ARTHRITIS & RHEUMATISM Vol. 58, No. 4, April 2008, pp 1107–1115 DOI 10.1002/art.23407 © 2008, American College of Rheumatology
Requirement of Toll-like Receptor 7 for Pristane-Induced Production of Autoantibodies and Development of Murine Lupus Nephritis Emina Savarese,1 Christian Steinberg,1 Rahul D. Pawar,2 Wolfgang Reindl,1 Shizuo Akira,3 Hans-Joachim Anders,2 and Anne Krug1 Objective. The detection of high titers of antibodies against small nuclear ribonucleoproteins (snRNP) is a diagnostic finding in patients in whom systemic lupus erythematosus (SLE) is suspected. Endogenous RNA molecules within snRNP trigger Toll-like receptor 7 (TLR-7) activation in B cells and dendritic cells, leading to anti-snRNP antibody production, which is associated with the development of immune complex nephritis in SLE. The purpose of this study was to investigate the role of TLR-7 in anti-snRNP antibody production and renal disease in SLE induced by an exogenous factor in the absence of genetic predisposition, using the pristane-induced murine lupus model. Methods. Serum autoantibodies, IgG isotypes, and cytokine levels in pristane-treated wild-type and TLR-7–deficient mice were analyzed by enzyme-linked immunosorbent assay. Histopathologic changes in
mouse kidneys were determined by light immunofluorescence microscopy. Cell subsets in splenocytes and peritoneal lavage cells from the mice were examined by flow cytometry. Results. We found that anti-snRNP antibody production induced by pristane treatment was entirely dependent on the expression of TLR-7, whereas anti– double-stranded DNA antibody production was not affected by a lack of TLR-7. Impaired anti-snRNP antibody production in TLR-7–deficient mice was paralleled by lower levels of glomerular IgG and complement deposits, as well as less severe glomerulonephritis. Conclusion. TLR-7 is specifically required for the production of RNA-reactive autoantibodies and the development of glomerulonephritis in pristane-induced murine lupus, a model of environmentally triggered SLE in the absence of genetic susceptibility to autoimmunity. Specific interference with TLR-7 activation by endogenous TLR-7 ligands may therefore be a promising novel strategy for the treatment of SLE.
Presented by Ms Savarese in partial fulfillment of the requirements for a PhD degree, Technical University Munich, Munich, Germany. Ms Savarese’s work was supported by German Research Foundation grant KR2199/1-3. Mr. Steinberg’s work was supported by German Research Foundation grant SFB 455. Dr. Anders’ work was supported by German Research Foundation grants AN372/8-1 and GRAKO 1202. Dr. Krug’s work was supported by German Research Foundation grant KR2199/1-3 and SFB 455. 1 Emina Savarese, Dipl. Biol., Christian Steinberg, Wolfgang Reindl, MD, Anne Krug, MD: Department of Internal Medicine II, Klinikum Rechts der Isar, Technical University Munich, Munich, Germany; 2Rahul D. Pawar, MPharm, Hans-Joachim Anders, MD: Medizinische Poliklinik, Ludwig Maximilans University, Munich, Germany; 3Shizuo Akira, MD, PhD: Research Institute for Microbial Diseases, Osaka University, Osaka, Japan. Dr. Anders has received consulting fees, speaking fees, and/or honoraria (more than $10,000) from Noxxon Pharma Ltd., Berlin, Germany. Address correspondence and reprint requests to Anne Krug, MD, II Medizinische Klinik, Klinikum Rechts der Isar, Technical University Munich, Trogerstrasse 32, Room 1.04, D-81675 Munich, Germany. E-mail:
[email protected]. Submitted for publication September 14, 2007; accepted in revised form December 14, 2007.
The immunologic hallmark of systemic lupus erythematosus (SLE) is the presence of high levels of circulating antinuclear autoantibodies, which develop several years before manifestation of the disease and which are thought to be critically involved in its pathogenesis (1). Recent data indicate that Toll-like receptors (TLRs) recognizing self nucleic acids may be critically involved in breaking peripheral tolerance against nuclear antigens and allowing the generation of a destructive autoimmune response in SLE. By activating Tolllike receptor 9 (TLR-9) and TLR-7/8, endogenous DNA and RNA sequences contained within nuclear autoantigens in circulating autoimmune complexes act as “autoadjuvants” for efficient priming and boosting of the autoimmune response in SLE (2). By simultaneous 1107
1108
engagement of B cell receptor and TLR in B cells, the DNA- or RNA-containing autoantigens directly boost the proliferation of autoreactive B cells and their differentiation into plasma cells, producing pathogenic classswitched antibodies against nuclear antigens (3–6). Full activation and differentiation of B cells to plasma cells is additionally supported by dendritic cells (DCs). DCs are also activated via TLR-9 by DNAcontaining immune complexes and via TLR-7 by RNPcontaining immune complexes, leading to the production of proinflammatory mediators (interleukin-6 [IL-6], tumor necrosis factor ␣, IL-12, and chemokines) as well as the up-regulation of costimulatory and class II major histocompatibility complex molecules (7–10). Plasmacytoid DCs specifically respond with the release of type I interferon (IFN), an amplifier of autoimmune responses, which promotes DC maturation, T cell activation, as well as differentiation