Glucocorticoid-induced leucine zipper is downregulated in human alveolar macrophages upon Toll-like receptor activation

Authors


Abstract

Induction of the glucocorticoid-induced leucine zipper (GILZ) by glucocorticoids plays a role in their antiinflammatory action, whereas GILZ expression is reduced under inflammatory conditions. The mechanisms regulating GILZ expression during inflammation, however, have not yet been characterized. Here, we investigated GILZ expression in human alveolar macrophages (AMs) following Toll-like receptor (TLR) activation. Macrophages were shown to predominantly express GILZ transcript variant 2. Lipopolysaccharide-treated AMs, THP-1 cells, and lungs of lipopolysaccharide-exposed mice displayed decreased GILZ protein and mRNA levels. The effect was strictly dependent on the adapter molecule MyD88, as shown by using specific ligands or a knockdown strategy. Investigations on the functional significance of GILZ downregulation performed by GILZ knockdown revealed a proinflammatory response, as indicated by increased cytokine expression and NF-κB activity. We found that TLR activation reduced GILZ mRNA stability, which was mediated via the GILZ 3′-untranslated region. Finally, involvement of the mRNA-binding protein tristetraprolin (TTP) is suggested, since TTP overexpression or knockdown modulated GILZ expression and TTP was induced in a MyD88-dependent fashion. Taken together, our data show a MyD88- and TTP-dependent GILZ downreg-ulation in human macrophages upon TLR activation. Suppression of GILZ is mediated by mRNA destabilization, which might represent a regulatory mechanism in macrophage activation.

Introduction

Inflammatory processes in the lung play an important role in different diseases, such as viral and bacterial infections. Moreover, noninfectious diseases such as chronic obstructive pulmonary disease (COPD) are characterized by chronic inflammatory processes. Asthma and allergic diseases are other examples of chronic inflammatory lung diseases with major clinical relevance [[1, 2]].

Toll-like receptors (TLRs) represent an important family of innate immune receptors, which act as a first line of defense of host immunity against various pathogens. Presently, ten human TLRs are known, which recognize pathogen-associated molecular patterns including bacterial cell wall components, such as lipoproteins (TLR1/2 or TLR1/6 dimers) or lipopolysaccharide (LPS, TLR4), bacterial flagellin (TLR5), viral RNA (TLR3, 7 and 8), as well as bacterial DNA (TLR9) [[3]]. Except for TLR3, all TLRs signal via the adapter molecule MyD88. The MyD88-independent pathway used by TLR3 can also be utilized by TLR4. The different signaling pathways employ different protein kinase complexes and show distinct cytokine profiles, but all activate NF-κB [[4]]. In addition to their role in pathogen defense, TLRs also play a dominant role in innate responses to noninfectious immunostimulatory components present in our environment [[2, 5-7]], making them important mediators in atopic reactions.

Glucocorticoid-induced leucine zipper (GILZ) was first described in 1997 as a glucocorticoid-inducible gene [[8]]. Glucocorticoids are widely used for the treatment of inflammatory pathologies, including inflammatory lung diseases, such as asthma. The induction of GILZ has been suggested to contribute to the antiinflammatory actions of glucocorticoids in different cells, such as T cells, mast cells, and epithelial cells [[9-13]]. The binding to and inhibition of NF-κB by GILZ seems essential for its antiinflammatory action [[9, 10, 14, 15]].

Despite the well-studied role of GILZ upon glucocorticoid treatment [[16]], rather little is known about regulation of GILZ during inflammatory cell activation. Recently, Eddleston et al. reported that cytokines can reduce GILZ mRNA levels in epithelial cells [[10]]. Furthermore, GILZ was reported to be downregulated or even absent in inflammatory diseases such as Crohn's disease, tuberculosis, or chronic rhinosinusitis [[14, 17]]. These findings suggest an active suppression of GILZ in inflammation.

The highest expression of GILZ is found in the lung, and macrophages have been reported to represent the major source of human GILZ [[14, 18]]. This suggests that GILZ plays a pivotal role in respiratory immune homeostasis. Therefore, we aimed at understanding mechanisms regulating GILZ expression in human alveolar macrophages (AMs) under inflammatory conditions, that is, upon TLR activation.

Results

GILZ downregulation upon TLR4 activation in AMs and in vivo

As limited data are available in the literature on expression and regulation of GILZ in AMs, we determined its constitutive expression in these cells. GILZ mRNA expression was determined by quantitative RT-PCR in human in vitro differentiated macrophages as well as THP-1 monocytic cells as positive controls. Additionally, we examined GILZ expression in interstitial macrophages (IMs), which represent another pivotal lung macrophage cell population [[5, 19]]. GILZ mRNA levels were threefold higher in both AMs and IMs compared with blood monocyte-derived macrophages as well as differentiated or nondifferentiated THP-1 cells (Fig. 1A). As the primers used for real-time RT-PCR by us and others [[10]] are not specific for a single transcript variant, we analyzed the expression of GILZ transcript variants in AMs by semiquantitative RT-PCR. In addition, we used four other human cell types, that is, differentiated THP-1 macrophages, in vitro differentiated macrophages, human umbilical vein endothelial cells (HUVECs), and human lung epithelial cells (16HBE14o). Transcript variant 1 specific primers revealed a weak signal in HUVECs, whereas the other cell types barely showed any amplification product (Fig. 1B). Transcript variant 3 specific primers amplified products for all cell types tested. However, employing primers specific for transcript variant 2 gave by far the strongest signal, indicating that GILZ2 is the major GILZ isoform in all cell types investigated.

