Whole transcriptome single-cell RNA-sequencing

B16 melanoma cells were injected s.c. on the posterior flank of B6 mice (see the section “Animals and cell culture”) and tumors were treated with L-PM and as control with NL-PM (see the section “Melanoma growth inhibition with MDL-loaded polypept(o)ide micelles”). At day 15 single-cell suspensions from NL-PM and L-PM treated tumors (n = 3) were generated by tissue digestion using murine tumor dissociation Kit and gentleMACS^TM dissociator (Miltenyi Biotec). Tumor-infiltrating CD45⁺ immune cells were enriched by murine CD45 (TIL) MicroBeads (Miltenyi Biotec) using MACS technology. Cells were washed with FACS buffer (PBS + 0.5% BSA + 5 mM EDTA) and stained with fixable viability dye (eBioscience) and CD45 antibody (eBioscience) for 30 min at 4 °C. To enrich viable CD45⁺ cells were sorted using an Aria III flow cytometer (BD Bioscience). The purity of sorted cells was >95%.

Single cells were counted and captured using the BD Rhapsody Single-Cell Analysis System following the manufacturer’s guidelines (BD Biosciences). Whole transcriptome analysis (WTA) library prep and Sample Tag library prep were generated following the BD Rhapsody System mRNA WTA and Sample Tag Library Preparation Protocol (BD Biosciences). Samples were sequenced using the Illumina NovaSeq 6000 Sequencing System (Novogene, Cambridge, UK) and a 150PE strategy. The resulting raw data were preprocessed according to the Illumina standard protocol and transcript alignment, counting, and demultiplexing were performed using the BD Rhapsody WTA analysis pipeline yielding 4388 called putative cells overall with a mean of 160,000 aligned reads per cell. Pipeline output RSEC Mols per Cell files were imported in Partek Flow software (Version 9.0) and cells that contained <500 detected genes and a percentage of mitochondrial counts higher than 15 were removed from subsequent analysis. Single-cell counts were normalized by Partek Flows recommended normalization order (CPM (counts_per_million), Add:1 and log2). For dimensionality reduction PCA was used, followed by graph-based clustering (using the SmartLocalMoving (SLM) clustering algorithm) and t-SNE algorithm for visualization. Populations shown in 2D t-SNE projections were annotated according to ImmGen’s (www.immgen.org) data browser. To compare an equal number of total cells in NL-PM- and L-PM- treated samples, cell counts were randomly downsampled (leading to 2820 cells in total, 1410 cells per each group NL-PM and L-PM). To study anti-inflammatory signature (gene set comprises Ptgs2, Vegfa, Egfr, Arg1, Ccl22, Ccl17, Il10, Il12a) in macrophages and neutrophils, AUCell analysis was performed in Partek Flow (Aibar et al. (2016) AUCell: Analysis of ‘gene set’ activity in single-cell RNA-seq data). To define exhausted T cells, expression of Pdcd1(PD-1), Havcr2 (TIM3), Ctla4, Tigit, and Lag3 was analyzed in both groups) and visualized by a heatmap after hierarchical clustering (cluster distance metric = average linkage; point distance metric = Euclidean) and a 2D scatter plot.