Therapeutic effect of cytotoxic T lymphocyte antigen 4/immunoglobulin on a murine model of primary biliary cirrhosis


  • Potential conflict of interest: Dr. Nadler owns stock in Bristol-Myers Squibb.

  • Financial was support provided by a grant from the National Institutes of Health (DK067003).


Collectively, the data in both humans and murine models of human primary biliary cirrhosis (PBC) suggest that activated T cells, particularly CD8 T cells, play a critical role in biliary cell destruction. Under physiological conditions, T-cell activation involves two critical signals that involve the major histocompatibility complex and a set of costimulatory molecules, which include a receptor on T cells termed cytotoxic T lymphocyte antigen 4 (CTLA-4). Germane to the studies reported herein, signaling by CTLA-4 has the potential to modulate costimulation and induce inhibitory signals. In this study, we have taken advantage of our well-defined murine model of PBC, in which mice are immunized with 2-octynoic acid coupled to bovine serum albumin (2OA-BSA), leading to the production of high-titer antimitochondrial autoantibodies (AMAs) and portal cellular infiltrates. To investigate the potential of CTLA-4-Ig (immunoglobulin) as an immunotherapeutic agent, we treated mice both before and after induction of autoimmune cholangitis. First, we demonstrate that CTLA-4-Ig treatment, begun 1 day before 2OA-BSA immunization, completely inhibits the manifestations of cholangitis, including AMA production, intrahepatic T-cell infiltrates, and bile duct damage. However, and more critically, treatment with CTLA-4-Ig, initiated after the development of autoimmune cholangitis in previously immunized mice, also resulted in significant therapeutic benefit, including reduced intrahepatic T-cell infiltrates and biliary cell damage, although AMA levels were not altered. Conclusion: These data suggest that an optimized regimen with CTLA-4-Ig has the potential to serve as an investigative therapeutic tool in patients with PBC. (HEPATOLOGY 2013)

Primary biliary cirrhosis (PBC) is a chronic autoimmune liver disease characterized by nonsuppurative destructive cholangitis and cholestasis, as well as progressive development of fibrosis and cirrhosis, leading, eventually, to liver failure.1 Previous studies have suggested a critical involvement of autoreactive T cells in the pathogenesis of human PBC.2-4 Using various animal models, we have demonstrated that CD8 T cells play a critical role in the pathogenesis of PBC.5-7 Therefore, we submit that successful therapy of PBC, utilizing immunological approaches, requires control of T-cell activation and proliferation, their recruitment to the liver, and the secretion of proinflammatory cytokines by these cellular infiltrates.8-12

Under physiological conditions, T-cell activation involves two critical signals, one being the presentation of a peptide epitope in the context of class I or II major histocompatibility complex (MHC) by an antigen-presenting cell (APC) to the T-cell receptor.13-15 The second signal is delivered by the costimulators, CD80/CD86, on the APC to their receptor CD28 on the T cell,16, 17 which is required to sustain the APC/T-cell formation of an immunological synapse, leading to significant enhancement of T-cell activation. In contrast, cytotoxic T lymphocyte antigen 4 (CTLA-4), another receptor expressed by T cells, is associated with inhibitory properties.18, 19 CTLA-4 binds to CD80/CD86 with a much higher affinity than CD2820 and functions to attenuate T-cell activation by inhibiting costimulation and transmitting inhibitory signals to T cells. This results in decreased cytokine production, inhibition of cell-cycle progression, down-modulation of T-cell receptor (TCR) signaling, and decreased activation of B cells and macrophages.21-26

These properties of CTLA-4 prompted us to explore the potential of CTLA-4-based therapy for PBC utilizing our murine model in which mice immunized with 2-octynoic acid–conjugated bovine serum albumin (2OA-BSA) develop a PBC-like cholangitis.7, 27, 28 CTLA-4-Ig (immunoglobulin) is a soluble recombinant human fusion protein comprised of the extracellular domain of human CTLA-4 linked to a modified portion of the Fc domain of human IgG-1, in which the sequences involved in Fc receptor (FcR) binding and complement activation have been eliminated, preventing antibody (Ab)-dependent cellular cytotoxicity and complement-mediated activity.29-31 CTLA-4-Ig reversibly binds to both human and murine CD80/86 by its CTLA-4 portion, thereby preventing CD80/86 from interacting with CD28 and thereby inhibiting the delivery of the second signal required for full T-cell activation. Herein, we demonstrate that CTLA-4-Ig treatment, started 1 day before 2OA-BSA immunization, prevents the clinical and histological manifestations of cholangitis, including the inhibition of serum antimitochondrial Ab (AMA), intrahepatic inflammatory change, and bile duct damage. Of importance is our finding that CTLA-4-Ig treatment, initiated after the onset of established disease in previously 2OA-BSA-immunized mice, led to a significant reduction in levels of intrahepatic infiltrating pathogenic effector CD4 and CD8 T cells as well as the severity of biliary cell damage. Our data suggest that an optimized regimen with CTLA-4-Ig (commercially known as Abatacept; Bristol-Myers Squibb) is a potential therapeutic candidate for patients with PBC.


