CBX7 functions as a methylation-dependent inducer of gene transcription and regulator of cytosolic signaling in lymphoid cells

Abstract

Chromobox protein-7 (CBX7) functions as a gene repressor. We, unexpectedly, found that CBX7 formed a methylation-dependent transcriptional complex, which induced gene transcription by binding to cytokine gene promoters. CBX7 translocated to the cytosol and formed a methylation-dependent signaling complex with c-Raf, MAPK (mitogen-activated protein kinase) kinase 1/2 (MEK1/2), and casein kinase-2 (CK2)–α to generate and sustain extracellular signal–regulated kinase 1/2 (ERK1/2) signaling. CBX7 is an allergen-inducible gene. Genetic and pharmacologic interventions established an essential role for CBX7 for the production of cytokines by mouse and asthmatic patient lymphoid cells, and for the induction of allergic asthma in multiple mouse models. RNA sequencing demonstrated a large-scale loss and gain in gene transcription in Cbx7^−/− T cells. The top down-regulated pathways included cytokine-cytokine receptor interaction, asthma, and T helper cell differentiation. CBX7 induction of the transcriptional activation complex and methylation of the ERK1/2 signalosome was specific for lymphoid cells as they were absent in epithelial cells. Our studies established a previously unknown paradigm of CBX7-generated methylation-dependent signaling complexes regulating inflammation.

CBX7 positively regulates cytokine expression by recruiting transcriptional activators and enhancing ERK1/2 activation in T cells.

INTRODUCTION

Chromobox protein7 (CBX7) is a member of the phylogenetically ancient polycomb repressor complex-1 (PRC1) (1–4). This multimolecular repressor complex and its sister complex, PRC2, are conserved chromatin modifiers that control gene transcription to promote developmental pattern formation and organogenesis (5). PRC2 contains histone methyl transferases—EED, SUZ12, and EZH1/2—which induce di- and trimethylation of histone 3 Lysine 27 (H3K27). CBX7 is a methyl reader, which mainly recognizes trimethylated H3K27 (H3K27me3) and recruits the other member of PRC1 complex. The PRC1 E3 ubiquitin ligase RING1 A/B then ubiquitinates H2A K119 (6). This ubiquitination results in chromatin compaction and gene repression. CBX7 positively regulated stem cell self-renewal and differentiation, neuron regeneration, and tissue repair (2, 7, 8). Cbx7 knockout (KO) mice were prone to developing liver and lung cancer, especially in aged mice (9). They had malformation of dentin and bones. CBX7 was considered an oncogene for certain cancers, especially myeloid leukemia and gastric cancer (10, 11). Some recent studies demonstrated that individual members of the PRC complex functioned as inducers of gene transcription instead of repression under certain circumstances (12–14). Gao et al. showed that autism susceptibility candidate-2 (AUTS2), a noncanonical member of PRC1, recruited p300 and induced gene transcription, while the associated casein kinase-2 (CK2) inhibited the repressive function of PRC1 (12). Bmi1 and Mel18 positively regulated T helper 2 cell (T_H2) differentiation and type-2 allergic inflammation through multiple mechanisms (15–17). CBX2 facilitated T cell development in the thymus (18). CBX4 positively regulated thymic epithelial proliferation and development and, secondarily, promoted thymopoiesis (19). However, CBX7’s role as a positive regulator of transcription has not been reported yet.

In addition to their nuclear function, certain PRC members function in the cytosol. The Tarakhovsky laboratory reported the presence of EZH1/2 in the cytosol of T cells (20, 21). EZH1/2 interacted with CD3ε and positively regulated extracellular signal–regulated kinase 1/2 (ERK1/2) but not p38 mitogen-activated protein kinase (MAPK), nuclear factor κB, or calcium signaling. Consequently, EZH1/2-deficient cells did not produce interleukin-2 (IL-2) and failed to proliferate. The translocation of CBX7 to the cytosol and its role in the cytoplasm are yet to be explored, and in particular, the role of CBX7 in lymphoid cells and in asthma is unknown. For this reason, we examined the function of CBX7 in cells and processes relevant for allergic inflammation. Here, we showed that CBX7 bound to methylated MAPK signaling intermediates in the cytosol stabilized the signaling complex and regulated T cell receptor (TCR)–induced gene transcription.

RESULTS

Inhibition of CBX7 function reduces the lymphocyte capacity to produce type-2 cytokines

We compared the transcriptomic profile of the CBX family members in the lung tissue from an allergen-induced asthma model in mice by microarray, which was done as a part of another project as described previously (22). The expression of Cbx1 and Cbx7 was significantly up-regulated in the allergen-induced asthma model (Fig. 1A). The foregoing suggests that Cbx7 is an allergen-inducible gene. The expression of CBX7 mRNA was up-regulated in peripheral blood mononuclear cells (PBMCs) from asthmatic patients compared to disease controls (Fig. 1B). We investigated the role of TCR stimulation on CBX7 expression. CBX7 protein expression was increased after TCR stimulation of CD4 T cells with anti-CD3/28 antibodies for 72 hours as shown by Western blotting (Fig. 1C). The latter implies a role for TCR in allergen induction of CBX7. These observations prompted us to investigate the role of CBX7 in asthma. We focused on lymphoid cells. We first used the pharmacological inhibitor of CBX7, MS37452, which selectively binds to the chromo domain of CBX7 to competitively inhibit its binding to methylated proteins (23). We cultured purified lineage-negative (Lin⁻) innate lymphoid cells (ILCs) or PBMCs from asthmatic donors with increasing doses of MS37452 (abbreviated as CBX7-inh) or dimethyl sulfoxide (DMSO; vehicle: Veh) for 72 hours and analyzed cytokine production by flow cytometry (FCM). The CBX7-inh blocked IL-4 expression in ILCs in a dose-dependent manner (Fig. 1, D and E), as well as IL-5 and IL-13 expression by CD4 T cells and Lin⁻ cells from PBMCs (fig. S1, A to G). To validate the FCM data, we measured IL-5 and IL-13 protein levels by enzyme-linked immunosorbent assay (ELISA) in culture supernatants of PBMCs that were incubated with increasing doses of the CBX7-inh (fig. S1, H and I). The inhibitor significantly reduced the levels of IL-5 and IL-13. To further examine these results, we cultured PBMCs from 16 asthmatic patients either with Veh or the CBX7-inh (100 μM) for 72 hours and analyzed cytokine production by FCM. We also measured IL-5 and IL-13 proteins levels in culture supernatants by ELISA. The CBX7-inh reduced IL-4⁺ and IL-5⁺Lin⁺ (contains T cells) as well as Lin⁻ [contains ILCs; (24)] (Fig. 1, F to J). The IL-5 and IL-13 protein levels were significantly lower in the culture supernatant from CBX7-inh–treated cells compared to Veh-treated cells (Fig. 1, K and L). The gating strategy and isotype controls for the flow analysis are shown in (fig. S1, A to C). Last, we examined the effect of the inhibitor on type-2 cytokine production by airway ILC2s from patients with asthma. The CBX7-inh inhibited IL-5 expression in bronchoalveolar lavage (BAL) ILCs (Fig. 1, M and N). Prolonged culture of PBMCs with the CBX7-inh increased cell death (fig. S1J). Pharmacological inhibitors sometimes have nonspecific activities. To examine the specific effect of CBX7, we tested three different short hairpin RNAs (shRNAs) in a green fluorescent protein (GFP)–expressing lentiviral vector for knockdown of CBX7. All three shRNAs (CBX7-sh1-3) reduced the expression of the CBX7 protein and decreased the frequency of IL-4⁺ CD4 T cells (Fig. 1, O to Q). These results suggest a positive role of CBX7 in type-2 cytokine production by blood and BAL lymphoid cells from patients with asthma.

Fig. 1. CBX7 regulates cytokine production by lymphoid cells.

