Phospho-kinase profile of triple negative breast cancer and androgen receptor signaling

Cell culture media and supplements (fetal bovine serum (FBS), glutamine, penicillin/streptomycin) were purchased from Invitrogen (Gaithersburg, MD). The anti-pErk1/2, anti-Erk2, anti-PDGFR β, anti-pPDGFR β, anti-EGFR and anti-pEGFR antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-Akt and anti-AR antibodies were from Cell Signalling Technologies (Beverly, MA). The anti-phosphorylated Akt (Serine 473) antibody was generated against the sequence RPHFPQFpS473YSAS (p = phosphorylated) [20].

All cell lines were cultured at 37°C in a humidified atmosphere in the presence of 5% CO₂–95% air. Cells were grown in DMEM or in RPMI medium containing a high glucose concentration (4,500 mg/liter) supplemented with antibiotics (penicillin at 100 U/ml, streptomycin at 100 μg/ml), glutamine 2 mM and 10% FBS. Cells were treated with different drugs: PD98059 (50 μM), BEZ235 (0.5 μM), BIC (20 μM), Imatinib (10 μM), Lapatinib (10 μM) and DHT (50nM).

Human samples were obtained from the tumor bank of the Salamanca University Hospital and Albacete University Hospital following institutional and ethical guidelines. All patients signed the study consent form. TNBC samples were defined as those with HER2 negative by inmunohistochemistry (IHC), HER2 of 0 or 1+ or a negative fluorescence in situ hybridization (FISH). HER2 amplification (evaluated by FISH DAKO HER2 FISH pharmDxTM Kit, DakoCytomation, Glostrup, Denmark A/S) was defined as a HER2-chromosome 17 ratio of < .2.0, as required by guidelines. Hormone receptor (HR) negative was defined as follow <10% positive cells by IHC for both estrogen receptor (ER) and progesterone receptor (PR) following recommendations before 2011.

Frozen human samples were inspected by haematoxylin-eosin staining for epithelial tumor content by analysis of two slices at each end of the tumor. Only samples containing 70% epithelial tumoral cells were selected for Western analyses. The tumors were minced, washed with phosphate-buffered saline (PBS), and homogenized in ice-cold lysis buffer (140 mM NaCl; 10 mM EDTA; 10% glycerol; 2% Triton X-100; 20 mM Tris pH 7.0; pepstatin, 10 mM; aprotinin, 10 mg/ml; leupeptin, 10 mg/ml; PMSF, 1 mM; beta-glycerophosphate, 25 mM; sodium fluoride, 10 mM; and sodium orthovanadate, 10 mM) with a tight-fitting Dounce homogenizer. This homogenate was centrifuged at 10,000 g for 20 minutes at 4 uC, and the supernatants were transferred to new tubes.

Two commercial arrays were used for the studies; the human phospho-RTK array kit (R&D Systems, Abingdon, United Kingdom) and the PathScan RTK Signaling Antibody Array Kit (Cell Signaling). The Image J 1.44 software (National Institute of Health, Bethesda, MD, USA) was used for the quantification of the different RTKs in the human phospho-RTK array kit, and the Odyssey V3.0 program for the quantification of the cell signalling intermediates.

Cells were washed with phosphate-buffered saline and lysed in ice-cold lysis buffer (140 mM NaCl, 10 mM ethylenediaminetetraacetic acid, 10% glycerol, 1% Nonidet P-40, 20 mM Tris, pH 7.0, 1 μM pepstatin, 1 μg/ml aprotinin, 1 μg/ml leupeptin, 1 mM phenylmethyl sulphonyl fluoride (PMSF), and 1 mM sodium orthovanadate) [21]. Lysates were centrifuged at 10,000 × g at 4°C for 10 min, and supernatants were transferred to new tubes. Samples were then boiled in electrophoresis sample buffer and placed on 6%–15% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) gels, depending on the molecular weight of the proteins to be analyzed. After electrophoresis, proteins in gels were transferred to polyvinylidene difluoride (PVDF) membranes (Millipore Corporation). Membranes were blocked in Tris-buffered saline with Tween 20 (TBST) (100 mM Tris [pH 7.5], 150 mM NaCl, 0.05% Tween 20) containing 1% of bovine serum albumin for 1 hour and then incubated with the corresponding antibody for 2–16 hours. After washing with TBST, membranes were incubated with HRP-conjugated anti-mouse or anti-rabbit secondary antibodies for 30 minutes and bands were visualized by using ECL Western Blotting Detection System (GE Healthcare, Buckinghamshire, United Kingdom).

