Posttranscriptional Suppression of Interleukin-6 P
http://www.100md.com
病菌学杂志 2005年第1期
Cardiff School of Biosciences, Cardiff University
Section of Infection and Immunity, University of Wales College of Medicine, Cardiff, Wales, United Kingdom
ABSTRACT
Human cytomegalovirus (HCMV) has evolved multiple strategies for suppression of the antiviral response of the infected cell. DNA array technology has revealed that HCMV clearly regulates host gene expression during the course of a productive infection by enhancing, sustaining, or suppressing steady-state levels of cellular transcripts. Interleukin-6 (IL-6) is a pleiotropic cytokine that plays a central role in the immune response to infection. Here we report a detailed study of the effects of HCMV infection on IL-6 expression by human fibroblasts. UV-inactivated virus was found to induce high levels of IL-6 mRNA and protein expression, and IL-6 mRNA remained abundant in cells 16 h after inoculation even though the level of ongoing IL-6 transcription was not significantly enhanced. In lytic HCMV infections, the onset of viral gene expression resulted in two apparently antagonistic effects on IL-6 expression: (i) transcriptional activation, mediated at least in part by the IE2p86 protein, and (ii) posttranscriptional suppression mediated by destabilization of IL-6 mRNA. Transcriptional activation was outweighed by the suppressive effect, such that cells undergoing productive infection produced less IL-6 than cells challenged with inactivated virus. Suppression of IL-6 expression was independent of the viral IL-10 homologue, cmvIL-10. Destabilization of IL-6 mRNA was observed to coincide with the enhanced expression and aberrant intracellular localization of HuR, an mRNA-binding protein known to interact with IL-6 and other mRNAs containing 3' AU-rich elements. Our data suggest a novel mechanism for gene regulation by HCMV at the posttranscriptional level.
INTRODUCTION
Human cytomegalovirus (HCMV) is a ubiquitous, clinically important herpesvirus. Following primary infection, HCMV persists throughout the lifetime of the host, during which the virus must avoid elimination by host immune mechanisms. Thus, HCMV has become a paradigm for viral immune evasion (1, 35, 44). In recent years a remarkable series of HCMV immune evasion or modulation systems have been elucidated, including the following: down-regulation of surface MHC-I expression (1), evasion of natural killer (NK) cell-mediated killing (60), and altered expression of immune signaling molecules and interferon response genes (1, 6, 47). Furthermore, these studies have shown that the virus has acquired or evolved in its genome a number of homologues of immune regulators, including chemokines and chemokine receptors (35, 58). The assumption that such molecules play a direct role in persistence and pathogenesis in the host is supported by the fact that attenuated laboratory strains (AD169 and Towne) lack a set of additional genes, associated with immunomodulatory functions, that are present in virulent clinical or low-passage isolates of HCMV (10, 42). Thus, HCMV has a direct impact on normal immune signaling, to promote an infection associated with a wide range of life-threatening conditions in the immune-compromised host (41).
Interleukin-6 (IL-6) plays a central role in both innate and acquired immune responses. IL-6 is the predominant inducer of the acute-phase response, an innate immune mechanism which is triggered by infection and inflammation (45). IL-6 also plays multiple roles during subsequent development of acquired immunity against incoming pathogens, including regulation of cytokine and chemokine gene expression, stimulation of antibody production by B cells (45), regulation of macrophage and dendritic cell differentiation (12), and response of regulatory T cells to microbial infection (40). In addition to these roles in pathogen-specific inflammation and immunity, IL-6 levels are elevated in chronic inflammatory conditions, such as rheumatoid arthritis, and indeed antibodies against IL-6 and the soluble form of its receptor have been used therapeutically for this condition (30). The inflammatory effects of IL-6 have also been implicated as a factor in the failure of organ and tissue grafts; moreover, elevated levels of the cytokine have been reported to accompany HCMV replication in transplanted lungs and bone marrow during episodes of inflammation or rejection (18, 28, 48, 55, 56). Given the ability of HCMV to persist in the infected host, it might be anticipated that the virus has evolved mechanisms to modulate the expression or signaling of IL-6 as part of the viral armory of immune evasion strategies. However, the effects of productive HCMV infection on IL-6 expression are not well understood.
Transcription of the IL-6 gene can be induced by several cellular factors in response to a range of physiological stimuli; moreover, the mRNA is subject to extensive regulation at the posttranscriptional level. Some inducers, such as IL-1, induce both transcription and stabilization of IL-6 mRNA, whereas tumor necrosis factor alpha induces transcription but not stabilization of the mRNA (39). Regulation of IL-6 mRNA stability is mediated through the 3' untranslated region (UTR), which contains a number of AU-rich elements (AREs). AREs occur in the mRNAs of many genes involved in inflammation, immune signaling, and cell growth and proliferation (4). Their presence in an mRNA generally promotes its degradation in the unstimulated cell but can confer stability in response to appropriate—typically inflammatory—stimuli such as IL-1; these effects on mRNA stability are mediated by a number of both positive (stabilizing) and negative (destabilizing) proteins in the cell, although the precise mechanism of action of ARE-binding proteins has not yet been elucidated fully.
The effects of HCMV infection on cellular gene expression are initiated with binding of the virion to the cell surface (22), with interaction between glycoprotein B and Toll-like receptor 2 playing a major role in this process (13). The binding of virions, or even glycoprotein B in isolation, leads to the induction of inflammatory cytokines and interferon-responsive genes (3, 47, 61). IL-6 mRNA expression is up-regulated by HCMV infection in the absence of de novo expression of viral genes (6, 8, 24, 62, 63), although the IL-6 promoter does not contain recognized interferon response elements. However, virion binding activates multiple intracellular signal transduction pathways, including the phosphatidylinositol kinase, mitogen-associated protein/extracellular signal-regulated kinase, and protein kinase C pathways, all of which lead to the activation of nuclear factors such as NF-B and p38 (13, 22), which are known inducers of IL-6 gene transcription (45). Since the prolonged expression of inflammatory and antiviral genes would likely be detrimental to viral replication, HCMV has evolved mechanisms to counter the host cell's innate antiviral defenses: another virion protein, the matrix protein pp65, has recently been shown to inhibit expression of many of the interferon-responsive genes that are induced by HCMV infection (5). Nevertheless, wild-type HCMV virions still induce up-regulation of a substantial number of proinflammatory genes, showing that the presence of pp65 is not sufficient to completely overcome the innate immune response. Indeed, elevated levels of prostaglandins appear to be required to activate viral gene expression (64). Thus, the induction of many inflammatory cytokines and interferon response genes is transient (6, 27), and de novo expression of virus-encoded genes acts in concert with incoming pp65 to suppress cellular antiviral gene functions (5).
While HCMV infection is known to produce profound changes in the regulation of cellular genes, published studies provide only limited insight into the control of IL-6 gene expression in the productively infected cell. Although the 72-kilodalton HCMV immediate-early (IE) protein 1 (IE1p72) can mediate transcriptional activation of the IL-6 promoter in monocytic cells (24), expression of IE genes is limited in undifferentiated monocytes (49), and it is not known whether such activation occurs during productive infection. Moreover, most previous work has been carried out in the presence of serum, which has profound effects on the expression of cytokines and induces IL-6 expression in fibroblasts (29). In this study, we analyzed in detail the regulation of IL-6 expression in HCMV-infected primary human fibroblasts, in both the absence and the presence of serum. Fibroblasts remain the best-characterized cell system for the study of productive HCMV replication and are a potential source of IL-6 during localized episodes of viral replication in vivo. Our data reveal a hitherto-unsuspected level of complexity in the regulation of IL-6 expression in the infected cell and an apparently novel mode of gene regulation by HCMV at the posttranscriptional level.
MATERIALS AND METHODS
Cells. Human fetal foreskin fibroblasts (HFFF) and human glioblastoma astrocytoma cells (U373-MG) were obtained from the European Cell and Culture Collection (Salisbury, United Kingdom) and were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS; Invitrogen, Paisley, United Kingdom). All cells were tested periodically for mycoplasma contamination using the Mycoplasma Plus PCR primer set (Stratagene, La Jolla, Calif.) and found to be negative.
Virus culture and infection. The attenuated HCMV strain Towne and the low-passage strain Toledo were gifts of E. Mocarski (Stanford University, Stanford, Calif.). Strain Merlin, a low-passage HCMV strain with wild-type gene content, has been described elsewhere (17). A Toledo mutant lacking the early (?) 2.7-kb RNA gene (?2.7 gene) and the corresponding revertant virus have been described previously (33). Where indicated, virus was inactivated by UV irradiation (at 254-nm wavelength) using a biocidal lamp for 30 min at a distance of 20 cm; the amount of UV-inactivated virus used in experiments is expressed in terms of the titer prior to inactivation.
Analysis of IL-6 protein and mRNA in infected fibroblast cultures. HFFF were seeded in tissue culture vessels (Nunc; Fisher Scientific, Loughborough, United Kingdom) at 50% confluence and allowed to adhere overnight in DMEM-10% FCS. Except where indicated, cells were then were starved in serum-free medium for 24 h prior to infection or other challenge. Cells were inoculated either with active HCMV, typically at a multiplicity of infection (MOI) of 1 or as indicated, or with the same amount of UV-inactivated virus; mock-infected cultures received a volume of DMEM-10% FCS equal to that of the viral inoculum in each experiment. For the assay of IL-6 protein, culture medium was clarified by centrifugation (1,200 x g for 5 min) and stored at –20°C prior to analysis by IL-6-specific enzyme-linked immunosorbent assay (Quantikine kit; R&D Systems, Abingdon, United Kingdom). For the assay of IL-6 and ?-actin mRNA by reverse transcription (RT)-PCR, cells were harvested and resuspended in RNAlater reagent (Ambion, Austin, Tex.) prior to purification of RNA. After removal of the storage reagent, total RNA was isolated using the RNeasy Mini kit (QIAGEN, Crawley, United Kingdom), and yields were estimated by densitometry after gel electrophoresis. Equal amounts of RNA were used as templates for cDNA synthesis using SuperScript II reverse transcriptase (Invitrogen) primed with random hexamers (Amersham Biosciences, Amersham, United Kingdom) as recommended by the manufacturers. PCRs were carried out using a forward primer specific for exon 1 of the IL-6 gene (RT primer 1, 5'-CAGCTATGAACTCCTTCTCCAC-3') and a reverse primer specific for exon 2 (RT primer 2, 5'-GAGATGCCGTCGAGGATGTAC-3'); spliced ?-actin cDNA was amplified using primers 5'-ACAACGGCTCCGGCATGTGCAA-3' (forward, exon 1) and 5'-CCATGTCGTCCCAGTTGGTGAC-3' (reverse, exon 2). PCR products were analyzed by agarose gel electrophoresis and quantified by densitometry. In all RT-PCR analyses, PCRs were determined as being in the exponential phase of amplification by sampling reactions at different cycle numbers and plotting the increase in band intensity against cycle number.
Filtration and transfer of conditioned medium. In experiments involving transfer of conditioned medium, the medium from donor cultures was removed at the time indicated and filtered through syringe-mounted 0.1-μm-pore-size filters (Pall Gelman Laboratory, Ann Arbor, Mich.). An aliquot was retained for analysis of IL-6 by enzyme-linked immunosorbent assay, and the conditioned medium was then transferred onto recipient cell cultures, which had either received no prior treatment or had been challenged with UV-inactivated HCMV. In all cases, donor and recipient cell monolayers were washed briefly with phosphate-buffered saline (PBS) between removal of the original medium and addition of fresh or conditioned medium as appropriate.
Plasmids and transfections. Luciferase reporter constructs containing the human IL-6 promoter (p1168hu.IL6P-luc+, p234hu.IL6P-luc+, p110hu.IL6P-luc+, and p50hu.IL6P-luc+) were gifts of G. Haegeman (University of Gent, Ghent, Belgium) (57). Plasmids pcDNA3IE72 and pcDNA3IE86, for expression of HCMV (strain AD169) IE1p72 or IE2p86, respectively, have been described previously (7).
HFFF were transfected using Effectene reagent (QIAGEN); U373-MG cells were transfected using Transfast reagent (Promega, Madison, Wisc.). Where indicated, cells were infected with HCMV 24 h posttransfection and harvested 40 h posttransfection for assay of reporter gene activity. Luciferase activity was assayed using the Luciferase Assay System (Promega) and measured on a TD-20/20 luminometer (Turner Designs, Sunnyvale, Calif.).
Nuclear run-on assays. HFFF were seeded, serum starved, and inoculated as described above for analysis of IL-6 protein and mRNA. Transcription of the endogenous IL-6 gene was assayed using an RT-PCR-based technique, essentially as described by Hildebrandt and Neufer (26). Briefly, cells were harvested at 16 h postinfection (hpi); after gentle cell lysis, the nuclei were isolated, washed, and stored at –80°C in storage buffer (75 mM HEPES [pH 7.9], 40% [vol/vol] glycerol, 60 mM KCl, 15 mM NaCl, 5 mM magnesium acetate, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM dithiothreitol, 0.5 mM spermidine, 0.15 mM spermine, 2 μg of aprotinin/ml, 2 μg of leupeptin/ml). After thawing, each sample was divided in two; one aliquot was mixed with reaction buffer containing nucleoside triphosphates (NTPs; final concentrations: ATP, 1.2 mM; CTP, GTP, and UTP, 0.6 mM each), and the other received control buffer lacking nucleotides; run-on reaction mixtures were then incubated for 15 min at 22°C. RNA was isolated and used for cDNA synthesis as described above; control reactions were performed without reverse transcriptase in each case. Spliced IL-6 and ?-actin cDNA products were amplified using the primers described above; unspliced IL-6 cDNA products were amplified using RT primer 1 and a reverse primer specific for intron 1 of the IL-6 gene (RT primer 3, 5'-GGGTGACTGACAGCACAGCTG-3'). PCR products were quantified by densitometry following agarose gel electrophoresis.
RNA stability assays. Replicate cultures of serum-starved fibroblasts were inoculated at an MOI of 1 with either HCMV or UV-inactivated HCMV, as described above. At 16 hpi, transcription was inhibited by the addition of actinomycin D (Sigma-Aldrich, Poole, United Kingdom) to a final concentration of 5 μg/ml; cultures were harvested either immediately following actinomycin D treatment (0-h time point) or at 4 or 8 h after treatment. cDNA was prepared, and levels of IL-6 and ?-actin mRNA were analyzed using exon-specific primers as described above; the same cDNA samples were used to analyze cyclooxygenase-2 (COX-2) expression, using the following PCR primers: 5'-GTCCACCAACTTACAATGCTG-3' (forward) and 5'-ACATCATGTTTGAGCCCTGG-3' (reverse).
Western blotting. Cell lysates for analysis by Western blotting were resolved by denaturing sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using standard techniques and electroblotted onto polyvinylidene difluoride membranes (Hybond-P; Amersham Biosciences). Membranes were blocked in Tris-buffered saline (10 mM Tris-HCl [pH 8.0], 150 mM NaCl) containing 0.05% (vol/vol) Tween-20 and 5% (wt/vol) dried milk for 1 h at room temperature and then probed sequentially with primary and horseradish peroxidase (HRP)-conjugated secondary antibodies, diluted in a mixture of Tris-buffered saline containing 0.05% (vol/vol) Tween-20 and 5% milk, as indicated. After washing, membrane-bound HRP activity was detected by enhanced chemiluminescence (SuperSignal West Pico reagent; Perbio, Tattenhall, United Kingdom). Monoclonal antibody MAB810, specific for HCMV IE1p72/IE2p86, was from Chemicon (Temecula, Calif.); antibodies specific for HuR (sc-5261) or -tubulin (sc-5286) and HRP-conjugated anti-mouse immunoglobulin G (IgG; sc-2005) were from Santa Cruz Biotechnology (Santa Cruz, Calif.).