of B cells to autoantibodyproducing plasma cells (8–11). Type I IFN specifically enhances the response to endogenous TLR-7 ligands by up-regulating TLR-7 expression in B cells and DCs (4,9,12). Results of recent in vivo studies provide evidence of a critical role of TLR-7 in the generation of anti–small nuclear RNP (anti-snRNP) antibodies and the development of glomerulonephritis in mouse models of spontaneous SLE. Overexpression of TLR-7 in mice carrying the Yaa mutation leads to enhanced anti-snRNP autoantibody responses, which are associated with more severe disease in 2 mouse models of SLE (13,14). Furthermore, anti-snRNP antibody production and disease activity were found to be reduced in TLR-7– deficient mice backcrossed to the MRL/Mplpr/lpr or the 564Igi transgenic mouse strains, which spontaneously develop lupus (15,16). In contrast, TLR-9–deficient MRL/Mplpr/lpr mice developed more severe disease, suggesting a so far unexplained dominant protective function of TLR-9 (15). It is still unclear how TLR-7 influences autoantibody production and disease development in lupus syndromes induced by environmental factors, which are also known to trigger SLE in humans. To investigate the requirement of TLR-7 for induction of SLE-like disease in genetically unaltered mice without complex preexisting immunologic derangements, we chose to study the pristane-induced mouse model of SLE, comparing TLR7–deficient and C57BL/6 wild-type (WT) mice. Intraperitoneal injection of the hydrocarbon oil pristane into several different inbred strains of mice (including C57BL/6) initially leads to the development of a chronic granulomatous peritonitis involving the
SAVARESE ET AL
formation of ectopic lymphoid tissue, a phenomenon that has also been described in several autoimmune diseases (17). Pristane-induced apoptosis of peritoneal cells provides an abundant source of nuclear antigens presented in the context of chronic inflammation (18). Mice injected with pristane sequentially produce antibodies against single-stranded DNA, double-stranded DNA (dsDNA), Sm, snRNP, and Su autoantigens, similar to human lupus patients, and develop immune complex–mediated glomerulonephritis at ⬃6 months after pristane injection, followed by the development of arthritis at later time points (19). Pristane-induced SLE develops in the BALB/c and C57BL/6 mouse strains, which are not genetically susceptible to autoimmune disease. Thus, the model resembles drug-induced lupus occurring in humans. In contrast to drug-induced lupus, however, pristane-injected mice develop a broader repertoire of autoantibodies and irreversible organ damage. We investigated the requirement of TLR-7 for the production of anti-snRNP/Sm antibodies and the development of glomerulonephritis in mice with pristane-induced lupus, which serves as a model of SLE disease triggered by environmental factors in the absence of genetically determined susceptibility to SLE. MATERIALS AND METHODS Mice. C57BL/6 mice were purchased from Harlan Winkelmann (Borchen, Germany). The TLR-7⫺/⫺ mice (backcrossed to the C57BL/6 background for at least 7 generations) have been described elsewhere (20). Experiments were performed in accordance with the German animal care and ethics legislation and were approved by the local government authorities. At 3 months of age, C57BL/6 WT mice (14 females) and TLR-7⫺/⫺ mice (9 females and 4 males) were injected intraperitoneally with a single dose of 500 l of pristane (2,6,10,14-tetramethylpentadecane; Sigma-Aldrich Chemie, Steinheim, Germany). No significant differences were observed between the male and female mice in the TLR-7⫺/⫺ group with regard to disease activity, antinuclear antibody production, or immunologic parameters. Detection of serum autoantibodies and IgG isotypes. Antinuclear antibodies were detected by incubating serum samples at a dilution of 1:20 on HEp-2 cell–coated slides (BioSystems, Barcelona, Spain) followed by fluorescein isothiocyanate (FITC)–conjugated goat anti-mouse IgG antibody (Invitrogen, Karlsruhe, Germany) and analyzed as described elsewhere (21). For the Crithidia luciliae assay, sera were diluted 1:20 and applied to slides coated with fixed C luciliae (Bio-Rad, Redmond, WA). Binding to C luciliae kinetoplast was detected with FITC-conjugated goat anti-mouse IgG (1:1,000 dilution; Invitrogen, San Diego, CA). Staining with 4⬘,6-diamidino-2-phenylindole (Vector, Burlingame, CA) allowed colocalization with kinetoplast dsDNA. Kinetoplast
REQUIREMENT OF TLR-7 FOR PRISTANE-INDUCED LUPUS
