Introduction
Acquired resistance to hormone therapy remains a major challenge in the treatment of estrogen receptor positive (ER(+)) metastatic breast cancers. Previous studies have demonstrated that ER (+) breast cancer can escape anti-estrogen actions by up-regulating other signaling pathways involved in cell survival and proliferation. Enhanced signaling via growth factor receptors, such as EGFR [
1] and HER2 [
2], has been implicated in acquired resistance to endocrine therapy. Activation of downstream intracellular signaling like the MAPK pathway and the PI3K/Akt pathway has also been linked to hormone resistance [
3,
4]. The cross-talk between ER and such alternative signaling pathways are believed to enable breast cancer to evade the antiproliferative effects of anti-estrogens [
5]. This knowledge has led to numerous treatment strategies combining endocrine and targeted inhibitor therapies. However, early clinical trials of EGFR- and ERBB2-targeted inhibitors (for example, gefitinib, erlotinib, trastuzamab, and lapatinib) or m-TOR inhibitors (everolimus and temsirolimus) in combination with endocrine therapies have yielded mixed results [
6‐
8]. It is likely that cross-talk and negative feedback loops may result in cellular resistance to individual inhibitors [
9]. Additional therapies targeting converging points of shared signaling pathways, such as MYC and cyclin D1-CKD4, may be more effective at blocking proliferation in resistant breast cancers [
10].
Current understanding of endocrine resistance mechanisms is largely based on the study of relatively few genes. Integrative approaches that examine gene expression in the genomic and proteomic context may lead to the discovery of previously unconsidered mechanisms for the modulation of therapeutic responses. The current study employed a quantitative proteomic strategy to capture global changes in protein expression in a tamoxifen resistant cell line derived from the wild type MCF-7 parental cells.
In vitro studies of tamoxifen resistance have provided valuable foundational data that can be translated into
in vivo and clinical applications [
11‐
13]. The most widely used and best characterized cell line for study of acquired tamoxifen resistance has been the MCF-7 variants, from which much of our current understanding of the mechanisms of hormone resistance has derived [
13,
14]. While numerous earlier studies in other laboratories have demonstrated that tamoxifen resistant breast cancer cell lines were generated by long term exposure of MCF-7 cells to 10
-6 to 10
-7M 4-OH Tam over a period of 6 to 12 months, adaptive signatures of the resulting resistant phenotypes may vary with different experimental conditions employed. For example, EGFR expression was reported to be 10-fold higher in one tamoxifen-resistant model [
14] but not in other models [
15,
16]. It has also been shown [
13] that use of dextran coated charcoal-stripped (DCC) serum in tamoxifen treatment may introduce, in addition to adaptive changes of the cells in response to tamoxifen, effects of long term estrogen deprivation (LTED), thus complicating the interpretation of molecular signals of resistance development for tamoxifen. Moreover, in estrogen deprived medium, tamoxifen can act as an agonist [
17] towards ER, adding another complicating factor to the mechanistic interpretation of tamoxifen resistance. We used a phenol-red free DMEM medium containing 5% FBS so that the background estrogen level is in a range that is unlikely to induce adaptive changes due to estrogen deprivation and to minimize the agonistic action of tamoxifen in ER(+) breast cancer cells.
In this study, we examined global proteomic alterations of the tamoxifen resistant cell line vs the parental MCF-7 cells using an isobaric labeling approach combined with a high resolution tandem mass spectrometry instrument for relative quantitative analysis. Our proteomics data demonstrated extensive adaptive changes in the proteome involving hundreds of significantly up- and down-regulated proteins. In particular, results from this study revealed the overexpression of multiple tumorigenic, pro-metastatic proteins and the down-regulation of ER mediated signaling pathways. These findings provide novel insights into the complex events of the adaptive signaling network occurring during the acquisition of tamoxifen resistance in breast cancer cells and highlight the role of S100P in conferring both resistance and enhanced migration.
Materials and methods
Cell culture
MCF-7 cell line was purchased from ATCC (ATCC #HTB-22, Manassas, VA, USA), and routinely cultured in phenol red-free DMEM medium supplemented with 5% FBS, 4 mM glutamine, 1 mM sodium pyruvate, 100 IU/mL penicillin, 100 μg/mL streptomycin and 0.25 μg/mL amphotericin. Tamoxifen resistant variant cells (MCF-7-TamR) derived from MCF-7 cells were continuously cultured in the medium as described above containing additional 10-7 M 4-OH Tam (Sigma-Aldrich, St Louis, MO, USA) for at least six months, along with the parental MCF-7-cells under identical culture conditions except that the control cells were treated with 0.1% ethanol. The two cell lines were grown side by side at all times. Cultures were maintained in 5% carbon dioxide at a temperature of 37°C.
Cell growth assay
For growth assay in the presence of 10-7 M 4-OH Tam, MCF-7 cells cultured with 10-7 M 4-OH Tam for zero to six months were plated in six-well plates at a density of 50,000 in each well in 5% FBS DMEM medium. The cells were then treated with 10-7 M 4-OH Tam for five days, while equal treatment volumes of ethanol were used as a vehicle control. Cell numbers were counted with a Coulter instrument (Beckman-Coulter, Indianapolis, IN, USA). The ratio of 4-OH Tam treated cell numbers to vehicle treated cell numbers was defined as survival ratio. Experiments were conducted in triplicate and data represented as mean ± SD.
