Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The development of lung airway and vascular structures proceeds via a series of epithelial-mesenchymal cell interactions, each regulated by growth factors, their receptors, and components of the extracellular matrix. The fibroblast growth factors (FGFs) comprise a multigene family of structurally related members (FGFs 1 to 9), some of which bind heparan sulfate proteoglycans to form highly mitogenic complexes within the extracellular matrix (1). A number of studies have highlighted the importance of FGFs and their receptors (FGFRs) in lung morphogenesis. FGF1, for example, has been shown to stimulate lung epithelial cell proliferation and airway bud formation (2). Another important member of the FGF family in fetal lung development, FGF7, is detected within the mesenchyme by reverse transcriptase-polymerase chain reaction (RT-PCR), and antisense oligonucleotides added to lung explants inhibit branching (3). In lung epithelial cultures, FGF7 stimulation of epithelial cell proliferation results in cyst formation (2), and there is lung malformation in transgenic mice overexpressing FGF7 (4). The most widely distributed of the FGFs is FGF2 (basic FGF), a mitogen with a role in processes as diverse as neuronal cell survival and angiogenesis.

In the present study, we focus on the role of FGF2 and the FGFRs in late fetal and early postnatal lung development, that is, from embryonic Day 18 (E18) to postnatal Day 28 (P28). Embryonic Day 18 represents the transition between the pseudoglandular and canalicular stage (E18 to E19), the latter being characterized by extensive cell proliferation as bronchioles form by continuously dividing into smaller canals, the appearance of a primitive alveolar region, and an increase in vascularization (5, 6). Further thinning of the interstitium and formation of primitive alveoli, or saccules, characterize the saccular stage (E20 to E22) (7, 8). Alveolarization begins on the fourth or fifth postnatal day (9), when a spurt in multiplication greatly increases alveolar surface area, such that by 8 d after birth there is a threefold increase in alveolar concentration (11). The larger pulmonary arteries formed during the embryonic period continue to branch into smaller and more numerous units, especially between P8 and P11, mirroring alveolar formation (11).

The FGF family acts through high-affinity tyrosine kinase receptors. Four FGFR genes have been cloned from human, rodent, and chicken cDNA libraries (12) encoding 135- to 150-kD glycoproteins with tyrosine-specific protein kinase activity (15) designated FGFR1 (flg), FGFR2 (bek), FGFR3, and FGFR4. These genes are expressed in several different spliced forms, both membrane-bound and secreted. In the adult, mRNAs for each receptor are expressed by lung cells; FGFR4, however, appears to be restricted to the lung, and so perhaps has a particular functional relevance for this tissue (16). At an earlier stage of development than studied here (i.e., between E9.5 and E16.5), FGFR is expressed by tracheal lung bud mesenchyme (17), and between E13 and E22 is expressed by airway epithelial cells (especially at branch sites) and interstitial cells. Reactivity increases in the pseudoglandular stage, is variable in the canalicular stage, and is absent by the saccular stage (18). FGFR2 is first expressed by cells of the tracheal epithelium and airway buds and later by bronchiole epithelial cells (19). Between E12 and E17 there is diffuse reactivity for the two spliced forms of FGFR2 in the lung (17). The receptor for FGF7 is the FGFR2 IIIb spliced form, which is expressed at Day 11 (2), when FGFR4 is located in lung buds at distal but not proximal sites. The importance of FGFRs in lung development also has been demonstrated in transgenic mice, in which a block in FGFR2 function by a dominant negative mutation resulted in a profound defect in branching and epithelial differentiation (20).

