Genome-wide transcriptional profiling of mononuclear phagocytes recruited to mouse lungs in response to alveolar challenge with the TLR2 agonist Pam3CSK4

Maciej Cabanski, Jochen Wilhelm, Zbigniew Zasłona, Mirko Steinmüller, Ludger Fink, Werner Seeger, Jürgen Lohmeyer

Abstract

Compared with the Toll-like receptor 4 (TLR4) ligand LPS restricted to Gram-negative bacteria, few studies have addressed induction of lung inflammation and concomitant leukocyte recruitment in response to TLR2 ligands. This study is the first report showing that selective TLR2 stimulation by its ligand Pam3-Cys-Ser-Lys-Lys-Lys-Lys-OH (Pam3CSK4) within the alveolar compartment promoted lung inflammation in mice and induced the migration of circulatory immune cells including mononuclear phagocytes into the inflamed alveolar space. By using the transgenic CX3CR1+/GFP mouse strain for high-purity sorting of circulating and alveolar recruited mononuclear phagocytes together with SMART preamplification and whole genome oligonucleotide microarray techniques, we found that alveolar trafficking of mononuclear phagocytes was associated with profound changes of their gene expression profiles (∼900 differentially regulated genes postrecruitment). In particular, alveolar recruited mononuclear phagocytes showed upregulated transcripts of genes encoding cytokines/chemokines and pattern recognition receptor (PRR)-associated molecules. Notably, we observed a dynamic change of the genetic program of recruited mononuclear phagocytes obtained from bronchoalveolar lavage fluid at different time points (24 vs. 48 h) post-Pam3CSK4 challenge. In early alveolar recruited mononuclear phagocytes, mRNA levels of both proinflammatory (e.g., TNF-α, CCL2, and IL-6) and central anti-inflammatory/ proresolution [e.g., IL-1-receptor antagonist (IL-1RN), CD200 receptor (CD200R), IL-1 receptor-associated kinase (IRAK-M), IL-10, and Bcl-2-associated X protein (Bax)] mediators were found to be highly upregulated simultaneously. In corresponding cells recruited until later time points, transcript levels of anti-inflammatory/proresolution molecules persisted at the same level, whereas mRNA levels of proinflammatory mediators were found to decline. Collectively, our in vivo study identifies genetic programs by which alveolar recruited mononuclear phagocytes may contribute to the development and termination of pneumonia caused by Gram-positive bacteria.

  • Toll-like receptor

host sensing of pathogens invading the alveolar space is crucial to alert the pulmonary innate immune response. Toll-like receptors (TLRs) are one of the major pattern recognition receptor (PRR) classes involved in the detection of evolutionary conserved microbial components called pathogen-associated molecular patterns (PAMPs; Ref. 18). Ligand binding of TLRs leads to activation of inflammatory signal transduction pathways such as MAPKs and NF-κB. In consequence, a variety of inflammatory mediators including cytokines, chemokines, and other molecule classes are produced that promote the recruitment of immune cells to the site of infection (1, 3). We (1416) recently showed that alveolar challenge with bacteria or bacterial PAMPs was characterized by an early neutrophil influx followed by delayed recruitment of peripheral blood (PB) mononuclear phagocytes into the alveolar compartment.

Mononuclear phagocytes play a pivotal role in lung host defense to pathogens by activating both innate and adaptive immunity (24). Under noninflammatory steady-state conditions, circulating mononuclear phagocytes such as monocytes enter the alveolar air space to mature into resident alveolar macrophages (AMs) or myeloid dendritic cells (DCs) (6, 12). Bacterial infections or deposition of PAMPs in the alveolar compartment, however, result in heavily increased mononuclear phagocyte trafficking to the site of inflammation (24, 30). Before becoming macrophages or DCs, recruited mononuclear phagocytes may modulate local host-defense responses by secretion of inflammatory mediators, phagocytosis, and antigen presentation (24).

Recently, our group (30) addressed changes in the gene expression profile of mononuclear phagocytes in response to alveolar challenge with the TLR4 ligand LPS. Compared with the LPS restricted to Gram-negative bacteria, a limited number of studies have addressed the induction of lung inflammation and the concomitant leukocyte recruitment in response to PAMPs signaling via TLR2. Considering the fact that Gram-positive bacteria lacking LPS but expressing TLR2 ligands are the main cause of community-acquired pneumonia (CAP), there is urgent need to elucidate the activation programs of resident and lung-recruited immune cells including mononuclear phagocytes in response to TLR2 ligands. In particular, in vivo experimental data on global activation programs of mononuclear phagocytes recruited to the lungs are rare.

