INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pulmonary surfactant, a mixture of phospholipids and surfactant-specific proteins, is synthesized in type II alveolar pneumocytes. It is well recognized that surfactant has a significant role in maintaining openness of pulmonary alveoli. It was initially acknowledged by Macklem and colleagues (1) that surfactant might also be important for maintenance of patency in narrow conducting airways. When they measured the pressure required to open terminal airways closed by liquid, they found that it could be lowered by the detergent Tween, but not by pulmonary surfactant. However, following more recent studies, it is now appreciated that pulmonary surfactant, as suspected by Macklem and colleagues, has the important role of maintaining patency in narrow conducting airways (2).

Airway inflammation is a key feature of asthma that contributes to disease severity and symptoms. Surfactant normally prevents moisture from accumulating in the airways' most narrow sections and in this way maintains patency of conducting airways (4, 5). Airway inflammation is associated with leakage of plasma proteins into the airway lumen (6), and the proteins will inhibit the normal surfactant function (12). Airway inflammation, which is an integral component of asthma, can conceivably be associated with surfactant dysfunction, and thus cause a certain proportion of the terminal conducting airways to become blocked by moisture accumulating in the most narrow sections. Airway resistance would increase because of this enhanced blockage of the airway. The surfactant dysfunction would thus augment the airway resistance caused by an inflammatory swelling of the airway wall and by smooth muscle contraction. Pulmonary surfactant is furthermore of interest in cases of asthma because it has an immune modulatory effect. It reduces the activation of nuclear factor kappa B (NF-B) that is critical for the transcription of many cytokines (16).

Pulmonary surfactant function in narrow conducting airways can be evaluated with a capillary surfactometer (CS) (4, 5). This instrument, briefly described under METHODS, shows how a surfactant preparation is capable of maintaining patency during a 2-min period of airflow through a glass capillary.

Antigen challenge in atopic subjects produces an airway inflammation that is particularly intense when the antigen is administered via a local segmental route (17). This report investigates how the surfactant ability to maintain patency was affected by a segmental antigen challenge in atopic subjects. With a technique previously described (17), bronchoscopy was performed and bronchoalveolar lavage (BAL) was carried out twice, 5 min and then 48 h after segmental bronchoprovocation (SBP) with antigen in one segment and with saline in another. BAL fluids (BALF) were tested for their surfactant function with the CS. A control group of healthy volunteers underwent bronchoscopy and BAL without challenge. BALF samples from the control group were also evaluated for cell count, differential and were studied with the CS.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Twenty allergic patients were recruited to participate in this study. All had a positive prick skin test to house dust mite, ragweed, or cat, as well as a history of nasal congestion and sneezing when exposed to these aeroallergens. Some of these subjects had occasional episodes of wheezing with exposure to these antigens and they rarely used inhaled -agonists. However, all of them had normal lung function tests and none of them were regularly using asthma medications. The subjects were stable at the time of the study and none of them smoked or had been affected by a respiratory infection during the month preceding enrollment. Six healthy volunteers were also enrolled. They had no history of respiratory symptoms and skin tests to common aeroallergens were negative. The study was approved by the University of Wisconsin Committee for Human Subjects. All participants gave written informed consent upon entering the study.

Study Design

After evaluation of medical history, physical examination, spirometry, and prick skin tests to common aeroallergens, the subjects returned for histamine and antigen bronchoprovocation on two separate days. Pulmonary functions were calculated from a flow-volume loop (Med-Graphic, St. Paul, MN). The best of three acceptable trials was used to obtain the FEV₁ and FVC. To establish antigen responsiveness and determine antigen dose for the SBP, each subject underwent a graded nebulized antigen challenge performed as previously described (18). Briefly, a baseline spirometry value was determined and was redetermined after 10 min, after five breaths of the diluent solution. If there was no significant response to the diluent challenge (FEV₁ remained within 10% of baseline), five breaths of an endotoxin-free antigen extract was given and spirometry was repeated after 10 min. Consecutively higher antigen doses were given until the FEV₁ fell by > 20% from the postdiluent value. Antigen PD₂₀ (APD₂₀) was calculated by linear interpolation of the dose-response curve using a computer program (PD20; Madison Scientific Software, Wexford, PA) and expressed in cumulative breath units (CBU).

