Autoimmune disease in MRL-Faslpr mice is characterized by fatal nephritis, systemic pathology, and autoantibodies, mimicking human lupus. We previously reported that 1) intrarenal IL-12 elicits nephritis by fostering the accumulation of intrarenal IFN-γ-secreting T cells, and 2) MRL-Faslpr mice deficient in the IFN-γ receptor were spared from nephritis. Therefore, we hypothesized that eliminating IL-12 in MRL-Faslpr mice reduces IFN-γ-secreting cells and thereby prevents systemic pathology. For this purpose, we constructed an IL-12p40-deficient MRL-Faslpr(IL-12−/−) strain. We determined that glomerular and interstitial, but not perivascular, renal pathology were decreased in IL-12−/− mice vs the wild-type (WT) strain (5 mo of age). Similarly, systemic pathology (lung, lacrimal and salivary glands, skin, and lymphadenopathy) was diminished. The intrarenal accumulation of T cells (CD4+, CD8+, CD4−CD8−B220+) and macrophages was dramatically reduced in IL-12−/− MRL-Faslpr kidneys. We determined that there were fewer IFN-γ transcripts (>70%) in the IL-12−/− protected kidneys compared with the WT kidneys. Similarly, cells propagated from IL-12−/− MRL-Faslpr kidneys generated substantially less IFN-γ when stimulated with IL-12 and IL-18 compared with those from WT kidneys, and we detected fewer CD8 and B220 T cells producing IFN-γ in these IL-12−/− MRL-Faslpr kidneys. Of note, survival was modestly extended in the IL-12−/− MRL-Faslpr mice. While lung and lacrimal and salivary gland pathology remained reduced in moribund IL-12−/− MRL-Faslpr mice, renal pathology and IFN-γ expression were equivalent to those in the WT strain. Thus, we suggest that IL-12 is a therapeutic target for multiple tissues in lupus; however blocking IL-12 alone is not sufficient to confer enduring protection from lupus nephritis.
The MRL-Faslpr mouse has a systemic autoimmune disease mimicking human lupus, characterized by inflammation in multiple tissues, skin, glands, lungs, and joints; massive lymphadenopathy; and splenomegaly (1, 2, 3). Kidney disease in the MRL-Faslpr strain is fatal and is dependent on the infiltration of T cells regulated by cytokines. Multiple T cell populations accumulate in the MRL-Faslpr kidney with nephritis, including CD4, CD8, and B220-positive T cells that lack CD4 and CD8 determinants, termed double-negative (DN)3 T cells (4, 5, 6). Furthermore, we have established that in the absence of CD4 and/or CD8 and DN T cell populations, MRL-Faslpr mice are totally protected from kidney disease (7). Thus, the regulation of T cells in the kidney is critical in the pathogenesis of autoimmune disease.
Cytokines, including IL-12 and IFN-γ, promote autoimmune disease in the MRL-Faslpr strain (5). MRL-Faslpr mice lacking IFN-γR or IFN-γ are spared from kidney disease and have a prolonged life span (8, 9). Since T cells within the kidney are a rich source of IFN-γ, cytokines that induce the production of IFN-γ are potential therapeutic targets for lupus (5, 10, 11, 12).
IL-12 released within the kidney in MRL-Faslpr mice may provide the signal for T cells to destroy the kidney. This concept is based on the following: 1) IL-12 generated by APC regulates T cells (13, 14, 15); 2) IL-12 promotes T cell proliferation (16, 17); 3) IL-12 stimulates T cells to generate IFN-γ (18); and 4) IL-12 is a prerequisite for committing CD4 T cells into a T cell subset (Th1) that is associated with disease in MRL-Faslpr mice (19, 20, 21). Furthermore, we have determined that local and systemic provision of IL-12 via gene transfer promotes IFN-γ-dependent nephritis in MRL-Faslpr mice (5), and IL-12 recruits CD4, CD8, and DN T cells into the kidney, activates kidney-infiltrating T cells to produce IFN-γ, and promotes the expansion of CD4 T cells within MRL-Faslpr kidneys (5). In addition, glomerulonephritis in young MRL-Faslpr mice is accelerated by injection of IL-12 (22). Taken together, these data suggest that IL-12 is required for autoimmune kidney disease.
IL-12 is a heterodimeric structure consisting of two disulfide-bonded subunits, p40 and p35 (23, 24, 25). Both subunits are required for the biological activity of IL-12 (26, 27). The introduction of a null mutation into the IL-12p40 gene of normal (C57BL/6) mice results in these mice being unable to produce bioactive IL-12, yet does not compromise their development (28). The primary defect in the IL-12p40−/− mice is the reduced generation of IFN-γ (28). IFN-γ production in IL-12p40−/− mice is impaired, but not completely lacking, in response to endotoxin (28). Similarly, IL-12p40−/− mice generate a reduction, but not an elimination, of IFN-γ+ CD4+ T cells following exposure to nonlethal infections with intracellular pathogens or repeated immunization with a pathogen extract (29). Thus, IL-12 diminishes the generation of IFN-γ-producing cells in response to immunological stimuli.
