Introduction

Surfactant protein C (SP-C) is composed, depending on the species, of 33–35 amino acid residues and is one of the most hydrophobic proteins in the proteome, accounting for 4 % of phospholipids by weight in large aggregate of premature infants [1, 2]. Its extreme hydrophobicity derives from the presence of a hydrophobic membrane spanning domain, a hydrophobic N-terminal domain [3], and two palmitoyl residues bound to cysteines 5 and 6 [4].

SP-C is expressed primarily in type II alveolar epithelial cells and is synthesized as a precursor protein (proSP-C) that is converted, by a series of processing steps, into mature SP-C. Mature SP-C is stored in the lamellar bodies and finally secreted into the alveolar spaces.

SP-C is the most specifically expressed surfactant protein in type II cells [5]. As a result, measuring the levels of SP-C could be used to gauge the ability of the lung epithelial cells to produce surfactant.

SP-C knockout mice grow normally without any apparent pulmonary abnormalities. They have normal alveolar structure, lamellar bodies, tubular myelin, and surfactant saturated phosphatidylcholine pool size, indicating that SP-C has no role in the intracellular biosynthesis, packaging, and secretion of surfactant. However, SP-C null mice generate a surface film that is unstable at low lung volumes [6]. Furthermore, SP-C has a role in maintenance of lung function during hyperoxia [7]. Lu and co-workers have recently observed a reduction in SP-C expression in infants dying from respiratory distress syndrome who were ventilated at elevated oxygen concentrations [8].

Mutations in the SP-C gene have been associated with familiar interstitial lung disease and adult pulmonary fibrosis [9, 10]. A mutation in the SP-C gene has been recently identified in a young woman with bronchiectasis and severe respiratory insufficiency [11–13]. Marked heterogeneity in severity, age of presentation, and time of presentation of the pulmonary disease may be influenced by genetic background as reported in a study with SP-C knockout mice where a strain differences was observed [14].

Current data on the turnover of SP-C are available only in animals [15–17]. Stable isotope tracer has been used by our group to trace surfactant protein B (SP-B) metabolism [18, 19], therefore the goal of this study was to demonstrate the feasibility of measuring SP-C metabolism in infants and adults using 13C-leucine as tracer.

Methods

Materials

Acetic acid, hydrochloric acid, HPLC grade chloroform, hexane, and methanol were obtained from Carlo Erba Reagents (Milan, Italy). Analytical grade trifluoroacetic acid and trifluoroacetic anhydride were purchased from Sigma Chemicals (Milan, Italy). [113C]l-leucine with isotopic purity of >99 % was obtained from Cambridge Isotope Laboratories (Andover, MA, USA). All solvent used were tested for aminoacids contamination.

Subjects

SP-C kinetics was studied in four infants and eight adults who were admitted to the neonatal intensive care unit and to the intensive care unit, City Hospital of Padova, respectively. All patients had no lung disease and needed mechanical ventilation after major surgery or for neurological failure.

The local ethics committee approved the protocol and informed consent was obtained from all patients or, in the case of infants, from both parents. These patients were also part of two larger studies done to assess the turnover of SP-B [20, 21].

Study design

All infants received a 24-h constant i.v. infusion of 2 mg/kg/h of 113C-leucine dissolved in saline. Adults received a 24-h constant i.v. infusion of 1 g 113C-leucine (about 42 mg/h independently of body weight).

Baseline EDTA-blood and tracheal aspirates (TA) samples were obtained prior to the administration of the isotope. To establish the isotopic enrichment of leucine at plateau, blood samples were collected every 6 h during the infusion period (24 h). TAs were obtained every 6 h until 72 h. Blood was centrifuged at 1,300×g for 10 min; plasma was isolated and stored at −80 °C until analysis. TAs were performed according to a standardized procedure [22] and centrifuged at 400×g for 10 min; supernatant was stored at −80 °C.

Isolation of SP-C from TA

Preliminary phosphorous assay [23] was made on the TA sample in order to avoid overloading of solid-phase extraction columns. Aliquots of TAs containing 200 μg of phospholipids were extracted according to a modified Bligh and Dyer method [19]. The organic phase was dried under nitrogen, resuspended in 1 ml of chloroform and applied to a Bond Elute NH2 column previously activated with 3–5 ml of chloroform. Fractions were eluted with 3 ml of the following chloroform/methanol/glacial acetic acid mixtures: 20:1:0, 9:1:0, 4:1:0, 4:1:0.025, 3:2:0, 1:4:0, and 1:9:0 [18].

