|Home | About | Journals | Submit | Contact Us | Français|
Generation of reactive oxygen intermediates (ROI) following antigen receptor ligation is critical to promote cellular responses. However, the effect of antioxidant treatment on humoral immunity during a viral infection was unknown. Mice were infected with lymphocytic choriomeningitis virus (LCMV) and treated with Mn(III)tetrakis(4-benzoic acid)porphyrin chloride (MnTBAP), a superoxide dismutase mimetic, from days 0 to 8 postinfection. On day 8, at the peak of the splenic response in vehicle-treated mice, virus-specific IgM and IgG antibody-secreting cells (ASC) were decreased 22- and 457-fold in MnTBAP-treated animals. By day 38, LCMV-specific IgG ASC were decreased 5-fold in the bone marrow of drug-treated mice, and virus-specific antibodies were of lower affinity. Interestingly, antioxidant treatment had no effect on the number of LCMV-specific IgG memory B cells. In addition to decreases in ASC, MnTBAP treatment decreased the number of functional virus-specific CD4+ T cells. The decreased numbers of ASC observed on day 8 in drug-treated mice were due to a combination of Bim-mediated cell death and decreased proliferation. Together, these data demonstrate that ROI regulate antiviral ASC expansion and have important implications for understanding the effects of antioxidants on humoral immunity during infection and immunization.
Antibodies are a critical component of the immune system's defense to infectious microorganisms. In order to initiate an antibody response to a pathogen, naïve B cells must first be activated through recognition of antigen by the B cell receptor (BCR). Following antigen stimulation, activated B cells enlist cognate CD4+ T cell help to stimulate clonal expansion (1). Upon activation and proliferation, B cells embark on two distinct differentiation pathways (2). First, the initial production of antibody to a pathogen is accomplished through the differentiation of activated B cells into extrafollicular plasmablasts (3). These short-lived cells are essential in generating low-affinity antibodies early during the infection. However, to generate long-lived humoral immunity, activated B cells must migrate to the germinal center, undergo affinity maturation by somatic hypermutation, and undergo isotype switching to produce memory B cells or antibody-secreting plasma cells (ASC) (3). Memory B cells are long-lived and rapidly respond to pathogen re-encounter by proliferating and differentiating into ASC (4). High-affinity, long-lived ASC migrate to the bone marrow, where they continuously secrete antibody and persist for a year or more in mice (5) and decades in humans (6). Therefore, determining the factors that modulate these pathways is critical not only for understanding the generation and maintenance of serological memory but also for optimizing vaccines and therapeutics for autoimmune disorders.
Following antigen receptor ligation, reactive oxygen intermediates (ROI) are generated and required for B cell function (7–9). Previous work has demonstrated that antioxidant treatment decreased lipopolysaccharide (LPS)-induced B cell proliferation in vitro (10, 11). Singh and colleagues (12) provided the first piece of evidence that ROI produced following B cell activation are critical for calcium flux and amplification of early BCR-induced signals. Consistent with this idea, B cells deficient in ROI-generating proteins exhibit decreased Syk and Akt phosphorylation following activation (7). Additionally, we have previously shown that the first oxidation product of cysteine, sulfenic acid, is a critical oxidative modification required for the induction of capacitative calcium entry and maintenance of tyrosine phosphorylation following BCR ligation (8). Furthermore, ROI have also been implicated in the humoral immune response in vivo. Studies using loss-of-function mutants in the voltage-gated proton channel HVCN1 by Capasso et al. (7) show that ROI act as positive regulators in B cell responses to both T-cell-dependent and -independent model antigens. Taken together, these findings suggest ROI generated following antigen receptor ligation act as positive mediators of B cell responses. In contrast, other studies suggest that the overproduction of ROI associated with chronic viral infections, autoimmune disorders, and cancer plays a role in lymphocyte dysfunction (13–15). Due to the potential role of ROI in pathology, antioxidant supplementation has risen, with 48% of people ingesting one daily (16). The contribution of ROI and antioxidant ingestion to the humoral immune response to a physiological stimulus such as a viral infection is unknown.
