|Home | About | Journals | Submit | Contact Us | Français|
B cells play critical roles in the pathogenesis of lupus. To examine the influence of B cells on disease pathogenesis in a murine lupus model, New Zealand Black and New Zealand White F1 hybrid (NZB/W) mice were generated that were deficient for CD19 (CD19−/− NZB/W mice), a B cell-specific cell surface molecule that is essential for optimal B cell signal transduction. The emergence of anti-nuclear antibodies was significantly delayed in CD19−/− NZB/W mice in comparison with wild type NZB/W mice. However, the pathologic manifestations of nephritis appeared significantly earlier and survival was significantly reduced in CD19−/− NZB/W mice in comparison with wild type mice. These results demonstrate both disease promoting and protective roles for B cells in lupus pathogenesis. Recent studies have identified a potent regulatory B cell subset (B10 cells) within the rare CD1dhiCD5+ B cell subset of the spleen that regulates acute inflammation and autoimmunity through the production of IL-10. In wild type NZB/W mice, the CD1dhiCD5+B220+ B cell subset that includes B10 cells was increased by 2.5-fold during the disease course, while CD19−/− NZB/W mice lacked this CD1dhiCD5+ regulatory B cell subset. However, the transfer of splenic CD1dhiCD5+ B cells from wild type NZB/W mice into CD19−/− NZB/W recipients significantly prolonged their survival. Furthermore, regulatory T cells were significantly decreased in CD19−/− NZB/W mice, but the transfer of wild type CD1dhiCD5+ B cells induced Treg cell expansion in CD19−/− NZB/W mice. These results demonstrate an important protective role for regulatory B10 cells in this systemic autoimmune disease.
Systemic lupus erythematosus (SLE) is a prototypic multisystem autoimmune disease characterized by the production of autoantibodies and the involvement of most organ systems (1). Recent studies have demonstrated a critical role for B cells in SLE pathogenesis (2–4). In addition to autoantibody production, abnormal B cell activities or functions such as cytokine production and Ag presentation are likely to contribute to SLE development. Indeed, B cell-targeted therapies including mAbs to CD20, CD22, and BAFF are currently under evaluation in the treatment of human SLE patients (5–8).
B cell activation depends on BCR-generated signals during immune responses to self and foreign antigens (9). Cell surface and intracellular molecules that inform B cells of their microenvironment, such as CD19, CD22, Fc receptors, and TLRs, also play critical roles in controlling B cell responses (10). Among these molecules, CD19 serves as a positive response regulator that amplifies the strength and duration of BCR and other signaling events by regulating Src-family protein tyrosine kinases, and other effector molecules (11–19). CD19 is a 95-kDa member of the Ig superfamily and is expressed on B cells and potentially follicular dendritic cells. CD19-deficient (CD19−/−) mice are hyposensitive to a variety of transmembrane signals (20, 21), while B cells from transgenic mice that overexpress CD19 are hyperresponsive to transmembrane signals and generate autoantibodies spontaneously (22, 23), suggesting that altered CD19 function or expression can influence B cell susceptibility to autoimmunity (24). Thereby, selective targeting of CD19 might be a less-invasive B-cell-directed strategy for treating SLE rather than total B cell depletion.
As a well-established murine lupus model, New Zealand Black (NZB) and New Zealand White (NZW) F1 hybrid mice (NZB/W mice) spontaneously develop a SLE-like disease in which IgG anti-dsDNA autoantibody production is associated with immune complex-mediated glomerulonephritis (25). Aged NZB/W mice have increased numbers of splenic CD23loCD21hi marginal zone B cells as well as increased numbers of peritoneal B220intCD5+ B1 cells, although their significance in the pathogenesis has been unclear (26–29). Recent studies have identified a phenotypically unique subset of spleen “regulatory B cells” that share phenotypic markers with both B-1 and marginal zone B cells (30–33). A portion of these rare CD1dhiCD5+ B cells are competent for IL-10 production and have therefore been called B10 cells (34). B10 cells and potentially other regulatory B cell subsets negatively regulate inflammation and autoimmune disease in mice, including contact hypersensitivity, experimental autoimmune encephalomyelitis (EAE), inflammatory bowel diseases, and arthritis (30–40). Both contact hypersensitivity responses and EAE are augmented in CD19−/− mice due to the absence of B10 cells (34, 41, 42). Thus, while B cells and autoantibodies play major pathogenic roles in NZB/W mice, B cells may also contribute to the suppression of the disease. In this context, we assessed the effect of CD19 deficiency on disease initiation and progression in NZB/W mice.
