SUPPLEMENTARY INFORMATION
MiR-16 mediates trastuzumab and lapatinib response in ErbB-2-positive breast and gastric cancer via its novel targets CCNJ and FUBP1.
Leandro Venturutti, Rosalía I. Cordo Russo, Martín A. Rivas, María F. Mercogliano, Franco Izzo, Robert H. Oakley, Matías G. Pereyra, Mara De Martino, Cecilia J. Proietti, Patricio Yankilevich, Juan C. Roa, Pablo Guzmán, Eduardo Cortese, Daniel H. Allemand, Tim H-M. Huang, Eduardo H. Charreau, John A. Cidlowski, Roxana Schillaci, Patricia V. Elizalde.
SUPPLEMENTARY FIGURE LEGENDS
Supplementary Figure S1. (a)TZ effects on cell proliferation of ErbB-2-positive BC cell lines. Cells were treated with TZ for the stated times and proliferation was measured by [3H]-Thymidine incorporation, setting the value of untreated cells to 1. (b) ErbB-2 silencing induces miR-16 upregulation and growth arrest in TZ- and lapatinib-sensitive BC cells. Cells were transfected with siRNAs targeting ErbB-2 or a Control siRNA. Left panel: miR-16 levels were determined by RT-qPCR. The fold change of miR-16 expression levels was calculated by normalizing the absolute levels of miR-16 to those of U6 snRNA, setting the value of Control-siRNA-transfected cells to 1. Middle panel: cell proliferation was measured by [3H]-Thymidine incorporation. Right panel: control of ErbB-2 inhibition by siRNAs. (c and d)Cells were treated with TZ for the stated times, and cell proliferation was measured as in (a). (e and f) TZ induces CCNE downregulation only in TZ-sensitive BC cells. CCNE levels were analyzed by WB in protein extracts from cells treated with TZ for the stated times or from tumors from experiments in Figures 1b and 1d. Shown are two representative tumors from each group. ß-tubulin expression was used as loading control. Numbers under each blot represent the corresponding densitometric quantification. Images shown in (b), (e) and (f) are representative of a total of three experiments, all with similar results. Data shown in (a-d) represent the mean ± s.e.m. of three independent experiments.Related to Figure 1.
SupplementaryFigure S2. (a)Pre-miR-16 transfection efficiency in BC cells.Cells were transfected for 48 h as stated and miR-16 levels were assessed by RT-qPCR. (b) MiR-16 inhibitsin vitro growth of ErbB-2-positive BC cells. Cells were transfected with pre-miR-16 or pre-miR-Control for 48h and proliferation was measured by cell count. Results are presented as inhibition percentage, setting the value of non-transfected cells to 100%.(c) Pre-miR-16 transfection efficiency in GC cells. Cells were transfected for 48 h as statedand miR-16 levels were measured as in (a). Data shown in (a-c) represent the mean ± s.e.m. of three independent experiments.(d)Efficacy of the pre-miR delivery method used in in vivo studies. Nude mice were inoculated subcutaneously with JIMT-1 cells (3x106/mice). Once tumors reached 100mm3, animals (n=3 per group) received a single intratumoral pre-miR-16 or pre-miR-Control injection. Animals were sacrificed 72 h later and miR-16 levels were determined by RT-qPCR in total RNA extracts from tumor cells. Results are representative of a total of three experiments, all with similar results.Related to Figure 2.
Supplementary Figure S3. (a) Lapatinib induces miR-16 upregulation and cell growth arrest in lapatinib-sensitive BC cells. Cells were treated with lapatinib for 24 h. Left panel: miR-16 levels were determined by RT-qPCR. Right panel: cell proliferation was measured by [3H]-Thymidine incorporation. (b) WB analysis of whole cell lysates,performed with the indicated phospho (p) antibodies. Filters were reprobed with the respective total antibodies. (c) Cells were transfected with miR-16 inhibitors or control, and then treated with lapatinib for 24 h. Proliferation was assessed as in (a). (d) Lapatinib-resistant cells were treated with lapatinib for 24 h. Left panels: miR-16 levels were determined by RT-qPCR. Right panels: cell proliferation was measured as in (a). (e) WB analysis was performed as in (b). Data shown in (a), (c) and (d)represent the mean ± s.e.m. of three independent experiments. *P<0.05; ***P<0.001. Images shown in (b) and (e) are representative of a total of three experiments, all with similar results.Related to Figure 3.
