Purpose: hMena (ENAH), a cytoskeleton regulatory protein involved in the regulation of cell motility and adhesion, is overexpressed in breast cancer. The aim of this study was to define at what stage of breast carcinogenesis hMena is overexpressed and to correlate hMena overexpression with established prognostic factors in breast cancer, focusing on human epidermal growth factor receptor-2 (HER-2).

Experimental Design: hMena expression was assessed immunohistochemically in a prospective cohort of cases (n = 360) encompassing a highly representative spectrum of benign breast diseases associated with different risk of transformation, in situ, invasive, and metastatic tumors. Correlations with conventional pathologic and prognostic variables, such as proliferation index, hormonal receptor status, and HER-2 overexpression, were also evaluated. In vitro experiments were done to study the effect of neuregulin-1 and Herceptin treatments on hMena expression.

Results: hMena protein is undetectable in normal breast and is weakly expressed in a small percentage of low-risk benign diseases (9%), but displays a progressive and significant increase of positivity in benign lesions at higher risk of transformation (slightly increased risk 43%; moderate increased risk 67%), in in situ (72%), invasive (93%), and metastatic breast cancer (91%). A significant direct correlation with tumor size (P = 0.04), proliferation index (P < 0.0001), and HER-2 overexpression (P < 0.0001) and an inverse relationship with estrogen (P = 0.036) and progesterone receptors (P = 0.001) are found in invasive carcinomas. In vitro experiments show that neuregulin-1 up-regulates, whereas Herceptin down-regulates, hMena expression.

Conclusions: Our data provide new insights into the relevance of actin-binding proteins in human breast carcinogenesis and indicate hMena overexpression as a surrogate indicator in breast disease management.

Management of breast cancer, and other malignancies, is likely to benefit from the identification of early markers of transformation. In this context, major efforts are ongoing in the areas of genomic and proteomic profiling aimed at identifying genetic or biochemical markers related to proliferative lesions heralding the development of breast cancer. In view of the increasing evidence that the host immune response contributes to the editing of the tumor phenotype (1), we have recently isolated, by serologic analysis of cDNA expression libraries (SEREX), hMena (ENAH) protein, the human orthologue of murine Mena, which is overexpressed in over 70% of primary breast cancers (2). Mena belongs to the Ena/VASP protein family, which, by controlling the geometry of the actin filament network (3, 4), represents key regulator molecules of cell movement and shape in a large variety of cell types and organisms (5). It has been suggested that Ena/VASP proteins are members of the adherens junction structures and are required for the actin dynamics necessary to seal membranes into epithelial sheets (6), a process frequently deregulated during transformation. Of relevance, cytoskeleton regulatory proteins have been found to be up-regulated in the early stages (7) and in the progression of breast cancer (8). Furthermore, the murine Mena is overexpressed in a subpopulation of breast tumor cells with highly invasive properties (9). In view of the above, in the present study, we have addressed the question of whether hMena may represent an early marker of breast carcinogenesis, a process arising via a series of intermediate hyperplastic lesions through various grades of atypia (10). To this end, we have analyzed by immunohistochemistry hMena expression in a prospective series of cases encompassing a highly representative spectrum of benign and malignant breast lesions. This information has been correlated in in situ and invasive breast cancer, with conventional pathologic variables (i.e., tumor size and nodal status) and with known prognostic factors, such as proliferation index, human epidermal growth factor receptor-2 (HER-2) overexpression, and hormonal receptor status (11).

Data presented herein show that hMena protein, although not expressed in normal breast, is detectable in a small percentage of low-risk proliferative lesions, with a progressive increase of positivity in benign breast lesions at higher risk of transformation in in situ and invasive cancers. In the latter, a significant direct correlation was found between hMena, tumor size, proliferation index, and HER-2 overexpression whereas an inverse relationship was evidenced with estrogen receptor (ER) and progesterone receptor (PgR).

Furthermore, in vitro studies suggest that stimulation by neuregulin-1 (NRG1) or inhibition by Herceptin of HER-2 activity affects hMena expression.

Patients and tissue specimens. A consecutive series of 74 patients (median age, 49 years; range, 19-84) with benign breast diseases, 10 normal breasts sampled from mastoplasty, 65 patients with ductal carcinoma in situ (DCIS), and 163 with invasive breast cancer of different histotypes, surgically treated at the Regina Elena Cancer Institute (Rome, Italy) between 2000 and 2004, were prospectively collected for immunohistochemical studies with informed consent of the patients. Furthermore, 48 synchronous and metachronous breast metastases, sampled from different body sites (axillary node, n = 15; supraclavicular node, n = 3; mediastinic node, n = 1; lung, n = 11; brain, n = 8; pleura, n = 4; ovary, n = 2; liver, n = 2; and bone, n = 2), were obtained. Detailed clinicopathologic information of the DCIS and invasive breast cancer patients are listed in Tables 1 and 2, respectively.

Table 1.

Clinicopathologic characteristics of 65 DCIS

CharacteristicsNo. cases (%)
No. patients 65 
Median age, y (range), 59 (38–85)  
Histotype  
    Low grade 13 (20) 
    Intermediate grade 8 (12) 
    High grade 44 (68) 
ER  
    Negative 26 (40) 
    Positive 39 (60) 
PgR  
    Negative 32 (49) 
    Positive 33 (51) 
Ki67  
    Negative (<15) 36 (55) 
    Positive (>15) 29 (45) 
HER-2 overexpression  
    Negative 26 (40) 
    Positive 39 (60) 
CharacteristicsNo. cases (%)
No. patients 65 
Median age, y (range), 59 (38–85)  
Histotype  
    Low grade 13 (20) 
    Intermediate grade 8 (12) 
    High grade 44 (68) 
ER  
    Negative 26 (40) 
    Positive 39 (60) 
PgR  
    Negative 32 (49) 
    Positive 33 (51) 
Ki67  
    Negative (<15) 36 (55) 
    Positive (>15) 29 (45) 
HER-2 overexpression  
    Negative 26 (40) 
    Positive 39 (60) 
Table 2.

