CD4+,CD25+ T regulatory cells may offer opportunities to intervene in the course of autoimmune disease. We wished to evaluate their potential for influencing systemic and chronic joint inflammation by investigating their involvement in collagen-induced arthritis (CIA).
We depleted DBA/1 mice of CD25+ regulatory cells by injection of a depleting monoclonal antibody specific for CD25 14 days before a single immunization with type II collagen (CII) in Freund's complete adjuvant. CD4+,CD25+ T cells were adoptively transferred to some groups of mice during immunization. Mice were then scored for signs of arthritis, and blood was taken periodically to measure the amounts of CII-specific antibodies. Splenocytes of treated mice were examined in vitro to determine the effects of depletion on proliferation to CII and control antigens.
CD25+ cell–depleted DBA/1 mice had significantly more severe disease than control mice following collagen immunization. The magnified severity was also accompanied by higher antibody titers against collagen, and in vitro tests showed increased proliferation of collagen-specific T cells. Adoptively transferring CD4+,CD25+ T cells into depleted mice was shown to reverse the heightened severity. Control mice, which were depleted and immunized with the neoantigen keyhole limpet hemocyanin (KLH), had neither an increased antibody response toward KLH nor an augmented proliferative response, indicating that CD25+ cell depletion preferentially affects immunity against self antigen.
These results establish a link between CD4+,CD25+ regulatory cells and CIA and provide a rationale for investigating CD4+,CD25+ T regulatory cells in the treatment and prevention of arthritis.
A subset of CD4+ T cells expressing CD25, the α chain of the interleukin-2 (IL-2) receptor, is a potent regulatory T cell population. These T cells have shown, both in vitro and in vivo, properties that are characteristic of T regulatory cells. For example, they are anergic to T cell receptor stimulation (1, 2) and have been reported to be capable of suppressing the proliferation of other T cells through membrane-bound transforming growth factor β (3), IL-10 secretion (4–6), or down-regulation of costimulatory molecules on antigen-presenting cells (7). Given their potency as regulators, it is not surprising that they play an important role in the control of autoimmunity. This role is exemplified best by experiments involving reconstitution of immunodeficient nude mice with CD4+ cells that were depleted of CD25+ cells. CD4+,CD25− reconstituted nude mice develop various organ-specific autoimmune diseases, such as gastritis, oophoritis, orchitis, and thyroiditis (8). Inclusion of the CD4+,CD25+ subset in the nude mice prevents the onset of these diseases. The protective value of CD4+,CD25+ cells against organ-specific autoimmunity has also been shown in several other models of autoimmunity, such as autoimmunity caused by neonatal thymectomy performed 3 days after birth (9) or inflammatory bowel disease caused by reconstitution of SCID mice with CD45RBhigh,CD4+ T cells (10).
The above examples demonstrate the role of these cells in the prevention of several autoimmune diseases in immunocompromised mice. However, the involvement of intrinsic CD25+ regulatory cells in rheumatic disease models (e.g., collagen-induced arthritis [CIA]), which are induced through immunization of susceptible mice with self antigen or homologous antigen, has not been thoroughly investigated.
To address this issue, we examined the regulation of CIA by CD25+ cells. CIA is induced in susceptible mice having specific major histocompatibility complex class II alleles, H2-Aq and H2-Ar, by immunization with type II collagen (CII) emulsified in Freund's complete adjuvant (CFA). Between 3 and 4 weeks following immunization, swollen limbs appear with a pathology that resembles rheumatoid arthritis (RA), thus making CIA an appropriate model for studying fundamental and applied aspects of this human disease.