Supplementary Figure S4. (a) Primary cultures of C4HD BC tumors were transfected with pre-miR-16 (n=3) or pre-miR-Control (n=3) for 48 h and miR-16 levels were assessed by RT-qPCR. (b) Gene expression profile of ErbB-2-positive BC cells overexpressing miR-16.Genome-wide gene expression studies were performed using samples in (a) and whole mouse genome multiplex format oligo arrays. Heat Map visualization shows the expression profile of the top 400 most significantly down- and upregulated genes in the microarray studies. Variables were standardized with mean 0 and standard deviation 1 before hierarchical clustering was performed. (c) Validation of microarray results for CCNJ and FUBP1. Primary cultures of C4HD tumors were transfected with pre-miR-16 or pre-miR-Control and mRNA levels of the gene of interest (GOI) were assessed by RT-qPCR. Fold change of mRNA levels was calculated by normalizing the absolute levels of GOI mRNA to those of GAPDH, used as an internal control, setting the value of pre-miR-Control-transfected cells to 1. Data shown in (c) represent the mean ± s.e.m. of three independent experiments.Related to Figure 5.
Supplementary Figure S5.(a)Nuclear FUBP1 expression levels in ErbB-2-positive BC tumor samples were evaluated by IHC and scored considering intensity and percentage of positive cells, as described in SupplementaryMaterials and Methods. Shown are representative images fromtumor samples displaying different nuclear staining intensities. (b) Validation of the FUBP1 antibody used for IHC. Left panel: FUBP1 (green) was localized by IF and confocal microscopy in SKBR-3 cells transfected with FUBP1 or Control siRNAs. Nuclei were stained with propidium iodide (red). Right panels: quantitative analysis of FUBP1 staining. Total and nuclear fluorescence intensities (integrated intensity per unitarea) were quantified in 50 cells from each treatment and are plotted as mean± s.d.Related to Figure 8.
SUPPLEMENTARYMATERIALS AND METHODS
Antibodies
The following antibodies were used for Western blot: anti-CCNE1 (M-20, sc-481), anti-ErbB-2 (C-18, sc-284), anti-pErk (E-4, sc-7383) and anti-Erk2 (C-14, sc-154), from Santa Cruz Biotechnology (Santa Cruz, CA, USA); anti-c-Myc (D84C12), anti-pSer473-Akt and anti-Akt, from Cell Signaling Technology (Beverly, MA, USA); anti-CCNJ (ab104895), anti-FUBP1 (ab28732)and anti-pSer62-c-Myc (ab78318),from Abcam (Cambridge, MA, USA); anti-β-tubulin (T0198),from Sigma-Aldrich (Saint Louis, MO, USA); HRP-conjugated secondary antibodies, from Vector Laboratories (Burlingame, CA, USA). Antibodies used for chromatin immunoprecipitation (ChIP) assays were: anti-FUBP1 (ab28732), anti-c-Myc (N-262, sc-764X, Santa Cruz),anti-acetyl-Histone H4 (06-866, Millipore, Temecula, CA, USA) and anti-trimethyl-Histone H3 (Lys9) (07-442, Millipore). Rabbit IgG (Sigma-Aldrich) was used as negative control.
Reagents
Trastuzumab (TZ, Herceptin™) was from Genentech Inc. (San Francisco, CA, USA). Lapatinib (Tykerb™) was from GSK (Brentford, Middlesex, UK). LY294002 was purchased from Sigma-Aldrich. The MEK 1/2 inhibitor U0126 was purchased from Cell Signaling.
Cell proliferation
Proliferation assays were performed in complete growth media. Cell proliferation was evaluated by incorporation of 1μCi [3H]-thymidine (New England Nuclear, DuPont, Boston, MA, USA; specific activity 20 Ci/mmol) during the last 16 hours of treatment, as previously described.1 Proliferation was also assessed by counting cells in the presence of Trypan Blue 48 h after transfection with pre-miR-Control or pre-miR-16.
siRNA transfections
DharmaFECT (Dharmacon, Lafayette, CO, USA) was used for siRNA transfection.siRNAs targeting ErbB-2, c-Myc, CCNJ or FUBP1 were synthesized by Dharmacon. Experiments were performed with at least two different siRNA sequences for each protein, but we here presented results obtained with only one of them. siRNA sequences were as follows: ErbB-2 siRNA #1, 5’-GGACGAAUUCUGCACAAUG-3’; ErbB-2 siRNA #2, 5’-GACGAAUUCUGCACAAUGG-3’; CCNJ siRNA #1, 5’-GAACGACUGUUGAUCGCUC-3’; CCNJ siRNA #2, 5’-UGGGUUGUAUGACUAAUAU-3’; FUBP1 siRNA #1, 5’-GACAACAAGCAGCCUAUUA-3’; FUBP1 siRNA #2, 5’-GACAAACCUCUUAGGAUUA-3’; c-Myc siRNA #1, 5’-GAAACGACGAGAACAGUUG-3’; c-Myc siRNA #2, 5’-GGACACACAACGUCUUGGA-3’. A control siRNA oligonucleotide from Dharmacon that does not target any known mammalian gene was used as a negative control.
