Data suggests that modulation of FcγRIIB expression represents a significant risk factor for the development of autoimmunity. In this study, we investigated this notion in mice that possess genetics permissible for the development of autoimmunity. To this end, Mrl-MpJ Fcgr2b−/− mice were monitored for the development of autoreactivity. We found that FcγRIIB deficiency led to chronic B cell activation associated with increased germinal center and plasma cell accumulation in the spleen. Likewise, Mrl-MpJ Fcgr2b−/− mice exhibited significant serum IgG reactivity against DNA. We further analyzed the IgG isotype contribution to the anti-dsDNA response and found increases in all subtypes with the exception of IgG3. In particular, we found large increases in IgG1 and IgG2b autoreactivity correlating with significant increases in immune complex deposition and kidney pathology. Finally, we found dendritic cells derived from Mrl-MpJ Fcgr2b−/− mice greatly increased IL-12 expression upon coincubation with apoptotic thymocytes compared with wild-type controls. The results indicate that FcγRIIB is an important regulator of peripheral tolerance and attenuation of the inhibitory signal it provides enhances autoimmune disease on susceptible backgrounds. Additionally, the data indicates FcγRIIB function has a significant impact on APC activity, suggesting a prominent role in dendritic cell activity in response to interaction with particulate autoantigens.
Systemic lupus erythematosus (SLE)3 is an enigmatic disease most likely representing a spectrum of disorders with a common endpoint of autoantibody production (in particular, anti-dsDNA). Investigation into the underlying genetics contributing to disease development and characteristics have revealed a myriad of genes that likely cooperate in an epigenetic fashion to lead to disease manifestation in response to environmental triggers.
In particular, genetic association studies have repeatedly linked members of the Fcγ Ab receptor family with an increased risk of developing SLE and disease severity. For example, decreased copy numbers (<2) for FCGR3B are associated with a significantly increased risk of developing systemic autoimmunity associated with glomerulonephritis in rats and humans while increased copy number (>2) was associated with a slight protection from disease (1). Likewise, polymorphisms in a related FcγR gene, FCGR3A, resulting in homozygosity for a high affinity allele are linked with a rapid progression to end stage renal disease. In contrast, a lower affinity allele is associated with development of SLE in general, suggesting a dichotomy between the roles for individual FcγRs in induction of lupus and effector mechanisms (2).
FCGR2B encodes a single chain receptor that possesses an inhibitory ITIM motif in the cytoplasmic domain. Unique in its inhibitory capacity among the classical Fc receptors, FcγRIIB is widely expressed on cells of hematopoietic origin. Furthermore, FcγRIIB has been found to play a critical role in modulation of immunoreactivity in the macrophage, dendritic cell (DC), and B cell compartments. For example, FcγRIIB is a potent regulator of BCR-mediated signals and thus may influence B cell division, class switching, and maturation to plasma cells (3). Based on this potential, it has been suggested that this molecule may play a prominent role in the maintenance of peripheral tolerance.
Reports have repeatedly found associations between modulation in FcγRIIB expression or signaling potential and risk for developing systemic autoimmunity (i.e., SLE) in mouse and man. For example, a polymorphism resulting in an isoleucine to threonine substitution in the transmembrane region associated with a loss of function is significantly associated with SLE in multiple ethnic groups (4, 5, 6, 7). Likewise, a FcγRIIB haplotype influencing promoter activity has been associated with modulated FcγRIIB expression and SLE development (8, 9). Interestingly, recent studies indicate memory B cells of some SLE patients are defective in up-regulation of FcγRIIB expression, suggestive of decreased B cell regulatory control that may contribute to disease (10). In a similar vein, Enyedy et al. (11) found defects in FcγRIIB inhibitory capacity in lupus patients compared with normal controls. Thus reduction of the constraints placed on B cell activity by FcγRIIB may be a central feature in systemic autoimmune disease.
In mice, the evidence is quite strong that FcγRIIB plays an important role in the maintenance of tolerance. A 13-nucleotide deletion polymorphism proximal to the transcriptional start site is a characteristic shared by the autoimmune prone strains NOD, BXSB, NZB, and MRL (12, 13). This deletion lead to reduced FcγRIIB expression on activated macrophages and germinal center (GC) B cells (12, 14). Furthermore, C57BL6 mice that are congenic for the polymorphism show higher serum Ab titers post immunization and higher levels of spontaneous serum anti-DNA IgG autoantibodies, demonstrating that reduction of surface FcγRIIB in mice has significant functional consequences (15). Likewise, functional inactivation of Fcgr2b renders previously resistant mouse strains susceptible to induced autoimmunity in several models (16, 17).
It appears that Fcgr2b deletion does not affect immunity in a dominant fashion as, in general, deficient mice do not show gross defects in immune-reactivity. However, in the presence of epigenetic modifiers found on the C57BL/6 strain, a fulminate, fully penetrate lupus-like disease develops with high titer autoantibodies, prominent glomerulonephritis, and a significant reduction of life span (18). Genomic mapping of backcrossed progeny revealed three loci that interact with the FcγRIIB deficiency in the development of autoantibodies, splenomegaly, and kidney damage (19). Thus, FcγRIIB may behave as a true susceptibility marker for lupus in the C56BL/6 mouse, epigenetically interacting with several chromosomal loci to lead to disease manifestation in the absence of inhibitory signaling.
To test this hypothesis, we generated the mouse strain Mrl-MpJ Fcgr2b−/− using a speed congenics approach. The Mrl strain was chosen, as they posses some mild immunological defects (i.e., the presence of low titer anti-dsDNA autoantibodies and low levels of glomerulonephritis) in aged mice (>1 year). Furthermore, it is clear that the background is permissive for autoimmune disease development, as it is the parental strain of the well-studied spontaneous lupus model Mrllpr.
