Suppression of Glomerulonephritis in Lupus-Prone NZB × NZW Mice by RN486, a Selective Inhibitor of Bruton's Tyrosine Kinase




Bruton's tyrosine kinase (BTK) plays a critical role in B cell development and function. We recently described a selective BTK inhibitor, RN486, that blocks B cell receptor (BCR) and Fcγ receptor signaling and is efficacious in animal models of arthritis. The aim of this study was to examine the potential efficacy of BTK in systemic lupus erythematosus (SLE), using an NZB × NZW mouse model of spontaneous SLE.


Mice received RN486 or its vehicle (administered in chow) at a final concentration of 30 mg/kg for 8 weeks, starting at 32 weeks of age.


The administration of RN486 completely stopped disease progression, as determined by histologic and functional analyses of glomerular nephritis. The efficacy was associated with striking inhibition of B cell activation, as demonstrated by a significant reduction in CD69 expression in response to BCR crosslinking. RN486 markedly reduced the secretion of IgG anti–double-stranded DNA (anti-dsDNA) secretion, as determined by enzyme-linked immunosorbent and enzyme-linked immunospot assays. Flow cytometric analysis demonstrated depletion of CD138highB220low plasma cells in the spleen. RN486 inhibited secretion of IgG anti-dsDNA but not IgM anti-dsDNA, suggesting that pharmacologic blockade of BTK resembles the reported transgenic expression of low levels of endogenous BTK in B cells. In addition, RN486 may also impact the effector function of autoantibodies, as evidenced by a significant reduction in immune complex–mediated activation of human monocytes in vitro and down-regulation of the expression of macrophage-related and interferon-inducible genes in both the kidneys and spleens of treated mice.


Collectively, our data suggest that BTK inhibitors may simultaneously target autoantibody-producing and effector cells in SLE, thus constituting a promising therapeutic alternative for this disease.

Systemic lupus erythematosus (SLE) is an extremely heterogeneous disease. Among its numerous manifestations, lupus nephritis is the most common. Lupus nephritis is present in more than half of the patient population and significantly reduces survival rates, particularly in African Americans ([1, 2]).

The role of B cells in the pathogenesis of SLE has been repeatedly demonstrated ([3]). Most novel treatments of SLE target B cells, and the results of such treatment have been marginal yet encouraging so far ([4-6]). Recommendations for the management of SLE considering such novel therapeutic approaches were recently published by the American College of Rheumatology, as well as the Joint European League Against Rheumatism and European Renal Association–European Dialysis and Transplant Association ([7, 8]), highlighting the fact that no data are yet available for lupus nephritis, and that further research addressing this unmet medical need is required.

Bruton's tyrosine kinase (BTK) is a Tec family kinase that plays a critical role in B cell development and function ([9]). Although BTK is expressed by multiple cell types, including macrophages, B cells (particularly mature B cells) are unique in that they rely on BTK signaling for survival ([10]). BTK deficiency in mice and humans leads to X-linked immunodeficiency and X-linked agammaglobulinemia, respectively ([11]), and reduced responsiveness to B cell receptor (BCR) ([12]), Toll-like receptor (TLR) ([13, 14]), Fcγ receptor (FcγR), and chemokine receptor stimulation ([15, 16]). With respect to autoimmunity, it was recently reported that BTK controls the autoreactive phenotype of B cells in mice, that its overexpression in vivo leads to systemic autoimmunity, and that no anti–double-stranded DNA (anti-dsDNA) antibodies are present in BTK-deficient mice ([12, 17]), thus making BTK an attractive target for the treatment of lupus nephritis.

Therefore, we investigated the therapeutic effect of 6-cyclopropyl-8-fluoro-2-(2-hydroxymethyl-3-{1-methyl-5-[5-(4-methyl-piperazin-1-yl)-pyridin-2-ylamino]-6-oxo-1,6-dihydro-pyridin-3-yl}-phenyl)-2H-isoquinolin-1-one (RN486), a highly selective inhibitor of BTK, on disease progression in NZB × NZW (NZB/NZW) mice. In NZB/NZW mice, a lupus-like syndrome that is similar to human SLE develops spontaneously. These mice express susceptibility genes that have been linked to human lupus, develop serologic abnormalities such as autoantibodies to dsDNA and other nuclear antigens, and present with immune complex deposition and progressive proteinuria that lead to lethal kidney pathology. Additionally, disease development is markedly accelerated in female NZB/NZW mice compared with their male littermates ([18, 19]). NZB/NZW mice have intrinsic B cell defects, and the increased number of antibody-secreting cells (ASCs) has been attributed to increased plasma cell survival in the spleen accompanied by decreased migration of plasma cells to the bone marrow ([20, 21]). Treatment with a B cell–depleting agent, when initiated at symptom onset, prolongs survival and delays the appearance of proteinuria in this model ([22]). At the renal level, however, it is clear that mechanisms beyond B cell alterations play a key role in end-organ damage (including most notably mononuclear cell infiltration and activation), which is associated with poor outcomes in human and mouse models ([19, 23-27]) and has the potential of being affected by BTK inhibition.

