Bromodomain protein Brd3 promotes Ifnb1 transcription via enhancing IRF3/p300 complex formation and recruitment to Ifnb1 promoter in macrophages

As members of bromodomain and extra-terminal motif protein family, bromodomain-containing proteins regulate a wide range of biological processes including protein scaffolding, mitosis, cell cycle progression and transcriptional regulation. The function of these bromodomain proteins (Brds) in innate immune response has been reported but the role of Brd3 remains unclear. Here we find that virus infection significantly downregulate Brd3 expression in macrophages and Brd3 knockout inhibits virus-triggered IFN-β production. Brd3 interacts with both IRF3 and p300, increases p300-mediated acetylation of IRF3, and enhances the association of IRF3 with p300 upon virus infection. Importantly, Brd3 promotes the recruitment of IRF3/p300 complex to the promoter of Ifnb1, and increases the acetylation of histone3/histone4 within the Ifnb1 promoter, leading to the enhancement of type I interferon production. Therefore, our work indicated that Brd3 may act as a coactivator in IRF3/p300 transcriptional activation of Ifnb1 and provided new epigenetic mechanistic insight into the efficient activation of the innate immune response.

transcription), suggesting that they possess histone chaperone activity 14 . However, these two proteins are not simply redundant. Except the interaction with histones, Brd3 could also combine with transcription factors, such as GATA1 and promote its chromatin occupancy at erythroid target genes 15 . Brd4 has been found acting as a co-activator for the transcriptional activation of NF-κ B 16 , suggesting that BETs might participate in immune response [17][18][19] .
In our previous effort to identify molecules selectively involved in the regulation of innate immune response against viral infection 20 , we found Brd3 decreased nearly 2 folds after VSV infection in macrophages by genome-wide screening. Together with the data mining results of the GEO profiles that reveals Brd3 downregulation after various virus infection (see Results), these evidences strongly suggested that Brd3 may be involved in the process of virus-triggered immune response. In this study, the function of Brd3 in virus-initiated immune response was addressed. We demonstrated that Brd3 is an indispensable molecule for macrophages to produce IFN-β after virus infection. It can interact with IRF3/p300 complex, and enhances their recruitment to the Ifnb1 promoter after viral infection. We further demonstrate that Brd3 increases the acetylated histone3/histone4 within the Ifnb1 promoter. Therefore, our work revealed Brd3 as a positive regulator in the production of IFN-β in response to viral infection, and provided new mechanistic insight into the efficient activation of the innate immune response.

Virus infection down regulates Brd3 expression in macrophages.
We first examined the expression pattern of Brd3 in mouse normal tissues and immune cells by RT-PCR. As shown in Fig. 1a, Brd3 was ubiquitously expressed in various mouse tissues, including immune organs such as the thymus, bone marrow, and spleen. Further detection of Brd3 expression in immune cells revealed that Brd3 was also expressed in various immune cells including macrophages and NK cells (Fig. 1a).
Our systemic analysis identified several genes of BET family that were downregulated significantly in macrophages in response to VSV infection, including Brd3 20 . Data mining of Brd3 expression pattern in NCBI GEO profiles also revealed similar data including measles virus brain infection model in mice (GDS4553/1450902_at/ Brd3), pandemic H1N1 influenza virus infections of human bronchial epithelial cells (GDS4855/203825_at/ BRD3), and treatment with Hepatitis C virus core protein of the hepatocyte cell line (GDS2239/212547_at/ BRD3), indicating the decreased expression of Brd3 after various virus infection. We then wondered whether (a), Total RNA was extracted from different mouse tissues and immune cells, 1 μ g RNA was used to perform reverse transcription-PCR. Real time-PCR was performed to analysis the Brd3 mRNA expression level, Actb was used as a control. Then the reaction product was analysed by agarose electrophoresis. (b), Mouse peritoneal macrophages were infected with HSV (MOI = 10), VSV (MOI = 10), SeV (MOI = 10) for the indicated times. The Brd3 mRNA expression level was detected by Q-PCR. The results were presented as fold expression of Brd3 mRNA to that of Actb. (c), Mouse peritoneal macrophages were treated as in b, cells were lysed and subjected to Western blot analysis(40 μ g) with the indicated antibodies. Data are representative of three independent experiments with similar results (a and c), and data are shown as means ± SD of three independent experiments. *P < 0.05(analysis of variance (ANOVA)).
various virus infection would affect the expression of Brd3 in macrophages. Primary peritoneal macrophages were treated with HSV, VSV and SeV for different times, respectively. The result showed that Brd3 mRNA levels significantly decreased in virus-stimulated macrophages (Fig. 1b). Virus infection also markedly down regulated Brd3 protein expression in a time-dependent manner (Fig. 1c). These data indicate that Brd3 might be involved in virus triggered immune response.

