Histone Acetylation and Flagellin Are Essential for Legionella pneumophila-Induced Cytokine Expression¹ Free

The Gram-negative Legionella pneumophila is the causative agent of Legionnaires’ disease, a severe community-acquired pneumonia, and the second most commonly detected pathogen in pneumonia that is admitted to intensive care units in industrialized countries (1, 2). Legionella bacteria are facultative intracellular parasites of amoeba replicating in natural and man-made aquatic environments (1). Infection of humans is observed after inhalation of contaminated water aerosol droplets. Essential results about L. pneumophila pathogenesis were obtained by analyzing infection of protozoa or immune cells like macrophages. However, although Legionella bacteria replicate efficiently within lung epithelial cells (3, 4) and recent studies pointed to the lung epithelium as an important sentinel and effector system of innate immunity (5, 6, 7, 8), little is known about the consequences of pulmonary epithelial cell infection with L. pneumophila. Upon infection, activated lung epithelium may contribute to the regulation of the immune response by the liberation of numerous proinflammatory mediators (9, 10). As an example, liberation of the CXC chemokine IL-8 plays an important role for the recruitment of immune cells, like macrophages and granulocytes, to the site of infection in the lung (11).

After recognition of pathogens by transmembraneous and cytosolic pattern recognition receptors (6, 8), complex networks of kinases and transcription factors initiate the expression of proinflammatory cytokines (6, 7, 8, 9).

However, recent evidence suggested that the host DNA has to be prepared for the binding of transcription factors and the basic transcription machinery. At nontranscribed genes, tight binding of the DNA to histone octamers obstructs binding of transcription factors (12, 13). Loosening of DNA-histone interactions and subsequent unmasking of transcription factor binding sites is controlled by specific covalent modifications of accessible N-terminal histone tails (14). In particular, phosphorylation at Ser¹⁰ and acetylation at Lys¹⁴ on H3 and acetylation of H4 seems to have a special impact on gene regulation. For example, it was found that LPS stimulation of dendritic cells induced p38 MAPK-dependent phosphorylation at Ser¹⁰ and acetylation at Lys¹⁴ on H3 and acetylation on H4 specifically occurring at IL-8, and MCP-1, but not at TNF-α or MIP-1α, genes (14). In endothelial cells, viable Listeria monocytogenes-related expression of IL-8, but not of IFN-γ, depended on modifications of H3 and H4 (12, 15). Although the host response in pneumonia is characterized by massive cytokine production (10), and altered histone modifications were observed in diseased lungs (16), it is not known how histone modifications contribute to innate immune regulation in the lung.

In this study, we demonstrated that infection of lung epithelial cells with L. pneumophila induced acetylation and phosphorylation of histones both globally and at the IL-8 promoter and specifically modified histone deacetylases (HDAC)⁴ /histone acetylase (HAT) composition at the gene promoter. Inhibition of HDACs increased Legionella-related IL-8 expression whereas blocking of HATs inhibited IL-8 expression. Thus, histone modifications may contribute to the regulation of proinflammatory gene expression in infected lung cells.

HAM’s F12 (PAA) was obtained from Invitrogen Life Technologies. Protease inhibitors, Triton X-100, and Tween 20 were purchased from Sigma-Aldrich. SB202190 was obtained from Calbiochem (Merck), IκB kinase (IKK)-nemo binding domain (NBD) was obtained from Biomol, trichostatin A (TSA) was obtained from Sigma-Aldrich, and anacardic acid and recombinant were obtained from flagellin from Alexis. All other chemicals used were of analytical grade and obtained from commercial sources.

Alveolar epithelial cell line A549 was purchased from American Type Culture Collection (ATCC) and cultured in HAM’s F12 with l-glutamine, 10% FCS without antibiotics as described (4, 17, 18). The primary human small airway epithelial cells (SAEC) were obtained from Clonetics/Cambrex (SAEC System; Cambrex) and cultured in a SAEC Bullet kit (Clonetics/Cambrex) according to the supplier’s instruction.

