Transcriptional profiling has identified five breast cancer subtypes, of which the basal epithelial is most aggressive and correlates with poor prognosis. These tumors display a high degree of cellular heterogeneity and lack established molecular targets, such as estrogen receptor-α, progesterone receptor, and Her2 overexpression, indicating a need for definitive diagnostic markers. We present evidence that nestin, a previously described marker of regenerative cells in diverse tissues, is expressed in the regenerative compartment of the normal human mammary gland. Colocalization studies indicate two distinct populations of mammary epithelia that express nestin: one expressing cytokeratin 14 (CK14) and ΔN-p63 and another expressing desmin. Immunohistochemical analysis indicates that ΔN-p63 and nestin are coordinately expressed during pregnancy in the murine mammary gland. In the embryonal carcinoma cell line NT2/D1, ectopic ΔN-p63-α disrupts retinoic acid–induced differentiation, thereby preserving expression of nestin; however, small interfering RNA–mediated ablation of nestin is insufficient to promote differentiation, indicating that whereas nestin may identify cells within the regenerative compartment of the mammary gland, it is insufficient to block differentiation and preserve replicative capacity. Immunohistochemical analysis of basal epithelial breast tumors, including those shown to carry BRCA1 mutations, indicates robust expression of nestin and CK14, punctate expression of p63, and low to undetectable levels of desmin expression. Nestin was not detected in other breast cancer subtypes, indicating selectivity for basal epithelial breast tumors. These studies identify nestin as a selective marker of the basal breast cancer phenotype, which displays features of mammary progenitors. [Cancer Res 2007;67(2):501–10]

Transcriptional profiling of large cohorts of human breast cancers has resulted in the identification of five distinct subtypes (1). These studies indicate that between 17% and 37% of human breast tumors display a basal epithelial phenotype (24). These tumors are highly aggressive, are poorly differentiated, and lack molecular targets for endocrine or anti-Her2 therapy. The basal phenotype is associated with an early age of onset and short times to relapse and disease progression (5). A disproportionately high number of these tumors are detected during normal screening mammography intervals (6), reflecting the aggressive nature of these tumors. For these reasons, the basal epithelial subtype contributes disproportionately to breast cancer mortality (7). Improved understanding of the etiology of these tumors may help to identify selective markers and therapeutic targets that will improve detection, diagnosis, and treatment of the basal epithelial breast cancer subtype.

Throughout reproductive life, the epithelial portion of the mammary gland undergoes multiple successive regenerative cycles characterized by cellular proliferation and terminal differentiation (8). During pregnancy, expansion of the epithelial compartment is thought to be initiated by the mitotic division of the mammary stem cells, resulting in production of a pool of mammary progenitors. The progenitor population enters a highly proliferative state that drives this expansion and is followed by cellular differentiation. The regenerative cycle ends with extensive apoptosis and tissue remodeling during post-lactation involution. In non-pregnant females, a less-pronounced cycle of proliferation and cell death occurs with each menstrual cycle (9). Continuous regenerative cycles within the epithelial portion of the mammary gland depend upon a population of mammary stem cells that retain their proliferative capacity and resist terminal differentiation. These features confer a prolonged replicative life span, suggesting that mammary stem cells may be capable of accumulating and harboring mutations and propagating these mutations into the progenitor pool. This further suggests that mammary stem cells may be the sites of breast cancer initiation, which is supported by the finding that the mammary stem cell fraction is amplified in the MMTV-Wnt1A murine breast cancer model (10). This observation is consistent with other studies that indicate breast cancer initiation is a condition of poorly regulated self-renewal (11) and has focused attention upon the oncogenic potential of genetic pathways that regulate self-renewal.

To repopulate the mammary gland and to achieve cellular diversity during each regenerative cycle, while preserving self-renewing capacity for subsequent regenerative cycles, the mitotic offspring of mammary stem cells must execute the decision to preserve or forfeit self-renewing capacity. There is abundant evidence that the p53 family member TP63 plays a critical role in this decision (1215). The gene encoding TP63 uses proximal and distal promoters to produce trans-activating (TA-p63) and NH2-terminally deleted (ΔN-p63) isoforms (16). Other studies suggest that in the adult mammary gland, p63 is required for the preservation of self-renewal and cellular stasis (17). Mutations in TP63 have been shown to underlie a broad spectrum of syndromes that have in common defects in cellular stasis of a variety of epithelial and apocrine structures (14). These defects are believed to result from a genetic program of non-regenerative differentiation that ultimately leads to the depletion of the regenerative compartment. Specifically relevant to the establishment and preservation of the mammary regenerative compartment are mutations within the α-specific COOH terminus of TP63 that underlie limb-mammary syndrome (OMIM# 603543; ref. 18). Patients bearing these mutations display severe to complete hypoplasia of the mammary epithelia. Consistent with these phenotypes, targeted ablation of TP63 in the mouse resulted in hypoplasia of both embryonic and adult epithelial and apocrine structures (13, 15). These studies indicate that TP63 is required for the establishment and preservation of the regenerative compartments of multiple epithelial structures. Additionally, studies using isoform-specific antibodies indicate that ΔN-p63 predominates in the basal epithelia of the mammary gland (19), which is a component of the mammary regenerative compartment. This indicates that ΔN-p63 may play an important role in the maintenance and preservation of cellular stasis during the mammary regenerative cycle.