NZB, NZW, and C57BL/6 mice were purchased from Japan SLC, Inc. (Shizuoka, Japan). CD19−/− mice were generated as described (21) and backcrossed onto a C57BL/6 genetic background ≥12 times. CD19−/− mice were also backcrossed 12 times onto the NZB or NZW genetic backgrounds to obtain CD19−/− NZB mice and CD19−/− NZW mice. Female NZB/W mice were generated by mating female NZB and male NZW mice. Female CD19−/− NZB/W mice were generated by mating female CD19−/− NZB and male CD19−/− NZW mice. Mice were housed in a specific pathogen-free barrier facility. All procedures were approved by the Animal Committee of International Medical Center of Japan.
Serum samples were obtained from NZB/W mice and CD19−/− NZB/W mice every 2 weeks for determining serum IgG ANA levels. To determine ANA positivity, serum was diluted 1:100 and added to fixed HEp-2 cell ANA slides (MBL, Nagoya, Japan) with FITC-conjugated goat anti-mouse IgG (H+L) (ICN Biomedical, Inc. Costa Mesa, CA) used as the indirect immunofluorescence detection reagent at predetermined optimal concentrations. Also, sera at 12, 20, 28, and 36 wk of age were diluted 1:40, 80, 160, 320, 640, 1280, and 2560 to determine ANA titers, and were assessed as above. Immunofluorescence staining of the slides was evaluated on a fluorescent microscope at 400x. The serum levels of IgG anti-dsDNA Abs were measured using dsDNA-coated 96-well ELISA plates (MESACUP; MBL). Sera were diluted 1:100, added to the ELISA plates, and allowed to react for an hour at room temperature. Subsequently, the plates were washed three times before adding predetermined optimal concentrations of HRP-conjugated anti-mouse IgG Ab (Cappel, MP Biomedical, Irvine, CA). Ab binding was evaluated using TMB substrate (Bethyl Laboratories, Inc., Montgomery, TX), with the reactions stopped using 1N H2SO4, and read at a wavelength of 450 nm. A high-titer serum was plated in serial dilutions on each plate for quantification. The OD units were arbitrary determined by taking a ratio between the OD value obtained for the test sample and that obtained for the high-titer sample at the same dilution.
Proteinuria was evaluated using Nephrosticks L (Bayer Medical, Tokyo, Japan). Kidneys were harvested from NZB/W and CD19−/− NZB/W mice, and bisected. The specimens were either fixed in 4% formalin for routine histology with hematoxylin and eosin (H&E) and Periodic Acid Schiff (PAS) staining or flash-frozen in OTC compound (Sakura Fineteck, Torrance, CA) for the detection of glomerular immune-complex deposits. The H&E and PAS stained sections were scored for interstitial and glomerular disease, as described previously (43), in a blinded manner. Cryostat-cut tissue sections from frozen samples were fixed in acetone for 5 min, and were incubated with 10% normal rabbit serum in PBS (10 min, 37°C) to block nonspecific staining. The tissue sections were incubated sequentially (20 min, 37°C) with predetermined optimal concentration of FITC-conjugated goat anti-mouse IgG (H+L) Ab (ICN Biomedicals, Inc.). The stained sections were read on a fluorescent microscope at 400x, and images were captured with a constant exposure time of 0.5 s. Mean fluorescence was calculated from captured images; three representative glomeruli per mouse were outlined, and mean pixel intensity was calculated with Adobe Photoshop®.
B cells were purified from single cell splenocyte suspensions by removing T cells with anti-Thy-1.2 Ab-coated magnetic beads (Dynal, Lake Success, NY). B cell suspensions were always >95% B220+, as determined by flow cytometry analysis. B cells were resuspended (2×107/ml) in RPMI 1640 medium containing 5% FCS at 37°C. The cells were stimulated with goat anti-mouse IgM Ab F(ab’)2 fragments (40 μg/ml; Cappel), and subsequently lysed in buffer containing 1% NP-40, 150 mM NaCl, 50 mM Tris-HCl (pH 8.0), 1 mM Na orthovanadate, 2 mM EDTA, 50 mM NaF, and protease inhibitors. Protein concentrations were determined by light absorbance at 280 nm. The obtained lysates were subjected to SDS-PAGE with subsequent electrophoretic transfer to nitrocellulose membranes. These membranes were incubated with anti-phospho Akt Ab (Ser473; Cell Signaling, Beverly, MA), anti-active ERK Ab (Promega, Madison, WI), or anti-active JNK Ab (Promega), followed by incubation with HRP-conjugated donkey anti-rabbit IgG Abs (Jackson ImmunoResearch Laboratories, West Grove, PA). These blots were developed using an enhanced chemiluminescence kit (Pierce, Rockford, IL). To verify the presence of equivalent amounts of protein in each lane, the blots were stripped and reprobed with anti-ERK2 Ab (Santa Cruz Biotechnology).