Figure 1.

Constitutive expression and downregulation of GILZ by LPS. (A) RNA was isolated from AMs, IMs, in vitro differentiated macrophages (MΦs), THP-1 cells and PMA-differentiated, THP-1 macrophage-like cells (THP-1 MΦs), and real-time RT-PCR analysis for GILZ was performed. Data were normalized to β-actin. Data show means + SEM of data pooled from at least five independent experiments. Primary cells were from at least five different donors. (B) GILZ transcript variants (GILZ 1–3) were detected by semiquantitative RT-PCR in cDNA obtained from different cell types. β-Actin served as a loading control. One representative experiment of three is shown. (C and D) AMs were treated (C) with LPS for 2 h at the indicated concentrations or (D) with 100 ng/mL LPS for the indicated time points followed by real-time RT-PCR analysis for GILZ and TNF-α. Data are normalized to β-actin and represent means ± SEM of data pooled from four independent experiments with cells from four different donors. *p < 0.05 compared with GILZ in untreated controls, #p < 0.05 compared with TNF-α in untreated controls, Student's t-test. (E) AMs were treated with LPS (100 ng/mL) for the indicated times. Western blot analysis was performed using tubulin as a loading control. One representative blot is shown. GILZ signal intensities were quantified and normalized to tubulin values (bottom). Data are expressed as percentage of untreated control (Co) and show means + SEM of data pooled from three independent experiments with cells from three different donors. (F and G) In vitro differentiated macrophages were treated with CHX (5 μg/mL) for the indicated times followed by western blot analysis for GILZ using tubulin as a loading control. (F) GILZ signal intensities were quantified and are expressed as percentage of 0 h values. Data represent means ± SEM of values obtained from independent experiments with cells from three different donors. (G) One experiment representative of three independent experiments is shown.

We then examined expression of GILZ in AMs by an inflammatory stimulus, that is, upon TLR4 activation. GILZ was quickly downregulated after LPS addition. This was accompanied by activation of AMs, as indicated by a significant induction of TNF-α mRNA upon LPS challenge (Fig. 1C and D). GILZ mRNA levels were already downregulated to 20% of untreated controls at an LPS concentration of 10 ng/mL and were not further decreased by increasing LPS concentrations (Fig. 1C). Maximal LPS-mediated reduction of GILZ mRNA occurred at 2 h (Fig. 1D).

Next, we tested whether GILZ protein was reduced as well. A protein band corresponding to GILZ isoform 2 with a molecular weight of ∼17 kDa was readily detectable by western blot in AM lysates. LPS reduced GILZ protein expression that became evident already after 4 h (Fig. 1E). Similar results were obtained with in vitro differentiated macrophages as well as the THP-1 cell line (data not shown). To investigate whether a transient downregulation of GILZ mRNA was likely to affect GILZ protein levels, we determined GILZ protein half life in in vitro differentiated macrophages using cycloheximide (CHX) (Fig. 1F and G). Densitometric analysis (Fig. 1F) showed that GILZ had a protein half life of about 2 h, suggesting that GILZ protein can be downregulated rapidly once its mRNA levels decrease.

We then sought to check for an in vivo relevance for these findings. GILZ protein as well as mRNA levels were significantly decreased in lungs of LPS-treated mice compared with control animals (Fig. 2A–C), which corroborated our in vitro findings. As reported previously [[20]], LPS-mediated pulmonary inflammation was confirmed by induction of TNF-α mRNA in whole lung lysates (Fig. 2D).

Figure 2.

(A–D) In vivo downregulation of GILZ by LPS. C57BL/6 mice were injected intraperitoneally with 50 μg LPS diluted in physiological salt solution (n = 5) or an equal volume of salt solution only (Co, n = 4). After 4 h of injection, mice were sacrificed and the lungs were removed to determine GILZ levels. (A) Lung tissue was lysed and equal protein amounts were assessed by western blot analysis using tubulin as a loading control. One representative blot out of three is shown. (B) Relative GILZ signal intensities were quantified and normalized to tubulin values. Data are expressed as percentage of Co values and show means + SEM. *p < 0.05 compared with Co. (C and D) RNA was isolated from lung tissue and real-time RT-PCR analysis for (C) GILZ and (D) TNF-α was performed. Data were normalized to 18S and show means + SEM. *p < 0.05 compared with Co. (E–J) GILZ downregulation amplifies an inflammatory response. (E–G) In vitro differentiated macrophages were transfected with either control siRNA (siCo) or GILZ siRNA (siGILZ) and an NF-κB-driven luciferase reporter construct. (E) Cells were harvested 20 h posttransfection. Western blot analysis for GILZ was performed using tubulin as a loading control. One representative blot is shown. Values indicate GILZ signal intensities + SEM from four independent experiments with cells from different donors after normalization to tubulin values. (F and G) NF-κB activity in siGILZ versus siCo cells. Data show means + SEM of n = 30 replicates pooled from five independent experiments with cells from different donors, each performed with n = 6. (F) Baseline luciferase activity 20-h posttransfection. RLU: relative luminescence units; n.s.: not significant (p > 0.05). (G) Cells were either left untreated or incubated with LPS (100 ng/mL, 4 h) 16 h after transfection. LPS-induced NF-κB activity (fold induction compared with equally transfected untreated cells) is expressed as percentage of LPS-treated siCo transfected cells. *p < 0.05 (H–J) In vitro differentiated macrophages were transfected with siCo or siGILZ for 19 h and either left untreated or treated with LPS (100 ng/mL) for an additional hour. (H) GILZ, (I) TNF-α, and (J) IL-8 expression were measured by real-time RT-PCR. Data are normalized to β-actin and represent means + SEM of n = 3 replicates from three independent experiments, *p < 0.05. p-Values were calculated by Student's t-test.

