Wnt/ß-catenin signaling is strongly implicated in neoplasia, but the role of this pathway in human breast cancer has been controversial. Here, we examined Wnt/ß-catenin pathway activation as a function of breast cancer progression, and tested for a relationship with HER2/neu expression, using a human tissue microarray comprising benign breast tissues, ductal carcinoma in situ (DCIS), and invasive carcinomas. Cores were scored for membranous ß-catenin, a key functional component of adherens junctions, and for nucleocytoplasmic ß-catenin, a hallmark of Wnt/ß-catenin pathway activation. Only 82% of benign samples exhibited membrane-associated ß-catenin, indicating a finite frequency of false-negative staining. The frequency of membrane positivity was similar in DCIS samples, but was significantly reduced in carcinomas (45%, P<0.001), consistent with loss of adherens junctions during acquisition of invasiveness. Negative membrane status in cancers correlated with higher grade (P = 0.04) and estrogen receptor-negative status (P = 0.03), both indices of poor prognosis. Unexpectedly, a substantial frequency of nucleocytoplasmic ß-catenin was observed in benign breast tissues (36%), similar to that in carcinomas (35%). Positive-staining basal nuclei observed in benign breast may identify putative stem cells. An increased frequency of nucleocytoplasmic ß-catenin was observed in DCIS tumors (56%), suggesting that pathway activation may be an early event in human breast neoplasia. A correlation was observed between HER2/neu expression and nucleocytoplasmic ß-catenin in node-positive carcinomas (P = 0.02). Furthermore, cytoplasmic ß-catenin was detected in HER2/neu-induced mouse mammary tumors. The Axin2NLSlacZ mouse strain, a previously validated reporter of mammary Wnt/ß-catenin signaling, was utilized to define in vivo transcriptional consequences of HER2/neu-induced ß-catenin accumulation. Discrete hyperplastic foci observed in mammary glands from bigenic MMTV/neu, Axin2NLSlacZ mice, highlighted by robust ß-catenin/TCF signaling, likely represent the earliest stage of mammary intraepithelial neoplasia in MMTV/neu mice. Our study thus provides provocative evidence for Wnt/ß-catenin signaling as an early, HER2/neu-inducible event in breast neoplasia.
Citation: Khalil S, Tan GA, Giri DD, Zhou XK, Howe LR (2012) Activation Status of Wnt/ß-Catenin Signaling in Normal and Neoplastic Breast Tissues: Relationship to HER2/neu Expression in Human and Mouse. PLoS ONE 7(3): e33421. doi:10.1371/journal.pone.0033421
Editor: Mikhail V. Blagosklonny, Roswell Park Cancer Institute, United States of America
Received: December 13, 2011; Accepted: February 14, 2012; Published: March 23, 2012
Copyright: © 2012 Khalil et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by NIH grant R01 CA131219 (L.R.H.) and Department of Defense Breast Cancer Research Program grant W81XWH-04-1-0619 (L.R.H.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The goal of this study was to investigate the activation status of Wnt/ß-catenin signaling in human breast neoplasia, and to test for a potential relationship between Wnt/ß-catenin pathway activation and expression of human epidermal growth factor receptor 2 (HER2/neu). ß-catenin protein exists in two discrete functional pools in epithelial cells. Membrane-associated ß-catenin is an integral component of the adherens junctions linking membrane-localized E-cadherin to the actin cytoskeleton via alpha-catenin. In contrast, ß-catenin protein accumulates in the cytoplasm and nucleus in response to canonical Wnt signaling, the best characterized pathway regulated by Wnt proteins . Thus the presence of nucleocytoplasmic ß-catenin is considered a hallmark of canonical Wnt pathway activation. Stabilized ß-catenin drives transcriptional activation of multiple protumorigenic genes via interaction with TCF/Lef family transcription factors (http://www.stanford.edu/group/nusselab/cgi-bin/wnt/target_genes). The key role of Wnt/ß-catenin signaling in stem cell biology provides another mechanism by which this signaling axis may impact tumorigenesis , .
The role of Wnt/ß-catenin signaling in human breast cancer has been subject to much debate , . The first mammalian Wnt gene, Wnt1, was originally identified as a locus activated by retroviral insertion of mouse mammary tumor virus (MMTV), and transgenic Wnt1 overexpression was subsequently shown to drive mammary tumor formation in mice –. However, historical failure to identify substantial frequencies of Wnt ligand overexpression in human breast tumors hindered appreciation of the relevance of Wnt signaling to the human disease. Renewed interest followed the identification of ß-catenin/TCF complexes as functional mediators of Wnt-induced transcription.
Striking frequencies of aberrant nucleocytoplasmic ß-catenin accumulation have now been recorded in multiple human neoplastic conditions, most notably in colorectal cancers. Direct elucidation of the likely contribution of Wnt ligand overexpression to pathway activation in human cancer specimens has been hampered by a dearth of immunohistochemistry-compatible anti-Wnt antibodies. However, mutation of pathway components, including the APC, Axin, and CTNNB1 genes (encoding ß-catenin), leading to ß-catenin stabilization, and hence activation of the Wnt/ß-catenin pathway, is now recognized as a common event in human tumorigenesis , . Because such mutations are comparatively rare in human breast carcinomas, excepting fibromatoses and metaplastic tumors –, multiple groups have sought evidence of pathway activation by interrogating the subcellular localization of ß-catenin protein in human breast cancers. Conflicting results from these predominantly immunohistochemical studies have again kindled controversy, with some groups reporting that high proportions of breast cancers have nucleocytoplasmic ß-catenin – but other investigators failing to detect substantial frequencies . Further confusion emanates from several studies which elected not to separately assess the functionally distinct membrane and nucleocytoplasmic ß-catenin pools –.
Our goal in the present study was to further investigate the frequency of Wnt/ß-catenin signaling pathway activation in breast neoplasia using ß-catenin immunohistochemistry (IHC) to analyze a human breast tissue microarray (TMA) and to systematically catalogue the prevalence of both membrane-associated and nucleocytoplasmic ß-catenin protein. We sought to characterize the status of Wnt/ß-catenin signaling as a function of breast cancer progression by quantifying and comparing the proportion of tissue samples that exhibited nucleocytoplasmic ß-catenin in benign breast tissue, ductal carcinoma in situ tumors, and invasive carcinomas. An additional goal was to assess the relationship between canonical Wnt pathway activation and HER2 overexpression based on several lines of evidence indicating that Wnt and epidermal growth factor receptor (EGFR) signaling pathways can interact –. Data presented herein suggest that Wnt/ß-catenin pathway activation may be an early event in breast neoplasia, and may be driven, at least in part, by HER2/neu expression.
