Tamoxifen accelerates proteasomal degradation of O6-methylguanine DNA methyltransferase in human cancer cells

Abstract

Tamoxifen, a synthetic triphenyl-ethylene compound, is a member of a class of anticancer drugs known as selective estrogen receptor modulators. It may block tumor growth by mimicking estrogen and binding to the estrogen receptors, preventing cancerous growth. Clinical studies have demonstrated that a combination chemo/hormonal therapy regimen with tamoxifen and O⁶-alkylating drugs increased the tumor response rate in cancer patients. The mechanism of action of this combined regimen remains undefined. In this study, we demonstrated that treatment of human colorectal HT-29 carcinoma cells with tamoxifen decreased the repair activity and expression level of O⁶-methylguanine DNA methyltransferase (MGMT) protein in a concentration- and time-dependent manner. This inhibition was also shown in other malignant human cells, regardless of their estrogen receptor status. Furthermore, MGMT inactivation by tamoxifen was associated with a significantly increased susceptibility of cells to 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU). No alteration in MGMT mRNA levels was observed in tamoxifen-treated cells. The half-life of MGMT protein was markedly decreased in the presence of tamoxifen. Tamoxifen-induced MGMT degradation could be blocked by MG-132, a proteasome inhibitor. An increased level of ubiquitinated MGMT protein was found after tamoxifen treatment. We conclude that tamoxifen decreased the MGMT protein level by accelerating protein degradation through the ubiquitin-dependent proteasomal pathway. These findings provide a strong rationale for combined chemo/hormonal therapy with tamoxifen and BCNU in the treatment of human cancers. © 2007 Wiley-Liss, Inc.

Tamoxifen, a synthetic triphenyl-ethylene compound, is a member of a class of anticancer drugs known as selective estrogen receptor modulators (SERMs). It is used to treat patients with all stages of hormone-responsive breast cancer.1 It has also been shown to prevent breast cancer in women who are at high risk for this disease, by mimicking estrogen and binding to the estrogen receptor (ER).2 Although the primary mechanism of action of tamoxifen is believed to be through inhibition of the ER, research over the years has indicated that additional, non-ER-mediated mechanisms exist. These include modulation of signaling proteins, such as protein kinase C (PKC),3, 4 calmodulin,3, 5 transforming growth factor-β3 and insulin-like growth factor I.3, 6 The antitumor effect of tamoxifen is thus believed to be a combination of ER-mediated and non-ER-mediated mechanisms.5 In addition, tamoxifen has been used in the treatment of malignancies other than breast carcinomas, including hepatocellular, colorectal, ovarian, pancreatic, and renal cell carcinomas, malignant gliomas and melanomas.

O⁶-methylguanine-DNA methyltransferase (MGMT), a ubiquitous DNA repair protein, is responsible for removal of alkyl adducts from the O⁶-position of guanine in a reaction that transfers the alkyl group from the DNA to an internal cysteine residue in the MGMT, thus restoring the integrity of the DNA.7, 8 This action inactivates one MGMT protein molecule for each lesion repaired and makes MGMT a suicide protein, because alkylated MGMT is then degraded through the ubiquitin-dependent proteasomal pathway.9 The alkylated base adduct is generated in DNA either endogenously or after exposure to alkylating carcinogens and antitumor drugs with methylating and chloroethylating properties, such as chemotherapeutic 2-chloroethyl-N-nitrosourea (CNU) derivatives [e.g., 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU)]8, 10 and monofunctional triazenes [e.g., dacarbazine (DTIC)].11 Although O⁶-methylguanine is a relatively minor component of methylated base adducts, it is a major mutagenic and carcinogenic lesion because it can pair not only with cytosine but also with thymine during DNA replication, leading to a GC to AT transition mutation.12, 13 The CNUs induce O⁶-chloroethylguanine in DNA, which can ultimately lead to the formation of stable, cytotoxic DNA crosslinks.14 Repair of O⁶-alkylguanine adducts by MGMT is thus important in protecting cells against the mutagenic, carcinogenic and cytotoxic effects of O⁶-alkylating agents.15, 16

The activity of MGMT has been determined in a broad range of human tumors and compared with the corresponding normal tissues, such as brain, colon, ovary, testis and breast.17 Increased expression of MGMT is associated with resistance of tumor cells to DTIC, temozolomide and BCNU.8, 16, 18 Conversely, MGMT-deficient cells are very sensitive to O⁶-alkylating agents.8, 19, 20 Thus, finding ways of controlling MGMT expression, which could enhance the cytotoxicity of O⁶-alkylating agents toward cancer cells, are of significant clinical interest.

