Document resource
Objectives Data regaring the involvement of myeloid-derived suppressor cells (MDSC) in systemic lupus erythematosus (SLE) remain limited and conflicting. We hypothesized that SLE MDSC are dysfunctional—losing suppressive capacity or becoming pro-inflammatory—driven by distinct immunometabolic programs. We aimed to quantify circulating MDSC and subsets, relate them to disease activity, test their impact on T-cell proliferation, assess their metabolic profle, and compare single-cell transcriptomes with healthy donors.Methods We conducted a prospective single-center study of adults meeting ACR/EULAR 2019 and/or SLICC criteria (SLEDAI-2K > 1) at baseline (M0), 3 months (M3), and 6 months (M6). PBMC were analyzed by 27-parameter flow cytometry. MDSC were gated as CD19–CD3–HLA-DR–CD11bhiCD33+ and assigned to monocytic (M-), granulocytic (G-), or early (E-) subsets according to the expression of CD14 and CD15. MDSC were purified by magnetic sorting (HLA-DR and lineage depletion, CD11b selection) and co-cultured with anti-CD3/CD28-activated T cells. An intracellular panel quantified metabolic proteins (HK1, GLUT1, CPT1A, PRDX2, ASS1, ACAC, ATP5A). A public single-cell RNA-seq database (162 SLE/99 healthy donors) was re-analyzed focusing on myeloid HLA-DR low cells. This study was approved by local ethics committee, and registered on ClinicalTrials.com ( NCT05424627).Results We enrolled 21 SLE patients (19 women; median age 39) and 8 controls. Total MDSC were increased in SLE (median 0.87% vs 0.27% of viable PBMC; p=0.005), mainly due to M-MDSC (0.08% vs 0.02%; p=0.03). Longitudinally, MDSC rose at M3 and M6, largely driven by G-MDSC expansion. After Benjamini–Hochberg correction, CD38 intensity on G-MDSC correlated with articular involvement and inflammatory arthralgia ( rho =0.79; q<0.05). Purified SLE MDSC generally failed to suppress—and sometimes enhanced—T-cell proliferation; suppression appeared in one patient at M3 after cyclophosphamide. Compared to functionally suppressive MDSC obtains from other settings (peripheral stem cell graft, in vitro induced MDSC), SLE MDSC showed an increase in glycolytic pathway (HK1) and decrease in antioxidant programs (PRDX2). scRNA-seq identified reduced IL10/CCL2/CCL20 and increased RRAD/CRYM/APOC1.Conclusions In active SLE, circulating MDSC expand yet are largely non-suppressive/pro-inflammatory, with an unbalance between glycolytic and antioxidant programs. They also display a distinct transcriptomic signature, consistent with metabolic rewiring and impaired anti-inflammatory programs. Restoring glycolysis–antioxidant balance may offer therapeutic benefit.