4T1 tumor model used for batch correction testing

Control group 4T1 breast cancer tumors from female BALB/c mice from two batches were used to test the robustness of the batch correction algorithms. The 4T1 cell line was purchased from ATCC and cultured in a humidified atmosphere (37°C, 5% CO₂). The 4T1 cells were cultured in RPMI-1640 medium (Sigma-Aldrich) with 10% FBS, 2 mM L-glutamine and pen/strep. MycoAlert (Lonza, LT07-318) was continuously used to confirm that the cell line was mycoplasma free throughout the study. Female BALB/c mice (Envigo) were fed a standard chow diet. At 4-6 weeks, 2.0×10⁵ 4T1 cells were mixed 1:1 with BD Matrigel Matrix Growth Factor Reduced (BD Biosciences) and injected into the right mammary fat pad.

Upon sacrifice, tumors were harvested and dissociated using a mouse tumor dissociation kit (Miltenyi Biotec, 130-096-730). After dissociation, erythrocytes were lysed with Red Cell Lysis Buffer (Miltenyi Biotec, 130-094-183), according to the manufacturer's protocol.

The dissociated 4T1 tumor cells were stained with cisplatin for viability with RPMI1640 (10% FBS, 0.25 μM cisplatin, RT) for 5 min and fixed with 1.6% PFA (electron microscopy grade, Electron Microscopy Sciences, 15710) before freezing. After 10 min incubation, samples were centrifuged and the supernatant was aspirated before being placed in −80°C freezer until analysis.

The panel contained 18 mass-tagged antibodies with the same target as the seven experimental batches. Many of the same antibody targets were on different channels (marked as ‘sdc’ in Table 4) and one was a different clone. A similar protocol, detailed below, was used for staining and cells were run on the same Helios CyTOF machine.

Prior to staining, samples were thawed at RT and resuspended in cell washing buffer (CWB) [DPBS (Thermo Fisher Scientific, 14040-133) with 1% BSA, 0.02% NaAzide and 0.025% DNaseI (Sigma-Aldrich, DN25-1G)]. Cells were counted using the Countess automated cell counter and ∼3×10⁶ cells per sample were used for barcoding. Counted cells were washed in 1× Maxpar® Barcode Perm Buffer (perm buffer; Fluidigm, 201057) twice, before being resuspended in 1× perm buffer (195 µl). Samples were then mixed with 5 µl barcoding solution (Fluidigm, 201060), or 1 µl (500 µM) Intercalator-103Rh (Fluidigm, 201103A) for the control, and incubated for 30 min at RT. After incubation, cells were washed in CWB twice, then pooled with CWB before washing in Maxpar^® Cell Staining Buffer (CSB; Fluidigm, 201068).

Surface antibodies were diluted in CSB, while antibodies targeting intracellular proteins were diluted in CSB/perm (10% 10× perm buffer). Samples were blocked on ice for 10 min using anti-CD16/CD32 (75 µl per 3×10⁶ cells) diluted in CSB. The surface staining antibody cocktail was prepared to total 40 µl per 3 million cells. Samples were incubated with the antibody cocktail for 30 min (RT) before washing in CWB with a 10 min (RT) incubation. Samples were then washed in PBS (2 mM EDTA) and fixed with 200 µl 2% PFA per 3 million cells for 30 min (RT, in the dark). Fixed cells were then washed in CSB/perm twice and blocked with anti-CD16/CD32 in CSB/perm as described above. Samples were incubated with the intracellular staining antibody cocktail for 30 min (RT) before being washed in CWB and subsequently in 2 mM EDTA/PBS. Cellular DNA was stained by incubating the samples in Ir191/193 Intercalator (0.33 µl 500 µM Intercalator-Ir per 2×10⁶ cells, Fluidigm, 201192B) mixed with 2% PFA (1 ml per 20×10⁶ cells) overnight (4°C). The following day, samples were centrifuged and resuspended in CWB with 10 min incubation (RT). Samples were then washed in PBS (2 mM EDTA) and kept on ice until the mass cytometer was ready. Before acquisition, an aliquot of cells was washed in MilliQ water four times (400 g, 5 min, RT). The aliquot was then resuspended in 1× EQ Four Element Calibration Beads solution (Fluidigm, 201078) to a final concentration of ∼1-2×10⁶ cells/ml, and strained through a 40 µm cell strainer. The cells were then acquired on the Helios mass cytometer (Flow Cytometry Core Facility, Department of Clinical Science, University of Bergen). All centrifugation steps were performed with a swing-bucket rotor at 900 g for 5 min at RT unless otherwise specified.