Control of Innate Immune Signaling and ... - Journal of Virology

Control of Innate Immune Signaling and Membrane Targeting by the Hepatitis C Virus NS3/4A Protease Are Governed by the NS3 Helix ␣0 Stacy M. Horner, Hae Soo Park, and Michael Gale, Jr. Department of Immunology, University of Washington, Seattle, Washington, USA

Hepatitis C virus (HCV) infection is sensed in the host cell by the cytosolic pathogen recognition receptor RIG-I. RIG-I signaling is propagated through its signaling adaptor protein MAVS to drive activation of innate immunity. However, HCV blocks RIG-I signaling through viral NS3/4A protease cleavage of MAVS on the mitochondrion-associated endoplasmic reticulum (ER) membrane (MAM). The multifunctional HCV NS3/4A serine protease is associated with intracellular membranes, including the MAM, through membrane-targeting domains within NS4A and also at the amphipathic helix ␣0 of NS3. The serine protease domain of NS3 is required for both cleavage of MAVS, a tail-anchored membrane protein, and processing the HCV polyprotein. Here, we show that hydrophobic amino acids in the NS3 helix ␣0 are required for selective cleavage of membrane-anchored portions of the HCV polyprotein and for cleavage of MAVS for control of RIG-I pathway signaling of innate immunity. Further, we found that the hydrophobic composition of NS3 helix ␣0 is essential to establish HCV replication and infection. Alanine substitution of individual hydrophobic amino acids in the NS3 helix ␣0 impaired HCV RNA replication in cells with a functional RIG-I pathway, but viral RNA replication was rescued in cells lacking RIG-I signaling. Therefore, the hydrophobic amphipathic helix ␣0 of NS3 is required for NS3/4A control of RIG-I signaling and HCV replication by directing the membrane targeting of both viral and cellular substrates.

H

epatitis C virus (HCV) is a major human pathogen and the leading cause of liver disease and liver cancer. Similar to other positive-strand RNA viruses, HCV harnesses cellular membranes as a scaffold for genome replication, protein translation, and assembly (21). During HCV infection, the cytosolic RNA sensor protein RIG-I detects specific sequences in the viral RNA and signals through its adaptor protein, MAVS (28, 34). MAVS is the scaffold for a macromolecular signaling complex that assembles on innate immune synapses that are formed by mitochondrionassociated endoplasmic reticulum (ER) membrane (MAM) interactions with peroxisomes and mitochondria (9), with MAVS localized to all three of these subcellular locations (6, 9, 29). These interactions all drive intracellular innate immune signaling that culminates in the production of type I interferon (IFN), inflammatory cytokines, and IFN-stimulated genes to limit virus replication and spread (8, 11). However, HCV controls this downstream signaling, including IRF-3 activation, through HCV NS3/4A protease-mediated cleavage of MAVS on the MAM, which releases it from this intracellular membrane and also disrupts MAVS oligomerization (1, 7, 9, 14, 17, 20). The 9.6-kb single-stranded RNA genome of HCV encodes a single polyprotein that is co- and posttranslationally processed by cellular and viral proteases, including the NS3/4A protease complex, into the structural and nonstructural proteins of the virus (21). The multifunctional NS3/4A protease processes the nonstructural proteins of the HCV polyprotein and is important for RNA replication and virion production (22). NS3 contains a protease domain at its amino terminus and an NTPase/RNA helicase domain at its carboxy terminus. NS3 interactions with its cofactor, the 54-amino-acid NS4A polypeptide, anchor the NS3/4A protease complex to intracellular membranes and stimulate maximum NS3 protease activity and stability (4, 36). Both NS3 and NS4A have membrane-targeting domains, an amphipathic ␣-helix at the amino terminus of NS3, and a transmembrane ␣-helix at the amino terminus of NS4A (4). The NS4A protein

3112

jvi.asm.org

localizes to the membranes of the ER and the MAM at junctions with mitochondria and/or peroxisomes (9, 23, 36). The HCV replicase complex, which includes the NS3 protein, localizes to an ER subcompartment (23–25, 35–36). Because of the essential role of the NS3/4A protease in HCV replication and control of innate immune signaling, it has emerged as a target for HCV therapy with the first round of protease inhibitors now approved for clinical use with the current standard of care, pegylated interferon and ribavirin, as a part of a novel HCV therapy. Here, we examine how membrane targeting by NS3/4A governs both viral and host processes, including HCV polyprotein processing, viral RNA replication, and MAVS cleavage and control of HCV innate immune evasion. MATERIALS AND METHODS Cell culture and viruses. Huh7 and Huh7.5 cells are human hepatoma cells that have been described previously (30). These cells, as well as HEK293 cells, were propagated in Dulbecco’s modified Eagle’s medium (Cellgro) supplemented with 10% fetal bovine serum (HyClone), as described previously (30). Sendai virus strain Cantell was obtained from Charles River Laboratories. Immunoblotting and immunoprecipitation. Cells were lysed in a modified radioimmunoprecipitation assay (RIPA) buffer (10 mM Tris [pH 7.5], 150 mM NaCl, 0.5% sodium deoxycholate, and 1% Triton X-100) supplemented with protease inhibitor cocktail (Sigma) and phosphatase inhibitor cocktail II (Calbiochem). After harvest, protein was resolved by SDS-PAGE, transferred to nitrocellulose membranes in a 25 mM Tris-192 mM glycine-0.01% SDS buffer and blocked in 5% milk–

Received 3 November 2011 Accepted 3 January 2012 Published ahead of print 11 January 2012 Address correspondence to Michael Gale, Jr., [email protected]. Copyright © 2012, American Society for Microbiology. All Rights Reserved. doi:10.1128/JVI.06727-11

