SEARCH

SEARCH BY CITATION

Keywords:

  • fibronectin;
  • alveolar development;
  • bronchopulmonary dysplasia;
  • innate immunity;
  • chorioamnionitis

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We tested the hypothesis that innate immune signaling in utero could disrupt the structural development of the fetal lung, contributing to the pathogenesis of bronchopulmonary dysplasia. Injection of Escherichia coli lipopolysaccharide (LPS) into the amniotic fluid of E15 BALB/cJ mice increased the luminal volume density of fetal mouse lungs at embryonic day (E) 17 and E18. LPS also increased luminal volume and decreased distal lung branching in fetal mouse lung explants. This effect required NF-κB activation and functional Toll-Like Receptor 4. Airway branching may require fibronectin-dependent epithelial–mesenchymal interactions, representing a potential target for innate immune signaling. Anti-fibronectin antibodies and LPS both blocked distal lung branching. By immunofluorescence, fibronectin localized to the clefts between newly formed airways but was restricted to peripheral mesenchymal cells in LPS-exposed explants. These data suggest that LPS may alter the expression pattern of mesenchymal fibronectin, potentially disrupting epithelial–mesenchymal interactions and inhibiting distal airway branching and alveolarization. This mechanism may link innate immune signaling with defects in structural development of the fetal lung. Developmental Dynamics 233:553–561, 2005. © 2005 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Lung development in mammals is a continuous process beginning early in gestation and continuing until after birth (Ten Have-Opbroek, 1981). After formation of the bronchi from the trachea, conducting airways bifurcate in an exact pattern. Branching of the conducting airways of the proximal or bronchial lung are completed by 16 weeks of gestation in humans (equivalent of embryonal day 15 in mice; Perl and Whitsett, 1999). Formation of the conducting airways during early lung development requires both physical interactions and bidirectional signaling between epithelia and adjacent lung mesenchyme (Shannon and Hyatt, 2004). Proximal lung branching requires the extracellular matrix protein fibronectin, which may promote the migration of mesenchymal cells into the clefts between branches (Sakai et al., 2003). After conducting airway formation, the distal airways form and branch during the canalicular phase of development from 17 to 23 weeks of human gestation (approximately embryonic day [E] 16–18 in mice; Inselman and Mellins, 1981; Ten Have-Opbroek, 1981, 1991). Distal airway branching has been less studied than proximal lung development; some molecular signals guiding the development of terminal airways into eventual alveoli may overlap with those of early lung development (Perl and Whitsett, 1999). Distal airway branching and alveolar septation form the gas-exchange unit of the lung, allowing respiration and survival after birth.

Due to advances in mechanical ventilation and surfactant replacement therapy, premature infants frequently survive when born after 23–25 weeks gestation (Draper et al., 2003), before the complete formation of alveolar structures. Distal lung development in these extremely premature infants, therefore, must continue after birth (Hislop, 2002). Postnatal development of the lung in premature infants frequently occurs abnormally, leading to bronchopulmonary dysplasia, a disease characterized by hypoxia, carbon dioxide retention, and altered pulmonary mechanics (Bancalari et al., 2003). Upon histological examination, the lungs of infants with bronchopulmonary dysplasia frequently contain fewer, larger alveoli with decreased alveolar septation and abnormal alveolar–capillary structures (Coalson, 2003). Understanding the basic mechanisms of distal lung development is critical to better prevent and treat bronchopulmonary dysplasia.

