Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Immunol. Author manuscript; available in PMC 2010 June 15.
Published in final edited form as:
PMCID: PMC2741092

The Immunoregulatory Enzyme IDO Paradoxically Drives B Cell-Mediated Autoimmunity1


Rheumatoid arthritis (RA) is a chronic and debilitating inflammatory autoimmune disease of unknown etiology. As with a variety of autoimmune disorders, evidence of elevated tryptophan catabolism has been detected in RA patients, indicative of activation of the immunomodulatory enzyme indoleamine 2,3-dioxygenase (IDO). However, the role that IDO plays in the disease process is not well understood. The conceptualization that IDO acts solely to suppress effector T cell activation has led to the general assumption that inhibition of IDO activity should exacerbate autoimmune disorders. Recent results in cancer models, however, suggest a more complex role for IDO as an integral component of the inflammatory microenvironment necessary for supporting tumor outgrowth. This has led us to investigate the involvement of IDO in the pathological inflammation associated with RA. Using the K/BxN murine RA model and IDO inhibitor 1-methyl-tryptophan (1MT), we found that inhibiting IDO activity had the unexpected consequence of ameliorating, rather than exacerbating arthritis symptoms. 1MT treatment led to decreased autoantibody titers, reduced levels of inflammatory cytokines, and an attenuated disease course. This alleviation of arthritis was not due to an altered T cell response, but rather resulted from a diminished autoreactive B cell response, thus demonstrating a previously unappreciated role for IDO in stimulating B cell responses. Our findings raise the question of how an immunosuppressive enzyme can paradoxically drive autoimmunity. We suggest that IDO is not simply immunosuppressive, but rather plays a more complex role in modulating inflammatory responses, in particular those that are driven by autoreactive B cells.

Keywords: B cells, Rheumatoid Arthritis, Autoantibodies, Inflammation, Indoleamine 2,3-dioxygenase, 1-methyl tryptophan


Rheumatoid arthritis (RA)3 is an inflammatory autoimmune disease characterized by chronic inflammation of the synovial joints, eventually leading to a progressive and debilitating destruction of cartilage and bone (1). K/BxN mice spontaneously develop a joint inflammatory disease that shares many characteristics with human RA, including cellular infiltrates, pro-inflammatory cytokines, autoantibodies, and cartilage and bone destruction (2, 3). This model uses a T cell receptor transgene, KRN, that when present in a genetic background expressing the I-Ag7 MHC Class II molecule, leads to the development of arthritis (2). Arthritis can be induced either spontaneously, by breeding KRN with mice expressing I-Ag7 (K/BxN model), or by transferring serum from arthritic mice into any naïve strain of mice (serum transfer model) (3). In K/BxN mice, the autoreactive T and B cells both recognize the glycolytic enzyme glucose-6-phosphate-isomerase (GPI) as an autoantigen and disease severity correlates with rising titers of anti-GPI Ig in the serum (36). However, as in human RA, the factors responsible for triggering the initiating autoimmune response in K/BxN mice are unknown.

Indoleamine-2,3-dioxygenase (IDO) is an IFN-γ inducible enzyme, that catalyzes the initial and rate limiting step in the degradation of tryptophan (7, 8). An immunoregulatory role for IDO was suggested by the observation that administration of the bioactive IDO inhibitor, 1-methyl-tryptophan (1MT) (9), elicited MHC-restricted, T cell-mediated rejection of allogeneic mouse concepti (10, 11). IDO has also been shown to be a critical driver of immune escape in cancer (12). This, coupled with data that IDO could suppress activation of effector T cells in vitro (13), led to the concept of IDO as an immunosuppressive actor involved in the establishment of acquired peripheral immune tolerance.

If IDO were simply immunosuppressive, then it would be expected to play an inhibitory role in autoimmune responses. Indeed, this is consistent with some reports using 1MT in the context of inducible mouse models of autoimmunity, including experimental autoimmune encephalomyelitis (EAE), collagen induced arthritis (CIA), and trinitrobenzene sulphonic acid (TNBS) induced colitis (1416). However, other data, such as that reported in a mouse model of inflammatory airway disease, show IDO can also play an activating role in driving TH2-mediated inflammatory responses (17). These data appear to be more in line with the countervailing hypothesis that increased IDO activity may, in some instances, contribute positively to inflammatory responses. This may be the more relevant model with regard to autoimmunity in humans as elevated tryptophan degradation has been shown to correlate with disease activity in both RA and systemic lupus erythematosus (SLE) patients (18, 19).

The first direct evidence that IDO could contribute to inflammatory disease pathology was the recent finding that elevated IDO is an integral component of the severe cutaneous inflammation produced by topical application of PMA, necessary for supporting tumor outgrowth (20). Along these lines, we report here finding that IDO activity is also elevated in the serum of K/BxN mice at the earliest stages of joint inflammation. Importantly, the onset of arthritis was delayed and disease severity alleviated by treatment of these mice with the IDO inhibitory compound 1MT at this early stage of disease progression. In contrast, if 1MT was administered after this timepoint, it was no longer effective in treating joint inflammation. The alleviation of joint inflammation with 1MT was not due to a reduction in T regulatory cells or an altered T helper cell cytokine profile, but resulted from a diminished autoreactive B cell response. These results provide the first indication that IDO can contribute to the development of autoimmune disease pathology by supporting the activation of autoreactive B cells, adding to the growing body of evidence that IDO does not simply counteract inflammation through its ability to suppress T cells, but rather is a key constituent in the complex milieu of factors that shape the inflammatory environment.

Materials and Methods


KRN TCR Tg mice on a C57BL/6 background have been described (2). NOD mice were purchased from Jackson laboratories. To obtain arthritic mice, homozygous KRN Tg C57BL/6 mice were crossed with NOD mice yielding KRN (C57BL/6 × NOD)F1 mice designated K/BxN. To obtain arthritic KRN B6.g7 mice, KRN C57BL/6 mice were crossed with C57BL/6 mice expressing the IAg7 MHC Class II molecule to yield KRN C57BL/6 mice expressing IAb/g7 (termed KRN B6.g7). All mice were bred and housed under specific pathogen free conditions in the animal facility at the Lankenau Institute for Medical Research. Studies were performed in accordance with National Institutes of Health and Association for Assessment and Accreditation of Laboratory Animal Care guidelines with approval from the LIMR Institutional Animal Care and Use Committee.

