|Home | About | Journals | Submit | Contact Us | Français|
Surfactant dysfunction was studied in C57BL/6 (B6), B6.SP-A−/−, and B6.iNOS−/− mice with pulmonary mycoplasma infection (107 colony-forming units). Cell-free bronchoalveolar lavage (BAL) from uninfected B6.SP-A−/− versus B6 mice had a reduced content of very large aggregates (VLA) and an increase in intermediate large aggregates (ILA), with no difference in total large aggregates (LA = VLA + ILA). However, LA from uninfected B6.SP-A−/− versus B6 mice contained less protein and were more sensitive to inhibition by serum albumin and lysophosphatidylcholine in pulsating bubble studies in vitro. Infection with Mycoplasma pulmonis caused significant lung injury and surfactant abnormalities in B6.SP-A−/−, B6.iNOS−/−, and B6 mice at 24, 48, 72 h after infection compared with uninfected mice of the same strain. Analyses of time-pooled data indicated that mycoplasma-infected B6.SP-A−/− and B6.iNOS−/− mice had significantly lower levels of LA and higher protein/phospholipid ratios in BAL compared with infected B6 mice. Infected B6.iNOS−/− versus B6 mice also had increased minimum surface tensions on the pulsating bubble and decreased levels of surfactant protein (SP)-B in BAL. These results indicate that pulmonary mycoplasma infection in vivo causes lung injury and surfactant abnormalities that are dependent in part on iNOS and SP-A. In addition, SP-A deficiency modifies surfactant aggregate content and lowers the inhibition resistance of LA surfactant in vitro compared with congenic normal mice.
Mycoplasmas account for a large fraction of pneumonias, and may cause severe systemic problems, such as arthritis. In this article, we show that mycoplasmas damage the pulmonary surfactant, an essential component of normal lung function.
Mycoplasma pneumoniae (mycoplasmas) accounts for a significant fraction (10–20%) of community-acquired pneumonias in the general population, as well as being a frequent cause of tracheobronchitis and other upper respiratory symptoms (1). Mycoplasma-induced pneumonia was once thought to be relatively uncommon in children under five and in older adults, but is now recognized as affecting a broad spectrum of patient age groups (1). In addition to causing a primary pneumonia, mycoplasmas are known to exacerbate the pathophysiology of asthma and chronic obstructive pulmonary disease (COPD) (2). Less common, but severe associated conditions precipitated by mycoplasma infection include acute lung injury (ALI), acute respiratory distress syndrome (ARDS), pericarditis, myocarditis, hemolytic anemia, and encephalitis (3).
Infection of C57BL/6 (B6) mice with Mycoplasma pulmonis reproduces the essential features of human respiratory mycoplasmosis. Alveolar macrophage (AM) activation has been demonstrated to be essential for the killing of M. pulmonis in vitro (4). Both surfactant protein (SP)-A and inducible nitric oxide synthase (iNOS) are thought to be important in pulmonary defenses against respiratory pathogens such as mycoplasmas. The present paper investigates lung surfactant abnormalities in mice with respiratory mycoplasma infection, and examines the specific importance of SP-A and iNOS in mycoplasma-induced surfactant dysfunction through the study of B6.SP-A−/− and B6.iNOS−/− mice. In addition, experiments also assess the specific aggregate composition, surface activity, and inhibition resistance of surfactant from uninfected B6.SP-A−/− mice compared with uninfected B6 mice.
Lung SP-A plays multiple roles in pulmonary biology, and related protein forms are present in all vertebrate classes, including most primitive amphibious fish (5). SP-A has calcium-dependent activity in increasing the aggregation and molecular order of phospholipids, including promoting tubular myelin formation in lung surfactant in conjunction with SP-B (6, 7). SP-A also enhances the ability of lung surfactant extracts to resist biophysical inhibition by plasma proteins in vitro (7–10). As a member of the collectin family of host defense proteins, SP-A also contributes to innate pulmonary immunity (11–13). We have previously studied M. pulmonis infection in B6 mice and showed that SP-A is necessary for maximal mycoplasma killing (4, 14, 15) by AM by upregulating the production of NO and reactive oxygen-nitrogen intermediates in these cells (14, 15). Both SP-A−/− and iNOS−/− mice have a decreased ability to clear intratracheally instilled mycoplasmas. However, SP-A and iNOS-dependent abnormalities in pulmonary surfactant have not been investigated in detail in mycoplasma infection as is done here using congenic B6.SP-A−/− and B6.iNOS−/− as well as B6 mice. Studies in mycoplasma-infected mice test the hypothesis that a deficiency in SP-A or iNOS leads to a significant increase in the severity of mycoplasma-induced surfactant dysfunction in the respective knockout models compared with B6 mice. An additional hypothesis tested was that surfactant from uninfected B6.SP-A−/− mice will have an altered distribution of VLA and ILA compared with uninfected B6 mice, and will be more sensitive to biophysical inhibition by lysophosphatidylcholine (LPC) and albumin in vitro.
