Intratracheal bleomycin in rats is associated with respiratory distress of uncertain etiology. We investigated the expression of surfactant components in this model of lung injury. Maximum respiratory distress, determined by respiratory rate, occurred at 7 days, and surfactant dysfunction was confirmed by increased surface tension of the large-aggregate fraction of bronchoalveolar lavage (BAL). In injured animals, phospholipid content and composition were similar to those of controls, mature surfactant protein (SP) B was decreased 90%, and SP-A and SP-D contents were increased. In lung tissue, SP-B and SP-C mRNAs were decreased by 2 days and maximally at 4–7 days and recovered between 14 and 21 days after injury. Immunostaining of SP-B and proSP-C was decreased in type II epithelial cells but strong in macrophages. By electron microscopy, injured lungs had type II cells lacking lamellar bodies and macrophages with phagocytosed lamellar bodies. Surface activity of BAL phospholipids of injured animals was restored by addition of exogenous SP-B. We conclude that respiratory distress after bleomycin in rats results from surfactant dysfunction in part secondary to selective downregulation of SP-B and SP-C.
- surface tension
- surfactant hydrophobic proteins
- lung injury
in animal models, acute lung injury of various etiologies results in respiratory distress in association with an inflammatory response. Similarly, human diseases such as respiratory distress syndrome in preterm infants (31), acute respiratory distress syndrome in older children and adults (57), and pulmonary infections such as Pneumocystis carinii (2) and respiratory syncytial virus (54) all involve inflammation and respiratory distress. The pathophysiology of respiratory distress in these conditions involves a deficiency and/or dysfunction of pulmonary surfactant producing generalized atelectasis, intrapulmonary arteriovenous shunt, and hypoxemia (30).
Surfactant is a complex mixture of phospholipids, neutral lipids, and proteins that reduces surface tension at the air-liquid interface in alveoli, thereby preventing alveolar collapse (29). Phospholipids constitute ∼70% of surfactant, consisting largely of phosphatidylcholine (PC), phosphatidylglycerol (PG), and phosphatidylinositol (PI) (5). Phospholipid interaction with specific surfactant proteins (SPs), in particular SP-B, has been shown to be critical for normal alveolar function (60). Transgenic mice lacking the SP-B gene (14) and human neonates with a congenital deficiency of SP-B (43) develop normally in utero but fail to expand their lungs at birth and die secondary to respiratory failure. In both conditions, abnormal processing of SP-C accompanies the primary gene defect and results in combined SP-B and SP-C deficiency.
Intratracheal instillation of bleomycin in rats is an established model of lung injury that includes an early inflammatory response (4–7 days) and subsequent fibrosis (7–28 days), after which slow resolution occurs (37). Inflammation after bleomycin injury consists of macrophage accumulation that occurs in association with maximal expression of transforming growth factor (TGF)-β1 and tumor necrosis factor (TNF)-α at 7 days (39, 40). Pressure-volume studies have indicated surfactant dysfunction during the early inflammatory phase after bleomycin-induced lung injury, whereas restriction of lung volumes during the latter phase of this disease results from a loss of tissue elasticity as a result of fibrosis (44, 52). Importantly, TGF-β1 downregulates the expression of SP-A, -B, and -C and phospholipid synthesis in human fetal lung explants (9), and TNF-α decreases the expression of SP-B and SP-C but not SP-A when administered to mice in vivo (46).
We therefore hypothesized that the respiratory distress and surfactant dysfunction after bleomycin injury would be associated with a deficit of critical surfactant components. There are conflicting data regarding the changes in phospholipids and SP-A after bleomycin injury, and there are no reports of SP-B and SP-C expression in rats during the early phase of injury. Two murine studies (16, 17) and one rat study (19) found increased SP-A, -B, and -C during the late recovery phase. In the current report, we examined all surfactant components and surface activity after bleomycin injury, focusing specifically on the inflammatory phase of the disease when surfactant dysfunction is most evident. A preliminary report of these data has been published previously (49).
