Epcam apc

Trypsinized (Biological Industries, Cat: BI03-052-1B) cells were collected and washed with ice-cold PBS once. Next, cells were fixed with 4% Paraformaldehyde (Sigma, Cat:158127) in PBS for 20 minutes at room temperature and centrifuged at 1500 rpm for 5 minutes; following by resuspension in stain buffer (BD, Cat: 554656) in a concentration scale of 1 × 106 cells/ml. Following antibodies were used as described; EpCAM-APC (BD, 347200) (1:100 v/v), CD133-PE (BioLegend, 372804) (1:100 v/v), CD44-FITC (Miltenyi Biotec, Cat: 130-095-195) (1:10 v/v) and CD90-FITC (Miltenyi Biotec, Cat: 130-095-403) (1:10 v/v). For unstained controls, IgG1 Isotypes; IgG1-FITC (Immunostep, Cat: ICIGG1F-100), IgG1-PE (Immunostep, Cat: ICIGG1PE-50), and IgG1-APC (BD, Cat: 555751); were used as 1:20 (v/v). For staining, cells were incubated with antibodies for 30 minutes in room temperature at dark and washed once with staining buffer. Stained cells were analyzed on NovoCyte Flow Cytometer System (Acea) and analysis was performed via NovoExpress Software (Supplementary Fig. S1).

Patient-derived mammary tissues (n = 3) were grown in hydrogels for 12 days in the presence or absence of DDR1 inhibitor (2 μM). Cells were isolated from hydrogels using collagenase followed by trypsin, as described above for colony and mammosphere assays. Cells were incubated for 90 min on ice with the following antibodies: EpCAM-PE (347198, BD Biosciences, clone EBA-1, 1:5), EpCAM-APC (347200, BD Biosciences, clone EBA-1, 1:5), CD49f-FITC (555735, BD Biosciences, clone G0H3, 1:5). Subsequent analyses on luminal and basal cell populations were performed using the following antibodies: JAG1-PE (94449, Cell Signaling Technology, clone D4Y1R 1:50) or DDR1-PE (ab253251, Abcam, clone 51D6, 1:20). DDR1+ and Jagged-1+ population gating was determined based on fluorescence minus one (FMO) controls. Flow cytometry was performed on BD LSR II flow cytometer at the Tufts Flow Cytometry Core. Flow cytometry data were collected using FACS Diva, version 6.2. Flow cytometry data analysis was performed on FlowJo software v10.7. All analyses were performed in compliance with MiFlowCyt standards^{34 (link)}. The complete report is available upon request. A figure exemplifying the gating strategy for flow cytometry experiments in this study is provided in the Supplementary Information (Supplementary Fig. 6).