A total of 54 paired fresh primary tumor and metastatic lymph node tissues were collected from Hunan Cancer Hospital and The Affiliated Cancer Hospital of Xiangya School of Medicine, Central South University (Changsha, China) between November 2013 and March 2014. All the female patients were initially diagnosed with breast cancer, and did not receive radiotherapy or chemotherapy prior to surgery (modified radical mastectomy). Each axillary lymph node tissue sample was divided into two parts: The first was used for H&E staining, and the second was stored in liquid nitrogen. A total of 23 paired axillary lymph nodes were identified to be positive.

A total of 207 tissue samples were obtained from the female patients with breast disease who underwent surgery (including radical mastectomy, modified radical mastectomy and local excision) in Hunan Cancer Hospital (Changsha, China) between May 1996 and March 2008 and used for immunohistochemical staining. Among them, 15 patients were diagnosed with BBD, including breast fibrocystic changes, atypic proliferation and fibroadenoma, and 192 patients were diagnosed with breast cancer with (n=109) or without (n=83) regional lymph node metastases. All tissue slides were examined by two pathologists in the Pathology Department of Hunan Cancer Hospital. The pathological parameters of the patients are summarized in Table I. This study was approved by the Research Ethics Committee of Hunan Cancer Hospital and informed consent was obtained from all of the patients.

The iTRAQ proteomic analysis was performed using the Fitgene iTRAQ Proteomics Platform (http://www.fitgene.com) according to the manufacturer's protocol. The tissue was placed in liquid nitrogen and ground it until it broke down. A total of 50 mg smashed tissue was added into a 1.5 ml eppendorf tube with 300 µl radioimmunoprecipitation assay lysate by pipette and agitated repeatedly. The lysate cells release a viscous material which was sonicated using ultrasound (5%, 1 sec on, 1 sec off, 10 sec on ultrasound, cooled on ice and repeated 8 times), and then centrifuged (4°C, 14,000 × g, 20 min) to obtain the supernatant. The prepared lysates (200 µg) were prepared with 4 µl reducing reagent (dithiotreitol; United States Biological, Salem, MA, USA) for 1 h at 60°C and then blocked with 2 µl cysteine Triethylammonium bicarbonate (TEAB; Sigma-Aldrich; Merck KgaA, Darmstadt, Germany) for 10 min at room temperature. Following centrifugation (14,000 × g, 20 min, room temperature) and solution collection, 100 µl 1M TEAB was added and discarded and the solution at the bottom of the tube following centrifugation (14,000 × g, 20 min) was collected, which was repeated 3 times. Subsequently, trypsin (trypsin to protein mass ratio, 1:100) was added and after 2 h trypsin was added again (trypsin to protein mass ratio, 1:50). 1M TEAB was added to make the volume of solution to 50 µl and the cell samples were stored overnight at 37°C. The digested peptide solution was mixed, centrifuged (14,000 × g, 4°C, 20 min), 50 µl 1 M TEAB was added, and sample were centrifuged again (14,000 × g, 4°C, 20 min) to obtain 100 µl digested sample. A total of 50 µl prepared protein sample and iTRAQ reagents were mixed according to the manufacturer's protocol, and the Dionex Ultimate 3000 system (Thermo Fisher Scientific, Inc., Waltham, MA, USA), was used to analyze the labeled protein sample. Protein identification and quantification were achieved by searching the UniProt database (18). Proteomics profiling and database searching based on TripleTOF^® 5600+ System (SCIEX, Framingham, MA, USA) and 6 ProteinPilot 4.0 (SCIEX) were performed according to the manufacturer's protocol.