L. pneumophila sg1 130b wild-type (ATCC BAA-74; provided by N. P. Cianciotto, Northwestern University Medical School, Chicago, IL), Corby wild-type, and a Corby flaA mutant, defective in flagellin (both provided by K. Heuner, Würzburg University, Würzburg, Germany), were routinely grown on buffered charcoal-yeast extract agar for 2 days at 37°C and subsequently inoculated in plain HAM’s F12 to an OD at 660 nm (OD₆₆₀) of 0.2–0.4. A549 cells or SAEC (10⁵/ml) were infected with 10⁵–10⁸ CFU/ml L. pneumophila, i.e., a multiplicity of infection of 1:1–1:1000, for the indicated durations in 1 ml of epithelial cell growth medium. Extracellular bacteria were not routinely killed with antibiotics or washed away. L. pneumophila strains did not significantly grow in epithelial cell growth medium as controlled by serial dilutions plated on buffered charcoal-yeast extract agar. Bacterial density was checked by determining the OD₆₆₀ with a Beckman spectrophotometer DU520 (Beckman Coulter) as described (4, 17, 19).

A549 cells (10⁶) were detached, nucleofected with 2 μg of short silencing RNA (siRNA) (Amaxa Cell Line Nucleofector kit), grown for 48 h in 12-well plates, and then stimulated as indicated. The following siRNAs were purchased from Ambion (Applied Biosystems/Ambion): HDAC1 S1: sense CCAAGUACCACAGCGAUGATT, antisense UCAUCGCUGUGGUACUUGGTC; HDAC1 S2: sense GCUUUAACCUGCCUAUGCUTT, antisense AGCAUAGGCAGGUUAAAGCTC; HDAC5 S1: sense CCUGAUGUGGUCCUAGUCUTT, antisense AGACUAGGACCACAUCAGGTG; and HDAC5 S2: sense GCCUGGUGCUGGAUACAATT, antisense UUUGUAUCCAGCACCAGGCTC.

Total RNA from A549 cells was isolated with the RNeasy Mini kit (Qiagen) and reverse transcribed using AMV reverse transcriptase (Promega). The generated cDNA was amplified by semiquantitative RT-PCR using specific primers as described (4): HDAC1 (sense 5′-CACCCGGAGGAAAGTCTGTTA-3′, antisense 5′-TCTTCCAGGCCGTCACCAT-3′), HDAC2 (sense 5′-ATTACTGATGCTTGGAGGAGG-3′, antisense 5′-ACCACTGTTGTCCTTGGATTT-3′), HDAC3 (sense 5′-ATGACGGTGTCCTTCCACA-3′, antisense 5′-TCATAGGTCAGGAGGTCTGCA-3′), HDAC4 (sense 5′-TGGAGCTGCTGAATCCTGC-3′, antisense 5′-TCATCTTTGGCGTCGTACAT-3′), HDAC5 (sense 5′-AACCATCCTCCTTGGAAATCCTG-3′, antisense 5′-TCCTTTGACTTCGACAAGAGG-3′), HDAC6 (sense 5′-AACCAGGCAGCGAAGAAGTA-3′, antisense 5′-ATAAGACTGTGCTGGGCGTGA-3′), HDAC7 (sense 5′-TGCTCCTCTACGGCACCAA-3′, antisense 5′-TACCTCATCCACAGCCCCACT-3′), HDAC8 (sense 5′-AGTCGCTGGTCCCGGTTTA-3′, antisense 5′-TGAATGCGTCTTCTACACCAT-3′), GAPDH (sense 5′-CCACCCATGGCAAATTCCATGGCA-3′, antisense 5′-TCTAGACGGCAGGTCAGGTCC ACC-3′) and IL-8 (sense 5′-CTAGGACAAGAGCCAGGAAGA- 3′, antisense 5′-AACCCTCTGCACCCAGTTTTC-3′), IL-8 sense 5′-CTAGGACAAGAGCCAGGAAGA-3′.

