Main

Innate defense of the airways represents the first line of defense before immune-mediated mechanisms come into play. The upper and conducting airways are protected from microbial colonization by filtering mechanisms, the mucociliary escalator, and sneezing and coughing reflexes. Beyond the respiratory bronchioles, the alveoli are protected primarily by alveolar macrophages and by nonimmune humoral factors such as defensins, lysozyme (1), lactoferrin, and pulmonary surfactant.

Surfactant is a phospholipid-rich film that lines the alveolar epithelium and plays a critical role in preventing alveolar collapse at end expiration. In addition to surface tension-reducing properties, the lipid component of surfactant may contribute to host defense by forming a barrier that impedes attachment of pathogens to epithelial cells and facilitates clearance via the mucociliary escalator (for review see Refs.2 and3). Surfactant phospholipids may also modulate inflammatory responses associated with airway infection (4, 5). The hydrophilic surfactant-associated proteins, SP-A and SP-D, are more directly involved in host defense. In vitro, SP-A acts as an opsonin for various pathogens including bacteria, viruses, and fungi (6–9). SP-A also binds directly to alveolar macrophages stimulating chemotaxis (10) and enhancing pathogen killing through generation of bactericidal oxygen radicals (11, 12). SP-A levels in bronchoalveolar fluid are lower in patients with bacterial or viral pneumonia and in patients with cystic fibrosis whose airways are chronically colonized by bacteria (13, 14). SP-A -/- mice which lack detectable SP-A RNA and protein are more susceptible to GBS and Pseudomonas aeruginosa infection further supporting an important role for SP-A in host defense in vivo (15, 16). As for SP-A, SP-D binds to and aggregates some bacteria and viruses, enhances uptake of pathogens by alveolar macrophages and neutrophils (8) and elicits release of bactericidal oxygen radicals by alveolar macrophages (11). The role of SP-D in bacteria clearance in vivo has yet to be demonstrated.

In contrast to SP-A and SP-D, there is no evidence that the hydrophobic surfactant proteins, SP-B and SP-C, facilitate clearance of airway pathogens or interact with macrophages. However, a recent report suggested that the 79 amino acid SP-B mature peptide inhibited bacterial growth in vitro (17). In this study, the role of SP-B in bacterial clearance was tested by challenging transgenic mice, in which SP-B levels were either increased or decreased, with common airway pathogens.

MATERIALS AND METHODS

Mice

Generation of SP-B+/− mice has previously been described (18). Relative to wild-type littermates, lung tissue from SP-B+/− mice contains half as much SP-B mRNA, as assessed by S1 nuclease analyses, and half as much mature SP-B protein in bronchoalveolar lavage fluid, as assessed by ELISA. The generation of SP-B overexpressing mice has previously been reported (19). These mice contain 2- to 3- fold more SP-B in bronchoalveolar lavage fluid relative to wild-type littermates but are normal with respect to birth weight, lung weight, lung structure, lung function, somatic growth, and reproductive capacity. SP-B+/− mice were crossed with SP-B overexpressing mice (mSP-B+/+:hSP-B) to generate SP-B hemizygous mice that also express the human SP-B transgene (mSP-B+/−:hSP-B). Siblings with the mSP-B+/−:hSP-B genotype were mated to generate WT (mSP-B+/+), SP-B+/− and SP-B overexpressing (mSP-B+/+:hSP-B+) littermates for use in this study. All mice were housed under pathogen-free conditions and were studied under a protocol approved by the Institutional Animal Care and Use Committee. For each experiment, 5- to 6-wk-old (20–25 g body weight) littermates were used.

SP-B Protein Analysis

Bronchoalveolar lavage was performed by tracheal cannulation after intraperitoneal injection of pentobarbital. Lungs were lavaged with three 1-mL aliquots of PBS. SP-B concentration in pooled samples was determined by ELISA with a polyclonal antibody directed against mature bovine SP-B (19). Purified bovine SP-B was quantitated from amino acid analysis and used to generate a standard curve. SP-B protein was normalized to total lavage protein, as determined by bicinchoninic assay (20), and expressed as ng/mg total protein.