We examined the role of CD4+,CD25+ T regulatory cells in CIA by depleting CD25+ cells in susceptible DBA/1 mice 14 days before immunization with CII. Previous studies in mice and rhesus monkeys showed that depletion of CD25+ cells near the time of or after immunization, without first allowing the antibody to be eliminated from the subject's body, resulted in lesser severity due to depletion of activated T cells needed to cause the disease (11, 12). We found that depletion, when administered well before CII immunization, significantly increased severity and incidence of the disease. Titers of antibodies formed against collagen were also significantly raised in depleted mice. Similarly, in vitro examination of the proliferative capacity of splenocytes showed increased CII-specific proliferation. The clinical effects of CD25+ depletion could be reversed by adoptively transferring CD4+,CD25+ T cells isolated from naive mice. When the same depletion experiments were performed with a control antigen, keyhole limpet hemocyanin (KLH), an enhanced KLH-specific immunity was not seen, indicating that the enhanced immune reactivity resulting from depletion of CD25+ cells was limited to a subset of antigens. These findings indicate that CD4+,CD25+ regulatory cells play a role in the control of systemic autoimmune disease, and that the loss of this particular cell population enhances the clinical symptoms of chronic arthritis.
MATERIALS AND METHODS
Animals and antibody preparation.
The study was approved by the Experimental Animal Commission of the Leiden Academic Hospital. Male DBA/1J mice age 7–9 weeks were obtained from Charles River ('s-Hertogenbosch, The Netherlands) and kept conventionally. Treatment and maintenance were in accordance with the national guidelines for animal care.
Rat anti-murine CD25 was produced by culturing the cell line PC61 (American Type Culture Collection, Rockville, MD) in Iscove's modified Dulbecco's medium (BioWhittaker, Verviers, Belgium) supplemented with 1 unit/ml penicillin, 1 μg/ml streptomycin, 2 μML-glutamine, 20 μM β-mercaptoethanol, and 4% fetal calf serum (FCS). Using a protein G column (Roche, Basel, Switzerland), the CD25-specific monoclonal antibody (mAb) was purified from the supernatant according to the manufacturer's protocol.
Induction of CIA and evaluation of arthritis.
Bovine CII (BII; Chondrex, Redmond, WA) was dissolved in 0.1M acetic acid solution overnight at 4°C at a concentration of 2 mg/ml. The dissolved BII (100 μg BII/mouse) was emulsified with an equal volume of CFA (Difco, Detroit, MI), and 100 μl was injected subcutaneously into the base of the tail. Two weeks after immunization, mice were examined 3 times weekly. The presence of arthritis was determined by examining the appearance of the front and hind paws. Severity was graded for each paw using an established scoring system: 0 = normal joint, 1 = one or two swollen joints, 2 = more than two swollen joints, 3 = extreme swelling of the entire paw and/or ankylosis. An arthritis score was assigned to each mouse by summing the scores of each paw. The mice were killed upon obtaining two paws with a maximum score of 3.
Administration of anti-CD25 and flow cytometric analysis.
CD25+ cells were depleted by intraperitoneal injections of CD25-specific, rat IgG1 mAb. The mAb (400 μg/mouse) was dissolved in phosphate buffered saline (PBS) and administered on days −28, −24, −21, and −14 (except where noted) before immunization with 100 μg antigen (CII or KLH) in CFA. Control animals received injections of rat IgG (Jackson ImmunoResearch, West Grove, PA).
Depletion was verified by staining blood lymphocytes with antibodies specific for CD4+ and CD25+ (BD PharMingen, San Diego, CA). Using heparin-coated tubes, blood was collected through a small incision in the lateral tail vein. The blood cells were washed with fluorescence-activated cell sorting (FACS) buffer (PBS/0.5% bovine serum albumin) and stained in the dark for 30 minutes on ice. Subsequently, the cells were washed once with FACS buffer and then treated with FACS Lysing Solution (BD, Franklin Lakes, NJ), which lyses the erythrocytes and fixes the other cells. After washing 3 more times, the lymphocytes were analyzed using a FACScalibur and CellQuest software (both from BD).
Isolation and transfer of CD4+,CD25+ T regulatory cells.
