The alveolar basement membrane contains a variety of extracellular matrix (ECM) molecules, including laminin and sulfated glycosaminoglycans of proteoglycans. These mixtures exist within microdomains of differing levels of sulfate, which may specifically interact to be key determinants of the known capacity of the type II cell to respond to certain growth factors. Isolated type II cells were exposed to either acidic fibroblast growth factor (FGF-1), basic fibroblast growth factor (FGF-2), or keratinocyte growth factor (KGF; FGF-7) on culture wells precoated with laminin alone or in combination with chondroitin sulfate (CS), high-molecular-weight heparin, or their desulfated forms. Desulfated heparin significantly elevated FGF-1- and FGF-2-stimulated DNA synthesis, whereas desulfated CS and N-desulfated heparin elevated FGF-7-stimulated DNA synthesis by type II cells on laminin substrata. When FGF-1 was mixed into the various test matrix substrata, DNA synthesis was significantly increased in all cases. These results demonstrated that decreased levels of sulfate in ECM substrata act to upregulate responses to heparin-binding growth factors by alveolar epithelial cells on laminin substrata.
- sulfated proteoglycans
- basement membrane
- acidic fibroblast growth factor
- basic fibroblast growth factor
- keratinocyte growth factor
laminin is an established component of the alveolar basement membrane (ABM) (7) and is biosynthesized by type II cells in vitro (21). These cells also actively synthesize a variety of other extracellular matrix (ECM) molecules as well, including fibronectin, type IV collagen, heparan sulfate proteoglycan, chondroitin sulfate (CS)-dermatan sulfate proteoglycan, glycosaminoglycans (GAGs) associated with a number of uncharacterized proteoglycans (for review, see Ref. 5), and entactin (30). Those known to be a part of the ABM are laminin (7), heparan sulfate proteoglycan (Perlecan) (25), CS, CS proteoglycan (Bamacan), entactin (25), and type IV collagen (2). The relationships between the alveolar epithelial cells and their underlying ABMs are important ones and are likely to influence all their various functions (1, 3, 31-33).
Laminin appears to play a prominent role in maintaining cellular phenotype and functional character as evidenced by type II cell behavior in vitro on purified laminin surfaces and laminin-enriched matrix substrata such as are found in Engelbreth-Holm-Swarm (EHS) sarcoma matrix (22). Importantly, laminin has also been shown to inhibit incorporation of thymidine into cellular DNA in serum-containing cultures compared with uncoated surfaces (22). The latter study also showed that EHS matrix (<60% laminin) was even more inhibitory to thymidine incorporation. The reason for this “added” inhibition was not clear but could be related to the basic complex nature of ABM matrix structure (7, 23, 36). As visualized immunocytochemically, laminin appears homogeneous and symmetrically distributed within the ABM (7), which differs markedly from the heterogeneous and asymmetrical distribution of sulfated macromolecules found there. Specifically, the ABM microdomains beneath type II cells are distinctly less sulfated than the same region adjacent to type I cells (8, 11, 23, 35). These anionic sites were shown to be chiefly associated with heparan sulfates (36), which raised the question of their functional significance with regard to type I and type II cells. Recent evidence (26) has indicated that CS and heparin can reduce the capacity of the type II cell to incorporate thymidine in response to heparin-binding growth factors in vitro, whereas their chemically desulfated forms restore this capacity. These data would seem to support the notion that sulfate per se can modulate type II cell responses to select growth factors (23), but, given the complex structure of the ABM in situ, it is likely that the interactions between soluble or bound growth factors, ECMs, and cell surfaces are multiple in nature. A key question therefore becomes, how do molecules such as laminin and associated (bound) sulfated proteoglycans (29), which have the tendency to inhibit DNA synthesis, behave in combination with specific growth factors that promote DNA synthesis in alveolar type II cells? The aim of this study was to determine whether laminin added to collagen substrata in combination with native or desulfated CS or heparin modified isolated rat type II cell responses to heparin-binding growth factors.
