IFN- Induces Early Lethal Lupus in Preautoimmune (New Zealand Black x New Zealand White)F1 but Not in BALB/c Mice
http://www.100md.com
免疫学杂志 2005年第5期
Abstract
Recent studies indicate that IFN- is involved in pathogenesis of systemic lupus erythematosus. However, direct proof that IFN- is not only necessary, but also sufficient to induce lupus pathogenicity is lacking. In this study, we show that in vivo adenovector-mediated delivery of murine IFN- results in preautoimmune (New Zealand Black (NZB) x New Zealand White (NZW))F1, but not in normal, mice, in a rapid and severe disease with all characteristics of systemic lupus erythematosus. Anti-dsDNA Abs appeared as soon as day 10 after initiation of IFN- treatment. Proteinuria and death caused by glomerulonephritis occurred in all treated mice within, respectively, 9 and 18 wk, at a time when all untreated (NZB x NZW)F1 did not show any sign of disease. IFN- in vivo induced an overexpression of B lymphocyte stimulator in circulation at similar levels in both the preautoimmune and the normal mouse strains. All effects elicited by IFN- were dose dependent. (NZB x NZW)F1 infused with purified murine IFN- also showed acceleration of lupus. Thus, prolonged expression of IFN- in vivo induces early lethal lupus in susceptible animals.
Introduction
Systemic lupus erythematosus (SLE)5 is a prototype autoimmune disease characterized by a break of tolerance to nuclear components and profound alterations of the immune system (1, 2, 3, 4, 5). Dysregulation in activation of many cells of the immune system, in particular B and T lymphocytes, dendritic cells, and macrophages, may result in the development of pathogenic autoimmune response. Cytokines have been extensively studied in both human and murine lupus. In particular, the contribution of IL-10 and IFN- in lupus pathogenesis has been demonstrated (6, 7). Recently, type I IFNs (IFN-) have gained considerable interest as candidates in the etiopathogenic process of lupus (8, 9). Increased serum IFN- in SLE patients that correlates with both disease activity and severity has long been known (10, 11, 12). Moreover, early reports showed that a high proportion of mononuclear cells from SLE patients with IFN--positive or -negative sera expressed IFN--inducible 2-5A synthetase (13) and Mx proteins (14). The observations that IFN- treatment of patients with cancer or infectious diseases can cause development of a broad spectrum of different autoantibodies sometimes associated with overt organ-specific and systemic autoimmune diseases and more specifically anti-dsDNA autoantibodies and SLE have led to possible causative link between IFN- and lupus (15, 16). In the case of those patients developing SLE, discontinuation of IFN- treatment and prednisone administration led to remission of the disease (17).
The origin of, and mechanisms involved in, IFN- production in lupus remain obscure. Anti-dsDNA/DNA immune complexes and/or apoptotic bodies have been shown to induce IFN- production by plasmacytoid dendritic cells and might therefore constitute an amplification loop in disease (18, 19). A recent study showed that IFN- in SLE serum induced the differentiation of monocytes into dendritic cells with autoantigen-presenting capacity (20), supporting the concept that promoting dendritic cell maturation is a major mechanism by which IFN- may contribute to lupus pathogenesis (21, 22). The importance of IFN- in lupus etiology has been further substantiated by gene array analysis in which up-regulation of numerous IFN-dependent genes in PBMCs from lupus patients was reported, constituting an overall IFN signature in SLE (23, 24). However, all this body of work is correlative of, but does not demonstrate a direct cause-effect relationship between type I IFN and SLE.
Inasmuch as safe and efficient anti-murine IFN- (mIFN-) are not currently available, the best demonstration of the role of IFN in SLE has recently been provided by the analysis of type I IFN receptor-deleted mice. Indeed, crosses between lupus-prone and type I IFN receptor knockout mice have shown beneficial effects on the spontaneous course of disease (25, 26), providing evidence that IFN- is an important mediator in lupus disease. Yet, these studies do not prove that type I IFNs are sufficient to induce disease. In this regard, early studies showed that treatment of lupus-prone mice with systemic injections of semipurified preparations of IFN- for several months only led to modest acceleration in autoantibody production and severity of glomerulonephritis (27, 28). Whether this discrepancy is due to the purity, doses, timing of injections, or severity of mouse lupus models is questioned. Therefore, the direct importance of type I IFNs in lupus pathogenicity still remains an open question. In the present study, we re-evaluated the effects of IFN- in vivo on the course of lupus response and asked whether IFN- can break tolerance and initiate a de novo autoimmune reaction, but also enhance an existing, but silent autoimmune process and precipitate a clinically overt disease. Our data show that continuous in vivo expression of IFN- precipitates in preautoimmune (New Zealand Black (NZB) x New Zealand White (NZW))F1 mice, but not in normal BALB/c mice, the autoimmune process, and kidney damage, leading to premature death from severe immune complex glomerulonephritis. Our data support the notion that sustained IFN- in vivo results, in susceptible individuals, in a brisk disease that has all characteristics of SLE.
Materials and Methods
Generation of IFN- adenovirus (Adv)
The recombinant Adv vector containing the mIFN- subtype 5 cDNA under the control of the CMV promoter/enhancer (AdCMVmIFN5) was obtained from Qbiogene. The virus was cloned by plaque assay on 293 cells and functionally tested for mIFN-5 protein expression by ELISA (PBL Biomedical Laboratories) before being amplified. AdCMVpLpA lacking a cDNA insert was used as a negative control. Viruses were propagated on 293 cells and purified by two rounds of cesium chloride density gradient centrifugations. The recombinant virus titer was determined by OD at 260 nm (1A260 = 1.1 x 1012 particles/ml), and aliquots were stored at –80°C in a buffer containing 10 mM Tris (pH 8.0), 2 mM MgCl2, and 5% sucrose. For all in vivo studies, aliquots of virus preparations were diluted in saline before injection.
Mice and IFN- treatment
Female NZB and male NZW were purchased from Harlan, and a colony of (NZB x NZW)F1 mice (NZB/W) was developed in specific pathogen-free barrier facility. In some instances, experiments were conducted on NZB/W purchased from Harlan. BALB/c mice were used as nonautoimmune controls. Female mice (8–16 mice/group) were treated at 9–11 wk of age with a single i.v. injection in the retro-orbital plexus of 1 x 1010 IFN- Adv particles. Controls received the same amount of control Adv particles. In some experiments, mice were injected with gradual doses of IFN- Adv ranging from 0.03 x 1010 to 1 x 1010 viral particles (vp). Alternatively, mice were treated with successive 14-day periods of continuous infusion of purified mouse IFN-A (PBL) administered by micro-osmotic pump (model 1002; ALZET). Previous report showed that delivery of IFN- using a micro-osmotic pump achieved rapid and stable expression of IFN- in vivo (29). Osmotic pumps were loaded under sterile conditions with 105 U of IFN-A (sp. act.: 4.3 x 106 U/mg) in 100 μl of PBS containing 0.1% BSA, and were implanted in s.c. tissue of the back. Control animals received pumps filled with PBS/BSA. Every 14 days for a total delivery of 8–10 wk, the pump was replaced by a new one containing the same amount of IFN-A. To test whether the adenovector might play an adjuvant role in IFN--mediated effects on autoimmunity, mice implanted with pumps filled with IFN-A were injected or not with 1010 vp of control, null Adv i.v., once, at the time the first pump was implanted. IFN-A and 0.1% BSA solutions contained endotoxin levels <0.40 EU/ml (corresponding to 0.04 EU/per pump), as assessed using the Limulus amebocyte lysate detection kit (BioWhittaker). All in vivo experiments received approval from Baylor Research Institute Institutional Animal Care and Use Committee.
ELISAs
Mouse IFN-. Serum IFN- levels were measured by ELISA (PBL) following the manufacturer’s instructions. The detection limit was 12.5 pg/ml.
Mouse B lymphocyte stimulator (BLyS) ELISA. Serum BLyS levels were measured by ELISA (Alexis), following the manufacturer’s recommendation. Ninety-six-well microtiter plates (Nunc Immuno Maxisorp) were coated overnight at 4°C with 5 μg/ml anti-mouse BLyS mAb (clone 5A8). Plates were then blocked with PBS (pH 7.2) supplemented with 10% FCS (PBS/FCS) for 2 h at room temperature (RT). Sera serially diluted in PBS/10% FCS/0.05% Tween 20 (1/2–1/10) or standard rBLyS (serially diluted in PBS/FCS from 4 to 0.0625 ng/ml) were incubated for 2 h at RT. After washing, biotin-conjugated anti-BLyS mAb (clone 1C9) was added, followed by incubation with avidin HRP, and plates were developed with ABTS substrate (Southern Biotechnology Associates). The OD was read at 405 nm using a Spectra Max 190 spectrophotometer (Molecular Devices), and the concentration of BLyS in samples was determined based on the standard curve.
Total Igs. Ninety-six-well microtiter plates were coated overnight at 4°C with affinity-purified goat anti-mouse IgG (Jackson ImmunoResearch Laboratories). Plates were then blocked with PBS (pH 7.2) supplemented with 10% FCS for 2 h at RT. Serial dilutions of kidney eluates (1/200–1/3,200), sera (1/70,000–1/350,000), or standard IgG (0.05–0.0006 μg/ml) (Southern Biotechnology Associates) were incubated for 2 h at RT. After washing, peroxidase-conjugated goat anti-mouse IgG (Sigma-Aldrich) was added, and plates developed with ABTS substrate (Southern Biotechnology Associates). The OD was read at 405 nm, and the concentration of IgG in samples was determined based on the IgG standard curve.
Anti-ssDNA and anti-dsDNA Abs. Serum autoantibody levels were assessed by ELISA, as previously described (30). Briefly, ELISA plates pretreated with poly(L-lysine) (100 μg/ml) were coated with phage dsDNA (5 μg/ml in PBS, pH 7.2) (Roche Diagnostics). ssDNA was prepared by boiling phage dsDNA (250 μg/ml) for 10 min and immediately chilling on ice, and was plated at 5 μg/ml in PBS. After overnight incubation at 4°C, plates were blocked with PBS/10% FCS or, alternatively, a PBS/20% solution of blocker casein (Pierce) to block samples from the ALZET experiments (i.e., serum samples might contain Abs developed against BSA added to stabilize IFN-). Sera (starting from 1/100 up to 1/1500 dilution) or kidney eluates (1/4–1/264) were incubated for 2 h at room temperature. Bound IgG was detected with peroxidase-conjugated goat anti-mouse IgG (Fc specific) (1/1000) (Sigma-Aldrich) added to the plates for 90 min. Binding was measured by adding ABTS substrate solution, and the OD was read at 405 nm. A positive control
NZB/W serum was run in serial dilutions on each plate to allow standardization. For kidney eluates, raw OD was converted to arbitrary units (AU)/ml by setting the OD of the 1/100 dilution of the standard serum to 1000 AU/ml.
In some experiments, sera diluted 1/10 were assessed for anti-dsDNA Abs using Crithidia luciliae immunofluorescence detection (Immunoconcepts).
Evaluation of renal disease
Proteinuria. Urine was tested for proteinuria using dipstick (Chemstrip 2 GP; Roche Diagnostics). Proteinuria scored as + for levels of 30 mg/dl, ++ for 100 mg/dl, and +++ for levels 500 mg/dl. Mice were considered to have proteinuria if two consecutive urine samples scored 2+.
Renal eluates. Four individual IFN--treated NZB/W premortem mice (harvested, respectively, at 52, 60, 66, and 90 days after initiation of IFN- treatment), four age-matched control Adv NZB/W, and four proteinuric, unmanipulated NZB/W (43 wk of age) were used for elution of Igs from the kidneys. Renal elution was performed, as described (31), with slight modifications. Briefly, blood vessels were perfused via the left ventricle with 20 ml of ice-cold PBS. Kidneys were harvested, the medulla removed, and the cortex was weighed, minced, and washed three times with cold PBS. Kidneys were resuspended into 5 M urea/0.15 M glycine, pH 2.8, and the suspensions were sonicated and slowly stirred at 4°C overnight. Kidney eluates were collected after centrifugation (20,000 x g x 30 min) and filtered first through a 70-μm cell strainer (BD Discovery Labware), followed by a passage though an Acrodisc syringe filter (PALL). Eluates were then neutralized with 1/10 (v/v) 2 M Tris, pH 8.0, and dialyzed extensively against PBS. Renal eluates were aliquoted and stored at –80°C until use.
