Legionella pneumophila induces human beta Defensin-3 in pulmonary cells

Background Legionella pneumophila is an important causative agent of severe pneumonia in humans. Human alveolar epithelium and macrophages are effective barriers for inhaled microorganisms and actively participate in the initiation of innate host defense. The beta defensin-3 (hBD-3), an antimicrobial peptide is an important component of the innate immune response of the human lung. Therefore we hypothesize that hBD-3 might be important for immune defense towards L. pneumophila. Methods We investigated the effects of L. pneumophila and different TLR agonists on pulmonary cells in regard to hBD-3 expression by ELISA. Furthermore, siRNA-mediated inhibition of TLRs as well as chemical inhibition of potential downstream signaling molecules was used for functional analysis. Results L. pneumophila induced release of hBD-3 in pulmonary epithelium and alveolar macrophages. A similar response was observed when epithelial cells were treated with different TLR agonists. Inhibition of TLR2, TLR5, and TLR9 expression led to a decreased hBD-3 expression. Furthermore expression of hBD-3 was mediated through a JNK dependent activation of AP-1 (c-Jun) but appeared to be independent of NF-κB. Additionally, we demonstrate that hBD-3 elicited a strong antimicrobial effect on L. pneumophila replication. Conclusions Taken together, human pulmonary cells produce hBD-3 upon L. pneumophila infection via a TLR-JNK-AP-1-dependent pathway which may contribute to an efficient innate immune defense.


Background
Legionella pneumophila is the causative agent of Legionnaires' disease, a severe pneumonia with high mortality [1]. The bacterium enters the human body by aerosol droplets and successfully establishes itself in macrophages and the alveolar epithelium, which normally offer an efficient barrier against infections [2]. Among the various putative virulence factors of this pathogen that have been identified to date, the type II (Lsp) and IVB (Dot/Icm) secretion system enable the bacteria to export proteins and therefore activates diverse cell signaling pathways [3,4]. Furthermore, bacterial cytoplasm membrane components, flagellin, and bacterial DNA, all major pathogen-associated factors of L. pneumophila, which activate innate immune response of alveolar epithelium as well as in macrophages [5][6][7]. L. pneumophila can be detected by means of toll-like receptors (TLR) or cytosolic pathogen pattern recognition receptors. Indeed it has been demonstrated that the atypical Legionella LPS can be recognized by TLR2, flagellin through TLR5 and DNA via TLR9 [5][6][7]. To clear L. pneumophila from the lung, a functionally intact innate immune system must be present. There is increasing evidence that human βdefensins (hBDs), a family of endogenous, cationic antimicrobial and immunomodulatory peptides secreted at epithelial mucosal and macrophages are critical components of host defense [8,9]). The human β-defensin (hBD) family comprises multiple members. While hBD-1 is constitutively expressed [9], production of hBD-2 and hBD-3, can be induced upon stimulation with bacteria and/or cytokines [10][11][12]. hBD-3 is, unlike other hBDs, a saltinsensitive defensin with a broad antimicrobial activity against e.g. multidrug-resistant nosocomial strains [10][11][12][13]. It has been reported that hBD-3 is expressed by pulmonary epithelial cells and increases in respiratory tract and serum of patients with bacterial pneumonia [14]. Consequently, the antibacterial properties of hBD-3 have attracted the attention of researchers in the field of pulmonary diseases.
In the study presented, we demonstrate that L. pneumophila induced hBD-3 in alveolar epithelium and macrophages. The hBD-3 expression was controlled by TLR2, TLR5 and TLR9, as well as activation of JNK and AP-1 (c-Jun) whereas NF-κB was not required. Also, recombinant hBD-3 elicited a strong antimicrobial effect on L. pneumophila. Moreover, inhibition of hBD-3 expression increased the L. pneumophila intracellular growth in pulmonary epithelium. Thus, hBD-3 production by pulmonary cells may contribute to the host response in Legionnaires' disease. Our results may significantly contribute to the understanding of the pathogenesis of Legionnaires' disease.

Cell lines Alveolar macrophages
Human alveolar macrophages were obtained by bronchoalveolar lavages (BAL) of patients from routine diagnostic. The study was approved by the local ethics committee of the Charité-Universitätsmedizin Berlin in accordance with the ethical rules stated in the Declaration of Helsinki. Cells were recovered, washed twice in cold PBS, then resuspended at 10 6 /ml in RPMI 1640 (PAA, Pasching, Austria) with 10% heat-inactivated FCS and antibiot-ics. Alveolar macrophages were placed into 24-well tissue culture plates and allowed to adhere for 2 h. The monolayers were then washed three times to remove nonadherent cells and antibiotics, and cultured in RPMI 1640 with 10% heat-inactivated FCS until infections.

