Murine Norovirus Replication Induces G₀/G₁ Cell Cycle Arrest in Asynchronously Growing Cells

ABSTRACT

Many viruses replicate most efficiently in specific phases of the cell cycle, establishing or exploiting favorable conditions for viral replication, although little is known about the relationship between caliciviruses and the cell cycle. Microarray and Western blot analysis of murine norovirus 1 (MNV-1)-infected cells showed changes in cyclin transcript and protein levels indicative of a G₁ phase arrest. Cell cycle analysis confirmed that MNV-1 infection caused a prolonging of the G₁ phase and an accumulation of cells in the G₀/G₁ phase. The accumulation in G₀/G₁ phase was caused by a reduction in cell cycle progression through the G₁/S restriction point, with MNV-1-infected cells released from a G₁ arrest showing reduced cell cycle progression compared to mock-infected cells. MNV-1 replication was compared in populations of cells synchronized into specific cell cycle phases and in asynchronously growing cells. Cells actively progressing through the G₁ phase had a 2-fold or higher increase in virus progeny and capsid protein expression over cells in other phases of the cell cycle or in unsynchronized populations. These findings suggest that MNV-1 infection leads to prolonging of the G₁ phase and a reduction in S phase entry in host cells, establishing favorable conditions for viral protein production and viral replication. There is limited information on the interactions between noroviruses and the cell cycle, and this observation of increased replication in the G₁ phase may be representative of other members of the Caliciviridae.

IMPORTANCE Noroviruses have proven recalcitrant to growth in cell culture, limiting our understanding of the interaction between these viruses and the infected cell. In this study, we used the cell-culturable MNV-1 to show that infection of murine macrophages affects the G₁/S cell cycle phase transition, leading to an arrest in cell cycle progression and an accumulation of cells in the G₀/G₁ phase. Furthermore, we show that MNV replication is enhanced in the G₁ phase compared to other stages of the cell cycle. Manipulating the cell cycle or adapting to cell cycle responses of the host cell is a mechanism to enhance virus replication. To the best of our knowledge, this is the first report of a norovirus interacting with the host cell cycle and exploiting the favorable conditions of the G₀/G₁ phase for RNA virus replication.

INTRODUCTION

Many viruses are able to alter the host cell cycle to favor their own replication or have adapted to replicate most effectively in specific cell cycle stages (1–4). The cell cycle is a series of events that describe the growth and division of a cell. In the first gap phase (G₁), there is a high rate of translation as components required for DNA synthesis are created. Cells can enter gap 0 (G₀) during G₁ phase, if mitogenic stimulants are not present where the cell metabolic rate is low. After G₁, cells enter synthesis (S) phase, where the genome of the cell is replicated. Following this is gap 2 (G₂) phase, where protein synthesis and the metabolic rate are again high. Finally, in mitosis (M) phase, the DNA chromatids and cell contents are split into two daughter cells. Transition through and between phases is highly controlled by multiple regulators. Positive regulators include cyclins and cyclin-dependent kinases (CDK). To activate a CDK, binding of a partner cyclin is required to form a complex and activate its kinase activity (reviewed in reference 5). Progression through G₁ phase is controlled by retinoblastoma (Rb) protein and its bound E2F transcription factor. Rb protein is phosphorylated by binding of D-type cyclins to CDK4/6, driving cell cycle progression through early G₁ phase. S phase entry and progression are governed by cyclins A and E, which both associate with CDK2 and further phosphorylate Rb, releasing E2F and driving DNA synthesis (6). The cyclin B family of proteins associate with CDK1 and direct progression into M phase. Negative regulators of the cell cycle include cyclin-dependent kinase inhibitors (CDI) that inhibit cyclin-CDK complexes. CDI can be split into two families, the Ink4 family and the Cip/Kip family, which have a wide range of inhibitory effects.

Viruses can either delay, arrest, or progress the cell cycle through direct interactions of viral proteins or by inducing an innate immune response to viral replication. For example, DNA viruses, including human papillomavirus, adenovirus, and simian virus 40, encode proteins that promote S phase progression to encourage viral genome replication (7–9). Retroviruses such as human immunodeficiency virus type 1 encode proteins which cause a cell cycle arrest in the G₂/M phase (10). Interestingly, RNA viruses that replicate in the cytoplasm can also affect the cell cycle. Murine coronavirus promotes cell accumulation in the G₀/G₁ phase, while hepatitis C virus induces a G₂/M phase arrest in infected cells (1, 11).

Noroviruses are members of the family Caliciviridae and are nonenveloped RNA viruses that cause gastroenteritis in animals and humans. The inability to culture human norovirus in a cell line has limited research and understanding of the viral replication cycle. Recently, an in vitro model for human norovirus was developed in B cells using enteric bacteria as a stimulatory factor for norovirus infection (12).