Several lines of evidences have demonstrated that tamoxifen could sensitize human cancer cells to O⁶-alkylating agents.21, 22, 23 Various mechanisms underlying the described synergistic antitumor effects of these O⁶-alkylating agents and tamoxifen have been proposed. One possible explanation is that tamoxifen-induced synchronization of ER-positive cells into the G₁ phase of the cell cycle subsequently leads cells to becoming more sensitive to CNUs because CNUs preferentially kill cells in the G₁ phase.24 Other possible explanations include that tamoxifen might inhibit the function of calmodulin and particulate-associated PKC activity.22, 23 Yet, there may exist other possible mechanisms to explain this synergism.

In this study, we demonstrate for the first time that tamoxifen is able to decrease the level of MGMT protein, which subsequently lead to enhanced cytotoxicity of BCNU toward human cancer cells. The mechanism of this decreased level of MGMT protein involves tamoxifen accelerating the degradation of the MGMT protein through the ubiquitin-dependent proteasomal pathway. These results provide a strong rationale for combined chemo/hormonal therapy with tamoxifen and BCNU in the treatment of human cancers.

AP-1, activator protein-1; BCNU, 1,3-bis(2-chloroethyl)-1-nitrosourea; CHX, cycloheximide; CNUs, 2-chloroethyl-N-nitrosourea derivatives; DTIC, dacarbazine; E1, ubiquitin-activating enzyme; E2, ubiquitin-conjugating enzyme; E3, ubiquitin-ligating enzyme; ER, estrogen receptor; ERCC1, excision-repair cross-complementary 1; FBS, fetal bovine serum; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; IP, immunoprecipitation; MGMT, O⁶-methylguanine-DNA methyltransferase; PBS, phosphate-buffered saline; PKC, protein kinase C; RT-PCR, reverse transcription-polymerase chain reaction; SERMs, selective estrogen receptor modulators; t_1/2, half-life; XRCC1, X-ray repair cross complementing 1.

Material and methods

Materials

Tamoxifen and a monoclonal antibody for α-tubulin were purchased from Sigma Chemical Co. (St. Louis, MO). Mouse monoclonal anti-MGMT antibody (clone MT5.1) and anti-excision-repair cross-complementary 1 (ERCC1) antibody (clone 8F1) were purchased from BD Pharmingen (San Diego, CA). Rabbit polyclonal anti-MGMT antibody was purchased from Abcam (Cambridge, UK). Mouse monoclonal anti-X-ray repair cross complementing 1 (XRCC1) antibody (clone 33-2-5) was purchased from NeoMarkers (Fremont, CA). Monoclonal antibodies to ubiquitin and a horseradish peroxidase-conjugated secondary antibody were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Cell culture reagents were obtained from Gibco-BRL Life Technologies (Gaithersburg, MD). N-[³H]-methyl-N-nitrosourea (specific activity 18.6 Ci/mmol) was purchased from Amersham Pharmacia Biotech (Buckinghamshire, UK). All other chemicals were from Merck (Darmstadt, Germany) or Sigma Chemical and were standard analytic grade or higher.

Cell culture

Human colorectal carcinoma HT-29 cells, melanoma A375 cells and H460 nonsmall cell lung cancer cells were maintained in RPMI-1640 medium supplemented with 5% fetal bovine serum (FBS). Human breast carcinoma MCF-7 cells and hepatocellualr carcinoma Hep G2 cells were maintained in MEM medium supplemented with 10% FBS. Human melanoma A2058 cells and breast carcinoma BT-474 cells were maintained in DMEM medium supplemented with 10% FBS. Mycoplasma contamination was routinely monitored in our laboratory, and only mycoplasma-free cultures were used in our study.

Clonogenic survival assay

Cells in logarithmic growth phase were cultured in 6-well plates (250 cells/well) for 24 hr. Next, cells were treated with various concentrations of drugs for the indicated times. Cells were then washed with prewarmed phosphate-buffered saline (PBS) twice and maintained in a drug-free complete medium for 9–12 days. At the end of the incubation period, cells were fixed and stained with 50% ethanol containing 0.5% methylene blue for 30 min, and then washed with water. The number and size of methylene blue-stained colonies were then recorded. The assays were carried out in triplicate. Data were expressed as means ± standard deviations. Student's t-test was used to compare the mean of each combined group with that of the BCNU-alone group. We considered p values of <0.05 to be statistically significant.

