Menin Controls the Memory Th2 Cell Function by Maintaining the Epigenetic Integrity of Th2 Cells

Antigen recognition through TCR induces naive CD4 T cells to differentiate into several distinct Th cell subsets including Th1, Th2, and Th17 cells (1). IFN-γ–producing Th1 cells are essential for protection against intracellular pathogens; IL-4–, IL-5–, and IL-13–producing Th2 cells are required for immunity to parasite infection; and IL-17–producing Th17 cells are crucial for antifungal immunity. These Th cells are also involved in the pathogenesis of the inflammatory diseases (2). For example, Th1 cells associate with tissue-specific autoimmune diseases, Th2 cells are responsible for allergic diseases, and Th17 cells are involved in the bowel and neuroinflammatory diseases (3). Several lineage-specifying transcription factors that are required for Th cell differentiation have been identified (4). T-box transcription factor TBX21 encoded by the Tbx21 gene, GATA binding protein 3 (GATA3), and RAR-related orphan receptor γ encoded by the Rorc gene have been proposed as master transcription factors for Th1, Th2, and Th17 cell differentiation, respectively. In some cases, the continual expression of the transcription factors that direct differentiation are required to maintain cellular identity (5).

Epigenetic histone modifications play crucial roles in the induction of the lineage-specific transcription factor and cytokine gene expression. Transcribed genes are often associated with increased levels of trimethylation of lysine 4 on histone H3 (H3K4me3), which is catalyzed by the Trithorax group (TrxG) proteins originally discovered in Drosophila melanogaster (6). In mammals, six H3K4 methylases have been identified (6). Menin is a highly specific partner for mixed-lineage leukemia MLL1/2-containing H3K4 methyltransferase complex (7). Human Menin protein is encoded by the MEN1 gene, and the mutation of this gene causes multiple endocrine neoplasia type 1 (7). In addition, roles of Menin in Th cells also have been reported (8). For instance, the binding of the Menin/TrxG complex is required for the maintenance of Gata3 expression and Th2 cytokine production in established Th2 cells (9), and the same mechanism was also recently found to function in human Th2 cells (10). Menin is also essential for induction of Il17a expression in differentiating Th17 cells and is required for the maintenance of Rorc expression in differentiated Th17 cells (11). Menin-dependent Bach2 expression has been shown to be important for preventing cellular senescence in activated CD4 T cells (12). Regarding memory T cells, H3K4 methylation is shown to be responsible for their rapid response to the Ags, suggesting that TrxG proteins represent an essential mechanism of transcriptional maintenance in memory T cell response (13–15). Few studies address this point, however, so how TrxG proteins regulate the memory T cell response remains unclear.

In this study, we found that Menin was dispensable for the acquisition of Th2 cell identity, but it was required for the maintenance of Th2 cell identity during multiple rounds of Ag stimulation. Consequently, Th2 cell–induced airway inflammation was attenuated in the absence of Menin. The binding of Menin to the Gata3 gene locus was crucial for the maintenance of Th2 cell identity. We also found that Menin maintained the memory Th2 (mTh2) cell function during the long-term resting phase, and that Menin deficiency resulted in the attenuation of the mTh2 cell–mediated type 2 immune response via Ag challenge in vivo. Thus, this study reveals a mechanism by which Menin regulates the maintenance of Th2 cell identity in vitro and in vivo.

C57BL/6, BALB/c, and BALB/c-nu/nu mice were purchased from CLEA (Tokyo, Japan). Men1^fl/fl mice (16) were purchased from The Jackson Laboratory and backcrossed at Chiba University to a C57BL/6 or BALB/c background >10 times. CD4-Cre transgenic (Tg) mice were purchased from Taconic Farms. CD4-Cre–driven conditional knockout (KO) mice for the Men1 gene were used as Menin-deficient mice. All mice used in this study were maintained under specific pathogen-free conditions and ranged from 6 to 8 wk of age. All experimental protocols using mice were approved by the Chiba University animal committee. All animal care was performed in accordance with the guidelines of Chiba University.

The Abs used for the chromatin immunoprecipitation (ChIP) assay were anti-acetylhistone H3-K9 (06-599; Millipore), anti-trimethylhistone H3-K4 (AR-0169; LP Bio), and anti-Menin (A300-105A; Bethyl). The Abs used for cytoplasmic and cell surface staining were as follows: FITC-conjugated anti–IFN-γ mAb (clone XMG1.2; BioLegend), allophycocyanin-conjugated anti–IL-4 mAb (clone 11B11; BioLegend), allophycocyanin-conjugated anti–IL-5 mAb (clone TRFK5; BioLegend), PE-conjugated anti–IL-13 mAb (eBio13A; eBioscience), Alexa 647–conjugated anti-GATA3 (560068; BD PharMingen), BV421-conjugated anti-CD4 mAb (clone GK1.5; BioLegend), FITC- or allophycocyanin-conjugated anti-DO11.10 TCR mAb (clone KJ1-26; eBioscience), and PE-conjugated anti–thymic stromal lymphopoietin receptor (anti-TSLPR) mAb (FAB5461P; R&D Systems). The Abs used for cell culture were anti-TCRβ mAb (clone H57-597) and anti-CD28 mAb (clone 37.51; BioLegend).

