Introduction

The conversion of HMG-CoA to mevalonate (MVA) is catalyzed by HMG-CoA reductase, the rate-limiting enzyme of the mevalonate pathway (Figure 1). MVA constitutes the precursor of isoprene units incorporated into sterol and non-sterol compounds such as cholesterol, dolichol, ubiquinone (coenzyme Q), isopentenyladenine, geranylgeranyl pyrophosphate (GGPP), and farnesyl pyrophosphate (FPP).1 These end products of the pathway are important and necessary for diverse cellular functions. Cholesterol is a component of cellular membrane structure, as well as a precursor for biosynthesis of steroid hormones and bile acid.2,3 Dolichol, in its phosphorylated form, works as a carrier molecule of oligosaccharides in N-linked protein glycosylation.4 Ubiquinone functions as an electron acceptor in the mitochondrial respiratory chain,5 as well as an antioxidant with an important function in the inhibition of lipid peroxidation.6 Isopentenyladenine is an essential substrate for the modification of some tRNAs.1 Geranylgeranyl transferases and farnesyl transferase utilize GGPP and FPP, respectively, for post-translational isoprenylation of protein.7 There are a large number of proteins that are modified by prenylation including Ras, nuclear lamins, and many small GTP-binding proteins such as members of the Rab, Rac, and Rho families.8 Inhibition of the mevalonate pathway by HMG-CoA reductase inhibitors results in depletion of mevalonate, the direct product of the enzyme reaction, and thereby prevents the biosynthesis of the many diverse downstream products of this key biochemical pathway.

Figure 1
figure 1

The mevalonate pathway.

Full size image

We and others have reported that the HMG-CoA reductase inhibitor lovastatin induces apoptosis of a variety of tumor cells including acute myelogenous leukemia,9 medulloblastoma,10 mesothelioma,11 and neuroblastoma.12 Recently, we have shown that lovastatin also induces significant apoptosis in a large number of tumor-derived cell lines including juvenile monomyelocytic leukemia, pediatric solid malignancies (rhabdomyosarcoma, medulloblastoma) and squamous cell carcinoma of the cervix and of the head and neck.13 The antiproliferative properties of lovastatin suggest that this agent may be an effective anticancer drug.9,14 Indeed, we have recently reported that lovastatin controlled leukemic blast cell counts in an elderly female patient with relapsed AML.15

The mechanism underlying lovastatin-induced apoptosis of malignant cells remains unclear. As there are several end products in the mevalonate pathway, it is important to address which branch of this pathway is essential for lovastatin-induced cell death. In this article, we address this issue and show that the inhibition of the mevalonate pathway, specifically the geranylgeranylation of proteins is critical for lovastatin-induced apoptosis in both established and primary AML cell cultures.

Materials and methods

Tumor cells and culture conditions

The AML cell lines employed in the present study are OCI-AML-2, OCI-AML-3, OCI-AML-5 (hereafter referred to as AML-2, AML-3 and AML-5, respectively) and NB-4 as previously described.9 These cells were cultured in alpha-minimal essential media (α-MEM; Princess Margaret Hospital Media Service, Toronto, ON, Canada) with 10% fetal bovine serum (Sigma, St Louis, MO, USA) and antibiotics in a humidified atmosphere of 5% CO2 at 37°C. Media for AML-5 cells was supplemented with 10% 5637 conditioned media.

Compounds and treatment

Lovastatin was kindly provided by Apotex (Mississauga, Ontario, Canada) and prepared as previously described.9 Since lovastatin from Merck Research Laboratories (Montreal, Quebec, Canada) was used in our previous studies, we compared the two sources of lovastatin. Their chemical structures are identical and no difference in their biological activity was evident (data not shown). AML cell lines were exposed to 20 μM lovastatin for 48 h to induce apoptosis according to previous results.9 Concomitantly, these cells were co-incubated with mevalonate, cholesterol, squalene, lanosterol, desmosterol, dolichol, dolichol phosphate, isopentenyladnine, ubiquinone, FPP or GGPP (all purchased from Sigma).

Geranylgeranyl transferase inhibitors (GGTI-298 and GGTI-2166) and farnesyl transferase inhibitor (FTT-277) were kind gifts of Dr S Sebti (University of South Florida, Tampa, FL, USA). These compounds were dissolved in dimethyl sulfoxide (DMSO) containing 10 mM DTT as suggested by the supplier.

