|Home | About | Journals | Submit | Contact Us | Français|
The Forkhead Box f1 (Foxf1) transcriptional factor (previously known as HFH-8 or Freac-1) is expressed in endothelial and smooth muscle cells in the embryonic and adult lung. To assess effects of Foxf1 during lung injury, we used CCl4 and butylated hydroxytoluene (BHT) injury models. Foxf1+/− mice developed severe airway obstruction and bronchial edema, associated with increased numbers of pulmonary mast cells and increased mast cell degranulation after injury. Pulmonary inflammation in Foxf1+/− mice was associated with diminished expression of Foxf1, increased mast cell tryptase, and increased expression of CXCL12, the latter being essential for mast cell migration and chemotaxis. After ovalbumin (OVA) sensitization and OVA challenge, increased lung inflammation, airway hyperresponsiveness to methacholine, and elevated expression of CXCL12 were observed in Foxf1+/− mice. During lung development, Foxf1+/− embryos displayed a marked increase in pulmonary mast cells before birth, and this was associated with increased CXCL12 levels in the lung. Expression of a doxycycline-inducible Foxf1 dominant-negative transgene in primary cultures of lung endothelial cells increased CXCL12 expression in vitro. Foxf1 haploinsufficiency caused pulmonary mastocytosis and enhanced pulmonary inflammation after chemically induced or allergen-mediated lung injury, indicating an important role for Foxf1 in the pathogenesis of pulmonary inflammatory responses.
Foxf1 haploinsufficiency caused pulmonary mastocytosis and enhanced pulmonary inflammation after chemically induced or allergen-mediated lung injury, indicating an important role for Foxf1 in the pathogenesis of pulmonary inflammatory responses.
Mast cells are highly specialized cells that play important roles in adaptive and innate immunity, immunoglobulin E–mediated allergy, and inflammatory responses in the lung (1, 2). Immediately after exposure to allergens, mast cells release a variety of inflammatory mediators, including histamine, heparin, prostaglandins, and mast cell proteases, causing airway constriction, blood vessel dilatation and permeation, tissue swelling, hyperoxia, and eosinophil infiltration, contributing to pathogenesis of bronchial asthma (1, 2). Mast cells store tumor necrosis factor (TNF) and are thus primed to trigger TNF-mediated inflammatory responses (3). Mast cells also produce IL-4, IL-6, and IL-8, which play important roles in acute and chronic lung inflammation (2, 4). Mast cell recruitment is mediated by various factors, including platelet-derived growth factor (PDGF), vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), transforming growth factor β (TGF-β), and stromal cell–derived factor-1 (SDF-1 or CXCL12) (4–6). CXCL12 is produced by fibroblasts and endothelial cells and binds to its receptor, CXCR4, present on the surface of mast cells. CXCL12 stimulates mast cell migration and IL-8 production (4, 7). Interestingly, activated mast cells are found in excessive numbers in tissues in heterogeneous clinical disorders associated with mastocytosis (8). Although a subset of patients with mastocytosis displayed activating mutations in c-KIT gene, little is known about transcriptional and signaling pathways contributing to molecular pathogenesis of this disease (8).
The Forkhead Box (Fox) family of transcription factors shares homology in the winged helix DNA-binding domain (9) and its members play important roles in cellular proliferation, differentiation, and metabolic homeostasis (10, 11). Forkhead Box f1 (Foxf1) is expressed in capillary endothelial cells, fibroblasts, and peribronchial smooth muscle cells of the embryonic and adult lung and other organs (12, 13). We and others have previously generated mice with targeted disruption of the Foxf1 gene and demonstrated that Foxf1−/− embryos die by 8.5 days post coitum (dpc) due to defective vasculogenesis in the yolk sac and allantois (13, 14). Haploinsufficiency of the Foxf1 gene in Foxf1+/− mice causes perinatal pulmonary hemorrhage, and severe defects in alveolarization and vascularization, as well as fusion of lung lobes and arteries (13, 15, 16). Perinatal lethality from pulmonary hemorrhage was observed in half of newborn Foxf1+/− mice that displayed the most severe reduction in pulmonary Foxf1 levels (13). Interestingly, the other half of the newborn Foxf1+/− mice displayed compensatory increases of Foxf1 mRNA levels in the lung, exhibited diminished alveolar septation without pulmonary hemorrhage, and survived postnatally (13). Adult Foxf1+/− mice had normal life spans and exhibited normal lung morphology in adulthood, suggesting that these mice compensated for the alveolar septation defect (17). However, in response to butylated hydroxytoluene (BHT)-mediated lung injury, the Foxf1+/− mice exhibited severe defects in lung repair and died from pulmonary hemorrhage (17).
