Weak acids induce PGE2 production in human oesophageal cells: novel mechanisms underlying GERD symptoms

The role of weak acids with pH values in the range of 4–7 has been implicated in the symptoms of gastroesophageal reflux disease (GERD). Prostaglandin E2 (PGE2) is associated with heartburn symptom in GERD patients; however, the precise productive mechanisms remain unclear. In this study, we revealed that exposure to weak acids increases PGE2 production with a peak at pH 4–5, slightly in human normal oesophageal cells (Het-1A), and robustly in oesophageal squamous carcinoma cells (KYSE-270). Release of PGE2 from the oesophageal mucosa was augmented by weak acid treatment in rat. Chenodeoxycholic acid (CDCA), a bile acid, upregulated cyclooxygenase-2 (COX-2) expression in Het-1A and KYSE-270 and induced PGE2 production in KYSE-270 cells. Weak acid-induced PGE2 production was significantly inhibited by cytosolic phospholipase A2 (cPLA2), ERK, and transient receptor potential cation channel subfamily V member 4 (TRPV4), a pH-sensing ion channel, inhibitors. Hangeshashinto, a potent inhibitor of COX-2, strongly decreased weak acid- and CDCA-induced PGE2 levels in KYSE-270. These results indicated that weak acids induce PGE2 production via TRPV4/ERK/cPLA2 in oesophageal epithelial cells, suggesting a role in GERD symptoms like heartburn. Interventions targeting pH values up to 5 may be necessary for the treatment of GERD.

Gastroesophageal reflux disease (GERD) is an inflammatory disease of the upper gastrointestinal tract characterised by heartburn and acid regurgitation. Damaged oesophageal mucosa caused by acid reflux may develop into Barrett's oesophagus, which is a major risk factor for oesophageal adenocarcinoma 1,2 . Since the extent of mucosal damage is proportional to the acid reflux time in patients with GERD 3-6 , reflux with pH values < 4 is defined as "acid reflux" and is used as an index for the diagnosis and treatment of GERD. Advanced techniques for the measurement of oesophageal pH have shown that reflux events with values other than pH < 4, such as weak acids (i.e., pH 4-7), are possible responsible substances related to various GERD symptoms 7 . Several studies have reported that weak acids contribute to the pathogenesis of GERD symptoms in patients with proton pump inhibitor (PPI)-refractory GERD, which are considered problematic among the majority of clinicians [8][9][10][11][12] . Despite recent evidence for the role of weak acid reflux, the degree of risk and mechanisms by which it leads to GERD-related symptoms remain unclear.
Prostaglandin E 2 (PGE 2 ), an inflammatory mediator, is involved in various inflammatory diseases. Cyclooxygenase 2 (COX-2), a rate-limiting enzyme in PGE 2 production, is overexpressed in dysplastic lesions of the oesophagus. In GERD patients, nonsteroidal anti-inflammatory drugs effectively inhibit heartburn symptoms 13,14 . Indeed, increased PGE 2 production by acid exposure has previously been reported in healthy volunteers 15 . However, although bile induces PGE 2 production in oesophageal epithelial cells by enhancing COX-2 expression [16][17][18] , few studies have examined the mechanisms of acid-stimulated PGE 2 production. Especially, detailed investigations of the relationship between pH and PGE 2 production in oesophageal epithelial cells are lacking.
In this study, we used several types of oesophageal epithelial cells and an animal model to investigate extracellular pH-dependent PGE 2 production and the underlying mechanisms, with a focus on weak acids in comparison with bile acid stimulation. We also examined the effect of a Kampo medicine, hangeshashinto (HST), which reportedly alleviates reflux-related symptoms in patients with PPI-refractory GERD 19 , on PGE 2 production. Weak acid treatment induces PGE 2 production from rat oesophageal mucosa in vivo. Experiments were conducted using an animal model to examine whether weak acid also induces PGE 2 production in vivo; we found that perfusion of a weak acid into the oesophageal lumen significantly increased PGE 2 release in the perfusate of rats (Fig. 2a). After exposure to a weak acid, the oesophageal mucosa exhibited a marked increase in PGE 2 production capacity compared with that in the PBS-treated group (Fig. 2b).