0022-538X/12/$12.00

Journal of Virology p. 3112–3120

HCV NS3 Helix ␣0 in NS3/4A Function

phosphate-buffered saline with 0.1% Tween 20 (PBS-T) buffer. For NS4A immunoblotting (IB), transfer onto Immobilon-P membranes (Millipore) was done in buffer without SDS. After washing, membranes were incubated with species-specific horseradish peroxidase-conjugated antibodies (Jackson ImmunoResearch) followed by treatment of the membrane with ECL⫹ (GE Healthcare) and imaging on X-ray film. For immunoprecipitation (IP), cells were lysed in IP buffer (10 mM Tris [pH 7.5], 150 mM NaCl, and 1% Triton X-100, in some cases supplemented with 10% glycerol). Protein was immunoprecipitated and captured on anti-Flag affinity gel (Sigma). Bound protein complexes were washed in IP buffer and then eluted using SDS loading buffer (0.125 M Tris-HCl [pH 6.8], 4% SDS, 20% glycerol, 0.004% bromophenol blue, 10% ␤-mercaptoethanol). For NS3/NS4A immunoprecipitations, protein was immunoprecipitated with either anti-NS3 or anti-NS4A overnight and bound to protein G plus agarose (Calbiochem). The following antibodies were used for immunoblot analysis: antiHCV NS3 1B6 (gift from D. Moradpour), anti HCV-NS4A (Virogen), anti-HCV polyclonal sera (gift from W. Lee), anti-Flag M2 (Sigma), antiMyc (Abcam), anti-tubulin (Sigma), and anti-Cardif (for MAVS; Axxora). Immunofluorescence analysis and confocal microscopy. Cells were stained with 2 nm of Draq5 (Axxora) according to the manufacturer’s protocol before fixation. Cells were fixed in 4% paraformaldehyde, permeabilized, and immunostained as described previously (16). For immunofluorescence, we used anti-HCV NS3 (Novocastra) and anti-calnexin (Enzo Life Sciences). Samples were imaged on a Zeiss 510 Meta confocal microscope with a 63⫻ oil immersion lens with a numerical aperture of 1.4 at the Nyquist limit. Images were collected as z stacks with a focal step size of 0.43 ␮m and processed with ImageJ. z stacks were deconvolved using AutoDeblur (Media Cybernetics). Plasmids and transfections. The plasmids used in this study have been described previously and are as follows: pEF-NS3, pEF-NS3/4A, and pEF-Flag-SC-protease (wild type [wt] and S139A) (10); pEFBos-N-RIG-I (38); and pMyc-MAVS (wt and C508Y) and pEFTak-Flag-MAVS (17). DNA transfections were done using Lipofectamine 2000 (Invitrogen) or FuGENE 6 (Roche). IFN-␤ promoter luciferase assays were conducted as described previously (30). The NS3/4A helix ␣0 mutant containing five mutations, namely, L13A, L14A, I17A, I18A, and L21A, was created using recombinant PCR with overlapping mutagenic primers (5=-GCgctagtgG CgGCgcagccaGCtGCgcctcgcgtctgttgggagtaggc-3=; 5=-GCaGCtggctgcGCc GCcactagcGCcacaggccgggacaggaaccagg-3=; uppercase represents nucleotide mutations) and subcloned into the EcoRI and XbaI restriction sites in the pEF expression vector. pEF-NS3/4A R24/26A and pEF-Flag-SCprotease R24/26A, to substitute arginine at positions 24 and 26 of NS3 with alanine, as well as pEF-NS3 L14A, pCDNA-Myc MAVS C283Y, and pCDNA-Myc MAVS C452Y, were created using site-directed mutagenesis. To create the expression constructs containing Flag HCV NS4B/5A and Flag HCV NS5A/B, PCR was used to amplify HCV genotype 1b NS4B/5A or NS5A/5B. The PCR product was digested with NotI and PmeI and subcloned into the pEF-Tak expression vector (27). The pEFTakMAVS amino acid (aa) segment from 284 to 540 was created by PCR to amplify the truncated region followed by subcloning into the NotI and PmeI sites of pEF-Tak. The sequences of oligonucletotide duplexes and primers used are available upon request. All DNA sequences were verified by DNA sequencing. HCV replicons. RNA was in vitro transcribed (MEGAscript; Ambion) from ScaI-digested HP-HCV replicon plasmid DNA, either wt, ⌬NS5B to inactivate the polymerase (31), or harboring the NS3 ␣0 mutations (L13A, L14A, I17A, I18A, L21A) and purified free of DNA. One microgram of the RNA was mixed with 4 ⫻ 106 Huh7 or Huh7.5 cells in PBS and electroporated at 960 ␮F and 250 V by using a Gene Pulser system (Bio-Rad). Cells were plated into 100-mm dishes at 2 ⫻ 105, 2 ⫻ 104, or 2 ⫻ 103 cells per dish, along with 2 ⫻ 105 cells that had been electroporated with ⌬NS5B replicon RNA. The next day, cells were washed twice with PBS and 400 ␮g/ml of G418 was added to the culture media to begin selection.

March 2012 Volume 86 Number 6

Crystal violet staining was performed after 3 weeks of G418 selection. Colonies from replicate plates were counted to determine the relative transduction efficiency of the corresponding HCV replicon RNA. Relative transduction efficiency is expressed as the percentage of colonies that were stably transduced in Huh7 cells relative to wt as compared to the percentage of colonies that were stably transduced in Huh7.5 cells relative to wt. For the ␣0 mutant, all remaining electroporated cells were plated in a 150-mm dish for G418 selection. The sole resulting colony was picked and expanded. RNA extracted (RNeasy kit; Qiagen) from this clonal cell population was reverse transcribed into cDNA (Advantage cDNA polymerase; Clontech), and the NS3 region was cloned and sequenced to deduce the nucleotide sequence of NS3 in this replicon RNA. HCV infection. HCV was produced from a molecular clone of HCV genotype 2a corresponding to the sequence of strain JFH1 and synthesized by Blue Heron, Inc. The corresponding cDNA was cloned into pENTR, allowing production of the full-length HCV RNA from the T7 promoter to make pSJ (for synthetic JFH1). pSJ/NS3 ␣0 containing alanine substitution of the hydrophobic amino acids in the NS3 helix ␣0 mutations in JFH1 (L13A, L14A, I17A, V18A, V19A, M21A) was created using recombinant PCR with overlapping mutagenic primers (5=-GCcGCgggcgccGC agCCgCgagtGCgacggggcgtgacaggacagaa-3=; 5=-cGCactcGcGGctGCggcgc ccGCgGCgcctcgtgtttgctgggcataag-3=) and subcloned into the SpeI and NotI restriction sites in pSJ. To generate genomic sJFH1 and sJFH1/NS3 ␣0 RNA, the pSJ and pSJ/NS3 ␣0 plasmids were linearized at the 3= end of the HCV cDNA by XbaI digestion. The linearized DNA was then purified and used as a template for in vitro transcription. RNA was purified free of DNA and transfected into Huh7.5 cells via electroporation, as above. HCV infection was assessed by focus forming assay, immunostaining, and immunoblotting for HCV proteins and by quantitative real-time (qRT)PCR for intracellular RNA.

