Published ahead of print on April 5, 2007, doi:10.1165/rcmb.2005-0345RC
© 2007 American Thoracic Society DOI: 10.1165/rcmb.2005-0345RC
G1 Phase Cell Cycle Arrest Induced by SARS-CoV 3a Protein via the Cyclin D3/pRb PathwayDepartment of Pathophysiology, Beijing Institute of Radiation Medicine, and State Key Laboratory of Pathogen and Biosecurity, Department of Pathogenic Molecular Biology, Beijing Institute of Microbiology and Epidemiology, Beijing, China Correspondence and requests for reprints should be addressed to Yuwen Cong, Ph.D., Department of Pathophysiology, Beijing Institute of Radiation Medicine, No. 27 Taiping Road, Beijing, 100850, China. E-mail: congyw{at}nic.bmi.ac.cn
SARS-CoV 3a is a structural protein, mainly localizing to Golgi apparatus and co-localizing with SARS-CoV M in co-transfected cells. Here we observed that transient expression of 3a inhibited cell growth and prevented 5-bromodeoxyuridine incorporation, suggesting that 3a deregulated cell cycle progression. Cell cycle analysis demonstrated that 3a expression was associated with blockage of cell cycle progression at G1 phase in HEK 293, COS-7, and Vero cells 24–60 h after transfection. Mutation analysis of 3a revealed that C-terminal region (176 aa  274 aa), including a potential calcium ATPase motif, was essential for induction of cell cycle arrest. Topological analysis showed that 3a predominantly located in Golgi apparatus, with its N-terminus residing in the lumen (Nlum) and C-terminus in the cytosol (Ccyt). Analyzing the cellular proteins involving in regulation of cell cycle progression, we demonstrated that 3a expression was correlated with a significant reduction of cyclin D3 level and phosphorylation of retinoblastoma (Rb) protein at Ser-795 and Ser-809/811, not with the expression of cyclin D1, D2, cdk4, and cdk6 in 293 cells. Increases in p53 phosphorylation on Ser-15 were observed in both SARS-CoV M and 3a transfected cells, suggesting that it might not correlate with the 3a-induced G0/G1 phase arrest. The reduction of cyclin D3 level and phosphorylation of Rb were further confirmed in SARS-CoV infected Vero cells. These results indicate that SARS-CoV 3a protein, through limiting the expression of cyclin D3, may inhibit Rb phosphorylation, which in turn leads to a block in the G1 phase of the cell cycle and an inhibition of cell proliferation.
Key Words: SARS-CoV 3a • growth inhibition • G1 phase cell cycle arrest • cyclin D3 • pRb
Severe acute respiratory syndrome (SARS), caused by SARS coronavirus (SARS-CoV), is a life-threatening emerging infectious disease originating from Guangdong Province, China (1, 2). SARS-CoV, a distant member of Group 2 coronaviruses, has recently been identified as the etiologic agent of SARS (3–5). Analysis of the nucleotide sequence of the SARS-CoV genome showed that it is nearly 30 kb in length and organized with the gene order that is characteristic for coronavirus [5'-replicase (rep), Spike (S), Envelope (E), Membrane (M), Nucleocapsid (N)-3']. The genome of SARS-CoV also contains nine genes specifying so-called "accessory proteins" located between S and E, and M and N genes. All these ORFs are predicted to be expressed from eight subgenomic mRNAs identified in SARS-CoV–infected Vero cells, and remarkably, up to four of the eight SARS-CoV subgenomic mRNAs may be functionally bicistronic (6). The exact roles of these SARS-CoV–specific accessory proteins are poorly understood. Previous studies demonstrated that accessory protein genes of other coronaviruses, which vary in size, sequence, and position in the genome, are dispensable for virus replication, at least in cell culture. Still, they may be important for virus–host interaction in the in vivo situation. For example, mutants or deletion of one of these genes, such as the 7b gene of feline coronavirus and gene 3 of swine enteric and respiratory coronavirus, have been reported to relate to reduced virulence and pathogenesis (7, 8), indicating a possible in vivo function. The product of the SARS-CoV 3a gene (CDS: 25252–26074), also referred to as ORF3, X1, and U274 in other articles, was identified separately in the SARS-CoV–infected cells, lung specimen from a patient with SARS, and crude virions (9–11). Previously, it was reported that 3a protein is located in the Golgi apparatus, and the second or third trans-membrane regions are responsible for the Golgi localization (12). The integral membrane protein 3a interacts with other structural proteins, such as S, M, and E proteins, as well as nonstructural protein U122, in infected or transfected Vero E6 cells (9). Recently published papers showed that the 3a gene product is a structural protein of SARS-CoV (13, 14). Overall, these findings suggest that the 3a protein is an important protein in the viral life cycle. In this study, we first present evidence that overexpression of the 3a gene can inhibit cell growth and block cell cycle progression at the G1 phase. The domain responsible for these functions was further identified through construction of a series of truncated mutants of the 3a gene. The mechanism behind the G1 phase arrest principally involved a decrease in expression of cyclin D3 and phosphorylated retinoblastoma (Rb) protein. These results suggested that the 3a protein plays important roles in the SARS-CoV life cycle and virus-induced pathogenesis.