For dose-dependent proliferation assays, MCF-7-TamR and MCF-7-control cells were seeded in 96-well plate with a density 3,000 per well and treated with varying concentrations (10-7 to 10-5 M) of 4-OH tamoxifen for five days; 0.1% ethanol was used as a vehicle control. Alamar Blue dye (Invitrogen, Grand Island, NY, USA) was added and incubated for 2 h at 37°C, protected from light. A Synergy 2 microplate reader (BioTek, Winooski, VT, USA) was used to record fluorescence using an excitation wavelength at 560 nm and emission wavelength at 590 nm. The ratio of 4-OH Tam-treated cell fluorescence intensity to that of vehicle treated cells was determined as the survival ratios in triplicate experiments. Data were represented as mean ± SD.
Cell lysis
MCF-7-TamR and MCF-7-control cells were cultured to 80% confluent in the medium as described above, and washed with cold Hank's Buffered Salt Solution (HBSS) for three times, then collected with a cell scraper. NP40 cell lysis buffer (Invitrogen) containing additional 1 mM of phenylmethylsulfonyl fluoride (PMSF) and protease inhibitor cocktail (Sigma) was used to extract total cellular proteins. The concentration of proteins was measured with BCA assay (Pierce Biotechnology, Rockford, IL, USA). The cell lysis was stored at -80°C before further processing.
Trypsin digestion
Protein samples were digested with sequencing grade modified trypsin (Promega Corp. Madison, WI, USA) according to the manufacturer's instructions. Briefly, to aliquots of 100 μg of protein sample was added 45 μL of 200 mM triethyl ammonium bicarbonate (TEAB) and the final volume was adjusted to 100 μL with ultrapure water. A total of 5 μL of 200 mM tris(2-carboxyethyl)phosphine (TCEP) was added and the resulting mixture was incubated for 1 h, then 5 μL of 375 mM iodoacetamide was added and the mixture was incubated for 30 minutes without light. After incubation, 1 mL of pre-chilled acetone was added and the precipitation was allowed to proceed overnight. The acetone-precipitated protein pellets were suspended with 100 μL of 200 mM TEAB and 2.5 μg of trypsin was added to digest the sample overnight at 37°C.
Tandem mass tags TMT6 (Thermo Scientific, Rockford, IL, USA) with different molecular weights (126 to approximately 131 Da) were applied as isobaric tags for relative and absolute quantification. According to the manufacturer's protocols, the digested samples were individually labeled with TMT6 reagents for 1 h as follows: three 100-μg aliquots of digested MCF-7-control peptides were each labeled with a different isobaric tag (TMT126, 127 and 128, respectively). Likewise, 100-μg aliquots of peptides from MCF-7-TamR cells were labeled with TMT129, 130, and 131 mass tags, respectively. The labeling reaction was quenched with 5% hydroxylamine. Finally, the six labeled peptide aliquots were combined for subsequent fractionation.
Fractionation of labeled peptide mixture using a strong cation exchange column
The combined TMT labeled peptide mixture was fractionated with a strong cation exchange column (SCX) (Thermo Scientific) on a Shimadzu 2010 HPLC equipped with a UV detector (Shimadzu, Columbus, MD, USA). Mobile phase consists of buffer A (5 mM KH2PO4, 25% acetonitrile, pH 2.8) and buffer B (buffer A plus 350 mM KCl). The column was equilibrated with Buffer A for 30 minutes before sample injection. The mobile phase gradient was set as follows at a flow rate of 1.0 mL/minute: (a) 0 to 10 minutes: 0% buffer B; (b) 10 to 40 minutes: 0% to 25% Buffer B, (c) 40 to 45 minutes: 25% to 100% Buffer B; (d) 45 to 50 minutes: 100% buffer B; (e) 50 to 60 minutes: 100% to 0% buffer B; (f) 60 minutes to 90 minutes: 0% buffer B. A total of 60 fractions were initially collected, lyophilized and combined into 15 final fractions based on SCX chromatographic peaks.
Desalination of fractionated samples
A C18 solid-phase extraction (SPE) column (Hyper-Sep SPE Columns, Thermo-Fisher Scientific, Waltham, MA, USA) was used to desalt all collected fractions. The combined 15 fractions were each adjusted to 1-mL final volume containing 0.25% (v/v in water) trifluoroacetic acid (TFA, Sigma). The C18 SPE columns were conditioned before use by filling them with 1 mL acetonitrile and allowing the solvent to pass through the column slowly (approximately three minutes). The columns were then rinsed three times with 1 mL 0.25% (v/v in water) TFA solution. The fractions were loaded on to the top of the SPE cartridge and allowed to elute slowly. Columns were washed four times with 1-mL 0.25% TFA aliquots before the peptides were eluted with 3 × 400 μL of 80% acetonitrile/0.1% formic acid (aqueous).
LC-MS/MS analysis on LTQ-Orbitrap
Peptides were analyzed on an LTQ-Orbitrap XL instrument (Thermo-Fisher Scientific) coupled to an Ultimate 3000 Dionex nanoflow LC system (Dionex, Sunnyvale, CA, USA). High mass resolution was used for peptide identification and high energy collision dissociation (HCD) was employed for reporter ion quantification. The RP-LC system consisted of a peptide Cap-Trap cartridge (0.5 × 2 mm) (Michrom BioResources, Auburn, CA, USA) and a pre-packed BioBasic C18 PicoFrit analytical column (75 μm i.d. × 15 cm length, New Objective, Woburn, MA, USA) fitted with a FortisTip emitter tip. Samples were loaded onto the trap cartridge and washed with mobile phase A (98% H2O, 2% acetonitrile and 0.1% formic acid) for concentration and desalting. Subsequently, peptides were eluted over 180 minutes from the analytical column via the trap cartridge using a linear gradient of 6 to 100% mobile phase B (20% H2O, 80% acetonitrile and 0.1% formic acid) at a flow-rate of 0.3 μL/minute using the following gradient: 6% B for 5 minutes; 6 to 60% B for 125 minutes; 60 to 100% B for 5 minutes; hold at 100% B for 5 minutes;100 to 6% B in 2 minutes; hold at 6% B for 38 minutes.