Kidney histology. Kidneys fixed with Formalin were embedded in paraffin, and sections were stained with periodic acid Schiff (PAS) and read blind by a pathologist (A. Weinberg). For IgG and C3 staining, frozen sections (4–5 μm) were fixed with acetone, washed with PBS, and blocked with PBS/10% normal rat serum/5% FCS. Sections were stained with a mix of FITC-conjugated rat anti-mouse IgG1, IgG2a, Ig2b, and IgG3 mAbs (each at 1/100 dilution) (BD Pharmingen) for 1 h at RT. For complement staining, sections were first blocked with PBS/10% goat serum/5% FCS and incubated with FITC-conjugated goat F(ab')2 to mouse C3 (1/100) (Cappel, MP Biomedicals). Sections were mounted with fluoromount (Vector Laboratories) and analyzed by microscopy.
Statistics
Results were compared using Mann-Whitney U test or the log-rank test for proteinuria and survival. A p value of <0.05 was considered significant.
Results
Long-lasting IFN- delivery in vivo precipitates clinical signs of lupus in susceptible mice
We investigated the effects of prolonged mouse IFN- administration in NZB/W and BALB/c mice using Adv-mediated gene transfer technology. A replication-deficient murine rIFN- Adv was administered once at 1010 vp/mouse to young, preautoimmune NZB/W and age-matched control BALB/c mice. Control groups for each strain consisted of mice treated with PBS or with 1010 vp of control Adv. Mice were bled at days 3, 10, and 42 after Adv delivery. As shown in Fig. 1A, blood concentrations of IFN- at day 3 reached median levels of 714 and 614 pg/ml in NZB/W (n = 33) and BALB/c (n = 21) mice, respectively (NS). On day 10, serum IFN- levels dropped to 195 pg/ml in NZB/W and 258 pg/ml in BALB/c mice (NS). IFN- was still present in blood 42 days after administration of IFN- Adv (133 pg/ml in NZB/W and 76 pg/ml in BALB/c (NS)). No IFN- was found in circulation of mice injected with control Adv. Thus, IFN- Adv treatment provides sustained IFN- concentrations in blood of both lupus-prone and nonautoimmune mice.
FIGURE 1. In vivo Adv-mediated delivery of mIFN- induces the brisk occurrence of proteinuria and death in NZB/W lupus-prone mice. A, Ten-week-old BALB/c mice (open) and NZB/W mice (filled) were injected with 1 x 1010 control Adv vp (, ) or with 1 x 1010 IFN- Adv vp (, ?). Mice were bled at days 3, 10, and 42 posttreatment, and IFN- serum levels were measured by ELISA. Each dot represents an individual mouse, and lines show median values. Results from 33 NZB/W and 20 BALB/c mice are shown. B, Upper panel, Proportion of mice without proteinuria (1+). Lower panel, Survival rate. n = 16 in each experimental group for each mouse strain.
Yet, the clinical outcome was strikingly different (Fig. 1B). BALB/c mice injected with IFN- Adv showed no sign of proteinuria or premature death. In sharp contrast, young NZB/W mice became rapidly proteinuric. As early as 28 days after IFN- treatment, 4 of 16 mice already had heavy proteinuria (>500 mg/dl; +++) and 3 of 16 had mild proteinuria (100 mg/dl; ++). At day 35, >80% of IFN--treated NZB/W mice were proteinuric; 9 of 16 (56%) had levels >500 mg/dl (+++) and 4 of 16 (25%) had 100 mg/dl (++). All mice were severely proteinuric 63 days after initiation of IFN- treatment (i.e., mice were then 19 wk old), while the first mice that became proteinuric in the NZB/W control groups were 25 wk old. Thus, prolonged expression of IFN- in vivo induces early proteinuria in NZB/W lupus-prone mice.
IFN- precipitated the death of NZB/W mice. The first death occurred at day 40 after initiation of IFN- treatment. Fifty percent (8 of 16) of IFN--treated NZB/W mice died by day 66, and all mice were dead by day 130 (i.e., at 29 wk of age). At this age, all control NZB/W (32 of 32) were alive. Thus, sustained IFN- in vivo accelerates renal damage and death in genetically susceptible animals.
IFN- induces glomerulonephritis in NZB/W mice
We next characterized the kidney alterations elicited by IFN- in NZB/W mice. Kidneys of IFN--treated NZB/W mice with heavy proteinuria (>500 mg/dl; 40 days posttreatment) looked pale and enlarged, features not observed in NZB/W control groups. Histological examination revealed a severe diffuse proliferative glomerulonephritis with either segmental or global involvement of all glomeruli. Mesangial hypercellularity and sclerosis were associated with glomerulus enlargement and segmental or global thickening of capillary basement membranes. Wire loop lesions and segmental glomerulosclerosis were easily identified (Fig. 2A). Many capillary lumens were filled with monocytes, lymphocytes, and rare neutrophils. Trichrome stain demonstrated large numbers of deposits in the mesangium and subendothelial regions as well as intramembranous and epimembranous deposits (data not shown). These features were qualitatively and quantitatively similar to the lesions that developed spontaneously in NZB/W mice (our results not shown) (32). Kidneys from untreated and control Adv-injected NZB/W looked normal (Fig. 2A). Immunofluorescence staining of glomeruli from IFN--treated NZB/W showed that IgG and C3 intensely decorated the mesangium and peripheral capillary loops in a granular pattern. Again, NZB/W control groups showed neither IgG nor C3 deposits. Histological and immunofluorescence analysis of BALB/c kidneys demonstrated complete lack of kidney lesions whether mice were treated with IFN- or control Adv.
FIGURE 2. In vivo Adv-mediated delivery of mIFN- induces the rapid development of immune complex glomerulonephritis. A, Kidney sections prepared from proteinuric NZB/W mice 40 days post-IFN- Adv treatment, or from age-matched unmanipulated and control Adv mice, were stained with PAS (upper panel), a mixture of FITC-conjugated anti-IgG1, anti-IgG2a, anti-IgG2b, and anti-IgG3 Abs (middle panel), or with FITC-conjugated anti-C3 Abs (lower panel). Results shown are from one mouse in each experimental group, and staining is representative of three mice/group, run in parallel. Note the deposits along the capillaries and in the mesangium (arrows). B, Eluates from kidneys of four control Adv mice () (negative control); four sick, unmanipulated, 43-wk-old, NZB/W ( ) (positive control); and four individual proteinuric IFN--Adv-treated NZB/W mice (harvested at 52, 60, 66, and 90 days posttreatment) (?) were tested for total IgG and for anti-nuclear IgG autoantibody specificity. Thick horizontal lines represent mean values. Vertical lines represent positive SD.
Lupus nephritis might be due to deposition of immune complexes composed of anti-dsDNA Abs and DNA (see for review Ref. 33). We therefore eluted Igs from kidneys of heavily proteinuric, IFN--treated NZB/W and tested them for autoantibody specificity. Kidneys of IFN--treated NZB/W mice contained amounts of total IgG (5.5 ± 4.5 μg per 100 mg of cortex (mean ± SD)) comparable to those found in 43-wk-old, spontaneously proteinuric NZB/W (7.4 ± 4.1; NS) (Fig. 2B). Kidney eluates from IFN--treated NZB/W contained anti-ssDNA and anti-dsDNA IgG autoantibodies at levels comparable to those found in spontaneously proteinuric NZB/W. For instance, anti-dsDNA IgG were 13.8 ± 9.9 (mean ± SD) in IFN--treated NZB/W vs 14.5 ± 15.9 AU/μg IgG in NZB/W (NS). Thus, histology, deposition in glomeruli of IgG and complement C3, along with the presence of kidney-associated IgG with anti-nuclear specificity, suggest that IFN- promotes a typical lupus glomerulonephritis in susceptible animals.
Anti-dsDNA IgG develop specifically in IFN--treated NZB/W mice
Treatment of NZB/W and BALB/c mice with control Adv did not affect the levels of circulating IgG when compared with untreated animals (Fig. 3). In contrast, administration of IFN- Adv induced in both strains a sharp increase of circulating IgG. In BALB/c mice, total IgG increased by 3-fold at day 10 (p < 0.0001 vs control Adv) and by 1.6-fold at day 42 (p = 0.02). Likewise, there was a 2.5-fold increase in IgG in NZB/W, at day 10 (p < 0.00005), that returned to basal values at day 42 (Fig. 3A). Note that basal IgG levels in control NZB/W are 3 times as high as in BALB/c. IFN- induced an increase in BALB/c IgM, while IgM decreased in NZB/W (data not shown). Serum anti-nuclear autoantibody reactivity showed striking differences between BALB/c and NZB/W mice (Fig. 3, B and C). Twenty-one days after IFN- treatment (e.g., at the onset of proteinuria in NZB/W), serum anti-ssDNA IgG were augmented in both strains, but at a much higher level in NZB/W than in BALB/c mice (0.694 ± 0.35; OD ± SD; in NZB/W (p < 0.01 vs controls) vs 0.209 ± 0.057 in BALB/c (p < 0.05 vs controls)). Remarkably, IFN- promoted the development of anti-dsDNA IgG autoantibodies exclusively in the lupus-prone strain; five of nine NZB/W mice had anti-dsDNA values equal or above threshold (established as the mean ± 3 SD in controls), while none of the IFN--treated BALB/c were positive. Thus, IFN- promotes production of anti-dsDNA autoantibodies in NZB/W, albeit at low levels.
FIGURE 3. In vivo Adv-mediated delivery of mIFN- induces production of anti-dsDNA Abs specifically in NZB/W mice. BALB/c and NZB/W mice, either unmanipulated ( ) or injected with 1 x 1010 control Adv vp ( ) or 1 x 1010 IFN- Adv vp (), were bled at days 10 and 42 after initiation of treatment. A, Total serum IgG was measured by ELISA. Values represent the mean ± SEM from 8–10 mice in each group. *, p < 0.05; **, p < 0.01; ****, p < 0.0001. B and C, Sera from mice 21 days posttreatment were tested for anti-nuclear IgG autoantibodies. Each dot represents an individual mouse. The middle panel (purple) shows anti-ssDNA and anti-dsDNA IgG values in sera of unmanipulated 25- to 28-wk-old, proteinuric (++/+++) NZB/W mice (n = 9–11). Thin black horizontal lines represent mean values. Gray horizontal lines represent the threshold values set as the mean OD ± 3 SD in untreated mice.
BLyS is increased in serum of IFN--treated mice
Overexpression of BLyS, an essential component of B cell homeostasis, results in SLE-like and Sjogren-like phenotype in mice (34, 35, 36, 37, 38). Likewise, NZB/W mice have increased circulating BLyS during the onset and progression of lupus, and its neutralization with soluble transmembrane activator and CAML interactor-Ig fusion protein inhibits the development of proteinuria and prolongs survival of the animals (34). Our results (Fig. 4) show that treatment of NZB/W and BALB/c mice with control Adv did not affect the levels of circulating BLyS when compared with untreated animals. In contrast, administration of IFN- Adv induced in both strains increased levels of BLyS in circulation as soon as day 24 posttreatment. Circulating BLyS further increased at day 48 to reach in NZB/W mice as well as in BALB/c mice levels observed in sick 10-mo-old NZB/W. Thus, IFN- treatment induces sustained BLyS concentrations in blood of both lupus-prone and nonautoimmune mice.
FIGURE 4. In vivo Adv-mediated delivery of mIFN- induces prolonged expression of BLyS in the sera of NZB/W and BALB/c mice. Ten-week-old BALB/c (open) and NZB/W mice (filled) were injected with 1 x 1010 control Adv vp (, ) or with 1 x 1010 IFN- Adv vp (, ?). Mice were bled at days 21 and 42 posttreatment, and BLyS serum levels were measured by ELISA. Each dot represents an individual mouse, and lines show median values. *, p < 0.05; **, p < 0.01; ***, p < 0.001; **** p < 0.00001; *****, p < 0.00001. The right panel shows BLyS values in sera of 1-mo-old unmanipulated () and 10-mo-old proteinuric () NZB/W mice (n = 7–12).
Very low doses of IFN- can induce rapid lethal lupus
Young NZB/W mice were administered progressively lower doses of IFN- Adv down to 0.03 x 1010 vp. As shown in Fig. 5A, the levels of circulating IFN- detected at days 3 and 10 after initiation of treatment indeed gradually decreased with decreasing doses of IFN- Adv. Overall, treated mice can be divided into two groups: a high IFN- Adv group (e.g., 0.3–1 x 1010 Adv vp range) and a low IFN- Adv group (e.g., 0.03–0.1 x 1010 Adv vp range). Indeed, there was no statistical difference in the levels of circulating IFN- at day 10 elicited by the different IFN- Adv doses within each of the two groups.
