Free access
Research Article
26 January 2015

Mutations across Murine Hepatitis Virus nsp4 Alter Virus Fitness and Membrane Modifications


A common feature of infection by positive-sense RNA virus is the modification of host cell cytoplasmic membranes that serve as sites of viral RNA synthesis. Coronaviruses induce double-membrane vesicles (DMVs), but the role of DMVs in replication and virus fitness remains unclear. Coronaviruses encode 16 nonstructural proteins (nsps), three of which, nsp3, nsp4, and nsp6, are necessary and sufficient for DMV formation. It has been shown previously that mutations in murine hepatitis virus (MHV) nsp4 loop 1 that alter nsp4 glycosylation are associated with disrupted DMV formation and result in changes in virus replication and RNA synthesis. However, it is not known whether DMV morphology or another function of nsp4 glycosylation is responsible for effects on virus replication. In this study, we tested whether mutations across nsp4, both alone and in combination with mutations that abolish nsp4 glycosylation, affected DMV formation, replication, and fitness. Residues in nsp4 distinct from glycosylation sites, particularly in the endoplasmic reticulum (ER) luminal loop 1, independently disrupted both the number and morphology of DMVs and exacerbated DMV changes associated with loss of glycosylation. Mutations that altered DMV morphology but not glycosylation did not affect virus fitness while viruses lacking nsp4 glycosylation exhibited a loss in fitness. The results support the hypothesis that DMV morphology and numbers are not key determinants of virus fitness. The results also suggest that nsp4 glycosylation serves roles in replication in addition to the organization and stability of MHV-induced double-membrane vesicles.
IMPORTANCE All positive-sense RNA viruses modify host cytoplasmic membranes for viral replication complex formation. Thus, defining the mechanisms of virus-induced membrane modifications is essential for both understanding virus replication and development of novel approaches to virus inhibition. Coronavirus-induced membrane changes include double-membrane vesicles (DMVs) and convoluted membranes. Three viral nonstructural proteins (nsps), nsp3, nsp4, and nsp6, are known to be required for DMV formation. It is unknown how these proteins induce membrane modification or which regions of the proteins are involved in DMV formation and stability. In this study, we show that mutations across nsp4 delay virus replication and disrupt DMV formation and that loss of nsp4 glycosylation is associated with a substantial fitness cost. These results support a critical role for nsp4 in DMV formation and virus fitness.


RNA viruses modify host cytoplasmic membranes for the formation of viral replication complexes (13). Coronaviruses (CoVs) induce substantial membrane rearrangements, including a reticulovesicular network composed of two types of membrane modifications, double-membrane vesicles (DMVs) and convoluted membranes (CM). The reticulovesicular network is contiguous with the endoplasmic reticulum (ER) membranes and is the site of viral RNA synthesis (4). CoV genomes, which are 27 to 32 kb, encode two replicase/transcriptase open reading frames (ORFS), orf1a and orf1b, which are translated into the polyproteins 1a and 1ab incorporating nonstructural proteins 1 to 16 (nsps 1 to 16). The replicase polyproteins are cleaved by virus-encoded proteases within nsp3 (PLP1/2) and nsp5 (3CLpro). All nsps tested to date have been shown to colocalize by immunoelectron microscopy to DMVs and/or CM and by immunofluorescence at probable sites of viral RNA synthesis (411).
Nonstructural proteins 3, 4, and 6 are integral membrane proteins, and the topology of these proteins has been determined in vitro for both severe acute respiratory syndrome-CoV (SARS-CoV) and murine hepatitis virus (MHV) (1113). nsp3 contains two transmembrane domains, nsp4 contains four transmembrane domains, and nsp6 contains six transmembrane domains (1417). SARS-CoV nsp3, nsp4, and nsp6 are necessary and sufficient for double-membrane vesicle formation. When expressed alone, nsp3 causes membrane proliferation, and nsp6 induces single-membrane vesicles. When coexpressed, nsp3 and nsp4 together have the capacity to pair membranes; however, all three proteins are required for formation of DMVs (3). Likely, interactions among these proteins mediate membrane modifications. Hagemeijer et al. demonstrated by using immunoprecipitation and protein complementation assays that MHV nsp4 forms homotypic and heterotypic interactions with the transmembrane domains of nsp3 and nsp6 (18). The region of nsp4 that interacts with nsp3 and nsp6, as well as the regions of these proteins required for membrane modifications, remains unknown.
nsp4 contains four transmembrane domains and three loop regions. Loops 1 and 3 are ER luminal, whereas loop 2 and the N and C termini are cytosolic (13, 14). nsp4 is likely required for replication, as demonstrated by the inability to recover a virus with an nsp4 deletion (5). Deletion of transmembrane 4 (TM4) and truncation of the C terminus of nsp4 allow recovery of viable mutants; however, these viruses have altered viral replication and RNA synthesis (5). Additionally, charge-to-alanine mutagenesis of nsp4 identified viral mutants having a range of replicative capacities, ranging from wild-type (WT)-like replication to delayed exponential replication with a 3-log10 reduction in peak viral titers (5).
MHV nsp4 is N-linked glycosylated at two asparagine residues (N176 and N237) within loop 1, as demonstrated by endo-β-N-acetylglucosaminidase H (endo-H) sensitivity (1, 12, 13). Glycosylation has many functions, including protein folding, sorting, and trafficking (19). SARS-CoV nsp4 also has been demonstrated to be glycosylated in vitro at an atypical glycosylation site, N131IC (13). All coronavirus nsp4 proteins tested to date contain at least one predicted glycosylation site; however, the location and number of the glycosylation sites vary (data not shown). Our lab previously generated an MHV nsp4 mutant lacking both glycosylation sites (N176A/N237A), referred to as the DGM (double-glycosylation mutant). The DGM virus exhibited delayed exponential replication and a reduction in RNA synthesis and formed DMVs with aberrant morphologies (1). These observations suggested a direct link between the capacity of the virus to induce stable DMVs and virus fitness. However, recently, the relationship of DMV size and number to virus fitness has been questioned (20). Since nsp4 is directly involved in DMV formation, we sought to test the role of the DGM and other mutations across nsp4 on the stability and number of DMVs. Additionally, we determined the fitness of these viruses over multiple passages compared to wild-type and the DGM viruses. Our data demonstrate that mutations across nsp4 domains, in addition to glycosylation mutants, result in alterations of DMV morphology and number and that loss of nsp4 glycosylation is clearly associated with fitness cost when directly competed with WT virus.