The LTQ-Orbitrap tandem mass spectrometer was operated in a data-dependent mode. Briefly, each full MS scan (60,000 resolving power) was followed by six MS/MS scans where the three most abundant molecular ions were dynamically selected and fragmented by collision-induced dissociation (CID) using a normalized collision energy of 35%, and the same three molecular ions were also scanned three times by HCD-MS2 with collision energy of 45%. MS scans were acquired in profile mode and MS/MS scans in centroid mode. LTQ-Orbitrap settings were as follows: spray voltage 2.0 kV, 1 microscan for MS1 scans at 60, 000 resolution (fwhm at m/z 400), microscans for MS2 at 7,500 resolution (fwhm at m/z 400); full MS mass range, m/z 400 to 1,400; MS/MS mass range, m/z 100 to 2,000. The "FT master scan preview mode", "Charge state screening", "Monoisotopic precursor selection", and "Charge state rejection" were enabled so that only the 2+, 3+ and 4+ ions were selected and fragmented by CID and HCD.
Database search and TMT quantification
The protein search algorithm used was Mascot (Matrix Science, Boston, MA, USA). Mascot format files were generated by the Proteome Discoverer 1.2 software (Thermo-Fisher Scientific) using the following criteria: database, IPI_Human.fasta.v3.77; enzyme, trypsin; maximum missed cleavages, 2; Static modifications, carbamidomethylation (+57 Da), N-terminal TMT6plex (+229 Da), lysyl TMT6plex (+229 Da). Dynamic modifications, N-terminal Cln- pyro-Glu(+17Da); methionine oxidation (+16 Da); STY phosphorylation (+80 Da); MS peptide tolerance was set at 15 ppm; MS/MS tolerance at 0.05 Da. Peptides reported by the search engine were accepted only if they met the false discovery rate of P < 0.05 (target decoy database). For TMT quantification, the ratios of TMT reporter ion abundances in MS/MS spectra generated by HCD (up to six reporter ions ranging from m/z 126.12 to m/z 131.14) from raw data sets were used to calculate fold changes in proteins between control and treatment.
Quantitative RT-PCR
Confirmation of selected targets identified in proteomic analysis
Total RNA from MCF-7-TamR and control cells was extracted using a PureLink total RNA purification system (Invitrogen) and quantitatively analyzed with a nanodrop spectrophotometer (Thermo Scientific). The reverse transcription was carried out with a SuperScript first-strand synthesis system (Invitrogen) using Oligo(dT)12-18 primers. The primer pairs used to amplify the genes were designed using the online tool of Oligo Perfect Designer (Invitrogen), and beta actin (actb) was employed as an internal standard. Primer specificity was confirmed by BLAST analysis. For real-time PCR analyses, a MyiQ real time PCR detection system (BioRad, Hercules, CA, USA) and a SYBR GreenER qPCR supermix kit (Invitrogen) were used as follows: 50°C for 2 minutes, 95°C for 8 minutes and 30 seconds, and 50 cycles (15 seconds at 95°C, 1 minute at 60°C). The data were analyzed with a normalized gene expression method (ΔΔCt) using the iQ5 Optical System Software (BioRad), and the gene actb was used as a reference for normalization. All experiments were repeated three times independently.
ER regulated gene transcripts
MCF-7-control or MCF-7-TamR cells were seeded at a density of 2 × 10
6 cells per 25 cm
2 culture flask in phenol red-free 5% FBS-DMEM. On the following day, cells were washed in PBS and media were changed to phenol red-free media supplemented with 5% CS-DMEM and grown to 50 to 80% confluency for 48 h before treatment with vehicle (DMSO), 17β-estradiol (100 pM), or tamoxifen (100 nM). RNA was extracted using QiaShredders (QIAGEN, Valencia, CA, USA) and purified on RNeasy columns (QIAGEN) according to the manufacturer's protocol. RNA quality and concentration were determined by absorbance at 260 and 280 nm. Then 2 μg of total RNA was reverse transcribed using the iScript kit (Bio-Rad Laboratories). The levels of ERα, PgR and SDF-1 transcripts were determined using real-time quantitative PCR. The primer sequences are as follows (sense and antisense, respectively): PgR, 5'-TACCCGCCCTATCTCAACTACC-3', 5'-TGCTTCATCCCCACAG-ATTAAACA-3'; SDF-1, 5'-AGTCAGGTGGTGGCTTAACAG-3', 5'-AGAGGAGGTGAAGGCAGTGG-3'; and ER
α
, 5'-GGCATGGTGGAGATCTTCGA-3', 5'-CCTCTCCCTGCAGATTCATCA-3', Actin, 5'- TGA GCG CGG CTA CAG CTT -3', 5'-CCTTAATGTCACACACGATT-3'. The PCR reaction was carried out as follows: step 1: 95°C 3 minutes; step 2: for 40 cycles 95°C 20 seconds, 60°C 1 minute; step 3: 70°C 10 seconds, held at 4°C. Each reaction tube contained: 12.5 μL 2 × SYBR Green supermix + 6.5 μL nuclease-free water + 1 μL 0.1 μg/μL primer (pair) + 5 μL cDNA (0.2 μg/μL). Genes were amplified in triplicate. Data were analyzed by comparing relative target gene expression to actin control. Relative gene expression was analyzed using 2-ΔΔCt method [
21].