To determine the functional relevance of GILZ downregulation, we used siRNA technology to knock down its expression. As shown in Fig. 2E and H, transfection of in vitro differentiated macrophages with GILZ siRNA reduced its mRNA and protein expression compared with GILZ levels seen in control siRNA transfected cells. We then assessed whether downreg-ulation of GILZ could induce NF-κB activation, which was investigated by using a luciferase NF-κB reporter construct. We observed no differences in baseline NF-κB activity between siGILZ- and control siRNA-transfected macrophages (Fig. 2F). However, GILZ knockdown resulted in an increase of LPS-induced NF-κB activity (Fig. 2G) and also in an amplification of LPS-mediated TNF-α and IL-8 mRNA induction (Fig. 2I and J), suggesting that the absence of GILZ drives a proinflammatory response.

MyD88-dependent GILZ downregulation

Reduced GILZ expression following TLR4 activation suggested that cell activation via TLRs generally decreased GILZ levels. We therefore employed other TLR ligands and determined GILZ mRNA levels.

All TLR ligands mimicking bacterial or viral molecular patterns, that is, Pam3CSK4 (TLR1/2 ligand), flagellin (TLR5 ligand), FSL-1 (TLR2/6 ligand), imiquimod (TLR7 ligand), ssRNA40 (TLR8 ligand), and bacterial DNA (TLR9 ligand), reduced GILZ mRNA levels in AMs (Fig. 3A). GILZ downregulation was paralleled by a marked activation of cells, as assessed by TNF-α induction (Fig. 3B).

Figure 3.

MyD88-dependent GILZ downregulation. (A and B) AMs were left untreated (Co) or treated with either Pam3CSK4 (250 ng/mL), flagellin (250 ng/mL), FSL-1 (250 ng/mL), imiquimod (7.5 μg/mL), ssRNA40 (7.5 μg/mL), or bacterial DNA (20 μg/mL), followed by real-time RT-PCR analysis for (A) GILZ or (B) TNF-α. Data are normalized to β-actin and represent means + SEM of data pooled from four independent experiments performed with cells derived from four different donors. *p < 0.05 compared with Co. (C and D) AMs were left untreated (Co) or (C) treated with Poly(I:C) for 2 h at the indicated concentrations or (D) with 10 μg/mL Poly(I:C) for the indicated time points followed by real-time RT-PCR analysis for GILZ and TNF-α. Data are normalized to β-actin and represent means ± SEM of data pooled from six independent experiments with cells obtained from six different donors. *p < 0.05 compared with 0 h values. (E) PMA-differentiated THP-1 cells were transfected with either siCo or MyD88 siRNA (siMyD88) or left untreated (Co) for 72 h, followed by immunoblotting for MyD88 using tubulin as a loading control. Data show one representative experiment. Relative MyD88 signal intensities were quantified and normalized to tubulin values. Data are expressed as percentage of Co values and show means + SEM of four independent experiments. (F) siMyD88, siCo, or Co cells were treated with LPS (100 ng/mL) or Pam3CSK4 (100 ng/mL) for 2 h, and GILZ mRNA was assessed by real-time RT-PCR. Data are normalized to β-actin and represent means + SEM of three independent experiments performed in triplicate, *p < 0.05. p-Values were generated by Student's t-test.

Treatment of AMs with the synthetic analogue of double-stranded RNA and TLR3 ligand Poly(I:C) also increased TNF-α mRNA levels. TNF-α induction was not due to activation of cytosolic Poly(I:C)-responsive elements [[21]], as inhibition of TLR3-dependent signaling by chloroquine completely abrogated cell activation (data not shown). Interestingly, Poly(I:C) did not alter GILZ mRNA expression at any of the concentrations tested (Fig. 3C). As the lack of effect of Poly(I:C) on GILZ mRNA might be due to a different time course transmitted via the MyD88-independent pathway compared with the other TLR ligands utilizing MyD88, we performed time-course studies with the TLR3 ligand. However, no decrease of GILZ mRNA levels could be observed at any time point, although cells were highly activated as suggested by the marked induction of TNF-α after treatment (Fig. 3D).

These findings indicated that downregulation of GILZ mRNA requires the adapter molecule MyD88. To specify the role of MyD88, we performed an siRNA-mediated MyD88 knockdown (Fig. 3E) and checked the ability of LPS and Pam3CSK4 to decrease GILZ mRNA levels in PMA-differentiated THP-1 cells. Treatment of control siRNA cells with LPS or Pam3CSK4 reduced GILZ mRNA levels to an extent comparable with those detected in untransfected control cells. This effect was significantly abrogated in MyD88 siRNA cells (Fig. 3F). These data demonstrate that GILZ mRNA downregulation upon TLR activation requires the adapter molecule MyD88.