MGMT activity assay

MGMT activity was measured by transfer of [³H]-labeled methyl groups from the O⁶-position of guanine in DNA to the MGMT protein, as described previously.25 Cell extracts were prepared by sonication (5 cycles for 15 sec) in MGMT assay buffer (50 mM Tris-HCl (pH 8.3), 1 mM EDTA, 1 mM dithiothreitol and protease inhibitor mixture (1 mM PMSF, 1 μg/mL pepstatin and 50 μg/mL leupeptin)) followed by centrifugation (10,000g, 10 min). Extracts (400 μg protein) were supplemented with [³H]-DNA enriched for O⁶-methylguanine [200 μg; 5,000 cpm (1 pmol)] incubated for 60 min at 37°C. Reactions were quantitated after acid hydrolysis of the DNA substrate, collection of protein precipitates and radioactivity counting.

Reverse transcription-polymerase chain reaction

Total RNA was extracted from HT-29 cells using TRIZOL reagent (Invitrogen, Carlsbad, CA, U.S.A) according to the manufacturer's protocol. RNA was reverse-transcribed using SuperScript II Rnase H reverse transcriptase (Invitrogen) according to the manufacturer's instructions. Human MGMT cDNA was amplified using the following primer pairs: forward primer 5′-AAGGATCCCCGTTTGCGACTTGGTACTT-3′ and reverse primer 5′-CGACGATATCAAGCGGCCGCCCGATGCAGTGTTACACG-3′. PCR amplification was performed under the following conditions: preincubation was performed at 94°C for 2 min, followed by 30 cycles of 94°C (30 sec), 64°C (45 sec), and 72°C (1 min), with a final extension at 72°C for 7 min. Human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA was used as an internal control. The amplified DNA sizes were 704 bp for MGMT and 450 bp for GAPDH. The PCR products were analyzed by electrophoresis on 1% agarose gels and visualized by ethidium bromide staining under UV light.

Northern blot analysis

Total RNA was isolated from HT-29 cells by the TRIZOL RNA isolation method (Invitrogen). RNA (20 μg per lane) was subjected to 1.2% agarose formaldehyde gel electrophoresis and transferred to a Hybond N⁺ nylon membrane. Membranes were UV crosslinked to immobilize the RNA. MGMT or GAPDH probes were labeled with ³²P using the random primer labeling kit (Stratagene, La Jolla, CA). For prehybridization, membranes were placed in Quick hybrid solution (GE Healthcare Bio-Sciences Corp., Piscataway, NJ) at 65°C for 1–2 hr. Probes were added and hybridized to RNA overnight. Membranes were then washed 3 times with a solution of 2× SSC, 0.1% SDS at room temperature, and washed 3 times again with a solution of 1× SSC and 0.05% SDS at 50°C. Wrapped membranes were exposed to X-OMAT film at −70°C. The expression mRNA level of MGMT was calculated as the ratio of the radioactivity in these bands relative to that of the GAPDH bands.

Immunoprecipitation

Cells were initially seeded at a density of ∼2–4 × 10⁶ in 150-mm² dishes. After treatment for the indicated times with the drugs of interest, adherent cells were washed twice with PBS, gently scraped from the dishes, centrifuged, lysed in ice-cold lysis buffer (30 mM Tris (pH 7.5), 1 mM EDTA, 2.5 mM sodium fluoride, 5 mM sodium pyrophosphate, 1% NP40, 0.5 mM dithiothreitol, 1 mM sodium vanadate, 1 mM leupeptin, 1 mM pepstatin, 1 mM PMSF and 10 mM N-ethylmaleimide) with sonication (5 cycles for 15 sec on ice, tune for mininum: 20, amplitude: 40, Vibracell™, SONICS & MATERIALS, Danbury, CT), followed by centrifugation at 4°C (12,000g, 20 min). The resulting supernatant was used as the protein source. Protein concentrations were measured using the BCA protein assay kit (Pierce Biotechnology, Rockford). For immunoprecipitations (IPs), cell extracts (1,000 μg of total protein in 250 μL) were precleaned with protein A/G PLUS-Agarose (Santa Cruz Biotechnology) at 4°C for 30 min, then centrifuged (2,000g, 1 min) and the supernatant was transferred to a fresh tube and further incubated with MGMT antibodies (rabbit, polyclonal; Abcam) at a concentration of 2 μg/mL and mixed overnight at 4°C. Protein A/G Sepharose was added to each tube and they were incubated for a further 4 hr at 4°C. The immunoprecipitates were washed 3 times with Tris-buffered saline and suspended in nonreducing SDS gel loading buffer following subjection to Western blot analysis.

