Pulmonary host defense employs a combination of biochemical and biophysical activities to recognize, inactivate, and mediate clearance of environmental agents as well as modulate the overall response to such challenge. Dysregulation of the inflammatory arm of this response is associated with chronic lung diseases (CLD) including cystic fibrosis and chronic obstructive lung disease. Although mechanisms mediating immunoregulation are incompletely characterized, decrements in levels of the nonciliated secretory cell product Clara cell secretory protein (CCSP) in numerous CLD and identification of proinflammatory state in mice homozygous for a null allele of the CCSP gene (CCSP−/−) suggest a central role for the nonciliated secretory cell in this process. In an effort to determine the molecular basis for immunoregulatory defects associated with CCSP deficiency, we utilized difference gel electrophoresis in combination with matrix-assisted laser desorption ionization time-of-flight to compare the proteomes of wild-type and CCSP−/− mice. We demonstrate a shift in the isoelectric point of the immunomodulatory protein annexin A1 (ANXA1) to more acidic isoforms in CCSP−/− mice. Similar ANXA1 mRNA and protein abundance in wild-type and CCSP−/− tissue and identical localization of ANXA1 protein to alveolar macrophages and the ciliary bed of ciliated cells demonstrated that CCSP deficiency was associated exclusively with altered posttranslational modification of ANXA1. These results suggest that both long- and short-range paracrine signaling between nonciliated secretory cells and cells of the immune system and epithelium impact modification of cell type-specific proteins and implicate nonciliated secretory cells in a regulatory axis that might integrate critical aspects of host defense.
- Cara cell secretory protein
- chronic lung disease
efficacy of mucosal defense mechanisms is dependent on selectivity of the system for pathogenic agents and appropriate regulation of the magnitude and duration of the response. In the gut, identification of anti-inflammatory molecules derived from various tissue compartments as well as leukocytes has led to the concept that cross talk between epithelial, vascular, mesenchymal, and inflammatory system cells creates an “anti-inflammatory network” that is critical to balanced activation and termination of the host defense response (9, 12, 16). Dysregulation of feedback mechanisms that are an essential component of the anti-inflammatory network are postulated to contribute to hyperinflammatory states associated with inflammatory diseases. Thus disruption of regulatory cascades has the potential to transform a beneficial response into a pathological situation that contributes to initiation and exacerbation of chronic disease.
Continual exposure of the pulmonary mucosa to inhaled microorganisms, particles, and gases has necessitated development of host defense mechanisms that are highly responsive to infection or injury but are effectively attenuated following clearance of the pathogen, withdrawal of the irritant, or repair of the lesion. Multiple tissues and cell types contribute to the waxing and waning process that constitutes the innate response to environmental agents in the lung. Traditional roles for the airway epithelium in host defense include function as a dynamic barrier to inhaled agents and contribution to the biochemical and mechanical aspects of mucociliary clearance. In addition, epithelial irritation and injury results in rapid alterations in the spectrum of proteins secreted into the airway lumen. Epithelial-derived host defense proteins include antioxidant, bacteriostatic, and bacteriolytic proteins as well as chemokines and cytokines that recruit immune system cells and modulate the immune response (20). The vital role for appropriately regulated epithelial responses in mucosal defense is highlighted by the potentially life-threatening sequelae of chronic inflammatory diseases such as cystic fibrosis (CF) and chronic obstructive pulmonary disease (COPD) (19, 25). In such individuals, deviations in the normal inflammatory response have been correlated with a decline in the most abundant secreted product of nonciliated secretory cells, Clara cell secretory protein (CCSP/CC16), and increased abundance of annexin A1 (ANXA1)/lipocortin1 (2, 10, 27, 30, 38). Both of these proteins have been proposed to exhibit immunomodulatory properties, yet their role in modulation of airway inflammation has not been determined.
