Neurogastroenterology & Motility Neurogastroenterol Motil (2014)

doi: 10.1111/nmo.12334

Natural compounds boldine and menthol are antagonists of human 5-HT3 receptors: implications for treating gastrointestinal disorders € J. WALSTAB ,* C. WOHLFARTH ,* R. HOVIUS ,† S. SCHMITTECKERT ,* R. R OTH ,* , € H. B ONISCH ¶ & B. NIESLER * **

,

** F. LASITSCHKA ,‡ M. WINK ,§

*Institute of Human Genetics, Department of Human Molecular Genetics, Heidelberg University, Heidelberg, Germany †Laboratory of Protein Engineering, Ecole Polytechnique Federale de Lausanne, Lausanne, Switzerland ‡Institute of Pathology, Heidelberg University, Heidelberg, Germany §Department Biology, Institute of Pharmacy and Molecular Biotechnology, Heidelberg, Germany ¶Institute of Pharmacology and Toxicology, University of Bonn, Bonn, Germany **nCounter Core Facility, Institute of Human Genetics, University of Heidelberg, Heidelberg, Germany

Key Messages

• Due to the limited availability of specific drugs for treating gastrointestinal (GI) disorders, many patients rely on • • •

alternative medicines. Impaired 5-HT3 receptor function appears to be implicated in disease pathology. The aim was to study the pharmacological impact of boldine and menthol, natural compounds used to combat symptoms of GI disorders, on human 5-HT3 receptors. We performed radioligand binding, Ca2+ influx measurement, and a membrane potential assay on HEK293 cells heterologously expressing human 5-HT3A- and 5-HT3AB receptors. Boldine competitively and (-)-menthol noncompetitively inhibited the activation of human 5-HT3 receptors in the low and middle micromolar range, respectively. Both compounds were more potent antagonists at 5-HT3A- vs 5-HT3AB receptors. Expression analyses on mRNA- (nCounter measurement) and protein level (immunofluorescence) revealed co-expression of 5-HT3A- and 5-HT3B subunits in relevant human gut layers. The gained results provide first evidence for physiologically relevant GI receptors as targets for the natural compounds boldine and menthol and its plant extracts in the treatment of GI disorders.

Abstract Background Impaired 5-HT3 receptor function is likely involved in the pathogenesis of functional gastrointestinal disorders (FGID) and 5-HT3 receptor antagonists are effective treatments for chemotherapyinduced nausea and vomiting (CINV) and irritable bowel syndrome (IBS). The monoterpene alcohol menthol and the aporphine alkaloid boldine combat

symptoms of gastrointestinal diseases; both interact with other members of the Cys-loop ligand-gated ion channel family and may therefore also act on 5-HT3 receptors. Methods The impact of boldine and menthol on human recombinant homomeric 5-HT3A- and heteromeric 5-HT3AB receptors in HEK293 cells was determined by radioligand binding, a luminescencebased Ca2+ assay, and a membrane potential assay. 5-HT3 protein and mRNA expression was assessed in human colon tissue. Key Results Boldine and menthol inhibited the 5-HT-induced activation of 5-HT3 receptors in the low and middle micromolar range, respectively. Boldine was a competitive antagonist of both receptors being 6.5- to 10-fold more potent at 5-HT3A- vs 5-HT3AB receptors. Menthol non-competitively and stereoselectively inhibited both receptors:

Address for Correspondence PD Dr. Beate Niesler, Institute of Human Genetics, Department of Human Molecular Genetics, Im Neuenheimer Feld 366, Heidelberg 69120, Germany. Tel: 0049-6221-56 35274; fax: 0049-6221-56 5155; e-mail: [email protected] Received: 30 October 2013 Accepted for publication: 25 February 2014

© 2014 John Wiley & Sons Ltd

1

J. Walstab et al.

Neurogastroenterology and Motility

revealed beneficial effects of peppermint oil to improve abdominal symptoms of IBS.11–13 Its pharmacological effect may be attributed to the main component, the monoterpene alcohol ( )-menthol. Menthol exhibits a spasmolytic action in the gastrointestinal (GI) tract, which had been attributed to the inhibition of voltagegated Ca2+ channels.14–16 Menthol can also interact with ligand-gated ion channels; it is a positive modulator of c-aminobutyric acid type A (GABAA)- and glycine receptors17 and a negative modulator of neuronal nicotinic acetylcholine (nACh) receptors.18 The aporphine alkaloid boldine is structurally different from menthol. It is the main alkaloid of the Chilean boldo tree but is also present in other members of the Monimiaceae (e.g. Boldea fragrans), and other plant families.19 Boldine has been traditionally used for treating GI disorders especially by South American natives and it also mediates a smooth muscle relaxation in the GI tract.20,21 Among its known pharmacological targets are metabotropic dopamine- and 5-HT2A receptors, and ligand-gated neuronal nACh receptors.22–25 Both menthol and boldine were shown to interact with ligand-gated ion channels and seem to be effective against symptoms of FGIDs. Therefore, the aim of this study was to analyze a putative interaction of these compounds with human 5-HT3 receptors and if this was the case, to test for a possible discrimination between homomeric 5-HT3A- and heteromeric 5HT3AB receptors. Complementing 5-HT3 protein and mRNA expression analyses of human colon samples should substantiate the potential relevance of the interaction in the human gut and for the treatment of neurogastrointestinal disorders.