We describe differential expression of the four FGFR genes in late fetal and early neonatal rat lung, using Northern blot analysis and in situ hybridization. We show that each receptor is important during this stage of lung development and we correlate the temporal patterns of FGF1 and FGF2 mRNA expression with receptor expression. FGF1 is the universal ligand for all spliced forms of the four FGF receptors, whereas FGF2 binds to the c-spliced forms of FGFR1 to 3 and FGFR4 (21). In contrast to the FGFRs, we detected only low amounts of FGF1 and 2 mRNAs in the fetal lung and immediately after birth, although immunohistochemistry detected FGF2 protein at this time. FGF1 and FGF2 mRNA, and FGF2 protein as identified by enzyme-linked immunosorbent assay (ELISA) and mitogenic activity, accumulated over time in the postnatal lung, whereas immunohistochemistry detected persistent expression of FGF2 protein in some cell populations and decreased expression in others.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Lung Tissue

Viral-antibody-free Sprague-Dawley rats (Charles River Laboratories, Kingston, NY) were timed as pregnant from the morning after mating (E0). Fetal lungs were obtained at five time points, after Cesarean section (E18 to E22). The day after birth was designated P0. Lungs from rat pups were obtained at P0, P2, P6, P10, P14, P21, and P28, and from adult male rats (6 mo old). For RNA extraction the lungs were immediately frozen in liquid nitrogen (at each time point, respectively, n embryos = 60, 31, 20, 21, and 22 and n pups = 28, 14, 11, 8, 5, 6, and 6). For in situ hybridization studies or immunocytochemistry, the lungs of the smallest fetuses (E18; n = 12) were simply immersed in fixative for 2 h (4% paraformaldehyde in phosphate-buffered saline [PBS]) whereas lungs from later fetuses (E19 to E22; at each time point, respectively, n = 8, 5, 5, and 6), pups (P0 to P28; at each time point, respectively, n = 7, 7, 5, 5, 3, 2, 2, and 2), and adult rats (n = 4) were inflated with fixative via the trachea. They were then immersed in fixative for a further 2 h and embedded in paraffin wax or cryoprotected (16% sucrose for 16 h at 4°C) and flash-frozen (70°C).

For heparin-affinity chromatography, lungs obtained from fetuses (E20; n = 10) and pups (P28; n = 2), and brain tissue from adult rats (n = 2) were frozen, thawed, and homogenized at 4°C in buffer (20 mM Tris-HCl, 0.1% 3-[(3-cholamidopropyl)dimethyl ammonio]-1-propane sulfonate [CHAPS], 1 mM dithiothreitol [DTT], and 2.0 M NaCl). Cells were lysed by 10 passes in a Dounce homogenizer and the lysate frozen (70°C), thawed, and sonicated on ice. Unbroken cells were removed by centrifugation and the supernatant diluted to a final concentration of 0.1 M NaCl.

cDNA Probes

Mouse FGF1, 3, 4, and 5 were obtained from J. Hebert and G. Martin, University of California at San Francisco (22); rat FGF2 was as described (23); FGF7 was from Clive Dickson, Imperial Cancer Research Fund, London (24); mouse FGFR1 and FGFR3 cDNAs were obtained from David Ornitz, Washington University Medical School, St. Louis, MO (25); mouse FGFR2 was from Claudio Basilico, New York University, New York, NY (26); and rat FGFR4 was from Robert Horlick, DuPont Merck, Wilmington, DE (27). For Northern blot analysis, cDNA probes were generated by labeling with ³²P uridine triphosphate (UTP) (6,000 Ci/mM; DuPont NEN, Boston, MA) by random hexamer priming to a specific activity of 2 × 10⁹ cpm/ mg. For the RNase protection assay, a 350-bp Kpn1- EcoRI fragment of rat FGF2 cDNA (28) was cloned into pBluescript (Stratagene, La Jolla, CA). The plasmid was linearized with BamHI and an antisense riboprobe generated by labeling with ³²P UTP (800 Ci/mM). Full-length transcripts of 175 bp were purified from an 8% polyacrylamide urea gel. For in situ hybridization, FGFR1 to 4 cDNA templates were cloned into pBluescript to synthesize single-strand sense or antisense transcripts. Riboprobes were generated using standard procedures (29), the transcribed probes being hydrolyzed (100 mM bicarbonate buffer, pH 10.2) to generate 150- to 200-bp fragments.