Arthritis Induction by K/BxN Serum Transfer

Serum was collected and pooled from 8 wk old arthritic K/BxN mice. To induce arthritis, 150μl serum was injected i.p. into naive C57BL/6 mice on day 0. Arthritis induced by this method is transient, beginning 48 hr after serum transfer and resolving 2–3 weeks later (3).

Administration of 1MT

Mice were given 400 mg/kg/dose (100μl total volume) D/L-1MT (Sigma) diluted in Methocel/Tween (0.5% Tween 80, 0.5% methylcellulose (v/v in water; Sigma) twice daily by oral gavage (p.o.) using a curved feeding needle (20G × 1 ½ in; Fisher) as described (21). The p.o. dose of 1MT (400mg/kg) was selected based on PK studies done previously in the lab. A dose titration of 1MT (50mg/kg – 800mg/kg) performed at 1 hour postadministration (Cmax for 1MT) showed that the maximum amount of 1MT in the serum (99.1 ± 6.4 μM) is achieved at the 400mg/kg dose. We have confirmed, using K/BxN mice, that the maximum level of 1MT in the serum of this strain is also achieved at 400mg/kg and that increasing the dosage to 800mg/kg does not further increase the serum level of 1MT. Control mice were given an equal volume of carrier alone (Methocel/Tween). 1MT was administered on a b.i.d. schedule, once in the morning and once in the evening Monday – Friday. For the initial experiments, 1MT (2mg/ml) was also administered in water bottles on the weekends. No difference was seen between mice that received 1MT via water bottles on weekends and mice that were dosed only Monday – Friday. Therefore, in later experiments, mice were only given 1MT by oral gavage Monday – Friday. For the mice that spontaneously develop arthritis (K/BxN and KRN B6.g7), 1MT treatment was started at 21 days of age. For mice whose arthritis was induced by serum transfer, 1MT treatment was started 6h prior to the administration of arthritic serum.

Arthritis incidence

The two rear ankles of K/BxN or KRN B6.g7 mice were measured starting at weaning (3 wk of age). Measurement of ankle thickness was made above the footpad axially across the ankle joint using a Fowler Metric Pocket Thickness Gauge. Ankle thickness was rounded off to the nearest 0.05mm. At the termination of the experiment, ankles were fixed in 10% buffered formalin for 48 hrs, decalcified in 14% EDTA for 2wks, embedded in paraffin, sectioned, and stained with H&E.



Mice were bled once a week between 3 and 11 weeks of age. Sera were stored at −20°C before analysis. Serum samples were plated at an initial dilution of 1:100 and diluted serially 1:5 in Immunlon II plates coated with GPI-his (2μg/ml). Recombinant GPI-his protein was generated and purified as described previously. The serum titer was defined as the reciprocal of the last dilution that gave an OD>3x background.


Tissue culture supernatants were plated in triplicate undiluted on Immulon II plates coated with unlabeled donkey anti-mouse total Ig (Jackson Immunoresearch). Purified mouse IgM, IgG1, IgG2a, IgG2b, and IgG3 (Southern Biotechnology Associates) were used to generate standard curves. The total amount of Ig in the supernatant was calculated from the standard curve using Prism 4 software (GraphPad Software, Inc.).

For all ELISAs

Donkey anti-mouse total Ig- HRP (Jackson Immunoresearch), goat anti-mouse IgM-HRP, IgG1-HRP, IgG2a-HRP, IgG2b-HRP, IgG2c-HRP, or IgG3-HRP (Southern Biotechnology Associates) were used as secondary Abs. Ab was detected using ABTS substrate (Fisher).

ELISpot Assay

Spleen or LN cells were plated at 4 × 105 cells per well and diluted serially 1:4 in Multiscreen HA mixed cellulose ester membrane plates (Millipore) coated with GPI-his (2μg/ml). The cells were incubated on the Ag-coated plates for 4hr at 37°C. The Ig secreted by the plated cells was detected by Alkaline Phosphatase (AP)-conjugated goat anti-mouse total Ig secondary Ab (Southern Biotechnology Associates) and visualized using NBT/BCIP substrate (nitroblue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate; Sigma).

Cytokine Secretion

Cells from the draining LNs of carrier or control-treated K/BxN mice were harvested and cultured in either media alone or PMA (50 ng/ml) + ionomycin (500 ng/ml) for 16 h. The supernatants were then harvested and analyzed for the levels of IL-2, IL-4, IL-5, IL-6, IL-12, TNFα, IFNγ, and MCP-1 by cytometric bead array (BD Biosciences) and IL-17 by ELISA.

Cytometric Bead Array

For the cytometric bead array, samples were stained according to manufacturers instructions and analyzed on a FACSCanto II flow cytometer (BD Biosciences) using FACSDIVA software (BD Biosciences). Cytokine concentrations were calculated by comparing to standard curves using CBA analysis software (BD Biosciences).

Anti-IL-17 ELISA

Tissue culture supernatants were plated in duplicate on anti-IL-17-coated wells (clone TC11-18H10). IL-17 was detected with a biotinylated anti-IL-17 secondary Ab (clone TC11-8H4.1, BD Biosciences), followed by Streptavidin-HRP. ABTS was used as a substrate. The amount of IL-17 in the supernatant was calculated by comparison to a standard curve generated with recombinant IL-17 (R&D) using Prism software (GraphPad Software, Inc).

Western Blotting

Joint draining LNs (popliteal, axillary, and brachial) were harvested from K/BxN mice aged 4 wk (pre-arthritic, early arthritic), 6wk (acute arthritic), and 8wk (chronic arthritic). Epididymis from K/BxN and IDO deficient mice were used as positive and negative controls, respectively. Pooled LNs from 2 mice each or epididymus were frozen in LN2, mechanically disrupted with a mortar and pestal, and lysed in RIPA buffer containing protease and phosphatase inhibitors. 1 mg total protein for each LN sample and 50 μg from each epididymis sample were immunoprecipitated with 1μg affinity-purified polyclonal rabbit α-IDO1 (gift of R. Metz, NewLink Genetics), separated by SDS-PAGE and blotted to Immobilon-NC membrane (Millipore). The blots were then incubated with 2.5 μg monoclonal Rat α-IDO (clone mIDO-48; BioLegend), and detected with HRP-conjugated anti-Rat Ig (Southern Biotechnology Associates) using ECL reagent (Pierce) according to manufacturer’s instructions.