Ab, antibody; AMA, antimitochondrial Ab; APC, antigen-presenting cell; BSA, bovine serum albumin; CFA, complete Freund's adjuvant; CTLA-4, cytotoxic T lymphocyte antigen 4; ELISA, enzyme-linked immunosorbent assay; FCA, flow cytometric analysis; FcR, Fc receptor; H&E, hematoxylin and eosin; IFA, incomplete Freund's adjuvant; Ig, immunoglobulin; IL, interleukin; IP, intraperitoneally; MHC, major histocompatibility complex; MNCs, mononuclear cells; 2OA-BSA, 2-octynoic acid–conjugated BSA; PBC, primary biliary cirrhosis; PBS, phosphate-buffered saline; PDC-E2, pyruvate dehydrogenase complex/dihydrolipoyl transacetylase; TCR, T-cell receptor.

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 (Davis, CA). All studies conducted with the use of these animals were approved by the animal care and use committee at the University of California at Davis.

Induction of Cholangitis.

2OA, a synthetic chemical mimic of lipoic acid/lysine located within the inner domain of pyruvate dehydrogenase complex/dihydrolipoyl transacetylase (PDC-E2), was coupled to BSA, as previously described.7 For the induction of autoimmune cholangitis, 100 μg of 2OA-BSA conjugate (in 50 μL of phosphate-buffered saline; PBS) were emulsified with 50 μL of complete Freund's adjuvant (CFA; containing 1 mg/mL of Mycobacterium tuberculosis strain H37RA; Sigma-Aldrich, St. Louis, MO) and injected intraperitoneally (IP) into 6 week-old female B6 mice. Additionally, mice received 100 ng of pertussis toxin (List Biological Laboratories, Campbell, CA) in 100 μL of PBS (IP) at the time of, and 2 days after, the initial immunization with 2OA-BSA. After 2 weeks, mice were reboosted with 100 μg of 2OA-BSA in 50 μL of PBS emulsified with 50 μL of incomplete Freund's adjuvant (IFA; Sigma-Aldrich) administered IP.

CTLA-4-Ig Treatment.

Abatacept (BMS-188667, CTLA-4-Ig), utilized for the studies reported herein, was a gift from Bristol-Meyers Squibb (Clinton, NY). Each vial of CTLA-4-Ig (250 mg/vial) for injection was reconstituted with 10 mL of sterile water to yield a concentration of 25 mg/mL and further diluted with PBS to 4 mg/mL. Mice were injected IP three times a week at a dosage of 20 mg/kg body weight. Preventative treatment was started 1 day before the first immunization with 2OA-BSA and continued thrice-weekly for 8 weeks. Therapeutic treatment was started 8 weeks after the first immunization with 2OA-BSA and continued thrice-weekly for 4 weeks.

Detection of anti-PDC-E2 Abs.

Serum was collected every 2 weeks after 2OA-immunization and stored at −80°C. Levels of anti-PDC-E2 autoAbs in serum were measured by enzyme-linked immunosorbent assay (ELISA), as previously described,7, 32 using purified recombinant PDC-E2 to coat the ELISA plates. Serum samples were tested at a dilution of 1:250.

Cell Preparation and Flow Cytometry Analysis.

