Deregulated expression of the PCPH proto-oncogene in rat mammary tumors induced with 7,12-dimethylbenz a anthracene

Abstract.

We performed a study on the expression of thePCPH protein in samples corresponding to normal, pre-malignant and malignant stages of the human mammarygland by using protocols of immunohistochemistry andWestern blot analysis with anti-PCPH specific antibodies.Results obtained from the immunohistochemical studyshowed that PCPH was undetectable in samples of normalbreast and of benign diseases, with the exception of glandspresenting apocrine metaplasia, in which an intense PCPHstain was observed both in the basal cytoplasm of the secretorycells and in the apocrine secretion. On the contrary, an intenselabeling was observed in the cytoplasm of neoplastic cells insamples of both ductal and lobular carcinoma

in situ, withthis immunostaining increasing even further in samples ofinfiltrating carcinoma, both ductal and lobular. Western blotanalyses of the same set of samples detected a 47 kDa formas the main PCPH polypeptide present in all cases studied.However, whereas this 47 kDa polypeptide was the onlyPCPH form detected in normal and pre-malignant samples,multiple forms could be detected in carcinoma samples,indicating the presence of altered PCPH polypeptides at thesedisease stages. These results were in agreement with thosefrom the immunohistochemical study and together indicatedthat PCPH protein expression represents a good molecularmarker to follow the process of human breast carcinogenesis.Furthermore, these results suggested that characterization of

the pattern and level of PCPH expression may be a usefultool for early identification of breast cancers.

Introduction

Breast cancer is the most common neoplastic disease diagnosedamong women, and the second leading cause of cancer-relateddeath. The development of breast cancer has been postulatedto be a multi-step process that follows a defined sequence ofqualitatively different events, namely progression of ductalhyperplasia and atypical ductal hyperplasia, which representthe initial stages of neoplastic growth, to carcinoma in situ,invasive carcinoma, and ultimately metastasis, as it has beendocumented for a number of other malignancies (1-3).However, the possibility that normal cells give rise directlyto ductal carcinoma in situ or invasive ductal carcinoma hasnot been definitely ruled out (2,3). Aberrant gene expressionis the hallmark of transformed cells. Alteration in gene copynumber is one of the most important mechanisms leadingto deregulated gene expression and neoplastic transformation.A variety of human cancers carry specifically amplifiedoncogenes with high copy numbers (4). The magnitude ofamplification was found to correlate in different tumors withproliferative activity and aggressive behavior (5-8). Althoughamplification and deletion are the most common mechanismleading to gene deregulation in breast cancer (9), severalother molecular mechanisms are also known to contribute tothe disease onset and progression (10).

The PCPH oncogene was identified by its frequentactivation in Syrian hamster fetal cells during in vitro protocolsof chemical carcinogenesis (11,12). The PCPH gene ishighly conserved in eukaryotes, from yeast to human cells,and it is expressed in most normal adult tissues in Syrianhamsters, mice, rat and humans (13-16). In addition, Westernblot analyses of extracts from a variety of human tumor celllines detected alterations in the expression of the PCPHprotein that were consistent with aberrations of either thesplicing of the PCPH transcript or the post-translationalprocessing of the PCPH protein (15,17).

A potential association of PCPH with at least a subset ofbreast tumors characterized by the over-expression of theerbB-2 receptor (18) was initially suggested by the fact that

INTERNATIONAL JOURNAL OF ONCOLOGY 25: 821-830, 2004

8 2 1

Deregulated expression of the

PCPHproto-oncogene in human breast cancers

MARIA JOSÉ BLÁNQUEZ1, MARIA ISABEL ARENAS2, ISABEL CONDE2,

OSCAR M. TIRADO3, RICARDO PANIAGUA2 and VICENTE NOTARIO3

1Department of Anatomy, Veterinary School, Universidad Complutense; 2Department of Cell Biology and Genetics,

Universidad de Alcalá, Alcalá de Henares, Madrid, Spain; 3Laboratory of Experimental Carcinogenesis,

Department of Radiation Medicine, Georgetown University Medical Center, Washington, DC, USA

Received April 1, 2004; Accepted May 21, 2004

_________________________________________

Correspondence to: Dr Vicente Notario, Laboratory ofExperimental Carcinogenesis, Department of RadiationMedicine, Georgetown University Medical Center, ResearchBuilding, Room E215, 3970 Reservoir Road, NW, Washington,DC20057-1482, USAE-mail: [email protected]

Key words: proto-oncogene, ductal and lobular dysplasia, carcinomain situ, protein expression

heterologous expression of the PCPH oncoprotein in severalmammalian cell culture systems resulted in the reproducible,concomitant elevation of the expression levels of both theerbB-2 receptor and its putative neuregulin ligand (19,20).Furthermore, analyses of PCPH expression in benign andmalignant rat mammary tumors (16) showed that, relative tothe normal mammary gland, the expression levels of PCPHdecreased markedly in benign tumors and became nearlyundetectable in malignant tumors, and that, as in the case ofhuman tumor cell lines (15,17) and of dysplastic lesions ofthe human larynx (21), these changes were accompaniedby the detection of multiple PCPH polypeptide forms inWestern blot analyses (16). However, the status of PCPHexpression in normal and pathological human breast tissueshas not been established to date. Consequently, we studiedthe expression of PCPH in different samples of normaltissue, benign proliferative diseases and carcinoma of thehuman mammary gland in order to elucidate whetheralterations in PCPH expression may be also associated withthese pathologies in humans.