Although BTK blockade has not yet been tested in human lupus, its beneficial effect in the MRL mouse model of lupus has already been reported ([28]), even though the mechanism of action was not explored in detail. In the current study, we reproduce that finding in the NZB/NZW mouse strain and expand the observations, demonstrating that the efficacy of BTK blockade depends on its ability to deplete splenic plasma cells (CD138+B220low). Splenic plasma cells have been shown to be the major producers of anti-dsDNA antibodies in NZB/NZW mice ([20]) and to decrease the expression of renal macrophage and interferon (IFN)–related genes, suggesting that reduced kidney macrophage accumulation and activation are attributable to impaired antibody effector function, which we demonstrate in vitro. We further demonstrate impairment in immune complex–mediated activation, using human monocytes treated with BTK, and decreased macrophage and T cell infiltration into the kidneys.


In vivo studies

All in vivo procedures were approved by the Hoffmann-La Roche Institutional Animal Care and Use Committee and were performed according to the 1996 Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources). Eight-week-old female NZB/NZW mice (The Jackson Laboratory) were acclimated for a minimum of 3 days. Body weight, proteinuria, and serum anti-dsDNA antibody levels were determined once every other week, starting at 18 weeks of age. At 32 weeks of age, the mice were randomized according to their proteinuria scores to receive either regular chow (n = 12) or chow containing RN486 formulated at a final concentration equivalent to 30 mg/kg/day (n = 12) (Research Diets). This amount is the equivalent of 0.225 mg/gm of chow and was previously shown to be adequate for inhibiting BTK-dependent CD69 expression ([16]). Two mice in the vehicle group and 1 mouse in the treatment group died before the study was terminated; therefore, data obtained 8 weeks after enrollment correspond to 10 mice in the vehicle group and 11 mice in the treatment group.

Evaluation of proteinuria and urine chemistry

Proteinuria was evaluated weekly using Multistix 10 SG reagent strips (Siemens) and scored as follows: 0 = trace, 1 = 30 mg/dl, 2 = 100 mg/dl, 3 = 300 mg/dl, and 4 = ≥2,000 mg/dl. These data were confirmed at the end of the study by determining total urinary protein, using an automated turbidimetric method in 24-hour urine samples.

Histologic and immunohistochemical analyses

Sections (4 μm) were cut from paraffin-embedded kidney tissue, fixed in formalin, and stained with hematoxylin and eosin, periodic acid–Schiff, or Masson's trichrome. The sections were graded by a board-certified pathologist (NK) for active and chronic renal lesions, using a 0–5 scale reflecting the severity and extent of lesions, as follows: 0 = no abnormality, 1+ = minimal, 2+ = slight, 3+ = mild, 4+ = moderate, and 5+ = marked. The scoring method (modified from ref.[29]) was used to evaluate the histologic changes most commonly associated with lupus nephritis, including glomerular lesions (glomerular cellularity, necrosis, and glomerulosclerosis), tubulointerstitial lesions (tubular basophilia, tubular atrophy, protein casts, interstitial fibrosis), inflammatory infiltrates (perivascular, periglomerular, or peripelvic), and arterial or arteriolar vasculitis. Three scores (glomerular activity score, interstitial activity score, and inflammatory activity score) were compiled to determine a total score for renal pathology.

Immunohistochemistry was performed on frozen kidney sections using goat polyclonal antibodies for IgG (Jackson ImmunoResearch), IgM (Abcam), and C3 F(ab′)2 (MP Biomedicals), rat monoclonal antibody for F4/80 (Serotec) and IgD (BD Biosciences), or rabbit monoclonal antibody for CD3 (Abcam). The tissue specimens were counterstained with hematoxylin. Immunolabeled sections were scanned at 20× magnification, using an Aperio ScanScope Slide Scanner, and areas positive for IgG and other proteins were semiautomatically quantified using Definiens Tissue Studio software. Data are expressed as the percentage of the positively labeled area (area of diaminobenzidine labeling/tissue reference area).

