ARTHRITIS & RHEUMATISMVol. 58, No. 5, May 2008, pp 1433–1444DOI 10.1002/art.23428© 2008, American College of Rheumatology

An Orally Bioavailable Spleen Tyrosine Kinase Inhibitor DelaysDisease Progression and Prolongs Survival in Murine Lupus

Frances Rena Bahjat,1 Polly R. Pine,1 Andrea Reitsma,1 Gail Cassafer,2 Muhammad Baluom,1

Sunny Grillo,1 Betty Chang,1 Fei Fei Zhao,1 Donald G. Payan,1

Elliott B. Grossbard,1 and David I. Daikh2

Objective. To assess whether R788, an orallybioavailable small molecule inhibitor of spleen tyrosinekinase (Syk)–dependent signaling, could modulate dis-ease in lupus-prone (NZB � NZW)F1 (NZB/NZW) micevia inhibition of Fc receptor (FcR) and B cell receptorsignaling.

Methods. R788 was administered to NZB/NZWmice before and after disease onset. Proteinuria, bloodurea nitrogen levels, and autoantibody titers were ex-amined periodically, and overall survival and renalpathologic features were assessed following long-termtreatment (24–34 weeks). The distribution and immu-nophenotype of various splenic T cell and B cell sub-populations were evaluated at the time of study termi-nation. Arthus responses in NZB/NZW mice pretreatedwith R788 or Fc-blocking antibody (anti-CD16/32) werealso examined.

Results. When R788 was administered prior to orafter disease onset, it delayed the onset of proteinuriaand azotemia, reduced renal pathology and kidneyinfiltrates, and significantly prolonged survival oflupus-prone NZB/NZW mice; autoantibody titers were

minimally affected throughout the study. Dose-dependent reductions in the numbers of CD4� acti-vated T cells expressing high levels of CD44 or CD69were apparent in spleens from R788-treated mice. Min-imal effects on the numbers of naive T cells expressingCD62 ligand and total CD8� T cells per spleen wereobserved following long-term drug treatment. R788 pre-treatment resulted in reduced Arthus responses inNZB/NZW mice, similar to results obtained in micepretreated with FcR-blocking antibody.

Conclusion. We demonstrate that a novel Syk-selective inhibitor prevents the development of renaldisease and treats established murine lupus nephritis.These data suggest that Syk inhibitors may be oftherapeutic benefit in human lupus and related disor-ders.

Systemic lupus erythematosus (SLE) is an auto-immune disease characterized by excessive and chronicactivation of inflammatory pathways via immune com-plex deposition, complement activation, and Fc receptor(FcR) ligation, often leading to significant clinical pa-thologies including glomerulonephritis, immune throm-bocytopenic purpura, hemolytic anemia, polyserositis,and vasculitis in humans. Mapping studies have revealeda region on telomeric chromosome 1 in mice and thesyntenic region in humans that often show strong linkageto lupus (1–6). Observations in nonautoimmune micecarrying chromosome 1 regions from lupus-prone strainshave further elucidated that genes in this region encod-ing Fc�RII (CD32) and Fc�RIII (CD16) are importantrisk factors for disease (7–9).

FcR�-deficient (NZB � NZW)F1 (NZB/NZW)mice generate and deposit immune complex and activatecomplement but are protected from severe nephritis,suggesting that the downstream inflammatory events aredependent on FcR� expression (4). FcR� deficiency

1Frances Rena Bahjat, PhD, Polly R. Pine, PhD, AndreaReitsma, Muhammad Baluom, PhD, Sunny Grillo, Betty Chang, PhD,Fei Fei Zhao, Donald G. Payan, MD, Elliott B. Grossbard, MD: RigelPharmaceuticals, South San Francisco, California; 2Gail Cassafer,David I. Daikh, MD, PhD: University of California, San Francisco, andSan Francisco VA Medical Center, San Francisco, California.

Dr. Bahjat, Dr. Pine, Ms Reitsma, Dr. Baluom, Ms Grillo, Dr.Chang, Ms Zhao, Dr. Payan, and Dr. Grossbard hold stock and/orstock options in Rigel Pharmaceuticals. Rigel Pharmaceuticals, Inc.,owns patents and patent applications covering various aspects of andrelating to R788.

Address correspondence and reprint requests to David I.Daikh, MD, PhD, University of California, San Francisco, and SanFrancisco VA Medical Center, 4150 Clement Street, Immunology/Arthritis 111R, San Francisco, CA 94121. E-mail: david.daikh@ucsf.edu.

Submitted for publication August 1, 2007; accepted in revisedform January 28, 2008.

1433

reduces experimental hemolytic anemia and thrombocy-topenia in mice (10) and modulates development ofanti–glomerular basement membrane (anti-GBM)antibody–induced glomerulonephritis (11). Studies us-ing bone marrow chimera strategies to produce micewith FcR� deficiency in specific cell compartments haveshown that hematopoietic cell expression of FcR, andnot kidney mesangial cell expression, is required forspontaneous disease development in lupus-prone NZB/NZW mice as well as for development of anti-GBM–induced nephritis, suggesting that inhibition of FcRsignaling in circulating leukocytes would potentially besufficient to ameliorate tissue-specific inflammation(12,13). NZB/NZW mice given soluble Fc�RIII dem-onstrate reduced anti–double-stranded DNA (anti-dsDNA) antibody levels, decreased proteinuria, andprolonged survival (14), providing significant rationalefor targeting FcR pathways for the treatment of lupus.

B cells also play a significant role in lupus patho-genesis via production of autoantibodies, resulting inantigen–antibody immune complex formation and stim-ulation of FcR. B cell receptors and FcR share acommon signaling pathway, and both are transmem-brane molecules containing immunoreceptor tyrosine–based activation motif (ITAM) or immunoreceptortyrosine–based inhibition motif (ITIM), depending onthe receptor type, which become phosphorylated by Srcfamily kinases (15,16). The dual phosphorylation ofITAMs provides for high-affinity interaction with spleentyrosine kinase (Syk), a molecule that is crucial for thesignaling of B cell receptors and activating FcR. Becausethe pathogenesis of SLE has been associated with FcRexpression and autoreactive B cells, we hypothesizedthat inhibition of Syk using a recently described orallybioavailable small molecule Syk-selective inhibitor(R406) (17), prepared as a prodrug (R788), wouldameliorate disease in spontaneous lupus–prone mice.

MATERIALS AND METHODS

Animals. Female NZB/NZW and 4-month-old maleBALB/c mice were obtained from The Jackson Laboratory(Bar Harbor, ME) and acclimated for several months. Closelyage-matched NZB/NZW mice were monitored for protein-uria beginning at 5–6 months of age and randomized intogroups at the start of the study. Temperature was maintainedat 72 � 5°F, with relative humidity at 35–70% and a 12-hourlight/dark cycle. Mice were given rodent chow and water adlibitum. All studies were approved by the Rigel InstitutionalAnimal Care and Use Committee.

Drug formulation. R788 was prepared in 0.1% car-boxymethylcellulose sodium, 0.1% methylparaben, 0.02%propylparaben/H2O vehicle. R788 prodrug is converted in vivo

to the active metabolite, R406, and little R788 reaches thesystemic circulation.

