B Lymphocyte Depletion by CD20 Monoclonal Antibody Prevents Diabetes in Nonobese Diabetic Mice despite Isotype-Specific Differences in FcγR Effector Functions

Mice

Wild-type C57BL/6J, NOD/Lt, and NOD/LtSz-scid/scid (NOD.scid) mice were obtained from The Jackson Laboratory. FcγRI/III^−/− C57BL/6 mice were generated in our laboratory by crossing FcγRI^−/− and FcγRIII^−/− mice (25). All mice were housed in a specific pathogen-free barrier facility. These studies were approved by the Animal Care and Use Committee of Duke University (Durham, NC) or University of North Carolina (Chapel Hill, NC).

Antibodies

Mouse CD20-specific mouse mAbs MB20-11 (IgG2c), MB20-18 (IgG2b), and MB20-1 (IgG1) were as described (23). IgG2a and IgG1 isotype switch variants of the MB20-18 mAb were produced by transient cotransfection of 293T cells with expression plasmids containing engineered H chain and native L chain cDNAs. Abs were subsequently purified from culture supernatants using protein A or G affinity chromatography columns as suggested by the manufacturer (Amersham). The 9G8.1 hamster anti-mouse FcγRIV mAb (provided by J. Ravetch, Rockefeller University, New York, NY) was as described (34). FITC-conjugated mAb reactive with L-selectin (CD62L; clone LAM1-116) was as described (35). Other mAbs used included: B220 mAb (RA3-6B2), CD1d (1B1), CD5 (53-7.3), CD11b (M1/70), CD16/32 (2.4G2), CD21 (7G6), CD24 (M1/69), CD25 (PC61), CD28 (37.51), CD40L (MR1), CD44 (IM7), CD69 (H1.2F3), and OX40 (OX-86) mAbs (BD Pharmingen). CD3 (145-2C11), CD4 (L3T4), CD8α (53-6.7), and CD19 (1D3) mAbs were obtained from eBioscience. Intracellular staining used mAbs reactive with CTLA-4 (UC10-4F10-11; BD Pharmingen) and Foxp3 (FJK-16s; eBioscience) and the Cytofix/Cytoperm kit (BD Pharmingen). Isotype-specific anti-mouse secondary Abs were obtained from Southern Biotechnology Associates. Background staining was assessed using nonreactive, isotype-matched control mAbs (Caltag Laboratories).

Cell isolation and immunofluorescence analysis

Single-cell suspensions of bone marrow (bilateral femurs), spleen, and peripheral lymph node (paired axillary and inguinal) lymphocytes were generated by gentle dissection. To isolate peritoneal cavity leukocytes, 10 ml of cold 0.2% BSA in PBS buffer was injected into the peritoneum followed by gentle massage of the abdomen. For two- or three-color immunofluorescence analysis, single-cell suspensions (1 × 10⁶ cells) were stained at 4°C using predetermined optimal concentrations of mAb for 20 min as described (36). Blood erythrocytes were lysed after immunofluorescence staining using FACS lysing solution (BD Biosciences). Cells with the forward and side light scatter properties of lymphocytes were analyzed using a FACScan flow cytometer (BD Biosciences).

Islets were purified as described (37). Briefly, pancreata from NOD mice were perfused with 2 mg/ml collagenase P (Roche) and digested for 20 min at 37°C. Islets were purified from digested tissues using a Ficoll gradient and then handpicked. Purified islets were dissociated into a single-cell suspension using enzyme-free cell dissociation solution (Sigma-Aldrich), washed, and cultured overnight in RPMI 1640 complete medium (Mediatech) containing 10% FBS (Sigma-Aldrich), 2 mM glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin (Mediatech), 55 μM 2-ME (Invitrogen Life Technologies), and 4 ng/ml recombinant murine IL-2 (PeproTech).

For FcγRIII and FcγRIV expression analysis, macrophages were elicited by i.p. injection of 1 ml of 3% w/v thioglycolate (Sigma-Aldrich). After 4 days, peritoneal cavity cells were collected by lavage or isolated from single-cell splenocyte populations (34).

Immunotherapy

Sterile CD20 and isotype control mAbs (1–250 μg) in 300 μl of PBS were injected i.v. into the lateral tail vein. Blood leukocyte numbers were quantified by hemocytometer after red cell lysis, with blood and tissue B220⁺ cell frequencies determined by immunofluorescence staining with flow cytometry analysis as described (23, 25). In some cases, 1 ml of 3% thioglycolate solution was injected i.p. into mice 1 day before mAb treatment. For B cell depletion analysis in NOD mice, 5-wk-old female NOD mice were treated with MB20-11 mAb (250 μg) three times at 2-wk intervals. For long-term disease incidence experiments, 5- or 15-wk-old female NOD mice were treated with three MB20-11 mAb (250 μg) injections at 2-wk intervals.

Adoptive transfer experiments

Unfractionated splenocyte suspensions from NOD and C57BL/6 mice were labeled with 0.1 and 1.0 μM Vybrant CFSE, respectively, according to the manufacturer’s instructions (Invitrogen Life Technologies). The relative frequency of B220⁺ cells among CFSE-labeled splenocytes was determined by immunofluorescence staining with flow cytometry analysis. Subsequently, equal numbers of CFSE-labeled B220⁺ C57BL/6 and NOD splenocytes (4 × 10⁷ total) were injected i.v. into NOD and C57BL/6 recipients 1 day before i.v. injection of either MB20-11 or control mAb (250 μg/mouse). After 1 day, spleen and lymph node cells were harvested with numbers of CFSE-labeled B220⁺ cells determined by immunofluorescence staining with flow cytometry analysis.