Anti-dsDNA antibody enzyme-linked immunosorbent assay (ELISA).

Anti-dsDNA was analyzed by ELISA, using a commercially available kit (Shibayagi). Briefly, serial dilutions of serum were incubated on dsDNA-coated ELISA plates for 2 hours at room temperature. After washing, horseradish peroxidase–conjugated goat anti-mouse IgG (Shibayagi) or IgM (SouthernBiotech) was added for 2 hours at room temperature. Tetramethylbenzidine was used for detection. All analyses were performed in duplicate, and the concentrations of IgG anti-dsDNA antibodies are expressed as mU/ml.

Enzyme-linked immunospot (ELISpot) assay

Ninety-six–well plates (BD Biosciences) were coated with 100 μg/ml of sonicated calf thymus DNA (Sigma-Aldrich) in phosphate buffered saline (PBS) and blocked with 5% fetal bovine serum (FBS) in PBS for 90 minutes. Splenocytes or bone marrow cells were isolated and plated in serial dilutions overnight at 37°C. Plates were washed with PBS/0.01% Tween 20 followed by distilled water to ensure complete cell lysis. Biotinylated anti-mouse IgM or IgG antibody (1:1,000 dilution in 1% FBS in PBS) was added and incubated for 3 hours at room temperature. After washing, 100 μl/well of a solution of alkaline phosphatase–conjugated streptavidin (R&D Systems) in 1% FBS in PBS was added and incubated for 2 hours at room temperature. A BCIP/nitroblue tetrazolium chromogen solution (R&D Systems) was added and incubated at room temperature until visible spots were present. Spots were imaged and counted using a CTL-ImmunoSpot analyzer (Cellular Technology Limited) and expressed as the total number of spots/1 × 106 cells.

Induction of CD69 expression in murine whole blood

Heparinized mouse blood (50 μl/well) was incubated with a rat anti-mouse IgD antibody (F[ab′]2; Accurate) at 60 μg/ml in 50 μl/well of RPMI medium. After 4 hours, samples were labeled for flow cytometry, using the following antibodies: fluorescein isothiocyanate–conjugated CD3, allophycocyanin-conjugated CD45R (B220), and phycoerythrin (PE)–conjugated CD69 (BD Biosciences) for 30 minutes. Red blood cells were then lysed using BD FACS Lysing Solution (BD Biosciences), and the plate was washed with BD PharmingenStain Buffer (BD Biosciences). Samples were analyzed using a BD FSRFortessa flow cytometer (BD Biosciences) and FlowJo software (Tree Star).

Flow cytometric analysis

Splenocytes and bone marrow cells were isolated from freshly harvested minced spleens by passing the tissue through a nylon mesh, followed by lysis of the red blood cells using BD FACS Lysing solution (BD Biosciences) and staining with PE-conjugated anti-CD138 and Pacific blue–labeled anti-CD45R (B220) (BD Biosciences). Samples were acquired using a BD LSRFortessa flow cytometer (BD Biosciences) and analyzed with FlowJo software (Tree Star).

Immune complex–mediated stimulation of human monocytes

Human serum albumin (HSA) immune complexes (HSA–anti-HSA) were generated in high-binding 96-well flat bottomed plates by adding 200 μl/well HSA at 20 μg/ml carbonate buffer, pH 9.6, overnight at 4°C. After a 1-hour blockade in 10% ultralow IgG FBS, goat anti-HSA in PBS was added (200 μl/well of 10 μg/ml) and incubated for 30 minutes at 37°C. Stimulation was initiated by adding monocytes (freshly isolated from normal human blood by negative selection using RosetteSep [StemCell Technologies]) to plates coated with immune complexes at 0.3 × 106 monocytes/200 μl/well and incubated overnight at 37°C. BTK inhibition was achieved by preincubating monocytes with specified concentrations of RN486 in 0.1% DMSO or DMSO alone for 1 hour at 37°C before adding them to wells with immune complexes. Monocytes added to wells with HSA alone were used to determine the background response. Supernatants were collected and assessed by ELISA for tumor necrosis factor α (TNFα) release.