Airway epithelial cells
Primary human small airway epithelial cells (SAEC) were obtained from Cambrex (Cambrex, Taufkirchen, Germany) and cultured according to the supplier's instructions. The alveolar epithelial cell line A549 was purchased from DSMZ (Braunschweig, Germany) and cultured in HAM'S F 12 (PAA, Pasching, Austria) with L-glutamine, 10% FCS without antibiotics.

Infection with bacterial strains
L. pneumophila sg1 strain 130b (ATCC BAA-74, kindly provided by N. P. Cianciotto, Chicago, USA), strain JR32 (wild type), JR32ΔdotA deficient in dot/icm, encoding a protein essential for the type IVB secretion system (kindly provided by H. Shuman, New York, USA), strain Corby (wild type), CorbyΔflaA deficient in flagellin as well as its type II secretion system knock out Cor-byΔlspDE [3] were routinely grown on buffered charcoalyeast extract (BCYE) agar for 2 days at 37°C before used [22]. Heat inactivation of L. pneumophila was accomplished in a water bath at 56°C for 30 min. No live bacteria were detected after this suspension was plated onto agar plates. The used cells were infected with L. pneumophila with a multiplicity of infection (MOI) of 10 at 37°C and 5% CO 2 .

Replication assay
To address the antimicrobial activity of hBD-3 towards L. pneumophila, 10 5 cfu/ml bacteria were suspended in HAM's 12 without supplements and recombinant hBD-3 was added. Bacteria growth was enumerated by plating suspension on agar plates as colony-forming-units (cfu). For the intracellular replication assay, A549 cells were infected with L. pneumophila 130b. After 2 h, cells were washed with PBS and remaining extracellular bacteria were killed by the cell-impermeable antibiotic gentamycin (50 μg/ml, Invitrogen, Karlsruhe, Germany) for 1 h. Afterwards cells were washed twice with PBS (time point 0) and incubated in HAM's 12 medium without supplements with recombinant hBD-3. Intracellular bacteria were enumerated by lysed cell suspension plated on agar plates as cfu.

Isolation of bacterial DNA
Bacterial DNA from Legionella was prepared using a Qiagen construct kit protocol for endotoxin-free isolation of bacterial DNA (Qiagen, Hilden, Germany) as described elsewhere [23].

hBD-3 ELISA
A549, SAEC and alveolar macrophages were infected as indicate, supernatants were collected. Samples were 10 times concentrated with Amicon Ultra-15 Centrifugal Filter Unit with Ultracel-3 membrane according to manufacturer instructions (Millipore GmbH, Schwalbach, Germany) and processed for hBD-3 quantification by ELISA (Phoenix Pharmaceuticals Inc., Belmont, CA, USA). Since the probes were 10 times concentrated, we presented in the figures the concentration of hBD-3 corresponding to a non concentrated probe.

Western Blot
A549 cells were transfected or infected as indicated. Cells were lysed in buffer containing Triton X-100, subjected to SDS-PAGE and blotted on Hybond-ECL membrane (Amersham Biosciences, Freiburg, Germany). Immunodetection of target proteins was carried out with specific antibodies against p-JNK, JNK, c-Jun, β-actin (antibodies were purchased from Santa Cruz Biotechnologies, Santa Cruz, CA, USA) and p-c-Jun (Ser 73) (purchased from Cell Signaling, Frankfurt, Germany). In all experiments, β-actin was detected simultaneously to confirm equal protein load.

Chromatin immunoprecipitation
A549 cells were infected with L. pneumophila 130b. Cells were processed for chromatin immunoprecipitation (ChIP) as described elsewhere [21]. hbd-3 promoter DNA was amplified by PCR using RedTaq polymerase (Qiagen, Hilden, Germany). PCR products were separated by agarose gel electrophoresis and detected by ethidium bromide staining of gels. Equal amounts of input DNA were confirmed by gel electrophoresis. For immunoprecipitation, the antibodies used were purchased from Santa Cruz Biotechnology (polymerase II, c-Jun), Frankfurt, Germany. The following promoter-specific primers for hbd3 were used: sense 5'-TCCCAGAACTAACACAC-CCTT-3' and antisense 5'-TTCCAGCCACAGCTG-CAATT-3'.

