INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lung surfactant, a complex mixture of phospholipids (PL) and at least four apoproteins, reduces surface tension at the alveolar surface, promoting lung expansion during inspiration and preventing lung collapse during expiration. Surfactant is synthesized in the type II cells lining the alveoli, stored in lamellar bodies, and secreted into the alveolar lining fluid, where the lamellar bodies are transformed into tubular myelin, the source of PL for the surface-active monolayer at the air-water interface. Surfactant is cleared from the airways through reuptake and recycling by alveolar type II cells. During its complex life cycle, lung surfactant converts from highly surface active, large aggregates to less surface active, smaller aggregates (1, 2). This process is probably regulated by a 68-kD serine protease (surfactant convertase) produced by type II cells and alveolar macrophages (AM) (9), and depends on repeated expansion and contraction of the air-fluid interface (3, 4). Although other enzymes such as phospholipases can also inhibit surfactant function (7), and surfactant convertase may not be the only enzyme involved in the extracellular metabolism of surfactant, convertase has been shown to be highly specific for surfactant subtype conversion (3). Sequencing of purified surfactant convertase has shown it to be a serine-active carboxylesterase (10, 11). In vitro and in vivo studies indicate that surfactant convertase is sensitive to inhibition by ₁-antitrypsin (₁-AT) (5, 6).

₁-AT is the principal antiprotease in both the circulation and the lung, and is normally found in the alveolar fluid lining layer. The primary function of this 52-kD serine protease inhibitor is to inhibit neutrophil elastase, but it also inhibits other serine proteases, such as trypsin, chymotrypsin, and thrombin (8). Increasing alveolar levels of ₁-AT by intratracheal instillation of exogenous ₁-AT delays the conversion of large to small surfactant aggregates (5), suggesting that ₁-AT affects the extracellular metabolism of surfactant by inhibiting the serine protease present in surfactant.

We hypothesized that after its instillation into the airways, the addition of ₁-AT to exogenous surfactant would reduce conversion of the latter from large to small aggregates by inhibiting serine protease activity in the alveolar space, and would further improve the effect of the surfactant on lung function. We added ₁-AT to three types of surfactant preparations, as follows: (1) a palmitic acid-containing PL mixture; (2) PL mixture with synthetic surfactant proteins B (SP-B) and C (SP-C) (BC mixture); and (3) a natural bovine surfactant (Survanta). We then tested whether the addition of ₁-AT improved lung function in ventilated, surfactant-deficient rats.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Materials

Peptide synthesis reagents were purchased from Applied Biosystems (Foster City, CA), high-performance liquid chromatography (HPLC) solvents from Fisher Chemical Co. (Pittsburgh, PA), and all other chemicals from Sigma Chemical Co. (St. Louis, MO) and Aldrich Chemical Co. (Milwaukee, WI). Dipalmitoyl phosphatidylcholine (DPPC) and 1-palmitoyl-2-oleoyl phosphatidyl-glycerol (POPG) were obtained from Avanti Polar Lipids (Alabaster, AL). All surfactant preparations were used at a lipid concentration of 2.5 mg/ml in surface activity studies, and of 25 mg/ml in in vivo studies. Adult male Sprague-Dawley rats weighing 200 to 220 g were obtained from B & K Universal Inc. (Fremont, CA).

Synthesis and Purification of SP-B and SP-C

Full-length SP-B_1-78 (B) and palmitoylated SP-C_1-35 (C), based on the human sequences, were synthesized on a 0.25-mmol scale with a model 431A peptide synthesizer (Applied Biosystems-Perkin Elmer, Foster City, CA), using FastMoc chemistry (12). The surfactant proteins were then purified and characterized as described previously (13, 14).

Experimental Surfactant Preparations

Surfactant preparations used in these experiments included: (1) a PL mixture consisting of DPPC, POPG, and palmitic acid (PA) (DPPC:POPG:PA = 69:22:9 wt/wt/wt); (2) a synthetic surfactant prepared by mixing 3% synthetic SP-B and 1% synthetic SP-C into the PL mixture (BC mixture); and (3) Survanta, a modified bovine lung extract approved for clinical use, which contains SP-B and SP-C (Beractant; Ross Laboratories, Columbus, OH). Surfactant preparations (25 mg of PL/ml) were enriched with ₁-AT from human plasma (Sigma) at a dose of 100 mg/3 ml by simple mixing.