WT virus, previously reported mutant viruses, cells, and antibodies.

Recombinant MHV A59 (GenBank accession number AY910861) virus was used as a WT control virus for all experiments. The double-glycosylation mutant (DGM; N176A/N237A), VUJS11 (K44A/D47A), VUJS17 (E226A/E227A), and N258T viruses were previously described (1, 5, 21). Delayed brain tumor (DBT) cells and baby hamster kidney (BHK) cells were grown in Dulbecco's modified Eagle medium (DMEM; Gibco) supplemented with 10% fetal bovine serum, 1% HEPES, 1% penicillin-streptomycin, and 0.1% amphotericin B (complete DMEM). Medium for the BHK-MHV receptor (MHVR) cells was supplemented with G418 (Mediatech) at 0.8 mg/ml to maintain selection of MHVR. Rabbit polyclonal antibodies were used for biochemical studies and immunofluorescence directed at the viral protein, nsp4 (VU158) (5).


In order to introduce substitutions into nsp4, an MHV reverse genetics system was used. Briefly, the MHV genome was divided into seven plasmids, with nsp4 spanning two of these plasmids (pCR-XL-pSMART B and pCR-XL-pSMART C). Nucleotides 8721 to 9555 of the MHV A59 clone were located in fragment B, and nucleotides 9556 to 10208 were in fragment C. Substitutions were introduced into the B or C fragment by PCR mutagenesis using the QuickChange method (Stratagene) and the primers listed in Table 1. Changes to the manufacturer's protocol include the use of Pfu Turbo and the following PCR conditions: initial denaturation at 95°C for 2 min, followed by 16 cycles of denaturation at 95°C for 30 s, annealing at temperatures dependent on the primers for 1 min, and extension for 10 min at 72°C. All B and C fragment plasmids containing mutations in nsp4 were sequenced to ensure that PCR amplification did not introduce additional changes in the coding region.
TABLE 1 Primers used for alternate NXS/T mutagenesis
Primer nameSequenceaPurpose
K85S sense5′-CCGCAACTCTTTCGCTTGTCCTG-3′Mutagenesis for K85S
K85S antisense5′-CAGGACAAGCGGAAGAGTTGCGG-3′Mutagenesis for K85S
P106N antisense5′-CTTAAAACTGTGGTGTTAACATTAAATAAG-3′Mutagenesis for P106N
V129N sense5′-CTACTGATAGCAACCAGTGTTACACGC-3′Mutagenesis for V129N
V129N antisense5′-GCGTGTAACACTGGTTGCTATCAGTAG-3′Mutagenesis for V129N
Q130I sense5′-CTACTGATAGCAACATATGTTACACGC-3′Mutagenesis for Q130I
Q130I antisense5′-GCGTGTAACATATGTTGCTATCAGTAG-3′Mutagenesis for Q130I
C133N sense5′-GATAGCGTGCAGAACTACACGCCAC-3′Mutagenesis for C133N
C133N antisense5′-GTGGCGTGTAGTTCTGCACGCTAG-3′Mutagenesis for C133N
V214N sense5′-GTGCGTGTTAACCGCACTCGC-3′Mutagenesis for V214N
V214N antisense5′-GCGAGTGCGGTTAACACGCAC-3′Mutagenesis for V214N
E355N sense5′-CAACACTTATATTCGAAGGG-3′Mutagenesis for E355N
E355N antisense5′-CCCTTCGAATATAAGTGTTG-3′Mutagenesis for E355N
N479S sense5′-CATAATAATGGTTCCGATGTTCTC-3′Mutagenesis for N479S
N479S antisense5′-GAGAACATCGGAACCATTATTATG-3′Mutagenesis for N479S
Bold letters denote nucleotides used to introduce mutations.

Virus recovery.

Viruses containing the nsp4 mutations were generated using a reverse genetics system for MHV A59 as described by Yount et al. (22) and modified by Denison et al. (23) and Sparks et al. (5). Briefly, the MHV A59 genome was divided into seven cDNA fragments, which were digested using the appropriate restriction enzymes. These digested fragments were then ligated at 16°C overnight before the DNA was purified, in vitro transcribed, and electroporated into BHK-MHVR cells along with N gene transcripts. Electroporated cells were cocultured with DBT cells and incubated at 37°C until cytopathic effects (CPE) were seen. The cytopathic effect see in MHV-infected cells is the formation of multinucleated giant cells (syncytium formation). The virus produced from electroporated cells (passage 0 [P0]) was passaged onto uninfected DBT cells to generate a P1 stock virus that was used for all experiments. The P1 virus was sequenced across the region of the nsp3 to nsp6 genes to ensure that no additional mutations were present in the membrane proteins. If the in vitro-transcribed genome did not produce virus on the first attempt, virus assembly was attempted at least two additional times.

Reverse transcription-PCR (RT-PCR) and sequencing.

Total intracellular RNA was isolated using TRIzol (Invitrogen) and the manufacturer's protocol. Viral RNA was then reverse transcribed using Superscript III (Invitrogen) and random hexamers (Roche). The coding sequences of nsp3 through nsp6 were amplified by PCR to generate four amplicons covering nucleotides 2389 to 13219. The amplicons generated were directly Sanger sequenced to analyze retention of the engineered mutations and absence of additional mutations.

Viral replication assay.

DBT cells were infected with either WT or nsp4 mutant viruses at a multiplicity of infection (MOI) of 1 PFU/cell and absorbed for 30 min. Cells were then washed twice with phosphate-buffered saline (PBS), medium was replaced, and cells were incubated at 37°C. Supernatants were sampled from 30 min to 24 h postinfection (p.i.), and viral titers were determined by plaque assay, as previously described (24).

Radiolabeling of viral proteins and protein immunoprecipitation.

DBT cells were infected at an MOI of 10 PFU/cell or mock infected. At 4 h p.i., the medium was replaced with DMEM lacking cysteine and methionine and supplemented with 20 μg/ml actinomycin D (ActD; Sigma). At 5 h p.i., [35S]methionine/cysteine ([35S]Met-Cys) was added to the cells and monitored for CPE. When cells were 90 to 100% involved in CPE, lysates were harvested in lysis buffer (1% NP-40, 0.5% sodium deoxycholate, 150 mM NaCl, and 50 mM Tris, pH 8.0). Lysates were immunoprecipitated as previously described (1).