C57BL/6NCr (B6) mice were obtained from the National Cancer Institute (Frederick, MD). C57BL/6 SP-A−/− (B6.SP-A−/−) N10 mice were generated from 129 × BS SP-A−/− mice provided by Drs. Whitsett and Korfhagen (University of Cincinnati, Ohio) (16) by backcrossing for at least 10 generations onto the B6 background (15). Mice were bred at the University of Alabama at Birmingham (UAB), and genotype was characterized from tail DNA as described previously (15). C57BL (C57BL/6J-Nos2tm1Lau) transgenic mice lacking the gene for iNOS (B6.iNOS−/−) mice were obtained from the Jackson Laboratory (Bar Harbor, ME) (14) and were bred at UAB in autoclaved microisolator cages (Lab Products, Maywood, NJ). Mice of both sexes were used, and animals were maintained in a sterile environment and provided with autoclaved food (Agway, Syracuse, NY) and water ad libitum until being studied at 8–12 wk of age (20–25 g body weight). All mice were monitored and found to be negative for murine pathogens by the Health Surveillance Facility at UAB (17).
Animal experiments were performed according to protocols and guidelines approved by the Institutional Animal Care and Use Committee guidelines at the University of Alabama at Birmingham (UAB). For mycoplasma infection, mice were inoculated intranasally with the UAB CT strain of M. pulmonis (18), while control mice received an equal 50-μl intranasal volume of sterile broth. An inoculate dose of 107 colony-forming units (cfu) was confirmed by serial dilution and plating on agar for 7 d (4). The dose of 107 cfu was chosen because exposure to lower amounts of mycoplasma did not result in a reproducible level of injury to the pulmonary surfactant system in B6 mice. All mouse strains were studied at 24, 48, and 72 h after inoculation with M. pulmonis.
After being killed at a designated time point, mice were lavaged with a total of 3 ml of 0.15 M NaCl, given in three equally divided 1-ml aliquots. Recovered aliquots of lavage fluid were pooled on ice, centrifuged immediately at 150 × g for 10 min to remove cells, and the cell-free supernatant was then frozen at −20°C for biochemical and biophysical analyses.
In the majority of studies, cell-free bronchoalveolar lavage (BAL) was centrifuged at 12,500 × g for 30 min to sediment the total large aggregate (LA) fraction. In the subset of experiments investigating aggregate subfractions, cell-free BAL was initially centrifuged at 1,500 × g for 10 min to pellet very large aggregates (VLA) and the supernatant was then re-centrifuged at 12,500 × g for 30 min to pellet intermediate large aggregates (ILA). The fraction of smaller aggregates remaining in the supernatant after centrifugation at 12,500 × g for 30 min was also examined in terms of its content relative to LA, VLA, and ILA.
Total phospholipid in cell-free BAL and aggregate fractions was measured by the assay of Ames (19), and total protein was determined using the colorimetric assay of Lowry and coworkers (20) modified by the addition of 15% SDS to allow accurate quantitation of protein in the presence of lipid. Total hydrophobic surfactant protein was measured by the modified Lowry assay following the extraction of cell-free BAL or centrifuged aggregates into chloroform:methanol by the method of Bligh and Dyer (21). The composition of phospholipid classes in cell-free BAL and LA was determined on the basis of phosphate measurements after extraction and one-dimensional thin-layer chromatography (TLC) on 250-μm thick silica gel G (Analtech, Newark, DE) using a solvent system of chloroform:methanol:2-propanol:triethylamine:water (30:9:25:25:7 by volume) (22).
The overall surface tension lowering ability of surfactant dispersions was measured during cycling at a physiologic rate of 20 cycles/min at 37 ± 0.5°C on a pulsating bubble surfactometer (General Transco, Largo, FL) (7, 23). Cell-free BAL or centrifuged aggregates were evaporated under nitrogen and then resuspended with hand vortexing in 0.15 M NaCl plus 2 mM CaCl2 at the desired phospholipid concentration (1.0 or 2.0 mg/ml). When present, inhibitors were added to resuspended cell-free BAL or surfactant aggregates and incubated for 30 min at 37°C before surface activity measurements. Substances studied for inhibitory effects on the activity of surfactant from uninfected B6.SP-A−/− versus B6 mice were bovine serum albumin (BSA) (Fraction V, 3 mg/ml; Sigma Chemical Company, St. Louis, MO) or C18:1 lysophosphatidylcholine (LPC) (0.125 mg/ml; Sigma Chemical). Aliquots of surfactant or surfactant/inhibitor mixtures were introduced into a 40-μl plastic sample chamber mounted on the pulsator unit of the surfactometer, and an air bubble (communicating with ambient air) was formed and repetitively oscillated between maximum and minimum radii of 0.55 and 0.4 mm at a rate of 20 cycles of compression–expansion per min. The pressure drop across the air–water interface of the bubble was measured by a precision pressure transducer, and surface tension at minimum bubble radius (minimum surface tension) was calculated as a function of time of pulsation using the Laplace equation for a spherical interface (7, 24).
