Tyrosine Kinases As a Strategy to Abrogate Antiestrogen Resistance in Human Breast Cancer 1

4436s Vol. 7, 4436s-4442s, December 2001 (Suppl.) Clinical Cancer Research

Inhibition of erbB Receptor (HER) Tyrosine Kinases as a Strategy to Abrogate Antiestrogen Resistance in Human Breast Cancer 1

Hirokazu Kurokawa and Carlos L. Arteaga 2 duced growth of established MCF-7/HER2 xenografts in Departments of Medicine and Cancer Biology and Vanderbilt-Ingram athymic nude mice. Similar results have been obtained with Cancer Center, Vanderbilt University School of Medicine, Nashville, ZD1839 ("Iressa"), an epidermal growth factor receptor Tennessee 37232 (HER1) tyrosine kinase inhibitor. Taken together, these data suggest that exogenous inhibitors of the HER-signaling net- Abstract work and other mitogenic pathways can abrogate or delay the emergence of antiestrogen resistance, thus providing an It has been proposed that binding of ligand to the estrogen receptor (ER) releases its association with tran- evaluable therapeutic strategy in human breast carcinoma. scriptional corepressors, allowing the ER to recruit coactivators, which possess histone acetylase activity, and induce Introduction transcription of gene promoters containing estrogen re- The classical ER 3 oL belongs to the superfamily of nuclear sponse elements. It has also been proposed that the anties- steroid receptors that function as transcription factors on binding trogen tamoxifen recruits transcriptional corepressors to the by their respective ligands (1). The ER contains 595 amino acids AF-2 region of the hormone-binding domain of the ER, thus blocking ER-mediated transcription. The ER cross-talks with a central DNA-binding domain and a COOH-terminal, with a number of mitogenic signaling pathways and second hormone-binding domain. More recently, a smaller 530-amino messengers, like the epidermal growth factor receptor, the acid ER~3 was cloned (2). Binding of estradiol to ER induces insulin-like growth factor-I receptor, mitogen-activated pro- conformational changes in the receptor and ER dimerization tein (MAP) kinase, phosphatidylinositol-3 kinase/Akt, do- with subsequent high-affinity binding to DNA through their pamine, and cyclic AMP. Some of these molecules may: (a) DNA-binding domain at specific EREs present in the promoter support ligand-independent ER transcription; (b) increase of estrogen-responsive genes. ER-induced gene transcription is the association of ER with coactivators of transcription; mediated by two distinct transactivation domains: the hormone- and/or (c) reduce the antiestrogen-induced association of ER independent AF-1 region in the NH 2 terminus and the hormone- with corepressors. These events either alone or in combina- dependent AF-2 region in the COOH-terminal region of the tion may result in hormone independence and/or antiestro- receptor (3, 4). Studies with ER mutants have shown that, gen resistance. We have examined whether signaling by although AF-1 and AF-2 are required for maximal ER transcrip- HER2/neu (erbB-2) receptor tyrosine kinase, which can in- tional activity in most cells, certain gene promoters, sometimes duce antiestrogen resistance, can also disrupt the tamoxifen- as a function of their cellular context, can be independently induced interaction of ER with transcriptional corepressors. transactivated by AF-1 or AF-2 alone (5). This promoter and/or Notably, tamoxifen-induced association of ER with the tran- cell dependence of ER-mediated transcription appears to be scriptional corepressors N-CoR or SMRT was reduced in modulated by the recruitment of coactivator and corepressor HER2-overexpressing breast tumor cells but not in cells proteins, which serve as signaling intermediates between the with low HER2 levels. Small molecule inhibitors of the DNA-bound receptor dimer and the basal transcriptional ma- HER2 kinase or MAP extracellular signal-regulated kinase chinery (Refs. 1, 6, and references therein). These coactivators 1/2 or dominant-negative MAP extracellular signal-regu- and corepressors of transcription possess intrinsic histone acety- lated kinase 1/2 constructs restored the inhibitory effect of lase or histone deacetylase activities, respectively. Coactivators tamoxifen on both ER-mediated transcription and tumor can acetylate histones, loosen their interaction with DNA and, cell proliferation. Treatment with both tamoxifen and the therefore, expose motifs required for gene transcription. On the small molecule HER1/2 kinase inhibitor AG1478 reduced other hand, corepressor molecules can silence steroid-respon- mitogen-activated protein kinase activity and markedly re- sive genes by binding to the unliganded receptor, deacetylating histones, and potentially keeping the DNA in a conformation that occludes coactivator binding sites. Presented at the First International Conference on Recent Advances and Future Directions in Endocrine Therapy for Breast Cancer, June 21-23, 2001, Cambridge, MA. Supported in part by NIH Grant R01 CA80195, Department of Defense Breast Cancer Research Program Grant DAMD 17-98-1-8262, and Vanderbilt-Ingram Cancer Center sup- 3 The abbreviations used are: ER, estrogen receptor; ERE, estrogen port Grant CA68485. response element; SERM, selective estrogen receptor modulator; EGFR, 2 To whom requests for reprints should be addressed, at Division of epidermal growth factor receptor; EGF, epidermal growth factor; Hematology-Oncology, Vanderbitt University School of Medicine, 777 MAPK, mitogen-activated protein kinase; PI3K, phosphatidylinositol-3 Preston Residential Building, Nashville, TN 37232-6307. Phone: (615) kinase; MEKK, mitogen-activated protein kinase kinase kinase; MEK, 936-3524; Fax: (615) 936-1790; E-mail: carlos.arteaga@mcmail. mitogen-activated protein/extracellular signal-regulated kinase kinase; vanderbilt.edu. JNK, c-Jun-NHz-terminal kinase.