Immunofluorescence assay.

DBT cells on glass coverslips were mock infected or infected with WT or nsp4 mutant viruses. At 7 h p.i., medium was aspirated, and cells were fixed in 100% methanol at −20°C. Cells were then rehydrated in PBS for 10 min and blocked in PBS containing 5% bovine serum albumin. For indirect immunofluorescence, cell were incubated with nsp4 in wash solution (1% bovine serum albumin and 0.05% NP-40 in PBS) containing 2% normal goat serum for 45 min. Cells were then washed three times for 5 min each time in wash solution. Cells were then incubated with goat anti-rabbit Alexa-488 at 1:1,000 for 30 min at room temperature. Anti-nsp8 directly conjugated to Alexa-546 was used for direct immunofluorescence, as previously described (1). Cells were incubated with anti-nsp8 at a 1:200 dilution for 30 min at room temperature. Cells were then washed three times in wash solution for 5 min each before a final wash in PBS for 30 min. PBS was then aspirated and replaced with distilled water. Coverslips were mounted to glass slides using Aqua-Poly/Mount (Polysciences) and visualized using a Zeiss LSM 510 Meta inverted confocal microscope with a 40× oil immersion lens. Images were merged and processed using Adobe Photoshop CS5.


Total intracellular RNA was extracted in TRIzol (Invitrogen) using the manufacturer's protocol. Viral RNA (1 μg) was then reverse transcribed using Superscript III (Invitrogen) and random hexamers (Roche). Quantitative RT-PCR (qRT-PCR) was performed as previously described (25). Briefly, quantitative PCR (qPCR) was performed on the RT product using an Applied Biosciences 7500 real-time PCR system with Power SYBR green PCR master mix (Life Technologies). Values were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH). A Kruskal-Wallis test was used to determine statistical differences between WT and mutant levels of RNA synthesis.


DBT cells were mock infected or infected with WT or nsp4 glycosylation mutant viruses at an MOI of 5 PFU/cell in three 60-mm dishes per sample and incubated at 37°C. At 8 h p.i., medium was aspirated, and cells were washed once with PBS. The cells were then fixed in 2% glutaraldehyde for 10 min, scraped off the dishes, and pelleted. The initial 2% glutaraldehyde was aspirated, fresh 2% glutaraldehyde was added to the fixed cells for 1 h and aspirated, and then fresh glutaraldehyde was added to the fixed cells for overnight incubation at 4°C. Cells were washed three times in PBS, transferred to 1% osmium tetroxide in distilled H2O (diH2O) for 1 h, and washed three times in diH2O. Cells were stained en bloc in 1% aqueous uranyl acetate for 1 h and washed three times in diH2O. Dehydration of cells was carried out gradually using a graded series of ethanol and increasing the times each sample remained in solution, starting with 30%, followed by 50%, 70%, 95%, and finally absolute ethanol. Propylene oxide was used as a transitional solvent to replace the dehydration solution. Cells were transferred to a 1:1 PolyEmbed 812-propylene oxide mixture for 1 h and then placed in pure PolyEmbed 812 for four exchanges over 36 h. Pure resin specimens were then transferred into capsules containing fresh resin and finally placed into an oven overnight to polymerize. Ultrathin serial sections (50 to 60 nm) from polymerized blocks were obtained using a Leica UCT ultracut microtome (Leica Microsystems, Vienna, Austria), transferred to Formvar-coated slot grids, and examined using a Phillips/FEI T-12 transmission electron microscope (TEM; FEI Company, Hillsboro, OR) equipped with an ATM XR41-S side-mounted charge-coupled-device (CCD) camera (2,000 by 2,000 pixels; Advanced Microscopy Techniques Corp., Woburn, MA). For mock-infected cells, fields were chosen randomly. For virus-infected cells, cells were imaged if there were signs of infection, which included DMVs, convoluted membranes, and virions. DMVs were categorized as either normal or aberrant (inner membrane appears collapsed). For statistical analysis, the area of cytoplasm was calculated using ImageJ software (26), and the number of DMVs per area of cytoplasm was determined. A Kruskal-Wallis test was used to determine statistical significance of DMVs per area among the viruses.

Competition assay.

DBT cells were infected at a final MOI of 0.1 PFU/cell with the WT and DGM, with the DGM and DGM-K85S, and with the DGM and N258T viruses at ratios of 1:1. Supernatants were collected for passage, and total RNA was harvested in TRIzol. Supernatants were passaged three times. RNA extraction and RT-PCR were performed as described above. The nsp4 amplicon electropherograms were then analyzed using MacVector, version 13, to measure the area under the curve for each mutation. Then for each virus pair, the percentage of virus A nucleotides compared to virus B nucleotides was calculated. At least three nucleotide positions per pair of viruses was analyzed and averaged to calculate the final percentages. There were at least two biological replicates for each competition.


Mutations across nsp4 impair DMV morphology and numbers.