Taken together, we hypothesize that blocking IL-12 reduces lupus nephritis, intrarenal IFN-γ-generating cells, and systemic pathology in MRL-Faslpr mice. In this study we determined that there is a delay in renal disease (pathology, proteinuria) and systemic pathology (skin, lungs, lacrimal and salivary glands, lymphadenopathy) in IL-12p40 genetic deficient MRL-Faslpr mice (IL-12−/− MRL-Faslpr) compared with the wild-type (WT) strain. We detected fewer macrophages and T cells (CD4, CD8, and DN) in the IL-12−/− MRL-Faslpr kidneys compared with those from the WT strain (5 mo of age). In the absence of intrarenal IL-12, we detected a reduction in intrarenal IFN-γ-generating B220 and CD8 T cells. Eliminating IL-12 extended survival and protected the lungs, skin, and lacrimal and salivary (submandibular) glands from progressive injury. However, deleting IL-12 did not confer enduring protection from renal disease. This suggests that sufficient IFN-γ-generating cells remained, resulting in fatal disease. Taken together, eliminating IL-12 delays tissue injury in MRL-Faslpr mice. We suggest that an optimal therapeutic strategy should block IL-12 along with perhaps other IFN-γ-generating molecules.
Materials and Methods
MRL/MpJ++ (MRL++), MRL/MpJ-Faslpr/Faslpr(MRL-Faslpr), C3H/Fej, and C57BL/6, IL-12p40−/− mice on a C57BL/6 background were purchased from The Jackson Laboratory (Bar Harbor, ME), and housed and bred in our pathogen-free animal facility.
Generating IL-12-deficient MRL-Faslprmice
We constructed an IL-12p40-deficient MRL-Faslpr strain (IL-12−/− MRL-Faslpr) using a backcross-intercross scheme. MRL-Faslpr mice were bred with IL-12−/− (C57BL/6) mice to yield heterozygous F1 offspring. We intercrossed F1 mice and screened the progeny by PCR amplification of tail genomic DNA for the Faslpr mutation and for IL-12 using specific primers. After five generations of backcross-intercross matings, this breeding scheme generated a colony of MRL-Faslpr mice homozygous and heterozygous for the disrupted IL-12p40 gene. We analyzed the B5 generation as we have previously established that there are sufficient MRL-Faslpr background genes even by the B3 generation to result in phenotypic changes characteristic of the MRL-Faslpr strain. Identification of the disrupted and nondisrupted IL-12p40 in experimental mice was determined by PCR using tail genomic DNA. The DNA was assessed by PCR using oligonucleotide primers that recognize either the normal IL-12p40 gene (sense, 5′-AGT GAA CCT CAC CTG TGA CAC G-3′; antisense, 5′-TTT TTA CTG TGG ATG GAT-3′) or the neo gene (sense, 5′-ATT GAA CAA GAT GGA TTG CAC-3′; antisense, 5′-TCT TCG TCC AGA TCA TCC T-3′). Gel analysis of the PCR products identified the IL-12p40 and neo gene fragments at 500 bp.
We validated the expression of IL-12p40 when mice were sacrificed by measuring IL-12p40 in the serum using the Duo Set ELISA for mouse IL-12p40 (R&D System, Minneapolis, MN) and by measuring IL-12p40 transcripts in the kidney using RT-PCR (28).
Urinary protein levels in IL-12−/− MRL-Faslpr and IL-12+/+ MRL-Faslpr mice were assessed semiquantitatively by dipstick analysis (Albustick; Bayer, Elkhart, IN) on a monthly basis beginning at 2 mo of age. On the day of analysis, dipstick proteinuria measurements were made in individual mice in the morning and repeated in the evening. If the measurements were inconsistent, the mouse was reassessed on the following day. The urinary protein (milligrams per deciliter)/creatinine (milligrams per deciliter) ratio was assessed in urine specimens as a more accurate measurement for proteinuria. Urine was collected from individual mice housed in metabolic cages for 24 h. The protein/creatinine ratio was measured at Children’s Hospital (Boston, MA).
Survival was assessed in the IL-12−/− and IL-12+/+ MRL-Faslpr strains and included moribund mice.
Lymphadenopathy (cervical, brachial, and inguinal) was assessed monthly beginning at 3 mo of age. We palpated the lymph node and scored them using a grade of 0–3 (0 = none; 1 = small, at one site; 2 = moderate, at two sites; 3 = large, at three or more different sites).
Skin lesions were assessed monthly beginning at 3 mo of age. We scored the skin lesions by gross pathology using a grade of 0–3 (0 = none; 1 = mild (snout and ears); 2 = moderate, <2 cm (snout, ears, and intrascapular); and 3 = severe, >2 cm (snout, ears, and intrascapular)).
Spleen enlargement was assessed at the time of sacrifice or death. We compared spleen weights in IL-12+/+ and IL-12−/− MRL-Faslpr mice.
The kidneys were fixed in 10% formalin for 3 h at 4°C. Paraffin sections (4 μm) were stained with H&E and the periodic acid-Schiff reagent. We evaluated glomerular pathology by assessing 20 glomerular cross-sections (gcs) per kidney and scored each glomerulus on a semiquantitative scale: 0 = normal (35–40 cells/gcs); 1 = mild (glomeruli with few lesions, with slight proliferative changes, and mild hypercellularity (41–50 cells/gcs), and/or minor exudation); 2 = moderate (glomeruli with moderate hypercellularity (50–60 cells/gcs), including segmental and/or diffuse proliferative changes, hyalinosis, and/or moderate exudates); and 3 = severe (glomeruli with segmental, or global sclerosis, and/or exhibiting severe hypercellularity (>60 cells/gcs), necrosis, crescent formation, and/or heavy exudation). Damaged tubules (percentage; consisting of dilation and/or atrophy and/or necrosis) were determined in 200 randomly selected renal cortical tubules per kidney (magnification, ×400). Perivascular cell accumulation was determined semiquantitatively by scoring the number of cell layers surrounding the majority of vessel walls on a 0–3 scale (0 = none; 1 = <5 cell layers; 2 = 5–10 cell layers; 3 = >10 cell layers). We evaluated renal pathology using coded slides.