Molecular Basis for the ER Agonist/Antagonist Role A SERM sensitive estrogen of Tamoxifen The antiestrogen tamoxifen antagonizes the transcription of genes that require the AF-2 region for ER-mediated transactivation. However, for gene promoters in which AF-2 function is dispensable and whose transcription is driven by the AF-1 Coactlvatom 7~s domain, tamoxifen can function as an agonist (5, 7). This ability of tamoxifen to behave as an AF-2 antagonist and AF-1 agonist NH2-I A/B I C I D I ElF I.COOH explains its antagonistic effect in breast epithelium, as well as its agonistic effect in other tissues, such as endometrium, bone, and AF-1 DBD AF-21HBD liver. In addition to blocking AF-2-dependent gene transactivation, in some experimental systems, tamoxifen has been shown I transcription to repress ER-mediated transcription by recruiting corepressor proteins to the receptor (8, 9). In other experiments, overexpres- B SERM resistant sion of transcriptional coactivators, like SRC-1 or L7/SPA, results in enhancement of tamoxifen-stimulated ER reporter (~j estrogen activity (8, 10). Because of their ability to either repress or activate the transcription of estrogen target genes, mixed antag- tamoxifen / onist/agonists like tamoxifen are now more appropriately re- ferred to as SERMs. The data summarized above imply that the net cellular response to SERMs may depend on the tissue repertoire, avail- NH2-1 AJB I C IDI ElF I_COOH ability, and/or relative amounts of transcriptional coactivators and corepressors. Therefore, high levels of coactivators and/or low levels of corepressors may be permissive for the agonist AF-1 DBD AF-21HBD activity of SERMs. These results suggest that the recruitment of I coactivators or corepressors in different ratios to the ER-ERE ~- transcription complex could introduce a potential element of transcriptional Fig. 1 A molecular hypothesis to support sensitivity or resistance to control and diversity that may be more marked in some estro- SERMs. In tamoxifen-sensitive tumor cells (A), corepressors predomi- nate over coactivators in tamoxifen-occupied ERs. In this setting, the gen-responsive genes than others. Finally, these data have led to tamoxifen-mediated inhibitory effect on AF-2 is dominant, resulting in the evaluable hypothesis that de novo or acquired abundance of a blockade of hormone-induced gene transcription. With progression to coactivators and/or the reduction or loss of transcriptional core- SERM resistance (B), the ER coactivator:corepressor ratio is reversed, pressors in breast tumor cells may result in hormone independ- resulting in a loss of tamoxifen-mediated inhibition of AF-2 and/or ence and antiestrogen resistance (11). This hypothesis, illus- unmasking of its agonistic effect on AF-1 function (modified from Takimoto et al., Ref. 11). trated in Fig. 1, is supported by a recent experiment reported by Lavinsky et al. (9). On prolonged treatment with tamoxifen, ER-positive MCF-7 xenografts in athymic nude mice developed resistance to treatment and grew during continuous tamoxifen well as that induced by tamoxifen-occupied ERs (13). Neuro- therapy. Tumor levels of N-CoR were reduced in the tamoxifen- transmitters like dopamine or SKF-38393 can induce ER re- resistant xenografts, suggesting that reduced levels of the core- porter activity in the absence of added steroids (14). Exogenous pressor may enable tamoxifen to stimulate ER-mediated tran- EGF or overexpression of the EGFR-homologous HER2/neu scription through unopposed AF-1 function (9). Interestingly, (erbB2) proto-oncogene can reduce tamoxifen-induced associ- 4-hydroxytamoxifen failed to activate ER-dependent transcrip- ation of ER with corepressors and thus limit the growth-inhib- tion in wild-type mouse embryonic fibroblasts but acted as a full itory action of tamoxifen against breast carcinoma cells (9, 15). agonist in N-CoR-null mouse embryonic fibroblasts. Introduc- Arouica et al. (16) reported that insulin-like growth factor-I is tion of an N-CoR expression vector was sufficient to reverse this also a potent activator of ERa and that this response could be effect (12), further implying that N-CoR is required for the blocked by pure antiestrogens, such as ICI 164,384. Activation ER-antagonistic effect of 4-hydroxytamoxifen and, second, that of the Ras/MAPK signaling pathway, a frequent alteration in the ability of tamoxifen to function as an ER agonist may transformed cells, has been shown to phosphorylate Serll8 in require functional inactivation of N-CoR. the ER, leading to hormone-independent activation of the ER with loss of the inhibitory effect of tamoxifen on ER-mediated Cross-Talk between ER and Peptide Growth Factor transcription (17, 18). Of note, EGF blocked the tamoxifen- Receptor Signaling Pathways induced association of wild-type ER with N-CoR but failed to A number of mitogenic signaling pathways have been block the association of a mutant S 118A ER with the transcrip- shown to interact with ER transcriptional complexes and induce tional corepressor (9), implicating EGF-activated MAPK as the hormone-independent ER activation or, in some cases, limit or signal mediating the inhibition of tamoxifen action. Overexpres- reverse tamoxifen-induced repression of ER transcription. Ac- sion of PI3K and the PI3K-dependent serine/threonine kinase tivators of protein kinase A or inhibitors of protein phosphatases Akt activates ERa in the absence of added estrogen (19). Al- are able to stimulate hormone-independent ER transcription, as though PI3K increased the activity of both AF-1 and AF-2