Materials and Methods
All mice were housed in pathogen-free rooms in filter-topped cages at the Laboratory Animal Research Center at the New York Blood Center. This facility is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care, and operates in accordance with Federal (PHS Policy on the Human Care and Use of Animals, Guide for the Use and Care of Laboratory Animals, Animal Welfare Act), State and local laws and regulations. All mice were used in accordance with protocols approved by the Institutional Animal Care and Use Committees of both the New York Blood Center (Protocol Number 266) and Weill Cornell Medical College (Protocol Numbers 0808-787A, 0055-11). Mice received food and water ad libitum.
To analyze ß-catenin protein immunohistochemically in human breast tissues we used the 2nd Generation Breast Cancer Progression Tissue Microarray developed by the National Cancer Institute (NCI) Cancer Diagnosis Program (CDP). The CDP assembled a collection of 339 breast tissue specimens arrayed on three separate slides (http://cdp.nci.nih.gov/breast/progression_cs2.html). The intended tissue representation was: 69 cores of benign breast tissue (normal or hyperplastic), 31 ductal carcinoma in situ (DCIS) specimens, and 239 invasive breast cancers with a principal histology of ductal carcinoma. Of the benign samples, 43 were from women without breast cancer and 26 were from individuals with breast cancer represented elsewhere on the TMA. Of the DCIS samples, 15 were from women without breast cancer and 16 were from individuals with breast cancer represented elsewhere on the TMA. Of the invasive carcinomas, there were 80 from patients that were node-negative at diagnosis, 80 from patients that were node-positive but not metastatic at diagnosis, and 79 from patients with metastatic disease at diagnosis. The TMA slides purchased from the NCI CDP contained coded human biological specimens lacking personal identifiers such that they could not be linked to specific individuals by the research team. Therefore this study did not constitute human subjects research, and Institutional Review Board approval was not required.
Three freshly-cut serial sections were obtained from each of the three TMA blocks. Two of the three sections from each block were stained in duplicate for ß-catenin protein by IHC as described below. The third intermediate section was stained in parallel omitting primary antibody to provide a negative control. Histopathology of each core on slides stained with ß-catenin antibody was evaluated by a breast pathologist (D.D.G.).
Immunohistochemistry Staining and Scoring
ß-catenin IHC was performed using Clone 14 anti-ß-catenin antibody (BD Transduction Labs) as previously described . Scoring was completed by a specialist breast pathologist (D.D.G.) and a scientist (L.R.H.) blinded to the clinical information; consensus was reached by simultaneous examination using a dual-headed microscope. ß-catenin signal intensity was separately scored in two cellular compartments, membrane and nucleocytoplasmic, which represent distinct functional pools of ß-catenin. The scoring system for each compartment was: 0, no ß-catenin staining; +/−, weak signal; 1+, clear uniform signal; 2+, strong uniform signal; 3+, extremely strong uniform signal. Mean values for each sample were generated from the data for duplicate cores. Membrane and nucleocytoplasmic signals were considered separately when assessing correlations with clinicopathological parameters. For this purpose, cores were assigned as negative (score = 0) or positive (score = +/−, 1+, 2+ or 3+) for each functional pool of ß-catenin.
HER2 IHC was performed on an additional set of TMA slides by the MSKCC Pathology Dept using Ventana's PATHWAY anti-HER2 antibody (clone 4B5), an FDA-approved monoclonal for immunohistochemical detection of HER2 protein in breast cancer tissue. Slides were scored by a breast pathologist (D.D.G.) in accordance with the American Society of Clinical Oncology/College of American Pathologists guideline recommendations for HER2 testing in breast cancer .
Mouse strains, breeding, tissue harvesting and processing
MMTV/NDL mice express a mammary-targeted, mutationally activated HER2/neu allele (NDL, Neu Deletion mutant), under the control of the mouse mammary tumor virus (MMTV) long terminal repeat, that induces mammary hyperplasia and tumorigenesis . Tissue sections were prepared from formalin-fixed, paraffin-embedded MMTV/NDL mammary glands (MGs) generated during a previous study , and were subjected to ß-catenin IHC as described above.
The Axin2NLSlacZ strain (Axin2LacZ) provides a useful reporter of in vivo ß-catenin/TCF signaling. Axin2 is upregulated in response to canonical Wnt/ß-catenin signaling, and functions as a negative feedback regulator , . Axin2NLSlacZ mice have a bacterial ß-galactosidase (ß-gal; LacZ) expression cassette “knocked-in” to the endogenous Axin2 locus, such that ß-gal activity provides a surrogate for Wnt/ß-catenin signaling . We have previously established that this strain functions appropriately as a reporter of in vivo ß-catenin/TCF pathway activation in mouse mammary gland by demonstrating increased ß-gal activity in response to expression of a Wnt1 transgene .
The MMTV/neu strain (FVB/N-Tg(MMTVneu)202Mul/J; The Jackson Laboratory) expresses a wildtype HER2/neu allele. MMTV/neu females develop palpable mammary tumors with a latency of 2–18 months, with subsequent lung metastases . Mammary tumors in this strain bear mutationally activated alleles of the HER2/neu transgene, suggesting that mutational activation of the transgene may be a prerequisite for tumor formation , .
Here, Axin2+/NLSlacZ mice were interbred with homozygous MMTV/neu animals. Abdominal (#4) mammary glands were harvested post-mortem from both MMTV/neu and bigenic MMTV/neu, Axin2+/NLSlacZ female offspring, stained with X-gal to detect ß-gal activity and wholemounted, as previously described . Abdominal MGs from Pea3NLSlacZ females were harvested and stained in parallel to provide a positive control for LacZ staining. Pea3NLSlacZ mice (Pea3LacZ) have a ß-gal expression cassette “knocked-in” to the endogenous Pea3 gene, which is expressed predominantly in the myoepithelial compartment in murine mammary gland , . Genotyping of Axin2LacZ and Pea3LacZ mice was performed as described previously .
