Abstract
Humoral responses to nonproteinaceous Ags (i.e., T cell independent [TI]) are a key component of the early response to bacterial and viral infection and a critical driver of systemic autoimmunity. However, mechanisms that regulate TI humoral immunity are poorly defined. In this study, we report that B cell–intrinsic induction of the tryptophan-catabolizing enzyme IDO1 is a key mechanism limiting TI Ab responses. When Ido1−/− mice were immunized with TI Ags, there was a significant increase in Ab titers and formation of extrafollicular Ab-secreting cells compared with controls. This effect was specific to TI Ags, as Ido1 disruption did not affect Ig production after immunization with protein Ags. The effect of IDO1 abrogation was confined to the B cell compartment, as adoptive transfer of Ido1−/− B cells to B cell–deficient mice was sufficient to replicate increased TI responses observed in Ido1−/− mice. Moreover, in vitro activation with TLR ligands or BCR crosslinking rapidly induced Ido1 expression and activity in purified B cells, and Ido1−/− B cells displayed enhanced proliferation and cell survival associated with increased Ig and cytokine production compared with wild-type B cells. Thus, our results demonstrate a novel, B cell–intrinsic, role for IDO1 as a regulator of humoral immunity that has implications for both vaccine design and prevention of autoimmunity.
Introduction
In immunity, B cells occupy a unique niche, providing the cellular pool from which Ab-producing plasma cells form as well as serving as APCs, thus playing a central role in humoral and cellular immunity. Marginal zone (MZ) and B-1 B cell subsets are specialized “innate-like” B cell populations that show limited BCR diversity and are the first to respond to blood-borne Ags (1, 2). The initial MZ B cell response to infection occurs without T cell help, resulting in rapid production of low-affinity, cross-reactive IgM and IgG3 for early pathogen neutralization. This is followed by an adoptive response by follicular (FO) B cells generating Ag-specific, high-affinity IgG Abs through a germinal center reaction with T cell help (3, 4). Both MZ cells and B-1 B cells express poorly diversified germline-encoded BCRs polyreactive against microbial and self-antigen (1). MZ B cells also express greater levels of TLRs compared with FO B cells, ensuring their high responsiveness to ligands such as LPS- and CpG-eliciting humoral immunity (5–7). Upon recognizing such innate components, along with BCR engagement, MZ B cells become activated and robustly proliferate to form foci of plasma cells in the extrafollicular regions of the spleen (8). The MZ B cell response is further enhanced by cytokines and other poorly defined mechanisms mediated by macrophages, dendritic cells, and neutrophils (9–11).
Most low-affinity extrafollicular plasma cells derived from MZ B cells are short-lived, normally undergoing apoptosis within a few days (12, 13), although they can also generate memory cells providing a long-lasting primary Ab response against polysaccharide Ags (14). A greater extrafollicular response is favorable for immunity against bacterial and viral infection; however, a dysregulated MZ B cell response is implicated as a cause of autoimmunity (15, 16). For example, abnormal MZ B cell migratory properties and prolonged survival of short-lived plasma cells in mice have been linked to the development of lupus-like autoimmune disease in association with autoreactive Ab production (12, 17, 18). Moreover, immunoregulatory functions of IL-10–producing regulatory B cells in the MZ B cell population are a key cellular regulator of autoimmune disease pathogenicity (19). Thus, the innate B cell response is a key driver of protective immunity to infection, yet the possibility for autoimmunity necessitates strict regulation of B cell responses to TI Ags.
IDO is an intracellular, tryptophan-metabolizing enzyme that drives immune regulation in a variety of settings, including cancer and autoimmunity (20, 21). IDO is present as two isoforms (IDO1 and IDO2) that evolved as a result of gene duplication (22); however, the two IDO isoforms are induced by different stimuli, with the Ido1 gene exhibiting responsiveness primarily to immunologic signals (23). In most cell types, IDO1 is not expressed under normal physiologic conditions, but inflammatory cues, including type I and II IFN stimulation, rapidly induces IDO1 activity in dendritic cells, macrophages, and some stromal cell populations (21, 24, 25). IDO1 inhibits naive T cell proliferation and survival and promotes differentiation and activation of regulatory T cells driving immune suppression and, ultimately, stable tolerance (26, 27). We have previously shown that apoptotic cell-driven IDO1 activity in the spleen is required to halt autoantibody responses and establish tolerance to self-antigens (21). Given the close mechanistic relationship between MZ B cells and humoral autoimmunity, we hypothesized that IDO1 may play a regulatory role in extrafollicular B cell responses to TI Ags (15, 16).
In this study, we identified IDO1 induction in B cells as a negative regulatory mechanism of the TI humoral immune response. We further demonstrate that B cell–intrinsic IDO1 activity is mechanistically required to regulate the magnitude of extrafollicular responses by limiting proliferation and survival associated with Ab production (IgM, IgG1, and IgG3) and cytokine secretion (IL-6 and IL-10).
Materials and Methods
Mice
Female 8- to 10-wk-old C57BL/6J (B6), B6.Ido1−/−, B6.μMT−/−, B6.Ifnar1−/−, and B6.Ifnγr1−/− mice were obtained from The Jackson Laboratory and maintained under specific pathogen-free conditions in the Georgia Regents University animal facilities in accordance with Institutional Animal Care and Use Committee guidelines.
Immunization
For T-independent immunization, mice were injected i.v. with 20 μg 4-hydroxy-3-nitrophenylacetic acid (NP) hapten conjugated to NP-Ficoll (aminoethyl carboxymethyl–Ficoll) or 10 μg NP-LPS (Biosearch Technologies) in PBS. For T-dependent immunization, mice were injected i.p. with 50 μg NP-OVA (Biosearch Technologies) emulsified with Imject Alum (Thermo Scientific). For immunizations with CpG oligonucleotides, mice were injected i.p. with 50 μg class B CpG oligonucleotides (ODN 1826; InvivoGen) mixed with 50 μg NP-OVA in 200 μl PBS.