In Vitro Surface Activity

Changes in surface tension of the various surfactant preparations were measured during compression on unbuffered 0.9% NaCl at room temperature in a modified Langmuir-Wilhelmy balance (KimRay; Greenfield Surfactometer, Oklahoma City, OK) (15, 16). Samples containing 25 µg of PL were loaded onto a saline subphase in a 51.5 cm² rectangular Teflon trough. Compression of the surface film from 100% to 15% of the total area was done five times, with a cycle time of 90 s. Minimum and maximum surface tension data were collected from the first and fifth run. Four measures were made for each data point.

Animal Protocol

Adult rats were anesthetized with 35 mg/kg of pentobarbital sodium and 80 mg/kg of ketamine by intraperitoneal injection. After placement of a tracheal cannula (I.D. = 2.0 mm), the rats were supported with a rodent ventilator (Harvard Apparatus, South Natick, MA) with 100% O₂, a tidal volume of 7.5 ml/kg, and a ventilatory rate of 60 breaths/min. Mean airway pressure was an electronic mean of airway pressures sampled at the top of the tracheal cannula. An arterial line was placed in the abdominal aorta for serial measurements of pH and blood gases, and the rats were paralyzed with 2 mg/kg of pancuronium bromide intravascularly. Only rats with Pa_O2 values > 400 mm Hg during ventilation with 100% O₂ were included in the experiments. Their lungs were gently lavaged from eight to 12 times with 8 ml of 0.9% NaCl warmed to body temperature in each lavage. Ten minutes after the Pa_O2 in 100% O₂ fell below 100 mm Hg, one of the experimental surfactants was instilled intratracheally as a quick bolus at a dose of 100 mg PL/kg body weight, with the rats in a supine position and without sighing. Rats were ventilated for 60 min, and had their arterial blood gases measured at 15-min intervals. At 60 min, ventilation settings were changed by adding 3 cm of positive end-expiratory pressure (PEEP) for 5 min. After further arterial blood gas measurement, the rats were killed with 100 mg/kg pentobarbital sodium intravascularly, and were exsanguinated and had their lungs degassed in situ. A pressure-volume curve was generated in situ for each pair of rat lungs, to define lung mechanics. Lungs were inflated and deflated with a closed chest, using a bidirectional Harvard pump (Harvard Apparatus Inc., Holliston, MA) coupled to a 50-ml glass syringe, and pressure was continuously recorded on a Gould multichannel recorder (Gould Instrument Systems Inc., Cleveland, OH). Each pressure-volume curve was corrected for the compliance of the system by subtracting the pressure-volume curve of the pump/syringe unit before generating each curve. Absence of air leaks was assessed by verifying that lung volumes changed by less than 0.1 ml/min over a period of 3 min at 30 cm H₂O pressure. The lung volume measured at a pressure of 5 cm H₂O was used as an index of stability at low lung volumes (V₅), and the lung volume measured at a pressure of 30 cm H₂O was used as an index of stability at high lung volumes (V₃₀). The lungs were then gently relavaged three times with 8 ml of 0.9% NaCl warmed to body temperature. Each treatment group consisted of eight animals. Previous work (17) has shown that lung-lavaged rats treated with an air placebo instead of surfactant have persistently low Pa_O2 values in this experimental setting. All experiments were performed humanely and with the approval of the animal care and use committee of our institution.

Lung Lavages

The first three lung lavages done on each rat to induce surfactant deficiency (pretreatment lavage), and the three lung lavages done at the end of the experiment (posttreatment lavage), were pooled. Protein content of the pre- and posttreatment lavages was measured with a modified Lowry assay (18). Lipid content of the lung lavages was determined by Fourier transform infrared (FTIR) spectroscopy (19), using a Mattson Research Series FTIR spectrometer (Mattson Instruments, Madison, WI) fitted with a deuterated triglycerine sulfate (DTGS) detector. Lavage material was air-dried onto a 50 mm × 20 mm × 3 mm, 45° germanium attenuated total reflectance (ATR) crystal (Spectra Tech, Stamford, CT). The ATR crystal was mounted in a Model 304 variable ATR accessory (Spectra Tech), and spectra were averaged from 64 scans at a gain of 4 and a spectral resolution of 2 cm¹. The relative amount of lipid was estimated from analysis of the C-H stretching region (3,000 to 2,800 cm¹). Surfactant particle conversion was measured by analysis of the particle size distribution in the lung lavages, diluted 1:10 with phosphate-buffered saline (pH = 7.3), by dynamic light scattering (20), using a Microtrac Ultrafine Particle Analyzer (UPA; Leeds & Northrup, North Wales, PA) (21).