Livers and spleens were harvested immediately after sacrifice. First, livers were perfused with PBS containing 0.2% BSA (0.2% BSA/PBS), passed through a 100-μm nylon cell strainer (BD Biosciences, San Jose, CA) and resuspended in 0.2% BSA/PBS. Hepatocytes were removed as pellets after centrifugation (at 75 g for 1 minute), and the remaining suspended cells were collected. Spleens were disrupted between two glass slides and suspended in 0.2% BSA/PBS. Mononuclear cells (MNCs) from the liver and spleen were isolated by gradient centrifugation using Histopaque-1.077 (Sigma-Aldrich). For flow cytometry, cells were first incubated with monoclonal Ab against FcR for blocking nonspecific binding (BioLegend, San Diego, CA) for 10 minutes at 4°C and then stained with combinations of fluorochrome-conjugated Abs, including fluorescein-isothiocyanate–conjugated anti-CD44 (BioLegend), phycoerythrin-conjugated anti-CD62L (BioLegend), allophycocyanin-conjugated anti-TCR-β (eBioscience, San Diego, CA), peridinin chlorophyll protein complex–conjugated anti-CD8α (BioLegend), and allophycocyanin/cyanin 7–conjugated anti-CD4 (BioLegend) for phenotypic analysis of T-cell subsets. Multicolor flow cytometric analyses (FCAs) were performed using a FACScan flow cytometer (BD Biosciences) upgraded by Cytec Development (Fremont, CA) to allow for five-color analysis. Acquired data were analyzed with CellQuest software (BD Biosciences).

Histopathologic Staining.

Livers and spleens were harvested immediately after sacrifice and aliquots fixed in 10% buffered formalin at room temperature for 2 days, embedded in paraffin, and cut into 4-μm sections for routine hematoxylin (DakoCytomation, Carpinteria, CA) and eosin (American Master Tech Scientific, Lodi, CA) (H&E) staining. Standard pathologic evaluation utilizing light microscopy was performed, and relative levels of biliary cell damage were recorded on coded H&E-stained sections. Each section was scored (as either 0 = no pathologic change; 1 = minimal; 2 = mild; 3 = moderate; or 4 = severe pathology) by a pathologist, as previously described.7, 33, 34

Statistical Analysis.

Statistical differences between groups were determined using a two-tailed unpaired t test and a chi-square test for independence. All results were expressed as mean ± standard error. The Prism statistical package (GraphPad Software Inc., La Jolla, CA) was used. A P value of <0.05 was considered statistically significant.


CTLA-4-Ig treatment Before Immunization With 2OA-BSA.

In efforts to investigate the effects of CTLA-4-Ig treatment on preventing 2OA-BSA-induced autoimmune cholangitis, groups of B6 mice were untreated (controls) or treated with CTLA-4-Ig before immunization with 2OA-BSA. As expected, control mice developed significant levels of serum autoAbs to PDC-E2. In contrast, anti-PDC-E2 autoAbs were not detected in serum from mice treated with CTLA4-Ig (Fig. 1), indicating that the administration of CTLA-4-Ig before 2OA-BSA immunization prevented the production of AMAs. Coincidentally, whereas cholangitis, exemplified by lymphoplasmacytic infiltration and bile duct damage, was readily found in livers of control mice at 8 weeks after 2OA immunization, liver tissues from mice pretreated with CTLA-4-Ig showed a significant decrease in the extent of liver inflammation (Fig. 2A). Furthermore, interlobular bile duct damage was not observed in any of the treated mice (Fig. 2B). Of note, administration of CTLA-4-Ig alone without immunization with 2OA-BSA did not induce any detectable pathologic changes in mouse liver (data not shown). Finally, FCA was also performed on intrahepatic MNCs isolated from livers of control and CTLA-4-Ig-treated mice at 8 weeks post-2OA-BSA immunization. Results of such analysis revealed that whereas the former showed marked increases (P < 0.001) in the frequency and absolute numbers of effector (CD44+CD62L) CD4 and CD8 T cells, the latter demonstrated marked decreases in such infiltrates (see Fig. 3). These data indicate that CTLA4-Ig treatment started before 2OA-BSA immunization completely prevents loss of tolerance and subsequent autoimmune cholangitis.

Figure 1.

CTLA-4-Ig treatment inhibited production of PDC-E2-specific autoAbs induced by 2OA-BSA immunization. CTLA-4-Ig treatment was started 1 day before initial 2OA-BSA immunization and continued for 8 weeks (2OA+CTLA). Untreated mice (2OA) were utilized as a control group. Serum samples were collected at different time points, diluted 1:250, and tested for anti-PDC-E2 reactivity by ELISA assay. OD, optical value. Each group included 8 mice.

Figure 2.

