Effect of culture/microexpansion in various conditions on virus-specific CD8+ T cells

To investigate the conditions for the optimal expansion ot T cell therapeutic products, we used a microexpansion platform to screen various cytokine cocktails (no cytokines, IL-4 + IL-7, IL-15, IL-15 + IL-4, IL-15 + IL-6, and IL-15 + IL-7). We then performed gene expression profiling analysis using data obtained from RNA-seq of 4 different donors under 6 different culture/microexpansion conditions Gene expression profiling analysis (fpkm) of RNA-seq data for CD8+ T cells expanded under six culture conditions: (1) no cytokines, (2) IL-4 + IL-7, (3) IL-15, (4) IL-15 + IL-4, (5) IL-15 + IL-6, and (6) IL-15 + IL-7. BMCs from 4 donors were thawed and incubated with media containing overlapping peptide libraries containing SARS-CoV-2 structural proteins for spike, membrane, envelope, and nucleocapsid (A&A Peptide, San Diego, CA, USA). Pulsed PBMCs were plated in 96-well plates at 200,000 cells/well in media with one of six cytokine conditions, specifically IL-4 (400 IU/mL) + IL-7 (10 ng/mL), IL-7 (10 ng/mL) + IL-15 (5 ng/mL), IL-4 (400 IU/mL) + IL-15 (5 ng/mL), IL-6 (100 ng/mL) + IL-15 (5 ng/mL), IL-15 alone (5 ng/mL), or no cytokines (R&D Systems, Minneapolis, MN, USA). Cells were split on day 7 of culture and supplemented with additional media and cytokines. On day 10, expanded VSTs from 4 donors were re-stimulated with SARS-CoV-2 pepmixes encompassing spike, nucleocapsid, membrane, and envelope with CD28/CD89d (BD Biosciences) and incubated for 4 h. Cell viability was assessed using Live-Dead-near-infrared (IR). Cells were surface stained with fluorophore-conjugated antibodies against CD8-BV605, CD8-BV421, CD19-FITC, CD16-FITC, and CD14-FITC (Biolegend) and then sorted for CD8+ and CD8+ cells. Cells were not stained with CD3+ antibodies to prevent non-specific cell activation. Cells were then frozen and preserved in Trizol (Invitrogen, Waltham, MA, USA). Total RNA was extracted from 10,000–50,000 cells, and mRNA sequencing libraries were generated (NEBNext), Novaseq6000 was used for 50-cycle single-end read sequencings, which were processed with bcl2fastq 2.17.1 to generate FastQ files. Reads were mapped to the human transcriptome hg19 using STAR 2.7.0; gene expression values (fragments per kilobase exon per million mapped reads [fpkm]) were calculated and normalized with RSEM 1.3.0.