- Email: [email protected]
Plasma membrane calcium-ATPase 2 and 4 in human breast cancer cell lines
BBRC Biochemical and Biophysical Research Communications 337 (2005) 779–783 www.elsevier.com/locate/ybbrc
Plasma membrane calcium-ATPase 2 and 4 in human breast cancer cell lines q Won Jae Lee, Sarah J. Roberts-Thomson, Gregory R. Monteith * The School of Pharmacy, The University of Queensland, Brisbane, Qld 4072, Australia Received 16 September 2005 Available online 28 September 2005
Abstract There is evidence to suggest that plasma membrane Ca2+-ATPase (PMCA) isoforms are important mediators of mammary gland physiology. PMCA2 in particular is upregulated extensively during lactation. Expression of other isoforms such as PMCA4 may influence mammary gland epithelial cell proliferation and aberrant regulation of PMCA isoform expression may lead or contribute to mammary gland pathophysiology in the form of breast cancers. To explore whether PMCA2 and PMCA4 expression may be deregulated in breast cancer, we compared mRNA expression of these PMCA isoforms in tumorigenic and non-tumorigenic human breast epithelial cell lines using real time RT-PCR. PMCA2 mRNA has a higher level of expression in some breast cancer cell lines and is overexpressed more than 100-fold in ZR-75-1 cells, compared to non-tumorigenic 184B5 cells. Although differences in PMCA4 mRNA levels were observed between breast cell lines, they were not of the magnitude observed for PMCA2. We conclude that PMCA2 mRNA can be highly overexpressed in some breast cancer cells. The significance of PMCA2 overexpression on tumorigenicity and its possible correlation with other properties such as invasiveness requires further study. Ó 2005 Elsevier Inc. All rights reserved. Keywords: Calcium; PMCA; Mammary gland; Lactation; Breast cancer; mRNA; Transcription
Free intracellular calcium (Ca2+) controls many essential cellular processes. The fidelity of Ca2+ signaling depends on homeostatic mechanisms that maintain steep Ca2+ gradients across biological membranes [1–3]. Mechanisms that lower the intracellular Ca2+ concentration ([Ca2+]i) prevent excessive rises in [Ca2+]i and thus circumvent cellular toxicity [1–3]. Plasma membrane Ca2+-ATPases (PMCAs) represent one such mechanism and belong to a family of P-type Ca2+-ATPases that specifically and actively transport Ca2+ [4]. PMCAs are important modulators of intracellular Ca2+ signals as they can affect the
q This work was supported by grants from the Queensland Cancer Fund (Q34-02), the National Health and Medical Research Council (NHMRC) of Australia (102426), and a Dora Lush (Biomedical) Research Scholarship awarded to W.J.L. by the NHMRC. * Corresponding author. Fax: +61 7 3365 1688. E-mail address: [email protected] (G.R. Monteith).
0006-291X/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2005.09.119
dynamics of Ca2+ signals by modulating amplitude and duration [5,6]. PMCA isoforms (PMCA1–4) are encoded by genes located on separate chromosomes [4]. PMCA1 and PMCA4 are expressed ubiquitously while PMCA2 and PMCA3 are generally found in excitable tissues [7]. PMCAs differ in their biochemical properties, with PMCA2 having the greatest affinity for calmodulin, which is the main activator of PMCA-mediated Ca2+ transport [8,9]. The distinctive features of PMCAs suggest that the expression of different PMCA isoforms is matched to suit the requirements for maintaining Ca2+ homeostasis in particular cell types or tissues and thus support the cell or tissueÕs cellular and physiological functions [10]. The mammary gland experiences extensive Ca2+ flux during lactation [11,12], yet our understanding of the relative importance of PMCAs in milk production and their potential involvement in other key processes in mammary gland
780
W.J. Lee et al. / Biochemical and Biophysical Research Communications 337 (2005) 779–783
physiology is limited. Rat mammary gland expresses PMCA1, 2, and 4 [13,14]. PMCA2 expression undergoes significant induction after parturition, then rapidly increases as lactation progresses and remains elevated at levels that are almost 100-fold higher compared to levels during pregnancy [13,14]. Moreover, null mutations in PMCA2 substantially lower milk Ca2+ concentrations in lactating mice [15]. Thus, there is strong evidence supporting PMCA2 as a major mechanism for the enrichment of Ca2+ into milk and thus extracellular Ca2+ homeostasis. Dynamic and coordinated regulation of PMCA isoform expression appears to be a crucial regulatory mechanism for supporting the unique physiology of the mammary gland [13,14]. Other PMCA isoforms may also control breast development, architecture, and post-lactational involution in humans, since particular PMCA isoforms mediate proliferation, differentiation, and apoptosis in other cell culture model systems [16–18]. Little is known about exactly how PMCA isoforms are regulated in the mammary gland to meet different physiological circumstances, for example, formation of glandular structures versus secretory acini. Deregulation of coordinated and dynamic PMCA isoform expression in the mammary gland is hypothesized to contribute to pathological consequences such as mammary gland tumorigenesis. We previously reported that the relative expression of PMCA1 mRNA is significantly higher in tumorigenic MCF-7 and MDA-MB-231 human breast cancer cell lines, compared to non-tumorigenic MCF-10A human breast epithelial cells [19]. We have also shown that expression of stable antisense targeting PMCA inhibits the proliferation of MCF-7 cells [20]. To further explore the plausibility that deregulation of PMCA isoform expression is characteristic of mammary gland tumorigenesis and given that transcriptional modulation represents the most likely means for PMCA regulation in the course of changing physiological demands [21–23], we compared the levels of PMCA2 and PMCA4 mRNA in a bank of tumorigenic and non-tumorigenic human breast epithelial cell lines. Materials and methods Cell culture. The tumorigenic human breast cancer cell lines MCF-7 and MDA-MB-231, and the non-tumorigenic human breast epithelial cell line MCF-10A were cultured and plated for total RNA isolations as previously described [19]. The tumorigenic human breast cancer cell lines ZR-75-1, T-47D, BT-483, and SK-BR-3, and the non-tumorigenic human breast epithelial cell lines 184A1 and 184B5 were a generous gift from Professor Rob Sutherland (Garvan Institute, Sydney, Australia). The aforementioned tumorigenic cell lines were grown in basal DMEM containing phenol red (JRH Biosciences, Brooklyn, Victoria, Australia), and supplemented with 10% FBS (JRH Biosciences), 2 mM L-glutamine (Invitrogen, Mount Waverly, Victoria, Australia) and 100 U/ml–100 lg/ ml penicillin G–streptomycin (Invitrogen). 