In experiments testing the direct effects of M. pulmonis on surface activity, BAL from uninfected mice of a given strain was exposed to 2 × 107 cfu/ml of mycoplasma organisms in vitro for 48 h at room temperature. BAL was then centrifuged at 12,500 × g for 30 min to pellet LA, and minimum surface tension was measured on the pulsating bubble apparatus (37°C, 20 cycles/min, 50% area compression).
SP-B levels were specifically examined in BAL for each mouse strain at 48 h after inoculation with mycoplasmas (107 cfu/mouse). SP-B in cell-free BAL was detected as described by Griese and colleagues (25). For SP-B detection, the protein content each sample was determined with the BioRad Protein Assay Kit (BioRad, Richmond, CA) based on the method of Bradford (26). Five micrograms of total protein were separated under nonreducing conditions on 12% Bis-Tris gels using a Criterion Cell and Blotter System (BioRad, Carlsbad, CA). Two gels were prepared in parallel for each sample, and after electrophoresis one was stained with SYPRO Ruby protein gel stain (Invitrogen, Carlsbad, CA) and the other subjected to Western blotting. For immunodetection, the proteins in the gels were transferred onto a PVDF membrane with a Criterion Blot module in a wet buffer system (all from BioRad). Overnight blotting was with the primary rabbit polyclonal anti–SP-B antibody (1:1,000 dilution; Chemicon International, Inc., Temecula, CA) and horseradish peroxidase–conjugated goat anti-rabbit polyclonal anti-IgG (1:10,000 dilution; Amersham Biosciences, Piscataway, NJ). We used an ECL Detection system to detect SP-B, which appeared as a band at 18–20 kD.
All data are expressed as mean ± 1 SEM. The pooled variance t test was used to assess statistical significance for statistical analysis of discrete data on biochemical composition and aggregate content. Quantitative measures (surface tension, phospholipids levels, etc.) data were analyzed between mouse strains by one-way ANOVA with Tukey's HSD (Honest Significant Difference) post hoc procedure to identify points of significant difference. In some cases, data were analyzed by ANOVA followed by Tukey's multi-group comparison of the means after log conversion or by Kruskal Wallis ANOVA and Pearson's correlation of the means for nonparametric data (Analytical Software, St. Paul, MN). In all statistical analyses, differences were considered significant if the P value was < 0.05.
In an initial set of studies, cell-free BAL obtained from groups of 10–15 uninfected B6.SP-A−/− mice was pooled and assessed for aggregate composition compared to pooled lavage from uninfected B6 mice. Uninfected B6.SP-A−/− had a reduced percentage of BAL phospholipid in the VLA fraction sedimenting at 1,500 × g for 10 min as compared with B6 mice (31 ± 6% versus 52 ± 6% of total BAL phospholipid, P < 0.05) (Figure 1). However, this decrease in VLA in uninfected B6.SP-A−/− mice was offset by an increase in ILA obtained by centrifuging the supernatant of the 1,500 × g spin at 12,500 (i.e., aggregates sedimenting between 1,500 × g and 12,500 × g). B6.SP-A−/− mice had 40 ± 6% of total BAL phospholipid in the ILA pellet compared to only 22 ± 3% for B6 mice (Figure 1; P < 0.01). Due to the offsetting changes in VLA and ILA, there was no difference between uninfected B6.SP-A−/− and B6 mice in the content of total large surfactant aggregates (LA = VLA + ILA) sedimenting from BAL at 12,500 × g. Consistent with this, uninfected B6.SP-A−/− and B6 mice also had no difference in smaller aggregates remaining in the supernatant after centrifugation at 12,500 × g (aggregates sedimenting at >12,500 × g, Figure 1).