Exposure to inflammatory stimuli may contribute to the pathogenesis of bronchopulmonary dysplasia (Speer, 2003). Chorioamnionitis, or inflammation of the amniotic membranes, placenta, and/or uterus, is a frequent cause of preterm delivery and premature rupture of the amniotic membranes (Romero et al., 2002). Infants born in the presence of chorioamnionitis have a lower risk of surfactant deficiency and respiratory distress syndrome in the immediate perinatal period (Watterberg et al., 1996). Despite their improved lung function initially, these premature infants are at increased risk of developing bronchopulmonary dysplasia and other significant morbidities (Jobe, 2003). These clinical observations raise the possibility that prenatal exposure to inflammation may accelerate alveolar type II cell maturation and surfactant production but lead to defective development of distal lung structure (Jobe, 2003). We have developed a fetal mouse model of chorioamnionitis to investigate the molecular mechanisms linking the innate immune system and Toll-Like Receptor (TLR) signaling to the abnormalities of structural lung development seen in bronchopulmonary dysplasia. TLRs directly recognize foreign and microbial pathogens in the lung and other tissues, leading to activation of the innate immune response and release of inflammatory cytokines. Lipopolysaccharide (LPS) from Gram-negative bacteria activates innate immunity through TLR4. Unlike adaptive immunity, this response is genetically encoded and does not require antigen presentation (Beutler et al., 2003). By using this model, we have shown previously that signaling through TLR4 and NF-κB increased alveolar type II cell maturation in both fetal mouse lungs and fetal mouse lung explants (Prince et al., 2004). Whereas these findings might contribute to improved lung function after delivery, they did not explain how chorioamnionitis could lead to bronchopulmonary dysplasia. Because TLR/NF-κB signaling appeared to influence fetal lung development in the absence of inflammatory cell recruitment or injury, we hypothesized that innate immune signaling in the fetal lung could directly disrupt structural development of the distal lung.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

To initially test if innate immune signaling could disrupt fetal lung development, we developed a mouse model of chorioamnionitis, directly injecting Escherichia coli LPS into the amniotic fluid of E15 fetal BALB/cJ mice. The fetal mice, therefore, were exposed to LPS during the period of distal airway development. LPS did not significantly alter the weight of E16–E18 lungs in these animals, and other internal organs were grossly normal. Mice exposed to LPS in utero were live born and survived the perinatal period. To determine whether LPS altered structural development of the lung, we performed morphometric analysis of fetal mouse lungs exposed to LPS. After fixation, sectioning, and staining of the fetal lungs at E17 and E18 (48 hr and 72 hr after injection, respectively), we observed dilated distal airways in LPS-exposed animals compared with saline-injected controls (Fig. 1A–F). The luminal volume density in fetal lungs exposed to LPS was increased at both time points (Fig. 1G,H). LPS also appeared to decrease the number of branches forming from distal airspaces (Fig. 1D,F). To test if the observed changes in fetal lung development were specific for LPS signaling through TLR4, we performed identical experiments in C.C3H-Tlr4Lpsd mice, a congenic BALB/cJ strain possessing a loss of function mutation in TLR4 (Vogel et al., 1994). LPS injection did not increase luminal volume density or branching in E18 fetal C.C3H-Tlr4Lpsd lungs (Fig. 1I–K).

thumbnail image

Figure 1. Lipopolysaccharide (LPS) increased distal airway luminal volume in fetal mouse lungs. A–F: Either saline (control; A,C,D) or LPS (B,E,F) was injected into the amniotic fluid of embryonic day (E) 15 mice. Fetal mouse lungs were obtained at E17 (A,B) and E18 (C–F), processed, sectioned, and stained with hematoxylin and eosin. LPS appeared to increase the size of the distal airway lumen at both time points. Areas outlined by boxes in C and E are magnified in D and F, showing decreased airway branching in LPS-exposed lungs. G,H: Luminal volume density was measured in multiple sections as described in the Experimental Procedures section. Three lungs from each of five saline-injected and five LPS-injected litters were measured at both E17 (G) and E18 (H; *P < 0.05). I–K: LPS did not alter fetal lung structure in C.C3H.Tlr4Lpsd mice. I,J: Hematoxylin and eosin-stained C.C3H.Tlr4Lpsd E18 fetal lung sections were photographed by brightfield microscopy. Representative sections are shown. K: Multiple sections of four lungs from each of three saline-injected and three LPS-injected litters were analyzed to measure luminal volume density. No significant changes were observed after LPS injection in C.C3H.Tlr4Lpsd mice.

