Breast cancer cell lines
The MDA-MB-231 cell line is ER, PR, and E-cadherin negative and expresses mutated p53 and was isolated from an invasive ductal carcinoma [6]. The MDA-MB-231 genome clusters with basal subtype of breast cancer. The cells lack HER2 and are commonly used as a model for triple-negative breast cancer. MDA-MB-231 cells are known to seed to lymph nodes upon orthotopic transplantation in mice. Various cell lines derived from MDA-MB-231 were selected for specific metastatic predilection among which line MDA-MB-231-B2M2 that preferentially metastasizes to brain [7]. The SK-BR-3 (SKBR3) cells were isolated in 1970 from pleural effusion of a breast adenocarcinoma patient. This cell line overexpresses Her2 and is a preferential model for Her2 positive breast cancer with metastatic potential without organ-specificity [8] .
Cell line culture procedure
Human astrocytes (ScienCell) were cultured in astrocyte medium (AM, ScienCell) supplemented with 1% astrocyte growth factors (AGS, ScienCell), 2% fetal bovine serum (FBS, ScienCell), and 1% penicillin and streptomycin (P/S, ScienCell). Human umbilical vein endothelial cells (HUVECs, ScienCell) were cultured in endothelial cell medium (ECM, ScienCell) supplemented with 1% endothelial cell growth factors (ECGS, ScienCell), 5% FBS, and 1% P/S. Human astrocytes and HUVECs were used between passage 2 and 5. Breast cancer cell lines were cultured in RPMI-1640 with L-glutamine (BioWhittaker®) medium supplemented with 10% FBS and 1% P/S, and 1% gentamicin was added to the breast cancer cells. All cell lines were cultured at 37˚C, in a humidified incubator with 5% CO2.
Activation and Expansion of T-lymphocytes
Human bulk T lymphocytes isolated from healthy donors were expanded on irradiated allogeneic feeder cells, consisting of a mixture of 40-Gy gamma-irradiated peripheral blood mononuclear cells, Epstein-Barr Virus (EBV)-transformed B-lymphoblast cell lines BSM (also known as GM06821, GLCneg/HLA-A2pos) and APD (also known as GM06817, EADneg/HLA-A1pos) cells. Cells were cultured in RPMI-Hepes (Gibco) with 1% L-glutamine (BioWhittaker®), 1% P/S and 6% human serum albumin (HSA), in combination with phytohemagglutinin-L (PHA-L, Sigma) and IL-2 in a 96-well flat-bottomed plate for 6-7 days at 37˚C, in a humidified incubator with 5% CO2. After incubation, cells were harvested, centrifuged, and cultured in RPMI-Hepes medium supplemented with 6% HSA and IL-2 (360 IU/mL, Chiron, Amsterdam, The Netherlands).
Construction of the in vitro blood-brain barrier (BBB) model
The specifications of the BBB in vitro model were published previously by us [5, 9]. HUVECs were co-cultured with human astrocytes on the opposite sides of a transwell insert [9]. Inserts with pore sizes of 3.0 μm were used. In order to prevent direct cell-cell contact for part of the experiments, a 0.4 µm pore membrane physical barrier between both cell types was applied, or following incubation of the tumor cells with the T lymphocyte CM.
Functional studies using the in vitro BBB
To investigate the influence of T lymphocytes on the ability of the various cancer cells to cross the BBB, three parallel sets of experiments were conducted: 1) cancer cells were co-cultured with T lymphocytes in a 3:1 ratio to investigate the effects of direct cell-cell contacts; 2) cancer cells were cultured with cell-free conditioned media (CM) from activated T lymphocytes to investigate the effects of activated T lymphocyte-derived soluble factors; 3) cancer cells cultured without the addition of T lymphocytes, or T lymphocyte CM, were used as controls. Co-culturing was performed for five consecutive days at 37˚C, with 5% CO2. The culturing procedure and visualization of passed cells are described previously [5]
Cell pellet preparation for LC-MS and RT-PCR measurements
Co-culturing experiments were repeated twice for proteomics and genomics measurements. Following the first five days of the co-culture procedure, the T lymphocytes were removed by three consecutive vigorous PBS-washing steps. The tumor cells were trypsinized, collected into sterile tubes, and washed with PBS in six consecutive steps. During the last wash step, cells were collected in 1.5 ml Eppendorf tubes, centrifuged for five minutes at 1,000 g, and the presence of residual T cells was excluded by screening for CD3, CD4, and CD8. The supernatant was removed and the cell pellets were snap-frozen in dry ice and stored at -80˚C until further use.