We found that Mrl-MpJ Fcgr2b−/− mice developed signs of autoimmunity, including chronic B cell activation, serum reactivity to several autoantigens, and kidney pathology by 8 mo of age. Interestingly, the serum autoantibody profile was imperfectly reflected in isotype distribution of the immune complexes (IC) found in the kidney suggestive of enrichment for nephrophillic autoantibodies in the target organ. Furthermore, for several subclasses of IgG there was a shift toward localization in the tubules. This correlated with the increased tubular atrophy observed in Mrl Fcgr2b−/− mice. In agreement with previous studies, a primary defect in the Mrl Fcgr2b−/− mice seemed to be at the transition from GC B cell to plasma cell (PC) as the deletion of FcγRIIB led to a >8-fold increase in PC accumulation. We also found these mice had defects in the response to apoptotic cells, as injection of apoptotic thymocytes i.v. significantly increased the number of class-switched IgG2a and IgG3 positive PC in the spleen of Fcgr2b−/− mice, whereas the response was minimal in control mice. When we explored this concept further using bone marrow derived-dendritic cells (BMDC), we found that the BMDC from FcγRIIB-deficient mice responded to coincubation with apoptotic cells with increased production of IL-12 compared with Mrl Fcgr2b+/+ or +/− BMDC. Opsonization of the apoptotic cells had no effect in the case of Fcgr2b−/− BMDC, whereas it significantly enhanced the cytokine response of +/− and +/+ BMDC. Thus, it appears that the response is maximal in Fcgr2b−/− BMDC irrespective of opsonin conditions illustrative of a potential defect in APC activity in FcγRIIB-deficient mice.
Materials and Methods
Mrl Fcgr2b−/− mice were generated using speed congenics as previously described (20). Briefly, C57BL/6 Fcgr2b−/− mice were crossed with the Mrl-MpJ strain (Jackson ImmunoResearch Laboratories). The resulting F1 progeny were again crossed to Mrl-MpJ mice, the resultant F2 progeny were screened for 103 satellite markers spanning the genome via PCR, and those with the highest percentage of Mrl genomic sequence and heterozygous disruption of Fcgr2b were selected for continued breeding. This process was repeated for four more generations, at which point the genome was greater than 99.9% Mrl-MpJ in origin. The resultant F6 generation was then intercrossed to result in the Fcgr2b−/− homozygous genotype.
Ab reagents and flow cytometry
For flow cytometry experiments, splenic single-cell suspensions were generated as described previously from 10-mo-old female mice (21). To assay for T cell activation, 106 cells were stained with 1 μg of anti-CD4 allophycocyanin, 1 μg anti CD8 FITC, and 1 μg anti-CD69 PE. To examine B cell activation, 106 splenocytes were stained with 1 μg anti-B220 FITC and 1 μg of anti-CD24 PE. To determine GC B cell numbers, 106 splenocytes were stained with 1 μg anti-B220 allophycocyanin, 1 μg anti-FAS PE, and a 1/1000 dilution of anti-GL7 FITC. To examine the formation of PC, splenic suspensions of 106 cells were stained with 1 μg anti-CD138 PE and 1 μg anti-B220 FITC. PC were defined as the B220low CD138+ population for the purposes of this study. For the analysis of BMDC cultures activation status and purity, 106 BMDC cells were incubated with 1 μg anti-CD11c allophycocyanin, and 1 μg anti-CD80 FITC, 1 μg anti-CD86 PE, or 1 μg anti-MHC class II FITC (I-Ak).
In one set of experiments, 5×106 splenocytes were stained with anti-CD138 PE and anti-B220 allophycocyanin as described above. The cells were then fixed and permeabilized using the cytofix/cytoperm kit according to the manufacturer’s directions (BD Pharmingen). The cells were then stained with a 1/1000 dilution of goat polyclonal FITC-conjugated IgG anti-mouse IgG1, 2a, 2b, or 3 Abs or control goat IgG FITC conjugate (Bethyl Laboratories).
For flow cytometric analysis, at least 105 events (106 in the case of the intracellular staining of PC) were collected on a FACSCalibur flow cytometer (BD Biosciences) and all results were analyzed using the public domain software winmdi. All Abs were purchased from BD Pharmingen unless otherwise indicated.
Assays for serum autoantibodies have been described previously (21). Briefly, Immulon II plates (Dynatech Laboratories) precoated with BSA were coated with 50 μg/ml calf dsDNA (Sigma-Aldrich). To assay for serum autoantibody levels, 100 μl of whole blood was collected from mice by retro-orbital bleed and the serum was separated using blood collection micro tubes according to the manufacturer’s directions (Sarstedt). The serum was diluted 1/100 and assayed for autoantigen reactivity against the plates described above by incubation for 2 h at room temperature. Bound IgG was detected using a goat polyclonal HRP-anti-mouse IgG, IgG1, IgG2a, IgG2b, or IgG3 detection Ab (Bethyl Laboratories) and visualized at 450 nm using TMB substrate (Kirkegaard & Perry Laboratories) according to the manufacturers directions. ELISA to detect anti-cardiolipin, glomerular basement membrane (GBM), glycoprotein G1, and IgG rheumatoid factor IgG autoreactivity were performed using precoated plates (Bethyl Laboratories) according to the manufacturer’s instructions. All samples were assayed in duplicate.