Nestin is an intermediate filament protein that is expressed in regenerative compartments of the central nervous system and other tissues (20). It has been shown to colocalize with p63, as detected by the pan-p63 monoclonal antibody 4A4, within a subset of the limbal epithelia of the cornea (21). This suggests that nestin may also be expressed in the regenerative compartment of epithelial structures and is consistent with multiple findings indicating that nestin is a marker of neural progenitors (22, 23). In this article, we report that nestin is expressed in two distinct cell types within the basal/myoepithelial layer of the mammary gland. Two-color immunofluorescence identifies a cell type in which nestin is coexpressed with cytokeratin 14 (CK14) and p63 and a second that is positive for desmin. Analysis of nestin and ΔN-p63 expression in the murine mammary gland during pregnancy indicates overlapping patterns of temporal and cell type–specific regulation. We provide evidence that ΔN-p63-α is sufficient to block cellular differentiation and preserve nestin expression in the embryonal carcinoma cell line NT2/D1, and that small interfering RNA (siRNA)–mediated repression of nestin in NT2/D1 was insufficient to promote differentiation. Taken together, these observations suggest that nestin is expressed in the regenerative compartment in the mammary gland. Immunohistochemical analysis indicates that nestin is robustly expressed in basal epithelial breast tumors [defined here as triple negative for the estrogen receptor-α (ERα), the progesterone receptor (PR), and Her2] and undetectable in breast tumors representing other molecular classifications. We also present data indicating that nestin is strongly expressed in BRCA1-associated breast tumors, which is consistent with the finding that BRCA1-associated tumors possess a basal phenotype (3, 24). Further analysis of the triple-negative tumors indicates high levels of CK14, variable expression of p63, and undetectable levels of desmin, suggesting that these tumors arose from components of the basal epithelia that express nestin, CK14, and p63. This signature was absent in other breast cancer subtypes, indicating that nestin expression identifies the poor prognosis basal epithelial subtype. These studies also indicate that the highly aggressive and poorly differentiated basal breast cancer subtype displays biochemical features of the regenerative compartment of the mammary gland.

Human tissue samples. Normal human mammary gland samples derived from reduction mammoplasty were identified from archived samples within the Tissue and Tumor Bank at Dartmouth Hitchcock Medical Center. For breast tumors representing diverse subtypes, the files of the Department of Pathology at Dartmouth Hitchcock Medical Center were reviewed to identify formalin-fixed, paraffin-embedded (FFPE) samples representing tumors that were either ER/PRHer2, ER/PR/Her2+ [by fluorescence in situ hybridization (FISH)], or ER+/PR+. Identified tumors were evaluated to ensure that sufficient tissue existed within the paraffin blocks. Identification and collection of tissues and tumors were conducted in strict adherence with regulations related to the protection of patient identity. BRCA1-associated tumors were identified and selected from archived material obtained through the Family Risk Assessment Program at the Fox Chase Cancer Center (FCCC). These tumor tissue samples were derived from patients that had undergone genetic testing through the Clinical Molecular Genetics Laboratory at FCCC and were found to be carriers of a deleterious BRCA1 mutation.

Murine tissue samples. C57/Black6 mice were mated at 8 weeks of age, and pregnancy was confirmed by the presence of the anogenital plug. Murine mammary glands were harvested by gross dissection at days 3, 7, 14, and 20 of pregnancy. Glands were fixed in Bouin's solution, dehydrated though a series of graded alcohols, and embedded in paraffin. Four- to 5-μm sections were prepared and fixed onto frosted glass microscope slides. Immunohistochemistry was done as described below.