Bacteria

Stock cultures of Pseudomonas aeruginosa and group B Streptococci used in this study were clinical isolates provided by Dr. A. M. LeVine (Children's Hospital Medical Center, Cincinnati, OH).

GBS.

To minimize variation in virulence, all the bacteria used in this study were selected from aliquots of the same passage which had been frozen at −70°C in 20% glycerol/PBS. For each experiment, an aliquot of bacteria was thawed, plated on tryptic soy/5% defibrinated sheep agar plates and subsequently inoculated in 4 mL Todd-Hewitt broth (Difco, Detroit, MI). Bacteria were grown for 14–16 h at 37°C with continuous shaking. The bacteria were harvested from the broth by centrifugation at 200 ×g for 10 min, washed, and resuspended in sterile PBS at a concentration of 106 or 104/100 μL. The concentration of the inoculum was verified by quantitative culture on the sheep blood agar plates. The dose required to kill 50% of FVB/N mice infected with GBS (LD50) was determined to be 108 CFU/mL at 48 h post-infection.

P. aeruginosa.

P. aeruginosa obtained from a single passage was stored in aliquots at −70°C in 20% glycerol/2× yeast tryptone broth. For each experiment, an aliquot of bacteria was plated on 2XYT agar followed by inoculation into 4 mL 2XYT broth. Preparation of the inoculum was carried out as described for GBS; the concentration of the inoculum was verified by quantitative culture on 2XYT plates. The LD50 was 109 CFU/mL at 48 h post-infection.

Mice were anesthetized with isofluorane and their trachea exposed through an anterior midline incision. A dose of 104 or 106 CFU (GBS) or 107 CFU (P. aeruginosa) suspended in 100 μL of sterile PBS was delivered just beneath the cricoid cartilage. The incision was sealed by applying one drop of surgical glue (Nexabrand, Veterinary Products Laboratories, Phoenix, AZ). As a control, 100 μL nonpyrogenic PBS were similarly instilled into mice from each genotype.

To assess bacterial clearance, at 6 h (GBS and P. aeruginosa) or 24 h (P. aeruginosa) postinfection, mice were anesthetized with intraperitoneal pentobarbital, exsanguinated by transecting the abdominal aorta, and lung and splenic tissues were harvested, weighed, and subsequently homogenized in sterile PBS. Serial dilutions of homogenates were plated on blood agar (GBS) or 2XYT (Pseudomonas) plates and incubated at 37°C overnight. Viable pathogen counts in the lung and spleen were estimated from the number of colonies after 24 h of quantitative culture and expressed as CFU/g of tissue.

Statistical Analysis

Differences between groups were assessed by 2-way analysis of variance, and differences between means were assessed by contrast comparisons and the Student-Newman-Keuls test (StatView, Abacus Concepts, Inc. Berkeley, CA). Data are expressed as mean ± SD.

RESULTS

Alveolar SP-B Levels.

To minimize differences related to genetic background, wild-type, hemizygous, and SP-B overexpressing mice from the same litter were used for these experiments. The levels of surfactant protein B in BAL fluid from each of the three genotypes selected for this study were assessed by ELISA. SP-B levels in hemizygous mice (mSP-B±) were half of the level in wild-type mice (20 ± 4.1 ng/mg total protein versus 45.1 ± 5.7 ng/mg, p= 0.017) although the level of SP-B in overexpressing mice (94.2 ± 11.2 ng/mg) was more than 2-fold higher than wild-type littermates (p= 0.01) and nearly 5-fold higher than hemizygous littermates (p= 0.0008) (Fig. 1). Western analyses confirmed the results of the ELISA and indicated that the mature peptide was the only form of SP-B detected in the BAL from all three genotypes (not shown). Mice from each group were otherwise similar with respect to body and lung weights.