To ensure the reliability and stability of the reported data, the following steps were performed for data quality control. First, prior to database searching, ‘Run False Discovery Rate Analysis’ was selected in the software ProteinPilot (ProteinPilot™ Software, version 4.5) for FDR control (SCIEX). Second, according to the SCIEX manufacturer's protocol, the software was set up as Unused ≥1.3 to ensure credibility ≥95%. Third, results identified by the reverse database were removed. Fourth, PMLN proteins with abnormal levels of quantified expression were excluded, specifically, when the expression of protein S100-A8 in PMLN was <0.05-fold or >20-fold compared with that in PBT. Finally, proteins with abnormal quantification between technical repetition and biological repetition were removed. A >1.5-fold change of protein expression was considered to indicate a significant difference between fresh PBT and fresh PMLN tissue. Gene ontology (GO) analysis of protein S100-A8 was performed, including biological process (GO ID: 0034641, 0009058, 0007165, 0006950, 0008219, 0002376, 0009056, 0042592, 0007010, 0048870, 0040011, 0051604, 0007155, 0040007), cellular component (GO ID: 0005634, 0005737, 0005829, 0043226, 0005576, 0005886, 0005856, 0005615) and molecular function (GO ID: 0043167, 0008092, 0008289), by searching GO browsers.

Paraffin-embedded tissue (4 µm thick, fixed in 10% neutral formalin for 12–24 h at room temperature) sections were heated at 60°C for 2 h, and then dewaxed by soaking in dimethylbenzene for 10 min twice, followed by rinsing in a graded ethanol series for 5 min at each concentration. Subsequently, the slides were immersed in citrate retrieval solution (trisodium citrate 3 g, citric acid 0.4 g, pH 6.0, 1,000 ml; Fuzhou Maixin Biotech Co., Ltd., Fuzhou, China), heated using a microwave and cooled to room temperature. A solution of 3% H₂O₂ was added to the tissue samples for 15 min at room temperature, and the slides were subsequently treated with ultra V block (TL-125-QHD; Thermo Fisher Scientific, Inc.) for 15 min at room temperature to block nonspecific binding. Subsequently, an anti-protein S100-A8 antibody (1:100; ab92331; Abcam, Cambridge, UK) was added to the tissue sections, and the slides were incubated at 4°C overnight. The following day, sections were warmed at room temperature for 1 h and incubated with primary antibody amplifier (TL-125-QHD) for 15 min at room temperature. Following this, the tissue samples were incubated with HRP polymer (TL-125-QHD) for 15 min at room temperature. Next, 3,3′-Diaminobenzidine (DAB; ready-to-use; cat no. TL-125-QHD; Thermo Fisher Scientific, Inc.) was used to stain the tissue sections for a few seconds at room temperature and evaluated by light microscopy (Olympus Corporation, Tokyo, Japan). Finally, the tissue sections were washed with water and stained with hematoxylin. Negative controls were performed by replacing DAB with PBS.

IHC staining was evaluated by two independent pathologists (Pathology Department of Hunan Cancer Hospital and The Affiliated Cancer Hospital of Xiangya School of Medicine, Central South University, Changsha, China). A total immunostaining score was calculated as the proportion score × the intensity score. The proportion score described the estimated fraction of positive-stained tumor cells as follows: 0, 0–4%; 1, 5–25%; 2, 26–50%; 3, 51–75%; 4, >75%. The intensity score represented the estimated staining intensity as follows: 0, no or marginal staining; 1, weak; 2, moderate; 3, intense. The total score ranged from 0–12, with 0 representing negative, 1–4 representing weak positive, 6–8 indicating moderate positive and 9–12 describing an intense positive score (19,20).

The age and follow-up time of patients is presented as the median (range). Patients were divided into positive and negative groups according to the total immunostaining score for protein S100-A8. All statistical analyses were performed using SPSS version 20.0 (IBM SPSS, Armonk, NY, USA). McNemar's test was used to examine the difference in S100-A8 protein expression between the positive rate of the PBT and PMLN FFPE tissues. Additionally, the χ² test was utilized to analyze the difference in the positive rate of protein S100-A8 expression between the BBD, NMBT, PBT and PMLN tissues. Associations between protein S100-A8 positive expression and the clinicopathological characteristics of 192 patients were also examined using the χ² test. The age and Survival rates of patients with breast cancer with protein S100-A8 positive or negative expression were calculated by using Kaplan-Meier analysis, and were examined using the log-rank test. The Cox proportional hazards model was utilized to determine the association between protein S100-A8 and breast cancer prognosis. P<0.05 was considered to indicate a statistically significant difference.

