Abstract
Patients with active systemic lupus erythematosus (SLE) have signs of an ongoing IFN-α production, that may be of pathogenic significance in the disease. We previously showed that SLE patients have an IFN-α-inducing factor in blood, probably consisting of complexes containing anti-DNA Abs and immunostimulatory DNA. The DNA component could be derived from apoptotic cells, because SLE patients have been reported to have both increased apoptosis and reduced clearance of apoptotic cell material. In the present study, we therefore investigated whether apoptotic cells, together with IgG from SLE patients, could act as an IFN-α inducer in normal PBMC in vitro. We found that apoptotic cells of the myeloid leukemia cell line U937 as well as four other cell lines (MonoMac6, H9, Jurkat, U266) could induce IFN-α production in PBMC when combined with IgG from SLE patients. The IFN-α production by PBMC was much enhanced when PBMC were costimulated by IFN-α2b. The ability of IgG from different SLE patients to promote IFN-α induction by apoptotic U937 cells was associated with the presence of anti-ribonucleoprotein Abs, but not clearly with occurrence of anti-DNA Abs. These results suggest that apoptotic cells in the presence of autoantibodies can cause production of a clearly immunostimulatory cytokine, which is IFN-α. This mechanism for induction of IFN-α production could well be operative also in vivo, explain the IFN-α production seen in SLE patients, and be important in the pathogenesis of SLE.
Systemic lupus erythematosus (SLE)2 is a multisystem autoimmune disease characterized by production of autoantibodies against DNA and nucleic acid-associated proteins (1). The type I IFN system may be important in the autoimmune process because increased levels of circulating IFN-α and of IFN-α-inducible cellular proteins are found in SLE patients and correlate to disease activity (2, 3, 4, 5, 6, 7, 8). Furthermore, patients with nonautoimmune disorders treated with IFN-α frequently develop antinuclear Abs, Abs to native DNA, and occasionally SLE (9, 10, 11, 12). Interestingly, we recently identified an IFN-α-inducing factor (SLE-IIF) in sera from SLE patients, especially those with active disease (13). Given the important immunoregulatory functions of type I IFN (14, 15, 16, 17, 18, 19), the SLE-IIF could impair induction of self-tolerance and promote autoimmunity via produced IFN-α. Such an action is rendered even more likely by the fact that the SLE-IIF appears to selectively activate the natural IFN-producing cells (natural IPC) (13), which have a phenotype resembling that of immature dendritic cells (DC) of the DC2 type (13, 20, 21, 22) that are potent stimulators of immune responses (23).
The SLE-IIF was shown to consist of complexes of anti-DNA Abs and DNA (13), the latter resembling hypomethylated immunostimulatory (is) DNA in function (24). The DNA component may be derived from apoptotic cells, because SLE patients have been reported to have both increased apoptosis and reduced clearance of apoptotic cell material (25, 26, 27, 28). In the present study, we therefore investigated whether apoptotic cells, together with IgG from SLE patients, could act as an IFN-α inducer in normal PBMC in vitro. We found that apoptotic cells of the U937 myeloid leukemia cell line could induce IFN-α production in PBMC when combined with IgG from SLE patients. The ability of IgG from different SLE patients to promote IFN-α induction by apoptotic U937 cells appeared related to presence of anti-ribonucleoprotein (RNP) Abs, but not to anti-DNA Abs. This mechanism for induction of IFN-α production may explain the IFN-α production seen in SLE patients in vivo and could be important in the pathogenesis of the disease.
Materials and Methods
Patients and Controls
A total of 22 SLE patients (20 female and 2 male) with a median age of 42.5 years (range 14–68 years) and a mean duration of disease of 10.5 years (range 1–49 years) were included in the study. The median American College of Rheumatology index (29) for the patients was 6 (range 4–10). Disease activity was measured by the SLE disease activity index (SLEDAI) (30), where complement levels and anti-DNA Abs were excluded. Citrated plasma samples were obtained, converted to serum by addition of 1 M CaCl2, and stored at –80°C. A large plasma sample was collected by plasmapheresis of a 16-year-old female patient (SLE 1), with a SLEDAI score of 10. Plasma samples from four normal blood donors, median age 28 years (range 20–31 years), served as controls. The study protocol was approved by the Committee of Ethics, Faculty of Medicine, Uppsala University.
