Functional characterization of serum- and growth factor-induced phenotypic changes in intact bovine tracheal smooth muscle

Culturing of bovine tracheal smooth muscle (BTSM) strips for up to 8 days in the presence of 10% foetal bovine serum caused a time-dependent (t 1/2 =2.8 days) decrease in maximal contraction (E max ) to methacholine compared to serum-deprived controls (E max =74±4% at day 8). A reduced E max was also found using insulin-like growth factor-1 (30 ng ml −1 ) and platelet-derived growth factor (30 ng ml −1 ), but not using epidermal growth factor (10 ng ml −1 ) (E max =83±3, 67±8, 100±4%, respectively). Similar serum and growth factor-induced changes in E max were found for KCl-induced contraction (65±9, 80±7, 64±11% and 107±2%, respectively).

A role for extracellular matrix components has also been implicated in regulating airway smooth muscle phenotype, showing that fibronectin and collagen type I favoured progression towards a synthetic phenotype, whereas laminin and matrigel (a solubilized basement membrane matrix) strongly inhibited the progression towards a synthetic phenotype ( Hirst et al., 2000a ). Knowing that cell to cell contacts and extracellular matrix components are preserved in intact smooth muscle, we were interested to learn about the occurrence of functional changes in intact bovine tracheal smooth muscle (BTSM). We evaluated the contractile properties of BTSM strips exposed to serum and to different growth factors (platelet-derived growth factor (PDGF), insulin-like growth factor-1 (IGF-1) and epidermal growth facor (EGF)) in relation to their mitogenic potencies. In addition, we investigated the role of PI-3 kinase and ERK1/2 signalling pathways in the phenotypic changes in contractility in response to stimulation with PDGF.

Dulbecco’s modification of Eagle’s Medium (DMEM) and methacholine hydrochloride were obtained from ICN Biomedicals (Costa Mesa, CA, U.S.A.). Foetal bovine serum, NaHCO 3 solution (7.5%), HEPES solution (1 M), sodium pyruvate solution (100 mM), non-essential amino acid mixture, gentamycin solution (10 mg ml −1 ), penicillin/streptomycin solution (5000 u ml −1 /5000 μg ml −1 ) and amphotericin B solution (250 μg ml −1 ) (Fungizone) were obtained from Gibco BRL Life Technologies (Paisley, U.K.). Platelet-derived growth factor AB (PDGF-AB, human recombinant), epidermal growth factor (EGF, human recombinant), insulin-like growth factor 1 (IGF-1, human recombinant), insulin (from bovine pancreas), apotransferrin (human), soybean trypsin inhibitor and (−)isoprenaline hydrochloride were from Sigma Chemical Co. (St. Louis, MO, U.S.A.). PD 98059 and LY 294002 were obtained from Tocris Cookson Ltd. (Bristol, U.K.). L(+)ascorbic acid was from Merck (Darmstadt, Germany). [methyl- 3 H]-thymidine (specific activity 25 Ci mmol −1 ) was obtained from Amersham (Buckinghamshire, U.K.) Papain and Collagenase P were from Boehringer (Mannheim, Germany). All other chemicals were of analytical grade.

BTSM cells were plated in 24-well cluster plates at a density of 50,000 cells per well directly after isolation and were allowed to attach overnight in 10% FBS containing medium. Cells were washed twice with sterile phosphate buffered saline (PBS, composition (mM): NaCl 140.0; KCl 2.6; KH 2 PO 4 1.4; Na 2 HPO 4 .2H 2 O 8.1; pH 7.4) and made quiescent by incubation in FBS-free medium, supplemented with apo-transferrin (5 μg ml −1 ), ascorbate (100 μM) and insulin (1 μM) for 72 h. Cells were then washed with PBS and stimulated with mitogens in FBS- and insulin-free medium for 28 h, the last 24 h in the presence of [ 3 H]-thymidine (0.25 μCi ml −1 ). After incubation the cells were washed twice with PBS at room temperature and once with ice-cold 5% trichloroacetic acid (TCA). Cells were treated with this TCA-solution on ice for 30 min and subsequently the acid-insoluble fraction was dissolved in 1 ml NaOH (1 M). Incorporated [ 3 H]-thymidine was quantified by liquid-scintillation counting using a Beckman LS1701 β-counter.