Therefore, in the present study, we examined the global gene expression profiles of mononuclear phagocytes recruited from PB to the alveolar space following alveolar deposition of the TLR2 ligand Pam3-Cys-Ser-Lys-Lys-Lys-Lys-OH (Pam3CSK4) in transgenic CX3CR1+/GFP mice suited to identify and sort PB and alveolar recruited mononuclear phagocytes to high purity. Compared with circulating mononuclear phagocytes, ∼900 genes were found to be differentially expressed in alveolar mononuclear phagocytes obtained from bronchoalveolar lavage (BAL) fluid (BALF) after intratracheal challenge with Pam3CSK4. Among these genes, ∼700 were found to be more active in alveolar recruited mononuclear phagocytes including not only pro- but also prominent anti-inflammatory mediators such as IL-1 receptor-associated kinase (IRAK-M), CD200 receptor (CD200R), and IL-10. Whereas the induced mRNA levels of selected proinflammatory mediators were found to decrease during later inflammation phases (48 vs. 24 h postchallenge), the increased transcript levels of central anti-inflammatory mediators including IL-1-receptor antagonist (IL-1RN) persisted. Alveolar trafficking of mononuclear phagocytes was also associated with marked upregulation of genes encoding enzymes implicated in PG and leukotriene (LT) metabolism such as cyclooxygenase (COX)-2 and of genes encoding molecules involved in tissue repair such as arginase 1, transforming growth factor-β induced (TGFβI), and fibronectin (FN1). Collectively, our results identify distinct genetic programs in mononuclear phagocytes recruited to the alveolar space at different time points post-Pam3CSK4 challenge by which these cells may contribute to both the development and resolution of lung inflammation.

MATERIALS AND METHODS

Animals.

CX3CR1GFP/GFP mice were generated on a mixed C57BL/6 × 129/Ola background by targeted disruption of CX3CR1 gene, as described previously (10). Parent CX3CR1GFP/GFP and CX3CR1+/+ mice were bred to yield heterozygous CX3CR1+/GFP offspring. Wild-type C57BL/6 mice (WT) were purchased from Charles River (Sülzfeld, Germany). All mice used in the study were between 8 and 12 wk old and were kept under specified pathogen-free conditions with free access to water and food. Experimental protocols involving animals were approved by institutional and local government committees.

Reagents.

Synthetic TLR2 agonist bacterial lipoprotein Pam3CSK4 was purchased from EMC Microcollections (Tübingen, Germany). RNeasy Micro Kit for total RNA isolation was purchased from Qiagen (Hilden, Germany). Whole Mouse Genome (44K) 60-mer oligonucleotide spotted microarray slides were obtained from Agilent Technologies (Böblingen, Germany). Cy3 and Cy5 dyes were purchased from PerkinElmer (Rodgau-Jügesheim, Germany). Random hexamers primers were purchased from Boehringer Mannheim (Mannheim, Germany). Moloney murine leukemia virus (MMLV) RT and RNase inhibitor were obtained from Promega (Mannheim, Germany). SMART amplification kit was purchased from Clontech (Saint-Germain-en-Laye, France). Platinum SYBR Green I qPCR SuperMix-UDG and DTT were purchased from Invitrogen (Karlsruhe, Germany).

Treatment of animals.

Intratracheal administration of Pam3CSK4 ligand was performed as described previously for LPS installation (31). Briefly, WT or CX3CR1+/GFP mice were anesthetized with tetrazoline hydrochloride and ketamine, and the trachea was exposed. Subsequently, catheter (Abbot, Wiesbaden, Germany) was inserted into trachea, and Pam3CSK4 (50 μg per mouse dissolved in total volume of 60 μl) was installed under stereomicroscopic control (MS5; Leica Microsystems, Wetzlar, Germany). After installation, wounds were closed, and mice were allowed to recover with free access to food and water.

Collection and analysis of blood and BALF.

Mice were killed with an overdose of isoflurane (Forene; Abbott, Wiesbaden, Germany), and blood and BALF collection was performed as described previously (31). Briefly, blood was collected from the vena cava inferior. The lysis of red blood cells was performed using ammonium chloride solution. For BAL, trachea was exposed, and a small incision was made to insert a shortened 21-gauge cannula connected to a 1-ml insulin syringe, followed by repeated intratracheal instillations of 0.5-ml aliquots of PBS (pH 7,2, supplemented with 2 mM EDTA). BAL was performed until a total volume of 5 ml was recovered. BALF was spun for 10 min at 1,400 rpm at 4°C, and cells were resuspended in 1 ml of RPMI 1640 medium containing 10% FCS and l-glutamine.

To study the inflammatory BALF profile, the cells were counted with a hemocytometer, and quantification of macrophages and polymorphonuclear neutrophils was done on differential cell counts of Pappenheim-stained cytocentrifuge preparations as recently described (31).

Flow cytometric analysis and cell sorting.