In this study, we examined the effect of antioxidant treatment on humoral immune responses during acute viral infection. Mice were treated with Mn(III)tetrakis(4-benzoic acid)porphyrin chloride (MnTBAP), a superoxide dismutase mimetic, from days 0 to 8 during acute lymphocytic choriomeningitis virus (LCMV) infection. We found that early (day 5) expansion of virus-specific IgM and IgG ASC was not affected, but at the peak of the splenic response, MnTBAP-treated mice had 22- and 457-fold-lower numbers of virus-specific IgM and IgG ASC, respectively. When the long-lived ASC in the bone marrow were assessed at day 38, LCMV IgG ASC were found to be decreased 5-fold in the drug-treated mice. Interestingly, the number of virus-specific IgG memory B cells was not affected. In addition, MnTBAP treatment decreased the number of polyfunctional virus-specific CD4+ T cells. The decrease in virus-specific ASC and CD4+ T cells was due to a combination of Bim-mediated cell death and decreased proliferation. Together, these data demonstrate an important role for ROI in controlling the expansion of antigen-specific ASC generated during viral infection and provide a better understanding of how high-dose antioxidant supplementation could affect humoral immune responses following infection and immunization.
Female C57BL/6 mice, 6 to 8 weeks old, were purchased from the National Cancer Institute (Frederick, MD). Bim (Bcl2l11)−/− mice were purchased from Jackson Laboratories. Mice were infected with 2 × 105 PFU of LCMV Armstrong intraperitoneally and used at the time points indicated below. Virus was grown and quantitated as previously described (17). All studies were approved by the Institutional Animal Care and Use Committee (IACUC) of Wake Forest School of Medicine.
MnTBAP treatment was performed as previously described (18). Briefly, MnTBAP (Calbiochem) was dissolved in 0.1 N NaOH as a 6-mg/ml stock and diluted in sterile phosphate-buffered saline (PBS) to give an intraperitoneal dose of 5 mg/kg of body weight. Four hours later, mice were infected with LCMV Armstrong. A maintenance dose was administered to the mice every 24 h for the duration of treatment.
Spleens were removed from mice following cervical dislocation. After the spleen was teased apart on a wire mesh screen, red blood cells were osmotically lysed using ACK lysis buffer (Lonza). Splenocytes were resuspended in complete medium containing RPMI 1640 supplemented with 10% fetal calf serum (FCS; HyClone), l-glutamine (HyClone), penicillin-streptomycin (Cellgro), nonessential amino acids (Gibco), and 2-mercaptoethanol (Gibco). To isolate bone marrow cells, both femurs were flushed with complete medium. Red blood cells were osmotically lysed as described above, and then cells were washed and counted.
In this study, the following antibodies were used: rat anti-mouse CD4-peridinin chlorophyll protein (CD4-PerCP), rat anti-mouse CD44-allophycocyanin (CD44-APC), rat anti-mouse CD62L-APC-Cy7, rat anti-mouse CXCR5-biotin, rat anti-mouse GL7-fluorescein isothiocyanate (GL7-FITC), rat anti-mouse CD138-phycoerythrin (CD138-PE), rat anti-mouse B220-APC-Cy7, rat anti-mouse B220-APC, rat anti-mouse CD19-APC, hamster anti-mouse PD-1-PE, hamster anti-mouse CD95-PE, anti-mouse CD150-Pacific blue surface, anti-mouse IgD-eFluor 450, rat anti-mouse gamma interferon (IFN-γ)-FITC, rat anti-mouse tumor necrosis factor alpha (TNF-α)-PE, and rat anti-mouse interleukin 2 (IL-2)-APC. All antibodies were purchased from BD Pharmingen except for CD19, IgD (eBioscience), and CD150 (BioLegend). Surface staining was performed by incubating cells in a 1:100 dilution of antibody in 2% FACS buffer (PBS plus 2% FCS) for 30 min on ice. Cells were washed three times with FACS buffer and fixed in 2% paraformaldehyde (Sigma-Aldrich, St. Louis, MO). After three washes, intracellular cytokine staining was performed using the BD Cytofix/Cytoperm kit according to the manufacturer's protocol. Samples were acquired on a BD FACSCanto instrument and analyzed using FlowJo software (TreeStar, San Francisco, CA).