(A) The expression of mRNA for the CBX7 (1–8) family proteins in the lung of allergen (Alternaria) or saline-sensitized mouse models of asthma (n = 4 mice per group). (B) CBX7 mRNA expression in PBMCs analyzed by reverse transcription quantitative real-time PCR. (C) Western blots show CBX7 expression in CD4 T cells after 72 hours of anti-CD3/28 antibody stimulation. The numbers below the blot indicate CBX7 levels normalized to actin (n = 3). (D and E) Effect of the CBX7 inhibitor (CBX7-inh) MS37452 on IL-4⁺ ILCs. Blood Lin⁻ cells were cultured with IL-2/IL-33 (20 ng/ml each) and increasing doses of the CBX7-inh for 72 hours. IL-4⁺ ILCs were measured by FCM (N = 4). (F to J) PBMCs from asthmatics were cultured either with DMSO (Veh) or the CBX7-inh (100 μM) for 72 hours. Representative flow plots and graphs show the expression of IL-4 and IL-5 in Lin⁻ cells and Lin⁺ cells (n = 16). (K and L) IL-5 and IL-13 protein levels in culture supernatants from the experiments for FCM [(H) to (L)]. (M and N) BAL cells were cultured with the CBX7-inh (100 μM) for 72 hours. IL-5 expression by Lin⁻IL-7Rα⁺CRTH2⁺ ILC2s was measured by FCM (n = 8). (O) CD3/28–stimulated PBMCs were transduced with a bicistronic GFP expressing pZMP lentiviral vector containing nontargeting (con) shRNA or CBX7 targeting shRNA—sh1, sh2, and sh3—for 72 hours and analyzed by Western blotting. Numbers below the CBX7 blot indicate the relative CBX7 levels compared to con-sh, which was normalized to β-actin (n = 3). (P and Q) Flow plots and bar graphs show the effect of CBX7 knockdown with CBX7-sh2 on IL-4⁺ CD4 T cells (n = 3). P values are shown above the graphs. The Wilcoxon signed-rank test was used to calculate the P value for data presented in (G) to (N). FSW, forward scatter width; CRTH2, chemoattractant receptor homologous molecule expressed on T helper type 2 cells.

The CBX7 inhibitor abrogates allergen-induced airway hyperreactivity and type-2 inflammation in an allergen-induced innate (ILC-dependent) model of asthma

ILC2s and T_H2 cells are important for the development of asthma. Next, we investigated the effect of the CBX7 inhibitor in an ILC-dependent asthma model in wild-type (WT) mice and in Rag1^−/− mice. The latter mice are devoid of T and B lymphocytes. We pretreated WT with either Veh or CBX7-inh followed by intranasal administration of the Alternaria allergen extract daily for 5 days and analyzed the phenotype on day 8 (Fig. 2A). The CBX7-inh–treated mice showed significantly reduced airway hyperreactivity (AHR) (Fig. 2B), lung inflammation and mucus production (Fig. 2, C and D), and influx of eosinophils in the lung tissue (Fig. 2, E and F). There was no change in neutrophils (Fig. 2G). ILC2s from the lungs of CBX-inh–treated mice produced significantly lower levels of IL-5 and IL-13 (Fig. 2, H to J). Similarly, we investigated effect of the CBX7-inh in Rag1^−/− mice (Fig. 2K). Similar to the WT mice, Rag1^−/− mice that were treated with the CBX7-inh showed reduced AHR (Fig. 2L), BAL eosinophils (Fig. 2M), and IL-5 and IL-13⁺ ILCs (Fig. 2, N to P). The levels of IL-4, IL-5, and IL-13 were significantly reduced in the BAL from the inhibitor-treated group (Fig. 2, Q to S). These results suggest that a functional CBX7 is required for ILC2-driven AHR and type-2 inflammation. We tested the safety and toxicity of the CBX7 inhibitor by treating the WT mice with either the CBX7-inh or Veh for 10 consecutive days and then monitored the body weight and examined the tissue morphology for any visible abnormalities in the kidney and liver by hematoxylin and eosin (H&E). We did not find any significant differences in the CBX7-inh–treated group as compared to the Veh-treated group (fig. S2, A and B).

Fig. 2. The CBX7 inhibitor MS37452 abrogates allergic asthma in an ILC-dependent mouse model.

A schematic diagram of allergen exposure to C57BL/6J WT. (A) WT B6 mice were pretreated with the Veh (DMSO) or the CBX7-inh (319 μg per dose) and 1 hour later, sensitized intranasally (IN) with the Alternaria (Alt) allergen extract (10 μg/ml) for 5 consecutive days. Treatment with the CBX7-inh reduced AHR (B), lung inflammation and mucus production (C and D), and the numbers of lung infiltrating eosinophils (E to G) and IL-5 and IL-13 secreting ILC2s (H to J). (K) A schematic diagram of allergen exposure to Rag1^−/− mice. The CBX7-inh reduced AHR (L), lung infiltrating eosinophils (M), the frequency of IL-5⁺ and IL-13⁺ ILCs in the lung (N to P), and the levels of IL-4, IL-5, and IL-13 in BAL (Q to S). D, day; BM, basement membrane; RL, lung resistance; Sal, saline.

Genetic ablation of Cbx7 abrogates type-2 inflammation in a combined ILC and T cell model of asthma

Next, we studied the effect of genetic ablation of Cbx7 on allergen-induced airway inflammation. We generated Cbx7^−/− mice using frozen sperm from MMRRC (Mutant Mouse Resource and Research Center). Western blot of spleen cells confirmed the absence of the Cbx7 protein in Cbx7^−/− mice (Fig. 3A). We studied the effect of Cbx7 genetic ablation in a modified asthma model, which involves both ILCs and T cells. Mice were sensitized intranasally with the Alternaria allergen extract once daily for 5 days (elicits an innate immune response), rested for 3 weeks, and then challenged for 3 consecutive days (elicits a T cell response) as in Fig. 3B. AHR to methacholine (Mch) (Fig. 3C), eosinophils in lung digest (Fig. 3D), peribronchial and perivascular inflammation, mucus production, and collagen deposition (Fig. 3, F to I) were reduced in the Cbx7^−/− group compared to the Cbx7^+/+ group. There was a low level Ly6-G⁺CD11b⁺ neutrophilic inflammation, which was modestly but significantly increased in Cbx7^−/− mice (Fig. 3E). We observed a mild decrease in survival of lung CD45⁺ cells in Cbx7^−/− mice (fig. S3D). Both CD4 T cells and ILC2s from the lung of Cbx7^−/− mice produced significantly lower levels of IL-5 and IL-13 compared to the Cbx7^+/+ mice (Fig. 3, J to M). The FCM gating strategy, isotype and FMO controls, and representative flow plots are shown in fig. S3, A to C. BAL cells from Cbx7^−/− mice showed a decrease in CD45⁺CD4⁺IL-13⁺ cells compared to Cbx7^+/+ mice (Fig. 3, N and O). Similarly, the BAL IL-4, IL-5, and IL-13 protein levels, measured by ELISA (Fig. 3, P to R), and mRNA levels from the lung tissue, measured by quantitative polymerase chain reaction (qPCR) (Fig. 3S), were reduced in the Cbx7^−/− group. The Cbx7^−/− group also had reduced levels of Il1β and the mucin gene Muc5ac compared to the Cbx7^+/+ group (fig. S3, E and F). However, there was no significant change in the mRNA levels of the mucosal cytokines Il25, Il33, and Tslp, and the chemokines Cxcl1 and Ccl20 between these two groups of mice (fig. S3, F and G). The results underscore the function of CBX7 as an inducer of type-2 cytokines, allergic inflammation, and experimental asthma in mice.

Fig. 3. Genetic ablation of CBX7 inhibits features of allergic asthma in a combined ILC2- and T_H2-dependent mouse model.

(A) Western blots showing the CBX7 protein expression in splenocytes from Cbx7^+/+ and Cbx7^−/− mice (n = 4). (B) A schematic of the timeline of Alternaria (Alt) allergen sensitization and challenge of Cbx7^+/+ or Cbx7^−/− mice in a combined ILC2- and T_H2-dependent model of asthma. Genetic ablation of Cbx7 resulted in reduced AHR (C), lung infiltrating eosinophils (D) and neutrophils (E), peribronchial and perivascular inflammation, mucus production, and airway remodeling (F to I), IL-5⁺ and IL-13⁺ CD4 T cells and ILCs (J to M) in the lung, IL-13⁺ CD4 T cells in BAL (N and O), and IL-4, IL-5, and IL-13 proteins in BAL (P to R), and their mRNA (S). (T to X) A schematic of adoptive transfer of allergen-sensitized splenic CD4 T cells from Cbx7^+/+ and Cbx7^−/− mice to Rag2:Il2rγC KO mice. Mice that received CD4 T cells from Cbx7^−/− mice had reduced AHR (U), lung eosinophils (V), and IL-4⁺ and IL-5⁺ CD4 T cells [(W) and (X)]. FSC, Forward scatter; GAPDH, Glyceraldehyde-3-phosphate dehydrogenase.

