Effect of an exogenous trigger on the pathogenesis of lupus in NZB Ф NZWF1 mice.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 46, No. 8, August 2002, pp 2235–2244 DOI 10.1002/art.10441 © 2002, American College of Rheumatology Effect of an Exogenous Trigger on the Pathogenesis of Lupus in (NZB ⫻ NZW)F1 Mice Hideo Yoshida,1 Minoru Satoh,1 Krista M. Behney,1 Chee-Gun Lee,2 Hanno B. Richards,1 Victoria M. Shaheen,1 Jun-Qi Yang,3 Ram R. Singh,3 and Westley H. Reeves1 Objective. This study examined the interactions between exogenous and endogenous factors shaping the phenotype of lupus in autoimmune (NZB ⴛ NZW)F1 mice exposed to pristane, a model environmental trigger. Methods. Frequencies of various autoantibodies in untreated NZB/NZW mice were determined by various means (immunoprecipitation, enzyme-linked immunosorbent assay [ELISA], Crithidia luciliae kinetoplast staining). Pristane or saline was administered intraperitoneally to 9–12-week-old NZB/NZW mice, followed by serial studies of autoantibodies, total Ig levels (ELISA), and proteinuria (dipstick). Results. Besides antichromatin/DNA responses, NZB/NZW mice spontaneously produced novel autoantibodies against the double-stranded RNA binding protein RNA helicase A (RHA). In contrast, NZB/NZW mice (n ⴝ 70) did not produce autoantibodies against the nuclear RNP (nRNP), Sm, Ro, or La antigens. Pristane exposure synergistically activated the production of antichromatin/DNA antibodies and dramatically accelerated renal disease. Production of anti-nRNP/Sm and Su autoantibodies also was induced, indicating that the unresponsiveness of NZB/NZW mice to these antigens can be overcome. Curiously, pristane treatment did not enhance the production of anti-RHA, suggesting that these autoantibodies are regulated differently than anti-DNA/chromatin and Sm. In contrast to previous reports that suggest a critical role of deficient interleukin-12 (IL-12) production in the pathogenesis of lupus, there was overproduction of IL-12 in the peritoneal cavity of pristane-treated NZB/NZW mice, and their spleen cells also produced large amounts of IL-12. Conclusion. These data lead us to propose that environmental influences exacerbate autoimmune manifestations in genetically lupus-susceptible mice through their stimulatory effects on proinflammatory cytokines, such as IL-12. Lupus is an autoimmune disorder of protean manifestations thought to result from both hereditary immunoregulatory defects and poorly characterized environmental factors such as ultraviolet radiation or chemicals (1–3). Multiple genetic loci contribute to the pathogenesis of lupus in (NZB ⫻ NZW)F1, (NZB/ NZW3), and other lupus-prone mice (4,5). However, little is known about the mechanisms by which environmental factors influence the disease. An inducible lupus syndrome with diseasespecific autoantibodies (anti–double-stranded DNA [anti-dsDNA], anti–nuclear RNP/Sm [anti-nRNP/Sm], anti–ribosomal P), arthritis, and nephritis develops in non–autoimmune-prone mice treated with pristane, a hydrocarbon derived from the metabolism of chlorophyll (6,7). BALB/c, B6, and virtually all other immunocompetent mice are susceptible to pristane-induced lupus, but the associated autoantibodies and clinical manifestations differ from strain to strain (8). The development of lupus following pristane treatment is cytokine dependent. Induction of anti-DNA/chromatin antibodies and glomerulonephritis by pristane is abrogated in mice lacking interleukin-6 (IL-6) or interferon-␥ (IFN␥) (9,10). In contrast, the antinRNP/Sm autoantibody frequency is markedly reduced in IFN␥-deficient mice, but not in IL-6–deficient strains. Supported by research grants R01-AR44731, R01-AI44074, and R01-AR47322 from the USPHS. 1 Hideo Yoshida, MD, Minoru Satoh, MD, PhD, Krista M. Behney, BS, Hanno B. Richards, MD, Victoria M. Shaheen, MS, Westley H. Reeves, MD: Department of Medicine, University of Florida, Gainesville; 2Chee-Gun Lee, PhD: New Jersey Medical School, University of Medicine and Dentistry of New Jersey, Newark; 3 Jun-Qi Yang, PhD, Ram R. Singh, MD: Department of Internal Medicine, University of Cincinnati, Cincinnati, Ohio. Address correspondence and reprint requests to Westley H. Reeves, MD, Division of Rheumatology and Clinical Immunology, University of Florida, 1600 SW Archer Road, PO Box 100221, Gainesville, FL 32610-0221. E-mail: email@example.com. Submitted for publication February 6, 2002; accepted in revised form April 22, 2002. 2235 2236 YOSHIDA ET AL The goal of this study was to explore the interactions between exogenous and endogenous factors shaping the phenotype of lupus, using pristane as a model environmental trigger of disease in autoimmune-prone NZB/NZW mice. These mice spontaneously develop a strong autoantibody response against DNA and chromatin antigens as well as polyclonal hypergammaglobulinemia and severe immune complex–mediated glomerulonephritis (4,11,12). It has been suggested that they do not produce anti-Sm (11,13,14), but other reports are contradictory (15–17). Since there are surprisingly few systematic studies of autoantibody specificities in NZB/ NZW mice, we examined both the baseline serologic phenotype and the effect of pristane treatment on this phenotype. The present study verifies the absence of anti-nRNP and Sm, as well as anti-Su, Ro/SSA, and La/SSB, autoantibody production in NZB/NZW mice. We show for the first time that NZB/NZW mice produce autoantibodies against anti–RNA helicase A (RHA) and that the serologic and clinical phenotype of lupus in this strain is modified substantially by pristane exposure. This may prove valuable for modeling the role of exogenous factors in systemic lupus erythematosus (SLE). MATERIALS AND METHODS Mice. Four-week-old female NZB/NZW mice from The Jackson Laboratory (Bar Harbor, ME) were housed in a virus-free animal facility in barrier cages at the University of North Carolina, Chapel Hill. At 9–11 weeks of age, 23 mice received 0.5 ml of pristane (Sigma, St. Louis, MO) intraperitoneally (IP). Controls (n ⫽ 