We also determined nonspecific airway response to histamine challenge. It was conducted in a similar fashion (18) to antigen challenge, except that the interval between each consecutive dose was 5 min. The provocative histamine concentration leading to 20% fall in FEV₁ (PC₂₀) was calculated and expressed in a manner similar to that used for the APD₂₀. Subjects who did not show a 20% fall in FEV₁ after the final dose of histamine (25 mg/ml) were assigned a PC₂₀ of 25 mg/ml for purpose of data analysis.

One month later, two bronchoscopies that were 2 d apart were performed in each of the allergic subjects. At the first bronchoscopy, a bronchopulmonary segment was identified and SBP with 0.9% NaCl solution was performed. This served as the control segment. BAL was performed 5 min after the SBP with saline. For BAL three or six 40-ml aliquots of warm (37° C) solution of 0.9% NaCl were instilled and the liquid was recovered sequentially, using gentle suction. In a second segment SBP with relevant antigen was carried out using 10% of the APD₂₀, and BAL performed 5 min after the antigen in identical fashion to that used in the saline-challenged segment. A second bronchoscopy was done 48 h later when the two challenged segments were identified and again lavaged. For each subject the same volume of 0.9% NaCl was used for the BAL in each of the two segments. The BALF recovery averaged 75% from all segments.

BALF Analysis

BALF from each segment was pooled after passing through gauze and centrifuged (400 × g for 10 min) to sediment BALF cells. The supernatant was removed and kept at 70° C for later analysis. Aliquots of the supernatants were shipped in frozen condition to Buffalo, New York, for blind analysis with the CS. The cell pellet was washed in phosphate-buffered saline (PBS) and suspended in Hanks' balanced salt solution. Cells were counted with a hemocytometer and cell suspensions were adjusted to a final concentration of 2 × 10⁶ cells/ml. Cytocentrifuge slides were air dried, fixed in methanol, and stained (Diff-Quik; Scientific Products, Chicago, IL). For the differential cell counts we categorized 300 cells as either macrophages, eosinophils, lymphocytes, neutrophils, or epithelial cells on the basis of staining characteristics and morphologic criteria (19, 20). Total proteins in the BALF were determined with the method of Lowry and coworkers (21).

Pulmonary Surfactant Study

The particular ability of BALF to maintain airway patency was tested with a CS. BALF surfactant had been collected with a technique that caused it to be so diluted that it had to be concentrated 25 times before it could be examined with the CS. The required concentration was obtained five times by centrifugation and five times by evaporation. The cell-free BALF was first centrifuged for 1 h at 40,000 × g at 4° C. Supernatant, corresponding to 80% of the centrifuged fluid, was removed. The remaining 20%, generally 100 µl, containing the large aggregate surfactant, was exposed to vacuum (80 kPa) in a small plastic test tube. Through a very fine glass capillary a stream of dry nitrogen was sucked into the tube without noticeably reducing the vacuum. In 20 to 30 min the liquid was desiccated and 20 µl of the supernatant previously obtained was added. The mixture was carefully stirred before it was evaluated with the CS.

The principle of the CS has previously been described in detail (4, 5) and the instrument has been used for an evaluation of BALF from animals (22). The instrument has a glass capillary, with a narrow section where the width (interior diameter 0.3 mm) is simulating a terminal conducting human airway (27). In this narrow section of the capillary, 0.5 µl of the liquid sample to be evaluated is deposited. The liquid blocks the capillary lumen, but as pressure is raised at one end of the capillary, where it is continuously measured, the liquid is extruded. If the sample contains well-functioning pulmonary surfactant, it will not return to the narrow section and the steady airflow, which is maintained for 120 s after the liquid was first extruded, will not meet resistance and for that reason a pressure of zero is continuously recorded. If, on the other hand, the surfactant is too diluted or is not functioning well, the liquid may return to the narrow section one or more times. As soon as a blocking liquid column is formed, the recorded pressure is raised while the liquid is being extruded. A microcomputer calculates the percentage of the 120-s study period that the capillary is open to a free airflow. For each sample evaluated a new capillary is used. Figure 1 shows an evaluation with the CS of a well-functioning surfactant preparation at a phospholipid concentration of 1 mg/ml. The figure also shows the moderate dysfunction caused by the addition of a large quantity of albumin, a relatively weak surfactant inhibitor. The percentage of the study time, 120 s, that the capillary remained open for a free airflow was printed on the recording paper (Figure 1). A value used for this study was the mean of five assays of a sample. After the evaluation with the CS, the remains of the concentrated BALF and the supernatant were frozen for further studies.