ANXA1 is a 37-kDa protein that is abundantly expressed in cells critical to the innate immune response, notably blood leukocytes, macrophages, and epithelial cells (1). Putative roles for ANXA1 in modulation of inflammation stem from the finding that ANXA1 antagonizes neutrophil extravasation (8). In lipopolysaccharide-exposed macrophages, the anti-inflammatory effects of ANXA1 may be mediated by ANXA1-dependent alterations in the activity of proinflammatory enzymes including inducible nitric acid synthetase, cPLA2, and potentially in cyclooxygenase-2 (33). In these cell types, interactions between ANXA1 and target proteins are regulated through receptor tyrosine kinase or PKC-mediated phosphorylation of specific serine, tyrosine, and threonine residues (9). Alternative posttranslational modifications to ANXA1 result in generation of isoforms with distinct isoelectric point (pI). Interestingly, lymphocytes from individuals with Fragile X syndrome express multiple pH isoforms of ANXA1 that are not phosphorylation intermediates and may be a consequence of hyperacetylation of the NH2-terminal domain (24, 37). Within the lung epithelium, ANXA1 is expressed specifically in ciliated cells of the conducting airway (26), colocalizes with the chloride transporter, CF transmembrane conductance regulator (CFTR), and is severely downregulated in the context of CF disease in humans and in CFTR knockout mice (2).
To define roles for CCSP in airway homeostasis, mice have been established that are homozygous for a null allele of the CCSP gene (CCSP−/−). Even though CCSP−/− mice have no overt lung phenotype in the steady state (35), challenge studies have revealed elevated epithelial cell susceptibility and exaggerated inflammatory responses to inhaled oxidant pollutants (21, 22), an elevated inflammatory response to challenge with either respiratory viruses or live Pseudomonas aeruginosa (17, 42), and enhanced epithelial remodeling in an ovalbumin model of allergic inflammation (43). Moreover, CCSP deficiency in CCSP−/− mice is associated with changes in nonciliated secretory cell secretory function and the steady-state abundance of specific mRNA species within the lung including those coding for secretory proteins (SCGB3A2) and immunoglobulin A (IgA) (11, 36, 45). These reports suggest that CCSP−/− mice phenocopy aspects of chronic lung diseases seen in human patients with COPD, asthma, and CF and support a role for CCSP and/or altered secretory cell function in the pathophysiology of these diseases.
The elevation of IgA and the greater inflammatory response of CCSP−/− mice to injury suggested that CCSP deficiency results in a proinflammatory state. The present study uses difference gel electrophoresis (DIGE) (39) to further explore the molecular basis for immunoregulatory defects associated with CCSP deficiency. Our initial comparison of the total lung proteomes of wild-type and knockout mice by DIGE revealed one consistent, significant difference. ANXA1 undergoes posttranslational modifications in knockout mice that shift the pI of ANXA1 from a single protein found at ca. pH 7.0 to multiple more acidic forms separated by ∼0.2–0.4 pH units. We demonstrate that CCSP-dependent modifications to ANXA1 are common to airway ciliated cells and macrophage populations from lung and the peritoneum. These findings suggest that CCSP functions as a biochemical link between functionally distinct airway epithelial cells and among cells of the innate immunity network. Disruption of this paracrine signaling network in the context of CCSP deficiency coupled with secondary effects on ANXA1 function may contribute to hyperinflammation in chronic lung disease.
Wild-type 129 SvEv (Taconic, Germantown, NY) and C57Bl/6 (Jackson Laboratories, Bar Harbor, ME) mice were maintained as inhouse breeding colonies within the University of Pittsburgh barrier facility. Two lines of mice harboring a null allele of the Scgb1A1 gene (formerly termed CCSP), 129P2/OlaHSD*129P3/J-Scgb1A1tmBrst (129 CCSP−/−, Mouse Genome Informatics no. 2684333), and B6.129OlaHSD*129P3/J-Scgb1A1tmbrst (B6 congenic CCSP−/−), were maintained in a homozygous condition through interbreeding in the same facility. All animals were maintained on a 12-h light/dark cycle and were allowed food and water ad libitum. Animals were monitored quarterly for known mouse pathogens and entero- and ectoparasites and were consistently disease-free. Female mice between 2 and 4 mo were used for all experiments. All procedures involving animal use were reviewed and approved by the University of Pittsburgh Institutional Animal Care and Use Committee.
Wild-type and CCSP−/− mice of either the 129 or B6 congenic background were anesthetized with 15 ml of 2.5% avertin/kg body wt and exsanguinated. Left lung lobes were homogenized with a Polytron homogenizer in 0.5 ml of denaturing lysis buffer (7 M urea, 2 M thiourea, 4% CHAPS, 10 mM DTT, 10 mM HEPES, pH 8.0) on ice. Particulate matter was removed by centrifugation (12,000 g, 15 min, 4°C). Supernatants were aliquoted, frozen in liquid nitrogen, and stored at −80°C.
