Objective Systemic Lupus Erythematosus (SLE) is a complex autoimmune disease with a diverse pathophysiology triggered by the loss of self-tolerance, resulting in inflammation and tissue damage in multiple organs. The use of animal models has been instrumental in advancing our understanding of SLE. The elimination of the requirement for animal models in the FDA’s drug approval process prompts a reevaluation of their role. We aim to define a molecular framework for the future design of preclinical studies in SLE, helping researchers in the mouse model and time point selection based on the molecular pathway to be studied in relationship with the human disease.

Methods To address this, a 4 timepoint longitudinal study was conducted using four different spontaneous SLE mouse models: MRL/lpr, NZB/W, BXSB.Yaa, and TLR7.Tg6. RNA-Seq from whole-blood, spleen, and kidney, flow-cytometry from spleen, cytokines and autoantibodies in serum were profiled. All this data was characterized using longitudinal differential expression analysis and time-point cross-sectional analysis. Molecular and flow-cytometry data from blood from PRECISESADS SLE patients was integrated by means of the MEFISTO software with mouse data, allowing us to integrate direct relationships between human and mouse along the progression of disease.

Results In the spleen, lymphopenia mimicking the human and senescent-related B cells were in common between mouse models, while neutrophil and cytotoxic lymphocyte expansions differentiated the models. TLR7-involved models showed important transcriptional dysregulation in the spleen as compared with non-TLR7 models, whose dysregulations were primarily in the kidney. Molecular changes were associated with severity across all the models, with splenic changes preceding kidney damage and suggesting the ideal moment to study early immune pathogenic processes.

Human and mouse molecular integration showed a shared interferon-mediated pathway with MRL/lpr and TLR7.Tg6 models associated with hematological and constitutional clinical domains, but at different timepoints. Neutrophil-mediated inflammatory pathway was shared between humans, MRL/lpr, BXSB.Yaa and TLR7.Tg6 models and related to renal and cutaneous clinical human domains.

Conclusion The findings provide a valuable framework for future experimental design involving SLE mouse models, demonstrating their continued usefulness in investigating the elusive pathogenesis and heterogeneity of SLE.