staining intensity was quantified using a semiquantitative scoring system (0–4 scale). For detection of anti-dsDNA antibodies by enzymelinked immunosorbent assay (ELISA), MaxiSorp plates (Nunc, Wiesbaden, Germany) were coated with 0.1% poly-L-lysine (Trevigen, Gaithersburg, MD), followed by dsDNA (calf thymus DNA; Sigma-Aldrich Chemie) in phosphate buffered saline at 3 g/ml. For detection of anti-snRNP/Sm antibodies by ELISA, serially diluted serum samples were added to Nunc MaxiSorp plates that had been coated with 5 g/ml of U1 snRNP (Arotec Diagnostics, Wellington, New Zealand). Bound anti-dsDNA and anti-snRNP/Sm antibodies were detected by incubation with horseradish peroxidase–conjugated goat anti-mouse IgG (heavy and light chain; 1:500 dilution (SouthernBiotech, Birmingham, AL). ABTS (Roche Diagnostics, Mannheim, Germany) was used as substrate. Purified monoclonal anti-Sm antibody (Y12; kindly provided by Iain Mattaj, EMBL, Heidelberg, Germany) and antinucleosome antibody (PR1-3; kindly provided by Marc Monestier, Temple University School of Medicine, Philadelphia, PA) were used to create standard curves. For quantification of total IgG isotypes (IgG1, IgG2c, IgG2b, IgG3), we used ELISA kits from Bethyl Laboratories (Montgomery, TX). Kidney histopathologic assessment and immunofluorescence analysis. Formalin-fixed paraffin-embedded kidney sections (2 m) were stained with periodic acid–Schiff. The severity of kidney disease was graded by an observer (H-JA) who was blinded to the genotype of the mice; a glomerulonephritis activity score (range 0–24) developed for the assessment of lupus nephritis in humans (22) was used. For quantitative analysis, glomerular cells in 10 cortical glomeruli per section were counted. Cryosections were stained with FITClabeled goat anti-mouse IgG antibody (1:100 dilution; Jackson ImmunoResearch, West Grove, PA) or FITC-labeled goat anti-mouse C3c (1:200 dilution; Nordic, Tilburg, The Netherlands). Semiquantitative scoring of IgG and complement C3 deposits (0–3 scale) was performed on 15 cortical glomerular sections as described elsewhere (21). Flow cytometry. Splenocytes and peritoneal lavage cells were stained for fluorescence-activated cell sorter (FACS) analysis using FITC-labeled anti-CD4, phycoerythrin (PE)–labeled anti-CD4, allophycocyanin (APC)–labeled antiCD3, PE-labeled anti-CD8, PE-labeled anti-CD44, PE-labeled anti-CD69, FITC-labeled anti-CD25, FITC-labeled antiCD19, APC-labeled anti-B220, APC-labeled anti-CD11c, and PE-labeled anti-CD86 (BD Biosciences, Heidelberg, Germany). Plasmacytoid DCs were stained with FITC-labeled anti-BST2 and biotinylated anti–Siglec H antibody (kindly provided by Marco Colonna, Washington University School of Medicine, St. Louis, MO) followed by APC-labeled Streptavidin (Molecular Probes, Eugene, OR). Propidium iodide was added to exclude dead cells from analysis. Cells were analyzed with a FACSCalibur flow cytometer (BD Biosciences, San Jose, CA). Splenocyte stimulation, cytokine ELISAs, and quantitative reverse transcription–polymerase chain reaction. Splenocytes were isolated and cultured in complete medium at a density of 3 ⫻ 106 cells/ml for 24 hours, as described previously (23). TLR-7 ligand R848 (3 M; InvivoGen, San Diego, CA), lipopolysaccharide (LPS; 1 g/ml) (Sigma-Aldrich Chemie), and CpG 2216 (0.5 M; MWG Biotech, Martinsried, Germany) were used as stimuli. Concentrations of IL-6 and IL-12p40 were
1109
Figure 1. Serum antinuclear antibodies in pristane-treated wild-type (WT) and TLR-7⫺/⫺ mice. Antinuclear antibodies in the sera of 14 WT mice and 13 TLR-7⫺/⫺ mice at 7 months after pristane injection were determined by immunofluorescence staining on HEp-2 cells. A, Representative photomicrographs of WT and TLR-7⫺/⫺ mouse sera staining (original magnification ⫻ 400). B, Frequency of specific staining patterns in WT and TLR-7⫺/⫺ mouse sera. ⴱ ⫽ P ⬍ 0.05 versus WT mice, by Fisher’s exact test. Homogen ⫽ homogeneous.
measured by ELISA using matched antibody pairs and standards (from BD Biosciences), following a standard ELISA protocol. For quantification of gene expression levels, RNA was isolated from spleen tissue using TRIzol (Invitrogen), treated with DNase (Fermentas, St. Leon-Rot, Germany), and reverse transcribed using Superscript III and oligo(dT) primer (Invitrogen) to generate complementary DNA. TaqMan polymerase chain reaction was performed using commercial primers and probes (Applied Biosystems, Foster City, CA) or primer sequences generated with the use of Primer Express software. Relative expression of messenger RNA (mRNA) was calculated by normalization to -actin or hypoxanthine guanine phosphoribosyltransferase expression using the ⌬⌬Ct method. Statistical analysis. Results are shown as the mean ⫾ SD. Data were analyzed using Student’s unpaired 2-tailed t-test for comparison between 2 groups (WT versus TLR-7⫺/⫺ mice). For comparison of the frequency of specific antinuclear antibody staining patterns between 2 groups, Fisher’s exact test was used.
RESULTS Selective impairment of anti-snRNP autoantibody production in the absence of TLR-7 in the pristane-induced mouse model of SLE. Mice treated with pristane produce significant amounts of antinuclear
1110
SAVARESE ET AL
Figure 2. Quantification of anti–small nuclear RNP (anti-snRNP)/Sm and anti–double-stranded DNA (anti-dsDNA) antibodies in the sera of pristane-treated wild-type (WT) and TLR-7⫺/⫺ mice. Sera from 14 WT and 13 TLR-7⫺/⫺ mice were examined. A, Anti-snRNP antibodies in individual WT and TLR-7⫺/⫺ mouse sera at different time points after pristane injection, as determined by enzyme-linked immunosorbent assay (ELISA). B, Anti-dsDNA antibodies in WT and TLR-7⫺/⫺ mouse sera at 7 months after pristane injection, as determined by ELISA (left) and by Crithidia luciliae kinetoplast staining (score 0–4) (right). Values are the mean and SD. C, IgG isotypes in WT and TLR-7⫺/⫺ mouse sera at 7 months after pristane injection, as determined by ELISA. Values are the mean and SD. ⴱ ⫽ P ⬍ 0.05 versus WT mice, by Student’s t-test.