Determination of SP-C fractions purity

Western blot analysis performed previously by our group on a bronchoalveolar lavage sample containing 150 μg of phospholipids revealed that SP-C started to be recovered with the mixture 3:2:0 but most of the protein eluted with the fraction 1:4:0 (antibody anti SP-C was provided by Altana Pharma AG, Konstanz, Germany) [18]. Selective elution of SP-C in fractions 1:4:0 and 1:9:0, derived from SPE elutions (see above), was further confirmed by high-resolution mass spectrometry. To that end, 4:1:0, 4:1:0.025, 3:2:0, 1:4:0, and 1:9:0 fractions were individually loaded onto a C4 reversed-phase LC column (3.9 × 15 mm, Deltapak C4, 100 A, Waters, MA, USA) installed on a nano HPLC Dionex Ultimate 3000 interfaced with a Orbitrap LTQ XL (Thermo Fisher Scientific, MA, USA). Isocratic chromatographic conditions were adopted and a constant flow of chloroform/methanol was used. High-resolution direct analysis of intact SP-B and SP-C was performed in positive ionization mode based on the deconvolution of the most abundant charged states of the intact proteins. Were considered the triple, fourth, and fifth charge states for SP-C and eleventh, tenth, ninth, eighth, and seventh for SP-B.

SP-C kinetics

To study the incorporation of 13C leucine into SP-C, 1:4:0 and 1:9:0 fractions were pooled together, evaporated to dryness, and hydrolysed to free amino acids for 24 h at 100 °C in 500 μl of 6 N HCl.

Amino acids from acid hydrolysis were then derivatized into their oxazolinone derivatives [24]. To determine leucine plasma enrichment, 100 μl of plasma were deproteinized with sulphosalicilic acid (6 % w/v) and plasma amino acids were derivatized according to Husek [25].

Both free plasma leucine and SP-C leucine enrichments were measured on a 6890N gas chromatographer coupled to an Agilent 5973i mass spectrometer (Agilent Technologies Italia SpA, Cernusco sul Naviglio, Italy).

Oxazolinone derivatives were analyzed by negative ion chemical ionization. Leucine was separated on a 50 m × 0.20 mm × 0.33 μm Ultra 2 fused silica capillary column (Agilent technologies, Italia, SpA). Oven temperature was programmed for 1 min at 80 °C, increased from 80 to 180 °C at 6 °C min−1, increased from 180 to 280 °C at 30 °C min−1, and held at 280 °C for 3 min. The ion monitored for leucine was m/z 209 while for 113C-leucine it was m/z 210. Results were expressed as mole percent excess (MPE) referring to a calibration curve for 113C-leucine ranging from 0 to 5.23 MPE for oxazolinone derivatives and from 0 to 9.94 MPE for ethyl chloroformate derivatives.

Data analysis

All kinetic measurements were performed between 6 and 24 h assuming a 113C-leucine steady state. The assumption was based on the following considerations: (1) in all patients, plasma 13C-leucine enrichment reached steady state within 6 h from the start of the isotope infusion; (2) the slope of the enrichment curve over time did not deviate significantly from zero between 6 and 24 h.

Fractional synthesis rate (FSR), secretion time (ST), and peak time (PT) were calculated as previously reported [21, 18]. Briefly, FSR was calculated by dividing the slope of the linear increase of the enrichment of SP-C by the plasma steady-state enrichment of free 13C-leucine.

ST of SP-C was defined as the time-lag between the start of the precursor infusion and the appearance of the enriched product. The time-lag was calculated by plotting the regression line for the linear increasing part of the enrichment versus time curve and extrapolating it to baseline enrichment.

PT was the time required to reach maximum enrichment of SP-C 113C-leucine. Kinetic variables were expressed as individual values and as median and interquartile range or as mean and standard deviation as appropriate.

Results

Patients

We observed no side effects after the infusion of 13C-leucine. Clinical details of the study patients are showed in Table 1. All patients were clinically stable during the study period. No one had lung disease or signs of infection. Five patients required mechanical ventilation for more than 48 h for major surgery (two adults and three infants). Three patients needed mechanical ventilation for impaired neurological function (two adults and one infant), and four adult patients were on spontaneous breathing via a tracheostomy.

