Protease‐Activated Receptor 2 Activation Provokes an Increase in Intracellular Calcium and Serotonin Secretion in a Human Enteroendocrine Cell Line

ABSTRACT

The P‐STS human ileal enteroendocrine tumor cell line responds with an increase in intracellular calcium and serotonin secretion to acetylcholine and histamine. Here we show that the cells react similarly to the protease‐activated receptor 2 (PAR2) agonists trypsin and SLIGRL‐NH2 peptide. The calcium increase induced by both agonists is inhibited by the PAR2 antagonist I‐191. PAR2‐IN‐1, another PAR2 antagonist, did not inhibit the response to the agonist peptide. Trypsin can also be looked upon as a surrogate for mast cell tryptase which cleaves PAR2 at the same site as trypsin. As mast cells may secrete tryptase simultaneously with histamine in close proximity to enteroendocrine cells, we tested whether trypsin and histamine might induce mutual desensitization. Histamine did not desensitize the response to trypsin and trypsin did not desensitize the response to histamine or acetylcholine. Further known effects of short‐time incubation with trypsin, namely phosphorylation of p38 mitogen‐activated protein kinase and activation of the nuclear factor κB pathway, were not detected in P‐STS cells. In conclusion, our findings indicate that serotonin secretion by enterochromaffin cells in response to PAR2 activation might contribute to gastrointestinal symptoms after mast cell activation by food allergens or irritable bowel syndrome. Our data suggest that histamine and mast cell tryptase may have at least additive effects on serotonin secretion.

Summary

• The protease‐activated receptor 2 (PAR2) agonists trypsin and SLIGRL‐NH2 peptide induced an increase in intracellular calcium in the human P‐STS ileal enteroendocrine cell line.

• This increase was inhibited by the PAR2 antagonist I‐191, but not by PAR2‐IN‐1, another antagonist.

• As expected from the calcium response, trypsin evoked serotonin secretion.

• No p38 mitogen‐activated protein kinase or nuclear factor κB signaling was detected.

• Trypsin and histamine did not evoke mutual desensitization of the calcium response.

• This suggests that histamine and the PAR2 agonist tryptase secreted by activated mast cells could additively contribute to serotonin secretion from enteroendocrine cells in their vicinity and cause diarrhea in food allergy and irritable bowel syndrome.

1. Introduction

Protease‐activated receptors (PARs) are G‐protein coupled receptors (GPCRs) activated by protease cleavage of their extracellular N‐terminus [1, 2]. When cleaved at the canonical site, PARs activate themselves with the newly formed N‐terminus. The gastrointestinal tract is exposed to a high number of luminal and mucosal proteases. These include digestive enzymes, proteases released by commensal microbiota and immune cells and extravascular clotting factors (reviewed in Pontarollo et al. [3]). PAR2, one of 4 known PARs, is involved in glandular secretion in the mouth, stomach and pancreas [4] and is expressed on the apical and basolateral side of enterocytes of the intestinal tract [5]. It can be activated by trypsin produced in the pancreas or by intestinal epithelial cells [3]. Mast cell tryptase, neutrophil elastase, some proteases produced by gut bacteria and the clotting factors FXa, FVIIa and tissue factor can also cleave at the canonical site and activate PAR2.

Activation of PAR2 on enterocytes has been implicated in ion secretion, in increasing intestinal permeability and in mediating epithelial inflammation [3, 4, 6], suggesting that this receptor and its agonists may play a role in intestinal diseases. Increased expression of trypsin, tryptase and PAR2 and increased levels of histamine have been found in colonic mucosa biopsies of irritable bowel syndrome (IBS) patients [7, 8, 9]. In line with this, supernatants from intestinal mucosa of IBS patients induced visceral hypersensitivity, a feature of IBS, in mice [10] and activated rat dorsal root ganglion neurons and neurons in human colonic biopsies, symptoms inhibited by protease inhibitors and anti‐histamines [8, 9]. Trypsin‐3, produced by intestinal epithelial cells and apparently mainly secreted to the basolateral side of the epithelium, is overexpressed in IBS patients and may be the main endogenous PAR2 agonist in the context of visceral hypersensitivity [11].