Cytokine ELISA

Cytokine analysis was performed using mouse interleukin-10 (IL-10), IL-12/IL-23p40, and TNFα ELISA DuoSet kits (R&D Systems) according to the manufacturer's instructions, using a 1:5 dilution of serum obtained at the time of enrollment and a 1:100 dilution of serum obtained at the time of study termination for IL-10 and IL-12p40, respectively.

Statistical analysis

All graphing and statistical analyses were performed using GraphPad Prism for Windows. P values less than 0.05 were considered significant.


Effect of RN486 treatment on lupus nephritis in NZB/NZW mice

At 32 weeks of age, when mild proteinuria (average score of 2) had developed in NZB/NZW mice, the mice were randomized into 2 groups to receive either vehicle- or RN486-formulated chow. Vehicle-treated mice showed normal progression of proteinuria scores (up to 300 mg/dl) until the time at which they were killed. In contrast, RN486-treated mice did not display further progression of proteinuria, and proteinuria levels in the treatment group were significantly lower at the end of the study than at the time of enrollment (Figure 1A, left). These results were confirmed by analysis of total urinary protein, using an automated urine chemistry method at the end of the study (Figure 1A, right).

Figure 1.

RN486 suppresses nephritis in NZB × NZW (NZB/NZW) mice. NZB/NZW mice (32 weeks old) were randomized into 2 groups of at least 10 mice each. The first group received 30 mg/kg of RN486 in chow ad libitum, and the second group received vehicle-containing chow for 8 weeks. A, Left, Proteinuria scores as measured by dipstick at the indicated time points. ∗ = P < 0.05; ∗∗ = P < 0.01; ∗∗∗ = P < 0.001 versus vehicle, by two-way analysis of variance. Right, Total urinary protein as determined using an automated turbidimetric method, in 24-hour urine samples at week 8. Each data point represents a single mouse; horizontal lines show the mean. B, Kidney sections from vehicle-treated and RN486-treated mice, showing histologic differences between the 2 groups. Hematoxylin and eosin (H&E) staining revealed an uneven renal cortical surface with enlarged glomeruli, tubular basophilia, distension of tubular lumina, protein casts, and inflammatory infiltrates in vehicle-treated mice. Periodic acid–Schiff (PAS) staining confirmed the expansion of glomeruli in vehicle-treated mice, with epithelial or endothelial deposits and cellular infiltrates. IgG, IgM, and C3 immunohistochemistry demonstrated little to no glomerular immunolabeling in RN486-treated mice compared with vehicle-treated mice. Macrophage infiltration and T cell infiltration were analyzed using F4/80 and CD3 antibodies, respectively. C, Semiautomated quantification of IgG immunohistochemistry findings (left) and histopathology scores (right). Values are the mean ± SEM. In A (right) and C, each data point represents a single mouse; horizontal lines show the mean.

Figure 1B illustrates the histologic differences between kidney sections from vehicle-treated mice and RN486-treated mice. On average, the severity and extent of lesions (including glomerulosclerosis and IgG, IgM, and C3 glomerular deposition) were greater in vehicle-treated mice compared with RN486-treated mice. For example, 9 of 10 mice receiving vehicle had moderate-to-marked glomerular IgG deposits, whereas only 2 of 11 mice receiving RN486 had similar severity scores. On the opposite end of the spectrum, only 1 of 10 vehicle-treated mice had slight IgG deposits, compared with 5 of 11 RN486-treated mice with either no or slight IgG renal glomerular deposits.

Semiautomated analysis of the sections was performed to corroborate the histopathology scores, antibody deposition, and cellular infiltration (Figure 1C). This analysis demonstrated significant differences between vehicle-treated and RN486-treated mice (for IgG, mean ± SD 3.632 ± 0.68 versus 0.086 ± 0.030 [P = 0.0006] [Figure 1C, left]; for IgM, 4.2 ± 0.7 versus 1.9 ± 0.5 [P = 0.0121]; for C3, 5.3 ± 1.0 versus 1.0 ± 0.5 [P = 0.0009]). Consequently, the total score for renal histopathology was significantly reduced in RN486-treated mice compared with vehicle-treated mice (mean ± SEM 18.33 ± 0.781 versus 31.63 ± 1.658 [P = 0.0001]) (Figure 1C, right).