Using murine norovirus 1 (MNV-1) as a model, replication of noroviruses can be studied in cell culture. Previous studies have demonstrated that MNV-1 can induce apoptosis through modulation of regulatory proteins (13, 14). Cross talk between apoptosis and the cell cycle occurs due to the overlap in regulatory mechanisms. However, no viruses in the family Caliciviridae have been investigated for their ability to affect the cell cycle. Analysis of microarray data from MNV-1-infected RAW264.7 cells showed dysregulation of transcripts involved in cell cycle regulation as well as fluctuations in pathways involved in DNA replication (15, 16). Therefore, it was considered likely that MNV-1 affects the cell cycle in infected cells.

In this study, we show that MNV-1 infection of RAW264.7 and RAW-Blue cells altered expression of key cell cycle regulatory molecules and caused an accumulation of cells in the G₀/G₁ phase of the cell cycle. Furthermore, the conditions created by infection aid MNV-1 replication, as cells progressing through the G₁ phase supported MNV-1 replication over cells in other phases of the cell cycle.

MATERIALS AND METHODS

Cells and viruses. (i) Bioinformatic analysis and quantitative real-time PCR.

RAW264.7 cells (obtained from ATCC) were cultured in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Gaithersburg, MD) containing penicillin (100 U/ml), streptomycin (0.1 mg/ml) (Life Technologies), and 5% heat-inactivated fetal bovine serum (Thermo Fisher Scientific). Cells were passaged every 48 h and were incubated at 37°C in 5% CO₂. Murine norovirus 1 (CW1-P3) (17) was generated through reverse genetics as previously described (18) and propagated in RAW264.7 cells. Cell debris was removed through centrifugation, and the supernatant (unpurified MNV-1) was collected.

(ii) Cell cycle analysis.

RAW-Blue cells (mouse leukemic monocyte macrophage cell line) (InvivoGen, San Diego, CA) were cultured in DMEM (Life Technologies, Gaithersburg, MD) containing penicillin (100 U/ml), streptomycin (0.1 mg/ml), Normocin (100 μg/ml), zeocin (200 μg/ml) (Life Technologies), and 10% heat-inactivated fetal bovine serum (FBS) (Thermo Fisher Scientific). Cells were passaged every 48 h, and cells were incubated at 37°C in 5% CO₂. MNV-1 was propagated in RAW-Blue cells and purified by ultracentrifugation through a 30% (wt/vol) sucrose cushion at 112,700 × g. The pellet was resuspended in Dulbecco's phosphate-buffered saline (dPBS) and filter sterilized through a 0.45-μm filter to create a purified MNV-1 stock.

Infections.

Unless otherwise stated, cells were infected in medium free of antibiotics with MNV-1 at a multiplicity of infection (MOI) of 5 for viral effects on the cell cycle and an MOI of 1 for cell cycle effects on MNV-1 replication. After 1 h of virus adsorption, the medium was removed, and fresh cell medium containing 10% FBS was added. Cells were harvested at various times postinfection (p.i.) for cell cycle analysis or Western blot probing.

Expression analysis by microarray.

Asynchronously growing RAW264.7 cells were seeded in 6-well plates at 8 × 10⁵ cells/well and then mock infected or infected with MNV-1 at an MOI of 1. At 18 h p.i., RNA was extracted using the Qiagen RNeasy minikit as per the manufacturer's instructions. Transcriptomic analysis was conducted on duplicate infected and mock-infected samples in a manner similar to that previously described (16). GenePattern procedures were used as previously described (19). Raw intensities were normalized using robust multiarray average (RMA) and quantile normalization using the ExpressionFileCreator and then preprocessed using PreprocessDataset. Comparison was done using ComparativeMarkerSelection. Means and ratios of means of infected to mock-infected cells were determined (Table 1), and the false discovery rate (FDR) was estimated by ComparativeMarkerSelection using the procedure of Benjamini and Hochberg (20). Ratios were not calculated when the mock-infected values were low (<50). The normalized data and ratios were visualized in PathVisio using WikiPathways (e.g., WP190 cell cycle) (21). The sets with highest and lowest ratios representing genes of increased or decreased expression were analyzed for “Functional annotation clustering” using the DAVID platform (22, 23) and a Mus musculus background. A Benjamini-Hochberg correction was used to correct for multiple testing using Swiss-Prot or Gene Ontology (GO) terms. The clustering process was used to group terms with similar groups of genes, e.g., Swiss-Prot keyword “cell division” and GO biological process “cell cycle.”

TABLE 1.