Kynurenine Assay

Serum was collected from pre-arthritic, early arthritic, acute arthritic, and chronic arthritic K/BxN mice. The serum was diluted in water (1:4 v/v), deproteinated and analyzed by HPLC coupled to electrospray ionization tandem mass spectroscopy (LC/MS/MS) analysis as described (22). Quantitation of kynurenine was based on analysis of 2 daughter ions.

T regulatory cell assay

Joint draining LNs were harvested from carrier or 1MT-treated K/BxN mice. 1×106 cells were stained with anti-CD4-PE-Cy7 (GK1.5) and anti-CD25-APC (3C7; BioLegend), fixed/permeabilized with Fixation/Permeabilization solution, and then stained intracellularly with anti-Foxp3-PE (FJK-16s; eBiosciences). Samples were analyzed on a FACSCanto II flow cytometer (BD Biosciences) using FACSDIVA software (BD Biosciences). Data was analyzed using Cellquest software (BD Biosciences). Gating on live lymphocytes was based on forward and side scatter, with 100,000 events collected for each sample.

In Vitro Stimulation

B cells from non-tg C57BL/6 spleens were purified using the MACS separation system with paramagnetic anti-CD43 beads (Miltenyi Corporation). Cell purity was >95%. In some experiments, the cells were also labeled with 5μM CFSE (Invitrogen). 106 B cells were cultured for 3 days in a 1ml total volume in either media alone (Iscoves DME, 10% FCS, 5×10−5 M 2-mercaptoethanol), 0.5–25 μg/ml LPS (Escherichia coli strain 0111:B4; Sigma Aldrich), 1–50μg/ml goat anti-mouse IgM (Fab′)2 (Jackson Immunoresearch), or 100–2000ng/ml anti-CD40 (clone 1C10; R&D Systems) + 50ng/ml IL-4 (eBioscience). The IDO inhibitor D/L-1MT (in DMSO + 0.1N HCl) or vehicle alone was added at a final concentration of 100μM.

Statistical Analysis

Statistical significance was determined using an unpaired Student’s t test or the Mann-Whitney nonparametric test and Instat Software (GraphPad Software, Inc)


1MT delays the rate of arthritis development and alleviates its severity

To begin to define the role that IDO activation plays in shaping the inflammatory autoimmune response in K/BxN mice, we inhibited its activity pharmacologically using the racemic mix of 1MT (D/L-1MT). K/BxN mice were given 1MT, or carrier alone starting at 21 days of age, prior to the initiation of arthritis. Mice were monitored for arthritis development by measuring joint inflammation by the change in ankle thickness and synovial proliferation and inflammatory cell infiltrates by histology (Fig. 1). Consistent with previous observations in untreated animals (5), K/BxN mice that received carrier alone developed severe inflammation in their front and rear paws that began between 28 and 35 days of age. In contrast, the rate of inflammation was slower and severity reduced in mice that had IDO activity blocked with 1MT (Fig. 1a). The incidence of arthritis was also reduced in 1MT-treated mice (Fig. 1a). At the termination of the experiment (7wk of age), rear ankles were harvested and examined for histological evidence of arthritis by staining with hematoxylin and eosin (Fig. 1b). Carrier-treated mice showed classic signs of arthritis, with a greatly expanded synovium, panus formation, and inflammatory cell infiltrates. Similar to what was observed by measuring ankle thickness, joints from 1MT-treated mice showed a reduction in the severity of arthritis with minimal synovial expansion and fewer infiltrating inflammatory cells. Consistent with the reduction in arthritis, 1MT treatment also reduced the levels of the inflammatory cytokines MCP-1 and IL-6 produced in the draining LNs. Levels of IL-10 were also reduced (Fig. 1c).

Figure 1
1MT inhibits arthritis development and inflammatory cytokine production

The K/BxN model is an F1 between C57BL/6 and NOD. Several immune abnormalities have been described in NOD mice, including decreased numbers of T regulatory cells and reduced APC function (23, 24). Additionally, young female NOD mice have been reported to have a defect in IDO activation, attributable to a peroxynitrite-induced blockade of IFNγ signaling in dendritic cells (25). These abnormalities could complicate the interpretation of the 1MT studies in K/BxN mice. Therefore, we repeated the 1MT experiments using KRN B6.g7 mice. KRN B6.g7 mice are C57BL/6 mice that express both the KRN TCR tg and the IAg7 MHC Class II molecule necessary for KRN T cell activation, but lack the rest of the NOD-associated genes (2). KRN B6.g7 mice were given 1MT or carrier alone, starting at 3 weeks of age and monitored for arthritis development (Fig. 2a). Similar to what was seen in K/BxN mice, 1MT delayed the rate of onset and attenuated the severity of arthritis in KRN B6.g7 mice. Therefore, the anti-arthritic effect of 1MT is independent of the NOD genetic background.

Figure 2
1MT exerts its effect early in the response

1MT exerts its inhibitory effect early in the response

Arthritis in K/BxN mice occurs in two phases, the initiation phase that is dependent on GPI-specific T and B cells (3), and the effector phase that occurs once anti-GPI Ab is produced and is dependent upon neutrophils, macrophages, and mast cells (2628). 1MT inhibition of IDO could affect one or both of these stages to block arthritis development. In the K/BxN model, these two phases can be separated experimentally by transferring serum from arthritic mice to naïve non-tg recipients, bypassing the T and B cell dependent initiation phase (29). To test whether 1MT could inhibit the effector stage, arthritis induced by serum transfer was compared in carrier and 1MT-treated C57BL/6 mice (Fig 2b). Carrier-treated mice began to develop inflammation in their front and rear paws 2 days after serum transfer, with a peak of inflammation 8 days post-transfer. 1MT-treated mice developed arthritis with identical kinetics, demonstrating that 1MT was unable to inhibit the effector phase of arthritis.

To begin to define the mechanism by which 1MT inhibited the initiation phase of arthritis development in K/BxN mice, we determined when 1MT administration was required during the course of arthritis to produce an efficacious response. First, we tested whether 1MT was able to inhibit an ongoing arthritic response. K/BxN mice were allowed to develop arthritis and then given 1MT or carrier alone (Fig. 2c). When administered at this stage, 1MT had no significant effect on arthritis development, corroborating the serum transfer experiment, and thus indicating that 1MT needed to be present earlier in the response to have its anti-arthritic effect. Next, we tested whether treatment with 1MT during the establishment of disease would be sufficient to produce an anti-arthritic effect. To accomplish this, K/BxN mice were given 1MT or carrier alone for 10 days, at which time treatment was stopped and mice were monitored for arthritis development (Fig. 2d). When administered for just this short time period, 1MT was able to significantly attenuate arthritis in the K/BxN mice, demonstrating that blocking IDO activity with 1MT during the initiation phase of the response was sufficient to affect the course of arthritis.

