Proteomic analysis of nipple aspirate fluid from women with early-stage breast cancer using isotope-coded affinity tags and tandem mass spectrometry reveals differential expression of vitamin D binding protein

NAF samples were obtained from 18 women with stage I or II unilateral invasive breast carcinoma who presented to The University of Texas M. D. Anderson Cancer Center's Nellie B. Connally Breast Center. NAF was also collected from 4 healthy volunteers not related to the patients. Individuals were eligible to participate if they had not previously undergone subareolar surgery, which might have disrupted the terminal ductal system. All participants gave written informed consent to undergo bilateral nipple aspiration. The study protocol was approved by M. D. Anderson Cancer Center's institutional review board.

Before aspiration was attempted, the nipple was initially cleansed with a small amount of Omniprep paste (D.O. Weaver and Co., Aurora, CO) to remove any keratin plugs and then wiped with an alcohol pad. A small amount of lotion was placed on the breast, and the breast was gently massaged from the chest wall toward the nipple for 1 minute. Ductal fluid was collected as previously described by Sartorius et al. [29]. In brief, nipple aspiration was accomplished using a handheld suction cup (Product Health, Inc., Menlo Park, CA) similar to the nonpowered breast pumps commonly used to express milk from lactating women. With the suction cup over the nipple, the plunger of the syringe was withdrawn to the 5- to 10-mL level until ductal fluid was visualized. The fluid droplets were collected into a 10-μl graduated micropipette (Drummond Scientific Co., Broomall, PA). In the case of women with breast cancer, NAF was collected separately from the tumor-bearing and disease-free breasts. The volumes of NAF were recorded for each patient and each breast.

Immediately after collection, the NAF samples were rinsed into centrifuge tubes containing sterile phosphate-buffered saline supplemented with the protease inhibitors 4-(2-aminoethyl)-benzenesulfonylfluoride HCl (0.2 mM), leupeptin (50 μg/mL), aprotinin (2 μg/mL), and dithiothreitol (0.5 mM). The samples were then centrifuged at 1500 RPM for 10 minutes, and the supernatants were collected and suspended in a 500 μl PBS buffer with the previously mentioned protease inhibitors.

ICAT analysis was performed using a Cleavable ICAT Reagent Kit (Applied Biosystems, Foster City, CA) according to the manufacturer's guidelines with some modifications (Figure 1) [30]. For ICAT analysis, the NAF protein samples from the tumor-bearing breasts and disease-free breasts of patients with early-stage breast cancer were respectively pooled in a normalized fashion. One hundred micrograms of each NAF protein pool was separately reduced in 80 μL of denaturing buffer at 37°C for 2 hours with 5 mM TCEP and alkylated at 37°C for 2 hours in the dark with ICAT-heavy (NAF from tumor-bearing breasts) and ICAT-light (NAF from disease-free breasts) reagent. After alkylation, ICAT-heavy and ICAT-light reactions were mixed together and concentrated by centrifugal evaporation, SDS-PAGE loading buffer was added, and the sample was separated by SDS-PAGE. The gel was cut into 14 slices and subjected to in-gel trypsin digestion at 37°C for 20 hours. After extraction, the ICAT-labeled peptides were purified using the ICAT Avidin Buffer Pack and Avidin Affinity Cartridge kit (Foster City, CA) according to the manufacturer's guidelines. Briefly, the peptide mixture was dried by vacuum centrifugation and resolubilized in the loading buffer of the kit. The peptide mixture was loaded in the Avidin Affinity Cartridge and washed twice with two kinds of wash buffer (first wash in 2× PBS followed by a second wash in 50 mM NH₄HCO₃ in 20% MeOH) to reduce the salt concentration and remove nonspecifically bound peptides. The ICAT-labeled peptides were then eluted with the elution buffer and dried by vacuum centrifugation. The dried peptides were cleaved from the biotin portion of the ICAT reagent with the cleaving reagents at 37°C for 2 hours. The ICAT-labeled peptides were then dried by vacuum centrifugation and resolubilized in 2% formic acid for liquid chromatography and MS.

Liquid chromatography and MS were performed using an LCPackings nano-LC system consisting of an Ultimate pump, Famos autosampler, and Switchos column-switching device for on-line desalting (Dionex Corp., Sunnyvale, CA). The Switchos was fitted with a C18 reversed phase column (Pepmap C18, 300 μmIDx1 mm,) and a C18 reversed phase nanocolumn (Pepmap C18, 75 μmIDx10 mm). Reversed phase solvents were 0.005% heptafluorobutyric acid in either water and 2% acetonitrile or 80% acetonitrile. Gradient elutions were performed over 90 minutes from about 5% acetonitrile to about 40% followed by flushing with 72% acetonitrile and reequilibration for the next injection. Spectra were acquired on a QqTOF quadrupole/time-of-flight mass spectrometer (Qstar Pulsar-I, Applied Biosystems/MDS Sciex, Foster City, CA). Qstar was externally calibrated, using the doubly and triply charged peaks of the peptide neurotensin. One survey MS scan was followed by two MS/MS scans on the two most intense ions from the MS spectrum, omitting previously selected targets.