The lungs, lacrimal glands, and salivary glands were fixed in 10% formalin for 3 h at 4°C. Paraffin sections (4 μm) were stained with H&E and evaluated by light microscopy. The perivascular leukocyte infiltrations in the lungs were analyzed by morphomeric analysis. We measured the leukocyte infiltrates surrounding 10 random vessels (0 = none; 1 = <3 layers in <50%; 2 = 3–6 cell layers >50%; 3 = >6 layers). Peribronchiolar leukocyte infiltration was determined by semiquantitatively scoring the number of cells surrounding 10 random bronchi (0 = none; 1 = <3 layers >50% bronchi; 2 = 3–6 layers >50% bronchi; 3 = >6 layers >50% bronchi). We scored lacrimal gland pathology on a scale of 0–3 (grade 0 = no inflammatory cells; grade 1 = the presence of at least one focus; grade 2 = multiple foci; grade 3 = multiple foci plus evidence of gland destruction). We scored salivary gland inflammation on a scale of 0–3 (grade 0 = no inflammatory cells; grade 1 = few perivascular and periductal inflammatory infiltrates (<100 cells); grade 2 = moderate number of perivascular and periductal inflammatory infiltrates (=100–500 cells); grade 3 = extensive inflammation with large inflammatory foci (>500 cells)).
Cryostat-sectioned kidneys were stained for macrophages with CD68 (Serotec, Oxford, U.K.), and T cells were stained with anti-CD4, anti-CD8, and anti-B220 rat-anti-mouse mAb (BD PharMingen, San Diego, CA) according to a previously described immunoperoxidase method (30). The immunostaining was analyzed by counting for the presence of CD68, CD4, CD8, and B220-positive cells in 10 randomly selected high power fields. We determined that B220-positive cells in the kidney are unique DN T cells and are not B cells (5, 8, 31).
IgG and C3 deposits within renal glomeruli
Kidney cryostat cross-sections (4 μm thick) were stained with FITC-conjugated goat anti-mouse IgG and FITC-conjugated goat IgG fraction of mouse C3 (Cappel, Malvern, PA) for 30 min at 37°C. The fluorescence intensity within the peripheral glomerular capillary walls and the mesangium were scored on a scale of 0–3 (0 = none; 1 = weak; 2 = moderate; 3 = strong). At least 10 glomeruli/section were analyzed.
Identifying cytokine transcripts in the kidney and lymph nodes (PCR)
Taq activation for 15 min at 95°C, then 40 cycles with heating to 95°C for 15 s and cooling to 60°C for 1 min. The mRNA levels were normalized to those of GAPDH. The PCR primers used were as follows: GAPDH: sense, 5′-CAT GGC CTC CAA GGA GTA AG-3′; antisense, 5′-CCT AGG CCC CTC CTG TTA TT-3′; IL-18: sense, 5′-AGT GAA CCC CGA ACC AGA CT-3′; antisense, 5′-TTC AGG TGG ATC CAT TTC CT-3; IFN-γ: sense, 5′-AAC GCT ACA CAC TGC ATC TTG G-3′; antisense, 5′-GCC GTG GCA GTA ACA GCC-3′; CSF-1: sense, 5′-ACC TGG CAA GGG ACT CAC TA-3′; antisense, 5′-CAG GCT CTC TTC TTG GGA AA-3′; GM-CSF: sense, 5′-GGC AAT TTC ACC AAA CTC AA-3′; antisense, 5′-TGA AAT CG CAT AGG TGG TA-3; IL-4: sense, 5′-TCA TCG GCA TTT TGA ACG AG-3′; antisense, 5′-CGT TTG GCA CAT CCA TCT CC-3; and TNF-α: sense, 5′-TCA GCC GAT TTG CTA TCT CA-3′; antisense, 5′-TGG AAG ACT CCT CCC AGG TA-3.
Intrarenal IFN-γ protein (ELISA)
At the time of sacrifice, kidneys were removed and gently dissociated into single-cell suspensions. Red blood cells were removed by treatment with ACK lysing buffer (BioSource International, Camarillo, CA). Cell suspensions were resuspended in complete medium (RPMI 1640 medium supplemented with 10% FBS, 50 μM 2-ME, 50 μg/ml penicillin/streptomycin, and 2 mM l-glutamine; Cellgro, Herndon, VA).
To stimulate IFN-γ production, cell suspensions were cultured with recombinant murine IL-12 (10 μg/ml; PeproTech, Rocky Hill, NJ) and IL-18 (100 μg/ml; Endogen, Woburn, MA) for 48 h. IFN-γ in the supernatant was evaluated by ELISA according to published methods (18). Briefly, Maxisorp (Nunc, Naperville, IL) were coated with purified rat anti-mouse (clone R46A2; 1–5 μg/ml, 4°C overnight). After washing the plates, standards and samples (diluted 1/5) were added and incubated (4°C overnight). Wells were washed, and IFN-γ levels were determined by biotinylated rat anti-mouse IFN-γ Ab (clone XMG1.2) and a peroxidase visualization system. The Abs and reagents in these assays were purchased from BD PharMingen with the exception of avidin-peroxidase (Sigma-Aldrich, St. Louis, MO).