For CXCR5, secondary stains were performed. Briefly, cells were incubated in a 1:100 dilution of CXCR5-biotin in FACS buffer for 30 min at room temperature. Following three washes with FACS buffer, cells were stained with PE-Cy7-labeled streptavidin (Invitrogen) for 30 min on ice. Cells then were washed three times with FACS buffer and fixed in 2% paraformaldehyde.
Intracellular IL-21 staining was performed following surface staining as previously described by Johnston et al. (19). Briefly, samples were fixed and permeabilized using BD Cytofix/Cytoperm and incubated with recombinant human IL-21 R subunit Fc chimera (R&D Systems) for 60 min on ice. After two washes, samples were stained with APC-conjugated F(ab′)2 fragment goat anti-human IgG Fcγ for 30 min on ice. Samples were washed twice with PermWash buffer and twice with FACS buffer and then fixed in 2% paraformaldehyde.
IgM and IgG LCMV-specific ASC were quantified using nitrocellulose bottom 96-well plates (Millipore, Billerica, MA) coated with LCMV-infected BHK cell lysate as previously described (20). Briefly, coated plates were washed with PBS containing 0.1% Tween 20 (PBS-T) followed by three washes with PBS. The plates were blocked with complete medium for 1 h at room temperature. Threefold dilutions of cells were incubated on the blocked plate for 5 h at 37°C. Following washes, samples were incubated with biotin-conjugated goat anti-mouse IgM(μ) or IgG(γ) (Invitrogen) for a minimum of 12 h at 4°C. The plates were washed with PBS-T and incubated with horseradish peroxidase-conjugated avidin-D (Vector Laboratories) for 1 h at room temperature. After washes, the plates were developed using hydrogen peroxide chromogen substrate. The membranes were dried, and spots were enumerated.
At the time points indicated below, serum from infected mice was collected, and the LCMV-specific antibody titer was quantified by enzyme-linked immunosorbent assay (ELISA) as previously described (17). The LCMV-specific antibody titer is expressed as the reciprocal of the highest dilution showing a reading for optical density at 492 nm (OD492) greater than three times the average OD492 reading of background levels.
LCMV-specific IgG affinity was determined by a modification of an ELISA method previously described by Bates et al. (21). Briefly, 96-well plates were coated overnight with LCMV-infected BHK cell lysate. The plates were washed three times with PBS containing 0.05% Tween 20 and blocked with 200 μl of PBS supplemented with 10% FCS for 2 h at room temperature. Following three washes, serum samples were added to the wells and incubated overnight at 4°C. The wells were washed five times, and various concentrations of sodium thiocyanate (NaSCN) were added to the plate for 15 min. Samples were washed and incubated with horseradish peroxidase-conjugated anti-mouse IgG (SouthernBiotech) for 2 h at room temperature. Wells were developed with tetramethylbenzidine (Sigma-Aldrich) for 30 min. The reaction was stopped with 2 N H2SO4, and absorbance was measured at 450 nm. The affinity was defined as the concentration of NaSCN required to produce a 50% reduction in absorbance. Preliminary studies were conducted in order to determine that the optimal serum dilution was 1:51,200. At this dilution, no protein precipitate formed during the development of the assay.
LCMV-specific memory B cells were quantified by a modification of a limiting dilution method described previously (22). Briefly, 5-fold dilutions of splenocytes from infected mice were tested in 12-well replicates. Splenocytes were cultured in a 96-well flat-bottom plate for 5 days at 37°C in the presence of 1 × 106 irradiated (1,200 rads) sex-matched feeder splenocytes, 0.6 μg of R595 lipopolysaachride (Alexis Biochemicals), and 20 μl of concanavalin A supernatant in complete medium for a total volume of 200 μl. After 5 days of polyclonal stimulation, cells were harvested, washed, and transferred to a nitrocellulose bottom 96-well plate coated with LCMV-infected BHK cell lysate. LCMV-specific ELISPOT assays were performed as previously described (20).
Mice were administered 0.8 mg/ml of bromodeoxyuridine (BrdU) in the drinking water beginning at day 6 of LCMV Armstrong infection. BrdU staining was performed as previously described by Tebo et al. (23). Samples were acquired on a BD FACSCanto instrument and analyzed using FlowJo software.
Data from vehicle- and MnTBAP-treated mice were analyzed using a two-tailed Student t test. A P value of ≤0.05 was considered significant.