Cells were plated in 24-well plates at 15,000–20,000 cells/well and cultured overnight in DMEM or RPMI + 10% FBS. The next day medium was replaced with DMEM or RPMI containing the different drugs. Cell proliferation was analyzed days later by an MTT-based assay as described [22]. Unless otherwise indicated, the results are presented as the mean ± standard deviation (SD) of quadruplicates of a representative experiment that was repeated at least three times.

Total RNA was extracted with RNeasy Mini Kit (Quiagen) as recommended by the supplier. cDNAs were synthesized from 5 μg of total RNA by using RevertAid H Minus First Stand cDNA Synthesis Kit (Fermentas) in a total volume of 20 μl. Reverse transcription was performed at 65° for 5 min followed by 60 min at 42°. PCR reactions were performed with 2 μl cDNA in a 50 μl volume and using Taq Polymerase (Biotools) and the following primer pairs (fragments size indicated in brackets): AR (247 bp) forward 5′-CTCACCAAGCTCCTGGACTC-3′ and reverse 5′-CAGGCAGAAGACATCTGAAAG -3′, β-actin (661 bp), forward 5′-TGACGGGGTCACCCACACTGTGCCCATCTA-3′ and reverse 5′-CTAGAAGCATTTGCGGTGGACGATGGAGGG-3′. PCR cycle conditions were 1 min at 95°C, 1 min at 55°C, and 1 min at 72°C for AR and β-actin. The number of cycles was 35 and the products were resolved on a 1% agarose gel and visualized with ethidium bromide.

The design of this study is approved by the Committee of Ethics in Research of the University hospital of Albacete (Reference PI-2010/017).

We used a panel of eight TNBC cell lines to evaluate the expression of the AR. Five out of eight cell lines expressed the AR, except for HBL100, HCC3153, and MDA-MB231. To corroborate these findings at a genetic level we extracted gene expression data from public libraries and evaluated the expression of the AR gene in a panel of cell lines. BT549 and HS578T (claudin-low subtype) had the highest level, HCC3153 and HCC70 (basal-like subtype) the lowest, and MDA-MB231 (claudin-low subtype) and MDA-MB435 (claudin-low subtype), had low to moderate levels (Figure 1A).

We next evaluated the expression of the AR in 16 human samples from triple negative breast cancer patients. As can be seen in Figure 1B the expression of the AR was observed in 5 out of 16 tumors (samples number: 2428, 37, 795, 278, 1603). Of note in three of them (samples number: 2428, 278 and 1603) the expression of the AR was high.

Given the fact that one of the mechanisms that control the expression of the AR in prostate cancer is linked with kinase signaling pathways, we evaluated if the activation of membrane RTKs or their downstream pathways could be associated with the expression of the AR. For this purpose, we analyzed the expression of 42 RTKs and downstream effectors using phospho-specific antibody arrays in the 16 human samples from TNBC patients. Two examples are shown in Figure 2A and all date available in Additional file 1: Figure S1.

Two RTKs, the PDGFRβ and EGFR, were phosphorylated in an important number of patients (more than 20%), and their activation status correlated at a significant level with the expression of the AR measured by western-blot (Figure 2A-B) (Mann–Whitney test: EGFR and AR: p = 0.03; PDGFRβ and AR: p = 0.004). We did not identify any association between the activation status of other RTKs and the expression of the AR (data not shown). Using a similar approach we studied the kinase activation profile of proteins involved in the intracellular signaling mediated by RTKs (Figure 2A, right arm). Of note pAKT (pAKTS473, pAKTT308) and pErk1/2 were expressed in more than 40% of the samples. Indeed, pAKTS473 was expressed in around 90% of the tumors. Additional file 1: Figure S1 describes the composition of both arrays and the activation profile of these kinases. No differences were observed between any of the intracellular signaling proteins- AKT or Erk1/2- and the expression of the AR (Additional file 2: Figure S2, upper panel). This lack of association could be explained by the limited sample size used and the wide expression of these activated kinases (pAKTS473 is expressed in 90% of samples and pErk1/2 in more than 40%). However, some samples with a strong activation of pAKTS473 also showed higher expression of the AR including samples BT278, BT1603 and BT2428 (Additional file 2: Figure S2, middle and lower panel).