Immunohistochemistry. Tumors representing the basal breast cancer subtype were identified by negativity for ERα, PR, and Her2. Additionally, paraffin blocks representing tumors of the Her2 subtype (ER/PR/Her2+ by FISH) and tumors with a luminal epithelial phenotype (ER+/PR+) were identified for comparative analysis. Additionally, FFPE samples of normal human mammary gland derived from reduction mammoplasty were used for analysis of marker expression in the normal mammary gland. Briefly, 4-μm-thick series sections were cut and applied to charged glass slides (Superfrost Plus). Sections were deparaffinized in xylene, rehydrated through a series of graded alcohol, placed in 10 mmol/L citrate buffer (pH 6), and underwent antigen retrieval in a microwave oven for 15 min. Endogenous peroxidase was blocked with 0.3% hydrogen peroxide in distilled water for 10 min. Samples stained for nestin, desmin, and p63 were blocked in 5% donkey serum in 0.1% Triton X-100 in PBS. Samples stained with CK5 and CK14 were similarly blocked with 5% horse serum. All blocking was done for 30 min at 37°C. Immunohistochemistry was done using an avidin-biotin peroxidase system. The following primary antibodies were incubated for 45 min at 37°C: nestin (1:50, clone C-20; Santa Cruz Biotech, Santa Cruz, CA) or nestin (1:50, clone 10C2; Santa Cruz Biotech), desmin (1:50, clone Y-20; Santa Cruz, Santa Cruz, CA), pan-p63 (1:100, clone 4A4; BD Biosciences, San Diego, CA), CK14 (1:100, clone LL002; Neomarkers, Fremont, CA), and CK5 (1:25, clone XM26; Neomarkers). Detection of ΔN-p63 isoforms was conducted using a rabbit polyclonal antibody raised during these studies and diluted 1:200. Following washes in PBST, a biotinylated secondary antibody (Vector Laboratories, Inc., Burlingame, CA) was applied for 30 min at 37°C. Detection of nestin and desmin staining was blocked in 5% donkey anti-goat IgG (1:200). Detection of pan-p63, CK5, and CK14 was blocked with horse anti-mouse IgG (1:400). Detection of the ΔN-p63–specific staining was achieved with a goat anti-rabbit IgG (1:400). Slides were incubated with streptavidin-linked peroxidase (1:400; Vector Laboratories) for 30 min at 37°C and developed with 3,3′-diaminobenzidine tetrahydrochloride staining kit (Vector Laboratories). Mayer's hematoxylin was applied as a counterstain.

Two-color immunofluorescence. The pretreatment and preparation of slides, including deparaffinization, rehydration, and antigen retrieval, were identical to the protocols for immunohistochemistry. Blocking serum was applied for removing nonspecific binding accordingly. Nestin or desmin costained with pan-p63, CK5, and CK14 were blocked with 5% donkey serum. Then, the samples were incubated with two different primary antibodies together (nestin or desmin + pan-p63, CK5, CK14). Following a series of wash steps in PBST, the slides were incubated with two different Alexa Fluor–conjugated secondary antibodies (1:200; Invitrogen, Carlsbad, CA) together. Nestin (C-20) or desmin costained with nestin (10C2), pan-p63 (4A4), CK5, and CK14 were blocked with donkey anti-goat-Alexa Fluor 594 IgG and donkey anti-mouse-Alexa Fluor 488. Slides were mounted with Vectorshield (Vector Laboratories) mounting solution with 4′,6-diamidino-2-phenylindole (DAPI) and imaged by fluorescence microscopy and confocal microscopy. Double staining of nestin plus desmin, CK14 + pan-p63 (4A4) is modified slightly due to identical host species of primary antibodies. The staining with primary antibodies was conducted sequentially. The first staining of the primary antibody was done as regular immunohistochemistry protocol. The only difference was to use Avidin-Alexa Fluor 488 (1:200; Invitrogen) to replace the streptavidin-peroxidase complex reagent followed by microwaving in 10 mmol/L citrate buffer (pH 6) for 15 min to remove nonspecific binding. Following treatment, slides were stained with the secondary primary antibody and detected with Alexa Fluor 594–conjugated IgG accordingly.

Cell culture. The embryonal carcinoma cell line NT2/D1 was cultured as previously described, and differentiation studies were conducted by treating cells with either 0.01% DMSO (vehicle) or 1 μmol/L retinoic acid (RA). Cells were transfected using Fugene (Roche, Indianapolis, IN) according to the manufacturer's protocol. Ectopic expression of ΔN-p63-α was achieved by adenoviral infection. siRNA-mediated repression of nestin was achieved by tranfection of a pool of four nestin-specific siRNAs (ONTARGETplus SMARTpool, Dharmacon, Lafayette CO). A similar pool of four nonspecific siRNAs was used as negative control (ONTARGETplus SMARTpool, Dharmacon). siRNAs were transfected in amounts recommended by the manufacturer. Transfections were done using Oligofectamine (Invitrogen) according to manufacturer's protocol.

Northern analysis. RNA was isolated using the RNeasy system (Qiagen, Valencia, CA). Seven to 10 μg RNA was loaded per lane, and Northern blotting was conducted as previously described. Nestin mRNA was detected using an ∼1-kb EcoRI fragment derived from an IMAGE clone # 5493839 containing nestin cDNA sequences. RA receptor β (RARβ) mRNA was detected using a 615-bp EcoRI fragment derived from the RARβ cDNA. Northern blotting procedures were as previously described (25).

A2B5 immunodetection. Following treatment with DMSO or RA, NT2/D1 cells were washed twice in cold PBS and incubated with an undiluted cell culture supernatant from a hybridoma that produces an anti-A2B5 immunoglobulin. Incubation was carried out for 30 min at 37°C, and samples were washed twice in PBS for 10 min each. Following washing, a 1:200 dilution of a goat anti-mouse-IgG linked to Alexa Fluor 488 was incubated with the samples for 15 min at 37°C. Samples were washed, and nuclei were stained with DAPI before fluorescence imaging.