Previous studies have determined that superoxide generated following naïve BCR ligation is critical for cysteine oxidation (8) and signal transduction pathways (7). Therefore, we hypothesized that following in vivo activation, B cells generate ROI, which regulate their responses. To address this hypothesis, we wanted to determine if ROI levels increase in B cells following in vivo activation with a physiological stimulus such as viral infection. Superoxide production was assessed using the fluorescent dye dihydroethidium (HE), a cell-permeant dye that upon oxidation intercalates into DNA. Unactivated B cells were identified as B220+ IgD+ cells, while activated/memory phenotype B cells were identified as B220+ IgDlo cells (19). Figure 1A demonstrates that unactivated B cells from naïve mice contain basal levels of superoxide, with a mean fluorescence intensity (MFI) of 3,225. However, on day 8 after LCMV infection, superoxide levels in activated/memory phenotype B cells had increased to an MFI of 6,377 (Fig. 1B). In contrast, when we assessed the superoxide production in MnTBAP-treated mice on day 8 postinfection, superoxide levels in activated/memory phenotype B cells were found to be modestly reduced, to an MFI of 5,579 (Fig. 1B). When data from multiple mice were normalized to unactivated, naïve B cells, treatment with MnTBAP modestly reduced the levels of superoxide in activated/memory phenotype B cells (fold increases were 1.90 for vehicle-treated cells and 1.50 for MnTBAP-treated cells) (Fig. 1C). Thus, superoxide production increases in B cells following in vivo activation during acute viral infection, and MnTBAP treatment lowers the level of ROI produced.
Because ROI production following receptor ligation promotes signal transduction, we hypothesized that lowering ROI with MnTBAP would decrease B cell responses following viral infection. To address this hypothesis, we administered the antioxidant MnTBAP to mice infected with LCMV Armstrong from days 0 to 8 postinfection. In both vehicle- and MnTBAP-treated mice, LCMV Armstrong induces an acute infection that peaks to similar levels at day 3 and is cleared in 9 to 10 days (18). It also induces a large expansion of B cells that differentiate into antigen-specific ASC which secrete virus-specific antibody for the life of the animal (5). The initial ASC response occurs in the spleen at day 8 postinfection; however, by day 30, 90% of the cells will home to the bone marrow (5). Early during infection, on day 5, vehicle- and drug-treated mice had similar numbers of virus-specific IgM and IgG ASC in the spleen (Fig. 2A and andB).B). At the peak of the response, on day 8 postinfection, antigen-specific IgM and IgG ASC were decreased 22- and 457-fold, respectively, compared to levels in control mice. Virus-specific ASC located in the blood and lymph nodes were also reduced in the MnTBAP-treated mice, suggesting that the cells were not localized to a different tissue (data not shown). ASC were lower at day 10 postinfection; however, the virus-specific IgM and IgG ASC responses in the spleen were similar in the two groups by day 15 postinfection. When the long-lived LCMV-specific ASC population was analyzed on day 38 postinfection, there was a significant reduction in the IgG (5-fold) but not the IgM response in MnTBAP-treated mice (Fig. 2C and andD).D). Additionally, we detected similar kinetics and reductions in the polyclonal plasmablast (CD19hi-int B220hi-int CD138+ IgD−) pool between vehicle- and MnTBAP-treated mice (Fig. 2E). Because virus-specific ASC were decreased with antioxidant treatment, we determined whether total IgG ASC was altered. After 8 days of vehicle or MnTBAP treatment of uninfected mice, there was no significant difference in the number of polyclonal ASC (Fig. 2F). During infection, the total number of IgG-secreting ASC increased to 1.72 × 106 in vehicle-treated mice, and this was reduced 4-fold following MnTBAP administration. Taken together, these data suggest that MnTBAP is not inherently cytotoxic to ASC.
In addition to examining the number of LCMV-specific ASC, we determined whether MnTBAP treatment affected the generation of virus-specific IgG memory B cells. These quiescent and recirculating cells respond rapidly to antigen re-encounter by differentiating into virus-specific ASC (5). Interestingly, the numbers of antigen-specific IgG memory B cells were equivalent between the two groups at day 38 postinfection (Fig. 2G). Therefore, treatment with MnTBAP reduces the number of antigen-specific ASC while having no effect on the generation of memory B cells.