regions, Akt enhanced the activity of only the hormone-indepen- tumor cells. MCF-7 human breast cancer cells transfected with dent AF-1 domain. Moreover, Akt overexpression protected either a full-length HER2 cDNA or with ectopic heregulin-131, MCF-7 breast cancer cells from tamoxifen-induced apoptosis (19). the HER3/4 ligand that activates HER2, lose sensitivity to Other nonreceptor kinases downstream of neurotransmitter tamoxifen or estrogen dependence (15, 25-27). Several clinical or (HER) erbB receptors have been implicated in dysregulation studies have shown that metastatic tumors with high HER2 of ER function. Activation of the Src nonreceptor tyrosine expression and/or high circulating levels of the HER2 ectodo- kinase, which is elevated in the majority of human breast can- main exhibit a statistically lower clinical response rate and/or a cers, markedly enhances the transcriptional activity of the unli- shorter duration of response to antiestrogen therapy (28-36), ganded ER and the tamoxifen-ER complex. This effect appeared further associating high tumor levels of the proto-oncogene with in part mediated by the ability of Src to increase the AF-1- resistance to endocrine therapy. Two studies, however, failed to coactivating function of CREB-binding protein and the glu- show a reduced clinical response to tamoxifen in HER2-over- cocorticoid receptor-interacting protein (20). Expression of expressing tumors (37, 38). The first of these studies, however, wild-type MEKK1 and an active Racl increased the activity of did show a statistically shorter time to tamoxifen treatment the unliganded ER. Moreover, expression of constitutively ac- failure in patients bearing cancers with -->10% HER2-positive tive MEKK1 increased the agonistic activity of 4-hydroxyta- ceils by immunohistochemistry compared with patients whose moxifen to a level comparable to that of 17[3-estradiol and fully tumors exhibited lower levels of HER2 (37). In the second blocked its antagonistic activity (21). Finally, phosphorylation study, patients with lymph node-positive, nonmetastatic breast of the SMRT corepressor by MEKK1 and MEK1 inhibited the cancer were assigned to tamoxifen therapy or not by physician ability of SMRT to tether to its transcription partners. These discretion after completion of standard adjuvant chemotherapy. kinases also induced a redistribution of SMRT from the nucleus The lack of randomization and the immediately preceding ther- to a perinuclear/cytosolic compartment, further contributing to apy potentially confound the rigorous interpretation of the re- loss of its corepressor function (22). These results suggest that suits of this trial (38). The discrepancies between these last two different elements of non-ER mitogenic signaling pathways can reports and the former group of studies (28-36) could be ex- modulate ER function in some cases by modifying steady-state plained by the variable and sometimes low number of patients in levels and/or the posttranslational state of coactivators/corepres- these studies, different antiestrogen therapies used, variable sors of ER transcription. In general, the identification of the methods to assess HER2 overexpression, and the failure to specific steroid-responsive genes whose expression is altered as rigorously exclude ER-negative tumors, among other potential a result of the cross-talk between ER and other signal transduc- inconsistencies. Nonetheless, a recent meta-analysis of ~ 1200 tion systems requires further investigation. Nonetheless, these breast cancer patients, 71% ER positive and 6.2% ER unknown, data suggest that, in the context of parallel oncogenic signaling showed a 2.8 odds ratio of treatment failure in tumors with high pathways, the estradiol- or SERM-bound ER-ERE-coactivator- levels of HER2 compared with low-expressing tumors, as de- corepressor complex can behave as a platform for multiple fined by clinical progression within 6 months of initiation of inputs leading to transcriptional diversity and the selection of endocrine therapy (39). hormone-independent phenotypes. Inhibition of HER2 and MAPKs Restores Tamoxifen HER (erbB) Tyrosine Kinase Receptors and Action against Antiestrogen-resistant Breast Tumor Cells Antiestrogen Resistance Two studies have reported the ability of antibodies against The HER2/neu (erbB2) transmembrane receptor tyrosine the ectodomain of HER2 to increase the inhibitory effect of kinase has been causally associated with hormone independence tamoxifen or the pure antiestrogen ICI 182,780 against breast and antiestrogen resistance of human breast cancer cells. This tumor cells with or without HER2 overexpression (40, 41). We kinase is the protein product of the HER2 proto-oncogene and a have extended these reports by studying the mechanisms by member of the EGF receptor (HER1, erbB 1) family of trans- which HER2 signaling confers antiestrogen resistance to ER- membrane tyrosine kinases, which also includes HER3 (erbB3) positive breast tumor cells. These results (15) will be summa- and HER4 (erbB4). On binding of ligands to the EGFR, HER3, rized next. Forced expression of HER2 in MCF-7 breast cancer or HER4, HER2 is recruited as the preferred partner of these cells resulted in MAPK hyperactivity and tamoxifen resistance. ligand-bound receptors into active, phosphorylated hetero- Inhibition of HER2 and MAPKs with the HER1/2 kinase inhib- dimeric complexes, which activate several signaling pathways itor AG1478 (42) and the MEK1/2 inhibitor U0126, respec- involved in the proliferation and survival of tumor cells (Ref. tively, as well as by transient transfection of dominant-negative 23, 24, and references therein). Cytoplasmic signal transducers MEK1/2 constructs, restored the inhibitory effect of tamoxifen induced by the HER network include PLC-v 1, Ras-Raf-MEK- on ER-reporter activity and colony survival. Similar results were MAPKs, PI3K and Akt, Src, the stress activated protein kinases, observed in BT-474 human breast cancer cells, which naturally PAK-JNKK-JNK, and the signal transducers and activators of exhibit HER2 gene amplification. Both AG1478 and U0126 also transcription (23, 24). Of note, aberrant activation of these mole- restored the tamoxifen-mediated association of ER with the cules has been shown to dysregulate ER transcription (above). corepressors N-CoR (15) and SMRT 4 in the antiestrogen-resist- HER2 can transform normal mammary epithelial cells and is over- ant MCF-7/HER2-18 cells. Finally, only treatment with the expressed in a cohort of breast tumors, where it is associated with a more virulent behavior and poor patient outcome (24). Multiple lines of experimental evidence suggest that overexpression of HER2 confers antiestrogen resistance to breast 4 C. L. Arteaga, unpublished observations.

