17β-Estradiol Induces Down-Regulation of Cap43/NDRG1/Drg-1, a Putative Differentiation-Related and Metastasis Suppressor Gene, in Human Breast Cancer Cells

Purpose: Cap43 is known as a nickel- and calcium-inducible gene. In the present study, we examined whether 17β-estradiol (E₂) could affect the expression of Cap43 in breast cancer.

Experimental Design: Real-time PCR, immunoblotting, and immunocytochemistry were used to examine the expression of Cap43 and estrogen receptor-α (ER-α) in breast cancer cell lines. MDA-MB-231 and SK-BR-3 cell lines were transfected with ER-α cDNA to establish cells overexpressing ER-α. Immunohistochemistry was used to evaluate the expression of the Cap43 protein in breast cancer patients (n = 96), and the relationship between Cap43 expression and clinicopathologic findings was examined.

Results: Of the eight cell lines, four expressed higher levels of Cap43 with very low levels of ER-α, whereas the other four expressed lower levels of Cap43 with high ER-α levels. Treatment with E₂ decreased the expression of Cap43 dose-dependently in ER-α-positive cell lines but not in ER-α-negative lines. Administration of antiestrogens, tamoxifen and ICI 182780, abrogated the E₂-induced down-regulation of Cap43. Overexpression of ER-α in both ER-α-negative cell lines, SK-BR-3 and MDA-MB-231, resulted in down-regulation of Cap43. Immunostaining studies showed a significant correlation between Cap43 expression and the histologic grade of tumors (P = 0.0387). Furthermore, Cap43 expression was inversely correlated with the expression of ER-α (P = 0.0374).

Conclusions: E₂-induced down-regulation of Cap43 seems to be mediated through ER-α-dependent pathways in breast cancer cells both in culture and in patients. Cap43 has potential as a molecular marker to determine the therapeutic efficacy of antiestrogenic anticancer agents in breast cancer.

Cap43 has been identified as a nickel- and calcium-inducible gene (1). The Cap43, a 43-kDa protein, has three unique 10–amino acid tandem-repeat sequences at its carboxyl terminus and is phosphorylated by protein kinase A (2). The Cap43 gene is identical to the N-myc downstream regulated gene 1 (NDRG1), a human homocysteine-inducible gene (3), and to the differentiation-related gene-1 (4). Cap43 expression is reduced in tumor cells (RTP/Rit42; ref. 5). Furthermore, the expression of Cap43 is markedly influenced by several stimuli, including oxidative stress, metal ions, hypoxia, phorbol esters, vitamins A and D, steroids, homocysteine, and tunicamycin as well as oncogenes (N-myc and C-myc) and tumor suppressor genes (p53 and VHL; refs. 1, 3, 5–10).

Although several studies have elucidated various characteristics of the Cap43, its exact function remains unclear. Cap43 is expressed in various organs, including the prostate, ovary, colon, and kidney, and its expression is dynamically changed during postnatal development in the kidney, brain, liver, and nerves (3, 11–13). These studies, which suggest the involvement of Cap43 in organ maturation and differentiation, have recently prompted detailed investigations of its role in the neuronal system. Cap43 was originally shown to be responsible for Charcot-Marie-Tooth disease type 4D; mutations in Cap43 are commonly identified in this hereditary neuropathy of the motor and sensory systems. Okuda et al. have recently established Cap43 knockout mice that exhibit Schwann cell dysfunction, suggesting that Cap43 is essential for the maintenance of the myelin sheaths in peripheral nerves (14). Consistent with this study, Hirata et al. have reported that Cap43 plays an important role in the terminal differential of Schwann cells during nerve regeneration (15).

