Previous Article | Next Article ![]()
Journal of Virology, June 2006, p. 5644-5650, Vol. 80, No. 11
0022-538X/06/$08.00+0 doi:10.1128/JVI.02400-05
Copyright © 2006, American Society for Microbiology. All Rights Reserved.
Heidi Palosaari,1,
Jean-Patrick Parisien,1,
Patricia Devaux,2
Roberto Cattaneo,2
Toru Ouchi,3 and
Curt M. Horvath1*
Department of Medicine and Department of Biochemistry, Molecular Biology, and Cell Biology, Northwestern University, Evanston, Illinois, and Department of Medicine, Evanston Northwestern Healthcare, Evanston, Illinois 60208,1 Molecular Medicine Program, Mayo Clinic College of Medicine, Rochester, Minnesota 55905,2 Department of Oncological Sciences, Mount Sinai School of Medicine, New York, New York 100293
Received 15 November 2005/ Accepted 11 March 2006
|
|
|---|
|
|
|---|
In cases of viral oncogenesis, p53 has been demonstrated to interact with proteins encoded by tumor-associated viruses, many times providing a host evasion mechanism for bypassing cell cycle arrest and/or apoptosis (22). Destruction of this important cancer elimination checkpoint abrogates tumor suppressor activity in the infected cells and contributes to the oncogenic properties of tumor viruses. Connections between the p53 family and innate antiviral immunity have been recently established. Evidence indicates p53 pathways may be components of the antiviral responses mediated by interferon (IFN) (17, 35). Transcription of the p53 gene is induced by IFN, accompanied by an increase in p53 protein levels that prime cells toward enhanced p53 responses. The p53 protein can be activated in response to further virus infection, inducing a cellular apoptotic response that can restrict virus replication. Thus, there exists a relationship between death-inducing stress responses and innate antiviral immunity that transcends the tumor suppression properties of p53. This connection is also supported at the level of gene expression, as several gene targets of the IFN system are also subject to regulation by the p53 family (e.g., see references 2, 13, 24, and 25). Furthermore, influenza virus induces p53 activity, including p53-dependent apoptosis (37, 38), and cells deficient in p53 are defective in IFN responses (12, 38). In some cases, the viral protein that targets p53 is also linked to evasion of host innate immunity. For example, human papillomavirus E6 protein can enhance the ubiquitylation and degradation of p53 (32) and also can inhibit the induction of IFN through interaction with IFN regulatory factor 3 (31). Altogether, these findings support a role for the p53 family pathways in cellular regulation of virus infections.
Measles virus, a member of the Morbillivirus genus in the Paramyxoviridae family, is an enveloped negative-strand RNA virus that causes an acute illness, typically during childhood. Measles infections can promote a virus-induced immune suppression that creates susceptibility to opportunistic infections, resulting in over 1 million deaths annually worldwide (21). Measles virus disengages the host innate and adaptive antiviral immune systems in several ways. Like many other paramyxoviruses, measles virus encodes a protein, termed "V," that functions as an inhibitor of host innate antiviral responses. Paramyxovirus V proteins are key mediators of host evasion and have been demonstrated to disengage IFN signal transduction, to prevent IFN biosynthesis, to regulate host range restriction, and to interfere with virus-induced cell death (4, 5, 9, 10, 27, 28, 30, 45, 46). For example, recombinant simian virus 5 (SV5) viruses with specific defects in V protein expression are potent inducers of cell death, while the wild-type virus is not (10, 34, 43, 44). The measles virus V protein binds to a number of cellular proteins, including the IFN signaling proteins, STAT1 and STAT2 (26), allowing for efficient evasion of the host's IFN-induced antiviral immune response. Here, we demonstrate that the p53 family is also a target for measles virus V host interference that enables evasion of programmed cell death.
Paramyxovirus V proteins exert anticellular actions through protein-protein interactions (1, 11, 14, 15). Affinity chromatography was employed previously to identify cellular proteins that interact with measles virus V (26, 36), and two prominent species were identified as p53 and p73 by specific immunoblotting (Fig. 1). A control FLAG-tagged green fluorescent protein (GFP) did not bind to p53 or p73, and the V protein from parainfluenza virus 5 (SV5) precipitated only a trace of p53 and no p73. However, the FLAG-measles virus V eluate contained both p53 and p73
(Fig. 1). These results indicate that both p53 and p73
are able to associate specifically with the measles virus V protein.
