Induction of Dendritic Cell Differentiation by IFN-α in Systemic Lupus Erythematosus

See allHide authors and affiliations

Science  16 Nov 2001:
Vol. 294, Issue 5546, pp. 1540-1543
DOI: 10.1126/science.1064890


Dendritic cells (DCs) are important in regulating both immunity and tolerance. Hence, we hypothesized that systemic lupus erythematosus (SLE), an autoimmune disease characterized by autoreactive B and T cells, may be caused by alterations in the functions of DCs. Consistent with this, monocytes from SLE patients' blood were found to function as antigen-presenting cells, in vitro. Furthermore, serum from SLE patients induced normal monocytes to differentiate into DCs. These DCs could capture antigens from dying cells and present them to CD4-positive T cells. The capacity of SLE patients' serum to induce DC differentiation correlated with disease activity and depended on the actions of interferon-α (IFN-α). Thus, unabated induction of DCs by IFN-α may drive the autoimmune response in SLE.

Lupus (SLE) is a systemic autoimmune disease characterized by a waxing and waning course and the involvement of multiple organs, including skin, kidneys, and central nervous system (1). Although SLE etiopathology is poorly understood, there is likely a role for environmental triggers, for instance viruses, acting in the context of susceptibility genes (2,3). Now, SLE therapy is based on nonspecific immunosuppression and symptom control; however, no cure for SLE has been found, and that is particularly needed for children, whose disease often progresses to chronic renal failure and/or death (1). Among the immunological features of SLE are high titers of autoantibodies predominantly specific for DNA and nucleosomes (4, 5). Paradoxically, polyclonal hypergammaglobulinemia, high plasma cell numbers, and an increased frequency of pre–germinal center B cells (6,7) coincide with considerable B lymphopenia. Likewise, autoreactive T cells are found in the blood despite T lymphopenia (8, 9). The systemic autoimmune response that characterizes SLE might be explained by alterations in the function of dendritic cells (DCs), because these cells are key regulators of the immune system. DCs are professional antigen-presenting cells (APCs) capable of inducing activation of naı̈ve T cells (10, 11) and of stimulating B cell growth and differentiation (12, 13). We therefore investigated whether blood from pediatric SLE patients displayed any abnormalities in DC function.

We found that CD11c+ myeloid DCs (mDCs) were markedly decreased in SLE blood [∼44% decrease (14)] but not in a control group of patients with juvenile dermatomyositis (JDM) (P = 0.1) who undergo a similar regimen of steroid therapy (14). Monocytes, which are the precursors of mDCs, were only 20% decreased (14,15). Although normal monocytes are not efficient stimulators of naı̈ve allogeneic CD4+ T cell proliferation [mixed lymphocyte reaction (MLR)], those derived from a proportion of SLE patients were able to induce strong MLR (Fig. 1A), a property that can normally be used to define DCs (16). This led us to the hypothesis that SLE patients' serum might favor differentiation of monocytes into DCs. Consistent with this, exposure of normal monocytes to such serum resulted, within 24 to 72 hours, in the clustering and formation of cells with morphology typical of DCs (Fig. 1B). These cells displayed a pattern of surface markers consistent with mature DC differentiation, including a down-regulation of CD14, along with an increased expression of major histocompatibility complex (MHC) class II, costimulatory molecules CD80 and CD86, and CD83 (Fig. 1C). Furthermore, normal monocytes cultured with SLE serum induced a strong MLR, whereas those cultured with autologous serum did not (Fig. 1D) (16). When serum samples from 20 different patients were tested for DC-inducing activity (16), 12 of them generated DCs able to induce proliferation of allogeneic T cells (17). Meanwhile, other samples of allogeneic serum from eight healthy controls and from patients with other systemic autoimmune diseases (i.e., JDM and juvenile arthritis, n = 5) were unable to do so (17). This capacity to induce DC differentiation correlated with disease activity according to SLEDAI (SLE disease activity index; P = 0.03; Fig. 1E) (16). Finally, normal monocytes cultured with SLE patients' serum, but not those cultured with autologous serum, were able to capture dying allogeneic cells (Fig. 2, A and B) and to present their antigens to autologous normal CD4+ T cells leading to their proliferation (Fig. 2C) (15). Thus, morphology, phenotype, and antigen-presenting capacity indicate that serum from SLE patients induces monocytes to differentiate into DCs. It is noteworthy that, in the context of SLE, these DCs present antigens from captured apoptotic cells (18).

Figure 1

(A) Monocytes from a proportion of SLE patients induce proliferation of allogeneic CD4+ T cells. Lowercase letters refer to patients. (B) SLE patients' serum induces monocytes to differentiate into cells with DC properties. Monocytes cultured with SLE patients' serum (left), but not with autologous serum (right), cluster within 12 to 24 hours, and, within 24 to 72 hours, clustered cells display fine cytoplasmic projections (Giemsa staining, middle). (C) Flow cytometry analysis of surface markers. (D) Monocytes cultured with serum from SLE patients induce proliferation of naı̈ve CD4+CD45RA+ T cells. Representative of three experiments. (E) Induction of DC differentiation is related to disease activity. Monocytes cultured with 20 different samples of serum from SLE patients were tested for induction of T cell proliferation. GM-CSF and IL-4 cultured monocytes were used as a 100% “gold standard.” The allostimulatory activity of monocytes cultured with SLE patients' serum is expressed as a percentage of T cell proliferation induced by GM-CSF/IL-4 monocytes (horizontal axis) and related to disease activity (SLEDAI, vertical axis). SLEDAI consists of nine groups of clinical and laboratory criteria that include assessment of organ systems: CNS, vascular, renal, musculoskeletal, serosal, dermal, immunologic, and hematologic. Simple regression analysis (r 2 = 0.23).

