On the mechanism of induction of dna synthesis in cultured arterial smooth muscle cells by leukotrienes: Possible role of prostaglandin endoperoxide synthase products and platelet-derived growth factor

Endothelial cell injury, platelet adhesion/degranulation, invasion of the subendothelium by leukocytes from the blood and smooth muscle cells (SMCs) from the media, and intimal SMC proliferation are important early events in atherogenesis (Ross, 1986; Schwartz et al. 1986). During the course of this process, a complex interplay between different cell types and a multitude of cellular products takes place. The eicosanoids represent one group of biologically active molecules that are of interest in this context. These compounds, including prostaglandins (PGs), leukotrienes (LTs) and monohydroxy acids (HETEs), are synthesized from arachidonic acid via the prostaglandin endoperoxide synthase and lipoxygenase pathways, and have been implicated in various aspects of vascular homeostasis (Curtis-Prior, 1988; Watkins et al. 1989). During recent years, increasing attention has been paid to the possible role of eicosanoids in the regulation of differentiated properties and proliferation of arterial SMCs (Pomerantz and Hajjar, 1989; Thyberg et al. 1990b). PGI₂, PGE₂ and PGF_2α are the principal arachidonic acid metabolites of vascular tissues and are synthesized by endothelial cells as well as SMCs (Needleman et al. 1986; Smith, 1986; Powell and Funk, 1987). Leukotrienes are formed by granulocytes and monocytes/macrophages (Samuelsson, 1983). In addition, endothelial cells, platelets and SMCs may convert LTA₄ produced by granulocytes and monocytes/macrophages into LTB₄ and/or LTC₄; the latter substance may be further metabolized into LTD₄ and LTE₄ (Feinmark and Cannon, 1986,1987; Claesson and Haeggström, 1988; Edenius et al. 1988; Claesson et al. 1990). Hydroxyeicosatetraenoic acids (HETEs) are synthesized by leukocytes and a large variety of other cell types, in general via the lipoxygenase pathway. However, endothelial cells and SMCs form HETEs also via the prostaglandin endoperoxide synthase pathway (Spector et al. 1988). Both the prostaglandins and the leukotrienes are believed to exert their effects via specific receptors, but the responsible signal transduction machinery is in most cases poorly known (Cristol et al. 1989; Halushka et al. 1989).

With regard to proliferation of vascular SMCs, eicosanoids have been reported to exert both stimulatory and inhibitory effects. Platelet-derived growth factor (PDGF), a potent serum mitogen for cells of mesenchymal origin (Raines et al. 1990), was demonstrated to stimulate prostaglandin production by SMCs (Coughlin et al. 1980) and 3T3 fibroblasts (Shier, 1980; Rozengurt et al. 1983b; Habenicht et al. 1985). Moreover, exogenous PGE₁, PGE₂ and PGI₂ were found to inhibit the growth of SMCs cultured in the presence of serum, whereas PGF_2n mostly lacked distinct effect (Huttner et al. 1977 ; Cornwell et al. 1979; Pietilà et al. 1980; Sjölund et al. 1984; Smith et al. 1984; Loesberg et al. 1985; Owen, 1986; Morisaki et al. 1988a,b;Uehara et al. 1988). Likewise, PGE₁ and PGE₂ blocked induction of DNA synthesis by PDGF in serum-starved cells, whereas PGF_2α again was without effect (Nilsson and Olsson, 1984). On the other hand, PGE₁ showed a growth-promoting effect when added to serum-starved cells together with insulin (Owen, 1986). In a similar manner, PGE₁ and PGF_2α. were found to act synergistically with insulin to induce DNA replication in 3T3 fibroblasts (Jimenez de Asua et al. 1975; Otto et al. 1982; Rozengurt et al. 1983a), whereas endogenously formed PGE₂ inhibited the growth of polyoma virus-transformed 3T3 fibroblasts (Lindgren et al. 1979). These results indicate that prostaglandins affect cellular proliferation in a multifaceted way, depending on the type of cell and the experimental conditions.