Download figure to PowerPoint

Injection of LPS into the amniotic fluid establishes a local inflammatory response, recruiting maternal leukocytes to the uterine wall (Prince et al., 2004). Although we did not detect gross histological damage or inflammatory infiltrate in hematoxylin and eosin-stained sections of LPS-exposed fetal lungs, cytokines from the maternal or fetal and systemic circulations could alter fetal lung development. The fetal mouse lung expresses TLR4 on the surface of conducting airway epithelia and in the fetal lung mesenchyme adjacent to alveoli (Prince et al., 2004). Therefore, the fetal mouse lung may be capable of responding to LPS directly without systemic influences. We, therefore, tested the direct effects of LPS on cultured E16 fetal lung explants in the absence of either fetal or maternal systemic responses. Similar to our in vivo data, LPS increased the luminal airway volume of cultured explants (Fig. 2). LPS decreased epithelial but not interstitial volume density in fetal lung explants (Fig. 2A). LPS did not appear to change cellular proliferation in fetal lung explants as measured by 5-bromo-2-deoxyuridine incorporation (not shown). These findings suggested that LPS directly altered structural development of the fetal lung.

thumbnail image

Figure 2. Lipopolysaccharide (LPS) increased distal airway luminal volume in fetal mouse lung explants. Embryonic day (E) 16 fetal mouse lungs were minced and cultured as explants for 72 hr. The explants were then fixed, processed, and stained with hematoxylin and eosin. A: Luminal, epithelial, and interstitial volume densities were measured from random sections (*P < 0.05; n = 12). B–E: Representative photomicrographs from control explants (B,D) or explants exposed to LPS (C,E). Areas outlined by boxes in B and C are magnified in D and E, showing decreased branching of distal airways and prominent mesenchymal cells surrounding the airway in LPS-exposed explants. F,G: Anti-vimentin labeling in control and LPS-exposed explants. Mesenchymal cells in control and LPS-exposed explants were labeled with a goat anti-vimentin polyclonal antibody (magenta) and visualized by confocal microscopy. Nuclei were labeled with SYTO 13 (yellow-green).

Download figure to PowerPoint

When examined by higher magnification, the newly formed airways along the periphery of LPS-exposed explants appeared smaller than controls (Fig. 2D,E). The mesenchyme surrounding newly formed airways also was more prominent in LPS-exposed explants. Immunofluorescent labeling with anti-vimentin antibodies more clearly differentiated mesenchymal cells from adjacent epithelia (Fig. 2F,G). Confocal microscopy sections revealed more mesenchymal cells adjacent to LPS-exposed peripheral airways compared with controls. The morphology of these distal airways and the surrounding mesenchyme suggested LPS disrupted distal airway branching, resulting in a more simplified, dilated airspace.

By visualizing cultured explants with light microscopy, we quantified distal airway formation and branching over time (Fig. 3A,B). Within 72 hr of culture, the number of peripheral airway branches nearly doubled in control explants, whereas LPS inhibited any increase in branch number (Fig. 3C). We also measured airway complexity by determining whether visualized airways went through multiple generations of branching (complex) or simply branched from a central airway without further division (not complex). LPS decreased the percentage of complex airways in fetal lung explants (Fig. 3D). Decreased branching with LPS did not appear to be due to inhibition of explant growth, as LPS also decreased the number of peripheral branches when normalized to explant area (Fig. 3E). LPS, therefore, inhibited distal lung branching without the contribution of maternal and fetal systemic immune responses. LPS could mediate these effects by directly signaling the TLR4/NF-κB pathway in the fetal lung.

