|Home | About | Journals | Submit | Contact Us | Français|
Lymphocytic choriomeningitis virus (LCVM) nucleoprotein (NP) counteracts the host type I interferon (IFN) response by inhibiting activation of the IFN regulatory factor 3 (IRF3). In this study, we have mapped the regions and specific amino acid residues within NP involved in its anti-IFN activity. We identified a region spanning residues 382 to 386 as playing a critical role in the IFN-counteracting activity of NP. Alanine substitutions at several positions within this region resulted in NP mutants that lacked the IFN-counteracting activity but retained their functions in virus RNA synthesis and assembly of infectious particles. We used reverse genetics to rescue a recombinant LCMV strain carrying mutation D382A in its NP [rLCMV/NP*(D382A)]. Compared to wild-type (WT) LCMV, rLCMV/NP*(D382A) exhibited a higher level of attenuation in IFN-competent than IFN-deficient cells. In addition, A549 cells infected with rLCMV/NP*(D382A), but not with WT LCMV, produced IFN and failed to rescue replication of the IFN-sensitive Newcastle disease virus.
Arenaviruses cause chronic infections of rodents with a worldwide distribution (3). Infected rodents move freely in their natural habitat and may invade human dwellings. Humans are usually infected through mucosal exposure to aerosols or by direct contact of skin abrasions with infectious material. Several arenaviruses, chiefly Lassa virus (LASV), cause hemorrhagic fever disease in humans and represent a serious public health problem in their areas of endemicity (3, 31, 39). Moreover, compelling evidence indicates that the worldwide-distributed prototypic arenavirus lymphocytic choriomeningitis virus (LCMV) is a neglected human pathogen of clinical significance (23, 32) and poses a special threat to immunocompromised individuals, as illustrated by cases of transplant-associated infections by LCMV with fatal outcomes (10, 36, 40).
Mice persistently infected with LCMV exhibit relative high levels of virus replication in many tissues, but only very modest levels of type I interferon (IFN-I) are detected in sera from these mice (4). These findings would suggest that LCMV is able to modulate the innate defense responses of its natural host, the mouse, to favor virus long-term persistence. In addition, evidence indicates that morbidity and mortality associated with LASV infection involve a failure of the host's innate immune response to restrict virus replication and to facilitate the initiation of an effective adaptive immune response (31).
Mammals have developed pathogen receptor recognition (PRR) sensors at both the cell membrane (Toll-like receptors [TLRs]) and cytoplasm (retinoic-acid-inducible gene I [RIG-I] and melanoma differentiation-associated gene 5 [MDA5]). These sensors recognize specific viral components referred to as pathogen-associated molecular patterns and lead to activation of the transcription factor IFN regulatory factor 3 (IRF3), IRF7, or both, which together with AP1 and NF-κB transcription factors promote a robust induction of IFN-I production and IFN-stimulated genes (ISGs), which play a key role in the host antiviral innate immune response (2, 14, 22). To counteract this, viruses have developed a plethora of strategies to disrupt the IFN-mediated antiviral defense of the host. Accordingly, viral gene products with anti-IFN activity are often major virulence determinants. Mutations that functionally impair these viral IFN-counteracting factors often cause attenuation both in vitro and in vivo, as illustrated by influenza and Ebola viruses containing mutant NS1 and VP35, respectively, proteins unable to inhibit IRF3 activation (7, 15, 17, 18, 20, 43).
The IRF3 pathway is one of the cellular factors most commonly targeted by viruses to inhibit the IFN response (8, 12, 13, 16, 45, 46). Viral IFN antagonist proteins can target IRF3 via different mechanisms, including inhibition of IRF3 activation, inhibition of the transcriptional function of activated IRF3, or targeting of proteins that interact with IRF3 (16). We have documented that the nucleoprotein (NP) of the prototypic arenavirus LCMV counteracts the IFN-I response during viral infection based on its ability to prevent the activation and nuclear translocation of IRF3 and subsequent induction of IFN-I production and ISGs (30). This anti-IFN activity is shared by all arenaviruses so far examined, with exception of the New World arenavirus Tacaribe virus (TCRV) (26, 29). Here we present mutation-function studies that mapped critical domains and amino acid residues within NP that are critical for its anti-IFN activity. We identified NP mutants that lost their anti-IFN activity while retaining their activity in virus RNA replication and gene expression in a minigenome (MG) rescue assay. Segregation between the roles played by NP in counteracting induction of IFN-I and viral RNA synthesis allowed us to rescue a recombinant LCMV (rLCMV) strain with a mutated NP impaired in its anti-IFN activity.
