The forkhead transcription factor Foxj1 inhibits spontaneous autoimmunity, in part by antagonizing NF-κB activation in T cells. We demonstrate here that Foxj1 also inhibits humoral immune responses intrinsically in B cells; Foxj1 deficiency in B cells results in spontaneous and accentuated germinal center formation, associated with the development of pathogenic autoantibodies and accentuated responses to immunizations—all reflecting excessive activity of NF-κB and its target gene IL-6, and correlating with a requirement for Foxj1 to regulate the inhibitory NF-κB component IκBβ. Thus, Foxj1 restrains B cell activation and the maturation of humoral responses, demonstrating a critical role for at least this forkhead transcription factor in the regulation of B lymphocyte homeostasis.

Systemic autoimmune syndromes like lupus reflect pathogenic, hyperactive T and B cell responses that culminate in the production of pathogenic autoantibodies, such as anti- dsDNA (1). Recent studies suggest that deficient functions in forkhead transcription factors, such as Foxj1 and Foxo3a, predispose to such diseases, because their activities are significantly diminished in lupus lymphocytes and their deficiency results in spontaneous, multisystem autoimmune syndromes characterized by autoreactive, hyperactivated T cells (2, 3). Presumably, such hyperactivated T cells can promote pathogenic autoantibody production by overexpressing costimulatory molecules and/or effector cytokines, resulting in excessive autoreactive B cell activation. However, both Foxj1 and Foxo3a are also expressed in B cells (2), and a potential immunoregulatory role for the forkhead genes in B cells remains as yet undefined.

Pathogenic autoantibody responses require somatic hypermutation and class switching of natural autoantibodies, such as in the context of T-dependent germinal centers, to develop high-affinity, pathogenic activities, e.g., the production of anti-dsDNA from germline-encoded anti-ssDNA specificities (1, 4). Most circulating natural autoantibodies are produced by natural Ab-secreting B cells, such as B1 cells, or by plasma cells that have not undergone either affinity maturation or class switch recombination (5, 6). As a result, the entry of an autoreactive B cell, when activated, into a germinal center reaction, therefore, may contribute to and/or underlie systemic autoimmune diseases like lupus (4, 6).

To investigate the role of Foxj1 in B cells, we initiated a series of experiments, including adoptive transfers, to isolate and elucidate the B cell-intrinsic function(s) of Foxj1, particularly its relationship to Ag-specific Ab responses. We find that Foxj1 modulates germinal center B cell formation in vivo via its ability to inhibit the NF-κB-regulated cytokine IL-6 (7, 8). Foxj1 appears to prevent inappropriate B cell responses at least in part by antagonizing NF-κB target genes like IL-6, which would otherwise propagate pathogenic autoimmune responses through dysregulated B cell hyperactivation.

129, BALB/c, C57BL/6, C57BL/6-IgHa, C57BL/6-CD45.1, BXSB, F1(NZW × NZB), MRL/+, MRL/lpr, and athymic C57BL/6-nu/nu mice (The Jackson Laboratory), Rag-2−/− (Taconic Farms), and C57BL/6 Foxj1 −/− mice (9) were maintained under specific pathogen-free conditions at the Washington University School of Medicine. As judged by microsatellite markers, mice mutant for Foxj1, which is located on chromosome 11 at 78.0 cM, were homozygotic for C57BL/6 loci, including marker D11Mit333 (66.0 cM; 11qter is ∼80.0 cM), indicating a <15 cM residual 129 contribution on chromosome 11. Foxj1 +/+ and −/− fetal liver chimeras (FLCs)3 were generated in irradiated Rag-2-deficient hosts as previously described (2). For B cell-only chimeras (BOC), fetal livers from Foxj1 +/+ vs −/− (CD45.2+IgHb) embryos were adoptively transferred into irradiated C57BL/6-nu/nu hosts, and reconstitution of the peripheral B cell lineage was allowed for 8–12 wk. Then, splenic B cell populations were purified by negative selection against CD43 (Miltenyi Biotec), and adoptively transferred into C57BL/6-IgHa (for serological studies) or C56BL/6-CD45.1+ (for flow cytometric studies) animals (one spleen equivalent, ∼30–40 million B cells, per recipient), and animals were studied 1–2 wk thereafter, as indicated in the text. All experiments were performed in compliance with the relevant laws and institutional guidelines, as overseen by the Animal Studies Committee of the Washington University School of Medicine.

Immunizations with 4-hydroxy-3-nitrophenylacetyl (NP)-chicken γ globulin and NP-Ficoll were performed with 50 μg of Ag in PBS (without adjuvant) administered i.p. as previously described (4). Where indicated, animals were treated with neutralizing anti-IL-6 (MP5-20F3; BD Pharmingen) or control rat IgG Ab, 5 mg i.p. three times per week (10); or with RELA/p65 antisense (5′-GAAACAGATCGTCCATGGT) or mismatch (5′-GGAACAGTTCGTCTATGGC) oligonucleotides, 800 μg i.v. daily (11). ELISA-based serological assessments, including determination of anti-ssDNA using sheared salmon sperm DNA, anti-dsDNA using Crithidia luciliae immunofluorescence, IgK rheumatoid factor (RF) activity using Igλ IgGs, and anti-hapten responses using NP-BSA and TNP-BSA, were performed as described (4, 12). Germinal centers were identified by staining frozen spleen sections with peanut agglutinin (PNA)-FITC, as previously described (4).