Determination of autoantibodies
Specific autoantibodies directed toward Smith (Sm), RNP or Sjögrens Syndrome antigens A and B (SSA and SSB, also termed Ro and La) (1), were determined by investigations with immunodiffusion (Auto I.D. plates; Immuno Concepts, Sacramento, CA) yielding immunological identity reactions with predefined autoantibody containing samples. IgG anti-cardiolipin (anti-CL) Abs were determined using a commercial ELISA (Autozyme; Cambridge Life Sciences, Ely, Cambridgeshire, U.K.). Abs against dsDNA were detected by indirect immunofluorescence using Crithidia luciliae (Immuno Concepts) or when indicated by an anti-dsDNA Ab ELISA kit (Dako, Glostrup, Denmark).
Preparation of IgG and removal of anti-DNA Ab
Sera were prefiltered by using 0.45-μm filters (Acrodisc; Gelman Sciences, Ann Arbor, MI) and then treated for 1 h at 37°C with equal volumes of DNAse I (2000 U/ml; Boehringer Mannheim, Mannheim, Germany) in 100 mM Tris-HCl (pH 7.5) and 10 mM MnCl2 to eliminate endogenous DNA. Subsequently, IgG was purified on protein G Sepharose (Amersham Pharmacia Biotech, Uppsala, Sweden) as recommended by the manufacturer. The IgG containing eluates were dialyzed against RPMI 1640 medium (ICN Biomedical, Costa Mesa, CA) supplemented with penicillin (60 μg/ml), streptomycin (100 μg/ml), l-glutamine (2 mM), and HEPES (20 mM) and was used at a concentration of 1 mg/ml in the cultures.
Abs against dsDNA and ssDNA were removed by separation of IgG from patient SLE-1 on dsDNA or ssDNA cellulose columns (Worthington, Freehold, NJ) using uncoupled cellulose columns as control (31). The effluents were desalted and transferred to RPMI 1640 medium, supplemented as described above, using PD-10 columns (Amersham Pharmacia Biotech).
Human monoclonal anti-dsDNA and anti-ss/dsDNA Abs (MER-2 and MER-3; Serotec, Oxford, U.K) were dialyzed against RPMI 1640 medium and used in cultures at concentrations of 7.5 μg/ml and 2.5 μg/ml, respectively.
Culture and treatment of cell lines
The monocytic cell lines U937 and MonoMac6, the T cell lines Jurkat and H9, and the B cell line U266 were cultured in RPMI 1640 medium supplemented with 5% FCS (10% for U266) (Myoclone; Life Technologies, Paisley, U.K.), penicillin (60 μg/ml), streptomycin (100 μg/ml), l-glutamine (2 mM), and HEPES (20 mM), at 37°C in 7% CO2. The cells were treated at 1 × 106 cells/ml by UV light (254 nm, 60 mJ). The U937 cells were also treated by 12 μM etoposide (Sigma, St. Louis, MO) for 3 or 6 h or by 1 μg/ml anti-Fas mAb (clone CH-11; Immunotech, Marseilles, France) for 1 h. Cells treated with etoposide and anti-Fas mAbs were then washed once. The treated U937 cells were then cultured for 4 or 24 h (UV light), for 1 or 18 h (etoposide), and for 3 or 21 h (anti-Fas). The other cell lines were cultured for 4 h after UV light treatment. For all cultures, 96-well round-bottom plates (Nunclon; Nunc, Roskilde, Denmark) were used. The final concentration of U937 cells used in the cultures was 0.5 × 106 cells/ml, which was optimal for IFN-α induction.
The irreversible caspase inhibitor zVAD-fmk (Calbiochem, Cambridge, MA) was used at a concentration of 50 μM to inhibit apoptosis in some experiments.
No mycoplasma could be detected in any of the cell lines by staining with bisbenzimide (Hoechst no. 33258; Sigma). The U937 cells were also negative when assessed by PCR using mycoplasma group-specific primers complementary to the 16S rRNA genes (32) or by semi-nested PCR using primers complementary to universal regions of bacterial 16S rRNA genes (33).
Herpes simplex virus
The HSV was prepared and UV inactivated as described before (13) and used as a control IFN-α inducer at a final concentration of 2 × 107 PFU/ml in the cultures.