The ability of ASC to produce antibodies is often the immune system's first line of defense to an invading pathogen. As shown in Fig. 2H, we compared the virus-specific antibody titers in vehicle- and MnTBAP-treated mice at day 38 postinfection. In vehicle-treated mice, the serum LCMV-specific IgG titer reached 10,000. When titers in MnTBAP-treated mice were examined, they were found to be reduced to 2,500. In addition to total titer, it is critical to measure the affinity of the antibodies generated. On day 38 postinfection, serum samples from vehicle- and MnTBAP-treated mice were collected, and antibody affinity was determined by the concentration of sodium thiocyanante (NaSCN) required to reduce the antigen-bound antibody absorbance by 50% as measured by ELISA. The average LCMV-specific antibody affinity from vehicle-treated mice was 1.3 M NaSCN, while samples from MnTBAP-treated mice had a significant reduction (0.99 M) in antibody affinity (Fig. 2I). Thus, MnTBAP treatment reduces the affinity and titer of LCMV-specific IgG antibody.
Previous studies have shown that germinal center B cell formation is critical to the differentiation of long-lived, high-affinity ASC (24). Because MnTBAP treatment decreased the overall titer and affinity of the anti-LCMV antibody, we determined whether the number of germinal center B cells was altered. Figure 3 shows that early during infection, on day 5, the numbers of germinal center B cells (B220+ GL7+ Fas+) were roughly equivalent in the two groups of mice (1.53 × 105 for vehicle versus 1.14 × 105 for MnTBAP). However, in vehicle-treated mice, the number of germinal center B cells expanded to 1.36 × 106 cells at day 8 and remained elevated through day 15 postinfection. This result contrasts with that for the MnTBAP-treated mice, which had a substantial reduction in germinal B cells at days 8 and 10 postinfection. However, by day 15 postinfection, germinal center B cell numbers in MnTBAP-treated mice reached a similar peak. By day 38 postinfection, the numbers of germinal center B cells declined and were equivalent between the two groups. Therefore, treatment with MnTBAP alters the number of germinal center B cells and the kinetics of their formation.
Cognate CD4+ T cell help is required for effective B cell responses to T-cell-dependent antigens such as LCMV. Because the number of virus-specific ASC was decreased, we determined whether CD4+ T cell help was altered in MnTBAP-treated mice. To measure functional CD4+ T cells, we performed intracellular cytokine staining for IFN-γ, TNF-α, and IL-2 following in vitro GP61-80 stimulation. Early during the effector response, on day 5, we observed similar percentages of cells producing cytokines in vehicle- and MnTBAP-treated mice (Fig. 4A and andB).B). By day 8 postinfection, 16.03% of CD4+ T cells produced only IFN-γ, 10.6% produced both IFN-γ and TNF-α, and 8.35% produced IFN-γ and IL-2. MnTBAP-treated mice showed a decrease in the percentage of cells producing all cytokines such that only 2.16% produced IFN-γ, 1.28% produced both IFN-γ and TNF-α, and 1.25% produced IFN-γ and IL-2. Similar trends were observed at days 10 and 15 postinfection. Analysis of the memory phase revealed that MnTBAP mice had a decrease in the percentage of cells producing IFN-γ and IL-2. In accordance with the percentage data, we observed similar reductions in the total number of functional CD4+ T cells during the effector (day 8, 13-fold) and memory (day 38, 2-fold) phases in MnTBAP-treated mice (Fig. 4C). Furthermore, MnTBAP-treated mice exhibited a reduction in the percentage of IFN-γ cells also producing TNF-α and IL-2 at day 15 (Fig. 4D and andE).E). Taken together, antioxidant treatment decreases the number and polyfunctionality of antigen-specific effector and memory CD4+ T cells.