Indirect immunofluorescence. Serum-starved fibroblasts were grown on sterilized glass coverslips and inoculated with HCMV as appropriate. At 40 hpi, medium was removed, and cell monolayers were washed with PBS. Cells were fixed in PBS-2% (wt/vol) formaldehyde for 15 min, washed, and then permeabilized in PBS-0.2% (vol/vol) Triton X-100 for 5 min. After washing, coverslips were incubated with primary antibody (either HuR-specific IgG1 [sc-5261; Santa Cruz] or normal mouse IgG1 [sc-2025; Santa Cruz]) at a concentration of 1 μg/ml in PBS-10% (vol/vol) FCS. Coverslips were washed and then incubated with Texas Red-conjugated goat anti-mouse IgG (ab6787; Abcam; 5 μg/ml in PBS-10% FCS). Unbound sites on the secondary antibody were blocked by incubation of coverslips in PBS-10% FCS-10% (vol/vol) mouse serum (Abcam) for 30 min, prior to incubation with fluorescein isothiocyanate (FITC)-conjugated mouse antibody specific for HCMV pUL44 (a gift of Richard Greaves), diluted in PBS-10% FCS-10% mouse serum containing 100 ng of 4',6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich)/ml. After further washing, coverslips were mounted onto glass slides in glycerol containing 2% (wt/vol) 1,4-diazabicyclol[2.2.2]octane (DABCO; Sigma-Aldrich) and sealed with clear nail polish. Slides were viewed on a Leica (Milton Keynes, United Kingdom) DM IRBE fluorescence microscope, and images were captured using Openlab software (Improvision, Coventry, United Kingdom).
RESULTS
Induction of IL-6 expression by HCMV is transient in productive infection. Previous reports have shown that fibroblasts challenged with HCMV express elevated levels of IL-6 protein (8, 34) and mRNA (6, 8, 62, 63); there is, however, little data regarding the continued regulation of IL-6 expression as infection progresses through the lytic cycle. To discriminate between the effects induced by virion binding and/or uptake and de novo expression of HCMV gene products, cells were infected with HCMV or UV-inactivated HCMV, and the levels of secreted IL-6 were measured. HCMV and UV-inactivated virus both induced the release of substantial quantities of IL-6 (Fig. 1A), with higher levels in cultures challenged with UV-inactivated virus than in those infected with HCMV at 16 and 40 hpi. This induction of IL-6 is not mediated by autoregulation, since neither uninfected nor HCMV-infected fibroblasts express a functional IL-6 receptor (S. A. Jones, personal communication). The abundance of IL-6 mRNA at 16 hpi correlates with the level of IL-6 in the medium; IL-6 mRNA levels were up-regulated by HCMV infection yet were approximately twofold more abundant in cells challenged with UV-inactivated virus (Fig. 1B). However, by 40 hpi, the level of IL-6 mRNA had declined sharply in productively infected cells, whereas the message remained abundant in cells challenged with UV-inactivated virus. IL-1? induced high levels of both IL-6 protein and mRNA (Fig. 1A and B, respectively), confirming that cells responded normally to a known regulator of IL-6 expression. IL-6 expression was stimulated in a manner similar to that seen with the attenuated Towne strain by two low-passage, pathogenic strains of HCMV, Toledo and Merlin (Fig. 1C). The above experiments were performed under serum-free conditions, to eliminate the confounding effects of serum on cytokine expression; however, experiments performed in the presence of 10% serum also showed that while HCMV infection induced IL-6 expression from fibroblasts, higher levels of IL-6 were induced by UV-inactivated virus (Fig. 1D). The induction of IL-6 required a direct interaction between virion and cell, as removal of virus particles from the inoculum by filtration through a 0.1-μm-pore-size filter inhibited the production of IL-6, whereas use of a filter which allowed the passage of virions had no effect (Fig. 1E).
The results implicate a factor associated with virion binding and/or uptake being responsible for the sustained up-regulation of IL-6 production, mediated at least in part by an increase in the steady-state level of IL-6 mRNA. This effect is countered, albeit not completely, by the onset of HCMV lytic gene expression. To determine whether increased numbers of virion particles could overcome the down-regulation of IL-6 secretion observed in lytic infection, we challenged cells at a higher MOI with either HCMV or UV-inactivated HCMV alone or a mixture of both active and inactivated virus. While cells challenged with UV-inactivated virus alone again expressed much higher levels of IL-6 than productively infected cells (Fig. 1E), cells receiving a mixed inoculum (equivalent to a total of 10 PFU/cell) expressed a level of IL-6 which was similar to that seen with active virus alone and was greatly reduced compared with that expressed by cells receiving the same amount of UV-irradiated virus. Thus, the onset of viral gene expression results in suppression of continued IL-6 expression by the infected cell; the higher levels of IL-6 seen in cultures challenged with UV-inactivated virus are due to the lack of such suppression and are not an artifact of UV irradiation of the inoculum. Consistent with this, we observed no IL-6 expression after mock infection of control cultures using a UV-irradiated inoculum (data not shown). Furthermore, at high multiplicities of infection the virus restricts further accumulation of IL-6 between 16 and 40 hpi, whereas cultures challenged with inactivated virus continue to accumulate IL-6 in the medium during this period, consistent with the hypothesis that suppression of IL-6 secretion is an active process requiring HCMV lytic gene expression.
Induction of IL-6 expression by HCMV is not mediated by soluble factors. The results above show that induction of IL-6 requires contact between virion and cell. However, it is not clear whether this contact induces IL-6 expression directly by intracellular mechanisms (e.g., by activation of signal transduction pathways which lead to transcriptional up-regulation of the IL-6 promoter by cellular factors or by direct activation of the IL-6 promoter by a virion transactivator) or HCMV induces the release of other soluble factors into the medium, which subsequently trigger IL-6 expression in an autocrine or paracrine manner. To address this, virus-free tissue culture supernatants were collected from HCMV-infected cells and cells challenged with UV-inactivated HCMV (donors) at 6 or 16 hpi and transferred onto untreated recipient cells (Fig. 2). In all cases, the concentrations of IL-6 observed in recipient cell cultures, subsequent to incubation with the transferred medium, were very similar to those in the donor medium at the time of transfer; i.e., medium from virus-challenged cells possessed little ability to induce IL-6 expression de novo. Furthermore, medium from virus-challenged cells failed to induce significant expression of IL-6 mRNA in recipient cultures (Fig. 2B and D). IL-6 continued to be secreted from donor cells challenged with UV-inactivated HCMV after replacement of the conditioned medium with fresh DMEM at 6 or 16 hpi, whereas cells undergoing active infection produced no significant amounts of IL-6 after removal of the original medium (Fig. 2A and B, respectively). In contrast, medium from donor cells treated with IL-1? induced high levels of IL-6 mRNA and protein expression in recipient cultures, whereas the donor cells failed to produce significant amounts of IL-6 after removal of IL-1?. These data do not support a mechanism whereby HCMV induces up-regulation of IL-6 expression by an indirect pathway involving the prior release of soluble factors into the medium. As an additional test, tissue culture supernatants from this experiment were assayed for a number of known inducers of IL-6. IL-1?, tumor necrosis factor alpha, and interferon-? could not be detected in any samples following virus infection; low levels of transforming growth factor ?1 (TGF-?1) were detected at 40 hpi, but not reproducibly at 16 hpi, and at similar levels in cells inoculated with HCMV or UV-inactivated HCMV (data not shown).
Suppression of IL-6 expression is not mediated by an autocrine pathway in HCMV-infected fibroblasts. The results above show that induction of IL-6 expression is not mediated primarily by an autocrine or paracrine mechanism, but they do not address the suppression of IL-6 secretion that is observed subsequent to the onset of lytic gene expression. One potential candidate which might mediate suppression is the HCMV homologue of IL-10 (cmvIL-10), and indeed this has been shown to inhibit lipopolysaccharide-mediated induction of IL-6 by monocytes (51) and signal via the cellular IL-10 receptor (31). However, inhibition of IL-10 signaling by use of a neutralizing antibody against the IL-10 receptor had no significant effect on the levels of IL-6 induced by HCMV or UV-inactivated HCMV (Fig. 3).
To test further whether cmvIL-10 or another soluble factor might be responsible for the suppression of IL-6, conditioned medium from HCMV-infected donor cells was transferred at 16 hpi onto recipient cells that had been challenged 6 h previously with UV-inactivated HCMV. Recipient cultures were then incubated for a further 16 h. If the conditioned medium contained a soluble suppressor, then subsequent IL-6 secretion by the recipient cultures should have been inhibited. However, recipient cultures receiving medium from HCMV-infected cells produced an amount of IL-6, after transfer of medium, similar to that produced by control cultures receiving either fresh DMEM or conditioned medium from mock-infected cells (Table 1). Moreover, donor cells infected at an MOI of 3 produced less IL-6 than those infected at an MOI of 1 (Table 1), consistent with active suppression of IL-6 expression by HCMV infection; however, there was little difference in the amount of IL-6 produced by the respective recipient cultures (Table 1, recipient cultures 1 and 2). These results indicate that the suppression of IL-6 secretion is not mediated by a soluble factor released into the culture medium but is exerted within the infected cell, possibly at the transcriptional, posttranscriptional, or translational level.
HCMV activates the IL-6 promoter through core elements. HCMV infection up-regulates the steady-state levels of IL-6 mRNA (Fig. 1B). The human IL-6 promoter is recognized to contain multiple positive regulatory elements and at least one negative regulatory element, the RBP site (summarized schematically in Fig. 4A). To further delineate the mechanism by which HCMV up-regulates IL-6 expression, a series of reporter plasmids containing truncated versions of the IL-6 promoter were employed in transient transfection experiments in permissive U373-MG cells. Productive HCMV infection resulted in strong activation of the IL-6 promoter at 16 hpi that was dependent only on the core promoter; a construct lacking all regulatory elements upstream of the major TATA motif gave the same level of expression as the longest promoter construct tested (Fig. 4B). Furthermore, challenge with UV-irradiated virus resulted only in weak, albeit detectable, activation of IL-6 promoter constructs; this modest response was essentially abolished by deletion of sequences between nucleotides –110 and –50 encoding the NF-B and C/EBP binding sites. Control experiments demonstrated that IL-1? activated the expression of plasmids containing at least 110 nucleotides of promoter sequence upstream of the transcription start site, whereas there was no effect on the shortest construct (–50), which lacked all upstream regulatory elements (data not shown). Transfection experiments were repeated with fibroblasts at both 16 h and 40 hpi to emulate the conditions used to monitor IL-6 protein and mRNA levels (Fig. 1). Although the transfection efficiency was much reduced relative to that in U373-MG cells (as indicated by lower basal and activated luciferase levels), both full-length and core promoter constructs were again strongly activated in HCMV-infected fibroblasts at 16 hpi, whereas there was little or no detectable activation by UV-irradiated virus (Fig. 4C). Moreover, transcriptional activity was sustained, and in fact increased, at 40 hpi.
The ability of HCMV to activate through core promoter elements has also been observed for the c-myc, c-fos, hsp70, and TGF-?2 cellular promoters and the human immunodeficiency virus type 1 long terminal repeat (9, 25, 59). In these studies, activation of the core promoter was shown to be mediated by the HCMV immediate-early proteins. Furthermore, the increased promoter activation observed in HCMV-infected fibroblasts at 40 hpi compared to that at 16 hpi (Fig. 4C) correlated with accumulation of the major immediate-early proteins IE1p72 and IE2p86 (data not shown). We therefore tested whether HCMV IE1p72 or IE2p86 could activate the IL-6 promoter in permissive cells. A plasmid encoding IE2p86 strongly up-regulated the IL-6 promoter in U373-MG cells, and this was independent of upstream regulatory elements (Fig. 4D). Thus, activation by both HCMV and IE2p86 shared a requirement only for the core IL-6 promoter, suggesting that IE2p86 may be the dominant activator of the IL-6 promoter during productive infection. Comparable results were also obtained from experiments performed with primary fibroblasts and nonpermissive CV-1 cells (data not shown).
In contrast to the strong activation by IE2p86, the effects of IE1p72 were weak, variable, and possibly cell type specific. In U373-MG, the weak activation induced by IE1p72 was relatively independent of upstream promoter elements (Fig. 4E). In CV-1 cells, however, while IE1p72 was able to up-regulate the longest IL-6 promoter construct to an extent similar to that seen in U373-MG, there was no significant or reproducible effect on any of the truncated promoter constructs (data not shown). Consistent with this, there was some evidence that IE1p72 could augment, weakly, IE2p86-mediated activation of the IL-6 core promoter in U373-MG (Fig. 4F) but not CV-1 cells (data not shown).
Transcriptional activation of the endogenous IL-6 gene in productive infection. The strong activation of reporter gene expression from the IL-6 promoter in transfection assays by productive HCMV infection or IE2p86 alone (Fig. 4) would appear superficially to be inconsistent with the observation that lytic gene expression suppresses steady-state levels of IL-6 mRNA and IL-6 secretion (Fig. 1). To address this, nuclear run-on assays were performed to examine the level of ongoing transcription from the endogenous IL-6 gene at 16 hpi in fibroblasts; the RT-PCR-based method utilized distinguishes between spliced and unspliced transcripts using different primer sets (Fig. 5A). A low level of spliced but not unspliced IL-6 mRNA was detected in the nuclei of mock-infected cells, and its abundance did not increase in the run-on assay (i.e., in the presence of NTPs); thus, there was no significant ongoing IL-6 transcription in these cells. There was also little or no ongoing transcription from the ?-actin gene during the run-on assay in any of the samples; this was expected since uninfected cells were quiescent and HCMV infection induces only a modest increase in ?-actin mRNA (25). However, there was a significant increase in the levels of both spliced and unspliced IL-6 transcripts in the nuclei of productively infected cells during the run-on assay (Fig. 5B to D), whereas there was little or no increase in transcript levels in nuclei from cells challenged with UV-inactivated virus. These effects were reproducible, and data from a second experiment are combined in Fig. 5E. Thus, both DNA transfection experiments (Fig. 4) and nuclear run-on assays are consistent with the ability of productive HCMV infection, but not UV-inactivated virus challenge, to induce significant up-regulation of transcription from the IL-6 promoter at 16 hpi.
The onset of viral gene expression results in increased turnover of IL-6 mRNA. The above data suggest that transcription from the IL-6 promoter is activated first transiently by a virion factor (either directly or by the activation of signal transduction cascades leading to activation of a cellular transcription factor), then subsequently by one or more HCMV transcriptional activators expressed de novo, likely including IE2p86. However, IL-6 mRNA levels were lower in productively infected cells than in cells challenged with UV-inactivated virus (Fig. 1B), implying that the rate of IL-6 mRNA turnover was enhanced during productive infection. To test this hypothesis, ongoing transcription was blocked with actinomycin D at 16 hpi, and cells were analyzed to determine the level of IL-6 mRNA remaining at intervals following the block. In cells harvested immediately after actinomycin D treatment, the relative level of IL-6 mRNA was significantly higher in cells inoculated with UV-inactivated virus than in those infected with HCMV (Fig. 6A), as seen for steady-state levels of mRNA. In productively infected cells the level of IL-6 mRNA declined rapidly and was barely detectable 8 h after the block; by contrast, in the abortive infection, IL-6 mRNA persisted at around 70% of the original level even 8 h after the block (Fig. 6B). Decay appeared to show first-order exponential kinetics (Fig. 6C), giving a half-life for IL-6 mRNA during productive HCMV infection of approximately 77 min, compared with approximately 13 h for that in cells challenged with UV-inactivated virus. The long half-life of IL-6 mRNA in cells challenged with inactivated virus is consistent with our previous data (Fig. 1B), which showed that the message remained abundant in cells at 40 hpi even though there was little or no ongoing transcription after 16 hpi (Fig. 4 and 5).
The stability of IL-6 mRNA is regulated at least in part by AREs in the 3' UTR. Interestingly, COX-2 mRNA, which is also subject to ARE-mediated regulation (15, 36, 46, 53), was extremely stable in cells challenged with UV-inactivated virus over an 8 h period but underwent significant degradation in productively infected cells over the same period; however, the kinetics of COX-2 mRNA decay in HCMV-infected cells appeared to be more complex than that for IL-6 (Fig. 6B). In contrast, mRNA for ?-actin, which is not subject to ARE-mediated turnover, remained stable in both productively and abortively infected cells during the experiment (Fig. 6A). These results suggest that HCMV may potentially encode a mechanism to preferentially accelerate the turnover of ARE-containing mRNAs.
