Single-cell data derived from metastatic murine liver tissues from the B16-spleen-liver melanoma liver metastasis model (three biological replicates (n1, n2, n3)).

Description

Methods

Single-cell analysis applying single-nuclei RNA-sequencing (snRNA-seq)

Snap frozen tissue samples were dissociated for generating single nuclei gene expression libraries following customized protocols from 10X Genomics. Nuclei isolation was performed using the Chromium Nuclei Isolation Kit with RNase Inhibitor (10xGenomics, PN-1000494) according to the user guide CG000505, Chromium Nuclei Isolation Kit, UG, RevA. In brief, 200 μl of lysis buffer was added to the sample and dissociated with pestles provided by the vendor. After adding additional 300 μl of lysis buffer, the tissue was homogenized by pipette-mixing and incubated on ice for 11 10min. Afterward, 500 μl of the suspension was loaded onto Nuclei Isolation Columns 12 and centrifuged for 20s at 16,000 rcf, 4°C, followed by vortexing the flow-through for 13 10s and centrifugation for 3min at 500 rcf, 4°C in a swinging bucket rotor. Supernatant 14 was removed and the pellet was gently resuspended in 500 μl of Debris Removal 15 Buffer before centrifugation for 10min at 700 rcf, 4°C. After removal of the supernatant, 16 the pellet was washed twice with 1ml of Wash Buffer followed by centrifugation for 17 5min at 500 rcf, 4°C. For multiplexing a customized approach based on barcoding with 18 cholesterol-modified oligos (CMOs)56 was applied to pool 6 samples per single-nuclei 19 sequencing reaction. For this, nuclei were subsequently incubated after the first wash 20 with 0.2μM unique CMO Anchor-Barcode and 0.2μM Co-Anchor and thereafter 21 washed twice with 1ml of Wash Buffer by centrifugation for 5min at 500 rcf, 4°C. The 22 final nuclei pellet was taken up in 50μl Resuspension Buffer (10x Genomics). Nuclei 23 were counted using Acridine Orange/Propidium Iodide Stain (Logos, F23001) and an 24 automated dual fluorescence cell counter (Logos, LUNA-FL™). Equal numbers of 25 nuclei were pooled per multiplexed reaction resulting in a total of 20,000 nuclei per individual 10x Genomics Chromium Next GEM Single Cell 3' Kit v3.1 (PN- 1000268) 1 reaction. The pooled nuclei were loaded on a Chromium Next GEM Chip G (PN-2 1000120) and run on the Chromium iX instrument as instructed by the manufacturer. 3 

Single-cell RNA-seq libraries were generated according to the manufacturer’s 4 instructions aiming for a maximum cell recovery of 10,000 cells. Library concentrations 5 were quantified with the Qubit 2.0 Fluorometric Quantitation system (Life 6 Technologies, Carlsbad, CA, USA) and the size distribution was assessed using 7 the 2100 Bioanalyzer instrument (Agilent, Santa Clara, CA, USA). Libraries were 8 sequenced by the Biomedical Sequencing Facility at the CeMM Research Center for 9 Molecular Medicine of the Austrian Academy of Sciences on a NovaSeq 6000 10 instrument (Illumina, San Diego, CA, USA). 11 

After quality control, the raw sequencing reads were aligned to the mouse reference 12 genome GRCm38, by application of CellRanger (10X genomics) in order to obtain 13 feature-barcode matrices. Raw count matrices were analyzed using R package Seurat 14 v. 5.1.0 Cells with greater than 10% mitochondrial RNA content, less than 200 genes 15 detected or greater than 5000 genes detected were excluded from analysis. Individual 16 samples were normalized with the standard workflow (NormalizeData, 17 FindVariableFeatures, ScaleData). Following the calculation of principal components, 18 samples were integrated with Harmony. Clustering was performed applying the 19 Louvain algorithm (n = 30 PCs, resolution = 1.0). Differentially expressed genes 20 between all clusters were identified using the function "FindAllMarkers" with a LogFC 21 threshold > 0.25. Annotation of cell types and clusters was performed using canonical 22 marker genes, Enrichr-based interpretation of marker gene lists and by comparing the 23 gene signatures with the well-annotated metastatic melanoma cell clusters in mice and 24 humans defined by the Marine group57,58. Significant differentially expressed genes 25 between the individual groups were calculated for annotated cell-subclusters or combined subclusters of a celltype, respectively, using the “FindMarkers” function of 1 Seurat. Differentially expressed genes with an adjusted p-value < 0.1 were regarded 2 as statistically significant.