Brd3 is indispensable for the production of IFN-β in virus-infected macrophages.
To investigate the role of Brd3 in virus-induced innate immune response, CRISPR-Cas9 technology was used to knockout Brd3 expression in RAW264.7 cells (Supplementary Fig. 1a). Then we used this cell line to examine the production of pro-inflammatory cytokines and type I interferon triggered by virus infection. The results showed that, after virus infection Brd3 knockout (Brd3-ko) cells showed significantly decreased IFN-β production both in mRNA and protein levels (Fig. 2a,b), and, to a much lesser extent, IL-6 production ( Supplementary Fig. 1b). However, the production of pro-inflammatory cytokine TNF-α was not affected (Supplementary Fig. 1c). Different clones of Brd3-ko cells showed similar results ( Supplementary Fig. 2a-c) and increased virus replicates ( Supplementary Fig. 2d). The results in primary murine macrophages further confirmed that Brd3 has a positive role in IFN-β production ( Supplementary Fig. 3). To confirm the increased production of IFN-β mediated by Brd3, Brd3-ko cells were transfected with Myc-tagged Brd3 expression vector for 24 hours (Supplementary Fig. 4) and then infected with viruses. As shown in Fig. 2c, the virus-induced production of IFN-β was restored by reintroducing Brd3 expression into Brd3-ko cells. These results indicate that Brd3 can promote IFN-β expression in virus activated macrophages.
Furthermore, we evaluated the effect of Brd3 on the phenotype of macrophages. A Flow cytometry analysis of macrophages transfected with specific Brd3 si-RNA showed that the expression of CD11b and F4/80 was barely affected by Brd3 knocked down in macrophages before or after VSV infection(data not shown), indicating that Brd3 has no influence on macrophage phenotype.
Brd3 associates with IRF3 and promotes IRF3-mediated IFN-β production. We then explored the underlying mechanism for increasing virus-induced IFN-β expression by Brd3. Virus-triggered signaling, especially the TBK1-IRF3 signal pathway which is important for the production of IFN-β were under first consideration. As shown in Fig. 3a, Brd3 knockout hardly had any effect on VSV-triggered phosphorylation of TBK1, p65 or IRF3. And the translocation of p65 and IRF3 into the nucleus was not affected either by Brd3 knockout after virus infection or LPS challenge (Fig. 3b). It's been reported that Brd3 mainly resides in nucleus 21 . We also confirmed the predominant nuclear localization of Brd3 in Raw 264.7 macrophages (Fig. 3c), indicating (d), RAW264.7 cells were infected with VSV (MOI = 10) for the indicated times. Cell lysates (500 μ g) were immunoprecipitated with Brd3 antibody and then immunoblotted with the indicated antibodies. Similar results were obtained in three independent experiments. (e), HEK293T cells were co-transfected with 100 ng IRF3-wide type expressing plasmid; 50 ng IFN-β luciferase reporter plasmid; and 5 ng pTK-Renilla-luciferase reporter plasmid together with 50, 100, or 150 ng of Brd3-expressing plasmid. Total amounts of plasmid DNA were equalized using an empty control vector. After 24 hours of culture, luciferase activity was measured and normalized by Renilla luciferase activity. (f), HEK293T cells were co-transfected with 50 ng IFN-β luciferase reporter plasmid; 5 ng pTK-Renilla-luciferase reporter plasmid together with 50, 100, or 150 ng of Brd3expressing plasmid and detected as in (e). Data are shown as means ± SD of three independent experiments. *P < 0.05(ANOVA).