In contrast to (+)-menthol, ( )-menthol was significantly more potent toward 5-HT3A- vs 5-HT3AB receptors. We show co-expression of 5-HT3A and 5-HT3B subunits in the human gut epithelium, the lamina propria, the myenteric plexus, and the muscular cell layer. Conclusions & Inferences The demonstrated 5-HT3 inhibitory effects may be relevant for boldine’s and menthol’s alleviating properties on FGID and may encourage clinical studies with the compounds or the plant extracts for CINV and IBS treatment. The found receptor-discriminative properties make boldine and ( )-menthol to potentially useful tools for analyzing structural differences between these receptor subtypes. Keywords 5-HT3 receptor, aequorin, irritable bowel syndrome, ligand-gated ion channel, natural compound. Abbreviations: [Ca2+]i, cytosolic Ca2+ concentration; CINV, chemotherapy-induced nausea and vomiting; FGID, functional gastrointestinal disorders; GABAA, c-aminobutyric acid type A; GI, gastrointestinal; IBS, irritable bowel syndrome; MP, membrane potential.

INTRODUCTION Serotonin type 3 (5-HT3) receptors are Cys-loop ligandgated ion channels composed of five subunits surrounding a cation-permeable (Na+, Ca2+, K+) pore.1 Five human 5-HT3 subunit-encoding genes have been cloned: HTR3A-E encoding the subunits 5-HT3A, B, C, D, and E.2–4 The best characterized receptor subtypes are homomeric 5-HT3A- and heteromeric 5-HT3AB receptors that can be discriminated, for example, by their channel properties and the potency of the physiological agonist 5-HT.2,5 Besides the well-known efficacy of 5-HT3 receptor antagonists for treating chemo-/radiotherapy-induced nausea and vomiting, several studies showed that they can also combat symptoms of neurogastrointestinal and psychiatric disorders.6 Disturbances in 5-HT3 receptor function seem to be involved in the pathoetiology of neurogastrointestinal and psychiatric disorders.6,7 Although 5-HT3 receptor antagonists, for example alosetron, are effective treatments of irritable bowel syndrome (IBS), they are not approved in Europe because of rare cases of serious ischemic colitis.8 Due to the limited availability of specific drugs for treating functional gastrointestinal disorders (FGID), many patients rely on alternative medicines.9 In a recent study, peppermint oil reduced intragastric pressure and proximal phase contractility in healthy volunteers.10 Randomized-controlled trials

MATERIALS AND METHODS Chemicals and drugs Coelenterazine h was from PJK (Kleinblittersdorf, Germany). Boldine, 5-hydroxytryptamine creatinine sulfate (5-HT, serotonin), ( )- and (+)-menthol were from Sigma-Aldrich (Munich, Germany). [3H]GR65630 ([3H]-3-(5-methyl-1H-imidazol-4-yl)-1-(1methyl-1H-indol-3-yl)-1-propanone, specific activity 83.8 Ci mmol/L) was from PerkinElmer (Boston, MA, USA).

Expression constructs The human 5-HT3A- and 5-HT3B subunit-encoding cDNAs from HTR3A (AJ003079) and HTR3B (AF080582) were cloned into pcDNA3 (Invitrogen). The 5-HT3A cDNA was subcloned into pcDNA5/FRT/TO (Invitrogen) yielding 5-HT3A_pcDNA5/FRT/ TO allowing inducible receptor expression in Flp-InTM T-RexTM293 cells (Invitrogen). 5-HT3A- and 5-HT3B cDNAs were also subcloned into pBudCE4.1 (Invitrogen) yielding pBudCE4.1

2

© 2014 John Wiley & Sons Ltd

Boldine and menthol inhibit human 5-HT3 receptors

5-HT3AB for expression of the 5-HT3A- and 5-HT3B subunit driven by human elongation factor 1a-subunit and human cytomegalovirus immediate-early promoter, respectively. Apoaequorin cDNA (L29571) was subcloned from cytAEQ/pcDNA1 into pcDNA3.1/zeo(+) (both Invitrogen).

speed Vmax of signal change using SoftMax Pro (Molecular Devices). Pharmacological curves and parameters were calculated with GraphPad Prism 5.0 (GraphPad Software Inc., San Diego, CA, USA) or IGOR Pro (WaveMetrics Inc., Portland, OR, USA). The pIC50 values were converted to Ki values using the equation of Cheng and Prusoff.29 Concentration–response data were fitted using one- and two-component sigmoidal equations. The best fitting model was determined by the extra sum-of-squares F-test. The pEC50 values for boldine obtained from Schild experiments were plotted against the antagonist concentration and fitted by non-linear regression according to the method of Lew and Angus.30 An F-test was performed to assess the conformity of the data to a model of simple competition. Data are presented as the means  SEM. Statistical analysis was performed with Student’s t-test and, when appropriate, one-way ANOVA followed by Tukey’s posttest. Differences were considered significant at p < 0.05.

Cell culture and transfection Human embryonic kidney (HEK) 293 cells (ATCC, Manassas, VA, USA) were grown in DMEM +10% fetal calf serum (FCS) +100 U/ mL penicillin +100 lg/mL streptomycin. Transfection was performed as previously described.26 Ca2+ influx and radioligand binding: Cells stably expressing 5-HT3A receptors and apoaequorin (HEK293 5-HT3A/apoaequorin) were recently established26 and cultured in the above stated medium containing geneticin (600 lg/mL) and zeocin (500 lg/mL) for selection. For measuring 5-HT3AB receptors, cells were transiently transfected with 5-HT3A- and 5-HT3B cDNAs (ratio 1 : 4) and, for aequorin assays, additionally with apoaequorin cDNA.

5-HT3 receptor protein expression analysis: immunofluorescence experiments

Membrane potential (MP) assay: For generation and cultivation of cell lines (i) with inducible expression of 5-HT3A (Flp-InTM TRexTM-293 5-HT3A) and (ii) stably expressing 5-HT3A- and B subunits (HEK293 pBud 5-HT3AB) see supplemental data.