After spleen B cells had been stimulated with goat anti-mouse IgM F(ab’)2 Ab and lysed as described above, the lysates were analyzed using ProFluor® Src-Family Kinase Assays (Promega) according to manufacturer’s protocol. The lysates were mixed with Src-family kinase R110 substrate, with ATP added to initiate the kinase reaction. After incubating the plate at room temperature for 60 min, protease solution was added to each well and incubated for 60 min at room temperature. After terminating the protease reaction, the fluorescence of the liberated R110 was read at a wavelength of 525 nm. The fluorescence of each well inversely relates to kinase activity within the cell lysate. The kinase activity of wild type B cells stimulated for 3 min was defined as 100%.
Spleen cells (1×107/ml) in RPMI 1640 medium containing 5% BSA and 10 mM HEPES buffer were loaded with 1 μM Fluo-4 (Molecular Probes, Eugene, OR) at 37°C for 30 min. The cells were washed and stained with PE-Cy5-conjugated anti-B220 Ab for 20 min on ice and washed. The fluorescence ratio (525/405 nm) of B220+ cells was determined using an Epics Altra flow cytometer (Beckman Coulter, Miami, FL) with fluorescence intensity shown on a 4-decade log scale. Fluorescence contours are shown as 50% log density plots. Positive and negative populations of cells were determined using unreactive isotype-matched Abs (SouthernBiotech, Birmingham, AL) as controls for background staining. Baseline fluorescence ratios were collected in real time for 1 min before goat anti-mouse IgM F(ab’)2 Ab fragments (Cappel) were added. The results were plotted as fluorescence ratios at 10-second intervals, with increasing fluorescence ratios indicating increased [Ca2+]i.
Eight-week-old mice were immunized i.p. with 100 μg of 2,4-dinitrophenylated keyhole limpet hemocyanin (DNP-KLH, LSL, Tokyo, Japan) in complete Freund’s adjuvant and were boosted 21 days later with 100 μg of DNP-KLH in incomplete Freund’s adjuvant. The mice were bled before and after immunizations. Serum DNP-specific Ab titers were measured by adding diluted sera to ELISA plates coated with DNP-BSA (5 μg/ml) for an hour at room temperature. After washing the plates five times, bound Abs was detected using HRP-conjugated goat anti-mouse IgM or anti-mouse IgG1 Ab (SouthernBiotech) at predetermined optimal concentrations. The ELISA plates were developed using TMB substrate (Bethyl Laboratories, Inc.), stopped with 1N H2SO4 and read at 450 nm wavelength.
The following mAbs were used: FITC-, PE-, and PE-Cy5- conjugated anti-mouse B220 (CD45R, RA3-6B2; BD PharMingen, San Diego, CA), FITC-conjugated CD19 (MB19-1; BD PharMingen), FITC-conjugated CD1d (1B1; BD PharMingen), PE-Cy5-conjugated CD4 (H129.19; BD PharMingen), PE-conjugated CD5 (53-7.3; BD PharMingen), and FITC-conjugated anti-Thy1.2 (30-H12; BD PharMingen) mAbs.
Single cell spleen suspensions were stained for two/three-color immunofluorescence analysis at 4°C using Abs at predetermined optimal concentrations for 20 min as described (14). Cell numbers were counted using a hemocytometer, with relative lymphocyte percentages among viable cells (based on scatter properties) determined by flow cytometry analysis. Erythrocytes were lysed after staining using FACS™ Lysing Solution (BD Biosciences, San Jose, CA). A PE-conjugated anti-mouse/rat/human FOXP3 Flow Kit (clone 150D; BioLegend, San Diego, CA) was used to detect intracellular Foxp3 expression by regulatory T (Treg) cells according to the manufacturer’s protocol. The labeled cells were analyzed on an Epics Altra flow cytometer (Beckman Coulter) with fluorescence intensity shown on a 4-decade log scale. Positive and negative populations of cells were identified using unreactive isotype-matched Abs (SouthernBiotech) as controls for background staining.