In this study, the proportion of samples with positive membrane and nucleocytoplasmic ß-catenin staining in samples of various clinical pathological categories were summarized and compared. Specifically, the associations between ß-catenin staining and the type of tissue sample were examined using mixed-effects logistic regression to take into account possible within-subject correlation because some subjects provided samples of more than one tissue type. Pair-wise comparison of ß-catenin staining positivity between any two tissue types of interest were performed using simultaneous inference methods for general parametric models . P-values were adjusted for multiple comparisons using the conservative Bonferroni method. Fisher's exact test was used to examine the association between ß-catenin staining and clinicopathological parameters within the invasive cancer cases, and to explore the association between membrane and nucleocytoplasmic ß-catenin in samples of the same tissue type.
In this study, we used IHC to assess ß-catenin protein levels in human breast tissues on the 2nd Generation Breast Cancer Progression TMA generated by the NCI Cancer Diagnosis Program (see Materials and Methods for composition of the TMA set). Meaningful data were obtained from 90% of the 339 cores (Table 1). 19 cores did not contain the stated tissue type, and were excluded from subsequent analyses. A further 15 cores were not assessable because of poor sample integrity. Therefore only 305 cores were further analyzed. Notably, although 97% of the invasive breast cancer cores (231/239) yielded data, only 58% of DCIS cores (18/31) were assessable. Clinical characteristics of the assessed invasive breast cancer cases are summarized in Table 2.
We independently scored the signal intensities for two discrete functional pools of ß-catenin, membrane (MB) and nucleocytoplasmic (NC), for each core. Nuclear and cytoplasmic signals were combined into a single entity because accumulation at both these subcellular locations is indicative of canonical Wnt signaling pathway activation. Strikingly in this respect, Kim et al. reported an absolute correlation between nuclear and cytosolic ß-catenin protein levels in all of nine breast cancer cell lines tested using a systematic fractionation approach . In our study, ß-catenin signals were detected predominantly in epithelial cells, with negligible staining of stromal components. Normal colon tissue included on the TMA served as a positive control for ß-catenin staining. As expected, colonocytes exhibited distinct membrane staining, with no cytoplasmic or nuclear staining (Figure 1A). We anticipated that normal breast tissues would similarly exhibit membrane-localized ß-catenin, given that ß-catenin is an important junctional component in epithelial tissues. However, only 82% of benign breast samples (46/56) had detectable membrane staining (e.g. Figure 1B), with a mode value of +/− (Figure 2A), and almost one-fifth of benign breast tissues (n = 10) had essentially no detectable membrane signal despite exhibiting overtly normal epithelial morphology (e.g. Figure 1C).
The 2nd Generation Breast Cancer Progression TMA purchased from the NCI Cancer Diagnosis Program was subjected to ß-catenin IHC using BD Transduction Labs anti-ß-catenin antibody Clone 14 as previously described  and counterstained with hematoxylin. As a control, serial sections were stained in parallel omitting primary antibody (right-hand panels). (A) Normal colon. (B–E) Normal breast. (F) Ductal carcinoma in situ. The image in (E) illustrates the positive nuclear ß-catenin staining observed in myoepithelial cells in some benign breast cores (inset, enlargement of boxed area; red arrowheads indicate positively-staining nuclei).
Each core was assigned a separate score (0, +/−, 1+, 2+, or 3+) for membrane ß-catenin signal (A) and nucleocytoplasmic ß-catenin signal (B), and the percentage of cores with each score was separately calculated for benign breast tissue, DCIS, and invasive carcinomas. Of note, the staining pattern denoted as +/− in this study appeared visually similar to the staining pattern scored as 1 by Khramtzov and colleagues . A statistically significant reduction in membrane signal was observed in invasive cancers relative to benign breast tissues (P<0.001). A numerical increase in the proportion of DCIS tumors with nucleocytoplasmic ß-catenin compared with normal breast tissue was observed.
In DCIS tumors the frequency of positive membrane ß-catenin staining was similar to that in benign breast tissue (89% vs. 82%; P = 0.72) (Figure 2A; Table 3). In contrast, the proportion of invasive breast cancers with membrane positivity was significantly decreased: only 45% of invasive samples (104/231) exhibited detectable membrane ß-catenin (P<0.001 compared with either benign or DCIS). Additionally, when cases were sub-divided into node-negative, node-positive or metastatic at diagnosis, a trend of lower odds of having positive membrane staining in higher stage cancer was observed (Table 3). Examples of invasive cancer staining patterns are shown in Figure S1.
Nucleocytoplasmic staining was not detected in the majority of benign breast samples (Figure 2B). However, approximately one quarter had a +/− nucleocytoplasmic score (15/56), and five samples had stronger staining (e.g. Figure 1D). In total, nucleocytoplasmic staining was detected in 36% (20/56) of benign samples (Table 3; Figure 2B). These findings contrasted with our expectation that benign breast tissue would exhibit little or no nucleocytoplasmic ß-catenin. Similar frequencies of nucleocytoplasmic ß-catenin were observed in invasive carcinomas (82/231; 35%; P = 1). Interestingly however, the proportion of DCIS tumors positive for nucleocytoplasmic ß-catenin was quantitatively increased relative to benign tissue (10/18; 56%; Figure 2B; Table 3) (e.g. Figure 1F).
A striking positive correlation was identified between membrane and nucleocytoplasmic ß-catenin positivity in individual samples (Table 4). For carcinoma cores, having positive ß-catenin membrane staining increased the odds of having positive nucleocytoplasmic signal in the same core by 15-fold (95% CI = (7.3, 32.6), P<0.001, Fisher's exact test). The two parameters were not associated in DCIS cores, but there was a correlation between membrane and nucleocytoplasmic ß-catenin positivity in benign breast tissue samples (P = 0.01, Fisher's exact test). In order to control for effects of this correlation, we conducted an additional analysis restricted to those samples with positive membrane staining (n = 166). In this subset, the proportion of samples with positive nucleocytoplasmic staining was increased for both DCIS (56%, 9/16) and invasive cancer samples (65%, 68/104; P = 0.03) relative to benign breast tissue (43%, 20/46).