Transcript changes for cell cycle and nucleotide metabolism regulators

Transcript	Transcript level ina:		Ratiob
	MNV-infected cells	Mock-infected cells
Cyclin A2 (Ccna2)	154 ± 37	1,611 ± 49	0.10*
Cyclin B1 (Ccnb1)	148 ± 20	1,154 ± 24	0.13*
Cyclin B2 (Ccnb2)	106 ± 16	1,225 ± 27	0.09*
Cyclin C (Ccnc)	145 ± 30	145 ± 9	1.00
Cyclin D1 (Ccnd1)	3,211 ± 33	2,099 ± 2	1.53
Cyclin D3 (Ccnd3)	325 ± 49	412 ± 58	0.79
Cyclin E1 (Ccne1)	162 ± 35	280 ± 14	0.58*
Cyclin E2 (Ccne2)	51 ± 18	260 ± 33	0.20*
Thymidine kinase 1 (TK1)	50 ± 10	936 ± 51	0.05*
Ribonucleotide reductase M2 (RRM2)	199 ± 22	2,011 ± 212	0.10*
Deoxyuridine triphosphatase (DUT)	50 ± 27	687 ± 92	0.07*

^aData are means (the “±” values indicate ranges).

^bBenjamini and Hochberg corrected. *, FDR < 0.001 (FDR was estimated by ComparativeMarkerSelection using the procedure of Benjamini and Hochberg [20]).

Synchronization of cells.

Subconfluent cultures of RAW-Blue cells were synchronized to the G₀ phase by serum deprivation. Approximately 1.5 × 10⁶ cells were seeded into 25-cm² flasks and maintained in FBS-free medium for 72 h. For G₁ phase arrest, cells were seeded at approximately 8 × 10⁵ cells/well in 6-well plates or 2.0 × 10⁶ cells in 25-cm² flasks and treated with N-butyrate (B5887; Sigma) at 3 mM for 20 h. For G₂ phase arrest, cells were seeded at 1.5 × 10⁶ in 25-cm² flasks and treated with 100 μM genistein (G6649; Sigma) for 48 h. For M phase arrest, cells were seeded at 8 × 10⁵ cells per well in 6-well plates or 2.5 × 10⁶ cells in 25-cm² flasks and treated with nocodazole (M1404; Sigma) at 50 ng/ml for 10 h. To generate a G₁ phase progressing population (designated G₁>), approximately 3.5 × 10⁶ cells were seeded in 25-cm² flasks then synchronized to M phase with nocodazole treatment; the monolayer was washed 3 times, and complete medium was added. After 3 h of release, the now-G₁ phase cells were plated at 1 × 10⁶ cells per well in a 6-well plate.

Cell cycle analysis by flow cytometry.

The percentage of cells in each phase of the cell cycle was determined through propidium iodide (P4170; Sigma) staining and fluorescence-activated cell sorting (FACS) analysis of the DNA content. Cells were harvested and fixed in 3 ml ice-cold 70% ethanol. After >12 h, the cells were washed in dPBS and resuspended in staining buffer (50 μg/ml propidium iodide and 0.1 mg/ml RNase A [R4875; Sigma] in dPBS) for 45 min at 37°C in 5% CO₂. Stained cells were then washed in dPBS and analyzed using FACS (BD FACSCalibur or BD Fortessa). At least 10,000 cells were counted for each sample, and data were analyzed with MODfit L.T. 3.0 software (Verity Software House).

Plaque assays.

Crystal violet staining was used to quantify viral titers. The cell culture medium was collected postinfection, and plaque assays were performed as previously described (24).

Western blot analysis.

MNV-1-infected and mock-infected cells were collected postinfection and washed twice in dPBS. Cells were lysed directly in 25 μl dPBS and 25 μl sample buffer (120 mM Tris-HCl [pH 6.8], 5% SDS, 10% 2-mercaptoethanol, 20% glycerol, 0.01% bromophenol blue) and stored at −80°C until analyzed. Proteins were transferred to nitrocellulose membranes (Amersham Hybond-C Extra; GE Healthcare) and detected with the corresponding primary and secondary antibodies. Fluorescence of infrared secondary antibodies was detected and quantified with an Odyssey Fc imaging system (LI-COR).

Antibodies.

The following antibodies were used in Western blot analysis. Primary antibodies were cyclin D1 (ab16663; Abcam), cyclin E (E-4) (sc-25303; Santa Cruz), cyclin A (H-432) (sc-751; Santa Cruz), cyclin B1 (AB2949; Abcam), cyclin B2 (AB18250; Abcam), actin (I-19) (sc-1616; Santa Cruz), and anti-MNV-1 capsid (25). Secondary antibodies were 680RD donkey anti-goat IgG (926-68074; LI-COR), 800CW donkey anti-rabbit IgG (926-32213; LI-COR), and 800CW donkey anti-mouse IgG (926-32212; LI-COR).

Statistical and densitometric analysis.

Western blots are shown for one of three independent experiments. Band analysis for each protein was quantified using Image Studio Lite software, and the results are presented as means and standard deviations (SD) of the three experiments. Statistics were analyzed using Student's t test. P values of <0.05 were considered statistically significant. Each protein quantification was first normalized against actin loading, before comparisons for changes (recorded as n-fold) were made.

Microarray data accession number.

The complete raw data set was submitted as cel files to the GEO database under accession number GSE61562.

RESULTS

Bioinformatic analysis of the cell cycle.