MATERIALS AND METHODS
Cell preparation. Type II pneumocytes were isolated from 200- to 240-g Fischer CDF (F-344) rats (Charles River Laboratories, Wilmington, MA) by the panning method of Dobbs et al. (4). Briefly, animals were anesthetized by intraperitoneal injection of Nembutal containing heparin and perfused via the hepatic portal vein with HEPES-saline solution until the lungs were free of blood. The lungs were then dissected free and lavaged via the trachea with a HEPES-saline solution containing glucose and EDTA. The lungs were then continuously insufflated via the trachea with 4.3 U of elastase (Worthington Biochemicals, Freehold, NJ; Boehringer Mannheim Biochemicals, Indianapolis, IN) at 5 cmH2O for 20 min, and the reaction was stopped with fetal bovine serum. The lungs were minced and filtered sequentially through two layers of 165-μm silk mesh followed by four layers of 42-μm silk mesh, and the recovered cells were centrifuged and panned on IgG-coated bacteriological plastic culture dishes for 1 h at 37°C in a 10% CO2incubator. Adherent cells were discarded, and nonadherent cells were examined with a modified Papanicolaou stain and judged to be 90–95% type II cells as evidenced by lamellar body-like structures and cell size. These cells were 80–95% viable as defined by trypan blue exclusion.
Matrix substrata preparation.Twenty-four-well tissue culture plates (Becton Dickinson, Lincoln Park, NJ) were first coated with 6 μg/cm2 of type I collagen (Sigma, St. Louis, MO), and the coating was allowed to dry at room temperature overnight in a laminar flow hood. This treatment was designed to optimize adherence of the ECM molecules to the culture surface. Optimal ECM concentration was established in a previous study (26), so the focus in these experiments was on the impact each specific ECM combination had on the growth factor conditions. Designated chambers were then coated with 16 μg/cm2 of laminin (Collaborative Research Products, Bedford, MA) either alone or in combination with 35 μg/cm2 of CS [type A; major repeating disaccharide, β-glucuronic acid-[1→3]-N-acetyl-β-galactosamine-4-sulfate-[1→4], 70% pure (remaining is chondroitin 6-sulfate); Sigma], high-molecular-weight heparin (13,500–15,000 mol wt; Calbiochem, La Jolla, CA), or their specifically desulfated forms (see below). In selected experiments, 100 ng/ml of either acidic fibroblast growth factor (FGF-1) or basic fibroblast growth factor (FGF-2) were mixed with the matrix substrata preparations (final concentration of 15 ng/cm2). ECMs were evaporated to dryness in a laminar flow hood, covered, and stored at 3°C until use. In separate experiments, laminin was added to ECM substrata at concentrations that varied from 0.1 to 1,000 μg/ml (0.016–160 μg/cm2).
The CS and high-molecular-weight heparin were subjected to solvolytic desulfation with dimethyl sulfoxide (18). The treated and nontreated samples were dialyzed against deionized water, and the content of covalently bound sulfate was measured with a Dionex conductivity detector (17) after hydrolysis in 6 N HCl vapor at 109°C for 6 h and separation on an anion-exchange column (Dionex, Atlanta, GA). A sample without hydrolysis served as the blank to correct for free sulfate. Only ECM components that were >85% desulfated were used in testing type II cell responses. Specifically, N-desulfated high-molecular-weight heparin was purchased from Sigma.