To assess the cell-intrinsic role of CD4 T cells independent of other lymphoid and myeloid immune cells, we adoptively transferred Alt-sensitized splenic CD4⁺ T cells from Cbx7^+/+ and Cbx7^−/− mice to Rag2^−/−γC^−/− mice lacking T, B, and natural killer cells and ILCs. Recipient mice were intranasally challenged with Alt (Fig. 3T) and assessed for AHR and IL-5 and IL-13 production by CD4 T cells from the lung digest and eosinophil influx into the airways. Consistent with our previous results, mice receiving Cbx7^−/− CD4⁺ T cells showed reduced AHR and type-2 cytokine production by CD4 T cells and decreased eosinophils in the BAL (Fig. 3, U to X). These data suggest that a T cell–intrinsic CBX7 defect alone can impair type-2 cytokine production, eosinophilic inflammation, and AHR.

CBX7 directly binds to the proximal promoters, conserved noncoding sequences, and the enhancer region of the cytokine genes in activated lymphocytes

Results so far show that CBX7 positively regulates cytokine production in lymphoid cells. To explore its mechanism, we examined the binding of CBX7 to the promoters and regulatory regions of the cytokine genes in activated CD4 T cells by chromatin immunoprecipitation (ChIP) assay. In our initial screening, we examined multiple sites as shown in fig. S4A. Some of the sites (IL4p3, IL4p4, IL4p5, IL5p2, and CNS3) did not show any significant binding and were not used in subsequent experiments. We also excluded the sites that nonspecifically bound the control immunoglobulin G (IgG). The ChIP assay with CD3/28-activated CD4 T cells showed direct binding of CBX7 to the proximal promoter regions of IL4, IL5, GATA3, and IFNG genes (Fig. 4A). Similarly, CBX7 bound to the conserved noncoding sequence (CNS) of the type-2 gene cluster and the IL17 gene. The binding of CBX7 to the CDKN2A (p16) gene was used as a positive control (Fig. 4B) (23). These results demonstrate that CBX7 directly binds to the promoter and regulatory sequences of cytokine genes and drives their expression in activated lymphoid cells. CBX7 is recruited to H3K27me3, a repressive chromatin marker. We examined the epigenetic modification of the foregoing cytokine gene loci by performing ChIP with antibodies against H3K27me3, H3K4me3 (a chromatin marker for transcriptionally active genes), and H3K27ac (a poised enhancer marker). We observed a significant enrichment of H3K27me3 and H3K4me3 markings at the site of the cytokine promoter and regulatory sequences (Fig. 4C). As anticipated, we observed no H3K27ac marking of the promoter sites and modest H3K27ac marking of the CNS2 site (Fig. 4C). The dual chromatin marking with H3K27me3 and H3K4me3 is considered a sign of bivalent genes that are in a poised state (25). Of interest, GATA3 showed selective marking with H3K4me3 (Fig. 4D). Conversely, IL17CNS2, a conserved CNS, showed selective marking with H3K27me3 (Fig. 4E).

Fig. 4. CBX7 binds to cytokine gene promoters and interacts with transcriptional coactivators.

(A and B) Blood CD4 T cells were stimulated with anti-CD3/28 antibodies for 72 hours. ChIP-grade anti-CBX7 and control rabbit IgG were used to immunoprecipitate the bound chromatin. DNA was used for qPCR using primers for the proximal promoter, the enhancer region, and the conserved CNSs for type-2 genes. Fold enrichment is shown as the percentage of input (1:100 dilution), *P ≤ 0.05 (n = 3). (B) Primers from the promoter of the p16 gene were used as a positive control. (C to E) Distribution of various histone marks on cytokine genes and regulatory regions. ChIP-grade H3K27me3, H3K4me3, H3K27ac, or con-IgG were used for ChIP of CD3/28-stimulated CD4 T cells (n = 3). (F) Interaction of CBX7 with P-300, CK2-α, and Pol-II. Cell lysate from stimulated CD4 T cells was immunoprecipitated with an anti-CBX7 or control IgG (con) antibodies and Western blotted for P-300, Pol-II, and CK2-α (n = 3). IP, immunoprecipitation; LE, long exposure; SE, short exposure; WCL, whole cell lysate. (G and H) Cell lysate from stimulated CD4 T cells was immunoprecipitated with an anti-P300 antibody, Western blotted for CBX7 and P-300, and reprobed for acetyl lysine. (I) Primary HBECs were cultured with and without EGF (10 ng/ml) for 48 hours. The cell lysate was immunoprecipitated with the anti-CBX7 antibody and Western blotted for Pol-II, CK2-α, and MEK2. n = 3. IgG-H, IgG heavy chain; IgG-L, IgG light chain.

CBX7 selectively interacts with the transcriptional activators CK2, acetylated P300 (an acetyl transferase), and polymerase II in lymphoid cells

Next, we explored CBX7 interaction with transcriptional coactivators. CD4 T cells from patients with asthma were stimulated with anti-CD3/28 antibodies for 48 hours. We immunoprecipitated CBX7 and probed for its interaction with transcriptional activators by Western blotting. We found that CBX7 interacted with polymerase II (Pol-II), P-300, and CK2-α (Fig. 4F) in activated CD4 T cells. We confirmed the interaction between CBX7 and P-300 by reverse coimmunoprecipitation (Fig. 4G). Autoacetylation of P-300 enhances its acetyltransferase activity, binding of proteins, and transcription-enhancing activities (26). We analyzed the acetylation status of CBX7-bound P-300 and observed that CBX7 bound to acetylated P-300 (Fig. 4H). Next, we examined the effect of the CBX7 inhibitor on the acetylation of histone-3 at Lys²⁷ (H3K27ac) in CD3/28-stimulated CD4 T cells, which is a sign of open chromatin and a prerequisite for transcriptional activation. The anti-CD3/28 stimulation of CD4 T cells resulted in increased level of H3K27ac compared to unstimulated CD4 T cells (fig. S4B). CBX7 inhibitor–treated cells showed inhibition of acetylation of histone-3 compared to the Veh-treated cells. The CBX7 inhibitor variably affected the global acetylation status of proteins in the cell. We investigated whether the CBX7 interaction with the aforementioned molecules was specific for lymphoid cells. We immunoprecipitated CBX7 from epidermal growth factor (EGF)–treated or untreated primary human bronchial epithelial cells (HBECs) and probed its interaction with Pol-II and CK2-α by Western blotting. We could not detect these interactions in HBECs (Fig. 4I). These results suggest that CBX7 functions as an inducer of gene transcription in lymphoid cells.

CBX7 and MEK2 physically interact, colocalize, and positively regulate each other

We and others have previously reported that the ERK activator MAPK kinase 1/2 (MEK1/2) translocates to the nucleus and regulates gene transcription in lymphocytes (27–29). Severe asthma is frequently associated with steroid resistance (27, 28, 30). MEK2 and CBX7 were identified as steroid-resistant genes in a genome-wide screening of steroid-resistant leukemic cells (31). For this reason, we explored the interaction of CBX7 with MEK2. We studied the colocalization of CBX7 with MEK2 in human blood CD4 T cells 48 hours after stimulation. Upon anti-CD3/28 stimulation, a significant amount of CBX7 colocalized with MEK2 in the nucleus (Fig. 5A). Likewise, CBX7 physically interacted with MEK2 upon anti-CD3/28 stimulation as demonstrated by coimmunoprecipitation studies (Fig. 5B). We examined the mutual regulation of CBX7 and MEK. The expression of mRNA for Mek1 and Mek2 in the lung from Cbx7^−/− mice was significantly less compared to that from Cbx7^+/+ mice (Fig. 5C). Next, we examined the effect of pharmacological inhibitors on CBX7 and MEK2 levels in human lung ILCs, which were isolated from cadaveric donors. Inhibition of CBX7 and MEK by MS37542 and trametinib, respectively, resulted in decreased frequency of MEK2⁺ and CBX7⁺ human lung ILCs (Fig. 5, D to F). These results prompted us to examine whether MEK1/2 regulates CBX7 interaction with the transcriptional activators. The MEK inhibitor trametinib significantly reduced the interaction of CBX7 with P-300 and Pol-II (Fig. 5G).