antibodies, including anti-snRNP/Sm and anti-dsDNA antibodies. To investigate the requirement of TLR-7 for antinuclear antibody production in this model, antinuclear antibodies in the sera of WT and TLR-7⫺/⫺ mice were detected by HEp-2 cell staining at 7 months after pristane injection and were classified according to their staining pattern (Figure 1). All of the 14 sera from pristane-treated WT mice produced inhomogenous or speckled nuclear staining of different intensities and patterns. Distinct and dense speckled staining patterns, which are typically produced by anti-snRNP/Sm autoantibodies (among others), were observed in the sera of 9 of the 14 WT mice (Figure 1A). Sera from the TLR-7⫺/⫺ mice produced mainly homogenous nuclear staining, without speckled staining patterns (Figure 1A). Only 4 of the 13 sera from TLR-7⫺/⫺ mice generated detectable speckled nuclear staining of low density superimposed with homogenous nuclear staining in HEp-2 cells. Some homogenous nuclear staining was detectable in all of the sera from WT and TLR-7⫺/⫺ mice, whereas cytoplasmic staining was found
in 40% of the sera from the WT and TLR-7⫺/⫺ mice. Thus, specifically, the speckled nuclear staining pattern was found more frequently in WT mouse sera than in TLR-7⫺/⫺ mouse sera (Figure 1B), which could be due in part to a higher proportion of WT mice producing snRNP/Sm-reactive antibodies. To compare anti-snRNP/Sm antibody production on a quantitative level during the development of pristane-induced autoimmunity, concentrations of snRNP/Sm-specific antibodies were quantified in the sera of WT and TLR-7⫺/⫺ mice by ELISA at 2, 4, 5, 6, and 7 months after pristane injection. As shown in Figure 2A, significant amounts of anti-snRNP/Sm antibodies were detectable in 50% of the sera from WT mice at 6–7 months after pristane injection, whereas none of the TLR-7–deficient mice produced significant levels of anti–U1 snRNP/Sm antibodies at any of the time points studied (⬍25 g/ml). Anti-dsDNA antibodies, however, were produced at similar levels in WT and TLR-7⫺/⫺ mice at 7 months after pristane injection, as demonstrated by ELISA and by C luciliae kinetoplast staining
REQUIREMENT OF TLR-7 FOR PRISTANE-INDUCED LUPUS
Figure 3. Glomerular deposition of IgG and complement in pristanetreated wild-type (WT) and TLR-7⫺/⫺ mice. Deposition of A, IgG and B, complement C3c in kidney cryosections from 14 WT and 13 TLR-7⫺/⫺ mice at 7 months after pristane injection was determined by immunofluorescence staining (right) and semiquantitative analysis (0–3 scale) (left). Shown are representative photomicrographs of WT and TLR-7⫺/⫺ mouse kidney staining (original magnification ⫻ 200). Glomerular deposition scores are the mean and SD. ⴱ ⫽ P ⬍ 0.05 versus WT mice, by Student’s t-test.
(Figure 2B). Lower levels of total IgG2c, but similar levels of IgG1, IgG2b, and IgG3 were found in the sera of TLR-7⫺/⫺ mice as compared with the sera from WT mice, indicating that TLR-7 deficiency had only a minor impact on IgG class switching (Figure 2C). Taken together, these results show that in the pristane-induced lupus model, a lack of TLR-7 specifically prevents the production of anti-snRNP/Sm antibodies, whereas the production of anti-dsDNA antibody and total IgG are not affected. Role of TLR-7 in glomerular IgG deposition, complement activation, and glomerulonephritis in mice with pristane-induced lupus. Similar to human SLE, deposition of circulating autoimmune complexes and complement factors in glomerular capillaries leads to the development of glomerulonephritis in pristaneinduced experimental lupus. We therefore investigated whether the lack of substantial anti-snRNP/Sm antibody production in TLR-7⫺/⫺ mice would be accompanied by lower glomerular IgG and complement deposition, leading to less severe renal disease. The extent of glomerular IgG and complement C3c deposition was quantified using a semiquantitative scoring system in WT and TLR-7⫺/⫺ mice at 7 months after pristane injection. As shown in Figure 3, IgG and complement C3c were clearly detectable in the glomeruli of WT mice, but
1111
TLR-7⫺/⫺ mice showed significantly lower levels of glomerular IgG and complement C3c deposits. We used a histopathology score developed for the assessment of lupus nephritis in humans to quantify glomerulonephritis activity. WT mice developed glomerular disease with hypercellularity, indicating proliferation (Figure 4B, top), and occasionally showed necrosis of the glomerular capillaries (Figure 4B, bottom). Glomerulonephritis was significantly milder in TLR-7⫺/⫺ mice than in WT mice (mean ⫾ SD glomerulonephritis activity score 2.7 ⫾ 1.8 versus 8.1 ⫾ 2.8) (Figure 4A). Thus, the impaired production of snRNP-reactive autoantibodies and other autoantibodies producing speckled nuclear staining in TLR-7–deficient mice was associated with significantly less severe immune complex glomerulonephritis in mice with pristane-induced lupus. Comparable cellular activation and cytokine expression in pristane-treated WT and TLR-7–deficient mice. B lymphocytes and DCs are directly activated by endogenous TLR-7 ligands to produce antibodies, secrete cytokines, and up-regulate costimulatory molecules. Activation of TLR-7 in B cells and DCs may indirectly lead to expansion and differentiation of Th1 cells as well as suppression of regulatory T cells, which further promotes the autoimmune response. We therefore investigated whether the impaired anti-snRNP antibody production and lower disease severity in pristane-
Figure 4. Pristane-induced glomerulonephritis in wild-type (WT) and TLR-7⫺/⫺ mice. Periodic acid–Schiff–stained, paraffin-embedded kidney sections from 14 WT and 13 TLR-7⫺/⫺ mice obtained at 7 months after pristane injection were examined. A, Glomerulonephritis activity scores in WT and TLR-7⫺/⫺ mouse kidney sections (0–24 scale; see Materials and Methods for details). Values are the mean and SD. ⴱ ⫽ P ⬍ 0.05 versus WT mice, by Student’s t-test. B, Representative photomicrographs of WT and TLR-7⫺/⫺ mouse kidney sections. Glomerular hypercellularity due to proliferation (arrow, top left) and focal segmental necrosis of glomerular capillaries (arrow, bottom left) are seen in WT mice (original magnification ⫻ 200).
