Ret inhibition decreases growth and metastatic potential of estrogen receptor positive breast cancer cells

We show that elevated levels of Ret receptor are found in different sub-types of human breast cancers and that high Ret correlates with decreased metastasis-free survival. The role of Ret in ER+ breast cancer models was explored combining in vitro and in vivo approaches. Our analyses revealed that ligand-induced Ret activation: (i) stimulates migration of breast cancer cells; (ii) rescues cells from anti-proliferative effects of endocrine treatment and (iii) stimulates expression of cytokines in the presence of endocrine agents. Indeed, we uncovered a positive feed-forward loop between the inflammatory cytokine IL6 and Ret that links them at the expression and the functional level. In vivo inhibition of Ret in a metastatic breast cancer model inhibits tumour outgrowth and metastatic potential. Ret inhibition blocks the feed-forward loop by down-regulating Ret levels, as well as decreasing activity of Fak, an integrator of IL6-Ret signalling. Our results suggest that Ret kinase should be considered as a novel therapeutic target in subsets of breast cancer.

A. Western bot (WB) analysis on lysates from MCF7/Aro, MCF7 and T47D cells was carried out with a Ret antibody (anti-human Ret antibody from Cell Signaling) in the upper panels. In the lower panel a WB analysis on extracts from E10.5 mouse embryos and virgin mouse mammary gland tissue (M.Gland), as well as lysates from MCF7 and J110 cells was carried out with a second Ret antibody that recognizes human and mouse Ret (Santa Cruz). The electrophoretic motility of mouse and human Ret can be seen in J110 and MCF7 cells. Tubulin was used as a loading control. B. Total RNA from mouse embryos and J110derived tumors was isolated and RNA levels of the Ret signaling partners (Ret, GDNF, GFRα1) was analyzed by semi qRT-PCR using specific primers. C. IL6 levels (pg/ml) were measured by ELISA in 4-day conditioned medium (CM) of MCF7 cultures treated as indicated. Results represent the mean ± s.d. of triplicate determinations from triplicate wells. n.d.: not detected. D. Steroid-deprived MCF7, MCF7/Aro or J110 cells (left panel) were treated for 3 days with 10nM 4A or 10nM E2 as indicated. Lysates were analyzed by WB with the indicated antibodies. E. J110 cells were serum-deprived and then seeded into the upper chamber of a transwell. The lower wells contained 0.5% FBS control medium or IL6 (100ng/ml). Migrated cells were fixed and stained. Data shown are the mean of 3 independent experiments and error bars represent s.e.m. *p<0.05 by t test.
A. Microarray slides were co-hybridized with cDNA from 6-day treated MCF7/Aro cells cultured in the 6 indicated conditions (as in Fig 2F). For each condition, triplicates were tested and independent hybridizations were performed. The results of the contrasts between different conditions (Contrast) were obtained. A name (Name) was assigned to identify each contrast in the posterior analysis. The number of annotated genes (n° of genes) altered for each contrast is indicated. B. MCF7/Aro cultures were treated under the same conditions used in the microarray. The mRNA levels of the indicated genes were analyzed by qRT-PCR using specific primers. The values were normalized to actin expression and represented as fold induction relative to expression in Δ4A-treated cells. Columns, means of the values ± s.e.m. from triplicates. C. MCF7 cells were serum-deprived and seeded into the upper chamber of a transwell. The lower wells contained 0.5% FBS medium (Control) plus IL6 (100ng/ml), CXCL10 (100ng/ml), CXCL11 (100ng/ml), the combination (CXCL10/11+IL6) or 10% FBS. The mean of 3 independent experiments and error bars represent s.e.m. *p<0.05 by t test. D. J110 cells were steroid-deprived and then treated for 3 days, with 10nM E2 ± 10nM Tamoxifen (E2+Tam) or ± 100nM Fulvestrant (E2+Ful). Total RNA was isolated and IL6 mRNA levels were analyzed by qRT-PCR. The values were normalized to 18S mRNA expression and represented as fold induction relative to expression in steroid conditions. Columns, means of the values ± s.e.m. from triplicates. **p<0.01 and ***p<0.001 by t test. Fig S4. A. MCF7 cells were pre-incubated with DMSO or NVP-AST487 (100nM), then treated 15 min with IL6 (100ng/ml) or GDNF (10ng/ml). Lysates were analyzed by WB with the indicated antibodies. B. Serum-deprived MCF7 cells were pre-incubated with DMSO or NVP-AST487 (100nM), then seeded into the upper chamber of a transwell. Lower wells contained 0.5% FBS alone (Control) or supplemented with GDNF (10ng/ml) or IL6 (100ng/ml). Migrated cells were fixed, stained and counted. Data shown are the mean of 3 independent experiments; error bars represent s.e.m. *p<0.05 ***p<0.001 ANOVA using t test. C. MCF7 cells were pre-incubated with DMSO or NVP-BBT594 (50nM), then treated 15 min with IL6 (100ng/ml) or GDNF (10ng/ml). A WB analysis was carried out with the indicated antibodies. gp130 IPs were probed with a pY-specific antibodies and then for a gp130. Fig S5.