24) received 0.5 ml of phosphate buffered saline (PBS) IP. Sera were collected 2 weeks before injection, and monthly thereafter. Proteinuria was measured with Albustix (Miles, Elkhart, IN). Additional female mice (9 pristane- and 9 PBS-treated, ages 9–11 weeks; The Jackson Laboratory) housed in a virusfree conventional colony in barrier cages at the University of Florida were studied 3 months after treatment for histologic features and cytokine production. Untreated female NZB/ NZW mice (n ⫽ 37; The Jackson Laboratory) housed in nonbarrier cages in a conventional animal facility at the University of Cincinnati were bled at ages 5–10 months. The respective institutions’ animal care and use committees approved this study. Immunoprecipitation. Autoantibodies to cellular proteins in murine sera from all 3 institutions were analyzed at the University of Florida using a standardized immunoprecipitation procedure (6). 35S-radiolabeled K562 cell extract was immunoprecipitated using 3 l serum. Specificities were verified by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) using human prototype sera for antinRNP/Sm and anti-Su. Human anti-RHA serum was from a patient with SLE. The University of Florida Institutional Review Board approved the studies using human sera. Double-stranded RNA binding proteins were isolated using poly(IC) agarose (Pharmacia, Piscataway, NJ). A volume of 120 l beads (50% [volume/volume] in Tris buffered saline [TBS]/Tween 20 [150 mM NaCl, 20 mM Tris HCl, 0.1% Tween 20 (pH 7.5)] with 3 mM MgCl2) was washed 3 times. Beads were then incubated with cell extract from 107 35S-methionine– labeled K562 cells in TBS/Tween 20 plus phenylmethylsulfonyl fluoride and aprotinin for 1 hour at 4°C, and washed 6 times with 1 ml TBS/Tween 20 plus 3 mM MgCl2. Bound proteins were eluted with 600 l of 1.5M NaCl NET/NP40 (1.5M NaCl, 2 mM EDTA, 20 mM Tris [pH 7.5]) for 20 minutes at 4°C, and an equal volume of water was added (final concentration 0.75M NaCl). The eluate was immunoprecipitated with 3 l of human or murine autoimmune sera or normal human serum for 30 minutes at 4°C, washed 3 times with 0.5M NaCl NET/NP40, and once with NET before SDS-PAGE and autoradiography (6). Immunoblotting. IgG from 3 l of human or murine sera containing possible autoantibodies to RHA, or control serum, was crosslinked to protein A–Sepharose beads using dimethyl pimelimidate (18). Proteins immunoprecipitated from K562 cell extract (2 ⫻ 107 cell equivalents) were fractionated by SDS-PAGE and transferred to nitrocellulose. One filter was probed with rabbit anti-RHA serum at a dilution of 1:3,000 (19,20). A second was probed with preimmune rabbit serum. Blots were developed with 1:2,000 alkaline phosphatase–labeled goat anti-rabbit IgG (␥ and light chain– specific; Southern Biotechnology, Birmingham, AL) and developed using the Western-Star chemiluminescent system (Tropix, Bedford, MA). Immunoblot analysis of the fine specificity of antinRNP/Sm antibodies was carried out using murine sera as described (21). Monoclonal antibodies (mAb) 2.73 (anti– U1–70 kd) (22), 7-13 (anti–Sm D) (23), and 9A9 (anti–U1-A plus U2-B⬙) (24) were used as standards. Ig and autoantibody levels. Immunologic tests were performed at the University of Florida using standardized and previously published protocols. Levels of IgG1, IgG2a, IgG2b, IgG3, and IgM were determined as described using sera diluted 1:200,000 and/or 1:500,000 (25). Enzyme-linked immunosorbent assays (ELISAs) for antichromatin and anti–singlestranded DNA (anti-ssDNA) antibodies were performed using sera diluted 1:500 (7,26). IgM and IgG antichromatin antibodies were considered positive when sample absorbance was higher than the mean optical density from blank wells plus 3SD Anti-ssDNA antibodies were considered positive when sample absorbance was higher than the mean optical density plus 3SD using sera from 8 healthy female BALB/c mice. Anti-dsDNA antibodies were detected by Crithidia luciliae kinetoplast staining and quantified by titration emulation (Image Titer; Rhigene, Des Plaines, IL) as described (27). Cytokine production in vitro and in vivo. Three months after PBS or pristane treatment (6 months of age), the NZB/NZW mice were killed and the peritoneal cavity was lavaged with 5 ml of Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum (FBS) and 10 units/ml heparin using a 5-ml plastic syringe and 18-gauge needle. Single-cell suspensions of peritoneal and spleen cells were treated with Tris NH4Cl to lyse erythrocytes. Cells were EFFECT OF PRISTANE ON LUPUS IN MICE Table 1. 2237 Autoantibodies in sera obtained from 70 untreated or PBS-treated NZB/NZW mice* No. of mice Age range, months Housing Anti-dsDNA (by Crithidia luciliae assay), no. (%) Anti-ssDNA (by ELISA), no. (%) Antichromatin (by ELISA), no. (%) Anti-nRNP/Sm (by immunoprecipitation) Anti-Su (by immunoprecipitation) Anti-Ro/SSA (by immunoprecipitation) Anti-La/SSB (by immunoprecipitation) Anti–ribosomal P (by immunoprecipitation) Anti–RNA helicase A (by immunoprecipitation), no. (%)‡ University of North Carolina University of Florida University of Cincinnati 24 6–7 Virus free, barrier 7 (29) 24 (100) 23 (96) 0 0 0 0 0 6 (25) 9 6–7 MHV free, barrier 4 (44) 5 (56) 9 (100) 0 0 0 0 0 0 37 5–10 Conventional 30/34 (88)† 33/34 (97)† 33/34 (97)† 0 0 0 0 0 4 (11) * PBS ⫽ phosphate buffered saline; MHV ⫽ mouse hepatitis virus; dsDNA ⫽ double-stranded DNA; ssDNA ⫽ single-stranded DNA; ELISA ⫽ enzyme-linked immunosorbent assay; nRNP ⫽ nuclear RNP. † Only 34 of the 37 mice were tested for anti-DNA and antichromatin due to insufficient quantities of 3 sera. ‡ The 140-kd autoantigen was shown by immunoprecipitation to be RNA helicase A (see Figure 1). resuspended at 106/ml in DMEM plus 10% FBS and cultured for 48 hours in 24-well culture plates (Costar, Cambridge, MA) without stimulation or in the presence of 10 g/ml lipopolysaccharide (LPS; from Salmonella minnesota Re 595; Sigma). Culture supernatants were stored at ⫺80°C until assayed. ELISAs for IL-4, IL-6, IL-12, tumor necrosis