Natural antibody to apoptotic cell membranes inhibits the proinflammatory properties of lupus autoantibody immune complexes

Abstract

Objective

Naturally-arising IgM antibodies (NAbs) to apoptotic cell (AC) determinants are present from birth and can be further induced by AC challenge. In systemic lupus erythematosus, lower anti-AC NAb levels have been associated with higher disease activity. We have recently shown that a prototypic AC-specific NAb IgM can suppress pro-inflammatory responses to purified agonists for Toll-like receptors, and also block the in vivo induction of IgG-immune complex (IC)-induced arthritis. IgG autoantibody-complexes with nuclear antigens, which activate dendritic cells (DCs), have been implicated in autoimmune pathogenesis. Here, we sought to investigate potential roles of such NAbs for regulating immune-complex mediated DC activation, which is believed to be involved in disease initiation and perpetuation.

Methods

Bone-marrow-derived myeloid DCs were stimulated with ICs composed of IgG-autoantibody-chromatin or IgG-autoantibody-RNA. Outcome was evaluated based on production of inflammatory cytokines by ELISA, and expression of co-stimulatory molecules, which are markers of DC activation, by flow cytometry. MAPK activation was evaluated by phospho-flow and immunofluorescence microscopy.

Results

Anti-AC NAb IgM dose-dependently suppressed both DNA- and RNA-IC-induced IL-6 and DNA-IC-induced TNF-α production, as well as RNA-IC-induced upregulation of CD86 and CD40 on DCs. NAb IgM-mediated inhibition was associated with suppression of IC-mediated p38-MAPK activation and nuclear localization.

Conclusions

We demonstrated a direct in vitro inhibitory effect of the NAb IgM on inflammatory responses induced by IgG-nucleic acid ICs. These findings contribute to emerging evidence that regulatory NAbs to apoptotic-cell determinants may oppose the influence of pathogenic lupus autoantibody-ICs and thereby play roles in the maintenance of immune homeostasis.

Introduction

Lupus IgG-autoantibodies can form immune complexes (ICs) with DNA/protein or RNA/protein macromolecular complexes released from dying cells. Following internalization via FcγRs, these ICs can engage endosomal TLRs causing cellular activation that drives autoimmune pathogenesis (reviewed in (1, 2)). While first discovered in SLE, this paradigm of IC-mediated stimulation may also be involved in self-perpetuation of other inflammatory autoimmune diseases, including rheumatoid arthritis (3).

In health, immense numbers of cells die each day from senescence and injury. If not efficiently cleared, these apoptotic cells (ACs) progress to necrosis and release inflammatory factors and autoantigens, which may drive autoimmune responses. Efficient AC clearance is therefore a fundamental function of the innate immune system, and a range of overlapping pathways exist to aid this function.

Apoptotic clearance is promoted by a primordial class of innate-like IgM natural autoantibodies (NAbs), which specifically recognize cell-membrane determinants that include phosphorylcholine (PC), a phospholipid headgroup that becomes exposed and accessible for immune recognition from early stages of apoptotic death (4-7). These anti-AC NAbs arise spontaneously early in life without defined immune exposure, and levels are increased in mice by experimental exposure to dying cells ((6) & reviewed in (8)). Antibodies that recognize PC are also commonly found in humans (9), and a specific B-cell origin has recently been reported (10). Higher anti-AC NAb IgM levels are reported to correlate with fewer cardiovascular events in general populations and correlate with lower autoimmune disease activity, less organ damage and lower incidence of cardiovascular events in lupus patients (11,12).

Recent studies in murine models demonstrated that NAb IgM to ACs can have two major regulatory functions: i) to enhance the clearance of ACs by phagocytes; (ii) to suppress in vivo and in vitro pro-inflammatory responses in innate cells, which has previously been demonstrated with a range of purified TLR agonists. Moreover, anti-AC NAb IgM infusions have been shown to suppress or prevent disease induction in murine arthritis models (5, 6)

Dendritic cells (DCs) play active roles in regulating the balance between tolerance and autoimmunity. To further investigate the mechanistic basis by which NAbs to ACs may modulate processes that otherwise contribute to the breakdown of tolerance, we evaluated the effect of this class of regulatory NAb on the responses of bone marrow-derived DCs (BM-DCs) induced by stimulation with nucleic acid-containing ICs.