RESULTS

Helix ␣0 of NS3 is required for MAVS cleavage and RIG-I pathway control by HCV NS3/4A. The amphipathic helix ␣0 at the amino terminus of NS3 (Fig. 1A) has a surface-exposed hydrophobic patch (Fig. 1B) that is highly conserved among HCV genotypes and confers membrane localization to the NS3/4A protease (4). To define how the NS3 helix ␣0 directs proteolytic targeting of MAVS by NS3/4A on intracellular membranes, we eliminated the surface-exposed hydrophobicity of the helix ␣0 (Fig. 1B) through alanine mutagenesis (␣0 mutant) and examined the ability of the encoded protein to target and cleave endogenous MAVS in human hepatoma (Huh7) cells, a hepatocyte cell line that supports HCV infection (33). These mutations have been shown to eliminate the interaction of the NS3 helix ␣0 with intracellular membranes (4). Confocal microscopy analysis revealed that the loss of the hydrophobic membrane-targeting amino acids in the NS3 helix ␣0 altered NS3/4A from a punctate pattern to diffuse staining throughout the cell that does not overlap with the ER marker calnexin (Fig. 1C), confirming that the NS3 helix ␣0 is a membrane-targeting domain (4). Coexpression analysis of NS3/4A constructs with Myc-tagged MAVS demonstrated that while wt NS3/4A could cleave MAVS, the ␣0 mutant was deficient in MAVS cleavage, although not to the same extent as NS3 containing a mutation that inactivated the protease active site (S139A; compare ␣0 to SA) (Fig. 1D). These analyses also demonstrate the requirement for coexpression of NS3 with NS4A for MAVS cleavage, as NS3 alone was unable to cleave MAVS (Fig. 1D). Proteins targeted near mitochondria often contain basic residues in proximity to a hydrophobic membrane-targeting domain that are essential for the preservation of membrane association (26). Therefore, we mutated the two arginine residues located downstream of helix ␣0 of NS3 (R24, R26)

jvi.asm.org 3113

Horner et al.

FIG 1 Helix ␣0 of HCV NS3 is required for MAVS cleavage and RIG-I pathway control by HCV NS3/4A. (A) Diagram of the sequence of the amino terminus of NS3 (HCV Con1 clone; GenBank accession no. AJ238799). Amino acids are numbered with respect to NS3 and the HCV polyprotein (top row). Functional domains are labeled. The hydrophobic amino acids in the NS3 helix ␣0 are shown in red. (B) Ribbon representation of NS3 aa 12 to 23 using the coordinates provided by Yao et al. (37) (PDB entry 1CU1) with the side chains of the surface-exposed hydrophobic amino acids depicted and labeled. (C) Confocal micrographs of Huh7 cells expressing wt NS3/4A or NS3/4A with mutation of the hydrophobic amino acids in the helix ␣0 to alanine (␣0 mutant) that were immunostained with anti-NS3 (green) and anti-calnexin (red). Nuclei were stained with Draq5 (blue). Bar: 10 ␮m. (D to F) Immunoblot analysis of extracts from cells expressing the indicated NS3/4A constructs and Myc-MAVS (D), Flag-NS5A/B (E), or Flag-NS4B/5A (F). Arrows mark full-length (FL) and cleaved (C) MAVS in panel D and full-length and cleaved HCV polyprotein in panels E and F. The constructs used express either NS3, wt NS3/4A, NS3/4A S139A (SA), NS3/4A ␣0 mutant (␣0), or NS3/4A R24/26A (RA). (G) IFN-␤ promoter reporter luciferase expression of Huh7 cells transfected with increasing amounts of NS3/4A expression plasmids and then mock or SenV infected. Values are mean ⫾ standard deviation (SD) (n ⫽ 3). *, P ⱕ 0.003 by unpaired Student’s t test. (Inset) Immunoblot for NS3 protein expression from this assay. (H) Immunoblot analysis of anti-NS3 or anti-NS4A immunoprecipitated extracts, as well as input, from cells expressing wt NS3/4A, ␣0 mutant NS3/4A, or empty vector (V).