Western blotting. Western analysis of TA-p63-γ was done as previously described (26). ΔN-p63–specific Western blotting was conducted by blocking filters in TBST-5% milk for 1 h at room temperature. The rabbit anti-ΔN-p63 antibody was incubated with the filter for 1 h at room temperature in TBST-1% milk. Following a series of washes in TBST, a horseradish peroxidase–linked goat anti-rabbit secondary antibody was used to detect the rabbit IgG, and detection was carried by chemiluminescence.

Nestin is expressed in the basal epithelial layer of the normal mammary gland. Several studies have indicated that ΔN-p63 is highly expressed in the basal/myoepithelial layer of the mammary gland and other epithelial structures and is required for the preservation of self-renewal (17, 19, 27). Studies of the limbal epithelia of the cornea indicate that ΔN-p63-α is expressed in the self-renewing population, and this expression is repressed as cells forfeit their self-renewing capacity, enter a stage of transient amplification, and achieve terminal differentiation (28). Other studies show colocalization of nestin and p63 in a subset of cells within the limbal epithelia (21). This study suggested that nestin may be expressed in the regenerative compartment of the mammary gland. FFPE samples of normal human mammary gland tissue were sectioned and subjected to immunohistochemical analysis of the expression of nestin, the basal epithelial marker CK14, and p63. A goat polyclonal antibody directed against human nestin detected robust expression of nestin (Fig. 1A) in the subluminal compartment of the mammary gland. Similar studies using a mouse monoclonal antibody directed against a distinct epitope of nestin resulted in an identical pattern of expression (Fig. 1B). Immunohistochemical analysis of CK14 and p63 (Fig. 1C and D) confirms that each is present in the subluminal epithelia of the mammary duct. Interestingly, we noted that CK14 expression was restricted to the columnar basal epithelia of the ducts and was present at low to undetectable levels in the lobules (Fig. 1C,, right). Two-color immunofluorescence using antibodies directed against two distinct epitopes of nestin resulted in a high degree of overlap, which increases confidence that the protein being detected is nestin (Fig. 1E). In mammary ducts and lobules, nestin was detected in two morphologically distinct cell types in the subluminal compartment (Fig. 1A , inset). The first is a subset of columnar basal epithelia (indicated by black arrowheads) in which cytoplasmic nestin staining surrounds the nucleus. A second filamentous cell type (indicated by red arrowheads) that is distributed along the periphery of the duct also stained positively for nestin. Importantly, regions within ducts and lobules are identifiable, in which the filamentous nestin-positive cell type is distinct and physically separate from the nestin-positive columnar epithelia. This observation indicates that nestin is expressed in two morphologically distinct subtypes in the subluminal compartment of the human mammary gland and raises the possibility that nestin may colocalize with CK14 and p63.

Nestin independently colocalizes with basal/myoepithelial markers. To determine if nestin is expressed in the subluminal layer of the mammary gland, two-color immunofluorescence was used to measure the degree of colocalization between nestin and p63 or CK14. Results indicate colocalization of nestin with p63 (Fig. 2A,, left), and confocal imaging indicates that the cytoplasmic nestin staining surrounds the nuclear p63 staining in multiple cells (Fig. 2A,, right). To confirm these studies, we sought to show that nestin is coexpressed with the basal epithelial marker CK14. Consistent with previous studies, two-color immunofluorescence indicates that CK14 and p63 are coexpressed in cells within the columnar basal epithelia of the mammary gland (Fig. 2B,, left). This result is also consistent with a recent study showing that CK14 is a direct transcriptional target of ΔN-p63 isoforms (29). Confocal imaging (Fig. 2B,, right) clearly shows the cytoplasmic staining of CK14 surrounding the nuclear staining of p63. Two-color immunofluorescence also indicates that colocalization of nestin and CK14 are coexpressed in these cells (Fig. 2C). We also noted that no p63 or CK14 expression was detected in the nestin-positive filamentous cells that appear along the periphery of the ducts and lobules. The location and morphology of these cells suggests that they represent myoepithelia, a myoepithelial precursor, or myofibroblasts. Other studies have noted that nestin is coexpressed with the striated muscle neurofilament protein desmin in regenerating skeletal muscle (30). To determine if the filamentous nestin-positive cells were myoepithelial, two-color immunofluorescence was conducted to determine if nestin was coexpressed with desmin in these cells. Results (Fig. 2D) indicate that nestin and desmin are colocalized in the filamentous myoepithelial cells. Additionally, two-color immunofluorescence indicates no overlap between desmin and p63 (Fig. 3E) or desmin and CK14 (Fig. 3F). Taken together, these studies indicate that nestin is expressed in two morphologically and biochemically distinct cell types in the subluminal compartment of the mammary gland. The colocalization of nestin with p63 coupled to the abundant evidence that nestin and p63 are expressed in the regenerative compartments of diverse tissues may be consistent with a role for nestin in the regenerative compartment of the mammary gland.

