INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Surfactant treatment of infants with respiratory distress syndrome (RDS) decreases morbidity and mortality and is now the standard of care. Nevertheless, many infants of very low birth weight progress to having bronchopulmonary dysplasia (BPD), and airway samples from infants destined to have BPD have increased numbers of proinflammatory cells and cytokines within the first days after birth and initiation of ventilation (1). The surfactants used to treat preterm infants with RDS are isolated from animal lungs, and the organic solvent extraction and purification steps used in their isolation remove the innate host-defense protein surfactant protein A (SP-A) while leaving variable amounts of SP-B and SP-C. SP-A is a member of the collectin family of proteins that bind carbohydrates and lipopolysaccharides (2). Although SP-A binds to phospholipids and is essential for tubular myelin formation, SP-A-ablated mice have normal surfactant function, metabolism, and tolerance to exercise and hyperoxia (3). However, SP-A-deficient mice are more susceptible to bacterial and viral infections (6, 7). SP-A enhances the efficiency of clearance and killing of bacteria, fungi, and viruses in vitro and in SP-A- deficient mice that are treated with SP-A (7, 8). Immunomodulatory functions of SP-A in the airways include binding and modulation of processing of endotoxin, modulation of cytokine production by alveolar macrophages (AM) (9, 10), and interaction with the endotoxin receptor CD14 (11).

SP-A has been proposed as being either pro- or antiinflammatory on the basis of its properties in vitro (12). There is no information about host-defense effects of exogenous SP-A in the preterm lung, in which SP-A is normally present in very small amounts. The initiation of ventilation with mechanical ventilation induces transcription of proinflammatory cytokines in preterm lamb lungs (15, 16). This cytokine induction is presumed to be injurious, because preterm fetuses exposed to chorioamnionitis and increased levels of tumor necrosis factor (TNF)-, interleukin (IL)-1, IL-6, and IL-8 have an increased risk of developing BPD, as do infants with increased levels of the same cytokines in samples of bronchoalveolar lavage fluid (BALF) taken within the first days of life (1, 17). In preterm infants, low concentrations of SP-A in the airways correlated with increased mortality and with a higher rate of BPD (18). In the preterm baboon model of BPD, SP-A was present in lung tissue, but was present at very low levels in BALF, demonstrating an abnormality of secretion of SP-A (19).

The host-defense functions of the preterm fetal lung differ from those of the adult lung because the fetal lung contains virtually no granulocytes or mature macrophages (1). Multiple immune defects have been described for white blood cells from infants, and levels of innate host-defense molecules such as lysozyme and defensins also are low (20). We hypothesized that SP-A would alter the proinflammatory response of the preterm lung to mechanical ventilation. To test this hypothesis, we treated preterm lambs with an effective synthetic surfactant that contained human recombinant SP-C, or with the same surfactant supplemented with 3% sheep SP-A, and evaluated lung function, surfactant components, and multiple local and systemic indicators of inflammation.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

SP-A and SP-C Surfactant

SP-A was isolated from alveolar washes of healthy adult sheep by butanol extraction (21). After extraction of whole surfactant with butanol, insoluble proteins were resuspended in octylglucopyranoside, and SP-A was solubilized in 5 mM Tris-HCl in water (pH 7.4). SP-A was dialyzed for 48 h against the same buffer to remove residual octylglucopyranoside, and the purification was verified by sodium dodecylsulfate-polyacrylamide gel (SDS-PAGE) analysis. The SP-A used to treat the lambs was tested for bacterial endotoxin contamination with the Limulus amebocyte lysate assay (Sigma Chemical Co., St. Louis, MO), and no endotoxin was detected. The treatment dose of 3 mg/kg SP-A in 100 mg/kg of recombinant human SP-C surfactant (Byk Gulden, Konstanz, Germany) was similar to the amount of SP-A and surfactant recovered from alveolar lavage of full-term animals. The SP-C surfactant contained 2% recombinant human SP-C in phospholipids (dipalmitoylphosphatidylcholine and palmitoyloleoylphosphatidylglycerol in a 70:30 [wt/wt] ratio) and 5% palmitic acid (22). This SP-C surfactant was very effective when tested in preterm lambs and in models of lung injury (22, 23). Currently, SP-C surfactant is being tested in clinical trials for the treatment of acute respiratory distress syndrome (ARDS) (24). The recombinant SP-C consists of the 34-amino-acid human SP-C peptide, altered by the replacement of cysteine in positions 4 and 5 with phenylalanine, and of methionine in position 32 by isoleucine. The altered sequence of recombinant SP-C improves solubility and prevents the aggregation of native SP-C, but maintains the functions of the dipalmitoylated form of SP-C (25). This recombinant protein has clearance kinetics from the airways and lungs of rabbits and preterm lambs that are similar to those of native SP-C (22). The SP-C surfactant was resuspended in 0.9% NaCl to a final concentration of 25 mg/ml.