Our previous studies demonstrated that Foxf1 is expressed in hepatic stellate and endothelial cells (18, 19). After carbon tetrachloride (CCl4) injury, Foxf1+/− livers exhibited diminished activation of the hepatic stellate cells, associated with a significant reduction in α-smooth muscle actin, type I collagen, notch-2, and IP-10 proteins, indicating an important role for Foxf1 in liver repair and function (18).
In this article, we have demonstrated increased numbers of pulmonary mast cells and susceptibility of Foxf1+/− mice to pulmonary injury and inflammation. Increased numbers of pulmonary mast cells, increased mast cell degranulation, and bronchial edema caused airway obstruction in CCl4-treated Foxf1+/− mice. Increased pulmonary inflammation in Foxf1+/− mice was associated with increased pulmonary expression of mast cell tryptase and CXCL12, the latter of which is essential for mast cell migration and chemotaxis. After ovalbumin (OVA) sensitization and OVA challenge, Foxf1+/− mice displayed increased lung inflammation, airway hyperresponsiveness to methacholine, and elevated expression of tryptase and CXCL12. Furthermore, increased numbers of mast cells and elevated CXCL12 levels were found in lungs of Foxf1+/− embryos, suggesting that Foxf1 haploinsufficiency causes a genetic predisposition to inflammatory lung diseases. Finally, using primary endothelial cells containing a doxycycline-inducible Foxf1 dominant-negative transgene, we demonstrated that Foxf1 deficiency increased CXCL12 expression in vitro. Foxf1 haploinsufficiency caused abnormal accumulation of mast cells during lung development and lung injury, contributing to genetic predisposition to chemically induced and antigen-mediated lung inflammation.
Foxf1+/− mice in which the Foxf1 allele was disrupted by an in-frame insertion of a nuclear localizing β-galactosidase (β-Gal) gene were bred for 10 generations into the Black Swiss mouse genetic background (13). Carbon tetrachloride (CCl4; Sigma, St. Louis, MO) was dissolved in mineral oil at a 1:20 ratio vol/vol, and a single intraperitoneal injection of CCl4 (0.5 μl of CCl4/1 g of body weight) was administered to male Foxf1+/− mice or their wild-type (WT) littermates as described (18, 20). In separate experiments, Foxf1+/− and WT mice were injected intraperitoneally with BHT (300 mg/kg body weight) to induce lung injury as described (12, 17, 21). To determine statistical significance of any observed differences, we used five Foxf1+/− and WT mice in each group. Mice were killed by CO2 asphyxiation, and the left lung lobe was used to prepare total RNA. Right lung lobes were fixed overnight in 4% paraformaldehyde (PFA) at 4°C and then paraffin embedded.
Foxf1+/− and WT mice were immunized with 10 μg of OVA mixed with 1.125 mg of aluminum hydroxide (Imgect Alum; Pierce Chemical Co., Rockford, IL) on Days 0, 7, and 14 as described previously (22). On Days 21 to 24, mice were exposed to aerosolized OVA (1%) or saline for 40 minutes (22). Twenty-four hours after the last OVA challenge, mice were killed, and the mouse lungs were used for paraffin embedding or preparation of total lung RNA.
Bronchoalveolar lavage (BAL) fluid was collected either 24 hours after the final dose of OVA or 18 hours after CCl4 treatment. BAL was performed by delivering 0.8 ml cold PBS into the airway through a trachea cannula and gently aspirating the fluid. The lavage was repeated three times to recover a total volume of 2 to 3 ml. The cells were stained with Trypan blue to determine viability and with Turk solution to obtain total nucleated cell counts using a hemocytometer. Cytospin (Cytospin 2; Shandon, Waltham, MA) slides were prepared from the BAL fluid and were then fixed and stained using Diff-Quick reagent (Dade Diagnostics, Deerfield, IL). Differential cell counts were determined by counting 300 cells per each slide using standard morphologic criteria in a single-blind method. Five mice were used for each group.
The concentrations of IL-4, IL-5, and IFN-γ in BAL fluid were measured using a Mouse Cytokine CBA kit (BD Biosciences, San Diego, CA) according to the manufacturer's recommendations. The concentrations of CXCL12 in BAL fluid and serum was determined by a quantitative immunoassay from R&D Systems (Minneapolis, MN).