Cell Culture and Transfection Human embryonic kidney cell lines, HEK 293, and African green monkey kidney cell lines, Vero E6 and COS-7, were cultured in Dulbecco's modified Eagle's medium (DMEM) (Gibco BRL, Grand Island, NY) supplemented with 10% FBS. Cultures were incubated at 37°C in a humidified environment with 5% CO2. When cell density in a culture plate reached 70% confluence, the cells were transfected with different plasmid DNAs using Lipofectamine 2000 (Invitrogen, Carlsbad, CA), following the protocol provided by the manufacturer. Briefly, the total amount of DNA transfected into the cells in each well was adjusted to 2 µg/ml by using empty pCMV-myc vector. Cells were incubated with transfection mixtures for 5 h and then replaced with fresh medium.
Construction of Expressing Vectors of SARS-CoV 3a and its Mutants
Growth Curves and 3-(4,5-Dimethylthiazolyl-2)-2,5-Diphenyl Tetrazolium Bromide Assay HEK 293 cells seeded in a 24-well plate (Costar, Cambridge, MA) were transfected with 3a/pCMV-myc and pCMV-myc in triplicate. At 12-h intervals after transfection, cells were rinsed with PBS and a single cell suspension was prepared by trypsinization. Viable cells, which are resistant to trypan blue staining, were counted with a hemacytometer chamber. To avoid bias, counting was done blindly by two individuals for each sample. Cells transfected with at least three independent clones were analyzed. For the 3-(4,5-dimethylthiazolyl-2)-2,5-diphenyl tetrazolium bromide (MTT) assay, HEK 293 cells seeded in a 96-well plate (Costar) were transfected with different concentrations of 3a/pCMV-myc and pCMV-myc plasmids in triplicate. After 48 h, each well was supplemented with 20 µl of MTT solution (5 mg/ml), and incubated for 3 h. The medium was removed and 200 µl of DMSO were added to each well. Then the plate was vibrated to dissolve crystals and the absorbance (O.D.) at 570 nm was measured. The experiments were independently repeated three times.
Confocal Microscopy Analysis For 5-bromodeoxyuridine (BrdUrd) incorporation, Vero E6 and COS-7 cells were transfected with 3a/pCMV-myc and M/pCMV-myc. At 24 h after transfection, cells on glass cover slips were incubated with 10 µmol/liter BrdUrd for 4 h at 37°C and fixed with 100% methanol at 4°C for 10 min. Incorporated BrdUrd was exposed by treatment with 2 M hydrochloric acid at 37°C for 2 h, followed by neutralization in 0.1 M borate buffer (pH 8.5). After washing in PBS, cells were permeabilized in 0.1% Triton X-100/PBS for 5 min and incubated with anti-BrdUrd (1:100; Sigma, St. Louis, MO) and anti-myc (1:100; Cell Signaling, Beverly, MA) antibodies for 1 h. Images were viewed and collected with a confocal fluorescence microscope connected to a Bio-Rad Radiance 2100 laser scanner (Bio-Rad, Richmond, CA).
