These authors contributed equally to this work.
Autoimmune, Cholestatic and Biliary Disease
B cell depletion therapy exacerbates murine primary biliary cirrhosis†
Article first published online: 28 DEC 2010
Copyright © 2010 American Association for the Study of Liver Diseases
Volume 53, Issue 2, pages 527–535, February 2011
How to Cite
Dhirapong, A., Lleo, A., Yang, G.-X., Tsuneyama, K., Dunn, R., Kehry, M., Packard, T. A., Cambier, J. C., Liu, F.-T., Lindor, K., Coppel, R. L., Ansari, A. A. and Gershwin, M. E. (2011), B cell depletion therapy exacerbates murine primary biliary cirrhosis. Hepatology, 53: 527–535. doi: 10.1002/hep.24044
Potential conflict of interest: Dr. Kehry owns stock and intellectual property rights in and is an employee of Biogen Idec.
- Issue published online: 27 JAN 2011
- Article first published online: 28 DEC 2010
- Accepted manuscript online: 18 OCT 2010 11:16AM EST
- Manuscript Accepted: 29 SEP 2010
- Manuscript Received: 2 APR 2010
- National Institutes of Health. Grant Number: DK067003
Primary biliary cirrhosis (PBC) is considered a model autoimmune disease due to the clinical homogeneity of patients and the classic hallmark of antimitochondrial antibodies (AMAs). Indeed, the presence of AMAs represents the most highly directed and specific autoantibody in autoimmune diseases. However, the contribution of B cells to the pathogenesis of PBC is unclear. Therefore, although AMAs appear to interact with the biliary cell apotope and contribute to biliary pathology, there is no correlation of disease severity and titer of AMAs. The recent development of well-characterized monoclonal antibodies specific for the B cell populations, anti-CD20 and anti-CD79, and the development of a well-defined xenobiotic-induced model of autoimmune cholangitis prompted us to use these reagents and the model to address the contribution of B cells in the pathogenesis of murine PBC. Prior to the induction of autoimmune cholangitis, mice were treated with either anti-CD20, anti-CD79, or isotype-matched control monoclonal antibody and followed for B cell development, the appearance of AMAs, liver pathology, and cytokine production. Results of the studies reported herein show that the in vivo depletion of B cells using either anti-CD20 or anti-CD79 led to the development of a more severe form of cholangitis than observed in control mice, which is in contrast with results from several other autoimmune models that have documented an important therapeutic role of B cell–specific depletion. Anti-CD20/CD79–treated mice had increased liver T cell infiltrates and higher levels of proinflammatory cytokines. Conclusion: Our results reflect a novel disease-protective role of B cells in PBC and suggest that B cell depletion therapy in humans with PBC should be approached with caution (HEPATOLOGY 2011:53:527-535)
Although the role of B cells in autoimmunity has historically been associated with the ability to produce autoantibodies,1 it is now clear that B cells are involved in multiple mechanisms beyond antibody secretion, including regulatory function.2, 3 Indeed, B cells efficiently present antigens,4 act as costimulators during the initiation of immune responses,5-7 and secrete cytokines.3, 8-10 Not surprisingly, this increased awareness of the importance of B cells in the pathogenesis of autoimmunity has led to the development of novel B cell–targeted biological therapies.11-15
Primary biliary cirrhosis (PBC) is considered a model autoimmune disease highlighted by the presence of high titers of antimitochondrial antibodies (AMAs) against the E2 subunit of the pyruvate dehydrogenase complex (PDC-E2), which is found in 95% of patients16-20 and is considered the most specific autoantibody in human autoimmune disease. Interestingly, there is no correlation of disease severity with the titer of AMA,21, 22 and recent studies in the transforming growth factor β receptor II dominant negative (dnTGF-βRII) murine model of PBC23 have suggested that whereas depletion of B cells in adult mice worsens liver disease,24 similar depletion of B cells in young dnTGF-βRII mice has a marginal beneficial clinical effect.24
Until recently, our understanding of PBC has been limited by the absence of appropriate animal models. Based upon a rigorous quantitative analysis of the epitope of PDC-E2, our laboratory has identified several organic compounds that resemble the immunodominant epitope of PDC-E2. In particular, 2-octynoic acid (2OA), a compound found in perfumes, lipstick, and many common food flavorings, reacts equally or even better than lipoic acid to AMAs.25-26 Importantly, immunization with 2OA when coupled with bovine serum albumin (BSA), induces high-titer AMAs and portal inflammation strikingly similar to human PBC.27-29 We report herein that treatment of this xenobiotic induced murine model of human PBC with either anti-CD20 or anti-CD79 monoclonal antibodies (mAbs) exacerbates liver pathology, even though it successfully depletes B cells and diminishes the production of AMAs. These findings have important clinical implications for the treatment of PBC and other autoimmune diseases in which B cell regulatory function may be critical.