Western blot
MCF-7-control or MCF-7-TamR cells were seeded in 10 cm2 plates at a density of 60 to 70% confluence (5 to 10 × 106 cells) and were allowed to grow for three days until they approached 80 to 90% confluence. The media was then removed and the cells were scraped into 1 mL of PBS plus 3 mM EDTA. The cell suspensions were spun for five minutes at 2,000 × g and the supernatant was aspirated. The cell pellets were lysed by vortexing in 200 μL of M-PER mammalian protein extraction buffer (Pierce, cat. # 78501) containing protease and phosphatase inhibitors (Sigma, cat. #'s P1860-1ML, P0044, and P5726). The samples were then spun in a microcentrifuge for five minutes at 12,000 × g and the supernatants were collected. Protein concentrations were determined using a nanodrop spectrophotometer (Thermo Life Sciences) and 50 μg of total protein was loaded and run on a 4 to 12% polyacrylamide gel (Invitrogen). The gels were blotted onto nitrocellulose using the iblot transfer system (Invitrogen). The blots were blocked for one hour at room temperature in 1 × TBST (Affymetrix, Santa Clara, CA, USA, cat # 77500 5 LT) containing 5% non-fat milk. The blots were then washed in 1 × TBST and were incubated overnight at 4°C in 10 mL of primary antibody at a 1:500 dilution in 5% BSA/TBST (Sigma cat # A7906-1 KG). Blots were then washed in 1 × TBST and incubated with infrared-labeled secondary antibodies (LiCor) for 30 minutes at room temperature. The blots were then washed in 1 × TBST and scanned using the Odyssey infrared imaging system (LiCor, Lincoln, NE USA). Bands were quantified using the Odyssey software (LiCor) and normalized to bands corresponding to the housekeeping Rho-GDI protein. Four independent samples were prepared for each cell line. Paired t test analyses were performed for each protein using Origin 8.5.1 software (control vs TamR), and P-values < 0.05 were considered significant.
Transwell migration assay
Migration assays were performed following the manufacturer's instructions (BD Falcon, Sparks, MD, USA). Briefly, MCF-7-control or MCF-7-TamR cells were seeded at a density of 2.5 × 104 in 500 μL serum-free and phenol red-free media in the upper chamber of a 24-well transwell system. Phenol red-free DMEM supplemented with FBS (5%) was used as a chemoattractant in the lower wells. After 24 h, membranes were scrubbed, fixed with 10% phospho-buffered formalin, permeabilized with 100% ice-cold methanol, and stained with 0.1% crystal violet in 20% methanol. Membranes were removed and mounted on glass slides for visualization by light microscopy. Data are represented as a percent of the migrated MCF-TamR cells per 100 × field of view (100×) ± SEM for triplicate experiments.
MCF-7 cells overexpressing S100P
Construction of S100P lentiviral vector
The S100P gene was generated by elongating RT-PCR using a Superscript III one-step RT-PCR system (Invitrogen) with the following primers: S100P-F (sense) 5'-CGC CAC CAT GAC GGA ACT AGA GAC AGC C-3' and S100P-R (antisense) 5'-GGA TCC TCA TTT GAG TCC TGC CTT CTC-3'. The RT-PCR reaction was carried out as follows: step 1: 45°C for 30 minutes and 94°C for 2 minutes; step 2: 35 cycles at 94°C for 15 sec, 51°C for 30 sec and 72°C for 1 minute; step 3: 72°C for 5 minutes and held at 4°C. The PCR product was cloned using a TA Cloning kit (Invitrogen). The S100P lentiviral vector (pLenti6/S100P) was constructed by digesting vector pLenti6 (Invitrogen) with EcoR I and BamH I for insertion of the S100P gene.
MCF-7-S100P cell line stably overexpressing S100P
To produce S100P-overexpressing lentivirus, the 293FT cells were co-transfected with expression construct (pLenti6/S100P) and the optimized packaging mix (ViraPower Packaging mix, Invitrogen) from a lentiviral expression system (Invitrogen). The transfection was carried out by incubating cells overnight at 37°C in a CO2 incubator using a Lipofectamine 2000 reagent (Invitrogen). Media were replaced in 24 hours and the virus-containing supernatants were harvested and centrifuged at 48 to 72 hours. MCF-7 cells were grown to 30 to 50% confluent, and the culture medium was replaced with viral supernatants as obtained previously. Polybrene was added for the overnight viral transfection. Subsequently, medium was replaced every 2 to 3 days with antibiotic (Blasticidin) and the selection process continued for a total of 10 to 12 days. The stable MCF-7-S100P cell line was cultured in phenol red free DMEM medium with 5% FBS, and the S100P expression was checked with Western blot.
Bioinformatics were performed on significantly altered proteins. This was determined by two parameters: one is having an analytical replication
P-value of < 0.05 and the second is determined by the ratio value. The standard deviation (SD) of all the ratios in the control sample was determined and then significance was defined as (1 +/- 2SD) [
22‐
24]. Classification of proteins was determined by the web program PANTHER [
25]. The proteins were analyzed for over expression of gene ontology terms in the categories of pathways, molecular function and biological process. Pathway mapping was done using Pathvisio 2.0.11, a tool for visualizing and editing biological pathways [
26]. The ratio data of the significant proteins were loaded into Pathvisio and used to map onto preloaded pathways from Wikipathways [
27] and KEGG [
28‐
30]. The pathway thus created was heavily modified from KEGG pathway 04810, "Regulation of actin cytoskeleton" in
Homo sapiens.