A.
Groups of J110-tumor bearing mice were randomized and treated once daily with vehicle (10%EtOH in corn oil) or tamoxifen (100µg/d) for 3 weeks. Tumor volume was determined every 2 days. Points represent mean ± s.e.m. B-C. Groups of J110-tumor bearing mice were randomized and treated with vehicle (10%EtOH in corn oil), or tamoxifen (100ug/d). After one week (arrow), the tamoxifen-treatment group was randomized to continue with tamoxifen alone plus oral vehicle (Nmethylpyrrolidone/PEG300) or combined with the Ret inhibitor NVP-AST487 (50mg/Kg/day) (Tamoxifen+AST487). *p<0.05 by t test. In B, tumor weight at the end of the experiment is shown (n=8-13). Bars represent mean ± s.e.m.; In C, tumor volume was determined every 2 days. Points represent mean ± s.e.m. A representative experiment of two is shown. D. Tumors from individual mice in each treatment group shown in C were harvested 8 hours after the last treatment (vehicle, tamoxifen and tamoxifen+AST). Tumor lysates from 3 independent mice per treatment group were analyzed by WB using the indicated antibodies. Right, quantification of the indicated phospho-protein/protein was performed in 4-10 independent tumors for each group from 2 independent experiments using imageJ. Data shown the mean ± s.e.m. *p<0.05 or **p<0.01 by Mann-Whitney test. Fig S6. A-B. Tumors from mice treated as described in Fig 5C were harvested 8 hours after the final treatment. Paraffin blocks were prepared and sections were stained for proliferation (phospho-histone H3, pH3, 400X) or apoptosis (cleaved caspase 3, CC3, 200X) and quantified in B. pH3-positive and -negative cells were manually counted and expressed as the percentage (%) of positive cells/total cells counted per field. The proportion of the tumor area showing CC3 immunoreactivity was quantified using ImageJ. For each group n=4 tumors were examined and 10-20 fields per tumor were analyzed. *p<0.05 or **p<0.01 by t test for pH3. Representative pictures are shown in the upper panels. Scale bars: 12 and 24μm for pH3 and CC3, respectively. Table S1. Correlation of Ret-score with established clinical and histopathological parameters of breast cancer patients. The number of patients (n) and their relative frequencies (%) in the indicated molecular subtypes (Voduc et al, 2010)