Treatments. Lupus mice were treated twice daily withR788 or vehicle by oral gavage, with an 8-hour intervalbetween the morning and evening treatments. For Arthusstudies, mice received R788 orally 30 minutes prior to intra-dermal antibody challenge or were intraperitoneally injectedwith NA/LE rat anti-mouse CD16/32 antibody (Fc Block[clone 2.4G2]; BD Biosciences, San Jose, CA) in 200 �l sterile0.9% normal saline, �18 hours prior to challenge.

Spontaneous lupus efficacy modeling. PrediseasedNZB/NZW mice (urinary protein level �100 mg/dl) ages 6–7months were enrolled and randomized according to the levelof proteinuria (n � 30 per group). The mice were treated forup to 240 days with vehicle or R788 (orally, twice daily) andmonitored until moribund or until �80% mortality was ob-served in the naive cohort. Urinary protein levels, blood ureanitrogen (BUN) levels, and anti-dsDNA antibody titers weremeasured periodically and when the mice were killed. Surviv-ing mice were killed over 2 consecutive days (days 239 and240), 1–2 hours postdose. On day 239, 10 mice per group wererandomly selected for spleen immunophenotyping, and theremaining mice were killed on day 240. At least 2 independentstudies were performed, and representative data are shown.Similar studies were performed using mice with establishedproteinuria, defined as urinary protein levels of 100–300 mg/dl(relative score 2�). Results from a single long-term (175 days)survival study (18 vehicle-treated mice, 20 mice receiving 20mg/kg of R788, and 20 mice receiving 40 mg/kg of R788) anda 30-day treatment study (15–16 mice/group) are reported.

Proteinuria and BUN. Mice were bled by tail nick, andblood was used for semiquantitative Azostix analysis (Bayer,Elkhart, IN), which was performed by assessing the color of thestrip compared with a given standard. Values of 5–15 mg/dl,15–26 mg/dl, 30–40 mg/dl, or 50–80 mg/dl were assigned.Proteinuria was similarly assessed in mouse urine using Al-bustix (Bayer). Urinary protein concentrations of trace (�30mg/dl), 1 (30� mg/dl), 2 (100� mg/dl), or 3 (300� mg/dl) wereassigned.

Serum anti-dsDNA titers. Diluted sera were appliedto poly[dA-dT]–coated plates (Sigma, St. Louis, MO) andincubated for 1 hour at room temperature prior to washing.Peroxidase-conjugated goat anti-mouse IgG Fc (Cappel,West Chester, PA) secondary antibody was added at 1 �g/mlin 2% bovine gamma globulin/1% bovine serum albumin(Sigma) for 1 hour at room temperature prior to washing.O-phenylenediamine substrate (Invitrogen/Zymed, South SanFrancisco, CA) in citrate–phosphate buffer (pH 5.0) and0.03% hydrogen peroxide were added, and samples were readat 490 nm on a spectrophotometer (BioTek Instruments,Winooski, VT).

Kidney histopathology. Kidneys were fixed in 10%buffered formalin, paraffin-embedded, sectioned, and stainedwith hematoxylin and eosin (H&E) for scoring. Kidney histo-pathology (n � 11/group) was graded by a single blindedinvestigator (DD) using a systematic scoring system, as previ-ously described (18). Damage indices included the following:glomerular sclerosis (0–4-point scale, where 0 � no sclerosisand 4 � complete sclerosis), tubular dilatation, and perivascu-lar lymphocytic infiltration. The extent of glomerular sclerosisin a given mouse was expressed as the average score deter-

1434 BAHJAT ET AL

mined from 25 individual glomeruli and as the percentage ofseverely sclerosed glomeruli (number of glomeruli [from atotal of 25] with a sclerosis score of 3 or 4). The extent oftubular dilatation was expressed as the percentage of the totaltubular area in a complete sagittal section that containeddilated tubules. The extent of perivascular infiltration wasexpressed as the number of vessels with large perivascular in-filtrates in a complete sagittal section. H&E-stained sampleswere also independently scored by a second blinded patholo-gist using a graded system, with similar results (data notshown).

Immunohistochemical staining. One kidney from eachmouse was embedded in OCT compound and frozen in liquidnitrogen. Immunohistochemical analysis was performed on10–12-�m cryostat sections using standard avidin–biotin com-plex techniques after blocking endogenous biotin. Target cellswere identified using either horseradish peroxidase–conjugated secondary antibody visualized by deposition ofdiaminobenzidine or by indirect immunofluorescence offluorescein isothiocyanate (FITC)– or phycoerythrin (PE)–conjugated secondary antibody. Tissue sections (n � 15/group)were stained with the following monoclonal antibodies: anti-CD3 (145-2C11; BD Biosciences), anti-CD4 (RM4-5; BDBiosciences), anti-B220 (RA3-6B2; Caltag, South San Fran-cisco, CA), anti-CD11b (M1/70; BD Biosciences), and anti-CD11c (HL3; BD Biosciences). B220-, CD4-, and CD11c-expressing cells were scored as the average number of positivecells in 25 high-power fields. CD11b-expressing cells weregraded as 0 (no positive staining) to 4�. Indirect immunoflu-orescence was performed for IgG (M30215; Caltag/Invitrogen,South San Francisco, CA), IgM (M31515; Caltag/Invitrogen),and C3 (HyCult, Uden, The Netherlands), using biotinylatedanti-rat Ig (Vector, Burlingame, CA) and FITC-conjugatedstreptavidin (HyCult).

Platelet count determinations. Analyses were per-formed on mouse blood collected in dipotassium EDTA–coated Microtainer tubes (BD Biosciences) at study termina-tion, using a Cell-Dyn 3700 analyzer (Abbott, Abbott Park, IL).

Spleen flow cytometric analyses. Splenocyte single-cell suspensions were prepared by mechanical disruption inice-cold staining buffer (phosphate buffered saline [PBS]without calcium and magnesium supplemented with 2%heat-inactivated fetal bovine serum [FBS]), passed through a70-�m filter, washed, and resuspended. Red cell lysis wasperformed using 1� PharmLyse (BD Biosciences). Cell countswere determined by hemocytometer or Cell-Dyn, and 1E6cells were aliquoted. Prior to surface staining, purified ratanti-mouse CD16/32 antibody (Fc Block; BD Biosciences)was added to each sample to block nonspecific FcR binding.Cells were stained with antibodies specific for CD4, CD8,CD19, CD21, CD23, CD25, CD44, CD62 ligand (CD62L),CD69, CD86, CD138, GL7, and class II major histocompati-bility complex (I-Ad) (BD Biosciences) conjugated with allo-phycocyanin, biotin, fluorescein, PE, or peridinin chlorophyllA protein–Cy5.5 conjugate. Fluorochrome-conjugated strepta-vidin secondaries were used for biotin conjugates. The optimalworking concentrations were determined prior to use. Germi-nal center, class-switched B cells (CD19�/IgM� GL7�),follicular cells (CD19�/CD21intermediateCD23high), marginalzone cells (CD19�/CD21highCD23low), newly formed cells(CD19�/CD21lowCD23low), and plasma cells (CD19�

CD138�) were determined by gating, as indicated. Sampleswere acquired on a FACSCalibur (BD Biosciences), with50,000 events collected per sample, and analysis was per-formed using FlowJo software (Tree Star, San Carlos, CA).Absolute numbers were calculated using population frequen-cies (%) and total cells/spleen for each animal and areexpressed as the group mean � SEM.