(Fig. 1A) to investigate if they contributed to the membrane-targeting and proteolytic activities of NS3. However, these mutations did not affect cleavage of MAVS by NS3/4A (RA mutant) (Fig. 1D). To ascertain the impact of the helix ␣0 motif on the proteolytic efficiency of viral substrates, we examined the role of the NS3 helix ␣0 in cleavage of HCV polyprotein substrates. We expressed the NS3/4A constructs with Flag-tagged HCV polyprotein substrates, Flag-NS4B/5A and Flag-NS5A/5B, in Huh7 cells and assessed the processing of each by immunoblot analysis. The NS4B/5A junction is known to be membrane associated and requires NS3 interaction with NS4A for efficient processing, while the NS5A/5B junction is not membrane associated and can be cleaved by NS3 alone (2, 15). Both wt NS3/4A and the ␣0 mutant efficiently cleaved the NS5A/B junction to liberate NS5A (Fig. 1E). Moreover, mutation of the NS3 helix ␣0 did not affect processing of the

3114

jvi.asm.org

cis cleavage site between NS3 and NS4A (compare SA to ␣0) (Fig. 1E, top). However, mutation of NS3 helix ␣0 reduced the efficiency of processing of the NS4B/5A polyprotein, as compared to both the wt and RA mutants of NS3/4A (Fig. 1F). Sendai virus (SenV) is an RNA virus that triggers the RIG-I pathway in a similar fashion as HCV (28, 30). Both wt NS34/A and the NS3/4A RA mutant were able to block this SenV-mediated signaling to the IFN-␤ promoter (Fig. 1G). Consistent with a reduced cleavage efficiency of MAVS, the NS3/4A helix ␣0 mutant failed to block SenV-mediated signaling to the IFN-␤ promoter, similar to the NS3/4A SA active-site mutant (Fig. 1G). Thus, the hydrophobic composition of helix ␣0 is important for proteolytic targeting of membrane-anchored substrates, including MAVS and the HCV NS4B/5A polyprotein, as well as for control of RIG-I pathway signaling by NS3/4A during virus infection.

Journal of Virology


FIG 2 Helix ␣0 positions the HCV NS3 protease active site at intracellular membranes for substrate cleavage. (A) Confocal micrographs of Huh7 cells expressing wt SC-protease or SC-protease ␣0 mutant that were immunostained with anti-Flag (SC-protease; green) and anti-calnexin (red). Nuclei were stained with Draq5 (blue). Bar: 10 ␮m. (B to D) Immunoblot analysis of extracts from cells expressing SC-protease constructs and Myc-MAVS (B), Flag-NS5A/B (C), or FlagNS4B/5A (D). Arrows mark full-length (FL) and cleaved (C) MAVS in panel B and full-length and cleaved HCV polyprotein in panels C and D. In panel B, the asterisk indicates an alternate form of MAVS likely modified by phosphorylation (32). In panel D, the asterisk marks an alternate SC-protease cleavage product of NS4B/5A.

NS4A interacts with specific residues in the amino terminus of NS3, but it is not known to specifically interact with the NS3 helix ␣0 (18). However, because the NS3 helix ␣0 mutant demonstrated reduced ability to process the NS4B/5A polyprotein junction (which requires NS4A interaction with NS3 for efficient processing [2, 15]), we hypothesized that this construct may have a reduced interaction with NS4A compared to wt NS3. Results from coimmunoprecipitation assays of NS3 and NS4A showed that the NS3 helix ␣0 mutant did not stably interact with NS4A as compared to wt NS3 (Fig. 1H), demonstrating that the NS3 helix ␣0 is required for efficient interaction of NS3 with NS4A. Although we did not detect a stable interaction between NS4A and the NS3 helix ␣0 mutant, it is likely that some transient interaction occurs between these two proteins, as the NS3 helix ␣0 mutant was able to partially cleave the NS4B/5A polyprotein junction, albeit with reduced efficiency, and this cleavage requires NS3 interaction with NS4A. Helix ␣0 of NS3 required for membrane-associated proteolysis by NS3/4A. The NS3 cofactor NS4A contains a transmembrane domain at its amino terminus that confers membrane targeting (4) and is known to stimulate NS3 protease activity, especially for membrane-anchored substrates (2, 15). Therefore, to isolate the function of the NS3 helix ␣0 from that of the NS4A transmembrane domain in conferring membrane targeting and function of the NS3 protease, we utilized a single-chain protease construct of NS3/4A, which lacks the transmembrane domain of NS4A but contains the central 12 amino acids of NS4A fused to the NS3 protease domain (SC-protease) (10). Use of the SC-protease eliminates any contribution of the NS4A transmembrane domain to the targeting function of the protein while maintaining interaction of the NS4A amino acids essential for full protease activity with NS3 by directly tethering them onto the amino terminus xof NS3. As shown in Fig. 2A, the NS3 helix ␣0 mutations in the SCprotease altered its localization from a punctate pattern to diffuse staining throughout the cell that did not overlap with the ER


marker calnexin (Fig. 2A). Immunoblot analysis of extracts from cells expressing the SC-protease and Myc-tagged MAVS found that the SC-protease ␣0 mutant was unable to cleave MAVS at the canonical C508 cleavage site (Fig. 2B). Furthermore, while the loss of the NS3 helix ␣0 in the SC-protease had no effect on proteolytic processing of NS5A/B (Fig. 2C), it ablated processing of the NS4B/5A polyprotein at the canonical cleavage site, which was instead processed at an alternate site to liberate a protein product of higher mass than the canonical NS4B product (Fig. 2D; indicated by the star). Examination of the NS4B/5A polyprotein sequence revealed an alternative consensus serine protease cleavage sequence within NS5A (DPSHIT2AETA; P6, P-1, and P1 residues are in bold) that is likely the alternative site of cleavage of the SC-protease helix ␣0 mutant. Taken together with the previous data, these data demonstrate that in the context of either the SCprotease or the NS3/4A protease complex, the NS3 helix ␣0 is required for cleavage of MAVS and the NS4B/5A polyprotein substrate but not the NS5A/5B polyprotein substrate. Further loss of hydrophobicity in the NS3 helix ␣0 prevents the cleavage of membrane-anchored substrates at membrane-proximal cleavage sites while redirecting proteolytic function to non-membraneassociated cleavage sites of specific substrates. SC-protease cleaves MAVS at an alternate site to block RIG-I pathway signaling. Surprisingly, we found that even though the SC-protease helix ␣0 mutant failed to cleave MAVS at the canonical C508 cleavage site (Fig. 2B), it blocked SenV-mediated signaling to the IFN-␤ promoter at similar levels as the wt SC-protease, while the SC-protease harboring the protease active-site mutation (S139A) did not block this signaling (Fig. 3A). This result was different than that seen with full-length NS34/A, where helix ␣0 mutation blocked signaling regulation by NS3/4A (Fig. 1G). As NS3/4A has been reported to block MAVS oligomerization, which is essential for downstream signaling to IFN-␤ (1), we hypothesized that the SC-protease helix ␣0 mutant, unlike the NS3/4A helix ␣0 mutant, was forming a stable complex with MAVS that would fully block MAVS oligomerization and subsequent down-

jvi.asm.org 3115

Horner et al.

