Determining the genetic basis of the lupus immune response
Much previous research has identified several genetic risk factors for developing lupus; but has focused on studying very diverse patient populations. Some patients may have systemic lupus, while others only have cutaneous (skin) lupus or lupus nephritis (inflammation of the kidneys). While this allows scientists to identify genetic factors common across lupus types, there are likely important risk factors unique to each type. Even within a specific type of lupus, significant differences in the immune response can occur. Dr. Riquelme’s research has been investigating the differences present in systemic lupus and found that patients can be differentiated into three groups based on their immune response. However, the reasons for these differences are not clear since the genes associated with these differences are not known. In her study, Dr. Riquelme will identify these genes and determine what function they play in altering the immune response during lupus. In addition, she will observe whether these genes are associated with changes in the treatment success for each group of patients.
What this study means for people with lupus
Dr. Riquelme’s study will help determine which genes lead to the differences in immune response observed in different lupus patients. Her research may also help determine which are the best therapies for specific groups of people with lupus.
Lupus is a very heterogeneous disease and the identification of groups of patients that share features that may allow understand their pathogenesis and improve their treatment has been a matter of much study and debate. On the one hand, clinical and routine laboratory information has been used to classify patients with SLE, but these have been of no use to define groups of patients, only to group them as one disease. New molecular and bioinformatics tools are needed to sort out the heterogeneity. We recently described a new stratification of lupus based on longitudinal transcriptome data from two sets of patients, a pediatric and an adult set. The stratification is based on the identification of features through the correlation of gene expression with the activity score available for the data, in this case, the SLEDAI. In that work we clearly showed that in two clusters of lupus patients, disease activity was followed by the positive correlation of gene expression derived from neutrophils, while in the 3rd cluster, lymphocytes and their expressed genes showed a positive correlation with disease activity. The first cluster was associated with an increased development of proliferative nephritis. The clusters had clear clinical differences as well as clearly different molecular patterns, although in two of the clusters, neutrophil signatures predominated. In one of the clusters of neutrophils, interferon-induced genes correlated positively with disease activity, despite almost all patients having a basal interferon signature, showing that the mechanisms behind disease activity are different between patients. One of us (Dr Joaquín Dopazo, co-PI) has recently demonstrated the enormous potential of mechanistic models of cell functional activity for predicting different relevant disease endpoints in cancer as well as new therapeutic targets. Moreover, we have shown how machine learning (ML) algorithms can be used over omic data and mechanistic models for high precision drug repurposing. So we aim at using innovative methods of a. stratifying lupus patients, b. identify the genetic characteristics of each group of patients, and c. identify functional pathways and new drug targets. We propose the following hypothesis: That different genetic loci and different gene-environment interactions contribute to each of the disease clusters, as susceptibility loci and quantitative trait loci, both as eQTLs and/or meQTLs with a differential genetic and environmental contribution. Hence the following aims: Specific aim 1: To identify the genetic loci relevant for each cluster of lupus patients. Specific aim 2: To identify the functionality of the pathways differentiating the groups of patients. Using transcriptome data we will model the activity of the pathways and their involvement in the definition of the different patient clusters, along with the functional consequences. We will find the optimal drug (or drug combination) for each cluster.