ΔN-p63 and nestin are coordinately regulated in the mouse mammary gland during pregnancy. Immunohistochemical analysis of p63 expression using the commercially available 4A4 monoclonal antibody indicates high levels of p63 expression in the subluminal compartment of the mammary gland. Because this antibody was raised against a region of p63 that is common to all p63 isoforms, it fails to distinguish between TA-p63 and ΔN-p63. To address this, we raised a rabbit polyclonal antibody against a 14-amino-acid peptide that is encoded by the ΔN-p63–specific exon 3′. Affinity-purified fractions of this serum are capable of detecting ΔN-p63-α but not TA-p63-γ when each was overexpressed in H1299 cells (Fig. 3A). Immunohistochemical analysis of normal human mammary gland using this antibody indicated that ΔN-p63 isoforms predominate in the subluminal epithelia of the mammary gland (Fig. 3B). This finding is consistent with previous studies (19) and is also consistent with the finding that ΔN-p63-α is the dominant isoform in both stem cells and transient amplifying cells of the limbal epithelia of the cornea (28). These studies coupled to the observation that p63 and nestin colocalize in the subluminal epithelia of the mammary gland suggest that nestin and ΔN-p63 may be coordinately regulated during the regenerative cycle of the mammary gland. To test this, mammary glands were harvested from mice at time points during pregnancy and analyzed for expression of ΔN-p63 and nestin. Results (Fig. 3C) of this study indicated that the expression of nestin and ΔN-p63 was undetectable during early pregnancy (P3 and P7) but was detected in the columnar cells of the subluminal compartment at day 14 of pregnancy (P14). Additionally, it was observed that by late pregnancy (P20), nestin expression was restricted to the myofibroblasts, and ΔN-p63 was undetectable. Importantly, staining for CK14 indicates that the basal epithelial cells are present throughout the time course of this pregnancy (Fig. 3C , right). These studies indicate that nestin and ΔN-p63 are coordinately regulated in the columnar basal epithelia during pregnancy in the mouse mammary gland.

ΔN-p63-α blocks cellular differentiation and preserves nestin expression. Coordinated expression and colocalization of nestin and ΔN-p63 coupled to the established role for ΔN-p63 in the preservation of self-renewing capacity in the mammary gland and other epithelial structures indicate that nestin is expressed in cells with self-renewing capacity. This further suggests that expression of nestin may be differentially regulated during the preservation or forfeiture of self-renewal. To test this, we employed the NT2/D1 embryonal carcinoma cell culture model, in which robust levels of nestin are observed to decline in response to all-trans-RA–induced cellular differentiation (31). Additionally, we noted that these cells did not express measurable levels of any ΔN-p63 isoform (data not shown). Treatment of NT2/D1 cells with 0.01% DMSO (vehicle) or 1 μmol/L RA indicated that RA-induced differentiation repressed nestin expression at 72 and 96 h posttreatment and activated expression of RARβ, a canonical RA target gene (Fig. 4A). Consistent with these results, RA-induced differentiation leads to immunodetection of A2B5, a glycolipid that is a well-characterized marker of NT2/D1 differentiation (refs. 3133; Fig. 4B). To determine if ectopic ΔN-p63-α is sufficient to block RA-induced repression of nestin expression, NT2/D1 cells were transfected with either green fluorescent protein (GFP; control) or ΔN-p63-α followed by treatment with either 0.01%DMSO (vehicle) or 1 μmol/L RA. RNA was collected at 0, 24, 48, 72, and 96 h posttreatment and subjected to Northern analysis. Results (Fig. 4C) indicate that GFP had no effect on repression of nestin or induction of RARβ by RA. In contrast to GFP, ectopic ΔN-p63-α was sufficient to block RA-induced repression of nestin expression. Importantly, the induction of RARβ by RA was still observed in the presence of ΔN-p63-α, but at somewhat lower levels, suggesting that ΔN-p63-α may elevate the threshold necessary for RA-induced differentiation. Immunodetection of A2B5 (Fig. 4D,, top) and nestin (Fig. 4D,, bottom) under identical conditions indicated that ectopic expression of ΔN-p63-α blocked differentiation of NT2/D1 cells by RA (Fig. 4D) and preserved expression of nestin. This result indicates that ΔN-p63-α opposes RA-mediated differentiation signals and is consistent with the role of ΔN-p63-α in the preservation of self-renewal. It also implies that nestin expression is reflective of the regenerative capacity of cells.

The observation that ΔN-p63-α was able to block the repression of nestin during differentiation suggested that nestin expression may also oppose cellular differentiation. To test this, a pool of four nestin-directed siRNAs or a pool or four negative control siRNAs was transfected into NT2/D1 cells. Northern analysis indicates that nestin expression was substantially reduced by the nestin-specific pool at 24 h but began to recover by 48 h after transfection (Fig. 5A). Following transfection under these conditions, cells were treated with 0.01% DMSO or 1 μmol/L RA, and immunodetection of A2B5 was measured at 0, 24, 48, 72, 96, and 120 h posttreatment. Results (Fig. 5B) indicated that siRNA-mediated repression of nestin over this time course was insufficient induce the A2B5 display or accelerate RA-induced A2B5 display. These results indicate that whereas repression of nestin occurs during differentiation induced by RA, direct ablation of nestin is insufficient to induce differentiation as measured by immunodetection of A2B5. Taken together, these studies indicate that ΔN-p63-α is able to oppose RA-induced cellular differentiation, thereby preserving expression of nestin and are consistent with previous work indicating that nestin expression is a feature of undifferentiated cells.