A549 cells or SAEC were infected with L. pneumophila as indicated and then subjected to ChIP assay as previously described (4, 12, 18) using anti-p65 (C-20), anti-HDAC1, -CREB binding protein (CBP), -p300 (Santa Cruz Biotechnology) and anti-P(Ser¹⁰)/Ac(Lys¹⁴)-H3, anti-Ac-H4, anti-HDAC5 (Upstate Biotechnology), or anti-RNA polymerase II (Pol II; Santa Cruz Biotechnology) Abs (4, 12, 18). The IL-8 enhancer region was amplified by PCR using HotStarTaq polymerase (Qiagen) and specific primers as followed: sense 5′-GAATCCACGGATACAGAACCT-3′, antisense 5′-TTGACAACACGAACAGTGTCG-3′. PCR amplifications of the total input DNA in each sample is shown as a control. As control primers covering a region 1-kb upstream, we used the following: sense 5′-ATCATGGGTCCTCAGAGGTCAGAC-3′, antisense 5′-GGTGGGAGGGAGGTGTTATCTAATG-3′ (20).

Confluent A549 cells or SAEC were infected or stimulated as indicated in a humidified atmosphere. After incubation, supernatants were collected and processed for IL-8 quantification by sandwich ELISA as described previously (4, 12, 18).

For determination of HDAC expression, A549 cells were infected as indicated, washed twice, and nuclear extracts of A549 cells were separated by SDS-PAGE and blotted as described previously (4). Membranes were exposed to Abs specific to HDAC1, 2 (Santa Cruz Biotechnology), 5 (Upstate Biotechnology), or USF2 (Santa Cruz Biotechnology), respectively, and subsequently incubated with secondary Abs (IRDye 800-labeled anti-mouse, or Cy5.5-labeled anti-rabbit, respectively). Simultaneous detection of USF2 and respective HDAC by using an Odyssey infrared imaging system (LI-COR) confirmed equal protein loading as previously described (18).

For histone analysis, cells were lysed and H₂SO₄ was added to a final concentration of 0.2 M as described previously (12). Protein was precipitated with 50% trichloroacetic acid in a final concentration of 20%, subjected to SDS-PAGE, and blotted on Hybond-ECL membrane (Amersham Biosciences). Immunodetection was conducted with Abs specifically detecting Ac-H4 and P-Ser¹⁰/Ac-Lys¹⁴-H3 (Upstate Biotechnology). In all experiments, ERK2 (Santa Cruz Biotechnology) was detected simultaneously to confirm equal protein loading. Proteins were visualized by incubation with secondary IRDye 800- or Cy5.5-labeled Abs, respectively (Odyssey infrared imaging system; LI-COR).

Data are shown as means ± SEM of at least three independent experiments. A one-way ANOVA was used for data in Figs. 1, C and D; 3; 5,D; 6, A–C; 7,A; and 8, A and B. Main effects were then compared by a Newman-Keul’s posttest. A value of p < 0.05 was considered to be significant and indicated by asterisks or double crosses (if not indicated otherwise, test was performed vs control).

Because IL-8 is of particular importance for the recruitment of immune cells into the lung during pneumonia (11), we analyzed IL-8 expression as a model cytokine in Legionella-infected A549 cells. Human alveolar epithelial A549 cells were infected with L. pneumophila strain 130b and IL-8 mRNA (Fig. 1, A and B) and protein (Fig. 1, C and D) expression was analyzed. Legionella induced, concentration- (10⁵–10⁷ CFU/ml) and time-dependently, IL-8 mRNA and protein expression in A549 cells.

Because modifications of histones, in particular acetylation of H4 and phosphorylation/acetylation of P-Ser¹⁰/Ac-Lys¹⁴-H3, seem to contribute to the regulation of inflammatory genes like IL-8 (12), we assessed global, genome-wide histone modifications in L. pneumophila-infected cells by Western blot (Fig. 2 A). L. pneumophila (10⁷ CFU/ml) time-dependently induced acetylation of H4 and phosphorylation/acetylation of P-Ser¹⁰/Ac-Lys¹⁴-H3, but showed no alteration of Lys9-H3 dimethylation. The effect of L. pneumophila on H3 modification was comparable to effects caused by 0.1 ng/ml HDAC inhibitor TSA (1 h) as positive control, but weaker than the TSA effects observed at H4.