Splenocytes and superficial inguinal and mesenteric lymph nodes were isolated from naive mice. Cells were enriched for CD4+ cell populations by first staining the cells with anti-CD8 (Sanver Tech, Heerhugowaard, The Netherlands) and anti-mouse class II (BD PharMingen). Class II+, CD8+, and B cell populations were then eliminated by panning on plates coated with goat anti-mouse IgG (ITK Diagnostics BV, Uithoorn, The Netherlands). CD25+ cells were isolated from this population by first staining with fluorescein isothiocyanate (FITC)–conjugated anti-CD25 mAb (BD PharMingen) followed by incubation with magnetic-activated cell sorting anti-FITC beads (Miltenyi Biotec, Auburn, CA). CD4+,CD25+ T cells were selected on an (LS) column, and the flow-through was collected as CD4+,CD25− T cells.
Isolated cells were activated by overnight incubation on 24-well plates coated with 2 μg/ml anti-CD3 (145-2C11) and with IL-2 (100 units/ml; Sanver Tech) added to RPMI medium supplemented with 1 unit/ml penicillin, 1 μg/ml streptomycin, 20 mML-glutamine, 50 μM β-mercaptoethanol, and 8% FCS. After harvesting, dead cells were removed using Lympholyte-M (Cedarlane, Hornby, Ontario, Canada). CD4+,CD25+ cells were determined to be >95% pure. Cells (0.5 × 106/mouse) were suspended in PBS and injected intravenously on the same day as immunization.
Collagen-specific cell proliferation and suppression assay.
Splenocytes were isolated from mice immunized with either CII or KLH (PerbioScience, Etten-Leur, The Netherlands) emulsified in CFA. Cells were restimulated in triplicate with 20 μg/ml antigen and cultured with Dulbecco's modified Eagle's medium (Life Technologies, Paisley, UK) supplemented with 1 unit/ml penicillin, 1 μg/ml streptomycin, 20 mML-glutamine, 50 μM β-mercaptoethanol, 10 mM HEPES, and 5% FCS in 96-well round-bottom plates at a concentration of 0.5 × 106/well. Proliferation was measured 3–5 days later by addition of 0.5 μCi/well of 3H-thymidine. Values shown represent the average of the triplicates with the medium values subtracted.
After isolating and activating CD4+,CD25− and CD4+,CD25+ T cells, cells were cultured in 96-well roundbottom plates in RPMI medium with equal numbers of freshly isolated splenocytes. Proliferation was stimulated by adding phytohemagglutinin (PHA) at a dilution of 1:400. 3H-thymidine incorporation was measured after 3–4 days of culture. Values shown represent the triplicate average.
Measurement of serum antigen-specific antibodies.
Antibodies were measured by enzyme-linked immunosorbent assay. Immuno-Maxisorp 96-well plates (Nunc, Roskilde, Denmark) were coated with BII (Chondrex) or KLH overnight at 4°C. After washing with PBS/0.5% Tween 20, plates were blocked with PBS/10% milk for 2 hours at 4°C. Serially diluted mouse serum was then incubated on the washed plates overnight at 4°C. Plates were subsequently treated with one of the following detection antibodies: horseradish peroxidase (HRP)–conjugated anti-mouse polyvalent immunoglobulins (Sigma, St. Louis, MO), HRP-conjugated anti-mouse IgG1, or HRP-conjugated anti-mouse IgG2a (Southern Biotechnology, Birmingham, AL). Detection was performed using 3,3′,5,5′-tetramethylbenzidine as a substrate (Sigma).
The optical density (OD) was measured at 450 nm using a microplate reader (Wallac, Gaithersburg, MD) and the reader's software (MultiCalc; Wallac). Antibody units for BII were determined using a reference serum created from pooled sera of arthritic mice and assigned an arbitrary level of collagen-specific antibodies. Antibody titers for KLH were determined by making titration curves for each sample. The specific titer for each sample was found by choosing the half-maximal OD value in the linear section of the curve and then finding the corresponding titer.
Radiographs were taken of formalin-fixed forelimbs. The limbs were radiographed using X-OMAT, TL2 film (Eastman Kodak, Rochester, NY) to determine bone destruction. The images were enlarged using a microscope.