Figure 1
figure 1

SP-B in BAL fluid. SP-B mature peptide in BAL fluid obtained from 6-wk-old wild-type, hemizygous and SP-B overexpressing littermates was quantitated by ELISA. SP-B concentration (ng) was normalized to total protein in bronchioalveolar lavage (mg). Data are mean ± SEM. *Hemizygous mice versus WT mice, p= 0.02; **hemizygous versus overexpressors, p= 0.0008. n = Number of animals/experiment.

Full size image

Bacterial clearance from the airways.

To determine whether alveolar SP-B levels were correlated with bacterial clearance, GBS or P. aeruginosa was instilled into the airway of mice from each group. All mice survived until they were killed at either 6 (GBS and P. aeruginosa) or 24 h (P. aeruginosa) postinfection. Lungs and spleens were harvested, weighed, and homogenized in sterile PBS and equal volumes of homogenate plated for each genotype. Bacterial burden was assessed as the number of CFU/g of lung tissue after overnight incubation at 37°C.

Six hours after infection with intratracheally administered 106 CFU of GBS, the number of CFU/g lung tissue from wild-type mice was 7.2 ± 2.1 × 106 compared with 11.9 ± 3.8 × 106 for hemizygous (p= 0.54) and 15.9 ± 5.7 × 106 for SP-B overexpressing mice (p= 0.30 relative to wild-type) (Fig. 2). The incidence of systemic dissemination of GBS, as assessed by growth of colonies on plates inoculated with splenic homogenates, was not correlated with SP-B protein levels (60% in WT, 50% in hemizygous and 60% in SP-B overexpressing mice). There was also no significant difference in bacterial clearance or systemic dissemination among the groups when a lower dose of GBS (104) was administered (9.5 × 103 ± 0.8 CFU/g lung tissue in SP-B overexpressors compared with 11.2 × 103 ± 1.32 in hemizygous mice, p= 0.22 and 4.1 × 103 ± 0.3 in wild-type mice, p= 0.53). Systemic dissemination was not significantly different among groups but was reduced relative to the 106 CFU dose (wild-type 28%, hemizygous 14%, overexpressors 14%).

Figure 2
figure 2

Clearance of GBS from the airway at 6 h postinfection. Bacteria counts were determined by quantitative cultures of lung homogenates 6 h after intratracheal instillation of 106 CFU of GBS, as described in “Materials and Methods.” Data are CFU/g lung tissue ± SEM. WT versus hemizygous mice, p= 0.80; hemizygous versus overexpressors, p= 0.25. n = Number of animals/experiment.

Full size image

Six hours after instillation of P. aeruginosa, the number of CFU/g lung tissue from wild-type mice was 4.7 ± 1.1 × 107 compared with 6.0 ± 0.8 × 107 (p= 0.35) for hemizygous mice and 3.9 ± 0.7 for SP-B overexpressing mice (p= 0.55) (Fig. 3). Similar to results for GBS, systemic dissemination of P. aeruginosa in each group was not correlated with SP-B levels (50% in WT, 45% in hemizygous, and 42% in SP-B overexpressing mice).

Figure 3
figure 3

Clearance of P. aeruginosa from the airway at 6 h postinfection. Bacteria counts were determined by quantitative culture of lung homogenates 6 h after intratracheal instillation of 107 CFU of P. aeruginosa, as described in “Materials and Methods.” Data are CFU/g lung tissue ± SEM. WT versus hemizygous, p= 0.35; Hemizygous versus overexpressors, p= 0.83. n = Number of animals/experiment.

Full size image

Twenty-four hours after intratracheal infection with P. aeruginosa, quantitative culture of lung tissues from wild-type mice contained 9.2 ± 1.1 × 106 CFU/g compared with 6.4 ± 1.3 for hemizygous mice (p= 0.35) and 8.5 ± 1.8 × 106 for SP-B overexpressing mice (p= 0.78) (Fig. 4). As for the 6-h time point, systemic dissemination of P. aeruginosa was not correlated with SP-B protein levels (67% in WT, 60% in hemizygous, and 55% in SP-B overexpressing mice).

