Z-VAD-FMK (S7023), Erlotinib (S7786), BKM120 (S2247), Afatinib (S1011), Alpelisib (S2814) and sodium carboxymethyl cellulose (S6703) were purchased from Selleck Chemicals. Fetal bovine serum (FBS) (631106) was obtained from Takara Bio. 4-20% precast polyacrylamide gels (5671094) were purchased from Bio-Rad. Immobilon PVDF transfer membrane (IPVH00010), mini protease inhibitor cocktail (11836153001) and propidium iodide (P4864) were purchased from Millipore-Sigma. PBS (21-031-CV), EGF (354042) and Phenol Red-Free LDEV-Free Matrigel (356237) were purchased from Corning. Sodium orthovanadate (P0758S) was from Fisher Chemicals. NP40 Cell Lysis Buffer (FNN0021), Penicillin-Streptomycin (15140-122), tissue extraction reagent (FNN0071) and RPMI 1640 medium (11875-093) were obtained from ThermoFisher Scientific.
Antibodies
Anti-phospho-HER3 Y1289 (4791S), anti-HER3 (12708S), anti-EGFR (2232), anti-PTEN (9559S), anti-phospho-EGFR Y1068 (2234), anti-phospho-AKT S473 (9271S), anti-AKT (9272S), anti-phospho-MAPK T202/Y204 (9101), anti-phospho-P70S6K T389 (9234S), anti-P70S6K (9202S), anti-phospho-S6 S235/236 (4856S), anti-S6 (2217S) and anti-PI3K (4249S) were purchased from Cell Signaling Technology. anti-HSC70 (sc-7298) and anti-ERK2 (D-2, sc-1647) antibodies were purchased from Santa Cruz Biotechnology. Goat Anti-Mouse IgG HRP conjugate (172-1011) and Goat Anti-Rabbit IgG HRP conjugate (1721019) were purchased from BioRad.
Cell lines
MDA-MB-468, BT20, MDA-MB-231 were obtained from ATCC, maintained in RPMI 1640 medium with 10% FBS and 1% penicillin/streptomycin, and grown at 37 °C, 5% CO. These cell lines were recently authenticated and tested for mycoplasma contamination.
Caris data
Next-generation sequencing (NGS)
NGS was performed on genomic DNA using the NextSeq (592-whole gene targets; Agilent Technologies, Santa Clara, CA, USA) or NovaSeq 6000 (>700 genes at high coverage/read-depth and >20,000 genes at lower depth; Agilent Technologies, Santa Clara, CA, USA) platforms (Illumina, Inc., San Diego, CA, USA) as described previously [38]. The copy number alteration (CNA) of each exon is determined by normalizing the sequencing depth of each exon divided by the average sequencing depth of the sample and comparing it to the pre-calibrated mean of normalized values in the training data (The mean values are re-calibrated every 60 days with up to 10,000 samples). A gene was defined as amplified by having segment level of copies ≥4.5.
Survival analysis
Caris CodeAI™ clinico-genomic database containing insurance claims data were used to calculate real-world overall survival (OS) from tissue collection to last contact. Kaplan-Meier curves were generated to calculate molecularly defined patient cohorts as previously described [39, 40].
cBioPortal data
Datasets from TCGA, METABRIC and MSKCC were searched for EGFR amplification, as well as mutations in PIK3CA, PIK3R1, PIK3R2, PIK3R3, PTEN, AKT1, AKT2, AKT3, RPS6KB1, RPS6KB2 and RPS6 in cBioPortal (cbioportal.org) [41,42,43]. Amplification was defined by GISTIC2.0 score = 2 (CN ≥ 4-5) [44].
Western blot
Cell lysates were collected, prepared and immunoblotted as previously described, and repeated at least 3 times [45]. Densitometric analysis of band intensities was calculated using ImageJ and data are presented as average ± SEM.
Fluorescence in situ hybridization (FISH)
To prepare metaphase chromosomes, cells were treated with 0.02 mg/ml Colcemid (Invitrogen, Grand Island, NY) for 1 h, suspended in a hypotonic solution for 20 min, fixed using a methanol/acetic acid mixture (1:3 ratio) and dropped onto slides in a cytogenetic drying system (Thermotron, Holland, MI). Dual color FISH was performed with custom probes for EGFR and centromere 7 from Cytotest (Rockville, MD), labeled with Dyomics dyes (Jena, Germany). Slides were pepsin treated, washed in 1× PBS, dehydrated in an ethanol series, and left to air-dry. Co-denaturation was performed at 72 °C for 2 min, followed by overnight hybridization at 37 °C. The slides were detected in 2× SSC/0.3% Nonidet P-40 for 2 min at 48 °C, and 2× SSC/0.1% Nonidet P-40 for 1 min at room temperature and were mounted with DAPI antifade solution (Vector Laboratories, Burlingame, CA). Images were captured using a Leica DM-RXA fluorescence microscope (Leica, Wetzlar, Germany) equipped with a 40× objective and custom optical filters (Chroma, Bellows Falls, VT).
