| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Cell and Tumor Biology |
Departments of 1 Experimental Therapeutics; 2 General Internal Medicine, Ambulatory Treatment and Emergency Care; 3 Molecular Therapeutics; and 4 Endocrine Neoplasia and Hormonal Disorders, University of Texas M.D. Anderson Cancer Center, Houston, Texas
Requests for reprints: Sai-Ching J. Yeung, Department of Endocrine Neoplasia and Hormone Disorders, University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030. E-mail: syeung{at}mdanderson.org or Jingxuan Pan, E-mail: jpan{at}mdanderson.org.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
FTIs were originally developed to inhibit the mitogenic function of ras oncogenes. However, the antineoplastic activity of FTIs also results, at least in part, from inhibition of the farnesylation of proteins other than Ras (3, 4). More than 10 uncharacterized isoprenylated proteins are affected by FTIs (5). Among the known farnesylated proteins other than Ras, lamins A and B, Rap2, RhoB and RhoE, and inositol triphosphate 5-phosphatase type I may be relevant in intracellular signaling and apoptosis. FTI treatment can decrease farnesylated RhoB and increase geranylgeranylated RhoB to inhibit cell cycling and reverse malignant transformation selectively in cancer cells (6). Geranylgeranylated RhoB is required for the antineoplastic action of FTIs (7). Cyclin B1 suppression may be an important step in FTI-induced apoptosis that is mediated by geranylgeranylated RhoB (8). Inhibition of the PI-3 kinase/Akt-mediated cell survival and adhesion pathway by FTIs has been shown to play a role in the induction of apoptosis (9). Heat shock protein 70 may also be a target for induction of apoptosis by FTIs as well (10, 11).
The FTI manumycin A is a natural product of Streptomyces parvulus (12), and a selective FTI (IC50 = 5 µmol/L) that inhibits geranylgeranyl transferase only at high concentrations (IC50 = 180 µmol/L). Manumycin inhibits farnesyltransferase through competition with the farnesyl PPi group (13), a mechanism of action different from peptide mimicry of the CAAX motif by FTIs, such as SCH66336(lonafarnib), FTI-276, and L-744832 (14). We previously showed that the combination of manumycin and paclitaxel had a synergistic antineoplastic effect on anaplastic thyroid cancer, one of the most aggressive solid tumors (15). In the course of this investigation (15–17), we discovered that manumycin induced a paradoxical increase in [3H]thymidine incorporation (data not shown) similar to that induced by hydroxymethylglutaryl CoA-reductase inhibitors (18, 19). DNA repair was proposed as a possible explanation for the paradoxical increase in [3H]thymidine incorporation induced by lovastatin (18). Recently, FTIs were observed to induce reactive oxygen species (ROS) in chronic myelogenous leukemia cells (20).
On the basis of our findings and others, we hypothesized that FTIs induced ROS that damage DNA, leading to DNA damage responses. We tested this hypothesis by evaluating the effects of manumycin on generation of ROS, DNA damage, and DNA-damage-sensing kinases, the effect of quenching ROS on these responses to manumycin, and the relevance of ROS generation to the overall cytotoxicity of manumycin. The phenomena of ROS generation and DNA damage were also examined for four FTIs in a variety of human cancer cell lines of different tissue origins. Because RhoB is inducible after DNA damage (21) and required to mediate apoptosis in malignant cells after DNA damage (22), we also investigated whether FTIs indirectly increased RhoB via ROS-induced DNA damage.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell culture. MCF-7 (human breast cancer) and HCT116 (human colon cancer) cells were obtained from the American Type Culture Collection (ATCC, Manassas, VA) and incubated in DMEM plus 10% fetal bovine serum (FBS). Anaplastic thyroid cancer cells, ARO (23) and KAT-4 (24), were gifts from Dr. J.A. Fagin (University of Cincinnati, Cincinnati, OH) and Dr. K.B. Ain (University of Kentucky, Lexington, KY), respectively. These cells were cultured in RPMI 1640 supplemented with bovine serum (10%). Human primary lung fibroblast cells CCD-37 (CRL-1496) were obtained from ATCC and cultured in DMEM supplemented with 15% FBS. AT22IJE-T cells [ataxia-telangiectasia mutated protein (ATM)–/– fibroblasts] and AT22IJE-TpEBS7-YZ5 cells (ATM+/+, ATM–/– fibroblasts complemented with ATM cDNA; ref. 25) were gifts from Dr. Y. Shiloh (Tel Aviv University, Tel Aviv, Israel). These ataxia telangiectasia cells were cultured in DMEM supplemented with heat-treated bovine serum (20%). M059J and M059K cells (26) were cultured in 
-medium supplemented with heat-treated bovine serum (20%). U2OS ATM- and Rad3-related protein (ATR)-kinase dead-inducible cells, a gift from Dr. S.L. Schreiber (Harvard University, Cambridge, MA), were grown in DMEM plus 10% FBS containing 200 µg/mL G418 and 50 µg/mL hygromycin. Cells were induced with 1 µg/mL doxycycline for 2 days before exposure to experimental drug treatments (27).
