Bromodomain-containing protein 4 regulates interleukin-34 expression in mouse ovarian cancer cells

Background Interleukin (IL)-34 acts as an alternative ligand for the colony-stimulating factor-1 receptor and controls the biology of myeloid cells, including survival, proliferation, and differentiation. IL-34 has been reported to be expressed in cancer cells and to promote tumor progression and metastasis of certain cancers via the promotion of angiogenesis and immunosuppressive macrophage differentiation. We have shown in our previous reports that targeting IL-34 in chemo-resistant tumors in vitro resulted in a remarkable inhibition of tumor growth. Also, we reported poor prognosis in patients with IL-34-expressing tumor. Therefore, blocking of IL-34 is considered as a promising therapeutic strategy to suppress tumor progression. However, the molecular mechanisms that control IL-34 production are still largely unknown. Methods IL-34 producing ovarian cancer cell line HM-1 was treated by bromodomain and extra terminal inhibitor JQ1. The mRNA and protein expression of IL-34 was evaluated after JQ1 treatment. Chromatin immunoprecipitation was performed to confirm the involvement of bromodomain-containing protein 4 (Brd4) in the regulation of the Il34 gene. Anti-tumor effect of JQ1 was evaluated in mouse tumor model. Results We identified Brd4 as one of the critical molecules that regulate Il34 expression in cancer cells. Consistent with this, we found that JQ1 is capable of efficiently suppressing the recruitment of Brd4 to the promotor region of Il34 gene. Additionally, JQ1 treatment of mice bearing IL-34-producing tumor inhibited the tumor growth along with decreasing Il34 expression in the tumor. Conclusion The results unveiled for the first time the responsible molecule Brd4 that regulates Il34 expression in cancer cells and suggested its possibility as a treatment target.


Background
Tumor microenvironment (TME) is composed of various types of cells including not only tumor cells but also immune cells, fibroblasts, mesenchymal, and endothelial cells. Cytokine is known as one of the important components for interactions and communications between tumor and non-tumor cells. Some cytokines promote tumor progression while others have anti-tumor effects. For example, interleukin (IL)-10 is involved in suppression of immune response via inhibiting the expression of major histocompatibility complex and costimulatory molecules on antigen-presenting cells which sometimes support tumor growth. On the contrary, IL-12 acts as an inducer of interferon-γ in antitumor immune responses [1][2][3]. Among various cytokines, IL-34 is a novel cytokine that was first identified in 2008 as an alternative ligand to colonystimulating factor-1 (CSF-1) for the CSF-1 receptor (CSF-1R) [4]. IL-34 has been reported to play crucial roles in TME. To date, IL-34 expression was observed in various types of tumors such as lung, liver, or colon cancer [5][6][7]. It has been shown that IL-34 expression was upregulated in cancer cells upon stimulation with anticancer drugs that was implicated with cancer cells' acquisition of resistance to chemotherapeutic treatment [8]. Additionally, although immune checkpoint inhibition therapy using anti-programmed death-1 antibody has shown dramatic effects, considerable cases with therapeutic resistance have been reported [9,10]. Among those, IL-34 has been suggested as a driver molecule of the resistance by inducing immune suppressive macrophages [11]. It has been reported that the macrophages generated by stimulation through CSF-1R signaling have the potential to induce T cell exhaustion and dysfunction [12]. There is another clinical fact that IL-34 expression in cancer correlates with poor prognosis and higher disease stage in several types of cancers such as brain, lung, ovarian cancers, and melanoma [13]. Therefore, it seems important to understand the expression mechanism of IL-34 in cancer cells. According to a previous report, IL-34 production is regulated through NF-κB or c-Jun N-terminal kinase signaling pathway [14]. However, the transcriptional regulation of IL-34 has not been identified. Consequently, this study aims to clarify the transcriptional regulator that controls Il34 expression.
There is a report indicating that the canonical promoter of IL34, as well as of CSF1R, is rich in putative RUNX1 binding sites in melanoma [15]. It has been noted that RUNX1 recruits the transcription regulators cyclindependent kinase (CDK), bromodomain-containing protein 4 (BRD4), the mediator complex, and the looping factor LIM domain binding 1 [16]. Among them, BRD4 is one of the components of bromodomain and extraterminal domain (BET) family with BRD2, BRD3, and BRDT. It has been reported that the BET family protein regulates cellular proliferation and cytokine production [17][18][19]. In detail, BRD4 has two different roles that plays as a transcriptional regulator and an epigenetic regulator (histone reader). As a transcriptional regulator, there is a report indicating that BRD4 interacts with RNA polymerase II (POL II) or P-TEFb complex (CDK9 and CycT1) [20]. As a histone reader, it was suggested that BRD4 recognizes histone lysine acetylation by recruiting additional chromatin modifiers [21,22]. Because inhibition of BRD4 impedes growth of cancer cells, targeting BRD4 has recently emerged as a promising anticancer approach [20]. However, it has not been explored whether expression of IL-34 is regulated by these factors, such as BRD4. JQ1, a low molecular compound which inhibits the binding of BET family members, including Brd4, to their targets, has been reported to modulate transcription of oncogenes such as c-Myc [23]. Moreover, JQ1 is known to inhibit multiple targets including TNF-alpha, Il6, and Mcp1 both in vitro and in vivo [24]. Thus, JQ1 is now widely used in cancer research [25,26], and its administration either alone or in combination with other anticancer agents has exhibited efficient suppression of a variety of tumors [27,28].
In this brief report, we show for the first time that BET inhibitor JQ1 could reduce Il34 expression in two IL-34 highly producing cancer cells, such as OV3121-ras4 and HM-1. Brd4 was recruited on Il34 promotor region which was blocked by JQ1, leading to downregulation of IL-34 production. When JQ1 was administered into an in vivo tumor model, both Il34 expression and tumor growth were suppressed. These results unveil the mechanism of IL-34 expression, which implicates a novel TME-targeting cancer therapy.