combination of tamoxifen and the HER2 kinase inhibitor re- HER21neu duced tumor MAPK activity and completely prevented growth of HER2-overexpressing MCF-7/HER2-18 xenografts in athy- JNK ~""/ ~~PI3K-Akt estrogen mic mice in the absence of any host toxicity. In this study, several results implicated MAPK signaling as src MAPK tamoxifen/ the mediator of antiestrogen resistance in the HER2-overexpressing tumor cells: (a) overexpression of HER2 in MCF-7 cells resulted in both activation of MAPK and antiestrogen resistance; (b) inhibition of MAPK with U0126 enhanced the s,!sJ ability of tamoxifen to inhibit both ER-mediated transcription NH2-[ A/B I C IDI ElF I-COOH and colony formation of MCF-7/HER2-18 and BT-474 cells; (c) dominant-negative mutants of MEK1 and MEK2 also enhanced AF-1 DBD AF-2/HBD the inhibitory effect of tamoxifen on ER-mediated transcription; [ 9 transcription and (d) AG1478 markedly reduced active MAPK in MCF-7/ HER2-18 xenografts, and this reduction was temporally associ- Fig. 2 Potential mechanisms of HER2/neu-mediated resistance to tamoxifen. Activation of the ras/MAPK signaling pathway leads to ated with tamoxifen-induced growth restraint of tumors in vivo. ligand-independent phosphorylation of Serl 18 in ERa. This modifica- MAPK has been shown to posttranslationally modify the tion dampens the ability of tamoxifen to recruit the transcriptional ER by phosphorylating Serll8 (17, 18). We were unable to corepressors N-CoR and SMRT to the ER and/or unmasks the agonistic detect ER phosphorylation at Serl 18. However, we have trans- effect of the SERM. This represses the ability of tamoxifen to inhibit hormone-induced gene transcription and/or enhances tamoxifen-in- fected wild-type ER and S 118A ER-luciferase constructs into duced transcription of AF-l-dependent promoters. Other molecules ER-negative, HER2-overexpressing SKBR-3 breast cancer activated by the overexpressed HER2 proto-oncogene product, such as cells. In these cells, tamoxifen failed to inhibit hormone- Src, PI3K, Akt, and JNK, may also be involved in the dysregulation of induced, wild-type ER reporter activity but completely inhibited ER transcription. hormone-induced S118A ER luciferase expression. These re- suits clearly suggest that Serll8 phosphorylation is indeed involved in HER2-mediated repression of tamoxifen action. Neither HER2 overexpression nor inhibition of HER2 and positive breast cancers that overexpress HER2. We should note, MAPK altered ER levels, estradiol-induced ER reporter activity, however, that HER2 is a member of a highly interactive signal- or estradiol-induced proliferation, suggesting that the restoration ing network (23, 24) and that other members of this network can of tamoxifen action was independent of ER levels or their also mediate per se escape from antiestrogen therapy, e.g., intrinsic responsiveness to hormones. Of note, however, tamox- transfection of EGFR into hormone-dependent MCF-7 and ifen-induced association of ERa with N-CoR was markedly ZR75-1 breast cancer cells also induced tamoxifen resistance reduced in HER2-overexpressing cells compared with controls. (44, 45). Furthermore, ER-positive, EGFR-positive tumors ex- Inhibition of HER2 or MAPK restored the antiestrogen-induced hibit a lower response to antiestrogens compared with ER- association of ER with N-CoR. This result implies that anties- positive, EGFR-negative tumors (46). For the other two mem- trogen resistance in HER2-overexpressing tumor cells can, at bers of this receptor family, HER3 and HER4, there are no least in part, be explained by unrestrained MAPK signaling that reports associating them with antiestrogen resistance in breast alters the association of ER with corepressors of transcription cancer. In a recent report, Gee et al. (47) showed that MCF-7 (Fig. 2). The molecular mechanisms by which MAPK regulates cells selected in vitro for resistance to tamoxifen acquired mark- these protein-protein associations require further study. In a edly elevated levels of EGFR and HER2 compared with wild- more recent report, however, another HER2-transfected MCF-7 type antiestrogen-responsive controls. Tamoxifen resistance did cell line (27) exhibited significant down-regulation of ER pro- not occur if the selection was done in the presence of the small tein. Inhibition of MEK or MAPK restored ER expression and, molecule EGFR tyrosine kinase inhibitor ZD1839 ("Iressa"). In presumably, sensitivity to antiestrogens (43). This result sug- addition, low concentrations of ZD1839 inhibited EGFR and gests that HER2 hyperactivity may be able to subvert the an- HER2 phosphorylation and tumor cell growth (47). Taken to- tiestrogen response by additional mechanisms involving ER gether, these data provide a rationale for blocking not only down-regulation. Nonetheless, the data presented imply that: (a) HER2 but also other receptors and/or signal transducers in the MAPK may play a causal role in HER2-induced tamoxifen HER network to maximally abrogate antiestrogen resistance. In resistance in ER-positive breast tumor cells; and (b) that exog- addition, these results suggest that the abrogation of antiestrogen enous inhibitors of HER2 signal transduction can partially ab- resistance by blockade of the HER network may not be limited rogate the antiestrogen-resistant phenotype, providing an evalu- to HER receptor-overexpressing tumor cells. The possibility that able therapeutic strategy in hormone-dependent breast cancer. the clinical response of ER-positive tumors with low levels of EGFR and HER2, such as MCF-7 cells, may also benefit from Clinical Strategies to Inhibit HER (erbB) Receptor blockade of the HER network has important implications for the Signaling in ER-positive Mammary Tumors selection of patients into trials testing this approach (see below). The causal association of HER2 with antiestrogen resist- One anti-EGFR therapeutic strategy is based on the pre- ance, the preclinical data summarized above, and the availability mise that EGFR receptors with mutations in the ATP-binding of HER2 inhibitors provide a strong basis for the testing of site lack tyrosine kinase function and do not display a full range combinations of HER2 inhibitors with antiestrogens in ER- of ligand-induced biochemical responses (23), implying that the