thumbnail image

Figure 3. Lipopolysaccharide (LPS) inhibited distal airway branching in fetal mouse lung explants through TLR4/NF-κB signaling. A–C: BALB/cJ fetal lung explants were cultured in the absence or presence of LPS (250 ng/ml). Individual explants were imaged by phase microscopy over 72 hr. The number of peripheral branches around each explant was counted at 24 hr, 48 hr, and 72 hr of culture. LPS prevented the increase in peripheral branching seen in controls (*P < 0.05; n = 9). D: Branch complexity (% of complex branches/total branches) in control and LPS-treated explants was calculated by digitally mapping and analyzing the distal airways of explants cultured for 72 hr. LPS-exposed airways had lower complexity than controls (*P < 0.05; n = 12). E: To account for potential changes in explant growth, the number of branches at 72 hr of culture was normalized to area. Explants were cultured in the presence of LPS (250 ng/ml) with or without the NF-κB inhibitor parthenolide (1 μM; *P < 0.05; n = 19 for control and LPS, n = 14 for LPS + parthenolide). F–H: The number of peripheral branches/area was also measured in E16 explants obtained from C.C3H-Tlr4Lps-d mice (a congenic strain bearing a mutation in the tlr4 gene) and cultured for 72 hr in the presence or absence of LPS. Phase images are shown in F and G. H: LPS did not change the number of branches per area in 72 hr explants cultured for 72 hr (n = 12).

Download figure to PowerPoint

We next tested if the changes measured in fetal lung explants exposed to LPS required NF-κB activation. We previously determined that the sesquiterpene lactone parthenolide inhibited NF-κB activation in fetal lung explants after LPS exposure (Prince et al., 2004). In explants cultured with parthenolide (1 μM) in the medium, LPS did not decrease the number of distal airway branches/area (Fig. 3E). These results suggested the effects of LPS on distal airway branching were mediated by NF-κB activation. We then tested the requirement of the LPS receptor TLR4 on distal airway branching by culturing fetal lung explants from C.C3H-Tlr4Lpsd mice. Fetal lung explants from C.C3H-Tlr4Lpsd mice branched with similar efficiency as BALB/cJ explants, but LPS did not alter their branching (Fig. 3F–H). These findings increased the likelihood that LPS decreased airway branching through stimulating the TLR4/NF-κB pathway.

LPS alters cellular adhesion and migration of circulating immune cells (Hazeki et al., 2003; Roman et al., 2004). We hypothesized that LPS might disrupt similar processes in the fetal lung during distal development. Fibronectin mediates cell adhesion and spreading and is required for branching morphogenesis in the submandibular gland and during development of proximal conducting airways (Sakai et al., 2003). We, therefore, tested if fibronectin was also required for distal lung branching and if LPS could alter its expression. To examine the role of fibronectin in distal airway branching, we used antibodies to block fibronectin-dependent interactions in cultured fetal mouse lung explants. Anti-fibronectin antibodies inhibited branching in a concentration-dependent manner (Fig. 4); nonimmune rabbit IgG had no effect. We also tested the effects of supplemental fibronectin on distal airway development. Addition of fibronectin to the culture medium increased airway branching and promoted thinning of the mesenchyme adjacent to airways (Fig. 5). Fibronectin also rescued the effects of LPS on fetal lung explants (Fig. 5C–F). Supplemental fibronectin, therefore, increased peripheral branch formation in both control and LPS-exposed explants. Fibronectin, therefore, appeared to play a critical role in distal airway development similar to its involvement in proximal airway branching.

thumbnail image

Figure 4. Anti-fibronectin antibodies inhibited distal airway branching in fetal mouse lung explants. A–C: Embryonic day 16 BALB/cJ fetal mouse lung explants were cultured in the presence of nonimmune rabbit IgG (25 μg/ml; A) or increasing concentrations of polyclonal rabbit anti-fibronectin (FN) antibodies (B,C). After 72 hr of culture, the explants were fixed, stained with DAPI (4′,6-diamidine-2-phenylidole-dihydrochloride) for visualization of the distal airways, and imaged by fluorescence microscopy. D: Increasing concentrations of anti-FN antibodies inhibited distal airway branching. After 72 hr of culture, the number of distal branches per area was measured (*P < 0.05; n = 12).

