The uptake of natural lung surfactant into differentiated type II cells may be used for the targeted delivery of other molecules. The fluorescent anion pyranine [hydroxypyren-1,3,6-trisulfonic acid, sodium salt (HPTS)] was incorporated into a bovine surfactant labeled with [3H]dipalmitoylphosphatidylcholine ([3H]DPPC). The uptake of [3H]DPPC and of HPTS increased with time of incubation and concentration, decreased with the size of the vesicles used, and was stimulated by 8-bromo-cAMP and partially inhibited by hypertonic sucrose. However, the amount of HPTS uptake was ∼100 times smaller than that of [3H]DPPC. This large difference was due to a more rapid regurgitation of some of the HPTS from the cells but not to leakage from the surfactant before uptake. The acidification of the internalized surfactant increased linearly over 90 min to 7.13, and after 24 h, a pH of 6.83 was measured. In conclusion, after internalization of a double-labeled natural surfactant, the lipid moieties were accumulated in relation to the anions, which were targeted to a compartment not very acidic and in part rapidly expelled from the cells.
- 8-hydroxypyren-1,3,6-trisulfonic acid
- intracellular pH
- microporous membranes
alveolar type ii cells are highly differentiated, polarized epithelial cells that are mainly responsible for the synthesis, the uptake, and the recycling of pulmonary surfactant (31). These functions are cell specific and tightly associated with the degree of the differentiation and possibly with the expression of the surfactant-specific proteins and of specific receptors on the cell surface (8, 9). Isolated type II cells cultured on plastic (6, 17) are nonresponsive to agonists, which effectively stimulate the uptake process in vivo or in the model of the isolated perfused lung (12). When cells are cultured on microporous membranes, they are well polarized and differentiated and respond to various agonists (6). Established cell lines from pulmonary epithelial cells also differ significantly with respect to the uptake of surfactant from type II cells in primary culture (19) and thus cannot be used. We therefore selected the system of differentiated type II cells cultured on microporous membranes for the study of the uptake processes.
A complete natural surfactant was used to overcome the complex interactions that exist between the different surfactant components, e.g., the inhibition of the surfactant protein (SP) A effect by SP-D (21) or of the SP-C effect by SP-B or SP-B plus SP-A (19). This surfactant was labeled in the main lipid component, dipalmitoylphosphatidylcholine (DPPC). As a second label, a small organic anion, 8-hydroxypyren-1,3,6-trisulfonic acid, sodium salt, pyramine (HPTS), was used. This is a water-soluble, pH-dependent fluorophore that is readily encapsulated into liposomal and surfactant vesicles and that reports intravesicular changes in pH (7). Low pH excitation at 403 nm and high pH excitation at 455 nm result in intense fluorescent emissions. Excitation of HPTS at 413 nm, the isobestic point, is pH independent, and the resulting fluorescence intensity allows the estimation of the total amount of cell-associated HPTS. The delivery of a pH reporter together with the surfactant to lung lamellar bodies and other key intracellular organelles may yield information on the pH in the compartments related to surfactant uptake and its intracellular routing. The other advantage of HPTS was that it represents a prototype for small organic anions. Encapsulation of these into the surfactant might be used for targeted and enhanced delivery into alveolar type II cells. As such, its behavior may be an example for the handling of this class of molecules by type II cells. In this study, the uptake of a natural surfactant and of the small organic anion HPTS into differentiated type II epithelial cells was directly compared with respect to their dependency on dose, on time, and on inhibitory and stimulatory factors to yield new insights into the variation of the pH in compartments related to the surfactant uptake. The results allow an estimate of the efficacy of the uptake processes and of the differential intracellular handling of the two labels.
MATERIALS AND METHODS
Preparation of cells.
Type II cells were isolated from the lungs of adult male Sprague-Dawley rats (200–300 g; Charles River, Sulzfeld, Germany) by the elastase digestion and immunoglobulin G panning, as detailed previously (16). The freshly isolated cells were resuspended in DMEM containing 10% FBS, 10 mg/ml streptomycin, and 100 U/ml penicillin and were plated at a density of 4.5 × 106 cells/insert of a microporous membrane (3-μm pore size, 106pores/cm2, surface area 4.7 cm2, 15-μm thickness, polyester; Costar, Bodenheim, Germany). The cells were added in a volume of 1.5 ml to the insert, which was in a dish containing 0.9 ml of culture medium. For comparison, cells were also cultured on membranes with 0.4-μm pores and also directly plated on 35-mm plastic dishes as described before (16). After culture for 18–20 h at 37°C in a humidified atmosphere of 94–95% air and 5–6% CO2, resulting in a pH of 7.4 of the medium, at least 95% of the cells were identifiable as type II cells by staining with phosphine 3R.