FIG 3 SC-protease cleaves MAVS at an alternate site to block RIG-I pathway signaling. (A) IFN-␤ promoter reporter luciferase expression of 293 cells transfected with increasing amounts of SC-protease expression plasmids and then mock or SenV infected. Values are mean ⫾ SD (n ⫽ 3) of one of four replicate experiments. (Inset) Immunoblot for Flag-SC-protease protein expression from this assay. (B) Immunoblot analysis of anti-Myc immunoprecipitations and input from 293 cells expressing Flag-MAVS, Myc-MAVS, and increasing amounts of indicated SC-protease constructs (wt, S139A, or ␣0) or vector. (C) Entire panel of the Flag-MAVS immunoblot from panel B is shown. Arrows mark full-length (FL), cleaved (C; at C508), and alternate cleaved (C-2) forms of MAVS. The asterisk indicates an alternate form of MAVS not resulting from cleavage. (D) The sequence of the HCV polyprotein cleavage sites, as well as known and possible cleavage sites of MAVS, are shown, with the P1 and P1= residues shown in boldface. (E) Immunoblot analysis of MAVS cleavage at alternate sites by the SC-protease. Blots were probed with anti-Myc (for MAVS) and anti-Flag (for SC-protease). Arrows mark full-length (FL) and cleaved (C, at C508; C-2, at C283) MAVS. (F) IFN-␤ promoter reporter luciferase expression of 293 cells following MAVS wt overexpression alone or with increasing doses of MAVS aa 284 to 540. Values are mean ⫾ SD (n ⫽ 3).

stream signaling to IFN-␤. Indeed, we have shown previously that both SC-protease wt and S139A are able to form a stable complex with MAVS (10). Using a MAVS oligomerization assay that measures interaction between overexpressed Flag-tagged MAVS and Myc-tagged MAVS, we found that expression of the SC-protease with the helix ␣0 mutation only partially blocked MAVS oligomerization, as compared to the complete block in MAVS oligomerization by the wt SC-protease (Fig. 3B). Therefore, the SC-

3116

jvi.asm.org

protease helix ␣0 mutant block of RIG-I pathway signaling is not solely due to the inhibition of MAVS oligomerization. Upon closer analysis of the immunoblots measuring cleavage of MAVS in these assays, we found that while the wt SC-protease cleaves MAVS at the canonical C508 cleavage site and the helix ␣0 mutant of the SC-protease does not, it appeared that both of these constructs cleaved MAVS at a second site, releasing a MAVS cleavage of a lower molecular weight (Fig. 3C and E). This secondary cleav-

Journal of Virology


age of MAVS occurs only upon coexpression of the SC-protease but not with the wt NS3/4A protein, which cleaves MAVS only at C508 (references 14 and 20 and data not shown). Analysis of the MAVS protein sequence revealed additional cysteine residues as candidate serine protease cleavage sites (Fig. 3D). To test if these cysteine residues were the alternate MAVS cleavage sites by the SC-protease, we inactivated them by tyrosine mutation, coexpressed the MAVS constructs with the SC-protease constructs, and monitored cleavage by immunoblot assay. The C508Y mutation in MAVS abolished SC-protease cleavage at the canonical NS3/4A cleavage site (C508) but not at the secondary cleavage site (Fig. 3E). The C452Y mutation in MAVS did not prevent cleavage at the secondary site (data not shown). However, the C283Y mutant prevented MAVS cleavage at the secondary site by both the wt and helix ␣0 mutant SC-protease, thereby identifying C283 as a secondary cleavage site in MAVS by the SC-protease (Fig. 3E). As MAVS lacking a transmembrane domain can function as a dominant negative factor in MAVS pathway signaling via oligomerization of MAVS CARD domains (Y. M. Loo and M. Gale, Jr., unpublished results), we hypothesized that the cleavage at C283 by the SC-protease helix ␣0 mutant liberated a fragment of MAVS that acted as a dominant negative to block RIG-I pathway signaling to IFN-␤. To test this notion, we expressed wt MAVS, along with increasing doses of the truncated MAVS 284-to-540 fragment, and measured signaling to the IFN-␤ promoter in luciferase assays. Indeed, while wt MAVS overexpression induced signaling to the IFN-␤ promoter, coexpression with MAVS 284 to 540 abrogated this effect (Fig. 3F), demonstrating that it acts as a dominant negative to block MAVS signaling. Therefore, while the SCprotease helix ␣0 mutant does not cleave MAVS at C508, it does cleave MAVS at C283, and this resulting cleavage prevents RIG-I pathway signal transduction. Further, as the helix ␣0 mutant does not cleave MAVS at C508, which is positioned at intracellular membranes next to the transmembrane domain of MAVS (aa 514 to 535), our results identify that the NS3 helix ␣0 is required to direct the protease active site of NS3/4A for cleavage of the host immune protein MAVS at amino acids located near the membrane. Helix ␣0 of NS3 is required for HCV replication. To directly test the role of the NS3 helix ␣0 in HCV replication, we engineered the mutant helix ␣0 motif into an HCV genotype 1B subgenomic replicon encoding a G418-selectable marker (the HP replicon) (31) and assessed its ability to mediate persistent replication in Huh7 or Huh7.5 cells transfected with in vitro-transcribed replicon RNA and subjected to G418 selection. While Huh7 cells have an intact RIG-I pathway, Huh7.5 cells lack RIG-I function and are highly permissive for HCV replication (30). Neither Huh7 nor Huh7.5 cells express functional TLR3 (12). While G418-resistant colonies arose in both Huh7 and Huh7.5 cells transduced with the parental replicon RNA, no G418-resistant colonies arose in either Huh7 or Huh7.5 cells transduced with the NS3 helix ␣0 mutant replicon RNA, demonstrating that the hydrophobic amino acids in this domain are essential for HCV replication (Fig. 4A). Moreover, an HCV genotype 2A strain JFH1 engineered with the mutant NS3 helix ␣0 motif also failed to replicate to a level that supported productive infection in Huh7.5 cells after direct transfection of cells with viral RNA (data not shown). HCV replicon RNA harboring NS3 mutations at R24 and R26 displayed reduced transduction efficiency in both Huh7 and Huh7.5 cells (Fig. 4A). To test if any of the hydrophobic amino acids in the NS3