Cell culture. Isolated cells were seeded at 1 × 103/mm3on the various substrata in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum (Life Technologies, Gaithersburg, MD) overnight in a 10% CO2incubator at 37°C. In selected experiments, 100 ng/ml of FGF-1 or FGF-2 were mixed with matrix substrata beforehand or were added to the medium at seeding. Optimal growth factor concentrations had been established (15, 20) and confirmed in our preparations (26). Attached cells (>85%) were washed with serum-free hormonally defined (SFHD) medium (12) and then incubated with the same medium with or without one of the following growth factors: 100 ng/ml of FGF-1 (bovine, Collaborative Research Products), 100 ng/ml of FGF-2 (bovine, Collaborative Research Products), 10 ng/ml of keratinocyte growth factor (KGF; FGF-7; R&D Systems, Minneapolis, MN), 10 ng/ml of transforming growth factor (TGF)-α (R&D Systems), or 10 ng/ml of TGF-β (R&D Systems). The latter two were utilized as growth factor controls, and serum-free medium without growth factor added served as an additional control. The levels chosen for each were designed to optimize cell responses and were fixed so as to focus on variations in response to the ECM substrata outlined above. To the SFHD medium with or without growth factors, 1 μCi/ml of [3H]thymidine (final concentration) was added, and cells were cultured for 48 h. Cell lysates were then processed and counted in a scintillation counter according to standard methods (22). The remaining volume was used for determining the protein content of each sample. Treatment groups were organized and analyzed in quadruplicate. Experimental protocols were repeated a minimum of seven times. In parallel preparations, some cells were fixed with either 1% paraformaldehyde or acid alcohol (70% ethanol and 1% HCl) for histochemical localization of type II cell-specific antigens (26).
Data analysis. Means ± SE for the disintegrations per minute per microgram of protein were developed for the various treatment groups (in quadruplicate for each experiment). Different treatment groups within experiments were compared with laminin control substrata by analysis of variance (ANOVA, E-Z Stat, Trinity Software, Campton, NH) within experiments.P values < 0.05 were considered significant.
Data are also expressed as a percentage of control value (laminin treatment alone for each growth factor treatment) so that The significance of a percent value above or below 100% (control) was determined by the P value (<0.05) established by the raw data. The results presented in graphic form were representative experiments that were consistent with or closely resembled the outcomes of at least five or more repeats of the same protocol.
Cell adhesion assay. To ensure that the starting cell numbers for the various treatments were comparable, we evaluated adhesion according to standard methods (34). After the 24-h seeding period in serum-containing medium, cells on the various matrices were washed with PBS. Attached cells were then either treated with additional medium-growth factor combinations or terminated. After an additional 48 h, attached cells in the medium-growth factor combinations were terminated by fixation with 3.7% Formalin for 8 min at room temperature and stained with 1% toluidine blue in PBS for 2 h at room temperature. Cells were then washed three times, air-dried, and solubilized with 2% SDS for 30 min at 37°C. Absorbance of the solubilized cells was measured in a spectrophotometer at 650 nm. Nonadherent cells were similarly quantitated after centrifugation and identical treatment.
For the short duration of these experiments (72 h total), no changes in phenotype were detected for the isolated type II cells under any of the various treatments as defined by alkaline phosphatase and surfactant protein A histochemistry (26).
Isolated type II cells seeded on laminin plus desulfated heparin exposed to FGF-1 showed a significant increase (232% of control values) in incorporation of thymidine compared with laminin alone (Fig.1). Untreated heparin caused a modest reduction in thymidine uptake that was not significant compared with laminin controls. Similarly, laminin plus desulfated heparin substrata and FGF-2 stimulated a significant increase (256% of control values) in thymidine incorporation by type II cells (Fig.2), with the modest increases in other groups not significant. Cells exposed to FGF-7 showed significant increases in thymidine uptake on substrata containing either laminin plus desulfated CS (174% of control values) or laminin plus N-desulfated heparin (190% of control values) (Fig.3); those containing laminin plus heparin showed decreased thymidine uptake, but the decrease was not significant.
For comparison, TGF-α was used as a non-heparin-binding growth factor, with unexpected results. Type II cell responses to TGF-α on the substrata used in this study were all stimulatory above laminin alone. Specifically, laminin plus CS, laminin plus desulfated CS, and laminin plus N-desulfated heparin stimulated dramatic increases in thymidine uptake (561, 456, and 437% of laminin controls, respectively; Fig. 4). Thymidine incorporation results from experiments with cells exposed to either serum-free medium without growth factors or TGF-β (additional growth factor control) were extremely low, and comparisons between groups were not significant (data not shown).