Enzyme-linked immunosorbent assay (ELISA)
The production of IL-34 in cell lines was measured with ELISA using LEGEND MAX TM Mouse IL-34 ELISA kit with pre-coated plates (clone: Poly5193) (BioLegend, CA, USA). Culture supernatants were collected at 48 h after seeding the cells at a density of 2 × 10 5 cells/well of 96-well plate. Absorbance at a test wavelength of 450 nm and a reference wavelength of 570 nm was measured by using a Multiskan FC (Thermo Fisher Scientific, MA, USA).

Cell viability assay
Cell viability was evaluated using the MTT Cell count kit (Nacalai Tesque). Absorbance at a test wavelength of 570 nm and a reference wavelength of 670 nm was measured by using a Multiskan FC (Thermo Fisher Scientific, MA, USA). Cell proliferation was observed for up to 2 days.

Quantitative reverse transcription PCR analysis
Total RNAs were extracted using TriPure Isogen Reagent (Roche Diagnostics, Mannheim, Germany), and 1 μg of total RNAs was used for first-strand cDNA synthesis using ReverTraAce (TOYOBO, Osaka, Japan). qRT-PCR was performed on cDNA products using Fast SYBR Green PCR Master Mix (Thermo Fisher Scientific), and samples were run on Applied Step One realtime PCR system (Thermo Fisher Scientific). The thermal cycling conditions were composed of 95°C for 20 s followed by an initial denaturation step at 95°C for 3 min, 40 cycles at 95°C for 3 s, and 60°C for 30 s. The experiments were carried out in triplicate. The relative quantification in gene expression was determined using the 2 -ΔΔCt method and normalized by Gapdh. Primers were as follows: Gapdh (forward: 5′-TCAAATGGGG TGAGGCCGGT-3′ and reverse: 5′-TTGCTGACAA TCTTGAGTGA-3′) and Il34 (forward: 5′-CTTTGG GAAACGAGAATTTGGAGA-3′ and reverse: 5′-GCAA TCCTGTAGTTGATGGGGAAG-3′). All experiments were performed in triplicate for each sample.
The ChIP signal was normalized to input. Three biological replicates were performed for each experiment.