Associations between ß-catenin signal and clinicopathological parameters were assessed for the invasive cancer cases (Table 5). Decreased membrane ß-catenin positivity was associated with higher grade (P = 0.04) and estrogen receptor (ER) negative status (P = 0.03), both indices of poor prognosis. No statistically significant associations were observed between nucleocytoplasmic ß-catenin positivity and individual clinicopathological parameters when all invasive tumors were considered together (Table 5). However, stratification into node-negative, node-positive and metastatic revealed an association between nucleocytoplasmic ß-catenin and HER2/neu positivity in node-positive cases. Nucleocytoplasmic ß-catenin was detected in 67% (8/12) of HER2/neu-expressing node-positive breast carcinomas, but only in 30% (20/66) of HER2/neu-negative node-positive cases (P = 0.02). When all invasive cases were considered together, the data were also suggestive of a potential interaction between HER2/neu and ß-catenin stabilization: nucleocytoplasmic ß-catenin was detected in 43.8% of HER2/neu-expressing breast carcinomas, but only in 34.2% of HER2/neu-negative cases (Table 5; not significant). In DCIS tumors, 100% of HER2/neu-positive cases (3/3) exhibited nucleocytoplasmic ß-catenin, whereas only 50% of HER2/neu-negative samples (7/14) were nucleocytoplasmic ß-catenin-positive, but statistical power was limited by small sample size.
To further investigate the potential relationship between HER2/neu and canonical Wnt signaling in vivo in mammary tissues, we assessed ß-catenin localization and function in mammary glands (MGs) from HER2/neu transgenic mice. Firstly, ß-catenin IHC was performed on MG sections from MMTV/NDL mice, which develop multiple DCIS-like lesions in each MG due to expression of a mammary-targeted, mutationally activated HER2/neu allele , . Diffuse cytoplasmic ß-catenin staining was observed in DCIS lesions in MMTV/NDL MGs (Figure 3), reminiscent of the staining pattern observed in human DCIS tumors (Figure 1F). This staining pattern contrasted with the predominantly membranous signal that we observed in mammary epithelium from wildtype mice .
Mammary gland tissue sections from virgin female MMTV/NDL mice were subjected to ß-catenin IHC as previously described  and counterstained with hematoxylin. As a control, serial sections were stained in parallel omitting primary antibody (right-hand panel). Diffuse cytoplasmic ß-catenin was detected in DCIS-like lesions in MMTV/NDL MGs.
To test the in vivo functional significance of HER2/neu-induced ß-catenin redistribution, we employed a ß-catenin/TCF reporter strain, Axin2NLSlacZ, which has a ß-gal expression cassette “knocked-in” to the Axin2 locus . We previously established the utility of this strain for detecting ß-catenin/TCF-dependent transcription in mouse mammary gland by demonstrating dramatically increased ß-gal reporter activity in Axin2NLSlacZ MGs in response to expression of Wnt1, which activates canonical Wnt signaling . In the present study, we analyzed MGs from bigenic MMTV/neu, Axin2+/NLSlacZ mice, to test the ability of HER2/neu to drive ß-catenin/TCF-dependent transcription in vivo. Discrete regions of robust ß-gal activity were detected in the mammary epithelium of bigenic MMTV/neu, Axin2+/NLSlacZ MGs which coincided with focal regions of hyperplasia (Figure 4A–G). Positively staining regions were of similar size in all cases, with clearly demarcated boundaries (e.g. Figure 4A–D, F, G; marked by red arrowheads). The cellular staining profile contrasted markedly with that detected in Pea3+/NLSLacZ mice (Figure 4H), in which ß-gal expression is predominantly restricted to the myoepithelial layer , . Comparison of these staining patterns (e.g. Figure 4G vs 4H) suggests that the Axin2LacZ-positive nuclei in bigenic MMTV/neu, Axin2+/NLSlacZ MGs correspond to luminal epithelial cells, the compartment in which the neu transgene is expressed.
Abdominal (#4) mammary glands were harvested post-mortem from virgin female mice that were MMTV/neu or bigenic MMTV/neu, Axin2+/NLSlacZ. MGs were stained with X-gal and wholemounted as previously described . Pea3+/NLSlacZ (Pea3LacZ) samples were processed in parallel for comparison. (A–D) Wholemounted glands from bigenic MMTV/neu, Axin2NLSlacZ mice (38–44 weeks old), viewed at 4× magnification. (E,F) Higher power images of lesion seen in Panel A, viewed at 20× magnification. (G) Higher power image of lesion seen in Panel C, viewed at 20× magnification. (H) Wholemounted gland from Pea3+/NLSlacZ virgin female (39 weeks old) stained in parallel with the specimen in Panels C/G, viewed at 20× magnification. Eight focal lesions displaying intense ß-gal activity, all of similar size, were identified in MGs from 26 bigenic MMTV/neu, Axin2+/NLSlacZ mice. By contrast, no comparable lesions were observed in MGs from MMTV/neu mice lacking the Axin2LacZ allele (not shown). Additionally, in our previous study of Axin2LacZ mice, no such lesions were observed in tissues from Axin2LacZ animals lacking a tumor-promoting transgene . Strikingly, each discrete region of robust ß-gal activity in bigenic MMTV/neu, Axin2NLSlacZ MGs coincided with a focus of hyperplastic morphology markedly dissimilar to the normal-looking morphology of the immediately adjacent ductal structures. Also notable were the clearly demarcated boundaries between positive and negatively stained epithelium (examples are marked with red arrowheads in Panels A, C, F, G). The cellular staining profile in bigenic MMTV/neu, Axin2NLSlacZ MGs contrasted with the myoepithelial pattern characteristic of Pea3LacZ MGs (compare Panels G and H), suggesting that Axin2LacZ is expressed in luminal cells within the HER2/neu-induced hyperplastic lesions.
In this study we analyzed ß-catenin protein levels in membrane and nucleocytoplasmic compartments in human breast tissues (normal breast, DCIS and invasive cancers) with the goal of better understanding the activation status of Wnt/ß-catenin signaling in human breast cancer, given conflicting findings from previous studies , –. Membrane ß-catenin, a key component of adherens junctions, was undetectable in almost one fifth of normal breast tissue cores despite overtly normal epithelial morphology (Figure 2A), suggesting a substantial inherent frequency of false negatives which may reflect loss of ß-catenin antigenicity during tissue processing or storage. Of note, we observed a marked correlation between positivity for membrane and nucleocytoplasmic ß-catenin in individual cores of normal breast tissue (Table 4), again consistent with reduced ß-catenin antigenicity in some samples. This systematic analysis of normal breast tissues stained simultaneously on a TMA highlights the potential for underestimating ß-catenin signal in IHC-based analyses. Nevertheless, we observed a significantly higher frequency of samples lacking discernible membrane ß-catenin among the invasive cores relative to the benign breast samples (55% versus 18%; P<0.001 versus benign), consistent with previous reports of reduced membranous ß-catenin in human breast cancers , , , , , . Loss of adherens junctions has been suggested to contribute to the invasive phenotype , . Consistent with this model, there were significant correlations between reduced frequencies of membrane ß-catenin in carcinoma samples and two indices of poor prognosis, higher grade (P = 0.04) and ER-negative status (P = 0.03).