Identifying IFN-γ-expressing cells in the kidney (flow cytometry)
We removed the kidneys and gently homogenized them, lysed RBC with ACK lysing buffer (BioSource International), and washed the remaining cells in PBS. These cells were cultured in RPMI containing PMA (1 ng/ml; Sigma-Aldrich) and ionomycin (500 ng/ml; Sigma-Aldrich) for 4 h. Monensin (3 μM; Sigma-Aldrich) was added during the last 3 h. We washed the cells in PBS, suspended them in FACS buffer (PBS, 5% FBS, and 0.09% NaN3
Serum Ig profile
We measured total IgG and IgM, and IgG1, IgG2a, IgG2b and IgG3 isotypes by ELISA. Plates were coated overnight at 4°C with goat anti-mouse Ig capture Abs (Southern Biotechnology Associates, Birmingham, AL) in PBS. The wells were blocked for 2 h with 3% BSA/PBS. We added Ig standards to the plate using a series of 3-fold dilutions, and assessed serum samples using serial dilutions starting at 1/100. Standards and serum samples were incubated overnight at 4°C, and bound Ig was detected with goat anti-mouse detection Abs conjugated with HRP (Southern Biotechnology Associates) and 3,3′,5,5′-tetramethylbenzidine peroxidase substrate (Kirkegaard & Perry Laboratories, Gaithersburg, MD). The absorbance was measured at 450 nm.
In addition, we measured the anti-dsDNA Ab isotypes (IgG1, IgG2a, IgG2b, IgG3) in the IL-12−/− and IL-12+/+ MRL-Faslpr mice at five dilutions (1/50 through 1/12,150) as previously reported (32).
The data represent the mean ± SEM. We determined statistical significance using ANOVA and/or Student’s t test. We compared the survival curves using the log-rank two-tailed test.
Glomerular and tubular, but not perivascular, pathology is decreased in IL-12−/− MRL-Faslprkidneys
To evaluate renal pathology we sacrificed groups (n = 5) of IL-12−/− MRL-Faslpr and IL-12+/+ MRL-Faslpr mice at 5 mo of age. The glomerular pathology in IL-12−/− MRL-Faslpr mice was reduced compared with that in age- and sex-matched WT mice (p ≤ 0.001; Fig. 1⇓, a and d, arrowheads). In particular, there was a reduction in glomerular hypercellularity, glomerulosclerosis, and crescent formation. However, glomerular pathology was not entirely prevented; there was an increase in glomerular pathology in the IL-12−/− MRL-Faslpr mice compared with the age- and sex-matched mice with normal kidneys (MRL++, C3H/Fej; p ≤ 0.001). Similarly, there was less tubular pathology, including atrophy, dilation, and casts (asterisk), in IL-12−/− MRL-Faslpr mice compared with the WT strain (p ≤ 0.001; Fig. 1⇓, b and d). In fact, the IL-12−/− MRL-Faslpr and normal (C3H/Fej, MRL++) tubular pathologies were equivalent. In contrast, the extent of perivascular pathology, consisting of a cuffing of leukocytes around vessels, was not diminished in the IL-12−/− MRL-Faslpr strain compared with the WT strain (Fig. 1⇓, c and e).
Reduced leukocytes in the glomeruli and interstitium in IL-12−/− MRL-Faslprkidneys
We compared leukocytes including macrophages (CD68), CD4 and CD8 T cells, and B220 cells in IL-12−/− MRL-Faslpr kidneys at 5 mo of age with age- and sex-matched WT kidneys. There were far fewer macrophages (>50%), CD4 and CD8 T cells, and B220 cells in the IL-12−/−MRL-Faslpr glomeruli and interstitium compared with the WT kidneys (Fig. 2⇓, a and b). The numbers of CD4, CD8, and B220 cells were reduced to normal levels within glomeruli in IL-12−/− MRL-Faslpr kidneys compared with the WT strain. In contrast, the number of macrophages in the glomeruli and interstitium in IL-12−/− MRL-Faslpr mice were not reduced to normal levels (MRL++, C3H/Fej; Fig. 2⇓, a and c). Similarly, the numbers of CD4, CD8, and B220 cells remained above normal in the IL-12−/− MRL-Faslpr interstitium (Fig. 2⇓, b and c). Thus, deleting IL-12 reduces, but does not entirely prevent, intrarenal leukocytic accumulation.
Diminished Ig and C3 in glomeruli of IL-12−/− MRL-Faslprkidneys
The deposition of IgG and C3 in glomeruli is a characteristic feature of renal disease in MRL-Faslpr mice (33, 34, 35). IL-12+/+ MRL-Faslpr have IgG and C3 deposits within the peripheral glomerular capillary loops and the mesangium at 5 mo of age. We detected a decrease in IgG and C3 in glomeruli (periphery, mesangium) in IL-12−/− MRL-Faslpr mice compared with the WT strain (Fig. 3⇓, a and b, respectively; p ≤ 0.001).