Spleen B cells and T cells were purified with B220 mAb- and Thy1.2 mAb-coated microbeads (Miltenyi Biotech, Auburn, CA) by positive selection following the manufacturer’s instructions. In addition, CD1dhiCD5+ B cells were isolated from purified B cell preparations using an Epics Altra flow cytometer (Beckman Coulter) with purities of 85–95%. These cells were homogenized in Isogen S (Wako, Tokyo, Japan), with total RNA isolated following the instructions of the user’s manual. Total RNA was reverse transcribed to cDNA using a Reverse Transcription System with random hexamers (Promega). Quantitative RT-PCR was performed using the TaqMan® system (Applied Biosystems, Foster City, CA) with analysis using an ABI Prism 7000 Sequence Detector (Applied Biosystems) according to the manufacturer’s instructions. TaqMan® probes and the primers for IL-10 and GAPDH were purchased from Applied Biosystems. Relative expression of the real-time PCR products was determined using the ΔΔCT technique. B cells from 28-wk-old wild type NZB/W mice were used as the calibrator. Briefly, each set of samples was normalized using the difference in threshold cycle (CT) between the target gene and housekeeping gene (GAPDH): ΔCT = (CT target gene – CT GAPDH). Relative mRNA levels were calculated by the expression 2−ΔΔCT, where ΔΔCT=ΔCT sample − ΔCT calibrator. Each reaction was performed in triplicate at least.
Serum IL-10 levels were measured using mouse IL-10 ELISA kits (Invitrogen, Carlsbad, CA) according to the manufacturer’s protocol. Briefly, diluted sera were added to a 96-well plate pre-coated with anti-mouse IL-10 Abs, incubated for 90 min at 37°C, and washed with buffer four times. Following the addition of biotin-conjugated anti-IL-10 Ab, the plate was incubated for 45 min at 37°C. The plates were then washed and incubated with HRP-conjugated streptavidin for 45 min at 37°C. The ELISA was developed using stabilized chromogen, terminated with stop solution, and read at a wavelength of 450 nm.
IL-10 present in tissue culture supernatant fluid was also quantified using the same assays. Splenic B cells from wild type NZB/W mice were purified with B220 mAb-coupled microbeads (Miltenyi Biotech). CD1dhiCD5+ B cells were isolated using an Epics Altra flow cytometer (Beckman Coulter). Isolated CD1dhiCD5+ B cells as well as CD5− B cells (3×105) were cultured in 200 μl of RPMI 1640 medium containing 10% FBS, 10 mM HEPES, 55 μM 2-ME, 200 μg/ml penicillin, and 200 U/ml streptomycin (all from Gibco, Carlsbad, CA) in 96-well flat-bottom tissue culture plates at 37°C with 5% CO2 in the presence of LPS (10 μg/ml, Escherichia coli serotype 0111: B4, Sigma-Aldrich, St. Louis, MO). After culture for 72 hr, IL-10 concentrations in culture supernatant fluid were quantified. All assays were carried out using triplicate samples.
Splenic B cells were purified using B220 mAb-coupled microbeads (Miltenyi Biotech) from 20-wk-old wild type NZB/W mice. Spleen CD1dhiCD5+ B cells were isolated using an Epics Altra flow cytometer (Beckman Coulter) with purities of 85–95%. After isolation, 2×106 CD1dhiCD5+B220+ or CD5−B220+ B cells were transferred intravenously into 20-wk-old CD19−/− NZB/W mice (n=15), with nephritis and survival monitored. In some mice, spleen Treg cell numbers were assessed at 24 wk of age.
ANA, proteinuria, and survival data were analyzed using Kaplan-Meier curves and the log rank test. Unless indicated otherwise, comparisons between groups were made using the Mann-Whitney U-test. A p value <0.05 was considered to be statistically significant.
The age of ANA production in wild type and CD19−/− NZB/W mice was compared using a fluorescent ANA assay with HEp-2 cells as substrates. Serum ANA was first detected in NZB/W mice between 16 to 24 wk of age. However, the appearance of serum ANA was significantly delayed in CD19−/− NZB/W mice (p<0.001, Fig. 1A). ANA titers were also significantly lower in CD19−/− NZB/W mice than in wild type NZB/W mice at all the ages examined (Fig 1B). ANA staining had a homogenous to speckled nuclei staining pattern, with no difference observed between wild type and CD19−/− mouse sera (data not shown). The development of autoantibodies to dsDNA was also delayed in CD19−/− NZB/W mice as determined by ELISA (Fig. 1C). While most CD19−/− NZB/W mice eventually produced anti-dsDNA autoantibodies, their mean serum titers were significantly lower than those of wild type NZB/W mice after 20 wk of age. Consequently, CD19 expression positively regulates autoantibody production in NZB/W mice.