Differences in disease severity were analyzed with a Student's unpaired 2-tailed t-test. Incidence was compared using the chi-square test. Survival curves were compared using the log rank test. Statistical significance of antibodies was determined by the Mann-Whitney test. P values less than 0.05 with a 95% confidence interval were considered significant.
Onset of severe arthritis hastened by CD25+ regulatory cell depletion.
To study the possibility of CD25+ regulatory cells influencing the progression of CIA, we depleted DBA/1 mice of CD25-expressing cells using a depleting anti-CD25 mAb from the cell line PC61. Complete depletion of CD25-expressing cells is normally seen 3 days after a single injection of anti-CD25, and 50% of the cells are recovered 3 weeks later, with full recovery typically seen at 4 weeks (Figure 1A). Due to the risk of depleting recently activated, CII-specific, pathogenic T cells needed to cause CIA, we elected to immunize mice with CII emulsified in CFA, 2 weeks after the last injection of the CD25-specific mAb. Following this protocol, depleted mice, when analyzed near the time of immunization, had a reduction of 80% in their CD4+,CD25+ T cell population compared with control mice (Figures 1B and C). Directly after immunization, the amount of CD25+ cells increased (data not shown), indicating that activated cells were not being depleted in mice which were injected previously with anti-CD25. In contrast, when mice were depleted 3 days before immunization, a decrease in disease severity was noted (Figure 2A), suggesting that recently activated, pathogenic T cells are indeed depleted when anti-CD25 is given shortly before vaccination.
CD25+ cell–depleted mice immunized with CII 2 weeks after the last injection of anti-CD25 showed a significantly higher incidence of disease 4–6 weeks after immunization compared with undepleted mice treated with the control antibody (Figure 2B). After this period, the dissimilarity in incidence between the groups disappeared due to the development of disease in control mice normally observed after a single immunization with CII in CFA.
Likewise, CD25+ cell–depleted mice immunized with CII showed significantly enhanced severity compared with control mice (Figure 2C) or compared with mice only depleted of CD25+ cells, which did not develop disease (data not shown). The increased severity, measured by visual examination of the paws, was also associated with severe bone destruction (Figure 3).
These effects on the severity of disease caused by depletion of CD25+ cells were reversed when CD4+,CD25+ T cells from naive mice were adoptively transferred to mice (Figures 4A and B). Using a standard in vitro assay, these CD4+,CD25+ T cells were found to be capable of suppressing PHA-stimulated T cell proliferation (Figure 4C). Taken together, these results show that depletion of CD25+ cells increases early arthritis incidence as well as disease severity following immunization with CII, indicating that CD25+ T regulatory cells play a role in the control of CIA.
CII-specific antibody titers increased by CD25+ cell depletion.
It has been shown previously that B cell–deficient mice are resistant to CIA (13), indicating that collagen-specific antibodies are crucial to disease induction. Therefore, we wished to investigate whether the antibody response toward collagen was enhanced following depletion of CD25+ cells and immunization with CII. Accordingly, we analyzed the titers of CII-specific antibodies in the sera of the mice at several time points after vaccination. Fifteen days after immunization, the titers of CII-specific antibodies (combined IgG, IgA, and IgM) in CD25+ cell–depleted mice were significantly higher than those in undepleted control mice (Figures 5A and B).
Since mice depleted of CD25+ cells can spontaneously produce autoantibodies, which is reflected by an increase of total IgG (14, 15), we were interested to see whether spontaneous, collagen-specific antibodies also developed in CIA-susceptible mice after depletion. Collagen-specific antibodies did not develop in DBA/1 mice receiving only injections of CD25-specific mAb (Figure 5A) or in depleted mice immunized with the control antigen, KLH in CFA (data not shown), indicating that CII-specific antibodies are not produced spontaneously as a result of depletion. These results imply that the marked increase of CII-specific antibodies was dependent on depletion of CD25+ regulatory cells followed by immunization with CII.
