Meta-Analysis from Gene Expression Profiles of Lupus Affected Tissues Reveals Novel Immune Cell Contributions

Abstract

Abstract Immunologic mechanisms causing tissue damage in autoimmune diseases such as SLE are not fully understood. To gain additional insight, gene expression profiles obtained from lupus affected skin, synovium and kidney were obtained and compared to meta-analyzed data obtained from active lupus B, T and myeloid cells. More than 300 arrays from lupus patients and appropriate controls were analyzed to determine differentially expressed (DE) genes (8279 discoid lupus skin, 5465 synovium, 6381 glomerulus, 5587 tubulointerstitum). Notably, the majority of lupus affected tissue DE genes were detected in more than one tissue and 439 were differentially expressed in all tissues. A variety of approaches assessed the molecular pathways and cellular phenotypes accounting for these common lupus affected tissue DE genes. Curated STRING-based interaction analysis identified a number of pathways including co-stimulation of T cells, activation of B and myeloid cells, antigen presentation, TLR signaling and p38 activation; from a total of 193 IPA-documented pathways, 59 were common to all tissues including p38 signaling, TLR signaling, maturation of dendritic cells, B cell activation and ICOS-ICOSL in T cells. Novel bioinformatics approaches documented that more than 50% of the DE genes in the tissues were associated with immune cell function and 11–18% were unique to the immune system. Further analysis using LINCS and additional novel software highlighted specific molecular pathways of immune cell activation including the JAK/STAT and the IL12 pathways. These results demonstrate the value of comprehensive application of orthogonal curated bioinformatics tools in identifying the role of immune cells in lupus pathogenesis and tissue damage.