The human breast cancer SUM1315 cell line was provided by Dr Stephen Ethier (University of Michigan, Ann Arbor, MI, USA) were labeled with green fluorescent protein (GFP), as previously described (13). GFP-labeled SUM1315 cells (G1315) were cultured in Gibco Dulbecco's modified Eagle's medium (DMEM; Thermo Fisher Scientific, Inc.) supplemented with 10% Gibco fetal bovine serum (FBS; Thermo Fisher Scientific, Inc.) and 1% penicillin-streptomycin solution (Gibco; Thermo Fisher Scientific, Inc.) in a humidified atmosphere of 5% CO₂ at 37°C.

Female severe combined immunodeficiency (SCID) CB-17IcrCrl-scidbgBR mice aged 4–6 weeks were purchased from the Model Animal Research Center of Nanjing University (Nanjing, Jiangsu, China). The SCID mice were kept under specific pathogen-free, temperature-controlled conditions. Cages, bedding and drinking water were autoclaved and changed regularly. Food was sterilized by irradiation. The mice were maintained in a daily cycle of 12 h period of lightness and darkness. In total, 6 mice were randomly divided into two groups to establish the HB and SUB mouse models.

Normal human breast tissues were obtained in May 2013 from elective breast reduction mammoplasty surgery of 3 patients at the The First Affliated Hospital of Nanjing Medical University (Nanjing, Jiangsu, China). Sample collection was performed in accordance with the ethical guidelines of the Declaration of Helsinki (14), and approved by the Ethics and Research Committee of the First Affiliated Hospital with Nanjing Medical University (Nanjing, Jiangsu, China). Breast tissues were sliced under sterile conditions into pieces ~4×4x4 mm in size. Three pieces were selected randomly for histological examination to exclude primary malignant disease. Small pieces of the human breast tissues were placed in ice-cold PBS until implantation into SCID mice. Implantation was finished within 6 h of mammoplasty surgery. Prior to implantation, the mice were anesthetized by intraperitoneal injection with 1% pentobarbital sodium (10 µl/g body weight; Sigma-Aldrich, Steinheim, Germany). Surgical procedures were modified from a previous method (15), which transplanted human gastric tissue, whereas in the present study, human mammary tissue was transplanted. Briefly, 5-6-mm scalpel incisions were made in the skin of the left mid-dorsal flank of mice, through which 5 human breast tissue fragments were implanted subcutaneously. The mice received gentamycin in the drinking water (800,000 U/l) until 1 week following the implantation. All in vivo experiments were conducted according to the Guide for the Care and Use of Laboratory Animals (16) and approved by the Animal Care and Use Committee of Nanjing Medical University.

Human breast cancer cells (~80% confluent) were cultured in fresh DMEM supplemented with 10% FBS and 1% penicillin-streptomycin solution for 24 h. The cells were harvested with 0.25% tyrosine and 0.02% disodium EDTA (Gibco; Thermo Fisher Scientific, Inc.), washed in medium, counted, and re-suspended in PBS. Human breast cancer cells (5×10⁵) in 0.2 ml PBS were injected into the implanted human breast tissues ~1 week subsequent to implantation of the human breast tissue. Equal amounts of tumor cells were injected into the left mid-dorsal flank subcutaneously to form the SUB control group.

All mice were sacrificed 5 weeks subsequent to injection of human breast cancer cells. Orthotopic tumor masses were collected under sterile conditions. One section was used for histological examination and another section was used for primary culture. Primary culture was conducted as previously described (13). Briefly, sections of the orthotopic tumor masses (2×2x2 mm) were washed in sterile PBS with 1% penicillin-streptomycin three times, and broken into small tumor pieces using scissors (1–2 Hz for 10 min) in 0.2 ml PBS. PBS (3 ml) was then added to the mixture, and the samples were centrifuged at 129 × g for 5 min in a centrifuge (Eppendorf, Hamburg, Germany). Subsequent to discarding the supernatant, the sedimentary tumor pieces were re-suspended in 5 ml DMEM and transferred to 10 cm dishes. Another 10 ml DMEM was added into the dishes for routine cell culture 16 h later. Following 5–7 days of culture, breast cancer cells from small tumor pieces adhered to dishes and the small tumor pieces were discarded. GFP-labeled primary cultured tumor cells were grown to ~80% confluency and harvested for fluorescence activated cell sorting (FACS) 1 week later. FACS analysis was conducted according to the manufacturer's protocol (FACSCalibur; BD Biosciences, Franklin Lakes, NJ, USA). The primary cultured human breast cancer cells were finally obtained, as follows: primary cultured G1315 cells from HB mouse models (pri-HB-1315); and primary cultured G1315 cells from SUB mouse models (pri-SUB-1315).