To assess renal disease in NZB/W mice, the relationship between proteinuria and IgG deposition in the basement membranes of glomeruli were first investigated. Protein levels >300 mg/dl in urine correlated with histological nephritis based on H&E and PAS staining in both wild type and CD19−/− NZB/W mice (Fig. 2A). Therefore, urinary protein excretion >300 mg/dl was defined as proteinuria onset. Proteinuria was monitored every 2 wk in wild type and CD19−/− NZB/W mice. Proteinuria developed slightly but significantly earlier in CD19−/− NZB/W mice than wild type NZB/W mice (p<0.05 from 23 to 32 wk; Fig. 2C). Pathological examination of the kidneys from 32-wk-old mice revealed that glomerulonephritis and interstitial nephritis developed both in wild type and CD19−/− NZB/W mice. Glomerulonephritis and interstitial nephritis tended to be even more severe in CD19−/− NZB/W mice than wild type mice (Fig. 2B, D). The deposition of IgG in the basement membrane of glomeruli was also observed in both wild type and CD19−/− mice. The fluorescence intensity of glomerular IgG staining was also slightly higher in CD19−/− mice, although the difference was not statistically significant (Fig. 2B, D). Glomerular IgG deposition was even detected in the kidneys of CD19−/− NZB/W mice that had been found to be ANA-negative (data not shown). Therefore, CD19−/− NZB/W mice developed glomerulonephritis earlier than wild type NZB/W mice, despite their low frequency and titers of anti-dsDNA Abs (Fig. 1).
Wild type NZB/W mice begin to succumb to disease at ~25 wk of age (Fig. 2C), following the development of nephritis. By contrast, CD19−/− NZB/W mice begin to succumb to disease at ~20 wk of age, consistent with their accelerated proteinuria development. Median survival in CD19−/− NZB/W mice was significantly shorter in comparison with wild type NZB/W mice (30 wk v.s. 35 wk; p<0.05). Death in CD19−/− NZB/W mice followed the development of nephritis, although some mice did not have detectable ANA or minimal anti-dsDNA Abs (data not shown). Collectively, CD19 expression negatively regulates the development of renal disease, which accelerated mortality.
Since CD19 deficiency generally leads to an “immunodeficient” B cell phenotype in both mice and humans (20, 21, 44), the finding that CD19 deficiency accelerated disease progression in NZB/W mice was paradoxical. Therefore, it was determined whether strain differences might result in an unanticipated phenotype for B cells from CD19−/− NZB/W mice. Cell-surface CD19 expression on B cells from the blood, spleen, and lymph nodes was identical between C57BL/6 and NZB/W mice (data not shown). In functional studies, IgM ligation generated augmented intracellular calcium responses by splenic B cells from NZB/W mice relative to C57BL/6 mice (Fig. 3A), consistent with previous reports of polyclonal B cell activation in NZB/W mice (45–48). When wild type and CD19−/− NZB/W B cells were compared, IgM-induced intracellular calcium responses were delayed in CD19−/− B cells (Fig. 3B), which is consistent with results obtained with CD19−/− B cells from C57BL/6 mice (49). IgM-induced Src-family kinase activation and Akt phosphorylation were also significantly reduced in CD19−/− NZB/W B cells in comparison with B cells from wild type NZB/W mice (Fig. 3C, D), as previously reported for CD19−/− B cells from C57BL/6 x 129 mice (14, 50). Impaired ERK and JNK activation were also observed in CD19−/− NZB/W B cells in comparison with wild type NZB/W B cells (Fig. 3D). The proliferation of B cells cultured in the presence of F(ab’)2 anti-IgM Abs was also reduced by CD19-deficiency in NZB/W mice (data not shown). In vivo, the influence of CD19-deficiency on humoral immune responses in NZB/W mice was assessed by immunizing mice with DNP-KLH, a T cell-dependent Ag. Following immunizations, the primary and secondary IgM and IgG1 responses in CD19−/− NZB/W mice were significantly lower than in wild type NZB/W mice (Fig. 3E). Thus, CD19-deficiency in NZB/W mice results in B cell defects that are identical to those reported for CD19−/− mice on nonautoimmune backgrounds. This explains the impaired autoantibody production in CD19−/− NZB/W mice, but not the dissociated acceleration of nephritis progression.
