Introduction

Abstract Introduction References

Despite remarkable progress in its treatment, respiratory distress syndrome (RDS) remains the most important cause of neonatal mortality and morbidity. A deficit in pulmonary surfactant is thought to be one of the major underlying causes of RDS. Surfactant, which is produced by alveolar epithelial type II cells, consists of both phospholipids and proteins; it forms a continuous film over the luminal aspect of the alveolar epithelial cells, allowing surface contact to be broken and the alveolus to expand and fill with inspired air (1). The major component, and the one responsible for surfactant's unique biophysical properties, is disaturated phosphatidylcholine (DSPC). The four specific surfactant proteins, SP-A, -B, -C, and -D, have a variety of functions, not all of which are related to phospholipids (1). Respiratory function at birth depends on a sufficient amount of lung surfactant. Although the control of surfactant synthesis in the developing fetal lung is multifactorial and is not completely understood, mesenchymal- epithelial interactions appear to be of major importance (2). Therefore, there is considerable interest in identifying regulatory substances that mediate mesenchymal control of epithelial maturation.

Growth factors have been shown to influence various aspects of lung development (2). One of the most recently considered is keratinocyte growth factor (KGF), otherwise designated fibroblast growth factor (FGF)-7. KGF was initially isolated from a fetal lung-derived fibroblast line (3). It is expressed in stromal cells, whereas its receptor KGF-R or FGF-R2 IIIb (4) is found only in epithelia, including lung epithelium (5). KGF is a mitogen for type II cells in the developing and the adult lung (6), and KGF-R plays a key role in early lung branching morphogenesis (9, 10). Overexpression of KGF in vivo (11) and exogenous KGF in vitro (12, 13) both disturbed lung morphogenesis, but seemed to delay type II cell maturation in the former instance and conversely to enhance it in the latter. It is not known whether KGF has a role in the control of surfactant synthesis, especially during late gestational lung maturation, when surfactant accumulates in preparation for extrauterine life. Recently it was suggested that KGF may increase the surfactant pool in the newborn rabbit in vivo, but only whole-tissue DSPC was found to be increased without regard to the nature of the containing cell type, whereas no increase in alveolar DSPC was observed (14). Regarding SPs, contradictory findings have been reported. In isolated adult rat type II cells, exogenous KGF stimulated the expression of SP-A and SP-B at a pretranslational level, but had little effect on SP-C expression (15). For the developing lung, if KGF has been reported repeatedly to enhance SP-B expression, some investigators using the whole lung primordium have shown no effect on SP-A and strong stimulation of SP-C expression (12), whereas others using the epithelial structure separated from mesenchyme conversely found advanced SP-A expression but reduced SP-C expression (13). It should be emphasized, however, that these studies were conducted with the use of embryonic anlages explanted at a stage far earlier than the period of surfactant accumulation.

To explore this question further, we therefore examined the effects of KGF on the rate of surfactant phospholipid synthesis, particularly that of DSPC, and on SP expression by fetal rat type II cells isolated during their maturation period. Cells were cultured on Engelbreth- Holm-Swarm (EHS) tumor basement membrane matrix, a condition shown to be optimal for maintaining the type II cell phenotype (16, 17) and which has been used for studying KGF effects on adult type II cells (15) as well as on the isolated epithelial lung primordium (13). KGF was compared with epidermal growth factor (EGF), which is known to stimulate surfactant synthesis in vivo and in vitro (18, 19); with acidic FGF (aFGF or FGF-1), which binds KGF-R with high affinity (4); and with hepatocyte growth factor/scatter factor (HGF/SF), which is a mitogen for type II cells (6, 20). We show that KGF is a stimulator for all surfactant components and may therefore participate in the mediation of mesenchymal cell-stimulating activity.

	Materials and Methods

Animals

Virgin Wistar female rats (Charles River, Saint-Aubin-Lès-Elbeuf, France) were mated overnight in the laboratory. The following morning was designated Day 0 of gestation. Pregnancy was checked by palpation 14 d later (term is 22 d). Rats were anesthetized with pentobarbital and laparotomized, and uterine horns were removed en bloc and transferred to a laminar-flow hood where fetuses were extracted and fetal lungs removed under aseptic conditions.