Proteinuria is reduced in IL-12−/− MRL-Faslprmice
Urinary protein levels in IL-12−/− and IL-12+/+ MRL-Faslpr mice, which were undetectable at 2 mo of age, progressively increased from 3–5 mo of age (Fig. 4⇓). By comparison, although urinary protein levels in the IL-12−/− MRL-Faslpr strain rose above normal with advancing age, they remained lower than those in the WT strain. Proteinuria was more accurately evaluated by determining the protein/creatinine ratios from 24-h urine collections. The protein/creatinine ratio was reduced in the IL-12−/− MRL-Faslpr strain compared with the WT strain at 5 mo of age (Fig. 4⇓). However, the urinary protein level in IL-12−/− MRL-Faslpr mice was elevated compared with that in MRL-Faslpr mice with normal kidneys (2 mo of age). Taken together, deleting IL-12 in MRL-Faslpr mice retards, but does not prevent, proteinuria.
Lymphadenopathy, skin lesions, and lung and lacrimal and salivary gland pathology are suppressed in IL-12−/− MRL-Faslprmice
Lymphadenopathy, skin lesions, and splenomegaly are characteristic features of MRL-Faslpr mice (36). Lymphadenopathy progressively increased between 3–5 mo of age in IL-12−/− and IL-12+/+ MRL-Faslpr mice; however, in IL-12−/− mice, the severity was reduced compared with that in WT animals (Fig. 5⇓a). Similarly, skin lesions in the IL-12−/− MRL-Faslpr strain were reduced compared with those in the WT strain (Fig. 5⇓b). Conversely, splenomegaly (as measured by weight) did not differ in IL-12−/− and WT mice at 5 mo of age (0.4 ± 0.0 g (n = 8) and 0.5 ± 0.1 g (n = 6), respectively; p ≥ 0.27).
Several tissues (lungs, lacrimal and salivary glands), in addition to the kidney, were targeted for autoimmune destruction in the MRL-Faslpr strain. MRL-Faslpr mice developed peribronchiolar and perivascular pathology consisting of leukocytic infiltrates that were predominantly macrophages (6). The lung pathology in the IL-12−/− MRL-Faslpr strain was substantially reduced (>50%) compared with that in the IL-12+/+ MRL-Faslpr strain as evaluated at 5 mo of age (Fig. 5⇑c). Similarly, the MRL-Faslpr mice spontaneously developed inflammation in the lacrimal and salivary (submandibular) glands, resembling Sjögren’s syndrome, a condition that results in dry eyes and mouth (20, 37). Lacrimal and salivary pathology in MRL-Faslpr mice consisted of focal increases in leukocytes. There was a decline in the numbers of leukocytes in lacrimal and salivary glands in the IL-12−/− MRL-Faslpr mice compared with the WT strain (Fig. 5⇑d). These leukocytes were predominantly CD4, with lesser amounts of CD8 and even fewer macrophages and B220 cells (n = 5/group; data not shown). Therefore, MRL-Faslpr mice lacking IL-12 are initially protected from lung, lacrimal gland, and salivary gland pathology.
Survival is extended in IL-12−/− MRL-Faslprmice; renal disease is substantial in moribund IL-12−/− MRL-Faslprmice, while lung and lacrimal and salivary gland pathology and skin lesions remain protected
Survival in IL-12−/−MRL-Faslpr mice was modestly extended compared with that in the WT strain (Fig. 6⇓a, ∗, p ≤ 0.05). To determine whether tissue pathology remained suppressed in IL-12−/− MRL-Faslpr compared with IL-12+/+ MRL-Faslpr moribund mice (n = 10 and n = 5, respectively), we evaluated kidney, lacrimal and salivary gland, lung, and skin pathology (Fig. 6⇓b). Renal pathology in IL-12−/− MRL-Faslpr mice did not remain substantially diminished and became almost as severe as that in the WT strain (p > 0.16). However, the reduction in lung pathology, skin lesions and gland pathology was enduring (Fig. 6⇓b). The decreased skin lesions in the IL-12−/− MRL-Faslpr mice was impressive. Both the incidence and the severity of gross skin lesions in the IL-12−/− MRL-Faslpr mice were minimal compared with those in the WT strain (Fig. 6⇓b). While only a small portion (33%) of the IL-12−/− MRL-Faslpr strain had skin lesions, all of the IL-12+/+MRL-Faslpr strain were afflicted. In addition, the skin lesions evident in IL-12−/− MRL-Faslpr mice were tiny (grade 0.3 ± 0.2) compared with the prominent lesions in the WT strain (grade 2.2 ± 0.5; p ≤ 0.0004). This suggests that eliminating IL-12 in MRL-Faslpr mice retards, but is not sufficient to halt, renal disease. On the other hand, eliminating IL-12 does provide enduring protection from lung and lacrimal and salivary gland pathology and skin lesions in the MRL-Faslpr strain.
IFN-γ is decreased in IL-12−/− MRL-Faslprkidneys, while IL-18 remains unchanged
IFN-γ is required for kidney disease in MRL-Faslpr mice (8, 9, 38). To determine whether IFN-γ is reduced in IL-12−/− MRL-Faslpr kidneys, we probed for the expression of IFN-γ transcripts in the renal cortex using real-time PCR. There was a dramatic decrease in IFN-γ transcript expression in IL-12−/− MRL-Faslpr kidneys compared with the WT mice at 5 mo of age (Fig. 7⇓a). However, IFN-γ expression in IL-12−/− MRL-Faslpr kidneys was elevated above normal values (C3H/Fej kidneys).