Total RNA was isolated from primary cells using TRIzol^® reagent (Thermo Fisher Scientific, Inc.) combined with the Qiagen RNeasy Lipid Tissue kit protocol (Qiagen, Valencia, CA, USA), according to the manufacturer's recommendation. RNA concentrations and the A260 nm/A280 nm ratio were assessed using a NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific, Inc., Waltham, MA, USA).

The HumanHT-12 v4 expression microarray (Illumina, Inc., San Diego, CA, USA) platform was used to measure expression levels in each unpooled specimen, according to the manufacturer's instructions. Briefly, subsequent to purification of RNA using an RNeasy Mini kit, 500 ng of total RNA was amplified and biotin-labeled with the Ambion Illumina^® Total Prep RNA Amplification kit (Thermo Fisher Scientific, Inc.). Labeled cRNAs were hybridized to the Illumina HumanHT-12 V4 expression BeadChip (Illumina, Inc.) and imaged using an Iscan system. Raw data was obtained using GenomeStudio Software (Illumina, Inc.).

RT-qPCR was used to verify the gene expression profile results. In total, 10 target genes were selected. PCR for the reference gene β-actin and target genes was performed for each cDNA sample, as in the microarray. All PCR reactions were performed on the Eppendorf MasterCycler RealPlex (Eppendorf) using the fluorescent SYBR Green I methodology with SYBR^® Premix Ex Taq™ (Takara, Dalian, Liaoning, China), according to the manufacturer's instructions. The thermal cycling conditions consisted of 10 min at 95°C and 40 cycles of 10 sec denaturation at 95°C, 15 sec annealing at 58°C, and 45 sec extension at 72°C. The primer sequences for RT-qPCR are listed in Table I. The mRNA quantities were analyzed in triplicate and normalized against β-actin as a control gene. The results are expressed as relative gene expression using the 2^−ΔΔCq method (17).

All data were analyzed using the SPSS software, version 12.0 (SPSS, Chicago, IL, USA). Paired t-test was used to screen differential genes in the microarray. Student's t-test was used for the analysis of the qPCR results. P<0.05 was considered to indicate a statistically significant difference.

A total of 94 genes were differentially expressed in the pri-HB-1315 cells cultured from the HB model and the pri-SUB-1315 cells cultured from the SUB model (Table II). Gene ontology (GO) analysis (18,19) was applied to obtain insight into the biology associated with the stereotypic differences between the pri-HB-1315 and pri-SUB-1315 cells. Significant enrichment of 16 molecular functions, 6 distinct biological processes and 2 cellular components was noted (P<0.05; Table III). The Kyoto Encyclopedia of Genes and Genomes (20) yielded 7 pathways with significant differences (P<0.05; Table IV).

A total of 10 genes were selected based upon microarray data and biological significance. Consistent with the microarray results, RT-qPCR confirmed the differential expression of 8 genes, with only 2 genes not demonstrating differential expression (Table V). Genes with significantly increased expression in the pri-HB-1315 group included SH3-domain binding protein 5 (BTK-associated), ribosomal protein L32 (RPL32), B-cell CLL/lymphoma 10 (BCL10), leucine-rich repeats and transmembrane domains 1 and KIAA1618 (P<0.05), while bromodomain PHD finger transcription factor (BPTF), tripartite motif containing 36 (TRIM36) and fibroblast growth factor 13 expression was significantly decreased in the pri-HB-1315 group (P<0.05). The other 2 genes, coiled-coil domain containing 113 and PWWP domain containing 2A, showed no significant difference in expression between the two groups.