Table 1 Clinical characteristics of adult (panel a) and infants (panel b) patients
Full size table

SP-C purity

TAs were centrifuged to sediment cells and the chloroformic extract of the resulting supernatant was fractionated with Bond Elute NH2 columns as previously described [18]. Fractions eluted with mixtures of choloform:methanol:acetic acid 4:1:0, 4:1:0.025 (containing SP-B), and 1:4:0 and 1:9:0 (containing SP-C) were analyzed with LC-HRMS analysis. HR-MS spectra of intact (not subjected to hydrolysis) SP-B and SP-C are shown in Figs. 1 and 2, respectively. The LC column and the mobile phase choice were optimized for highly sensitive and selective SP-B and SP-C detection. A reversed-phase C4 column was chosen based on the size and on the high hydrophobicity of the two proteins since the aim of the investigation was the analysis of the two surfactant intact proteins. Isocratic conditions successfully resolved SP-B and SP-C within the TA samples and the purity of the peaks could be easily estimated. The most critical electrospray source parameters affecting ionization were the sheath and auxiliary gases which were experimentally tuned according to the signal intensity obtained for each protein molecule. A specific tune page was then set for selective detection of each proteins. For the purpose of the method, we aimed at determining the purity of the fractionated TA samples with respect to SP-B and SP-C. There are no standard reference surfactant protein B and C which would enable the determination of the analytical limit of detection of surfactant proteins within TA, and therefore a relative semiquantitation based on detectable peaks area was performed proving the application of the optimized parameters for each protein.

Fig. 1
figure 1

Panel A, HRMS spectrum of human SP-B homodimer at 3.09 min. Panel B, extracted ion chromatogram of [SP-B + 10H]10+ charge state

Full size image
Fig. 2
figure 2

Panel A, HRMS spectrum of human dipalmitoyl SP-C at 4. 02 min. Panel B, extracted ion chromatogram of [SP-C + 4H]4+ charge state

Full size image

Within fractions 1:4:0 and 1:9:0, we did not find any measurable amount of SP-B, which is primarily eluted in fractions 4:1:0 (>85 %) and 4:1:0.025 (15 %).

Kinetic data

In 4 of the 12 patients, the signal of 13C leucine SP-C was too weak (signal-to-noise ratio for m/z 210 below 300) for determining the MPE ratio. Moreover, in one infant, two of the TA samples collected had an insufficient amount of SP-C. Therefore, leucine-SP-C kinetics was successfully measured only in three of the four infants and in four of the eight adults.

Plasma leucine enrichment achieved the plateau within 6 h from the start of the study (Electronic Supplementary Material Fig. S1). Isotopic equilibrium was maintained during the remaining 18 h of 13C-leucine infusion. Mean enrichment value was 6.2 ± 1.0 MPE in infants and 3.8 ± 1.2 MPE in adults. The mean value was 6.3 ± 1.2 and 5.1 ± 0.8 MPE in infants and adults where SP-C kinetics was successfully measured, whereas it was 5.9 (patient 12) and 3.1 ± 0.2 (adults) MPE in those patients where SP-C kinetics could not be measurable.

We found a coefficient of variation <3 % for each plasma and TA sample measured in triplicate. Figure 3 reports SP-C isotopic enrichment curves in our patients together with SP-B curves obtained in the same patients in previous studies [18, 21]. As shown in Fig. 3, incorporation of 13C leucine into SP-C increased until 20–25 h after the start of isotope infusion with a PT of 24 h [20–27] like in SP-B. The maximum enrichment of 13C leucine was higher for SP-B than for SP-C. Labeled SP-C appeared in the alveoli in 6.0 (4.7–11.5) h with a time-lag similar of that of SP-B. Both hydrophobic surfactant proteins were turned over within 48 h. The FSR of these proteins was rapid and we calculate that 15.7 (14.1–27.5) % of SP-C was newly synthesized each day; this value was about one third of that of SP-B. All kinetics results are shown in Table 2.

Fig. 3
figure 3

13C leucine SP-C (empty circles) and 13C leucine SP-B (black squares) enrichment curves in adult patients (patients 1 to 4), and infants (patients 9, 10, and 11) with no lung diseases receiving a 24-h infusion of 13C-leucine. SP-B data are from [18, 21]. MPE mole percent excess

Full size image
Table 2 Kinetics data [18, 21]
Full size table

Discussion

Stable isotopes have been used to study the turnover of surfactant components [18, 26]. We now add the turnover of SP-C with the hope that this may help to clarify the pathophysiology of lung disease.