MATERIALS AND METHODS
Model of Bleomycin Injury
Six-week-old (200–250 g) male Sprague-Dawley rat littermates (Charles River Breeding Laboratories, N. Wilmington, MA) were housed in the Animal Care Facility of The Children's Hospital of Philadelphia under standard conditions with free access to food and water. All animal experimental protocols were reviewed and approved by the Animal Care and Use Committees of both the Children's Hospital of Philadelphia and The University of Pennsylvania. Anesthesia consisted of ketamine-xylazine-atropine (16:8:0.01 mg/kg) injected intraperitoneally. Under sterile conditions, the trachea was visualized through a vertical incision in the neck. With the use of an insulin syringe, 250 μl of either human clinical-grade, sterile, and lipopolysaccharide-negative saline or 8.0 U/kg of bleomycin sulfate (Bristol Myers Squibb, Princeton, NJ) in 250 μl of saline were injected into the trachea. The incision was closed with surgical clips, and the animals were allowed to recover. Mortality in experimental animals was ∼5%, occurring only at the time of administration of bleomycin.
Measurement of Respiratory Rates
Respiratory rates were determined before institution of anesthesia and at intervals after intratracheal treatments under conditions that did not startle or arouse the rats. Two independent observers counted the respiratory rate of each animal for at least three separate 15-s intervals, with a minimum of five animals examined for each condition at each time point.
Processing of Bronchoalveolar Lavage Fluid and Lung Tissue
Animals were studied 7 days after intratracheal treatments. Bronchoalveolar lavage (BAL) with saline to a total of 36 ml/kg was performed on each animal. Total fluid recovery was consistently 80–90% in all animals. The BAL fluid was immediately centrifuged at 500 g for 10 min to remove all cells and cellular debris. An aliquot of cell-free BAL fluid was taken, and the large-aggregate surfactant-rich fraction was isolated by centrifugation (27,000g for 60 min). The pellet was resuspended in 154 mmol/l NaCl, 1.5 mmol/l CaCl2, and 10 mmol/l Tris-Cl (pH 7.4) and washed by centrifugation.
The pulmonary artery was perfused with PBS to remove all blood from the lungs. The heart, mediastinal structures, and trachea were dissected free from the lungs. The left lung was inflated to a pressure of 25 cmH2O with 1% paraformaldehyde and was placed in 10% neutral formalin for fixation before being processed for paraffin sectioning or in 1% paraformaldehyde for frozen sectioning. The three lobes of the right lung, dissected separately, were immediately frozen at −70°C for further analysis.
Determination of Phospholipid Content
The phospholipid content and composition of the surfactant-enriched BAL fluid was determined in 12–15 animals for each condition. Lipids were extracted from aliquots of the BAL fluid by the method of Bligh and Dyer (11). Individual phospholipids (PC, PI, and PG) were separated by TLC using activated silica plates (Whatman, Clifton, NJ) in a chloroform-methanol-petroleum ether-acetic acid-boric acid solution (80:40:60:20:3.6 vol/vol/vol/vol/wt), as described by Gilfillan et al. (24). Phospholipid phosphorus was determined by spectrophotometric determination using the method of Wells and Dittmer (58).
Determination of Surface Activity
The concentration of phospholipids in the sedimentable large-aggregate fraction was adjusted to 1.0 mg/ml. In preliminary studies, this concentration of phospholipid was consistently associated with low minimum surface tension in normal rat lavage samples (data not shown). Surface tension determinations were made at 37°C in humidified air on a pulsating bubble surfactometer (Electronetics, Buffalo, NY). The bubble radius of 0.40 mm was maintained for 15 s. Thereafter, the radius was varied between 0.4 and 0.55 mm at a frequency of 0.33 Hz for 5 min. The pulsating bubble surfactometer recorded data every 0.05 s, which were stored in Excel (Microsoft, Seattle, WA). Minimum and maximum surface tensions and hysteresis curves were determined for each sample.
In experiments to determine the contribution of serum proteins and/or SP-B deficiency to the loss of surface activity of BAL fluid after bleomycin injury, lipid extracts were obtained using the method of Bligh and Dyer (11), dried under nitrogen, and resuspended in chloroform. The phospholipid content was adjusted to 1 mg/ml and, in one-half of the sample, purified bovine SP-B in a chloroform-methanol solution (the kind gift of Drs. Karina Rodriquez and Fred Possmayer, University of Western Ontario, London, Ontario, Canada) was added at a final concentration of 1%. Both the reconstituted and original samples were then sonicated, dried, and resuspended, and surface tension activity was determined by the bubble surfactometer as described above.