NP-specific ELISAs
ELISAs were performed to measure NP-specific serum Ab responses. Ninety-six–well flat-bottom plates were coated, overnight, with 5 μg/ml NP-BSA (Biosearch Technologies) diluted in PBS at 4°C. Plates were washed three times (PBS, 0.1% Tween 20; Acros Organics) followed by blocking (PBS, 1% BSA) for a minimum of 2 h at room temperature. Samples were prepared with serial dilutions of sera using blocking buffer and incubated for 2 h at room temperature. Plates were washed five times and incubated with HRP-conjugated anti-mouse isotype-specific detection Ab (Bethyl Laboratories, Montgomery, TX) for 1 h at room temperature. After five washes, plates were developed by addition of TMB (Kirkegaard and Perry Laboratories, Gaithersburg MD) and the absorbance values were read at 450/570 nm. Mouse reference serum with defined concentrations of IgG and IgM (Bethyl Laboratories) was used to quantify absolute concentrations of NP-specific Abs in the serum.
Affinity measurements
To examine the affinity maturation, NP-specific ELISAs were performed using plates coated with either NP1–9-BSA or NP>20-BSA (Biosearch Technologies) diluted in PBS at 5 μg/ml. The ratio of high affinity (NP1–9-BSA) to low affinity (NP>20-BSA) for a given sample was determined as the OD titer of serum Abs bound to NP1–9-BSA, divided by the OD titer of Abs bound to NP>20-BSA.
NP-Specific ELISPOTs
NP-specific Ab-secreting cells (ASCs) were measured in 96-well flat-bottom EIA/RIA high-binding plates coated overnight at 4°C with 5 μg/ml NP-BSA diluted in PBS. Plates were washed three times (PBS, 0.1% Tween 20) followed by blocking (PBS, 1% BSA) for minimum of 2 h at room temperature. Cells were isolated from the spleen 7 d after immunization and seeded in duplicate at a starting density of 107 total viable cells per 100 μl in the first well, and 2-fold serial dilutions were performed down the plate. Cells were incubated at 37°C in 5% CO2 for 6 h in complete RPMI 1640 (Sigma-Aldrich). Plates were then washed twice with double-distilled H2O for 5 min at room temperature followed by three washes (PBS, 0.1% Tween 20). Secreted Abs were detected by incubating plates with an isotype-specific AP-conjugated goat anti-mouse Ig (SouthernBiotech) diluted in blocking buffer for 1 h at room temperature. After five washes the plates were developed overnight at 4°C with 5-bromo-4-chloro-3-indolyl phosphate/NBT substrate (Life Technologies). Plates were washed twice with double-distilled H2O and allowed to dry at room temperature. The plates were scanned using an ELISPOT machine (Cellular Technology, Immunospot) and the developed spots were counted visually from the scanned images.
Immunofluorescence
Spleens were harvested from the mice after NP-Ficoll immunization at indicated times, snap frozen in liquid nitrogen, and frozen in Tissue-Tek OCT compound (Sakura Finetek) at −80°C. For immunofluorescence of IDO1, 5-μm frozen splenic sections were immediately fixed in −20°C methanol. The sections were blocked with PBS containing 1% nonfat milk (Sigma-Aldrich) and stained with 4 μg/ml polyclonal rabbit anti-mouse IDO1 (a gift from Dr. David Munn) in PBS containing 1% nonfat milk. After extensive washing with TBS plus 0.05% Tween 20 (Acros Organics), the sections were stained with a 1:400 dilution of Cy3-labeled goat anti-rabbit IgG (Jackson ImmunoResearch Laboratories), allophycocyanin-labeled anti-B220 (eBioscience), or FITC-labeled anti-CD138 (BD Pharmingen). To examine Ab response to NP-Ficoll immunization, sections were incubated with anti-IgM (BioLegend) and anti-IgG3 (SouthernBiotech) in blocking buffer for 60 min at room temperature in the dark. For stimulated and unstimulated purified B cells, cytospins were prepared and fixed with −20°C methanol. Cells were made permeable using Perm Buffer III (BD Biosciences) and stained for IDO1 as explained above and anti-B220 (eBioscience). Sections and cytospins were mounted with ProLong Gold antifade with DAPI (Invitrogen). Fluorescent images were captured using a Zeiss LSM 510 Meta confocal microscope equipped with 405-, 488-, 561-, and 633-nm lasers.
Flow cytometry
To determine percentages of immune cell subsets, single-cell suspensions from spleen, bone marrow and peritoneal lavage were incubated with anti-B220, anti-CD5, anti-Thy1.2, anti-CD4, anti-F4/80, anti-CD11c, anti-CD93 (AA4.1) (all from eBioscience), anti-CD21, anti-CD24, anti-CD8, anti-CD11b (all from BD Biosciences), anti-CD23, anti-IgM, anti-CD40, and anti-Ly6G (all from BioLegend). For flow cytometric study, >105 events were collected on an LSR II flow cytometer and all results were analyzed with FlowJo software (Tree Star).
