George Sentis, Chrysoula Kalantzi, Sofia Flouda, Aggelos Banos, Panagiotis Garantziotis, Theodora Manolakou, Katerina Chatzidionysiou, Noemin Kapsala, Antigone Pieta, Katerina Chavatza, Konstantinos Triantafyllias, Anastasia Filia, George Bertsias, Ioannis Parodis, Dimitrios T Boumpas, Dionysis Nikolopoulos
Metabolic -rather than inflammatory-pathways are dominant in LA and may account in part for suboptimal responses to existing therapies. LA encompasses molecular subtypes, characterized by distinct pathophysiologic aberrancies potentially reversible by unique compounds.
BACKGROUND: Lupus arthritis (LA) is a major determinant of clinical outcomes and treat-to-target strategies in systemic lupus erythematosus (SLE), yet its molecular basis remains elusive.
METHODS: Whole-blood RNA sequencing from 170 SLE [67 with active arthritis but quiescent disease in other organs; 42 with only mucocutaneous activity (mcSLE); 61 clinically inactive (iSLE)], 19 active rheumatoid arthritis (RA), and 72 healthy donors. Pathway enrichment, unsupervised WGCNA, deconvolution, k-means clustering, and druggability analyses were applied.
RESULTS: Compared to mcSLE, LA patients demonstrated upregulation of ribosome/metabolism, while inflammatory pathways were significantly upregulated in mcSLE. Comparison between LA and RA revealed 769 differentially expressed genes, indicating upregulation of ribosome/metabolism and oxidative phosphorylation (OXPHOS) pathways in LA. Activated NK cells were significantly increased in LA compared to RA and mcSLE. Within SLE patients, interferon pathways were less upregulated in LA. Unsupervised analysis showed LA to be strongly linked to upregulation of "OXPHOS/metabolism" module and downregulation of the "Interferon" module. In contrast, mcSLE and iSLE were associated with "Interferon" and "T-cell" modules, respectively. RA was strongly associated with the "Innate immunity" module. LA patients were categorized into three distinct molecular clusters dominated by OXPHOS/metabolism pathways ("metabolism" cluster; n = 14), inflammatory pathways ("inflammation" cluster; n = 29), and both metabolic and inflammatory pathways ("mixed" cluster; n = 24). Druggability analysis revealed differential anticipated benefit from different compounds for each LA-cluster.
CONCLUSIONS: Metabolic -rather than inflammatory-pathways are dominant in LA and may account in part for suboptimal responses to existing therapies. LA encompasses molecular subtypes, characterized by distinct pathophysiologic aberrancies potentially reversible by unique compounds.