Download figure to PowerPoint

thumbnail image

Figure 5. Fibronectin (FN) augmented distal airway branching in fetal mouse lung explants. A,B: Supplementation of the cell culture medium with fibronectin (50 μg/ml) promoted thinning of the distal interstitium and division of distal airways. C–E: Addition of fibronectin increased the expansion and branching of distal airways in lipopolysaccharide (LPS) -treated explants. The interstitium surrounding the LPS-treated distal airways also appeared thinner when cultured with exogenous FN. F: Peripheral airway branches were counted, showing increased airway branching in control (ctrl) and LPS-treated explants after 72 hr of culture (*P < 0.05; n = 8).

Download figure to PowerPoint

We then tested if LPS could alter fibronectin expression in the developing lung, leading to altered epithelial–mesenchymal interactions and abnormal lung branching. We localized fibronectin expression in control and LPS-treated fetal mouse lung explants by immunolabeling and fluorescence microscopy. In control explants, fibronectin protein was detected in the mesenchyme, primarily in the clefts between distal airways (Fig. 6). LPS dramatically altered the localization of fibronectin in fetal lung explants. Instead of being found in the clefts between distal airways, fibronectin labeling was detected along the perimeter of the explants in a circumferential pattern (Fig. 6C). The total amount of fibronectin seen by immunoblot was not significantly different between control and LPS-treated explants (Fig. 6D). LPS did not change the immunolocalization pattern in C.C3H.Tlr4Lpsd explants (Fig. 6E,F), suggesting the alterations after LPS exposure required signaling through TLR4.

thumbnail image

Figure 6. Fibronectin (FN) immunolocalization and expression in fetal mouse lung explants. A: Embryonic day (E) 16 fetal lung explants were cultured for 72 hr, fixed, and labeled with anti-fibronectin antibodies. The nuclei were labeled with DAPI (4′,6-diamidine-2-phenylidole-dihydrochloride). Fluorescence images were obtained on whole mount specimens (×20 objective). Fibronectin labeling is shown as magenta staining in the clefts between distal airways and on the surface of mesenchymal cells. B,C: Fibronectin staining in control and lipopolysaccharide (LPS) -treated explants. E16 fetal lung explants were cultured for 72 hr in the absence (B) or presence (C) of LPS (250 ng/ml). The explants were then fixed and labeled with anti-fibronectin antibodies. Whereas staining in control explants was seen in the clefts between airways, LPS-treated explants expressed fibronectin only in the mesenchyme lining the explant edge. Fibronectin staining was dramatically diminished in the airway clefts of LPS-treated explants. Representative images are shown. D: Fibronectin immunoblot of fetal lung explant homogenates. Fetal mouse lung explants were cultured in the absence or presence of LPS (250 ng/ml) for 72 hr, lysed, and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblotted using a rabbit polyclonal antibody against fibronectin. No significant difference between control and LPS samples was detected (n = 4). E,F: Fibronectin staining in fetal lung explants from C.C3H.Tlr4Lpsd mice. Explants were culture for 72 hr in the absence (E) or presence (F) of LPS (250 ng/ml). Fibronectin was labeled and imaged as in B,C.

Download figure to PowerPoint

Confocal sectioning at high magnification further demonstrated the effects of LPS on fibronectin immunostaining in developing fetal lung explants (Fig. 7). In controls, fibronectin was assembled on the surface of mesenchymal cells located within the clefts between developing airways (Fig. 7A–D). In LPS-treated explants, fibronectin was assembled only on mesenchymal cells located at the explant margin, several cell layers away from the airway epithelium (Fig. 7E–H). To determine whether LPS could directly alter the fibronectin expression in fetal lung mesenchyme, we examined fibronectin expression in cultured primary fetal lung mesenchymal cells treated with LPS. In control cells, fibronectin localized to thin filaments on the cell surface (Fig. 7I,J). LPS treatment altered the fibronectin expression pattern in mesenchymal cells, forming thick linear immunoreactive structures at the cell periphery (Fig. 7K,L). Cells exposed to LPS also appeared more elongated than control cells. These data suggest that LPS may cause differences in fibronectin assembly at the cell surface of fetal lung mesenchymal cells.