View larger version (57K): [in this window] [in a new window]   Figure 1.   CLSE at a concentration of 1 mg/ml will keep the capillary open during the 120-s study period. The images of the capillary, showing how the 0.5-µl surfactant sample is being pushed to the right, are correlated to the pressure tracing, as indicated by the arrows. The amphipathic surfactant molecules form a film at the air-liquid interface of the large meniscus. It develops a high surface pressure as it is compressed, causing the meniscus to become flat. Once the sample has been extruded from the narrow section of the capillary it does not return. Therefore, the flow of air through the capillary is meeting no resistance; pressure remains at zero. The lower tracing is showing the effect of adding a large quantity of albumin to the CLSE. This causes the surfactant to be partly inhibited. As a result the large meniscus is hemispherical and the liquid returns to the narrow section three times, reducing the percentage of time when pressure zero is recorded.

A reduced patency was most often caused by water-soluble inhibitors and for that reason an attempt was made to remove them by washing the BALF surfactant. This was done by adding a large volume of saline solution to the concentrated surfactant. After careful stirring, the mixture was once again centrifuged for 1 h at 40,000 × g and 4° C. A volume of supernatant equal to the amount of saline solution that had previously been added was removed and, after vigorous stirring, the sample was reevaluated with the CS.

Another way of determining if the supernatant contained inhibitors, was to use it for a dilution of calf lung surfactant extract (CLSE, kindly donated by ONY Inc., Amherst, NY). The original concentration of CLSE is 35 mg/ml and for these experiments it was diluted to 1 mg/ml. The diluted liquid sample was then evaluated in the CS. If saline solution was used for such a dilution, CLSE maintained an "openness" of 100%. A reduced capacity to maintain patency demonstrated the action of inhibitors.

Data and Statistical Analysis

The data were generally not normally distributed, and for this reason were expressed as medians with 25 to 75% interquartiles. For analysis of BALF surfactant and cellular data, analysis of variance (ANOVA) on ranks (Kruskal-Wallis) was used with an appropriate post hoc testing (Dunnett's method). A nonparametric t test (Mann-Whitney) was used for an assessment of whether the BAL fluid's value of "percent open" had been altered in a statistically significant way when the fluid originated from a lung segment that had been challenged 48 h previously with antigen or with saline solution, respectively. The correlation between surfactant function and cellular data or protein concentration was evaluated with a Spearman rank test. A p value less than 0.05 was considered to show a significant difference.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Response to BAL

The atopic patients had normal pulmonary function at baseline; however, compared with the healthy volunteers, they showed increased nonspecific bronchial hyperresponsiveness (Table 1). The procedures were tolerated well and generated no change in pulmonary function (data not shown).

BAL Characteristics

BAL yielded a median volume return of 74 to 80% in atopic subjects (Table 2). As in previous studies (17, 19, 20), antigen SBP was associated with a significant cellular inflammation 48 h after the challenge. This included a roughly threefold increase in the number of BAL cells, an 80-fold increase in the proportion of and total number of eosinophils in the BALF, a reduction in the proportion of alveolar macrophages, but no change in the proportion of lymphocytes. A minimal, but statistically significant, increase in the percentage of BAL neutrophils and a reduction in epithelial cells were seen in saline as well as in antigen-challenged segments. No significant changes were noted in the number of epithelial cells. Total BALF proteins increased threefold after antigen challenge and they doubled after saline challenge (Table 2, Figure 2).

