We recently showed that the farnesyltransferase inhibitor FTI-277 blocks interleukin 1β (IL-1β)-induced nitric oxide production in pulmonary vascular smooth muscle cells (SMC), whereas the geranylgeranyltransferase inhibitor GGTI-298 enhances this effect. Here we show that IL-1β and platelet-derived growth factor (PDGF) stimulate superoxide production by pulmonary vascular SMC and that this effect is blocked by both FTI-277 and GGTI-298, suggesting that farnesylated and geranylgeranylated proteins are required for superoxide production. We also show that FTI-277 and GGTI-298 block superoxide production stimulated by constitutively active mutant H-Ras. Furthermore, superoxide production by IL-1β, PDGF factor, and constitutively activated Ras is blocked by diphenyleneiodonium, implicating NAD(P)H oxidase as the generating enzyme. Given the role of oxidant radicals in vascular reactivity and injury, the action of both FTI-277 and GGTI-298 in suppressing superoxide generation by an inflammatory cytokine as well as by a potent smooth muscle mitogen may be therapeutically useful.
- platelet-derived growth factor
isoprenoid lipids are critically required to attach small G proteins to the plasma membrane (32). We have developed compounds that inhibit prenylation in a specific manner by targeting the prenyltransferase enzymes responsible for attaching the C15 lipid, farnesyl, or C20, geranylgeranyl (13, 14, 20, 27, 29). The farnesyltransferase-specific inhibitor FTI-277 blocks oncogenic Ras signaling (13), transformation, and tumor growth in nude mice (27), and the geranylgeranyltransferase I-specific inhibitor GGTI-298 blocks platelet-derived growth factor (PDGF)-dependent tyrosine phosphorylation of the PDGF receptor (16), induces the expression of the cyclin-dependent kinase inhibitor p21WAF (31), blocks cells in the G1 phase of the cell cycle (30), and induces apoptosis (17). Because unregulated cell growth and phenotypic change are also features of vascular pathologies such as atherosclerosis and restenosis (2) and pulmonary hypertension (26) in which smooth muscle is the principal cell type involved, we have recently evaluated the potential for prenyltransferase inhibitors to act as therapeutic agents against smooth muscle. We have shown that treatment of pulmonary vascular smooth muscle cells (SMC) with FTI-277 blocks interleukin IL-β (IL-1β)-stimulated expression of inducible nitric oxide synthase (NOS-2) (7). This is consistent with other work suggesting a role for farnesylated Ras in this transduction pathway (24). More intriguingly, however, GGTI-298 superinduces IL-1β-mediated NOS-2 expression and increases the levels of nitric oxide released (7). This suggests that a geranylgeranylated protein normally suppresses NOS-2 expression. This is an important finding because NOS-2-derived nitric oxide has been shown to play a key role in suppressing aberrant smooth muscle proliferation (22). Nitric oxide is not the only radical species known to play a key role in SMC physiology. Smooth muscle-derived superoxide radicals also have been implicated in altered vascular reactivity and hyperplastic growth (18, 19, 21). We recently showed that SMC possess a superoxide-generating system that can be activated by IL-1β (5). Indeed, IL-1β stimulates cogeneration of nitric oxide and superoxide, resulting in formation of the peroxynitrite radical (5). Production of superoxide is sensitive to diphenyliodonium (DPI) and quinacrine, suggesting that NADPH and/or NADH oxidases are involved. These two flavin enzymes are also responsible for superoxide production by rat aortic smooth muscle (8), whereas NADPH oxidase has been implicated in superoxide and hydrogen peroxide generation by endothelial cells (33). Little is known, however, of the regulation of superoxide production by resident vascular cells. In the neutrophil, on the other hand, the superoxide-generating enzyme NADPH oxidase is well characterized and shown to form a transmembrane complex consisting of flavocytochrome-b 558, phox proteins 47 and 67, and the small GTP binding protein Rac (1). Only when all components are assembled at the plasma membrane does superoxide production occur. Production can be suppressed by the inhibitor of cholesterol synthesis lovastatin (4). This is consistent with a role for a small GTP-binding protein because lovastatin inhibits the mevalonic acid pathway responsible for production of isoprenoid lipids. The aim of the current study then was to investigate the effects of our two prenyltransferase inhibitors, FTI-277 and GGTI-298, on the production of superoxide by pulmonary vascular SMC. The results show that not only IL-1β but also PDGF stimulates SMC to produce superoxide and that this process is inhibited by both FTI-277 and GGTI-298. Because PDGF is a prototype of Ras-dependent signaling (9), we have stably transfected SMC with a constitutively active mutant, H-Ras, which we have subsequently shown to produce elevated levels of superoxide in the absence of cytokine or growth factor stimulation. Here, too, superoxide production was also inhibited by FTI-277, consistent with the role of the transfected H-Ras, a farnesylated protein. The fact that GGTI-298 also inhibited production underlines its value as an alternative inhibitor of superoxide production in blocking the action of a geranylgeranylated protein downstream. However, it also emphasizes the potential of GGTI-298 to suppress the production of superoxide while enhancing that of NOS-2-derived nitric oxide.
