Introduction
Surfactant complex is a mixture of phospholipids and proteins. Phosphatidylcholine is the major surfactant phospholipid and among the several species, PC16:0/16:0 is the principle PC, thought to be actively involved in surface reduction at the air-liquid interface [
1]. In patients with respiratory failure secondary to acute respiratory distress syndrome (ARDS), the lavaged surfactant complex show compositional derangement and lacks adequate surface activity [
2],[
3]. Consequently, these changes are likely to contribute to the detrimental clinical features of severe hypoxemia, poor lung compliance and lung atelectasis, which are characteristic of ARDS. However, the exact underlying mechanisms for this dysregulation in surfactant metabolism is not known. The plausible hypothesises are complex and includes decreased synthesis/ secretion, increased breakdown by hydrolysis or oxidation, increased clearance and or functional inhibition by invasion of non-surfactant material into the alveolus [
4]. Human in-vivo studies exploring these mechanisms are lacking and consequently, there remains a significant gap in the understanding of surfactant metabolism during acute alveolar injury.
ARDS is a heterogeneous disease process with significant variations in aetiology, pathophysiology, response to therapy and outcome. Furthermore, it imposes significant morbidity and mortality and lack of defined pharmacotherapies to moderate disease process remains an enduring challenge. Current diagnostic definitions fail to identify a homogenous population, and are limited by the lack of specificity and information regarding possible underlying genophenotypes. The disappointing results from surfactant replacement clinical trials in adults, reiterate the necessity for in-vivo human studies to explore the exact underlying mechanisms leading to the disordered surfactant metabolism evident in ARDS. Such studies may enable an individualised treatment strategy in the future after biochemical phenotyping of patients according to their intrinsic surfactant metabolism.
Traditional assessments of pulmonary surfactant consisted of isolation with sequential bronchoalveolar lavage fluid (BALF) and subsequent separation of surfactant phospholipids by liquid chromatography [
3],[
5]. Such studies provided significant insights into surfactant compositional abnormalities in patients with ARDS. However, quantitative phospholipid assessments were unpredictable due to variability in surfactant recovery by BALF. Furthermore, such approaches lack specific information regarding the dynamics of surfactant synthesis and turnover. Stable isotope labelling coupled with analytical methods of electrospray ionization mass spectrometry (ESI-MS/MS), provides a novel approach to the dynamic assessment of surfactant metabolism in addition to highly specific molecular analysis of surfactant phospholipids [
6]. Consequently, we used this methodology to characterise surfactant phosphatidylcholine kinetics in patients with ARDS.
Methods
Study population
Following approval from National Research Ethics Committee (10/WNo01/52) and University Hospital Southampton Research and Development Department (CRI0244), patients with ARDS were identified and recruited from general intensive care unit at University Hospital Southampton NHS Foundation Trust. The diagnosis of ARDS was based on the American European Consensus Conference diagnostic criteria [
7]. All eligible patients were enrolled after informed assent was obtained from patient's next of kin within 72 hours of onset of ARDS. All patients were sedated and required invasive mechanical ventilation. The control group consisted of nine healthy volunteers without any prior history of lung diseases and were all non-smokers.
Methyl-Dcholine chloride
Deuterated
methyl-D
9 choline chloride is a stable isotope of choline, which can be used to probe dynamic PC synthetic pathways in-vivo and has been successfully applied to quantify surfactant PC flux via the CDP-choline pathway in healthy human volunteers and in animal models [
8],[
9]. Following informed assent or consent, patients and controls had an intravenous infusion of
methyl-D
9 choline chloride (3.6 mg/kg) for a three hour period.
Sample collection and processing
Bronchoalveolar lavage fluid (BALF) was obtained via a fibre-optic bronchoscope (Olympus BF P60). All patients were intubated with an endotracheal tube (size 7-9) and mechanically ventilated. Prior to the bronchoscopy, patients were pre-oxygenated with 100% oxygen and this was continued throughout the procedure. Patients were given additional sedation if required to facilitate the procedure. Bronchoscopy was not performed on patients with inspired oxygen (Fi02) of >80% and was abandoned, when there was a significant desaturation defined as pulse oximetry oxygen saturation of <85%. Small volumes of BALF were collected by suctioning after instillation of 10-20 mls of warm sterile saline from a single lobar bronchus (either from middle or lower lobes). This was then rotated (in the order of right middle lobe, left middle lobe, right lower lobe and left lower lobe) for the subsequent lavages to minimise theoretical concerns, that repeated lavage of the same lobe could be detrimental to surfactant concentration or function in that lobe. This small volume BALF was performed at baseline before the administration of methyl-D9-choline chloride infusion, and at 6, 12, 24, 48, 72, 96 hours after the infusion. The reason for the frequent sampling recovery in the early stages was to ensure adequate analysis is performed at this stage, which could be utilised to target therapeutic interventions. For healthy volunteers, small volume BALF was obtained at 24 and 48 hours after methyl-D9-choline infusion.