Chemicals and surfactant.
DMEM, Earle's balanced salt solution [EBSS (in mM): 116.4 NaCl, 1.8 CaCl2, 5.4 KCl, 0.8 MgSO4, and 1 NaH2PO4, pH 7.4] and FBS were purchased from GIBCO BRL (Karlsruhe, Germany); porcine pancreatic elastase was from Elastin Products (Owensville, MO); [3H]DPPC was from NEN (Frankfurt, Germany); and Optifluor was from Packard Instruments (Gronningen, The Netherlands). The following chemicals were purchased from Sigma Chemical (Deisenhofen, Germany): rat IgG, brefeldin A, cytochalasin D, 8-bromo-cAMP (8-Br-cAMP), ATP, and nigericin. HPTS, p-xylene-bis-pyridinium bromide (DPX), and rhodamine-1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine (DHPE) were from Molecular Probes (Leiden, The Netherlands). Polyclonal antibodies, primarily raised in rabbits against human SP-A and SP-B isolated from lavages of patients with alveolar proteinosis and against human recombinant SP-C and cross-reacting with bovine SP-A, SP-B, and SP-C, were a gift from Byk-Gulden (Konstanz, Germany). A natural surfactant preparation was prepared from lavages of bovine lungs by differential centrifugation on a sucrose gradient and was provided by Thomae (Biberach, Germany). A single lot was stored in appropriate aliquots for the experiments. All other biochemicals were from Merck (Darmstadt, Germany) or Serva (Heidelberg, Germany) at the highest purity available.
Labeling and characterization of the surfactant preparation.
The surfactant suspension (0.3 ml, 3.9 mmol of DPPC) was mixed with 1 ml of butanol containing 0.9 MBq of [3H]DPPC for 30 min on a vortexer (Baxter, Munich, Germany). In some experiments, rhodamine-DHPE at a final concentration of 2% of the total phospholipids by weight was used instead of [3H]DPPC. The mixture was thoroughly dried under a stream of nitrogen, followed by reduced pressure to remove traces of the organic solvent. The dried mixture was rehydrated in 2.5 ml of HPTS solution (75 mM HPTS, 77 mM NaCl, rest as in EBSS, but without HEPES and glucose, pH 7.4, same osmolarity as EBSS with glucose and HEPES) and sonicated with a probe sonifier HD800 equipped with an MS73D tip (Bandelin, Berlin, Germany) at 20% maximum output for 5–7 min at room temperature. The suspension was then cycled in polystyrol tubes (Falcon 2054; Becton Dickinson) for 3 h at 37°C at 40 rpm to achieve a reproducible mixture of surfactant vesicles of various sizes. After removal of very large aggregates by centrifugation (3,000 g, 8 min, 4°C), free HPTS was separated from HPTS incorporated in the surfactant liposomes by filtration over Sephadex G25 M PD-10 columns (Pharmacia, Freiburg, Germany) equilibrated with EBSS with 5.5 mM glucose and 25 mM HEPES. A first fraction (eluting volume 1.5–4 ml) contained larger particle fractions, and a second fraction (elution volume 4–6.5 ml) contained smaller particles. Free HPTS eluted between 8 and 12 ml. The exact size range differed somewhat from preparation to preparation and was characterized each time by laser light scattering (Lo-Sizer; Malvern, Seefeld, Germany) using the polydisperse fitting procedures described below. For comparison, some samples were also quantitated by negative staining and electron microscopy (24). Droplets of the surfactant suspension were placed on Parloidin- and carbon-coated grids, the surfaces of which were treated with poly-l-lysine, and excess surfactant suspension was drawn off with filter paper. After negative staining with phosphotungstic acid, electron micrographs were taken (TEM 900; Zeiss, Oberkochen, Germany), and liposome sizes were determined from printed enlargements by manual measurement of 1,000 liposomes for each of three preparations (24). Electron micrographs of the type II cells were prepared from araldite-embedded samples as described (15). The surface activity of this double-labeled surfactant was very high, as assessed in the pulsating bubble surfactometer (minimum surface tension at 1 mg/ml phospholipid after 5 min of pulsation was ∼0 ± 0 mN/m). The phospholipid composition was assessed by high-performance TLC (30). The content of the SP-A, SP-B, and SP-C was determined by ELISA (14, 27) and verified by semiquantitative SDS-electrophoresis (NuPage; Novex, San Diego, CA), immunoblotting (Immobilon P; Millipore, Bedford, MA), and detection by the antisera with the enhanced chemiluminescence assay (Amersham Life Science, Buckinghamshire, UK) with goat horseradish peroxidase-conjugated anti-rabbit polyclonal anti-IgG (Sigma) using purified bovine SPs as standards (SP-A, gift of Dr. W. Bernhard, University of Hannover, Germany, and SP-B and SP-C, gifts of Dr. R. Schmidt, University of Giessen, Germany).