Figure 5 compares the raw data of the thymidine incorporation by type II cells exposed to the various growth factor and matrix substrata conditions. Although all growth factor concentrations used were considered optimal for thymidine incorporation by published standards, the responses on the substrata used were highly varied. For instance, laminin alone and laminin plus heparin exposures resulted in the lowest thymidine incorporation with all growth factors, with the notable exception of TGF-α. Responses to FGF-1, FGF-7, and TGF-α were generally high, with the former two highest on desulfated matrices and lowest on sulfated matrices. FGF-2 followed a pattern similar to the other heparin-binding growth factors but at a generally lower level of incorporation. TGF-α was surprising with its very low response to laminin alone but with highly elevated responses to all matrix substrata combinations with laminin regardless of sulfation.
Type II cells were exposed to different concentrations of laminin substrata alone in the presence of 10 ng/ml of KGF to determine whether there was a dose relationship with its effects on thymidine incorporation (Fig. 6). A slight elevation over collagen alone was observed at 0.016 μg/cm2 (0.1 μg/ml) of laminin, followed by a steady decrease in thymidine incorporation toward 160 μg/cm2. Laminin is typically used as a substrate in tissue culture between 0.16 and 16 μg/cm2 (the latter used in the present experiments). These levels are generally considered to yield optimal cell attachment with acceptable levels of growth.
The impact of the timing and mode of growth factor addition on thymidine uptake by type II cells was tested by adding either FGF-1 or FGF-2 to the seeding medium or to the preformed culture substrata. The results were particularly profound for FGF-1, which stimulated 150–190% increases in thymidine uptake over cells for which FGF-1 was mixed with laminin alone (Fig. 7). Unlike the results seen in Fig. 1, in which FGF-1 was added only after 24 h of seeding in conventional serum-containing medium (no additional growth factors added), all matrix substrata were significantly stimulatory of thymidine uptake regardless of sulfation. This changed somewhat when FGF-1 was added to the medium instead of to the matrix, which resulted in significant elevations in thymidine uptake only in the desulfated CS and N-desulfated heparin groups (165 and 150% of laminin controls), although all were elevated. The effects were such that the typical reduction in thymidine uptake expected from sulfated matrix substrata was lost. The impact of this approach was less profound for FGF-2, with a significant increase observed only in the desulfated CS group (530% of laminin control value; Fig.8). When these growth factors were mixed with collagen only (no laminin or sulfated ECMs), significant elevation in thymidine uptake only occurred with FGF-1 and FGF-2 when added to the medium (Fig. 9), a result that was the reverse of that seen with laminin mixtures.
To attempt to define the influence the various treatments had on adhesion, we exposed the cells to standard serum-containing medium on the various ECM substrata, followed by growth factor addition at 24 h. At 24 h, laminin plus desulfated heparin was the only matrix to give modest but significant enhancement of adherence (137% of laminin alone). The addition of FGF-1 enhanced adherence between the 24- and 48-h cultures on laminin plus CS and laminin plus desulfated CS, whereas FGF-7 depressed adherence on laminin plus heparin and laminin plus N-desulfated heparin (Fig. 10). The other growth factors tested did not influence the adherence of type II cells on the various substrata in a significant fashion.
Laminin has previously been shown to support type II cell attachment and maintenance of phenotype, including cuboidal shape (22). In addition to being an important point of attachment for epithelial cells, specific domains of the laminin molecule act to structurally link or bind other components of the ABM, including heparan sulfate proteoglycans (14) and entactin (6). These structural relationships are well established, but their biological consequences may be much greater than presently appreciated. In other biological systems, for instance, FGF-2 is a survival factor for embryonal carcinoma cells cultured on plastic but becomes mitogenic for the same cells when cultured on laminin substrata (28). Laminin has also been shown to directly affect gene expression as demonstrated for the β-casein gene in mammary epithelium in the presence of lactogenic hormones (35). These data could be relevant to the current study, which was undertaken to determine whether the capacity of type II cells to incorporate thymidine into DNA in response to selected growth factors was affected by the presence of laminin alone or in combination with selected ECMs with varying levels of sulfate.