Furthermore, treatment with RN486 also resulted in a dramatic decrease in macrophage infiltration in RN486-treated mice compared with vehicle-treated mice, as determined using F4/80 antibodies (mean ± SEM 5.0 ± 1.0% versus 1.8 ± 1.0%; P = 0.02). The histopathologic examination of vehicle-treated kidney sections showed that periglomerular immunolabeling of macrophages was much more prominent than tubulointerstitial immunolabeling. In RN486-treated mice, there were few periglomerular F4/80-positive macrophages but only individual cell labeling within the cortical and medullary interstitium as well as the renal pelvis (results not shown). With regard to T cell infiltration, although there was specific and distinct CD3+ immunolabeling of round cells within glomeruli, free in the cortical and medullary interstitium and especially surrounding blood vessels in the renal pelvis, the intensity of labeling was weak, precluding semiautomated quantification to corroborate the histopathologic examination. However, using a grading scale of 1–5, similar to that used for a regulatory assessment (where 1 = minimal, 2 = slight, 3 = mild, 4 = moderate, and 5 = marked), 8 of 10 vehicle-treated mice were graded as ≥3 for CD3 glomerular immunolabeling, compared with only 3 of 11 RN486-treated mice.

RN486-induced inhibition of B cell activation and diminished IgG anti-dsDNA antibody secretion

Using BCR-dependent CD69 expression as a pharmacodynamic marker as previously described ([16]), we demonstrated that when RN486 was used at a dose that was effective for inhibiting the development of proteinuria, an average plasma level of 434.4 ng/ml was achieved (Figure 2A, left). At this level, RN486 completely inhibited the induction of CD69 expression on CD45R+ (B220) cells in response to BCR crosslinking with anti-IgD, as determined by flow cytometry (Figure 2A, right), which is indicative of robust B cell suppression.

Figure 2.

RN486 inhibits B cell activation and autoantibody secretion in mice. A, Left, Level of RN486 that is effective for inhibiting the development of proteinuria. Right, Complete inhibition of CD69 expression on B220 (CD45R+) cells in response to B cell receptor crosslinking with anti-IgD, as determined by flow cytometry. Symbols represent individual data points; horizontal lines show the mean. B and C, Left, Serum concentrations of IgG (B) and IgM (C) anti–double-stranded DNA (anti-dsDNA) in mice treated with vehicle or RN486. Middle, Numbers of IgG (B) and IgM (C) anti-dsDNA antibody-secreting cells in the spleen. Right, total numbers of IgG (B) and IgM (C) antibody–secreting cells. Values are the mean ± SEM. In A, B (middle and right), and C (middle and right), each symbol represents a single mouse; horizontal lines show the mean. ∗ = P < 0.05; ∗∗∗ = P < 0.001 versus vehicle. ELISpot = enzyme-linked immunospot.

To examine the impact of RN486-induced B cell suppression on the production of pathogenic autoantibodies, we evaluated the concentrations of anti-dsDNA antibodies in blood as well as anti-dsDNA–producing cells in the spleen and bone marrow. Consistent with the proteinuria data, treatment with RN486 completely prevented further increases in anti-dsDNA antibody levels, while vehicle-treated mice showed a typical progressive increase in anti-DNA antibody levels. The serum concentrations of IgG anti-dsDNA became significantly different between the 2 groups beginning at week 6 after enrollment (Figure 2B, left). In line with this result, RN486 dramatically reduced the number of IgG anti-dsDNA ASCs in the spleen, as determined by ELISpot assay using total splenocytes obtained at the end of the study (Figure 2B, middle), without changing the total number of cells secreting IgG antibodies (Figure 2B, right). In contrast, RN486 failed to modify serum levels of IgM anti-dsDNA antibodies or the number of splenic IgM anti-dsDNA ASCs (Figure 2C, left and middle). Similarly, the number of total IgM ASCs or the corresponding serum titers were not altered by treatment (Figure 2C, middle and right). As opposed to the effect of RN486 on the spleen, it had no significant impact on the number of bone marrow anti-dsDNA and total IgG and IgM ASCs (Figure 3).

Figure 3.

No significant effect of RN486 on the numbers of IgG (A) or IgM (B) anti-dsDNA antibodies or total IgG (A) or IgM (B) antibody-secreting cells in the bone marrow of mice, as determined by ELISpot assay. Each symbol represents a single mouse; horizontal lines and error bars show the mean ± SEM. See Figure 2 for definitions.