Statistical methods
Data are shown as means ± SEM of at least three independent experiments. A one-way ANOVA test was used for data of figure 1D -F; Figure 2; Figure 3A and B; Figure  4A and 4C. The main effects were then compared by a Newman-Keuls' post-test. A two way ANOVA test was used for data of Figure 3C, D and 3E and Figure 5 and main effects were compared by a Bonferroni post-test. P < 0.05 was considered to be significant and indicated by asterisks or H-Keys. If not indicated otherwise, test was performed vs. control (*) or stimulated probe vs. inhibitor treated probe (#).

L. pneumophila induced hBD-3 expression of human alveolar epithelial cells and macrophages
SAEC, A549 cells as well as primary human alveolar macrophages were infected with L. pneumophila strain 130b or stimulated with TNFα as control for different periods of time and analyzed by RT-PCR and ELISA ( Figure 1). An increased mRNA level of hBD-3 was observed in primary lung cells (SAEC) (figure 1A), in A549 cells (figure 1B) and in primary alveolar macrophages ( figure 1C). Furthermore, a strong hBD-3 release could be demonstrated by ELISA in all three cell types infected with L. pneumophila (figure 3D -F). This suggests an important role of hBD-3 for the L. pneumophila-induced innate immune response of the lung. In A549, a time dependent increase of hBD-3 release (figure 1E) was observed. Next we addressed the mechanisms involved in L. pneumophilainduced hBD-3 expression by using A549 cells.

The type II and IV secretion systems of L. pneumophila are not essential for hBD-3 release of alveolar epithelial cells
The type II (lsp) as well as the Dot/Icm type IVB secretion system is known to be important pathogen factors of L. pneumophila [3,4]. Infection of A549 cells with L. pneumophila strains JR32 and Corby induced comparable hBD-3 release to strain 130b (figure 1B, 2) suggesting that induction of this peptide is common in L. pneumophila infection of alveolar epithelium. Furthermore, no signifi-cant differences could be observed among the effects of JR32ΔdotA (figure 2) and CorbyΔlspDE (figure 2) deletion mutants and wild-type strains with respect to hBD-3 liberation. These data suggest that the types II as well as the type IVB secretion system and their effector molecules are not involved in L. pneumophila-induced hBD-3 release. Furthermore, data indicates that intracellular replication of L. pneumophila appears not to be required for L. pneumophilainduced release of hBD-3, because the presence of the type IVB secretion system was shown not to be important.

hBD-3 release induced by L. pneumophila in alveolar epithelial cells is controlled by TLR2, TLR5 , and TLR9
Following our previous and present observations that L. pneumophila strongly activates lung epithelial cells [20,21,24], we tested the hypothesis that TLR2, TLR5, and TLR9 might be essential for the observed hBD-3 induction. To address this issue we stimulated SAEC with the TLR2 ligand Malp-2, the TLR5 ligand flagellin and non-methylated CpG-motifs (ODN) as ligand for TLR9. Incubation of SAEC with all agonists stimulated the release of hBD-3 ( figure 3A). Furthermore we assessed the role of flagellin for the L. pneumophila induced hBD-    3D). Furthermore we addressed the role of TLR9 sensing nucleic acids in this process. Therefore A549 cells were transfected with specific TLR9-siRNA and afterwards incubated with purified L. pneumophila-DNA and ODN or infected with L. pneumophila ( figure 3E). L. pneumophila-DNA and ODN strongly induced hBD-3 release to the same extent as the wild type strain and liberation of this peptide was reduced in all cells transfected with specific-TLR9-siRNA. In all RNAi experiments TNFαrelated hBD-3 production (figure 3) was not reduced. Overall, our observations suggest an important role for TLR2, TLR5 and TLR9 with respect to the hBD-3 induction observed in alveolar epithelium.