Total phospholipid in cell-free BAL was equivalent for uninfected B6.SP-A−/− and B6 mice (0.18 ± 0.01 mg/ml and 0.17 ± 0.02 mg/ml, respectively). In addition, the phospholipid class distributions and total protein levels in whole cell-free BAL did not differ between these mice (Tables 1 and and2).2). However, centrifuged LA from uninfected B6.SP-A−/− mice had a significantly lower total protein content compared with uninfected B6 mice (4.9 ± 1.7 versus 9.3 ± 1.4% by weight relative to phospholipid, P < 0.05 Table 2). The lower protein content of LA surfactant from uninfected B6.SP-A−/− versus B6 mice in Table 2 is consistent with the lack of SP-A in the knockout animals. LA from uninfected B6.SP-A−/− and B6 mice had equal levels of chloroform-extracted hydrophobic protein indicative of SP-B/C (Table 2). Additional analyses not included in Table 2 showed that VLA and ILA from uninfected B6.SP-A−/− also had a reduced total protein content as compared with B6 mice (6.4 ± 1.0 versus 8.2 ± 0.6% by weight for VLA and 4.7 ± 1.1 versus 7.5 ± 0.6% by weight for ILA, respectively; P < 0.05). These findings of reduced protein in VLA and ILA from uninfected B6.SP-A−/− versus B6 mice are again consistent with the lack of SP-A in the knockout animals.
The surface activity of each form of surfactant studied (VLA, ILA, LA, or whole BAL) for uninfected B6.SP-A−/− and B6 mice increased as surfactant phospholipid concentration increased from 1 mg/ml to 2 mg/ml (Figure 2). However, none of the surface tension lowering curves for comparable forms of surfactant (VLA, LA, ILA, or whole BAL) differed significantly between uninfected B6.SP-A−/− and B6 mice by ANOVA at either phospholipid concentration (Figures 2A–2D). At a fixed phospholipid concentration, the relative surface activity of the different forms of surfactant studied from both uninfected B6.SP-A−/− and B6 mice was ordered as: VLA > LA > ILA > whole BAL (Figure 2). This behavior is consistent with the known concentration-dependent activity of lung surfactants and the fact that larger aggregate forms of surfactant exhibit the greatest surface activity (see Refs. 7, 27, and 28 for review).
Resuspended BAL and LA from uninfected B6.SP-A−/− versus B6 mice had less ability to resist inhibition when incubated with either BSA (3 mg/ml; Figure 3) or C18:1 LPC (0.125 mg/ml; Figure 4). For both types of mice, the greatest inhibitor-induced reductions in surface activity were found at a low surfactant concentration of 1 mg phospholipid/ml (Figures 3A and and4A),4A), and inhibition was decreased at a higher concentration of 2 mg phospholipid per milliliter (Figures 3B and and4B).4B). However, at either phospholipid concentration, whole BAL and LA from uninfected B6.SP-A−/− versus B6 mice had elevated minimum surface tensions in the presence of BSA or LPC (Figures 3 and and44).
Following studies of surfactant aggregates and their activity in pooled surfactant from uninfected B6.SP-A−/− versus B6 mice, remaining experiments focused on the effects of mycoplasma infection on lung injury and surfactant abnormalities in B6.SP-A−/−, B6.iNOS−/−, and B6 mice. All strains of mice infected with intranasal mycoplasma (107 cfu) had evidence of lung injury and surfactant dysfunction (Table 3). Mycoplasma-infected B6, B6.SP-A−/−, and B6.iNOS−/− mice had higher levels of BAL total protein, higher protein/phospholipid ratios in BAL, and lower levels of LA in BAL at all time points of injury studied (24, 48, 72 h) compared with uninfected mice of the same strain (Table 3). However, no significant intrastrain differences were observed for any BAL parameter across the three time points of injury studied (Table 3). Thus, data for BAL parameters at all three injury times were combined for each mouse strain and analyzed statistically to examine differences between mycoplasma-infected and uninfected mice. Analyses of time-pooled data for each mouse strain were consistent with the results already noted at individual times in Table 3—that is, mycoplasma-infected B6.SP-A−/−, B6.iNOS−/−, and B6 mice all had significant decreases in LA in BAL (Figure 5), increased total protein levels in BAL (Figure 6), and increased protein/phospholipid ratios in BAL (Figure 7), compared to uninfected mice of the same strain. Minimum surface tension values on the bubble surfactometer were also elevated after 0.25 min and 20 min of pulsation for resuspended LA's from infected B6.SP-A−/−, B6.iNOS−/−, and B6 mice compared with uninfected mice of the same strain (Figures 8A and 8B).
In addition to documenting lung injury and surfactant dysfunction in mycoplasma-infected B6.SP-A−/−, B6.iNOS−/−, and B6 mice compared to uninfected same-strain controls, analyses of time-pooled injury data also identified significant differences between mouse strains in responses to mycoplasma infection. One-way ANOVA analysis followed by post hoc Tukey's HSD testing indicated that mycoplasma-infected B6.SP-A−/− and B6.iNOS−/− mice had significantly decreased levels of LA in BAL (Figure 5), and increased ratios of protein to phospholipid in BAL (Figure 7), compared with mycoplasma-infected B6 mice. Infected B6.SP-A−/− mice also had increased levels of total protein in BAL compared with infected B6 mice (Figure 6). In addition, resuspended LA from infected B6.iNOS−/− mice had increased minimum surface tensions after 0.25 min or 20 min of bubble pulsation compared with infected B6 mice (Figure 8). Differences in LA surface activity for infected B6.SP-A−/− mice versus B6 mice approached, but did not reach, statistical significance (P = 0.06; Figure 8).
