Originally published In Press as doi:10.1074/mcp.M500348-MCP200 on November 28, 2005.
Molecular & Cellular Proteomics 5:462-483, 2006.
© 2006 by The American Society for Biochemistry and Molecular Biology, Inc.
Research
Apocrine Cysts of the Breast
Biomarkers, Origin, Enlargement, and Relation with Cancer Phenotype*
Julio E. Celis
,
,¶,
Pavel Gromov,,
José M. A. Moreira,,
Teresa Cabezón,,
Esbern Friis,||,
Ilse M. M. Vejborg,**,
Gottfried Proess,,
Fritz Rank, and
Irina Gromova,
From the Danish Centre for Translational Breast Cancer Research (DCTB) and Department of Proteomics in Cancer, Institute of Cancer Biology, Danish Cancer Society, Strandboulevarden 49, || Department of Breast and Endocrine Surgery, ** Radiology Department, Mammography Division, and Department of Pathology, the Centre of Diagnostic Investigations, Rigshospitalet, DK-2100 Copenhagen, Denmark and Eurogentec, Parc Scientifique du Sart Tilman, 4102 Seraing, Belgium
 |
ABSTRACT
|
|---|
Up to one-third of women aged 30–50 years have cysts in their breasts and are presumed to be at increased risk of developing breast cancer. Here we present an extensive proteomic and immunohistochemistry (IHC) study of breast apocrine cystic lesions aimed at generating specific biomarkers and elucidating the relationship, if existent, of apocrine cysts with cancer phenotype. To this end we compared the expression profiles of apocrine macrocysts obtained from mastectomies from high risk cancer patients with those of cancerous and non-malignant mammary tissue biopsies collected from the same patients. We identified two biomarkers, 15-hydroxyprostaglandin dehydrogenase and 3-hydroxymethylglutaryl-CoA reductase, that were expressed specifically by apocrine type I cysts as well as by apocrine metaplastic cells in type II microcysts, terminal ducts, and intraductal papillary lesions. No expression of these markers was observed in non-malignant terminal ductal lobular units, type II flat cysts, stroma cells, or fat tissue as judged by IHC analysis of matched non-malignant tissue samples collected from 93 high risk patients enrolled in our cancer program. IHC analysis of the corresponding 93 primary tumors indicated that most apocrine changes have little intrinsic malignant potential, although some may progress to invasive apocrine cancer. None of the apocrine lesions examined, however, seemed to be a precursor of invasive ductal carcinomas, which accounted for 81% of the tumors analyzed. Our studies also provided some insight into the origin, development, and enlargement of apocrine cysts in mammary tissue. The successful identification of differentially expressed proteins that characterize specific steps in the progression from early benign lesions to apocrine cancer opens a window of opportunity for designing and testing new approaches for pharmacological intervention, not only in a therapeutic setting but also for chemoprevention, to inhibit cyst development as both 15-hydroxyprostaglandin dehydrogenase and 3-hydroxymethylglutaryl-CoA reductase are currently being targeted for chemoprevention strategies in various malignancies.
The human breast presents many benign lesions that involve both the glandular and stromal tissues. These include fibrocystic changes, benign breast tumors, and breast inflammatory disease (1). Fibrocystic changes affect more than 50% of women during their lifetime and are comprised of cystic dilation of ducts, apocrine metaplasia of ductal epithelium, fibrosis, adenosis, and intraductal epithelial proliferation (Ref. 2 and references therein).
Cysts are fluid-filled sacs that usually develop in the upper half of the breast and are most common in women between 30 and 50 years of age (3). They are also found in menopausal women on hormone replacement therapy. Cysts start as a microscopic dilation of the ductules (microcyst) but can enlarge and reach sizes of a couple of centimeters in diameter (macrocysts) causing pain and discomfort due to the tension effect on the surrounding stroma. In addition, cysts may rupture, eliciting chronic inflammation. Draining of the cystic fluid by fine needle aspiration often alleviates the symptoms, but although cysts may disappear following aspiration, in about a third of the cases they recur (4).
Cysts are divided into apocrine (type I, acidophil cells), flattened (type II, basophil cells), and intermediate types based on the K+/Na+ concentration ratio (5–9). Type I cysts have a higher K+/Na+ ratio (
1.5) (5), are lined by metaplastic apocrine secretory epithelium that resembles sweat glands, and exhibit tight cell-cell junctions. Apocrine changes can be complex and usually show apical snouts that are shed off to the lumen of the duct (10). In some cases, cells within type I cysts present transitions from cuboidal to flat apocrine epithelia, most likely representing differences in their stage of differentiation and/or metabolic activity. Breast lesions undergoing apocrine differentiation in most cases do not express estrogen receptor (ER)1-
or progesterone receptor (PR) but are often positive for the androgen receptor (11–14). At present, little is known about the molecular mechanisms underlying apocrine differentiation of mammary lesions.
Type II cysts on the other hand have an electrolyte composition more similar to serum (K+/Na+ ratio <1.5) and are lined by flat epithelial cells showing open cell-cell junctions (15). Contrary to the epithelia of normal breast ductules that show heterogeneous and lower levels of ER- expression, flat type II microcysts are composed of contiguous cells that stain strongly for ER- and PR (16, 17). Flat type II microcysts are believed to be derived from a cystic transformation of the glandular cells that become columnar followed by dilation of the ductules due to secretion (18). To date little is known about the mechanism(s) that leads to cyst development and enlargement, although there is some information indicating that their development is regulated by hormones.
The protein, hormone, and electrolyte content of breast cyst fluid has been the subject of numerous studies that have shown that this fluid contains a variety of biologically active substances including steroids (19–23) and cytokines and growth factors (insulin-like growth factor-binding protein-3, tumor necrosis factor (TNF)-, transforming growth factor- and -ß, interleukin (IL)-1, and IL-6)) (24–28) as well as proteins such as prostate-specific antigen (9, 29–31), proteases (kallikrein 6 and pepsinogen C) (32, 33), c-ErbB-2 oncoprotein (ErbB-2 protein) (34), and carcinoembryonic antigen (34). Zinc-2-glycoprotein (35, 36), apolipoprotein D (apoD) (37, 38), also known as gross cystic disease fluid protein (GCDFP)-24, and GCDFP-15 (39, 40) are the three most abundant proteins present in breast gross cystic disease fluid. Whether breast cysts are active endocrine glands or inactive reservoirs of hormones, growth factors, and other compounds is at present unknown.
For some time, microcysts were considered as a normal stage in the development and involution of the breast (16, 18, 41), but lately several publications have associated them with the transition of mammary epithelium from a benign to malignant phenotype (16, 17, 42, 43). In addition, clinical follow-up studies have indicated that the presence of cysts increases the risk for subsequent development of breast cancer (8, 44–47). In particular, women with type I apocrine cysts have been reported to have a 2–4 times higher risk for breast cancer development as compared with women with type II cysts (23, 48). The latter observations have been contested by Dixon et al. (47), who reported that all women with breast cysts are at increased risk of breast cancer irrespective of the cyst type. Clearly the results are ambiguous at this point as different authors have described apocrine cysts as either representing a risk factor, a non-obligate precursor lesion of apocrine carcinoma, or benign lesions with no correlation with malignancy (3, 49–54).
