Effect of Docosahexaenoic Acid on Ca2+ Signaling Pathways in Cerulein-Treated Pancreatic Acinar Cells, Determined by RNA-Sequencing Analysis

Intracellular Ca2+ homeostasis is commonly disrupted in acute pancreatitis. Sustained Ca2+ release from internal stores in pancreatic acinar cells (PACs), mediated by inositol triphosphate receptor (IP3R) and the ryanodine receptor (RyR), plays a key role in the initiation and propagation of acute pancreatitis. Pancreatitis induced by cerulein, an analogue of cholecystokinin, causes premature activation of digestive enzymes and enhanced accumulation of cytokines and Ca2+ in the pancreas and, as such, it is a good model of acute pancreatitis. High concentrations of the omega-3 fatty acid docosahexaenoic acid (DHA) inhibit inflammatory signaling pathways and cytokine expression in PACs treated with cerulein. In the present study, we determined the effect of DHA on key regulators of Ca2+ signaling in cerulein-treated pancreatic acinar AR42 J cells. The results of RNA-Sequencing (RNA-Seq) analysis showed that cerulein up-regulates the expression of IP3R1 and RyR2 genes, and that pretreatment with DHA blocks these effects. The results of real-time PCR confirmed that DHA inhibits cerulein-induced IP3R1 and RyR2 gene expression, and demonstrated that DHA pre-treatment decreases the expression of the Relb gene, which encodes a component of the nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) transcriptional activator complex, and the c-fos gene, which encodes a component of activator protein-1 (AP-1) transcriptional activator complex. Taken together, DHA inhibits mRNA expression of IP3R1, RyR2, Relb, and c-fos, which is related to Ca2+ network in cerulein-stimulated PACs.


Introduction
Acute pancreatitis is an inflammatory disease that results in organ dysfunction. This disease is caused by abnormal activation and release of digestive enzymes and it is characterized by increased cytokine release and oxidative stress [1]. Elevated levels of inflammatory cytokines and free radical products are commonly observed in animal models of acute pancreatitis as well as in individuals inflicted by the disease [2,3]. It has been previously proposed that the excessive rise in the intracellular Ca 2+ concentration ([Ca 2+ ]i) observed in pancreatic acinar cells (PACs), the functional units of the exocrine pancreas, triggers the onset of acute pancreatitis [4]. The disruption of calcium signaling in PACs that occurs during acute pancreatitis is well characterized [5,6].
The gastrointestinal peptide cholecystokinin (CCK) causes the release of pancreatic digestive enzymes and the growth of the normal pancreas. However, an abnormally high level of plasma CCK has been reported to occur in patients with chronic pancreatitis [7] and acute pancreatitis [8]. When cholecystokinin (CCK) binds to the CCK receptor, which is a G protein-coupled receptor, IL-1β [32,33]. In cerulein-induced acute pancreatitis, DHA suppresses IL-1β and IL-6 gene expression by inhibiting AP-1 activation [34].
Therefore, we hypothesize that DHA inhibits Ca 2+ signaling and its related signaling pathways in cerulein-stimulated rat pancreatic acinar AR42J cells. To identify the regulators involved in cerulein-mediated changes in Ca 2+ signaling, and to examine the effect of DHA treatment on cerulein-induced alterations in Ca 2+ signaling and related pathways, we carried out RNA-Sequencing (RNA-Seq) analysis on the AR42J cells.
DHA was dissolved in ethanol (0.5 M solution) (Sigma, St. Louis, MO, USA) and cerulein was dissolved in PBS containing 0.1% BSA (10 −4 M) [29]. AR42J cells (1 × 10 5 /mL) were treated with a vehicle 0.5 M ethanol (designated as 'none') or DHA (50 µM) dissolved in 0.5 M ethanol (designated as 'DHA') for 6 h. For cerulein treatment, the cells were pretreated with 0.5 M ethanol (designated as 'cerulein') or DHA (50 µM) dissolved in 0.5 M ethanol for 2 h and then stimulated with cerulein (10 −8 M) (designated as 'cerulein + DHA') for 4 h. The indicated dose and incubation time of DHA were chosen based on our previous study reporting anti-oxidative and anti-inflammatory effects of DHA on cerulein-stimulated AR42J cells [29]. Further, 50 µM of DHA alone used in the present study had no effect on cell response such as cytokine expression and inflammatory signaling in AR42J cells [29] and cell viability in pancreatic cancer PANC-1 cells for 24 h culture [35].