Methacholine challenge was performed either 24 hours after the final dose of OVA or 18 hours after CCl4-treatment. Pulmonary function testing was performed using a computer-controlled small-animal ventilator (SAV) (Flexivent; Scireq, Montreal, PQ, Canada) as previously described (22). Briefly, mice were anesthetized with 30 mg/kg xylazine and 80 mg/kg ketamine intraperitoneally, and the trachea was cannulated with an 18-gauge metal needle connected to the SAV. Regular mechanical ventilation was applied, and animals were ventilated quasi-sinusoidally at a frequency of 120 breaths/minute at a tidal volume of 6 ml/kg. The expiratory valve of the SAV allowed the animal to empty passively through a water trap adjusted to maintain a positive end-expiratory pressure of 2.0 cm H2O. This positive end-expiratory pressure was shown to be optimal for the determination of methacholine-induced changes in respiratory system resistance (22). Increasing doses of methacholine (1.25–20 mg/ml) were delivered at 5-minute intervals using inhalation method.
Sections of paraffin-embedded lung tissue were stained with hematoxylin and eosin or used for immunohistochemical staining with mouse monoclonal antibodies against mast cell tryptase (clone AA1; Lab Vision, Fremont, CA). Antibody–antigen complexes were detected using anti-mouse antibody conjugated with alkaline phosphatase (AP) and BCIP/NBT substrate (all from Vector Labs, Burlingame, CA) as described (12). Lung sections were counterstained with nuclear fast red (Vector Labs).
Total mouse RNA was prepared by an acid guanidium-thiocyanate-phenol-chloroform extraction method using RNA-STAT-60 (Tel-Test “B” Inc., Friendswood, TX) from mouse lungs or primary lung endothelial cells. Primers specific for mouse tryptase (5′-gccaatgacacctactggatgc-3′ and 5′-gggtttgtgagtttcagcaggg-3′) were used to amplify tryptase cDNA fragment from total lung RNA and then clone it into the pCRII plasmid to synthesize antisense RNA probes. RNase protection assay was performed with [32P] UTP-labeled antisense RNA synthesized from mouse tryptase cDNA as described previously (23). RNA probe hybridization, RNase One (Promega, Madison, WI) digestion, electrophoresis of RNA protected fragments, and autoradiography were performed as described previously (13). Gene expression levels were determined by analysis of phosphorimager scans using the ImageQuant program (Amersham Biosciences, Inc., Piscataway, NJ) as described previously (24, 25). Hybridization signals for mouse cyclophilin and ribosomal L32 protein mRNAs were used for normalization.
Hybridization of the Applied Biosystems Mouse Genome Survey Microarray (P/N 4345065; Applied Biosystems, Framingham, MA) was performed using 2 μg of total lung RNA in the Functional Genomic Facility of the University of Chicago. For this assay, each sample contained a mixture of equal amounts of RNA from three distinct Foxf1+/− or WT mice, each treated with CCl4 for 16 hours. Images were analyzed using the Applied Biosystems 1700 Chemiluminescent Microarray Analyzer software v 1.1.
Total RNA was digested with RNAse-free DNase I to remove contaminating genomic DNA, and then purified using the Qiagen RNeasy Micro Kit (Cat. # 74004; Qiagen, Germantown, MD). We used the Bio-Rad cDNA Synthesis Kit (Bio-Rad, Hercules, CA) containing both oligo-dT and random hexamer primers to synthesize cDNA from 10 μg of total RNA as described (26). Quantitative RT-PCR primers for mouse cyclophilin were described previously (26). The following sense and anti-sense primers were used for amplification: mast cell tryptase, 5′-caccctgtccccctacct-3′ and 5′-atgtccttcattcccagcac-3′; Foxf1, 5′-cctggagcagccatacctt-3′ and 5′-taagatcctccgcctgttgt-3′; CXCL12, 5′-ctcctgtagaatggagccaga-3′ and 5′-ttcgatcagagcccatagag-3′; endomucin, 5′-caactacggcatgttttcca-3′ and 5′-gaggaaccaacacaatttcca-3′; mast cell protease -5, 5′-ttcatctgctgctccttctcctgg-3′ and 5′-ggctttctccttcaacttcagtagc-3′. Mouse cyclophilin was used for normalization control as described previously (25).
