Introduction

Phospholipase A2 (phosphatide 2-acylhydrolase, EC 3.1.1.4) belongs to a superfamily of enzymes which catalyze the hydrolysis of ester bonds at the sn-2 position of membrane phospholipids [1]. The secretory type of this enzyme (sPLA2) is well known to play a key role in the lung injury pathophysiology [2]. sPLA2 is a soluble molecule, mainly produced by alveolar macrophages and secreted into the alveoli in which it is active [1, 2]. The enzyme is involved in inflammation pathways, through the formation of eicosanoids and other inflammatory mediators [1, 3]. sPLA2 is also responsible for surfactant dysfunction by generating lyso-phopshatidylcholine (lyso-PC) and therefore reducing the tensioactive properties of surfactant [4, 5]. This has been widely recognized as a crucial point in the acute respiratory distress syndrome (ARDS) [3, 6], while a sPLA2 role is now being investigated in some obstructive lung diseases, as well [7]. Pancreatic sPLA2 is also involved in meconium aspiration syndrome (MAS), a severe form of neonatal lung injury, being responsible for surfactant inactivation [8, 9].

Bile acids (BA) have been shown to be harmful for the lung: their intratracheal injection in animals produced an injury histologically similar to the hyaline membrane disease [10]. Several years ago they have been associated with aspiration-related ARDS in surgical patients [11, 12], while we recently demonstrated their role in neonatal bile acid pneumonia, a respiratory distress form in which newborn lungs are challenged with high concentrations of BA coming from the circulation [1315]. Moreover, some data strongly support the causative role of BA in determining bronchiolitis obliterans in patients who received lung transplant and suffered from gastro-esophageal reflux [16, 17]. BA are also present in meconium and they are likely to play a role in the development of MAS. Mechanisms of BA lung toxicity have been poorly studied and several pathways may be involved: cellular apoptosis, pH-related enzymatic damage, host defense reduction and increased susceptibility to infections have been suggested [18].

Surfactant inactivation is another supposed mechanism [19, 20] but the interaction BA–sPLA2 has never been investigated. We hypothesized that BA can increase sPLA2 activity leading to surfactant deactivation. We aimed to verify this hypothesis in an extracellular environment representing an in vitro model of aspiration.

Materials and methods

Patients and broncho-alveolar lavage collection

During a 8-month period all infants who were admitted to our Pediatric Intensive Care Unit (PICU) and required broncho-alveolar lavage (BAL) for routine microbiologic surveillance were eligible for the study. To be enrolled, babies had to have healthy lungs in physiological conditions; in details they had to meet the following criteria: (1) age ≤6 months, (2) need for mechanical ventilation for reasons other than any lung disease, (3) normal appearing chest X-rays and (4) need for oxygen <25%.

Exclusion criteria were as follows: (1) prematurity (gestational age <37 weeks), (2)major congenital malformations, (3) any evidence of respiratory disease, (4) proven lower tract respiratory infection in the last month, (5) need for thoracic surgery and (6) any evidence of inflammation.

Non-bronchoscopic BAL was performed, according to the recommendations of the European Respiratory Society [21] by instillation into the endotracheal tube of two aliquots of 1 ml/kg 0.9% NaCl, warmed at 37°C, followed by suctioning. Blood stained BALs were excluded from further analysis. The first collected aliquot was used for microbiological tests, while the second was centrifuged to obtain supernatant, as previously published [15]. BAL supernatant were immediately frozen at −80°C until used. All infants were ventilated using Servo-I ventilator (Maquet Critical Care, Solna, Sweden) providing volume controlled ventilation. BAL was carried out within 1 h from the intubation and clinical data were registered in real time; 15′ before BAL 0.5 ml blood was also drawn using indwelling arterial lines, for routine biochemical parameter check, including urea measurement.

BAL is routinely carried out for microbiological surveillance in our PICU; no blood samples were collected exclusively for the purposes of this study and no changes in the routine clinical assistance were provided. Being a completely non-invasive study, our Institutional Review Board approved a waiver of consent.

Study design and materials

Broncho-alveolar lavage was assayed for total sPLA2 activity in basal condition and then samples were randomly assigned to receive BA at 1 and 5 μmol/l [19] or an equivalent volume of normal saline and enzyme activity was re-measured. Randomization was carried out using a computer-generated randomisation table.

Human BA were purchased in lyophil serum (TBA controls, Sentinel Diagnostic Milano, Italia) as a chemical standard at the concentration of 42.3 μmol/l. This solution was microbiologically screened to ensure its sterility.