Next, we analyzed histone modifications at the IL-8 gene promoter by ChIP (Fig. 2 B). A549 cells were infected with L. pneumophila (10⁷ CFU/ml) for 30 min to 4 h. L. pneumophila time-dependently increased acetylation of H4 and phosphorylation/acetylation of P-Ser¹⁰/Ac-Lys¹⁴-H3. Acetylation/phosphorylation of histones was accompanied by recruitment of NF-κB p65/RelA, which was known to be critical for IL-8 expression (7). Moreover, gene transcription was further indicated by recruitment of Pol II at the gene promoter. Overall, L. pneumophila infection resulted in phosphorylation/acetylation of histones in A549 cells. To validate the specificity of these findings, we used primers covering a region 1 kb upstream of the transcription start site of IL-8 (20), but we could not detect significant binding of Pol II or p65 (data not shown).

Next, we wondered whether inhibition of HDACs by TSA or blocking of HATs by anacardic acid impacts on L. pneumophila-related IL-8 expression (Fig. 3,A). We increased global histone acetylation by incubation of A549 cells with TSA (0.1 and 1 ng/ml, 1 h) which did not induce IL-8 secretion per se. Preincubation with TSA and subsequent infection with L. pneumophila increased IL-8 release synergistically (Fig. 3 A).

Suppression of histone acetylation by blocking of HATs via anacardic acid, in contrast, concentration-dependently (0.1–10 μM, 1 h) reduced L. pneumophila-related IL-8 expression (Fig. 3 B) indicating that histone acetylation promotes IL-8 expression in L. pneumophila-infected epithelial cells.

HDACs are known to be negative regulators of gene transcription and their expression may be altered in inflamed lung cells and lung tissue (16, 21). In unstimulated A549 cells, mRNA for class I HDAC1, 2, 3, and 8, as well as for class II HDAC4, 5, 6, and 7 was detected by RT-PCR (data not shown). L. pneumophila induced, concentration (10⁵–10⁸ CFU/ml)- and time (1–7 h)-dependently, the expression of HDAC5 and IL-8 (Fig. 4, A and B, respectively), but not of HDAC1 and HDAC2 mRNA (Fig. 4) and the other investigated HDACs (data not shown). The induction of HDAC5 was also observed at the protein level in nuclear extracts of L. pneumophila-infected A549 cells (Fig. 4 C).

Next, we analyzed the binding of histone-modifying proteins at the IL-8 gene promoter by ChIP (Fig. 5 A). A549 cells were infected with L. pneumophila (10⁷ CFU/ml) for 30 min to 4 h. Legionella time-dependently increased recruitment of the highly related HATs CBP and p300 (12) while HDAC1 and HDAC5 binding was temporarily reduced at the IL-8 gene promoter in L. pneumophila-infected cells. In particular, there is strong recruitment of HDAC5 over time at the promoter although still high acetylation was detected.

To assess the relevance of HDAC1 and HDAC5 binding to the IL-8 promoter, we performed a gene knockdown of these enzymes by nucleofection with two different siRNA sequences, respectively (Fig. 5, B and C). However, although protein levels of HDAC1 and HDAC5 were reduced after 48 h (Fig. 5, B and C) and 72 h (data not shown), we found no effect on L. pneumophila-induced IL-8 release by A549 cells (Fig. 5 D).

A549 lung epithelial cells could be activated by purified L. pneumophila flagellin (Fig. 6,C). We wondered whether the proinflammatory activation of A549 cells by living L. pneumophila depended on bacterial flagellin expression. A549 cells were infected with L. pneumophila Corby wild-type and Corby flagellin deletion mutants. These mutants induced lower levels of IL-8 release (Fig. 6, A and B), histone modification, and Pol II recruitment to the IL-8 promoter (Fig. 6 D).