Posttranscriptional regulation of GILZ

The observed downregulation of GILZ mRNA might either occur via transcription or mRNA stability. Since effects on GILZ mRNA were fast, we hypothesized destabilization of GILZ mRNA upon TLR activation. In order to determine potential changes in GILZ mRNA half-life, we incubated AMs with the transcription inhibitor actinomycin D in the presence or absence of LPS and determined changes in GILZ mRNA levels. Control cells displayed a half life of GILZ mRNA of 2.3 h, whereas it was reduced to 1.3 h upon LPS (Fig. 4A). Similar results were obtained when repeating the experiments with in vitro differentiated macrophages (Fig. 4B), indicating that the reduction of GILZ mRNA half-life after LPS treatment is not restricted to a single macrophage type.

Figure 4.

GILZ mRNA destabilization upon TLR activation. (A) AMs or (B) in vitro differentiated macrophages (Møs) were either left untreated or treated with actinomycin D (10 μg/mL) in the absence or presence of LPS (100 ng/mL). RNA was isolated at the indicated time points, and real-time RT-PCR analysis for GILZ was performed. Data are normalized to β-actin and represent means ± SEM of data pooled from five (A) or three (B) independent experiments with cells obtained from different donors. Curve fitting was done using Origin software, *p < 0.05. (C) THP-1 cells were transfected with a luciferase construct fused to the 3′UTR of GILZ or the empty luciferase-encoding vector. Transfected cells were treated with actinomycin D (10 μg/mL) alone or in combination with Pam3CSK4 (1 μg/mL) for the times indicated. The amount of luciferase and β-actin mRNA was determined by real-time RT-PCR, and luciferase was normalized to β-actin. Data are presented as the ratio of GILZ 3′UTR luciferase expression relative to empty vector luciferase expression in equally treated cells (set as 100%). Data represent the means ± SEM obtained from five independent experiments, *p < 0.05. p-Values were generated by Student's t-test.

As mRNA stability is mostly dependent on the 3′-untranslated region (3′UTR) of the respective mRNA [[22]], we aimed to determine the role of the GILZ 3′UTR on mRNA stability. Thus, THP-1 cells were transfected with a luciferase construct containing the 3′UTR of GILZ. As we observed a low responsiveness of undifferentiated THP-1 cells toward LPS (data not shown), Pam3CSK4 was chosen to activate cells. Treatment with Pam3CSK4 significantly destabilized luciferase mRNA containing the GILZ 3′UTR, indicating an involvement of the GILZ 3′UTR in GILZ mRNA destabilization (Fig. 4C).

Involvement of tristetraprolin (TTP) in GILZ mRNA destabilization

A prime mechanism of 3′UTR-dependent mRNA destabilization is mediated via RNA-binding proteins [[22]]. TTP is an RNA-binding protein critically involved in the regulation of many inflammation-associated mediators. Our data indicate that TTP is induced in AMs in a MyD88-dependent fashion, since its upregulation was only seen after LPS or Pam3CSK4, but not after Poly(I:C) treatment (Fig. 5A). To evaluate whether TTP plays a role in regulating GILZ expression, we transiently overexpressed TTP in THP-1 cells (Fig. 5C). We then determined expression of TNF-α mRNA, whose regulation by TTP is well described [[23]], similar to the recently identified TTP target IL-10 mRNA [[24]]. Overexpression of TTP significantly decreased TNF-α and IL-10 mRNA, thus confirming TTP functionality (Fig. 5B). Interestingly, these findings were accompanied by decreased GILZ expression both at mRNA and protein level, indicating a TTP-mediated GILZ downregulation (Fig. 5C and D).

Figure 5.

MyD88-dependent TTP induction and influence of TTP overexpression on GILZ expression. (A) AMs were either left untreated or treated with LPS (100 ng/mL), Pam3CSK4 (100 ng/mL), or Poly(I:C) (10 μg/mL) for 8 h. Western blot analysis for TTP was performed using tubulin as a loading control. One representative experiment is shown. Values indicate relative TTP signal intensities + SEM of two independent experiments performed in triplicate with cells obtained from different donors after normalization to tubulin values. (B) THP-1 cells were transfected with either control (Co-V) or TTP expression (TTP-V) vector and harvested after 24 h, followed by real-time RT-PCR analysis for TNF-α, IL-10, and GILZ. Data are normalized to β-actin and represent means + SEM of four independent experiments each performed in triplicate. *p < 0.05 compared with Co-V. n.s: not significant. (C and D) Total protein was isolated from Co-V or TTP-V transfected and untransfected (Co) THP-1 cells and assessed by western blot analysis for GILZ and TTP. Tubulin served as a loading control. (C) One experiment representative of four is shown. (D) TTP and GILZ signal intensities were quantified and relative signal intensities were calculated by normalization to tubulin values. Data are expressed as percentage of values for untransfected cells (not shown) and represent means + SEM of data pooled from four independent experiments. *p < 0.05 compared with Co-V. p-Values were calculated by Student's t-test.