184A1 and 184B5 cells were cultured in phenol red containing basal MCDB 170 medium (Invitrogen), with the inclusion of 100 U/ml–100 lg/ml penicillin G–streptomycin and serum-free supplements provided by the manufacturer that were used at the recommended dilution (Invitrogen). ZR-75-1, T-47D, BT-483, SKBR-3, 184A1, and 184B5 cells were routinely passaged once a week or when approximately 90% confluent and subcultured into T-75 cm2 or T-
150 cm2 cell culture flasks (in the case of the non-tumorigenic cells) at a split ratio of 1:5–1:10. Cultures were harvested for isolation of total RNA when they reached confluence. All cell lines were maintained in a humidified 5% CO2/95% air incubator at 37 °C. RNA isolation. Total RNA was isolated using RNeasy mini or midi columns (Qiagen, Doncaster, Victoria, Australia) according to the manufacturerÕs instructions. This procedure included an on-column DNA digestion step using RNase-free DNase I (Qiagen). Total RNA concentrations were quantified by measuring UV absorbance at 260 nm. Real time RT-PCR. Taqman real time RT-PCR assays were used to determine the expression of PMCA2 and PMCA4 mRNA, relative to the 18S rRNA. For the relative quantification of PMCA2 mRNA (GenBank Accession No. NM_001683) primers and probe were designed using Primer Express software, version 1.5 (Applied Biosystems, Scoresby, Victoria, Australia). The primers had the following sequences: forward primer 5 0 -ATCACCCGGCAGCCTCTT-3 0 and reverse primer 5 0 TTGGAGTAAAAGTCGCTGTTGGT-3 0 , and probe sequence was: 5 0 CCGAGCAAGGACCG-3 0 . The Taqman Gene Expression Assay Hs00608066_m1 (Applied Biosystems) was used to quantitate PMCA4 mRNA, and primers and probe for 18S rRNA were also purchased from Applied Biosystems. For all reactions quantitating 18S rRNA, primers and probe were included at 50 nM and were cycled in separate tubes at a 1:1000 dilution of total RNA using the same method as for each specific target. PMCA2 mRNA was quantitated using a one-step RT-PCR procedure and the TaqMan one-step RT-PCR master mix reagents kit (Applied Biosystems), primers (900 nM), probe (200 nM), and total RNA (100 ng for PMCA2, 0.1 ng for 18S rRNA). Reactions were cycled using an ABI PRISM 7700 Sequence Detector (Applied Biosystems) with the following thermal cycling conditions: 48 °C for 30 min, 95 °C for 10 min, followed by 40 cycles at 95 °C for 15 s, and 60 °C for 1 min. First strand cDNA synthesis required for real time PCR detection of the PMCA4 amplicon was prepared using the Omniscript RT kit according to the vendorÕs instructions (Qiagen). Real time PCR amplifications for PMCA4 mRNA and 18S rRNA amplicons were then set up as separate reactions in TaqMan universal PCR master mix (Applied Biosystems). Each tube for real time PCR detection of PMCA4 mRNA contained cDNA template equivalent to 25 ng of total RNA and for 18S rRNA 0.025 ng total RNA. Thermal cycling was completed using an Applied Biosystems 7500 Real Time PCR System with 95 °C for 10 min and then 40 cycles at 95 °C for 15 s and 60 °C for 1 min. Data and statistical analysis. Analysis of PMCA2 mRNA real time RTPCR data was carried out using the comparative CT method [19,24,25]. PMCA4 mRNA real time RT-PCR data were analyzed using the relative standard curve method. See User Bulletin #2: ABI PRISM 7700 Sequence Detection System December 11, 1997 (updated 10/2001) from Applied Biosystems for further information. Data points represent the mean ± the standard error of the mean (SEM). N values are given in the figure legends for each corresponding graph. Means were compared with one-way analysis of variance (ANOVA) and subsequent TukeyÕs pairwise comparison tests using Prism Version 4.03 (GraphPad Software, San Diego, CA). Differences were considered to be statistically significant if P values were less than 0.05 (P < 0.05).
Results and discussion PMCA2 and PMCA4 mRNA expression in MCF-7, MDAMB-231, and MCF-10A cells Breast cancer cell lines express multiple PMCA mRNA isoforms and our laboratory has previously observed that relative PMCA1 mRNA expression is greater in tumorigenic MCF-7 and MDA-MB-231 cells compared to nontumorigenic MCF-10A cells [19]. To see if there are also differences in the expression of other PMCA mRNA
W.J. Lee et al. / Biochemical and Biophysical Research Communications 337 (2005) 779–783
PMCA2 and PMCA4 mRNA expression in other breast epithelial cell lines
9
Subconfluent
*
Confluent
*
8 7 6 5 4 3 2 1 0 -1 -2
MCF-7
MDA-MB-231
MCF-10A
Fig. 1. Relative PMCA2 mRNA levels in MCF-7, MDA-MB-231, and MCF-10A cells, normalized to confluent MCF-10A cells and expressed as DDCT values. Bars represent the mean ± SEM (n = 4, 4, 3, 3, 4, and 4 for samples left to right). The asterisk (*) denotes a significant difference compared to confluent MCF-10A cells using a TukeyÕs test.
2.5 Subconfluent
10
2.0
1.5
1.0
9 8 7
*
6
*
5 4 3 2 1 0 -1
0.5
MCF-7
MDA-MB-231
MCF-10A
Fig. 2. Relative PMCA4 mRNA levels in MCF-7, MDA-MB-231, and MCF-10A cells, normalized to confluent MCF-10A cells and expressed as fold change. Bars represent the mean ± SEM (n = 4 for all samples). No significant differences were found between all pairwise comparisons using TukeyÕs tests.