Recent studies from our laboratory have pointed out that leukotrienes influence the growth of vascular SMCs in a dual way. First, LTB₄, LTC₄, LTD₄ and LTE₄ all stimulated the transition of the cells from a contractile to a synthetic phenotype (Palmberg et al. 1989), i.e. the overall change in differentiated properties that precedes the onset of cellular proliferation early in primary culture (Thyberg et al. 19905). Second, LTB₄, LTC₄ and LTD₄ were found to induce DNA synthesis in serum-starved secondary cultures (Palmberg et al. 1987). Treatment of the cells with prostaglandin synthesis inhibitors blocked the effect of LTB₄, but not that of LTC₄ and LTD₄ suggesting that LTB₄ acted via a prostaglandin endoperoxide synthase product. In the present report, the mechanism of action of leukotriene-induced DNA synthesis in vascular SMCs has been explored in further detail. Special attention was paid to the possible role of endogenously formed prostaglandin endoperoxide synthase products as mediators of the LTB₄ effect.

Bovine serum albumin (BSA), indomethacin and acetylsalicylic acid were purchased from Sigma Chemical Company (St Louis, MO, USA), Ham’s medium F-12, newborn calf serum (NCS) and collagenase from Gibco BRL (Paisley, Scotland), trypsin (1:250) from Difco Laboratories (Detroit, MI, USA), and cell culture plastic from Nunc (Roskilde, Denmark). The culture medium was routinely supplemented with 10 mw Hepes/10 mM Tes (pH 7.3), 50 μg ml⁻¹ L-ascorbic acid and 50 μg ml⁻¹ gentamycin sulphate (medium F-12). Synthetic LTB₄, LTC₄, LTD₄ and LTE₄ were gifts from Dr T. Miyamoto, Ono Pharmaceutical Company (Osaka, Japan). PGF₁,, PGE₂, PGI₂, 6-keto-PGF_1α, 15(S)HETE and 5(S)HETE were purchased from Biomol (PA, USA). The leukotriene receptor antagonists FPL 55712 and ICI 198615 were obtained from Fisons (Loughborough, England) and ICI (Macclesfield, England), respectively. [³H]thymidine (5 Ci mmol⁻¹) and a radioimmunoassay kit for 15(S)HETE were purchased from Amersham International (Amersham, England) and [³H]arachi-donic acid, [³H]PGF_2α and radioimmunoassay kits for PGE₂ and PGI₂ from New England Nuclear (Dreieich, FRG). PDGF was isolated to homogeneity from human platelets and labelled with ¹²⁵I to a specific activity of about 30000ctsmin⁻¹ ng⁻¹ using the chloramine T method (Heldin et al. 1981, 1987). Antibodies against PDGF were produced in rabbits and immunoglobulins were purified from immune sera by chromatography on a column of protein A-Sepharose (Heldin et al. 1981).

SMCs were isolated from the aortic media of male Sprague Dawley rats (300–400 g) by digestion with 0.1% collagenase in medium F-12/0.1% BSA as described (Palmberg et al. 1987; Thyberg et al. 1990a). After rinsing and counting, the cells were seeded into 80 cm² plastic flasks (3– 10⁴ cells cm⁻²) and grown in medium F-12 with 10% NCS. The cultures were kept at 37 °C in a humidified atmosphere of 5% CO₂ in air and medium was changed every second day. At confluence, the cells were detached by exposure to 0.12% trypsin and 0.02% EDTA in calcium- and magnesium-free phosphate-buffered saline (PBS; pH 7.3) and reseeded in: 100-mm plastic dishes (6× 10³ cells cm^-2) for collection of conditioned media and assays of leukotriene production and arachidonic acid metabolism; 60-mm dishes (1×10⁴ cells cm⁻²) for measurement of prostaglandin and 15(S)HETE production; 35-mm dishes (l× 10⁴ cells cm⁻²) for analysis of prostaglandin metabolism; and 24-well multidishes on glass coverslips (1×10⁴ cells cm⁻²) for assay of DNA synthesis. The cells were grown to subconfluence in medium F-12 with 10 % NCS and growth-arrested by incubation in medium F-12/0.1% BSA for 48 h.