Materials and Methods
Female C57BL/6J (B6) mice were obtained from The Jackson Laboratory (Bar Harbor, ME) and maintained in ventilated cages under specific pathogen-free conditions at the animal facilities of the University of California at Davis. The Animal Care and Use Committee in University of California Davis approved all studies.
To deplete B cells in vivo, four independent groups of 6-week-old mice were injected intraperitoneally weekly with either sterile murine immunoglobulin (IgG) 2a anti-mouse CD20 antibody (n = 8) (250 μg/250 μL in phosphate-buffered saline [PBS]), hamster IgG2 anti-mouse CD79b antibody (n = 10) (1 mg/100 μL in PBS), or isotype-matched control mAbs. The anti-mouse CD20 IgG2a (Biogen Idec, San Diego, CA) and the Armenian hamster anti-mouse CD79b IgG used herein have been described elsewhere.30, 31 The non–cross-reactive mouse anti-human CD20 antibody (250 μg/250 μL in PBS) and an Armenian hamster normal IgG (1 mg/mL) (Innovative Research, Novi, MI) were used as controls.
One week after the beginning of B cell depletion therapy, autoimmune cholangitis was induced as described.24 Briefly, a 2OA-BSA conjugate (100 μg/100 μL in PBS) was injected intraperitoneally with complete Freund's adjuvant (Sigma-Aldrich, St. Louis, MO) containing 1 mg/mL of Mycobacterium tuberculosis strain H37RA, and was subsequently boosted every 2 weeks with 2OA-BSA in incomplete Freund's adjuvant (Sigma-Aldrich). Additionally, mice received 100 ng of pertussis toxin (List Biological Laboratories, Campbell, CA) at the time of initial immunization with 2OA-BSA in Complete Freund's Adjuvant.
Peripheral blood samples from individual mice were obtained from the tail vein prior to the initiation of treatment with mAbs (baseline) and then at 2-week intervals. Sera was collected prior to mAb treatment, 1 week afterward, and thereafter every 4 weeks, and stored at −70°C until use. Animals were sacrificed at 15 weeks of age.
Detection of Anti–PDC-E2 Antibodies.
Serum titers of anti–PDC-E2 autoantibodies were measured by way of enzyme-linked immunosorbent assay using our well-standardized recombinant autoantigens.32
Mononuclear Cell Preparation and Flow Cytometry.
Peripheral blood mononuclear cells were isolated from heparinized murine blood using Accupaque (Accurate Chemical & Scientific Company, Westbury, CT) to assess levels of B cells. Cells were preincubated with anti-mouse FcR blocking reagent and then incubated at 4°C with a predetermined optimum concentration of antigen-presenting cell (APC)-conjugated anti–T cell receptor β (BioLegend), phycoerythrin-conjugated anti-mouse IgM (Caltag), and fluorescein isothiocyanate–conjugated anti-CD19 (BioLegend); B cell frequency was then determined by way of flow cytometry. The liver and spleens were collected from mice immediately following sacrifice, and single-cell mononuclear cell suspensions were prepared for multicolor flow analysis as described.23
Histopathology and Immunohistochemistry.
Immediately after sacrifice, liver and spleen tissues were harvested and fixed in 10% buffered formalin, embedded in paraffin, and cut into 4-μm sections for routine hematoxylin (DakoCytomation, Carpinteria, CA) and eosin (American Master Tech Scientific, Lodi, CA) staining. Evaluation under light microscopy and scoring of liver inflammation was performed on coded hematoxylin and eosin–stained sections of liver using a set of three indices by a blinded pathologist (K. T.); indices included degrees of portal inflammation, parenchymal inflammation, and bile duct damage.33 Phenotypic analysis of bile duct damage was performed as described.34
Serum Liver Enzyme Quantification.
Serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and alkaline phosphate (ALP) levels were measured using the Roche Diagnostics COBAS INTEGRA 400 Plus (Indianapolis, IN).
Serum levels of the proinflammatory cytokines interleukin (IL)-6, IL-10, monocyte chemotactic protein-1 (MCP-1), interferon-γ (IFN-γ), tumor necrosis factor α, and IL-12p70 were quantified using the BD Cytometric Bead Array Mouse Inflammatory Kit (BD Biosciences) as described.35 The serum samples were loaded onto the plate neat. The samples were then analyzed using the commercial protocol using a FACScan flow cytometer (BD Immunocytometry, San Jose, CA) and BD Cytometric Bead Array software.
The statistical difference between groups was determined using a two-tailed Mann-Whitney nonparametric test with 95% confidence interval. All results are expressed as the mean ± SEM. The Prism statistical package (GraphPad Software Inc, La Jolla, CA) was used. P < 0.05 was considered statistically significant.
Anti-CD20 and Anti-CD79 Antibodies Deplete B Cells.
To assess the efficacy of the B cell depletion, the frequency of CD19+ cells in peripheral blood was quantified by flow cytometry. The vast majority of B cells were depleted 1 week after the beginning of antibody administration in mice treated with either anti-CD20 (Fig. 1A-C) or anti-CD79 (Fig. 1B-D), whereas control mice exhibited no detectable changes in B cell frequency. Indeed, the frequency of CD19+ cells in peripheral blood mononuclear cells from anti-CD79–treated mice was 9.60% versus 46.89% in controls (P < 0.001) after 1 week of treatment, and 0.34% in anti-CD20–treated mice versus 32.63% in control mice (P < 0.001). Importantly, both treated groups consistently displayed marked depletion of B cells after 8 weeks of therapy. B cell depletion was also assessed in the livers and spleens of the anti-CD20–treated and anti-CD79–treated mice (Table 1). Again, these mice demonstrated a decrease in B cells compared with control mice.
|Control mAb||CD20 mAb||P Value||Control mAb||CD79 mAb||P Value|
|CD19+ TCRβ− cells (×106)|
|Liver||1.07 ± 0.11||0.01 ± 0.01||<0.002||1.68 ± 0.41||0.11 ± 0.02||<0.001|
|Spleen||32.01 ± 3.65||0.06 ± 0.01||<0.001||53.05 ± 4.91||2.92 ± 0.32||<0.001|
|CD3+ T cells (×106)||0.68 ± 0.53||1.04 ± 0.12||0.011||1.97 ± 0.18||4.32 ± 0.50||0.001|
|CD3+ CD4+ (×106)||0.26 ± 0.02||0.38 ± 0.02||0.017||0.71 ± 0.07||1.23 ± 0.13||0.005|
|CD3+ CD8+ (×106)||0.14 ± 0.04||0.25 ± 0.08||0.030||0.63 ± 0.07||0.91 ± 0.08||0.027|
|CD4+ CD62L− CD44hi (×106)||0.09 ± 0.01||0.13 ± 0.02||0.026||0.27 ± 0.03||0.64 ± 0.11||0.004|
|CD8+ CD62L− CD44hi (×106)||0.05 ± 0.01||0.14 ± 0.06||0.034||0.15 ± 0.02||0.36 ± 0.06||0.007|
|CD3+ NK1.1+ (×106)||0.33 ± 0.04||0.39 ± 0.08||NS||0.46 ± 0.06||1.00 ± 0.14||0.002|
|CD3- NK1.1+ (×106)||0.24 ± 0.02||0.27 ± 0.04||NS||0.43 ± 0.05||0.50 ± 0.04||NS|
B Cell Depletion Abolishes PDC-E2–Specific Antibody Production.
The effect of B cell depletion on serum reactivity against PDC-E2 was assessed at weeks 4 and 8 after the first immunization with 2OA. Whereas mice treated with control antibodies produced high titers of PDC-E2–specific antibodies, sera from mice depleted of B cells showed undetectable levels of PDC-E2 reactive antibodies (P < 0.0001) (Fig. 2).