Clinicopathologic characteristics of 163 invasive breast carcinomas

CharacteristicsNo. cases (%)
No. patients 163 
Median age, y (range), 57 (28–88)  
Histotype  
    Invasive ductal carcinoma 148 (91) 
    Invasive lobular carcinoma 11 (7) 
    Invasive papillary carcinoma 4 (2) 
Histologic grade  
    Grade 1 21 (13) 
    Grade 2 104 (64) 
    Grade 3 38 (23) 
Lymph node status  
    Negative 87 (53) 
    Positive 76 (47) 
Tumor size  
    T1 102 (63) 
    T2 46 (28) 
    T3,4 15 (9) 
ER  
    Negative 38 (23) 
    Positive 125 (77) 
PgR  
    Negative 63 (39) 
    Positive 100 (61) 
Ki67  
    Negative (<15) 97 (59) 
    Positive (>15) 66 (41) 
HER-2 overexpression  
    Negative 124 (76) 
    Positive 39 (24) 
CharacteristicsNo. cases (%)
No. patients 163 
Median age, y (range), 57 (28–88)  
Histotype  
    Invasive ductal carcinoma 148 (91) 
    Invasive lobular carcinoma 11 (7) 
    Invasive papillary carcinoma 4 (2) 
Histologic grade  
    Grade 1 21 (13) 
    Grade 2 104 (64) 
    Grade 3 38 (23) 
Lymph node status  
    Negative 87 (53) 
    Positive 76 (47) 
Tumor size  
    T1 102 (63) 
    T2 46 (28) 
    T3,4 15 (9) 
ER  
    Negative 38 (23) 
    Positive 125 (77) 
PgR  
    Negative 63 (39) 
    Positive 100 (61) 
Ki67  
    Negative (<15) 97 (59) 
    Positive (>15) 66 (41) 
HER-2 overexpression  
    Negative 124 (76) 
    Positive 39 (24) 

Benign breast diseases were grouped according to the relative risk of developing invasive carcinoma using The Cancer Committee of the College of the American Pathologists criteria (12). Because, as expected, single specimens contained more than one type of benign lesion for each patient, hMena expression was evaluated in all the different histopathologic lesions identified morphologically in the same section. On the basis of these criteria, we managed to analyze 165 different lesions in 74 patients.

As described in Table 3, hMena protein expression was evaluated in 84 lesions morphologically defined as “no increased risk” (NR; 32 adenosis, 15 duct ectasia + cysts, 17 fibroadenomas, 3 fibrosis, and 17 simple apocrine metaplasia); 66 as “slightly increased risk” (SR; 12 sclerosing adenosis, 5 complex fibroadenomas, 45 hyperplasia without atypia, 4 intraductal papillomas); and 15 as “moderate increased risk” (MR; 11 atypical ductal hyperplasia; and 4 atypical papilloma; ref. 12).

Table 3.

hMena protein expression in 165 benign breast lesions with different risk of progression sampled from 74 patients

Benign breast diseasesSingle lesionTwo lesionsThree lesions
 14 NR 3 NR/MR 8 NR/NR/NR   
 3 SR 18 SR/SR 11 NR/NR/SR   
 1 MR  5 NR/NR/MR   
   7 NR/SR/SR   
   2 NR/SR/MR   
 
 

 
2 NR/MR/MR
 
  
Total patients (N = 74) 18 patients 21 patients 35 patients   

 
     
Benign breast lesions No. lesions (N = 165) hMena score
 
   

 

 
0 (%) 1+(%) 2+(%) 3+(%) 
NR      
    Adenosis 32 25 
    Duct ectasia + cysts 15 15 
    Fibroadenoma 17 17 
    Fibrosis 
    Simple apocrine metaplasia 17 16 
    Total lesions 84 76 (91) 6 (7) 2 (2) 0 (0) 
SR      
    Sclerosing adenosis 12 10 
    Complex fibroadenoma 5* 
    Hyperplasia without atypia 45 24 
    Intraductal papilloma 
    Total lesions 66 38 (57) 9 (14) 10 (15) 9 (14) 
MR      
    Atypical ductal hyperplasia 11 
    Atypical papilloma 
    Total lesions 15 5 (33) 1 (7) 3 (20) 6 (40) 
Benign breast diseasesSingle lesionTwo lesionsThree lesions
 14 NR 3 NR/MR 8 NR/NR/NR   
 3 SR 18 SR/SR 11 NR/NR/SR   
 1 MR  5 NR/NR/MR   
   7 NR/SR/SR   
   2 NR/SR/MR   
 
 

 
2 NR/MR/MR
 
  
Total patients (N = 74) 18 patients 21 patients 35 patients   

 
     
Benign breast lesions No. lesions (N = 165) hMena score
 
   

 

 
0 (%) 1+(%) 2+(%) 3+(%) 
NR      
    Adenosis 32 25 
    Duct ectasia + cysts 15 15 
    Fibroadenoma 17 17 
    Fibrosis 
    Simple apocrine metaplasia 17 16 
    Total lesions 84 76 (91) 6 (7) 2 (2) 0 (0) 
SR      
    Sclerosing adenosis 12 10 
    Complex fibroadenoma 5* 
    Hyperplasia without atypia 45 24 
    Intraductal papilloma 
    Total lesions 66 38 (57) 9 (14) 10 (15) 9 (14) 
MR      
    Atypical ductal hyperplasia 11 
    Atypical papilloma 
    Total lesions 15 5 (33) 1 (7) 3 (20) 6 (40) 
*

Positivity of complex fibroadenomas is restricted to the hyperplastic component.

DCIS and invasive breast cancer were classified according to the WHO Classification of Tumors (13) and were graded according to Bloom and Richardson.