HCMV infection induces relocalization of a regulator of ARE-containing mRNAs. The mechanism by which AREs promote mRNA turnover has not yet been fully elucidated; however, it has been reported that proteasome activity is required for ARE-mediated turnover of mRNA (32). Therefore, we tested the effect of the proteasome inhibitor MG132 on IL-6 expression by HCMV-infected cells by inoculating cells with either HCMV or UV-inactivated virus and adding the inhibitor at various time points (0, 2, or 4 hpi). In addition, control cultures were treated with IL-1?. Cells and culture medium harvested at 16 hpi were then assayed for IL-6 and IE gene expression. Treatment with MG132 resulted in strong inhibition of IL-6 expression when applied at 0 or 2 h postchallenge, and to a lesser extent when applied at 4 h, and this inhibition was observed with all stimuli (Fig. 7A). Furthermore, MG132 inhibited the expression of IE proteins in cells challenged with active virus (Fig. 7C), and as for IL-6 expression, this effect was strongest when MG132 was applied at the same time as the viral inoculum. Similar effects were observed when cultures were grown in the presence of serum, though as expected, the basal and induced levels of IL-6 expression were much higher (Fig. 7B). The effects of MG132 are discussed below; however, since viral gene expression was required for suppression of IL-6 but MG132 also inhibited the expression of viral proteins, it was not possible to use this approach directly to determine whether the increased turnover of IL-6 mRNA in productively infected cells was dependent on proteasome activity.
As an alternative approach, we asked whether productive HCMV infection could alter the level or activity of regulators of ARE-mediated mRNA turnover. The best-characterized positive regulator of ARE-mediated mRNA stability is HuR, a member of the Elav family of RNA-binding proteins. HuR stabilizes mRNAs as a result of direct binding to ARE motifs (4), including those found in IL-6 and COX-2 mRNAs (38, 46); overexpression of HuR stabilizes ARE-containing mRNAs in vivo (21, 43, 46), whereas inhibition of HuR activity by siRNA has the opposite effect (46, 53). However, in HCMV-infected cells the total level of HuR protein was increased compared to that in either mock-infected cells or those challenged with UV-inactivated virus (Fig. 8A). This increase was evident by 24 hpi and became more pronounced at later times of infection. As observed for the phenomenon of IL-6 suppression, induction of HuR by active HCMV infection was also evident in cultures grown in 10% serum (Fig. 8A, right). The level of HuR in cells treated with IL-1? was similar to that seen in mock-infected cells and those challenged with inactivated virus (data not shown), indicating that an increase in the overall level of HuR was not required for high levels of IL-6 expression.
Although overexpression of HuR has previously been correlated with stabilization of ARE-containing mRNAs, we observed that the half-lives of IL-6 and COX-2 mRNA were much reduced in HCMV-infected cells despite elevated levels of HuR (Fig. 8A). Two forms of HuR were apparent in Western transfer experiments, and while these may possibly represent different phosphorylation states of the protein, HCMV infection did not significantly alter the relative proportion of these forms. However, HCMV infection had a profound effect on the distribution of HuR within intracellular compartments. Consistent with published studies, HuR was found predominantly in the nuclei of mock-infected cells (21, 38) and in cells challenged with IL-1? or UV-inactivated HCMV (Fig. 8B); nuclear localization of HuR is therefore compatible with both the lack of IL-6 expression from quiescent cells and induced expression. Cells undergoing productive HCMV infection were identified by staining for the viral DNA polymerase accessory protein pUL44; in such cells there was significant cytoplasmic staining of HuR (Fig. 8B, panels i), whereas uninfected cells in the same sample retained HuR in the nucleus (Fig. 8B, panels i). The cytoplasmic localization of HuR was more pronounced at a higher multiplicity of infection (Fig. 8B, panels ii), with some cells appearing to exclude HuR from the nucleus (Fig. 8B, panels ii). Cytoplasmic staining was not due to cross-reactivity with viral proteins, as staining using an isotype-matched primary antibody control did not show cytoplasmic Texas Red fluorescence in HCMV-positive cells at a multiplicity of either 1 (Fig. 8B, panels vi) or 3 (data not shown). (Modest cross-reactivity with a component of viral replication centers could be seen in infected cells; this appeared to be due to binding by the secondary antibody, as independent experiments using a different FITC-conjugated antibody with the same primary antibodies showed no such cross-reactivity [data not shown]).
HuR redistribution is independent of HCMV ?2.7. Since suppression of IL-6 was not mediated by a soluble factor in the culture medium, precluding a role for cmvIL-10 (Fig. 3), we asked what other viral gene product(s) might be responsible. Interestingly, inspection of the sequence of the HCMV ?2.7 RNA transcript, which is abundantly expressed from the duplicated IRL4/TRL4 genes of HCMV strain AD169 at early times of infection (50), reveals a number of AREs and ARE-like motifs in the 3' region of the molecule. The function of the ?2.7 gene is not yet known; moreover, although the overall gene sequence is strongly conserved in other strains of HCMV, the open reading frame present in the AD169 ?2.7 transcript is not (33), suggesting that the RNA itself may be the functional gene product. This led us to speculate that ?2.7 RNA might act as a decoy or competitor for HuR, preventing HuR from binding to and stabilizing IL-6 mRNA. To test this possibility, cells were infected with wild-type HCMV Toledo (which carries a single copy of the ?2.7 RNA gene), a mutant lacking the ?2.7 RNA gene, and a revertant strain. We anticipated that if the ?2.7 RNA was responsible for inhibition of HuR activity and/or destabilization of IL-6 mRNA, infection with a mutant lacking this gene would express IL-6 at greatly increased levels, due to the high level of ongoing transcription in productively infected cells (Fig. 4 and 5). However, the level of IL-6 produced by infected cells was essentially the same for all three viruses (Fig. 8C) and was strongly suppressed in comparison to that in cultures challenged with UV-inactivated viruses. As observed previously (Fig. 1B), the level of IL-6 mRNA correlated with the level of IL-6 protein in the medium (Fig. 8D); furthermore, all three virus strains showed similar ability to induce HuR during productive, but not abortive, infection (Fig. 8E), as observed for wild-type strain Towne. Similar experiments were carried out using wild-type strain AD169 and mutants lacking one or both copies of the ?2.7 RNA gene; as seen with viruses in the Toledo background, these showed no differences in their ability to suppress IL-6 expression or induce HuR during active infection (data not shown). We conclude that while both attenuated and low-passage strains of HCMV are able to suppress IL-6 expression at the posttranscriptional level and induce the overexpression of HuR, the ?2.7 RNA is not required for these phenomena. Moreover, since the Towne and Toledo strains of HCMV exhibited similar effects on IL-6 and HuR expression, we conclude that such effects are not exerted by the products of the additional genes present in the Toledo strain.
DISCUSSION
The binding and fusion of HCMV virions with the cell surface, combined with the delivery of virion components and tegument proteins in particular, to the infected cell is known to initiate the activation of multiple signal transduction pathways and the induction of inflammatory and antiviral gene expression. Consistent with these observations, challenge with UV-inactivated HCMV particles is sufficient to elicit a strong induction of IL-6 expression (8). Transcriptional activation of the IL-6 promoter by virion particles appears to be transient and mediated largely through the NF-B pathway; deletion of the sequence containing the NF-B binding site abolished the ability of UV-irradiated virus to activate the promoter (Fig. 4B). Moreover, Carlquist et al. reported that an antagonist of NF-B inhibited IL-6 expression in response to challenge with inactivated HCMV, while we observed that MG132, which inhibits NF-B activation by preventing proteasomal degradation of IB, also strongly reduced IL-6 expression in response to both IL-1? and viral challenge (Fig. 7A and B). Interestingly, we observed that MG132 also markedly inhibited the expression of HCMV IE genes (Fig. 7C). Since the HCMV major IE enhancer contains multiple NF-B sites, it is possible that reduced IE expression is a direct consequence of inhibition of this transcription factor. Indeed, a recent report has shown that both MG132 and aspirin inhibit IE gene expression in HCMV-infected fibroblasts, while inhibition of the NF-B activation pathway decreases expression from the major IE promoter (16). However, our experiments do not rule out the possibility that the onset of viral gene expression requires other events which are dependent on proteasomal activity, and such putative events are not addressed by our work.
The effect of UV-inactivated HCMV on IL-6 expression resembled that resulting from IL-1 challenge. While this inflammatory cytokine up-regulates transcription from the IL-6 promoter, the effect is only modest, and the large increase in IL-6 protein expression that occurs in response to IL-1 is predominantly due to extensive stabilization of the IL-6 mRNA (45). Similarly, we observed that UV-inactivated HCMV was only a weak activator of the IL-6 promoter, but the half-life of IL-6 mRNA was greatly prolonged in abortively infected cells, allowing these cells to maintain synthesis of IL-6 protein over an extended period. The similar effect of IL-1 and HCMV virions is consistent with the hypothesis that virion interaction with the host cell induces an inflammatory or antiviral response.
In contrast to the effects of inactivated virions, the onset of HCMV gene expression appears to curtail IL-6 expression by the productively infected cell (Fig. 1). Paradoxically, transcription initiation from the IL-6 promoter was up-regulated in productively infected cells; however, the reduced level of IL-6 mRNA and protein can be accounted for by the greatly increased turnover of IL-6 mRNA in productively infected cells compared to those challenged with inactivated virions. Additionally, we found no evidence for accumulation or retention of IL-6 protein within productively infected cells (data not shown), which might otherwise account for lower levels in the culture medium. Thus, whereas inactivated virions induce an inflammatory state characterized by induction and stabilization of the IL-6 mRNA, the onset of HCMV lytic gene expression counteracts this effect by accelerating the turnover of IL-6 mRNA.
It is well established that the stability of IL-6 mRNA is regulated extensively through elements in the 3' UTR and in particular AREs. Interestingly, the turnover of another ARE-containing mRNA, that of COX-2, was also accelerated in productively infected cells compared with those challenged with inactivated virions (Fig. 6). Moreover, this accelerated decay coincided with aberrant localization of HuR, an ARE-binding protein which binds to both IL-6 and COX-2 mRNAs and stabilizes ARE-containing mRNAs in the cell. The increased turnover of ARE-containing mRNAs occurred despite an elevated level of HuR protein in the HCMV-infected cell, whereas studies in uninfected cells have shown that HuR overexpression results in stabilization of such mRNAs. Conversely, inhibition of HuR expression by siRNA leads to decreased expression of COX-2 (46, 53), suggesting that the increased turnover of IL-6 and COX-2 mRNAs that we observe during productive HCMV infection might be due to inhibition of the mRNA-stabilizing activity of HuR. Furthermore, we found no change during HCMV infection in the expression of BRF-1 (data not shown), an ARE-binding protein which destabilizes ARE-containing mRNAs and thus acts in a manner antagonistic to HuR (52).
It is generally accepted that HuR functions by binding to target mRNAs in the nucleus and protecting them against degradation prior to translation; thus, displacement of HuR from its normal targets would promote turnover of these mRNAs. One possible mechanism by which HCMV might achieve this is by expression of a decoy RNA. Interestingly, the small nuclear RNAs of herpesvirus saimiri have been shown to bind HuR in vitro (37) and in vivo (14); however, their expression does not appear to affect the levels of cellular mRNAs (14). Moreover, deletion of the HCMV gene encoding the ?2.7 RNA, a candidate decoy which contains multiple AREs and ARE-like motifs near the 3' end, had no effect on the ability of active HCMV infection to suppress IL-6 expression. An alternative mechanism for HuR displacement might be by interaction with a viral protein, which interferes with mRNA binding and/or the subcellular localization of HuR; this might explain the aberrant localization of HuR we observed in productively infected cells. Notably, the increase in total HuR levels seen during infection is reminiscent of the overexpression of p53 that occurs when this protein is inactivated either by mutation or by interaction with viral oncoproteins. Whether or not HuR is engaged in novel interactions in the infected cell is clearly an intriguing subject for further research.
The database of ARE-containing mRNAs (http://rc.kfshrc.edu.sa/ared) contains over 900 entries (2); therefore, if HCMV infection were to result in inhibition of HuR activity, the effect on cellular gene expression might be expected to be widespread. Interestingly, of the mRNAs identified by Browne et al. (6) as being more abundant in cells challenged with UV-inactivated virus than in those undergoing productive infection, around 30% are found in the ARE database. Moreover, the supplementary data of Browne et al. (available at http://www.molbio.princeton.edu/labs/shenk/browneetal2001/) shows that both IL-6 and COX-2 mRNAs are more abundant in cells challenged with inactivated virus, consistent with our observations. HCMV may therefore have evolved a mechanism for the suppression of inflammatory gene expression that is mediated through the AREs of target mRNAs, to complement other mechanisms that inhibit interferon-mediated induction of gene expression (5). The susceptibility of different ARE-containing mRNAs to such a mechanism would be expected to vary widely, however. AREs are heterologous in sequence and number (2, 54) and vary in their responsiveness to different regulatory pathways. For example, a global study of the stability of ARE-containing mRNAs in monocytes showed that inhibition of p38 mitogen-associated protein kinase activity greatly reduced the half-life of COX-2 mRNA but had little effect on the turnover of IL-6 mRNA (23). Since HCMV infection is known to up-regulate p38 activity, this may explain the different kinetics of IL-6 and COX-2 mRNA decay in HCMV-infected cells following actinomycin D treatment (Fig. 6).
A final question posed by our data is, why does HCMV both activate IL-6 expression at the level of transcription initiation and suppress expression at the posttranscriptional level? Given the wide spectrum of biological effects of IL-6, it is conceivable that under certain circumstances the expression of IL-6 could be favorable for viral replication or persistence and that continued transcription might allow the infected cell to produce IL-6 protein under conditions that stabilize the IL-6 mRNA and thus counteract suppression. For example, IL-6 antagonizes TGF-?-induced apoptosis in hepatocytes (11) and might potentially play a role as an antiapoptotic factor during HCMV infection of this particular cell type. However, it seems more likely that suppression of IL-6 expression plays a role in immune evasion and the ability of HCMV to persist in the infected host, particularly if the suppressive phenomenon we observe for IL-6 forms part of a wider mechanism for down-regulation of inflammatory and immunoregulatory genes. Moreover, our preliminary data indicate that hepatocytes, which are key mediators of the acute-phase response, produce lower amounts of IL-6 during productive HCMV infection than cells challenged with inactivated virus (data not shown). Viewed in this context, transcriptional activation of the IL-6 promoter by HCMV would appear to be an unwanted activity. However, activation appears to be largely mediated by the essential IE2p86 protein; it may be that the molecular mechanisms by which IE2p86 up-regulates the IL-6 promoter are also required for activation of HCMV promoters, and therefore such functions would be indispensable for viral replication, forcing the virus to evolve distinct mechanisms to counteract such unwanted side effects.
In summary, our results demonstrate that expression of IL-6, and potentially other inflammatory genes, is subject to suppression at the posttranscriptional level by a virally encoded factor that is expressed at IE or early times of infection. This factor remains unidentified, but our results rule out cmvIL-10, the ?2.7 RNA, and the products of genes that are deleted in the attenuated Towne strain of HCMV, suggesting that the factor responsible plays an important role in viral replication both in vivo and in vitro. Accelerated turnover of IL-6 mRNA coincides with relocalization of HuR, suggesting that HCMV may target HuR activity as part of a mechanism for down-regulation of ARE-containing mRNAs, thus overcoming the antiviral state induced in the cell by virion binding and entry. Interestingly, recent reports indicate that the vhs gene product of herpes simplex virus type 1 preferentially targets ARE-containing mRNAs for degradation (19) and that herpes simplex virus type 1 infection results in the cytoplasmic accumulation of cellular proteins that promote the turnover of AU-rich mRNAs (20). Thus, selective destabilization of ARE-containing mRNAs by various mechanisms may be a common theme in the strategies employed by herpesviruses to disrupt cellular gene expression and enable persistence in the immune-competent host.
ACKNOWLEDGMENTS
This work was supported by a Career Establishment grant (G9801388) from the UK Medical Research Council to R.C. and a Cooperative Group Grant (G9827961) from the MRC to R.C., G.W., and other members of the Cooperative Group in Pathogenesis of Persistent Human Virus Infections.
We are grateful to Richard Greaves, Guy Haegeman, and John Sinclair for plasmid and antibody reagents, to Simon Jones and Suzanne Hurst for unpublished data from IL-6R assays, to Sin Llewellyn-Lacey and the late Vicky Dickinson for technical assistance, and to Simon Jones and Mark Stinski for helpful discussions.
REFERENCES
Alcami, A., and U. H. Koszinowski. 2000. Viral mechanisms of immune evasion. Trends Microbiol. 8:410-418.