Recombinant mouse IL-4 was purchased from PeproTech. The OVA peptide (residues 323–339; ISQAVHAAHAEINEAGR) was synthesized by BEX (Tokyo, Japan).

Effector Th2 (eTh2) cells were generated as previously described (11). Splenic naive CD4 T cells (CD44^loCD62L^hi) from wild-type (WT) or Menin-deficient mice were purified using a magnetic cell sorter (autoMACS; Miltenyi Biotec) yielding a purity of >98%. Purified naive CD4 T cells were stimulated with immobilized anti-TCRβ mAb (clone H57-597; 2.5 μg/ml) in the presence of IL-2 (25 U/ml), IL-4 (100 U/ml; PeproTech), and anti-CD28 mAb (clone 37.51; 1 μg/ml) for Th2 cell differentiation.

Splenic CD4 T cells were stimulated under Th2 culture conditions for 5 d in vitro. The Th2 cells were further cultured in vitro for another 2 d in the absence of any exogenous cytokines. The cultured Th2 cells were then restimulated with immobilized anti-TCRβ mAbs (clone H57-597; 2.5 μg/ml) with IL-2 (25 U/ml) and anti–IL-4 mAb (clone 11B11) for 5 d. This cycle was repeated two times (Th2-2nd to Th2-3rd).

mTh2 cells were generated as previously described (15). In brief, splenic CD4 T cells from DO11.10 OVA-specific TCR Tg mice were stimulated with OVA peptides (Loh15, 1 μM) plus irradiated (3500 rad) Thy-1.2–depleted splenic cells under Th2-culture conditions for 6 d in vitro. The cultured eTh2 cells (3 × 10⁷) were transferred i.v. into BALB/c or BALB/c-nu/nu recipient mice. Four weeks after cell transfer, splenic CD4⁺ and DO11.10 TCR⁺ cells were purified using a cell sorter (BD FACSAria) and then used as mTh2 cells.

For the PCR studies, total RNA was isolated using the TRIzol reagent (Invitrogen), and cDNA was synthesized using oligo(dT) primers and Superscript II RT (Invitrogen). Quantitative RT-PCR was performed using the StepOnePlus Real-Time PCR System via the comparative threshold cycle method. The primers and TaqMan probes used for the amplification and detection of the indicated genes were purchased from Roche (sequences are available in Supplemental Table I). The expression of the target genes was normalized to those of the Hprt signals.

To knock down the Gata3 genes, we used the Mouse T cell Nucleofector Kit (Amaxa) in accordance with the manufacturer’s protocol. The small interfering RNAs (siRNAs) used for the knockdown of Gata3 (s66482) and the negative control (AM4635) were purchased from Thermo Fisher Scientific.

The pMX-human GATA3 (hGATA3)-IRES-hNGFR plasmid was generated as previously described (17). The infected cells were enriched using a cell sorter (BD FACSAria) with anti-hNGFR (clone C40-1457; BD PharMingen) and subjected to a quantitative RT-PCR assay.

ChIP experiments for Menin, H3K9ac, H3K4me3, and control Ab were carried out using Dynabeads (Invitrogen) as previously described (18). In brief, 1–10 × 10⁶ cells were fixed with 1% paraformaldehyde at 37°C for 10 min. Cells were sedimented, washed, and lysed with SDS lysis buffer (50 mM Tris-HCl, 1% SDS, 10 mM EDTA, 1 mM PMSF, 1 μg/ml aprotinin, and 1 μg/ml leupeptin). The lysates were sonicated to reduce the DNA lengths to between 200 and 1000 bps. The soluble fraction was diluted in ChIP dilution buffer and incubated with Ab conjugated with Dynabeads protein A and G overnight at 4°C. The immune complexes were then captured using a magnet and washed with low-salt, high-salt, LiCl, and Tris–EDTA wash buffer. Enriched chromatin fragments were eluted with elution buffer (0.1 M NaHCO₃ containing 1% SDS). The eluted material was incubated at 65°C for 6 h to reverse the formaldehyde cross-links and treated with RNase A (10 μg/ml) and Proteinase K (40 μg/ml). DNA was extracted with a QIAquick PCR purification kit (QIAGEN). The total input DNA (cellular DNA without immunoprecipitation) was purified in parallel. A real-time quantitative PCR (qPCR) analysis was performed using the StepOnePlus Real-Time PCR System via the comparative cycle threshold method with TaqMan probes and primers (sequences are available in Supplemental Table I). To calculate the enrichment of each protein to a particular target DNA, we divided the values obtained for each target by the amount of the corresponding target in the input fraction. All of the results are expressed as percentages of input DNA.