Endothelial cells (ECs) were isolated from 10-d-old WT and transgenic TetO-Foxf1 DN mice using a tissue digestion, binding with endothelial-specific Pecam-1 antibodies followed by EC purification using immunomagnetic beads as described (27, 28). The purity of pulmonary ECs was 95.1 ± 2.7%, as demonstrated by cytochemical staining with fluorescein isothiocyanate (FITC)-conjugated isolectin B4, a specific marker of endothelial cells (13).
The TetO-Foxf1 dominant-negative (DN) transgene contains the CMV-Tet operator (TetO) promoter, which drives the expression of a T7-tagged Foxf1 DNA binding domain fused to engrailed transcriptional repression domain (29). Cultured ECs were transiently transfected with 6× Foxf1-TATA-luciferase (LUC) reporter construct (30) and CMV-Foxf1 expression plasmid using Fugene 6 reagent (Roche, Indianapolis, IN) as described previously (29, 30). A CMV-Renilla construct was used as an internal control to normalize transfection efficiency. Two hours after transfection, ECs were infected at a multiplicity of infection (MOI) of 100 ifu per cell with adenovirus containing Tetracycline activator (Ad-TA, Tet-off system) or with control LacZ adenovirus (Ad-LacZ) as described (19, 29). Dual luciferase assays (Promega) were performed 48 h after the adenoviral infection as described previously (19, 30). In separate experiments, WT and transgenic ECs were infected with either Ad-TA or Ad-LacZ and then used for preparation of total RNA or for immunofluorescent staining. ECs were fixed with 10% paraformaldehyde and then stained with mouse monoclonal antibodies against T7 followed by anti-mouse antibody conjugated with TRITC as described (29).
The Student's t test was used to determine statistical significance. P values less than 0.05 were considered significant. Values for all measurements were expressed as the mean ± SD.
To assess whether Foxf1 insufficiency influenced lung injury, we used CCl4 and BHT injury models. In studies of CCl4 toxicity in Foxf1+/− mice, approximately 60% of the Foxf1+/− mice died within the initial 24 hours after CCl4 treatment (Figure 1A). While CCl4 is hepatotoxic, serum levels of liver aminotransferases, alkaline phosphatase, bilirubin, and albumin were similar in CCl4-treated Foxf1+/− and WT mice (data not shown), indicating that the increased mortality of Foxf1+/− mice during the initial 24 hours of CCl4 treatment did not result from liver failure. Since Foxf1 is essential for formation of pulmonary capillaries (13), we considered the possibility that capillary insufficiency and lung hemorrhage associated with partial loss of Foxf1 function may contribute to the observed increased early mortality in Foxf1+/− mice. However, at the light microscopic level, no detectable changes were observed in the vascular bed of CCl4-treated Foxf1+/− lungs compared with WT lungs (data not shown).
In contrast to WT mice, CCl4-treated Foxf1+/− lungs displayed severe narrowing of large airways (Figures 1B-1I), leading to airway occlusion (Figures 1J and 1K). The average diameter of pulmonary bronchioles in CCl4-treated Foxf1+/− mice was significantly decreased compared with pulmonary bronchioles in either WT mice or vehicle-treated Foxf1+/− mice (Figure 1N). Furthermore, the airway obstruction in CCl4-treated Foxf1+/− mice was associated with increased airway resistance as demonstrated by airway responsiveness to methacholine (Figure 1O). These results suggested that CCl4-induced bronchial obstruction contributed to the increased mortality of Foxf1+/− mice.
In contrast to WT mice, CCl4 treatment caused a leukocyte infiltration in Foxf1+/− lungs, which was mainly observed in interstitial tissue surrounding airways and pulmonary blood vessels (Figures 1D, 1E, 1H, and 1I). BAL fluid of CCl4-treated Foxf1+/− mice contained increased numbers of eosinophils and macrophages when compared with BAL of either WT mice or vehicle-treated Foxf1+/− mice (Figure 2A). Lymphocyte and neutrophil numbers were unaltered in BAL of Foxf1+/− mice (Figure 2A). Increased inflammation in CCl4-treated Foxf1+/− lungs was associated with increased concentrations of IL-4, IL-5, and IFN-γ proteins in BAL fluid (Figure 2B). These results demonstrated that CCl4 treatment induced pulmonary inflammation, leukocyte infiltration, and cytokine production in Foxf1+/− lungs.
