Proteomic Profiling and Interactome Analysis of ER-Positive/HER2/ neu Negative Invasive Ductal Carcinoma of the Breast: Towards Proteomics Biomarkers

Abstract

Breast cancer, especially ER positive/HER2/neu negative IDC, is the predominant subtype of invasive ductal carcinoma. Although proteomic approaches have been used towards biomarker discovery in clinical breast cancer, ER positive/HER2/neu negative IDC is the least studied subtype. To discover biomarkers, as well as to understand the molecular events associated with disease progression of estrogen receptor positive/HER2/neu negative subtype of invasive ductal carcinoma, differential protein expression profiling was performed by using LC-MS^E (MS at elevated energy). A total of 118 proteins were identified, of which 26 were differentially expressed. These identified proteins were functionally classified and their interactions and coexpression were analyzed by using bioinformatic tools PANTHER (Protein Analysis THrough Evolutionary Relationships) and STRING (Search Tool for the Retrieval of Interacting Genes). These proteins were found to be upregulated and were involved in cytoskeletal organization, calcium binding, and stress response. Interactions of annexin A5, actin, S100 A10, glyceraldehyde 3 phosphate dehydrogenase, superoxide dismutase 1, apolipoprotein, fibrinogen, and heat shock proteins were prominent. Differential expression of these proteins was validated by two-dimensional gel electrophoresis and Western blot analysis. The cluster of these proteins may serve as a signature profile for estrogen receptor positive/ HER2/neu negative subtype.

Introduction

Cancer is a disease with dynamic changes in the proteome resulting from cumulative alteration in gene expression, mutations, and post-translational modifications, which leads to deregulation of intracellular signaling pathways allowing cancer cells to evade signals that are involved in normal functioning of cell (Hanahan et al., 2000). Amongst different cancers, breast cancer is one of the leading causes of death in women worldwide, with 458,400 estimated deaths in 2008 accounting for 23% of the total cancer deaths (Jemal et al., 2011). Breast cancer is a heterogeneous disease characterized by a wide spectrum of histological, pathological, clinical, and molecular signatures. Invasive ductal carcinoma (IDC) is one of the major breast cancer characterized by expression of hormone and growth factor receptors, such as estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth receptor (HER2/neu). Based on the expression of these receptors, IDC has been classified into five types viz. luminal A (ER positive/ HER2/neu negative), luminal B (ER-positive/HER2/neu positive), HER2 type, triple negative/ basal like (negative for HER2, ER and PR), and normal like, of which luminal A and B types accounts for 60% of the tumors (Sorlie et al., 2001; van't Veer et al., 2002). Further, luminal A subtype is categorized into four groups: a] (HER2/neu negative, ER positive, PR positive), b] (HER2/neu negative, ER negative, PR positive), c] (HER2/neu negative, ER positive, PR negative), and d] (HER2/neu negative, ER negative, PR negative) (Perou et al., 2000; Sorlie et al., 2001).

Proteomics-based studies enable comparative and quantitative investigation of cellular protein expression, which helps in discovery of biomarkers, drug targets, and understanding signaling pathways that govern tumorigenesis (Simpson et al., 2001). In view of this, many comparative proteomic studies in breast cancer have been reported (Zhang et al., 2005; Rezaul et al., 2010). Although HER2/neu positive tumors are more aggressive phenotype, they account for only 25% to 30% of all IDCs (Slamon et al., 1989), whereas the ER positive/HER2/neu negative subgroup numerically comprises about two-thirds of all breast cancers and in fact is the dominant subtype. Very few proteomics studies have been carried out in ER positive/HER2/neu negative IDC breast clinical tissues till now. Only one such study has been reported, where upregulation of peptidyl prolyl cis-trans isomerase B, GDP inhibitor, and tropomyosin alpha 4 were observed consistently (Weitzel et al., 2010). Therefore, it is important to study this subtype of IDC, which may be useful in better cancer management. Here we have studied the proteomic profile of ER positive/HER2/neu negative IDCs. Further interaction and coexpression of proteins were analyzed by using bioinformatic tools. In addition, the expressions of key differentially expressed proteins were validated by two-dimensional electrophoresis and Western blot analysis.

Materials and Methods

Chemicals

All chemicals were procured from Sigma-Aldrich (St. Louis, MO, USA), unless mentioned. Glu-fibrinopeptide and polyethylene glycol were procured from Waters (Waters Corporation, Milford, USA). All the primary antibodies were purchased from Abcam (Abcam, UK).

Clinical breast tumor tissue

A total of six malignant breast tissue biopsy specimens and corresponding adjacent normal breast tissue from the same patients without any signs of malignant transformation were obtained from Jehangir Hospital, Pune, India. Collected samples were immediately snap-frozen in liquid nitrogen and were stored at −80°C until further use. All these cases were diagnosed between the year 2006 and 2008; and had given their informed consent in accordance to the guidelines of the hospital/institutional scientific and ethical committee approval prior to inclusion into the study. The percentage of epithelial cells in tumor and normal tissue with more than 90% were subjected to proteomic analysis. Clinical information about the samples is listed in Table 1 and histology microphotographs are shown in Figure 1.

FIG. 1.

Hematoxylin and eosin stained histological 40× microphotographs showing lymph node metastasis and tumor necrosis.

Table 1.

Clinical Information about IDC Samples

		Hormone receptor status
Sl. No	HER2/neu status	Estrogen	Progesterone	TNM status	Tumor grade
1	Negative	Positive	Positive	T4N1Mx	II, Score 6
2	Negative	Positive	Positive	T3N3Mx	II, Score 6
3	Negative	Positive	Positive	T1N1Mx	II, Score 6
4	Negative	Positive	Positive	NA^*	II, Score 7
5	Negative	Positive	Positive	T1N0Mx	II, Score 6
6	Negative	Positive	Positive	NA^*	II, Score 6

NA^*, not available.