Breast tumors with a basal epithelial phenotype express nestin, ΔN-p63, and CK14. The tumor stem cell theory of breast carcinogenesis implies that breast cancers may initiate within a population of cells capable of self-renewal (34). This may further suggest that highly aggressive and poorly differentiated tumors display features of stem cells. Based on this, we hypothesized that the poorly differentiated basal epithelial breast cancer subtype would express nestin and other markers of the regenerative compartment of the mammary gland. Tumors that were triple negative for ERα, PR, and Her2 were selected for analysis of nestin, p63, and CK14 expression. For each marker, positivity was defined as detectable expression within the tumor and not merely at the periphery of the tumor. For example, a ductal carcinoma in situ (DCIS), in which the in tact basal/myoepithelial layer stains positive but the cells within the core of the DCIS do not, would be scored as negative. Results indicate that in 14 of 16 triple-negative tumors, nestin expression was readily detectable. Figure 6A shows a representative sample of nestin staining of a triple-negative tumor. Similarly, 16 of 16 basal breast tumors were positive for CK14 and indicated a range of signal intensities from intermediate to high. Figure 6B shows a representative sample of high-intensity CK14 staining of a basal epithelial tumor. Analysis of p63 expression in these tumors indicated that 8 of 16 tumors were positive and displayed a range of expression patterns from punctate (Fig. 6C,, left) to intermediate (Fig. 6C,, middle) to uniform (Fig. 6C , right). Finally, analysis of these tumors indicated that none showed any evidence of desmin expression (data not shown). These studies indicate that basal epithelial breast tumors are positive for nestin, CK14, and p63 and negative for desmin and suggest that basal breast tumors have a phenotype that is similar to cells within the regenerative compartment of the mammary gland.

To determine if nestin is a selective marker of the basal breast cancer subtype, we evaluated 16 tumors that were representative of the Her2 subtype (ER/PRHer2+ by FISH) and 16 tumors with a luminal epithelial phenotype (ER+/PR+). Under the same conditions in which nestin was detected in the basal tumors, we failed to detect nestin in these other subtypes. Consistent with our analysis of the basal breast tumors, a positive signal for nestin was defined as immunodetectable expression within the tumor. Nestin expression was detected at the periphery of ducts and in DCIS but not in the tumor itself, and these tumors were scored as negative. These studies (summarized in Table 1), although limited by low sample numbers, indicate that nestin may be a selective marker of the basal breast cancer subtype.

Nestin expression is common in BRCA1-associated tumors. Global transcriptional profiling of human breast cancers has indicated that tumors that are associated with mutations in BRCA1 cluster within the basal breast cancer subtype (3). We therefore sought to determine if expression of nestin was detectable in BRCA1-associated tumors. Immunohistochemical analysis of these tumors with either the goat anti-nestin polyclonal antibody or the mouse anti-nestin monoclonal antibody indicated robust nestin (Fig. 6D) expression in six of eight individual cases. Similar to the basal breast tumors studied above, expression of CK14 and p63 was also detected in the BRCA1-associated tumors. These studies indicate that nestin expression is correlated with BRCA1 associated tumors and are consistent with the finding that BRCA1-associated tumors are classified as a basal breast cancer. This data coupled to our studies indicating that in normal mammary gland nestin is coexpressed with markers of the basal/myoepithelia may indicate that BRCA1-associated tumors display a progenitor-like phenotype. They further suggest that nestin may be a marker of BRCA1-associated tumors.

We describe here the identification of nestin as a selective marker of the basal breast cancer subtype. We present evidence that nestin is expressed in two morphologically and biochemically distinct subtypes within the basal/myoepithelial layer of the normal human mammary gland. In one of these cell types, nestin is coexpressed with ΔN-p63, which coupled to the role of ΔN-p63 in preservation of self-renewal, suggesting that nestin may be expressed in the regenerative compartment within the mammary gland. Analysis the mammary glands of pregnant mice indicated that the timing and cell type specificity of nestin and ΔN-p63 expression are coordinately regulated. Our data indicate that ΔN-p63-α is sufficient to block RA-induced differentiation of NT2/D1 cells, and that this ability is linked to the preservation of nestin expression. Although the precise relevance of this observation to the relationship between ΔN-p63 isoforms and nestin in mammary progenitors is unclear, this finding does indicate that ΔN-p63-α is able to preserve a state of dedifferentiation in RA-treated NT2/D1 cells and in doing so preserves expression of nestin. Breast tumors representing the basal breast cancer subtype (ER/PR/Her2) express robust levels of nestin and CK14 and display a diverse pattern of ΔN-p63 expression, suggesting a basal/myoepithelial phenotype. This is likely to be consistent with the aggressive nature and poorly differentiated phenotype of the basal epithelial class of breast tumors. Our studies also identify nestin as a potential target for molecular detection and diagnosis of breast cancers with a basal phenotype, including those with known BRCA1 mutations.