Pathology and immunohistochemistry
Kidneys were fixed with 10% formalin in PBS for 48 h and embedded in paraffin. Kidney sections were stained with Periodic acid Schiff reagent (PAS) to visualize basement membrane structure. To assay for IC and complement deposition, kidneys were embedded in Tissue-Tek OCT compound (Sakura) and snap frozen. Sections (5 μm) were air-dried, fixed with cold acetone, and stained with FITC-conjugated anti-mouse IgG or anti-C3 (Sigma-Aldrich). To assay for IgG subclass deposition, frozen kidney sections were incubated with goat anti-mouse IgG1, 2a, 2b, or 3 (Bethyl Laboratories) for 1 h followed by incubation with FITC-conjugated anti-goat Abs (Sigma-Aldrich). Proteinuria was determined with Chemistrip 2GP dip sticks for urinalysis (Roche). Mice were considered to have significant kidney disease when at least 100 mg/dL of protein was detected in the urine. Glomerulonephritis was scored on PAS-stained sections using a 1–4 scale based on the percentage of glomeruli affected. The grading scale was 1 + 0–15%, 2 + 16–40%, 3 + 41–70%, and 4+ >70% of glomeruli affected. The lesions graded included thickening of the mesangium, noticeable increases in both mesangial and glomerular cellularity with/without superimposed inflammatory exudates and capsular adhesions, glomerulosclerosis, and cast formation. Scoring was performed on at least 200 glomeruli in a 40× field per kidney.
For immunohistochemistry of the spleen snap frozen tissue was sectioned and fixed as described above. The spleen sections were stained with anti-Marco FITC (Serotech), anti-B220 biotin, anti-B220 FITC, or anti-CD4 biotin followed by incubation with alkaline phosphatase-conjugated anti-FITC and HRP-conjugated anti-biotin (Serotech) Abs. After extensive washing in PBS, the Ab staining was visualized using a DAB and BCIP/NBT substrate kit (Vector Laboratories) according to the manufacturer’s directions. For peanut agglutinin (PNA) staining, sections were incubated with PNA-FITC (Vector Laboratories) followed by administration of the AP-conjugated anti-FITC as described above.
Apoptosis induction and BMDC culture
To generate apoptotic thymocytes, the thymuses of 8-wk-old female Mrl-MpJ mice were collected and forced through 70-μm nylon mesh filters (Falcon) to produce single-cell suspensions. RBC were removed via hypotonic lysis, and cell concentration was adjusted to 107 cells/ml in RPMI 5 media. Staurosporine (Sigma-Aldrich) was added to a concentration of 10 μM and the cells were incubated at 37°C for 6 h in 5% CO2. Examination via annexin V and propidium iodide staining revealed that by 6 h, greater than 95% of the thymocytes were apoptotic vs necrotic defined as AnnexinV+ propidium iodide− (data not shown). To further confirm that the cells were apoptotic, intracellular staining was performed to test for the presence of cleaved caspase 3 using an activated caspase 3-FITC Ab (BD Pharmingen). We found that upon 6 h of incubation with staurosporine, greater than 90% of the thymocytes were positive for the activated caspase 3 isoform indicative of initiation of the apoptotic program. The cells were washed extensively with ice cold PBS and resuspended at a concentration of 5×107 cells/ml in PBS. For i.v. injections, 200 μl of the apoptotic thymocyte suspension was injected via the lateral tail vein.
In one series of experiments, thymocytes were opsonized with either heat inactivated pooled normal mouse (Mrl-MpJ) serum or mouse serum from Mrl-MpJ Fcgr2b−/− mice with demonstrated high reactivity against dsDNA. For the opsonization, apoptotic thymocytes were placed in RPMI 5 media containing 20% mouse serum at a concentration of 107 cells/ml. After an incubation of 30 min at 37°C, the cells were washed two times with prewarmed (37°C) PBS. The binding of mouse IgG to the apoptotic cells was confirmed via staining with anti-mouse IgG FITC and analysis via FACS.
For BMDC culture, 8–12-wk-old female Mrl-MpJ mice of the genotypes indicated were sacrificed and the bone marrow was collected from the long bones of the hind limbs as previously described. The cells were washed once in PBS and resuspended at a cell density of 2×106 cells/ml in RPMI 5 (5% FCS, 50 μm 2-ME, 2 mM glutamine, 100 U/ml penicillin, and 100 mg/ml streptomycin) and plated for 2 h at 37°C to remove adherent cells. The non-adherent cells were collected and replated in 6-well culture plates at 4×106 cells/well in RPMI 5 supplemented with 10 ng/ml GM-CSF (PeproTech) for 7–10 days. To confirm the identity and maturation status of the BMDC, they were examined via FACS as described above.
In coculture experiments, 105 BMDC placed in 96-well plates with an equal number of apoptotic thymocytes (either nonopsonized or opsonized as described above) in 200 μl of RPMI 5. After 24 h of coculture, the culture media was collected and assayed for the presence of IL-12p70, TNF-α, IL-10, and IL-6 via ELISA using commercially available kits (R&D Systems).
Effect of FcγRIIB deletion on lymphocyte activation and splenic architecture
To determine whether FcγRIIB deletion affected immune system homeostasis in Mrl FcγRIIB−/−, splenic lymphocytes were assayed via FACS to ascertain levels of expansion and activation compared with the wild-type Mrl-MpJ strain. At 8 mo of age, there was significant splenomegaly in Mrl Fcgr2b−/− mice (176 mg ± 24 mg) compared with the parental strain (76 mg ± 8 mg). Furthermore, we found that CD4+ T cell activation was slightly increased, as measured by the percentage of CD69+CD4+ splenic T cells, although the difference was not statically significant (data not shown). In contrast to Mrllpr mice, we did not detect the presence of the CD4/CD8 double positive cells characteristic of the lupus-prone strain.
When we examined activation in the B cell compartment, we found, similar to C57BL/7 Fcgr2b−/− mice, a substantial portion of the B220+ B cell population had down-regulated CD24 expression in Mrl Fcgr2b−/− mice indicative of chronic B cell activation (Fig. 1, a and c). Accordingly, there was a 50% increase in the number of GL7+Fas+ GC B cells when compared with age-matched MrL-MpJ mice (2.5% ± 0.4 for Mrl Fcgr2b−/− vs 1.36% ± 0.3 for Mrl-MpJ mice; Fig. 1,c). It is noteworthy that the Mrl-MpJ parental strain possessed levels of B cell activation that were significantly higher than age-matched C57BL/6 mice (Fig. 1, a and c), suggestive of the underlying immunological abnormalities in the Mrl strain. The increase in GC B cells was reflected in immunohistochemical analysis, where we found an increase in GC numbers and size in Mrl Fcgr2b−/− mice compared with Mrl controls as revealed by PNA staining (Fig. 1 b).