5 4B 18
18
4A
1
-3 -B R
SK
T4
D 47 T-
5 -7 ZR
0.0
83
-2
-1
Relative PMCA4 mRNA levels (fold change, normalized to confluent MCF-10A)
Confluent
*
11
Relative PMCA2 mRNA levels (-∆∆CT, normalized to 184B5)
isoforms in the above cell lines, we performed real time RT-PCR for PMCA2 and PMCA4 mRNA. Fig. 1 shows the results for relative PMCA2 mRNA expression in subconfluent and confluent cultures of MCF-7, MDA-MB231, and MCF-10A cells. Relative PMCA2 mRNA levels within a cell line did not differ significantly with the state of confluence. PMCA2 mRNA expression was, however, significantly greater in the confluent breast cancer cell lines MCF-7 and MDA-MB-231 compared to confluent normal breast MCF-10A cells (Fig. 1). In contrast, relative PMCA4 mRNA expression in subconfluent and confluent MCF-7, MDA-MB-231, and MCF-10A cells showed no significant differences either within or between cell lines (Fig. 2).
We then assessed relative PMCA2 and PMCA4 mRNA levels in a larger array of tumorigenic ZR-75-1, T-47D, BT-483, and SK-BR-3 cells and two other nontumorigenic 184A1 and 184B5 cell lines. Non-tumorigenic 184A1 and 184B5 cells did not differ significantly from each other in their relative expression of PMCA2 mRNA (Fig. 3). Except for SK-BR-3 cells, relative PMCA2 mRNA expression was significantly higher in all the tumorigenic cell lines tested, when compared to 184B5 cells (Fig. 3). Different levels of PMCA2 mRNA expression were observed within the group of tumorigenic cell lines examined, with ZR-75-1 cells expressing the highest level of PMCA2 mRNA. The ZR-75-1 breast cancer cells express a more than 100-fold greater level of PMCA2 mRNA compared to the normal 184B5 cell line ðfold change ¼ 2DDCT Þ. Whereas housekeeping PMCA isoforms such as PMCA4 may predominately influence processes like pregnancy-induced mammary gland development or cell proliferation in the breast [14,20], the evidence for PMCA2 in the mammary gland suggests that PMCA2 is primarily a mechanism for bulk Ca2+ transport during lactation for the enrichment of milk with Ca2+ [13–15]. Since lactation is protective against breast cancer in premenopausal women [26], the finding that PMCA2 is overexpressed in tumorigenic cell lines (Figs. 1 and 3) as well as being present in ZR-75-1 at levels 100-fold greater than the non-tumorigenic
B
Relative PMCA2 mRNA levels (-∆∆CT, normalized to MCF-10A confluent)
11 10
781
Fig. 3. Relative PMCA2 mRNA levels in an extended panel of tumorigenic (ZR-75-1, T-47D, BT-483, and SK-BR-3) and non-tumorigenic (184A1 and 184B5) breast epithelial cell lines. Bars represent the mean ± SEM (n = 3 for all samples) and are indicative of two independent experiments. The asterisk (*) denotes a significant difference compared to 184B5 cells using a TukeyÕs test.
782
W.J. Lee et al. / Biochemical and Biophysical Research Communications 337 (2005) 779–783
cell lines is intriguing and suggests that some breast cancers may be characterized by the overexpression of calcium transporters associated with lactation. Indeed, Stat5a is a protein involved in mammary gland development and milk protein expression during lactation [27,28] that may also be involved in breast cancer [29,30]. Our studies indicate that the increased levels of PMCA2 in some breast cancer cell lines are not correlated with estrogen receptor status, since the estrogen receptor negative cell line MDA-MB-231 had elevated levels of PMCA2. PMCA4 mRNA expression in all the cell lines examined varied modestly with a narrower dynamic range than that seen for PMCA2 mRNA. Indeed, the greatest difference between two breast cell lines for PMCA4 was approximately 8-fold, whereas for PMCA2 it was more than 100-fold. However, there was a modest but a significantly lower level of PMCA4 mRNA in the tumorigenic samples when compared to the 184B5 non-tumorigenic cell line (Fig. 4). The differences in PMCA4 mRNA levels seen in this study could be of functional importance, given that a reduction in PMCA4 expression is correlated with an inhibition of MCF-7 breast cancer cell line proliferation [20] and may suggest a slightly greater sensitivity to selective PMCA4 inhibition in tumorigenic cell lines. However, the modest differences between cell lines in PMCA4 may reflect a more housekeeping role for PMCA4 compared with PMCA2, which has a highly restricted tissue distribution [7] and is the PMCA isoform with the greatest change in expression associated with lactation [13,14]. The potential biological significance of PMCA2 overexpression in breast tumorigenesis as well as its impact on 2.5 Relative PMCA4 mRNA levels (fold change, normalized to 184B5)
* 2.0
1.5
1.0
*
*
0.5
*
* 5 18
4B
1 4A 18
-3 K S
B
T-
-B
48
R
3
D 47 T-
ZR
-7
5-
1
0.0
Fig. 4. Relative PMCA4 mRNA levels in an extended panel of tumorigenic (ZR-75-1, T-47D, BT-483, and SK-BR-3) and non-tumorigenic (184A1 and 184B5) breast epithelial cell lines. Bars represent the mean ± SEM (n = 3 for all samples) and are indicative of two independent experiments. The asterisk (*) denotes a significant difference compared to 184B5 cells using a TukeyÕs test.
