Inhibition of the BET family of epigenetic reader proteins: A novel principle for modulating gene expression in IgE‐activated mast cells

Abstract Introduction The BET family of bromodomain‐containing proteins constitute epigenetic readers that bind to acetylated lysine residues of core histones, thereby translating epigenetic histone marks to effects on gene expression. BET inhibitors are currently emerging as promising therapeutic agents for treatment of various pathological conditions. Here, we explored the potential of using BET inhibition to modulate IgE‐mediated responses in mast cells. Methods We assessed the effects of BET inhibitors PFI‐1, I‐BET151, and I‐BET762 on responses downstream of mast cell activation through IgE receptor cross‐linking. Results BET inhibitors were neither toxic for mast cells (at doses up to 20 μM), nor did they prevent IgE‐mediated mast cell degranulation. However, we found that BET inhibition, in particular by I‐BET151, suppressed IL‐6 gene expression and IL‐6 protein release in response to IgE‐mediated mast cell activation. This was observed in both bone marrow‐derived mast cells (BMMCs) and in mature peritoneal‐cell derived mast cells. Further analysis showed that BET inhibition also suppressed the expression of a number of additional genes of those that were upregulated by IgE receptor cross‐linking, including IL‐3, IL‐7R, CCR1, and ADAMTS9. However, BET inhibition was selective, i.e., several genes that were upregulated by IgE receptor cross‐linking were not affected by BET inhibitors. Conclusions These findings suggest that BET inhibition can interfere with the upregulated expression of selected genes in mast cells activated by IgE receptor cross‐linking. Further, our findings introduce the concept of utilizing epigenetic mechanisms for modulating mast cell function in the context of IgE‐driven disease.


Introduction
Mast cells are well known for their detrimental impact on allergic disease, but are also important players in a range of additional pathologies [1]. The classical mode of mast cell activation, most notably in the context of allergic reactions, is through binding of multivalent antigen to IgE molecules bound to the high affinity IgE receptor, FceRI, on the mast cell surface. The ensuing IgE receptor cross-linking will lead to extensive release of preformed mediators from granular stores, that is degranulation, but IgE-mediated activation also results in the synthesis of numerous lipid-derived proinflammatory mediators and expression of a panel of genes coding for pro-inflammatory compounds [2,3]. These latter include genes coding for cytokines, chemokines, and various growth factors [3]. Altogether, mast cell activation through IgE receptor cross-linking will thus lead to the release of a wide array of pro-inflammatory compounds, which collectively gives rise to a powerful inflammatory reaction.
Based on the powerful impact of mast cells on allergic and other pathological conditions, there is a great demand for therapeutic strategies that can counteract harmful mast cell activities. Such strategies include the use of various mast cell stabilizers and anti-IgE therapy to prevent mast cell degranulation, as well as antagonists to the individual mediators that are released following mast cell degranulation [4]. However, there is also need for strategies that can interfere with gene expression patterns that are induced upon IgE-mediated mast cell activation.
Histone modification by various mechanisms, including lysine acetylation, methylation, phosphorylation, and sumoylation is now established as a major epigenetic mechanism with wide consequences for gene expression [5,6]. The Bromodomain and extraterminal domain (BET) family of bromodomain proteins bind specifically to acetylated lysine residues in core histones, thereby transmitting the signal imposed by histone acetylation into effects on gene expression [7]. Based on this principle, inhibitors of BET proteins have been developed to provide a novel means of directly modulating effects on gene expression. Such inhibitors have subsequently been evaluated in diverse settings and are currently regarded as promising therapeutic agents in various pathological conditions, including cancer, sepsis, and autoimmunity [8][9][10][11][12]. However, the possibility of using BET inhibitors for modulating mast cellmediated events has hitherto not been explored.
Here, we investigated the effect of BET inhibition on mast cell function. Our findings identify BET inhibition as a novel means of modulating the expression of selected genes in response to IgE-mediated mast cell activation. Hence, the use of BET inhibitors could represent a new strategy to directly interfere with gene expression patterns in pathological settings where mast cells are involved.

Cell culture
Bone marrow-derived mast cells (BMMCs) were obtained by culturing bone marrow cells from the femur and tibia of C57BL/6 mice as previously described [13]. The animal experiments were approved by the local ethical committee (Uppsala djurf€ ors€ oksetiska n€ amnd; approval no C 31/14). Peritoneal cell-derived mast cells (PCMCs) [14] were obtained by maturation of cells from peritoneal lavage of C57BL/6 mice as described [13].