To support these results and assess the influence of TTP on TLR-mediated GILZ regulation, we knocked down TTP expression using two different short hairpin RNA (shRNA) constructs in THP-1 cells. As a control, cells were transfected with a construct expressing an shRNA targeting the luciferase gene (shLuc) (Fig. 6A). Subsequently, cells were treated with Pam3CSK4 in order to induce MyD88-dependent GILZ downregulation. As previously seen in AM, LPS and Pam3CSK4 treatment induced TTP in control transfected THP-1 cells, whereas TTP induction upon TLR activation was undetectable upon TTP knockdown (Fig. 6E and F). As expected, Pam3CSK4 upregulated TNF-α as well as IL-10 mRNA, which was significantly enhanced by the knockdown of TTP (Fig. 6B and C). shRNA-mediated downregulation of TTP also abrogated GILZ downregulation upon activation with Pam3CSK4 both at mRNA and protein level (Fig. 6D, E, and 6G). We detected similar effects for GILZ protein when challenging cells with LPS (Fig. 6E and 6G).

Figure 6.

Impact of TTP knockdown on TLR-mediated decrease in GILZ expression. THP-1 cells were transfected with a control shRNA construct (shLuc) or one of two shRNA vectors targeting TTP (shTTP1, shTTP2). (A) Cells were lysed 24 h after transfection, and TTP was assessed by western blot analysis using tubulin as a loading control. One representative experiment is shown. Values indicate TTP signal intensities + SEM of data obtained from four independent experiments after normalization to tubulin values. (B and D) After 22 h of transfection, shLuc, shTTP1, or shTTP2 cells were either treated with Pam3CSK4 (100 ng/mL) or left untreated for 2 h and real-time RT-PCR for (B) TNF-α, (C) IL-10, or (D) GILZ mRNA was performed. Values are normalized to β-actin and represent means + SEM of four independent experiments each performed in triplicate. Data are expressed as x-fold (B and C) or percentage (D) of values for untreated shLuc cells, *p < 0.05. (E–G) 20 h after transfection, shLuc, shTTP1, or shTTP2 cells were treated with LPS (100 ng/mL) or Pam3CSK4 (100 ng/mL) for 4 h or left untreated. Western blot analysis for GILZ and TTP was performed using tubulin as a loading control. (E) One experiment representative of four is shown. (F) TTP or (G) GILZ signal intensities were quantified and normalized to tubulin values. Data are expressed as percentage of shLuc-transfected control and represent means + SEM of data pooled from four independent experiments, *p < 0.05. p-Values were calculated by Student's t-test.

Discussion

Lung macrophages contribute to pulmonary inflammation associated with asthma, COPD, or allergic diseases [[25, 26]]. Whereas the scientific interest in reducing macrophage activation mostly focuses on attenuating inflammatory mediators [[27]], little is known about regulation of antiinflammatory regulators, such as GILZ.

Glucocorticoids are currently the most widely used class of drugs in the treatment of inflammatory lung diseases. They are considered to exert their actions to a significant extent via induction of GILZ [[10, 16, 28]]. GILZ mediates its antiinflammatory properties mainly by direct binding to the NF-κB subunit p65, thereby preventing NF-κB-driven transcription [[9, 14, 15]]. Moreover, GILZ was reported to inhibit AP-1 activation by interaction with c-fos and c-jun subunits and to interfere with Raf/Ras signaling [[12, 28]]. Thus, GILZ expression attenuates inflammatory processes.

Here, we present the first report of constitutive GILZ expression in primary human lung macrophages, both in AMs and in IMs. Interestingly, although IMs express higher levels of IL-10 than AMs [[5, 19]] and IL-10 induces GILZ expression [[14]], expression levels do not differ between AMs and IMs. The transcript variants of GILZ described in the literature have as yet only been reported for the murine system [[29, 30]]. Still, alignment sequences can be found in databases also for postulated human GILZ transcript variants (GILZ transcript variant 1: NM_198057.2; transcript variant 2: NM_004089.3; transcript variant 3: NM_001015881.1). We therefore examined GILZ transcript variant expression and found that GILZ2 was the predominant transcript variant in all cell types tested. The human GILZ2 protein with a molecular weight of ∼17 kDa corresponds to the dexamethasone-inducible isoform exhibiting the NF-κB-inhibitory action described for this antiinflammatory mediator [[10, 15, 18]], whereas the function of human GILZ1 and GILZ3 presently remains elusive.

Regulation of GILZ during inflammation is poorly understood. TLRs play a key role in both infectious and noninfectious inflammatory lung diseases, as they are capable of sensing different microbial as well as endogenous molecules that are released after cell damage [[31]]. When we treated AMs with the TLR4 agonist LPS, both GILZ mRNA and protein levels rapidly decreased. Interestingly, we also detected significantly decreased GILZ mRNA and protein levels in lungs of LPS-exposed mice. In line with reports of decreased GILZ expression in inflammatory bowel disease, tuberculosis, and chronic rhinosinusitis [[14, 17]], these observations point to the in vivo relevance of our findings.

To assess functional implications of GILZ downregulation under inflammatory conditions, we used siRNA to knock down GILZ in in vitro differentiated macrophages. We found indeed that absence of GILZ enhanced NF-κB activity and TNF-α mRNA induction upon LPS-treatment. Our findings are in line with previous reports of increased cytokine production upon IL-1β-mediated activation in GILZ-silenced lung A549 cells [[13]], indicating that decreased levels of GILZ within the cell promote an inflammatory response.

