Allergic TH2 Response Governed by B-Cell Lymphoma 6 Function in Naturally Occurring Memory Phenotype CD4+ T Cells

Transcriptional repressor B-cell lymphoma 6 (Bcl6) appears to regulate TH2 immune responses in allergies, but its precise role is unclear. We previously reported that Bcl6 suppressed IL-4 production in naïve CD4+ T cell-derived memory TH2 cells. To investigate Bcl6 function in allergic responses in naturally occurring memory phenotype CD4+ T (MPT) cells and their derived TH2 (MPTH2) cells, Bcl6-manipulated mice, highly conserved intron enhancer (hcIE)-deficient mice, and reporter mice for conserved noncoding sequence 2 (CNS2) 3′ distal enhancer region were used to elucidate Bcl6 function in MPT cells. The molecular mechanisms of Bcl6-mediated TH2 cytokine gene regulation were elucidated using cellular and molecular approaches. Bcl6 function in MPT cells was determined using adoptive transfer to naïve mice, which were assessed for allergic airway inflammation. Bcl6 suppressed IL-4 production in MPT and MPTH2 cells by suppressing CNS2 enhancer activity. Bcl6 downregulated Il4 expression in MPTH2 cells, but not MPT cells, by suppressing hcIE activity. The inhibitory functions of Bcl6 in MPT and MPTH2 cells attenuated allergic responses. Bcl6 is a critical regulator of IL-4 production by MPT and MPTH2 cells in TH2 immune responses related to the pathogenesis of allergies.

Transcriptional repressor B-cell lymphoma 6 (Bcl6) appears to regulate TH2 immune responses in allergies, but its precise role is unclear. We previously reported that Bcl6 suppressed IL-4 production in naïve CD4 + T cell-derived memory TH2 cells. To investigate Bcl6 function in allergic responses in naturally occurring memory phenotype CD4 + T (MPT) cells and their derived TH2 (MPTH2) cells, Bcl6-manipulated mice, highly conserved intron enhancer (hcIE)-deficient mice, and reporter mice for conserved noncoding sequence 2 (CNS2) 3′ distal enhancer region were used to elucidate Bcl6 function in MPT cells. The molecular mechanisms of Bcl6-mediated TH2 cytokine gene regulation were elucidated using cellular and molecular approaches. Bcl6 function in MPT cells was determined using adoptive transfer to naïve mice, which were assessed for allergic airway inflammation. Bcl6 suppressed IL-4 production in MPT and MPTH2 cells by suppressing CNS2 enhancer activity. Bcl6 downregulated Il4 expression in MPTH2 cells, but not MPT cells, by suppressing hcIE activity. The inhibitory functions of Bcl6 in MPT and MPTH2 cells attenuated allergic responses. Bcl6 is a critical regulator of IL-4 production by MPT and MPTH2 cells in TH2 immune responses related to the pathogenesis of allergies.
In some experiments, anti-IL-4 mAbs or anti-IFN-γ mAbs were added to the TH0 condition cultures. On days 3 and 5, activated naïve T cells and MPT cells were stimulated with rIL-2 (25 U/ mL) and rIL-7 (10 U/mL) following primary stimulation. NATH2 cells were further cultured with IL-7 for 21 days to yield NATH2 cell-derived memory-like TH2 (NAM-LTH2) cells, which have a functional phenotype similar to NATH2 cell-derived memory (NAMTH2) cells in vivo (15). Some MPT cells were cultured in the presence of IL-33 (0-100 ng/mL) with or without IL-7 for the appropriate times as shown in each experiment prior to analysis of chromatin immunoprecipitation (ChIP) assays and the effect of TCR stimulation on cytokine production.
Fluorescence-activated cell sorting (Facs) analysis As previously described (15,17), T cells with or without 8 h of restimulation were treated with monensin (2 µM) for the last 3 h, followed by staining with an appropriate combination of FITCconjugated anti-KJ1-26, APC-conjugated anti-CD44, and PerCPconjugated anti-CD4 mAbs. For staining, cells were washed once with FACS buffer (PBS with 3% fetal calf serum and 0.1% sodium azide) and then permeabilized with Perm2 (BD Biosciences) for 10 min at room temperature, followed by two washes in FACS buffer. Finally, cells were stained with an appropriate combination of anti-IFN-γ-APC and anti-IL-4-PE for 30 min at room temperature, washed, and resuspended in FACS buffer for analysis.

cytokine concentrations
IL-4, IL-5, and IL-13 levels in the culture supernatants of cells that were stimulated for 48 h in bronchoalveolar lavage fluid (BALF) were determined using ELISA kits (R&D Systems, Minneapolis, MN, USA). IgE anti-OVA Abs were detected using a mouse anti-OVA IgE Antibody Assay Kit (Chondrex, Redmond, WA, USA).