Nitric oxide flow cytometry assay. Nitric oxide (NO) was directly measured using a membrane-permeable diaminofluorescein NO sensor dye (28). Briefly, cells were preincubated with the NO sensor dye in a proprietary kit purchased from BD Clontech (Palo Alto, CA) for 30 minutes and then exposed to treatment solutions for 6 hours. The cells were then collected by trypsinization, pelleted by centrifugation, resuspended in PBS, and analyzed by flow cytometry.
Measurement of intracellular superoxide anion. Cells were grown in 96-well plates and treated with manumycin for specified durations. Then, the cells were incubated with 1 µmol/L dihydroethidium for 10 minutes at 37°C in the dark (29). Cells were rinsed twice with PBS. One hundred microliters of PBS were then added to each well. Ethidium levels were immediately measured with a fluorescent plate reader (SpectraFluor Plus, Tecan, Switzerland) with excitation wavelength at 430 nm and emission wavelength at 590 nm. Alternatively, the level of intracellular superoxide in some samples was measured using dihydroethidine by flow cytometry as described previously (30).
Sulforhodamine B protein biomass assay. Cells were grown in 96-well plates. After treatments, the cells were fixed with cold 50% trichloroacetic acid 50 µL per well for 1 hour at 4°C. After fixation, plates were washed several times with tap water and air dried. Trichloroacetic acid–fixed cells were stained for 20 minutes with sulforhodamine B (0.4% w/v in 1% acetic acid). The unbound dye was washed out with 1% acetic acid after staining. Two hundred microliters 10 mmol/L Tris base (pH 10.5) were added to each well of each dried plate. Absorbance was measured using a 96-well plate reader at 570 nm.
Preparation of total cell lysates. Control or drug-treated cells were rinsed with PBS and then lysed with radioimmunoprecipitation assay buffer (1x PBS, 1% NP40, 0.5% sodium deoxycholate, 0.1% SDS) or NETN lysis buffer [150 mmol/L NaCl, 1 mmol/L EDTA, 20 mmol/L Tris (pH 8.0)], 0.5% NP40 containing freshly added 10 mmol/L glycerophosphate, 1 mmol/L Na3VO4, 10 µg/mL pepstatin A, 5 µg/mL aprotinin, 10 µg/mL leupeptin, 10 mmol/L NaF, and 1 mmol/L phenylmethylsulfonyl fluoride. The DNA in the lysate was sheared by sonication with eight 1-second bursts at medium power.
SDS-PAGE and immunoblotting. SDS-PAGE and immunoblotting was done using standard methods as described previously (15). The protein concentrations of samples were measured using a modified Lowry method (DC Protein Assay, Bio-Rad Laboratories, Hercules, CA). Equal amounts of total protein from each sample were loaded onto the SDS-polyacrylamide gel.
Pulsed-field gel electrophoresis. ARO cells treated with manumycin and control cells were washed and suspended with PBS after trypsin-EDTA treatment. Approximately 2 x 106 ARO cells were washed and resuspended in cold PBS and quickly mixed with an equal volume of 1% low-melting-point agarose prepared in PBS. The mixture was immediately poured into molds. The cells embedded in agarose were digested in lysis buffer containing 0.5 mol/L EDTA (pH 8.0), 10 mmol/L Tris, 1% (w/v) Sarkosyl, and 0.1 mg/mL proteinase K at 50°C for 24 hours, followed by washing with TE buffer [10 mmol/L Tris-HCl (pH 8.0), 1 mmol/L EDTA, and 0.2 mg/mL RNase A] for 3 hours with buffer changed each hour. The agarose plugs containing purified DNA were kept in a storage buffer [10 mmol/L Tris-HCl and 50 mmol/L EDTA (pH 8.0)] at 4°C. Then, the cellular DNA in the plugs was analyzed by pulsed-field gel electrophoresis (CHEF-DR II, Bio-Rad Laboratories) at 200 V with a switch time of 50 seconds for 16 hours at 7°C. The electrophoresis buffer contained 50 mmol/L Tris-borate (pH 8.2) and 1 mmol/L EDTA. After electrophoresis, gels were stained with ethidium bromide and then photographed on a UV transilluminator.