factor ␣ (TNF␣), and IFN␥ were performed using rat mAb pairs (PharMingen, San Diego, CA) following the manufacturer’s instructions. After incubation with biotinylated cytokine-specific antibodies, 100 l/well of 1:1,000 streptavidin–alkaline phosphatase (Southern Biotechnology) was added for 30 minutes at 22°C and the reaction was developed with p-nitrophenyl phosphate substrate. Standard curves were generated using recombinant cytokines (PharMingen). Cytokine concentrations were calculated using a 4-parameter logistic equation as part of the Softmax Pro 3.0 ELISA plate reader software (Molecular Devices, Sunnyvale, CA). Statistical analysis. Levels of Ig and autoantibodies were compared using the Mann-Whitney test. The development of proteinuria (% of mice with ⱖ300 mg/dl) and the survival rates were evaluated by a log rank test (LifeTest Procedure; SAS Institute, Cary, NC). RESULTS Serologic phenotype of untreated NZB/NZW mice. Consistent with prior observations (28,29), serologic studies of 70 untreated or PBS-treated NZB/NZW mice, ages 5–10 months, housed in 3 different institutions revealed that, depending on the animal colony, 29–88% had anti-dsDNA (Crithidia luciliae assay) and 56–100% had anti-ssDNA (ELISA), whereas virtually all had anti-chromatin antibodies (ELISA) (Table 1). When screened for anti-nRNP/Sm antibodies by a sensitive immunoprecipitation technique, all 70 sera from animals at 3 different institutions were negative (Table 1). In addition, all of the sera were negative for anti-Ro/ SSA, anti-La/SSB, anti-Su, and anti–ribosomal P autoantibodies. Although many specificities were not produced by NZB/NZW mice, autoantibodies against a protein of ⬃140 kd were detected in up to 25% of the sera. This protein comigrated with a 140-kd protein recognized by certain human lupus sera (Figure 1A). The proteins immunoprecipitated by human autoimmune serum and 2 NZB/NZW sera were transferred to nitrocellulose membrane and probed with a rabbit antiserum specific for RHA (Figure 1B). The rabbit antiserum bound intensely to the 140-kd protein, whereas control (preimmune) rabbit serum did not. The dsRNA binding protein RHA had a molecular weight of ⬃140 kd (20,30). In addition to its size and reactivity with rabbit anti-RHA antibodies, the 140-kd protein bound specifically to poly(IC)–agarose (Figure 1A), as expected for a dsRNA binding protein. Although the frequency of these autoantibodies varied from group to group, the production of anti-RHA was a consistent serologic manifestation in mice from 4 separate experiments at 2 institutions (University of North Carolina and University of Cincinnati) (Table 1). Anti-RHA was not detected in mice housed at the University of Florida, possibly due to the small number evaluated (n ⫽ 9). The spectrum of autoantibodies in NZB/NZW mice differed substantially from that in mice with pristane-induced lupus. Pristane stimulates the production of anti-nRNP/Sm, Su, and ribosomal P in a wide variety of mouse strains (8), but anti-RHA, anti-Ro/ SSA, and anti-La/SSB have not been reported. In contrast, anti-DNA and antichromatin antibodies are pro- 2238 Figure 1. Spontaneous production of autoantibodies to RNA helicase A (RHA) by NZB/NZW mice. A, 35S-labeled K562 cell extract was immunoprecipitated using human serum containing autoantibodies against a 140-kd protein (RHA), with human serum containing anti-NF90/NF45/p110 autoantibodies (NF45/90), and with sera from 2 phosphate buffered saline–treated NZB/NZW mice (B/W #1, B/W #2). Immunoprecipitation with normal mouse serum (NMS) also is shown. A protein of ⬃140-kd immunoprecipitated by the 3 sera is indicated to the left, as are the double-stranded RNA binding proteins NF90, NF45, and p110. On the right, poly(IC) binding proteins from whole cell extract eluted from the poly(IC)–agarose affinity matrix at 0.15M and 0.5M NaCl were analyzed by sodium dodecyl sulfate– polyacrylamide gel electrophoresis. Immunoprecipitation using the 1.5M NaCl poly(IC) eluate (after washing the beads with Tris buffered saline, 0.1% Tween 20, 3 mM MgCl2) is shown in the last 4 lanes. Three of the prominent proteins in the 1.5M eluate were immunoprecipitated by anti-NF90/NF45/p110 antiserum. The fourth was immunoprecipitated by human anti-RHA antiserum and by sera from the NZB/NZW mice (B/W #1, B/W #2). Positions of the molecular weight standards are indicated on the right. B, Proteins immunoprecipitated by human (h-RHA) or murine (B/W #1, B/W #2) sera were transferred to nitrocellulose and probed with polyclonal rabbit anti-RHA antiserum (anti-RHA) or with preimmune rabbit serum (control). Position of the RHA is indicated on the left and that of the molecular weight markers on the right. duced in both spontaneous (NZB/NZW) and pristaneinduced lupus. It was of interest, therefore, to see if pristane could alter the pattern of autoantibody production and clinical features of NZB/NZW mice. YOSHIDA ET AL Acceleration of the anti-DNA/chromatin response and nephritis. Consistent with previous reports, NZB/NZW mice spontaneously developed IgM antissDNA and antichromatin antibodies before they were treated with PBS or pristane at 9 weeks of age (Figure 2). At that time, the frequencies of IgM anti-ssDNA and IgM antichromatin autoantibodies were 58% and 33%, respectively. In contrast, IgG anti-ssDNA and IgG antichromatin antibodies were not produced until after 6 months of age (4 months after PBS treatment). By 8 months of age, their frequencies were 100% and 96%, respectively. Pristane greatly accelerated the onset and increased the levels of IgM and IgG anti-ssDNA and antichromatin autoantibodies at 1–4 months (P ⬍ 0.05 versus PBS-treated group, by Mann-Whitney test). Although antichromatin is typical of the autoimmune syndrome of NZB/NZW mice, anti-dsDNA is more closely linked to glomerulonephritis. Consistent with previous observations, anti-dsDNA antibodies were detected at high titers (Crithidia luciliae assay) in sera from 29% of 5-month-old PBS-treated NZB/NZW mice (Figure 3A). However, the