B cell isolation and in vitro stimulation
For B cell enrichment, splenocytes were incubated with an Ab mixture containing biotin-labeled anti-CD90, anti-CD11c, and anti-CD11b (eBioscience). Anti-biotin microbeads (Miltenyi Biotec) were used to deplete CD90+ T cells, CD11c+ dendritic cells, and CD11b+ macrophages using MACS separation. Purified B cells were left unstimulated or were stimulated with anti-IgM (Jackson ImmunoResearch Laboratories) at 10 μg/ml plus IL-4 (PeproTech) at 20 ng/ml, anti-CD40 (BD Biosciences) at 10 μg/ml plus IL-4 at 20 ng/ml, LPS (Sigma-Aldrich; 055:B5) at 5 μg/ml, and CpG (InvivoGen; ODN 1826) at 5 μg/ml. The kynurenine/tryptophan ratio was determined with HPLC analysis of culture medium as previously described (28). Cytokine protein concentrations in the culture medium were measured using Ready-SET-Go! ELISA kits (eBioscience). Cell proliferation was determined by FACS analysis after labeling cells with 0.5 μM CFSE (Invitrogen). Cell survival was quantified on a flow cytometer by examining annexin V staining. To test the surface expression profiles of activation markers, B cells were stained with anti-CD86 (BD Pharmingen), anti–MHC class II (eBioscience), anti-IgM (BioLegend), anti-CD44 (BD Pharmingen), anti-IgM (BioLegend), and anti-CD40 (BioLegend). Events were collected on an LSR II flow cytometer and all results were analyzed with FlowJo software (Tree Star).
Semiquantitative PCR
RNA from sorted cells was purified using RNeasy RNA purification kits (Qiagen), and 250 ng RNA was reverse transcribed using a random hexamer cDNA reverse transcription kit (Clontech). For the PCR, 1 μl cDNA was amplified with primers as previously described (21, 29). PCR was completed using iQ SYBR Green supermix (Bio-Rad) and on an iQ5 real-time PCR detection system (Bio-Rad), and results were analyzed with the accompanying software according to the manufacturer’s instructions.
Western blotting
Western blot analysis was done as previously described (28). Specific Abs for following molecules were used: cleaved caspase-3, cleaved caspase-9, cleaved PARP, β-actin (all from Cell Signaling Technology), cleaved caspase-8 (R&D Systems), Bak (Upstate Cell Signaling Solutions), and Bcl-2 (BD Transduction Laboratory).
Statistical analysis
Means, SDs, and unpaired Student t test results were used to analyze the data. When comparing two groups, a p value ≤0.05 was considered to be significant.
Results
IDO1 inhibits the TI Ab response
We have previously shown that IDO1 is induced in the MZ of the spleen in response to apoptotic cell exposure and deficiency in IDO1 activity correlated with a rapid increase in serum autoantibodies to DNA after challenge with a bolus of apoptotic cells i.v. (21). Because MZ B cells are a significant source of humoral anti-DNA reactivity (15), we hypothesized IDO1 may regulate innate (i.e., MZ) B cell activation in the spleen. Because MZ B cells are important for the response to TI Ags, we immunized Ido1 knockout (Ido1−/−) and B6 mice with the TI type II Ag NP-Ficoll and the TI type I Ag NP-LPS, or the T cell-dependent (TD) protein Ag NP-OVA. B6 mice responded rapidly to TI or TD Ags with increased Ag-specific IgM, IgG1, and IgG3 anti-NP Abs after immunization (Fig. 1A); moreover, Ido1−/− mice exhibited a significant increase in anti-NP IgM, IgG1, and IgG3 responses to TI Ags at day 7 relative to controls (Fig. 1A). In contrast, the lack of IDO1 did not increase Ag-specific Ab responses to NP-OVA, suggesting that the effect was specific for TI responses (Fig. 1A). TI Ags such as NP-Ficoll induce MZ B cell responses that are relatively short-lived and of low affinity (2, 30). Thus, we tested whether IDO1 deficiency impacted the longevity or affinity of the TI Ab response. Compared to the humoral NP response in control B6 mice, Ido1−/− mice showed significantly augmented IgM, IgG1, and IgG3 responses to NP-Ficoll at all time points examined (Fig. 1B). However, the kinetics of the Ab response were similar between control and Ido1−/− mice.
IDO1 limits TI Ab responses. (A) B6 and Ido1−/− mice were injected with NP-Ficoll, NP-LPS, and NP-OVA as described in 2Materials and Methods, and serum NP-specific IgM, IgG1, and IgG3 Ab responses were measured by ELISA. (B) B6 and Ido1−/− mice were injected with NP-Ficoll as in (A). Serum NP-specific IgM, IgG1, and IgG3 Ab responses were measured at indicated times. Datum points are the mean value for five mice per group, and error bars represent the SD. Experiment was repeated four times with similar results. *p < 0.05, **p ≤ 0.01 (as determined by Student t test).
IDO1 limits TI Ab responses. (A) B6 and Ido1−/− mice were injected with NP-Ficoll, NP-LPS, and NP-OVA as described in 2Materials and Methods, and serum NP-specific IgM, IgG1, and IgG3 Ab responses were measured by ELISA. (B) B6 and Ido1−/− mice were injected with NP-Ficoll as in (A). Serum NP-specific IgM, IgG1, and IgG3 Ab responses were measured at indicated times. Datum points are the mean value for five mice per group, and error bars represent the SD. Experiment was repeated four times with similar results. *p < 0.05, **p ≤ 0.01 (as determined by Student t test).
When we compared splenic B cell populations in B6 and Ido1−/− mice we found no difference in MZ or follicular B cell numbers (Fig. 2A). Likewise, there was no difference in splenic B1 to B2 B cell ratios, and the numbers of transitional and mature B cells were similar (Fig. 2A). Similarly, there were no observable differences in splenic T cells (Fig. 2C) or myeloid cells (Fig. 2C), and B cell populations in the peritoneal cavity (Fig. 2D) and bone marrow (Fig. 2E) were comparable. Furthermore, serum IgM, IgG1, and IgG3 concentrations were unchanged in naive Ido1−/− mice compared with control mice (Fig. 2F). Thus, the augmented Ab response to TI Ags in Ido1−/− mice was likely not due to difference in baseline B cell numbers, but rather to the magnitude of the B cell response.