IDO is most active early in the arthritic response

1MT was only effective if administered early in the arthritic response, and this short-term exposure was sufficient to attenuate arthritis development even if the treatment was stopped. This suggested that IDO expression or activity was essential during the initiation of arthritis and either was inhibited or was unnecessary once arthritis had been established. To evaluate IDO expression during the course of arthritis, joint-draining LNs were harvested from K/BxN mice at the very onset of arthritis development (pre/early), during the aggressive inflammatory stage (acute), or during established arthritis (chronic). IDO protein was immunoprecipitated from whole LN lysates and detected by western blotting (Fig. 3a). IDO protein was detectable at the earliest stage of arthritis and remained present throughout arthritis development, ruling out differential protein expression as an explanation for 1MT’s effects. To measure IDO activity, serum samples from K/BxN mice were analyzed for kynurenine levels (Fig. 3b), a breakdown product indicative of IDO-mediated tryptophan catabolism. A spike in serum kynurenine levels was detected in mice just after the onset of arthritis. Kynurenine levels decreased once arthritis became established in the acute and chronic stages. These data indicate that, although IDO was expressed throughout the course of arthritis development, it was most active at the initiation of the arthritic response. Therefore, 1MT was effective at inhibiting arthritis development only during the time when IDO was most active, at the initiation of the autoimmune response.

Figure 3
IDO is expressed throughout, but has highest activity early in arthritis

1MT does not affect TH1/TH2/TH17 cytokine responses

One potential mechanism by which 1MT could inhibit arthritis development early in the autoimmune response is by skewing the cytokine profile of the T cells responsible for its initiation. IDO has been shown to be a modifier of the cytokine profile of T cell responses in vitro (30). The precise mechanism by which IDO exerts this effect is unknown, but is thought to involve tryptophan depletion and/or tryptophan catabolites, which have been shown to preferentially affect the survival of TH1, but not TH2 cells (31). Their effect on TH17 cells has not been assessed.

TNFα and more recently, IL-17 have been proposed to be key cytokines involved in both human and other mouse models of RA (3235). The cytokine profile of the anti-GPI T cells driving the arthritogenic response in K/BxN mice is not well characterized, although IL-4 is thought to be important to the disease process. This is because serum anti-GPI Abs are predominantly IgG1, an IL-4 related isotype, and K/BxN mice deficient in IL-4 show reduced disease (36). To determine which cytokines were expressed during arthritis development in K/BxN mice and whether 1MT altered this cytokine profile, we measured the cytokine profile secreted by T cells in the LNs draining the arthritic joints by ELISA and/or the flow cytometry-based cytometric bead array (Fig. 4). In cultures from control-treated K/BxN mice, the TH1 cytokine IFNγ was detected at high levels and TH2 cytokine IL-5 was present at a low level. IL-17 and TNFα were both present at high levels, a profile associated with TH17 cells. This is consistent with the increasing evidence implicating both TH1 and TH17 cells as being promoters of autoimmunity (37). In contrast to what was predicted by the serum anti-GPI IgG1 profile, levels of IL-4 were barely detectable above background. Surprisingly, 1MT did not affect this cytokine profile. High levels of IFNγ, IL-17, and TNFα, a low level of IL-5, and almost no IL-4 were secreted by LN cells from 1MT-treated K/BxN mice (Fig. 4). Together, these data indicate that 1MT does not inhibit arthritis development by skewing the TH cytokine profile of the autoimmune response.

Figure 4
1MT treatment does not alter TH1/TH2/TH17 cytokine levels

1MT-treated K/BxN mice have a normal frequency of T regulatory cells

T regulatory cells (Tregs) have been shown to be important in controlling the aggressiveness of autoimmunity in the K/BxN model (38, 39). Therefore, treatment with the IDO inhibitor 1MT, which we have shown elicited a reduction in the level of IL-6 in the joint-draining LNs of K/BxN mice (Fig. 1C), might be inhibiting arthritis development through suppression of Tregs, given that Treg development is promoted by IDO and inhibited by IL-6 (40, 41). To test this, the percentage of Tregs was quantitated by flow cytometry and compared in 1MT vs. carrier-treated K/BxN mice (Table I). Tregs were elevated in the dLN compared to the non-dLN in carrier-treated K/BxN mice (p=0.03) as has been reported for non-treated K/BxN mice (38, 42). Treg frequencies were also increased in the dLNs of 1MT-treated K/BxN mice (p<0.0001). However, there was no difference in the percentage of Tregs in either dLNs (p=0.8) or non-dLNs (p=0.3) in 1MT compared to carrier-treated mice. Therefore, decreased disease activity in 1MT-treated mice could not be attributed to a reduction in the percentage of Tregs.

Table I
1MT treatment does not affect the percentage of Tregs

1MT inhibits the anti-GPI B cell response

Antibodies that recognize the glycolytic enzyme GPI are the key effector molecules in the disease process in K/BxN mice (4). Previously, we demonstrated that this pathogenic anti-GPI B cell response was focused to the LNs draining the arthritic joints (5). To address if the reduced arthritis in 1MT-treated mice was the result of a diminished GPI-specific B cell response, we measured the titers of anti-GPI Ig in the serum of 1MT- and carrier-treated mice (Fig. 5). Titers of anti-GPI Ig (Fig. 5a) were significantly lower in 1MT compared to carrier treated mice, particularly those of the IgG1 isotype required for disease initiation. The effect of 1MT appeared to be specific to GPI-reactive B cells, as total serum Ig levels were not different between carrier and 1MT-treated animals (Fig. 5b).

