Background/Purpose: Systemic Lupus Erythematosus (SLE) is a clinically heterogeneous disease with few treatment options. Current treatments are not fully effective and show highly variable responses. Drug repurposing is based on the comparison of gene expression signatures of blood, PBMCs or separated cell types from cases with a disease with gene expression signatures made with various cancer cell lines treated with different drugs. It has been an effective technique for the identification of new therapeutic approaches. We present a systematic drug repurposing analysis based on gene expression signatures derived from blood cells of SLE patients to discover potential new drug candidates and target genes.

Methods: We collected and processed a compendium of gene expression data sets of SLE from the NCBI Gene Expression Omnibus database. We selected data sets of adult and juvenile SLE cases from different micro-array platforms to obtain a heterogeneous group of data from which to identify a common signature. We used R to process each data set independently, making normalization, transformation to logarithmic scale, quality control, and differential expression analysis. The genetic signatures obtained were queried independently on the Lincscloud database, which contains tens of thousands of drug and genetic perturbation-caused profiles, and obtained a list of drugs and genes based on similarity scores for each SLE signature. A positive similarity score represented a genetic pattern similar to the SLE signature, and a negative similarity score represented an inverse genetic profile. A negative score was therefore considered to revert the genetic profile of SLE. The median of the similarity score of each independent experiment was then calculated to obtain a unique list of genes and drugs. Drug targets were annotated using a file created in R from three drug databases to classify the drugs into groups. The Enrichr web tool was used to obtain the pathways involved when analyzing the knock-in and knock-down experiments.

Results: We obtained drugs never or little studied to treat SLE and drugs undergoing intensive research. Results were not dependent of cell type or SLE heterogeneity, as we selected genes conserved across all SLE signatures queried. Phospoinositol 3 kinase (PI3K) and mammalian targets of Rapamycin (mTOR) inhibitors were the most significant groups of drugs obtained. When we analyzed the biological pathways obtained with the knock-in and knock-down experiments we found pathways impaired in SLE, such as the interferon and immune signaling pathway, the translational process, but also, pathways related to the PI3K signaling pathway or the Insulin signaling pathway. So, these results are complementary and consistent with those observed with the drugs. PI3K acts in important processes related to the development of SLE, such as immune signaling, the lymphocyte differentiation and proliferation or apoptotic processes.

Conclusion: Our results suggest that PI3K inhibitors affecting biological pathways impaired in SLE are the best potential therapeutic option. This work has received support from the EU/EFPIA Innovative Medicines Initiative Joint Undertaking (PRECISESADS, grant n. 115565).

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/in-silico-drug-repurposing-analysis-supports-phosphoinositol-3-kinase-inhibitors-as-treatment-of-lupus/