Morphologic examination of HE-stained lung sections and immunostaining with tryptase antibody revealed increased numbers of mast cells in CCl4-treated Foxf1+/− lungs (Figures 3A–3F). Tryptase-positive mast cells were observed in alveolar, peribronchial, and perivascular regions (Figures 3G–3I). Increased tryptase immunoreactivity was also observed in the lumen of Foxf1+/− pulmonary airways (Figure 3F), a finding consistent with activation of mast cells (31, 32). Consistent with increased tryptase staining in CCl4-injured Foxf1+/− lungs, tryptase mRNA levels were increased 7-fold (Figure 3J). Furthermore, numbers of tryptase-positive mast cells in liver, kidney, and intestine of CCl4-treated Foxf1+/− mice remained unaltered (data not shown), indicating that the CCl4-mediated inflammation was restricted to the Foxf1+/− lung. Thus, haploinsufficiency of the Foxf1 gene caused bronchial obstruction and increased numbers of activated mast cells after CC14-induced injury, perhaps contributing to bronchial edema and mortality of CCl4-treated Foxf1+/− mice. Interestingly, numbers of tryptase-positive mast cells were increased in lungs of Foxf1+/− mice before injury (Figure 3D). Mast cells were not detected in normal lung tissue from untreated WT mice (Figure 3A). Thus, Foxf1 influences mast cell accumulation in the lung that may in turn influence susceptibility to lung injury.
To identify genes influenced by Foxf1 during lung injury, we performed hybridization analysis of Applied Biosystems mouse expression arrays with cDNA probes generated from total lung RNA prepared from CCl4-treated Foxf1+/− and CCl4-treated WT mice (see GEO database for a complete list of genes with altered expression levels in CCl4-treated Foxf1+/− lungs; www.ncbi.nim.nih.gov/geo; accession # GSE11112). This analysis revealed that CCl4-treated Foxf1+/− lungs exhibited a 2.5-fold reduction in Foxf1 mRNA levels (data not shown), consistent with diminished Foxf1 expression during BHT-mediated lung injury (17). Expression of several inflammation-associated genes was increased in Foxf1+/− mice after CCl4 treatment (data not shown), a finding consistent with increased inflammation in CCl4-treated Foxf1+/− mice. The up-regulated genes included endomucin and stromal cell–derived factor 1 (SDF-1 or CXCL12), the latter of which is known to regulate mast cell migration and chemotaxis (4).
Quantitative real-time RT-PCR (qRT-PCR) analysis was used to confirm that the CCl4-treated Foxf1+/− lungs exhibited diminished expression of Foxf1, whereas CXCL12 and endomucin mRNAs were increased in Foxf1+/− lungs (Figure 3K). Increased levels of CXCL12 protein were also found in BAL fluid of CCl4-treated Foxf1+/− mice (Figure 3L). CXCL12 expression was not altered in the liver (Figure 3K) or blood serum of Foxf1+/− mice (Figure 3L). Furthermore, CXCL12 mRNA and protein levels were similar in vehicle-treated WT and Foxf1+/− lungs (Figures 3K and 3L), indicating that the CXCL12 expression was induced by CCl4-mediated pulmonary inflammation.
To test whether lung injury responses were specific for CCl4, we tested whether Foxf1+/− mice were susceptible to BHT-induced lung injury. BHT was given as a single intraperitoneal injection to Foxf1+/− and WT male mice (12, 17). Total lung RNA was prepared 2 or 6 days after BHT injection and used for qRT-PCR. After injury, consistent with previous studies (17), Foxf1 mRNA was decreased in BHT-injured Foxf1+/− lungs compared with control WT lungs (Figure 4). Increased pulmonary inflammation with bronchial edema and increased CXCL12 mRNA were observed after BHT exposure in Foxf1+/− mice (Figure 4) (17, and data not shown). Mast cell tryptase and endomucin expression were also increased in lungs of BHT-injured Foxf1+/− mice (Figure 4), indicating the presence of activated mast cells and elevated mucus production.