Protein digestion and proteomics measurements
Frozen cell pellets were re-suspended in 50 μL 0.1% RapiGest buffer (Waters Corporation, Milford, MA) and cells were disrupted by external sonification for one minute at 70% amplitude at a maximum temperature of 25˚C (Branson Ultrasonics, Danbury, CT). A reducing agent, 0.5 M DL-dithiothreitol (Sigma-Aldrich), was added to a final concentration of 5 mM and the samples were incubated at 60°C for 30 minutes. Iodoacetamide (Sigma-Aldrich) alkylating reagent was added to a final concentration of 15 mM and samples were kept in the dark for 30 minutes. For protein digestion to each sample, 0.5 µg trypsin gold, mass spectrometry grade (Promega, Madison, WI), was added and incubated overnight at 37 °C. To stop the digestion, the pH of the solution was adjusted to pH < 2 with trifluoroacetic acid (TFA, Sigma-Aldrich) to a final percentage of 0.5%. Protein samples were incubated at 37˚C for 30 to 45 minutes. Subsequently, these samples were spun at maximum speed for 40 minutes at 4˚C. The supernatant was transferred into sterile LC-vials to be measured with LC-MS, as previously described [5].
Proteomics data analysis
Samples were measured with LC-MS and raw data files from the Orbitrap Fusion mass spectrometer were processed as described before [5]. Scaffold software (version 4.8.3, Portland, OR) was used to identify peptides and proteins as previously described [5]. Identified proteins of each type of breast cancer cell lines, that were co-cultured with either activated T lymphocytes or T-lymphocyte CM were compared to those of cancer cells cultured without the addition of activated T lymphocytes or T-lymphocyte CM. The comparison was based on Anova analysis using Scaffold software, and the p values of all proteins were considered differentially expressed if p < 0.05.
The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE [10] partner repository with the dataset (identifier XXX).
The selection procedure of relevant proteins was based on significant expressional difference before and after T lymphocyte or CM contact and functional properties related to cell migration and/or cell locomotion.
siRNA Knockdown
A mix of four siRNA sequences that target CORO1A mRNA (5’-CCUCAAGGAUGGCUACGUA-3’, 5’-CCAUGACAGUGCCUCGAAA-3’, 5’-CCGCAAAGGCACUGUCGUA-3’ and 5’-GUGCAGUGUUCGUGUCGGA-3’) and scrambled non-targeting siRNAs (5’-UGGUUUACAUGUCGACUAA-3’, 5’-UGGUUUACAUGUUGUGUGA -3’, 5’-UGGUUUACAUGUUUUCUGA-3’ and 5’-UGGUUUACAUGUUUUCCUA-3’) were obtained from Dharmacon (GE health care, Netherland). Breast cancer cells were transfected following the manufacturer’s protocol (using Lipofectamine® RNAiMAX Reagent, Invitrogen). The efficiency of CORO1A knock-down was assessed after 24 and 72 hours, at mRNA and protein levels.
Immunofluorescence microscopy
Breast cancer cells were trypsinized and 5x104 cells were fixed in formaldehyde for 15 minutes and kept in 100% ethanol until cytocentrifugation and staining with an antibody against coronin-1A (1:250, sc-100925, Santa Cruz Biotechnology). Cells from two independent experiments were analyzed by confocal microscopy using a LSM700 Zeiss microscope (Zeiss, Oberkochen, Germany). Images were captured with a 20X lens and a 4x4 tile was made from each slide. Each image was divided in 16 individual microscopic fields from which signal intensity as a mean ± SD was determined using ImageJ software [11].