chromatin immunoprecipitation
The ChIP assay was performed as previously described (14,15). Protein and chromatin in TH cells were cross-linked by adding formaldehyde solution (Thermo Fisher Scientific, Waltham, MA, USA), after which the cells were lysed in SDS lysis buffer. Subsequently, precleared, sonicated chromatin and protein G agarose (Millipore) were incubated with specific Abs for the protein of interest or control IgG (rabbit). Some of the untreated chromatin was used as an input sample. qPCR was used to quantify the DNA region in the immune-precipitated chromatin and the input DNA. Relative ChIP DNA quantification was performed using the comparative Ct method. The Ct value of ChIP DNA was normalized to that of the input DNA using the Specifically, a fragment of d2EGFP cDNA was PCR amplified using an XhoI-anchored sense primer (underlined) (5′-CCG CTCGAGTCTAGAGGATCCACCGGTCGC-3′) immediately upstream of the XbaI site (+258) and an antisense primer with a SalI-anchored antisense primer (underlined) (5′-ACGC GTCGACTCTAGAGTCGCGGCCGCATC-3′) immediately down stream of the XbaI site (+1147) of pd2EGFP. The XhoI/SalI fragment of d2EGFP was subcloned into a T Easy vector (d2EGFP-T vector). The EcoRI-digested d2EGFP fragment was blunted and subcloned into a blunted NotI/NcoI-restricted pMX vector (pMX-d2EGFP). A genomic fragment of the Il4 promoter region was PCR amplified using the EcoRI-anchored sense primer (underlined) (5′-GAATTCCTCCACACTGATGCTGTAGTGC-3′) and XhoI-anchored antisense primer (underlined) (5′-CTCGAGG CTAACAATGCTGGC-3′). The subcloned Il4 promoter fragment was then digested with EcoRI and XhoI and subcloned into the restricted site of pMX-d2EGFP (pMX-Il4p-d2EGFP). An EcoRI and SalI fragment of pMX-Il4p-d2EGFP was then subcloned into the EcoRI/SalI-restricted pBABE delta Bll(−) to generate pBABE delta Bll(−)-Il4p-d2EGFP. The vector pBABE delta BII(−) is based on pBABEpuro, with further modifications to completely destroy the endogenous transcriptional regulatory sequences within the retroviral long terminal repeat (LTR). R and U5 are the intact R and U5 regions of MMLV, respectively, en. del. U3 is the SIN U3 found in proviral LTRs after integration of the virus into the host genome, and partial LTR denotes a transcription-competent part of the LTR that is used to drive transcription of the genomic viral RNA in the packaging cells. The hcIE genomic fragments were PCR amplified with the XhoI-anchored sense primer (underlined) (5′-CCGCTCGAGCCTTTCTGCCTGCTGCTCTG-3′) and SalI-anchored antisense primer (underlined) (5′-ACGCGTCGAC GAAAAGCAGGCAGTCTGGAG-3′).

retrovirus infection
Platinum-E packaging cells (32) were transfected with 1-1.5 µg of DNA of a retrovirus construct mixed with 6 µL of Fugene (Boehringer Mannheim). Virus supernatant was concentrated by centrifugation (8,000 × g, 16 h) and added to TH2 cell-inducing cultures on day 2. Intracellular cytokine staining or mean fluorescence intensity (MFI) analysis was performed on day 7 as described previously. Infected cells were subjected to FACS analysis of the intracellular fluorescence of d2EGFP 8 h after restimulation with plate-bound anti-CD3 mAbs.

OVA Challenge and Bronchoalveolar lavage (BAL)
TH2 cells (1.5 × 10 7 or 3 × 10 7 ) were injected intravenously into naïve wild-type (WT) BALB/c mice (day 0), followed by intratracheal challenge with 1% OVA solution (50 µL) twice (days 2 and 3), BAL three times (days 2, 7, and 12), and serum collection. On days 4 and 5, the transferred TH2 cells isolated from whole lungs and BALF were collected from the mice by instilling the lungs with 0.5 mL of PBS six times. Sera on day 14 were analyzed for OVA antigen-specific IgE Abs. In another experiment, a mixture of Bcl6-WT, Bcl6-TG, or Bcl6-KO KJ1-26 + MPT (2 × 10 6 cells) and Bcl6-WT KJ1-26 − naïve CD4 + T (5 × 10 6 cells) cells were intravenously transferred into BALB/c nu/nu mice (day 0). Subsequently, mice were sensitized via i.p. injection of 10 µg of OVA plus 1 mg of alum twice (days 1 and 6), followed by intratracheal challenge with OVA twice (days 16 and 17). BAL and pathology examination were performed (day 18), and transferred KJ1-26 − cells were isolated from spleens (day 16). The isolated cells were restimulated with plate-bound anti-CD3 mAbs to analyze cytokine production. The BALF supernatant was stored at −80°C. Each cell pellet was resuspended in PBS for counting and subjected to cytospin. Preparations on slides were stained with Diff-Quick (Sysmex International Reagents, Kobe, Japan) for the differential analysis of cell counts. After BAL, lungs were treated with collagenase II (1 mg/mL) for 30 min at 37°C, and leukocytes were isolated on a Percoll gradient.