Intracellular phospho flow cytometry. Splenocytesfrom female BALB/c mice (13–18 weeks old) were collectedin Vacutainer sodium heparin tubes (BD Biosciences).Spleens were processed as described above, resuspended inculture medium (RPMI 1640 containing 10% FBS [volume/volume]) and counted, and 2E6 cells were aliquoted. R406 wasdissolved in 100% DMSO and added to cells in triplicate (0.1%final DMSO concentration). Samples were incubated for 30minutes at room temperature. Cells were stimulated with goatF(ab�)2 anti-mouse IgM (catalog no. 1021; Southern Bio-technology, Birmingham, AL) at 10 �g/ml for 10 minutes at37°C, then washed and stained with goat anti-mouse FITC-conjugated IgM (catalog no. 1021-02; Southern Biotech-nology) and rat anti-mouse PE-conjugated CD45R/B220monoclonal antibody (clone RA3-6B2; BD Biosciences) for30 minutes at room temperature. Cells were fixed with 1�Phosflow Fix Buffer (BD Biosciences) for 10 minutes at 37°C,permeabilized with Phosflow Perm Buffer III on ice for30 minutes, rehydrated, washed with PBS plus 1% bovineserum albumin 3 times, then stained with AlexaFluor 647–conjugated anti–phospho-ERK-1/2 (T202/Y204) (20A; BDBiosciences).

Arthus reaction. Mice were given 0.2 ml of chickenegg albumin solution (2 mg/ml; antigen) (Sigma) preparedin saline containing 0.5% Evans blue dye, administered in-travenously. Mice were treated with R788 or vehicle 30 min-utes prior to intradermal challenge with 25 �g or 50 �govalbumin (OVA)–specific IgG antibody (C-6534; Sigma) in25 �l per mouse on one side and isotype control IgG (I-8140;Sigma) on the contralateral side. Edema was evaluated 4 hoursafter the OVA-specific antibody challenge by measuring thediameter of extravascular leakage of Evans blue dye on thereverse side of the injection site with calipers (edema area).Dye was extracted from a tissue punch biopsy specimen using1 ml of formamide at 80°C overnight. Evans blue dye wasmeasured spectrophotometrically (optical density at 610 nm)using a SpectraMax M5 (Molecular Devices, Sunnyvale, CA).Results are expressed as the difference between isotype andOVA IgG sites. Fc Block was administered �18 hours prior tochallenge.

Statistical analysis. Analyses were performed usingGraphPad Prism version 4.0 (GraphPad Software, San Diego,CA). For mice that died before study termination, the lastknown values for dsDNA, BUN, and urinary protein werecarried forward. Statistically significant differences were deter-mined using one-way analysis of variance with Tukey-Kramer’spost hoc comparison (for �3 groups) or a 2-tailed unpairedStudent’s t-test (for 2 groups). Renal pathology and infiltrateswere compared using a Mann-Whitney U test. Survival datawere analyzed using Kaplan-Meier plots, and significance wasdetermined using a Mantel-Haenszel log rank test. P valuesless than 0.05 were considered significant. The sample size forpERK data (n � 3) was insufficient to determine normality,and these statistics were omitted.

NOVEL Syk-SELECTIVE INHIBITOR PREVENTS AND TREATS MURINE LUPUS NEPHRITIS 1435

RESULTS

Effect of R788 treatment on disease progressionand survival of prediseased lupus-prone mice. Pro-longed R788 treatment of prediseased lupus-proneNZB/NZW mice reduced disease progression (Figure1). By study termination (day 240), only 2 of 29 micetreated with the 40-mg/kg dose of R788 demonstratedelevated proteinuria (urinary protein level 300 mg/dl),compared with 21 of 30 mice in the vehicle-treatedgroup (Figure 1a). All 29 mice treated with 40 mg/kg ofR788 survived until study termination (P � 0.0001versus vehicle-treated mice) (Figure 1b). Protectionwas extended to mice treated with intermediate dosesof R788, wherein 21 of 30 mice (70%) treated with

20 mg/kg of R788 and 19 of 30 mice (63%) treated with10 mg/kg of R788 survived until study termination(P � 0.08 and P � 0.12, respectively, versus vehicle-treated mice). Disease progressed (urinary proteinlevels 300 mg/dl) in �50% of mice treated with lowerdoses of R788, and 5 of 30 naive mice (17%) and 14 of30 vehicle-treated mice (47%) survived until study ter-mination.

BUN levels were also reduced in mice that re-ceived R788 treatment, and reduced body weight losseswere observed (data not shown). Regardless of R788treatment, accumulation of pathogenic anti-dsDNAautoantibodies was observed, and serum titers increased2–4-fold in all groups at study end when compared with

Figure 1. Effect of R788 treatment on disease progression in and survival of (NZB � NZW)F1 mice. Prediseased mice (proteinuria �100 mg/dl;n � 30 per group) were randomized. One animal from the group randomized to receive 40 mg/kg of R788 was killed on day 60 due to a necroticlesion of unknown origin on the nose. a, The percentage of mice with significant proteinuria (urinary protein level 300 mg/dl) was reducedfollowing 240 days of R788 treatment. b, Kaplan-Meier survival plots demonstrated significantly prolonged survival of prediseased mice treated with40 mg/kg of R788 for up to 240 days (for 40 mg/kg versus vehicle, P � 0.0001; for naive versus all other groups, P � 0.01; for 10 mg/kg and 20 mg/kgversus 40 mg, P � 0.003, by log rank test). c, Data representing the mean dilution required for detection of serum double-stranded DNA antibodiesshowed a trend for reduction with R788 treatment at a dose of 40 mg/kg compared with vehicle. Two independent studies were performed, andrepresentative data are shown. Titers at baseline were detectable at a 1:128 dilution. d, Platelet counts in whole blood from mice treated for 240 dayswith R788 were increased in a dose-dependent manner. Each symbol represents an individual mouse; bars show the mean � SEM.

1436 BAHJAT ET AL

baseline levels (Figure 1c). Total IgG titers in thevehicle-treated group and the group receiving 40 mg/kgof R788 increased from 2.6 mg/ml at study initiationto �4.5 mg/ml after 195 days of treatment, but nodifference was detected between the vehicle-treatedgroup and the group treated with 40 mg/kg of R788(mean � SEM on day 195 4.41 � 0.53 mg/ml versus4.87 � 0.37 mg/ml, respectively).

Platelet counts increased following prolongedR788 treatment, suggesting protection againstautoantibody-mediated platelet depletion (Figure 1d).These data were consistent with the inhibitory activity ofR788 observed in a mouse model resembling humanimmune thrombocytopenic purpura (19).

Following 34 weeks of R788 treatment, micedemonstrated reductions in glomerular and tubularchanges and reduced cellular infiltrates (Table 1 andFigure 2), and specific staining of B220�, CD4�,CD11b�, and CD11c� cells revealed reductions ofthese populations within the kidney glomeruli and inter-stitium compared with that observed in vehicle-treatedcontrols (Table 1 and Figure 2a). Glomerular deposi-tions of IgG, IgM, and C3 were present in treated mice,but all were significantly reduced in mice that received20 mg/kg of R788 compared with vehicle-treated con-trols (Table 1). Mice examined prior to the start of drugtreatment (baseline mice) showed no significant signs ofrenal disease.