BHK-21, Vero, A549, and 293T cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin. Sendai virus (SeV) strain Cantell and Newcastle disease virus (NDV) expressing the green fluorescent protein (NDV-GFP) were grown in 10-day-old embryonated eggs, as previously described (1, 37).
LCMV NP wild-type (WT) and mutant constructs were generated by PCR using specific primers (available upon request) and cloned into a modified pCAGGs expression plasmid containing a hemagglutinin (HA) tag at the carboxy-terminal end, pCAGGs HA-COOH (35). For the PCRs, a WT pCAGGs expression plasmid for LCMV NP was used as a template. pCAGGs expression plasmids for influenza virus NS1 and LASV NP have been previously described (30). Amino- and carboxy-terminal mutants (as well as internal deletion versions of the viral protein) were confirmed by sequencing and Western blotting (WB). Single amino acid substitutions for alanine or the indicated amino acids were generated by site-direct mutagenesis (Stratagene) on LCMV and LASV NP cDNAs. The IRF3-dependent reporter plasmid p55C1B-FF (49) was a gift from T. Fujita. The IFN-β red fluorescent protein-chloramphenicol acetyltransferase (RFP-CAT) reporter plasmid was previously described (30).
The GFP-tagged versions of LCMV NP were generated by amplifying the viral gene with PCR and cloning into a modified pCAGGs expression vector expressing GFP, pCAGGs-GFP. This plasmid contains two multicloning sites flanking GFP to make amino- or carboxy-GFP-tagged constructs. Fusions to GFP were confirmed by sequencing and by WB.
Activation of the IFN-β and IRF3-dependent reporter plasmids after SeV infection has been previously described (30). Briefly, 293T cells were cotransfected by a calcium phosphate method with 0.5 μg of the reporter plasmids and the indicated amounts of the LCMV NP expression vectors together with a simian virus 40 Renilla luciferase (SV40-RL) expression plasmid to normalize transfection efficiencies. After overnight transfection, cells were infected with SeV, and 16 h postinfection (p.i.), cell lysates were prepared for luciferase activity and protein expression. Luciferase activities were measured by using the dual-luciferase kit as recommended by the manufacturer (Promega). Reporter gene activation is expressed as induction (activation) over a noninduced empty-vector-transfected control. Protein expression was determined by WB with the indicated mono- or polyclonal antibodies.
Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and then transferred onto nitrocellulose filters. After blocking with 10% milk, filters were incubated with a polyclonal antibody against HA (Sigma) or a monoclonal antibody against GFP (Clontech) for 1 h. After three washes with phosphate-buffered saline containing 0.5% Tween 20, membranes were incubated with secondary horseradish peroxidase-conjugated anti-mouse or anti-rabbit immunoglobulin (Ig) antibodies (Amersham Biosciences). Protein expression was detected using an enhanced chemiluminescence kit from PerkinElmer.
BHK-21 cells were cotransfected with plasmids coding for the LCMV MG and the viral polymerase (L), together with the corresponding LCMV NP expression plasmids, as described previously (41, 42). WT LCMV NP expression plasmid was always included as a positive control, whereas transfection in the absence of L (−L) served as a negative control. At 60 h posttransfection, cell lysates were prepared for CAT assay.
To produce virus-like particles (VLPs), HEK-293T cells (1.5 × 106) growing in 35-mm-diameter wells were transfected using Lipofectamine 2000 with plasmids encoding NP (0.8 μg), L (1 μg), GP (0.4 μg), Z protein (0.1 μg), or T7 RNA polymerase (T7RP; 1 μg) and an LCMV MG (MG#7Δ2G) (0.5 μg) that allowed for T7RP-mediated intracellular synthesis of an LCMV MG RNA containing the open reading frame (ORF) of the CAT reporter gene in lieu of the virus NP ORF. Control transfections lacking Z protein received empty pCAGGS plasmid to keep constant the total amount of DNA transfected in each well. At 72 h posttransfection, supernatants (1.5 ml) were saved and cell lysates prepared for CAT assay. Aliquots of clarified supernatants (0.6 μl) were used to infect fresh monolayers of BHK-21 cells in 12-well plates. After 4 h, the inoculum was removed and cells were infected with LCMV at a multiplicity of infection (MOI) of 3. After 90 min of adsorption, the inoculum was removed and fresh medium was added. At 60 h p.i. with LCMV, cell lysates were prepared and analyzed in a CAT assay.