Surfactant uptake experiments.
After the overnight culture, the cells were rinsed three times with EBSS without serum, phenol red, or antibiotics, but containing 25 mM HEPES and 5.5 mM glucose, pH 7.4, and were incubated for 30 min at 37°C. In experiments with various test agents, these were already present during this period. Next, the surfactant mixture at the appropriate concentration in EBSS (37°C, pH 7.4) was added, both to the upper and lower compartments of the dish. After incubation for the indicated time periods, the reaction was stopped, the medium was removed rapidly, the cells were washed three times with ice-cold EBSS without glucose, and 0.54 mM EDTA was added. After 5 min, the cells were gently dislodged from the membranes of the culture inserts and washed three times in cold EBSS without glucose by centrifugation. The final pellet was resuspended, and HPTS fluorescence was determined by spectrofluorometry at 4°C, as described below. Another aliquot was transferred to a scintillation vial, 10 ml of Optifluor was added, and radioactivity was measured in a liquid scintillation counter. A third aliquot was used to determine protein content after digestion with 0.1 mM NaOH by a Coomassie blue binding method using an assay kit supplied by Bio-Rad (Munich, Germany). The specific activity of the surfactant preparation for each label ([3H]DPPC/nmol DPPC and HPTS/nmol DPPC) was used to calculate the uptake of the natural surfactant, which was then expressed in nanomoles of DPPC per unit of type II cell protein. In all the experiments, culture dishes incubated overnight with DMEM containing serum and antibiotics but no cells were treated in parallel exactly as described above. Radioactivity and fluorescent signal recovered from these dishes were routinely <1–2% of that of the dishes with cells and thus were not subtracted from the corresponding cell-associated marker. No correction for autofluorescence of unlabeled cells was necessary.
The cells, resuspended in EBSS without phenol red and glucose at 4°C, were continuously stirred in a microcuvette, and fluorescence was determined in a luminescence spectrometer (LS 50B; Perkin-Elmer,Überlingen, Germany). Excitation spectra (λex350–500 nm, bandwidth 5 nm, 240 nm/min) were generated by recording emission of fluorescence at 515 nm (bandwidth 5 nm). Spectra were smoothed by binomial fitting, peak heights (λex, 405, 413, and 455 nm) were measured, and the 403- to 455-nm fluorescence excitation ratio was calculated.
Amount of HPTS.
Excitation at 413 nm, the isobestic point, is pH independent, and the resulting fluorescence intensity allows the estimation of the total amount of cell-associated HPTS. For calibration of surfactant uptake assessed by HPTS, HPTS-labeled surfactant was serially diluted, and surfactant content was quantitated by phosphorus assay (10) of the surfactant after lipid extraction. Fluorescence was determined after the addition of Triton X-100 (2.5%). A linear relationship over the whole range of fluorescent intensities (0–1,000 arbitrary units) was established (r = 0.998). Autofluorescence of the cells was in the range of 5–12 arbitrary units (n = 20) and was negligible. For a double-labeled surfactant preparation containing a known amount of phospholipids (nmol of DPPC determined by phosphorus assay in the lipid extract of the surfactant), the “specific activities” were determined, e.g., the amount of HPTS (fluorescence at 413 nm) and the amount of [3H]DPPC (determined by scintillation counting) associated with this DPPC. The uptake of the surfactant vesicles into the type II cells was expressed as nanomoles of DPPC per milligram of type II cell protein, derived from either the amount of fluorescence measured (“marker HPTS”) or the radioactivity measured (“marker [3H]DPPC”). This allowed a direct comparison of the uptake of the two labels.
