| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
and IL-4 Through mRNA Stabilization: Involvement of the RNA-Binding Protein HuR 1
* Division of Allergy and Clinical Immunology, Johns Hopkins University, Baltimore, MD 21224;
Department of Molecular Genetics and Microbiology, Duke University Medical Center, Durham, NC, 27710;
National Institute of Aging, National Institutes of Health, Baltimore, MD 21224; and
University of Texas Houston Health Science Center Medical School, Houston, TX 77030
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and IL-4, regulate its production in airway epithelial cells by increasing eotaxin mRNA stability. In experiments using the transcriptional inhibitor actinomycin D, eotaxin mRNA half-life was significantly prolonged by cell stimulation with TNF- or IL-4, with the combination of the two cytokines being the most effective in extending the mRNA half-life. Involvement of the eotaxin 3' untranslated region in the mRNA-stabilizing effect was tested by transient transfection of a construct expressing a chimeric transcript carrying a serum-inducible 
-globin reporter linked to the eotaxin 3' untranslated region. The half-life of the chimeric mRNA was markedly increased in cells stimulated with TNF- and IL-4. Evidence that the mRNA-stabilizing protein HuR participated in the cytokine effect was obtained: first, HuR presence in the cytoplasm, believed to be required for HuR-mediated mRNA stabilization, increased in both transformed (BEAS-2B cell line) and primary bronchial epithelial cells following treatment with TNF- and IL-4. Second, endogenous eotaxin mRNA was found to bind to HuR in vivo, as detected by immunoprecipitation of HuR-containing messenger ribonucleoprotein complexes followed by real-time RT-PCR analysis; such association increased after cell treatment with TNF- and IL-4. Third, overexpression of HuR in BEAS-2B cells significantly increased the expression of eotaxin mRNA and protein. Our findings implicate mRNA stabilization in the cytokine-mediated increase in eotaxin expression and strongly suggest a role for HuR in this effect. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The C-C chemokine eotaxin exemplifies the complex and essential role that members of this gene family play in human diseases. Through binding to the chemokine receptor CCR3, eotaxin acts as a potent and specific eosinophil chemoattractant and activating stimulus. This characterizes eotaxin as a key mediator in allergic diseases, in which eosinophilic infiltration is a hallmark (4). Indeed, eotaxin expression is increased in asthmatic subjects and after allergen challenge, as well as after respiratory viral infections, a well-known trigger of asthma exacerbations (5, 6, 7, 8, 9). The production of eotaxin correlates with lung eosinophilia in human and in mouse models of allergic inflammation as well as with impaired lung function in asthma (7, 10, 11). In overexpression models, eotaxin synergizes with IL-5 in causing eosinophilia in the lung and in the gut (12, 13). Eotaxin is dysregulated in atopic dermatitis and allergic rhinitis and is critical for the establishment of gastrointestinal allergic hypersensitivity (14, 15, 16). Eotaxin also appears to be involved in biological functions beyond leukocyte chemotaxis and activation, since it induces CCR3-mediated angiogenesis and alters epithelial gene expression (17, 18). Overexpression of eotaxin and the CCR3 receptor has been found in atherosclerosis and Hodgkin’s lymphoma (19, 20).
Mounting evidence suggests that different types of inflammatory reactions trigger specific chemokine pathways, which appear to be tightly regulated by the inflammatory signals present in the tissue environment (21, 22). In the context of allergic inflammation, several studies indicate that the Th2-derived cytokines IL-4 and IL-13 are strong inducers of the CCR3 ligands eotaxin (23, 24) and eotaxin-3 (25, 26, 27) as well as the CCR4 ligands thymus and activation-regulated chemokine and macrophage-derived chemokine (28, 29), especially when used in association with TNF-. In addition, IL-4 potently inhibits production of the CXC chemokine IL-8 (30). These data suggest that Th2 products, by shaping a specific pattern of chemokine expression, could further amplify the recruitment of CCR3- and CCR4-bearing leukocytes such as eosinophils and Th2-type lymphocytes, especially if additional broad proinflammatory stimuli such as TNF- are present, as it occurs in asthma (31). Th2 products would concomitantly reduce signals for neutrophil influx, therefore creating a selective cytokine/chemokine feedback loop that would chronically sustain the mucosal inflammation (21, 32).
Previous studies aimed at investigating the molecular mechanisms of eotaxin induction by IL-4 and TNF- indicated that eotaxin expression was transcriptionally up-regulated by transcription factors STAT6 and NF-
B (24). Further analysis revealed that the transcriptional up-regulation induced by the combined stimulation with TNF- and IL-4 was additive, as compared with the effect of each cytokine alone (24). By contrast, combined treatment with both cytokines was strongly synergistic, causing increases in eotaxin production that were up to 11-fold higher than those achieved after each single cytokine challenge (23). These data suggest the existence of additional posttranscriptional regulatory pathways controlling eotaxin production. Additional lines of evidence indicate that TNF- and IL-4 could regulate eotaxin production through stabilization of its mRNA. First, the eotaxin mRNA contains sequences in its 3' untranslated region (UTR)
4 known to regulate mRNA turnover such as two adjacent AUUUA pentamers surrounded by an AU-rich region (see Fig. 1). Furthermore, we previously showed that in airway epithelial cells, inhibition of eotaxin expression by glucocorticoids was due, at least in part, to acceleration of eotaxin mRNA decay through a 3'UTR-mediated mechanism (23, 33). Moreover, combined treatment with TNF- and IL-4 has been shown to affect the stability of mRNAs encoding other proinflammatory molecules such as VCAM-1 (34).