View larger version (13K): [in this window] [in a new window]   View larger version (11K): [in this window] [in a new window]   Figure 2.   Surfactant function (A) and total proteins (B) observed in the BALF of normal volunteers (NV) and in the fluid from atopic subjects challenged with saline (S5 and S48) and with antigen (A5 and A48). S5 and A5 were obtained after 5 min, and S48 and A48 not until 48 h after saline and antigen challenge, respectively. Box plots indicate medians with the 25th and the 75th percentiles. The error bars depict the 10th and the 90th percentiles. *p < 0.05 compared with S5 and A5 (Dunnett's method).

Surfactant function in concentrated BALF. Five minutes after the challenge, the ability of the BALF to maintain patency was similar in saline and in antigen-challenged segments. Forty-eight hours later, the surfactant function in the saline-challenged segment had not changed, but in the antigen-challenged segments the value of percent open was significantly lower (p = 0.0385). Interestingly, in the BALF obtained 48 h after antigen challenge and concentrated 25 times, there was a significant negative correlation between total proteins and surfactant function expressed as percent open (Figure 3A). Furthermore, a similar but less striking correlation was seen between the number of eosinophils recruited to the airway and the value of percent open (Figure 3B). Airway function, expressed as FEV₁ percent predicted, FEV₁/FVC ratio, or FEF_25-75% predicted, did not show a significant correlation with pulmonary surfactant function.

View larger version (18K): [in this window] [in a new window]   Figure 3.   (A) The BALF obtained 48 h after antigen challenge showed a negative correlation between total proteins and surfactant function expressed as percent open (r = 0.62, p = 0.005). (B) There was also a negative correlation between total eosinophils and surfactant function (r = 0.46, p < 0.05).

Effect of water-soluble inhibitors. Figure 4 shows the effect of "washing" the surfactant. With few exceptions, the removal of water-soluble inhibitors restored a normal surfactant function of the concentrated BALF. This showed that the BALF, concentrated 25 times, had an adequate surfactant concentration and functioned normally, if only the inhibiting agents were removed. The action of those inhibitors was also unveiled when the supernatant from the high-speed centrifugation was used for a dilution of CLSE from 35 to 1 mg/ml (Figure 4). However, when the supernatant originated from BALF of healthy volunteers or from the fluid obtained immediately after the challenge, with saline or with antigen, the percent open value was optimal.

View larger version (13K): [in this window] [in a new window]   View larger version (14K): [in this window] [in a new window]   Figure 4.   The median, the 25th, and the 75th percentile of percent open exerted by the concentrated BALF 5 min (A) and 48 h (B) after antigen challenge (boxes to the left). The boxes in the middle show that after the "washing" procedure the surfactant in BALF functioned almost perfectly. The boxes to the right demonstrate that when supernatant of BALF obtained 5 min after the challenge was used to dilute CLSE, inhibitors were hardly detectable, but BALF obtained 48 h after antigen challenge caused significant CLSE dysfunction. The error bars depict the 10th and the 90th percentiles. The asterisks in (A) indicate a p value of < 0.005 versus BALF (Mann-Whitney rank sum test). The asterisks in (B) indicate a p value of 0.005 versus BALF, and the + indicates the same p value versus washed BALF (Mann-Whitney rank sum test).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study we evaluated the pulmonary surfactant function of atopic individuals with and without mild asthma. We compared the effect of a challenge with antigen in one lung segment with that of a saline challenge in another. The surfactant property that was specifically evaluated was its capability to maintain airway patency, and this ability was clearly hampered in the airway challenged with antigen. As expected, antigen induced a significant cellular inflammation in the airway with a predominant increase in the number of eosinophils. Furthermore, the antigen challenge was associated with a significant increase in the BALF protein levels. The inhibiting action of plasma proteins on surfactant function is well known (12) and might very well have been the major reason for the surfactant dysfunction seen in our subjects. When these proteins, together with other water-soluble inhibitors, were removed, the surfactant function returned to normal in almost every instance (Figure 4). Quite likely the inhibitors were the plasma proteins that had leaked into the airway in association with antigen-induced airway inflammation. Plasma proteins have a varying potency as inhibitors of surfactant. Some, such as albumin, which is likely to be the major airway invader, are relatively weak inhibitors whereas others, such as fibrinogen, are potent (13). It is therefore very appealing to analyze the invading proteins and identify the data of the most inhibiting components.