Polyclonal SP Antisera
Polyclonal SP-A and proSP-C antisera produced in rabbits have been described previously. 1) The epitope-specific antiserum anti-NPRO-SP-C recognizes a region of the rat pro-SP-C molecule near the amino terminus (Met10 to Glu23) (10). Anti-NPRO-SP-C does not recognize mature SP-C and does not cross-react with SP-A, SP-B, proSP-B, or rat serum proteins (8). 2) Polyclonal anti-SP-A antiserum (PA3) was produced in rabbits by injection of purified rat SP-A as previously described (56). PA3 recognizes the rat, human, and bovine forms of SP-A and does not cross-react with serum proteins or other SPs. 3) Polyclonal, monospecific anti-rat SP-B antiserum was the kind gift of Drs. Mora and Ingenito (Harvard Medical School, Boston, MA) and was prepared in a manner identical to that previously described for a polyclonal anti-bovine SP-B antisera (7). By Western blotting, this antiserum recognizes bovine, human, and rat SP-B and does not cross-react with SP-A, SP-C, rat serum proteins, or BSA. 4) The SP-D antibody, a rabbit polyclonal antibody raised against purified rat SP-D as described previously (45), was the kind gift of Dr. E. Crouch (Washington University School of Medicine, St. Louis, MO).
3C9 monoclonal antibody, the kind gift of Dr. Henry Shuman (Institute for Environmental Medicine, University of Pennsylvania, Philadelphia, PA), was raised by immunizing mice with the limiting membrane of alveolar type II cell lamellar bodies (63). The antibody recognizes a 180-kDa protein that localizes to multivesicular bodies, lamellar bodies, and regions of the plasma membrane in these cells (50).
This monoclonal antibody, raised against rat spleen cells, recognizes a single-chain glycoprotein of 90–100 kDa expressed predominantly on the lysosomal membranes of myeloid cells (Serotec, Oxford, UK) and is used to identify rat macrophages (20).
Indirect immunofluorescence for SPs was performed on 5-μm paraffin-embedded sections from at least three animals per condition with at least three sections examined per animal. Sections were first treated to remove paraffin and rehydrated. Nonspecific sites were blocked with PBS-10% goat serum-0.02% azide for 30 min at room temperature, and sections were incubated with primary antibody overnight at 4°C. The dilutions used for primary antibodies are as follows: SP-A 1:500, SP-B 1:500, proSP-C 1:300, and SP-D 1:500. Normal rabbit IgG (1:100 dilution) was used as a negative control. The next day, after washes, the sections were treated with 1 M glycine in PBS for 1 h at room temperature to block autofluorescence. A Cy3-conjugated goat anti-rabbit secondary antibody was applied at a dilution of 1:300 for 2 h in the dark at room temperature. Sections were then washed, air-dried, and mounted with Fluoromount-G (Southern Biotechnology Associates, Birmingham, AL). Fluorescence was viewed with epifluorescence at 510–560 nm on a Nikon TE300 inverted microscope with appropriate ultraviolet (UV) filters. Images were captured using a Hammamatsu digital camera using Metamorph Software (Universal Imaging, West Chester, PA). Confocal-microscopic images were obtained using a computer-interfaced, laser-scanning microscope (Leica TCS 4D) in the Confocal Core Facility at the Children's Hospital of Philadelphia. Simultaneous wavelength scanning allowed superimposition of FITC-labeled ED1 antibody (green) and SP or 3C9 antibodies (red) recognized with goat anti-rabbit Texas Red at wavelengths of 488 and 568 nm, respectively. Laser power was fixed at 75% for all image acquisition. Image output was at 1,024 × 1,024 pixels.