The rLCMV WT was generated on the basis of the rescue plasmids described previously (11, 44). The mutant rLCMV/NP*(D382A) was generated by using a pPOLI-S* plasmid carrying a mutation in LCMV NP that leads to amino acid substitution of aspartic acid for alanine at amino acid position 382. Briefly, BHK-21 cells were cotransfected with the expression plasmids supporting viral replication and transcription (pCAGGs NP and pCAGGs L) together with the plasmids encoding either the WT NP (pPOLI-S) or LCMV/NP*(D382A) (pPOLI-S*). After transfection, supernatants from transfected cells were passaged in fresh BHK-21 cells to grow the rescued viruses. Recombinant virus encoding the LCMV/NP*(D382A) was confirmed by reverse transcription-PCR (RT-PCR) and sequencing. LCMV titers were determined by immunofocus centers on Vero cells as described previously (6).
IFN-deficient (Vero) and IFN-competent (A549) cells were infected (MOI of 0.001) with either the rLCMV WT or rLCMV/NP*(D382A), and at the indicated times p.i., tissue culture supernatant (TCS) samples were collected and infectious virus was determined using an immunofocus assay. Each virus-cell-type infection was done in triplicate, and results for each independent infection were collected independently.
The NDV-GFP complementation bioassay was done as described previously (30). Briefly, IFN-competent A549 cells were infected (MOI of 3) with the rLCMV WT or rLCMV/NP*(D382A), and at 48 h p.i., infected cells were seeded into a M24 plate (1 × 105 cells/well) and subjected to a liposome-based DNA transfection protocol known to trigger production of IFN-I (30), and 24 h later infected with NDV-GFP (MOI of 1). At 24 h p.i. with NDV-GFP, cells were fixed in 2.5% formaldehyde and permeabilized with 0.1% NP-40. Cells were incubated with a guinea pig polyclonal serum to LCMV for 1 h, followed by incubation with red-conjugated anti-guinea pig goat IgGs. Samples were analyzed by fluorescence microscopy for LCMV (red) and NDV-GFP (green) infections. Cells were incubated with DAPI (4′,6-diamidino-2-phenylindole; Invitrogen) for nuclear chromatin staining.
Sequence homology among different arenavirus NPs was based on the following sequences: LCMV NP strain Armstrong (accession no. AY847350), LASV NP strain Josiah (accession no. AY628203), Junin arenavirus NP strain Candid #1 (accession no. AY746353), Machupo virus NP strain Carvallo (accession no. NC 004293), Latino virus NP strain Maru 10924 (accession no. AF512830), Pichindé virus NP strain AN3739 (accession no. NC 006447), Whitewater Arroyo virus (WWAV) NP strain AV 9310135 (accession no. AF228063), and TCRV NP strain TRVL 11573 (accession no. NC 004293).
BHK-21 cells were infected (MOI of 0.1) with TACV, and at 48 h p.i., RNA was isolated using the TRI reagent procedure and used to amplify the complete ORF of the virus NP by RT-PCR. PCR products coming from two independent RT reactions were ligated into pCRII, and upon bacterial transformation using Escherichia coli DH5-α, independent clones were selected for sequencing.
To gain insight about the regions within the LCMV NP required for the anti-IFN activity, we generated a collection of truncated N- and C-terminal NP deletion mutants, as well as internal deletion mutants, and assayed them for their ability to inhibit activation of an IRF3-dependent promoter.