In order to test whether nsp4 glycosylation regulates proper DMV formation independent of its position in the protein, we engineered mutations to introduce a glycosylation sequon, NX(S/T) at different locations in the nsp4 DGM (double-glycosylation mutant, N176A/N237A) background (Fig. 1). Two viruses were engineered to insert predicted glycosylation sites from other coronaviruses; V129N/Q130I is the glycosylation sequon of SARS-CoV, and N479S introduced the sequon present in both human CoV (HCoV) OC43 and bovine CoV (BCoV). All other mutations were inserted near highly conserved asparagine, serine, or threonine residues to introduce new potential NX(S/T) sequons (Fig. 1). Both strategies have successfully been used to introduce glycosylation sites into proteins (2730). We recovered three mutants in the DGM background and one in the WT background; two DGM viruses contained NX(S/T) mutations from other coronaviruses (DGM-V129N/Q130I and DGM-N479S) one DGM virus had a mutation that was introduced after a conserved asparagine (DGM-K85S). Additionally, N479S was recovered in the WT nsp4 background. Recovered viruses were sequenced across the region of nsp3 through nsp6 and contained no additional mutations. Recovery of all other engineered genomes was attempted at least three times with no signs of virus-induced CPE.
FIG 1 Engineering nsp4 mutants. (A) Proposed topology of nsp4: nsp4 has four transmembrane regions (TM1 to TM4) and three loops (loop 1 to 3). Mutations in nsp4 tested in this study are shown on the diagram. Red circles represent nonviable mutations; green circles represent viable viruses that were recovered in this study. The double-headed arrows represent native glycosylation sites within MHV nsp4. (B) nsp4 mutants were engineered with alternate NX(S/T) sites (gray double-headed arrows) in the DGM background. The locations of the native NX(S/T) sequons (black double-headed arrows) are shown. Viruses are identified with the introduced NX(S/T) site shown below the designation. *, previously recovered viruses (1).
To determine whether the mutations that introduce potential alternate glycosylation sequons restored proper DMV morphology, murine delayed brain tumor (DBT) cells were infected with WT or nsp4 mutant viruses and analyzed for membrane changes by TEM (Fig. 2). Mock-infected cells exhibited regular cellular architecture and normal organelles. WT-infected cells displayed swollen ER and Golgi compartments, as well as the presence of several virus-induced structures, including DMVs, convoluted membranes (CM), and vesicles containing newly formed virions (Fig. 2). The DMVs in WT-infected cells had membranes containing two lipid bilayers in close proximity to one another. Similarly, DGM-infected cells had swollen ER and Golgi compartments; however, the DMVs were aberrant. Aberrant DMVs in this study are defined as vesicles having inner membranes that are separated or collapsed away from the outer membrane. The degree of inner membrane collapse varied; however, the majority of aberrant DMV inner membranes were completely collapsed and appeared as electron-dense structures at one side of the vesicles. This is in contrast to the DMV artifacts with spiderweb-like content seen with standard electron microscopy (EM) of SARS-CoV-infected cells (7). All DMVs considered normal in this study had both lipid bilayers intact and contained no spiderweb-like contents. Each of the viruses that contained engineered NX(S/T) sequons exhibited aberrant DMVs similar to the DGM virus. The numbers of normal and aberrant DMVs were calculated for each virus. DGM-K85S- and DGM-V129N/Q130I-infected cells produced increased numbers of aberrant DMVs (88% and 78%, respectively) compared to DGM (67%) (Fig. 2 and Table 2). Cells infected with DGM-N479S and N479S viruses produced fewer aberrant DMVs (54% and 53%, respectively) than DGM but more aberrant DMVs than the WT (37%). Next, the number of DMVs per area of cytoplasm was calculated to determine whether the differences in total numbers of DMVs counted per sample corresponded to differences in total numbers of DMVs (Fig. 2). DGM, DGM-K85S, and N479S had significantly decreased numbers of DMVs compared to the WT (Fig. 2I). The differences in total numbers of DMVs may be due to direct effects of mutations in nsp4 or may reflect differences in the degree of viral replication occurring. These data demonstrate that inserting additional mutations within loop 1 of nsp4 exacerbates the aberrant DMV formation associated with the DGM phenotype. Additionally, these data demonstrate that mutations within the C terminus of nsp4 cause aberrant DMVs independent of mutations within loop 1.
FIG 2 Electron microscopic analysis of alternate NX(S/T) sequon mutants. (A to F) DBT cells were mock infected (A) or infected with the WT (B), DGM (C), DGM-K85S (D), DGM-V129N/Q130I (E), DGM-N479S (F), or N479S (G) virus at an MOI of 5 PFU/cell for 8 h before being fixed in 2% glutaraldehyde and processed for TEM. Black arrowheads indicate normal DMVs; the white arrowheads indicate aberrant DMVs. CM, convoluted membranes; N, nucleus; M, mitochondrion; E, endosome; VV, virions in vesicles; ER, endoplasmic reticulum. Scale bar, 500 nm. (H) Normal and aberrant DMVs were quantified for each virus. Percentages are shown as indicated on the figure. The total number of DMVs analyzed is shown above each bar. (I) The numbers of DMVs per area of cytoplasm were calculated and are shown on the graph. Each circle represents an individual field, and the bars show the means ± standard deviations. Fields were chosen based on the presence of signs of virus infection that included DMVs, convoluted membranes, and virions. A Kruskal-Wallis test was used to analyze for significant differences in numbers of DMVs per area of cytoplasm. *, P = 0.008; **, P < 0.0001 (compared to the WT).
TABLE 2 Quantification of normal and aberrant DMVs
VirusNo. of DMVs
TotalNormalAberrant (% of total)
WT623393230 (37)
DGM19865133 (67)
DGM-K85S34440304 (88)
DGM-V129I/Q130I577126451 (78)
DGM- N479S438202236 (54)
K44A/D47A308101207 (67)
E226A/E227A32696230 (71)
N258T412128284 (69)
N479S587310277 (53)

nsp4 NX(S/T) sequons do not complement defects in replication or nsp4 glycosylation.