In addition to activation of PAR2 expressed on enterocytes or neurons, PAR2 expressed on enteroendocrine cells of the intestinal epithelium might contribute to symptoms of IBS patients. Excessive serotonin secretion, a tissue hormone and neurotransmitter thought to increase gut motility and ion secretion [12], may be involved in diarrhea‐predominant IBS. Serotonin is primarily produced and secreted by enterochromaffin cells, a subset of enteroendocrine cells. The P‐STS cell line, isolated from a neuroendocrine tumor of the terminal ileum [13], appears to be a suitable model for human enterochromaffin cells. which have been characterized by their high expression of mRNA for tryptophan hydroxylase 1 and chromogranin A [14]. In early phases of cell culture P‐STS cells also expressed high levels of chromogranin A mRNA and protein as well as tryptophan hydroxylase 1 mRNA [15]. P‐STS cells respond to ACh and histamine with an increase in intracellular calcium [Ca²⁺] _i which results in secretion of serotonin [15]. In contrast to mouse enterochromaffin cells which reacted with high increases of [Ca²⁺] _i and serotonin release to 1 µM adrenaline [16], in P‐STS cells even 100 µM adrenaline induced only a weak [Ca²⁺] _i response and no serotonin secretion exceeding constitutive secretion [17].

In this study we aimed to establish whether P‐STS cells respond to the PAR2 agonist trypsin with an increase in [Ca²⁺] _I . and serotonin secretion. It has been shown that PAR2 signaling after cleavage at the canonical cleavage site, for example, by trypsin or mast cell tryptase, involves an increase in [Ca²⁺] _I in mouse dorsal root ganglion neurons and a release of calcitonin gene‐related peptide and substance P [18]. GPCRs may be influenced by signaling cascades following activation of GPCRs of other receptor families [19]. As increased concentrations of proteases and histamine often occur together, the reaction of P‐STS cells to exposure to the PAR2 agonist trypsin in the presence of histamine was also investigated. Histamine, which may be derived from histamine‐rich food, intestinal microbiota or mast cell secretion in food allergy or IBS, has been shown to induce an increase of [Ca²⁺] _i in P‐STS cells via the histamine 1 receptor (H1R) [20]. Furthermore, we investigated whether treatment with trypsin would induce stress signaling [21] in P‐STS cells. Phosphorylation of the stress‐related kinase p38 mitogen‐activated protein kinase (MAPK) or activation of the nuclear factor κB (NF‐κB) pathway after activation of PAR2 has been shown in a human keratinocyte cell line [22, 23], oral squamous carcinoma cells [24] and in some studies with intestinal epithelial cell lines [25, 26]. It is shown in this study that P‐STS cells react with a PAR2‐mediated increase of [Ca²⁺] _i to trypsin which is mediated by T‐type voltage‐gated calcium channels (VGCCs) and followed by serotonin secretion. After challenge with trypsin no desensitization to acetylcholine or histamine, the two other known agonists evoking serotonin secretion in P‐STS cells, was seen and histamine did not desensitize the cells with respect to trypsin treatment. These results suggest that PAR2‐activating proteases and histamine have at least additive effects on serotonin secretion from enterochromaffin cells. Stress signaling in response to trypsin or trypsin plus histamine was not detected.

2. Methods

2.1. Cell Culture

The P‐STS cell line, originally isolated from a WHO III neuroendocrine tumor of the terminal ileum [13], was grown in a 1:1 mixture of Ham's F12 nutrient mixture and medium M199 supplemented with 10% heat‐inactivated fetal calf serum (FCS), 100 U/mL penicillin G and 100 µg/mL streptomycin at 37°C in a humidified atmosphere containing 5% CO₂. HeLa H1 human cervix carcinoma cells were cultured in the same way in DMEM. Both cell lines were free of Mycoplasma contamination, as verified by staining with Hoechst dye 33342.

2.2. Reagents

ACh chloride, G‐418 sulfate solution, histamine dihydrochloride, hydrogen peroxide (30% w/w with stabilizers, freshly diluted before use) and nifedipine were from Sigma Aldrich (St. Louis, Missouri, USA); mibefradil dihydrochloride was from Tocris Bioscience (Bristol, United Kingdom); Gibco trypsin‐EDTA (0.05%) was from Thermo Fisher (Schwerte, Germany); the PAR2 agonist peptide SLIGRL‐NH2 was from HelloBio (Dunshaughlin, Ireland); I‐191 and PAR2‐IN‐1 were from MedChemExpress (Monmouth Junction, New Jersey, USA); Fluo‐4AM was from AAT Bioquest (Pleasanton, California, USA); rabbit anti‐phospho‐p38 antibody (Thr 180/Tyr 182) was from Santa Cruz Biotechnology (Dallas, Texas, USA); mouse anti‐human RelA/NF‐κB p65 (clone 532301) was from R&D Systems (Minneapolis, Canada), goat Alexa Fluor 488‐labeled secondary fluorescent antibodies were from Life Technologies (Carlsbad, California, USA) and the serotonin Elisa kit was from ImmuSmol (Bordeaux, France).