To globally investigate cellular changes in gene expression due to MNV-1 infection, asynchronous RAW264.7 cells were infected with virus, and changes were measured after 18 h. The levels of transcripts were compared between infected and mock-infected cells using Affymetrix 430 v2 microarrays. MNV-1 infection had significant effects on many transcripts. Of the 8,500 genes affected, fewer than 10% were ≥2-fold downregulated and ∼20% were upregulated. However, most transcripts were not significantly altered. Notably, transcripts encoding proteins in the cell cycle were reduced (e.g., cyclin B2, 10-fold), and immune response genes were highly increased (e.g., interleukin 18, over 10-fold). Those with 5-fold or greater changes in both directions were analyzed to identify enriched pathways or processes. Functional Annotation Clustering (DAVID) (22, 23) of ≥5-fold-downregulated transcripts generated 11 significant clusters of terms describing classes of genes that included terms with corrected P values of <1 × 10⁻⁶ (Table 1). The transcripts most significantly reduced included the Gene Ontology (GO) Biological Pathways (BP) terms “Cell cycle” (97 genes; corrected P value, 3.1 ×10⁻⁵³) and similar protein terms (Swiss-Prot) or GO terms (data not shown). Additional clusters of reduced expression included terms related to the cellular component “Chromosome” (68 genes; P, 3.3 × 10⁻⁴²) and “Spindle” and to the molecular function “ATP binding.” All 11 clusters related to cell division. This indicates that infection affects the mitotic cell cycle transcripts. Of the most reduced transcripts, several regulators of the cell cycle stood out (Fig. 1, dark green). The reduction of cyclin A, E, and B transcripts suggests that MNV infection impairs cell cycle progression (Table 1). The expression of several genes associated with the biological process “G₁/S phase transition” (GO:0000082) are reduced >2-fold, including but not limited to Cdc25a, Cdca5, Cdk2, Cdk4, Rb1, Skp2, cyclin A2, cyclin E1, and cyclin E2. Fewer genes associated with the “G₂/M transition” term (GO:0000086) were significantly reduced (Fig. 1).

FIG 1.

Changes in the expression of cell cycle-related genes during MNV-1 infection. Asynchronously growing RAW264.7 cells were mock infected or MNV-1 infected for 18 h prior to RNA extraction, and gene expression was analyzed by Affymetrix Mouse 430 2.0 microarrays (as described in Materials and Methods). Only the genes in the “Cell cycle (Mus musculus)” pathway (WP190) are shown; the full set of data is available from NCBI GEO (accession no. GSE61562). Genes with higher mRNA expression are shown in orange to red (2-fold), and those with lower levels are in shades of green (5-fold). Those in darkest green are reduced >5-fold. Visualization used PathVisio (21). When multiple Affymetrix probe sets correspond to a gene, more than one bar is shown. Several of the downregulated cyclin genes (designated with the format Ccnxx [e.g., Ccna2 and Ccnb2]) showed highly significant reductions and were chosen for further study.

MNV-1 infection affects cyclin expression.

An effect on cells at an RNA transcript level is likely to change expression levels of proteins. Based on the microarray data showing changes to cell cycle regulatory protein transcripts, we asked if this affected cyclin levels in MNV-1-infected cells. Asynchronously growing RAW-Blue cells were infected at an MOI of 5 and harvested at various times postinfection for Western blot analysis. After immunofluorescence band analysis, each protein signal was normalized to actin, and the change was calculated (n-fold) for each time point between MNV-1-infected and mock-infected cells. MNV-1 infection caused a 0.59-fold decrease in cyclin A expression at 12 h postinfection. Cyclin B2 expression was decreased 0.56-fold at 9 h postinfection, while cyclin D1, E, and B1 showed no differences between asynchronously growing mock-infected and infected cells (Fig. 2). This correlates with the transcriptomic data (Fig. 1), where cyclin A2 and B2 decreased significantly at an mRNA level. Although there is some variation in protein expression triplicates (indicated by the error bars in Fig. 2), the results show a significant decrease in both cyclin A and B2 expression despite fluctuations in cyclin expression in mock-infected populations. Changes to cyclin levels are supported by microarray results and were further confirmed in later experiments. We postulated that the decrease in A and B cyclins could indicate that cells were accumulating in the G₀/G₁ phase of the cell cycle.

FIG 2.

Effects of MNV infection on host cyclin expression. (A) Asynchronously growing RAW-Blue cells were mock infected (Mock) or infected with MNV at an MOI of 5. At the indicated times p.i., whole-cell lysates were collected and subjected to Western blot analysis for probing with cyclin antibodies. The data are from one of three experiments. Actin was used as a loading control. Cyclin D1 and cyclin E bands migrate as doublets around 36 kDa and 50 kDa, respectively. Cyclin B1 migrates at 60 kDa, cyclin B2 at 45 kDa, and cyclin A at 54 kDa. (B) Cyclin levels were quantified with Image Studio Lite (LI-COR) and normalized against actin, and results are presented as means and SD from three experiments. Statistical significance was compared to the corresponding value at 0 h p.i. for each cyclin, *, P < 0.05; **, P < 0.01.