Likewise, there was a large accumulation of CD138+ PC in the spleen of Mrl Fcgr2b−/− mice similar to what has been observed previously in C57BL/6 Fcgr2b−/− mice, suggestive of a defect in control of transition from GC B cell to PC in the spleen (4.7% ± 1.1 B220lowCD138+ vs 0.5% ± 0.2 B220lowCD138+ for age-matched Mrl-MpJ mice; Fig. 1, a and c). Overall, we found no gross morphological alterations in splenic architecture (Fig. 1 b). Thus, the data indicates that FcγRIIB deficiency impacts B cell function in Mrl mice leading to increased B cell activation indicative of potential autoimmune reactivity.
Serum autoimmunity in Mrl Fcgr2b−/− mice
To ascertain the level of serum autoreactivity, samples were collected from 9- to 11-mo-old female Mrl Fcgr2b−/−, Mrl-MpJ, and C57BL/7 mice. The serum was evaluated for the presence of reactivity against dsDNA (Fig. 2,a), cardiolipin, GBM (Fig. 2,b), glycoprotein G1, and IgG rheumatoid factor via ELISA. Although aged C57BL/6 mice showed low reactivity to any Ag tested in either the IgM or IgG isotypes (Fig. 2), the Mrl-MpJ mice exhibited low to moderate levels of anti-dsDNA Abs in the IgG compartment, particularly in the IgG2a and IgG3 subclasses. This is in accord with the increased basal autoreactivity described for this strain.
Deletion of FcγRIIB led to a 2-fold increase in the OD for anti-dsDNA IgG in Mrl mice (mean value of 0.77 vs 0.38 for Mrl-MpJ mice; Fig. 2,a). In contrast, while there was a trend toward increased IgM anti-dsDNA reactivity, the change was not significant in accordance with previous observations in C57BL/6 Fcgr2b−/− mice. When we examined the contribution of IgG subclasses to the anti-DNA reactivity, it was revealed that there were large increases in IgG1 and IgG2b anti-dsDNA levels and a lower but significant increase in the IgG2a subclass (Fig. 2 a). Conversely, there was little change in anti-dsDNA reactivity in the IgG3 compartment in the absence of FcγRIIB.
Likewise there was a considerable increase in IgG reactivity to the autoantigens cardiolipin and GBM in the Fcgr2b−/− mice compared with Mrl control animals (Fig. 2 b). However, the anti-cardiolipin reactivity was not found to be associated with RBC destruction or immune thrombocytopenia (data not shown). Finally, there was no detectable reactivity to the autoantigen glycoprotein G1 or IgG rheumatoid factor activity. Therefore, it appears that the loss of inhibitory signaling in Mrl Fcgr2b−/− mice lead to enhanced serum autoreactivity to a variety of self Ags indicative of active autoimmune disease.
IC deposition in the kidney of Mrl Fcgr2b−/− mice
We next examined the kidneys of 10-mo-old female Mrl Fcgr2b−/−, Mrl-MpJ, and C57BL/6 mice for the presence of IC and complement fixation in the glomeruli. To this end, sections of kidney were stained with FITC-labeled anti-mouse IgG, complement component C3, IgG1, 2a, 2b, or 3 Abs, and visually examined for staining patterns.
There was punctuate basal staining for IgG in the glomeruli of C57BL/6 mice, which is typical for many strains of nonautoimmune mice. This was not associated with significant complement fixation in the kidney, further indicating the nonpathogenic nature of the IC found in C57BL/6 sections (Fig. 3,a). In Mrl-MpJ mice, there were increased levels of IC detectable in the glomeruli when compared with C57BL/6 mice. The pattern of staining was primarily mesangial with little to no linear staining of the capillary network. However, like C57BL/7 kidney sections, there was little complement fixation in the glomeruli. In contrast, Mrl Fcgr2b−/− mice had significant increases in IgG deposition compared with both C57BL/6 and Mrl-MpJ mice (Fig. 3 a). The staining was primarily mesangial in Mrl Fcgr2b−/− kidney sections with the majority of the glomerular surface staining positive for IgG IC contrasting with the Mrl-MpJ sections where the IgG staining typically covered less than half of the surface area of the glomeruli. Furthermore, there was significant complement fixation in the Mrl Fcgr2b−/− glomeruli.
When IgG subclass of the IC was examined in C57BL/6 mice, it was determined that the background IC deposition was exclusively contributed to by the IgG2b isotype (Fig. 3,b). When Mrl-MpJ kidneys were examined, we found significant staining for the IgG1, IgG2a, and IgG2b subclasses and weak staining for IgG3 (Fig. 3 b). This corresponded to the isotype distribution for serum autoreactivity to dsDNA in Mrl-MpJ mice, indicative of the correlation between serum anti-DNA reactivity and renal involvement that has been observed in SLE. When kidney sections from Mrl Fcgr2b−/− mice were examined for IgG subtype contribution to IC deposition, we found significant increases in several IgG subclasses. Furthermore, the glomerular staining in Mrl Fcgr2b−/− sections was uniformly high for all IgG subclasses and appeared primarily mesangial, similar to the staining pattern for total IgG.