Our studies demonstrated that haploinsufficiency of the Foxf1 gene causes an increase in the number of pulmonary mast cells, and renders the mice sensitive to bronchial inflammation and airway obstruction after CCl4 and BHT injury. Since increased numbers of mast cells were found in lungs of untreated Foxf1+/− mice, we considered that Foxf1 haploinsufficiency might contribute to antigen-mediated lung inflammation similar to that occurring in asthma. To test this, Foxf1+/− and WT mice were sensitized with three injections of OVA and then challenged with either aerosolized OVA or saline (22). Consistent with previous studies (22), OVA challenges induced inflammation in interstitial tissue surrounding airways and pulmonary blood vessels in control and Foxf1+/− mice (Figure 5A). Inflammation was more severe in Foxf1+/− mice, as demonstrated by an increased number of inflammatory cells in the lung tissue (Figure 5A) and BAL fluid (Figure 5B). BAL fluid of OVA-treated Foxf1+/− mice contained a significant increase in numbers of eosinophils, whereas numbers of macrophages, lymphocytes, and neutrophils remained unaltered compared with OVA-treated WT mice (Figure 5B). OVA-treated Foxf1+/− mice also displayed increased concentrations of IL-4 and IL-5 proteins in BAL fluid compared with OVA-treated WT mice (Figure 5C). Consistent with the increased pulmonary inflammation in Foxf1+/− lungs, airway resistance during methacholine treatment was significantly increased (Figures 5D and 5E). Increased inflammation in Foxf1+/− lungs was associated with reduced Foxf1 mRNA and increased expression of mast cell tryptase (Figure 5G), as well as with increased CXCL12 mRNA (Figure 5G) and protein levels (Figure 5F). These results are consistent with BHT and CCl4-mediated lung injury in Foxf1+/− mice. Taken together, these lung injury studies demonstrate that Foxf1 haploinsufficiency causes an increased susceptibility to either antigen-mediated or chemically-induced lung inflammation.
Next we determined whether increased CXCL12 expression and increased mast cell numbers occurred in Foxf1+/− embryonic lungs before birth. Paraffin lung sections from WT and Foxf1+/− E18.5 embryos were used for immunohistochemical staining with antibodies specific to mast cell protease 7 (tryptase), a known marker for mast cells (33). Although tryptase staining was not detected in WT lungs, increased numbers of tryptase-positive mast cells were observed in lungs from Foxf1+/− embryos at E18.5 (Figure 6C). Likewise, tryptase and mast cell–specific protease 5 mRNAs were increased in RNase protection assays (Figure 6A) and RT-PCR analysis (Figure 6B), confirming increased numbers of mast cells in Foxf1+/− lungs. Increased tryptase was observed in Foxf1+/− mice as early as E15.5 (Figure 6A). Furthermore, CXCL12 mRNA was increased in lungs from Foxf1+/− embryos at E15.5 and E18.5 (Figure 6D) and was correlated with elevated tryptase expression (Figure 6A) and increased numbers of pulmonary mast cells (Figure 6C). Since CXCL12 is essential for migration of mast cells into lung tissue (4), our results suggest that increased CXCL12 expression may contribute to the accumulation of mast cells seen in Foxf1+/− embryonic lungs. Interestingly, accumulation of mast cells was specific for Foxf1+/− lung tissue because normal mast cell numbers and tryptase mRNA levels were observed in embryonic Foxf1+/− livers and intestine (data not shown). These results suggest that Foxf1 haploinsufficiency causes pulmonary mastocytosis and increased CXCL12 expression during lung development.
The Foxf1 transcription factor is not expressed in pulmonary mast cells or their precursors (18, 34), but is co-expressed with CXCL12 in pulmonary endothelial cells (18, 34), supporting the concept that pulmonary mastocytosis is related to migration/ engraftment of mast cell precursors into the lung rather then changes in differentiation or growth of hematopoetic precursors per se. To determine whether Foxf1 directly affects CXCL12 expression in endothelial cells, we used primary lung ECs prepared from either WT or TetO-Foxf1 DN transgenic mice. The TetO-Foxf1 DN mouse line uses the CMV-Tet operator (TetO) promoter to drive expression of a T7-tagged Foxf1 dominant-negative (Foxf1 DN) fusion protein (29). Purity of pulmonary ECs was 95.1 ± 2.7%, as demonstrated by cytochemical staining with FITC-conjugated isolectin B4 (Figures 7A and 7B), a specific marker of endothelial cells (13).