To identify whether direct exposure to mycoplasmas contributed significantly to lung surfactant activity deficits in the absence of an inflammatory response, BAL samples from uninfected B6.SP-A−/−, B6.iNOS−/−, and B6 mice were exposed to M. pulmonis in vitro (2 × 107 cfu/ml) for 48 h. Mycoplasma-exposed BAL were then centrifuged to pellet large surfactant aggregates, and surface tension–lowering ability was assessed in comparison to control (unexposed) aggregates. As shown in Table 4, direct exposure of surfactant from B6 mice to mycoplasma organisms did not alter surface activity. In addition, only minor decreases in surface activity were apparent for mycoplasma-exposed surfactant from B6.SP-A−/− and B6.iNOS−/− mice. All surfactant samples exposed to mycoplasma in vitro reached minimum surface tension values of < 1 mN/m after 20 min of bubble pulsation, with the exception of B6.iNOS−/− mice (which still reached a very low mean minimum surface tension of 1.7 mN/m; Table 4).
Western blotting studies of BAL using specific anti–SP-B antibodies revealed a single band just below 20 kD, corresponding to the dimer form of SP-B. Respiratory mycoplasma infection caused a significant increase in the level of SP-B in BAL from B6 mice compared to uninfected same-strain controls at 48 h after inoculation (Figure 9). SP-B levels in BAL were unchanged for mycoplasma-infected B6.SP-A−/− mice at the 48-h time point compared with uninfected B6.SP-A−/− mice. In contrast, there was a significant decrease of about 40% in the levels of SP-B in BAL from mycoplasma-infected B6.iNOS−/− mice compared with uninfected controls at 48 h after inoculation with mycoplasma (Figure 9). It should be stressed that equal amounts of proteins (5 μg) were loaded in each lane. Because of the very low levels of SP-B in mouse BAL, we were unable to detect it when we stained the 12% Bis-Tris gels with SYPRO Ruby protein gel stain. We thus normalized the digitized signal of the SP-B band seen in the Western blot to a protein of about 85 kD seen when gels were stained with SYPRO Ruby. The results were identical to those shown in Figure 9. Thus, we feel confident that the decrease of SP-B levels seen in the BAL of iNOS−/− mice was not due to unequal loading of the gels.
The results of this study demonstrate that pulmonary mycoplasma infection in congenic B6.SP-A−/−, B6.iNOS−/−, and B6 mice leads to surfactant dysfunction, with decreases in both the content and surface activity of large surfactant aggregates in BAL. Mycoplasma-infected B6.SP-A−/−, B6.iNOS−/−, and B6 mice all had decreased levels of LA, increased levels of total protein, and increased protein/phospholipid ratios in BAL compared with uninfected mice of the same strain (Table 3, Figures 5–7).). Resuspended LA from mycoplasma-infected B6.SP-A−/−, B6.iNOS−/−, and B6 mice also had decreased surface activity (increased minimum surface tension values) during pulsation on a bubble surfactometer compared with uninfected same-strain controls (Figure 8). Additional analyses of time-pooled data combining results at 24, 48, and 72 h after mycoplasma infection for each mouse strain documented significant interstrain differences in the severity of mycoplasma-induced surfactant dysfunction and lung injury. Infected B6.SP-A−/− and B6.iNOS−/− mice had significantly lower levels of LA and higher protein/phospholipid ratios in BAL compared with infected B6 mice (Figures 5 and and7).7). Mycoplasma-infected B6.SP-A−/− mice also had increased levels of protein in BAL compared with infected B6 mice (Figure 6), and LA from infected B6.iNOS−/− mice had increased minimum surface tensions compared with infected B6 mice after 0.25 or 20 min of bubble pulsation (Figure 8). Infected B6.iNOS−/− mice had significantly decreased levels of SP-B in BAL compared with uninfected same-strain controls, while levels of SP-B were unchanged for infected versus uninfected B6.SP-A−/− mice and significantly increased for infected versus uninfected B6 mice (Figure 9). These results are consistent with the overall interpretation that both SP-A and iNOS contribute to limiting lung surfactant abnormalities in the presence of mycoplasma infection.