Because lavage samples were obtained under standard conditions with consistent volume recovery between the various treatment groups, immunoblotting for SPs was performed using an equal volume (20 μl) of BAL fluid per lane. Samples were loaded on a 4–12% precast NuPAGE Bis-Tris polyacrylamide gel (Invitrogen, Carlsbad, CA). Proteins were electrophoresed in either MES-SDS running buffer for SP-B or in MOPS-SDS buffer for SP-A and SP-D under reducing conditions at 200 volts for 35–45 min per the manufacturer's instructions (Invitrogen). Proteins were then transferred from the gel to a nitrocellulose membrane using the XCell II Mini-Cell and sandwich blot module (Invitrogen) in Bicine-10% methanol-0.01% SDS transfer buffer (Invitrogen) at 30 volts for 1 h. Blots were blocked for 1 h at room temperature with 5% nonfat milk and then incubated with primary antibody overnight at 4°C, except for SP-B for which the blots were blocked overnight and incubated with primary antibody for 1 h at room temperature. Primary antibody dilutions used were SP-A 1:30,000, SP-B 1:5,000, and SP-D 1:2,000. Blots were then incubated with 1:5,000 goat anti-rabbit IgG-horseradish peroxidase for 1 h at room temperature. Signal was detected using the enhanced chemiluminescence kit (Amersham, Arlington Heights, IL), and blots were exposed to Kodak Biomax MS film. Blots were developed for varying lengths of time, and relative changes in protein content were determined by first scanning the blots using an Agfa Arcus II scanner and FotoLook SA scanning software into a Macintosh G3 Power PC computer. Semiquantitative densitometric analysis of bands was accomplished using MacBAS version 4.2 (Fujifilm) after subtraction of background density. Results are expressed as a percentage of control values and were calculated as the means ± SE of four animals per condition.
Total RNA was obtained from snap-frozen tissue maintained on ice during isolation. Tissue (∼250 mg wet wt) was mechanically homogenized in RNA Stat60 (Tel-Test, Friendswood, TX). RNA integrity was confirmed using 1% agarose gels. RNA samples were denatured for 15 min at 65°C in 1× MOPS (pH 7.0), 6.5% formaldehyde, and 50% formamide and rapidly chilled on ice. Total RNA (20 μg) from each sample was separated in 1% agarose gels. After electrophoresis, separated RNA was then transferred to Duralose UV nitrocellulose membranes (Stratagene) overnight with 1× high-efficiency transfer solution (Tel-Test). Membranes were baked for 1 h at 65°C, and RNA was fixed on the membrane by UV cross-linking.
The membrane was prehybridized for 2 h at 65°C in hybridization solution [0.5 M sodium phosphate, pH 7.5, 7% SDS, 1 mM EDTA, 1% BSA, 50 μg/ml poly(A)+ RNA, and 50 μg/ml of denatured and sheared salmon sperm DNA]. cDNA probes were labeled by random priming using the Ready-To-Go Kit (Pharmacia-Upjohn) per the manufacturer's instructions and were purified with a G-50 column. The 28S oligonucleotide probe was 5′-end labeled using a 5′-end-labeling protocol (35–50 ng of 28S oligonucleotide, 2 μl of T4 polynucleotide kinase, and 50 μCi of [γ-32P]ATP in 1× kinase buffer) at 37°C for 1 h per the manufacturer's instructions (Promega, Madison, WI). The probe was purified with a G-25 column (Boehringer Mannheim, Indianapolis, IN). Hybridization of membranes with 32P-labeled probes (1 × 106counts · min−1 · ml−1) was performed for 16–18 h at 65°C. The blots were then washed under high-stringency conditions (2 volumes, 500 ml each, of 0.2× saline-sodium citrate-0.1% SDS at 65°C each) and were developed using a PhosphorImager (Storm 840; Molecular Dynamics, Sunnyvale, CA). Semiquantitative densitometric analysis of bands was accomplished on a Macintosh G3 Power PC computer using MacBAS version 4.2 (Fujifilm) after subtraction of background density. Results were calculated as the degree of change from control values. The results of at least five animals per condition and time point are expressed as means ± SE and percentage of control.
cDNA and Oligonucleotide Probes
The cDNA probes used in Northern analyses were generated byEcoRI digestion of prokaryotic expression vectors, liberating the following full-length inserts: 1) rat SP-A cDNA insert in pGEM4Z (22), used as previously published (56); 2) rat SP-B cDNA in pGEM4Z vector (21); 3) rat SP-C cDNA in pGEM4Z backbone, used as previously published (23); and4) rat SP-D cDNA in BlueScript SKII plasmid, the kind gift of Dr. J. Fisher (University of Colorado, Denver, CO) as described previously (51). The oligonucleotide 5′-AACGATCAGAGTAGTGGTATTTCACC-3′ was used to hybridize to 28S RNA as a loading control. This sequence has been shown previously to be specific for 28S (4).
Group mean data were analyzed using Excel (Microsoft) and Statistica (StatSoft, Tulsa, OK) and are expressed as means ± SE. Comparisons of multiple groups using ANOVA were made with either Fisher's protected least significant difference or the Student-Newman-Keuls tests of normality. The level of statistical significance was P < 0.05.
