Cyclooxygenase-2 Expression in Human Breast Cancers and Adjacent Ductal Carcinoma in Situ

Tissue Samples.

We analyzed a nonconsecutive series of primary human breast carcinomas (n = 57), adjacent DCIS (n = 14) or DCIS alone (n = 2), and normal-appearing breast epithelia (n = 41) that were obtained from the surgical pathology laboratory at The University of Texas M. D. Anderson Cancer Center. Tissue specimens were formalin fixed, paraffin embedded, and represented different tumor stages [stage 1 (n = 15), stage 2 (n = 24), stage 3 (n = 12), stage 4 (n = 1), or unknown (n = 3)], according to the TNM classification (45) . Consecutive 4–6-μm tissue sections were cut from paraffin blocks and placed on polylysine-coated slides for IHC analysis. An additional nine pairs of frozen, invasive breast cancer and normal tissue samples were obtained from the Institutional Tissue Bank.

Tumor Cell Lines.

We analyzed the human breast adenocarcinoma cell lines MDA-MB-231, SK-BR-3, MCF-7, and MCF-7/HER2. MCF-7 cells had been stably transfected with the HER2 cDNA (gift from Dr. Mien Chi Hung, University of Texas M. D. Anderson Cancer Center). The HCA-7 colorectal carcinoma cell line was obtained from Dr. S. Kirkland (Imperial Cancer Research Fund, London, United Kingdom), and the HCT116 colon cancer cell line was from the American Type Culture Collection (Rockville, MD). All cells were grown in DMEM:F-12 (Life Technologies, Inc., Frederick, MD) supplemented with heat-inactivated 10% FCS and 10 units/ml penicillin and streptomycin. Cultured cells were maintained in a humidified environment at 37°C with 5% CO₂.

IHC.

Slides were deparaffinized, and endogenous peroxidase activity was blocked by incubation in 3% H₂O₂ in methanol for 30 min at room temperature. Sections were then microwaved in PBS (pH 7.4) for 4 min for antigen retrieval. Immunoreactivity was detected by an immunoperoxidase method (Vectastain Elite ABC; Vector Laboratories, Burlingame, CA). Nonspecific binding was blocked with avidin then biotin (Vector) for 15 min each. The primary MAb against human COX-2 (Cayman Chemical, Ann Arbor, MI) recognizes a 19-amino acid sequence at the COOH terminus that is absent in COX-1. This antibody was applied at a dilution of 1:500 for 2 h at room temperature. After slides were rinsed in PBS, the biotinylated secondary IgG antibody was applied for 30 min at room temperature. Slides were rinsed in PBS, and avidin conjugated to horseradish peroxidase (ABC reagent) was applied for 45 min at room temperature. The chromogen 3,3′-diaminobenzidine (Research Genetics, Huntsville, AL) was subsequently added, and the color reaction was observed with light microscopy. The reaction was stopped by immersing slides in deionized water; slides were then counterstained with hematoxylin and mounted.

As a control for nonspecific staining, a COX-2-specific blocking peptide (Cayman Chemical) derived from the human COX-2 cDNA sequence was mixed with the COX-2 MAb and applied to sections. In all cases, immunoreactivity was completely suppressed. As a negative control, the primary MAb was omitted and PBS was applied, with all other steps of the procedure included. The HCA-7 colon cancer cell line was used as a positive control for COX-2.

For HER2 immunostaining, a modification of the procedure used for COX-2 was used. Antigen retrieval was performed by microwave treatment of slides in citrate buffer (pH 6.0) for 4 min on the high setting and an additional 6 min at lower power. Slides were incubated with a monoclonal anti-c-erbB-2/HER2 antibody (clone e2-4001+B5; Neomarkers, Fremont, CA) at a 1:200 dilution. A series of breast cancers were stained with this MAb, and the results were compared with these same cases stained with another anti-c-erbB-2/HER2 MAb (Oncogene Science, Manhasset, NY). Identical results were found (data not shown). A human breast cancer with HER2 overexpression and amplification was included with all slide runs. Additional controls included SK-BR-3 and MDA-MB-231 cells. Negative control slides omitted the primary antibody but included all other steps of the procedure.

IHC Scoring.

For COX-2, weighted score was computed that represented the product of percentage of tumor cell positivity and intensity, as described previously (46) . Percent tumor cell positivity was categorized as follows: 0, <5%; 1, 5–25%; 2, 26–50%; 3, 51–75%, and 4, >75%. Staining intensity was graded as absent (0), weak (1+), medium (2+), or strong (3+). The intensity of the invasive tumor and of DCIS (when it existed on the same slide) were determined separately. On the basis of weighted scores, expression was categorized as negative or low (0–4), intermediate (5, 6, 7, 8) , or high (9, 10, 11, 12) . Localization of HER2 staining was described as cytoplasmic or membranous. Cases considered positive for HER2 had membranous staining in at least 10% of tumor cells and intensity scored as 0–3+. Before the study specimens were scored, representative sections displaying each of the four staining intensities were reviewed by two examiners.