B cell purification and stimulation was performed similarly to a previous study (13). Briefly, for FLCs and unmanipulated mice, naive-enriched B cells were purified from spleens by negative selection against CD43 (Miltenyi), and were further purified over a discontinuous Percoll gradient (70/66/60/50%), with resting B cells isolated from the 66–70 interface. For BOCs, CD45.2+CD43 B cells were purified by flow cytometry, followed by Percoll gradient purification. Cells were cultured in RPMI 1640 medium supplemented with 10% FCS (BioWhittaker) and 100 U of penicillin/streptomycin (Sigma-Aldrich), in the presence or absence of 25 μg/ml LPS (LPS; Sigma-Aldrich), 2 μg/ml anti-CD40 Ab (BD Pharmingen), 5–10 μg/ml anti-IgM (Jackson Immunologicals), 3 mM CpG-1 stimulatory phosphorothioate oligonucleotide 5′-TCCATGACGTTCCTGACGTT, 100 ng/ml IFN-γ, 5 ng/ml TGF-β, and/or 10 ng/ml IL-4 (PeproTech). Where indicated, phosphorothioate decoy or control decoy annealed NF-κB oligonucleotides, which inhibit the activity of all NF-κB subunits, were added at 10 μM (14). Real-time PCR detection of Foxj1, IL-6, CD80, CD86, ICOS, ICOSL, CD40, CD154, bcl-6, Blimp-1, IRF-4, Mitf, c-myc, Pax5, and Xbp1 were performed as described, with normalization against β-tubulin (13, 15). Western blot analyses of NF-κB activities and proteins were performed as previously described (2).

A promoter-reporter construct for IL-6 was constructed by PCR from C57BL/6 genomic DNA, using primers 5′-GGGGTACCATTCAAATCCTGTCATCCAGTAGAAGGGAG and 5′-GAAGATCTGAAAACCGGCAAGTGAGCAGATAGCACAGT, producing an ∼1257-bp fragment corresponding to the putative promoter (−1058/+199), flanked by KpnI and BglII restriction sites. The amplicon was cloned into the KpnI-BglII sites of TK-luc and then confirmed by routine sequencing, generating IL-6-luc. Reporter assays involved Dual-Luciferase (Promega) assays using M12 murine B cell lymphoma cells, electroporated in the presence of 10 μg of IL-6-luc or 20 μg of NF-κB-luc, 400 ng of pRL-CMV (Renilla luciferase control reporter; Promega), and 10 μg of pcDNA3 (Invitrogen Life Technologies) or pcDNA3-Foxj1, as described (2). Primary B cell transfection was performed by a modification of a previously described protocol (16): 107 cells, purified by negative selection and Percoll gradient centrifugation as described above, were incubated for 10 min at room temperature in 0.4-cm cuvettes in 400 μl of RPMI 1640 medium supplemented with 10% FCS, 100 μg of NF-κB-luc, and 1 μg of pRL-CMV. The cells were then electroporated with a Bio-Rad electroporation system at 280 V, 975 μF, incubated at room temperature for 5 min, and then cultured in 5–10 ml of RPMI 1640 medium supplemented with 10% FCS and 25 μg/ml LPS. Dual-Luciferase assays were then performed after 4–5 h of incubation at 37°C.

To assess for somatic hypermutation, NP-binding B cells from immunized FLCs were purified by flow cytometry using NP-FITC-BSA (Biosearch Technologies), and their genomic DNA amplified by PCR using VH186.2- and JH4-specific primers (17, 18). For BOCs, Foxj1 +/+ or −/− fetal livers (CD45.2+) were used first to chimerize nu/nu recipient mice, the splenic B cells of which were then adoptively transferred into wild-type C57BL/6-CD45.1+ recipient mice. After immunization, NP-binding, CD45.2+ B cells were purified by flow cytometry; representative rates of chimerization, subdivided by representative B cell subsets, are shown in Table I. The PCR products were cloned into pCR2.1-TOPO (Invitrogen Life Technologies), and their sequence was determined by routine sequencing.

Table I.

B cell characteristics of Foxj1 BOCsa

DonornSplenocytes (×10−6)B Cells (B220+)NP-BindingGC-Like (PNA+B220+)
Percentage of splenocytesChimeric (%)Percentage of splenocytesChimeric (%)Percentage of B cellsChimeric (%)
WT 67.0 ± 2.5 62.2 ± 1.5 2.2 ± 0.16 2.2 ± 0.32 2.6 ± 0.38 3.9 ± 1.1 2.7 ± 0.3 
KO 68.8 ± 3.8 64.6 ± 4.2 9.9 ± 0.76∗ 2.8 ± 0.13 5.6 ± 0.32b 20.6 ± 2.4b 52 ± 4.3b 
DonornSplenocytes (×10−6)B Cells (B220+)NP-BindingGC-Like (PNA+B220+)
Percentage of splenocytesChimeric (%)Percentage of splenocytesChimeric (%)Percentage of B cellsChimeric (%)
WT 67.0 ± 2.5 62.2 ± 1.5 2.2 ± 0.16 2.2 ± 0.32 2.6 ± 0.38 3.9 ± 1.1 2.7 ± 0.3 
KO 68.8 ± 3.8 64.6 ± 4.2 9.9 ± 0.76∗ 2.8 ± 0.13 5.6 ± 0.32b 20.6 ± 2.4b 52 ± 4.3b 
a

BOCs were generated first by chimerization of nu/nu animals by Foxj1 +/+ or −/− fetal livers (CD45.2), the B cells of which were then adoptively transferred into wild-type C57BL/6-CD45.1 animals. Shown are flow cytometric data (means ± SD) of representative spleens from such BOCs, quantifying the subpopulations indicated, after immunization with NP-Ficoll. Chimeric percentages indicate the percentage of CD45.2-staining cells of the subpopulations indicated.

b

Significant differences between wild-type (W/T) and knockout (K/O) donor cohorts (p < 0.001).