titers were dramatically higher in pristane-treated mice of the same age (P ⫽ 0.0004 versus PBS-treated controls, by Mann-Whitney test), as was their frequency (78% versus 29%; P ⬍ 0.002 by Fisher’s exact test). Similarly, renal disease was greatly accelerated by pristane treatment. As early as 3 Figure 2. Anti–single-stranded DNA (anti-ssDNA) and antichromatin autoantibody levels. IgM and IgG anti-ssDNA (top) and IgM and IgG antichromatin (bottom) antibodies were measured in 1:500 diluted sera from phosphate buffered saline-treated or pristane NZB/ NZW mice as a function of time after treatment. Individual data points represent the mean ⫾ SEM of 23 or 24 mice. Heat-denatured calf thymus DNA was used as an antigen for detecting anti-ssDNA antibodies; chicken chromatin was used as an antigen for detecting antichromatin antibodies. EFFECT OF PRISTANE ON LUPUS IN MICE Figure 3. Acceleration of anti–double-stranded DNA (anti-dsDNA) antibody production and nephritis by pristane. A, Anti-dsDNA antibody titers (Crithidia luciliae assay) in sera from phosphate buffered saline (PBS)–treated (n ⫽ 24) or pristane-treated (n ⫽ 23) NZB/NZW mice were quantified using titration emulation. The levels were significantly different between the 2 groups (P ⫽ 0.0004). B, The frequency of proteinuria ⱖ300 mg/dl was determined monthly after treating 8-weekold NZB/NZW mice with pristane (}; n ⫽ 12) or PBS (■; n ⫽ 12). There was a significant difference between the 2 groups (P ⫽ 0.01 by log-rank test). Data are representative of 2 separate experiments. C, Survival rates were determined in pristane-treated (}; n ⫽ 12) and PBS-treated (■; n ⫽ 12) NZB/NZW mice at ages 9–10 weeks. Survival of the pristane-treated group was reduced compared with that of the PBS controls (P ⫽ 0.006 by log-rank test). Data are representative of 2 separate experiments. 2239 months after pristane treatment, an increased frequency of severe proteinuria (ⱖ300 mg/dl) was observed in a group of pristane-treated mice compared with PBStreated NZB/NZW controls (Figure 3B). Over the ensuing 3 months, the acceleration of renal disease caused by pristane became increasingly apparent. The difference was significant 5 months after treatment (P ⫽ 0.01 by log-rank test). Pristane treatment also dramatically accelerated mortality (Figure 3C). Seventy-five percent of the pristane-treated mice died within 6 months, compared with 9% of the PBS-treated control mice (P ⫽ 0.006 by log-rank test). These data strongly suggest that pristane acts synergistically with the NZB/NZW genetic background to promote the development of anti-DNA/ chromatin antibodies and glomerulonephritis. Induction of anti-nRNP/Sm and anti-Su. NZB/ NZW mice do not produce anti-nRNP/Sm and Su autoantibodies spontaneously (Table 1), whereas the production of anti-RHA has not been reported in pristane-induced lupus. We next sought to determine 1) whether the production of “pristane-associated” autoantibodies could be induced in NZB/NZW mice and 2) whether, like anti-DNA and antichromatin antibodies, anti-RHA was accelerated by pristane treatment. Although not found in 70 control mice, antinRNP/Sm or Su autoantibodies were detected by immunoprecipitation in 4 of 32 (12.5%) pristane-treated mice (P ⫽ 0.008 versus controls, by Fisher’s exact test) (Table 2). Anti-nRNP/Sm autoantibodies appeared at high levels ⬃3 months after pristane treatment (Figure 4A). Immunoprecipitation of the U5 small nuclear RNP– specific 200-kd doublet (Figure 4A) indicated that antiSm antibodies were present. That possibility was supported further by the reactivity of sera with the B⬘/B and/or D proteins by Western blotting (Figure 4B). In 1 mouse (B/W F1 #1 in Figure 4) anti-nRNP preceded the onset of anti-Sm (U5–200-kd doublet), a pattern characteristic of pristane-induced lupus (21). The presence of anti-nRNP autoantibodies also was verified by West- Table 2. Frequency of IgG autoantibodies in NZB/NZW mice* No. positive by immunoprecipitation assay (%) Treatment No. of mice Anti-nRNP/Sm/Su Anti-RHA PBS or unmanipulated Pristane 70 32 0 4 (12)† 10 (14) 1 (3) * PBS ⫽ phosphate buffered saline; RHA ⫽ RNA helicase A. † P ⫽ 0.008 versus controls, by Fisher’s exact test. 2240 Figure 4. Induction of anti–nuclear RNP/Sm/Su (anti-nRNP/Sm/Su) autoantibodies by pristane. A, 35S-labeled K562 cell extract was immunoprecipitated with sera from 2 NZB/NZW mice that developed anti-nRNP/Sm autoantibodies following pristane treatment. Sera were tested 0, 1, 2, 3, and 4 months after treatment. Immunoprecipitates using normal mouse serum (NMS; negative control) and human anti-nRNP prototype serum (nRNP; positive control) are shown on the left. Positions of the U5–200-kd doublet, U1-A, Sm B⬘/B, U1-C, and Sm, D, E/F, and G proteins are indicated to the left. B, Immunoblot analysis of sera from the 2 anti-nRNP/Sm–positive NZB/ NZW mice, using affinity-purified U1 small nuclear RNP antigen. Positive controls include monoclonal antibodies 2.73 (anti–U1–70 kd), 9A9 (anti–U1-A ⫹ U2-B⬙), and 7-13 (anti–Sm D). The pattern with normal mouse serum (control) is also shown. Positions of the U1–70kd, U1-A, and Sm B⬘/B and D proteins are indicated to the left. ern blotting, which showed reactivity with the U1-A and U1–70-kd proteins (Figure 4B). In striking contrast to the acceleration of antiDNA/chromatin autoantibodies and the induction of anti-nRNP/Sm and Su, the production of autoantibodies to RHA was not enhanced by pristane treatment (Table 2). In control mice, the frequency of these autoantibodies was 14% (10 of 70), whereas in the pristane-treated group, the frequency was 3% (1 of 32). Together, these data indicate that pristane accelerates the production of anti-DNA and antichromatin antibodies and induces anti-Sm, anti-nRNP, and anti-Su autoantibodies in NZB/NZW mice. In contrast, pristane did not enhance the production of anti-RHA. Alteration of cytokine production in NZB/NZW mice by pristane. In view of the induction/acceleration of the production of IFN␥-dependent autoantibodies such as anti-nRNP/Sm and