MATERIALS AND METHODS
Isolation and culture of pulmonary vascular SMC. SMC were isolated from the distal vasculature of the lungs of adult male rats (Charles River Breeding Laboratories, Wilmington, MA) as described previously (11). Cells were grown in Dulbecco's modified Eagle's medium (DMEM; JRH Bioscience, Lenexa, KS) supplemented with 20% heat-inactivated low-endotoxin fetal bovine serum (HyClone, Logan, UT), penicillin (100 U/ml), streptomycin (100 μg/ml; Sigma, St. Louis, MO), and HEPES (United States Biochemical, Cleveland, OH). Cells were grown at 37°C in 95% air-5% CO2 and passaged by harvesting with trypsin-EDTA and seeding into 75-cm2 culture flasks. Cells were used for experiments between passages 3 and8.
Pretreatment of cells with Ras inhibitors. SMC were plated in growth medium at a density of 2 × 105 cells/well in 24-well plates and incubated overnight. Cells were transferred to serum-deprived basal medium (low glucose DMEM, 0.1% BSA, penicillin, and streptomycin) and treated with FTI-277 (5 μM) or GGTI-298 (10 μM) on each of 2 successive days.
Measurement of superoxide production. SMC at a density of 2 × 105 cells/well in 24-well plates were rinsed three times in PBS, which was then replaced by reaction buffer solution (in mM: 2 glucose, 1 CaCl2, 1.3 MgCl2, 4 KCl, 100 NaCl, and 10 phosphate buffer, pH 7.4). Ferricytochrome c(Sigma) was added at a final concentration of 70 μM in the presence or absence of superoxide dismutase (SOD, Sigma) at 40 μg/ml. When appropriate, PDGF-BB (GIBCO BRL) or IL-1β (recombinant human; National Cancer Institute, Bethesda, MD) was added to a total volume of 1 ml/well, and the cells were incubated for 2 h. The reaction was terminated by adding N-ethylmaleimide (1 mM). The absorbance of the supernatant was read spectrophotometrically at 550 nm. The difference in absorbance in the presence or absence of SOD was determined to measure release. is expressed as nanomoles per microgram of protein (15). To measure the superoxide-scavenging activity of FTI-277 and GGTI-298, we used the microplate assay method reported by Ewing and Janero (6). For the assay, 25-ml samples of compounds were pipetted into a microtiter well containing 200 ml of freshly prepared 0.1 mM EDTA, 62 mM nitro blue tetrazolium (NBT), and 98 mM NADH in 50 mM phosphate buffer, pH 7.4. The reaction was initiated by the addition of 25 ml of freshly prepared phenazine methosulfate (33 mM) in 50 mM phosphate buffer, pH 7.4, containing 0.1 mM EDTA. The absorbance at 600 nm was measured at the start of the reaction and again after 5 min of incubation at room temperature using a Dynatech MRX microplate reader.