FcγR sequencing and constructs

Total RNA was isolated from spleens using the RNeasy kit (Qiagen). Random hexamer primers (Promega) and SuperScript II RNase H-Reverse Transcriptase (Invitrogen Life Technologies) were used to generate cDNA as described (38). Primers were as follows: for FcγRI, forward 5′-aaggaattcg acattacatg attcttacca gctttggag-3′, C57BL/6, reverse 5′-attgtcgaca tccctttggt ataggtctcc tttcaaatg-3′, NOD, reverse 5′-attgtcgacc tcagaaacca aagggacttc taagttgta-3′; for FcγRIII, forward 5′-atagaattcg acacccagat gtttcagaat gcacactct-3′, reverse 5′-atagtcgacg atggatgggg tgtcacttgt cttgaggag-3′; for FcγRIV, forward 5′-atagaattca tctgcttcag cagcatgtgg cagctacta-3′, reverse 5′-atagtcgacc atgcgataag agctcacttg tcctgaggt-3′; for mouse FcγRI common γ-chain, (FcRγ), forward 5′-aatagatcta gcgcacccag gatgatctca gccgtgat-3′, reverse 5′-aagggatccg tctttaacgg agatggggac ctgccagtg-3′. Full-length FcγRI, FcγRIII, FcγRIV, and FcRγ cDNAs were subsequently cloned into the pIRES2-EGFP vector (BD Clontech) and sequenced completely in both 5′ and 3′ directions using the Applied Biosystems Dye Terminator Cycle Sequencing System. Three independent FcγRIV cDNA clones were sequenced. Sequences for NOD FcγRI were as published (39). C57BL/6 FcγRI sequences were as published (GenBank accession number AK033874). NOD and C57BL/6 FcγRIII sequences were as published (GenBank accession numbers AY897426 and AY897419). Our NOD FcγRIV sequences were deposited into GenBank (EU050648). C57BL/6 FcγRIV sequences were as published (40).

FcγR functional assays

COS cells were cultured in DMEM (Mediatech) supplemented with 10% FBS, 2 mM glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin. COS cells at 90% confluency were transiently cotransfected with FcγR (FcγRI, FcγRIII, or FcγRIV), and FcRγ expression plasmids by electroporation. Plasmid DNA (40 μg) was electroporated into 5 × 10⁶ cells in cultured medium using a Bio-Rad Genepulser at 250 V, 960 millifarad. FcγR transcript levels were assessed by visualizing GFP expression by the transfected cells. Transfected COS cells were transferred into 24-well plates 24 h later at 1 × 10⁵/well and cultured overnight. A total of 2 × 10⁷ A20 cells (mouse B lymphoma cell line) were preincubated in 200 μl of culture medium containing appropriate mAbs (20 μg/ml) for 30 min on ice. After washing three times with culture medium, mAb-coated A20 cells were added (10⁶ in 200 μl) onto the COS cell monolayers in 24-well plates with horizontal rotation at 4°C. After 60 min, unbound A20 cells were removed by extensive washing of the cell monolayers with culture medium before fixation using PBS containing 2.5% formaldehyde. A20 cell adhesion was evaluated by phase-contrast microscopy, with the number of bound A20 cells quantified for ≥10 random microscopic fields in each of three culture wells.

Calcium response measurements

Four days after thioglycolate injection, macrophages were purified from the peritoneal cavity by depleting CD19⁺ cells using MACS system following the manufacturer’s instructions (Miltenyi Biotec). Cells were resuspended (1 × 10⁷ cells/ml) in RPMI 1640 medium (Sigma-Aldrich) containing 5% FBS and 10 mM HEPES buffer (Invitrogen Life Technologies). The cells were loaded with 1 μM Indo-1 AM ester (Molecular Probes) for 30 min at 37°C, labeled for an additional 20 min with FITC-conjugated CD11b mAb, washed, and resuspended in warm tissue-culture medium (2 × 10⁶ cells/ml) for flow cytometry analysis. Baseline emission fluorescence ratios (405:525 nm) of CD11b⁺ gated cells were collected for 1 min before the addition of DNP-BSA/anti-DNP Ab immune complexes produced by incubating DNP-BSA (60 μg/ml; Biosearch Technologies) with rabbit anti-DNP IgG (90 μg/ml; Novus Biologicals) at 37°C for 2 h. Fluorescence ratios were plotted at 20-s intervals.

Diabetes and insulitis assessment

Diabetes was monitored weekly by measuring urine glucose levels with Diastix (Bayer). Mice were diagnosed as diabetic when urine glucose levels exceeded 0.25% for two successive measurements. Urine glucose levels of 0.25% were equivalent to blood glucose values of ≥250 mg/dl as determined by an Autokit Glucose CII assay (Wako Chemicals). Insulitis was assessed by histology as described (41). Briefly, pancreata were embedded and frozen in OCT and 5 μM tissue sections were stained using H&E. Insulitis severity was scored as no infiltration, peri-insulitis (islets surrounded by a few lymphocytes), or intrainsulitis (lymphocytic infiltration into the interior of islets) as described (41).