Involvement of cyclooxygenases in PGE 2 production in oesophageal epithelial cells. KYSE-
270 was used to investigate the mechanism by which PGE 2 is produced by "weak acid stimulation" and pH 4.5 since there was a sufficient PGE 2 production without prominent cell death. COX-2 mRNA levels in KYSE-270 cells were markedly higher than those in Het-1A cells, whereas cyclooxygenase-1 (COX-1) expression was slightly higher in KYSE-270 than in Het-1A (Fig. 3a). In a western blot analysis, COX-2 protein expression levels in KYSE-270 cells were also markedly higher than those in Het-1A cells (Fig. 3b). In KYSE-270 cells, the induction of PGE 2 production by pH 4.5 medium and CDCA was inhibited by the selective COX-2 inhibitor NS-398 (  (Fig. 4a). A similar phenomenon was observed in Het-1A cells (Fig. 4b). We then assessed the involvement of phospholipase A2 (PLA2), a synthetic enzyme of the cyclooxygenase substrate. The selective cytosolic PLA2 (cPLA2) inhibitor pyrrophenone inhibited

Involvement of MAPKs in PGE 2 production induced by pH 4.5 medium and CDCA. Since
MAPKs are involved in COX-2 expression and PLA2 activation 21-23 , we assessed their roles in PGE 2 production to clarify the mechanisms underlying the observed effects of weak acids and bile acid. Among several MAPK inhibitors, we found that an ERK inhibitor (FR180204) prominently suppressed PGE 2 production induced by pH 4.5 (FR180204 [0.2 μmol/L]; P = 0.00000020, FR180204 [01 μmol/L]; P = 0.00000020) but it had little effect on CDCA induced production (FR180204 [0.2 μmol/L]; P = 0.98, FR180204 [01 μmol/L]; P = 0.00044) (Fig. 5a). ERK phosphorylation level increase 5 min following treatment with pH 4.5 ( Supplementary Fig. S2); simultaneously, cPLA2 phosphorylation levels were also increased (Fig. 5b). Moreover, ERK inhibitor (FR180204) suppressed cPLA2 phosphorylation under stimulation at pH 4.5 (Fig. 5c), indicating, at least in part, the involvement of ERK/cPLA2 in the PGE 2 production by weak acid stimulation. Extracellular acid-sensing mechanism in oesophageal epithelial cells. In addition to the activation of ERK, an increase in intracellular calcium is essential for cPLA2 activation 24 . We confirmed that intracellular calcium increased in response to weak acid but not CDCA ( Supplementary Fig. S3). We also investigated the expression levels of ion channels associated with acid-sensing and the regulation of calcium influx in normal cells and carcinoma cells. We found that TRPV4 expression was substantially higher in KYSE-270 cells than in Het-1A cells (P = 0.0012) (Fig. 6a). Moreover, two TRPV4 inhibitors, RN-1734 and HC067047, suppressed  (Fig. 6b). ERK phosphorylation was slightly increased by TRPV4 agonist treatment ( Supplementary Fig. S4), indicating the involvement of TRPV4/ERK/cPLA2 pathway in the weak acid-induced PGE 2 production, at least in part.

