Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Lovastatin is a well known antihypercholesterolemic agent, which inhibits 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, the rate-limiting enzyme of cholesterol biosynthesis. HMG-CoA reductase synthesizes mevalonate from HMG-CoA; subsequently, mevalonate is either converted into cholesterol through multiple enzymatic steps or used in the production of isoprenoids such as ubiquinone, dolichol, isopentenyl-t-RNA, and prenylated proteins (Goldstein and Brown, 1990). It has been reported that noncholesterol metabolites converted from mevalonate were required for cell proliferation (Quesney-Huneeus et al., 1983), suggesting that mevalonate may play a general role in maintaining cellular functions. Recently, it has been reported that isopentenyl adenine could partially substitute for mevalonate in initiating DNA replication (Siperstein, 1995) and that isoprenoids were required for the protein prenylation (Zhang and Casey, 1996). Prenylation at or near the carboxyl cysteine residue is an important post-translational modification required for the full activity or activation of various membrane-bound proteins mediating signal transduction. Indeed, lovastatin has been reported to inhibit signal transduction induced by various factors such as insulin (Xu et al., 1996), epidermal growth factor, and insulin-like growth factor I (Vincent et al., 1991). These reports suggest that an activity level of HMG-CoA reductase may affect various cellular functions such as DNA synthesis, cell proliferation, and signal transduction.

The use of HMG-CoA reductase inhibitors in various diseases is under investigation. Négre-Aminou et al. (1997) reported that HMG-CoA reductase inhibitors suppressed platelet-derived growth factor- or basic fibroblast growth factor-induced DNA synthesis in smooth muscle cells prepared from human arteries, which might be responsible for the preventive effect of lovastatin on the restenosis after angioplasty in the atherosclerotic lesions. A similar effect of lovastatin was reported in mesangial cells of the kidney (O'Donnell et al., 1993), which might be beneficial in preventing progressive glomerular disease. In addition, lovastatin induced cell death in tumor cells, such as malignant mesothelioma cells (Rubins et al., 1998), promyelocytic HL-60 cells (Pérez-Sala and Mollinedo, 1994), and human malignant glioma cells (Jones et al., 1994). Inhibition of cholesterol biosynthesis also might be beneficial in suppression of the progression of neurodegenerative diseases. Bochelen et al. (1995) reported that oxysterols, potent inhibitors of cholesterol biosynthesis, reduced reactive astrogliosis in the injured rat brain. In addition, Pahan et al. (1997) reported that lovastatin inhibited lipopolysaccharides-induced nuclear factor-B activation, cytokine gene expression, and nitric oxide production in astrocytes. Although these reports suggested the importance of the cholesterol biosynthetic pathway on maintaining cellular functions, it was still unclear whether such effects of HMG-CoA reductase inhibitors were due to the suppression of cell growth or induction of cell death. It might be also possible that the effects of HMG-CoA reductase inhibitors are cell type specific. In various tumor cells, complete loss or marked impairment of feedback regulation for cholesterol biosynthesis has been reported (Siperstein, 1995). In this study, we demonstrate that the inhibition of HMG-CoA reductase can induce both suppression of cell growth and induction of cell death in C6 glial cells depending on the intracellular levels of mevalonate. C6 glial cells, cloned from a rat glial tumor induced by N-nitrosomethylurea, exhibit the normal feedback regulation of cholesterol biosynthesis (Volpe and Hennessy, 1977). Our results also indicate that the type of cell death induced by inhibition of HMG-CoA reductase is apoptosis that requires both active RNA and protein synthesis.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture and Media. C6 glial cells were kindly provided by Dr. Y. S. Kim (Seoul National University College of Medicine, Seoul, Korea). Cells were maintained in Dulbecco's modified Eagle's medium containing 10% (v/v) FBS, nonessential amino acids, penicillin (final concentration, 50 units/liter), and streptomycin (final concentration, 50 µg/liter) at 37°C (all from Sigma Chemical Co., St. Louis, MO). Lovastatin was kindly provided by Choong Wae Pharmaceutical Co. (Seoul, Korea) and was prepared as described by Kita et al. (1980). Cycloheximide (Sigma Chemical Co.) was dissolved in PBS, pH 7.4, to a concentration of 10 mg/ml. Actinomycin D (Sigma Chemical Co.) was dissolved in ethanol to a concentration of 5 mg/ml. Mevalonic acid lactone was purchased from Sigma Chemical Co. and dissolved in PBS, pH 7.4, to a concentration of 100 mM.