1112
Figure 5. Splenocyte phenotype and cytokine production in wild-type (WT) and TLR-7⫺/⫺ mice at 7 months after pristane injection. Determinations were made in 14 pristane-treated WT mice, 13 pristane-treated TLR-7⫺/⫺ mice, and 3 untreated WT mice. A, Spleens and mesenteric lymph nodes (LN) from WT and TLR-7⫺/⫺ mice were removed and weighed separately. B–D, Splenocytes were obtained from WT and TLR-7⫺/⫺ mice and stained with fluorescence-labeled antibodies against lineage-specific cell surface markers to measure the percentages of CD4⫹ T cells (T helper cells), CD8⫹ T cells (cytotoxic T lymphocytes), and B cells (CD19⫹,B220⫹) (B), CD11c⫹ dendritic cells (DCs) and Siglec H⫹,BST2 plasmacytoid DCs (PDC) (C), as well as CD44⫹,CD4⫹ T cells and CD25⫹,CD4⫹ T cells (D). E and F, The expression of early activation marker CD69 in CD4⫹ and CD8⫹ T cells (E) and of costimulatory molecule CD86 in B cells and DCs (F) is shown as the mean fluorescence intensity (MFI). G, Splenocytes from untreated WT mice and from pristane-treated WT and TLR-7⫺/⫺ mice were stimulated for 24 hours with specific Toll-like receptor ligands R848, lipopolysaccharide (LPS), CpG, or medium alone (–), and levels of interleukin-6 (IL-6) (left) and IL-12 (right) were determined by enzyme-linked immunosorbent assay. Values are the mean and SD. ⴱ ⫽ P ⬍ 0.05 versus WT mice, by Student’s t-test.
treated TLR-7⫺/⫺ mice was associated with specific changes in these different immune cell compartments or in their cytokine production capacity. At 7 months after pristane injection, spleen and lymph node weights were similar in WT and TLR-7⫺/⫺ mice (Figure 5A), and there were only minor differences in the composition of splenocytes between the 2 groups. TLR-7⫺/⫺ mice showed a lower percentage of CD4⫹ T
SAVARESE ET AL
lymphocytes and a lower expression of CD69 than did WT mice, but the difference was marginal (Figures 5B and E). The percentage of CD11c⫹ DCs (but not plasmacytoid DCs) was slightly decreased, whereas the percentage of B lymphocytes was increased, in the absence of TLR-7 (Figures 5B and C). However, no significant differences in the expression levels of the costimulatory molecule CD86 on the surface of B cells or DCs were observed (Figure 5F). The composition and activation of peritoneal lavage cells were also similar in WT and TLR-7⫺/⫺ mice (data not shown). Similarly, we found no difference in the composition and activation of splenocytes and peritoneal lavage cells during the priming phase of the autoimmune response at 6 weeks after pristane injection (data not shown). Splenocytes isolated from WT and TLR-7⫺/⫺ mice at 7 months after pristane injection (Figure 5G) or at 6 weeks after pristane injection (data not shown) produced similar amounts of IL-6 and IL-12 after stimulation with LPS or CpG 2216. TLR-7⫺/⫺ splenocytes failed to respond to TLR-7 ligand R848, as was expected. We also found no significant differences in mRNA expression of inflammatory cytokines (IL-6, IL-12, and IL-17), interferon-induced genes (interferon regulatory factor 7 and IFI204), the B cell–activating factor BAFF, the Th1 lineage regulator T-bet, or the Th2 lineage regulator GATA-3 in the spleens of WT mice compared with TLR-7⫺/⫺ mice at 6 weeks or 7 months after pristane injection (data not shown). Defective anti-snRNP antibody production and reduced disease activity in response to pristane exposure in the absence of TLR-7 did not correlate with relevant changes in immune cell activation or inflammatory cytokine production during the induction phase or the effector phase of pristane-induced lupus. Thus, TLR-7 plays a specific role in the production of anti-snRNP antibodies, which is associated with the development of lupus nephritis. DISCUSSION In this study, we investigated the requirement of TLR-7 for autoantibody production and lupus nephritis development in the pristane-induced mouse model of SLE. Our results show that activation of TLR-7 by endogenous TLR-7 ligands, such as U1 snRNA, is essential for the production of pathogenic antisnRNP/Sm antibodies and is associated with the development of immune complex nephritis in response to pristane treatment, without the need for a genetically susceptible host. TLR-7 has a highly selective function in
REQUIREMENT OF TLR-7 FOR PRISTANE-INDUCED LUPUS