We wondered whether the proinflammatory activation of A549 cells by living L. pneumophila was dependent on MAPK p38 or the canonical NF-κB pathway. Therefore, we preincubated A549 lung epithelial cells before L. pneumophila infection with a specific chemical inhibitor of p38 MAPK (SB201290) or a cell-permeable peptide inhibitor of the IKK complex (IKK-NBD). Both inhibitors strongly reduced L. pneumophila-induced IL-8 release (Fig. 7,A) as well as recruitment of p65 and histone H3 phosphorylation/acetylation at the endogenous IL-8 promoter (Fig. 7 B).

To validate the main findings on primary cells, we made use of commercially available SAEC. Lower infection doses of L. pneumophila induced activation of SAEC (Fig. 8 and data not shown). We increased global histone acetylation by incubation of SAEC with TSA (0.01 ng/ml, 1 h) which only insignificantly induce IL-8 secretion per se. Preincubation with TSA and subsequent infection with L. pneumophila increased IL-8 release synergistically (Fig. 8,A). Suppression of histone acetylation by blocking of HATs via anacardic acid (1 μM, 1 h) in contrast reduced L. pneumophila-related IL-8 expression (Fig. 8,B) indicating that histone acetylation promotes IL-8 expression in primary lung epithelial cells infected by L. pneumophila. Next, we analyzed histone modifications at the IL-8 gene promoter by ChIP (Fig. 8 C). SAEC were infected with L. pneumophila (10⁶ CFU/ml) for 30 min to 3 h. L. pneumophila time-dependently increased recruitment of p300, acetylation of H4, and phosphorylation/acetylation of P-Ser¹⁰/Ac-Lys¹⁴-H3. Moreover, gene transcription was further indicated by recruitment of Pol II at the gene promoter. Therefore, principal effects as well as time courses were comparable between A549 cells and SAEC.

In the present study, we found that L. pneumophila strain 130b-induced expression of the chemoattractant IL-8 is controlled by histone modifications in the human alveolar type II epithelial cell line A549 and primary human SAEC. L. pneumophila induced global, genome-wide histone modifications as well as acetylation of H4 and phosphorylation/acetylation of Ser¹⁰/Lys¹⁴-H3 specifically at the IL-8 gene promoter. IL-8 release and histone modifications were dependent on p38 MAPK and the canonical NF-κB pathway. Although inhibition of HATs reduced IL-8 expression, blocking of HDACs further increased IL-8 expression. Accordingly, CBP and p300 were quickly recruited to the promoter, whereas HDAC1 and 5 binding first decreased and later on reappeared at the IL-8 promoter. Interestingly, HDAC5 expression was specifically induced by L. pneumophila; knockdown of HDAC1 or HDAC5, however, did not alter IL-8 release. Finally, histone acetylation, IL-8 promoter activation, and IL-8 release depended in part on L. pneumophila flagellin expression.

During airborne lung infections, the airway and lung epithelium comprise the first host barrier against invading pathogens and is thereby directly exposed to pathogens and their products. In addition to acting as a mechanical barrier, liberation of agents regulating inflammation by the epithelium may contribute significantly to the initiation and regulation of host innate immune response (5). L. pneumophila is known as one of the most important pathogens causing pneumonia in humans (1) and is shown to activate pulmonary tissue cells (3, 22). In a recent study, we demonstrated that L. pneumophila infection of A549 cells induced liberation of CC, CXC, Th1, and Th2 cytokines (4). However, little is known about the mechanisms leading to L. pneumophila-related cytokine expression per se and particularly in lung epithelial cells. Therefore, we decided to analyze mechanisms regulating L. pneumophila-related CXC chemokine IL-8 expression in more detail. IL-8 is a potent neutrophil recruiting and activating factor known to be of particular importance in inflammatory lung diseases (23, 24).