Western blot analysis

Crude cellular extracts were prepared for Western blot analysis as described previously.26 Immunoreactive signals were detected using the Western Blot Chemiluminescent Reagent Plus (Perkin Elmer Life Sciences, Boston, MA). Band-specific intensity was quantitated using an AlphaImager 2000 system.

Results

Decreased level of MGMT protein and activity following tamoxifen treatment

Colorectal carcinoma HT-29 cells are proficient for MGMT. To assess the effect of tamoxifen on MGMT activity and protein level, HT-29 cells were treated with various concentrations of tamoxifen for the indicated times and subjected to Western blot and MGMT activity analysis. As shown in Figure 1a, tamoxifen reduced the activity of MGMT in HT-29 cells in a concentration-dependent manner. Similar to MGMT activity, tamoxifen provoked a decrease in MGMT protein level as shown by Western blot analysis (Fig. 1b). However, there was no change in the expression level of DNA repair proteins ERCC1 and XRCC1 after tamoxifen treatment (Fig. 1b). In addition, a time-dependent decrease in the level of MGMT protein in cells treated with tamoxifen was noted (Fig. 1c). The MGMT protein level began to decline in cells treated with 2.5 μM tamoxifen for 24 hr, and was further reduced to a minimum after treatment with 5 μM tamoxifen for 72 hr (Fig. 1c).

Figure 1. Effects of tamoxifen on the expression level and functional activity of MGMT in HT-29 cells. (a) Dose effect of tamoxifen on regulation of MGMT functional activity. Cells were treated with various concentrations of tamoxifen for 24 hr prior to the measurement of MGMT functional activity. Bars depict the means of triplicates from 2 independent experiments (±S.D). (b) Dose effect of tamoxifen on regulation of MGMT protein level. Cells were treated with various concentrations of tamoxifen for 24 hr. Cell lysates were prepared and subject to Western blot for MGMT, ERCC1, XRCC1 and α-tubulin. (c) Time effect of tamoxifen on regulation of MGMT protein level. Percent of MGMT protein level was expressed as the mean ±S.D. of the 3 independent experiments.

Enhancement of BCNU cytotoxicity toward HT-29 cells by tamoxifen

We next used clonogenic survival assay to evaluate whether the decreased level of MGMT protein impacts the cellular sensitivity to BCNU. For this purpose, we determined the expression level of MGMT protein in several human cancer cells. In contrast to HT-29 cells that are proficient in MGMT expression, melanoma A2058 cells are deficient in MGMT expression (data not shown). Thus, we selected HT-29 and A2058 cells for this study. HT-29 cells were treated with various concentrations of tamoxifen for either 24 hr (schedule 1) or 48 hr (schedule 2), before treatment with different concentrations of BCNU for 1 hr. As shown in Figure 2a, in cells pretreated with 2.5 μM tamoxifen, there was significantly enhanced cytotoxicity of BCNU in HT-29 cells (p < 0.05). This enhanced effect was more pronounced with schedule 2 than schedule 1. The addition of 2.5 μM tamoxifen caused a reduction in BCNU LC₅₀ values from 64 to 47 μM (schedule 1) and 24 μM (schedule 2) in MGMT-proficient HT-29 cells (Table I). Similar result was obtained with another MGMT-proficient cell line, A375 melanoma (data not shown). This finding indicated that tamoxifen enhanced BCNU cytotoxicity in MGMT-proficient cells. There was no alteration in BCNU sensitivity after tamoxifen treatment in MGMT-deficient A2058 melanoma cells (Fig. 2b), the BCNU LC₅₀ values were about 10 μM in both of single agent or combination treatment (Table I).

Figure 2. Cytotoxic effect of combined tamoxifen with BCNU against MGMT-proficient and -deficient cells. (a) Enhancement of BCNU cytotoxicity toward MGMT-proficient HT-29 cells by tamoxifen. (b) No alterattion of BCNU cytotoxicity toward MGMT-null A2058 cells. In brief, cells were treated with 2.5 μM tamoxifen for 24 hr (schedule 1) or 48 hr (schedule 2) before treatment with different concentrations of BCNU for 1 hr. Cells were then grown in fresh medium for 9–12 days, and survival was tested by clonogenic survival assay. Colonies (>50 cells) were counted manually and expressed as percentage of survival in number of colonies relative to control. Result was expressed as the mean ± SD of the 3 independent experiments. *Significantly different from the BCNU-alone group at p < 0.001 with Student's t-test.