helix ␣0 domain provided a more dominant contribution to either HCV replication or HCV control of innate immunity, we mutated each individual hydrophobic amino acid in this domain to alanine (L13A, L14A, I17A, I18A, L21A), transduced the resulting replicon RNA into Huh7 or Huh7.5 cells, and examined the number of G418-resistant colonies that arose following selection. Remarkably, mutation of the individual hydrophobic amino acids in the NS3 helix ␣0 to alanine did not affect HCV RNA replication in Huh7.5 cells, as these constructs displayed a similar number of G418-resistant colonies as the parental HCV replicon RNA (Fig. 4A, bottom). However, in Huh7 cells, mutation of any of the individual hydrophobic amino acids in the NS3 helix ␣0 resulted in fewer G418-resistant colonies than seen for the parental replicon RNA (Fig. 4A, top). These results demonstrate that in the absence of efficient substrate targeting by the NS3/4A protease, the host cell can impart RIG-I-dependent suppression of viral RNA replication, and reveal that the hydrophobic amino acids in the NS3 helix ␣0 play a role in RIG-I pathway targeting and evasion during HCV replication. Whereas we observed no G418-resistant colonies after 3 weeks of selection of Huh7 cells transfected with the NS3 helix ␣0 mutant replicon RNA in two separate experiments, a sole colony of G418resistant Huh7.5 cells supporting HCV RNA replication eventually grew. This colony was expanded and characterized. We found that both MAVS cleavage and HCV polyprotein processing occurred efficiently in these cells and paralleled levels observed in cells harboring the parental HCV replicon (Fig. 4B). We determined the viral RNA sequence from these cells and found that the predominant HCV RNA species was retained at only one of the originally engineered mutations (L14A) in the NS3 helix ␣0 motif, while the other four mutations reverted back to the parental NS3 sequence (Fig. 4C). Introduction of this mutation (L14A) into the NS3/4A expression construct maintained HCV polyprotein processing by NS3/4A at the NS3/4A, NS5A/B, and NS4B/5A junctions (Fig. 4D and E). Taken together, our data demonstrate that the overall hydrophobic nature of NS3 helix ␣0 is required for efficient HCV RNA replication and contributes to viral RNA replication fitness under pressure from RIG-I signaling of MAVSdependent immune defenses imposed by the host cell. DISCUSSION

Our results demonstrate that the hydrophobic amino acids in the amphipathic NS3 helix ␣0 are required for cleavage of membraneanchored substrates, including the viral HCV NS4B/5A polyprotein and the host cell immune adaptor protein MAVS. Loss of hydrophobicity in the NS3 helix ␣0 results in an inability to establish persistence by HCV RNA replicons and infection by HCV. Further, loss of this domain prevents NS3/4A from controlling RIG-I pathway signaling during RNA virus infection. Therefore, the NS3 helix ␣0 governs control of both viral and cellular processes through targeting of membrane-anchored substrates. The NS3 helix ␣0 is required to position the protease active site to properly cleave membrane-anchored viral and host substrates. Indeed, SC-protease constructs lacking the hydrophobic amino acids in the NS3 helix ␣0 cleave the membrane-anchored substrates NS4B/5A and MAVS at alternate, membrane-distal sites. This selectivity in the cleavage of membrane-anchored substrates could be due to the ability of the NS3 helix ␣0 to sense or induce membrane curvature (19). Our data indicate that the NS4A tar-

jvi.asm.org 3117

Horner et al.

FIG 4 Helix ␣0 of NS3 is required for RNA replication of HCV subgenomic replicon. (A) RNA was in vitro transcribed from the parental HCV-HP replicon construct (wt), HP replicon harboring the NS3 helix ␣0 mutations (␣0), or individual NS3 amino acid mutations and electroporated into Huh7 and Huh7.5 cells,

followed by plating on 100-mm-diameter dishes in serial dilutions, (2 ⫻ 105, ⫻104, and ⫻103 cells) and crystal violet staining after 3 weeks of G418 selection. The relative transduction efficiency (mean ⫾ SD, from 2 experiments) denotes the percent colonies in Huh7 cells relative to wt as compared those in Huh7.5 cells. (B) A single colony was expanded from Huh7.5 cells replicating the HP replicon with the NS3 helix ␣0 mutations (referred to as ␣0REV). Extracts from Huh7.5 cells harboring the wt, the ␣0REV, or no HP replicon (⫺) were assayed by immunoblotting for expression of MAVS and HCV proteins (using patient antisera). Arrows mark full-length (FL) and cleaved (C) MAVS; small arrows mark HCV proteins. (C) The hydrophobic amino acids in wt NS3 helix ␣0 are underlined (top) and the introduced NS3 helix ␣0 mutations are in bold (middle). The sequence of the ␣0 revertant (bottom) shows reversion to the parental NS3 amino acids except at L14A. (D and E) Immunoblot analysis of extracts from 293 cells expressing the indicated NS3 constructs and HCV NS5A/B (D) or NS4B/5A (E) using anti-NS3 and anti-Flag (for Flag-tagged HCV polyprotein substrates). Arrows mark full-length and cleaved HCV polyprotein.