Materials and Methods

Antibodies

E06 anti-AC NAb IgM was purified and essentially endotoxin-free, as described (5, 6, 13), from a B-cell hybridoma isolated without immunization or in vitro stimulation from the spleen of a hyperlipidemic apolipoprotein-E deficient mouse (gift of J. Witztum, UCSD). This multimeric IgM expresses antibody genes identical to the prototypic B-1 cell NAb clone, T15, defined by canonical heavy and light chain rearrangements without hypermutation (13). The anti-chromatin IgG hybridoma, PL2-3, was provided by Dr. Monestier (14) and anti-RNA IgG hybridoma, BWR4, was a gift from Dr. Eilat (15). IgG mAbs from culture supernatants were purified by Protein A chromatography (Repligen) following established protocols.

Nucleic acids

Primary necrotic cell extracts were obtained by freezing/thawing splenocytes suspended in RPMI 1640 supplemented with 10% FBS followed by filtration through 0.22 μM membranes. Alternately, ‘B’ Type phosphorothioate CpG ODN 1668 (16) 5′ TCCATGACGTTCCTGACGTT 3′ (Bioneer Inc) was used as a TLR9 agonist in some experiments. ssRNA40, a GU-rich single-stranded RNA phosphorothioate oligonucleotide (Invivogen) provided a source of purified RNA.

IC preparation

Chromatin-ICs were formed either by addition of the anti-chromatin IgG directly to the culture or by pre-incubation with primary necrotic cell extracts (at 6.25 % of final volume) for 1-2 hrs at 37°C (adapted from (17)). RNA-ICs were formed by incubating BWR4 with ssRNA40, in the presence of Protector RNAase Inhibitor (Roche), for 30 mins at 4°C.

DC stimulation assays

C57BL/6 bone marrow cells were cultured, as described (17), in complete media supplemented with GM-CSF (6.67 ng/ml) and IL-4 (0.4 ng/ml) for 3 days. On the 4^th day, an equal amount of media, supplemented as above, was added. On day 6, BM-DCs were purified with anti-CD11c magnetic beads (Miltenyi Biotec) to >94% CD11c+ purity, as previously described (5, 6). BM-DCs were then distributed in flat-bottomed TC-treated plates (Corning) in media supplemented with 10% low IgG FBS (Gibco), GM-CSF and IL-4 (eBioscience). Where indicated, we added CpG ODN 1668 (0.5 μM), IgM (40 μ/ml), C1q (30 μ/ml) (Quidel) and/or dexamethasone (1 μM) (Sigma-Aldrich), or concentrations as described. In some experiments, to enhance cytokine responses, cells were pre-incubated with CD40L–CD8 fusion protein and supernatant from anti-CD8 B-cell hybridoma, 5 3-6.72 (American Type Culture Collection), followed by addition of pre-formed ICs and/or antibodies, as described (17). In all experiments, conditions were performed in triplicate.

Apoptotic cell assays

C57BL/6 thymocytes were treated with 1μM dexamethasone for 8 hrs to induce apoptosis. Apoptotic or freshly-isolated thymocytes were incubated with biotinylated anti-AC NAb or isotype control IgM, followed by anti-chromatin or isotype control IgG antibodies at 37°C. Cells were then washed twice and incubated with Streptavidin-Alexa Fluor^R 647 (Invitrogen) and goat anti-mouse IgG-FITC (Jackson Immunoresearch) antibodies for 30 mins at 4°C. For flow cytometry, cells were additionally stained with Annexin-V PE and 7AAD (BD Biosciences) following manufacturer’s protocols and data was collected on a FACSCalibur (BD Biosciences). For microscopy, cells were incubated with Hoechst 34580 (Invitrogen) for 15 min at 4°C and mounted in suspension with glycerol. Images were immediately acquired at RT on an Applied Precision Personal DC live cell imaging system, using the 60X lens and 1.6X auxillary magnification and the CoolSnap HQ2 CCD camera operated by Softworx software. Acquired images were deconvolved using Softworx and further analyzed by Image J. Composite sections were assembled using Adobe Photoshop and Illustrator.