Fig. 5. CBX7 and MEK2 physically interact with each other.

(A) Blood CD4 T cells from an asthmatic patient were stimulated with anti-CD3/28 antibodies for 48 hours and immunostained for CBX7 and MEK2. The overlay (merged) confocal image shows colocalization (yellow) in the nucleus (n = 3). (B) Coimmunoprecipitation of CBX7 with MEK2. Cell lysate from stimulated blood CD4 T cells was immunoprecipitated with an anti-MEK2 antibody and Western blotted for CBX7 (n = 3). (C) mRNA expression of Mek1 and Mek2 in the lung tissue from Cbx7^+/+ or Cbx7^−/− mice (n = 5). (D to F) Effect of the MEK inhibitor trametinib and the CBX7 inhibitor MS37452 on lung CBX7⁺ ILCs and MEK2⁺ ILCs, respectively. Lin⁻CD45⁺ cells were isolated from a cadaveric human lung of a healthy donor and cultured with the MEK or CBX7 inhibitors for 72 hours. CBX7⁺/MEK⁺ cells were quantified by FCM (n = 4). (G) CD4 T cells were stimulated with medium or anti-CD3/28 antibodies for 36 hours and then treated with DMSO (Veh) or the MEK inhibitor trametinib (MEK-inh) for 12 hours. The cell lysate was immunoprecipitated with an anti-CBX7 antibody and Western blotted for P-300 and Pol-II (n = 3). (H) Cytosolic translocation of CBX7 and colocalization with MEK2. Blood CD4 T cells were stimulated with anti-CD3/28 antibodies for the indicated time points and immunostained for CBX7 and MEK2. (I) Subcellular localization of CBX7. Subcellular fractions of stimulated CD4 T cells were Western blotted. Subcellular compartment markers are shown on the right-hand side. (J) A comparison of the two isoforms of CBX7 in nonstimulated and stimulated CD4 T cells. ATPase, adenosine triphosphatase.

CBX7 translocates to the cytoplasm upon TCR ligation

Some nuclear localized repressors translocate to and function in the cytosol (20, 32). We asked whether CBX7 had cytosolic activities. We analyzed the spatiotemporal location of CBX7 by immunofluorescence staining and confocal microscopy following TCR stimulation (with anti-CD3/CD28 antibodies) of blood CD4 T cells. CBX7 (stained green) was mostly localized to the nucleus at baseline. Upon CD3/28 stimulation, we observed cytosolic translocation and increased colocalization (yellow) of CBX7 (green) with MEK2 (red) at 5- and 30-min time points (Fig. 5H). In agreement with our immunofluorescence studies, subcellular fractionation of anti-CD3/28–stimulated CD4⁺ T cells revealed that CBX7 was present in the cytosolic, membranous, and nuclear fractions (Fig. 5I). Cho and colleagues recently reported two distinct isoforms of CBX7, p36 and p22, and described a preferential localization of the p22 form in the cytosol (33). We examined the expression of the CBX7 isoforms in blood lymphocytes and observed a very prominent expression of the p36 isoform as compared to the p22 isoform (Fig. 5J). We did not observe any increase in expression of the p22 isoform upon TCR stimulation. Hence, the presence of CBX7 in the cytosol upon TCR stimulation is unlikely due to enhanced expression of the p22 isoform.

CBX7 regulates ERK1/2 MAPK signaling

To investigate the function of cytoplasmic CBX7, we examined the activation of ERK1/2 in spleen CD4 T cells from Cbx7^+/+ and Cbx7^−/− mice. The activating phosphorylation level of ERK1/2 was significantly reduced in Cbx7^−/− cells (Fig. 6, A to C). The p-ERK1/2 level was also reduced in CD4⁻ cells (CD8 and other non-CD4 cells) and CD19⁺ B cells from Cbx7^−/− mice (fig. S5, A to F). To further confirm these results, we purified CD4⁺ T cells from Cbx7^+/+ and Cbx7^−/− mice and examined the activation of MEK1/2 and ERK1/2 by Western blotting. As shown in Fig. 6D, the activation of MEK1/2 and ERK1/2 were compromised in CD4⁺ T cells from Cbx7^−/− mice. The phosphorylation status of the downstream components of MEK-ERK signaling such as MSK1 and RSK90 and the upstream activator c-Raf was also reduced in Cbx7^−/− cells. LCK was phosphorylated at baseline in Cbx7^−/− cells, perhaps as a compensatory mechanism (Fig. 6D). We examined phosphorylation of ERK1/2 ex vivo. CD4⁺ T cells from BAL obtained from Cbx7^−/− mice, sensitized and challenged with allergen as indicated in Fig. 3A, showed reduced ERK1/2 activation compared to CD4⁺ T cells from Cbx7^+/+ mice (Fig. 6, E to G). We further confirmed the role of CBX7 in ERK1/2 signaling by pharmacological inhibition of CBX7 in vitro in human blood and BAL CD4 T cells. The CBX7-inh–treated cells showed reduced activation of ERK1/2 (Fig. 6, H and I). These results demonstrate a critical role of CBX7 in ERK1/2 signaling in the cytosol.

Fig. 6. CBX7 positively regulates ERK1/2 signaling.

(A to C) Comparison of phospho-ERK1/2 (p-ERK) expression in splenocytes from Cbx7^+/+ and Cbx7^−/− mice. Splenocytes were stimulated with anti-CD3/28 antibodies as indicated, stained with antibodies against CD4 and p-ERK1/2, and analyzed by FCM (A). [(B) and (C)] The panels show cumulative data for % p-ERK⁺ CD4 T cells and their geometric mean florescence intensity (MFI), n = 3. (D) Representative Western blots of CD4 T cells for phosphorylated MAPK signaling pathway proteins. Purified CD4 T cells from Cbx7^+/+ or Cbx7^−/− mice were stimulated with anti-CD3/28 antibodies as indicated and Western blotted for phosphorylated and total (up- and downstream) ERK1/2 pathway signaling molecules, n = 3. (E) A representative flow cytogram showing p-ERK⁺ CD4 T cells in BAL from Alt-sensitized and challenged Cbx7^+/+ and Cbx7^−/− mice as indicated in Fig. 3A. (F and G) Cumulative data for percentage and MFI of p-ERK⁺ CD4 T cells from BAL, n = 5. (H and I) The effect of the CBX7 inhibitor on ERK1/2 activation (p-ERK1/2) in CD4 T cells from the human blood and BAL. Human blood or BAL cells were pretreated with 200 μM of the CBX7-inh or Veh for 4 hours followed by anti-CD3/28 antibody stimulation as indicated. p-ERK⁺ CD4 T cells were quantified by FCM. p-ERK levels were normalized to respective Veh-treated samples at each time point (n = 5).