Table I. LC₅₀ Values of BCNU on Clonogenic Survival Assays in both MGMT-Proficient and -Deficient Human Cancer Cells

^a For this experiment, tamoxifen-alone only produced a rather weak colony forming inhibition (≥80% colony forming ability was obtained when 2.5 μM tamoxifen was present in both HT-29 and A2058 cells for 24 and 48 hr). Therefore, we used the concentration of tamoxifen up to 2.5 μM to evaluate the combined effect of tamoxifen and BCNU. ^b Cells were treated with 1.25 or 2.5 μM tamoxifen for 24-hr (schedule 1) or 48 hr (schedule 2) before treatment with various concentrations of BCNU for 1 hr. Cell viability was determined by clonogenic survival assay. The LC50 value resulting from 50% inhibition of colony forming ability was calculated. Each value represents the mean ± SD of 3 independent experiments. * Significantly different from the BCNU-alone group at p < 0.001 with Student's t-test.
Treatments
BCNU		+	+	+
Tamoxifen (μM)		0	1.25	2.5
Schedule 1
HT29 (colorectal)	MGMT-proficient	64.8 ± 1.4	55.7 ± 2.5*	47.4 ±3.5*
A2058 (melanoma)	MGMT-deficient	10.9 ± 2.7	10.4 ± 3.1	9.2 ± 2.5
Schedule 2
HT29 (colorectal)	MGMT-proficient	64.6 ± 1.8	50.8 ± 3.8*	23.8 ±2.7*
A2058 (melanoma)	MGMT-deficient	10.2 ± 2.9	9.7 ± 2.7	9.0± 3.5

Reduction in MGMT protein level by tamoxifen is found in both ER-negative and -positive cells

Because tamoxifen is a SERM, we investigated whether β-estradiol could also decrease MGMT expression. As shown in Figure 3a, the levels of MGMT remained unchanged in HT-29 cells treated with 25 or 50 μM β-estradiol. Indeed, there was only a 10% decrease in MGMT level in cells treated with 100 μM β-estradiol. Furthermore, we found that the synthetic glucocorticoid hormone, dexamethasone, increased MGMT expression levels in a concentration-dependent manner in HT-29 cells (Fig. 3a).

Figure 3. (a) Effect of hormonal agents on MGMT protein level of HT-29 cells. Cells were treated with various concentrations of tamoxifen, β-estradiol and dexamethasone for 24 hr. MGMT expression was examined by Western blot analysis. (b) Relationship between ER status and tamoxifen induced the decrement of MGMT protein level. ER-positive MCF and BT-474 cells, and ER-negative HT-29, A375, Hep G2 and H460 cells were treated with various concentrations of tamoxifen for 24 hr. The expression level of MGMT protein was examined by Western blot analysis. α-Tubulin has been used as internal control, and the result demonstrated that there is no change in the expression level of internal control after treatment with tamoxifen (data not shown).

We further examined the effect of tamoxifen on the regulation of MGMT protein levels in ER-positive and ER-negative human cancer cells. For this purpose, we chose two ER-positive cells, including breast carcinoma MCF-727 and BT-474,28 and 4 ER-negative cells, including HT-29 (colorectal carcinoma),27 A375 (melanoma),29 Hep G2 (hepatocellular carcinoma)30 and H460 (nonsmall cell lung cancer)31 on this study. As shown in Figure 3b, tamoxifen was able to decrease MGMT protein levels in both ER-negative (HT-29, A375, Hep G2 and H460) and ER-positive (MCF-7 and BT-474) cells (Fig. 3b).

No change in MGMT mRNA levels in tamoxifen-treated cells

We investigated whether the tamoxifen-induced decrease in MGMT protein levels and activity could be the result of decreased levels of the corresponding mRNA. Reverse transcription-polymerase chain reaction (RT-PCR) and Northern blot analysis were performed with total RNA extracted from HT-29 cells immediately after the end of tamoxifen treatment. The result demonstrates that no change was seen in the MGMT mRNA level in cells treated with concentrations of tamoxifen up to 25 μM (Fig. 4a, RT-PCR data; Fig. 4b, Northern blot result).

Figure 4. RT-PCR and Northern blot analysis of MGMT mRNA levels in HT-29 cells after tamoxifen treatment. Cells were exposed to various concentrations of tamoxifen for 24 hr. MGMT mRNA levels were evaluated at the end of treatment. (a) RT-PCR analysis of MGMT mRNA expression in HT-29 cells. Total RNA isolated from each group was subjected to semiquantitative RT-PCR as described in methods. Human GAPDH cDNA was used as an internal control. (b) Northern blot analysis of expression of MGMT mRNA in HT-29 cells. Total RNA was isolated from each group and probed with ³²P-labeled MGMT cDNA. GAPDH hybridization was used as a loading control. Result shown is a representative data from 2 independent experiments.