Sample preparation

For analysis of LC-MS^E (MS at elevated energy), pooled samples of six IDC clinical tissues and a pooled six adjacent control tissue samples were analyzed in three technical replicates. For two-dimensional electrophoresis (2DE) analysis and Western blot analysis, six individual IDC clinical samples and a pooled adjacent control tissue samples were independently analyzed. The frozen clinical IDC breast tissue samples were thoroughly washed with phosphate buffered saline (PBS), followed by homogenization by using hand-held tissue tearor (BioSpec Pvt. Ltd., USA) in protein extraction buffer containing 8 M urea, 2 M thiourea, 4% CHAPS, 70 mM DTT, 1% N-Deconoyl N-methylglucamide and 50 μL of mammalian protease inhibitor cocktail. Protein extract was centrifuged at 50,000 g for 1 h at 20°C. Protein concentration was determined by using a quick start Bradford protein assay (Bio- Rad, Hercules, CA, USA).

In-solution tryptic digestion for LC-MS^E analysis

Extracted protein sample was precipitated by TCA/acetone. 10 μg of complex protein mixture was washed and solubilized in 50 mM ammonium bicarbonate buffer containing 0.1% RapiGest (Waters Corporation, MA, USA). Protein solution was subjected to reduction by using 100 mM DTT at 60°C for 15 min, followed by alkylation by using 200 mM iodoacetamide in dark at room temperature for 30 min. The proteins mixture was digested overnight with proteomic grade trypsin at 37°C.

LC-MS^E analysis

Prior to LC-MS^E acquisition, for proteome quantification predigested internal standard (yeast alcohol dehydrogenase, Waters Corporation) was spiked with 50 fmol per 2 μL of protein digest. The 2 μL of protein digest with final concentration of 100 ng/μL was analyzed by LC-MS^E workflow by using nanoACQUITY (UPLC) online coupled to Q-TOF mass spectrometer, Synapt HDMS system (Waters Corporation) equipped with a nanolockspray ion source with flow rate of 300 nL/min (external lockmass standard: Glu-fibrinopeptide (GFP) (m/z 785.8426)). Sampling of the lock spray channel was performed every 30 s, and the system was tuned for a minimum resolution of 10000 and calibrated using GFP infusion. Both tumor and corresponding control digests were acquired in three technical replicates. Peptide samples were injected online onto a 5 μm Symmetry C18 trapping column (180 μm×2 cm length) at a flow rate of 15 μL/min. Peptides were separated by in-line gradient elution onto BEH (Bridged Ethyl Hybrid) 130 C18 1.7 μM×75 μM×150 mm nanoACQUITY analytical column, at a flow rate of 300 nL/min using a linear gradient from 5% to 40% B over 35 min (A. 0.1% formic acid in water, B. 0.1% formic acid in acetonitrile). Acquisition was performed in positive V mode in a mass range of 50–1990 m/z with a scan time of 1 s with alternating low (5 eV) and high (20 to 40 eV) collision energy. A capillary voltage of 3.2 kV, source temperature of 80°C, and cone voltage of 32 V were maintained during the analyses.

Data processing and database searching

LC-MS^E data were processed by using ProteinLynx Global Server (PLGS version 2.4. Waters Corporation) to generate charge state reduced and deisotoped precursor mass lists as well as associated product ion mass lists for subsequent protein identification and quantification. The processed data were allowed to search against reviewed human subset of UniProt database.

Prior to searching, the internal standard yeast alcohol dehydrogenase sequence (UniProt) was added to the human database. Default search parameters were used, including the “automatic” setting for mass accuracy, a minimum of 1 peptide match per protein, a minimum of 3 consecutive product ion matches per protein, a minimum of 7 total product ion matches per protein, and 2 missed tryptic cleavage site were allowed. A fixed modification of carbamidomethylation of cysteine residues and variable modifications as oxidation of methionine residue were used during search. False positive rate (FPR) for protein identification was set to 4%, and the absolute protein quantification functionality was enabled using yeast alcohol dehydrogenase as an internal standard.

Protein–protein interaction and coexpression analysis of IDC proteome

IDC proteome identified by LC-MS^E were uploaded onto STRING (Search Tool for the Retrieval of Interacting Genes) database search (Szklarczyk et al., 2011). STRING database version 9.0 (http://www.string-db.org) covers 5,214,234 proteins from more than 1100 completely sequenced organisms. It provides comprehensive protein–protein interaction and coexpression analysis with a confidence score using both experimental evidence and predicted interaction information.

Functional analysis of IDC proteome

IDC proteome identified by LC-MS^E were analyzed by PANTHER (Protein Analysis THrough Evolutionary Relationships; http://www.pantherdb.org) (Mi et al., 2010). This software allows predicting protein functions using experimental evidence and evolutionary relationships. The classified proteins were then categorized and visualized by a biological process, molecular function, protein class, and pathway ontology terms.

2DE and image analysis

For 2DE analysis, 11 cm nonlinear 4–7pH gradient IPG strips (Bio-Rad Laboratories, Richmond, CA, USA) were rehydrated with protein (270 μg), solubilized in rehydration buffer 8 M urea, 2 M thiourea, 30 mM Tris-HCl, 4% CHAPS, 70 mM DTT, 0.11% N-deconoyl N-methylglucamide, 0.1% 3-(decyldimethylammonio) propane sulfonate, and 0.1% of ampholytes (GE Healthcare Bio-Sciences, Sweden). First dimension isoelectric focusing was performed using Protean IEF cell (Bio-Rad Laboratories) followed by reduction with DTT (2%) and alkylation with iodoacetamide (2.5%) prepared in 6 M urea, 2% SDS, 0.375 M Tris-HCl (pH 8.8), 20% glycerol. Second dimension separation on to 12.5% SDS PAGE was performed using either SE 600 Ruby (GE Healthcare Bio-Sciences, Sweden) or Protean II xi Cell (Bio-Rad Laboratories), and resolved proteins were visualized by staining with CBB R250. Gel images were acquired by using GS800 calibrated densitometer (Bio-Rad Laboratories) and image analysis was performed by using PDQuest Advanced software (Bio-Rad Laboratories).