IDO1 deficiency does not affect B cell development. Single-cell suspensions of spleen, peritoneal lavage, and bone marrow from B6 and Ido1−/− mice (four mice per group) were generated and analyzed by FACS analysis. (A) Representative dot plots of splenic B cell subsets: MZ B (CD21highCD23low), FO B cells (CD21+CD23high), B1 (CD5lowIgM+), B2 (CD5−IgM+), transitional T1 (IgMhighCD23low), T2 (IgMhighCD23high), and T3 (IgMlowCD23high) B cells in the B220+ B cell populations. (B) Splenic populations of T cells (Thy1+), CD4+, and CD8+ T cells. (C) Splenic myeloid cell populations identified by surface markers indicated. (D) Representative dot plots of peritoneal cavity cell subsets: B1 (CD5lowIgM+) and B2 (CD5−IgM+) B cells. (E) Representative dot plots of B cell subsets in the bone marrow: pro–B and pre–B (IgM−B220low), immature B (IgMlowB220low), mature B (IgMlowB220high), and transitional B cells (IgMhighB220low–high). For (A) to (E), bars represent the mean value for quadruplicate samples (±SD). For analysis of transitional B cell subsets in (A), cells were gated on B220+CD93+ splenic B cells. (F) Baseline serum Ab levels of IgM, IgG, IgG1, and IgG3 in B6 and Ido1−/− mice. Datum points are the mean value for four mice per group, and error bars represent the SD. Experiment was repeated twice with similar results.
IDO1 deficiency does not affect B cell development. Single-cell suspensions of spleen, peritoneal lavage, and bone marrow from B6 and Ido1−/− mice (four mice per group) were generated and analyzed by FACS analysis. (A) Representative dot plots of splenic B cell subsets: MZ B (CD21highCD23low), FO B cells (CD21+CD23high), B1 (CD5lowIgM+), B2 (CD5−IgM+), transitional T1 (IgMhighCD23low), T2 (IgMhighCD23high), and T3 (IgMlowCD23high) B cells in the B220+ B cell populations. (B) Splenic populations of T cells (Thy1+), CD4+, and CD8+ T cells. (C) Splenic myeloid cell populations identified by surface markers indicated. (D) Representative dot plots of peritoneal cavity cell subsets: B1 (CD5lowIgM+) and B2 (CD5−IgM+) B cells. (E) Representative dot plots of B cell subsets in the bone marrow: pro–B and pre–B (IgM−B220low), immature B (IgMlowB220low), mature B (IgMlowB220high), and transitional B cells (IgMhighB220low–high). For (A) to (E), bars represent the mean value for quadruplicate samples (±SD). For analysis of transitional B cell subsets in (A), cells were gated on B220+CD93+ splenic B cells. (F) Baseline serum Ab levels of IgM, IgG, IgG1, and IgG3 in B6 and Ido1−/− mice. Datum points are the mean value for four mice per group, and error bars represent the SD. Experiment was repeated twice with similar results.
To test whether the increased magnitude of Abs in Ido1−/− mice was associated with changes in affinity maturation, NP-Ficoll and NP-LPS immunized mouse serum samples were tested for the ratio of NPhigh (NP1–9-BSA)/NPlow (NP>20-BSA) IgM and IgG as a measure of affinity (31). The anti-NP response did not show evidence of affinity maturation during a 28-d period as expected in the absence of T cell help (Supplemental Fig. 1; see Ref. 31). Similarly, IDO1 deficiency did not affect the NPhigh/NPlow affinity ratio for either IgM or IgG isotypes, indicating that the increased humoral NP reactivity was not due to either prolonged responses or increased affinity maturation (Supplemental Fig. 1).
B cell–intrinsic IDO1 deficiency compromises TI Ab response
Immunization with NP-Ficoll is characterized by robust B cell proliferation and formation of ASCs confined to extrafollicular spaces in the red pulp of the spleen (8, 32). We tested whether the enhanced Ab response in Ido1−/− mice is the result of augmented B cell proliferation and formation of ASCs. For this, we measured the frequency of Ag-specific ASCs 7 d after NP-Ficoll immunization by ELISPOT. Paralleling serum Ab results, Ido1−/− mice showed significant increases in the frequency of splenic anti-NP IgM- and IgG3-producing ASCs (Fig. 3A). We also found large clusters of IgM+ and IgG3+ ASCs within the splenic extrafollicular spaces of Ido1−/− mice compared with B6 controls, which exhibited a much smaller cluster formation (Fig. 3B). There were no NP-specific IgM and IgG3 ASCs in the lymph nodes after immunization, suggesting a specific regulatory role for splenic IDO1 in TI Ab responses (data not shown).
IDO1 is induced by TI Ag immunization and inhibits extrafollicular Ab response in a B cell intrinsic manner. B6 and Ido1−/− mice were immunized with NP-Ficoll i/v as in Fig. 1. (A) Frequency of splenic NP-specific IgM and IgG3 ASCs was measured by ELISPOT 7 d after immunization. Bars are the mean value for five mice per group (±SD). (B) Spleen sections from B6 and Ido1−/− mice 7 d after immunization with NP-Ficoll were examined by immunofluorescence staining for IgM+ and IgG3+ foci. B220 staining was used to depict the orientation of the follicles. Original magnification ×200. (C) B6 mice were immunized with NP-Ficoll or NP-OVA and spleens were collected on day 3 and day 5, respectively. Spleen sections were examined for expression of IDO1 (red) and CD138 (green), and spleen lysates for mRNA levels of Ido1 by semiquantitative PCR. Original magnification ×400. (D) Purified B cells (2.5 × 107) from spleens of B6 and Ido1−/− mice were adoptively transferred i.v. to μMT−/− mice and 1 d later recipients were immunized with NP-Ficoll i.v. Serum NP-specific IgM, IgG1, and IgG3 Ab responses were measured at indicated times. Experiment was repeated at least three times with similar results. *p < 0.05, **p ≤ 0.01 (as determined by Student t test).