Sanger sequencing
Sanger DNA sequencing of genomic DNA from the MDA-MB-231, MDA-MB-468 and BT20 cell lines for PIK3CA mutations was performed using the Big Dye Terminator v.1.1 Cycle Sequencing Kit (Applied Biosystems, CA) according to the manufacturer's specifications and run on an ABI 3130xl or 3730 Genetic Analyzer (Applied Biosystems, CA) at the CCR Genomics Core at the National Cancer Institute.
Cell death
Cell death was determined by CytoTox-Glo (G9291) as directed by Promega or by propidium iodide uptake. Propidium iodide (PI) staining, imaging and quantification was performed using the BioTek Cytation 1 Imaging Reader from Agilent (Santa Clara, CA) and Gen5 Image Prime 3.14 software (Agilent). For CytoTox Glo, cells were pretreated with 20 µM ZVAD for 1 h, followed by co-treatment with DMSO, 10 µM erlotinib, 1 µM BKM120 or the combination for 48 h. Dead cell percentage was then calculated according to the manufacturer instructions. For PI, 5000 cells were plated in a 96-well plate and treated the next day with experimental conditions in 1% FBS RPMI, along with PI (1:3000 dilution). Dead cell percentage was calculated by dividing dead cell number (PI stained) by brightfield cell count number and multiplying by 100 and normalized relative to the Day 0 dead cell percentage. In experiments using ZVAD-FMK, cells were pretreated +/- 20 µM ZVAD-FMK for 1 h, followed by co-treatment with experimental conditions.
Cell viability
10,000 cells were plated in a 96-well white walled plate (3610, Corning) and treated the following day with experimental conditions for 48 h in 1% FBS RPMI. Viability was determined by CellTiter-Glo 2.0 (Catalog #G9242) as directed by Promega.
Cell cycle
Cells were plated in 6-well plates (3 × 10 cells/well) and treated with EGFR and/or PI3K inhibitors for 48 h. DNA content was measured by flow cytometry as previously described, using a BD LSRFortessa™ (BD Biosciences) flow cytometer and analyzed using FlowJo® software (FlowJo LLC) [46, 47].
Animal studies
Sixty female athymic nude mice between 5 and 6 weeks in age obtained from Charles River Laboratories were housed and treated according to approved NCI-ACUC guidelines and in accordance with an approved NCI Animal Care and Use Committee (IACUC number WMB-004). 5 million BT20 cells suspended in matrigel (1:1 with PBS) were injected subcutaneously into the mammary fat pad with a 25 G needle (Terumo Medical Care Solutions; SS-01T2516). When tumors were 100 mm the mice were randomized into four groups of 15. The mice were randomized so each group included similar average tumor size, with similar standard deviation. 0.5% sodium carboxymethyl cellulose (CMC) in water (w/v) was used as control, and 20 mg/kg afatinib or 20 mg/kg alpelisib were suspended in 0.5% CMC. Control or 20 mg/kg alpelisib, 20 mg/kg afatinib, or the combination of both was given by oral gavage (1 inch, 22 G needle) once a day for five days each week for the duration of the study. Animals were not included in treatment or analysis if the tumors did not reach 100 mm at the time of randomization. After the first four days of treatment, four mice per group were sacrificed, tumors were excised and proteins were harvested in Tissue Extraction Reagent I (Invitrogen-FNN0071) according to the manufacturers protocol. Tumor caliper measurements and body weight were measured twice a week for the duration of the study, as required by humane endpoints for all mice. Investigators were blinded to the tumor measurements. Tumor volumes were calculated according to the formula V = ½ (length × width). When maximum tumor size for each animal was reached (20 mm in any direction) the mice were humanely euthanized.
Statistical analysis
Student's t test with 2 tailed comparisons assuming equal variance, and one-way or two-way ANOVA were performed as indicated. P-values of ≤0.05 were considered significant. Synergy scores were calculated using SynergyFinder.org [48]. In vitro experiments consisted of triplicates per treatment group, with at least three experimental repeats performed (with specific repeat numbers indicated in figure legends). In vivo experiments contained n = 15 per treatment group.