SP-C accounts for 2.1 and 1.3 % of total phospholipids in surfactant large aggregates obtained from term infants and adults with no lung diseases, while it is barely detectable in samples obtained from premature infants within the first days of life increasing to 4 % over subsequent week of life [2]. Existing evidence indicates that SP-C may not be critical for respiratory adaptation at birth but is important for normal postnatal lung function in special conditions like hyperoxia, infection, or lung immaturity [27, 8, 7, 28].

Studying the renewal of SP-C in newborn and adult subjects, we found that FSR, ST, and PT assumed a rather wide spectrum of values, possibly due to the heterogeneity of patients. Since all these patients have also been studied for SP-B and disaturated phosphatidylcholine (DSPC) kinetics [20, 21], we compared the turnover of SP-B and SP-C. Our study shows that labeled SP-C appears in the lungs after the same time lag of SP-B. This implies that these proteins have similar secretion pathways. We could also observe that SP-C turnover rate is lower compared to the one found for SP-B in adult and in infant patients (Table 2). SP-C FSR ranged from 22 to 55 % and from 6 to 31 % of that of SP-B in adults and in infants, respectively. Only patient 4 renewed the SP-C pool faster than that of SP-B (66.3 vs. 47.0 %/day). Kristensen et al. has recently found that proteins of the same functional class seems to have similar synthesis and degradation rates [29], so, since SP-B and SP-C are packaged together into the lamellar bodies, differences in turnover may be due to a different destiny of these two proteins after secretion into the alveolar spaces.

There are no data on the turnover of SP-C in humans and previous works on the clearance of SP-C administered into the trachea of mice and rabbits have given different results. During short-time experiments (0–4 h) in which labeled surfactant was administered into the trachea of newborn rabbits, Baritussio et al. [15] found that SP-C is lost from the airways and appears in the lamellar bodies more rapidly than DSPC. On the opposite in studies performed over tens of hours, Ikegami and colleagues reported similar rates of removal of DSPC and SP-C from adult rabbit lungs (half life of about 11 h), whereas in mice, SP-C was cleared at a slower rate (half life 28 h) than DSPC (half life 12 h) [17]. One possible explanation of these discrepancies is that short studies may reflect events happening in the alveolar spaces, while longer studies tend to reflect events happening in the whole lung.

In this study, the percentage of failures in obtaining workable enrichment curves was 25 % in infants, and 50 % in adults. Interestingly, the adults with lower plasma 113C-leucine enrichment were the ones in whom SP-C kinetics could not be assessed. Considering the dose of isotope, infants received 2 mg/kg/h of 113C-leucine, whereas adults received a fixed dose not calculated on their body weight (0.6 ± 0.2 mg/kg/h). Thus, it is possible that a different isotope dosing could allow measuring SP-C kinetics in more adult subjects. Alternatively, the use of a different isotope, like [5,5,5-2H3] leucine, could improve the detection of the enriched signal. The triply-labeled amino acid have a lower background signal (given by the natural abundance) than the single-labeled one, leading to a better precision of the leucine isotope ratio measurement. Limits of this study were also the fact that some TAs did not contain enough material for analysis and the low frequency of sampling that could be increased at early time points. To improve the SPs recovery from TAs, we have now standardized the suctioning technique and decreased the dilution of the sample. It is noteworthy that lung fluid obtained by bronchoalveolar lavage (BAL) reflects better the composition of the alveolar surface and could contain a higher amount of surfactant protein [30] than that obtained by TA that collects surfactant present in the upper ways. Unfortunately, BAL is an invasive procedure and is not acceptable for ethical reason to repeat a BAL procedure every 6 h as per requested by our study design. On the other hand, tracheal aspiration is a less invasive technique and is often performed several times daily as part of the care of intubated patients. Moreover, the surfactant pool accessible by tracheal aspiration has been recently demonstrated to have a similar phospholipids composition and to be on isotopic equilibrium with the alveolar surfactant pool [31], thus making the tracheal aspirate samples a reliable substitute of bronchoalveolar lavage sample for kinetic study.

Further studies are needed to optimize the dose or the type of the isotope and to better characterize protein recovery.

In summary, we showed that it is feasible to measure SP-C kinetics in vivo in humans by stable isotope techniques. Future studies will focus on the relationship between the nature and extent of selected lung disease and the turnover of SP-C to better understand the role of this unique lung protein.