Foxj1 is prominently expressed in naive B cells, where, like T cells, it is significantly down-regulated in response to activation (Ref. 2 and Fig. 1,A, p < 0.001 comparing any stimulation condition to no treatment). In addition, naive, resting B cells from prediseased lupus-prone mice expressed significantly lower levels of Foxj1 (Fig. 1 B; p < 0.001 comparing MRL/+, BXSB, or NZB/W to 129, BALB, or C57BL/6). Although we could not completely eliminate the possibility that such results simply reflect a global B cell hyperactivation seen in such autoimmune mouse strains (1), these findings in highly purified, resting B cells raised the possibility that loss-of-function in Foxj1 might contribute to B cell hyperactivation and/or tolerance loss in autoimmunity, as it appears to do for T cells (2).

FIGURE 1.

Expression pattern of Foxj1 in B cells. Foxj1 expression was assessed by real-time PCR on cDNA from C57BL/6 naive B cells treated with the indicated stimuli in vitro for 24 h (A) or naive B cells purified from 6-wk-old mice of the indicated strains (B). Error bars indicate SDs for individually tested cells from three animals.

FIGURE 1.

Expression pattern of Foxj1 in B cells. Foxj1 expression was assessed by real-time PCR on cDNA from C57BL/6 naive B cells treated with the indicated stimuli in vitro for 24 h (A) or naive B cells purified from 6-wk-old mice of the indicated strains (B). Error bars indicate SDs for individually tested cells from three animals.

Close modal

However, analysis of Foxj1−/− FLCs revealed normal numbers of B1, marginal zone, T1, T2, and follicular B cells, and also demonstrated that overall titers of spontaneous anti-DNA and RF autoantibodies are generally comparable to their +/+ counterparts (Fig. 2,A), prompting our prior conclusion that Foxj1 plays a dispensable role in B cell homeostasis (2). Nonetheless, close further inspection of the data in repeated experiments revealed that Foxj1 −/− sera consistently exhibited higher spontaneous activities of both anti-DNA and RF, at least as judged by OD values (Fig. 1 A, comparing OD values of −/− vs +/+ sera at 1/100–1/300 dilution). This consistent, albeit subtle, dissociation between titers and OD activities suggested that the Foxj1 −/− autoantibodies were, in fact, of higher affinity, as opposed to higher quantity. Indeed, all Foxj1−/− sera tested positive for anti-dsDNA activity by Crithidia luciliae immunofluorescence, in contrast to Foxj1+/+ sera (14 of 14 vs 0 of 12 sera, respectively; p < 0.0001). Because, in general, only high-affinity anti-dsDNA, but not lower affinity anti-ssDNA, Abs are capable of exhibiting Crithidia immunoreactivity (19) and of mediating immune-complex disease (20), these findings together suggested the development of only natural, low-affinity autoimmunity in Foxj1+/+ FLCs, but the maturation of this response to a high-affinity, potentially pathogenic autoimmunity in Foxj1−/− FLCs; thus, Foxj1 might also play a negative regulatory role in B cells, as it does in T cells (2).

FIGURE 2.

Ab responses in the absence of Foxj1. “Standard” FLCs (AC) were generated by chimerization of Rag-deficient animals by Foxj1 +/+ or −/−fetal livers; BOCs (DF) were generated first by chimerization of nu/nu animals by Foxj1 +/+ or −/− fetal livers (IgHb), the B cells of which were then adoptively transferred into wild-type C57BL/6-IgHa animals. By ELISA, sera from these chimeras were assessed for the presence of spontaneous IgG (FLC, A) or IgGb (BOC, D) anti-DNA or IgK RF autoantibodies at the indicated dilutions; hapten-specific IgG (FLC, B) or IgGb (BOC, E) isotype Abs 35 days after immunization at the indicated dilutions, or affinity maturation of IgG (FLC, C) or IgGb (BOC, F) anti-hapten responses at the times indicated after immunization by relative reactivity against NP3-BSA vs NP34-BSA at 1/100 dilution. In A and D, B and E, and C and F, n = 4, 3, and 5, respectively, of each genotype, representative of at least three trial experiments containing the same number of animals. For reference, sera from 12-wk-old lupus-prone MRL/lpr mice developed anti-DNA and RF activities of 1.8–2.0 OD405 at 1/100 dilution, with titers of at least 1/2000 (data not shown). Dashed lines (A, B, D, and E) indicate the threshold for positivity, as determined by three SDs above the average OD405 generated by sera at 1/100 dilution from nonautoimmune BALB/c (A and D, which have lower spontaneous autoantibody activity than C57BL/6 animals) or unimmunized Foxj1 +/+ FLC mice (B and E).

FIGURE 2.