IDO1 is induced by TI Ag immunization and inhibits extrafollicular Ab response in a B cell intrinsic manner. B6 and Ido1−/− mice were immunized with NP-Ficoll i/v as in Fig. 1. (A) Frequency of splenic NP-specific IgM and IgG3 ASCs was measured by ELISPOT 7 d after immunization. Bars are the mean value for five mice per group (±SD). (B) Spleen sections from B6 and Ido1−/− mice 7 d after immunization with NP-Ficoll were examined by immunofluorescence staining for IgM+ and IgG3+ foci. B220 staining was used to depict the orientation of the follicles. Original magnification ×200. (C) B6 mice were immunized with NP-Ficoll or NP-OVA and spleens were collected on day 3 and day 5, respectively. Spleen sections were examined for expression of IDO1 (red) and CD138 (green), and spleen lysates for mRNA levels of Ido1 by semiquantitative PCR. Original magnification ×400. (D) Purified B cells (2.5 × 107) from spleens of B6 and Ido1−/− mice were adoptively transferred i.v. to μMT−/− mice and 1 d later recipients were immunized with NP-Ficoll i.v. Serum NP-specific IgM, IgG1, and IgG3 Ab responses were measured at indicated times. Experiment was repeated at least three times with similar results. *p < 0.05, **p ≤ 0.01 (as determined by Student t test).
Ido1 mRNA was induced 11-fold in splenic whole-tissue preparations 3 d after immunization with NP-Ficoll; however, there was no induction in mice immunized with NP-OVA (Fig. 3C). IDO1 protein was confined to the splenic extrafollicular spaces of the red pulp after immunization (Fig. 3C), which is also the site of ASC accumulation (Fig. 3B). Moreover, when splenic sections were costained with the plasma blast marker CD138 we found IDO1 protein colocalized with this cellular population (Fig. 3C). This led us to hypothesize that B cells may be a source of IDO1 after immunization, regulating humoral responses in a cell-intrinsic fashion. To test this, we adoptively transferred equal numbers of purified B cells from wild-type or Ido1−/− mice into B cell–deficient (μMT−/−) mice and assessed their response against NP-Ficoll immunization. Consistent with our serum analysis data (Fig. 1), μMT mice receiving Ido1−/− B cells showed significantly increased concentrations of serum anti-NP IgM, IgG1, and IgG3 7 d after immunization compared with wild-type B cell–recipient μMT−/− mice (Fig. 3D). These results suggest that IDO1 activity within the B cell compartment is a negative regulator specific for TI antigenic B cell responses.
B cell IDO1 inhibits proliferation and promotes apoptosis in response to TI antigenic stimuli
To investigate the mechanism of IDO1 induction and its regulatory function during TI responses, we activated purified splenic B cells with either the TI type I Ags LPS and CpG, anti-IgM BCR crosslinking (to mimic TI-2 Ag responses), or CD40 ligation (to mimic TD Ag responses) in vitro. B cells were enriched from the spleen through negative selection with >90% cell purity as explained in 2Materials and Methods (for purity, see Supplemental Fig. 2). Both LPS and CpG directly induced transcription of enzymatically active Ido1 as evidenced by the >90- and >30-fold induction in mRNA levels, respectively (Fig. 4A), and the >10- and >4-fold increase in the kynurenine/tryptophan ratio (Fig. 4B). Similarly, BCR crosslinking induced both Ido1 expression and kynurenine production, albeit to a lower level than observed with TLR ligands (Fig. 4A, 4B). In contrast, CD40 crosslinking had no effect on IDO1 activity (Fig. 4A, 4B). This was consistent with immunostaining of B cells in cytospin preparations that were IDO1+ after activation by TLR ligands or BCR crosslinking, but IDO1− in groups activated by CD40 ligation (Fig. 4C). Because IDO1 is expressed in response to IFNs at sites of inflammation (24, 25), we hypothesized that type I or type II IFNs may drive induction of IDO1 in B lymphocytes following TLR ligation. In agreement with this, both LPS and CpG stimulation led to the upregulation of Ifn-β1 and Ifn-γ at 72 h in activated B cells (Supplemental Fig. 3A). We next stimulated IFN-αR1 or IFN-γR1 knockout B cells (which are nonresponsive to type I and type II IFNs, respectively) with LPS and CpG. Interestingly, Ido1 mRNA levels in IFN-αR1 and IFN-γR1 knockout B cells were comparable to controls (Supplemental Fig. 3B). Thus, these findings suggest IDO1 expression is directly induced in B cells through activation signals independent of IFN signaling.
B cells express IDO1 in response to TI antigenic stimuli. Purified B cells from B6 mice were treated with anti-IgM, anti-CD40, LPS, and CpG for 72 h as indicated in 2Materials and Methods. (A) mRNA levels of Ido1 were measured with semiquantitative PCR and normalized to β-actin. (B) Culture supernatants were examined by HPLC to measure concentrations of tryptophan and kynurenine expressed in this study at the kynurenine/tryptophan ratio. (C) B cell cytospins were stained for IDO1 (red) and B220 (green) and visualized using immunofluorescence. Original magnification ×600. All bars represent the mean value for five samples per group (±SD). Experiment was repeated three times with similar results. **p ≤ 0.01 (as determined by Student t test).