To activate the TetO-Foxf1 DN transgene in cell culture, transgenic ECs were infected for 48 hours with either CMV-Tetracycline activator adenovirus (Ad-TA) or LacZ control adenovirus (AdLacZ). Immunofluorescent staining with T7 antibody demonstrated that the Foxf1 DN protein was induced by Ad-TA infection (Tet-off system) in transgenic TetO-Foxf1 DN ECs, but was not in Ad-TA–infected wild type ECs (Figure 7E) or in transgenic TetO-Foxf1 DN ECs infected with AdLacZ adenovirus (Figure 7D). Cotransfection studies with the CMV-Foxf1 expression vector and the 6× Foxf1-TATA-luciferase reporter plasmid showed that Foxf1 DN protein inhibited Foxf1 transcriptional activity in transgenic TetO-Foxf1 DN ECs when they were infected with Ad-TA adenovirus (Figure 7G). In contrast, both AdLacZ-infected transgenic ECs and Ad-TA–infected WT ECs displayed normal Foxf1 transcriptional activity compared with uninfected ECs (Figure 7G). Thus, conditional expression of the Foxf1 DN transgene in ECs reduced the Foxf1 transcriptional activity. Decreased Foxf1 activity in TetO-Foxf1 DN ECs was associated with increased CXCL12 mRNA as demonstrated by qRT-PCR (Figure 7H). Thus CXCL12 expression is enhanced by inhibition of Foxf1 in cultured pulmonary endothelial cells.
Foxf1 is an important transcriptional regulator expressed in the developing mesenchyme and its derivatives: fibroblasts, endothelial cells, and smooth muscle cells (13, 14, 18). Although previous studies demonstrated that Foxf1 was critical for vascular integrity during lung development and lung injury (13, 17), molecular mechanisms underlying the Foxf1 function in the lung remain largely unknown. In this study, we demonstrated that lungs of Foxf1+/− embryos exhibited increased numbers of mast cells and increased expression of tryptase, mast cell protease-5, and CXCL12 chemokine, the latter of which is known to promote mast cell migration and chemotaxis (4). These results suggest that haploinsufficiency of the Foxf1 gene increases pulmonary mast cell numbers during embryonic lung development, perhaps through a CXCL12-dependent mechanism. Activation of mast cells causes blood vessel dilatation and inhibits blood coagulation due to release of histamine and heparin, respectively (1, 2). Because a majority of Foxf1+/− mice exhibited a perinatal lethal phenotype due to pulmonary hemorrhage (13), it is tempting to speculate that mast cell–derived mediators contribute to the pulmonary hemorrhage seen in newborn Foxf1+/− mice (13).
In the present study, we demonstrated an increased susceptibility of Foxf1+/− mice to both chemical and allergen-mediated lung inflammation. In studies of CCl4 toxicity, severe airway obstruction and bronchial edema in Foxf1+/− mice preceded the onset of severe hepatic injury (18), suggesting that the liver injury does not cause mortality in Foxf1+/− mice. Pulmonary inflammation was associated with elevated tryptase and increased numbers of mast cells. Since degranulation of mast cells is known to cause the release of tryptase and histamine into the airways enhancing inflammation and leading to bronchial edema and bronchoconstriction (31, 32, 35, 36), our results suggest that CCl4 injury caused degranulation of pulmonary mast cells, contributing to bronchial edema and lethality of the Foxf1+/− mice from airway obstruction. Although CCl4-treated Foxf1+/− mice developed respiratory symptoms before death, it is not sufficient to conclude that the lethality occurred solely due to mast cell degranulation and pulmonary inflammation. Liver injury can also contribute, at least in part, to the lethality observed in CCl4-treated Foxf1+/− mice. In fact, abnormal degranulation of mast cells in Foxf1+/− lungs may be triggered by systemic release of cytokines, chemokines, and/or enzymes from injured liver. We previously demonstrated that hepatic expression of IFN-β, IFN-γ, TNF-α, and IL-6 were not changed in CCl4-treated Foxf1+/− mice, whereas the Foxf1+/− livers produced increased levels of TGF-β1 (18). Since TGF-β1 is a potent chemoattractant for mast cells (6), increased TGF-β1 activity in CCl4-treated Foxf1+/− mice may contribute to abnormal mast cell activation and/or recruitment in the lung. Furthermore, since Foxf1 is highly expressed in peribronchial smooth muscle cells (12, 13), it is also possible that Foxf1 haploinsufficiency influences muscular constriction in Foxf1+/− airways, thus contributing to airway obstruction and lethality.
Our previous studies demonstrated that Foxf1+/− mice died of pulmonary hemorrhage after BHT lung injury (17). This mortality with Foxf1+/− lung injury was associated with a significant reduction in pulmonary Foxf1 levels and diminished expression of genes critical for lung repair (17). In present study, we also found that BHT-treated Foxf1+/− lungs displayed elevated numbers of pulmonary mast cells and increased tryptase levels, causing airway obstruction. These results suggest that mast cell degranulation and bronchial edema contributes to the lethality in BHT-treated Foxf1+/− mice.