Northern Blot Analysis and RNase Protection Assay

RNA was extracted from frozen tissue by homogenization in 4 M guanidinium isothiocyanate and sedimentation through a cesium chloride gradient. Total RNA (20 µg) was denatured by heating in 2.2 M formaldehyde and 50% formamide in sample buffer containing ethidium bromide. The samples were fractionated using thin surface tension gels and RNA was transferred from the gel to nitrocellulose (28). Blots were washed in low-stringency buffer (2× saline sodium citrate [SSC], 0.1% sodium dodecyl sulfate [SDS], at room temperature 2 times for 15 min each) and in high-stringency buffer (0.1× SSC, 0.1% SDS, at 65°C, 2 times for 15 min each). Equal lane loading was assessed by subsequent hybridization of the filter to a rat cDNA for cyclophilin. For RNase protection assay, riboprobes were hybridized to 50 µg of total RNA (45°C overnight). Unprotected RNA was digested (37°C for 30 min) with RNase A (2 µg/ml; Life Technologies Inc., Gaithersburg, MD) and proteinase K (1 µg/ml; Sigma Chemical Co., St. Louis, MO). The protected fragments were analyzed in an 8% polyacrylamide urea gel. ³²P UTP-labeled -actin was added to each reaction to demonstrate equal amounts of RNA. RNA isolated from a C6 glioma cell line was used as a positive control (30).

In Situ Hybridization and Immunohistochemistry

Frozen sections (5 µm) were prehybridized (31), hybridized to alkaline hydrolyzed ³⁵S-labeled riboprobes generated from FGFR cDNAs (31), and stained with 10× diluted Giemsa (Rowley Biochemical Institute, Rowley, MA). No grains localized to tissues using the sense probe. For immunohistochemistry, deparaffinized sections (5 µm) were quenched in 3% hydrogen peroxide in absolute methanol (10 min at room temperature), washed in PBS in absolute methanol (10 min at room temperature) and washed in PBS. After incubation in bovine testicular hyaluronidase (50 µg/ml in PBS at 37°C for 15 min; Sigma), the sections were washed in 0.5% bovine serum albumin in PBS (5 min), and nonspecific antibody binding was blocked by incubation with normal goat serum (10% for 10 min, Histostain kit; Zymed, South San Francisco, CA). They were incubated overnight with antibovine FGF2 immunoglobulin G (4°C, 200 µg/slide; Upstate Biotechnology Inc., Lake Placid, NY) (18), washed and incubated with goat antibovine biotinylated antibody (15 min at room temperature) followed by a streptavidin peroxidase conjugate (15 min at room temperature), and stained with hematoxylin. Immunoreactive sites were stained red and cell nuclei blue. No immunoreactive sites were identified when the primary antibody was omitted.

Heparin-Affinity Chromatography

Heparin-affinity chromatography was performed using a heparin Sepharose 6B column (Pharmacia, Piscataway, NJ). The column was equilibrated with buffer (20 mM Tris-HCl, pH 7.5, 0.1 M NaCl, 0.1% CHAPS, and 1 mM DTT), and the tissue lysate from lung at E20 and P28 applied (in the same buffer). The column was washed extensively with buffer, and fractions were eluted with a linear gradient from 0.6 to 2.0 M NaCl. The growth factor activity in fractions was assessed by the ability to stimulate the incorporation of [³H]thymidine into quiescent Balb/c 3T3 cells in a 96-well plate (1,000 cells/well). Cells were incubated with column extracts (40 to 50 h), fixed in methanol, and lysed, and the label was counted. One unit of growth factor activity was defined as the amount of extract required to stimulate half-maximal DNA synthesis (23). FGF1 activity eluted at 1.0 M NaCl and, as previously described, FGF2 activity eluted at 1.5 M NaCl (23). No mitogenic activity was identified as eluting earlier in the gradient (i.e., from 0.6 to 1.0 M NaCl). Total protein was measured by a standard assay (BCA; Pierce, Rockford, IL), and immunoreactive FGF was detected using an ELISA specific for FGF2 (R&D Systems, Minneapolis, MN). Mitogenic activity eluting in the 1.5-M NaCl fraction from the lung at E20 and P28, or from adult rat lung or brain, were applied to the well and the assay was carried out according to the manufacturer'sinstructions. The antihuman antibody cross-reacted with rat FGF2 protein from brain tissue, providing a positive control.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Temporal Expression of FGF Receptor Genes