B cells express IDO1 in response to TI antigenic stimuli. Purified B cells from B6 mice were treated with anti-IgM, anti-CD40, LPS, and CpG for 72 h as indicated in 2Materials and Methods. (A) mRNA levels of Ido1 were measured with semiquantitative PCR and normalized to β-actin. (B) Culture supernatants were examined by HPLC to measure concentrations of tryptophan and kynurenine expressed in this study at the kynurenine/tryptophan ratio. (C) B cell cytospins were stained for IDO1 (red) and B220 (green) and visualized using immunofluorescence. Original magnification ×600. All bars represent the mean value for five samples per group (±SD). Experiment was repeated three times with similar results. **p ≤ 0.01 (as determined by Student t test).
To further test the impact of B cell–intrinsic IDO1 activity on function, we examined Ab and cytokine production by B cells after activation. Consistent with our in vivo results, Ido1−/− B cells demonstrated a significant increase in Ab and cytokine production after activation with TLR ligands or BCR crosslinking but exhibited no difference after CD40-driven activation when compared with B6 B cells (Fig. 5A). These results are consistent with a mechanistic role for IDO1 in regulating responses that activate B cells independent of T cells.
B cell–intrinsic IDO1 activity negatively regulates Ab responses and drives apoptosis. Purified splenic B cells were treated with anti-IgM, anti-CD40, LPS, and CpG for 72 h as indicated in 2Materials and Methods. (A) Culture supernatants were measured by ELISA for Ab (IgM, IgG1, IgG3) and cytokine production (IL-10, IL-6). (B) mRNA levels of cyclin D3 and blimp-1 were measured by semiquantitative PCR and normalized to β-actin. (C) Representative Western blot analysis of pro- and antiapoptotic proteins indicated. Freshly purified live B cells were used as control. Blots are representative for five samples per group. (D) FACS analysis of B cell survival. B cells were stained with annexin V and propidium iodine to determine the extent of B cell apoptosis. For all graphs, bars represent the mean value for five samples per group (±SD). Experiment was repeated three times with similar results. *p < 0.05 (as determined by Student t test).
B cell–intrinsic IDO1 activity negatively regulates Ab responses and drives apoptosis. Purified splenic B cells were treated with anti-IgM, anti-CD40, LPS, and CpG for 72 h as indicated in 2Materials and Methods. (A) Culture supernatants were measured by ELISA for Ab (IgM, IgG1, IgG3) and cytokine production (IL-10, IL-6). (B) mRNA levels of cyclin D3 and blimp-1 were measured by semiquantitative PCR and normalized to β-actin. (C) Representative Western blot analysis of pro- and antiapoptotic proteins indicated. Freshly purified live B cells were used as control. Blots are representative for five samples per group. (D) FACS analysis of B cell survival. B cells were stained with annexin V and propidium iodine to determine the extent of B cell apoptosis. For all graphs, bars represent the mean value for five samples per group (±SD). Experiment was repeated three times with similar results. *p < 0.05 (as determined by Student t test).
IDO1 negatively regulates B cell proliferation and survival
TLR engagement induces B cell activation and proliferation (5, 6). Because Ido1−/− B cells produce more Abs upon TLR ligation, we hypothesized that cell-intrinsic IDO1 activity negatively regulates B cell survival and proliferation. Therefore, we examined the mRNA levels of cyclin D3 as a proliferation marker and blimp-1 as a plasmablast differentiation marker in B cells after stimulation. In agreement with our hypothesis, Ido1−/− B cells exhibited significantly increased cyclin D3 and blimp-1 mRNA expression after either TLR ligand or BCR crosslinking-driven activation compared with wild-type B cells (Fig. 5B). Furthermore, utilizing a CFSE-based proliferation assay, we observed that Ido1−/− B cells proliferated more robustly compared with wild-type B cells after TI stimulation, as evidenced by the presence of higher frequencies of cells within the fractions with more cell divisions (Fig. 6). In stark contrast, when B cells were activated by CD40 ligation, IDO1 deficiency had no impact on proliferation, supporting the in vivo data and suggesting that IDO1 plays a specific role in regulation of BCR- or TLR-driven B cell proliferation (Fig. 6).
IDO1 inhibits B cell proliferation. Purified B cells were treated with anti-IgM, anti-CD40, LPS, or CpG for 72 h as indicated in 2Materials and Methods. Representative histograms of B cell proliferation as determined by FACS analysis of CFSE dye dilution in vitro. Bar graphs to the left represent quantitative assessment of percentage cell proliferation in each division for five samples per group (±SD). Experiment was repeated three times with similar results. *p < 0.05 (as determined by Student t test).
IDO1 inhibits B cell proliferation. Purified B cells were treated with anti-IgM, anti-CD40, LPS, or CpG for 72 h as indicated in 2Materials and Methods. Representative histograms of B cell proliferation as determined by FACS analysis of CFSE dye dilution in vitro. Bar graphs to the left represent quantitative assessment of percentage cell proliferation in each division for five samples per group (±SD). Experiment was repeated three times with similar results. *p < 0.05 (as determined by Student t test).