Next, we sought to determine whether the viruses with aberrant DMVs were associated with altered replication kinetics. DBT cells were mock infected or infected with viruses indicated in Fig. 3 at an MOI of 1 PFU/cell. At indicated times, supernatants were sampled, and titers were determined by plaque assay (Fig. 3). WT virus began exponential replication between 4 and 6 h p.i. and achieved a peak titer at 10 h p.i. The DGM virus replicated with delayed kinetics, began exponential replication between 6 to 8 h p.i., and achieved a WT-like peak titer at 12 h p.i. N479S replicated indistinguishably from the WT, and the DGM-K85S and DGM-N479S viruses replicated indistinguishably from the parental DGM virus. The DGM-V129N/Q130I mutant displayed delayed exponential replication and peak titer compared with both the WT and the parental DGM viruses.
FIG 3 Replication kinetics and glycosylation of alternate NX(S/T) sequon viruses. (A) DBT cells were infected with the indicated viruses at an MOI of 1 PFU/cell. Supernatants were sampled from 0 to 16 h p.i., and titers were determined by plaque assay. Error bars represent the standard errors of the means of three replicates plated in duplicate. (B) DBT cells were infected with the indicated viruses at an MOI of 10 PFU/cell. At 4 h p.i., cells were starved in DMEM lacking Met and Cys and treated with ActD for 1 h before being radiolabeled with [35S]Met-Cys. At 7 h p.i., lysates were harvested and immunoprecipitated with antibodies specific for nsp4 and treated in the presence or absence of endo-H. Proteins were resolved by SDS-PAGE (n ≥ 2). DGM-VQ/NI, DGM-V129N/Q130I.
These data suggested that either glycosylation at alternate residues did not complement the replication delay of DGM or that these nsp4 proteins were not glycosylated. To determine the glycosylation status of the mutant nsp4 proteins, virus-infected cells were radiolabeled, and cell lysates were immunoprecipitated with antibodies specific for nsp4, followed by treatment with endo-H to remove N-linked glycans (Fig. 3). WT nsp4 migrated at 44 kDa in the absence of endo-H and displayed a mobility shift to 39 kDa following endo-H treatment, indicative of the removal of two glycans, each with an expected mobility shift of 2.5 kDa. The N176A virus was glycosylated at one site (42 kDa), and DGM lacked glycosylation and migrated at 39 kDa. In contrast, all of the alternate NX(S/T) mutant viruses in the DGM background migrated at 39 kDa regardless of endo-H treatment, indicating that the mutant nsp4 proteins were not glycosylated. The N479S nsp4 protein migrated at 44 kDa in the absence of endo-H and shifted to 39 kDa in the presence of endo-H, demonstrating that this nsp4 protein is glycosylated at the native glycosylation sites only. While we can draw no conclusions from an inability to recover virus mutants, we did observe that all recovered viruses failed to confer glycosylation at the introduced sequon, whereas compared with a large number of mutations that are tolerated in nsp4, several new NX(S/T) sequons in loop 1 could not be recovered after multiple attempts. The results led us to consider the possibility that sequons capable of glycosylation may not be tolerated in loop 1.

nsp4 NX(S/T) mutants have WT-like RNA synthesis.

Since the differences in total DMV numbers demonstrated by EM may reflect overall differences in viral replication, next we assessed the RNA synthesis capacities of these viruses. DBT cells were infected at an MOI of 1 PFU/cell for 10 h. Total cellular RNA was harvested in TRIzol, and qRT-PCR was performed to amplify nsp10 and GAPDH. All viruses had RNA synthesis levels indistinguishable from the WT level (Fig. 4). It is likely that, similar to the replication kinetics, if time points were taken, there would have been a delay in RNA synthesis. These data suggest that the viruses have no significant decrease in viral replication at 10 h p.i. and that aberrant DMVs are capable of supporting RNA synthesis.
FIG 4 RNA synthesis of nsp4 NX(S/T) mutants. DBT cells were infected with the indicated viruses at an MOI of 1 PFU/cell for 10 h. Cells were then harvested in TRIzol, and genomic RNA was extracted. Genomic RNA levels were determined by qRT-PCR using primers specific to Orf1a. RNA levels were normalized using the 2−ΔCT (where CT is threshold cycle) method to endogenous GAPDH expression. Mean values ± standard errors of the means are shown (n ≥ 3). RNA synthesis levels were not significantly different from the WT level according to a Kruskal-Wallis test.

nsp4 NX(S/T) mutants localize to the replication complex.

In order to determine whether differences in total numbers of DMVs correspond to differences in replication complex formation and overall protein levels, DBT cells on coverslips were infected with the WT or the nsp4 NX(S/T) viruses at an MOI of 5 PFU/cell for 6.5 h (Fig. 5). Cells were then fixed in methanol and stained with antibodies specific to nsp4 and nsp8, a marker for the replication complex. In WT-infected cells, nsp4 and nsp8 colocalized at punctate perinuclear foci. All mutant nsp4 proteins extensively colocalized with nsp8 at perinuclear foci. These data demonstrate that the NX(S/T) mutant nsp4 proteins localize to the replication complex. Additionally, there is no apparent correlation between aberrant DMV formation and the level of replication complex formation or protein expression. Collectively, the RNA synthesis and immunofluorescence data indicate that the difference in total numbers of DMVs does not reflect overall replication or protein expression within these infected cells.
FIG 5 nsp4 localization in NX(S/T) mutant-infected cells. DBT cells were infected with the indicated viruses at an MOI of 5 PFU/cell for 6.5 h. Cells were then fixed in 100% methanol and stained with antibodies specific for nsp4 (green) and nsp8 (red). Yellow pixels represent colocalization of red and green pixels. Scale bar, 20 μm.

Substitutions in nsp4 loop 1 alter replication kinetics and DMV morphology independent from glycosylation status.