Table I Small molecule inhibitors of the EGFR tyrosine kinase in clinical development EGFR IC5o HER2 IC5o Compound (~M) a (~M) b Ref. Source CI1033 (PD183805) 0.0014 0.009 49, 50 Pfizer ZD 1839 ("Iressa") 0.033 > 3.7 51, 56 AstraZeneca OSI-774 0.02 0.35 52 OSI/Genentech EKB-569 0.039 • 0.007 1.3 + 0.16 53 Wyeth-Ayerst PKI- 166 (CGP59326) 0.025 NRC 54 Novartis GW2016 0.01 0.009 55 Glaxo SmithKline Concentration required to inhibit 50% of the EGFR kinase in vitro. b Concentration required to inhibit 50% of the HER2 kinase in vitro. c NR, not reported.

receptor's tyrosine kinase is essential for transformation. There- lOOO fore, in an attempt to generate antireceptor compounds, small 8oo MDA-361 molecules that compete for the Mg-ATP binding site of the catalytic domain of the EGFR tyrosine kinase have been iden- "~ 600 tified by random screening of natural or synthetic compound libraries for specific EGFR kinase inhibitory activity (48). Table 1 shows a (partial) list of the small-molecular-weight EGFR 2O0 tyrosine kinase inhibitors currently in clinical development. The 0 ~

concentrations of these compounds that inhibit the EGFR kinase 1000 in vitro are in the nM or sub-nM range, whereas, with some MCF-71HER2 exceptions, the concentrations required to inhibit the >80% "~ I ,T, homologous HER2 kinase are logarithms higher (Table 1), 9[ ,oo1 | supporting their overall EGFR specificity. This homology 8 4o0 among HER receptors like EGFR and HER2, however, has been exploited in some cases for the generation of bifunctional in- 2OO hibitors like CI1033, EKB-569, and GW2016. Chemical mod- 0 ification of some of these structures has led to the generation of 4-OH Tam (gMI 0 0 0.1 0.1 1 1 irreversible inhibitors that bind covalently to specific cysteine ZD1839 (~M) 0 1 0 1 0 1 residues in the ATP-binding pocket of either EGFR or HER2, Fig. 3 ZD 1839 enhances tamoxifen-mediated inhibition of tumor col- such as CI1033 and EKB-569. In principle, a bifunctional in- ony survival. MDA-361 or MCF-7/HER2-18 cells were plated as a single-cell suspension in 0.8% agarose/10% fetal calf serum/25 mM hibitor with low IC5o s against both EGFR and HER2 might be Hepes in 35-mm dishes (3 • 104 cells/dish), as described previously more attractive to block both kinases in tumor cells. However, (42). Where indicated, 4-OH tamoxifen and/or ZD1839 were added at recent data with the specific EGFR inhibitor ZD1839 have the time of plating. Controls contained 0.1% ethanol and 0.1% DMSO shown that it can effectively inhibit HER2 phosphorylation in to correct for the respective diluents of 4-OH tamoxifen and ZD1839. vivo at concentrations that would not have been predicted from Dishes were incubated in a humidified CO 2 incubator at 37~ and colonies measuring ->50 ~M in diameter were counted after 7 days using its reported IC5o against the HER2 kinase in vitro shown in an OMNICON 3800 Tumor Colony Analyzer (Biologicals, Gainesville, Table 1 (56). This result suggests that the EGFR may be a FL). Bar, the mean colony number of four dishes • SD. These results transactivating kinase for the ligandless HER2 receptor in tumor were reproduced in each cell line in two independent experiments. cells. It implies further that EGFR-specific inhibitors are a defensible strategy to block the HER2 kinase in tumor cells that also express EGFR. To support this speculation, we have tested the EGFR first-line therapy in patients with metastatic ER- or Progesterone kinase inhibitor ZD1839 in combination with 4-hydroxytamox- receptor-positive breast tumors and any levels of EGFR or ifen against MDA-361 and MCF-7/HER2 breast cancer cells in HER2. To address whether inhibition of EGFR/HER2 reverses a colony-forming assay. These cells express EGFR and high acquired antiestrogen resistance, patients progressing on the levels of HER2 but are only modestly growth inhibited by high SERM or aromatase inhibitor arm can be switched to the com- concentrations (1 ~M) of 4-hydroxytamoxifen. In both tumor bination arm of such study and evaluated for clinical benefit cell lines, the inhibitory effect of tamoxifen on colony formation (Fig. 4). The choice of the small molecule kinase inhibitor may was increased by the addition of 1 iXM ZD1839 (Fig. 3). At this depend on the toxicity profile of the small molecule, its ability concentration, ZD1839 inhibited HER2 phosphorylation with- to inhibit both EGFR (HER1) and HER2 in vivo, and its phar- out altering HER2 levels as measured by phosphotyrosine and macokinetic properties, as well as other issues beyond the scope HER2 immunoblot analyses of HER2 precipitates (data not of this report. Nonetheless, it is not unreasonable to expect that shown). Based on these cumulative data, there is a strong blockade of peptide growth factor receptor and other oncogenic rationale for randomized Phase III trials of an antiestrogen or an signaling pathways will have a profound impact in the natural aromatase inhibitor with or without HER kinase inhibitors as history of hormone-dependent breast cancer.