Mast cell activation
Mast cells (BMMCs or PCMCs, at 1 Â 10 6 cells/ml) were sensitized overnight with IgE anti-DNP at 0.1 mg/ml. The next day, cells were washed twice and suspended in fresh media or Tyrode's buffer (130 mM NaCl, 5 mM KCl, 1.4 mM CaCl 2 , 1 mM MgCl 2 , 5.6 mM glucose, 10 mM HEPES, and 0.1% BSA, pH 7.4) and stimulated for 1-24 h with DNP-HSA (0.5 mg/ml). In samples where BET inhibitors were evaluated, mast cells were incubated for 1 h with BET inhibitor (PFI-1, I-BET151, or I-BET762) at various concentrations, followed by activation with DNP-HSA. For cytokine/chemokine release and Western blot analysis, activation with DNP-HSA was performed for 4 or 24 h, while for mRNA expression, mast cells were activated for 1 h.

Cytotoxicity test
Mast cells (BMMCs or PCMCs, at 10 6 cells/ml) were incubated with serial dilutions of BET inhibitors (ranging from 10 nM to 20 mM) for 1 h. Cell viability was measured using the Cell Titer-Blue 1 cell viability assay (Promega-Invitrogen, Carlsbad, CA) according to the recommendations of the manufacturer.

Cytokine/chemokine release
Supernatants from activated mast cells were analyzed for levels of IL-6 and TNFa using ELISA kits as detailed by the manufacturer (eBiosciences, San Diego, CA).

Beta-hexosaminidase assay
Supernatants from activated mast cells were analyzed for their content of b-hexosaminidase activity as described [15] as a measure of mast cell degranulation.
Total RNA isolation, quantitative real time RT-PCR, and gene array analysis Total RNA isolation and quantitative real time RT-PCR (qPCR) was performed as described [13]. Hypoxanthine guanine phosphoribosyl transferase (Hprt) was used as housekeeping gene [13]. Primers for IL-3, À15 and 7r, CSF1, CCR1, and NLRP3 are described in Table 1. Primers for Nr4a3 [16], ADAMTS9 [13], IL-6 and À13 [17], and Cpa3 [18] were as described. Data were calculated as DDC t relative to HPRT, where values of the IgE-sensitized sample were set as 1. In samples, where no signal corresponding to the sensitized controls (IgE only) were detected, values for the IgE þ DNP group were set as 1. Gene array analysis was performed using Affymetrix GeneChip 1 expression arrays (GeneChip 1 Mouse Gene 1.0 ST Array), as described previously [19].

Statistical analysis
Data shown are means AE standard error of the mean (SEM). Statistical analyses were performed by using GraphPad Prism 7.0 (GraphPad Software) and paired Student's t-test for two-tailed distributions. Differences were considered significant if the P-values were 0.05 or less.

BET inhibitors are non-cytotoxic for mast cells
To evaluate the impact of BET inhibition on mast cell functionality, we first assessed whether BET inhibitors are cytotoxic to mast cells (BMMCs). At doses up to 20 mM, neither of three tested BET inhibitors (PFI-1, I-BET151, I-BET762) were cytotoxic to mast cells after 1 h of incubation (Fig. 1A), and only limited cytotoxicity at concentrations above 50 nM were seen after 24 h of incubation (not shown). For the continuation of the study, BET inhibitors were used at concentrations and incubation times that were non-toxic to mast cells (250 nM; 1 h).

BET inhibition does not affect IgE-mediated mast cell degranulation
Next, we assessed whether BET inhibition can interfere with mast cell degranulation in response to IgE receptor crosslinking. To this end, mast cells were pre-loaded with anti-DNP IgE overnight, followed by IgE cross-linking by addition of DNP-HSA, either in the presence or absence of BET inhibitors. As seen in Figure 1B, IgE receptor cross-linking resulted in a robust degranulation as measured by b-hexosaminidase release. However, neither of the BET inhibitors used caused any detectable reduction of this response. Hence, BET inhibition does not affect IgEmediated mast cell degranulation.