Patient survival analysis
An online database [
31] was used to assess relevance of significantly changed protein expressions to relapse-free survival. The database was established using gene expression data and survival information on 1,809 patients downloaded from Gene Expression Omnibus (GEO) (Affymetrix HGU133A and HGU133+2 microarrays, Santa Clara, CA, USA). Briefly, single or multiple genes were entered into the database to obtain Kaplan-Meier survival plot where the number-at-risk was indicated below the main plot. Hazard ratio (and 95% confidence intervals) and logrank
P were calculated and displayed on the webpage. For the genes listed in Tables
1 and
2, their effects on relapse-free survival (RFS) were calculated and listed. Positive logrank
P-values indicate positive correlation (that is, either overexpression or down-regulation of a gene correlates with decreased survival) and negative logrank
P-values indicate negative correlation (that is, either up- or down-regulation of a gene is associated with increased survival).
Table 1
Selected up-regulated proteins in tamoxifen resistant breast cancer cells
IPI00025311 | 584 | 61.7 | Breast carcinoma-amplified sequence 1 | bcas1 | 11.28 | 1.3E-06 | -3.2E-5 |
IPI00017526 | 95 | 10.4 | Protein S100-P |
s100p
| 5.20 | 2.3E-08 | 1.7E-6 |
IPI00218831 | 218 | 25.7 | Glutathione S-transferase Mu 1 | gstm1 | 3.70 | 8.6E-07 | -3.8E-11 |
IPI00183695 | 97 | 11.2 | Protein S100-A10 |
s100a10
| 3.15 | 1.0E-07 | 2.3E-5 |
IPI00922108 | 1002 | 111.1 | Integrin alpha-V | itgav | 2.88 | 7.8E-06 | 2.0E-9 |
IPI00021267 | 976 | 108.2 | Ephrin type-A receptor 2 |
epha2
| 2.81 | 3.1E-05 | -2.0E-9 |
IPI00027341 | 348 | 38.5 | Macrophage-capping protein | capg | 2.80 | 2.3E-07 | 3.9E-5 |
IPI00106687 | 222 | 25.7 | Latexin | lxn | 2.69 | 1.3E-05 | 0.2 |
IPI00013895 | 105 | 11.7 | Protein S100-A11 |
s100a11
| 2.51 | 3.0E-05 | 5.1E-10 |
IPI00455315 | 339 | 38.6 | Annexin A2 |
anxa2
| 2.41 | 2.7E-08 | 0.93 |
IPI00297910 | 323 | 35.7 | Tumor-associated calcium signal transducer 2 |
tacstd2
| 2.19 | 1.1E-06 | 0.53 |
IPI00903145 | 583 | 68.5 | Radixin |
rdx
| 2.17 | 2.4E-06 | 0.25 |
IPI00219301 | 332 | 31.5 | Myristoylated alanine-rich C-kinase substrate |
marcks
| 2.14 | 6.9E-06 | -0.17 |
IPI00853146 | 167 | 19.2 | Caveolin | cav | 2.08 | 1.0E-03 | -4.2E-8 |
IPI00215995 | 1051 | 116.5 | Integrin alpha-3 | itga3 | 2.00 | 2.0E-04 | -3.9E-9 |
IPI00010414 | 329 | 36.0 | PDZ and LIM domain protein 1 | pdlim1 | 1.94 | 1.8E-06 | -0.009 |
IPI00795633 | 448 | 52.3 | Clusterin | clu | 1.88 | 1.5E-04 | -0.02 |
IPI00010214 | 104 | 11.7 | Protein S100-A14 | s100a14 | 1.87 | 2.1E-04 | 0.96 |
IPI00465431 | 250 | 26.1 | Galectin-3 |
lgals3
| 1.79 | 2.5E-06 | 0.28 |
IPI00219219 | 135 | 14.7 | Galectin-1 |
lgals1
| 1.78 | 2.0E-04 | 0.008 |
IPI00018364 | 183 | 20.5 | Ras-related protein Rap-2b | rap2b | 1.71 | 7.5E-05 | 6.7E-6 |
IPI00016485 | 394 | 42.7 | Protein phosphatase slingshot homolog 3 | ssh3 | 1.68 | 1.1E-03 | -2.4E-11 |
IPI00217563 | 798 | 88.4 | Integrin beta-1 | itgb1 | 1.64 | 3.8E-08 | 6.9E-12 |
IPI00180240 | 44 | 5.1 | Thymosin beta-4-like protein 3 | tmsl3 | 1.63 | 3.7E-07 | 1.7E-5 |
IPI00641181 | 195 | 19.5 | MARCKS-related protein | marcksl1 | 1.55 | 3.6E-06 | 0.011 |
IPI00759759 | 327 | 37.1 | Epidermal growth factor receptor kinase substrate 8-like protein 2 | eps8l2 | 1.53 | 2.4E-03 | -0.13 |
IPI00021828 | 98 | 11.1 | Cystatin-B |
cstb
| 1.52 | 4.2E-04 | 4.7E-5 |
IPI00001871 | 340 | 36.5 | PRKC apoptosis WT1 regulator protein | pawr | 1.47 | 1.9E-04 | 0.23 |
IPI00843975 | 586 | 69.4 | Ezrin |
ezr
| 1.41 | 4.0E-04 | 1.8E-5 |
IPI00847442 | 92 | 10.1 | FK506 binding protein12 | fkbp12 | 1.40 | 1.4E-05 | 0.0011 |
IPI00022462 | 760 | 84.8 | Transferrin receptor protein 1 | tfrc | 1.40 | 1.2E-05 | 0.00049 |
IPI00005585 | 124 | 13.7 | Tax1-binding protein 3 | tax1bp3 | 1.39 | 5.5E-06 | 6.7E-6 |
IPI00016179 | 98 | 11.5 | Protein S100-A13 |
s100a13
| 1.39 | 7.6E-09 | 0.15 |
IPI00005202 | 247 | 26.2 | Membrane-associated progesterone receptor component 2 | pgrmc2 | 1.38 | 2.2E-04 | 3.6E-5 |
IPI00015148 | 184 | 20.8 | Ras-related protein Rap-1b | rap1b | 1.36 | 5.7E-06 | 2.3E-5 |