Preparation and culture of PBMC
Human PBMCs were prepared by Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) density gradient centrifugation of buffy coats from normal blood donors. The cells were washed in PBS four times and stored frozen at –80°C in FCS containing 10% DMSO. Before using, the PBMCs were thawed rapidly, washed twice in RPMI 1640 medium, and cocultured at 2 × 106 cells/ml with U937 cells (0.5 × 106 cells/ml) in RPMI 1640 medium supplemented as above, but with 3.75% FCS and 500 U/ml recombinant IFN-α2b (Intron-A; Schering-Plough, Bloomfield, NJ), if not otherwise indicated. Triplicate cultures with final volumes of 100 μl/well in 96-well round-bottom plates (Nunc) were incubated for 24 h at 37°C and 7% CO2. The PBMCs were initially selected for a good IFN-α production in response to HSV and SLE serum.
Apoptosis assays
Apoptosis was detected by annexin V or TUNEL staining. In brief, the former was performed by mixing 50 μl U937 cells (1 × 106 cells/ml) with 2.5 μl FITC-labelled recombinant annexin V and 5 μl propidium iodide (50 μg/ml). After 15 min incubation in the dark at room temperature, 400 μl of annexin V binding buffer (10 mM Hepes/NaOH, pH 7.4, 140 mM NaCl, 2.5 mM CaCl2) was added. The TUNEL assay was done using the APO-Direct kit according to the manufacturer’s description, detecting DNA breaks by incorporation of FITC-labeled dUTP and total DNA by propidium iodide (PI) staining. All apoptosis assay reagents were obtained from Pharmingen (San Diego, CA). Analysis of stained cells were performed by a FACScan flow cytometer and the CellQuest software (Becton Dickinson, San Jose, CA).
Immunoassay for IFN-α
The levels of IFN-α in culture supernatants were determined by dissociation-enhanced lanthanide fluoroimmunoassay (DELFIA) as described (34), with modifications. Briefly, microtiter plates were coated with the anti-IFN-α mAb LT27:293, which detects the majority of IFN-α subtypes but not the IFN-α2b used for costimulation in the IFN-α induction cultures. Samples and standard were then coincubated with the europium-labeled LT27:297 anti-IFN-α mAb for 1 h at 37°C in the LT27:293-coated immunoplates. The detection level was 2 U/ml. The IFN-α standard was calibrated against the National Institutes of Health reference leukocyte IFN-α GA-23-902-530.
Statistics
Data are expressed as means ± SD. The significance of differences was determined by the Wilcoxon signed rank sum test or the Mann-Whitney U tests as indicated.
Results
Apoptotic U937 cells combined with either SLE serum or IgG induce IFN-α production in PBMCs
We asked if apoptotic U937 cells alone, or when combined with SLE sera, could trigger IFN-α production in normal PBMCs. The latter were cocultured for 24 h with UV-treated or control-untreated U937 cells and sera from 22 SLE patients, and the IFN-α levels were then measured.
A total of 14 of 22 SLE sera alone induced a production of >25 U/ml IFN-α in PBMC (Table I⇓), which was in accordance with earlier findings (13). Apoptotic U937 cells combined with SLE sera increased the IFN-α production significantly (p < 0.001), while untreated U937 cells with the same SLE sera had a much smaller stimulatory effect (p < 0.05). Normal control sera could not induce IFN-α production irrespective of addition of apoptotic U937 cells or not (results not shown). These results and our previous observations (13, 24) suggested the possibility that autoantibodies from SLE sera together with apoptotic cells are capable of inducing IFN-α.
Sera from SLE patients together with UV-treated U937 cells induce IFN-α production in normal PBMCa
We then investigated whether the component in SLE sera that triggered IFN-α production in combination with apoptotic U937 cells consisted of Ig. The SLE sera were therefore DNAse treated, and IgG were subsequently purified on a protein G column. This procedure also had the advantage of destroying the DNA that is an essential part of the endogenous IFN-α inducer in SLE (13). In addition to UV light, we used etoposide and anti-Fas Abs as inducers of apoptosis in U937 cells. Purified SLE IgG, but not control IgG, in combination with all three types of apoptotic U937 cells resulted in a strong IFN-α production in normal PBMC (Table II⇓). The IFN-α levels attained were comparable to those caused by an optimal concentration of the potent viral IFN-α inducer HSV.