Ab-specific immunoprecipitates and total input DNA samples were prepared using a NEBNext ChIP-Seq Library Prep Reagent Set for Illumina. Adaptor-ligated DNA was recovered using AMPure XP Beads. This DNA was then amplified by 15 cycles of PCR and again recovered using AMPure XP Beads. Fifty cycles of sequencing reaction were performed on an Illumina HiSeq1500. Read sequences (50 bp) were then aligned to the mm10 mouse reference genome (University of California, Santa Cruz, July 2011) using Bowtie. Each aligned read sequence was extended to 120 bp to efficiently detect duplicate reads aligned to identical locations. These 120-bp tags were used for further analyses (Bed file). SICER (19) was used for peak calling and the visualization of binding, with the parameters set as follows: window_size = 200, gap_size = 200, and false discovery rate = 0.01.

RNA-sequencing (RNA-seq) was carried out as previously described (18). Total cellular RNA was extracted with TRIzol reagent (Invitrogen). For cDNA library construction, we used TruSeq RNA Sample Prep Kit v2 (Illumina) in accordance with the manufacturer’s protocol. Sequencing the library fragments was performed on the HiSEquation 1500 System. For data analyses, read sequences (50 bp) were aligned to the mm10 mouse reference genome (University of California, Santa Cruz, December 2011) using the Bowtie (version 0.12.8) and TopHat (version 1.3.2) software programs. Fragments per kilobase of exon per million mapped reads for each gene were calculated using the Cufflinks (version 2.0.2) software programs. Downregulated genes were selected with the following criteria: 1) absolute fragments per kilobase of exon per million mapped reads >1 in WT cells, and 2) fold change of expression <0.5.

The ChIP-Seq data sets of Menin are available in the Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/geo) under accession number GSE99261.

For the intracellular staining of IFN-γ, IL-4, or GATA3, eTh2 cells or purified mTh2 cells were stimulated with immobilized anti-TCRβ mAb (clone H57-597; 2.5 μg/ml) and anti-CD28 mAb (clone 37.51; 1 μg/ml) for 4 h. The cells were then fixed with 4% paraformaldehyde for 10 min at room temperature and permeabilized in a permeabilizing solution (50 mM NaCl, 5 mM EDTA, 0.02% NaN₃ [pH 7.5]) containing 0.5% Triton X-100 for 10 min on ice. After blocking with 3% BSA in PBS for 15 min, the cells were incubated on ice for 45 min with anti–GATA3-Alexa 647 (560068; BD PharMingen). The cells were washed with FACS buffer at the end of each step. The Abs used for cell surface staining were as follows: BV421-conjugated anti-CD4 mAb (clone GK1.5; BioLegend) and allophycocyanin-conjugated anti-DO11.10 TCR mAb (clone KJ1-26; eBioscience). The flow cytometric analysis was performed on a FACSCalibur or FACSVerse instrument (BD Biosciences, San Jose, CA), and the results were analyzed with the FlowJo software program (BD Biosciences).

The purified mTh2 cells were restimulated with immobilized anti-TCRβ mAb (clone H57-597; 2.5 μg/ml) and anti-CD28 mAb (clone 37.51; 1 μg/ml) for 24 h. IL-4 production in cell culture supernatants was analyzed by a standard sandwich ELISA protocol using Ab pairs clone BVD4-1D11 and clone BVD6-24G2-biotin (BD Biosciences). IL-13 was measured using a DuoSet mouse IL-13 ELISA set (R&D Systems).

Naive CD4 T cells from WT or Menin-deficient DO11.10 OVA-specific TCR Tg mice were stimulated with an OVA peptide plus APC under Th2-culture conditions for 6 d in vitro. The cultured cells were harvested and administered i.v. through the tail vein to WT BALB/c mice (1 × 10⁶ or 3 × 10⁶ cells per mouse, day 0) 1 d before the first airway challenge with aerosolized OVA. After cell transfer, the mice were exposed to allergen challenges via the airway on days 2 and 4, and assays were conducted on day 6.

BALB/c WT or Menin-deficient mice were immunized with OVA (i.p.) on days 0 and 7, and were challenged with OVA inhalation on days 14 and 16. Bronchoalveolar lavage (BAL) fluid was collected 24 h after the last OVA inhalation.

BALB/c nu/nu mice were administered Th2 cells. The mice were sacrificed on day 30, and the mTh2 cells obtained by cell sorter (BD FACSAria) were transferred into BALB/c recipient mice. The recipient mice were then subsequently challenged with OVA inhalation on days 1, 3, 8, and 10 after cell transfer. Assays were conducted on day 11.

The collection of BAL fluid was performed 24 or 48 h after the last OVA challenge, as described previously. A total of 100,000 viable BAL cells was cytocentrifuged onto slides by a Cytospin 4 (Thermo Electron, Waltham, MA) and stained with Diff-Quik Stain Kit (Sysmex). Two hundred leukocytes were counted on each slide. The cell types were identified using morphological criteria.

Mice were sacrificed 24 h after the last OVA inhalation, and the lungs were fixed in 10% (v/v) formalin. The samples were sectioned and stained with H&E for the examination of pathological changes under a light microscope at ×200.