Delivery and Ventilation of Preterm Lambs

Preterm lambs were delivered at 130 d to 132 d gestational age by cesarean section, as previously described (term is 150 d) (22, 26). Each pregnant ewe was preanesthetized with ketamine (20 mg/kg intramuscularly) and acepromazine (0.3 mg/kg), and then given spinal-epidural anesthesia (10 ml of 2% lidocaine and 0.5% bupivacaine [1:1, vol/ vol]). After exposure of the head, the preterm lamb was given ketamine (10 mg/kg intramuscularly) and acepromazine (0.1 mg/kg intramuscularly). A tracheal tube of 4.5-mm I.D. was inserted and tied into the trachea after local anesthesia of the anterior neck with 2% lidocaine. Fluid from fetal airways was removed with a syringe. After the umbilical cord was cut, the newborn lamb was delivered and weighed. Ventilation was started with the following settings: fraction of inspired oxygen (FI_O2) = 1.0, respiratory rate (RR) = 60 breaths/min, inspiratory time (TI) = 0.4 s, positive end-expiratory pressure (PEEP) = 4 cm H₂O, and a peak inspiratory pressure (PI_max) sufficient to yield a tidal volume (VT) of 6 to 8 ml/kg, but with PI_max limited to 40 cm H₂O. VT was monitored continuously (CP-100; Bicore Monitoring Systems, Anaheim, CA). A size 5 French catheter was advanced into the aorta via an umbilical artery, and a 10 ml/kg transfusion of filtered fetal blood collected from the placenta was administered within 10 min of delivery. After 15 min of ventilation, each preterm lamb was randomly assigned to treatment with 100 mg/kg SP-C surfactant (n = 7; SP-C surfactant group) or 100 mg/kg SP-C surfactant plus 3 mg/kg SP-A (n = 7; SP-A + SP-C-surfactant group). The surfactant treatment was given in two equal aliquots, with the lamb in the left lateral and right lateral positions, respectively (26). The initial ventilation for 15 min without surfactant treatment was used to initiate the injury process in the preterm lungs, and to mimic the clinical sequence for resuscitation followed by surfactant treatment of preterm infants. After surfactant treatment, the ventilation pressure was adjusted to achieve an arterial carbon dioxide tension (Pa_CO2) of 45 to 60 mm Hg and an arterial oxygen tension of 150 to 200 mm Hg with a VT of 6 to 8 ml/kg. VT was measured at 15 min, 45 min, 2 h, 4 h, and 6 h with a pneumotachometer, and dynamic compliance (Cdyn) was calculated as VT normalized to body weight and divided by the ventilatory pressure (PI_max  PEEP) (22, 26). The ventilatory efficiency index (VEI) was calculated as VEI = 3,800/(respiratory rate × [PI_max  PEEP] × Pa_CO2], where 3,800 is a CO₂ production constant ([ml × mm Hg]/[kg × min]) (25). Complete blood counts and differentials were done on from cord blood and on peripheral blood at 6 h. The arterial catheter was used for blood gas analysis, pH measurement, and blood pressure recording, and to infuse 10% dextrose (100 ml/kg/d). Rectal temperature was monitored and kept at 38° to 39° C with heating pads and radiant heat. Supplemental ketamine (10 mg/kg intramuscularly) and acepromazine (0.1 mg/kg intramuscularly) were administered to suppress spontaneous breathing. At 2 h of afterbirth, a sample of BALF was obtained by wedging a feeding tube in an airway and infusing 1 ml of saline, followed by aspiration. After 6 h, each animal was deeply anesthetized with 25 mg/kg pentobarbital given intravenously, and was ventilated briefly with 100% oxygen. The endotracheal tube was clamped for 3 min to permit oxygen absorption. The animal was exsanguinated by cutting the abdominal aorta.