Ca2+ signaling in mammary gland epithelial cells is uncertain. Our study does further highlight the isoform-specific differences in PMCA expression in epithelial cells of the mammary gland. Future studies are required to determine if relative levels of PMCA isoforms correlate to clinical outcomes or sensitivity to specific treatment regimes. Further studies are also required to examine the mechanisms of PMCA isoform transcriptional and translational regulation particularly in mammary gland epithelial cells during lactation and breast cancer. Acknowledgments We thank Dr. Jodie Robinson and Nicola Holman for assistance with methodology. References [1] E. Carafoli, Intracellular calcium homeostasis, Annu. Rev. Biochem. 56 (1987) 395–433. [2] E. Carafoli, L. Santella, D. Branca, M. Brini, Generation, control, and processing of cellular calcium signals, Crit. Rev. Biochem. Mol. Biol. 36 (2001) 107–260. [3] M.J. Berridge, P. Lipp, M.D. Bootman, The versatility and universality of calcium signalling, Nat. Rev. Mol. Cell Biol. 1 (2000) 11–21. [4] E. Carafoli, Biogenesis: plasma membrane calcium ATPase: 15 years of work on the purified enzyme, FASEB. J. 8 (1994) 993–1002. [5] M. Brini, D. Bano, S. Manni, R. Rizzuto, E. Carafoli, Effects of PMCA and SERCA pump overexpression on the kinetics of cell Ca(2+) signalling, EMBO. J. 19 (2000) 4926–4935. [6] M. Brini, L. Coletto, N. Pierobon, N. Kraev, D. Guerini, E. Carafoli, A comparative functional analysis of plasma membrane Ca2+ pump isoforms in intact cells, J. Biol. Chem. 278 (2003) 24500–24508. [7] E.E. Strehler, D.A. Zacharias, Role of alternative splicing in generating isoform diversity among plasma membrane calcium pumps, Physiol. Rev. 81 (2001) 21–50. [8] H. Hilfiker, D. Guerini, E. Carafoli, Cloning and expression of isoform 2 of the human plasma membrane Ca2+ ATPase. Functional properties of the enzyme and its splicing products, J. Biol. Chem. 269 (1994) 26178–26183. [9] N.L. Elwess, A.G. Filoteo, A. Enyedi, J.T. Penniston, Plasma membrane Ca2+ pump isoforms 2a and 2b are unusually responsive to calmodulin and Ca2+, J. Biol. Chem. 272 (1997) 17981–17986. [10] G.R. Monteith, B.D. Roufogalis, The plasma membrane calcium pump-a physiological perspective on its regulation, Cell Calcium 18 (1995) 459–470. [11] D.B. Shennan, M. Peaker, Transport of milk constituents by the mammary gland, Physiol. Rev. 80 (2000) 925–951. [12] R.L. Horst, J.P. Goff, T.A. Reinhardt, Calcium and vitamin D metabolism during lactation, J. Mammary Gland Biol. Neoplasia 2 (1997) 253–263. [13] T.A. Reinhardt, R.L. Horst, Ca2+-ATPases and their expression in the mammary gland of pregnant and lactating rats, Am. J. Physiol. 276 (1999) C796–C802. [14] T.A. Reinhardt, A.G. Filoteo, J.T. Penniston, R.L. Horst, Ca(2+)ATPase protein expression in mammary tissue, Am. J. Physiol. Cell Physiol. 279 (2000) C1595–C1602. [15] T.A. Reinhardt, J.D. Lippolis, G.E. Shull, R.L. Horst, Null mutation in the gene encoding plasma membrane Ca2+-ATPase isoform 2 impairs calcium transport into milk, J. Biol. Chem. 279 (2004) 42369–42373. [16] M. Husain, L. Jiang, V. See, K. Bein, M. Simons, S.L. Alper, R.D. Rosenberg, Regulation of vascular smooth muscle cell proliferation by plasma membrane Ca(2+)-ATPase, Am. J. Physiol. 272 (1997) C1947–C1959.
W.J. Lee et al. / Biochemical and Biophysical Research Communications 337 (2005) 779–783 [17] P.C. Brandt, J.E. Sisken, R.L. Neve, T.C. Vanaman, Blockade of plasma membrane calcium pumping ATPase isoform I impairs nerve growth factor-induced neurite extension in pheochromocytoma cells, Proc. Natl. Acad. Sci. USA 93 (1996) 13843–13848. [18] K. Paszty, A.K. Verma, R. Padanyi, A.G. Filoteo, J.T. Penniston, A. Enyedi, Plasma membrane Ca2+ ATPase isoform 4b is cleaved and activated by caspase-3 during the early phase of apoptosis, J. Biol. Chem. 277 (2002) 6822–6829. [19] W.J. Lee, S.J. Roberts-Thomson, N.A. Holman, F.J. May, G.M. Lehrbach, G.R. Monteith, Expression of plasma membrane calcium pump isoform mRNAs in breast cancer cell lines, Cell Signal 14 (2002) 1015–1022. [20] W.J. Lee, J.A. Robinson, N.A. Holman, M.N. McCall, S.J. RobertsThomson, G.R. Monteith, Antisense-mediated inhibition of the plasma membrane calcium-ATPase suppresses proliferation of MCF-7 cells, J. Biol. Chem. 280 (2005) 27076–27084. [21] D. Guerini, E. Garcia-Martin, A. Gerber, C. Volbracht, M. Leist, C.G. Merino, E. Carafoli, The expression of plasma membrane Ca2+ pump isoforms in cerebellar granule neurons is modulated by Ca2+, J. Biol. Chem. 274 (1999) 1667–1676. [22] D. Guerini, X. Wang, L. Li, A. Genazzani, E. Carafoli, Calcineurin controls the expression of isoform 4CII of the plasma membrane Ca(2+) pump in neurons, J. Biol. Chem. 275 (2000) 3706–3712. [23] L. Li, D. Guerini, E. Carafoli, Calcineurin controls the transcription of Na+/Ca2+ exchanger isoforms in developing cerebellar neurons, J. Biol. Chem. 275 (2000) 20903–20910.
783
[24] S.J. Roberts-Thomson, N.A. Holman, F.J. May, W.J. Lee, G.R. Monteith, Development of a real-time RT-PCR assay for plasma membrane calcium ATPase isoform 1 (PMCA1) mRNA levels in a human breast epithelial cell line, J. Pharmacol. Toxicol. Methods 44 (2000) 513–517. [25] T.M. Mamic, N.A. Holman, S.J. Roberts-Thomson, G.R. Monteith, PMCA1 mRNA expression in rat aortic myocytes: a real-time RT-PCR study, Biochem. Biophys. Res. Commun. 276 (2000) 1024–1027. [26] P.A. Newcomb, B.E. Storer, M.P. Longnecker, R. Mittendorf, E.R. Greenberg, R.W. Clapp, K.P. Burke, W.C. Willett, B. MacMahon, Lactation and a reduced risk of premenopausal breast cancer, N. Engl. J. Med. 330 (1994) 81–87. [27] L. Hennighausen, G.W. Robinson, K.U. Wagner, X. Liu, Developing a mammary gland is a stat affair, J. Mammary Gland Biol. Neoplasia 2 (1997) 365–372. [28] X. Liu, G.W. Robinson, K.U. Wagner, L. Garrett, A. WynshawBoris, L. Hennighausen, Stat5a is mandatory for adult mammary gland development and lactogenesis, Genes Dev. 11 (1997) 179– 186. [29] I. Cotarla, S. Ren, Y. Zhang, E. Gehan, B. Singh, P.A. Furth, Stat5a is tyrosine phosphorylated and nuclear localized in a high proportion of human breast cancers, Int. J. Cancer 108 (2004) 665–671. [30] L. Shan, M. Yu, B.D. Clark, E.G. Snyderwine, Possible role of Stat5a in rat mammary gland carcinogenesis, Breast Cancer Res. Treat. 88 (2004) 263–272.