B cell survival after activation is determined by a balance between the expression of proapoptotic and antiapoptotic proteins (33). Because TLR ligation prevents spontaneous apoptosis (7), we next explored whether IDO1 affects the viability of activated B cells. Compared to wild-type B cells, Ido1−/− B cells exhibited a reduction in caspase activation, as there was a demonstrable decreases in the presence of cleaved caspase-3, caspase-8, caspase-9, and PARP as determined by Western blot (Fig. 5C). Similarly, there was a decrease in the presence of the Bcl-2 family member and BAK, an initiator of the apoptotic program (Fig. 5C) (34). In contrast, LPS-stimulated Ido1−/− B cells exhibited an increase in the presence of the antiapoptotic protein Bcl-2 compared with controls; however, there was no detectible difference in Bcl-2 protein bands in Ido1−/− B cells stimulated with CpG compared with wild-type B groups. These results were consistent with flow cytometry analysis of annexin V staining, which revealed a significant reduction in Ido1−/− B cell apoptosis after TI stimulation compared with controls (Fig. 5D). Furthermore, there were no differences between Ido1−/− B cells and wild-type B cells in surface expression of costimulatory and accessory molecules (CD86, MHC class II, membrane-bound IgM, and CD44) after TI activation, suggestive of equivalent activation after stimulation (Supplemental Fig. 4). Taken together, these findings clearly demonstrate that intrinsic IDO1 mechanistically regulates proliferation and survival of activated B cells in response to TI Ags.
Innate signals regulate B cell responses to TD Ags
Humoral immunity to microbial infection or sterile inflammation associated with necrotic and apoptotic cell death will likely be the result of a combination of TI-1, TI-2, and TD antigenic responses. Thus, we hypothesized that although IDO1 appeared to be induced primarily by TI Ags and innate stimuli, it may play a role in responses to TD Ags in situations with strong BCR crosslinking or in the presence of TLR ligands. To test this we stimulated purified B cells with a combination of LPS, BCR crosslinking, and CD40 ligation. This combination drove a 275-fold increase in Ido1 mRNA (Fig. 7A) (>2-fold higher induction than that seen by LPS alone, as shown in Fig. 4A). B cell proliferation in wild-type B cells after LPS/anti-BCR/anti-CD40 was vigorous and ∼50% of the B cells exhibited CFSE dye dilution (Fig. 7B). In contrast, Ido1−/− B cells showed a significant increase in proliferative capacity, as >90% of the B cells had undergone at least seven divisions after LPS/anti-BCR/anti-CD40 stimulation, suggestive of a significantly exaggerated B cell response (Fig. 7B). In agreement with this, Ido1−/− B cells exhibited highly significant increases in Ig (Fig. 7C) and cytokine (Fig. 7D) production after LPS/anti-BCR/anti-CD40 stimulation compared with Ido1 wild-type B cells. This result suggests that BCR and TLR coligation in the presence of T cell help strongly potentiates IDO1-dependent regulatory responses.
CD40-, BCR-, and TLR-driven stimulation act in a synergistic fashion to induce IDO1-dependent regulatory activity in B cells. Purified splenic B cells from mice with the indicated genotype were treated with anti-IgM, anti-CD40, and LPS in combination for 72 h as indicated in 2Materials and Methods. (A) mRNA levels of Ido1 were measured by semiquantitative PCR and normalized to β-actin. (B) Representative histograms of B cell proliferation as determined by FACS analysis of CFSE dye dilution in vitro. Bar graphs to the right represent quantitative assessment of percentage cell proliferation in each division for five samples per group. (C and D) Culture supernatants were measured by ELISA for Ab (IgM, IgG1, IgG3) and cytokine production (IL-10, IL-6). For all graphs, bars represent the mean value (±SD). Experiment was repeated three times with similar results. **p < 0.01 (as determined by Student t test).
CD40-, BCR-, and TLR-driven stimulation act in a synergistic fashion to induce IDO1-dependent regulatory activity in B cells. Purified splenic B cells from mice with the indicated genotype were treated with anti-IgM, anti-CD40, and LPS in combination for 72 h as indicated in 2Materials and Methods. (A) mRNA levels of Ido1 were measured by semiquantitative PCR and normalized to β-actin. (B) Representative histograms of B cell proliferation as determined by FACS analysis of CFSE dye dilution in vitro. Bar graphs to the right represent quantitative assessment of percentage cell proliferation in each division for five samples per group. (C and D) Culture supernatants were measured by ELISA for Ab (IgM, IgG1, IgG3) and cytokine production (IL-10, IL-6). For all graphs, bars represent the mean value (±SD). Experiment was repeated three times with similar results. **p < 0.01 (as determined by Student t test).
To test this in vivo, we immunized mice with the TD Ag NP-OVA using CpG containing oligonucleotides as an adjuvant. As also shown in Fig. 1, Ido1−/− mice showed a similar response to i.p. immunization with either NP-OVA alone or in combination with aluminum hydroxide (alum) compared with controls (Fig. 8). In contrast, when CpG containing oligonucleotides was used as an adjuvant, Ido1−/− mice showed a significantly higher humoral response to antigenic challenge compared with B6 mice (Fig. 8). Alum and CpG oligonucleotides drive immunity via distinct innate pathways, with alum activating the NLRP3 inflammasome (35), whereas CpG oligonucleotides induce TLR9/MyD88-driven responses (36). Thus, this result, coupled with the in vitro findings, suggests that innate signals driving strong BCR crosslinking or TLR activation will regulate B cell responses to both TI and TD Ags by IDO1 induction. In contrast, inflammasome activation (at least by alum) appears unable to induce IDO1 activity and, by extension, regulate B cell responses. However, as a whole the data indicate that IDO1-mediated regulation of B cell responses is mechanistically independent of T cells.
IDO1 regulates humoral responses to TD Ags in the presence of TI agonists. Mice were immunized i.p. with 50 μg OVA alone (in PBS), with alum, or with 50 μg CpG motif containing oligonucleotides, and serum was collected at the times indicated. All bars represent the mean value for five mice per group (±SD). Experiment was repeated three times with similar results. *p < 0.05, **p ≤ 0.01 (as determined by Student t test).