Cell Isolation and Culture

For studying surfactant protein expression, immature cells were isolated from the lungs of 18-d-old rat fetuses, that is, at a developmental stage that immediately precedes the start of surfactant synthesis (2). The number of cells recoverable at this stage was insufficient, however, to allow the experiments relative to phospholipids to be performed. For this purpose, cells were isolated from the lungs of 20-d-old rat fetuses, that is, at a developmental stage that precedes the phase of maximum surfactant accumulation (2). Fetal lung cells were enzymatically dispersed, and type II cells were purified by differential adhesion to plastic and low-speed centrifugation (21). Type II cells were resuspended in minimum essential medium (MEM) + 10% fetal bovine serum, counted, seeded (0.5 × 10⁶ cells/cm²) in multiwell plastic culture plates coated with EHS basement membrane matrix prepared in the lab (22) or on other substrata as indicated in RESULTS, and allowed to adhere overnight under air-CO₂ (95%-5%). Cells were then rinsed and experimental media were introduced. Experiments were conducted in a defined culture medium (hereafter designated control medium) designed on the basis of previous experiments (23), in which growth factors were added at various concentrations. This consisted of Dulbecco's modified MEM (GIBCO-BRL, Eragny, France) containing Na pyruvate and enriched with nonessential amino acids, transferrin (5 mg/liter), biotin (100 mg/liter), Na selenite (10 mg/liter), FeCl₃ (1 mg/liter), ZnCl₂ (30 mg/liter), MnCl₂ (10 mg/liter), CuSO₄ (25 mg/liter), penicillin (100,000 IU/ liter), and streptomycin (100 mg/liter). All additives were cell-culture tested (Sigma, L'Isle d'Abeau-Chesnes, France). Cells were removed from the culture substratum by dispase (Sigma; 10 mg/ml) at 37°C, or (for enzyme assays) in Matrisperse (Becton Dickinson, Le Pont de Claix, France) at 0°C, and then pelleted (4,000 × g, 10 min).

Growth Factors

Human recombinant KGF and aFGF were purchased from Pepro Tech Europe, Inc. (London, UK). EGF was a natural factor extracted from mouse submaxillary glands (Serva, Heidelberg, Germany). Human recombinant HGF/SF was provided by Sigma.

DNA Assay

Cell DNA was determined fluorometrically (23) using the fluorescent dye Hoechst 33258 ( excitation: 350 nm;  emission: 455 nm). In the instance of nuclei pelleted from the isolated surfactant fraction, DNA was determined by the colorimetric diphenylamine method (24).

Incorporation of [³H]Choline and [³H]Acetate

For evaluating the rate of phospholipid synthesis, type II cells were exposed either to [methyl ³H]choline chloride (0.5 mCi/ml [18.5 kBq/ml], 80.7 Ci/mmol [3.03 TBq/mmol]), or to [³H]acetate (2 mCi/ml [74 kBq/ml], 2.18 Ci/mmol [80.7 GBq/mmol]; Amersham, Les Ulis, France) in defined medium. At the end of the incorporation period, the radioactive medium was removed and cells were rinsed twice with phosphate-buffered saline before treatment with dispase. The cell pellet either had its lipids extracted by chloroform/methanol/water, 1:2:0.8 (vol/vol/vol) in the presence of unlabeled lung tissue homogenate (carrier tissue), or the surfactant fraction (SF) was extracted.

Surfactant Fraction

A density-gradient fractionation technique adapted for surfactant extraction from cultured cells and quantitative analysis (25) was used. The fraction isolated by using this method has been shown to collect essentially lamellar bodies, the intracellular form of surfactant, and to exhibit all the characteristic biochemical and biophysical features of surfactant (26). In brief, cell homogenates prepared in the absence (for phospholipid assay) or in the presence (for incorporation studies) of carrier fetal lung tissue in isotonic Tris/ethylenediaminetetraacetic acid/NaCl (TEN) buffer were deposited on 0.75 M sucrose and centrifuged for 1 h at 48,000 × g. Material at the interface was collected, diluted with TEN, deposited on a discontinuous sucrose gradient (0.25 M and 0.68 M), and centrifuged for 1 h at 64,000 × g. The final SF was banded over 0.68 M sucrose. The so-called residual fraction (RF) consisted of pellets and supernatants of the two centrifugation steps, and contained all of the cell membrane compartments other than lamellar bodies. Lipids were extracted as described, from both fractions.