Ras/Rap1A immunoblotting. Cells were plated in T25 flasks at a density similar to that used for the superoxide assay and treated with FTI-277 and GGTI-298 as previously described. After incubation, medium was removed and cells were washed with ice-cold PBS (pH 7.5) and then lysed in a small volume of lysis buffer containing 50 mM HEPES (pH 7.5), 100 mM NaCl, 5 mM MgCl2, 1% Triton X-100, 10% glycerol, 2 mM NaVO4, 10 μg/ml soybean trypsin inhibitor, 10 μg/ml aprotinin, 25 μg/ml leupeptin, 2 mM phenylmethylsulfonyl fluoride, and 6.4 mg/ml Sigma-104R phosphatase substrate (Sigma). Whole cell lysate was loaded on a 12.5% SDS-polyacrylamide gel and separated electrophoretically. Eluted proteins were transferred overnight to nitrocellulose membrane at 100 mA and then blocked in a solution of PBS (pH 7.5) containing 0.1% Tween 20 and 5% non-fat dried milk for 1 h. Membranes were exposed to primary antibodies against Ras (Y13–238, 50 μg/ml) or Rap1A (0.1 μg/ml; Santa Cruz Biotechnology, Santa Cruz, CA). Results were visualized using horseradish peroxidase-conjugated IgG secondary antibodies at a concentration of 1:1,000, chemiluminescence (ECL; Amersham, Buckinghamshire, UK), and exposure to X-ray film.
Stable transfection of SMC. We used a pSV2neo plasmid containing a 6.6-kb mutant Val 12 H-ras genomic fragment from human bladder carcinoma (23) purchased from American Type Culture Collection. Subconfluent monolayers of cells were transfected using cationic liposomes (Lipofectin, GIBCO BRL) at a 5:1 ratio of liposomes to plasmid DNA for 5 h. Control cells were transfected with the pSV2neo empty plasmid. After transfection, cells were washed and transferred to growth medium. Because the vector includes the Neo gene, we selected nonclonally for positively transfected cells by incubation in G-418 (150 μg/ml) over a 3-wk period. This procedure was performed twice on separate cultures.
Statistical analysis. Determinations of superoxide levels were made in triplicate. Superoxide data are given as the means ± SD of at least three determinations and were analyzed by ANOVA with Fisher's correction. Differences were considered significant at P < 0.05.
We evaluated the effects of inhibition of protein prenylation on the formation of superoxide, an oxidant radical that is cogenerated with nitric oxide in response to IL-1β (5). Subconfluent monolayers of SMC were pretreated with FTI-277 or GGTI-298 on each of 2 successive days before IL-1β. Cellular lysates were analyzed by Western blots using antibodies against Ras and Rap1A as described in materials and methods. Inhibition of processing was detected by a band shift from the fully processed lower band to an upper unprocessed band (Fig. 1). In cells treated with vehicle, only the fully processed forms of Ras and Rap1A were evident. FTI-277 inhibited the processing of Ras but had little effect on the processing of the exclusively geranylgeranylated Rap1A. Conversely, GGTI-298 inhibited the processing of Rap1A but had no effect on the processing of Ras (Fig. 1). To determine the effects of the inhibitors on superoxide production, the cell- conditioned medium collected over 2 h from the same experimental protocols previously described was analyzed for SOD inhibitable-superoxide production by a cytochromec reduction assay. Control cells treated with vehicle released superoxide levels of 0.09 nmol/μg protein, and superoxide production was stimulated 3.5-fold by IL-1β. However, both FTI-277 and GGTI-298 significantly reduced IL-1β-stimulated superoxide levels to baseline (Fig. 2). In a separate, cell-free assay monitoring NBT reduction by superoxide produced by an aerobic mixture of NADH and phenazine methosulfate, we showed that neither peptidomimetic reduced superoxide levels to this degree: at the concentrations applied to the cells, FTI-277 reduced superoxide-dependent NBT reduction by 8% and GGTI-298 by 16% (data not shown).