Pressure-Volume Curve and Lung Processing

The thorax of the lamb was opened, the lungs were inflated with 40 cm H₂O pressure for 1 min, and maximal lung volume was recorded (26). The pressure was sequentially lowered to 20, 15, 10, 5, and 0 cm H₂O, and lung volumes were recorded 30 s after each pressure was reached. Volumes were corrected for the compliance of the system. Lungs were then removed from the thorax. Pieces of the right lower lobe were immediately frozen in liquid nitrogen for RNA isolation. Alveolar washing was done on the left lung by filling the lung with 0.9% NaCl at 4° C until it was visually distended, and was repeated five times (22, 26). The washes were pooled and aliquots were saved for measurement of saturated phosphatidylcholine (SatPC), total protein, cell number and differential count, surfactant protein quantification, and flow cytometric analysis. Cell pellets were used for hydrogen peroxide assay and RNA isolation. The lung was homogenized in 0.9% NaCl and SatPC was measured in aliquots of homogenate.

Measurement of SatPC and Total Protein

SatPC was isolated from chloroform-methanol (2:1) extracts by neutral alumina column chromotography after exposure to osmium tetroxide, and was quantified by phosphorus assay (25). Total protein in alveolar washes was measured as previously described (26).

Alveolar Cells

Collected alveolar washings were centrifuged at 500 × g for 10 min and the pellet was resuspended in phosphate-buffered saline (PBS). Total cells were stained with tryptan blue and counted. Differential cell counts were performed on cytospin preparations after staining with Diff-Quik (American Scientific Products, San Diego, CA). In order to assess the activation state of the cells recruited to the airways, hydrogen peroxide was measured with an assay based on the oxidation of ferrous iron (Fe²⁺) to ferric iron (Fe³⁺) by hydrogen peroxide under acidic conditions (Bioxytech H₂O₂-560 Assay; OXIS International, Portland, OR).

Aliquots of alveolar cells were incubated on ice with monoclonal antibodies (primary antibody) against ovine CD11b (M subunit of integrin CR3), CD14 (receptor for complex of lipopolysaccharide and lipopolysaccharide binding protein), and CD44 (proteoglycan link protein). The cell pellet was washed twice with PBS to remove unbound antibody, and was incubated with phycoerythrin (PE)-labeled F(ab')₂ anti-IgG fragments (secondary antibody) in the dark on ice. Control staining was performed with isotype antibodies and with secondary antibody alone to obtain background fluorescence. Cells were washed twice, resuspended in PBS, kept on ice, and immediately analyzed on a fluorescence-activated cell sorter (FACS) (Calibur; Becton-Dickinson, Inc., Mountain View, CA). All antibodies were purchased from Serotec (Raleigh, NC).

Western Blots for SP-B and SP-D

SDS-PAGE electrophoresis was done with 8 to 16% gradient Tris/ glycine gels (Novex, San Diego, CA). Samples containing 3 nmol of SatPC were electrophoresed along with ovine SP-B or murine SP-D standards. Proteins were transferred to nitrocellulose paper (Schleicher & Schuell, Keene, NH) for immunoblot analysis with rabbit anti- SP-B or anti-SP-D serum. The nitrocellulose membrane was first blocked with 5% bovine serum albumin (wt/vol) (Sigma Chemicals) in Tris-buffered saline, pH 7.4, containing 0.1% (vol/vol) Tween-20 (5). The antibody was diluted and incubated overnight. Horseradish peroxidase-conjugated goat antirabbit immunoglobulin (Calbiochem, La Jolla, CA) was used as secondary antibody, and enhanced chemiluminescence substrates (Amersham, Arlington Heights, IL) were used to develop the immunoblots. Estimates of the amount of SP-B and SP-D were made by calculation of the relative band densities, using Alpha-Imager 2000 Documentation and Analysis Software (Alpha Innotech, San Leandro, CA).

