Cells were cultured at 37°C, 5% CO2 in DMEM (HeLa, Hct116, HEY, T24), RPMI1640 (AGS, Ht1080, Panc1, RPE) or McCoy 5A medium (USMCC1, U2OS) supplemented with 10% fetal bovine serum (FBS) (Wisent). The AGS, USMCC1, HEY, HT1080, and HeLa cell lines were obtained from ATCC in 2018, 2016, 2013, 2010, and 2009, respectively, and were maintained between passages 4 and 15. The U2OS line was a kind gift from Laurence Pelletier Laboratory, Lunenfeld Tanenbaum Research Institute, Sinai Health System, Toronto, Canada (obtained in 2013 and maintained between passages 6 and 10). Panc1, Hct116, T24 and RPE lines were a kind gift from Daniel Schramek lab, Lunenfeld Tanenbaum Research Institute, Sinai Health System, Toronto, Canada (obtained in 2020 and maintained between passages 8 and 28). The cell lines have not been further authenticated in our laboratory. For drug treatment experiments, AGS cells were exposed to centrinone B (5690, TOCRIS Bioscience) dissolved in DMSO, at a final concentration of 500 nmol/L in co-culture medium, 24 h before the peritoneal tissue is placed in the MatTek box.
Creation of stable cell lines expressing wild-type CDH1
AGS cells, derived from primary gastric adenocarcinoma, do not express CDH1 protein due to a truncating mutation (T578fs)34. The parental AGS cell line was infected with lentivirus to stably express V5-CDH1, or V5 alone, using gateway constructs (pLEX_306m, Addgene). Lentiviruses were produced as described40 and used to infect cells for 24 h, followed by selection with puromycin (2 µg/ml) for 7 days. To confirm protein expression, cell extracts were separated by SDS-PAGE, transferred to PVDF membranes, blocked with 5% milk-PBS-0.1% Tween, probed with anti-CDH1 antibody (BD Transduction Laboratories, 610182, 1:1000) at 4°C overnight, followed by secondary antibody bound to HRP (GE Healthcare) and detected using SuperSignal West Femto Maximum Sensitivity Substrate (34095; Thermo Fisher Scientific).
Fluorescence-activated cell sorting
On day 1, 24 h after exposing AGS cells to peritoneal tissue in a MatTek dish, culture medium was collected and suspension cells were treated with RBC lysis buffer (NH4Cl (155 mM), KHCO3 ( 10mM), EDTA (0.1mM) pH 7.6). After clearance and resuspension in flow cytometry sorting buffer (Hanks Balanced Salt Solution, 25 mM HEPES pH 7.0, 2 mM EDTA, 1% Fetal Bovine Serum) with 4′,6-diamidino-2- phenylindole (DAPI, 0.2 µM) at a concentration of 5 X106 cells per ml, counting based on flow cytometry, with gating parameters that separated non-AGS cells, and live-dead assay were performed. AGS cells that remained adherent to the backwell were harvested by trypsinization and treated similarly for counting and analysis of living dead. Samples were filtered through 40 μm nylon mesh to remove large aggregates. Stained samples were immediately analyzed on a Fortessa 20 flow cytometer. Data were processed with FlowJo software.
Histological and immunohistochemical (IHC) evaluation
All protocols were approved by Sinai Health System Research Ethics Boards. Tissue samples were fixed in 10% formalin for 24 h at room temperature. Serial sections were made at three levels, 50 mm apart. Each tissue edge was stained red with tissue marking dye to orient the sample for paraffin embedding (Fig. S1a). Peritoneal tissue strips were embedded in paraffin by aligning the three strips parallel to each other with the mesothelial layer facing to one side (Fig. S1b). Sections (5 μm) were stained with hematoxylin and eosin (H&E) according to standard protocols using the Veristain Gemini Automated Slide Stainer (Thermo Scientific) at the University of Toronto Dentistry Pathology Core, and examined with a Leica DMR upright microscope. For IHC staining, slides were stained with CDX2 1:50 or D2-40 1:50 antibodies using the fully automated Dako Omnis platform (integrated deparaffinization and recovery), using the detection kit Dako OMNIS (Ref: GV800). Slides were exposed to high pH for 15 min, antibody incubation for 10 min, polymer detection for 20 min, substrate chromogen for 5 min, and finally Dako hematoxylin for 3 min. This IHC treatment was performed in the Department of Pathology at the Hospital for Sick Children in Toronto.
Cell staining for live cell imaging
AGS cells plated on MatTek dishes were stained using CellTracker™ CM-DiI (ThermoFisher, C7000). AGS cells were first plated on MatTek dishes for 18 h before staining. The medium was removed from the cells and the cells were incubated with a dye (1 μg/μl) for 30 min.
Cell morphology assessment
Images of cells grown to ~80% confluence on the 96-well plates were captured by INCell Analyzer 2000. FIJI was used to analyze the images and determine the cell axial ratio (width/length). 30×105 AGS cells expressing V5 or V5-CDH1 were imaged per well, 20 fields per well.
Migration of scratch wounds
1×106 cells were seeded in 6-well plates and starved of serum for 18 h before scraping. The scratches were made manually using a P10 pipette tip, and the migration was assayed for 24 h using the INCell6000 analyzer equipped with a motorized stage and a live cell device. (37°C heated and humidified chamber with 5% CO2) with a 6-well plate adapter. Data acquisition was carried out continuously over the durations indicated. Images were collected with a 10X objective. All hardware and image capture requirements were made possible and images analyzed using MATLAB 9.4 2018.
Image analysis was performed by measuring the total wound area in three fields per condition, at t = 0, 8, 16 and 24 h. The residual wound area was expressed as a percentage of the original wound, which was itself highly reproducible, subtracted from 100 to give the % healing at each time point, and compared between groups by the t-test.
Measurement of cell viability and proliferation
AGS cells expressing V5 or V5-CDH1 were seeded on a 96-well black polystyrene plate with a transparent flat bottom (Corning) and cultured in RPMI medium supplemented with 10% FBS at 37 °C, 5% CO2. Cell viability (1 × 105 cells spread) and proliferation (5 × 104 plated cells) tests were performed at 24 h and 48 h. Cell viability was assessed using the LIVE/DEAD® Viability/Cytotoxicity Kit (ThermoFisher), according to the manufacturer’s instructions. For cell proliferation, cells were incubated at room temperature with Hoechst (1:800) for 15 min. For both assays, cells were then visualized using the INCell6000 analyzer with a 96-well plate adapter. Images were collected with a 20X objective. Absolute cell numbers were counted using the Columbus Image Data Storage and Analysis System (PerkinElmer). Viability was measured as the number of live cells divided by the total number of cells, both dead and alive, in a well. Proliferation was assessed by measuring the number of cells per well.
Statistics and reproducibility
Statistical significance and p the values were evaluated by analysis of variance with Bonferroni and Student correction you trials. Error bars reflect SEM. The number of replicates and sample size are indicated in the respective figure legend for each figure, with the n value corresponding to independent experiments. All analyzes were performed using R statistical software (v4.1.2; R Core Team 2021)44.