BET inhibition suppresses IL-6 expression and release in response to IgE receptor cross-linking
Earlier reports have shown that BET inhibition can suppress IL-6 expression in response to TLR4 ligation in macrophages [8,20]. In order to evaluate if BET inhibitors can Table 1. Primers used for qPCR. Target  Forward primer  Reverse primer   IL-3  ACCCACCGTTTAACCAGAACG  TGGACAGGTTTACTCTCCGAA  IL-15  CATCCATCTCGTGCTACTTGTGTT  CATCTATCCAGTTGGCCTCTGTTT  IL-7r  CCTTCGAAACTCCAGAACCCA  GACTAGGCCATACGACAGGTT  CSF1  AGTCTTGCTGACTGTTGGGG  TGTCTGTCCTCATCCTGGGT  CCR1  AGGCCCAGAAACAAAGTCTG  AGCAGTCTTTTGGCATGGAGT  NLRP3 CGAGACCTCTGGGAAAAAGCT GCATACCATAGAGGAATGTGATGTACA modulate gene expression in IgE-activated mast cells, we therefore first focused on IL-6. As depicted in Figure 2A, IgE receptor cross-linking caused a strong upregulation of IL-6 gene expression. Moreover, it was seen that BET inhibition resulted in a profound reduction of this response. In particular, I-BET151 proved to be a highly potent inhibitor of IL-6 expression. In contrast, although TNFa gene expression was profoundly upregulated in response to IgE receptor cross-linking in mast cells, neither of the BET inhibitors was capable of suppressing the TNFa induction (Fig. 2B). In agreement with these data, ELISA measurements showed that BET inhibition also suppressed the release of IL-6 protein in response to IgE receptor cross-linking ( Fig. 2C and D), whereas no effect on the release of TNFa was seen ( Fig. 2E and F).
BET inhibition modulates IL-6 expression in peritoneal cell-derived mast cells Although BMMCs are commonly used for studies of mast cell function, they are less mature than those mast cells that are found in vivo. To test the impact of BET inhibition on fully mature mast cells, we therefore developed PCMCs [14]. Importantly, PCMCs have previously been shown to represent fully differentiated mast cells of connective tissue subtype [14]. As for BMMCs, neither of the BET inhibitors at any of the tested concentrations showed cytotoxicity for PCMCs, when incubated for 1 h with the cells (Fig. 3A). Also after prolonged incubation (24 h), undetectable or only marginal cytotoxicity was seen at concentrations above 100 nM (data not shown). In agreement with the effects seen on BMMCs, BET inhibition did not affect the degranulation of PCMCs in response to IgE receptor crosslinking (Fig. 3B). In further accordance with the effects on BMMCs, BET inhibition suppressed the expression of IL-6 mRNA in response to IgE receptor cross-linking and inhibited the release of IL-6 protein in PCMCs ( Fig. 3C and D). It was also seen that BET inhibition down-regulated the baseline

Mast cells and BET inhibition
expression of IL-6. As for BMMCs, BET inhibitor I-BET151 was more effective than PFI-1 and I-BET762. A dose response experiment showed that optimal inhibition of IL-6 gene expression was obtained at $250 nM of I-BET151 (Fig. 3E).
BET inhibition represses the expression of selected genes that are upregulated by IgE receptor cross-linking To provide a deeper insight into whether BET inhibition affects gene expression in IgE-activated mast cells beyond the effects on IL-6 (see Figs. 1 and 2), we performed a gene array analysis to search for additional genes affected by BET inhibitors. To this end, we performed a gene array analysis of PCMCs activated by IgE receptor cross-linking, either in the absence or presence of BET inhibitor I-BET151. As expected, this analysis suggested that a large number of proteinencoding genes were upregulated in response to IgE receptor cross-linking (Suppl . Table S1). Further, the gene array data indicated that only selected genes were substantially affected by BET inhibition (Suppl . Table S2). These included IL-6, NLRP3, IL-7R, IL-3, CCR1, ADAMTS9, IL-15, and IL-13. To confirm the findings obtained through the gene array screen, we used quantitative real time PCR (qPCR). As shown in Figure 4, the qPCR analysis indeed revealed a robust induction of the IL-3 (Fig. 4A), IL-13 (Fig. 4B), IL-7R Fig. 4D), CSF1 (Fig. 4E), CCR1 (Fig. 4F), and ADAMTS9 ( Fig. 4H) genes in PCMCs subjected to IgE receptor crosslinking. We also noted that Nr4a3, a gene that previously was shown to be one of the most highly induced genes in mast cells activated by various means [16], was strongly upregulated in IgE-activated PCMCs (Fig. 4G). Out of these genes, the expression of IL-3, IL7R (coding for the IL-7 receptor), and CCR1 (receptor for chemokines CCL3, À5, À7, and À23) was suppressed by BET inhibition. Also the expression of NLRP3 (a key inflammasome component [21]) was suppressed by BET inhibition, although the qPCR analysis did not support upregulated NLRP3 expression after IgE receptor cross-linking (Fig. 4J). As for IL-6 (see Fig. 3), BET inhibitor I-BET151 was more effective than the other BET inhibitors. Notably, although the qPCR analysis confirmed an upregulated expression of IL-13, CSF1, and Nr4a3 after IgE receptor cross-linking, no effect of BET inhibition was seen on the expression of any of these genes. For ADAMTS9 (a metalloprotease), a trend of suppression was seen but this did not reach statistical significance. Further, BET inhibition did not affect the expression of the IL-15 gene (Fig. 4B). We also tested the effect of BET inhibition on Cpa3, a secretory granule-contained protease that serves as a major mast cell marker [2]. Cpa3 expression was not affected by IgE receptor cross-linking and, moreover, there was no effect of BET inhibition on Cpa3 expression (Fig. 4I).
To extend these observations, we additionally evaluated whether BET inhibition had an impact on the expression of the corresponding genes in BMMCs, that is in less mature mast cells. As seen in Fig. 5, IgE receptor cross-linking caused upregulation of the genes coding for IL-3 (Fig. 5A), IL-13 (Fig. 5B), CSF1 (Fig. 5E), CCR1 (Fig. 5F), Nr4a3 (Fig. 5G), ADAMTS9 (Fig. 5H), and NLRP3 (Fig. 5J) along with a trend of upregulated expression of the Il-7R gene (Fig. 5D), that is similar findings as for PCMCs. However, BMMCs were generally less responsive to BET inhibition in comparison with PCMCs, with only the expression of the ADAMTS9 (Fig. 5H) and IL-15 (Fig. 5C) genes being significantly repressed by any of the BET inhibitors. However, trends of inhibition were seen for the IL-7R (Fig. 5D) and CCR1 (Fig. 5F) genes. As in PCMCs, Cpa3 expression was neither affected by IgE receptor cross-linking, nor by BET inhibition (Fig. 5I).