We are part of a large multidisciplinary long term research initiative, the Danish Centre for Breast Cancer Research (DCTB), aimed at the comprehensive analysis of breast cancer lesions using multiple experimental paradigms from genomics, proteomics, and transcriptomics. Our strategy is based on the study of mammary tumors and matched benign fresh tissues collected from mastectomies of high risk cancer patients along with the integration of the "omic" and clinical datasets (55–58). Gross pathological inspection of the mastectomy specimens often revealed the presence of macrocysts of various sizes (Fig. 1) that were localized at varying distances to the tumor. Considering that apocrine metaplasia occurs in a spectrum of breast epithelial lesions (12, 60) and taking advantage of the unique possibility of having access to matched tissue samples collected from the same patient as well as our experience in biomarker discovery (61–64), we carried out a comprehensive proteomic study of apocrine cysts with the aim of generating specific biomarkers for breast apocrine metaplasia that could provide some insight into their origin and mechanisms underlying their development and enlargement as well as to enlighten their relationship, if existent, with cancer phenotype.
View larger version (48K):
[in this window]
[in a new window]
|
FIG. 1. Overview of the various steps involved in the analysis of cyst tissue and fluid. A similar procedure was used to analyze non-malignant and tumor specimens. As seen from the keratin 19 staining of macrocyst 81, the cyst is composed of several microcysts.
|
|
|
EXPERIMENTAL PROCEDURES
|
|---|
Patient Selection
Thirty-six women with primary, operable, high risk2 invasive breast cancer were selected for this prospective study (DCTB patient numbers 59–94 in Table I). Patients had no previous surgery to the breast and did not receive preoperative treatment. They underwent mastectomy, including axillary dissection. Data concerning age of the patient, size of the tumor, histological malignancy grade, axillary nodal status, and ER and PR status are given in Table I. The Her-2neu status was also recorded but is not included in Table I. Paraffin-embedded tissue blocks from additional 57 high risk invasive breast cancer patients enrolled in the DCTB program were used for validation purposes. These included 50 invasive ductal carcinomas, six lobular carcinomas, and one mucinous lesion. Archival paraffin-embedded tissue samples from six patients diagnosed with pure apocrine carcinoma were obtained from the Rigshospitalet tissue bank. The project was approved by the Scientific and Ethical Committee of the Copenhagen and Frederiksberg Municipalities (KF 01-069/03).
Sample Collection and Handling
Tissue specimens were collected from the Pathology Department at Rigshospitalet within 20–30 min after surgery and were rapidly transported on ice (average transport time was 15–20 min) to the Institute of Cancer Biology for further processing. Nine of the patients (patient numbers 65, 68, 74, 79, 80, 81, 86, 90, and 94; Table I) presented macrocysts ranging in size from a few millimeters to about 0.3 cm (Fig. 1). All nine patients contained more than one macrocyst. Non-malignant and tumor tissue was collected from all 36 patients, and a fraction of each sample was stored at –80 °C for gel analysis. A similar tissue block was paraffin-embedded for immunohistochemistry (IHC) analysis.
Preparation of Tissue and Cyst Fluid Samples for Two-dimensional Gel Electrophoresis
Cysts were cleaned of surrounding tissue, and the cyst fluid was aspirated using an elongated Pasteur pipette (Fig. 1). Because the volume collected was very small (5–50 µl), the samples were not centrifuged. Empty macrocysts were washed with PBS to minimize traces of cyst fluid and were stored at –80 °C. Twenty to 30, 6-µm cryostat sections of frozen macrocyst tissue (Fig. 1) were resuspended in 0.1 ml of CBL-1 lysis solution (Zeptosens AG, Witterswil, Switzerland) and were kept at –20 °C until used (56). The first and last sections of each sample were used for immunofluorescence analysis using keratin 19 antibodies as this epithelial marker is ubiquitously expressed by mammary epithelial cells (Fig. 1) (65). The availability of these pictures greatly facilitated the interpretation of the two-dimensional (2D) PAGE studies as it gave a rough estimate of the ratio of glands/cyst/tumor to stromal tissue. The same procedure was applied to both non-malignant and tumor tissue. 2D gels were analyzed using the PDQUEST software from Bio-Rad.
Cyst fluid samples were freeze-dried and resuspended in lysis solution prior to electrophoresis. A few milliliters of breast gross cystic disease fluid were also collected from four patients bearing palpable cysts but deemed free of breast cancer. In the latter set of cases, the samples were centrifuged at 5000 rpm for 20 min at 4 °C, and the supernatant was kept at –80 °C until used.
Two-dimensional Gel Electrophoresis
Tissue samples and freeze-dried fluids resuspended in lysis solution were subjected to IEF 2D PAGE as described previously (66, 67). Between 5 and 40 µl of sample were applied to the first dimension, and at least three IEF gels were run for each sample. Proteins were visualized using a silver staining procedure compatible with mass spectrometry analysis (68).
Protein Identification by Mass Spectrometry
In-gel Digestion Protocol—
Protein spots were excised from the dry gels followed by rehydration of gel plugs in water for 30 min at room temperature. The gel pieces were detached from the cellophane film, rinsed twice with water, and cut into about 1-mm2 pieces with subsequent additional washes. Proteins were "in-gel" digested with bovine trypsin (unmodified, sequencing grade; Roche Diagnostics) for 8 h as described by Shevchenko et al. (69). The reaction was stopped by adding TFA (up to 0.4%) followed by shaking for 20 min at room temperature to increase peptide recovery. In most cases, peptides were analyzed using the supernatant. In the few cases where the amount of peptides was too low or when no conclusive identification was achieved by peptide fingerprinting, the remaining amount of supernatant (approximately 10 µl) as well as the peptides additionally extracted from the gel pieces with 1% TFA and 50% ACN were concentrated on micro-ZipTipµ-C18 columns in accordance with the manufacturer’s instructions (Millipore).
Probe Preparation and Acquisition of the MALDI-TOF Spectra—
Samples were prepared for analysis by applying 2 µl of digested supernatant or microcolumn-eluted material on the surface of a 400/384 AnchorChip target (Bruker Daltonics, GmbH) followed by co-crystallization with 0.3 µl of -cyano matrix. After drying, the droplets were washed twice with 0.5% TFA to remove contamination from the samples.