Growth-arrested secondary cultures were prepared as described above. The cells were exposed to 1 μCi ml⁻¹ of [³H]thymidine in experimental media for 24 h and fixed in 3% cacodylate-buffered glutaraldehyde. After dehydration in ethanol, the coverslips were mounted on glass slides, dipped in Kodak NTB2 emulsion, exposed at 4°C for 2 days, developed in Kodak D-19, and stained with 1 % Methylene Blue. The percentage of labelled nuclei was determined by counting 500 cells on each coverslip.

Subconfluent cultures were growth-arrested by incubation in medium F-12/0.1 % BSA for 48 h. The cells were then incubated with leukotrienes for another 48 h. Media were collected and centrifuged at 200g for 5 min to remove particulate material. The cells were detached by trypsinization and counted in a haemocytometer. Control media were prepared from cell-free dishes in the same manner. To assay for the presence of DNA synthesisstimulating activity, 75% conditioned medium was mixed with 25 % fresh medium and used directly in a test system of growth-arrested secondary cultures.

Binding experiments were performed on confluent cell layers in 12-well multidishes (Nistér et al. 1984; Sjölund et al. 1988). The cells were serum-starved for 12 h before binding. Cultures were rinsed with ice-cold PBS (pH 7.3) containing 0.1% BSA (binding medium) and exposed to conditioned media or varying amounts of PDGF (diluted in binding medium) for 180 min at 4°C. After rinsing with binding medium, the cells were exposed to ¹²⁵I-labelled PDGF (2 ng ml⁻¹) in binding medium for 45 min at 4°C. They were then rinsed five times with binding medium and lysed in 1 % Triton X-100/10 % glycerol/20 mM Hepes (pH 7.4). Finally, radioactivity was determined in an LKB-Wallac gamma counter. Cell numbers were determined in parallel dishes.

Serum-starved SMCs showed a low inherent rate of DNA replication and responded in a dose-dependent manner to stimulation with leukotrienes (B₄, C₄ and D₄) or PDGF, with labelling indices of 30–50 % and 50–80 %, respectively. The prereplicative lag phase was 16–18 h with LTB₄,12–14 h with LTC₄,14–16 h with LTD₄, and 10–12 h with PDGF (data not shown; cf. Palmberg et al. 1987). Indomethacin blocked induction of DNA synthesis by LTB₄ but not that by LTC₄, LTD₄ or PDGF. Moreover, the effects of LTC₄ and LTD₄ were inhibited by the cysteinyl-containing leukotriene receptor antagonists FPL 55 712 (≥10⁻⁷ M) and ICI 198615 (≥10⁻¹² M), while the effect of LTB₄ remained unaffected (Fig. 1A,B). These findings support our earlier notions that the leukotrienes act at an earlier stage in the mitogenic signalling pathway than PDGF, that different leukotrienes, at least partly, have different mechanisms of action, and that the effect of LTB₄ may be mediated via a prostaglandin endoperoxide synthase product (Palmberg et al. 1987).

To link further the induction of DNA replication by LTB₄ with synthesis of prostaglandin endoperoxide products, the relationship in time between these processes was studied. The results indicated that indomethacin could be added 8–10 h after LTB₄ with full inhibition. Thereafter, the effect gradually disappeared, and by 16 h it had vanished completely (Fig. 2). Arachidonic acid (20:4) -but not linoleic acid (18:2) -stimulated initiation of DNA synthesis when added to growth-arrested SMCs in serum-free medium (maximum effect at picomolar concentrations), and the effect was blocked by indomethacin (Fig. 3). Next, various metabolites of arachidonic acid were tested as possible mediators of the LTB₄ effect. When added alone, PGF_2α, PGE₂ and PGI₂ all induced DNA synthesis (maximum effects at nanomolar concentrations), whereas 6-keto-PGF_1α. was ineffective (Figs 4A, 5A). On the other hand, when added together with PDGF, PGE₂ and PGI₂ partially inhibited the effect of this mitogen, whereas no effects of PGF_2α and 6-keto-PGF_lα. were observed (Figs 4B, 5B). The monohydroxy acid 15(S)HETE -but not 5(S)HETE -was mitogenic for the SMCs under serum-free conditions (maximal effect at nanomolar concentrations; Fig. 6A). Indomethacin did not influence the cellular response to 15(S)HETE and this latter substance did not interfere with the effect of PDGF (Fig. 6B).