For the RT-PCR analysis, one sample t test was used to determine whether log₂(fold change) was equal to zero (p values are available in Supplemental Table II). For the noninferiority test, the lower confidence limit was set as log₂(0.9). All other analyses were performed using Student t test.

Menin is an essential component of the MLL1/2 complex that possesses methyltransferase activity for histone H3K4 (20). We previously showed that once Th2 cell differentiation takes place, GATA3 and IL-4 expression are maintained via the recruitment of Menin to the Gata3 gene locus, even in the absence of IL-4–mediated STAT6 activation (9). However, whether Menin plays a role in Th2 cell–mediated immune responses is unclear.

We first used a model of airway inflammation in which Th2 cells are key mediators of eosinophilic inflammation and pathology to assess the role of Menin in Th2 cell–mediated airway inflammation. We adoptively transferred eTh2 cells generated from DO11.10 TCR Tg WT or Menin-deficient (Menin KO) mice into syngeneic BALB/c recipient mice and assessed the acute airway inflammation mediated by eTh2 cells (Fig. 1A). At 2 and 4 d after cell transfer, BALB/c recipient mice were challenged by OVA inhalation, and BAL fluid samples were collected for an analysis. The total number of infiltrating leukocytes in the BAL fluid was significantly decreased (p < 0.05) in the group that received Menin-deficient Th2 cells (Fig. 1B, 1C). Furthermore, we detected a large increase in the number of eosinophils in the BAL fluid from mice receiving WT Th2 cells that was not observed in the mice receiving Menin-deficient Th2 cells.

Our previous studies demonstrated that a Menin deficiency results in the impaired accumulation of Th17 or nonpolarized Th cells in the lung (11, 12). To determine whether the impaired ability of transferred Menin-deficient Th2 cells to induce airway inflammation was due to the impaired accumulation of Th2 cells, we examined the numbers of transferred Th2 cells in the lungs of the recipient mice. We found that the number of CD4⁺ DO11.10.TCR⁺ cells in the lungs of mice receiving 1 × 10⁶ WT Th2 cells was higher than in mice receiving 1 × 10⁶ Menin-deficient Th2 cells, but comparable with that in mice receiving 3 × 10⁶ Menin-deficient Th2 cells (Fig. 1D). This result suggests that the accumulation in the lung is also impaired in Menin-deficient Th2 cells. However, this result also suggests that we can examine the role of Menin in the Th2 cell–mediated immune responses by adjusting the cell numbers for adaptive transfer. Eosinophilic airway inflammation was more severe in mice receiving 1 × 10⁶ WT Th2 cells than mice receiving 3 × 10⁶ Menin-deficient Th2 cells (Fig. 1B, 1C). We next examined the cytokine expression profiles of the transferred Th2 cells that engrafted to the lung after OVA challenge (Fig. 1E, 1F). A dramatic reduction in the proportion of IL-4–, IL-5–, and IL-13–producing CD4⁺ DO11.10.TCR⁺ cells was observed in the lung in the absence of Menin. We therefore concluded that the impaired ability of transferred Menin-deficient cells to induce airway inflammation was due to the decreased function of Th2 cells, as well as the impaired accumulation of Th2 cells in the lung. We then examined the airway inflammation directly in CD4-specific Menin-deficient mice and also observed attenuation in the inflammation in the absence of Menin (Fig. 1G, 1H). These results suggest that Menin is required for inducing type 2 immune responses in vivo.

The earlier results also raise the question of why Menin deficiency resulted in an attenuation in airway inflammation despite the induction of Th2 cell differentiation not being affected (Fig. 1A). To address this issue, we examined expression profiles of 32 Th2-specific inducible genes (Asb2, Ccnjl, Ccr8, Crem, Cyp11a1, Dusp4, Ecm1, Epas1, F2r, GATA3, Grtp1, Gzma, Il13, Il1r2, Il24, Il4, Il5, Itgb3, Jdp2, Mapk12, Nfil3, Oit3, Penk, Plcd1, Ptgir, Rnf128, S100a1, Spry2, Tanc2, Tmtc2, Tnfrsf8, and Tube1) that we have previously reported (21). We excluded the Oit3 gene, because its expression was too low to evaluate, and analyzed the other 31 Th2-specific inducible genes.