Between E18 and P28, each FGFR gene was expressed in the developing rat lung at some time: FGFR1 RNA expression was relatively low from E18 until P21 and highest at P28 (Figure 1a); FGFR2 RNA expression, first detected at E18, increased between E20 and P0 (Figure 1b) and then decreased between P0 and P2 before increasing to its highest postnatal level at P28; FGFR3 RNA expression (Figure 1c) was low at E18, increased between E19 and E22, reached a peak early in the postnatal period (P2 to P10) and then declined; and FGFR4 RNA expression, first detected at E19 (Figure 1d), increased at P0, and, after an initial fall, was again increased at P10 and remained high. For FGFR1 to 3, a major transcript was detected at 4.4 kb. In addition, a second band was detected at 6.0 kb, during the time of peak expression of FGFR2 (i.e., at E20 and E22). Similarly, for FGFR3, this band was detected faintly at P2 and P6, and more obviously at P10. In the adult rat lung also, each receptor gene was expressed, with a transcript size between 4.3 and 4.4 kb (data not shown). FGFR4, however, was transcribed as a 3.2-kb mRNA. The results shown in Figures 1a to 1d are summarized in Table 1. FGFR2 and FGFR4 expression levels were highest overall, FGFR2 being highest between the end of the saccular stage and birth, and FGFR4 late in the postnatal period. FGFR1 expression was greatest late in the postnatal period and FGFR3 expression greatest in the 10 d after birth.

View larger version (64K): [in this window] [in a new window]   View larger version (69K): [in this window] [in a new window]   View larger version (82K): [in this window] [in a new window]   View larger version (75K): [in this window] [in a new window]   Figure 1.   Temporal expression of FGFR1 to 4 RNA in developing rat lung. (a) FGFR1 RNA is detected from E18, and its level remains relatively constant until rising at P28; (b) FGFR2 RNA is detected by E18; its level increases between E20 and birth and declines after birth (P2) to less than one quarter of the late postnatal level (P28); (c) FGFR3 RNA is barely detectable at E18 but rises from E19 to peak levels in the days after birth (P2 to P10); (d) FGFR4 RNA is barely detectable before birth (E18 to E22) but increases after birth (P0), is higher 10 d later (P10), and remains high late in the postnatal period (P14 to P28). Equal lane loading was assessed by subsequent hybridization of the same filter to a rat cDNA for cyclophilin (CP).

Spatial Expression of FGF Receptor Genes

At each time point, cells expressing each FGFR were distributed throughout developing lung tissue. The typical pattern of expression is described, illustrated at selected time points, and summarized in Table 2. In the fetal and postnatal lung, FGFR1 mRNA was expressed predominantly by the smooth muscle cells of large arteries (Figure 2a), and by airway epithelial cell populationssites of expression being identified along large airways where increased expression occurred at bifurcations (Figure 2b). In the early postnatal period in particular. FGFR2 mRNA was expressed by epithelial cells of large airways and by developing alveolar wall cells (Figure 2c); FGFR3 mRNA was highly expressed by esophageal epithelial cells (Figure 2d) and by alveolar cells, with expression levels higher in discrete regions of the alveolar wall (Figure 2e). Postnatally, FGFR4 was widely expressed by the cells of large vessels, large airways, and cells of developing alveolar walls (Figure 2f). Thus, FGFR1 was expressed predominantly by cells in extra-alveolar regions, FGFR3 predominantly by cells in intra-alveolar ones, and FGFR2 and FGFR4 by cells in both regions.