Cytokine Messenger RNA

Total RNA was isolated from a portion of the right lower lobe and from cell pellets of the tracheal aspirate and alveolar washings by guanidium thiocyanate-phenol-chloroform extraction (27). Ribonuclease (RNAse) protection assays were performed with total RNA from lung tissue and cell pellets (28). Briefly, RNA transcripts of ovine interleukin (IL)-1, IL-6, IL-8, IL-10, tumor necrosis factor (TNF)-, and ovine ribosomal protein L32 as a reference RNA were synthesized with [³²P]uridine triphosphate (Life Sciences Products, Boston, MA), using SP6 or T7 polymerase (Ribonuclease Protection Assay III; Ambion, Austin, TX). Aliquots of 10 µg RNA were incubated with excess radiolabeled probes for cytokines and L32 at 55° C for 18 h. Remaining single-stranded RNA was digested with RNAse A/RNAse T1 (Ambion). After inactivation and precipitation, protected fragments were electrophoresed on a 6% polyacrylamide-urea (8 mol/L) sequencing gel and visualized by autoradiography. Densities of the protected bands were quantified on a Phosphorimager (Molecular Dynamics Inc., Sunnyvale, CA), using ImageQuant software (Molecular Dynamics).

Surfactant Protein mRNA

The relative abundance of surfactant protein mRNA was measured with the S1 nuclease protection assay as previously described (27). Briefly, an excess of linearized probes for SP-A, SP-B, SP-C, and L32, all 5' end -labeled with [³²P], were hybridized at 56° C with 3 µg of total RNA from lung tissue. SP-D was detected in a separate hybridization using 10 µg RNA. The SP-D probe was a kind gift of Dr. M. Hallman of the University of Oulu, Finland. After incubation with S1 nuclease, the protected fragments were resolved on a 6% polyacrylamide-urea (8 mol/L) sequencing gel, visualized by autoradiography, and quantified on a Phosphorimager with ImageQuant software.

Lung Morphology/Inflammation Score

The right upper lobe was inflation fixed with 10% formalin at 30 cm H₂O pressure. Paraffin tissue sections of 5 µm thickness, stained with hematoxylin and eosin, were graded for the degree of inflammation in a blinded fashion by scoring three sections from each animal as zero if they had no inflammatory cells in tissue or airspaces, 1 if they had a few cells, 2 if they exhibited moderate cell infiltration, and 3 if they had a large number of inflammatory cells in airspaces and tissue (27). Airspaces and tissue were assessed separately. Average scores were calculated for each animal.

Data Analysis

Results are given as mean ± SEM. Comparisons between SP-A- treated lambs and SP-C-surfactant-treated animals were made with two-tailed t tests. Significance was accepted at p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lung Function

The seven lambs randomized to SP-C surfactant treatment weighed 3.4 ± 0.2 (mean ± SEM) kg, and the seven lambs randomized to SP-A + SP-C-surfactant treatment weighed 3.3 ± 0.3 kg. There were no differences between the groups in cord blood gas or pH values. Values of Cdyn were similar for the two groups of lambs at 15 min of age (Figure 1A). After surfactant treatment, Cdyn increased in both groups. A VT of 7.8 ± 0.2 ml/kg were used to maintain target PCO₂ values of 45 to 60 mm Hg for both groups. Over the 6-h period of ventilation, there were no significant differences in ventilation between the SP-A + SP-C-surfactant-treated group and the SP-C-surfactant-treated group, as evaluated with the ventilation efficiency index, a measure of the efficiency of gas exchange (Figure 1B). Sequential blood gas and pH values also were not different. Lung volumes for the deflation limbs of the pressure- volume curves of the two groups of animals were similar after 6 h of ventilation (Figure 1C). Therefore, the presence of SP-A had no effect on lung function.

View larger version (18K): [in this window] [in a new window]   Figure 1.   Sequential measurements of lung function in premature lambs treated with SP-C surfactant alone or with ovine SP-A (SP-A + SP-C surfactant). (A) Cdyn. Initial Cdyn was similar for both groups. After surfactant treatment, Cdyn increased in both groups. (B) The ventilation efficiency index (VEI) was similar in the two study groups over the ventilation period of 6 h. (C) Deflation limbs of pressure-volume curves. After 6 h of ventilation there were no significant differences in the maximal lung volumes of the SP-A + SP-C-surfactant-treated and the SP-C-surfactant-treated animals.

Surfactant Measurements

The amounts of SatPC in alveolar washes and in homogenized total lungs, measured after 6 h of ventilation, were not different in the SP-A + SP-C-surfactant and the SP-C-surfactant- treated groups (Table 1). The relative amounts of SP-B and SP-D were measured by Western blotting. There was no difference between SP-A + SP-C-surfactant-treated and SP-C-surfactant-treated animals after 6 h. The steady-state levels of mRNAs for SP-A, SP-B, SP-C, and SP-D were not affected by treatment with SP-A (Figure 2), although the increase in mRNA for SP-D in SP-A-treated animals approached significance (p = 0.07).