Adoptive transfer of diabetogenic splenocytes into NOD.scid mice

Splenocytes (1 × 10⁷) isolated from recent diabetic NOD female mice (diabetogenic splenocytes) were mixed with equal numbers of splenocytes from untreated, CD20, or control Ab-treated (250 μg/mouse) mice in 200 μl of PBS. Cell mixtures were injected i.v. into 6- to 8-wk-old NOD.scid recipients. Mice were monitored weekly for urine glucose.

In vitro T cell proliferation assays

After 14 days of mAb treatment, CFSE-labeled (0.1 μM) splenocytes or lymph node lymphocytes were plated in 24-well plates (2 × 10⁶ cells) in 2 ml of RPMI 1640 complete medium with soluble CD3 mAb (2 μg/ml; 145-2C11; BD Pharmingen) plus CD28 mAb (2 μg/ml; 37.51, BD Pharmingen). After 70 h of incubation, the cultured cells were stained with CD4 or CD8 mAb to identify CFSE-labeled CD4⁺ and CD8⁺ T cells by flow cytometry.

Statistical analysis

All data are shown as means ± SEM. The Student t test was used to determine the significance of differences between sample means. Disease incidence curves were compared using the log-rank test.

Defective B cell depletion in NOD mice after CD20 mAb treatment

In contrast with efficient B cell depletion in C57BL/6 mice using the MB20-11 (IgG2c) CD20 mAb, initial studies in NOD mice revealed less efficient B cell depletion (data not shown). Therefore, three CD20 mAbs representative of each IgG isotype, MB20-1 (IgG1), MB20-18 (IgG2b), and MB20-11, were assessed for their ability to deplete mature bone marrow, blood, and tissue B cells in NOD and C57BL/6 mice over a range of mAb concentrations (2–250 μg/mouse). In both mouse strains, the CD20 mAbs displayed a dose-dependent hierarchy of B cell depletion efficiencies over 7 days of treatment, while isotype- and dose-matched control mAbs were without effect (Fig. 1⇓A). At each mAb dose, the IgG2c MB20-11 mAb depleted B cells most efficiently, followed by the IgG1 CD20 mAb, with the IgG2b CD20 mAb being the least effective, as described (23, 25). Similar results were obtained for peritoneal cavity B2 cells and lymph node B cells (data not shown). As in C57BL/6 mice, bone marrow pre-B and immature (B220^low) B cells were not depleted efficiently by CD20 mAbs (Table I⇓) due to absent or low CD20 expression, as described (24, 25). MB20-11 mAb potency was not an individual characteristic of this mAb, but reflected the potency of IgG2a/c CD20 mAbs in general (23). The IgG2a and IgG2c isotypes (42) are indistinguishable in their FcγR specificities (25). Thereby, when the IgG2b MB20-18 mAb was engineered to contain the IgG2a H chain, it and the IgG2c MB20-11 mAb were able to deplete B cells equally over a range of mAb doses (Fig. 1⇓B). Thus, CD20 mAbs have characteristic depletion capabilities depending on their isotypes, with IgG2a/c mAbs having greater potency than IgG1 or IgG2b CD20 mAbs (23).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Reduced B cell depletion in NOD mice relative to C57BL/6 (B6) mice. A, Bone marrow (mature, B220^high), blood (B220⁺ cells/ml), and spleen (mature, CD24⁺CD21⁺B220⁺) B cell numbers were determined 7 days after MB20-1, MB20-18, MB20-11, or isotype-matched control (CTL) mAb treatment of NOD or C57BL/6 mice. B, Bone marrow (mature), blood, and spleen B cell numbers were determined 7 days after IgG2c MB20-11 (○) or IgG2a MB20-18 (•), or isotype-matched control (CTL) mAb treatment of C57BL/6 mice. A and B, Values (±SEM) represent the percentage of B cells present in CD20 mAb-treated mouse tissues (n = 3) relative to control mAb-treated littermates (n = 3) at each mAb dose evaluated. Significant differences between sample means are indicated (∗, p < 0.05; ∗∗, p < 0.01).

Although each CD20 mAb-depleted significant numbers of B cells in C57BL/6 mice, each mAb was significantly less efficient in NOD mice (Fig. 1⇑A). The MB20-11 mAb given at 10 μg/mouse depleted ∼98% of circulating and spleen B cells in C57BL/6 mice by day 7. By contrast, NOD mice required a 25-fold higher dose of mAb (250 μg/mouse) to remove only 90% of blood and 70% of spleen B cells. Likewise, the MB20-1 IgG1 and MB20-18 IgG2b mAbs depleted 80–90% of spleen B cells when used at 100 μg in C57BL/6 mice, whereas this mAb dose depleted only 30–60% of B cells in NOD mice. Thus, B cells in NOD mice were significantly more resistant to CD20 mAb treatment compared with C57BL/6 mice, and this deficiency was not limited to individual CD20 mAb isotypes.

