SRA

Purpose: To identify potential underlying molecular programs associated with the observed dysfunctional TCRTg101 phenotype, RNA sequencing (seq) was performed on TCRTg101 isolated from livers of leukemia-bearing mice at several time points following their adoptive transfer (days 6-7, 13-14, and 167-18), andas well as on TCRTg101 isolated from control (leukemia-free) analysis Methods: TCRTg101 in livers and spleens of leukemia-bearing mice were isolated at various time points following adoptive transfer (day 0, day 6-7, day 13-14, day 16-18) by FACS and were re-suspended in Trizol (Life Technologies). TCRTg101 RNA was isolated via chloroform extraction. Low input RNA sequencing was performed in the University of Chicago Genomic. Genes with fewer than 10 reads in at least 6 samples were filtered out, resulting in a dataset of 11,164 genes. Differential gene expression analysis was performed on raw aligned read counts using DESeq2 (Love et al., 2014), with batch effects accounted for in the design formula. Genes were considered to be differentially expressed if they had an adjusted p-value < 0.05 using a Benjamini-Hochberg test (FDR). Counts per gene were regular log (rlog)-transformed, and batch effects were removed using the removeBatchEffect function from limma (Ritchie et al., 2015) for principal component analysis (PCA), sample clustering based on Euclidean distance, and heatmaps depicting z-scores of gene expression. rlog-transformed, batch-corrected counts were also used for weighted gene correlation network analysis (WGCNA) (Langfelder and Horvath, 2008) while additionally filtering out the 50% of genes with the lowest variance to reduce noise, resulting in a set of 5,582 genes. Adjacency was determined using a signed analysis and a soft thresholding power of 14, which was determined by scale-free fit index (Zhang and Horvath, 2005). 12 clusters of genes were initially identified, 2 of which contained a combined 4,269 genes (76.5% of the dataset), and roughly corresponded to genes whose expression increased or decreased over the experimental time course. Of the other 10 clusters, 2 were identified as containing genes which were transiently upregulated or downregulated. These two clusters contained 448 genes, combined. The remaining 865 genes were assigned to eight different clusters, which appeared to be the result of high variance within sample groups, either due to low overall gene expression or low outlier values. No conclusions were drawn regarding these clusters due to uncertainty in the data. Core Facility on the Illumina HiSeq 2500 platform in two batches. Reads were mapped onto the University of California Santa Cruz mouse genome using kallisto (Bray et al., 2016) Results: In total, 4,075 genes were found to be significantly differentially expressed across all pairwise comparisons Conclusion:Dysfunctional TCRTg101 acquire a transcriptional program canonically associated with T cell exhaustion Overall design: RNA-seq