RN486-induced depletion of splenic plasma cells

In line with the ELISpot data, flow cytometric analysis of total splenocytes demonstrated a statistically significant decrease in the number of total B220+ cells, accompanied by almost complete depletion of CD138highB220low plasma cells (Figures 4A–D). This effect was not observed in bone marrow (Figures 4E and F). As expected ([30]), BAFF levels were significantly increased in the treated mice (results not shown). Plasma cell depletion was not associated with changes in the splenic follicular microarchitecture, as determined by immunohistochemistry using IgD and B220 (results not shown). The location and intensity of IgD and B220 immunolabeling were not significantly altered between treatment groups, although there was a slight decrease in the number of IgD-labeled B cells in RN486-treated spleens.

Figure 4.

RN486 induces plasma cell depletion. A and B, Percentages of B220+ cells (A) and CD138+B220low cells (B) in the spleens of mice treated with vehicle or RN486. C and D, Depletion of CD138highB220low plasma cells in RN486-treated mice (D) compared with vehicle-treated mice (C). E and F, Percentages of B220+ cells (E) and CD138+B220low cells (F) in the bone marrow of mice treated with vehicle or RN486. In A, B, E, and F, each symbol represents a single mouse; horizontal lines and error bars show the mean ± SEM.

Changes in gene expression profiles and cytokine release

In addition to its effect on B cell function, BTK inhibition may also inhibit the effector function of autoantibodies via inhibition of immune complex–mediated stimulation of FcγRs, which leads to macrophage activation and IFN-inducible gene transcription. To examine the potential relevance of BTK on these mechanisms, we performed reverse transcription–polymerase chain reaction to analyze the expression of 20 genes related to B cell differentiation, macrophage activation, and IFN-inducible transcription ([19, 24, 25, 31-34]).

In the spleen, which is an important organ involved in the development of short-lived plasma cells, RN486 down-regulated the expression of several genes related to B cell differentiation, namely, Xbp1, Blimp1, CD138, and Aid, although the down-regulation was significant only for Blimp1 and Aid. In addition, RN486 also reduced the expression of IFN-inducible genes (Figure 5A). However, consistent with previous reports, the IL-10 gene was most significantly down-regulated, and this was confirmed at the protein level by ELISA (Figure 5B). Thus, we observed that treatment with RN486 prevented an increase in IL-10 that was evident in the vehicle control group at 8 weeks postenrollment.

Figure 5.

Changes in gene expression and cytokine secretion in vivo in response to RN486. Gene expression was analyzed by real-time polymerase chain reaction. The expression levels of target genes were normalized to the levels of the reference genes 18S ribosomal RNA, β-actin, GAPDH, and β-glucuronidase. A, Expression of the indicated genes in the spleens of RN486-treated mice relative to that in the spleens of vehicle-treated mice. B, Decreased expression of the interleukin-10 (IL-10) gene, as confirmed by enzyme-linked immunosorbent assay using serum samples obtained from vehicle-treated and RN486-treated mice at the time of study entry (week 0) or termination (week 8). Each symbol represents an individual mouse; horizontal lines and error bars show the mean ± SEM. C, Expression of the indicated genes in the kidneys of RN486-treated mice relative to that in the kidneys of vehicle-treated mice. The shaded areas in A and C represent the assay detection limits. P values were determined using Student's t-test. ∗ = P < 0.05.

Compared with gene expression in the spleen, gene expression in the kidney was affected to a greater extent, especially the expression of monocyte/macrophage-related genes, including those associated with the so-called “hybrid” population of macrophages described in lupus nephritis ([24, 25]), such as Itgam (CD11b), Il34, CD16 (Fcgr1), and Cx3cr1, among others; IFN-inducible genes such as Socs3, Mx2, Isg20, Ifit2, and Oas2; genes related to kidney injury such as Lcn2 (NGAL) and Cxcl10; as well as the plasma cell marker Cd138 and Btk (Figure 5C).

RN486 regulates similar pathways in immune complex–stimulated human monocytes and NZB/NZW mice

The gene expression data obtained in mice suggested that in addition to B cell suppression, RN486 halts the progression of lupus nephritis by inhibiting the effector function of immune complexes in monocytes. To test that hypothesis, and to determine whether the same genes modulated by RN486 in the animal model are susceptible to regulation in human cells, we analyzed gene expression in peripheral blood monocytes purified from healthy donors and stimulated with immune complexes in vitro. Figure 6A shows the dose-dependent inhibition of TNFα secretion in human monocytes stimulated with HSA immune complexes, with a 50% inhibition concentration of 11 nM. These monocytes were subjected to gene expression analysis by microarray.

Figure 6.