2.3. [Ca²⁺]i Imaging

For [Ca²⁺]i imaging P‐STS cells were grown to a density of about 1000 cells per well. The cells were stained with Fluo‐4AM for 45 min at room temperature in serum‐free medium. The medium was exchanged for 300 µL of fresh serum‐free medium and the plate was placed on an inverted microscope Axio Observer Z1 equipped with a high‐resolution AxioCam MRc 5 camera (Carl Zeiss, Jena/Göttingen, Germany). After 25 min of incubation at room temperature photographs were taken every 10 s (200‐fold magnification, program Axiovision) before and after careful addition of 50 or 100 µL of medium control or agonists in pure medium. The median fluorescence intensity of the cells in one field of view per well was determined with Image J.

2.4. Measurement of Serotonin Secretion

Chromogranin A‐transfected P‐STS cells [27] were grown in 24‐well plates to a density of about 50% under the same conditions as untransformed P‐STS cells. For preservation of the chromogranin A expression plasmid 500 µg/mL G‐418 sulfate was added to the culture medium. The cells were washed with 300 µL pure medium and incubated for 10 min in 300 µL of the same medium. The medium was removed and the cells were incubated for 3 min in 100 µL fresh medium. Then medium control or trypsin was added in a volume of 10 µL to a final trypsin concentration of 600 nM. It was deemed possible from the [Ca²⁺]i imaging experiments that there would be no increase in serotonin secretion in many wells despite addition of a higher trypsin concentration of 600 nM. Therefore, to increase the likelihood to detect an agonist‐induced serotonin secretion in the sample by ELISA, two wells were treated in parallel and 60 µL of their supernatants were collected and combined after incubation for 5 min at room temperature. The supernatants were kept at 4°C for 1 h to allow sedimentation of cells that might have been detached and collected. Forty‐five microliters of the supernatants without cells were then added to 5 µL 10% stabilizer of the ImmuSmol Serotonin Research ELISA kit (ImmuSmol SAS, Bordeaux, France) and 20 µL of this solution was analyzed as described by the manufacturer, adding 20 µL medium with stabilizer also to the standards. In samples to which trypsin had been added as PAR2 agonist, residual cells in the supernatant might have been lysed by trypsin during the incubation at 4°C, thereby increasing the serotonin concentration of supernatants from these samples. To evaluate whether this could possibly have influenced the results, we measured the serotonin concentration in supernatants from control cells to which 600 nM trypsin had been added before putting them to 4°C for 1 h. The same supernatants with or without added trypsin contained 0.28 ± 0.01 and 0.46 ± 0,016 ng/ml serotonin, respectively, excluding cell lysis by trypsin at 4°C.

2.5. Imaging of RelA/NF‐κB p65 and Phospho‐p38

P‐STS cells were grown on cover slips in 24‐well plates and were treated with agonists for 15 or 30 min at 37°C in medium with or without FCS as described in the legend of Figure 6. For analysis of phospho‐p38 the cells were then immediately fixed with 4% paraformaldehyde in PBS. For imaging of RelA/NF‐κB p65 the cover slips were incubated for additional 3 h in complete medium at 37°C before fixation. The cells were then permeabilized for 5 min with 0.2% v/v Triton X‐100 in PBS and blocked with 10% FCS in PBS. After incubation of with antibodies against NF‐κB p65 or phospho‐p38 for 1 h at room temperature, followed by Alexa Fluor 488‐labelled secondary antibodies, nuclei were stained with Hoechst dye 33342 in PBS. The cover slips were mounted in a 9:1 (v/v) mixture of glycerol and 1 M Tris‐HCl (pH 8.6). containing 25% diazabicyclo(2,2,2)octane (Merck) and were viewed under a Zeiss Axioplan 2 fluorescence microscope using Axiovision software at 400‐fold magnification.

Figure 6.