RNA Isolation. The single-step method of RNA isolation by acid guanidinium thiocyanate/phenol/chloroform extraction, as described by Chomczynski and Sacchi (1987), was used. C6 glial cells were lysed in a solution containing 4 M guanidinium thiocyanate, 2.5 mM sodium citrate (pH 7.0), 0.5% sodium lauryl sarcosyl, and 0.1 M 2-mercaptoethanol. The cell lysate was made 0.2 M with respect to sodium acetate, pH 4.0. After adding a 1/5 volume of chloroform, the total RNA was extracted with 1 volume of water-saturated acid phenol, pH 5.0. After precipitation with 2-propanol, pelleted RNA was washed with 80% ethanol, vacuum dried, and resuspended in 0.05 M Tris·HCl, pH 7.5, 0.01 M MgCl₂, and 6 µg of RNase-free DNase and 10 units of RNasin (both from Promega, Madison, WI). The samples were incubated at 37°C for 45 min, and total RNA was extracted by using acid phenol as described. RNA was precipitated with ethanol, dried under a vacuum, and resuspended in diethyl pyrocarbonate-treated water. The samples were stored at 70°C.

Reverse Transcription-Polymerase Chain Reaction (PCR) Procedure. cDNA synthesis was performed using 0.75 µg of total RNA and the Perkin-Elmer first-strand cDNA synthesis kit (Roche Molecular Systems Inc., Branchburg, NJ). An HMG-CoA reductase 5' primer corresponding to bases 967 to 986 (5'-GTGGTTACCCTGAGCTTAGC-3') and a 3' primer corresponding to bases 1409 to 1428 (5'-CGGGATGTGCTTGGCATTGA-3') based on the sequence of Syrian hamster HMG-CoA reductase cDNA were used for HMG-CoA reductase PCR. A 5' primer corresponding to bases 102 to 122 (5'-CGTGGGCCGCCCTAGGCACCA-3') and a 3' primer corresponding to bases 323 to 344 (5'-TTGGCCTTAGGGTTCAGGGGGG-3') were used for -actin PCR, which were designed based on the sequence of rat -actin gene. PCR samples were amplified using the Perkin Elmer PCR kit (Roche Molecular Systems Inc.) for 25 cycles. In each cycle, DNA strands were melted at 94°C for 60 s, annealed at 60°C for 60 s, and extended at 72°C for 60 s. The PCR sample was analyzed on 1.5% agarose gel retaining ethidium bromide at a concentration of 0.5 µg/ml.

Cell Proliferation Assay. C6 glial cells were seeded at a density of 5 × 10³ cells/well in 100 µl of culture medium in a 96-well plate (Falcon; Becton Dickinson, Franklin Lakes, NJ). Lovastatin and/or mevalonic acid lactone was added to the medium at the appropriate concentrations, and cells were cultured for the indicated time periods. Cell proliferation kit II (Boehringer-Mannheim Biochemica, Mannheim, Germany) was used to measure the number of cells. This is a colorimetric assay based on tetrazolium salt [XTT; sodium 3'-(1-(phenyl-amino-carbonyl)-3,4-tetrazolium)-bis(4-methoxy-6-nitro)-benzene sulfonic acid hydrate]. The tetrazolium salt XTT is cleaved to formazan by the succinate tetrazolium reductase that is active only in the mitochondria of the viable cell (Jones et al., 1994). After adding tetrazolium salt XTT to the medium, the levels of formazan in the medium were measured at 490 nm by SpectraMAX340 spectrophotometric microplate reader (Molecular Devices, Menlo Park, CA). The numbers of cells were expressed as mean ± S.E., and statistical analysis of difference was carried out by two-tailed Student's t test. All values represent at least three independent experiments performed in triplicate.