Decreased B cell depletion in NOD mice is not intrinsic to B cells

Whether attenuated B cell depletion in NOD mice resulted from abnormal or low-density CD20 expression by unique B cell subsets was assessed. All blood and peripheral lymph node B220⁺ B cells expressed CD20, with comparable expression densities in NOD and C57BL/6 mice (Fig. 2⇓A). Splenic marginal zone B cells (CD21^highCD1d⁺B220⁺), T1 cells (CD21⁻CD24^highB220⁺), T2 cells (CD21^highCD24^highB220⁺), and mature B cells (CD21⁺CD24⁺B220⁺) expressed CD20 at similar higher densities in both NOD and C57BL/6 mice (Fig. 2⇓B). CD20 was also expressed at comparable developmental stages and at similar levels by bone marrow immature IgM⁺B220^low and mature IgM⁺B220^high B cells (Fig. 2⇓C). Peritoneal cavity B1a (CD5⁺CD11b⁺B220^low), B1b (CD5⁻CD11b⁺B220^low), and B2 (CD5⁻B220^high) cells in NOD and C57BL/6 mice also expressed CD20 similarly (Fig. 2⇓D). Thus, altered CD20 expression did not explain decreased B cell depletion in NOD mice.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Reduced B cell depletion in NOD mice is not intrinsic to B cells. A–D, CD20 expression by B cells in NOD mice assessed by immunofluorescence staining using MB20-18 (thin line for C57BL/6, thick line for NOD mice) or isotype control (dashed line) mAbs. Single-cell suspensions of lymphocytes isolated from (A) blood and peripheral lymph nodes (PLN), (B) spleen, (C) bone marrow, and (D) the peritoneal cavity were examined by four-color flow cytometry analysis. Spleen T1, T2, mature (M), and marginal zone (MZ) B cells were defined by relative HSA and CD21 expression densities. Bone marrow pro/pre-B, immature, and mature B cells were identified by relative IgM and B220 densities. Peritoneal B1a, B1b, and B2 cells were defined by CD11b, CD5, and B220 densities. All results represent those obtained with more than or equal to three 8-wk-old NOD or C57BL/6 mice. E and F, Splenocytes from NOD and C57BL/6 mice were labeled with CFSE at different intensities, mixed equally, and injected i.p. into (E) C57BL/6 or (F) NOD recipients before treatment with MB20-11 or control (CTL) mAb (250 μg). After 24 h, spleen and peripheral lymph nodes were isolated and stained for B220 expression before flow cytometry analysis. The gates show CFSE^high-labeled B220⁺ and B220⁻ lymphocytes from C57BL/6 donor mice and CFSE^low-labeled cells from NOD donor mice isolated from recipient mice. Bar graphs indicate relative ratios (±SEM) of cells from NOD and C57BL/6 donors within the CFSE-labeled B220⁺ and B220⁻ lymphocyte populations. Results represent those obtained with more than or equal to three mouse pairs.

Whether CD20 mAb resistance was an intrinsic property of NOD B cells was assessed by labeling NOD and C57BL/6 splenocytes with CFSE at different intensities before transfer into NOD or C57BL/6 recipients. One day following CD20 or control mAb treatments of recipient mice, the relative frequencies of CFSE-labeled B220⁺ and B220⁻ lymphocytes in the spleen and lymph nodes of recipient mice were quantified by flow cytometry analysis. B220⁻ splenocytes served as internal controls for relative numbers of transferred NOD or C57BL/6 splenocytes and to control for any variability in numbers of i.v. injected lymphocytes. CD20 mAb treatment depleted NOD and C57BL/6 B cells equally, with equivalent ratios of CFSE-labeled B220⁺ NOD and C57BL/6 splenocytes in C57BL/6 (Fig. 2⇑E) or NOD recipients (Fig. 2⇑F). Because the assay times were short and no differences were observed between C57BL/6 and NOD recipients with either C57BL/6 and NOD donor cells, histocompatability Ag differences were unlikely to influence the results. Thus, intrinsic B cell abnormalities did not explain decreased B cell depletion in NOD mice.

Whether CD20 mAb resistance reflected an inability to deplete specific B cell subsets was assessed in NOD and C57BL/6 mice after CD20 mAb treatment. One week following CD20 or control mAb treatment of NOD mice, tissue B cell numbers were quantified by flow cytometry analysis. The majority of splenic B cells (B220⁺), marginal zone B cells (CD21^highCD1d⁺B220⁺), T1 cells (CD21⁻CD24^highB220⁺), T2 cells (CD21^highCD24^high B220⁺), and mature B cells (CD21⁺CD24⁺B220⁺) were depleted in both NOD and C57BL/6 mice (Table I⇑). Most peripheral lymph node B cells (B220⁺) were also depleted in both NOD and C57BL/6 mice. Peritoneal cavity B1a (CD5⁺ CD11b⁺B220^low), B1b (CD5⁻CD11b⁺B220^low), and B2 (CD5⁻B220^high) cells in NOD and C57BL/6 mice were also depleted similarly. Furthermore, B220⁻ lymphocyte numbers were not affected by CD20 mAb treatment. Thus, decreased B cell depletion in NOD mice was a general defect that was not restricted to specific B cell subsets.