View larger version (28K): [in this window] [in a new window]   Figure 2.   (A) Representive S1 nuclease protection assays for SP-A, SP-B, SP-C, and SP-D mRNAs and ribosomal protein L32 mRNA. (B) Quantification for surfactant protein mRNA. Intensity of hybridization of mRNA was related to that of L32 mRNA a ribosomal protein mRNA for each reaction. SP-A treatment did not significantly change mRNAs of surfactant proteins.

Indicators of Inflammation

The total protein in alveolar washes was low in both study groups, and was not typical of lung injury (26) (Table 1). The total cell number in alveolar washes of the SP-A + SP-C-surfactant-treated group was increased by 4.9-fold above the cell numbers for SP-C-surfactant-treated animals (Figure 3A). The recruited cells consisted predominantly of neutrophils and macrophages/monocytes (Figure 3B). Hydrogen peroxide was measured in aliquots of 1 × 10⁶ cells, to evaluate metabolic activation. In cells from SP-A + SP-C-surfactant-treated lambs, the average hydrogen peroxide production was 15 µM per 10⁶ cells, as compared with 42 µM per 10⁶ cells in SP-C-surfactant-treated animals (Figure 3C). Stimulation with phorbol myristate acetate did not significantly increase hydrogen peroxide production. Therefore, cells from the alveolar washes of SP-A + SP-C-surfactant-treated animals produced significantly less hydrogen peroxide than did those from SP-C-surfactant-treated animals.

View larger version (19K): [in this window] [in a new window]   Figure 3.   Cell numbers in alveolar washes, differential counts, and hydrogen peroxide production. (A) The total number of nucleated cells recovered from alveolar washes per unit body weight was 4.9-fold greater in SP-A + SP-C-surfactant-treated than in SP-C-surfactant- treated animals. (B) Absolute number of lymphocytes, neutrophils, and monocytes in alveolar washes. The recruited cells in SP-A + SP-C-surfactant-treated animals consisted mainly of neutrophil granulocytes. The numbers of lymphocytes and monocytes recruited to the airway of SP-A + SP-C-surfactant-treated animals were slightly higher than for SP-C-surfactant-treated animals. (C) Quantification of hydrogen peroxide in alveolar washes. Hydrogen peroxide was measured in aliquots of 1 × 106 cells with and without stimulation with phorbol myristate acetate (PMA, 5 nM). Cells of SP-A-treated animals produced significantly less hydrogen peroxide both before and after stimulation with PMA than did cells of SP-C-treated animals. *p < 0.05 versus SP-C surfactant.

Cells from alveolar washes were analyzed for the expression of intercellular adhesion molecule-1 receptor (CD11b) and proteoglycan link protein (CD44), both of which molecules are involved in the vascular-to-tissue migration of neutrophils and monocytes to sites of inflammation (29). The cells from SP-A-treated animals expressed significantly higher levels of both proteins as measured by the percentages of CD11b- and CD44-positive cells (representive samples are shown in Figure 4A) and by mean fluorescence intensity (Figure 4B). The expression of endotoxin receptor CD14 was low and not different between the two study groups.

View larger version (26K): [in this window] [in a new window]   Figure 4.   Flow-cytometric analysis of cells from alveolar washes. Cells from SP-A + SP-C-surfactant-treated animals expressed greater amounts of CD11b and CD44. The expression of CD14 was low in both groups and unchanged after SP-A + SP-C surfactant treatment *p < 0.05 versus SP-C surfactant.