In the IgG1 compartment, there was increased glomerular deposition in Mrl Fcgr2b−/− sections compared with Mrl-MpJ kidney. Moreover, the Fcgr2b−/− mice had greater IgG1 and IgG2a deposition in the tubules compared with Mrl-MpJ and C57BL/6 mice. In particular, IgG2a staining in the tubules was uniformly strong through out the sections and of nearly equal intensity to the staining occurring in the glomeruli (Fig. 3,b). Interestingly, when we examined IgG2b content of the IC we found no difference between the Mrl Fcgr2b−/− and Mrl control mice (Fig. 3,b). This is in stark contrast to the serum IgG2b anti-DNA levels (Fig. 2,a) indicative of a dichotomy between serum autoreactivity and Ab localization in target organs. Likewise, there was a large increase in IgG3 deposition in the kidneys of Mrl Fcgr2b−/− mice, which was not reflected in the serum autoimmune profile (Fig. 2 a). Recent reports have suggested that the kidney can serve as a depot enriching several-fold the levels autoantibodies present over serum concentrations (22). A similar phenomenon may be occurring in this case with enrichment of particular pathogenic subclasses, most notably IgG3, in the glomeruli of Mrl Fcgr2b−/− mice.
Kidney disease in Mrl Fcgr2b−/− mice
Although Mrl-MpJ mice show increased propensity to develop anti-dsDNA autoantibodies late in life, in general they do not develop significant kidney disease. To examine the impact that FcγRIIB deficiency had on kidney structure and function, we examined paraffin-embedded, PAS-stained sections from 10-mo-old Mrl Fcgr2b−/− mice to assess what, if any, pathology had occurred. Age- and sex-matched Mrl-MpJ mice showed some instances of mild glomerular hypercellularity, but this was not associated with significant nephritis, glomerulosclerosis, or tubular involvement (Fig. 4,a). In contrast, Mrl Fcgr2b−/− mice presented with severe kidney pathology including extensive hyperproliferation in the glomeruli, glomerulosclerosis with crescent formation, tubular and glomerular cast formation, and prominent interstitial nephritis (Fig. 4,a). When the kidney sections were scored for levels of renal involvement, we found all Mrl Fcgr2b−/− mice examined had extensive glomerular alteration scoring the maximum on the scale (average of 4, Fig. 4,b). In contrast, the glomerulonephritis score in Mrl-MpJ mice was much more mild (average of 1.4, Fig. 4,b). Finally, when we determined protenuria as a measure of kidney disease, we found all of the Mrl Fcgr2b−/− mice had moderate to severe protenuria whereas the Mrl-MpJ mice had no evidence of protein in the urine (Fig. 4 c). When other tissues were examined for development of inflammatory disease (i.e., liver, lung, heart, and salivary glands), we found no evidence of pathology indicating the kidney is the primary target organ in Mrl Fcgr2b−/− mice. This, in contrast to C56BL/6 Fcgr2b−/− mice that exhibit vasculitis in the lung and inflammation of several organs, is illustrative of the influence of background genetics on the characteristics of autoimmunity that develops in Fcgr2b−/− animals.
Surprisingly, Mrl Fcgr2b−/− mice did not show increased mortality compared with Mrl-MpJ mice at 1 year, despite significant kidney pathology (data not shown). This could be related to the ability of Mrl-MpJ mice to completely heal some damaged tissues with the absence of pathologic scar tissue formation and essentially a complete restoration of tissue structure and function (23, 24, 25). In particular, the Mrl-MpJ mouse has the ability to heal damage with the absence of collagenous tissue deposition. Thus, Mrl-MpJ mice may possess some inherent resistance to inflammatory pathology associated with autoimmune disease.
Fcgr2b deficiency and the response to apoptosis
The data presented herein and elsewhere supports the notion that a major defect in FcγRIIB-deficient animals lies in the B cell compartment, specifically in the transition from GC centrocyte to PC. However, the wide distribution of FcγRIIB expression lends to the possibility that other cell types may influence the development of disease. In particular, several recent studies illustrate the importance of FcγRIIB in the regulation of DC function (26, 27). Thus, to gain a more complete understanding of the mechanisms behind lupus development in the absence of FcγRIIB-mediated inhibition, we examined the behavior of DCs in Mrl Fcgr2b−/− mice. Furthermore, as a major defect in SLE is in the response to and clearance of apoptotic cell debris, we chose to examine APC response to apoptotic thymocytes under normal and autoimmune conditions.
As a preliminary experiment, 107 thymocytes rendered apoptotic were transferred i.v. to Mrl Fcgr2b−/− and Mrl-MpJ mice, and 3 days later, the splenic B cell population was examined via FACS for the development of primary plasma cell foci. The rationale for this experiment is based on earlier studies indicating that under normal conditions a single bolus of apoptotic cells delivered to an immuno-competent animal will provoke little to no response and repeated administration of apoptotic material is necessary to generate autoimmunity (28). Our hypothesis was that in the absence of inhibitory FcγRIIB signaling, this restraint might be removed allowing for a response to single administrations of apoptotic thymocytes. Previous experiments found that CFSE-labeled apoptotic thymocytes transferred i.v. localized primarily to the marginal zone of the spleen, thus, we expected the bulk of any response to be manifested in this organ (29).