anti-dsDNA, we examined whether pristane-treated NZB/NZW mice displayed evidence of Th1 cytokine overproduction (Figure 5). LPSstimulated IL-12 production by spleen cells from PBStreated mice by was relatively low (Figure 5A), although it was greater than the levels seen in PBS-treated B6 or BALB/c mice (Satoh M: unpublished observations). In YOSHIDA ET AL contrast, spleen cells from pristane-treated mice produced large amounts of IL-12 (P ⬍ 0.05 versus PBS controls, by Mann-Whitney test) (Figure 5A). IL-4 was produced in LPS-stimulated spleen cell cultures from PBS-treated NZB/NZW mice, but was reduced in pristane-treated mice (Figure 5A). Peritoneal lavage fluid from pristane-treated mice also contained large amounts of IL-12 (P ⬍ 0.005 versus controls, by Mann-Whitney test) (Figure 5B), indicating that pristane enhances the production of this cytokine in vivo. Peritoneal lavage fluid from both pristane- and PBS-treated mice contained little IL-6 or TNF␣. Although there was a trend toward increased levels of IFN␥ with pristane treatment, this did not reach statistical significance. Consistent with the effect of IL-12 in priming T Figure 5. Alteration of cytokine balance by pristane. A, Increased lipopolysaccharide (LPS)–stimulated interleukin-12 (IL-12) production. Spleen cells were isolated from pristane-treated or phosphate buffered saline (PBS)–treated NZB/NZW mice and stimulated in vitro with LPS (10 g/ml). Levels of IL-12 and IL-4 in the culture supernatants were determined 48 hours later. IL-4 production was lower and IL-12 higher in cultures of cells from pristane-treated mice (P ⬍ 0.05 by Mann-Whitney test). B, The peritoneal cavities of pristane-treated mice and PBS-treated controls were lavaged with 5 ml of Dulbecco’s modified Eagle’s medium, and the fluid was tested for IL-12, IL-6, tumor necrosis factor ␣ (TNF␣), and interferon-␥ (IFN␥) (by enzymelinked immunosorbent assay-treated ELISA). Higher IL-12 levels were found in the peritoneal lavage fluid from pristane-treated mice compared with the controls (P ⬍ 0.005 by Mann-Whitney test). C, Levels of total IgG2a and IgG1 in the sera of pristane-treated and PBStreated mice were measured by ELISA 3 months after treatment, and the ratio of the levels for each mouse was determined. The mean IgG2a; IgG1 ratio in pristane-treated mice was higher than that in PBS-treated mice (P ⬍ 0.001 by Mann-Whitney test). RHA ⫽ RNA helicase A; nRNP ⫽ nuclear RNP. D, Total IgG1, IgG2a, IgG2b, IgG3, and IgM levels were determined by ELISA as a function of time after treatment with pristane or PBS. EFFECT OF PRISTANE ON LUPUS IN MICE cells for IFN␥ production and inhibiting IL-4 (31), the production of IFN␥ versus IL-4–dependent Ig isotypes was altered dramatically by pristane treatment. There was a marked difference in the ratio of total IgG2a to IgG1 in pristane- versus PBS-treated NZB/NZW mice (P ⬍ 0.001 by Mann-Whitney test) (Figure 5C). Interestingly, the 7 anti–RHA–positive mice (6 control, 1 pristane-treated) all had low IgG2a/IgG1 ratios, whereas 3 of the 4 anti-nRNP/Sm and Su-positive mice (2 anti-nRNP/Sm, 2 Su) had high ratios (P ⬍ 0.05 by Mann-Whitney test). As seen in non–autoimmuneprone mice treated with pristane, total levels of the IFN␥-dependent Ig isotype IgG2a increased dramatically in pristane-treated mice versus controls (P ⬍ 0.05 at 3 and 4 months by Mann-Whitney test) (Figure 5D). The IL-4–dependent isotype IgG1 also increased, but less markedly and with different kinetics: IgG2a peaked at 4 months and then declined precipitously, whereas IgG1 reached a plateau at 5 months. These data strongly suggest that pristane skews cytokine production toward Th1 predominance, which might be expected to promote the formation of anti-nRNP/Sm and Su autoantibodies (10). DISCUSSION Susceptibility to lupus is thought to result from the action of poorly defined environmental triggers on a susceptible genetic background (1,12). Multiple genes contribute independently to the phenotype of lupus exhibited by lupus-prone mice, such as NZB/NZW (5,12). The ability to induce lupus-like manifestations in a wide variety of mice (BALB/c, B6, SJL, CBA, etc.) with pristane suggests that nearly any genetic background may be permissive, given a sufficiently powerful inciting event (7,8). To model how environmental stimuli influence the onset of disease in a susceptible host, we examined the interaction of pristane with the lupusprone NZB/NZW genetic background. Although in this mouse model, the hydrocarbon oil was administered into the peritoneal cavity, human exposure to mineral oil through the respiratory or gastrointestinal tract also induces inflammatory disease with pathologic changes similar to those seen in pristane-treated mice (32–34). Oil-based adjuvants also are used in human and veterinary vaccines (35). Hydrocarbon exposure synergistically activated the production of antichromatin and anti-DNA antibodies and accelerated the onset of renal disease. However, the serologic phenotype of the disease was altered with the appearance of anti-nRNP/Sm and Su 2241 autoantibodies, specificities that are not produced spontaneously by NZB/NZW mice. NZB/NZW mice spontaneously produced novel autoantibodies against the dsRNA binding protein RHA. Curiously, pristane treatment did not enhance anti-RHA production despite dramatically stimulating the production of other autoantibodies, which suggests that anti-RHA is regulated differently than anti-DNA/chromatin. The Sle1 locus, which regulates susceptibility to antichromatin responses (28), may not control anti-RHA autoantibody production, or else additional loci may be required in concert, a hypothesis that should be testable using Sle1 congenic strains. NZB/NZW mice produce autoantibodies primarily against chromatin (DNA and histones) (13). We did not detect anti-Sm, nRNP, Ro, La, Su, or ribosomal P autoantibodies in 70 NZB/NZW mice, a finding in agreement with some studies (14,29) but contradicting other studies in which these specificities were detected by ELISA or immunoblot