In a low pH environment, fluorescent emission of HPTS excited at 403 nm is intense, whereas excitation at 455 nm only gives a weak signal. At a high pH environment, the opposite is true. Therefore, the ratio of 403 to 455 nm can be used to monitor intracellular pH, and between a pH of 5 and 7.4, it was found to be indirectly but linearly related to pH. For calibration, the cells were loaded overnight in the presence of 4.7 mM free HPTS and processed as in the uptake experiments. The final cell pellet was resuspended in K+-nigericin buffer (12 mM NaCl, 130 mM KCl, 1.8 mM CaCl2, 0.8 mM MgSO4, 5.5 mM glucose, 6 mM HEPES, and 7 μM nigericin), the pH was adjusted to the desired value, and HPTS fluorescence was determined. This calibration curve was used for the estimation of the intracellular pH under various experimental conditions.
Exposure to NH4.
To modulate intracellular pH, 10 mM NH4Cl was added to the cells. NH3 freely diffuses into the cells, combines with H+, and cannot leave the cell because of its charge, and the intracellular pH increases. After washing off the ammonia, these processes are reversible. The pH changes can be monitored from the changes of the HPTS fluorescence measured at 403 and 455 nm.
Determination of the leakage of HPTS from the vesicles.
Mixing DPX directly with HPTS quenches the HPTS fluorescence completely. HPTS-containing vesicles were incubated in EBSS containing 50 mM DPX and 90 mM NaCl. Leakage of HPTS from the surfactant vesicles was monitored by the loss of fluorescence, as HPTS leaving the cells was quenched by DPX. The addition of Triton X-100 led to complete leakage and total loss of fluorescence. In a second approach to detect free HPTS, HPTS incorporated in liposomes was separated from leaked-out, free HPTS by column chromatography, as described above for the preparation of the HPTS-labeled surfactant.
The cells were incubated for 0 or 240 min either with the HPTS-surfactant mixture (200 μM) or with free HPTS (0.47 mM). After washing of the cells on the inserts, the insert membranes were cut out and fixed on the bottom of a temperature-controlled glass slide by a coverslip on the inverted phase-contrast microscope (IMT 2; Olympus, Hamburg, Germany). Fluorescence of HPTS was visualized by excitation at 450–490 nm, emission at 515 nm, and a dichroic mirror at 505 nm (XF19 filter cube; Omega Optical, Brattleboro, VT). For the rhodamine-labeled surfactant, a G filter cube with excitation at 450–550 nm, emission at 590 nm, and a dichroic mirror at 570 nm was used (Olympus). Images were recorded by a SITS camera (Hamamatsu, Herrsching, Germany), stored on a videotape, and digitized and processed with National Institutes of Health (NIH) image software (NIH, Bethesda, MD). NIH image is a public-domain (http://rsb.info.nih.gov/nih-image/about.html) image processing and analysis program of which version 1.58 for the Macintosh was used.
Cells were cultured overnight in the presence of 306 ± 16 μM surfactant vesicles (size 912 ± 125 nm) labeled with [3H]DPPC and HPTS. After three rinses with EBSS, the dishes were incubated with fresh EBSS without phenol red and incubated for the indicated time periods. The medium was removed, the cells were washed three times by centrifugation (180 g, 10 min, 4°C), and the amounts of cell-associated [3H]DPPC, HPTS, and protein content were quantified.
Determination of surfactant particle size.
The size of surfactant particles was determined by photon correlation spectroscopy using a laser-equipped Lo-Sizer (Malvern). A multimodal exponential sampling algorithm was employed for the calculation of mean vesicle diameters and width after 10 min of measurements at an angle of 90° and a temperature of 37°C.
Lactate dehydrogenase assay.
The rate of lactate dehydrogenase release in medium was monitored to assess cellular integrity. The cells were cultured as in the secretion experiments, after which lactate dehydrogenase activity in the cells and medium was assayed spectrophotometrically by following the disappearance of NADH at 340 nm.
Type II cells were isolated from the pooled lungs of four to five rats and distributed among the various treatment groups. Two dishes with cells and two blank dishes without cells per group were used in each uptake experiment. Each dish was processed separately, and the values were averaged to yield a single data point per group per experiment. All data are means ± SE from the number of individual experiments indicated. The data were analyzed statistically as indicated inresults.
Characterization of the type II cells cultured on different supports and of the surfactant preparation.