Figure 2

Monocytes cultured with SLE patients' serum present antigens from apoptotic cells to autologous CD4+ T lymphocytes. (A) Flow cytometry phagocytosis assay with HLA-DR fluorescein isothiocyanate (FITC)-stained monocytes cultured with serum from SLE patients and DNA-labeled (7AAD), killed allogeneic cells. Phagocytosis is determined as the percentage of double fluorescence. (B) Monocytes cultured with SLE patients' serum present antigens from apoptotic cells to autologous CD4+ T cells as indicated by cell proliferation (means ± SD from triplicates). Representative of three experiments.

We next investigated the nature of the SLE serum factor(s) that might cause monocyte differentiation into DCs. Several cytokines can induce the differentiation of monocytes into DCs including granulocyte/macrophage colony-stimulating factor (GM-CSF), interleukin 4 (IL-4), IFN-α, and CD40-L (11). Among those, relatively large amounts of IFN-α and CD40-L have been found in the serum of many SLE patients, and their titers correlate with the titers of anti-DNA antibodies and disease activity (19–22). As shown in Fig. 3A, DC induction, determined by the capacity of normal monocytes cultured with SLE patients' serum to stimulate MLR, was inhibited by neutralizing antibodies against IFN-α (23). In contrast, blocking CD40-L, IL-4, or GM-CSF with relevant antibodies did not alter the DC-inducing capacity of serum from SLE patients (14). These results point to IFN-α as the major factor in SLE serum that mediates monocyte differentiation to DCs. Accordingly, the addition of IFN-α to normal serum induced monocytes to become potent stimulators of MLR (Fig. 3B) (23). Finally, DC induction correlated with increased IFN-α in serum (P < 0.001) (Fig. 3C) (23). Consistent with this, the ability of monocytes derived from SLE patients to stimulate MLR correlated with IFN-α in serum as well (17).

Figure 3

Induction of DC differentiation is IFN-α dependent. (A) IFN-α neutralizing antibody (10 μg/ml) blocks the induction of DCs by SLE patients' serum. Representative of three experiments. (B) Spiking autologous serum with 100 UI/ml IFN-α results in the induction of monocyte differentiation to DCs as determined by MLR capacity. (C) The capacity of monocytes cultured with serum from SLE patients to stimulate MLR is expressed as in Fig. 1E and related, in simple regression (r 2 = 0.70), to the serum levels of IFN-α (range of ELISA: 12.5 to 500 pg/ml).

Finally, we investigated the cellular source of IFN-α in SLE patients' blood. Potentially, IFN-α could originate from CD11cCD123+ plasmacytoid DCs (pDCs), a unique subset of DCs, which are a major source of IFN-α in the blood (24). Unexpectedly, pDC numbers were markedly reduced in the blood of SLE patients (Fig. 4A). This may have reflected an accelerated migration of pDC from the blood into tissues (25). Although low in numbers, the CD123+ cells isolated from SLE patients' blood produced normal levels of IFN-α in response to viral triggering (Fig. 4B). We also examined whether pDCs were the only source of IFN-α in SLE patients. Depletion of CD123+ cells from normal peripheral blood mononuclear cells (PBMCs) resulted in a 95% reduction of IFN-α release upon viral triggering (Fig. 4C). In contrast, depletion of CD123+cells from PBMCs from SLE patients resulted in an average 57% reduction of IFN-α release upon viral triggering (Fig. 4C). Thus, other cells within PBMCs are capable of IFN-α production in SLE.

Figure 4

(A) Pediatric SLE patients have lower numbers of HLA-DR+CD11cCD123+ pDCs. Numbers of pDCs × 103/ml blood (means ± SEM;t test). (B) CD123+ cells from either normal blood or SLE blood release similar amounts of IFN-α after stimulation with influenza virus [multiplicity of infection (MOI) of 2]. (C) PBMCs from SLE patients retain the capacity to produce IFN-α after depletion of CD123+cells. PBMCs from four normal donors and from five SLE patients, either total or depleted of CD123+ cells (by flow sorting), were cultured overnight with influenza virus (MOI 2). Percentage of IFN-α production reflects the IFN-α produced by PBMCs depleted of CD123+ cells as compared with the amounts from total PBMCs (unpaired t test on log-transformed data and nonparametric Mann-Whitney test).

In conclusion, patients with SLE display major alterations in DC homeostasis. Whereas one DC subset, i.e., pDCs, is dramatically reduced in the blood, normally quiescent monocytes act as DCs in these patients. This is caused by circulating IFN-α that induces monocytes to differentiate into DCs. Such DCs might very efficiently capture apoptotic cells and nucleosomes, present in SLE patients' blood (26, 27). Subsequent presentation of autoantigens to CD4+ T cells could thus initiate the expansion of autoreactive T cells, followed by differentiation of autoantibody-producing B cells. Autoantibodies could form immune complexes with circulating nucleosomes, which could potentially sustain IFN-α production (28). High circulating IFN-α could also explain the puzzling T and B lymphopenia found in SLE (29) and the presence of CD38+ pre–germinal center B cells in SLE patients' blood (7, 30). Taken together, our results suggest that IFN-α might represent a potential target for therapeutic intervention in SLE.

  • * These authors contributed equally to this work.

  • To whom correspondence should be addressed. E-mail: j.banchereau{at}


View Abstract

Navigate This Article