. We report the recovery from Vero cells [interferon (IFN) nonproducing cells] of a recombinant SV5 (rSV5) that lacks the V protein C-terminal specific domain (rSV5VDeltaC). In Vero cells rSV5VDeltaC forms large plaques and grows at a rate and titer similar to those of rSV5. In BHK or CV-1 cells rSV5VDeltaC forms small plaques and grows poorly. However, even when grown in Vero cells rSV5VDeltaC reverts to pseudo-wild-type virus in four to five passages, indicating the importance of the V protein for successful replication of SV5. Whereas rSV5 grows in many cell types with minimal cytopathic effect (CPE), rSV5VDeltaC causes extensive CPE in the same cell types. To overcome the antiviral state induced by IFN, many viruses have evolved mechanisms to counteract the effects of IFN by blocking the production of IFN and abrogating IFN signaling. Whereas rSV5 blocks IFN signaling by mediating the degradation of STAT1, rSV5VDeltaC does not cause the degradation of STAT1 and IFN signaling occurs through formation of the ISGF3 transcription complex. Furthermore, we find that rSV5 infection of cells prevents production of IFN-beta. The transcription factor IRF-3 which is required for transcription of the IFN-beta gene is not translocated from the cytoplasm to the nucleus in rSV5-infected cells. In contrast, in rSV5VDeltaC-infected cells IRF-3 is localized predominantly in the nucleus and IFN-beta is produced. By using ectopic expression of IRF-3, it was shown that after dsRNA treatment and expression of the V protein IRF-3 remained in the cytoplasm, whereas after dsRNA treatment and expression of the P protein (which lacks the C-terminal cysteine-rich domain) IRF-3 was localized predominantly in the nucleus. Thus, SV5 blocks two distinct pathways of the innate immune response, both of which require the presence of the C-terminal specific cysteine-rich domain of the multifunctional SV5 V protein.

In this article we show that the paramyxovirus SV5 is a poor inducer of interferon-beta (IFN-beta). This inefficient induction is a consequence of the expression of an intact viral V protein. In the absence of the viral V protein cysteine-rich C-terminal domain, IFN-beta mRNA is strongly induced and the transcription factors NF-kappaB and IRF-3 are activated significantly. The V protein can work in isolation from SV5 to block intracellular dsRNA signaling. The mechanism of block to dsRNA signaling is distinct from that previously observed for blocking IFN signaling in that proteolysis of candidate factors cannot be detected, and furthermore, the respective blocks require distinct protein domains. Blocking of the induction of IFN-beta by dsRNA requires the C-terminal cysteine-rich domain, a feature that is highly conserved among paramyxoviruses. We demonstrate that the V proteins from other paramyxoviruses have equivalent functions and speculate that limiting the yield of IFN-beta during infection may be a general property of paramyxoviruses.

CPI(+) and CPI(-) are two canine isolates of simian virus 5 (SV5). CPI(+) was originally isolated from the cerebrospinal fluid of a dog with temporary posterior paralysis and CPI(-) was recovered at 12 days p.i. from the brain tissue of a dog experimentally infected with CPI(+). We have previously shown that the V protein of SV5 blocks interferon (IFN) signalling by targeting STAT1 for degradation. Here we report that whilst CPI(+) targets STAT1 for degradation, CPI(-) fails to and as a consequence, CPI(+) blocks IFN signalling but CPI(-) does not. Three amino acid differences in the P/V N-terminal common domain of the V protein are responsible for the observed difference in the abilities of CPI(+) and CPI(-) to block IFN signalling. In cells persistently infected with CPI(-) the virus may become repressed in response to IFN, under which circumstances virus glycoproteins are lost from the surface of infected cells and virus nucleocapsid proteins accumulate in cytoplasmic inclusion bodies. We suggest that in vivo cells infected with IFN-resistant viruses (in which there would be continuous virus protein synthesis) may be more susceptible to killing by cytotoxic T cells than cells infected with IFN-sensitive viruses (in which virus protein synthesis was repressed), and a model of virus persistence is put forward in which there is alternating selection of IFN-resistant and IFN-sensitive viruses depending upon the state of the adaptive immune response.