Flow Cytometric Cell Cycle Analysis
Western Blot Analysis
When it was necessary to reprobe the membrane with another antibody, the membrane was stripped with stripping buffer (2% SDS, 100 mM
Infection of Vero E6 Cells with SARS-CoV
RT-PCR Assay
Statistical and Densitometric Analysis
Cell Growth Is Inhibited in Transfected Cells Expressing the SARS-CoV 3a Protein In this study, the 3a gene of SARS-CoV (ZJ01, AY297028), was cloned into the pCMV-myc vector and expressed in HEK 293 cells as described before (12). We have observed that HEK 293 cells transfected with 3a/pCMV-myc grow slower than pCMV-myc–transfected cells. We thus speculated that expression of 3a gene may inhibit cell proliferation. Compared with exponentially growing control cells transfected with pCMV-myc, cells expressing the 3a gene showed a significant decrease in cell number after 24 h (Figure 1A). In the 3a/pCMV-myc–transfected cells, less than 4.8% of the cells were dead, which indicated that the reason for the decrease in cell number was not the cell death, but may be the cell growth inhibition induced by 3a gene expression. The above experiments were repeated with MTT assay, a more sensitive colorimetric test to monitor the cell proliferation. As shown in Figure 1B, the growth inhibition of 3a/pCMV-myc–transfected HEK 293 cells was significantly dependent on the dose of plasmid used for transfection, whereas the growth of 3a/pCMV-myc–transfected HEK 293 cells was marginally inhibited at the highest dose of plasmid (1.0–2.0µg/ml). Based on its apparent ability to inhibit cell growth, 1.0 µg/ml 3a/pCMV-myc plasmid was used for transfection in our subsequent experiments. To address the mechanism of 3a on cell growth inhibition, cell DNA synthesis was further measured by BrdUrd incorporation (15). In view of the propensity of HEK 293 cells easily to form cell's dumpling after treatment with 2 M hydrochloric acid, COS-7 cells were used in the assay. As shown in Figure 1C, 75% of myc-3a–negative COS-7 cells had incorporated BrdUrd, while almost all myc-3a–expressing cells had little or no BrdUrd incorporation (only 2.85% positive cells). Similar results were obtained in myc-3a–transfected Vero E6 cells (data not shown). As a control, both myc-M (the membrane protein of SARS-CoV)–positive and –negative cells had similar rates of BrdUrd incorporation (63% versus 68%). These data indicate that 3a expression inhibits cells growth and prevented cell cycle entry into S phase.
Induction of G1 Phase Cell Cycle Arrest by the SARS-CoV 3a Protein Cell cycle deregulation is a common response of host cells to many virus infections, and some viral proteins have been shown to be efficient in inducing cell cycle arrest, such as orf-a of feline immunodeficiency virus and nonstructural protein p28 of murine hepatitis coronavirus (MHV) (18, 19). Flow cytometry is a rapid, quantitative, multiparameter cell analysis based on the measurement of visible and fluorescent light emission. Using the myc-tag to identify the transfected (myc-tag–positive cells) and nontransfected cells (myc-tag–negative cells) in 3a/pCMV-myc–transfected HEK 293 cells (15), the cell cycle of the two populations was analyzed, and the transfection efficiency of 3a/pCMV-myc was revealed to be 20.2% by flow cytometry. As shown in Figure 2A, 67% of the cells positive for 3a gene expression were in the G1 phase, whereas only 50% of the 3a-negative cells were in G1 phase at 24 h after transfection. To further confirm the effect of 3a gene expression, HEK 293 cells were transfected with M/pCMV-myc and GFP/pCMV-myc. The transfection efficiencies of myc-GFP and myc-M expression plasmids were found to be 35.7% and 34.0%, respectively. As shown in Figure 2A, the myc-M– and myc-GFP–positive and –negative populations had similar cell cycle progression profiles, indicating that the expression of myc-M or myc-GFP had little effect on the cell cycle. As controls, HEK 293 control cells and pCMV-myc–transfected HEK 293 cells had similar cell cycle profiles, with 50% of the cells in the G1 phase (data not shown). To test whether the G1 phase arrest induced by 3a gene expression was cell line specific or not, pCMV-myc and 3a/pCMV-myc were separately transfected into COS-7 and Vero E6 cells. The transfection efficiencies in the two cell types were 18%. As in the case of 293 cells, transfection with pCMV-myc had little effect on cell cycle profiles of COS-7 and Vero E6 cells (data not shown). In both cell lines, G1 phase arrest induced by 3a gene expression was observed, which was as obvious as that in 293 cells, suggesting that a common mechanism in the different cell lines is involved in 3a protein–induced G1 phase arrest (Figure 2B).