We next determined whether this effect of prenylation inhibitors on superoxide formation was specific to IL-1β. We therefore treated SMC with a known smooth muscle mitogen, PDGF (0–10 ng/ml), and determined the levels of superoxide in the culture medium as described in materials and methods. Figure3 A shows that PDGF-BB caused a concentration-dependent increase in SMC-derived superoxide anion during a 2-h incubation. Pretreatment of cells with FTI-277 and GGTI-298 blocked production of superoxide stimulated by PDGF (Fig. 3 B). Thus FTI-277 and GGTI-298 were able to block superoxide production by both a cytokine (IL-1β) and a mitogen (PDGF).
The effect of FTI-277 and GGTI-298 implicated prenylated proteins in the transduction pathway mediating superoxide production. We considered Ras a likely candidate because it is prenylated and it is the prototypic mediator of PDGF signaling. To test its involvement in superoxide production, we stably transfected SMC with a GTP-locked, constitutively active form of Ras (Val 12 H-Ras mutant). In cells transfected with vector alone (pSV2neo), only one form of Ras was expressed (Fig. 4 A, lane 1). In cells transfected with the mutant H-Ras, an additional, more slowly migrating form of Ras was expressed more abundantly than the endogenous form (Fig. 4 A, lane 4). FTI-277 was again applied at 5 μM, a concentration that we have previously found completely inhibits mitogen-activated protein kinase activation in H-Ras-transformed NIH/3T3 cells. At this concentration, FTI-277 inhibited lane 4 processing of both endogenous wild-type Ras and mutant H-Ras [comparelanes 1 and 2 for endogenous wild-type Ras andlanes 4 and 5 for endogenous (wild-type) as well as exogenous mutant H-Ras]. GGTI-298 had little effect on processing of the endogenous or mutant forms of Ras. In the absence of mitogen or cytokine stimulation, the mutant H-Ras transfectants produced levels of superoxide similar to those of wild-type cells stimulated with PDGF-BB or IL-1β and significantly higher than the levels of superoxide produced by cells transfected with the pSV2neo empty plasmid vector (Fig. 4 B). Production of superoxide, however, was suppressed by pretreatment with FTI-277, consistent with a role for the exclusively farnesylated H-Ras mutant. GGTI-298, however, also suppressed production, suggesting the involvement of a geranylgeranylated protein downstream of H-Ras.
Because stably transfected cells exhibit an altered phenotype and potentially increased metabolism that can be manifested in elevated superoxide production, we wished to confirm that the Ras-dependent superoxide was produced by the same enzyme as in wild-type cells. Accordingly, we measured superoxide production by PDGF-stimulated wild-type cells (Fig. 5 A) and by the mutant H-Ras-transfected cells (Fig. 5 B) in the presence of DPI or quinacrine, the two inhibitors previously shown to suppress IL-1β-stimulated superoxide production in wild-type cells (4). Both inhibitors significantly reduced the superoxide production to baseline. In the cell-free NBT reduction assay, DPI had no nonspecific scavenging effect; quinacrine, however, had a pronounced effect, reducing superoxide-dependent NBT reduction by more than 80% (data not shown). The results with DPI are therefore a more reliable indication of flavin enzyme inhibition and suggest that NAD(P)H oxidase(s) was responsible for the superoxide production seen.
This paper has shown that production of superoxide by pulmonary vascular SMC is stimulated by IL-1β, PDGF, and constitutively active Ras and is blocked by prenyltransferase inhibitors. In all cases, the superoxide appears to originate from DPI-sensitive enzyme(s) likely to be similar to the NAD(P)H oxidase noted in rat aortic SMC (8).
In a recent study we treated SMC similarly obtained from the lung with our prenyltransferase inhibitors at the same concentrations used here and showed that GGTI-298, but not FTI-277, in serum-free medium induced apoptosis in 30% of the cells. Whereas serum inhibited this effect, GGTI-298 nevertheless inhibited serum-dependent DNA synthesis and cell proliferation; FTI-277, on the other hand, did not (25). Thus GGTI-298 exclusively inhibits growth and induces apoptosis, whereas both compounds suppress superoxide production.