View larger version (156K): [in this window] [in a new window]   View larger version (136K): [in this window] [in a new window]   View larger version (194K): [in this window] [in a new window]   View larger version (170K): [in this window] [in a new window]   View larger version (161K): [in this window] [in a new window]   View larger version (173K): [in this window] [in a new window]   Figure 2.   Spatial expression of FGFR1 to 4 mRNA in developing rat lung. (a and b, E20) FGFR1 is shown localized to smooth muscle cells of a large artery (a, lumen marked by asterisks) and to epithelial cells lining a large airway at the branch point (b, arrows), while the cells of small vessels and airway buds are negative. (c, P0) FGFR2 localized to alveolar wall cells (arrows). (d and e, P4) FGFR3 localized to epithelial cells of the esophagus (d, lumen marked by asterisks), and to alveolar wall cells, including discrete cell clusters randomly distributed in the alveolar wall (e, marked by asterisks) where the label intensity is increased (focused to show increased site of expression). (f, P14) FGFR4 localized to cells of a large vessel (lumen marked by asterisks). Bar = 200 µm (a-c, e, and f ) and 300 µm (d).

Temporal Expression of FGF Ligand Genes

FGF1 mRNA was not detected in fetal rat lung (Figure 3, E18 to E22). Postnatally, a single transcript (4.6 kb) was identified by Northern blot analysis, expression being low at first (P6) and higher later (P28, Figure 3). As measured by RNase protection analysis, FGF2 expression was low in the fetal lung and immediately after birth (Figure 4), and increased to its highest level at P28. We did not detect expression of FGF ligands 3 to 5 and 7 at any time in the fetal or postnatal lung (data not shown). In normal adult lung, each of seven FGF2 mRNAs (6.0, 3.7, 2.5, 1.8, 2.6, 1.4, and 1.0 kb) and FGF1 mRNA (4.6 kb) were detected (data not shown).

View larger version (72K): [in this window] [in a new window]   Figure 3.   Temporal expression of FGFR1 RNA in developing rat lung. FGF1 RNA is not detected between E18 and E22; a single transcript is expressed at low levels from P6, and expression increases slightly to P28. Equal lane loading was assessed by subsequent hybridization of the same filter to a rat cDNA for cyclophilin (CP).

View larger version (67K): [in this window] [in a new window]   Figure 4.   Temporal expression of FGF2 RNA in developing rat lung. FGF2 expression, detected by RNase protection analysis, is seen only as a faint band before birth, whereas after P7 the level of expression increases with age; a C6 glioma cell line expressing a large amount of FGF2 RNA provides a positive control.

Identification of FGF2 Activity by Mitogenic Activity and ELISA

Mitogenic activity for Balb/c 3T3 cells eluting at 1.5 M NaCl from a heparin-sepharose-affinity column was not detected in the lung at E20 but was present by P28 (9 U/mg protein). The more sensitive ELISA technique, however, detected protein at E20 (11 pg/mg protein) as well as in adult lung (420 pg/mg protein). This is in agreement with the levels of mRNA measured by RNase protection analysis (Figure 4). Levels of FGF2 in adult lung were lower than those found in adult brain (513 pg/mg protein).

Spatial Expression of FGF2 Protein

No FGF2 protein was detected in lung without first treating the tissue sections with hyaluronidase. After treatment, sites of FGF2 immunoreactivity were detected at all time points, the extent and location varying with the stage of lung development.