FIGURE 5. Outcome of disease is dependent on the dose of administered mIFN- Adv: dose-response effects. Ten-week-old NZB/W mice were injected with gradual doses of IFN- Adv. A, Mice were bled at days 3 and 10 after treatment, and serum IFN- levels were measured by ELISA. *, p < 0.05; **, p < 0.01. B, Proportion of mice without proteinuria (1+). Black line, 1 x 1010 IFN- Adv vp (n = 16); dark gray line, 0.3 x 1010 (n = 16); gray line, 0.1 x 1010 (n = 16); light gray line, 0.03 x 1010 (n = 8). Note the delay in occurrence of proteinuria as the dose of IFN- Adv decreases. Significant statistical differences are p = 0.05 for 1 vs 0.1; p < 0.02 for 1 vs 0.03; p < 0.04 for 0.3 vs 0.03, x 1010 Adv vp. Dashed line represents proteinuria in control Adv NZB/W mice (n = 8). C, Survival rate. Note the delay in occurrence of death as the dose of IFN- Adv decreases. Significant statistical differences are p = 0.002 for 1 vs 0.1; p = 0.005 for 1 vs 0.03; p = 0.005 for 0.3 vs 0.1; and p = 0.004 for 0.3 vs 0.03, x 1010 Adv vp. Dashed line represents survival rate in control Adv NZB/W mice (n = 8). D and E, Ten-week-old NZB/W mice were injected with 0.3–1 x 1010 IFN- Adv vp (high IFN-) (D) or with 0.03–0.1 x 1010 IFN- Adv vp (low IFN-) (E), and serum was analyzed for anti-dsDNA Abs at indicated times. Each line represents one mouse, and the results show 16 individual mice in each group. Note that most of the mice present a peak in anti-DNA preceding the occurrence of death. , Represent median values.
Remarkably, progressively lower doses of IFN- Adv progressively delayed occurrence of proteinuria and death (Fig. 5, B and C). Occurrence of proteinuria (Fig. 5B) apparently reflected the differences in IFN- levels observed between the high and low IFN- groups. Fifty percent of the mice (16 of 32) treated with high IFN- became proteinuric within 27 days after initiation of treatment, while 50% of the mice (12 of 24) treated with low IFN- became proteinuric within 34 days (p < 0.004). Survival was even a more sensitive index of in vivo IFN- levels as decreases in IFN- Adv doses significantly delayed occurrence of death (Fig. 5C). Overall, all mice treated with high IFN- were dead by day 90–103, while mice treated with low IFN- were dead by day 135–154 (p < 0.0001). Remarkably, the lowest dose of IFN- Adv (0.03 x 1010 vp), which elicited levels of circulating IFN- that were undetectable 10 days after initiation of treatment (Fig. 5A), yet induced proteinuria and death in all treated animals (Fig. 5, B and C).
Thus, severity of lupus depends upon the dose of IFN-, although minute amounts of IFN- precipitate a lethal lupus.
Occurrence and levels of circulating anti-dsDNA Abs are dependent upon the dose of IFN- Adv (Fig. 5, D and E). Anti-dsDNA IgG appeared as soon as day 10 after initiation of IFN- treatment in the high IFN- group; 9 of 16 mice were positive (e.g., above threshold; mean ± 3 SD in controls), while only 1 of 16 mice was positive in the low IFN- group (p < 0.05). Later on, between days 10 and 38, 15 of 16 mice in the high IFN- group and 16 of 16 in the low IFN- group became positive, although at lower levels than that observed in sick, unmanipulated NZB x NZW mice (see Fig. 3C). Notably, anti-dsDNA reached high titers between days 52 and 94 in animals that survived in the low IFN- group (Fig. 5E).
Sustained release of low doses of purified IFN- induces lupus disease
Although the above data demonstrate that occurrence of disease is related to the dose of IFN- in vivo, it is possible that the adenovector plays a role of adjuvant for IFN-. Thus, young NZB/W mice were infused with a low dose of purified IFN- using micro-osmotic pumps. Continuous delivery of IFN- (105 U/per pump) did not yield measurable levels of IFN- in blood (data not shown). Yet, this low dose of IFN- induced substantial levels of anti-dsDNA IgG Abs in seven to eight of nine mice, as illustrated by positive Crithidia immunofluorescence staining and ELISA detection (Fig. 6A and Table I). Remarkably, all mice, except one (eight of nine), became proteinuric within 39–78 days after initiation of treatment with purified IFN- (Table I), while all controls at that time were devoid of proteinuria. These proteinuric mice showed glomerulonephritis with IgG and C3 deposits in the mesangium and along the basement membrane (Fig. 6B). There was no difference in the occurrence of anti-dsDNA IgG or of proteinuria whether or not mice had also received the control adenovector. Thus, IFN- alone induces a dramatic acceleration of lupus in susceptible individuals.
FIGURE 6. Continuous infusion of IFN- in vivo precipitates lupus disease. Ten-week-old NZB/W mice were implanted s.c. with micro-osmotic pumps containing either PBS (n = 9) or mIFN--A (n = 9). Fourteen-day calibrated pumps were replaced three to four times to deliver constant amounts of IFN--A in vivo for a total period of 8–10 wk. A, Immunofluorescence in Crithidia (sera diluted 1/10). B, Kidney sections of mice infused with IFN--A for 56 days (8 wk) and harvested at day 60. Upper panel, PAS staining. Middle panel, Sections from control or IFN--A-treated mice were stained with a mixture of FITC-conjugated anti-IgG1, anti-IgG2a, anti-IgG2b, and anti-IgG3. Lower panel, Staining with anti-C3.
Table I. Continuous delivery of low doses of purified IFN- induces precipitation of lupus
Discussion
The present study demonstrates that the prolonged expression of IFN- in vivo leads to dramatic acceleration of lupus manifestations in young lupus-prone mice. Using Adv-mediated gene transfer of mouse IFN-5, elevated serum levels of IFN- were rapidly obtained, which persisted for at least 42 days.
This prolonged expression of IFN- in vivo results, in genetically susceptible preautoimmune lupus mice, in a brisk disease that has all characteristics of SLE. Anti-dsDNA Abs appeared in circulation as soon as day 10 after initiation of IFN- treatment and mice rapidly died of lupus glomerulonephritis, as illustrated by glomerular hypercellularity, heavy capillary and mesangial IgG and C3 deposits (32, 39, 40, 41), and the presence of anti-nuclear autoantibodies in kidney eluates (31, 42). Disease acceleration mediated by IFN- is striking as proteinuria and death occurred 12 and 21 wk earlier, respectively, than in untreated NZB/W mice (our colony) (32, 39).
Systemic administration of purified preparations of mIFN- (three times per week from birth/1 mo to 33–37 wk of age) has been shown to induce only a slight acceleration of disease in NZB/W and NZB lupus-prone strains (27, 28, 43). In particular, IFN- did not accelerate the development of proteinuria, and acceleration of death was modest. These studies using semipurified preparations of type I IFN (e.g., containing both IFN- and IFN-) did not directly attribute to IFN- the key role in the acceleration of the disease. Furthermore, the effect of a prolonged expression of IFN- in a normal genetic background was not evaluated. Our study addresses both points. It is very likely that the difference with our results holds in both the duration of, and persistence in, expression of IFN- in vivo. Indeed, it has been shown that 8 h after a single injection of IFN-, 90% of the IFN- is removed from the circulation (29). Thus, our model, by favoring IFN- impregnation in vivo, might represent a model of IFN- dysregulation seen in human lupus (10, 11, 12, 13, 23, 24).
IFN- in vivo enhances total IgG Abs both in BALB/c and NZB/W mice, a result in accordance with recent report (44). Despite the induction of anti-ssDNA IgG autoantibodies in BALB/c mice, IFN- failed, however, to induce the appearance of anti-dsDNA IgG and glomerulonephritis, suggesting that IFN- is not sufficient to induce a full SLE-like phenotype in a normal genetic background. We cannot, however, rule out that if followed for several months after initiation of IFN- treatment, BALB/c mice would not have shown signs of SLE-like disease. The human studies have also suggested that IFN- treatment of patients with cancer or infectious diseases can induce full-blown SLE in certain susceptible individuals (15, 16). Several predisposing loci have been identified for several lupus strains (reviewed in Refs. 3 , 5). However, genetic basis conferring to certain animals the susceptibility for IFN- to trigger autoimmunity has yet to be determined. Recently described IFN-inducible gene, ifi202, might be, among other susceptibility genes, a potential candidate that accounts for the observed IFN--mediated disease (45). The influence of the genetic background appears to be crucial as it has been very recently reported that type I IFNs suppress autoimmunity in MRLlpr/lpr mice (46).
IFN- in vivo induced the expression of BLyS to a similar level in both the autoimmune and the normal strains. BLyS is a member of the TNF ligand family that is a potent coactivator of B cells in vitro and in vivo (37). Mice transgenic for BLyS develop symptoms characteristic of SLE, such as the presence of high levels of rheumatoid factors, circulating immune complexes, anti-DNA autoantibodies, and Ig deposition in the kidneys, and suggest that dysregulation of BLyS expression may be a critical element in the chain of events leading to autoimmunity (34, 35, 36, 37, 38). Consistent with a possible role for BLyS in Ab-mediated autoimmune diseases, in patients with SLE, rheumatoid arthritis, and Sjogren’s syndrome, an association was found between elevated levels of BLyS in the blood and the severity of disease (38, 47, 48). Given our finding that mice treated with IFN- develop sustained levels of BLyS in circulation, it is tempting to suggest that high BLyS levels observed in human SLE are a consequence of an overproduction of IFN- in vivo. However, overexpression of both BLyS and IFN- does not lead to the development of a lupus-like phenotype in BALB/c mice. Several explanations can be put forward to explain such an absence of effect. Among those, the genetic background (e.g., most of BLyS transgenic mice are on BL/6 background (35, 37, 38), time for development of autoimmunity (e,g., several months in BLyS Tg mice), and duration and levels of BLyS expression in blood of Tg mice may explain why BALB/c behave as a BLyS-resistant mouse strain in our experimental conditions. It is also possible that the association of IFN- and BLyS precipitates autoimmunity in lupus-prone mice, while, on the contrary, IFN- would counteract BLyS-mediated autoimmunity in normal mice. These hypotheses are currently under investigation.
In regard with the severity of IFN--induced glomerulonephritis in NZB/W mice, it is intriguing that these mice injected with high IFN- Adv had low levels of serum anti-dsDNA Abs. A first explanation is that IFN- induces glomerulonephritis independently of anti-dsDNA autoantibodies, as it has been recently reported in different autoimmune models (49, 50) and in glomerular lesions induced by IFN at birth in normal mice (51). In this regard, it is possible that IFN- might have a direct deleterious effect on NZB/W kidneys, as proposed previously to explain IFN--mediated glomerular lesions in newborn normal mice (52). A second explanation is that in the brisk IFN--driven lupus process, nephritogenic Abs rapidly deposit into the kidneys or are lost in the urine during renal flare, and might therefore become barely detectable in circulation (53, 54, 55). The presence of anti-dsDNA Abs in substantial amounts in kidneys of IFN--treated NZB/W mice would rather favor the second hypothesis, although the two mechanisms may not be mutually exclusive.
If Adv gene transfer technology has major advantages (long-term protein expression, low cost), it, however, may induce the generation of cellular immune responses, inflammation (56), and, possibly, tissue damage (57). The Adv infection might indeed provide an adjuvant role in the acceleration of SLE observed upon IFN- delivery, and this has implications for the role of infection in triggering lupus disease. However, our data provide several lines of evidence to assign little role of the virus in the brisk lupus disease we see upon prolonged in vivo expression of IFN-. First, it is worth mentioning that serum IFN- elicited by the highest dose of IFN- Adv did not induce any premature death or hepatic toxicity (data not shown). Second, none of BALB/c or NZB/W mice injected with 1 x 1010 vp of empty Adv developed any biological nor clinical sign of lupus. Third, we could not detect the presence of anti-Adv Abs (as measured by ELISA) (58) in kidney eluates from 1 x 1010 IFN- Adv-treated NZB/W mice (data not shown). These Abs, complexed or not with adenoviral particles, could very well have participated in the IFN--driven immune complex glomerulonephritis (59). Finally, a strong line of evidence is provided by our observation that continuous infusion of rmIFN- also accelerated lupus autoimmunity. Altogether, this suggests that IFN- per se is the driving force leading to the early expression of lupus manifestations.
It is intriguing that, at variance with SLE patients, lupus-prone mice, of the several strains tested, do not have circulating IFN- (our unpublished results) (26). It is possible, however, that serum IFN- levels are below the threshold of detection and/or IFN- is preferentially expressed into (inflamed) target tissues, as it has been recently shown in cutaneous lupus erythematosus lesions (60). Nonetheless, it is important to stress that SLE patients show a generalized up-regulation of IFN-related genes regardless of the presence or absence of IFN- in their sera (23, 24).