Immunohistochemistry. hMena, ERα, PgR, HER-2, and Ki67 expression was assessed by indirect immunoperoxidase staining. Breast specimens were fixed for 18 to 24 hours in 4% buffered formaldehyde and then processed through to paraffin wax. Immunohistochemical staining was carried out on 5-μm-thick paraffin-embedded tissues. Sections were harvested on SuperFrost Plus slides (Menzel-Glaser, Braunschweig, Germany). The deparaffinized and rehydrated sections were pretreated in a thermostatic bath at 96°C for 40 minutes in 10 mmol/L citrate buffer (pH 6). Sections were incubated with Mena monoclonal antibody (mAb clone 21, BD Transduction, San Jose, CA; 2.5 μg/mL) that specifically recognizes hMena and does not cross-react with other members of Ena/VASP family proteins for 60 minutes at room temperature. Thirty-minute incubation at room temperature was used for anti-ERα mAb (clone 6F11, Novocastra, UCS Diagnostics, Rome, Italy), anti-PgR mAb (clone 1A6, UCS Diagnostics), anti-Ki67 mAb (clone MIB-1, DakoCytomation, Milan, Italy), and the anti-HER-2 polyclonal antibody (A0485, DakoCytomation). The immunoreactions were revealed using horseradish peroxidase-LSAB2 system (DakoCytomation) using 3-amino-9-ethylcarbazole (DakoCytomation) as chromogenic substrate for ERα, PgR, Ki67, HER-2, and 3,3′-diaminobenzidine (DakoCytomation) for hMena. All sections were slightly counterstained with Mayer's hematoxylin and mounted in Permount or in aqueous mounting medium (UCS Diagnostics) according to the type of chromogenic substrate used. The intensity of hMena staining, detected in the cytoplasm, was scored from 0 to 3+ according to the following criteria: no staining, score 0; weak diffuse cytoplasmic staining in <10% of neoplastic cells, score 1+; moderate cytoplasmic staining in 10% to 70%, score 2+; strong cytoplasmic staining in >70% of the tumor cells with or without a juxtamembrane reinforce, score 3+. The staining was completely abolished by the preincubation with a lysate of a hMena-transfected breast cancer cell line (Supplementary Fig. S1). ERα and PgR were considered positive when >10% of the neoplastic cells showed a nuclear immunoreactivity. Fifteen percent of positive nuclei was the cutoff for Ki67. HER-2 overexpression was determined as defined in the HercepTest kit guide: scores 0 or 1+ were considered negative, 2+ weak positive, and 3+ strong positive. To qualify for 2+ scoring, complete membrane staining of >10% of tumor cells at a moderate intensity had to be observed. In the majority of 3+ cases, at least 80% of the tumor cells showed an intense and homogeneous cell membrane staining.

Evaluation of the immunohistochemical results was done independently and in blinded manner by three investigators (M. Mottolese, A. Di Benedetto, and F. Novelli).

Statistical analysis. The χ2 test for trend was used to verify whether hMena expression is associated with benign lesions with an increased risk of transformation.

The Pearson's χ2 test was used to assess in in situ and invasive breast carcinomas, the correlation between hMena staining intensity, and the relevant clinical and biopathologic variables.

The overall threshold of significance was 0.05 for both tests.

Cell cultures and in vitro treatment. The human breast carcinoma cell lines BT474 and MCF-7, purchased from the American Type Culture Collection (Rockville, MD), were cultured in DMEM supplemented with 10% FCS, glutamine, at 37°C under 5% CO2-95% air. Human breast cancer cell line DAL (2) and hMena-transfected DAL were used for experiments showed in the Supplementary Fig. S1. For NRG1 treatment, exponentially growing cells were cultured for 24 hours in medium without serum and then incubated with the same medium containing 10 ng/mL NRG1 (HRG-β1, R&D Systems, Minneapolis, MN) for 12, 24, or 48 hours. Herceptin (Roche, Monza, Italy) for clinical and in vitro use was stored at 4°C and adjusted to the final concentration of 25 or 50 μg/mL with culture medium. For all experiments, exponentially growing cells were exposed to the treatment for 24, 48, or 72 hours.

Treated and control cells were washed and processed according to the experiment to be done.

Western blot. Cells were rinsed thrice with ice-cold PBS and lysed on ice with 100 μL/well of lysis buffer containing 50 mmol/L HEPES, 150 mmol/L NaCl, 10% glycerol, 5 mmol/L EDTA, 1.5 mmol/L MgCl2, and 1%Triton with protease inhibitors. Lysates were centrifuged and supernatant was determined for protein quantification using BCA Protein Assay Reagent (Pierce, Rockford, IL). Thirty micrograms of protein extracts were resolved on 10% polyacrylamide gel and transferred to nitrocellulose membrane (Amersham Biosciences, Little Chalfont, United Kingdom). Blots were blocked for 1 hour with 3% skimmed milk in TBST and probed with 10 μg/mL anti-Mena mAb (clone 21, BD Transduction) or 10 μg/mL anti-HER-2 rabbit antibody (C-18, Santa Cruz Biotechnology, Santa Cruz, CA) in 3% skimmed milk/TBST for 1 hour. After three washes of 15 minutes each, blots were incubated with the appropriate secondary antibody conjugated with horseradish peroxidase for 1 hour and then washed again thrice. The protein signals were detected by ECL kit (Amersham Biosciences). For actin signal, blots were reprobed with 1 μg/mL monoclonal anti-actin, mouse ascites fluid clone AC-40 (Sigma-Aldrich, Milan, Italy).

Flow cytometry. Tumor cells were removed from plates using PBS with 2 mmol/L EDTA. After washing thrice with PBS, cells were stained for HER-2 using 10 μg/mL of two different antibodies (W6-100 and W6/800; ref. 14) recognizing two distinct epitopes of the oncoprotein extracellular domain, at 4°C for 30 minutes, followed by three washes and by secondary labeling with FITC-conjugated goat anti-mouse antibody (ICN Biomedicals, Irvine, CA) done at 4°C for 30 minutes. Isotype-matched mouse IgG were used as control.

After PBS washes, cells underwent cytofluorimetric analysis by FACScan (Becton-Dickinson, Milan, Italy).

hMena expression in benign breast lesions with different risk of progression. To evaluate at what stage of breast carcinogenesis hMena protein overexpression occurs, we analyzed 165 benign breast lesions sampled from 74 patients grouped according to the criteria defined by the Cancer Committee of the College of the American Pathologists (12). In Table 3, we describe the different lesions and their combination in each patient's specimen. In particular, we found in 18 patients a single lesion, in 21 patients two different lesions, and in 35 patients three different lesions.