Lactate Dehydrogenase Assay. C6 glial cells were seeded at a density of 5 × 10³ cells/well in 100 µl of culture medium in a 96-well plate (Falcon; Becton Dickinson). Lovastatin and/or mevalonic acid lactone were added to the medium at the appropriate concentrations. The level of lactate dehydrogenase (LDH) activity in the medium was measured by using the LDH-LQ kit (Asan Pharmaceutical Co., Seoul, Korea). This is a colorimetric assay based on nitrotetrazolium blue. The nitrotetrazolium blue is reduced to the diformazen in the presence of NADH and 1-methoxy phenazine methosulfate. LDH catalyzes the reaction between lactate and NAD to produce pyruvate and NADH. At time points after treatment, each 50 µl was removed from the culture medium, and the level of formazan after reaction was measured at 570 nm by using an Ultraspec 3000 spectrophotometer (Pharmacia Biotech, Cambridge, UK). The levels of LDH activity were expressed as mean ± S.E., and statistical analysis of difference was carried out by two-tailed Student's t test. All values represent at least three independent experiments.

DNA Fragmentation Assay. To examine the fragmentation of DNA, a modified method of Lyons et al. (1992) was used. Cells were harvested by trypsinization and centrifuged at 800g for 10 min. After resuspension of cells in 250 µl of TE buffer containing 10 mM Tris·HCl, pH 8.0, and 1 mM EDTA, 250 µl of lysis buffer was added. Lysis buffer contains 5 mM Tris, pH 8.0, 20 mM EDTA, and 0.5% Triton X-100. The sample was centrifuged at 13,000 rpm for 15 min in the microcentrifuge (Eppendorf, Hamberg, Germany), and the DNA in the supernatant was ethanol precipitated overnight. The sample was digested with RNase (Boehringer-Mannheim) at a final concentration of 1.5 mg/ml for 1 h at 37°C and with proteinase K (Boehringer-Mannheim) at a final concentration of 1 mg/ml for additional 2 h at 37°C. The sample was analyzed on an 1.5% agarose gel.

Comet Assay. Comet assay was done as described by Green et al. (1996). C6 glial cells were harvested by trypsinization and were diluted to the density of 3 × 10⁶ cells/ml. Each 85 µl of sample was mixed with 85 µl of 1% low-melting agarose solution at 37°C. Then 75 µl of the mixture was transferred onto the 0.5% agarose bed set on the slide. After covering the sample mixture with 75 µl of 0.5% agarose solution, the slide was placed in a staining rack containing cold lysis buffer and stored at 4°C for 2 h. Lysis buffer was composed of 2.5 M NaCl, 100 mM EDTA, 10 mM Tris base, pH 10.0, 1% (v/v) Triton X-100, and 10% (v/v) dimethyl sulfoxide. After placing the slide in the gel electrophoresis box, the sample was electrophoresed at 20 V for 24 min. The electrophoresis buffer was composed of 300 mM NaOH and 1 mM EDTA and stored at 4°C. After electrophoresis, the slide was rinsed dropwise with 1 ml of neutralization buffer containing 0.4 M Tris base, pH 7.5. The DNA in the sample was stained by adding 60 µl of 0.002% ethidium bromide solution onto the slide, and the comet representing DNA fragmentation was examined immediately in an inverted fluorescence microscope with ultraviolet excitation at 340 to 380 nm.