Previous studies have demonstrated that the heparin-binding growth factors FGF-1, FGF-2, and FGF-7 (KGF) stimulate isolated type II cell incorporation of thymidine into DNA (15, 16, 20), and FGF-1 and FGF-2 have been immunolocalized in the type II cell and ABM, respectively (24). The effects of FGF-1 and FGF-2 on bromodeoxyuridine (BrdU) incorporation by isolated type II cells were shown to be inhibited by sulfated CS and low- and high-molecular-weight heparins, and this phenomenon was reversed by their chemical desulfation (26). In the latter study, in which the “base” substratum was collagen type I, laminin was shown to modestly support BrdU incorporation by type II cells exposed to FGF-1 and FGF-2. Under the conditions used in the present study, type II cell incorporation of thymidine into DNA was not supported by laminin alone in the presence of most heparin-binding growth factors tested (FGF-1, FGF-2, and FGF-7), and higher concentrations tended to be even less supportive (see Fig. 6). This would seem to support previous observations on the effects of laminin on type II cell DNA synthesis (22). In the present study, desulfated high-molecular-weight heparin combined with laminin gave significant increases in thymidine incorporation compared with laminin substrata alone in the presence of FGF-1 or FGF-2, whereas desulfated CS and N-desulfated heparin significantly enhanced FGF-7-stimulated DNA synthesis. These modulatory effects support a previous report (26) that indicated that desulfated CS supported BrdU incorporation into DNA when combined with type I collagen in the presence of FGF-2. Taken together, these observations support the notion that the undersulfated microdomains of adult ABMs beneath type II cells (8, 11, 23, 36) could act to support or promote responses to growth factors.
The aim of the present study was to develop an artificial combination of ECM components within substrata that mimicked the in vivo state of the ABM by including laminin with mixtures (high and low sulfate content) used in a previous study (26). Clearly, the data indicate that a lack of support of laminin for DNA synthesis (22) is dramatically overcome by selected desulfated GAGs in the presence of certain FGFs. The importance of sulfated molecules in the activity of FGF-1 and FGF-2 is well established (10, 19). FGF-1 and FGF-2 bind to cell surfaces via a dual-receptor system consisting of a low-affinity binding site that is a heparan sulfate and a high-affinity binding site (FGF receptor) that is linked to tyrosine kinase (10, 13). Successful activation of the FGF ligand by the low-affinity binding site is crucial for binding the high-affinity site and appears to be related to sulfation (9, 10). This dependency has been shown to be different between FGF-1 and FGF-2, with the former requiring a high content of 6-O-sulfate groups and the latter a fully N-sulfated decasaccharide enriched in 2-O- and 6-O-sulfated disaccharide units (10). Similarly, Guimond et al. (9) prevented FGF-2 activation and 3T3 fibroblast mitogenesis by selective 2-O- and 6-O-desulfation of heparin. It would be expected that the molecular constructs on the FGF ligand that interact with these low-affinity “activating” sites on the cell surface would be probable targets for competitive inhibition by CS and high-molecular-weight heparin used in the present study and as suggested previously (26). As noted above, however, these molecules were not inhibitory in the mixtures used here, whereas their desulfated forms were actually stimulatory. Although the mechanism(s) of this action is not clear from the data, it is possible that the reduced sulfate environment or the sugar sequences within the desulfated molecules themselves act in concert with cell-surface heparan sulfate proteoglycans (low-affinity receptor) to facilitate the necessary ligand-receptor interaction critical for subsequent signaling events. The fact that the sulfated forms of these GAGs were not as obviously inhibitory of DNA synthesis as previously described (26) could be explained by the possibility that the inherently “retarding effects” of laminin were maximal and could not be enhanced further by additional factors in the mixture. The role of laminin, if any, in modulating heparin-binding growth factor activity at the cell surface appears to be altered in a profound way by desulfated GAGs. These data, at the very least, demonstrate that complex mixtures of ECMs in the ABM, with their varying levels of sulfate, function in concert with one another, perhaps in sometimes subtle ways, to modulate type II cell responses to growth factors.