As we speculated that downregulation of GILZ was not an effect specific for TLR4 activation, we also explored effects of other TLR ligands. Indeed, activation of all TLRs, with the exception of TLR3, diminished GILZ mRNA in AMs. As TLR3 represents the only TLR that does not utilize the adaptor molecule MyD88 for signal transduction [[3]], we examined the role of MyD88 in GILZ regulation by MyD88 knockdown. GILZ mRNA downregulation upon TLR activation was markedly, although not completely, abrogated in MyD88 knockdown cells. The partial restoration of GILZ mRNA levels rather occurs due to residual MyD88 than additional MyD88-independent mechanisms, as suggested by the experiments with receptor-specific ligands.

TLR-dependent signaling in response to bacterial as well as viral infections in the respiratory tract was shown to critically depend on MyD88 [[31]]. MyD88 knockout mice show an impaired expression of proinflammatory mediators and a reduced pathogen clearance after infection with Klebsiella pneumoniae [[32]]. They are also highly susceptible toward Legionella pneumophila, Streptococcus pneumoniae, as well as Haemophilus influenzae infections [[33-35]]. The potential inability to downregulate the expression of the immunosuppressant factor GILZ might contribute to an insufficient host response in those mice.

Gene expression can be regulated by both transcriptional and posttranscriptional mechanisms. Post-transcriptional modulation of gene expression is commonly procured by changes in mRNA stability, which can be modulated by various extracellular stimuli. Regulation of mRNA stability permits cells to rapidly decrease or increase mRNA levels, and therefore provides a mechanism for fast and tight regulation of protein production. In this manner, cells are able to dynamically change gene expression profiles depending on environmental challenges [[36]]. Thus, we speculated whether mRNA destabilization might account for TLR-mediated GILZ mRNA downregulation and found that GILZ mRNA half-life was indeed significantly decreased after LPS treatment.

The rate of mRNA decay is often determined by cis-regulatory elements containing AU-rich elements (AREs) found in the 3′UTR. These AREs mostly induce destabilization of the respective mRNA and are found in mRNAs coding for numerous cellular regulators, such as COX-2, IL-10, TNF-α, and a number of other cytokines involved in the inflammatory response [[23, 37-39]].

No real consensus sequence has yet been precisely defined for any class of ARE, but a domain containing several copies of the AUUUA pentamer embedded in a U-rich region is classified as ARE class I [[40, 41]]. The GILZ 3′UTR is shared by all human transcript variants. It contains only one copy of the AUUUA motif, which is not situated within a U-rich context and therefore does not account for an ARE class I [[41, 42]]. However, a U-rich domain can be found 269 bp upstream of the polyadenylation signal and might represent a class III ARE. Class III AREs are only loosely defined as U-rich regions without AUUUA pentamers, and the best documented example can be found within the 3′-UTR of c-jun [[38, 39]]. When we assessed the effect of the GILZ 3′UTR on mRNA stability by transfecting THP-1 cells with a luciferase construct containing the GILZ 3′UTR, we observed that GILZ 3′UTR-luciferase mRNA was indeed significantly destabilized upon TLR activation.

Several proteins are known to bind to AU-rich or U-rich elements, and many of them have been shown to regulate mRNA stability [[41, 42]]. These include the TTP family of zinc-finger RNA-binding proteins [[43]]. LPS has been shown previously to increase TTP mRNA and protein levels, for example, in THP-1 cells or human primary monocytes [[44, 45]]. Our data show that this is also true for AMs. Moreover, our observations reveal that TTP induction upon TLR activation occurs in a MyD88-dependent manner, which suggests a possible role of TTP in GILZ mRNA decay. Both TTP overexpression and knockdown experiments in THP-1 cells confirmed this assumption. IL-10 mRNA was only recently shown to be regulated by TTP in murine macrophages [[24]], although these observations have not been described for human cells yet. In contrast, TNF-α mRNA represents the best characterized TTP target [[23, 43]]. IL-10 and TNF-α were both downregulated when overexpressing TTP in THP-1 monocytes. Vice versa, knockdown of TTP and subsequent treatment with Pam3CSK4 led to higher induction of TNF-α as well as IL-10 in shTTP-transfected cells compared with control cells. Importantly, TTP overexpression also decreased GILZ amounts, both at mRNA and protein level. In addition, our results demonstrate that attenuated TTP expression by using RNAi abrogates TLR1/2- as well as TLR4-mediated GILZ downregulation. These observations indicate an involvement of TTP in the regulation of GILZ expression.

In summary, we present the first observation of GILZ expression in primary human lung macrophages. We also show that the expression of GILZ in AMs is actively downregulated by TLR activation in a MyD88-dependent fashion. Finally, we demonstrate that the RNA binding protein TTP is critical in TLR-mediated GILZ downregulation. GILZ has emerged as a potential target for the treatment of inflammatory lung diseases. Understanding its regulation under inflammatory conditions might help to develop strategies to influence its expression.