We generated 50-amino-acid progressive deletions in the N terminus of LCMV NP (Fig. (Fig.1A)1A) by PCR and cloned the mutated LCMV NP genes into a modified pCAGGs expression plasmid (pCAGGs COOH-HA) that provided a C-terminal HA tag to facilitate detection of the mutant constructs. Previously we showed that addition of an HA tag at the C terminus of NP did not affect either its anti-IFN or replication activity. We observed some differences in expression levels among N-terminal deletion NP mutants within the first 400 amino acid residues. However, all tested NP mutants were readily detected by WB and reached expression levels similar or above to those obtained with NP WT (Fig. (Fig.1B).1B). Exceptions were mutants ΔN450 and ΔN500, whose expression levels were undetectable by WB. To examine the ability of these NP N-terminal deletion mutants to inhibit activation of an IRF3-dependent promoter, we cotransfected 293T cells with each of the indicated mutants together with an IRF3-dependent luciferase reporter plasmid (p55C1B-FF), and also SV40-RL as an internal transfection control. At 24 h posttransfection, cells were infected with SeV, and 24 h later, we prepared cell extracts to analyze promoter activation (Fig. (Fig.1C).1C). Our results revealed that the N terminus (residues 1 to 350) of LCMV NP did not play a critical role in the anti-IFN activity of NP. We tested the activity of the same mutants in our LCMV MG rescue system (Fig. (Fig.1D).1D). Notably, all NP mutants failed to promote replication and expression of the LCMV MG, suggesting that integrity of the N terminus of NP is required for its role in viral RNA synthesis.
The inability to detect expression of mutants ΔN450 and ΔN500 by WB prevented us from conclusively determining that NPs lacking the N-terminal 450 (or more) amino acid residues were unable to inhibit SeV-mediated activation of an IRF3-dependent promoter. However, we detected readily expression of mutant ΔN400, which was unable to inhibit activation of the IRF3 (Fig. (Fig.1C).1C). This result indicated that amino acid residues between positions 350 and 400 were important in the anti-IFN activity of NP, and thereby we predicted that mutants ΔN450 and ΔN500 lacked also the anti-IFN activity. To confirm this prediction, we generated C-terminal GFP fusion proteins for mutants ΔN450 and ΔN500. Expression of both ΔN450/GFP and Δ500/GFP proteins was readily detected (data not shown), but neither ΔN450/GFP nor Δ500/GFP inhibited activation of the IRF3 (data not shown) or IFNb (data not shown) promoters. In contrast, and consistent with our previous findings (Fig. (Fig.1C),1C), both ΔN300/GFP and ΔN350/GFP proteins inhibited activation of both promoters to levels comparable to those of WT NP/GFP (data not shown). These results indicated that the region spanning residues 350 to 558 was required for the IFN-counteracting activity of NP.
We next examined the ability of C-terminal mutants of LCMV NP to inhibit an IFN-responsive promoter and to support viral replication in the LCMV MG assay. For this, we generated a collection of C-terminal mutants (Fig. (Fig.2A)2A) and tested them for their ability to inhibit activation of the IRF3-dependent promoter. All mutants were expressed to levels similar to, or higher than those of WT NP (Fig. (Fig.2B).2B). Deletion of five C-terminal amino acids (ΔC5) did not affect the anti-IFN (Fig. (Fig.2C)2C) or RNA synthesis (Fig. (Fig.2D)2D) activities. In ΔC5 transfected cells, an additional truncated form of NP was readily detected. The source of this truncated NP was not determined, but it did not appear to affect the anti-IFN and replication activities of ΔC5. Deletions of more than five amino acids in the C-terminal part of LCMV NP eliminated the NP's ability to inhibit activation of the IRF3-dependent promoter (Fig. (Fig.2C)2C) and to promote RNA replication and expression of an LCMV MG (Fig. (Fig.2D).2D). These results indicated that, with the exception of the last five C-terminal amino acids, the integrity of the C-terminal part of NP was required for both its anti-IFN activity and roles in virus RNA replication and transcription.
To further characterize the NP requirements to counteract the IFN response, we generated a series of internal deletion mutants within the C-terminal region of NP (Fig. (Fig.3A).3A). We first confirmed that all of these NP mutants could be expressed to levels comparable to those of WT NP (Fig. (Fig.3B).3B). We then tested these mutants in our reporter assays. Consistent with previous results (Fig. (Fig.1),1), an NP deletion mutant lacking residues 350 to 400 (Δ350-400) lost its IFN-counteracting activity (Fig. (Fig.3C).3C). However, an NP deletion mutant lacking residues 350 to 360 (Δ350-360) retained the ability to inhibit activation of an IRF3-dependent promoter to levels comparable to those seen with WT NP, whereas an NP deletion mutant lacking residues 360 to 370 (Δ360-370) inhibited promoter activation to a slightly lesser extent. All other C-terminal internal deletion mutants tested were unable to inhibit activation of the reporter plasmid. These findings provided further evidence of the critical role played by the C-terminal region of NP in inhibiting induction of IFN-I. All NP mutants tested failed to promote RNA replication and expression of the LCMV MG (Fig. (Fig.3D),3D), further reinforcing the view that the role of NP in viral RNA synthesis requires structural integrity of most of the NP.