The majority of the NR lesions (91%), including adenosis, ductal ectasia, cysts, fibroadenomas, fibrosis, and simple apocrine metaplasia, were hMena negative. The NR lesions only presented a weak/moderate hMena positivity (score 1+/2+) in a small percentage (9%) of cases. Of interest, these NR lesions hMena positive (8 of 84 cases) were never found in the single lesions, but only in combination with SR or MR lesions. In SR lesions, including sclerosing adenosis, complex fibroadenomas, hyperplasias without atypia, intraductal papillomas, the percentage of hMena-negative lesions decreases to 57%, and in MR lesions, which include atypical duct hyperplasias and atypical papillomas, to 33%. hMena score 3+ is evidenced in 14% of SR lesions and this percentage increased up to 40% in the atypical lesions that are included in MR benign diseases. As illustrated in Fig. 1, the percentage of hMena positive lesions (score 3+) is significantly higher in SR and MR than in NR lesions (χ2 for trend, P < 0.0001). Furthermore, none of the SR and MR lesions, displaying a 3+ immunohistochemistry score, showed a juxtamembrane reinforcement. All the 10 normal breasts sampled from mastoplasty were hMena negative. Representative cases of hMena immunohistochemical expression in benign breast diseases are shown in Fig. 2.

Fig. 1.

hMena expression in benign lesions with different risk of progression. White columns, hMena-negative lesions (score 0); black columns, hMena-positive lesions (score 3+). The percentage of hMena-positive benign lesions is significantly and progressively higher from NR to SR and MR. *, P < 0.0001 according to nonparametric χ2 test for trend.

Fig. 1.

hMena expression in benign lesions with different risk of progression. White columns, hMena-negative lesions (score 0); black columns, hMena-positive lesions (score 3+). The percentage of hMena-positive benign lesions is significantly and progressively higher from NR to SR and MR. *, P < 0.0001 according to nonparametric χ2 test for trend.

Close modal
Fig. 2.

hMena protein detection in benign breast diseases by immunohistochemistry. A, no immunostaining was evidenced in a NR lesion represented here by a fibroadenoma. B, a ductal hyperplasia without atypia, considered a SR lesion, shows a weak, but homogeneous, cytoplasmic immunoreactivity (score 1+). C, florid hyperplasia without atypia (SR), composed of cells with indistinct cell margins and variation in shape and size of nuclei, showing a cytoplasmic immunostaining score of 2+. Note the small central area with a moderate hyperplasia (arrow) presenting a weaker staining (score 1+). D, atypical hyperplasia with a strong (3+) hMena positivity. Magnifications, ×20 (A); ×40 (B-D).

Fig. 2.

hMena protein detection in benign breast diseases by immunohistochemistry. A, no immunostaining was evidenced in a NR lesion represented here by a fibroadenoma. B, a ductal hyperplasia without atypia, considered a SR lesion, shows a weak, but homogeneous, cytoplasmic immunoreactivity (score 1+). C, florid hyperplasia without atypia (SR), composed of cells with indistinct cell margins and variation in shape and size of nuclei, showing a cytoplasmic immunostaining score of 2+. Note the small central area with a moderate hyperplasia (arrow) presenting a weaker staining (score 1+). D, atypical hyperplasia with a strong (3+) hMena positivity. Magnifications, ×20 (A); ×40 (B-D).

Close modal

hMena overexpression in DCIS and invasive breast cancer: relationship with pathologic variables. As reported in Table 4, hMena immunoreactivity was detected in 72% (47 of 65) of DCIS independently of the score. Of interest, the intensity of hMena immunohistochemical positivity was related to tumor grade, low-grade DCIS being strongly positive in 30% (15% score 2+ and 15% score 3+), whereas intermediate and high-grade DCIS in 63% (25% score 2+, 38% score 3+) and 66% (25% score 2+, 41% score 3+) of the cases, respectively. Representative hMena staining in DCIS is reported in Fig. 3A and B.

Table 4.

hMena protein expression in 65 DCIS and 163 invasive breast carcinomas

HistotypeNo. caseshMena score
0 (%)1+ (%)2+ (%)3+ (%)
In situ carcinomas      
    Low grade 13 5 (39) 4 (31) 2 (15) 2 (15) 
    Intermediate grade 1 (12) 2 (25) 2 (25) 3 (38) 
    High grade 44 12 (27) 3 (7) 11 (25) 18 (41) 
    Total cases 65 18 (28) 9 (14) 15 (23) 23 (35) 
Invasive carcinomas      
    Invasive ductal carcinoma 148 9 (6) 33 (22) 54 (37) 52 (35) 
    Invasive lobular carcinoma 11 1 (9) 4 (37) 3 (27) 3 (27) 
    Invasive papillary carcinoma 1 (25) 1 (25) 2 (50) 0 (0) 
    Total cases 163 11 (7) 38 (23) 59 (36) 55 (34) 
HistotypeNo. caseshMena score
0 (%)1+ (%)2+ (%)3+ (%)
In situ carcinomas      
    Low grade 13 5 (39) 4 (31) 2 (15) 2 (15) 
    Intermediate grade 1 (12) 2 (25) 2 (25) 3 (38) 
    High grade 44 12 (27) 3 (7) 11 (25) 18 (41) 
    Total cases 65 18 (28) 9 (14) 15 (23) 23 (35) 
Invasive carcinomas      
    Invasive ductal carcinoma 148 9 (6) 33 (22) 54 (37) 52 (35) 
    Invasive lobular carcinoma 11 1 (9) 4 (37) 3 (27) 3 (27) 
    Invasive papillary carcinoma 1 (25) 1 (25) 2 (50) 0 (0) 
    Total cases 163 11 (7) 38 (23) 59 (36) 55 (34) 
Fig. 3.

hMena expression in DCIS, invasive, and metastatic breast carcinomas by immunohistochemistry. A, high-grade DCIS with features of comedonecrosis, a recognized precursor lesion for the development of invasive breast cancer, shows a strong cytoplasmic immunoreactivity score of 3+. B, an intermediate-grade DCIS with a hMena immunoreactivity score of 3+ compared with negative normal ducts (arrow). C to E, three representative cases of invasive ductal carcinomas displaying immunoreactivity scores of 1+ (C), 2+ (D), and 3+(E) respectively. F, a metastatic lymph node displaying a strong hMena immunostaining score of 3+ with juxtamembrane reinforcement in the neoplastic cells. Normal lymphocytes are completely negative. Magnifications, ×20 (A and B); ×40 (C-F).