Statistical Analysis

Surface activity data are presented as mean ± SEM, with a minimum of four measures for each data point. The arterial/alveolar (a/A) PO₂ ratio was calculated from the Pa_O2, Pa_CO2, and fraction of inspired oxygen (FI_O2) values and was used to express oxygenation. Oxygenation and data from the pressure-volume curves are given as mean ± SEM, with eight rats in each experimental group. Mean airway pressures are described as mean ± SD. A two-tailed t test was used to compare two groups. The significance of differences between multiple experimental groups was compared through one-way analysis of variance (ANOVA). In those cases in which the F test showed a significant difference (p < 0.05) among groups, comparisons of different groups were made with the Student-Newman-Keuls test. A value of p < 0.05 was considered as indicating a significant difference.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All six surfactant preparations had in vitro minimum surface tensions < 10 mN/m, except for the PL mixture + ₁-AT. Hystereses with the preparations varied between 17 and 22 cm² on the first isotherm, except for the PL mixture (9 cm²) (Figure 1). Reproducibility of the hystereses (fifth versus first isotherm) ranged from 54 to 101% as follows: PL mixture = 77%, PL mixture + ₁-AT = 54%, BC = 65%, BC + ₁-AT = 76%, Survanta = 89%, and Survanta + ₁-AT = 101%. The addition of ₁-AT did not appear to interfere with the characteristic hysteresis for any preparation in the presence of native or synthetic surfactant proteins. However, the dispersion of the pure PL mixture showed a significant difference from the other preparations in the breadth of hysteresis and in the minimum surface tension produced. In the presence of ₁-AT, a distinct collapse plateau appeared at about 50% surface compression, indicating the formation of an insoluble surface monolayer at about 20 mN/m, a much higher minimum surface tension than for any of the other dispersions, with a value typical of compressed films of unsaturated PL. ₁-AT alone did not show surface activity.

View larger version (33K): [in this window] [in a new window]   Figure 1.   In vitro surface activity of the six surfactant preparations (PL mixture, PL mixture + 1-AT, Survanta, Survanta + 1-AT, BC mixture, and BC mixture + 1-AT) tested on a modified Langmuir/Wilhelmy balance.

Groups of eight rats were lavaged and treated with one of the six surfactant preparations or with a similar dose of ₁-AT alone. Additional control groups included ventilated unlavaged rats (hyperoxia controls) and ventilated lavaged rats that did not receive any surfactant or ₁-AT therapy (No Rx group). The number of lavages needed to reduce Pa_O2 < 100 mm Hg varied from six to 24 (median = 10). Before the lavage procedure, Pa_CO2 was 26 ± 1 mm Hg, Pa_O2 was 474 ± 6 mm Hg, and the a/A ratio was 0.689 ± 0.008. After lung lavage, these values were Pa_CO2 = 51 ± 2 mm Hg, Pa_O2 = 75 ± 2 mm Hg, and a/A ratio = 0.113 ± 0.003, respectively. Treatment with a bolus of surfactant led to a quick and sustained increase in a/A ratios (Figure 2). The BC surfactant preparation was more effective in improving oxygenation than was Survanta (p < 0.02), and Survanta outperformed the PL-mixture surfactant (p < 0.05). Addition of ₁-AT to the PL mixture and Survanta increased their effect on oxygenation (p < 0.005 and p < 0.05, respectively) and erased their differential effects, but addition of ₁-AT to the BC preparation did not significantly increase a/A ratios over those attained with the BC preparation alone. The oxygenation effect of postlavage treatment with ₁-AT alone was similar to that with instillation of the PL-mixture. The a/A ratio for rats in the No Rx group did not change significantly over time. Application of PEEP at 60 min after the various treatment modalities increased a/A ratios in all groups, but did not change the intergroup differences (Figure 2).

View larger version (26K): [in this window] [in a new window]   Figure 2.   Mean ± SEM a/A PO2 ratios after lung lavage and 15, 30, 45, 60, and 65 (+ PEEP) min after intratracheal instillation of PL mixture, PL mixture + 1-AT, Survanta, Survanta + 1-AT, BC mixture, BC mixture + 1-AT, or 1-AT only in rats (n = 8 per group) ventilated with 100% oxygen. Controls included lavaged but untreated rats (Lavage, No Rx) and unlavaged untreated rats (Hyperoxia controls). (*p < 0.05 versus all other experimental groups; #PL mixture and 1-AT groups: p < 0.05 versus Survanta, Survanta + 1-AT, BC mixture, BC mixture + 1-AT, and PL mixture + 1-AT; p < 0.05 versus BC mixture, BC mixture + 1-AT, Survanta + 1-AT, and PL mixture + 1-AT).

