Bulk T cell receptor repertoire sequencing of blood, stimulated PBMC, and biopsies from the tuberculin skin test in people with latent tuberculosis infection

We used the tuberculin skin test (TST) as human challenge model to study the temporal evolution of the T cell response to Mycobacterium tuberculosis (Mtb) antigens in vivo. Study participants comprised healthy HIV seronegative adults, 18-60 years of age. Latent tuberculosis (TB) infection was defined as immune memory for Mtb-specific antigens identified by positive peripheral blood IFNg release assays, but no clinical or radiological evidence of active TB. Two units tuberculin each were injected intradermally into the contralateral forearms of participants. After 2-3 days (n=17) or 7 days (n=165), 3 mm skin punch biopsies were taken from the injection sites and processed for high-throughput T cell receptor sequencing. The time points reflect maximum clinical inflammation (day 2-3) and maximum T cell infiltration (day 7) of the TST. TST samples were compared to whole blood samples (n=20). In addition, peripheral blood mononuclear cells (PBMC) from a subset of participants (n=12) were stimulated in vitro with one of 10 microgram/ml purified protein derivative of Mtb (PPD), 100 microgram/ml tetanus toxoid (TT), or control media for 6 days. Up to 5 replicate samples were collected per participant for each stimulus. PBMC samples were used to define expanded TCRs as antigen-reactive and those were subsequently quantified in paired TST samples. This submission comprises FASTQ data for alpha and beta chain sequences for 313 samples from a total of 175 individuals. In the early version of the experimental sequencing pipeline (Oakes et al. 2017, Front Immunol), alpha and beta chain samples were indexed separately and read 1 (R1) and read 2 (R2) files were merged for each sample. In the later version of the protocol (Nageswaran et al. 2022, Methods in Molecular Biology), alpha and beta chain samples were indexed with the same index pair (and are therefore contained within the same demultiplexed FASTQ files), and R1 and R2 files were provided separately. Both formats can be processed with Decombinator into separate alpha and beta chain data (https://github.com/innate2adaptive/Decombinator).