![]() View larger version (49K): [in a new window] |
FIG. 1. Measles virus V protein binds to p53 and p73. Immunoaffinity purification of FLAG-tagged GFP, measles virus V (MeV), or SV5 V (SV) expressed in human 293T cells with FLAG M2 affinity resin was carried out as described previously (39). Whole-cell lysates and FLAG tripeptide eluates were subject to immunoblotting (IB) with antiserum for p53, p73, or FLAG epitope as indicated.
|
![]() View larger version (32K): [in a new window] |
FIG. 2. Measles virus V targets p53 and p73 DNA binding domains. (A) FLAG-tagged p53 (i) and the indicated fragments (ii, iii, and iv) illustrated in the box were expressed in 293T cells along with hemagglutinin (HA)-tagged measles virus V (MeV), and complexes were captured by immunoprecipitation (IP) with FLAG M2 affinity resin. Whole-cell lysates and FLAG tripeptide eluates were subjected to immunoblotting (IB) for HA tag to detect measles virus V or FLAG tag to detect p53. An asterisk indicates a nonspecific cross-reactivity smear due to closely migrating fragment ii. (B) HA- or FLAG-tagged p73 (v) or p73 DNA binding domain fragment (vi) illustrated in the box was expressed with FLAG- or HA-tagged measles virus V. The left side shows proteins before immunoaffinity precipitation. The right panel shows proteins collected with FLAG M2 affinity resin, eluted with FLAG tripeptide, and subjected to immunoblotting for HA (top panel) or FLAG (bottom panel) as indicated.
|
Although p53 and p73 mediate distinct functions in mammals, they overlap in their abilities to induce programmed cell death. To test the ability of measles virus V to interfere with apoptosis, parental 2fTGH fibrosarcoma cells and measles virus V-expressing derivatives were generated using replication-defective lentivirus vectors. These cell lines are defective in IFN signaling, consistent with prior conclusions regarding inhibition of JAK-STAT signaling (Fig. 3A) in the presence of measles virus V (26). These cell lines were left untreated or were treated with doxorubicin (Dox) to induce the DNA damage response (47). The cellular DNA content was analyzed by propidium iodide staining and flow cytometry (Fig. 3B). DNA fragmentation was induced by Dox, but the measles virus V-expressing cells were protected from apoptosis compared to wild-type cells, as indicated by the reduction in sub-G0 DNA. To verify this result at the molecular level, immunoblotting was used to detect cleavage of the poly(ADP-ribose) polymerase (PARP), another hallmark of apoptosis (Fig. 3C). In 2fTGH cells, the 89-kDa PARP cleavage fragment was initially detected after 20 h of treatment with 0.5 µg/ml Dox and accumulated thereafter. In contrast, expression of measles virus V delayed initial detection of cleaved PARP to 24 h, indicating delayed apoptosis induction.
![]() View larger version (20K): [in a new window] |
FIG. 3. Measles virus V protein expression suppresses cell death and alters accumulation of proapoptotic mRNAs. (A) Functional expression of measles virus V in stable cell lines. Human fibrosarcoma cell line 2fTGH and two measles virus V-expressing derivatives made using recombinant lentivirus vectors as described previously (40) using the pLenti system were subjected to IFN- -responsive luciferase assays to verify the ability of measles virus V to suppress IFN signaling. Cells were treated with 1,000 U/ml IFN- for 6 h beginning 24 h after reporter transfection and then lysed. (B) Analysis of DNA fragmentation. Human fibrosarcoma cell line 2fTGH and the measles virus V-expressing derivative were left untreated () or were treated with Dox (+; 1 µg/ml) for 24 h and then stained with propidium iodide. The fraction of cells exhibiting sub-G0 DNA content is indicated as percent apoptosis. Average values with standard deviations are graphed (n = 3). (C) Analysis of PARP cleavage. 2fTGH cells or measles virus V-expressing cells were treated with doxorubicin (0.5 µg/ml) for the times (h) indicated, and whole-cell extracts were subjected to immunoblotting with antiserum for PARP or FLAG (to detect measles virus V protein). (D) Analysis of mRNA accumulation in response to Dox treatment. Focused gene expression DNA arrays for human p53-related genes were obtained from SuperArray Biosciences and probed and analyzed according to the manufacturer's recommendations using total RNA from 2fTGH and measles virus V-expressing derivative with and without Dox treatment. Cells were treated for 16 h with 1 µg/ml doxorubicin or left untreated, and data are the average increase (fold) by Dox treatment of three independent experiments.