hBD-3 expression in L. pneumophila -infected cells requires AP-1 (c-Jun)
Recent studies showed that activation of NF-κB and/or AP-1 (c-Jun) controls hBD-3 expression [15][16][17][18][19]. Therefore, to further investigate a possible role of NF-κB activation for L. pneumophila-dependent hBD-3 release, we pre-incubated A549 cells with the proteasome inhibitor MG-132 to prevent IκB degradation. In addition we made pneumophila-infected A549 cells whereas interleukin-8 release was significantly reduced (data not shown). Our data demonstrated that activation of NF-κB by L. pneumophila was not important for hBD-3 release in lung epithelium. Activation of JNK is considered to participate in the regulation of inflammatory processes in alveolar epithelial cells [25]. Phosphorylation of the JNK kinase by L. pneumophila infection of epithelial cells was detected 1 h after infection, increased up to 2 h, and decreased slightly at 4 h (figure 4B). Since JNK-mediated phosphorylation enhances the ability of c-Jun, a component of the AP-1 transcription factor, to activate transcription [20], we inhibited this kinase by pre-incubation of A549 cells with a JNK inhibitor (JNK II) before infection with L. pneumophila 130b. As shown in figure 4C, inhibition of JNK abolished L. pneumophila induced hBD-3 release. These observations suggest that JNK is important for the L. pneumophila-induced hBD-3 release in lung epithelial cells.
To further investigate the role of AP-1, we addressed the AP-1 subunit c-Jun activation following infection of alveolar epithelial cells with L. pneumophila. We observed a time dependently increased phosphorylation at serine 73 of the AP-1 subunit c-Jun ( figure 4D).
To characterize the mechanism by which AP-1 (c-Jun) contributes to L. pneumophila-induced hBD-3 expression, the recruitment of c-Jun to the hbd-3 promoter were evaluated by ChIP assay. We observed an increase of the AP-1 subunit c-Jun binding to the hbd-3 promoter (figure 4E) whereas the NF-κB subunit p65 was not recruited (data not shown). An increased binding of the RNA polymerase II (Pol II) to the hbd-3 promoter was indicative for the subsequent activation of the hbd-3 gene in infected A549 cells. These experiments confirm that the JNK-AP-1 (c-Jun) pathway controls hBD-3 expression in L. pneumophila infected alveolar epithelial cells.
The chemical inhibitors used in these experiments did neither reduce viability and proliferation of the A549 cells nor induces morphological signs of cytotoxicity, or alterations of bacterial growth within the time-frame tested (data not shown). L. p. + 5ng/ml rhBD-3 + 10ng/ml rhBD-3 + 15ng/ml rhBD-3 + 0.5µg/ml Erythr.

hBD-3 elicited a strong antimicrobial effect towards L. pneumophila
To study the susceptibility of L. pneumophila to hBD-3, we incubated the wild type 130b in suspension with increasing concentrations of recombinant hBD-3 and a cfu assay was performed. As control for inhibition of replication we used the antibiotic erythromycin [26]. hBD-3 efficiently inhibited replication of this strain of Legionella in all used concentrations ( figure 5A). Next we elucidated if hBD-3 has an antimicrobial effect towards intracellular Legionella growth. Therefore we infected A549 cells with L. pneumophila for a replication assay. Treatment with hBD-3 reduced the replication of the intracellular bacteria ( figure 5B). Finally, the role of endogenous hBD-3 for intracellular replication of Legionella was tested in A549 cells transfected with hBD-3-specific siRNA or control siRNA. Knockdown of hBD-3 was confirmed by ELISA and RT-PCR (data not shown). The intracellular replication of L. pneumophila 130b was enhanced in cells transfected with specific hBD-3 siRNA compared to cells transfected with control siRNA (figure 5C), suggesting the importance of this peptide in Legionella-induced innate immune response.