Pre-miR and anti-miR transfections
miRNA precursors and inhibitors were obtained from Life Technologies and were used in accordance with manufacturer's instructions. Briefly, pre-miR-16 (AM17100) or a pre-miR-Control (AM17110) that does not form any known mammalian miRNA, or anti-miR-16 (MH10339), or anti-miR-Control (#4464076) were transfected using siPORT NeoFx transfection reagent (Ambion).
Plasmids transfections and luciferase assays
X-tremeGENE HP (Roche Biochemicals, Indianapolis, IN, USA) was used for plasmids transfection.pBABE-Myc-S62D, a phospho-mimicking mutant c-Myc expression vector, which carries a serine-to-aspartic acid substitution at codon 62 and acts as an aberrantly stable variant, was a kind gift from Dr PJ Hurlin (Oregon Health and Science University, Portland, OR, USA), and so was the pBABEpuro empty vector.2Luciferase constructs were bought from SwitchGear Genomics (Menlo Park, CA, USA) and contain the wild-type human CCNJ 3' UTR or FUBP1 3’ UTR downstream of the RenSP luciferase reporter gene, in the pSGG_3'UTR vector (luc-3’CCNJ-WT or luc-3’FUBP1-WT, respectively), or an EMPTY multiple cloning site (pLightSwitch_3UTR-Empty). A firefly luciferase expression plasmid (pSGG-3’UTR), used as control, was also purchased from SwitchGear Genomics. Additional constructs carrying a mutated miR-16 binding site (luc-3’CCNJ-MUT and luc-3’FUBP1-MUT), shown in Figure 5A, were synthetized by ACGT Corporation (Toronto, Ontario, Canada). For luciferase assays, cells were co-transfected with pSGG-3’UTR, which encodes Firefly luciferase, and a construct carrying the intact or mutated CCNJ or FUBP1 3’UTR region, or an empty plasmid. Cells were then treated with TZ or lapatinib, or re-transfected with miR-16 precursors or inhibitors. Firefly and Renilla luciferase activities were measured consecutively (Dual Luciferase Reporter, Promega, Madison, WI, USA). The former was used for normalization.
Western blots
Sodium dodecyl sulphate-polyacrylamide gel electrophoresis and immunoblots were performed as described.3Experiments exploring Akt, Erk or c-Myc phosphorylation levels were repeated three to five times.Signal intensities of phospho-proteins were analyzed by densitometry using Image J software (National Institutes of Health) and normalizedto total protein bands.Experimentsassessing total protein content were also repeated three to five times and signal intensities werenormalized to β-tubulin bands, used as loading control. Data analysis in BT-474, SKBR-3 and NCI-N87 cells showed a significant decrease in c-Myc, pAkt, pErk, CCNJ, FUBP1 and CCNE levels by TZ treatment at 24 and 48 h (p<0.001). Likewise, there was a significant reductionin c-Myc, pAkt, pErk, CCNJ, FUBP1 and CCNE levels in TZ-treated BT-474 xenografts, compared with IgG-treated animals (p<0.001). On the contrary, TZ treatment for 24 or 48h did not induce significant changes in c-Myc, pAkt, pErk, CCNJ, FUBP1 or CCNE expression levels in JIMT-1, BT-474 HR, HCC-1569 and SNU-I cells. In addition, there was no difference in CCNJ, FUBP1 or CCNE levels between TZ-treated and IgG-treated JIMT-1 xenografts. TZ treatment did significantly reduced pErk and p-c-Myc (S62) after 15, 30 and 60 minutes in BT-474 cells (p0.001), but not in JIMT-1 cells.Reconstitution assays in BT-474 cells showed that TZ treatment effectively reduced c-Myc levels in TZ-treated pBABE-puro-transfected cells, compared with untreated cells or TZ-treated pBABE-puro-MycS62D-transfected cells(p0.001). However, there was no significant difference in c-Myc levels between TZ-treated pBABE-puro-MycS62D-transfected cells and untreated cells.Treatment of BT-474, BT-474 HR and JIMT-1 cells with LY294002 significantly reduced pAkt and c-Myc levels (p0.001). In a similar manner, treatment with U0126 in these cell lines resulted in reduced pErk and c-Myc levels (p0.001). Lapatinib treatment at 1μM significantly inhibited pAkt, pErk and c-Myc in BT-474, SKBR-3 and BT-474 HR cell lines (p0.001), but not in JIMT-1 cells. Moreover, lapatinib at this concentration induced CCNJ and FUBP1 inhibition in BT-474 and BT-474 HR cells (p0.001), but not in HCC-1569 or JIMT-1 cells. On the contrary, lapatinib treatment at 5 μM did reduce c-Myc, pAkt, pErk, CCNJ and FUBP1 levels in JIMT-1 cells (p0.001).Transfection with pre-miR-16 significantly inhibited CCNJ and FUPB1 expression levels in all the studied cell lines, compared with pre-miR-Ctrl transfected cells (p0.001). Similarly, CCNJ and FUBP1 expression levels were significantly lower upon pre-miR-16 intratumoral injection in JIMT-1 xenografts (p0.001). Similar data analysis revealed thatanti-miR-16 transfection induced a significant increase in CCNJ and FUBP1 expression levels in all tested cell lines, compared with anti-miR-Ctrl transfected cells (p0.001). ErbB-2 silencing in BT-474 cells significantly reduced c-Myc, pAkt and pErk levels, when compared to control-siRNA transfected cells (p0.001). FUBP1 silencing in BT-474 cells led to a significant reduction in c-Myc and CCNJ levels when compared to control-siRNA transfected cells (p0.001). Differences between groups were analysed by unpaired 2-tailed Student’s t test or one-way ANOVA test.