The degradation of leukotrienes in serum-free media was examined to determine if the stimulatory action on DNA synthesis of conditioned media from cultures incubated with leukotrienes was mediated by intact leukotrienes or other products made by the cells (see below). Tritium-labelled LTB₄ (Ing; 100000 cts min⁻¹) and LTC₄ (Ing; 100 000 cts min⁻¹) were added to 10 ml of serum-free medium without cells. The sample was divided in two parts: 5 ml of medium was immediately mixed with 5 volumes of ethanol and the residual 5 ml was incubated at 37 °C in an atmosphere of 5% CO₂ in air for 72 h prior to addition of ethanol. The products were isolated and analysed by reverse-phase HPLC as described (Palmberg et al. 1987). Fig. 7 shows the distribution of radioactivity in the chromatographic fractions. After 72 h of incubation, less than 5 and 10 % of the added radioactivity coeluted with standards of LTB₄ and LTC₄, respectively (Fig. 7B), compared with the sample that was not incubated at 37 °C (Fig. 7A). These results show that the tritium-labelled LTB₄ and LTC₄ were almost completely and nonenzymically degraded during 72 h of incubation with cell-free medium, suggesting that the leukotrienes did not mediate the effect of the conditioned media on DNA synthesis. In contrast, Fitzpatrick et al. (1983) found LTB₄, LTC₄ and LTD₄ to be stable for several days in aqueous solution at physiological pH. Hence, it is possible that components in the cell culture medium are responsible for bringing about the decomposition of the leukotrienes (10⁻¹¹M) observed in our experiments.

Radioimmunological analyses of conditioned media revealed that the SMCs were able to produce PGI₂ (determined as the stable metabolite 6-keto-PGF_1α), PGE₂ and 15(S)HETE, and that the production of these substances was blocked by indomethacin (Table 1). In the case of 15(S)HETE, the block was not complete, probably due to the fact that this substance also may be formed via the lipoxygenase or the monooxygenase pathway (Needleman et al. 1986; Spector et al. 1988). No distinct increase in the net release of prostaglandins and 15(S)HETE into the media during the 48-h period of conditioning was recorded after stimulation of the cells with LTB₄ (Table 1).

Cultures were incubated for 30 min with [³H]arachidonic acid (0.33 μCiml⁻¹) and calcium ionophore A23187 (1.0 μM) in PBS. The buffer was collected and mixed with 5 volumes of ethanol. The samples were evaporated and extracted with diethyl ether at pH 3.4. The lipid extracts were dissolved in methanol:H₂O:acetic acid (60:40:0.01, by vol.) and further analysed by reverse-phase HPLC. Collection of the eluted material in 1ml fractions and determination of radioactivity demonstrated that the cells produced radiolabelled material co-eluting with standards of 6-keto-PGF_1α, PGE₂ and PGF_2α (Fig. 8).

To assay for release of mitogenic material by SMCs incubated with leukotrienes, media were conditioned for 48 h and then transferred to growth-arrested cultures together with tritiated thymidine for another 24 h. Subsequent autoradiographic analysis demonstrated that media from cultures treated with LTB₄, LTC₄ or LTD₄ all contained stimulatory material (Fig.9). The activity induced by LTB₄ was abolished by treatment with indomethacin during the conditioning period. Furthermore, addition of antibodies against PDGF during the mitogenic assay blocked the activity of conditioned media from LTB₄-treated cultures. On the other hand, indomethacin and antibodies to PDGF had no effect on the activity of conditioned media from cultures exposed to LTC₄ and LTD₄. The leukotrienes themselves were inactivated during the incubation (see above), and media incubated for 48 h in the absence of cells did not induce DNA replication above control levels (Fig. 9). Analysis of the conditioned media using a receptor competition assay with radioiodinated PDGF, showed a 40–50 % stimulatory effect of LTB₄ on the release of PDGF-like material by the SMCs (Table 2).

