INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Meconium aspiration is a common cause of respiratory distress in term and post-term infants. Meconium causes an acute decrease in gas exchange and dynamic lung compliance (1, 2), associated with alveolar edema and pulmonary hypertension (3). Together with mechanical obstruction (4) and chemical pneumonitis (5), inactivation of surfactant (6) has been suggested as a pathophysiologic mechanism of meconium aspiration syndrome (MAS). Meconium inhibits pulmonary surfactant in vitro (2, 6), in situ in excised canine lung lobes (7), and inhibition of endogenous surfactant has been demonstrated in experimental MAS in piglets (1). Furthermore, albumin and protein in tracheal aspirate are increased in piglets aspirating meconium (1). Albumin, hemoglobin, and fibrinogen inhibit surfactant in vitro (8, 9). Thus, in MAS, surfactant is inhibited not only by meconium but also secondarily by plasma proteins leaking into the airways. Similar mechanisms with serum inhibition of surfactant might play a role in respiratory distress in infants (10) and in adults (11, 12).

In vitro studies have shown that the inhibitory effect of meconium on surfactant may be reduced by increasing doses of surfactant (6). Treatment with surfactant improves ventilation efficiency in experimental models of MAS in rats (13) and in newborn rabbits (14). In a randomized controlled study, Findlay and coworkers demonstrated that repeated doses of surfactant to term infants with MAS improved oxygenation, and decreased severity of pulmonary morbidity and hospitalization time (15). The improvement in oxygenation was far more prominent after the second and third dose of surfactant, in comparison with the moderate effect of the first dose. This is in accordance with the in vitro findings that high doses of surfactant are required to overcome inhibition by meconium (6).

The mechanisms of meconium inhibition of surfactant are unknown. It has been shown that meconium destroys the fibrillary structure of surfactant (16, 17), and that meconium inhibition is reversed by nonionic polymers (18).

We have previously found that meconium could exert prooxidant action in the presence of iron (19). Oxidative stress and free radicals have been reported to destroy surfactant function (20), so we hypothesized that iron would aggravate meconium-dependent inhibition of surfactant. However, we found that FeCl, abolished the inhibitory effect of meconium on surfactant under in vitro conditions. In the present report, we describe this unexpected observation together with additional experiments clarifying possible mechanisms involved.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Materials

First-passed meconium was collected from 20 healthy term infants in the nursery. The meconium was pooled, mechanically dissolved, lyophilized, frozen in aliquots at 20° C, thawed in room temperature, and suspended in normal saline (110 mg/ml). Water used in buffers or as solute was deionized. Modified bovine lung surfactant, supplemented with dipalmitoylphosphatidylcholine (DPPL), palmitic acid, and tripalmitin (Beractant [Survanta]; Abbott Laboratories IL, Wiesbaden, Germany), modified porcine lung surfactant (Curosurf; Chiesi Pharmaceutici, Parma, Italy), and modified bovine lung lavage surfactant extract (Alveofact; Boehringer Ingelheim Pharma KG, Biberach, Germany) were kindly provided by the manufacturers. FeCl₃, FeSO₄, and deferoxamine mesylate were purchased from Sigma Chemical Co. (St. Louis, MO). Magnesium sulfate (MgSO₄) (Ferak, Berlin, Germany), and CuCl₂ (Merck AG, Darmstadt, Germany) were also used.

Pulsating Bubble Measurements

Alveofact, Curosurf, and Survanta were suspended in 0.9% NaCl, at a concentration of 5 mg/ml (21) unless otherwise stated, with or without FeCl₃ (final concentration 0.3 mM) and meconium (2.75 mg/ml). The mixtures were incubated at 37° C for 10 min before surface tension measurements. Surface properties were determined with a Pulsating Bubble Surfactometer (PBS; Electronetics Corporation, Buffalo, NY). Approximately 20 µl of sample was filled in a plastic chamber and a bubble was created. The surface area of the bubble was compressed by 50% between 0.4 and 0.55 mm at a frequency of 40 cycles/min, and the contractile force at the air-liquid interface was determined. Unless otherwise stated, surface tensions at minimum and maximum bubble size (_min and _max) after 5 min of pulsation are given. Surface tension measurements with the PBS were also performed on 2.5 mg/ml of Curosurf with or without 5% ethylenediaminetetraacetic acid (EDTA)-plasma and with or without FeCl₃.

pH Measurements

pH was measured with a PHM 80 portable pH meter (Radiometer Analytical A/S, Copenhagen, Denmark). The pH meter was calibrated immediately before use. pH was measured in at least two samples of surfactant from different vials, and average values are presented.