Evaluation of cell proliferation and apoptosis

In vitro effect on proliferation of leukemic cells was determined using the 3-4,5-dimethylthiazolyl-2,2,5-diphenyl tetrazolium bromide (MTT) assay as described previously.9

Apoptosis induced by lovastatin was quantified by the TUNEL and Annexin V assays. For the TUNEL assay, cells were collected, washed with PBS, fixed in 4% formaldehyde and suspended in 70% ethanol. The cells were stored at −20°C and analyzed within a week. After washing twice with PBS, 5 × 105 cells were labeled with 0.02 mM biotin-dUTP and 12.5 U TdT enzyme in a 1× reaction buffer (200 mM potassium cacodylate, 25 mM Tris-HCl, 25 μg/ml bovine serum albumin, pH 6.6), 2.5 mM CoCl2 and 0.01 mM dTTP (Roche Molecular Biochemicals, Laval, QC, Canada) for 45 min at 37°C. Thereafter, the cells were washed once and labeled with avidin-FITC at room temperature for 60 min. After washing with PBS, the cells were analyzed with a FACScalibur cytometer (Becton Dickinson, San Jose, CA, USA). For the Annexin V assay, cells were harvested, washed once with ice-cold PBS. 1 × 105 cells were stained with Annexin V-FITC (PharMingen, San Diego, CA, USA) and analyzed by flow cytometry.

Immunoblot for isoprenylation analysis

Cells were co-incubated with 20 μM lovastatin and FPP or GGPP for 48 h at the indicated concentrations or exposed to the transferase inhibitors, GGTI-298 or FTI-277 for 48 h. Following incubation, the cells were harvested, washed with ice-cold PBS, and lysed in 50 mM HEPES, pH 7.5, 10 mM NaCl, 1% Triton X-100, 10% glycerol, 5 mM MgCl2, 1 mM EDTA, 2 mM Na3VO4, 10 μg/ml soybean trypsin inhibitor, 6.4 mg/ml phosphatase substrate, 25 μg/ml leupeptin, 10 μg/ml aprotinin, and 2 mM phenylmethylsulfonyl fluoride.16 Protein extracts were collected after centrifugation at 14 000 r.p.m., 4°C for 20 min. Equivalent amounts of protein (40 μg/lane) were resolved on a 12% SDS-polyacrylamide gel and thereafter transferred electrophoretically on to Immobilon-P transfer membrane (Millipore Corporation, Bedford, MA, USA). Western blotting was performed using a polyclonal rabbit anti-Rap1A antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) or a mouse monoclonal anti-Rab5 antibody (Transduction Laboratories, Lexington, KY, USA). Detection was achieved using a horseradish peroxidase-labeled secondary antibody in an ECL detection system (NEN Life Science, Boston, MA, USA).

Results

MVA or GGPP abrogate lovastatin-induced cytotoxicity in AML-3 cells

There are at least six pathways downstream of mevalonate (Figure 1). It is important to delineate which pathway(s) is essential for lovastatin-induced cytotoxicity. The cytotoxic effects of lovastatin and the inhibition of this cytotoxicity by the various products of the MVA pathways were evaluated using the MTT assay. In AML-3 cells, the cytotoxicity triggered by 20 μM lovastatin was inhibited following exposure to MVA in a dose-dependent manner. The concentration range of MVA tested was 0.1–200 μM and the inhibitory effect was first evident at 5 μM and maximal at 100 μM (Figure 2a). To test whether this prevention of cytotoxicity is specific for HMG-CoA reductase inhibitors, AML-3 cells were co-incubated with cytosine arabinoside or daunorubicin and MVA, however MVA could not block the cytotoxic effect of these chemotherapeutic agents (data not shown).

Figure 2
figure 2

The effect of mevalonate and its products on lovastatin-induced cytotoxicity in AML-3 cells. Cells were co-incubated with 20 μM lovastatin and individual compounds: (a) mevalonate; (b) cholesterol, squalene, lanosterol or desmosterol; (c) dolichol or dolichol phosphate; (d) ubiquinone; (e) isopentenyladenine; (f) FPP, GGPP or FPP+GGPP at the indicated concentrations for 48 h. Cytotoxic effects, expressed relative to solvent control, were measured using the MTT assay. Data are presented as the means ± s.d. of three independent experiments.