Separation, Assay, and Determination of Radioactivity of Phospholipids

A trace amount of [¹⁴C]dipalmitoylphosphatidylcholine (113 mCi/mmol [4.2 GBq/mmol]; Amersham) was added at the time of lipid extraction for determination of recovery. For assessment of choline incorporation and determination of DSPC content, DSPC was extracted either from whole-cell extracts or from fractions by the osmium tetroxide method and thin-layer chromatography separation on silica gel 60 chromatography plates (Merck, Darmstadt, Germany) in chloroform/methanol/water, 65:25:4 (vol/vol/ vol) (27). To measure acetate incorporation, individual phospholipids were separated from the SF by one-dimensional thin-layer chromatography in a solvent system of chloroform/ hexane/methanol/glacial acetic acid/water, 12:7:4:3:0.3 (vol/ vol/vol/vol/vol) (28). Phospholipids identified by comparison with standards run in parallel were eluted from gels by chloroform/methanol/water 1:2:0.8 (vol/vol/vol), dried, and redissolved in chloroform/methanol, 2:1 (vol/vol), and activity was counted in Optiscint scintillation cocktail (EEG Instrument, Evry, France) with the aid of an LKB rack  counter, using a double-channel disintegrations per minute (dpm) program. Total phospholipid and DSPC contents were determined through phosphate determination in mineralized samples (25).

Choline Phosphate Cytidylyltransferase and Fatty Acid Synthase Assays

Choline phosphate cytidylyltransferase (CPCT) (EC 2.7.-7.15), the enzyme that catalyzes the principal rate-limiting step in the biosynthesis of phosphatidylcholine (PC) (29), was assayed in cultured-cell homogenates by measuring the rate of incorporation of phosphoryl[¹⁴CH₃]choline (50 mCi/mmol [1.85 GBq/mmol]; DuPont NEN, Les Ulis, France) into cytidine triphosphate (CTP) to form cytidine diphosphate (CDP)-choline according to Weinhold and colleagues (30). Fatty acid synthase (FAS) (EC 2.3.1.85) was assayed spectrophotometrically by measuring the rate of nicotinamide adenine dinucleotide phosphate (NADPH) disappearance at 340 nm (31). Activities were normalized to the protein concentration in homogenates as measured with the Bradford assay kit of Bio-Rad (Ivry sur Seine, France) and bovine serum albumin as a standard.

Isolation of RNA and Northern Blot Analysis

Total RNA was isolated from pelleted cells using the guanidium isothiocyanate procedure (32). RNA, 20 to 30 µg, was fractionated by electrophoresis through 1% agarose/2.2 M formaldehyde gels and blotted onto nylon membranes (Gene Screen, DuPont NEN). The CPCT probe was a 950-base pair (bp) insert encoding rat liver CPCT (33). The FAS probe was obtained from clone FASg57pB1.8 encoding rat spleen FAS (gift from Dr. S. D. Clarke, Colorado State University, Fort Collins, CO). SP-A, SP-B, and SP-C complementary DNA (cDNA) probes were amplified by reverse transcription-polymerase chain reaction (RT-PCR) from RNA extracted from adult rat lung. The primers used were: for SP-A, 5'-GGAAGCCCTGGGATCCCTGGA-3' and 5'-TGGTGTGGTTGACCATGGGTC-3' amplifying a 552-bp fragment between positions 88 and 639; for SP-B, 5'-GCTACTGCTCCTTCCTAC-3' and 5'-AGAGGTGTGGGGTTTGGA-3' amplifying a 1,107-bp fragment between positions 30 and 1,136; and for SP-C, 5'-ATGGGTAGCAAAGAGGTA-3' and 5'-GAGTATGGACAGGAGCAG-3' amplifying a 673-bp fragment between positions 19 and 691. PCR products were separated by agarose gel electrophoresis and eluted from gel using GenElute Spin Columns (Supelco, Bellefonte, PA). Probes were labeled with [-³²P]dCTP (ICN, Irvine, CA) using the Rediprime DNA labeling system from Amersham, and purified on NucTrap probe purification columns (Stratagene, Cambridge, UK). Conditions for prehybridization, hybridization, and stripping of blots were as previously described (34). To allow correction for variations in the amount of RNA loaded, the blots were also hybridized with an 18S ribosomal RNA (rRNA) probe. Autoradiographs were made by exposing blots at 80°C to X-ray film (Reflection; DuPont NEN) with intensifying screens. Quantification of signal intensity was performed by densitometric analysis of autoradiograms using the NIH Image program.

Statistical Analysis

Data are presented as mean ± SEM. Multiple comparison of mean values was made by analysis of variance (ANOVA) (Fisher's protected least significant difference test). Two-variable comparison was made by two-tailed t test. In both instances, P = 0.05 was considered as the limit of statistical significance.