We first assessed whether the expression of the Th2-specific genes was changed after multiple rounds of TCR stimulation using Th2-1st and Th2-3rd cells. Only five genes (Il5, Il13, Rnf128, Tanc2, and Cyp11a), four of which have been reported to be associated with Th2 cell–mediated inflammation (22–24), showed a significant increase in the expression in Th2-3rd cells (Fig. 2A). Indeed, multiple rounds of differentiation are required for generating IL-5–producing human Th2 cells (25). In contrast, three genes (Tube1, Ccnjl, and Grtp1) were downregulated in Th2-3rd cells compared with Th2-1st cells. The expression of the other 23 genes was maintained in Th2-3rd cells upon multiple rounds of TCR stimulation. We next examined whether Menin deficiency affected the expression of these genes in Th2-1st or Th2-3rd cells. Interestingly, six genes (Spry2, Tmtc2, F2r, Gzma, Jdp2, and Itgb3) showed a significant increase (p < 0.05) in mRNA expression in Menin-deficient Th2-1st cells (Fig. 2B, left). In contrast, two genes (Tube1 and Grtp1) showed a significant decrease in mRNA expression in Menin-deficient Th2-1st cells (p < 0.05). Consistent with a previous report, in Th2-1st cells, expression levels of Gata3 and Il4 did not show a significant change in the absence of Menin (9). In the context of Th2-3rd cells, 14 genes (Il5, Il13, Spry2, Tanc2, Il4, Cyp11a1, Ccr8, Il24, Gata3, Gzma, S100a1, Mapk12, Tube1, and Asb2) showed a significant decrease (p < 0.05) in mRNA expression in Menin-deficient cells (Fig. 2B, middle). Representative expression profiles for these genes are shown in Fig. 2C. In contrast, two genes (Spry2 and Gzma) showed a significant increase in mRNA expression in Menin-deficient Th2-3rd cells (p < 0.05). Using a retroviral gene transduction system, we next assessed whether the enforced expression of the hGATA3 gene could restore the dysregulation of the 14 genes (Il5, Il13, Spry2, Tanc2, Il4, Cyp11a1, Ccr8, Il24, Gata3, Gzma, S100a1, Mapk12, Tube1, and Asb2) observed in Menin-deficient Th2-3rd cells (Fig. 2B, right). hGATA3 protein has the same ability as mouse GATA3 protein to induce Th2 cytokine expression (17). Eight genes (Il5, Il13, Tanc2, Cyp11a1, Il24, Gata3, Gzma, and Tube1) were restored by the introduction of the hGATA3 gene, indicating that the dysregulation of these genes was dependent on the decreased expression of GATA3. In addition, an RNA-seq analysis revealed that Menin deficiency reduced the expression of 83 genes in Th2-1st cells, whereas 1033 genes were downregulated in Menin-deficient Th2 cells (Fig. 2D). These results suggest that a major role of Menin is maintaining the Th2-specific gene expression in differentiated Th2 cells after multiple rounds of TCR stimulation. The results obtained from in vitro experiments also suggest that decreased expression of Th2-specific genes in the absence of Menin is associated with the attenuation of airway inflammation in vivo.

Next, we examined whether dysregulation of the Th2-specific gene expression in Menin-deficient Th2-3rd cells was dependent on the lack of Menin, decreased expression of GATA3, or both. The Th2-specific genes were classified into four groups according to a flowchart in which the gene expression profiles under three conditions [i.e., genetic deletion of Menin, Gata3 siRNA treatment (21), and retroviral gene transduction of hGATA3] were taken into consideration (Fig. 3A). In the final step of this classification, we compared the expression of the selected eight genes between hGATA3-introduced Menin KO and Mock-introduced WT cells using a noninferiority test. Consequently, Asb2, Ccr8, Gzma, Il4, Il5, Il13, Il24, Mapk12, Tanc2, and Tube1 were classified as the genes controlled by both GATA3 and Menin (group 1). Interestingly, only Gzma was negatively regulated by Menin, whereas the other nine genes were positively regulated. Although the Gzma gene was downregulated by Gata3 siRNA treatment, the forced expression of hGATA3 also reduced the Gzma expression for some unknown reason. Seven genes (Crem, Cyp11a1, F2r, Nfil3, Ptgir, Rnf128, and Tmtc2) were found to be dependent on GATA3, and all of them were downregulated by Gata3 siRNA treatment (group 2). The Cyp11a1 gene was the only gene that showed a noninferior expression (i.e., lower confidence limit > log₂[0.9]) in hGATA3-introduced cells compared with mock-introduced control cells and was classified into group 2. Two genes (Spry2 and S100a) were found to be controlled in a Menin-dependent and GATA3-independent manner (group 3). With 11 genes (Ccnjl, Dusp4, Ecm1, Epas1, Grtp1, Il1r2, Itgb3, Jdp2, Penk, Plcd1, and Tnfrsf8), neither the effects of Gata3 knockdown nor Menin deficiency were statistically significant (group 4). We also performed ChIP-sequencing analysis for Menin and detected a strong signal at the Gata3 gene locus (Fig. 3B), indicating that Menin directly regulates the maintenance of Gata3 expression via binding to the Gata3 gene locus. Direct binding of Menin was observed in all group 1 genes except Asb2 and Mapk12, indicating that Menin has a positive effect on the expression of these genes. Among group 3 genes, Spry2 and S100a1 showed binding signals for Menin, with Spry2 negatively regulated by Menin, whereas S100a1 was positively regulated. Menin binding was also detected for the group 2 and 4 genes, on which the effect of Menin deficiency was not statistically significant.