After the removal of mucosa and connective tissue, tracheal smooth muscle was chopped using a McIlwain tissue chopper, three times at a setting of 500 μm and three times at a setting of 100 μm. Tissue particles were washed two times with the medium mentioned above, supplemented with 0.5% FBS. Enzymatic digestion was performed in the same medium, supplemented with collagenase P (0.75 mg ml −1 ), papain (1 mg ml −1 ) and soybean trypsin inhibitor (1 mg ml −1 ). During digestion, the suspension was incubated in an incubator shaker (Innova 4000) at 37°C, 55 r.p.m. for 20 min, followed by a 10 min period of shaking at 70 r.p.m. After filtration of the obtained suspension over 50 μm gauze, cells were washed three times in medium supplemented with 10% FBS.

Tissue strips, collected from suspension culture flasks, were washed with several volumes of KH-buffer pregassed with 5% CO 2 and 95% O 2 , pH 7.4 at 37°C. Subsequently, strips were mounted for isometric recording (Grass force-displacement transducer FT03) in 20 ml water-jacked organ baths, containing KH-buffer at 37°C, continuously gassed with 5% CO 2 and 95% O 2 , pH 7.4. During a 90 min equilibration period, with washouts every 30 min, resting tension was graduallly adjusted to 3 g. Subsequently, muscle strips were precontracted with 20 and 30 mM isotonic KCl solutions. Following two wash-outs, basal smooth muscle tone was established by the addition of 0.1 μM isoprenaline and tension was re-adjusted to 3 g, immediately followed by two changes of fresh KH-buffer. After another equilibration period of 30 min cumulative concentration response curves (CRCs) were constructed to stepwise increasing concentrations of isotonic KCl (5.6–50 mM) or methacholine (1 nM–100 μM). When maximal KCl or methacholine-induced tension was obtained, the strips were washed several times and basal tone was established using isoprenaline (10 μM).

Bovine tracheae were obtained from local slaughterhouses and rapidly transported to the laboratory in Krebs-Henseleit (KH) buffer of the following composition (mM): NaCl 117.5, KCl 5.60, MgSO 4 1.18, CaCl 2 2.50, NaH 2 PO 4 1.28, NaHCO 3 25.00 and glucose 5.50, pregassed with 5% CO 2 and 95% O 2 ; pH 7.4. After dissection of the smooth muscle layer and careful removal of mucosa and connective tissue, tracheal smooth muscle strips were prepared while incubated in gassed KH-buffer at room temperature. Care was taken to cut tissue strips with macroscopically identical length (1 cm) and width (2 mm). Tissue strips were washed once in sterile Dulbecco’s modification of Eagle’s medium (DMEM), supplemented with NaHCO 3 (7 mM), HEPES (10 mM), sodium pyruvate (1 mM), nonessential amino acid mixture (1 : 100), gentamicin (45 μg ml −1 ), penicillin (100 u ml −1 ), streptomycin (100 μg ml −1 ) and amphotericin B (1.5 μg ml −1 ). Next, tissue strips were transferred into suspension culture flasks and a volume of 7.5 ml medium, with or without 10% foetal bovine serum (FBS), was added per tissue strip. Strips were maintained in culture in an incubator shaker (37°C, 55 r.p.m.) for a maximum of 8 days, refreshing the medium on day 4. When applied, growth factors (PDGF, IGF-1, EGF) or kinase inhibitors (PD 98059, LY 294002, given 30 min prior to the growth factors) were added in a small volume (7.5 μl per tissue strip). Culture flasks containing kinase inhibitors were protected from light during the whole experiment. Occasionally, some strips were used for contraction experiments immediately after preparation.