Immunoblotting.

Frozen tissue samples were homogenized in Tris-HCl buffer (pH 7.4) containing 0.5% NP40 and protease inhibitors (Boehringer Mannheim, Indianapolis, IN). Cultured cell lines were processed according to standard procedures, and a cell lysate was prepared. Protein extraction was performed on tissue and cell lysates, and the protein concentration was determined in soluble supernatants in each sample. We then loaded 40 μg of protein per lane onto a SDS-polyacrylamide gel, and proteins were then transferred onto an Immobilon membrane (Millipore, Bedford, MA). Membranes were blocked with 5% nonfat dry milk. Blots were incubated overnight at 4°C with MAb against COX-2 (Cayman Chemical) at a dilution of 1:1000 and actin (Oncogene Science, Uniondale, NY). The level of protein expression was measured with the enhanced chemiluminescence detection system (Amersham, Arlington Heights, IL) according to the manufacturer’s instructions. Actin levels were analyzed as controls for protein loading.

RT-PCR.

Total RNA was extracted and purified from human breast tumors and paired normal tissues using the RNeasy mini Kit (Qiagen, Valencia, CA) according to the manufacturer’s instructions. Total RNA (1 μg) was converted to cDNA by Superscript^II reverse transcriptase (Life Technologies, Inc., Rockville, MD) in a total reaction volume of 20 μl. The cDNA was PCR-amplified with AmpliTaq DNA polymerase (Perkin-Elmer/Applied Biosystems, Foster City, CA). Relative expression of COX-2 mRNA was analyzed using quantitative multiplex RT-PCR with β-actin as an internal control. The reaction mixture contained 1 unit of Hotstart polymerase (Qiagen, Hilden, Germany), 10 mm deoxynucleotide triphosphates, 1 μl of DMSO, 50 ng of forward (F) and reverse (R) primers, and 1 μl of cDNA in a final volume of 10 μl. The primer sequences and PCR product sizes were as follows: for COX-2, (F) 5′-GCTGAGCCATACAGCAAATCC-3′, (R) 5′-GGGAGTCGGGCAATCATCAG-3′ (386-bp product); for β-actin, (F) 5′-GTTGCTATCCAGGCTGTGC-3′ and (R) 5′-GCATCCTGTCGGCAATGC-3′ (540-bp product). The reaction was performed according to the following program: 15 min at 94°C for activating the polymerase, 38 cycles (28 cycles for β-actin) at 94°C for 30 s for denaturation, 1 min at 61°C for annealing, and extension for 1 min at 72°C, followed by an additional extension step at 72°C for 7 min. The PCR products were run on a 2.0% agarose gel and then stained with 0.5 μg/ml ethidium bromide. Stained bands on the gel were visualized under UV light and then photographed.

Patient Characteristics.

We evaluated archival tissue specimens from 57 women with invasive breast cancer, as shown in Table 1 ⇓ . Age at diagnosis ranged from 25 to 78 years (median age, 47 year), and 36 (63%) women were less than age 50. Eighty percent of the tumors were infiltrating ductal adenocarcinomas. Adjacent DCIS was found in 13 of the 57 cases. Two patients had exclusively DCIS and no invasive cancer. Patient characteristics, including tumor stage, nuclear grade, hormonal receptor status, and preoperative treatment history, are outlined in Table 1 ⇓ . At a median follow-up of 19 months (range, 1–86 months), all but two patients were alive.

COX-2 Immunoreactivity.

Cytoplasmic COX-2 expression was detected at an intermediate or high level by a semiquantitative IHC method in neoplastic epithelial cells in 18 of 42 (43%) primary breast cancers (Table 2) ⇓ . Negative or low-level staining for COX-2 was observed in 57% of these invasive tumors. COX-2 was detected at an intermediate or high level in 10 of 16 (62.5%) cases of DCIS adjacent to cancer. The two cases of DCIS alone showed an intermediate level of COX-2. COX-2 staining was granular and localized to the tumor cytoplasm (Fig. 1) ⇓ . COX-2 expression was less heterogeneous in DCIS cases than in invasive cancers. Within the same tissue sections, COX-2 expression in invasive breast tumors and adjacent DCIS were highly correlated (P = 0.019). COX-2 was expressed in benign ductular epithelial cells adjacent to tumor in 81% of cases (Table 2) ⇓ . Staining in normal-appearing epithelia was often focal and of similar or reduced intensity relative to neoplastic epithelia. Stromal staining for COX-2 in fibroblasts and other interstitial cells was infrequently detected. Specificity of staining was confirmed by use of a COX-2 sequence-specific blocking peptide that completely suppressed staining in all cases.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

A and B, COX-2 staining is detected in the cytoplasm of tumor cells within DCIS. Normal breast ductular epithelium (dotted arrow) shows reduced intensity of COX-2 staining relative to DCIS (solid arrow). C, invasive breast carcinoma lacking COX-2 staining. D, infiltrating ductal adenocarcinoma of the breast with a high level of COX-2 expression. E, membranous HER2 staining (3+) in an infiltrating ductal breast carcinoma. F, SK-BR-3 breast cancer cells overexpress HER2 in the cell membrane in contrast to MDA-MB-231 cells (inset).

