Surgically resected 211 breast cancer tissues were examined in the present study. All cases underwent surgical resection at Fukushima Medical University between 2002 and 2005. Information regarding age, TNM stage (the 7th classification), ER, PgR and HER2, and pathological diagnosis including lymphatic and venous invasion, were retrospectively collected. The patient median age was 55 years. These cases included 85 stage I cases, 120 stage II cases and 6 stage III cases. The present study was approved by the Ethics Committee of Fukushima Medical University. Written informed consent was obtained from all participants.

Formalin-fixed paraffin-embedded (FFPE) sections were immunostained for ANXA1, ER, PgR and HER2, and evaluated for staining intensity. Briefly, breast cancer specimens were fixed in a 10% formalin and embedded in paraffin. They were cut into thin (4 μm) sections and stained with hematoxylin and eosin or other antibodies. Analysis of ER, PgR and HER2 was performed by the Ventana HX automatic system BenchMark (Ventana Medical Systems SA, Illkirch Cedex, France). Antibodies used for IHC staining were as follows: anti-ANXA1 (clone 29; BD Biosciences, San Jose, CA, USA), anti-ER (SP1), anti-PgR (1E2), anti-HER2 (4B5) (all from Ventana Medical Systems SA). The expression of ANXA1 protein was evaluated by the ratio of the numbers of positively stained cells to those negatively stained. Tissues were dichotomized as positive (≥5% staining) or negative (<5% staining) for ANXA1. ER, PgR and HER2 expression levels were also dichotomized as positive or negative as previously described (28).

The breast cancer cell lines were originally obtained from the American Type Culture Collection (Rockville, MD, USA) and were cultured in the recommended media with 10% fetal bovine serum. These monolayer cells were maintained in a 37°C incubator with 5% CO₂. Cells were checked regularly under a light microscope and were subcultured when reaching 80 to 90% confluence. For hypoxia exposure, each cell type was cultured for 24 h in a modulator incubator chamber (Billups-Rothenberg, Del Mar, CA, USA) at 37°C with 1% O₂, 5% CO₂ and 94% N₂. To mimic hypoxia, the cells were cultured for 24 h with 100 μM cobalt chloride (CoCl₂) (Sigma-Aldrich, St. Louis, MO, USA).

Cells were washed twice with ice-cold PBS and immediately scraped. The cells were then centrifuged at 15,000 rpm to pellet cellular debris and were stored overnight at −80°C. The protein lysates were collected using ice-cold RIPA buffer containing Halt Protease Inhibitor Single-Use Cocktail (Thermo Scientific, San Jose, CA, USA). The protein concentration was determined by the Bradford method (Bio Rad, Hercules, CA, USA). Samples for SDS-PAGE were prepared by mixing aliquots of the protein with Novex Tris-Glycine SDS sample buffer (Invitrogen, Carlsbad, CA, USA) and heated at 100°C for 3 min. Protein samples were run on 10% Bis-Tris gels at 100 V for 90 min with MES SDS running buffer (Invitrogen). For western blotting, the gels were electro-transferred to a nitrocellulose membrane using the iBlot Dry Blotting System (Invitrogen). The proteins were blocked using Starting Block (PBS) Blocking Buffer (Pierce, Rockford, IL, USA) and were detected using anti-ANXA1, anti-hypoxia-inducible factor-1α (HIF-1α) (BD Novus Bioscience Biologicals, Littleton, CO, USA), anti-β-actin (Santa Cruz Biotechnology, Dallas, TX, USA) and a goat anti-mouse secondary antibody phosphatase (Novagen, Billerica, MA, USA). Western blottings were then incubated with Super Signal West Pico detection system (Pierce) and detected using LAS-4000 IR MultiColor (Fuji, Tokyo, Japan).

The total RNA was extracted from the cells using TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. Complementary DNA (cDNA) was synthesized from 5 μg of total RNA with a random hexamer using the SuperScript III First-Strand Synthesis kit (Invitrogen). These cDNAs were used for the measurement of gene expression with a 7500 Real-time PCR system (Applied Biosystems, Foster City, CA, USA) using TaqMan probes, and experiments were performed in triplicate to the blinded patient information. β-actin was used as an internal control. Expression assays were purchased from Applied Biosystems: ANXA1 (Hs00167549_ m1) and β-actin (Hs99999903_ m1). Normalized ANXA1 gene expression was calculated using the 2^−ΔΔCT method as previously described (29).