	Results

DNA Synthesis

Type II cells cultured in control medium did not increase in number: for 10⁶ seeded cells, the amount of DNA extracted after 48 h (6.6 ± 2.4 mg) was not different from the theoretical DNA content of 10⁶ rat cells (6.2 pg per diploid cell in this species [35]). As shown in Table 1, all growth factors studied tended to increase DNA synthesis. The difference with defined medium was significant, however, only for KGF 50 ng/ml (+41%) and for aFGF 50 ng/ml (+47%). In the instance of cultures used for surfactant content determination (see below), DNA amount determined by the diphenylamine method was found to be unchanged in the presence of KGF 25 and 50 ng/ml or of EGF 20 and 50 ng/ml (not shown).

[³H]Choline Incorporation into Whole-Cell DSPC

In control medium without growth factor, choline incorporation was sustained over 48 h. The incorporation rate remained almost constant over 24 h (about 950 dpm/10⁶ cells/h) and declined slightly between 24 and 48 h. Time-response study indicated that KGF (50 ng/ml) induced a stimulation that became detectable from time 8 h (8%) and increased gradually (40% at 24 h and 200% at 48 h). Dose-response study at 48 h for the various growth factors is presented in Figure 1. Incorporation increased linearly with KGF concentration between 1 and 25 ng/ml (5.3 × 10¹¹ to 2.6 × 10⁹ M, r = 0.78, P < 0.001), then plateaued. Maximal incorporation was 3.2 times that in control medium. EGF also induced a dose-dependent response between 1 and 20 ng/ml (1.7 × 10¹⁰ to 3.3 × 10⁹ M), but its relative level was only 1.8 times that in control medium, even though maximal incorporation was reached at a similar molar concentration for EGF and KGF. aFGF induced a dose-response stimulation from 50 to 200 ng/ml (3.2 × 10⁹ M to 12.9 × 10⁹ M). At the latter concentration, the incorporation was 175 ± 18% of that in control medium alone (P < 0.001, not shown in Figure 1). HGF/SF had no significant effect upon tritiated choline incorporation.

View larger version (19K): [in this window] [in a new window]   Figure 1.   Dose-response curve of the various growth factors for [3H]choline incorporation into whole type II cell DSPC at 48 h. Control medium is defined medium without added growth factor. Basal incorporation in control medium was 30,400 ± 800 dpm/106 cells, n = 24. Mean ± SEM of six culture experiments. Multiple comparison by ANOVA. Significant difference from control medium for: (a) P < 0.05, (b) P < 0.01, (c) P < 0.001. Significant difference from the preceding growth factor concentration for: (d) P < 0.05.

Because the EHS matrix contains a number of factors (22) capable of influencing DSPC synthesis, we examined the effect of KGF on [³H]choline incorporation in cells maintained on inert substrata. On plastic, KGF 10 and 25 ng/ml had no effect, whereas KGF 50 ng/ml increased incorporation by only 11% (33,405 ± 1,290 versus 30,035 ± 850 dpm/10⁶ cells, n = 6, P < 0.05). By contrast, on collagen I membrane (Transwell coll; Costar, Cambridge, MA), a substratum previously reported to support the expression of type II cell differentiated features (36), KGF 25 ng/ml increased [H³]choline incorporation by 21% (18,045 ± 1,045 versus 14,890 ± 945 dpm/10⁶ cells, n = 6, P < 0.025).

[³H]Choline Incorporation into Surfactant DSPC

DSPC is the major surfactant component but is also present, although in smaller proportion, in other cell membrane compartments. We restricted our investigation to the effects of the various growth factors on the synthesis of DSPC in the surfactant compartment by studying the SF separated from the other cell compartments (RF). Growth factors were tested only at concentrations that had markedly increased choline incorporation in whole-cell DSPC.

In control medium without growth factor, [H³]choline incorporation into SF and RF represented 26% and 74%, respectively. KGF 25 and 50 ng/ml increased incorporation into DSPC 2.6 and 4.7 times, respectively, in SF (significant difference between both concentrations), but only 1.3 and 1.6 times in RF (Figure 2). Incorporation was therefore enhanced by KGF two to three times more in surfactant DSPC than in nonsurfactant DSPC, and about 50% of total incorporation was retrieved in each fraction. EGF 20 and 50 ng/ml both stimulated incorporation about twice in SF and 1.5 times in RF (Figure 2). aFGF, tested only at 100 ng/ml, increased incorporation 1.7-fold in SF, but there was no significant change in RF (Figure 2).