In light of the fact that nucleocytoplasmic ß-catenin is indicative of activated Wnt/ß-catenin signaling, our analysis of nucleocytoplasmic ß-catenin signal yielded somewhat unexpected data. Specifically, we observed a finite frequency of positive nucleocytoplasmic staining in normal human breast tissues (36%), with 5 of the 56 assessable cores exhibiting substantial signal (Figs. 1D & 2B). These findings contrast with a recent study which did not detect significant nucleocytoplasmic ß-catenin in normal breast . We hypothesized that Wnt/ß-catenin pathway activation observed in benign breast tissue could result from precancerous changes in morphologically normal breast, and further reasoned that breast tissue adjacent to invasive tumors would be more likely to contain such protumorigenic molecular alterations. However, the frequency of nucleocytoplasmic ß-catenin positivity was similar in benign tissue irrespective of the presence or absence of adjacent invasive disease (with invasive disease, 31.6% NC-positive; without invasive disease, 37.8% NC-positive, P = 0.77). Thus the basis for ß-catenin stabilization in these samples remains unclear. Notably, the vast majority of previous IHC studies of ß-catenin in breast cancer focused exclusively on malignant specimens.
Intriguingly, some benign breast cores exhibited discrete staining of nuclei in the basal layer of normal-looking epithelium (e.g. Figure 1E, inset, arrowheads). This could reflect pathway activation in these cells by endogenous Wnt ligands, since several Wnt genes are expressed in human and mouse breast tissues . Notably, activation of canonical signaling in response to transgenically expressed Wnt1 ligand in vivo in mouse MG is predominantly restricted to the myoepithelial compartment , , reflecting the myoepithelial expression profile of Wnt co-receptors low density lipoprotein receptor-related protein (LRP) 5 and 6 in post-natal MG , . The basally restricted expression pattern of LRP5/6 may also provide a partial explanation for the observation that nucleocytoplasmic ß-catenin accumulation is selectively enriched in basal-like breast cancers , .
We speculate that cells positive for nucleocytoplasmic ß-catenin in the myoepithelial layer may include mammary stem cells, based both on location and on substantial data implicating canonical Wnt signaling in stem cell maintenance , , . Intriguingly, Zeng and Nusse recently reported evidence for active Wnt/ß-catenin signaling in a subset of basally-located epithelial cells in murine MG which may overlap with the mammary stem cell compartment . Further evidence for a relationship between “stemness” and Wnt signaling is provided by the recently reported association between nuclear/cytoplasmic ß-catenin and the putative stem cell surface marker phenotype CD44+/CD24− in human breast carcinomas .
In our study, DCIS tumors exhibited the highest frequency of nucleocytoplasmic ß-catenin. Our analysis of Wnt signaling in DCIS was limited by the number of samples that yielded meaningful data (n = 18; Table 1). Nevertheless, we saw a substantially increased frequency of nucleocytoplasmic ß-catenin positivity in DCIS cores relative to that in benign breast samples (56% v. 36%) (Figure 2B, Table 3), suggesting the possibility that Wnt/ß-catenin signaling is activated early during breast neoplasia and could be a target for prophylactic intervention as in colorectal cancer .
In invasive carcinomas, we detected nucleocytoplasmic ß-catenin in 35% of samples (Figure 2B), which is within the range detected in other studies , –, . Notably, and consistent with observations from other investigators , , , , we did not see profound levels of nuclear ß-catenin comparable to those reported for colorectal cancers (Figure S1). The intense nuclear signal commonly observed in colorectal tumors may reflect a specific response to APC mutation, the causal event in the majority of human colorectal cancers. In contrast, commensurate increases in nuclear and cytoplasmic pools are frequently elicited by other stimuli of the Wnt/ß-catenin pathway, including Wnt1 transgene expression in mouse MG . Intriguingly, marked cytosolic accumulation of ß-catenin was observed in mitotic cells in some carcinoma samples (Figure S1G, inset), consistent with published data implicating ß-catenin in mitosis –.
No correlations were detected between pathway activation and any clinicopathological parameter, when invasive cancers were analyzed as a single population. However, a positive correlation between HER2/neu expression and nucleocytoplasmic ß-catenin was observed in node-positive cases (P = 0.02). It is unclear whether this reflects an intrinsic biologic phenomenon specific to the node-positive subset. Nevertheless, the observed association is of particular interest because EGFR family members have the capacity to modulate ß-catenin phosphorylation, localization and transcriptional activity –. In apparent contradiction, other investigators have reported an inverse correlation between HER2/neu expression and nuclear/cytosolic ß-catenin , . Conversely, Lopez-Knowles and colleagues identified a relationship between increased cytoplasmic ß-catenin and HER2/neu positivity .
To further investigate the relationship between HER2/neu expression and canonical Wnt signaling, we analyzed mammary tissues from HER2/neu transgenic mice. Consistent with our observation of an increased frequency of nucleocytoplasmic ß-catenin in human DCIS relative to benign breast epithelium (Figure 2B; Table 3), we detected nucleocytoplasmic ß-catenin in murine mammary DCIS tumors induced by overexpression of a mutationally activated HER2/neu transgene (NDL; Figure 3). Functional significance of nucleocytoplasmic ß-catenin in HER2/neu-overexpressing breast neoplasias was suggested by data obtained using the Axin2LacZ reporter strain, a validated tool for visualizing ß-catenin/TCF signaling in vivo in murine breast tissue. Discrete regions of intense ß-gal reporter activity were observed to colocalize with foci of hyperplastic epithelium in MGs from bigenic MMTV/neu, Axin2LacZ mice (Figure 4A–G). Notably, all lesions were approximately similar in size and had clearly delineated boundaries, leading us to speculate that these hyperplastic foci with robust ß-catenin/TCF signaling could represent clonal populations with distinct biology from the immediately adjacent ductal epithelium. Importantly in this respect, formation of mammary tumors in the MMTV/neu strain is associated with mutational activation of the (initially) wildtype neu transgene, which may be an obligate prerequisite to tumor formation , . Thus, we speculate that the hyperplastic foci that we observed may be clonal groups of cells bearing a mutated neu allele, and potentially the earliest manifestation of mammary intraepithelial neoplasia (MIN) in this strain. Irrespective of the mutational status of the neu allele, these observations provide intriguing evidence of ß-catenin/TCF transcriptional activity in HER2/neu-expressing MIN lesions, and suggest that HER2/neu-induced activation of canonical Wnt signaling may be an early event in breast neoplasia. Consistent with this model, the elevated frequency at which nucleocytoplasmic ß-catenin was detected in human DCIS specimens compared with invasive carcinomas in our study could reflect the relatively high frequency of HER2/neu overexpression in human DCIS lesions compared with frank cancers –.