We next examined the roles of Menin in mTh2 cells. We first assessed whether the expression of the Th2-specific genes was maintained during mTh2 cell formation. Comparison of mRNA levels between eTh2 and mTh2 cells revealed that expression was decreased in most genes (Fig. 4A). Among them, 16 genes (Ptgir, Nfil3, Ccnjl, Il13, Il5, Ecm1, Asb2, Tube1, Tmtc2, Il1r2, Jdp2, Tnfrsf8, S100a1, Gzma, Il24, and Cyp11a1) showed a significant decrease in the mRNA expression in mTh2 cells (p < 0.05). These results indicate that mTh2 cells have less potential to cause inflammation than eTh2 cells.

We next examined whether Menin deficiency affected the expression of these genes in mTh2 cells. In the absence of Menin, the number of mTh2 cells was significantly reduced compared with the numbers in WT mice and was too small to perform a molecular analysis when using BALB/c mice as recipients (Supplemental Fig. 1A). To obtain sufficient cell numbers, we used BALB/c-nu/nu mice as recipients. Among the 31 Th2-specific genes, 4 (Mapk12, Gata3, Il4, and Il13) genes were found to be significantly decreased in Menin-deficient mTh2 cells, whereas 1 (Gzma) gene was found to be significantly increased (Fig. 4B). For the Gata3 transcripts, the first alternative exons (1a and 1b) are known to be spliced to a common exon 1, which contains the translation start site (26, 27). We assessed the levels of transcripts of both exon 1a-1 and exon 1b-1 in WT and Menin-deficient mTh2 cells (Fig. 4C). Consistent with the result of Gata3 mRNA, the level of the exon 1b-1 transcript was decreased in Menin-deficient mTh2 cells. We also found that the level of the exon 1a-1 transcript was decreased in Menin-deficient mTh2 cells. However, the exon 1b-1 transcript was much more abundant than that of exon 1a-1 both in WT and Menin-deficient mTh2 cells (9, 27). These results suggest that Gata3 mRNA expression is mostly dependent on the transcription started from the exon 1b, which is epigenetically regulated by Menin in mTh2 cells. Consistent with the findings from an mRNA analysis, GATA3 protein expression was significantly decreased in Menin-deficient mTh2 cells when compared with WT controls (Fig. 4D). In addition, a dramatic reduction in the production of IL-4 and IL-13 was observed in Menin-deficient mTh2 cells (Fig. 4E). To further clarify the possible mechanism by which Menin regulates the Th2-specific gene expression, we next assessed the binding of Menin and the histone modification states around the Gata3 and Th2 cytokine gene loci by ChIP assays. Menin binding was detected not only in the downstream regions of the transcription start site of the Gata3 gene but also in the upstream regions of the transcription start site, consistent with the results previously obtained in eTh2 and differentiated Th2 cells (Fig. 4F) (9). In the absence of Menin, the levels of H3K9ac and H3K4me3 were decreased at the upstream regions of the Gata3 proximal promoter (probes [−5] and [−4]), whereas weaker effects were observed at the gene body (probes [+0], [+3], and [+17]; Fig. 4G, upper panels). We also observed the accumulation of Menin at the Th2 cytokine gene loci and found that Menin deficiency affected the histone modifications at the Th2 cytokine gene loci (Fig. 4G, lower panels). These results indicate that the Menin-dependent high-level expression of GATA3 proteins was required for Th2 cytokine-mediated type 2 immune responses. Menin deficiency itself or its induced Gata3 downregulation seemed to be sufficient for altering the histone modification degree at the Th2 cytokine gene loci and reducing the expression of Th2 cytokine genes.

We next assessed the airway inflammation in BALB/c recipient mice mediated by mTh2 cells (Supplemental Fig. 1B). OVA inhalation was performed on days 31 and 33 after transfer, and BAL samples were collected for an analysis. The numbers of total infiltrating leukocytes and eosinophils in the BAL fluid was significantly decreased (p < 0.05) in the group receiving Menin-deficient Th2 cells (Supplemental Fig. 1C). As mentioned earlier, the number of mTh2 cells was reduced in the absence of Menin (Supplemental Fig. 1A). To examine whether the attenuation of the airway inflammation in the absence of Menin was due to the reduced mTh2 cell number or to the decreased mTh2 cell function, we first generated WT or Menin-deficient mTh2 cells in BALB/c-nu/nu mice. After the isolation of mTh2 cells, the same numbers of WT or Menin-deficient mTh2 cells were transferred into BALB/c recipient mice, which were subjected to OVA inhalation (Fig. 5A). In this experimental setting, the number of CD4⁺ DO11.10.TCR⁺ cells accumulated in the lungs was comparable between the WT and Menin-deficient groups (Fig. 5B). Thus, we can examine the role of Menin in the maintenance of the Th2 cell function regardless of the cell survival or migration. Consequently, the number of eosinophils in the BAL fluid was significantly decreased in the group that received Menin-deficient mTh2 cells (Fig. 5C). Consistent with these findings, the deletion of Menin in mTh2 cells resulted in the diminished infiltration of mononuclear cells around the peribronchiolar and perivascular regions of the lungs (Fig. 5D). The proportion of IL-4–, IL-5–, and IL-13–producing CD4⁺ DO11.10.TCR⁺ cells was also decreased in the lung after OVA inhalation in the group that received Menin-deficient mTh2 cells (Fig. 5E, Supplemental Fig. 1D).