Recovered BALF and blood cells were pooled from 10 to 12 CX3CR1+/GFP mice to obtain sufficient cell numbers for flow sorting. Before cell sorting, blood and BALF samples were filtered through a 40-μm cell strainer. The green fluorescent protein (GFP)-positive mononuclear phagocytes were counted and sorted using a high-speed FACSVantage SE flow cytometer (BD Biosciences) as described previously (30). In short, sorting of GFP-positive PB and alveolar mononuclear phagocytes was performed first according to their different forward scatter area (FSC-A) vs. side scatter area (SSC-A) characteristics, and second FSC-A vs. fluorescence 1 area (FL1-A) [fluorescence intensity emitted at 525 nm (F525) ± 15 nm; FITC-A/GFP] characteristics, and finally according to FL1 vs. fluorescence 2 area (FL2-A) characteristics [F575 ± 25 nm; phycoerythrin area (PE-A)]. These settings allowed for the exclusion of lymphocytes, AMs, and neutrophils in analyzed samples. The purity of sorted GFP-positive cells was ≥98%. In addition, Pappenheim staining verified a monocytic cell population of isolated GFP-positive cells (data not shown).

Surface expression of CD200R on alveolar and PB mononuclear phagocytes was analyzed 24 and 48 h after pulmonary deposition of Pam3CSK4 using anti-CD200R-PE (OX-110; Serotec) antibody at 1:20 dilution. After staining at 4°C in the dark for 20 min, cells were washed three times in PBS/2 mM EDTA/0.5% FCS and subsequently subjected to flow cytometric analysis using a FACSCanto flow cytometer (BD Biosciences) and gating as described above.

Isolation of total RNA.

Total cellular RNA was isolated from 30,000 to 40,000 freshly sorted GFP-positive mononuclear phagocytes using the RNeasy Micro Kit. RNA was eluted with 12 μl of RNase-free H2O. RNA was quantified spectrophotometrically (Nanodrop ND-100), and the quality was assessed by the 18S/28S rRNA bands in capillary electrophoresis (2100 Bioanalyzer; Agilent Technologies).

Microarray analysis.

Four independent microarray experiments were performed using groups of 10–12 animals. Each time, BALF and blood cells were pooled to obtain sufficient number for flow sorting, followed by RNA isolation. Purified total RNA was subjected to amplification and Cy labeling using the SMART kit following the kit instructions. Briefly, this method creates full-length cDNA from all oligo(A) RNA with introduced specific PCR primer binding sites at either end. One priming site is introduced by the oligo(dT) primer, and the other one is linked by template-switching of the MMLV RT: the RT adds a stretch of Gs to the end of the cDNA when the end of the RNA template is reached. A C-rich oligodeoxyribonucleotide containing the PCR primer sequence binds to this stretch. The RT then switches the template and produces a sense cDNA strand that now contains the binding site for the PCR primer at either side. These full-length cDNAs are subsequently amplified by PCR. In preliminary experiments, the number of amplification cycles was determined so that the PCR is not run until the plateau. All samples were finally amplified for 24 cycles, followed by 4 cycles with aminoallylated dCTP. After purification, monoreactive Cy dyes (Amersham, Little Chalfont, United Kingdom) were coupled to the incorporated amino groups.

Cy3- and Cy5-labeled cDNAs were hybridized overnight to 44K 60-mer oligonucleotide spotted microarray slides (4x44k whole mouse genome, G4122F; Agilent Technologies). Hybridization and subsequent washing and drying of the slides were performed following the Agilent Technologies hybridization protocol. The total number of 4 hybridizations (all based on biologically independent samples) were performed in a balanced dye-swap design.

Dried slides were scanned using the GenePix 4100A scanner (Axon Instruments). Image analysis was performed with GenePix Pro 5.0 software, and calculated values for all spots were saved as GenePix results files. Stored data were evaluated using the R software (http://www.r-project.org) and the limma package from BioConductor (7, 26). The spots were weighted for subsequent analyses according to the spot intensity, homogeneity, and saturation. The spot intensities were corrected for the local background using the method of Edwards (5) with an offset of 64 to stabilize the variance of low-intensity spots. The M/A data were LOESS normalized before averaging (28). Genes were ranked for differential expression using a moderated t-statistic (27). Candidate lists were created on the basis of a 5% false discovery rate (FDR). In addition, a 10% FDR evaluation was performed (data not shown).

Gene Set Enrichment Analysis (GSEA) was applied using the function geneSetTest to identify pathways with enriched numbers of regulated genes (20, 26). Pathway annotation was taken from Kyoto Encyclopedia of Genes and Genomes (KEGG) database package. Only pathways with >3 genes were considered. The tests were preformed based on the t-statistics including both up- and downregulated genes. P values were adjusted for multiple testing using the procedure of Benjamini and Hochberg (2).

Identification of important transcription factor binding sites (TFBSs) was performed using Distinct Regulatory Elements of coregulated genes (DiRE; Ref. 9). Briefly, the software predicts function-specific regulatory elements consisting of clusters of specifically associated TFBSs within a set of coregulated genes. The analysis was done independently for candidate genes (selected with 5% FDR) with ≥2-fold higher expression in each group (PB and alveolar mononuclear phagocytes).

Raw data was submitted to the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) database. The data set is available under acc. no. GSE11978.

Relative mRNA quantification by real-time PCR.