P-p38 detection

Isotype IgM- or AC-specific NAb-treated BM-DCs were stimulated with anti-chromatin IgG (8 μg/ml) or CpG ODN B (0.5 μM) for 30 mins. For microscopy, cells were added to poly-L-Lysine treated coverslips (BD Biosciences) and allowed to adhere for 10 minutes, followed by fixation with 4% paraformaldehyde 10 mins at RT, blocking with PBS 5% goat serum 0.3% Triton X and incubation with rabbit anti-P-p38 mAb (Thr180/Tyr182) (Cell Signaling) ON at 4°C in PBS 1% BSA 0.3% Triton X following the manufacturer’s protocol. Cells were then washed and incubated with goat anti-rabbit IgG-FITC (Santa Cruz Biotech) in PBS 1% BSA 0.3% Triton X, followed by addition of Hoechst 34580 (Invitrogen) in PBS. Coverslips mounted in Prolong Gold Antifade (Invitrogen) were visualized at RT on an Applied Precision Personal DC live cell imaging system, using the 60X lens and the CoolSnap HQ2 CCD camera operated by Softworx software. Acquired images were further analyzed by Image J. Composite sections were assembled using Adobe Photoshop and Illustrator. Fluorescence intensity of nuclear P-p38 was calculated using an Image J macro (kindly written by Yan Deng, NYU Microscopy Core). Hoechst 34580 staining was used to define the nuclear region of interest for calculation of mean fluorescent intensity (MFI) of P-p38 in the FITC channel.

Flow cytometry

Following FcγR blockade with anti-CD16/32 mAb (2.4G2), cells were stained with anti-CD11c (HL3), anti-CD86 (GL1) and anti-CD40 (MR1) antibodies (BD Biosciences) as described (5, 6). Data were collected on a FACSCalibur (BD Biosciences) and analyzed with Flowjo (Tree Star). To detect intracellular P-p38 by flow cytometry, BM-DCs stimulated in culture for 30 mins were then fixed for 10 mins with 4% paraformaldehyde (Electron Microscopy Sciences), permeabilized with Permeabilzation Buffer II (BD Biosciences) and then incubated with anti-P-p38 rabbit mAb (Cell Signaling) followed by goat anti-rabbit IgG–FITC (Santa Cruz Biotechnology), according to manufacturer’s instructions. Data were collected on a FACSCalibur (BD Biosciences) and analyzed with Flow Jo (Tree Star). In phospho-flow studies, DCs were also gated on MHC II^hi subset (M5/114.15.2, anti-I-A/I-E mAb-Alexa-Fluor^R 647) (Biolegend) which identifies those most responsive to TLR stimulation (5).

ELISA

To assess binding interactions of the IgM with necrotic cell extracts, ELISA high binding capacity half-area plates (Costar) precoated with anti-AC NAb IgM (3 μg/ml) were blocked with PBS 1% BSA and incubated for 1 hr at RT with shaking with the indicated dilutions of primary necrotic cell extracts at6.25% combined either with the anti-chromatin IgG, the isotype control (2 μg/ml), the antibodies alone, or with phosphorylcholine (PC)-BSA-biotin. Plates were washed, then developed with goat anti-mouse IgG γ-specific Ab HRP-conjugate (Jackson Immunoresearch) to detect bound IgG, or with streptavidin-polyHRP80 (Fitzgerald Industries International) to detect PC-biotin. Alternately, plates were pre-coated with goat anti-mouse IgG (Southern Biotech), blocked and then incubated with the necrotic cell extract/IgG mixtures as above. Plates were then washed and incubated anti-AC NAb IgM (at 2 μg/ml) for 1 hr at RT with shaking, then washed and developed with goat anti-mouse IgM (μ-specific) HRP-conjugate (Jackson Immunoresearch).