When CD138+ PCs were examined for IgG subclass representation among the population, in general, we observed that 15–20% of the splenic PC population was IgG producing (Table I) in Mrl-MpJ mice and disruption of FcγRIIB did not perturb the percentages to an appreciable degree. Furthermore, administration of apoptotic thymocytes had no effect on the percentages of IgG1, 2a, 2b, or 3+ CD138+ PC in the parental Mrl strain in agreement with the observed lack of reactivity to apoptotic cells described above. In contrast, in the absence of FcγRIIB, administration of apoptotic thymocytes lead to substantial increases in the percentage of IgG2a+ and IgG3+ PC compared with basal levels (Table I). In particular, the increase in IgG3+ PCs was greater than 3-fold over background levels. This data suggest that the lack of FcγRIIB leads to a defective primary response to apoptotic material. Furthermore, the pattern of PC formation was indicative of an inflammatory reaction as IgG2a and IgG3 class switching are driven by inflammatory cytokine production. Accumulated data indicates DCs regulate class switching and B cell activation in the GC via contact and cytokine-dependent mechanisms (30, 31, 32). Thus, we examined whether deficiency in FcγRIIB lead to defective dendritic cell/apoptotic cell interactions in vitro.
|.||+Apoptotic Thymocytes .||.||.||.|
|.||MRL FcγRIIB−/− .||Mrl MpJ .||MrL FcγRIIB−/− .||Mrl MpJ .|
|IgG1||3.4% ± 1.9||4.1% ± 1.3||1.8% ± 2.6||4.3% ± 3.6|
|IgG2a||9.8% ± 3.5b||5.1% ± 0.7||4.87% ± 1.9||5.96% ± 2.5|
|IgG2b||5.3% ± 0.5||5.8% ± 3.5||4.33% ± 3.1||5.67% ± 1.1|
|IgG3||14.5% ± 3.3c||3.9% ± 1.4||3.36% ± 3.0||5.11% ± 0.9|
|.||+Apoptotic Thymocytes .||.||.||.|
|.||MRL FcγRIIB−/− .||Mrl MpJ .||MrL FcγRIIB−/− .||Mrl MpJ .|
|IgG1||3.4% ± 1.9||4.1% ± 1.3||1.8% ± 2.6||4.3% ± 3.6|
|IgG2a||9.8% ± 3.5b||5.1% ± 0.7||4.87% ± 1.9||5.96% ± 2.5|
|IgG2b||5.3% ± 0.5||5.8% ± 3.5||4.33% ± 3.1||5.67% ± 1.1|
|IgG3||14.5% ± 3.3c||3.9% ± 1.4||3.36% ± 3.0||5.11% ± 0.9|
A total of 107 apoptotic thymocytes were administered i.v. to 12-wk-old female Mrl-MpJ and Mrl Fcgr2b−/− mice as described. Three days post-administration, splenocytes were examined for the numbers of class-switched, CD138+B220low plasma cells. Numbers represent the percentage of CD138+B220low cells, which was positive for the intracellular presence of each IgG sub-type indicated for five animals ± the SD. Value of p was determined via the unpaired student’s t test.
p = 0.02,
p < 0.01.
To examine the impact of FcγRIIB deficiency on DC responsiveness, BMDC were generated as described in Materials and Methods. The BMDC cultures were greater than 90% CD11chigh, MHCIIint, CD40+, and CD86+ at 6 days post culture initiation, indicating that the cells were of the DC lineage (data not shown).
When the BMDC were incubated with “naked” (i.e., nonopsonized) apoptotic thymocytes at a 1:1 ratio, there was no significant production of IL-10 or TNF-α (data not shown). Likewise, the BMDC produced very low levels of IL-6 in response to the coculture (Fig. 5,a). In contrast, BMDC from Mrl Fcgr2b−/− mice produced significant amounts of IL-12 in response to coculture with apoptotic thymocytes (141 ± 20 pg/ml; Fig. 5,b). This level of production was 3-fold higher than the IL-12 secretion observed from Mrl Fcgr2b−/+ (57.7 ± 6 pg/ml) or Mrl Fcgr2b+/+ (58 ± 9.8 pg/ml) BMDC (Fig. 5,b). That the increased activation occurred in the absence of mouse Abs indicated that FcγRIIB plays a role in maintenance of the basal nonreactive state in monocytic DC populations. We next investigated the effect thymocyte opsonization would have on BMDC IL-12 production under both normal and autoimmune serum conditions. To test this, apoptotic thymocytes were incubated in media containing either 20% normal mouse serum (NMS) or 20% serum obtained from mice with demonstrated high anti-chromatin reactivity (autoimmune mouse serum [AMS]; Fig. 5) as indicated in Materials and Methods. The opsonized thymocytes were incubated at a 1:1 ratio with the BMDC, and 24 h post culture initiation cytokine levels were determined. We found opsonization had no effect on IL-10 or TNF-α production (data not shown). There was a slight decrease in IL-6 production in the NMS-opsonized groups (Fig. 5,a) that was further diminished in the AMS-opsonized group (Fig. 5,a). Furthermore, while the decrease in IL-6 production was significant after opsonization of the apoptotic particles, there was no significant difference between the different BMDC cultures (Fig. 5,a). In contrast, opsonization enhanced IL-12 synthesis significantly in Mrl-MpJ Fcgr2b+/− and Mrl-MpJ Fcgr2b+/+ BMDC cocultures (Fig. 5,b). When apoptotic thymocytes were opsonized with NMS, IL-12 production by Mrl-MpJ Fcgr2b+/− BMDC increased 2-fold. A similar increase was observed in Mrl-MpJ Fcgr2b+/+ BMDC cultures, although the increase was significantly lower than that observed in both Mrl-MpJ Fcgr2b+/− and Mrl-MpJ Fcgr2b−/− BMDC cultures (Fig. 5,b). In the presence of AMS, IL-12 production was 3-fold higher than that seen in the absence of opsonization (Fig. 5 b) for both Mrl-MpJ Fcgr2b+/− and Mrl-MpJ Fcgr2b+/+ BMDC cultures although the Mrl-MpJ Fcgr2b+/− BMDC culture continued to exhibit higher secretion of IL-12 in comparison to Fcgr2b+/+ BMDC cultures. In contrast, in the absence of FcγRIIB, the BMDC failed to increase IL-12 production in the presence of either NMS- or AMS-opsonized thymocytes although the level of IL-12 was already significantly higher than that of the other groups of BMDC. Thus, it is likely that in the absence of FcγRIIB, IL-12 synthesis was already at a near maximal level, indicative of the central role FcγRIIB plays in the regulation of DC activity. It is of interest to note that the heterozygous Fcgrb+/− BMDC appeared to express IL-12 at a level that was intermediate between the Fcgr2b−/− and +/− BMDC cocultures after opsonization of the thymocytes yet fails to respond significantly in the presence of bare apoptotic particles. This is in accord with the in vivo observation that C57BL/6 Fcgr2b+/− mice fail to exhibit any of the immunological abnormalities of FcγRIIB-deficient mice, suggesting that a complete absence of FcγRIIB is necessary to develop lupus in mice and indicating that at least part of the defect lies in DC responsiveness to apoptotic cell interactions.