Recent work has suggested a role for type I IFNs in murine lupus. Indeed, NZB congenic lupus-prone mice and C57BL/6 lpr/lpr mice lacking IFNAR-1, the -chain of the common receptor for type I IFNs, have reduced lupus-like disease (25, 26). Our present work not only adds evidence to the concept that type I IFNs are major effectors in the pathogenesis of lupus, but also directly demonstrates a major role of the genetic background conferring to IFN- its pathogenic role.
In conclusion, the present study provides the direct demonstration that prolonged in vivo expression of IFN- induces, in susceptible individuals, a brisk disease that has all characteristics of SLE. Moreover, we have established a new experimental model in which lupus disease is amenable in a few weeks and which will permit to dissect further the etiology and the cellular and molecular mechanisms that initiate and sustain the lupus autoimmune response.
Disclosures
The authors have no financial conflict of interest.
Acknowledgments
We thank Dr. R. Gerard (University of Texas Southwestern, Dallas, TX) for his precious advice on adenovectors. We also thank C. Joseph, S. Clayton, and F. Marches for their expert technical help. We are also grateful to Dr. C. Mohan for his help in kidney elution technique and for providing us with lupus mouse sera. We thank Drs. V. Pascual, K. Palucka, I. Gresser, A. Dalloul, and D. Emilie for stimulating discussions and critical reading of the manuscript.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 J.B. is the recipient of the Max and Gayle Clampitt chair of Immunology Research. This work was supported by postdoctoral fellowship from the Assistance Publique des Hopitaux de Paris (to A.M.).
2 J.B. and S.K. codirected this work and are co-corresponding authors.
3 Address correspondence and reprint requests to Dr. Jacques Banchereau, Baylor Institute for Immunology Research, 3434 Live Oak, Dallas, TX 75204. E-mail address: jacquesb{at}baylorhealth.edu
4 S.K. is on leave to Institut National de la Santé et de la Recherche Médicale Unité 131, Institut Paris-Sud des Cytokines, Clamart, France. E-mail address: sophie.koutouzov{at}inserm.ipsc.u-psud.fr
5 Abbreviations used in this paper: SLE, systemic lupus erythematosus; Adv, adenovirus; BLyS, B lymphocyte stimulator; mIFN, murine IFN; PAS, periodic acid-Schiff; RT, room temperature; vp, viral particle; AU, arbitrary units.
Received for publication September 2, 2004. Accepted for publication October 26, 2004.
References
Lipsky, P. E.. 2001. Systemic lupus erythematosus: an autoimmune disease of B cell hyperactivity. Nat. Immunol. 2:764.
Desai-Mehta, A., L. Lu, R. Ramsey-Goldman, S. K. Datta. 1996. Hyperexpression of CD40 ligand by B and T cells in human lupus and its role in pathogenic autoantibody production. J. Clin. Invest. 97:2063.
Marrack, P., J. Kappler, B. L. Kotzin. 2001. Autoimmune disease: why and where it occurs. Nat. Med. 7:899.
Shlomchik, M. J., J. E. Craft, M. J. Mamula. 2001. From T to B and back again: positive feedback in systemic autoimmune disease. Nat. Rev. Immunol. 1:147.
Wakeland, E. K., K. Liu, R. R. Graham, T. W. Behrens. 2001. Delineating the genetic basis of systemic lupus erythematosus. Immunity 15:397
Llorente, L., W. Zou, Y. Levy, Y. Richaud-Patin, J. Wijdenes, J. Alcocer-Varela, B. Morel-Fourrier, J. C. Brouet, D. Alarcon-Segovia, P. Galanaud, et al 1995. Role of interleukin 10 in the B lymphocyte hyperactivity and autoantibody production of human systemic lupus erythematosus. J. Exp. Med. 181:839.
Lawson, B. R., G. J. Prud’homme, Y. Chang, H. A. Gardner, J. Kuan, D. H. Kono, A. N. Theofilopoulos. 2000. Treatment of murine lupus with cDNA encoding IFN-R/Fc. J. Clin. Invest. 106:207
Stewart, T. A.. 2003. Neutralizing interferon as a therapeutic approach to autoimmune diseases. Cytokine Growth Factor Rev. 14:139.
Pascual, V., J. Banchereau, A. K. Palucka. 2003. The central role of dendritic cells and interferon- in SLE. Curr. Opin. Rheumatol. 15:548.
Ioannou, Y., D. A. Isenberg. 2000. Current evidence for the induction of autoimmune rheumatic manifestations by cytokine therapy. Arthritis Rheum. 43:1431.
Ronnblom, L. E., G. V. Alm, K. E. Oberg. 1990. Possible induction of systemic lupus erythematosus by interferon- treatment in a patient with a malignant carcinoid tumor. J. Intern. Med. 227:207
Vallin, H., A. Perers, G. V. Alm, L. Ronnblom. 1999. Anti-double-stranded DNA antibodies and immunostimulatory plasmid DNA in combination mimic the endogenous IFN- inducer in systemic lupus erythematosus. J. Immunol. 163:6306.
Bave, U., G. V. Alm, L. Ronnblom. 2000. The combination of apoptotic U937 cells and lupus IgG is a potent IFN- inducer. J. Immunol. 165:3519.
Blanco, P., A. K. Palucka, M. Gill, V. Pascual, J. Banchereau. 2001. Induction of dendritic cell differentiation by IFN- in systemic lupus erythematosus. Science 294:1540.
Steinman, R. M., D. Hawiger, M. C. Nussenzweig. 2003. Tolerogenic dendritic cells. Annu. Rev. Immunol. 21:685.
Banchereau, J., V. Pascual, A. K. Palucka. 2004. Autoimmunity through cytokine-induced dendritic cell activation. Immunity 20:539.
Baechler, E. C., F. M. Batliwalla, G. Karypis, P. M. Gaffney, W. A. Ortmann, K. J. Espe, K. B. Shark, W. J. Grande, K. M. Hughes, V. Kapur, et al 2003. Interferon-inducible gene expression signature in peripheral blood cells of patients with severe lupus. Proc. Natl. Acad. Sci. USA 100:2610
Bennett, L., A. K. Palucka, E. Arce, V. Cantrell, J. Borvak, J. Banchereau, V. Pascual. 2003. Interferon and granulopoiesis signatures in systemic lupus erythematosus blood. J. Exp. Med. 197:711.
Braun, D., P. Geraldes, J. Demengeot. 2003. Type I interferon controls the onset and severity of autoimmune manifestations in lpr mice. J. Autoimmun. 20:15.
Jovelin, F., G. Mostoslavsky, Z. Amoura, H. Chabre, D. Gilbert, D. Eilat, J. F. Bach, S. Koutouzov. 1998. Early anti-nucleosome autoantibodies from a single MRL+/+ mouse: fine specificity, V gene structure and pathogenicity. Eur. J. Immunol. 28:3411.
Xie, C., Z. Liang, S. Chang, C. Mohan. 2003. Use of a novel elution regimen reveals the dominance of polyreactive antinuclear autoantibodies in lupus kidneys. Arthritis Rheum. 48:2343
Dixon, F. J., M. B. Oldstone, G. Tonietti. 1971. Pathogenesis of immune complex glomerulonephritis of New Zealand mice. J. Exp. Med. 134:65s.
Foster, M. H., V. R. Kelley. 1999. Lupus nephritis: update on pathogenesis and disease mechanisms. Semin. Nephrol. 19:173.
Gross, J. A., J. Johnston, S. Mudri, R. Enselman, S. R. Dillon, K. Madden, W. Xu, J. Parrish-Novak, D. Foster, C. Lofton-Day, et al 2000. TACI and BCMA are receptors for a TNF homologue implicated in B-cell autoimmune disease. Nature 404:995
Mackay, F., S. A. Woodcock, P. Lawton, C. Ambrose, M. Baetscher, P. Schneider, J. Tschopp, J. L. Browning. 1999. Mice transgenic for BAFF develop lymphocytic disorders along with autoimmune manifestations. J. Exp. Med. 190:1697
Khare, S. D., I. Sarosi, X. Z. Xia, S. McCabe, K. Miner, I. Solovyev, N. Hawkins, M. Kelley, D. Chang, G. Van, et al 2000. Severe B cell hyperplasia and autoimmune disease in TALL-1 transgenic mice. Proc. Natl. Acad. Sci. USA 97:3370.
Mackay, F., P. Schneider, P. Rennert, J. Browning. 2003. BAFF and APRIL: a tutorial on B cell survival. Annu. Rev. Immunol. 21:231.
Groom, J., S. L. Kalled, A. H. Cutler, C. Olson, S. A. Woodcock, P. Schneider, J. Tschopp, T. G. Cachero, M. Batten, J. Wheway, et al 2002. Association of BAFF/BLyS overexpression and altered B cell differentiation with Sjogren’s syndrome. J. Clin. Invest. 109:59
Lambert, P. H., F. J. Dixon. 1968. Pathogenesis of the glomerulonephritis of NZB/W mice. J. Exp. Med. 127:507.
Granholm, N. A., T. Cavallo. 1990. Mechanism of localization of immune complexes in NZB/W mice with early nephritis. Autoimmunity 8:17.
Sasaki, T., A. Hatakeyama, S. Shibata, H. Osaki, M. Suzuki, K. Horie, Y. Kitagawa, K. Yoshinaga. 1991. Heterogeneity of immune complex-derived anti-DNA antibodies associated with lupus nephritis. Kidney Int. 39:746.
Ebling, F., B. H. Hahn. 1980. Restricted subpopulations of DNA antibodies in kidneys of mice with systemic lupus: comparison of antibodies in serum and renal eluates. Arthritis Rheum. 23:392.
Sergiescu, D., I. Cerutti, E. Efthymiou, A. Kahan, C. Chany. 1979. Adverse effects of interferon treatment on the life span of NZB mice. Biomedicine 31:48.
Le Bon, A., G. Schiavoni, G. D’Agostino, I. Gresser, F. Belardelli, D. F. Tough. 2001. Type I interferons potently enhance humoral immunity and can promote isotype switching by stimulating dendritic cells in vivo. Immunity 14:461.
Rozzo, S. J., J. D. Allard, D. Choubey, T. J. Vyse, S. Izui, G. Peltz, B. L. Kotzin. 2001. Evidence for an interferon-inducible gene, Ifi202, in the susceptibility to systemic lupus. Immunity 15:435.
Hron, J. D., S. L. Peng. 2004. Type I IFN protects against murine lupus. J. Immunol. 173:2134.
Zhang, J., V. Roschke, K. P. Baker, Z. Wang, G. S. Alarcon, B. J. Fessler, H. Bastian, R. P. Kimberly, T. Zhou. 2001. Cutting edge: a role for B lymphocyte stimulator in systemic lupus erythematosus. J. Immunol. 166:6.
Ho, A., L. S. Magder, S. G. Barr, M. Petri. 2001. Decreases in anti-double-stranded DNA levels are associated with concurrent flares in patients with systemic lupus erythematosus. Arthritis Rheum. 44:2342.
Macanovic, M., M. B. Hogarth, P. J. Lachmann. 1999. Anti-DNA antibodies in the urine of lupus nephritis patients. Nephrol. Dial. Transplant. 14:1418.
Isenberg, D. A., M. R. Ehrenstein, C. Longhurst, J. K. Kalsi. 1994. The origin, sequence, structure, and consequences of developing anti-DNA antibodies: a human perspective. Arthritis Rheum. 37:169.
Yang, Y., F. A. Nunes, K. Berencsi, E. E. Furth, E. Gonczol, J. M. Wilson. 1994. Cellular immunity to viral antigens limits E1-deleted adenoviruses for gene therapy. Proc. Natl. Acad. Sci. USA 91:4407.
Wright, N. G., H. J. Cornwell. 1983. Experimental canine adenovirus glomerulonephritis: histological, immunofluorescence and ultrastructural features of the early glomerular changes. Br. J. Exp. Pathol. 64:312.
Gahery-Segard, H., F. Farace, D. Godfrin, J. Gaston, R. Lengagne, T. Tursz, P. Boulanger, J. G. Guillet. 1998. Immune response to recombinant capsid proteins of adenovirus in humans: antifiber and anti-penton base antibodies have a synergistic effect on neutralizing activity. J. Virol. 72:2388
Wright, N. G., W. I. Morrison, H. Thompson, H. J. Cornwell. 1973. Experimental adenovirus immune complex glomerulonephritis. Br. J. Exp. Pathol. 54:628.