Discussion
The development of an excellent surgical animal model for GERD has contributed to the elucidation of the pathophysiological mechanisms underlying GERD 25 , but many unresolved issues remain. In particular, there has been little progress in research on the effect of reflux materials with a range of pH values on GERD symptoms owing to the limitations of animal models for such detailed investigations. In the present study, we demonstrated for the first time that PGE 2 production increases in oesophageal epithelial cells in response to a narrow pH range of pH 4-5 via TRPV4/ERK/cPLA2 (Fig. 8). These results indicate that weak acids (i.e. pH 4-5) could contribute to GERD symptoms like heartburn via PGE 2 production. PGE 2 is involved in the induction of heartburn symptoms 14,26 . Interestingly, heartburn symptoms were most frequently reported when weak acid was refluxed, especially at pH 5, in patients with PPI-refractory GERD 10,11 . Moreover, the administration of weak acids (pH 4-5) induced heartburn symptoms in nearly 50% of patients with GERD symptoms 27 . Although increase in PGE 2 production by oesophagus acid exposure is reported in healthy volunteers, the relation between PGE 2 production and precise extracellular pH has not been fully investigated 15 . In the present study, we demonstrated that weak acids, at pH 4-5, significantly induced the production of PGE 2 in human oesophageal squamous epithelial cell carcinoma (KYSE-270). Similar results were obtained in normal oesophageal epithelial squamous cells (Het-1A) and normal rat oesophageal mucosa (in vivo), suggesting a widely applicable phenomenon in oesophagus epithelial cells. Our results indicate that excessive PGE 2 production by oesophageal epithelial cells induced by weak acids (pH 4-5) may explain heartburn symptoms observed in patients with PPI-refractory GERD. Moreover, we found that PGE 2 production increased as pH decreased from 4.7, peaked at pH 4.4, and gradually reduced thereafter due to increased cytotoxicity in KYSE-270 cells. Until now, acid reflux in the oesophagus with pH values < 4 has been a focus of GERD diagnosis, and reducing the reflux time with pH < 4 has been considered important in PPI therapy 28 . However, our data suggest that careful attention should be paid not only to acid reflux with pH < 4 but also to weak acid reflux with pH 4-5. PGE 2 is suggested to be involved in the exacerbation of various gastrointestinal cancers, including oesophageal cancer 29,30 . However, there are no reports regarding the possible involvement of weak acid reflux in oesophageal cancer. In this study, weak acid stimulation significantly induced PGE 2 production in human oesophageal squamous epithelial cell carcinoma (KYSE-270) but not oesophagus adenocarcinoma cells (FLO-1 and KYAE-1). Although weak acid reflux may play a role in exacerbating oesophageal cancer through PGE 2 production in the oesophageal mucosa, further in vivo investigations are required to verify the involvement of weak acids in oesophageal cancer.
In this study, we showed that PGE 2 production in response to pH 4.5 is mediated by cPLA2 activation, since its inhibitor suppressed PGE 2 production induced by a weak acid in KYSE-270 cells. ERK activation and calcium influx are considered important for cPLA2 activation 24,31 ; thus, we also confirmed that intracellular calcium levels were increased by weak acidification of the medium, and weak acid-induced cPLA2 phosphorylation was  Recently, not only acid reflux but also bile acid reflux has been attracting attention in GERD pathogenesis 32 . Interestingly, bile acid-induced PGE 2 production was not affected by cPLA2 and ERK inhibitors despite the ERK activation, and bile acid stimulation had no effect on calcium influx. Bile acids strongly induce COX-2 expression and PGE 2 production in oesophageal epithelial cells 18 , which is consistent with our results in KYSE-270 cells. In Het-1A cells, PGE 2 production was not elevated by CDCA stimulation, although COX-2 expression was significantly increased. This might be attributed to the significantly lower basal and induced expression levels of COX-2 compared to those in KYSE-270 cells. Interestingly, pH 4.5 stimulation induced PGE 2 production without the induction of COX-2 expression in both Het-1A cells and KYSE-270 cells, indicating that the mechanisms underlying PGE 2 production differ between weak acid and bile acid stimulation.
