Pancreas-Specic ARID1A Deciency Promotes Acinar-to-Ductal Metaplasia and Elevates Interleukin-6 Expression in Experimental Pancreatitis Model

Background: Although role of ARID1A in pancreatic homeostasis and tumorigenesis has been recently described using genetically engineered mouse (GEM) models, whether ARID1A plays a role in pancreatic inammation and regeneration remains to be explored. Methods: Pancreas-specic Arid1a-decient GEM model (Arid1a def ) was generated by Ela1-Cre/ERT2 mice crossing with Arid1a / mice and characterized histologically. In physiological and inammatory conditions, serum amylase and lipase activity were measured to investigate effects of Arid1a deciency on pancreatic secretion function. Histology analysis of pancreas was used to evaluate pancreatic lesions and recovery. Ex vivo primary acinar cell culture was employed to study acinar-to-ductal metaplasia (ADM) process. In HPNE cells, ARID1A knockdown and histone acetyltransferases inhibitors were used to explore epigenetic regulation on interleukin-6 (IL6) expression. Chromatin immunoprecipitation (ChIP) and quantitative real-time PCR were performed to analyze on IL6 promoters. Results: Arid1a deciency promoted formation of ductal cysts characterized as silenced acinar genes and activated duct genes. Arid1a-decient acinar cells were more inclined to trans-differentiation to ductal cells in cerulein-induced acute pancreatitis (AP) model. Expression analysis of proinammatory cytokines reveals that ARID1A deciency led to increased IL-6 expression in mice acinar cells and HPNE cells. ARID1A-associated histone acetylation partially involved in epigenetic regulation of IL-6. Conclusion: These results demonstrate ARID1A is involved in cerulein-induced AP development by mediating pro-inammatory cytokines IL-6 and suggest that ARID1A-containing SWI/SNF complex is an epigenetic regulator of acute pancreatitis.


Background
Pancreatitis is an in ammatory disease of highly variable severity, beginning with an injury to the pancreas and resulting in an in ammatory response [1]. Genetically engineered mouse models (GEMM) can be a useful tool to study the pathophysiology of pancreatitis. During the in ammatory cascade, response genes seem to be activated by signaling molecules or cytokines [2]. In addition to intricate balance between proin ammatory cytokine (IL-1beta, TNF-alpha, IL-6, IL-8, and platelet activating factor) and anti-in ammatory cytokine (IL-10, TNF-soluble receptors and IL-1 receptor antagonist) [3], epigenetic control mechanisms are being attached great importance in the control of the in ammatory cascade.
Chromatin remodeling complex are associated with epigenetic regulation of both early and late proin ammatory genes in acute pancreatitis, therefore play an important role in the control of the in ammatory cascade [4]. Chromatin remodeling complexes are a group of epigenetic regulators that play a key role in orchestrating chromatin architecture and gene expression by directly altering the assembly of nucleosomes as well as the accessibility of transcription factor to DNA. In particular, the SWI/SNF complex control transcription by activating promoter/enhancer regions by partially regulating acetylated histone H3K27 (K3K27ac) [5,6], where histone acetylation marks are is catalyzed by histone acetyltransferases (HATs) and histone deacetylases (HDACs). More importantly, HATs and HDACs play a central role in transcriptional regulation of genes involved in the in ammatory response [7,8].