In vivo tumor assay
All experimental animals were maintained in our specific animal facility according to institutional guidelines, with protocols which have been approved by the institutional animal care and use committee of Institute for Genetic Medicine of Hokkaido University. B6C3F1 (Japan SLC, Shizuoka, Japan) female mice (6-8 weeks old) were injected subcutaneously with Il34 WT HM-1 or Il34 OE HM-1 cell lines (1 × 10 6 cells). After 7 days, mice were divided into two groups randomly. JQ1 was dissolved in DMSO. The two groups were treated with vehicle control or JQ1 (50 mg/kg 3 times a week) for 3 weeks. During treatment, tumor size was measured 3 times a week, and tumor volumes were calculated as length×width×height. Tumors were then removed from sacrificed mice, photographed, and analyzed by qPCR.

Statistical analysis
Significance was determined by Student's t test. p value was considered statistically significant when < 0.05.

Results
BET inhibitor JQ1 suppresses IL-34 expression in IL-34producing cancer cell lines In order to find possible candidates that could regulate IL-34 expression, we have at first taken a glance at transcriptional changes of IL34 in human cancer cell lines and a tumor tissue treated with several low molecular inhibitors. Data were gathered from GEO profile of NCBI database [29], and we found that among those only BET inhibitor JQ1 showed downregulation of IL34 expression (reference series: GSE51020) (Supplementary figure 1). Thus, we attempted to analyze IL34 expression which seemed to be regulated by JQ1.
We then tested whether JQ1 could really inhibit IL-34 expression in cancer cells. Murine ovarian cancer cell lines, OV3121, OV3121-RAS4, and HM-1, were treated with 10 μM of JQ1 for 2 days. OV3121-RAS4 and HM-1 spontaneously expressed IL-34 while OV3121 did not. Quantitative PCR and ELISA were performed to analyze mRNA and protein expression, respectively. As expected, when treated with JQ1, mRNA as well as protein expression of IL-34 was strongly inhibited in OV3121-RAS4 and HM-1 cells (Fig. 1a), whereas no change was observed in OV3121 cells (Fig. 1a). MTT assay was performed to ensure that IL-34 reduction was not due to decrease of cell viability (Fig. 1a). Cells were observed microscopically, and there were no obvious morphological changes according to the treatment with JQ1 (Fig. 1b). These findings indicate that, for the first time, IL-34 expression can be downregulated with a BET inhibitor JQ1 at both mRNA and protein levels.

JQ1 suppresses Brd4 occupancy at Il34 gene promoter region
It has been shown that JQ1 binds to BRD4 with high affinity and inhibits its binding to RNA POL II in malignant cells [20,30]. Thus, it is required to confirm whether BRD4 can be recruited to the IL34 promoter region. We carried out chromatin immunoprecipitation (ChIP) to confirm the involvement of Brd4 in the regulation of the Il34 gene. According to the Ensembl Genomes database [31], the chromosomal location of mouse Il34 is from 110,741,829 to 110,805,949 on chromosome 8. Untranslated regions are located in exon 1, 2, and 7, and coding domain sequences are located in exon 2 to 7. We determined four possible regions for the Il34 promoter (R1 to R4) (Fig. 2a). ChIP analysis indicated that Brd4 occupancy was enriched at the R2 region that was significantly suppressed by JQ1 treatment (Fig. 2b). These data indicate that Brd4 is directly involved in Il34 gene transcription and regulates Il34 expression mainly by binding to the R2 region.
As Brd4 is known to bind acetylated lysine of H3 [20], we analyzed histone acetylation on several regions of the Il34 gene including the R2 region which was suggested as promoter region (Fig. 2b) and the other regions (R3 and R4). Histone 3 lysine 27 acetylation (H3K27ac) modification was detected in all regions, and JQ1 treatment tended to slightly increase rather than decrease it (Fig. 2c). Further, histone H3 trimethylated at lysine 4 (H3K4me3) modification was analyzed. Similar to the results of H3K27ac modification analysis, H3K4me3 modification was detected at all regions tested, and JQ1  treatment tended to slightly increase rather than decrease it (Fig. 2d). Collectively, our results suggest that regulation of Il34 expression by JQ1 is mainly exerted by inhibition of Brd4 binding to the promoter region rather than by histone modification.