To confirm results obtained by microarrays, the regulation of a subset of genes was analyzed by real-time quantitative PCR as described previously (30). Four to five independent experiments were performed using groups of 10–12 animals. Each time, BALF and blood cells were pooled to obtain sufficient number for flow sorting followed by RNA isolation. In short, total RNA was reverse-transcribed using MMLV RT. Real-time quantitative PCR was performed using the 7900 HT (Applied Biosystems) and Platinum SYBR Green I qPCR SuperMix-UDG for detection of PCR products. Cycling conditions were as following: 1 cycle of 95°C for 5 min, followed by 45 cycles of 95°C for 10 s, 60°C for 30 s, and 72°C for 10 s. After cycling process, melting curve analysis and agarose gel electrophoresis (in selected cases) was performed to confirm the exclusive amplification of the expected PCR.

Relative gene expression is expressed as change in cycle threshold (ΔCT) with mouse porphobilinogen deaminase (PBGD) serving as reference gene: ΔCT = CT(PBGD) − CT(target gene) (higher values indicate higher mRNA levels). Differential gene expression between conditions is expressed as ΔΔCT what corresponds to the log2 fold difference in mRNA levels between the conditions compared (13). The oligonucleotide primer pairs were designed using Primer Express 2.0 (Applied Biosystems) and were synthesized by Metabion (Table 1).

View this table:
Table 1.

Primers used for real-time PCR validation of the transcriptional profiles

Statistical analysis.

The data for BALF cell profile are presented as means ± SD. Differences between treatments were analyzed using Dunnett's post hoc test after significant ANOVA (P ≤ 0.05).

ΔCT and ΔΔCT values are given as means ± SE of ≥4 independent experiments. Differences in mean ΔCT values were tested by 2-tailed t-tests (paired or unpaired where appropriate). A value of P ≤ 0.05 was considered significant.

Mean fluorescence intensities (MFIs) for CD200R-PE antibody staining (MFI CD200R-PE − MFI isotype-PE control) are given as means ± SE of 5 mice per treatment group. Differences in mean MFI values were tested by 2-tailed t-tests. A value of P ≤ 0.05 was considered significant.

RESULTS

Intratracheal deposition of Pam3CSK4 induces lung inflammation and alveolar trafficking of circulating mononuclear phagocytes.

To determine the capacity of the synthetic TLR2 agonist Pam3CSK4 to induce lung inflammation in vivo, WT C57BL/6 mice received intratracheal application of Pam3CSK4 (50 μg per mouse), and BALF cellular constituents were monitored at various time points postchallenge. Pulmonary administration of Pam3CSK4 provoked rapid recruitment of neutrophils into the alveolar compartment that was detectable as early as at 6 h, peaked at ∼24 h after challenge, and significantly declined until 48 h. The number of BALF macrophages was found to be slightly increased at 24 h postchallenge without reaching statistical significance (Fig. 1). To further assess alveolar trafficking of mononuclear phagocytes, we employed transgenic CX3CR1+/GFP mice expressing GFP under control of the CX3CR1 promoter as described before (10, 30). This approach allowed identifying and further high-purity sorting of GFP-positive mononuclear phagocytes from the PB and of freshly recruited mononuclear phagocytes from BALF as described in materials and methods. Heterozygous CX3CR1+/GFP mice were either left untreated or received intratracheal application of the TLR2 ligand Pam3CSK4 (50 μg per mouse), and recruited mononuclear phagocytes were assessed in BALF after 24 or 48 h. As illustrated in Fig. 2, very few GFP-positive cells (0.3–0.5%) were detected in BALF collected from untreated mice. In contrast, at 24 h and even more pronounced at 48 h after intratracheal Pam3CSK4 application, we observed significantly increased proportions of GFP-positive cells in BALF samples (1.4–1.7 and 2.5–3.5%, respectively). These data demonstrate that intratracheal application of the TLR2 ligand Pam3CSK4 induced enhanced trafficking of circulating mononuclear phagocytes into the alveolar space.

Fig. 1.

Inflammatory cell accumulation in the bronchoalveolar compartment in response to alveolar deposition of the Toll-like receptor 2 (TLR2) ligand Pam3-Cys-Ser-Lys-Lys-Lys-Lys-OH (Pam3CSK4). C57BL/6 WT mice were either left untreated (0 h) or challenged intratracheally with Pam3CSK4 (50 μg per mouse) for 6, 12, 24, or 48 h. After indicated times, bronchoalveolar lavage fluid (BALF) cells were isolated, and microscopic differentiation and quantification of macrophages and alveolar recruited neutrophils was performed as described in materials and methods. Means of different time points (5 mice per treatment group) were compared using the ANOVA test as described in materials and methods. *P ≤ 0.01. n.d, Not detected.

Fig. 2.