geting subunit of the NS3/4A protease complex also contributes to this process, as even the wt SC-protease, which lacks the NS4A transmembrane anchor, cleaves MAVS at an alternate site (C283) distal from the MAVS transmembrane anchor. Therefore, amino acids within both the NS4A transmembrane anchor and the NS3 interacting domain (aa 21 to 32), as well as the hydrophobic amino acids in the NS3 helix ␣0, contribute to canonical MAVS cleavage and disruption of RIG-I pathway signaling. These results support the model by Brass et al. (4), who performed structural

3118

jvi.asm.org

analysis of the NS3 helix ␣0 and the NS4A transmembrane ␣-helix and proposed that both of these domains are required to properly position the protease active site at the membrane surface for the cleavage of substrates. It is also possible that the NS3 helicase domain, which is not present in the SC-protease, could contribute to this process, as functional communication between the NS3 helicase and protease domains has been reported (3). Our data indicate that the membrane association of the NS3 helix ␣0 is essential to establish HCV RNA replication, even in

Journal of Virology


Huh7.5 cells which do not have functional RIG-I signaling. This is not surprising, considering that the NS3 helix ␣0 mutant has a reduced capacity for processing of the HCV NS4B/5A polyprotein and reduced interaction with NS4A. Further, the fact that the only HCV RNA replicon we obtained after transduction of Huh7.5 cells with HCV RNA containing the five mutations in the NS3 helix ␣0 domain reverted four of these five amino acids back to their exact wt amino acid sequence highlights the essential nature of the hydrophobicity in this domain for NS3 function. By inactivating single hydrophobic amino acids, we were also able to show that the entire hydrophobic nature of the NS3 helix ␣0 is also required for the control of RIG-I pathway signaling in the context of HCV RNA replication. These mutant RNA replicons (L13A, L14A, I17A, I18A, and L21A) were established at equal frequency in the Huh7.5 cells; however, they displayed reduced transduction efficiency in the Huh7 cells, in which the HCV RNA activates RIG-I pathway signaling. Interestingly, we were unable to demonstrate an effect on MAVS cleavage or RIG-I pathway control by overexpressed NS3/4A proteins containing these mutations in cell culture assays (data not shown). Therefore, we hypothesize that to establish RNA replication, there is likely a race between HCV and the host, wherein any small defects in MAVS cleavage or in viral RNA replication/polyprotein processing by NS3/4A will compromise the ability of the viral RNA to establish productive replication during innate immune pressure. The multifunctional HCV NS3/4A protease has emerged as the target of the first set of direct-acting antiviral drugs approved for HCV therapy. In addition to MAVS, NS3/4A has also been reported to cleave the host cell proteins TRIF (13) and T-cell protein phosphatase (5). While it remains to be seen exactly how the cleavage of these molecules impacts viral persistence in vivo, it is clear that the HCV protease has the ability to cleave multiple cellular targets, likely to promote its replication and viral pathogenesis in the host, and these antiviral agents, which act directly against NS3/ 4A, will also inhibit these processes. Examination of the spatialtemporal regulation of NS3/4A-mediated cleavage of viral polyprotein versus that of host cell proteins, as well as the selectivity of these processes, will likely reveal the roles of the host cell protein cleavages in the virus life cycle. Taking these findings together with our previous studies and those of others (4, 9, 36), we propose a model wherein MAVS cleavage by NS3/4A is governed through a bipartite process. In this model, NS4A serves as the primary membrane subcellular targeting subunit to bring NS3/4A to the MAM for the cleavage of MAVS on the MAM, which is directed by the amphipathic helix ␣0 of NS3. The diverse roles of NS3 helix ␣0 in HCV replication and in regulating RIG-I signaling through MAVS cleavage highlight the importance of this domain for productive HCV infection. ACKNOWLEDGMENTS We thank the indicated colleagues for antibodies, C. L. Johnson and K. Chang for technical assistance, C. Wilkins for discussion and critical reading of the manuscript, and the W. M. Keck Center for Advanced Studies in Neural Signaling for microscopy assistance. This work was supported by NIH grants AI060389, AI40035, and DA024563, the Burroughs Wellcome Fund (M.G.), and the Irvington Institute Fellowship Program of the Cancer Research Institute (S.M.H.).