The combination of apoptotic U937 cells and IgG from SLE patients induce production of IFN-αa
As controls, apoptotic U937 cells were shown to have no significant effect on HSV-induced IFN-α response in PBMC (Table II⇑). The protein G column effluent was unable to induce IFN-α together with apoptotic U937 cells (Fig. 1⇓). Furthermore, viable or apoptotic U937 cells alone were unable to produce IFN-α upon addition of SLE-IgG, HSV, or IFN-α2b (results not shown). PBMCs alone, without addition of U937 cells, produced little or no IFN-α, regardless of the presence of SLE-IgG (Fig. 2⇓).
Anti-DNA Abs are not required for the ability of the combination of SLE-IgG and apoptotic U937 cells to cause IFN-α production in normal PBMC. The IgG was purified from one SLE serum on a protein G column and further separated on a dsDNA and/or a ssDNA cellulose column. Effluents (see Materials and Methods) in combination with untreated (□) or UV-treated (▪) U937 cells were added to PBMC cultures and levels of IFN-α (U/ml; means ± SD) at 24 h were determined by immunoassay. Data from one of two experiments with similar results.
The effect of the caspase inhibitor zVAD-fmk on the ability of the combination of SLE IgG and apoptotic U937 cells to induced IFN-α production in PBMC. Apoptosis was induced in U937 cells by UV, anti-Fas Abs, or etoposide in the absence (□) or presence (▨) of zVAD-fmk (see Table I⇑ and Figs. 1⇑ and 2⇓), using untreated cells as control. Normal PBMC were then cultured with the combination of treated or control U937 cells and SLE IgG. Levels of IFN-α (U/ml; means ± SD) were measured by immunoassay after 24 h. PBMC alone denotes cultures without U937 cells. Representative results from one of three experiments are shown.
Taken together, our data clearly demonstrate that the IFN-α-inducing property was restricted to the combination of SLE-IgG and apoptotic U937 cells. We further examined whether the ability to induce IFN-α production was unique to U937 cells or present also in other cell lines. As shown in Table III⇓, MonoMac6, H9, Jurkat, and U266 cells caused IFN-α production in PBMC when combined with SLE IgG. Furthermore, all cell lines, except H9, induced more IFN-α when treated with UV light.
Cell lines inducing IFN-α production by PBMC in presence of SLE IgG
Costimulatory activity of IFN-α2b
In the present investigation, IFN-α2b was added to the cultures to enhance the IFN-α production of PBMCs. The IFN-α2b increased IFN-α production by PBMCs induced by UV-treated U937 cells and SLE-IgG in a dose-dependent manner (Fig. 3⇓). The increase in IFN-α production was 5-fold in cultures with 500 U/ml of IFN-α2b compared to cultures without IFN-α2b. Therefore, this IFN-α2b concentration was used in all PBMC cultures to achieve maximal IFN-α production. The cytokines GM-CSF and IFN-γ that in other systems had costimulatory activity (35) were without costimulatory effect either alone or in combination with IFN-α (results not shown). It should be noted that the immunoassay for IFN-α does not detect the added costimulatory IFN-α2b.
IFN-α2b increases the IFN-α production by normal PBMC induced by the combination of SLE-IgG and apoptotic U937 cells. The PBMC were cultured with different concentrations of recombinant IFN-α2b and the combination of SLE-IgG and either untreated U937 cells (□) or UV-treated U937 cells (▪). The levels of IFN-α (U/ml; mean ± SD) were measured after 24 h by immunoassay. Representative results from one of four experiments.
Inhibition of apoptosis in U937 cells reduces their capacity to induce IFN-α production
To verify that U937 cells had to be apoptotic to induce IFN-α production, the caspase inhibitor zVAD-fmk was added to the cells before apoptotic treatment. By adding zVAD-fmk to UV light or etoposide-treated U937 cells, the level of apoptosis (annexin V-positive and PI-negative cells) was reduced by 90% and 94%, respectively (Fig. 4⇓A). In contrast, zVAD-fmk did not affect the level of annexin V-positive and PI-negative staining in U937 cells treated with anti-Fas, but did decrease the number of apoptotic cells determined by the TUNEL technique (Fig. 4⇓B).
The effect of the caspase inhibitor zVAD-fmk on apoptosis induced in U937 cells. The U937 cells were treated by UV light (60 mJ), anti-Fas Abs (1 μg/ml) for 1 h, or etoposide (12 μM) for 3 h, in the absence (▨) or presence (□) of 50 μM zVAD-fmk. After 4 h of culture, apoptotic U937 cells were detected by annexin V and PI staining (A) or by TUNEL staining (B) as described in Materials and Methods. Annexin V-positive, PI-negative cells were considered apoptotic in A. The percentage of positive cells was determined by flow cytometry, sample size 10,000 cells. Data from one of two experiments with similar results.