Full size image

In the cholesterol biosynthesis pathway, four products, ie squalene, lanosterol, desmosterol and cholesterol, were added to the culture medium at concentrations between 0.1 and 200 μM.17,18,19 None of these compounds were able to block cytotoxicity induced by lovastatin (Figure 2b). Other mevalonate metabolites that had no protective effects on lovastatin-induced cytotoxicity, included dolichol or dolichol phosphate (1–100 μg/ml; Figure 2c), ubiquinone (1–50 μM; Figure 2d), or isopentenyladenine (1–200 μg/ml; Figure 2e). Preincubation with these compounds for 24 h prior to lovastatin exposure also did not inhibit lovastatin-induced cytotoxicity (data not shown).

To determine whether protein isoprenylation is critical to the cytotoxic effects of lovastatin, AML-3 cells were exposed to solvent control or a broad concentration range of GGPP or FPP. Co-incubation with 20 μM lovastatin and GGPP (0.01–10 μM) blocked the loss of MTT activity caused by lovastatin exposure in a dose-dependent manner. As shown in Figure 2f, this protective effect was detectable from 0.1 μM GGPP (17% of control) and reached maximal level at 1 μM (69% of control). Interestingly, at the same concentration range, FPP had only partial protective effects. The inhibition of lovastatin-induced cytotoxicity by FPP was observed at 0.5 μM (20% of control) and maximal protection occurred with 10 μM (64% of control). When both GGPP and FPP were added concomitantly to the cell culture containing lovastatin, the MTT activity of AML-3 cells was 19% at 0.1 μM FPP and GGPP, and reached 80% at 0.75 μM FPP and GGPP.

MVA and GGPP also block lovastatin-induced cytotoxicity in other AML cells

We have reported that other AML cells share similar sensitivity to lovastatin including AML-2, AML-5 and NB-4.9 To determine if lovastatin-induced cytotoxicity in these leukemic cells shared a similar mechanism as in AML-3, each cell line was co-incubated with 20 μM lovastatin and MVA, squalene, GGPP or FPP and the effects were assessed by the MTT assay. MVA (Figure 3a) blocked the lovastatin-induced cytotoxicity in AML-2, AML-5 and NB-4 cells. In contrast, squalene had no inhibition of lovastatin effect in these cell lines (Figure 3b). GGPP (Figure 3c) reversed lovastatin-induced cytotoxicity in these cell lines, while FPP only partially inhibited the effect of lovastatin (Figure 3d). Similar to the effects in the AML cell lines, MVA and GGPP blocked lovastatin-induced cytotoxicity and FPP partially blocked lovastatin-induced cytotoxicity in AML patient samples (data not shown).

Figure 3
figure 3

The effects of mevalonate and downstream products (squalene, GGPP and FPP) on lovastatin-induced cytotoxicity in additional AML cell lines. AML cell lines AML-2, AML-5 and NB-4, were exposed to 20 μM lovastatin and mevalonate or its products. (a) MVA; (b) squalene; (c) GGPP; (d) FPP at indicated concentrations for 48 h. Cell survival was assayed by MTT. The results are shown as means ± s.d. of three independent experiments.

Full size image

Lovastatin-induced apoptosis is abrogated by MVA or GGPP in AML-3 cells

Previously, we have reported that the cytotoxicity in response to lovastatin is due to the induction of apoptosis.9 Here, we determined whether the reversal of MTT activity by MVA or GGPP is due to the inhibition of lovastatin-induced apoptosis in AML-3 cells. Two traditional apoptosis assays were employed: the TUNEL method, which detects DNA-strand breaks in the apoptotic cells20 and the Annexin V apoptosis assay, which measures the translocation of phosphatidylserine from the inner face of the plasma membrane to the cell surface soon after the induction of apoptosis.21 The percentage of apoptotic cells quantified varies slightly between the two types of apoptosis assays due to the principle of the assay. The results from the TUNEL method showed that exposure to 20 μM lovastatin for 48 h triggered significant apoptosis (66.8%) of AML-3 cells (Figure 4a) and MVA inhibited this lovastatin-induced apoptosis starting at a concentration of 5 μM (Figure 4b, upper panel). When the MVA concentration was increased to 55 μM or greater, complete inhibition was observed (Figure 4a, b, upper panel). When AML-3 cells were coincubated with 0.1–10 μM GGPP, dose-dependent inhibition of apoptosis was evident (Figure 4b, middle panel). Inhibition of apoptosis was initially seen at 0.1 μM, 50% inhibition of apoptosis was achieved at approximately 0.2 μM and complete inhibition was achieved at 2.5 μM or greater (Figure 4a, b, middle panel). Similarly, FPP (Figure 4b, lower panel) also inhibited lovastatin-induced apoptosis in a dose-dependent manner, but with less potency than GGPP (Figure 4b, middle panel). FPP inhibited 50% apoptotic death at 2.5 μM and complete inhibition was not observed at the tested concentrations (Figure 4a, b, lower panel). As seen in Figure 5, the results from the Annexin V assay were similar to those from the TUNEL method. MVA (Figure 5a, b, upper panel) or GGPP (Figure 5a, b, middle panel) block lovastatin-induced apoptosis completely at 85 μM or 2.5 μM, respectively, while FPP had only partial preventive effects (Figure 5a, b, lower panel). Similar analysis of the additional AML cell lines and primary AML patient samples was consistent with the AML-3 results (data not shown).