Farkas, L., K. Beiske, F. Lund-Johansen, P. Brandtzaeg, F. L. Jahnsen. 2001. Plasmacytoid dendritic cells (natural interferon-/-producing cells) accumulate in cutaneous lupus erythematosus lesions. Am. J. Pathol. 159:237.(Alexis Mathian Arthur Wei)
Recent studies indicate that IFN- is involved in pathogenesis of systemic lupus erythematosus. However, direct proof that IFN- is not only necessary, but also sufficient to induce lupus pathogenicity is lacking. In this study, we show that in vivo adenovector-mediated delivery of murine IFN- results in preautoimmune (New Zealand Black (NZB) x New Zealand White (NZW))F1, but not in normal, mice, in a rapid and severe disease with all characteristics of systemic lupus erythematosus. Anti-dsDNA Abs appeared as soon as day 10 after initiation of IFN- treatment. Proteinuria and death caused by glomerulonephritis occurred in all treated mice within, respectively, 9 and 18 wk, at a time when all untreated (NZB x NZW)F1 did not show any sign of disease. IFN- in vivo induced an overexpression of B lymphocyte stimulator in circulation at similar levels in both the preautoimmune and the normal mouse strains. All effects elicited by IFN- were dose dependent. (NZB x NZW)F1 infused with purified murine IFN- also showed acceleration of lupus. Thus, prolonged expression of IFN- in vivo induces early lethal lupus in susceptible animals.
Introduction
Systemic lupus erythematosus (SLE)5 is a prototype autoimmune disease characterized by a break of tolerance to nuclear components and profound alterations of the immune system (1, 2, 3, 4, 5). Dysregulation in activation of many cells of the immune system, in particular B and T lymphocytes, dendritic cells, and macrophages, may result in the development of pathogenic autoimmune response. Cytokines have been extensively studied in both human and murine lupus. In particular, the contribution of IL-10 and IFN- in lupus pathogenesis has been demonstrated (6, 7). Recently, type I IFNs (IFN-) have gained considerable interest as candidates in the etiopathogenic process of lupus (8, 9). Increased serum IFN- in SLE patients that correlates with both disease activity and severity has long been known (10, 11, 12). Moreover, early reports showed that a high proportion of mononuclear cells from SLE patients with IFN--positive or -negative sera expressed IFN--inducible 2-5A synthetase (13) and Mx proteins (14). The observations that IFN- treatment of patients with cancer or infectious diseases can cause development of a broad spectrum of different autoantibodies sometimes associated with overt organ-specific and systemic autoimmune diseases and more specifically anti-dsDNA autoantibodies and SLE have led to possible causative link between IFN- and lupus (15, 16). In the case of those patients developing SLE, discontinuation of IFN- treatment and prednisone administration led to remission of the disease (17).
The origin of, and mechanisms involved in, IFN- production in lupus remain obscure. Anti-dsDNA/DNA immune complexes and/or apoptotic bodies have been shown to induce IFN- production by plasmacytoid dendritic cells and might therefore constitute an amplification loop in disease (18, 19). A recent study showed that IFN- in SLE serum induced the differentiation of monocytes into dendritic cells with autoantigen-presenting capacity (20), supporting the concept that promoting dendritic cell maturation is a major mechanism by which IFN- may contribute to lupus pathogenesis (21, 22). The importance of IFN- in lupus etiology has been further substantiated by gene array analysis in which up-regulation of numerous IFN-dependent genes in PBMCs from lupus patients was reported, constituting an overall IFN signature in SLE (23, 24). However, all this body of work is correlative of, but does not demonstrate a direct cause-effect relationship between type I IFN and SLE.
Inasmuch as safe and efficient anti-murine IFN- (mIFN-) are not currently available, the best demonstration of the role of IFN in SLE has recently been provided by the analysis of type I IFN receptor-deleted mice. Indeed, crosses between lupus-prone and type I IFN receptor knockout mice have shown beneficial effects on the spontaneous course of disease (25, 26), providing evidence that IFN- is an important mediator in lupus disease. Yet, these studies do not prove that type I IFNs are sufficient to induce disease. In this regard, early studies showed that treatment of lupus-prone mice with systemic injections of semipurified preparations of IFN- for several months only led to modest acceleration in autoantibody production and severity of glomerulonephritis (27, 28). Whether this discrepancy is due to the purity, doses, timing of injections, or severity of mouse lupus models is questioned. Therefore, the direct importance of type I IFNs in lupus pathogenicity still remains an open question. In the present study, we re-evaluated the effects of IFN- in vivo on the course of lupus response and asked whether IFN- can break tolerance and initiate a de novo autoimmune reaction, but also enhance an existing, but silent autoimmune process and precipitate a clinically overt disease. Our data show that continuous in vivo expression of IFN- precipitates in preautoimmune (New Zealand Black (NZB) x New Zealand White (NZW))F1 mice, but not in normal BALB/c mice, the autoimmune process, and kidney damage, leading to premature death from severe immune complex glomerulonephritis. Our data support the notion that sustained IFN- in vivo results, in susceptible individuals, in a brisk disease that has all characteristics of SLE.
Materials and Methods
Generation of IFN- adenovirus (Adv)
The recombinant Adv vector containing the mIFN- subtype 5 cDNA under the control of the CMV promoter/enhancer (AdCMVmIFN5) was obtained from Qbiogene. The virus was cloned by plaque assay on 293 cells and functionally tested for mIFN-5 protein expression by ELISA (PBL Biomedical Laboratories) before being amplified. AdCMVpLpA lacking a cDNA insert was used as a negative control. Viruses were propagated on 293 cells and purified by two rounds of cesium chloride density gradient centrifugations. The recombinant virus titer was determined by OD at 260 nm (1A260 = 1.1 x 1012 particles/ml), and aliquots were stored at –80°C in a buffer containing 10 mM Tris (pH 8.0), 2 mM MgCl2, and 5% sucrose. For all in vivo studies, aliquots of virus preparations were diluted in saline before injection.
Mice and IFN- treatment
Female NZB and male NZW were purchased from Harlan, and a colony of (NZB x NZW)F1 mice (NZB/W) was developed in specific pathogen-free barrier facility. In some instances, experiments were conducted on NZB/W purchased from Harlan. BALB/c mice were used as nonautoimmune controls. Female mice (8–16 mice/group) were treated at 9–11 wk of age with a single i.v. injection in the retro-orbital plexus of 1 x 1010 IFN- Adv particles. Controls received the same amount of control Adv particles. In some experiments, mice were injected with gradual doses of IFN- Adv ranging from 0.03 x 1010 to 1 x 1010 viral particles (vp). Alternatively, mice were treated with successive 14-day periods of continuous infusion of purified mouse IFN-A (PBL) administered by micro-osmotic pump (model 1002; ALZET). Previous report showed that delivery of IFN- using a micro-osmotic pump achieved rapid and stable expression of IFN- in vivo (29). Osmotic pumps were loaded under sterile conditions with 105 U of IFN-A (sp. act.: 4.3 x 106 U/mg) in 100 μl of PBS containing 0.1% BSA, and were implanted in s.c. tissue of the back. Control animals received pumps filled with PBS/BSA. Every 14 days for a total delivery of 8–10 wk, the pump was replaced by a new one containing the same amount of IFN-A. To test whether the adenovector might play an adjuvant role in IFN--mediated effects on autoimmunity, mice implanted with pumps filled with IFN-A were injected or not with 1010 vp of control, null Adv i.v., once, at the time the first pump was implanted. IFN-A and 0.1% BSA solutions contained endotoxin levels <0.40 EU/ml (corresponding to 0.04 EU/per pump), as assessed using the Limulus amebocyte lysate detection kit (BioWhittaker). All in vivo experiments received approval from Baylor Research Institute Institutional Animal Care and Use Committee.
ELISAs
Mouse IFN-. Serum IFN- levels were measured by ELISA (PBL) following the manufacturer’s instructions. The detection limit was 12.5 pg/ml.
Mouse B lymphocyte stimulator (BLyS) ELISA. Serum BLyS levels were measured by ELISA (Alexis), following the manufacturer’s recommendation. Ninety-six-well microtiter plates (Nunc Immuno Maxisorp) were coated overnight at 4°C with 5 μg/ml anti-mouse BLyS mAb (clone 5A8). Plates were then blocked with PBS (pH 7.2) supplemented with 10% FCS (PBS/FCS) for 2 h at room temperature (RT). Sera serially diluted in PBS/10% FCS/0.05% Tween 20 (1/2–1/10) or standard rBLyS (serially diluted in PBS/FCS from 4 to 0.0625 ng/ml) were incubated for 2 h at RT. After washing, biotin-conjugated anti-BLyS mAb (clone 1C9) was added, followed by incubation with avidin HRP, and plates were developed with ABTS substrate (Southern Biotechnology Associates). The OD was read at 405 nm using a Spectra Max 190 spectrophotometer (Molecular Devices), and the concentration of BLyS in samples was determined based on the standard curve.
Total Igs. Ninety-six-well microtiter plates were coated overnight at 4°C with affinity-purified goat anti-mouse IgG (Jackson ImmunoResearch Laboratories). Plates were then blocked with PBS (pH 7.2) supplemented with 10% FCS for 2 h at RT. Serial dilutions of kidney eluates (1/200–1/3,200), sera (1/70,000–1/350,000), or standard IgG (0.05–0.0006 μg/ml) (Southern Biotechnology Associates) were incubated for 2 h at RT. After washing, peroxidase-conjugated goat anti-mouse IgG (Sigma-Aldrich) was added, and plates developed with ABTS substrate (Southern Biotechnology Associates). The OD was read at 405 nm, and the concentration of IgG in samples was determined based on the IgG standard curve.
Anti-ssDNA and anti-dsDNA Abs. Serum autoantibody levels were assessed by ELISA, as previously described (30). Briefly, ELISA plates pretreated with poly(L-lysine) (100 μg/ml) were coated with phage dsDNA (5 μg/ml in PBS, pH 7.2) (Roche Diagnostics). ssDNA was prepared by boiling phage dsDNA (250 μg/ml) for 10 min and immediately chilling on ice, and was plated at 5 μg/ml in PBS. After overnight incubation at 4°C, plates were blocked with PBS/10% FCS or, alternatively, a PBS/20% solution of blocker casein (Pierce) to block samples from the ALZET experiments (i.e., serum samples might contain Abs developed against BSA added to stabilize IFN-). Sera (starting from 1/100 up to 1/1500 dilution) or kidney eluates (1/4–1/264) were incubated for 2 h at room temperature. Bound IgG was detected with peroxidase-conjugated goat anti-mouse IgG (Fc specific) (1/1000) (Sigma-Aldrich) added to the plates for 90 min. Binding was measured by adding ABTS substrate solution, and the OD was read at 405 nm. A positive control
NZB/W serum was run in serial dilutions on each plate to allow standardization. For kidney eluates, raw OD was converted to arbitrary units (AU)/ml by setting the OD of the 1/100 dilution of the standard serum to 1000 AU/ml.
In some experiments, sera diluted 1/10 were assessed for anti-dsDNA Abs using Crithidia luciliae immunofluorescence detection (Immunoconcepts).
Evaluation of renal disease
Proteinuria. Urine was tested for proteinuria using dipstick (Chemstrip 2 GP; Roche Diagnostics). Proteinuria scored as + for levels of 30 mg/dl, ++ for 100 mg/dl, and +++ for levels 500 mg/dl. Mice were considered to have proteinuria if two consecutive urine samples scored 2+.
Renal eluates. Four individual IFN--treated NZB/W premortem mice (harvested, respectively, at 52, 60, 66, and 90 days after initiation of IFN- treatment), four age-matched control Adv NZB/W, and four proteinuric, unmanipulated NZB/W (43 wk of age) were used for elution of Igs from the kidneys. Renal elution was performed, as described (31), with slight modifications. Briefly, blood vessels were perfused via the left ventricle with 20 ml of ice-cold PBS. Kidneys were harvested, the medulla removed, and the cortex was weighed, minced, and washed three times with cold PBS. Kidneys were resuspended into 5 M urea/0.15 M glycine, pH 2.8, and the suspensions were sonicated and slowly stirred at 4°C overnight. Kidney eluates were collected after centrifugation (20,000 x g x 30 min) and filtered first through a 70-μm cell strainer (BD Discovery Labware), followed by a passage though an Acrodisc syringe filter (PALL). Eluates were then neutralized with 1/10 (v/v) 2 M Tris, pH 8.0, and dialyzed extensively against PBS. Renal eluates were aliquoted and stored at –80°C until use.