Histologic Examination
After BAL, the left lobes of lungs were extracted, washed with PBS, and fixed in 4% formaldehyde in sodium phosphate buffer for more than 2 days at room temperature. After fixation, lungs were embedded in paraffin and stained with hematoxylin and eosin. Images of each tissue section were captured using a Zeiss Axioscope 2 microscope equipped with a video camera (AxioCam ERc5s, Carl Zeiss, Jena, Germany) and processed using Axiovision V.4 software (Carl Zeiss).  resUlTs

Bcl6 represses il-4 Production by MPT cells
Splenic CNS2-active MPT cells were detected as a GFP + subpopulation in reporter gene TG mice (CNS2-GFP-TG) on each Bcl6 genotype background (28) (Figure 1A). Unfortunately, offspring from CNS2-GFP-TG mice on the Bcl6-KO background could not be obtained ( Figure 1B). Although the percentages of GFP + cells were similar between Bcl6-TG and Bcl6-WT mice (Figure 1C), the IL-4 + MPT cell frequency ( Figure 1D) and MFI of CNS2-GFP in MPT cells ( Figure 1E) were inversely correlated with Bcl6 levels. GFP + MPT cells displayed significant Il4 expression, which was lower in Bcl6-TG cells than in WT cells ( Figure 1F). Il4 expression was extremely low in the GFP − population regardless of Bcl6 levels. The absolute numbers and percentages of IL-4 + MPT cells were also negatively associated with Bcl6 levels (Figure 1G), whereas the absolute numbers of GFP + MPT cells ( Figure 1H) and MPT cells ( Figure 1I) among all CD4 + T cells  were positively correlated with Bcl6 levels. Therefore, Bcl6 may be involved in Il4 downregulation in MPT cells and MPT cell survival and maintenance. Because it has been reported that the TH2 and TH1 conditions are promotive and inhibitory, respectively, on the maintenance of Bcl6-WT CNS2-GFP + MPT cells (28), we analyzed the effect of Bcl6 on the maintenance of CNS2-GFP + MPT cells in each culture setting ( Figure S1 in Supplementary Material). Regarding the maintenance of GFP + cells, a promoting effect of the TH2 condition and inhibitory effect of TH1 condition were observed regardless of the Bcl6 genotype, whereas Bcl6 appears to function as a suppressor for CNS2 activity.

Bcl6 represses Il4 expression in T h 2-Primed MPT cells
To investigate the function of Bcl6 in the differentiation of MPT cells into TH cell lineages following TCR stimulation, MPT cells expressing a clonotypic TCR (KJ1-26 + ) from the spleens of Bcl6-TG, Bcl6-KO, and Bcl6-WT DO11.10 TG mice were cultured under conditions driving them toward the TH0, TH1, or TH2 phenotype, followed by intracellular IL-4 analysis after restimulation with anti-CD3 mAbs (Figures 2A,B). Under the TH0 condition, Bcl6 decreased IL-4 production in a concentration-dependent manner, and high Bcl6 expression facilitated IFN-γ induction during TH1 phenotype differentiation. Under the TH1 condition, Bcl6 deficiency in MPT cells preserved IL-4 production, although its level was lower than that under the TH0 condition. Under the TH2 condition, Bcl6 negatively regulated MPT cell-derived TH2 (MPTH2) cell differentiation but not NATH2 differentiation, as previously reported (15) (Figures 2A,B), although Bcl6 could suppress the initial IL-4 production by naïve CD4 T cells under the TH0 condition even when blocking the effects of IFN-γ ( Figure  S2 in Supplementary Material). Because Bcl6 appears to promote IFN-γ production, which may indirectly affect IL-4 induction, we analyzed a mixed culture of Bcl6-WT MPT cells with either Bcl6-TG or Bcl6-KO cells under the TH0 condition. Bcl6-KO MPT cells caused WT cells to skew clearly toward the TH2 phenotype with reduced TH1 skewing, whereas Bcl6-TG cells promoted slight TH skewing ( Figure S3 in Supplementary Material), indicating that increased IL-4 production in Bck6-KO MPT cells autoaccelerates TH2 cell differentiation by preventing TH1 cell differentiation.
Thus, Bcl6 appears to promote IFN-γ production by inhibiting IL-4 production rather than inhibiting IL-4 production via the promotion of IFN-γ production.
To confirm the suppressive effects of Bcl6 on TH2 cytokine genes in MPTH2 cells, Bcl6-WT MPTH2 cells were treated with a Bcl6 inhibitor (15), followed by restimulation with anti-CD3 mAbs. Bcl6 inhibition augmented IL-4 production but not IFN-γ production (Figures 2C,D). TH2 cytokine gene expression was upregulated by the inhibitor without changes in Gata3 expression (Figure 2E), indicating that Bcl6 suppresses Il4 expression in developing and differentiated MPTH2 cells.