Animal studies support the concept that pulmonary surfactant plays an important role in the maintenance of patency in small airways (22, 28). They show that with the inflammatory changes that develop after a respiratory syncytial virus (RSV) infection (23, 24) an ozone exposure (25, 26), or an antigen challenge (28), surfactant loses some of its ability to promote patency, a failure that might promote increased airway resistance. Our study supports the hypothesis that this scenario is also true in the human lung. We therefore propose that a disturbance in surfactant function might be an important mechanism contributing to the airway obstruction observed during an allergen-induced asthma attack.

The primary reason for the surfactant dysfunction is probably the inflammatory reaction, with associated leakage of plasma proteins into the airway lumen. Those proteins will have a varying degree of surfactant-inhibiting effect. However, the apoprotein, surfactant protein A (SP-A), will partly protect surfactant from this inhibition (32). An analysis of the SP-A concentration in patients with asthma reported by van de Graaf and colleagues (33) showed that the SP-A concentration was less in asthmatics than in control subjects.

Airway inflammation with infiltrating cells, edema, secretions, edematous swelling of the airway wall, and contraction of smooth muscles have all been proposed as potential mechanisms of airway obstruction during an asthma exacerbation. However, several recent studies suggest that a disturbed surfactant function will also contribute to this process (22-26, 28- 31). In a study by Kurashima and colleagues (34), the surface activity of sputum collected serially from patients during asthma attacks was compared with that of normal subjects and patients with stable asthma. There were no significant differences in minimal surface tension, phospholipid content, and total protein levels between sputum from normal subjects and those with stable asthma. However, during the acute attack, minimal surface tension, total protein, and total protein/phospholipid ratio all increased. These parameters improved during recovery from the asthma exacerbation. The investigators concluded that the deteriorating surface properties of airway fluid contribute to the early phase of an asthma attack and might be involved in the pathogenesis of airway obstruction in asthma. A clinical pilot study showed that inhalation of aerosolized surfactant during an acute asthma attack had a beneficial effect on FEV₁ and arterial oxygen tension (35). It has been demonstrated that when the surfactant function is inhibited by plasma proteins it can be restored by increasing the surfactant concentration (5). This makes it appealing to seek to restore a surfactant function, inhibited by invading proteins, by supplying well-functioning pulmonary surfactant.

Although our examinations showed that the airway challenged with antigen resulted in surfactant dysfunction, no significant abnormalities were detected with spirometry performed in these patients after the challenge. However, only a small portion of the lung airway had been affected by the segmental challenge, therefore, even significant changes in the challenged segments can be missed with spirometry. In future studies, evaluating segmental airway resistance, in vivo, after segmental challenge, would be of great interest.

The BALF had to be concentrated before it could be evaluated with the CS. To obtain the required 25-fold concentration of the surfactant we used two methods: centrifugation and evaporation. The first method concentrated large surface active aggregates but not proteins and other inhibiting agents. To increase their concentration six times we completely dehydrated 100 µl of the samples concentrated by centrifugation and then added 20 µl of the supernatant that contained the inhibiting agents. With this procedure that increased the surfactant concentration 25 times and the concentration of water-soluble surfactant inhibitors six times, we were able to see the effect of the inhibitors in the BALF obtained 48 h after antigen challenge. After the "washing" procedure the inhibiting effect disappeared almost completely, which demonstrated that the concentration of surfactant was adequate, but its activity was inhibited by a water-soluble agent.

The presence of these inhibitors was also confirmed when the BALF supernatant was used for a dilution of CLSE. When it came from BALF obtained 48 h after antigen SBP, it gave a value of percent open that was significantly lower than when it came from BALF obtained only 5 min after the challenge or if the segment was challenged with saline.

In summary, our study has shown that when an atopic individual is challenged with an appropriate antigen, airway inflammation, which is associated with a surfactant dysfunction, will develop. This dysfunction, which was related to water-soluble inhibitors, can reduce the surfactant ability to maintain airway patency and cause a certain number of terminal conducting airways to become blocked by liquid columns. This could be one factor contributing to the increase in airway resistance that develops during an asthma attack.