CD19 expression is critical for regulatory B10 cell development in C57BL/6 mice (34, 40, 41, 51). Therefore, the development of the spleen CD1dhiCD5+ B cell subset, which includes B10 cells, was assessed in NZB/W mice. A spleen CD1dhiCD5+B220+ B cell subset was identified in NZB/W mice, that was increased in 28-wk-old wild type NZB/W mice when compared with 12-wk-old mice (0.9±0.2% at 12 wk and 2.3±0.5% of B220+ cells at 28 wk). By contrast, splenic CD1dhiCD5+ B cells were virtually absent in CD19−/− NZB/W mice at both 12- and 28-wks of age (0.07±0.03% at 12 wk and 0.13±0.03% at 28 wk, p<0.05 v.s. wild type mice at each equivalent age; Fig. 4A and Table I). CD19−/− NZB/W mice also had reduced numbers of splenic marginal zone B cells with a CD23loCD21hi phenotype as well as reduced numbers of peritoneal B1 cells with a B220intCD5+ phenotype (Table I), both of which increase with age in NZB/W mice (29).
Since IL-10 production is the hallmark of B10 cells, IL-10 secretion by CD1dhiCD5+ B cells was investigated. At 12 wk of age, IL-10 mRNA expression in splenic B cells was comparable between wild type and CD19−/− NZB/W mice. IL-10 mRNA levels of splenic B cells from wild type NZB/W mice were increased by 2.5-fold at 28 wk of age in comparison with those at 12 wk (Fig. 4B, left). IL-10 mRNA levels in splenic B cells from CD19−/− NZB/W mice remained unaltered at 28 wk of age. While B10 cells are not only IL-10 secreting B cells in the spleen, increased numbers and enhanced activation of B10 cells may at least partially contribute to the increase of IL-10 expression in splenic B cells from wild type mice since CD1dhiCD5+ B cells from wild type NZB/W mice produced augmented IL-10 levels at 28 wk compared with CD1dintCD5− B cells (p<0.05; Fig. 4B, right). IL-10 secretion from B cells was 11.2-fold higher in wild type mice by 12 wk of age and 11.4-fold at 28 wk of age, respectively, than in CD19−/− mice (p<0.05 for each; Fig. 4C, left). When splenic B cells from wild type NZB/W mice were separated into CD1dhiCD5+ cells and non-CD1dhiCD5+ cells, CD1dhiCD5+ cells secreted 4- to 5-fold more IL-10 in comparison with non-CD1dhiCD5+ B cells (p<0.05; Fig. 4C, right). Thus, CD1dhiCD5+ B cells were increased in number and produced significant levels of IL-10 during disease, while these cells were severely reduced in CD19−/− NZB/W mice at all time points. Additionally, serum IL-10 concentrations increased during disease progression in wild type NZB/W mice, but remained significantly lower in CD19−/− NZB/W mice (p<0.01 at 12, 20, and 28 wk; Fig. 4D). Thereby, modest IL-10 production and the absence of CD1dhiCD5+ B10 cells offered an explanation for accelerated disease in CD19−/− NZB/W mice.
To determine whether the absence of regulatory B10 cells in CD19−/− NZB/W mice explains their accelerated disease progression, CD1dhiCD5+B220+ B cells from 20-wk-old wild type NZB/W mice were transferred into CD19−/− NZB/W mice of the same age. As a control, spleen CD5−B220+ follicular B cells were also transferred into CD19−/− NZB/W mice. The transfer of wild type CD1dhiCD5+B220+ B cells into CD19−/− NZB/W mice normalized nephritis onset (p<0.05 at 23 wk, Fig. 5A) and prolonged survival until 35 wk of age (p<0.05, Fig. 5B) to the extent seen in wild type NZB/W mice. In fact, CD19−/− NZB/W mice that received wild type CD1dhiCD5+ B cells lived even longer than wild type NZB/W mice (median survival, 37 wk v.s. 35 wk). Nephritis and survival were not significantly altered in CD19−/− NZB/W mice that received wild type CD5−B220+ cells. Also, the adopted transfer of CD1dintCD5+ B cells did not improve nephritis or survival significantly (data not shown). Thereby, spleen regulatory CD1dhiCD5+ B cells can inhibit lupus progression when transferred into CD19−/− NZB/W mice, demonstrating that this subset normally inhibits disease initiation in NZB/W mice.