Effects of R788 treatment on spleen immuno-phenotype and B cell activation. Following prolongedR788 treatment of prediseased mice, the absolute num-ber and proportion of various splenic subpopulationswere affected (Table 2 and Figure 2b), as compared withvehicle-treated controls. CD19� cell numbers were re-duced, while the proportion per spleen remained un-changed. Class-switched germinal center, marginal zone,newly formed, and follicular B cell populations werereduced in number following treatment with �20 mg/kgR788, as were CD19� class II MHC� CD86� cells(Table 2). Direct inhibitory effects of R406, the activemetabolite of R788, on B cell receptor signal transduc-tion were demonstrated by reduced phosphorylation ofERK, a key signaling intermediate downstream from Syk(Figure 2c). Ex vivo pretreatment of BALB/c spleno-cytes with R406 dose-dependently reduced p-ERK levelsin splenic B220�/IgM� B cells stimulated with anti-IgM, with an approximate median effective dose of 1 �M(Figure 2d). Importantly, micromolar levels of R406 areachievable in mice following administration of 20mg/kg of R788.

Total CD8� cell counts per spleen were notdifferent between vehicle-treated and R788-treatedmice, despite a 50% reduction in spleen weight (data notshown) and cellularity (Table 2). The proportion ofCD8� cells increased by 3-fold, whereas no change inthe proportion of CD4� cells per spleen was observed

Table 1. Results of renal histopathology and immunohistochemistry*

Treatment

None Vehicle

R788

10 mg/kg 20 mg/kg 40 mg/kg

Glomerular sclerosis score 2.1 � 0.6 2.1 � 0.7 1.2 � 0.3† 1.3 � 0.4† 1.0 � 0.0‡Severe sclerosis 31 � 25 31 � 24 1 � 4† 7 � 14† 0 � 0‡Tubular dilatation 29 � 27 18 � 19 5 � 11† 3 � 3† 0 � 0‡Perivascular infiltrates 4.7 � 3.6 5.5 � 3.9 1.8 � 2.1§ 3.0 � 5.0§ 0.1 � 0.3‡B220� 2.0 � 4.2 2.7 � 4.3 0.5 � 0.6 0.5 � 1.8 0.0 � 0.0§CD4� 1.6 � 0.7 1.3 � 0.8 1.3 � 0.8 0.8 � 0.6§ 0.4 � 0.4‡CD11b� 2.0 � 0.9 2.1 � 0.9 1.3 � 0.6† 1.2 � 0.6§ 0.7 � 0.3‡CD11c� 8.9 � 4.1 8.2 � 7.2 2.7 � 3.2 4.2 � 3.1 0.5 � 1.6‡IgG� 3.0 � 0.9 2.6 � 0.8 2.0 � 1.0 1.1 � 0.9‡ 1.3 � 1.0‡IgM� 3.1 � 0.9 2.9 � 0.6 2.3 � 0.6‡ 1.8 � 0.8‡ 2.0 � 0.8‡C3� 3.5 � 1.5 2.7 � 1.1 2.0 � 1.1 1.6 � 0.8† 1.5 � 0.8‡

* Values are the mean � SD. Values for glomerular sclerosis are based on a 0–4 scale (0 � no sclerosis).Values for severe sclerosis are the percent of glomeruli (from a total of 25) with a glomerular sclerosisscore of 3 or 4. Values for tubular dilatation are the percent of the total tubular area in a complete sagittalsection that contained dilated tubules. Values for B220�, CD4�, and CD11c� are the number of cells in25 high-power fields. Values for CD11b� are based on a 0–4 scale (0 � no positive staining). Values forIgG�, IgM�, and C3� are based on a 0–5 scale (0 � no deposition).† P � 0.01 versus vehicle, by Mann-Whitney U test.‡ P � 0.001 versus vehicle, by Mann-Whitney U test.§ P � 0.05 versus vehicle, by Mann-Whitney U test.

NOVEL Syk-SELECTIVE INHIBITOR PREVENTS AND TREATS MURINE LUPUS NEPHRITIS 1437

after treatment with 40 mg/kg R788 compared withvehicle (Figure 2b), resulting in normalization of theCD4:CD8 T cell ratio (Table 2). R788 treatment at a

dose of 40 mg/kg decreased the number of activatedCD4� T cells (CD4�CD69� and CD4�CD62L�CD44high populations) by 70% (Table 2). Treatment

Figure 2. Effect of R788 treatment on renal pathologic features in (NZB � NZW)F1 mice. a, Kidneys from mice that were killed when moribundand those surviving to study termination were collectively evaluated, revealing dose-dependent reductions in glomerular and tubular effects followingR788 treatment. Representative hematoxylin and eosin (H&E)–stained sections are shown. Bar � 30 �m. b, Spleens from surviving mice following239 days of treatment or mice with established proteinuria after 30 days of treatment were examined by flow cytometry for changes in the proportionsof T cell subpopulations. Data are from an individual animal and are representative of the mean for each group. Representative dot plots (CD4versus CD8) are shown in the top row; numbers in the upper right quadrants denote the percentages of total viable gated lymphocytes for eachquadrant. All histograms include CD4-gated events; the numbers in the histograms are the percentage of positive cells of the total number of CD4�gated events. c, BALB/c mouse splenocytes stimulated with anti-IgM were evaluated by intracellular phospho flow cytometry for the percentage ofpositive p-ERK in B220�/IgM� gated B cells. Results are representative of duplicate experiments. d, Ex vivo pretreatment of BALB/c mousesplenocytes with R406 reduced p-ERK levels in splenic B220�/IgM� B cells stimulated with anti-IgM, in a dose-dependent manner. Values are themean and SD results from triplicate measurements for a single experiment.

1438 BAHJAT ET AL

with 40 mg/kg R788 decreased the proportions of CD69-and CD44-expressing CD4� cells, while the proportionof naive CD4� T cells expressing CD62L was increased�2-fold compared with that in vehicle-treated controls(Figure 2b).

Mean spleen cell counts in mice with establisheddisease at the time of study initiation were reducedfollowing 30 days of oral R788 treatment at a dose of40 mg/kg twice daily (Table 2). Although mean spleenweights were decreased by 50%, values did not return tobaseline levels. After 30 days of treatment, mean countsof naive T cells expressing CD62L were virtually identi-cal between the vehicle-treated group and the groupreceiving 40 mg/kg of R788 (data not shown), whereas a70% reduction in the number of cells expressing highlevels of CD44 was observed (Table 2). In our studies,CD4�CD25�CD62L� cell counts in spleens from micetreated for 30 days with 40 mg/kg of R788 were reduced

by �30% (P 0.05 versus vehicle), and similar datawere also observed following 239 days of treatment with40 mg/kg of R788 in prediseased mice (Table 2), sug-gesting that this subpopulation containing autoimmuneregulatory cells (20–24) was not depleted.