Materials and methods

Pathological specimens. Breast dysplastic samples werecollected from 54 patients diagnosed by the Pathology Serviceof the Hospital Príncipe de Asturias of Alcalá de Henares.Glandular breast lesions were classified as follows: 22 casesdiagnosed of ductal carcinoma (6 ductal carcinoma in situand 16 infiltrating ductal carcinoma), 13 cases of lobularcarcinoma (3 lobular carcinoma in situ and 10 infiltratinglobular carcinoma), 4 cases diagnosed of mixed carcinoma(infiltrating ductal and lobular carcinoma) and 15 casesfollowed established criteria of benign proliferative disease(ductal and lobular hyperplasia, apocrine metaplasia, fibro-adenoma and fibrocystic changes). Peri-menopausal womenwith an average 45.81±13.46 years of age presented benignproliferative disease and in situ carcinomas, while theinfiltrating and invasive tumors derived from post-meno-pausal women with higher average age (60.46±14.95). Allinfiltrating tumor samples were classified by the TNM system(22). After surgery, the hormonal status of each infiltratingand invasive carcinomatous lesion was determined. Removalof tissues was carried out with the consent of the patients'relatives and approval by the Ethics Committee of the Hospital.Each tissue sample was divided into two approximatelyequal portions: one portion was immediately processed forimmunohistochemistry, and the other was snap-frozen inliquid nitrogen and maintained at -80˚C for Western blotanalysis.

Immunohistochemistry. Tissue samples were fixed in 10%(v/v) formaldehyde in phosphate-buffered saline (pH 7.4) for24 h, dehydrated, and embedded in paraffin. Sections (5 µm-thick) were processed following the avidin-biotin-peroxidasecomplex (ABC) method (23). Following deparaffination,sections were hydrated, incubated for 30 min in 0.3% H2O2 inmethanol to reduce endogenous peroxidase activity and, toretrieve the antigen, incubated in 0.1 M citrate-buffer (pH 6.0),for 5 min, in a conventional pressure cooker (24). Afterrinsing in TBS buffer, the slides were incubated with normal

goat serum (NGS) at a 1:5 dilution in TBS, for 30 min, toprevent non-specific binding of the first antibody. Thereafter,the primary anti-PCPH antiserum (25) was applied, at a dilutionof 1:4000 in TBS/NGS, at 4˚C, overnight. Afterwards, thesections were washed twice in TBS and then incubatedwith goat anti-rabbit biotinylated immunoglobulins (Dako,Barcelona, Spain) at 1:500 dilutions. After 1 h of incubationwith the secondary antibody, the sections were incubatedwith a standard streptavidin-biotin complex (Dako) anddeveloped with either the Histostain-SP or the Histostain-SBPsystem (Zymed Laboratories, San Francisco, CA). In allcases, slides were lightly counterstained with hematoxylin.Sections of breast samples processed identically, but notincubated with the primary antibodies, were used as negativecontrols. Positive controls included specimens from normalhuman tissues (larynx, uterus, cervix, testes, skin and liver)known to express PCPH (13-15,21).

Immunoblotting. For Western blot analysis, each of the tissuesamples were homogenized in 0.5 M Tris-HCl buffer (pH 7.4)containing 1 mM EDTA, 12 mM 2-mercaptoethanol, 1 mMbenzamidine, and 1 mM phenylmethylsulphonyl fluoride(PMSF), and supplemented with a cocktail of proteaseinhibitors (10 mM iodoacetamide, 0.01 mg/ml soybean trypsininhibitor, and 1 µl/ml leupeptin) and phosphatase inhibitors(10 mM NaF and 1 mM sodium orthovanadate), in the presenceof 0.5% Triton X-100. Homogenates were centrifuged for10 min at 10,000 rpm. After boiling for 2 min, protein aliquots(20 µg) were electrophoresed in 9% polyacrylamide/SDSminigels. Resolved proteins were transferred for 4 h at0.25 A to nitrocellulose membranes (0.2 µm pore size) thatwere then blocked for 1 h with 5% blotto in 0.05 M Tris-HCl,and incubated overnight with the PCPH-antibody (15,17) at1:10000 in TBS. After extensive washing with TBS/Tween-20,the membranes were incubated for 1 h with a horseradishperoxidase-labeled rabbit anti-goat secondary antibody(Chemicon, UK) diluted 1:4000 in TBS; after intensivewashing, the membranes were developed with an enhancedchemiluminescence (ECL) kit, following the proceduredescribed by the manufacturer (Amersham, Buckinghamshire,UK). Blots were stripped and re-probed with an anti-humanß-actin monoclonal antibody (Amersham) to control forequal sample loading.