CD4+Foxp3+ Treg cell numbers increase in wild type NZB/W mice during disease (Fig. 6A) as published (52). The CD4+Foxp3+ Treg cell subset made up 2.4±0.7% of splenic Thy1.2+ T cells (2.4±0.8 x106 cells) in 12-wk-old wild type NZB/W mice. Treg cell frequencies increased to 8.5±1.7% (17.0±1.9 x106 cells) in wild type NZB/W mice that had developed ANA and proteinuria at 28 wk of age. While spleen Treg cell frequencies were comparable between wild type and CD19−/− NZB/W mice at 12 wk of age (2.1±0.8%; 1.8±0.8 x106 cells in CD19−/− mice), there was not a significant increase in the Treg subset in CD19−/− mice at 28 wk of age (2.8±0.9%; 2.8±0.7 x106 cells; p<0.05 v.s. wild type mice at 28 wk). IL-10 production is also a characteristic of Treg cells (53). While T cells from 28-wk-old wild type NZB/W mice expressed ~10-fold higher IL-10 mRNA levels than 12-wk–old mice, IL-10 production from splenic T cells was below detectable levels in CD19−/− NZB/W mice at 28 wk of age (p<0.05 v.s. wild type mice; Fig. 6B). Thereby, Treg cell numbers and IL-10 production by splenic T cells was decreased in CD19−/− NZB/W mice. Since CD19 expression is restricted to the B cell lineage, these results suggest that B cells can influence Treg cell development and/or activation.
To determine whether the absence of CD1dhiCD5+ B cells in CD19−/− NZB/W mice influences Treg cell expansion, CD1dhiCD5+ B cells from 20-wk-old wild type NZB/W mice were transferred into CD19−/− NZB/W mice of the same age. Four weeks after cell transfers, the numbers and percentages of Treg cells in CD19−/− NZB/W mice given wild type CD1dhiCD5+ B cells were significantly higher than in age-matched CD19−/− NZB/W mice (5.0±1.1 x106 cells v.s. 1.9±0.3 x106 cells, p<0.05; 6.1±.1.2% v.s. 2.3±0.2% of Th1.2+ T cells, p<0.05; Fig. 6C). Thereby, B10 cells or other CD1dhiCD5+ regulatory B cells are likely to play a critical role in Treg cell expansion in NZB/W mice.
Nephritis and death were accelerated in CD19−/− NZB/W mice relative to wild type NZB/W mice (Fig. 2) despite B cell hyporesponsiveness and their “immunodeficient” phenotype (Fig. 3) of CD19−/− mice (20, 21). These unexpected findings were due to the virtual absence of B10 cells in CD19−/− NZB/W mice (Fig. 4) as described previously for C57BL/6 CD19−/− mice (34, 51). This was confirmed by the adoptive transfer of splenic CD1dhiCD5+ B cells from wild type NZB/W mice into CD19−/− NZB/W mice, which significantly prolonged their survival and demonstrated an important protective role for regulatory B10 cells in this systemic autoimmune disease. Consistent with these observations, B cell depletion by CD20 mAb treatment eliminated 99% of B10 cells and accelerated disease development in young NZB/W mice as demonstrated in the companion paper to these studies (Haas, K. M., R. Watanabe, T. Matsushita, H. Nakashima, N. Ishiura, H. Okochi, M. Fujimoto, and T. F. Tedder. Protective and pathogenic roles for B cells during systemic autoimmunity in NZB/W F1 mice. Submitted). These studies thereby demonstrate protective roles for B cells in lupus pathogenesis.
CD19 expression had both protective and disease promoting roles in lupus pathogenesis in NZB/W mice. CD19-deficiency significantly delayed the generation of ANA, especially anti-dsDNA Abs, in this lupus-prone mouse strain (Fig. 1). Unexpectedly, however, the manifestation of nephritis was paradoxically accelerated by the loss of CD19, although the difference was rather modest (Fig. 2). This result paralleled enhanced mortality in CD19−/− NZB/W mice. This discrepancy mirrors the findings of Mohan and colleagues in transgenic mice that overexpress CD19 and expressed the Sle1 lupus susceptibility locus (54). In this case, CD19 overexpression augmented humoral autoimmunity, but did not accelerate mortality or clinical evidence of renal dysfunction. Consistent with this, B cells from these CD19-transgenic mice are hyper-responsive to transmembrane signals, but have significantly increased B10 cell numbers (21, 34, 51). Thereby, CD19 expression positively correlates with autoantibody production, but is likely to have opposing roles during autoimmune disease by regulating B10 cell development. That severe glomerulonephritis can occur in the absence of ANA, including anti-DNA Abs, and that autoreactive B cells can exert pathogenic effects independent of Ab secretion has also been demonstrated in other lupus-prone mouse strains (55–57). Thus, the severe renal disease observed in CD19−/− NZB/W mice is likely to result from B cell functions other than autoantibody secretion. These studies demonstrate that this B cell function is attributable in part to the “suppressive” role of B10 cells that normally negatively regulate disease progression.
