The serum proteomic analysis was performed with 2DE. Two iterations showed the reliability and reproducibility of the experimental results. More than 100 protein spots were thoroughly separated, although highly abundant proteins interfered with the map (Figure 1A,B). Comparing the 2DE map of the breast cancer serum sample with that of the non-cancer serum sample, we observed three protein spots in the breast cancer serum sample that showed a two-fold or greater change in expression level (Figure 1A,B). Protein spots 1, 2, and 3 were increased in the breast cancer sample by 21.9-, 4.2- and 2.03-fold, respectively (Figure 1C).

The three protein spots that were significantly increased in the breast cancer serum sample gel were excised and analyzed with MALDI-TOF-MS to obtain the peptide sequences (Figure 2).

Using the NCBI GenBank database, we matched the peptides to three protein sequences (Figure 2). The proteins were identified as the human crystal structure of the Fxia catalytic domain in complex with ecotinm84r (spot 1 in Figures 1 and 2A), the haptoglobin alpha (2FS)-beta precursor (HP) (spot 2 in Figures 1 and 2B) and ILP-2 (spot 3 in Figures 1 and 2C) (see Table 1 for details).

Of the three proteins, only ILP-2 has been previously associated with cancer [16]; thus, we focused further studies on ILP-2. To validate the ILP-2 expression change in the breast cancer serum sample, we performed Western blot analysis on 10 breast cancer serum samples and 10 non-cancer serum samples (Figure 3A). Eight of the 10 breast cancer serum samples had increased ILP-2 levels compared to the non-cancer serum samples (Figure 3B). After normalizing the data to GAPDH levels, the average level of ILP-2 was 0.86 in the breast cancer serum samples and 0.19 in the non-cancer serum samples. The ILP-2 levels in the breast cancer serum samples were significantly higher than in the non-cancer serum samples using Student’s t-test (Figure 3C).

Serum samples from 35 subjects with breast cancer, other types of cancer, galactophore hyperplasia, breast cancer post-surgery or without cancer were analyzed with 2DE (Figure 4). ILP-2 was present in the serum samples from breast cancer, breast cancer post-surgery and other cancer subjects, whereas there was no ILP-2 present in the serum samples from healthy subjects or subjects with galactophore hyperplasia. Moreover, sera from subjects with breast cancer had increased ILP-2 levels compared to the other serum samples.

An ELISA was performed to further validate the ILP-2 differences in the serum samples of 35 subjects (Figure 5). Although ILP-2 was present in all five types of serum samples, the sera from breast cancer subjects had significantly increased ILP-2 levels compared to the other serum samples. Furthermore, sera from post-surgery breast cancer patients had significantly less ILP-2 protein than the other serum samples, including the samples from breast cancer patients.

ILP-2 is the most recent apoptosis inhibitor to be identified [14,15] and has only been detected in the testis and in lymphoblastoid cells [14]. There are no previous studies on ILP-2 in breast cancer tissues and in sera from breast cancer patients.

From a medical perspective, a serum sample is the most informative proteome source, because numerous cells communicate through the secretion of soluble proteins into the blood [17,18]. Furthermore, serum analysis is a convenient method to diagnose breast cancer with a peripheral blood sample [19]. The correlation between ILP-2 and breast carcinogenesis and the potential of ILP-2 as a serological biomarker for breast cancer remain unclear.

We performed 2DE and MALDI-TOF analysis on 400 pooled samples of sera from breast cancer patients and 40 pooled samples of sera from healthy control subjects, and we identified ILP-2 expression in the breast cancer serum samples (Figures 1 and 2, Table 1). ILP-2 was present in the breast cancer serum samples, but not in the control serum samples.

Western blot analysis was performed on 10 breast cancer serum samples and 10 non-cancer serum samples to validate the correlation between ILP-2 and breast cancer (Figure 3). The results indicate increased levels of ILP-2 in the breast cancer serum samples compared to the non-cancer serum samples.