IPI00291175 | 1066 | 116.6 | Vinculin | vcl | 1.35 | 2.3E-05 | -0.11 |
IPI00375426 | 323 | 36.2 | Cathepsin H | ctsh | 1.34 | 3.6E-04 | -0.44 |
IPI00011285 | 714 | 81.8 | Calpain-1 catalytic subunit | capn1 | 1.33 | 6.1E-06 | -0.22 |
IPI00020599 | 417 | 48.1 | Calreticulin | calr | 1.31 | 3.3E-06 | 0.67 |
IPI00009342 | 1657 | 189.1 | Ras GTPase-activating-like protein IQGAP1 | iqgap1 | 1.29 | 7.9E-07 | 1.0E-4 |
IPI00010080 | 527 | 58.0 | Serine/threonine-protein kinase OSR1 | oxsr1 | 1.28 | 5.0E-04 | 0.49 |
IPI00478231 | 193 | 21.8 | Transforming protein RhoA |
rhoa
| 1.28 | 4.8E-07 | 0.016 |
IPI00012011 | 166 | 18.5 | Cofilin-1 |
cfl1
| 1.28 | 1.5E-05 | -7.9E-8 |
IPI00025084 | 268 | 28.3 | Calpain small subunit 1 | capn2 | 1.26 | 1.4E-05 | -1.6E-12 |
IPI00013808 | 911 | 104.8 | Alpha-actinin-4 | actn4 | 1.24 | 1.5E-04 | 0.64 |
IPI00000513 | 821 | 90.9 | E-cadherin | cdh1 | 1.24 | 6.6E-05 | 1.5E-7 |
IPI00220739 | 195 | 21.7 | Membrane-associated progesterone receptor component 1 | pgrmc1 | 1.23 | 1.1E-03 | 5.7E-7 |
IPI00028091 | 418 | 47.3 | Actin-related protein 3 | actr3 | 1.23 | 4.8E-04 | 2.4E-14 |
IPI00217519 | 206 | 23.6 | Ras-related protein Ral-A | rala | 1.21 | 3.6E-03 | < 1.0E-16 |
IPI00922213 | 1014 | 111.2 | fibronectin 1 | fn1 | 1.19 | 3.4E-02 | 0.46 |
IPI00220847 | 1745 | 194.3 | Integrin beta-4 | itgb4 | 1.19 | 6.3E-04 | -0.12 |
IPI00016786 | 191 | 21.2 | Cell division control protein 42 homolog | cdc42 | 1.15 | 5.2E-05 | -4.3E-10 |
Table 2
Selected down-regulated proteins in tamoxifen resistant breast cancer cells
IPI00218414 | 260 | 29.2 | Carbonic anhydrase 2 |
ca2
| -2.70 | 1.3E-05 | 0.79 |
IPI00032808 | 219 | 24.3 | Ras-related protein Rab-3D | rab3d | -2.36 | 4.5E-06 | 1.2E-4 |
IPI00472076 | 175 | 19.4 | tumor protein D53 | tpd52l1 | -2.03 | 7.5E-06 | 0.66 |
IPI00012866 | 480 | 55.7 | RAC-alpha serine/threonine-protein kinase | akt1 | -1.95 | 3.4E-06 | -0.028 |
IPI00647268 | 189 | 21.6 | Ras homolog gene family, member C | rhoc | -1.86 | 6.4E-06 | -0.019 |
IPI00025318 | 114 | 12.8 | SH3 domain-binding glutamic acid-rich-like protein | sh3bgrl | -1.75 | 2.4E-05 | 3.8E-5 |
IPI00550020 | 102 | 11.5 | Parathymosin |
ptms
| -1.74 | 5.2E-06 | 0.095 |
IPI00186008 | 291 | 33.0 | PCTP-like protein |
stard10
| -1.74 | 9.7E-04 | NA |
IPI00011564 | 198 | 21.6 | Syndecan-4 | sdc4 | -1.68 | 2.1E-04 | NA |
IPI00019502 | 1960 | 226.4 | Myosin-9 | myh9 | -1.63 | 5.8E-06 | -0.085 |
IPI00011696 | 845 | 98.3 | Proto-oncogene vav | vav1 | -1.58 | 4.8E-02 | 6.2E-11 |
IPI00022283 | 84 | 9.1 | Trefoil factor 1 | tff1 | -1.54 | 1.5E-04 | 1.5E-5 |
IPI00550900 | 172 | 19.6 | Translationally-controlled tumor protein | tpt1 | -1.51 | 1.4E-06 | 1.1E-10 |
IPI00019345 | 184 | 21.0 | Ras-related protein Rap-1A | rap1a | -1.50 | 1.2E-05 | -9.5E-8 |
IPI00479997 | 149 | 17.3 | Stathmin | stmn1 | -1.47 | 1.5E-05 | -2.9E-12 |
IPI00011229 | 412 | 44.5 | Cathepsin D |
ctsd
| -1.45 | 5.9E-07 | 0.56 |
IPI00216319 | 246 | 28.2 | 14-3-3 protein eta |
ywhah
| -1.45 | 5.5E-04 | -0.066 |
IPI00217975 | 586 | 66.4 | Lamin-B1 | lmnb1 | -1.43 | 2.3E-05 | 0.32 |
IPI00414676 | 724 | 83.2 | Heat shock protein HSP 90-beta | hsp90ab1 | -1.40 | 5.9E-07 | 5.6E-7 |
IPI00246975 | 225 | 26.5 | Glutathione S-transferase Mu 3 | gstm3 | -1.37 | 8.9E-07 | 0.008 |
IPI00009607 | 183 | 20.7 | Ras-related protein Rap-2c | rap2c | -1.36 | 2.4E-03 | -3.1E-6 |
IPI00020904 | 358 | 40.9 | Serine/threonine-protein kinase PRKX | prkx | -1.33 | 6.5E-03 | 0.7 |
IPI00000041 | 196 | 22.1 | Rho-related GTP-binding protein RhoB | rhob | -1.33 | 2.0E-03 | 0.0032 |
IPI00003815 | 204 | 23.2 | Rho GDP-dissociation inhibitor 1 | arhgdia | -1.27 | 1.5E-03 | 0.0042 |
IPI00955014 | 297 | 34.1 | cell division control protein 2 homolog | cdk1 | -1.26 | 6.1E-04 | NA |
IPI00419235 | 229 | 26.9 | Glutathione S-transferase Mu 5 | gstm5 | -1.25 | 1.3E-04 | < 1E-16 |
IPI00941907 | 350 | 38.4 | Serine-threonine kinase receptor-associated protein | strap | -1.25 | 1.7E-05 | -5.4E-9 |
IPI00297261 | 435 | 49.9 | Tyrosine-protein phosphatase non-receptor type 1 | ptpn1 | -1.25 | 4.3E-04 | 1.1E-11 |