|
Although the molecular mechanisms of ARE-dependent mRNA turnover are not yet fully understood, studies in the last decade have led to the identification of several proteins that bind to AREs and either accelerate mRNA decay or increase mRNA stability (38). Among the proteins that enhance mRNA stability is HuR, an ubiquitously expressed member of the Hu family of RNA-binding proteins that is homologous to the Drosophila embryonic lethal abnormal vision protein family (50). It has been shown that HuR binds in vitro to AREs in mRNAs encoding key regulators of proliferation and stress responsiveness such as p21, cyclin A, cyclin B1, and c-fos, as well as critical mediators of inflammatory reactions such as TNF-, GM-CSF, cyclooxygenase 2, and vascular endothelial growth factor (50). Upon HuR binding, its targets display increased mRNA stability and increased translation of the mRNA transcripts (51, 52, 53). The mRNA-stabilizing function of HuR has been also documented using chimeric mRNAs that bear AREs (43).
Taken together, these studies prompted us to investigate whether TNF- and IL-4 treatment was capable of regulating eotaxin mRNA turnover. Our study focused on the expression of eotaxin by airway epithelial cells, which constitute a key source of several chemokines in the lung. In particular, epithelial cell-derived eotaxin has been shown to correlate with eosinophil influx in the airway mucosa and with bronchial hyperreactivity in asthmatics (54, 55).
Using the airway epithelial cell line BEAS-2B and primary bronchial epithelial cells (PBECs), we show that eotaxin mRNA is indeed stabilized following treatment with TNF- and IL-4 and we provide evidence that the eotaxin 3'UTR and the RNA-binding protein HuR are involved in this process. Importantly, treatment with TNF- and IL-4 induced the cytoplasmic localization of HuR, a process that is linked with activation of HuR function, and increased the association of eotaxin mRNA with HuR in vivo. These findings, along with the observation that HuR overexpression significantly elevates eotaxin production, strongly support the notion that production of eotaxin triggered by TNF- and IL-4 relies on the stabilization of the eotaxin mRNA and implicate the eotaxin 3'UTR and HuR in this process.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
PBECs were isolated by Pronase digestion from bronchi of cadaveric lungs as described previously (56). The purity of these preparations was routinely confirmed by immunohistochemical staining for cytokeratin (data not shown). PBECs were cultured on collagen-coated flasks and maintained in serum-free LHC-9 medium (Biofluids, Rockville, MD). The BEAS-2B cell line, derived from human tracheal epithelium transformed by an adenovirus 12-SV40 hybrid virus (57), was generously supplied by Dr. C. Harris (National Institutes of Health, Bethesda, MD) and was maintained in F12/DMEM (Invitrogen, Frederick, MD), containing 5% heat-inactivated FCS, 2 mM L-glutamine, penicillin (100 U/ml), and streptomycin (100 mg/ml). This medium is referred to as "complete medium". The NIH 3T3 mouse fibroblast cell line (American Type Culture Collection, Manassas, VA) was cultured in DMEM (Life Technologies/Invitrogen) supplemented with 10% heat-inactivated bovine calf serum. Cultures of PBECs were used during the first passage only, while BEAS-2B cells were used from passages 36 to 45. All cells were cultured at 37°C in humidified air containing 5% CO2.
Plasmid constructs and transfection
The pBBB plasmid has been previously described (58). The pBBB-eotaxin construct was generated by amplification of a BEAS-2B genomic template using BglII-tailed primers: AACCTCATTATCAGTCCA (forward), AGAGGAGAGGGGGGAACAT (reverse) (Bioserve Biotechnologies, Laurel, MD). After sequence verification, the PCR product encompassing the AU-rich region of the eotaxin 3'UTR from +426 to +647 of the eotaxin mRNA (accession number NM_002986) was subcloned into the unique BglII site of pBBB, and its correct orientation was verified by restriction enzyme digestion. G10 epitope-tagged HuR was generated as previously described (59), with the substitution of the full-length HuR open reading frame.
In transfection experiments, NIH 3T3 or BEAS-2B cells were seeded in six-well plates in their respective medium at a density of 
250,000/well and transfected 24 h later, when cells reached 50–60% confluence, using the nonliposomal cationic vehicle FuGene (Roche Applied Science, Indianapolis, IN) according to the manufacturer’s specifications. Briefly, 3 µl of FuGene/sample was resuspended (5 min at room temperature) in serum-free Opti-MEM (Invitrogen) and allowed to complex with 1 µg of plasmid DNA for 15 min at room temperature. The plasmid/FuGene mixture was then overlayed on the cells in a final volume of 2 ml of complete medium. After incubation for 24 h at 37°C, cells were treated according to the experimental protocols. The efficiency of the transient transfection of BEAS-2B cells, assessed with a CMV promoter-driven expression vector coding for green fluorescence protein (Promega, Madison, WI), was 22% of total cell count performed by nuclear staining with Hoechst dye.