The results of the present and earlier investigations (Palmberg et al. 1987) indicate that leukotrienes induce DNA replication in rat aortic SMCs with a longer lag phase than the polypeptide mitogen PDGF (LTB₄> LTD₄>LTC₄>PDGF). Moreover, the effects of LTC₄ and LTD₄ -but not that of LTB₄ -were blocked by the cysteinyl-containing leukotriene receptor antagonists FPL 55 712 and ICI 198615. These findings are in agreement with the hypothesis that LTB₄ and cysteinyl-containing leukotrienes act via distinct receptors (Cristol et al. 1989; Halushka et al. 1989). High-affinity binding sites for LTD₄ have been demonstrated in the plasma membrane of tracheal SMCs and shown to be involved in an agonist-dependent activation of phosphoinositide hydrolysis and calcium ion mobilization (Mong et al. 1988a,b; Crooke et al. 1989). These processes also appear to be activated by LTB₄ and LTC₄ (Cristel et al. 1989; Halushka et al. 1989), and are well established as parts of the signal transduction machinery for a large variety of growth factors and hormones (Rana and Hokin, 1990). Nevertheless, their exact roles in mitogenesis are still unclear.

As noted previously, the growth-promoting effect of LTB₄ was blocked by prostaglandin synthesis inhibitors like indomethacin and acetylsalicylic acid, whereas the effects of LTC₄ and LTD₄ remained unaffected (Palmberg et al. 1987). Neither was there any influence of prostaglandin synthesis inhibitors on PDGF-induced DNA replication, as confirmed by recent studies on 3T3 fibroblasts (Handler et al. 1990). These observations further emphasize that LTB₄ and the cysteinyl-containing leukotrienes have different mechanisms of action and suggest a role for prostaglandin endoperoxide synthase products in mediating the effect of LTB₄. In the present report, this possibility was explored in some further detail by examining the ability of the SMCs to synthesize prostaglandins and monohydroxy acids on their own and the ability of a few major prostaglandins and related compounds to induce DNA replication.

First, a closer look into the time dependence of the effect of indomethacin revealed that it inhibited induction of DNA replication by LTB₄ if added within 8–10 h after the leukotriene. This suggests a role for prostaglandin endoperoxide synthase products in an intermediate phase of the signalling pathway for LTB₄, beginning about 8–10 h before the cells enter the S phase. However, the possibility cannot be excluded that such products might be manufactured also during other parts of the prereplicative period. The lag in time between the start of exposure to LTB₄ and the inhibitory action of indomethacin on induction of DNA replication may be due to the possibility that LTB₄ acts indirectly via one or several other substances that first must be produced. These substances may, for example, be involved in up-regulation of the enzyme systems responsible for release of arachidonic acid from phospholipids and eicosanoid biosynthesis. PDGF is one compound of potential interest in this context. As further discussed below, it is produced by SMCs under certain circumstances. Moreover, studies on 3T3 fibroblasts have indicated that the stimulatory effect of PDGF on prostaglandin synthesis is related to increased phospo-lipase A₂ activity (Shier, 1980) as well as increased PGG/PGH and PGI₂ synthase activities (Habenicht et al. 1985; Goerig et al. 1988; Lin et al. 1989).