Results

Immunohistochemical characterization of samples of normaland benign lesions. Negative controls were performed onserial sections of tumor samples by incubation withoutprimary antibody. No background immunoreaction to PCPHwas observed in any of these controls (Fig. 1a). Samplesderived from normal, PCPH-expressing human tissues, such asthe one from human skin (Fig. 1b), were used as positivecontrols. As expected these samples showed an intensestaining for PCPH, which in the case of the skin waslocalized in the cytoplasm of the keratinocytes (Fig. 1b). Innormal breast samples and in the normal acini and ductspresent in samples with benign hyperplasia, PCPH immuno-staining was not observed in either secretory or myoepithelialcells. However, in these regions, intense staining was observed

BLÁNQUEZ et al: PCPH EXPRESSION IN HUMAN BREAST NEOPLASIA8 2 2

INTERNATIONAL JOURNAL OF ONCOLOGY 25: 821-830, 2004 8 2 3

Figure 1. (a), Negative control section from ductal carcinoma in situ showing no immunoreaction when it was incubated without the primary antibody. x200.(b), Positive control section from human skin. An intense reaction to PCPH antibody was observed in the cytoplasm of keratinocytes (asterisk). Star, dermis.x400. (c), Normal mammary gland showing a negative immunoreaction for PCPH in the ducts (asterisk) and the acini (star). Blood cells are positive (arrow).x200. (d), No labeling for PCPH was observed in ductal hyperplasia. x300. (e), Negative immunoreaction to PCPH antibody in lobular hyperplasia. Arrow,myoepithelial cells showing no reaction to PCPH. x300. (f), Negative reaction to PCPH-antiserum in fibroadenoma. x200. (g), Apocrine metaplasia showingan intense granular PCPH immunoreactivity in the basal cytoplasm of the epithelial cells and in the blebs detached of the apical cytoplasm (arrow). x300.

BLÁNQUEZ et al: PCPH EXPRESSION IN HUMAN BREAST NEOPLASIA8 2 4

Figure 2. (a), Ductal carcinoma in situ with solid pattern: an intense immunoreaction to PCPH was detected in the cytoplasm of neoplastic cells, whilemyoepithelial cells showed no reaction (arrow). x400. (b), Ductal carcinoma in situ with cribiform pattern: intense PCPH immunostaining was observed in thecytoplasm of the neoplastic cells. x300. (c), Ductal carcinoma in situ with papillary pattern: intense PCPH-labeling can be observed in the epithelial cellscytoplasm. x300. (d), Acinus of apocrine carcinoma. Immunoreaction to PCPH can be observed in the basal cytoplasm and in the blebs detached of the apicalcytoplasm (arrow). x300. (e), Infiltrating ductal carcinoma. PCPH immunoreaction was located in both the plasma membrane and the cytoplasm with a patchypattern. x400. (f), Infiltrating lobular carcinoma with inflammatory component showing an intense immunolabeling to PCPH in the cytoplasm of theneoplastic cells. Arrow, inflammatory cells. x200. Inset showing at high magnification a strong PCPH-labeling in the cytoplasm of monocytes. x400.

in blood cells (Fig. 1c), but the stroma was always negativefor PCPH staining.

The benign lesions examined included cases of: i) ductaland lobular hyperplasia, characterized by proliferatingepithelial and myoepithelial cells with the appearance ofsecondary lumens (ductal hyperplasia) as well as aggregatesof compact proliferating tubules lined by epithelial and myo-epithelial cells with sparse intervening stroma (lobular hyper-plasia) (Fig. 1d and e); ii) fibroadenoma with intracanalicularpattern where the stroma exhibited a radial growth withdeposition of fibers perpendicular to the epithelial elements(Fig. 1f); and iii) fibrocystic changes consisting of groups ofcystically dilated, round to ovoid spaces lined by slightlyattenuated epithelial and myoepithelial cell layers. Similar tothe case of the normal acini (Fig. 1c), immunostaining forPCPH was negative in all benign samples studied (Fig. 1d-f).

The only exception corresponded to zones presenting apocrinemetaplasia. In these apocrine glands, an intense immuno-reaction to PCPH was observed in both the apical and thebasal cytoplasm with a granular pattern. This labeling wasstronger in the blebs becoming detached from the apicalcytoplasm (Fig. 1g). In all these pathologies, the stroma wasalways negative.