To determine whether IFN-γ protein is diminished in the kidney, we evaluated intrarenal IFN-γ at 5 mo of age. IL-18 and IL-12 act synergistically to produce IFN-γ (39). Freshly isolated IL-12−/− MRL-Faslpr kidney cells stimulated with IL-12 and IL-18 (48 h) produced substantially less IFN-γ compared with WT cells, as measured by ELISA (Fig. 7⇑b). Thus, there were fewer IFN-γ-generating cells in IL-12−/− MRL-Faslpr kidneys compared with WT kidneys.
To determine whether IFN-γ increases with advancing renal pathology, we analyzed IL-12−/− and IL-12+/+ MRL-Faslpr moribund mice. We probed the kidneys from these mice for IFN-γ transcripts. Intrarenal IFN-γ transcripts in moribund IL-12−/− MRL-Faslpr were equivalent to those in moribund WT mice (0.82 ± 0.25 and 0.58 ± 0.12, respectively; n = 6/group; p > 0.42). Thus, there was an increase in IFN-γ transcripts with advancing renal disease and death.
IL-12−/− MRL-Faslprkidneys have a reduction in B220 and CD8 IFN-γ-producing cells compared with the WT strain
To determine the cell source of IFN-γ in the kidney, we analyzed IFN-γ in T cells (CD4, CD8, B220), macrophages (CD68), and TEC in freshly isolated cells in the IL-12−/− and IL-12+/+ MRL-Faslpr kidneys. We determined that in IL-12+/+ MRL-Faslpr kidneys, TEC cells produced the most IFN-γ (p ≤ 0.02), while B220-bearing cells and macrophages produced more IFN-γ (p ≤ 0.008) than CD8 T cells, By comparison, CD4 T cells produced the least amount of IFN-γ (p ≤ 0.008; Fig. 7⇑c). We detected a decrease in IFN-γ in the cells bearing the B220 and CD8 markers in IL-12−/− vs IL-12+/+ MRL-Faslpr kidneys (Fig. 7⇑c). We have previously determined that B220-expressing cells in the kidney are the unique T cell population bearing CD4 and CD8 determinants and lacking the B cell determinants, CD21/35 (8). Thus, the reduction in IFN-γ produced by B220 and CD8 T cells appeared to be responsible for the decrease in intrarenal IFN-γ in IL-12−/− MRL-Faslpr mice.
Nephritogenic cytokines (CSF-1, GM-CSF, and IL-4) are reduced in IL-12−/− MRL-Faslprkidneys compared with WT kidneys
CSF-1 (40), GM-CSF (40), and IL-4 (41) are cytokines that promote kidney disease in MRL-Faslpr mice. We determined that these nephritogenic cytokines transcripts were reduced in IL-12−/− MRL-Faslpr kidneys (5 mo of age) compared with WT kidneys (p ≤ 0.01; Fig. 8⇓a). Although there was a decline in TNF-αtranscripts, another nephritogenic cytokine (42), in the IL-12−/− MRL-Faslpr kidneys, this decrease was not significant (p ≥ 0.14; Fig. 8⇓a).
IFN-γ, TNF-α, GM-CSF, and IL-4 transcripts remain similar in the IL-12−/− and IL-12+/+ MRL-Faslprlymph nodes
We examined the Th1 cytokines (IFN-γ, TNF-α and GM-CSF) and the Th2 cytokine (IL-4) in the lymph nodes of IL-12−/− MRL-Faslpr and WT mice. We did not detect a difference in Th1 (IFN-γ (p ≥ 0.06), TNF-α (p ≥ 0.19), GM-CSF (p ≥ 0.43), and the Th2 cytokine, IL-4 (p ≥ 0.26), in IL-12−/− and IL-12+/+ MRL-Faslpr lymph nodes at 5 mo of age (Fig. 8⇑b).
Circulating isotypes were not diminished in IL-12−/− MRL-Faslprmice
We determined whether the decrease in Ig deposits in the kidney was related to a decline in circulating Ig isotypes. Although there was a modest reduction in total IgG in IL-12−/− MRL-Faslpr serum, it was not significant (Fig. 9⇓a). We did not detect a difference in IgG1, IgG2a, IgG2b, IgG3, and IgM levels in the circulation of IL-12−/− MRL-Faslpr compared with the WT strain (Fig. 9⇓a). To determine whether there was an alteration in autoantigen-specific serum Ig isotypes, we measured the levels of dsDNA-specific Ig subclasses. We did not detect a difference in the dsDNA-specific IgG isotypes (IgG1, IgG2a, IgG2b, IgG3) in the IL-12−/− and IL-12+/+ MRL-Faslpr serum compared with the WT strain evaluated at five dilutions (1/50 through 1/12,150; Fig. 9⇓b, 1/150 dilution). Taken together, these findings indicate that the decrease in IgG in IL-12−/− MRL-Faslpr glomeruli is not related to a reduction of these Igs or dsDNA-specific Igs in the circulation.