With regard to the IFN-α production, zVAD-fmk markedly inhibited the ability of U937 cells treated by UV, etoposide, or anti-Fas to induce IFN-α production in PBMC in the presence of SLE-IgG by 86%, 86%, and 55%, respectively (Fig. 2⇑). In contrast, the zVAD-fmk did not inhibit the HSV-induced IFN-α production in PBMC (results not shown).
Other autoantibodies than anti-dsDNA Abs are required for apoptotic U937 cells to induce IFN-α production in PBMCs
The combination of anti-dsDNA Abs and DNA is required for the ability of SLE sera to induce IFN-α production in normal PBMCs in vitro (13, 24). Therefore, we asked whether anti-dsDNA Abs in SLE-IgG were also necessary for apoptotic U937 cells to induce IFN-α production.
Purified SLE-IgG from patients with ≥40 IU/ml anti-dsDNA Abs, together with apoptotic U937 cells, induced IFN-α production in normal PBMCs (Table IV⇓). However, IgG from one patient without significant levels of anti-DNA Abs (<8 IU/ml) was also stimulatory. In this small patient population, the IFN-α-inducing capacity of IgG from patients with active disease as measured by SLEDAI score was higher compared to IgG from patients in remission (p = 0.014, Mann-Whitney U test).
Induction of IFN-α production in PBMC by apoptotic U937 cells and IgG from different SLE patientsa
To further demonstrate that anti-DNA Abs were not required for the IFN-α production induced by apoptotic U937 cells, SLE-IgG was passed over dsDNA columns, ssDNA columns, or both. In all cases, this resulted in ∼95% depletion of anti-dsDNA Abs (results not shown). Still, such SLE-IgG had an intact ability to cause IFN-α synthesis together with UV-treated U937 cells (Fig. 1⇑). Furthermore, monoclonal anti-dsDNA and anti-ssDNA Abs, in the presence of apoptotic U937 cells, were unable to induce IFN-α (results not shown). Consequently, anti-DNA Abs are not required for the induction of IFN-α synthesis in the present experimental system. Therefore, we attempted to clarify whether other autoantibodies in SLE patients were involved in the induction of IFN-α production in the presence of apoptotic U937 cells. This was done by comparing the ability of SLE sera (same as in Table I⇑) with or without a certain Ab to induce IFN-α production in normal PBMC, either alone or together with UV-treated U937 cells. The IFN-α levels caused by SLE sera alone were subtracted from the IFN-α levels caused by the combination of sera and UV-treated cells. This was done to remove the impact of the previously described SLE-IIF, which is dependent on anti-dsDNA Abs (24) and thus reveal the effect of the apoptotic U937 cells and other autoantibodies. As shown in Table V⇓, the ability of different SLE sera to induce IFN-α production in PBMC, when combined with apoptotic U937 cells, was clearly associated with the occurrence of RNP Abs but not anti-SSA, anti-SSB, anti-Sm, or anti-CL Abs. As expected, the IFN-α production caused by the SLE serum alone correlated only to the presence of anti-dsDNA Abs.
The occurrence of different autoantibodies in SLE sera is related to the ability of SLE sera alone or combined with UV-treated U937 cells to cause IFN-α production in normal PBMC
Discussion
Apoptosis is usually considered to be a controlled process to eliminate damaged cells without causing an inflammatory response. In fact, apoptotic cells have been shown to induce production of the anti-inflammatory cytokines IL-10 and TGF-β1 that can decrease production of proinflammatory cytokines such as TNF-α, IL-1β, and IL-12 (36, 37). Our finding that the combination of apoptotic U937 cells and IgG from SLE patients can induce IFN-α production in human PBMC is therefore of considerable interest, because IFN-α is a proinflammatory cytokine that can promote autoimmunity (9, 10, 11, 12, 38).