Plasma membrane calcium-ATPase 2 and 4 in human breast cancer cell lines q Won Jae Lee, Sarah J. Roberts-Thomson, Gregory R. Monteith * The School of Pharmacy, The University of Queensland, Brisbane, Qld 4072, Australia Received 16 September 2005 Available online 28 September 2005
Abstract There is evidence to suggest that plasma membrane Ca2+-ATPase (PMCA) isoforms are important mediators of mammary gland physiology. PMCA2 in particular is upregulated extensively during lactation. Expression of other isoforms such as PMCA4 may influence mammary gland epithelial cell proliferation and aberrant regulation of PMCA isoform expression may lead or contribute to mammary gland pathophysiology in the form of breast cancers. To explore whether PMCA2 and PMCA4 expression may be deregulated in breast cancer, we compared mRNA expression of these PMCA isoforms in tumorigenic and non-tumorigenic human breast epithelial cell lines using real time RT-PCR. PMCA2 mRNA has a higher level of expression in some breast cancer cell lines and is overexpressed more than 100-fold in ZR-75-1 cells, compared to non-tumorigenic 184B5 cells. Although differences in PMCA4 mRNA levels were observed between breast cell lines, they were not of the magnitude observed for PMCA2. We conclude that PMCA2 mRNA can be highly overexpressed in some breast cancer cells. The significance of PMCA2 overexpression on tumorigenicity and its possible correlation with other properties such as invasiveness requires further study. Ó 2005 Elsevier Inc. All rights reserved. Keywords: Calcium; PMCA; Mammary gland; Lactation; Breast cancer; mRNA; Transcription
Free intracellular calcium (Ca2+) controls many essential cellular processes. The fidelity of Ca2+ signaling depends on homeostatic mechanisms that maintain steep Ca2+ gradients across biological membranes [1–3]. Mechanisms that lower the intracellular Ca2+ concentration ([Ca2+]i) prevent excessive rises in [Ca2+]i and thus circumvent cellular toxicity [1–3]. Plasma membrane Ca2+-ATPases (PMCAs) represent one such mechanism and belong to a family of P-type Ca2+-ATPases that specifically and actively transport Ca2+ [4]. PMCAs are important modulators of intracellular Ca2+ signals as they can affect the
q This work was supported by grants from the Queensland Cancer Fund (Q34-02), the National Health and Medical Research Council (NHMRC) of Australia (102426), and a Dora Lush (Biomedical) Research Scholarship awarded to W.J.L. by the NHMRC. * Corresponding author. Fax: +61 7 3365 1688. E-mail address: [email protected] (G.R. Monteith).
0006-291X/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2005.09.119
dynamics of Ca2+ signals by modulating amplitude and duration [5,6]. PMCA isoforms (PMCA1–4) are encoded by genes located on separate chromosomes [4]. PMCA1 and PMCA4 are expressed ubiquitously while PMCA2 and PMCA3 are generally found in excitable tissues [7]. PMCAs differ in their biochemical properties, with PMCA2 having the greatest affinity for calmodulin, which is the main activator of PMCA-mediated Ca2+ transport [8,9]. The distinctive features of PMCAs suggest that the expression of different PMCA isoforms is matched to suit the requirements for maintaining Ca2+ homeostasis in particular cell types or tissues and thus support the cell or tissueÕs cellular and physiological functions [10]. The mammary gland experiences extensive Ca2+ flux during lactation [11,12], yet our understanding of the relative importance of PMCAs in milk production and their potential involvement in other key processes in mammary gland
780
W.J. Lee et al. / Biochemical and Biophysical Research Communications 337 (2005) 779–783
physiology is limited. Rat mammary gland expresses PMCA1, 2, and 4 [13,14]. PMCA2 expression undergoes significant induction after parturition, then rapidly increases as lactation progresses and remains elevated at levels that are almost 100-fold higher compared to levels during pregnancy [13,14]. Moreover, null mutations in PMCA2 substantially lower milk Ca2+ concentrations in lactating mice [15]. Thus, there is strong evidence supporting PMCA2 as a major mechanism for the enrichment of Ca2+ into milk and thus extracellular Ca2+ homeostasis. Dynamic and coordinated regulation of PMCA isoform expression appears to be a crucial regulatory mechanism for supporting the unique physiology of the mammary gland [13,14]. Other PMCA isoforms may also control breast development, architecture, and post-lactational involution in humans, since particular PMCA isoforms mediate proliferation, differentiation, and apoptosis in other cell culture model systems [16–18]. Little is known about exactly how PMCA isoforms are regulated in the mammary gland to meet different physiological circumstances, for example, formation of glandular structures versus secretory acini. Deregulation of coordinated and dynamic PMCA isoform expression in the mammary gland is hypothesized to contribute to pathological consequences such as mammary gland tumorigenesis. We previously reported that the relative expression of PMCA1 mRNA is significantly higher in tumorigenic MCF-7 and MDA-MB-231 human breast cancer cell lines, compared to non-tumorigenic MCF-10A human breast epithelial cells [19]. We have also shown that expression of stable antisense targeting PMCA inhibits the proliferation of MCF-7 cells [20]. To further explore the plausibility that deregulation of PMCA isoform expression is characteristic of mammary gland tumorigenesis and given that transcriptional modulation represents the most likely means for PMCA regulation in the course of changing physiological demands [21–23], we compared the levels of PMCA2 and PMCA4 mRNA in a bank of tumorigenic and non-tumorigenic human breast epithelial cell lines. Materials and methods Cell culture. The tumorigenic human breast cancer cell lines MCF-7 and MDA-MB-231, and the non-tumorigenic human breast epithelial cell line MCF-10A were cultured and plated for total RNA isolations as previously described [19]. The tumorigenic human breast cancer cell lines ZR-75-1, T-47D, BT-483, and SK-BR-3, and the non-tumorigenic human breast epithelial cell lines 184A1 and 184B5 were a generous gift from Professor Rob Sutherland (Garvan Institute, Sydney, Australia). The aforementioned tumorigenic cell lines were grown in basal DMEM containing phenol red (JRH Biosciences, Brooklyn, Victoria, Australia), and supplemented with 10% FBS (JRH Biosciences), 2 mM L-glutamine (Invitrogen, Mount Waverly, Victoria, Australia) and 100 U/ml–100 lg/ ml penicillin G–streptomycin (Invitrogen). 184A1 and 184B5 cells were cultured in phenol red containing basal MCDB 170 medium (Invitrogen), with the inclusion of 100 U/ml–100 lg/ml penicillin G–streptomycin and serum-free supplements provided by the manufacturer that were used at the recommended dilution (Invitrogen). ZR-75-1, T-47D, BT-483, SKBR-3, 184A1, and 184B5 cells were routinely passaged once a week or when approximately 90% confluent and subcultured into T-75 cm2 or T-