In contrast with these studies, Stein et al. have identified Cap43 as a gene that is up-regulated by p53 and have also shown that Cap43 is required for p53-dependent apoptosis, thereby indicating that Cap43 is a p53 target gene (16). Furthermore, Kim et al. have reported that the Cap43 is associated with microtubules in the centrosome and participates in the spindle checkpoint in a p53-dependent manner, suggesting that Cap43 may play a key role in the regulation of microtubule dynamics (17). Overexpression of the Cap43 gene also inhibits growth in colon cancers as well as metastasis in prostate and colon cancer cells in an animal model (18, 19), suggesting that Cap43 suppresses metastasis. Low levels of Cap43 expression in breast cancer cells are closely correlated with poor clinical outcomes (20). Cap43 thus seems to be closely associated with the differentiation and/or malignant states of human cancers.

In the present study, we examine how the expression of the Cap43 gene could be modulated during therapeutic treatment by antiestrogenic drugs and discuss the potential of Cap43 as a molecular target for the therapeutic efficacy of antiestrogenic anticancer agents in breast cancer.

Cells and cell culture. Human breast cancer cells SK-BR-3, MDA-MB-231, T47D, MCF-7, and CRL1500 were obtained from the American Type Culture Collection (Manassas, VA). YMB-1 and OCUB-M were obtained from the Japanese Collection of Research Bioresources (Osaka, Japan) and Riken/Wako (Osaka, Japan), respectively. A tamoxifen-resistant cell line, R-27, was established from MCF-7 (21). All cell lines were grown in phenol red–free McCoy's (PromoCell GmbH, Heidelberg, Germany) and Leibovitz's L-15, RPMI 1640, and α-MEM containing 10% fetal bovine serum, 100 units/mL penicillin G sodium, and 100 μg/mL streptomycin sulfate.

Estrogen receptor-α overexpression in estrogen receptor–negative cell lines. Two estrogen receptor-α (ER-α)–negative cell lines, SK-BR-3 and MDA-MB-231, were transfected with either pRc/CMV vector or ER-α plasmid DNA (a kind gift from Dr. Shin-ichi Hayashi, Department of Molecular Medical Technology, Faculty of Medicine, Tohoku University, Sendai, Japan) using LipofectAMINE Plus (Invitrogen Life Technologies, Inc., Gaithersburg, MD). Three days after transfection, 500 μg/mL G418 disulfate (Nacalai Tesque, Kyoto, Japan) was added to the growth medium. The resulting G418-resistant cells were propagated to generate stable cell lines for use in further studies.

Cell proliferation assay. Cells were plated at 2 × 10⁴ per well in 12-well dishes. After 48 hours, cells were rinsed and incubated for a further 24 hours in same medium supplemented with 5% double charcoal-stripped serum CSS. Cells were then washed with serum-, estrogen-, and phenol red–free medium and then incubated in medium supplemented with 5% CSS in the presence or absence of 17β-estradiol E₂ (Sigma-Aldrich Co., St. Louis, MO) with or without tamoxifen (Calbiochem, La Jolla, CA). The cells were trypsinized and counted at 0, 2, 4, and 6 days after incubation with a coulter counter (Beckman Coulter, Miami, FL).

Immunocytochemistry. Cells were trypsinized and plated on glass coverslips in six-well plates. Then, cells were rinsed with PBS and fixed in 4% paraformaldehyde/PBS for 30 minutes at room temperature. Cells were permeabilized with solution containing 5% bovine serum albumin, 0.2% Triton X-100 in PBS for 90 minutes at room temperature. After 1 hour of blocking with 2% goat serum, the cells were incubated overnight with rabbit polyclonal anti-Cap43 (1:1,000; developed in our laboratory; ref. 10). Cells were then rinsed and incubated with goat anti-rabbit IgG and 1 μg/mL Alexa Fluor 546 (Molecular Probes, Eugene, OR) for 60 minutes at room temperature. Nuclear staining was carried out using 4′,6-diamidino-2-phenylindole (1:1,000, Dojindo, Kumamoto, Japan). Coverslips were mounted on slide glasses using gel mount and viewed using an Olympus BX51 florescence microscope (Olympus, Tokyo, Japan).