IPI00019812 | 499 | 56.8 | Serine/threonine-protein phosphatase 5 | ppp5c | -1.15 | 4.4E-04 | 7.4E-12 |
Discussion
We have established a tamoxifen-resistant breast cancer cell line obtained under an FBS-containing medium condition to minimize adaptive cellular changes in response to LTED. Indeed, earlier studies have shown that LTED leads to enhanced expression of the estrogen receptor [
42] or EGFR [
14], which are not usually observed in tamoxifen resistant cell lines cultured in normal FBS medium [
13,
15]. In the MCF-7-TamR cell line obtained in this study after six months of 4-OH tamoxifen treatment, the estrogen receptor was significantly down-regulated but retained viable function (Figure
5). Current understanding of endocrine resistance depicts a progressive, stepwise process in response to anti-estrogen challenge where breast cancer cells evolve from an estrogen-dependent phenotype to a non-responsive one and eventually to a stage of estrogen independence. Our results indicate the tamoxifen resistant cells appear to be at a stage of minimized estrogen responsiveness without complete loss of ER. Previous studies of tamoxifen resistance using
in vitro models suggest translocation of ER from nucleus to membrane, facilitating crosstalk with growth factor receptors and enhancing the non-genomic signaling of the ER. In these reports, the total ER levels remain largely unchanged [
13,
15,
42]. On the other hand, complete loss of ER expression has occurred when MCF-7 cells became resistant to the pure antiestrogen, fulvestrant [
43‐
45].
This
in vitro behavior is also consistent with clinical observations that tamoxifen resistant tumors may still respond to fulvestrant [
46,
47] and that only 15 to 30% of patients present with complete loss of ER at time of relapse [
11,
48,
49]. The down-regulation of ER mediated signaling pathways in our MCF-7-TamR cells is corroborated by proteomic evidence that showed suppressed expression levels of cathepsin D and TFF1/PS2 and was confirmed by Western blot analysis showing diminished ER protein expression. PgR, an ER dependent gene, was also found significantly down-regulated (> 1,000-fold, Figure
5A) by RT-PCR analysis. On the other hand, E
2 stimulation did induce a 50-fold increase in PgR expression from its greatly suppressed basal level (Figure
5B) in the resistant cells.
In ER positive breast cancer cells, estrogen signaling is the main mediator of proliferation and tumor progression. Adaptation to tamoxifen challenge which blocks ER signaling must involve activation of alternative survival signaling to sustain growth and circumvent the apoptotic effect of tamoxifen. As demonstrated in numerous
in vitro and
in vivo studies on the mechanisms of tamoxifen resistance, tumor cells recruit a remarkably wide variety of signaling pathways to achieve the resistant outcome [
50,
51], including cross talk with EGFR and Her2 [
52,
53], and enhanced nongenomic signaling accompanied by translocation of ER [
54,
55]. Our study identified several proteins that are known to promote tumorigenesis and progression but their roles in tamoxifen resistance have not been explored. In particular, the up-regulation of S100P revealed a previously unknown link between tamoxifen resistance and the small calcium binding protein. S100P is a ligand for the receptor for advanced glycation end product (RAGE). Binding of the Ca
2+ activated S100P homodimer to RAGE has been shown to promote cancer cell proliferation via the ERK1/2 and NFκB signaling pathways [
56‐
58]. S100P was found to co-immunoprecipitate with RAGE and its action on cell survival and proliferation could be blocked by RAGE inhibitors [
56].
The forced overexpression of S100P in the tamoxifen sensitive MCF-7 cell line increased its resistance to tamoxifen significantly (Figure
8B), confirming the role of S100P in acquired tamoxifen resistance. Our results suggest that, as the ER-regulated proliferation pathway was severely suppressed after prolonged exposure to tamoxifen, the S100P-RAGE signaling via activation of ERK1/2 and possibly NF-κB is increased as a compensatory mechanism of cell proliferation and survival.