Tissue from an unaffected colon sample (male cecum adenocarcinoma patient, 72-year-old) was obtained from the GEZEH tissue bank and experiments were approved by the local Ethics Committee of the Heidelberg University.

Radioligand binding experiments

Immunolabeling of 5-HT3A- and 5-HT3B subunits in tissue sections and fluorescence imaging were performed as described previously.31 We used the following antibodies: primary: rabbit anti-5-HT3A31 and rabbit anti-5-HT3B (ab39629; Abcam; Cambridge, UK), mouse anti-HuC/D (A21272, clone 16A11; Invitrogen), secondary: Alexa Fluorâ 488 goat anti-mouse IgG (A11029) and Alexa Fluorâ 568 goat anti-rabbit IgG (A11036; Invitrogen).

Preparation of crude membranes and [3H]GR65630 binding was performed as previously described.26,27

Aequorin luminescence assay The aequorin assay was performed as previously described.26,28

5-HT3 receptor mRNA expression analysis

Membrane potential assay

Laser microdissection and pressure catapulting Tissue from three unaffected colon samples (male colon adenocarcinoma patient, female cecum adenocarcinoma patient, male hyperplastic polyposis coli patient; age: 55–80 years) was obtained from the GEZEH tissue bank and experiments were approved by the local Ethics Committee of the Heidelberg University. Laser microdissection and pressure catapulting were performed as described previously.31

Typically, 25 000–35 000 cells in 100 lL DMEM/F12+ GlutaMAXTM-I+ 10% newborn calf serum were seeded per well of a 96-well plate (Isoplate 96 FTC; Perkin Elmer, Boston, MA, USA). 5-HT3A receptor expression was initiated after 16 h by adding tetracycline to 2 lg/mL. After 48 h, medium was replaced by 100 lL dye solution (1 vial Red component of MP assay kit [Molecular Devices, Sunnyvale, CA, USA] in 30 mL Dulbecco’s phosphate buffered saline containing CaCl2 and MgCl2 [D-PBS++; Sigma-Aldrich]). After incubation for 30 min at 37 °C, fluorescence intensity at 565 nm upon excitation at 530 nm (emission cut off 550 nm) was measured every 1.7 s during 200 (5-HT-activation) or 300 (competition) sec using a FLEXstation (Molecular Devices). 5-HT-evoked receptor activation was quantified by adding at 20 s 50 lL of serial dilutions of 5-HT in D-PBS++. For antagonist competition, 50 lL of serial dilutions of antagonist in D-PBS++ were added at 20 s, followed by addition of 50 lL 5-HT at 120 s.

nCounter measurement of microdissected samples Total RNA from microdissected colon tissue was isolated using the guanidium thiocyanate-phenol-chloroform extraction method32 according to the manufacturer’s instructions (peqGOLD TriFastTM kit; PeqLab Biotechnologie GmbH, Erlangen, Germany). For precipitation, 2 lL of Pellet Paintâ Co-Precipitant (Novagen; Merck Group, Darmstadt, Germany) per sample were used. Up to 100 ng of total RNA were analyzed by nCounter gene expression analysis (Nanostring, Seattle, WA, USA) with a customized codeset (Table SD1) as recommended by the manufacturer. Counts for each gene were derived by subtracting the average +2SD of background counts (negative controls) from the total counts. Expression of HTR3A (NM_000869.5) and HTR3B (NM_006028.3) was normalized to the geometric mean expression of the six reference genes ARF1 (NM_001024227.1), PGK1 (NM_000291.2), SDHA (NM_004168.2) SNX17 (NM_014748.2), TBP (NM_003194. 4), and UBB (NM_018955.2).33

Data analysis Relative light units for 5-HT-induced increases of the cytosolic Ca2+ concentration ([Ca2+]i) in the aequorin assay were obtained by subtraction of baseline luminescence from the 5-HT-induced peak maximum luminescence. In the MP assay, time responses of 5-HT-induced MP changes were evaluated from the maximal

© 2014 John Wiley & Sons Ltd

3

J. Walstab et al.

Neurogastroenterology and Motility

Table 1 Pharmacological parameters of serotonin and natural compounds at 5-HT3A- and 5-HT3AB receptors expressed in HEK293 cells (Fig. S1 and 2) 5-HT3A

Compound

5-HT3AB

pIC50  SEM (IC50 [lmol/L])

pKi  SEM (Ki [lmol/L])

nH  SEM

n

pIC50  SEM (IC50 [lmol/L])

pKi  SEM (Ki [lmol/L])

nH  SEM

n

6 9 6 6 7

   

2+

Ca influx assay 5-HT 5.76  0.07 (1.75)† Boldine

6 6

     

0.10 (2.07)†,‡ 0.17 (38.37) 0.05 (25.70) 0.05 (13.12)§ 0.06 (34.51)** 0.06 (24.89)

5.38  0.05 (4.14)*** 5.24  0.05 (5.70)§,*** n.a. n.a.

1.94  0.56 1.37  0.36 1.72  0.19 1.34  0.19§ 1.55  0.21 1.52  0.32

0.21 0.06 0.57 0.77

5 3 3 3

6.12 4.15 2.70 3.09

   

0.11 0.20 0.15 0.12

n.a. 4.45  0.20 (35.50)* n.a. n.a.

1.41 0.91 1.89 0.98

0.27* 0.33 0.79 0.39

5 3 3 3

2.78  0.31 1.00  0.04 –

4 3 3

6.14  0.06 (0.72) 5.56  0.06 (2.74) –

6.54  0.06 (0.29)*** 5.96  0.06 (1.09) –

1.19  0.03* 1.09  0.05 –

5 3 3

1.82  0.27

6

6.38  0.06 (0.42)

1.96  0.28

8

(17.62) (17.70)

n.a. n.a.