Electron-microscopic analysis of type II cells cultured for 18–20 h on 3-μm polyester microporous membranes revealed that the cells had the morphological characteristics of differentiated type II pneumocytes, with a clear polarization of the organelles, the nucleus at the basal side, and lamellar bodies apically. This was not the case when the cells were cultured on membranes with 0.4-μm pores or on tissue culture plastic (data not shown). The cells cultured on 3-μm pores were also functionally superior. Uptake of lipid after 90 min was increased about threefold compared with cells cultured on 0.4-μm membranes (3.8 ± 0.1 vs. 1.2 ± 0.1 nmol DPPC/mg protein). In addition, the cells cultured on 0.4-μm membranes did not respond to secretagogues (see below).
The natural surfactant double labeled with [3H]DPPC and HPTS was highly surface active and had a phospholipid composition that was characteristic of pulmonary surfactant (68 ± 2% phosphatidylcholine, 10 ± 1% phosphatidylglycerol, and 2.8 ± 0.7% sphingomyelin,n = 5). The protein-to-phospholipid ratio was 0.18 ± 0.02 μg/μg, the SP-A-to-phospholipid ratio was 0.14 ± 0.03 μg/μg, the SP-B-to-phospholipid ratio was 0.0015 ± 0.0003 μg/μg, and the SP-C-to-phospholipid ratio was 0.04 ± 0.01 μg/μg (each n = 6). The first fractions eluting from the column to prepare the surfactant particles containing the larger particles (larger than ∼1,000 nm) and the second fractions containing the smaller particles (smaller than ∼1,000 nm) were not different in protein-to-phospholipid ratio, the SP-A-to-phospholipid ratio, and the SP-C-to-phospholipid ratio. On average, the SP-B-to-phospholipid ratio was larger in the first fractions (0.0037 ± 0.0005 μg/μg) compared with the second fractions (0.0012 ± 0.0002 μg/μg, n = 6, P< 0.05). The size of the vesicles was determined by laser light scattering for each preparation to define the particles to which the cells were exposed with respect to size. The variation from experiment to experiment was not very large, as indicated by the mean and SE of the particle sizes given in the legends to Figs. 1-7. These size measurements were in close agreement with those obtained from electron microscopic images. Particles in both fractions were mostly liposomes of unilamellar appearance [first fraction 475 ± 28 nm (interquartile range 180–950), second fraction 1,083 ± 42 nm (interquartile range 600–1,400); n = 3]. Corresponding size by laser light scattering for these preparations was 495 ± 54 and 1,024 ± 100 nm (mean ± width of peak). In addition to particle size, the lipid concentration was assessed for each preparation and taken into consideration when setting up the experiments.
Uptake of natural surfactant assessed with [3H]DPPC as the marker.
Uptake of [3H]DPPC by the cells was time dependent (Fig.1 A) and concentration dependent (Fig. 2 A). There was a dependency on vesicle size, with the vesicles smaller than 1,000 nm being taken up more avidly (Fig.3 A). Because of partial inhibition of uptake by cytochalasin D and hypertonic sucrose, uptake was shown to be dependent on actin filaments and endocytosis via coated pits, processes that are blocked by these treatments (Table1). However, other mechanisms must be operating, since even a combination of the two treatments inhibited uptake by <50% (Table 1). Stimulation of the cells with 8-Br-cAMP, but not with the purinergic agonist ATP, stimulated uptake ∼1.5-fold, demonstrating agonist regulation of uptake (Table 1).
Uptake of free HPTS and of HPTS encapsulated in natural surfactant.
Free HPTS was taken up ∼100 times less than HPTS incorporated in the surfactant vesicles at the same concentration (Table2). After 24 h, the pH indicated by HPTS was significantly lower when free HPTS was used instead of HPTS incorporated in surfactant vesicles. This was not due to changes of the fluorescence response of HPTS in the presence of the lipids or cells because the same results were obtained when the lipid membranes were disrupted by treatment with Triton X-100 (data not shown). Thus HPTS leaking from surfactant vesicles could not significantly contribute to HPTS uptake, and HPTS in surfactant vesicles was likely to be located in an intracellular compartment different from that of free HPTS.
A striking finding of these experiments was the 10- to 100-fold lower apparent uptake of the surfactant vesicles when HPTS was used as a marker instead of [3H]DPPC. Although the general characteristics of time and concentration dependency and the dependency on size and on the modulation by blockers and secretagogues remained (Figs. 1 B and 2 B and Table 2), the ratio between the two markers was not constant, suggesting a differential handling of HPTS and the natural surfactant.