Bcl6 negatively regulates the histone Modification of T h 2 cytokine loci in MPT h 2 cells
Because unprimed MPT cells express higher Bcl6 levels than naïve CD4 + T cells (29), Bcl6 expression levels in the MPT and MPTH2 cells of CNS2-GFP-TG mice with Bcl6-WT background were analyzed at rest (Figure 3A). Bcl6 expression in GFP + MPT and GFP − MPTH2 cells was increased by sevenfold and threefold, respectively, compared with that in GFP + MPTH2 cells. NATH2 cells had markedly lower Bcl6 expression than GFP + MPTH2 cells. Bcl6 expression in GFP + MPT cells was slightly increased compared with that in GFP − MPT cells. Consistent with the mRNA levels, Bcl6 protein expression was lower in GFP + MPTH2 cells than in GFP − MPTH2 cells (Figure 3B). Bcl6 protein levels in MPT cells from Bcl6-WT mice were higher than those in MPTH2 cells, whereas the protein levels in GFP + MPT cells were slightly higher than those in GFP − MPT cells. To address Bcl6 function, TH2 cytokine production by MPTH2 cells from Bcl6-WT-CNS2-GFP-TG mice was analyzed. TH2 cytokine protein ( Figure 3C) and transcript levels ( Figure 3D) were significantly greater in the GFP + population than in the GFP − population following stimulation, implying that Bcl6 function may be inhibited depending on its quantity and/or quality and that this inhibition may be involved in TH2 cytokine production in MPTH2 cells. Conversely, IFN-γ protein ( Figure 3C) and transcript levels ( Figure 3D) were undetectable and minimal, respectively, in both the GFP + and GFP − populations. Because Bcl6 binds to BSs (except BSIL13) (Figure 3E, top) and thereby reduces TH2 cytokine production in NAMTH2 cells (15), Bcl6 binding to each site in MPTH2 cells was analyzed by ChIP (Figure 3E, bottom). In Bcl6-WT and Bcl6-TG MPTH2 cells, Bcl6 binding was observed at all BS sites excluding BSIL13, BS1, and BS2. GFP + cells had significantly less Bcl6 binding than GFP − cells among Bcl6-WT and Bcl6-TG MPTH2 cells, whereas Bcl6 binding was augmented in Bcl6-TG MPTH2 cells. Thus, Bcl6 repressor functions may be regulated qualitatively (e.g., its binding ability) and quantitatively by its binding to TH2 cytokine gene foci. To investigate the effects of Bcl6 of STATs on histone modification in these foci, ChIP was performed for STAT5 and STAT6 binding to BSs and for histone H3 acetylation in MPTH2 cells (Figure 3F). STAT6 binding was marginal, whereas STAT5 binding was significantly decreased depending on Bcl6 levels, as indicated by attenuated histone acetylation.

Bcl6 represses Il4 expression by Binding to cns2 in MPT h 2 cells
B-cell lymphoma 6, but not STAT proteins, binds to BS7 (Figures 3D,E) in the major Il4 regulatory region. Although no significant Bcl6-mediated interaction was observed between BS7 in CNS2 regarding Il4 regulation in NAMTH2 cells (15), CNS2 enhancer activity may be suppressed by Bcl6 through BS7 binding. FACS analysis indicated that GFP MFI levels related to CNS2 activation in MPT cells, including at two mutated sites, namely, BS7 (1) and (2) (Figure 4A), were inversely correlated with Bcl6 levels ( Figure 1E). Therefore, the role of Bcl6 in enhancing activity in MPTH2 cells from Bcl6-WT or Bcl6-KO mice was investigated using a retrovirus reporter gene transfer vector ( Figure 4B) designed to assess Il4 promoter (p) activity by measuring the MFI for d2EGFP, a reporter protein, following stimulation with anti-CD3 and anti-CD28 mAbs (Figures 4C,D). Additionally, a CNS2 sequence containing WT or mutated BS7, that is, Mu-BS7 (1)-CNS2 and Mu-BS7 (2)-CNS2, were inserted downstream of d2EGFP ( Figure 4B). The MFI for d2EGFP with CNS2-WT elements in Bcl6-KO cells was higher than that in Bcl6-WT cells. The MFI was augmented by mutations in both BS7 (1) and (2) in Bcl6-WT cells, whereas that of Bcl6-KO cells was not significantly changed (Figures 4C,D). Thus, Bcl6 mediated CNS2 suppression in MPTH2 cells and presumably in unprimed MPT cells.