In addition, we investigated whether the expression of the cytokine/chemokine receptors known to be associated with Th2 memory function (Crlf2, Ccr4, Il1rl1, and Tnfrsf9) was affected by Menin deficiency (28, 29). The expression of Crlf2 encoding TSLPR was decreased in Menin-deficient mTh2 cells isolated from the spleen (Supplemental Fig. 1E). We found that the Menin deficiency had little effect on the other cytokine/chemokine receptor genes we analyzed (Ccr4, Il1rl1, and Tnfrsf9). We also assessed the protein expression of TSLPR in the transferred CD4⁺ DO11.10.TCR⁺ cells that engrafted in the lung after OVA challenge (Fig. 5F, Supplemental Fig. 1F). The expression of TSLPR was decreased in Menin-deficient mTh2 cells in the lung. These results indicate that the Menin-mediated expression of cytokines and cytokine receptor expression is also required for mTh2 cell–dependent immune response and airway inflammation in vivo.

Our previous work has established that the Menin/TrxG complex plays a critical role in the maintenance of Gata3 expression in differentiated Th2 cells (9). Menin also regulates both the induction and the maintenance of Th17 cell functions (11). We extended these studies, and in this article report a crucial role for Menin in the regulation of Th2 cell identity using the expression profiles of Th2-specific genes as indicators. Our results show that Menin bound to the Gata3 gene locus and maintained Gata3 expression and Th2 cell identity after multiple rounds of Ag stimulation. In cases of a long interval between initial Ag exposure (priming) and secondary or greater Ag exposure (rechallenging), Menin deficiency resulted in decreased mTh2 cell function in vivo. These results seem to point to an important role for the Menin-Gata3 axis in the maintenance of Th2 cell identity and Th2 cytokine-dependent pathology.

We first used a very simple airway inflammation model in which the recipient mice with eTh2 cells were exposed to allergen challenges twice (30). We observed attenuation in the airway inflammation in the mice that received Menin-deficient Th2 cells. This result was at first unexpected, because the induction of Th2 cell differentiation was not affected by the absence of Menin (9). However, taking into consideration that eTh2 cells were exposed to allergen challenges twice in this model, this model may reflect the results obtained with multiple rounds of TCR stimulation in vitro. Indeed, dramatic effects of Menin deficiency on Th2 cell functions were observed after multiple rounds of TCR stimulation, consistent with the results obtained in the airway inflammation model. We also observed attenuation in the airway inflammation in the absence of Menin when we examined airway inflammation directly in CD4-specific Menin-deficient mice. These results point to an important role for Menin in the maintenance of Th2 cell identity after multiple rounds of exposure to the Ag.

The effect of Menin deficiency on airway inflammation observed in this experimental setting may be partly explained by the decreased number of transferred cells caused by cellular senescence (12). However, under an experimental condition in which a comparable number of transferred cells accumulated in the lung, the attenuation of airway inflammation was also observed in mice transferred with Menin-deficient Th2 cells. We also performed similar experiments using WT and Menin-deficient mTh2 cells and obtained similar results. Thus, Menin deficiency reduced the number of transferred cells by inducing early senescence; however, when we compensated for the effects of the difference in cell numbers between WT and Menin-deficient mice, airway inflammation was also attenuated in the absence of Menin. These results indicate that Menin is involved in the maintenance of the Th2 function.

Regarding cell survival, the phenotypes of the Menin-deficient cells are similar to those observed in cells lacking Polycomb group (PcG) proteins. Indeed, the deletion of Bmi1 or Ezh2 results in the reduction of the memory T cell number (31, 32). Why does Menin deficiency impair the memory T cell survival as well as PcG deficiency does, despite Menin being a component of the TrxG complex, which has functions opposite those of the PcG proteins? Is there a shared mechanism underlying cell apoptosis between Menin deficiency and PcG deficiency? A genome-wide gene expression analysis based on DNA microarray and RNA-seq revealed that the upregulation of the Gzma gene expression is one of the dramatic changes observed in both Menin-deficient and PcG-deficient cells (31, 33). Granzyme A is reported to be associated with cell-intrinsic apoptosis (34). In cells that lack PcG proteins, cellular senescence and apoptosis are primarily induced by the upregulation of the apoptotic genes, such as Cdkn2a, Pmaip1, and Bcl2l11 (32, 35, 36). In cases of Menin deficiency, the decreased expression of the Bach2 gene is a possible mechanism underlying early senescence and cellular apoptosis (12). Together, PcG and Menin prevent apoptosis by regulating the proper expression of Cdkn2a and Bach2, respectively. In addition, granzyme A is also reported to act as an endogenous modulator of proinflammatory cytokine secretion (37). Thus, the increased expression of the Gzma gene may be associated with the senescence-associated secretory phenotype (SASP), in which cytokine hyperproduction is observed (38). Even though the SASP is well-defined in tumor cells and stromal cells, further studies are needed to elucidate the roles of the SASP in T cells.