Sites of FGF2 protein expression are illustrated in Figure 5, and the relative intensity of expression of FGF2 protein in developing lung is summarized in Table 3. At E18, FGF2 protein localized to the epithelial basement membrane of large airways and the branch points of airway buds, whereas epithelial cells were immunonegative (Figures 5a and 5b). FGF2 was expressed by developing smooth muscle cells in the walls of large airways and mesenchymal cells adjacent to airway buds (Figures 5a and 5b). In large vessels, FGF2 protein was expressed by mesenchymal cells (putative smooth muscle cells) adjacent to the endothelium (Figure 5b). Pleural cells were weakly immunopositive. Between E20 and E22, this pattern of expression persisted and FGF2 protein localized to the epithelial basement membranes of distal airways and to cells forming the developing saccules (Figure 5c). At this time, FGF2 protein also localized to putative smooth muscle cells of smaller airways and vessels, and its expression increased in pleural cells.

View larger version (127K): [in this window] [in a new window]   Figure 5.   Spatial expression of FGF2 in developing lung (after hyaluronidase treatment immunoreactive sites for FGF2 appear red). (a and b, E18) Low magnification of the lung at the transition from pseudoglandular to canalicular stage, showing high expression of FGF2 in the condensed mesenchyme surrounding airway epithelial tubes (a). Higher magnification (b) illustrating immunonegative cells lining epithelial tubes (lumen marked by asterisks) and localization of FGF2 to the basement membrane of these cells (small arrows), to mesenchymal cells between epithelial tubes, and to putative smooth muscle cells of vessels (large arrows). (c-e) FGF2 is shown localized to the basement membrane of airway epithelial tubes (arrows) and to cells in the developing alveolar region at the beginning of the saccular stage (E20, c); and at birth (P0, d) to alveolar cells (arrows), to smooth muscle cells of airway associated vessels (e, small arrows) and to the basement membrane and associated smooth muscle cells, but not epithelial cells, of a terminal bronchiole (large arrow). (f-h) At P28, FGF2 expression persists but is low in cells forming the expanded alveolar region and in the basement membrane and associated smooth muscle cells of a bronchiole (arrows) and in smooth muscle cells of associated vessels (asterisk, f ), but is intense in smooth muscle cells of large vessels (g) and in mesothelial and associated pleural cells (arrows, h). Bar = 200 µm (a) and 50 µm (b-h).

At P0, alveolar walls were thinner and there was evidence of septation: the overall pattern and amount of expressed FGF2 protein resembled that at E22 (Figures 5d and 5e). This pattern persisted at P7, although the amount of FGF2 localized to airway smooth muscle cells and to many cells in the alveolar region was less. At P28 the smooth muscle cells of large vessels were still strongly positive (Figures 5f and 5g). In large and small bronchioles, less FGF2 localized to smooth muscle cells and to the basement membranes of alveolar cells than at earlier time pointsthe decrease in the cells of the alveolar region reflecting both distribution and intensity such that FGF2 expression was associated only with basement membrane fragments. Pleural cells remained intensely positive (Figure 5h).

In contrast to lung, FGF2 protein was detected without hyaluronidase treatment in the epithelium of the esophagus (Figure 6). After treatment, immunoreactivity was lost, whereas the basement membrane of the smooth muscle cells of the esophagus was now positive.

View larger version (159K): [in this window] [in a new window]   Figure 6.   Spatial expression of FGF2 protein in developing rat esophagus (without hyaluronidase treatment immunoreactive sites for FGF2 appear red). FGF2 expression by esophageal epithelial cells (arrows) at P28.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Between E18 and P28, lung growth includes increase in airway and vessel length and distal branching, as well as the formation and increase in the size of alveoli and of the associated capillary network. Because the FGFRs are transcribed at relatively high levels immediately before birth, it appears that it is the fine regulation of the FGF ligands that regulate growth and morphogenesis at this time. This appears not to be the case in postnatal lung development, when FGF1 and 2 become abundant. High levels of expression of receptors and ligands at this late time could reflects a more general role for the FGFs in homeostasis and regulation of continued lung growth.