|
To test the ability of measles virus V to specifically influence p53 transcriptional activity, reporter gene assays were carried out in human 293T cells. Expression of p53 can potently activate a luciferase reporter gene carrying a multimeric canonical p53/p73 response element. Expression of measles virus V did not diminish the transcriptional response due to p53 expression (Fig. 4A). In fact, enhanced reporter gene activity was observed when measles virus V was coexpressed with p53 in this assay. To validate this result, endogenous p53 activity was induced with a specific p53 agonist, Nutlin 3 (42), which disrupts the p53 binding site of Mdm2 (Fig. 4B). By activating p53 accumulation, Nutlin 3 treatment induced accumulation of Mdm2 mRNA (a p53 gene target) regardless of measles virus V expression. This result confirms that p53 activity is not inhibited by measles virus V protein, excluding p53 as an endogenous target for measles virus V protein interference.
![]() View larger version (27K): [in a new window] |
FIG. 4. Measles virus V selectively inhibits p73-mediated transcription and PUMA expression. (A) Effects of measles virus V (MeV) on p53-mediated gene regulation. Cells were cotransfected with p53 luciferase reporter plasmid, p53 expression vector, and measles virus V expression vector in the ratios indicated and assayed 36 h later. The data plotted are normalized to a cotransfected Renilla luciferase control, expressed as percent p53 activity alone (n = 3). (B) Activity of endogenous p53 on Mdm2 mRNA accumulation. 2fTGH and two measles virus V-expressing derivative clones were treated with the indicated concentrations of Nutlin 3 for 12 h, and then total RNA was processed for reverse transcription-PCR (RT-PCR) with specific primers for Mdm2 and GAPDH. (C) Measles virus V suppresses p73-mediated gene regulation. Cells were cotransfected with p73 reporter plasmid, p73 expression vector, and measles virus V expression vector in the ratios indicated and assayed 36 h later. Data plotted are normalized to a cotransfected Renilla luciferase control, expressed as percent p73 activity alone (n = 3). (D) Analysis of measles virus V effects on PUMA protein expression. 2fTGH and measles virus V-expressing derivative were treated with Dox (0.5 µg/ml) for the times indicated, and lysates were subjected to immunoblotting with specific antisera for PUMA, p21, -actin, and FLAG (to detect measles virus V).
|
Expression of the PUMA protein is regulated by p73 during apoptosis induction (18). To determine the effect of measles virus V on endogenous p73 responses, the abundance of PUMA protein was tested by immunoblotting (Fig. 4D). PUMA was readily detected in parental 2fTGH cells, and its abundance remained relatively steady during 10 h of doxorubicin treatment. In contrast, detection of PUMA was diminished in cells harboring measles virus V protein. No changes were observed in the abundance of mRNAs encoding p21WAF1 or
-actin. Together, these data support the concept that p73 is a target of measles virus V protein interference with apoptosis.
To assess whether measles virus V protein acts as an inhibitor of cell death in a natural context, recombinant-derived wild-type and V-deficient measles viruses (named MVvac and MVvacVko, respectively) (3) were grown and their titers were determined in Vero cells for use in assays for virus-induced cytopathic effects (Fig. 5). It has been previously documented that apoptosis has a significant contribution to measles virus-induced cytopathicity (6). 2fTGH cells were inoculated with equal amounts of serially diluted viruses and assayed 3 days later by fixation and staining with methylene blue. While both viruses induced dose-dependent cytopathic effects, the V-deficient measles virus was more cytopathic than the wild type. To reduce any potential complications due to IFN production, Vero cells, a line that does not produce IFN, were used in the same assay. Again, the V-deficient measles virus was a more potent inducer of cytopathic effect than wild-type virus. Together, these results provide genetic support for the concept that the measles virus V protein is a modulator of virus-induced cell death and are in agreement with prior studies of paramyxovirus V proteins.
![]() View larger version (14K): [in a new window] |
FIG. 5. Measles virus lacking V protein is more cytopathic. (A) Immunoblot analysis of the parental virus MVvac and its derivative, MVvacVko. HeLa cells were infected with parental virus MVvac (lane 4) and its derivative, MVvacVko (lane 1). Antisera against P/V proteins (upper panel) and against the N protein (middle panel) were used. Anti-ß-actin (lower panel) was used as a loading control. Mock, uninfected cell control (lanes 2 and 3). (B) 2fTGH cells (top two panels) and Vero cells (bottom two panels) were plated in a 96-well dish and then infected with a serial 1:2 dilution of MVvac or MVvacVko virus in 50 µl of serum-free Dulbecco's modified Eagle's medium, beginning with 1 x 105 PFU in well 2. One hundred fifty microliters Dulbecco's modified Eagle's medium plus 5% serum was added at 2 h postinfection, and plates were incubated at 37°C for 72 h prior to fixation and staining with 3% methylene blue in 50% ethanol. C, control uninfected cells.