P‐STS cells react with p38 MAPK phosphorylation and NF‐κB migration into the nucleus to H₂O₂ but not to 600 nM trypsin. (a) NF‐κB p65 migrated into the nucleus in some cells (white arrows) after addition of H₂O₂ as a positive control (final concentration 1 mM, incubation at 37°C for 15 min in medium without FCS or for 30 min in complete medium, followed by 3 h incubation in complete medium after removal of H₂O₂). After treatment with 600 nM trypsin for 15 min at 37°C in medium without FCS in the presence or absence of 10 µM histamine, followed by 3 h in complete medium, no accumulation of p65 in the nucleus was detected. The images shown are representative for 4 experiments conducted with trypsin, 4 with trypsin plus histamine, 3 with H₂O₂ in medium without FCS and 2 with H₂O₂ in complete medium. Final concentrations are indicated. b) Phosphorylated p38 MAPK is detected in the cytoplasm in P‐STS cells after treatment with 10 mM H₂O₂ for 15 min at 37°C, but no phosphorylation was detected with 1 mM H₂O₂ or 600 nM trypsin in the presence of absence of 10 µM histamine. High levels of phosphorylated p38 MAPK are detected in the cytoplasm of apoptotic cells (white arrows). The images shown are representative for 4 experiments conducted with trypsin, 4 with trypsin plus histamine, 3 with 1 mM and 3 with 10 mM H₂O₂. All treatments were conducted in medium without FCS. The specificity of the antibody was confirmed with HeLa cells, where p38 MAPK is detected in the nucleus after treatment with 10 mM H₂O₂ for 15 min.

2.6. Statistical Analysis of [Ca²⁺] _i Imaging and Serotonin Secretion Experiments

Increases in fluorescence calculated from the [Ca²⁺] _i imaging experiments were not normally distributed for P‐STS cells but rather bimodal. Therefore, the statistical significance of the difference of two mean values could not be calculated directly with a parametric test, but was calculated by the unpaired two‐tailed t‐test assuming unequal variances after ranking the data of the two groups tested, as described [27]. The same method was applied to the serotonin secretion experiment. Ranking the data reduced the power of the t‐test to detect significance, because it caused an underestimation of the highest fluorescence values obtained. To increase statistical reliability and strength, in every experiment the different treatments were conducted in parallel, resulting in equal sample numbers in each treatment group. Up to 3 such experiments were combined, that is, conducted together on the same cell culture plate. As shown in previous work [27], it was appropriate to consider each well as independent data point for the statistics. Significant differences are indicated by stars (*p ≤ 0.05, **p ≤ 0.01). Means +/− standard deviation are shown.

3. Results

3.1. Trypsin Evokes an Increase in [Ca²⁺] _i and Serotonin Secretion in P‐STS Cells

Challenge of P‐STS cells with 150 nM trypsin caused an increase in [Ca²⁺] _i within 10 to 30 s in about half of the samples (Figure 1a,b). Trypsin concentrations of 30 nM or lower were also tested (6 samples per trypsin concentration), but did not induce any visible increase in [Ca²⁺] _i within the 50 s observation period (Figure 1b). The [Ca²⁺] _i response of P‐STS cells to 150 nM trypsin was abolished by heat‐inactivation of the trypsin solution, indicating that enzymatic activity of trypsin ‐ presumably PAR2 cleavage—was required for this response (Figure 1c). An increase in [Ca²⁺] _i is known to provoke secretion in neuroendocrine cells [29] and this has also been confirmed for P‐STS cells [27]. Concordantly, the trypsin‐induced increase in [Ca²⁺] _i provoked serotonin secretion from P‐STS cells. The difference in serotonin concentration in the medium between trypsin‐treated cells and control cells with basal serotonin secretion almost reached statistical significance in 14 experiments (p = 0.062, Figure 1d).

Figure 1.

P‐STS cells respond with an increase in [Ca²⁺] _i and serotonin secretion to trypsin. Trypsin treatment was performed at room temperature where its activity is about 35% lower than at 37°C [28]. (a) Imaging of [Ca²⁺] _I with Fluo4‐AM after addition of 150 nM trypsin showing lag times of about 10 and 30 s (upper and middle lane, respectively) before the increase in [Ca²⁺] _I or no reaction within 50 s (lower lane). (b) Time courses of [Ca²⁺] _I after addition of trypsin (final concentrations as indicated, n = 2 for each concentration). One typical experiment out of three is shown. (c) Incubation of trypsin at 99°C for 15 min inhibits the increase in [Ca²⁺] _I caused by 150 nM trypsin. The increases in fluorescence 30 s after substance addition are shown. With heated trypsin the increase in fluorescence occurs immediately after the addition of the trypsin solution without any further increase (not shown), indicating that it is due to an increase in absorbance of the medium and not to residual trypsin activity. (d) Serotonin content in the supernatant of chromogranin A‐transfected P‐STS cells after treatment with 600 nM trypsin for 5 min at room temperature.