Our studies investigating the importance of SP-A and iNOS in uninfected and mycoplasma-infected mice used congenic animals all having the B6 background. Previous investigations of surfactant function in SP-A–deficient mice have been performed using animals on the outbred Black Swiss background (16, 29). While the use of outbred mice is often justified as being representative of the human (outbred) population, there are inherent differences within outbred mouse populations that complicate interstrain comparisons such as those involving responses to mycoplasma-induced lung injury (15). Mouse strains differ dramatically in resistance to infection with M. pulmonis, with B6 mice being comparatively resistant to disease (30). In-depth knowledge about the specific impact of murine genetic background on surfactant function is lacking, and SP-A and iNOS knockout mice backcrossed onto the B6 background (N10) were thus compared directly with B6 (SP-A+/+) mice to control for unknown genetic factors that might impact the surfactant system.
The surface activity deficits measured here for LA surfactant from mycoplasma-infected B6.SP-A−/−, B6.iNOS−/−, and B6 mice (Figure 8) most likely underestimate the severity of functional deficits actually present in the alveoli of infected animals in vivo. Activity studies in mycoplasma-infected animals used LA obtained by centrifugation of cell-free BAL at 12,500 × g, a centrifugation force that does not sediment free proteins (plasma- or cell-derived proteins) in whole BAL. Due to material constraints in processing mouse lavages, centrifuged LA were resuspended for activity studies in 0.15 M NaCl + 2 mM CaCl2 as opposed to the original BAL supernatants that became depleted during assays and handling. For mycoplasma-infected animals, original BAL samples contained substantial amounts of injury-induced free protein from the alveolar spaces (Figure 6), and this free protein was thus not present when LA surface activity was measured (Figure 8). Since plasma- and cell-derived proteins are known to be inhibitory to lung surfactant activity (see Refs. 7 and 31 for review), it is highly probable that surface activity deficits in all injured mice would have been more severe than shown in Figure 8 if all the original BAL protein had been present during bubble measurements. This effect would be expected to be most pronounced for BAL from mycoplasma-infected B6.SP-A−/− mice, which had the largest amounts of total protein originally present in whole BAL (Figure 6). Future physiological studies of pulmonary mechanics in different strains of mycoplasma-infected mice would be helpful in assessing the functional importance of the surface activity differences reported here.
Direct exposure to mycoplasmas did not cause significant reductions in the surface activity of lavaged lung surfactant in vitro (Table 4), indicating that the surfactant abnormalities found during mycoplasma infection were largely associated with the underlying lung injury. The surface activity detriments measured in resuspended LA from mycoplasma-infected mice (Figure 8) could result from several mechanisms. Although free proteins in BAL remain in the supernatant during centrifugation at 12,500 × g as noted above, recent work in rodents with aspiration lung injury (32) has demonstrated that some plasma- or cell-derived proteins in BAL become associated with or incorporated into surfactant aggregates so as to impair surface activity. In addition, surfactant aggregates from animals with lung injury can also exhibit changes in lipids and proteins as a result of interactions with inhibitors such as phospholipases, proteases, or reactive oxygen/nitrogen species. For example, degradation of surfactant phospholipids in injured lungs generate increased levels of inhibitory lysophosphatidylcholine in surfactant aggregates that can impair surface activity, as has been reported in rats with aspiration (32). The specific protein and lipid composition of LA from mycoplasma-infected animals was not assessed in the current study in order to avoid pooling of BAL samples that would have hampered other assessments of strain-dependent lung injury (aggregate composition was assessed only in pooled lavage from uninfected B6.SP-A−/− and B6 mice in a subset of studies in Tables 1 and and22 and in Figure 1).
Although comparisons between mycoplasma-infected B6.SP-A−/−, B6.iNOS−/−, and B6 mice (Figures 5–8)) implicate SP-A and iNOS as being important in mitigating some aspects of surfactant dysfunction and lung injury in this condition, additional mediators and factors are almost certainly also involved. Our prior studies have investigated mycoplasma lung injury in iNOS-deficient mice (15, 33, 34), with an emphasis on the roles of reactive nitrogen species in inflammatory tissue injury as well as alterations in cellular sodium transport. Infection of mice with M. pulmonis induces an infiltration of polymorphonuclear cells (PMNs) in the airways and alveoli, together with an increase in alveolar macrophages (AMs) (e.g., 14, 15, 33, 34). Prior work on innate host defenses against respiratory mycoplasmas has identified the alveolar macrophage as a primary effector cell in early mycoplasma killing in vivo (30) and in vitro (4). In vitro studies with activated alveolar macrophages from mycoplasma-resistant B6 mice indicate that SP-A and nitric oxide (NO) are important in mycoplasma killing (4). NO and other reactive nitrogen species including peroxynitrite are still present in B6.SP-A−/− and B6.iNOS−/− mice. Our experiments did not specifically assess physiological detriments in B6.SP-A−/− mice associated with the loss of the immunoprotective effects of SP-A as a member of the collectin family of host defense proteins (11–13, 35). Decreased clearance of mycoplasmas in B6.SP-A−/− and/or B6.iNOS−/− mice may have contributed to lung injury and surfactant abnormalities, although neither mouse strain exhibited a substantial increase in lung injury severity over time after infection (e.g., at 24, 48, and 72 h in Table 3).