Immunohistochemical characterization of malignant samples.In samples that presented ductal carcinoma in situ, a positivereaction to PCPH antibody was observed in the whole tumoralarchitecture, regardless of whether it corresponded to a solidstructure (Fig. 2a), a cribiform pattern characterized by theproliferation of a uniform population of cells forming asieve-like arrangement with secondary lumens rounded andrigid bridges (Fig. 2b), or a papillary pattern with epithelialtufts projecting into the lumen (Fig. 2c). Only one of thesecases showed no reaction for PCPH (Table I). These differenttumoral structures showed undisrupted basal membranes,maintaining the myoepithelial cells that, as in the case ofbenign pathologies, showed no reaction with the PCPH-antibody (Fig. 2a-c). In samples with ducts presenting apocrinecarcinoma, some of the neoplastic cells contained intracyto-plasmic granules that showed an intense immunostaining forPCPH (Fig. 2d). In addition, a strong PCPH-label was alsovisualized in the apical cytoplasmic protrusions (Fig. 2d),similar to the staining encountered in zones of apocrinemetaplasia observed in benign samples (compare Fig. 2dwith Fig. 1f).

In samples of infiltrating ductal carcinoma, solid clustersof tumor tissue were observed. In these clusters, similar towhat was observed in ductal carcinoma in situ, PCPH immuno-reaction was located in the cytoplasm of the neoplasticcells, but the staining was relatively more intense, especiallyin what appeared as intracytoplasmic granular structures(Fig. 2e). Only one of these cases was negative for PCPHstaining (Table II). In samples of infiltrating ductal carcinomapresenting an inflammatory component, the cytoplasm ofthe inflammatory cells showed a strong PCPH staining(Fig. 2f).

In samples of lobular carcinoma in situ, the acini wereformed by a uniform population of cells proliferating in asolid occlusive manner. In all samples, these cells showed anintense PCPH-labeling in their cytoplasm (Fig. 3a and b).This same immunostaining pattern was also observed in mostof the samples of infiltrating lobular carcinoma (Fig. 3c and d),which were characterized by the presence of cohesive andgenerally small cells diffusely invading the stroma. Onlythree of the cases of infiltrating lobular carcinoma werenegative for PCPH (Table II). The reactivity to PCPH insamples of infiltrating lobular carcinoma was consistentlystronger in areas close to the adipose tissue (Fig. 3e). A signet-ring cell component was encountered in 2 of the 10 patientspresenting infiltrating lobular carcinoma. The cells in thesestructures presented intracytoplasmic lumens that causedcompression of the nuclei to a thin crescent morphology. Anintense immunoreaction to PCPH was observed in theremnant cytoplasm (Fig. 3d).

Results from the immunohistochemistry study aresummarized in Tables I and II, which also include clinico-

INTERNATIONAL JOURNAL OF ONCOLOGY 25: 821-830, 2004 8 2 5

Table I. Clinicopathological data of the 24 patients with benignpathologies and in situ carcinomas and relative expressionlevels of PCPH.–––––––––––––––––––––––––––––––––––––––––––––––––Case Age Diagnostic PCPH no. (years) expression–––––––––––––––––––––––––––––––––––––––––––––––––1 35 Adenosis -

2 47 Apocrine metaplasia +

3 42 Ductal hyperplasia -

4 51 Ductal hyperplasia -

5 46 Ductal hyperplasia -

6 53 Ductal hyperplasia -

7 37 Ductal hyperplasia -

8 46 Ductal hyperplasia -

9 58 Ductal hyperplasia -

10 25 Ductal hyperplasia -

11 16 Fibroadenoma -

12 59 Fibroadenoma -

13 91 Ductal and lobular hyperplasia -

14 54 Ductal and lobular hyperplasia -

15 50 Ductal and lobular hyperplasia -

16 54 DCIS +/++

17 45 DCIS ++

18 42 DCIS -

19 72 DCIS ++

20 50 DCIS ++

21 65 DCIS ++

22 53 LCIS ++

23 46 LCIS ++

24 78 LCIS ++–––––––––––––––––––––––––––––––––––––––––––––––––DCIS, ductal carcinoma in situ. LCIS, lobular carcinoma in situ.Immunoreaction intensity: -, negative; +, moderate; ++, intense;+++, strong.–––––––––––––––––––––––––––––––––––––––––––––––––

pathological data of the samples, which are arranged accordingto the favorable (benign pathologies and carcinoma in situ,Table I) or unfavorable (infiltrating carcinomas, Table II)prognosis of the cases.

Expression of PCPH polypeptides in normal and neoplastichuman breast samples. Extracts prepared from all normal,benign and malignant human breast samples were tested forPCPH expression in Western blot analyses with anti-PCPH

antiserum. Fig. 4 shows the results from a blot arranged to berepresentative of the various patterns of PCPH polypeptideexpression detected in our study. Results demonstrate theexistence of profound quantitative and qualitative differencesin PCPH expression between samples corresponding tocases of benign or malignant pathologies. Expression of apolypeptide of 47 kDa, a size matching that deduced from thehuman PCPH cDNA sequence (15), was a feature commonto all samples tested. However, while this 47 kDa species

BLÁNQUEZ et al: PCPH EXPRESSION IN HUMAN BREAST NEOPLASIA8 2 6

Table II. Clinicopathological data of the 30 patients with breast infiltrating carcinomas and relative expression levels of PCPH.–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––Case Age TNM Nodal Disease-free Tumor ER PR PCPH no. (years) status status (months) types expression–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––1 49 T2N1M0 1/7 18 Ductal +++ +++ +