In addition, the up-regulation of the anti-apoptotic protein CLU can be viewed as another possible survival pathway contributing to tamoxifen resistance. Previous reports have implicated CLU up-regulation as a general defense mechanism of cancer cells toward cytostatic drugs [
59‐
61]. Under cell stress, such as treatment with trastuzamab in breast cancer cells, or following androgen ablation in prostate cancer cells, significant increase in CLU expression was associated with activation of alternative signaling [
62,
63].
Another significantly up-regulated protein, EphA2, may contribute to the survival of tamoxifen resistant cells. The EphA2 expression level in breast cancer cells has been found inversely related to ER expression [
64,
65]. This is consistent with our RT-PCR and Western blot results where ER was significantly down-regulated (Figures
4 and
5). EphA2-transfected cells demonstrated increased growth
in vitro and form larger and more aggressive tumors
in vivo [
66]. Moreover, EphA2 overexpression decreased the ability of tamoxifen to inhibit breast cancer cell growth and tumorigenesis [
67,
68]. The finding in this study that EphA2 was overexpressed in a tamoxifen resistant cell line confirms the involvement of the receptor tyrosine kinase in the development of tamoxifen resistance in breast cancer.
As the cells adapt to the inhibitory effects of tamoxifen, the acquired resistance appears to transform the breast cancer cells into a more aggressive phenotype with increased motility. Indeed, many of the overexpressed proteins thought to regulate growth and proliferation in our TamR cells have also been implicated in promoting cancer cell migration and invasion. Gene Ontology and KEGG pathway analyses collectively using proteomic data suggest that regulation of actin cytoskeleton may be responsible for driving the motility of TamR cells. The novel role of S100P in the regulation of cytoskeleton dynamics was highlighted in the pathway map (Figure
6) in which S100P was involved in the interactions with ezrin [
69], a membrane/F-actin cross-linking protein implicated in tumor metastasis [
70‐
73], and with the scaffolding protein IQGAP1 [
74], known to promote cell motility and invasion [
75]. To confirm the involvement of S100P in regulation of tamoxifen induced cell motility, we conducted functional studies of S100P by overexpressing the protein in the parental MCF-7 cells and observed increased motility in MCF-7-S100P cells as a result (Figure
8C). Moreover, our proteomic finding that both ezrin and IQGAP1 were up-regulated in the tamoxifen resistant cells (1.43- and 1.29-fold, respectively) provided additional evidence for the involvement of S100P in motility enhancement and suggests that the mechanism of action may involve the ezrin and IQGAP1 pathways.
Finally, overexpression of S100P and its role in mediating tamoxifen resistance and cell motility also bear clinical relevance. Using a GEO gene expression database from 1,809 breast cancer patients, the Kaplan-Meier survival plots demonstrate the prognostic relevance of S100P overexpression on patient survival. Overexpression of S100P is predictive of lower relapse free survival (P = 1.76e-6) and significantly correlated with decreased distant metastasis free survival (P = 0.029). Furthermore, truly prognostic patient group, that is, systematically untreated breast cancer patients with higher levels of S100P tend to have shorter relapse free period (P = 0.017). Finally, S100P up-regulation appears to be significantly associated with reduced survival in ER(+) but not in ER(-) breast cancer patients.
Conclusion
Using a quantitative proteomic approach we have identified and verified key adaptive protein changes that are involved in the development of tamoxifen resistance. Long term treatment with 4-hydroxytamoxifen significantly suppressed ER-regulated signaling pathways in MCF-7 breast cancer cells. This was demonstrated in the marked down-regulation of ER dependent genes, including PgR, PS2, and SDF-1. In response, alternative survival signaling was activated that appeared to involve the up-regulation of multiple proteins. This was reflected in the global proteomic changes that included the increased expression of TROP2, CLU, MARCKS, and S100 family proteins. In particular, we identified S100P, an EF-hand calcium binding protein previously implicated in breast and other solid tumors, as a significant player in conferring tamoxifen resistance and cell motility. Overexpression of S100P in the hormone sensitive parental MCF-7 cells significantly increased resistance to tamoxifen. The mechanism of S100P action may involve its interaction with the receptor RAGE, leading to sustained survival and proliferation.
Proteomic analysis of MCF-7-TamR cells also revealed a critical phenotypic transformation of the cells towards an increased migratory capacity, consistent with most clinical outcomes where tumor invasion and metastasis follow the acquired hormone resistance in patients. The enhanced cell motility in the tamoxifen resistant cells appeared to be driven by the cytoskeletal dynamics where S100P played an important role. This was supported by the observation that overexpressing S100P in MCF-7 cells significantly increased cell migration. Additional evidence comes from proteomic data where up-regulation of multiple proteins in a coordinated signaling network may regulate the actin cytoskeleton dynamics as depicted in our proposed pathway model. Specifically, we observed the up-regulation of EphA2, RhoA, ITGB1, vinculin, ezrin, and radixin, which are key proteins contributing to the increased cell motility in a tamoxifen resistant phenotype by promoting actin fiber polymerization, filopodia formation, and cell contractability.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
CZ cultured cell lines, performed survival assays and proteomic sample preparation, RT-PCR, interpreted data and drafted the manuscript. QZ performed lentiviral transduction and subsequent functional studies of cell survival and migration, and contributed to drafting the revised manuscript. LR performed migration assays. IT performed bioinformatics analysis, while MRB performed Western blotting. QZ carried out HPLC-MS/MS based protein identification and database search. EM performed RT-PCR and SE performed colony assays. BMC and MEB participated in experimental design and interpretation, and critically revised the manuscript. GW designed the study, and drafted and critically revised the manuscript. All authors have read and approved the final manuscript.