In-gel tryptic digestion

Excised protein gel spots were destained in solution containing 50 mM ammonium bicarbonate/50% acetonitrile. Gel spots were dehydrated in 100% acetonitrile followed by drying in a vacuum concentrator. Gel spots were reduced for one hour in 10 mM DTT at 45°C and subsequently alkylated with 55 mM iodacetamide in the dark at room temperature. Gel spots were again dehydrated, dried, and rehydrated in proteomic grade trypsin solution at 20 μg/mL enzyme concentrations. After about 20 h of incubation at 37°C, the peptides were extracted with 5% formic acid in 50% acetonitrile. Supernatants of protein digest were dried in vacuum centrifuge at room temperature. Evaporated digest samples were stored at −80°C until further analysis.

MALDI MS/MS analysis

MALDI MS/MS analysis was performed on the SYNAPT HDMS system (Waters Corporation). Peptide digests were reconstituted in 5% ACN containing 0.1% TFA and premixed with equal volume of 10% CHCA matrix and applied on to 96-well MALDI plate. All samples were acquired on 200 Hz solid state UV laser in V mode by MassLynx 4.1. The quadrupole profile was set to 500 m/z for 5% of the scan time and then ramped to 1500 m/z for remaining period of the scan. The instrument mass calibration was performed by using polyethylene glycol. Proteins were identified by MALDI survey method, involving peptide mass fingerprinting (PMF) and MS/MS. In the MS survey method, spectra were recorded in the mass range of 800–4,000 m/z for 60 s. MS/MS analysis was performed in a data-dependent manner for the top 7 peptides with higher relative intensity for 30 s each and product ion mass range was set to 100 to 1,500 m/z. For protein identification, data were processed by PLGS software, searched against a UniProt human database with mass tolerance set to 100 ppm, carbamidomethylation as fixed modification and methionine oxidation as variable modification. Further, amino acid sequences corresponding to tryptic peptide masses were identified and subjected to BLAST homology searches to rule out alternative protein identifications.

Western blot analysis

The total protein extract (10 μg) from both normal and IDC samples were resolved onto 12.5% SDS-PAGE gels by using Mini PROTEAN^™ Tetra cell (Bio-Rad, Hercules, CA, USA). The proteins were then transferred on to PVDF membranes and blocked for 1 h with 5% membrane blocking agent (GE Healthcare UK Limited) prepared in TBS, and then incubated with antibodies either 1:3000 annexin A5 (ab14196, Abcam), 1:500 HSP70 (ab31010, Abcam), 1:5000 GAPDH (ab8245, Abcam), 1:3000 alpha-1-antitrypsin (ab7633, Abcam), 1:4000 fibrinogen (ab6666, Abcam), 1:4000 apolipoprotein A1 (ab7613, Abcam), 1:5000 SOD2 (ab16956, Abcam), 1:5000 S100 A10 (ab52272, Abcam), or 1:7000 beta-actin (A1978, Sigma-Aldrich), followed by 1:5000 of the appropriate biotin conjugated secondary antibody and 1:2000 streptavidin conjugated HRP. Protein bands were visualized using SIGMAFAST^™ DAB peroxidase substrate. Blot images were acquired by using GS800 calibrated densitometer (Bio-Rad) and images were analyzed by using Quantity One software (Bio-Rad). In addition, Western blots were stained with coomassie stain to assess protein load in Western blot analysis as described by Welinder et al. (2011).

Statistical analysis

LC-MS^E data were analyzed according to Cheng et al. (2009). Scatter plot comparison of log2 transformed peptide peak intensities obtained from technical replicates of representative four proteins produced a correlation coefficient of 0.981 (Supplementary Fig. S1: Supplementary data are available online at www.liebertpub.com/omi). A Students t-test was performed for the proteins that showed fold change of 1.5 and above. The fold change was considered significant if p value is less than 0.05 by Students t-test. For 2DE analysis, the density of each protein was expressed as mean±SD. The statistical significance was established by Student's t-test. Differences were considered significant if p<0.05.

Results

Protein identification by LC-MS^E method

A total of 118 protein hits were identified from a pooled sample of six IDC tissues by LC-MS^E method and details are listed in Table 2 and Supplementary Table 1. Of these, 26 proteins (22%) were overexpressed in tumor in comparison with the pooled control tissue samples; details of overexpressed proteins are listed in Table 2. Average fold change of overexpressed protein was 2.01 with expression range between 1.5 to 3.24. Annotated MS/MS spectra of two peptides for each of the differentially expressed proteins, as well as unique peptide sequence have been provided in Supplementary Files 1 and 2.

Table 2.

LC-MS^E Method Identification of Proteins Differential Expression in IDC Compared to Control Tissue

Sl. N	Accession number ^a	Protein name	FE ^b	MW ^c (Da)	PI ^d	Plgs ^e score	SC ^f
1	P60709	Actin cytoplasmic 1	3.24	41709	5.14	9550	70.13
2	P08758	Annexin A5	2.86	35914	4.73	1987	35.63
3	P01009	Alpha 1 antitrypsin	2.79	46707	5.24	2091	30.38
4	P68363	Tubulin alpha 1B chain	2.68	50119	4.76	1252	19.51
5	P04792	Heat shock protein beta 1	2.32	22768	5.96	3264	57.56
6	P21695	Glycerol 3 phosphate dehydrogenase	2.29	37543	5.76	714	35.24
7	P02675	Fibrinogen beta chain	2.19	55892	8.25	4540	45.21
8	O14558	Heat shock protein beta 6	2.15	17124	5.95	722	35.63
9	P00915	Carbonic anhydrase 1	2.14	28852	6.66	1210	27.20
10	Q6NZI2	Polymerase I and transcript release factor	1.96	43449	5.34	1839	27.44
11	P12110	Collagen alpha 2 VI chain	1.92	108511	5.79	960	14.62
12	P04075	Fructose bisphosphate aldolase A	1.91	39395	8.06	167	25.00
13	P15090	Fatty acid binding protein adipocyte	1.90	14709	7.04	4375	31.82
14	P68104	Elongation factor 1 alpha 1	1.89	50109	9.34	186	4.33
15	P07585	Decorin	1.86	39721	8.68	1096	29.25
16	Q9NZN4	EH domain containing protein 2	1.79	61122	6.00	501	23.57
17	Q96Q06-2	Isoform 2 of Perilipin 4	1.77	140588	9.17	561	30.08
18	P30043	Flavin reductase	1.74	22105	7.49	816	31.55
19	P30041	Peroxiredoxin 6	1.69	25019	5.96	860	29.91
20	O60814	Histone H2B type 1 K	1.69	13881	10.7	6150	41.27
21	P02679	Fibrinogen gamma chain	1.66	51478	5.24	3826	38.19
22	Q06830	Peroxiredoxin 1	1.64	22096	8.24	1870	40.20
23	P61981	14-3-3 protein gamma	1.58	28284	4.61	508	8.91
24	P02647	Apolipoprotein A I	1.57	30758	5.43	510	36.70
25	P04083	Annexin A1	1.56	38689	6.64	2955	39.31
26	P62937	Peptidyl prolyl cis trans isomerase A	1.53	18000	7.85	3639	60.61