Preparation of Total RNA Extracts and Library Construction
Total RNA was extracted from cells harvested from culture dishes using TRI reagent (Molecular Research Center, Cincinnati, OH, USA) and then purified using the RNeasy MinElute Cleanup Kit (Quiagen, Valencia, CA, USA) according to the manufacturer's instructions. The quality of the total isolated RNA was determined by measuring the concentration of total RNA in the extracts from three replicates, using NanoQuant Infinite M200 instrument (Tecan, Männedorf, Switzerland). The total RNA from each sample was pooled for RNA-Seq library construction using TruSeq RNA Sample Prep Kit (Illumina, San Diego, CA, USA). Briefly, cDNA libraries were prepared by purification from 1 µg total RNA, random fragmentation, and reverse transcription.

RNA-Sequencing and Bioinformatics Analysis
The total RNA library was subjected to transcriptome sequencing. The sequencing was carried out with Macrogen (www.macrogen.co.kr; Seoul, Korea). CLRNASeqTM software (Chunlab, www.chunlab.com; Seoul, Korea) was used for the initial data processing. Raw RNA-Seq reads were trimmed with a quality cutoff of Q30 and the gene expressions were quantified using Cufflinks (https://omictools.com/cufflinks-tool). The RNA-Seq data were normalized by using the relative log expression (RLE) value.
Changes in gene expression of the cells with neither DHA treatment nor cerulein stimulation (treated with 0.5 M ethanol vehicle only) (none), those treated with cerulein, those with DHA treatment alone, and those pretreated with DHA and treated with cerulein (cerulein + DHA) were compared. Differential gene expression analysis was performed using the RNA-Seq analysis module of the CLRNASeq software, with a cutoff set at gene expression of >100 and p-value of <0.05. Transcripts with fold change of >1.5 were included as differentially expressed genes (DEGs). After identifying the DEGs, gene ontology (GO) analysis was performed using the DAVID bioinformatics program (https:

Statistical Analysis
Statistical analysis was performed using the edgeR test and the one-way ANOVA technique. The results were expressed as the mean ± S.E. of three independent experiments. A p-value of 0.05 or less was considered statistically significant.

DHA Inhibits Cerulein-Induced Changes in the Transcriptomic Profile of AR42J Cells
In order to identify substances modulated by cerulein and to examine the effect of DHA on cerulein-induced transcriptomic alterations, RNA-Seq analysis was performed. RNA-Seq analysis identified 339 genes with differential expression between non-stimulated and cerulein-stimulated cells, 181 genes differentially expressed between vehicle-treated and DHA-treated cells, and 116 differentially expressed genes between cerulein-stimulated cells and DHA-treated and cerulein-stimulated cells. Among these genes, 76 genes showed significant differential expression between non-stimulated, cerulein-stimulated, and DHA-treated and cerulein-stimulated cells, with gene expression levels of >100, fold changes of >1.5, and p values of <0.05. The heatmap of the filtered DEGs is shown in Figure 1. Out of the significant DEGs, four genes showed differential expression following DHA treatment compared to non-treated cells: Pqlc3 known as a protein coding gene, Gtpbp6 whose function is not yet known, Slc15a1 required for transport of peptides across membranes, and an antioxidant enzyme gene NQO1.