ARID1A, a component of SWI/SNF chromatin remodeling complex, has been shown to play important roles in tissue homeostasis and regeneration. ARID1A is the most frequently mutated at variable frequencies across molecular and histological subtypes of cancer [9][10][11]. Within the context of functional models in vitro and in vivo, ARID1A was involved in multiple biological processes, such as proliferation, differentiation, senescence, apoptosis, metastasis, and angiogenesis [12][13][14][15][16][17]. Intriguingly, several investigations reveal important roles of ARID1A inactivation in in ammation-driven tumorigenesis. In human ovarian tissue, ARID1A mutations seem to be an important early event in the malignant transformation of endometriosis to endometrioid and clear cell carcinomas [18]. It is noteworthy that ARID1A inactivation in cooperation with PIK3CA activation promote tumorigenesis through protumorigenic in ammatory cytokine signaling, and proposed that ARID1A protects against in ammationdriven tumorigenesis [19]. In addition, hepatocyte-speci c Arid1a knockout could result in mouse steatohepatitis and tumor development [20]. ARID1A plays important roles in tissue homeostasis and regeneration [21][22][23]. Nevertheless, it remains unclear whether ARID1A de ciency is implicated in pancreatitis and tissue regeneration after injury. Interleukin 6 (IL-6), as a multifunctional cytokine, is closely linked to the most burdened exocrine pancreatic diseases including acute pancreatitis, chronic pancreatitis and pancreatic cancer [24]. Epigenetic mechanisms in uencing IL-6 expression in acute pancreatitis are poorly understood.
Notably, our previous study in vitro suggests that ARID1A knockdown promotes malignant transformation through inhibiting oncogene-induced senescence in human pancreatic cell lines [16].
Furthermore, evidences from both mice PDAC and pancreatitis patients display pancreatitis-induced in ammation can inhibit oncogene-induced senescence [25], prompting us to hypothesize that ARID1A functions in in ammation during pancreatic lesions. Pancreatic acinar cells are with high plasticity, and have capability to dedifferentiate or transdifferentiate to a progenitor-like phenotype that express ductal markers, in a process termed acinar-to-ductal metaplasia (ADM) during pancreatic in ammation or injury in mouse and human tissue [26]. The present work focuses on acute pancreatitis as an experimental model of acute in ammation. In this study, we investigated the roles of Arid1a de ciency in mice pancreas under physiological condition and cerulein-induced acute pancreatitis using a conditional acinar-speci c Arid1a knockout mice model. We further observed the effect of Arid1a de ciency on ADM, proin ammatory cytokines, and epigenetic regulatory mechanisms in the control of the in ammatory cascade. Our ndings highlight ARID1A-mediated epigenetic regulation on in ammatory response and provide an anti-in ammation approach based on epigenetic inhibitors targeting against ARID1A de cient patients.

Materials And Methods
Generation of conditional acinar-speci c loss of Arid1a Mice Cerulein-induced acute pancreatitis (AP) 1 week after the mice were administrated with tamoxifen, acute pancreatitis (AP) was induced by intraperitoneal injections of cerulein (Sigma-Aldrich, Steinheim, Germany) eight times hourly for two consecutive days with a total of 0.1 µg/g body weight per mouse, as previously described [27]. The nal day of injection was considered as day 0. Blood was collected at 1, 3, 5, 10 hours, 1, 5 and 7 days after the last cerulein injection to examine the serum amylase and lipase levels.
Tissue processing and histology Mice were euthanized at indicated time points, pancreata were dissected in ice-cold PBS and separated into multiple fragments. Formalin-xed, para n-embedded (FFPE) pancreas tissue was cut into 6-8 µm sections and stained with H&E. Pancreatic damage was semi-quantitatively assessed by scoring 3-5 random slides per mouse according to quantitative method described in the references [28][29][30]. Brie y, the semi-quantitative scoring system for edema and acinar necrosis were graded from 0 to 3 according to the following criteria: edema: 0, absent; 1, focally increased between lobules; 2, diffusely increased between lobules; and 3, acini disrupted and separated; and acinar necrosis: 0, absent; 1, periductal necrosis (5%); 2, focal necrosis (5-20%); and 3, diffuse parenchymal necrosis (20-50%). Total pancreatitis was calculated by grade of in ammation × distribution of in ammation. The scoring was performed in a blinded manner by an experienced pathologist.