CBX7 is required for a methylation-dependent molecular complex formation with MEK2 and c-Raf

Since CBX7 interacted with MEK1/2 and promoted ERK1/2 signaling, we examined CBX7 interaction with c-Raf, an upstream activator of MEK1/2. Coimmunoprecipitation studies demonstrated that CBX7 interacted with c-Raf and MEK1/2 (Fig. 7A). Since CBX7 is a methyl reader and binds to methylated histone, we explored the possibility that MEK1/2 and c-Raf were methylated, which enabled their interaction with CBX7. We immunoprecipitated nonstimulated or anti-CD3/28–stimulated (24 hours) CD4 T cells with anti-MEK2 antibody. Immunoprecipitated MEK2 was methylated as demonstrated by Western blotting for methyl-lysine (methyl-K) (Fig. 7B). Immunoprecipitation with the methyl-K antibody confirmed MEK2 methylation. We also found c-Raf in the methyl-K immunoprecipitate (Fig. 7C). STAT3 (signal transducer and activator of transcription 3) undergoes methylation and served as a positive control for the experiment (34). Next, we immunoprecipitated c-Raf and confirmed its methylation by Western blotting for methyl-K (Fig. 7D). We confirmed c-Raf–CBX7 interaction (Fig. 7A) by reprobing the c-Raf immunoprecipitation, which showed that c-Raf was methylated and interacted with CBX7 (Fig. 7D). To examine the importance of methylation for protein-protein interaction in the ERK signalosome, we treated anti-CD3/28–stimulated CD4 T cells with the EZH2 methyl transferase inhibitor, GSK503. GSK503 significantly reduced c-Raf methylation (Fig. 7E). This inhibition of methylation was sufficient to completely inhibit the colocalization of MEK2 with CBX7 as demonstrated by immunofluorescence staining and confocal microscopy (Fig. 7, F and G). Since CBX7 binds to methylated proteins, we explored the possibility of CBX7 functioning as an adaptor to stabilize the multimolecular complex of MEK1/2 and c-Raf upon TCR stimulation. To test this, we used the cell lysate from anti-CD3/28–stimulated CD4 T cells obtained from Cbx7^+/+ and Cbx7^−/− mice for immunoprecipitation. Immunoprecipitates of MEK2 (Fig. 7H) and c-Raf (Fig. 7I) from Cbx7^−/− cell lysates showed reduced interaction among the foregoing molecules. c-Raf was shown to be phosphorylated by CK2-α (35). We showed that CBX7 interacted with CK2-α (Fig. 4D). The c-Raf interaction with CK2-α was attenuated in Cbx7^−/− T cells (Fig. 7I). The foregoing results suggested that CBX7 functioned as an adaptor to form and stabilize the signaling complex for optimal and durable signaling. To validate the relevance of methylation-facilitated interaction of CBX7 with the ERK1/2 signalosome in the airway tissue from asthmatic patients, we immunostained endobronchial biopsy samples from patients with uncontrolled, severe persistent asthma and disease controls. We found increased expression of CBX7, MEK2, and methyl-K in infiltrating peribronchial inflammatory cells and their colocalization in the tissue from patients with asthma as compared to disease controls (Fig. 7J and fig. S6B). The interaction of CBX7 with MEK2 and the methylation of the latter were specific for lymphoid cells as they were not detected in EGF-treated HBECs (Figs. 4H and 7K).

Fig. 7. CBX7 interacts with methylated MEK2 and c-Raf.

(A) Interaction of CBX7 with c-Raf. Mouse splenic CD4 T cells were stimulated with anti-CD3/28 antibodies for 24 hours, immunoprecipitated with an anti-CBX7 antibody, and Western blotted for c-Raf and MEK1/2. (B and C) Methylation of MEK2. Cell lysate from nonstimulated and stimulated CD4 T cells was immunoprecipitated with a MEK2 antibody and Western blotted for methyl-K (n = 3). (D) Interaction of methylated c-Raf with CBX7. Cell lysate from stimulated CD4 T cells was immunoprecipitated with a c-Raf antibody and Western blotted for methyl-K and CBX7. (E) Effect of the methyl transferase EZH2 inhibitor GSK503 on c-Raf methylation. Stimulated CD4 T cells were treated with DMSO (Veh) or GSK503 (20 μM) for 12 hours, immunoprecipitated with a c-Raf antibody, and Western blotted for methyl-K. (F) Effect of the EZH2 inhibitor on colocalization of CBX7 (green) with MEK2 (red). Human blood CD4 T cells were treated with DMSO or GSK503 (20 μM) for 4 hours, stimulated with anti-CD3/28 antibodies (Stim) for the indicated time and immunostained for CBX7 and MEK2. (G) The bar graph shows quantification of colocalization (n = 3). (H and I) CBX7 stabilizes MEK and c-Raf interaction. Splenic CD4 T cells from Cbx7^+/+ and Cbx7^−/− mice were stimulated with anti-CD3/28 antibodies for 24 hours, immunoprecipated with anti-MEK1/2 and c-Raf antibodies, and Western blotted for c-Raf and MEK1/2, respectively. The c-Raf immunoprecipitate was additionally Western blotted for CK2-α. (J) Immunostaining of endobronchial biopsy specimens from a patient with asthma and a disease control subject for CBX7 (red), MEK2 (magenta), and methyl-K (green). Insets show the enlarged view of the marked area of the image. The side graphs show colocalization coefficient of the indicated immune-staining pair and intensity of CBX7 expression (n = 5, 10 cells per sample). (K) Absence of methylated MEK2 in stimulated airway epithelial cells. Primary HBECs were cultured with and without EGF (10 ng/ml) for 48 hours, immunoprecipitated with an anti-MEK2 antibody, and Western blotted for trimethyl lysine and CBX7. ut, untreated.

CBX7 deficiency causes a genome-wide transcriptional loss and gain in T cells

To assess the role of CBX7 as a transcriptional activator in T cells, we performed RNA sequencing (RNA-seq) of anti-CD3/28–stimulated spleen CD4 T cells from Cbx7^+/+ and Cbx7^−/− mice. Cells were stimulated for 6 hours. A differential gene expression analysis showed 1593 and 1270 genes were down- and up-regulated in Cbx7^−/− T cells. The down-regulation of a large number of genes validated that CBX7 functioned as an inducer of gene transcription. Some of the top down-regulated genes included cytokines (Il13, Il4, Il2, Ifng, and Csf2), chemokines (Ccl1, Ccl3, and Ccl4), transcription factors (Tbx21), adhesion molecules such as CD160, and interferon (IFN)–regulated genes (Isg15, Isg20, Oas3, and Ifih1) (Fig. 8, A and B). We validated the RNA-seq data for the type-2 cytokine genes, Il4, Il5, and Il13, in purified CD4 T cells by qPCR (Fig. 8, C to E). Some of the top up-regulated genes included Synj2 (synaptojanin-2), Erdr1 (erythroid differentiation regulator 1), Acss2 (acetyl coenzyme synthetase-2), and St8sia1 (also known as gangliosidase 3 synthase). Synj2 is important for transforming growth factor–β signaling. Erdr1 induces T cell apoptosis through Fas/Fasl. Acss2 is important for autophagy and lysosomal biogenesis. St8sia1 inhibits Jak/Stat signaling. We confirmed the up-regulation of these molecules in Cbx7^−/− cells by analyzing PBMCs from asthmatic patients treated with CBX7-inh for 24 hours. Inhibition of CBX7 up-regulated the mRNA expression of ERDR1, SYNJ2, and ACSS2 (Fig. 8, F to H). The likely outcome of their increased expression is T cell inhibition, which needs to be addressed in the future. A Kyoto encyclopedia of genes and genomes (KEGG) pathway analysis of the bulk RNA-seq data revealed that pathways related to asthma, IL-17 signaling, T_H1 and T_H2 differentiation, and TCR signaling were down-regulated in stimulated CD4 T cells from Cbx7^−/− mice (Fig. 8, I and J).

Fig. 8. Results of bulk RNA-seq of CD4 T cells.

(A) Spleen CD4 T cells from Cbx7^+/+ and Cbx7^−/− mice were stimulated with anti-CD3/28 antibodies for 6 hours. RNA was isolated and processed for bulk RNA-seq. The volcano plot shows down- and up-regulated genes in CD4 T cells from CBX7 KO and WT mice. (B) A heatmap of top 50 DEGs (differentially expressed genes) of stimulated CD4 T cells from WT and CBX7 KO mice. (C to E) Validation of the RNA-seq data for select type-2 genes by qPCR using RNA from stimulated CD4 T cells (n = 6). (F to H) Effect of the CBX7 inhibitor MS37452 on expression of mRNA for ERDR1, SYNJ2, and ACSS2 in PBMCs from asthmatic donors (n = 10). (I and J) KEGG pathway enrichment analysis of differentially expressed genes using Enrichr. The bar charts show the most relevant and significantly enriched terms. The x axis represents the −log₁₀ of the enrichment P value. The y axis represents the enriched terms in KEGG databases. KEGG pathway enrichment analysis of differentially expressed genes were performed using Enrichr under default parameters (49).