+ HER kinase References ER+ or PgR+ Antiestrogen inhibitor at 1. McKenna, N., Lanz, R., and O'Malley, B. Nuclear receptor regula- Post-menopausal / or SERM progression tors: cellular and molecular biology. Endocr. Rev., 20: 321-344, 1999. HER1/2 any ~ 2. Kuiper, G., Enmark, K., Pelto-Huikko, M., et al. Cloning of a novel Metastatic disease Antiestrogen or SERM First line therapy + estrogen receptor expressed in rat prostate and ovary. Proc. Natl. Acad. Sci. USA, 93: 5925-5930, 1996. HER kinase inhibitor 3. Kumar, V., Green, S., Stack, G., et al. Functional domains of the Fig. 4 Flow diagram of randomized Phase III trials of antiestrogens human oestrogen receptor. Cell, 51: 941-951, 1987. (aromatase inhibitors or pure antiestrogens) or SERMs +_ HER tyrosine 4. Webster, N. J. G., Green, S., Tasset, D., et al. The transcription kinase inhibitors in patients with hormone-dependent breast cancer. activating function located in the hormone-binding domain of the estrogen receptor is not encoded by a single exon. EMBO J., 8: 1441- 1446, 1989. 5. Tzukerman, M. T., Esty, A., Santisomere, D., et al. Human estrogen receptor transactivational capacity is determined by both cellular and Open Discussion promoter context and mediated by two functionally distinct intramolec- Dr. Kent Osborne: You didn't mention tamoxifen. When ular regions. Mol. Endocrinol., 8: 21-20, 1994. you use tamoxifen with one of these kinase inhibitors, you have 6. Horwitz, K. B., Jackson, T. A., Bain, D. L., et al. Nuclear receptor the same antiestrogen effects as with an aromatase inhibitor, but coactivators and corepressors. Mol. Endocrinol., 10:1167-1177, 1996. you still have the advantages on bone and other parameters. 7. McDonnell, D., Clemm, D., Hermann, T., et al. Analysis of estrogen receptor function in vitro reveals distinct classes of antiestrogens. Mol. Dr. Arteaga: One reason is that fewer and fewer patients Endocrinol., 9: 659-668, 1995. with metastatic disease have tamoxifen as first-line therapy. 8. Jackson, T. A., Richer, J. K., Bain, D. L., et al. The partial agonistic Certainly that has hurt recruitment for a trial we have at Vander- activity of antagonist-occupied steroid receptors is controlled by a novel brit. We recently closed a Vanderbilt Cancer Center Network hinge domain-binding coactivator L7/SPA and the corepressors N/CoR trial of tamoxifen plus Herceptin in ER+ HER2/neu-positive and SMRT. Mol. Endocrinol., 11: 693-705, 1997. metastatic disease. The trial accrued poorly until we changed 9. Lavinsky, R. M., Jepsen, K., Heinzel, T., et al. Diverse signaling pathways modulate nuclear receptor recruitment of N-CoR and SMRT tamoxifen for an aromatase inhibitor. complexes. Proc. Natl. Acad. Sci. USA, 95: 2920-2925, 1998. If the aromatase inhibitors have Dr. Stephen Johnston: 10. Smith, C. L., Nawaz, Z., and O'Malley, B. W. Coactivator and the response rate that Dr. Ellis cites, 88% in ER+, HER1/2+ corepressor regulation of the agonist/antagonist activity of the mixed disease, maybe the control arm is going to be have as high a antiestrogen, 4-hydroxytamoxifen. Mol. Endocrinol., 11: 657-666, 1997. response rate as the experimental arm. 11. Takimoto, G. S., Graham, J. D., Jackson, T. A., et al. Tamoxifen Dr. Arteaga: That is possible, but I doubt it. In the pre- resistant breast cancer: coregulators determine the direction of transcription by antagonist-occupied steroid receptors. J. Steroid Biochem. Mol. clinical models, the response to estrogen deprivation in HER2- Biol., 69: 49-50, 1999. overexpressing tumors is dramatic but very short-lived. 12. Jepsen, K., Hermanson, O., Onami, T. M., et al. Combinatorial roles Dr. Johnston: That's why progression-free response is the of the nuclear receptor corepressor in transcription and development. important end point. Cell, 102: 753-763, 2000. Dr. Osborne: If resistance to tamoxifen is frequently re- 13. Fujimoto, N., and Katzenellenbogen, B. S. Alteration in the agonist/ lated to its intrinsic "agonist" activity, it might be augmented by antagonist balance of antiestrogens by activation of protein kinase A the growth factor pathways after a period of treatment. If you signaling pathways in breast cancer cells: antiestrogen selectivity and promoter dependence. Mol. Endocrinol., 8: 296-304, 1994. could overcome that by concomitant blockade, then tamoxifen 14. Smith, C. L., Conneely, O. M., and O'Malley, B. W. Modulation of plus kinase inhibitor would be a better combination to use in the ligand-independent activation of the human estrogen receptor by hor- adjuvant patients than aromatase inhibitor or antigrowth factor mone and antihormone. Proc. Natl. Acad. Sci. USA, 90: 6120-6124, 1993. inhibitor, because of the bone-sparing effects and the other 15. Kurokawa, H., Lenferink, A. E. G., Simpson, J. F., et al. Inhibition advantages that tamoxifen offers. So tamoxifen may be a worse of HER2/neu (erbB-2) and mitogen-activated protein kinases enhances drug right now because of its intrinsic agonist activity. If you tamoxifen action against HER2-overexpressing, tamoxifen-resistant breast cancer cells. Cancer Res., 60: 5887-5894, 2000. can get rid of that agonist activity with Iressa or another newer 16. Aronica, S. M., and Katzenellenbogen, B. S. Stimulation of estro- agent, then you've got an ideal treatment. gen receptor-mediated transcription and alteration in the phosphoryla- Dr. Matthew Ellis: The experiment is a good one, and in tion state of the rat uterine estrogen receptor by estrogen, cyclic AMP, fact this might be something you could do in the perioperative and insulin-like growth factor-I. Mol. Endocrinol., 7: 743-752, I993. treatment-naive setting, comparing a week or 2 of tamoxifen 17. Kato, S., Endoh, H., Masuhiro, Y., et al. Activation of the estrogen versus tamoxifen plus tyrosine kinase inhibitor to get some data receptor through phosphorylation by mitogen-activated protein kinase. Science, 270: 1491-1494, 1995. on cell growth and apoptosis. But it is tough to accrue patients 18. Bunone, G., Briand, P. A., Miksicek, R. J., et al. Activation of the to tamoxifen trials in the metastatic setting, because many unliganded estrogen receptor by EGF involves the MAP kinase pathway patients are becoming resistant to tamoxifen during their adju- and direct phosphorylation. EMBO J., 15: 2174-2183, 1996. vant treatment or soon after. 19. Campbell, R. A., Bhat-Nakshatri, P., Patel, N. M., et al. Phosphati- Dr. Johnston: You may not need that many patients, dylinositol 3-kinase/Akt-mediated activation of estrogen receptor oL: a though. There really is a major effect on time to progression. new model for anti-estrogen resistance. J. Biol. Chem., 276:9817-9824, 2001. You may not need the power if you have good clinical response 20. Feng W., Webb P., Nguyen P., et al. Potentiation of estrogen numbers. It may be difficult to do, but it seems to me a very receptor activation function 1 (AF-1) by Src/JNK through a serine important question. 118-independent phosphorylation. Mol. Endocrinol., 15: 32-45, 2001.

