Collection of the supplementary materials for the paper "Fibroblast-derived thrombospondin-1 drives macrophage polarization in advanced human co-culture models"

The repository contains anonymized processed results of the high-throughput experiments from the paper "Fibroblast-derived thrombospondin-1 drives macrophage polarization in advanced human co-culture models" Bulk RNA-Seq – Transwell & Direct Contact Human healthy donor-derived macrophages were co-cultured in either a transwell insert in a 3D matrix, or in direct contact with patient-derived colorectal cancer (CRC) tissue-matched cancer-associated fibroblasts (CAF) and adjacent normal-derived fibroblasts (NF). After 72 hours of co-culture, cells were physically separated by removing the transwell insert (transwell system) or by CD11b enrichment (direct-contact system). RNA was isolated using the Qiagen RNeasy Kit and submitted for sequencing. Macrophages from the transwell system were derived from 3 different healthy donors. Macrophages in the direct co-culture system were derived from the same donor; however, the fibroblasts were sourced from different patient NF/CAF pairs. Provided data: Bulk RNA-Seq/Direct contact/direct-kallisto-counts.tsv – count matrix calculated from the kallisto (Bray et al., 2016) output by summing up and rounding estimated counts per transcript to the gene level (direct-contact system); Bulk RNA-Seq/Direct contact/direct-kallisto-tmps.tsv – TPM matrix calculated from the kallisto (Bray et al., 2016) output by summing up TPMs per transcript to the gene level (direct-contact system); Bulk RNA-Seq/Direct contact/direct-metadata.tsv – metadata with samples description for the direct-contact system; Bulk RNA-Seq/Transwell/transwell-kallisto-counts.tsv – count matrix calculated from the kallisto (Bray et al., 2016) output by summing up and rounding estimated counts per transcript to the gene level (transwell system); Bulk RNA-Seq/Transwell/transwell-kallisto-tmps.tsv – TPM matrix calculated from the kallisto (Bray et al., 2016) output by summing up TPMs per transcript to the gene level (transwell system); Bulk RNA-Seq/Transwell/transwell-metadata.tsv – metadata with samples description for the transwell system. scMultiome Tissue-matched patient-derived organoids (PDOs) and tumoroids (PDTs) were co-cultured with corresponding cancer-associated fibroblasts (CAFs) or normal fibroblasts (NFs) and healthy donor-derived macrophages (refer to the advanced human co-culture model) in a collagen: Matrigel 3D matrix. After 3 days of co-culture, cells were extracted from the 3D matrix using collagenase/TrypLE E digestion, the pellet was snap-frozen, and the nuclei were isolated using the Chromium Nuclei Isolation Kit according to the manufacturer's protocol. Viably frozen corresponding CRC tissue was processed using the same kit. Then, multiplexing using cholesterol-modified oligos (CMOs) was applied to pool up to 4 samples per single-nuclei sequencing reaction. CMO 1 (healthy tissue) and CMO 2 (tumor tissue) didn't pass QC, and only CMO 3 (healthy triple-culture) and CMO 4 (tumor triple-culture) were included in the final analysis. Provided data: scMultiome/sample_cmo_mapping.csv – metadata table with multiplexed samples description; scMultiome/cell-hashing-assignment.csv – table with assignment probabilities of each individual cell to different samples; scMultiome/atac_fragments.tsv.gz – fragment file from Cell Ranger ATAC containing all 4 multiplexed samples (doesn't pass the QC and therefore isn't used for the analysis in the paper); scMultiome/rna_filtered_feature_bc_matrix.h5 – .h5-container with filtered RNA UMI count matrix from Cell Ranger containing all 4 multiplexed samples (there is no cells passed QC and assigned to CMO 1). scMultiome/TC_combined.h5ad – .h5ad-file with processed RNA from triple-culture samples (CMO 3 and CMO 4); scMultiome/TC_combined_macrophages.h5ad – .h5ad-file with processed RNA from triple-culture samples (CMO 3 and CMO 4), containing only macrophages. Bulk RNA-Seq – Recombinant Thrombospondin-1 Study (rTSP1) Primary M-CSF-differentiated macrophages from six healthy donors (3 males, 3 females) were treated with either 10 ng/ml or 1 µg/ml recombinant thrombospondin-1 (rTSP1) for three days. RNA was isolated using the Qiagen RNeasy Kit and submitted for sequencing. Provided data: Bulk RNA-Seq/rTSP1/rTSP1-hisat2-counts.tsv – count matrix calculated based on the HISAT2 (Kim et al., 2019) alignment; Bulk RNA-Seq/rTSP1/rTSP1-hisat2-fpkm.tsv – FPKM matrix calculated based on the HISAT2 (Kim et al., 2019) alignment; Bulk RNA-Seq/rTSP1/rTSP1-metadata.tsv – metadata table with samples description. Lipidomics Primary M-CSF-differentiated macrophages from four healthy donors (2 males, 2 females) were treated with either 10 ng/ml or 1 µg/ml recombinant thrombospondin-1 (rTSP1) for three days. Pellets were snap-frozen, and lipids were extracted according to the method of Matyash et al. Donors 1 and 2 are males; donors 3 and 4 are females. Provided data: Lipidomics/lipidomics-negmode-normalized.csv – the file contains untargeted LC-MS lipidomics data acquired in negative ionization mode, with lipid species annotated in MS-DIAL and quantified by integrated peak area in Skyline. Each row corresponds to a single lipid species in a single sample, with raw area (Area), normalized area (Area_norm), and log₂-transformed normalized area (Area_log). The experimental design comprises four donors (D1–D4) of primary human cells across three conditions (control, treated at dose 1, treated at dose 10), encoded in Group...18. The Area_log values serve as input for autoscaling and PLS-DA with VIP-based feature ranking.