Taken together, these results showed that the natural surfactant was internalized in differentiated type II cells by a regulated process at a relatively high rate and that the uptake of the small anion HPTS was substantially increased by encapsulation in the surfactant. However, unexpectedly, the rate of surfactant lipid uptake by far exceeded that of HPTS uptake.
Possible reasons responsible for this discrepancy between the two labels, e.g., leakage of HPTS from the surfactant vesicles during the process of internalization or a preferential excretion of HPTS from the cells, were addressed in further experiments.
Leakage of HPTS from the surfactant vesicles.
Although the label [3H]DPPC is localized in all the lipid membranes of the surfactant preparation, HPTS is dissolved only in the internal watery contents of the vesicles and as such may leak out of the vesicles. The surfactant mixture was labeled with HPTS and DPX (50 mM) to quench the HPTS signal. Incubation of this mixture with type II cells did not induce a significant leakage of HPTS, e.g., no increased signal from HPTS via reduced DPX quenching was observed (4 experiments, data not shown).
Leakage was further quantitated by other experiments in which free and encapsulated HPTS were separated by gel chromatography after incubation of the HPTS-surfactant in the presence or absence of cells. No free HPTS that was liberated from the vesicles could be detected (data not shown). In summary, leakage could not account for the observed differences between uptake of HPTS and [3H]DPPC.
Preferential and rapid reexcretion of HPTS from the cells.
After incubation of the type II cells with double-labeled surfactant, the secretion of 3H label and of HPTS from the cells was monitored (Fig. 4). Initially, the cells lost HPTS much more rapidly; from 15 min onward, the kinetics were identical (Fig. 4). These data suggested that there may be an active transporter for HPTS or similar small anions operating in these cells. In separate experiments, probenecid was used, which is an organic anion transport inhibitor (28, 29). In the presence of probenecid, net uptake in the type II cells was significantly increased (Fig. 5). These data clearly suggested the presence of an outward transporter for HPTS or similar anions in type II cells. Thus such a transport mechanism may contribute to the observed differences between the uptake of HPTS and [3H]DPPC.
The intracellular location of HPTS was proven by exposure of the cells to NH4Cl, which resulted in a rapid increase of the pH reported, from pH 7.14 ± 0.01 to pH 7.35 ± 0.01, which was completed within 2 min. After washing of the cells, the original pH of 7.14 ± 0.02 was rapidly reestablished (n = 2).
HPTS reported a continuous decrease of pH starting from 7.4 to ∼7.1 after 90 min, having reached a minimum of 6.7 after 24 h of culture (Fig. 1 D). The pH was not changed by increased amounts of HPTS internalized in surfactant vesicles or by uptake of vesicles of different size. Neither stimulation of the cells by ATP or 8-Br-cAMP nor inhibition of uptake by hypertonic conditions or cytochalasin D changed the pH reported by HPTS. Similar to a short-term exposure to NH4Cl (see above), continuous NH4Cl exposure of the cells increased the pH. Finally, even after up to 75% of the initial cellular HPTS content was secreted, the pH reported by HPTS remained constant, indicating that the compartment in which the majority of the HPTS was localized and from which it was secreted remained unchanged (Fig. 4).
Fluorescence microscopy of the cells cultured on the microporous membranes also clearly demonstrated significantly increased uptake of HPTS encapsulated in natural surfactant compared with free HPTS (Fig.6). The distribution of the fluorescence appeared to colocalize only in part with the lamellar body inclusions of the cells. In addition, the cytoplasm surrounding the lamellar bodies was fluorescent (Fig. 6). This pattern was more evident in experiments with a fluorescently double-labeled surfactant. HPTS was encapsulated in a surfactant preparation labeled in the lipid moiety with rhodamine-DHPE instead of [3H]DPPC. After uptake, the rhodamine-DHPE fluorescence was mainly, but not entirely, colocalized with the HPTS fluorescence (data not shown). The same was observed in experiments in which the cells were incubated overnight with the double-labeled surfactant and monitored for the release of the two markers (Fig. 7). The rhodamine-DHPE fluorescence had a distribution similar to that of the lamellar bodies in the phase-contrast images, whereas the HPTS fluorescence was localized in both the lamellar body compartment and additional cytoplasmatic compartments. With ongoing time and secretion, the HPTS was lost from the latter more rapidly, and the colocalization with the lamellar bodies became more apparent (Fig. 7).