Defective FcγRI function in NOD mice

Because CD20 mAb-mediated B cell depletion is FcγR dependent (23), FcγR expression and amino acid sequences were examined in NOD mice. Thioglycolate-elicited peritoneal CD11b⁺F4/80⁺ macrophages from NOD mice expressed FcγRII/III epitopes identified by the 2.4G2 mAb at 40% lower levels than macrophages from C57BL/6 mice as determined by immunofluorescence staining with flow cytometry analysis (Fig. 3⇓A). A smaller but reproducible decrease in FcγRII/III expression by spleen monocytes from NOD mice was also observed relative to spleen monocytes from C57BL/6 mice. However, FcγRIII transcript levels in NOD and C57BL/6 peritoneal macrophages were similar when analyzed by semiquantitative RT-PCR analysis (data not shown). Thus, reduced 2.4G2 mAb staining in NOD mice was most likely due to reduced FcγRII expression, as described (43). By contrast, FcγRIV expression by peritoneal and spleen CD11b⁺F4/80⁺ monocytes was similar in NOD and C57BL/6 mice, as determined using the 9G8.1 mAb (34). Thus, reduced ADCC in NOD mice was not due to reduced FcγRIII or FcγRIV expression.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

FcγR expression and function in NOD mice. A, FcγRII/III and FcγRIV expression by spleen and thioglycolate-elicited peritoneal CD11b⁺F4/80⁺ monocytes from NOD (thick line) and C57BL/6 (thin line) mice. Dashed lines indicate control mAb staining. Representative results from one of three experiments is shown. B, Predicted FcγRIII and FcγRIV protein sequences for NOD and C57BL/6 mice. Identical residues are denoted with a dash. Differences in amino acid sequences are boxed. Residue numbering is as published (GenBank accession numbers: AY897419, AY897426, NM_144559, and EU050648). Extracellular domain 1 (EC1) and the cytoplasmic (CYTO) domains are shown. C and D, FcγR function. Binding of CD20 mAb-coated or untreated (Medium) A20 cells to COS cells transiently cotransfected with FcγR (FcγRI, FcγRIII, FcγRIV) and FcRγ-subunit cDNAs. Values represent mean (±SEM) numbers of A20 cells bound to COS cells in at least 30 fields from three independent experiments. Significant differences between sample means for CD20 mAb-coated and untreated A20 cells are indicated (∗, p < 0.05; ∗∗, p < 0.01). D, Representative A20 cell binding to COS cells transfected with FcγRI from NOD or C57BL/6 mice. E, IgG-immune complexes induce equal [Ca²⁺]_i response in Indo-1-loaded macrophages from NOD (thick line) and C57BL/6 (thin line) macrophages. Changes in [Ca²⁺]_i were quantified by flow cytometry with immune complexes added at the time point indicated (arrow). These results are representative of three experiments.

FcγR polymorphisms were also examined in NOD mice. NOD FcγRI transcripts differed from C57BL/6 transcripts by 24 nucleotides within coding regions, 17 of which encoded different amino acids including a four amino acid insertion between domains 2 and 3, and a frame shift that lead to premature truncation of the cytoplasmic domain (data not shown), as described (44, 45). FcγRIII and FcγRIV polymorphisms were also observed in coding regions from NOD mice. Four amino acid substitutions were identified in the FcγRIII extracellular and cytoplasmic domains (Fig. 3⇑B), as described (GenBank accession number: AY897426). One polymorphic residue was identified in FcγRIV, which led to an amino acid change in extracellular domain 1. FcγRI, FcγRIII, and FcγRIV require FcRγ for assembly and ADCC effector mechanisms (32). However, NOD and C57BL/6 FcRγ transcripts were identical (data not shown), despite the amino acid sequence differences found in FcγRI, FcγRIII, and FcγRIV.

To determine whether the FcγR sequence differences were functionally significant, COS cells transiently expressing FcγRs from NOD and C57BL/6 mice were analyzed for their ability to bind mouse A20 B lymphoma cells coated with CD20 mAbs in vitro. A20 cell binding to COS cells was visualized by light microscopy. COS cells expressed equivalent levels of NOD and C57BL/6 FcγRI transcripts as verified by semiquantitative RT-PCR analysis (data not shown). Equivalent FcγRIII and FcγRIV transcription by transfected COS cells was verified by flow cytometry analysis, as each FcγR expression plasmid contained an internal ribosomal entry site-GFP expression cassette downstream of the FcγR cDNA insert. Specifically, 50–60% of the COS cells expressed GFP, with similar fluorescence intensities. COS cells expressing C57BL/6 FcγRI bound IgG2a CD20 mAb-bearing A20 cells at significant levels (127 ± 13 cells/field), whereas cells expressing NOD FcγRI did not bind IgG2a CD20 mAb (9 ± 1 cells/field; p < 0.001, Fig. 3⇑, C and D). C57BL/6 FcγRI did not bind IgG1 or IgG2b CD20 mAbs in vitro. By contrast, COS cells expressing FcγRIII from either NOD or C57BL/6 mice bound A20 cells coated with IgG1, IgG2a, or IgG2b CD20 mAbs similarly. COS cells expressing NOD or C57BL/6 FcγRIV also bound IgG2a and IgG2b isotype CD20 mAbs equally, but did not bind IgG1 CD20 mAb. Thus, NOD FcγRI did not bind IgG2a mAbs efficiently when compared with C57BL/6 FcγRI, while NOD and C57BL/6 FcγRIII and FcγRIV were equivalent in their CD20 mAb-binding properties.

Although the pathways regulating macrophage FcγR signaling during ADCC are poorly defined, immune complex-mediated FcγR cross-linking induces a slow increase in intracellular calcium levels ([Ca²⁺]_i) as described (32, 46). Therefore, DNA-BSA/anti-DNP IgG immune complex-induced changes in [Ca²⁺]_i by peritoneal macrophages from NOD and C57BL/6 mice were compared to determine whether global defects in FcγR signaling explained reduced ADCC in NOD mice. However, induced [Ca²⁺]_i levels were similar if not identical in both NOD and C57BL/6 mice, suggesting intact FcγR signaling in NOD mice (Fig. 3⇑E).

