Methods:
Ethics statement: The animal study protocol was approved by The Institutional Animal Care and Use Committee of BIDMC Institutional Animal Care and Use Committee (IACUC), protocol number 086-2011, in accordance with the Public Health Service (PHS) Policy on Humane Care and Use of Laboratory Animals and the  Guide for the Care and Use of Laboratory Animals (NIH), for the protection of Vertebrates Used for Scientific Purposes (Scientific Procedures) Act 1986 as well as the ARRIVE guidelines. All animal studies were performed in BIDMC.
Breast cancer cell lines: Normal immortalized breast epithelial cells (MCF-10A) and the human breast cancer cell lines MDA-MB-231 were purchased from American Type Culture Collection (ATCC, Manassas, VA). MDA-MB-231 cells are TNBC cells lacking PR, ER and HER-2. The breast tumor metastatic brain variants include the MDA-231-BrM2 tagged to GFP as well as the murine mammary tumor cells TGL-4T1-BrM5 which specifically form metastatic brain tumors (Bos, Zhang et al., 2009; Rodriguez, Jiang et al., 2014; Avraham, Jiang et al., 2014) and generate multifocal lesions in the cerebrum, the cerebellum, and the brainstem, with features typical of brain metastasis in cancer patients.
MAGL inhibitors: MAGL inhibitor AM9928 was synthesized and characterized at the Center for Drug Discovery (CDD), Northeastern University. AM9928 inhibits human MAGL (hMAGL) with an IC50 value of 8.9 nM (Tyukhtenko, Ma et al., 2020; Tyukhtenko, Karageorgos et al., 2016; Tyukhtenko, Rajarshi et al., 2018). AM9928 lacked any affinity for the cannabinoid receptors CB1 and CB2 (Tyukhtenko, Ma et al., 2020).
Human Brain Microvascular Endothelial Cells : Human Brain Microvascular Endothelial Cells (HBMEC), and all HBMEC culture reagents, were purchased from Cell Systems (Kirkland, WA). Cells were maintained according to the manufacturer’s protocol and were cultured in CSC Complete Media on Attachment Factor pre-coated surfaces. Cells were grown to confluence and were maintained for 3-5 days for complete tight junction formation.
HBMEC and TNBC co-culture assay: Approximately 40,000 fluorescent-labeled MDA-MB-231 or MDA-MB-BrM2 cells were seeded over a confluent layer of HBMEC grown on glass coverslips. Cells were then allowed to attach at 37 °C in 5% CO2 overnight and incubation was ended by removing media, rinsing 2-3X with PBS and immediately adding 2% paraformaldehyde. For HBMEC treatment, the HBMEC monolayer was pre-incubated with either vehicle control or with MAGL inhibitor AM9928 (100 ng/mL or otherwise as indicated), or with vehicle control for 2 hours in media with 0.1% serum. This was followed by removal of the inhibitor and rinsing with PBS prior to addition of the breast cancer cells as to observe the effects of the inhibitor on HBMEC only as compared with controls.
Adhesion assay of MDA-MB-BrM2 cells: Approximately 40,000 fluorescent-labeled MDA-MB-231 or MDA-MB-BrM2 cells were seeded over a confluent layer of HBMEC grown on glass coverslips. The cells were then allowed to attach at 37 °C in 5% CO2 overnight and incubation was stopped by removing media, rinsing 2-3X with PBS and immediately adding 2% paraformaldehyde. For MAGL inhibitor treatment, HBMEC monolayer was pre-incubated with either vehicle control or with the MAGL inhibitor (100 ng/mL) for 2 hours in serum-starving conditions. This was followed by removal of the inhibitor and rinsing with PBS prior to addition of the breast cancer cells as to observe the effects of the inhibitor on HBMEC only. After samples preparation, the average number of attached labeled tumor cells was calculated across all samples.
Endothelial transmigration assay: HBMECs were dispersed using the CSC Passage Reagent Group kit (Cell Systems), counted, and resuspended in the appropriate volume of CSC complete media. Approximately 100,000 HBMECs were seeded into 24-well transwell inserts with 8μm pores (Costar Corp.) pre-coated with fibronectin and maintained for 5 days to allow for complete tight junction formation. The media was replaced every 2 days with fresh CSC complete media. For the assay, the apical (upper) and basal (lower) chamber media was removed. Approximately 40,000 fluorescent-labeled MDA-MB-BrM2 cells or MDA-MB-231 cells in 100μL aliquots were seeded into the apical chamber in serum-starved conditions (0.5% serum) in CSC media and allowed 5 hours to transmigrate to the basal side of the transwell. All basal chambers contained complete media to provide a chemotactic gradient for the breast cancer cells to follow. The HBMEC monolayer was pre-treated for 2 hours prior to the addition of breast cancer cells, either with vehicle control or with the MAGL inhibitor. After 5 hours, the apical side of the transwell filter was gently wiped clean with a cotton swab to remove all non-migrating cells and the basal side was dipped into PBS (with Ca2+ and Mg2+) and then fixed in 3.7% formaldehyde solution for 10 minutes. The number of transmigrated cells was determined by averaging the total number of green, fluorescent cells in 10 random fields at 200X magnification under a fluorescent microscope.