Figure 5
1MT inhibits the anti-GPI B cell response

To determine if the reduced serum anti-GPI levels were due to 1MT treatment affecting the location or magnitude of the anti-GPI B cell response, the number of antibody secreting cells (ASCs) was quantitated in the spleen, dLNs, and non-dLNs of carrier vs. 1MT-treated K/BxN mice (Fig. 5c). GPI-reactive ASCs were present in all three locations in carrier-treated K/BxN mice, with the largest number in the dLNs. Numbers of anti-GPI ASCs were significantly reduced in both the spleen and dLN of 1MT-treated mice. This reduction in serum anti-GPI titers and numbers of anti-GPI ASCs was found whether the 1MT was administered continuously (long-term) or was stopped at the onset of arthritis (short-term). Together, these data demonstrate that inhibition of IDO activity with 1MT at the earliest stage of the autoimmune response is sufficient to diminish the subsequent pathogenic anti-GPI B cell response and result in reduced arthritis.

B cells, like most antigen presenting cells, express IDO. Therefore, 1MT could be acting directly on the B cells or inhibiting their activation through an indirect mechanism. To distinguish between these two possibilities, the ability of 1MT to affect B cell activation and differentiation was measured in vitro (Fig. 6). Purified B cells were labeled with CFSE and cultured in vitro with media alone, LPS, anti-IgM (Fab′)2 fragments, or the T cell mimic anti-CD40 + IL-4 in the presence or absence of 1MT. B cells proliferated robustly to LPS, anti-IgM, and anti-CD40 + IL-4. 1MT had no effect on either the number of cells dividing, or the number of cell divisions (Fig. 6a). The effect of 1MT on Ab secretion was determined by measuring the amount of Ig secreted into the culture supernatants (Fig. 6b). Ig was detected at equal levels in cultures with and without 1MT. Therefore, at least in vitro, 1MT does not affect directly affect B cell activation. These data suggest that 1MT affects the pathogenic B cell response in K/BxN mice through an indirect mechanism.

Figure 6
1MT does not directly affect B cell activation in vitro


The role that IDO plays in regulating immune responses has been the subject of intense investigation. The bulk of the literature has focused on investigating the suppressive effects of IDO activity, predominantly on the activation of T cells (43). The prevailing theory is that IDO expressed by dendritic cells inhibits T cell activation, either directly or indirectly by driving the development of Tregs (30, 44, 45). In contrast to IDO’s effect on T cell responses, the role that IDO may play in B cell responses has not been evaluated. In this study, we show that administration of 1MT to K/BxN mice reduced inflammatory cytokines and autoantibodies, resulting in an attenuated course of arthritis. Surprisingly, no difference was detected in the percentage of Tregs, nor in the levels of TH1/TH2/TH17 cytokines. Instead, the main effect of 1MT appeared to be to suppress the autoreactive B cell response. Our findings suggest that IDO is not simply an immunosuppressive enzyme, but rather plays a more complex role that includes supporting the establishment of B cell-mediated inflammatory responses.

RA patients show evidence of elevated IDO activity that correlates with disease activity (18, 46), but it has been unclear what relevance this has, if any, to the autoimmune response. In K/BxN mice, IDO activity was highest at the initiation of arthritis, and treatment with the pharmacological inhibitor of IDO, 1MT, at this early stage delayed the development of arthritis and reduced disease severity. Importantly, 1MT exposure was required only during the initiation of arthritis to exert its protective effect. In fact, starting 1MT treatment after disease initiation was no longer effective. There is precedence for short-term exposure to 1MT having a lasting effect on immune cell function (20, 47). Therefore, in K/BxN mice, we suggest that IDO plays an activating role in establishing the autoreactive B cell profile at the onset of the autoimmune response. If IDO activity is inhibited at this critical stage, the autoreactive B cell profile is not established and subsequent joint inflammation and damage is reduced.

Although 1MT was effective at alleviating arthritis, 1MT treatment did not completely prevent arthritis development, as most mice developed an attenuated course of disease. This study, like most in the literature, used a pharmacological agent to inhibit IDO activity. A potential caveat of pharmacological inhibitors is that they may not be fully effective at inhibition or may have off-target effects. Indeed, 1MT can also inhibit the IDO-related enzyme IDO2 (48). Additionally, there may be an underlying biological difference between constitutive loss of IDO due to genetic ablation versus acute loss through pharmacologic inhibition. This is consistent with studies in pregnancy and tumor models in which compensatory mechanisms for maintaining tolerance that apparently come into play in IDO-deficient mice are not as effectively engaged following IDO inhibitor treatment (49, 50). To address these possibilities, it will be important to evaluate the impact of genetic loss of IDO and/or IDO2 on the development of arthritis in the K/BxN model.

1MT has been used in several other inflammatory disease models with conflicting results. 1MT exacerbated disease in experimental autoimmune encephalomyelitis (EAE) and trinitrobenzene sulphonic acid (TNBS) induced colitis (14, 16). In collagen induced arthritis (CIA), one study showed accelerated disease upon administration of 1MT (15). However, another study showed 1MT had no effect on its own, but did reverse the protective effect of immunotherapy with an antibody to the B7 family molecule 4-1BB (51). In contrast to the disease exacerbating effect of 1MT in these models, 1MT administration was protective in a mouse model of allergic airway inflammation (17). In this case, the disease-initiating TH2 response was inhibited in 1MT-treated mice, suggesting that IDO normally promotes TH2-mediated inflammatory responses. A similar response was reported in vitro where IDO was shown to inhibit TH1 responses and promote TH2 responses (31). We have likewise shown in this study that 1MT treatment was also protective against joint inflammation in K/BxN mice. However, in this case, the protective effect of 1MT was not due to a skewing of the TH cytokine profile. The autoimmune response in K/BxN mice exhibited characteristics of TH1, TH2, and TH17 responses (Fig. 4 and ref. (36)) and 1MT treatment did not affect this cytokine profile.

In contrast to the TH1/TH2/TH17 cytokines, cytokines associated with inflammation, MCP-1, IL-6, and IL-10, were reduced in 1MT-treated mice. MCP-1, a cytokine that plays a key role in recruiting monocytes into sites of inflammation, has been shown to be elevated in RA patients (52). Likewise, IL-6, a cytokine thought to induce inflammatory joint destruction through the recruitment and induction of inflammatory TH17 cells, is also elevated in RA patients (53, 54). MCP-1 and IL-6 were also both elevated in control-treated and reduced in 1MT-treated K/BxN mice. Therefore, the alleviation of inflammation in 1MT-treated K/BxN mice was reflected in a reduction of contributory cytokines. At this point, it is not known whether 1MT treatment caused a reduction in IL-6 and MCP-1 levels directly, or if the reduced levels were simply a consequence of the overall reduced inflammatory response. IL-10 levels have also been shown to increase in response to inflammation, however, unlike MCP-1 and IL-6, IL-10 serves to dampen the response (55). Although the significance of the reduced IL-10 levels is not clear, it may also be the result of the overall reduction in the inflammatory response in the 1MT-treated mice.