Discussion
BET inhibitors have previously been assessed in a variety of settings and are currently emerging as promising therapeutic agents (see Introduction). However, the possibility of using this type of inhibitors to intervene with mast cell functions has, to our knowledge, not been addressed before. Here, we show that gene expression in mast cells activated by IgE receptor cross-linking can be modulated by inhibition of the BET proteins. One important implication of these findings is that responses downstream of IgE-mediated mast cell activation can be influenced by epigenetic mechanisms. In line with our findings, it was recently shown that interference with BRD4, one of the BET family members can suppress growth and induce apoptosis in neoplastic mast cells [22]. Hence, modulation of BET proteins could potentially represent a novel means of modulating mast cell-mediated events in a variety of settings.
It is apparent from the present study that BET inhibitors do not interfere with mast cell degranulation. It is therefore unlikely that BET inhibition has effects on the immediate response downstream of mast cell activation, that is events related to effects of the preformed mast cell mediators released from secretory granules. Instead, BET inhibition was shown to modulate gene expression patterns in response to IgE receptor cross-linking, thus implicating BET inhibitors as possible agents to modulate events occurring in the late phase responses following mast cell activation. However, it was noted that BET inhibition also suppressed the baseline expression of several of the assessed genes. Hence, BET-mediated regulation of gene expression in mast cells is not restricted to situations in which they are activated.
One of the main findings was that BET inhibition potently suppressed the upregulation of IL-6, both at the gene and protein level. This is thus in analogy with previous studies demonstrating that BET inhibitors can suppress IL-6 expression in macrophages stimulated with TLR4 ligands [8,20]. Based on the findings presented here, we may propose that BET inhibition can be used as a therapeutic strategy to modulate IL-6 gene expression in mast cells. Importantly, mast cellderived IL-6 has been implicated in a variety of pathological contexts, including atherosclerosis [23], stroke pathology [24], arthritis [25], Th17 skewing [26], mastocytosis [27], and diabetes [28]. Hence, modulation of IL-6 expression by BET inhibition could have a therapeutic potential in such settings.
Besides the IL-6 gene, we here identified a number of additional genes whose expression in IgE-activated mast cells was affected by BET inhibition. These included IL-3, IL-7R, and CCR1, all of which have established and important functions in inflammatory responses. Of note, IL-3 is a potent growth factor for mast cells [29], and the suppression of IL-3 expression by BET inhibition might thus interfere with IL-3-driven mast cell proliferation. Potentially, reduction of mast cell numbers by inhibition of BET proteins could thus have beneficial effects under circumstances when mast cells account for aggravating activities, for example in allergic asthma and mastocytosis. We also found that BET inhibition suppressed the expression of NLRP3. It is thus possible that BET inhibition could represent a novel strategy for interfering with inflammasome activity. In addition, the expression of ADAMTS9, a protease that recently was identified as a mast cell signature gene [30] and to be highly upregulated by IgE receptor cross-linking [13], was suppressed in mast cells subjected to BET inhibition. As ADAMTS9 is implicated in connective tissue remodeling events [31], BET inhibition could thus counteract such processes orchestrated by mast cells.
In summary, this study identifies BET inhibitors as novel agents with capacity to modulate mast cell gene expression. Importantly, the use of BET inhibition to modulate gene expression in mast cells constitutes a novel principle for intervening with mast cell function, in which gene expression is directly targeted rather than being modulated by targeting of signaling molecules that act upstream of the gene transcription apparatus.

SUPPORTING INFORMATION
Additional supporting information may be found in the online version of this article at the publisher's web-site Table S1. Genes upregulated in peritoneal cell-derived mast cells activated by IgE receptor crosslinking. Table S2. Genes downregulated by BET inhibitor 151 in peritoneal cell-derived mast cells activated by IgE receptor crosslinking.