In our study, only the combination of SLE-IgG and apoptotic U937 cells, but not either component alone, had IFN-α-inducing activity. A likely reason for the effect of SLE-IgG is the presence of autoantibodies. A role for anti-DNA Abs were first considered, because we previously demonstrated an endogenous circulating IFN-α inducer in SLE patients consisting of anti-DNA Abs and DNA as essential components (13, 24). Surprisingly, when IgG fractions from different SLE patients were compared, we were unable to connect the IFN-α production caused by apoptotic U937 cells and SLE-IgG to the occurrence of anti-DNA Abs. Furthermore, removal of anti-DNA Abs by absorption to DNA-cellulose had no effect on the IFN-α-inducing capacity. In contrast, there was a clear association between presence of anti-RNP Abs in SLE sera and ability of the sera to induce IFN-α production when combined with apoptotic U937 cells. No such association could be detected for the other Abs investigated: anti-SSA, anti-SSB, anti-CL, anti-Sm, and anti-dsDNA. Further studies are necessary to clarify whether it is the anti-RNP Abs or coexisting Abs with other specificities that are involved in the IFN-α production.
With regard to the role of the U937 cells in the IFN-α production, apoptosis was important because pretreatment of these cells with the apoptosis inducers etoposide or UV light greatly increased their IFN-α-inducing capacity. Furthermore, this increase was inhibited by pretreatment of U937 cells with the broad-spectrum caspase inhibitor zVAD-fmk. However, somewhat discrepant results were obtained with anti-Fas Ab-treated U937 cells, because the zVAD-fmk only caused a partial reduction of their IFN-α-inducing capacity. The latter finding may be related to our observation that zVAD-fmk inhibited cell death in a different way in anti-Fas-treated U937 cells compared to UV light or etoposide-treated cells. Thus, the zVAD-fmk treatment clearly inhibited the DNA fragmentation process triggered by all three apoptosis inducers, as determined by the TUNEL method. Paradoxically, the number of annexin V-positive and PI-negative cells did not decrease among anti-Fas-treated U937 cells. This may indicate that zVAD-fmk failed to inhibit early apoptotic events in anti-Fas-treated U937 cells, because annexin V binds to phosphatidylserine and indicates an early phase of apoptosis (39). However, we observed that the number of possibly necrotic annexin/PI double-positive cells increased (results not shown). The latter may represent the results of a Fas-mediated triggering of a necrotic pathway when apoptosis is repressed, as has been reported in murine L929 cells transfected with the human Fas gene (40). While our results indicate that cells dying by apoptosis can contribute to the induction of IFN-α production, it is therefore not possible to exclude that other forms of cell death may also be relevant.
We previously demonstrated that SLE-IIF in serum contains DNA and anti-DNA Abs, and that such Abs from SLE patients convert plasmid isDNA into a strong inducer of IFN-α synthesis in normal PBMC (24). Furthermore, the presence of isDNA sequences in serum of SLE patients have been identified by molecular cloning and a pathogenic role suggested (41, 42, 43), and similar isDNA motifs consisting of unmethylated CpG palindromes have been shown to induce IFN-α production (44, 45, 46, 47, 48). For these reasons, we speculate that isDNA could be generated by the U937 cells and be the actual IFN-α inducer in PBMC in the present study. Such isDNA could be present as naked DNA, nucleosomes, or larger chromatin fragments. It is relevant that nucleosomes can be internalized in cells by Abs to histones or DNA, have biologic activities such as causing cell proliferation and IgG synthesis (49, 50, 51), and could possibly also induce IFN-α gene expression when they have a high content of hypomethylated CpG-rich DNA. However, our findings in the present study that anti-DNA Abs were not active together with apoptotic U937 cells actually argues against a role for free DNA or nucleosomes in the induction of IFN-α production in the present experimental system. The explanation could be that the apoptotic DNA is encapsulated in cells or apoptotic bodies and therefore not accessible to anti-DNA Abs and that other Abs therefore are relevant.
Alternatively, another type of IFN-α inducer than isDNA could be important. For instance, the suggestion that anti-RNP Abs are involved points to the possibility that RNA/protein complexes released by apoptotic cells (52) could also act as an IFN-α inducer. In fact, dsRNA is the prototype IFN-α inducer and RNA with such activity may actually be present in normal eukaryotic cells (14, 53). Consequently, we cannot exclude that at least two different IFN-α inducers are present in SLE patients. One inducer consists of anti-dsDNA Abs and isDNA and the other depends on Abs of other specificities, perhaps anti-RNP. The components of the latter inducer, including any RNA responsible for triggering IFN-α gene expression, are now being identified.