Immunofluorescence staining. Cells were seeded on coverslips in six-well plates and allowed to grow overnight. Cells were exposed to manumycin (54 µmol/L) for 1 hour and washed with PBS twice and then cells were fed with fresh medium. After another incubation for 2 hours, cells were fixed with 3% paraformaldehyde for 20 minutes. Then, the cells were permeabilized in 1% Triton X-100 and 0.5% NP40 for 20 minutes. Samples were blocked with 5% goat serum and then incubated with primary antibody for 2 hours and secondary (FITC) antibody for 1 hour. Cells were then stained with 4',6-diamidino-2-phenylindole (DAPI) to permit visualization of nuclear DNA. The coverslips were mounted onto glass slides with VectaShield antifade (Vector Laboratories, Burlingame, CA), and immunofluorescence was visualized by using a Leica DM LB fluorescence microscope. Images were captured with a Kodak digital imaging system.
DNA damage detection. The cell-free DNA damage detection assay (the 3D assay) was done as described by Barret et al. (31) with modifications. Purified genomic DNA (1 µg/mL, 50 µL) from KAT-4 cells was bound to poly-L-lysine–coated microplates. DNA was then treated for 3 hours with manumycin (54 µmol/L) or paclitaxel (22 µmol/L). UV light–damaged genomic DNA was used as a positive control. The treated DNA and control DNA were rinsed twice with PBS plus Tween 20 (PBST) and incubated for 3 hours at 37°C with 50 µL reaction mixture containing 30 µg HeLa nuclear extract (Promega, Madison, WI), 70 mmol/L KCl, 40 mmol/L HEPES-KOH (pH 7.6), 5 mmol/L MgCl2, 0.5 mmol/L DTT, 10 mmol/L phosphocreatine, 2.5 µg creatine phosphokinase, 2 mmol/L EGTA, 18 µg bovine serum albumin (BSA), 0.4 µmol/L dGTP, 0.4 µmol/L dCTP, 0.4 µmol/L dTTP, and 0.4 µmol/L biotin-14-dATP (Invitrogen). After being washed thrice with PBST plus 0.1% BSA, the DNA was incubated with streptavidin peroxidase and BSA diluted in PBST for 30 minutes. Chemiluminescence substrate (50 µL; Dura Substrate, Pierce Chemical Co., Rockford, IL) was added to each well after rinsing. The emitted light was measured with a Tecan Fluoroscan Plus plate reader in the luminescence mode.
Small interfering RNA transfection. SiGENOME SMARTpool duplexes against ß-subunit of farnesyltransferase (FNTB) and small interfering RNA (siRNA) control duplex against Luciferase GL2 were purchased from Dharmacon RNA Tech. (Lafayette, CO). Transfection of the synthesized siRNA duplex of either anti-FNTB or the control into KAT-4 cells was done using LipofectAMINE 2000 (Invitrogen, Carlsbad, CA) according to the manufacturer's instruction. Briefly, cells were grown in a six-well plate to 40% confluence and then incubated with a mixture of a 200 pmol siRNA duplex and 8 µL LipofectAMINE 2000. After 4 hours, FBS was added to a final concentration of 10% (v/v). A second transfection was done 24 hours later. Experimental drug treatments were carried out 24 hours after the second transfection.
Quantitative real-time reverse transcription-PCR. DNA-free total RNA was prepared with RNeasy minicolumns using the on-column DNase digestion step (Qiagen, Valencia, CA). One-step quantitative real-time reverse transcription-PCR (RT-PCR) was done with ABI Prism 7900HT instrument (Applied Biosystems, Inc., Foster City, CA) according to the manufacturer's recommended protocol. The reaction was carried out in a 25-µL reaction volume in 96-well plates using the specific primers and probes, TaqMan Gene Expression Assays (FNTB Assay ID: Hs00157635_ml) and internal control of glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Each sample was assayed in duplicate in each independent experiment. The comparative CT method (
CT) for relative quantitation of gene expression was used to analyze the mRNA expression. The mean cycle threshold (CT) value for FNTB and GAPDH across replicates of each sample was calculated. CT (sample) = CT (target) – CT (GADPH), CT = CT (sample) – the baseline's CT (siRNA control), and comparative expression level = 2–CT. The results shown are representative of at least three independent experiments.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
6 hours of drug exposure and then decreased gradually. In KAT-4 cells, the average fluorescence values after 4, 6, and 8 hours of treatment were significantly higher than the fluorescence values in control cells (P < 0.01, one-way ANOVA, post hoc comparisons, Tukey's test; Fig. 1A). In all the figures in this paper, error bars represented the 95% confidence intervals. A similar time profile of superoxide induction was observed in ARO cells (data not shown). Manumycin (54 µmol/L, 6 hours) also induced superoxide production in ARO, MCF-7, and HCT116 cells (data not shown), suggesting this effect did not depend on the original tissue origins of the cell lines.
|
We also investigated whether manumycin affected the generation of NO, another type of ROS. We used a proprietary membrane-permeable compound that fluoresces bright green upon binding NO to measure NO production in a fluorescence flow cytometry assay. Three independent experiments were done. Representative fluorescence distribution plots for the ARO and KAT-4 cell lines were shown in Fig. 1C. Manumycin (54 µmol/L) for 6 hours significantly increased the average green fluorescence (i.e., NO production; P < 0.01, Student's t test) in both cell lines. When a cutoff value of fluorescence was defined such that 95% of the untreated control cells were below the cutoff value, the percentage of cells stained positive (higher fluorescence than the cutoff value) by the NO sensor dye was well above 50% in both cell lines.