Bakheet, T., B. R. G. Williams, and K. S. A. Khabar. 2003. ARED 2.0: an update of AU-rich element mRNA database. Nucleic Acids Res. 31:421-423.
Boyle, K. A., R. L. Pietropaolo, and T. Compton. 1999. Engagement of the cellular receptor for glycoprotein B of human cytomegalovirus activates the interferon-responsive pathway. Mol. Cell. Biol. 19:3607-3613.
Brennan, C. M., and J. A. Steitz. 2001. HuR and mRNA stability. Cell. Mol. Life Sci. 58:266-277.
Browne, E. P., and T. Shenk. 2003. Human cytomegalovirus UL83-coded pp65 virion protein inhibits antiviral gene expression in infected cells. Proc. Natl. Acad. Sci. USA 100:11439-11444.
Browne, E. P., B. Wing, D. Coleman, and T. Shenk. 2001. Altered cellular mRNA levels in human cytomegalovirus-infected fibroblasts: viral block to the accumulation of antiviral mRNAs. J. Virol. 75:12319-12330.
Bryant, L. A., P. Mixon, M. Davidson, A. J. Bannister, T. Kouzarides, and J. H. Sinclair. 2000. The human cytomegalovirus 86-kilodalton major immediate-early protein interacts physically and functionally with histone acetyltransferase P/CAF. J. Virol. 74:7230-7237.
Carlquist, J. F., L. Edelman, D. W. Bennion, and J. L. Anderson. 1999. Cytomegalovirus induction of interleukin-6 in lung fibroblasts occurs independently of active infection and involves a G protein and the transcription factor, NF-B. J. Infect. Dis. 179:1094-1100.
Caswell, R. C., C. Hagemeier, G. P. Hayhurst, T. Kouzarides, and J. H. Sinclair. 1993. Mechanisms of transactivation of cellular promoters by the HCMV major immediate early proteins involves direct interaction with cellular transcription factors, p. 13-16. In S. Michelson and S. A. Plotkin (ed.), Multidisciplinary approaches to understanding cytomegalovirus disease. Elsevier, Amsterdam, The Netherlands.
Cha, T. A., E. Tom, G. W. Kemble, G. M. Duke, E. S. Mocarski, and R. R. Spaete. 1996. Human cytomegalovirus clinical isolates carry at least 19 genes not found in laboratory strains. J. Virol. 70:78-83.
Chen, R. H., M. C. Chang, Y. H. Su, Y. T. Tsai, and M. L. Kuo. 1999. Interleukin-6 inhibits transforming growth factor-beta-induced apoptosis through the phosphatidylinositol 3-kinase/Akt and signal transducers and activators of transcription 3 pathways. J. Biol. Chem. 274:23013-23019.
Chomarat, P., J. Banchereau, J. Davoust, and A. K. Palucka. 2000. IL-6 switches the differentiation of monocytes from dendritic cells to macrophages. Nat. Immunol. 1:510-514.
Compton, T., E. A. Kurt-Jones, K. W. Boehme, J. Belko, E. Latz, D. T. Golenbock, and R. W. Finberg. 2003. Human cytomegalovirus activates inflammatory cytokine responses via CD14 and Toll-like receptor 2. J. Virol. 77:4588-4596.
Cook, H. L., H. E. Mischo, and J. A. Steitz. 2004. The Herpesvirus saimiri small nuclear RNAs recruit AU-rich element-binding proteins but do not alter host AU-rich element-containing mRNA levels in virally transformed T cells. Mol. Cell. Biol. 24:4522-4533.
Dean, J. L. E., S. J. Sarsfield, E. Tsounakou, and J. Saklatvala. 2003. p38 mitogen-activated protein kinase stabilizes mRNAs that contain cyclooxygenase-2 and tumor necrosis factor AU-rich elements by inhibiting deadenylation. J. Biol. Chem. 278:39470-39476.
DeMeritt, I. B., L. E. Milford, and A. D. Yurochko. 2004. Activation of the NF-B pathway in human cytomegalovirus-infected cells is necessary for efficient transactivation of the major immediate-early promoter. J. Virol. 78:4498-4507.
Dolan, A., C. Cunningham, R. D. Hector, A. F. Hassan-Walker, L. Lee, C. Addison, D. J. Dargan, D. J. McGeoch, D. Gatherer, V. C. Emery, P. D. Griffiths, C. Sinzger, B. P. McSharry, G. W. G. Wilkinson, and A. J. Davison. 2004. Genetic content of wild-type human cytomegalovirus. J. Gen. Virol. 85:1301-1312.
Duncan, S. R., I. L. Paradis, S. A. Yousem, S. L. Similo, W. F. Grgurich, P. A. Williams, J. H. Dauber, and B. P. Griffith. 1992. Sequelae of cytomegalovirus pulmonary infections in lung allograft recipients. Am. Rev. Respir. Dis. 146:1419-1425.
Esclatine, A., B. Taddeo, L. Evans, and B. Roizman. 2004. The herpes simplex virus 1 UL41 gene-dependent destabilization of cellular RNAs is selective and may be sequence-specific. Proc. Natl. Acad. Sci. USA 101:3603-3608.
Esclatine, A., B. Taddeo, and B. Roizman. 2004. Herpes simplex virus 1 induces cytoplasmic accumulation of TIA-1/TIAR and both synthesis and cytoplasmic accumulation of tristetraprolin, two cellular proteins that bind and destabilize AU-rich RNAs. J. Virol. 78:8582-8592.
Fan, X. C., and J. A. Steitz. 1998. Overexpression of HuR, a nuclear-cytoplasmic shuttling protein, increases the in vivo stability of ARE-containing mRNAs. EMBO J. 17:3448-3460.
Fortunato, E. A., A. K. McElroy, I. Sanchez, and D. H. Spector. 2000. Exploitation of cellular signaling and regulatory pathways by human cytomegalovirus. Trends Microbiol. 8:111-119.
Frevel, M. A., T. Bakheet, A. M. Silva, J. G. Hissong, K. S. Khabar, and B. R. Williams. 2003. p38 mitogen-activated protein kinase-dependent and -independent signaling of mRNA stability of AU-rich element-containing transcripts. Mol. Cell. Biol. 23:425-436.
Geist, L. J., and L. Y. Dai. 1996. Cytomegalovirus modulates interleukin-6 gene expression. Transplantation 62:653-658.
Hagemeier, C., S. M. Walker, P. J. Sissons, and J. H. Sinclair. 1992. The 72K IE1 and 80K IE2 proteins of human cytomegalovirus independently trans-activate the c-fos, c-myc and hsp70 promoters via basal promoter elements. J. Gen. Virol. 73:2385-2393.
Hildebrandt, A. L., and P. D. Neufer. 2000. Exercise attenuates the fasting-induced transcriptional activation of metabolic genes in skeletal muscle. Am. J. Physiol. Endocrinol. Metab. 278:E1078-E1086.
Hirsch, A. J., and T. Shenk. 1999. Human cytomegalovirus inhibits transcription of the CC chemokine MCP-1 gene. J. Virol. 73:404-410.
Humbert, M., R. M. Delattre, S. Fattal, B. Rain, J. Cerrina, P. Dartevelle, G. Simonneau, P. Duroux, P. Galanaud, and D. Emilie. 1993. In situ production of interleukin-6 within human lung allografts displaying rejection or cytomegalovirus pneumonia. Transplantation 56:623-627.
Iyer, V. R., M. B. Eisen, D. T. Ross, G. Schuler, T. Moore, J. C. Lee, J. M. Trent, L. M. Staudt, J. Hudson, Jr., M. S. Boguski, D. Lashkari, D. Shalon, D. Botstein, and P. O. Brown. 1999. The transcriptional program in the response of human fibroblasts to serum. Science 283:83-87.
Jones, S. A., and S. Rose-John. 2002. The role of soluble receptors in cytokine biology: the agonistic properties of the sIL-6R/IL-6 complex. Biochim. Biophys. Acta 1592:251-263.
Kotenko, S. V., S. Saccani, L. S. Izotova, O. V. Mirochnitchenko, and S. Pestka. 2000. Human cytomegalovirus harbors its own unique IL-10 homolog (cmvIL-10). Proc. Natl. Acad. Sci. USA 97:1695-1700.
Laroia, G., B. Sarkar, and R. J. Schneider. 2002. Ubiquitin-dependent mechanism regulates rapid turnover of AU-rich cytokine mRNAs. Proc. Natl. Acad. Sci. USA 99:1842-1846.
McSharry, B. P., P. Tomasec, M. L. Neale, and G. W. G. Wilkinson. 2003. The most abundantly transcribed human cytomegalovirus gene (? 2.7) is non-essential for growth in vitro. J. Gen. Virol. 84:2511-2516.
Michelson, S., J. Alcami, S.-J. Kim, D. Danielpour, F. Bachelerie, L. Picard, C. Bessia, C. Paya, and J.-L. Virelizier. 1994. Human cytomegalovirus infection induces transcription and secretion of transforming growth factor ?1. J. Virol. 68:5730-5737.
Mocarski, E. S., Jr. 2002. Immunomodulation by cytomegaloviruses: manipulative strategies beyond evasion. Trends Microbiol. 10:332-339.
Mukhopadhyay, D., C. W. Houchen, S. Kennedy, B. K. Dieckgraefe, and S. Anant. 2003. Coupled mRNA stabilization and translational silencing of cyclooxygenase-2 by a novel RNA binding protein, CUGBP2. Mol. Cell 11:113-126.
Myer, V., S. Lee, and J. Steitz. 1992. Viral small nuclear ribonucleoproteins bind a protein implicated in messenger RNA destabilization. Proc. Natl. Acad. Sci. USA 89:1296-1300.
Nabors, L. B., G. Y. Gillespie, L. Harkins, and P. H. King. 2001. HuR, a RNA stability factor, is expressed in malignant brain tumors and binds to adenine- and uridine-rich elements within the 3' untranslated regions of cytokine and angiogenic factor mRNAs. Cancer Res. 61:2154-2161.
Ng, S., Y. Tan, and G. Guy. 1994. Differential induction of the interleukin-6 gene by tumor necrosis factor and interleukin-1. J. Biol. Chem. 269:19021-19027.
Pasare, C., and R. Medzhitov. 2003. Toll pathway-dependent blockade of CD4+CD25+ T cell-mediated suppression by dendritic cells. Science 299:1033-1036.
Pass, R. F. 2001. Cytomegalovirus, p. 2675-2705. In D. M. Knipe and P. M. Howley (ed.), Fields virology, 4th ed., vol. 2. Lippincott Williams & Wilkins, Philadelphia, Pa.
Penfold, M. E., D. J. Dairaghi, G. M. Duke, N. Saederup, E. S. Mocarski, G. W. Kemble, and T. J. Schall. 1999. Cytomegalovirus encodes a potent alpha chemokine. Proc. Natl. Acad. Sci. USA 96:9839-9844.
Pioli, P. A., B. J. Hamilton, J. E. Connolly, G. Brewer, and W. F. Rigby. 2002. Lactate dehydrogenase is an AU-rich element-binding protein that directly interacts with AUF1. J. Biol. Chem. 277:35738-35745.
Ploegh, H. L. 1998. Viral strategies of immune evasion. Science 280:248-253.
Richards, C. D. 1998. Interleukin-6, p. 87-108. In A. Mire-Sluis and R. Thorpe (ed.), Cytokines. Academic Press, San Diego, Calif.
Sengupta, S., B.-C. Jang, M.-T. Wu, J.-H. Paik, H. Furneaux, and T. Hla. 2003. The RNA-binding protein HuR regulates the expression of cyclooxygenase-2. J. Biol. Chem. 278:25227-25233.
Simmen, K. A., J. Singh, B. G. Luukkonen, M. Lopper, A. Bittner, N. E. Miller, M. R. Jackson, T. Compton, and K. Fruh. 2001. Global modulation of cellular transcription by human cytomegalovirus is initiated by viral glycoprotein B. Proc. Natl. Acad. Sci. USA 98:7140-7145.
Simmons, P., K. Kaushansky, and B. Torok-Storb. 1990. Mechanisms of cytomegalovirus-mediated myelosuppression: perturbation of stromal cell function versus direct infection of myeloid cells. Proc. Natl. Acad. Sci. USA 87:1386-1390.
Sinclair, J., and P. Sissons. 1996. Latent and persistent infections of monocytes and macrophages. Intervirology 39:293-301.
Spector, D. H. 1996. Activation and regulation of human cytomegalovirus early genes. Intervirology 39:361-377.
Spencer, J. V., K. M. Lockridge, P. A. Barry, G. Lin, M. Tsang, M. E. Penfold, and T. J. Schall. 2002. Potent immunosuppressive activities of cytomegalovirus-encoded interleukin-10. J. Virol. 76:1285-1292.
Stoecklin, G., M. Colombi, I. Raineri, S. Leuenberger, M. Mallaun, M. Schmidlin, B. Gross, M. Lu, T. Kitamura, and C. Moroni. 2002. Functional cloning of BRF1, a regulator of ARE-dependent mRNA turnover. EMBO J. 21:4709-4718.
Subbaramaiah, K., T. P. Marmo, D. A. Dixon, and A. J. Dannenberg. 2003. Regulation of cyclooxgenase-2 mRNA stability by taxanes: evidence for involvement of p38, MAPKAPK-2, and HuR. J. Biol. Chem. 278:37637-37647.
Tebo, J., S. Der, M. Frevel, K. S. A. Khabar, B. R. G. Williams, and T. A. Hamilton. 2003. Heterogeneity in control of mRNA stability by AU-rich elements. J. Biol. Chem. 278:12085-12093.
Torok-Storb, B., M. Boeckh, C. Hoy, W. Leisenring, D. Myerson, and T. Gooley. 1997. Association of specific cytomegalovirus genotypes with death from myelosuppression after marrow transplantation. Blood 90:2097-2102.
Torok-Storb, B., L. Bolles, M. Iwata, K. Doney, G. E. Sale, T. A. Gooley, and R. Storb. 2001. Increased prevalence of CMV gB3 in marrow of patients with aplastic anemia. Blood 98:891-892.
Vanden Berghe, W., S. Plaisance, E. Boone, K. De Bosscher, M. L. Schmitz, W. Fiers, and G. Haegeman. 1998. p38 and extracellular signal-regulated kinase mitogen-activated protein kinase pathways are required for nuclear factor-B p65 transactivation mediated by tumor necrosis factor. J. Biol. Chem. 273:3285-3290.
Vink, C., M. J. Smit, R. Leurs, and C. A. Bruggeman. 2001. The role of cytomegalovirus-encoded homologs of G protein-coupled receptors and chemokines in manipulation of and evasion from the immune system. J. Clin. Virol. 23:43-55.
Walker, S., C. Hagemeier, J. G. Sissons, and J. H. Sinclair. 1992. A 10-base-pair element of the human immunodeficiency virus type 1 long terminal repeat (LTR) is an absolute requirement for transactivation by the human cytomegalovirus 72-kilodalton IE1 protein but can be compensated for by other LTR regions in transactivation by the 80-kilodalton IE2 protein. J. Virol. 66:1543-1550.
Wang, E. C. Y., B. McSharry, C. Retiere, P. Tomasec, S. Williams, L. K. Borysiewicz, V. M. Braud, and G. W. G. Wilkinson. 2002. UL40-mediated NK evasion during productive infection with human cytomegalovirus. Proc. Natl. Acad. Sci. USA 99:7570-7575.
Yurochko, A. D., and E. S. Huang. 1999. Human cytomegalovirus binding to human monocytes induces immunoregulatory gene expression. J. Immunol. 162:4806-4816.
Zhu, H., J. P. Cong, G. Mamtora, T. Gingeras, and T. Shenk. 1998. Cellular gene expression altered by human cytomegalovirus: global monitoring with oligonucleotide arrays. Proc. Natl. Acad. Sci. USA 95:14470-14475.
Zhu, H., J. P. Cong, and T. Shenk. 1997. Use of differential display analysis to assess the effect of human cytomegalovirus infection on the accumulation of cellular RNAs: induction of interferon-responsive RNAs. Proc. Natl. Acad. Sci. USA 94:13985-13990.