1.36  0.17 2.16  0.59

(0.22)† (11.61) (179.41) (369.64)

n.a. 5.26  0.08 (5.50) n.a. n.a.

2.13 0.64 1.45 1.89

   

(0.31) (3.45)

7.06  0.05 (0.09) 5.97  0.03 (1.07) –

5.94  0.06 (1.14)

( )-Menthol 4.75  0.06 (+)-Menthol 4.75  0.09 Membrane potential assay 5-HT 6.66  0.01 Boldine 4.96  0.08 ( )-Menthol 3.75  0.27 (+)-Menthol 3.43  0.12 [3H]GR65630 binding 5-HT 6.51  0.05 Boldine 5.46  0.03 ( )-Menthol –

5.69 4.42 4.59 4.88 4.46 4.60

n.a.

(0.77)*** (71.01) (2041.90)* (819.38)

n.a.

SEM, standard error of the mean; pIC50 (Ki), -logIC50 (Ki); nH, Hill coefficient; n.a., not applicable. †pEC50  SEM (EC50 [lmol/L]) values for the agonist 5-HT; ‡biphasic curve; shown are pEC50H/pEC50L, -log EC50 of the high (H) or low (L) affinity part of the curve; §boldine inhibition of 5-HT3AB in the presence of 50 lmol/L 5-HT; significant differences between 5-HT3A- and 5-HT3AB receptors: *p < 0.05, **p < 0.01, ***p < 0.001 (Student’s t-test).

binding on the receptor. The calculated pKi values of boldine for the two 5-HT concentrations were not significantly different and could therefore be pooled (5.33  0.04, n = 15; Table 1). The Ki of boldine was about ten-fold lower on the 5-HT3A- compared to that on the 5-HT3AB receptor. To reveal a putative stereoselective effect of menthol, as has been demonstrated for GABAA receptors,17 we used the enantiomers ( )- and (+)-menthol for the assay (Fig. 1B and C). The determined IC50 values of (+)-menthol on both receptor subtypes were virtually similar ranging around 20 lmol/L (Table 1). In contrast, ( )-menthol was significantly more potent in inhibiting 5-HT3A- compared to 5-HT3AB receptors (p < 0.001), demonstrating a stereoselective inhibition.

RESULTS Inhibition of 5-HT-induced Ca2+ influx through human 5-HT3 receptors by boldine and menthol enantiomers In the aequorin assay, 5-HT induced concentrationdependent increases in [Ca2+]i on coelenterazine h-loaded stable HEK293 5-HT3A/apoaequorin cells and HEK293 cells transiently expressing 5-HT3A/B subunits and apoaequorin (see supplemental data, Fig. S1; Table 1). The tested compounds concentration-dependently inhibited 5-HT-induced Ca2+ influx through 5-HT3A- and 5-HT3AB receptors (Fig. 1A). The compounds did not elicit a Ca2+ response on their own and they did not directly influence [Ca2+]i of the cells as revealed by measuring remaining aequorin luminescence upon cell lysis (data not shown). The pharmacological parameters derived from the inhibition curves are depicted in Table 1. The (p)IC50 values of boldine were converted to (p)Ki values according to the equation of Cheng and Prusoff. This was not applicable for menthol, as further experiments showed that it is a non-competitive inhibitor (Fig. 1C). The boldinemediated inhibition of Ca2+ influx through 5-HT3AB receptors was recorded in the presence of either 50 or 200 lmol/L 5-HT (Fig. 1A). The derived pIC50 values significantly differed from each other (p < 0.001), indicating a competition of boldine and 5-HT for the

Influence of boldine and ( )-menthol on the concentration–response relationship of 5-HT-induced Ca2+ influx To get insights into the mode of inhibitory action of boldine and ( )-menthol on human 5-HT3 receptors, 5-HT concentration–response experiments in the presence of varying concentrations of boldine or a defined concentration of ( )-menthol were performed. At homomeric 5-HT3A receptors, increasing concentrations of boldine led to a parallel right shift of the 5-HT concentration–response curve without changing the maximum response (Fig. 1B, Table 2). This indi-

4

© 2014 John Wiley & Sons Ltd

Boldine and menthol inhibit human 5-HT3 receptors

A

B

C

Figure 1 Inhibition of 5-HT-induced Ca2+ influx through 5-HT3A- and 5-HT3AB receptors by boldine and menthol enantiomers. 5-HT-induced aequorin luminescence in response to an increased cytosolic Ca2+ concentration ([Ca2+]i) was measured in coelenterazine h-loaded HEK293 cells heterologously expressing apoaequorin and human 5-HT3 receptors. The respective antagonist was present 10 min before and during 5-HT application. (A) Concentration-dependent inhibition of 5-HT3 receptors. 5-HT concentrations for receptor activation were: 3 lmol/L (5-HT3A), 50 or 200 lmol/L (5-HT3AB). Data are expressed as percentages of the response to 5-HT in the absence of the antagonist (means  SEM of 3–10 independent experiments). Half-maximal inhibition values and corresponding Ki values are shown in Table 1. Concentration-dependent 5-HTinduced Ca2+ influx in the absence or presence of increasing concentrations of boldine (B) or the estimated IC50 of ( )-menthol (C). Data are expressed as percentages of the response to 10 lmol/L (5-HT3A) or 300 lmol/L (5-HT3AB) 5-HT in the absence of the antagonist (means  SEM of 3–7 independent experiments). Parameters derived from these curves are summarized in Table 2.