Immunofluorescence
BEAS-2B cells and PBEC were seeded at 40–50% confluency on sterile coverslips and incubated in complete medium overnight. Following treatment, cells were fixed for 15 min in PBS containing 4% paraformaldehyde for a period of 15 min, followed by permeabilization in PBS containing 0.4% TritonX-100 for an additional 15 min. After incubation in blocking buffer (PBS containing 2% BSA and 0.1% Tween 20) for 16 h, coverslips were incubated in a 1/500 dilution of mouse anti-HuR Ab (Santa Cruz Biotechnology, Santa Cruz, CA) in blocking buffer for 1 h. Following washes in blocking buffer, samples were incubated with a mixture of horse anti-mouse Texas Red (1/200; The Jackson Laboratory, Bar Harbor, ME) and Hoechst 33342 (1/5,000; Molecular Probes, Eugene, OR) for 1 h. After washes with blocking buffer, coverslips were mounted in Vectashield (Vector Laboratories, Burlingame, CA). An Axiovert 200 M microscope (x63 lens; Zeiss, Oberkochen, Germany) was used for visualization using separate channels for the analysis of phase-contrast images, red fluorescence, and blue fluorescence. Images were then processed with the Zeiss AxioVision program, version 3.0.
RNA isolation and analysis
Total RNA was extracted using the TRIzol reagent (Invitrogen) as described elsewhere (60). Cytoplasmic RNA was isolated by the RNAeasy kit (Qiagen, Valencia, CA). RNA concentration was measured spectrophotometrically and reading was validated by comparison of the intensity of ethidium bromide-stained 28S and 18S RNA bands in 1% agarose gels loaded with 1 µg of total RNA/condition. Northern blot analysis was conducted as previously described (23). Briefly, aliquots of 15 µg of total RNA (10 µg of cytoplasmic RNA for the experiments with NIH cells) were separated by electrophoresis on a 1% agarose/6% formaldehyde gel, then blotted onto a Genescreen Plus nylon membrane (PerkinElmer Life Sciences, Boston, MA). The membrane was prehybridized and subsequently hybridized in PerfectHyb Plus solution (Sigma-Aldrich, St. Louis, MO) with 1 x 106 cpm/ml cDNA probes labeled with [-32P]dATP by random priming (Promega). After stringent washing, membranes were visualized by autoradiography and the intensity of bands, which were recorded by a Kodak 120 Electrophoresis Documentation and Analysis System (Kodak, Rochester, NY), was quantified densitometrically using NIH Image software (Public Software). The DNA probes used were a 260-bp BamHI fragment of the eotaxin coding region (61), a commercial probe for the housekeeping gene GAPDH (Clontech Laboratories, Palo Alto, CA), and for the transcriptional pulsing analysis, cDNAs encompassing the full-length coding regions of rabbit -globin and rat -globin (58).
For real-time RT-PCR analysis, reverse transcription of total RNA from epithelial cells was performed using the components of the Gene Amp kit (PerkinElmer Life Sciences) and amplification of eotaxin mRNA was performed using the TaqMan reagent kit (PerkinElmer Life Sciences) following the manufacturer’s instructions for both procedures. The PCR primers for eotaxin were AGGAGAATCACCAGTGGCAAA (forward) and GGAATCCTGCACCCACTTCTT (reverse); the probe sequence was TCCCCAGAAAGCTGTGATCTTCAAGACC. The reaction yielded a 103-bp DNA fragment. Primers for the housekeeping gene GAPDH were GAAGGTGAAGGTCGGAGTC (forward) and GAAGATGGTGATGGGATTTC (reverse); the probe sequence was JOE-CAAGCTTCCCGTTCTCAGCC-TAMRA. The reaction yielded a 226-bp DNA fragment. The conditions for the reaction for both primer sets were: 50 °C, 2 min; 95 °C, 10 min (hot-start to activate the Taq polymerase), then 95 °C, 15 s; 60 °C, 60 s for 40 cycles. Resulting PCR products were measured and elaborated by the sequence detector ABI 7700 (PerkinElmer Biosystems). Primers were designed by Primer Express software (PerkinElmer). All samples were run in triplicate. For the detection of eotaxin mRNA after immunoprecipitation (IP) of messenger ribonucleoproteins (mRNPs), changes between untreated and cytokine-treated cells among different IP conditions were based on the shift of threshold cycle (CT), the fractional cycle number at which the amplified target reaches a significant threshold. The higher the starting copy number of the target, the sooner a significant increase in signal will be observed as lower CT number. For detection of eotaxin mRNA in cells overexpressing HuR, quantitation of gene expression by real-time RT-PCR was evaluated using the Comparative CT Method as per PerkinElmer’s guidelines. Briefly, eotaxin gene expression was expressed as a function of cell treatment. For each experiment, the "calibrator" sample was the cDNA from untreated cells. All quantitations were also normalized to an endogenous control, the housekeeping gene GAPDH, to account for variability in the initial concentration of RNA and in the conversion efficiency of the reverse transcription reaction. The analysis of the relative quantitation required calculations based on the CT as follows: 1) 
CT, the difference between the mean CT values of the samples evaluated with eotaxin-specific primers and those of the same samples evaluated with GAPDH-specific primers; 2) CT, the difference between the CT values of the samples and the CT value of the calibrator sample; and 3) 2-CT, which yields the relative mRNA units representing the fold induction over the control. The same quantitation method was used to calculate eotaxin mRNA decay after actinomycin D (Act D) treatment in cells overexpressing HuR.