The acquired MS/MS spectra were automatically compared against the National Center for Biotechnology Information (NCBI) database using the ProICAT program (Applied Biosystems/MDS Sciex) allowing 1 missed cleavage (trypsin) and 0.3 and 0.5 d windows for MS/MS and MS resp. Modifications allowed were fairly typical: ICAT on Cys, Met oxidation. The same program was also used to calculate ICAT ratios. The ICAT experiment by design produced fewer peptides, but the requirement of containing Cys increases the confidence of the one peptide hits when manually curated (as these were). We extracted tables of high-confidence (99%) matched peptides with their ratios and used them for further calculations. The quality of peptide matches of interest was checked manually.

Protein samples (40 μg) were separated on a 10% SDS-PAGE gel and transferred (170 V for 45 minutes) to nitrocellulose filters (pure nitrocellulose; pore size, 0.45 μm; Bio-Rad, CA). Blots were blocked overnight at 4°C in 5% (wt/vol) skim milk-TTBS (0.1% Tween 20-Tris-buffered saline). The next morning the membranes were incubated for 1 hour in 5% skim milk in PBS with polyclonal anti-vitamin D-binding protein (Biodesign International, Saco, ME) diluted 1:1000. Membranes were then incubated with goat anti-rabbit secondary antibody (Santa Cruz, CA) for 1 hour at 4°C. The bound antibody was visualized using the ECL-Plus Detection kit (Amersham Pharmacia Biotech, Piscataway, NJ) and the bands were analyzed by video imaging and densitometry.

Densitometry measurements were analyzed and statistical significance was determined using independent Student's t-test. A standard cut-off of 15–20% change is commonly utilized to denote differential expression. In the current experiments, a peptide ratio of 1.56 correlated with one standard deviation above the median ratio.

NAF from the tumor-bearing breasts versus disease-free breasts of 10 patients with unilateral early stage breast cancer was assessed by ICAT analysis. Using the ProICAT bioinformatics program, 353 peptides were identified with a high level of confidence (> 99%) on the basis of matching of tandem mass spectra of peptides to peptide sequences in the NCBI database. Approximately equal numbers of peptides were up- versus down-regulated (Figure 2). Following manual review of all peptide match spectra and exclusion of redundant matches, 39 proteins were identified that had sufficient quality to justify assignment (Table 1). Twenty proteins (51%) were identified based on two or more peptide matches; 19 proteins (49%) were identified based on a single peptide match.

As previously noted, peptides from tumor-bearing breasts had a heavy tag, and peptides from disease-free breasts had a light tag. Of the 39 proteins, 19 had an ICAT ratio of Heavy:Light (H:L) > 1.0, 16 proteins had an ICAT ratio of H:L < 1.0, and 1 protein had an ICAT ratio of H:L = 1.0. In three cases, a peptide signal could only be convincingly detected in one, not both states (diseased versus not), so no ratio could be calculated. Proteins that were differentially expressed between the NAF from tumor-bearing breasts and the NAF from disease-free breasts included, among others, alpha2HS-glycoprotein (H:L ratio 0.63), lipophilin B (H:L ratio 1.42), beta-globin (H:L ratio 1.98), hemopexin (H:L ratio 1.73), and vitamin D-binding protein precursor (H:L ratio 1.82) (Table 1). For the one-peptide hit of interest, Vitamin-D binding protein, a second peptide, not containing Cys was identified using Mascot (data not shown).

The finding that ICAT identified vitamin D-binding protein as being overexpressed in NAF from the tumor-bearing breasts of women with early-stage breast cancer was particularly interesting (Figures 3 and 4). Previous studies [31‐34] have suggested that the vitamin D receptor may be overexpressed in breast tumors. We therefore sought to examine further the expression of vitamin D-binding protein in a separate, independent set of NAF specimens to validate the ICAT vitamin D findings.

To assess the correspondence of the ICAT vitamin D-binding protein data with an independent assay of protein abundance, we performed Western blot analysis on NAF obtained from a separate cohort of 12 women (4 healthy volunteers and 8 women with early-stage breast cancer). On western blot analysis, vitamin D-binding protein expression in the pooled samples of NAF specimens from women with early-stage breast cancer was twice that in the NAF specimens from healthy volunteers (P = 0.04; standard error = 9703.7) (Figure 5). Interestingly, the relative difference in expression was similar to the abundance ratio derived from ICAT analysis (1.82). Although the data derived from ICAT and pooled western analyses were similar, western blot analysis of individual NAF specimens showed variation in vitamin D-binding protein expression among both healthy volunteers and women with early-stage breast cancer (Figure 5). Menopausal status did not affect expression of vitamin D-binding protein expression in NAF from women with early-stage breast cancer (P = 0.57; standard error = 14425.3).