The hyperoxia controls had lung volumes exceeding those of lavaged rats rescued from respiratory failure by surfactant treatment, whereas rats in the No Rx group had the lowest lung volumes (Figure 3). Rats treated with the PL mixture + ₁-AT or with Survanta + ₁-AT had lower V₃₀ values than those treated with the corresponding surfactants without ₁-AT (p < 0.005 and p < 0.001, respectively), whereas the differences for the BC mixture and the BC mixture + ₁-AT were not significant (Figure 3). V₅ values were lower in rats treated with the BC mixture + ₁-AT or Survanta + ₁-AT than in those treated with the BC mixture or Survanta alone (p < 0.02 and p < 0.01, respectively).

View larger version (24K): [in this window] [in a new window]   Figure 3.   Deflation limbs of the pressure-volume curves (mean ± SEM) of rats ventilated for 1 h with 100% oxygen following lung lavage and treatment with PL mixture, PL mixture + 1-AT, Survanta, Survanta + 1-AT, BC mixture, BC mixture + 1-AT, or 1-AT only. Controls included lavaged untreated rats (Lavage, No Rx) and unlavaged, untreated rats (Hyperoxia controls). (**p < 0.005: PL mixture + 1-AT versus PL mixture; *p < 0.02: BC mixture + 1-AT versus BC mixture; p < 0.01 Survanta + 1-AT versus Survanta).

Mean airway pressure values needed to deliver a tidal volume of 7.5 ml/kg increased sharply after lung lavage, and continued to rise in the No Rx group (Figure 4). Mean airway pressures were lower after surfactant or ₁-AT instillation than in the No Rx group. Addition of ₁-AT to the PL mixture and to Survanta further decreased mean airway pressures in these groups.

View larger version (23K): [in this window] [in a new window]   Figure 4.   Average mean airway pressures (MAP: mean ± SD) of rats ventilated with 100% oxygen before (prelavage) and 5 min after completion of lung lavage (postlavage), and of rats ventilated for 1 h with 100% oxygen after lung lavage without further treatment (No Rx) or lung lavage followed by instillation of PL mixture, PL mixture + 1-AT, Survanta, Survanta + 1-AT, BC mixture, BC mixture + 1-AT, or 1-AT only. (*p < 0.05 versus all other groups; p < 0.05 versus PL mixture; #p < 0.05 versus Survanta).

Recovery of lavage fluid during the first and the final lavages was > 95%. The total protein concentration in the final lavages of all groups (2,093 ± 89 µg/ml) was higher than in the first lavages (292 ± 28 µg/ml) (p < 0.001), without significant intergroup differences. Total lipids in the final lavages were uniformly 62% higher than in the first lavages, which was again irrespective of the surfactant preparation used.

Particle sizes in the pretreatment lavages were uniformly distributed in the > 1 µm range, but decreased to < 1 µm in posttreatment lavage material from rats that did not receive ₁-AT or surfactant following lavage (Figure 5). Among rats treated with one of the three unmodified surfactants, the rank order for average particle size was BC mixture > Survanta > PL mixture. Lavages from rats treated with surfactant + ₁-AT showed a major increase in average particle size over those from rats treated with the same surfactant without ₁-AT (Figure 5). Particle sizes in lavaged rats treated with ₁-AT alone only showed a uniform distribution in the > 0.8 µm range. Particle sizing by dynamic light scattering was compared with the classic centrifugation approach to small and large aggregate quantification. Figure 6 shows the mean particle size distribution of the pellets (large aggregate forms) and supernatants (small aggregate forms) obtained by centrifugation for 15 min at 40,000 × g of samples of pretreatment (n = 7) and posttreatment lavages (n = 6).

View larger version (26K): [in this window] [in a new window]   Figure 5.   Average surfactant particle size distribution obtained by dynamic light scattering in pretreatment lavages (Prelavage) and in posttreatment lavages of rats treated with PL mixture, PL mixture + 1-AT, Survanta, Survanta + 1-AT, BC mixture, BC mixture + 1-AT, or 1-AT only. Data for untreated lavaged rats (Lavage, No Rx) are included for comparison.