analysis (16,17). Although the reason for this discrepancy is uncertain, differences in the housing and/or autoantibody assays used in previous studies versus those used in the present study may be responsible. It is difficult to ascribe this to housing differences because mice housed at 3 different institutions did not spontaneously produce anti-nRNP/Sm autoantibodies. Assay reliability is another possible explanation for the discrepancy. The specificity of anti-Sm antibodies for SLE has been established by double immunodiffusion, counterimmunoelectrophoresis, and immunoprecipitation (36–38), whereas the diagnostic specificity of the antibodies detected by ELISA or immunoblotting has not been established. Since immunoprecipitation generally correlates favorably with double immunodiffusion but is more sensitive (39), it seems unlikely that the absence of anti-nRNP/Sm autoantibodies was a consequence of an insensitivity of the assay. While we cannot exclude the presence of additional low-affinity or lowtiter autoantibodies, insofar as could be determined here, the spontaneous autoantibody response of NZB/ NZW mice is restricted to chromatin and RHA. Pristane is powerfully synergistic with the genetic background in promoting anti-DNA/chromatin responses and renal disease, and it induces de novo production of anti-nRNP/Sm and Su. In contrast, antiRHA was not enhanced (and was possibly inhibited), suggesting that its regulation differs from that of other autoantibodies. There is mounting evidence that subsets of autoantibodies are differentially regulated. The absence of functional Fas or Fas ligand promotes anti- 2242 DNA/chromatin responses in B6 mice (40), but unlike the situation in NZB/NZW mice, pristane treatment has no effect on these responses (26). Pristane’s synergy with anti-DNA/chromatin responses in NZB/NZW mice and lack of effect in lpr or gld mice suggest that there is more than one pathway to antichromatin production. The spontaneous production of autoantibodies to the dsRNA binding protein RHA stands in sharp contrast to the absence of other “typical” lupus autoantibodies in NZB/NZW mice. Double-stranded RNA binding proteins are recently described autoantigens in pristane-induced lupus and SLE (41,42). The novel 140-kd protein immunoprecipitated by sera from ⬃10– 20% of NZB/NZW mice was identified on the basis of biochemical and immunologic criteria. RHA is a dsRNA binding protein controlled by IFN regulatory factor (30) that interacts with adenovirus VA RNAII (20), plays a role in the nuclear export of viral RNA (43), and facilitates Rev-mediated gene expression and nuclear export of RNA in human immunodeficiency virus– infected cells (44). Consistent with a recent report (42), autoantibodies to this protein also were found in certain human lupus sera (Figure 1A). However, we have not detected anti-RHA autoantibodies in other murine models of spontaneous lupus (MRL, BXSB), and have only rarely encountered anti-RHA either spontaneously or following pristane treatment in non–autoimmuneprone mice (Satoh M, et al: unpublished observations). The different frequencies of anti-RHA autoantibodies in untreated NZB/NZW mice housed at different institutions suggest that undefined environmental factors can affect the spontaneous production of autoantibodies in lupus-prone mice. Although the mice in 2 of the animal colonies (University of Florida and University of North Carolina) were housed in barrier cages and were free of conventional murine viruses, including mouse hepatitis virus, it is difficult to completely control all environmental variables, which may include lighting, diet, bedding, stress, and the microbial flora. Indeed, anti-nRNP/Sm and anti-Su autoantibody production in pristane-induced lupus is enhanced in conventionally housed mice compared with mice housed under specific pathogen–free conditions (25). It will be of interest to further define these effects, which may imply a greater role of the environment in the pathogenesis of spontaneous lupus than has been generally supposed. We hypothesize that environmental variables may influence the levels of cytokines that promote autoantibody formation in pristane-induced lupus, such as IL-6, IFN␥, and IL-12 (9,10). IL-12 was the major cytokine overproduced both YOSHIDA ET AL in vitro and in vivo following pristane treatment (Figure 5). Interestingly, NZB/NZW, MRL/lpr, and other lupus mice exhibit a defect in macrophage IL-12 production prior to the onset of disease (45). Monocyte-derived macrophages from SLE patients also have defective IL-12 production (46). In NZB/NZW mice, a defect in the initiation, but not the propagation, of Th1 responses may play a fundamental role in lupus by fostering polyclonal hyperexpansion of autoreactive lymphocytes and the establishment of a Th1-deficient environment promoting autoantibody formation (45). However, large amounts of IL-12 were detected in the peritoneal cavity of pristane-treated mice, and their spleen cells produced high levels of LPS-stimulated IL-12 (Figure 5) in association with a dramatic enhancement of anti-DNA/ chromatin autoantibody formation and glomerulonephritis. Our data suggest that the defective IL-12 production reported in murine and human lupus is not fundamental to the disease process. Indeed, IFN␥ may promote autoimmunity in NZB/NZW mice, since IFN␥ receptor–deficient NZB/NZW mice have less severe disease (47), and its neutralization with anti-IFN␥ mAb (48) or soluble IFN␥ receptor (49) prevents renal disease. Pristane does not induce anti-DNA/chromatin antibodies and nephritis in either IL-6 or IFN␥– deficient BALB/c mice (9,10). The greatly increased IL-12 production in pristane-treated NZB/NZW mice versus controls could promote Th1 responses (31). Two lines of evidence suggest this is the case: 1) IL-4 production in vitro was reduced following pristane treatment, presumably due to IFN␥ and/or IL-12 production; and 2) there was greatly increased production of the IFN␥dependent (50) IgG2a isotype. Since IgG2a is the predominant