View larger version (56K): [in this window] [in a new window]   Figure 2.   Effects of KGF, EGF, and aFGF on incorporation of [3H]choline into DSPC of SF and RF at 48 h. Concentration of growth factors is in nanograms per milliliter. Mean ± SEM of six culture experiments. Multiple comparison by ANOVA. Significant difference from control medium (defined medium without growth factors) for: (a) P < 0.05, (b) P < 0.01, (c) P < 0.001. Significant difference from KGF 25 ng/ml for: (d) P < 0.05, (e) P < 0.01.

[³H]Acetate Incorporation into Surfactant Phospholipids and Neutral Lipids

The incorporation of labeled acetate into the SF of cultured cells was measured to determine whether the studied growth factors enhanced only the CDP-choline pathway, that is, the pathway that allows the condensation of activated choline with diacylglycerols (29), or whether they also affected the synthetic rate of precursors of the lipid moiety of various surfactant phospholipids (Table 2). KGF (25 ng/ml) considerably enhanced incorporation into total PC (10.8 times), DSPC (11.8 times), phosphatidyl-ethanolamine (PE, 5.9 times), sphingomyelin (Sm, 6.2 times), phosphatidylinositol (4.7 times), and other lipids including neutral lipids (2.2 times). EGF (20 ng/ml) significantly increased incorporation only into PC (5.1 times), DSPC (5.3 times), PE (2.0 times), and Sm (3.5 times). aFGF (100 ng/ml) significantly increased incorporation only into PC (2.7 times), DSPC (2.6 times), and Sm (1.7 times).

Surfactant Content of Cultured Cells

To determine whether the increased incorporation of labeled precursors corresponded to an actual increase of surfactant content of cultured cells, SF was isolated and its total phospholipid (TPL) and DSPC contents were determined. Only KGF and EGF effects were studied. KGF 25 and 50 ng/ml and EGF 20 ng/ml both increased surfactant TPL and DSPC about twofold on a per-cell basis; the increase with EGF 50 ng/ml was not statistically significant (Figure 3).

View larger version (42K): [in this window] [in a new window]   Figure 3.   TPL and DSPC content of surfactant material extracted from type II cells cultured for 48 h either in control medium or in the presence of KGF (25 or 50 ng/ml), or of EGF (20 or 50 ng/ml). Mean ± SEM of six culture experiments. Multiple comparison by ANOVA. Significant difference from control medium for *P < 0.05.

CPCT and FAS Activity and Gene Expression

Enzyme activity and gene expression were studied only for KGF 50 ng/ml and EGF 20 ng/ml. In control medium, CPCT activity was about 150 dpm incorporated per milligram protein per hour. As shown in Figure 4, CPCT activity was enhanced 27% by KGF after 12 and 24 h, and had returned to control level at 48 h. EGF increased CPCT activity 34% after 12 h of exposure, but this stimulation was no longer sustained after 24 h. Densitometric analysis of Northern blots revealed no change in CPCT messenger RNA (mRNA) level in the presence of either KGF or EGF for 24 h (1.5-kb transcript: 105.5 ± 13.9 and 98.3 ± 4.9% of control value; 5.5-kb transcript: 106.0 ± 23.8 and 99.0 ± 9.1% of control value for KGF 50 ng/ml and EGF 20 ng/ml, respectively, n = 4).

View larger version (12K): [in this window] [in a new window]   Figure 4.   Changes over 48 h in the activity of CPCT in type II cells cultured in the presence of either KGF (50 ng/ml) or EGF (20 ng/ml). Results are expressed as percentage of the activity in control medium. Mean ± SEM of six samples at each time point. Multiple comparison by ANOVA. Significant difference from control medium for *P < 0.05; **P < 0.01.

In control medium, FAS activity was about 4 pmol NADPH used per microgram protein per min. As shown in Figure 5A, after 24 h it was increased 64% by KGF but not significantly changed by EGF. At 12 h, however, FAS activity was increased 41% by EGF (P < 0.05) but was not yet significantly changed by KGF (not shown). KGF 50 ng/ml induced a twofold increase in FAS mRNA level after 24 h, whereas there was no significant change in the presence of EGF 20 ng/ml (Figure 5B).