Western blot analysis and IP of endogenous mRNP complexes
For Western blot, cells were harvested by trypsinization and lysed for 5 min on ice in 10 mM Tris (pH 7.4), 150 mM NaCl, 5 mM EDTA, 10% (v/v) glycerol, 1% Triton X-100, 1 mM PMSF, 10 mg/ml aprotinin, 1 mM sodium orthovanadate, 1 mM leupeptin, and 1 mM pepstatin A. Insoluble cell debris were removed by centrifugation (14,000 rpm, 5 min), and cleared cell lysates were boiled for 5 min in SDS sample buffer (2% SDS, 50 mM Tris (pH 6.8), 100 mM DTT, 0.1% bromphenol blue, and 10% glycerol). Total protein was measured by the Bradford assay, then separated (10 µg/lane) by electrophoresis on a 12% SDS-polyacrylamide minigel. Following transfer of samples, membranes (polyvinylidene difluoride;, Bio-Rad, Hercules, CA) were blocked in PBS containing 4% BSA and 0.1% Tween 20 for 1 h at room temperature with continuous shaking, then incubated with 2 µg/ml mouse monoclonal anti-HuR (Santa Cruz Biotechnology) or 0.3 µg/ml mouse monoclonal anti--actin (Santa Cruz Biotechnology) in 1x PBS containing 1% Tween for 1 h at room temperature with continuous shaking. After washes with 1x PBS containing 0.1% Tween 20 (10 min, room temperature), membranes were incubated for 1 h at room temperature with a HRP-conjugated goat anti-mouse secondary Ab. After a final washing step (1x PBS containing 0.1% Tween 20, 10 min, room temperature), immunoreactive bands were visualized by ECL (Amersham, Arlington Heights, IL).
For experiments aimed at detecting eotaxin mRNA in mRNPs, cells were treated using a modification of a described method (62). Briefly, cells were harvested, counted, and the same number of cells per condition (30 x 106) was pelleted and resuspended in approximately two cell pellet volumes of polysome lysis buffer containing 100 mM KCl, 5 mM MgCl2, 10 mM HEPES (pH 7.0), 0.5% Nonidet P-40 with 1 mM DTT, 100 U/ml RNaseOUT (Life Technologies, Rockville, MD), 0.2% vanadyl-ribonucleoside complex (Life Technologies), 0.2 mM PMSF, 1 mg/ml pepstatin A, 5 mg/ml bestatin, and 20 mg/ml leupeptin added fresh immediately before use. Cell lysates were then frozen at -80°C for storage. At the time of use, cell lysates were thawed and centrifuged at 16,000 x g for 10 min at 4°C. The mRNP cell lysate contained 30 mg/ml total protein (see regular BEAS-2B lysates). For IP, protein A-Sepharose beads (Sigma-Aldrich) were swollen 1:1 v/v in NT2 buffer (50 mM Tris, (pH 7.4), 150 mM NaCl, and 1 mM MgCl2/0.05% Nonidet P40) supplemented with 5% BSA. A 100-µl aliquot of the preswollen protein A bead slurry was used for IP reaction and incubated for 4 h at room temperature with excess immunoprecipitating Ab (30 µg), using either a mouse mAb specific for HuR, 3A2 (obtained from a hybridoma, a generous gift from Dr. J. Steitz, Yale University, New Haven, CT) (63) or an IgG1 isotype control Ab (B&D Life Sciences, San Diego, CA). The Ab-coated beads were washed with ice-cold NT2 buffer and resuspended in 900 µl of NT2 buffer supplemented with 100 U/ml RNaseOUT, 0.2% vanadyl-ribonucleoside complex, 1 mM DTT, and 20 mM EDTA. The beads were vortexed briefly, and 100 µl of the mRNP cell lysate was added and immediately centrifuged. A 100-µl aliquot was removed to represent total cellular RNA. The IP reactions were tumbled at room temperature for 2 h and then washed six times with ice-cold NT2 buffer. Washed beads were resuspended in 100 µl of NT2 buffer supplemented with 0.1% SDS and 30 µg of proteinase K, incubated for 30 min at 55°C, then extracted using phenol-chloroform-isoamylalcohol, and precipitated in ethanol.
Eotaxin protein measurement in cell supernatants
Secreted eotaxin was detected in the supernatant of BEAS-2B cells using a sensitive and specific ELISA (R&D Systems, Minneapolis, MN).