Materials and methods

Materials

Cell media, fetal calf serum (FCS), human serum, penicillin, streptomycin, and glutamine were from PAA (Pasching, Austria). TLR ligands (Pam3CSK4, Poly(I:C), LPS, flagellin, FSL-1, imiquimod, and ssRNA40) and the psiRNA-h7SKGFPzeo vector were obtained from Invivogen (San Diego, CA, USA). RT-PCR primers and dual-labeled probes were obtained from Eurofins MWG Operon (Ebersberg, Germany). Taq-Polymerase (5 U/μL), 10× Taq buffer, and the dNTP mix were from Genscript (Piscataway, NJ, USA). SMART pool siRNA targeting GILZ or MyD88 mRNA and control siRNA were purchased from Dharmacon (Chicago, IL, USA). GILZ stealth® siRNA and the respective control siRNA, Lipo-fectamine LTX reagent, Lipofectamine 2000, and Opti-MEM I reduced serum medium were from Invitrogen (Carlsbad, CA, USA). The Amaxa® Human Macrophage Nucleofector® Kit and the Nucleofector® II device were obtained from Lonza (Cologne, Germany). The pGL4.32(luc2P/NF-κB-RE/Hygro) vector and 5× passive lysis buffer were from Promega (Madison, WI, USA). Goat anti-GILZ antibody (sc-26518) was obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA), rabbit anti-MyD88 and rabbit anti-TTP antibodies were purchased from Abcam (Cambridge, MA, USA). Mouse antitubulin was from Sigma-Aldrich (St. Louis, MO, USA). IRDye© 680 or IRDye© 800 conjugated secondary antibodies were from Rockland (Gilbertsville, PA, USA) or LI-COR Biosciences (Lincoln, NE, USA). Restriction enzymes were obtained from New England Biolabs (Ipswich, MA, USA). Other chemicals were purchased either from Sigma-Aldrich or Roth (Karlsruhe, Germany) unless marked otherwise.

Mice

C57BL/6 mice (5 weeks old) were injected intraperitoneally either with 50 μg LPS diluted in saline or the equal amount of salt solution only. After 4 h of injection, mice were sacrificed and the lungs were removed to determine GILZ levels. All animal procedures were performed in accordance with the local animal welfare committee (permission no. H1 2.4.2.2, 13/2009).

Cell culture

Lung macrophages

Human AMs and IMs were isolated as described previously [[19, 46]]. The use of human material for isolation of primary cells was reviewed and approved by the local Ethics Committees (State Medical Board of Registration, Saarland, Germany; permission no. 213/06). The informed consent of all participating subjects was obtained.

In vitro differentiated macrophages

Monocytes were isolated from healthy adult blood donors (Blood Donation Center, Saarbrücken, Germany) as described in [[46]] and differentiated into macrophages by 20% FCS or 10% human serum for 7 days. Macrophages used for electroporation were cultured in Teflon bags (Biofolie 25; Heraeus, Hanau, Germany) instead of plates.

Cell lines and HUVECs

THP-1 monocytic cells were grown and differentiated as described previously [[46]]. 16HBE14o- lung epithelial cells and HUVECs were cultured as described in [[47]].

Semiquantitative RT-PCR and real-time RT-PCR

RNA isolation and reverse transcription were performed as described previously [[19, 48]]. Primer and probe sequences for real-time RT-PCR are given in Tables 1 and . Cloned standards were run alongside the samples to generate a standard curve. All human transcripts were determined using TaqMan probes [[19, 46]]. For luciferase and murine GILZ and 18S PCR, quantification was performed by SYBR Green as described in [[48]]. Expression of the human GILZ transcript variants was analyzed by semiquantitative RT-PCR. GILZ1, GILZ2, and GILZ3 PCR products were amplified in samples containing equivalent amounts of β-actin mRNA, as determined by real-time RT-PCR. Forty cycles of PCR were performed with the transcript variant specific primers listed in Table 1, resulting in PCR products of ∼240 bp. The amplified products were subjected to gel electrophoresis, and PCR product identity was confirmed by sequencing (MWG Eurofins sequencing service, Ebersberg, Germany).

Table 1. Primer sequences as used for semiquantitative RT-PCRand RT-PCR
 Primer sense, 5′→3′Primer antisense, 5′→3′
hu GILZTCCTGTCTGAGCCCTGAAGAGAGCCACTTACACCGCAGAAC
hu GILZ1TCCACAGGTCCGCCCCAATCCGCGCCTAAAGCCAGCCACCTT
hu GILZ2TCGAGTCACTTCCCTTCAACCATGGGGGTCTGATACATTTC
hu GILZ3GAATTCCTAGCTAGCTTCAGAGCAGACAGGACTGGAACTTCTC
hu IL-8TGCCAGTGAAACTTCAAGCAATTGCATCTGGCAACCCTAC
hu IL-10CAACAGAAGCTTCCATTCCAAGCAGTTAGGAAGCCCCAAG
hu TNF-αCTCCACCCATGTGCTCCTCACTCTGGCAGGGGCTCTTGAT
hu β-ActinTGCGTGACATTAAGGAGAAGGTCAGGCAGCTCGTAGCTCT
mu GILZGGGATGTGGTTTCCGTTAAAATGGCCTGCTCAATCTTGTT
mu 18SGCGCTTCTCTTTCCGCCAAGCTCTCCGACACCTCTT
mu TNF-αCCATTCCTGAGTTCTGCAAAGGAGGTAGGAAGGCCTGAGATCTTATC
IuciferaseCAACAGAAGCTTCCATTCCAAGCAGTTAGGAAGCCCCAAG
Table 2. Probe sequences as used for RT-PCR
 Probe, 5′ FAM à 3′ BHQ1
hu GILZTCCCGAATCCCCACAAGTGCCCGA
hu IL-8CAGACCCACACAATACATGAAGTGTTGA
hu IL-10AGCCTGACCACGCTTTCTAGCTGTTGAG
hu TNF-αCACCATCAGCCGCATCGCCGTCTC
hu β-ActinCACGGCTGCTTCCAGCTCCTC