FcγRI is required for optimal IgG2a CD20 mAb-mediated B cell depletion in vivo

Using a panel of CD20 mAbs of different isotypes, previous studies demonstrated that IgG2a CD20 mAbs required FcγRIV and potentially FcγRI for B cell depletion in vivo, while FcγRIII deficiency was without effect (23, 25). IgG2b CD20 mAbs preferentially used FcγRIV, and IgG1 mAbs used FcγRIII exclusively. To rule out the influence of mAb binding or affinity/avidity differences, the MB20-18 mAb was used to generate IgG1, IgG2a, and IgG2b mAbs with identical V regions so that the influence of mAb isotype on FcγR use could be assessed directly. Under these conditions, the IgG2a MB20-18 mAb was most potent for depleting B cells in C57BL/6 mice, while the IgG1 MB20-18 mAb was more potent than the IgG2b MB20-18 mAb (Fig. 4⇓).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

FcγRI is required for IgG2a CD20 mAb-mediated B cell depletion in vivo. Blood and spleen B cell depletion in C57BL/6 FcγRI/III^−/− (filled circles), C57BL/6 (open circles) and NOD (shaded circles) mice. FcγRI/III^−/− and C57BL/6 mice were treated using the IgG2a MB20-18, IgG2b MB20-18, and IgG1 MB20-18 CD20 mAbs, while NOD mice were treated using the IgG2c MB20-11, IgG2b MB20-18, and IgG1 MB20-1 mAbs. Values (±SEM) indicate the percentage of B cells present in mAb-treated mice (n = 6 for 10 μg/mouse, n = 3 for other doses) relative to control (CTL) mAb-treated littermates at each mAb dose evaluated. Significant differences between sample means of FcγRI/III^−/− and C57BL/6 mice (∗, p < 0.05), or FcγRI/III^−/− and NOD mice (†, p < 0.05) are indicated.

FcγRI/III^−/− double-deficient C57BL/6 mice were also generated and compared with wild-type mice for CD20 mAb-mediated B cell depletion using the IgG2a, IgG1, and IgG2b MB20-18 mAbs. Thereby, mice expressing only FcγRIV were compared directly with mice expressing FcγRI, FcγRIII, and FcγRIV. For the IgG2a CD20 mAb at 250 μg/mouse, B cell depletion was complete in both wild-type and FcγRI/III^−/− mice (Fig. 4⇑). However, at lower IgG2a CD20 mAb doses, the IgG2a mAb depleted significantly more B cells in wild-type mice. For example, IgG2a CD20 mAb (10 μg/mouse) only reduced circulating and spleen B cell numbers by 67 and 52% in FcγRI/III^−/− mice, respectively, while reducing B cell numbers by >95% in wild-type mice. Thus, CD20 mAb-mediated B cell depletion was significantly less efficient when only FcγRIV was expressed, demonstrating that FcγRI and/or FcγRIII were also required for optimal B cell depletion. Circulating and spleen B cells were depleted similarly by IgG2b CD20 mAb in both wild-type and FcγRI/III^−/− mice, confirming that IgG2b CD20 mAbs can use FcγRIV efficiently in the absence of FcγRI. Spleen B cells were not depleted by the IgG1 CD20 mAb in FcγRI/III^−/− mice, consistent with IgG1 mAbs using FcγRIII exclusively for CD20 mAb-mediated B cell depletion. However, the IgG1 CD20 mAb depleted 40–60% of circulating B cells in FcγRI/III^−/− mice, suggesting that FcγR-dependent and -independent pathways contribute to circulating B cell clearance by CD20 mAb, as described (25). When blood and tissue B cell depletion in NOD and FcγRI/III^−/− mice were compared over a range of IgG2a/c and IgG2b CD20 mAb doses, B cell depletion was similar at most mAb doses (Fig. 4⇑). Thus, the absence of normal FcγRI function in NOD mice explains the significant in vivo defect in IgG2a CD20 mAb-mediated B cell depletion.