Fig. 3.

hMena expression in DCIS, invasive, and metastatic breast carcinomas by immunohistochemistry. A, high-grade DCIS with features of comedonecrosis, a recognized precursor lesion for the development of invasive breast cancer, shows a strong cytoplasmic immunoreactivity score of 3+. B, an intermediate-grade DCIS with a hMena immunoreactivity score of 3+ compared with negative normal ducts (arrow). C to E, three representative cases of invasive ductal carcinomas displaying immunoreactivity scores of 1+ (C), 2+ (D), and 3+(E) respectively. F, a metastatic lymph node displaying a strong hMena immunostaining score of 3+ with juxtamembrane reinforcement in the neoplastic cells. Normal lymphocytes are completely negative. Magnifications, ×20 (A and B); ×40 (C-F).

Close modal

In invasive carcinomas, hMena immunohistochemical reactivity, independently of the score and histotype, was detected in 152 of 163 (93%) tumors. In the ductal type, a score 2+/3+ was observed in 106 of 148 cases (72%), whereas in the lobular and papillary type in 8 of 15 (53%).

In the tumors displaying a 3+ score, a juxtamembrane reinforcement of the staining was observed in 1 of 24 (4%) high-grade DCIS and in 24 of 55 (44%) invasive carcinomas, respectively. Figure 3C to E shows representative cases of the different hMena scores observed in invasive breast tumors.

When we analyzed the relationship between hMena and conventional pathologic variables in invasive breast cancer, a significant correlation was found with tumor size. In fact, as reported in Table 5, hMena-negative tumors were uniformly distributed in T1, T2, and T3/4 carcinomas, whereas protein overexpression (score 3+) was more frequent in the T3,T4 lesions (66%) compared with the T1 (28%) and T2 (37%) lesions (P = 0.04). In addition, we found hMena expression to be slightly correlated to lymph node involvement (P = 0.06; data not shown).

Table 5.

Relationship between tumor size and hMena score in 163 invasive breast carcinomas

Tumor size
hMena score
0 (%)1+ (%)2+ (%)3+ (%)Total cases
T1 8 (8) 30 (29) 36 (35) 28 (28) 102 
T2 2 (4) 7 (15) 20 (44) 17 (37) 46 
T3,4 1 (7) 1 (7) 3 (20) 10 (66) 15 
Total 11 (7) 38 (23) 59 (36) 55 (34) 163 
Tumor size
hMena score
0 (%)1+ (%)2+ (%)3+ (%)Total cases
T1 8 (8) 30 (29) 36 (35) 28 (28) 102 
T2 2 (4) 7 (15) 20 (44) 17 (37) 46 
T3,4 1 (7) 1 (7) 3 (20) 10 (66) 15 
Total 11 (7) 38 (23) 59 (36) 55 (34) 163 

NOTE: P = 0.04.

hMena expression in breast cancer metastases. As summarized in Table 6, 14 of 15 (93%) synchronous axillary metastases and 25 of 33 (76%) metachronous distant metastases (lung 11, brain 8, pleura 4, ovary 2, liver 2, bone 2, and supraclavicular and mediastinic nodes 4) overexpressed (score 2+/3+) hMena protein. Of interest, 85% of metastatic lesions with a 3+ score presented a juxtamembrane reinforcement of the immunostaining as exemplified in Fig. 3F.

Table 6.

hMena protein expression in 48 breast cancer metastases sampled from different body sites

Site of metastasesNo. metastaseshMena score
0 (%)1+ (%)2+ (%)3+ (%)
Lymph nodes* 19 1 (5) 1 (5) 5 (27) 12 (63) 
Lung 11 2 (18) 2 (18) 3 (27) 4 (36) 
Brain 0 (0) 0 (0) 3 (37) 5 (63) 
Pleura 1 (25) 1 (25) 0 (0) 2 (50) 
Ovary 0 (0) 0 (0) 0 (0) 2 (100) 
Liver 0 (0) 1 (50) 1 (50) 0 (0) 
Bone 0 (0) 0 (0) 2 (100) 0 (0) 
Total cases 48 4 (8) 5 (11) 14 (29) 25 (52) 
Site of metastasesNo. metastaseshMena score
0 (%)1+ (%)2+ (%)3+ (%)
Lymph nodes* 19 1 (5) 1 (5) 5 (27) 12 (63) 
Lung 11 2 (18) 2 (18) 3 (27) 4 (36) 
Brain 0 (0) 0 (0) 3 (37) 5 (63) 
Pleura 1 (25) 1 (25) 0 (0) 2 (50) 
Ovary 0 (0) 0 (0) 0 (0) 2 (100) 
Liver 0 (0) 1 (50) 1 (50) 0 (0) 
Bone 0 (0) 0 (0) 2 (100) 0 (0) 
Total cases 48 4 (8) 5 (11) 14 (29) 25 (52) 
*

Fifteen axillary, 3 supraclavicular, and 1 mediastinic.

Correlation between hMena expression, proliferation index, hormonal receptor status, and HER-2 overexpression. The extent of hMena expression in the 65 DCIS and the 163 invasive breast carcinomas was compared with Ki67, ERα, PgR, and HER-2 status, biomarkers known to be associated to therapy response and patient's survival in this neoplasia (11). As graphically reported in Fig. 4, a significant relationship occurred between the hMena score and the proliferation index evaluated by means of Ki67 antibody both in DCIS (P = 0.01) and in invasive (P < 0.0001) breast carcinomas.

Fig. 4.

Relationship between hMena score and proliferation index in DCIS and invasive breast carcinomas. White and black columns, the percentage of DCIS and invasive carcinomas, respectively, with elevated proliferative index evaluated by means of Ki67 antibody (cutoff >15%) for each hMena score. hMena score and elevated proliferation index are significantly associated both in DCIS (P = 0.01) and in invasive (P < 0.0001) breast carcinomas. *, °, P according to nonparametric Pearson's χ2 test.