Altogether, these results revealed distinct NP requirements for viral RNA synthesis and inhibition of IFN production. The C-terminal part of LCMV NP (amino acids 370 to 553) was critical in counteracting the IFN response, whereas the majority of the viral protein sequence, except for the five most-C-terminal amino acids, was required for the roles of NP in virus RNA replication and gene expression. These findings raised the possibility that these two protein functions could be separated and therefore the feasibility of generating a mutant NP fully competent in replication yet not able to counteract the IFN response, as found with the WT form of the naturally occurring TCRV NP (29).
We used sequence identity among known arenaviral NPs to identify candidate amino acid residues that could play a critical role in the IFN-I-counteracting activity of NP (data not shown). Based on this information, we generated a collection of single amino acid substitutions, as well as small deletions, involving amino acid residues predicted to be important for the IFN-I-counteracting activity of NP and tested them regarding their ability to counteract the IFN response and to promote replication and expression of an LCMV MG (Table (Table1).1). We were particularly interested in identifying NP mutants that retained their activity in RNA synthesis but lost their ability to counteract the IFN response, since these types of mutants (NP*) would represent excellent candidates for the generation of rLCMVs to study the contribution of the anti-IFN activity of NP in the context of the natural virus infection. Compared to WT NP, alanine substitutions at positions 382 and 385, and to lesser degree also 386, significantly affected the anti-IFN function of NP (Fig. (Fig.4A).4A). Mutations D382A and R386A did not affect significantly the NP's ability to promote replication and expression of the LCMV MG, whereas mutation G385A reduced the NP's activity in the MG rescue assay to about 40% of that of WT NP (Fig. (Fig.4B).4B). In contrast, NP* I383A and E384A exhibited WT NP activity in its anti-IFN (Fig. (Fig.4A)4A) and RNA synthesis (Fig. (Fig.4B)4B) functions, whereas NP* Δ382-385 failed to perform both NP functions (Table (Table1).1). All mutants examined within the NP region spanning residues 382 to 386 were expressed to similar levels as WT NP (Fig. (Fig.4C).4C). To further confirm that residues 382 and 385 played a critical role in the IFN-counteracting activity of LCMV-NP, but not in the NP roles associated with viral RNA synthesis and gene expression, we generated alternative amino acid substitutions and deletions at these positions and characterized them functionally (Fig. (Fig.5A).5A). Additional amino acid substitutions at positions 382 (D382M and D382E) and 385 (G385M) did significantly affect the NP anti-IFN activity (Fig. (Fig.5B)5B) without disrupting NP's activity in virus RNA replication and gene expression (Fig. (Fig.5C).5C). Single amino acid deletions at positions 382 and 385 resulted in the lost of the NP IFN-counteracting activity and also in the NP's ability to mediate replication and expression of an LCMV MG. Notably, the double mutant NP* D382A G385A lacked or was impaired in its anti-IFN activity but retained the ability to promote replication and expression of an LCMV MG (Fig. 5B and C).
These findings led us to examine whether the same amino acid residues 382 and 385 played also a critical role in the anti-IFN activity exhibited by the LASV NP. For this, we generated the corresponding alanine substitutions in LASV NP and tested them in our plasmid-based IFN reporter assay (Fig. (Fig.6).6). LASV NP* D382A and G385A were expressed to levels comparable to the LASV WT NP, but both were unable to inhibit SeV-mediated activation of the IFN-responsive promoter (Fig. (Fig.66).
These results demonstrated that the IFN-counteracting activity of NP could be segregated from its replication activity, and therefore it would be possible to generate recombinant viruses with mutated NP proteins that are fully competent in replication yet not able to counteract the IFN response.