PcG complexes are known to be classified into two canonical types: Polycomb repressive complex (PRC) 1 and PRC2. Whereas the PRC1 component Bmi1 has a limited number of target genes and its major role is repressing the apoptotic gene expression, the PRC2 component Ezh2 has a broader range of target genes and more diverse functions (18). For example, Ezh2 is involved in the repression of the lineage-specifying genes (Tbx21 and Gata3) in Th cells (33) and the memory-signature genes in CD8 T cells (39). Ezh2 is also shown to be important for Foxp3-mediated gene regulation in regulatory T cells (40, 41). In contrast, TrxG proteins, including Menin, participate in the maintenance of the high expression of lineage-specifying genes (Gata3 and Rorc) rather than the induction of these genes (9, 11). In addition, the TrxG-mediated maintenance of H3K4 methylation is important for the recall responses of memory T cells (15). Some reports show that permissive histone markers, such as H3K4me3 or H3K9ac, are responsible for the quick cytokine production of memory T cells upon antigenic stimulation (14, 42). This study provides new evidence that Menin mediates the epigenetic memory of the T cell–dependent immune responses.

When studying DNA-binding proteins that have broad target genes, researchers often face difficulty in interpreting the phenotypes of KO mice, because many targets are supposed to be dysregulated. In particular, the experimental protocol should be considered carefully when analyzing DNA-binding proteins whose deletion affects the cell survival. Our previous studies have shown that Menin deficiency results in fewer transferred CD4⁺ cells in vivo (11, 12). In this study, we wanted to assess whether Menin was involved in the maintenance of the Th2 functions. Therefore, we compensated for the effects of the difference in the cell survival between WT and Menin-deficient cells, and established our experimental conditions very carefully, wherein a comparable number of transferred cells accumulated in the lung.

Even in this experimental setting, the proportions of IL-4–, IL-5–, and IL-13–producing mTh2 cells were decreased in the lung after OVA inhalation. Our interpretation of these experiments is that Menin is indispensable for the maintenance of the Th2 cytokine production ability, and that the deletion of Menin results in a decreased ability to produce these cytokines. We suspect this because it is becoming apparent that Th cells display plasticity and that the plasticity is controlled by epigenetic regulators (43). For example, Ezh2-mediated and Suv39h1-dependent H3K9me3 have been shown to be responsible for the maintenance of the stability of Th2 cells (33, 44). In addition, Th2 cells can acquire the ability to produce IFN-γ, indicating that there is some degree of functional plasticity in Th2 cells (45). Another potential explanation for the decreased Th2 cytokine production under conditions of Menin deficiency is the lack of a positive feedback loop of IL-4 (46). It is also possible that the deletion of Menin results in IL-4–producing cell-specific apoptosis. However, the Menin expression is comparable between Th1 and Th2 cells (17), indicating that the Menin-dependent regulation of the cell survival is not likely controlled by a subset-specific mechanism. A recently developed epigenome editing technique will be helpful for studying DNA-binding proteins that have broad targets (47). With this technique, site-specific alternation in epigenetic marks can be achieved. Thus, researchers can focus on the targets that they want to analyze, leaving the other target genes intact.

In mTh2 cells, a lack of Menin resulted in a significant decrease in the Gata3, Il4, and Il13 expression, but not the Il5 expression. This may be because of the increased expression of Il5 in Menin-deficient eTh2 cells, although this increase was not statistically significant in this study. The hyperproduction of IL-5 caused by a Menin deficiency has also been reported in a study, suggesting that cellular senescence associated with Bach2 downregulation is responsible for the increased expression of Il5 (12). Taken together, these findings suggest that Menin deficiency potentially increases the Il5 expression but actually decreased the Gata3 expression in mTh2 cells, ultimately resulting in slightly decreased Il5 expression as a final output.

In Menin-deficient mTh2 cells, a reduction in the H3K9ac and H3K4me3 levels was observed; however, this reduction was mild. One possible explanation for this mild reduction is the redundant function of H3K4 methyltransferases. As mentioned earlier, six methyltransferases have been identified in mammalian cells (6). Menin has been reported to be a specific component of the MLL1/2-containing complex, and the other four methyltransferases (MLL3, MLL4, SET1A, and SET1B) can compensate for the function of the Menin-deficient MLL1/2 complex. A reduction in the H3K9ac and H3K4me3 levels was obvious in the upstream region of the Gata3 proximal promoter, where the displacement of PcG by TrxG proteins was observed during Th2 cell differentiation (9). In this study, we detected Menin binding at the region between the distal and proximal promoters of the Gata3 gene locus in mTh2 cells. Taken together, these findings suggest that Menin binding to the region between the distal and proximal promoters of the Gata3 gene locus is important for maintaining the Gata3 expression in mTh2 cells.

We thank Dr. Atsushi Iwama, Dr. Motohiko Oshima, Dr. Yutaka Suzuki, Dr. Atsunori Saraya, and Dr. Kaoru Sugaya for excellent experimental suggestions.

The authors have no financial conflicts of interest.