Having demonstrated that mutations in nsp4 loop 1 exacerbate the aberrant DMV morphology associated with the DGM phenotype, we tested whether any mutation in nsp4 loop 1, independent of glycosylation potential, impacted replication and DMV morphology/numbers. nsp4 loop 1 mutants previously reported from our lab were tested (Fig. 6); these consist of mutants K44A/D47A (VUJS11), E226A/E227A (VUJS17), and N258T (5, 21). In contrast to WT MHV-A59, all three mutant viruses exhibited a range of replication defects. The N258T mutant was minimally delayed but achieved WT-like titers, the K44A/D47A mutant was more delayed in both exponential replication and peak titer, and the E226A/E227A mutant showed a significant delay but achieved WT-like peak titer (Fig. 6B). We then tested for the glycosylation status of the mutant nsp4 proteins by radiolabeling and immunoprecipitation. All nsp4 loop 1 mutant nsp4 proteins migrated at 44 kDa in the absence of endo-H and at 39 kDa in the presence of endo-H (Fig. 6C). These results show that all tested nsp4 loop 1 mutants are glycosylated at both native asparagine residues. Thus, the observed replication defects were not due to changes in the glycosylation status of the protein. Finally, we examined the DMV morphology and numbers in mutant virus-infected cells (Fig. 7 and Table 2). Similar to the N479S mutation alone, each of the nsp4 mutant virus-infected cells produced aberrant DMVs. The percentage of aberrant DMVs was calculated for each virus and showed an increase compared to the WT level. All of the nsp4 loop 1 mutant viruses produced numbers of aberrant DMVs similar to level in the DGM mutant. Infection with K44A/D47A and E226A/E227A also resulted in significantly fewer DMVs than infection with WT virus. In contrast, N258T produced a similar number of DMVs as the WT. Thus, our results indicate that in addition to mutations that affect glycosylation of nsp4, substitutions at other locations in nsp4 loop 1 result in altered DMV morphology.
FIG 6 Replication and glycosylation of nsp4 mutants. (A) Schematic of nsp4 showing the location of the nsp4 mutations. (B) DBT cells were infected with the indicated viruses at an MOI of 1 PFU/cell for 24 h. At the indicated time points, supernatants were sampled, and titers were determined by plaque assay. Error bars represent the standard errors of the means of three replicates plated in duplicate. (C) DBT cells were infected with the indicated viruses at an MOI of 10 PFU/cell. Cells were starved with DMEM lacking Cys and Met in the presence of ActD for 1 h prior to being radiolabeled with [35S]Cys-Met for 2 h before lysates were harvested. Lysates were immunoprecipitated with antibodies specific to nsp4 in the presence or absence of endo-H, and proteins were resolved by SDS-PAGE (n = 2).
FIG 7 EM of nsp4 mutant-infected cells. (A to F) DBT cells were mock infected (A) or infected with the WT (B), DGM (C), K44A/D47A (D), E226A/E227A (E), or N258T (F) virus for 8 h before fixation in 2% glutaraldehyde and processed for TEM. Black arrowheads indicate normal DMVs, and white arrowheads indicate aberrant DMVs. N, nucleus; M, mitochondrion; VV, virion in vesicles; E, endosome. (G) Normal and aberrant DMVs were quantified, and results are shown as percentages of total DMVs, as indicated on the figure. The total number of DMVs analyzed is shown above each bar. (H) The total number of DMVs per area of cytoplasm was calculated. Circles represent the number of DMVs per area of cytoplasm for a single field. Bars represent the means ± standard deviations. A Kruskal-Wallis test was used to analyze for significant differences in numbers of DMVs per area of cytoplasm. *, P = 0.008; **, P < 0.0001 (compared to the WT).

Loss of nsp4 glycosylation results in decreased virus fitness.

We next tested the effect of nsp4 mutations on virus competitive fitness. In order to test the fitness of nsp4 mutant viruses compared to the WT or the DGM, DBT cells were coinfected at a ratio of 1:1 and passaged three times at 37°C (Fig. 8A). The N258T and N479S viruses competed equally with the WT and were maintained at about 50% of the population. In contrast, viruses lacking glycosylation of nsp4 exhibited profoundly decreased fitness compared to the WT. We next competed DGM-K85S and DGM-N479S with the DGM to test the effects of additional mutations within the DGM virus on fitness (Fig. 8B). When K85S was introduced into DGM nsp4 loop 1 (DGM-K85S), virus fitness was further deceased relative to that of the DGM. However, DGM-N479S competed equally with DGM, and each remained at approximately 50% of the population. The results suggest that the decrease in fitness of DGM-N479S compared to the WT is likely due to the DGM phenotype and that the N479S mutation does not alter virus fitness. Collectively, these data demonstrate that loss of nsp4 glycosylation is associated with a substantial decrease in virus fitness that can then be further decreased by introduction of additional mutations into DGM nsp4 loop 1.
FIG 8 Competition assay of nsp4 mutants. DBT cells were infected with the indicated pairs of viruses at a total MOI of 0.1 PFU/cell at a ratio of 1:1. When cells were at least 50% involved in cytopathic effect, supernatants were collected, and cell monolayers were harvested in TRIzol. Supernatants were used for subsequent passages, for a total of three passages. Total RNA was extracted, and nsp4 amplicons were generated by RT-PCR and sequenced. For residues of interest, the area under the peak was calculated using MacVector, version 13. Then, the percentage of nucleotides of virus A to virus B was calculated and plotted on the graph. Error bars represent standard errors of the means (n ≥ 2).


The role of host cytoplasmic membrane modifications in coronavirus replication and fitness is a subject of increasing investigation and interest for understanding replication, pathogenesis, and possible pathways for broad-spectrum inhibition of coronavirus infection. Recently, nsp3, nsp4, and nsp6 have been shown to mediate membrane modifications during individual expression and coexpression in cells (3). Further, recent studies with a novel inhibitor have shown that drugs targeting coronavirus membrane modifications are associated with profoundly impaired replication (31). Studies have shown that nsp4 is necessary for DMV formation and that mutations at nsp4 glycosylation sites result in aberrant DMVs (1, 3). In the current study, we show that mutations across nsp4, independently or in combination with mutations that abolish glycosylation, cause or exacerbate defects in DMV formation. Further, we show that loss of nsp4 glycosylation is associated with a substantial fitness cost.

Intact nsp4 is required for proper DMV morphology.

We previously reported that mutations in nsp4 loop 1 were important for efficient RNA synthesis (5). We also described glycosylation of nsp4 at N-linked glycosylation sites in loop 1 and the negative effect on DMV formation by elimination of these sites (1). Our current results indicate that multiple residues and domains within nsp4, even in the C-terminal non-TM domain, impact DMV morphology rather than glycosylation alone. In addition, mutations in other regions of nsp4 loop 1 exacerbated changes in membrane structures of the DGM virus. Angelini et al. demonstrated that when coexpressed, nsp3 and nsp4 have the capacity to pair membranes (3). It is likely that nsp3 and nsp4 interact across the DMV lipid bilayer to hold both membranes in close proximity. In this model, any perturbation of the nsp3-nsp4 interaction could result in separations of the inner and outer membranes. Our present results in fact support that idea that it is possible to dysregulate nsp4 functions in DMV formation by mutations at several locations in nsp4 loop 1, as well as other locations, including the C-terminal 10-kDa region, not predicted to be luminal or transmembrane. Since the mutation in the C terminus induced aberrant DMV formation, this would suggest that the nsp4 C terminus is also involved in DMV formation or stability after formation. We previously have demonstrated that the nsp4 C-terminal 10-kDa portion of the protein is dispensable for viral replication in culture (5). The crystal structures of the C termini of the nsp4s for MHV-A59 and feline CoV have been determined and are structurally conserved (32, 33). The N479S residue maps to a surface-exposed loop in the MHV A59 C terminus crystal structure. The DMV morphology phenotypes were the same between N479S and DGM-N479S even though the viruses exhibited differences in replication kinetics (Fig. 2 and 3). This suggests that even mutations that do not affect viral replication can alter DMV morphology and numbers. Our data, in combination with other studies, suggest that DMV presence and not morphology is critical for efficient viral replication (31, 34).