Kidney histology. Kidneys fixed with Formalin were embedded in paraffin, and sections were stained with periodic acid Schiff (PAS) and read blind by a pathologist (A. Weinberg). For IgG and C3 staining, frozen sections (4–5 μm) were fixed with acetone, washed with PBS, and blocked with PBS/10% normal rat serum/5% FCS. Sections were stained with a mix of FITC-conjugated rat anti-mouse IgG1, IgG2a, Ig2b, and IgG3 mAbs (each at 1/100 dilution) (BD Pharmingen) for 1 h at RT. For complement staining, sections were first blocked with PBS/10% goat serum/5% FCS and incubated with FITC-conjugated goat F(ab')2 to mouse C3 (1/100) (Cappel, MP Biomedicals). Sections were mounted with fluoromount (Vector Laboratories) and analyzed by microscopy.
Statistics
Results were compared using Mann-Whitney U test or the log-rank test for proteinuria and survival. A p value of <0.05 was considered significant.
Results
Long-lasting IFN- delivery in vivo precipitates clinical signs of lupus in susceptible mice
We investigated the effects of prolonged mouse IFN- administration in NZB/W and BALB/c mice using Adv-mediated gene transfer technology. A replication-deficient murine rIFN- Adv was administered once at 1010 vp/mouse to young, preautoimmune NZB/W and age-matched control BALB/c mice. Control groups for each strain consisted of mice treated with PBS or with 1010 vp of control Adv. Mice were bled at days 3, 10, and 42 after Adv delivery. As shown in Fig. 1A, blood concentrations of IFN- at day 3 reached median levels of 714 and 614 pg/ml in NZB/W (n = 33) and BALB/c (n = 21) mice, respectively (NS). On day 10, serum IFN- levels dropped to 195 pg/ml in NZB/W and 258 pg/ml in BALB/c mice (NS). IFN- was still present in blood 42 days after administration of IFN- Adv (133 pg/ml in NZB/W and 76 pg/ml in BALB/c (NS)). No IFN- was found in circulation of mice injected with control Adv. Thus, IFN- Adv treatment provides sustained IFN- concentrations in blood of both lupus-prone and nonautoimmune mice.
FIGURE 1. In vivo Adv-mediated delivery of mIFN- induces the brisk occurrence of proteinuria and death in NZB/W lupus-prone mice. A, Ten-week-old BALB/c mice (open) and NZB/W mice (filled) were injected with 1 x 1010 control Adv vp (, ) or with 1 x 1010 IFN- Adv vp (, ?). Mice were bled at days 3, 10, and 42 posttreatment, and IFN- serum levels were measured by ELISA. Each dot represents an individual mouse, and lines show median values. Results from 33 NZB/W and 20 BALB/c mice are shown. B, Upper panel, Proportion of mice without proteinuria (1+). Lower panel, Survival rate. n = 16 in each experimental group for each mouse strain.
Yet, the clinical outcome was strikingly different (Fig. 1B). BALB/c mice injected with IFN- Adv showed no sign of proteinuria or premature death. In sharp contrast, young NZB/W mice became rapidly proteinuric. As early as 28 days after IFN- treatment, 4 of 16 mice already had heavy proteinuria (>500 mg/dl; +++) and 3 of 16 had mild proteinuria (100 mg/dl; ++). At day 35, >80% of IFN--treated NZB/W mice were proteinuric; 9 of 16 (56%) had levels >500 mg/dl (+++) and 4 of 16 (25%) had 100 mg/dl (++). All mice were severely proteinuric 63 days after initiation of IFN- treatment (i.e., mice were then 19 wk old), while the first mice that became proteinuric in the NZB/W control groups were 25 wk old. Thus, prolonged expression of IFN- in vivo induces early proteinuria in NZB/W lupus-prone mice.
IFN- precipitated the death of NZB/W mice. The first death occurred at day 40 after initiation of IFN- treatment. Fifty percent (8 of 16) of IFN--treated NZB/W mice died by day 66, and all mice were dead by day 130 (i.e., at 29 wk of age). At this age, all control NZB/W (32 of 32) were alive. Thus, sustained IFN- in vivo accelerates renal damage and death in genetically susceptible animals.
IFN- induces glomerulonephritis in NZB/W mice
We next characterized the kidney alterations elicited by IFN- in NZB/W mice. Kidneys of IFN--treated NZB/W mice with heavy proteinuria (>500 mg/dl; 40 days posttreatment) looked pale and enlarged, features not observed in NZB/W control groups. Histological examination revealed a severe diffuse proliferative glomerulonephritis with either segmental or global involvement of all glomeruli. Mesangial hypercellularity and sclerosis were associated with glomerulus enlargement and segmental or global thickening of capillary basement membranes. Wire loop lesions and segmental glomerulosclerosis were easily identified (Fig. 2A). Many capillary lumens were filled with monocytes, lymphocytes, and rare neutrophils. Trichrome stain demonstrated large numbers of deposits in the mesangium and subendothelial regions as well as intramembranous and epimembranous deposits (data not shown). These features were qualitatively and quantitatively similar to the lesions that developed spontaneously in NZB/W mice (our results not shown) (32). Kidneys from untreated and control Adv-injected NZB/W looked normal (Fig. 2A). Immunofluorescence staining of glomeruli from IFN--treated NZB/W showed that IgG and C3 intensely decorated the mesangium and peripheral capillary loops in a granular pattern. Again, NZB/W control groups showed neither IgG nor C3 deposits. Histological and immunofluorescence analysis of BALB/c kidneys demonstrated complete lack of kidney lesions whether mice were treated with IFN- or control Adv.
FIGURE 2. In vivo Adv-mediated delivery of mIFN- induces the rapid development of immune complex glomerulonephritis. A, Kidney sections prepared from proteinuric NZB/W mice 40 days post-IFN- Adv treatment, or from age-matched unmanipulated and control Adv mice, were stained with PAS (upper panel), a mixture of FITC-conjugated anti-IgG1, anti-IgG2a, anti-IgG2b, and anti-IgG3 Abs (middle panel), or with FITC-conjugated anti-C3 Abs (lower panel). Results shown are from one mouse in each experimental group, and staining is representative of three mice/group, run in parallel. Note the deposits along the capillaries and in the mesangium (arrows). B, Eluates from kidneys of four control Adv mice () (negative control); four sick, unmanipulated, 43-wk-old, NZB/W ( ) (positive control); and four individual proteinuric IFN--Adv-treated NZB/W mice (harvested at 52, 60, 66, and 90 days posttreatment) (?) were tested for total IgG and for anti-nuclear IgG autoantibody specificity. Thick horizontal lines represent mean values. Vertical lines represent positive SD.
Lupus nephritis might be due to deposition of immune complexes composed of anti-dsDNA Abs and DNA (see for review Ref. 33). We therefore eluted Igs from kidneys of heavily proteinuric, IFN--treated NZB/W and tested them for autoantibody specificity. Kidneys of IFN--treated NZB/W mice contained amounts of total IgG (5.5 ± 4.5 μg per 100 mg of cortex (mean ± SD)) comparable to those found in 43-wk-old, spontaneously proteinuric NZB/W (7.4 ± 4.1; NS) (Fig. 2B). Kidney eluates from IFN--treated NZB/W contained anti-ssDNA and anti-dsDNA IgG autoantibodies at levels comparable to those found in spontaneously proteinuric NZB/W. For instance, anti-dsDNA IgG were 13.8 ± 9.9 (mean ± SD) in IFN--treated NZB/W vs 14.5 ± 15.9 AU/μg IgG in NZB/W (NS). Thus, histology, deposition in glomeruli of IgG and complement C3, along with the presence of kidney-associated IgG with anti-nuclear specificity, suggest that IFN- promotes a typical lupus glomerulonephritis in susceptible animals.
Anti-dsDNA IgG develop specifically in IFN--treated NZB/W mice
Treatment of NZB/W and BALB/c mice with control Adv did not affect the levels of circulating IgG when compared with untreated animals (Fig. 3). In contrast, administration of IFN- Adv induced in both strains a sharp increase of circulating IgG. In BALB/c mice, total IgG increased by 3-fold at day 10 (p < 0.0001 vs control Adv) and by 1.6-fold at day 42 (p = 0.02). Likewise, there was a 2.5-fold increase in IgG in NZB/W, at day 10 (p < 0.00005), that returned to basal values at day 42 (Fig. 3A). Note that basal IgG levels in control NZB/W are 3 times as high as in BALB/c. IFN- induced an increase in BALB/c IgM, while IgM decreased in NZB/W (data not shown). Serum anti-nuclear autoantibody reactivity showed striking differences between BALB/c and NZB/W mice (Fig. 3, B and C). Twenty-one days after IFN- treatment (e.g., at the onset of proteinuria in NZB/W), serum anti-ssDNA IgG were augmented in both strains, but at a much higher level in NZB/W than in BALB/c mice (0.694 ± 0.35; OD ± SD; in NZB/W (p < 0.01 vs controls) vs 0.209 ± 0.057 in BALB/c (p < 0.05 vs controls)). Remarkably, IFN- promoted the development of anti-dsDNA IgG autoantibodies exclusively in the lupus-prone strain; five of nine NZB/W mice had anti-dsDNA values equal or above threshold (established as the mean ± 3 SD in controls), while none of the IFN--treated BALB/c were positive. Thus, IFN- promotes production of anti-dsDNA autoantibodies in NZB/W, albeit at low levels.
FIGURE 3. In vivo Adv-mediated delivery of mIFN- induces production of anti-dsDNA Abs specifically in NZB/W mice. BALB/c and NZB/W mice, either unmanipulated ( ) or injected with 1 x 1010 control Adv vp ( ) or 1 x 1010 IFN- Adv vp (), were bled at days 10 and 42 after initiation of treatment. A, Total serum IgG was measured by ELISA. Values represent the mean ± SEM from 8–10 mice in each group. *, p < 0.05; **, p < 0.01; ****, p < 0.0001. B and C, Sera from mice 21 days posttreatment were tested for anti-nuclear IgG autoantibodies. Each dot represents an individual mouse. The middle panel (purple) shows anti-ssDNA and anti-dsDNA IgG values in sera of unmanipulated 25- to 28-wk-old, proteinuric (++/+++) NZB/W mice (n = 9–11). Thin black horizontal lines represent mean values. Gray horizontal lines represent the threshold values set as the mean OD ± 3 SD in untreated mice.
BLyS is increased in serum of IFN--treated mice
Overexpression of BLyS, an essential component of B cell homeostasis, results in SLE-like and Sjogren-like phenotype in mice (34, 35, 36, 37, 38). Likewise, NZB/W mice have increased circulating BLyS during the onset and progression of lupus, and its neutralization with soluble transmembrane activator and CAML interactor-Ig fusion protein inhibits the development of proteinuria and prolongs survival of the animals (34). Our results (Fig. 4) show that treatment of NZB/W and BALB/c mice with control Adv did not affect the levels of circulating BLyS when compared with untreated animals. In contrast, administration of IFN- Adv induced in both strains increased levels of BLyS in circulation as soon as day 24 posttreatment. Circulating BLyS further increased at day 48 to reach in NZB/W mice as well as in BALB/c mice levels observed in sick 10-mo-old NZB/W. Thus, IFN- treatment induces sustained BLyS concentrations in blood of both lupus-prone and nonautoimmune mice.
FIGURE 4. In vivo Adv-mediated delivery of mIFN- induces prolonged expression of BLyS in the sera of NZB/W and BALB/c mice. Ten-week-old BALB/c (open) and NZB/W mice (filled) were injected with 1 x 1010 control Adv vp (, ) or with 1 x 1010 IFN- Adv vp (, ?). Mice were bled at days 21 and 42 posttreatment, and BLyS serum levels were measured by ELISA. Each dot represents an individual mouse, and lines show median values. *, p < 0.05; **, p < 0.01; ***, p < 0.001; **** p < 0.00001; *****, p < 0.00001. The right panel shows BLyS values in sera of 1-mo-old unmanipulated () and 10-mo-old proteinuric () NZB/W mice (n = 7–12).
Very low doses of IFN- can induce rapid lethal lupus
Young NZB/W mice were administered progressively lower doses of IFN- Adv down to 0.03 x 1010 vp. As shown in Fig. 5A, the levels of circulating IFN- detected at days 3 and 10 after initiation of treatment indeed gradually decreased with decreasing doses of IFN- Adv. Overall, treated mice can be divided into two groups: a high IFN- Adv group (e.g., 0.3–1 x 1010 Adv vp range) and a low IFN- Adv group (e.g., 0.03–0.1 x 1010 Adv vp range). Indeed, there was no statistical difference in the levels of circulating IFN- at day 10 elicited by the different IFN- Adv doses within each of the two groups.