IDO1 regulates humoral responses to TD Ags in the presence of TI agonists. Mice were immunized i.p. with 50 μg OVA alone (in PBS), with alum, or with 50 μg CpG motif containing oligonucleotides, and serum was collected at the times indicated. All bars represent the mean value for five mice per group (±SD). Experiment was repeated three times with similar results. *p < 0.05, **p ≤ 0.01 (as determined by Student t test).
Discussion
During an immune response, B cell activities are tightly controlled to avert undesirable or excessive reactivity, such as autoantibody production. T cell–dependent FO B cells differentiate into long-lived plasma cells and memory B cells, which is instructed by cognate T cell help in a multiple stepwise manner through MHC class II–restricted Ag presentation (3, 4). In contrast, upon engagement of TLR, or crosslinking of BCRs, innate B cells function independently of T cell activity (5, 6). Such a spontaneous differentiation program is fundamental for innate B cells to provide immediate humoral protection against infection. However, very little is known about the mechanisms by which innate B cells regulate their activation and Ig production. In this study, we found that IDO1, an enzyme that drives the oxidative tryptophan metabolism pathway, is induced in activated B cells specifically in response to TI Ags and negatively regulates B cell responses.
Although IDO1 induction in B cells has never been closely examined before, a previous report found the presence of functionally inactive IDO1 in human B cells in response to CD40L and IFN-γ stimulation (37). However, the role of IDO in B cell responses in these environmental settings was never elucidated. In this context, to our knowledge, the present study is the first to demonstrate that intrinsic IDO1 activity regulates proliferation and survival of activated B cells and controls the magnitude of plasma cell differentiation in response to TI Ags.
IDO1 can be induced in a variety of cell types through various stimuli in inflammatory and immunosuppressive conditions (21, 25). The expression of IDO1 in many cases has been shown to be mediated by type I and type II IFNs, which can drive IDO1 expression in an autocrine or paracrine manner (24, 25). Interestingly, we found that IDO1 induction in B cells appears to be triggered independently of type I and type II IFN signaling (Supplemental Fig. 3). Thus, our results suggest that Ido1 expression may be directly triggered by TLRs or strong BCR engagement by Ficoll in vivo.
There are at least two potential mechanisms by which the IDO1 pathway can regulate B cell activity. The best known mechanism of action for IDO-mediated suppression is by intracellular and microenvironmental tryptophan consumption coupled with activation of the integrated stress response kinase, general control nonderepressible 2 (GCN2) (26). Previous studies have demonstrated that IDO1-driven activation of the GCN2 signal drove proliferative arrest in naive T cells and inflammatory T cell apoptosis while at the same time promoting Foxp3+ regulatory T cell differentiation and activation (26). Similarly, arginase-driven metabolism of arginine inhibited naive T cell proliferation by GCN2-dependent downregulation of the TCR ζ-chain and cyclin D3 (38). We found that Ido1 deletion led to a significant increase in cyclin D3 expression (as shown in Fig. 4A) and reduction in proapoptotic protein cleavage (as shown in Fig. 4C). This is consistent with results by Jiang and Wek (39) who reported that GCN2 activation drove caspase-3 cleavage and apoptosis in fibroblasts. IDO1 can also regulate adaptive immunity by tryptophan metabolite–driven activation of the aryl hydrocarbon receptor and downstream signaling events (27). In particular, N-formylkynurenine has been identified as a ligand for the aryl hydrocarbon receptor (27, 40). However, we have found that kynurenines do not alter B cell responses to either NP-Ficoll or NP-LPS in vitro or in vivo, suggesting that GCN2 signals may be the dominant mechanism by which IDO1 regulates TI immune responses.
There have been several recent several studies reporting the role of IDO in B cell response in both nominal and autoimmune environments (41–43). However, these studies were limited to the use of the IDO antagonist D1MT to inhibit IDO activity rather than knock out mice. For instance, previous studies reported that D1MT administration reduced B cell–mediated inflammatory responses and suppressed inflammatory cytokines and autoantibody production in a mouse model of arthritis (41, 43). Moreover, D1MT administration reduced humoral immune response to hepatitis B surface Ag vaccine in BALB/c mice, suggesting, as a whole, that IDO activity is required for B cell responses (44). However, a recent report suggested deletion of Ido2 phenocopied the effect of D1MT with a diminished autoantibody response and reduced arthritis pathology in the KRN.g7 arthritis model, suggesting a differential role for IDO1 and IDO2 in arthritis (42). Thus, it is likely that the role of IDO1 and IDO2 in immunity will be context specific depending on the relative role of B and T cells in the immune response, the stimulus driving IDO activity, and tissue-specific factors.
Based on our results, we propose that a primary role for B cell–derived IDO1 in humoral responses is to restrict B cell hyperactivity to self-TI Ags (such as DNA). This would be in agreement with our previous report that found that apoptotic cell challenge in Ido1−/− mice led to a rapid increase in anti-DNA IgG (21). Thus, it is likely that IDO1 activity, induced as a consequence of innate inflammatory responses, is a key feedback mechanism preventing excessive humoral autoimmunity and limiting the extent of B cell proliferation. In conclusion, this study describes a novel role of IDO1 in regulation of TI B cell responses. The present study, to our knowledge for the first time, defines functional IDO1 activity in B cells and suggests it may be a therapeutic target to augment Ab responses against encapsulated bacteria or to regulate autoimmunity.
Acknowledgements
We thank Drs. David Munn and Andrew Mellor for comments and advice during the development of this research report.
Footnotes
This work was supported by National Institute of Allergy and Infectious Diseases Grants AI099043 and AI105500 (to T.L.M.).
The online version of this article contains supplemental material.
References
Disclosures
The authors have no financial conflicts of interest.