different from unity (1.21  0.28). In the latter, the EC50 value, derived from a monophasic fit of the 5-HT concentration–response curve in the absence of boldine (=8.24 lmol/L), was used for non-linear Schild analysis. Application of the estimated IC50 of ( )-menthol revealed a non-competitive inhibition of 5-HT3A- and 5-HT3AB receptors as the antagonist did not change 5-HT potency but significantly reduced the 5-HTinduced maximum response (Fig. 1C; Table 2). The significantly higher EC50 value of 5-HT that was determined for the low affinity fraction of the concentration–response curve on 5-HT3AB receptors in the presence of ( )-menthol (Table 2) might be due to a certain impreciseness of the biphasic curve fit.

cates a competitive mode of action, which is supported by non-linear regression according to Lew and Angus revealing a Schild slope not significantly different from unity (0.76  0.16). At 5-HT3AB receptors, the biphasic 5-HT concentration–response curve became more monophasic and steeper with increasing concentrations of boldine, indicating the predominant inhibition of the lower high affinity part of the curve, represented by homomeric 5-HT3A receptors (Fig. 1B, Table 2). This result confirms the differential potency of boldine for the inhibition of 5-HT3A- and 5-HT3AB receptors. Increasing concentrations of boldine led to a parallel right shift of the 5-HT concentration–response curve without reducing the maximum response. Thus, boldine also competitively inhibits 5-HT3AB receptors, which is supported by a Schild slope not significantly

© 2014 John Wiley & Sons Ltd

5

J. Walstab et al.

Neurogastroenterology and Motility

Table 2 Pharmacological parameters of 5-HT concentration–response curves in the absence or presence of varying concentrations of boldine or a fixed concentration of ( )-menthol derived from Ca2+ influx measurement (Fig. 1B and C) 5-HT3A

5-HT3AB

Antagonist

lmol/L

pEC50  SEM (EC50 [lmol/L])

Control

0

5.76  0.07 (1.75)

Boldine

1 5 10 15 30

5.38 5.02 4.80 5.74

( )-Menthol

   

0.01 0.04 0.03 0.08

(4.16)** (9.48)*** (15.70)*** (1.83)

nH  SEM

Emax  SEM (% of 10 lmol/L response)

n

1.82  0.27

111.7  5.9

6

1.47 1.76 2.38 1.80

   

0.05 0.19 0.35 0.40

96.0 96.1 87.2 84.5

   

0.4 11.3 7.6 6.1*

3 3 3 5

pEC50  SEM (EC50 [lmol/L])

5.69 4.42 4.94 4.69 4.33

    

0.10 0.17 0.06 0.01 0.07

nH  SEM

(2.07) (38.37)† (11.56) (20.23)* (46.34)***

1.94 1.37 1.05 1.57 1.31

5.53  0.08 (2.99) 3.98  0.06 (104.47)†,*

    

0.56 0.36 0.03 0.13** 0.08

1.57  0.39 3.49  4.05

Emax  SEM (% of 300 lmol/L response)

n

103‡

6

94.2  1.5 95.0  3.5 97.5  8.9

6 3 4

79‡

6

For the statistical analysis of pEC50 and nH of 5-HT on 5-HT3AB receptors in the presence of boldine, values were compared to the 1 lmol/L concentration of boldine. pEC50, -log EC50; SEM, standard error of the mean; nH, Hill coefficient; Emax, serotonin-induced maximum response. † biphasic curve; shown are pEC50H/pEC50L, -log EC50 of the high (H) or low (L) affinity part of the curve, ‡top of biphasic curve was constrained to this value as this yielded the best fit. Significant difference compared to control: *p < 0.05, **p < 0.01, ***p < 0.001 (one-way ANOVA followed by Dunnett’s posttest).

A

B

C

Figure 2 Concentration-dependent inhibition of 5-HT-induced 5-HT3A- and 5-HT3AB receptor-mediated membrane potential changes by boldine (A) and menthol enantiomers (B and C). 5-HT-induced membrane potential changes in response to cation flux through 5-HT3 receptors was measured in HEK293 cells heterologously expressing human 5-HT3 receptors (fluorescence assay). For receptor activation, the EC50 concentrations of 5-HT, that is, 0.22 lmol/L (5-HT3A) or 0.77 lmol/L (5-HT3AB), were used. The respective antagonist was present 100 s before and during 5-HT application. Data are expressed as percentages of the response to 5-HT in the absence of the antagonist (means  SEM of three independent experiments). MP, membrane potential.

of ( )-menthol was significantly higher on the 5-HT3Avs the 5-HT3AB receptor (p < 0.05), the pIC50 values for (+)-menthol were virtually similar at both receptor subtypes. This demonstrates that ( )-menthol but not (+)-menthol clearly discriminates between the two receptor subtypes being 11-fold more potent toward the homomeric 5-HT3A receptor.

Inhibition of 5-HT-induced 5-HT3 receptormediated MP changes by boldine and menthol enantiomers To validate the data derived from Ca2+ influx measurement, we performed a MP assay quantifying the total net ion flux through the receptor. 5-HT concentration-dependent activation of 5-HT3 receptors in FlpInTM T-RexTM-293 5-HT3A cells and HEK293 pBud 5-HT3A/B cells revealed comparable results with the aequorin assay (Table 1). Concentration-dependent inhibition of 5-HT-evoked activation of 5-HT3A- (5-HT 0.22 lmol/L) and 5-HT3AB receptors (5-HT 0.77 lmol/L) by boldine and menthol enantiomers qualitatively confirmed the aequorin assay results (Fig. 2; Table 1). The Ki of boldine was about 6.5-fold lower on the 5-HT3A- compared to that on the 5-HT3AB receptor (p < 0.05). Although the pIC50