DHA Suppresses Cerulein-Induced Alteration in the Calcium Signaling Pathway, Determined by Functional Annotation and Pathway Analysis of DEGs
The DEGs listed in Table 1 were used as inputs to perform GO term analysis within the DAVID bioinformatics tool suite. The annotation results were categorized under biological process, molecular function, and cellular function. Gene annotation analysis revealed that the DEGs in the cells treated with cerulein are mostly annotated 'biological process' among the three GO categories. The common GO terms were metabolic pathways, oxytocin signaling pathway, pancreatic secretion, regulation of actin cytoskeleton, MAPK signaling pathway, cAMP signaling pathway, calcium signaling pathway, and apoptosis. The functions of the impacted pathways and the names of the key pathway genes are listed in Table 2. Next, KEGG pathway analysis was performed to identify the physiological pathways in AR42J cells treated with cerulein. DEGs listed in Table 1, whose expression level was increased by cerulein but decreased by DHA, were assessed. Among these, we focused on Ca 2+ signaling pathway. The expression of the respective genes encoding the calcium signaling mediators RyR2 and IP3R1 gene expression was notably up-regulated in cells treated with cerulein, but down-regulated in cells pre-treated with DHA prior to exposure to cerulein. KEGG pathway analysis revealed that the effect of DHA on cerulein-stimulated PACs may involve alteration of several signaling pathways that are downstream of calcium release from the ER, such as the MAPK pathway, protein kinase A (PKA) pathway, nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) pathway, and AP-1 pathway (Figure 2). Mylk2 is involved in Ca2+ signaling pathway according to the KEGG Mapper, but it has its main role in vascular contraction that is different from pancreatic acinar cell function. Thus, some genes including Myk2, whose functions are unrelated to pancreatic function, were not included in Figure 2.

DHA Inhibits Expression of RyR2, IP3RI, Relb, and c-fos Genes in Cerulein-Treated AR42J Cells, Determined by Real-Time PCR
In order to validate the inhibitory effect of DHA on the cerulein-induced alterations identified by RNA-Seq analysis, several candidate genes were selected from those assigned to the calcium signaling pathway and subjected to real-time PCR confirmation. The results are in agreement with the RNA-Seq results. Cerulein induced expression of the genes RyR2 and IP3R1 by up to 3-fold, which was suppressed by pretreatment of DHA ( Figure 3A). Cerulein increased expression of the Relb gene, a member of the NF-κB gene family, and c-fos gene, which encodes a component of the AP-1 complex ( Figure 3B). DHA inhibited cerulein-induced expression of the Relb and c-fos genes ( Figure 3B). DHA treatment alone did not affect mRNA expression of RyR2 , IP3R1, RelB, and c-fos in PACs (Figure 3).

DHA Inhibits Expression of RyR2, IP3RI, Relb, and c-fos Genes in Cerulein-Treated AR42J Cells, Determined by Real-Time PCR
In order to validate the inhibitory effect of DHA on the cerulein-induced alterations identified by RNA-Seq analysis, several candidate genes were selected from those assigned to the calcium signaling pathway and subjected to real-time PCR confirmation. The results are in agreement with the RNA-Seq results. Cerulein induced expression of the genes RyR2 and IP3R1 by up to 3-fold, which was suppressed by pretreatment of DHA ( Figure 3A). Cerulein increased expression of the Relb gene, a member of the NF-κB gene family, and c-fos gene, which encodes a component of the AP-1 complex ( Figure 3B). DHA inhibited cerulein-induced expression of the Relb and c-fos genes ( Figure 3B). DHA treatment alone did not affect mRNA expression of RyR2, IP3R1, RelB, and c-fos in PACs (Figure 3). by RNA-Seq analysis, several candidate genes were selected from those assigned to the calcium signaling pathway and subjected to real-time PCR confirmation. The results are in agreement with the RNA-Seq results. Cerulein induced expression of the genes RyR2 and IP3R1 by up to 3-fold, which was suppressed by pretreatment of DHA ( Figure 3A). Cerulein increased expression of the Relb gene, a member of the NF-κB gene family, and c-fos gene, which encodes a component of the AP-1 complex ( Figure 3B). DHA inhibited cerulein-induced expression of the Relb and c-fos genes ( Figure 3B). DHA treatment alone did not affect mRNA expression of RyR2 , IP3R1, RelB, and c-fos in PACs (Figure 3).