The G1 phase arrest is a crucial DNA damage checkpoint, which acts as an important safeguard for genomic stability. Cells in G1 phase arrest may go into apoptosis, or recover from the G1 phase to enter into the S phase (15, 20). To observe the denouement of the G1 phase arrest induced by the 3a protein, cell cycle analysis was performed in 3a/pCMV-myc–transfected HEK 293 cells from 24–60 h after transfection. A significant increase in the percentage of cells in the G1 phase, and concomitant reduction in percentage of cells in S phase, was noticeable in 3a protein–expressing cells between 24 and 60 h after transfection relative to control cells, with a maximum induction of G1 phase arrest being observed at 24 h after transfection (Figure 2C). After 24 h, although there was some increases in the number of cells in S phase, a G1 phase arrest was still noticeable in 3a protein–expressing cells, indicating that this arrest was not due to a cell density effect. Western blotting assay showed that the expression levels of myc-3a protein in the transfected cells were similar between 24 and 60 h after transfection (Figure 2D). The reasons for the increase in the number of cells in S phase at the later times might be that some cells expressing lower level of myc-3a could slowly enter into S phase. Sub-G1 phases, representing one type of cell apoptosis, were observed from 24 to 60 h after transfection, but were not over 6% of analyzed cells. These data support the idea that the 3a protein is not an inductor of apoptosis, but of cell cycle arrestor.
Cellular Localization and Induction of G1 Phase Arrest of Truncated 3a Protein Mutants
Furthermore, a cell cycle analysis was performed as before in HEK 293 cells. At 24 h after transfection with 3a-mutants/pCMV-myc and pCMV-myc plasmids, cells were collected and analyzed by flow cytometry. The transfection efficiencies of myc-3a, D77–274, D99–274, D176–274, and D1–222 in the samples analyzed were 18.4, 29.2, 22.1, 18.6, and 41.2%, respectively. As shown in Figure 3C, when compared with transfection-negative cells, mutants D77–274, D99–274, D176–274, and 3a all had a similar ability to induce G1 phase cell cycle arrest, but D1–222 had not. It was concluded that the C-terminal domain of 3a may be responsible for its cell cycle arrest induction. Recent articles reported that the C-terminal domain of the 3a protein contains a Yxx  motif (160 aa–173 aa), which was important for the internalization of the 3a protein from the plasma membrane (9), and a potential calcium ATPase motif (200 aa–274 aa) (9, 10). The C-terminally truncated mutant (D1–222), containing the Yxx motif (160 aa–173 aa), could not induce G1 phase arrest, which further suggested that the potential calcium ATPase motif is important for the 3a protein to induce cell cycle arrest.
Topology of the SARS-CoV 3a Protein in the Golgi Apparatus
G1 Phase Arrest Induced by the SARS-CoV 3a Protein via Cyclin D3/pRb Pathway Rb phosphorylation is a critical step in the G1 to S phase transition. Phosphorylation of Rb is regulated primarily by complexes of cyclin D associated with CDK4/6 and later by cyclin E associated with CDK2. To understand the mechanism of 3a-induced G0/G1 phase arrest, we first examined the phosphorylation status of Rb in transfected cells by Western blotting analysis. The total HEK 293 cells transfected with pCMV-myc, 3a/pCMV-myc, and M/pCMV-myc plasmids were collected at 24 h after transfection and analyzed with antibodies against Rb phosphorylation on Ser-795 and Ser-807/811. As shown in Figure 5A, myc-3a and myc-M were expressed at the expected molecular mass, and when compared with control and myc-transfected cells, the Rb phosphorylation on Ser-795 and Ser-807/811 was down-regulated in myc-3a– and myc-M–transfected cells and was more significant in myc-3a–transfected cells, indicating that the expression of 3a inhibits Rb phosphorylation and block cell cycle progression at G0/G1 phase. The phosphorylation status of Rb was further studied in 3a/pCMV-myc–transfected cells. Expression of myc-3a was observed at all the time points with the highest level at 24 h after transfection, whereas the phosphorylation of Rb on Ser-795 and Ser-807/811 decreased gradually after transfection. In bisacrylamide cross-linked gels, hyperphosphorylated Rb migrates slowly, while hypophosphorylated and nonphosphorylated Rb comigrate and appear as a more rapidly migrating band (18, 20). Using an anti-Rb antibody, the phosphorylation status of Rb was further determined. In control pCMV-myc and M/pCMV-myc–transfected cells, the majority of Rb appeared as a slowly migrating band, while in 3a/pCMV-myc–transfected cells, the rapidly migrating band was detected from 24 h after transfection, indicating hypophosphorylation of Rb. These results suggested that the 3a protein arrests cell proliferation by regulating Rb phosphorylation (Figure 5A).
