Previous Article | Next Article ![]()
Journal of Virology, July 2005, p. 9320-9324, Vol. 79, No. 14
0022-538X/05/$08.00+0 doi:10.1128/JVI.79.14.9320-9324.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
Alla Y. Karpova,
Maren Kops,
and
Peter M. Howley*
Department of Pathology, Harvard Medical School, 77 Avenue Louis Pasteur, Boston, Massachusetts 02115
Received 17 March 2005/ Accepted 23 March 2005
|
|
|---|
|
|
|---|
) and IFN-ß. This signaling results in activation of IFN-stimulated genes (ISGs) that mediate the effects of IFN. IFN regulatory factors (IRFs) are a family of nine cellular factors that bind to consensus IFN-stimulated response elements (ISREs) and induce other ISGs (15, 24).
IRF-3 is unique among IRFs because it is constitutively expressed and is posttranslationally activated immediately following infection by several classes of viruses (11, 27). Upon viral infection, Tank binding kinase 1 and I

kinase 1 (8, 22) phosphorylate the C-terminal residues required for activation of cytoplasmic IRF-3 (14, 21). IRF-3 then translocates to the nucleus, where it potentiates a transcription complex composed of CREB binding protein or p300, ATF/c-jun, and NF
B, activating the transcription of IFN-ß (10, 14, 27).
Infection by many viruses leads to apoptosis by two major pathways that initiate downstream of a death signal; however, the mechanism that triggers programmed cell death is not completely understood. Type I IFNs have been implicated as essential mediators of apoptosis (23); however, cell death after viral infection can also be initiated independently of IFN (28). Overexpression of a constitutively active IRF-3 mutant leads to apoptosis in Jurkat and 293 cells, and expression of wild-type IRF-3 can augment virus-induced apoptosis (10). Interestingly, the role of IRF-3 in virus-induced apoptosis was found to be both IFN and p53 independent (28). In this study, we explored the mechanism by which IRF-3 might mediate the apoptotic response to Sendai virus (SeV) infection.
SeV infection leads to transcriptional upregulation of TRAIL. Heylbroeck et al. (10) and Weaver et al. (28) have implicated caspase 8 as a downstream effector in SeV-induced apoptosis. To identify the SeV-induced apoptotic signals upstream of caspase 8, we infected both primary (human foreskin keratinocytes [HFK]) and tumor (human colonic adenocarcinoma cells [HT-29]) cell lines with SeV and analyzed mRNA expression profiles by RNase protection using probe sets for mRNAs in the tumor necrosis factor (TNF) family. Total cell mRNA was isolated at various times from control or SeV-infected cells, and the RNase protection products were resolved by polyacrylamide gel electrophoresis. Figure 1A and B show results from HFK cells. The induction of IFN-ß mRNA served as a positive control (24, 25), with a strong upregulation observed at 4 h postinfection (Fig. 1B). Interestingly, for the hApo3d probe set, only the TRAIL (TNF-related apoptosis-inducing ligand) mRNA was strongly up-regulated post SeV infection at 4 h, similar to IFN-ß mRNA (Fig. 1A). In addition, we noted the expression of its signaling receptor, DR5, prior to viral infection at 0 h, and that was further upregulated, albeit not to the high levels seen for TRAIL mRNA (Fig. 1A).
![]() View larger version (76K): [in a new window] |
FIG. 1. SeV induces TRAIL in HFK cells (A and B). HFK cells were infected with 200 hemagglutination (HA) units/ml SeV, and RNA was isolated at various time points postinfection. RNase protection was performed with the hApo3d (A) and hCK3 (B) probe sets. TRAIL induction (normalized to glyceraldehyde-3-phosphate dehydrogenase [gapdh]), 12-fold; DR5 induction (normalized to glyceraldehyde-3-phosphate dehydrogenase), 3-fold. Full-length unprotected TRAIL, DR5, and IFN-ß probes are underlined, and lines drawn to protected fragments. TRAIL mRNA is induced by SeV in HT-29 cells (C). HT-29 cells were infected with 200 HA units/ml SeV, and RNA was isolated at various time points postinfection. RNase protection was performed with the hApo3c probe set. TRAIL induction (normalized to L32), eightfold; DR5 induction (normalized to L32), threefold. SeV and IRF-3 induce TRAIL in HEC-1B cells (D). HEC-1B cells were transfected with either pcDNA or pcDNA-IRF-3. At 24 h posttransfection, cells were infected with SeV and RNA was isolated at various time points. RNase protection was performed with the hApo3d probe set. Full-length unprotected TRAIL and DR5 probes are underlined, and lines drawn to protected fragments.
|
To determine whether IRF-3 might play a role in SeV-induced TRAIL expression, we performed RNase protection analysis using IRF-3-transfected cells. Previous experiments that demonstrated a role for IRF-3 in SeV-induced apoptosis had also been performed with IRF-3-transfected cells (10, 28). To avoid possible secondary effects arising from IFN signaling, HEC-1B cells (which lack the IFN receptor) (28) were transfected with either the vector control or an IRF-3 expression plasmid 24 h prior to infection with SeV. As shown in Fig. 1D, TRAIL mRNA was strongly induced in the IRF-3-transfected cells and not in the vector-transfected cells, indicating that IRF-3 contributes to SeV activation of TRAIL transcription.