150 cm2 cell culture flasks (in the case of the non-tumorigenic cells) at a split ratio of 1:5–1:10. Cultures were harvested for isolation of total RNA when they reached confluence. All cell lines were maintained in a humidified 5% CO2/95% air incubator at 37 °C. RNA isolation. Total RNA was isolated using RNeasy mini or midi columns (Qiagen, Doncaster, Victoria, Australia) according to the manufacturerÕs instructions. This procedure included an on-column DNA digestion step using RNase-free DNase I (Qiagen). Total RNA concentrations were quantified by measuring UV absorbance at 260 nm. Real time RT-PCR. Taqman real time RT-PCR assays were used to determine the expression of PMCA2 and PMCA4 mRNA, relative to the 18S rRNA. For the relative quantification of PMCA2 mRNA (GenBank Accession No. NM_001683) primers and probe were designed using Primer Express software, version 1.5 (Applied Biosystems, Scoresby, Victoria, Australia). The primers had the following sequences: forward primer 5 0 -ATCACCCGGCAGCCTCTT-3 0 and reverse primer 5 0 TTGGAGTAAAAGTCGCTGTTGGT-3 0 , and probe sequence was: 5 0 CCGAGCAAGGACCG-3 0 . The Taqman Gene Expression Assay Hs00608066_m1 (Applied Biosystems) was used to quantitate PMCA4 mRNA, and primers and probe for 18S rRNA were also purchased from Applied Biosystems. For all reactions quantitating 18S rRNA, primers and probe were included at 50 nM and were cycled in separate tubes at a 1:1000 dilution of total RNA using the same method as for each specific target. PMCA2 mRNA was quantitated using a one-step RT-PCR procedure and the TaqMan one-step RT-PCR master mix reagents kit (Applied Biosystems), primers (900 nM), probe (200 nM), and total RNA (100 ng for PMCA2, 0.1 ng for 18S rRNA). Reactions were cycled using an ABI PRISM 7700 Sequence Detector (Applied Biosystems) with the following thermal cycling conditions: 48 °C for 30 min, 95 °C for 10 min, followed by 40 cycles at 95 °C for 15 s, and 60 °C for 1 min. First strand cDNA synthesis required for real time PCR detection of the PMCA4 amplicon was prepared using the Omniscript RT kit according to the vendorÕs instructions (Qiagen). Real time PCR amplifications for PMCA4 mRNA and 18S rRNA amplicons were then set up as separate reactions in TaqMan universal PCR master mix (Applied Biosystems). Each tube for real time PCR detection of PMCA4 mRNA contained cDNA template equivalent to 25 ng of total RNA and for 18S rRNA 0.025 ng total RNA. Thermal cycling was completed using an Applied Biosystems 7500 Real Time PCR System with 95 °C for 10 min and then 40 cycles at 95 °C for 15 s and 60 °C for 1 min. Data and statistical analysis. Analysis of PMCA2 mRNA real time RTPCR data was carried out using the comparative CT method [19,24,25]. PMCA4 mRNA real time RT-PCR data were analyzed using the relative standard curve method. See User Bulletin #2: ABI PRISM 7700 Sequence Detection System December 11, 1997 (updated 10/2001) from Applied Biosystems for further information. Data points represent the mean ± the standard error of the mean (SEM). N values are given in the figure legends for each corresponding graph. Means were compared with one-way analysis of variance (ANOVA) and subsequent TukeyÕs pairwise comparison tests using Prism Version 4.03 (GraphPad Software, San Diego, CA). Differences were considered to be statistically significant if P values were less than 0.05 (P < 0.05).
Results and discussion PMCA2 and PMCA4 mRNA expression in MCF-7, MDAMB-231, and MCF-10A cells Breast cancer cell lines express multiple PMCA mRNA isoforms and our laboratory has previously observed that relative PMCA1 mRNA expression is greater in tumorigenic MCF-7 and MDA-MB-231 cells compared to nontumorigenic MCF-10A cells [19]. To see if there are also differences in the expression of other PMCA mRNA
W.J. Lee et al. / Biochemical and Biophysical Research Communications 337 (2005) 779–783
PMCA2 and PMCA4 mRNA expression in other breast epithelial cell lines
9
Subconfluent
*
Confluent
*
8 7 6 5 4 3 2 1 0 -1 -2
MCF-7
MDA-MB-231
MCF-10A
Fig. 1. Relative PMCA2 mRNA levels in MCF-7, MDA-MB-231, and MCF-10A cells, normalized to confluent MCF-10A cells and expressed as DDCT values. Bars represent the mean ± SEM (n = 4, 4, 3, 3, 4, and 4 for samples left to right). The asterisk (*) denotes a significant difference compared to confluent MCF-10A cells using a TukeyÕs test.
2.5 Subconfluent
10
2.0
1.5
1.0
9 8 7
*
6
*
5 4 3 2 1 0 -1
0.5
MCF-7
MDA-MB-231
MCF-10A
Fig. 2. Relative PMCA4 mRNA levels in MCF-7, MDA-MB-231, and MCF-10A cells, normalized to confluent MCF-10A cells and expressed as fold change. Bars represent the mean ± SEM (n = 4 for all samples). No significant differences were found between all pairwise comparisons using TukeyÕs tests.
5 4B 18
18
4A
1
-3 -B R
SK
T4
D 47 T-
5 -7 ZR
0.0
83
-2
-1
Relative PMCA4 mRNA levels (fold change, normalized to confluent MCF-10A)
Confluent
*
11
Relative PMCA2 mRNA levels (-∆∆CT, normalized to 184B5)
isoforms in the above cell lines, we performed real time RT-PCR for PMCA2 and PMCA4 mRNA. Fig. 1 shows the results for relative PMCA2 mRNA expression in subconfluent and confluent cultures of MCF-7, MDA-MB231, and MCF-10A cells. Relative PMCA2 mRNA levels within a cell line did not differ significantly with the state of confluence. PMCA2 mRNA expression was, however, significantly greater in the confluent breast cancer cell lines MCF-7 and MDA-MB-231 compared to confluent normal breast MCF-10A cells (Fig. 1). In contrast, relative PMCA4 mRNA expression in subconfluent and confluent MCF-7, MDA-MB-231, and MCF-10A cells showed no significant differences either within or between cell lines (Fig. 2).