Bcl6 represses Il4 expression by Binding to hcie in MPT h 2 cells
Another Il4 regulatory region, HS2 (1.2 kbp) located in intron 2, is a critical regulatory region for GATA3 binding-mediated Il4 expression in NATH2 cells in HS2-KO mice (33) (Figure 5A   in MPTH2 cells was not significantly affected by Bcl6 levels (Figure 5A, bottom). We investigated the enhancer activity using a reporter construct, uncovering that Bcl6 inhibited hcIE function in MPTH2 cells ( Figures S4A-C

in Supplementary Material).
Similarly, ChIP demonstrated that GATA3 binding to G3 was increased in CNS2-active GFP + MPTH2 cells compared with that in unprimed GFP + MPT cells and was significantly attenuated in Bcl6-TG background cells ( Figure 5B). Thus, Bcl6 repressed Il4 expression by downregulating GATA3-mediated hcIE activity in MPTH2 but not MPT cells.  To further examine the role of hcIE in TH2 cytokine production, we generated hcIE-KO mice and observed markedly diminished IL-4 production in hcIE-KO NATH2 and NAMTH2 cells (15). Intracellular cytokine analysis revealed a similar frequency of IL-4 + populations in unprimed MPT cells in WT and hcIE-KO background mice, whereas IL-4 + MPTH2 cell development was impaired without changes in Gata3 expression following hcIE deletion (Figures 5C,D). Bcl6 binding was augmented at BS4, BS5, and BS6 in intron 2 but not at CNS2 (BS7) in hcIE-KO MPTH2 cells compared with that in hcIE-WT background cells (Figure 5E), indicating that hcIE activity dampens Bcl6-mediated suppressor activity for intron 2 except at the CNS2 region.

Bcl6 suppresses initial il-4 Production in MPT cells and T h 2 cell Differentiation
Because IL-4 production by MPT cells plays an important role in NATH2 cell differentiation (28), to address the effects of Bcl6 on MPT cell function, Bcl6-WT-naïve KJ1-26 + CD4 + T cells were cocultured with KJ1-26 − MPT cells from Bcl6-TG, Bcl6-KO, or Bcl6-WT mice in the presence of soluble anti-CD3 and CD28 mAbs and irradiated CD11c + DCs as the TH0 condition. Differentiation of IL-4-producing KJ1-26 + NATH2 cells varied inversely with Bcl6 levels in KJ1-26 − MPT cells, whereas IFN-γ-producing NATH1 cells differentiated in the opposite direction (Figures 6A,B). Because MP cell-derived IFN-γ may affect NATH2 cell differentiation, we analyzed the TH2 skewing of naïve CD4 + T cells cocultured with MPT cells by excluding the effect of endogenous IFN-γ. Although TH2 skewing became prominent in the coculture in the presence of anti-IFN-γ Abs regardless of the Bcl6 genotype, the skewing was still suppressed in the presence of Bcl6-TG MPT cells. Therefore, Bcl6 plays an important role in suppressing MPT cell function to skew naïve CD4 + T cells toward the TH2 phenotype (Figures 6A,B). Furthermore, regardless of the Bcl6 genotype, intrinsic IL-4 in MPT cells was involved in preserving the TH2 cell phenotype ( Figures S5A,B in the Supplementary Material).
As CNS2-active MPT cells are essential for inducing TH2 responses following immunization in an allergic murine model (28), we examined Bcl6 function in the MPT cell-induced response during the development of allergic immunity in BALB/c nu/nu mice undergoing adoptive transfer of Bcl6-WT-naïve CD4 + T cells (KJ1-26 + ) and MPT cells (KJ1-26 − ) from each respective Bcl6 genotype. Following OVA challenge in the mice, the numbers of all inflammatory cells, neutrophils, eosinophils (left), and KJ1-26 + T cells (right) in whole lung tissues were significantly increased, being inversely correlated with Bcl6 levels in the transferred MPT cells (Figures 7A,B). In BALF from the recipients, the TH2 cytokine concentrations of IL-4, IL-5, and IL-13, but not IFN-γ, were decreased after the last OVA challenge, with this effect being dependent on Bcl6 levels in the transferred MPT cells (Figure 7C). In KJ1-26 + T cells (naïvederived TH cells) from the spleens of recipients after the last OVA challenge, TH2 cytokine mRNA expression (Il4, Il5, and Il13) was decreased depending on Bcl6 levels in the transferred MPT cells (Figure 7D). OVA-specific IgE levels in the sera were increased, in accordance with increased cytokine production after the last challenge ( Figure 7E). This finding indicates that Bcl6 suppressed the development of allergic inflammation by reducing MPT cell function to facilitate NATH2 cell differentiation.