Bleomycin Injury Results in Surfactant Dysfunction
To define the effects of intratracheal instillation of bleomycin on pulmonary function, we determined the changes in respiratory rates of animals over the first 21 days after injury. Untreated animals and those given intratracheal saline served as controls. Increased respiratory rates were noted by 2 days after bleomycin injury and were maximal at 7 days (172 ± 0.1% of control values; Fig.1 A). At that time, bleomycin-injured animals had retractions and cyanosis. Respiratory rates decreased thereafter and were normal by 21 days after injury. The respiratory rates of control and saline-treated animals were similar and did not change during the 21-day protocol.
To confirm surfactant dysfunction, we determined the surface activity of surfactant-enriched fractions of BAL fluid using a pulsating bubble surfactometer. The total phospholipid (TPL) content of each sample was adjusted to 1 mg/ml. Cycling of the artificial bubble for 5 min was required to obtain stable surface tension measurements (data not shown). Lavage fluid from bleomycin-injured animals had significantly higher minimum surface tension compared with saline-treated controls (20 ± 0.4 vs. 1.5 ± 0.5 mN/m,n = 5, P < 0.05, ANOVA; Fig.1 B). Similar results were obtained using uninjured controls (1.6 ± 1.6 mN/m, n = 5, P < 0.05). Collectively, these data show that symptomatic respiratory distress in bleomycin-treated rats was associated with abnormal surface tension, indicative of surfactant dysfunction at 7 days after injury.
Phospholipid Content and Composition
We determined the TPL, PC, PG, and PI contents of a surfactant-enriched fraction isolated from BAL fluid. At 7 days, there were no statistically significant differences in the contents of TPL, PC, PG, or PI in any of the groups examined (Fig.2), suggesting that surfactant dysfunction was not the result of reduced surfactant phospholipid. The sum of PC, PG, and PI contents was consistently lower than the TPL content, since other phospholipids such as phosphatidylethanolamine, phosphatidylserine, and sphingomyelin were not determined in these studies.
Immunoblot analysis of SPs in BAL fluid from control and bleomycin-injured animals is shown in Fig.3 A. By densitometry, mature 8-kDa SP-B was decreased by 90% compared with that in saline-treated animals (Fig. 3 B). In contrast, both SP-A (2-fold) and SP-D (1.5-fold) were increased in the BAL fluid of injured animals relative to those in saline-treated controls (Fig. 3 B). Some degradation of SP-A was noted in BAL samples from bleomycin-injured animals (Fig. 3 A). Similar results were found in immunoblots of large-aggregate surfactant obtained by centrifugation. NPRO-SP-C antiserum failed to detect any bands in immunoblots because this antiserum does not recognize mature SP-C (data not shown).
We also examined the changes in the steady-state mRNA contents of SP-A, -B, -C, and -D in control and bleomycin-treated lungs 7 days after treatment (Fig. 4 A). After correction for loading variability and expression of data as a percentage of normal, uninjured control values, SP-A mRNA content was unchanged after bleomycin but slightly increased in saline-treated animals (P = 0.05, Fig. 4 B). SP-D mRNA content was slightly but significantly increased equally in both saline- and bleomycin-treated animals (P < 0.05, Fig.4 A). SP-B and SP-C mRNAs were similar in saline and control animals but were decreased by 50 and 60%, respectively, after bleomycin (P < 0.001, Fig. 4 B). Because SP-A and SP-B are expressed in both bronchiolar Clara and alveolar type II cells, we examined SP-C mRNA levels over time to determine type II cell-specific effects of bleomycin injury (Fig. 4 C). At all time points, saline-treated animals showed no differences from control animals (data not shown). The decrease in the steady-state content of SP-C mRNA was evident as early as 2 days after bleomycin and was maximally decreased at 4–7 days. SP-C mRNA levels were greater than control levels by 21 days, suggestive of type II cell hyperplasia known to occur in this model (Fig. 4 C). Changes in SP-B mRNA over time after injury were similar to those of SP-C mRNA (data not shown).