Another important finding of our studies is that Foxf1+/− mice displayed increased lung inflammation and airway hyperresponsiveness to methacholine after OVA sensitization and OVA challenge. Increased tryptase levels were also observed in OVA-challenged Foxf1+/− lungs, suggesting that antigen-mediated response can trigger mast cell degranulation in Foxf1+/− lungs. This may enhance pulmonary inflammation and airway hyperresponsiveness in Foxf1+/− lungs. Because increased numbers of mast cells were observed in untreated Foxf1+/− lungs, Foxf1 haploinsufficiency may contribute to a genetic predisposition to antigen-mediated lung inflammation and asthma. Interestingly, an increased number of activated mast cells is also characteristic of systemic mastocytosis, a heterogeneous group of human disorders characterized by myeloproliferative syndrome and elevated tryptase levels in multiple organs (8). Since ablation of a single allele of the Foxf1 gene caused increased numbers of mast cells, it will be of interest to assess status of the Foxf1 gene in human mastocytosis.
Although mechanisms underlying increased numbers of mast cells in Foxf1+/− lungs remain unclear, our results suggest that Foxf1 is involved in regulation of mast cell homeostasis or migration. The Foxf1 protein is a transcription factor, which is expressed in pulmonary fibroblasts, ECs, and smooth muscle cells, but not in hematopoetic cells (12, 13). Using either Foxf1 antibody or LacZ staining detecting a β-galactosidase reporter knocked into endogenous Foxf1 gene locus (13), we were unable to detect the Foxf1 in mast cells, basophils, or their precursors in the bone marrow or embryonic liver (18, 34, and data not shown). Therefore, Foxf1 regulates mast cell homeostasis or recruitment by an indirect mechanism, possibly involving pulmonary endothelial cells, a major cell type expressing Foxf1 protein in the lung (12, 13).
In the present study, we used an inducible Foxf1 dominant-negative transgene (Foxf1 DN) (29) to inhibit Foxf1 function in primary lung ECs prepared from either WT or TetO-Foxf1 DN transgenic mice. Our results demonstrated that Foxf1 depletion in cultured ECs induces the expression of CXCL12, a chemokine known to promote mast cell migration and chemotaxis (4). Thus, Foxf1 negatively regulates CXCL12 expression in cultured endothelial cells in vitro. Consistent with this hypothesis, increased expression of CXCL12 was found in Foxf1+/− lungs after either chemically induced (CCl4 and BHT) or antigen-mediated (OVA) lung injury, correlating with increased number of mast cells and elevated tryptase levels in Foxf1+/− lungs. Interestingly, five potential forkhead-binding sites are present in the −2 kb promoter region of the mouse CXCL12 gene (−936, −1,219, −1,335, −1,419, and −1,807), raising the possibility that Foxf1 may bind to this promoter region and act as a repressor of CXCL12 gene transcription. Alternatively, since a repressor function has not previously been reported for the Foxf1 transcription factor, the Foxf1 protein may indirectly regulate the CXCL12 gene. It is also possible that the absence of Foxf1 allows other transcriptional activators, more potent then Foxf1, to bind the promoter region, thus inducing CXCL12 gene expression.
In summary, Foxf1+/− mice displayed severe airway obstruction, elevated numbers of pulmonary mast cells, and increased tryptase levels after various forms of lung injury. Pulmonary inflammation in Foxf1+/− mice was associated with diminished expression of Foxf1, and increased expression of CXCL12, the latter of which is essential for mast cell migration and chemotaxis. Increased numbers of mast cells, and elevated CXCL12 levels, were found in lungs of Foxf1+/− embryos. Foxf1 depletion in cultured endothelial cells led to a significant increase in CXCL12 expression. Our results suggest that Foxf1 haploinsufficiency causes genetic predisposition to inflammatory lung diseases in mice due, at least in part, to increased recruitment of mast cells to the lung and increased CXCL12 production by pulmonary endothelial cells.
The authors thank S. Ramakrishna, D. Malin, and I. M. Kim for technical assistance and Lloyd Graf Jr. for useful comments. The authors dedicate this manuscript to the memory of Robert H. Costa, a pioneer in discovery and characterization of Forkhead transcription factors.
This work was supported by American Heart Association Grant 0335036N (to V.V.K.), Research Grant 6-FY2005–325 from the March of Dimes Birth Defects Foundation (to V.V.K.), and US Public Health Service Grant HL 84151 (to V. V. K.).
Originally Published in Press as DOI: 10.1165/rcmb.2008-0044OC on April 17, 2008
Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.