View larger version (20K): [in this window] [in a new window]   Figure 6.   Average surfactant particle size distribution obtained by dynamic light scattering of the pellets (large aggregate forms) and supernatants (small aggregate forms) obtained by centrifugation for 15 min at 40,000 × g of pretreatment (A) lavages (n = 7) and posttreatment (B) lavages (n = 6).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study was unique in that we measured lung function (oxygenation and respiratory system compliance) and surfactant aggregate size in ventilated, lavaged, surfactant-deficient rats treated with various types of surfactant with and without added ₁-AT. We used a protein-free, palmitic acid-containing PL mixture, PL with synthetic SP-B and SP-C (BC mixture), and a bovine surfactant extract with SP-B and SP-C (Survanta). These three preparations were chosen because they represent the various classes of surfactant currently available. Previous experience with these surfactants in the surfactant-deficient rat has shown that they improve oxygenation in the rank order: BC mixture > Survanta > PL mixture (14, 17). The addition of ₁-AT resulted in a dramatic improvement in oxygenation in surfactant-deficient rats treated with the PL mixture or Survanta, but not in rats treated with the BC mixture. Improved oxygenation was associated with a reduction in mean airway pressure and an increase in surfactant aggregate size in posttreatment lung lavages, but a decrease in postmortem lung volume. The decrease in lung volume was probably the result of the presence of considerable amounts of protein in the airways after treatment with a surfactant containing ₁-AT. The increase in surfactant aggregate size after treatment with surfactants containing ₁-AT is in agreement with the observation by Gross and colleagues (5) that ventilated, lavaged mice respond to intratracheal instillation of 5 mg of ₁-AT by generating less light-subtype and more heavy-subtype surfactant. The collected data suggest a positive role for ₁-AT in preventing exogenous surfactant degradation in the lower airways.

Two findings in our study were peculiar. First, the in vitro surface activity of the PL mixture + ₁-AT did not correlate with its in vivo effect on oxygenation. Although its hysteresis area was double that of the PL mixture alone, the minimum surface tension of the PL mixture + ₁-AT far exceeded 10 mN/m. Although a definitive argument cannot be made on the basis of the available data, we speculate that of the components of the PL mixture annealed to ₁-AT, POPG is preferentially bound and is presented to the interface, thus producing the characteristic isotherm. Alternatively, DPPC might be bound to ₁-AT and removed from the monolayer, but this possibility is less likely in view of the protective effect of ₁-AT on lung surfactant. Second, the oxygenation and pressure-volume curve data in the study did not correspond, since oxygenation improved whereas lung volume decreased when ₁-AT was added to a surfactant preparation. This phenomenon might have been due to the generation of intrinsic PEEP by the increased protein content of an ₁-AT-containing surfactant, and a concomitant increase in small-airway resistance.

The average surfactant aggregate size in posttreatment lavages was larger in the presence of ₁-AT, and this was associated with improvement in oxygenation. Inhibition by ₁-AT of convertase activity and the conversion of large to small aggregates is associated with an accumulation of large surfactant aggregates, but does not affect the amount of lipids in lung lavage material, which might suggest that the presence of ₁-AT does not lead to an accelerated reuptake of small surfactant aggregates by alveolar type II cells.

This study had several limitations. The dosage of ₁-AT (133 mg/kg) was extrapolated from the dose used by Gross and colleagues in mice (5), and presents a significant protein load to the surfactant preparations (used at 100 mg PL/kg), with the potential for direct surfactant inhibition. Lower doses of ₁-AT may be equally or more effective, and will have to be investigated. Also, surfactant aggregate sizing is usually done with a centrifugation protocol (22) rather than with dynamic light scattering, which is a standard technique in the study of liposomes (20, 23). We have not extensively investigated how particle sizing with laser beam dynamic light scattering compares with the classical centrifugation protocol, but in a small pilot study done with pre- and some posttreatment lung lavages, we found a reasonable correlation of results with the two methods. Rats were ventilated for 1 h after surfactant instillation. Although this is a limited period, surfactant therapy leads to rapid changes in lung function, which have reached their maximum response at that time.

We conclude that the addition of ₁-AT to various surfactant preparations improves oxygenation in ventilated, lavaged, surfactant-deficient rats. Improved oxygenation was associated with an increase in surfactant aggregate size in posttreatment lung lavages, and with a decrease in lung volume. The former was probably due to inhibition of surfactant convertase activity by ₁-AT, and the latter may have been due to surfactant inhibition by the protein load presented by ₁-AT. The equilibrium between these two effects seems to work in favor of enhanced oxygenation in the presence of exogenous ₁-AT.