IRF-3 transactivates the TRAIL promoter following SeV infection.
A TRAIL luciferase reporter construct was cotransfected with plasmids expressing members of the IRF family of transcription factors into HEC-1B cells 18 h prior to SeV infection. The functionality of all expression constructs was verified using an IFN-ß luciferase reporter (data not shown). In addition, two IRF-3 mutants (IRF-3 5D and IRF-3
N60) were tested for the ability to transactivate the TRAIL promoter. Constitutively active IRF-3 5D contains phosphomimetic substitutions of aspartic acid in the carboxy-terminal cluster of serines and threonines (14). IRF-3
N60 lacks 60 amino acids of the DNA binding domain (IRF-3
N60) and functions in a dominant negative fashion (28). Luciferase levels were measured 16 h after SeV infection (Fig. 2). IRF-3 activated the TRAIL promoter about sevenfold after infection compared to the uninfected IRF-3-expressing cells. Neither IRF-1, IRF-2, nor IRF-7 activated the TRAIL promoter reporter (5). As anticipated, IRF-3 5D activated the TRAIL promoter whereas IRF-3
N60 inhibited transactivation of the TRAIL promoter. These data demonstrated that IRF-3 was the only IRF family member tested that could activate the 1371 TRAIL promoter.
![]() View larger version (13K): [in a new window] |
FIG. 2. IRF-3 activates transcription of the TRAIL promoter. HEC-1B cells were transfected with 3 µg of the indicated IRF expression vectors and 1 µg of TRAIL promoter-luciferase reporter (TRAIL 1371) and IFN-ß-luciferase reporter, which were kindly provided by B. Mark Evers (26) and John Hiscott (21), respectively. At 18 h posttransfection, cells were infected with 200 HA units/ml of SeV for 18 h, and luciferase assays were performed with the cell lysates. Gray bars represent luciferase activity of uninfected cells, and black bars represent activity of SeV-infected cells. Error bars represent the results of experiments performed in duplicate. Results are normalized to Renilla luciferase amounts.
|
B, AP-1, NFAT, Sp1, Oct-1, and IRFs (9, 26). The locations of these elements in the TRAIL promoter are shown in Fig. 3.
![]() View larger version (57K): [in a new window] |
FIG. 3. The human TRAIL promoter sequence contains a putative ISRE. The human TRAIL promoter was analyzed using the TRANSFAC transcription factor binding site program. Underlined are putative transcription factor binding sites, and an arrow indicates the transcriptional start site. Putative ISRE sites are bolded.
|
IRF-3 binds the TRAIL promoter following SeV infection in HT-29 cells.
To determine whether IRF-3 binds to the upstream TRAIL ISRE identified by TRANSFAC analysis, we performed no-shift (Novagen) analysis. No-shift assays are similar to electrophoretic mobility shift assays but use an enzyme-linked immunosorbent assay-based format. We prepared nuclear extracts from HT-29 cells that were either untreated or SeV infected. Extracts were incubated with a TRAIL ISRE (5' GCTTCTTTCAGTTTCCCTCCTTTCCAACGA) or ISG15 ISRE (5' ATGCCTCGGGAAAGGGAAACCGAAACTGAAGCCA) double-stranded biotinylated oligonucleotide and transferred to a streptavidin-coated plate. Bound transcription factors were detected using IRF-3-, IRF-7-, and ISGF3
-specific antibodies (Santa Cruz), followed by a secondary antibody conjugated to horseradish peroxidase. The complexes were detected by a chromogenic reaction with a tetramethylbenzidine substrate. Figure 4A and B demonstrate that IRF-3 bound both the TRAIL and ISG15 ISREs threefold and sixfold, respectively, above the background following SeV infection. No significant binding of IRF-7 or ISGF3
was observed with either oligonucleotide. An unrelated TNF-
-specific antibody showed no increase in signal intensity with either oligonucleotide using extract from SeV-infected cells (data not shown).