Although increase in IL-8 release after L. pneumophila stimulation was strong and significant, absolute concentrations were quite low. This is in part due to moderate bacterial concentrations used in the experiments. In additional experiments, SAEC were found to be more sensitive than A549 cells to L. pneumophila infection and produced higher IL-8 levels with lower infection concentrations. Nevertheless, the results regarding TSA and anacardic acid effects on IL-8 expression were qualitatively comparable. Teruya et al. (25) found similar IL-8 levels after infecting A549 or NCI-H292 cells. Therefore, observed IL-8 levels are comparable to other in vitro studies of primary lung epithelial cells and cell lines. In addition, Tateda et al. (26) reported mRNA expression of the functional IL-8 homolog KC in Legionella-infected mice lungs (RT-PCR) and observed expression of ∼2–3 ng/ml KC protein expression in infected mouse lung tissue (peak, 24 h p.i.). Furthermore, anti-KC as well as anti-CXCR2 (KC receptor) increased mortality in Legionella-infected mice indicated a biological effect of this molecule in Legionella pneumonia. Overall, these data underline a biological significance of IL-8/KC in Legionella pneumonia, although further in vivo studies are needed to clarify this point.

Increasing evidence indicates that histone modifications may be important for the transcriptional activity state of genes by loosening the DNA-histone interaction and unmasking of transcription factor binding sites (27). In chromatin, 146 bp of DNA are wrapped in 1.65 turns around a histone octamer (H2A, H2B, H3, H4) (13). Transcription repression or gene activation is regulated by specific covalent modifications of accessible N-terminal histone tails (28, 29) including acetylation (mostly lysine), phosphorylation (serine/threonine), and methylation (lysine) (30, 31).

In L. pneumophila-infected lung cells, we observed increased global, genome-wide acetylation of H4, and phosphorylation/acetylation of Ser¹⁰/Lys¹⁴-H3 whereas no effect on dimethylation of Lys9-H3 could be detected. Induced modifications could also be specifically observed at the IL-8 gene promoter. In line with this notion, acetylation of H4 and phosphorylation/acetylation of Ser¹⁰/Lys¹⁴-H3 was also observed in L. monocytogenes-infected human endothelial cells (12, 15) and in LPS-exposed dendritic cells (14), suggesting that these modifications may be of general importance in cells exposed to proinflammatory agents.

The acetylation status of histones is controlled by a delicate balance between HATs increasing histone acetylation, thereby reducing DNA-histone binding and facilitating gene transcription and HDACs acting in the opposing way (32, 33). Recruitment of the highly related HATs p300 and CBP to the IL-8 gene promoter was observed in L. pneumophila-infected epithelial cells. In addition, inhibition of global HAT activity by anacardic acid concentration-dependently reduced L. pneumophila-induced IL-8 expression. Together, these data indicate that HAT activity contributes to chromatin remodeling required for Legionella-related IL-8 expression. Recruitment of CBP to the IL-8 gene promoter was also observed in L. monocytogenes-infected endothelial cells (12) and seems to play a role in papilloma virus (34) and TNF-α-related IL-8 expression (35), suggesting that p300/CBP is critical for the induction of this cytokine.

NF-κB p65/RelA is known to be a critical transcription factor for IL-8 expression in general and in infected lung epithelial cells in particular (7, 36) and is recruited to the IL-8 gene promoter in L. pneumophila-infected A549 cells (this study). Inhibition of the canonical NF-κB pathway by IKK-NBD reduced L. pneumophila-induced IL-8 release and histone modifications. But NF-κB does not appear to be the only driver of IL-8 expression induced by L. pneumophila. Another important player could be AP-1 because it is known to participate in IL-8 gene regulation in general (37) and in bacterial infections in particular (9).

However, in our study, HAT binding and modifications of histones H3 and H4 at the endogenous IL-8 promoter could be detected earlier than NF-κB recruitment. This is in accordance with a study by Saccani et al. (14) in primary human dendritic cells: phosphorylation of H3 at the IL-8 promoter was noticed 30 min after LPS stimulation, but NF-κB was recruited 60 min after stimulation, indicating that histone modification enhances accessibility of the IL-8 promoter for NF-κB. Nevertheless, it cannot be ruled out that we, and Saccani, failed to detect a small amount of early recruited NF-κB initiating p300/CBP recruitment and chromatin preparation (38, 39). The mechanism of promoter targeting in these inflammatory signaling events remains to be elucidated in depth in further studies. In addition to their ability to acetylate promoter proximal nucleosomal histones, CBP and p300 can form complexes with Pol II and basal transcription factors like TATA-binding protein indicating the complex arrangements at the promoter (40, 41).