RN486 inhibits immune complex (IC)–induced monocyte activation in human cells. A, Dose-dependent inhibition of tumor necrosis factor α (TNFα) secretion in human monocytes stimulated with human serum albumin (HSA) immune complexes, as determined by enzyme-linked immunosorbent assay. Monocytes were isolated from healthy volunteers and incubated in the presence or absence of RN486 at the indicated concentrations for 1 hour before being exposed to HSA immune complexes overnight. Bars show the mean ± SEM. IC50 = 50% inhibition concentration. B and C, Heatmaps showing genes that were modulated by immune complex stimulation of monocytes isolated from 9 subjects (S1–S9). Monocytes treated with RN486 were subjected to microarray analysis and compared with untreated and vehicle-treated monocytes from the same donors. B shows 2 sets of genes that were modulated by immune complex stimulation of monocytes. C shows genes analyzed in NZB × NZW (NZB/NZW) mice. A selected subset of differentially expressed genes containing some of the genes analyzed by reverse transcription–polymerase chain reaction in tissue from NZB/NZW mice (in addition to BTK, which was not differentially expressed) was used to plot the heatmaps. The colors shown are based on the Z score of the expression level, i.e., for the expression level of each gene, the color is determined by subtracting the mean expression value and scaled by dividing by the standard deviation.

Figure 6B represents 2 sets of genes that were modulated by immune complex stimulation of monocytes, as determined by microarray analysis. Treatment of these samples with RN486 significantly reversed that phenotype, with some variability between donors. This phenomenon was also visible when specifically looking at the genes analyzed in NZB/NZW mice, as shown in Figure 6C. Three of the genes that were most highly induced by immune complex stimulation were Ccl2 (MCP1), Itgam (CD11b), and Csf1, all of which were restored by RN486 to the level observed in untreated cells. Finally, immune complex stimulation of monocytes led to down-regulation of the IFN-inducible genes ([35]) Isg20, Oas2, Ifit2, Mx2, Irf7, and Cxcl10. Incubation with RN486 reversed this phenomenon and in some cases resulted in up-regulation of these genes, reminiscent of the results obtained in spleens from NZB/NZW mice treated with RN486.


Using RN486, a selective BTK inhibitor, we demonstrate that pharmacologic inhibition of BTK impairs the 2 most critical events in the pathogenesis of lupus nephritis, namely, autoantibody production and immune complex–dependent activation of monocytes ([23, 27, 36]), leading to suppression of lupus nephritis in NZB/NZW mice.

RN486 impairs autoantibody secretion by significantly depleting CD138+B220low cells in the spleen without significantly affecting those in bone marrow. Even though we did not directly investigate the proliferation status of this splenic plasma cell population, these CD138+B220low cells may correspond to short-lived plasma cells, which are considered to be the main producers of anti-dsDNA antibodies in NZB/NZW mice. This result is consistent with previously reported data showing that most plasma cells resistant to therapy with glucocorticoids or cyclophosphamide in NZB/NZW mice reside in bone marrow ([37]).

In contrast to its marked inhibition of IgG anti-dsDNA, RN486 spared IgM-secreting plasma cells. The number of IgM ASCs was slightly although not statistically significantly higher in mice treated with RN486 compared with vehicle-treated mice. These results are consistent with published literature showing restoration of IgM (but not IgG) anti-dsDNA secretion and increased IgM anti-dsDNA levels in mice expressing a transgene that drives 25% of endogenous BTK (BTKlow) levels in B cells in 56R.Btk−/− mice ([10, 17, 38]), suggesting that pharmacologic inhibition of BTK resembles the reported transgenic expression of low levels of BTK in B cells. A contributing explanation might be the fact that marginal zone B cells are more sensitive than follicular B cells to TLR-induced differentiation, and although BTK participates in TLR signaling, it is not required for the development of marginal zone B cells, which are the main producers of IgM ([14, 39, 40]). Also, BCR signaling, or TLR-9–induced IL-10 expression in 56R.Btk−/− mice expressing the BTKlowtransgene, is sufficient for the activation and differentiation of anti-DNA B cells, particularly IgM-producing marginal zone B cells ([17]).