JQ1 treatment suppresses Il34 expression in vivo and impedes tumor growth
JQ1 has been reported to inhibit growth of several types of tumors in vivo [25,26]. However, it has not been known whether the anti-tumor effect of JQ1 was mediated by IL-34 suppression. In our previous research, we have shown that blockade of IL-34 could significantly suppress the growth of IL-34-producing tumors [8]. Therefore, we next sought to test the effect of JQ1 on tumor growth in vivo.
HM-1 is known as aggressive and highly metastatic tumor cells with poor prognosis [32]. As shown above, HM-1 cells spontaneously produce IL-34 (Fig. 1a). We designated an intact HM-1 cell line as Il34 wild-type line (Il34 WT ). And we generated Il34-overexpressing cell line (Il34 OE ) driven by EF1α promotor whose activity is not inhibited by BET inhibitors [33]. These cell lines were treated with JQ1 in vitro, and the Il34 expression was efficiently suppressed in a dose-dependent manner only in Il34 WT but not in Il34 OE (Fig. 3a). Additionally, we carried out ChIP analysis for H3K27ac to demonstrate histone acetylation in the introduced EF1α promoter. H3K27ac modification was detected in the introduced EF1α promoter in Il34 OE , and JQ1 treatment did not affect it (Supplementary figure 2). These results indicate that only endogenous Il34 expression was regulated by Brd4.
Then, tumor cells were inoculated into syngeneic B6C3F1 mice 1 week before the onset of treatment, then tumor-bearing mice were randomly divided into two groups (control or 50 mg/kg of JQ1, 3 times a week) (Fig. 3b). Tumor size was measured at the indicated time points during treatment (Fig. 3b). Notably, there was no significant difference in the tumor growth of Il34 OE HM-1 tumors regardless of JQ1 treatment. In contrast, JQ1 treatment significantly suppressed the tumor growth in the Il34 WT HM-1 group (Fig. 3c). We collected the tumor tissues at the endpoint of observation to analyze Il34 mRNA expression. Consistent with the in vitro findings, Il34 expression in Il34 WT , but not in Il34 OE , HM-1 tumor was remarkably reduced by the JQ1 treatment (Fig. 3e). Based on these results, it is strongly suggested that in vivo treatment with BET inhibitor JQ1 suppressed IL-34 expression in TME which leads to the anti-tumor effect.