We examined whether the ROS generation by FTIs was related to inhibition of farnesyltransferase. The two members of the prenyltransfase family, farnesyltransferase and geranylgeranyl transferase-1, are heterodimeric proteins that share an identical -subunit but have different ß-subunits (2). Therefore, we used siRNA targeting the FNTB to evaluate the effects of loss of farnesyltransferase on ROS generation. Transfection of KAT-4 cells with anti-FNTB siRNA decreased the target mRNA level by 90% when measured at 48 hours after the first transfection (Fig. 1D). However, suppressing the FNTB mRNA level by 90% was not sufficient to induce ROS (Fig. 1E). After a subsequent 6-hour exposure to FTIs (SCH66336 FTI-276, and L-744832), there were significant decreases in superoxide levels in cells transfected with anti-FNTB siRNA treated with an FTI compared with the respective cells transfected with control siRNA treated with the same FTI (P < 0.05, Student's t test; Fig. 1E). However, the FTI-elicited superoxide generation was not attenuated proportionally like the 90% of inhibition of mRNA expression. These results suggested that inhibition of farnesyltransferase alone was not sufficient to generate ROS (in other words, the induction of ROS by FTIs was largely independent of farnesyltransferase) and that farnesyltransferase contributed partially to the generation of ROS by FTI in a yet unknown manner.
Farnesyltransferase inhibitors induced DNA double-strand breaks. ROS are well known to damage a variety of biomolecules: proteins, RNA, and DNA. To evaluate whether manumycin could lead to DNA damage, we looked for direct evidence of DSBs by pulsed-field gel electrophoresis (PFGE). ARO cells were exposed to manumycin (54 µmol/L) for 3 hours and harvested, and the degree of fragmentation in the genomic DNA (i.e., DSB) was assessed using PFGE. As shown in Fig. 2A, DSBs were evident in manumycin-treated cells but not in DMSO-treated control cells. In all the figures in this paper, the black dots indicated the presence of a drug or treatment. The presence of a pan-caspase inhibitor or a caspase-3 inhibitor did not interfere with DSB formation induced by manumycin (Fig. 2A), indicating that the formation of DSBs was not due to internucleosomal DNA cleavage in the apoptotic process.
|
-H2AX), a subclass of eukaryotic histones that are part of the chromatin structure (32). Detection of -H2AX is probably the most specific and efficient technique identified to date for detecting DSBs in cells (33–35). Paclitaxel is an antitubulin antineoplastic agent that has activity against anaplastic thyroid cancer (24) and does not damage DNA. We compared paclitaxel and manumycin with respect to DNA breaks and damage secondary to apoptosis and cell death. We examined the formation of -H2AX in KAT-4 and ARO cells exposed to control medium (0.1% DMSO), manumycin alone, paclitaxel alone, or both drugs. After 3 and 6 hours of treatment, cells were harvested and cell lysates were analyzed by immunoblotting with the antibody specifically recognizing Ser39-phosphorylated H2AX. As shown in Fig. 2B, H2AX was phosphorylated in cells treated with manumycin or manumycin plus paclitaxel. The ratio of -H2AX to total histone 2A (H2A, as estimated by densitometric scanning of immunoblots) was 12 times as high in manumycin-treated cells as in control cells at 3 hours of treatment and 16 times as high in manumycin-treated cells as in control cells at 6 hours of treatment. The addition of paclitaxel, which enhanced apoptosis (15), did not enhance H2AX phosphorylation by manumycin, this again suggesting that the DSBs formed were not the result of apoptosis. The failure of paclitaxel to induce -H2AX was in agreement with its mechanism of antineoplastic action. To further investigate whether the DSBs were the result of apoptosis, we treated ARO cells with manumycin or manumycin plus paclitaxel in the presence of an inhibitor of caspase-3 and then assayed for DSBs by immunodetection of -H2AX. As shown in Fig. 2C, the phosphorylation of H2AX after manumycin or manumycin plus paclitaxel treatments were not dependent on caspase-3 activation. This result was consistent with the finding with PFGE (Fig. 2A) and indicated that the DSBs induced by manumycin did not result from apoptosis.