Cell lines and cell culture conditions
MCF7 and T47D human breast carcinoma lines were cultured in DMEM or RPMI 1640, respectively. The J110 mouse mammary cancer cell line was obtained from Dr. Myles Brown (Dana Farber Cancer Institute, Boston); cells were cultured in DMEM-F12. MCF7 aromatase expressing cells (MCF7/Aro) cells were cultured in MEM EBS (Amimed). Supplements included 10% FCS (MCF7 and T47D) or 5% FCS (J110), 2mmol/L Lglutamine, 100IU/ml penicillin, 100ug/ml streptomycin or 50ug/ml gentamycin (J110) and, for MCF7/Aro, 1mmol/L sodium pyruvate and 1% nonessential amino acids were added. MCF7 and T47D Ret knockdown cells and control cells were generated and analyzed previously (Boulay et al, 2008). Cells were steroid deprived using phenol red-free medium supplemented with 10% charcoal-stripped FCS (Hyclone) for 2-4 days before initiating treatment. For the transcriptome analysis, treatment of MCF7/Aro cells was initiated 2 days post-seeding and the cells were treated with inhibitors (100nM) and ligands (10ng/ml GDNF and/or 1nM Δ4A) every second days for 6 days. To evaluate proliferation, 0.1x10 6 MCF7/Aro cells were seeded into 12-well plates. Steroid-deprived cells were treated with 1nM Δ4A, with letrozole, fulvestrant or tamoxifen (100nM) alone or in combination with 10ng/ml of GDNF every second day for 6 days. Cells were harvested by trypsinization, resuspended in PBS and counted using a Vi-CELL XR automated viable cell analyzer (Beckman).

Microarray
RNA from long-term (6 days) MCF7/Aro-treated cells (triplicate experiments) was obtained using the RNeasy Mini Kit (Qiagen) and hybridized to Affymetrix Human Gene 1.0 Array (Affymetrix) according to the standard Affymetrix protocols. Data analysis and gene filtering were done using R/Bioconductor (Gentleman et al, 2004). Signal condensation was done using only the RMA from the Bioconductor Affy package. Differentially expressed genes were identified using the empirical Bayes method (F test) implemented in the LIMMA package and adjusted with the false discovery rate method (Wettenhall & Smyth, 2004). Probe sets with a contrast signal adjusted to p value of <0.01 and log 2 fold-change of ≥1.0-fold in linear space were selected. The microarray data have been submitted to the Gene Expression Omnibus (GEO) http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?token=tfefhkmyummagli&acc=GSE41405) and assigned the identifier GSE41405. Pathway analysis was done using Ingenuity Systems software (Life technologies).

Quantitative and semi-quantitative RT-PCR
RNA was extracted using the RNeasy Mini Kit (Qiagen), then DNase-treated RNA (3µg) was reversed transcribed with Ready-to-go You-Prime First-Strand Beads (GE Healthcare), diluted (1:3) and 1-4 µl of cDNA dilution was used for the PCR reactions. Quantitative Real-Time PCR was carried out using specific primers (Supporting Information Table S5) with StepOne Real-Time PCR System (Applied Biosystems) instrument and software. For each primer set, a calibration curve was done and used for the calculations. Expression values were normalized to actin, cytokeratin 18 or 18S values. Semi-quantitative PCR was performed according to standard protocols using 22-26 amplification cycles.
Following 30 minutes on ice, lysates were cleared by centrifugation and the supernatant collected and stored at -80°C. Before freezing, a sample aliquot was diluted at 1:1000 in water for determination of the protein concentration with a commercially available protein Bradford assay (Bio-Rad) using bovine serum albumin as standard. Fifty g of each sample was separated by SDS-PAGE and blotted onto a PVDF immobilon membrane (Millipore). After 1 hour blocking with 20% horse serum (HS) in PBS, and 0.1% Tween 20, filters were probed overnight (4°C) with specific primary antibodies (dissolved 1:1000 in 10% HS in PBS 0.1% Tween 20). The antigen-antibody complexes were visualized using horseradish peroxidase-conjugated anti-mouse or rabbit IgG secondary antibody (Amersham, GE Healthcare) and the enhanced chemiluminescence was detected using ECL Western blotting System (Amersham, GE Healthcare). The membrane was stripped and re-probed.
For IPs, 400 µg of cell lysates were incubated overnight with specific antibodies. Immune complexes were collected with protein A-Sepharose (Sigma) and washed three times with lysis buffer. Precipitated proteins were released by boiling in sample buffer and subjected to SDS-PAGE. The proteins were blotted and analyzed as described (Badache & Hynes, 2001).

IL6-specific ELISA
4 x 10 5 MCF7/Aro cells were plated onto 60 mm dishes and after 3 days, cultures were rinsed with PBS, steroid-deprived medium was added and 24 hours later, treatments were begun. Conditioned medium was harvested 4 days later and assayed for IL6 protein levels using the DuoSet Human IL6 ELISA (R&D System) according to the manufacturer instructions.