thumbnail image

Figure 7. Fibronectin expression in fetal mouse lung explants. A: Fibronectin was immunolocalized using a laser-scanning confocal microscope. Fibronectin labeling was pseudocolored magenta and nuclei were stained with DAPI (4′,6-diamidine-2-phenylidole-dihydrochloride). Images were collected at high magnification (×100 objective). z sections were obtained at 0.2-μm intervals throughout the depth of the explant (total of 150–180 sections). Fibronectin expression was detected predominantly on cells within the mesenchyme between developing airways. The airway lumen is identified with an asterisk (*).B,C:xz projections show fibronectin adjacent to distal airways. D:yz projection. E–H: Fibronectin expression pattern in lipopolysaccharide (LPS) -treated explant after 72 hr of culture as seen by confocal microscopy. E: Representative xy section shows fibronectin at the outer edge of the explant. F,G:xz projections reveal fibronectin at the explant edge but not within the mesenchyme adjacent to airways (*). H:yz projection of fibronectin in LPS-treated explant. I–L: Fibronectin expression in primary fetal lung mesenchymal cells. I,J: Representative control cells labeled with anti-fibronectin antibodies and imaged by immunofluorescence. Fibronectin (magenta) localized to thin, filamentous structures on the cell surface (arrowheads). K,L: Cells treated with LPS for 72 hr show thick, long fibronectin-containing structures at the cell margins (arrows).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Abnormal branching and septation of the distal airways after preterm delivery may lead to defective alveolar formation and the development of bronchopulmonary dysplasia (Coalson, 2003). The clinical observation that premature infants exposed to chorioamnionitis are at increased risk of bronchopulmonary dysplasia suggests that inflammatory signals can alter lung development (Watterberg et al., 1996). We have tested the role of innate immune signaling on the structural development of the distal fetal mouse lung. By using an in vivo model of chorioamnionitis and cultured fetal lung explants, we found that LPS altered distal airway formation, preventing branching of distal airways into complex structures. These effects of LPS required signaling through TLR4 and NF-κB. In searching for candidate targets for the effects of LPS, we found that fibronectin expression within the fetal lung mesenchyme was misplaced in explants after LPS treatment, possibly contributing to altered airway structure. This link between innate immune signaling and the regulation of epithelial–mesenchymal interactions could identify a mechanism in the pathogenesis of bronchopulmonary dysplasia.

We speculate that abnormal interactions between LPS-treated mesenchymal cells and epithelia prevented normal airway branching. Previous studies using animal models of bronchopulmonary dysplasia and human autopsy specimens have described alterations in lung cell matrix proteins, cellular differentiation, and alveolar structure (Bland et al., 2003; Coalson, 2003; Kaarteenaho-Wiik et al., 2004). LPS appears to also cause alveolar simplification in fetal sheep, possibly arising from decreased distal airway branching and septation (Willet et al., 2000).

The importance of this study is the connection between the innate immune system and abnormal structural development in the fetal mouse lung. Our data suggest that innate immune signaling alters fibronectin expression and/or assembly. Fibronectin is an important cellular matrix protein, required for branching in submandibular gland, kidney, and early embryonal lung (Jiang et al., 2000; Sakai et al., 2003). Expression of fibronectin adjacent to the basolateral surface of airway epithelia displaces homotypic epithelial cell–cell interactions, allowing migration of mesenchymal cells into newly formed clefts within the epithelia (Sakai et al., 2003). We propose that LPS disrupts the normal interactions between epithelia and mesenchyme, leading to abnormal structural development in the fetal lung.

The fibronectin expression pattern in LPS-treated explants was clearly different than in controls, with very little fibronectin expression detected adjacent to the airway epithelia. We have yet to determine the exact mechanism by which LPS alters fibronectin expression or if this change is completely responsible for the abnormal airway branching seen with LPS. TLR/NF-κB signaling could either change which cells express fibronectin or alter the expression of fibronectin receptors, which facilitate cell surface fibronectin assembly. Mesenchymal cells exposed to LPS were more elongated and incorporated fibronectin into thicker, more linear structures. These alterations resulted from either changes in mesenchymal cell differentiation or fibronectin biosynthesis and assembly. Differences in mesenchymal cell phenotype could lead to abnormal cell migration, apoptosis, or proliferation.