Isolation of Exosomes: The culture supernatants of MDA-MB-231 and MDA-MB-BrM2 cells untreated or treated with AM9928 for 24hours, at approximately 65% confluence and were harvested after 16 hours conditioning in serum-free media. Cells and debris were cleared from the supernatants by centrifugation (500 g, 10 min) followed by filtration using 0.22-micron filters (Millipore Inc.). Exosomes were prepared from cell-free supernatants using the Exosome Isolation Kit (Kit# EIK-01, Creative Biolabs, NY). The quantitative and qualitative analysis of exosomes were performed on double sandwich enzyme-linked immunoassay using the Total Exosome Capture & Quantification Kit (Kit#EQK-04Creative Biolabs).
Tissue processing and immunostaining: Methods were detailed previously in our studies (17,18). Briefly, brains were removed and divided sagittally. One hemi‐brain was fixed in 4% paraformaldehyde, pH 7.4, at 4°C for 48 h and sectioned using a Vibratome 2000 (Leica, Germany), while the other hemi‐brain was snap‐frozen and stored at −70°C for protein analysis or for BBB permeability analysis. Hippocampal brain tissues were dissected and post‐fixed, and treated with primary antibodies overnight at 4°, followed by the corresponding fluorescent secondary antibody (AlexaFluor 488 or Alexa Fluor 594, 1:1000 dilution; Invitrogen). The following antibodies were used: antibodies for ZO‐1 (1:500 dilution), claudin‐5 (1:500 dilution; Invitrogen), CD31 (1:500), Pan‐cytokeratin antibodies (1:500 dilution), GFP antibodies (1:50 dilutions; Abcam and Santa Cruz) and their respective controls. All slides/sections were processed and analyzed under the same standardized conditions. The immunolabelled blind‐coded sections were imaged with the LSCM (MRC1024, Bio-Rad).
In vivo effects of MAGL on BBB integrity and TNBC colonization in brain: Female (6wks) BALB/c mice were purchased from Jackson Laboratories (Bar Harbor, ME). The mice were housed at an AAALAC-accredited facility at Beth Israel Deaconess Medical Center, Boston. The mice were handled in accordance with the animal care policy of Harvard Medical School. Mice were euthanized humanely by CO2 inhalation in the end of the experiments following treatment and tumor samples were harvested for further study as described below.
We have selected AM9928 for our in vivo studies based on its prolonged target engagement (Tyukhtenko, Ma et al., 2020). We expected that the prolong inhibitory effect of AM9928 would be translatable to a longer pharmacodynamic effect when compared to other MAGL inhibitors such as AM4301. To that end, we studied the effects of AM9928 on BMEC-TJs and tumor colonization in the brain using the spontaneous breast cancer metastasis mouse model (syngeneic) of mammary tumor cells. We used GFP-4T1-BrM5 cells which migrate to the brain (Bos, Zhang et al., 2009; Rodriguez, Jiang et al., 2014; Avraham, Jiang et al., 2014) and form breast metastasis in the brains of Balb/c mice. We administered GFP-4T1-BrM5 tumor cells (5x104 cells) into the mammary fat pads of Balb/c mice. More than 70% of these mice developed mammary tumors within 3 weeks, while brain microtumors were developed in about 5 weeks. Here, following administration of GFP–4T1-BrM5 cells, mice were injected with AM9928 (10 mg/kg, i.v.) or the vehicle control twice a week for 3 weeks (10 mice/group/treatment/experiment). Tumor sizes in the mammary fat pads were measured using calipers, and the volume was calculated, using the formula: V = 0.52 × (length) × (width)2. The BBB integrity at the end of the experiments was analyzed by Evan blue test in control groups and in mice treated with AM9928 as compared to vehicle control. Briefly, Evans’s blue (EB) (Sigma Chemical Co. St. Louis, MO. USA.) dye (25% in 0.9% NaCl solution) was intravenously injected at dose of 25 mg/kg under anesthesia. One hour after the injection, animals were sacrificed. Brains were weighed, clipped, and individually placed within formamide p.a. (2mL/brain). The content of dye extracted from each brain was determined by spectrophotometer (Photometer 4010, Boehringer) at 620nm and compared to standard graph created through the recording of optical densities from serial dilutions of EB in 0.9% NaCl solution. For in vivo  imaging, spectral fluorescence images were obtained using a Maestro‐based imaging system (CRI Inc., Woburn, MA, USA) (Rodriguez, Jiang et al., 2014; Avraham, Jiang et al., 2014). Five sets of side‐by‐side whole‐brain images of animals with tumors from the GFP–4T1BrM5 cells treated with vehicle control or with AM9928 were obtained. In the end of the experiment, mice were sacrificed, and brain and mammary fat pad tissues were collected for further analysis.
Statistical analysis: All values are expressed as the mean ± SEM of the mean and are representative of two to three independent experiments as indicated. P values were calculated using unpaired two-tailed Student’s tests, built in the GraphPad Prism Software for evaluation of statistical significance.