Ab responses in the absence of Foxj1. “Standard” FLCs (AC) were generated by chimerization of Rag-deficient animals by Foxj1 +/+ or −/−fetal livers; BOCs (DF) were generated first by chimerization of nu/nu animals by Foxj1 +/+ or −/− fetal livers (IgHb), the B cells of which were then adoptively transferred into wild-type C57BL/6-IgHa animals. By ELISA, sera from these chimeras were assessed for the presence of spontaneous IgG (FLC, A) or IgGb (BOC, D) anti-DNA or IgK RF autoantibodies at the indicated dilutions; hapten-specific IgG (FLC, B) or IgGb (BOC, E) isotype Abs 35 days after immunization at the indicated dilutions, or affinity maturation of IgG (FLC, C) or IgGb (BOC, F) anti-hapten responses at the times indicated after immunization by relative reactivity against NP3-BSA vs NP34-BSA at 1/100 dilution. In A and D, B and E, and C and F, n = 4, 3, and 5, respectively, of each genotype, representative of at least three trial experiments containing the same number of animals. For reference, sera from 12-wk-old lupus-prone MRL/lpr mice developed anti-DNA and RF activities of 1.8–2.0 OD405 at 1/100 dilution, with titers of at least 1/2000 (data not shown). Dashed lines (A, B, D, and E) indicate the threshold for positivity, as determined by three SDs above the average OD405 generated by sera at 1/100 dilution from nonautoimmune BALB/c (A and D, which have lower spontaneous autoantibody activity than C57BL/6 animals) or unimmunized Foxj1 +/+ FLC mice (B and E).

Close modal

To gain insight into the mechanisms by which Foxj1 might regulate B cells, we turned to model immunogens and assessed the response of Foxj1 −/− FLCs to immunization with the T-independent (Ti) Ag, NP-Ficoll, and the T-dependent (Td) Ag, NP-chicken γ globulin. In response to NP-Ficoll, Foxj1 −/− FLCs mounted exaggerated anti-hapten responses, with higher titers (Fig. 2,B; median titers 1/400 vs 1/1600 on TNP3-BSA, respectively; p < 0.001). Interestingly, Foxj1 −/− Ti anti-NP responses also appeared to be of higher affinity than Foxj1+/+ responses, as judged by the relative activity of their sera against TNP3-BSA vs TNP34-BSA (Fig. 2,C; p < 0.0001 comparing Foxj1 +/+ to −/− affinities at days 28 and 35). Analogous findings were observed during Td immunization, which also elicited higher titer anti-hapten responses in Foxj1−/− than Foxj1+/+ FLCs (Fig. 2,B and data not shown; median titers 1/8100 vs 1/900 on TNP3-BSA, respectively; p < 0.001) and appeared to be associated with a more rapid onset of affinity maturation in Foxj1−/− FLCs (Fig. 2 C; p < 0.001 comparing Foxj1 +/+ to −/− affinities at day 14, p < 0.01 at day 28). Therefore, we conclude that Ag-specific humoral immune responses are accentuated in the absence of Foxj1.

For evidence that somatic hypermutation, in fact, could be responsible for the apparently increased affinity maturation in Foxj1−/− FLCs, we analyzed IgH sequences. NP-binding B cells were purified by flow cytometry from Foxj1 +/+ and −/− FLCs 35 days after immunization with NP-Ficoll, their genomic DNA was amplified by PCR using VH182.6-specific primers, and sequences were determined by routine cloning and sequencing (Fig. 3). Strikingly, mutations were abundant in sequences from Foxj1−/−, but not Foxj1+/+ FLCs, particularly the CDR1 and CDR2 regions, reflecting an ∼10-fold increase in somatic hypermutation frequency per base pair (Fig. 3, p < 0.0001).

FIGURE 3.

Somatic hypermutation in the absence of Foxj1. NP-binding B cells were purified by flow cytometry from Foxj1 −/− vs Foxj1 +/+ FLCs. A, Sequences of 10 individual clones obtained 35 days after immunization with NP-Ficoll from Foxj1 −/− vs +/+ FLCs are shown, representative of 20 total clones sequenced for each genotype. Shaded regions are marked as corresponding to CDR1 and CDR2. The germline VH186.2 sequence is shown. B, Mutation frequency was calculated as the percentage of point mutations observed per base pair. Similar results were obtained with BOCs, analyzing NP-binding CD45.2+ B cells (BOC in B).

FIGURE 3.

Somatic hypermutation in the absence of Foxj1. NP-binding B cells were purified by flow cytometry from Foxj1 −/− vs Foxj1 +/+ FLCs. A, Sequences of 10 individual clones obtained 35 days after immunization with NP-Ficoll from Foxj1 −/− vs +/+ FLCs are shown, representative of 20 total clones sequenced for each genotype. Shaded regions are marked as corresponding to CDR1 and CDR2. The germline VH186.2 sequence is shown. B, Mutation frequency was calculated as the percentage of point mutations observed per base pair. Similar results were obtained with BOCs, analyzing NP-binding CD45.2+ B cells (BOC in B).

Close modal

Because somatic hypermutation, affinity maturation, as well as isotype switching of B cells can generally be attributed to germinal centers, these findings suggested that Foxj1 deficiency might accentuate humoral responses by promoting germinal center formation. Indeed, while Foxj1+/+ FLCs failed to develop significant numbers of germinal centers, either spontaneously or in response to Ti Ag immunization (Fig. 4, A and B, and data not shown; p < 0.001 comparing PBS- or NP-Ficoll-immunized Foxj1−/− to their +/+ counterparts), Foxj1−/− chimeras developed significant numbers of spontaneous germinal centers, which were dramatically increased in number upon Ti Ag immunization (p < 0.0001, comparing PBS- to NP-Ficoll-immunized Foxj1−/− mice to each other or to Foxj1+/+ mice). Altogether, these findings (Figs. 2–4) indicate that Foxj1 antagonizes B cell responses, as characterized by (auto)Ag-specific Abs, somatic hypermutation, and germinal centers.

FIGURE 4.