The generation of rLCMV carrying a mutated NP* lacking the ability to counteract the IFN-I response would allow us to assess the biological implications of the NP IFN-counteracting activity in the context of the natural course of virus infection. Successful rescue of rLCMV/NP* could be complicated if the mutated NP* was also affected in any of the other NP activities, including RNA replication and particle formation. Results from the MG rescue system identified several NP* mutants that lost their anti-IFN activity while retaining WT activity in promoting LCMV MG replication and expression. These results, however, did not rule out whether these NP* mutants could be impaired in promoting generation of infectious virus particles. To address this issue, we tested NP* D382A and G385A in their ability to generate infectious VLPs (27, 38, 41). NP* D382A and G385A were able to produce infectious VLPs to levels similar to those of NP WT (Fig. (Fig.7).7). Based on these findings, and as a proof of concept, we attempted to rescue an rLCMV carrying the mutation D382A. For this, we used reverse genetics procedures to rescue infectious rLCMV entirely from cloned cDNAs (11, 44). Briefly, we transfected BHK-21 cells with plasmids that allowed for RNA polymerase I-mediated (pPol) intracellular synthesis of the viral L and S genome RNA species, together with plasmids expressing the minimal viral trans-acting factors NP (pCAGGs-NP) and L polymerase (pCAGGs-L) required for virus RNA replication and transcription. The construct driving expression of the S segment (pPoI-S/NP*D382A) was engineered to incorporate mutation D382A within the NP gene. At 60 h posttransfection, cell culture supernatants were used to infect fresh BHK-21 cells to prepare stocks of the rescued virus.
To characterize the interaction between rLCMV/NP*(D382A) and the host IFN-I response, we compared the growth kinetics of LCMV WT and rLCMV/NP*(D382A) in IFN-deficient (Vero) and IFN-competent (A549) cell lines (Fig. (Fig.8A).8A). In several independent infections at very low MOI (0.001), we observed that in IFN-deficient Vero cells, LCMV/NP*(D382A) tended to grow to slightly lower titers than WT LCMV, whereas in IFN-competent A549 cells, LCMV/NP*(D382A) growth tended to be further impaired compared to that of WT LCMV (Fig. (Fig.8A).8A). Likewise, Lipofectamine-DNA (LF/DNA) transfection of A549 cells infected with LCMV/NP*(D382A), but not with WT LCMV, induced production of IFN-I that resulted in an antiviral state that inhibited multiplication of NDV-GFP (Fig. (Fig.8B).8B). Notably, and physiologically more relevant, in the absence of LF/DNA transfection, A549 cells infected with rLCMV/NP*(D382A), but not with WT LCMV, exhibited increased resistance to infection with NDV-GFP. This result indicated that rLCMV/NP*(D382A), without requiring an additional stimulus, triggered the IFN response, which provided further evidence of the critical role of NP in the arenavirus's ability to counteract the IFN response of the cell. Consistent with these findings, Vero cells treated with TCS from LCMV/NP*(D382A)-infected A549 cells, but not from WT LCMV-infected A549 cells, developed an antiviral state (Fig. (Fig.8C8C).
Viral antagonists of the IFN-I response include structural and nonstructural proteins. Some of these viral anti-IFN proteins are dispensable for viral replication in IFN-compromised systems, whereas others are key components of the viral replication machinery. In the latter case, mutations affecting the anti-IFN activity of the protein may compromise its functions in virus multiplication, thus making it difficult to assign specific functions to specific amino acid residues.
We have documented that TACV NP is severely impaired in its ability to inhibit induction of IFN-I compared to NPs from many other arenaviruses (29). This observation suggested the possibility of segregating the roles of the arenavirus NP in RNA synthesis and formation of infectious virus particle from its counteracting IFN activity. In the present work, we used mutation-function studies to identify specific amino acid residues within LCMV NP that if mutated impaired the NP anti-IFN activity without affecting the functions of NP in RNA synthesis and formation of infectious virus particles. Our results from mutation-function studies identified residues 382, 385, and, to a lesser degree, 386, as critically required for the anti-IFN activity of NP but dispensable for the roles of NP in virus RNA replication and generation of infectious virus particles. Consistent with results from cell-based functional assays, rLCMV/NP*(D382A) was attenuated in IFN-competent cells, and rLCMV/NP*(D382A)-infected A549 cells were able to induced an antiviral state that inhibited subsequent infection with the IFN-sensitive NDV.