Statistical analysis
Data were analyzed using the ANOVA test with Fisher’s post hoc analysis. Comparisons among groups were further analyzed with the Kruskal-Wallis nonparametric test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and IL-4 induce stabilization of eotaxin mRNA in a 3'UTR-dependent fashion
The presence of a distinct ARE in 3'UTR of the eotaxin mRNA (Fig. 1) and earlier observations suggesting that transcriptional activation could not fully account for the strong increase in eotaxin production after treatment with TNF- and IL-4 (23, 24, 64) prompted us to evaluate whether treatment with TNF- and IL-4 stabilized the eotaxin mRNA. As the initial approach, we examined the mRNA decay using a method that employs the transcriptional inhibitor Act D. BEAS-2B cells and PBECs were cultured for 18 h with medium alone, TNF- (50 ng/ml), IL-4 (50 ng/ml), or the combination of TNF- plus IL-4 for 18 h. Cells were then either collected (time 0 control) or treated with 3 µg/ml Act D to block further transcription of mRNA and then harvested 4, 8, and 24 h thereafter. Eotaxin mRNA was detected by Northern blot analysis and its expression was normalized to the 18S rRNA and further compared with the GAPDH mRNA. The half-life of the eotaxin mRNA was then calculated as the time period necessary to reduce the amount of eotaxin mRNA in Act D-treated cells to 50% of the original eotaxin mRNA abundance at time 0. Densitometric analysis of these experiments (Fig. 2) shows that cytokine treatment significantly increased the stability of eotaxin mRNA. In BEAS-2B cells, the mRNA half-life increased from 2 h in unstimulated cells to 5 h after TNF- treatment and to 7 h after IL-4 treatment. The most striking increase in half-life was seen after challenge with the combination of the two cytokines, which prolonged the half-life of the mRNA up to 11 h (Fig. 2A). Similarly, in PBECs the addition of IL-4 markedly increased TNF--induced eotaxin mRNA stability (Fig. 2B). In these cells, eotaxin mRNA was undetectable in the untreated and IL-4 treatment groups. In neither cell system did cytokine treatment influence the decay of GAPDH mRNA (Fig. 2).
|
-globin as reporter gene. This gene is under the transcriptional control of the serum-inducible c-fos promoter, which has the unique ability of driving gene transcription only for a short period of time after serum stimulation. This transient expression scheme mimics the effect of transcriptional inhibitors without perturbing other cellular components. The BEAS-2B cell line could not be used for the transcriptional pulsing assay, as the immortalization by SV40 interfered with the proper shutoff of the c-fos promoter driving the expression of the transfected chimeric construct. Therefore, we conducted the experiments in NIH 3T3 cells, which have been extensively used for this assay (65). NIH 3T3 cells were transiently transfected with pBBB-eotaxin, serum-starved for 24 h, and subsequently pulsed with serum; no other treatments were added to the culture media during the initial 60 min after serum stimulation to exclude any influence of cytokines on the transcription of the reporter construct. Cells were then treated with cytokines, and cytoplasmic RNA was extracted 2, 4, and 8 h later and examined by Northern blot analysis using a -globin-specific probe. mRNA signals were normalized to those of an heterologous mRNA (rat -globin) expressed by a cotransfected control plasmid (pSV-GAPDH) that was not influenced by the experimental conditions (58). Equal loading of RNA among lanes, as well as RNA integrity, were monitored by ethidium bromide staining (data not shown). As demonstrated in Fig. 3, densitometric analysis of Northern blots showed that the half-life of -globin mRNA increased in all treatment groups. In agreement with the results obtained with Act D (Fig. 2), this effect was more pronounced in cells treated with the combination of TNF- plus IL-4, which resulted in a 2.4-fold increase of the reporter mRNA half-life.
|
The RNA-binding protein HuR has been shown to mediate increased mRNA stability and thereby increased expression of several early response and proinflammatory genes such as c-fos, GM-CSF, TNF-, and vascular endothelial growth factor (50), all of which possess AREs in the 3'UTR of their mRNAs. Based on these observations, we set out to investigate whether eotaxin mRNA, which bears a distinct ARE in its 3'UTR (Fig. 1), was a target of binding by HuR and whether HuR could thus be involved in the eotaxin mRNA stabilization induced by cytokine treatment. Since HuR-driven mRNA stabilization appears to require the translocation of HuR to the cytoplasm (43), we first evaluated whether cytokine treatment caused an increase in the levels of cytoplasmic HuR. Following treatment of BEAS-2B cells and PBECs with either medium, TNF-, IL-4, or TNF- plus IL-4 for 3 and 18 h, the subcellular localization of HuR was assessed by immunofluorescence (66). We chose to evaluate the subcellular distribution of HuR after 18 h of cytokine treatment because at this time cytokine-induced eotaxin mRNA abundance has reached steady-state levels (see Fig. 2) and, therefore, we reasoned that posttranscriptional modifications might be ongoing at that time. In keeping with earlier reports (43), the HuR signal was primarily nuclear in all treatment groups (Fig. 4); however, the combination of TNF- and IL-4 markedly increased cytoplasmic HuR abundance in both cell systems. In BEAS-2B cells this increase was readily detectable after 18 h of treatment (Fig. 4A), while in PBECs, cytoplasmic staining for HuR was prominent at 3 h following the combined cytokine treatment (Fig. 4B), and was also detectable in cells stimulated with either TNF- or IL-4 alone (data not shown). In both cell types, control hybridizations using the secondary Ab alone were negative (data not shown).