Western blot analysis. After treatment with E₂, tamoxifen, ICI 182780 (Nacalai Tesque), or nickel chloride (NiCl₂; Nacalai Tesque), 200 μL lysis buffer [0.2% NP40, 225 mmol/L NaCl, 25 mmol/L Tris (pH 7.4)] was added. The cells were harvested, and the cell slurry was sonicated briefly before centrifugation at 15,000 × g for 15 minutes at 4°C. The supernatant was collected, and 50 μg aliquots of protein were loaded into each well, separated using SDS-PAGE, and transferred to Immobilon membranes (Millipore, Bedford, MA). After transfer, blots were incubated with blocking solution and probed with antibodies. The antibodies used were as follows: rabbit polyclonal antibody directed against Cap43 (produced in our laboratory; ref. 10), rabbit polyclonal anti-ER-α antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), and anti-β-actin (Sigma-Aldrich). The relative expression of each protein was calculated using the NIH Image Analysis Program version 1.62 (NIH, Bethesda, MD).

RNA extraction and cDNA synthesis. Total RNA was extracted using ISOGEN-LS reagent (Nippon Gene, Toyama, Japan) and digested with DNase I (Sigma-Aldrich). Total RNA (1 μg) was then reverse transcribed using random hexamer priming and SuperScript II reverse transcriptase (Toyobo, Osaka, Japan).

Real-time PCR. cDNA (100 ng) was amplified in a real-time PCR using SYBR Green Mix (PE Applied Biosystems, Warrington, United Kingdom) and 200 nmol/L primer for Cap43. The real-time PCR reactions were done in an ABI PRISM Model 7700 Sequence Detector (Applied Biosystems, Foster City, CA) under the following conditions: 50°C for 2 minutes, 95°C for 1 minute followed by 40 cycles at 95°C for 15 seconds and 60°C for 1 minute. The sequences of primers were as follows: Cap43 forward primer 5′-AGGCGGACATTCTGGAAATG-3′ and reverse primer 5′-CGGTACTTCCCCAGCACACTT-3′, pS2 forward primer 5′-TTTGGAGCAGAGAGGAGGCAATGG-3′ and reverse primer 5′-TGGTATTAGGATAGAAGCACCAGGG-3′, and C-myc forward primer 5′-AGGCGCGCGTAGTTAATTCAT-3′ and reverse primer 5′-CGCCCTCTGCTTTGGGA-3′. Glyceraldehyde-3-phosphate dehydrogenase and β-actin were used as reference genes.

Immunohistochemistry. Tissue sections were taken from 96 breast cancer patients who underwent radical surgery (Department of Surgery, Kurume University Hospital, Kurume, Japan) between 1995 and 1999. The 4-μm tissue sections were deparaffinized for 15 minutes at 85°C and the slides were heated in a microwave oven in a CC1 buffer for 60 minutes. The sections were stained using the Benchmark (IHC Automated Systems, Tucson, AZ) with rabbit polyclonal anti-Cap43 (9, 10), anti-ER-α, anti–progesterone receptor (PgR), anti–epidermal growth factor receptor (EGFR), and anti-HER-2 antibody. All antibodies else than Cap43 (produced in our laboratory) were purchased from Ventana Medical Systems (Tucson, AZ). The samples were viewed using an Olympus BX51 florescence microscope.

The extent to which Cap43 proteins were stained in immunohistologic studies was analyzed to compare the strength of Cap43 expression in cancer cells with that in normal glands: cancer cells that expressed Cap43 stronger than normal glands appeared positive for staining and normal glands expressing Cap43 stronger than cancer cells were negative for staining. The extent to which ER-α or PgR proteins were stained in immunohistologic studies was defined by the percentage of cells with strongly stained nuclei: ≥10% defined a gland as positive for ER-α or PgR and ≤9% defined it as negative. The immunohistochemical expression of EGFR and HER-2 was categorized into four groups: score 0, no staining at all or membrane staining in <10% of the tumor cells; score 1+, faint/barely perceptible partial membrane staining in >10% of the tumor cells; score 2+, weak to moderate staining of the entire membrane in >10% of the tumor cells; score 3+, strong staining of the entire membrane in >10% of the tumor cells. The extent of immunohistologic staining for EGFR and HER-2 was defined as follows: scores of 2+ or 3+ were regarded as positive and scores of 0 or 1+ were regarded as negative. The positive cells were counted by two experienced observers who were blinded to the condition of the patients. The relationship between Cap43 and each clinicopathologic finding (age, tumor size, menopausal status, histologic grade, lymph node metastasis, and expression of EGFR, HER-2, ER-α, and PgR) and postoperative survival was examined.

Statistical analysis. The χ² test, Fisher's exact probability test, and Student's t test were used for statistical analyses. In patients undergoing resection, the relationships between Cap43 expression and prognosis were examined by the Kaplan-Meier method (22) and the univariate relationship between prognostic factors and overall survival rate was assessed by the log-rank test. P < 0.05 was regarded as statistically significant.

Comparison of cellular levels of Cap43 and ER-α in human breast cancer. All eight breast cancer cell lines used in this study were screened for Cap43 expression using immunoblotting and Immunocytochemistry (Fig. 1). SK-BR-3, MDA-MB-231, CRL1500, OCUB-M, and R-27 showed relatively high levels of Cap43 protein expression, whereas its expression was decreased in T47D, MCF-7, and YMB-1 (Fig. 1A). ER-α was expressed in T47D, MCF-7, YMB-1, and R-27, but its expression was negligible in SK-BR-3, MDA-MB-231, CRL1500, and OCUB-M. Consistent with the results from Western blot analysis (Fig. 1A), immunocytochemical analysis of the eight cell lines showed the expression of Cap43 in SK-BR-3, MDA-MB-231, CRL1500, OCUB-M, and R-27 (Fig. 1B). A tamoxifen-resistant cell line, R-27, showed a much higher expression of Cap43 protein than its parental counterpart cell line, MCF-7 (Fig. 1A and B). We conducted further studies to examine whether the expression of ER-α in these cell lines, and the addition of E₂, tamoxifen, and ICI 182780, to the culture could modulate the expression of Cap43. Three ER-α-positive (T47D, MCF-7, and R-27) and two ER-α-negative (SK-BR-3 and MDA-MB-231) cell lines were selected for further study.

Cell growth dependence on E₂ and the effect of tamoxifen. We used an E₂-depleted culture condition to examine whether the addition of E₂ to the five breast cancer cell lines tested in this experiment promoted E₂-dependent growth. Previous studies have shown that the growth of human breast cancer cell lines depends on the presence of 10⁻⁸ to 10⁻⁹ mol/L E₂ in the culture medium (23, 24). Of the five cell lines, the addition of 10⁻⁸ mol/L E₂ enhanced growth in the ER-α-positive cell lines (T47D, MCF-7, and R-27). However, neither of the ER-α-negative lines, SK-BR-3 and MDA-MB-231, showed enhanced growth on the addition of E₂ to the culture (Fig. 2). Tamoxifen at 10⁻⁶ mol/L inhibited E₂-induced cell growth in T47D and MCF-7 cell lines, resulting in a degree of growth similar to that observed in the absence of E₂. However, tamoxifen could not inhibit E₂-induced cell growth in R-27 cells (Fig. 2).

Effects of E₂ with or without tamoxifen or ICI 182780 on Cap43 expression. We next used real-time PCR and Western blotting to examine whether E₂ could modulate expression of Cap43 in cultured breast cancer cells. As shown in Fig. 3A, expression of Cap43 mRNA was markedly down-regulated by exogenous addition of E₂ for 24 hours in the ER-α-positive lines, T47D, MCF-7, and R-27, but not in the ER-α-negative lines, SK-BR-3 and MDA-MB-231 (Fig. 3A). Furthermore, E₂ down-regulated expression of Cap43 in a dose-dependent manner in the three ER-α-positive cell lines. Tamoxifen at 10⁻⁶ mol/L or ICI 182780 at 10⁻⁷ mol/L almost completely abrogated the E₂-induced down-regulation of the Cap43 gene in two ER-α-positive lines, T47D and MCF-7, in the presence of various doses of E₂ (10⁻¹² to 10⁻⁶ mol/L). However, the abrogatory effect of tamoxifen was hardly observed in the tamoxifen-resistant R-27 line in comparison with its parental MCF-7 cells (Fig. 3A). ICI 182780 at 10⁻⁷ mol/L almost completely abrogated the E₂-induced down-regulation of the Cap43 in R-27 cells, whereas tamoxifen at 10⁻⁶ mol/L could not abrogate the E₂-induced down-regulation of the Cap43 when in the presence of E₂ at 10⁻⁸ and 10⁻⁶ mol/L, respectively. In contrast, no marked change in Cap43 mRNA levels in SK-BR-3 and MDA-MB-231 was evident when these cell lines were treated with E₂ in the absence or presence of tamoxifen and ICI 182780 (Fig. 3A). Consistent with the effects on mRNA levels of Cap43, Western blot analysis also showed that the expression of Cap43 protein was decreased in E₂-treated T47D, MCF-7, and R-27 lines and tamoxifen at 10⁻⁶ mol/L or ICI 182780 at 10⁻⁷ mol/L abrogated the E₂-induced down-regulation of Cap43 in these three ER-α-positive cell lines. By contrast, neither treatment with E₂ alone nor E₂ with tamoxifen or ICI 182780 affected cellular Cap43 protein levels in the SK-BR-3 or MDA-MB-231 lines (Fig. 3B).

Previous studies have shown that the expression of the pS2 gene is profoundly affected by E₂ in MCF-7 cells (23–25). The results from the present study are consistent with these findings, showing that administration of E₂ markedly increases expression of pS2 mRNA and that administration of tamoxifen or ICI 182780 blocks this E₂-induced stimulatory effect in the T47D, MCF-7, and R-27 cell lines (Fig. 3C). The inhibitory effect of tamoxifen on the E₂-induced stimulation of pS2 mRNA expression in the R-27 line was much less than that in its parental MCF-7 cells. By contrast, ICI 182780 at 10⁻⁷ mol/L more greatly abrogated the E₂-induced down-regulation of Cap43 than tamoxifen at 10⁻⁶ mol/L in R-27 cells. Expression of pS2 mRNA was not affected by either E₂ alone or E₂ and tamoxifen in the SK-BR-3 and MDA-MB-231 (Fig. 3C).

Up-regulation of the Cap43 gene by nickel in all breast cancer cell lines. Cap43 was originally isolated as a gene induced by nickel compounds and its expression is highly susceptible to the presence of nickel (1, 9). We examined whether exposure to nickel at 1 mmol/L for 24 hours could specifically alter the expression of Cap43 in the five breast cancer cell lines. Exposure to nickel markedly enhanced cellular Cap43 mRNA levels in all five lines (Fig. 4A). The marked increase in the levels of Cap43 mRNA seen in these cell lines was observed irrespective of the expression of ER-α (Fig. 4A). Western blot analysis also showed significant increases in the levels of Cap43 protein in nickel-treated cell lines (Fig. 4B). Cap43 protein levels were higher in the two untreated ER-α-negative lines, SK-BR-3 and MDA-MB-231, than in the ER-α-positive lines. However, expression of Cap43 protein was further enhanced in these ER-α-positive cell lines in response to nickel.

Effect of E₂ with or without tamoxifen or ICI 182780 on C-myc mRNA expression. C-myc as well as N-myc is known to regulate expression of Cap43 gene (3). Expression of C-myc is highly susceptible to E₂. As shown in Fig. 5, expression of C-myc mRNA was markedly up-regulated by addition of E₂ in the ER-α-positive lines, T47D, MCF-7, and R-27. However, there was no change in C-myc expression in the ER-α-negative cell lines, SK-BR-3 and MDA-MB-231. Similarly, tamoxifen or ICI 182780 blocks this E₂-induced stimulatory effect in T47D and MCF-7 cells (Fig. 5). Compared with the inhibitory effect of tamoxifen on the E₂-induced C-myc up-regulation in MCF-7 cells, tamoxifen showed only a slight, if any, effect on the E₂-induced C-myc up-regulation in R-27 cells. By contrast, almost complete inhibition by ICI 182780 was observed on the E₂-induced up-regulation of C-myc in R-27 cells as well as MCF-7.

Overexpression of ER-α in ER-α-negative cells down-regulates Cap43 expression. We examined whether the E₂-induced down-regulation of Cap43 was specifically mediated through its interaction with ER-α. We introduced ER-α cDNA into the ER-α-negative lines, MDA-MB-231 and SK-BR-3, and established six cell lines, MDA/ER-1, MDA/ER-2, MDA/ER-3, SK-BR/ER-1, SK-BR/ER-2, and SK-BR/ER-3. We also isolated transfectants of the vector alone (MDA/Vec-1 and SK-BR/Vec-1). ER-α gene expression was observed in the cDNA transfectants from the SK-BR-3 and MDA-MB-231 lines (Fig. 6A and B). Three ER-α-expressing cell lines derived from SK-BR-3 and MDA-MB-231 showed a marked decrease in their expression of the Cap43 protein in comparison with their vector counterparts (Fig. 6C and D). Taken together, these data consistently indicated a close association between Cap43 gene expression and E₂-ER-α signaling in human breast cancer cells.

Expression of Cap43 and ER-α in clinical samples of human breast cancer. We next performed immunohistochemical analysis to examine whether the expression of Cap43 was associated with ER-α expression in tissue samples from breast cancer patients. Ninety-six breast cancer patients were included in this study. The relationship between Cap43 expression and clinicopathologic findings is shown in Table 1. There was no significant correlation between the expression of Cap43 and age, tumor size, menopausal status, lymph node metastasis, EGFR expression, or HER-2 expression. However, because Cap43 expression was detected in 42% of grade 1 and 2 and 71% of grade 3 tumors, there was a significant correlation between Cap43 expression and tumor grade (P = 0.0387; Table 1). Univariate analysis for 5-year postoperative overall survival done on these patients showed that there was no significant difference in Cap43 expression according to the postoperative prognosis (P = 0.345).

Immunohistochemical analysis showed that breast cancers were variously positive and negative for the expression of Cap43 and ER-α. Figure 7 shows representative immunohistochemical data from two breast cancer patients. The tissue shown from case 1 with high expression of Cap43 and negligible expression of ER-α was therefore scored positive for the expression of Cap43 and negative for ER-α expression. By contrast, case 2 showed high expression of ER-α and negligible expression of Cap43, indicating that it was positive for ER-α and negative for Cap43. Positive expression of both Cap43 and ER-α was detected in 20 of 63 (31.7%) patients, and 43 of 63 (68.3%) patients were classified as being Cap43 negative and ER-α positive (Table 2). Cap43 expression was thus generally decreased in breast cancer cells from ER-α positive patients, and Cap43 expression was increased in breast cancer cells of ER-α-negative patients, indicating that the expression of Cap43 is inversely correlated with the expression of ER-α in breast cancer patients (P = 0.0374). On the other hand, there was no relationship between PgR and Cap43 expression in breast cancer patients (P = 0.8405; Table 2).

In our present study, we observed that Cap43 expression levels were inversely correlated with expression levels of ER-α in all seven human breast cell lines (Fig. 1). One tamoxifen-resistant line, R-27, which was derived from the MCF-7 line, however, showed expression of both Cap43 and ER-α. The addition of E₂ was found to markedly down-regulate the expression of the Cap43 gene in ER-α-positive cell lines but not in ER-α-negative lines (Fig. 3A and B). Because the expression of a representative ER-responsive gene, pS2, could be modulated by E₂ only in ER-α-positive lines (Fig. 3C), we concluded that ER-dependent signaling operated in these ER-α-positive cell lines but not in the ER-α-negative lines. Furthermore, overexpression of ER-α in ER-α-negative cell lines induced down-regulation of both protein and mRNA levels of Cap43. Exposure to nickel, however, markedly increased the expression of the Cap43 gene in both ER-α-positive and ER-α-negative lines (Fig. 4), suggesting that the E₂-induced specific down-regulation of the Cap43 gene depends on ER-α. Taken together, these studies indicate that the presence of the functional ER-α could be required for E₂-induced down-regulation of Cap43.

We also showed that coadministration of tamoxifen or ICI 182780 abrogated the E₂-induced down-regulation of the Cap43 gene in ER-α-positive lines. Expression of the Cap43 gene is thus modulated in response to E₂ or antiestrogen possibly through the ER-α expressed in human breast cancer cells. Tamoxifen at 10⁻⁶ mol/L or ICI 182780 at 10⁻⁷ mol/L almost completely abrogated the E₂-induced down-regulation of Cap43 gene in ER-α-positive breast cancer cell lines, MCF-7 and T47D. The abrogatory effect of tamoxifen seemed to be much less in a tamoxifen-resistant subline, R-27, compared with the parental counterpart, MCF-7, in the presence of E₂ (10⁻⁸ to 10⁻⁶ mol/L). However, ICI 182780 could almost completely abrogate the E₂-induced down-regulation of Cap43 gene in R-27 cells. Moreover, the abrogatory effect of 10⁻⁶ mol/L tamoxifen on expression of the E₂-sensitive pS2 gene also seemed to be much less than that of ICI 182780 at 10⁻⁷ mol/L when R-27 cells were exposed to E₂ at 10⁻⁸ mol/L (Fig. 3C). ICI 182780 showed ∼10-fold higher antiestrogenic activity in the regulation of Cap43 as well as pS2 gene compared with tamoxifen, and ICI 182780 could thus overcome tamoxifen resistance in breast cancer cells, consistent with previous reports (26, 27). Cap43 could be a molecular target for the functional hormone-dependent cell growth signal of breast cancers and also a target that is useful to determine the therapeutics efficacy of antiestrogenic anticancer agents.

Expression of Cap43 gene is negatively regulated by myc gene (3). We examined the effect of E₂ on C-myc expression, and E₂-induced up-regulation of C-myc gene was observed only in ER-α-positive breast cancer cell lines. E₂ thus up-regulated C-myc and down-regulated Cap43 possibly through the E₂-ER-α pathway. E₂-induced down-regulation of Cap43 gene might be mediated through up-regulation of C-myc. However, further study should be required to determine how C-myc is involved in the E₂-induced down-regulation of Cap43 gene. Oncogenes, tumor suppressor genes, and several physiologic stimuli are known to modulate expression of Cap43 (see Introduction). Recent studies have shown that Cap43 is a p53 target gene (16, 17). Cap43 inhibited polyploidy in p53-negative cancer cell lines and increased the cell population at M phase when exposed to Taxol, a microtubule inhibitor, indicating that Cap43 plays a role in the p53-dependent mitotic spindle checkpoint (17). Stein et al. have also reported that Cap43 is necessary for p53-dependent apoptosis (16). Of the breast cancer cell lines used in this study, SK-BR-3, MDA-MB-231, and T47D have mutant p53 and MCF-7 has wild-type p53 (IARC TP53 Mutation Database, http://www-p53.iarc.fr/). Both ER-α-positive and ER-α-negative breast cancer cell lines with wild-type p53 showed a marked increase in Cap43 expression when exposed to doxorubicin, an anticancer agent that mediates cytotoxicity through p53.⁶

Nickel also promoted the increased expression of Cap43 in all ER-α-positive and ER-α-negative lines (Fig. 4). The up-regulation of Cap43 by nickel or doxorubicin occurred irrespective of the presence of p53. Tamoxifen- and ICI 182780–induced abrogation of the E₂-induced down-regulation of Cap43 might also occur irrespective of the presence of the p53 pathway but does depend on the presence of ER-α.

Cap43 is a putative metastasis suppressor gene in human colon and prostate cancer, and its expression in these cancers is closely correlated with the prognosis of patients (18, 19). We also observed an inverse correlation between Cap43 expression and the prognosis of patients with pancreatic cancers.⁷

Immunohistochemical studies by Bandyopadhyay et al. (20) on tissue from 85 breast cancer patients have shown that patients positive for Cap43 have a significantly more favorable prognosis than those with reduced expression of Cap43. The Cap43 protein was detected in normal mammary gland cells in all 85 breast cancer patients, but its expression was significantly reduced in the tumor cells of 30% of patients (20). In our present study, immunohistochemical analysis of breast cancers showed an inverse correlation between Cap43 expression and histologic grade or ER-α expression (Tables 1 and 2). However, there was no apparent correlation between Cap43 expression and other clinical and pathologic features, including lymph node metastasis and prognosis (Tables 1 and 2). At present, the reasons underlying discrepancies on the correlation between Cap43 expression and prognosis in breast cancer patients are unclear. Such discrepancies might be due to differences in the background of patients and the methods and evaluation of immunohistochemical staining used. There are also conflicting studies regarding the effects of Cap43 expression levels in colon cancer. Cap43 has been reported to be down-regulated in colon cancers with low levels of metastasis (4, 18). By contrast, a study by Wang et al. reported that Cap43 expression was well correlated with the progression of colon cancer (28). Further studies are required to understand the pathways by which Cap43 protein could modulate malignant characteristics and mechanisms, including tumor progression in breast cancer.

The action of estrogen is mediated by nuclear-localizing ER through the regulation of target gene transcription (i.e., genomic signaling). On the other hand, recent studies presented another pathway of so-called nongenomic signaling that is mediated by activation of membrane-associated ER. This nongenomic signaling events by membrane-associated ER could be also involved in some physiologic functions of E₂ (29–33). The E₂ concentrations that cause the maximal effect on the down-regulation of Cap43 in this study might be rather higher than physiologically most effective concentrations, suggesting that this E₂-induced down-regulation of Cap43 could be mediated through nongenomic signaling by membrane-associated ER. We, however, observed that most of ERs were localized in nucleus of the cells and we did not observe any apparent differences in the membrane localization of ER when the cells exposed to 10⁻¹² to 10⁻⁶ mol/L E₂ by immunocytochemistry (data not shown). It is also known that nongenomic signaling by membrane-associated ER was not blocked by the pure antiestrogen ICI 182780 (29, 34, 35). Consistent with these findings, treatment with ICI 182780 effectively blocked the E₂-induced down-regulation of Cap43 in breast cancer cells (Fig. 3). Taken together, it seems unlikely that membrane-associated ER plays a critical role in the E₂-induced down-regulation of Cap43 gene in our present study. However, further study is required to understand how E₂-dependent down-regulation of Cap43 is associated with genomic signaling.

In conclusion, the expression of Cap43, a putative differentiation- and metastasis-related gene, was greatly modulated by E₂ and/or antiestrogen in ER-α-positive breast cancer cells. Therefore, decreasing the susceptibility of Cap43 to antiestrogens would be a potential therapeutic strategy for the treatment of breast cancer.

We thank Kenji Nakano and Fumihito Hosoi for fruitful discussions.