Figure 4
figure 4

Evaluating the protective effects of MVA, GGPP or FPP on lovastatin-induced apoptosis in AML-3 cells using the TUNEL assay. (a) Flow cytometry analysis showed the populations of apoptotic and non-apoptotic cells in AML-3 cells. The cells were treated with ethanol (control), 20 μM lova, 20 μM lova + 100 μM MVA, 20 μM lova + 10 μM GGPP or 20 μM lova + 10 μM FPP for 48 h. The percentage of cells staining as TUNEL positive is indicated in each histogram. (b) AML-3 cells were co-incubated with 20 μM lovastatin and 0.1–100 μM MVA (upper), or 0.1–10 μM GGPP (middle), or 0.1–10 μM FPP (lower) for 48 h. The percentage TUNEL-positive cells are plotted and results shown are means ± s.d. (n = 3). Ap, apoptosis; lova, lovastatin.

Full size image
Figure 5
figure 5

Analyzing the effects of MVA, GGPP or FPP on lovastatin-induced apoptosis in AML-3 cells by Annexin V assay. (a) Flow cytometry analysis showed the distribution of apoptotic (upper panel) and non-apoptotic (lower panel) AML-3 cells. The cells were treated as described in Figure 4a and then assayed for the presence of Annexin V. The percentage of apoptotic, Annexin V-positive cells is shown in the upper panel of each histogram. (b) Concentration-dependent inhibition of lovastatin-induced apoptosis in AML-3 cells by MVA (upper), GGPP (middle), or FPP (lower). The cells were treated as described in Figure 4b and then assayed for Annexin V. Data are presented as the means ± s.d. of three independent experiments.

Full size image

Geranylgeranyl transferase inhibitors can induce apoptosis in AML cells

To further investigate the role of protein isoprenylation in lovastatin-induced apoptosis of AML-3 cells, we assessed the effect of geranylgeranyl transferase and farnesyl transferase inhibitors (GGTI, FTI). Both TUNEL and Annexin V apoptosis assays described above were employed to assess apoptosis. The TUNEL assay revealed that GGTI-298 induced apoptosis in a dose dependent manner in AML-3 cells (Figure 6a). Significant and comparable apoptosis was observed when AML-3 cells were exposed to 20 μM lovastatin or 20 μM GGTI-298 for 48 h (Figure 6b and c). By contrast, FTI-277 produced only a weak induction of apoptosis in AML-3 cells. Another geranylgeranyl transferase inhibitor GGTI-2166, which is more selective for geranylgeranyl transferase I,22 induced apoptosis in AML-3 cells to the same extent as GGTI-298 (data not shown).

Figure 6
figure 6

Apoptosis triggered by GGTI-298 in AML-3 cells. (a) Cells were cultured in the presence of 1–20 μM GGTI-298 or FTI-277 for 48 h and analyzed using the TUNEL assay. (b and c) AML-3 cells were exposed to ethanol, 20 μM lova, 20 μM GGTI-298 or 20 μM FTI-277 for 48 h. Apoptotic cells were quantitated by the TUNEL or Annexin V assay. Results shown are based on three independent experiments and presented as means ± s.d. Ap, apoptosis; lova, lovastatin.