Discussion
Acute pancreatitis occurs upon autodigestion of the pancreas by digestive enzymes and the induction of the inflammatory response. Disturbances in calcium homeostasis is fundamental to the pathophysiology of acute pancreatitis [36,37]. Calcium homeostasis mediators IP3Rs and RyRs have long been implicated in acute pancreatitis. Inhibitors of IP3Rs and RyRs have proven to be effective in reducing zymogen activation, acinar cell death, and proinflammatory cytokine generation in experimental acute pancreatitis [38,39]. IP3Rs and RyRs trigger the initial release of Ca 2+ from internal stores such as the ER, and supraphysiological stimulation results in a persistent increase in Ca 2+ release. This leads to depletion of the internal Ca 2+ stores, which in turn activates the CRAC channels.
Compared to other types, PACs are relatively ineffective in maintaining low intracellular Ca 2+ extrusion and are especially mal-equipped for [Ca 2+ ]i clearance [36,40]. In the absence of voltagegated Ca 2+ channels, calcium signaling primarily depends on internal Ca 2+ stores [41]. Therefore, PACs are particularly vulnerable to the elevation in intracellular Ca 2+ spikes.
In the present study, RNA-Seq analysis of cerulein-treated AR42J cells suggests that cerulein may increase the level of intracellular calcium by up-regulating genes encoding the calcium releasemediating channels, IP3R1 and RYR2. Pretreatment of cells with DHA decreases the expression of the IP3R1 and RYR2 genes up-regulated by cerulein. Therefore, DHA may normalize the ceruleininduced abnormal Ca 2+ wave in AR42J cells. Functional analysis of molecular pathways suggests that cerulein, via transcriptional regulation of Ca 2+ release channels, may modulate signaling regulators related to the MAPK, NF-κB, and AP-1 pathways, which may be suppressed by DHA in PACs.
Experimental validation using real time-PCR confirmed the effect of DHA in reducing ceruleininduced increased expression of the IP3R1 and RYR2 genes. The results obtained from real time-PCR analysis also demonstrate that cerulein increases expression of genes encoding the transcription factors NF-κB and AP-1, and that DHA reduces cerulein-induced expression of the Relb gene, which encodes a component of the NF-κB complex, and the c-fos gene, which encodes a component of the AP-1 complex.