isotype in glomerular deposits of NZB/NZW mice (13), IL-12 could promote disease by enhancing the production of IFN␥ and the expression of pathogenic IgG2a anti-dsDNA antibodies (13). The exacerbation of nephritis by pristane also might be a direct effect of IL-12 (and/or IFN␥) on the kidney, leading to the formation of epithelial crescents (51), which have been reported in pristane-induced lupus (7). Alternatively, increased IL-12 production could promote glomerulonephritis by enhancing nitric oxide biosynthesis (52). An IFN␥-rich environment seems to favor the production of anti-nRNP/Sm and anti-dsDNA antibodies (10,53). Mice that produced anti-nRNP/Sm and Su had a higher IgG2a/IgG1 ratio than those that produced anti-RHA (Figure 5C), suggesting that antiRHA antibodies may be produced under the influence EFFECT OF PRISTANE ON LUPUS IN MICE of a different set of cytokines than anti-DNA/chromatin or anti-nRNP/Sm/Su. Bacterial DNA stimulates the production of a spectrum of cytokines similar to that induced by pristane (54). Like pristane treatment, immunization of NZB/ NZW mice with Escherichia coli dsDNA accelerates the production of anti-dsDNA antibodies while promoting a shift toward Th1 cytokine production (54). But in contrast to pristane, bacterial DNA reduces the severity of glomerulonephritis (55). Moreover, bacterial DNA has not been reported to induce anti-nRNP/Sm or Su autoantibodies. Thus, despite similarities in the cytokines induced by these 2 exogenous immune stimulants, there also may be significant differences in their effects. These differences could have implications for understanding how various environmental exposures play a role in shaping the phenotype of lupus. ACKNOWLEDGMENTS We are grateful to Dr. Alan Hutson (Biostatistics Department, University of Florida, Gainesville) for statistical assistance and Dr. Michael Mathews (New Jersey Medical School, University of Medicine and Dentistry of New Jersey, Newark) for helpful discussions. We thank Drs. Robert A. Eisenberg and Philip L. Cohen (University of Pennsylvania, Philadelphia) and Dr. Walther J. van Venrooij (University of Nijmegen, The Netherlands) for providing monoclonal antibodies. REFERENCES 1. Vyse TJ, Kotzin BL. Genetic susceptibility to systemic lupus erythematosus. Annu Rev Immunol 1998;16:261–92. 2. Dowdy MJ, Nigra TP, Barth WF. Subacute cutaneous lupus erythematosus during PUVA therapy for psoriasis: case report and review of the literature. Arthritis Rheum 1989;32:343–6. 3. Hess E. Drug-related lupus. N Engl J Med 1988;318:1460–2. 4. Morel L, Mohan C, Yu Y, Croker BP, Tian N, Deng A, et al. Functional dissection of systemic lupus erythematosus using congenic mouse strains. J Immunol 1997;158:6019–28. 5. Morel L, Tian XH, Croker BP, Wakeland EK. Epistatic modifiers of autoimmunity in a murine model of lupus nephritis. Immunity 1999;11:131–9. 6. Satoh M, Reeves WH. Induction of lupus-associated autoantibodies in BALB/c mice by intraperitoneal injection of pristane. J Exp Med 1994;180:2341–6. 7. 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. 8. Satoh M, Richards HB, Shaheen VM, Yoshida H, Shaw M, Naim JO, et al. Widespread susceptibility among inbred mouse strains to the induction of lupus autoantibodies by pristane. Clin Exp Immunol 2000;121:399–405. 9. Richards HB, Satoh M, Shaw M, Libert C, Poli V, Reeves WH. Interleukin-6 dependence of anti-DNA antibody production: evidence for two pathways of autoantibody formation in pristaneinduced lupus. J Exp Med 1998;188:985–90. 2243 10. Richards HB, Satoh M, Jennette JC, Croker BP, Reeves WH. Interferon-␥ is required for lupus nephritis in mice treated with the hydrocarbon oil pristane. Kidney Int 2001;60:2173–80. 11. Theofilopoulos AN, Kono DH. Murine lupus models: genespecific and genome-wide studies. In: Lahita RG, editor. Systemic lupus erythematosus. San Diego: Academic Press; 1999. p. 145–81. 12. Morel L, Rudofsky UH, Longmate JA, Schiffenbauer J, Wakeland EK. Polygenic control of susceptibility to murine systemic lupus erythematosus. Immunity 1994;1:219–29. 13. Hahn BH. Animal models of systemic lupus erythematosus. In: Wallace DJ, Hahn BH, Quismorio FP, Klinenberg JR, editors. Dubois’ lupus erythematosus. Baltimore (MD): Williams & Wilkins; 1996. p. 339–80. 14. Riemekasten G, Kawald A, Weiss C, Meine A, Marell J, Klein R, et al. Strong acceleration of murine lupus by injection of the SmD183–119 peptide. Arthritis Rheum 2001;44:2435–45. 15. Eisenberg RA, Tan EM, Dixon FJ. Presence of anti-Sm reactivity in autoimmune mouse strains. J Exp Med 1978;147:582–7. 16. Suen JL, Wu CH, Chen YY, Wu WM, Chiang BL. Characterization of self-T-cell response and antigenic determinants of U1A protein with bone marrow-derived dendritic cells in NZB ⫻ NZW F1 mice. Immunology 2001;103:301–9. 17. Deshmukh US, Lewis JE, Gaskin F, Dhakephalkar PK, Kannapell CC, Waters ST, et al. Ro60 peptides induce antibodies to similar epitopes shared among lupus-related autoantigens. J Immunol 2000;164:6655–61. 18. Schneider C, Newman RA, Sutherland DR, Asser U, Greaves MF. A one-step purification of membrane proteins using a high efficiency immunomatrix. J Biol Chem 1982;257:10766–9. 19. Nakajima T, Uchida C, Anderson SF, Lee CG, Hurwitz J, Parvin JD, et al. RNA helicase A mediates association of CBP with RNA polymerase II. Cell 1997;90:1107–12. 20. Liao HJ, Kobayashi R, Mathews MB. Activities of adenovirus virus-associated RNAs: purification and characterization of RNA binding proteins. Proc Natl Acad Sci U S A 1998;95:8514–9. 21. Satoh M, Langdon JJ, Hamilton KJ, Richards HB, Panka D, Eisenberg RA, et al. Distinctive immune response patterns of human and murine autoimmune sera to U1 small nuclear ribonucleoprotein C protein. J Clin Invest 1996;97:2619–26. 22. Billings PB, Allen RW, Jensen FC, Hoch SO. Anti-RNP monoclonal antibodies derived from a mouse strain with lupus-like autoimmunity. J Immunol 1982;128:1176–80. 23. Bloom DD, Davignon JL, Retter MW, Shlomchik MJ, Pisetsky DS, Cohen PL, et al. V region gene analysis of anti-Sm hybridomas from MRL/Mp-lpr/lpr mice. J Immunol 1993;150:1591–610. 