Fluorescence Microscopic Assay. Fluorescence microscopic assay was used as described by Hoorens et al. (1996) to determine cell viability. C6 glial cells were cultured for 24 or 48 h, and the percentage of viable and dead cells were estimated after staining cells with Hoechst 33342 (HO 342; Sigma Chemical Co.) and propidium iodide (PI; Sigma Chemical Co.). HO 342 freely enters cells with intact membrane as well as cells with damaged membrane and stains DNA blue. PI, a highly polar dye that is impermeable to cells with intact membrane, stains DNA red. HO 342 was added to the medium to a final concentration of 5 µg/ml, and cells were incubated at 37°C for 15 min. After adding PI to the medium to a final concentration of 5 µg/ml, the cells were examined immediately using inverted fluorescence microscope with ultraviolet excitation at 340 to 380 nm. Viable or dead cells were identified by nuclei with blue (HO 342) or red (HO 342 plus PI) fluorescence. When necessary, floating cells in the medium and cells harvested by trypsinization were collected together and centrifuged at 800g for 10 min. After resuspending cells in the medium containing HO 342 and PI, the sample was mounted on the microscopic slide and examined. In each condition and experiment, at least 100 cells were counted. The percentage of viable and dead cells was expressed as mean ± S.E., and all values represent at least three independent experiments.

Electron Microscopic Assay. After trypsinization, suspended cells were centrifuged at 800g and fixed in Karnovsky's solution containing 0.1 M cacodylate, pH 7.4, 2% glutaraldehyde, 2% paraformaldehyde, and 0.5% CaCl₂. Samples were postfixed in cacodylate-buffered osmium tetroxide (1.33%), dehydrated sequentially using 50 to 100% ethanol, immersed in propylene oxide for 10 min, and embedded in EPON mixture. Thin sections were stained with uranyl acetate and lead citrate and then examined using a Philips CM-10 electron microscope (Philips Electron Optics, Eindhoven, Holand).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Lovastatin Increases HMG-CoA Reductase mRNA Levels in C6 Glial Cells. Previous studies on various cell systems such as hepatocyte, fibroblasts (Goldstein and Brown, 1990), and kidney cells (Choi et al., 1993) have demonstrated that lovastatin, an HMG-CoA reductase inhibitor, increases HMG-CoA reductase mRNA levels. This increase is likely due to a reduction of intracellular sterols and cholesterol, which mediates the negative feedback regulation of HMG-CoA reductase gene expression (Sakai et al., 1998). In this study, the levels of HMG-CoA reductase mRNA increased when C6 glial cells were treated with 10 µM lovastatin for 24 h, which indicated the presence of negative feedback regulation of HMG-CoA reductase gene expression in C6 glial cells (Fig. 1). There was no significant increase in HMG-CoA reductase mRNA levels when cells were cultured with 1 µM lovastatin (Fig. 1). As a control, the -actin mRNA levels were also measured. -Actin is a constitutively expressed protein, and its mRNA level had been shown to remain relatively unchanged under similar conditions.

View larger version (39K): [in this window] [in a new window]   Fig. 1.   Lovastatin increased HMG-CoA reductase mRNA levels in C6 glial cells. Cells were cultured in media containing lovastatin at the indicated concentrations for 24 h and the levels of HMG-CoA reductase mRNA were determined. M, 100-bp PCR marker; HMGR, HMG-CoA reductase mRNA; -actin, -actin mRNA.

Lovastatin Inhibits Proliferation and Induces Cell Death in C6 Glial Cells. It was previously demonstrated that lovastatin inhibited cell proliferation in A172 cells, U87-MG cells, human malignant glioma cells (Jones et al., 1994), and human promyelocytic HL-60 cells (Pérez-Sala and Mollinedo, 1994). To observe whether lovastatin has similar effects on C6 glial cells, cells were cultured in media containing various concentrations of lovastatin, and cell numbers were determined at 24 or 48 h after lovastatin treatment. As shown in Fig. 2, lovastatin at concentrations above 10 µM reduced the number of cells at 24 h. At 48 h after treatment, lovastatin reduced the number of cells in a dose-response manner from 1 to 25 µM. To exclude the possibility that lovastatin exerted direct cytotoxic effects on the C6 cells, cells were simultaneously treated with exogenous mevalonate, the product of HMG-CoA reductase. As shown in Fig. 2, mevalonate at a concentration of 300 µM prevented the cytotoxic effects of lovastatin completely. When cells were cultured in the medium containing 25 µM lovastatin and 100 µM mevalonate, the number of cells at 48 h was almost the same as the number at 24 h (Fig. 2). In the presence of 25 µM lovastatin and 30 µM mevalonate, the number of cells at 48 h decreased markedly compared with the number at 24 h (Fig. 2). These results suggest that mevalonate at a concentration of 100 µM can prevent lovastatin-induced cell death, whereas it cannot restore the proliferation of C6 glial cells after lovastatin treatment.