Analysis of p63 expression in basal epithelial breast tumors indicated that 8 of 16 samples were observed to express p63, whereas none of the Her2-associated or luminal tumor types showed any detectable expression of p63. The expression pattern of p63 varied from punctate to uniform, indicating that expression of p63 isoforms is unlikely to be a clinically useful marker. Although it is unclear if p63 expression identifies tumor stem cells, that p63 is required for self-renewal may indicate that these cells retain the capacity for self-renewal. Further analysis of mammary tumor stem cells will be required to determine if p63 expression underlies the self-renewing capacity of tumor stem cells in the basal epithelial subtype.

Our study indicates that nestin is expressed in the basal/myoepithelial layer of the mammary gland and is a selective marker of the basal epithelial breast cancer subtype. Although the precise function of nestin remains to be elucidated, several studies indicate that it may play a role in the regulation of mitosis (20, 35, 36). These findings coupled to the use of nestin as a marker of neural progenitors and the nestin promoter to selectively target neural progenitors suggest a role for nestin in the progenitor pools of diverse tissues. Our finding that siRNA-mediated repression of nestin was insufficient to promote cellular differentiation indicates that nestin is unable to actively preserve self-renewing capacity in stem cell populations. Additionally, the observation that ectopic ΔN-p63-α blocks differentiation thereby preserving nestin expression coupled to the studies indicating a potential role in the regulation of mitosis is consistent with the conclusion that nestin is expressed in a population of cells with retained proliferative capacity. We believe this is consistent with the poorly differentiated phenotype of the basal epithelial subtype of breast tumors. Additionally, the finding that nestin expression was restricted to the most aggressive and least differentiated breast tumor subtype coupled to the finding that ectopic expression of ΔN-p63-α was sufficient to preserve nestin expression during differentiation may suggest that the degree of progenitor-like features correlates with the aggressiveness and differentiation state of the tumors. This would further imply that the presence of nestin within a tumor might correlate with poor clinical prognosis and is consistent with the clinical features of the basal epithelial breast cancer phenotype. Further analysis of the function of nestin in mammary progenitors may provide greater insight into self-renewal processes. Additionally, larger retrospective studies will be necessary to evaluate the prognostic significance of nestin.