The critical role played by residues 382 and 385 in the IFN-I-counteracting activity of NP was further supported by the observation that NP with substitutions, other than alanine, at these amino acid positions, as well as mutant Δ382-386, resulted in NPs lacking the anti-IFN-I activity. Moreover, LASV NP* D382A and G385A lost also their ability to inhibit induction of IFN-I in cell-based assays. Intriguingly, substitutions at positions 383 and 384 affected neither the role of NP in virus RNA replication and gene expression nor its anti-IFN activity. The motif DIEG, spanning residues 382 to 385 of LCMV-NP, is highly conserved between Old World and New World arenaviruses (data not shown). Our results support a critical role of D382 and G385 in the arenavirus NP anti-IFN activity. Interestingly the previously published sequence for TACV NP (accession no. AAA47903) showed the sequence DIEDLQLD for residues 382 to 389, suggesting that a G385D substitution in TACV NP could have contributed to its decreased anti-IFN activity. In addition, the motif LQL corresponding to residues 386 to 388 in TACV NP is not found in any of the other known arenavirus NP sequences. Because of the potential implication of this observation, we revisited the sequence of TACV NP. For this, we used RNA isolated from TACV-infected cells to amplify via RT-PCR the complete NP ORF. Sequencing of multiple independent clones revealed the sequence DIEGPPTD for residues 382 to 389 of TACV NP, a sequence that was also present in the pC-TACV-NP expression plasmid used in our cell-based assays (29). The conservation of the DIEG motif in TACV NP, which is impaired in its anti-IFN activity, indicates that other amino acid residues located outside residues 382 to 386 should contribute to the IFN-I-counteracting activity of NP. Our mutation-function analysis did not involve a saturation mutagenesis of the region spanning residues 370 to 553, and thereby at the present time, we cannot assess whether, in addition to the DIEG motif, a specific combination of amino acid residues within the C-terminal part of NP are also required for the anti-IFN-I activity of NP. Likewise, we have documented that the NPs of the New World arenavirus WWAV and LCMV exhibit similar levels of anti-IFN activity, despite WWAV NP having the substitution P516A that we found to affect the anti-IFN activity of LCMV NP. The significant degree of genetic diversity among arenavirus NPs, together the current lack of knowledge about the molecular mechanisms by which NP inhibits induction of IFN-I production, make it difficult to determine whether other amino acid residues with the C terminus of the NP of WWAV could compensate for the presence of A instead of P at position 516.
The rescued rLCMV/NP*(D382A) exhibited phenotypic features consistent with predictions based on the results we obtained in cell-based assays using reporter gene expression as a readout. Mutation D382A was associated with some degree of impaired virus multiplication in IFN-deficient cells, but this growth impairment tended to be more accentuated in IFN-competent A549 cells (Fig. (Fig.8A).8A). In some experiments, titers of infectious virus in LCMV/NP*(D382A)-infected A549 cells at 72 h p.i. rose closer to those observed in A549 cells infected with LCMV WT. It remains to be determined whether this finding reflected the emergence over time of revertant viruses with WT NP anti-IFN activity. Likewise, compared to LCMV WT, rLCMV/NP*(D382A) exhibited also a reduced ability to rescue productive multiplication of rNDV-GFP in A549 cells even in the absence of the antiviral state induced by LF/DNA transfection (Fig. (Fig.8B).8B). Moreover, Vero cells treated with TCS from A549 cells infected with rLCMV/NP*(D382A), but not with the LCMV WT, developed an antiviral state reflected by their increased resistance to infection with vesicular stomatitis virus (Fig. (Fig.8C).8C). This antiviral state was likely triggered by the presence of IFN-I in TCS from rLCMV/NP*(D382A)-infected A549 cells because treatment of the TCS with a neutralizing antibody to IFN-I prevented the establishment of the antiviral state in Vero cells (Fig. (Fig.8C8C).
As with many other viruses, LCMV infection induces an early and transient burst of IFN-I production that peaks shortly before the peak of production of infectious progeny. This observation would appear to be in contradiction to our findings that LCMV NP exhibits a robust inhibitory effect on the induction of IFN-I. In this regard, it is worth noting that a similar situation is found during infection with influenza virus and Ebola virus, despite the fact that these two viruses are known to encode strong anti-IFN-I factors (influenza virus NS1 and Ebola virus VP35 proteins). Notably, influenza and Ebola viruses with mutated NS1 and VP35, respectively, lacking their anti-IFN activities were found to be highly attenuated in both cultured cells and animal models of infection (7, 17-19). In the case of LCMV, the contribution of different PRR sensors to virus recognition has not been elucidated. Likewise, the identity of the specific cell types responsible for the early burst of IFN-I production in LCMV-infected mice has not been unequivocally established, but evidence indicates the involvement of a complex network of cellular interactions, including dendritic cells (DCs) (33, 34), and the specialized organization of the splenic marginal zone (5, 28).