|
plus IL-4 treatment of epithelial cells induced eotaxin mRNA stabilization in a 3'UTR-dependent manner and that it caused HuR translocation to the cytoplasm, we next investigated whether such treatment might enhance in vivo binding of HuR to the endogenous eotaxin mRNA. We examined this possibility by IP of endogenous mRNPs using either a mouse mAb specific for HuR (3A2) or an isotype-matched Ab from lysates of BEAS-2B cells stimulated with or without TNF- plus IL-4 (62, 63). The immunoprecipitated mRNA pools were then assayed for eotaxin mRNA by real-time RT-PCR, as explained in Materials and Methods. No amplification of eotaxin mRNA was detected in the control Ig-immunoprecipitated samples (n = 3, 45 ± 0, mean ± SEM) and a very weak signal was present in only one of three HuR-immunoprecipitated samples (n = 3, 43.3 ± 1.67). In contrast, Fig. 5A shows that eotaxin mRNA was detectable in the RNA samples of TNF- plus IL-4-stimulated cells obtained by IP with anti-HuR. In three experiments (Fig. 5B), the average CT of samples obtained with IP with anti-HuR was 43 ± 2.2 in unstimulated cells and 32 ± 2 in cells treated with TNF- plus IL-4; the average CT of samples obtained with IP with an isotype-matched irrelevant Ab was 44 ± 0.6 in unstimulated cells and 39 ± 2 in cells treated with TNF- plus IL-4. The lower CT number obtained for the samples subjected to IP using the anti-HuR Ab indicates the presence of a higher starting copy number of the eotaxin mRNA. The lower CT number (i.e., the increased detection) obtained in cytokine-treated samples immunoprecipitated with the control Ab reflects an expected increase in the background binding of eotaxin mRNA to the IP beads. This effect is likely due to the sharp rise in the amount of eotaxin mRNA present in the extracts, as expression of eotaxin mRNA—which is minimal in unstimulated epithelial cells—is strongly up-regulated upon treatment with TNF- plus IL-4 (see also Fig. 2). However, the net shift of the mean CT over the background signal given by the irrelevant Ab was 1.7 in unstimulated cells and 7.1 in samples stimulated with TNF- plus IL-4. Therefore, stimulation resulted in a net cycle gain of 5.4, indicating a 25.4 = 42.2-fold enrichment in eotaxin mRNA over unstimulated samples in the anti-HuR-immunoprecipitated samples. The specificity of the anti-HuR IP was confirmed by Western blot analysis to monitor the HuR presence in the mRNPs (Fig. 5C).
|
and IL-4, alone or in combination, and cells were harvested for eotaxin mRNA and protein analysis. As shown in Fig. 6B, real-time RT-PCR measurement of eotaxin mRNA in BEAS-2B cells transiently overexpressing HuR revealed a significant increase in cells treated with TNF- alone (4.6 ± 0.9-fold over empty vector (EV) transfected, p < 0.0001) or in combination with IL-4 (2.3 ± 0.2-fold over EV transfected, p < 0.05) in comparison to the eotaxin levels induced by cytokine treatment in cells transfected with the EV control. To investigate whether the increase in steady-state levels of eotaxin mRNA was due to a stabilization of the mRNA, we cultured BEAS-2B cells overexpressing HuR in the presence of Act D according to the protocol used in the experiments shown in Fig. 2. In two real-time RT-PCR experiments, eotaxin mRNA abundance 24 h after treatment of mock- and HuR-transfected cells with Act D was 56 and 93% of that in untreated populations, respectively, indicating that overexpression of HuR increased the stability of cytokine-induced eotaxin mRNA.
|
plus IL-4 (3.1 ± 0.3-fold, p < 0.05).