Since the growth factor-induced depression of BTSM contraction depended on the proliferative effect of the growth factors used, we explored the involvement of proliferation-associated signalling pathways in growth factor-induced depression of maximal contractility to methacholine. For this purpose, PD 98059 (30 μM), an inhibitor of the MAPK/ERK pathway or LY 294002 (10 μM), an inhibitor of PI-3 kinase were added to the culture medium 30 min before the addition of PDGF (10 ng ml −1 ). In agreement with the above observations, PDGF alone reduced the maximum methacholine-induced contraction ( ). However, in the presence of PD 98059, no measurable change in contractility was induced by PDGF ( ). PD 98059 by itself did not significantly affect maximal methacholine-induced contraction or sensitivity (E max =20.5±3.0 and 22.1±2.1 g and pEC 50 =7.29±0.08 and 7.15±0.10 for PD 98059 and vehicle treated strips, respectively). Also in the presence of LY 294002, PDGF did not induce a measurable change in methacholine-evoked contraction ( ). However, pretreatment with LY 294002 by itself induced a significant depression of maximum contraction compared to vehicle-treated strips (E max =18.9±2.0 and 22.1±2.1 g for LY294002 and vehicle treated strips, respectively, P<0.05). The potency of methacholine-induced contraction remained unaltered (pEC 50 =7.29±0.07 and 7.15±0.10 for LY294002 and vehicle treated strips, respectively).

As indicated in and , the effects of growth factors on maximal contraction of BTSM strips to methacholine or KCl depended on the nature and the concentration of the growth factor used. The proliferative responses of isolated BTSM cells, measured as [ 3 H]-thymidine incorporation, were also dependent on the nature and concentration of the growth factors under investigation, with a reciprocal rank order of potency ( ). A strong correlation was observed between the efficacy of growth factors to induce depression of maximal methacholine or KCl-induced contraction of BTSM strips and the proliferative potency of these factors in BTSM cells (r=0.97, P=0.002 and r=0.93, P=0.007 for methacholine and KCl-induced contraction, respectively; )

To establish whether the growth factor-induced effects on BTSM contractility were limited to methacholine-induced contraction, contraction to KCl was also determined in BTSM strips pre-treated with growth factors for 8 days. As observed for methacholine, maximal KCl-induced contraction was sensitive to pretreatment with IGF-1 (30 ng ml −1 ) and PDGF (30 ng ml −1 ), but not EGF (10 ng ml −1 ). Moreover, changes in maximal KCl-induced contraction depended on the concentration of PDGF used. The potency (EC 50 ) of KCl was unchanged after treatment with growth-factors, irrespective of the growth factor applied. Interestingly, changes in KCl and methacholine-induced contraction were similar, both qualitatively and quantitatively ( ).

The ability to alter contractile behaviour was not limited to FBS treatment. Treatment for 8 days with IGF-1 (30 ng ml −1 ) and PDGF (30 ng ml −1 ) were also effective in inducing a statistically significant depression of the E max for methacholine-induced contraction compared to vehicle-treated BTSM strips ( ). No effect was observed for EGF (10 ng ml −1 ), however. The PDGF-induced decrease in E max was concentration-dependent ( ). As with FBS-treated BTSM strips, the pEC 50 of methacholine-induced contraction was unaltered upon treatment with the above mentioned growth factors ( ).

The development of the differences in maximal methacholine-induced contraction between BTSM strips cultured in FBS-enriched and in FBS-free media was time-dependent ( ). The difference in E max between FBS-treated and FBS-free conditions was maximal at day 8, with a half-maximal effect after 2.8 days ( ). Comparison with results obtained with freshly used strips shows that the E max of strips maintained in FBS-free and FBS-enriched media diverged in time ( ).

Maximal methacholine-induced contractile force (E max ) of BTSM strips cultured for up to 8 days in the presence of 10% FBS was significantly lower compared to strips cultured in FBS-free medium (E max =17.2±0.8 and 22.2±2.2 g, respectively; P<0.05; ). No difference in the sensitivity to methacholine was found (pEC 50 =6.99±0.10 and 6.99±0.05 for FBS-treated and FBS-deprived BTSM strips, respectively).