2 41 T3N1M0 7/14 22 Ductal ++/+++ ++/+++ +

3 56 T4N2M0 5/12 24 Ductal ++ + ++

4 65 T1N2M0 0/12 - Ductal - - ++

5 30 T2N1M0 - - Ductal - - ++

6 53 T2N1M0 13/20 31 Ductal +++ - ++

7 58 T2N0M0 0/7 54 Ductal +++ +++ -

8 43 T2N0M0 0/12 55 Ductal - - +

9 91 T3N0M0 0/10 36 Ductal - - ++

10 78 T2N1M0 3/8 55 Ductal ++ - +++

11 74 T2N0M0 0/7 54 Ductal +++ ++ ++

12 72 T2N2M0 14/14 36 Ductal +++ +++ ++

13 71 T1N1M0 1/10 54 Ductal +++ - ++

14 59 T2N0M0 0/10 55 Ductal +++ - ++

15 75 T1N1M0 2/14 51 Ductal ++ ++ ++

16 70 T1N0M0 0/16 24 Ductal +++ +++ ++

17 78 T4N2M0a 15 Lobular +++ +++ +

18 56 T2N1M0 3/9 12 Lobular - ++ +

19 72 T1N1M0 1/3 45 Lobular +++ +++ -

20 35 T2N1M0 6/10 35 Lobular +++ +++ ++

21 67 T1N0M0 0/18 45 Lobular +++ +++ -

22 82 T2N0M0 0/6 42 Lobular + + ++

23 50 T1N0M0 0/25 41 Lobular +++ +++ -

24 77 T2N2M0 9/12 41 Lobular +++ ++ ++

25 54 T1NxM0 - 49 Lobular +++ ++ +++

26 64 T1N0M0 0/16 36 Lobular + - +++

27 48 T1N1M0 1/15 26 Mixed - ++ ++

28 46 T1N0M0 0/14 51 Mixed ++ ++ ++

29 54 T2N1M0 1/4 29 Mixed +++ +++ ++

30 46 T2N1M0 2/7 26 Mixed +++ +++ ++–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––T, primary tumor. N, regional lymph nodes. M, distant metastases. ER, estrogen receptor. PR, progesterone receptor. aMassive infiltration. Itwas not possible to determinate the total number of affected nodules. -, not determinated. Immunoreaction intensity: -, negative; +, moderate;++, intense; +++, strong.–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

INTERNATIONAL JOURNAL OF ONCOLOGY 25: 821-830, 2004 8 2 7

Figure 3. (a), Lobular carcinoma in situ showing an intense reaction to PCPH. Asterisk, normal duct. x200. (b), High magnification of acini from lobularcarcinoma in situ showing a strong label with PCPH antibody in the cytoplasm of neoplastic cells. x400. (c), Lobular infiltrating carcinoma. The neoplasticcells were positive for PCPH. x300. (d), An intense labeling for PCPH was observed in the ‘indian files’ pattern of lobular infiltrating carcinoma. x300. (e), Lobularcarcinoma infiltrating in the adipose tissue showing an intense immunoreaction to PCPH. x200. (f), Signet-ring cell carcinoma. x200. The inset corresponds toa high magnification of the cells included in the rectangle, showing the intracytoplasmic lumens, the nuclei, and the positive reaction for PCPH in the cyto-plasm. x400.

was the only PCPH polypeptide detectable in samples frombenign cases (Fig. 4, lanes 1-3), carcinoma-derived samplesexpressed a variety of PCPH polypeptides, including somespecies of low molecular mass, ~15-30 kDa, and other larger(>60 kDa) PCPH polypeptides (Fig. 4, lanes 4-8). In general,the more aggressive the tumor, the greater the overall intensity(Fig. 4, lane 6 vs. lanes 4 and 5) and the complexity of thepattern of PCPH polypeptides (Fig. 4, lanes 6-8 vs. all others)detected in the corresponding samples.

Discussion

This report represents the first study of the possible involvementof the PCPH gene and gene product in the development ofhuman breast carcinoma. Data from our immunohistochemistryand Western blot analyses demonstrated the existence of aremarkable association between the presence of alterationsin PCPH expression and the advanced malignant stage ofthe breast carcinomas. Most normal mouse, rat and humantissues express either of two major PCPH polypeptide formsof 47 kDa or 27 kDa (14,15,21). The 47 kDa species matchesthe size of the translation products deduced from the openreading frames of the PCPH proto-oncogene cDNAs (14,15)and is more ubiquitously expressed, whereas the 27 kDaform has been detected as the only major PCPH polypeptidein normal rat tissues such as the adipose tissue, the skin and,more importantly, the mammary gland (16). The alterationsin PCPH expression observed in Western blot analyses ofextracts from human breast tissues revealed a common patternin the different types of carcinoma (ductal and lobular)characterized by the appearance of PCPH forms in the rangeof 15-30 kDa. In addition, the presence of polypeptides larger