Mice deficient in TLR3 exhibit normal both innate and adaptive immune responses to LCMV (9), and TLR8 is not functional in mice, while TLR9 recognizes unmethylated CpG DNA associated with DNA viruses. Recent evidence has indicated that recognition of LCMV by plasmacytoid DCs (pDCs) via TLRs and signaling via the adaptor MyD88 play a critical role in production of IFN-I in LCMV-infected mice (24). However, other studies have shown that production of IFN-I is not impaired in pDC-depleted mice (5). It is plausible that this burst of IFN-I production is driven by cells that are not infected with LCMV but which are responding to signals produced by cells targeted by LCMV at the very early times of infection. The initially infected cell types could include fibroblasts and monocytes capable of producing IFN upon sensing viral RNA. The IFN-I-counteracting activity of NP might contribute to the modulation of the kinetic magnitude or duration of this early IFN-I response, which may influence both virus multiplication and propagation, as well as the ensuing host immune responses and thereby the outcome of infection.
It should be noted that arenaviruses, including LCMV, are maintained in their natural hosts via vertical transmission, which results in lifelong persistence due to T-cell tolerance. This chronic infection is associated with relative high levels of virus RNA replication in many cell types capable of recognizing viral RNA via both membrane-associated (TLR) and cytosolic (RIG-I and MDA5) PRR, which should result in a robust induction of IFN-I production. However, only very modest levels of IFN-I were detected in the sera of these mice (4). More recently, we have used DNA array-based approaches to compare brain gene expression profiles between congenitally persistently infected and control mice (25). Despite high viral loads in brains of persistently infected mice, we detected changes in the host's central nervous system gene expression for only 75 genes (56 increased and 19 reduced). Notably, the majority of the induced genes were ISGs, but changes in IFN-I mRNA were below detection levels. Moreover, arenaviruses appear to be relatively resistant to the antiviral effects of IFN-I. These findings would suggest that during its natural persistent state, LCMV is sensed by the host cell PRR sensors (both TLR and cytosolic RIG-I/MDA5) that normally detect the presence of replicating RNA viruses leading to a robust IFN-I response. However, LCMV is able to modulate this response and thereby prevent the deleterious consequences for the host associated with chronic production of high levels of IFN-I. This, in turn, would facilitate a balance between host and pathogen that may be required for long-term persistence.
The use of rLCMV/NP*(D382A) infection of the mouse using already well-characterized different experimental settings should allow us to gain a detailed understanding of the in vivo implications of the NP-mediated anti-IFN activity. It needs to be emphasized that similarly to the situation observed for many other viruses, arenavirus virulence is likely to be a polygenic trait, and therefore, the ability of LCMV NP to interfere with induction of type I IFN may be a necessary, but not sufficient, factor in arenavirus pathogenesis, including Lassa fever and other arenaviral hemorrhagic fever disease. The elucidation of the mechanisms underlying the IFN-counteracting activity of arenavirus NP should contribute to a better understanding of the pathogenesis and immunogenicity of arenavirus infections. This, in turn, could lead to the development of better antiviral drugs and vaccines to combat arenaviruses pathogenic to humans.
We thank R. Cadagan for technical support and W. Cardenas and T. Baas for advice and helpful discussion. We also thank T. Fujita for the IRF3-responsive reporter plasmid, p55C1B-FF.
The work of A.G.-S. was partially supported by CIVIA, an NIH-NIAID center grant (U19 AI62623), the Northeast Biodefense Center (NIH-NIAID grant U54AI57158), and DoD grant W81XWH-07. The work of J.C.T. was partially supported by NIH-NIAID grants (RO1AI47140 and R56AI077719-01A1). The work of L.M.-S. was partially supported by an NIH-NIAID grant (R56A1077719-01A1).
Published ahead of print on 26 August 2009.