Rb Ser-795 and Ser-809/811 are targets of active cyclin D/CDK4/6 complexes (23). The observed decrease in its phosphorylation prompted us to examine the expression of cyclin D1, D2, D3, cdk4, and cdk6 in 3a/pCMV-myc–transfected cells by Western blots. No significant changes were observed in the levels of cdk4 and cdk6 between pCMV-myc, M/pCMV-my, and 3a/pCMV-myc. And the expression of cdk4 and cdk6 were not obviously different from 12–48 h after transfection with 3a/pCMV-myc (Figure 5B). In contrast, cyclin D3 was decreased in both myc-3a– and myc-M–expressed cells and became more significantly so in myc-3a–expressed cells when compared with controls. Furthermore, the level of cyclin D3 was significantly decreased from 12–48 h after transfection with 3a/pCMV-myc, which was consistent with the changes of Rb phosphorylation (Figure 5B). The expression of cyclin D1 and D2 was not observed in HEK 293 transfected with 3a/pCMV-myc, M/pCMV-myc, and pCMV-myc, respectively (data not shown).
In addition, cyclin D3 expression and Rb phosphorylation were examined in SARS-CoV–infected Vero E6 cells. In this experiment, no cytopathic effects (CPU) were observed at 6 h after infection, while at CDK-inhibitors (CKIs) are well known to interfere with cell cycle progression and to cause phase-specific cycle arrest. These inhibitors perturb the phosphorylation process by directly interacting with their target proteins, such as cyclins or CDKs. As shown in Figure 5E, p53 phosphorylation on Ser-15 was up-regulated in both 3a and M gene–expressed cells and gradually increased in a time-dependent manner in 3a/pCMV-myc–transfected cells, while the protein levels of p53 were unaffected by 3a and M gene expression. As no significant difference on p53 phosphorylation was observed between 3a and M gene–expressed cells, and the levels of p21Cip and p27Kip, members of the Cip/Kip subfamily, were unchanged by 3a or M expression (data not shown), it is proposed that p53 phosphorylation might be induced by higher expression of foreign proteins at the Golgi apparatus and not associated with 3a-induced G1 phase arrest. According to current concepts, the cell cycle commitment after restriction point passage requires the sustained stimulation by mitogens of the synthesis of labile D-type cyclins, which associate with cyclin-dependent kinase (CDK) 4/6 to phosphorylate Rb family proteins and sequester the CDK inhibitor such as p21(WAF1) and p27kip1. Cyclin D3 is expressed in nearly all proliferating cells and has shown the most broad expression pattern of the three D-type cyclins (cyclin D1, D2, and D3). Lin and coworkers reported that the Cdk6–cyclin D3 complex is unique among the D cyclin and kinase combinations in the ability to promote the cell cycle start, evading the inhibition by p27(KIP1) and p21(CIP1) with a resemblance to viral cyclin-bound Cdk6 (24). PTEN tumor suppressor gene, PKA activation, and some anti-tumor agents such as glucocorticoids were reported to have arrested cell cycle progression in G1 phase by decreasing cyclin D3 mRNA levels and/or by inducing its proteasomal degradation. Furthermore, enforced expression of cyclin D3 abrogated the PTEN-induced cell cycle arrest, while silencing cyclin D3 by RNA interference further inhibited S phase entry, indicating a key role for cyclin D3 repression in these events (25, 26). The decreases in the protein level of cyclin D3 were observed in the measles virus (MV) and MHV-infected cells and may play some roles for the virus in inducing G1 cell cycle arrest (20, 27). In this article, we first reported that the expression of SARS-CoV 3a gene arrested cell cycle progression in G1 phase by a significant down-modulation of cyclin D3. It would be interesting to define the pathway for 3a protein to decrease the expression of cyclin D3. The role of cell cycle arrest induced by the 3a