The Inflammatory versus Constitutive Trafficking of Mononuclear Phagocytes into the Alveolar Space of Mice Is Associated with Drastic Changes in Their Gene Expression Profiles ¹

Mononuclear phagocytes are known to contribute to both acute and chronic inflammatory diseases of the lung, including acute respiratory distress syndrome (ARDS) ³, bronchiolitis obliterans, and idiopathic pneumonia syndrome (1, 2, 3, 4, 5). In addition, we recently demonstrated that monocytes may act as regulators of the neutrophilic response in a mouse model of acute lung inflammation (6, 7). However, the molecular mechanisms and potential candidate genes that might regulate the accessory function of mononuclear phagocytes in acute lung inflammation have not been characterized. The lack of potent purification protocols allowing the isolation of monocytic cells from both peripheral blood and alveolar compartments may partially explain the lack of currently available data addressing the changes in gene expression profiles of these cells transmigrating from the vascular into the alveolar compartment under baseline vs inflammatory conditions.

Previous studies from our laboratory made use of the lipophilic intravital dye, PKH26-PCL, to discriminate resident alveolar macrophages (strongly PKH26 positive) from newly alveolar-recruited monocytic cells (PKH26 dull) to study their recruitment pathways during lung inflammation (6, 8, 9). However, this method did not stain circulating blood monocytes to allow their subsequent purification for molecular and functional characterization. Therefore, in the current study, we made use of a novel transgenic mouse strain (CX₃CR1^+/GFP) that allows the identification and subsequent FACS of both circulating and alveolar recruited mononuclear phagocytes to determine changes in their gene expression profiles during their recruitment into the alveolar compartment under both baseline and acute lung inflammatory conditions. In CX₃CR1^+/GFP mice, one allele for the gene encoding CX₃CR1, the receptor for the membrane-tethered chemokine fractalkine (CX3CL1, Fkn) is replaced by the gene encoding GFP (10). Because the transgene GFP is under the control of the CX₃CR1 gene promoter and because CX₃CR1 is homogeneously expressed on circulating monocytes but not on resident alveolar macrophages, CX₃CR1^+/GFP mice were used in the current study to track, sort, and genotype mononuclear phagocytes during their migration into the alveolar air space and, at the same time, allowing their clearcut discrimination from differentiated, resident alveolar macrophages and most other inflammatory elicited leukocyte subsets.

DiVa-assisted FACS analysis of bronchoalveolar lavage (BAL) fluid cells collected from CX₃CR1^+/GFP mice enabled us for the first time to detect, sort, and transcriptionally profile constitutively migrated alveolar mononuclear phagocytes under noninflammatory conditions. In addition, we demonstrate that alveolar mononuclear phagocyte recruitment in response to the monocyte chemoattractant, CCL2, is associated with profound changes in their gene expression profiles. Finally, we provide evidence that mononuclear phagocytes recruited into the lungs of mice in response to CCL2 in the presence of low endotoxin challenge exhibit an activated, highly proinflammatory genotype, characterizing these cells as powerful cellular contributors of the lung inflammatory response. The current technical approach may help to identify candidate genes regulating the functional role of mononuclear phagocytes in both acute and chronic lung inflammatory conditions.

Heterozygous (CX₃CR1^+/GFP) mice were generated on a mixed C57BL/6 × 129/Ola genetic background. BALB/c CX₃CR1^+/GFP mice were derived from repeated backcrosses (N9) into the BALB/c background (10). Parent CX₃CR1^GFP/GFP and CX₃CR1^+/+ mice were bred to yield offspring with the CX₃CR1^+/GFP genotype, which were then used in all the experiments at 8–12 wk of age. All mice were bred and kept under specific pathogen free conditions with free access to food and water. All animal experiments were approved in accordance with the guidelines of our Institutional Animal Care and Use Committee.