MNV-1 infection causes cells to accumulate in G₀/G₁ phase of the cell cycle.

Based on the observation that MNV-1 infection dysregulates regulatory cell cycle RNA transcripts, as well as expression of host cell cyclins, we asked if MNV-1 affects the cell cycle during an infection. Asynchronous RAW-Blue cells were mock infected or MNV-1 infected at an MOI of 5, and at selected times postinfection, cells were collected and analyzed by flow cytometry to compare cell cycle profiles. At 9 h postinfection, there was a 24% increase in the G₀/G₁ phase of the cell cycle and a 26% reduction in the proportion of cells in S phase in MNV-1-infected cells compared to mock-infected cells (Fig. 3). These results are indicative of an arrest at the G₁/S restriction point.

FIG 3.

MNV infection induces the accumulation of cells in G₀/G₁ phase of the cell cycle. (A) Asynchronously growing RAW-Blue cells were mock infected (Mock) or infected with MNV at an MOI of 5. Cells were collected postinfection (p.i.) at the times indicated for FACS analysis of the cell cycle. The data presented are representative of one of three experiments. (B) The histograms from A were analyzed with MODfit LT 3.0, and the percentage of cells in each phase of the cell cycle is shown. The results are means and SD from three experiments. Statistical significance was determined for comparisons between mock-infected and MNV-1-infected cells for each time point. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Progression of MNV-1-infected cells from G₁/S phase is reduced.

To better understand the G₀/G₁ arrest, synchronized cells were monitored for cell cycle progression through the G₁/S restriction point. With all synchronization treatments, dosage and treatment times were first optimized to minimize effects on cell viability and viral infection. Cells were synchronized to G₁ with N-butyrate treatment and then stimulated to reenter division through replacement of N-butyrate medium. Cells were incubated for 3 h in N-butyrate-free medium prior to infection at an MOI of 5, and at selected times postrelease, cells were collected and analyzed by flow cytometry. MNV-1 infection caused a reduction in S phase entry, compared to mock-infected cells (Fig. 4). In mock-infected cells, the G₀/G₁ population decreases and the S phase population increases after serum addition, as cells reenter cell replication. In contrast, MNV-1-infected cells remained mostly in G₀/G₁ phase after serum addition, indicating that the viral infection induces a G₀/G₁ phase accumulation through either a decrease in progression through G₁/S or a prolonging of G₀ or G₁ phase.

FIG 4.

MNV infection reduces cell cycle progression from G₁ to S phase. (A) N-Butyrate-treated RAW-Blue cells arrested in G₁ were released from arrest by 3 washes and the addition of complete medium containing 10% FBS to the cells. The cells were mock infected (Mock) or infected with MNV at an MOI of 5 after 3 h postrelease. After 1 h of virus absorption, complete medium was added and cell cycle profiles were taken at the indicated times p.i. for FACS analysis. The data are from one of three experiments. (B) The histograms were analyzed using MODfit LT 3.0, and the percentage of cells in each phase of the cell cycle is shown. The results are means and SD from three experiments. Statistical significance was determined for comparisons between mock-infected and MNV-1-infected cells for each time point. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

MNV-1 infection does not influence late-phase cell progression from M phase into G₁ phase.

Another possible explanation for the G₀/G₁ phase accumulation seen during MNV-1 infection is faster progression through the later phases of the cell cycle through to G₀/G₁. To test this, cells were arrested in early M phase with nocodazole treatment. Synchronized cells were then infected, and fresh medium was added to induce cell cycle progression. Cell cycle profiles were collected postrelease and analyzed with flow cytometry. In both mock-infected and MNV-1-infected cell populations, the G₂/M population progressed through the cell cycle into the G₀/G₁ phase after 2 h postrelease. This is shown by a 40% drop in the G₂/M population and a simultaneous 40% increase in the G₀/G₁ population in both mock-infected and MNV-1-infected populations (Fig. 5). Results show that MNV-1 infection had no influence on progression of the cell cycle through the M phase to the G₀/G₁ phase.

FIG 5.

MNV infection has no effect on cell cycle progression from M phase through to G₀/G₁ phase. (A) RAW-Blue cells were synchronized to M phase of the cell cycle by nocodazole treatment. Synchronized cells were mock infected (Mock) or MNV infected at an MOI of 5. After 1 h of viral absorption, the nocodazole-containing medium was added back to the cells, and the virus was given 3 h of replication time. At 3 h p.i., the cell monolayer was washed 3 times with complete medium, and cell cycle profiles were taken at the indicated times postrelease for FACS analysis. The data are from one of three experiments. (B) The histograms were analyzed by MODfit LT 3.0, and the percentage of cells in each phase of the cell cycle is shown. The results are means and SD from three experiments.

MNV-1 infection inhibits cyclin A expression.