B Cell–Depleted Mice Develop More Aggressive PBC-Like Liver Disease.
Liver sections from mice treated with anti-CD20 (Fig. 3A) and anti-CD79 (Fig. 4A) demonstrated a marked increase of cellular infiltrates in the portal tract and around the interlobular bile ducts, as well as an overall marked increase in liver inflammation compared with their respective controls (data not shown). Increased infiltration of lymphocytes or mononuclear cells surrounding damaged bile ducts was frequently observed in portal areas. The degrees of portal tract inflammation plotted individually are shown in Figs. 3B and 4B. Furthermore, bile duct damage was observed, and epithelioid granulomas were scattered within some portal tracts and also in hepatic parenchyma. Bile duct damage was studied by immunostaining with anti-CK22 (Fig. 5). Histological findings characteristic of PBC-like disease, including interlobular bile duct damage and nonsuppurative destructive cholangitis, were readily noted in the liver tissues from B cell–depleted mice.
B Cell Depletion by Anti-CD20 and Anti-CD79 Increases the Number of CD4+ and CD8+ T Cell Infiltrates in the Liver.
To clarify whether T cell infiltration was affected by B cell depletion with anti-CD20 and anti-CD79, total T cell numbers in the liver and spleen were quantified by way of flow cytometric analysis (Table 1). The number of CD3+ T cells, as well as absolute number of liver CD4+ and CD8+ T cells, was significantly increased in livers of B cell–depleted mice compared with control groups. Importantly, the absolute number of both CD4+ and CD8+ T cells that coexpressed CD44 but lacked CD62L, the phenotype of mouse effector memory T cells, was increased over controls in the liver: CD4+ CD62L− D44+ cells, 0.136 ± 0.05 versus 0.09 ± 0.02 (× 106) (P = 0.0262) in CD20-treated mice versus controls and 0.64 ± 0.11 versus 0.27 ± 0.03 (×106) (P = 0.004) in CD79-treated mice versus controls; CD8+ CD62L− CD44+ cells, 0.14 ± 0.06 versus 0.05 ± 0.01 (× 106) (P = 0.0348) in CD20-treated mice versus controls and 0.362 ± 0.06 versus 0.158 ± 0.029 (× 106) (P = 0.007) in CD79-treated mice versus controls. As expected, there was no difference in the phenotypic distribution of mononuclear cells in the spleen (data not shown).
B Cell–Depleted Mice Have Elevated Levels of Liver Enzymes.
ALT, AST, and ALP levels were measured to assess the correlation between serum biochemical values and histopathological abnormalities. As seen in Table 2, mice treated with anti-CD79 produced higher levels of ALT, AST, and ALP compared with that of control mice (P < 0.05). Mice treated with anti-CD20 demonstrated a significant increase in AST levels only (P < 0.05).
The levels of serum inflammatory cytokines in B cell–depleted mice were higher compared with control mice starting at 4 weeks of 2OA immunization (Fig. 6A). In particular, serum levels of IFN-γ in the CD20- and CD79-treated mice were significantly higher compared with sera from control mice (P = 0.046 and P = 0.011, respectively). Similarly, serum levels of MCP-1 in CD20- and CD79-treated mice were also significantly higher than sera from control mice (P = 0.0017 and P = 0.042, respectively) at 8 weeks after cholangitis induction. As expected, IL-10, in part secreted by B cells, was lower in B cell–depleted mice (Fig. 6B).