Germinal center formation in the absence of Foxj1 requires the RELA component of NF-κB and IL-6. Foxj1 +/+ vs −/− FLCs were immunized with vehicle (PBS) vs TNP-Ficoll, and their spleens were assessed 21 days later for the presence of germinal centers by PNA staining: A shows representative staining patterns, summarized in B as the average number of germinal centers observed per spleen section (at least 20 sections per spleen were analyzed). C, Similar studies were performed in BOCs (see Fig. 2). Similar results were obtained with NP-Ficoll (data not shown). D, Foxj1 −/− BOCs were immunized with NP-Ficoll, and concomitantly treated with antisense or mismatch control oligonucleotides against the NF-κB RELA component (p65), or a neutralizing Ab against murine IL-6 or isotype control. Spleens were assessed 21 days later for the presence of germinal centers by PNA staining. At the end of the experiment, (E) lysates of whole spleens were analyzed by Western blot for RELA and p50 expression levels.

FIGURE 4.

Germinal center formation in the absence of Foxj1 requires the RELA component of NF-κB and IL-6. Foxj1 +/+ vs −/− FLCs were immunized with vehicle (PBS) vs TNP-Ficoll, and their spleens were assessed 21 days later for the presence of germinal centers by PNA staining: A shows representative staining patterns, summarized in B as the average number of germinal centers observed per spleen section (at least 20 sections per spleen were analyzed). C, Similar studies were performed in BOCs (see Fig. 2). Similar results were obtained with NP-Ficoll (data not shown). D, Foxj1 −/− BOCs were immunized with NP-Ficoll, and concomitantly treated with antisense or mismatch control oligonucleotides against the NF-κB RELA component (p65), or a neutralizing Ab against murine IL-6 or isotype control. Spleens were assessed 21 days later for the presence of germinal centers by PNA staining. At the end of the experiment, (E) lysates of whole spleens were analyzed by Western blot for RELA and p50 expression levels.

Close modal

Because the humoral responses of Foxj1−/− FLCs might simply be a secondary effect of the hyperactivated Foxj1−/− Th cells (2), we developed BOCs by performing fetal liver chimerization in athymic nu/nu host mice, creating animals containing Foxj1 +/+ vs −/− B cells (IgHb) that develop in the absence of T cell help. B cells from these animals were then adoptively transferred into wild-type C57BL/6-IgHa mice, which were subsequently immunized with Ti Ags and/or otherwise subjected to the assays in this study. These Foxj1−/− BOCs recapitulated the Ab phenotypes of the Foxj1−/− FLCs, including accentuated anti-hapten titers and affinity, as well as germinal center formation, in response to NP-Ficoll immunization (as judged by IgHb-allotype-specific Ig assays; Figs. 2–4). In addition, all Foxj1 −/− BOCs developed Crithidia-positive anti-dsDNA Abs, in contrast to Foxj1+/+ counterparts, as judged 16 wk after reconstitution (10 of 10 vs 0 of 12 sera, as judged by IgHb-allotype-specific IgG assays; p < 0.0001). Thus, Foxj1 is an intrinsic antagonist of B cell activation, and in its absence, B cells undergo uncontrolled activation.

Foxj1 is known to repress NF-κB activation in T cells (2), and several members of the NF-κB pathway play critical roles in germinal center formation and humoral immunity (21). Consistent with this, naive B cells from Foxj1−/− BOCs animals possess enhanced spontaneous NF-κB, but not NFAT, activity compared with their Foxj1+/+ counterparts, as demonstrated by luciferase reporter assays (Fig. 5,A, p < 0.001). Because Foxj1 regulates NF-κB activity in T cells via IκBβ (2), we speculated that a similar mechanism would account for NF-κB hyperactivity in Foxj1−/− B cells. Indeed, Foxj1 −/− B cells were selectively deficient in IκBβ, as judged by both Western blot and real-time PCR, in contrast to other NF-κB family members (Fig. 5, B and C, and data not shown; p < 0.001 comparing IκBβ mRNA in Foxj1 −/− vs +/+ samples). Because Foxj1 can transactivate the IκBβ promoter (2), such findings together suggest that Foxj1 is required in B cells to regulate IκBβ expression and antagonize NF-κB activity. Indeed, in vivo treatment of Foxj1−/− BOCs with antisense oligonucleotides against the RELA (p65) NF-κB subunit inhibited spontaneous and immunization-induced germinal center formation (Fig. 4, D and E, p < 0.0001 comparing p65 antisense- to p65 missense-treated animals), suggesting that NF-κB hyperactivity, in fact, accounted for the hyperactive B cell phenotype of Foxj1 deficiency.

FIGURE 5.

Foxj1 regulates NF-κB activity via IκBβ. A, Spontaneous NF-κB and NFAT activity was assessed in resting B cells from Foxj1−/− vs Foxj1+/+ BOCs via NF-κB- and NFAT-luciferase reporters. B, Western blot analysis of various NF-κB pathway components was performed on resting B cells from Foxj1−/− vs Foxj1+/+ BOCs. C, Real-time PCR analysis of mRNA of the indicated IκB subunits was performed on resting B cells from Foxj1−/− vs Foxj1+/+ BOCs. For A and C, SDs reflect triplicate samples, representative of at least three separate experiments. For B, results reflect combined B cell extracts from at least three animals of each genotype, representative of at least three separate experiments.

FIGURE 5.

Foxj1 regulates NF-κB activity via IκBβ. A, Spontaneous NF-κB and NFAT activity was assessed in resting B cells from Foxj1−/− vs Foxj1+/+ BOCs via NF-κB- and NFAT-luciferase reporters. B, Western blot analysis of various NF-κB pathway components was performed on resting B cells from Foxj1−/− vs Foxj1+/+ BOCs. C, Real-time PCR analysis of mRNA of the indicated IκB subunits was performed on resting B cells from Foxj1−/− vs Foxj1+/+ BOCs. For A and C, SDs reflect triplicate samples, representative of at least three separate experiments. For B, results reflect combined B cell extracts from at least three animals of each genotype, representative of at least three separate experiments.