Lung histology was unaffected by saline instillation. However, 7 days after intratracheal bleomycin, there were large, patchy areas of consolidation and atelectasis with thickened alveolar walls, decreased alveolar septa and an alveolar cell infiltrate. In addition, patchy emphysematous areas were seen (Fig.5). Macrophage accumulation, as identified by positive staining with the ED1 antibody (Fig.6), was noted in the alveolar interstitium and airspaces of injured lungs. Using double-label confocal microscopy, all four SPs were colocalized with the ED1 marker, suggesting that these proteins were also found in macrophages responding to lung injury (data not shown, see Fig. 6). In addition, bleomycin injury resulted in decreased to absent staining of epithelial cells for both SP-B and proSP-C in areas of lung injury (Fig. 5). SP-D staining in normal and saline-treated lungs was noted in the epithelial cells of larger airways and in type II epithelial cells in alveoli. After bleomycin injury, SP-D staining continued in epithelial cells (Fig. 5).
To confirm the identity of cells as macrophages and the presence of type II cell-specific proteins within them, sections of lung double labeled for a lamellar body membrane-specific antigen (3C9 antibody) and the macrophage marker ED1 showed staining with 3C9 in type II cells in saline-treated lungs (Fig. 6). Interestingly, at the dilutions used in our experiments, ED1 did not recognize resident alveolar macrophages in control sections. In contrast, lung sections from bleomycin-injured animals showed increased 3C9 staining that colocalized with ED1 in macrophages responding to injury (Fig. 6).
We used electron microscopy to examine alveolar structure and lamellar body morphology after bleomycin injury (Fig.7). Control and saline-treated lungs had normal architecture consisting of close apposition of type I epithelial and endothelial cells with alveolar airspaces (Fig. 7,A and B). Type II alveolar epithelial cells had apical microvilli and well-formed lamellar bodies. After bleomycin injury, the alveolar interstitium was greatly thickened and collagen deposition was noted. Alveolar type II cells were identified by the presence of microvilli, but these cells lacked definable lamellar bodies (Fig. 7, C and D). In more distal areas of lung that were relatively spared of injury, type II cells had few to no lamellar bodies (Fig. 7 E). Activated macrophages, identified by their vacuolated cytoplasm and numerous lamellapodia, were found in the alveolar airspaces with lamellar bodies within their cytoplasm, suggesting the scavenging of type II epithelial cells and/or their contents (Fig. 7 F).
Reconstitution of Surface Activity of BAL Fluid by Exogenous SP-B
The loss of surface activity of BAL fluid after bleomycin injury could result from SP-B deficiency and/or the presence of inhibitory serum proteins. To eliminate the possible inhibitory influence of serum proteins contaminating BAL fluid and thus adversely affecting surface tension measurements, phospholipids were extracted from large-aggregate surfactant pellets from control animals and injured animals at 7 days. The minimum surface tension of the extracted samples from injured animals (as determined by the bubble surfactometer) remained high, suggesting that inhibitory serum proteins were not responsible for the loss of surfactant function (Fig. 8). Addition of exogenous SP-B (1%) to the phospholipid extract reconstituted a low minimum surface tension comparable to control samples, indicating that addition of SP-B alone was sufficient to restore surfactant function.
Respiratory distress is a consistent finding after bleomycin-induced lung injury and is associated with decreased lung volumes that result from surfactant dysfunction and/or fibrosis. In earlier studies, lungs expanded with air and saline to generate pressure-volume curves were examined at intervals after bleomycin injury (44, 52). Decreased compliance, consistent with surfactant dysfunction, was noted in the first 7–10 days after injury, whereas tissue forces were responsible for the loss of lung volume at later times. Our observations confirm these findings by showing that a surfactant-enriched fraction of BAL fluid from bleomycin-injured animals had decreased ability to lower surface tension in a bubble surfactometer compared with that in controls. In addition, we found that respiratory distress after bleomycin was temporally associated with a specific, transient downregulation of SP-B and SP-C gene expression. The 90% decrease in mature SP-B protein content of BAL fluid after injury was associated with a marked reduction of lamellar bodies in identifiable type II epithelial cells. The ability of exogenously added SP-B to restore surface activity to surfactant phospholipids from injured animals suggests that a deficiency of hydrophobic SPs is the critical determinant of respiratory distress in this model.