the initiation and propagation of anti-snRNP/Sm antibody production in this mouse model of inducible SLE. No relevant differences were observed between TLR7⫺/⫺ and WT mice with regard to the activation of T cells, B cells, and DCs in the spleen and peritoneal cavity or the expression of inflammatory cytokines and transcription factors in the spleen. We conclude that TLR-7 plays a critical and specific role in the development of SLE induced by an exogenous factor, even in the absence of a genetic predisposition to autoimmunity. Using the pristane-induced model of SLE, we were able to very clearly show that the production of anti-snRNP/Sm antibodies is entirely dependent on TLR-7 expression. In the MRL/Mplpr/lpr model of lupus, anti-snRNP/Sm antibody production was also blocked in the absence of TLR-7, but the frequency of WT mice producing anti-snRNP/Sm antibodies was very low (15). In the study by Berland et al (16), the concentration of a specific autoantibody reacting with several nuclear autoantigens (RNA, DNA, and nucleosomes) in the 564 immunoglobulin-transgenic mice was significantly reduced, but not abrogated, in the absence of TLR-7. In our study as well as in the study by Christensen et al (15), the production of anti-dsDNA autoantibodies was not affected by TLR-7 deficiency. Thus, we conclude that TLR-7 is specifically required for the production of anti-snRNP/Sm antibody during the development of lupus. Anti-snRNP/Sm antibody production was absent in TLR-7⫺/⫺ mice from early time points on. Therefore, TLR-7 seems to be involved not only in the amplification of the anti-snRNP autoantibody response, but also in its initiation. It is not yet clear which type of antigenpresenting cells, whether B cells or DCs, is initially activated by endogenous TLR-7 ligands and effectively presents the autoantigen to initiate snRNP-reactive B cell and T cell responses. Autoreactive B cells can directly bind snRNP that is present within apoptotic material and transport it to the endosomal compartment where TLR-7 is localized (4,6). Dual activation of snRNP-reactive B cells via the B cell receptor and TLR-7 may then lead to the initial breaking of tolerance and clonal B cell expansion (4,5). It has been observed in vitro that activation of TLR-7 in plasmacytoid and conventional DCs by U1 snRNA contained within U1 snRNP requires delivery to the endosomal compartment (9,10). This is achieved naturally only by the formation of autoantibody immune complexes, which are internalized by Fc receptors (8–11). Thus, the activation of DCs by endogenous TLR-7 ligands in vivo seems to require the prior production of specific autoantibodies by B
1113
lymphocytes. Therefore, the major role of plasmacytoid DCs and conventional DCs activated by immune complexes containing endogenous TLR-7 ligands lies in the amplification of the autoimmune response, which involves type I IFN induction in plasmacytoid DCs (24). Type I IFN signaling specifically enhances the response of B lymphocytes and DCs to TLR-7 stimulation by up-regulating the expression of TLR-7, thus further amplifying the autoimmune response (4,9,12). In the pristane-induced lupus model, we observed only minor differences in the overall expansion and activation of immune cells isolated from the spleen or peritoneal cavity of TLR-7⫺/⫺ mice as compared with WT mice. In the MRL/Mplpr/lpr model, TLR-7⫺/⫺ mice showed lower lymph node and spleen weights and a reduced number of double-negative T lymphocytes and memory CD4⫹ T cells. Nonspecific activation of B cells, production of total IgG2a and IgG3, and class II MHC expression in plasmacytoid DCs were also reduced in the absence of TLR-7 (15). Differences in immune cell activation are strongly amplified in the MRL/Mplpr/lpr model, which is characterized by massive lymphoproliferation, high concentrations of serum autoantibodies, and fatal autoimmune disease. In the pristane-induced SLE model, initiation and amplification of the autoimmune response seem to be more restricted to the local environment, for example, to the newly formed lymphoid tissue within lipogranulomas induced in the peritoneal cavity by pristane treatment (17). TLR-7– dependent production of RNA-reactive autoantibodies and development of nephritis did not correlate with systemic inflammatory responses in the pristane-induced mouse model, thus supporting the specific role of TLR-7 in anti-snRNP antibody production. Our results show that low or absent antisnRNP/Sm antibody production in TLR-7–deficient mice is associated with lower immunoglobulin deposition and less severe glomerulonephritis in the pristaneinduced model of lupus. This association has also been observed in the MRL/Mplpr/lpr model of murine lupus. The ameliorating effect of TLR-7 deficiency on lupus nephritis, however, was more pronounced in the pristane-induced SLE model described in the present study than in the MRL/Mplpr/lpr model (15), possibly due to a higher percentage of WT mice in the pristane model producing anti-snRNP/Sm antibodies. In the MRL/Mplpr/lpr mouse, the presence of anti-snRNP/Sm antibodies correlated with increased immune activation and more severe disease (15). Similarly, in humans with SLE, the presence of anti-snRNP antibodies correlated with a higher level of IFN␣ in the serum and a higher
1114