Previous studies have demonstrated the requirement of T follicular helper (TFH) cells for the development and maintenance of germinal center B cells (19, 25, 26). Because the number of germinal center B cells was significantly decreased in MnTBAP-treated mice, we determined whether the number of TFH cells was altered with antioxidant treatment. TFH cells were identified as activated (CD44hi CD62Llo) CD4+ T cells that express high levels of CXCR5 and low levels of SLAM (27). Figure 5A shows that the numbers of TFH cells were similar for vehicle- and MnTBAP-treated mice early during infection (day 5). However, at days 8 and 10 postinfection, we detected fewer TFH cells in the drug-treated mice relative to control mice. Similar results were observed when we assessed TFH cells by Bcl6 and CXCR5 staining (data not shown). Prior studies have demonstrated that TFH cells within the germinal center (GC TFH) have a greater ability to provide help to B cells (28, 29). Early during infection, at day 5, vehicle- and MnTBAP-treated mice contained similar numbers of this population in the spleen; however, by days 8 and 10 postinfection, the numbers were significantly decreased, 8.3- and 4.6-fold, respectively, in drug-treated mice (Fig. 5B). Although the numbers were decreased at days 8 and 10 postinfection in MnTBAP-treated mice, this population increased in drug-treated mice to levels similar to those in the control mice by day 15 postinfection.
TFH cell-derived IL-21 is critical for proliferation and survival of germinal center B cells, isotype switching, and ASC differentiation (30, 31). In Fig. 5C, we compared the abilities of CD4+ T cells from vehicle- and MnTBAP-treated mice to produce IL-21 when stimulated with phorbol myristate acetate (PMA) and ionomycin in vitro. Early during infection, on day 5, MnTBAP-treated mice had numbers of IL-21-producing CD4+ T cells similar to those of the control mice. At day 8 postinfection, the number of functional cells was significantly reduced in the drug-treated mice (1.28 × 106 for vehicle versus 1.85 × 105 for MnTBAP). Similar trends were observed at day 10 postinfection. However, when we assessed the cells on day 15 postinfection, we found that the numbers of CD4+ T cells producing IL-21 were similar for the two groups. Taken together, these results demonstrate that antioxidant treatment decreases the number and function of TFH cells at days 8 and 10 postinfection. However, this loss was temporary, because MnTBAP-treated mice recovered by day 15 postinfection.
Cell number is a balance of proliferation and cell death. While we observed similar numbers of virus-specific ASC and CD4+ T cells in vehicle- and MnTBAP-treated mice at day 5 postinfection, there was a significant reduction in drug-treated mice by day 8. To determine whether proliferation was altered in the MnTBAP-treated mice, we administered 0.8 mg/ml of bromodeoxyuridine (BrdU) in the drinking water of both groups of mice at day 6 postinfection. A mouse receiving water alone was used as a control. This time point was selected because it was an intermediate point between days 5 and 8 postinfection, when we observed a decrease in cell number. Spleens were harvested from mice at day 7 postinfection and stained for surface markers and intracellular BrdU. In vehicle-treated mice, 38.6% of plasmablasts were BrdU+, compared to 13.9% in MnTBAP-treated mice (Fig. 6A). Similar trends were observed in the activated CD4+ T cell compartment (47.3% for vehicle versus 16.3% for MnTBAP) (Fig. 6B). Data from multiple mice revealed a significant decrease in the percentages of plasmablasts and activated CD4+ T cells that were BrdU+ in the MnTBAP-treated mice (Fig. 6C). These results indicate that plasmablasts and activated CD4+ T cells in MnTBAP-treated mice underwent reduced proliferation.
In addition to measuring proliferation, we assessed the contribution of cell death by examining whether mice deficient in Bim (Bcl2l11), a BH3-only proapoptotic member of the Bcl-2 superfamily, could rescue the cell number defect we observed in the MnTBAP-treated mice at day 8 postinfection. Wild-type and Bcl2l11−/− mice were treated with vehicle or MnTBAP and then infected with LCMV Armstrong. The humoral and cell-mediated responses were measured at day 8 postinfection. As we previously observed, wild-type MnTBAP-treated mice had significant decreases in the numbers of splenic virus-specific IgM and IgG ASC (Fig. 6D) and functional antigen-specific CD4+ T cells (Fig. 6E) compared to vehicle-treated mice. When Bcl2l11−/− mice were examined, both the IgM and IgG virus-specific ASC were found to be increased in vehicle-treated mice compared to the levels observed in wild-type animals. Strikingly, loss of Bim rescued much of the decrease in ASC observed at day 8 in MnTBAP-treated mice. Similar trends were observed when we analyzed the number of antigen-specific CD4+ T cells (Fig. 6F). Since the loss of Bim rescued most of the antigen-specific ASC and CD4+ T cells in MnTBAP-treated mice, we wanted to determine whether viral replication was altered. Similar to our prior studies (18), we observed that wild-type vehicle-treated mice had viral titers of 1.5 × 104 PFU/g in their spleens at day 8 postinfection. This contrasted with findings for MnTBAP-treated wild-type mice, in which viral titers were 3.8 × 105 PFU/g. Treatment with MnTBAP delays viral clearance from day 9 to 10 postinfection in wild-type mice (18). When vehicle- and MnTBAP-treated Bcl2l11−/− mice were examined, viral titers were found to be 3.2 × 104 and 5 × 104 PFU/g, respectively (Fig. 6G). Importantly, titers were not significantly altered with treatment in Bcl2l11−/− mice, demonstrating that the rescue in LCMV-specific ASC was not due to differences in viral load. Taken together, these results demonstrate that the alteration in antigen-specific ASC and activated CD4+ T cells in drug-treated mice is due to a combination of decreased proliferation and increased cell death.