Atlas Mouse 1.2 II nylon array membranes, BD Supersmart mRNA amplification kit and BD Atlas SMART fluorescent probe amplification kit were purchased from BD Biosciences. Nylon membrane processing was performed according to the manufacturer’s instructions. The DNA isolation and Qiaquick PCR purification kit were purchased from Qiagen. The murine CCL2 protein, the homologue of the human MCP-1 gene product (JE/MCP-1), was purchased as a recombinant protein preparation from PeproTech and was routinely ascertained to be free of endotoxin as analyzed with the Coatest amoebocyte lysate assay (detection limit < 10 pg/ml; Chromogenix). Escherichia coli LPS (0111:B4) was purchased from Sigma-Aldrich. Roti-Quick kit for total RNA isolation was purchased from Carl Roth. SYBR green 1 kit was purchased from Eurogentec. Random hexamer primers were purchased from Boehringer Mannheim. Moloney murine leukemia virus-RT and recombinant RNase inhibitor were purchased from Promega. AmpliTaq polymerase, 10× PCR buffer containing 15 mM MgCl₂, and dNTPs were purchased from Applied Biosystems. DTT was purchased from Invitrogen Life Technologies.

RNA profiling of trafficking mononuclear phagocytes was performed in three different treatment groups: mice in group 1 were left untreated to monitor the constitutive mononuclear phagocyte trafficking into the lungs of mice in the absence of any previous manipulation. Mice in group 2 received intratracheal applications of CCL2 (50 μg) alone for 24 h to study changes in RNA expression profiles occurring in mononuclear phagocytes accumulating within the lungs of mice in response to the major monocyte chemoattractant, CCL2. Mice in group 3 received combined intratracheal applications of CCL2 (50 μg) in the presence of low doses of endotoxin (10 ng) for 24 h. The combined application of CCL2+LPS has recently been shown to induce an ARDS-like acute lung inflammatory response with monocytes acting as facilitators of the developing neutrophilic alveolitis (7, 11). All treatment protocols were done essentially as described recently (6, 7, 8, 9, 11). Briefly, CX₃CR1^+/GFP mice were anesthetized with tetrazoline hydrochloride and ketamine, and the trachea was exposed by surgical resection. Intratracheal instillation of CCL2 in the absence (group 2) or presence (group 3) of endotoxin (10 ng/mouse) was performed under stereomicroscopic control using a 29 gauge Abbocath, which was inserted into the trachea. After instillations, wounds were closed with sterile sutures. Mice were allowed to recover from anesthesia and subsequently returned to their cages with free access to food and water.

Twenty-four hours after intratracheal instillations, animals were sacrificed with an overdose of isoflurane (Forene; Abbott Laboratories). Blood and BALF were collected as described earlier (6, 7, 8, 9, 11).

A high-throughput FACSVantage SE flow cytometer (BD Biosciences) equipped with a DiVa sort option and an argon ion laser operating at 488 nm excitation wavelength and a laser output of 200 mW were used for the sorting of peripheral blood and BALF mononuclear phagocytes. Blood and BALF specimen were filtered through a 40-μm cell strainer (BD Biosciences) before cell sorting. Flow cytometric data of GFP-positive peripheral blood and alveolar mononuclear phagocytes from the various treatment groups were acquired on five-decade log-scale dot plots displaying forward scatter (FSC) area vs side scatter area and fluorescence 1 area vs fluorescence 2 area characteristics, respectively. First, hierarchy sort gates were set in FSC vs side scatter dot plots to exclude lymphocytes; second, hierarchy sort gates specific for GFP-expressing mononuclear phagocytes were set according to FSC area vs FL1 (F525 ± 15 nm; FITC/GFP) characteristics; and third, hierarchy sort gates were set according to FL1 vs FL2 characteristics (F575 ± 25 nm), thus allowing the exclusion of both alveolar macrophages and neutrophils.