View larger version (31K): [in this window] [in a new window]   Fig. 2.   Lovastatin-induced inhibition of C6 glial cell proliferation was prevented by mevalonate. A and B, cells were cultured in media containing lovastatin at the indicated concentrations. C and D, cells were cultured in media containing 25 µM lovastatin and mevalonate at the indicated concentrations. At 24 (A and C) or 48 (B and D) h after treatment, cell numbers were determined. **p < .01, ***p < .001.

Lovastatin Inhibits Cell Proliferation and Induces Cell Death. Lovastatin might decrease the number of C6 glial cells via two different mechanisms. One was the inhibition of cell proliferation, and the other was the induction of cell death. It has been reported that lovastatin arrested cell cycle at early G₁ phase (Hengst and Reed, 1996). The interruption of cell cycle transition may induce cell death as reported by Colombel et al. (1992). To determine whether lovastatin induced cell death, the level of LDH activity in the medium was measured. As shown in Fig. 3, the level of LDH activity increased slightly but not significantly at 24 h after 25 µM lovastatin treatment, whereas it increased markedly at 48 h. These results suggest that cell membrane remained intact until 24 h after lovastatin treatment. Figure 3 also demonstrates that the cytotoxic effects of lovastatin were prevented by 100 µM mevalonate.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Mevalonate prevented the lovastatin-induced increase of LDH activity in the medium. C6 glial cells were cultured in media containing 25 µM lovastatin and mevalonate at the indicated concentrations. The level of LDH activity was measured at 24 (A) or 48 (B) h after treatment. **p < .01, ***p < .001.

Lovastatin Induces Chromatin Condensation in C6 Glial Cells. Cell death occurs by two processes: necrosis and apoptosis (Columbano, 1995). In the process of apoptosis, chromatin condensations at the nuclear edge or apoptotic bodies are frequently observed (Dini et al., 1996). The fluorescence microscopic assay by HO 342 and PI showed nuclear fragments, which seemed to contain chromatin clumps, in lovastatin-treated C6 glial cells at 36 h after treatment (Fig. 4). In cells treated with lovastatin for 24 h, nuclear fragments could not be observed by fluorescence microscopy, but electron microscopic assay showed the condensed chromatin protruding from the nuclear membrane (Fig. 4), which indicated that lovastatin-treated cells were at the early stage of chromatin condensation (Dini et al., 1996). These results indicate that the type of lovastatin-induced cell death was apoptosis, which had already begun at 24 h after lovastatin treatment.

View larger version (99K): [in this window] [in a new window]   Fig. 4.   Lovastatin induced chromatin condensation and nuclear fragmentation in C6 glial cells. A and B, at 24 h after lovastatin treatment, cellular structures were examined by electron microscopy. C-F, at 36 h after lovastatin treatment, cells were stained with HO 342 and PI and examined by fluorescence microscopy. A, C, and D, untreated. B, E, and F, 25 µM lovastatin. Original magnification: A, 13,220×; B, 17,800×; C and E, 200×; and D and F, 400×.

Lovastatin Induces Internucleosomal DNA Fragmentation in C6 Glial Cells. Among the characteristics of the apoptosis, the most prominent biochemical feature is the formation of DNA fragments of approximately 180 bp or DNA laddering (Majno and Joris, 1995). When C6 glial cells were cultured in media containing various concentrations of lovastatin for 24 h, DNA laddering was observed at concentrations above 10 µM (Fig. 5). The data from the comet assay showed that the number of cells containing fragmented DNA increased at 24 h after 25 µM lovastatin treatment (Fig. 6). Figure 6 also shows that mevalonate at concentrations above 100 µM blocked lovastatin-induced DNA fragmentation completely, whereas mevalonate at a concentration of 30 µM could not. This suggests that lovastatin-induced DNA fragmentation was due to the depletion of intracellular mevalonate.