Serial numbers (Sl. N) from 1–26 proteins were differentially expressed and are arranged from highest fold-expression to lowest. ^aAccession number of protein is from UniProt database; ^bFE, fold expression calculated as ratio of tumor to control samples; ^cMW, molecular weight of protein in daltons; ^dPI, protein isoelectric point; ^ePlgs score, PLGS score (software generated score that defines the confidence of protein identification); ^fSC, sequence coverage (identified peptides sequence coverage with protein's full sequence in database).

Protein–protein interaction and coexpression study

The IDC proteome interaction was generated as evidence-based prediction (Fig. 2a). The global STRING generated protein network showed high connectivity among stress-related, cytoskeletal, and glycolysis pathway proteins. The prominent interaction network was found to be within proteins including SOD1, SOD2, HSP70, thioredoxin, peroxiredooxin1, triosphosphate isomerase, GAPDH, annexins, beta-actin, and apolipoprotein A1. Further, IDC interactome was analyzed for coexpression mapping (Fig. 2b) where S100 A10 coexpress with annexins. The significant coexpression was found between aldolase and GAPDH, as well as SOD1 and peroxiredoxin 1. Additionally, actin was also found to be coexpressed with tubulin, cofilin, aldolase and GAPDH.

FIG. 2.

Prediction of (a) protein interactions and (b) coexpression by STRING database search, respectively.

Functional analysis of IDC proteome

The majority IDC proteomes were related to ontology of cytoskeletal, stress response, and binding. In biological processes, the proteins related to metabolism, stress response, signal transduction, and developmental process were significantly enriched (Fig. 3a). In molecular function, the proteins belonging to structural molecule activity, protein binding, oxidoreductase activity, receptor activity, and hydrolase activity were selectively enhanced emphasizing on the biological process they are involved (Fig. 3b). In the protein class, cargo protein, cytoskeletal protein, signaling molecule, receptor, transferase, structural protein, and isomerase classes were enriched (Fig. 3c). In the pathway category, cytoskeletal regulation, glycolysis, integrin signaling pathways were enriched (Fig. 3d).

FIG. 3.

PANTHER classification of IDC proteins based on (a) biological process, (b) molecular functions, (c) protein class, and (d) molecular pathways.

Validation experiments

Protein identification by 2DE method

In the IDC proteome, a total of 30 proteins were identified, of which 25 proteins were found to be differentially regulated across all six samples (Fig. 4a and 4b). Average fold changes of upregulated proteins was found to be within the range of 1.5 to 3.31. Amongst differentially expressed proteins, 24 proteins were upregulated and one protein was downregulated, the details are listed in Table 3. Only five proteins including alpha-1-antitrypsin, actin, fibrinogen, annexin A5, and apolipoprotein A1 were found to be differentially expressed in both LC-MS^E and 2DE analysis. Some of these proteins were further validated by Western blotting.

FIG. 4.

(a) Localization of the 30 selected spots on the 2DE image. Spot numbers correspond to serial numbers in Table 3. (b) Cropped images from 2DE gels demonstrating the differential expression of selected proteins listed in Table 3. Circles indicate the spots of interest. Differential expression of proteins was found to be statistically significant with p value less than 0.05.

Table 3.

2DE Method Information of Protein Identification and Differential Expression in IDC in Comparison with Pooled Control Samples

Sl No	Protein	ACCN	Ratio (T/C)	Mw(kDa)/PI	PLGS Score	Sequence Coverage
1	HSP70	P08107	3.31	70.0/5.3	7.8	33%
2	HSP70	P08107	3.31	70.0/5.2	7.8	33%
3	Serotransferrin	P02787	-	77.0/6.7	9.6	37%
4	Serotransferrin	P02787	-	77.0/6.7	9.6	35%
5	Serum albumin	P02768	-	69.3/5.9	8.1	14%
6	Serum albumin	P02768	-	69.3/5.9	9.4	29%
7	Serum albumin	P02768	-	69.3/5.9	8.1	22%
8	Vitamin D binding protein	P02774	1.5	52.9/5.2	5.7	11%
9	Vitamin D binding protein	P02774	1.5	52.9/5.2	6.8	13%
10	Alpha 1 antitrypsin	P01009	1.99	46.7/5.2	8.4	32%
11	Alpha 1 antitrypsin	P01009	1.97	46.7/5.2	8.5	41%
12	Thioredoxin isoform2	Q96J42-2	1.5	38.0/4.6	6.4	10%
13	Haptoglobin	P00738	1.51	45.1/5.1	3.0	36%
14	Actin	P60709	2.33	41.7/5.1	8.7	22%
15	Actin	P60709	2.33	41.7/5.1	8.7	19%
16	Fibrinogen gamma chain	P02679-2	1.5	49.4/5.6	6.2	24%
17	Fibrinogen gamma chain	P02679-2	1.5	49.4/5.6	7.1	33%
18	45 kDa Ca binding protein	Q9BRK5	1.5	41.7/4.6	5.8	16%
19	Mg transporter NIPA3	Q6NVV3	1.78	44.6/5.6	6.9	14%
20	ADP ribosyl hydrolase	Q8NDY3	0.53	40.0/5.6	6.6	15%
21	Annexin A5	P08758	2.77	35.9/4.7	9.4	50%
22	Isoform 3 of tropomyosin beta chain	P07951-3	2.21	28.6/4.4	6.6	50%
23	Isoform 3 of tropomyosin beta chain	P07951-3	1.5	28.6/4.4	2.3	35%
24	Putative selection and upkeep of intraepithelial T cells protein1	A8MVG2	1.5	25.3/5.5	7.3	9%
25	Complement C1q tumor necrosis factor related protein 6	Q9BXI9	1.56	28.6/5.6	6.3	13%
26	Transcription elongation factor A	Q5H9L2	1.59	23.2/4.5	6.2	13%
27	Apolipoprotein A I	P02647	1.5	30.7/5.4	8.9	54%
28	Glycodelin isoform 2	P09466-2	1.57	18.2/4.9	6.3	55%
29	Transthyretin	P02766	1.54	15.8/5.4	7.4	9%
30	Galectin	P09382	1.5	14.7/5.1	6.7	56%