that Brd3 may mainly function in nucleus in macrophages, to regulate IFN-β transcription. The multiple functional BETs can interact with transcription factors 16 . Given the important role of IRF3 in the production of type I interferon 22 , we investigated whether Brd3 positively regulated IFN-β production by interacting with IRF3. Immunoprecipitation assays with lysates from RAW 264.7 cells that express Brd3 and IRF3 endogenously were performed to test whether these two proteins can interact with each other. As shown in Fig. 3d, IRF3 could be detected in Brd3 immunoprecipitates and the interaction was increased after virus infection. To further investigate the effect of Brd3 association with IRF3, Myc-tagged Brd3 expression vector were co-transfected with Flag-tagged IRF3 expression vector and IFN-β luciferase reporter plasmid in HEK293T cells to detect the luciferase activity of IFN-β by dual-luciferase reporter assay. The results showed that IRF3-mediated reporter activity of IFN-β was significantly enhanced by Brd3 in a dose-dependent manner (Fig. 3e), whereas Brd3 overexpression alone couldn't affect the IFN-β reporter activity (Fig. 3f). These results suggest that Brd3 could associate with IRF3 in nucleus and promote IRF3-mediated IFN-β activation.
Brd3 enhances p300-mediated acetylation of IRF3. Besides being phosphorylated which is required for the activation, IRFs could also be acetylated (26,27) although their acetylation has not been well studied. As the bromodomains in Brd3 has the ability to recognize acetylated proteins, we then examined the acetylation status of IRF3 after virus infection. The result showed that overexpression of Brd3 increased the acetylation level of IRF3 after VSV infection (Fig. 4a). It is reported that acetyltransferase p300 may have the ability to acetylate IRF3 23,24 . We further co-transfected Flag-tagged IRF3, p300 and Myc-tagged Brd3 expression vectors into 293T cells. 48 hours later cells were infected with VSV for 8 hours and then subjected to IP and then Western blot analysis. As shown in Fig. 4b, the acetylation level of IRF3 was increased when p300 were co-expressed, and this modification has been reinforced when Brd3 was overexpressed. However, overexpression of Brd3 alone had minimal effect on the acetylation level of IRF3. These results suggest that Brd3 can promote p300-mediated acetylation of IRF3.
Brd3 associates with p300 and enhances the interaction between IRF3 and p300. After being phosphorylated in the cytoplasm, IRF3 would translocate into the nucleus and interact with p300 to initiate the transcription of the type I IFN and ISG genes 22 . Although IRF3 displays relatively strong DNA binding activity, its transactivation appears to be dependent on p300 which can acetylate histones and facilitate the transcription process. Considering that Brd3 could interact with IRF3 and enhance p300-mediated acetylation of IRF3, we speculated that Brd3 could also be involved in the formation of IRF3/p300 complex. Co-IP experiments showed that endogenous p300 could be detected in Brd3 immunoprecipitants and the association could also be detected in unstimulated RAW264.7 cells (Fig. 5a). We further investigated the role of Brd3 in binding with IRF3/p300 complex. For endogenous interaction of IRF3 and p300, co-IP analysis were performed in lysates of Brd3-ko and control cells, and result showed that the interaction between IRF3 and p300 was attenuated in Brd3-ko cells compared with control cells after VSV infection (Fig. 5b). Furthermore, we transfected Flag-tagged IRF3, p300 and Myc-tagged Brd3 expression vectors into 293T cells and immunoprecipitated Flag for Western blot analysis. As shown in Fig. 5c, binding of p300 and IRF3 in 293T cells was markedly enhanced by Brd3 overexpression. Together, these experiments provide evidence that Brd3 interacts with IRF3/p300 complex and promotes the association between IRF3 and p300, suggesting that they may act as a complex to co-regulate gene transcription.

Ifnb1. Our observation that Brd3 increased virus-induced IFN-β production independent of NF-κ B and
TBK1-IRF3 signaling pathways implied that Brd3 may mediate gene-specific transcription regulation in gene locus of Ifnb1. To this end, ChIP assay was performed to evaluate the effect of Brd3 on IRF3 and p300 recruitment to the Ifnb1 promoter. As shown in Fig. 6a, virus-induced binding of IRF3 and p300 to the gene promoter of Ifnb1 decreased significantly in Brd3-ko cells, suggesting that Brd3 enhances virus-triggered IRF3 and p300 recruitment to Ifnb1 promoter.