In this study, we utilized the antioxidant MnTBAP to examine the role of ROI in the humoral immune response during a viral infection. Here, we report five novel observations. First, in response to viral infection, activated/memory phenotype B cells increased their levels of ROI. Second, antioxidant treatment reduced proliferation and increased Bim-mediated cell death of virus-specific IgM and IgG ASC during the effector phase. Third, antioxidant treatment decreased the number and functionality of virus-specific effector and memory CD4+ T cells. Fourth, MnTBAP treatment decreased the affinity of the anti-LCMV IgG antibody at day 38 postinfection. Fifth, antioxidant treatment had no effect on the number of virus-specific IgG memory B cells generated. Taken together, these results demonstrate that ROI generated in response to an acute viral infection are critical in controlling both the antiviral humoral and CD4+ T cell responses.
What are the roles of ROI in B cell responses? Prior work from our laboratory and others has demonstrated that ROI are generated in vitro following BCR ligation by antibodies (7–9) or antigen (8). Here, we extend these studies, demonstrating for the first time that ROI are elevated in activated B cells directly ex vivo following acute viral infection. There have been conflicting reports regarding the role of these molecules in B cell responses. Addition of antioxidants, which lower ROI, decreases B cell proliferation in vitro. This suggests that ROI act in a positive manner to promote B cell proliferation. The role of ROI in B cell responses in vivo is less clear. Richards and Clark (9) utilized mice deficient in the gp91 subunit of the NADPH oxidase complex and demonstrated that T-cell-independent responses to model antigens were increased, while T-cell-dependent responses were unaffected. These results suggest that superoxide and other ROI negatively regulate responses. Using HVCN1-deficient mice, which have decreased ROI production, Capasso and colleagues (7) demonstrated that T-cell-dependent responses were decreased following 4-hydroxy-3-nitrophenyl acetyl-keyhole lympet hemocyanin (NP-KLH) immunization. This supports a positive role for ROI in promoting B cell responses in vivo. We addressed the role of ROI in a controlled window from days 0 to 8 postinfection through antioxidant administration. Directly ex vivo, we observed a 22% decrease in ROI levels in activated B cells and a 457-fold decrease in LCMV-specific IgG-secreting ASC by day 8. Although we were able to modestly reduce ROI in B cells, it is possible that decreasing ROI in other cell types could contribute to the effects we observed. Indeed, in our prior studies of antioxidant treatment on CD8+ T cell responses (18), we demonstrated that MnTBAP treatment lowered ROI in activated T cells. Interestingly, our data presented here, along with our prior studies of CD8+ T cells (18), demonstrate that the initial phase of the antiviral response (~5 days) is refractory to the effects of antioxidants. This contrasts with in vitro studies in which activation and early proliferation are suppressed. One potential explanation is that in vitro, superphysiological (millimolar) concentrations of antioxidant are utilized to suppress activation. In vivo, it may take several days to reach a concentration of antioxidant sufficient to alter responses. Taken together, the decreased LCMV-specific ASC we observed in MnTBAP-treated mice is consistent with a model in which ROI are positive regulators of B cell responses in vivo.