FIGURE 5. Outcome of disease is dependent on the dose of administered mIFN- Adv: dose-response effects. Ten-week-old NZB/W mice were injected with gradual doses of IFN- Adv. A, Mice were bled at days 3 and 10 after treatment, and serum IFN- levels were measured by ELISA. *, p < 0.05; **, p < 0.01. B, Proportion of mice without proteinuria (1+). Black line, 1 x 1010 IFN- Adv vp (n = 16); dark gray line, 0.3 x 1010 (n = 16); gray line, 0.1 x 1010 (n = 16); light gray line, 0.03 x 1010 (n = 8). Note the delay in occurrence of proteinuria as the dose of IFN- Adv decreases. Significant statistical differences are p = 0.05 for 1 vs 0.1; p < 0.02 for 1 vs 0.03; p < 0.04 for 0.3 vs 0.03, x 1010 Adv vp. Dashed line represents proteinuria in control Adv NZB/W mice (n = 8). C, Survival rate. Note the delay in occurrence of death as the dose of IFN- Adv decreases. Significant statistical differences are p = 0.002 for 1 vs 0.1; p = 0.005 for 1 vs 0.03; p = 0.005 for 0.3 vs 0.1; and p = 0.004 for 0.3 vs 0.03, x 1010 Adv vp. Dashed line represents survival rate in control Adv NZB/W mice (n = 8). D and E, Ten-week-old NZB/W mice were injected with 0.3–1 x 1010 IFN- Adv vp (high IFN-) (D) or with 0.03–0.1 x 1010 IFN- Adv vp (low IFN-) (E), and serum was analyzed for anti-dsDNA Abs at indicated times. Each line represents one mouse, and the results show 16 individual mice in each group. Note that most of the mice present a peak in anti-DNA preceding the occurrence of death. , Represent median values.
Remarkably, progressively lower doses of IFN- Adv progressively delayed occurrence of proteinuria and death (Fig. 5, B and C). Occurrence of proteinuria (Fig. 5B) apparently reflected the differences in IFN- levels observed between the high and low IFN- groups. Fifty percent of the mice (16 of 32) treated with high IFN- became proteinuric within 27 days after initiation of treatment, while 50% of the mice (12 of 24) treated with low IFN- became proteinuric within 34 days (p < 0.004). Survival was even a more sensitive index of in vivo IFN- levels as decreases in IFN- Adv doses significantly delayed occurrence of death (Fig. 5C). Overall, all mice treated with high IFN- were dead by day 90–103, while mice treated with low IFN- were dead by day 135–154 (p < 0.0001). Remarkably, the lowest dose of IFN- Adv (0.03 x 1010 vp), which elicited levels of circulating IFN- that were undetectable 10 days after initiation of treatment (Fig. 5A), yet induced proteinuria and death in all treated animals (Fig. 5, B and C).
Thus, severity of lupus depends upon the dose of IFN-, although minute amounts of IFN- precipitate a lethal lupus.
Occurrence and levels of circulating anti-dsDNA Abs are dependent upon the dose of IFN- Adv (Fig. 5, D and E). Anti-dsDNA IgG appeared as soon as day 10 after initiation of IFN- treatment in the high IFN- group; 9 of 16 mice were positive (e.g., above threshold; mean ± 3 SD in controls), while only 1 of 16 mice was positive in the low IFN- group (p < 0.05). Later on, between days 10 and 38, 15 of 16 mice in the high IFN- group and 16 of 16 in the low IFN- group became positive, although at lower levels than that observed in sick, unmanipulated NZB x NZW mice (see Fig. 3C). Notably, anti-dsDNA reached high titers between days 52 and 94 in animals that survived in the low IFN- group (Fig. 5E).
Sustained release of low doses of purified IFN- induces lupus disease
Although the above data demonstrate that occurrence of disease is related to the dose of IFN- in vivo, it is possible that the adenovector plays a role of adjuvant for IFN-. Thus, young NZB/W mice were infused with a low dose of purified IFN- using micro-osmotic pumps. Continuous delivery of IFN- (105 U/per pump) did not yield measurable levels of IFN- in blood (data not shown). Yet, this low dose of IFN- induced substantial levels of anti-dsDNA IgG Abs in seven to eight of nine mice, as illustrated by positive Crithidia immunofluorescence staining and ELISA detection (Fig. 6A and Table I). Remarkably, all mice, except one (eight of nine), became proteinuric within 39–78 days after initiation of treatment with purified IFN- (Table I), while all controls at that time were devoid of proteinuria. These proteinuric mice showed glomerulonephritis with IgG and C3 deposits in the mesangium and along the basement membrane (Fig. 6B). There was no difference in the occurrence of anti-dsDNA IgG or of proteinuria whether or not mice had also received the control adenovector. Thus, IFN- alone induces a dramatic acceleration of lupus in susceptible individuals.
FIGURE 6. Continuous infusion of IFN- in vivo precipitates lupus disease. Ten-week-old NZB/W mice were implanted s.c. with micro-osmotic pumps containing either PBS (n = 9) or mIFN--A (n = 9). Fourteen-day calibrated pumps were replaced three to four times to deliver constant amounts of IFN--A in vivo for a total period of 8–10 wk. A, Immunofluorescence in Crithidia (sera diluted 1/10). B, Kidney sections of mice infused with IFN--A for 56 days (8 wk) and harvested at day 60. Upper panel, PAS staining. Middle panel, Sections from control or IFN--A-treated mice were stained with a mixture of FITC-conjugated anti-IgG1, anti-IgG2a, anti-IgG2b, and anti-IgG3. Lower panel, Staining with anti-C3.
Table I. Continuous delivery of low doses of purified IFN- induces precipitation of lupus
Discussion
The present study demonstrates that the prolonged expression of IFN- in vivo leads to dramatic acceleration of lupus manifestations in young lupus-prone mice. Using Adv-mediated gene transfer of mouse IFN-5, elevated serum levels of IFN- were rapidly obtained, which persisted for at least 42 days.
This prolonged expression of IFN- in vivo results, in genetically susceptible preautoimmune lupus mice, in a brisk disease that has all characteristics of SLE. Anti-dsDNA Abs appeared in circulation as soon as day 10 after initiation of IFN- treatment and mice rapidly died of lupus glomerulonephritis, as illustrated by glomerular hypercellularity, heavy capillary and mesangial IgG and C3 deposits (32, 39, 40, 41), and the presence of anti-nuclear autoantibodies in kidney eluates (31, 42). Disease acceleration mediated by IFN- is striking as proteinuria and death occurred 12 and 21 wk earlier, respectively, than in untreated NZB/W mice (our colony) (32, 39).
Systemic administration of purified preparations of mIFN- (three times per week from birth/1 mo to 33–37 wk of age) has been shown to induce only a slight acceleration of disease in NZB/W and NZB lupus-prone strains (27, 28, 43). In particular, IFN- did not accelerate the development of proteinuria, and acceleration of death was modest. These studies using semipurified preparations of type I IFN (e.g., containing both IFN- and IFN-) did not directly attribute to IFN- the key role in the acceleration of the disease. Furthermore, the effect of a prolonged expression of IFN- in a normal genetic background was not evaluated. Our study addresses both points. It is very likely that the difference with our results holds in both the duration of, and persistence in, expression of IFN- in vivo. Indeed, it has been shown that 8 h after a single injection of IFN-, 90% of the IFN- is removed from the circulation (29). Thus, our model, by favoring IFN- impregnation in vivo, might represent a model of IFN- dysregulation seen in human lupus (10, 11, 12, 13, 23, 24).
IFN- in vivo enhances total IgG Abs both in BALB/c and NZB/W mice, a result in accordance with recent report (44). Despite the induction of anti-ssDNA IgG autoantibodies in BALB/c mice, IFN- failed, however, to induce the appearance of anti-dsDNA IgG and glomerulonephritis, suggesting that IFN- is not sufficient to induce a full SLE-like phenotype in a normal genetic background. We cannot, however, rule out that if followed for several months after initiation of IFN- treatment, BALB/c mice would not have shown signs of SLE-like disease. The human studies have also suggested that IFN- treatment of patients with cancer or infectious diseases can induce full-blown SLE in certain susceptible individuals (15, 16). Several predisposing loci have been identified for several lupus strains (reviewed in Refs. 3 , 5). However, genetic basis conferring to certain animals the susceptibility for IFN- to trigger autoimmunity has yet to be determined. Recently described IFN-inducible gene, ifi202, might be, among other susceptibility genes, a potential candidate that accounts for the observed IFN--mediated disease (45). The influence of the genetic background appears to be crucial as it has been very recently reported that type I IFNs suppress autoimmunity in MRLlpr/lpr mice (46).
IFN- in vivo induced the expression of BLyS to a similar level in both the autoimmune and the normal strains. BLyS is a member of the TNF ligand family that is a potent coactivator of B cells in vitro and in vivo (37). Mice transgenic for BLyS develop symptoms characteristic of SLE, such as the presence of high levels of rheumatoid factors, circulating immune complexes, anti-DNA autoantibodies, and Ig deposition in the kidneys, and suggest that dysregulation of BLyS expression may be a critical element in the chain of events leading to autoimmunity (34, 35, 36, 37, 38). Consistent with a possible role for BLyS in Ab-mediated autoimmune diseases, in patients with SLE, rheumatoid arthritis, and Sjogren’s syndrome, an association was found between elevated levels of BLyS in the blood and the severity of disease (38, 47, 48). Given our finding that mice treated with IFN- develop sustained levels of BLyS in circulation, it is tempting to suggest that high BLyS levels observed in human SLE are a consequence of an overproduction of IFN- in vivo. However, overexpression of both BLyS and IFN- does not lead to the development of a lupus-like phenotype in BALB/c mice. Several explanations can be put forward to explain such an absence of effect. Among those, the genetic background (e.g., most of BLyS transgenic mice are on BL/6 background (35, 37, 38), time for development of autoimmunity (e,g., several months in BLyS Tg mice), and duration and levels of BLyS expression in blood of Tg mice may explain why BALB/c behave as a BLyS-resistant mouse strain in our experimental conditions. It is also possible that the association of IFN- and BLyS precipitates autoimmunity in lupus-prone mice, while, on the contrary, IFN- would counteract BLyS-mediated autoimmunity in normal mice. These hypotheses are currently under investigation.
In regard with the severity of IFN--induced glomerulonephritis in NZB/W mice, it is intriguing that these mice injected with high IFN- Adv had low levels of serum anti-dsDNA Abs. A first explanation is that IFN- induces glomerulonephritis independently of anti-dsDNA autoantibodies, as it has been recently reported in different autoimmune models (49, 50) and in glomerular lesions induced by IFN at birth in normal mice (51). In this regard, it is possible that IFN- might have a direct deleterious effect on NZB/W kidneys, as proposed previously to explain IFN--mediated glomerular lesions in newborn normal mice (52). A second explanation is that in the brisk IFN--driven lupus process, nephritogenic Abs rapidly deposit into the kidneys or are lost in the urine during renal flare, and might therefore become barely detectable in circulation (53, 54, 55). The presence of anti-dsDNA Abs in substantial amounts in kidneys of IFN--treated NZB/W mice would rather favor the second hypothesis, although the two mechanisms may not be mutually exclusive.
If Adv gene transfer technology has major advantages (long-term protein expression, low cost), it, however, may induce the generation of cellular immune responses, inflammation (56), and, possibly, tissue damage (57). The Adv infection might indeed provide an adjuvant role in the acceleration of SLE observed upon IFN- delivery, and this has implications for the role of infection in triggering lupus disease. However, our data provide several lines of evidence to assign little role of the virus in the brisk lupus disease we see upon prolonged in vivo expression of IFN-. First, it is worth mentioning that serum IFN- elicited by the highest dose of IFN- Adv did not induce any premature death or hepatic toxicity (data not shown). Second, none of BALB/c or NZB/W mice injected with 1 x 1010 vp of empty Adv developed any biological nor clinical sign of lupus. Third, we could not detect the presence of anti-Adv Abs (as measured by ELISA) (58) in kidney eluates from 1 x 1010 IFN- Adv-treated NZB/W mice (data not shown). These Abs, complexed or not with adenoviral particles, could very well have participated in the IFN--driven immune complex glomerulonephritis (59). Finally, a strong line of evidence is provided by our observation that continuous infusion of rmIFN- also accelerated lupus autoimmunity. Altogether, this suggests that IFN- per se is the driving force leading to the early expression of lupus manifestations.
It is intriguing that, at variance with SLE patients, lupus-prone mice, of the several strains tested, do not have circulating IFN- (our unpublished results) (26). It is possible, however, that serum IFN- levels are below the threshold of detection and/or IFN- is preferentially expressed into (inflamed) target tissues, as it has been recently shown in cutaneous lupus erythematosus lesions (60). Nonetheless, it is important to stress that SLE patients show a generalized up-regulation of IFN-related genes regardless of the presence or absence of IFN- in their sera (23, 24).