DISCUSSION

Here, we showed that the PRC protein CBX7, unexpectedly, functioned as an organizer of methylation-dependent transcriptional and signaling complexes in the nucleus and cytosol. By organizing these signaling complexes, CBX7 functioned as an inducer of gene transcription in the nucleus and a facilitator of ERK1/2 signaling in the cytosol. CBX7 directly bound to the promoter and enhancer sites of key cytokines, IL-4, IL-5, IL-17A, and IFN-γ, and the type-2 transcription factor GATA3. CBX7 recruited the transcriptional activators Pol-II, acetylated P300, and CK2-α in activated lymphocytes (12), which helped explain its function as an inducer of gene transcription in lymphoid cells. CBX7 formed a multimolecular complex with the ERK1/2 upstream kinases c-Raf and MEK in a methylation-dependent manner and facilitated their activation. The foregoing activating functions were specific for lymphoid cells and did not occur in nonlymphoid cells, where CBX7 acted as a repressor. We validated the importance of CBX7 as a positive regulator of cytokine production in lymphoid cells using a pharmacologic inhibitor, shRNA-mediated knockdown, and germline ablation of the gene. We demonstrated our findings in vitro in isolated lymphoid cells, and in vivo in two different mouse models of asthma. We confirmed our results in vitro in lymphoid cells isolated from the blood and the lung from asthmatic patients, and ex vivo in endobronchial biopsy samples from patients with asthma. Thus, we provided a previously unidentified paradigm of CBX7 function where it functioned as an activator of gene transcription in the nucleus and regulator of signal propagation in the cytosol.

CBX7 was previously studied for activities outside chromatin remodeling. It directly bound DNA and RNA sequences (36–38). CBX7 suppressed p16/CDKN2A and Ink4a/Arf (39). Some PRC proteins function independently of their repressive function. RING1, MEL18, BMI1, EZH2, Eed, and YY1 bound to the promoter and enhancer sites of IL4 and IFNG genes in T cells (15, 17). This was associated with cytokine gene transcription.

H3K4me3 and H3K27me3 mark the chromatin of genes that are transcriptionally active and repressed, respectively (25). H3K4me3 leads to an open conformation of the gene locus. In contrast, H3K27me3 induces a compact chromatin configuration leading to an inactive gene locus. Unexpectedly, our ChIP studies demonstrated simultaneous H3K4me3 and H3K27me3 chromatin markings of IL4 and IFNG promoters. H3K27Ac marks poised and active gene enhancers. We observed minimal H3K27Ac marking of the type-2 cytokine enhancer site (CNS2) and, expectedly, negligible H3K27Ac marking of the promoter sites. Dual H3K4me3 and H3K27me3 marking is considered a sign of bivalent promoters that are poised for activation (40). Bivalent promoters are characteristically found in high numbers in stem cells and cancer cells. Human and mouse embryonic stem cells have been reported to have 4979 and 3659 bivalent promoters, respectively (41). Another study reported 23,167 bivalent domains in the mouse brain tissue. Naïve CD8 T cells contain 589 bivalent genes (42). We speculate that CBX7 is active in bivalent domains in T cells.

CBX7 is primarily localized to the nucleus. A recent study has identified a 22-kDa isoform of CBX7 that is localized to the cytosol (33). We examined its expression in lymphocytes and observed negligible expression when compared to the expression of its nuclear isoform (Fig. 5H). The nuclear isoform translocated to the cytosol upon T cell activation. Of note, EZH2, a PRC2 protein, was found to localize to the cytosol of lymphocytes and was important for TCR signaling (21).

We demonstrated methylation of c-Raf and MEK1/2 in lymphocytes. We believe that CBX7, as a methyl reader (36), bound methylated c-Raf and MEK1/2 and facilitated the formation of a multimolecular complex. CBX7 interacted with CK2-α, a protein kinase known to phosphorylate and activate Raf kinases (35). We believe that the CBX7 recruitment brought CK2-α to the ERK1/2 signalosome. CK2-α phosphorylated c-Raf. The latter initiated the downstream cascade of phosphorylation of MEK1/2 and ERK1/2. A previous study demonstrated a critical role for EZH1/2 in ERK1/2 and MEK1/2 phosphorylation and TCR-mediated T cell activation (21). EZH1/2 is known to methylate actin (43). In this study, we showed that the EZH1/2 inhibitor GSK503 prevented TCR-mediated c-Raf methylation in lymphocytes. We believe that EZH1/2 methylation of c-Raf led to CBX7 recruitment to the ERK1/2 signalosome. Methylation of proteins facilitates hydrophobic interactions and increases their stability and function (44). One plausible explanation for this effect is that methylation of lysine prevents its ubiquitination and degradation (45). Inhibition of methylation impaired the recruitment of CBX7 and CK2-α to c-Raf and MEK. Thus, our studies demonstrated a previously unidentified methylation-dependent mechanism of activation of the ERK1/2 signaling cascade.

CBX7 plays an important role in allergic inflammation and AHR in the mouse model of asthma. We demonstrated this role in vivo using two different approaches—germline genetic ablation and pharmacologic inhibition. CBX7 was important for an ILC2 as well as a T_H2/ILC2 models of asthma. This effect was likely due to reduced type-2 cytokine production by ILC2s and T_H2 cells, reduced T cell number, and impaired locomotion. Our RNA-seq data showed that Cbx7^−/− CD4 T cells had reduced expression of type-2 as well type-1 and type-3 cytokines, which was predictable from our ChIP studies. CBX7 functions as a repressor in many stromal and mucosal cells including epithelial cells and smooth muscle cells (39). We anticipated heightened epithelial production of type-2 cytokines and smooth muscle hypertrophy resulting in increased AHR. Instead, we observed decreased AHR and remodeling. Epithelial Il33 and Tslp production was not increased in Cbx7^−/− lungs and was comparable to that seen in Cbx7^+/+ lungs. The foregoing results suggested that the absence of the repressor function in Cbx7^−/− mucosal and stromal cells did not grossly affect the asthma phenotype that resulted from the reduced type-2 immune response in our experimental models. Adoptive transfer experiments suggested a critical role of CD4 T cell–intrinsic Cbx7 for induction of allergic inflammation. We observed elevated airway neutrophils in Cbx7^−/− mice. There was a concurrent modest but significant increase in hematopoietic cell death. We speculate that this increase in cell death contributed to elevated neutrophils. IL-17 pathway molecules regulate neutrophilic inflammation. The Il17 gene was positively regulated in T cells by CBX7. However, IL-17 could come from other sources including γδ T cells and nonlymphoid cells, which could also explain the increase in neutrophils.

In summary, CBX7, known primarily for its chromatin remodeling activity, functioned as an inducer of gene transcription in the nucleus and a methylation-dependent organizer of the ERK1/2 signalosome in the cytosol (Fig. 9). Through gene transcription and ERK1/2 signaling, CBX7 positively regulated ILC2- and T_H2-dependent allergic inflammation of the airways. Our studies established a previously unidentified paradigm of methylation-driven transcriptional and signaling complexes regulating immune cell function and inflammatory processes. Multiple pharmacological inhibitors of CBX7 have been identified. We demonstrated the efficacy of the CBX7 inhibitor MS37452 in inhibiting inflammatory cytokine production by mouse and human lymphoid cells in vitro and ex vivo, and in blocking experimental asthma in mice in vivo. The foregoing findings establish the foundation for clinical trials of CBX7 inhibitors in human asthma and other inflammatory conditions.

Fig. 9. A schematic diagram illustrating the molecular mechanisms of CBX7 function in CD4 T cells.

Left: During TCR activation, CBX7 recruits transcriptional activators to promoters and regulatory regions of inflammatory genes, driving their transcription. Simultaneously, a subset of CBX7 translocates to the cytoplasm, where it stabilizes the ERK1/2 signaling complex by binding methylated signaling intermediaries, leading to enhanced ERK1/2 activation. Activated ERK1/2 then translocates to the nucleus to further amplify gene transcription via a positive feedback loop. Right: Loss of CBX7 impairs this process.

MATERIALS AND METHODS

Study design

The goal of this study was to determine the role of CBX7 in lymphoid cells and decipher the molecular mechanism of its function in allergic inflammation in human asthma and in mouse models of asthma. The experimental design involved human blood, BAL cells, isolated lung cells from human cadaveric lungs, and age-matched evaluation of Cbx7^+/+ and control mice paired with FCM, immunofluorescence, immunoprecipitation, bulk RNA-seq, shRNA-mediated gene knockdown, target specific pharmacological inhibitors, and cellular and biochemical analyses. Blinding was not used in this study. The health status of each mouse was monitored daily, and mice that met the predefined study endpoints were humanely euthanized. Numbers of mice and replicates are indicated in figure legends. Few samples were excluded from the analysis, where we could not retrieve the tissue or could not detect the expression of investigated molecules. For other assays, at least two to three independent experiments were performed, with biological replicates or triplicates.