Zhu, H., J.-P. Cong, D. Yu, W. A. Bresnahan, and T. E. Shenk. 2002. Inhibition of cyclooxygenase 2 blocks human cytomegalovirus replication. Proc. Natl. Acad. Sci. USA 99:3932-3937.(Claire Gealy, Marian Dens)
Section of Infection and Immunity, University of Wales College of Medicine, Cardiff, Wales, United Kingdom
ABSTRACT
Human cytomegalovirus (HCMV) has evolved multiple strategies for suppression of the antiviral response of the infected cell. DNA array technology has revealed that HCMV clearly regulates host gene expression during the course of a productive infection by enhancing, sustaining, or suppressing steady-state levels of cellular transcripts. Interleukin-6 (IL-6) is a pleiotropic cytokine that plays a central role in the immune response to infection. Here we report a detailed study of the effects of HCMV infection on IL-6 expression by human fibroblasts. UV-inactivated virus was found to induce high levels of IL-6 mRNA and protein expression, and IL-6 mRNA remained abundant in cells 16 h after inoculation even though the level of ongoing IL-6 transcription was not significantly enhanced. In lytic HCMV infections, the onset of viral gene expression resulted in two apparently antagonistic effects on IL-6 expression: (i) transcriptional activation, mediated at least in part by the IE2p86 protein, and (ii) posttranscriptional suppression mediated by destabilization of IL-6 mRNA. Transcriptional activation was outweighed by the suppressive effect, such that cells undergoing productive infection produced less IL-6 than cells challenged with inactivated virus. Suppression of IL-6 expression was independent of the viral IL-10 homologue, cmvIL-10. Destabilization of IL-6 mRNA was observed to coincide with the enhanced expression and aberrant intracellular localization of HuR, an mRNA-binding protein known to interact with IL-6 and other mRNAs containing 3' AU-rich elements. Our data suggest a novel mechanism for gene regulation by HCMV at the posttranscriptional level.
INTRODUCTION
Human cytomegalovirus (HCMV) is a ubiquitous, clinically important herpesvirus. Following primary infection, HCMV persists throughout the lifetime of the host, during which the virus must avoid elimination by host immune mechanisms. Thus, HCMV has become a paradigm for viral immune evasion (1, 35, 44). In recent years a remarkable series of HCMV immune evasion or modulation systems have been elucidated, including the following: down-regulation of surface MHC-I expression (1), evasion of natural killer (NK) cell-mediated killing (60), and altered expression of immune signaling molecules and interferon response genes (1, 6, 47). Furthermore, these studies have shown that the virus has acquired or evolved in its genome a number of homologues of immune regulators, including chemokines and chemokine receptors (35, 58). The assumption that such molecules play a direct role in persistence and pathogenesis in the host is supported by the fact that attenuated laboratory strains (AD169 and Towne) lack a set of additional genes, associated with immunomodulatory functions, that are present in virulent clinical or low-passage isolates of HCMV (10, 42). Thus, HCMV has a direct impact on normal immune signaling, to promote an infection associated with a wide range of life-threatening conditions in the immune-compromised host (41).
Interleukin-6 (IL-6) plays a central role in both innate and acquired immune responses. IL-6 is the predominant inducer of the acute-phase response, an innate immune mechanism which is triggered by infection and inflammation (45). IL-6 also plays multiple roles during subsequent development of acquired immunity against incoming pathogens, including regulation of cytokine and chemokine gene expression, stimulation of antibody production by B cells (45), regulation of macrophage and dendritic cell differentiation (12), and response of regulatory T cells to microbial infection (40). In addition to these roles in pathogen-specific inflammation and immunity, IL-6 levels are elevated in chronic inflammatory conditions, such as rheumatoid arthritis, and indeed antibodies against IL-6 and the soluble form of its receptor have been used therapeutically for this condition (30). The inflammatory effects of IL-6 have also been implicated as a factor in the failure of organ and tissue grafts; moreover, elevated levels of the cytokine have been reported to accompany HCMV replication in transplanted lungs and bone marrow during episodes of inflammation or rejection (18, 28, 48, 55, 56). Given the ability of HCMV to persist in the infected host, it might be anticipated that the virus has evolved mechanisms to modulate the expression or signaling of IL-6 as part of the viral armory of immune evasion strategies. However, the effects of productive HCMV infection on IL-6 expression are not well understood.
Transcription of the IL-6 gene can be induced by several cellular factors in response to a range of physiological stimuli; moreover, the mRNA is subject to extensive regulation at the posttranscriptional level. Some inducers, such as IL-1, induce both transcription and stabilization of IL-6 mRNA, whereas tumor necrosis factor alpha induces transcription but not stabilization of the mRNA (39). Regulation of IL-6 mRNA stability is mediated through the 3' untranslated region (UTR), which contains a number of AU-rich elements (AREs). AREs occur in the mRNAs of many genes involved in inflammation, immune signaling, and cell growth and proliferation (4). Their presence in an mRNA generally promotes its degradation in the unstimulated cell but can confer stability in response to appropriate—typically inflammatory—stimuli such as IL-1; these effects on mRNA stability are mediated by a number of both positive (stabilizing) and negative (destabilizing) proteins in the cell, although the precise mechanism of action of ARE-binding proteins has not yet been elucidated fully.
The effects of HCMV infection on cellular gene expression are initiated with binding of the virion to the cell surface (22), with interaction between glycoprotein B and Toll-like receptor 2 playing a major role in this process (13). The binding of virions, or even glycoprotein B in isolation, leads to the induction of inflammatory cytokines and interferon-responsive genes (3, 47, 61). IL-6 mRNA expression is up-regulated by HCMV infection in the absence of de novo expression of viral genes (6, 8, 24, 62, 63), although the IL-6 promoter does not contain recognized interferon response elements. However, virion binding activates multiple intracellular signal transduction pathways, including the phosphatidylinositol kinase, mitogen-associated protein/extracellular signal-regulated kinase, and protein kinase C pathways, all of which lead to the activation of nuclear factors such as NF-B and p38 (13, 22), which are known inducers of IL-6 gene transcription (45). Since the prolonged expression of inflammatory and antiviral genes would likely be detrimental to viral replication, HCMV has evolved mechanisms to counter the host cell's innate antiviral defenses: another virion protein, the matrix protein pp65, has recently been shown to inhibit expression of many of the interferon-responsive genes that are induced by HCMV infection (5). Nevertheless, wild-type HCMV virions still induce up-regulation of a substantial number of proinflammatory genes, showing that the presence of pp65 is not sufficient to completely overcome the innate immune response. Indeed, elevated levels of prostaglandins appear to be required to activate viral gene expression (64). Thus, the induction of many inflammatory cytokines and interferon response genes is transient (6, 27), and de novo expression of virus-encoded genes acts in concert with incoming pp65 to suppress cellular antiviral gene functions (5).
While HCMV infection is known to produce profound changes in the regulation of cellular genes, published studies provide only limited insight into the control of IL-6 gene expression in the productively infected cell. Although the 72-kilodalton HCMV immediate-early (IE) protein 1 (IE1p72) can mediate transcriptional activation of the IL-6 promoter in monocytic cells (24), expression of IE genes is limited in undifferentiated monocytes (49), and it is not known whether such activation occurs during productive infection. Moreover, most previous work has been carried out in the presence of serum, which has profound effects on the expression of cytokines and induces IL-6 expression in fibroblasts (29). In this study, we analyzed in detail the regulation of IL-6 expression in HCMV-infected primary human fibroblasts, in both the absence and the presence of serum. Fibroblasts remain the best-characterized cell system for the study of productive HCMV replication and are a potential source of IL-6 during localized episodes of viral replication in vivo. Our data reveal a hitherto-unsuspected level of complexity in the regulation of IL-6 expression in the infected cell and an apparently novel mode of gene regulation by HCMV at the posttranscriptional level.
MATERIALS AND METHODS
Cells. Human fetal foreskin fibroblasts (HFFF) and human glioblastoma astrocytoma cells (U373-MG) were obtained from the European Cell and Culture Collection (Salisbury, United Kingdom) and were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS; Invitrogen, Paisley, United Kingdom). All cells were tested periodically for mycoplasma contamination using the Mycoplasma Plus PCR primer set (Stratagene, La Jolla, Calif.) and found to be negative.
Virus culture and infection. The attenuated HCMV strain Towne and the low-passage strain Toledo were gifts of E. Mocarski (Stanford University, Stanford, Calif.). Strain Merlin, a low-passage HCMV strain with wild-type gene content, has been described elsewhere (17). A Toledo mutant lacking the early (?) 2.7-kb RNA gene (?2.7 gene) and the corresponding revertant virus have been described previously (33). Where indicated, virus was inactivated by UV irradiation (at 254-nm wavelength) using a biocidal lamp for 30 min at a distance of 20 cm; the amount of UV-inactivated virus used in experiments is expressed in terms of the titer prior to inactivation.
Analysis of IL-6 protein and mRNA in infected fibroblast cultures. HFFF were seeded in tissue culture vessels (Nunc; Fisher Scientific, Loughborough, United Kingdom) at 50% confluence and allowed to adhere overnight in DMEM-10% FCS. Except where indicated, cells were then were starved in serum-free medium for 24 h prior to infection or other challenge. Cells were inoculated either with active HCMV, typically at a multiplicity of infection (MOI) of 1 or as indicated, or with the same amount of UV-inactivated virus; mock-infected cultures received a volume of DMEM-10% FCS equal to that of the viral inoculum in each experiment. For the assay of IL-6 protein, culture medium was clarified by centrifugation (1,200 x g for 5 min) and stored at –20°C prior to analysis by IL-6-specific enzyme-linked immunosorbent assay (Quantikine kit; R&D Systems, Abingdon, United Kingdom). For the assay of IL-6 and ?-actin mRNA by reverse transcription (RT)-PCR, cells were harvested and resuspended in RNAlater reagent (Ambion, Austin, Tex.) prior to purification of RNA. After removal of the storage reagent, total RNA was isolated using the RNeasy Mini kit (QIAGEN, Crawley, United Kingdom), and yields were estimated by densitometry after gel electrophoresis. Equal amounts of RNA were used as templates for cDNA synthesis using SuperScript II reverse transcriptase (Invitrogen) primed with random hexamers (Amersham Biosciences, Amersham, United Kingdom) as recommended by the manufacturers. PCRs were carried out using a forward primer specific for exon 1 of the IL-6 gene (RT primer 1, 5'-CAGCTATGAACTCCTTCTCCAC-3') and a reverse primer specific for exon 2 (RT primer 2, 5'-GAGATGCCGTCGAGGATGTAC-3'); spliced ?-actin cDNA was amplified using primers 5'-ACAACGGCTCCGGCATGTGCAA-3' (forward, exon 1) and 5'-CCATGTCGTCCCAGTTGGTGAC-3' (reverse, exon 2). PCR products were analyzed by agarose gel electrophoresis and quantified by densitometry. In all RT-PCR analyses, PCRs were determined as being in the exponential phase of amplification by sampling reactions at different cycle numbers and plotting the increase in band intensity against cycle number.
Filtration and transfer of conditioned medium. In experiments involving transfer of conditioned medium, the medium from donor cultures was removed at the time indicated and filtered through syringe-mounted 0.1-μm-pore-size filters (Pall Gelman Laboratory, Ann Arbor, Mich.). An aliquot was retained for analysis of IL-6 by enzyme-linked immunosorbent assay, and the conditioned medium was then transferred onto recipient cell cultures, which had either received no prior treatment or had been challenged with UV-inactivated HCMV. In all cases, donor and recipient cell monolayers were washed briefly with phosphate-buffered saline (PBS) between removal of the original medium and addition of fresh or conditioned medium as appropriate.
Plasmids and transfections. Luciferase reporter constructs containing the human IL-6 promoter (p1168hu.IL6P-luc+, p234hu.IL6P-luc+, p110hu.IL6P-luc+, and p50hu.IL6P-luc+) were gifts of G. Haegeman (University of Gent, Ghent, Belgium) (57). Plasmids pcDNA3IE72 and pcDNA3IE86, for expression of HCMV (strain AD169) IE1p72 or IE2p86, respectively, have been described previously (7).
HFFF were transfected using Effectene reagent (QIAGEN); U373-MG cells were transfected using Transfast reagent (Promega, Madison, Wisc.). Where indicated, cells were infected with HCMV 24 h posttransfection and harvested 40 h posttransfection for assay of reporter gene activity. Luciferase activity was assayed using the Luciferase Assay System (Promega) and measured on a TD-20/20 luminometer (Turner Designs, Sunnyvale, Calif.).
Nuclear run-on assays. HFFF were seeded, serum starved, and inoculated as described above for analysis of IL-6 protein and mRNA. Transcription of the endogenous IL-6 gene was assayed using an RT-PCR-based technique, essentially as described by Hildebrandt and Neufer (26). Briefly, cells were harvested at 16 h postinfection (hpi); after gentle cell lysis, the nuclei were isolated, washed, and stored at –80°C in storage buffer (75 mM HEPES [pH 7.9], 40% [vol/vol] glycerol, 60 mM KCl, 15 mM NaCl, 5 mM magnesium acetate, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM dithiothreitol, 0.5 mM spermidine, 0.15 mM spermine, 2 μg of aprotinin/ml, 2 μg of leupeptin/ml). After thawing, each sample was divided in two; one aliquot was mixed with reaction buffer containing nucleoside triphosphates (NTPs; final concentrations: ATP, 1.2 mM; CTP, GTP, and UTP, 0.6 mM each), and the other received control buffer lacking nucleotides; run-on reaction mixtures were then incubated for 15 min at 22°C. RNA was isolated and used for cDNA synthesis as described above; control reactions were performed without reverse transcriptase in each case. Spliced IL-6 and ?-actin cDNA products were amplified using the primers described above; unspliced IL-6 cDNA products were amplified using RT primer 1 and a reverse primer specific for intron 1 of the IL-6 gene (RT primer 3, 5'-GGGTGACTGACAGCACAGCTG-3'). PCR products were quantified by densitometry following agarose gel electrophoresis.
RNA stability assays. Replicate cultures of serum-starved fibroblasts were inoculated at an MOI of 1 with either HCMV or UV-inactivated HCMV, as described above. At 16 hpi, transcription was inhibited by the addition of actinomycin D (Sigma-Aldrich, Poole, United Kingdom) to a final concentration of 5 μg/ml; cultures were harvested either immediately following actinomycin D treatment (0-h time point) or at 4 or 8 h after treatment. cDNA was prepared, and levels of IL-6 and ?-actin mRNA were analyzed using exon-specific primers as described above; the same cDNA samples were used to analyze cyclooxygenase-2 (COX-2) expression, using the following PCR primers: 5'-GTCCACCAACTTACAATGCTG-3' (forward) and 5'-ACATCATGTTTGAGCCCTGG-3' (reverse).
Western blotting. Cell lysates for analysis by Western blotting were resolved by denaturing sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using standard techniques and electroblotted onto polyvinylidene difluoride membranes (Hybond-P; Amersham Biosciences). Membranes were blocked in Tris-buffered saline (10 mM Tris-HCl [pH 8.0], 150 mM NaCl) containing 0.05% (vol/vol) Tween-20 and 5% (wt/vol) dried milk for 1 h at room temperature and then probed sequentially with primary and horseradish peroxidase (HRP)-conjugated secondary antibodies, diluted in a mixture of Tris-buffered saline containing 0.05% (vol/vol) Tween-20 and 5% milk, as indicated. After washing, membrane-bound HRP activity was detected by enhanced chemiluminescence (SuperSignal West Pico reagent; Perbio, Tattenhall, United Kingdom). Monoclonal antibody MAB810, specific for HCMV IE1p72/IE2p86, was from Chemicon (Temecula, Calif.); antibodies specific for HuR (sc-5261) or -tubulin (sc-5286) and HRP-conjugated anti-mouse immunoglobulin G (IgG; sc-2005) were from Santa Cruz Biotechnology (Santa Cruz, Calif.).