We then assessed relative PMCA2 and PMCA4 mRNA levels in a larger array of tumorigenic ZR-75-1, T-47D, BT-483, and SK-BR-3 cells and two other nontumorigenic 184A1 and 184B5 cell lines. Non-tumorigenic 184A1 and 184B5 cells did not differ significantly from each other in their relative expression of PMCA2 mRNA (Fig. 3). Except for SK-BR-3 cells, relative PMCA2 mRNA expression was significantly higher in all the tumorigenic cell lines tested, when compared to 184B5 cells (Fig. 3). Different levels of PMCA2 mRNA expression were observed within the group of tumorigenic cell lines examined, with ZR-75-1 cells expressing the highest level of PMCA2 mRNA. The ZR-75-1 breast cancer cells express a more than 100-fold greater level of PMCA2 mRNA compared to the normal 184B5 cell line ðfold change ¼ 2DDCT Þ. Whereas housekeeping PMCA isoforms such as PMCA4 may predominately influence processes like pregnancy-induced mammary gland development or cell proliferation in the breast [14,20], the evidence for PMCA2 in the mammary gland suggests that PMCA2 is primarily a mechanism for bulk Ca2+ transport during lactation for the enrichment of milk with Ca2+ [13–15]. Since lactation is protective against breast cancer in premenopausal women [26], the finding that PMCA2 is overexpressed in tumorigenic cell lines (Figs. 1 and 3) as well as being present in ZR-75-1 at levels 100-fold greater than the non-tumorigenic
B
Relative PMCA2 mRNA levels (-∆∆CT, normalized to MCF-10A confluent)
11 10
781
Fig. 3. Relative PMCA2 mRNA levels in an extended panel of tumorigenic (ZR-75-1, T-47D, BT-483, and SK-BR-3) and non-tumorigenic (184A1 and 184B5) breast epithelial cell lines. Bars represent the mean ± SEM (n = 3 for all samples) and are indicative of two independent experiments. The asterisk (*) denotes a significant difference compared to 184B5 cells using a TukeyÕs test.
782
W.J. Lee et al. / Biochemical and Biophysical Research Communications 337 (2005) 779–783
cell lines is intriguing and suggests that some breast cancers may be characterized by the overexpression of calcium transporters associated with lactation. Indeed, Stat5a is a protein involved in mammary gland development and milk protein expression during lactation [27,28] that may also be involved in breast cancer [29,30]. Our studies indicate that the increased levels of PMCA2 in some breast cancer cell lines are not correlated with estrogen receptor status, since the estrogen receptor negative cell line MDA-MB-231 had elevated levels of PMCA2. PMCA4 mRNA expression in all the cell lines examined varied modestly with a narrower dynamic range than that seen for PMCA2 mRNA. Indeed, the greatest difference between two breast cell lines for PMCA4 was approximately 8-fold, whereas for PMCA2 it was more than 100-fold. However, there was a modest but a significantly lower level of PMCA4 mRNA in the tumorigenic samples when compared to the 184B5 non-tumorigenic cell line (Fig. 4). The differences in PMCA4 mRNA levels seen in this study could be of functional importance, given that a reduction in PMCA4 expression is correlated with an inhibition of MCF-7 breast cancer cell line proliferation [20] and may suggest a slightly greater sensitivity to selective PMCA4 inhibition in tumorigenic cell lines. However, the modest differences between cell lines in PMCA4 may reflect a more housekeeping role for PMCA4 compared with PMCA2, which has a highly restricted tissue distribution [7] and is the PMCA isoform with the greatest change in expression associated with lactation [13,14]. The potential biological significance of PMCA2 overexpression in breast tumorigenesis as well as its impact on 2.5 Relative PMCA4 mRNA levels (fold change, normalized to 184B5)
* 2.0
1.5
1.0
*
*
0.5
*
* 5 18
4B
1 4A 18
-3 K S
B
T-
-B
48
R
3
D 47 T-
ZR
-7
5-
1
0.0
Fig. 4. Relative PMCA4 mRNA levels in an extended panel of tumorigenic (ZR-75-1, T-47D, BT-483, and SK-BR-3) and non-tumorigenic (184A1 and 184B5) breast epithelial cell lines. Bars represent the mean ± SEM (n = 3 for all samples) and are indicative of two independent experiments. The asterisk (*) denotes a significant difference compared to 184B5 cells using a TukeyÕs test.