Close modal

Several NF-κB target genes could account for this phenotype, including costimulatory molecules such as members of the CD28/B7 system, or cytokines such as IL-6 (21). However, when assessed as freshly isolated naive cells or 4 h after LPS stimulation in vitro, Foxj1−/− B cells demonstrated 1.5-fold or less differences in RNA and/or protein levels of CD80, CD86, ICOS, ICOSL, CD40, CD154, TNF-α, LT-α, bcl-6, Blimp-1, IRF-4, Mitf, c-myc, Pax5, and Xbp1, at least as judged by real-time PCR and/or flow cytometry (our unpublished data). In contrast, they consistently expressed at least 3-fold more IL-6 RNA after 4 h of LPS stimulation (Fig. 6,A, p < 0.0001), and secreted 5- to 10-fold or more IL-6 in vitro (Fig. 6 B, p < 0.0001 comparing all LPS doses), compared with Foxj1+/+ B cells.

FIGURE 6.

Foxj1 regulates IL-6. A, Resting B cells from Foxj1 −/− vs +/+ BOCs were stimulated in vitro with 25 μg/ml LPS for 4 h, and RNA was analyzed for IL-6 expression by real-time PCR. B, Production of IL-6 by B cells from Foxj1 −/− vs +/+ B cell chimeras was assessed in culture supernatants by ELISA after 2 days in culture with LPS. Note that both parts of the figure are derived from a single experiment, with differences in y-axes used to emphasize the spontaneous (albeit low-level) production of IL-6 by Foxj1-deficient B cells in response to no or low-dose LPS stimulation. C, The activity of an IL-6 promoter-reporter construct was assessed in M12 B cell lymphoma cells in the presence (pcDNA-Foxj1) or absence (pcDNA) of Foxj1. D, The ability of NF-κB decoy oligonucleotides to inhibit IL-6 secretion was assessed in B cells from Foxj1 −/− vs +/+ BOCs stimulated with 2.5 μg/ml LPS for 2 days. SDs reflect triplicate samples, representative of at least three separate experiments.

FIGURE 6.

Foxj1 regulates IL-6. A, Resting B cells from Foxj1 −/− vs +/+ BOCs were stimulated in vitro with 25 μg/ml LPS for 4 h, and RNA was analyzed for IL-6 expression by real-time PCR. B, Production of IL-6 by B cells from Foxj1 −/− vs +/+ B cell chimeras was assessed in culture supernatants by ELISA after 2 days in culture with LPS. Note that both parts of the figure are derived from a single experiment, with differences in y-axes used to emphasize the spontaneous (albeit low-level) production of IL-6 by Foxj1-deficient B cells in response to no or low-dose LPS stimulation. C, The activity of an IL-6 promoter-reporter construct was assessed in M12 B cell lymphoma cells in the presence (pcDNA-Foxj1) or absence (pcDNA) of Foxj1. D, The ability of NF-κB decoy oligonucleotides to inhibit IL-6 secretion was assessed in B cells from Foxj1 −/− vs +/+ BOCs stimulated with 2.5 μg/ml LPS for 2 days. SDs reflect triplicate samples, representative of at least three separate experiments.

Close modal

Foxj1 inhibited the activity of an IL-6 promoter-luciferase reporter (Fig. 6,C, p < 0.0001 comparing pcDNA to pcDNA-Foxj1)—likely reflecting NF-κB inhibition (2). In addition, NF-κB decoy oligonucleotides, but not control mismatch oligonucleotides, inhibited the IL-6 hypersecretion of Foxj1−/− B cells, reducing the amounts to the levels secreted by Foxj1+/+ B cells (Fig. 6,D, p < 0.0001 comparing decoy- to mismatch oligonucleotide-treated Foxj1−/− samples). Finally, in vivo treatment of Foxj1−/− BOCs with neutralizing Ab against IL-6 inhibited spontaneous and immunization-induced germinal center formation (Fig. 4, D and E; p < 0.0001 comparing isotype- to anti-IL-6-treated animals). Thus, hyperactivity of B cells in Foxj1 deficiency results from overactivity of both NF-κB and IL-6.

A previous study has indicated that Foxj1 is required to prevent autoimmunity, because Foxj1-deficient FLCs develop spontaneous T cell-related end-organ inflammation, and defective expression of Foxj1 in T cells correlates with spontaneous lupus-like disease (2). The present findings extend the relevance of Foxj1 to B cell (humoral) autoimmunity, because Foxj1-deficient BOCs exhibit accentuated spontaneous autoantibody formation, anti-hapten responses in immunization and germinal center formation, and defective expression of Foxj1 in B cells correlates with spontaneous lupus-like disease (Figs. 1–4). Interestingly, the role of Foxj1 in B cells appears analogous to its role in T cells, antagonizing NF-κB and IL-6 (Figs. 5 and 6). Still, the present results suggest that Foxj1 deficiency does not break tolerance in B cells per se, in contrast to T cells (2), but rather amplifies ongoing Ag- or autoantigen-specific responses (Fig. 2), likely via amplified NF-κB and/or IL-6 activity.