than 60 kDa was a feature common to samples that presentedan inflammatory component. However, all samples fromhuman breast tissue expressed the 47 kDa form, the onlyone present in normal tissue and in cases of benign disease.These data are opposite to those reported from the analysis ofPCPH expression in the normal gland and in chemically-induced mammary tumors of the rat (16). These authorsobserved that the 27 kDa PCPH form is characteristic of thenormal rat mammary gland, whereas most tumors expressedthe 47 kDa polypeptide. Although both rat and humanmalignant tumors have in common the presence of aberrant(larger and smaller) PCPH polypeptides, results from thecurrent study demonstrate that PCPH expression is regulatednot only in a tissue-specific fashion, as it was shown earlier(14-16), but also in a species-specific manner.

A close agreement was observed between the resultsobtained by Western blot analysis and those obtained byimmunohistochemistry. Sections from samples of normalbreast tissue and of benign diseases showed negative immuno-reactivity to the PCPH antibody in both secretory and myo-epithelial cells of acini and in the epithelial cells of ducts.However, the 47 kDa PCPH polypeptide was detected inWestern blot analyses of extracts of samples from caseswith benign pathologies that did not stain for PCPH inimmunohistochemistry assays. This apparent contradictionmay be explained in several ways. On the one hand, thegenerally lower intensity of the 47 kDa form detected in benignsamples may be correlated with a different conformation ofthe PCPH polypeptide in the tissue sections that wouldmake it inaccessible to the antibody; on the other hand, it ispossible that the PCPH detected in Western blots may reflectthe intense PCPH labeling observed in blood cells present inthese samples or in the zones presenting apocrine metaplasia.The positive reaction observed by immunohistochemistryin apocrine metaplasia is not unexpected since there is amorphologic progression from benign metaplasia to apocrinecarcinoma in situ, which involves increasing cellular atypiaand proliferation (26,27). Furthermore, it has been reportedthat both breast apocrine metaplasia and apocrine carcinomacells produce proteins belonging to the ‘gross cystic diseasefluid protein’ (GCDFP) family, also known as prolactin-inducible (PIP) proteins, which are not produced by normalepithelial cells (28,29). However, an association of the PCPHprotein with these proteins has not been explored to date.

The multiple PCPH species observed in human breastcarcinoma by Western blot analysis and the expressionpatterns observed by immunohistochemistry might be relatedto genetic alterations reported for human tumors at the 14q24chromosomal region (30), where the PCPH gene is localized(15). Xie and coworkers (31) reported multiple genes whichwere over-expressed in breast cancer cells due to gains onseveral chromosomes, including the 14q21-q24 region. Thereare a number of reports describing chromosome 14 gains inhuman breast cancer (32-34), involving genes implicated incycle cell regulation (32,35,36) and metastasis potential(34,37). Thus, PCPH expression in human breast cancercould be associated with the progression stage of the trans-formation of normal epithelial cells into fully neoplastic,tumorigenic cells. However, although PCPH expressionincreased markedly in samples of carcinoma in situ, and even

BLÁNQUEZ et al: PCPH EXPRESSION IN HUMAN BREAST NEOPLASIA8 2 8

Figure 4. Immunoblot representative of the patterns of PCPH expressiondetected in the different types of normal, benign and malignant humanbreast samples. The mobility of protein standards is shown on the left. Lanes1-3, pattern observed with extracts from tissues that presented benigndiseases, showing only a polypeptide of about 47 kDa. Lanes 4 and 5,samples from ductal carcinoma showed two low molecular weight species inaddition to the 47 kDa polypeptide. Lane 6, samples of lobular carcinomashowed a pattern similar to that observed in lane 4, but the PCPH-relatedforms showed stronger intensity. Lanes 7 and 8, mixed breast carcinomasamples (ductal and lobular) with an inflammatory component showingmultiple PCPH-polypeptide species. After stripping, immunoreactivity withan anti-actin antibody was used as loading control (actin, bottom panel).

more strongly in invasive ductal and lobular carcinoma, thesechanges did not appear to correlate with clinicopathologicaldata, including tumor grade, nodal status, hormonal status,age and disease-free time. Nevertheless, it is noteworthy thatit has been recently demonstrated that chemotherapy-inducedapoptosis and Bcl-2 levels correlate with the response ofbreast cancer to chemotherapy (38). Because increasing thecellular resistance to apoptosis is an important componentof the transforming activity of the PCPH oncoprotein (39), itseems reasonable to hypothesize that there may be a correlationbetween the levels of Bcl-2 and the PCPH oncoprotein inbreast tumors. Indeed, the validity of this notion seems tobe supported by preliminary results (data not shown) fromexperiments ongoing in our laboratory, designed to test thishypothesis with cultured human breast tumor cell lines.

Overall our results strongly indicate that the PCPH proteinrepresents a good molecular marker to follow human breastcarcinogenesis. Furthermore, the detection of atypical PCPHpolypeptides in non-invasive breast tumors suggests thatimmunodetection of PCPH can be applied for the earlyidentification of malignant cells.