REFERENCES 1. Baril M, Racine ME, Penin F, Lamarre D. 2009. MAVS dimer is a crucial signaling component of innate immunity and the target of hepatitis C virus NS3/4A protease. J. Virol. 83:1299 –1311. 2. Bartenschlager R, Ahlborn-Laake L, Mous J, Jacobsen H. 1994. Kinetic and structural analyses of hepatitis C virus polyprotein processing. J. Virol. 68:5045–5055. 3. Beran RK, Pyle AM. 2008. Hepatitis C viral NS3-4A protease activity is enhanced by the NS3 helicase. J. Biol. Chem. 283:29929 –29937. 4. Brass V, et al. 2008. Structural determinants for membrane association and dynamic organization of the hepatitis C virus NS3-4A complex. Proc. Natl. Acad. Sci. U. S. A. 105:14545–14550. 5. Brenndorfer ED, et al. 2009. Nonstructural 3/4A protease of hepatitis C virus activates epithelial growth factor-induced signal transduction by cleavage of the T-cell protein tyrosine phosphatase. Hepatology 49:1810 –1820. 6. Dixit E, et al. 2010. Peroxisomes are signaling platforms for antiviral innate immunity. Cell 141:668 – 681. 7. Foy E, et al. 2003. Regulation of interferon regulatory factor-3 by the hepatitis C virus serine protease. Science 300:1145–1148. 8. Horner SM, Gale M, Jr. 2009. Intracellular innate immune cascades and interferon defenses that control hepatitis C virus. J. Interferon Cytokine Res. 29:489 – 498. 9. Horner SM, Liu HM, Park HS, Briley J, Gale M, Jr. 2011. Mitochondrial-associated endoplasmic reticulum membranes (MAM) form innate immune synapses and are targeted by hepatitis C virus. Proc. Natl. Acad. Sci. U. S. A. 108:14590 –14595. 10. Johnson CL, Owen DM, Gale M, Jr. 2007. Functional and therapeutic analysis of hepatitis C virus NS3.4A protease control of antiviral immune defense. J. Biol. Chem. 282:10792–10803. 11. Kawai T, et al. 2005. IPS-1, an adaptor triggering RIG-I- and Mda5mediated type I interferon induction. Nat. Immunol. 6:981–988. 12. Li K, Chen Z, Kato N, Gale M, Jr, Lemon SM. 2005. Distinct poly(I-C) and virus-activated signaling pathways leading to interferon-beta production in hepatocytes. J. Biol. Chem. 280:16739 –16747. 13. Li K, et al. 2005. Immune evasion by hepatitis C virus NS3/4A proteasemediated cleavage of the Toll-like receptor 3 adaptor protein TRIF. Proc. Natl. Acad. Sci. U. S. A. 102:2992–2997. 14. Li XD, Sun L, Seth RB, Pineda G, Chen ZJ. 2005. Hepatitis C virus protease NS3/4A cleaves mitochondrial antiviral signaling protein off the mitochondria to evade innate immunity. Proc. Natl. Acad. Sci. U. S. A. 102:17717–17722. 15. Lin C, Pragai BM, Grakoui A, Xu J, Rice CM. 1994. Hepatitis C virus NS3 serine proteinase: trans-cleavage requirements and processing kinetics. J. Virol. 68:8147– 8157. 16. Loo YM, et al. 2008. Distinct RIG-I and MDA5 signaling by RNA viruses in innate immunity. J. Virol. 82:335–345. 17. Loo YM, et al. 2006. Viral and therapeutic control of IFN-beta promoter stimulator 1 during hepatitis C virus infection. Proc. Natl. Acad. Sci. U. S. A. 103:6001– 6006. 18. Love RA, et al. 1996. The crystal structure of hepatitis C virus NS3 proteinase reveals a trypsin-like fold and a structural zinc binding site. Cell 87:331–342. 19. McMahon HT, Gallop JL. 2005. Membrane curvature and mechanisms of dynamic cell membrane remodelling. Nature 438:590 –596. 20. Meylan E, et al. 2005. Cardif is an adaptor protein in the RIG-I antiviral pathway and is targeted by hepatitis C virus. Nature 437:1167–1172. 21. Moradpour D, Penin F, Rice CM. 2007. Replication of hepatitis C virus. Nat. Rev. Microbiol. 5:453– 463. 22. Morikawa K, et al. 2011. Nonstructural protein 3-4A: the Swiss army knife of hepatitis C virus. J. Viral Hepat. 18:305–315. 23. Mottola G, et al. 2002. Hepatitis C virus nonstructural proteins are localized in a modified endoplasmic reticulum of cells expressing viral subgenomic replicons. Virology 293:31– 43. 24. Nomura-Takigawa Y, et al. 2006. Non-structural protein 4A of hepatitis C virus accumulates on mitochondria and renders the cells prone to undergoing mitochondria-mediated apoptosis. J. Gen. Virol. 87:1935–1945. 25. Pietschmann T, Lohmann V, Rutter G, Kurpanek K, Bartenschlager R. 2001. Characterization of cell lines carrying self-replicating hepatitis C virus RNAs. J. Virol. 75:1252–1264. 26. Rapaport D. 2003. Finding the right organelle. Targeting signals in mitochondrial outer-membrane proteins. EMBO Rep. 4:948 –952. 27. Saito T, et al. 2007. Regulation of innate antiviral defenses through a

jvi.asm.org 3119

Horner et al.

28. 29. 30. 31. 32.

shared repressor domain in RIG-I and LGP2. Proc. Natl. Acad. Sci. U. S. A. 104:582–587. Saito T, Owen DM, Jiang F, Marcotrigiano J, Gale M, Jr. 2008. Innate immunity induced by composition-dependent RIG-I recognition of hepatitis C virus RNA. Nature 454:523–527. Seth RB, Sun L, Ea CK, Chen ZJ. 2005. Identification and characterization of MAVS, a mitochondrial antiviral signaling protein that activates NF-kappaB and IRF 3. Cell 122:669 – 682. Sumpter R, Jr, et al. 2005. Regulating intracellular antiviral defense and permissiveness to hepatitis C virus RNA replication through a cellular RNA helicase, RIG-I. J. Virol. 79:2689 –2699. Sumpter R, Jr, Wang C, Foy E, Loo YM, Gale M, Jr. 2004. Viral evolution and interferon resistance of hepatitis C virus RNA replication in a cell culture model. J. Virol. 78:11591–11604. Vitour D, et al. 2009. Polo-like kinase 1 (PLK1) regulates interferon (IFN) induction by MAVS. J. Biol. Chem. 284:21797–21809.

3120

jvi.asm.org

33. Wakita T, et al. 2005. Production of infectious hepatitis C virus in tissue culture from a cloned viral genome. Nat. Med. 11:791–796. 34. Wilkins C, Gale M, Jr. 2010. Recognition of viruses by cytoplasmic sensors. Curr. Opin. Immunol. 22:41– 47. 35. Wolk B, Buchele B, Moradpour D, Rice CM. 2008. A dynamic view of hepatitis C virus replication complexes. J. Virol. 82:10519 –10531. 36. Wolk B, et al. 2000. Subcellular localization, stability, and trans-cleavage competence of the hepatitis C virus NS3-NS4A complex expressed in tetracycline-regulated cell lines. J. Virol. 74:2293–2304. 37. Yao N, Reichert P, Taremi SS, Prosise WW, Weber PC. 1999. Molecular views of viral polyprotein processing revealed by the crystal structure of the hepatitis C virus bifunctional protease-helicase. Structure 7:1353– 1363. 38. Yoneyama M, et al. 2004. The RNA helicase RIG-I has an essential function in double-stranded RNA-induced innate antiviral responses. Nat. Immunol. 5:730 –737.

Journal of Virology