Cytokine ELISAs

Cytokines were quantitated using matched capture and detection antibodies to IL-6 (Biolegend) and TNF-α (R&D Systems), as previously described (5)

Statistics

Two-tailed t tests were performed with Prism software (Graphpad). P< 0.05 was considered significant.

Results

NAb IgM inhibits lupus anti-chromatin IgG-immune complex-mediated responses

In recent studies, AC-specific NAbs were shown to block the inflammatory responses of macrophages and DCs to a range of TLR agonists, and could also suppress autoimmune disease induced by infusions of pathogenic IgG-autoantibodies (5). We therefore sought to evaluate whether these NAbs could affect autoantibody-IC-mediated DC responses. At first, we refined culture conditions for optimal activation by chromatin– IC. In pilot studies, based on viability cell counts we found that during overnight cultures a substantial number of purified BM-DCs (10-25%) died, with changes characteristic of apoptotic death ((6) and not shown). Based on the notion that DCs dying in culture may themselves provide a source of autoantigens for IC formation (1, 17), we directly added IgG anti-chromatin to DC cultures, and found this induced substantial IL-6 responses, but only limited amounts of TNF-α (Supplementary Figure 1A). In contrast, when the antibody was first pre-incubated with primary necrotic cell extracts much higher levels of TNF-α were detected, while the IL-6 levels were lower than that induced by addition of antibody directly to culture (Supplementary Figures 1A & B). To examine the potential role for Fc-receptor involvement, we added the inhibitory anti-FcγRII/III antibody (2.4G2) (18) and documented suppression of IC-mediated cytokine induction (Supplementary Figures 1B & D). These findings thereby confirmed the requirement of FcγR for anti-chromatin autoantibody-mediated induction of proinflammatory cytokines in our system.

In an earlier report, the prototypic T15/E06 anti-AC NAb IgM was shown to spontaneously form immune complexes with ACs and inhibit TLR responses of macrophages and DCs by a mechanism dependent on interactions with the PC-antigen binding site (5, 6). In the current studies, we found that this NAb IgM significantly inhibited chromatin IC-elicited DC production of IL-6 and TNF-α in a dose-dependent manner (Figures 1A & B), while isotype control had no effect (Figures 1A & B). In side-by-side comparisons, dexamethasone, a potent glucocorticoid inhibitor of TLR responses, and the NAb IgM at <40 μg/ml displayed comparable levels of inhibition (Figures 1C & D). Substantial, but lower, levels of this cytokine was induced by chromatin-ICs compared to the synthetic TLR9 agonist, CpG ODN B at 0.5 μM (Figure 1E). In agreement with previous findings, the NAb IgM inhibited IL-6 production induced by the CpG ODN by over 90%. Extracts from primary necrotic cells contain various damage-associated molecular patterns (DAMPs) that can activate TLRs (19). Indeed, we observed a small but significant stimulatory effect of the primary necrotic cell extracts on DCs, which was significantly inhibited by the anti-AC NAb IgM (Figure 1F).

Figure 1.