Image Analysis of Bubble Size and Number

Curosurf was suspended in normal saline at a concentration of 1 mg/ml with or without 0.55 mg/ml of meconium, with or without 0.06 mM FeCl₃. The samples were vortexed under standardized conditions to generate bubbles (22). One drop of the sample was put on a well slide and examined by light microscopy at a magnification of ×10. The microscope was connected to a VIDAS image analyzer (Kontron, Munich, Germany) via a TV camera. A specially designed program was used to measure the maximum and minimum diameters and to calculate the equivalent circle diameter of the bubbles. For each sample, 5 random fields in each of 3 wells were examined, corresponding to a total area of 4.35 cm². Proportion of microbubbles (diameter < 20 µm) was calculated as an average of the 15 areas. Mean and standard deviation for three independent samples are given. At a concentration of 1 mg/ml, the proportion of microbubbles in Curosurf is more than 90% in the absence of inhibitory material, yet the surfactant is sensitive to inactivation by, for example, albumin at serum concentration (P. Berggren and coworkers, unpublished data).

Measurements of Surface Spreading with Wilhelmy Balance

Curosurf (40 mg/ml) was incubated with meconium (14 mg/ml) and spreading rates of surfactant at 37° C were determined with a modified Wilhelmy balance (Biegler Electronic, Mauerbac, Austria) (23). For these measurements, the trough of the balance was filled with normal saline. The surface area of the saline hypophase was 69 cm². After calibration, a droplet (50 µl) of the test sample, diluted to a phospholipid concentration of 40 mg/ml, with or without 1.25 mM FeCl₃, was applied onto the surface, approximately 3 cm from the dipping platinum plate. Surface tension, reflected by the traction force on the dipping plate, was recorded continuously until an equilibrium state had been established. Spreading rate was defined as the time interval between application of the sample and the moment when the surface tension had dropped to a level of 30 mN/m.

Statistics

Values are given as mean with standard deviation. Comparisons of samples were performed with one-way analysis of variance (ANOVA), with Bonferroni correction. A repeated measure ANOVA analysis with Scheffe's post hoc test was used for comparison of the groups during the 200 cycles. Two-tailed p values are given. Analyses were performed with a GraphPad INSTAT tm/PC statistical package (San Diego, CA). For the repeated measurements, ANOVA StatView for Windows was used (SAS Institute Inc., Cary, NC).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pulsating Bubble Measurements

Meconium (2.75 mg/ml) caused significant (p < 0.001) inhibition of Curosurf and Survanta, but not of Alveofact at a phospholipid concentration of 5 mg/ml (Figure 1). Increasing the concentration of phospholipids to 10 mg/ml abolished inhibition of Curosurf by 2.75 mg/ml of meconium (_min: 3.8 ± 1.0 mN/m and _max: 35 ± 1.5 mN/m, n = 4).

View larger version (25K): [in this window] [in a new window]   Figure 1.   Surface tension at minimum (upper panel ) and maximum (lower panel ) bubble size measured with a Pulsating Bubble Surfactometer. 5 mg/ml of surfactant (Alv, Alveofact; Cu, Curosurf; Su, Survanta) was incubated with 2.75 mg/ml of meconium for 10 min before measurement. 0.3 mM FeCl3 was added to samples labeled "fe". *p < 0.05 versus surfactant without meconium and versus surfactant/ meconium/FeCl3.

Addition of 0.3 mM FeCl₃ completely reversed the inhibition exerted by meconium (Figure 1). Figure 2A demonstrates the dose-dependent effect of FeCl₃ on meconium inhibition of 5 mg/ml of Curosurf. Lowering the concentration of Alveofact to 2.5 mg/ml made this preparation susceptible to inhibition by meconium too, increasing _min from 6.2 ± 2.2 mN/m (n = 5) to 27 ± 11 mN/m (n = 5). Also this inhibition could be counteracted by FeCl₃ (_min: 5.2 ± 1.1 mN/m, n = 5).