The mRNAs for IL-1, IL-6, IL-8, TNF-, and IL-10 were measured in total RNA from cells recovered from BALF at 2 h, from cells recovered from alveolar washes, and from total lung homogenates after 6 h of ventilation. These mRNAs were virtually undetectable in fetal lungs at this gestational age (16). Low levels of proinflammatory cytokine mRNAs were detected in cell pellets from BALF samples after 2 h of ventilation, and no differences were observed between the two study groups of animals (Figure 5A). After 6 h of ventilation, there were significant differences between SP-A + SP-C-surfactant-treated and SP-C-surfactant-treated lambs in terms of cytokine expression in cells from alveolar washes (Figure 5B). Levels of IL-1 and IL-8 mRNAs were significantly higher in SP-A + SP-C-surfactant-treated animals than in SP-C-surfactant-treated animals. In lung tissue, the staining intensities of IL-1, IL-6, and IL-8 mRNA relative to that of ribosomal protein L32 mRNA were higher in SP-A + SP-C-surfactant- treated animals than in the SP-C-surfactant-treated group (Figure 5C). The level of IL-10 mRNA did not differ in the groups, and TNF- mRNA was not detected in either group. The pattern of mRNA levels for IL-1, IL-6, and IL-8 in alveolar cells at 6 h differed from the pattern found in lung tissue at 6 h. The ratio of IL-1 mRNA in SP-A + SP-C-surfactant- treated animals to that in SP-C-surfactant-treated animals increased from 3.2 in tissue to about 6.0 in the cells from alveolar washes (Figures 5B and 5C). For IL-8, the ratio of tissue mRNA of SP-A + SP-C-surfactant-treated to that of SP-C-surfactant-treated animals was about 3.2, and this ratio rose to 8.5 for cells from the respective two groups' alveolar washes. The level of IL-6 mRNA was approximately 2.6-fold greater in the lung tissue of SP-A + SP-C-surfactant-treated animals than in that of SP-C-surfactant-treated animals. In contrast, in cells from alveolar washes the level of IL-6 mRNA was lower in SP-A + SP-C-surfactant-treated than in SP-C-surfactant- treated animals (ratio 0.5).

View larger version (31K): [in this window] [in a new window]   Figure 5.   Quantification of cytokine mRNAs. (A) Proinflammatory cytokine mRNA in cells recovered at 2 h from BALF. After 2 h of ventilation, there was only moderate expression of proinflammatory cytokine mRNAs, with no differences between the SP-A + SP-C-surfactant- treated and SP-C-surfactant-treated groups. (B) Proinflammatory cytokine mRNA levels in cells from alveolar washes after 6 h of ventilation. After 6 h there was a much higher level of proinflammatory cytokine mRNA in cells from the airways. In SP-A + SP-C-surfactant-treated animals, levels of IL-1 and IL-6 were significantly higher than in SP-C-surfactant-treated animals. In SP-C-surfactant-treated animals the level of IL-6 mRNA was significantly higher than in SP-A + SP-C-surfactant-treated animals. (C) Expression of mRNA for IL-1, IL-6, IL-8, IL-10, and TNF- in 10 µg of total lung RNA relative to mRNA for ribosomal protein L32. In SP-A + SP-C-surfactant-treated animals the level of proinflammatory cytokines IL-1, IL-6, and IL-8 was significantly higher than in SP-C-surfactant-treated animals. The level of IL-10 mRNA did not differ in the SP-A + SP-C-surfactant-treated and SP-C-surfactant-treated groups. TNF- was not detectable in either group *p < 0.05 versus SP-C surfactant.

Lung pathology. Representative sections of lung tissue illustrate the increased numbers of inflammatory cells in SP-A + SP-C-surfactant-treated animals but show no other signs of injury (Figure 6). Inflammatory cells were scored in lung tissue and airspaces (Figure 7). There were significantly higher scores in SP-A + SP-C-surfactant-treated animals.

View larger version (92K): [in this window] [in a new window]   Figure 6.   Histology of lungs. Upper panel shows a representive section from SP-C-surfactant-treated group. Lower panel shows a section from SP-A + SP-C-surfactant-treated group. The difference was the presence of more inflammatory cells in lung tissue and airspace.

View larger version (17K): [in this window] [in a new window]   Figure 7.   Inflammatory scores in airspace and lung tissue. Treatment with SP-A + SP-C surfactant increased the inflammatory score in both lung tissue and airspace *p < 0.05 versus control.