Besides the NF-κB pathway and histone modifications, p38 MAPK is suggested to contribute to proinflammatory gene expression in L. pneumophila-infected cells (4, 19) and to histone modification in bacterial infection (12). In accordance, inhibition of p38 by SB201290 reduced IL-8 expression as well as phosphorylation/acetylation of Ser¹⁰/Lys¹⁴-H3 and NF-κB p65 recruitment at the il8 gene promoter in L. pneumophila-infected cells. Whether histone modifications are regulated independently or coordinately by IKK2 and p38 MAPK remains to be clarified in further studies.

In contrast, HDACs suppress gene transcription. The 18 human HDACs are grouped into four classes (16). In A549 cells, we detected mRNA for class I HDAC1, 2, 3, and 8, as well as for class II HDAC4, 5, 6, and 7. Despite increasing HDAC5 expression, we observed no alteration of the expression levels of other tested HDACs in L. pneumophila-infected A549 cells. At the IL-8 gene promoter, temporary disappearance of HDAC1 from the promoter was observed whereas HDAC5 binding increased. Moreover, inhibition of HDACs by TSA further increased IL-8 protein expression in L. pneumophila-infected lung cells, indicating an important role of HDACs in regulation of IL-8 expression. However, gene knockdown of HDAC1 and HDAC5 clearly reduced HDAC1/5 protein levels but did not alter L. pneumophila-induced IL-8 release by A549 cells.

Because at least 11 histone deacetylases act to regulate the expression of distinct subsets of inflammatory/immune genes (16), other HDACs may also be involved in the regulation of the IL-8 gene promoter and compensate for the knockdown of HDAC1 or HDAC5, respectively. Nevertheless, data obtained in patients suffering from chronic obstructive lung disease (16) demonstrating decreased HDAC activity and the observation that Moraxella catarrhalis decreased HDAC expression and activity (21) furthermore indicate that HDACs may function in a specific way in inflammatory lung diseases. Thus, our data give a first hint that these histone modifications might be biologically important for Legionella-related host cell activation—at least with respect to IL-8 expression—but extensive further experimentation is needed to clearly dissect the involved molecules.

Further studies are needed to dissect the signaling pathways mediating L. pneumophila-related histone modifications, and kinase and transcription factor activation. From the pathogen’s view, several pathogenic factors, like flagellin and type II- or type IV-secreted effectors may be contributing to the response (42, 43). On the host site, pattern recognition receptors involved in Legionella detection like TLR2 (44, 45), TLR5 (46, 47), or cytosolic nucleotide-oligomerization domain proteins (17) may also contribute to the response.

Accordingly, flagellin deletion mutants of L. pneumophila induced histone acetylation, IL-8 gene transcription, as well as IL-8 release to a lower extent than wild-type bacteria. Flagellin is a potent inflammatory stimulus present in the flagellar structure of many bacteria (48). A common dominant TLR5 stop codon polymorphism abolished flagellin signaling and was associated with susceptibility to Legionnaires’ disease (46). Additionally, it has been shown that cytosolic recognition of flagellin by mouse macrophages restricted L. pneumophila infection (49) and flagellin-deficient L. pneumophila mutants evaded caspase-1- and Naip5-mediated macrophage immunity (50). In accordance with our finding of reduced IL-8 release, Hawn et al. (47) recently demonstrated that L. pneumophila-infected mice lacking TLR5 displayed lower neutrophil recruitment into the lung.

In summary, this report demonstrates that L. pneumophila impacts on gene transcription of host cells by inducing histone acetylation leading to proinflammatory gene transcription.

We greatly appreciate the excellent technical assistance of Frauke Schreiber, Jacqueline Hellwig (Department of Internal Medicine/Infectious Diseases and Pulmonary Medicine, Charité, Berlin, Germany), and Kerstin Rydzewski (Robert Koch Institut, Berlin, Germany). Part of this work will be included in the doctoral thesis of J. Lorenz.

The authors have no financial conflict of interest.