Intrinsic ADCC defects in NOD mice

Spleen monocyte numbers were significantly reduced (60 ± 3%, p < 0.01, n = 5) in NOD mice compared with age-matched C57BL/6 mice. To assess whether monocyte activation could compensate for reduced monocyte numbers and enhance B cell depletion in vivo, NOD and C57BL/6 mice were each treated i.p. with thioglycolate to induce systemic macrophage activation and migration into the peritoneal cavity over a 24- to 48-h period, as described (47). One day after thioglycolate treatment, mice were given suboptimal doses of MB20-11 (25 μg), MB20-18 (100 μg), and MB20-1 (250 μg) mAb, with blood and tissue B cell numbers analyzed 7 days later. In mice treated with an IgG2c CD20 mAb, thioglycolate treatment did not enhance blood, spleen, or lymph node depletion in NOD mice, while B cells were depleted efficiently in C57BL/6 mice without thioglycolate treatment (Fig. 5⇓). Within the peritoneal cavity, thioglycolate treatment resulted in a 58% decrease in B cell numbers in NOD mice, with a 98% decrease in C57BL/6 mice relative to littermates that were not treated with thioglycolate. In mice treated with IgG1 and IgG2b CD20 mAbs, circulating B cells were depleted by 56% (IgG1) and 2% (IgG2b) without thioglycolate treatment, but 81% (IgG1) and 47% (IgG2b) with thioglycolate treatment at day 7. Thioglycolate treatment did enhance B cell depletion in lymph nodes of NOD mice, but most significantly reduced peritoneal cavity B cell numbers in both NOD and C57BL/6 mice. Thus, B cell depletion in both NOD and C57BL/6 mice was accelerated by monocyte activation regardless of CD20 mAb isotype, arguing that ADCC could be induced in NOD mice, although at reduced levels compared with C57BL/6 mice. Thus, reduced monocyte numbers and deficient FcγRI function are likely to explain reduced B cell depletion in NOD mice rather than defective monocyte activation.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Thioglycolate treatment augments B cell depletion in NOD mice. Blood (B220⁺), spleen (CD24⁺CD21⁺B220⁺), lymph node (B220⁺), and peritoneal cavity B2 (CD5⁻B220^high) cell depletion with (filled circles/squares/bars) or without (open circles/squares/bars) thioglycolate treatment. Mice were treated using the IgG2c MB20-11 (25 μg), IgG1 MB20-1 (250 μg), IgG2b MB20-18 (100 μg), or isotype- and dose-matched control mAbs. Values (±SEM) represent the percentage of B cells present in CD20 mAb-treated mice (n = 5 for IgG2c; n = 3 for IgG2b and IgG1) relative to control (CTL, n = 3) mAb-treated littermates. Significant differences between sample means of mice treated with and without thioglycolate are indicated: ∗, p < 0.05.

B cell depletion protects NOD mice against T1D

Despite reduced ADCC in NOD mice, B cells were nonetheless effectively depleted by high-dose IgG2c CD20 mAb treatment (Fig. 1⇑). Therefore, 5-wk-old female NOD mice were treated with CD20 mAb (250 μg) given biweekly over 6 wk (Fig. 6⇓A). Circulating B cell numbers remained reduced by >95% for up to 12 wk of age, but began to normalize thereafter. Circulating B cell numbers did not change in untreated littermates or littermates treated with an isotype-matched control mAb. Likewise, a single CD20 mAb treatment at 5 wk of age depleted the majority of spleen, pancreatic lymph node, and mesenteric lymph node B cells in 8-wk-old mice, with >95% fewer B cells within the islets of these mice (Fig. 6⇓B). Thus, B cell depletion was significant and durable in NOD mice following high-dose CD20 mAb treatment.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

B cell depletion protects NOD mice against T1D. A, Blood B cell depletion in female NOD mice. The frequency of blood B cells was measured over time in untreated NOD mice (squares, n = 4) or littermates treated three times (weeks 5, 7, and 9, arrows) with MB20-11 (circles; n = 5) or control (triangles; n = 5) mAb (250 μg). B, Tissue B cell depletion in female NOD mice. The frequency of B cells in spleen, PanLN (pancreatic lymph node), MLN (mesenteric lymph node), and islets was measured in 8-wk-old NOD mice (n = 2) 3 wk after a single injection of MB20-11 (▪) or control (▦) mAb (250 μg) or in untreated littermates (□). Significant differences between sample means are indicated: ∗∗, p < 0.01. C–F, B cell depletion delayed diabetes onset and progression in NOD mice. C–E, Five-week-old or, F, 15-wk-old female NOD female mice were treated biweekly three times (arrows) with MB20-11 (circles) or control (triangles) mAb (250 μg) and monitored for diabetes in comparison with untreated littermates (squares). C, Values represent results using groups of 11 littermate mice for the respective experimental groups. D, Pancreatic insulitis was assessed in 14-wk-old NOD mice 5 wk after the last injection of CD20 or control mAb compared with untreated littermates. A minimum of 20 islets was counted for each mouse, with five mice per group. E, Pancreatic lymphocyte frequencies were assessed in 17-wk-old NOD mice 8 wk after the last injection of CD20 or control mAb compared with untreated littermates. Results represent frequencies of CD19⁺ B cells and CD4⁺ or CD8⁺ T cells as assessed by flow cytometry. Islets were pooled from groups of four mice and cultured overnight in medium containing IL-2 (4 ng/ml). All data were normalized per 100 islets. F, Values represent results using seven MB20-11 and eight control mAb-treated littermates.

The impact of B cell depletion on diabetes progression was assessed using 5-wk-old NOD female mice that were untreated, or treated biweekly three times with high-dose CD20 or control mAb, with disease onset and incidence monitored for up to 53 wk. The majority of CD20 mAb treated NOD mice (7 of 11) remained diabetes free, while all of the untreated (n = 11) or control mAb-treated mice (n = 11) developed diabetes (Fig. 6⇑C). Moreover, B cell depletion significantly reduced the frequency of intrainsulitis when compared with untreated or control mAb-treated littermates when assessed at 14 wk of age (p < 0.05; Fig. 6⇑D). In CD20 mAb-treated mice, 32% of islets were free of insulitis with only 12% of the islets heavily infiltrated (e.g., >50% intrainsulitis). By contrast, the majority (e.g., >50%) of islets were heavily infiltrated in untreated or control mAb-treated mice. In a cohort of mice treated similarly, the number of lymphocytes within islets of 17-wk-old mice was reduced by 65% after CD20 mAb treatment (Fig. 6⇑E). However, the relative frequencies of CD19⁺ B cells and CD4⁺ and CD8⁺ T cells was similar in untreated, and CD20 or control mAb-treated mice (Fig. 6⇑E). Thus, the return of circulating B cells at 13 wk in NOD mice (Fig. 1⇑A) was likely to parallel the return of tissue B cells, as occurs in C57BL/6 mice (23, 24), with normalized B cell compartments by 17 wk. Nonetheless, the absence of B cells during the period of normal disease development in NOD mice significantly reduced the severity of insulitis and the frequency of lymphocytes migrating into islets, with most CD20 mAb-treated mice remaining disease free for up to 53 wk.