We have previously published measurements of several histologic lung lesion indices for different strains of mice infected with mycoplasmas under conditions identical to those used here (15, 33). Lungs from mycoplasma-infected mice were coded randomly and scored subjectively for lesion severity on the basis of: (1) neutrophilic exudates in the airway lumina; (2) hyperplasia-dysplasia of the mucosal epithelium; (3) peribronchial and perivascular lymphoid accumulations; and (4) inflammatory infiltrations within the alveoli. At 72 h after infection, B6.iNOS−/− mice had significantly higher indices for all lesion parameters compared with B6 mice (33). These results are consistent with the concept that pulmonary epithelial injury induced by the immune/inflammatory response contributes to surfactant abnormalities in B6.iNOS−/− mice. In contrast, our prior work did not identify significant differences between mycoplasma-infected B6 and B6.SP-A−/− mice in histologic indices of lung injury despite the presence of higher numbers of mycoplasmas within the lungs of the knockout mice (15). However, functional changes in injured lungs can precede the appearance of overt histologic changes (e.g., rabbits exposed to hyperoxia develop significant increases in alveolar permeability before the appearance of interstitial and alveolar edema or structural changes to the blood gas barrier based on histology [56, 57]). The importance of the immune/inflammatory response in mycoplasma-induced lung injury is indicated by our prior work showing that infection of wild-type but not MPO−/− mice with mycoplasmas leads to decreased sodium-dependent fluid clearance across their alveolar epithelium in vivo (58). Detailed future investigations of inflammatory mediator responses during pulmonary mycoplasma infection would be helpful in clarifying the contributions of the immune/inflammatory response to surfactant dysfunction in this condition.
To help understand specific actions of SP-A that might contribute to surfactant abnormalities, a subset of studies focused on specific aggregation and inhibition resistance characteristics in lavaged surfactant from uninfected B6.SP-A−/− versus B6 mice (Figures 1–4,, Tables 1 and and2).2). Pooled lavage from uninfected B6.SP-A−/− mice showed significant changes in the size-distribution of surfactant aggregates relative to uninfected congenic B6 mice. BAL from uninfected B6.SP-A−/− mice had a significantly reduced content of VLA sedimenting at 1,500 × g, but an increased content of ILA sedimenting between 1,500 × g and 12,500 × g (Fig. 1). The content of total large aggregates (LA = VLA + ILA) did not differ between uninfected B6.SP-A−/− and B6 mice (Figure 1), and LA content in these mice was also the same as in uninfected B6.iNOS−/− mice (Table 3). LA from B6.SP-A−/− mice had reduced levels of total protein compared to B6 mice (Table 2), but there were no differences in terms of phospholipid class distribution (Table 1) or content of hydrophobic protein (Table 2).
The finding of compensatory increases in ILA in BAL from uninfected B6.SP-A−/− versus B6 mice (Figure 1) extends earlier studies on the composition, activity, and metabolism of pulmonary surfactant in outbred SP-A−/− mice that reported a decreased content of BAL aggregates obtained by centrifugation at 40,000 × g with and without 0.8 M sucrose (16, 29, 36). The aggregates studied here (VLA, ILA, and LA) reflect microstructural populations discriminated primarily by size as opposed to density, and represent the most active fractions of normal native surfactant (e.g., 7, 27, 28). These aggregate forms are the basis for several active clinical exogenous surfactants such as Infasurf (CLSE), which is currently used in treating surfactant deficiency and dysfunction in humans (7). Changes in VLA and ILA in B6.SP-A−/− versus B6 mice could involve modifications in aggregate size/structure within the alveolar lumen, or aggregate changes associated with altered surfactant synthesis, reuptake, or recycling in type II pneumocytes that are normally regulated by SP-A (6, 7, 11, 37). It is also possible that some aggregate changes in SP-A−/− mice could involve contributions induced by SP-D, which is found in normal amounts in lung homogenates from these mice (16). SP-D does not participate in the biophysics of normal surfactant containing SP-A (7), and it is not present in normal type II cell lamellar bodies (38, 39). However, SP-D has structural similarities to SP-A, including an N-terminal collagenous domain and a C-terminal carbohydrate recognition domain (40, 41). SP-D has been reported to bind and agglutinate carbohydrates but not to affect the surface activity of phospholipids (42).