In summary, in this study we effected a systematic comparison of membrane and nucleocytoplasmic ß-catenin during breast cancer progression from benign breast to metastatic carcinoma. A limitation of the study was that outcome data were not available for the breast cancer cases, and thus prognostic significance of membrane and nucleocytoplasmic ß-catenin positivity could not be determined. However, data addressing prognostic significance has been provided by other studies , , , . Nevertheless, we made several notable findings, including identification of a substantial frequency of canonical Wnt signaling pathway activation in benign human breast tissues, an apparent increase in pathway activation in DCIS tumors, and a correlation between HER2/neu expression and nucleocytoplasmic ß-catenin in node-positive breast carcinomas. A causal relationship between HER2/neu and Wnt/ß-catenin signaling is suggested by our companion mouse studies. We observed ß-catenin accumulation in HER2/neu-induced mouse mammary tumors, and identified discrete regions of ß-catenin/TCF transcriptional activity coincident with early MIN lesions in HER2/neu transgenic breast. Our study thus provides provocative evidence for Wnt/ß-catenin signaling as an early, HER2/neu-inducible event in breast neoplasia.
ß-catenin staining patterns of human breast carcinomas. Images of seven invasive cancer cores demonstrating the range of observed staining patterns. Samples were scored as follows: (A) Membrane (MB), 0; Nucleocytoplasmic (NC), 0. (B) MB, +/−; NC, 0. (C) MB, 1+; NC, 0. (D) MB, 1+; NC, 1+. (E) MB, 2+; NC, 2+. (F) MB, 0; NC, 2+. The image in (G) illustrates the positive cytoplasmic ß-catenin staining observed in mitotic cells in some cores (inset, enlargement of boxed area). Left-hand panels; slides treated with primary (anti-ß-catenin) and secondary antibodies; right-hand panels, negative control slides treated with secondary antibody alone.
Conceived and designed the experiments: LRH. Performed the experiments: SK GAT DDG LRH. Analyzed the data: XKZ LRH. Wrote the paper: LRH. Developed the scoring system: DDG LRH.
- 1. Cadigan KM, Peifer M (2009) Wnt signaling from development to disease: insights from model systems. Cold Spring Harb Perspect Biol 1: a002881.
- 2. Nusse R (2008) Wnt signaling and stem cell control. Cell Res 18: 523–527.
- 3. Reya T, Clevers H (2005) Wnt signalling in stem cells and cancer. Nature 434: 843–850.
- 4. Howe LR, Brown AMC (2004) Wnt signaling and breast cancer. Cancer Biol Ther 3: 36–41.
- 5. Smalley MJ, Dale TC (2001) Wnt signaling and mammary tumorigenesis. J Mammary Gland Biol Neoplasia 6: 37–52.
- 6. Nusse R, van Ooyen A, Cox D, Fung YK, Varmus H (1984) Mode of proviral activation of a putative mammary oncogene (int-1) on mouse chromosome 15. Nature 307: 131–136.
- 7. Nusse R, Varmus HE (1982) Many tumors induced by the mouse mammary tumor virus contain a provirus integrated in the same region of the host genome. Cell 31: 99–109.
- 8. Peters G, Brookes S, Smith R, Dickson C (1983) Tumorigenesis by mouse mammary tumor virus: evidence for a common region for provirus integration in mammary tumors. Cell 33: 369–377.
- 9. Tsukamoto AS, Grosschedl R, Guzman RC, Parslow T, Varmus HE (1988) Expression of the int-1 gene in transgenic mice is associated with mammary gland hyperplasia and adenocarcinomas in male and female mice. Cell 55: 619–625.
- 10. Giles RH, van Es JH, Clevers H (2003) Caught up in a Wnt storm: Wnt signaling in cancer. Biochim Biophys Acta 1653: 1–24.
- 11. Polakis P (2000) Wnt signaling and cancer. Genes Dev 14: 1837–1851.
- 12. Candidus S, Bischoff P, Becker KF, Hofler H (1996) No evidence for mutations in the alpha- and beta-catenin genes in human gastric and breast carcinomas. Cancer Res 56: 49–52.
- 13. Furuuchi K, Tada M, Yamada H, Kataoka A, Furuuchi N, et al. (2000) Somatic mutations of the APC gene in primary breast cancers. Am J Pathol 156: 1997–2005.
- 14. Jonsson M, Borg A, Nilbert M, Andersson T (2000) Involvement of adenomatous polyposis coli (APC)/beta-catenin signalling in human breast cancer. Eur J Cancer 36: 242–248.
- 15. Kashiwaba M, Tamura G, Ishida M (1994) Aberrations of the APC gene in primary breast carcinoma. J Cancer Res Clin Oncol 120: 727–731.
- 16. Schlosshauer PW, Brown SA, Eisinger K, Yan Q, Guglielminetti ER, et al. (2000) APC truncation and increased beta-catenin levels in a human breast cancer cell line. Carcinogenesis 21: 1453–1456.
- 17. Sorlie T, Bukholm I, Borresen-Dale AL (1998) Truncating somatic mutation in exon 15 of the APC gene is a rare event in human breast carcinomas. Mutations in brief no. 179. Online. Hum Mutat 12: 215.
- 18. Webster MT, Rozycka M, Sara E, Davis E, Smalley M, et al. (2000) Sequence variants of the axin gene in breast, colon, and other cancers: an analysis of mutations that interfere with GSK3 binding. Genes Chromosomes Cancer 28: 443–453.
- 19. Geyer FC, Lacroix-Triki M, Savage K, Arnedos M, Lambros MB, et al. (2011) beta-Catenin pathway activation in breast cancer is associated with triple-negative phenotype but not with CTNNB1 mutation. Mod Pathol 24: 209–231.
- 20. Abraham SC, Reynolds C, Lee JH, Montgomery EA, Baisden BL, et al. (2002) Fibromatosis of the breast and mutations involving the APC/beta-catenin pathway. Hum Pathol 33: 39–46.