It is interesting to note that, at least in some circumstances, pathogenic autoantibody production may (4, 22) or may not (23, 24, 25) require germinal centers. Because different experimental systems were used in each of these studies, these apparently disparate conclusions may reflect a differential relevance of germinal centers to different types of autoimmune disease: some lupus-like syndromes, in fact, may result in pathogenic autoantibody production without the development of germinal centers. Therefore, Foxj1 deficiency might be found in only a subset of affected individuals or mice, and may only be relevant to specific disease subtypes. Nonetheless, the present findings correlate well with the previously demonstrated importance of the NF-κB pathway in the pathogeneses of several autoimmune syndromes (26); the role of IL-6 in autoreactive germinal centers (7, 8); the spontaneous germinal center formation that occurs in humoral autoimmune diseases, likely in response to endogenous autoantigens (22, 27); as well as the proposed utility of IL-6 blockade in the treatment of humoral autoimmune diseases like lupus (28). As such, our present findings raise the intriguing possibility that functional Foxj1 deficiency, by whatever genetic mechanism, could perhaps underlie a large proportion of autoimmune syndromes. Further studies that address how Foxj1, NF-κB, and IL-6 relate to B and T cell tolerance, therefore, are likely to be particularly enlightening.

Curiously, in addition to its demonstrated role in autoreactive germinal center formation (7, 8), IL-6 has been heavily implicated in plasma cell survival (29). The ability of Foxj1-deficient B cells to generate significantly enhanced anti-hapten titers (Fig. 2 B) are potentially consistent with the enhanced development and/or accumulation of Ag-specific plasma cells. However, the increased affinity of these specificities in the absence of Foxj1, associated with the development of significant numbers of germinal centers, suggests that the role of IL-6 here, in fact, may be 2-fold, both promoting germinal center formation as well as plasma cell survival, as suggested by studies with IL-6 transgenic mice (30).

Relatively little continues to be known regarding the role of the forkhead genes in B cells. Some studies have indicated that members of the Foxo forkhead subfamily may regulate apoptotic and/or proliferative responses in B cells (31, 32), but such studies have been primarily limited to transformed cultured cell lines, and/or overexpression studies in primary cells in vitro. Therefore, the present findings supplement this growing literature by demonstrating that at least the Foxj subfamily member Foxj1 indeed modulates B cell effector function in vivo. Continued investigation into the forkhead family, including members of other Fox subfamilies, therefore, will hopefully reveal additional insights into the mechanisms of B as well as T cell immunoregulation, as well as the relationship between their target genes and autoimmunity.

We thank Markus Neurath for advice regarding NF-κB antisense oligonucleotide strategies.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported in part by the Rheumatic Diseases, Siteman Cancer, Diabetes Research and Training, and the Digestive Diseases Research Core (DK52574) Centers of the Washington University School of Medicine, as well as grants from the National Institutes of Health (AI057471 and AI061478 to S.L.P.), Arthritis Foundation, and the Lupus Research Institute. S.L.P. is supported in part by an Arthritis Investigator Award from the Arthritis Foundation.

3

Abbreviations used in this paper: FLC, fetal liver chimera; BOC, B cell-only chimera; NP, 4-hydroxy-3-nitrophenylacetyl; PNA, peanut agglutinin; RF, rheumatoid factor.