BALF was transferred at 4°C to the processing laboratory. 10ul of Butylated Hydroxy Toluene (BHT) (20 g/L) solution was added to all samples as an anti-oxidant. Then the samples were filtered through a 100um nylon cell strainer and centrifuged at 100 x g for 20 seconds at 4°C. The resultant liquid material was further centrifuged at 400 g for 10 minutes at 4°C. The supernatant was aliquoted in eppendorfs and stored in a −80°C freezer.
Total lipid extraction was performed using the Bligh and Dyer method [
10]. Samples were made up to a volume of 800ul with 0.9% NaCl and dimyristol-PC (PC14:0/14:0) was added as internal standard (IS). One ml of chloroform and 2 mls of methanol were added to the samples, followed by further 1 ml of chloroform and 1 ml of distilled water. Sharper resolution was achieved by centrifuging at 400 g at 20°C for 20 minutes. The lower lipid rich layer was then removed and dried under a nitrogen concentrator at 40°C. Once dried further 1 ml of chloroform was added and dried again to be analysed by mass spectrometry.
Mass spectrometry analysis and data extraction
The dried phospholipid fraction was suspended in a mixture of methanol-butanol-water-concentrated NH
4OH (6:2:1.6:0.4 v/v), and was injected by syringe pump at a rate of 8ul/ml into the electrospray ionisation interface of a Xevo TQ mass spectrometer (Waters Corporation, UK) (ESI-MS/MS). The endogenous and newly produced PC species were calculated from MS/MS fragmentations of precursor ion m/z +184 and m/z + 193 respectively [
8],[
9]. Dedicated excel spread sheets were used to quantify ion peaks after corrections for
13C-isotopic effects.
Statistical analysis
Data were summarised by means, standard deviation (SD) and standard errors of means (SEM). The difference in composition in each group was examined by calculating the difference in means. Single comparisons were analysed by Student’s t-test and multiple comparisons by ANOVA of variance.
Discussion
This study conducted on a small number of ARDS patients illustrates the feasibility of performing stable isotope labelling in combination with ESI-MS/MS analytical methods to investigate surfactant phospholipid metabolism in a defined human patient cohort. This is the first study to characterise the molecular specificity of surfactant synthesis in adult ARDS population. The study group composed of patients with moderate to severe ARDS as defined by the degree of hypoxemia. Most patients (90%) had pneumonia as the single precipitating factor.
Despite the significant variability, ARDS patients had persistently lower BALF total PC concentrations compared to healthy controls. The BALF fractional PC16:0/16:0 absolute concentrations were also much lower in patients and only accounted for about 10% of the healthy controls. The decreased fractional PC16:0/16:0 content in our patient group is comparable with a previous study of patients with ARDS, where the disaturated PC content of from patients was only 17% of that of controls [
11]. Apparent differences with other previous reports in ARDS are largely due to a combination of less severe disease [
5] and alternative methodologies and analysis of purifying isolated surfactant fractions [
2],[
5],[
12]. This deficit in total PC and fractional PC16:0/16:0 concentration may be due to reduced synthesis, increased breakdown, or dilution by pulmonary oedema. Reduced synthesis in turn could be due to either destruction of AT-II cells or dysfunction of their synthesis and secretory mechanisms. However, this was coupled with increased
methyl-D
9 choline enrichments of both total PC and PC16:0/16:0 (Figure
5A and
5B). One possible explanation for this finding is that the rate of surfactant PC synthesis in remaining uninjured AT-II cells must be preserved or even enhanced, but that total surfactant PC production is indeed decreased probably due to decreased numbers of functional AT-II cells. This conclusion support and extend a previous study that employed in-vivo stable isotope labelling with deuterated water (
2H
20) to characterise patterns of enrichment of newly-synthesised fatty acids into disaturated PC in ARDS patients [
11]. This study also demonstrated increased fractional labelling of disaturated PC, coupled with decreased concentration, but provided no specific molecular species analysis. In this context, the interaction between rates of PC synthesis on the endoplasmic reticulum, its packaging into lamellar bodies and subsequent secretion will be critical. One realistic scenario, for instance, could be that decreased surfactant secretion from a depleted number of AT-II cells results in accumulation of labelled PC in existing AT-II cells, which could explain the conundrum of decreased concentration, but increased fractional enrichment of surfactant PC in ARDS patients.