Acknowledgments

This work was supported by USPHS Grant RO1-64472 fromthe National Cancer Institute.

References

1. Farber B: The multistep nature of cancer development. CancerRes 44: 4217-4223, 1984.

2. Page DJ and Dupont WD: Anatomic markers of humanpremalignancy and risk breast cancer. Cancer 66: 1326-1335,1990.

3. Russo J, Calaf G and Russo IH: A critical approach to themalignant transformation of human breast epithelial cells withchemical carcinogens. Crit Rev Oncog 44: 403-417, 1993.

4. Knuutila S, Bjorkqvist AM, Autio K, Tarkkanen M, Wolf M,Monni O, Szymanska J, Larramendy ML, Tapper J, Pere H,El-Rifai W, Hemmer S, Waseniius VM, Vidgren V and Zhu Y:DNA copy number amplifications in human neoplasms: reviewof comparative genomic hybridization studies. Am J Pathol 152:1107-1123, 1998.

5. Kreipe H, Feist H, Fischer L, Felgner J, Heidorn K, Mettler Land Parwaresch R: Amplification of c-myc but not of c-erB-2 isassociated with high proliferative capacity in breast cancer.Cancer Res 53: 1956-1961, 1993.

6. Amadori D, Maltoni M, Volpi A, Nanni O, Scarpi E, Renault B,Pellegata NS, Gaudio M, Magni E and Ranzani GN: Geneamplification and proliferative kinetics in relation to prognosisof patients with gastric carcinoma. Cancer 79: 226-232, 1997.

7. Press MF, Bernstein L, Thomas PA, Meisner LF, Zhou JY, Ma Y,Hung G, Robinson RA, Harris C, El-Naggar A, Slamon DJ,Phillips RN, Ross JS, Wolman SR and Flom KJ: HER-2/neugene amplification characterized by fluorescence in situhybridation: poor prognosis in node negative breast carcinomas.J Clin Oncol 15: 2894-2904, 1997.

8. Galanis E, Buckner J, Kimmel D, Jenkins R, Alderete B,O'Fallon J, Wang CH, Scheithauer BW and James CD: Geneamplification as a prognostic factor in primary and secondaryhigh-grade malignant gliomas. Int J Oncol 13: 717-724, 1998.

9. Bergstein K: Molecular alterations in breast cancer. In: BreastCancer: Molecular Genetics, Pathogenesis, and Therapeutics.Bowcock AM (ed.) Humana Press, Totowa, pp143-170, 1999.

10. Dankort DL and Muller WJ: Signal transduction in mammarytumorigenesis: a transgenic perspective. Oncogene 19: 1038-1044,2000.

11. Notario V, Castro R, Flessate DM, Doniger J and Di Paolo JA:Frequent activation of non-ras transforming sequences inneoplastic Syrian hamster cells initiated with chemical carcino-gens. Oncogene 5: 1425-1430, 1990.

12. Velasco JA, Castro R, Avila MA, Laborda J, Di Paolo JA,Cansado J and Notario V: Cph, a novel oncogene whichcooperates with H-ras in the transformation of NIH/3T3fibroblasts. Oncogene 9: 2065-2069, 1994.

13. Velasco JA, Zimonjic DB, Popescu NC, Cansado J, Di Paolo JA,Albor A and Notario V: Tissue-specific expression, evolutionaryconservation and localization of the cph proto-oncogene onSyrian hamster chromosome X. Oncogene 12: 2713-2717,1996.

14. Recio JA, Zambrano N, De la Peña L, Powers C, Siwarski D,Huppi K and Notario V: cDNA isolation, expression andchromosomal localization of the mouse Pcph proto-oncogene.Mol Carcinog 26: 130-136, 1999.

15. Recio JA, Zambrano N, De la Peña L, Reig JA, Rhoads A,Rouzaut A and Notario V: The human PCPH proto-oncogene:cDNA identification, primary structure, chromosomal mapping,and expression in normal and tumor cells. Mol Carcinog 27:1-8, 2000.

16. Solanas M, Escrich E, Rouzaut A, Costa I, Martínez A andNotario V: Deregulated expression of the PCPH proto-oncogenein rat mammary tumors induced with 7,12-dimethyl-benz[a]anthracene. Mol Carcinog 33: 219-227, 2002.

17. Rouzaut A, Recio JA and Notario V: Expression of the proteinproduct of the PCPH proto-oncogene in human tumor cell lines.Radiat Res155: 181-187, 2001.

18. Zhou BP and Hung MC: Dysregulation of cellular signaling byHER2/neu in breast cancer. Semin Oncol 30: 38-48, 2003.

19. Avila MA, Velasco JA, Cho C, Lupu R, Wen D and Notario V:Hyperactive autocrine loop mediated by a NDF-related factor inneoplastic hamster embryo fibroblasts expressing an activatedcph oncogene. Oncogene 10: 963-971, 1995.