View larger version (111K): [in this window] [in a new window]   Fig. 5.   Lovastatin induced internucleosomal DNA fragmentation in C6 glial cells. Cells were cultured in media containing lovastatin at the indicated concentrations. At 24 h, DNA fragmentation was examined. M, 100-bp marker.

View larger version (32K): [in this window] [in a new window]   Fig. 6.   Mevalonate prevented lovastatin-induced DNA fragmentation. C6 glial cells were cultured in media containing 25 µM lovastatin and mevalonate at the indicated concentrations. DNA fragmentation was examined after 24 h. A, agarose gel electrophoresis. B, comet assay.

Cycloheximide and Actinomycin D Block Lovastatin-Induced DNA Fragmentation in C6 Glial Cells. To determine whether lovastatin-induced DNA fragmentation required active mRNA or protein synthesis, the effects of cycloheximide, a translational inhibitor, and actinomycin D, a transcriptional inhibitor, on the lovastatin-induced apoptosis of C6 glial cells were examined. When cells were cultured in media containing either cycloheximide at a concentration of 0.5 µg/ml or actinomycin D at a concentration of 0.1 µg/ml, lovastatin-induced DNA fragmentation was prevented (Fig. 7).

View larger version (75K): [in this window] [in a new window]   Fig. 7.   A, time course of DNA fragmentation after lovastatin treatment in C6 glial cells. Cells were cultured in media containing 25 µM lovastatin. At the indicated time points, cells were harvested and DNA fragmentation was examined by agarose gel electrophoresis. B and C, cycloheximide or actinomycin D prevented lovastatin-induced DNA fragmentation. C6 glial cells were cultured in media containing 25 µM lovastatin and either cycloheximide (B) or actinomycin D (C) at the indicated concentrations. DNA fragmentation was examined at 24 h. M, 100-bp marker.

Point of Commitment to Apoptosis after Lovastatin Treatment in C6 Glial Cells. To determine the time course of lovastatin-induced DNA fragmentation, C6 glial cells were cultured in the medium containing 25 µM lovastatin, and DNA laddering was examined at various time intervals. As shown in Fig. 7, DNA laddering was observed at 12 h after lovastatin treatment and became more evident gradually until 24 h. To determine the point of commitment to apoptosis as the latest time at which lovastatin-treated cells could be rescued, mevalonate was added to the medium to a final concentration of 300 µM at different time points after lovastatin treatment. In this experiment, DNA laddering was examined at 24 h after lovastatin treatment. Mevalonate provided protection when added by 12 h (Fig. 8). When mevalonate was added to the medium at 16 h, the level of DNA fragmentation was apparently reduced (Fig. 8), which suggests the preventive role of mevalonate on the progression of apoptosis. Cycloheximide provided protection when added up to 8 h (Fig. 8). These data (Figs. 7C and 8) indicate that protein and RNA synthesis are required for the activation of lovastatin-induced DNA fragmentation and that lovastatin induces genuine apoptosis in C6 glial cells.

View larger version (57K): [in this window] [in a new window]   Fig. 8.   RNA and protein synthesis are required for lovastatin-induced apoptosis in C6 glial cells. Cells were cultured in media containing 25 µM lovastatin. A, mevalonate was added to the medium to a final concentration of 300 µM at the indicated time point after lovastatin treatment. DNA fragmentation was examined at 24 h. B, cycloheximide was added to the medium to a final concentration of 0.5 µg/ml at the indicated time point after lovastatin treatment. M, 100-bp marker; CON, no treatment.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Our results indicate that lovastatin can inhibit proliferation and induce apoptosis in C6 glial cells. In this study, the increase in cell number was retarded when measured at 24 and 48 h after lovastatin treatment (Figs. 2 and 3). Lovastatin treatment not only inhibited the proliferation of C6 cells but also induced cell death (Table 1 and Fig. 3). The cytotoxic effects of lovastatin were prevented by adding mevalonate to the medium, which indicates that such effects of lovastatin were due to the depletion of intracellular mevalonate.