Ca2+ signaling in mammary gland epithelial cells is uncertain. Our study does further highlight the isoform-specific differences in PMCA expression in epithelial cells of the mammary gland. Future studies are required to determine if relative levels of PMCA isoforms correlate to clinical outcomes or sensitivity to specific treatment regimes. Further studies are also required to examine the mechanisms of PMCA isoform transcriptional and translational regulation particularly in mammary gland epithelial cells during lactation and breast cancer. Acknowledgments We thank Dr. Jodie Robinson and Nicola Holman for assistance with methodology. References [1] E. Carafoli, Intracellular calcium homeostasis, Annu. Rev. Biochem. 56 (1987) 395–433. [2] E. Carafoli, L. Santella, D. Branca, M. Brini, Generation, control, and processing of cellular calcium signals, Crit. Rev. Biochem. Mol. Biol. 36 (2001) 107–260. [3] M.J. Berridge, P. Lipp, M.D. Bootman, The versatility and universality of calcium signalling, Nat. Rev. Mol. Cell Biol. 1 (2000) 11–21. [4] E. Carafoli, Biogenesis: plasma membrane calcium ATPase: 15 years of work on the purified enzyme, FASEB. J. 8 (1994) 993–1002. [5] M. Brini, D. Bano, S. Manni, R. Rizzuto, E. Carafoli, Effects of PMCA and SERCA pump overexpression on the kinetics of cell Ca(2+) signalling, EMBO. J. 19 (2000) 4926–4935. [6] M. Brini, L. Coletto, N. Pierobon, N. Kraev, D. Guerini, E. Carafoli, A comparative functional analysis of plasma membrane Ca2+ pump isoforms in intact cells, J. Biol. Chem. 278 (2003) 24500–24508. [7] E.E. Strehler, D.A. Zacharias, Role of alternative splicing in generating isoform diversity among plasma membrane calcium pumps, Physiol. Rev. 81 (2001) 21–50. [8] H. Hilfiker, D. Guerini, E. Carafoli, Cloning and expression of isoform 2 of the human plasma membrane Ca2+ ATPase. Functional properties of the enzyme and its splicing products, J. Biol. Chem. 269 (1994) 26178–26183. [9] N.L. Elwess, A.G. Filoteo, A. Enyedi, J.T. Penniston, Plasma membrane Ca2+ pump isoforms 2a and 2b are unusually responsive to calmodulin and Ca2+, J. Biol. Chem. 272 (1997) 17981–17986. [10] G.R. Monteith, B.D. Roufogalis, The plasma membrane calcium pump-a physiological perspective on its regulation, Cell Calcium 18 (1995) 459–470. [11] D.B. Shennan, M. Peaker, Transport of milk constituents by the mammary gland, Physiol. Rev. 80 (2000) 925–951. [12] R.L. Horst, J.P. Goff, T.A. Reinhardt, Calcium and vitamin D metabolism during lactation, J. Mammary Gland Biol. Neoplasia 2 (1997) 253–263. [13] T.A. Reinhardt, R.L. Horst, Ca2+-ATPases and their expression in the mammary gland of pregnant and lactating rats, Am. J. Physiol. 276 (1999) C796–C802. [14] T.A. Reinhardt, A.G. Filoteo, J.T. Penniston, R.L. Horst, Ca(2+)ATPase protein expression in mammary tissue, Am. J. Physiol. Cell Physiol. 279 (2000) C1595–C1602. [15] T.A. Reinhardt, J.D. Lippolis, G.E. Shull, R.L. Horst, Null mutation in the gene encoding plasma membrane Ca2+-ATPase isoform 2 impairs calcium transport into milk, J. Biol. Chem. 279 (2004) 42369–42373. [16] M. Husain, L. Jiang, V. See, K. Bein, M. Simons, S.L. Alper, R.D. Rosenberg, Regulation of vascular smooth muscle cell proliferation by plasma membrane Ca(2+)-ATPase, Am. J. Physiol. 272 (1997) C1947–C1959.
W.J. Lee et al. / Biochemical and Biophysical Research Communications 337 (2005) 779–783 [17] P.C. Brandt, J.E. Sisken, R.L. Neve, T.C. Vanaman, Blockade of plasma membrane calcium pumping ATPase isoform I impairs nerve growth factor-induced neurite extension in pheochromocytoma cells, Proc. Natl. Acad. Sci. USA 93 (1996) 13843–13848. [18] K. Paszty, A.K. Verma, R. Padanyi, A.G. Filoteo, J.T. Penniston, A. Enyedi, Plasma membrane Ca2+ ATPase isoform 4b is cleaved and activated by caspase-3 during the early phase of apoptosis, J. Biol. Chem. 277 (2002) 6822–6829. [19] W.J. Lee, S.J. Roberts-Thomson, N.A. Holman, F.J. May, G.M. Lehrbach, G.R. Monteith, Expression of plasma membrane calcium pump isoform mRNAs in breast cancer cell lines, Cell Signal 14 (2002) 1015–1022. [20] W.J. Lee, J.A. Robinson, N.A. Holman, M.N. McCall, S.J. RobertsThomson, G.R. Monteith, Antisense-mediated inhibition of the plasma membrane calcium-ATPase suppresses proliferation of MCF-7 cells, J. Biol. Chem. 280 (2005) 27076–27084. [21] D. Guerini, E. Garcia-Martin, A. Gerber, C. Volbracht, M. Leist, C.G. Merino, E. Carafoli, The expression of plasma membrane Ca2+ pump isoforms in cerebellar granule neurons is modulated by Ca2+, J. Biol. Chem. 274 (1999) 1667–1676. [22] D. Guerini, X. Wang, L. Li, A. Genazzani, E. Carafoli, Calcineurin controls the expression of isoform 4CII of the plasma membrane Ca(2+) pump in neurons, J. Biol. Chem. 275 (2000) 3706–3712. [23] L. Li, D. Guerini, E. Carafoli, Calcineurin controls the transcription of Na+/Ca2+ exchanger isoforms in developing cerebellar neurons, J. Biol. Chem. 275 (2000) 20903–20910.
783
[24] S.J. Roberts-Thomson, N.A. Holman, F.J. May, W.J. Lee, G.R. Monteith, Development of a real-time RT-PCR assay for plasma membrane calcium ATPase isoform 1 (PMCA1) mRNA levels in a human breast epithelial cell line, J. Pharmacol. Toxicol. Methods 44 (2000) 513–517. [25] T.M. Mamic, N.A. Holman, S.J. Roberts-Thomson, G.R. Monteith, PMCA1 mRNA expression in rat aortic myocytes: a real-time RT-PCR study, Biochem. Biophys. Res. Commun. 276 (2000) 1024–1027. [26] P.A. Newcomb, B.E. Storer, M.P. Longnecker, R. Mittendorf, E.R. Greenberg, R.W. Clapp, K.P. Burke, W.C. Willett, B. MacMahon, Lactation and a reduced risk of premenopausal breast cancer, N. Engl. J. Med. 330 (1994) 81–87. [27] L. Hennighausen, G.W. Robinson, K.U. Wagner, X. Liu, Developing a mammary gland is a stat affair, J. Mammary Gland Biol. Neoplasia 2 (1997) 365–372. [28] X. Liu, G.W. Robinson, K.U. Wagner, L. Garrett, A. WynshawBoris, L. Hennighausen, Stat5a is mandatory for adult mammary gland development and lactogenesis, Genes Dev. 11 (1997) 179– 186. [29] I. Cotarla, S. Ren, Y. Zhang, E. Gehan, B. Singh, P.A. Furth, Stat5a is tyrosine phosphorylated and nuclear localized in a high proportion of human breast cancers, Int. J. Cancer 108 (2004) 665–671. [30] L. Shan, M. Yu, B.D. Clark, E.G. Snyderwine, Possible role of Stat5a in rat mammary gland carcinogenesis, Breast Cancer Res. Treat. 88 (2004) 263–272.