Radioimmunological determinations and analyses of arachidonic acid metabolism showed that the SMCs were able to synthesize PGI₂ (measured as the stable metabolite 6-keto-PGF_1α), PGE₂, PGF_2α and 15(S)HETE. A similar spectrum of eicosanoid biosynthesis has been reported in cultured SMCs stimulated by various agents (Alexander and Gimbrone, 1976; Huttner et al. 1977; Larrue et al. 1980, 1983, 1984; Coughlin et al. 1980, 1981; Bailey et al. 1983; Blay and Hollenberg, 1989; Lang and Vallotton, 1989; Pomerantz and Hajjar, 1989; Brinkman et al. 1990; Chaudhari et al. 1990). However, we were not able to detect any distinct increase in the release of eicosanoids into the medium following exposure of the cells to LTB₄. This could be due to the possibility that the activated synthesis of these substances was limited in time and added little to the basic production during the periods of conditioning of the medium. Alternatively, the eicosanoids that were formed as a result of the stimulation with LTB₄ remained cell-associated and were not released into the medium. A third possibility is that the cells synthesized compounds not yet identified via the prostaglandin endoperoxide synthase pathway. Treatment with indomethacin only partially blocked production of 15(S)HETE, in accordance with the notion that this substance may be produced via the prostaglandin endoperoxide synthase pathway as well as via the lipoxygenase and monooxygenase pathways (Needleman et al. 1986; Spector et al. 1988).

To confirm that the compounds mentioned above may act as mediators of the mitogenic effect of LTB₄, arachidonic acid, linoleic acid and various prostaglandin endoperoxide synthase products were tested in the thymidine-incorporation assay. The results demonstrated that arachidonic acid, PGI₂, PGE₂, PGF_2α. and 15(S)HETE all induced DNA replication in the SMCs under serum-free conditions and already at picomolar concentrations, whereas linoleic acid, 6-keto-PGF_1α and 5(S)HETE were inactive. On the other hand, PGI₂ and PGE₂ partially reduced the response of the cells to PDGF. These findings support the idea that the effect of LTB₄ on entrance into and progression through the cell cycle is mediated via one or more prostaglandin endoperoxide products. In agreement with earlier studies (for references, see the Introduction), they further point out that such products may have both positive and negative influences on cellular proliferation, partly depending on the presence of other growth regulators. This may be the result of a cross-talk between different signal transduction pathways, as part of the complex biochemical events that takes place during the G₁ phase of the cell cycle (Wickremasinghe, 1988; Pardee, 1989).

As briefly referred to above, endogenous PDGF is a possible mediator of the mitogenic effect of LTB₄. Earlier investigations have indicated that this growth factor is produced by SMCs in a phenotype- and growth statedependent manner (Thyberg et al. 1990b). Here, conditioned media from cultures stimulated with LTB₄, LTC₄ and LTD₄ were observed to induce DNA synthesis in quiescent cells. The release of stimulatory material by LTB₄-treated cultures was blocked by addition of indomethacin to the medium during the conditioning or by addition of antibodies against PDGF during the subsequent thymidine incorporation assay. On the other hand, no effect of these treatments on medium from cultures stimulated with LTC₄ and LTD₄ was noted. In a similar manner, analysis of the conditioned media for the presence of material competing with radioiodinated PDGF for binding to specific cell surface receptors pointed out that LTB₄ but not LTC₄ provoked release of PDGF-like molecules.

Summing up, these findings indicate that LTB₄ and the cysteinyl-containing leukotrienes LTC₄ and LTD₄ induce DNA replication in rat aortic SMCs by distinct mechanisms. The effect of LTB₄ requires prostaglandin endoperoxide synthase activity, suggesting the involvement of a prostaglandin or a related product. In support of this idea, the cells were found to produce PGI₂, PGE₂, PGF_2α and 15(S)HETE and these substances were all shown to stimulate cell cycle progression. In addition, analyses of conditioned media suggest that LTB₄ stimulates release of PDGF or a PDGF-like molecule from the cells. These observations will form a basis for further exploration of the signal transduction machinery responsible for leukotriene-induced mitogenesis in vascular SMCs.

The authors thank Karin Blomgren, Hélène Ax:son Johnson, Inger Forsberg, and Barbro Nasman-Glaser for expert technical assistance and Dr Carl-Henrik Heldin for kindly providing PDGF and antibodies against PDGF. Financial support was obtained from the Swedish Medical Research Council, the Swedish Cancer Society, the Swedish Heart Lung Foundation, and the King Gustaf V 80th Birthday Fund.