cells and naM-lT h 2 cells on allergic responses
IL-4 levels were affected by Bcl6 in NAMTH2 cells, as previously reported (15). We focused on the functional difference in the spatiotemporal dynamics between MPTH2 and NAMTH2 cells.
In the current study, NAM-LTH2 cells were analyzed as memory cells derived from naïve CD4 + T cells. In the resting phase, MPTH2 cells constitutively express Il4, the expression of which is reduced in a Bcl6-dependent manner. Following 1 h of restimulation, Il4 expression in MPTH2 cells was increased to similar levels in each Bcl6 genotype, and the expression occurred earlier than that in Bcl6-WT-NAM-LTH2 cells. Il4 expression levels were decreased in most MPTH2 cells, but not Bcl6-KO cells, in a Bcl6-dependent manner at 8 h after restimulation ( Figure 8A).
In NAM-LTH2 cells, Il4 expression levels were low in the resting phase and increased after restimulation. The expression levels in Bcl6-WT-NAM-LTH2 cells were high, similar to those in Bcl6-KO MPTH2 cells at 8 h after restimulation ( Figure 8A). The protein levels of IL-4 and IL-5, but not of IL-13, were consistent with the Il4 expression pattern in each TH2 cell type ( Figure S6 in Supplementary Material). After adoptive transfer of each cell type (MPTH2 cells or NAM-LTH2 cells) with a DO11.10 genetic background into WT BALB/c nu/nu mice, cell migration into lung tissues following OVA antigen challenge was determined and presented as percentages ( Figure 8B) and absolute cell numbers (Figure 8C). Among Bcl6-WT cells, MPTH2 cells had greater migratory capability compared with NAM-LTH2 cells at 24 h. The migration of MPTH2 cells decreased sequentially, whereas that of NAM-LTH2 cells increased at 48 h. The migration of Bcl6-KO MPTH2 cells was further augmented compared with that of Bcl6-WT cells. Next, we assessed the role of Bcl6 in interactions between MPTH2 and Bcl6-WT-NAM-LTH2 cells during allergic responses. WT BALB/c mice were adoptively transferred with combinations of each type of KJ1-26 + TH2 cells and sequentially challenged with OVA (Figures 8D,E). When Bcl6-WT-NAM-LTH2 or Bcl6-WT MPTH2 cells were transferred, TH2 cytokine levels (IL-4, IL-5, and IL-13) in the BALF were similar among recipients, whereas Bcl6-KO MPTH2 cells induced a fourfold to sevenfold increase in TH2 cytokine levels. Combined transfer of Bcl6-WT-NAM-LTH2 and Bcl6-WT MPTH2 cells resulted in synergistic cytokine production, which was further augmented when Bcl6-KO MPTH2 cells were transferred instead of Bcl6-WT MPTH2 cells (Figure 8D). The numbers of inflammatory cells, including eosinophils and lymphocytes, in the BALF ( Figure 8E) were increased, in accordance with the increased production of cytokines, indicating that Bcl6 plays a critical role in regulating the functions of MPTH2 cells, which precede NAMTH2 cells in the development of local allergic pathology.

il-33 reinforces il-4 Production by MPT cells
Because we previously reported the effects of IL-33 on Bcl6mediated histone modification in memory TH2 cells to augment IL-4 production (15), we focused in this study on the effect of IL-33 on MPT cells. FACS analysis demonstrated no significant difference in the cell-surface expression of ST2, an IL-33R subunit on MPT cells, between Bcl6-TG and Bcl6-WT mice (Figures 9A,B). ST2 was preferentially expressed on GFP + MPT cells rather than GFP − cells. When MPT cells were cultured in the presence of IL-7 for 6 days followed by IL-33 administration (Figure 9C, top), the frequency ( Figure 9C) and absolute number (Figure 9D) of IL-4 + MPT cells increased in a concentration-dependent manner at 8 h following the last IL-33 dose. The effect of IL-33 on IL-4 + MPT cells was significantly reduced in Bcl6-TG cells compared with that in WT cells (Figures 9C,D). Consistent with the priming effect of IL-33, we observed elevated levels of histone acetylation at BS sites in the Il4 locus with increased STAT5 histone association and decreased Bcl6 histone association. These effects of IL-33 on histone modification were attenuated in Bcl6-TG cells (Figure 9E).