20. Velasco JA, Feijoo E, Avila MA and Notario V: Secretion ofneu differentiation factor-like polypeptides by cph-transformedfibroblasts: cloning and characterization of Syrian hamsterneuregulin cDNAs. Mol Carcinog 21: 156-163, 1998.

21. Blánquez MJ, Regadera J, Mariño J, Newman RE and Notario V:Gradual deregulation and loss of PCPH expression in theprogression of human laryngeal neoplasia. Mol Carcinog 35:186-195, 2002.

22. UICC: TNM Classification of Malignant Tumor. InternationalUnion Against Cancer, Geneva, 1968.

23. Hsu SM, Reiner L and Fanger H: Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase technique.A comparison between ABC and unlabeled antibody PAPprocedures. J Histochem Cytochem 29: 577-580, 1981.

24. Norton A, Jordan S and Yeomans P: Brief high-temperatureheat denaturation (pressure cooking): a simple and effectivemethod of antigen retrieval for routinely precessed specimens. JPathol 173: 371-379, 1994.

25. Tirado OM, Mateo-Lozano S, Sanders S, Dettin LE and Notario V:The PCPH oncoprotein antagonizes the proapoptotic role of themammalian target of rapamycin in the response of normalfibroblasts to ionizing radiation. Cancer Res 63: 6290-6298,2003.

26. Raju Y, Zarbo RJ, Kubus J and Schultz DS: The histologicspectrum of apocrine breast proliferations. Hum Pathol 24:173-181, 1993.

27. O'Malley FP, Page DL, Nelson EH and Du Pont WD: Ductalcarcinoma in situ of the breast with apocrine cytology. HumPathol 25: 164-168, 1994.

28. Mazoujian G, Pinkus GS, Davis S and Haagensen DE Jr:Immunohistochemistry of a gross cystic disease fluid protein(GCDFP-15) of the breast. A marker of apocrine epithelium andbreast carcinomas with apocrine features. Am J Pathol 110:105-112, 1983.

29. Haagensen DE Jr: Is cystic disease related to breast cancer? AmJ Surg Pathol 15: 687-694, 1991.

30. Mitelman F, Mertens F and Johansson B: A breakpoint map ofrecurrent chromosomal rearrangements in human neoplasia. NatGenet: 417-474, 1997.

31. Xie D, Jauch A, Miller CW, Bartram CR and Koeffler HP:Discovery of over-expressed genes and genetic alterations inbreast cancer cells using a combination of suppressionsubtractive hybridization, multiplex FISH and comparativegenomic hybridization. Int J Oncol 21: 499-507, 2002.

32. Rasmussen UB, Wolf C, Mattei MG, Chenard MP, Bellocq JP,Chambon P, Rio MC and Basset P: Identification of a newinterferon-alpha-inducible gene (p27) on human chromosome14q32 and its expression in breast carcinoma. Cancer Res 53:4096-4101, 1993.

INTERNATIONAL JOURNAL OF ONCOLOGY 25: 821-830, 2004 8 2 9

33. Adeyinka A, Mertens F, Idvall I, Bondeson L, Ingvar C,Mitelman F and Pandis N: Different patterns of chromosomalimbalances in metastasizing and non-metastasizing primarybreast. Int J Cancer 84: 370-375, 1999.

34. Martin MD, Fischbach K, Osborne CK, Mohsin SK, Allred DCand O'Connell P: Loss of heterozygosity events impeding breastcancer metastasis contain the MTA1 gene. Cancer Res 61:3578-3580, 2001.

35. Lehmann U, Glöckner S, Kleeberger W, Feist H, vonWasielewski R and Kreipe H: Detection of gene amplificationin archival breast cancer specimens by laser-assisted micro-dissection and quantitative real-time polymerase chain reaction.Am J Pathol 156: 1855-1864, 2000.

36. Bassal S, Nomura N, Venter D, Brand K, McKay MJ and vander Spek PJ: Characterization of a novel human cell-cycle-regulated homologue of Drosophila dlg1. Genomics 77: 5-7,2001.

37. O'Connell P, Fischbach K, Hilsenbeck S, Mohsin SK, Fuqua SA,Clark GM, Osborne CK and Allred DC: Loss of heterozygosityat D14S62 and metastatic potential of breast cancer. J NatlCancer Inst 91: 1391-1397, 1999.

38. Buchholz TA, Davis DW, McConkey DJ, Symmans WF,Valero V, Jhingran A, Tucker SL, Pusztai L, Cristofanilli M,Esteva FJ, Hortobagyi GN and Sahin AA: Chemotherapy-induced apoptosis and Bcl2 levels correlate with breast cancerresponse to chemotherapy. Cancer J 9: 33-41, 2003.

39. Velasco JA, Avila MA and Notario V: The product of the cphoncogene is a truncated, nucleotidebinding protein thatenhances cellular survival to stress. Oncogene 18: 689-701,1999.

BLÁNQUEZ et al: PCPH EXPRESSION IN HUMAN BREAST NEOPLASIA8 3 0