3.2. The [Ca²⁺] _i Response to Trypsin Is Mediated by PAR2 and T‐Type VGCCs

Trypsin is a known agonist for PAR2. As mentioned before, receptor activation occurs by the new N‐terminus of PAR2 formed by cleavage of the receptor by trypsin. In accordance with this mechanism, the agonist peptide SLIGRL‐NH₂ also evoked an increase in [Ca²⁺] _i in P‐STS cells (Figure 2a). This peptide mimics the cleaved N‐terminus of mouse PAR2 and is an agonist for human PAR2 with an EC₅₀ of 3.1 µM for a human colonocyte cell line [30]. The response to SLIGRL‐NH₂ was blocked by the PAR2 antagonist I‐191 [31], but not by PAR2‐IN‐1, which is commercially offered as alternative PAR2 antagonist (Figure 2b). I‐191 also inhibited the [Ca²⁺] _i response to trypsin (Figure 2c,d).

Figure 2.

P‐STS cells respond with an increase in [Ca²⁺] _i to the PAR2 agonist SLIGRL‐NH_2. (a) [Ca²⁺] _i responses to SLIGRL‐NH₂ (time courses of individual experiments are shown, n = 8). (b) The PAR2 antagonist I‐191 inhibits the [Ca²⁺] _i response to SLIGRL‐NH₂ (10 µM) while PAR2‐IN‐1 has no inhibitory effect (n = 4 for each inhibitor concentration, medium as well as SLIGRL‐NH2). (c) I‐191 also inhibits the [Ca²⁺] _i response to trypsin (110 nM, n = 12). (d) Inhibition of the [Ca²⁺] _i response to SLIGRL‐NH₂ and trypsin by pre‐incubation (5 min) with I‐191 (same experiments as shown in b and c). Relative increase in fluorescence 40 s (trypsin) or 50 s (SLIGRL‐NH₂) after addition of agonist. *p < 0.05 (two‐tailed binominal test).

Similar to the increase in [Ca²⁺] _i after challenge with ACh and histamine [20], the [Ca²⁺] _i response to trypsin was primarily mediated by T‐type VGCCs. This was indicated by strong inhibition by the T‐type channel inhibitor mibefradil, while no significant reduction of the [Ca²⁺] _i response to trypsin by the L‐type channel inhibitor nifedipine was seen (Figure 3).

Figure 3.

Ca²⁺ influx into the cytoplasm in response to trypsin occurs mainly through T‐type VGCCs in P‐STS cells. (a) Time courses of [Ca²⁺] _I after addition of trypsin (final concentration 150 nM, n = 10) with or without pre‐incubation (5 min, n = 10) with the T‐type VGCC inhibitor mibefradil (5 µM, n = 10) or the L‐type VGCC inhibitor nifedipine (5 µM). (b) Relative increase in fluorescence 40 s after addition of trypsin (same experiments as shown in a).

3.3. PAR2 Activation Does Not Desensitize P‐STS Cells Against Activation With Histamine or ACh

The [Ca²⁺] _i response to trypsin was not always immediate, but in some samples also occurred at a later time points, as seen from Figure 4a (no time points were taken between 50 and 100 s to avoid fluorescence bleaching). In some samples no response to trypsin was detected in the monitored field of view during the observation period of 160 s. The reason for this cannot be the state of the cell culture or differences in ambient temperature, as immediate or no response to the same trypsin concentration often occurred in parallel samples of the same experiment. A similar reaction pattern was also seen after challenge with histamine (see below). Apparently it is not specific for trypsin and probably due to the complex calcium response activation and inactivation processes occurring upon stimulation of cells with concentrations of agonists causing only submaximal stimulation [32, 33].

Figure 4.

PAR2 activation does not desensitize P‐STS cells against activation with histamine or ACh. (a–c) Time courses of [Ca²⁺] _I in different experiments after addition of trypsin (final concentration 150 nM) to Fluo4‐AM‐labelled cells, followed by addition of a) medium (n = 11, in three experiments the reaction to trypsin was delayed (open circles), (b) histamine (final concentration 10 µM, n = 8) or c) ACh (final concentration 0.1 µM, n = 6) 110 s later without trypsin removal. (d) Time courses of [Ca²⁺] _I experiments with immediate reaction to trypsin, including the experiments shown in a–c and an additional experiment with ACh added to a final concentration of 0.25 µM. (e and f) Cells activated by trypsin (150 nM) can respond to histamine (e, final concentration 10 µM) or Ach (f, final concentration 0.25 µM) added 110 s after trypsin.

Desensitization of individual P‐STS cells to histamine or ACh by pretreatment with trypsin as compared to pretreatment with medium was not observed (Figure 4b–d). From the images in Figure 4e,f it can be seen that the same cells, after having reacted to trypsin, reacted again with a strong increase in [Ca²⁺] _i to histamine or ACh only about 100 s later.

3.4. Activation With Histamine Does Not Desensitize P‐STS Cells Against PAR2 Activation

Like the [Ca²⁺] _i response to trypsin, the [Ca²⁺] _i response to histamine was sometimes delayed or no response was seen within 160 s (Figure 5a). Cells that had reacted within the first 30 s to histamine did not react again during the observation period of 160 s. However, when trypsin was added 110 s after challenge with histamine, some of the cells which before had reacted to histamine reacted within 10 s to trypsin stimulation (Figure 5b).

Figure 5.

Activation with histamine does not desensitize P‐STS cells against PAR2 activation. (a) Time courses of [Ca²⁺] _I in different experiments after addition histamine (final concentration 10 µM, n = 11) to Fluo4‐AM‐labelled cells. In 3 experiments the reaction to histamine was delayed (open circles). (b) Cells activated by histamine (10 µM) can respond to trypsin (e, final concentration 150 nM) added 110 s after histamine, which was still present.

3.5. P‐STS Cells React With p38 MAPK Phosphorylation and NF‐κB Migration Into the Nucleus to H₂O₂ but Not to 600 nM Trypsin

Neither treatment with trypsin alone nor trypsin plus histamine resulted in detectable activation of the NF‐κB pathway in P‐STS cells. Migration of p65 into the nucleus was detected after treatment with 1 mM H₂O₂, used as a positive control, in complete medium or in medium without FCS, while no nuclear fluorescence was seen after incubation with trypsin in the presence or absence of histamine (Figure 6a). Incubation with 10 mM H₂O₂, but not with 1 mM H₂O₂, for 15 min induced phosphorylation of p38, which was seen in the cytoplasm of P‐STS cells (Figure 6b). In accordance with literature [34], phosphorylated p38 was also detected in dying cells with fragmented nuclei. As in HeLa cells p38 MAPK was detected in the nucleus after treatment with 10 mM H₂O₂ for 15 min [35], we repeated the same experiment with HeLa cells to test the specificity of the antibody. Indeed, in H₂O₂‐treated HeLa cells the anti‐p38 MAPK antibody primarily stained the nuclei (Figure 6b, bottom), confirming this discrepancy of p38 MAPK behavior between the two cell lines.

4. Discussion

The P‐STS cell line has been cultivated from a human neuroendocrine tumor of the terminal ileum. P‐STS cells express tryptophan hydroxylase 1 and secrete serotonin and can therefore be considered as an enterochromaffin cell model. Serotonin secretion is triggered by an increase in [Ca²⁺] _I which can be induced by ACh via muscarinic 3 receptor or histamine via H1R. Histamine also synergistically enhanced the [Ca²⁺] _i response to ACh when added simultaneously or 5 to 10 min before ACh [27]

Here we show that P‐STS cells react with an increase in [Ca²⁺] _i to challenge with the PAR2 agonist trypsin or the PAR2 agonist peptide SLIGRL‐NH₂ and also secrete serotonin when treated with trypsin. The [Ca²⁺] _i response to both agonists was inhibited by the PAR2 antagonist I‐191, confirming PAR2 activation as the cause for the [Ca²⁺] _i increase. However, another compound sold by several companies as a PAR2 antagonist, PAR2‐IN‐1, did not inhibit the response to SLIGRL‐NH₂. In retrospect, neither in literature nor in the patent (WO2015048245A1) cited by the companies with respect to this compound, any indication for its PAR2 antagonist properties was found and it should therefore not be used as a PAR2 inhibitor.

PAR2 signaling can sensitize transient receptor potential channels, thereby increasing the release of neurotransmitters and contributing to neurogenic inflammation and pain [6, 36]. However, contrary to the findings in mice, where transient receptor potential ankyrin 1 (TRPA1) is present in about half of all enterochromaffin cells [37] and may mediate calcium entry [16], this ion channel is not prominent in human enterochromaffin cells [14] and apparently is not present in P‐STS cells [20]. P‐STS cells also showed no reaction to the transient receptor potential vanilloid 1 (TRPV1) agonist capsaicin and the transient receptor potential vanilloid 4 (TRPV4) agonist GSK‐1016790A (lit 14), receptors sensitized by PAR2 activation in other cell types [6]. Here we show that in P‐STS cells, calcium entry into the cytoplasm in response to PAR2 activation proceeds mainly via T‐type VGCCs, as has also been shown for the response to ACh and histamine [20].

It has been shown that cleavage of PAR2 at the canonical site by trypsin or tryptase induces receptor internalization, thereby markedly desensitizing the cell to subsequent PAR2 activation for some minutes [38]. With P‐STS cells at room temperature, the [Ca²⁺] _i response to trypsin in the field of view monitored was maximal 20 to 40 s after addition of the protease in most cases, but sometimes was delayed. To investigate whether the response to ACh or histamine is influenced by preceding challenge with trypsin, ACh or histamine was added 110 s after trypsin. This time point for addition of a second agonist was chosen because in two rat epithelial cell lines desensitization by trypsin to a second treatment with a PAR2 agonist has been shown to be maximal 2 min after the first challenge with trypsin [38]. We did not observe any desensitization of the cells to ACh or histamine added 110 s after trypsin (which was still present in the medium). Several cells that had immediately reacted with an increase in [Ca²⁺] _i to trypsin, reacted again to ACh or histamine. Likewise, the same cells reacted to histamine and then to trypsin added 110 s after challenge with histamine. Tryptase has been shown to enhance responses to histamine plus serotonin in human enteric neurons of the submucosa [39]. There might also be a tendency toward a synergistic effect of trypsin and histamine in P‐STS cells, but this could not be shown unequivocally here. In any case, no mutual desensitization between trypsin and histamine was seen and the response to trypsin did not inhibit a subsequent response to the pro‐secretory neurotransmitter ACh, indicating that tryptase and histamine secreted by mast cells will have at least additive effects on serotonin secretion.

To further investigate the reaction of P‐STS cells to these mast cell products which putatively play a role in irritable bowel syndrome, P‐STS cells were treated with 600 nM trypsin at 37°C for 15 min in the absence or presence of histamine. This treatment with a rather high concentration of trypsin and histamine caused neither detectable MAPK p38 phosphorylation nor migration of the p65 subunit of NF‐κB into the nucleus. As positive control for NF‐κB activation and p38 phosphorylation in P‐STS cells the cells were treated with H₂O₂. In previous experiments, NF‐κB p65 nuclear accumulation was detected after challenge with 1 mM H₂O₂, but not with Escherichia coli lipopolysaccharide, despite expression of toll‐like receptor 4 and CD14 in P‐STS cells [20, 40]. Phosphorylation of p38—which interestingly was seen in the cytoplasm in P‐STS cells but in the nucleus in HeLa cells ‐ was detectable in P‐STS cells after treatment for 15 min with 10 mM H₂O₂, but not with 1 mM H₂O₂. Strong p38 phosphorylation was also detected in dying P‐STS cells with fragmented nuclei. These results show that P‐STS cells respond to oxidative stress by H₂O₂ with NF‐κB activation and p38 phosphorylation, but these two stress response pathways are not readily induced by PAR2 activation or the mast cell product histamine. This is in contrast to Caco‐2 colorectal carcinoma cells, where both p38 phosphorylation and NF‐κB activation were induced by the PAR2 activating peptide SLIGRLKV‐NH2 [25].

In conclusion, our findings suggest that gastrointestinal symptoms after acute mast cell activation by food allergens or in irritable bowel syndrome may not only be caused by mast cell degranulation, but also by serotonin secretion from enterochromaffin cells in response to the mast cell products histamine and tryptase. In addition, trypsin secreted by the pancreas or intestinal epithelial cells and proteases produced by gut bacteria might induce excessive serotonin secretion.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.