The role of FcγRIIB in the development of autoimmune disease remains an inscrutable question. Several studies suggest a link between polymorphisms in the promoter or coding region of FCGR2B and increased risk of developing systemic autoimmune disease in the clinical setting. For example, the −343 C/C genotype is significantly associated with increased risk for the development of SLE in European populations (9). However, a significant problem with such association studies is the close linkage of the FCGR2 and FCGR3 genes on chromosome 1q23 that could result in false association to disease via linkage disequilibrium (LD). Indeed, meta-analysis suggests a very strong LD between FCGR2B and FCGR3B and somewhat weaker linkage between 2B and 3A (7). LD analysis indicated that each FcγR found to be associated with increased risk of SLE development did so in an independent manner, yet these studies also indicated that FcγRIIB might play a more prominent role in disease characteristics rather than the development of pathology in general (7). Therefore, it is unclear what influence the FcγRIIB may have on the maintenance of tolerance in man.
To better understand its regulatory role in immunostasis and epigenetic interaction with other susceptibility loci, we generated Mrl Fcgr2b−/− mice via a speed congenics approach. Previous studies had found that homozygous disruption of Fcgr2b only led to systemic autoimmunity in the presence of epigenetic modifiers termed sbb1, 2, and 3 found on chromosomes 9, 12, and 17, respectively, of the C57BL/6 genomic background (18, 19). The implication is modulation of FcγRIIB function can have significant impact on B cell immunostasis, but can only do so in the presence of other susceptibility loci impacting disparate mechanisms of tolerance. Thus, in effect, this would suggest that FcγRIIB behaves as a true susceptibility loci in the murine setting, influencing the development of autoimmunity under conditions of reduced signaling capacity, but unable to induce disease outright of its own accord. The Mrl mouse strain provides a useful genetic background to study this phenomenon of epistasis and autoimmunity due to the permissive genomic background for the development of autoimmunity. Although occurring in a much more mild form than the well-studied Mrllpr strain of mice, analysis indicates the presence of at least 12 susceptibility loci spanning the genome that contribute to disease susceptibility in Mrl-MpJ mice (33). Like the C57BL/6 Fcgr2b−/− strain, we found that Mrl Fcgr2b−/− mice developed prominent autoimmune disease. This included the development of high levels of serum reactivity to chromatin, IC deposition in the kidney, and chronic B cell activation. Furthermore, attenuation of FcγRIIB function resulted in the accumulation of plasma cells in the spleen of Mrl Fcgr2b−/− mice, indicating a primary mechanism by which FcγRIIB promotes tolerance is by regulating the transition from germinal center B cell to plasmablast (34). However, disease differed from that in C57BL/6 Fcgr2b−/− mice in several fundamental ways. For instance, we found no incidence of inflammatory disease in tissue other than the kidney whereas in C57BL/6 Fcgr2b−/− mice, there is considerable vasculitis in the lungs as well as frequent inflammation of other organs (18). Likewise, in Mrl Fcgr2b−/− mice, IC depositions in the kidney were found to contain all IgG isotypes with a particular accumulation of IgG3. This is in stark contrast to C57BL/6 Fcgr2b−/− kidney that develops accumulation primarily of IgG2b IC (35). Similarly, in C57BL/6 Fcgr2b−/− mice, IC deposition is limited primarily to the glomeruli, even in cases of severe pathology (unpublished observations) whereas in Mrl Fcgr2b−/− animals, Ab deposition could also be detected in the tubules.
Recently, the relative contribution of FcγRIIB in the development of murine lupus has been called into question. Fcgr2b lies in a region of murine chromosome 1 in close proximity to several potential lupus susceptibility loci including the gene encoding serum amyolid P (SAP), a serum pentraxin that may aid in chromatin removal and clearance of apoptotic cells from circulation (36, 37). Disruption of SAP led to defects in chromatin degradation and an increased incidence of anti-nuclear autoreactivity (38). However, later studies by Bygrave et al. (39) revealed the possibility that the autoimmunity that arises in the SAP−/− mice is due to the residual 129/sv background DNA. This argument was furthered by Dr. Edward Wakeland’s group who suggested the SLAM/CD2 loci, which is located at 1q23 adjacent to Fcgr2b, is responsible for the autoimmunity observed in female Fcgr2b−/− C57BL/6 congenic mice, as this line was bred from a founding strain with a mixed Sv129/C57BL6 genotype (18, 40). However, we dispute the “carry-over DNA” hypothesis for two reasons. First, as the suggestion has been that the SLAM/CD2 loci from the 129sv background is responsible for autoimmunity in knock-out mice for genes in the region of 1q23, before speed congenics was undertaken to generate Mrl Fcgr2b−/− mice, we performed fine mapping of the region and selected animals for crossing who possessed C57BL/6 genome at the SLAM loci eliminating this confounding factor. Secondly, in a previous publication, we demonstrated that increased expression of FcγRIIB, by means of retroviral transduction, prevented autoimmunity in C57BL/6 Fcgr2b−/− mice indicating a major role the receptor plays in the development of autoimmunity in mice (21). Thus, it is likely that the autoimmune phenotype observed is due to the FcγRIIB deficiency and not to residual 129sv sequence in the mice.