Discussion
IL-34 is known to affect proliferation and induction of therapeutic resistance in various types of tumors [5,7,8,11]. Consequently, the regulation of IL-34 expression could be considered as a new therapeutic method. However, the mechanism of IL-34 expression in cancer cells has not yet been clarified. It has been indicated that IL-34 can be induced by various stimuli such as chemotherapy, chemical stressors, pro-inflammatory cytokines, PAMPs, vitamin D, and viral infections in a wide range of cells via an NF-κB-mediated mechanism. On the other hand, IL-34 expression is suggested to be downregulated by other stimuli such as transforming growth factor-beta 1 and bone morphogenetic protein 2 [14]. More recently, IL-34 has been suggested to be highly expressed in human mesangial cells of lupus nephritis patients and is negatively regulated by the Wnt pathway [34]. However, these reports are limited to describe which stimulations or pathways relate to the expression of IL-34, and IL-34 gene transcription regulator has been unknown.
In this study, we identified for the first time a responsible transcription regulator that controls Il34 expression. We found that the BET inhibitor JQ1 efficiently suppressed Il34 expression in cancer cell lines (Figs. 1a  and 3a). JQ1 binds with a high affinity to the first bromodomain of BRD4 [30], and our results using mouse cell lines indicate that Brd4 is recruited to the promoter region of the Il34 gene and controls its expression.
Although JQ1 treatment in HM-1 cells clearly reduced Brd4 binding to the promoter region and Il34 Representative results as mean ± SEM (n = 3) from two independent expreriments. NS, not significant; **p < 0.01; Student's t test. c Percent of input values of H3K27ac in HM-1 cells analyzed by ChIP-qPCR. Genomic DNA from 10 μM of JQ1 or DMSO treated HM-1 cells was immunoprecipitated with anti-H3K27ac antibody or control IgG and amplified by qPCR with Il34 each region's primer pairs. Representative results as mean ± SEM (n = 3) from two independent expreriments. NS, not significant; **p < 0.01; Student's t test. d Percent of input values of H3K4me3 in HM-1 cells analyzed by ChIP-qPCR. Genomic DNA from 10 μM of JQ1 or DMSO treated HM-1 cells was immunoprecipitated with anti-H3K4me3 antibody or control IgG and amplified by qPCR with Il34 each region's primer pairs. Representative results as mean ± SEM (n = 3) from two independent expreriments. NS, not significant; *p < 0.05, **p < 0.01; Student's t test expression, H3K27ac and H3K4me3 status in the region were not significantly changed (Fig. 2). These results indicate that JQ1 exhibited its suppressive function on Il34 expression through inhibiting direct binding but not changing (decreasing) accessibility of Brd4 to the Il34 promoter. Brd4 is known to recruit transcription elongation factor P-TEFb and facilitates transcription elongation [20]. In Il34 gene expression, it is likely that the transcript elongation by Brd4 is impaired by JQ1 treatment.
It has been reported that Brd4 also affect enhancers [23,35] or super-enhancers [35,36]. In Il34 gene, several enhancer regions are suggested according to Ensembl database [37,38]; therefore, it is possible that JQ1 exerts its function as a therapeutic drug by regulating them in Il34 gene. Regarding the super-enhancers, according to SE analysis [39], there are seven super-enhancer regions in the IL34 gene in human cells, and BRD4 binding was observed in at least two regions. In mouse case, ten regions of Il34 super-enhancers have been speculated in a  [40]. Therefore, JQ1 may also inhibit Il34 super-enhancers, in addition to its promoters, and contribute to the therapeutic outcome by regulating Il34 expression.
There are three future tasks to understand the detailed mechanism of IL-34 expression. Firstly, investigation of condition which promotes the binding of Brd4 to Il34 promoter region is needed because the promoting process and factors are still unclear. As mentioned above, some reports suggested several pathways and transcription factors such as RUNX1 or NF-κB are important. Therefore, protein phosphorylation analysis of related factors could explain a process of IL-34 expression. Secondly, the data that BET inhibition did not completely suppress IL-34 expression (Figs. 1a and 3a) implies the existence of other pathways regulating IL-34 expression. Finally, although suppression effect of in vivo tumor growth by Brd4-specific inhibitor JQ1 was shown in Fig. 3c, it is necessary to investigate whether the growth suppression is a solely IL-34-dependent phenomenon because Brd4 regulates a wide range of tumor-associated transcription factors.
In conclusion, we identified here Brd4 as a responsible transcription regulator that controls Il34 expression in cancer cells. The findings may help establish a new therapeutic strategy against IL-34-producing cancer in which usually an immunosuppressive environment is observed. Future studies on primary cancer tissues and cells are needed to strengthen this conclusion.

Conclusion
According to the results, Il34 expression mechanism is suggested to be regulated by Brd4 in IL-34-producing cancer cells, which is efficiently suppressed with JQ1. Treatment of mice bearing IL-34-producing tumors with JQ1 shows anti-tumor effect.