Subsequently, only samples coming from neonates were assigned to receive an amount of poractant-α (Curosurf, Chiesi Farmaceutici, Parma, Italia) needed to reach a 10 mg/kg BAL concentration of phospholipids. This was chosen according to the values found in neonatal BAL after the surfactant replacement therapy for respiratory distress syndrome [22]. In these BALs, PLA2 activity was again measured after correction for the surfactant optical densities.

sPLA2 was measured using an ultrasensitive non-radioactive method on microplate (Assay Designs, Ann Arbor, MI, USA). This uses 1,2-dithio analog of diheptanoyl phosphatydilcholine that is a specific substrate for PLA2 [23]. Upon hydrolysis of the thiol ester bond at sn-2 position by sPLA2, free thiols are generated and react with 5,5′-dithio-bis-2-nitrobenzoic acid producing a sulfhydryl compound, detected calorimetrically at wavelength of 405 nm. The colorimetric assay was checked with standardized increasing sPLA2 activity and performed following manufacturer’s recommendations.

The intra-assay and the inter-assay variability were <5 and <9%, respectively. To normalize the dilution bias resulting from BAL fluid collection, a measurement of urea in serum and in BAL supernatants was performed in duplicate using a colorimetric assay on automatic analyzer (Urea Nitrogen on Olympus AU2700; Olympus Life Europe, Hamburg, Germany). The serum/supernatant urea ratio was used as a multiplying coefficient to express PLA2 activity [24]. All sPLA2 assays were performed in duplicate at a 37°C temperature and read within 10′ on photometer Viktor (Perkin Elmer, Turku, Finland). All results were corrected for dilution factors determined after the addition of BA and surfactant to each BAL sample.

Statistics

Non-Gaussian distribution of data was proven using Kolmogorov–Smirnov test, therefore data were expressed as median [interquartile range] and then the following analyses were performed. sPLA2 basal activity between neonates and infants was tested using Mann–Whitney U test; sPLA2 activity before and after the additions, in each one of the three experimental group, was analyzed with Wilcoxon signed rank test (two related variables, “before and after” design). In neonatal samples, which were subjected to multiple experimental addition (first BA/saline and then poractant-α), sPLA2 activity variation throughout the experiment was tested using Friedman Q test (three related variables, “before and after” design, in multiple time-points). Correlation between patients’ age, weight and sPLA2 activity was tested using Spearman analysis.

Testing was carried out using SPSS 15.0 for Windows (SPSS Inc., Chicago, IL, USA) and P values <0.05 were considered to be statistically significant.

Results

During the study period, 23 infants were enrolled, seven of which were neonates. Reasons for intubation were surgical procedures (17), intracranial hemorrhage (3), neonatal asphyxia (2), status epilepticus (1). All BAL samples resulted to be sterile and non-blood stained. Median age was 3 months [0.1–4], while median weight resulted to be 6 kg [3.9–6.5]. No significant differences in sPLA2 activity were seen between neonates and infants (neonates: 283.9 IU/ml [77.3–990.5]; infants: 136.7 IU/ml [3.7–373.5]; P = 0.278). No correlations were found between sPLA2 activity and the age or weight of infants.

Figure 1 reports the global sPLA2 activity in basal conditions and in the three experimental models. Adding 5 μmol/l BA significantly increased the enzyme activity (245.9 IU/ml [8.8–424.8] vs. 1361 IU/ml [601.8–2814.3]; P = 0.012), while insignificant changes were seen after addition of 1 μmol/l BA (238.1 IU/ml [10.6–757.8] vs. 312.7 IU/ml [28.6–1418.3]; P = 0.249) or normal saline (135.3 IU/ml [43.8–348] vs. 355 IU/ml [113.6–548.4]; P = 0.06).

Fig. 1
figure 1

Measured global activity of sPLA2 (expressed using serum/supernatant urea ratio) in basal conditions and after: a addition of normal saline (nine samples), b addition of 1 μmol/l of human bile acids (six samples), c addition of 5 μmol/l of human bile acids (eight samples). In model c sPLA2 activity is significantly higher than in basal conditions

Figure 2 shows the effect of surfactant on sPLA2 activity in BAL of neonates. Baseline activity was 283.9 IU/ml [77.3–990.5]. Because of the randomized allocation of BALs to different amounts of BA or normal saline, 5/7 neonatal samples received 5 μmol/l BA, while the remaining two BALs were given normal saline and the activity after the addition increased up to 598.2 IU/ml [367–1656.5]. The subsequent addition of surfactant reduced the sPLA2 activity almost to the baseline level: 118.1 IU/ml [86.4–131.3]. These enzyme activity levels (baseline, after BA addition, after poractant-α addition) resulted significantly different between them (P = 0.004).

Fig. 2
figure 2

Effect of poractant-α on the sPLA2 activity in BAL of newborn infants (seven samples). After measurement in basal conditions and the randomization, five samples received 5 μmol/l of human BA and two an equivalent volume or normal saline. Poractant-α was subsequently added and sPLA2 was again assayed. Enzyme activity is significantly different between the three experimental time-points. Boxes represent median and interquartile range, bars symbolize extreme values