View larger version (21K): [in this window] [in a new window]   Figure 2.   Surface tension at minimum (white bars) and maximum (black bars) bubble size measured with a Pulsating Bubble Surfactometer. 5 mg/ml of Curosurf was incubated with 2.75 mg/ml of meconium, and increasing amounts of FeCl3 (A) or acetic acid (B) were added 10 min before the measurements. -mec, Curosurf without meconium.

FeCl₃, without meconium, did not significantly alter _min of Survanta or Curosurf (this was not tested for Alveofact). Comparison of Curosurf, Curosurf/FeCl₃, and Curosurf/meconium/FeCl₃ from the fifth to the 200th cycle revealed no differences in _min between the groups at any time point (Figure 3).

View larger version (16K): [in this window] [in a new window]   Figure 3.   Surface tension at minimum bubble size measured with a Pulsating Bubble Surfactometer at cycles 5, 40, 80, 120, 160, and 200 after onset of pulsation (rate 40 cycles per min). Curosurf 5 mg/ml (open boxes), Curosurf 5 mg/ml and meconium 2.75 mg/ml (filled boxes), Curosurf 5 mg/ml, meconium 2.75 mg/ml, and 0.3 mM FeCl3 (filled circles), and Curosurf 5 mg/ml and 0.3 mM FeCl3 (open circles). There was a significant difference (p < 0.0001) between the Curosurf/ meconium group and the other groups. There were no significant differences between the other groups.

Reversing the order of addition of meconium and FeCl₃ (to surfactant) did not influence the result. The inhibition resulting from 1 h of incubation of meconium with Curosurf was reversed immediately by addition of FeCl₃. Doubling the concentration of meconium (5.5 mg/ml) did not cause further inhibition of 5 mg/ml of Curosurf (_min: 20 ± 1.2 mN/m and _max: 46 ± 4.0 mN/m, n = 3). This inhibition could also be reversed by addition of 0.3 mM FeCl₃ (_min: 3.4 ± 1.1 mN/m and _max: 41 ± 6.7 mN/m, n = 3).

Inhibition of 2.5 mg/ml of Curosurf by 5% EDTA-plasma (_min: 26 ± 3.4 mN/m and _max: 54 ± 4.4, n = 6) could also be reversed by 0.94 mM FeCl₃ (_min: 13 ± 11 and _max: 46 ± 10, n = 7, p < 0.05) and 1.9 mM FeCl₃ (_min: 7.2 ± 7.7 and _max: 39 ± 6.8, n = 5, p < 0.01).

The iron-chelator deferoxamine did not modify the effect of FeCl₃ on meconium inhibition of Curosurf (Table 1). Deferoxamine, a powerful chelator of ferric iron and known to inhibit iron-dependent lipid peroxidation and hydroxyl radical formation, did not neutralize the pH lowering effect of FeCl₃. Formation of the deferoxamine-FeCl₃ complex actually caused a slight lowering of pH in the FeCl₃ solution from 3.5 to 3.1. 

Meconium-induced inhibition of Curosurf at 5 mg/ml (_min: 21 ± 3.8 mN/m and _max: 51 ± 5.9 mN/m, n = 15), was also reduced by 0.3 mM CuCl₂ (_min: 5.5 ± 2.3 mN/m [p < 0.01] and _max: 39 ± 6.5 mN/m [p < 0.01], n = 4), and to some extent by 0.3 mM FeSO₄ (_min: 13 ± 8.9 mN/m [p < 0.01] and _max: 48 ± 8.3 mN/m [not significant, NS], n = 7), but not by 0.3 mM MgSO₄ (_min: 20 ± 6.9 mN/m and _max: 52 ± 8.8 mN/m, n = 7). The lack of effect of the iron chelator deferoxamine and the modest effect of FeSO₄ raised the question whether the observed contrainhibition could be mediated by a change in pH rather than by the ferric ion specifically.

pH Measurements and pH Adjustments

pH measured in the samples revealed that addition of meconium (pH 6.2) raised pH of Alveofact and Curosurf but had no such effect on Survanta (Table 2). FeCl₃ lowered pH of all three preparations. Addition of 0.3 mM CuCl₂ to Curosurf/ meconium resulted in a prominent decrease in pH (to 5.2). FeSO₄ (pH 6.1) had little effect, and MgSO₄ (pH 6.3) had no effect on pH, paralleling the minor and absent contrainhibitory properties of these compounds, respectively.