Serological examination of amylase, lipase, and IL-6 cytokine The serum activity of amylases and lipases were measured by enzyme dynamics chemistry using commercial kits, according to the manufacturer's instructions in a Roche/Hitachi modular analytics system (Roche, Mannheim, Germany). Serum IL-6 level was determined by enzyme-linked immunosorbent assay (ELISA) using a commercial kit (Quantikine; R&D Systems, Minneapoils, MN, USA).
Immunohistochemistry and immuno uorescence assay FFPE samples were cut into 5µm sections. After depara nization and rehydration by passage through xylene and a graded alcohol series, heat-induced antigen retrieval was achieved by cooking in antigen unmasking solution (Vector Laboratories). Endogenous peroxidase activity was inactivated by 0.3% hydrogen peroxide treatment for 10 min at room temperature. After blocking with TBS containing 3% BSA, sections were incubated with the primary antibody at 4˚C overnight. For immunohistochemistry, sections were incubated in secondary antibody for 1 hour and then developed using Avidin-Biotin Complex (ABC) and Diaminobenzidine (DAB) kit (Vector Laboratories) and counterstained with hematoxylin. For immuno uorescence assay, slides were subject to primary and secondary antibody incubations, counterstained with DAPI (Sigma-Aldrich; St. Louis, MO).
Isolation and culture of primary acinar cells Primary acinar cells were isolated from mice pancreas using a collagenase digestion as previously described procedure [31,32]. Brie y, pancreata were immediately removed and rinsed twice with ice-cold Hank's Balanced Salt Solution (HBSS, Thermo Fisher Scienti c). Pancreatic tissue was minced into 1-5 mm pieces and digested with collagenase P (Roche Applied Science, Mannheim, Germany) for 15 min at 37 ˚C. Isolated acini were washed three times in ice-cold HBSS containing 5% fetal bovine serum (FBS), and then ltered through 500 and 105-µm nylon meshes (Spectrum Laboratories, CA, USA). Acini were collected by centrifugation and re-suspend in 10 ml Waymouth medium (Invitrogen, Carlsbad, USA) containing 1% FBS, 0.1 mg/ml soybean trypsin inhibitor, 1 μg/ml dexamethasone. Culture dishes were coated with rat tail collagen type І (Sigma-Aldrich) in 0.02 N acetic acid, pH 3.67 solution for 1 hr at 37 ºC, and the isolated primary acini were seed into the prepared plates coated with matrix scaffolds in triplicates. Acinar cells were infected with lentivirus of interest and incubated for 3-5 hours before embedding in the collagen/waymouth media mixture. Numbers of ducts were counted under a microscope. Acinar explants were seeded in triplicates, and cells clusters were counted from at least 3 optical elds/well and reported as a percentage of acinar clusters and ring-like spheres.

RNA isolation and reverse transcription quantitative-PCR (RT-qPCR)
Total mRNA was isolated using Trizol reagent (Ambion) following manufacture's instruction. 1 μg RNA was reverse transcribed using oligo dT primer was performed with iScript TM gDNA Clear cDNA Synthesis Kit (BIO-RAD). Synthesized cDNAs were diluted 2-fold with water, and regions of interest were ampli ed by SYBR ® Green Real-Time PCR Master Mixes (Thermo Fisher Scienti c). Ct values of target genes were normalized using that of housekeeping geneGapdh. In addition, we employed a reference gene Rpl13a used for normalization in qRT-PCR analyses in AP mouse tissue [33]. The primer sequences used in the study are listed in Supplementary Table 1.
Flow cytometric analysis of cell proliferation and apoptosis Cell apoptosis was detected using the Annexin V-FITC/PI Apoptosis Detection Kit (BD Biosciences) following the manufacturer's instructions. Brie y, the cells were collected and suspended in binding buffer, followed by incubation with Annexin V-FITC and PI. The stained samples were loaded on FACSCalibur ow cytometer (BD Biosciences) and produced data were analyzed with CELLQuest software (BD Biosciences). Cell proliferation was analyzed using Click-iT® Plus EdU Proliferation Kits (ThermoFisher) according to the manufacturer's protocol.