Discussion

FBS treatment-induced decrease of contractile function has been demonstrated recently in organ-cultured vascular tissue. In cultured rat tail arterial smooth muscle, contractile responses to both noradrenaline and 60 mM K+ were diminished upon FBS-treatment for 4 days (Lindqvist et al., 1997). A similar effect has also been observed in guinea-pig ileum smooth muscle, which showed a decreased contractility to carbachol and 60 mM K+ after treatment with FBS-rich medium (Gomez & Sward, 1997). This loss of contractile function was explained by a continuously elevated intracellular [Ca2+]i upon culturing in FBS, thus leading to a decreased function of voltage-operated calcium channels (VOC) and/or a decreased sensitivity of the contractile apparatus for Ca2+. Effects of long-term exposure to FBS on contraction of organ cultured airway smooth muscle have not yet been described. FBS contains a variety of mitogenic stimuli and it causes contraction of canine ASM (Abdullah et al., 1994). Therefore, long-term effects of FBS are not necessarily the consequence of proliferative stimulation, but may also be caused by continuously elevated [Ca2+]i levels or stimulation of the contractile apparatus.

Our data show that pretreatment with both 10% FBS and the growth factors PDGF and IGF-1, followed by a washout period of over 120 min, induce a depression in maximal response to methacholine. The observed effects for 10% FBS are time-dependent. The diverging change of Emax in time ( ) implies that the differences in Emax on day 8 are the consequence of both deprivation-induced increases in Emax and FBS-induced decreases in Emax. It is crucial for the interpretation of our results, that KCl-induced contraction and methacholine-induced contraction were influenced similarly by long-term exposure to growth factors, both quantitatively and qualitatively. These results imply that KCl and methacholine-induced contraction were affected by a common mechanism. However, methacholine induces contraction through stimulation of PI-turnover and subsequent release of Ca2+ from intracellular stores and influx through receptor operated Ca2+-channels (Zaagsma et al., 1997). In contrast, KCl-induced contraction is fully dependent on VOC mediated Ca2+-influx and does not require receptor mediated stimulation of PI-turnover. Therefore, quantitative similarities between KCl and methacholine-induced contraction can be achieved only by affecting contraction downstream intracellular Ca2+-increases. Further evidence that changes in [Ca2+]i or in the sensitivity of the contractile apparatus to Ca2+ cannot explain the 10% FBS or growth factor induced effects, comes from the observation that the potency (pEC50) of methacholine is not affected by culturing in the presence of 10% FBS or growth factors. Methacholine requires only fractional stimulation of PI-turnover and subsequent elevations of [Ca2+]i for BTSM contraction (Hoiting et al., 1996; Meurs et al., 1988). Altering events upstream of changes in [Ca2+]i or in the sensitivity of the contractile apparatus for Ca2+, should therefore be accompanied by a shift in the concentration-response curve to methacholine.

Modulation towards a less contractile phenotype is associated with a decrease in contractile protein expression (Mitchell et al., 2000), although changes in cytoskeletal organization should also be considered as an explanation for the altered maximal contraction (Halayko & Solway, 2001). However, agonist-induced cytoskeletal reorganization occurs rapidly (within 5 min) in human ASM cells (Togashi et al., 1998). Considering the long-term nature of the 10% FBS-induced decline in maximal contractile response, growth-related changes towards a less contractile phenotype become the most likely explanation for the observed effects. Western analysis of proteins harvested from canine ASM cells in primary culture shows a decline in the expression of contractile proteins (e.g. smooth muscle myosin heavy chain, smooth muscle α-actin) in the period prior to proliferation (Halayko et al., 1996). We therefore hypothesized that the degree of decline in contractile response should be reciprocally related to the degree by which proliferation is stimulated as a consequence of phenotypic modulation. Indeed, a strong correlation exists between the proliferative responses of the applied growth factors on BTSM cells and the degree of depression in contractile response, both for KCl and methacholine. Since our proliferation data are consistent with earlier findings in BTSM cells (Kelleher et al., 1995), a similar relationship could be constructed using our contraction data and the proliferative data published by Kelleher et al. (1995) (r=0.98 and r=0.97 for methacholine and KCl, respectively). The relative inactivity of EGF, applied in its maximally mitogenic concentration in BTSM cells, is in agreement with the low proliferative efficacy as obtained by us and others (Kelleher et al., 1995).