Serial numbers in table corresponds to spot numbers in Figure 4a.

Western blot analysis

Western blot analysis was performed using antibodies against alpha1 antitrypsin, annexin A5, HSP70, apolipoprotein, fibrinogen, GAPDH, SOD2, beta-actin, and S100 A10. The differential expression of these proteins was found to be prominent across all tumor samples in comparison with the pooled control sample. Upregulation of these proteins in tumor was statistically significant with p value less than 0.01 as shown in Figure 5a and 5b.

FIG. 5.

(a) Western blot analysis of proteins (a. Alpha-1 antitrypsin, b. Annexin5, c. Apolipoprotein A1, d. Fibrinogen, e. GAPDH, f. SOD2, g. HSP70, h. beta-actin, and i. S100 A10) in clinical IDC breast tissue samples and adjacent control samples. Immunoblots images were acquired using GS800 calibrated densitometer and relative average expression levels of proteins were calculated (in error bar graph, 1 represents control and 2 represents breast tumor samples) with Quantity-One software (BioRad) and p values were less than 0.001. M, molecular weight marker (high range rainbow, product code RPN 756E, GE Healthcare Life Sciences). PC, pooled control breast tissue samples. T (T1, T2, T3, T4, T5, and T6), clinical IDC breast tissue samples. (b) Representative CBB stained blot (alpha 1 antitrypsin) shows the use of coomassie staining after the immunodetection of proteins on PVDF membranes to assess total protein load and the blotting efficiency. PC, pooled control breast tissue samples.

Discussion

Breast cancers are characterized based on protein expression profile, such as ER, PR hormone receptors, and HER2/neu oncogenes. Such characterization will aid in effective therapy (e.g., hormone therapy and Herceptin therapy). However, the lone protein expression status of ER or HER2/neu does not represent the ER dependent or HER2/neu dependent signaling pathways. Moreover, different subtypes of IDC differ in their morphology, clinical course, and response to therapy. Therefore, the objective of the present study was to identify differentially expressed proteins associated with ER positive and HER2/neu negative subtype that may serve as useful information in understanding the molecular mechanisms of carcinogenesis and progression. Widely used gel-based approaches have limitations such as being laborious and time consuming; several other liquid chromatography (LC) based quantification methods have been developed (Gygi et al., 1999; Ong et al., 2002; Ow et al., 2008). The evolved LC-MS^E is a parallel, unbiased, and continuous process where all peptide precursors are simultaneously separated and fragmented, regardless of intensity enabling identification of low abundant peptides that provide enhanced peptide coverage for identification. LC-MS^E technique uses an internal spiked standard with alternating cycles of scans at low and elevated collision energy to obtain both protein quantification and identification data set in a single LC-MS run (Silva et al., 2006).

LC-MS^E based quantification was performed to obtain comparative proteomic data set of six patients with ER positive, HER2/neu negative subtype IDC, and a corresponding pooled control sample. This study identified a total of a 118 proteins, of which 26 were differentially expressed. Out of these 26 proteins identified by LC-MS^E analysis, only five proteins including alpha-1-antitrypsin, actin, fibrinogen, annexin A5, and apolipoprotein A1 were also found to be differentially expressed in 2DE analysis with a similar trend of fold change as that of LC-MS^E (Table 3). Validation experiments employing Western blot analysis also confirmed the expression of proteins including beta-actin, alpha-1 antitrypsin, annexin A5, HSP70, apolipoprotein A1, fibrinogen, GAPDH, SOD2, and S100 A10. This proteomic study indicates upregulation of specific proteins involved in cytoskeletal organization, metabolism, and stress.

Cytoskeleton organization proteins

The actin and associated protein of cytoskeleton organization is critical for cell motility and migration. In this study, the IDC proteome consists of about 15% cytoskeletal-associated proteins, including beta actin, collagen alpha 1, collagen alpha 2, desmin, fibrinogen alpha, fibrinogen beta, fibrinogen gamma, gelsolin, lumican, mimecan, myosin, decorin, prolargin, profilin, transgelin, tubulin beta2A, tubulin beta, and vimentin. Beta actin is associated with cell migration in breast cancer (Vandermoere et al., 2007). Similarly, an abundance of fibrinogen and transgelin is associated with structural integrity and stabilization of actin cytoskeleton, respectively (Costantini et al., 1991; Gimona et al., 2003). Further, transgelin promotes migration and invasion of cancer stem cells (Lee et al., 2010). Many of these cytoskeletal proteins have been reported in IDC. However, the implication of cytoskeletal proteins in carcinogenesis is obvious in general, but their expression specificity for ER positive and HER2/neu negative IDC is not well defined in particular. In addition, actin binding protein tropomyosin mutations and their expressions are associated with neoplastic phenotype (Marleen et al., 2008), and actin-based cell migration is a result of complex interaction network of proteins in the tumor microenvironment (Marleen et al., 2008), which also supports the STRING interaction network and coexpression as shown in Figure 2a and 2b. Hydrolysis of extracellular matrix facilitates cell migration and evidently the hydrolase activity is significantly enriched in molecular function (Fig. 3b). The migratory mode from epithelial to mesenchymal transition is considerably dependent on cytoskeletal reorganization, and such reorganizations may be facilitated by coexpression (Fig. 2b) and expression levels of actin interacting proteins including cofilin, gelsolin, profiling, and transgelin. Many of these proteins are reported to be deregulated in different cancers (Marleen et al., 2008), indicating their potential role in regulation of actin-cytoskeleton.