Effect of R788 treatment on established diseasein lupus-prone mice. R788 treatment of mice withestablished disease resulted in delayed disease progres-sion and prolonged survival, despite the presence ofproteinuria at the time when treatment was initiated.Proteinuria and BUN levels were dose-dependently pre-served (Figures 3a and b). By the end of the study,following 175 days of treatment, �95% of mice treatedwith 40 mg/kg of R788 demonstrated proteinuria scoresthat were lower than the initial scores, compared with17% of mice in the vehicle-treated group. Treatmentwith R788 at both the 20 mg/kg and 40 mg/kg dosessignificantly prolonged survival compared with that

Table 2. Absolute numbers of splenic subpopulations*

PopulationBaseline(n � 10)

Prediseased mice, 239 days of treatment

Mice with establishedproteinuria, 30 days of

treatment

Vehicle(n � 10)

R788,10 mg/kg(n � 10)

R788,20 mg/kg(n � 8)†

R788,40 mg/kg(n � 10)

Vehicle(n � 15)

R788,40 mg/kg(n � 16)

Total count per spleen 51.2 � 4.5 88.6 � 22.4 75.2 � 6.9 71.0 � 13.0 29.8 � 1.3‡ 115.0 � 16.4 52.8 � 5.1§Total CD19� 23.31 � 3.17 41.69 � 9.36 34.35 � 5.09 29.68 � 7.41 14.47 � 0.77‡ 61.00 � 8.67 24.94 � 3.29§B cell:T cell ratio 1.42 � 0.16 1.69 � 0.24 1.65 � 0.35 1.52 � 0.50 1.33 � 0.24 1.57 � 0.12 1.09 � 0.10¶GC, class-switched 1.74 � 0.32 2.50 � 0.61 2.78 � 0.39# 2.07 � 0.69 0.57 � 0.09‡ 4.95 � 0.89 1.54 � 0.32¶Follicular 14.23 � 1.62 18.27 � 3.01 16.07 � 2.59# 11.97 � 2.92 5.01 � 0.64** 34.16 � 4.22 13.86 � 2.18§Marginal zone 4.26 � 0.69 5.28 � 1.26 3.92 � 0.93 4.90 � 1.42 3.12 � 0.41 4.18 � 0.68 3.03 � 0.56Newly formed 7.27 � 1.41 21.99 � 8.19 19.22 � 3.31 15.89 � 4.18 7.15 � 0.98 21.76 � 6.13 6.02 � 0.97‡CD19� MHCIIhigh CD86� 3.03 � 0.58 1.02 � 0.26†† 0.90 � 0.16†† 0.69 � 0.16†† 0.34 � 0.05†† 9.76 � 1.35 4.19 � 0.59Plasma cells 0.36 � 0.08 0.95 � 0.46 0.80 � 0.25 0.38 � 0.07 0.30 � 0.07 0.36 � 0.08 0.10 � 0.01¶APC (CD19�/MHCIIhigh) 5.41 � 1.10 4.50 � 2.24 2.73 � 0.43 1.72 � 0.56 0.47 � 0.05§§ 2.96 � 0.69 0.70 � 0.13¶CD4:CD8 ratio 2.71 � 0.35 2.85 � 0.50 3.36 � 0.52# 2.84 � 0.75 1.30 � 0.16 4.52 � 0.65 2.18 � 0.18§Total CD8� 5.57 � 0.59 7.08 � 1.23 6.10 � 0.93 5.65 � 0.84 5.13 � 0.64 7.03 � 0.58 7.14 � 0.71Total CD4� 13.72 � 0.93 21.01 � 7.94 17.89 � 2.34 15.07 � 4.03 5.90 � 0.48 30.59 � 4.83 14.83 � 1.33¶CD4�/CD25�CD69� 1.94 � 0.33 2.89 � 1.16 2.43 � 0.29 2.16 � 0.55 1.04 � 0.08 1.96 � 0.20 1.07 � 0.08¶CD4�/CD25�CD69� 1.00 � 0.14 4.47 � 1.71 4.21 � 0.91 3.14 � 1.19 0.45 � 0.09‡ 9.78 � 2.18 3.47 � 0.50¶CD4�/CD25�CD69� 1.42 � 0.12 2.75 � 1.40 2.12 � 0.47 1.26 � 0.39 0.42 � 0.05 4.72 � 0.96 1.55 � 0.21¶CD4�/CD62L�CD44high 9.30 � 1.10 14.16 � 6.35 11.92 � 2.18 8.17 � 2.45 2.52 � 0.23 21.75 � 4.92 7.20 � 0.93¶CD4�/CD25�CD62L� ND 0.74 � 0.18 0.65 � 0.10 0.47 � 0.07 0.47 � 0.03 1.44 � 0.10 0.96 � 0.07§

* Values are the mean � SEM absolute count � 106 cells/spleen. Prediseased mice were killed and evaluated at age 14 months; mice with establishedproteinuria were killed and evaluated at age 11 months. Baseline values were established at the time of study initiation in untreated 6-month-oldmice. GC � germinal center; APC � antigen-presenting cells defined by high class II major histocompatibility complex (MHCII) expression;CD62L � CD62 ligand; ND � not determined.† Two samples were excluded due to sample collection errors.‡ P � 0.05 versus vehicle.§ P � 0.001 versus vehicle.¶ P � 0.01 versus vehicle.# P � 0.05 versus 40 mg/kg R788.** P � 0.01 versus vehicle; P � 0.05 versus baseline.†† P � 0.001 versus baseline.§§ P � 0.05 versus baseline.

NOVEL Syk-SELECTIVE INHIBITOR PREVENTS AND TREATS MURINE LUPUS NEPHRITIS 1439

among vehicle-treated control mice (Figure 3c). Overall,55% of mice treated with 20 mg/kg of R788 and 85% ofmice treated with 40 mg/kg of R788 survived to day 175,compared with 28% of vehicle-treated mice.

Treatment with R788 after disease onset resultedin a decrease in the incidence and severity of lupus-associated renal changes compared with that amongvehicle-treated controls (data not shown). No micro-scopic evidence of renal changes was observed in 47% ofthe mice treated with 40 mg/kg of R788, while only 10%of vehicle-treated control mice showed this phenotype.Glomerular and tubular changes consistent with lupusnephritis were routinely observed in vehicle-treatedmice after 175 days compared with kidneys from base-line mice that demonstrated relatively infrequent andmild pathology. Mice with established proteinuria alsodemonstrated a 77% increase in the mean platelet countfollowing 30 days of treatment with 40 mg/kg of R788(Figure 3d).

Effect of R788 treatment on the FcR-dependentArthus response in normal and lupus-prone mice. Theresponse to FcR activation can be modeled in rodentsusing systemic administration of OVA followed by lo-calized administration of OVA-specific IgG, referred toas the reverse passive cutaneous Arthus reaction. Innonautoimmune male BALB/c mice, pretreatment withR788 administered at doses of 20 mg/kg and 40 mg/kg 30minutes prior to OVA-specific antibody challenge re-sulted in statistically significant reductions in inflamma-tory edema (Table 3). Pretreatment of NZB/NZW micewith R788 also resulted in dose-related reductions ininflammatory edema; however, doses of 80 mg/kg ofR788 were required to achieve statistical significance(Table 3). Preliminary data suggest that NZB and NZWparental strains similarly required higher doses of R788(data not shown) for inhibition and showed similarvariability in responsiveness. Importantly, CD16/32-specific antibody (Fc Block) resulted in significant inhi-bition of the Arthus response in both lupus-prone NZB/NZW and normal BALB/c mice (Table 3), consistentwith the severe impairment of the Arthus reactionobserved in �-chain– and Fc�RIII (CD16)–deficientmice (4,25–27).

DISCUSSION

Recent data indicate that activating FcR proteinsplay a pivotal role in the onset and progression ofimmune complex–mediated inflammation and diseasesusceptibility in SLE. Due to the B cell receptor and FcRshared signaling pathways involving Syk, we hypothe-

Figure 3. Effect of R788 treatment on disease progression in andsurvival of (NZB � NZW)F1 mice with established proteinuria. Micewith established urinary protein levels of �100 mg/dl (score � 2) andblood urea nitrogen (BUN) levels of �30 mg/dl were treated untilmoribund or study end (day 175) and evaluated for disease progressionand survival (18 mice in the vehicle group, 20 mice in the 20 mg/kggroup, and 20 mice in the 40 mg/kg group). a and b, R788 reduced thepercentage of mice with urinary protein levels 300 mg/dl (a) and thepercentage of mice with BUN levels �30 mg/dl (b). For mice that diedbefore study termination, the last known values were carried forward.c, Treatment with 20 mg/kg and 40 mg/kg of R788 for 175 dayssignificantly prolonged survival of mice with established proteinuria(P � 0.03 and P � 0.001 versus vehicle, respectively, by log rank test).d, Platelet counts were significantly increased in mice treated with 40mg/kg of R788 for 30 days (P � 0.002 versus vehicle; n � 15–16mice/group). Each symbol represents an individual mouse; bars showthe mean � SEM.