Mass spectrometry was performed using a Reflex IV MALDI-TOF mass spectrometer equipped with a Scout 384 ion source. All spectra were obtained in positive reflector mode with delayed extraction using an accelerating voltage of 28 kV. Each spectrum represented an average of 100–200 laser shots, depending on the signal-to-noise ratio. The resulting mass spectra were internally calibrated using the autodigested tryptic mass values (805.417/906.505/1153.574/1433.721/2163.057/2273.160) visible in all spectra. Calibrated spectra were processed by the Xmass 5.1.1 and BioTools 2.1 software packages (Bruker Daltonics, GmbH). All spectra were analyzed manually as described previously (57). No restriction on the protein molecular mass and taxonomy has been applied. A number of fixed modifications (acrylamide-modified cysteine, i.e. propionamide/carbamidomethylation) as well as variable ones (methionine oxidation and protein N terminus acetylation) were included in the search parameters. The peptide tolerance did not exceed 50 ppm, and as a maximum only one trypsin missed cleavage was allowed. The protein identifications were considered to be confident when the protein score of the hit exceeded the threshold significance score of 70 (p < 0.05) and no fewer than seven peptides were recognized. The database was checked for redundancy, and whenever possible Swiss-Prot accession numbers were assigned. The second search was performed for all identifications as follows. 1) The predicted peptide digest was compared with the experimental one to reveal additional peptides present within the spectra. 2) The unmatched molecular weight values from the initial search were applied for extra search with the same reproducibility requirements for identification of the second and the third component in the spot. Whenever the protein score hit was close to the threshold significance score of 70, the PSD was performed as an additional means to confirm the identity of the proteins identified by post-translational modification. The following PSD search parameters were used: peptide tolerance of 50 ppm and MS/MS tolerance of 1 Da without any restriction on the protein molecular mass and taxonomy. Because the amount of peptides extracted from the silver-stained gels did not yield overall peak intensities high enough to allow multiple peptide sequencing (prerequisite for conclusive PSD analysis), the identification of proteins was never made solely based on PSD analysis. The molecular weight and pI of the identified proteins were evaluated by analysis of mobility of the corresponding protein band in the 2D gel images. Positive protein identification was achieved in 80% of the cases with average sequence coverage in the range of 8–80% depending on the protein size and spot intensity.
Antibody Arrays
Detection of multiple cytokines present in the cyst fluid of one breast cancer patient and breast gross cystic disease fluid from four patients, free of cancer, attending the mammography clinic was done using array-based technology. For this purpose RayBioTM Cytokine Antibody Arrays 6.1 and 7.1 were purchased from RayBiotech, Inc. Arrays were incubated with either 25 µl (breast cancer patient) or 250 µl (mammography clinic patients) of cyst fluid at 4 °C overnight, and bound cytokines were detected according to the manufacturer’s instructions. The sensitivity of the cytokine antibody array ranges from 1–2000 pg/ml.
Antibodies
Anti-peptide antibodies against cathepsin D and apoJ were prepared by Eurogentec (Seraing, Belgium). Polyclonal rabbit antibodies against human 15-hydroxyprostaglandin dehydrogenase (15-PGDH) recovered from 2D gels were prepared as described previously (70). In addition, a commercially available rabbit polyclonal antibody against 15-PGDH (Cayman Chemicals, Ann Arbor, MI) was used. Monoclonal antibodies specific for psoriasin were described previously (71). Antibodies against 3-hydroxymethylglutaryl (HMG)-CoA reductase (CRL 18811) isolated from the A9 hybridoma cell line (ATCC, Manassas, VA) and the nuclear antigen RN3 were kindly provided by Richard J. Cenedella (Kirksville, MO) and Frans Ramaekers (Nijmegen, The Netherlands), respectively. Antibodies against cyclooxygenase 2 (COX-2), Ki67, ER-, PR, ErbB-2, androgen receptor (AR), and p53 were purchased from Dako Cytomation (Glostrup, Denmark). Antibodies against keratin 19 were from NeoMarker (Fremont, CA). The antibodies against apoD, GCDFP-15, and active caspase-3 were from Zymed Laboratories Inc. (San Francisco, CA), Oncogene Research Products (San Diego, CA), and Promega (Madison, WI), respectively. The specificity of the 15-PGDH, cathepsin D, apoJ, psoriasin, and keratin 19 antibodies was confirmed by 2D gel immunoblotting (72).
Immunohistochemistry
Following surgery, fresh tumor blocks were immediately placed in formalin fixative and paraffin-embedded for archival use. Four-micrometer sections were cut from the tissue blocks and mounted on Super Frost Plus slides (Menzel-Gläser, Braunschweig, Germany), baked at 60 °C for 60 min, deparaffinized, and rehydrated through graded alcohol rinses. Nonspecific staining of slides was blocked (10% fetal calf serum in PBS buffer) for 15 min, and endogenous peroxidase activity was quenched using 1% H2O2 in 99% ethanol for 30 min. Heat-induced antigen retrieval was performed by immersing slides in Tris-EDTA buffer (pH 9.0) and microwaving in a 750-watt microwave oven for 10 min. The slides were then cooled at room temperature for 20 min and rinsed abundantly in tap water. Antigen was detected with a relevant primary antibody followed by a species-matched suitable secondary antibody conjugated to a peroxidase complex (EnVision+, Dako Cytomation). Finally color development was done with enhanced 3,3'-diaminobenzidine (DAB+, Dako Cytomation) as a chromogen to detect bound antibody complex. Slides were counterstained with hematoxylin. Standardization of the dilution, incubation, and development times appropriate for each antibody allowed an accurate comparison of expression levels in all cases. Sections were imaged using either a standard bright field microscope (Leica DMRB) equipped with a high resolution digital camera (Leica DC500) or a motorized digital microscope (Leica DM6000B) controlled by Objective Imaging’s Surveyor software (Objective Imaging Ltd.) for automated scanning and imaging that enables tiled mosaic image acquisition (see Fig. 11). Original magnification for all images is 200x.
|
RESULTS
|
|---|
Characterization of the Proteome Profile of Apocrine Macrocysts—
Macrocysts, devoid of cyst fluid, were frozen in liquid nitrogen and sectioned serially into 30 6-µm-thick slices as described under "Experimental Procedures" and illustrated in Fig. 1. The first and last of the sections were kept for immunofluorescence analysis using keratin 19 antibodies to assess the type of cyst as well as the ratio of epithelial cells to stroma in the specimen. The remaining slices were dissolved in lysis solution for 2D PAGE analysis (56, 57).