As acetyl-lysine "readers", Brd2 and Brd3 have been found to couple histone acetylation to transcription. We did indeed detect Brd3 association with acetylated histone3 and histone4, even in unstimulated macrophages ( Supplementary Fig. 5). Given the fact that histone acetylation is implicated in regulating transcription of type I interferon during innate immune response, we therefore tested whether histone acetylation at the Ifnb1 promoter was affected in Brd3-ko cells after VSV infection. The result showed that, when Brd3 protein has been knocked out, the acetylated histone3 and histone4 within Ifnb1 promoter was reduced significantly comparing with control cells after virus infection (Fig. 6b). However, the acetylated histone3 and histone4 within Tnfa promoter showed no difference in Brd3-ko cells and control cells (Supplementary Fig. 6). Taken together, these results indicate that Brd3 promotes IRF3/p300 transcription activity by enhancing virus-triggered IRF3/p300 recruitment to Ifnb1 promoter and assisting the acetylation of histone3 and histone4, leading to the promotion of type I interferon production. Discussion IFN-β expression is dependent on the assembly of a transcription enhanceosome consisting of many transcription factors. IRF3 is one of the main transcription factors responsible for recruiting the histone acetyltransferases p300/CBP to the Ifnb1 enhanceosome 3 . In this work, we demonstrated that Brd3, as epigenetic regulators, has functionally important consequences for antiviral responses. Our results have shown that the virus-induced interaction of IRF3 and p300 was attenuated in Brd3-ko macrophages. Notably, knockout of Brd3 inhibited VSV-induced recruitment of both IRF3 and p300 to the Ifnb1 promoter. These data support the hypothesis that Brd3 associates with the IFN-β enhanceosome by interacting with IRF3 and p300 and enhances their recruitment to the Ifnb1 promoter to increase histone hyperacetylation and transcription of Ifnb1 in virus-triggered innate immune responses. Given that knockout of Brd3 inhibited the phosphorylation of neither TBK1 nor IRF3 and p65 and impaired recruitment of IRF3/p300 to the Ifnb1 promoter, we propose that Brd3 mediate virus-induced production of type I interferon by acting as a coactivating pathway for such interferon production.
A lot of observations have underscored the importance of the transcription factor IRF3 and the coactivator p300 in immune response against viral infection 3,25 . Our study demonstrated that Brd3 could positively regulate IRF3/p300 complex transcriptional activation toward virus-induced IFN-β production by interacting with and enhancing the formation of IRF3/p300 complex. As acety-lysine "readers", bromodomain-containing proteins recognize and decode the acetylated proteins to regulate their activities. Therefore, the increased Brd3 interaction with IRF3/p300 complex upon viral infection we detected here might possibly correlated with the acetylation status of either IRF3 or p300 (or both). It's been shown that although their acetylation has not been well studied, IRFs could also be acetylated 23,24 . The coactivator and acetyltransferase p300 may be among the potential acetyltransferases responsible for the IRF3 acetylation. Suhara et al. has demonstrated that IRF3 could be acetylated by p300 by in vitro experiments 23,24 . Here we found that Brd3 increased the acetylation level of IRF3 after VSV infection (Fig. 4a). When co-expressed with IRF3 and p300 in 293T cells, Brd3 promoted p300-mediated acetylation of IRF3 (Fig. 4b). Therefore, while Brd3 contains no HAT domain, it may promote the acetylation process by recruiting other acetyltransferase (e.g., p300) to substrate proteins like IRF3. Further experiments to determine whether Brd3 (or other bromodomain proteins) interacts with acetylated IRF3 via specific bromodomain and to identify the precise sites of IRF3 acetylation as well as a sequence motif that is specifically recognized by the bromodomains would be helpful to reveal the detailed mechanisms underlining the interaction of Brd3 with IRF3/ p300 complex.
In this work, we found that Brd3 could interact with IRF3, but not transcription factor NF-κ B (data not shown). However, there's report demonstrated that Brd4 could interact with NF-κ B via the binding of the bromodomain to the acetylated-Lys 310 residue on the RelA subunit 16 . The differences may partially due to the sequence flexibilities and structural diversities among BETs 21,26 . Whether Brd4 could interact with NF-κ B in the macrophage context remains to be established. The differentially regulation of transcription by different family members reflects the nonredudancy in the functions of bromodomain-containing proteins. This underscores the need for systematic genome-wide approaches to catalog genes associated and regulated by BETs, which will facilitate to describe the physiological and pathological roles of these proteins.