View larger version (21K): [in this window] [in a new window]   Figure 5.   Effects of KGF (50 ng/ml) and EGF (20 ng/ml) on FAS activity and gene expression in cultured type II cells at 24 h. (A) FAS activity. Results are expressed as percentage of the activity in control medium (mean ± SEM of six culture experiments). Multiple comparison by ANOVA. Significant difference from control medium for *P < 0.05. (B) Study by Northern blot hybridization analysis of relative level of FAS mRNA and of 18S rRNA using specific cDNA probes as indicated in MATERIALS AND METHODS. (B1) Representative Northern blot. (B2) Densitometric analysis of the exposed films. Results are expressed as percentage of the control value after normalization for RNA loading on the basis of hybridization of 18S rRNA (mean ± SEM of three culture experiments). Multiple comparison by ANOVA. Significant difference from control medium for **P < 0.01.

KGF Effects on SP Gene Expression

Densitometric analysis of Northern blots revealed that KGF increased the expression of all three SPs (SP-A, SP-B, and SP-C) at a pretranslational level (Figure 6). The increase was dose-dependent for SP-A and SP-B between 1 and 50 ng/ml KGF. At the latter concentration, it reached about 3 times for SP-B and about 3.5 times for SP-A. Regarding SP-C, KGF effect was biphasic: the strongest stimulation (2.2-fold) was observed for KGF 10 ng/ml, and there was a significant decline between 10 and 50 ng/ml, although the amount of SP-C mRNA remained significantly higher than the control value at the latter KGF concentration.

View larger version (28K): [in this window] [in a new window]   Figure 6.   Dose-response effect of KGF on surfactant protein gene expression by cultured type II cells in a 48-h exposure. Study by Northern blot hybridization analysis of relative level of mRNAs coding for the surfactant proteins SP-A, -B, and -C, and of 18S rRNA using specific cDNA probes as indicated in MATERIALS AND METHODS. Left-panel: representative Northern blots. From left to right, lanes are control, KGF 1, KGF 10, and KGF 50 ng/ml; for SP-A, arrows show the two transcripts at 1 and 1.7 kb. Right panel: densitometric analysis of the exposed films. Results are expressed as percentage of the control value after normalization for RNA loading on the basis of hybridization of 18S rRNA (mean ± SEM of four culture experiments). Multiple comparison by ANOVA. Significant difference from control medium for *P < 0.05; **P < 0.01; ***P < 0.001. Significant difference from the preceding KGF concentration for: +P < 0.05; ++P < 0.01; ¶P < 0.001. Note that for SP-B, the difference between KGF at 1 and 50 ng/ml is significant for P < 0.05.

As a first approach to the underlying mechanisms, cells were exposed for 12 h to actinomycin D (Act. D) in the absence or presence of KGF at the concentration that had led to the maximal increase of surfactant protein mRNAs (Figure 7). For SP-A and SP-C, the rate of disappearance of mRNAs was unchanged in the presence of KGF as compared with control medium, which suggests that the enhanced mRNA steady state induced by KGF may have resulted from a transcriptional effect. By contrast, the disappearance of SP-B mRNA in the presence of Act. D was prevented by KGF (no significant difference between mRNA level in the presence and absence of Act. D), which argues for an effect of KGF primarily on SP-B mRNA stability.

View larger version (36K): [in this window] [in a new window]   Figure 7.   Stability of surfactant protein mRNAs in type II cells cultured in the absence () or presence (+) of KGF. Cells were cultured first in control medium or in the presence of KGF for 48 h, then for an additional 12 h in the same medium without () or with (+) Act. D (5 µg/ml). KGF was used at concentrations that had produced the highest increase in former experiments (see Figure 6), specifically, 50 ng/ml for SP-A and SP-B and 10 ng/ml for SP-C. Expression of SP-A, -B, and -C mRNAs was analyzed by Northern blotting, and results were normalized for RNA loading on the basis of hybridization of 18S rRNA as in Figure 6. Densitometric analysis of exposed films is presented as percentage of control value (mean ± SEM of three culture experiments). Comparison between corresponding media without and with Act. D by two-tailed t test. Significant difference for *P < 0.05. The rate of SP-A and SP-C mRNA decay was similar in the absence or presence of KGF, whereas SP-B mRNA stability appeared to be increased by KGF.

	Discussion

Because it is well known that the control of surfactant synthesis in developing fetal lung is critical to normal lung function at birth, many investigators have sought the factors that might play important roles in the regulation of its synthesis. Our findings suggest that KGF may be such a factor. We found that it is a potent stimulus of surfactant phospholipid synthesis, particularly of the major surfactant component, DSPC, in fetal alveolar type II cells. We also report that KGF increases the expression of SP genes in these cells, including that for SP-C, for which contradictory data had been reported previously.