Proteins involved in stress response

In this study, upregulation of HSP27, HSP20, HSP70 cognate, heat shock factor binding protein 1, calreticulin, protein disulfide isomerase, superoxide dismutase Cu Zn, superoxide dismutase Mn mitochondrial, catalase, thioredoxin, peroxiredoxin 1, and peroxiredoxin 6 was observed in tumor tissues. Heat shock proteins (HSPs) during stress prevent aggregation and promote refolding of damaged proteins. HSPs are overexpressed in many of human cancers and are implicated in tumor cell proliferation, differentiation, invasion, and recognition by the immune system (Ciocca et al., 2005). In ER-positive breast cancers, ER-regulated HSP27 is implicated in co-localization with ER in the cell nucleus (Ciocca et al., 1990) and is associated with both good and poor prognosis, as well as been proposed to be a marker of estrogenic endometrial response (Ciocca et al., 1993). Further, protein disulfide isomerase (PDI) is overexpressed on the cancer cell surface due to oxidative stress (Uehara et al., 2000). Apart from catalysis of isomerase activity, PDI is a molecular chaperone assisting in protein folding (Puig et al., 1994; Quan et al., 1995). Generation of reactive oxygen species (ROS) has been implicated in the etiology of many human diseases, including cancer. Estrogen-induced ROS are associated with transducing signals through redox sensitive transcription factors including Trx-1 and NF-kB, which have been implicated in cell migration and invasion of the breast cancer (Okoh et al., 2010). Many proteins are expressed as a cellular defense mechanism against agents that induce oxidative stress, including superoxide dismutase (SOD). SOD Mn gene polymorphism is associated with increased breast cancer risk and survival (Mitrunen et al., 2001), and such biologically active SOD mimicking compounds would be valuable in understanding signal mechanisms implicated in the progression of disease (Jenney et al., 1999). The conversion of H₂O₂ into water and oxygen is catalyzed by catalase and peroxiredoxins, the latter dominates catalysis in cytosol (Finkel et al., 1998).

Calcium binding proteins

Oxidative stress response triggers the activation of different intracellular pathways, resulting significant changes in calcium ions, pH homeostasis, and production of lipid second messengers (Finkel et al., 1998). These changes are sensed by annexins, which interact with specific lipid and protein moieties at the plasma membrane, thus contributing to stress response via regulation of various signaling pathways (Monastyrskaya et al., 2009). In this study, annexin A1 and annexin A5 were upregulated and these proteins have been implicated in signal transduction pathways associated with inflammation, cell differentiation, and cell proliferation of both glandular epithelium and squamous epithelium (Lim et al., 2007). Annexin A1 is functionally involved in basal-like breast cancer metastasis by regulating TGF-β signaling and actin cytoskeletal reorganization (de Graauw et al., 2010). Annexin A5 is involved in the development of skeletal muscles (Arcuri et al., 2002), and proteomic studies with clinical breast cancer tissues have already reported the overexpression of annexin A1 and annexin A5 (Hondermarck et al., 2001; Weitzel et al., 2010). However, other reports indicate the distinctive loss of annexin A1 in breast cancer (Cao et al., 2008; Shen et al., 2005). Similar to annexins, S100 proteins are calcium binding proteins associated with intracellular and extracellular functions in diseases. These S100 proteins interact with many other proteins, including kinases and actin etc., and are involved in signaling to regulate cell proliferation and differentiation (Miwa et al., 2000). Coexpressed annexin A2 and S100 A10 form a heterotetrameric complex and interaction of cathepsin B with such complex on the surface of tumor cells could facilitate tumor invasion and metastasis (Mai et al., 2000). Further, the annexin A2 and S100 A10 complex is directly involved in reduction reactions and is a substrate for thioredoxin (Kwon et al., 2005). Interaction of S100 A10 and 14-3-3 proteins with BAD could attenuate the pro-apoptotic signals (Hsu et al., 1997), and annexin A1 through interaction with IKK (IkB kinase enzyme complex) is associated with the activation of transcription factor NF-kB-mediated metastasis in breast cancer (Bist et al., 2011). In the absence of HER2/neu, we speculate the interaction of S100 A10 and RAGE may play a key in tumorigenesis via ROS generation and NF-kB activation (Korwar et al., 2012).

Metabolic process

Elevated glycolysis is a characteristic phenomenon of tumors for energy requirement as survival, growth, and invasion is concerned. A significant percentage of proteins associated with metabolic processes were enriched in PANTHER classification by their biological process as shown in Figure 3a. Glycolytic proteins including ENOA, ALDOA, GAPDH, TPIS, and PGK1 were found to be upregulated in IDC tumors as reported in earlier studies (Hondermarck et al., 2001; Weitzel et al., 2010). Further, LDHB is overexpressed to produce NAD⁺, which is essential for glycolysis by converting pyruvate to lactate. These upregulated metabolic proteins have been related to additional functions in tumour, such as GAPDH, which is associated with oxidative stress, tubulin binding, mRNA stability, vesicular secretory transport, and DNA repair (Durrieu et al., 1987; Sirover et al., 2005; Zhou et al., 2008).

In conclusion, proteomic profiling of IDC resulted in the identification of 118 proteins, of which 26 proteins were differentially expressed. STRING interaction analysis revealed the prominent interaction of annexins with HSP27, beta-actin, S100 A10, and apolipoprotein A1. These proteins, along with S100 A10, may serve as a signature for IDC with ER positive/HER2/neu negative subtype.