Then, we performed 2DE and MALDI-TOF analysis on 35 samples of sera from healthy subjects and subjects with other types of cancers, galactophore hyperplasia, breast cancer post-surgery and breast cancer according to previously described methods (Figures 2C and 4). The serum samples from subjects with breast cancer, breast cancer post-surgery and other cancer types contained ILP-2, whereas no ILP-2 was measured in the serum samples from galactophore hyperplasia and healthy control subjects. Moreover, the breast cancer serum samples had increased ILP-2 protein levels compared to the other serum samples.

Thereafter, we used an ELISA to analyze the 35 serum samples (Figure 5). The non-cancer serum samples were collected before the ELISA analysis. PMSF was added to the sample at 1 mM concentration, and the mixture was frozen at −80 °C. Despite the presence of ILP-2 in the serum samples from galactophore hyperplasia, breast cancer post-surgery, other types of cancer and non-cancer subjects, the ILP-2 protein levels were increased in the serum samples from breast cancer patients.

This study was approved by the Gannan Medical University Ethical Committee. Written informed consent was obtained from all patients before their participation in this study.

Breast cancer serum samples were collected from Stage 3 breast cancer patients who were diagnosed in accordance with the breast cancer pathology standard. The same procedure was applied to the serum samples from subjects with other types of cancers (Table 2). The average age of the breast cancer patients was 45 years, and the average age of the women with other types of cancers was 40 years.

Whole blood (5 mL) was collected from the cubital vein and then transferred to a clear, empty tube. The sample was incubated at room temperature for 2 h and then centrifuged at 1500× g for 10 min at 4 °C. The supernatant (80 μL) was transferred into an Eppendorf (EP) tube and 1 mM phenylmethylsulfonylfluoride (PMSF, Sigma, St. Louis, MO, USA) was added. The sample was stored in a freezer at −80 °C [20].

The sera from 400 breast cancer patients and from 40 women without cancer were mixed separately. The mixtures comprised 50 μL serum from each sample. The mixed sera were centrifuged at 1500× g for 30 min at 4 °C. The supernatant was processed by using the Aurum™ Serum Protein Mini Kit (Bio-Rad, Hercules, CA, USA), according to the manufacturer’s instructions. The depleted sample (with albumin and IgG removed) was used for 2DE [21].

The protein concentration of the serum sample was quantified with the improved Bradford method [22].

Samples of 35 μL of breast cancer sera and non-cancer sera (the undiluted sera) were transferred into separate, clear EP tubes. Then, 400 μL of the hydration solution (7 M urea, 2 M thiourea, 2% CHAPS (3-((3-cholamidopropyl)dimethylammonio)-1-propanesulphonate), 0.2% (w/v) Bio-Lyte (Bio-Rad, Hercules, CA, USA), 2% SB3-10 (N-decyl-N,N-dimethyl-3-ammonio-1-propanesulphonate) and 65 mM DTT (dithiothreitol, Sigma, St. Louis, MO, USA) was added to the sera and incubated at 25 °C for 2 h. The mixture was then added into a hydrate groove. An immobilized 17 cm pH gradient strip (pH 3–10, Bio-Rad, Hercules, CA, USA) was placed in the mixture and rehydrated at 25 °C for 16 h [23]. The isoelectric focusing (IEF) process involved the use of a 250 V linear ramp for 1 h, a 1000 V linear ramp for 1 h, a 4000 V linear ramp for 2 h, an 8000 V linear ramp for 3 h, an 8000 V rapid ramp for 10 h and a 500 V rapid ramp for 30 min with a PROTEIN IEF System (Bio-Rad, Hercules, CA, USA) [24]. After the IEF, each gel strip was equilibrated with 6 mL buffer I (6 M urea, 30% glycerol, 2% SDS, 0.375 M Tris-HCl pH 8.8 and 0.18 g/strip DTT) for 15 min and then with 6 mL buffer II (6 M urea, 30% glycerol, 2% SDS, 0.375 M Tris-HCl pH 8.8 and 0.225 g/strip iodoacetamide) for 15 min. The equilibrated gel strip was placed on top of a 12% sodium dodecyl sulphate polyacrylamide gel and then sealed with an agarose overlay (Bio-Rad, Hercules, CA, USA). The sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) was performed with 12 mA/per gel for 30 min and then with 25 mA/per gel until the bromophenol blue reached the bottom of the gels [25]. The gels were then stained with a silver-staining procedure [26,27].