All macrocysts analyzed were of apocrine nature (type I), but only five specimens (two from patient 81, one from patient 86, and two from patient 94) exhibited abundant apocrine epithelium relative to stroma, yielding distinct protein patterns as judged by 2D PAGE analysis. The remaining cyst specimens contained too much connective tissue, and consequently their protein expression patterns were not informative. The expression profiles of the cysts from patients 81 and 94 were nearly identical as determined by PDQUEST analysis of their 2D gel patterns, whereas that of patient 86 exhibited a slightly different ratio of some proteins due to relatively higher levels of surrounding stroma. Fig. 2 shows representative silver-stained 2D gel of acidic (IEF) proteins of apocrine macrocysts from patients 81 (Fig. 2A) and 94 (Fig. 2B), both of which were located distant to the tumor in the mastectomized breast. Because most of the cells in apocrine macrocyst 81 corresponded to apocrine epithelium (see immunofluorescence picture in Fig. 1) and given the abundance of available tissue, which allowed us to perform both 2D PAGE and IHC analysis, we selected its protein expression profile as a reference map to single out markers of breast apocrine epithelia as well as for creating a database of apocrine cyst proteins identified by mass spectrometry (Table II). A total of 75 primary translation products, several of which were highly expressed in the apocrine cysts as compared with the corresponding non-malignant and tumor samples, are listed in alphabetical order in Table II together with the accession number, apparent molecular weight and pI, and gene map locus as well as molecular functions and biological processes according to the Human Protein Resource Database (www.hprd.org). Fifty-one of the identified proteins are known to be regulated by hormones, particularly androgens (36 proteins), and eight of these are known to interact directly with AR (Table II). Interestingly more than 50% of the identified proteins play a role in energy metabolism, particularly lipid biosynthesis, and about 45% are secreted or participate in transport processes. An annotated 2D gel image of all the proteins identified will be made available through our website (proteomics.cancer.dk).
View larger version (74K):
[in this window]
[in a new window]
|
FIG. 2. IEF 2D PAGE of whole protein extracts from breast apocrine macrocysts. A, apocrine macrocyst 81. B, apocrine macrocyst 94. Arrows indicate some proteins that are preferentially expressed by the apocrine cysts as compared with non-malignant and tumor tissues (see also Fig. 6). The position of -enolase is indicated in red as a reference as this protein and its modified forms migrate very close to HMG-CoA synthase.
|
|
As expected, the apocrine cysts expressed major cyst fluid proteins like GCFDP-15, apoD (37–40), and zinc-2-glycoprotein (35, 36), a fact that was confirmed by 2D PAGE analysis of cyst fluid 81 (Figs. 3A) as well as of the nine other cyst fluids investigated (a few examples are presented in Fig. 3, B and C). Loading of the gels with higher amounts of cyst fluid proteins allowed the additional identification of lower abundance proteins such as 15-PGDH, HMG-CoA synthase, cathepsin D, vitamin D-binding protein, hsp60, and a few others (Fig. 3D).
View larger version (95K):
[in this window]
[in a new window]
|
FIG. 3. IEF 2D PAGE of apocrine cyst fluid proteins. A, cyst fluid 81. B, cyst fluid 80. C, cyst fluid 65. D, higher load of proteins from cyst fluid 81. Major proteins are indicated for reference.
|
|
A number of studies have indicated that cytokines as well as growth factors are produced by cysts cells (24–28). These observations as well as compelling evidence suggesting a role in mammary epithelia development (73) prompted us to perform a cytokine and growth factor analysis of the cyst fluid. Although the amount of fluid collected from the apocrine macrocysts was generally too small to warrant reliable analysis using antibody arrays (56, 57), we were able to obtain a single cytokine pattern from a small sample collected from patient 94, revealing the presence of a number of cytokines and growth factors in the cyst fluid, including IL-1, IL-10, IL-12 p40, NAP-2, macrophage colony-stimulating factor, epidermal growth factor, basic fibroblast growth factor, RANTES (regulated on activation normal T-cell expressed and secreted), angiogenin, TNF-, eotaxin-3, angiopoietin, betacellulin, Fas, growth-related oncogene, IGFBP-3, macrophage migration inhibitory factor, MIP-1, TRAIL-R4, TIMP-1, TIMP-2, vascular endothelial growth factor, and vascular endothelial growth factor-D (Fig. 4). To substantiate this result we analyzed four additional fluid samples collected from large apocrine cysts of women free of breast cancer obtaining nearly identical cytokine patterns, thus corroborating the presence of many cytokines and growth factors in the cyst fluid of which only a few have been reported previously in the literature (6, 27, 28, 74–76). In addition, the 2D PAGE profiles of the four fluids were indistinguishable from those obtained with the cyst fluids of breast cancer patients (results not shown).
View larger version (56K):
[in this window]
[in a new window]
|
FIG. 4. Cytokine profiling of apocrine cyst fluids. Cyst fluid was collected from a cyst of one breast cancer patient and from large apocrine cysts of four patients, free of cancer, attending the mammography clinic. Cytokine-specific antibody arrays (RayBio Human Cytokine Array Series 1000, RayBiotech, Inc.) were incubated with cyst fluid, and detection of bound cytokines was performed according to the manufacturer’s instructions. Shown are the results obtained with the cyst fluid of the breast cancer patient (cyst 1) and one of the four cyst fluids collected from mammography clinic patients (cyst 2). The apocrine nature of the cells composing the cysts was confirmed by cytology (exemplified by hematoxylin staining of cyst 2, shown in rightmost panel). POS, positive; NEG, negative; PDGF, platelet-derived growth factor; RANTES, regulated on activation normal T-cell expressed and secreted; MCP, monocyte chemotactic protein; CNTF, ciliary neurotrophic factor; BMP, bone morphogenic protein; FGF, fibroblast growth factor; SDF, stromal cell-derived factor; M-CSF, macrophage colony-stimulating factor; GM-CSF, granulocyte-macrophage colony-stimulating factor; BDNF, brain-derived neurotrophic factor; BLC, B lymphocyte chemoattractant; SCF, stem cell factor; MDC, macrophage-derived chemokine; MIG,  interferon-induced monokine; MIP, macrophage inflammatory protein; IGFBP, insulin-like growth factor-binding protein; NT, neurotrophin; TARC, thymus and activation-regulated chemokine; PARC, pulmonary and activation-regulated chemokine; IFN, interferon; CTACK, cutaneous T-cell-attracting chemokine; ICAM, intercellular adhesion molecule; I-TAC, interferon--inducible T-cell chemoattractant; TECK, thymus-expressed chemokine; EGF-R, epidermal growth factor receptor; TRAIL, tumor necrosis factor-related apoptosis-inducing ligand; VEGF, vascular endothelial growth factor; GITR, glucocorticoid-induced TNF receptor; HCC, human CC chemokine; PIGF, phosphatidylinositol-glycan biosynthesis, class F protein; bFGF, basic fibroblast growth factor; uPAR, urokinase plasminogen activator surface receptor precursor; R, receptor; sTNF, soluble TNF; HGF, hepatocyte growth factor; MSP, macrophage-stimulating protein; GRO, growth-related oncogene; ENA-78, epithelial neutrophil-activating protein 78.
|
|
As a whole, our results point toward the apocrine cyst as being a gland capable of secreting a wide variety of macromolecules that may play key roles in the mechanisms underlying cyst evolution. Also the presence of inflammatory cytokines in the fluid may explain why rupture of the cysts can elicit inflammation. As expected from published data, apocrine cells in all type I macrocysts analyzed were ER--negative (Fig. 5A), PR-negative (not shown), AR-positive (Fig. 5B) (11–13), and ErbB-2-positive (Fig. 5C) (17) and exhibited a low proliferation index as judged by staining with the Ki67 antibody (Fig. 5D) (77).