- 21. Hayes MJ, Thomas D, Emmons A, Giordano TJ, Kleer CG (2008) Genetic changes of Wnt pathway genes are common events in metaplastic carcinomas of the breast. Clin Cancer Res 14: 4038–4044.
- 22. Gonzalez MA, Pinder SE, Wencyk PM, Bell JA, Elston CW, et al. (1999) An immunohistochemical examination of the expression of E-cadherin, alpha- and beta/gamma-catenins, and alpha2- and beta1-integrins in invasive breast cancer. J Pathol 187: 523–529.
- 23. Lin SY, Xia W, Wang JC, Kwong KY, Spohn B, et al. (2000) Beta-catenin, a novel prognostic marker for breast cancer: its roles in cyclin D1 expression and cancer progression. Proc Natl Acad Sci U S A 97: 4262–4266.
- 24. Ryo A, Nakamura M, Wulf G, Liou YC, Lu KP (2001) Pin1 regulates turnover and subcellular localization of beta-catenin by inhibiting its interaction with APC. Nat Cell Biol 3: 793–801.
- 25. Bukholm IR, Bukholm G, Nesland JM (2001) Coexpression of cyclin A and beta-catenin and survival in breast cancer patients. Int J Cancer 94: 148–149.
- 26. Prasad CP, Gupta SD, Rath G, Ralhan R (2007) Wnt signaling pathway in invasive ductal carcinoma of the breast: relationship between beta-catenin, dishevelled and cyclin D1 expression. Oncology 73: 112–117.
- 27. Khramtsov AI, Khramtsova GF, Tretiakova M, Huo D, Olopade OI, et al. (2010) Wnt/beta-catenin pathway activation is enriched in basal-like breast cancers and predicts poor outcome. Am J Pathol 176: 2911–2920.
- 28. Lopez-Knowles E, Zardawi SJ, McNeil CM, Millar EK, Crea P, et al. (2010) Cytoplasmic localization of beta-catenin is a marker of poor outcome in breast cancer patients. Cancer Epidemiol Biomarkers Prev 19: 301–309.
- 29. Karayiannakis AJ, Nakopoulou L, Gakiopoulou H, Keramopoulos A, Davaris PS, et al. (2001) Expression patterns of beta-catenin in in situ and invasive breast cancer. Eur J Surg Oncol 27: 31–36.
- 30. Chung GG, Zerkowski MP, Ocal IT, Dolled-Filhart M, Kang JY, et al. (2004) beta-Catenin and p53 analyses of a breast carcinoma tissue microarray. Cancer 100: 2084–2092.
- 31. Logullo AF, Nonogaki S, Pasini FS, Osorio CA, Soares FA, et al. (2010) Concomitant expression of epithelial-mesenchymal transition biomarkers in breast ductal carcinoma: association with progression. Oncol Rep 23: 313–320.
- 32. Bertolo C, Guerrero D, Vicente F, Cordoba A, Esteller M, et al. (2008) Differences and molecular immunohistochemical parameters in the subtypes of infiltrating ductal breast cancer. Am J Clin Pathol 130: 414–424.
- 33. Hoschuetzky H, Aberle H, Kemler R (1994) Beta-catenin mediates the interaction of the cadherin-catenin complex with epidermal growth factor receptor. J Cell Biol 127: 1375–1380.
- 34. Hazan RB, Norton L (1998) The epidermal growth factor receptor modulates the interaction of E-cadherin with the actin cytoskeleton. J Biol Chem 273: 9078–9084.
- 35. Bonvini P, An WG, Rosolen A, Nguyen P, Trepel J, et al. (2001) Geldanamycin abrogates ErbB2 association with proteasome-resistant beta-catenin in melanoma cells, increases beta-catenin-E-cadherin association, and decreases beta-catenin-sensitive transcription. Cancer Res 61: 1671–1677.
- 36. Adam L, Vadlamudi RK, McCrea P, Kumar R (2001) Tiam1 overexpression potentiates heregulin-induced lymphoid enhancer factor-1/beta -catenin nuclear signaling in breast cancer cells by modulating the intercellular stability. J Biol Chem 276: 28443–28450.
- 37. Schroeder JA, Adriance MC, McConnell EJ, Thompson MC, Pockaj B, et al. (2002) ErbB-beta-catenin complexes are associated with human infiltrating ductal breast and murine mammary tumor virus (MMTV)-Wnt-1 and MMTV-c-Neu transgenic carcinomas. J Biol Chem 277: 22692–22698.
- 38. Lu Z, Ghosh S, Wang Z, Hunter T (2003) Downregulation of caveolin-1 function by EGF leads to the loss of E-cadherin, increased transcriptional activity of beta-catenin, and enhanced tumor cell invasion. Cancer Cell 4: 499–515.
- 39. Graham NA, Asthagiri AR (2004) Epidermal growth factor-mediated T-cell factor/lymphoid enhancer factor transcriptional activity is essential but not sufficient for cell cycle progression in nontransformed mammary epithelial cells. J Biol Chem 279: 23517–23524.
- 40. Wang K, Ma Q, Ren Y, He J, Zhang Y, et al. (2007) Geldanamycin destabilizes HER2 tyrosine kinase and suppresses Wnt/beta-catenin signaling in HER2 overexpressing human breast cancer cells. Oncol Rep 17: 89–96.
- 41. Civenni G, Holbro T, Hynes NE (2003) Wnt1 and Wnt5a induce cyclin D1 expression through ErbB1 transactivation in HC11 mammary epithelial cells. EMBO Rep 4: 166–171.
- 42. Schlange T, Matsuda Y, Lienhard S, Huber A, Hynes NE (2007) Autocrine WNT signaling contributes to breast cancer cell proliferation via the canonical WNT pathway and EGFR transactivation. Breast Cancer Res 9: R63.
- 43. Baker R, Kent CV, Silbermann RA, Hassell JA, Young LJ, et al. (2010) Pea3 transcription factors and Wnt1-induced mouse mammary neoplasia. PLoS One 5: e8854.
- 44. Wolff AC, Hammond ME, Schwartz JN, Hagerty KL, Allred DC, et al. (2007) American Society of Clinical Oncology/College of American Pathologists guideline recommendations for human epidermal growth factor receptor 2 testing in breast cancer. J Clin Oncol 25: 118–145.