How do ROI positively regulate humoral immune responses in vivo? In this study, we observed decreases in LCMV-specific IgG ASC during both the effector and memory phases. Our BrdU incorporation experiments demonstrate that part of the expansion defect is due to decreased proliferation. Multiple checkpoints in the cell cycle are regulated by ROI, including the G1/S phase and G2/M transitions (32). Cyclin D1, a critical regulator of the G0/G1 transition, is sustained when Nox enzymes are overexpressed under low-serum conditions (33). Treatment with MnTBAP, which lowers superoxide, could lower cyclin D1 levels and decrease the expansion of both antiviral ASC and CD4+ T cells. Similarly, studies by Piganelli and colleagues (34) demonstrated that treatment with AEOL-10113, a superoxide dismutase mimetic, reduced CD4+ T cell proliferation in a diabetes model. Our results extend these findings to an acute infectious disease model and show that expansion of TFH cells is also inhibited by antioxidants. In addition to decreased proliferation, infection of Bim-deficient mice demonstrates that cell death is a major contributor to the expansion defect in ASC of MnTBAP-treated mice. What are the mechanisms by which Bim-mediated cell death is increased with drug treatment? Bim exerts prodeath effects through interactions with prosurvival proteins such as Bcl-2 and Mcl-1. Importantly, we did not observe differences in Bcl-2 levels in activated B cells in MnTBAP-treated mice (data not shown). Previous studies with multiple cell types have demonstrated that the mitogen-activated protein (MAP) kinase cascade is redox regulated (35–39). Additionally, it is well documented that MAP kinase-mediated phosphorylation of BimEL and its subsequent proteosomal degradation are critical to B and T cell survival upon stimulation (40). Therefore, alterations in MAP kinase signaling in the presence of MnTBAP could decrease BimEL phosphorylation, increasing its stability and rendering cells more susceptible to death. Lastly, our observations of decreased numbers of antigen-specific ASC, but not memory B cells, in MnTBAP-treated mice are similar to observations following antigen immunization of XBP-1−/− mice (41). Prior studies have clearly documented that oxidation of transcription factors such as NF-κB promotes DNA binding (42, 43). Thus, XBP-1 or other transcription factors known to be critical for ASC generation might be highly sensitive to antioxidants because of potential oxidation sites.
Because of their roles in multiple cell types, there is interest in manipulating ROI for therapeutic benefit. Previous studies determining the efficacy of antioxidants in various disorders, including coronary artery disease and nephropathy, have included administration of 600 mg of N-acetylcysteine to patients (44, 45). Based on this dosage, patients typically ingest ~5 to 6 mg per kilogram, which is comparable to the amount of MnTBAP we used in this study. A recent study documented that vitamin C, α-tocopherol, carotenes, selenium, and flavonoids are the most commonly ingested antioxidants in the United States (16). These common antioxidants are beneficial in the treatment and prevention of chronic disorders. For example, there is an inverse relationship between β-carotene and vitamin C ingestion and the incidence of breast cancer in premenopausal women (46, 47). Additionally, studies suggest that diets rich in flavonoids reduce the risk of cardiovascular disease in humans and rodent models (48, 49). Due to positive effects in various disease models, there are currently 1,986 clinical trials using antioxidants (www.clinicaltrials.gov). However, while antioxidant ingestion may play a role in the prevention of chronic disorders, our results have important implications for supplementation during acute immune responses. Our study demonstrates that antioxidant with superoxide dismutase activity has a negative effect on virus-specific ASC, antibody titer, and affinity and suggest that antioxidant supplements should be suspended during infection and immunization.
In conclusion, we demonstrate that treatment with the antioxidant MnTBAP reduces the expansion of virus-specific antibody-secreting cells through decreased proliferation and increased cell death during the effector phase. These effects were also evident during the memory phase, in which the number of LCMV-specific ASC was decreased in the bone marrow and the overall antibody titer and affinity were lower. In addition to effects on B cells, antioxidant treatment decreased the expansion of virus-specific effector and memory CD4+ T cells. Our results demonstrate that ROI positively regulate antibody-secreting cell responses during acute viral infection and suggest that antioxidant treatment could benefit diseases in which inappropriate generation of these cells occurs.
This work was supported by NIAID grant RO1-AI068952 to J.M.G. K.E.C. was supported by NIAID grant 5T32AI007401-20.
Published ahead of print 19 December 2012