We have used E. coli LPS in our model as a potent, well-described activator of innate immunity (Beutler and Poltorak, 2001; Akira and Takeda, 2004). Different microbial organisms can all activate the innate immune system through various TLRs (Beutler et al., 2003). Although the exact molecular response to different organisms may differ slightly, activation of NF-κB is a common signaling pathway for most innate responses (Hallman et al., 2001). Therefore, the various microbial agents causing chorioamnionitis each lead to NF-κB activation and would likely alter development in the fetal lung. We have not yet determined if the structural changes we observed after LPS exposure were mediated by downstream inflammatory cytokines and chemokines. Inflammatory mediators amplify the LPS-initiated signal by additionally activating NF-κB in cells not directly activated by LPS (Martin and Wesche, 2002; Suzuki et al., 2003). Identification of cytokines or signaling pathways contributing to defective lung development in bronchopulmonary dysplasia could provide potential therapeutic targets for the treatment and prevention of lung disease in premature infants.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Cell Culture and Reagents

Primary fetal lung mesenchymal cells were isolated by allowing attachment and outgrowth of cells from E16 fetal lungs onto tissue culture-treated plastic dishes. After subculture, cells were maintained in DMEM/10% fetal bovine serum. Parthenolide was purchased from Calbiochem. Polyclonal goat anti-vimentin IgG (C-20) was purchased from Santa Cruz Biotechnology. Fluorescent secondary antibodies were obtained from Kirkegaard-Perry Laboratories and Molecular Probes. DAPI (4′,6-diamidine-2-phenylidole-dihydrochloride) and SYTO 13 were purchased from Molecular Probes.

Polyclonal anti-fibronectin antibodies were raised in rabbits (Rockland Immunochemicals) using Freund's adjuvants (Colorado Serum). The animals were immunized by using 8–10 synthetic peptides whose sequences were derived from selected regions of the fibronectin protein. The synthetic peptides were pooled and conjugated to KLH (Calbiochem) using multiple cross-linking chemistries. Antibodies were affinity purified on columns of native human fibronectin. Human fibronectin was extracted from serum using columns of denatured gelatin bound to CNBr activated Sepharose-4B (Pharmacia) and arginine elution buffers (Vuento and Vaheri, 1978) and then purified by gel filtration using Sephacryl-S300 (Pharmacia). The purified antibodies were cross-absorbed against pepsin digested purified human collagen types I–VI and human placental laminin, concentrated and dialyzed into phosphate buffered saline.

Mouse Strains and Chorioamnionitis Model

BALB/cJ and C.C3H.Tlr4Lpsd mice were purchased from Jackson Laboratories and maintained in pathogen-free facilities. Breeding colonies were established for timed matings, with the day of vaginal plug was defined as E0. All animal procedures and protocols were reviewed and approved by the Institutional Animal Care and Use Committee at the University of Alabama at Birmingham.

For the establishment of chorioamnionitis in pregnant mice, E15 mice were anesthetized with sodium pentobarbital injection (50 mg/kg i.p.). The abdominal wall was infiltrated with 0.1–0.2 ml of bupivacaine. A 1-cm midline abdominal incision allowed externalization of the uterine horns. A fine-tipped pulled glass pipette was used for direct intra-amniotic injection of 5 μl of either sterile, endotoxin-free saline or LPS (20 ng/ml) into the amniotic sac of each fetus. After internalization of the uterus, the abdominal wall was sutured in two layers. Mice were returned to their cages, given food and water ad libitum, and monitored for signs of pain or distress. Death 24–72 hr after surgery allowed procurement of amniotic fluid, uterine wall with membranes, placenta, and intact lungs.

Fetal Mouse Lung Explant Isolation and Culture

E16 female mice were killed with sodium pentobarbital (150 mg/kg i.p.). Using sterile technique, the fetal mouse lungs were isolated and dissected free of surrounding structures. The lung tissue was minced into 0.5- to 1-mm3 cubes using dissecting scissors and cultured on an air–liquid interface using permeable supports (Costar Transwell) and serum-free DMEM. Explants were cultured at 37°C in 95% air/5% CO2 for up to 72 hr.