DiscUssiOn
The function of Bcl6 to regulated TH2 cytokine production is unclear. We found that Bcl6 negatively regulated IL-4 gene expression in MPT cells and their derived MPTH2 cells. Bcl6 inhibition significantly augmented IL-4 production by WT MPTH2 cells. Furthermore, IL-4 expression was reduced in T cell-specific Bcl6-TG MPT and Bcl6-TG MPTH2 cells, indicating a suppressive function of T cell-intrinsic Bcl6. CNS2 contains multiple putative binding sites for RBP-J, a critical modulator of notch signaling (34). CNS2 is regulated by notch signals to control initial IL-4 expression in MPT cells (28). We demonstrated that Bcl6 binds to CNS2, leading to suppression of its enhancer activity in MPTH2 cells. Bcl6 antagonizes notch-dependent transcription (35,36). However, Rbpj deletion does not alter epigenetic markers on the  CNS2 site in TFH cells (29). Thus, to elucidate the positive regulatory mechanism of the activation of CNS2, a target of Bcl6 in MPT cells, further analysis is required. GATA3 binding in the HS2 enhancer region is critical for NATH2 (15,35) and NAMTH2 cells (15). However, extremely low GATA3 expression might not be associated with IL-4 production in MPT cells. We demonstrated that GATA3-mediated hcIE activation is not essential for IL-4 production by MPT cells (Figures 5B-D). However, MPTH2 cell differentiation requires hcIE enhancer activity, which induces permissive histone modification of the Il4 locus by cooperating with STAT5 and GATA3 (37). Bcl6 directly bound to and interfered with hcIE function in MPTH2 cells. Accordingly, we suggest that diverse Bcl6 functions regulate IL-4 production in MPTH2 and MPT cells. The locus control region (LCR) at the Rad50 gene is also extremely important for TH2 cytokine expression. This region is considered to be involved in coordinating TH2 cytokine genes including IL-4. We previously reported the GATA3-binding site and Bcl6/STAT-binding sites in conserved regions (TH2LCR) in the Rad50 gene in another study (15). We also reported that Bcl6 binding in the LCR is augmented by disruption of hcIE in Il4, indicating that Bcl6-mediated TH2LCR organizes TH2 cytokine gene including IL-4. Therefore, TH2LCR may be implicated in Il4 regulation in CNS2-active MPT cells. To elucidate the role of TH2LCR, further studies using region-deficient mice are required.
B-cell lymphoma 6 has various regulatory functions associated with cell viability and cytokine production, although the detailed molecular mechanisms have not been clarified. We observed that CNS2-active MPT cells contained high Bcl6 levels that declined following augmented IL-4 production under TH2 priming conditions. Intriguingly, in Bcl6-WT MPTH2 cells, the CNS2-active population exhibited markedly lower Bcl6 levels and higher Il4 levels than the CNS2-inactive population. Greater Bcl6 mRNA levels in CNS2-active MPT cells than in the CNS2-inactive population in WT mice have been reported (29), whereas we observed slight differences in expression between these two populations. However, Bcl6 protein levels in CNS2-active Bcl6-WT MPT cells were inversely decreased relative to those in the CNS2-inactive MPT cells. Therefore, when pleiotropic Bcl6 effects are required in the same cellular environment, its function may be quantitatively controlled at transcriptional, translational, or post-transcriptional levels.
We previously demonstrated that TH2 cytokine genes are negatively regulated by Bcl6 through chromatin remodeling and that interactions between Bcl6 and STAT5 are physiologically implicated in histone modulation and consequently cytokine production in NAMTH2 cells rather than NATH2 cell differentiation (15). In a previous report, we advocated that STAT5 and GATA3 cooperate in permissive histone modification of the Il4 locus by binding to hcIE and that STAT5-and GATA3-mediated epigenetic activity of hcIE may be controlled by directly and/or indirectly preventing the Bcl6-mediated silencing. In addition, Bcl6 binding to BS4, BS5, and BS6 in the Il4 locus was augmented upon hcIE disruption in differentiating TH2 cells. Therefore, even in the presence of high levels of Bcl6, Bcl6-TG naïve CD4 + T cells could differentiate into TH2 cells under the TH2 full commitment condition. Conversely, when naïve Bcl6-TG, Bcl6-WT, and Bcl6-KO CD4 + T cells are stimulated under the TH0 condition, IL-4 production by restimulated CD4 + T cells was reduced in a Bcl6 level-dependent manner. Therefore, we propose that the repressor activity of Bcl6 in the Il4 locus including hcIE and CNS2 can be determined in functional balance with transcriptional activators, such as GATA3, STATs, and RBP-J, in both MPTH2 and NATH2 cells. Accordingly, both quantitative and qualitative Bcl6 functional modifications, such as reduced binding activity (15), may be implicated in the gene regulation of Il4. Notably, we observed that Bcl6 binding to the Il4 locus is higher in CNS2-GFP − Bcl6-TG MPTH2 cells than in GFP + Bcl6-TG cells. Because enhancers can generally regulate transcription by interacting with enhancers or promoters via chromatin looping mechanisms, we propose that CNS2 may also stimulate Il4 transcription via physical interactions with hcIE, which may influence and organize Bcl6/STAT binding in hcIE. Therefore, Bcl6 binding to the Il4 locus may exceed STAT5 binding via Bcl6-mediated inhibition of CNS2 activity.