1
Shlomchik, M. J., J. E. Craft, M. J. Mamula.
2001
. From T to B and back again: positive feedback in systemic autoimmune disease.
Nat. Rev. Immunol.
1
:
147
-153.
2
Lin, L., M. Spoor, A. J. Gerth, S. L. Brody, S. L. Peng.
2004
. Modulation of Th1 activation and inflammation by the NF-κB repressor Foxj1.
Science
303
:
1017
-1020.
3
Lin, L., J. D. Hron, S. L. Peng.
2004
. Regulation of NF-κB, Th activation, and autoinflammation by the forkhead transcription factor Foxo3a.
Immunity
21
:
203
-213.
4
Hron, J. D., L. Caplan, A. J. Gerth, P. L. Schwartzberg, S. L. Peng.
2004
. SH2D1A regulates T-dependent humoral autoimmunity.
J. Exp. Med.
200
:
261
-266.
5
Berland, R., H. H. Wortis.
2002
. Origins and functions of B-1 cells with notes on the role of CD5.
Annu. Rev. Immunol.
20
:
253
-300.
6
Calame, K. L., K. I. Lin, C. Tunyaplin.
2003
. Regulatory mechanisms that determine the development and function of plasma cells.
Annu. Rev. Immunol.
21
:
205
-230.
7
Kopf, M., S. Herren, M. V. Wiles, M. B. Pepys, M. H. Kosco-Vilbois.
1998
. Interleukin 6 influences germinal center development and antibody production via a contribution of C3 complement component.
J. Exp. Med.
188
:
1895
-1906.
8
Deng, C., E. Goluszko, E. Tuzun, H. Yang, P. Christadoss.
2002
. Resistance to experimental autoimmune myasthenia gravis in IL-6-deficient mice is associated with reduced germinal center formation and C3 production.
J. Immunol.
169
:
1077
-1083.
9
Brody, S. L., X. H. Yan, M. K. Wuerffel, S. K. Song, S. D. Shapiro.
2000
. Ciliogenesis and left-right axis defects in forkhead factor HFH-4-null mice.
Am. J. Respir. Cell Mol. Biol.
23
:
45
-51.
10
Starnes, H. F., Jr, M. K. Pearce, A. Tewari, J. H. Yim, J. C. Zou, J. S. Abrams.
1990
. Anti-IL-6 monoclonal antibodies protect against lethal Escherichia coli infection and lethal tumor necrosis factor-α challenge in mice.
J. Immunol.
145
:
4185
-4191.
11
Schlaak, J. F., A. P. Barreiros, S. Pettersson, P. Schirmacher, K. H. Meyer Zum Buschenfelde, M. F. Neurath.
2001
. Antisense phosphorothioate oligonucleotides to the p65 subunit of NF-κB abrogate fulminant septic shock induced by S. typhimurium in mice.
Scand. J. Immunol.
54
:
396
-403.
12
Peng, S. L., S. J. Szabo, L. H. Glimcher.
2002
. T-bet regulates IgG class switching and autoantibody production.
Proc. Natl. Acad. Sci. USA
99
:
5545
-5550.
13
Lin, L., A. J. Gerth, S. L. Peng.
2004
. Active inhibition of plasma cell development in resting B cells by microphthalmia-associated transcription factor.
J. Exp. Med.
200
:
115
-122.
14
Khaled, A. R., E. J. Butfiloski, E. S. Sobel, J. Schiffenbauer.
1998
. Use of phosphorothioate-modified oligodeoxynucleotides to inhibit NF-κB expression and lymphocyte function.
Clin. Immunol. Immunopathol.
86
:
170
-179.
15
Overbergh, L., D. Valckx, M. Waer, C. Mathieu.
1999
. Quantification of murine cytokine mRNAs using real time quantitative reverse transcriptase PCR.
Cytokine
11
:
305
-312.
16
Laurencikiene, J., V. Deveikaite, E. Severinson.
2001
. HS1,2 enhancer regulation of germline ε and γ2b promoters in murine B lymphocytes: evidence for specific promoter-enhancer interactions.
J. Immunol.
167
:
3257
-3265.
17
Peng, S. L., M. E. Robert, A. C. Hayday, J. Craft.
1996
. A tumor-suppressor function for Fas (CD95) revealed in T cell-deficient mice.
J. Exp. Med.
184
:
1149
-1154.
18
Muramatsu, M., K. Kinoshita, S. Fagarasan, S. Yamada, Y. Shinkai, T. Honjo.
2000
. Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme.
Cell
102
:
553
-563.
19
Aarden, L. A., E. R. de Groot, T. E. Feltkamp.
1975
. Immunology of DNA. III. Crithidia luciliae, a simple substrate for the determination of anti-dsDNA with the immunofluorescence technique.
Ann. NY Acad. Sci.
254
:
505
-515.
20
Hahn, B. H..
1998
. Antibodies to DNA.
N. Engl. J. Med.
338
:
1359
-1368.
21
Li, Q., I. M. Verma.
2002
. NF-κB regulation in the immune system.
Nat. Rev. Immunol.
2
:
725
-734.
22
Grammer, A. C., R. Slota, R. Fischer, H. Gur, H. Girschick, C. Yarboro, G. G. Illei, P. E. Lipsky.
2003
. Abnormal germinal center reactions in systemic lupus erythematosus demonstrated by blockade of CD154-CD40 interactions.
J. Clin. Invest.
112
:
1506
-1520.
23
Toellner, K. M., W. E. Jenkinson, D. R. Taylor, M. Khan, D. M. Sze, D. M. Sansom, C. G. Vinuesa, I. C. MacLennan.
2002
. Low-level hypermutation in T cell-independent germinal centers compared with high mutation rates associated with T cell-dependent germinal centers.
J. Exp. Med.
195
:
383
-389.
24
Weller, S., A. Faili, C. Garcia, M. C. Braun, F. F. Le Deist, G. G. de Saint Basile, O. Hermine, A. Fischer, C. A. Reynaud, J. C. Weill.
2001
. CD40-CD40L independent Ig gene hypermutation suggests a second B cell diversification pathway in humans.
Proc. Natl. Acad. Sci. USA
98
:
1166
-1170.
25
William, J., C. Euler, S. Christensen, M. J. Shlomchik.
2002
. Evolution of autoantibody responses via somatic hypermutation outside of germinal centers.
Science
297
:
2066
-2070.
26
Peng, S. L..
2004
. Transcription factors in the pathogenesis of autoimmunity.
Clin. Immunol.
110
:
112
-123.
27
Luzina, I. G., S. P. Atamas, C. E. Storrer, L. C. daSilva, G. Kelsoe, J. C. Papadimitriou, B. S. Handwerger.
2001
. Spontaneous formation of germinal centers in autoimmune mice.
J. Leukocyte Biol.
70
:
578
-584.
28
Cross, J. T., H. P. Benton.
1999
. The roles of interleukin-6 and interleukin-10 in B cell hyperactivity in systemic lupus erythematosus.
Inflamm. Res.
48
:
255
-261.
29
Iwakoshi, N. N., A. H. Lee, L. H. Glimcher.
2003
. The X-box binding protein-1 transcription factor is required for plasma cell differentiation and the unfolded protein response.
Immunol. Rev.
194
:
29
-38.
30
Kovalchuk, A. L., J. S. Kim, S. S. Park, A. E. Coleman, J. M. Ward, H. C. Morse, III, T. Kishimoto, M. Potter, S. Janz.
2002
. IL-6 transgenic mouse model for extraosseous plasmacytoma.
Proc. Natl. Acad. Sci. USA
99
:
1509
-1514.
31
Dijkers, P. F., R. H. Medema, J. W. Lammers, L. Koenderman, P. J. Coffer.
2000
. Expression of the pro-apoptotic Bcl-2 family member Bim is regulated by the forkhead transcription factor FKHR-L1.
Curr. Biol.
10
:
1201
-1204.
32
Yusuf, I., X. Zhu, M. G. Kharas, J. Chen, D. A. Fruman.
2004
. Optimal B-cell proliferation requires phosphoinositide 3-kinase-dependent inactivation of FOXO transcription factors.
Blood
104
:
784
-787.