When cultured on 3-μm microporous membranes, type II cells were well polarized, and the uptake process responded to stimulatory factors, e.g., 8-Br-cAMP or to inhibitory factors, e.g., cytochalasin D and hypertonic conditions that block coated-pit endocytosis. Although the latter treatment was the strongest inhibitor of uptake in this study, the degree of inhibition was smaller than that observed previously with liposomes in the absence of SPs (23). In the study of Muller et al. (23), cells cultured on plastic were used. Compared with cells cultured on plastic or compared with uptake of simple synthetic liposomes, the natural surfactant was internalized about two to three times better (3, 6, 11,17). Reasons for this include the more differentiated and polarized phenotype of the cells (6) and the presence of SP-A, -B, and -C in the natural surfactant (17, 20, 26,32). Stimuli that are known to increase the secretion of surfactant, like 8-Br-cAMP (2), ATP (16), and alkalization of the cytoplasm (2), had different effects on uptake. Although 8-Br-cAMP increased uptake, ATP had no effect, and alkalization with ammonia reduced the uptake of radiolabeled surfactant. There appears to be a complex relationship between uptake and secretatory processes. A blocker of the anterograde secretatory transport of material from the Golgi apparatus, brefeldin A, stimulated surfactant uptake. Because both the cellular synthesis of the major surfactant lipid phosphatidylcholine and its packaging into lamellar bodies were unaltered by brefeldin A (5), the inhibition of the delivery of the SPs from the Golgi to the cell membrane or to the lamellar bodies by this compound, as previously shown (1), might lead to an upregulated endocytosis of external surfactant. Alternative explanations, e.g., indirect effects via other mechanism, cannot be excluded. The smaller surfactant particles were preferentially taken up (Fig. 3, A and B). Especially from the HPTS data (correlation coefficients in the legend to Fig. 3), it appeared that there was almost an inverse linear relation between size and uptake. In a previous study on type II cells cultured on plastic, uptake of the smaller surfactant vesicles of ∼100-nm diameter was threefold higher than that of larger particles of ∼200–2,000 nm (18). Compositional differences between the vesicles were unlikely to be responsible for the observed size dependency because the ratios of protein, SP-A, and SP-C to phospholipids were the same for the smaller and the larger particles and the relative content of SP-B was somewhat higher in the larger ones. Based on previous results that the hydrophobic surfactant lipids, e.g., SP-B, increase lipid uptake in vitro, an enhanced uptake of the larger particles could have been anticipated (26). In addition, the relative contribution of SP-B to the total SPs was small. Taken together, the uptake of a natural surfactant occurred in a time-, size-, and dose-dependent manner, was blocked by intracellular alkalosis and inhibition of coated pit-dependent endocytosis, and was stimulated by 8-Br-cAMP and brefeldin A.
The natural surfactant complex may be used to enhance the delivery of small organic molecules to the cells. We choose the strongly fluorescent compound HPTS, which was also sensitive to pH-dependent changes, as both a potential marker for surfactant handling and a prototype for small organic ions. Although the internalization of HPTS was increased ∼100-fold compared with the nonencapsulated material, its delivery to the cells was still 30- to 300-fold less effective than that of the surfactant lipid components. There are several potential explanations for this discrepancy. First, HPTS was rapidly expelled from the cytoplasm by probenecid-sensitive transporters. Such a transporter for small organic anions has been described in epithelial cells of the kidney and recently also for alveolar type II cells (29). Support for this explanation comes from the initial more rapid and larger secretion of HPTS from the cells compared with the surfactant lipids and from the severalfold increase of the cell association of HPTS in the presence of probenecid. Second, leakage of HPTS from the interior of the surfactant vesicles during the process of internalization may result in lower levels of HPTS. If at all, this mechanism may contribute only to a rather (<10%) small extent to explain the difference between HPTS and surfactant lipid uptake, as indicated by the experiments that did not find much evidence for cell-induced leakage of the surfactant vesicles. Last, a microheterogeneity of the HPTS distribution within the natural surfactant might play a role. A preferential uptake of fractions other than the HPTS-containing vesicles, e.g., lamellar or disklike structures containing the lipid label [3H]DPPC but no HPTS, might occur. Such an explanation is hypothetical and unfortunately cannot easily be tested by experiments. Taken together, the delivery of small organic anions to alveolar type II cells can be increased considerably with natural surfactant as a vehicle to access the cells and by the use of organic anion transport inhibitors to prevent their efflux.