The siRNA oligonucleotides for ANXA1 (HSS100502 and HSS100503) and the control were purchased from Invitrogen. ANXA1 siRNA sequences were designed as follows: siRNA-1, 5′-CAACCAUCAUUGACAUUCUAACUAA-3′ and siRNA-2, 5′-GCCUUGCAUAAGGCCAUAAUGGUUA-3′. One day prior to the siRNA transfection, 5×10⁵ cells were plated in 6 cm plates. The cells were transfected with 40 nM of each siRNA using Lipofectamine RNAiMAX (Invitrogen) following the manufacturer’s instructions and harvested after 48 h.

The cell invasion assay was conducted using a 24-well BD Bio-Coat Tumor Invasion System (BD Biosciences). Cells were harvested and seeded (5×10³ cells) into the top chamber and incubated at 37°C for 22 h. The cells were post-stained with 4 μg/ml of Calcein-AM (Molecular Probes, Carlsbad, CA, USA) in Hank’s buffered salt solution at 37°C in 5% CO₂ for 1 h. The labeled cells that invaded the BD Matrigel Matrix and passed through the pores of the BD FluoroBlok membrane were detected by Varioskan Flash Multimode Reader (Thermo Scientific).

Statistical analysis was carried out with an unpaired Student’s t-test using GraphPad Prism v5.0 (GraphPad Software, Inc., La Jolla, CA, USA). P<0.05 was considered to indicate a statistically significant result.

The ANXA1 protein expression in the breast cancer tissues was examined by IHC staining using an anti-ANXA1 antibody. In the breast cancer tissues, the ANXA1 expression was mainly detected in the cytoplasm of tumor cells, interstitial cells and myoepithelial cells (Fig. 1). In contrast, in the adjacent normal breast tissues, the ANXA1 expression was detected in myoepithelial cells, but not in normal duct-epithelial cells.

Of the 211 cases, 31 (14.7%) showed positive ANXA1 staining in the breast cancer tissues. Such cases showed a positive correlation with triple-negative (ER-negative, PgR-negative and HER2-negative) expression (P=0.007) and venous invasion (P=0.028) (Table I). However, such cases did not show any correlation with age, TNM stage, lymph node metastasis, histology, ER and PgR expression and recurrence. These results suggest that positive expression of ANXA1 is associated with tumor cell invasion in breast cancer, which was particularly significant in the TNBC cases.

We next examined the ANXA1 expression in five human breast cancer cell lines, MCF-7, T-47D, SK-Br-3, ZR-75-1 and MDA-MB-231. IHC staining using the anti-ANXA1 antibody was performed for those cell lines. Positive staining of ANXA1 was detected in the cytoplasm of the MDA-MB-231 cells, while that of ANXA1 was not detected in the other four cell lines (Fig. 2A). We also performed IHC staining and evaluation for ER, PgR and HER2 in the same breast cancer cell lines. Positive staining of both ER and PgR was observed in the MCF-7 and T-47D cells. Strong positive staining of HER2 was detected in the SK-Br-3 cells and weak positive staining of HER2 was detected in the ZR-75-1 cells. As a result, the MDA-MB-231 cell line was found to be a TNBC with ANXA1-positive expression.

To confirm the ANXA1 expression by IHC evaluation in the cancer cell lines, we performed western blotting and qRT-PCR using the same breast cancer cell lines. Consistent with the IHC results, the ANXA1 protein was highly expressed in the MDA-MB 231 cells (Fig. 2B). The mRNA of ANXA1 was also highly expressed in MDA-MB-231 cells, suggesting that the ANXA1 expression was upregulated at the transcriptional level (Fig. 2C).

According to the comparisons of our IHC results and the clinicopathological factors, we hypothesized that ANXA1 is upregulated in TNBCs and is associated with cellular invasion. To investigate this hypothesis, we used gene knockdown technology and cell invasion assay. Both protein and mRNA expression of ANXA1 was downregulated by two independent siRNA oligonucleotides (siRNA-1 and -2) in the MDA-MB-231 cells which originally exhibited upregulated ANXA1 (Fig. 3A and B). While no morphological changes were observed in the ANXA1 knockdown cells, the ability of cellular invasion was attenuated in the ANXA1 downregulated cells. The invasion ratio was significantly decreased in the cells treated by both ANXA1 siRNA-1 (P<0.05) and siRNA-2 (P<0.01) (Fig. 3C).

To further assess the induction of ANXA1 during cancer progression, we evaluated ANXA1 induction under hypoxic conditions, which is one of the main characteristic features of a tumor. T-47D and SK-Br3 breast cancer cell lines were treated with hypoxia and the hypoxia mimic induced by CoCl₂. HIF-1α protein was induced in both cell lines (Fig. 4A). The ANXA1 protein was also induced by hypoxia in both cell lines. This induction of ANXA1 was also confirmed by examining mRNA expression (Fig. 4B). These results suggested that hypoxia affects tumor invasion through induction of ANXA1 in TNBC.