Full size image

GGPP reverses lovastatin-induced apoptosis through geranylgeranylation of proteins

To monitor protein geranylgeranylation in AML cells after exposure to lovastatin, GGTI-298 or FTI-277, we evaluated the presence of processed and unprocessed forms of proteins shown previously to be geranylgeranylated by McGuire et al,16 in a NIH3T3 cell system. The proteins analyzed include Rap1A and Rab5, shown to be modified by GGTase type I and II, respectively.23,24 In control untreated AML-3 cells, all Rap1A protein was processed, while Rab5 protein was in both processed and unprocessed forms (Figure 7, lane 1). Treatment with 20 μM lovastatin resulted in inhibition of geranylgeranylation of both of these proteins (Figure 7, lane 2). When co-incubated with 10 μM GGPP, processing of these proteins was recovered (Figure 7, lane 3). In contrast, co-incubation with 10 μM FPP in lovastatin-treated cells had no significant effects on the isoprenylation of Rap1A or Rab5 (Figure 7, lane 4). Moreover, the inhibition of protein geranylgeranylation in both Rap1A and Rab5 in AML-3 cells exposed to lovastatin was similarly evident after exposure to 20 μM GGTI-298 (Figure 7, lanes 2 and 5, respectively). However, no such inhibition was evident in the cells treated with FTI-277 (Figure 7, lane 6). To ensure that FTI-277 was indeed inhibiting protein farnesylation, an immunoblot of H-ras was performed and the expected alteration in protein farnesylation was evident in response to lovastatin, or FTI-277 exposure, or co-incubation with lovastatin and FPP (data not shown).

Figure 7
figure 7

GGPP rescues lovastatin-induced inhibition of protein geranylgeranylation in AML-3 cells. Cells were exposed to solvent control (lane 1), 20 μM lovastatin (lane 2), 20 μM lovastatin + 10 μM GGPP (lane 3), 20 μM lovastatin + 10 μM FPP (lane 4), 20 μM GGTI-298 (lane 5) and 20 μM FTI-277 (lane 6) for 48 h. The cells were then harvested and cell lysates were subjected to SDS-polyacrylamide gel and immunoblotted for Rap1A and Rab5 using respective antibodies. Detection was performed by the ECL system (for the details, see Materials and methods).

Full size image

Discussion

Mevalonate is the key intermediate in the isoprenoid synthetic pathway leading to the formation of sterols, polyprenols, ubiquinone, and the farnesyl and geranylgeranyl group of prenylated proteins.1 HMG-CoA reductase acts as the principle regulatory step in the mevalonate pathway. The inhibition of this enzyme results in the depletion of its direct product, mevalonate and its derivatives. Several investigations have shown that lovastatin inhibits mevalonate production and that mevalonate overcomes lovastatin-induced cell death in a variety of cell types, including colon cancer cells,25 medulloblastoma cells26 and C6 glial cells.27 We also found that mevalonate completely inhibited lovastatin-induced cell death in AML cell lines (Figure 2a, 3a, 4 and 5). Taken together, these results demonstrate that the cytotoxic effects of lovastatin are due to its ability to inhibit mevalonate synthesis. This prevention is specific to lovastatin effects since mevalonate was unable to prevent cell death in response to the cytotoxic drugs cytosine arabinoside and daunorubicin (data not shown).

Our present data clearly show that the depletion of mevalonate is responsible for lovastatin-induced apoptosis in AML cells, suggesting that blocking the production of specific mevalonate derivatives must be involved in this process. Cholesterol biosynthesis is the major metabolic branch in the mevalonate pathway and certain aspects of cholesterol metabolism appear to be relevant to cancer.28 It has been reported that SFK 104976, an inhibitor of lanosterol 14-α demethylase, induces apoptosis in HL-60 cells, a human acute myeloid cell line.29 In contrast, inhibition of cholesterol biosynthesis by squalene epoxidase inhibitors TU-2078 and NB-598 prevents apoptosis in L6 myoblasts30 and cholesterol itself induces apoptosis in human erythroleukemia K562 cells.31 Taken together, the relationship between the inhibition of cholesterol end products and cell death appears to be dependent on the intracellular cholesterol content rather than other end products in the mevalonate pathway, as shown in a study using a neuronal cell system and a HMG-CoA reductase inhibitor.32 In the present study, four metabolites of cholesterol synthesis, cholesterol, squalene, lanosterol and desmosterol, were supplemented in cell cultures treated with lovastatin. No protective effects against lovastatin cytotoxicity could be observed from any of these compounds in AML-3 cells (Figure 2b). Furthermore, squalene also failed to prevent lovastatin-induced cytotoxicity in other AML cell lines (Figure 3b). These results suggest that the inhibition of cholesterol production might not be critical to lovastatin-induced cytotoxicity in AML cells.