-H2AX organizes into discrete foci in the nucleus for DNA repair. Obvious -H2AX foci formation was observed in manumycin-treated ARO cells (Fig. 2D). The phosphorylation of H2AX and organization of -H2AX into nuclear foci not only indicated the existence of manumycin-induced DSBs but also revealed the activation of DSB detection mechanisms.
To ascertain whether manumycin led to DNA damage directly, we used the cell-free DNA damage detection assay (3D assay; ref. 36). Purified KAT-4 genomic DNA was incubated for 3 hours with control culture medium, manumycin, or paclitaxel. UV light–damaged genomic DNA was used as a positive control. The 3D assay did not detect significant damage in DNA treated directly with manumycin or paclitaxel, but did detect significant damage in UV-treated DNA (Fig. 2E). Therefore, direct incubation of DNA with manumycin was not sufficient to cause DNA damage detectable by the 3D assay.
To assess whether FTI-induced DSBs was a common phenomenon for different FTIs in different cell lines, additional FTIs were tested in two other different solid tumor cell lines. When treated with different FTIs, phosphorylation of H2AX was observed to increase significantly in HCT116 and MCF-7 cells (Fig. 2F). At the concentrations used, manumycin seemed to induce about twice as much phosphorylation of H2AX as the other FTIs. These data supported that DNA damage induced by FTIs might be a common phenomenon.
Manumycin induced DNA damage responses. DSBs pose a major threat to living cells and several mechanisms have evolved for repairing these lesions. Eukaryotes can repair DSBs by homologous recombination (HR) or nonhomologous end joining (NHEJ; ref. 37). HR requires interaction of the broken DNA molecule with an intact homologous copy and allows restoration of the original DNA sequence. NHEJ connects DNA ends irrespective of their nucleotide sequence; this mechanism of repair predominates in mitotic cells (38). Several proteins are involved in DSB repair: BRCA1 functions in HR, DNA-PK and Ku70 function in NHEJ (39), and the MRE11-NBS1-RAD50 complex functions in both HR and NHEJ. These proteins are activated by posttranslational modification (primarily phosphorylation). Therefore, we investigated the posttranslational modification of some of these key DNA damage response proteins after manumycin treatment.
The activation of BRCA1 by manumycin was evaluated by examining the relative mobility shift of the protein after posttranslational modification (phosphorylation) that activated the protein. In both KAT-4 and ARO cells, manumycin (54 µmol/L) induced a mobility shift of BRCA1 (Fig. 3A) characteristic of phosphorylated BRCA1. Mobility shift was not observed in the paclitaxel (22 µmol/L)-treated cells. The addition of paclitaxel did not affect the mobility shift of BRCA1 induced by manumycin.
|
NBS1 and BRCA1 nuclear foci were detectable by immunofluorescence microscopy in manumycin-treated ARO cells (Fig. 3B) and KAT-4 (data not shown). These findings of BRCA1 and NBS1 posttranslational modification and nuclear foci formation after exposure to manumycin, but not paclitaxel, suggested that manumycin induced DNA damage and that manumycin treatment affected key components of the HR repair pathway.
NHEJ differs from HR in that in NHEJ there is no requirement for an undamaged partner or for extensive homology between the two recombining ends. For NHEJ in mammalian cells, a key player in the initial recognition of the DNA lesion is DNA-PK (26), a member of the PIKK family of kinases. DNA-PK comprises a large catalytic subunit (DNA-PKcs) of 450 kDa and two smaller Ku subunits (Ku70 and Ku80). As shown in Fig. 3A, we observed posttranslational modification manifesting as a mobility shift of DNA-PK in manumycin-treated cells but not paclitaxel-treated cells. This result suggested that this key component of the NHEJ pathway was affected by manumycin. It was previously reported that Ku-dependent autophosphorylation of DNA-PKcs at Thr2609 was required for the repair of ionizing radiation–induced DSBs by NHEJ (41). Whether the DNA-PKcs phosphorylation caused by manumycin is autophosphorylation at this site remains to be tested. Nevertheless, these data together suggested that some key components of the HR and NHEJ pathways might be involved in the responses to DNA damage induced by manumycin.