CTLA-4-Ig treatment protected mice against 2OA-BSA-induced cholangitis. CTLA-4-Ig treatment was started 1 day before initial 2OA-BSA immunization and continued for 8 weeks. (A) H&E-stained representative liver sections from 2OA-BSA-immunized mice treated with CTLA-4-Ig, compared to controls. Red arrowhead: epithelioid granuloma; red arrow: interlobular bile duct damage. (B) Bile duct damage was evaluated as described on H&E-stained liver sections. Each group included 8 mice. *P < 0.05 (chi-square test for independence).

Figure 3.

CTLA-4-Ig treatment prevented intrahepatic infiltration of effector T cells. CTLA-4-Ig treatment was started 1 day before initial 2OA-BSA immunization. Frequency of the effector (CD44+CD62L) population in CD4 and CD8 T cells as well as number of these populations in livers were analyzed by FCA. Each group included 8 mice. *P < 0.05; **P < 0.01; ***P < 0.001 (two-tailed unpaired t test).

Therapeutic Efficacy of CTLA-4-Ig in 2OA-BSA-Induced Cholangitis in Mice.

In efforts to determine the therapeutic efficacy of CTLA-4-Ig, we utilized our standard protocol to induce cholangitis in groups of mice by 2OA-BSA immunization. Because disease onset was detected at 8 weeks after the initial immunization (as observed in Fig. 2), consistent with our previously data,7, 27, 28 we started treatment with CTLA-4-Ig at that time point. After 4 weeks of treatment, mice were sacrificed and examined for liver pathology. CTLA-4-Ig treatment significantly reduced the extent and severity of liver inflammation (Fig. 4A) and the extent of biliary cell damage (Fig. 4B), as compared to untreated mice. FCA of intrahepatic cellular infiltrates reflected a marked diminution in the frequency and absolute numbers of CD44+CD62L effector CD4 and CD8 T cells in CTLA-4-Ig treated mice, as compared to control mice (Fig. 5). Analysis of serum collected pre- and post-CTLA4-Ig treatment from these mice for levels of PDC-E2-specific autoAbs did not show any detectable difference in titers (Fig. 6).

Figure 4.

Effects of CTLA-4-Ig treatment on 2OA-BSA-induced cholangitis. 2OA-BSA-immunized mice were treated with CTLA-4-Ig for 4 weeks, started 8 weeks after initial immunization (2OA+CTLA). 2OA-BSA-immunized mice without CTLA-4-Ig treatment were utilized as the control group (2OA). (A) H&E-stained representative liver sections. (B) Scores of intrahepatic bile duct damage in 2OA-immunized mice with or without CTLA-4-Ig treatment. Each group included 12 mice. *P < 0.05 (chi-square test for independence).

Figure 5.

CTLA-4-Ig treatment reduced intrahepatic infiltration of effector T cells. 2OA-immunized mice were treated with CTLA-4-Ig for 4 weeks, started 8 weeks after initial immunization. Frequency of the effector (CD44+CD62L) population in CD4+ and CD8+ T cells as well as the number of these populations in livers were analyzed by FCA. Each group included 8-12 mice. *P < 0.05; **P < 0.01; ***P < 0.001 (unpaired t test).

Figure 6.

CTLA-4-Ig treatment started after onset of cholangitis did not affect serum levels of PDC-E2-specific autoAbs. CTLA-4-Ig treatment was started 8 weeks after initial 2OA-BSA immunization and continued for 4 weeks. Each group included 6-12 mice.


CTLA-4-Ig is a soluble, recombinant human fusion protein comprised of the extracellular domain of human CTLA-4, linked to a modified portion of the Fc domain of human IgG-1, which is devoid of FcR binding and complement activation activity.29-31 There is considerable conservation between murine and human CTLA-4, including in studies of the effects of graft-versus-host disease across the MHC barrier in mice.35

This agent mimics the action of native CTLA4 by binding to CD80/CD86 on APCs and competitively inhibits the essential CD28:CD80/CD86 costimulatory signal required for T-cell activation, resulting in down-regulation of subsequent immune-effector mechanisms.31, 36, 37 CTLA-4-Ig similarly inhibits the interaction between murine CD28 and its ligand, CD80/CD86. CTLA-4-Ig is a U.S. Food and Drug Administration–approved drug for patients with rheumatoid arthritis and juvenile idiopathic arthritis.38, 39 CTLA-4-Ig has also been shown to ameliorate autoimmunity in vivo during collagen-induced arthritis,40 experimental autoimmune encephalomyelitis,41 psoriatic arthritis,42 systemic lupus erythematosus,43 and ankylosing spondylitis,44 and has been shown to be effective in the prolongation of allograft and xenograft survival.45-48