protein was not determined in the life cycle of SARS-CoV, but was proposed from recent studies. Infection of MHV, a member of coronavirus family, was recently reported to result in inhibition of host cellular DNA synthesis and accumulation of cells in G1 phase in activating DBT and 17Cl-1 cells through inducing cyclin D2 and cyclin E degradation (28). The expression of nonstructural protein p28 of MHV was reported to induce G1 phase arrest in transfected cells and might be responsible for MHV to induce cell cycle arrest (18). Increasing data proposed that cell cycle arrest in the G1 phase might favor coronavirus replication and exacerbate virus-induced pathogenicity, especially in some aspect, for example increasing amounts of ribonucleotide pools for efficient coronavirus RNA synthesis, preventing the induction and execution of early cell death in infected cells, assisting in efficient coronavirus assembly, benefiting cap-dependent translation of coronavirus proteins, and decreasing the killing efficiency of coronavirus-infected cells by cytotoxic T cells (20, 28). In this article, we reported that expression of 3a could significantly inhibit cell growth and induce cell G1 phase arrest in different types of transfected cells and infected Vero E6 cells. This suggested that the 3a protein may favor SARS-CoV replication by inducing cell cycle arrest at the G1 phase, and moreover, plays an important role in SARS-CoV–induced pathogenesis.
The authors thank Prof. Eric J. Snijder (Leiden University Medical Center, The Netherlands), Prof. Milton Taylor (Indiana University, USA), Dr. Sherief (Rutgers), and Dr. Gang Li for critical reading of the manuscript; Associated-Prof. Zhou Tao for the assay of confocal microscopy; and Drs. Liu Hong-Yan, Li Su-Yan, and Feng Yan-Bin for the construction of some plasmids. The authors also thank Dr. Baochang Fan (Indiana Univetsity) for some help.
* These authors contributed equally to this work.  This work was supported by a grant from the Nature Sciences Foundation of China (30470093). Originally Published in Press as DOI: 10.1165/rcmb.2005-0345RC on April 5, 2007 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form September 7, 2005 Accepted in final form December 25, 2006
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
CLINICAL RELEVANCE
-actin, cyclin D1, cyclin D2, p53, phospho-p53 on Ser-15, Rb, phospho-Rb on Ser-795 and Ser-809/811, CDK4, CDK6, cyclin D3, cdc2, and cyclin A; all antibodies were purchased from Cell Signaling) and subsequently with horseradish peroxidase–conjugated secondary antibodies. Antibody detection was performed using an enhanced chemiluminescence (ECL) detection kit (Cell Signaling). 
motif, respectively. The amino acid positions for these domains are given below. The G1 phase cell cycle arrests of these 3a mutants are shown on the right (Y, yes; N, no). (B) Subcellular localization of 3a-truncated mutants. The plasmids of serial mutants of 3a (D77–274, D99–274, D176–274, and D1–222) and pDsRed-Golgi were co-transfected into HEK 293 cells separately. At 24 h after transfection, the cells on glass slips were cells were permeabilized with Triton X-100 and viewed with anti-myc antibody. The nucleus was stained with Hoechst 33342. Co-localizations of serial 3a-truncated mutants with Golgi marker were observed under a confocal fluorescence microscope. (C) The effects of 3a-truncated mutants on G1 phase cell cycle arrest. The plasmids of the serial 3a-truncated mutants were transfected into HEK 293 cells. At 24 h after transfection, samples were collected and stained with anti-myc antibody and PI. The DNA contents of cells were measured by flow cytometry as before. The cell cycle profiles of "myc-positive" and "myc-negative" cells are shown in the middle and right columns respectively. The experiments were independently repeated three times.