The most dramatic effect of 1MT treatment was the reduction observed in the autoreactive B cell response. Autoantibody secreting B cell numbers were significantly decreased and titers of anti-GPI Ab in the serum were greatly reduced in 1MT-treated K/BxN mice. A role for IDO in driving B cell responses has not been previously appreciated. B cells, like most APCs, express IDO and levels increase upon activation (unpublished observations). Our experiments do not distinguish between 1MT having a direct effect on autoreactive B cells or inhibiting their activation by an indirect mechanism. However, our in vitro experiments demonstrate that 1MT does not directly inhibit the activation of non-autoreactive B cells. IDO expression in another cell type could affect the environment required for efficient B cell activation and Ab secretion. Macrophages and dendritic cells, in particular plasmacytoid dendritic cells, have been implicated in the IDO-mediated suppression of T cells (13, 56, 57). However, we were unable to detect any difference in the percentage or activation status of these cells in carrier- vs. 1MT-treated K/BxN mice (unpublished observations). Future experiments will be directed at identifying the cell type(s) responsible for the 1MT-mediated suppression of arthritis in the K/BxN model.

Recent evidence has shown that topical application of the pro-inflammatory agent PMA drives IDO activity in the regional lymph nodes and that this was a key component of the inflammatory microenvironment required for supporting tumor outgrowth following carcinogen exposure (20). The elevation of IDO in response to PMA in these studies was interpreted as paradoxical because IDO was considered to be immunosuppressive and yet no indication that IDO was having a negative impact on the development or severity of PMA-driven inflammation was observed. In light of the current study, it is clear that categorizing IDO strictly as an immunosuppressive enzyme is an oversimplification and that its involvement in disease processes such as cancer and autoimmune disorders will be much more complex. In particular, its role in driving the activation of autoreactive B cells may have broad clinical implications for the future utility of IDO inhibitors as potential therapeutic agents.


The authors would like to thank Dr. Paul Allen (Washington University) for the KRN C57BL/6 and B6.g7 mice, Dr. Richard Metz (NewLink Genetics) for the anti-IDO1 polyclonal Ab and helpful discussions, and Dr. Lisa Laury-Kleintop (LIMR) for critical reading of the manuscript and thoughtful input.


1Grant support: This work is supported by the Lankenau Hospital Foundation (LM-N and GCP). NR is supported by a research assistantship fellowship from the Brook J. Lenfest Foundation. AJM is the recipient of grants from the DoD Breast Cancer Research Program (BC044350), the Concern Foundation, and the Lance Armstrong Foundation, and GCP is the recipient of NIH grants CA109542, CA82222 and CA100123.

3Nonstandard abbreviations: ASC, antibody secreting cell; dLN, draining LN; GPI, glucose-6-phosphate isomerase; IDO, indoleamine-2,3-dioxygenase; K/BxN, KRN (C57BL/6 x NOD)F1; 1MT, 1-methyl-tryptophan; RA, rheumatoid arthritis; Treg, T regulatory cell


A.J.M. and G.C.P. have intellectual property interests (patents, and/or license fees through the authors’ institutions) in the therapeutic use of IDO and IDO inhibitors in cancer. Additionally these same authors are members of the Scientific Advisory Board for NewLink Genetics Inc. and receive consulting income and/or have financial holdings from this source.