ChIP-qPCR
Chromatin immunoprecipitation (ChIP) was performed using the EZ ChIP TM Chromatin Immunoprecipitation Kit (Merck Millipore, Billerica, MA, USA) according to the manufacturer's instruction. Cells were cross-linked with 1% formaldehyde for 15 min and quenched using 125 mM glycine for 5 min.
Real-time quantitative PCR was using PCR master mix (Bio-Rad) and speci c primers for regulatory elements regions of human IL-6 listed in Supplementary Table 1.
Statistical Analysis.
All data from mice experiments are shown as means ± SEM. Statistical comparisons were performed by the two-tailed Student's t-test. Analyses were performed using GraphPad Software (San Diego, CA, USA). p < 0.05 was considered statistically signi cantly.

Results
Conditional Arid1a de ciency in mature acinar compartment had no detectable effects on pancreatic function To determine whether Arid1a de ciency contributes to pancreatic pathogenesis, we generated a tamoxifen-inducible pancreas-speci c Arid1a-de cient GEM model driven by Ela1-Cre/ERT2 targeting acinar cells only (See Methods). To induce Cre-mediated excision of Arid1a exon, mice aged 6-8 weeks were intraperitoneally injected with tamoxifen (Tam), and then mice were sacri ced at 9-week and 40week to harvest pancreata and blood samples, respectively (Fig. 1a). We rst evaluated e ciency of Cremediated Arid1a depletion at 1-week post-Tam treatment. Genotyping analysis of Arid1a alleles showed the 268-bp fragment of excised Arid1a loci was present in pancreata tissue of Arid1a / ; Ela1-Cre/ERT2 mice with Tam treatment, while only oxed-Arid1a alleles were observed in tail and pancreata samples of control mice injected with corn oil (vehicle control) ( Figure S1a). Moreover, Arid1a protein was obviously decreased when bulk pancreata were examined by western blotting ( Figure S1b). Henceforth, Arid1a / ; Ela1-Cre/ERT2 mice injected with tamoxifen were referred to as Arid1a def ; their counterpart treated with corn oil were designated as Arid1a +/+ . Arid1a def mice survived healthily until euthanasia when 40-weekold and exhibited normal body weight. During this time, no signs of any insu ciency in the pancreatic exocrine functions were found in Arid1a +/+ and Arid1a def mice, such as weight loss (Fig. 1b). Moreover, the pancreas/body weight ratio was not signi cantly different between the two groups of mice (Fig. 1c) at the 9-week (p = 0.485) and 40-week (p = 0.962) time points. Histologically, H&E analysis revealed no obvious difference in tissue structure of the pancreas between two groups (Fig. 1d). To further address the role of Aird1a gene in exocrine function of acinar cells, we examined amylase and lipase secretion in vivo over the course of speci ed time periods respectively. The blood amylase activity analysis showed no difference in serum amylase between Arid1a +/+ and Arid1a def mice at the 9-week (p = 0.622) and 40week (p = 0.790) time points (Fig. 1e). In addition, we did not observe signi cant difference of serum lipase between the two groups of mice at the 9-week (p = 0.883) and 40-week (p = 0.858) time points (Fig. 1f). Taken together, these results demonstrated that Arid1a de ciency in mature acinar cells had no obvious effects on pancreatic exocrine function.