Likwise, recent reports have called into question the correlation between anti-dsDNA Ab production and nephritis. Congenic dissection of the inbred lupus-prone strain NZM 2328 revealed that separate genomic loci are responsible for anti-dsDNA autoantibodies and glomerulonephritis (41). C57BL/6 mice congenic for the loci responsible for glomerulonephritis (NZM.C57Lc4) developed severe protenuria with a high penetrance, yet failed to exhibit significant serum anti-dsDNA titers. In contrast, these animals seemed to develop a response primarily against kidney constituents, suggesting renal-specific autoimmune reactions may drive disease. Our results revealed that the IgG isotype composition of the IC in the kidney (Fig. 3,b) was dissimilar to that observed in the serum anti-dsDNA autoreactivity (Fig. 2,a), whereas there was an increase in serum anti glomerular basement membrane IgG autoreactivity (Fig. 2 b). Thus, the data may further support the supposition that serum anti-dsDNA reactivity, while a suitable marker for the development of systemic autoreactivity, may not be directly applicable to target organ damage in the kidney. Moreover, these results further indicate that FcγRIIB plays a central role in the regulation of tolerance to several potential autoantigens.
Finally, if the IC deposition pattern is examined in age-matched Mrl-MpJ mice, it can clearly be discerned that IC deposition is similar in isotype composition and structural localization compared with Mrl Fcgr2b−/− animals (Fig. 3 b). Thus, it appears that removal of FcγRIIB constraints on the immune system amplify underlying autoreactivity driving already present immunological defects to pathologic autoimmunity. This hypothesis is supported by the observations of Bolland et al. (19) signifying that autoimmunity in C57BL/6 Fcgr2b−/− mice is driven by loci that control splenomegaly and autoantibody production working in cooperation with the FcγRIIB deficiency. Currently, our laboratory is generating congenics of several of the Mrl susceptibility loci to examine how each may interact with FcγRIIB affecting tolerance in vivo.
B cells are believed to be the primary immunological compartment affected by modulation of FcγRIIB expression. This is supported by data indicating that reduced FcγRIIB functional capacity leads to increased Ca2+ flux upon BCR crosslinking, enhanced Ab responses TD and TI Ags (42), accumulation of plasmablasts in autoimmune settings (Fig. 1 b, and Ref. 34), or, conversely, to protection from autoimmune disease when expression is enhanced (21). Indeed, when we administered apoptotic thymocytes to Mrl Fcgr2b−/− mice, we found that the mice responded with an abnormal accumulation of IgG2a+ and IgG3+ plasma cells, suggesting that the lack of FcγRIIB inhibition allows for inappropriate responsiveness to autoantigens.
However, several recent papers suggest that FcγRIIB can have a strong impact on DC function as well. For example, selective blockade of FcγRIIB in vitro led to spontaneous maturation of immature human DCs due to the presence of ICs in the culture media interacting with the activating FcγRIIA and IIIA (27). Furthermore, Dhodapkar et al. (26) demonstrated blockade of FcγRIIB induced the type I IFN signaling pathway in monocyte-derived DC. This is an interesting finding, as the induction of signal transduction seemed not to require the synthesis of type I IFN, suggesting a novel mechanism by which FcRs may contribute to autoimmunity. Our data supports this concept and suggest that FcγRIIB provides a central regulatory mechanism in the response of APCs toward potentially autoreactive particulate Ag.
This hypothesis was further strengthened by the observation that Mrl Fcgr2b−/− BMDCs responded with significant IL-12 production upon incubation with apoptotic thymocytes. Recent studies illustrated the central role IL-12 plays in the development of autoreactivity. Dai et al. (43) found that IL-12 produced by DCs is required for the development of lupus in murine models of the disease. In this study, disease induction was found to require the activity of both B and T cells and removal of IL-12 drastically reduced anti-nuclear Ab production as well as kidney pathology. Taken together with our results, this suggests that a central mechanism by which FcγRIIB may regulate autoimmunity is via the prevention of inappropriate activation of DCs upon encounter with endogenous particulate matter, thereby dampening inflammation and the activation/expansion of autoreactive lymphocyte populations.
DCs interact with opsonized apoptotic particles primarily via FcγR-dependent mechanisms (44). Furthermore, autoimmune sera from SLE patients enhances this effect, presumably due to high titers of anti-chromatin Abs (45). It can be envisioned that under these circumstances, FcγRIIB would be a critical roadblock to autoimmune disease progression; however, our observation that FcγRIIB−/− DCs responded with equal intensity to apoptotic cells independent of opsonization suggests that FcγRIIB may serve an essential role in negating DC activation upon interaction with apoptotic cells under normal circumstances. Experiments are underway to better define the role of FcγRIIB in DC-apoptotic cell interactions under normal and autoimmune conditions.
In conclusion, our report demonstrated the significant impact attenuation of FcγRIIB function can have in a susceptible genetic background reflecting observations in the clinical setting and further strengthening the supposition that FcγRIIB is an SLE susceptibility loci. Furthermore, the data presented herein suggest that FcγRIIB plays an essential role in regulation of APC-apoptotic cell interactions, which could have implications for the development of therapeutic modalities.
We thank Dr. Falk Nimmerjahn for stimulating discussions during the course of the study as well as Patrick Smith, Jose Pagan, and Peter Curry for expert technical assistance.
The authors have no financial conflict of interest.
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.
This work was supported by grants from the National Institutes of Health (to J.V.R.) and a national research service award (to T.L.M.).
Abbreviations used in this paper: SLE, systemic lupus erythematosus; IC, immune complex; GC, germinal center; PC, plasma cell; DC, dendritic cell; BMDC, bone marrow-derived dendritic cell; PNA, peanut agglutinin; GBM, glomerular basement membrane; PAS, Periodic acid Schiff reagent; NMS, normal mouse serum; AMS, autoimmune mouse serum; LD, linkage disequilibrium; SAP, serum amyolid P.