Fig. 4.

Relationship between hMena score and proliferation index in DCIS and invasive breast carcinomas. White and black columns, the percentage of DCIS and invasive carcinomas, respectively, with elevated proliferative index evaluated by means of Ki67 antibody (cutoff >15%) for each hMena score. hMena score and elevated proliferation index are significantly associated both in DCIS (P = 0.01) and in invasive (P < 0.0001) breast carcinomas. *, °, P according to nonparametric Pearson's χ2 test.

Close modal

Furthermore, as shown in Fig. 5A, in invasive carcinomas, ERα and PgR positivity was inversely correlated with the hMena score. In fact, although in our series the totality of hMena-negative carcinomas were ERα/PgR positive, this positivity progressively and significantly decreased in the lesions expressing hMena (score 1+, 2+, and 3+) both for ERα (from 100% to 67%, P = 0.036) and for PgR (from 100% to 44%, P = 0.001).

Fig. 5.

Relationship between hMena score, hormonal receptor status and HER-2 overexpression in invasive carcinomas. A, percentage of ERα- positive (white columns) and PgR-positive (black columns) breast cancer for each hMena score. A significant inverse relationship between hMena overexpression (score 1+, 2+, and 3+) and ERα/PgR positivity is observed (*ERα P = 0.036; *PgR P = 0.001). B, percentage of HER-2 negative (white columns) and positive (black columns) breast cancer for each hMena score are reported. A significant lower incidence of HER-2-negative cases is observed when hMena score increases (0, 100%; 1+, 87%; 2+, 82%; and 3+, 58%). Inversely, the percentage of HER-2-positive cases (2+ and 3+) progressively and significantly increases in relationship to the hMena score. *, P < 0.0001 according to nonparametric Pearson's χ2 test.

Fig. 5.

Relationship between hMena score, hormonal receptor status and HER-2 overexpression in invasive carcinomas. A, percentage of ERα- positive (white columns) and PgR-positive (black columns) breast cancer for each hMena score. A significant inverse relationship between hMena overexpression (score 1+, 2+, and 3+) and ERα/PgR positivity is observed (*ERα P = 0.036; *PgR P = 0.001). B, percentage of HER-2 negative (white columns) and positive (black columns) breast cancer for each hMena score are reported. A significant lower incidence of HER-2-negative cases is observed when hMena score increases (0, 100%; 1+, 87%; 2+, 82%; and 3+, 58%). Inversely, the percentage of HER-2-positive cases (2+ and 3+) progressively and significantly increases in relationship to the hMena score. *, P < 0.0001 according to nonparametric Pearson's χ2 test.

Close modal

In contrast, a significant direct correlation between hMena staining intensity and HER-2 status was observed. As illustrated in Fig. 5B, all the hMena-negative carcinomas (100%) were concomitantly HER-2 negative, whereas the percentage of HER-2-negative lesions was significantly and progressively lower in the group of hMena-positive tumors from score 1+ to score 3+ (1+, 87%; 2+, 82%; and 3+, 58%). Inversely, the percentage of HER-2-positive cases (2+ and 3+) progressively and significantly increases in relationship to the hMena score (P < 0.0001).

hMena expression is increased by NRG1 and reduced by Herceptin in vitro treatments. In view of the in vivo significant relationship between hMena and HER-2 overexpression (score 2+/3+), we evaluated whether activation or inhibition of HER-2 may affect hMena expression. To this end, we have treated MCF-7 and BT474 breast cancer cell lines with NRG1. NRG1 (10 ng/mL) treatment was accompanied with a significant increase of the hMena protein level in both cell lines as detected by Western blot (Fig. 6A). Therefore, we have treated the BT474 breast cancer cell line, which overexpresses the 90 kDa hMena and HER-2, with Herceptin at a concentration of 25 or 50 μg/mL. The HER-2 cell surface expression resulted in an ∼60% decrease as measured by cytofluorimetry (Fig. 6B). This effect, which was present at 24 hours and further increased at 48 hours of treatment, was paralleled by a decrease of hMena expression as measured by Western blot (Fig. 6C). Similar results were obtained by treating the SKBr3 breast cancer cell line with Herceptin, showing hMena expression level comparable with that of BT474 cells (data not shown). These results place hMena expression downstream of tyrosine kinase receptors.

Fig. 6.

hMena expression is increased by NRG1 treatment and reduced by Herceptin in breast cancer cell lines. A, BT474 and MCF-7 cells were serum-starved for 24 hours and stimulated with 10 ng/mL NRG1 for 12-24-48 hours. Lysates (30 μg) were analyzed by Western blot with the indicated antibodies. B, cytofluorimetric analysis of BT474 breast tumor cells untreated or treated for 24, 48, or 72 hours with the indicated amount of Herceptin (HRC). Staining was done using anti-HER-2 mAb (W6/100, 10 μg/mL) followed by FITC-conjugated goat anti-mouse. Columns, mean of fluorescence intensity for each sample corrected by subtracting background fluorescence given by binding of FITC-labeled isotype-matched control IgG; bars, SD for four independent experiments. C, evaluation of HER-2 and hMena expression in BT474 breast tumor cells untreated or treated for 24-48-72 hours with 50 μg/mL Herceptin by Western blot. Lysates (30 μg) were immunoblotted with the indicated antibodies.

Fig. 6.

hMena expression is increased by NRG1 treatment and reduced by Herceptin in breast cancer cell lines. A, BT474 and MCF-7 cells were serum-starved for 24 hours and stimulated with 10 ng/mL NRG1 for 12-24-48 hours. Lysates (30 μg) were analyzed by Western blot with the indicated antibodies. B, cytofluorimetric analysis of BT474 breast tumor cells untreated or treated for 24, 48, or 72 hours with the indicated amount of Herceptin (HRC). Staining was done using anti-HER-2 mAb (W6/100, 10 μg/mL) followed by FITC-conjugated goat anti-mouse. Columns, mean of fluorescence intensity for each sample corrected by subtracting background fluorescence given by binding of FITC-labeled isotype-matched control IgG; bars, SD for four independent experiments. C, evaluation of HER-2 and hMena expression in BT474 breast tumor cells untreated or treated for 24-48-72 hours with 50 μg/mL Herceptin by Western blot. Lysates (30 μg) were immunoblotted with the indicated antibodies.