In this study, the mRNA levels of HMG-CoA reductase increased at the cytotoxic concentrations of lovastatin. There is an inverse relationship between the levels of intracellular sterols and HMG-CoA reductase gene expression (Goldstein and Brown, 1990). Such a negative feedback regulation is mediated by sterol regulatory element-binding proteins (SREBPs), SREBP cleavage-activating protein, and uncharacterized proteases (Sakai et al., 1998). Although the exact level of intracellular sterols that can activate SREBPs remains undetermined, Sakai et al. (1998) showed that NH₂-terminal transcription factor domains could be cleaved from the SREBPs in the presence of 50 µM mevalonate and 50 µM compactin. Our data demonstrate the presence of negative feedback regulation of HMG-CoA reductase gene expression in C6 glial cells. In this experiment, the cytotoxic effect of 10 µM lovastatin at 48 h after treatment was reduced compared with the cytotoxic effect induced by 25 µM lovastatin (Fig. 2), although lovastatin at both concentrations could increase the levels of HMG-CoA reductase mRNA (Fig. 1). Considering the data shown in Table 1, it seems that 10 µM lovastatin was not sufficient to block the induced activity of HMG-CoA reductase completely, which allowed the cells to maintain low levels of mevalonate.

In this study, the IC₅₀ values of lovastatin on the proliferation of C6 glial cells could not be calculated because it appeared to be different depending on the time periods of lovastatin treatment. Lovastatin at a concentration of 10 µM inhibited cell proliferation in a time-dependent manner. At 24 h after treatment, lovastatin at a concentration of 10 µM could inhibit cell proliferation only partially because the cell number increased by approximately 3-fold relative to the cell number at the beginning. This increase corresponded to 56.8% of the control. However, 10 µM lovastatin inhibited the increase of cell number completely at 48 h after treatment (Fig. 2). In human promyelocytic HL-60 cells, the IC₅₀ value of lovastatin on DNA synthesis was reported as 10 µM as determined by [³H]thymidine incorporation at 24 h after treatment (Pérez-Sala and Mollinedo, 1994), which is in accordance with our data. It has been reported that it took approximately 8 h to induce the maximal level of HMG-CoA reductase mRNA after 25 µM lovastatin treatment (Choi et al., 1993). Considering that lovastatin-induced increase of HMG-CoA reductase mRNA is due to the negative feedback regulation by the depletion of sterols (Goldstein and Brown, 1990; Sakai et al., 1998), the delayed inhibition of lovastatin on C6 glial cell proliferation might be due to the time period needed for the deprivation of intracellular mevalonate. Our data suggest that the IC₅₀ value for lovastatin on the proliferation of C6 glial cells resides at a concentration range between 1 and 10 µM. Although it was problematic to determine the IC₅₀ value of lovastatin on C6 glial cell proliferation, the ED₅₀ value of mevalonate to prevent cytotoxic effects of lovastatin was approximately 30 µM when cells were cultured in the presence of 25 µM lovastatin for 48 h (Table 1).

In mouse L-M cells, a fibroblast cell line, the IC₅₀ value of lovastatin on the sterol biosynthesis from radiolabeled acetate was reported as approximately 0.02 µM (Alberts, 1988). A discrepancy in lovastatin concentrations between the inhibition of cholesterol biosynthesis and cellular proliferation has been reported by other researchers. For example, the IC₂₅ value of lovastatin on the DNA synthesis in the human vascular smooth muscle cells was 0.8 ± 0.2 µM. In contrast, the IC₅₀ value of lovastatin on the cholesterol synthesis was approximately 0.001 µM under the same experimental conditions (Négre-Aminou et al., 1997). These reports indicate that inhibitory concentrations of lovastatin on cell proliferation are 100- or 1000-fold higher than IC₅₀ value for sterol biosynthesis. This might be due to the different catalytic kinetics between intracellular enzymes using metabolites derived from mevalonate as substrates. In HeLa and CHO-K1 cells, lovastatin at concentrations of 2.5 to 5.0 µM was required to inhibit isoprenylation of p21^ras and prelamin A, whereas the IC₅₀ value for cholesterol synthesis was 0.01 µM in these cell lines (Sinensky et al., 1990). Although the reason for such discrepancy has not been elucidated, our data show that at least 10 µM lovastatin was required to deprive intracellular mevalonate in C6 glial cells as estimated by the levels of HMG-CoA reductase mRNA (Fig. 1).