MRL-Faslpr mice are an appealing model to identify therapeutic targets, since disease is spontaneous, predictable, and rapid and mimics the multifaceted tissue destruction characteristic of human lupus. We now report that MRL-Faslpr with a genetic deficiency in IL-12 have reduced kidney, lacrimal gland, and salivary gland pathology; diminished lymphadenopathy; and prolonged survival. However, while the protection from lung, skin, and lacrimal and salivary gland pathology was enduring, renal pathology continued to progress, albeit at a slower rate, in the IL-12−/− MRL-Faslpr strain. The decreased severity of renal disease was associated with fewer IFN-γ-expressing cells intrarenally in the IL-12−/− kidneys compared with the WT strain. This may result in part from fewer IFN-γ-generating CD8 and B220 T cells infiltrating the kidney. Furthermore, the diminished renal disease in IL-12−/− MRL-Faslpr mice correlated with a reduction in IgG and C3 deposits in renal glomeruli. Of note, systemic (lymph node) IFN-γ and other Th1 cytokines (GM-CSF, TNF-α) and the Th2 cytokine, IL-4, remained similar in IL-12−/− MRL-Faslpr compared with the WT strain. We suggest that IL-12 alone is a therapeutic target for lung, skin, lacrimal gland, and salivary gland pathology, but renal disease requires the blockade of IL-12 along with other with nephritogenic molecules, such as IFN-γ, to confer long term protection.
It is intriguing that deleting IL-12 has a differential impact on pathology in multiple tissues undergoing autoimmune destruction in MRL-Faslpr mice. While renal injury is transiently delayed, protection from skin, lung, lacrimal gland, and salivary gland disease is more enduring in this strain. The discordant impact on tissue pathology of a single gene deletion in the autoimmune MRL-Faslpr strain is in keeping with prior reports. For example, deleting β2-microglobulin dampens nephritis, but accelerates skin lesions, in MRL-Faslpr mice (43). Furthermore, we have noted an improvement in kidney, lung, lymph node, and skin pathology despite equivalent splenomegaly in the current study as well as in monocyte chemoattractant protein-1-deficient MRL-Faslpr mice (6). The simplest explanation for these tissue differences is the distinctive microenvironment that reflects differing functions. While skin and lungs are barrier sites with specialized immune systems designed to ward off a breach (44, 45, 46), the kidney has other complex functions; it is responsible for homeostasis. In addition, a huge amount of cardiac output (20%) is filtered through the kidney. Thus, it has greater exposure to circulating leukocytes than other tissues and is at increased risk for leukocyte-mediated tissue injury. Future studies are required to detail the elements within the microenvironment that contribute to the relative importance of IL-12 in autoimmune-mediated injury.
There is ample evidence to suggest that IL-12p40 is a pivotal inflammatory mediator in barrier tissues (respiratory tract, skin, and gut). IL-12p40 is produced by macrophages during fibrosis, and IL-12p40-deficient mice are resistant to silica-induced pulmonary inflammation and fibrosis (47). Similarly, airway epithelial cell production of IL-12p40 is a key intermediate for virus-inducible inflammation (48). In the skin, IL-12p40 is enhanced in keratinocytes following UV light exposure and herpes lesions (49, 50, 51). In the gut, blocking IL-12 abrogates mouse experimental colitis (52). Taken together, although we have not determined the precise reason for the enduring protection from skin, lung, and lacrimal and salivary gland pathology in IL-12−/− MRL-Faslpr mice, our data suggest that IL-12 is pivotal in the pathogenesis of these tissues.
Why is kidney disease in MRL-Faslpr mice initially delayed, but eventually progresses in the absence of IL-12? This finding suggests that IL-12 is instrumental in the initiation phase of the disease, but that other nephritogenic molecules contribute to renal disease. Our laboratory and others established that in the absence of IFN-γ or the IFN-γ receptor in MRL-Faslpr mice, kidney disease is prevented (8, 9, 53). Furthermore, we note that renal injury correlates with a reduction in intrarenal IFN-γ. Therefore, it seems plausible that other IFN-γ-generating cytokines along with IL-12 drive lupus nephritis in the MRL-Faslpr strain. IL-18 is one plausible candidate. IL-12 and IL-18 are similar in many ways. 1) IL-12 and IL-18 are up-regulated in the kidney and serum of MRL-Faslpr mice with lupus nephritis and further increase with advancing disease (5, 54). 2) IL-12 and IL-18 induce IFN-γ (28, 55). 3) IL-12 and IL-18 are produced in the kidney by TEC (54, 56). 4) Injecting IL-18 alone or with IL-12 promotes proteinuria and glomerulonephritis (57). However, it is important to note that optimal IFN-γ production requires IL-12 plus IL-18 (58, 59). This is based on the findings that IL-18 plus IL-12 synergistically increase IFN-γ production (60), and mice deficient in IL-18 have suppressed IFN-γ expression despite the presence of IL-12, and vice versa (61). Furthermore, it is possible that IL-18 is a more potent inducer of IFN-γ in the MRL-Faslpr strain than in other mouse strains, since an overexpression of the IL-18R accessory β-chain in this strain has been linked to the lymphocyte hyper-responsiveness and enhanced IFN-γ production (62). In the present study IL-12−/− MRL-Faslpr mice maintain increased levels of intrarenal IL-18, and we identified fewer IFN-γ-producing B220 and CD8 T cells in the IL-12−/− MRL-Faslpr kidney compared with the WT strain. Thus, we suggest that deleting IL-12 retards the rate of accumulating IFN-γ-generating B220 and CD8 T cells within the kidney and, in turn, renal disease. It is possible that as the numbers of intrarenal IFN-γ-generating cells induced by IL-18 and other molecules reaches a critical mass, an amplification loop is triggered, leading to kidney destruction. Of note, the decrease in IFN-γ appears to be restricted to end organs (kidney), since systemic (lymph node) IFN-γ is not altered in MRL-Faslpr mice lacking IL-12. Clearly, the relative contributions of IL-12 and other IFN-γ-generating cytokines in the multiple tissues targeted for destruction in MRL-Faslpr mice require further clarification.