Increased MGMT protein degradation in tamoxifen-treated cells

Because no change was seen in the level of MGMT mRNA in tamoxifen-treated cells, we investigated protein stability and measured the half-life (t_1/2) of the MGMT protein in tamoxifen-treated cells. To determine the change of MGMT protein stability, cells were treated with cycloheximide (CHX) to prevent new protein synthesis in presence or absence of tamoxifen. As shown in Figure 5a, in contrast to a slight decrease in the MGMT protein level in cells treated with CHX alone, cells treated with CHX in the presence of 5 μM tamoxifen showed accelerated degradation of MGMT protein in a time-dependent manner (Fig. 5a). The t_1/2 of MGMT in the absence or presence of tamoxifen was 35 and 23 hr, respectively (Fig. 5b).

Figure 5. Degradation of MGMT protein after incubation with tamoxifen. (a) kinetics of MGMT degradation in HT-29 cells treated with vehicle control (left panel) or 5 μM tamoxifen (right panel). Cells were treated with 50 μg/mL cycloheximide alone or cotreated with 5 μM tamoxifen for the indicated times. Cell lysates were prepared and subjected to Western blot for MGMT and α-tubulin. (b) Graphical representation of the relative band intensities of MGMT protein. The percentage of initial intensities of MGMT protein in cell extracts normalized with respect to α-tubulin amounts was plotted against times. Data were expressed as the mean ±SD of the 3 independent experiments.

Tamoxifen accelerates MGMT degradation via the ubiquitin-dependent proteasomal pathway

We next examined whether the reduced half-life of MGMT protein in the presence of tamoxifen was due to proteolytic degradation by proteasomes. CHX was used to block protein synthesis to confirm the involvement of proteasomes in the tamoxifen-induced acceleration of MGMT degradation. It has been shown that proteasome inhibitors possess not only proteasome inhibition, but also induced apoptotic cell death. Therefore, to evaluate the role of proteasome inhibitors on tamoxifen-induced MGMT degradation, high concentration of tamoxifen (25 μM) and short treatment period (0–12 hr) were applied in order to exclude killing effects. As shown in Figure 6, tamoxifen accelerated MGMT protein degradation in a time-dependent manner (Fig. 6a, central panel). The decreased stability of the MGMT protein induced by tamoxifen could be blocked by MG-132, a cell-permeable proteasome inhibitor (Fig. 6a, right panel).

Figure 6. Tamoxifen-induced MGMT degradation is mediated by an ubiquitin-dependent proteasomal pathway. (a) Involvement of proteasomes in the tamoxifen-induced acceleration of MGMT degradation. HT-29 cells were treated with 50 μg/mL cycloheximide in presence or absence MG-132 for 20 min. Then the cells were treated with 25 μM tamoxifen for an additional 3, 6, 9 and 12 hr. Cell lysates were prepared and subjected to Western blot for MGMT and α-tubulin. (b) Time course of MGMT ubiquitination induced by tamoxifen. Cells were first treated with MG-132 and then added tamoxifen for 1, 3, 6, 9 and 12 hr. Cell lysates were prepared and subjected to Western blot with the mouse monoclonal anti-ubiquitin (left panel) and anti-MGMT antibody (right panel). (c) Tamoxifen induced formation of ubiquitin-MGMT conjugates in HT-29 cells. Cells were in presence or absence 25 μM tamoxifen for 6 hr. Normalized amounts of lysates were processed for immunoprecipitation (IP) by using 1 μg/mL of rabbit polyclonal anti-MGMT antibody followed by Western blot with the mouse monoclonal anti-ubiquitin (left panel) and anti-MGMT antibody (right panel).

To examine the mechanism by which the ubiquitin-dependent proteasomal pathway is involved in the tamoxifen-induced regulation of MGMT degradation, we evaluated the effects of tamoxifen on cellular ubiquitination in general. To do that, it was necessary to stabilize ubiquitinated proteins by using MG-132, because ubiquitinated proteins would normally be immediately degraded by proteasomes. We determined a time-course effect to estimate the changes in MGMT ubiquitination. The result showed that very low band intensities of ubiquitinated proteins were observed in the absence of MG-132, whereas increased band intensities of ubiquitinated proteins were observed in the presence of MG-132 in tamoxifen-treated cells in a time-dependent manner (Fig. 6b, left panel). The image in the right panel of Figure 6b shows the reprobing for MGMT using the same membrane after stripping the ubiquitin antibody. We found that multiply conjugated high-molecular mass species could be recognized by the MGMT antibody also in a time-dependent manner, which appeared rapidly, at 3 hr after treatment.