Lung recruitment of mononuclear phagocytes in the absence or presence of the Pam3CSK4-induced inflammation. A representative fluorescence-activated cell sorter (FACS) profile of BALF cells collected from CX3CR1+/GFP mice either left untreated (left column) or intratracheally challenged with Pam3CSK4 (50 μg per mouse) for 24 h (middle column) or 48 h (right column). At the indicated time points, mice were killed, and BALF and blood (data not shown) were collected followed by fluorescence-activated cell sorting of peripheral blood (PB) and alveolar mononuclear phagocytes as described in materials and methods. Briefly, alveolar mononuclear phagocytes were gated according to forward scatter area (FSC-A) vs. side scatter area (SSC-A) (gate R1, living cells), FSC-A vs. fluorescence 1 area (FL1-A) [FITC-A; gate R2, green fluorescent protein (GFP)-positive mononuclear phagocytes], and FL1-A vs. fluorescence 2 area (FL2-A) [phycoerythrin area (PE-A); gate R3, GFP-positive mononuclear phagocytes among R2]. Data show 1 representative of 4 independent experiments.

Alveolar Pam3CSK4 challenge induces global genome expression changes in mononuclear phagocytes recruited to the alveolar space.

To investigate global changes in the gene expression profiles of mononuclear phagocytes recruited to Pam3CSK4-exposed lungs during maximal inflammation, 10–12 heterozygous CX3CR1+/GFP mice received intratracheal application of the TLR2 ligand Pam3CSK4 (50 μg per mouse). Twenty-four hours postchallenge, GFP-expressing mononuclear phagocytes were isolated in high purity from both PB and BALF, and comparative global gene expression profiles of circulating and alveolar recruited mononuclear phagocytes were determined using SMART amplification and whole mouse genome oligonucleotide arrays. We found that the global transcriptome of mononuclear phagocytes recruited to the alveolar compartment at 24 h in response to Pam3CSK4 was dramatically changed compared with circulating mononuclear phagocytes obtained at the same time point (Fig. 3). Approximately 900 genes differentially expressed between PB and alveolar recruited mononuclear phagocytes were identified (5% FDR), the majority of which (∼700) were upregulated in the alveolar mononuclear phagocyte population. This group contains a broad panel of genes encoding proinflammatory cytokines previously described to be robustly upregulated by different microbial components in various cell types such as IL-1α, TNF-α, IL-6, and IL-12A (Table 2). We additionally observed upregulated mRNA levels of chemokines such as CCL2 and CCL3, whereas the gene expression of the neutrophil and monocyte chemoattractant CXCL7 was significantly downregulated in alveolar mononuclear phagocytes.

Fig. 3.

Global gene expression differences of alveolar recruited vs. PB mononuclear phagocytes after intratracheal application of Pam3CSK4. The plot shows the average log2 signal ratio (M values) plotted against the average log2 of the spot intensity (A values). Positive (negative) M values indicate an up (down)-regulation on alveolar recruitment. The black dots mark candidate genes identified as differentially regulated by a moderated t-test using a false discovery rate of 5%.

View this table:
Table 2.

Differential gene expression between alveolar recruited mononuclear phagocytes and peripheral blood mononuclear phagocytes

Notably, this Pam3CSK4-elicited proinflammatory activation in freshly alveolar recruited mononuclear phagocytes coincided with pronounced induction of anti-inflammatory mediators such as IRAK-M and IL-10. Other upregulated genes related to limitation and resolution of inflammation were CD200R, formyl peptide receptor (FPR-1), IL-18 binding protein (IL-18BP), adenosine A2b receptor (ADORA2B), annexin 1 (ANXA-1) and Bcl-2-associated X protein (Bax). Moreover, alveolar mononuclear phagocytes recruited after intratracheal Pam3CSK4 application contained elevated mRNA levels of genes associated with an alternative activated macrophage phenotype such as arginase 1, TGFβI, and FN1.

In addition, alveolar trafficking of mononuclear phagocytes in response to Pam3CSK4 elicited marked induction of genes encoding enzymes involved in PG and LT metabolism such as COX-2 as 12-hydroxydehydrogenase (LTB4DH). Opposite downregulation was observed for arachidonate 12-lipoxygenase (ALOX12) transcript.

Finally, a large number of genes encoding surface receptors were differentially regulated in mononuclear phagocytes recruited to the lung by alveolar exposure to Pam3CSK4. Thus we observed elevated transcript levels in alveolar mononuclear phagocytes for PRR genes such as TLR2, CLEC4N, and scavenger receptor (SR)-A. Moreover, alveolar trafficking of mononuclear phagocytes was found to be associated with marked upregulation of some of G protein-coupled receptors such as GPR89, GPR137B, and GPR178.

GSEA analysis was used to determine which functional pathways are likely involved during alveolar recruitment of mononuclear phagocytes after Pam3CSK4 administration. Taking into consideration all genes selected with 5% FDR, we identified pathways as being predominantly related to inflammation and metabolism processes, including proteasome pathway, TLR signaling pathway, cytokine-cytokine receptor interaction pathway, and purine/pyrimidine metabolism pathway (Table 3). A list of differentially regulated genes of enriched pathways is included as Supplemental Table S1 (available in the data supplement online at the AJP-Lung Cellular and Molecular Physiology web site).

