Instrument: Illumina HiSeq 2500
Strategy: RNA-Seq
Source: TRANSCRIPTOMIC
Selection: cDNA
Layout: PAIRED
Construction protocol: Thymic stromal cells were isolated and sorted according to their cell surface phenotypes. Fragmented thymi were digested repeatedly for 15-20 min at 37 °C with 1mg/ml Collagenase/Dispase (Roche Diagnostic) and 100µg/ml DNaseI (Roche Diagnostic) in HBSS containing 2% FCS (Perbio), to obtain single cell suspensions. After the final digest, cells were pooled and labelled with biotinylated anti-EpCAM for positive enrichment by AutoMACS system (Miltenyi Biotec), and stained using the following directly labelled antibodies and reagents: FITC-anti-IAb (clone AF6-120.1, BioLegend), PE-anti-Ly51 (clone 6C3, BioLegend), PECy7-anti-CD45 (clone 30-F11, BioLegend), APC-Cy7-anti-I-A/E (clone M5/114.15.2, BioLegend), biotinylated anti-EpCAM (clone G8.8, DSHB, University of Iowa), Streptavidin-labelled PerCP-Cy5.5 (BioLegend) and Cy5-UEA1 (Vector Laboratories). The cells were exposed to 4’, 6-diamidino-2-phenylindole (DAPI) to identify dead cells and then sorted by flow cytometry (FACSAira, BD Biosciences). Mature, MHCII-high mTEC were sorted into media+FCS and stored on ice for 2-3 hours before Fluidigm C1 analysis. Single cells were isolated and amplified on the Fluidigm C1 platform at the Sanger Institute/EBI Single Cell Genomics Centre. Briefly, a pool of FACS-sorted TECs (220K cells/ml) were loaded on two (denoted by the prefixes "A" and "B") medium chips (designed for capture of cells between 10-17 µM in diameter). After loading and capture, chips were inspected under the microscope and the presence of a single cell at capture sites was confirmed. Cell lysis, cDNA synthesis and subsequent PCR amplification was performed using the SMARTer Ultra Low RNA Kit (Clontech) for the C1 platform as per the manufacturer’s instructions (Clontech/Fluidigm). The cell lysis buffer (20 µL) was supplemented with 1 µL of a 1:400 dilution of the External RNA Controls Consortium (ERCC) spike-in Mix 1 (Invitrogen). Indexed Libraries from the 96 samples harvested from each chip were prepared using the Nextera XT kit (Illumina). Multiplexed libraries from each chip were sequenced over two lanes on a HiSeq 2500 (Illumina) at the Wellcome Trust Sanger Institute run in fast mode to generate 100bp paired-end reads.