Total cellular RNA was isolated from sorted and highly purified (>98%) GFP-positive mononuclear phagocytes from the various treatment groups using a commercially available RNA isolation kit (Carl Roth). RNA quantification and purity was determined on an Agilent Bioanalyzer 2100 (Agilent Biosystems). Only those RNA preparations exceeding absorbance ratios of (A_{260/280 nm}) > 1.90 were further processed for amplification and real-time RT-PCR validation experiments. Because numbers of constitutively alveolar recruited GFP-positive mononuclear phagocytes contained in BALFs of untreated CX₃CR1^+/GFP mice amounted to only ∼0.3–0.5% of total BALF cellular constituents (∼1500 cells/mouse), RNA samples from mononuclear phagocytes of ∼100 untreated group 1 CX₃CR1^+/GFP mice (10–15 mice of group 2 + 3 mice) were pooled together and subsequently used for RNA amplification before gene expression profiling using a nylon cDNA array (12). Experimental details regarding synthesis of complementary DNA and real-time RT-PCR validation were performed according to recently described experimental protocols (13, 14, 15).

The data are presented as mean ± SEM. Differences in outcome variables between treatment groups were analyzed by one-way ANOVA followed by Scheffe’s post hoc analysis. Values of p < 0.05 were considered to be significant.

Because of their homogeneously expressed, intrinsic GFP label, peripheral blood monocytes from CX₃CR1^+/GFP mice were easily detectable by their increased fluorescence 1 characteristics in flow cytometry, which is consistent with recently published reports (Ref. 10 ; Fig. 1,A). To confirm the specificity of the chosen sorting gates, the gated GFP-positive peripheral blood leukocyte population was sorted by high-speed flow cytometry combining both FSC area vs FITC area characteristics (Fig. 1,A) and fluorescence 1 (FITC, GFP) vs fluorescence 2 characteristics (Fig. 1,B). Subsequently, flow-sorted cells were reanalyzed by FACS to assess the purity of the sorted cell populations, which was found to be consistently >98% (Figs. 1,C and 2,C). Flow-sorted cells were also subjected to Pappenheim staining, which clearly identified these cells to consist almost exclusively of peripheral blood monocytes, as judged by their typical monocytic morphology (Fig. 1 D). In addition, the sorted cells clearly stained positive for unspecific esterase (data not shown).

FACS analysis of BALF cells collected from untreated CX₃CR1^+/GFP mice consistently revealed a small but clearly identifiable GFP-expressing cell population amounting to 0.3–0.5% of total BAL cells that was easily distinguishable from GFP-negative resident alveolar macrophages (Fig. 2, A and B). High-speed cell sorting and subsequent Pappenheim and unspecific esterase staining of these cells again revealed a monocytic cell population and, thus, to the best of our knowledge, identifies these rare cells as contributors to either alveolar macrophage and/or dendritic cell (DC) homeostasis in the absence of lung inflammation (Fig. 2 D).

The intratracheal application of recombinant murine CCL2 in the absence (or presence) of endotoxin did not affect GFP expression levels and proportions of circulating mononuclear phagocytes, as shown by flow cytometry (Fig. 1, A–C, middle and right columns) but induced a strong increase in alveolar accumulating GFP-positive leukocytes (Fig. 2, A–C, middle and right columns), which is in agreement with recently published reports (6, 8, 9). As expected, sorting of these CCL2-elicited, GFP-positive mononuclear phagocytes and subsequent Pappenheim-staining identified these cells as highly purified (>98%) alveolar recruited monocytic cells (Fig. 2 D, middle and right photographs). These data expand previous observations and clearly demonstrate the feasibility to track and purify GFP-positive mononuclear phagocytes from both intravascular and intra-alveolar compartments under baseline and acute inflammatory conditions for subsequent gene expression profiling in CX₃CR1^+/GFP mice.

In the current study, we compared changes in gene expression profiles of circulating vs alveolar recruited mononuclear phagocytes from untreated, CCL2 alone- and CCL2+LPS-treated CX₃CR1^+/GFP mice. Using nylon array technology covering a set of 1176 spotted genes, RNA expression levels assessed in alveolar recruited mononuclear phagocytes are presented as mean fold regulation relative to the corresponding RNA expression levels detected in circulating mononuclear phagocytes of the same treatment groups from three independent experiments. In addition, RNA level changes for selected genes of interest were validated by real-time RT-PCR.