Changes to cyclin expression in RNA transcripts and protein levels as well as cell cycle profiles of infected cell populations suggest that MNV-1 impacts the G₁/S restriction point. Other RNA viruses, such as influenza virus and coronaviruses, are also known to target the G₁/S transition as a means of manipulating the cell into a favorable phase of the cell cycle for viral replication (2, 26). Based on these observations, we examined G₁/S cyclin expression of infected cells passing through the G₁/S restriction point. Cells were synchronized to the G₀ phase, infected, and simultaneously given serum to stimulate cell cycle progression. Cells reentered the cell cycle after 12 h of serum stimulation (Fig. 6A). Cyclin D1 levels decreased slightly in infected populations compared to mock-infected cells. At 15 h postinfection, there was a significant decrease in cyclin D1 levels. Cyclin E protein levels did not change after serum stimulation or postinfection. Cyclin A is not expressed in the quiescent G₀ population; however, its level gradually increased in mock-infected cells. After 12 h, mock-infected cells expressed significant levels of cyclin A while MNV-1 prevented the accumulation of cyclin A in infected cells. At 15 h postinfection, there was a 0.28-fold decrease in cyclin A expression between mock-infected and MNV-1-infected cells (Fig. 6C). These results combined with the lack of effect upon progression from G₂/M confirm that MNV-1 infection inhibits progression through the G₁/S restriction point, possibly through inhibition of cyclin A expression.

FIG 6.

Changes to G₁ cyclin expression in MNV-infected cells released from quiescence. Serum-starved (G₀) RAW-Blue cells were either mock infected (Mock) or MNV infected at an MOI of 5. After 1 h of virus absorption the cells were washed 3 times and complete medium was added. (A) At the indicated times p.i., mock-infected cells were harvested for FACS analysis of the cell cycle. MODfit LT 3.0 was used to calculate the percentage of cells in each cell cycle phase. (B) Mock-infected and MNV-infected cells were collected p.i. for Western blot analysis of cyclins D1, E, and A. The data shown are from one of three experiments. (C) Cyclin expression results from panel B were quantified by Image Studio Lite (LI-COR) and normalized against actin loading control. Data are means and SD from three experiments. Statistical significance was determined for comparison to the corresponding value at 0 h p.i. for each cyclin. *, P < 0.05.

MNV-1 replication is favored in cells progressing through G₁.

MNV-1 infection caused a prolonging of the G₀/G₁ phase in the host cell; however, it was not known if this is beneficial to viral replication. We compared MNV-1 replication in an unsynchronized population of cells to that in populations synchronized into G₀/G₁ and G₂/M phases by viral capsid protein expression and virus progeny titers. Cells were synchronized into G₀ with serum withdrawal, into G₁ with N-butyrate treatment, into G₂ with genistein treatment, and into M with nocodazole treatment, and a population of G₁ phase progressing cells (G₁>) was generated by removal of nocodazole from an M phase-synchronized population to produce a population of cells synchronized into early G₁ phase prior to infection. A G₁ progressing population was included because MNV-1 may not cause a cell cycle arrest in G₁ but rather induce a prolonged G₁ phase, and hence this population of cells could mimic the cell state during an infection. The synchronized and unsynchronized populations were seeded at equal cell densities and infected with MNV-1 at an MOI of 1. At the time of infection, cells were harvested, and cell cycle analysis was performed to show cell synchronization prior to infection. All cell populations were synchronized to their desired phases preinfection (Fig. 7A). At 9 h postinfection, cells had remained synchronized in their target cell phase during the course of infection (Fig. 7B). MNV-1 virus levels were highest in cell populations progressing through G₁>, with an approximately 2-fold increase in virus titer compared to that in an unsynchronized population. Synchronized populations in G₁, G₂, and M phases had no difference in MNV-1 replication compared to an unsynchronized population, while G₀ phase-arrested cells had the lowest levels of MNV-1 replication (Fig. 7C). Capsid expression was measured through Western blot analysis in each of the synchronized cell populations. Expression profiles matched that of viral titers, as MNV-1 had the highest capsid expression in G₁> cells at 2-fold that of expression levels in unsynchronized cells, while capsid expression was undetectable in the G₀ population, and G₁, G₂, and M phase populations had levels of capsid protein expression similar to that in an unsynchronized population (Fig. 7D).

FIG 7.