Arid1a de ciency promoted ADM formation upon TGFα or Cerulein (Cer) treatment During pancreas injury or in ammation, a reversible ADM metaplasia occurs and deservedly contributes to the regeneration of acinar structures [26]. Since ADM represents a cellular mechanism accounting for regeneration after in ammation or injury, thus we hypothesized that Arid1a plays a role in pancreatic epithelial plasticity mediated by acinar cell metaplasia. We employed ex vivo primary acini culture system to investigate the ADM process. To determine whether Arid1a de ciency contributes to ADM, we initially utilized a de ned explant model [34,35], in which acinar cells were seeded in collagen-coated plates and treated with TGFα to induce acinar trans-differentiation. First, we observed ADM of Arid1a / mouse primary acinar cells that were lentivirally-infected with GFP or Cre. Floxed-Arid1a exon were fully excised and Arid1a protein obviously decreased ( Figure S2a), and TGFα-induced formation of ductal structure was signi cantly increased to approximately 60% when Arid1a was excised by Cre ( Figure S2b, p < 0.01).
Furthermore, acini were also isolated from Arid1a +/+ and Arid1a def mice, and Arid1a de ciency not only had effects on the numbers of ducts formed, but also affected their size (Fig. 2a, p < 0.01). To further con rm acinar conversion to ductal phenotype, we analyzed expression pro ling of acinar and ductal genes using quantitative PCR analysis. When acinar cells were treated with TGFα for 5 days, Arid1a def cells displayed downregulated expression of acinar markers (Amylase, Lipase, Cpa1, Mist1, and Ptf1a) and upregulated expression of ductal genes (Cftr, Krt19, CaII, Nes, Sox9, and Hnf6) compared with Arid1a +/+ controls (Fig. 2b). As demonstrated by western blotting assay, the expression of amylase decreased whereas CK-19 remarkably increased in these Arid1a-de cient cells (Fig. 2c). Thus, these results suggest Arid1a de ciency promote ADM formation of primary acinar cell. To validate whether Arid1a de ciency contributes to ADM in vivo, we observed effects of Arid1a de ciency on ADM in acute pancreatitis (AP) model induced by Cer. Signi cantly, Arid1a de ciency promoted ADM formation of acinar cells in the AP model, as shown by quantitative analysis of H&E staining (Fig. 2d, p < 0.01) and IHC assay for CK-19 staining (Fig. 2e, p < 0.05). Taken together, these data indicate that Arid1a acts as a critical epigenetic mediator to facilitate metaplasia of acinar cells under the context of pancreas injury.

Arid1a de ciency in uenced cell proliferation and exocrine secretion of acinar cells
Amylase and lipase levels derived from pancreatic acinar cells are important index as a re ection of acinar cell function and pancreas regeneration after injury [3,36]. We measured serum amylase and lipase levels at various time points for the AP model, and dynamic monitoring showed a decreasing trend in both amylase and lipase levels of Arid1a def mice over the course of time (Fig. 3a), although signi cant differences of amylase levels between two groups were statistically present at 10-hour (p < 0.01) and 24hour (p < 0.05) time points after Cer treatment. In contrast, Arid1a de ciency sustained signi cant reduced level of lipase at these time points from 10-hour until 7-day (Fig. 3b, 5-day, p < 0.01; 10-hour, 1-day and 7day, p < 0.05). To observe distribution and degree of histologic lesions, we applied a de ned scoring system [29,37] to evaluate tissue injury and regeneration of the AP model. As a result, no signi cant differences between the two groups could be observed in the pancreas weight and edema ( Figure S3a-b). However, Arid1a-de cient acinar cells showed lower grade of necrosis, and a signi cant decrease of necrosis was observed at the 5th day of AP injury/regeneration model ( Figure S3c). Arid1a-de cient mice somehow displayed strong pancreatitis as compared with Arid1a +/+ mice, even though quanti cation assessment of the pancreatitis exhibited a signi cant increase at 3h after cerulein treatment ( Figure S3d). In sum, these histological analysis based on the scoring system were multifactorial and various in the severity of acute pancreatitis [38], which was di cult to unveil Arid1a's exact roles in murine AP model. To further observe effect of Arid1a de ciency on cell proliferation, we performed immuno uorescence for Ki-67 on AP tissue and EdU cell proliferation assays in primary explant culture of primary acinar cells.