Indirect immunofluorescence. Serum-starved fibroblasts were grown on sterilized glass coverslips and inoculated with HCMV as appropriate. At 40 hpi, medium was removed, and cell monolayers were washed with PBS. Cells were fixed in PBS-2% (wt/vol) formaldehyde for 15 min, washed, and then permeabilized in PBS-0.2% (vol/vol) Triton X-100 for 5 min. After washing, coverslips were incubated with primary antibody (either HuR-specific IgG1 [sc-5261; Santa Cruz] or normal mouse IgG1 [sc-2025; Santa Cruz]) at a concentration of 1 μg/ml in PBS-10% (vol/vol) FCS. Coverslips were washed and then incubated with Texas Red-conjugated goat anti-mouse IgG (ab6787; Abcam; 5 μg/ml in PBS-10% FCS). Unbound sites on the secondary antibody were blocked by incubation of coverslips in PBS-10% FCS-10% (vol/vol) mouse serum (Abcam) for 30 min, prior to incubation with fluorescein isothiocyanate (FITC)-conjugated mouse antibody specific for HCMV pUL44 (a gift of Richard Greaves), diluted in PBS-10% FCS-10% mouse serum containing 100 ng of 4',6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich)/ml. After further washing, coverslips were mounted onto glass slides in glycerol containing 2% (wt/vol) 1,4-diazabicyclol[2.2.2]octane (DABCO; Sigma-Aldrich) and sealed with clear nail polish. Slides were viewed on a Leica (Milton Keynes, United Kingdom) DM IRBE fluorescence microscope, and images were captured using Openlab software (Improvision, Coventry, United Kingdom).
RESULTS
Induction of IL-6 expression by HCMV is transient in productive infection. Previous reports have shown that fibroblasts challenged with HCMV express elevated levels of IL-6 protein (8, 34) and mRNA (6, 8, 62, 63); there is, however, little data regarding the continued regulation of IL-6 expression as infection progresses through the lytic cycle. To discriminate between the effects induced by virion binding and/or uptake and de novo expression of HCMV gene products, cells were infected with HCMV or UV-inactivated HCMV, and the levels of secreted IL-6 were measured. HCMV and UV-inactivated virus both induced the release of substantial quantities of IL-6 (Fig. 1A), with higher levels in cultures challenged with UV-inactivated virus than in those infected with HCMV at 16 and 40 hpi. This induction of IL-6 is not mediated by autoregulation, since neither uninfected nor HCMV-infected fibroblasts express a functional IL-6 receptor (S. A. Jones, personal communication). The abundance of IL-6 mRNA at 16 hpi correlates with the level of IL-6 in the medium; IL-6 mRNA levels were up-regulated by HCMV infection yet were approximately twofold more abundant in cells challenged with UV-inactivated virus (Fig. 1B). However, by 40 hpi, the level of IL-6 mRNA had declined sharply in productively infected cells, whereas the message remained abundant in cells challenged with UV-inactivated virus. IL-1? induced high levels of both IL-6 protein and mRNA (Fig. 1A and B, respectively), confirming that cells responded normally to a known regulator of IL-6 expression. IL-6 expression was stimulated in a manner similar to that seen with the attenuated Towne strain by two low-passage, pathogenic strains of HCMV, Toledo and Merlin (Fig. 1C). The above experiments were performed under serum-free conditions, to eliminate the confounding effects of serum on cytokine expression; however, experiments performed in the presence of 10% serum also showed that while HCMV infection induced IL-6 expression from fibroblasts, higher levels of IL-6 were induced by UV-inactivated virus (Fig. 1D). The induction of IL-6 required a direct interaction between virion and cell, as removal of virus particles from the inoculum by filtration through a 0.1-μm-pore-size filter inhibited the production of IL-6, whereas use of a filter which allowed the passage of virions had no effect (Fig. 1E).
The results implicate a factor associated with virion binding and/or uptake being responsible for the sustained up-regulation of IL-6 production, mediated at least in part by an increase in the steady-state level of IL-6 mRNA. This effect is countered, albeit not completely, by the onset of HCMV lytic gene expression. To determine whether increased numbers of virion particles could overcome the down-regulation of IL-6 secretion observed in lytic infection, we challenged cells at a higher MOI with either HCMV or UV-inactivated HCMV alone or a mixture of both active and inactivated virus. While cells challenged with UV-inactivated virus alone again expressed much higher levels of IL-6 than productively infected cells (Fig. 1E), cells receiving a mixed inoculum (equivalent to a total of 10 PFU/cell) expressed a level of IL-6 which was similar to that seen with active virus alone and was greatly reduced compared with that expressed by cells receiving the same amount of UV-irradiated virus. Thus, the onset of viral gene expression results in suppression of continued IL-6 expression by the infected cell; the higher levels of IL-6 seen in cultures challenged with UV-inactivated virus are due to the lack of such suppression and are not an artifact of UV irradiation of the inoculum. Consistent with this, we observed no IL-6 expression after mock infection of control cultures using a UV-irradiated inoculum (data not shown). Furthermore, at high multiplicities of infection the virus restricts further accumulation of IL-6 between 16 and 40 hpi, whereas cultures challenged with inactivated virus continue to accumulate IL-6 in the medium during this period, consistent with the hypothesis that suppression of IL-6 secretion is an active process requiring HCMV lytic gene expression.
Induction of IL-6 expression by HCMV is not mediated by soluble factors. The results above show that induction of IL-6 requires contact between virion and cell. However, it is not clear whether this contact induces IL-6 expression directly by intracellular mechanisms (e.g., by activation of signal transduction pathways which lead to transcriptional up-regulation of the IL-6 promoter by cellular factors or by direct activation of the IL-6 promoter by a virion transactivator) or HCMV induces the release of other soluble factors into the medium, which subsequently trigger IL-6 expression in an autocrine or paracrine manner. To address this, virus-free tissue culture supernatants were collected from HCMV-infected cells and cells challenged with UV-inactivated HCMV (donors) at 6 or 16 hpi and transferred onto untreated recipient cells (Fig. 2). In all cases, the concentrations of IL-6 observed in recipient cell cultures, subsequent to incubation with the transferred medium, were very similar to those in the donor medium at the time of transfer; i.e., medium from virus-challenged cells possessed little ability to induce IL-6 expression de novo. Furthermore, medium from virus-challenged cells failed to induce significant expression of IL-6 mRNA in recipient cultures (Fig. 2B and D). IL-6 continued to be secreted from donor cells challenged with UV-inactivated HCMV after replacement of the conditioned medium with fresh DMEM at 6 or 16 hpi, whereas cells undergoing active infection produced no significant amounts of IL-6 after removal of the original medium (Fig. 2A and B, respectively). In contrast, medium from donor cells treated with IL-1? induced high levels of IL-6 mRNA and protein expression in recipient cultures, whereas the donor cells failed to produce significant amounts of IL-6 after removal of IL-1?. These data do not support a mechanism whereby HCMV induces up-regulation of IL-6 expression by an indirect pathway involving the prior release of soluble factors into the medium. As an additional test, tissue culture supernatants from this experiment were assayed for a number of known inducers of IL-6. IL-1?, tumor necrosis factor alpha, and interferon-? could not be detected in any samples following virus infection; low levels of transforming growth factor ?1 (TGF-?1) were detected at 40 hpi, but not reproducibly at 16 hpi, and at similar levels in cells inoculated with HCMV or UV-inactivated HCMV (data not shown).
Suppression of IL-6 expression is not mediated by an autocrine pathway in HCMV-infected fibroblasts. The results above show that induction of IL-6 expression is not mediated primarily by an autocrine or paracrine mechanism, but they do not address the suppression of IL-6 secretion that is observed subsequent to the onset of lytic gene expression. One potential candidate which might mediate suppression is the HCMV homologue of IL-10 (cmvIL-10), and indeed this has been shown to inhibit lipopolysaccharide-mediated induction of IL-6 by monocytes (51) and signal via the cellular IL-10 receptor (31). However, inhibition of IL-10 signaling by use of a neutralizing antibody against the IL-10 receptor had no significant effect on the levels of IL-6 induced by HCMV or UV-inactivated HCMV (Fig. 3).
To test further whether cmvIL-10 or another soluble factor might be responsible for the suppression of IL-6, conditioned medium from HCMV-infected donor cells was transferred at 16 hpi onto recipient cells that had been challenged 6 h previously with UV-inactivated HCMV. Recipient cultures were then incubated for a further 16 h. If the conditioned medium contained a soluble suppressor, then subsequent IL-6 secretion by the recipient cultures should have been inhibited. However, recipient cultures receiving medium from HCMV-infected cells produced an amount of IL-6, after transfer of medium, similar to that produced by control cultures receiving either fresh DMEM or conditioned medium from mock-infected cells (Table 1). Moreover, donor cells infected at an MOI of 3 produced less IL-6 than those infected at an MOI of 1 (Table 1), consistent with active suppression of IL-6 expression by HCMV infection; however, there was little difference in the amount of IL-6 produced by the respective recipient cultures (Table 1, recipient cultures 1 and 2). These results indicate that the suppression of IL-6 secretion is not mediated by a soluble factor released into the culture medium but is exerted within the infected cell, possibly at the transcriptional, posttranscriptional, or translational level.
HCMV activates the IL-6 promoter through core elements. HCMV infection up-regulates the steady-state levels of IL-6 mRNA (Fig. 1B). The human IL-6 promoter is recognized to contain multiple positive regulatory elements and at least one negative regulatory element, the RBP site (summarized schematically in Fig. 4A). To further delineate the mechanism by which HCMV up-regulates IL-6 expression, a series of reporter plasmids containing truncated versions of the IL-6 promoter were employed in transient transfection experiments in permissive U373-MG cells. Productive HCMV infection resulted in strong activation of the IL-6 promoter at 16 hpi that was dependent only on the core promoter; a construct lacking all regulatory elements upstream of the major TATA motif gave the same level of expression as the longest promoter construct tested (Fig. 4B). Furthermore, challenge with UV-irradiated virus resulted only in weak, albeit detectable, activation of IL-6 promoter constructs; this modest response was essentially abolished by deletion of sequences between nucleotides –110 and –50 encoding the NF-B and C/EBP binding sites. Control experiments demonstrated that IL-1? activated the expression of plasmids containing at least 110 nucleotides of promoter sequence upstream of the transcription start site, whereas there was no effect on the shortest construct (–50), which lacked all upstream regulatory elements (data not shown). Transfection experiments were repeated with fibroblasts at both 16 h and 40 hpi to emulate the conditions used to monitor IL-6 protein and mRNA levels (Fig. 1). Although the transfection efficiency was much reduced relative to that in U373-MG cells (as indicated by lower basal and activated luciferase levels), both full-length and core promoter constructs were again strongly activated in HCMV-infected fibroblasts at 16 hpi, whereas there was little or no detectable activation by UV-irradiated virus (Fig. 4C). Moreover, transcriptional activity was sustained, and in fact increased, at 40 hpi.
The ability of HCMV to activate through core promoter elements has also been observed for the c-myc, c-fos, hsp70, and TGF-?2 cellular promoters and the human immunodeficiency virus type 1 long terminal repeat (9, 25, 59). In these studies, activation of the core promoter was shown to be mediated by the HCMV immediate-early proteins. Furthermore, the increased promoter activation observed in HCMV-infected fibroblasts at 40 hpi compared to that at 16 hpi (Fig. 4C) correlated with accumulation of the major immediate-early proteins IE1p72 and IE2p86 (data not shown). We therefore tested whether HCMV IE1p72 or IE2p86 could activate the IL-6 promoter in permissive cells. A plasmid encoding IE2p86 strongly up-regulated the IL-6 promoter in U373-MG cells, and this was independent of upstream regulatory elements (Fig. 4D). Thus, activation by both HCMV and IE2p86 shared a requirement only for the core IL-6 promoter, suggesting that IE2p86 may be the dominant activator of the IL-6 promoter during productive infection. Comparable results were also obtained from experiments performed with primary fibroblasts and nonpermissive CV-1 cells (data not shown).
In contrast to the strong activation by IE2p86, the effects of IE1p72 were weak, variable, and possibly cell type specific. In U373-MG, the weak activation induced by IE1p72 was relatively independent of upstream promoter elements (Fig. 4E). In CV-1 cells, however, while IE1p72 was able to up-regulate the longest IL-6 promoter construct to an extent similar to that seen in U373-MG, there was no significant or reproducible effect on any of the truncated promoter constructs (data not shown). Consistent with this, there was some evidence that IE1p72 could augment, weakly, IE2p86-mediated activation of the IL-6 core promoter in U373-MG (Fig. 4F) but not CV-1 cells (data not shown).
Transcriptional activation of the endogenous IL-6 gene in productive infection. The strong activation of reporter gene expression from the IL-6 promoter in transfection assays by productive HCMV infection or IE2p86 alone (Fig. 4) would appear superficially to be inconsistent with the observation that lytic gene expression suppresses steady-state levels of IL-6 mRNA and IL-6 secretion (Fig. 1). To address this, nuclear run-on assays were performed to examine the level of ongoing transcription from the endogenous IL-6 gene at 16 hpi in fibroblasts; the RT-PCR-based method utilized distinguishes between spliced and unspliced transcripts using different primer sets (Fig. 5A). A low level of spliced but not unspliced IL-6 mRNA was detected in the nuclei of mock-infected cells, and its abundance did not increase in the run-on assay (i.e., in the presence of NTPs); thus, there was no significant ongoing IL-6 transcription in these cells. There was also little or no ongoing transcription from the ?-actin gene during the run-on assay in any of the samples; this was expected since uninfected cells were quiescent and HCMV infection induces only a modest increase in ?-actin mRNA (25). However, there was a significant increase in the levels of both spliced and unspliced IL-6 transcripts in the nuclei of productively infected cells during the run-on assay (Fig. 5B to D), whereas there was little or no increase in transcript levels in nuclei from cells challenged with UV-inactivated virus. These effects were reproducible, and data from a second experiment are combined in Fig. 5E. Thus, both DNA transfection experiments (Fig. 4) and nuclear run-on assays are consistent with the ability of productive HCMV infection, but not UV-inactivated virus challenge, to induce significant up-regulation of transcription from the IL-6 promoter at 16 hpi.
The onset of viral gene expression results in increased turnover of IL-6 mRNA. The above data suggest that transcription from the IL-6 promoter is activated first transiently by a virion factor (either directly or by the activation of signal transduction cascades leading to activation of a cellular transcription factor), then subsequently by one or more HCMV transcriptional activators expressed de novo, likely including IE2p86. However, IL-6 mRNA levels were lower in productively infected cells than in cells challenged with UV-inactivated virus (Fig. 1B), implying that the rate of IL-6 mRNA turnover was enhanced during productive infection. To test this hypothesis, ongoing transcription was blocked with actinomycin D at 16 hpi, and cells were analyzed to determine the level of IL-6 mRNA remaining at intervals following the block. In cells harvested immediately after actinomycin D treatment, the relative level of IL-6 mRNA was significantly higher in cells inoculated with UV-inactivated virus than in those infected with HCMV (Fig. 6A), as seen for steady-state levels of mRNA. In productively infected cells the level of IL-6 mRNA declined rapidly and was barely detectable 8 h after the block; by contrast, in the abortive infection, IL-6 mRNA persisted at around 70% of the original level even 8 h after the block (Fig. 6B). Decay appeared to show first-order exponential kinetics (Fig. 6C), giving a half-life for IL-6 mRNA during productive HCMV infection of approximately 77 min, compared with approximately 13 h for that in cells challenged with UV-inactivated virus. The long half-life of IL-6 mRNA in cells challenged with inactivated virus is consistent with our previous data (Fig. 1B), which showed that the message remained abundant in cells at 40 hpi even though there was little or no ongoing transcription after 16 hpi (Fig. 4 and 5).
The stability of IL-6 mRNA is regulated at least in part by AREs in the 3' UTR. Interestingly, COX-2 mRNA, which is also subject to ARE-mediated regulation (15, 36, 46, 53), was extremely stable in cells challenged with UV-inactivated virus over an 8 h period but underwent significant degradation in productively infected cells over the same period; however, the kinetics of COX-2 mRNA decay in HCMV-infected cells appeared to be more complex than that for IL-6 (Fig. 6B). In contrast, mRNA for ?-actin, which is not subject to ARE-mediated turnover, remained stable in both productively and abortively infected cells during the experiment (Fig. 6A). These results suggest that HCMV may potentially encode a mechanism to preferentially accelerate the turnover of ARE-containing mRNAs.