In summary, our results indicate that treatment of airway epithelial cells with TNF- and IL-4 increases the stability of eotaxin mRNA and that the 3'UTR of this mRNA mediates the stabilizing effect. Furthermore, we demonstrate the involvement of HuR as a regulatory trans-acting factor in this process, since cytokine treatment induced HuR activation and binding to eotaxin mRNA and HuR overexpression significantly increased eotaxin mRNA and protein production.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and IL-4 significantly increase eotaxin mRNA stability. In fact, while in unstimulated cells the eotaxin mRNA is short-lived, with a half-life of 2 h, treatment with either TNF- or IL-4 induced up to a 3-fold extension of the eotaxin mRNA half-life. Combined treatment with the two cytokines showed even more dramatic mRNA stabilization, with a 5-fold increase in eotaxin mRNA half-life in the BEAS-2B cell line, a phenomenon that was also readily observable in primary airway epithelial cells. Changes in the stability of labile mRNAs are often in the range of 2- to 4-fold fluctuations (37). Although seemingly modest, mRNA half-life changes of this amplitude have been shown to result in >1000-fold differences in steady-state mRNA levels and correspondingly altered protein production (37). Considering that TNF- and IL-4 also activate transcription of the eotaxin gene (24), there are at least two converging mechanisms by which cytokines regulate eotaxin production. Therefore, the increased eotaxin mRNA stability demonstrated in the present investigation is likely to contribute crucially to the synergistic increase of eotaxin protein produced by epithelial cells upon treatment with this cytokine combination (23, 64). This is in keeping with mathematical models demonstrating that between two genes transcribed at the same rate, the one susceptible of regulation at the level of mRNA stability ultimately displays more abundant mRNA production (37).
Additional information on the mRNA regions involved in the mRNA-stabilizing effect of the cytokine treatment was obtained through use of a chimeric reporter gene created by subcloning the eotaxin 3'UTR into a reporter construct (65). Although at reduced scale, the results obtained with this model recapitulated those gained with Act D, as the combination of TNF- and IL-4 produced the highest degree of stabilization of the reporter mRNA. These data reveal the presence of functional regulatory regions within the eotaxin 3'UTR that are involved in the mRNA stabilization induced by cytokines. In ongoing studies, we are conducting a more detailed analysis of the eotaxin 3'UTR to identify the critical regions necessary for the stabilizing effect and to test, in particular, whether the AREs present in this region are essential in this process. Although we have identified the eotaxin 3'UTR as necessary in the stabilization of the eotaxin mRNA by cytokines, we cannot exclude the participation of more than one sequence in this process, similar to what has been demonstrated for other ARE-bearing mRNAs such as those encoding GM-CSF, IL-6, IL-2, or c-fos, where additional destabilizing elements that are structurally and functionally distinct from AREs were found in the 3'UTR as well as in the coding region and in the 5'UTR of the mRNA molecule (37, 69, 70).
Regarding the trans-acting factors that mediate eotaxin mRNA stabilization by cytokines, several lines of evidence provided in this study demonstrate the involvement of the RNA-binding protein HuR: 1) HuR, which is localized predominantly in the nucleus, translocates to the cytoplasm upon cytokine challenge in both PBECs and BEAS-2B cells; 2) endogenous cytoplasmic HuR and eotaxin mRNA form mRNP complexes in vivo and their abundance increases following treatment with TNF- plus IL-4; and 3) HuR-overexpressing cells exhibit increased levels of eotaxin mRNA and protein.
Export of HuR from the nucleus, where it is present in high concentration, to the cytoplasm appears to be essential for HuR-mediated mRNA stabilization (43). Although much still needs to be learned about this process, it is currently believed that after binding to its target mRNAs in the nucleus, the HuR-mRNA complex is transported to the cytoplasm where HuR protects the bound mRNA from degradation (43). Mounting experimental evidence suggests that HuR cytoplasmic localization is essential for its mRNA-stabilizing function, like reports showing that an increase in cytoplasmic HuR following cell challenge with several stressors, like UV light or heat shock, correlated with the increased stability of multiple HuR-target labile mRNAs (44, 66). In keeping with this model, our report indicates that cytokines can also regulate HuR subcellular distribution, thus implicating HuR-mediated processes in their ability to stabilize the eotaxin mRNA and, possibly, also other labile, ARE-bearing mRNAs encoding additional chemokines or inflammatory mediators. The changes in subcellular distribution of HuR observed are occurring within a time frame in which eotaxin mRNA is strongly increased after cytokine challenge and is likely to be subject to active posttranscriptional regulation. Interestingly, cytoplasmic staining for HuR was noticeable in primary cells treated with TNF- and IL-4 for 3 h after stimulation. Since the expression of eotaxin mRNA induced by TNF- and IL-4 is already almost maximal after 3 h (data not shown), it is likely that alteration of mRNA turnover regulates eotaxin expression from earlier time points.
Our data showing that HuR binds to eotaxin mRNA in vivo and that binding increases after cytokine challenge provide strong support to the hypothesis that HuR is involved in the stabilizing effect of TNF- and IL-4. It is important to emphasize the strength of the IP assay used to demonstrate the in vivo association of eotaxin mRNA to HuR. In this assay, the presence of endogenous eotaxin mRNA is detected within a mRNA pool that has been obtained by IP using a specific anti-HuR Ab when the mRNAs are still bound in their original conformation to HuR and possibly to other HuR-associated proteins. Therefore, this assay reveals that the eotaxin mRNA in its native conformation indeed interacts with HuR. In other words, this approach allows the investigator to assess the true complexes that actually form in the cell, as it takes into consideration the secondary and tertiary structures that allow native RNA molecules to effectively associate with target RNA-binding protein (71). A variety of in vitro-binding assays developed over the years to examine mRNP formation typically use partial transcripts that may or may not resemble the site of mRNA recognition by the RNA-binding protein in the cell. Therefore, these in vitro assays may not recapitulate the mRNP conformation existing in the cell as faithfully as the in vivo IP assay does. Thus, we have provided compelling evidence of HuR association to eotaxin mRNA after treatment with cytokines.