Anti-AC NAb IgM inhibits the induction by chromatin IC of proinflammatory cytokines. (A & B), To test the effect of the anti-AC NAb IgM on IC-stimulation of BM-DCs, cells were pre-incubated with indicated concentrations of NAb or isotype control IgM, before addition of anti-chromatin IgG directly to culture wells to elicit IL-6 (A) or chromatin ICs pre-formed by incubating this Ab with primary necrotic cell extracts to elicit TNF-α (B). (C & D), To compare the level of inhibition, IgMs or dexamethasone were added at the indicated concentrations to BM-DCs before addition of anti-chromatin IgG (<b>C</b>) or preformed chromatin-IC (D). (E), Comparison of IL-6 induction by anti-chromatin IgG or synthetic CpG Type B ODN and effects of anti-AC NAb or isotype IgMs. All conditions were supplemented with C1q. (F), Comparison of TNF-α induction by primary necrotic cell supernatants alone or in immune complex with anti-chromatin IgG, and effect of anti-PC IgM. Values are mean ± SEM of 3 replicate cultures. A-D are representative of six independent experiments. E-F are representative of two independent experiments. Unless indicated, comparisons were made to anti-chromatin IgG or anti-chromatin IgG/primary necrotic cell extract-stimulated replicates, *P < 0.05, ** P <0.005, ***P <0.0005.

NAb IgM inhibits lupus anti-RNA IgG-immune complex mediated responses

To test the relevance of this inhibitory pathway on responses induced by another type of IC, we performed equivalent studies with the anti-RNA IgG, BWR4, but found that neither direct addition to cultures, nor pre-incubation with necrotic cell extracts, formed ICs with stimulatory activity for DCs (data not shown). As we postulated that this reflected the breakdown of relevant nucleic acids by RNAses in culture, we therefore next generated ICs by pre-incubation of anti-RNA IgG with ssRNA40, a GU-rich synthetic phosphorothioate oligonucleotide, in the presence of an RNAse inhibitor. While neither the addition of the ssRNA40 nor the antibody alone had an effect, the preformed RNA-ICs elicited significant levels of IL-6 (Figure 2B), and anti-FcγRII/III treatment confirmed that this stimulation involved IC/FcγR-mediated mechanisms (Figure 2C). Importantly, the NAb IgM, as well as dexamethasone, but not IgM isotype control, again significantly inhibited IL-6 production elicited by RNA-ICs (Figure 2C).

Figure 2.

Anti-AC NAb IgM inhibits RNA-IC-induced cytokine production and costimulatory molecule up-regulation (A), BM-DC were cultured with ssRNA40, anti-RNA IgG or RNA-ICs formed by preincubation of anti-RNA IgG and ssRNA40. (B), Fc block (anti-CD16/32 mAb) was added to culture wells prior to treatment with RNA-ICs. (C), To test the capacity for inhibitory properties, indicated concentrations of anti-AC NAb IgM, an irrelevant isotype control IgM, or dexamethasone were added to culture wells before addition of pre-formed RNA-ICs. All wells were supplemented with C1q. Supernatants collected at 24 hrs were assayed for IL-6 (A-C). (D-F), BM-DC were stimulated for 24 hrs, then CD11c+ cells were evaluated for CD86 and CD40 expression. (D), % CD11c+ cells that are CD86hi (top) and CD40hi (bottom). (E & F), MFI values for CD86 (E) and CD40 (F) expression on the total CD11c+ DC population are shown. Values represent mean ± SEM of 3 replicate cultures. A-C are representative of three independent experiments. Comparisons were to RNA IC-stimulated replicates * P < 0.05, ** P <0.005, *** P <0.0005.

Costimulatory communication between APCs and T-cells is required to generate robust immune responses and has also been implicated in autoimmune pathogenesis. We therefore tested the effect of these IgG-ICs on DCs and found that stimulation with RNA-ICs increased DC membrane expression of both CD86 and CD40 (Figure 2D). Here again, in a dose-dependent manner, the anti-AC NAb IgM significantly inhibited both CD86 (Figure 2D & E) and CD40 expression (Figure 2D & F) on the surface of conventional DCs.