HPTS was also used to monitor the intracellular fate of the HPTS-surfactant mixture, as indicated from the observed changes in pH. A continuous drop from pH 7.40 to 7.13 indicated that the compartment reached after 30 min was somewhat, but not very, acidic. The addition of NH4Cl reversibly increased the pH and demonstrated the presence of active proton transporters in these compartments. Irrespective of the size of the surfactant structures used, the same compartment with respect to pH was targeted. However, when incubated overnight, the HPTS-surfactant mixture was deposited in a more acidic compartment with an average pH of 6.87 (Fig. 4 and Table 2). Interestingly, free HPTS alone localized in an even more acidic intracellular compartment, indicating that the surfactant component influenced the site of the delivery. The various stimulatory or inhibitory interventions that modulated the extent of the internalization, however, had no effect on intracellular pH, except for NH4Cl, which led to a modest alkalization of the cells.
Basal intracellular pH in cultured type II pneumocytes has been reported to be ∼7.17 (5), 7.19 (25), 7.23 (13), and even 7.50 (22), depending partly on the extracellular environment (25). The compartment into which the HPTS encapsulated in natural surfactant was internalized, was somewhat more acidic, and had a pH of ∼7.12 after 90 min of uptake; after 24 h, the pH of 6.87 was reported (Table 2). These data suggested that the compartment reached was not end lysosomal, since much lower pH values are maintained. The measurements are in agreement with our secretion studies (Fig. 4) in which a pH of ∼6.83 was measured. Likewise, not all, but a significant fraction, of the HPTS was localized in the lamellar body compartment under these conditions (Fig. 7 H). In isolated lamellar bodies, a pH of 6.1 or below may be generated; however, those measurements were made in an artificial environment (4).
Our fluorescence microscopy data suggested that, after 90 min, much of the HPTS-surfactant mixture taken up was localized in cellular organelles, with a distribution similar, but not completely identical, to that of the lamellar bodies. This suggested targeting of HPTS to lamellar bodies and, in addition, to other compartments, the exact nature of which was not further studied. In contrast, free HPTS was taken up much less and clearly was more diffusely distributed in the cytoplasm. Secretion from the type II cells over time was visualized with a surfactant double labeled with lipid-encapsulated HPTS and with fluorescent rhodamine-DHPE in the lipid moiety. After overnight incubation, the HPTS fluorescence was distributed both in circumscribed structures colocalizing with the lamellar bodies in the phase-contrast images and diffusely in the cytoplasm of the cells. Initially, the loss of HPTS occurred from the latter compartment(s) more rapidly, and its colocalization with the lamellar bodies over time became much more apparent, supporting the biochemical measurements (Figs. 4 and 7). In the same lines, the increased ratio of [3H]DPPC to HPTS with time may be explained from the secretion of material coming mainly from the lamellar bodies. The lower pH in these studies compared with the uptake experiments indicated that the proton pumps take some time to generate a low pH in these compartments after the delivery of large amounts of exogenous surfactant. The significant fraction of the HPTS encapsulated in surfactant, which colocalized with the rhodamine-labeled lipids and the lamellar bodies in the phase-contrast images, suggested that the pH reported was to a substantial portion related to the pH in the lamellar bodies. However, these microscopy data together with the biochemical results on the differential uptake and the initial rapid secretion of HPTS from type II cells suggested that the intracellular routing of HPTS and surfactant DPPC was in part different, precluding unbiased direct conclusions on the pH in the intracellular compartments used for surfactant delivery.
Taken together, the regulated uptake of natural surfactant in differentiated type II cells may be used for the targeted delivery of other small water-soluble molecules to these cells. However, a differential cellular handling of the individual components must be anticipated. In contrast to the lipid structures, a large amount of the water-soluble label of the surfactant vesicles was rapidly expelled from the cells by probenecid-sensitive transporters. The pH of the compartment into which the HPTS-labeled natural surfactant vesicles were transported could thus not be determined unequivocally with this approach. Although not yet available, a fluorescent pH reporter covalently linked to lipid surfactant components appears to be more promising, provided that it does not change intracellular routing.
This work was done as parts of the theses of Astrid Baatz and Barbara Deubzer.
This work was supported by the Deutsche Forschungsgemeinschaft Grants 970/3–2 and 3–3.
Address for reprint requests and other correspondence: M. Griese, Childrens' Hospital, Ludwig-Maximilians-Univ., Pettenkofer Strasse 8a, 80336 Munich, Germany (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2001 the American Physiological Society