Discussion

Our findings suggest that BA raise sPLA2 activity and this might have clinical implications for aspiration-related ARDS in adults [12] or bile acid pneumonia and MAS in neonates [1315]. Although bile acid pneumonia and MAS are quite rare they can have severe clinical pictures and the interaction BA–sPLA2 seems important for their development. Since surfactant administration is a standard therapy for neonatal respiratory distress syndrome, we choose to add exogenous surfactant only to BAL samples coming from newborn patients. This allowed us to study the effect of exogenous surfactant on the interaction BA–sPLA2. Surfactant was added to reach a phospholipids concentration equal to that found in babies receiving surfactant in clinical practice [22]. The surfactant addition was able to revert the BA-induced increment in sPLA2 activity, reducing it almost to the baseline. Surfactant addition provides large amounts of exogenous phospholipids which strongly affects the sPLA2 activity. Wide literature data have shown an inhibitory effect of surfactant phospholipids on sPLA2 expression in vitro [1, 3]. Our findings also suggest a mechanism of direct inhibition through an excess of substrate. In fact, the rate of the reaction catalyzed by sPLA2 is mainly determined by the bile/phospholipids ratio [25, 26] and the exogenous surfactant added to our model decreased this ratio, achieving an activity reduction.

Main limitations of our study are the relatively small sample size and the absence of cellular components in the model. Larger studies could identify BA threshold concentration for surfactant deactivation and explore further lung injury pathways involving alveolar macrophages. In fact, these cells are the main source of sPLA2 [3] and therefore are likely to play a role in the alveolar environment charged with BA. Moreover, as for all in vitro studies, our findings should be cautiously considered as they represent only a partial picture of the complex interactions occurring during the lung injury development.

The role of a raised sPLA2 activity should be interpreted according to some recent research data. In fact, sPLA2 is known to cause hydrolysis of phosphatydilcholine (PC) into lyso-PC [4]. This latter can inactivate surfactant: it is surface-active in the sense that it adsorbs at an air–water interface and high levels of lyso-PC may fluidize the surface film in the alveoli, destabilizing them at end-expiration [5]. Robertson et al. [5] have shown lyso-PC to be able to cause a significant elevation of minimum surface alveolar tension at concentrations ≥2.5 mg/ml, in animal models. Similarly, surfactant dysfunction has been recognized as a key mechanism for ARDS in adult patients, as well [3, 6].

We demonstrated a BA-induced sPLA2 activation only at high concentrations of BA. Almost 20 years ago Donoso et al. [19] investigated the interaction between BA and surfactant in rabbits, using identical amounts of BA. These concentrations are the same we observed as median and maximum BA values during neonatal bile acids pneumonia [15]. BA in similar ranges were also found in BAL of adults affected by post-transplant bronchiolitis obliterans following gastro-esophageal reflux [17] and in babies deceased for sudden infant death syndrome [27]. Consistently with our observations, Donoso et al. [19] found an increased surface tension in rabbits after BA addition. This is consistent with the observation of Kaneko et al. [10] who described hyaline membranes in animals after intratracheal injection of BA. Our present findings suggest that this surfactant inactivation, at high BA concentrations, is associated to the increased phospholipids breakdown caused by the raised sPLA2 activity.

The different effect observed at higher BA concentration is consistent with the available data derived from biochemical studies about the interaction BA–phospholipases A2: the mechanism of this interaction can be explained as an anionic activation of sPLA2 [28]. Amphipathic BA may enter and be embedded in a phospholipids layer, presenting a negative charge that can be used by sPLA2. The enzyme can therefore bind its substrate and may be optimally poised for the catalysis. This interaction is depicted in the Fig. 3 and was demonstrated for pancreatic phospholipase A2 in the context of lipid digestion [28]. In the alveoli, where the surfactant phospholipids concentration is high and no BA are present in normal conditions, it is likely that a quite large amount of BA (and so a higher BA/phospholipids ratio) would be necessary to significantly activate sPLA2 and therefore cause surfactant dysfunction. This is what happens when particularly high BA amounts reach the lungs coming from the meconium (in MAS), from the circulation (in bile acid pneumonia) or from aspiration (after gastro-esophageal reflux). High amounts of BA and pancreatic phospholipases are present in meconium [8, 9, 18]. In MAS lungs are therefore contemporarily challenged with both compounds and this may be related to the severity of MAS clinical pictures. In conclusion, BA seem to act as non-substrate cofactors of the physiological substrate in the reaction catalyzed by sPLA2, as it happens in the gastro-intestinal tract [28]. In the lungs the substrate is represented by the surfactant phospholipids instead of the alimentary lipids.

Fig. 3
figure 3

Hypothetical model of anionic activation of sPLA2 by BA. Adapted from a similar description provided in the context of lipid digestion (permission-licence obtained from Elsevier publishing group [28]). Two amphipatic BA are shown [BA] embedded on a glycerophospholipid layer [PL]. Negative charges on BA are indicated as the black parts of the molecule. The bound sPLA2 is represented with the enzyme surface [S] interacting with anions of BA and is therefore optimally poised for catalysis of phospholipids. Black arrows indicate the interaction between negative charges of BA and activation sites of the enzyme

Present data increase knowledge about surfactant physiology but the effect of BA on the lungs is surely complex. Future studies are warranted to clarify all mechanisms of BA toxicity and the role of a possible therapeutic inhibition of sPLA2.