FGFR1 and FGFR2 previously have been localized to developing epithelial and mesenchyme tissues, including lung structures, at an earlier time than studied here (i.e., between E9.5 and E16.5 [32]). We found preferential localization and abundance of each receptor at specific times between E18 and P28. The localization of FGFR1 indicates a role in large vessel and airway growth between the canalicular stage and first 4 wk of the postnatal period, with activity greatest postnatally. In particular, our results indicate a role in the proliferation of smooth muscle cells, a finding in keeping with lengthening and thickening of vessel walls, and in the localized proliferation of epithelial cell clusters responsible for the expansion of airways at branch sites.

We found striking similarities in the diffuse pattern of localization of FGFR2 and FGFR4 to epithelial cells of large airways and alveolar wall cells. The expression pattern of FGFR2 indicates an increased role for this receptor in the proliferation of epithelial cells immediately before birth and later in the postnatal period. We found FGFR4 expression, previously demonstrated in the lung at E14.5 (33), to be abundant at birth and between the 10th and 28th days of postnatal life. Receptor abundance at these times indicates a role of FGFR2 and 4 in the growth of existing airways and in modifying the distal lung template, as primitive alveoli develop immediately before birth and then increase in number and complexity.

Unlike FGFR1, FGFR2, and FGFR4, expression of FGFR3 has not been detected previously in developing epithelial or mesenchymal tissues, including lung structures. The timing and localization of FGFR3 sites in our study indicate its role in the development of alveoli at the time of birth and in the period of accelerated growth in the following 10 d (11). During this window in time, the number and clusters of alveolar cells expressing FGFR3 indicate general as well as focal points of growth, expression during the postnatal growth period being earlier than the time of FGFR4 expressionthe other FGFR expressed abundantly by these cells. Although we cannot exclude FGFR3 expression by alveolar cells throughout the developing lung, its focal pattern of expression favors a role in the rapid expansion of the alveolar region. Whether this relates only to growth associated with an increase in alveolar number or to increase in number as well as the expansion of existing structures remains speculative.

Our findings of little, if any, FGF1 mRNA expression in the canalicular or saccular stage are in agreement with other reports of the lack of expression of this protein during this stage of lung development, suggesting that this ligand contributes little to lung changes at this time (34, 35). Its increased expression in the postnatal period also agrees with reported protein expression levels (36). Northern blot analysis showed no evidence of expression of FGF3 to 5 or 7 mRNAs in the fetal or postnatal lung, but other studies have detected FGF7 by more sensitive techniques than Northern blot, such as RT-PCR, and so we cannot exclude that, in addition to FGF2, other FGF family members are present at this time. We found mesenchymal cells, putative smooth muscle cells, and pleural cells to be major sites of FGF2 protein localization. When present on the cell surface, or in the extracellular matrix, heparin is required for FGF ligand-receptor binding. Acting as stabilizing factors, heparin or heparan sulfate glycoprotein protects FGF2 (and FGF1) from inactivation and from degradation (37, 38). Our findings agree with those of Sannes and colleagues (36) for the need for hyaluronidase digestion to demonstrate FGF2 in lung tissue. Although hyaluronidase treatment likely identifies heparan sulfate-bound FGF2 in basement membranes, the significance of FGF2 localization in cells is less clear, but its accessibility suggests that it may be cell-surface bound (see Ref. 36).

These results establish the pattern of FGF receptor expression associated with developing lung structures and confirm the pattern of FGF2 ligand expression. To our knowledge, this is the first demonstration of the spatial and temporal distribution of FGF ligand and FGFR1 to 4 in the lung at a critical time of alveolar growth. We conclude that between the saccular and postnatal periods a differential pattern of expression of FGF receptor and ligand contributes to the development of proximal and distal epithelial and vascular structures supplying the alveolar region, to the formation of the alveolar-capillary membrane, and to the development of the pleura. In the immediate period before the onset of alveolar development (i.e., in the canalicular stage), we confirm the importance of FGF ligand in the development of large airways and vessels and smaller branching airways and we identify the FGF receptors associated with this stage of lung development.