Baseline alveolar mononuclear phagocyte recruitment observed in the lungs of untreated CX₃CR1^+/GFP mice was associated with only minor changes in gene expression profiles, and most of the differentially regulated genes (n = 13) were found to belong to the growth and transcription factor gene families (Table I; Fig. 3, A and B). This finding clearly suggests that the homeostatic process of constitutive mononuclear phagocyte extravasation into the alveolar compartment under noninflammatory conditions does not lead to a significant activation of the recruited cells. In contrast, when mice received a single intratracheal instillation of the monocyte chemoattractant, CCL2, which is the key factor eliciting inflammatory monocyte trafficking into the lung, we observed an overall ∼6-fold increase in the number of differentially regulated genes in alveolar vs circulating mononuclear phagocytes (n = 66) compared with constitutively migrated mononuclear phagocytes (Table I; Fig. 3,A). Interestingly, most of these genes belong to the cell adhesion molecule families, including CD63 and CD68, as well as the growth factor, cytokine, and chemokine families. Gene expression of GRO-1 encoding the neutrophil chemoattractant KC, monocyte chemoattractant CCL2, and platelet-derived growth factor-B were found to be up-regulated in CCL2-elicited alveolar mononuclear phagocytes compared with their circulating progenitors. CCL2-elicited alveolar mononuclear phagocytes also demonstrated strongly elevated gene expression levels of the matrix metalloproteinase MMP-10 and the lysosomal cathepsins K, L, and D (Table I). In contrast, the cathepsin L-specific inhibitor, cystatin F, was found to be strongly down-regulated in alveolar compared with circulating mononuclear phagocytes of the CCL2 treatment group, demonstrating that CCL2-elicited alveolar mononuclear phagocytes exhibit a differentially regulated spectrum of proteolytically active enzymes with established functions in extracellular matrix degradation, pathogen elimination, as well as Ag processing.

As anticipated, most drastic changes in gene expression profiles of alveolar vs circulating mononuclear phagocytes were observed in CX₃CR1^+/GFP mice cochallenged with CCL2 in the presence of low levels of endotoxin. This treatment regimen has been characterized recently to induce an ARDS-like acute lung inflammation (7, 16). A total of 70 genes was found to be differentially regulated within this treatment group, out of which 34 genes overlapped with CCL2 alone-treated mice (Table I; Fig. 3, A and B). Importantly, CCL2+LPS elicited alveolar mononuclear phagocytes again showed strongly up-regulated mRNA levels for neutrophil chemoattractants such as KC and inflammatory protein (IP)-10, whereas mRNA levels of chemokine receptors CX₃CR1, CXCR2, and CXCR3 were down-regulated in CCL2+LPS-elicited alveolar mononuclear phagocytes (Table I; Fig. 3, A and B). The lysosomal cathepsins K, L, and D were strongly up-regulated in CCL2+LPS-elicited mononuclear phagocytes, whereas the inhibitor cystatin F was not detectable by nylon array analysis (Table I).

To validate gene expression data by an independent method, a selected set of interesting genes was additionally evaluated by real-time RT-PCR without any further preamplification step. As shown in Table II, differential gene expression analysis by real-time RT-PCR of alveolar vs circulating mononuclear phagocytes largely matched the data obtained by nylon array analysis, albeit at a much higher sensitivity level when compared with the nylon array approach. This suggests that the array technique, although accurately detecting trends in altered gene expression profiles, may not be suitable to accurately quantify low levels of gene expression in unstimulated cells, such as constitutively alveolar recruited mononuclear phagocytes. In fact, real-time RT-PCR analysis confirmed that alveolar recruited compared with circulating mononuclear phagocytes of CCL2-treated CX₃CR1^+/GFP mice expressed significantly elevated mRNA levels of the neutrophil chemoattractants KC, MIP-2 (not included in the nylon array analysis), and IP-10. Particularly, the mRNA level of the major neutrophil chemoattractant, MIP-2, was found significantly up-regulated in alveolar vs circulating mononuclear phagocytes of CCL2+LPS-treated mice. In line with the findings of the nylon array analysis, real-time PCR confirmed a consistent decline in the mRNA levels of chemokine receptor CX₃CR1 in alveolar compared with circulating mononuclear phagocytes from untreated, CCL2 alone-, and CCL2+LPS-treated CX₃CR1^+/GFP mice. In contrast, drastically increased mRNA levels of the lysosomal cathepsins B, L, D, and K together with down-regulated cystatin F mRNA levels were observed only in alveolar mononuclear phagocytes of the CCL2 and CCL2+LPS treatment groups.