1. Feldmann M, Brennan FM, Maini RN. Rheumatoid arthritis. Cell. 1996;85:307–310. [PubMed]
2. Kouskoff V, Korganow AS, Duchatelle V, Degott C, Benoist C, Mathis D. Organ-specific disease provoked by systemic autoimmunity. Cell. 1996;87:811–822. [PubMed]
3. Korganow AS, Ji H, Mangialaio S, Duchatelle V, Pelanda R, Martin T, Degott C, Kikutani H, Rajewsky K, Pasquali JL, Benoist C, Mathis D. From systemic T cell self-reactivity to organ-specific autoimmune disease via immunoglobulins. Immunity. 1999;10:451–461. [PubMed]
4. Matsumoto I, Staub A, Benoist C, Mathis D. Arthritis provoked by linked T and B cell recognition of a glycolytic enzyme. Science. 1999;286:1732–1735. [PubMed]
5. Mandik-Nayak L, Wipke BT, Shih FF, Unanue ER, Allen PM. Despite ubiquitous autoantigen expression, arthritogenic autoantibody response initiates in the local lymph node. Proc Natl Acad Sci U S A. 2002;99:14368–14373. [PubMed]
6. Ji H, Gauguier D, Ohmura K, Gonzalez A, Duchatelle V, Danoy P, Garchon HJ, Degott C, Lathrop M, Benoist C, Mathis D. Genetic influences on the end-stage effector phase of arthritis. J Exp Med. 2001;194:321–330. [PMC free article] [PubMed]
7. Shimizu T, Nomiyama S, Hirata F, Hayaishi O. Indoleamine 2,3-dioxygenase. Purification and some properties. J Biol Chem. 1978;253:4700–4706. [PubMed]
8. Yoshida R, Hayaishi O. Induction of pulmonary indoleamine 2,3-dioxygenase by intraperitoneal injection of bacterial lipopolysaccharide. Proc Natl Acad Sci U S A. 1978;75:3998–4000. [PubMed]
9. Cady SG, Sono M. 1-Methyl-DL-tryptophan, beta-(3-benzofuranyl)-DL-alanine (the oxygen analog of tryptophan), and beta-[3-benzo(b)thienyl]-DL-alanine (the sulfur analog of tryptophan) are competitive inhibitors for indoleamine 2,3-dioxygenase. Arch Biochem Biophys. 1991;291:326–333. [PubMed]
10. Munn DH, Zhou M, Attwood JT, Bondarev I, Conway SJ, Marshall B, Brown C, Mellor AL. Prevention of allogeneic fetal rejection by tryptophan catabolism. Science. 1998;281:1191–1193. [PubMed]
11. Mellor AL, Sivakumar J, Chandler P, Smith K, Molina H, Mao D, Munn DH. Prevention of T cell-driven complement activation and inflammation by tryptophan catabolism during pregnancy. Nature immunology. 2001;2:64–68. [PubMed]
12. Muller AJ, Scherle PA. Targeting the mechanisms of tumoral immune tolerance with small-molecule inhibitors. Nat Rev Cancer. 2006;6:613–625. [PubMed]
13. Munn DH, Shafizadeh E, Attwood JT, Bondarev I, Pashine A, Mellor AL. Inhibition of T cell proliferation by macrophage tryptophan catabolism. J Exp Med. 1999;189:1363–1372. [PMC free article] [PubMed]
14. Sakurai K, Zou JP, Tschetter JR, Ward JM, Shearer GM. Effect of indoleamine 2,3-dioxygenase on induction of experimental autoimmune encephalomyelitis. J Neuroimmunol. 2002;129:186–196. [PubMed]
15. Szanto S, Koreny T, Mikecz K, Glant TT, Szekanecz Z, Varga J. Inhibition of indoleamine 2,3-dioxygenase-mediated tryptophan catabolism accelerates collagen-induced arthritis in mice. Arthritis research & therapy. 2007;9:R50. [PMC free article] [PubMed]
16. Gurtner GJ, Newberry RD, Schloemann SR, McDonald KG, Stenson WF. Inhibition of indoleamine 2,3-dioxygenase augments trinitrobenzene sulfonic acid colitis in mice. Gastroenterology. 2003;125:1762–1773. [PubMed]
17. Xu H, Oriss TB, Fei M, Henry AC, Melgert BN, Chen L, Mellor AL, Munn DH, Irvin CG, Ray P, Ray A. Indoleamine 2,3-dioxygenase in lung dendritic cells promotes Th2 responses and allergic inflammation. Proc Natl Acad Sci U S A. 2008;105:6690–6695. [PubMed]
18. Schroecksnadel K, Winkler C, Duftner C, Wirleitner B, Schirmer M, Fuchs D. Tryptophan degradation increases with stage in patients with rheumatoid arthritis. Clin Rheumatol. 2006;25:334–337. [PubMed]
19. Pertovaara M, Hasan T, Raitala A, Oja SS, Yli-Kerttula U, Korpela M, Hurme M. Indoleamine 2,3-dioxygenase activity is increased in patients with systemic lupus erythematosus and predicts disease activation in the sunny season. Clin Exp Immunol. 2007;150:274–278. [PubMed]
20. Muller AJ, Sharma MD, Chandler PR, Duhadaway JB, Everhart ME, Johnson BA, 3rd, Kahler DJ, Pihkala J, Soler AP, Munn DH, Prendergast GC, Mellor AL. Chronic inflammation that facilitates tumor progression creates local immune suppression by inducing indoleamine 2,3 dioxygenase. Proc Natl Acad Sci U S A. 2008;105:17073–17078. [PubMed]
21. Muller AJ, DuHadaway JB, Donover PS, Sutanto-Ward E, Prendergast GC. Inhibition of indoleamine 2,3-dioxygenase, an immunoregulatory target of the cancer suppression gene Bin1, potentiates cancer chemotherapy. Nat Med. 2005;11:312–319. [PubMed]
22. Amirkhani A, Heldin E, Markides KE, Bergquist J. Quantitation of tryptophan, kynurenine and kynurenic acid in human plasma by capillary liquid chromatography-electrospray ionization tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci. 2002;780:381–387. [PubMed]
23. Serreze DV, Gaskins HR, Leiter EH. Defects in the differentiation and function of antigen presenting cells in NOD/Lt mice. J Immunol. 1993;150:2534–2543. [PubMed]
24. Alard P, Manirarora JN, Parnell SA, Hudkins JL, Clark SL, Kosiewicz MM. Deficiency in NOD antigen-presenting cell function may be responsible for suboptimal CD4+CD25+ T-cell-mediated regulation and type 1 diabetes development in NOD mice. Diabetes. 2006;55:2098–2105. [PubMed]
25. Grohmann U, Fallarino F, Bianchi R, Orabona C, Vacca C, Fioretti MC, Puccetti P. A defect in tryptophan catabolism impairs tolerance in nonobese diabetic mice. J Exp Med. 2003;198:153–160. [PMC free article] [PubMed]
26. Wipke BT, Allen PM. Essential role of neutrophils in the initiation and progression of a murine model of rheumatoid arthritis. J Immunol. 2001;167:1601–1608. [PubMed]
27. Solomon S, Rajasekaran N, Jeisy-Walder E, Snapper SB, Illges H. A crucial role for macrophages in the pathology of K/B × N serum-induced arthritis. Eur J Immunol. 2005;35:3064–3073. [PubMed]
28. Lee DM, Friend DS, Gurish MF, Benoist C, Mathis D, Brenner MB. Mast cells: a cellular link between autoantibodies and inflammatory arthritis. Science. 2002;297:1689–1692. [PubMed]