Signi cantly, Arid1a-de cient led to a 2-fold increase of Ki-67 positive cells found in AP tissues (Fig. 3c, p < 0.01). Furthermore, we evaluated cell proliferation when primary acinar cells were treated with 10µM Edu for 3 hours. Flow cytometry assay showed that cells labeled by Edu were 28.3 ± 1.5% in primary acinar cells isolated from Arid1a +/+ mice at 3 independent experiments, whereas Edu-labeling cells were 37.1 ± 2.2% and signi cantly increased in cultured acinar cells from Arid1a def mice (Fig. 3d, p < 0.01). Taken together, these results implied that Arid1a de ciency partly impaired pancreatic exocrine function and promoted cellular proliferation under the pathological condition of acute pancreatic injury.

Arid1a de ciency led to enriched histone H3K27ac modi cation on IL-6 gene
To gain insight into the molecular mechanisms how Arid1a was involved in pancreatic in ammatory response, we analyzed cytokine expression levels in AP tissue. We performed qRT-PCR on a subset of cytokine genes, including proin ammatory cytokines IL-6, Ifnγ, Tnfα, Lbp, G-Csf, Ifnα, IL-1α, IL-1β, as well as anti-in ammatory cytokines IL-10, IL-13, Tgfβ, Csf. Among these cytokines examined, IL-6 and IL-10 mRNA were found to be signi cantly elevated in Arid1a def mice, however no signi cant differences were found for other cytokines examined ( Fig. 4a-b, *, p < 0.05; **, p < 0.01). Since IL-6 is strongly linked to the most burdened exocrine pancreatic diseases including acute pancreatitis, chronic pancreatitis and pancreatic cancer [24], we next try to address epigenetic mechanism of IL-6 regulation associated with Arid1a de ciency in in ammatory cytokine signaling. Quantitative analysis of serum IL-6 indeed supported that Arid1a de ciency increased serum level of IL-6 in vivo (Fig. 4c, **, p < 0.01).
To further validate the role of ARID1A de ciency in mediating IL-6 transcription, we established isogenic cell models which were derived from human HPNE with KRAS mutant (KRAS G12D ) as well as CRISPR/Cas9-mediated ARID1A knockout. We identi ed ARID1A wildtype (WT) and knockout (KO) subclones derived from single cell, 2 WT and 2 KO subclones were used to study ARID1A-associated regulatory effects on IL-6. Activation of the p42/p44 (Erk1/2) MAP kinase (MAPK) phosphorylation cascade, a major downstream effector of KRAS signaling as well as ARID1A knockout were con rmed by western blot assay (Fig. 5a). ARID1A knockout notably increased IL-6 expression at protein and mRNA level ( Fig. 5a-b), supporting that upregulation of IL mediated by loss of ARID1A function can be reproduced in this context of human cells. According to these observations, we hypothesized that ARID1A de ciency remodels H3K27ac status on IL-6 gene locus. To test this hypothesis, we rst analyzed the H3K27ac ChIP-seq datasets in 7 available human cell lines using a website tool from ENCODE at UCSC (https://genome.ucsc.edu/encode/). We found that H3K27ac binding sites conservatively enriched in the IL6 proximal and distal regions ( Figure S4a). Subsequently, we performed ChIP using H3K27ac antibody, and qPCR analysis revealed that a signi cant enrichment of H3K27 acetylation at the IL-6 distal regions upon ARID1A knockout (Fig. 5c). As known, histone acetylation regulation is speci cally controlled by histone deacetylases (HDACs) and histone acetyltransferases (HATs). Histone acetylation by HAT activates in ammatory genes, whereas increased HDAC activity represses in ammatory genes [39], which implies HATs play a leading role in the H3K27 acetylation enrichment on IL-6 gene. Thus, we employed CBP30 and JQ1, two common HAT inhibitors, to inhibit the catalytic activity of HATs [40,41]. Expectedly, HAT inhibition led to a signi cant reverse IL-6 expression in ARID1A knockout subline cells relative to wildtype cells ( Fig. 5d and Figure S4b, **, p < 0.01). Collectively, these results demonstrated that ARID1A knockout remodeled HATs-mediated histone acetylation on IL-6 gene, and thus elevated IL-6 expression.