1440 BAHJAT ET AL

sized that treatment with the selective Syk inhibitorR788 would result in a therapeutic benefit in lupus-prone mice via simultaneous inhibition of B cell receptorand FcR signaling. We demonstrate that administrationof R788 in lupus-prone mice prevents the developmentof renal disease and treats established murine lupus andresults in marked attenuation of renal disease, as dem-onstrated by reduced glomerular sclerosis and tubulardamage, proteinuria, and BUN levels.

The direct effects of R788 on specific cell popu-lations can be postulated based on known drug targets.The active metabolite of R788, referred to as R406,potently inhibited IgE- and IgG-specific immunecomplex–mediated FcR activation in primary humanmast cells and macrophages in vitro and in mouseasthma and collagen antibody–induced arthritis modelsand rat collagen-induced arthritis models in vivo (17,28–30). Here we show inhibition of FcR signaling in theArthus model following the administration of R788 inlupus-prone mice, similar to results obtained in micepretreated with FcR-blocking antibody.

An unexpected requirement for higher doses ofR788 and higher plasma R406 exposures was observed,compared with that needed for clinical efficacy in thelupus model. These results may be attributable to theacute nature of the Arthus model or the presence ofactive inflammation at the time of the assay. In supportof this hypothesis, NZB/NZW mice often demonstratedgreater baseline stimulation and more variability than

BALB/c mice, despite receiving the same doses ofantigen and antibody. In fact, in NZB/NZW mice, 50%less OVA-specific IgG was needed to achieve stimula-tion similar to that in BALB/c or NZB and NZWparental strains, with edema observed as early as 2 hourspost-challenge in NZB/NZW mice (data not shown). Analternate hypothesis is that targets other than FcR arepartially responsible for efficacy, and mechanisms yet tobe discovered may contribute to renal disease via Syk-mediated signaling. For example, a novel pathway wasrecently identified in fibroblast-like synoviocytes isolatedfrom patients with rheumatoid arthritis, in which R406was shown to inhibit tumor necrosis factor–inducedcytokine release via Syk inhibition (31), suggesting mul-tiple mechanisms by which R788 modulates inflamma-tory disorders.

R788 treatment resulted in reductions in overallspleen cellularity as well as reduced frequencies ofvarious splenic B cell and T cell subpopulations. In ourstudies, mice with renal disease typically had largerspleens than nondiseased mice, regardless of treatmentgroup. Increases in the total numbers of B cells and Tcells, germinal center class-switched B cells, CD4�CD44�CD62L� effector cells, as well as T cells and Bcells expressing CD69, have also been observed in aginglupus-prone mice (32). R788 treatment reduced all ofthese disease-related populations while maintaining thenumbers of naive T cells and total CD8� T cells.Preservation of CD8� T cell numbers may be important

Table 3. Effect of R788 on the Arthus response in normal and lupus-prone strains*

Treatment

BALB/c (NZB � NZW)F1

Edema area, mm2Dye extravasation,

OD610 Edema area, mm2Dye extravasation,

OD610

Vehicle 90.8 � 1.8 0.164 � 0.009 136.3 � 28.6 0.262 � 0.06R788

10 mg/kg 73.3 � 12.8 0.089 � 0.02† ND ND20 mg/kg 38.9 � 11.0† 0.032 � 0.009† 82.5 � 23.8 0.169 � 0.0340 mg/kg 20.3 � 14.0‡ 0.019 � 0.009‡ 77.6 � 7.6 0.163 � 0.0380 mg/kg ND ND 46.8 � 9.0 0.110 � 0.03†

Normal saline, 0.9% ND ND 119.1 � 4.9 0.222 � 0.004Fc Block

50 �g 41.0 � 6.2† 0.040 � 0.02† 36.8 � 4.1‡ 0.074 � 0.02‡100 �g ND ND 3.6 � 3.6§ 0.011 � 0.009§150 �g ND ND 18.3 � 4.7§ 0.029 � 0.01§

* Values are mean � SEM results from 6–8 mice/group for Fc Block and from 5–6 mice per group for R788. Fc Block wasadministered once intraperitoneally in a 200-�l volume. Studies using R788 were performed separately from studies using FcBlock in (NZB � NZW)F1 mice. In a separate study (data not shown), isotype control rat anti-mouse IgG2b antibody (noazide/low endotoxin) at a dose of 100 �g showed no effect compared with vehicle. OD610 � optical density at 610 nm; ND �not determined.† P � 0.05 versus vehicle.‡ P � 0.001 versus vehicle.§ P � 0.001 versus vehicle; P � 0.05 versus 50 �g Fc Block.

NOVEL Syk-SELECTIVE INHIBITOR PREVENTS AND TREATS MURINE LUPUS NEPHRITIS 1441

given the recent studies demonstrating the generation ofFoxp3-expressing CD8� T cells and the therapeuticpotential of CD8-mediated suppression in the NZB/NZW model (33).

Activation of splenic B cells from BALB/c micestimulated in vitro with anti-IgM was reduced by �50%at a concentration of 1 �M R406, which approximatesthe in vivo plasma drug concentrations achievable at adose of 40 mg/kg of R788; therefore, inhibition of B cellreceptor signaling would be anticipated in our lupusstudies. The global reduction of B cell populationsobserved in the spleen suggested drug-related effects,yet only modest changes in total dsDNA autoantibodytiters were apparent with prolonged drug treatmentcompared with vehicle treatment. The pharmacokineticsof R788 could be responsible for this differential effecton B cells in vivo, despite potent in vitro inhibition of Bcell signaling at relevant concentrations in vitro. Sus-tained drug exposures may be necessary for effects onautoantibody generation but not for global B cell effects.Despite the presence of autoantibodies, R788 treatmentwas efficacious, revealing a lack of correlation betweenautoantibody titers and renal disease, consistent withpublished studies showing that FcR�-deficient NZB/NZW mice generated and deposited immune complexbut did not develop severe nephritis (4). These datasuggest that R788 treatment may result in efficacy viainhibition of renal damage caused by events downstreamof immune complex deposition.