The IgG isotypes, IgG2a and IgG1, are associated with Th1- and Th2-mediated immune responses, respectively. Since CIA is considered to be a Th1-mediated disease, we wished to determine the effects of CD25+ cell depletion on the isotypes of the CII-specific antibody response. Therefore, we tested the sera collected on day 15 for IgG2a and IgG1 antibodies specific for collagen. The depleted mice had significantly increased CII-specific antibody titers of both of these isotypes (Figures 5C and D), indicating that depletion of CD25+ cells did not affect the class of the antibody response.
KLH-specific antibody titer not increased by CD25+ cell depletion.
Recent research has shown that CD4+,CD25+ T cells are preferentially reactive toward self antigen (16). Because the influence of CD25+ cells on the B cell responses directed against self antigens (such as CII) and foreign antigens (such as KLH) is poorly defined, we studied the effect of depletion on an antibody response against a neoantigen (KLH) to gain a better understanding of the ability of CD25+ regulatory cells to modulate B cell responses against neoantigens. We therefore depleted mice of CD25+ cells and immunized them with KLH. As shown in Figure 6, the amount of KLH-specific antibody in each group was similar 14 days after immunization. Measurements taken at later time points after immunization also showed no difference between the groups (data not shown). These results suggest that depletion of CD25+ cells does not lead to an increase in the antibody response toward neoantigens. Taken together, these findings imply that CD25+ regulatory cells preferentially control self antigen–directed antibody responses.
Increased CII-specific proliferation displayed by splenocytes from CD25+ cell–depleted, CII-immunized mice.
The observation that depletion of CD25+ cells augments the antibody response in immunized mice in an antigen-specific manner prompted us to examine whether CII-specific T cell responsiveness was also increased after immunization. To determine antigen-specific proliferation, splenocytes were collected from depleted and nondepleted mice several weeks after immunization with either CII or KLH. CD25+ cell–depleted, CII-immunized mice showed a large increase in CII-specific proliferation compared with nondepleted controls (Figure 7). In contrast, the proliferation against the neoantigen KLH, determined from depleted, KLH-immunized mice, was not enhanced compared with that in their undepleted counterparts (Figure 7). These results demonstrate that CD25+ depletion preferentially influences CII-specific immunity, whereas T cell reactivity against the neoantigen KLH remains unaffected.
Transfer of CD25− splenocytes into immunocompromised mice can result in the development of various organ-specific autoimmune diseases (15). The most prevalent disease induced in this manner is gastritis. A small minority of these mice (5–10%) can develop symptoms that resemble some aspects of polyarthritis, suggesting that CD25+ regulatory cells could be involved in the control of systemic autoimmune diseases like arthritis (15). We have now shown that depletion of CD25+ cells in immunocompetent mice rapidly leads to severe arthritis following injection of CII. Furthermore, adoptive transfer of CD4+,CD25+ T cells, but not CD4+,CD25− T cells, reverses the effects of CD25+ cell depletion, indicating that CD25+ T regulatory cells play a pivotal role in the control of systemic polyarthritis. Depletion of CD25+ T cells was associated not only with enhanced induction of disease and more severe disease, but also with the emergence of high CII-specific antibody titers within 2 weeks after CII administration, as well as with a pronounced CII-specific T cell response.
Although CD25 can also be expressed by other cells, such as activated B cells, macrophages, and subsets of dendritic cells, we consider it likely that the effects observed after anti-CD25 injection are mediated by depletion of CD4+,CD25+ T regulatory cells. Indeed, a strongly reduced frequency of these cells was noted at the time of CII injection, and adoptive transfer of these cells reversed the severe disease seen by depletion to a level of disease typically seen after the standard immunization protocol. Moreover, B cells (by production of pathogenic autoreactive antibodies), macrophages (by production of cytokines and tissue-degrading enzymes), and dendritic cells (by CII presentation to T cells) are all involved in disease induction and/or progression (for review, see ref. 17). Therefore, it is likely that depletion of CD25-expressing B cells, macrophages, and dendritic cells would temper disease instead of enhancing its severity.