Recent work has suggested a role for type I IFNs in murine lupus. Indeed, NZB congenic lupus-prone mice and C57BL/6 lpr/lpr mice lacking IFNAR-1, the -chain of the common receptor for type I IFNs, have reduced lupus-like disease (25, 26). Our present work not only adds evidence to the concept that type I IFNs are major effectors in the pathogenesis of lupus, but also directly demonstrates a major role of the genetic background conferring to IFN- its pathogenic role.
In conclusion, the present study provides the direct demonstration that prolonged in vivo expression of IFN- induces, in susceptible individuals, a brisk disease that has all characteristics of SLE. Moreover, we have established a new experimental model in which lupus disease is amenable in a few weeks and which will permit to dissect further the etiology and the cellular and molecular mechanisms that initiate and sustain the lupus autoimmune response.
Disclosures
The authors have no financial conflict of interest.
Acknowledgments
We thank Dr. R. Gerard (University of Texas Southwestern, Dallas, TX) for his precious advice on adenovectors. We also thank C. Joseph, S. Clayton, and F. Marches for their expert technical help. We are also grateful to Dr. C. Mohan for his help in kidney elution technique and for providing us with lupus mouse sera. We thank Drs. V. Pascual, K. Palucka, I. Gresser, A. Dalloul, and D. Emilie for stimulating discussions and critical reading of the manuscript.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 J.B. is the recipient of the Max and Gayle Clampitt chair of Immunology Research. This work was supported by postdoctoral fellowship from the Assistance Publique des Hopitaux de Paris (to A.M.).
2 J.B. and S.K. codirected this work and are co-corresponding authors.
3 Address correspondence and reprint requests to Dr. Jacques Banchereau, Baylor Institute for Immunology Research, 3434 Live Oak, Dallas, TX 75204. E-mail address: jacquesb{at}baylorhealth.edu
4 S.K. is on leave to Institut National de la Santé et de la Recherche Médicale Unité 131, Institut Paris-Sud des Cytokines, Clamart, France. E-mail address: sophie.koutouzov{at}inserm.ipsc.u-psud.fr
5 Abbreviations used in this paper: SLE, systemic lupus erythematosus; Adv, adenovirus; BLyS, B lymphocyte stimulator; mIFN, murine IFN; PAS, periodic acid-Schiff; RT, room temperature; vp, viral particle; AU, arbitrary units.
Received for publication September 2, 2004. Accepted for publication October 26, 2004.
References
Lipsky, P. E.. 2001. Systemic lupus erythematosus: an autoimmune disease of B cell hyperactivity. Nat. Immunol. 2:764.
Desai-Mehta, A., L. Lu, R. Ramsey-Goldman, S. K. Datta. 1996. Hyperexpression of CD40 ligand by B and T cells in human lupus and its role in pathogenic autoantibody production. J. Clin. Invest. 97:2063.
Marrack, P., J. Kappler, B. L. Kotzin. 2001. Autoimmune disease: why and where it occurs. Nat. Med. 7:899.
Shlomchik, M. J., J. E. Craft, M. J. Mamula. 2001. From T to B and back again: positive feedback in systemic autoimmune disease. Nat. Rev. Immunol. 1:147.
Wakeland, E. K., K. Liu, R. R. Graham, T. W. Behrens. 2001. Delineating the genetic basis of systemic lupus erythematosus. Immunity 15:397
Llorente, L., W. Zou, Y. Levy, Y. Richaud-Patin, J. Wijdenes, J. Alcocer-Varela, B. Morel-Fourrier, J. C. Brouet, D. Alarcon-Segovia, P. Galanaud, et al 1995. Role of interleukin 10 in the B lymphocyte hyperactivity and autoantibody production of human systemic lupus erythematosus. J. Exp. Med. 181:839.
Lawson, B. R., G. J. Prud’homme, Y. Chang, H. A. Gardner, J. Kuan, D. H. Kono, A. N. Theofilopoulos. 2000. Treatment of murine lupus with cDNA encoding IFN-R/Fc. J. Clin. Invest. 106:207
Stewart, T. A.. 2003. Neutralizing interferon as a therapeutic approach to autoimmune diseases. Cytokine Growth Factor Rev. 14:139.
Pascual, V., J. Banchereau, A. K. Palucka. 2003. The central role of dendritic cells and interferon- in SLE. Curr. Opin. Rheumatol. 15:548.
Ioannou, Y., D. A. Isenberg. 2000. Current evidence for the induction of autoimmune rheumatic manifestations by cytokine therapy. Arthritis Rheum. 43:1431.
Ronnblom, L. E., G. V. Alm, K. E. Oberg. 1990. Possible induction of systemic lupus erythematosus by interferon- treatment in a patient with a malignant carcinoid tumor. J. Intern. Med. 227:207
Vallin, H., A. Perers, G. V. Alm, L. Ronnblom. 1999. Anti-double-stranded DNA antibodies and immunostimulatory plasmid DNA in combination mimic the endogenous IFN- inducer in systemic lupus erythematosus. J. Immunol. 163:6306.
Bave, U., G. V. Alm, L. Ronnblom. 2000. The combination of apoptotic U937 cells and lupus IgG is a potent IFN- inducer. J. Immunol. 165:3519.
Blanco, P., A. K. Palucka, M. Gill, V. Pascual, J. Banchereau. 2001. Induction of dendritic cell differentiation by IFN- in systemic lupus erythematosus. Science 294:1540.
Steinman, R. M., D. Hawiger, M. C. Nussenzweig. 2003. Tolerogenic dendritic cells. Annu. Rev. Immunol. 21:685.
Banchereau, J., V. Pascual, A. K. Palucka. 2004. Autoimmunity through cytokine-induced dendritic cell activation. Immunity 20:539.
Baechler, E. C., F. M. Batliwalla, G. Karypis, P. M. Gaffney, W. A. Ortmann, K. J. Espe, K. B. Shark, W. J. Grande, K. M. Hughes, V. Kapur, et al 2003. Interferon-inducible gene expression signature in peripheral blood cells of patients with severe lupus. Proc. Natl. Acad. Sci. USA 100:2610
Bennett, L., A. K. Palucka, E. Arce, V. Cantrell, J. Borvak, J. Banchereau, V. Pascual. 2003. Interferon and granulopoiesis signatures in systemic lupus erythematosus blood. J. Exp. Med. 197:711.
Braun, D., P. Geraldes, J. Demengeot. 2003. Type I interferon controls the onset and severity of autoimmune manifestations in lpr mice. J. Autoimmun. 20:15.
Jovelin, F., G. Mostoslavsky, Z. Amoura, H. Chabre, D. Gilbert, D. Eilat, J. F. Bach, S. Koutouzov. 1998. Early anti-nucleosome autoantibodies from a single MRL+/+ mouse: fine specificity, V gene structure and pathogenicity. Eur. J. Immunol. 28:3411.
Xie, C., Z. Liang, S. Chang, C. Mohan. 2003. Use of a novel elution regimen reveals the dominance of polyreactive antinuclear autoantibodies in lupus kidneys. Arthritis Rheum. 48:2343
Dixon, F. J., M. B. Oldstone, G. Tonietti. 1971. Pathogenesis of immune complex glomerulonephritis of New Zealand mice. J. Exp. Med. 134:65s.
Foster, M. H., V. R. Kelley. 1999. Lupus nephritis: update on pathogenesis and disease mechanisms. Semin. Nephrol. 19:173.
Gross, J. A., J. Johnston, S. Mudri, R. Enselman, S. R. Dillon, K. Madden, W. Xu, J. Parrish-Novak, D. Foster, C. Lofton-Day, et al 2000. TACI and BCMA are receptors for a TNF homologue implicated in B-cell autoimmune disease. Nature 404:995
Mackay, F., S. A. Woodcock, P. Lawton, C. Ambrose, M. Baetscher, P. Schneider, J. Tschopp, J. L. Browning. 1999. Mice transgenic for BAFF develop lymphocytic disorders along with autoimmune manifestations. J. Exp. Med. 190:1697
Khare, S. D., I. Sarosi, X. Z. Xia, S. McCabe, K. Miner, I. Solovyev, N. Hawkins, M. Kelley, D. Chang, G. Van, et al 2000. Severe B cell hyperplasia and autoimmune disease in TALL-1 transgenic mice. Proc. Natl. Acad. Sci. USA 97:3370.
Mackay, F., P. Schneider, P. Rennert, J. Browning. 2003. BAFF and APRIL: a tutorial on B cell survival. Annu. Rev. Immunol. 21:231.
Groom, J., S. L. Kalled, A. H. Cutler, C. Olson, S. A. Woodcock, P. Schneider, J. Tschopp, T. G. Cachero, M. Batten, J. Wheway, et al 2002. Association of BAFF/BLyS overexpression and altered B cell differentiation with Sjogren’s syndrome. J. Clin. Invest. 109:59
Lambert, P. H., F. J. Dixon. 1968. Pathogenesis of the glomerulonephritis of NZB/W mice. J. Exp. Med. 127:507.
Granholm, N. A., T. Cavallo. 1990. Mechanism of localization of immune complexes in NZB/W mice with early nephritis. Autoimmunity 8:17.
Sasaki, T., A. Hatakeyama, S. Shibata, H. Osaki, M. Suzuki, K. Horie, Y. Kitagawa, K. Yoshinaga. 1991. Heterogeneity of immune complex-derived anti-DNA antibodies associated with lupus nephritis. Kidney Int. 39:746.
Ebling, F., B. H. Hahn. 1980. Restricted subpopulations of DNA antibodies in kidneys of mice with systemic lupus: comparison of antibodies in serum and renal eluates. Arthritis Rheum. 23:392.
Sergiescu, D., I. Cerutti, E. Efthymiou, A. Kahan, C. Chany. 1979. Adverse effects of interferon treatment on the life span of NZB mice. Biomedicine 31:48.
Le Bon, A., G. Schiavoni, G. D’Agostino, I. Gresser, F. Belardelli, D. F. Tough. 2001. Type I interferons potently enhance humoral immunity and can promote isotype switching by stimulating dendritic cells in vivo. Immunity 14:461.
Rozzo, S. J., J. D. Allard, D. Choubey, T. J. Vyse, S. Izui, G. Peltz, B. L. Kotzin. 2001. Evidence for an interferon-inducible gene, Ifi202, in the susceptibility to systemic lupus. Immunity 15:435.
Hron, J. D., S. L. Peng. 2004. Type I IFN protects against murine lupus. J. Immunol. 173:2134.
Zhang, J., V. Roschke, K. P. Baker, Z. Wang, G. S. Alarcon, B. J. Fessler, H. Bastian, R. P. Kimberly, T. Zhou. 2001. Cutting edge: a role for B lymphocyte stimulator in systemic lupus erythematosus. J. Immunol. 166:6.
Ho, A., L. S. Magder, S. G. Barr, M. Petri. 2001. Decreases in anti-double-stranded DNA levels are associated with concurrent flares in patients with systemic lupus erythematosus. Arthritis Rheum. 44:2342.
Macanovic, M., M. B. Hogarth, P. J. Lachmann. 1999. Anti-DNA antibodies in the urine of lupus nephritis patients. Nephrol. Dial. Transplant. 14:1418.
Isenberg, D. A., M. R. Ehrenstein, C. Longhurst, J. K. Kalsi. 1994. The origin, sequence, structure, and consequences of developing anti-DNA antibodies: a human perspective. Arthritis Rheum. 37:169.
Yang, Y., F. A. Nunes, K. Berencsi, E. E. Furth, E. Gonczol, J. M. Wilson. 1994. Cellular immunity to viral antigens limits E1-deleted adenoviruses for gene therapy. Proc. Natl. Acad. Sci. USA 91:4407.
Wright, N. G., H. J. Cornwell. 1983. Experimental canine adenovirus glomerulonephritis: histological, immunofluorescence and ultrastructural features of the early glomerular changes. Br. J. Exp. Pathol. 64:312.
Gahery-Segard, H., F. Farace, D. Godfrin, J. Gaston, R. Lengagne, T. Tursz, P. Boulanger, J. G. Guillet. 1998. Immune response to recombinant capsid proteins of adenovirus in humans: antifiber and anti-penton base antibodies have a synergistic effect on neutralizing activity. J. Virol. 72:2388
Wright, N. G., W. I. Morrison, H. Thompson, H. J. Cornwell. 1973. Experimental adenovirus immune complex glomerulonephritis. Br. J. Exp. Pathol. 54:628.
Farkas, L., K. Beiske, F. Lund-Johansen, P. Brandtzaeg, F. L. Jahnsen. 2001. Plasmacytoid dendritic cells (natural interferon-/-producing cells) accumulate in cutaneous lupus erythematosus lesions. Am. J. Pathol. 159:237.(Alexis Mathian Arthur Wei)