View larger version (137K):
[in this window]
[in a new window]
|
FIG. 5. IHC pictures of apocrine microcyst 65 immunostained with antibodies against ER- (A), AR (B), ErbB-2 (C), and Ki67 (D), respectively.
|
|
Identification of Markers of Apocrine Metaplasia—
Comparison of the protein profiles of apocrine macrocysts (Fig. 2) with those of tumor and non-malignant tissues, exemplified in Fig. 6, A–F, revealed a few major cyst proteins that were either absent or present at much lower levels (factor of 4 or more) in non-malignant (Fig. 6, A, C, and E) and tumor (Fig. 6, B, D, and F) tissues. In addition, our studies identified several proteins that were preferentially expressed by malignant cells in some of the tumors. One of these, psoriasin (62), is described in more detail below in the context of apocrine tumors.
View larger version (120K):
[in this window]
[in a new window]
|
FIG. 6. Silver-stained IEF 2D gels of whole protein extracts from non-malignant and tumor tissues. A, non-malignant tissue 81. B, tumor 81. C, non-malignant tissue 64. D, tumor 64. E, non-malignant tissue 67. F, tumor 67. Arrows indicate some proteins, or their positions, that are preferentially expressed by the apocrine macrocysts.
|
|
Putative apocrine cell biomarkers included 15-PGDH, HMG-CoA synthase, cathepsin D, ß-glucuronidase, GCFDP-15, apoJ, and apoD (Table II). Antibodies against 15-PGDH, cathepsin D, GCFDP-15, apoJ, and apoD were used to confirm the 2D gel results by IHC using paraffin sections from cyst 81 as well as non-malignant and tumor tissue obtained from the same patient. Fig. 7 shows representative IHC pictures of tissue sections from cyst, non-malignant tissue, and tumor 81 immunostained with 15-PGDH (Fig. 7, A–C), GCDFP-15 (Fig. 7, D–F), cathepsin D (Fig. 7, G–I), apoD (Fig. 7, J–L), and apoJ (Fig. 7, M–O) antibodies, respectively. The results showed a clear correlation between the gel-derived data and the expression levels observed by IHC. GCDFP-15, cathepsin D, apoD, and apoJ were expressed predominantly, but not solely, by the apocrine epithelia as variable levels of staining were also observed in some cases in the non-malignant glands (Fig. 7H), the surrounding stroma (Fig. 7E), and the tumor (Fig. 7, F and I). IHC analysis of all available macrocysts indicated that only 15-PGDH was expressed exclusively by the apocrine epithelial cells, a fact that was also corroborated by 2D PAGE analysis, underscoring its potential value as a marker for these lesions. A 2D gel immunoblot (non-equilibrium pH gradient electrophoresis and IEF) showing the antigen specificity of the 15-PGDH antibody is presented in Fig. 8. In addition, matching of IHC results for a given tissue specimen with the corresponding 2D PAGE gel pattern provided cross-validation of antigen specificity for IHC.
View this table:
[in this window]
[in a new window]
|
TABLE III Immunohistochemical detection of markers in pure apocrine carcinomas
Data were determined by IHC analysis as described under "Experimental Procedures." Neg, negative; Pos, positive.
|
|
View larger version (103K):
[in this window]
[in a new window]
|
FIG. 7. IHC pictures of apocrine macrocyst (A, D, G, J, and M), non-malignant tissue (B, E, H, K, and N), and tumor (C, F, I, L, and O) from patient 81 immunostained with antibodies against proteins preferentially expressed by apocrine macrocysts. A–C, 15-PGDH; D–F, GCDFP-15; G–I, cathepsin D; J–L, apoD; and M–O, apoJ.
|
|
View larger version (55K):
[in this window]
[in a new window]
|
FIG. 8. Specificity of the rabbit anti-15-PGDH polyclonal antibody as determined by 2D gel immunoblotting (non-equilibrium pH gradient electrophoresis and IEF) of whole protein extracts from a transitional cell carcinoma.
|
|
Validation of the 15-PGDH Biomarker in Apocrine Microcysts Present in a Large Sample Set of Non-malignant Breast Tissue—
To further validate 15-PGDH as a marker of apocrine cysts and apocrine metaplasia, we analyzed by IHC non-malignant tissue from all the patients enrolled in the study (Table I) as well as retrospective samples from 57 additional high risk breast cancer patients (see "Experimental Procedures"). As shown in Fig. 9A, which is representative of all the sections analyzed, the 15-PGDH antibody specifically stained apocrine microcysts. These were found in 11 of the prospective samples (patients 65, 68, 71, 74, 79, 80, 81, 82, 86, 90, and 94), and in nine of the retrospective ones. The antibody detected not only obvious apocrine microcysts, easily identified by plain histology, but also pinpointed areas within some microcysts where only a few cells showed apocrine differentiation (see Fig. 12B). No reactivity was observed with terminal ductal lobular units (TDLU, Fig. 9B), type II flat microcysts (Fig. 9C), stroma cells (Fig. 9, A–C), or fat tissue (Fig. 9D), validating 15-PGDH as a specific marker for detection of apocrine metaplasia in the breast. Essentially the same results were obtained using a commercially available 15-PGDH polyclonal antibody.
View larger version (102K):
[in this window]
[in a new window]
|
FIG. 9. IHC staining of apocrine lesions immunostained with the 15-PGDH antibody. A, apocrine macrocyst in patient 81. B, TDLU. C, type II flat microcyst. D, fat tissue.
|
|
View larger version (41K):
[in this window]
[in a new window]
|
FIG. 12. Apocrine metaplastic cells immunostained with antibodies against 15-PGDH. A, duct apocrine metaplasia in patient 81. B and C, intraductal apocrine papillary lesions in patient 81.
|
|
Identification of an Additional Biomarker: HMG-CoA Reductase—
Although we did not have access to appropriate antibodies against the putative apocrine marker HMG-CoA synthase (Fig. 2) (Ref. 78 and references therein), it was apparent from data in the literature that this enzyme is co-regulated with HMG-CoA reductase (79, 80), a key enzyme in the biosynthesis of mevalonate. Additionally Farmer et al. (81) using microarrays recently identified a new group of breast tumors, "molecular apocrine," that expressed genes associated with androgen signaling and lipid metabolism, including HMG-CoA reductase. Encouraged by these data we obtained antibodies against HMG-CoA reductase and performed a comprehensive IHC analysis on all the tissue specimens available to us. Indeed the antibody specifically detected apocrine breast metaplasias (Fig. 10A) just like the 15-PGDH probe. The staining showed a punctuated pattern in line with the peroxisomal localization of the enzyme where mevalonate is further converted to farnesyl diphosphate (78). No reactivity was observed with terminal ductal lobular units (Fig. 10B), type II microcysts (Fig. 10A), stroma cells (Fig. 10, A and B), or fat tissue (results not shown), suggesting that HMG-CoA reductase is also a bona fide biomarker of breast apocrine metaplasia. From the above results it is likely that HMG-CoA synthase may also represent a candidate marker of apocrine metaplasia. Further studies, however, are needed to confirm this possibility. Having successfully identified specific apocrine biomarkers we were then able to address some of the more unclear and contentious points in the biology of apocrine cysts, namely their origin and mechanisms underlying their development as well as their relationship, if existent, with the cancer phenotype.