Mice

All animal studies were performed in compliance with the protocol approved by National Jewish Health Institutional Animal Care and Use Committee (protocol no. AS2614). We generated Cbx7^−/− mice in the C57BL/6N strain in our mouse genetics core facility using Cbx7^−/− sperm [Cbx7^{tm1a(KOMP)Wtsi}] from MMRC. We used Cbx7^−/− and sex- and age-matched Cbx7^+/+ littermate WT controls as well as Rag1^−/− (B6.129S7-Rag1tm1Mom/J) and Rag2^−/−γC^−/− mice in our experiments.

Human subjects

Whole blood and BAL fluid were obtained, upon written informed consent, from allergic asthma and disease control patients, and healthy donors recruited at National Jewish Health. Allergic sensitivity was defined by positive skin test or the presence of IgE antibody against environmental allergens. Asthma was defined by the American Thoracic Society criteria (the presence of reversible airway obstruction and/or a positive Mch test). Uncontrolled asthma was defined by the Expert Panel Report 4 (EPR4) guidelines. Disease controls were patients with vocal cord dysfunction, gastroesophageal reflux disorder, chronic cough, urticaria, tracheobronchomalacia, and chronic aspiration. The human subject study was approved by the Institutional Review Board at National Jewish Health. The demographic and diagnostic characteristics of healthy donors and patients with asthma were detailed in table S1.

Isolation of PBMCs, Lin⁻ cells, culture, and FCM

PBMCs were isolated from EDTA-treated blood by density centrifugation with Histopaque (Sigma-Aldrich, 10771). Lin⁻ cells were isolated from PBMCs by negative selection using antibody-coated magnetic beads (lineage cell depletion kit, no. 130-092-211) from Miltenyi, Inc., San Diego, CA. PBMCs or Lin⁻ cells were cultured in RPMI 1640 plus 10% fetal bovine serum (FBS), with 1× penicillin and streptomycin for the indicated time mentioned in text. The following cytokines were used at 20 ng/ml concentration: IL-2 (no. 200-02) and IL-33 (no. 200-33) (all from PeproTech, Rocky Hills, NJ). PBMCs and Lin⁻ cells were cultured for 3 days unless otherwise stated in the text. Monensin (2 μM, BioLegend, Inc., San Diego, CA) was added to all cultures 4 hours before conclusion. The culture cells were pelleted and then stained for FCM.

Processing and culture of BAL cells for ILC2 studies

BAL cells were pelleted, washed, and cultured in 10% FBS containing RPMI medium either with the Veh control or the CBX7 inhibitor MS37452 for 72 hours followed by incubation with monensin (2 μM) for 4 hours, and then stained for FCM.

Lentiviral transduction for CBX7 knockdown in PBMCs

The lentiviral packaging plasmid psPAX2 (no. 12260) and the envelope plasmid pMD2G (no. 12259) were purchased from Addgene (MA, USA). shERWOOD UltramiR lentiviral vectors that harbored an shRNA sequence against human CBX7 and encoded ZsGreen (GFP) (CBX7 shRNA) and corresponding control lentiviral vectors that encoded ZsGreen (con-shRNA) were obtained from transOMIC Technologies (Huntsville, AL). The viral particle generation and subsequent transduction in PBMCs were done as mentioned previously (46).

Asthma model

For an acute conventional asthma model, 6- to 8-week-old female Cbx7^−/− and Cbx7^+/+ mice were intranasally exposed to 20 μg of the Alternaria extract on days 1 to 5, rested for 3 weeks, and subsequently challenged intranasally with three doses of the Alternaria extract as depicted in the Fig. 3A. Animals were rested for 2 days after the last challenge and then examined for AHR and lung inflammation. For studies involving the CBX7 inhibitor, we used an innate (ILC-dependent) asthma model (five exposures of the allergen followed by endpoint studies on day 8). Alternaria was previously shown to preferentially activate ILCs during this short exposure protocol (47, 48). One group of WT and Rag1^−/− mice received a daily dose of 15.95 mg/kg of CBX7 inhibitor 1 hour prior to intranasal administration of the Alternaria extract (10 μg per dose) for 5 consecutive days. Mice were rested for 2 days and subsequently examined for allergic inflammation.

For adoptive transfer experiment, 5 million splenic CD4 T cells from allergen-sensitized (10 μg intranasally for three doses per week for 2 consecutive weeks followed by isolation of CD4 T cells on day 14) Cbx7^+/+ and Cbx7^−/− mice were intravenously transferred into Rag2^−/−γC^−/− mice, and the recipient mice were challenged with allergen as described above for the acute asthma model.

AHR measurement

AHR was measured in response to Mch as described previously (22). Briefly, 72 hours after the last allergen dose, mice were deeply anesthetized by intraperitoneal injection of ketamine (180 mg/ml), xylazine (9 mg/ml), and acepromazine (4 mg/ml) prior to tracheotomy and measurement of AHR by FlexiVent (Scireq). Tracheotomozied mice were connected to a small animal ventilator by an 18-gauge cannula and were exposed to increasing Mch aerosol concentrations (6.25, 12.5, 25, 50, and 100 mg/ml). Baseline resistance was measured with aerosolized saline. Group averages were expressed as the fold increase over baseline resistance. Statistical significance was determined using Student’s t test.

Histology and morphometric analysis

Formalin-fixed and paraffin-embedded 5-μm lung tissue sections were deparaffinized using citrisolv and rehydrated with gradient alcohol. Sections were stained with H&E, periodic acid–Schiff (PAS) reagent, or Masson’s trichrome for the evaluation of inflammatory cell influx, mucus staining, and collagen deposition. Images were captured on a Nikon microscope and analyzed using the Metamorph software as described previously (46).

ELISA

Culture supernatant from PBMCs and the cell-free BAL fluid from Cbx7^+/+ and Cbx7^−/− mice were collected and analyzed using specific ELISA kits for human IL-5 and IL-13 (R&D DuoSet, kit catalog no. DY205 and no. DY213) and for mouse IL-4, IL-5, and IL-13 (R&D DuoSet, catalog no. DY404, no. DY405, and no. DY413) according to the manufacturer’s instructions.

Lung digestion for isolation of single-cell preparations

Mouse lungs were perfused with saline and then subjected to mechanical mincing followed by digestion at 37°C for 40 min with Liberase (50 μg/ml) (Roche, no. 05401127001) and deoxyribonuclease (1 μg/ml) in RPMI containing 1% FBS and 1% penicillin/streptomycin with intermittent shaking. The enzymatic reaction was stopped with the addition of 2 mM EDTA. Isolated cell suspensions were filtered through 70-μm filters followed by red blood cell lysis with ACK cell lysis buffer. Lysed cell membrane fragments were removed by filtration through a 40-μm filter. Cells were washed twice with 2% FBS containing Hanks’ balanced salt solution and immediately used for staining for FCM and analysis.

Flow cytometry

For intracellular staining, cells were incubated for 4 hours with monensin and brefeldin A (2 μM). Aliquots of 1 × 10⁶ cells were collected and washed with ice-cold phosphate-buffered saline (PBS) followed by staining with a viability dye (Zombie Aqua Fixable Viability Kit, no. 423101) to exclude dead cells. After live/dead cell staining, cells were incubated with an Fc blocker in fluorescence-activated cell sorting buffer for 15 min at 4°C followed by cell surface staining with fluorophore-conjugated antibodies against specific cell surface markers for 30 min at 4°C. We used the BD Cytofix/Cytoperm Fixation/Permeabilization Kit (catalog no. 554714) for cytokines and the Fixation/Permeabilization Buffer Set from BioLegend (catalog no. 425401) for p-ERK staining. The latter was performed with fluorescently labeled monoclonal antibodies. Data were acquired using either LSRII or LSRFortessa (BD Biosciences, San Jose, CA) flow cytometer and analyzed using FlowJo v10 software. Antibodies used for surface and intracellular staining were purchased from BioLegend, Inc. (San Diego, CA). Others were from eBioscience or R&D, Inc., unless otherwise stated. The complete list of antibodies used for FCM is mentioned in table S1.