Numerous reports have shown the importance of cell-surface N-linked oligosaccharides for maintaining various cellular functions in tumor cells.33 Inhibition in the synthesis of N-linked glycoproteins on the surface of tumor cells may affect proliferation of these cells, or lead these cells to be more prone to apoptotic cell death. Dolichol phosphate functions as a carrier of oligosaccharide units in the synthesis of N-linked glycoprotein.4 Inhibition of dolichol phosphate synthesis would result in an impairment of protein glycosylation. Depletion of mevalonate, caused by HMG-CoA reductase inhibitors, would depress the biosynthesis of dolichol phosphate and the rate of N-linked glycosylation. To test this possibility, both dolichol and dolichol phosphate were used to inhibit lovastatin-induced cytotoxicity in our experiments. Neither compound could prevent cell death caused by lovastatin (Figure 2c), suggesting that inhibition of dolichol biosynthesis may not account for the mechanism of lovastatin-induced cell death in AML cells.

In addition to cholesterol and dolichol, we have also shown that isopentenyladenine and ubiquinone were ineffective in preventing lovastatin-induced cytotoxicity in AML cells (Figure 2d and e). It has been reported that ubiquinone supplementation prevents lovastatin-induced myopathy.34 Since our present data showed that ubiquinone did not inhibit lovastatin-induced cytotoxicity, it may be a useful supplement for those patients unable to tolerate this side-effect associated with lovastatin. The failure of these various compounds to overcome cytotoxicity of lovastatin strongly suggests that other product(s) of mevalonate, most notably isoprenylated proteins, are involved in the control of cell survival.

Prenylated proteins are post-translationally modified at or near the carboxyl terminus by formation of cysteine thioethers with the isoprenoid lipid substrates, FPP or GGPP. Three distinct enzymes, farnesyl transferase (FPTase), geranylgeranyl transferase I (GGPTase I) and (GGPTase II) that transfer these isoprenoid substrates to recipient proteins have been identified.8 These transferases are dependent upon FPP and GGPP substrates to modify target proteins. These include small GTP-binding proteins, such as Ras, Rho, Rab, Rac, and Rap that are involved in important cellular functions, such as the regulation of proliferation, signal transduction, and cell death.35 Some investigations have shown that lovastatin induces apoptotic cell death through the block of geranylgeranylated and/or farnesylated proteins in different cell types. For example, GGPP or geranyl geraniol completely prevents lovastatin-induced apoptosis in mouse proximal tubular cells,18 colon cancer cells,25 prostate stromal cells,36 and rat pulmonary vascular smooth muscle cells.37

It has been shown that the geranylgeranylated Rho family members play important roles in cellular signal transduction and possess dual roles in cell proliferation and apoptosis.38,39 Signaling pathways driven by these proteins have been associated with the function and/or expression of molecules that regulate the apoptotic response. The activation of multiple signaling pathways by Rho proteins includes PI3K,40 several serine/threonine kinases,41 transcription factor NF-kB42 and Jun.43 Several members of the small GTP-binding proteins, including Rho, may play a role in the regulation of lovastatin-induced apoptosis in AML cells. However, it is estimated that 0.5 to 1% of all cellular proteins are geranylgeranylated44 and further studies are underway to identify the key substrates essential for this apoptotic response.

In the present study, we have shown that GGPP completely inhibited lovastatin-induced apoptosis in AML cells. Further support for a central role of GGPP in lovastatin-induced apoptosis in AML is evident by the ability of GGTIs to induce apoptosis in AML. The reversal of lovastatin-induced apoptosis in AML cells by GGPP may be due to the replenishment of the intracellular pool of GGPP that is depleted by lovastatin treatment. By contrast, FPP leads to a partial reversal of lovastatin-induced apoptosis. The inability of FTI-277 to induce apoptosis in AML cells confirms the minor role of FPP in lovastatin-induced apoptosis in AML. The partial protective effect of the addition of FPP may be due to the lack of isopentenyl-PP required to convert FPP to GGPP.45 Alternatively, proteins that are normally geranylgeranylated may be farnesylated under the conditions of intracellular GGPP shortage.46

Western blotting showed that lovastatin and GGTI-298 similarly blocked the geranylgeranylation of both Rap1A and Rab5 suggesting that inhibition of GGPTases contributes to the apoptotic response seen in AML cells. The addition of GGPP to AML cells exposed to lovastatin restored the presence of the processed forms of both substrates and was associated with inhibition of lovastatin-induced apoptosis. The role of GGPTase I and II in the AML cell is unknown and requires further investigation. Likewise, identification of the geranylgeranylated proteins that mediate the anticancer properties of lovastatin may lead to more refined therapeutic approaches.