We then performed IP analysis to evaluate the effects of tamoxifen on the ubiquitination of MGMT. Western blotting (Fig. 6c, left panel), with an antiubiquitin antibody using a membrane on which MGMT immunoprecipitated by the MGMT antibody had been loaded, showed that high-molecular mass species of MGMT protein could be recognized in cells not treated with tamoxifen. Ubiquitinated species of MGMT protein were also recognized by the anti-MGMT antibody (Fig. 6c, right panel). Notably, significant increased in the levels of the ubiquitin-MGMT conjugates were evident in tamoxifen-treated cells (Fig. 6c).

Discussion

Tamoxifen is being used successfully to treat patients with all stages of hormone-reponsive breast cancer.1 It is applied as a monotherapy or in combination with other anticancer agents to treat various kinds of human cancers.5, 21, 22, 23, 32 Accumulating evidence demonstrates that combinations of tamoxifen with O⁶-alkylating agents may improve not only the response rate, but also prolong overall survival times in patients with metastatic melanoma.22, 32 Moreover, results from studies on cell lines demonstrated that combining tamoxifen with CNUs lead to synergistic cytotoxic effects in both ER-positive and -negative cells.21, 23, 33, 34 Various mechanisms, including inhibition of ER signaling,21, 34 modulation of calmodulin function and inhibition of PKC activity,23 have been proposed to explain this synergism. However, there are other possible explanations and the true answer remains unknown.

In the present study, we demonstrated for the first time that tamoxifen can decrease the level of MGMT protein in several malignant human cells, including breast carcinoma MCF-7 and BT-474, colorectal carcinoma HT-29, melanoma A375, hepatocellular carcinoma Hep G2 and nonsmall cell lung cancer H460 cells, regardless of their ER status (Fig. 3b). However, there appear to be differences in the tamoxifen dose responses between the various cell lines, which probably due to intrinsic different sensitivity to tamoxifen in these tested cell lines. The cell line HT-29 is proficient for MGMT repairing guanine O⁶-alkylations.35 Therefore, to specifically evaluate the role of tamoxifen in MGMT inhibition, we chose this line for further study.

Because tamoxifen is a hormone antagonist, we also investigated the role of β-estradiol and dexamethasone in MGMT regulation for comparison. In contrast to the tamoxifen-induced decrease in MGMT protein levels, β-estradiol had almost no impact on the level of MGMT protein (Fig. 3a). Our data also showed that the synthetic glucocorticoid hormone, dexamethasone, could increase the levels of MGMT protein (Fig. 3a), which is in line with previous studies and believed to occur through the activation of glucocorticoid responsive elements in the promoter region of the MGMT gene.36, 37 Thus, the common practice of administering a glucocorticoid together with CNUs for the treatment of certain cancers needs to be reconsidered, because the glucocorticoid may actually reduce the therapeutic efficacy of CNUs. This warrants further analysis in the clinic.

It has been proposed that MGMT is an important determinant of BCNU resistance. Indeed, decreased MGMT activity can lead to enhanced cytotoxicity of BCNU.8, 19, 25, 38-41 We thus examined the combined effect of tamoxifen and BCNU in HT-29 cells. The results show that the level of MGMT protein decreased by 20 and 40% when cells were treated with 2.5 μM tamoxifen for 24 and 48 hr, respectively (Fig. 1c). Consistently, the cytotoxic effect of BCNU is significantly more pronounced in cells pretreated with 2.5 μM tamoxifen for 48 hr than for 24 hr in MGMT-proficient HT-29 cells (Fig. 2a). Our result also showed that tamoxifen is not impact on BCNU-induced cell killing in MGMT-deficient A2058 melanoma cells (Fig. 2b). Notably, Trump et al. reported that tamoxifen at 150 mg/m² given twice a day following a loading dose of 400 mg/m² resulted in plasma levels of tamoxifen of 4 μM, without dose-limiting toxicity.42 These results indicate that tamoxifen at a clinically achievable plasma concentration can enhance BCNU cytotoxicity by decreasing MGMT protein levels (Figs. 1c and 2a).