ATM and DNA-dependent protein kinase, but not ATR, mediated the manumycin-induced DNA damage responses. Next, we explored which DNA damage sensors were responsible for initiating the DNA damage response upon manumycin treatment. Members of the PIKK family, including DNA-PK, ATM, and ATR, are central to the entire DNA damage response (26, 37, 42). Therefore, we investigated whether PIKKs were required for manumycin-induced H2AX phosphorylation. Pretreatment and continued presence of wortmannin, which inhibits PIKK family members, at concentrations as low as 1 µmol/L, reduced manumycin-induced phosphorylation of H2AX (Fig. 4A) by a large extent in both KAT-4 and ARO cells. The reduction of H2AX phosphorylation by wortmannin suggested that the DNA damage response to manumycin required PIKKs. It is known that wortmannin is a relatively potent inhibitor of DNA-PK at 3.6 µmol/L, of ATM at 5.8 µmol/L, and of ATR at >100 µmol/L (43), suggesting that ATR activity is significantly less sensitive than DNA-PK or ATM to wortmannin. ATM and ATR exhibit similar sensitivities to caffeine, a purine analogue, with 50% inhibition of kinase activity at concentrations (IC50) of 0.2 and 1.1 mmol/L, respectively (44). In contrast, DNA-PK is relatively resistant to caffeine (IC50, 10 mmol/L; ref. 44). Our results (Fig. 4A) showed that caffeine slightly inhibited H2AX phosphorylation at 3 mmol/L and inhibited H2AX phosphorylation by >50% at 10 mmol/L. Our experimental results with wortmannin suggested that the functions of ATM and DNA-PK, but not ATR, were important in the DNA damage responses after manumycin treatment.
|
Immunofluorescence microscopy showed that manumycin induced H2AX-containing nuclear foci in ATM+/+ cells and DNA-PK wild-type cells but not in ATM–/– cells and DNA-PK mutants (Fig. 4D). The presence of DSBs in ATM–/– and DNA-PK mutant cells (MO59J) after manumycin treatment was confirmed by neutral Comet assay (data not shown) using a standard protocol with a kit obtained from R&D Systems (Minneapolis, MN). The induction of kinase-dead ATR by doxycycline did not affect the formation of H2AX-containing nuclear foci after manumycin treatment in U2OS cells (Fig. 4D). Taken together, these results suggested that ATM and DNA-PK, but not ATR, were the sensors and/or mediators of the DNA damage response upon manumycin treatment.
Manumycin-induced DNA damage via reactive oxygen species generation. Given that manumycin induced ROS generation and that ROS could damage DNA, we evaluated the role of ROS in mediating manumycin-induced DNA damage by using the reducing agent N-acetyl-L-cysteine (NAC) to quench ROS. The thiol NAC is a source of cysteine for the synthesis of the endogenous antioxidant glutathione (GSH); thus, NAC is postulated to protect cells by increasing intracellular GSH levels and scavenging ROS (45). We measured the generation of NO using a flow cytometry method as described earlier. In both ARO and KAT-4 cells, manumycin (54 µmol/L) for 6 hours increased the average green fluorescence (i.e., NO production), but NAC coadministered with manumycin completely blocked the increase in NO production. When a cutoff value of fluorescence was defined such that 95% of the untreated control cells were below the cutoff value, the percentage of cells stained positive (higher fluorescence than the cutoff value) by the NO sensor dye was 85% in the KAT-4 cells (Fig. 5A), and NAC blocked the increase in percentage of cells stained positive for NO. Similar results were observed with ARO cells (data not shown). Manumycin-induced increase in superoxide anion levels (as measured by ethidium fluorescence after binding with mitochondrial DNA; ref. 29) was also blocked by NAC in both ARO and KAT-4 cells (Fig. 5B).
|
The role of NO in ROS-induced DNA DSBs was further investigated using deferoxamine. In the presence of deferoxamine (5 mmol/L), NO generation was selectively blocked after treatment with manumycin (Fig. 5F) without blocking generation of superoxide (data not shown). In the presence of deferoxamine, manumycin still induced phosphorylation of H2AX, but not as much as in the absence of deferoxamine. This suggested that NO played a partial role but was not absolutely required in the DNA damage after manumycin treatment.
Quenching reactive oxygen species antagonized the inhibitory effect of manumycin on protein biomass. To evaluate the overall importance of ROS generation by manumycin in the antineoplastic effect against anaplastic thyroid cancer cells, we measured to what degree NAC (5 mmol/L) could reverse the inhibitory effect of manumycin (54 µmol/L, 48 hours) on protein biomass in KAT-4 and ARO cells. Protein biomass (relative to control) was measured by sulforhodamine B staining. NAC blocked most of the manumycin-induced decrease in protein biomass in ARO cells and completely blocked the decrease in protein biomass in KAT-4 cells (Fig. 6). These findings indicated that the generation of ROS played an important role in the cytotoxic effect of manumycin.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our results are consistent with the observation of the ROS induction by FTIs in chronic leukemia cells by Selleri et al. (20). ROS generation induced by FTIs may be a common phenomenon because this has been observed for five different FTIs in eight solid tumor cell lines and three leukemia cell lines (20).5
Inhibiting expression of farnesyltransferase by targeting FNTB transcription with siRNA was not sufficient to induce ROS production. Therefore, the ability of FTIs to induce ROS was probably not a direct result of farnesyltransferase inhibition. Although the amounts of ROS induced by different FTIs were attenuated in FNTB-silenced cells, this attenuation was small in magnitude when comparing with 90% transcriptional suppression in FNTB by siRNA. Moreover, the time course of changes documented in this report would precede the expected time course for a mechanism mediated through inhibition of protein farnesylation because for all but the farnesylated proteins with the shortest half-lives, 3 hours would be insufficient to produce significant effects on protein farnesylation. Taken together, ROS induction by FTIs is largely independent of farnesyltransferase inhibition. This complex relationship of farnesyltransferase and FTI-induced ROS production is yet to be defined.