Membrane modifications and virus fitness.

All published studies to date are similar in demonstrating that disruption of “normal” DMV formation is not necessarily lethal to virus replication and likely represents an evolutionarily optimized process for maximum organization of replication components (1, 20, 31, 34). However, the direct role attributed to DMVs in virus replication and fitness has recently been called into question by a study by Al-Mulla et al. which used temperature-sensitive (ts) mutants in multiple replicase proteins to study the relationship of DMV size and number to virus fitness (20). In our study, we tested whether mutations within nsp4, one of the proteins directly involved in DMV formation and stability, affect DMV morphology and viral fitness. Our results support the conclusions of Al-Mulla et al. by demonstrating that alterations in DMV morphology and total numbers are not associated with a fitness cost compared to WT virus (WT versus N258T or N479S) (Fig. 8). Previously we observed that mutation of the nsp4 glycosylation sites alters DMV morphology, and the current study extends the result by demonstrating that loss of nsp4 glycosylation is associated with a substantial fitness cost. We cannot conclude that the decreased fitness is due to changes in DMV morphology or numbers because mutations across nsp4 that do not alter virus fitness cause aberrant DMVs and decreased numbers, regardless of location.
The relationship of DMVs to viral replication remains complex. Multiple mutations in nsp4, particularly loop 1, have significant effects on viral replication. Additionally, a small-molecule inhibitor was identified that prevents coronavirus DMV formation, likely by targeting nsp6 (31). The inhibitor profoundly knocked down virus replication, demonstrating that DMVs are linked to virus replication. The mechanisms of DMV formation and how the small molecule inhibits DMV formation remain unknown. Therefore, understanding how viruses induce membrane modifications and form replication complexes will help in the designing of antivirals to target this process. This work emphasizes the role of nsp4 loop 1 in the proper formation of DMVs and identifies nsp4 glycosylation as a putative target for antiviral therapy.


We thank Janice Williams for TEM assistance and image analysis. We thank Megan Culler Freeman and Clint Smith for critical reviews of the manuscript.
This work was supported by the NIH grant RO1 AI50083 (M.R.D. and D.C.B.) from the National Institute of Allergy and Infectious Diseases. D.C.B. was supported by a Training Grant in Immunobiology of Blood and Vascular Systems through the Vanderbilt University School of Medicine (T32HL697659). Experiments were performed in part through the use of the VUMC Cell Imaging Shared Resource (supported by NIH grants CA68485, DK20593, DK58404, DK59637, and EY08126).