Real-time PCR

Total RNA was isolated from frozen lung samples stored in RNAlater or anti-CD3/28–stimulated CD4 T cells using RNeasy Plus Kit (no. 74134, QIAGEN), as per the manufacturer’s instructions. Aliquots of 0.5 to 1 μg of RNA per reaction were used to synthesize cDNA using the Verso cDNA synthesis kit (Thermo Fisher Scientific, no. AB-1453/B) according to the manufacturer’s instructions. Gene-specific PCR products were amplified using the qPCR SYBR Green Rox mix (Thermo Fisher Scientific, no. AB-4162/B). The levels of target gene expression were normalized to 18S expression using the 2^−ΔCt method. The primers used to quantitate expression of various genes are mentioned in table S1.

CD4⁺ T cell isolation and proliferation

Human and mouse CD4 T cells were purified using human CD4 T cell isolation kit (no. 17952) and the mouse CD4 T cell isolation kit (no. 19852A) from STEMCELL Technologies Canada Inc., Vancouver, Canada. Purified CD4 T cells were either cultured unstimulated or stimulated with plate-bound anti-CD3/28 antibodies in 10% FBS containing RPMI for the indicated period of time.

Immunoprecipitation and Western blotting

For coimmunoprecipitation, 10 × 10⁶ CD4 T cells per reaction were washed with ice-cold PBS and then lysed at 4°C for 30 min in lysis buffer containing 25 mM tris-HCl (pH 7.4), 150 mM NaCl, 1.0% Triton X-100, 1 mM phenylmethylsulfonyl fluoride (PMSF), 0.1% bovine serum albumin (BSA), 5 mM EDTA, and protease/phosphatase inhibitor cocktail (Sigma-Aldrich, no. PPC2020). Lysates were centrifuged at 12,000g for 15 min at 4°C, and the supernatant fraction was precleared with agarose/magnetic beads for 2 hours at 4°C followed by immunoprecipitation using the required antibody or normal IgG as a control antibody overnight. For some immunoprecipitation reaction, specific antibodies/con-IgG were preconjugated with magnetic beads according to the manufacturer’s protocol (Thermo Fisher Scientific). Then, 25 μl of protein A/G plus agarose beads/magnetic beads [from Cell Signaling Technology (CST)] was added to the cell lysates for another 2 hours at 4°C. The beads containing immune complexes were washed thrice with wash buffer (20 mM Hepes, pH 7.4, 0.1% Triton X-100, 125 mM NaCl, 10% glycerol, 1 mM PMSF, and protease inhibitors), and then, the bead-bound immunoprecipitates were either eluted with 0.1 M glycine (pH 2.5) or boiled in Laemmli sample buffer for SDS–polyacrylamide gel electrophoresis (SDS-PAGE). The samples were resolved on SDS-PAGE and transferred to a nitrocellulose/polyvinylidene difluoride membrane. The membrane was blocked with 5% Blotto for 1 hour and subsequently probed overnight with the indicated primary antibodies. Membranes were washed thrice with tris-buffered saline with tween 20 (TBST) for 7 min (each wash) and probed for 1 hour with respective horseradish peroxidase–conjugated secondary antibodies and washed thrice with TBST (7 min each wash). Bands were visualized with ECL reagent, and the signal was captured on an autoradiograph. Images were scanned, and densitometry was performed using the ImageJ software. Statistical significance was determined using Student’s t test. All antibodies used for immunoprecipitation and Western blotting are listed in table S2. For MEK2/MEK1/2 immunoprecipitation, we used both polyclonal and monoclonal antibodies from three different vendors (CST, Bethyl, and BD Biosciences). We also used multiple antibodies for CBX7 immunoprecipitation. We observed similar results with most of the antibodies.

Immunofluorescence and confocal microscopy

Stimulated CD4 T cells were washed with ice-cold PBS and fixed with 3.7% formaldehyde for 10 min and then cytospun on charged glass slides. Cells were permeabilized with PBS containing 0.5% Triton X-100 and 0.05% Tween 20 for 6 min followed by blocking with 2% BSA in PBS for 1 hour at room temperature (RT). Cells were incubated overnight at 4°C with antibodies against MEK2 (MEK2, no. 610236, BD Biosciences) and CBX7 (A302-525A, Bethyl Lab., Inc.) and CBX7 (E7N1W, no. 34547, CST), followed by host-specific fluorophore-conjugated secondary antibody for 45 min at RT. We validated the specificity of anti-CBX7 monoclonal antibody (CST) that recognizes both human and mouse CBX7 using lymphoid cells from Cbx7^−/− mice in immunostaining, which we used for the experiments in Fig. 7J and fig. S6B. Cbx7^−/− did not show any staining with this antibody (fig. S6A). Cells were mounted with 4′,6-diamidino-2-phenylindole–containing mounting medium to stain the nucleus. Imaging was performed using LSM 700 Meta NLO Confocal Microscope (Carl Zeiss, Germany) on 63× oil immersion objective lens. Two to three 0.33-μm middle optical z sections were projected, and colocalization was analyzed using ZEN blue software. Images were further processed using Adobe Photoshop software.

Chromatin immunoprecipitation

ChIP was performed using the SimpleChIP Enzymatic Chromatin IP Kit (CST, 9003) as per the instructions of the manufacturer. Briefly, 10 μg of chromatin from anti-CD3/28–stimulated CD4 T cells was incubated with antibodies against CBX7 (ab21873), H3K27me3, H3K4me3, and H3K27ac or control rabbit IgG overnight followed by qPCR to quantify the corresponding genomic region across select cytokine gene loci. List of primers used is in the table S2.

Next-generation sequencing of the transcriptome (RNA-seq)

We used RNA-seq to assess the role of Cbx7 as a transcriptional activator in medium and anti-CD3/28–stimulated splenic CD4 T cells from Cbx7^+/+ and Cbx7^−/− mice. Cell samples were collected from a total of 16 mice (four mice per group) following stimulation for 6 hours. Total RNA was isolated using the standard RNA isolation kit from QIAGEN (no. 74134, Valencia, CA). The isolated total RNA was processed for next-generation sequencing library construction as developed in the NJH Genomics Facility for analysis with an Illumina NovaSeq 6000 (San Diego, CA). A Kapa Biosystems (Wilmington, MA) KAPA Hyper mRNA-Seq kit for whole transcriptome libraries was used to primarily target all poly(A) RNA. Briefly, library construction started from isolation of total RNA species, followed by mRNA [poly(A)] isolation, first- and second-strand cDNA synthesis, adaptor ligation, amplification, and cluster generation. Once validated, the libraries were sequenced as barcoded-pooled samples and run on the NovaSeq 6000 on an S4 flowcell with 2 × 150–base pair sequencing chemistry. Following library build, genes with less than or equal to one count per million reads in less than three mice were removed. Last, variance stabilizing transformation and differential gene expression analysis were performed in succession on the filtered genes using R packages DESeq2 and Limma, respectively. A total of 11,090 genes passed the filtering step. KEGG pathway enrichment analysis of differentially expressed genes was performed using Enrichr under default parameters.

Statistical analyses

Statistical analyses were performed, and data figures were prepared with GraphPad Prism software (v7; GraphPad Software, San Diego, CA). To calculate statistically significant differences between groups, we used two-tailed Student’s t tests for comparisons between two groups, and one-way analysis of variance (ANOVA), followed by Tukey’s post hoc test when comparing more than two groups for mouse data. Statistical significance was analyzed by Mann-Whitney U test or Wilcoxon signed-rank test for human data. Data are presented as means ± SEM. Statistical significance is indicated by the following annotations: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001, or P values are indicated above the graphs.

Study approval

All animal experiments were approved by the Institutional Animal Care and Use Committee, National Jewish Health, protocol no. AS2614-11-23. Protocols for blood and BAL studies from patients with asthma and disease control were approved by the Institutional Review Board (protocol no. HS-1700-528, HS-3153, and HS-2918) and were conducted in accordance with the principles of good clinical practice. All patients provided written informed consent.

Supplementary Materials

This PDF file includes:

Figs. S1 to S6

Tables S1 and S2