Breast cancer tissue sample selection
Primary ER- breast cancer FFPE samples from patients who developed brain metastasis (n = 23) and from patients who developed metastases to other organs, excluding brain metastases (n = 29) were collected. None of the breast cancer patients received neoadjuvant therapy (clinical information of n= 20 samples has been previously provided in [5], and from the validation set n = 32 in [12]). This study was approved by the Medical Ethics Committee of the Erasmus Medical Center, Rotterdam, The Netherlands (MEC 02·953) and performed in adherence to the Code of Conduct of the Federation of Medical Scientific Societies in the Netherlands (http://www.fmwv.nl/).
Morphological assessment of breast cancer samples
Hematoxylin and eosin (H&E)-stained 5-µm-thick sections from each sample, prepared before and after sectioning for RNA isolation and IHC, were evaluated by an experienced pathologist. Tumor cell areas were selected microscopically and a minimum content of 60% tumor tissue was taken as a threshold for RNA isolation.
Reverse transcription-polymerase chain reaction (RT-PCR)
Total RNA was extracted from frozen cell pellets using the RNeasy Plus Micro kit (Qiagen, Hilden, Germany) and from FFPE patient material, extracted from 10-15 five μm sections (depending on the size of the sample) using the RNeasy FFPE Micro kit (Qiagen, Hilden, Germany), according to manufacturer’s protocols. The quantity and quality of the isolated RNA were assessed (2100 Bioanalyser, Agilent Technologies, Santa Clara, USA). Samples were excluded if the yield did not reach a minimum of 100 ng RNA/μl. Reverse transcription was performed using the RevertAid H-Minus first-strand cDNA synthesis kit (Thermo Scientific, Vilnius, Lithuania) according to the manufacturer’s protocol. Quantitative real-time PCR (RT-PCR) was performed using TaqMan Master Mix (Applied Biosystems, Austin, USA) on the 7500 RT-PCR system, v.2.3 (Applied Biosystems, Foster City, USA). The following commercially available exon-spanning TaqMan Gene Expression Assays (Applied Biosystems, California, USA) were used: CORO1A (Hs00200039_m1), HPRT1, exon 2-3 (Hs02800695_m1) and HMBS, exon 13-14 (Hs00609296_g1). HPRT1 and HMBS were used as reference genes. The relative quantification of target gene expression was performed using the 2ΔΔCt comparative method.
Immunohistochemistry
Anti-coronin-1A (1:2000, sc-100925, Santa Cruz Biotechnology) antibody was used according to the manufacturer’s instructions. All IHC slides were scanned on a Nano-zoomer 2.0HT scanner (40x magnification, Hamamatsu Photonics, Japan) and four tumor regions of interest (ROI’s) per slide were randomly selected. ROI’s were evaluated with a semi-quantitative IHC method [23]. In short, ImageJ Fiji 1.52p software (USA) was used to deconvolute the selected immunostained ROI’s and convert the slides into gray shades. Finally, intensity scores are based on the mean grey value per ROI’s and total area intensity ratio of each specific IHC coronin-1A staining. All four ROI’s per slide were implemented in the evaluation.
Publicly available mRNA expression data assessment and proteomics data availability
Validation of mRNA CORO1A levels was performed using a primary breast cancer dataset in the publicly available repository Human Cancer Metastasis Database (HCMDB; http://hcmdb.i-sanger.com/index). HCMDB consists of 124 previously published transcriptome datasets collected from Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA). RNA expression status of CORO1A was evaluated in 204 GEO primary breast cancer samples available in the National Centre for Biotechnology Information (NCBI) under GEO accession number GSE12276. The Log2 Medi-an-Centered ratio was used to evaluate the differential expression of CORO1A in primary breast cancer of patients who developed brain metastasis compared to patients who developed metastasis to other organs.
The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE [20] partner repository. Dataset information is publicly available under GEO accession number GSE12276, EXP00013 [7]. The data presented in this study are available in this article (and supplementary material).
Statistical analysis
Prism 5.0 (GraphPad Software) was used to perform statistical tests. A two-tailed Student’s t-test or ANOVA was used to determine differences in sample means. Data are presented as means ± SD. In all statistical analyses, a p-value < 0.05 was considered statistically significant. Unless otherwise stated, in vitro experiments were repeated independently three times.