21. Lee, H., Jiang, F., Wang, Q., et al. MEKK1 activation of human 40. Witters, L. M., Kumar, R., Chinchilli, V. M., et al. Enhanced estrogen receptor cx and stimulation of the agonistic activity of 4- anti-proliferative activity of the combination of tamoxifen plus HER-2- hydroxytamoxifen in endometrial and ovarian cancer cells. Mol. Endo- neu antibody. Breast Cancer Res. Treat., 42: 1-5, 1997. crinol., 14: 1882-1896, 2000. 41. Kunisue, H., Kurebayashi, J., Otsuki, T., et al. Anti-HER2 antibody 22. Hong, S-K., and Privalsky, M. L. The SMRT corepressor is regu- enhances the growth inhibitory effect of anti-oestrogen on breast cancer lated by a MEK-1 kinase pathway: inhibition of corepressor function is cells expressing both oestrogen receptors and HER2. Br. J. Cancer, 82: associated with SMRT phosphorylation and nuclear export. Mol. Cell. 46-51, 2000. Biol., 20: 6612-6625, 2000. 42. Lenferink, A. E. G., Busse, D., Flanagan, W. M., et al. ErbB2/neu 23. Yarden, Y., and Sliwkowski, M. X. Untangling the erbB signalling kinase modulates cellular p27 Kpl and cyclin D1 through multiple sig- network. Nature Rev. Mol. Cell Biol., 2: 127-137, 2001. naling pathways. Cancer Res., in press, 2001. 24. Moniola, A. O., Neve, R. M., Lane, H. A., et al. The erbB signaling 43. Oh, A. S., Lorant, L. A., Holloway, J. N., Miller, D. L., Kern, F. G., network: receptor heterodimerization in development and cancer. and E1-Ashry, D. Hyperactivation of MAPK induces loss of ER cx EMBO J., 19: 3159-3167, 2000. expression in breast cancer ceils. Mol. Endocrinol. 15: 1344-1359, 25. Benz, C. C., Scott, G. K., Sarup, J. C., et al. Estrogen-dependent, 2001. tamoxifen-resistant tumorigenic growth of MCF-7 cells transfected with 44. Miller, D. L., el-Ashry, D., Cheville, A. L., et al. Emergence of HER2/neu. Breast Cancer Res. Treat., 24: 85-95, 1992. MCF-7 cells overexpressing a transfected epidermal growth factor re- 26. Pietras, R. J., Arboleda, J., Reese, D. M., et al. HER-2 tyrosine ceptor (EGFR) under estrogen-depleted conditions: evidence for a role kinase pathway targets estrogen receptor and promotes hormone-inde- of EGFR in breast cancer growth and progression. Cell Growth Differ., pendent growth in human breast cancer cells. Oncogene, 10: 2435- 5: 1263-1274, 1994. 2446, 1995. 45. van Agthoven, T., van Agthoven, T. L., Portengen, H., et al. Ectopic 27. Liu, Y., E1-Ashry, D., Chert, D., et al. MCF-7 breast cancer cells expression of epidermal growth factor receptors induces hormone inde- overexpressing transfected c-erbB-2 have an in vitro growth advantage pendence in ZR-75-1 human breast cancer cells. Cancer Res., 52: in estrogen-depleted conditions and reduced estrogen-dependence and 5082-5088, 1992. tamoxifen-sensitivity in vivo. Breast Cancer Res. Treat., 34:97-117, 1995. 46. Salomon, D. S., Brandt, R., Ciardiello, F., et aL Epidermal growth factor-related peptides and their receptors in human malignancies. Crit. 28. Harris, A. L., Nicholson, S., Sainsbury, J. R. C., et al. Epidermal Rev. Oncol. Hematol., 19: 183-232, 1995. growth factor receptors in breast cancer: association with early relapse and death, poor response to hormones and interactions with neu. J. 47. Gee, J. M., Hutcheson, I. R., Knowlden, J. M., et al. The EGFR- Steroid Biochem., 34: 123-131, 1989. selective tyrosine kinase inhibitor ZD1839 ("Iressa") is an effective inhibitor of tamoxifen-resistant breast cancer growth. Proc. Am. Soc. 29. Wright, C., Nicholson, S., Angus, B., et al. Relationship between c-erbB-2 protein product expression and response to endocrine therapy Clin. Oncol., 20: 282, 2001. in advanced breast cancer. Br. J. Cancer, 65:118-121, 1992. 48. AI-Abeidi, F., and Lam, K. S. Development of inhibitors for protein 30. Borg, A., Baldetorp, B., Fernt, M., et al. ERBB2 amplification is tyrosine kinases. Oncogene, 19: 5690-5701, 2000. associated with tamoxifen resistance in steroid-receptor positive breast 49. Smaill, J. B., Rewcastle, G. W., Loo, J. A., et al. Tyrosine kinase cancer. Cancer Lett., 81: 137-144, 1994. inhibitors. Irreversible inhibitors of the epidermal growth factor 31. Leitzel, K., Teramoto, Y., Konrad, K., et al. Elevated serum c- receptor: 4-(phenylamino)quinazoline and 4-(phenylamino)pyrido(3,2- erbB-2 antigen levels and decreased response to hormone therapy of d)pyrimidine-6-acrylamides bearing additional solubilizing functions. breast cancer. J. Clin. Oncol., 13:1129-1135, 1995. J. Med. Chem., 43: 1380-1397, 2000. 