Conflicting and limited data exist in the literature as to the changes in phospholipid synthesis, content, and composition after bleomycin lung injury in rodents. Some authors have shown increased phospholipid content, in particular of disaturated PC (DSPC), with either a slight reduction or no change in PG (41, 52). Others, however, have shown either no change in DSPC and decreased PG (35) or decreased DSPC and PG (44). With the use of hamster lung slices, the [14C]acetate incorporation in TPL, PC, and neutral lipids was decreased in bleomycin-treated animals compared with saline controls 2 days after injury (25). However, this group reported increased incorporation of [14C]choline in PC in the same model (34). In our studies, the content and composition of surfactant phospholipids in large-aggregate surfactant was unaffected 7 days after bleomycin. Whether more profound changes in the steady-state content and composition of surfactant phospholipids occur at other time points after bleomycin is currently unknown, and the effect of injury on phospholipid synthesis and turnover remains to be clarified.
In vitro, both SP-B and SP-C promote stability of the phospholipid monolayer and reduce surface tension (15). Strong evidence exists for the critical role of SP-B in surface tension-reducing properties of surfactant in vivo. Mice with targeted disruption of the SP-B gene (14) and infants with congenital deficiency of SP-B (43) develop fatal respiratory distress in the newborn period. In addition, mice given a monoclonal antibody to SP-B also develop respiratory failure (48). The role of SP-C in surface activity in vivo remains uncertain. Because congenital SP-B deficiency is associated with abnormal processing of SP-C, there is in effect a combined deficiency of SP-B and -C (14, 43). Recently, Glasser et al. (28) reported that mice with targeted disruption of the SP-C gene are viable and do not develop respiratory distress. However, Belgian White and Blue calves with normal lung SP-B but markedly low SP-C expression develop respiratory distress syndrome at birth (18). It is likely, therefore, that a critical amount of total hydrophobic SP is required for surface activity.
Atochina et al. (2) and Beers et al. (6) have reported a similar selective decrease in SP-B and SP-C gene expression in association with loss of surface activity in a mouse model ofP. carinii pneumonia. In addition, Ingenito et al. (36) have shown selective SP-B deficiency in relation to surfactant dysfunction after pulmonary challenge with lipopolysaccharide. The similarities in these responses suggest that common mechanisms likely exist for selective decreases in SP-B and SP-C gene expression observed after varied insults to the lung.
In addition to decreased synthesis of SP-B and SP-C, several other mechanisms could contribute to the surfactant deficiency noted after injury. First, we have noted lamellar bodies and intense immunoreactivity for SPs in macrophages accumulating in the airspaces of bleomycin-injured animals. The most likely explanation of these findings is that damaged, sloughed type II epithelial cells are engulfed by alveolar macrophages. However, the contents of PC, SP-A, and SP-D in BAL fluid are unchanged by bleomycin and many intact type II cells remain, albeit lacking lamellar bodies. Second, lung injury is associated with the accumulation of a protein exudate that inactivates surfactant. However, elevated minimum surface tensions also occurred with phospholipid extracts of BAL surfactant that are devoid of most contaminating serum proteins. A third mechanism that could contribute to surfactant dysfunction is oxidative/nitrative modification of SPs, with loss of function. In vitro tyrosine nitration of SP-A by peroxynitrite alters lipid binding activity, and in vitro nitration of SP-B results in a loss of surface activity, possibly resulting from either degradation or aggregation (32). In these studies, the authors found that in vitro nitration of SP-B resulted in an inability to detect the protein by immunoblot. Our finding that SP-B mRNA is decreased by 50%, whereas SP-B protein is decreased by 90%, may be explained by protein modifications of SP-B that affect its detection on immunoblots. The degradation products noted for SP-A in immunoblots could also represent the effect of oxidative and nitrative stresses that occur after injury. These possibilities are currently under investigation.
There have been several studies of SP expression after bleomycin lung injury. Horiuchi et al. (35) found that although elastic recoil was decreased after bleomycin, there was no change in the expression of SP-A at any time point studied. Furthermore, Kasper et al. (38) could not correlate SP-A localization with the effects of bleomycin injury in rats. Increased SP-A content has also been reported after several forms of acute lung injury, including intratracheal bleomycin (55). Interestingly, Daly et al. (16, 17) noted increased expression of SP-A, -B, and -C mRNAs in epithelial cells 30 days after bleomycin injury of mice. Deterding et al. (19) also showed similar increases in SP mRNAs in the later phase of bleomycin injury in rats. However, type II cell hyperplasia is known to occur at this time after injury, and these authors did not examine the expression of SPs during the initial phase after lung injury when significant surfactant dysfunction occurs.