Gadlage MJ, Sparks JS, Beachboard DC, Cox RG, Doyle JD, Stobart CC, Denison MR. 2010. Murine hepatitis virus nonstructural protein 4 regulates virus-induced membrane modifications and replication complex function. J Virol 84:280–290.
Miller S, Krijnse-Locker J. 2008. Modification of intracellular membrane structures for virus replication. Nat Rev Microbiol 6:363–374.
Angelini MM, Akhlaghpour M, Neuman BW, Buchmeier MJ. 2013. Severe acute respiratory syndrome coronavirus nonstructural proteins 3, 4, and 6 induce double-membrane vesicles. mBio 4(4):e00524-13.
Knoops K, Kikkert M, van den Worm SHE, Zevenhoven-Dobbe JC, van der Meer Y, Koster AJ, Mommaas AM, Snijder EJ. 2008. SARS-coronavirus replication is supported by a reticulovesicular network of modified endoplasmic reticulum. PLoS Biol 6:e226.
Sparks JS, Lu X, Denison MR. 2007. Genetic analysis of murine hepatitis virus nsp4 in virus replication. J Virol 81:12554–12563.
Ulasli M, Verheije MH, de Haan CAM, Reggiori F. 2010. Qualitative and quantitative ultrastructural analysis of the membrane rearrangements induced by coronavirus. Cell Microbiol 12:844–861.
Snijder EJ, van der Meer Y, Zevenhoven-Dobbe J, Onderwater JJM, van der Meulen J, Koerten HK, Mommaas AM. 2006. Ultrastructure and origin of membrane vesicles associated with the severe acute respiratory syndrome coronavirus replication complex. J Virol 80:5927–5940.
Bost AG, Carnahan RH, Lu XT, Denison MR. 2000. Four proteins processed from the replicase gene polyprotein of mouse hepatitis virus colocalize in the cell periphery and adjacent to sites of virion assembly. J Virol 74:3379–3387.
Brockway SM, Clay CT, Lu XT, Denison MR. 2003. Characterization of the expression, intracellular localization, and replication complex association of the putative mouse hepatitis virus RNA-dependent RNA polymerase. J Virol 77:10515–10527.
van der Meer Y, Snijder EJ, Dobbe JC, Schleich S, Denison MR, Spaan WJ, Locker JK. 1999. Localization of mouse hepatitis virus nonstructural proteins and RNA synthesis indicates a role for late endosomes in viral replication. J Virol 73:7641–7657.
Gosert R, Kanjanahaluethai A, Egger D, Bienz K, Baker SC. 2002. RNA replication of mouse hepatitis virus takes place at double-membrane vesicles. J Virol 76:3697–3708.
Clementz MA, Kanjanahaluethai A, O'Brien TE, Baker SC. 2008. Mutation in murine coronavirus replication protein nsp4 alters assembly of double membrane vesicles. Virology 375:118–129.
Oostra M, Hagemeijer MC, van Gent M, Bekker CPJ, Lintelo te EG, Rottier PJM, de Haan CAM. 2008. Topology and membrane anchoring of the coronavirus replication complex: not all hydrophobic domains of nsp3 and nsp6 are membrane spanning. J Virol 82:12392–12405.
Oostra M, Lintelo te EG, Deijs M, Verheije MH, Rottier PJM, de Haan CAM. 2007. Localization and membrane topology of coronavirus nonstructural protein 4: involvement of the early secretory pathway in replication. J Virol 81:12323–12336.
Oostra M, de Haan CAM, de Groot RJ, Rottier PJM. 2006. Glycosylation of the severe acute respiratory syndrome coronavirus triple-spanning membrane proteins 3a and M. J Virol 80:2326–2336.
Baliji S, Cammer SA, Sobral B, Baker SC. 2009. Detection of nonstructural protein 6 in murine coronavirus-infected cells and analysis of the transmembrane topology by using bioinformatics and molecular approaches. J Virol 83:6957–6962.
Kanjanahaluethai A, Chen Z, Jukneliene D, Baker SC. 2007. Membrane topology of murine coronavirus replicase nonstructural protein 3. Virology 361:391–401.
Hagemeijer MC, Ulasli M, Vonk AM, Reggiori F, Rottier PJM, de Haan CAM. 2011. Mobility and interactions of coronavirus nonstructural protein 4. J Virol 85:4572–4577.
Helenius A, Aebi M. 2004. Roles of N-linked glycans in the endoplasmic reticulum. Annu Rev Biochem 73:1019–1049.
Al-Mulla HMN, Turrell L, Smith NM, Payne L, Baliji S, Zust R, Thiel V, Baker SC, Siddell SG, Neuman BW. 2014. Competitive fitness in coronaviruses is not correlated with size or number of double-membrane vesicles under reduced-temperature growth conditions. mBio 5(2):e01107-13.
Beachboard DC, Lu X, Baker SC, Denison MR. 2013. Murine hepatitis virus nsp4 N258T mutants are not temperature-sensitive. Virology 435:210–213.
Yount B, Denison MR, Weiss SR, Baric RS. 2002. Systematic assembly of a full-length infectious cDNA of mouse hepatitis virus strain A59. J Virol 76:11065–11078.
Denison MR, Yount B, Brockway SM, Graham RL, Sims AC, Lu X, Baric RS. 2004. Cleavage between replicase proteins p28 and p65 of mouse hepatitis virus is not required for virus replication. J Virol 78:5957–5965.
Kim JC, Spence RA, Currier PF, Lu X, Denison MR. 1995. Coronavirus protein processing and RNA synthesis is inhibited by the cysteine proteinase inhibitor E64d. Virology 208:1–8.
Smith EC, Blanc H, Vignuzzi M, Denison MR. 2013. Coronaviruses lacking exoribonuclease activity are susceptible to lethal mutagenesis: evidence for proofreading and potential therapeutics. PLoS Pathog 9:e1003565.
Schneider CA, Rasband WS, Eliceiri KW. 2012. NIH Image to ImageJ: 25 years of image analysis. Nat Methods 9:671–675.
Hresko RC, Kruse M, Strube M, Mueckler M. 1994. Topology of the Glut 1 glucose transporter deduced from glycosylation scanning mutagenesis. J Biol Chem 269:20482–20488.
Chang XB, Hou YX, Jensen TJ, Riordan JR. 1994. Mapping of cystic fibrosis transmembrane conductance regulator membrane topology by glycosylation site insertion. J Biol Chem 269:18572–18575.
Popov M, Tam LY, Li J, Reithmeier RA. 1997. Mapping the ends of transmembrane segments in a polytopic membrane protein. Scanning N-glycosylation mutagenesis of extracytosolic loops in the anion exchanger, band 3. J Biol Chem 272:18325–18332.
Vagin O, Turdikulova S, Sachs G. 2005. Recombinant addition of N-glycosylation sites to the basolateral Na,K-ATPase β1 subunit results in its clustering in caveolae and apical sorting in HGT-1 cells. J Biol Chem 280:43159–43167.
Lundin A, Dijkman R, Bergström T, Kann N, Adamiak B, Hannoun C, Kindler E, Jónsdóttir HR, Muth D, Kint J, Forlenza M, Müller MA, Drosten C, Thiel V, Trybala E. 2014. Targeting membrane-bound viral RNA synthesis reveals potent inhibition of diverse coronaviruses including the Middle East respiratory syndrome virus. PLoS Pathog 10:e1004166.
Xu X, Lou Z, Ma Y, Chen X, Yang Z, Tong X, Zhao Q, Xu Y, Deng H, Bartlam M, Rao Z. 2009. Crystal structure of the C-terminal cytoplasmic domain of non-structural protein 4 from mouse hepatitis virus A59. PLoS One 4:e6217.
Manolaridis I, Wojdyla JA, Panjikar S, Snijder EJ, Gorbalenya AE, Berglind H, Nordlund P, Coutard B, Tucker PA. 2009. Structure of the C-terminal domain of nsp4 from feline coronavirus. Acta Crystallogr D Biol Crystallogr 65:839–846.
Prentice E. 2004. Coronavirus replication complex formation utilizes components of cellular autophagy. J Biol Chem 279:10136–10141.

Information & Contributors


Published In

cover image Journal of Virology
Journal of Virology
Volume 89Number 415 February 2015
Pages: 2080 - 2089
Editor: S. Perlman
PubMed: 25473044


Received: 25 September 2014
Accepted: 24 November 2014
Published online: 26 January 2015


Request permissions for this article.



Dia C. Beachboard
Department of Pathology, Microbiology, and Immunology, Vanderbilt University Medical Center, Nashville, Tennessee, USA
The Elizabeth B. Lamb Center for Pediatric Research, Vanderbilt University Medical Center, Nashville, Tennessee, USA
Jordan M. Anderson-Daniels
Department of Pathology, Microbiology, and Immunology, Vanderbilt University Medical Center, Nashville, Tennessee, USA
Mark R. Denison
Department of Pathology, Microbiology, and Immunology, Vanderbilt University Medical Center, Nashville, Tennessee, USA
Department of Pediatrics, Vanderbilt University Medical Center, Nashville, Tennessee, USA
The Elizabeth B. Lamb Center for Pediatric Research, Vanderbilt University Medical Center, Nashville, Tennessee, USA


S. Perlman


Address correspondence to Mark R. Denison, [email protected].

Metrics & Citations


Note: There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.

Citation counts come from the Crossref Cited by service.


If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.

View Options

Figures and Media






Share the article link

Share with email

Email a colleague

Share on social media

American Society for Microbiology ("ASM") is committed to maintaining your confidence and trust with respect to the information we collect from you on websites owned and operated by ASM ("ASM Web Sites") and other sources. This Privacy Policy sets forth the information we collect about you, how we use this information and the choices you have about how we use such information.
FIND OUT MORE about the privacy policy