Footnotes

Acknowledgment

The authors are grateful to Dr. Rama Sivaram for helping in procuring clinical breast cancer samples from Jehangir Hospital, Pune. This work was supported by CSIR network project NWP0004. We would also like to thank Dr. Surekha Zingde, ACTREC, for improving the quality of the manuscript, Dr. Uttara Joshi, Dr. N S. Jalnapurkar for helping in histological analysis, and Dr. Jomon Joseph, NCCS, for his generous donation of anti-actin antibody. AMK would like to thank UGC, New Delhi for research scholarship, Yashwant Kumar for helping in acquiring LC-MS^E data, and Dr. Bhushan B. Dholakia for critical reading of manuscript.

Author Disclosure Statement

The authors have no conflict of interest to state.

References

Arcuri

, Giambanco

, Bianchi

, Donato

. 2002. Annexin V, annexin VI, S100A1 and S100B in developing and adult avian skeletal muscles. Neuroscience, 109:371–388.

Bist

, Leow

, Phua

et al. 2011. Annexin-1 interacts with NEMO and RIP1 to constitutively activate IKK complex and NF-kappa B: Implication in breast cancer metastasis. Oncogene, 30:3174–3185.

Cao

, Li

, Edelweiss

et al. 2008. Loss of annexin A1 expression in breast cancer progression. Appl Immunohistochem Mol Morphol, 16:530–534.

Cheng

, Blackburn

, Lin

, Goshe

, Williamson

. 2009. Absolute protein quantification by LC/MS(E) for global analysis of salicylic acid-induced plant protein secretion responses. J Proteome Res, 8:82–93.

Ciocca

, Stati

, Decastro

MMA

. 1990. Colocalization of estrogen and progesterone receptors with an estrogen-regulated heat-shock protein in paraffin sections of human breast and endometrial cancer tissue. Breast Cancer Res Treat, 16:243–251.

Ciocca

, Oesterreich

, Chamness

, Mcguire

, Fuqua

SAW

. 1993. Biological and clinical implications of heat-shock protein 27000 (Hsp27). A review. J Natl Cancer Inst, 85:1558–1570.

Ciocca

, Calderwood

. 2005. Heat shock proteins in cancer: Diagnostic, prognostic, predictive, and treatment implications. Cell Stress Chaperones, 10:86–103.

Costantini

, Zacharski

, Memoli

, Kisiel

, Kudryk

, Rousseau

. 1991. Fibrinogen deposition without thrombin generation in primary human breast-cancer tissue. Cancer Res, 51:349–353.

de Graauw

, van Miltenburg

, Schmidt

et al. 2010. Annexin A1 regulates TGF-beta signaling and promotes metastasis formation of basal-like breast cancer cells. Proc Natl Acad Sci, 107:6340–6345.

10.

Durrieu

, Berniervalentin

, Rousset

. 1987. Microtubules bind glyceraldehyde-3-phosphate dehydrogenase and modulate its enzyme-activity and quaternary structure. Arch Biochem Biophys, 252:32–40.

11.

Finkel

. 1998. Oxygen radicals and signaling. Curr Opin Cell Biol, 10:248–253.

12.

Gimona

, Kaverina

, Resch

, Vignal

, Burgstaller

. 2003. Calponin repeats regulate actin filament stability and formation of podosomes in smooth muscle cells. Mol Biol Cell, 14:2482–2491.

13.

Gygi

, Rist

, Gerber

, Turecek

, Gelb

, Aebersold

. 1999. Quantitative analysis of complex protein mixtures using isotope-coded affinity tags. Nat Biotechnol, 17:994–999.

14.

Hanahan

, Weinberg

R.A.

2000. Hallmarks of cancer. Cell, 100:57–70.

15.

Hondermarck

, Vercoutter-Edouart

, Revillion

et al. 2001. Proteomics of breast cancer for marker discovery and signal pathway profiling. Proteomics, 1:1216–1232.

16.

Hsu

, Kaipia

, Zhu

, Hsueh

AJW

. 1997. Interference of BAD (Bcl-xL/Bcl-2-associated death promoter)-induced apoptosis in mammalian cells by 14-3-3 isoforms and P11. Mol Endocrinol, 11:1858–1867.

17.

Jemal

, Bray

, Center

, Ferlay

, Ward

, Forman

. 2011. Global cancer statistics. CA Cancer J Clin, 61:69–90.

18.

Jenney

, Verhagen

MFJM

, Cui

, Adams

MWW

. 1999. Anaerobic microbes: Oxygen detoxification without superoxide dismutase. Science, 286:306–309.

19.

Korwar

, Bhonsle

, Chougale

et al. 2012. Analysis of AGE modified proteins and RAGE expression in invasive ductal carcinoma. Biochem Biophys Res Commun, 16:490–494.

20.

Kwon

, Yoon

, Jeong

, Rhee

, Waisman

. 2005. Annexin A2-S100A10 heterotetramer, a novel substrate of thioredoxin. J Biol Chem, 280:23584–23592.

21.

Lee

, Han

, Park

, Song

, Kim

. 2010. Transgelin promotes migration and invasion of cancer stem cells. J Proteome Res, 9:5108–5117.

22.

Lim

LHK

, Pervaiz

. 2007. Annexin 1: The new face of an old molecule. FASEB J, 21:968–975.

23.

, Finley

, Waisman

, Sloane

. 2000. Human procathepsin B interacts with the annexin II tetramer on the surface of tumor cells. J Biol Chem, 275:12806–12812.

24.

Marleen Van Troys

, Vandekerckhove

, Ampe

. 2008. Actin-Binding Proteins and Disease. New York: Springer.

25.

, Dong

, Muruganujan

, Gaudet

, Lewis

, Thomas

. 2010. PANTHER version 7: Improved phylogenetic trees, orthologs and collaboration with the Gene Ontology Consortium. Nucleic Acids Res, 38:D204–D210.

26.

Mitrunen

, Sillanpaa

, Kataja

et al. 2001. Association between manganese superoxide dismutase (MnSOD) gene polymorphism and breast cancer risk. Carcinogenesis, 22:827–829.

27.

Miwa

, Uebi

, Kawamura

. 2008. S100-annexin complexes. Biology of conditional association. FEBS J, 275:4945–4955.

28.