Although the mode of action of CD4+,CD25+ T regulatory cells is still poorly understood, the observation that these cells can modulate CIA is intriguing. The immune response against the neoantigen KLH at both the T and B cell levels remains unaffected by the depletion of CD25+ cells, emphasizing that loss of CD4+,CD25+ T regulatory cells does not enhance the immune responses for a wide range of antigens (16). CIA is crucially dependent on both antigen-specific T and B cells; among other functions, T cells are thought to be involved in the provision of help required for the production of autoantibodies. These autoantibodies are most likely involved in the end-stage effector mechanism by massively recruiting leukocytes such as macrophages and neutrophils through complement activation and Fc receptor triggering, eventually resulting in a massive production of cytokines, growth factors, and tissue-degradative enzymes (18, 19). Therefore, it is conceivable that the enhanced CII-specific antibody response is the most prominent factor responsible for disease aggravation (20).
The fact that enhanced CII-specific antibody titers are found in CD25+ cell–depleted animals could be explained by the augmented magnitude of the CII-specific T helper response, resulting in an increased helper activity for B cell function. However, these results could also be explained by the lack of direct B cell inhibition after depletion of CD25+ T regulatory cells. Indeed, it has been described that the activation of B cells leads to the production of the chemokine CCL4, which recruits regulatory T cells, resulting in the inhibition of B cell activity (21). The elimination of CCL4 or CD25+ T cells was shown to prevent this inhibition, implying that regulatory T cells can directly modulate B cell activity.
The finding that increased CII-specific antibody titers are present in CD25+ cell–depleted mice 2 weeks after immunization, when no joint inflammation is detectable, indicates that CD25+ regulatory cells are involved in the control of disease induction. However, these cells could also be involved in control of disease progression. For example, activated macrophages also produce CCL4 (21), which would enable them to recruit CD25+ T regulatory cells. Because macrophages play a central role in the effector stage of disease, it is possible that T regulatory cells could control disease progression by inhibiting macrophage function. It is important to determine whether T regulatory cells can temper disease progression as well as the processes leading to disease initiation, since this would offer the opportunity to modulate established arthritis by harnessing CD25+ T regulatory cells.
Currently, it is not known whether regulatory T cells also play a role in the prevention and/or inhibition of RA. CD4+,CD25+ T cells are present at relatively low frequencies in the synovium of RA patients, but these cells may be recently activated pathogenic CD4+ T cells (22). Furthermore, the phenotype, cytokine-production profile, regulatory properties, or antigen specificity of CD4+,CD25+ T cells in general, at various stages of disease, is not known. Therefore, it will be important to determine whether CD4+,CD25+ T cells with an immunosuppressive phenotype are lost during the progression of RA, and to determine whether abnormal regulatory circuits are playing a role in induction and/or progression of RA.
Previously, the effects of systemic anti-CD4 mAb therapy have been studied in RA patients with relatively little success (23). Currently, therapies using anti-CD25 are being tested against established arthritis in rhesus monkeys (12). Unlike the studies with anti-CD4, these studies are proving that anti-CD25 has the potential to treat chronic arthritis. With both therapies, however, the possibility exists that crucial CD4+,CD25+ T regulatory cells are also being depleted. Because the consequences of CD4+,CD25+ T cell depletion are not clearly established in humans, interventions involving these mAb should be performed with some caution, since these therapies may harm the patient's ability to control autoimmune disease.
We have observed that CD25+ T regulatory cells are involved in the control of systemic joint inflammation. This finding offers a rationale for analyzing whether these cells are also employed to guard against systemic arthritis in humans and for determining whether they can be utilized in therapies for chronic polyarthritis.
We would like to thank Astrid van Halteren for her critical comments on this manuscript, the staff of the Leiden University Medical Center animal facility for their excellent work, and the members of the Autoimmunity section of the Immunohematology and Blood Transfusion department for their helpful commentary.