As fractional catabolic rates could not be calculated for either ARDS patients or healthy controls, no conclusion can be inferred about the potential contribution of increased surfactant catabolism to decreased BALF PC concentration in ARDS patients. However, any substantial increased surfactant catabolism would have been expected to cause a more rapid decline in PC enrichment in ARDS patients compared with control subjects. While such a comparison was not possible in the present study, the duration of labelling of sputum PC in previous studies [
8],[
13] was very similar to the enrichment profiles of BALF PC in ARDS patients shown in Figure
5. Consequently, the most likely cause of the lower BALF PC concentration in ARDS patients was a significantly decreased rate of total alveolar surfactant synthesis.
The newly synthesised surfactant fraction in ARDS patients also revealed abnormal PC composition, with diminished levels of PC16:0/16:0 and increased fractional compositions of unsaturated PC species, when compared to the endogenous composition. We have previously demonstrated in healthy volunteers that, during earlier time points following
methyl-D
9-choline infusion, the newly synthesised surfactant PC complex is different to that of endogenous composition [
8],[
13]. Furthermore, it takes up to 48 hours for the labelled
methyl-D
9-PC16:0/16:0 to achieve equilibrium with unlabelled PC16:0/16:0 percentage composition. If the acyl- remodelling mechanisms described previously [
8],[
9],[
13] are complete before surfactant secretion, the labelled PC16:0/16:0 composition should be at equilibrium with endogenous composition at all-time points. As our data shows this not be the case, our results are compatible with the previous healthy volunteer study conclusion, that not all alveolar surfactant is acyl- remodelled prior to secretion [
8],[
13]. Despite similarities, ARDS patients had much lower fractional composition of labelled PC16:0/16:0 at all- time points compared to healthy controls. In-addition, three patients had comparatively much lower fractional labelling compared to the endogenous fraction of PC16:0/16:0 at 48 hours. These findings indicate, that despite an increase in fractional synthesis, the surfactant PC produced has abnormal composition with some patients showing altered acyl-remodelling mechanisms of PC16:0/16:0.
Previous surfactant related clinical studies used gas chromatographic separation methods to quantify palmitic acid (16:0) and used the latter as a surrogate for PC16:0/16:0 [
2],[
5]. This methodology always produced higher fractional compositional values (70-80%) for palmitic acid (16:0), which may have come from PC species other than PC16:0/16:0, and have no clinical relevance. In contrast, stable isotope labelling of surfactant precursors in combination of ESI-MS/MS analytical approach, is a novel technique to study dynamic phospholipid flux with precise molecular analysis. This facilitated for the first time the assessment of surfactant PC molecular synthesis via CDP-choline pathway in adult ARDS population.
This is the first study to use small volume BALF to investigate surfactant metabolism in patients with ARDS. In the past, investigators have used quantitative BALF to investigate compositional analysis in ARDS. Although in general this is a safe procedure, repetitive large volume BALF may theoretically deplete alveolar surfactant and may produce a negative clinical impact. Moreover, anecdotally desaturations are much more common during quantitative BALF. The surfactant PC composition in our control group is similar to previous publications, which has used large volume BALF for molecular analysis [
12],[
14]. This suggests, at least in physiological conditions small volume BALF is comparable to quantitative BALF for the study of surfactant phospholipid composition and metabolism. Small volume BALF was tolerated by all, without any significant immediate complications. Despite no cardiovascular or respiratory compromise, all patients needed additional sedation, on average of 40 mgs of propofol or 2 mgs of midazolam to perform this procedure.
As this was a pilot study with small number of patients, no clinical correlations or outcome inferences were made. The study was further limited by the lack of ventilated control patients, who are more representative as a comparable population, than healthy adults. Despite these limitations, the demonstration of the existence of variable surfactant composition, synthesis and metabolism among patients indicate, the presence of underlying phenotypes of surfactant metabolism. Future similar studies if adequately powered, may help to explore clinical correlations.
Competing interests
ADP has no direct financial interest in the work presented in this study; his surfactant research programme is supported by kind donation of a therapeutic surfactant from Chiesi for a clinical trial. The remaining authors declare that they have no competing interests.
Authors' contributions
AD co-ordinated the study, contributed to the study design and performed all clinical and laboratory procedures, calculated the results and drafted the manuscript; VG co-ordinated the laboratory procedures; RC contributed to the clinical procedures and manuscript revision; MPWG contributed to the study design, clinical procedures and manuscript revision; ADP contributed to the study design, data interpretation and manuscript revision. All authors read and approved the final manuscript.