Our results validated a link between the FTI-induced ROS and the manumycin-induced DNA damage. We assessed DSBs by PFGE and -H2AX. PFGE is a classic assay, whereas -H2AX is a recently established approach for assessing DSBs. Both PFGE and -H2AX (immunoblots and immunofluorescence results) provided clear evidence of DSBs following manumycin exposure (Fig. 2A-D). Quenching ROS with NAC significantly suppressed manumycin-mediated DSB formation (Fig. 5) and decreased protein biomass (Fig. 6). The involvement of ATM and DNA-PK in the response to manumycin (Fig. 4A-D) also supported the existence of DSBs. Although manumycin induced the generation of NO and superoxide, we identified that the selective blocking of NO generation by deferoxamine only slightly attenuated the H2AX phosphorylation by manumycin. This suggested that NO is not absolutely required for DNA DSBs but contributed to DSBs. It has been reported that peroxynitrite (ONOO–), formed by a fast reaction of NO with superoxide in cells simultaneously producing NO and superoxide, is a powerful agent of DNA damage (46). Thus, NO might enhance the DNA-damaging effects of superoxide after manumycin treatments.
Our results did not support direct damage of DNA by manumycin. Direct incubation of DNA with manumycin was not sufficient to cause DNA damage detectable by the 3D assay (Fig. 2E). Further, the blocking of caspases did not prevent manumycin-induced DSBs; in other words, DSB formation was not caused by caspase activation as a result of apoptosis (Fig. 2A and C). In contrast, NAC blocked manumycin-induced DSBs, indicating that DSB formation by manumycin depended on ROS.
Our data showed that ATM and DNA-PK, but not ATR, were involved in sensing the ROS-induced DNA damage. Two lines of evidence supported this. First, wortmannin pretreatment remarkably diminished the manumycin-induced DSBs (assayed by phosphorylation of H2AX; Fig. 4A) in a dose-dependent manner. ATM and DNA-PK are relatively sensitive to wortmannin, whereas ATR is relatively resistant to wortmannin. Second, induction of kinase-dead ATR cells with doxycycline did not affect manumycin-induced DSB formation (Fig. 4C and D), whereas inhibition of ATM and DNA-PK function significantly reduced the manumycin-induced DSB formation (Fig. 4B and D). However, it remains unclear why both ATM and DNA-PK are required. One possibility is that mutual interaction between ATM and DNA-PK contributes to the initiation of DNA damage response.
We also revealed that BRCA1 and NBS1 are involved in the DNA damage response to manumycin. Brca1 is a tumor suppressor gene that is mutated at a high frequency in hereditary breast and ovarian cancers (47). Following exposure to ionizing radiation, hydroxyurea, or UV irradiation, BRCA1 undergoes phosphorylation and forms nuclear foci (42). Association of BRCA1 with the DNA repair protein RAD51 further shows that BRCA1 participates in a pathway (or pathways) associated with HR. BRCA1 is the central part of a large multisubunit protein complex of tumor suppressors, DNA damage sensors, and signal transducers known as the BRCA1-associated genome surveillance complex (48). Among the DNA repair proteins in the complex are ATM, BLM, MSH2, MSH6, MLH1, the RAD50-MRE11-NBS1 complex, and the RFC1-RFC2-RFC4 complex. BRCA1, BLM, and the RAD50-MRE11-NBS1 complex colocalize to large nuclear foci after DNA damage. It has been suggested that the BRCA1-associated genome surveillance complex might serve as a sensor of abnormal DNA structures, a regulator of DNA repair, or both (48). Our data showed that BRCA1 phosphorylation and foci formation occurred after manumycin treatment (Fig. 3A and B). Similarly, NBS1 phosphorylation on Ser343 was also observed in manumycin-treated cells (Fig. 3A and B). BRCA1 has been reported to modulate response to genotoxic chemotherapy. Reconstitution of wild-type BRCA1 in BRCA1-mutant cells (HCC 1937) via transfection resulted in a reduced level of apoptosis after treatment with cisplatin, bleomycin, or etopside (49). In addition, BRCA1-deficient mouse fibroblast cells were more sensitive to irinotecan, topotecan, etopside, doxorubicin, carboplatin, or oxaliplatin than BRCA1-proficient cells (50). Therefore, our findings may imply that silencing the repair proteins BRCA1 and NBS1 may enhance the cytotoxicity of manumycin in tumor cells.