Preclinical models
Animals were randomly assigned to treatment groups employing a parallel groups design. Briefly, each mouse from a single group was assigned to receive one treatment using a table of random numbers. Tumor volume and growth rate were determined as described4, and were assessed by a researcher blinded to the experimental arm.BT-474 cells (2x107/mouse) were injected subcutaneously (s.c.) in 2-month-old virgin female NIH(S)-nude mice (La Plata University, Buenos Aires, Argentina) concomitantly with a contralateral 0.72 mg estradiol pellet. Once tumors reached 150mm3, mice were administered TZ (5 mg/kg, intraperitoneally (i.p.)) or human IgG (5 mg/kg, i.p.) as control, twice a week. JIMT-1 cells (3x106 /mouse) were injected s.c. in nude mice. Once tumors reached 100mm3, mice were administered TZ or IgG twice a week, as stated above, and received pre-miR-16 or pre-miR-Ctrl intratumoral injections thrice a week.
Intratumoral pre-miR injections
For each injection, 8 ug of pre-miR-Ctrl or pre-miR-16 were complexed with 0.8 μl siPORT NeoFx reagent in 50 μlsterile phosphate-buffered saline (PBS).Palpable tumors were slowly injected thrice weekly withthe pre-miR solution using 33 guage needles. Mice were killed one day after the last treatment.
Establishment of BT-474 HR cell line
BT-474 HR cell line was obtained by prolonged in vivo treatment of BT-474 tumor xenografts with TZ. BT-474 tumor xenografts were stablished in nude mice as stated. Once tumors reached a volume of 300mm3, mice were treated twice a week with TZ (5 mg/kg) diluted in sterile PBS by i.p. injection. A tumor thatrecurred in the presence of maintained therapy with TZ was harvested, minced, and digested with 0.25% trypsin in DMEM containing antibiotics and DNase for 30 min at room temperature.BT-474 HR cells were then established according to a previously described method,5 and maintained for at least 40 passages in PRMI 10% SFB supplemented with TZ (10μg/ml).
Gene expression profile studies
C4HD tumor line displays high levels of estrogen and progesterone receptors, overexpresses ErbB-2 and ErbB-3, exhibits low ErbB-4 levels and lacks EGFR expression.6Gene expression analysis was conducted using Agilent Whole Mouse Genome 4x44 multiplex format oligo arrays (014868) (Agilent Technologies, Santa Clara, CA, USA) following the Agilent 1-color microarray-based gene expression analysis protocol. Starting with 300ng of total RNA, Cy3 labeled cRNA was produced according to manufacturer’s protocol. For each sample, 1.65µg of Cy3 labeled cRNAs were fragmented and hybridized for 17 hours in a rotating hybridization oven. Slides were washed and then scanned with an Agilent Scanner. Data was obtained using the Agilent Feature Extraction software (v9.5), using the 1-color defaults for all parameters. The software performed error modeling, adjusting for additive and multiplicative noise. Spots with signal intensities twice above the local background, not saturated and not flagged by Agilent Feature Extraction software were considered reliable. Pre-processing of data and quality control were conducted using R statistical software via the Agi4x44PreProcess (Pedro Lopez-Romero; PreProcessing of Agilent 4x44 array data. R package version 1.22.0) from Bioconductor.7Briefly, extracted intensities were background-corrected using the normexp method and were thenquantile normalized and log2-transformed.To determine genes that were differentially expressed between experimental conditions, we used the Limma Bioconductor package,8 applying an empirical Bayes function to moderate t-statistics obtained from the analysis, and the Benjamini-Hochberg multiple test correction to reduce the number of false positives. Probes with adjusted p-value <1% and additionally a fold change exceeding 1.9 in absolute value were selected as relevant.The expression heat map was built with differentially expressed genesusing CIMminer.9Cluster analysis of samples was performed using the Euclidean distance method and the average linkage cluster algorithm.