View this table:
Table 3.

Identification of functional pathways using GSEA analysis

To discover putative transcription factors (TFs) associated with TLR2 ligand-induced recruitment of mononuclear phagocytes, we searched for overrepresented promoter sites among genes differentially expressed using DiRE server. For the analysis, we considered only genes with a coefficient value ≥1 or less than or equal to −1. Table 4 lists the 10 TFs with the highest importance found in regulatory elements among differentially upregulated (higher expression in alveolar mononuclear phagocytes) as well as downregulated genes (higher expression in PB mononuclear phagocytes). The top ranked TFs were NF-κBp65 and NANOG in alveolar recruited cells and ETS1 and BACH2 in PB mononuclear phagocytes.

View this table:
Table 4.

Enriched putative transcription factors among differentially expressed genes after alveolar recruitment of mononuclear phagocytes

To verify the microarray findings, we performed quantitative real-time PCR for selected genes preferentially related to inflammatory processes without SMART preamplification. In addition, the IL-1RN gene was included that was differentially expressed in the microarray analysis when evaluated on the basis of 10% FDR (data not shown). The results obtained by real-time PCR largely matched with the microarray-based gene expression profiles (Fig. 4). However, real-time PCR data did not provide any evidence for differential expression of FN1 and ANXA-1. The mRNA levels measured by real-time PCR tended to be higher than gene expression levels determined by arrays, which is expected and due to the preamplification and the lower dynamic range of the microarray signals (35). The mRNA levels of IL-10 were below the detection limit in PB mononuclear phagocytes but could be consistently detected in alveolar mononuclear phagocytes in four independent experiments, indicating the upregulation of these genes on recruitment. Vice versa, ALOX12 mRNA levels found to be below the detection limit in alveolar mononuclear phagocytes in four separate experiments were consistently detectable in peripheral mononuclear phagocytes indicating ALOX12 downregulation on recruitment. Of interest, IL-1RN mRNA levels were below the detection limit in PB mononuclear phagocytes but could be consistently detected in alveolar mononuclear phagocytes, although this gene was not identified as differentially regulated by microarray analysis when performed on the basis of a 5% FDR.

Fig. 4.

Validation of differential gene expression profiles of alveolar recruited vs. PB mononuclear phagocytes following alveolar deposition of Pam3CSK4. CX3CR1+/GFP mice were either left untreated or intratracheally administrated with Pam3CSK4 (50 μg per mouse) for 24 h. Four independent experiments were performed using groups of 10–12 animals. Each time, BALF and blood cells were pooled to obtain sufficient number for flow sorting followed by RNA isolation, and the relative mRNA level of each gene was determined using real-time PCR as described in materials and methods. The coefficients values from microarray analysis (gray bars) were compared with ΔΔCT values from real-time PCR (white bars). Microarray and PCR data are given as means ± SE. Mean change in cycle threshold (ΔCT) of mononuclear phagocytes isolated from blood and alveolar compartment were compared using t-test. *P ≤ 0.05. ANXA-1, annexin 1; Bax, Bcl-2-associated X protein; IRAK-M, IL-1 receptor-associated kinase; FPR-1, formyl peptide receptor; LCN2, lipocalin 2; COX-2, cyclooxygenase-2; LTB4DH, 12-hydroxydehydrogenase; SPP1, secreted phosphoprotein 1; FN1, fibronectin.

Attenuation of proinflammatory gene expression levels in alveolar mononuclear phagocytes recruited during late inflammation resolution phase.

In our model of TLR2 ligand Pam3CSK4-initiated lung inflammation, we observed a significant decrease of recruited neutrophils together with an increase of mononuclear phagocyte infiltrates at 48 h compared with 24 h postchallenge (Figs. 1 and 2). Since this shift is considered to be a key histological hallmark of inflammatory infiltrate resolution, we decided to determine whether mononuclear phagocytes recruited until 48 h postchallenge differentially regulate inflammatory mediators compared with corresponding cells recruited until 24 h. Therefore, we monitored changes in the gene expression profiles of molecules related to inflammation of recruited mononuclear phagocytes recovered from BALF after 48 h compared with 24 h after intratracheal application of Pam3CSK4.

Interestingly, we noticed that mRNA levels of proinflammatory molecules such as TNF-α, IL-6, CCL2, and IL-12A were significantly lower in recruited mononuclear phagocytes isolated from BALF at 48 h compared with 24 h, whereas the data did not provide evidence for differential expression of gene with key anti-inflammatory and resolution function such as IRAK-M, IL-10, IL-1RN, Bax, and ANXA-1 (Fig. 5).

Fig. 5.

Time-dependent changes in the gene expression profile of alveolar recruited mononuclear phagocytes. Mice received intratracheal application of Pam3CSK4 (50 μg per mouse) 24 or 48 h before analysis. Four to five independent experiments were performed using groups of 10–12 animals. Each time, BALF and blood cells were pooled to obtain sufficient number for flow sorting followed by RNA isolation, and the relative mRNA expression of each gene was determined using real-time PCR as described in materials and methods. Real-time quantitative PCR data (ΔCT) are given as means ± SE calculated from ≥4 individual experiments. Mean ΔCT of samples collected at different time points posttreatment were compared using t-test. *P ≤ 0.05.