It has been shown that RhoB is one of the important mediators of FTI effect on apoptosis and proliferation inhibition (51). The essential role of RhoB has been validated by the knockout mouse studies (7). Geranylgeranylated RhoB may be a mediator of the effect of FTI on apoptosis and growth inhibition (6). Consistent with the other studies that DNA damage agents caused elevation of RhoB level (21, 22), we found that RhoB protein was increased after treatment with FTIs. Quenching ROS did prevent induction of RhoB by the FTIs, suggesting that RhoB was induced by FTI-induced ROS-mediated DNA damage. Linking FTI-induced DNA damage to RhoB added another section to the web of mechanism of action of FTIs.
In summary, FTIs induced ROS generation. ROS-mediated DNA damage led to formation of DSBs. DSB responses were then activated; ATM and DNA-PK but not ATR were the DNA damage sensors activated by manumycin. RhoB, an important mediator of FTI antineoplastic action, was induced as part of the DNA damage responses. This is the first report of DNA damage induced by FTIs via generation of ROS. Based on the findings in this paper and studies of other investigators, we proposed a model showing the relationship of FTI, ROS, DNA damage, DNA damage responses, and RhoB in the induction of growth inhibition and apoptosis of cancer cells (Fig. 7B). Future work will evaluate DNA repair mechanisms as potential therapeutic targets for combination therapy with manumycin. For instance, FTIs may be tested in combination with DNA repair inhibitors. The mechanism-based rational design of combination cancer chemotherapy might lead to improved survival of cancer patients.
|
Acknowledgments |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank Dr. Y. Shiloh for providing the ATM–/– and ATM+/+ cell lines, Dr. S.L. Schreiber for providing the ATR-kinase-dead doxycycline-inducible cell line, Dr. K. Ain for providing the KAT-4 cells, and Dr. J. Fagin for providing the ARO cells.
|
Footnotes |
|---|
5 Pooling our data in this paper, our unpublished data, and the data by Selleri et al. 
Received 8/ 2/04. Revised 2/10/05. Accepted 2/24/05.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. Lampiasi, A. Azzolina, N. D'Alessandro, K. Umezawa, J. A. McCubrey, G. Montalto, and M. Cervello Antitumor Effects of Dehydroxymethylepoxyquinomicin, a Novel Nuclear Factor-{kappa}B Inhibitor, in Human Liver Cancer Cells Are Mediated through a Reactive Oxygen Species-Dependent Mechanism Mol. Pharmacol., August 1, 2009; 76(2): 290 - 300. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Sun, X. Xu, Y. Wang, X. Shen, Z. Chen, and J. Yang Mycoplasma pneumoniae Infection Induces Reactive Oxygen Species and DNA Damage in A549 Human Lung Carcinoma Cells Infect. Immun., October 1, 2008; 76(10): 4405 - 4413. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Kiziltepe, T. Hideshima, L. Catley, N. Raje, H. Yasui, N. Shiraishi, Y. Okawa, H. Ikeda, S. Vallet, S. Pozzi, et al. 5-Azacytidine, a DNA methyltransferase inhibitor, induces ATR-mediated DNA double-strand break responses, apoptosis, and synergistic cytotoxicity with doxorubicin and bortezomib against multiple myeloma cells Mol. Cancer Ther., June 1, 2007; 6(6): 1718 - 1727. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Pan, A. Quintas-Cardama, H. M. Kantarjian, C. Akin, T. Manshouri, P. Lamb, J. E. Cortes, A. Tefferi, F. J. Giles, and S. Verstovsek EXEL-0862, a novel tyrosine kinase inhibitor, induces apoptosis in vitro and ex vivo in human mast cells expressing the KIT D816V mutation Blood, January 1, 2007; 109(1): 315 - 322. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Yu, F. Wan, S. Dutta, S. Welsh, Z. Liu, E. Freundt, E. H. Baehrecke, and M. Lenardo Autophagic programmed cell death by selective catalase degradation PNAS, March 28, 2006; 103(13): 4952 - 4957. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. D. Basso, P. Kirschmeier, and W. R. Bishop Thematic review series: Lipid Posttranslational Modifications. Farnesyl transferase inhibitors J. Lipid Res., January 1, 2006; 47(1): 15 - 31. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Pan and S.-C. J. Yeung Recent Advances in Understanding the Antineoplastic Mechanisms of Farnesyltransferase Inhibitors Cancer Res., October 15, 2005; 65(20): 9109 - 9112. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
indicates an inhibitory effect. RhoB-F, farnesylated RhoB; RhoB-G, geranylgeranylated RhoB.