Addition of 4.3 mM acetic acid (_min: 5.9 ± 2.5 mN/m and _max: 34 ± 1.9 mN/m, n = 5) to Curosurf/meconium mimicked the contrainhibitory effect of FeCl₃. As for FeCl₃, the effect of acetic acid was dose-dependent (Figure 2B). Combining measurements on Curosurf/meconium mixed with FeCl₃, acetic acid, CuCl₂, FeSO₄, and MgSO₄ revealed that adjusting pH of the Curosurf/meconium mixture from approximately 6.2 to approximately 5 to 5.5 gradually reversed the inhibitory effect of meconium (Figure 4).

View larger version (9K): [in this window] [in a new window]   Figure 4.   Surface tension at minimum bubble size in samples containing 5 mg/ml of Curosurf and 2.75 mg/ml meconium plotted against pH of the sample after addition of FeCl3, CuCl2, FeSO4, MgSO4, or acetic acid. Each dot represents the mean of 4 to 10 observations.

Image Analysis of Microbubbles in Surfactant

Surfactant function was also assessed with computer-aided image analysis of bubble size and number of bubbles in vortexed samples. Addition of 0.55 mg/ml meconium to 1 mg/ml of Curosurf reduced the total number of bubbles in 15 fields (4.35 mm²) from 1,748 ± 481 to 180 ± 166. This effect was reversed by addition of 0.06 mM FeCl₃ (Figure 5). The diameters of the few bubbles that were formed in the presence of meconium were not changed compared with those of Curosurf without meconium (Figure 6). Representative microphotographs of microbubbles in various samples are shown (Figure 7). The microscopic appearance of Curosurf is not changed when FeCl₃ is added without meconium present (Figure 7B).

View larger version (15K): [in this window] [in a new window]   Figure 5.   Number of bubbles in 15 light-microscopic fields measured with image analysis in samples of Curosurf (1 mg/ml) with or without meconium (0.55 mg/ml) and FeCl3 (0.06 mM). In each sample, bubbles were studied in three different droplets. Mean and SD from 2 to 3 independent samples are given. *p < 0.05 versus Curosurf and versus Curosurf + meconium + FeCl3.

View larger version (18K): [in this window] [in a new window]   Figure 6.   Percent of microbubbles (diameter < 20 µm [black bars] and diameter < 10 µm [white bars]) in samples with Curosurf (1 mg/ml), with or without meconium (0.55 mg/ml) and FeCl3 (0.06 mM). In each sample, bubbles were evaluated in 15 fields from three different droplets. Mean and SD from three independent samples are given. There were no statistical differences between the groups.

View larger version (94K): [in this window] [in a new window]   View larger version (82K): [in this window] [in a new window]   View larger version (40K): [in this window] [in a new window]   View larger version (58K): [in this window] [in a new window]   Figure 7.   Representative microscopic fields showing microbubbles in samples with (A) Curosurf (1 mg/ml), (B) Curosurf (1 mg/ml) and FeCl3 (0.06 mM); (C ) Curosurf (1 mg/ml) and meconium (0.55 mg/ml); and (D) Curosurf (1 mg/ml), meconium (0.55 mg/ml), and FeCl3 (0.06 mM).

Wilhelmy Balance Recordings

None of the samples (n = 5) containing Curosurf (40 mg/ml) and meconium (14 mg/ml) reached a surface tension less than 30 mN/m during the period of observation (mean 70 s) (Figure 8). When 1.25 mM FeCl₃ was added to meconium and Curosurf, surface tension dropped to 30 mN/m in 20 ± 6.9 s (n = 5). Curosurf alone reached 30 mN/m after 6.6 ± 3.6 s (n = 5) and the corresponding time interval for Curosurf mixed with FeCl₃ was 7.5 ± 13 s (n = 5).