Our finding that BAL and LA obtained from uninfected B6.SP-A−/− versus B6 mice were more sensitive to inhibition by LPC and BSA in vitro (Figures 3 and and4)4) is consistent with previous studies showing that SP-A improves inhibition resistance in hydrophobic lung surfactant extracts (8–10). SP-A has also been shown to mitigate surfactant dysfunction from peroxynitrite (43) and Cu-Zn superoxide dismutase (44), and to enhance phospholipid adsorption and film respreading in vitro (45–49). The effects of SP-A in enhancing inhibition resistance likely involve cooperative interactions with SP-B in forming tubular myelin (50, 51), plus the facilitation of less specific phospholipid aggregation in the aqueous phase (e.g., 51–55). It has previously been reported that minimum surface tension was increased in films of lung surfactant from SP-A−/− mice compared with Black Swiss mice in the presence of albumin on the Wilhelmy balance during cycling at slow rate (3 min/cycle) (29). Functional assessments in the present study utilized a pulsating bubble surfactometer, which measures a more physiologically relevant combination of adsorption and dynamic surface activity at a cycling rate (20 cycles/min) and area compression (50%) reflective of the lungs in vivo (7, 23, 24). The decreased inhibition resistance found for surfactant from B6.SP-A−/− mice in the presence of albumin and LPC in pulsating bubble studies (Figures 3 and and4)4) is consistent with this being a contributor to surfactant activity deficits in mycoplasma-infected B6.SP-A−/− mice (Figure 8).
Although surfactant from uninfected B6.SP-A−/− mice had less resistance to inhibition by albumin and LPC in vitro, LA from these animals maintained a normal phospholipid composition and a normal total hydrophobic protein content (Tables 1 and and2).2). Due to this normal content of active hydrophobic components, as well as a compensatory increase in ILA (Figure 1), surfactant from B6.SP-A−/− mice maintained near-normal surface activity in the absence of inhibitors (Figure 2). This finding is consistent with previous studies showing that lung compliance is effectively normal in uninfected outbred SP-A−/− mice (16). However, in the presence of mycoplasma infection, significant differences in the level of an important specific surfactant protein (SP-B) were found in BAL from B6.SP-A−/− and B6.iNOS−/− mice compared to infected B6 mice (Figure 9). At 48 h after infection with mycoplasma, infected B6 mice had a significant increase in SP-B in BAL compared with uninfected controls, while levels of SP-B were unchanged in infected B6.SP-A−/− mice and significantly decreased in infected B6.iNOS−/− mice compared to uninfected same-strain controls (Figure 9). Multiple studies have documented that SP-B is the most active of the hydrophobic surfactant proteins in facilitating adsorption and dynamic surface activity in endogenous surfactant (see Refs. 7 and 31 for review). Because of its combination of nonpolar and polar residues, amphipathic SP-B can interact with both the fatty chains and headgroups of phospholipid molecules. It is very likely that mycoplasma-induced reductions in SP-B in infected iNOS−/− mice, in particular, were an important contributor to the decreased surface activity found in LA from these animals (Figure 8).
In summary, the present study has demonstrated that mycoplasma-infected B6.SP-A−/−, B6.iNOS−/−, and B6 mice all had lung injury with increased protein and protein/phospholipid ratios in BAL, depleted LA surfactant, and impaired surface activity compared with uninfected mice of the same strain. Moreover, mycoplasma-infected B6.SP-A−/− and B6.iNOS−/− mice had more severe lung injury and surfactant dysfunction compared with B6 mice based on analyses of time-pooled data from 24, 48, and 72 h after infection, indicating a role for SP-A and iNOS in this disease. Abnormalities in lung surfactant in infected B6.iNOS−/− mice included a significantly decreased level of SP-B in BAL, which correlated directly with decreased surface activity in surfactant aggregates in BAL from these mice. An additional likely contributor to surface activity deficits in mycoplasma-infected B6.SP-A−/− mice is that lavaged surfactant from SP-A–deficient mice is shown here to have a decreased ability to resist biophysical inactivation by albumin and lysophosphatidylcholine when examined in vitro.
Ms. Glenda Davis (University of Alabama at Birmingham) provided outstanding technical support with infecting and lavaging mice. The authors also thank Dr. M. F. Beers (University of Pennsylvania) for many helpful discussions concerning the SP-B measurements, and Ms. A. McCole for editorial assistance.
The authors gratefully acknowledge the support of the National Institutes of Health through grants HL-56176 (R.H.N., Z.W., P.R.C.), HL-31197 (S.M.) and HL-51173 (S.M.), as well as the support of the American Lung Association through grant RG9928-N (J.M.H.-D.).
Originally Published in Press as DOI: 10.1165/rcmb.2006-0049OC on August 17, 2006
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.