Furthermore, the significant increase of eotaxin mRNA and protein production after cytokine stimulation that we observed in HuR-overexpressing epithelial cells provides strong additional evidence of a biological outcome for the postulated interaction, in line with other reports on increased expression of HuR targets in cells overexpressing HuR (67, 72). The increased stability of eotaxin mRNA in cells overexpressing HuR suggests that HuR is involved in the cytokine-induced increase of the eotaxin mRNA stability as an mRNA-stabilizing factor; however, since mRNA turnover and translation are considered to be linked (38), it remains to be explored whether eotaxin production is further regulated by cytokines at a translational level and whether HuR plays any role in this process, as it has been shown for other HuR targets (73).
The identification of the eotaxin mRNA as a novel target for binding and stabilization by HuR opens the way for additional studies on the control of mRNA turnover and other mechanisms of posttranscriptional regulation of chemokine genes. The methodology used in our single-gene analysis to identify HuR association with eotaxin mRNA in vivo was previously used in other cell systems for the global identification of entire mRNA subsets bound to a given RNA-binding protein (59). In neuronal cells, Tenenbaum et al. (59) identified different mRNA subsets that were targets of specific mRNA-binding proteins (HuB, eIF-4E, and PABP), and found that target mRNA subsets varied in response to cellular changes. These findings led to the hypothesis that the coordinate expression of genes that are structurally, and possibly functionally, related may be achieved also through an orchestrated posttranscriptional regulation via binding of groups of mRNAs to a common RNA-binding factor (74). This analysis, termed "ribonomics," seeks to identify the subsets of mRNAs that associate with a given RNA-binding protein under specific conditions. We are currently adopting this powerful approach to identify the subset of mRNAs with which HuR associates in response to cytokine stimulation using large-scale cDNA arrays. We are particularly interested in examining whether mRNAs encoding other chemokines, which are themselves structurally and functionally related, are joint targets of HuR-mediated stabilization in response to cytokine treatment in airway epithelial cells.
The precise mechanisms whereby cytokine treatment induces the cytoplasmic localization of HuR and its enhanced association with eotaxin mRNA remain unclear. Increased activity of the AMP-activated kinase (AMPK) has been found to inhibit HuR transport to the cytoplasm while, conversely, inhibition of AMPK has been shown to induce the cytoplasmic localization of HuR (66). It will be of interest to investigate whether cytokine treatment blocks AMPK function, thereby leading to increased HuR abundance in the cytoplasm and heightened association with eotaxin mRNA.
The identification of critical regulatory pathways of chemokine expression has been actively pursued to design pharmacological interventions aimed at disrupting the inflammatory cascade that the chemokine network sustains (3). Our study reveals that mRNA stability is an important regulatory mechanism in eotaxin generation by epithelial cells and underscores its potential for future development of therapeutic approaches. Finally, broadening our knowledge on mRNA-binding factors involved in these processes and their modulation by cytokines or other inflammatory signals, as reported here, could lead to the identification of posttranscriptional regulatory pathways possibly shared by multiple chemokine genes and potentially misregulated in allergic diseases.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 U.A. and S.L.C. contributed equally to the work presented in this manuscript.
3 Address correspondence and reprint requests to Dr. Cristiana Stellato, Johns Hopkins Asthma and Allergy Center, 5501 Hopkins Bayview Circle, Room. 1A.12A, Baltimore, MD 21224. E-mail address: stellato{at}jhmi.edu
4 Abbreviations used in this paper: UTR, untranslated region; Act D, actinomycin D; ARE, adenylate- and uridylate-rich element; IP, immunoprecipitation; mRNP, messenger ribonucleoprotein; PBEC, primary bronchial epithelial cell; CT, threshold cycle; EV, empty vector; AMPK, AMP-activated kinase.
Received for publication March 17, 2003. Accepted for publication August 13, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B and STAT6 in human airway epithelial cells. J. Immunol. 163:6876.
on macrophage-derived chemokine production: an amplification circuit of polarized T helper 2 responses. Blood 92:2668.: the role of this multifunctional cytokine in asthma. Immunol. Cell Biol. 79:132.[Medline]
in combination with either IL-4 or IL-13 through increased mRNA stability. Am. J. Pathol. 154:1149. plus fibronectin. J. Immunol. 166:4658.
, IgG). Note that following IP with anti-HuR, in this experiment the average CT (see Materials and Methods) decreased by 3.6 cycles in comparison to the CT value obtained after IP with the IgG control. The lower CT number obtained for the samples subjected to IP using the anti-HuR Ab indicates the presence of a higher starting copy number of the eotaxin mRNA. Lower panel, Visualization of eotaxin mRNA PCR product detected in cell extracts subjected to IP using the anti-HuR Ab in BEAS-2B cells either untreated or treated with TNF-
) or the IgG control Ab (
) in the unstimulated and TNF-
) or the EV control (pRL-CMV,