Previous studies have shown that TRPV4 is involved in the regulation of intracellular calcium in oesophagus epithelial cells, although its physiological role remains unclear 33,34 . TRPV4 is reportedly activated by changes in not only temperature and osmotic pressure but also extracellular pH 35 . Particularly, TRPV4 begins to be activated below pH 6 and most potent activation were observed at around pH 4 36 . In the present study, we showed that TRPV4 is more highly expressed in oesophageal epithelial squamous cell carcinoma cells (KYSE-270) than in www.nature.com/scientificreports/ normal oesophageal cells. Two kinds of TRPV4 inhibitors significantly suppressed PGE 2 production induced by pH 4.5. We also confirmed that TRPV4 agonist increased ERK phosphorylation, which is essential for cPLA2 activation, suggesting a novel physiological role of TRPV4 in oesophageal epithelial cells. TRPV1, which may be a crucial factor involved in oesophageal hyperesthesia, plays a role in pH sensing of weak acid 37,38 . However, TRPV1 inhibitor did not suppress weak acid-induced PGE 2 production in our study. The suppressive effect of the TRPV4 inhibitor on PGE 2 production was limited and further investigations of other acid-sensing mechanisms are warranted. HST, a Japanese traditional medicine (Kampo medicine), contains the extracts of seven medicinal herbs and has been approved by Japanese Ministry of Health, Labour and Welfare for clinical use 39 . HST is used for the treatment of inflammatory diarrhoea, gastritis, and heartburn and is effective for the treatment of stomatitis and diarrhoea via the reduction of PGE 2 production [40][41][42][43][44] . Combined treatment with PPI and HST is effective for alleviating heartburn symptoms in patients with PPI-refractory GERD to the same extent as a double dose of PPI 19 . Furthermore, HST significantly inhibited carcinogenesis in a surgical rat reflux model 45 . In this study, we showed that HST suppressed both weak acid-and bile acid-induced PGE 2 production in oesophagus epithelial cells. These findings suggest that oesophageal PGE 2 suppression could relieve the clinical symptoms of www.nature.com/scientificreports/ PPI-refractory GERD in patients exhibiting weak acid reflux. HST suppresses PGE 2 production by inhibiting COX-2 activity, not COX-1 activity, as confirmed using recombinant proteins 46 . Moreover, components of HST, especially ginger-derived, are reported to suppress COX-2 activity in several studies 42,47,48 . Additionally, we showed that HST prevented AA-induced PGE 2 production only in KYSE-270 cells, but not in Het-1A cells with low COX-2 expression. The suppressive effect of the COX-2 inhibitor NS-398 on PGE 2 production was stronger in KYSE-270 cells than in Het-1A cells, suggesting that HST supressed COX-2-dependent PGE 2 production. Excessive PGE 2 production is involved in pain and tumour progression, while a small amount of PGE 2 derived from COX-1 is important in tissue repair and gastrointestinal mucosal protection 30,49 . Thus, the selective effect of HST on COX-2 may contribute to the alleviation of GERD without disrupting mucosal protection. In the future, we hope to further examine the safety of HST and to clarify the efficacy of HST in patients with GERD in greater detail.
In summary, we demonstrated that weak acid reflux induced PGE 2 production by oesophagus epithelial cells via a unique mechanism, which may be involved in the pathogenesis of refractory GERD. Therefore, pH correction up to 5 in patients with GERD may prevent the heartburn symptoms. Six-week-old male Sprague-Dawley rats weighing 180-200 g were purchased from Charles River Laboratories Japan, Inc. (Kanagawa, Japan). All animals were housed in a room with controlled ambient temperature (23 ± 3 °C), humidity (50 ± 20%), and lighting (12 h light-dark cycle) conditions. The animals were provided with ad libitum water and a standard laboratory animal diet (MF; Oriental Yeast Co., Ltd., Tokyo, Japan).

Tissue repair Mucosal protection
Bile acid ERK Figure 8. A graphical hypothesis of this study. Weak acid exposure at pH 4-5 induces PGE 2 production via TRPV4/ERK/cPLA2. Under normal conditions, oesophageal epithelial cells produce only a small amount of PGE 2 in response to weak acid stimulation, which may contribute to biological protection. Meanwhile, in cases of high expression of COX-2 due to stimulations, such as CDCA, weak acid exposure at pH 4-5 produces excess PGE 2 , which may trigger various symptoms including heartburn and oesophageal cancer development. HST is effective in treating various GERD symptoms due to excessive PGE 2 production by suppressing COX-2dependent production. PGE 2 prostaglandin E 2 ; TRPV4 transient receptor potential vanilloid 4; ERK extracellular signal-regulated kinase; cPLA2 cytosolic phospholipase A2; COX-2 cyclooxygenase-2; CDCA chenodeoxycholic acid; HST hangeshashinto.