COX-2 mRNA by RT-PCR.

COX-2 mRNA expression was analyzed in paired tumor and normal tissue specimens from breast cancer patients (n = 9), using COX-2 specific primers and RT-PCR. As shown in Fig. 2 ⇓ , COX-2 was detected in tumor and normal breast tissue from all nine patients. The level of COX-2 mRNA in tumors was shown to exceed that in normal tissues in eight of these paired specimens from the same patients.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

RT-PCR analysis of COX-2 mRNA expression in infiltrating ductal adenocarcinomas of the breast. Total RNA was extracted from tumors (T) and their paired normal (N) tissues. PCR product sizes (in bp) are 386 for COX-2 and 540 for β-actin; Lane M, DNA markers.

HER2/neu Expression and Gene Amplification.

HER2 was analyzed by IHC in all breast cancer cases. Membranous staining of 3+ intensity was detected in 17 of 55 (31%) invasive breast cancers (Fig. 1E) ⇓ , and was 2+ in 4 and 1+ in 5 cases. Results for HER2 gene amplification by FISH were available on 20 cases. Five of 7 (72%) tumors with 3+ membranous staining and all 2+ tumors (n = 4) were amplified. HER2 gene amplification was not found in the cases that stained 1+ or were negative for HER2. A significant correlation was observed between membranous HER2 staining and gene amplification (P = 0.003). Cytoplasmic HER2 staining was detected in 15 of 55 (27%) cancers; 23 (42%) cases were negative for HER2. Cytoplasmic staining was granular, and the intensity was 2+ or 3+. HER2 was not amplified in these cytoplasmic or the negative cases.

Association of COX-2 with Clinicopathological Variables.

COX-2 protein expression in invasive breast cancers or adjacent DCIS did not significantly correlate with any of the clinicopathological variables examined, including hormonal status. No difference in the frequency of COX-2 expression was found in tumors from patients receiving neoadjuvant therapy (n = 16) versus untreated patients (P = 0.16). No association was found between COX-2 and HER2 immunostaining or gene amplification.

COX-2 in Breast Carcinoma Cell Lines.

Human breast cancer cell lines (SK-BR-3, MDA-MB-231, and MCF-7) were analyzed for COX-2 and HER2 expression by immunoblotting. We studied the estrogen-dependent, poorly invasive, and nonmetastatic MCF-7 cell line (47) and the estrogen-independent, highly invasive, and metastatic MDA-MB-231 cell line (48) . SK-BR-3 cells are less invasive and estrogen dependent (49) . SK-BR-3 and MCF-7 cells were found to lack COX-2 proteins, and MDA-MB-231 cells displayed a low level of COX-2 (Fig. 3, A and B) ⇓ . To determine whether HER2 can regulate the expression of COX-2, we analyzed MCF-7 cells stably transfected with the HER2. cDNA MCF-7/HER2 transfectants were found to express COX-2 in contrast to parental cells, suggesting induction of COX-2 by HER2 (Fig. 3B) ⇓ . As shown previously, SK-BR-3 cells overexpressed HER2 in the cellular membrane (Fig. 1F) ⇓ and displayed HER2 amplification (Fig. 3A ⇓ ; Ref. 49 ); MDA-MB-231 cells expressed a low level of HER2 proteins (Ref. 48 ; Fig. 3A ⇓ ). The HCA-7 colorectal cancer cell line is known to constitutively express COX-2 and was used as a positive control (50) . The COX-2-negative HCT116 colon cancer cell line (8) was used as a negative control for COX-2 protein expression. HCA-7 cells were found to overexpress COX-2 and HER2 proteins (Fig. 3A) ⇓ . HCT116 cells lacked COX-2 but expressed HER2 proteins (Fig. 3A) ⇓ . Immunostaining of these cell lines was concordant with immunoblotting results (data not shown). Neither HCA-7 nor HCT116 cells had HER2 amplification (Fig. 3A) ⇓ .

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

A, immunoblot of COX-2 (70 kDa) and HER2 (185 kDa) protein levels in human breast (SK-BR-3 and MDAMB-231) and colorectal (HCA-7 and HCT-116) carcinoma cell lines. HCA-7 cells constitutively express COX-2 (positive control) and HCT116 cells lack COX-2 proteins (negative control). HER2 gene amplification results, determined by fluorescence in situ hybridization (FISH), are shown for these same cell lines. B, MCF-7 breast cancer cells stably transfected with the HER-2 cDNA show induction of COX-2 proteins in contrast to parental MCF-7 cells. C, Densitometry analysis of COX-2 protein expression relative to β-actin in MCF-7/HER2-transfected cells compared with parental MCF-7 cells.