Discussion
Acute pancreatitis occurs upon autodigestion of the pancreas by digestive enzymes and the induction of the inflammatory response. Disturbances in calcium homeostasis is fundamental to the pathophysiology of acute pancreatitis [36,37]. Calcium homeostasis mediators IP3Rs and RyRs have long been implicated in acute pancreatitis. Inhibitors of IP3Rs and RyRs have proven to be effective in reducing zymogen activation, acinar cell death, and proinflammatory cytokine generation in experimental acute pancreatitis [38,39]. IP3Rs and RyRs trigger the initial release of Ca 2+ from internal stores such as the ER, and supraphysiological stimulation results in a persistent increase in Ca 2+ release. This leads to depletion of the internal Ca 2+ stores, which in turn activates the CRAC channels.
Compared to other types, PACs are relatively ineffective in maintaining low intracellular Ca 2+ extrusion and are especially mal-equipped for [Ca 2+ ]i clearance [36,40]. In the absence of voltage-gated Ca 2+ channels, calcium signaling primarily depends on internal Ca 2+ stores [41]. Therefore, PACs are particularly vulnerable to the elevation in intracellular Ca 2+ spikes.
In the present study, RNA-Seq analysis of cerulein-treated AR42J cells suggests that cerulein may increase the level of intracellular calcium by up-regulating genes encoding the calcium release-mediating channels, IP3R1 and RYR2. Pretreatment of cells with DHA decreases the expression of the IP3R1 and RYR2 genes up-regulated by cerulein. Therefore, DHA may normalize the cerulein-induced abnormal Ca 2+ wave in AR42J cells. Functional analysis of molecular pathways suggests that cerulein, via transcriptional regulation of Ca 2+ release channels, may modulate signaling regulators related to the MAPK, NF-κB, and AP-1 pathways, which may be suppressed by DHA in PACs.
Experimental validation using real time-PCR confirmed the effect of DHA in reducing cerulein-induced increased expression of the IP3R1 and RYR2 genes. The results obtained from real time-PCR analysis also demonstrate that cerulein increases expression of genes encoding the transcription factors NF-κB and AP-1, and that DHA reduces cerulein-induced expression of the Relb gene, which encodes a component of the NF-κB complex, and the c-fos gene, which encodes a component of the AP-1 complex.
Several studies have suggested that a link exists between the prolonged increase in [Ca 2+ ]i and up-regulation of the genes encoding NF-κB and AP-1. Elevated [Ca 2+ ]i and Ca 2+ -dependent activation of PKC may stimulate activation of AP-1 and NF-κB. The causal association of NF-κB activation and Ca 2+ signaling is well known [42]. Stimulation with high-dose cerulein activates NF-κB in the pancreatic acini via increased [Ca 2+ ]i and activation of PKC [43]. Calcium blockers and chelators abrogate cerulein-induced NF-κB activation and IκB degradation in PACs [44].
Abnormal Ca 2+ signaling is also indicated to be relevant to AP-1 activation [45]. Carbachol, a secretagogue similar to CCK, increases expression of the genes c-fos and c-jun, which encode AP-1. The increased expression is mediated by changes in [Ca 2+ ]i, calmodulin, PKC, and MAPK [46]. Ramnath et al. [47] demonstrated that the rise in [Ca 2+ ]i and Ca 2+ -dependent activation of PKC mediates AP-1 activation and subsequent chemokine production in substance P-induced experimental acute pancreatitis. Chelating cytosolic calcium, inhibition of PLC, and inhibition of Ca 2+ -dependent PKC each block substance P-induced [Ca 2+ ]i elevation, activation of AP-1, and chemokine production in PACs.
DHA has been suggested to positively modulate Ca 2+ signaling in the ER [48]. Begum et al. [49] revealed that DHA inhibits excess Ca 2+ release from the ER and further represses store-operated Ca 2+ overload in astrocytes, specifically by blocking IP3R. In cardiac myocytes, DHA significantly suppresses the Ca 2+ spark from SR, and inhibited RyR activity [50,51].
DHA treatment also appears to be effective in suppressing inflammatory mediators. DHA decreases production of pro-inflammatory cytokines, such as TNF-α, IL-6, and IL-1β, by inhibiting phosphorylation of NF-κB subunit p65 in bovine mammary epithelial [52] and cardiac cells [53]. In ischemic rats, DHA decreases the level of pro-inflammatory cytokines and chemokines, and inhibits phosphorylation of c-jun and AP-1 DNA binding activity [54].
As shown in Table 2, c-fos is involved in cAMP signaling and MAPK signaling, while Relb is relevant to MAPK signaling. c-Fos gene contains a Ca 2+ /cAMP response element (Ca 2+ /CRE), and its expression and post-translational modification is, in part, affected by Ca 2+ and cAMP level [55]. The elevation of cAMP further enhances expression of c-fos in cardiac myocytes [56] and myeloid leukemia cells [57]. c-Fos also possesses a serum response element (SRE), and MAPK activation induces c-fos expression via SRE [58,59]. Moreover, Relb is a member of the NF-κB family, which demands synergistic activation of MAPK signaling pathways for symptomatic and pathologic inflammatory response [60,61]. Therefore, cerulein-induced increase in c-fos and Relb may be related to activation of cAMP-PKA pathway, MAPK pathway, NF-κB, and AP-1 as shown in Figure 2. Thus, the inhibitory effect of DHA on Ca 2+ signaling may also be associated with suppression of PKA, MAPK, NF-κB, and AP-1 pathways. Further studies should be performed to assess the effects of DHA on PKA, MAPK, NF-κB, and AP-1 pathways in cerulein-stimulated PACs.
The main findings of this study are that DHA inhibits mRNA expression of IP3R1, RyR2, Relb, and c-fos, which is related to Ca 2+ signaling in cerulein-stimulated PACs. Further studies should be performed exploring whether DHA affects RyR, IP3R, Relb, and c-FOS at post-transcriptional level and their functional roles involving the calcium network to determine the effect of DHA on cerulein-induced acute pancreatitis.

Conclusions
DHA inhibits mRNA expression of IP3R1 and RyR2 which are upregulated in cerulein-stimulated PACs. Cerulein increases expression of the Relb gene, a member of the NF-κB gene family, and c-fos gene, a component of the AP-1 complex. DHA inhibits cerulein-induced expression of the Relb and