Previous studies have demonstrated that the promoter of the human MGMT gene contains two putative activator protein (AP)-1 sites and AP-1 involves in regulation of MGMT gene expression. AP-1 activity is a target for PKC-regulated changes in gene expression. Boldogh et al. demonstrated that the PKC activator phorbol-12-myristate-13-acetate increased the level of MGMT mRNA.43 Additionally, the PKC inhibitor 1-(5-isoquinoline sulfonyl)-2-methylpiperazine-HCl eliminated MGMT activation, suggesting the therapeutic significance of PKC-mediated MGMT modulation.43 It has been reported that tamoxifen is a potent PKC inhibitor.4 To gain insights into the mechanism by which tamoxifen decreases MGMT protein level, we thus evaluated the level of MGMT mRNA in tamoxifen-treated cells. Northern and RT-PCR analysis of tamoxifen-treated cells indicated that the levels of MGMT mRNA were not detectably affected by tamoxifen treatment (Figs. 4a and 4b), suggesting that the reduction in the level of MGMT protein in tamoxifen-treated cells was not a consequence of decreased levels of the corresponding mRNA.

Because there was no change in the level of MGMT mRNA in tamoxifen-treated cells, we proposed that loss of MGMT protein might occur through a posttranscriptional mechanism. Indeed, the loss of MGMT protein could result from reduced MGMT protein synthesis and/or accelerated MGMT protein degradation. We thus investigated the stability of MGMT protein and measured its t_1/2 in tamoxifen-treated and -untreated cells. Our data show that there was a slight decline in MGMT protein in cells treated with CHX alone, which is consistent with previous finding that wild-type MGMT protein has a long t_1/2 in cells unexposed to alkylating agents.8, 9 In contrast, the t_1/2 of MGMT protein was reduced after tamoxifen treatment (Fig. 5). These findings suggest that tamoxifen is able to accelerate MGMT protein degradation.

Protein degradation through the ubiquitin-dependent proteasomal pathway has been implicated in cellular protein destruction. In this pathway, a series of ubiquitin moieties become covalently attached to the free amino groups of proteins destined for catabolism in a sequence of ATP-dependent reactions catalyzed by ubiquitin-activating (E1), ubiquitin-conjugating (E2) and ubiquitin-ligating enzymes (E3). The multiubiquitin chain then serves as a reusable recognition signal for selective proteolysis of these target proteins by a large (26S) ATP-dependent proteasome complex.44, 45, 46, 47 MG-132 is a cell-permeable proteasomal inhibitor that specifically blocks the activity of the 26S proteasome. Thus, it causes accumulation of ubiquitinated proteins that would otherwise be degraded by proteasomes.

Accordingly, we preincubated cells with the specific proteasomal inhibitor MG-132 to assess whether the degradation of MGMT accelerated by tamoxifen mediated through the ubiquitin-dependent proteasomal pathway. Our data show that the addition of MG-132 to cells treated with tamoxifen and CHX could almost completely reverse the decrease in the level of MGMT protein (Fig. 6a), implicating that the proteasome pathway is involved in accelerated MGMT degradation induced by tamoxifen.

Because proteasomal degradation of proteins often requires ubiquitination, two mechanisms that may contribute to the degradation of MGMT: (i) tamoxifen induces ubiquitination of the MGMT protein or (ii) ubiquitination occurs in an MGMT-associated manner involving a factor that assists in targeting MGMT to the proteasome for degradation. Our results reveal that tamoxifen is able to induce the accumulation of high-molecular-weight polyubiquitinated forms of MGMT protein (Figs. 6b and 6c), which results in accelerated degradation of the MGMT protein.

The molecular mechanism by which the MGMT protein becomes a target for ubiquitinaiton by tamoxifen is not clear. It is well characterized that ubiquitin–protein ligation requires the sequential action of E1, E2 and E3, and the selectivity of protein degradation is determined mainly at the stage of ubiquitin ligation. Thus, one possibility is that tamoxifen activates an as yet unidentified E3 ubiquitin-protein ligase for MGMT. Niture et al. recently demonstrated that MGMT is associated with F-box proteins as well as proteasome components.48 Thus, further studies are required to address these mechanisms.

In conclusion, our study provides the first evidence that tamoxifen is able to decrease the MGMT level in cancer cells by accelerating protein degradation through the ubiquitin-dependent proteasomal pathway. Furthermore, tamoxifen at a serum concentration that is clinically achievable can enhance the cytotoxicity of BCNU toward cancer cells. These findings provide a rationale for clinical trials in colon cancer with the MGMT modulator, tamoxifen, to increase the therapeutic response to BCNU.