We then studied the expression of RTKs and downstream pathways in the panel of TNBC cell lines. EGFR was activated in HCC1187, BT 549, HCC 70 and HCC 3153; and PDGFR was activated in HS578 and HCC 1187. All these cell lines described, except HCC3153, expressed the AR (Figure 2C).

In addition we studied downstream activated pathways including AKT-mTOR and Erk1/2. All cell lines had activation of pAKT except MDA-MB231. pS6 and pErk1/2 were expressed in all cell lines at different amount levels (Figure 2C). The high presence of phosphorylated AKT and Erk1/2 matched the results observed using human samples, suggesting that these cell lines could be used as a representative in vitro model. However, the increased existence of activated AKT and Erk1/2 in most of these cell lines made difficult to identify any association between the expression of the AR and the activation of these pathways.

Given the association observed between some RTKs and downstream pathways with the expression of the AR in human samples and cell lines, we evaluated if the pharmacological inhibition of these receptors could modify the expression of the AR.

For this purpose we used two cell lines; BT549, with constitutive activation of EGFR; and HS578T, with activation of PDGFRβ. Both cell lines have activation of AKT, S6 and Erk1/2, being HS578T a cell line with more activation of Erk1/2. Treatment with imatinib mesylate, a PDGFβ inhibitor, do not decreased the amount of the AR in HS578T; and a similar effect was observed for lapatinib, an EGFR inhibitor, in BT549 (Figure 3A).

As EGFR and PDGFR signal through downstream pathways, mainly the PI3k-mTOR and the Erk1/2 pathway, and these routes have also been implicated in the androgen-independent control of the AR in prostate cancer, we evaluated if the inhibition of these central nodes could have more effect on the expression of the AR than individual inhibition of RTKs. Using the same two models, we observed that the administration of PD98059, a MEK inhibitor that inhibits Erk1/2, did not reduce the amount of the AR (Figure 3A). By contrast the PI3K-mTOR inhibitor BEZ235 reduced substantially the amount of the AR in both cell lines (Figure 3A).

We next explored the action of the anti-androgen bicalutamide when combined with inhibitors of EGFR, PDGFRβ and the PI3K-mTOR and Erk1/2 pathways. Interestingly we observed that the concomitant administration of bicalutamide with EGFR, PDGFRβ and MEK inhibitors reduced the amount of the AR compared to each agent alone (Figure 3A). This finding was not observed when combining a PI3K-mTOR inhibitor with bicalutamide.

We decided next to evaluate the effect on proliferation of pharmacological inhibitors of these receptors and pathways alone or in combination with antiandrogens. Bicalutamide alone reduced the MTT uptake in both cell lines (Figure 3B). Interestingly, in both cell lines the combination of bicalutamide with MEK inhibition was able to produce a greater anti-proliferative effect than each agent given alone (p < 0.05). Interestingly, this combination effect was not observed for BEZ235. In HS578T, imatinib was unable to reduce the MTT uptake but increased the anti-proliferative effect of bicalutamide; and in BT549 lapatinib augmented the action of bicalutamide in a similar manner (Figure 3B). A similar effect was observed counting cells (Figure 3C).

We then evaluated the action of DHT and bicalutamide on the expression of AR and how the inhibition of these kinases can modulate its expression. Interestingly administration of DHT increased the amount of the AR in HS578T and BT549 (Figure 4A). Administration of bicalutamide did not reduce the expression of the AR when given alone or in combination with DHT in BT549. However, in HS578T bicalutamide partially prevented the amount increase of the AR in the presence of DHT (Figure 4A). The amount of the AR gene measured by polymerase chain reaction (PCR), were not modified by the administration of DHT suggesting that this is a posttranscriptional mechanism (Figure 4B). When we evaluated the action of Erk1/2 and PI3K inhibitors on the expression of AR after stimulation with DHT, we observed that the concomitant administration of BEZ235 was not able to reduce AR expression as observed previously when given alone (Figure 3). In the presence of DHT, only bicalutamide is able to reduce the expression of AR produced by DHT (Figure 4A). Similarly, BEZ235 or PD98059 given alone; or BEZ235 and PD98059 in combination with bicalutamide did not modify the amount of the AR gene (Figure 4B).