G₁ phase progressing cells promote the highest viral replication. RAW-Blue cells were synchronized in flasks treated by serum starvation (G₀) for 72 h, treated with 3 mM N-butyrate (G₁) for 20 h, treated with 50 ng/ml nocodazole for 10 h and then released for 2 h (G₁>), treated with 100 μM genistein (G₂) for 48 h, or treated with 50 ng/ml nocodazole (M) for 10 h. The synchronized and unsynchronized cells were then scraped and centrifuged, and the supernatant was discarded. Unsynchronized and G₁-, G₁>-, G₂-, and M-synchronized cells were resuspended in complete medium and G₀ cells in medium with no FBS. Cells were then seeded at 1 × 10⁶ cells per well into 6-well plates and infected with MNV at an MOI of 1 for 1 h. After absorption, fresh synchronizing agent was added to synchronized cells at the same concentration that was used originally. At 9 h postinfection, cells were harvested and analyzed. (A and B) Cell cycle profiles pre- and postinfection were determined by FACS analysis. The histograms generated were analyzed with MODfit LT 3.0, and the percentage of cells in each phase of the cell cycle is shown. (C) MNV capsid expression was determined by Western blot analysis (top); actin was included as a loading control. Capsid levels from three experiments were quantified with Image Studio Lite (LI-COR) and normalized against the results for actin. All results are presented as a percentage of the MNV-1-infected unsynchronized control population (bottom). Statistical significance was determined for comparison to the value for the MNV-1-infected unsynchronized population. *, P < 0.05; **, P < 0.01; ****, P < 0.0001. (D) MNV progeny in the supernatant were titrated by plaque assay, and quantitative analysis data, in PFU, are shown as a percentage of the value for the unsynchronized MNV-1-infected control. The results are means and SD from three experiments. Statistical significance was determined for comparison to the value for the MNV-1-infected unsynchronized control population. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

DISCUSSION

This study explored the relationship between the cell cycle and murine norovirus replication. The results show that MNV-1 affects key cell cycle regulatory proteins at the level of expression in RAW264.7 cells, as determined by RNA transcript analysis. Changes to cell cycle regulators were then analyzed in another macrophage cell line (RAW-Blue cells) to confirm the effect of MNV-1 on the cell cycle in a related but distinct cell lineage. For both cell lines, similar MNV-1 growth kinetics were seen, with progeny virus detected after 6 h postinfection (data not shown). The influence of MNV-1 infection upon transcripts was consequently shown to affect the level of cyclin protein in infected RAW-Blue cells. FACS analysis of MNV-1-infected RAW-Blue cells confirmed an increase in the proportion of cells in the G₀/G₁ phase of the cell cycle consistent with the observed change in cyclins. Cell cycle progression through the G₁/S boundary was reduced, with the proportion of cells in the S phase declining due to the G₀/G₁ population not proceeding into S phase. MNV-1 infection had no effect on progression of cells through later phases of the cell cycle, as infected cells moving through G₂/M phases showed progression identical to that of the mock-infected population. Collectively, these results show that MNV-1 replication in RAW-Blue cells causes an accumulation of cells in the G₀/G₁ phase due to a prolonging of the G₁ phase or an arrest of cell cycle progression through the G₁/S restriction point.

The effect of MNV-1 infection on cyclin expression matches what is expected in a population of cells progressing through the G₁ phase. The B family of cyclins are highly expressed during cell cycle progression during late S phase until late mitosis (27, 28). A decrease in cyclin B2, as seen during MNV-1 infection from transcriptome data and protein quantification, would suggest a decrease in cell cycle progression through G₂/M phase. This is consistent with the effect of MNV-1 on the cell cycle, as cyclin B2 expression would be low in a G₁ phase-arrested population due to decreased S phase and G₂/M phase cells. The D family cyclin levels are highly expressed during the G₂ phase through to the end of the G₁ phase (29). MNV-1 infection had no significant effect on cyclin D1 expression during infection of an asynchronous population. In cells released from a G₀ arrest, a slight reduction in cyclin D1 was observed in MNV-1-infected cells at 15 h postinfection. Cyclin E is involved in the transition between the G₁ and S phases of the cell cycle. Although MNV-1 infection caused changes to cyclin E transcript levels (Table 1), MNV-1 had no observed effect on cyclin E protein levels during infection of asynchronous cells or on cells released from quiescence. Changes in mRNA transcripts do not always correlate to changes in protein levels, as there are multiple layers regulating protein translation. Further, the signal strength for cyclin E1 and E2 transcripts was low compared to that of other cyclins (Table 1), leading us to conclude that MNV-1 does not decrease cyclin E levels. Lastly, A cyclins are involved in several phases of the cell cycle, but predominantly S phase, where they are involved in entry and progression through S phase and into G₂ (30). MNV-1 infection caused a reduction in cyclin A expression in asynchronous culture and prevented expression in cells released from quiescence. The downregulation of cyclin A strongly supports the G₁ phase accumulation seen during infection. We propose that the inhibition of cyclin A expression is one probable cause of the observed MNV-1 induction of the G₁/S phase arrest. The role of cyclin D1 in MNV-1-induced cell cycle arrest is clouded by contradictory microarray results and only partial inhibition of protein expression very late in infection. We propose that the arrest induced by MNV-1 is late in the G₁ phase, where cyclin D1 would be expressed and where viral replication is at its highest. A range of factors regulates cell replication in addition to cyclins, including transcriptional and posttranslational modification, protein localization, CDK, and CDI. The mechanism by which MNV-1 induced a cyclin A protein decrease and consequent G₁/S phase cell cycle arrest is not known.