IgM anti-AC NAb and anti-chromatin IgG do not compete in their binding interactions

We next examined whether the inhibitory activities of the NAb might be related to a potential ability to interfere with the binding of prototypic pathogenic anti-chromatin IgG lupus autoantibody to its antigens in the experimental culture system. Because both antibodies bind apoptosis-associated antigens, we first sought to rule out that these autoantibodies might display competition in their binding interactions. We studied freshly isolated or dexamethasone-treated apoptotic thymocytes by flow cytometry and microscopy, and found that the NAb IgM, the anti-chromatin IgG, and the isotype controls, all failed to bind healthy thymocytes (Annexin-V-7AAD-) when examined either by flow cytometry (Figure 3A) or microscopy (Figure 4A and data not shown). In contrast, both the NAb IgM and the anti-chromatin IgG, but not the isotype controls, predominantly recognized a subset of late apoptotic cells (Annexin-V+ 7-AAD+) (Figures 3C & 4C). Importantly, binding of the lupus IgG antibody to this cell population was unaffected by the presence of the NAb IgM (Figure 3D), indicating that the IgM did not impede binding of the anti-chromatin IgG to its target antigen. By flow cytometry, both antibodies bound the 7AAD+ late apoptotic cell population (Figure 3C & D). However, consistent with earlier reports that the PC neo-antigens recognized by the NAb IgM can be exposed early in the apoptotic death process (4), when ACs from these culture conditions were examined by microscopy, the anti-AC NAb IgM bound to a subset of early apoptotic cells (Figure 4B and Supplementary Figure 3A) in a cell membrane-associated binding pattern (4, 6). In cells displaying pronounced nuclear condensation, a marker of a later stage apoptotic death, we detected binding by both the anti-AC NAb IgM and the anti-chromatin IgG (Figures 4C and Supplementary Figure 2B). On these ACs, the anti-chromatin IgG antibody bound in a linear fashion along the cell membrane, indicating that nuclear material had translocated to the outside of the cell, which has previously been reported for other anti-DNA antibodies (20). Most importantly, these two different antibodies bound to discrete and non-overlapping sites within the same ACs (Figure 4C and Supplementary Figure 2B).

Figure 3.

Binding of IgG anti-chromatin autoantibody to apoptotic thymocytes is not blocked by the anti-AC NAb IgM. C57BL/6 thymocytes were either freshly isolated or cultured with dexamethasone (1μM) for 8 hrs to induce apoptosis. (A-C), Binding of anti-AC NAb IgM and anti-chromatin IgG (black line) or isotype controls (gray line) to freshly isolated healthy (A), early-apoptotic (B) or secondary-necrotic (C) thymocytes was examined by flow cytometry. (D), To investigate whether either antibody blocked binding of the other, binding of anti-chromatin IgG in the absence (gray line) or presence (black line) of anti-AC NAb IgM (D, left) and binding of NAb IgM in the absence (gray line) or presence (black line) of anti-chromatin antibody (D, right) was examined.

Figure 4.

Anti-AC NAb IgM and anti-chromatin IgG recognize discrete non-overlapping sites in apoptotic cells. (A-C), A glycerol-mounted suspension of freshly isolated (A) or dexamethasone-treated early apoptotic (B) or late apoptotic (C) thymocytes was examined for binding to anti-chromatin antibody (green) and anti-AC NAb IgM (red) by fluorescent microscopy. Scale bar = 2 μM. Representative images from two of 3 experiments are shown.

As an alternate approach to understand the potential physical relationships between the autoantigenic sites recognized by this NAb IgM and the lupus-associated IgG autoantibodies, we performed immunoassays on ICs that were formed by premixing anti-chromatin IgG with primary necrotic cell extracts (as described in Figure 1B). In the first approach, the NAb IgM was immobilized on wells of ELISA plates and incubated with the necrotic cell/anti-chromatin mixtures (or controls). With the use of a final IgG-specific detection step, we found no evidence of detectable levels of the IgG lupus anti-chromatin antibody, which were associated with necrotic cell material that remained bound to the NAb IgM (Supplementary Figure 3A). By contrast, as a positive control, the NAb IgM was shown to strongly bind to a biotinylated form of PC-albumin conjugate (Supplementary Figure 3A).