Treatment with RN486 resulted in a significant decrease in IL-10 at the gene and protein levels. This cytokine is induced by TLR stimulation and is known to be modulated by BTK ([40]). The effect on IL-10 is also reminiscent of the correlation between the restoration of anti-dsDNA production in 56R.Btklowmice and the partial rescue of CpG DNA–induced IL-10 secretion ([17]) and is also consistent with other evidence demonstrating that BTK deficiency leads to decreased IL-10 secretion by multiple cell types ([41, 42]). IL-10 secretion by B cells in response to TLR-9 ligands is significantly increased in B cells from NZB/NZW mice, and antibodies against IL-10 are known to be protective in the NZB/NZW mouse model of lupus ([42]). Additionally, IL-10–deficient mice develop more severe lupus ([43]). It would be important to determine the consequences of this IL-10 depletion at earlier stages of the disease ([22]). Collectively, our findings suggest that RN486 blocks both BCR and TLR signaling in response to nucleic acid autoantigens.

The finding that the expression of some genes was altered by RN486 treatment shed some light on its mechanism of action. For instance, consistent with its potent B cell–suppressive effect in the spleen, RN486 significantly down-regulated Blimp1 and Aid, which are essential for B cell maturation and somatic hypermutation and are implicated in the loss of tolerance leading to autoimmunity. In agreement with our findings, a previous study showed that deletion of BTK attenuates the induction of Aid by TLR-7 stimulation in the presence of IL-4 ([44]). BTK has also been proposed to regulate class switching of DNA-reactive cells by regulating the TLR-9–induced expression of Aid, which is a potential additional explanation for the selective decrease in IgG production by BTK inhibition ([12]).

Finally, aside from reducing autoantibody production, RN486 treatment may also lead to depletion of ASCs in the kidney, as suggested by the down-regulation of plasma cell–related genes such as Pou2af1 and CD138 in mice treated with RN486. The significant down-regulation of several T cell–related genes and decreased T cell infiltration into the kidney point to a potential decrease in T cell assistance as one of the mechanisms behind the decreased antibody production, as a consequence of an overall decreased inflammatory phenotype.

In addition to being associated with decreased IgG deposition, a key driver of nephritis, RN486 is associated with profound inhibition of cellular infiltration and activation. This effect of RN486 is supported by the observed suppression of a wide range of inflammatory genes in the kidney. This suppression was significantly greater in magnitude than that in the spleen and is potentially mediated by inhibition of IgG formation and deposition. Because a clear pathogenic role for macrophages in kidney injury has been established, with Itgam constituting one of the most reliable predictors of lupus nephritis in the NZB/NZW mice and other lupus models ([23, 24]), this result suggests a relationship between the efficacy of RN486 and altered macrophage function ([45, 46]). In addition, most other inflammatory genes, including chemokines, cytokine signaling (SOCS3) and IFN-inducible genes (e.g., Mx2, Isg20, Ifit2, and Cxcl10), along with Lcn2 (NGAL), a biomarker for renal tissue damage, were also significantly down-regulated by RN486 in the kidney.

Finally, in addition to the decrease in IgG deposition, the efficacy of RN486 may be associated with blockade of immune complex–dependent activation of macrophages via FcγRs. In support of this hypothesis, we demonstrated that RN486 decreases the expression of numerous genes in immune complex–stimulated human monocytes isolated from healthy volunteers, most of which were also regulated by BTK inhibition in vivo in NZB/NZW mice. This is consistent with recent reports on the cross-species similarities in transcriptional networks between this animal model of lupus and the human disease ([19]) and enhances the potential translational significance of our findings.

In conclusion, we demonstrate that selective pharmacologic inhibition of BTK, which is expressed by all myeloid cells, suppresses kidney disease in NZB/NZW mice by depleting anti-dsDNA autoantibody–producing plasma cells in the spleen and blocking the effector function of immune complexes on inflammatory cells such as monocytes. Thus, the fact that BTK blockade impacts numerous cell types (e.g., B cells, plasmablasts, plasma cells, and monocytes) simultaneously might represent an advantage over other B cell–depleting therapies for the treatment of lupus nephritis, and further investigation is required.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Mina-Osorio had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Mina-Osorio, LaStant, Keirstead, Stefanova, Garrido, Giron, Kim, Patel, Woods, Ramanujam, DeMartino, Narula, Xu.

Acquisition of data. Mina-Osorio, LaStant, Keirstead, Whittard, Ayala, Stefanova, Garrido, Dimaano, Hilton, Giron, Hang, Postelnek, Kim, Min, Woods, Ramanujam, Xu.

Analysis and interpretation of data. Mina-Osorio, LaStant, Keirstead, Stefanova, Garrido, Dimaano, Giron, Lau, Patel, Woods, Ramanujam, DeMartino, Xu.


All of the authors are employees of Hoffmann-La Roche.