|
Protein interaction analysis implicated p53 and p73 as potential targets and demonstrated that measles V can interact with the conserved p53 family DNA binding domain. Mechanistic studies determined that the transcriptional activity of p73 is specifically inactivated by measles virus V protein. Measles virus V protein expression induces a delay in PARP cleavage in response to genotoxic stress, and suppression of p73-dependent transcription results in a substantial inhibition of steady-state PUMA expression in the measles virus V-containing cells, providing a plausible explanation for the antiapoptotic actions of measles virus V.
Enhanced cytopathicity was observed during infections with V-deficient measles virus. While many factors can contribute to the cytopathic effect assay, measles virus has been previously demonstrated to induce cytopathic effects via characteristic apoptotic pathways (6). Although previous evaluation of recombinant measles viruses did not uncover antiapoptotic effects attributable to V protein in cultured cells (33), V-deficient mutant viruses had altered growth and virulence properties in animal models (20, 29, 41). The exact reason for this discrepancy is not readily apparent from the literature but might reflect differences in methodology or cell lines used. A more trivial explanation could relate to the reported inability of the Edmonston tag strain V protein to antagonize IFN responses compared with other measles strains (23), possibly pointing toward loss-of-function mutations in the V protein (33). Notably, the increased cytopathicity of V-deficient measles virus appears to be less pronounced than that reported for V-deficient SV5 (10, 43, 44), and this could reflect the presence of measles virus C protein, which might also participate in modulation of host responses to infection. Further experimentation will be required to evaluate the effects of C protein in the assays for cell death carried out in the stable cell lines, as C protein expression has not been assessed. However, the coprecipitation analysis is strictly specific for V protein and the recombinant viruses retain C protein expression. Irrespective of potential C protein contributions, it is a common observation that genetic disruption of paramyxovirus V proteins results in viruses with enhanced cell death capacities, suggesting that avoidance of cell death is a general property of the paramyxovirus V protein.
Despite its ability to bind p53, no inhibition of p53-mediated gene regulation was observed in measles virus V-expressing cells. The underlying reason for this discrepancy has not been revealed by these studies, but this finding indicates specificity in measles virus V host interference. The results do not rule out the possibility of a functional physical association with p53 that does not influence transcriptional activity or coprecipitation of p53 via a bridging factor. As measles virus V protein interacts with the most highly conserved region shared by p53 and p73, the DNA binding domain, it would be reasonable to suspect the coprecipitation with p53 might be a result of molecular mimicry due to structural similarities between p73 and p53 in this domain. Narrowing the precise mode of p73 inhibition will require further investigation, but the association with the DNA binding domain suggests that measles virus V might interfere with p73-promoter contacts or alter essential coactivator recruitment. Regardless of the mechanism, the observed specificity of viral antagonism is similar to that observed for viral p53 inhibitors from adenovirus (E1B 55K), SV40 (T antigen), and human papillomavirus (E6). While all of these viral oncoproteins are effective p53 inhibitors, none of them targets p73 (16). Together with our observation of specific viral p73 antagonism, these results support divergent functions of these two family members with respect to virus replication.
The V proteins from a number of paramyxovirus species have been demonstrated to prevent innate antiviral immunity by distinct mechanisms. These include targeting JAK-STAT signal transduction downstream of interferon signals (reviewed in reference 11), inhibition of IFN biosynthesis in response to virus infection (30), and suppression of cell death (10, 43, 44). While measles virus V binds to both p53 and p73 DNA binding domains, transcriptional activity of p73 is specifically blocked by expressing measles virus V cDNA, and this is reflected in the regulation of the proapoptotic target gene, PUMA. While the interference with p73 function represents only one participant in the regulation of virus-induced cell death that might explain the function of V protein in antiapoptosis, there are likely additional components that contribute to the phenotype. However, these findings suggest a novel aspect of the p73 pathway leading to PUMA expression as a determinant of innate antiviral immunity and provide the first description of p73 antagonism by a negative-strand RNA virus.
This work was supported by NIH grants AI50707 to C.M.H. and AI63476 to R.C.
C.D.C., H.P., and J.-P.P. contributed equally to this study. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»