Discussion
In the study presented, we demonstrate that L. pneumophilainduced hBD-3 expression was dependent on TLR2, TLR5 and TLR9 in human alveolar epithelial cells.
The activation of JNK-pathway was identified to play a key role for L. pneumophilatriggered hBD-3 release. Furthermore, detailed analysis of signal transduction pathways provides evidence that activation of AP-1 subunit c-Jun but not NF-κB plays a key role in L. pneumophilamediated hBD-3 release ( figure 6). We demonstrated also that hBD-3 elicited an antimicrobial effect towards L. pneumophila. We also showed that recombinant hBD-3 decreased replication of Legionella more efficient in lower concentrations than the antibiotic erythromycin used in treatment of Legionnaires' disease [26]. The mechanism by which defensins kill or inactivate bacteria is not precisely understood but is generally thought to be related to a perforation of the peripheral microbial membrane [12]. A recently published study showed a co-localisation of endogenous hBD-2 with the bacterial cell wall of extra-and intracellular replicating Mycobacterium tuberculosis in A549 cells [27]. For hBD-3 a similar antimicrobial mechanism can be assumed since a keratinocyte cell line engineered to overexpress hBD-3 demonstrated significant antimicrobial activity against Staphylococcus aureus [28]. On the other hand, hBD-3 can activate the NF-κB pathway via a TLR-triggered mechanisms [29]). This may induce secondary effector molecules which may reduce intracellular replication of Legionella. Since we observed in our study an antimicro-bial effect within four hours, we presume that hBD-3 kills L. pneumophila via direct perforation of the bacterial membrane.
The expression of hBD-3 in respiratory cells, especially in infections of the lung, is not well understood and was so far rather investigated in studies of oral infections as well as in epithelium of skin and intestine [12,16,17,30,31]. Our data showed that L. pneumophila infection of pulmonary epithelium and alveolar macrophages led to increased mRNA levels of hBD-3 and a strong secretion of this peptide. Since different isolates of L. pneumophila serogroup1 were found to induce a comparable release of hBD-3, it is likely that hBD-3 production is a common response of lung epithelial cells to L. pneumophila infection. This assumption is supported by a study which showed increased hBD-3 concentrations in respiratory tract and serum of patients suffering bacterial pneumonia [14]. In this study, hBD-3 exhibited a strong antibacterial effect on Staphylococcus aureus, Escherichia coli and Pseudomonas aeruginosa [14]. Since we also demonstrated that hBD-3 has an antimicrobial effect towards L. pneumophila and this peptide do orchestrate the recruitment of alveolar macrophages to the site of infection [32], we assume that this defensin might be important for immune response in infectious diseases of the lung.
Pulmonary epithelial cells may detect L. pneumophila by TLRs [5][6][7]. In accordance we demonstrated that recognition of L. pneumophila by TLR2, TLR5 and TLR9 was essential for the production of hBD-3. In recently published studies, it was shown that hBD-3 expression was induced TLR2-dependent in skin epithelial cells [16,33,34]. To our knowledge, our study showed for the first time the induction of hBD-3 via activation of TLR5 and TLR9.
In mice pneumonia studies, TLR2, TLR5 and TLR9 were required for effective innate immune responses against L. pneumophila [5,7,35]. Since we demonstrated that inhibition of all three TLRs decreases L. pneumophila-induced hBD-3 release, we assume that these receptors might be essential for antimicrobial innate immune response in Legionella infections. Interestingly, hBD-3 liberation was not reduced in infections with type II or IVB secretion system mutant strains, suggesting that recognition of bacterial membrane component via TLR2, recognition of Legionella-flagellin via TLR5 and/or nonmethylated bacterial DNA through TLR9 might be the major pathways for L. pneumophila induced hBD-3.
A complex signaling network regulates the expression of inducible hBD-3 [16,17,33]. In L. pneumophilainfected lung epithelial cells, we noted that the bacteria induced a JNK-dependent release of hBD-3. This prompted to the assumption that JNK related transcription factor AP-1 is involved in this process. Indeed, we showed that AP-1 subunit c-Jun was recruited to hbd-3 promoter. Our data are in line with previous studies which demonstrated a strong AP-1 dependency of hBD-3 induction [16,17]. In contrast, a recently reported role of NF-κB in the regulation of the hBD-3 gene in keratinocytes has been published [15]. Since an NF-κB binding site is also located on hbd-3 promoter, we tested if a binding of this transcription factor could be observed in L. pneumophilainfected alveolar epithelial cells and failed to confirm a binding of NF-κB subunit p65 to this structure (data not shown). This strengthens our data that mainly AP-1 was essential for L. pneumophila induced release of hBD-3 in pulmonary cells.
Functional studies and signal transduction experiments of this study were performed in A549 cells. Since infection with L. pneumophila and stimulation with TLR agonists showed comparable expression of hBD-3 in A549 cells and primary SAEC, we assume similarities in regulation of hBD-3 between both cell types. Furthermore, previous studies by our and other groups demonstrated analogous behavior of A549 cells and the primary SAEC in release of cytokines, defensins, prostaglandin E2 as well as in production of reactive oxygen species after infection [20,21,24,36,37]. Nevertheless, A549 cells are a tumor cell line and further studies in vivo are needed to dissect the signaling pathways mediating L. pneumophila-related defensin release.

L. pneumophila
In conclusion, we found that L. pneumophilatriggered hBD-3 release is TLR2, TLR5-and TLR9-dependent hBD-3 in human pulmonary epithelial cells. Expression of hBD-3 included activation of the JNK-AP-1 (c-Jun) pathway, whereas NF-κB was not essential for this process in A549 cells. Since control of the immune response is crucial to assure bacterial clearance and to prevent excessive tissue damage in pneumonia, the mechanism described above could be important for the host defense in Legionnaires' disease.