Immune complexes can be internalized by den-dritic cells (DCs) via FcR, leading to Syk-dependentmaturation and antigen presentation (34,35). Althoughquite Syk-selective, R406 also inhibits flt-3 signaling invitro (17), and R788 inhibits flt-3 in vivo (36). Flt-3 is apotent hematopoietic growth factor known to promotethe differentiation and mobilization of plasmacytoidDCs (37), a cell population recently postulated to play arole in lupus (38–42). Prolonged flt-3 inhibition couldresult in reduced numbers of DCs due to the require-ment of this growth factor for differentiation and ho-meostasis (37,43); therefore, the efficacy we observed inlupus mice could be partially attributable to inhibition offlt-3. Consistent with this hypothesis, R788 treatmentreduced CD11c� infiltrates in the kidneys of lupus miceand reduced the numbers of CD19�/class II MHChigh

cells in the spleen (Table 2).Disease in mice depends on CD4� T cells, as

evidenced by the fact that anti-CD4 monoclonal anti-body treatment ameliorates lupus in mice (44), in whichautoreactive T cells appear at �4 months of age, pre-ceding the appearance of autoantibody-secreting B cells

(45). Hyperactive T cells have been described in humanSLE, and this phenotype was recently attributed to analternate T cell receptor (TCR)–signaling pathway in-volving FcR� and Syk (46,47). These data imply thatR788 could have direct effects specifically on lupus Tcells, which may not signal through the canonical TCRpathway involving ZAP-70 (48). Although the involve-ment of Syk in murine T cell signaling has not beendemonstrated, T cells from lupus-prone mice demon-strate diminished interleukin-2 levels, consistent withthe human phenotype (46,48,49).

Initial results from phase II trials in patients withimmune thrombocytopenic purpura showed that R788(fostamatinib disodium) increased platelet counts in amajority of the patients treated (50). Additionally, amulticenter phase II study to evaluate the safety andefficacy of R788 for the treatment of patients withrelapsed and refractory B cell non-Hodgkin’s lymphomawas initiated this year, with recent in vitro data demon-strating that R406 selectively kills primary B cell tumorsvia Syk inhibition (51). Certainly, the aberrant immuno-phenotypes identified in patients with SLE and lupus-prone mice suggest that Syk-dependent signaling maycontribute to the many events leading to end-organdamage and disease in SLE, and treatment with R788may ameliorate such conditions. We have demonstratedthat in vivo Syk inhibition delayed disease progressionand prolonged survival in a murine model of lupus,suggesting that further clinical development of R788 forthe treatment of human SLE is warranted.

AUTHOR CONTRIBUTIONS

Dr. Daikh had full access to all of the data in the study andtakes responsibility for the integrity of the data and the accuracy of thedata analysis.Study design. Bahjat, Pine, Baluom, Chang, Grossbard, Daikh.Acquisition of data. Bahjat, Reitsma, Cassafer, Baluom, Grillo, Chang,Zhao, Daikh.Analysis and interpretation of data. Bahjat, Pine, Reitsma, Baluom,Chang, Zhao, Payan, Daikh.Manuscript preparation. Bahjat, Pine, Reitsma, Baluom, Chang,Payan, Grossbard, Daikh.Statistical analysis. Bahjat, Pine, Reitsma, Baluom, Daikh.

REFERENCES

1. Edberg JC, Langefeld CD, Wu J, Moser KL, Kaufman KM, KellyJ, et al. Genetic linkage and association of Fc� receptor IIIA(CD16A) on chromosome 1q23 with human systemic lupus ery-thematosus. Arthritis Rheum 2002;46:2132–40.

2. Salmon JE, Millard S, Schachter LA, Arnett FC, Ginzler EM,Gourley MF, et al. Fc � RIIA alleles are heritable risk factors forlupus nephritis in African Americans. J Clin Invest 1996;97:1348–54.

3. Wu J, Edberg JC, Redecha PB, Bansal V, Guyre PM, Coleman K,

1442 BAHJAT ET AL

et al. A novel polymorphism of Fc�RIIIa (CD16) alters receptorfunction and predisposes to autoimmune disease. J Clin Invest1997;100:1059–70.

4. Clynes R, Dumitru C, Ravetch JV. Uncoupling of immune com-plex formation and kidney damage in autoimmune glomerulone-phritis. Science 1998;279:1052–4.

5. Rao S, Olson JM, Moser KL, Gray-McGuire C, Bruner GR, KellyJ, et al. Linkage analysis of human systemic lupus erythematosus–related traits: a principal component approach. Arthritis Rheum2001;44:2807–18.

6. Ahmad YA, Bruce IN. Genetic epidemiology: systemic lupuserythematosus [review]. Arthritis Res 2001;3:331–6.

7. Morel L, Mohan C, Yu Y, Croker BP, Tian N, Deng A, et al.Functional dissection of systemic lupus erythematosus using con-genic mouse strains. J Immunol 1997;158:6019–28.

8. Morel L, Blenman KR, Croker BP, Wakeland EK. The majormurine systemic lupus erythematosus susceptibility locus, Sle1, is acluster of functionally related genes. Proc Natl Acad Sci U S A2001;98:1787–92.

9. Mohan C, Morel L, Yang P, Watanabe H, Croker B, Gilkeson G,et al. Genetic dissection of lupus pathogenesis: a recipe fornephrophilic autoantibodies. J Clin Invest 1999;103:1685–95.

10. Clynes R, Ravetch JV. Cytotoxic antibodies trigger inflammationthrough Fc receptors. Immunity 1995;3:21–6.

11. Park SY, Ueda S, Ohno H, Hamano Y, Tanaka M, Shiratori T,et al. Resistance of Fc receptor-deficient mice to fatal glomerulo-nephritis. J Clin Invest 1998;102:1229–38.

12. Tarzi RM, Davies KA, Robson MG, Fossati-Jimack L, Saito T,Walport MJ, et al. Nephrotoxic nephritis is mediated by Fc�receptors on circulating leukocytes and not intrinsic renal cells.Kidney Int 2002;62:2087–96.

13. Bergtold A, Gavhane A, D’Agati V, Madaio M, Clynes R.FcR-bearing myeloid cells are responsible for triggering murinelupus nephritis. J Immunol 2006;177:7287–95.

14. Watanabe H, Sherris D, Gilkeson GS. Soluble CD16 in thetreatment of murine lupus nephritis. Clin Immunol Immunopathol1998;88:91–5.

15. Koretzky GA, Boerth NJ. The role of adapter proteins in T cellactivation [review]. Cell Mol Life Sci 1999;56:1048–60.

16. Cohen-Solal JF, Cassard L, Fridman WH, Sautes-Fridman C. Fc �receptors [review]. Immunol Lett 2004;92:199–205.

17. Braselmann S, Taylor V, Zhao H, Wang S, Sylvain C, Baloum M,et al. R406, an orally available spleen tyrosine kinase inhibitorblocks fc receptor signaling and reduces immune complex-medi-ated inflammation. J Pharmacol Exp Ther 2006;319:998–1008.

18. Cunnane G, Chan OT, Cassafer G, Brindis S, Kaufman E, Yen TS,et al. Prevention of renal damage in murine lupus nephritis byCTLA-4Ig and cyclophosphamide. Arthritis Rheum 2004;50:1539–48.

19. Crow AR, Siragam V, Freedman J, Grossbard EB, Masuda ES,Pine P, et al. Inhibition of immune thrombocytopenic purpura(ITP) by an orally bioavailable inhibitor of Syk kinase activity[abstract]. Blood 2005;106:2165.

20. You S, Slehoffer G, Barriot S, Bach JF, Chatenoud L. Unique roleof CD4�CD62L� regulatory T cells in the control of autoimmunediabetes in T cell receptor transgenic mice. Proc Natl Acad SciU S A 2004;101 Suppl 2:14580–5.

21. Wolf D, Hochegger K, Wolf AM, Rumpold HF, Gastl G, Tilg H,et al. CD4�CD25� regulatory T cells inhibit experimental anti-glomerular basement membrane glomerulonephritis in mice. J AmSoc Nephrol 2005;16:1360–70.