Fibronectin Immunoblotting

Fetal lung explants cultured for 72 hr in the absence or presence of LPS were lysed using a Dounce homogenizer in RIPA buffer (150 mM NaCl, 25 mM Tris pH 7.4, 0.5% sodium deoxycholate, 1% Nonidet P-40, 0.1% SDS) containing a protease inhibitor cocktail (Roche). Protein concentrations were measured using BCA reagents (Pierce). Ten micrograms of total protein was separated on 6% sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels and transferred to polyvinylidene difluoride membranes (Millipore). Membranes were blocked with 5% bovine serum albumin and incubated overnight with rabbit anti-fibronectin antibodies. Immunoreactive signal was developed using an anti-rabbit peroxidase secondary and chemiluminescence (Amersham). Blots were exposed on X-ray film and quantified using a CCD camera and densitometry (GelLogic, Kodak).

Microscopy and Immunostaining

Images of fetal lung explants were obtained using an inverted phase Axiovert 25 with a color CCD camera (Q Imaging). Images were saved as either TIFF or high-quality JPEG files for image analysis. For immunolabeling, intact fetal lung explants were fixed in 4% paraformaldehyde and permeabilized with 0.2% Triton X-100. Nonspecific binding was reduced with SuperBlock (Pierce). Primary antibodies were incubated with the explants overnight at 4°C. After washing, Texas Red or Alexa 594 anti-rabbit secondary antibodies were bound at room temperature. The explants were then treated with DAPI for nuclear labeling and mounted between glass slides and 0.17-mm coverslips using Vectashield anti-fade aqueous mounting medium. The explants were not removed from their permeable supports for labeling or imaging. Wide-field immunofluorescent images were captured by using a Leica DM RXA2 and a monochromatic CCD camera (Hamamatsu). Confocal images were obtained using a Leica DMIRB laser-scanning confocal microscope. For three-dimensional (3-D) reconstruction and rendering, 0.2-μM sections were imaged in the z-axis. Approximately 100 sections were typically obtained for 3-D rendering of an immunostained fetal lung explant.

Image Analysis and Statistics

For presentation of immunofluorescent data, monochromatic images of DAPI-stained nuclei and Texas Red- or Alexa-labeled antibodies were merged. Fluorescent red layers were pseudocolored magenta using Photoshop (Adobe). For measurement and calculation of volume density, random sections through paraffin-embedded fetal lungs and fetal lung explants were stained with hematoxylin and eosin and imaged by brightfield microscopy. By measuring the fraction of each field that was occupied by distal airway lumen in multiple sections, the volume density of the distal airway was measured. Copies of each image were printed and overlaid with a 10 × 10 morphometry grid. Volume density was then measured in quadruplicate for each image as described (Weibel et al., 1966). Measurements were confirmed using a luminal field fraction digital image analysis protocol to measure volume density of distal airways in multiple adjacent sections (Histometrix, Kinetic Imaging). Brightfield images of fetal lung explants were imported into Histometrix for analysis of peripheral branch number. For branch complexity, the distal airways were mapped in Photoshop and imported into Image (Scion). The background image was removed by using the threshold function. The mapped airways were converted to binary mode and then skeletonized for complexity analysis. The percentage of airway branches that underwent multiple generations of branching was calculated for each image. All statistical data were exported into Excel spreadsheets (Microsoft). For analysis of confocal z-series images, virtual sectioning was performed in LCS Lite (Leica). Images were saved as TIFF files and processed in Photoshop. All comparison images were processed identically. Data are presented as mean ± SEM. Unpaired Student's t-tests were performed for statistical comparisons.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We thank our colleagues for their helpful comments and suggestions. L.S.P. is a fellow of the Parker Francis Families Foundation. H.I.D. is a NIH/UAB Summer Research Fellow.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES