Caspase 3 in Breast Cancer

738 Vol. 9, 738–742, February 2003 Clinical Cancer Research

Norma O’Donovan, John Crown, Helen Stunell, progression. Secondly, most cytotoxic anticancer agents cause Arnold D. K. Hill, Enda McDermott, tumor regression, at least in part, by inducing apoptosis. There- Niall O’Higgins, and Michael J. Duffy1 fore, defects in apoptosis may cause drug resistance and result in treatment failure. Departments of Medical Oncology [N. O’D., J. C.] and Nuclear Although multiple genes are involved in apoptosis, the key Medicine [M. J. D.], St. Vincent’s University Hospital, and Department of Surgery [H. S., A. D. K. H., E. M., N. O’H., M. J. D.] mediators of the process are the caspases. Caspases are aspar- and Conway Institute of Biomolecular and Biomedical Science tate-specific cysteine proteases, which cleave their substrates on [A. D. K. H., M. J. D.], University College Dublin, Dublin 4, Ireland the carboxyl side of the aspartate residue (3, 4). Currently at least 14 different caspases are known to exist, of which two- thirds play a role in apoptosis. The caspases involved in apo- ABSTRACT ptosis can be divided into two main groups, the initiator Purpose: An inability to undergo apoptosis is widely caspases (e.g., caspases 8, 9, and 10) and the downstream thought to contribute to both tumorigenesis and tumor pro- effector caspases (e.g., caspases 2, 3, 6, and 7). It is the members gression. One of the key mediators of apoptosis is the thiol of the latter group that degrade multiple cell proteins and are protease caspase 3. In this investigation, caspase 3 mRNA responsible for the morphological changes in apoptosis. and protein expression in breast cancer was examined. Caspase 3 is the most widely studied of the effector Experimental Design: Caspase 3 was measured at the caspases. It plays a key role in both the death receptor pathway, mRNA level using reverse transcription-PCR and at the initiated by caspase 8, and the mitochondrial pathway, involving protein level using both Western blotting and activity assays. caspase 9. In addition, several studies have shown that caspase Levels of apoptosis were determined using an ELISA, which 3 activation is required for apoptosis induction in response to detects nucleosomes released during DNA fragmentation. chemotherapeutic drugs e.g., taxanes, 5-fluorouracil, and doxo- Results: Relative levels of caspase 3 mRNA were similar rubicin (5–8). Caspase 3 is produced as an inactive 32-kDa ؍ ؍ in breast carcinomas (n 103), fibroadenomas (n 25), proenzyme, which is cleaved at an aspartate residue to yield a ؍ and normal breast tissues (n 6). However, levels of both 12-kDa and a 17-kDa subunit. Two 12-kDa and two 17-kDa the precursor and active forms of caspase 3 were signifi- subunits combine to form the active caspase 3 enzyme. Caspase cantly higher in carcinomas compared with both fibroade- 3 cleaves a wide range of cellular substrates including structural ؍ ؍ nomas (P 0.0188) and normal breast tissues (P 0.0002). proteins (e.g., lamins) and DNA repair enzymes [e.g., poly- Levels of apoptosis were also highest in the carcinomas and (ADP-ribose) polymerase]. It also activates an endonuclease ؍ correlated significantly with active caspase 3 levels (r caspase-activated DNAse, which causes the DNA fragmentation ؍ 0.481; P 0.0003). In the carcinomas, expression of caspase that is characteristic of apoptosis (3). 3 showed no significant relationship with either tumor size, Compared with other genes involved in apoptosis (e.g., p53 tumor grade, nodal status, or steroid receptor status but was and the Bcl-2 family), relatively little work has been carried out significantly higher in ductal carcinomas than in lobular on caspase expression in breast cancer. The aim of this study ؍ carcinomas (P 0.0188). was therefore to investigate caspase 3 expression and its rela- Conclusions: We conclude that rates of apoptosis as tionship to apoptosis in breast cancer. measured by both caspase 3 activation and nucleosome re- lease are higher in breast cancer than in nonmalignant breast tissue. This finding would appear to conflict with the MATERIALS AND METHODS widely held belief that apoptosis is reduced in malignancy. Sample Processing. Primary breast carcinoma, breast fi- The proliferation:apoptosis ratio, however, may be higher in broadenoma, and normal breast tissue samples were obtained at carcinomas than in the corresponding normal tissue. the time of surgery. Before processing, necrotic tissue was removed. The tissue samples were then snap frozen in liquid INTRODUCTION nitrogen and stored at –80°C. Tissue samples were homoge- nized using a Mikro-Dismembrator U (Braun Biotech Interna- Apoptosis or programmed cell death is important in ma- tional, Melsungen, Germany), to yield a fine powder. An aliquot lignancy for two reasons (1, 2). Firstly, suppression of apoptosis of the powder was extracted with 50 mM Tris buffer (pH 7.4) appears to be a critical event in both cancer initiation and 2 containing 1 mM monothioglycerol and assayed for ER and PR by ELISA (Abbott Diagnostics, North Chicago, IL). The cutoff point for ER and PR was 250 fmol/g wet weight tissue. Received 3/23/02; revised 9/10/02; accepted 10/9/02. 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. 2 The abbreviations used are: ER, estrogen receptor; PR, progesterone 1 To whom requests for reprints should be addressed. Phone: 353-1- receptor; RT-PCR, reverse transcription-PCR; GAPDH, glyceraldehyde-3- 2094607; Fax: 353-1-2696018; E-mail: [email protected]. phosphate dehydrogenase; AFC, aminotrifluoromethyl-coumarin.

Table 1 Levels and frequency of caspase 3 mRNA expression in normal breast tissue, fibroadenomas, and primary breast carcinomas, as determined by semiquantitative RT-PCR Tissue type n Mean Median % positive Normal 6 0.201 0.225 83.3 Fibroadenoma 25 0.212 0.163 56.0 Primary carcinoma 103 0.253 0.225 72.8 Values are expressed in arbitrary absorbance units, relative to GAPDH.

RNA Extraction. RNA extractions were performed using an RNace Total Pure kit (Bioline, Randolph, MA). RNA Fig. 1 Western blot analysis to measure expression of procaspase 3 (32 integrity was assessed by gel electrophoresis, and concentrations kDa) and the large subunit of active caspase 3 (17 kDa). Measurement ␤ were measured spectrophotometrically. of -actin was used to control for equal loading of protein samples. RT-PCR Analysis. cDNA was synthesized from 1 ␮gof ␮ total RNA, using 50 M oligo(dT)12–18 primers (Promega, Mad- ison, WI), 0.4 mM deoxynucleotide triphosphates (Promega), 1ϫ Moloney murine leukemia virus buffer, and 100 units of dopropyl)dimethylammonio]-1-propanesulfonic acid, and 10% Moloney murine leukemia virus reverse transcriptase (Invitro- sucrose (pH 7.2)] at 37°C for 100 min. Fluorescence readings gen, Carlsbad, CA). PCR was performed using primers for the were measured on a Fluoroskan Ascent FL (Thermo Lab- housekeeping gene GAPDH and caspase 3 (Maxim Biotech, San systems, Helsinki, Finland) using an excitation wavelength of Francisco, CA). Briefly, the 25-␮l reaction mix contained 5 ␮l 390 nm and an emission wavelength of 510 nm. A standard of cDNA, 250 ␮M deoxynucleotide triphosphates, 1ϫ multiple curve was constructed using AFC, and caspase 3-like activity polymerase chain reaction primers, 1ϫ multiple polymerase values were expressed as nmol product formed (AFC)/mg pro- chain reaction buffer, and 1.25 units of Taq polymerase (Pro- tein/min. mega). Amplification was performed using 2 cycles of 1 min at Apoptosis ELISA. Levels of apoptosis were measured in 95°C and 4 min at 56°C followed by 28 cycles of 1 min at 94°C tumor samples using the Cell Death Detection ELISA (Roche, and 2.5 min at 56°C, with a final extension of 10 min at 72°C. Mannheim, Germany), which detects nucleosome fragments PCR products were visualized on a 2% agarose gel. Band released into the cytoplasm during apoptosis (9, 10). In brief, 30 ␮ intensities were measured by densitometry using the Eagle Eye g of cytosolic extracts, in duplicate, were incubated in strepta- gel documentation system (Stratagene, La Jolla, CA) and ex- vidin-coated microtiter plates with the biotin-conjugated anti- pressed as arbitrary units, relative to GAPDH. histone antibody and the peroxidase-conjugated anti-DNA anti- Caspase 3 Western Blotting. Cytosolic protein extracts body for2hatroom temperature, followed by washing and were prepared by extraction in Tris (pH 7.4) containing 0.1% colorimetric detection using ABTS substrate. Absorbances were Triton X-100. Protein concentrations were determined using the measured at 405 nm (reference wavelength, 490 nm). A stand- BCA protein assay (Pierce, Milwaukee, WI). Fifty ␮g of protein ard curve was constructed using the nucleosome solution sup- from each sample were separated on a 12% polyacrylamide gel. plied with the kit. Apoptosis values were expressed as concen- ␮ After electrophoresis, protein was transferred to a nitrocellulose tration of nucleosomes ( g/mg total protein). membrane that was blocked for1hatroom temperature in 1ϫ Statistical Analysis. Nonparametric Spearman rank cor- ␹2 Tris-buffered saline containing 5% skimmed milk powder, 1% relations, Mann-Whitney t tests, and tests were performed Ͻ BSA, and 0.05% Triton X-100. The membrane was then incu- using StatView 5.0.1. (SAS Institute Inc., Cary, NC). P 0.05 bated overnight at 4°C in the above blocking solution containing was regarded as statistically significant. rabbit anti-caspase 3 polyclonal antibody (1:1000; BD Trans- duction Laboratories, San Diego, CA). Washed membranes RESULTS were then probed with horseradish peroxidase-conjugated goat Caspase 3 mRNA Analysis. Table 1 shows the results of antirabbit secondary antibody, and detection was performed semiquantitative RT-PCR analysis of caspase 3 mRNA expres- using the chemiluminescent substrate Luminol (Santa Cruz Bio- sion in normal breast tissues, fibroadenomas, and primary breast technology, Santa Cruz, CA). Recombinant caspase 3 was used carcinomas. No significant difference in either the frequency or as a positive control, and MCF7 cells, which do not express the level of caspase 3 expression was found in the three different caspase 3, were used as a negative control. Membranes were tissue types. Caspase 3 mRNA expression in primary breast reprobed with a mouse anti-␤-actin antibody (Sigma, St. Louis, carcinomas did not correlate with tumor size, nodal status, MO) to control for equal loading of protein. Band intensities histological type, or ER status. However, caspase 3 mRNA was were measured by densitometry. detected more frequently and at significantly higher levels in Caspase Activity Assays. Cytosolic protein extract (100 PR-negative tumors than in PR-positive tumors (Mann Whitney ␮g) was incubated with 50 ␮M Ac-DEVD-AFC (Alexis Corp., test, P ϭ 0.0108). San Diego, CA) in caspase activity assay buffer [20 mM PIPES, Caspase 3 Protein Analysis. Caspase 3 protein expres- 100 mM NaCl, 10 mM DTT, 1 mM EDTA, 0.1% 3-[(3-cholami- sion was measured in breast tissues by Western blotting with an

Table 2 Levels and frequency of caspase 3 protein expression in Table 3 Relationship between caspase 3 protein expression and normal breast tissue, fibroadenomas, and primary carcinomas, as current prognostic indicators for breast cancer determined by Western blotting n Procaspase 3 Active caspase 3 Tissue type n Mean Median % positive Size Procaspase 3 Յ2 cm 19 0.674 0.003 Normal 11 0.269 0.074 100 Ͼ2 cm 62 0.659 0 Fibroadenoma 23 0.485 0.412 100 Nodes Primary carcinoma 83 0.746 0.674a,b 100 Negative 37 0.556 0 Active caspase 3 Positive 41 0.715 0.001 Normal 11 0.004 0 18 Grade Fibroadenoma 23 0.005 0 44 1 3 0.598 0.011 Primary carcinoma 83 0.016 0 49 2 29 0.645 0.001 3 42 0.691 0.001 a P ϭ 0.0188 compared with fibroadenomas. Histology b P ϭ 0.0002 compared with normal breast tissue. Lobular 6 0.342 0 Values are expressed in arbitrary optical density units, relative to Ductal 67 0.677a 0.002b ␤-actin. ER status Negative 28 0.659 0 Positive 52 0.621 0.001 PR status antibody that reacts with both the inactive procaspase 3 (32 Negative 30 0.710 0 kDa) and the large subunit (17-kDa) of active caspase 3 (Fig. 1). Positive 45 0.645 0 Procaspase 3 was detected in all of the samples tested (Table 2). a P ϭ 0.0209, Mann-Whitney U test. However, the levels of procaspase 3 were significantly higher in b P ϭ 0.0152, Mann-Whitney U test. primary carcinomas than in either fibroadenomas (Mann Whit- ney test, P ϭ 0.0188) or normal breast tissues (Mann Whitney test, P ϭ 0.0002). Levels of active caspase 3 (17-kDa subunit) ϭ ␮ also tended to be higher in the primary carcinomas (n ϭ 83; in the primary carcinoma samples (mean 3.1 g nucleo- ϭ ϭ ␮ mean ϭ 0.016 absorbance units) than in either fibroadenomas somes/mg protein; n 58) and fibroadenomas (mean 3.2 g ϭ (n ϭ 23; mean ϭ 0.004 absorbance units) or normal breast nucleosomes/mg protein; n 21) than in the normal breast ϭ ␮ ϭ tissues (n ϭ 11; mean ϭ 0.005 absorbance units), although these tissues (mean 0.7 g nucleosomes/mg protein; n 10). No differences did not reach statistical significance (Table 2). In significant correlation was found between either caspase 3 addition, the 17-kDa active subunit of caspase 3 constituted a mRNA or procaspase 3 protein and levels of apoptosis. In larger proportion of the total caspase 3 protein in the carcinoma contrast, active caspase 3, whether measured by Western blot- ϭ ϭ ϭ ϭ samples (1.85%) and the fibroadenomas (2.32%) than in the ting (r 0.481; P 0.0003; n 63) or an activity assay (r ϭ ϭ normal breast tissues (0.3%). 0.352; P 0.0279; n 40), correlated significantly with No significant relationship was observed between caspase apoptosis rates (Fig. 3). No relationship was observed between 3 protein levels and either tumor size, nodal status, grade, or the levels of apoptosis and tumor size, nodes, histology, or steroid receptor status (Table 3). A significant correlation was steroid receptor status. observed between caspase 3 protein and histological tumor type. Although the numbers of lobular tumors tested were low (n ϭ DISCUSSION 6), levels of procaspase 3 were significantly higher in the ductal A failure to undergo apoptosis is widely believed to be a tumors (n ϭ 67) than in the lobular group (Mann Whitney test, key event in cancer formation and progression (1, 2). Using two P ϭ 0.0209; Fig. 2). Furthermore, active caspase 3 was not indices of apoptosis, i.e., caspase 3 protein levels and nucleo- detected in any of the lobular tumors but was found in 38 of 67 some concentration, we showed that breast carcinomas exhib- (56.7%) ductal tumors. Caspase 3 protein levels showed no ited higher rates of apoptosis than either fibroadenomas or significant relationship with the relative levels of caspase 3 normal breast tissue. Other investigators have also reported mRNA. increased rates of apoptosis in breast carcinomas compared with Caspase 3 Activity Assays. Caspase 3-like activity was control nonmalignant breast tissue. Using the terminal de- measured using the fluorometric substrate Ac-DEVD-AFC. oxynucleotidyl transferase-mediated nick end labeling method Caspase 3 activity was not detected in any of the normal breast to measure apoptotic rates, Wong et al. (11) found that the mean tissues (n ϭ 8) or fibroadenomas (n ϭ 20) tested. In contrast, of apoptotic index in cancer was 4.3%, whereas the corresponding the 49 primary carcinomas tested, 11 were positive (22.4%) for value for the adjacent normal ductal cells was 0.04%. Also using caspase 3 activity (mean, 0.008; range, 0–0.097). Levels of the terminal deoxynucleotidyl transferase-mediated nick end caspase 3 activity in the primary carcinomas (n ϭ 49) correlated labeling assay, Vakkala et al. (12) reported that the mean with the levels of both procaspase 3 and active caspase 3 apoptotic index was 0.14 in benign epithelial hyperplasias, 0.17 measured by Western blotting [Spearman rank correlations: P ϭ in atypical hyperplasias, 0.61 in in situ carcinomas, and 0.94 in 0.0162 (r ϭ 0.347) and P ϭ 0.0169 (r ϭ 0.345), respectively]. invasive carcinomas. Apoptosis Assays. Apoptosis levels were measured us- Whereas apoptosis rates appear higher in invasive breast ing the Cell Death Detection ELISA (Roche), which measures cancers than in corresponding nonmalignant breast tissue, pro- nucleosomes released into the cytoplasm when DNA fragmen- liferation rates are also generally increased in the malignant tation occurs during apoptosis. Levels of apoptosis were higher tissues (13). Indeed, Wong et al. (11) found a significant cor-

Fig. 2 Levels of procaspase 3 protein in ductal and lobular tumors, Fig. 3 Levels of the 17-kDa subunit of active caspase 3 in apoptosis- measured by Western blotting. Boxes represent the 75th percentile with negative and apoptosis-positive primary carcinomas. Boxes represent the median indicated, and the bars represent the 90th percentile. the 75th percentile with the median indicated, and the bars represent the 90th percentile. Median levels of active caspase 3 in apoptosis-negative samples were zero. relation between the apoptotic index and proliferation rates in invasive breast carcinomas. Furthermore, Mommers et al. (14) potential marker for predicting response or resistance to chemo- showed that whereas both mitotic and apoptotic indices were therapeutic agents in breast cancer. Indeed, recent data from higher in invasive breast cancer compared with either hyperpla- breast cancer cell lines support this hypothesis. For example, sias or well-differentiated in situ cancers, the mitotic/apoptotic Blanc et al. (19) showed that caspase 3 was essential for index was increased in the invasive carcinomas. Recently, Zhao procaspase 9 processing and cisplatin-induced apoptosis in et al. (15) also showed that both the mitotic index and the MCF7 breast cancer cells. Using the same cell lines, Yang et al. apoptotic index are increased in ductal carcinoma in situ. Parton (20) reported that transfection with cDNA for caspase 3 led to et al. (16) also found a significant correlation between apoptosis doxorubicin- and etoposide-induced apoptosis. Future work and proliferation in breast tumors before treatment. These stud- should therefore investigate whether caspase 3 levels in breast ies, when taken together, suggest that both apoptosis and pro- carcinomas can predict clinical response/resistance to these liferation are increased in breast cancer. agents. Although higher levels of caspase 3 protein were found in malignant versus nonmalignant breast tissue in the present in- REFERENCES vestigation, no significant relationship was found between 1. Kerr, J. F., Winterford, C. M., and Harmon B. V. Apoptosis. Its caspase 3 protein levels and either tumor size, nodal status, or significance in cancer and cancer therapy. Cancer (Phila.), 73: 2013– steroid receptor status. Using immunohistochemistry, Vakkala 2026, 1994. et al. (12) found no significant association between caspase 3 2. Evan, G. I., and Vousden, K. H. Proliferation, cell cycle and apop- and either tumor grade, ER status, or PR status. However, tosis in cancer. Nature (Lond.), 411: 342–348, 2001. Mommers et al. (14) reported that the mean apoptotic index was 3. Stennicke, H. R., and Salvesen, G. S. Properties of the caspases. Biochim. Biophys. Acta, 1387: 17–31, 1998. approximately 3-fold higher in poorly differentiated breast can- 4. Thornberry, N. A., and Lazebnik, Y. Caspases: enemies within. cers than in well-differentiated breast cancers. In this study, Science (Wash. DC), 281: 1312–1316, 1998. higher levels of caspase 3 protein were found in ductal cancers 5. Keane, M. M., Ettenberg, S. A., Nau, M. M., Russell, E. K., and than in lobular breast cancers. Previously, lobular cancers were Lipkowitz, S. Chemotherapy augments TRAIL-induced apoptosis in shown to have a lower apoptotic index than ductal tumors (17). breast cell lines. Cancer Res., 59: 734–741, 1999. These findings suggest that the mechanisms of apoptosis may be 6. Bellarosa, D., Ciucci, A., Bullo, A., Nardelli, F., Manzini, S., Maggi, different in ductal and lobular breast carcinomas. C. A., and Goso, C. Apoptotic events in a human ovarian cancer cell line Our results show that measurement of active caspase 3, exposed to anthracyclines. J. Pharmacol. Exp. Ther., 296: 276–283, 2001. whether by Western blotting or by a catalytic activity assay, 7. Kottke, T. J., Blajeski, A. L., Martins, M., Mesner, P. W., Jr., correlates with apoptosis rates as detected by an ELISA for Davidson, N. E., Earnshaws, W. C., Armstrong, D. K., and Kaufmann, nucleosome fragments. In contrast, neither procaspase 3 nor S. H. Comparison of paclitaxel-, 5-fluoro-2Ј-deoxyuridine-, and epider- caspase 3 mRNA levels correlated with nucleosome fragment mal growth factor (EGF)-induced apoptosis. J. Biol. Chem., 274: levels. These results suggest that whereas measurements of 15927–15936, 1999. caspase 3 at either the mRNA or precursor protein level may 8. Suzuki, A., Kawabata, T., and Kato, M. Necessity of interleukin-1␤ converting enzyme cascade in taxotere-initiated death signaling. Eur. indicate a potential for apoptosis, measurement of the active J. Pharmacol., 343: 87–92, 1998. protein is necessary to detect ongoing apoptosis. 9. Bonfoco, E., Krainc, D., Ankarcrona, M., Nicotera, P., and Lipton, Because caspase 3 is a critical mediator of apoptosis (18) S. A. Apoptosis and necrosis: two distinct events induced, respectively, and correlates with apoptotic rates in breast cancer, it is a by mild and intense insults with N-methyl-D-aspartate or nitric oxide/

superoxide in cortical cell cultures. Proc. Natl. Acad. Sci. USA, 92: apoptotic cells in ductal carcinoma in situ of the breast. Breast Cancer, 7162–7166, 1995. 9: 118–126, 2002. 10. Kikuchi, S., Hiraide, H., Tamakuma, S., and Yamamoto, M. Expres- 16. Parton, M., Krajewski, S., Smith, I., Krajewska, M., Archer, C., sion of wild-type p53 tumor suppressor gene and its possible involvement Naito, M., Ahern, R., Reed, J., and Dowsett, M. Coordinate expression in the apoptosis of thyroid tumors. Surg. Today, 27: 226–233, 1997. of apoptosis-associated proteins in human breast cancer before and 11. Wong, S. C. C., Chan, J. K. C., Lee, K. C., and Hsiao, W. L. W. during chemotherapy. Clin. Cancer Res., 8: 2100–2108, 2002. Differential expression of p16/p21/p27 and cyclin D1/D3, and their 17. Kruger, S., Fahrenkrog, T., and Muller, H. Proliferative and apo- relationships to cell proliferation, apoptosis, and tumour progression in ptotic activity in lobular breast carcinomas. Int. J. Mol. Med., 4: 171– invasive ductal carcinoma of the breast. J. Pathol., 194: 35–42, 2001. 174, 1999. 12. Vakkala, M., Pa¨a¨ko¨, P., and Soini, Y. Expression of caspases 3, 6, 18. Slee, E. A., Adrain, C., and Martin, S. J. Serial killer: ordering and 8 is increased in parallel with apoptosis and histological aggres- caspase activation events in apoptosis. Cell Death Differ., 6: 1067– siveness of the breast lesion. Br. J. Cancer, 81: 592–599, 1999. 1074, 1999. 13. Bai, M., Agnantis, N. J., Kamina, S., Demou, A., Zagorianakou, P., 19. Blanc, C., Deveraux, Q. L., Krajewski, S., Ja¨nicke, R. U., Porter, Katsaraki, A., and Kanavaros, P. In vivo cell kinetics in breast carcino- A. G., Reed, J. C., Jaggi, R., and Marti, A. Caspase-3 is essential for genesis. Breast Cancer Res., 3: 276–283, 2001. procaspase-9 processing and cisplatin-induced apoptosis of MCF-7 14. Mommers, E. C. M., van Diest, P. J., Leonhart., A. M., Meijer, breast cancer cells. Cancer Res., 60: 4386–4390, 2000. C. J. L. M., and Baak, J. P. A. Balance of cell proliferation and apoptosis 20. Yang, X-H., Sladek, T. L., Liu, X., Butler, B. R., Froelich, C. J., and in breast carcinogenesis. Breast Cancer Res. Treat., 58: 163–169, 1999. Thor, A. D. Reconstitution of caspase 3 sensitizes MCF-7 breast cancer 15. Zhao, H., Morimoto, T., Sasa, M., Tanaka, T., and Izumi, K. cells to doxorubicin- and etoposide-induced apoptosis. Cancer Res., 61: Immunohistochemical expression of uPA, PAI-1, cathepsin D and 348–354, 2001.

Norma O'Donovan, John Crown, Helen Stunell, et al.

Clin Cancer Res 2003;9:738-742.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/9/2/738

Cited articles This article cites 20 articles, 8 of which you can access for free at: http://clincancerres.aacrjournals.org/content/9/2/738.full#ref-list-1

Citing articles This article has been cited by 5 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/9/2/738.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/9/2/738. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.