View larger version (55K):
[in this window]
[in a new window]
|
FIG. 10. IHC staining of apocrine lesions immunostained with the HMG-CoA reductase antibody. A, apocrine macrocyst in patient 81. B, TDLU.
|
|
Origin and Enlargement of Apocrine Microcysts—
It has been suggested that apocrine metaplasia of the epithelium lining the ductules may involve type II microcystic structures containing contiguous cells that stain strongly with ER- and PR antibodies (9, 53) (Fig. 11A). Because apocrine cells are ER-- and PR-negative and given the fact that we observed cystlike structures in which only a few cells expressed 15-PDGH (results not shown), we surmised that some type I cysts may be derived from type II lesions by mechanisms that could involve loss of receptor expression. Accordingly we screened the non-malignant tissue from 20 prospective and retrospective patients harboring apocrine microcysts for the presence of type II microcysts that have partially lost receptor expression as depicted in Fig. 11B. Subsequently we used the immunowalking approach (82), i.e. staining of consecutive sections with antibodies against 15-PDGH, PR, and ER-, in an effort to determine their relative expression in various areas of the same cyst. Although these structures were rare, we did identify a few, and a representative example from patient 45 (retrospective samples) is shown in Fig. 11, C–E. The relatively large type II microcyst depicted in Fig. 11C exhibits both PR-positive (lower half of the microcyst, open arrow) and -negative areas (upper half of the microcyst, closed arrow) that stained differentially with the 15-PGDH antibody (Fig. 11D). As expected, areas negative for PR stain positively for 15-PGDH (Fig. 11D, upper half of the microcyst), implying that this area of the type II lesion is undergoing apocrine metaplasia. Staining of the next section with the ER- antibody showed a good correlation with the PR staining (not shown). Interestingly, however, the small 15-PGDH-positive apocrine microcyst observed in Fig. 11, C and D, upper right corners (enclosed by a rectangle), showed groups of cells that expressed ER- (Fig. 11E, dotted boxes) but that were negative for PR (Fig. 11C, dotted boxes). From these data it seems likely that loss of PR expression precedes that of ER- receptor during apocrine microcyst development.
Data in the literature indicate that apocrine metaplasias may also arise from terminal ducts and intraductal papillary lesions. Indeed both the 15-PGDH and HMG-CoA reductase biomarkers identified these lesions (illustrated in Fig. 12, A–C, with the 15-PGDH antibody). In fact, some of the apocrine intraductal papillary lesions could not be recognized without the aid of the biomarker antibody probes (Fig. 12C).
Visual analysis of sections from macrocysts often showed apocrine intraductal papillary structures that stained positively with the 15-PGDH antibody and that protruded into the cyst cavity from opposite sides (Fig. 13A), suggesting that microcysts may combine or merge to generate larger cysts. Indeed careful IHC analysis of several areas of macrocyst 81, which was composed of several microcysts (Fig. 13A), revealed regions where apocrine cells from two adjacent microcysts had become very close to each other, separated only by a thin layer of stroma (Fig. 13B). Cells in the middle of the bridges showed loss of nuclear content as judged by hematoxylin counterstaining (Fig. 13B, arrow), a fact that was corroborated in different preparations by using a variety of antibodies against nuclear proteins (Fig. 13C). Loss of nuclear content is most likely not due to apoptosis because these cells did not stain with an antibody specific for active caspase-3 (data not shown) or show any of the characteristic alterations in cell morphology associated with programmed cell death. The data support a model for growth of apocrine cysts where coalescence of two microcysts with subsequent rupture of the connecting bridge(s), either by fluid pressure alone or in combination with peptidases present in the fluid (33), leads to cyst enlargement. Apocrine cells devoid of nuclear staining were also observed in apocrine cysts lined by a single layer of cuboidal cells (not shown), but the mechanism underlying this phenomenon remains unknown.
View larger version (130K):
[in this window]
[in a new window]
|
FIG. 13. Cyst structures that may be involved in microcyst enlargement. A, apocrine macrocyst 81 sections immunostained with antibodies against 15-PGDH. B and C, as above but stained with antibodies against COX-2 (B) and a nuclear antigen (RN3) (C), respectively. Apocrine cells devoid of nuclear content are indicated with arrows in B and C.
|
|
Apocrine Biomarkers and the Cancer Phenotype—
An important question that has been debated for some time concerns the relationship between apocrine changes and breast carcinoma (10, 17, 52, 53). Published studies have shown contradictory results as different authors have described apocrine lesions as either representing a risk factor, non-obligate precursor lesions of apocrine carcinoma, or benign lesions with no correlation with malignancy (10, 49–54, 60).
To shed some light into the problem and taking advantage of the fact that we had collected non-malignant and tumor tissue from the same patients, we analyzed by IHC the 93 sets of matched samples available to us using the 15-PGDH and HMG-CoA reductase antibodies with the aim of assessing the relationship, if any, between benign apocrine changes and breast carcinoma. Of all the lesions analyzed, including the retrospective samples, which consisted of 75 ductal carcinomas, 15 lobular carcinomas, one apocrine carcinoma, one mucinous carcinoma, and one tubular lesion, only tumor cells in patient 79, which presented an invasive apocrine carcinoma (Table I), expressed one of the apocrine markers, namely 15-PGDH. Expression of both markers, however, was detected in intraductal papillomas showing apocrine atypia that were observed in or very close to the invasive area of the ductal carcinoma in patient 71 (Table I). These results are described in detail below.
Tumor 79—
Tumor 79 expressed 15-PGDH (Fig. 14A), albeit at much lower levels as compared with the reference apocrine microcysts (Fig. 7A; see also Figs. 2 and 15A), but did not express detectable levels of HMG-CoA reductase (Fig. 14B). The lesion was p53-, ER-, and PR-negative (Table I), and about 50% of the tumor cells stained with the AR antibody (Fig. 14C) as expected for an apocrine carcinoma (81, 83, 84). Tumor 79 expressed a number of proteins not present in apocrine cysts or normal gland cells as judged by 2D PAGE (data not shown), and of these, psoriasin (62) was the most striking because this protein has been shown to be highly up-regulated in ductal carcinoma in situ (CIS) and associated with a worse prognosis in estrogen receptor-negative invasive ductal carcinomas (85, 86). Expression of psoriasin by tumor 79 was confirmed by IHC using a monoclonal antibody specific for this protein (Fig. 14D) (75).