STAT1 and STAT2 are cellular transcription factors involved in interferon (IFN) signaling and are thus critical for the IFN-induced antiviral state. We have previously shown that the paramyxovirus Simian Virus 5 (SV5) blocks both type I and type II interferon (IFN) signaling by targeting STAT1 for proteasome-mediated degradation. To determine whether this is a feature common to all Paramyxoviridae, we examined the abilities of SV5, Sendai virus (SeV), human respiratory syncytial virus (RSV), and human parainfluenza viruses types 2 and 3 (hPIV2 and hPIV3, respectively) to block interferon signaling. The results showed that in reporter assays SV5, SeV, and hPIV3 blocked both type I and type II IFN-signaling; hPIV2 blocked type I but not type II IFN-signaling; and RSV failed to block either type I or type II IFN-signaling. In agreement with these results, SV5 and SeV inhibited the formation of the ISGF3 and GAF transcription complexes (essential for type I and type II signaling, respectively). Surprisingly, although hPIV3 inhibited IFN-induction of the ISGF3 complex, GAF complexes were detected in hPIV3-infected cells. hPIV2 also blocked the formation of the ISGF3 complex but not the GAF complex, whereas RSV failed to block the induction of either complex. SV5 was the only virus that caused the degradation of STAT1. Indeed, in SeV- and hPIV3-infected cells STAT1 was phosphorylated on tyrosine 701 (Y701), a characteristic of IFN receptor activation. However, consistent with these viruses blocking IFN signaling downstream of receptor activation, there was a specific reduction in the levels of serine 727 (S727)-phosphorylated forms of STAT1alpha in SeV- and hPIV3-infected cells. In contrast both (Y701)- and (S727)-phosphorylated forms of STAT1 were detected in hPIV2-infected cells but there was a specific loss of STAT2. Both STAT1 (including Y701- and S727-phosphorylated forms) and STAT2 could readily be detected in RSV-infected cells. Despite not being able to block type I or type II IFN signaling, RSV was able to replicate in human cells that produce and respond to IFN, suggesting that RSV must have an alternative method(s) for circumventing the IFN response. These results demonstrate that, although interference with IFN signaling is a common strategy among Paramyxovirinae, distinct virus-specific mechanisms are used to achieve this end., (Copyright 2000 Academic Press.)

Using recombinant proteins extracted from mammalian cells, in a novel protein:protein binding assay, direct interaction of the nucleoprotein (NP) of simian virus 5 with the phosphoprotein (P) and V protein (V) was demonstrated. The amount of NP bound by V was found to be significantly less than that bound by P. Furthermore, preabsorption of NP with P removed the fraction of NP that could be bound by V, but preabsorption of NP with V did not remove all the NP that could be bound by P. These results suggested that V bound a subpopulation of the NP recognised by P. Further analysis revealed that P bound both soluble and homopolymeric forms of NP, while V bound only the soluble form; thus demonstrating that the binding sites on P and V, for soluble NP, are located within the N-terminal domain common to both P and V proteins. A monoclonal antibody, which recognised an epitope in the unique C-terminus of P, blocked the binding of P to polymeric NP but not to soluble NP. These results also suggest that there are two binding sites on NP for P, the site that interacts with the P/V common domain being either hidden or conformationally altered in polymeric NP.

The nucleotide sequence of the "P/V gene" of parainfluenza virus type 2 is presented. To determine the nature of any nontemplated additions of nucleotides that may arise during the synthesis of mRNA from this gene the polymerase chain reaction was used to amplify specific sequences of both genomic RNA and mRNA. These results demonstrated that the V protein is encoded by the genome while insertion of two nontemplated G residues are required to synthesize P mRNA. The predicted P protein has 44% identical amino acid homology with that of simian virus 5 and 37% with mumps virus; 25% of the amino acids is conserved between all three viruses.