The same method was applied to the 35 serum samples from the healthy control subjects and the subjects with other types of cancers, galactophore hyperplasia, breast cancer post-surgery and breast cancer.

Before analyzing the different proteins in gels, the 2DE experiment was repeated twice under the same conditions to confirm the 2DE analysis reliability and reproducibility [16,28–30].

Changes in the protein spots of the 2DE maps between the breast cancer serum sample and the non-cancer serum sample were analyzed with the Melanie 3 Viewer. The spots that exhibited at least a two-fold change between the sample gels were excised and then transferred to a clear EP tube. The proteins were digested with the in-gel trypsin method, and the extracted peptides were dissolved in an equivalent volume of an α-cyano-4-hydroxycinnamic acid solution. After air-drying, the samples were placed on a stainless steel plate and desalted with 0.1% trifluoroacetic acid, and the peptides were identified with the Voyager DE Pro matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrograph (Applied Biology System Company, Foster City, CA, USA) [31–33]. In the MALDI-TOF-mass spectrometry (MS) analysis, the extracted peptides were identified with the reflection mode and the positive ion spectrum. We applied adrenocorticotropic hormone for calibration and identified the peptide mass fingerprint using the Mascot Distiller (Matrix Science, London, UK). The detailed search parameters are listed in Table 3. The peptide match score was above 4, and the Mascot total score was above 64, which are regarded as statistically significant [34–36]. The same method was applied to the 35 serum samples from the healthy control subjects and the subjects with other cancer types, galactophore hyperplasia, breast cancer post-surgery and breast cancer.

A volume of 50 μL breast cancer serum or 50 μL non-cancer serum was separately diluted into 400 μL phosphate buffer. These sera mixtures were from 10 breast cancer patients and 10 healthy control subjects. The mixtures were centrifuged at 10,000× g for 20 min at 4 °C. The supernatant was collected, and then an equal volume of 2× loading buffer (100 mM Tris-HCl pH 6.8, 250 mM DTT, 10% β-ME, 4% SDS, 0.2% bromophenol blue and 20% glycerol) (Sigma, St. Louis, MO, USA) was added. The mixture was boiled for 10 min. A 20 μL/well sample was loaded onto 12% SDS-polyacrylamide gels [37]. The proteins were transferred to a polyvinylidene difluoride membrane [38]. Western blot analysis was performed by using a rabbit anti-human ILP-2 primary antibody (Abnova, Walnut, CA, USA) and a peroxidase-conjugated goat anti-rabbit IgG (Promega, Madison, WI, USA) [39]. An enhanced chemiluminiscence detection system (ECL-plus, Beytime, Haimen, China) was used to visualize the immunoreactive proteins via exposure to X-ray film (Tianguang, Tianjin, China). The rabbit anti-human ILP-2 (Promega, Madison, WI, USA) and rabbit anti-human GAPDH (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) antibodies were diluted at 1:500 and 1:250, respectively. Signal intensity was quantified through densitometry. The signals were normalized to GAPDH [40–42].

A total of 35 serum samples from healthy subjects and subjects with other cancer types, galactophore hyperplasia, breast cancer post-surgery and breast cancer were diluted 40,000-fold and then analyzed according to the kit instructions (USCNLIFE, Missouri, TX, USA) [43,44]. The ELISA was performed on every serum sample in triplicate. The optical density of each sample was measured with a microplate reader at 450 nm [45,46].

The results were presented as X ± SD. Statistical analysis was performed using Student’s t-test, analysis of variance (ANOVA) or Tukey’s test with SPSS 11.5. A value of p < 0.05 was considered statistically significant.