Decoding molecular heterogeneity in lupus nephritis: integrating single-cell and spatial multi-omics for endotype-driven precision therapeutics

Lupus nephritis (LN) develops in 40-60% of patients with systemic lupus erythematosus (SLE) and remains the leading cause of renal morbidity and progression to end-stage renal disease (ESRD). While the upstream immunological events initiating LN-including loss of self-tolerance, pathogenic autoantibody production, glomerular immune complex deposition, and type I interferon (IFN) activation-are broadly shared, accumulating evidence demonstrates that these convergent triggers give rise to molecularly heterogeneous tissue injury programs at the renal level. This fundamental divergence provides a mechanistic explanation for the inconsistent therapeutic responses observed with standardized immunosuppressive protocols. Advances in single-cell transcriptomics, spatial gene expression mapping, CITE-seq, and cross-platform multi-omic integration have deconvolved this heterogeneity into at least five recognizable tissue-level injury programs: (i) an IFN-γ-polarized macrophage-tubular epithelial cell co-activation circuit, (ii) a Tfh-B cell ectopic germinal center pathway, (iii) a neutrophil extracellular trap (NET)-driven inflammatory cascade, (iv) a podocyte-autonomous injury program, and (v) a TREM2+ macrophage-orchestrated fibrotic program. Concurrently, multiplexed urinary biomarker platforms are under development as non-invasive surrogates for tissue-level pathogenic activity. This review synthesizes current understanding of disease mechanisms, molecular profiling evidence, biomarker candidates, and mechanism-directed therapies corresponding to each tissue injury program. We map existing and investigational agents-including anifrolumab, obinutuzumab, voclosporin, CAR-T cell therapy, and complement inhibitors-to their mechanistic targets and delineate the design requirements for prospective biomarker-stratified trials essential to validating molecularly guided treatment selection.