To test whether CD20 mAb treatment at a later preclinical stage of T1D was also effective, 15-wk-old female NOD mice were treated three times with CD20 mAb (weeks 15, 17, 19), with diabetes monitored for up to 40 wk. B cell depletion did not prevent 15-wk-old mice from developing diabetes, but diabetes onset was significantly delayed following CD20 mAb treatment when compared with control mAb-treated littermates (p < 0.05, Fig. 6⇑F). Collectively, B cell depletion either prevented or delayed the progression of T1D in NOD mice, depending on the stage of β cell autoimmunity.

B cell depletion reduces T cell proliferation, but does not induce regulatory effector cells

Others have demonstrated generally impaired lymph node T cell activation in congenitally B cell-deficient NOD mice, suggesting a critical need for B cell costimulatory signals within these microenvironments (48). Therefore, the effects of B cell depletion on T cell numbers, phenotypes, and proliferative capacity was assessed in NOD mice given CD20 or control mAb at 5 wk of age. Two weeks after CD20 mAb treatment, B cell numbers were significantly reduced, but CD4⁺ and CD8⁺ T cell numbers were not changed in the spleen, or pancreatic and peripheral lymph nodes (Table I⇑). Similarly, CD4⁺ and CD8⁺ T cell expression of cell surface markers indicative of activation and memory cell development were not changed following CD20 mAb treatment for 2 wk. This included CTLA-4, CD28, CD44, CD62L, CD69, CD154, and OX-40 expression within the spleen, and peripheral and pancreatic lymph nodes (Fig. 7⇓A and data not shown). By contrast, the ability of NOD pancreatic and peripheral lymph node CD4⁺ and CD8⁺ T cells to proliferate in response to CD3 plus CD28 costimulation in vitro was significantly reduced by CD20 mAb treatment in vivo (Fig. 7⇓B). Furthermore, splenic CD4⁺, but not CD8⁺ T cell, proliferation in response to CD3 plus CD28 costimulation was also affected by CD20 mAb treatment, but to a lesser extent. Thus, induced B cell depletion did not have global effects on the numbers or phenotypes of either CD4⁺ or CD8⁺ T cells, but significantly reduced the proliferative capacity of CD4⁺ and CD8⁺ T cells within the lymph nodes of NOD mice.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

B cell depletion effects on T cell phenotypes, mitogen responses, and suppressor cell frequencies in NOD mice 14 days after CD20 or control mAb treatment. A, Spleen CD4⁺ and CD8⁺ T cell phenotypes after B cell depletion. Histograms represent T cell expression of activation molecules after CD20 (thick line) or control (thin line) mAb treatments. Dashed lines represent control mAb reactivity. B, CD4⁺ and CD8⁺ T cell proliferation after B cell depletion. Splenocytes, and peripheral (PLN) and pancreatic (PanLN) lymph node cells from CD20 or control mAb-treated mice were isolated 2 wk after mAb treatment, CFSE labeled, and stimulated in vitro with CD3 plus CD28 mAbs (thick line) or medium alone (dashed line) for 70 h. Diminished CFSE intensity indicative of proliferation was assessed for mAb-labeled CD4⁺ and CD8⁺ T cells by flow cytometry. Representative CFSE profiles are shown (left panels). Right panels, CFSE geometric mean fluorescence (control (□) and CD20 (▪) mAb-treated mice), which is inversely proportional to cell divisions. A and B, Histograms representative of results obtained with more than or equal to three mice in each group are shown; ∗, p < 0.05 and ∗∗, p < 0.01; CD20 vs control mAb-treated groups. C, Regulatory/suppressor cells were not induced by B cell depletion. Splenocytes from diabetic NOD female mice were isolated and mixed equally with splenocytes from CD20 (circles) or control (triangles) mAb-treated 14-wk-old NOD female mice. Splenocytes were transferred into groups of five 6–8-wk-old NOD.scid mice, with diabetes monitored weekly.

In lupus patients, B cell depletion is proposed to result in regulatory or suppressor T cell generation (49, 50). To assess this, splenocytes from diabetic NOD mice were mixed with splenocytes from CD20 or control mAb-treated NOD mice and coadoptively transferred into NOD.scid recipients. The adoptive transfer of splenocytes from diabetic and control mAb-treated NOD donors into NOD.scid mice resulted in median diabetes onset 5 wk postcell transfer. The time of onset and the frequency of diabetes were not significantly affected, however, in NOD.scid mice receiving splenocytes from CD20 mAb-treated and diabetic NOD mice (Fig. 7⇑C). Consistent with this, the numbers of spleen and lymph node CD25⁺FoxP3⁺CD4⁺ regulatory T cells were not changed in NOD (Table I⇑) or in C57BL/6 mice (19). Thereby, B cell depletion did not significantly promote regulatory/suppressor T cell function in NOD mice.