Please note that we have stopped the regular imports of Gene Expression Omnibus (GEO) data into ArrayExpress. This may not be the latest version of this experiment.
E-GEOD-23663 - Gene expression profiling of human CD8+ CD161hi and CD161lo memory and naïve T cell subsets
Released on 31 July 2011, last updated on 16 August 2011
Gene expression profiling of human CD8+ CD161hi and CD161lo central and effector memory and naïve T cell subsets. The mechanisms by which IL-17 secreting cells are regulated have not been completely elucidated. We previously identified a population of rhodamine-effluxing memory CD8+ T cells with high expression of CD161 that contributes to immune reconstitution after lymphopenia-inducing chemotherapy. Here we find that CD161hi CD8+ T cells share transcriptional programming with Th17 cells, but most do not secrete IL-17 or proliferate to stimulation through the T cell receptor (TCR). Transcriptional analysis of subsets identified by expression of CD161 and CD62L revealed a novel mechanism of TCR signaling pathway regulation in CD161hi CD8+ T cells that is distinct from that described in anergic or tolerant cells and renders them functionally dependent on costimulation through innate cytokine receptors or CD28. CD161hi CD8+ T cells, induced to proliferate by a TCR signal delivered with costimulation, demonstrated plasticity that was dependent on the nature of costimulation and resulted in expansion of IL-17 secreting cells that could not proliferate to a TCR signal alone or differentiation to Tc1-like cells that proliferated to TCR stimulation in the absence of costimulation. The data show an association between TCR signaling pathway downregulation and type 17 programming in CD161hi CD8+ T cells, whose dysregulation could mediate IL-17 dependent inflammatory diseases. T cell subsets were sort-purified from healthy adults and analyzed by gene expression profiling. Isolation: Effluxing CD161hi and non-effluxing CD161lo CD8+ TCM and TEM subsets were purified using magnetic bead separation and cell sorting to achieve >98% purity, as previously described (35). CD8+ T cells were positively selected using CD8 Microbeads (Miltenyi Biotec, Germany), loaded with Rh123 (Sigma, St. Louis, MO) and cultured for 30 min to allow Rh123 efflux, then labeled with fluorochrome-conjugated antibodies to CD4, CD16, TCRγδ, Vα24, CD8, CD95, CD62L and CD161. CD161hi and CD161lo TCM and TEM subsets were sort-purified on a FACS ARIA equipped with 405 nm, 488 nm and 633 nm lasers (BD Biosciences). CD161hi TCM and TEM subsets were identified as CD62L+/Rh123lo/CD161hi and CD62L-/Rh123lo/CD161hi events, respectively, in the CD4-/CD16-/TcRγδ-/Vα24-/CD8+/CD95+ population. CD161lo TCM and TEM subsets were identified as CD62L+/Rh123hi/CD161int/neg and CD62L-/Rh123hi/CD161int/neg events, respectively, in the CD4-/CD16-/TcRγδ-/Vα24-/CD8+/CD95+ population. Gene expression profiling: RNA was extracted from sort-purified subsets from 6 healthy individuals using the RNeasy Plus Minikit (Qiagen, Valencia, CA) and biotinylated, followed by generation of amplified cRNA for array hybridization using the Illumina TotalPrep RNA Amplification Kit (Applied Biosystems, Foster City, CA). Amplified biotinylated cRNA was then purified before quality control to ensure high quality cRNA was available for hybridization. Labeled cRNA was hybridized to Illumina WG-6 Expression BeadChips v3.0 (Illumina, San Diego, CA), which quantitate expression of 48,802 transcripts derived from the NCBI RefSeq (Build 36.2, Release 22) and UniGene databases (Build 199). BeadChips were washed before reading and image extraction, and then scanned on an Illumina BeadArray Reader. The resulting TIFF images were processed using GenomeStudio Gene Expression Module (GEM) software. Data quality was assessed using the Control Summary feature in GenomeStudio GEM. For a given analysis set, a GenomeStudio Probe-level Final Report was generated by combining the Sample Probe Profile and Control Probe Profile tables. The Final Report comprising the full dataset was initially processed using the Bioconductor package lumi by employing a background correction estimate and quantile normalization. A small adjustment (i.e. 20 counts) was made to the entire dataset to make all intensity signals non-negative and these values were log2-transformed. The dataset was initially filtered using the ‘shorth’ function of the Bioconductor package genefilter, resulting in retention of 31,084 of 48,802 original probes. Each pairwise comparison was further filtered by discarding probes whose signal intensity was less than a defined signal “noise floor” across all arrays in the data subset. This was achieved by calculating the median of the ‘negative control’ probe signals for each array, averaging these values, and setting the noise floor as 3 times this average. Differential gene expression was then determined using the Bioconductor package limma, and a false discovery rate (FDR) method applied to correct for multiple testing. Significant differential gene expression was defined by a log2 ratio > 0.585 (± 1.5-fold) and FDR (adjusted p value) < 0.05.
transcription profiling by array
Cameron Turtle <firstname.lastname@example.org>, Cameron J Turtle, Jeff Delrow, Ryan Basom, Stanley R Riddell