IL-10 is a pleiotropic cytokine with both immunosuppressive and immunostimulatory properties (53, 58). The role of IL-10 in lupus pathogenesis is complex, including the effects of high serum IL-10 levels in human SLE patients and lupus-prone mouse strains (59–63). For example, serum IL-10 levels positively correlate with SLE disease activity scores and anti-dsDNA autoantibody titers, but negatively correlate with C3 and C4 levels and lymphocyte counts (59, 64, 65). SLE patients also have significantly more IL-10 secreting mononuclear cells in their peripheral blood than normal controls, and disease severity correlates with increased numbers of circulating IL-10-secreting mononuclear cells (61). Furthermore, IL-10 production by B cells is higher for SLE patients than normal controls, and Ig production by SLE B cells is largely IL-10 dependent (60). Thereby, IL-10 can be pathogenic for lupus acceleration, but may also be produced to reduce already existing autoimmune inflammation. Various treatments targeting IL-10 against SLE have also shown contradictory results. For instance, IL-10 deficiency significantly enhances disease severity in MRL/lpr mice with increases in IFN-γ and IgG2a anti-dsDNA autoantibody production, which are suppressed by recombinant IL-10 treatment (66). In the current study, CD19 deficiency led to lower serum IL-10 levels in NZB/W mice throughout the disease course (Fig. 4D). By contrast, continuous anti-IL-10 mAb administration significantly delays disease development in NZB/W mice, which is attributed to increased TNF-α production (67). These contradictory findings are most likely explained by the fact that multiple cell types are capable of producing IL-10, including B cells. Thereby, the positive and negative regulatory roles of IL-10 are likely to differ depending on the cell source of IL-10, as well as the timing of its production, duration, and levels of IL-10 expression. Thus, B10 cell IL-10 production is but one component of a complex regulatory network that balances protective and pathogenic immune responses.
In addition to B10 cells and Ig secretion, B cells regulate immune responses through multiple mechanisms that have only recently been appreciated (68). B cells contribute to Ag-presentation, cytokine production, the regulation of lymphoid organogenesis, effector T cell differentiation, and dendritic cell function. It is also noteworthy that B cells have other critical roles in lupus, presumably through their interaction with T cells. For instance, B cell deficiency in MRL/lpr mice results in the complete absence of inflammatory T cell renal infiltration (69). B cell ablation in MRL/lpr mice using CD79 mAb decreases the relative abundance of CD4 memory T cells, and also reduces T cell infiltration into the kidneys (70). By contrast, MRL/lpr mice engineered to have B cells expressing surface-bound but not secretory Ig develop nephritis which is characterized by renal T cell infiltration (55). Thus, B cells play pathogenic roles via cytokine secretion or Ag presentation (71). Since lupus develops under the complex regulation of different B cell subsets and their functions, the selective targeting of B cell subsets may lead to promising therapies for this and other autoimmune disorders.
While the adoptive transfer of CD1dhiCD5+ B cells into CD19−/− NZB/W mice significantly improved survival, this treatment did not “cure” underlying disease (Fig. 5). Since CD19-positive transferred cells were detected in the spleen of CD19−/− NZB/W mice two weeks after injection, but not in five weeks (data not shown), this may be partly explained by the eventual rejection of CD19-expressing wild type B10 cells in CD19-deficient mice. However, this most likely reflects the complex etiology of the lupus-like diseases, and the involvement of multiple hematopoietic lineages in disease initiation and regulation. As an example of this, splenic T cell IL-10 mRNA levels were significantly reduced during the late stages of disease in CD19−/− NZB/W mice (Fig. 6B). The spleen CD4+Foxp3+ Treg cell subset was also significantly reduced in CD19−/− NZB/W mice, while Treg cells expanded during disease progression in wild type NZB/W mice (Fig. 6A). Consistent with this, the adoptive transfer of CD1dhiCD5+ B cells from wild type NZB/W mice significantly increased Treg cell numbers in CD19−/− NZB/W mice (Fig. 6C). These results indicate that CD19 expression by B cells or the presence or absence of B10 cells also has a significant influence on Treg cell development and/or activation in NZB/W mice that remains to be explored. Thus, effective treatments or a cure for lupus-like disease is likely to require the modulation of not only B cell and B10 cell functions, but also T cell and Treg cell functions that significantly modulate disease.
This work was supported by the Grant-in-Aid from the Ministry of Education, Science, and Culture of Japan (to RW and MF), and grants from the National Institutes of Health, USA (AI56363, CA105001, and CA96547 to TFT).