The rather dramatic, stimulatory effects of TGF-α on the various laminin-GAG substrata add another dimension to the discussion, although a somewhat puzzling one. Its effects on type II cell DNA synthesis were expected to be more modest, but this was not the case. The reason for the enhancing effects of the GAGs on TGF-α effects is not obvious from these experiments but, as suggested above, could reflect the complexity of receptor-ligand interactions at cell surfaces and their impact on signaling and other downstream events linked with DNA synthesis and proliferative events. These observations could not be readily explained and deserve further study.
It was of interest that the addition of FGF-1 to the preexisting substrata or the medium during seeding overwhelmed its more selective enhancement of thymidine incorporation by type II cells on laminin plus desulfated heparin. In fact, all ECM substrata showed increased thymidine incorporation compared with laminin alone. This could be explained by the possibility that FGF-1 can gain access sooner to or compete more effectively for its appropriate receptor complex by mixing with the substrata. By comparison, when mixed with the medium in soluble form, it was stimulatory in the laminin plus desulfated CS and laminin plus N-desulfated heparin groups. FGF-2 was only stimulatory when added in soluble form to the medium with the laminin plus desulfated CS. This effectively supports an earlier study (26) using BrdU incorporation as a measure of DNA synthesis.
Two important points arise from the adhesion studies. First, laminin plus desulfated heparin enhanced adherence of type II cells, which could have had an effect on the outcome of the thymidine incorporation results for FGF-1 and FGF-2 by ostensibly increasing the number of potential cells that could respond to the growth factors. Conversely, FGF-7 in the presence of laminin plus N-desulfated heparin had a low adherence (Fig. 9) yet high thymidine uptake (Fig. 3). These potential problems in interpretation are minimized by the normalization of thymidine incorporation with total protein.
An earlier report (26) presented data that support the notion that the heavily sulfated regions of the ABM, such as those found adjacent to type I cells, would be expected to retard or otherwise inhibit epithelial responses to heparin-binding growth factors. In combination with laminin, which characteristically does not support DNA synthesis by type II cells (22), these effects would predictably be even greater. This could explain the enhanced inhibitory effect of EHS matrix on thymidine incorporation by type II cells in the original study by Rannels et al. (22). The present study supports the converse of this corollary, wherein regions of low sulfate within the ABM, like those found beneath type II cells (8, 11, 23, 36), would be expected to promote or at least be less inhibitory to potential growth factor responses to heparin-binding growth factors (26), thus providing a “window” of controlled growth factor responsiveness within the alveolus. In further support of this paradigm are recent data demonstrating that type II cells maintained in culture for periods exceeding 7 days tend to biosynthetically add higher levels of sulfate to nearly all of ECM components, including a novel sulfated laminin, compared with a rather modest level of sulfation within the first 7 days (27). Cells maintained in this manner would be considered more “type I cell-like,” thus suggesting that type I cells may actively biosynthesize/maintain the heavily sulfated ABM with which they are known to be associated in vivo (8, 11, 23, 36). Such a biosynthetic scheme could act as a further mechanism to help maintain stable cell populations under normal conditions by allowing certain cell types of common lineage to divide as necessary while preventing excessive responsiveness to growth factors within a confined anatomic region.
The authors are indebted to the support and helpful discussions with Dr. Pi-Wan Cheng of the University of Nebraska Medical Center.
Address for reprint requests: P. L. Sannes, Dept. of Anatomy, Physiological Sciences, and Radiology, College of Veterinary Medicine, North Carolina State Univ., 4700 Hillsborough St., Raleigh, NC 27606.
These studies were supported by National Heart, Lung, and Blood Institute Grant HL-44497 and a grant from the State of North Carolina.
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. §1734 solely to indicate this fact.
- Copyright © 1998 the American Physiological Society