disease activity index (25). Production of the RNAreactive 564 antibody in the immunoglobulin transgenic 564Igi mouse line (on the C57BL/6 background) was sufficient to induce the development of lupus nephritis, suggesting a causal relationship between RNA-reactive autoantibody immune complexes and kidney disease (16). In the present study, pristane-treated WT mice, which produced high amounts of anti-snRNP antibodies, also had high activity of glomerulonephritis. However, some of the pristane-treated WT mice, which did not produce significant amounts of anti-snRNP/Sm antibodies, showed similar levels of glomerulonephritis activity, suggesting that the presence of anti-snRNP/Sm antibodies is not necessarily required for disease development in this model. One explanation could be that in addition to autoantibodies directed against snRNP autoantigen, which were detected by our ELISA (4,9,10), autoantibodies directed against other nuclear autoantigens that contain or are associated with TLR-7–stimulating RNA sequences (e.g., Ro Y RNA, pre–transfer RNA, pre–5S RNA, RNase P, RNA helicase A, RNA polymerase III) may also contribute to disease development (10,26). It is not entirely clear at present how the presence of RNA-reactive autoantibodies is connected to the development of lupus nephritis. Glomerular deposition of pathogenic antinuclear antibody immune complexes leads to the activation of the complement system as well as the activation of tissue-resident macrophages and DCs via the Fc␥ receptor type IV (27). Thus, an inflammatory immune response is generated locally in the kidney; cytokines and chemokines, such as monocyte chemoattractant protein 1, are produced, which further recruit immune cells, including DCs, macrophages, and activated T lymphocytes. Endogenous TLR-7 ligands within deposited autoimmune complexes may also directly activate tissue-resident and recruited hematopoietic inflammatory cells in the kidney during lupus nephritis. Renal mesangial cells, however, lack TLR-7 expression and are therefore not directly activated by endogenous TLR-7 ligands (28). The results presented herein provide further evidence of a pathogenic role of TLR-7 in SLE, even in the absence of specific genetic alterations or strong susceptibility to autoimmunity. TLR-7 (and possibly TLR-8 in the human system) may therefore be an attractive target for pharmacologic intervention in the treatment of lupus. Therapeutic agents that interfere with the activation or specific downstream signaling of TLR-7 would be useful not only for the prevention of SLE development, for which clinical opportunity seldom exists, but also for
SAVARESE ET AL
the treatment of ongoing disease, because amplification of the autoimmune response by TLR-7–activated DCs and plasmacytoid DCs, as well as target organ damage induced by inflammatory cytokines and effector cells triggered by DCs, could also be targeted later in the course of the disease, when autoantibodies are already present. Proof of principle for this therapeutic strategy has recently been provided by a study using TLR-7– antagonistic oligodeoxynucleotides in the MRL/Mplpr/lpr mouse model of lupus after the onset of autoimmunity (21). Given the ability to detect antinuclear antibodies years before the manifestation of SLE (1), administration of TLR-7 antagonists may potentially prolong the disease-free time interval in predisposed individuals producing anti-snRNP/Sm antibodies. ACKNOWLEDGMENTS We thank Bernadette Grohs, Dan Draganovic, Jana Mandelbaum, and Ewa Radamoska for excellent technical assistance. We would like to thank I. Mattaj, M. Monestier, and M. Colonna for providing antibodies. AUTHOR CONTRIBUTIONS Dr. Krug had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study design. Savarese, Krug. Acquisition of data. Savarese, Steinberg, Pawar, Reindl, Anders. Analysis and interpretation of data. Savarese, Steinberg, Anders, Krug. Manuscript preparation. Savarese, Krug. Statistical analysis. Savarese, Pawar. Provision of knockout mice. Akira.
REFERENCES 1. Arbuckle MR, McClain MT, Rubertone MV, Scofield RH, Dennis GJ, James JA, et al. Development of autoantibodies before the clinical onset of systemic lupus erythematosus. N Engl J Med 2003;349:1526–33. 2. Marshak-Rothstein A. Toll-like receptors in systemic autoimmune disease. Nat Rev Immunol 2006;6:823–35. 3. Fields ML, Metzgar MH, Hondowicz BD, Kang SA, Alexander ST, Hazard KD, et al. Exogenous and endogenous TLR ligands activate anti-chromatin and polyreactive B cells. J Immunol 2006; 176:6491–502. 4. Lau CM, Broughton C, Tabor AS, Akira S, Flavell RA, Mamula MJ, et al. RNA-associated autoantigens activate B cells by combined B cell antigen receptor/Toll-like receptor 7 engagement. J Exp Med 2005;202:1171–7. 5. Leadbetter EA, Rifkin IR, Hohlbaum AM, Beaudette BC, Shlomchik MJ, Marshak-Rothstein A. Chromatin-IgG complexes activate B cells by dual engagement of IgM and Toll-like receptors. Nature 2002;416:603–7. 6. Viglianti GA, Lau CM, Hanley TM, Miko BA, Shlomchik MJ, Marshak-Rothstein A. Activation of autoreactive B cells by CpG dsDNA. Immunity 2003;19:837–47. 7. Boule MW, Broughton C, Mackay F, Akira S, Marshak-Rothstein
REQUIREMENT OF TLR-7 FOR PRISTANE-INDUCED LUPUS
8.
9.
10.
11.
12.
13.
14. 15.
16. 17.