We also performed ELISA studies, using microtiter wells coated with anti-IgG antibodies, to which the necrotic cell/anti-chromatin antibody-ICs (or isotype control mixtures) were added. However, when the anti-AC NAb IgM antibody was later added, we were unable to detect an association of this IgM antibody (Supplementary Figure 3B) with the plate-bound anti-chromatin IgG, or with necrotic cell debris bound by the anti-chromatin IgG. Collectively, these data indicate that the anti-AC NAb IgM did not display detectable binding interactions with the anti-chromatin IgG antibody or IgG-bound autoantigen. Taken together, these findings indicate that the NAb IgM is unlikely to compete with the chromatin IgG and thus impede the formation of IgG-nucleic immune complexes.

NAb IgM inhibits p38 MAPK activation by lupus autoantibody-ICs

Inflammatory responses induced by stimulation via endosomal TLRs have been reported to involve the Mitogen Activated Protein Kinase (MAPK) cascade (reviewed in (21)), and in particular nucleic acid-ICs that cross-link membrane-associated BCR have been shown to activate the primary MAPK, p38 in B cells (22). As expected, in our studies we found that both the anti-chromatin ICs and CpG ODN induced p38 MAPK phosphorylation, when examined by flow cytometry (Figure 5). Overall, compared to stimulation with the CpG ODN, we found that a smaller proportion of DCs were responsive to IC induced P-p38 upregulation. Importantly, our studies documented that the anti-AC NAb IgM significantly inhibited both chromatin-IC as well as CpG-induced P-p38 upregulation. These inhibitory effects were demonstrated when DCs were examined altogether, and differences were more marked in MHC II^hi DC subsets with increased TLR responses (Figure 5).

Figure 5.

Anti-AC NAb IgM inhibits anti-chromatin IgG- and CpG-ODN-mediated induction of total intracellular P-p38. BM-DCs were cultured in complete media supplemented with C1q, and treated with isotype or anti-AC NAb IgM before stimulation with CpG ODN B (0.5 μM) or anti-chromatin IgG (8 μg/ml) for 30 mins. Total P-p38 was detected by phospho-flow as described (see methods). (A), Geometric MFI of P-p38 in size-gated, MHC II high population is shown. (B), MFI values from triplicate cultures in 2 independent experiments are shown. Values represent mean ± SEM. * P <0.05; *** P <0.0005. Results are representative of 3 independent experiments.

As transduction pathways triggered by activated p38 MAPK require specific spatiotemporal subcellular localization of the phosphorylated factor, we also examined these responses by immunofluorescence microscopy (Figure 6). These studies confirmed that exposure to the NAb IgM resulted in suppression of nuclear localization of the phosphorylated form of p38. Taken together, these studies confirm that anti-AC NAb-mediated inhibition of DC activation, based on induced surface expression of MHC and co-stimulatory molecules and production of proinflammatory cytokines, was associated with inhibition of p38 activation and nuclear localization

Figure 6.

Anti-AC NAb IgM reduces nuclear P-p38 in anti-chromatin IgG and CpG-ODN stimulated DCs. BM-DCs cultured in complete media with C1q and either isotype control or anti-AC NAb IgM before stimulation with CpG ODN B (0.5 μM) or anti-chromatin IgG (8 μg/ml) for 30 mins. (A), DCs attached to poly-L-lysine coated coverslips were stained for P-p38 FITC (green) and Hoechst 34580 (blue) to identify the nucleus and imaged by florescent microscopy as described (methods). DIC images were overlaid with P-p38 FITC and Hoechst 34580 using Photoshop software. Scale bar = 10 μM (B), Mean of p-p38 FITC fluorescence intensity in the nucleus was quantified for 30 cells in 2-3 randomly selected fields per condition using an Image J Macro, as described in methods. Values represent mean ± SEM. *** P <0.0005. Representative of 3 independent experiments.