In this study, lovastatin induced cell death in C6 glial cells, which was associated with chromatin condensation and internucleosomal DNA fragmentation. In human malignant glioma and HL-60 cells, lovastatin-induced DNA fragmentation was blocked by adding mevalonate to the medium (Jones et al., 1994), which is in agreement with our observations. Interestingly, concentrations of lovastatin that induce cytotoxic effects are different depending on the cell types. In glioma cells, DNA fragmentation was observed at 0.1 µM (Jones et al., 1994), whereas lovastatin at a concentration of 25 to 100 µM resulted only in reversible inhibition of DNA synthesis and proliferation in primary cultured glial cells (Langan and Slater, 1991). In our experiment, lovastatin at a concentration of 10 µM induced DNA fragmentation in C6 glial cells. These reports and our data suggest that the required level of intracellular mevalonate to prevent cells from death is different depending on the cell types.

Mevalonate is the precursor of various intracellular molecules such as cholesterol and isoprenoids. Besides cholesterol, which is the main component of cell membrane, two isoprenoids, farnesyl pyrophosphate and geranylgeranyl pyrophosphate, have important roles in the post-translational modification of various membrane-bound proteins mediating the signal transduction (Zhang and Casey, 1996). It seems probable that the cytotoxic effects of lovastatin were induced by nonspecific suppression of the activities of prenylated proteins. In this study, we could differentiate two different cytotoxic effects of lovastatin: inhibition of proliferation and induction of apoptosis. p21^Ras, the well known farnesylated protein involved in the G₁/S transition (Howe et al., 1993), and p27^Kip1, a cyclin-dependent kinase inhibitor (Hengst and Reed, 1996), may be the candidate proteins that are involved in the antiproliferative effect of lovastatin in C6 glial cells. It has been reported that lovastatin induced p27^Kip1 at the translational level (Hengst and Reed, 1996).

In this study, lovastatin-induced apoptosis was blocked by cycloheximide and actinomycin D. The effects of cycloheximide and actinomycin D in protecting cells from death are different in different systems (Gerschenson and Rotello, 1992). One widely accepted explanation for the protective effects of these antimetabolites is that they prevent the expression of proapoptotic genes (Qin et al., 1998). However, the induction of proapoptotic genes in the apoptosis has not been firmly established, and alternative explanations also seem possible. Recent reports suggest that the protective effects of these antimetabolites may not be related to their inherent actions on the protein or RNA synthesis (Morris and Geller, 1996; Furukawa et al., 1997). In addition, it was suggested that these antimetabolites just delayed the onset of chromatin degradation (Chow et al., 1995). Although the action mechanism of cycloheximide or actinomycin D has not been evaluated in our studies, it seems clear that these antimetabolites at appropriate concentrations could prevent or delay lovastatin-induced apoptosis in C6 glial cells. Interestingly, the time point at which cycloheximide or actinomycin D could rescue lovastatin-induced apoptosis was earlier than that at which mevalonate could rescue lovastatin-induced apoptosis. These data suggest that the labile protein promoting cell death was constitutively expressed in C6 glial cells and that this protein might be activated by depletion of intracellular sterols. It is also possible that lovastatin might inactivate protein or proteins suppressing apoptosis and that cycloheximide prevented apoptosis directly by changing the levels of labile protein or proteins promoting/suppressing apoptosis. The existence of such labile proteins was suggested in tumor necrosis factor--induced apoptosis in endothelial cells (Polunovsky et al., 1994), although these proteins remain to be identified. In either case, the mevalonate pathway seems to play a critical role in preventing apoptosis. In conclusion, our data demonstrate that lovastatin can inhibit cell proliferation and induce apoptosis in C6 glial cells, which highlights the importance of mevalonate pathway on the regulation of cell proliferation and prevention of apoptosis.