Several cytokines have been implicated as regulators of differentiation and function of alveolar type II cells (12). Maximum expression of inflammatory growth factors after bleomycin injury occurs in conjunction with the accumulation of macrophages in injured areas of the lung. Thus TGF-β1 and TNF-α expression is maximal 7 days after injury, corresponding to the time of maximal surfactant dysfunction. Relevant to the current study, TGF-β1 has been shown to inhibit phospholipid production by fetal rat type II cells in culture (53) and the expression of surfactant lipids and/or SPs in both human alveolar epithelial cell lines (61) and fetal lung explants (9). Overexpression of TGF-β1 in alveolar epithelial cells of transgenic mice decreased proSP-C expression (64), and these findings have been confirmed in embryonic cultures in vitro (13). Administration of anti-TGF-β1/2 antibodies (27) or decorin (26) to mice decreases collagen accumulation after intratracheal bleomycin. However, the effects of these treatments on respiratory distress, surfactant dysfunction, or SP-B gene expression are unknown. Collectively, these data suggest that TGF-β1 downregulation of SP-B and SP-C may contribute to the respiratory distress observed in the 1st wk after bleomycin injury.
TNF-α has been shown to decrease PC synthesis and content in vitro (1, 33) and to inhibit SP expression both in vitro and in vivo (3, 46, 47, 59, 62). Effects on SP-A may involve both gene transcription (62) and message stability (47). TNF-α decreases SP-C expression in cultured epithelial cells and after intratracheal administration to mice, at least in part, by affecting transcription of this gene (3). Furthermore, stimulation of endogenous TNF-α or intratracheal administration of this cytokine in mice results in downregulation of SP-B and SP-C but not of SP-A (46). It is likely that increased expression of other growth factors and cytokines after bleomycin injury may also contribute to the decreased expression of SP-B and SP-C and respiratory distress seen in these animals.
In summary, this report highlights the importance of the inflammatory response as a key early event in the respiratory distress noted after acute lung injury. This response involves increased expression of key growth factors and cytokines, such as TGF-β1 and TNF-α, that may mediate both surfactant deficiency as a result of decreased SP gene expression and increased collagen synthesis leading to fibrosis. Interventions to decrease the inflammatory process would therefore be predicted to limit both surfactant dysfunction and fibrosis after acute pulmonary damage. Key processes to the target may be the activation and mobilization of peripheral blood monocytes that subsequently become tissue macrophages, cells that are largely responsible for the increased expression of growth factors and cytokines observed after acute lung injury.
We are grateful to the following for the gifts of reagents used in this study: Drs. Karina Rodriquez and Fred Possmayer (Depts. of Obstetrics and Gynecology and Biochemistry, University of Western Ontario, London, Ontario, Canada) for bovine surfactant protein (SP) B, Drs. R. Mora and E. P. Ingenito (Harvard Medical School, Boston, MA) for anti-rat SP-B antibody, Dr. E. Crouch (Washington University, St. Louis, MO) for anti-rat SP-D antibody, Dr. Henry Shuman (University of Pennsylvania, Philadelphia, PA) for 3C9 antibody, and Dr. J. Fisher (University of Colorado, Denver, CO) for SP-D cDNA. We thank Gopi Mohan for technical assistance.
The research presented in this report was funded by grants from the Pennsylvania Thoracic Society/American Lung Association and the March of Dimes Foundation and by National Heart, Lung, and Blood Institute Grant HL-62472 (R. C. Savani), Grant HL-59867 (M. F. Beers), and Specialized Center of Research in Pathobiology of Lung Development and Bronchopulmonary Dysplasia Grant HL-56401 (L. W. Gonzales, S. H. Guttentag, M. F. Beers, and P. L. Ballard). In addition, we acknowledge funds from the Endowed Chair in Critical Care for purchase of the bubble surfactometer and from the Gisela and Dennis Alter Endowed Chair in Pediatrics (L. W. Gonzales, S. H. Guttenberg, and P. L. Ballard).
Address for reprint requests and other correspondence: R. C. Savani, Div. of Neonatology, Rm. 416, Abramson Research Center, Children's Hospital of Philadelphia, 34th St. and Civic Center Blvd., Philadelphia, PA 19104-4399 (E-mail:).
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 © 2001 the American Physiological Society