View larger version (14K): [in this window] [in a new window]   Figure 8.   Wilhelmy balance recordings. Curosurf (40 mg/ml) was incubated with or without meconium (14 mg/ml) and FeCl3 (1.25 mM) and applied as a droplet (50 µl) onto a hypophase of saline. Surface tension, reflected by the traction force on the dipping plate, was recorded. Each sample was tested five times and a median value used. Mean values from five different samples are shown. Filled boxes represent Curosurf and meconium; filled circles, Curosurf, meconium, and FeCl3; open boxes, Curosurf alone; and open circles Curosurf and FeCl3. There were no significant differences in time to reach surface tension 30 mN/m between the Curosurf, Curosurf/FeCl3, and Curosurf/meconium/FeCl3 group. The Curosurf/meconium group did not reach surface tension 30 mN/m during a mean observation period of 70 s.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The most striking finding of this study is the dose-dependent reversal of meconium inhibition of surfactant by FeCl₃ or acetic acid. Whether both acetic acid and FeCl₃ exert their contrainhibition through an adjustment of pH or represent two different mechanisms is not clear. However, the failure of the iron chelator deferoxamine to modify the contrainhibition exerted by FeCl₃, the lower effect of FeSO₄ (with less effect on pH than FeCl₃), and the observed correlation between pH and contrainhibitory capacity (irrespective of addition of acetic acid or FeCl₃) indicate a pH-associated mechanism. Furthermore, the rapid effect of FeCl₃, with the same results with incubation times of 10 and 60 min, argues against a breakdown of inhibitory substances in meconium.

Meconium increases the pH of both Alveofact and Curosurf, and the contrainhibition exerted by acidification might suggest that the inhibitory effect of meconium is caused simply by raising pH of the surfactant suspension. Amirkhanian and Merritt demonstrated that Survanta maintained its optimal surface properties at a pH range of 4.0 to 7.0 (24). In contrast to natural lung surfactant, which contains surfactant protein A (SP-A), further alkalization (pH > 7.2) of Survanta decreased its surface tension lowering abilities (24). However, our finding that Survanta has a high initial pH (6.7) which is actually slightly lowered (to 6.6) by addition of meconium, does not support elevation of pH as a central mechanism in meconium inhibition of surfactant.

Indeed, _min of Survanta was higher than that of Alveofact or Curosurf, but acidification of Survanta by addition of FeCl₃ (from pH 6.7 to 3.9) did not significantly alter its surface tension lowering abilities. Qanbar and coworkers previously found that acidification of DPPC from pH 6.5 to 3.5 resulted in enhanced adsorption and a decrease in the area compression required to reach near-zero surface tension (25). In our experiments, acidification of uninhibited surfactant did not seem to alter the "basic" quality of the respective surfactant preparations, neither with respect to adsorption kinetics nor with respect to _min after 5 min of pulsation.

Despite different basal values, the meconium-inhibited preparations responded similarly to acidification with reversal of inhibition exerted by meconium. Thus, the observed contrainhibitory effect seems to be due to either a change in the inhibitory substance or substances in meconium, altered susceptibility of surfactant to inactivation, or a change in the interaction between meconium and surfactant, rather than a more nonspecific basic improvement of surfactant function.

The mechanisms of meconium inhibition of surfactant are unknown. Meconium contains proteolytic enzymes (26) and breakdown of surfactant constituents has been suggested as a possible factor. Our results showing complete reversal of inhibition support the conclusion of Taeusch and coworkers that meconium does not destroy important constituents of the surfactant film (18). Recently, polymers such as dextrans, polyethylene glycols, and polyvinylpyrrolidones have been demonstrated to counteract inhibition of surfactant exerted by meconium, serum, lysophosphatidylcholine (18), or albumin (27). Taeusch and coworkers suggested that these polymers dissociate meconium-surfactant complexes, facilitating adsorption of surfactant components at the air-liquid interface. Our present observation that the number of bubbles in mixtures of surfactant and meconium was restored to normal after addition of FeCl₃ confirms the reversibility of meconium inhibition. The number of bubbles was significantly reduced by addition of meconium, in keeping with a pH-dependent complex formation of meconium and surfactant or one or more surfactant components. However, other explanations such as formation of mixed unstable films or competition between substances for adsorption at the air-liquid interface cannot be excluded. Addition of meconium to a bovine surfactant changes its ultrastructural appearance from lamellar rodlike micelles with open ends between the rods to spherical lamellar structure without open ends (16, 17). This observation could also fit a possible complex formation of constituents in meconium and surfactant.