We now report that IFN-γ is largely produced by intrinsic renal cells, TEC, in the MRL-Faslpr kidney, while the relative contributions of leukocytes is B220 and macrophages > CD8 T cells > CD4 T cells. Previous studies in an induced form of glomerulonephritis noted that TEC and CD8 T cells generate IFN-γ; in contrast, macrophages and CD4 T cells did not produce IFN-γ (12). On the other hand, macrophage production of IFN-γ has been linked to glomerulonephritis in MRL-Faslpr mice (63). Thus, the cell types producing IFN-γ and their contributions to nephritis may be dependent on the etiology of renal disease. We note that B220 and CD8 T cells produce diminished levels of intrarenal IFN-γ in IL-12−/− MRL-Faslpr compared with the WT strain. As B220 and CD8 T cells are required for renal disease in MRL-Faslpr mice (7), the generation of IL-12-dependent IFN-γ by these T cells may be instrumental in kidney disease. Additional studies are required to determine the exact contribution of each IFN-γ-producing cell population regulated by IL-12 in lupus nephritis.
If IL-12 promotes renal disease solely by increasing intrarenal IFN-γ, then a more direct therapeutic approach might be to directly target IFN-γ. However, it is not that simple. Unfortunately, eliminating IFN-γ is a double-edged sword. On the one hand, substantial evidence supports the concept that deleting IFN-γ prevents renal disease. Our laboratory and others report that genetic deletion of IFN-γ halts kidney disease and the systemic pathology in MRL-Faslpr mice (8, 9, 53). On the other hand, eliminating IFN-γ can promote renal disease (8, 31, 64, 65). Thus, blocking IL-12 maybe a less risky therapeutic for lupus nephritis than blocking IFN-γ, since there is no evidence that blocking IL-12 promotes nephritis. On the other hand, two strategies designed to block IL-12 in lupus mice have not markedly ameliorated nephritis. Provision of anti-IL-12 Ab in the NZB/W F1 hybrid mouse initiated before the onset of lupus inhibits IgG, anti-dsDNA Ab, but does not prevent nephritis (66). Similarly, IL-12p40 transgenic MRL-Faslpr mice that overexpress IL-12p40, a competitive inhibitor of IL-12, do not have substantial alterations in proteinuria, glomerulonephritis, and survival (67). Nevertheless, since beneficial cytokine blockade is dependent on the treatment initiation stage, the dose, and many other factors, it remains to be determined whether eliminating IL-12 reduces lupus.
The levels of systemic IFN-γ and pathology regulated by IL-12 do not necessary correlate. For example, despite improvement in systemic pathology (lung, lacrimal and salivary glands, skin, and lymphadenopathy), we did not note a reduction in IFN-γ in the lymph nodes of IL-12−/− MRL-Faslpr compared with the WT strain. In addition, there was a reduction in IFN-γ in the serum of IL-12p40 transgenic MRL-Faslpr mice, and yet no improvement in systemic pathology (lymphadenopathy, splenomegaly, glomerulonephritis) (67). While injecting IL-12 accelerates renal disease in MRL-Faslpr mice and increases serum IFN-γ, splenomegaly remains similar to that in control mice (22). Clearly, a more detailed analysis will be required to determine whether there is a central role of IL-12 in IFN-γ production.
We point out that other nephritogenic cytokine (CSF-1, GM-CSF, and IL-4) transcripts are diminished in IL-12−/− MRL-Faslpr kidneys. Thus, it is possible that IL-12 may be responsible for increasing each of these nephritogenic cytokines and, in turn, promoting lupus nephritis. Clearly, additional studies are required to determine the interrelationship of IL-12, IFN-γ, and other nephritogenic cytokines in lupus nephritis.
In conclusion, our data suggest that designing a strategy that blocks IL-12 and perhaps other IFN-γ-dependent pathways may be a therapeutic target to halt the progression of lupus nephritis.
We thank Dr. Kevin C. O’Connor for assistance with the dsDNA Ab isotype analysis, Dr. B. S. Vinay Dass for assistance with the ELISA experiments, Dr. Terry K. Means for PCR technical assistance, and Dr. Surya M. Nauli for assistance with the statistical analysis.
↵1 This work was supported by grants from the National Institutes of Health (DK36149 (to V.R.K.), DK56848 (to V.R.K.), DK52369 (to V.R.K.), and 1F32DK6029101 (to D.M.L.)) and the Alliance for Lupus Research (to V.R.K.).
↵2 Address correspondence and reprint requests to Dr. Vicki Rubin Kelley, Harvard Institute of Medicine, 77 avenue Louis Pasteur, Boston, MA 02115. E-mail address:
3 Abbreviations used in this paper: DN, double negative; gcs, glomerular cross-sections; TEC, tubular epithelial cell; WT, wild type.
- Received September 27, 2002.
- Accepted January 31, 2003.
- Copyright © 2003 by The American Association of Immunologists