HCMV infection induces relocalization of a regulator of ARE-containing mRNAs. The mechanism by which AREs promote mRNA turnover has not yet been fully elucidated; however, it has been reported that proteasome activity is required for ARE-mediated turnover of mRNA (32). Therefore, we tested the effect of the proteasome inhibitor MG132 on IL-6 expression by HCMV-infected cells by inoculating cells with either HCMV or UV-inactivated virus and adding the inhibitor at various time points (0, 2, or 4 hpi). In addition, control cultures were treated with IL-1?. Cells and culture medium harvested at 16 hpi were then assayed for IL-6 and IE gene expression. Treatment with MG132 resulted in strong inhibition of IL-6 expression when applied at 0 or 2 h postchallenge, and to a lesser extent when applied at 4 h, and this inhibition was observed with all stimuli (Fig. 7A). Furthermore, MG132 inhibited the expression of IE proteins in cells challenged with active virus (Fig. 7C), and as for IL-6 expression, this effect was strongest when MG132 was applied at the same time as the viral inoculum. Similar effects were observed when cultures were grown in the presence of serum, though as expected, the basal and induced levels of IL-6 expression were much higher (Fig. 7B). The effects of MG132 are discussed below; however, since viral gene expression was required for suppression of IL-6 but MG132 also inhibited the expression of viral proteins, it was not possible to use this approach directly to determine whether the increased turnover of IL-6 mRNA in productively infected cells was dependent on proteasome activity.
As an alternative approach, we asked whether productive HCMV infection could alter the level or activity of regulators of ARE-mediated mRNA turnover. The best-characterized positive regulator of ARE-mediated mRNA stability is HuR, a member of the Elav family of RNA-binding proteins. HuR stabilizes mRNAs as a result of direct binding to ARE motifs (4), including those found in IL-6 and COX-2 mRNAs (38, 46); overexpression of HuR stabilizes ARE-containing mRNAs in vivo (21, 43, 46), whereas inhibition of HuR activity by siRNA has the opposite effect (46, 53). However, in HCMV-infected cells the total level of HuR protein was increased compared to that in either mock-infected cells or those challenged with UV-inactivated virus (Fig. 8A). This increase was evident by 24 hpi and became more pronounced at later times of infection. As observed for the phenomenon of IL-6 suppression, induction of HuR by active HCMV infection was also evident in cultures grown in 10% serum (Fig. 8A, right). The level of HuR in cells treated with IL-1? was similar to that seen in mock-infected cells and those challenged with inactivated virus (data not shown), indicating that an increase in the overall level of HuR was not required for high levels of IL-6 expression.
Although overexpression of HuR has previously been correlated with stabilization of ARE-containing mRNAs, we observed that the half-lives of IL-6 and COX-2 mRNA were much reduced in HCMV-infected cells despite elevated levels of HuR (Fig. 8A). Two forms of HuR were apparent in Western transfer experiments, and while these may possibly represent different phosphorylation states of the protein, HCMV infection did not significantly alter the relative proportion of these forms. However, HCMV infection had a profound effect on the distribution of HuR within intracellular compartments. Consistent with published studies, HuR was found predominantly in the nuclei of mock-infected cells (21, 38) and in cells challenged with IL-1? or UV-inactivated HCMV (Fig. 8B); nuclear localization of HuR is therefore compatible with both the lack of IL-6 expression from quiescent cells and induced expression. Cells undergoing productive HCMV infection were identified by staining for the viral DNA polymerase accessory protein pUL44; in such cells there was significant cytoplasmic staining of HuR (Fig. 8B, panels i), whereas uninfected cells in the same sample retained HuR in the nucleus (Fig. 8B, panels i). The cytoplasmic localization of HuR was more pronounced at a higher multiplicity of infection (Fig. 8B, panels ii), with some cells appearing to exclude HuR from the nucleus (Fig. 8B, panels ii). Cytoplasmic staining was not due to cross-reactivity with viral proteins, as staining using an isotype-matched primary antibody control did not show cytoplasmic Texas Red fluorescence in HCMV-positive cells at a multiplicity of either 1 (Fig. 8B, panels vi) or 3 (data not shown). (Modest cross-reactivity with a component of viral replication centers could be seen in infected cells; this appeared to be due to binding by the secondary antibody, as independent experiments using a different FITC-conjugated antibody with the same primary antibodies showed no such cross-reactivity [data not shown]).
HuR redistribution is independent of HCMV ?2.7. Since suppression of IL-6 was not mediated by a soluble factor in the culture medium, precluding a role for cmvIL-10 (Fig. 3), we asked what other viral gene product(s) might be responsible. Interestingly, inspection of the sequence of the HCMV ?2.7 RNA transcript, which is abundantly expressed from the duplicated IRL4/TRL4 genes of HCMV strain AD169 at early times of infection (50), reveals a number of AREs and ARE-like motifs in the 3' region of the molecule. The function of the ?2.7 gene is not yet known; moreover, although the overall gene sequence is strongly conserved in other strains of HCMV, the open reading frame present in the AD169 ?2.7 transcript is not (33), suggesting that the RNA itself may be the functional gene product. This led us to speculate that ?2.7 RNA might act as a decoy or competitor for HuR, preventing HuR from binding to and stabilizing IL-6 mRNA. To test this possibility, cells were infected with wild-type HCMV Toledo (which carries a single copy of the ?2.7 RNA gene), a mutant lacking the ?2.7 RNA gene, and a revertant strain. We anticipated that if the ?2.7 RNA was responsible for inhibition of HuR activity and/or destabilization of IL-6 mRNA, infection with a mutant lacking this gene would express IL-6 at greatly increased levels, due to the high level of ongoing transcription in productively infected cells (Fig. 4 and 5). However, the level of IL-6 produced by infected cells was essentially the same for all three viruses (Fig. 8C) and was strongly suppressed in comparison to that in cultures challenged with UV-inactivated viruses. As observed previously (Fig. 1B), the level of IL-6 mRNA correlated with the level of IL-6 protein in the medium (Fig. 8D); furthermore, all three virus strains showed similar ability to induce HuR during productive, but not abortive, infection (Fig. 8E), as observed for wild-type strain Towne. Similar experiments were carried out using wild-type strain AD169 and mutants lacking one or both copies of the ?2.7 RNA gene; as seen with viruses in the Toledo background, these showed no differences in their ability to suppress IL-6 expression or induce HuR during active infection (data not shown). We conclude that while both attenuated and low-passage strains of HCMV are able to suppress IL-6 expression at the posttranscriptional level and induce the overexpression of HuR, the ?2.7 RNA is not required for these phenomena. Moreover, since the Towne and Toledo strains of HCMV exhibited similar effects on IL-6 and HuR expression, we conclude that such effects are not exerted by the products of the additional genes present in the Toledo strain.
DISCUSSION
The binding and fusion of HCMV virions with the cell surface, combined with the delivery of virion components and tegument proteins in particular, to the infected cell is known to initiate the activation of multiple signal transduction pathways and the induction of inflammatory and antiviral gene expression. Consistent with these observations, challenge with UV-inactivated HCMV particles is sufficient to elicit a strong induction of IL-6 expression (8). Transcriptional activation of the IL-6 promoter by virion particles appears to be transient and mediated largely through the NF-B pathway; deletion of the sequence containing the NF-B binding site abolished the ability of UV-irradiated virus to activate the promoter (Fig. 4B). Moreover, Carlquist et al. reported that an antagonist of NF-B inhibited IL-6 expression in response to challenge with inactivated HCMV, while we observed that MG132, which inhibits NF-B activation by preventing proteasomal degradation of IB, also strongly reduced IL-6 expression in response to both IL-1? and viral challenge (Fig. 7A and B). Interestingly, we observed that MG132 also markedly inhibited the expression of HCMV IE genes (Fig. 7C). Since the HCMV major IE enhancer contains multiple NF-B sites, it is possible that reduced IE expression is a direct consequence of inhibition of this transcription factor. Indeed, a recent report has shown that both MG132 and aspirin inhibit IE gene expression in HCMV-infected fibroblasts, while inhibition of the NF-B activation pathway decreases expression from the major IE promoter (16). However, our experiments do not rule out the possibility that the onset of viral gene expression requires other events which are dependent on proteasomal activity, and such putative events are not addressed by our work.
The effect of UV-inactivated HCMV on IL-6 expression resembled that resulting from IL-1 challenge. While this inflammatory cytokine up-regulates transcription from the IL-6 promoter, the effect is only modest, and the large increase in IL-6 protein expression that occurs in response to IL-1 is predominantly due to extensive stabilization of the IL-6 mRNA (45). Similarly, we observed that UV-inactivated HCMV was only a weak activator of the IL-6 promoter, but the half-life of IL-6 mRNA was greatly prolonged in abortively infected cells, allowing these cells to maintain synthesis of IL-6 protein over an extended period. The similar effect of IL-1 and HCMV virions is consistent with the hypothesis that virion interaction with the host cell induces an inflammatory or antiviral response.
In contrast to the effects of inactivated virions, the onset of HCMV gene expression appears to curtail IL-6 expression by the productively infected cell (Fig. 1). Paradoxically, transcription initiation from the IL-6 promoter was up-regulated in productively infected cells; however, the reduced level of IL-6 mRNA and protein can be accounted for by the greatly increased turnover of IL-6 mRNA in productively infected cells compared to those challenged with inactivated virions. Additionally, we found no evidence for accumulation or retention of IL-6 protein within productively infected cells (data not shown), which might otherwise account for lower levels in the culture medium. Thus, whereas inactivated virions induce an inflammatory state characterized by induction and stabilization of the IL-6 mRNA, the onset of HCMV lytic gene expression counteracts this effect by accelerating the turnover of IL-6 mRNA.
It is well established that the stability of IL-6 mRNA is regulated extensively through elements in the 3' UTR and in particular AREs. Interestingly, the turnover of another ARE-containing mRNA, that of COX-2, was also accelerated in productively infected cells compared with those challenged with inactivated virions (Fig. 6). Moreover, this accelerated decay coincided with aberrant localization of HuR, an ARE-binding protein which binds to both IL-6 and COX-2 mRNAs and stabilizes ARE-containing mRNAs in the cell. The increased turnover of ARE-containing mRNAs occurred despite an elevated level of HuR protein in the HCMV-infected cell, whereas studies in uninfected cells have shown that HuR overexpression results in stabilization of such mRNAs. Conversely, inhibition of HuR expression by siRNA leads to decreased expression of COX-2 (46, 53), suggesting that the increased turnover of IL-6 and COX-2 mRNAs that we observe during productive HCMV infection might be due to inhibition of the mRNA-stabilizing activity of HuR. Furthermore, we found no change during HCMV infection in the expression of BRF-1 (data not shown), an ARE-binding protein which destabilizes ARE-containing mRNAs and thus acts in a manner antagonistic to HuR (52).
It is generally accepted that HuR functions by binding to target mRNAs in the nucleus and protecting them against degradation prior to translation; thus, displacement of HuR from its normal targets would promote turnover of these mRNAs. One possible mechanism by which HCMV might achieve this is by expression of a decoy RNA. Interestingly, the small nuclear RNAs of herpesvirus saimiri have been shown to bind HuR in vitro (37) and in vivo (14); however, their expression does not appear to affect the levels of cellular mRNAs (14). Moreover, deletion of the HCMV gene encoding the ?2.7 RNA, a candidate decoy which contains multiple AREs and ARE-like motifs near the 3' end, had no effect on the ability of active HCMV infection to suppress IL-6 expression. An alternative mechanism for HuR displacement might be by interaction with a viral protein, which interferes with mRNA binding and/or the subcellular localization of HuR; this might explain the aberrant localization of HuR we observed in productively infected cells. Notably, the increase in total HuR levels seen during infection is reminiscent of the overexpression of p53 that occurs when this protein is inactivated either by mutation or by interaction with viral oncoproteins. Whether or not HuR is engaged in novel interactions in the infected cell is clearly an intriguing subject for further research.
The database of ARE-containing mRNAs (http://rc.kfshrc.edu.sa/ared) contains over 900 entries (2); therefore, if HCMV infection were to result in inhibition of HuR activity, the effect on cellular gene expression might be expected to be widespread. Interestingly, of the mRNAs identified by Browne et al. (6) as being more abundant in cells challenged with UV-inactivated virus than in those undergoing productive infection, around 30% are found in the ARE database. Moreover, the supplementary data of Browne et al. (available at http://www.molbio.princeton.edu/labs/shenk/browneetal2001/) shows that both IL-6 and COX-2 mRNAs are more abundant in cells challenged with inactivated virus, consistent with our observations. HCMV may therefore have evolved a mechanism for the suppression of inflammatory gene expression that is mediated through the AREs of target mRNAs, to complement other mechanisms that inhibit interferon-mediated induction of gene expression (5). The susceptibility of different ARE-containing mRNAs to such a mechanism would be expected to vary widely, however. AREs are heterologous in sequence and number (2, 54) and vary in their responsiveness to different regulatory pathways. For example, a global study of the stability of ARE-containing mRNAs in monocytes showed that inhibition of p38 mitogen-associated protein kinase activity greatly reduced the half-life of COX-2 mRNA but had little effect on the turnover of IL-6 mRNA (23). Since HCMV infection is known to up-regulate p38 activity, this may explain the different kinetics of IL-6 and COX-2 mRNA decay in HCMV-infected cells following actinomycin D treatment (Fig. 6).
A final question posed by our data is, why does HCMV both activate IL-6 expression at the level of transcription initiation and suppress expression at the posttranscriptional level? Given the wide spectrum of biological effects of IL-6, it is conceivable that under certain circumstances the expression of IL-6 could be favorable for viral replication or persistence and that continued transcription might allow the infected cell to produce IL-6 protein under conditions that stabilize the IL-6 mRNA and thus counteract suppression. For example, IL-6 antagonizes TGF-?-induced apoptosis in hepatocytes (11) and might potentially play a role as an antiapoptotic factor during HCMV infection of this particular cell type. However, it seems more likely that suppression of IL-6 expression plays a role in immune evasion and the ability of HCMV to persist in the infected host, particularly if the suppressive phenomenon we observe for IL-6 forms part of a wider mechanism for down-regulation of inflammatory and immunoregulatory genes. Moreover, our preliminary data indicate that hepatocytes, which are key mediators of the acute-phase response, produce lower amounts of IL-6 during productive HCMV infection than cells challenged with inactivated virus (data not shown). Viewed in this context, transcriptional activation of the IL-6 promoter by HCMV would appear to be an unwanted activity. However, activation appears to be largely mediated by the essential IE2p86 protein; it may be that the molecular mechanisms by which IE2p86 up-regulates the IL-6 promoter are also required for activation of HCMV promoters, and therefore such functions would be indispensable for viral replication, forcing the virus to evolve distinct mechanisms to counteract such unwanted side effects.
In summary, our results demonstrate that expression of IL-6, and potentially other inflammatory genes, is subject to suppression at the posttranscriptional level by a virally encoded factor that is expressed at IE or early times of infection. This factor remains unidentified, but our results rule out cmvIL-10, the ?2.7 RNA, and the products of genes that are deleted in the attenuated Towne strain of HCMV, suggesting that the factor responsible plays an important role in viral replication both in vivo and in vitro. Accelerated turnover of IL-6 mRNA coincides with relocalization of HuR, suggesting that HCMV may target HuR activity as part of a mechanism for down-regulation of ARE-containing mRNAs, thus overcoming the antiviral state induced in the cell by virion binding and entry. Interestingly, recent reports indicate that the vhs gene product of herpes simplex virus type 1 preferentially targets ARE-containing mRNAs for degradation (19) and that herpes simplex virus type 1 infection results in the cytoplasmic accumulation of cellular proteins that promote the turnover of AU-rich mRNAs (20). Thus, selective destabilization of ARE-containing mRNAs by various mechanisms may be a common theme in the strategies employed by herpesviruses to disrupt cellular gene expression and enable persistence in the immune-competent host.
ACKNOWLEDGMENTS
This work was supported by a Career Establishment grant (G9801388) from the UK Medical Research Council to R.C. and a Cooperative Group Grant (G9827961) from the MRC to R.C., G.W., and other members of the Cooperative Group in Pathogenesis of Persistent Human Virus Infections.
We are grateful to Richard Greaves, Guy Haegeman, and John Sinclair for plasmid and antibody reagents, to Simon Jones and Suzanne Hurst for unpublished data from IL-6R assays, to Sin Llewellyn-Lacey and the late Vicky Dickinson for technical assistance, and to Simon Jones and Mark Stinski for helpful discussions.
REFERENCES
Alcami, A., and U. H. Koszinowski. 2000. Viral mechanisms of immune evasion. Trends Microbiol. 8:410-418.