It was hypothesized that an arrest in the G₀/G₁ phase by MNV-1 infection could create a more favorable environment for MNV-1 replication. Both MNV-1 capsid protein production and MNV-1 progeny production were highest in cells progressing through the G₁ phase. This correlates with the effect that MNV-1 has on the cell cycle, as we propose that MNV-1 infection prolongs the G₁ phase of the cell cycle by inhibiting G₁/S transition, providing a more beneficial environment for MNV-1 replication. Numerous other RNA viruses cause arrests in the G₀/G₁ phase of the cell cycle. Influenza A virus causes an arrest in G₀/G₁ phase during its infection of A549 and A/WSN/33 cells through interaction of its NS1 protein with the RhoA/Rb pathway (1, 26). This arrest was also shown to benefit influenza A virus replication, as viral titers and viral protein production increase in G₀ phase-arrested cells. Murine coronavirus induces a G₀/G₁ phase arrest in 17CL-1 and DBT cells through inhibition of Rb phosphorylation. This is possibly caused by virus-induced degradation of host cyclins D2 and cyclin E, as well as a downregulation of CDK2 activity. MNV-1 similarly influences the host cell cycle, causing a G₀/G₁ phase arrest that benefits its own replication. However, it cannot be excluded that the effects on the cell cycle are host driven as a consequence of MNV-1 infection and that the virus has adapted to better replicate under these conditions.

The biological significance of MNV-1-induced cell cycle arrest or prolonging of the G₁ phase is supported by more efficient MNV-1 capsid production and progeny replication. The benefit to the virus may be explained by several factors. Ribonucleotides are precursors for deoxyribonucleotides whose levels change throughout cell division. Levels of ribonucleotides drop as cells enter S phase due to the increased demand for deoxyribonucleotides (31). A cell cycle arrest in the G₁ phase would provide increased amounts of ribonucleotide pools for MNV-1 genome synthesis. Expression of enzymes involved in the processing of ribonucleotides to deoxyribonucleotides is inhibited during MNV-1 infection. Thymidine kinase 1 (TK1), ribonucleotide reductase M2 (RRM2), and deoxyuridine triphosphatase (DUT) expression was decreased at a transcript level as shown by microarray results (Table 1). Nucleotide expression was further confirmed by quantitative reverse transcription-PCR (qRT-PCR) showing decreases in TK1, RRM2, and DUT (data not shown). An induced arrest at the G₁/S restriction point by MNV-1 would prevent a decrease in ribonucleotide levels, favoring the replication of MNV-1 RNA.

Different phases of the cell cycle have varying metabolic rates, with the G₁ phase having the highest translation efficiency (32, 33). The increased translation efficiency in the G₁ phase is exploited for recombinant protein expression by industries and may also be beneficial for MNV (34, 35). Translation efficiency of hepatitis C virus is shown to be greatest during the G₀/G₁ phase of the cell cycle (36). The G₁/S phase arrest induced by MNV may increase translation efficiency rates of MNV proteins and aid in MNV replication.

The recruitment of membranes for norovirus replication has been shown to be vital to the synthesis of viral proteins (37). The MNV-1 replication complexes associate with the endoplasmic reticulum and Golgi apparatus (38). During mitosis, the endoplasmic reticulum and the Golgi apparatus disassemble, and so the G₁/S arrest seen in MNV-1-infected cells may prevent impairment of membrane structures and reduced MNV-1 replication (39, 40).

An induced cell arrest has the potential to protect infected cells from the host immune system. Noncycling cells are also less likely to be killed by cytotoxic T cells, making a more persistent infection and longer shedding period (41). Several studies have described links between the cell cycle and apoptosis signaling. A delay in apoptosis is often seen following a cell cycle arrest (42, 43). Furthermore, induction of apoptosis often requires cell cycle progression (44). MNV-1 infection results in a cell cycle arrest for up to 12 h postinfection (data not shown). After 12 h postinfection, apoptosis is stimulated in RAW264.7 cells, illustrated by a downregulation of survivin (15). Thus, MNV-1-induced cell cycle arrest may prevent early death from apoptosis, allowing time for MNV replication.

The mechanism of the MNV-1 effect on the cell cycle is unknown. Microarray data for MNV-1 infection provide insights into the cell cycle response to MNV-1 infection and the global effects on cell cycle regulators. It is possible that the arrest is caused by MNV-1 proteins, but it is also plausible that this is a host response to viral infection and that MNV-1 has adapted to better replicate under these conditions. MNV-1 infection in RAW-Blue cells causes an accumulation of cells in the G₀/G₁ phase due to a decrease in G₁/S phase progression. Replication was furthermore favored in G₁, a phase that results from MNV-1 replication and is more likely to better support RNA virus replication. This interaction with the cell cycle phase identified for MNV-1 may be applicable to other caliciviruses.