Method: BM aspirates were obtained from SLE patients and age matched healthy controls (HC) (n=10). B cell, T cell and dendritic cell (DC) subsets from BM and paired peripheral blood (PBL) were analyzed by multiparameter flow cytometry for defined subset and effector molecules (B cell markers: CD19, CD27, IgD, MTG, CD24, CD38, IgM, CD10, CD23, 9G4, CXCR3, B220, CD95, CD21, CD24; T cell: CD3, CD4, CD45RA (memory), CCR7 (central memory), CXCR5 and ICOS (T follicular helper), CXCR3 (Th1), CCR4 (Th2), CD25 and FoxP3 (Treg), HLADR4; DC: CD11c, CD123, CCR7, CD40, CXCR3, HLADR, CCR5. PBMCs were stimulated with CpG 2216 or CpG 2006 for 2h at 37C, followed by protein transport inhibition (Brefeldin A) for 2h and staining for surface Lin1, CD304, CD123 and intracellular IFNa.
Result: CpG induced type I IFN production in pDCs in a dose dependent fashion. A higher proportion of BM pDCs produced IFNa compared to paired PB (%pDCs + for IFN: 5.2+1.5 vs. 0.9+0.5, 5ug/mL CpG 2006). Notably, significantly more pDCs from lupus BM produced IFNa compared to normal BM (71.9% vs. 5.2%). This was despite the fact that DC fractions in SLE BM and PB were lower than age matched controls (mDC in BM: 7.3+1.6 vs. 15.2+2.4%, p=0.03; pDC in BM: 0.50+0.11 vs. 0.95+0.28%, p=0.19). Phenotypic differences were observed in SLE pDCs with higher CXCR3 expression (p=0.05). The differences between T cell subsets in PB vs. BM in both SLE and HC were striking with significantly lower fractions of CD4+ T cells (e.g. 30.6+4.9 vs. 10.7+5.4% in SLE), Th1 cells (e.g 58.4+2.7 vs. 12.4+6.5% of memory T cells in HC), and Tregs (e.g. 2.0+0.4 vs. 0.6+0.2 in SLE) in BM (p=0.009-0.05) and altered Treg phenotype (high HLADR, CCR4 in BM). There were significantly lower CD4+ T cells and higher Tregs in SLE BM and PB. As expected BM samples had higher fractions of precursor B cell populations compared to paired PB (p=0.0001 for CD10+ B cells). BM naïve and transitional B cells had much lower expression of CD23 compared to PB in both SLE and HC (p=0.02-0.006). Notably, mature B cells in SLE BM were very abnormal compared to HC with an expansion of CD27-IgD- memory (p=0.008) and activated, effector memory B cell populations (e.g on the switched memory for SLE vs. HC CD95+% 42.2+18.3 vs. 20.7+6.2, p=0.05 and CD21-% 30.5+13.6 vs. 10.9+2.5, p=0.008).
Conclusion: These results suggest that the bone marrow is an important but previously unrecognized target organ in SLE. Circulating immune complexes and apoptotic fragments in SLE BM may serve as ligands for toll like receptors on pDCs, and along with disturbances in the T cell and DC compartment, contribute to aberrant IFN production. IFN production in the BM impacts B cell development and may contribute to an activated B cell compartment.
Disclosure: S. Nevarez, None; A. Palanichamy, None; A. Rabinovich, None; J. Barnard, None; J. Biear, None; B. Panepento, None; C. Wei, None; C. T. Ritchlin, None; J. Kobie, None; I. Sanz, None; J. H. Anolik, None.