In the current study, we used our animal model of xenobiotic-induced cholangitis to investigate the potential effects of CTLA-4-Ig, both as a preventative and as a therapeutic agent. Our data demonstrated that the development of both anti-PDC-E2 and liver pathology was successfully prevented in mice treated with CTLA-4-Ig 1 day before immunization with 2OA. Because CTLA-4 inhibits the costimulation of T cells required for T-cell activation, our results demonstrate the critical role of T-cell activation in the development of 2OA-induced autoimmune cholangitis, in which T cells serve as a pathogenic effector, as well as in the autoreactive B-cell response in which T-cell help is also essential. We thus proceeded to treat mice with established cholangitis at 8 weeks after 2OA-BSA immunization. Treatment with CTLA-4-Ig in this group not only resulted in reduced histological inflammatory changes in liver and biliary cell damage, but also reduced levels of intrahepatic pathogenic effector CD4 and CD8 T cells, indicating the potential therapeutic effect of CTLA-4-Ig.

Although CTLA-4-Ig treatment did not result in complete resolution of established cholangitis in our mouse model, the reduced frequency and absolute numbers of liver infiltrating pathogenic T cells and improved liver histology are important initial therapeutically beneficial findings. These data are consistent with the view that CTLA-4-Ig treatment inhibits the activation of naïve T cells, but not previously primed autoreactive memory T cells, because the latter have a lower threshold of activation and are less dependent on costimulatory signals for activation. However, it is important to keep in mind that activated T cells also need interaction with costimulatory molecules to provide survival signals for sustained effector T-cell responses. Thus, whereas the results of the current studies show significant therapeutic benefit, the studies only address the acute effects of CTLA-4-Ig therapy. Studies to determine whether discontinuing treatment will result in relapse of cholangitis, and whether a prolonged CTLA-4-Ig regimen (>4 weeks) will improve its therapeutic efficacy, need to be performed.

Although treatment with CTLA-4-Ig postimmunization reduced intrahepatic effector T-cell infiltration and biliary cell damage, these mice continued to manifest AMAs at levels comparable to control mice. Serum AMA is produced by long-lived plasma cells, which can be replenished from previously primed memory B cells with a lower threshold of activation.49, 50 This would explain the lack of effect of CTLA-4-Ig treatment on serum AMA levels in this group of mice. The differential response to CTLA-4-Ig treatment on T cells versus AMAs, as well as its effect on the therapeutic efficacy for PBC, should also be evaluated in future studies.

It is becoming increasingly clear that besides CTLA-4, there are a number of other inhibitory receptors that function to control immune responses, which include T-cell immunoreceptor with Ig and ITIM domains, programmed death-1, and T-cell Ig and mucin domain-containing molecule-3. At present, it is not clear whether there is a hierarchy among these molecules in terms of an ordered sequence by which these molecules act to regulate immune responses. In light of our previous studies on the role of inflammatory cytokines, such as interferon-gamma, interleukin (IL)-6, IL-12, and IL-23,51-53 in the pathogenesis of PBC (including autoAb production and biliary cell damages), it is reasonable to assume that CTLA-4 could influence the effect of these downstream inflammatory cytokines. However, it is clear that they do serve at different check points along the T-cell activation pathway, and it is likely that an optimized therapeutic approach may require a combination of therapeutic inhibitors to maintain long-term inhibition of autoimmune-effector mechanisms54 without compromising antiviral and tumor surveillance mechanisms.

With regard to CTLA-4, despite its extensive experimental and clinical use for the therapy of autoimmune disease and transplant rejection, there is a remarkable paucity of data on precisely how and when CTLA-4-Ig mediates its effect on T cells in vivo. In addition, the effect of CTLA-4-Ig on T-dependent B-cell responses are poorly understood and/or characterized. Although Abatacept has been demonstrated to have a significant therapeutic benefit in patients with rheumatoid arthritis,55, 56 an increased risk of infection, malignancy, and autoimmune events has been reported.57, 58 We reason that elucidation of the precise mechanisms by which CTLA-4-Ig mediate its effect will aid in defining the optimum therapeutic application of this unique immunomodulatory drug while minimizing in vivo toxicity and clinically nonbeneficial effects.


The authors thank Yugo Ando, Chen-Yen Yang, Kazuhito Kawata, and Hajime Tanaka for their technical support in this experiment. The authors also thank Ms. Nikki Phipps for her support in preparing this article.