View larger version (78K):
[in this window]
[in a new window]
|
FIG. 14. IHC pictures of invasive apocrine breast tumors. A–D, apocrine cancer 79 immunostained with antibodies against 15-PGDH (A), HMG-CoA reductase (B), AR (C), and psoriasin (D), respectively. E and G, apocrine carcinoma RH-21990 immunostained with 15-PGDH (E) and HMG-CoA reductase antibodies (G), respectively. F and H, CIS from carcinoma RH-21990 immunostained with 15-PGDH (F) and HMG-CoA reductase antibodies (H), respectively.
|
|
View larger version (73K):
[in this window]
[in a new window]
|
FIG. 15. Expression of 15-PGDH by apocrine tumors. Silver-stained IEF 2D gels of whole protein extracts from tumor 79 (A), tumor RH-9777 (B), tumor RH-21990 (C), and tumor RH-24678 (D) are shown.
|
|
Because the expression of 15-PGDH by tumor 79 suggested a link between apocrine metaplasia and the tumor phenotype, we analyzed six pure apocrine carcinomas by IHC for expression of the two apocrine-specific biomarkers, 15-PGDH and HMG-CoA reductase. Of these, five stained positively with the 15-PGDH antibody, which decorated both invasive and apocrine CIS cells in each one of the tumors (RH-34697, RH-21990, RH-9777, RH-10515, and RH-13319; Table III). These results were confirmed by 2D PAGE as shown in Fig. 15, B and C, which presents selected areas of silver-stained gels of whole protein extracts of apocrine tumors RH-9777 (Fig. 15B), RH-21990 (Fig. 15C), and RH-24678 (Fig. 15D), a lesion that did not stain with the 15-PGDH antibody (Table III). All five apocrine cancers were ER-- and PR-negative (Table III). Interestingly only one of the 15-PGDH-positive tumors (RH-34697) expressed HMG-CoA reductase both in the invasive and CIS apocrine cells, whereas another (RH-21990) expressed the enzyme solely in the CIS cells (Table III). Fig. 14 shows IHC pictures of apocrine tumor RH-21990, which stained positively with the 15-PGDH antibody (Fig. 14E) but was devoid of HMG-CoA reductase (Fig. 14G). Areas within the tumor exhibiting apocrine CIS cells that are thought to be a precursor of invasive disease stained positively with both antibodies (Fig. 14, F and H), suggesting that the expression of HMG-CoA reductase may be lost prior to 15-PGDH during cancer progression (see below). With one exception (RH-9777), all tumors expressing 15-PGDH also stained positively with the psoriasin antibody, including apocrine CIS cells (Table III), suggesting that this protein may be a valuable marker for late events in apocrine cancer progression, a process that may encompass benign metaplasia, hyperplasia, atypia, CIS, and invasive disease (54, 87, 88).
Tumor 71—
Immunohistochemical analysis of the tumor of patient 71, which was diagnosed as an invasive ductal carcinoma (Table I), and areas surrounding it revealed two types of intraductal papillary lesions with apocrine atypia that showed similar and yet distinct morphologies and phenotypes. Atypia a stained positively with the 15-PGDH (Fig. 16A) and HMG-CoA reductase (Fig. 16B) antibodies, exhibited some p53-positive cells (Fig. 16C), and expressed psoriasin (Fig. 16D). Atypia b, on the other hand, expressed similar levels of 15-PGDH (Fig. 16E), p53 (Fig. 16G), and psoriasin (Fig. 16H) but exhibited greatly reduced levels of HMG-CoA reductase (Fig. 16F). Because the analysis of the pure apocrine cancers suggested that the expression of HMG-CoA reductase may be lost prior to 15-PDGH reductase, we believe that there is a sequential relationship between the two atypias. The invasive component of the tumor was p53- and psoriasin-positive but did not show immunoreactivity with the 15-PGDH or the HMG-CoA reductase antibodies (results not shown). Whether tumor 71 was misclassified and corresponds to an apocrine lesion that has lost the expression of the two apocrine markers or corresponds to a mixed type is at present unknown, emphasizing the need to generate additional markers to cover all the intermediate stages involved in apocrine cancer progression.
View larger version (74K):
[in this window]
[in a new window]
|
FIG. 16. Apocrine atypias in patient 71. A–D, apocrine atypia a in patient 71 immunostained with antibodies against 15-PGDH (A), HMG-CoA reductase (B), p53 (C), and psoriasin (D). E–H, apocrine atypia b immunostained with antibodies against 15-PGDH (E), HMG-CoA reductase (F), p53 (G), and psoriasin (H), respectively.
|
|
|
DISCUSSION
|
|---|
Up to one-third of the women aged 30–50 years have cysts in their breasts (3), and both their frequency and their proposed association with an increased risk of breast cancer have underpinned the importance of investigating these lesions at the molecular level (53, 89). In particular, apocrine changes, which span from benign metaplasias to atypical apocrine hyperplasia (12, 53), have attracted much attention because apocrine epithelia may be a precursor in malignant transformation. Their relationship, however, with invasive breast cancer remains contentious despite much research effort (10, 49, 50, 52–54).
The available data indicate that most breast cysts arise from apocrine metaplastic lobules (3) and that type I apocrine cysts are more likely to progress to large lesions than type II flat microcysts (5). Currently, however, very little is known as to the molecular mechanisms underlying apocrine metaplasia, a process that entails the conversion of breast epithelial cells into sweat gland type cells (60). Breast apocrine cells are cytologically identical to cells in apocrine glands (60, 90) and exhibit rather large vesicular nuclei with prominent nucleoli and abundant eosinophilic cytoplasm that occasionally present apical snouts that slough off into the lumen. So far, no specific biomarker has been available to characterize breast apocrine lesions.
The study presented here was undertaken in an effort to dissect and gain a better understanding of the steps leading to breast apocrine metaplasia and ultimately intended to generate specific biomarkers that may enlighten its relationship with cancer phenotype. The strategy we applied was based on the comparison of the protein expression profiles of breast apocrine macrocysts with those of matched cancerous and non-malignant tissues from the same patient followed by biomarker identification, antibody preparation, and validation using a large sample size. Among all of the markers identified for which high quality antibodies could be obtained, only 15-PGDH and HMG-CoA reductase were expressed specifically by breast benign apocrine metaplasias arising from dilated type II microcysts, terminal ducts, and intraductal papillary lesions (Fig. 17).