Impact of boldine and ( )-menthol on [3H]GR65630 binding The extent of expression of 5-HT3A receptors in HEK293 5-HT3A/apoaequorin cells and of 5-HT3AB receptors in transiently transfected HEK293 cells was determined by saturation binding on membranes with the 5-HT3 receptor antagonist [3H]GR65630. Results were presented in a recent publication26 and were similar to those determined in a previous study.5 For

6

© 2014 John Wiley & Sons Ltd

Boldine and menthol inhibit human 5-HT3 receptors

A

B

C

Figure 3 [3H]GR65630 (0.2 nM) binding to membranes of HEK293 cells heterologously expressing 5-HT3A- or 5-HT3AB receptors in the presence of increasing concentrations of natural compounds. Data are expressed as percentages of [3H]GR65630 binding in the absence of the drug under study. Shown are the means  SEM of 3–6 independent experiments carried out in duplicate. Non-specific binding was determined on mock-transfected cells. Parameters derived from these curves are shown in Table 1.

line (Fig. 3A, Table 1), confirming the expression of 5HT3AB receptors in these cells. Boldine concentration-dependently displaced the radioligand from its binding site (Fig. 3B), whereas ( )-menthol did not reduce [3H]GR65630 binding up to a concentration of 1 mM (Fig. 3C). In contrast to its greater inhibition potency on 5-HT3A receptors, the determined affinities of boldine did not differ between the two receptor subtypes (Fig. 3B; Table 1).

competition experiments, the two-fold Kd concentration of 0.2 nM [3H]GR65630 was used. The expression of heteromeric 5-HT3AB receptors in transiently transfected HEK293 cells was examined by competition binding with the discriminating agonist 5-HT. The resulting inhibition curves revealed a significantly increased Ki and a reduced Hill slope for 5-HT on membranes of 5-HT3A/B-expressing cells compared to the parameters of the 5-HT3A receptor-expressing cell

B

A

C

Figure 4 Expression analysis of 5-HT3A and 5-HT3B subunits. (A) Immunocytochemical analysis of the immunoreactivity of 5-HT3 subunits within the human colon. Cross-sections of cryopreserved colon samples were stained by using respective antibodies. Staining of the mucosal gut layer (top panel): 5-HT3A and 5-HT3B are visible in enterocytes of the epithelial lining and in distinct cell populations within the lamina propria, presumably in immune cells. The immunoreactivities of 5-HT3A and 5-HT3B subunits (red fluorescence, Alexa Fluorâ568) are coincident with the staining for the neuronal marker HuC/D (green fluorescence, Alexa Fluorâ488) in the perikarya of neurons of the myenteric plexus within the muscular layer (bottom panel). Blue staining marks nuclei (Hoechst counterstain). The bar represents 500 lm. (B) Confirmation of HTR3 gene expression by nCounter analysis from human colon tissue. Microdissected tissue of the mucosal cell layer, the epithelium, the lamina propria, the muscular layer, and the myenteric plexus was analyzed. Both HTR3 genes are expressed to a different extent within enterocytes of the epithelium and in the lamina propria, in the muscular layer, and in the myenteric plexus. Values are means  SEM of three different individuals. (C) Schematic illustration of 5-HT3A and 5-HT3B subunit expression in different regions of the human colon (red label): within the gut mucosa (bottom): in enterocytes and immune cells of the lamina propria; within neurons of the myenteric plexus (top) and, as revealed by a previous study,45 within neurons of the submucosal plexus (middle).

© 2014 John Wiley & Sons Ltd

7

J. Walstab et al.

Neurogastroenterology and Motility

ical differences in the two assays. The receptordiscriminating effect of boldine was also obvious in our 5-HT concentration–response experiments in the presence of increasing concentrations of boldine on HEK293 cells expressing 5-HT3A/B subunits. We previously demonstrated that the biphasic concentration–response curve of 5-HT on these cells is presumably due to the co-expression of homomeric 5-HT3Anext to heteromeric 5-HT3AB receptors.5 Here, the presence of increasing boldine concentrations led to steeper and more monophasic curves and thus, to the complete inhibition of the lower high affinity part representing homomeric 5-HT3A receptors. In contrast to the results of the two functional assays, competition binding experiments with the 5-HT3 receptor antagonist [3H]GR65630 revealed a similar affinity of boldine toward both receptor subtypes. The same behavior was found in previous studies for d-tubocurarine, having a higher potency for the inhibition of 5-HT3A- vs 5-HT3AB receptors2 while exhibiting the same affinity toward both receptors.34 Our results, and those of other compounds34 binding to the orthosteric ligand binding site and having the same affinity for both 5-HT3 receptor subtypes, point to the fact that this binding site is not different between 5-HT3A- and 5-HT3AB receptors. This assumption is corroborated by a recent study, in which functional and ligand-binding studies on human 5-HT3AB receptors with mutated ligandbinding domains revealed that orthosteric ligands exclusively bind to an A-A interface.35 Consequently, pure competitive 5-HT3 receptor antagonists would not be able to differentiate between 5-HT3A- and 5-HT3AB receptors. Our study revealed an exclusively competitive mode of action for boldine, however, the interaction with an additional site other than the orthosteric binding site causing the receptor-differentiating effect cannot be ruled out for boldine and d-tubocurarine (also discussed in Ref. [36]). ( )-Menthol non-competitively impeded 5-HT3Aand 5-HT3AB receptor activation. In the Ca2+ assay, ( )-menthol but not (+)-menthol was significantly more potent in inhibiting the activation of 5-HT3Avs 5-HT3AB receptors; this stereoselective effect was previously shown for the potentiation of GABAA receptors.17 In the MP assay, the stereoselective inhibition was even more pronounced: ( )-menthol was about 11-fold more potent toward 5-HT3A- vs 5-HT3AB receptors, whereas (+)-menthol was only two-fold more potent toward 5-HT3A receptors. The observed stereoselectivity confirms a specific interaction with 5-HT3 receptors and argues against an exclusively non-specific effect of the lipophilic compound on membrane fluidity.37 The greater inhibition potency of ( )-menthol on