The MAPK/ERK pathway and the PI 3-kinase pathway were examined in order to gain insight in the mechanisms involved in the growth factor-induced change in phenotype. The inhibition of the MAPK/ERK-pathway using PD 98059 inhibited the development of contractile depression at a concentration that has been shown to fully inhibit PDGF-induced proliferation in BTSM cells (Karpova et al., 1997). This supports the hypothesis, that the development of a decreased contractile function is the consequence of proliferation-associated phenotypic modulation to a less contractile state. Interestingly, modulation of arterial smooth muscle cells towards a synthetic phenotype induced by FBS could also be inhibited by PD 98059, though only partially (Roy et al., 2001).

Inhibition of the PI 3-kinase pathway by LY 294002, applied during organ culture at a concentration known to reduce PDGF-mediated proliferation of BTSM cells to basal levels (Walker et al., 1998), inhibited the PDGF induced decrease of contractile response, suggestive for inhibition of modulation to a less contractile phenotype. In the absence of PDGF, LY 294002 treatment itself lowered maximal contraction to methacholine. This effect might be attributed to a diminished PI-3-kinase mediated maintenance of the cytoskeleton (Aplin et al., 1998; Rodriguez-Viciana et al., 1997). In addition, the effects of PI-3-kinase inhibition may also suggest an increase in apoptosis. This would imply apoptosis to be a significant regulatory mechanism in the intact BTSM strip. However, growth factors inhibit apoptosis, and this would induce enhancement rather than a depression of contractility. Although we do not know the exact nature of the LY 294002-induced effect, the inhibitory effect on PDGF-induced contractile depression is in agreement with the concept that proliferation and modulation to a less contractile state are stimulated simultaneously. Also, it implies a role for PI 3-kinase in functional modulation of BTSM.

Activation of the MAPK/ERK cascade is achieved via the PI 3-kinase pathway, but a redundant pathway stimulates MAPK/ERK independent of PI 3-kinase when a large number of receptors are activated (Duckworth & Cantley, 1997). Since LY 294002 and PD 98059 both totally inhibited the PDGF-induced change in phenotype, one might conclude that the pathway stimulated through PI-3-kinase, rather than the redundant MAPK/ERK-pathway, is responsible for the observed effects in phenotypic switching. Separate, parallel functioning pathways opposed to serially functioning pathways can, however, not be excluded: proliferation of BTSM cells induced by high concentrations of PDGF-BB has been reported to be inhibited completely by either PD 98059 or LY 294002, whereas PD 98059 does not inhibit PDGF-stimulated PI 3-kinase activity, nor does LY 294002 inhibit PDGF-stimulated MAPK activity (Walker et al., 1998).

Changes in phenotype may be relevant in asthma. Growth factors can be released in asthmatic airways from epithelial cells, inflammatory cells and ASM cells, or leaked into the airways as a consequence of extravasation, contributing to the formation of phenotypically altered myocytes with increased proliferative and synthetic capabilities. When extravasation ceases and growth factor concentrations return to basal levels redifferentiation to the contractile phenotype or even to a hypercontractile phenotype may occur as has been shown in cultured canine ASM cells (Halayko et al., 1999). Induction of synthetic smooth muscle phenotypes as postulated to occur in asthma (Hirst, 1996) is therefore not favourable and may contribute to typical features of asthmatic airways, such as hyperreactivity, irreversible airflow obstruction and the progressive increase in the severity of the disease.

In conclusion, BTSM contractility is sensitive to treatment with FBS or growth factors in vitro. Both receptor-dependent (methacholine) and receptor-independent (KCl) stimulation of BTSM, treated with FBS or growth factors, led to a decrease in maximal contraction to these agonists. This implies that changes occurred at the level of the contractile machinery. Moreover, the decline in contractile response correlates with the proliferative conditions of the culture medium and can be totally inhibited by PD 98059 and LY 294002. As a consequence, it seems that intact BTSM is sensitive to phenotypic changes and that these changes are regulated through signalling pathways involving both the MAPK/ERK and the PI-3 kinase-pathway.