Bakheet, T., B. R. G. Williams, and K. S. A. Khabar. 2003. ARED 2.0: an update of AU-rich element mRNA database. Nucleic Acids Res. 31:421-423.
Boyle, K. A., R. L. Pietropaolo, and T. Compton. 1999. Engagement of the cellular receptor for glycoprotein B of human cytomegalovirus activates the interferon-responsive pathway. Mol. Cell. Biol. 19:3607-3613.
Brennan, C. M., and J. A. Steitz. 2001. HuR and mRNA stability. Cell. Mol. Life Sci. 58:266-277.
Browne, E. P., and T. Shenk. 2003. Human cytomegalovirus UL83-coded pp65 virion protein inhibits antiviral gene expression in infected cells. Proc. Natl. Acad. Sci. USA 100:11439-11444.
Browne, E. P., B. Wing, D. Coleman, and T. Shenk. 2001. Altered cellular mRNA levels in human cytomegalovirus-infected fibroblasts: viral block to the accumulation of antiviral mRNAs. J. Virol. 75:12319-12330.
Bryant, L. A., P. Mixon, M. Davidson, A. J. Bannister, T. Kouzarides, and J. H. Sinclair. 2000. The human cytomegalovirus 86-kilodalton major immediate-early protein interacts physically and functionally with histone acetyltransferase P/CAF. J. Virol. 74:7230-7237.
Carlquist, J. F., L. Edelman, D. W. Bennion, and J. L. Anderson. 1999. Cytomegalovirus induction of interleukin-6 in lung fibroblasts occurs independently of active infection and involves a G protein and the transcription factor, NF-B. J. Infect. Dis. 179:1094-1100.
Caswell, R. C., C. Hagemeier, G. P. Hayhurst, T. Kouzarides, and J. H. Sinclair. 1993. Mechanisms of transactivation of cellular promoters by the HCMV major immediate early proteins involves direct interaction with cellular transcription factors, p. 13-16. In S. Michelson and S. A. Plotkin (ed.), Multidisciplinary approaches to understanding cytomegalovirus disease. Elsevier, Amsterdam, The Netherlands.
Cha, T. A., E. Tom, G. W. Kemble, G. M. Duke, E. S. Mocarski, and R. R. Spaete. 1996. Human cytomegalovirus clinical isolates carry at least 19 genes not found in laboratory strains. J. Virol. 70:78-83.
Chen, R. H., M. C. Chang, Y. H. Su, Y. T. Tsai, and M. L. Kuo. 1999. Interleukin-6 inhibits transforming growth factor-beta-induced apoptosis through the phosphatidylinositol 3-kinase/Akt and signal transducers and activators of transcription 3 pathways. J. Biol. Chem. 274:23013-23019.
Chomarat, P., J. Banchereau, J. Davoust, and A. K. Palucka. 2000. IL-6 switches the differentiation of monocytes from dendritic cells to macrophages. Nat. Immunol. 1:510-514.
Compton, T., E. A. Kurt-Jones, K. W. Boehme, J. Belko, E. Latz, D. T. Golenbock, and R. W. Finberg. 2003. Human cytomegalovirus activates inflammatory cytokine responses via CD14 and Toll-like receptor 2. J. Virol. 77:4588-4596.
Cook, H. L., H. E. Mischo, and J. A. Steitz. 2004. The Herpesvirus saimiri small nuclear RNAs recruit AU-rich element-binding proteins but do not alter host AU-rich element-containing mRNA levels in virally transformed T cells. Mol. Cell. Biol. 24:4522-4533.
Dean, J. L. E., S. J. Sarsfield, E. Tsounakou, and J. Saklatvala. 2003. p38 mitogen-activated protein kinase stabilizes mRNAs that contain cyclooxygenase-2 and tumor necrosis factor AU-rich elements by inhibiting deadenylation. J. Biol. Chem. 278:39470-39476.
DeMeritt, I. B., L. E. Milford, and A. D. Yurochko. 2004. Activation of the NF-B pathway in human cytomegalovirus-infected cells is necessary for efficient transactivation of the major immediate-early promoter. J. Virol. 78:4498-4507.
Dolan, A., C. Cunningham, R. D. Hector, A. F. Hassan-Walker, L. Lee, C. Addison, D. J. Dargan, D. J. McGeoch, D. Gatherer, V. C. Emery, P. D. Griffiths, C. Sinzger, B. P. McSharry, G. W. G. Wilkinson, and A. J. Davison. 2004. Genetic content of wild-type human cytomegalovirus. J. Gen. Virol. 85:1301-1312.
Duncan, S. R., I. L. Paradis, S. A. Yousem, S. L. Similo, W. F. Grgurich, P. A. Williams, J. H. Dauber, and B. P. Griffith. 1992. Sequelae of cytomegalovirus pulmonary infections in lung allograft recipients. Am. Rev. Respir. Dis. 146:1419-1425.
Esclatine, A., B. Taddeo, L. Evans, and B. Roizman. 2004. The herpes simplex virus 1 UL41 gene-dependent destabilization of cellular RNAs is selective and may be sequence-specific. Proc. Natl. Acad. Sci. USA 101:3603-3608.
Esclatine, A., B. Taddeo, and B. Roizman. 2004. Herpes simplex virus 1 induces cytoplasmic accumulation of TIA-1/TIAR and both synthesis and cytoplasmic accumulation of tristetraprolin, two cellular proteins that bind and destabilize AU-rich RNAs. J. Virol. 78:8582-8592.
Fan, X. C., and J. A. Steitz. 1998. Overexpression of HuR, a nuclear-cytoplasmic shuttling protein, increases the in vivo stability of ARE-containing mRNAs. EMBO J. 17:3448-3460.
Fortunato, E. A., A. K. McElroy, I. Sanchez, and D. H. Spector. 2000. Exploitation of cellular signaling and regulatory pathways by human cytomegalovirus. Trends Microbiol. 8:111-119.
Frevel, M. A., T. Bakheet, A. M. Silva, J. G. Hissong, K. S. Khabar, and B. R. Williams. 2003. p38 mitogen-activated protein kinase-dependent and -independent signaling of mRNA stability of AU-rich element-containing transcripts. Mol. Cell. Biol. 23:425-436.
Geist, L. J., and L. Y. Dai. 1996. Cytomegalovirus modulates interleukin-6 gene expression. Transplantation 62:653-658.
Hagemeier, C., S. M. Walker, P. J. Sissons, and J. H. Sinclair. 1992. The 72K IE1 and 80K IE2 proteins of human cytomegalovirus independently trans-activate the c-fos, c-myc and hsp70 promoters via basal promoter elements. J. Gen. Virol. 73:2385-2393.
Hildebrandt, A. L., and P. D. Neufer. 2000. Exercise attenuates the fasting-induced transcriptional activation of metabolic genes in skeletal muscle. Am. J. Physiol. Endocrinol. Metab. 278:E1078-E1086.
Hirsch, A. J., and T. Shenk. 1999. Human cytomegalovirus inhibits transcription of the CC chemokine MCP-1 gene. J. Virol. 73:404-410.
Humbert, M., R. M. Delattre, S. Fattal, B. Rain, J. Cerrina, P. Dartevelle, G. Simonneau, P. Duroux, P. Galanaud, and D. Emilie. 1993. In situ production of interleukin-6 within human lung allografts displaying rejection or cytomegalovirus pneumonia. Transplantation 56:623-627.
Iyer, V. R., M. B. Eisen, D. T. Ross, G. Schuler, T. Moore, J. C. Lee, J. M. Trent, L. M. Staudt, J. Hudson, Jr., M. S. Boguski, D. Lashkari, D. Shalon, D. Botstein, and P. O. Brown. 1999. The transcriptional program in the response of human fibroblasts to serum. Science 283:83-87.
Jones, S. A., and S. Rose-John. 2002. The role of soluble receptors in cytokine biology: the agonistic properties of the sIL-6R/IL-6 complex. Biochim. Biophys. Acta 1592:251-263.
Kotenko, S. V., S. Saccani, L. S. Izotova, O. V. Mirochnitchenko, and S. Pestka. 2000. Human cytomegalovirus harbors its own unique IL-10 homolog (cmvIL-10). Proc. Natl. Acad. Sci. USA 97:1695-1700.
Laroia, G., B. Sarkar, and R. J. Schneider. 2002. Ubiquitin-dependent mechanism regulates rapid turnover of AU-rich cytokine mRNAs. Proc. Natl. Acad. Sci. USA 99:1842-1846.
McSharry, B. P., P. Tomasec, M. L. Neale, and G. W. G. Wilkinson. 2003. The most abundantly transcribed human cytomegalovirus gene (? 2.7) is non-essential for growth in vitro. J. Gen. Virol. 84:2511-2516.
Michelson, S., J. Alcami, S.-J. Kim, D. Danielpour, F. Bachelerie, L. Picard, C. Bessia, C. Paya, and J.-L. Virelizier. 1994. Human cytomegalovirus infection induces transcription and secretion of transforming growth factor ?1. J. Virol. 68:5730-5737.
Mocarski, E. S., Jr. 2002. Immunomodulation by cytomegaloviruses: manipulative strategies beyond evasion. Trends Microbiol. 10:332-339.
Mukhopadhyay, D., C. W. Houchen, S. Kennedy, B. K. Dieckgraefe, and S. Anant. 2003. Coupled mRNA stabilization and translational silencing of cyclooxygenase-2 by a novel RNA binding protein, CUGBP2. Mol. Cell 11:113-126.
Myer, V., S. Lee, and J. Steitz. 1992. Viral small nuclear ribonucleoproteins bind a protein implicated in messenger RNA destabilization. Proc. Natl. Acad. Sci. USA 89:1296-1300.
Nabors, L. B., G. Y. Gillespie, L. Harkins, and P. H. King. 2001. HuR, a RNA stability factor, is expressed in malignant brain tumors and binds to adenine- and uridine-rich elements within the 3' untranslated regions of cytokine and angiogenic factor mRNAs. Cancer Res. 61:2154-2161.
Ng, S., Y. Tan, and G. Guy. 1994. Differential induction of the interleukin-6 gene by tumor necrosis factor and interleukin-1. J. Biol. Chem. 269:19021-19027.
Pasare, C., and R. Medzhitov. 2003. Toll pathway-dependent blockade of CD4+CD25+ T cell-mediated suppression by dendritic cells. Science 299:1033-1036.
Pass, R. F. 2001. Cytomegalovirus, p. 2675-2705. In D. M. Knipe and P. M. Howley (ed.), Fields virology, 4th ed., vol. 2. Lippincott Williams & Wilkins, Philadelphia, Pa.
Penfold, M. E., D. J. Dairaghi, G. M. Duke, N. Saederup, E. S. Mocarski, G. W. Kemble, and T. J. Schall. 1999. Cytomegalovirus encodes a potent alpha chemokine. Proc. Natl. Acad. Sci. USA 96:9839-9844.
Pioli, P. A., B. J. Hamilton, J. E. Connolly, G. Brewer, and W. F. Rigby. 2002. Lactate dehydrogenase is an AU-rich element-binding protein that directly interacts with AUF1. J. Biol. Chem. 277:35738-35745.
Ploegh, H. L. 1998. Viral strategies of immune evasion. Science 280:248-253.
Richards, C. D. 1998. Interleukin-6, p. 87-108. In A. Mire-Sluis and R. Thorpe (ed.), Cytokines. Academic Press, San Diego, Calif.
Sengupta, S., B.-C. Jang, M.-T. Wu, J.-H. Paik, H. Furneaux, and T. Hla. 2003. The RNA-binding protein HuR regulates the expression of cyclooxygenase-2. J. Biol. Chem. 278:25227-25233.
Simmen, K. A., J. Singh, B. G. Luukkonen, M. Lopper, A. Bittner, N. E. Miller, M. R. Jackson, T. Compton, and K. Fruh. 2001. Global modulation of cellular transcription by human cytomegalovirus is initiated by viral glycoprotein B. Proc. Natl. Acad. Sci. USA 98:7140-7145.
Simmons, P., K. Kaushansky, and B. Torok-Storb. 1990. Mechanisms of cytomegalovirus-mediated myelosuppression: perturbation of stromal cell function versus direct infection of myeloid cells. Proc. Natl. Acad. Sci. USA 87:1386-1390.
Sinclair, J., and P. Sissons. 1996. Latent and persistent infections of monocytes and macrophages. Intervirology 39:293-301.
Spector, D. H. 1996. Activation and regulation of human cytomegalovirus early genes. Intervirology 39:361-377.
Spencer, J. V., K. M. Lockridge, P. A. Barry, G. Lin, M. Tsang, M. E. Penfold, and T. J. Schall. 2002. Potent immunosuppressive activities of cytomegalovirus-encoded interleukin-10. J. Virol. 76:1285-1292.
Stoecklin, G., M. Colombi, I. Raineri, S. Leuenberger, M. Mallaun, M. Schmidlin, B. Gross, M. Lu, T. Kitamura, and C. Moroni. 2002. Functional cloning of BRF1, a regulator of ARE-dependent mRNA turnover. EMBO J. 21:4709-4718.
Subbaramaiah, K., T. P. Marmo, D. A. Dixon, and A. J. Dannenberg. 2003. Regulation of cyclooxgenase-2 mRNA stability by taxanes: evidence for involvement of p38, MAPKAPK-2, and HuR. J. Biol. Chem. 278:37637-37647.
Tebo, J., S. Der, M. Frevel, K. S. A. Khabar, B. R. G. Williams, and T. A. Hamilton. 2003. Heterogeneity in control of mRNA stability by AU-rich elements. J. Biol. Chem. 278:12085-12093.
Torok-Storb, B., M. Boeckh, C. Hoy, W. Leisenring, D. Myerson, and T. Gooley. 1997. Association of specific cytomegalovirus genotypes with death from myelosuppression after marrow transplantation. Blood 90:2097-2102.
Torok-Storb, B., L. Bolles, M. Iwata, K. Doney, G. E. Sale, T. A. Gooley, and R. Storb. 2001. Increased prevalence of CMV gB3 in marrow of patients with aplastic anemia. Blood 98:891-892.
Vanden Berghe, W., S. Plaisance, E. Boone, K. De Bosscher, M. L. Schmitz, W. Fiers, and G. Haegeman. 1998. p38 and extracellular signal-regulated kinase mitogen-activated protein kinase pathways are required for nuclear factor-B p65 transactivation mediated by tumor necrosis factor. J. Biol. Chem. 273:3285-3290.
Vink, C., M. J. Smit, R. Leurs, and C. A. Bruggeman. 2001. The role of cytomegalovirus-encoded homologs of G protein-coupled receptors and chemokines in manipulation of and evasion from the immune system. J. Clin. Virol. 23:43-55.
Walker, S., C. Hagemeier, J. G. Sissons, and J. H. Sinclair. 1992. A 10-base-pair element of the human immunodeficiency virus type 1 long terminal repeat (LTR) is an absolute requirement for transactivation by the human cytomegalovirus 72-kilodalton IE1 protein but can be compensated for by other LTR regions in transactivation by the 80-kilodalton IE2 protein. J. Virol. 66:1543-1550.
Wang, E. C. Y., B. McSharry, C. Retiere, P. Tomasec, S. Williams, L. K. Borysiewicz, V. M. Braud, and G. W. G. Wilkinson. 2002. UL40-mediated NK evasion during productive infection with human cytomegalovirus. Proc. Natl. Acad. Sci. USA 99:7570-7575.
Yurochko, A. D., and E. S. Huang. 1999. Human cytomegalovirus binding to human monocytes induces immunoregulatory gene expression. J. Immunol. 162:4806-4816.
Zhu, H., J. P. Cong, G. Mamtora, T. Gingeras, and T. Shenk. 1998. Cellular gene expression altered by human cytomegalovirus: global monitoring with oligonucleotide arrays. Proc. Natl. Acad. Sci. USA 95:14470-14475.
Zhu, H., J. P. Cong, and T. Shenk. 1997. Use of differential display analysis to assess the effect of human cytomegalovirus infection on the accumulation of cellular RNAs: induction of interferon-responsive RNAs. Proc. Natl. Acad. Sci. USA 94:13985-13990.
Zhu, H., J.-P. Cong, D. Yu, W. A. Bresnahan, and T. E. Shenk. 2002. Inhibition of cyclooxygenase 2 blocks human cytomegalovirus replication. Proc. Natl. Acad. Sci. USA 99:3932-3937.(Claire Gealy, Marian Dens)