In earlier reports, we and other groups uncovered that Bcl6 has no significant intrinsic function in the differentiation of naïve CD4 + T cells into TH1/TH2 cells in full commitment experiments in vitro. In later studies focusing on TFH cells, Bcl6 suppressed effector T cells, including TH1, TH2, and TH17 cells, resulting in the induction of TFH cell differentiation. The current study indicated that Bcl6 promotes IFN-γ production via by inhibiting IL-4 production in activated naïve CD4 + T cells and MPT cells in some experimental settings, rather than inhibiting IL-4 production by promoting IFN-γ production.
Contrarily, we previously reported that Bcl6 plays an important anti-apoptotic role in effector-derived memory precursor CD4 + T cells, suggesting that Bcl6 is involved in long-term memory T cell survival (17,30,38). We observed that the numbers of splenic MPT cells and, intriguingly, CNS2-active GFP + MPT cells were positively associated with intrinsic Bcl6 levels, whereas the MFI of GFP was reduced in Bcl6-TG cells. Recently, CNS2-active GFP + CD4 + T cells in secondary lymphoid tissues were found to have a high Bcl6 expression phenotype, similar to TFH cells (29). Bcl6 is a master regulatory factor for TFH cell differentiation. However, a substantial Bcl6-KO MPT cell population exists, and we suggested that CNS2-active MPT cells are not necessary as part of the TFH cell lineage. Although the molecular mechanism is unclear, Bcl6 may be implicated in, but not essential for, the development and/or maintenance of MPT and MPTH2 cells.
NAMTH2 cells have an important role in chronic allergic responses (15), although the relationship between NAMTH2 and MPTH2 cells is unclear. We observed that TH2 cytokine production peaked and declined earlier in Bcl6-WT-MPTH2 cells than in WT-NAM-LTH2 cells. Moreover, the migratory function of MPTH2 cells was superior to that of NAM-LTH2 cells, albeit due to an unknown mechanism. Because CNS2 and Il4 are constitutively activated in MPTH2 cells but not in NAMTH2 cells (15), MPTH2 cells might influence NAMTH2 cell function in chronic allergy. Accordingly, MPTH2 cells organize TH2 immune responses directly and/or indirectly by regulating NAMTH2 cell function, resulting in allergy enhancement. IL-4 production by CNS2-active MPT cells induced TH2 responses by inducing the differentiation of NATH2 cells from naïve CD4 + T cells and their self-differentiation into MPTH2 cells following immunization (28). We confirmed initial IL-4 production from MPT cells in this study. Because CNS2-active MPT cells do not belong to the TFH cell lineage derived from naïve CD4 + T cells (29) but they rather develop from selected thymocytes among those expressing other MHC class II markers (39), IL-4 + MPT cells might develop independently of naïve CD4 + T cells during thymic differentiation. In that case, sequentially differentiated MPTH2 cells as well as MPT cells contribute to the early pathology of some allergies.
When considering the nature of Bcl6 in MPT and MPTH2 cells in pathologic conditions, we should determine whether Bcl6 expression can be modified without artificial gene manipulation at both protein and RNA levels. Recently, we reported that a TH2-promoting factor, namely, IL-33-mediated breakdown of Bcl6 in NAMTH2 cells, is likely involved in allergies (15) given the effect of IL-33 on both MPT and NAMTH2 cells. Therefore, the IL-33/Bcl6 axis might participate in allergy pathology via the regulation of Il4 in MPT cells to promote disease development in MPTH2 and NAMTH2 cells, contributing to the maintenance and exacerbation of disease pathology.
In summary, the current study provides evidence for a novel role of Bcl6 in the functional regulation of MPT and MPTH2 cells, implying interplay between Bcl6 and transcriptional activators to promote the production of relevant TH2 cytokines, particularly IL-4. Thus, TH2 cell-promoting factors that suppress Bcl6 function may represent crucial therapeutic targets for TH2 cell-mediated diseases.

eThics sTaTeMenT
This study was carried out in accordance with the recommendations of the Chiba University Resolution on Use of Animals in Research. The protocol was approved by the Institutional Animal Care and Use Committee at Chiba University School of Medicine. The mice were maintained under specific pathogen-free conditions in the animal center of Chiba University Graduate School of Medicine.
aUThOr cOnTriBUTiOns MA and TO jointly designed the experiments and directed the study and wrote the manuscript. MA, TO, YK, JI, TT, NT,