Expression of 5-HT3A- and 5-HT3B subunits in human colon sections Expression of 5-HT3A- and B subunits was analyzed by immunofluorescence on human colon cryosections. Sections were either immunostained with rabbit anti-5HT3A- or anti-5-HT3B antibodies. To specify the nature of 5-HT3-positive cells, sections were co-stained with the neuronal marker antibody mouse anti-HuC/D. Immunoreactivities of 5-HT3A- and 5-HT3B subunits were coincident with anti-HuC/D in cell bodies of myenteric neurons. Both subunits were expressed in enterocytes of the colonic epithelium, in immune cells, for example macrophages, and in mast cells of the lamina propria. Light immunoreactivity was also detectable in muscular cells (Fig. 4A and C).

HTR3A- and HTR3B mRNA expression in microdissected human colon layers We performed complementing expression analyses of the 5-HT3 subunit genes HTR3A and HTR3B on mRNA level in colon tissue using the nCounter technology. The colonic mucosal epithelium and the muscularis externa were isolated by microdissection. The separated layers were analyzed confirming that HTR3A and HTR3B mRNA is present in the epithelium and the lamina propria. The transcription in the lamina propria correlated with the detected staining in immune cells, presumably in macrophages and mast cells (above section). For both genes, presence of mRNA was confirmed in the myenteric plexus and the muscular layer (Fig. 4B). The identity of microdissected tissue subregions was confirmed using specific marker genes for the respective subregions (Fig. SD2).

DISCUSSION We show that the aporphine alkaloid boldine and the monoterpene alcohol menthol inhibit the activation of human 5-HT3 receptors in the micromolar concentration range. According to our radioligand binding studies and concentration–response experiments of 5-HT in the presence of the respective antagonist, boldine behaves as a competitive antagonist, whereas menthol non-competitively inhibits the activation of homomeric 5-HT3A- and heteromeric 5-HT3AB receptors. In the Ca2+ assay, boldine was about ten-fold more potent in inhibiting the activation of 5-HT3A- compared to 5-HT3AB receptors. This is in line with the 6.5-fold greater potency toward 5-HT3A receptors determined in the MP assay. The generally lower IC50 values of boldine found in the MP assay may be due to method-

8

© 2014 John Wiley & Sons Ltd

Boldine and menthol inhibit human 5-HT3 receptors

of ( )-menthol in the gut can be expected. Therefore, the inhibition of 5-HT3 receptors by menthol in the middle micromolar range may be well involved in the beneficial effects of peppermint oil against the symptoms of IBS. To substantiate the potential relevance of the herein found interactions, we were the first to perform 5-HT3A- and 5-HT3B protein and mRNA expression analyses in human colon material. Previous analyses revealed expression of 5-HT3 subunits in defined human colon cell layers: co-expression of 5-HT3Aand 5-HT3B subunits within the submucosal plexus,45 co-expression of 5-HT3A- with 5-HT3C, D, and E subunits in cell bodies of myenteric neurons, and of 5-HT3A and 5-HT3D in the submucosal plexus.31 Our experiments confirmed the co-expression of 5-HT3Aand 5-HT3B subunits in all gut layers, that is, the epithelium, the lamina propria, the myenteric plexus, and the muscular cell layer. Consequently, our gained data may be of significant importance for physiologically relevant GI receptors that may indeed be targeted by menthol and boldine. In summary, we showed, using radioligand binding, Ca2+ influx and MP measurements, that the aporphine alkaloid boldine competitively and the monoterpene alcohol menthol non-competitively inhibit human 5-HT3 receptors. Our findings may help to explain the widespread medicinal use of ( )-menthol and peppermint oil to combat symptoms of GI disorders and may encourage studies with boldine for the treatment of FGIDs, emesis, or psychiatric disorders. The predominant inhibition of homomeric 5-HT3A receptors over heteromeric 5-HT3AB receptors by boldine and ( )-menthol makes them to potentially useful pharmacological tools to unravel the structural differences between these receptor subtypes.

5-HT3A- vs 5-HT3AB receptors probably results from the lower relative Ca2+ permeability of heteromeric vs homomeric receptors2: as ( )-menthol more potently blocks Ca2+ channels than Na+ channels,38 it may in turn preferably block the influx of Ca2+ through 5-HT3 receptors. This might explain its lower inhibition potency determined in the MP assay, measuring total net ion flux primarily led by Na+. It would further confirm menthol’s much lower inhibition potency on 5-HT3AB- vs 5-HT3A receptors in the MP assay, which is probably due to the comparatively smaller Ca2+ component of the recorded signal. Our findings of menthol acting as a non-competitive 5-HT3 receptor antagonist are corroborated by recent studies on murine cell lines39 or Xenopus oocytes40 published during the preparation of this manuscript. The determined affinity and inhibition potency of boldine toward human 5-HT3 receptors is in the range of its affinity for dopamine D1- and D2 receptors, 5-HT2A-, and neuronal nACh receptors.22–24 With regard to its 5-HT3 receptor antagonistic effect and the previously demonstrated affinities toward dopamine- and 5-HT2A receptors, boldine and boldo extracts may be effective against symptoms of psychiatric disorders as all mentioned receptors are involved in psychiatric diseases. This assumption is corroborated by the neuroleptic efficacy of boldine in mice.41 Besides its potential efficacy in the treatment of FGIDs, boldine may combat nausea and vomiting. Here, two main players are dopamine D2- and 5-HT3 receptors and receptor antagonists such as metoclopramide or setrons are among the most effective drugs to counteract these symptoms.6 Therefore, trials studying the effectiveness of boldine for treatment of neurogastrointestinal and psychiatric disorders are warranted. The determined 5-HT3 receptor inhibition potency of menthol in the middle to higher micromolar range fits to the values found for potentiation of GABAA- and glycine receptors17 and the inhibition of neuronal nACh receptors.18 Regarding menthol’s rather high IC50 on voltage-gated Ca2+ or Na+ channels of ~250– 500 lmol/L,15,42,43 the contribution of 5-HT3 receptors to its diverse pharmacological actions in vivo and especially to its relaxing effect in the GI tract is probable. A pharmacokinetic study involving 12 subjects revealed a maximum plasma concentration of 17 lmol/L following the oral application of 100 mg ( )-menthol.44 In a 4-week-randomized-controlled trial, treatment with 450 mg enteric-coated peppermint oil twice daily improved abdominal symptoms in patients with IBS13; here, even a higher plasma concentration and much higher local concentrations