Western blotting

Cells were suspended in lysis buffer (50 mM Tris-HCl, 1% (m/v) SDS, 10% (v/v) glycerol, 5% (v/v) β-mercaptoethanol, 0.004% (m/v) bromphenol blue) supplemented with a protease inhibitor mixture (Complete®, Roche Diagnostics, Basel, Switzerland) as described previously [[49]] and stored at –80°C. Murine lung tissue was disrupted using a Kontes tissue grinder and lysed accordingly. After sonication, lysates were boiled for 5 min. Subsequently, the samples were separated by SDS-PAGE on 15% gels and transferred to Immobilon FL-PVDF membranes (Millipore, Billerica, MA, USA). The membranes were blocked in blocking buffer for near-infrared western blotting for 1 h, incubated with primary antibodies for 3 h, followed by incubation with IRDye 680 or IRDye 800 conjugated secondary antibodies for 1.5 h as reported previously [[50]]. After washing, blots were scanned with an Odyssey Infrared Imaging System and relative protein amounts were determined using the Odyssey software.

Isolation of mycobacterial DNA

Mycobacteria (M. bovis BCG) were grown, harvested, and used for DNA isolation as described previously [[46]].

Plasmid generation and transfections

Two shRNA plasmids termed shTTP1 and shTTP2 encoding different siRNAs directed against the human TTP mRNA were created by cloning double-stranded oligonucleotides (shTTP1: 5′-ACCTCGGGATCCGACCCTGATGAATATCAAGAGTATTCATCAGGGTCGGATCCCTT-3′; shTTP2 as described by Fechir et al. [[51]]: 5′-ACCTCACAAGACTGAGCTATGTCGGATCAAGAGTCCGACATAGCTCAGTCTTGTTT-3′) into the BbsI sites of psiRNA-h7SKGFPzeo.

3′UTR-Luciferase reporter constructs were generated by fusing the GILZ 3′UTR sequence to a luciferase reporter gene. Human GILZ 3′UTR cDNA was amplified using the Expand High fidelity PCR System and the following oligonucleotides: 5′-GCCTACTAGTGCAGAGCCACTAAACTT G-3′ and 5′-AATAGAGCTCACTCTCACAAAACCCGCTAC-3′. The SacI/SpeI digested PCR product was inserted into the SacI/SpeI site of pMIR-REPORT–luc (Abgene, Epsom, UK) between the shLuc and the poly(A) signal.

The plasmids pZeo-hTTP-sense (TTP-V) and pZeo-hTTP-antisense (Co-V) were a kind gift from Hartmut Kleinert (Department of Pharmacology, Johannes Gutenberg University, Mainz, Germany).

Plasmid DNA for transfections was purified from overnight cultures by using an EndoFree Plasmid Maxi plasmid isolation kit (Qiagen, Hilden, Germany). Plasmids were introduced into THP-1 cells using Lipofectamine LTX transfection reagent according to the manufacturer's guidelines.

For knockdown of MyD88 by siRNA, cells were transfected with siMyD88 or control siRNA using Lipofectamine 2000 transfection reagent as recommended by the supplier.

Transfection of in vitro differentiated macrophages with siGILZ or control siRNA, either alone or in combination with the NF-κB reporter construct, was performed using the Nucleofector II device and the Human Macrophage Nucleofector Kit according to the manufacturer's recommendations.

NF-κB luciferase reporter assay

In vitro differentiated macrophages were cotransfected with the pGL4.32 [luc2P/NF-κB-RE/Hygro] vector and either GILZ or control siRNA. Cells were treated as indicated and harvested by addition of 1× passive lysis buffer 20 h after transfection. Luciferase activities were determined by addition of firefly luciferase substrate (470 μM D-luciferin, 530 μM ATP, 270 μM CoA, 33 mM DTT, 20 mM tricine, 2.67 mM MgSO4, and 0.1 mM EDTA, pH 7.8) and subsequent luminescence measurement using a POLARstar OPTIMA luminometer (BMG Labtech, Offenburg, Germany).

Statistics

All experiments were repeated at least three times. Mean values and standard error of the mean (SEM) were calculated from all data obtained in the respective experiment. Significant differences in the results were determined by the Student's t-test. (OriginPro 7.5G Software, OriginLabs, Northampton, MA, USA) as reported previously [[19, 46, 48]]. Differences were considered significant when p < 0.05.

Acknowledgements

We thank Hanno Huwer for providing human lung tissue, Sonja M. Kessler for delivering murine lung tissue, Rebecca T. Risch for providing HUVECs and 16HBE14o- cDNA, Sabine Meiser for help in plasmid generation, and Hagen von Briesen for advice in macrophage culture. This work was supported by the DFG (A.K. Kiemer: KI702, B. Brüne: BR999) and the LOEWE/OSF (B. Brüne: III L 4–518/55.004, 2009).

Conflict of interest

The authors declare no financial or commercial conflict of interest.

Abbreviations
3′UTR

3′-untranslated region

AM

alveolar macro-phage

Dex

dexamethasone

GILZ

glucocorticoid-induced leucine zipper

IM

lung interstitial macrophage

Pam3CSK4

N-palmitoyl-S-(2,3-bis(palmitoyloxy)-(2RS)-propyl)-(R)-Cys-(S)-Ser-(S)-Lys4 trihydrochloride

shRNA

short hairpin RNA

TTP

tristetraprolin

Ancillary