To gain further insight into changes in pattern recognition receptor (PRR) signaling molecules expression profiles of alveolar vs circulating mononuclear phagocytes of the various treatment groups, we additionally determined mRNA expression levels of TLR2, TLR4, CD14, MD-1, Ly78, and syndecan-4 in the corresponding mononuclear phagocyte populations. Interestingly, most drastic changes in PRR mRNA levels were noted for CD14, showing highest up-regulation in CCL2-recruited and significantly less increase in CCL2+LPS-elicited alveolar mononuclear phagocytes when compared with circulating mononuclear phagocytes, which is in agreement with recent data addressing CD14 cell surface protein expression on freshly alveolar recruited mononuclear phagocytes (8). Along the same line, we also found that PRR-signaling molecules other than CD14, such as TLR4 and syndecan-4, were significantly up-regulated in alveolar vs circulating mononuclear phagocytes recovered from CCL2- and CCL2+LPS-treated mice when compared with mononuclear phagocytes collected from untreated mice. Interestingly, TLR2 mRNA levels were found to be slightly down-regulated in alveolar vs peripheral blood mononuclear phagocytes of all three treatment groups, whereas expression levels of Ly78 remained largely unchanged, indicating differential regulation of various PRR molecules associated with the recruitment of mononuclear phagocytes.

In the current study, we exploited the endogenous GFP expression characteristics of transgenic CX₃CR1^+/GFP mice to identify and molecularly characterize mononuclear phagocytes before and after their spontaneous, as well as inflammatory, trafficking from the circulation into the alveolar compartment of the lung. Use of this novel mouse strain enabled us to identify cells with a monocytic phenotype within the alveolar compartment under baseline conditions, which may be involved in maintaining either alveolar macrophage and/or lung DC homeostasis. Combining DiVa-assisted FACS with nylon array gene expression analysis and real-time RT-PCR validation, we found that constitutively alveolar migrated mononuclear phagocytes showed a basal gene expression profile similar to circulating monocytes, including unchanged mRNA levels of major neutrophil chemoattractants, cell adhesion molecules, lysosomal cysteine proteases, and PRRs. The same approach revealed that mononuclear phagocytes recruited into the lungs of CX₃CR1^+/GFP mice in response to the major monocyte chemoattractant, CCL2 or CCL2, in the presence of low levels of endotoxin demonstrated drastic changes in their gene expression profiles compared with circulating mononuclear phagocytes, as reflected by increased mRNA levels of major neutrophil chemoattractants, LPS recognition molecules such as CD14 and TLR4, matrix metalloproteinases, and lysosomal cysteine proteases. The combination of flow cytometric and molecular approaches in the present study reveals that the leukocyte recruitment stimulus within the alveolar compartment strongly affects the gene expression profile of the recruited leukocyte subsets.

Various studies in the past few years have identified peripheral blood monocytes of mice not only to consist of different subpopulations, mainly depending on their Gr-1 expression profiles (Gr-1^high vs Gr-1^low), but also to exhibit the plasticity to differentiate into alveolar macrophages and/or pulmonary DCs upon their inflammatory recruitment to the lungs (6, 17, 18, 19). Because the rare cell population of GFP-positive mononuclear phagocytes detected in the lungs of untreated CX₃CR1^+/GFP mice showed a monocytic morphology similar to what has been described elsewhere to reflect monocytes/small macrophages (18) and stained positive for unspecific esterase (data not shown), it appears that these cells might contribute to alveolar macrophage homeostasis. In contrast, it cannot be excluded that these cells also contribute to the maintenance of pulmonary DCs, which are derived from the pool of circulating monocytes as well. Because of its rarity, this alveolar cell population was not extensively characterized with respect to its immunophenotype in the current study, and clarification of whether constitutively alveolar migrated monocytic cells primarily differentiate into macrophages or DCs will be subject to future investigations.

An important finding of the present study was that genes of the key CXC chemokines KC, CXCL2 (MIP-2), and CXCL10 (IP-10) were significantly up-regulated by a magnitude ranging from 10 to 33 in the CCL2+LPS treatment group, as compared with untreated animals. These data clearly support the concept that CCL2+LPS challenge provokes neutrophil chemotactic activities in alveolar-recruited mononuclear phagocytes, enabling them to act as facilitators in the development of neutrophilic alveolitis in acute lung injury and, at least in part, contribute to the strong neutrophil chemotactic activities observed in the BALFs of CCL2+LPS-treated mice (2, 6, 7, 16). This concept is further supported by the observation that S-100A9, a myeloid-related protein of the S100 protein family known to be expressed by monocytes/macrophages that exerts neutrophil chemotactic activities in inflamed tissues (20, 21, 22, 23, 24, 25), was also found to be significantly up-regulated in alveolar mononuclear phagocytes of CCL2 alone and CCL2+LPS-treated CX₃CR1^+/GFP mice.

Our study also addressed changes in gene expression levels of several PRR and associated signaling molecules, including TLR2 and TLR4, CD14, MD1, and syndecan 4. Most dramatic changes were seen in mRNA levels of TLR4, syndecan-4, and CD14. The strong up-regulation of CD14 (26, 27) by CCL2-elicited mononuclear phagocytes confirms earlier reports from our group and possibly reflects a priming of monocytes during the recruitment process with potential relevance for the pulmonary host defense (8). Syndecan-4, a transmembrane heparan sulfate proteoglycan of the syndecan family, was even more dramatically up-regulated in alveolar vs circulating mononuclear phagocytes of CCL2 and CCL2+LPS-treated but not untreated mice. Recently, Muramatsu and colleagues (28, 29) demonstrated that syndecan-4^−/− mice lack susceptibility to endotoxic shock. Thus, the strong up-regulation of syndecan-4 together with other important LPS-signaling molecules in alveolar mononuclear phagocytes supports the concept that freshly recruited mononuclear phagocytes may significantly contribute to the orchestration of acute lung inflammation. In this regard, the observed weak changes in mononuclear cell gene expression levels of the PRR MD-1 may well reflect the fact that only a 24-h time point posttreatment was analyzed.

Interestingly, we found significantly increased alveolar mRNA levels of lysosomal cathepsins together with decreased cystatin F mRNA levels in mononuclear phagocytes of the CCL2 and CCL2+LPS treatment groups compared with untreated mice. Cathepsins play important roles in phagolysosomal degradation, Ag processing, and tissue remodelling by professional phagocytes, including mononuclear phagocytes. These proteolytic enzymes have also been suggested to assist mononuclear cell transmigration (30, 31, 32, 33). Cystatin F belongs to the superfamily of endogenous intracellular thiol proteinase inhibitors (34, 35, 36). The reported findings of reciprocal expression patterns of cathepsins vs cystatins in alveolar recruited mononuclear phagocytes of the CCL2 and CCL2+LPS treatment groups support the potentially important contribution of these freshly immigrated cells in lung inflammatory remodeling processes. Interestingly, increased cathepsin B and K mRNA levels were also found to be elevated in mononuclear phagocytes in bleomycin-induced lung injury in rats and mice (37, 38). In addition, our study shows that the genes for cathepsins L, D, and K were among those few genes that were up-regulated even in alveolar mononuclear phagocytes compared with peripheral blood monocytes of untreated mice, indicating that constitutive trafficking of monocytic cells into the lungs to some extent alters the cathepsin gene expression pattern in these cells.

We and others recently demonstrated that F4/80-positive peripheral blood monocytes of mice consist of two principal subpopulations, Gr-1^high and Gr-1^low (6). Because these subsets were found to be equally recruited into the lungs of mice upon CCL2+LPS treatment, the currently used treatment regimen to recruit mononuclear phagocytes into the lungs of CX₃CR1^+/GFP mice did not allow us to specifically monitor subset-specific changes in gene expression profiles between Gr-1^high vs Gr-1^low monocyte subsets. However, a selective recruitment of Gr-1^high but not Gr-1^low CX₃CR1-expressing monocyte subsets was recently observed to occur in a model of thioglycolate-induced peritonitis (19). These data suggest that differences in experimental models and organs investigated and/or inflammatory stimuli applied may possibly affect the recruitment profiles of Gr-1^high vs Gr-1^low monocyte subsets into tissues, which may have implications on the overall inflammatory response.

Various steps in the recruitment process may potentially contribute to the activation of genetic programs in recruited mononuclear phagocytes, particularly 1) the CCL2-CCR2 interaction itself driving inflammatory monocyte recruitment, 2) additional molecular interactions during mononuclear phagocyte transendo/-epithelial migration, and 3) exposure of the recruited cells to an inflammatory altered alveolar microenvironment. In vitro exposure of isolated monocytes to CCL2 is known to activate genetic programs in monocytes but to a much more limited extent than that seen after CCL2-driven monocyte recruitment to the alveolar space in vivo (unpublished data). Therefore, we believe that additional activation signals that operate during the recruitment process are likely to play a key role in determining the phenotype of alveolar recruited mononuclear phagocytes, particularly in the CCL2 alone-treated group, which showed no significant changes in BALF cytokine profiles compared with untreated animals (16). The fact that several of the investigated genes, including cathepsins B and L, were found to be less fold up-regulated in alveolar mononuclear phagocytes of the CCL2+LPS as compared with the CCL2 alone-treated group might indicate a specific role of the alveolar micromilieu differentially modulating the expression kinetics of the investigated genes, including the cathepsins. Indeed, LPS activation of alveolar mononuclear phagocytes (particularly monocytes) may accelerate their transdifferentiation toward a “macrophage genotype” as is evident by the drastic up-regulation of cathepsin K transcript levels in the CCL2+LPS treatment group. LPS activation is also known to down-regulate expression of cell surface-bound receptor molecules, such as CCR2 (39). In fact, chemokine receptors such as CX₃CR1 were found to be most strongly and significantly down-regulated in alveolar mononuclear phagocytes collected from CCL2+LPS compared with CCL2 alone-treated mice. It is interesting to note that despite a strong down-regulation of the CX₃CR1 gene alveolar mononuclear phagocytes from all treatment groups displayed a strong and easily detectable GFP label because of the characteristically long half-life of the GFP protein (>24h), according to our previous reports (10).

In summary, the relative low proportion of circulating monocytes and the lack of sophisticated purification methods in mice are the main reasons why the proinflammatory molecular phenotype of this important leukocyte population is only poorly defined. In the current study, we demonstrate the feasibility to track, sort to high purity and transcriptionally profile peripheral blood and alveolar mononuclear phagocytes of CX₃CR1^+/GFP mice without prior in vivo manipulation. Based on this technique, this is the first study to identify and sort constitutively trafficking alveolar mononuclear phagocytes. Using gene array and real-time RT PCR we observed that the gene expression patterns of these cells differs drastically from that observed in CCL2-driven alveolar mononuclear phagocytes recruitment. These findings further support the concept that constitutive mononuclear phagocyte trafficking is tightly controlled to avoid inflammatory activation and may be linked to different molecular pathways enabling migration processes. In addition, our gene expression profile analysis shows that inflammatory-elicited mononuclear phagocytes activate genetic programs that render them potent contributors to the overall acute lung inflammatory response. Future studies using selective knockout mouse models and gene silencing techniques, as well as protocols to specifically purify monocyte subpopulations, will help to further refine the role of distinct gene products in subsets of mononuclear phagocytes and their relative contribution to the lung inflammatory response to microbial challenge.

We are grateful to the excellent technical support by Regina Maus, Petra Janssen, and Marlene Stein in preparing samples for flow cytometric sorting of mononuclear phagocytes and in lung histology experiments.

The authors have no financial conflict of interest.