22. Szanya V, Ermann J, Taylor C, Holness C, Fathman CG. Thesubpopulation of CD4�CD25� splenocytes that delays adoptivetransfer of diabetes expresses L-selectin and high levels of CCR7.J Immunol 2002;169:2461–5.

23. Fu S, Yopp AC, Mao X, Chen D, Zhang N, Chen D, et al. CD4�

CD25� CD62� T-regulatory cell subset has optimal suppressiveand proliferative potential. Am J Transplant 2004;4:65–78.

24. Scalapino KJ, Tang Q, Bluestone JA, Bonyhadi ML, Daikh DI.Suppression of disease in New Zealand Black/New Zealand Whitelupus-prone mice by adoptive transfer of ex vivo expanded regu-latory T cells. J Immunol 2006;177:1451–9.

25. Sylvestre D, Clynes R, Ma M, Warren H, Carroll MC, Ravetch JV.Immunoglobulin G-mediated inflammatory responses developnormally in complement-deficient mice. J Exp Med 1996;184:2385–92.

26. Takai T, Li M, Sylvestre D, Clynes R, Ravetch JV. FcR � chaindeletion results in pleiotrophic effector cell defects. Cell 1994;76:519–29.

27. Sylvestre DL, Ravetch JV. Fc receptors initiate the Arthus reac-tion: redefining the inflammatory cascade. Science 1994;265:1095–8.

28. Matsubara S, Li G, Takeda K, Loader JE, Pine P, Masuda ES, etal. Inhibition of spleen tyrosine kinase prevents mast cell activa-tion and airway hyperresponsiveness. Am J Respir Crit Care Med2006;173:56–63.

29. Matsubara S, Koya T, Takeda K, Joetham A, Miyahara N, Pine P,et al. Syk activation in dendritic cells is essential for airwayhyperresponsiveness and inflammation. Am J Respir Cell Mol Biol2006;34:426–33.

30. Pine PR, Chang B, Schoettler N, Banquerigo ML, Wang S, Lau A,et al. Inflammation and bone erosion are suppressed in models ofrheumatoid arthritis following treatment with a novel Syk inhibi-tor. Clin Immunol 2007;124:244–57.

31. Cha HS, Boyle DL, Inoue T, Schoot R, Tak PP, Pine P, et al. Anovel spleen tyrosine kinase inhibitor blocks c-Jun N-terminalkinase-mediated gene expression in synoviocytes. J Pharmacol ExpTher 2006;317:571–8.

32. Ramanujam M, Wang X, Huang W, Liu Z, Schiffer L, Tao H, etal. Similarities and differences between selective and nonselectiveBAFF blockade in murine SLE. J Clin Invest 2006;116:724–34.

33. Singh RP, La Cava A, Wong M, Ebling F, Hahn BH. CD8� Tcell-mediated suppression of autoimmunity in a murine lupusmodel of peptide-induced immune tolerance depends on Foxp3expression. J Immunol 2007;178:7649–57.

34. Nakashima K, Kokubo T, Shichijo M, Li YF, Yura T, YamamotoN. A novel Syk kinase-selective inhibitor blocks antigen presenta-tion of immune complexes in dendritic cells. Eur J Pharmacol2004;505:223–8.

35. Sedlik C, Orbach D, Veron P, Schweighoffer E, Colucci F,Gamberale R, et al. A critical role for Syk protein tyrosine kinasein Fc receptor-mediated antigen presentation and induction ofdendritic cell maturation. J Immunol 2003;170:846–52.

36. Pine PR, Bahjat R, Chang B, Taylor V, Markovstov V, Hitoshi Y,et al. An orally bioavailable inhibitor of FLT3 and Syk kinasesprevents tumor growth in subcutaneously implanted human tumorxenografts and promotes cell death of FLT3 mutant AML cells[abstract]. Blood 2005;106:243.

37. Brawand P, Fitzpatrick DR, Greenfield BW, Brasel K, Malisze-wski CR, De Smedt T. Murine plasmacytoid pre-dendritic cellsgenerated from Flt3 ligand-supplemented bone marrow culturesare immature APCs. J Immunol 2002;169:6711–9.

38. Ronnblom L, Eloranta ML, Alm GV. Role of natural interferon-�producing cells (plasmacytoid dendritic cells) in autoimmunity[review]. Autoimmunity 2003;36:463–72.

39. Ioannou Y, Isenberg DA. Current evidence for the induction ofautoimmune rheumatic manifestations by cytokine therapy [re-view]. Arthritis Rheum 2000;43:1431–42.

40. Preble OT, Black RJ, Friedman RM, Klippel JH, Vilcek J.Systemic lupus erythematosus: presence in human serum of anunusual acid-labile leukocyte interferon. Science 1982;216:429–31.

41. Mathian A, Weinberg A, Gallegos M, Banchereau J, Koutouzov S.IFN-� induces early lethal lupus in preautoimmune (New Zealand

NOVEL Syk-SELECTIVE INHIBITOR PREVENTS AND TREATS MURINE LUPUS NEPHRITIS 1443

Black x New Zealand White) F1 but not in BALB/c mice.J Immunol 2005;174:2499–506.

42. Santiago-Raber ML, Baccala R, Haraldsson KM, Choubey D,Stewart TA, Kono DH, et al. Type-I interferon receptor deficiencyreduces lupus-like disease in NZB mice. J Exp Med 2003;197:777–88.

43. Whartenby KA, Calabresi PA, McCadden E, Nguyen B, KardianD, Wang T, et al. Inhibition of FLT3 signaling targets DCs toameliorate autoimmune disease. Proc Natl Acad Sci U S A2005;102:16741–6.

44. Wofsy D, Seaman WE. Successful treatment of autoimmunity inNZB/NZW F1 mice with monoclonal antibody to L3T4. J Exp Med1985;161:378–91.

45. Peng SL, Craft J. The regulation of murine lupus [review]. Ann NY Acad Sci 1997;815:128–38.

46. Krishnan S, Warke VG, Nambiar MP, Tsokos GC, Farber DL.The FcR � subunit and Syk kinase replace the CD3 �-chain andZAP-70 kinase in the TCR signaling complex of human effectorCD4 T cells. J Immunol 2003;170:4189–95.

47. Krishnan S, Farber DL, Tsokos GC. T cell rewiring in differenti-ation and disease [review]. J Immunol 2003;171:3325–31.

48. Nambiar MP, Fisher CU, Kumar A, Tsokos CG, Warke VG,Tsokos GC. Forced expression of the Fc receptor �-chain rendershuman T cells hyperresponsive to TCR/CD3 stimulation. J Immu-nol 2003;170:2871–6.

49. Liossis SN, Ding XZ, Dennis GJ, Tsokos GC. Altered pattern ofTCR/CD3-mediated protein-tyrosyl phosphorylation in T cellsfrom patients with systemic lupus erythematosus: deficient expres-sion of the T cell receptor zeta chain. J Clin Invest 1998;101:1448–57.

50. Bussel JB, Schindler AM, Grossbard EB. R935788: a phase II,single center, open label, efficacy and safety, ascending dose, pilotstudy for the treatment of adult immune thrombocytopenic pur-pura (ITP) [abstract]. Blood 2007;110:1310.

51. Sharman JP, Irish J, Hatton O, Martinez O, Levy R. Syk kinaseinhibition by R406 inhibits BCR mediated B cell proliferation andinterferes with BCR signal transduction in primary tumor samples[abstract]. Blood 2007;110:802.

1444 BAHJAT ET AL