A previous study utilized the iTRAQ proteomic method to analyze differences in protein expression between PBT and PMLN tissues (11). A total of 4,837 proteins were identified, of which 643 were differentially expressed proteins, including 402 upregulated and 241 downregulated, in PMLN tissue, compared with PBT tissue. Protein S100-A8 was observed upregulated 1.72-fold in PMLN tissue, compared to PBT tissue. Further mass spectrometry analysis identified S100-A8 protein (Fig. 1). Based on the GO analysis performed in the current study (Table II), protein S100-A8 is involved in various biological processes, including cellular nitrogen compound metabolism, biosynthetic processes and signal transduction. Furthermore, GO analysis (http://www.uniprot.org/uniprot/P05109) revealed that it is an important constituent of numerous cellular components, including the nucleus, cytoplasm, cytosol, organelles, extracellular regions, plasma membrane and cytoskeleton. The molecular functions of protein S100-A8 include ion binding, cytoskeleton protein binding and lipid binding. This may suggest that protein S100-A8 serves a role in breast cancer metastasis, although further studies are required to elucidate this process.

IHC was utilized to examine the positive expression of protein S100-A8 in the NMBT, PBT and PMLN tissues, as well as the BBD control tissues. It was demonstrated that S100-A8 was located primarily in the cytoplasm of cancerous cells, whilst the nuclei of certain cancerous cells also exhibited sporadic, low-moderate expression of this protein (Fig. 2). In the BBD tissue, epithelial cells exhibited low or negative expression of protein S100-A8, whereas the staining of protein S100-A8 in the PBT tissue was more intense, compared with NMBT and PMLN tissues (Fig. 2). The majority of tissue types exhibited negative expression of protein S100-A8, and moderate or intense expression of protein S100-A8 was observed in BBD and NMBT tissues (Table III). A number of the PBT and PMLN tissues exhibited moderate staining (Table III). The positive expression percentages of protein S100-A8 in NMBT, PBT, PMLN and BBD control tissues were 37.3 (31/83), 34.8 (38/109), 18.4 (20/109) and 13.3% (2/15), respectively (Fig. 3). Compared with PMLN, the positive expression rate of protein S100-A8 in NMBT (P=0.002) and PBT (P<0.001) tissues was significantly higher. However, no significant differences were observed between any other types of tissue used in the study (BBD vs. NMBT, P=0.063; BBD vs. PBT, P=0.140; BBD vs. PMLN, P=1.000; PBT vs MNBT, P=0.643) (Fig. 3).

The association between protein S100-A8 and the clinicopathological features of 192 patients with breast cancer was further analyzed. The results indicated that the expression of this protein was associated with breast cancer histological type (P=0.022) and estrogen receptor expression (ER; P=0.009; Table IV). No significant correlation was observed between protein S100-A8 expression and patient age at diagnosis, histological grade, nodal status, tumor size, clinical stage, progesterone receptor (PR) expression, c-erbB2 expression, menstrual history (P>0.05; Table IV). Additionally, Kaplan-Meier analysis, followed by a log-rank test, was applied to identify the prognostic role of protein S100-A8. Although there was no significant difference observed in the overall survival rate between patients with breast cancer with negative or positive expression of protein S100-A8 (P=0.094), a possible association between the positive expression of protein S100-A8 and an increased survival time following surgery was observed (Fig. 4). According to the Cox proportional hazards model, histological grade (P=0.031) and nodal status (P=0.001) were risk factors for poor prognosis of patients with breast cancer (Table V).