PDGF has been shown to stimulate production of hydrogen peroxide in vascular SMC, as detected intracellularly by dichlorofluorescein-enhanced fluorescence microscopy (28). In SMC from the rat aorta, angiotensin II stimulates superoxide production (8). The fact that PDGF stimulates superoxide suggests that production of reactive oxygen species by smooth muscle is not solely dependent on inflammatory cytokines but is likely to occur under normal conditions.
FTI-277 inhibited processing of Ras but not of the exclusively geranylgeranylated Rap1A, indicating its selective action on farnesylation in SMC. GGTI-298, on the other hand, showed selective inhibition of Rap1A processing. Both compounds inhibited superoxide production, implicating farnesylated and geranylgeranylated proteins in the signaling pathway. To determine whether this also applied under conditions of receptor-independent activation by Ras, we transfected SMC with the GTP-locked, constitutively active H-Ras. Stably transfected cells were found to constitutively produce levels of superoxide that were appreciably elevated over levels released from cells transfected with the pSV2neo vector alone. This enhanced level of production disappeared when cells were treated with FTI-277. Because FTI-277 also inhibited the processing of the mutant Ras, we conclude that the enhanced superoxide production was attributable to H-Ras. GGTI-298 had no effect on the processing of the mutant Ras. Nevertheless, it inhibited superoxide production as effectively as FTI-277. This suggests that a geranylgeranylated, GGTI-298-sensitive protein lies downstream of Ras in the pathway regulating superoxide production. In NIH/3T3 cells transfected with the Val 12 mutant Ras, enhanced superoxide production was blocked by cotransfecting dominant negative Rac 1 (10). Because Rac is a geranylgeranylated protein, it is one of a number of candidate proteins targeted by GGTI-298 in our SMC. In neutrophils, the NADPH oxidase consists of multiple units, including the flavocytochrome (cytochrome b 558) itself, which needs to be complexed with Rac, a geranylgeranylated small G protein (3).
Neutrophil-derived superoxide is utilized for antimicrobial toxicity, whereas the superoxide generated by smooth muscle appears to have little effect on cellular viability (5). Instead, smooth muscle-derived superoxide has been implicated in vascular reactivity (18, 21) and growth (8, 12, 19). However, the fact that FTI-277 has no effect on DNA synthesis in these cells despite its inhibition of superoxide production suggests that superoxide is not a mandatory component of growth signaling.
The fact that PDGF-activated superoxide production is blocked by the prenyltransferase inhibitors is consistent with the role of Ras demonstrated in this study and with the prototypic role of this protein in mediating signaling from tyrosine kinase receptors. IL-1β, on the other hand, binds a nontyrosine kinase receptor. Nevertheless, its effect on superoxide production was also blocked by both inhibitors. We recently showed that IL-1β stimulated induction of NOS-2 and that subsequent production of nitric oxide by vascular SMC was blocked by farnesyltransferase inhibitors and suggested Ras as a likely mediator (7). GGTI-298, however, superinduced NOS-2 in IL-1β-stimulated cells (7), suggesting that geranylgeranylated proteins downstream of the IL-1β receptor exert positive effects on some downstream pathways and negative effects on others.
We thank Sothi Tran and Jennifer L. Litz for technical assistance.
Address for reprint requests and other correspondence: P. Davies, Inflammatory Diseases Research, DuPont Pharmaceutical, Wilmington, DE 19880–0400 (E-mail:).
This work was supported by a fellowship to A. Boota and a grant-in-aid to P. Davies from the American Heart Association, Pennsylvania Affiliate, and National Cancer Institute Grant CA- 67771 to S. M. Sebti.
Present address of S. Sebti: H. Lee Moffitt Cancer Center, Dept. of Biochemistry and Molecular Biology, Univ. of South Florida, 12902 Magnolia Dr., Tampa, FL 33612.
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 © 2000 the American Physiological Society