Close modal

It is becoming increasingly documented that changes of the cellular actin network represent a common effector event in malignant transformation and tumor progression (15). In fact, experimental evidences indicate that both processes are under the regulation of actin signaling proteins (i.e., Ras-family small GTPase Rho, Rac, and CDC42), which are components of oncogenic relevant signal transduction pathways (16).

In human breast cancer, numerous alterations in actin-binding proteins have been identified by proteomic scrutiny also in preinvasive in situ ductal carcinoma (7), demonstrating that remodeling of actin cytoskeleton represents an early event in breast carcinogenesis.

Through the analysis of the antibody repertoire specific for autologous breast cancer, we have previously identified hMena, a key regulatory protein of actin cytoskeleton dynamics. Because preliminary phenotypic analysis has shown that high levels of hMena are present in breast tumors of various histotype, in the present study we have tested the hypothesis whether this molecule may be an early marker of breast carcinogenesis and of tumor progression. With this aim, we have evaluated hMena expression in normal breast and in a wide spectrum of breast disorders, including benign, premalignant lesions, in situ, invasive, and metastatic carcinomas.

Collectively from our analysis, the following conclusions can be drawn: (a) hMena overexpression is an early event in breast cancer development, being undetectable in lesions with NR of transformation, although expressed in proliferative lesions with high risk of transformation, in DCIS and invasive carcinoma. (b) hMena up-regulation, which is present in over 90% of metastatic foci, correlates with established prognostic indicators of human breast cancer (i.e., proliferation index, HER-2 overexpression, and hormonal receptors status; ref. 17).

The increased use of mammography and fine needle aspiration has contributed greatly to the understanding that breast cancer develops through a well-defined, although not obligatory, sequence of histologic benign changes of normal epithelium (18). Still, the criteria to critically assess the risk of these lesions to turn into cancer rely on histologic classification (10).

In this study, we show through immunohistochemistry that hMena is not expressed in normal mammary epithelium and in the majority of benign lesions lacking association with risk of transformation with the exception of a faint staining in a small number of adenosis that were consistently associated with more complex glandular alterations. Of interest, atypical hyperplasia or atypical papilloma, two classic high-risk lesions, are hMena positive in 66% of the cases.

A critical component for the improvement of diagnostic and preventive interventions is the identification of molecular signatures in cells which are progressing to the first detectable stage of cancer (10). In this regard, the results of this study strongly suggest that hMena overexpression could represent an early marker of transformation and justify prospective studies to firmly establish this issue.

The analysis of a large cohort of DCIS and invasive carcinoma has shown a homogeneous and intense cytoplasmic staining in 58% of DCIS and in >70% of the invasive ductal carcinomas. These results, which confirm our previous work conducted in a smaller series of malignant tumors (2), are in agreement with the results of the proteomic analysis done by Wulfkuhle et al. (7), revealing alterations of actin-binding protein expression in DCIS. The finding that hMena is overexpressed in these early neoplastic lesions, which lack a motility signature, suggests that actin cytoskeleton regulatory proteins may also have a role in proliferation processes.

Although the molecular mechanisms connecting proliferation index and hMena overexpression still have to be elucidated, this link is suggested by the finding that hMena overexpression is present in a number of benign proliferative lesions and by its significant correlation with the proliferation index of DCIS and invasive tumors. Because major signal transduction pathways are involved in actin signaling (15), it is conceivable that hMena overexpression may be induced by mitogenic stimuli, contributing to cell proliferation through the rearrangement of actin cytoskeleton. The intense staining of hMena, often with a juxtamembrane reinforcement in synchronous and metachronous metastasis, is in line with data reported in murine breast tumor in which Mena overexpression characterizes the tumor motility molecular signature elicited by epidermal growth factor (9) and those cells in the primary cancer that are capable of penetrating blood vessel and forming metastasis (8). These findings were not unexpected because Mena protein, as well as the other members of the Ena/VASP family, localize to the focal adhesions, lamellipodial protrusions, and filopodia tips, where they control highly dynamic processes requiring a strict regulation of the actin cytoskeleton changes, such as fibroblast protrusion and migration (3), intracellular bacterial motility (19), and axon and dendritic guidance (20). A recent study on the gene expression signature in HER-2/neu-transgenic mice, a suitable model for mammary carcinogenesis, has shown Mena (ENAH) overexpression during HER-2-induced breast cancer progression (21), suggesting that signaling pathways related to different epidermal growth factor receptor family members may also modulate hMena. In line with these data, our results show a significant direct relationship between hMena and HER-2 overexpression. The high frequency of hMena up-regulation in primary breast tumors, on the other hand, clearly suggests that other pathways concur with the extent of hMena overexpression. HER-2 overexpression and lack of hormone receptors in invasive breast cancer are well-established predictors of the patient's poor outcome and response to treatment (11, 17). The strong association between hMena overexpression and these biological variables indicates that hMena protein expression may influence the biological behavior of breast tumors.

The observation that Mena is up-modulated in murine breast cancer cells by epidermal growth factor (9) and the present in vivo findings of the significant correlation between hMena and HER-2 levels suggest that hMena may couple tyrosine kinase receptor signaling to cytoskeleton. The experimental ground to this hypothesis is provided by our results on in vitro NRG1 and Herceptin treatments aimed to stimulate or reduce the HER-2 activity. In fact, the NRG1 factor up-regulates hMena in MCF-7 and BT474 breast cancer cell lines in agreement with results demonstrating that NRG1 by stimulating HER-2 activity induces breast cancer proliferation and progression (22) concomitantly with actin cytoskeletal changes (23). On the other hand, the Herceptin in vitro treatment of HER-2-overexpressing breast cancer cells, which results in a dose-dependent reduction of HER-2 expression, also down-modulates hMena.

Whether this effect is mediated through the two main pathways of HER-2 downstream signaling, namely mitogen-activated protein kinase and phosphatidylinositol 3-kinase/AKT remains to be elucidated.