Discussion
Epigenetic mechanisms, especially chromatin remodeling plays an emerging role in in ammation. Acute pancreatitis is a common digestive disease, and its onset is triggered by acinar events that induce autodigestion and acinar cell injury [42]. In ammatory processes have emerged as critical regulators of pancreatic tumorigenesis, and induction of pancreatitis accelerates PDAC development. Recurrent acute pancreatitis seems to increase the risk for developing pancreatic cancer. Epigenetic markers such as histone acetylation and methylation, as well as recruitment of SWI/SNF remodeling complex are associated with the upregulation with the upregulation of both early and late proin ammatory genes in acute pancreatitis [4]. As a subunit of the SWI/SNF chromatin remodeling complex, ARID1A is thought to impose upon the genetic code a chromatin structure characterized by chromatin accessibility, nucleosome position, and histone modi cation [10]. As known, ARID1A was recurrently found to be inactivated in human pancreatic ductal adenocarcinoma (PDAC). In this study, we revealed that Arid1a de ciency promoted endows acinar cells with more plasticity to stand up against pancreas injury. and enhanced pancreatitis-associated in ammation response through increased H3K27ac-mediated IL-6 expression, Recent studies have shown that role of ARID1A in tumorigenesis is highly dependent upon context, based on genetically engineered mouse (GEM) models. For instance, a hepatocyte speci c Arid1a knockout mouse model showed Arid1a de ciency were resistant to tumor initiation while ARID1A overexpression accelerated initiation. In contrast, Arid1a loss in established tumors accelerated progression and metastasis [17]. Another study revealed the paradoxical role of Arid1a in a mouse model with pancreas-speci c Arid1a loss [43], supporting the protective role of Arid1a de ciency in pancreatic injury. In ammation contributes to mouse PDAC, and therefore anti-in ammatory treatment may reduce the risk of developing PDAC [25]. Signi cantly, Interleukin 6 (IL-6) was acknowledged to participate in cell growth, differentiation and function adjustment and play a vital role in the in ammatory response and tumor growth in a paracrine fashion. IL-6 is required for the maintenance and progression of pancreatic cancer, therefore inhibition of IL-6 may have therapeutic potential for treatment of pancreatic cancer [44].
Acetylation of histone is associated with an "open" chromatin conformation representing activation of transcription, and Apparently, increased histone acetyltransferase activity or enriched local binding of regulatory elements will increase histone acetylation, leading to transcriptional activation. For example, the promoters of several pro-in ammatory cytokines (IL-1,IL-2, IL-8, and IL-12) were acetylated by CBP/p300, so that the cytokines are rapidly induced to response for in ammation [8].
In the study, we demonstrated that ARID1A could bind to distal and proximal regulatory elements of IL-6 gene, and thus affect their acetylation modi cation due to alternations of chromatin accessibility. Together, these evidences uncovered that epigenetic regulatory mechanism of IL-6 in pancreatic acinar cells. Hence, inhibition of IL-6 may have therapeutic potential for treatment of cancers characterized by oncogenic Ras mutations [45]. More practically, the highly selective inhibitors of the bromodomains of CBP/p300, JQ1 and CBP30, successfully inhibited IL-6 expression in human HPNE cells. recent investigation in macrophages demonstrated the increased binding of p300HAT to the promoter regions of IL-6 could elevate gene transcription through acetylation of histone

Availability of data and materials
The data used during the current study are available from the corresponding author upon reasonable request.
Ethical approval and consent to participate All procedures performed in the study involving animals were in accordance with the guidelines of the Institutional Animal Care and Use Committee at Taizhou University Hospital.