In contrast to Curosurf and Survanta, Alveofact at 5 mg/ml was not clearly inactivated by meconium. Seeger and coworkers reported similar differences for fibrinogen, hemoglobin, and albumin inhibition (28). Although there are several differences in phospholipid composition between the three different preparations, Seeger and coworkers emphasized differences in surfactant protein B (SP-B) content between Alveofact (1.7%), Curosurf (0.2%), and Survanta (< 0.1%) (28). SP-B-based surfactants exhibit markedly lower susceptibility to fibrinogen inhibition than SP-C-based preparations (29). Furthermore, monoclonal antibodies to SP-B cause inactivation of porcine surfactant (30). SP-B is an amphipathic cationic small protein, and human SP-B has nine positively charged and two negatively charged amino acids, giving a net positive charge of seven (31). The amphipathic structure and specific charge interactions between cationic residues and anionic lipids are important to film stabilization (31). Changes in pH might influence the ionization of zwitterionic phospholipids and the binding of surfactant proteins to lipid polar head-groups. If a minimum concentration of SP-B is required to form an efficient surfactant film, inhibition may occur by binding of SP-B to constituents in meconium.

As we found the same ranking of Alveofact, Curosurf, and Survanta in susceptibility to meconium inhibition as demonstrated by Seeger and coworkers for protein inhibition (28), we suspect that at least one common mechanism might be involved. We also found that inhibition of surfactant by plasma could, like inhibition by meconium, be reversed by ferric chloride. Meconium contains substantial amounts of protein, but rather small amounts of albumin. Furthermore, both bilirubin and the three predominant free fatty acids of meconium, oleic, stearic, and palmitic acid, have been shown to inhibit surfactant (32). Surfactant is inhibited both by the water-methanol soluble fraction (containing proteins and bilirubin) and by the chloroform soluble fraction (containing free fatty acids, triglycerides, and cholesterol) of meconium (2, 6). The chloroform soluble fraction contains the highest specific inhibitory activity (2). Whereas meconium retards adsorption (2), this parameter is not affected by oleic acid, which instead may inhibit surfactant by disrupting the rigid interfacial film (33). In contrast to inhibition exerted by meconium (2) and plasma proteins, the effect of oleic acid is not counteracted by increasing surfactant concentration (33).

Beneficial effects of surfactant therapy, increasing the alveolar pool size of surfactant beyond the threshold for inhibition by meconium, have been demonstrated in rabbits (14, 34), rats (13), and human babies (15, 35, 36). Surfactant lavage improves oxygenation of piglets after experimental meconium aspiration (37), and a similar effect has recently been demonstrated in adult rabbits and newborn rhesus monkeys subjected to bronchoalveolar lavage with diluted KL₄-surfactant (38). Whether buffering of surfactant at pH 5-5.5 could improve its in vivo performance in the presence of meconium has not been investigated. This pH is close to the pH of Curosurf and Alveofact; however, the buffer capacity of these preparations is very small. pH of alveolar lining fluid in rabbits is 6.9 (39). It is not known whether exogenous surfactant maintains an acidic pH in situ, or what buffer capacity this would require.

At present, we know that meconium inhibition of surfactant can be counteracted by surfactant proteins, polymers, and FeCl₃, acetic acid, or CuCl₂. Whether these mechanisms might be beneficial in clinical settings remains to be determined. However, the results of these in vitro studies may bring us closer to an understanding of the mechanisms of meconium inhibition of surfactant. Together with the study of Taeusch and coworkers (18), our results indicate that meconium inhibition is reversible and that breakdown of surfactant constituents is not a major factor. We suggest that meconium forms an aggregate with surfactant or binds to some constituents of surfactant and that these bindings may be counteracted by acidification.