We selected the CD200R gene product to evaluate whether its increased mRNA transcript levels in lung-recruited mononuclear phagocytes are correlated with increased CD200R protein levels. In agreement with the mRNA expression data, alveolar recruited mononuclear phagocytes expressed significantly higher amounts of CD200R protein on the cell surface than respective PB mononuclear phagocytes at 24 and 48 h after pulmonary Pam3CSK4 administration (Fig. 6).

Fig. 6.

Flow cytometric analysis of CD200 receptor (CD200R) surface expression on alveolar recruited vs. PB mononuclear phagocytes. Alveolar recruited and PB mononuclear phagocytes were stained by PE-labeled anti-CD200R antibody at 24 and 48 h after pulmonary deposition of Pam3CSK4 and gated as described in Fig. 2. Bars represent mean fluorescence intensity (MFI) of CD200R-stained cells (MFI CD200R-PE − MFI isotype-PE control) calculated from 5 mice per treatment group. The MFI differences between alveolar recruited and PB mononuclear phagocytes were compared using t-test. *P < 0.05.

DISCUSSION

In this study, we used high-purity flow sorting, SMART preamplification, and microarray transcriptome profiling to identify genes regulated in mononuclear phagocytes that are recruited to the alveolar space in response to intratracheal deposition of the TLR2 ligand Pam3CSK4. We identified ∼900 genes differentially expressed, mostly upregulated in alveolar recruited compared with PB mononuclear phagocytes, demonstrating that inflammatory mononuclear phagocyte trafficking into lung tissue was associated with a large-scale gene activation program. Notably, the mononuclear phagocyte gene clusters undergoing activation in response to intratracheal Pam3CSK4 application encoded both pro- and anti-inflammatory activities, however, with a profound profile change over time. Mononuclear phagocytes recruited within 24 h postchallenge showed simultaneous induction of both pro- and anti-inflammatory genes, whereas in mononuclear phagocytes isolated from BALF at 48 h postchallenge, mainly genes encoding anti-inflammatory mediators were found to persist in an upregulated state.

The induction of lung inflammation by applying the Gram-negative bacterial component LPS signaling via TLR4 into the respiratory tree has been thoroughly investigated (15, 30). In contrast, this study is the first report showing that selective TLR2 stimulation within the alveolar compartment likewise promoted lung inflammation in mice and induced the migration of circulatory immune cells including mononuclear phagocytes into the inflamed alveolar space. This model seems of considerable interest because most cases of CAP are caused by Gram-positive bacteria lacking LPS but expressing TLR2 ligands.

Under steady-state condition, half of the circulating mononuclear phagocytes leave the bloodstream each day to enter peripheral tissues and differentiate into tissue macrophages. Inflammatory stimuli such as invading pathogens or TLR ligands, however, elicit increased mononuclear phagocyte traffic to the sites of inflammation where the recruited cells can modulate local host responses (6). Here, we show that mononuclear phagocytes freshly recruited in response to the TLR2 ligand indeed express a broad array of proinflammatory molecules supporting the concept that these cells bear the potential to promote lung inflammation. However, Pam3CSK4 challenge was also found to induce pronounced anti-inflammatory mononuclear phagocyte responses with upregulation of IRAK-M, IL-10, IL-1RN, CD200R, ANXA-1, Bax, IL-18BP, and FPR-1 mRNA levels.

IRAK-M represents a central host strategy evolved to dampen inflammatory TLR signaling and IRAK-M gene deletion increases inflammation during bacterial infections (11). Augmented IRAK-M transcript levels found in alveolar recruited mononuclear phagocytes both at 24 and 48 h after Pam3CSK4 challenge and in circulating mononuclear phagocytes after 48 h compared with 24 h (data not shown) suggest that activation of IRAK-M may contribute to limit TLR2 ligand-induced inflammatory responses both in the circulation and in the lungs.

LPS-induced secretion of IL-10 limits the production of proinflammatory cytokines including TNF-α, IL-1, and IL-12, thereby promoting the termination of inflammation (4). IL-1RN also produced by monocytes/macrophages in response to LPS acts as a specific inhibitor of IL-1α/β signaling by competitive binding to the IL-1 receptor (21). The strong and up to 48-h persisting increase of mononuclear phagocyte IL-10 and IL-1RN transcripts observed after intratracheal Pam3CSK4 administration suggest a potential role for both molecules in limiting also TLR2-driven lung inflammatory responses.

CD200R is a membrane receptor almost exclusively expressed on myeloid cells including DCs and macrophages (19). Recently, Snelgrove and colleagues (29) reported that alveolar epithelial cells can suppress AM activity through the interaction of CD200R expressed on macrophages with its ligand CD200 expressed by airway epithelium. Increased levels of both CD200R transcripts and protein surface expression observed in alveolar recruited mononuclear phagocytes after Pam3CSK4 administration may indicate a possible function of CD200R in attenuating the TLR2-driven lung inflammatory cascade.