White blood cell count. Systemic effects of SP-A treatment were evaluated by white blood cell and differential counting of cells from cord blood and after 6 h of ventilation (Table 2). The cord blood had the typical lymphocytosis of fetal blood for both study groups. After 6 h of ventilation the number of leukocytes changed in both groups, without significant differences between the groups.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We treated preterm lambs with recombinant SP-C surfactant containing 3% ovine SP-A in order to evaluate the responses of the immature lung to SP-A treatment. This experiment was designed to initiate a proinflammatory response through a 15-min period of ventilation before surfactant treatment, so as to test whether SP-A would modulate the proinflammatory injury response of the preterm lung to ventilation. The presence of SP-A resulted in a 4.9-fold increase in cells in alveolar washes after 6 h of ventilation. The increased cells were predominantly granulocytes, expressing CD44 and CD11b adhesion molecules, that were recruited to the lungs. However, these granulocytes and the monocytes that accompanied them were not activated, because expression of the endotoxin receptor CD14 was not increased, and because hydrogen peroxide production was 3-fold lower for alveolar cells from SP-A + SP-C-surfactant-treated than for alveolar cells from SP-C-surfactant-treated lambs. Nevertheless, SP-A treatment was associated with increased mRNA levels for the proinflammatory cytokines IL-1, IL-6, and IL-8 in total lung, with no change in IL-10 or TNF- mRNA levels. SP-A treatment had no effect on lung function or measured concentrations of selected surfactant proteins, or on mRNA levels for the surfactant proteins. These results demonstrate unanticipated responses of the preterm lung to SP-A.

In our experiment, we used species-specific SP-A from healthy sheep. Butanol extraction was used for purification of the SP-A to maintain its protein function (21). On the basis of in vitro studies, SP-A has been labeled as either proinflammatory or antiinflammatory (11). Disparity in results with the protein may be partly explained by the use of SP-A from different sources and different purification methods. In some experiments, proinflammatory effects of SP-A were attributed to possible endotoxin contamination because SP-A binds endotoxin. Because the response of the immature lung to SP-A was predominantly proinflammatory, we carefully evaluated the possibility of endotoxin contamination of the SP-A used in our study. The SP-A tested negatively for endotoxin in the sensitive Limulus lysate assay, and we verified the earlier report that SP-A did not mask endotoxin in the assay (10). We treated three other lambs with the highest amount of Escherichia coli O55/B5 endotoxin that could be present in SP-A without being detected by the Limulus lysate assay. Lambs were also treated with the same SP-C surfactant as used in our study and which contained 0.1 ng/kg E. coli O55/B5 endotoxin. After 6 h of ventilation, results for most variables were similar to those for SP-C-surfactant-treated animals: the total number of cells per unit weight in alveolar washes was 1.4 ± 0.4 × 10⁷ cells/kg after endotoxin treatment, in comparison to 1.6 ± 0.3 × 10⁷ cells/kg in SP-C-surfactant-treated animals. The numbers of granulocytes were similar, at 6.5 ± 1.0 × 10⁶/kg after endotoxin treatment versus 6.8 ± 2.0 × 10⁶/kg in SP-C-surfactant-treated animals. The level of CD11b was 187 ± 33 mean fluorescence units (FU) in endotoxin-treated animals, versus 128 ± 43 FU in SP-C-surfactant-treated animals, and that of CD44 was 112 ± 21 FU in endotoxin-treated animals, versus 89 ± 16 FU in SP-C-surfactant-treated animals. Hydrogen peroxide production was similar, at 45 ± 4 µM per 1 × 10⁶ cells for endotoxin-treated animals, versus 42 ± 9 µM per 1 × 10⁶ cells for SP-C-surfactant-treated animals. However, the endotoxin receptor CD14 was increased on cells in the alveolar washes of endotoxin-treated animals, at 92 ± 11 FU, versus 59 ± 7 FU in SP-C-surfactant-treated animals (p < 0.05), and the mRNA level for prinflammatory cytokines was 10-fold higher in the total lung homogenates of endotoxin treated animals than in those of SP-C-surfactant-treated animals. The responses of the preterm lung to treatment with SP-A were distinct from the responses to endotoxin.

In adult animals, SP-A functions as a potent host-defense molecule and modulator of the inflammatory responses to pathogens. Mice that are SP-A deficient are unable to successfully phagocytose and kill bacteria and viruses (6, 7). The lungs become more inflamed in response to group B steptococcus, with greater recruitment of activated granulocytes that produce higher levels of proinflammatory cytokines, including TNF-, than is the case for the lungs of normal SP-A-sufficient mice (6). Treatment of adult SP-A-deficient mice with SP-A corrects the phagocytosis and killing defects and modulates the proinflammatory cytokine response (7).