In the experiments with SLE-IgG, we noted an association between clinical disease activity and ability of SLE-IgG to induce IFN-α production in PBMC together with apoptotic cells. This association could not be verified when using unfractionated SLE sera from the patients in Table I⇑ together with apoptotic U937 cells (results not shown). However, the possible association of IFN-α-inducing ability and disease activity should be evaluated in a larger patient population.
We demonstrated before that both the SLE-IIF (13) and plasmid DNA and anti-DNA Ab complexes (24) selectively induced IFN-α production in the natural IPC among PBMC. The phenotype of these cells resemble that of immature DC (20) and correspond to that later described for precursors of the DC2 (21, 22, 23). The same natural IPC are responsible for the IFN-α production caused by the combination of apoptotic U937 cells and SLE-IgG (U. Båve, H. Vallin, L. Rönnblom, and G. V. Alm, manuscipt in preparation). Because activation of IFN-α genes requires uptake of the IFN-α inducer (54), which in the present system appears to be DNA/RNA/protein complexes, it is possible that the same cell may both present autoantigens and produce immunostimulatory IFN-α in SLE, thus promoting the autoimmunization process.
We also found that the IFN-α production by PBMC cocultured with SLE-IgG and apoptotic cells was markedly enhanced by IFN-α2b. Such a costimulatory effect of type I IFN on the IFN-α response, termed priming, has previously been reported mainly for viral inducers and is under certain conditions necessary for IFN-αβ gene transcription (55, 56). Such prominent priming effects of type I IFN were also noted for the IFN-α production induced by SLE-IIF and by plasmid DNA/anti-DNA Abs (24). Accordingly, IFN-α production caused by especially viral infections in patients with inactive SLE might prime natural IPC to respond to complexes of IgG and material from apoptotic cells and therefore initiate more vigorous and sustained synthesis of IFN-α, which would increase disease activity. Such a mechanism could explain cases of disease relapses in SLE patients reported during viral infections (57, 58, 59).
The hallmarks of active SLE include increased levels of apoptotic cells (25, 26, 28), presence of autoantibodies against DNA and nucleic acid-associated proteins and several other autoantigens (1), and ongoing IFN-α production (2, 3, 5, 7, 8, 60, 61). Thus, all components seen in our in vitro system are present in SLE patients. Therefore, it is possible that the same mechanisms by which SLE-IgG and apoptotic cells induce IFN-α production in vitro also operates in vivo. Considering the immunostimulatory actions of IFN-α (14, 15, 16, 17, 18, 19) and its ability to promote autoimmunity in humans (9, 10, 11, 12, 38), the results of the present investigation can be crucial for understanding the etiology and pathogenesis of SLE. Therefore, it is obviously important to further define the identity and action of the active component(s) in the IgG fraction and in the apoptotic U937 cells.
Acknowledgments
We thank Anders Perers for excellent technical assistance, Lotta Karlnell for collecting sera from SLE patients, Drs. Helena Vallin and Anders Johannisson for invaluable technical advice, Dr. Bo Nilsson for autoantibody determinations, Dr. Karl-Erik Johansson for performing assays for mycoplasma, and Dr. Dan Grandér for kindly providing the H9 and U266 cell lines.
Footnotes
-
↵1 Address correspondence and reprint requests to Dr. Ullvi Båve, Immunology (V), Biomedical Center, P.O. Box 588, 75123 Uppsala, Sweden. E-mail address: Ullvi.Bave{at}medicin.uu.se
-
↵2 Abbreviations used in this paper: SLE, systemic lupus erythematosus; CL, cardiolipin; DELFIA, dissociation-enhanced lanthanide fluoroimmunoassay; DC, dendritic cell; IIF, IFN-α-inducing factor; IPC, IFN-α-producing cells; isDNA, immunostimulatory DNA; SLEDAI, SLE disease activity index. RNP, ribonucleoprotein; Sm, Smith; SSA and B, Sjögrens syndrome Ag A and B; PI, propidium iodide.
-
↵3 This work was supported by grants from the Tore Nilson Foundation, the 80 Years Foundation of King Gustaf V, the Swedish Rheumatism Foundation, the Tore and Wera Cornell Foundation, and the Swedish Medical Research Council.APOPTOTIC U937 CELLS COMBINED WITH SLE-IgG INDUCE IFN-α
- Received November 8, 1999.
- Accepted July 5, 2000.
- Copyright © 2000 by The American Association of Immunologists