A, Rifkin IR. Toll-like receptor 9-dependent and -independent dendritic cell activation by chromatin-immunoglobulin G complexes. J Exp Med 2004;199:1631–40. Means TK, Latz E, Hayashi F, Murali MR, Golenbock DT, Luster AD. Human lupus autoantibody-DNA complexes activate DCs through cooperation of CD32 and TLR9. J Clin Invest 2005;115: 407–17. Savarese E, Chae OW, Trowitzsch S, Weber G, Kastner B, Akira S, et al. U1 small nuclear ribonucleoprotein immune complexes induce type I interferon in plasmacytoid dendritic cells through TLR7. Blood 2006;107:3229–34. Vollmer J, Tluk S, Schmitz C, Hamm S, Jurk M, Forsbach A, et al. Immune stimulation mediated by autoantigen binding sites within small nuclear RNAs involves Toll-like receptors 7 and 8. J Exp Med 2005;202:1575–85. Bave U, Magnusson M, Eloranta ML, Perers A, Alm GV, Ronnblom L. Fc␥ RIIa is expressed on natural IFN-␣-producing cells (plasmacytoid dendritic cells) and is required for the IFN-␣ production induced by apoptotic cells combined with lupus IgG. J Immunol 2003;171:3296–302. Bekeredjian-Ding IB, Wagner M, Hornung V, Giese T, Schnurr M, Endres S, et al. Plasmacytoid dendritic cells control TLR7 sensitivity of naive B cells via type I IFN. J Immunol 2005;174: 4043–50. Pisitkun P, Deane JA, Difilippantonio MJ, Tarasenko T, Satterthwaite AB, Bolland S. Autoreactive B cell responses to RNArelated antigens due to TLR7 gene duplication. Science 2006;312: 1669–72. Subramanian S, Tus K, Li QZ, Wang A, Tian XH, Zhou J, et al. A Tlr7 translocation accelerates systemic autoimmunity in murine lupus. Proc Natl Acad Sci U S A 2006;103:9970–5. Christensen SR, Shupe J, Nickerson K, Kashgarian M, Flavell RA, Shlomchik MJ. Toll-like receptor 7 and TLR9 dictate autoantibody specificity and have opposing inflammatory and regulatory roles in a murine model of lupus. Immunity 2006;25:417–28. Berland R, Fernandez L, Kari E, Han JH, Lomakin I, Akira S, et al. Toll-like receptor 7-dependent loss of B cell tolerance in pathogenic autoantibody knockin mice. Immunity 2006;25:429–40. Nacionales DC, Kelly KM, Lee PY, Zhuang H, Li Y, Weinstein JS, et al. Type I interferon production by tertiary lymphoid tissue developing in response to 2,6,10,14-tetramethyl-pentadecane (pristane). Am J Pathol 2006;168:1227–40.
1115
18. Calvani N, Caricchio R, Tucci M, Sobel ES, Silvestris F, Tartaglia P, et al. Induction of apoptosis by the hydrocarbon oil pristane: implications for pristane-induced lupus. J Immunol 2005;175: 4777–82. 19. Satoh M, Kumar A, Kanwar YS, Reeves WH. Anti-nuclear antibody production and immune-complex glomerulonephritis in BALB/c mice treated with pristane. Proc Natl Acad Sci U S A 1995;92:10934–8. 20. Hemmi H, Kaisho T, Takeuchi O, Sato S, Sanjo H, Hoshino K, et al. Small anti-viral compounds activate immune cells via the TLR7 MyD88-dependent signaling pathway. Nat Immunol 2002;3: 196–200. 21. Pawar RD, Ramanjaneyulu A, Kulkarni OP, Lech M, Segerer S, Anders HJ. Inhibition of Toll-like receptor-7 (TLR-7) or TLR-7 plus TLR-9 attenuates glomerulonephritis and lung injury in experimental lupus. J Am Soc Nephrol 2007;18:1721–31. 22. Austin HA III, Muenz LR, Joyce KM, Antonovych TT, Balow JE. Diffuse proliferative lupus nephritis: identification of specific pathologic features affecting renal outcome. Kidney Int 1984;25: 689–95. 23. Krug A, French AR, Barchet W, Fischer JA, Dzionek A, Pingel JT, et al. TLR9-dependent recognition of MCMV by IPC and DC generates coordinated cytokine responses that activate antiviral NK cell function. Immunity 2004;21:107–19. 24. Christensen SR, Shlomchik MJ. Regulation of lupus-related autoantibody production and clinical disease by Toll-like receptors. Semin Immunol 2007;19:11–23. 25. Kirou KA, Lee C, George S, Louca K, Peterson MG, Crow MK. Activation of the interferon-␣ pathway identifies a subgroup of systemic lupus erythematosus patients with distinct serologic features and active disease. Arthritis Rheum 2005;52:1491–503. 26. Martin DA, Elkon KB. Autoantibodies make a U-turn: the toll hypothesis for autoantibody specificity. J Exp Med 2005;202: 1465–9. 27. Ehlers M, Fukuyama H, McGaha TL, Aderem A, Ravetch JV. TLR9/MyD88 signaling is required for class switching to pathogenic IgG2a and 2b autoantibodies in SLE. J Exp Med 2006;203: 553–61. 28. Patole PS, Pawar RD, Lech M, Zecher D, Schmidt H, Segerer S, et al. Expression and regulation of Toll-like receptors in lupus-like immune complex glomerulonephritis of MRL-Faslpr mice. Nephrol Dial Transplant 2006;21:3062–73.