29. Monach P, Hattori K, Huang H, Hyatt E, Morse J, Nguyen L, Ortiz-Lopez A, Wu HJ, Mathis D, Benoist C. The K/BxN mouse model of inflammatory arthritis: theory and practice. Methods in molecular medicine. 2007;136:269–282. [PubMed]
30. Mellor AL, Munn DH. IDO expression by dendritic cells: tolerance and tryptophan catabolism. Nature reviews. 2004;4:762–774. [PubMed]
31. Fallarino F, Grohmann U, Vacca C, Bianchi R, Orabona C, Spreca A, Fioretti MC, Puccetti P. T cell apoptosis by tryptophan catabolism. Cell Death Differ. 2002;9:1069–1077. [PubMed]
32. McInnes IB, Schett G. Cytokines in the pathogenesis of rheumatoid arthritis. Nature reviews. 2007;7:429–442. [PubMed]
33. Lubberts E, Koenders MI, Oppers-Walgreen B, van den Bersselaar L, Coenen-de Roo CJ, Joosten LA, van den Berg WB. Treatment with a neutralizing anti-murine interleukin-17 antibody after the onset of collagen-induced arthritis reduces joint inflammation, cartilage destruction, and bone erosion. Arthritis Rheum. 2004;50:650–659. [PubMed]
34. Chabaud M, Durand JM, Buchs N, Fossiez F, Page G, Frappart L, Miossec P. Human interleukin-17: A T cell-derived proinflammatory cytokine produced by the rheumatoid synovium. Arthritis Rheum. 1999;42:963–970. [PubMed]
35. Hsu HC, Yang P, Wang J, Wu Q, Myers R, Chen J, Yi J, Guentert T, Tousson A, Stanus AL, Le TV, Lorenz RG, Xu H, Kolls JK, Carter RH, Chaplin DD, Williams RW, Mountz JD. Interleukin 17-producing T helper cells and interleukin 17 orchestrate autoreactive germinal center development in autoimmune BXD2 mice. Nature immunology. 2008;9:166–175. [PubMed]
36. Ohmura K, Nguyen LT, Locksley RM, Mathis D, Benoist C. Interleukin-4 can be a key positive regulator of inflammatory arthritis. Arthritis Rheum. 2005;52:1866–1875. [PubMed]
37. Kikly K, Liu L, Na S, Sedgwick JD. The IL-23/Th(17) axis: therapeutic targets for autoimmune inflammation. Curr Opin Immunol. 2006;18:670–675. [PubMed]
38. Nguyen LT, Jacobs J, Mathis D, Benoist C. Where FoxP3-dependent regulatory T cells impinge on the development of inflammatory arthritis. Arthritis Rheum. 2007;56:509–520. [PubMed]
39. Kang SM, Jang E, Paik DJ, Jang YJ, Youn J. CD4+CD25+ regulatory T cells selectively diminish systemic autoreactivity in arthritic K/BxN mice. Mol Cells. 2008;25:64–69. [PubMed]
40. Fallarino F, Grohmann U, You S, McGrath BC, Cavener DR, Vacca C, Orabona C, Bianchi R, Belladonna ML, Volpi C, Santamaria P, Fioretti MC, Puccetti P. The combined effects of tryptophan starvation and tryptophan catabolites down-regulate T cell receptor zeta-chain and induce a regulatory phenotype in naive T cells. J Immunol. 2006;176:6752–6761. [PubMed]
41. Bettelli E, Carrier Y, Gao W, Korn T, Strom TB, Oukka M, Weiner HL, Kuchroo VK. Reciprocal developmental pathways for the generation of pathogenic effector TH17 and regulatory T cells. Nature. 2006;441:235–238. [PubMed]
42. Monte K, Wilson C, Shih FF. Increased number and function of FoxP3 regulatory T cells during experimental arthritis. Arthritis Rheum. 2008;58:3730–3741. [PubMed]
43. Mellor AL, Munn D, Chandler P, Keskin D, Johnson T, Marshall B, Jhaver K, Baban B. Tryptophan catabolism and T cell responses. Adv Exp Med Biol. 2003;527:27–35. [PubMed]
44. Moffett JR, Namboodiri MA. Tryptophan and the immune response. Immunol Cell Biol. 2003;81:247–265. [PubMed]
45. Puccetti P, Grohmann U. IDO and regulatory T cells: a role for reverse signalling and non-canonical NF-kappaB activation. Nature reviews. 2007;7:817–823. [PubMed]
46. Schroecksnadel K, Kaser S, Ledochowski M, Neurauter G, Mur E, Herold M, Fuchs D. Increased degradation of tryptophan in blood of patients with rheumatoid arthritis. J Rheumatol. 2003;30:1935–1939. [PubMed]
47. Grohmann U, Bianchi R, Orabona C, Fallarino F, Vacca C, Micheletti A, Fioretti MC, Puccetti P. Functional plasticity of dendritic cell subsets as mediated by CD40 versus B7 activation. J Immunol. 2003;171:2581–2587. [PubMed]
48. Metz R, Duhadaway JB, Kamasani U, Laury-Kleintop L, Muller AJ, Prendergast GC. Novel tryptophan catabolic enzyme IDO2 is the preferred biochemical target of the antitumor indoleamine 2,3-dioxygenase inhibitory compound D-1-methyl-tryptophan. Cancer Res. 2007;67:7082–7087. [PubMed]
49. Baban B, Chandler P, McCool D, Marshall B, Munn DH, Mellor AL. Indoleamine 2,3-dioxygenase expression is restricted to fetal trophoblast giant cells during murine gestation and is maternal genome specific. J Reprod Immunol. 2004;61:67–77. [PubMed]
50. Banerjee T, Duhadaway JB, Gaspari P, Sutanto-Ward E, Munn DH, Mellor AL, Malachowski WP, Prendergast GC, Muller AJ. A key in vivo antitumor mechanism of action of natural product-based brassinins is inhibition of indoleamine 2,3-dioxygenase. Oncogene. 2008;27:2851–2857. [PubMed]
51. Seo SK, Choi JH, Kim YH, Kang WJ, Park HY, Suh JH, Choi BK, Vinay DS, Kwon BS. 4–1BB-mediated immunotherapy of rheumatoid arthritis. Nat Med. 2004;10:1088–1094. [PubMed]
52. Melgarejo E, Medina MA, Sanchez-Jimenez F, Urdiales JL. Monocyte chemoattractant protein-1: A key mediator in inflammatory processes. Int J Biochem Cell Biol. 2009;41:998–1001. [PubMed]
53. Madhok R, Crilly A, Watson J, Capell HA. Serum interleukin 6 levels in rheumatoid arthritis: correlations with clinical and laboratory indices of disease activity. Ann Rheum Dis. 1993;52:232–234. [PMC free article] [PubMed]
54. Naugler WE, Karin M. The wolf in sheep’s clothing: the role of interleukin-6 in immunity, inflammation and cancer. Trends in molecular medicine. 2008;14:109–119. [PubMed]
55. Couper KN, Blount DG, Riley EM. IL-10: the master regulator of immunity to infection. J Immunol. 2008;180:5771–5777. [PubMed]
56. Fallarino F, Vacca C, Orabona C, Belladonna ML, Bianchi R, Marshall B, Keskin DB, Mellor AL, Fioretti MC, Grohmann U, Puccetti P. Functional expression of indoleamine 2,3-dioxygenase by murine CD8 alpha(+) dendritic cells. Int Immunol. 2002;14:65–68. [PubMed]
57. Baban B, Hansen AM, Chandler PR, Manlapat A, Bingaman A, Kahler DJ, Munn DH, Mellor AL. A minor population of splenic dendritic cells expressing CD19 mediates IDO-dependent T cell suppression via type I IFN signaling following B7 ligation. Int Immunol. 2005;17:909–919. [PubMed]