In vivo experiments
Mice were housed under hygienic conditions according to the Swiss guideline governing animal experimentation and experiments were approved by the Swiss veterinary authorities. Tumor growth and body weight were monitored every 2 days. Ret KD T47D cell lines and the control T47D cells were tested in female BALB/c nude mice, with 7-8 mice per group. E2 pellets, 0.025mg/90 days of release (Innovative Research of America) were implanted into the mice 1 week before 7.5 x 10 6 cells were injected. For the J110 model, 1 x 10 4 cells in 100µl of PBS were injected into the second mammary fad pad of 5 week old female FVB/N mice. When J110 tumors reached ~ 100mm 3 (10-12 days after injection) mice were randomized into treatment groups that received vehicle or NVP-AST487 as indicated. For the combination treatments, the mice were first randomized into three groups that received vehicle, NVP-AST487, fulvestrant or tamoxifen. Each group received the The corresponding vehicles were administrated in each group. After 10 days of treatment, fulvestrant-treated or tamoxifen-treated tumor-bearing mice were regrouped: one group received fulvestrant+AST487 or tamoxifen+AST487, the other continued on fulvestrant or tamoxifen alone. Single-agent and the combination treatment arms were followed for a total of 3 weeks.
NVP-AST487 was formulated by dissolving the powder in N-methylpyrrolidone/PEG300 (1:10 v/v). A fresh solution was prepared every other day and administered orally once daily (50mg/Kg/day). Fulvestrant and tamoxifen (Sigma) were prepared by dissolving the powder in 10% ethanol corn oil (Sigma) to obtain 1mg/100µl and 100µg/100µl, respectively. Fulvestrant was given three times per week by s.c. injection (3mg/week). Tamoxifen was given daily by i.p. injection (100µg/day). No changes in the body weight in the NVP-AST487, fulvestrant or the tamoxifen treated mice were measured. Tumor volumes were determined according to the formula: length x diameter 2 x π/6. Tumors were collected at the end of the experiment, 8 hours after the last drug administration and weighed. In order to visualize and count metastases, lungs were placed into Bouin's solution (fixation and staining). The number of nodules in the whole lung was obtained by counting. The metastatic index was calculated as the number of lung foci/tumor gram.

Immunohistochemistry
IHC analysis of J110 tumors for different markers was carried out in paraffin sections using the Ventana DiscoveryXT instrument (Roche Diagnostics) with the following antibodies: phosphoS10-Histone H3 9701 (pH3, Cell Signaling, 1:100, DAB Map XT protocol), D175cleaved caspase 3 9661 (CC3, Cell Signaling, 1:100, DAB Map XT protocol) and pY705Stat3 9145 (pStat3, Cell Signaling, 1:100, DAB Map XT protocol). pH3-positive and -negative cells were counted under the microscope (400X) in 10 fields corresponding to different areas of the tumor. Results were expressed as a % of pH3-positve cells/counted cells/field ± s.d. For CC3 pictures (200X) covering the whole area of the tumor were taken. For pStat3, pictures (400X) of non-necrotic areas avoiding tumor edge were taken. The quantification was done using ImageJ software and expressed as positive area fraction/field ± s.d. for each staining.
For Ret IHC in patient samples, antigen retrieval was 20 minutes in an autoclave in citrate buffer pH 6.0. Sections were exposed 10 minutes to 3% H 2 O 2 and sequentially blocked for Avidin/Biotin for 10 minutes (Avidin/Biotin blocking kit, SP-2001, Eubio). Then sections were blocked for 1 hour in normal goat serum and the primary Ret antibody (anti-Ret sc-167, Santa Cruz, 1:50 in PBS 1% BSA) was left on overnight at 4° C. Sections were washed in PBS and then exposed 30 minutes with an anti-rabbit biotinylated antibody (1:200 in PBS), then washed 10 minutes with strepavidin HRP (IDetect Super Stain System, IDST1007, Zamponi), washed, stained with DAB and hematoxilin as a counterstain.