As a positive regulator of IFN-β transcription activation, Brd3 expression was downregulated after virus infection, as demonstrated by genome-wide screening, data mining of the GEO profiles, as well as mRNA and protein level analysis of virus-infected macrophages (Fig. 1). This is different from other activators found participating in the production of IFN-β , which might be a fine-tuning mechanism possessed by the host to prevent the superfluous IFN-β production. Proper downregulation of the expression of pro-inflammatory cytokines is necessary for the host to avoid excessive cytotoxicity and tissue injury after infection. Diverse mechanisms in signal transduction 27 and PTM level 20,22 exist to allow well-timed attenuation of inflammatory reactions. The multiplicity of feedback regulation indicates the presence of divergent and maybe combined mechanisms that altogether assure timely cessation of cytokine induction. Brd3 enhancement of IRF3/p300 recruitment to the gene promoter may contribute to the collective inhibition by introducing a distinct mechanism that exerts at the level of Ifnb1 transcription.
Owing to the rapid development of both biological function and structural basis of BET proteins, the inhibitors of BETs have become a newly emerging therapeutic strategy for cancer and other diseases 8,28 . Collectively, our results raise the possibility that Brd3 expression and regulation has important consequences for Type I interferon responses and targeting Brd3 specifically may have benefits in viral infection. Plasmid Constructs. A recombinant expression vector encoding Myc-tagged Brd3 was constructed by PCR cloning into the pReceiver-M43 eukaryotic expression vector (Fulengene Genecopoeia). The pRL-TK-Renilla-luciferase and IFN-β luciferase reporter plasmids were described previously 29 . The p300 and Flag-tagged IRF3 expression vector was constructed by PCR cloning into the pcDNA3.1 eukaryotic expression vector (Invitrogen). All constructs were confirmed by DNA sequencing.

Mice, Antibodies and
Cell culture, and transfection. Thioglycolate-elicited mouse primary peritoneal macrophages were isolated and cultured as described previously 30 . Mouse macrophage cell line RAW 264.7 and human HEK293T cells were obtained from the American Type Culture Collection (Manassas, VA) and cultured as described 31 . RAW 264.7 cells stably overexpressing TAP-tagged IRF3 vectors which contains a calmodulin binding protein (CBP) epitope in TAP tag was established in our lab. For transient transfection of plasmids in HEK293T cells, jetPEI reagent was used according to manufacturer's instructions (Polyplus). RNA Interference. The specific siRNA for Brd3 (si-Brd3) and control siRNA (si-Non) were purchased from Gene Pharma and delivered into mouse peritoneal macrophages using INTERFERin reagent (Polyplus) as described previously 32 . si-Brd3 were synthesized as follows: 5′ -GGAGAGCUCUUCAGAUUCATT-3′ (sense), and 5′ -UGAAUCUGAAGAGCUCUCCTT-3′ (anti-sense).
Phenotype analysis. Cells were incubated for 15 min at 4 °C with fluorescein-conjugated monoclonal Abs in labeling solution. Fluorescein-conjugated isotype controls were used to establish the background. Flow cytometry analyses were conducted on LSR II (BD Biosciences). Data were analyzed with CELLQuest or FACSDiva software (BD Biosciences) as described previously 33 . CRISPR-Cas9-mediated gene knockout. CAS9 RAW 264.7 cells using jetPEI. Single transfected cells were sorted into individual wells in a 96-well plate using the MoFlo XDP, expanded, and screened by immunoblot.
ELISA. IFN-β , TNF-α and IL-6 levels in the supernatants were measured by ELISA kit (R&D) according to the manufacturer's instructions.
Immunoblot and immunoprecipitation. The nuclear and cytoplasmic extracts were prepared with NE-PER nuclear and cytoplasmic extraction reagents (Pierce). Immunoblot and immunoprecipitation were performed as described previously 31 .
Luciferase Reporter Gene Assay. HEK293T cells were cotransfected with a mixture of the IFN-β luciferase reporter plasmid, pRL-TK-Renilla-luciferase plasmid, and the appropriate additional constructs for 24 hours. Total amounts of DNA were equalized with empty control vector. Luciferase activities were measured using dual-luciferase reporter assay system (Promega) as described previously 32 . Data were normalized for transfection efficiency by comparing firefly luciferase activity with that of Renilla luciferase.