Monastyrskaya

, Babiychuk

, Draeger

. 2009. The annexins: Spatial and temporal coordination of signaling events during cellular stress. Cell Mol Life Sci, 66:2623–2642.

29.

Okoh

, Deoraj

, Roy

. 2011. Estrogen-induced reactive oxygen species-mediated signalings contribute to breast cancer. Biochim Biophys Acta, 1815:115–133.

30.

Ong

, Blagoev

, Kratchmarova

et al. 2002. Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol Cell Proteomics, 1:376–386.

31.

Ong

, Mann

. 2006. A practical recipe for stable isotope labeling by amino acids in cell culture (SILAC) Nature Protocols, 1:2650–2660.

32.

, Cardona

, Taton

et al. 2008. Quantitative shotgun proteomics of enriched heterocysts from Nostoc sp. PCC 7120 using 8-plex isobaric peptide tags. J Proteome Res, 7:1615–1628.

33.

Perou

, Sorlie

, Eisen

et al. 2000. Molecular portraits of human breast tumours. Nature, 406:747–752.

34.

Puig

, Gilbert

. 1994. Protein disulfide-isomerase exhibits chaperone and anti-chaperone activity in the oxidative refolding of lysozyme. J Biol Chem, 269:7764–7771.

35.

Quan

, Fan

, Wang

. 1995. Independence of the chaperone activity of protein disulfide-isomerase from its thioredoxin-like active-site. J Biol Chem, 70:17078–17080.

36.

Rezaul

, Thumar

, Lundgren

et al. 2010. Differential protein expression profiles in estrogen receptor-positive and -negative breast cancer tissues using label-free quantitative proteomics. Genes Cancer, 1:251–271.

37.

Shen

, Chang

, Chen

et al. 2005. Loss of annexin A1 expression in human breast cancer detected by multiple high-throughput analyses. Biophys Res Commun, 326:218–227.

38.

Silva

, Gorenstein

, Li

, Vissers

JPC

, Geromanos

. 2006. Absolute quantification of proteins by LCMSE. A virtue of parallel MS acquisition. Mol Cell Proteomics, 5:144–156.

39.

Simpson

, Dorow

. 2001. Cancer proteomics: From signaling networks to tumor markers. Trends Biotechnol, 19:S40–S48.

40.

Sirover

. 2005. New nuclear functions of the glycolytic protein, glyceraldehyde-3-phosphate dehydrogenase, in mammalian cells. J Cell Biochem, 95:45–52.

41.

Slamon

, Godolphin

, Jones

et al. 1989. Studies of the Her-2/Neu proto-oncogene in human breast and ovarian cancer. Science, 244:707–712.

42.

Sorlie

, Perou

, Tibshirani

et al. 2001. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci, 98:10869–10874.

43.

Szklarczyk

, Franceschini

, Kuhn

, Simonovic

, Roth

, Minguez

et al. 2011. The STRING database in 2011: functional interaction networks of proteins, globally integrated and scored. Nucleic Acids Res, 39:D561–D568.

44.

Tanaka

, Uehara

, Nomura

2000. Up-regulation of protein-disulfide isomerase in response to hypoxia/brain ischemia and its protective effect against apoptotic cell death. Biol Chem, 275:10388–10393.

45.

Vandermoere

, El Yazidi-Belkoura

, Demont

, Slomianny

, Antol

, Lemoine

et al. 2007. Proteomics exploration reveals that actin is a signaling target of the kinase Akt. Mol Cell Proteomics, 6:114–124.

46.

van't Veer

L. J.

, Dai

H. Y.

, van de Vijver

M. J.

, He

Y. D. D.

, Hart

A. A. M.

, Mao

et al. 2002. Gene expression profiling predicts clinical outcome of breast cancer. Nature, 415:530–536.

47.

Weitzel

L. R. B.

, Byers

, Allen

, Finlayson

, Helmke

S. M.

, Hokanson

J. E.

et al. 2010. Discovery and verification of protein differences between Er positive/ Her2/neu negative breast tumor tissue and matched adjacent normal breast tissue. Breast Cancer Res Treat, 124:297–305.

48.

Welinder

, Ekblad

2011. Coomassie Staining as Loading Control in Western Blot Analysis. J Proteome Res, 4:1416–1419.

49.

Zhang

D. H.

, Tai

L. K.

, Wong

L. L.

, Chiu

L. L.

, Sethi

S. K.

, Koay

E. S. C.

2005. Proteomic study reveals that proteins involved in metabolic and detoxification pathways are highly expressed in HER-2/neu-positive breast cancer. Mol Cell Proteomics, 4:1686–1696.

50.

Zhou

, Yi

, Stoffer

J. B.

, Bonafe

, Gilmore-Hebert

, McAlpine

et al. 2008. The multifunctional protein glyceraldehyde-3-phosphate dehydrogenase is both regulated and controls colony-stimulating factor-1 messenger RNA stability in ovarian cancer. Mol Cancer Res, 6:1375–1384.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.65 MB

0.00 MB

0.16 MB

0.07 MB

Proteomic Profiling and Interactome Analysis of ER-Positive/HER2/ neu Negative Invasive Ductal Carcinoma of the Breast: Towards Proteomics Biomarkers

Abstract

Abstract

Introduction

Materials and Methods

Chemicals

Clinical breast tumor tissue

Sample preparation

In-solution tryptic digestion for LC-MSE analysis

LC-MSE analysis

Data processing and database searching

Protein–protein interaction and coexpression analysis of IDC proteome

Functional analysis of IDC proteome

2DE and image analysis

In-gel tryptic digestion

MALDI MS/MS analysis

Western blot analysis

Statistical analysis

Results

Protein identification by LC-MSE method

Protein–protein interaction and coexpression study

Functional analysis of IDC proteome

Validation experiments

Protein identification by 2DE method

Western blot analysis

Discussion

Cytoskeleton organization proteins

Proteins involved in stress response

Calcium binding proteins

Metabolic process

Footnotes

Acknowledgment

Author Disclosure Statement

References

Supplementary Material

In-solution tryptic digestion for LC-MS^E analysis

LC-MS^E analysis

Protein identification by LC-MS^E method