View larger version (26K):
[in this window]
[in a new window]
|
FIG. 17. Expression of apocrine markers at various stages of apocrine carcinoma progression. Apocrine lesions span from apocrine metaplasia, apocrine atypia, CIS, and invasive apocrine carcinoma. The data concerning the apocrine CIS and the invasive apocrine carcinomas were derived from the IHC analysis of the six pure apocrine cancers. All other data were derived from the prospective and retrospective samples from the 93 high risk breast cancer patients. Markers in yellow are expressed already in benign metaplasias, whereas those indicated with red seem to be expressed at a later stage.
|
|
15-PGDH is a prostaglandin (PG)-degrading enzyme whose levels have been shown to be down-regulated in bladder (64), colon (91), and gastrointestinal (92, 93) cancers. Decreased expression of this protein, which has tumor suppressor activity in colorectal cancer (92, 93), is accompanied by increased expression of COX-2, a PG-synthesizing enzyme and a 15-PGDH physiological antagonist that has been implicated in tumor progression in various malignancies (94–99). Recently Muller-Decker et al. (100) reported that overexpression of COX-2 in keratin 5-positive cells of the mouse mammary gland caused an increase in PGE2, cystic duct dilations, adenosis, and fibrosis. Because stronger expression of COX-2 was observed in fibrocystic changes as compared with normal tissue, we used specific anti-COX-2 antibodies to validate the results in our collection of human breast tissue samples. Indeed IHC analysis of numerous macro- and microcysts showed that apocrine lesions (Fig. 18A; see also Fig. 17) but not type II flat microcysts (Fig. 18B) expressed this protein. The antibody decorated apocrine cell-cell contacts as well as the areas where these cells anchor to the basement membrane (Fig. 18A, arrow). IHC analysis of all the available samples showed that COX-2 was not expressed by stroma cells (Fig. 18B) and with very few exceptions by normal glands. Only in some cases was COX-2 expressed by non-apocrine tumors (results not shown), an observation that is in accordance with previously published data (98, 102). Hence breast apocrine metaplastic epithelia co-express two key antagonistic activities that modulate the expression of PGE2, a compound that has been involved in control of cell proliferation, migration, apoptosis, angiogenesis, and inflammation (103). The differential compartmentalization of COX-2 and 15-PGDH (compare Figs. 7A and 18A), however, suggests that their interplay with regard to PGE2 may be more complex than previously thought from the study of cell lines. In addition, our IHC studies of apocrine tumors failed to provide a clear correlation between the expression of the two proteins (Table III), suggesting that their interplay may be different in diverse pathological conditions.
HMG-CoA reductase is a key enzyme in cholesterol biosynthesis, and its product mevalonate has been shown to stimulate cell proliferation in breast cell lines (104, 105). Also there is evidence from in vivo and in vitro studies indicating that statins (105, 106), which are inhibitors of HMG-CoA reductase, may be effective in the treatment of malignancies by decreasing farnesylation and geranylgeranylation of key proteins such as the Ras and Rho family of GTPases, which are essential for cellular proliferation (107, 108). Our studies, however, showed abundant expression of both HMG-CoA reductase (Fig. 10A) and synthase (Fig. 2) in apocrine cyst cells that exhibit a very low proliferation index (Fig. 5D), implying that increased levels of these enzymes and perhaps increased levels of mevalonate may not be sufficient to stimulate cell growth. In fact, most of the invasive apocrine tumors analyzed were devoid of or showed very low levels of HMG-CoA reductase (Table III), a characteristic that may be distinctive to the progression of apocrine tumors.
IHC analysis of the 93 breast tumor samples using antibodies against the two specific apocrine biomarkers revealed that only one of the lesions, an invasive apocrine carcinoma in patient 79 (Table I), expressed 15-PGDH in the tumor cells. These results harmonize well with the fact that our set of tumors, which included 75 ductal carcinomas, 15 lobular carcinomas, one mucinous carcinoma, one tubular lesion, and one apocrine carcinoma, reflects the expected incidence of mammary tumor types prevalent in the general population (109, 110).
Expression of 15-PGDH by invasive apocrine cancer was further confirmed by IHC using archival paraffin-embedded samples from six patients diagnosed with pure apocrine carcinoma (Table III) as well as by 2D PAGE (Fig. 15). Five of these lesions expressed 15-PGDH both in the invasive and apocrine CIS cells in each tumor, providing for the first time a direct association between the phenotype of benign metaplasias and the apocrine cancer phenotype (Fig. 17). Only one of the 15-PGDH-positive pure carcinomas expressed HMG-CoA reductase both in the tumor and CIS cells, whereas another expressed the protein only in the apocrine CIS cells (Table III). These results imply that the expression of HMG-CoA reductase is lost earlier than 15-PGDH during apocrine cancer progression. This contention is further supported by the fact that HMG-CoA reductase-positive tumors constitute a subset of the 15-PGDH apocrine tumors (Table III) and by the analysis of two apocrine atypias, a and b, detected in patient 71 (Fig. 16), whose phenotypes were remarkably similar, although one of them, most likely the most advanced lesion (atypia b), expressed significantly lower levels of HMG-CoA reductase.
The IHC analysis of the pure apocrine carcinomas indicated that individual cancers may evolve through diverse molecular pathways because these lesions, CIS apocrine cells included, were either p53-positive or -negative and showed various combinations of phenotypes as far as 15-PGDH, HMG-CoA reductase, and COX-2 (Table III and Fig. 17) expression was concerned. Fig. 17 summarizes the protein expression phenotype of various lesions observed in this study that ranged from benign apocrine metaplasia to atypia, apocrine CIS, and invasive apocrine cancer. As a whole, our results indicate that most apocrine changes have little intrinsic malignant potential, although some lesions may progress to apocrine carcinoma. None of these lesions, however, seems to be a precursor of invasive ductal carcinomas, which accounted for 81% of the tumors analyzed in this study.
So far, there is no molecular marker that can be used to detect a precancerous state or identify which premalignant lesion(s) will develop into invasive breast cancer, but further proteomic analysis of apocrine cancers and benign metaplasias are expected to reveal additional biomarkers that may dissect all the intermediate stages involved in apocrine cancer progression. These studies, which require prospective samples, are currently underway in our laboratory. A corollary of this work is that the successful identification of specific apocrine biomarkers, which allowed the subsequent tackling of more unclear and contentious points in the biology of apocrine cysts, provides proof-of-concept for the validity of the strategic approach we took. In addition, our sample preparation methodology based on cryosectioning enabled us to overcome some of the limitations imposed by the availability of biological material inherent to certain lesions. Encouraged by these facts we are now applying the same experimental approach to the study of various biological problems relevant to the biology of breast cancer, most notably to investigate the progression of ductal carcinoma in situ, a type of premalignant mammary lesions, to invasive cancer.
Finally the identification of differentially expressed proteins that characterize specific steps in the progression from early benign lesions to apocrine cancer opens a window of opportunity for designing and testing new approaches for pharmacological intervention, not only in a therapeutic setting but also for chemoprevention, to inhibit cyst development. Enzymatic activities such as 15-PGDH, HMG-CoA reductase, and COX-2 are well known therapeutic targets (59, 103, 106, 111, 112) with pharmacological agents already available, and at least in the case of COX-2 one has reason to suspect a causal relationship to the development of fibrocystic changes (100). Statins, which are HMG-CoA reductase inhibitors, have well characterized efficacy and safety profiles (Ref. 106 and references therein), and it is therefore tempting to speculate that these drugs might be useful for preventing the development of breast cysts. Further studies into the functional role that the various biomarkers identified in this study play in cyst formation are warranted as they may i
TOP
ABSTRACT
EXPERIMENTAL PROCEDURES