In conclusion, the present study, aimed at evaluating the biological role of the recently described hMena protein in breast cancer, has shown that overexpression of this actin-binding molecule is associated to an increased risk of transformation as well as tumor progression, thus offering novel means of biological and clinical assessment of breast cancer.

Grant support: Italian Ministry of Health, Associazione Italiana per la Ricerca sul Cancro, Lega Italiana per la Lotta Contro i Tumori, Ministero dell'Istruzione, dell'Università e della Ricerca-Fondo per gli Investimenti della Ricerca di Base.

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.

Note: Supplementary data for this article are available at Clinical Cancer Research online (http://clincancerres.aacrjournals.org/).

F. Di Modugno and M. Mottolese contributed equally to this work.

We thank Dr. G. Zupi for her continuous support and helpful discussion, and P.I. Franke and M.V. Sarcone for the formal revision of the manuscript and for the secretarial assistance.

1
Dunn GP, Bruce AT, Sheehan KC, et al. A critical function for type I interferons in cancer immunoediting.
Nat Immunol
2005
;
6
:
722
–9.
2
Di Modugno F, Bronzi G, Scanlan MJ, et al. Human Mena protein, a SEREX-defined antigen overexpressed in breast cancer eliciting both humoral and CD8+ T-cell immune response.
Int J Cancer
2004
;
109
:
909
–18.
3
Bear JE, Svitkina TM, Krause M, et al. Antagonism between Ena/VASP proteins and actin filament capping regulates fibroblast motility.
Cell
2002
;
109
:
509
–21.
4
Barzik M, Kotova TI, Higgs HN, et al. Ena/VASP proteins enhance actin polymerization in the presence of barbed end capping proteins.
J Biol Chem
2005
;
280
:
28653
–62.
5
Krause M, Dent EW, Bear JE, Loureiro JJ, Gertler FB. Ena/VASP proteins: regulators of the actin cytoskeleton and cell migration.
Annu Rev Cell Dev Biol
2003
;
19
:
541
–64.
6
Vasioukhin V, Bauer C, Yin M, Fuchs E. Directed actin polymerization is the driving force for epithelial cell-cell adhesion.
Cell
2000
;
100
:
209
–19.
7
Wulfkuhle JD, Sgroi DC, Krutzsch H, et al. Proteomics of human breast ductal carcinoma in situ.
Cancer Res
2002
;
62
:
6740
–9.
8
Yang J, Mani SA, Donaher JL, et al. Twist, a master regulator of morphogenesis, plays an essential role in tumor metastasis.
Cell
2004
;
117
:
927
–39.
9
Wang W, Goswami S, Lapidus K, et al. Identification and testing of a gene expression signature of invasive carcinoma cells within primary mammary tumors.
Cancer Res
2004
;
64
:
8585
–94.
10
Hartmann LC, Sellers TA, Frost MH, et al. Benign breast disease and the risk of breast cancer.
N Engl J Med
2005
;
353
:
229
–37.
11
Konecny G, Pauletti G, Pegram M, et al. Quantitative association between HER-2/neu and steroid hormone receptors in hormone receptor-positive primary breast cancer.
J Natl Cancer Inst
2003
;
95
:
142
–53.
12
Fitzgibbons PL, Henson DE, Hutter RV. Benign breast changes and the risk for subsequent breast cancer: an update of the 1985 consensus statement. Cancer Committee of the College of the American Pathologists.
Arch Pathol Lab Med
1998
;
124
:
966
–78.
13
Tavassoli FA, Devilee P. Tumors of the breast. In: Tavassoli FA, Devilee P. Pathology and genetics. Tumours of the breast and female genital organs. Chapter 1. Lyon (France): IARC Press; 2003. p. 9–112.
14
Digiesi G, Giacomini P, Fraioli R, et al. Production and characterization of murine mAbs to the extracellular domain of human neu oncogene product GP185HER-2.
Hybridoma
1992
;
11
:
519
–27.
15
Rao JY, Li N. Microfilament actin remodeling as a potential target for cancer drug development.
Curr Cancer Drug Targets
2004
;
4
:
345
–54.
16
Lu QY, Jin YS, Pantuck A, et al. Green tea extract modulates actin remodeling via Rho activity in an in vitro multistep carcinogenic model.
Clin Cancer Res
2005
;
11
:
1675
–83.
17
Bast RC, Jr., Ravdin P, Hayes DF, et al. 2000 Update of recommendations for the use of tumor markers in breast and colorectal cancer: clinical practice guidelines of the American Society of Clinical Oncology.
J Clin Oncol
2001
;
19
:
1865
–78.
18
Ellsworth DL, Ellsworth RE, Liebman MN, Hooke JA, Shriver CD. Genomic instability in histological normal breast tissues: implications for carcinogenesis.
Lancet Oncol
2004
;
5
:
753
–8.
19
Auerbuch V, Loureiro JJ, Gertler FB, Theriot JA, Portnoy DA. Ena/VASP proteins contribute to Listeria monocytogenes pathogenesis by controlling temporal and spatial persistence of bacterial actin based motility.
Mol Microbiol
2003
;
49
:
1361
–75.
20
Menzies AS, Aszodi A, Williams SE, et al. Mena and vasodilator-stimulated phosphoprotein are required for multiple actin-dependent processes that shape the vertebrate nervous system.
J Neurosci
2004
;
24
:
8029
–38.
21
Astolfi A, Landuzzi L, Nicoletti G, et al. Gene expression analysis of immune-mediated arrest of tumorigenesis in a transgenic mouse model of HER-2/neu-positive basal-like mammary carcinoma.
Am J Pathol
2005
;
166
:
1205
–16.
22
Stove C, Bracke M. Roles for neuregulins in human cancer.
Clin Exp Metastasis
2004
;
21
:
665
–84.
23
Adam L, Vadlamudi R, Kondapaka SB, Chernoff J, Mendelsohn J, Kumar R. Heregulin regulates cytoskeletal reorganization and cell migration through the p21-activated kinase-1 via phosphatidylinositol-3 kinase.
J Biol Chem
1998
;
273
:
28238
–46.

Supplementary data