32. Carlomagno, C., Perrone, F., Gallo, C., et al. c-erbB2 overexpres- 50. Sherwood, V., Bridges, A. J., Denny, W. A., et al. Selective sion decreases the benefit of adjuvant tamoxifen in early-stage breast inhibition of heregulin-dependent tyrosine phosphorylation and cellular cancer without axillary lymph node metastases. J. Clin. Oncol., 14: signaling through erbB2, erbB3 and erbB4 by PD158780 and a new 2702-2708, 1996. irreversible inhibitor PD183805. Proc. Am. Assoc. Cancer Res., 40: 33. Yamauchi, H., O'Neill, A., Gelman, R., et al. Prediction of response 723, 1999. to antiestrogen therapy in advanced breast cancer patients by pretreat- 51. Baselga, J., and Averbuch. S. D. ZD 1839 ("Iressa") as an anticancer ment circulating levels of extracellular domain of the HER-2/c-neu agent. Drugs, 60: 33-40, 2000. protein. J. Clin. Oncol., 15: 2518-2525, 1997. 52. Moyer, J. D., Barbacci, E. G., Iwata, K. K., et al. Induction of 34. Berns, E. M., Foekens, J. A., van Staveren, I. L., et al. Oncogene apoptosis and cell cycle arrest by CP-358,774, an inhibitor of epidermal amplification and prognosis in breast cancer: relationship with systemic growth factor receptor tyrosine kinase. Cancer Res., 57: 4838-4848, treatment. Gene, 159: 11-18, 1995. 1997. 35. Archer, S. G., Eliopoulos, A., Spandidos, D., et aL Expression of 53. Torance, C. J., Jackson, P. E., Montgomery E., et al. Combinatorial ras p21, p53 and c-erbB-2 in advanced breast cancer and response to chemoprevention of intestinal neoplasia. Nat. Med., 6: 1024-1028, first line hormonal therapy. Br. J. Cancer, 72: 1259-1266, 1995. 2000. 36. Houston, S. J., Plunkett, T. A., Barnes, D. M., et al. Overexpression 54. Bruns, C. J., Solorzano, C. C., Harbison, M. T., et al. Blockade of of c-erbB2 is an independent marker of resistance to endocrine therapy the epidermal growth factor receptor signaling by a novel tyrosine in advanced breast cancer. Br. J. Cancer, 79: 1220-1226, 1999. kinase inhibitor leads to apoptosis of endothelial cells and therapy of 37. Elledge, R. M., Green, S., Ciocca, D., et al. HER-2 expression and human pancreatic carcinoma. Cancer Res., 60: 2926-2935, 2000. response to tamoxifen in estrogen receptor-positive breast cancer: a 55. Xia, W., Mullin, R., Keith B., et al. GW2016, a dual inhibitor of Southwest Oncology Group Study. Clin. Cancer Res., 4: 7-12, 1998. erbB-2 and EGFR tyrosine kinases: effects on receptor tyrosine auto- 38. Berry, D. A., Muss, H. B., Thor, A. D., et al. HER-2/neu and p53 phosphorylation, downstream signaling intermediaries and in vivo anti- expression versus tamoxifen resistance in estrogen receptor-positive, tumor activity. Proc. Am. Assoc. Cancer. Res., 42: 3635, 2001. node-positive breast cancer. J. Clin. Oncol., 18: 3471-3479, 2000. 56. Moulder, S. L., Yakes, F. M., Bianco, R., et al. Small molecule EGF 39. De Laurentiis, M., Arpino, G., Massarelli, E., et al. A metanalysis receptor tyrosine kinase inhibitor ZD 1839 ("Iressa") inhibits HER2/neu of the interaction between HER2 and the response to endocrine therapy (erbB-2) overexpressing breast tumor cells in vitro and in vivo. Cancer in metastatic breast cancer. Proc. Am. Soc. Clin. Oncol., 19: 300, 2000. Res. (In press), Dec. 15, 2001.

Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2001 American Association for Cancer Research. Inhibition of erbB Receptor (HER) Tyrosine Kinases as a Strategy to Abrogate Antiestrogen Resistance in Human Breast Cancer

Hirokazu Kurokawa and Carlos L. Arteaga

Clin Cancer Res 2001;7:4436s-4442s.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/7/12/4436s

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/7/12/4436s. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.