- 45. Siegel PM, Ryan ED, Cardiff RD, Muller WJ (1999) Elevated expression of activated forms of Neu/ErbB-2 and ErbB-3 are involved in the induction of mammary tumors in transgenic mice: implications for human breast cancer. Embo J 18: 2149–2164.
- 46. Howe LR, Chang SH, Tolle KC, Dillon R, Young LJ, et al. (2005) HER2/neu-induced mammary tumorigenesis and angiogenesis are reduced in cyclooxygenase-2 knockout mice. Cancer Res 65: 10113–10119.
- 47. Jho EH, Zhang T, Domon C, Joo CK, Freund JN, et al. (2002) Wnt/beta-catenin/Tcf signaling induces the transcription of Axin2, a negative regulator of the signaling pathway. Mol Cell Biol 22: 1172–1183.
- 48. Lustig B, Jerchow B, Sachs M, Weiler S, Pietsch T, et al. (2002) Negative feedback loop of Wnt signaling through upregulation of conductin/axin2 in colorectal and liver tumors. Mol Cell Biol 22: 1184–1193.
- 49. Guy CT, Webster MA, Schaller M, Parsons TJ, Cardiff RD, et al. (1992) Expression of the neu protooncogene in the mammary epithelium of transgenic mice induces metastatic disease. Proc Natl Acad Sci U S A 89: 10578–10582.
- 50. Siegel PM, Dankort DL, Hardy WR, Muller WJ (1994) Novel activating mutations in the neu proto-oncogene involved in induction of mammary tumors. Mol Cell Biol 14: 7068–7077.
- 51. Kurpios NA, MacNeil L, Shepherd TG, Gludish DW, Giacomelli AO, et al. (2009) The Pea3 Ets transcription factor regulates differentiation of multipotent progenitor cells during mammary gland development. Dev Biol 325: 106–121.
- 52. Hothorn T, Bretz F, Westfall P (2008) Simultaneous inference in general parametric models. Biom J 50: 346–363.
- 53. Kim SY, Dunn IF, Firestein R, Gupta P, Wardwell L, et al. (2010) CK1epsilon is required for breast cancers dependent on beta-catenin activity. PLoS One 5: e8979.
- 54. Dolled-Filhart M, McCabe A, Giltnane J, Cregger M, Camp RL, et al. (2006) Quantitative in situ analysis of beta-catenin expression in breast cancer shows decreased expression is associated with poor outcome. Cancer Res 66: 5487–5494.
- 55. Zschiesche W, Schonborn I, Behrens J, Herrenknecht K, Hartveit F, et al. (1997) Expression of E-cadherin and catenins in invasive mammary carcinomas. Anticancer Res 17: 561–567.
- 56. Berx G, Van Roy F (2001) The E-cadherin/catenin complex: an important gatekeeper in breast cancer tumorigenesis and malignant progression. Breast Cancer Res 3: 289–293.
- 57. Birchmeier W, Behrens J (1994) Cadherin expression in carcinomas: role in the formation of cell junctions and the prevention of invasiveness. Biochim Biophys Acta 1198: 11–26.
- 58. Prosperi JR, Goss KH (2010) A Wnt-ow of opportunity: targeting the Wnt/beta-catenin pathway in breast cancer. Curr Drug Targets 11: 1074–1088.
- 59. Teissedre B, Pinderhughes A, Incassati A, Hatsell SJ, Hiremath M, et al. (2009) MMTV-Wnt1 and -DeltaN89beta-catenin induce canonical signaling in distinct progenitors and differentially activate Hedgehog signaling within mammary tumors. PLoS ONE 4: e4537.
- 60. Badders NM, Goel S, Clark RJ, Klos KS, Kim S, et al. (2009) The Wnt receptor, Lrp5, is expressed by mouse mammary stem cells and is required to maintain the basal lineage. PLoS ONE 4: e6594.
- 61. Lindvall C, Zylstra CR, Evans N, West RA, Dykema K, et al. (2009) The Wnt co-receptor Lrp6 is required for normal mouse mammary gland development. PLoS ONE 4: e5813.
- 62. Lindvall C, Bu W, Williams BO, Li Y (2007) Wnt signaling, stem cells, and the cellular origin of breast cancer. Stem Cell Rev 3: 157–168.
- 63. Zeng YA, Nusse R (2011) Wnt proteins are self-renewal factors for mammary stem cells and promote their long-term expansion in culture. Cell Stem Cell 6: 568–577.
- 64. Clapper ML, Coudry J, Chang WC (2004) beta-catenin-mediated signaling: a molecular target for early chemopreventive intervention. Mutat Res 555: 97–105.
- 65. Sormunen RT, Leong AS, Vaaraniemi JP, Fernando SS, Eskelinen SM (1999) Immunolocalization of the fodrin, E-cadherin, and beta-catenin adhesion complex in infiltrating ductal carcinoma of the breast-comparison with an in vitro model. J Pathol 187: 416–423.
- 66. Olmeda D, Castel S, Vilaro S, Cano A (2003) Beta-catenin regulation during the cell cycle: implications in G2/M and apoptosis. Mol Biol Cell 14: 2844–2860.
- 67. McCartney BM, McEwen DG, Grevengoed E, Maddox P, Bejsovec A, et al. (2001) Drosophila APC2 and Armadillo participate in tethering mitotic spindles to cortical actin. Nat Cell Biol 3: 933–938.
- 68. Kaplan DD, Meigs TE, Kelly P, Casey PJ (2004) Identification of a role for beta-catenin in the establishment of a bipolar mitotic spindle. J Biol Chem 279: 10829–10832.
- 69. Bahmanyar S, Kaplan DD, Deluca JG, Giddings TH Jr, O'Toole ET, et al. (2008) beta-Catenin is a Nek2 substrate involved in centrosome separation. Genes Dev 22: 91–105.
- 70. Bartkova J, Barnes DM, Millis RR, Gullick WJ (1990) Immunohistochemical demonstration of c-erbB-2 protein in mammary ductal carcinoma in situ. Hum Pathol 21: 1164–1167.
- 71. Park K, Han S, Kim HJ, Kim J, Shin E (2006) HER2 status in pure ductal carcinoma in situ and in the intraductal and invasive components of invasive ductal carcinoma determined by fluorescence in situ hybridization and immunohistochemistry. Histopathology 48: 702–707.
- 72. Ramachandra S, Machin L, Ashley S, Monaghan P, Gusterson BA (1990) Immunohistochemical distribution of c-erbB-2 in in situ breast carcinoma–a detailed morphological analysis. J Pathol 161: 7–14.