24. Habets WJ, Hoet MH, De Jong BA, Van der Kemp A, Van Venrooij WJ. Mapping of B cell epitopes on small nuclear ribonucleoproteins that react with human autoantibodies as well as with experimentally-induced mouse monoclonal antibodies. J Immunol 1989;143:2560–6. 25. Hamilton KJ, Satoh M, Swartz J, Richards HB, Reeves WH. Influence of microbial stimulation on hypergammaglobulinemia and autoantibody production in pristane-induced lupus. Clin Immunol Immunopathol 1998;86:271–9. 26. Satoh M, Weintraub JP, Yoshida H, Shaheen VM, Richards HB, Shaw M, et al. Fas and Fas ligand mutations inhibit autoantibody production in pristane-induced lupus. J Immunol 2000;165: 1036–43. 27. Satoh M, Behney KM, Mizutani A, Richards HB, Hutson AD, Reeves WH. Titration emulation: a computer-assisted technique that simplifies the quantification of anti-dsDNA antibodies using the Crithidia luciliae assay. Lupus 2001;10:632–6. 28. Mohan C, Alas E, Morel L, Yang P, Wakeland EK. Genetic dissection of SLE pathogenesis: Sle1 on murine chromosome 1 leads to a selective loss of tolerance to H2A/H2B/DNA subnucleosomes. J Clin Invest 1998;101:1362–72. 2244 29. Theofilopoulos AN, Dixon FJ. Murine models of systemic lupus erythematosus. Adv Immunol 1985;37:269–390. 30. Lee CG, Eki T, Okumura K, da Costa Soares V, Hurwitz J. Molecular analysis of the cDNA and genomic DNA encoding mouse RNA helicase A. Genomics 1998;47:365–71. 31. Seder RA, Gazzinelli R, Sher A, Paul WE. Interleukin 12 acts directly on CD4⫹ T cells to enhance priming for interferon ␥ production and diminishes interleukin 4 inhibition of such priming. Proc Natl Acad Sci U S A 1993;90:10188–92. 32. Spickard A III, Hirschmann JV. Exogenous lipoid pneumonia. Arch Intern Med 1994;154:686–92. 33. Dincsoy HP, Weesner RE, MacGee J. Lipogranulomas in nonfatty human livers: a mineral oil induced environmental disease. Am J Clin Pathol 1982;78:35–41. 34. Landas SK, Mitros FA, Furst DE, LaBrecque DR. Lipogranulomas and gold in the liver in rheumatoid arthritis. Am J Surg Pathol 1992;16:171–4. 35. Podda A. The adjuvanted influenza vaccines with novel adjuvants: experience with the MF59-adjuvanted vaccine. Vaccine 2001;19: 2673–80. 36. Tan EM, Kunkel HG. Characteristics of a soluble nuclear antigen precipitating with sera of patients with systemic lupus erythematosus. J Immunol 1966;96:464–71. 37. Kurata N, Tan EM. Identification of antibodies to nuclear acidic antigens by counterimmunoelectrophoresis. Arthritis Rheum 1976;19:574–80. 38. Bernstein RM, Bunn CC, Hughes GR, Francoeur AM, Mathews MB. Cellular protein and RNA antigens in autoimmune disease. Mol Biol Med 1984;2:105–20. 39. Satoh M, Langdon JJ, Chou CH, McCauliffe DP, Treadwell EL, Ogasawara T, et al. Characterization of the Su antigen, a macromolecular complex of 100/102 and 200-kDa proteins recognized by autoantibodies in systemic rheumatic diseases. Clin Immunol Immunopathol 1994;73:132–41. 40. Cohen PL, Eisenberg RA. Lpr and gld: single gene models of systemic autoimmunity and lymphoproliferative disease. Annu Rev Immunol 1991;9:243–69. 41. Satoh M, Shaheen VM, Kao PN, Okano T, Shaw M, Yoshida H, et al. Autoantibodies define a family of proteins with conserved double-stranded RNA-binding domains as well as DNA binding activity. J Biol Chem 1999;274:34598–604. 42. Takeda Y, Caudell P, Grady G, Wang G, Suwa A, Sharp GC, et al. Human RNA helicase A is a lupus autoantigen that is cleaved during apoptosis. J Immunol 1999;163:6269–74. 43. Tang H, Gaietta GM, Fischer WH, Ellisman MH, Wong-Staal F. YOSHIDA ET AL 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. A cellular cofactor for the constitutive transport element of type D retrovirus. Science 1997; 276:1412–5. Li J, Tang H, Mullen TM, Westberg C, Reddy TR, Rose DW, et al. A role for RNA helicase A in post-transcriptional regulation of HIV type 1. Proc Natl Acad Sci U S A 1999;96:709–14. Alleva DG, Kaser SB, Beller DI. Intrinsic defects in macrophage IL-12 production associated with immune dysfunction in the MRL/⫹⫹ and New Zealand Black/White F1 lupus-prone mice and the Leishmania major-susceptible BALB/c strain. J Immunol 1998;161:6878–84. Horwitz DA, Gray JD, Behrendsen SC, Kubin M, Rengaraju M, Ohtsuka K, et al. Decreased production of interleukin-12 and other Th1-type cytokines in patients with recent-onset systemic lupus erythematosus. Arthritis Rheum 1998;41:838–44. Haas C, Ryffel B, Le Hir M. IFN-␥ receptor deletion prevents autoantibody production and glomerulonephritis in lupus-prone (NZB ⫻ NZW)F1 mice. J Immunol 1998;160:3713–8. Jacob CO, van der Meide PH, McDevitt HO. In vivo treatment of (NZB ⫻ NZW)F1 lupus-like nephritis with monoclonal antibody to gamma interferon. J Exp Med 1987;166:798–803. Ozmen L, Roman D, Fountoulakis M, Schmid G, Ryffel B, Garotta G. Experimental therapy of systemic lupus erythematosus: the treatment of NZB/W mice with mouse soluble interferongamma receptor inhibits the onset of glomerulonephritis. Eur J Immunol 1995;25:6–12. Snapper CM, Paul WE. Interferon-gamma and B cell stimulatory factor-1 reciprocally regulate Ig isotype production. Science 1987; 236:944–7. Kitching AR, Holdsworth SR, Tipping PG. Crescentic glomerulonephritis—a manifestation of a nephritogenic Th1 response? Histol Histopathol 2000;15:993–1003. Huang FP, Feng GJ, Lindop G, Stott DI, Liew FY. The role of interleukin 12 and nitric oxide in the development of spontaneous autoimmune disease in MRL/MP-lpr/lpr mice. J Exp Med 1996; 183:1447–59. Seery JP, Carroll JM, Cattell V, Watt FM. Antinuclear autoantibodies and lupus nephritis in transgenic mice expressing interferon gamma in the epidermis. J Exp Med 1997;186:1451–9. Gilkeson GS, Conover J, Halpern M, Pisetsky DS, Feagin A, Klinman DM. Effects of bacterial DNA on cytokine production by (NZB/NZW)F1 mice. J Immunol 1998;161:3890–5. Gilkeson GS, Ruiz P, Pippen AM, Alexander AL, Lefkowith JB, Pisetsky DS. Modulation of renal disease in autoimmune NZB/ NZW mice by immunization with bacterial DNA. J Exp Med 1996;183:1389–97.