Computational analysis of differentially regulated genes revealed the expected enrichment of genes involved in TLR signaling as well as cytokine-cytokine receptor interaction pathways. A second major category highlighted by GSEA was represented by metabolic pathways. In particular, we identified that alveolar recruitment of mononuclear phagocytes was associated with increased expression of genes encoding subunits of the proteasome. This multifunctional enzymatic complex is essential for a number of important cellular functions. In immune responses, proteasome is involved in proper antigen processing and degradation of TFs including NF-κB and hypoxia-inducible factor-1α (22, 32). It has been reported that activation of murine macrophages with LPS enhances the gene expression of proteasome subunits as well as proteasomal activity. Moreover, pretreatment of macrophages with the proteasome inhibitor lactacystin resulted in significant downregulation of LPS-regulated genes including TNF-α, IL-6, IL-12p40, and COX-2 (25). Therefore, differential mononuclear phagocyte expression of proteasome subunits may contribute to balancing lung inflammatory responses.

Our gene array data further revealed that inflammatory trafficking of mononuclear phagocytes was associated with profound regulation of genes involved in PG and LT metabolism such as COX-2. COX-2 catalyzes PG synthesis, including PGE2, commonly considered as central proinflammatory lipid mediator involved in the pathogenesis of several inflammatory diseases such as periodontitis and rheumatoid arthritis (17, 34). In the lungs, however, it has been reported to display anti-inflammatory and proresolution activities as well (33) and shown to downregulate the production of proinflammatory cytokines such as IL-12 or CCL2, thus inhibiting inflammatory cell recruitment (23). In addition, recruited mononuclear phagocytes may actively contribute to lowering the local levels of lung lipid inflammatory mediators by increased expression of the genes encoding the enzymes LTB4DH involved in deactivation of the polymorphonuclear neutrophil chemoattractant LTB4 and 15-hydroxyprostaglandin dehydrogenase (HPGD) promoting the deactivation of PGs and lipoxins.

Recently, macrophages have been assigned to two groups: classically activated with proinflammatory activities and alternatively activated with anti-inflammatory/repair capacities (8). In our study, we found that intratracheal administration of Pam3CSK4 induced sustained mRNA level elevation of genes preferentially associated with an alternatively activated macrophage phenotype known to play a major role in tissue repair during the healing phase by secretion of ECM components and release of anti-inflammatory mediators (8).

Recruitment of mononuclear phagocytes from the circulation to the alveolar space is a multistep process involving multiple cellular interactions. In addition, after trafficking to the air space, mononuclear phagocytes become exposed to the specialized alveolar microenvironment stimulated by PAMP-induced inflammation with the potential to further modulate the phenotype of the recruited cell population. Our global gene profiling approach does not allow assigning the observed changes in the genetic program of lung-recruited mononuclear phagocytes to any of these migratory steps. We also cannot discern whether the changes in gene activation found in alveolar recruited mononuclear phagocytes at 48 vs. 24 h are due to phenotype changes occurring in mononuclear phagocytes within the alveolar compartment or reflect the recruitment of functionally different mononuclear phagocytes during later stages that have acquired altered functional programs either within the vascular compartment or during the trafficking process. The fact, however, that the gene expression profiles of circulating mononuclear phagocytes obtained at 48 vs. 24 h also display a shifted profile suggests that indeed the extravasation of functionally modified mononuclear phagocytes may contribute to the altered gene expression profile of BALF mononuclear phagocytes at later time points after inflammatory challenge.

In conclusion, this is the first report that demonstrates the feasibility to evaluate cell-specific transcriptome profiles in highly purified flow-sorted mononuclear phagocytes from both PB and BALF in a mouse model of TLR2 ligand-induced lung inflammation. This approach identifies dynamic genetic program changes in mononuclear phagocytes recruited to the lung after alveolar deposition of TLR2 ligands that are also present in LPS-lacking Gram-positive bacteria. It further identifies TF binding motives that are possibly involved in the recruitment-induced gene expression changes in mononuclear phagocytes. This impacts on our understanding of how alveolar recruited mononuclear phagocytes may contribute to the development and termination of pneumonia caused by Gram-positive bacteria.

GRANTS

The current study has been supported by the German Research Foundation Grant 547, “Cardiopulmonary Vascular System,” the National Network on Community-Acquired Pneumonia (CAPNETZ), and the Bundesministerium für Bildung und Forschung (BMBF)-funded clinical research group in clinical infectiology, BMBF 01 KI 0770.

ACKNOWLEDGMENTS

We acknowledge the excellent technical support by Emma Braun, Petra Janssen, and Maria M. Stein.

Footnotes

  • *M. Cabanski and J. Wilhelm contributed equally to this study.

REFERENCES

View Abstract