Cells did not proliferate in control conditions. In the presence of growth factors, stimulation of DNA synthesis was lower than the stimulation of thymidine incorporation in adult rat lung type II cells cultured on plastic in the presence of serum (6). This is likely to result from the use of basement membrane matrix and serum-free medium. These conditions have effectively been shown to favor expression of type II cell differentiation markers at the expense of cell proliferation (16, 37).

KGF, aFGF, and EGF all increased choline incorporation into DSPC in a dose-dependent way. Furthermore, the effects were more marked for surfactant DSPC than for DSPC of other cell compartments, particularly for KGF. Although it has been reported to be a mitogen for type II cells and a synergistic factor with KGF in this respect (6, 20), HGF/SF failed to stimulate choline incorporation at molar concentrations for which the other growth factors had a stimulating effect. Consistent with incorporation data, KGF and EGF, which had the strongest effects, also increased the cell content in surfactant phospholipids. The somewhat higher rate of precursor incorporation as compared with the increase in surfactant content indicates, however, that incorporation studies led to an overestimate of surfactant synthesis rate, presumably because of changes in precursor pool size and/or increased surfactant turnover.

Recently, an increase of whole-lung DSPC was reported in the premature rabbit given KGF in vivo (14). In this investigation, however, no increase in alveolar DSPC was found, and it was not determined whether the increase of DSPC in lung tissue was due to an increase in type II cells. This prevents us from drawing conclusions about an actual increase of surfactant DSPC. By contrast, our findings clearly indicate that KGF markedly increased surfactant DSPC through direct action upon the alveolar type II cell.

KGF stimulated choline incorporation to a larger extent when cells were cultured on EHS matrix than on plastic or collagen. Plastic has long been recognized not to support the maintenance of type II cell differentiated features which, on the contrary, are well retained on EHS matrix (16, 17, 38, 39). This matrix has been shown to be an optimal support for type II cell marker expression, probably in part because it maintains type II cells in their cuboidal shape (38, 40). We employed Transwell collagen membrane as an alternative substratum because it has been reported to allow type II cells to remain cuboidal, as on EHS matrix, and to preserve their lamellar bodies (36). Cell response to KGF was effectively higher on this substratum than on plastic, which shows that not all basal membrane components or other factors present in EHS matrix are necessary for cell responsiveness to KGF. We did not examine basal membrane components such as laminin, collagen IV, or heparan sulfate proteoglycan because on the one hand, these have not been shown to achieve the type II cell-supporting properties of EHS matrix when used individually (41), and on the other hand, heparan sulfate was reported to inhibit the binding of KGF to its receptor (42).

From a mechanistic point of view, increased choline incorporation and increased DSPC synthesis correlated with enhanced activity of CPCT, the rate-limiting enzyme of the CDP-choline pathway. This increase in enzyme activity occurred in the absence of change in the steady state of CPCT mRNA, which suggests an activation of preexisting enzyme. CPCT has long been known to be activated by various metabolites, including phospholipids and fatty acids, both through translocation of soluble, less active enzyme to membranes, and through phosphorylation (43). By contrast, stimulation of FAS activity, which correlated with increased acetate incorporation into surfactant lipids, went along with increased FAS mRNA content of cultured cells, indicating that KGF controls FAS activity at a pretranslational level. The increase in CPCT activity may therefore be secondary to enhanced FAS expression and activity through the enhancement of fatty acid synthesis.

Comparing the effects of the various growth factors, although KGF, EGF, and aFGF stimulated choline incorporation in the nanomolar range of concentration, a response consistent with a possible physiologic role in vivo, at constant molar concentration KGF appeared about twice as active as EGF and several times more active than aFGF. Moreover, whereas KGF and EGF increased incorporation to a much larger extent than DNA synthesis and increased the surfactant content of cells, aFGF stimulated incorporation and DNA synthesis about the same. In other words, whereas KGF and EGF stimulated surfactant synthesis on a per-cell basis, the increase of choline incorporation induced by aFGF could be cell-number related. Strikingly, despite their differences in terms of DSPC synthesis, KGF and aFGF exhibited a similar efficiency in stimulating DNA synthesis. This may be explained by a control of phospholipid metabolism and of cell proliferation through different means of transduction, for instance, through the involvement of the KGF-R for phospholipid metabolism and by a splice variant of the FGF-R3 shown to be highly specific for aFGF (44) for proliferation. Comparing the effects of KGF and EGF, both factors increased CPCT and FAS activity to a similar extent despite their differences for precursor incorporation. Differences in final effect appear to result from a more prolonged stimulation by KGF. Thus, EGF does not seem to be less active than KGF at the same molar concentration, but seems to present a less sustained action, possibly as the consequence, for instance, of a shorter half-life in vitro.