![]() View larger version (15K): [in a new window] |
FIG. 4. IRF-3 binds to one of the TRAIL ISREs. No-shift assays were performed to detect IRF binding activity to either the TRAIL ISRE (A) or the ISG15 ISRE (B). Assays were performed with either no extract (white bar), uninfected HT-29 cell extract (gray bar), or SeV-infected HT-29 cell extract (black bar). Detection was performed with IRF-3-, IRF-7-, or ISGF3 -specific antibodies as indicated. Each set of reactions was performed in triplicate. (C) Same as in panel A but using mutant TRAIL ISRE oligonucleotides and detection performed with the monoclonal SL-12 IRF-3 antibody. Assays were performed with either no extract (white bar), uninfected HT-29 cell extract (gray bar), or SeV-infected HT-29 cell extract (black bar). Each set of reactions was performed in triplicate. HRP, horseradish peroxidase; wt, wild type.
|
To determine the responsiveness of IRF-3 to ISRE mutations, we constructed mutant versions of the full-length TRAIL promoter containing mutants 1 and 2. These reporters were transfected with an IRF-3 expression plasmid into HEC-1B cells, and luciferase assays were performed following SeV infection. Interestingly, the ISRE mutations significantly impaired both the basal and SeV-activated levels of the TRAIL reporters (data not shown). Thus, the mutations introduced into the TRAIL promoter ISRE may affect the interaction of cellular transcription factors involved in basal, as well as IRF-3-induced, transcription.
In this study, we have discovered that cells upregulate TRAIL in response to SeV infection and that IRF-3 regulates TRAIL at the transcriptional level, thus defining a novel transcriptional target for IRF-3. TRAIL has been shown to be part of the cellular response to infection by several viruses (for a review, see reference 4). Work by Clarke et al. (7) implicated TRAIL as a mediator of reovirus-induced apoptosis, and both TRAIL mRNA and TRAIL DR5 mRNAs are upregulated following infection by reovirus, as well as by cytomegalovirus (7, 20).
As many signaling pathways, including NF
B and ATF/c-jun, are also activated following SeV infection, it is likely that TRAIL induction, like IFN-ß, may result from the activation of several signaling pathways, not only IRF-3. Indeed, several potential response elements have been described within the TRAIL promoter, including those for CEBP, AP-1, GATA, NFAT, NF
B, and SP-1 (9, 26) in addition to the ISREs (this study and reference 3).
TRAIL transcriptional regulation has been largely studied in cells of hematopoietic origin. The IRF family member ISGF3
is critical for the regulation of murine TRAIL in natural killer cells (19). In Jurkat T cells, NF
B is a crucial regulator of TRAIL expression (3). In many other cell lines, TRAIL can be regulated by IFN-
/ß and IFN-
and by IRF-1 (6, 12, 17). It is therefore possible that different IRFs may be able to bind to ISRE upstream elements in the TRAIL promoter, depending on the cell type.
Interestingly, TRAIL has two decoy receptors (DcR1, DcR2) preferentially expressed on normal cells and two death receptors (DR4, DR5) preferentially expressed on tumor cells (1, 2, 16). We have shown that DR5 mRNA is expressed prior to and following viral infection. By expressing both TRAIL and its receptor, the virus-infected cell can ensure its elimination. Experiments are currently ongoing in our laboratory to determine the exact contributions of TRAIL and DR5 to SeV-induced apoptosis.
This study underscores the observation that IRF-3 engages multiple pathways in the cellular innate immune response, and it defines a novel target for IRF-3. These results highlight the important role of IRF-3 in the cellular host response to viral infection and also suggest a mechanism for virus-induced apoptosis, through the induction of TRAIL.
This work was supported by grant R01 AI42257 from the NIH to P.M.H. and by an individual NRSA fellowship (F32 CA 0936328) to J.R.K.
Present address: Synta Pharmaceuticals, 45 Hartwell Ave., Lexington, MA 02421. ![]()
Present address: Cold Spring Harbor Laboratory, 1 Bungtown Road, Cold Spring Harbor, NY 11724. ![]()
Present address: Meisemweg 15, 52078 Aachen, Germany. ![]()
|
|
|---|
B signaling reveals a novel role for NF-
B in the regulation of TNF-related apoptosis-inducing ligand expression. J. Immunol. 167:3164-3173.

and TBK1 are essential components of the IRF3 signaling pathway. Nat. Immunol. 4:491-496.[CrossRef][Medline]
enhances TRAIL-induced apoptosis through IRF-1. Eur. J. Biochem. 271:4222-4228.[Medline]
/ß. Eur. J. Immunol. 31:3138-3146.[CrossRef][Medline]
mediates a novel antiviral activity through dynamic modulation of TRAIL and TRAIL receptor expression. J. Immunol. 163:920-926.
-related pathway. Science 300:1148-1151.
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2010 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»