We demonstrate herein that acute B cell depletion in mice with otherwise normal immune systems exacerbates cholangitis induction. Liver cell infiltrates from B cell–depleted mice contained increased populations of CD4+ and CD8+ T cells and activated counterparts and exhibited increased serum levels of IFN-γ and MCP-1, but lower levels of IL-10, compared with B cell–sufficient mice. These findings should be considered in the context of recent studies on the role of B cells in a genetic animal model of PBC, the dnTGF-βRII mouse.23 Anti-CD20 therapy in young dnTGF-βRII mice attenuates liver damage but exacerbates inflammatory bowel diseases; however, B cell depletion in older mice does not modify the course of liver disease.34 Importantly, double transgenic mice with PBC-like disease and B cell depletion (Igμ−/−dnTGF-βRII mice) developed a more severe form of cholangitis.24 The present study demonstrates that B cell depletion immediately before disease induction in the xenobiotic induced murine model enhances disease severity in the absence of AMAs, suggesting that B cells play a critical regulatory role in the breakdown of tolerance against PDC-E2. These findings demonstrate that B cells play both positive and negative regulatory roles in the pathogenesis of PBC, and that the balance of these effects is likely to determine the severity of disease. In the study reported here, we deplete B cells before induction of cholangitis by xenobiotics. However, future studies on different timings of B cell depletion after induction of cholangitis by xenobiotics will be helpful to better define the role of B cells in the natural history of established disease.
A beneficial effect of anti-CD20 therapy has been reported in animal models of T cell–mediated disease, including experimental autoimmune encephalomyelitis (EAE),2 type 1 diabetes,36 and systemic sclerosis.37 It has been attributed primarily to reduced T cell activation; however, the reduction of antibody production may also have a beneficial effect.12 The importance of B regulatory cells has been suggested in several autoimmune diseases2, 38-39 and may reflect a role of IL-10-producing B cells as suppressors of autoimmune and inflammatory diseases.40, 41 A recent study reported that B regulatory cells predominantly control disease initiation in the EAE model, whereas T regulatory cells reciprocally inhibit late-phase disease.42 A proposed model for EAE may explain the role of B regulatory cells in PBC. Fillatreau et al.43 suggested that following immunization with autoantigen in complete Freund's adjuvant, activation of APCs through Toll-like receptor (TLR) 2, TLR4, and TLR9 (by mycobacterial ligands) stimulates the production of high levels of cytokines (IL-6, IL-12, IL-23) and drives the expansion of two auto-antigen–reactive CD4+ T cell populations: an auto-aggressive population and a T regulatory cell population. Concomitant stimulation of B cells through TLR2 or TLR4 early in the response induces production of IL-10, which has an inhibitory effect on the cytokine production by APCs, and eventually limits the initial expansion of the auto-aggressive cohort, ultimately leading to resolution of the disease. In the absence of B cells (or B cell–derived IL-10), the early expansion of the auto-aggressive population dominates, and the T regulatory cell cohort is unable to control this population. In accordance with this theory, we found lower levels of IL-10 in B cell–depleted mice.
The mechanisms responsible for exacerbation of cholangitis in B cell–depleted mice remain enigmatic, but the following data are relevant. First, B cell–depleted mice generate high levels of IFN-γ, a potent activator of the innate immune system.44 The innate immune system of patients with PBC demonstrates a higher reactivity than controls.45 Indeed, the frequency and absolute number of blood and liver resident natural killer cells are increased in patients with PBC, as is their cytotoxic activity and perforin expression46; moreover, peripheral monocytes from patients with PBC secrete higher levels of cytokines.47 Second, B cell depletion in the gut lymphoid tissue may increase portal venous concentrations of pathogen-associated molecular patterns with consequent activation of the innate immune system through TLR recognition. Third, evidence suggests that antigen-naïve B cells exert anti-inflammatory properties,48 which may inhibit APC maturation and proinflammatory differentiation49; in this regard, it has been demonstrated that dendritic cells from B cell–deficient mice produce higher levels of IL-12 and promote proinflammatory T cell differentiation.8
Thus, our findings suggest that B cell depletion therapy may be contraindicated in PBC. Indeed, we know that B cell depletion in dnTGF-βRII mice exacerbates inflammation of bile ducts.24 Although a biochemical benefit of anti-CD20 therapy in PBC patients refractory to ursodeoxycholic acid has been reported,50 we suggest that additional data regarding the role of B cells in human mucosal autoimmune diseases such as PBC are needed. Finally, our findings underscore the fact that unanticipated problematic issues can arise from the use of biologics in humans and thus the importance of trials in murine models and rigorous post marketing surveillance of such agents in humans.
- 50Rituximab for primary biliary cirrhosis (PBC) refractory to ursodeoxycholic acid (UDCA). HEPATOLOGY 2007; 46: 550. Abstract., , , , , , et al.