Figure 4
figure 4

Clearance of P. aeruginosa from the airway at 24 h postinfection. Bacteria counts were determined by quantitative cultures of lung homogenates 24 h after intratracheal instillation of 107 CFU of P. aeruginosa, as described in “Materials and Methods.” Data are CFU/g lung tissue ± SEM. WT versus hemizygous mice, p= 0.10; Hemizygous versus overexpressors, p= 0.43. n = Number of animals/experiment.

Full size image

DISCUSSION

The use of transgenic mice to manipulate SP-B levels offered several advantages over intratracheal instillation of SP-B. The mature SP-B peptide is extremely hydrophobic and aggregates extensively in aqueous solution, a property that would likely impede uniform distribution throughout the lungs and reduce potential antimicrobial activity. Administration of SP-B in liposomal form would decrease SP-B aggregation but introduce new variables related to changes in surfactant pool size and/or altered surfactant lipid composition at high doses of peptide. In the two transgenic mouse lines used in this study, surfactant pool size and lipid composition was unaltered and only SP-B peptide levels were affected.

Levels of mature SP-B peptide in BAL fluid of transgenic overexpressing mice were increased 5-fold relative to SP-B± littermates; alveolar surfactant lipid composition and pool size was not changed in either genotype (not shown). Increased SP-B protein levels in overexpressing mice were not accompanied by a concomitant change in bacterial clearance and decreased SP-B protein levels in SP-B± mice were not associated with increased susceptibility to bacterial infection. These results strongly suggest that levels of mature SP-B peptide do not influence bacterial clearance from the airway.

SP-B is a member of the saposin-like family of peptides that are characterized by a highly conserved pattern of intramolecular sulfhydryl bridges (21). The saposin-like family includes saposins A-D, the pore-forming peptide of Entamoeba histolytica, NK-lysin, acid sphingomyelinase, acyloxyacyl hydrolase, and several gene products of Caenorhabditis elegans (22–24). At least three of these peptides, NK-lysin, the pore-forming peptide and the amoeba pore homolog from C. elegans, exhibit significant antibacterial activity. The results of a recent study suggested that the mature SP-B peptide may also have antibacterial activity in vitro (17). Micromolar concentrations of a full length SP-B synthetic peptide inhibited growth of Escherichia coli in a dose-dependent manner when added directly to bacteria. In another report, SP-B was shown to attenuate lipopolysaccaride-induced nitric oxide production by rat alveolar macrophages (25). However, in the airspace mature SP-B peptide is always associated with surfactant phospholipids and the physiologic relevance of adding lipid-free SP-B to a bacterial suspension remains unclear. Consistent with this observation, SP-B in dodecylphosphocholine micelles did not exhibit detectable antibacterial activity toward E. coli or Bacillus megaterium (26). The latter in vitro study, taken together with the current in vivo study, argue against a significant role for mature SP-B peptide in innate host defense.

Surfactant protein B is synthesized in the distal airway epithelium as a 371 amino acid preproprotein in which the mature peptide is flanked by a 200 amino acid NH2-terminal and a 102 amino acid COOH-terminal propeptide. The mature peptide is secreted into the alveolar space after proteolytic cleavage of the flanking propeptides. The propeptides each contain a saposin-like domain that differs from the mature peptide in two important aspects: the saposin-like domains in the propeptides are significantly less hydrophobic than the mature peptide and lack the cysteine residue involved in homodimerization of mature SP-B. All SP-B proprotein is processed to mature peptide in type II cells. However, in Clara cells, SP-B proprotein is only processed to the Mr = 25 K processing intermediate and both the SP-B proprotein and 25 K form are secreted by Clara cells (unpublished). Our study cannot exclude the possibility that the saposin-like domains in the SP-B proprotein or processing intermediate (Mr = 25 K) are involved in bacterial clearance.