Bronchoalveolar and peritoneal lavage.
Groups of five C57Bl/6 wild-type or B6 congenic CCSP−/− mice were used for analysis of alveolar and peritoneal macrophage proteins. Mice were anesthetized as detailed above, and lungs were lavaged via a tracheal cannula with 2 × 1 ml of phosphate-buffered saline. The peritoneum was then lavaged by injecting 5 ml of phosphate-buffered saline, massaging the abdominal cavity, and recovering the fluid. Either peritoneal or bronchoalveolar lavage from animals of like genotype was pooled, and cells were recovered by centrifugation at 300 g, for 5 min, at 4°C. Cells were washed once with saline, and the pellet was aspirated to dryness. Cells were lysed in 100 μl of denaturing lysis buffer as described above. Debris was removed by centrifugation as described above.
Two-dimensional gel electrophoresis.
Cleared wild-type and CCSP−/− homogenates were thawed on ice, and protein concentrations were determined with the 2-D Quant Kit (Amersham). Aliquots containing 100 μg of wild-type protein were labeled with Cy3, and similar aliquots of CCSP−/− protein were labeled with Cy5 essentially as described by Unlu and colleagues (39). Dyes were obtained from the University of Pittsburgh Proteomics Core, and labeling was reversed to account for differences in dye labeling efficiency. The differentially labeled wild-type and knockout protein samples were combined and loaded onto either 18- or 24-cm Immobiline pH 4–10 Isoelectric Focusing (IEF) strips (Amersham Biosciences) that were preswollen overnight in lysis buffer plus 2% ampholytes (IPG buffer, Amersham). First-dimension gels were focused to 70,000-V hours on an IPGphor power source (Bio-Rad). Second-dimension linear 5–12% SDS-PAGE gels were run at 25 mA per gel for 30 min and then at 50 mA per gel for ∼5 h.
Gel images were acquired on Typhoon scanner (Amersham) and analyzed using Image J (http://rsb.info.nih.gov/ij/). In Image J, the Cy3 and Cy5 images were stacked, and a two-frame movie was evaluated visually for changes in spot intensity. At least three independent comparisons were performed to identify repeatable differences. Gels were poststained with colloidal Coomassie blue (Bio-Rad Bio-Safe), and proteins of interest were excised from the gel as 1.5-mm diameter plugs with a manual spot picker (OneTouch, The Gel).
Gel pieces were dehydrated at room temperature with 50% methanol/25 mM NH4HCO3 (2 × 100 μl wash) followed by 100% acetonitrile (1 × 50 μl wash). Dehydrated pieces were dried in a SpeedVac for 15 min at room temperature. Tryptic digests were performed with Promega Trypsin Gold, 20 ng/μl in 25 mM NH4HCO3 (10 μl) for 4 h at 42°C. Protein digests were removed to a new tube, and peptides were eluted twice with 50% acetonitrile/1% trifluoroacetic acid (2 × 50 μl). Recovered peptides were evaporated to dryness in a SpeedVac at room temperature. Dried peptides were dissolved in 50% acetonitrile/0.3% trifluoroacetic acid/1 mM citric acid, mixed with an equal volume of saturated α-cyano-4-hydroxy cinnamic acid. Multiple 0.75-μl aliquots were spotted on a 192-well matrix-assisted laser desorption ionization (MALDI) plate (Applied Biosystems).
Mass spectrometer analysis.
MALDI-time-of-flight (TOF) and MALDI-TOF-TOF analyses were performed on an ABI 4700 mass spectrometer at the University of Pittsburgh Integrated Core Facility. MALDI-TOF analysis was performed in the reflectron mode. An analysis of peak lists from MALDI-TOF and MALDI-TOF-TOF was performed with an inhouse version of Mascot (Matrix Science).
Western blot analysis.
Lung homogenate (20 μg) was resolved by reducing SDS-PAGE and was electrophoretically transferred to PVDF membrane. The membrane was blocked with 5% powdered milk in phosphate-buffered saline (blocking buffer) and probed using a rabbit polyclonal antibody against ANXA1 (1:4,000 in blocking buffer, Zymed Laboratories). Immunoreactive bands were detected with a goat anti-rabbit HRP conjugate (1:4,000 in blocking buffer, Invitrogen), and peroxidase reaction was developed using enhanced chemiluminescence (Amersham Pharmacia Biotech, Piscataway, NJ). Bands were detected using BioMax film (Kodak, Rochester, NY).
Quantitative PCR analysis.
Left lung lobes from three wild-type or three B6 congenic CCSP−/− mice were homogenized individually, and total cellular RNA was purified by the method of Chomczynski and Sacchi (6). RNA was quantified spectrophotometrically, and 400 ng was reverse transcribed using a First Strand Synthesis Kit (Invitrogen). Aliquots of wild-type, knockout, and reference cDNA were amplified using TaqMan 1000 RXN Gold with Buffer A Pack (Applied Biosystems) and Assays-On-Demand for ANXA1 and β-glucuronidase (Applied Biosystems). Cross threshold (CT) values were determined, and the ΔΔCT was calculated (18). The means ± SE are reported. Differences were compared by the Student's t-test. Differences were accepted at P = 0.05.
Dual immunofluorescence and confocal microscopy.
Lung tissue from wild-type or CCSP−/− mice was fixed by instillation of 10% neutral buffered formalin (NFB) at 10 cmH2O pressure through a tracheal cannula followed by immersion in 10% NFB for 24 h. The right lower lobe was embedded in paraffin, and 5-μm sections were taken through the major airway with the inclusion of minor daughter branches using a rotary microtome (Fisher Scientific, Pittsburgh, PA). Immunostaining was performed using a combination of airway-specific markers, CCSP (goat-anti-CCSP, 1:8,000, generated in this laboratory), and acetylated tubulin (mouse IgG2b-anti-AcT, 1:8,000, Sigma) in combination with ANXA1 (rabbit-anti-ANXA1, 1:200, Zymed Laboratories). Sections were cleared in xylenes and blocked for 1 h with 5% bovine serum albumin and fraction V protease-free in phosphate-buffered saline (5% BSA/PBS) and then incubated overnight at 4°C with primary antibody diluted in 5% BSA/PBS. Sections were washed with phosphate-buffered saline and stained with the appropriate secondary antibodies: donkey anti-goat Alexa 488 or donkey anti-goat Alexa 594 (1:500; Molecular Probes, Eugene, OR), donkey anti-rabbit Alexa 488 or donkey anti-rabbit Alexa 594 (1:500; Molecular Probes), and goat anti-mouse IgG2b Alexa 594 (1:500; Molecular Probes). Sections were mounted using Flouramount G (Southern Biotech, Birmingham, AL) containing 1 g/ml 4,6-diamidino-2-phenylindole dihydrochloride (Sigma). Immunoreactivity was not detected in the absence of primary antibodies. Images were captured using an Olympus Fluoview BX61 confocal microscope and assembled in Adobe Photoshop.
Epithelial Cell Cultures
Air-liquid interface cultures.
Groups of four to six C57BL/6 wild-type and B6 congenic CCSP−/− mice were anesthetized as detailed above, and the trachea was recovered. Tracheal tissue was digested with 0.15% pronase (Roche) in Ham's/F-12 (GIBCO) overnight at 4°C. Cells were recovered and cultured in mouse tracheal epithelial culture medium as detailed by You and colleagues (47) until individual cultures reached a resistance of ∼1,000 mΩ (5–6 days). At this point, the apical culture medium was removed, and the basal culture medium was replaced with 2% Nuserum. Cultures were fed on alternate days and allowed to differentiate for 9 days. Ciliogenesis was followed visually using a phase-contrast microscope and was used as an estimate of differentiation.
Whole mount immunofluorescence analysis.
Groups of three C57Bl/6 wild-type or B6 congenic CCSP−/− air-liquid interface cultures from two independent experiments were analyzed. The apical surface was washed with 200 μl of 1× phosphate-buffered saline, and the apical and basal compartments were filled with 10% neutral buffered saline. Cultures were fixed at 4°C for 30 min, washed with 1× phosphate-buffered saline, and permeabilized with PBS/BSA containing 0.1% Triton X-100 at room temperature for 30 min. Cellular antigens were detected by dual immunofluorescence and visualized by confocal microscopy as described above.
DIGE analysis of epithelial cultures.
Sets of six wild-type or B6 congenic CCSP−/− epithelial cultures were pooled for protein analysis. The apical surface of each culture was washed with 200 μl of phosphate-buffered saline, and the culture was placed on ice. Denaturing lysis buffer (100 μl) was added to the apical compartment, and the cells were disrupted with the broad end of a pipette tip. Lysates from cultures of like genotype were pooled, and debris was removed by centrifugation at 12,000 g for 15 min at 4°C. Supernatants were frozen in liquid nitrogen and stored at −80°C. Protein labeling and gel electrophoresis were preformed as described above.
Altered Posttranslational Modification of ANXA1 in Lung Tissue of CCSP−/− Mice
We previously demonstrated that nonciliated secretory cells of CCSP-deficient mice exhibit significant perturbations to their secretory apparatus (11). In the present study, we employed two-dimensional differential gel electrophoresis to more precisely define changes in protein abundance and distribution that accompany CCSP deficiency. Total protein recovered from lung tissue of wild-type C57Bl/6 mice and B6 congenic CCSP−/− (CCSP−/−) mice were conjugated to Cy3 and Cy5 dyes and protein profiles visualized by two-dimensional DIGE (2D-DIGE) using a linear 4–10 pH gradient in the first dimension followed by separation using linear 5–12% SDS-PAGE. Pairwise comparisons were performed using three independently generated lung protein samples from individual wild-type and CCSP−/− mice. Figure 1 includes a representative image of profiles generated comparing lung proteins from wild-type and CCSP−/− mice. Approximately 500 distinct protein species were resolved per gel. Differences in spot intensity that were consistently observed among gels comparing wild-type and CCSP−/− lung proteins were confirmed by dye substitution (data not shown). Protein identities of both validated and reference spots were determined by in-gel trypsin digestion followed by MALDI-TOF analysis. One series of protein species showing variability between lung protein samples from wild-type and CCSP−/− mice was found to be ANXA1 (Fig. 2, A–C, and Supplemental Data, which is available online at the AJP-Lung web site). The only ANXA1 species present within lung protein from wild-type mice had a pI of ∼7.0 and predicted molecular weight of 37 kDa (Fig. 1, A and C). In contrast, lung protein from CCSP−/− mice lacked the pI 7.0 isoform but harbored more acidic ANXA1 isoforms with pI values of 6.8 and 6.4 (Fig. 1, B and D). Differences between wild-type and CCSP−/− lung ANXA1 isoforms were retained following treatment of total lung homogenates with alkaline phosphatase (data not shown). Further analysis of ANXA1 protein abundance through Western blot analysis of total lung protein (Fig. 3A) and analysis of ANXA1 mRNA abundance through real-time quantitative RT-PCR amplification of total lung RNA (Fig. 3B) revealed no significant differences in ANXA1 gene expression between wild-type and B6 congenic CCSP−/− mice.
To independently verify the identity of protein species exhibiting differences between wild-type and CCSP−/− mice, unlabeled protein samples from homogenized lung tissue of C57Bl/6 WT and B6 congenic CCSP−/− were resolved by 2D gel separation and assessed by Western blot analysis. Consistent with data from MALDI-TOF analysis was the demonstration that blots of total lung protein from CCSP−/− mice included ANXA1-immunoreactive protein species with more acidic pI (Fig. 4B), whereas lung protein from wild-type mice contained the more basic isoform of ANXA1 only (Fig. 4A). To verify that differences in ANXA1 were dependent on CCSP and not due to strain-specific difference in ANXA1 posttranslational modification, unlabeled total lung protein from wild-type and CCSP−/− mice of the 129OlaHsd strain were compared by 2D gel electrophoresis followed by Western blot and immunodetection of ANXA1 isoforms. Even though strain-specific differences were observed in the abundance of ANXA1 within lung tissue samples, the same pattern of CCSP-dependent posttranslational modification of ANXA1 was observed (Fig. 4, C and D).
ANXA1 Localizes to the Apical Domain of Ciliated Cells Within Conducting Airways
Since differential posttranslational modification of ANXA1 has been linked with changes in its subcellular localization (34), we sought to determine the cellular and subcellular distribution of ANXA1 expression within airways of wild-type and CCSP−/− mice. ANXA1 was localized within airways by immunofluorescence microscopy, with colocalization for cell type-specific markers including CCSP to define secretory cells in airways of wild-type mice and acetylated tubulin (AcT) to define ciliated cells within airways of wild-type and CCSP−/− mice. In wild-type airways, ANXA1 immunoreactivity specifically localized to the apical membrane of a subset of airway epithelial cells that were positive for AcT immunoreactivity (Fig. 5, D–F) and was absent in CCSP-immunoreactive cells (Fig. 5, A–C). The cellular and subcellular distribution of ANXA1 in airways of CCSP−/− mice was identical to that observed for wild-type: ANXA1 immunoreactivity colocalized with that of AcT immunoreactivity within the ciliary bed of ciliated cells (Fig. 5, G–K). These data suggest that CCSP deficiency and associated changes to nonciliated secretory cells result in changes in the posttranslational modification but not the cellular pattern or subcellular distribution of ANXA1 within ciliated cells.
CCSP-Dependent Posttranslational Modification of Macrophage-Associated ANXA1
Secretory and ciliated cells share close proximity within airways that includes cell-cell contacts and the potential for junctional communication through tight junctions. However, resident myeloid cells and those recruited through local proinflammatory conditions are also known to harbor significant levels of ANXA1 that contribute to the total lung pool. We sought to determine whether myeloid cells maintained the same difference in ANXA1 posttranslational modification as was apparent within total lung protein. Two-dimensional DIGE was performed on total protein prepared from cells recovered in bronchoalveolar lavage from wild-type and CCSP−/− mice. Macrophages represented the principal cell type recovered within bronchoalveolar lavage from both wild-type and CCSP−/− mice, accounting for ∼98% of total lavagable cells in each case. Protein species separated from lavagable cell pellets included high levels of ANXA1 that were clearly detected by 2D-DIGE with subsequent confirmation of protein identity by Western blot analysis (Fig. 6). Lung macrophage-associated ANXA1 from wild-type mice was represented as a single basic isoform with a pI of 7.0 (Fig. 6A). In contrast, lung macrophage-associated ANXA1 from CCSP−/− mice was represented by primarily by the more acidic isoforms with pI of 6.8 and 6.4 (Fig. 6B). This pattern of ANXA1 pI isoforms was similar to that reported for lymphocyte-derived ANXA1 from Fragile X syndrome patients (37). The demonstration that macrophage-associated ANXA1 exhibits the same CCSP-influenced posttranslational modification as that observed within total lung protein implies that these influences are mediated through a paracrine mechanism.
Even though the lung is the principal site of CCSP expression and protein accumulation, CCSP is present at measurable levels within serum and is present within the glomerular filtrate before uptake and catabolism within the epithelium lining the proximal convoluted tubule (5, 14). We next sought to determine whether posttranslational modifications to ANXA1 were found on macrophages located in nonpulmonary sites. Peritoneal macrophages were recovered from peritoneal lavage fluid of wild-type and CCSP−/− mice, and their proteomes were evaluated by 2D-DIGE followed by MALDI-TOF identification of ANXA1 isoforms. As was the case for lung macrophages, peritoneal cells from wild-type mice exhibited a single ANXA1 isoform with a pI of ∼7.0 (data not shown). However, peritoneal cells from CCSP−/− mice displayed the same acidic isoforms of ANXA1 as had been demonstrated within lung tissue and isolated macrophages (data not shown). These data demonstrate that the paracrine influence of CCSP deficiency extends beyond the lung.
Posttranslational Modification of ANXA1 is an Intrinsic Property of the CCSP-Deficient Epithelium
Since ANXA1 isoforms detected in total lung homogenates (Figs. 1, 2, and 4) could reflect ANXA1 derived from ciliated epithelial cells (Fig. 5) or macrophages (Fig. 6), we assessed ANXA1 expression in differentiated airway epithelial cell cultures. Cultures derived from wild-type mice were harvested at the point of maximal ciliogenesis, the time point at which nonciliated secretory cell abundance was ∼10–15% of nucleated cells (47) (data not shown). ANXA1-immunoreactive protein was present at the apical surface of ciliated cells and colocalized with AcT immunoreactivity (Fig. 7, A–C). ANXA1 was not detected in other cell types (data not shown). When comparisons were made between cultures derived from wild-type and CCSP−/− mice, no differences were observed in the abundance of cilia or in the cellular or subcellular distribution of ANXA1 immunoreactivity (data not shown). This analysis indicated that expression of ANXA1 on ciliated cells was an intrinsic property of the airway epithelium and not a secondary consequence of interactions between the epithelium and inflammatory cells.
Two-dimensional DIGE was performed to determine the ANXA1 isoforms expressed on ciliated cells derived from wild-type and CCSP−/− mice. As was the case in total lung protein from wild-type mice, the predominant ANXA1 isoform in protein recovered from wild-type acute lung injury (ALI) cultures was the basic isoform (Fig. 8A). However, small amounts of the acidic isoforms were also detected. Incomplete modification of ANXA1 in wild-type cultures may be related to underrepresentation of CCSP-expressing cells in tracheal epithelial cell cultures (47). Total protein from ALI cultures of CCSP−/− epithelium lacked the basic isoform of ANXA1, yet possessed acidic ANXA1 isoforms previously identified within total lung tissue homogenate from CCSP−/− mice. These data support the conclusion that posttranslational modification of ANXA1 in ciliated cells was regulated in a CCSP-dependent fashion. The observation that the ANXA1 modification on ciliated cells was similar to that observed for macrophage cells suggests that alterations in ANXA1 posttranslational modification in CCSP−/− mice most likely resulted from short-range interactions between modified secretory cells and adjacent ciliated cells and long-range effects on distant inflammatory cell types.
We demonstrate that CCSP deficiency in mice results in changes to the charge characteristics of ANXA1 in ciliated cells and either pulmonary or peritoneal macrophages. Localization of ANXA1 to ciliated cells of the conducting airway epithelium and to resident macrophages suggests that interactions between these cell types and airway secretory cells are impacted directly by CCSP deficiency. This finding provides, until now, unappreciated links between secretory cell function and the molecular regulation of other lung cell types that fulfill roles in lung homeostasis and disease. Moreover, the finding of CCSP-dependent ANXA1 modifications in peritoneal macrophages suggests that airway epithelial remodeling has the potential to impact immunoregulation in other mucosal sites.
ANXA1 has been shown to undergo a number of distinct posttranslational modifications that have potential to alter its subcellular distribution and/or function. Phosphorylation of a tyrosine residue at position 21, NH2-terminal serine residues, and a COOH-terminal threonine residue, have been shown to be mediated through various receptor tyrosine kinases, PKC, or PKA, respectively (31, 41, 44). Modifications to the NH2-terminal domain influence calcium binding of ANXA1 and subsequently its ability to interact with plasma membranes and plasma membrane-associated EGF receptor (29, 44). However, changes in the subcellular distribution of ANXA1 were not observed between ciliated cells of wild-type and CCSP−/− mice. Moreover, our observation of changes in pI independently of changes in apparent molecular weight, and the resistance of these changes to phosphatase treatment, suggests that posttranslational modifications other than phosphorylation account for the observed differences in ANXA1 isoforms represented within lungs of wild-type and CCSP−/− mice. Other posttranslational modifications could include acetylation, sulfation, or glycosylation. NH2-terminal acetylation of ANXA1 has been suggested to regulate its interaction with cytoskeletal components such as the S100 calcium-binding family of proteins (15). Differences in N-linked glycosylation have also been shown to alter the charge characteristics of ANXA1, resulting in the generation of distinct isoforms that can be distinguished according to their pI (4). The potential for posttranslational modifications of ANXA1 to influence its ability to bind calcium, lipids, or other proteins, coupled with its localization at the base of cilia within ciliated cells, raises the possibility that altered Clara cell secretory function that results from CCSP deficiency may lead to altered ciliated cell function as a result of changes in the ligand-binding properties of ANXA1 (32). Further studies are needed to determine whether CCSP deficiency impacts ciliated and macrophage cell activity and whether this is mediated through changes in the posttranslational modification of ANXA1.
CCSP-dependent changes in the pI isoforms of ANXA1 present within airway epithelial cells and macrophages derived from either lung or peritoneal compartments suggest paracrine influences of CCSP deficiency on these cell types. The obvious candidate mediator of such a paracrine effect is CCSP itself. Like ANXA1, CCSP has been shown to bind both calcium and lipophilic compounds such as lipids, progestins, and certain metabolites of polychlorinated biphenyls. As such, CCSP deficiency may alter the bioavailability of these ligands, thus influencing the properties and functions of ANXA1. CCSP has also been shown to interact directly with cubulin, resulting in megalin-dependent internalization within epithelial cells of the proximal convoluted tubule (5). The cubulin/megalin-dependent catabolism of CCSP highlights the systemic distribution of CCSP that may be responsible for the observed posttranslational modifications to ANXA1 in peritoneal macrophages.
The demonstration that CCSP deficiency in CCSP−/− mice is also associated with perturbations in Clara cell secretory function and altered protein composition of airway lining fluid raises the possibility that other soluble factors may mediate these effects on ANXA1 (11, 45). Other candidate secretory products of Clara cells that may mediate these effects include SCGB3A2, another member of the secretoglobin family of proteins whose expression is elevated in lungs of CCSP−/− mice (45). SCGB3A2 has been shown to interact with the class A scavenger receptor, MARCO (3). Interestingly, interactions between scavenger receptors and other ligands, such as oxidized lipoproteins, have been shown to increase intracellular Ca2+ and activate a variety of cytosolic kinases and phospholipases via stimulation of pertussis toxin-sensitive G proteins (7, 28, 40, 46). Moreover, MARCO has been shown to mediate cytoskeletal rearrangements within dendritic cells, and its expression on macrophages is critical for appropriate trafficking of B cells to the marginal zone of the spleen (13, 23). These reports suggest that extracellular pools of secretoglobins such as SCGB3A2 and potentially CCSP can be sensed through interactions with scavenger receptors and that transduction of these signals impacts the functional properties of monocytic cells. Interestingly, we previously demonstrated that CCSP deficiency is associated with altered immunoglobulin production within lung tissue and increases in the abundance of IgA-expressing B cells in the peribronchial region of conducting airways (45). It is unknown whether SCGB3A2-MARCO interactions results in similar activation of signaling within lung macrophages or whether CCSP may exhibit a similar capacity for interaction with scavenger receptors. However, these reports raise the possibility that changes in the expression of secretoglobins, either CCSP or other members of the secretoglobin family such as SCGB3A2, may result in altered posttranslational modifications of ANXA1 within macrophages through a scavenger receptor-dependent pathway. The demonstration that this mode of epithelial-inflammatory cell cross talk is altered within lungs of CCSP−/− mice suggests that similar immunoregulatory changes may result from CCSP deficiency seen with chronic lung disease.
Roles for altered Clara cell secretory function in the setting of chronic lung disease have been suggested from studies demonstrating that the abundance of CCSP is typically reduced in either airway lining fluid or in the systemic circulation of patients with chronic lung disease. The basis for these changes to airways, be it chronic injury and phenotypic changes to Clara cells associated with their role as the principal bronchiolar progenitor cell, and/or altered differentiation of the epithelium leading to a phenotypic change within the secretory cell population, has not been well established. However, these changes to airways are associated with defects in mucociliary clearance and local inflammatory responses. The finding in the present study that CCSP deficiency leads to biochemical changes within ciliated cells of the airway epithelium both in vivo and in vitro provides a link between these disparate yet functionally coordinated processes. These data combined with the finding that decrements in CCSP protein levels are also associated with decreased abundance, degradation, and altered subcellular distribution of ANXA1 in ciliated airway cells of CF patients (2, 38) suggest that CCSP and ANXA1 may regulate critical aspects of mucociliary clearance. The demonstration that changes in ANXA1 modification are not restricted to ciliated cells of the surface epithelium but are also seen within lavagable macrophages of the steady-state lung and peritoneum provides further insight into mechanisms of cross talk between secretory cells and other lung cells. These data, when coupled with our earlier demonstration that CCSP deficiency sensitizes mice to injury and more profound inflammatory responses to inhaled oxidant gasses, suggest that CCSP deficiency leads to a proinflammatory state for which modifications to ANXA1 are a reflection. Further studies are needed to define CCSP-dependent components of the lung inflammatory response and whether changes to ANXA1 are causative.
This study was supported by National Institutes of Health Grants HL-64888, ES-08964, and HL-70575.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society