Grant support: U.S. Department of Defense Breast Cancer Research Program's Multidisciplinary Postdoctoral Training Award contract no. W81XWH-05-1-0350 (H. Li), Mary Kay Ash Charitable Foundation Scholars Award (J. DiRenzo), and National Cancer Institute grant 1RO1 CA108539 (J. DiRenzo).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Perou CM, Sorlie T, Eisen MB, et al. Molecular portraits of human breast tumours.
Nature
2000
;
406
:
747
–52.
2
Sorlie T, Perou CM, Tibshirani R, et al. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications.
Proc Natl Acad Sci U S A
2001
;
98
:
10869
–74.
3
van 't Veer LJ, Dai H, van de Vijver MJ, et al. Gene expression profiling predicts clinical outcome of breast cancer.
Nature
2002
;
415
:
530
–6.
4
West M, Blanchette C, Dressman H, et al. Predicting the clinical status of human breast cancer by using gene expression profiles.
Proc Natl Acad Sci U S A
2001
;
98
:
11462
–7.
5
Carey LA, Perou CM, Livasy CA, et al. Race, breast cancer subtypes, and survival in the Carolina Breast Cancer Study.
JAMA
2006
;
295
:
2492
–502.
6
Collett K, Stefansson IM, Eide J, et al. A basal epithelial phenotype is more frequent in interval breast cancers compared with screen detected tumors.
Cancer Epidemiol Biomarkers Prev
2005
;
14
:
1108
–12.
7
Sorlie T, Tibshirani R, Parker J, et al. Repeated observation of breast tumor subtypes in independent gene expression data sets.
Proc Natl Acad Sci U S A
2003
;
100
:
8418
–23.
8
Chepko G, Smith GH. Mammary epithelial stem cells: our current understanding.
J Mammary Gland Biol Neoplasia
1999
;
4
:
35
–52.
9
Smalley M, Ashworth A. Stem cells and breast cancer: a field in transit.
Nat Rev Cancer
2003
;
3
:
832
–44.
10
Shackleton M, Vaillant F, Simpson KJ, et al. Generation of a functional mammary gland from a single stem cell.
Nature
2006
;
439
:
84
–8.
11
Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF. Prospective identification of tumorigenic breast cancer cells.
Proc Natl Acad Sci U S A
2003
;
100
:
3983
–8.
12
Kaelin WG, Jr. The p53 gene family.
Oncogene
1999
;
18
:
7701
–5.
13
Mills AA, Zheng B, Wang XJ, Vogel H, Roop DR, Bradley A. p63 is a p53 homologue required for limb and epidermal morphogenesis.
Nature
1999
;
398
:
708
–13.
14
van Bokhoven H, McKeon F. Mutations in the p53 homolog p63: allele-specific developmental syndromes in humans.
Trends Mol Med
2002
;
8
:
133
–9.
15
Yang A, Schweitzer R, Sun D, et al. p63 is essential for regenerative proliferation in limb, craniofacial and epithelial development.
Nature
1999
;
398
:
714
–8.
16
Yang A, Kaghad M, Wang Y, et al. p63, a p53 homolog at 3q27–29, encodes multiple products with transactivating, death-inducing, and dominant-negative activities.
Mol Cell
1998
;
2
:
305
–16.
17
DiRenzo J, Signoretti S, Nakamura N, et al. Growth factor requirements and basal phenotype of an immortalized mammary epithelial cell line.
Cancer Res
2002
;
62
:
89
–98.
18
van Bokhoven H, Hamel BC, Bamshad M, et al. p63 Gene mutations in eec syndrome, limb-mammary syndrome, and isolated split hand-split foot malformation suggest a genotype-phenotype correlation.
Am J Hum Genet
2001
;
69
:
481
–92.
19
Nylander K, Coates PJ, Hall PA. Characterization of the expression pattern of p63 alpha and delta Np63 alpha in benign and malignant oral epithelial lesions.
Int J Cancer
2000
;
87
:
368
–72.
20
Wiese C, Rolletschek A, Kania G, et al. Nestin expression: a property of multi-lineage progenitor cells?
Cell Mol Life Sci
2004
;
61
:
2510
–22.
21
Seigel GM, Sun W, Salvi R, Campbell LM, Sullivan S, Reidy JJ. Human corneal stem cells display functional neuronal properties.
Mol Vis
2003
;
9
:
159
–63.
22
Dahlstrand J, Zimmerman LB, McKay RD, Lendahl U. Characterization of the human nestin gene reveals a close evolutionary relationship to neurofilaments.
J Cell Sci
1992
;
103
:
589
–97.
23
Lendahl U, Zimmerman LB, McKay RD. CNS stem cells express a new class of intermediate filament protein.
Cell
1990
;
60
:
585
–95.
24
Foulkes WD. BRCA1 functions as a breast stem cell regulator.
J Med Genet
2004
;
41
:
1
–5.
25
Harmes DC, Bresnick E, Lubin EA, et al. Positive and negative regulation of deltaN-p63 promoter activity by p53 and deltaN-p63-alpha contributes to differential regulation of p53 target genes.
Oncogene
2003
;
22
:
7607
–16.
26
Li N, Li H, Cherukuri P, Farzan S, Harmes DC, DiRenzo J. TA-p63-gamma regulates expression of DeltaN-p63 in a manner that is sensitive to p53.
Oncogene
2006
;
25
:
2349
–59.
27
Nylander K, Vojtesek B, Nenutil R, et al. Differential expression of p63 isoforms in normal tissues and neoplastic cells.
J Pathol
2002
;
198
:
417
–27.
28
Pellegrini G, Dellambra E, Golisano O, et al. p63 identifies keratinocyte stem cells.
Proc Natl Acad Sci U S A
2001
;
98
:
3156
–61.
29
Candi E, Rufini A, Terrinoni A, et al. Differential roles of p63 isoforms in epidermal development: selective genetic complementation in p63 null mice.
Cell Death Differ
2006
;
13
:
1037
–47.
30
Vaittinen S, Lukka R, Sahlgren C, et al. The expression of intermediate filament protein nestin as related to vimentin and desmin in regenerating skeletal muscle.
J Neuropathol Exp Neurol
2001
;
60
:
588
–97.
31
Przyborski SA, Morton IE, Wood A, Andrews PW. Developmental regulation of neurogenesis in the pluripotent human embryonal carcinoma cell line NTERA-2.
Eur J Neurosci
2000
;
12
:
3521
–8.
32
Andrews PW. Teratocarcinomas and human embryology: pluripotent human EC cell lines. Review article.
APMIS
1998
;
106
:
158
–67; discussion 67–8.
33
Wenk J, Andrews PW, Casper J, et al. Glycolipids of germ cell tumors: extended globo-series glycolipids are a hallmark of human embryonal carcinoma cells.
Int J Cancer
1994
;
58
:
108
–15.
34
Pardal R, Clarke MF, Morrison SJ. Applying the principles of stem-cell biology to cancer.
Nat Rev Cancer
2003
;
3
:
895
–902.
35
Sahlgren CM, Mikhailov A, Hellman J, et al. Mitotic reorganization of the intermediate filament protein nestin involves phosphorylation by cdc2 kinase.
J Biol Chem
2001
;
276
:
16456
–63.
36
Sahlgren CM, Mikhailov A, Vaittinen S, et al. Cdk5 regulates the organization of Nestin and its association with p35.
Mol Cell Biol
2003
;
23
:
5090
–106.