Discussion

In the current study, we demonstrated that a prototypic B-1 cell derived NAb, reactive with a determinant on AC membranes, can inhibit the capacity of both RNA- and chromatin-containing ICs to induce the production of pro-inflammatory cytokines, and it also blocked the maturation and differentiation of myeloid DCs based on expression of key co-stimulatory molecules involved in APC function. We also show that the mechanism of inhibition correlated with down-regulation of pro-inflammatory MAPK signaling. Moreover the potency of inhibition by the NAb was comparable in functional activity to a glucocorticoid, which is the most common class of agents used to treat inflammatory conditions.

In earlier studies, AC-specific regulatory NAbs were shown to enhance AC clearance and also blunt TLR-induced inflammation to purified ligands (5). Preincubation of the NAb with a PC-protein conjugate both blocked the binding of ACs and the inhibitory properties that were induced by these complexes. These findings supported the role of PC epitope-mediated interactions with AC membranes in these NAb regulatory activities (5, 6, 13).

Unbiased microarray surveys have previously suggested that this AC-specific NAb IgM is neither polyreactive nor does it bind to RNA- or DNA-containing ligands (5). In the current studies, we used three approaches to evaluate the potential relationship between self-antigens recognized by the NAb IgM and the lupus anti-chromatin IgG; flow cytometry, immunofluorescence microscopy, and ELISA. Our investigations found no evidence that the mechanism of action of the NAb IgM involves competitive inhibition with binding of IgG-autoantibodies to nuclear antigens. Furthermore, microscopy findings demonstrated that these two different classes of antibodies bound to discrete targets within apoptotic cells. We also investigated whether another antigen source, primary necrotic cell extracts, contained ligands for the NAb IgM by ELISA. Yet, in these studies we found no appreciable binding of the NAb to components of the necrotic cell extracts. Importantly, our signaling studies demonstrated that the influence on the stimulatory properties of pathogenic lupus autoantibody-ICs involved the inhibition by the NAb IgM of chromatin-IC-mediated induction of p38 MAPK phosphorylation. Taken together, the current evidence further supports the hypothesis that the regulatory activities of this AC-specific NAb, and similar antibodies in the circulation of neonatal and adult mice, arise from the capacity to facilitate complex interactions at the AC-DC interface, which can blunt TLR-induced inflammation via an inhibitory signaling pathway that antagonizes MAP kinase activation (((5, 6)) & manuscript submitted).

In health, IgM serves both as a membrane-receptor on B cells, and as a secreted primarily intravascular polymeric IgM molecule. Mice deficient in secreted IgM are more susceptible to infection, and are also prone to the development of IgG-autoantibodies and autoimmune disease (reviewed in (7)). Furthermore, lupus patients with reduced levels of anti-AC NAb IgM have more clinically active SLE and such individuals may also be at increased risk of cardiovascular events (11,12).

Nucleic acid-ICs are postulated to be key drivers in the self-reinforcing inflammatory cascades in SLE (1). Earlier evidence that a NAb can enhance AC clearance suggests that these IgM antibodies may in part contribute to immune homeostasis by reducing the formation of inflammatory ICs. In the current studies, we confirmed that the inhibitory activity of the NAb IgM did not correlate with a competition for the binding of the IgG-lupus autoantibody to its nuclear-associated antigen, even after translocation of these self-antigens during apoptotic cell death. Furthermore, our findings instead provide the first in vitro demonstration that this class of NAb to apoptotic cell membrane-associated determinants can directly antagonize the induction of an inflammatory signaling pathway by lupus IgG autoantibody-nucleic acid IC, suggesting that such regulatory functions could also exist in vivo. Overall, in conjunction with our findings, these studies support emerging evidence that NAbs to AC-membrane determinants, which are abundantly represented in the natural antibody repertoire from birth, may oppose the pathogenic roles of some IgG autoantibody-ICs, and provide another mechanism by which these regulatory NAbs can help to reinforce or in some cases reestablish immune homeostasis.