Scientific Reports
| (2020) 10:20775 | https://doi.org/10.1038/s41598-020-77495-z www.nature.com/scientificreports/ cannulated with airway management, after which warmed weak acid solution (pH 4.5) or PBS was perfused at approximately 500 μL/min for 1 h using a perfusion pump. The perfusate was drained outside from the cannula inserted into the oesophagus and emerged from the stomach and collected for the last 5 min. The samples were concentrated using a Sep-Pak C18 cartridge (Waters, Milford, MA, USA) to assess PGE 2 concentration.
To measure the capacity of PGE 2 production by the rat oesophageal mucosa, rats were euthanized after perfusion and the oesophageal mucosa was collected. The mucosa was divided into proximal, intermediate, and distal regions and incubated in F12/RPMI-1640 medium (1:1) for 2 h at 37 °C. PGE 2 concentration in the medium was then measured; PGE 2 production capacity was calculated by dividing the total amount of PGE 2 in the medium by the amount of total protein in each mucosal tissue. The data were shown as the average values of the three regions (proximal, intermediate, and distal) per hour.
Measurement of PGE 2 synthetic capacity using intact cells. Enzymatic activity related to PGE 2 synthesis was determined by measuring the accumulation of PGE 2 induced by arachidonic acid (AA; Wako Chemical) in the culture fluids. Briefly, Het-1A cells (5.0 × 10 4 /well) and KYSE-270 cells (2.5 × 10 4 /well) were cultured overnight in 96-well plates. The culture fluids were replaced with the same fresh medium containing HST or the COX-2 inhibitor NS-398 for 15 min, and AA was added to the culture medium at a final concentration of 3 μmol/L. PGE 2 concentrations were measured as described above after further incubation for 15 min.
Cytotoxicity assay. Cytotoxicity was assayed using the LDH-Cytotoxic Test Kit (Wako Chemical). The accurate estimation of LDH activity in acidic conditions is challenging; accordingly, cytotoxicity was evaluated by examining the amount of LDH remaining in living cells. After stimulation, the cells were solubilised by cell lysis buffer (Cell Signaling Technology, Danvers, MA, USA) containing 1% Triton X-100, and the supernatant was used for the LDH assay after centrifugation. Cytotoxicity was calculated using the following formula for relative LDH activity: Cell death (%) = 100 × [(b − a)/b], where a = absorbance at 560 nm for the test sample, b = control.
Cell viability after HST treatment was evaluated by the amount of LDH released in the medium. As a positive control, cells were treated with 0.2% Tween 20 for 15 min and the medium was collected. Cell viability was calculated using the following formula for relative LDH activity: viability (%) = 100 × [(c − a)/(c − b)], where a = test sample, b = blank well, and c = positive control. Absorbance was measured using a microplate reader SpectraMax Plus 384 (Molecular Devices, San Jose, CA, USA).
Cell metabolic activity assay. Cell metabolic activity was evaluated using a Cell Counting Kit-8 (CCK-8; Dojindo Laboratories, Kumamoto, Japan). After stimulation, 8 µL of CCK-8 reagent was added to 100 µL of culture medium, and the plates were incubated at 37 °C in an atmosphere of 5% CO 2 for 1 h. Cell metabolic activity is presented as the change in absorbance at 450 nm, as determined using a microplate reader (SpectraMax Plus 384).
Statistical analyses. Data are reported as means ± standard deviation. Student's or Aspin-Welch's t-test was performed for two-group comparisons, and the Tukey-Kramer or Dunnett test for multiple-group comparisons. Statistical differences were analysed using StatLight (Yukms Co. Ltd., Kawasaki, Japan). P < 0.05 was considered statistically significant.

Data availability
All data generated or analysed during this study are included in this published article.