In vitro studies have given conflicting results for the pro- and antiinflammatory properties of SP-A. Kremlev and colleagues found proinflammatory cytokine production after incubation of a monocytic cell line with SP-A (13, 30). Blau and coworkers reported that both SP-A and endotoxin stimulated the production of nitrite by AM (12). On the other hand, McIntosh and associates reported a downregulation of production of the proinflammatory cytokine TNF- by SP-A (14), and Borron and colleagues reported that SP-A had antiinflammatory effects on lymphocytes (31). In all of these experiments, mature AM and monocytic and lymphotic cell lines were used. There is no information about interactions of SP-A in-vitro with immature cells from animals.

On the basis of the effects of SP-A on mature macrophages and granulocytes, and the in vivo results with SP-A-deficient mice, we anticipated that SP-A treatment would suppress the granulocyte recruitment and decrease the proinflammatory cytokine transcription that accompanies mechanical ventilation after preterm birth. We chose to measure the cytokines IL-1, IL-6, IL-8, IL-10, and TNF- because these cytokines increase in SP-A-deficient mice (7), increase in the airways of infants developing bronchopulmonary dysplasia (BPD) (1), are increased in the amniotic fluid with chorioamnionitis (17), and contribute to lung injury in adult-lung models of lung injury (32). We found that SP-A induced granulocyte recruitment and increased the mRNAs of IL-1, IL-6, and IL-8, but did not increase the mRNA of IL-10 or TNF-. The hydrogen peroxide assay indicated that the granulocytes in alveolar washes were not activated. This unusual pattern of response is probably explained by the immaturity of the host-defense response of the preterm newborn (33). Granulocytes in term and preterm newborns have deficiencies in adherence, deformability, and locomotion in comparison with mature granulocytes (33), and chemotaxis of granulocytes from preterm infants to sites of inflammation is impaired irrespective of gestational age (34). Phagocytosis of opsonized particles is normal in term newborns, but the respiratory burst may be impaired (33). In most types of immunologic cell the second-messenger responses to exogenous stimuli are different from the responses measured in adult cells, partly explaining the immaturity of the host-defense response (20, 33). The present study suggests that SP-A is a potent neutrophil chemoattractant in the preterm lung. In our study a high dose of SP-A was locally administered to an immature lung with an immature immune system, in which other immunomodulatory molecules are present in low concentrations (33). SP-A functioned in the immature immune system as a potent neutrophil chemoattractant, with upregulation of the cell-surface molecules CD11b and CD44, which are involved in the recruitment of cells to sites of inflammation. The chemoatraction of neutrophils (35) and AM (36) by SP-A has previously been described in vitro. The present study shows that in vivo, SP-A is a potent neutrophil chemoattractant that efficiently recruits immature neutrophils and partially overcomes deficiencies of immature granulocytes. Our study evaluated a unique situation: the initiation of ventilation in a preterm, surfactant-deficient lung, and there is no comparable information for full-term or adult lungs.

The simplest interpretation of the findings in our study is that SP-A treatment of the preterm lung is predominantly proinflammatory and therefore harmful. However, this interpretation may be incorrect. The preterm lung has major defects in both immune and innate host defenses (1, 18). SP-A may promote host defenses as a chemoattractant by recruiting unactivated granulocytes to an alveolar environment without significant numbers of mature macrophages. Although low levels of SP-A were correlated with subsequent lung injury in preterm infants (18), there is no information about whether or how exogenous SP-A alters host defenses in the preterm newborn, or how SP-A might improve outcomes. Although supplementation of surfactant with SP-A can improve the lung function of preterm animals to a greater extent than can surfactant that lacks SP-A, these effects are modest and are not apparent in animals ventilated with sufficient PEEP (25). SP-A also can minimize inhibition of surfactant by proteinaceous pulmonary edema, but this effect will be of benefit only after severe lung injury has occurred, and injury in general suppresses SP-A synthesis (18, 37). Treatment with SP-A in our study was not associated with decreased mRNA levels for any of the surfactant proteins investigated, and TNF- mRNA did not increase.

The function of SP-A as a neutrophil chemoattractant in a preterm human ventilated lung is of particular interest with regard to ventilation of preterm human newborns. In these patients bacterial infections of the airways have been associated with BPD (38). Therefore, adding SP-A to surfactant treatment may overcome deficiencies in recruitment and adhesion of neutrophil granulocytes in clearing bacteria from the airways in preterm infants. SP-A treatment of the preterm lung may be beneficial because of induction of protective responses.