With regard to surfactant proteins, our study brings clarification to the understanding of KGF effects. KGF stimulated gene expression of all three SPs, SP-A, SP-B, and SP-C. Although this is consistent with all previous investigations for SP-B (12, 13, 15), puzzling findings have been reported for SP-A and SP-C. These discrepancies might be accounted for by differences in the experimental design, in the analytical method, and in the considered developmental stage. Thus, in isolated adult type II cells, SP-A expression was enhanced and SP-C expression was little affected (15), whereas in the cultured whole embryonic lung primordium, SP-C expression was reported to be enhanced and that of SP-A to be unchanged (12). Conversely, SP-C expression was found to be reduced in the cultured epithelial primordium separated from mesenchyme (13) and in vivo, in transgenic mice overexpressing KGF (11). Taking into account the biphasic effect of KGF on SP-C mRNA evidenced in the present study, the inhibiting effect of KGF observed by some could have been due to the use of high dosage in vitro (13) or to a high, uncontrolled level of expression in vivo (11). Our results are consistent with those of Shiratori and coworkers (12), who used KGF at a concentration (50 ng/ml) identical to our highest dosage and who similarly evaluated SP-C mRNA level by quantitative Northern blot analysis. In those studies in which SP-C expression was reported to be decreased, the assessment through PCR (13) or pro-SP-C immunostaining (11) was not really quantitative. Finally, it should also be pointed out that in both of the studies with explanted developing lung (12, 13), that of Cardoso and associates (13) was done at a time (14 d in the rat and 12 d in the mouse) far ahead of the stage at which surfactant accumulation starts. At these early stages of lung development, SPs and their mRNAs are detected only in trace amounts, and their rate of synthesis is considerably lower than in the last part of gestation. The present study is the first one to provide insights into the effects of KGF on developing type II cells at the actual time of surfactant accumulation. These cells appear to respond to KGF by enhanced synthesis of all surfactant components, including proteins. Moreover, we report the first attempt to explore the mechanism of action of KGF on SP gene expression. Although our results are still preliminary in this regard, the absence of change in the rate of SP-A and SP-C mRNA disappearance in the presence of KGF suggests that its effects are exerted at the transcriptional level for these genes, whereas the inhibited decline of SP-B mRNA argues for increased stability.

Although lung organogenesis was totally impaired in transgenic mice bearing a dominant negative KGF-R targeted on the lung epithelium (9), recent comparison studies of KGF with other factors of the FGF family have led to the conclusion that aFGF (13) and/or FGF-10 (45) would be the actual control factors of lung branching morphogenesis through this receptor. Although it induced an abnormal cystlike pattern of lung growth in vitro, in these studies KGF appeared to enhance type II cell morphologic maturation. Together with the present findings, this leads us to ascribe to KGF a role in the control of type II cell differentiation and maturation rather than in lung morphogenesis. Consistent with this assumption is the observation of increased mRNA levels for KGF and KGF-R in the rat lung during the perinatal period, when surfactant accumulation takes place (12, 46).

This raises the question of the biologic importance of KGF for lung maturation. Invalidation of the KGF gene effectively induced no obvious abnormality except in hair development, and no apparent respiratory difficulty in mice (47). It should be emphasized that surfactant was not explored in this model, however, and that the possibility of quantitative changes cannot be ruled out. These observations do not imply that KGF has no effective role in the control of type II cell maturation, but suggest a redundancy of mesenchymal factors involved in the process. Our results indicate that KGF, EGF, and to a lesser extent aFGF, would be such stimulating factors.

Thus far, glucocorticoid administration has been the approach of choice to accelerate lung maturation for preventing neonatal RDS. This treatment, however, may have adverse effects because glucocorticoids have been shown to decrease epithelial cell multiplication and to impair septation of alveolar sacs into definitive alveoli (48). Thus, KGF, which is simultaneously a mitogenic and a maturation-promoting factor for the alveolar epithelium, has potential as a therapeutic agent for the prevention and treatment of RDS. It may also be useful in the management of bronchopulmonary dysplasia, a frequent complication of RDS, and of acute RDS in the adult.