© 2014 John Wiley & Sons Ltd

ACKNOWLEDGMENTS We thank Michael Wilhelm for helpful discussion and Jutta Scheuerer for excellent technical assistance.

FUNDING This work was supported by the German Cancer Aid (108710 and 109226 to BN), the Prof. Karl und Gerhard Schiller-Stiftung (to BN) and the Swiss National Science Foundation (grant 133141 to RH). Work of FL was funded by DFG/SFB 938 TP Z2 (www. gezeh.de).

CONFLICTS OF INTEREST The authors have no competing interests.

9

J. Walstab et al.

Neurogastroenterology and Motility

AUTHOR CONTRIBUTION

manuscript; SS: performed experiments; RR: performed experiments, data analysis; FL: performed experiments; MW: contributed essential reagents, draft manuscript; HB: draft manuscript; BN: study design, grant supply, draft and final manuscript.

JW: study design, performed experiments, data analysis, draft and final manuscript; CW: performed experiments, data analysis; RH: grant supply, performed experiments, data analysis, draft

REFERENCES 1 Boess FG, Beroukhim R, Martin IL. Ultrastructure of the 5-hydroxytryptamine3 receptor. J Neurochem 1995; 64: 1401–5. 2 Davies PA, Pistis M, Hanna MC, Peters JA, Lambert JJ, Hales TG, Kirkness EF. The 5-HT3B subunit is a major determinant of serotoninreceptor function. Nature 1999; 397: 359–63. 3 Maricq AV, Peterson AS, Brake AJ, Myers RM, Julius D. Primary structure and functional expression of the 5HT3 receptor, a serotonin-gated ion channel. Science 1991; 254: 432–7. 4 Niesler B, Frank B, Kapeller J, Rappold GA. Cloning, physical mapping and expression analysis of the human 5HT3 serotonin receptor-like genes HTR3C, HTR3D and HTR3E. Gene 2003; 310: 101–11. 5 Walstab J, Hammer C, Bonisch H, Rappold G, Niesler B. Naturally occurring variants in the HTR3B gene significantly alter properties of human heteromeric 5-hydroxytryptamine-3A/B receptors. Pharmacogenet Genomics 2008; 18: 793–802. 6 Walstab J, Rappold G, Niesler B. 5-HT (3) receptors: role in disease and target of drugs. Pharmacol Ther 2010; 128: 146–69. 7 Niesler B. 5-HT(3) receptors: potential of individual isoforms for personalised therapy. Curr Opin Pharmacol 2011; 11: 81–6. 8 Chang L, Chey WD, Harris L, Olden K, Surawicz C, Schoenfeld P. Incidence of ischemic colitis and serious complications of constipation among patients using alosetron: systematic review of clinical trials and post-marketing surveillance data. Am J Gastroenterol 2006; 101: 1069–79. 9 Hussain Z, Quigley EM. Systematic review: complementary and alternative medicine in the irritable bowel syndrome. Aliment Pharmacol Ther 2006; 23: 465–71. 10 Papathanasopoulos A, Rotondo A, Janssen P, Boesmans W, Farre R, Vanden Berghe P, Tack J. Effect of acute peppermint oil administration on gastric sensorimotor function and

11

12

13

14

15

16

17

18

19

nutrient tolerance in health. Neurogastroenterol Motil 2013; 25: e263– 71. Kline RM, Kline JJ, Di Palma J, Barbero GJ. Enteric-coated, pH-dependent peppermint oil capsules for the treatment of irritable bowel syndrome in children. J Pediatr 2001; 138: 125–8. Liu JH, Chen GH, Yeh HZ, Huang CK, Poon SK. Enteric-coated peppermint-oil capsules in the treatment of irritable bowel syndrome: a prospective, randomized trial. J Gastroenterol 1997; 32: 765–8. Cappello G, Spezzaferro M, Grossi L, Manzoli L, Marzio L. Peppermint oil (Mintoil) in the treatment of irritable bowel syndrome: a prospective double blind placebo-controlled randomized trial. Dig Liver Dis 2007; 39: 530–6. Hills JM, Aaronson PI. The mechanism of action of peppermint oil on gastrointestinal smooth muscle. An analysis using patch clamp electrophysiology and isolated tissue pharmacology in rabbit and guinea pig. Gastroenterology 1991; 101: 55–65. Swandulla D, Carbone E, Schafer K, Lux HD. Effect of menthol on two types of Ca currents in cultured sensory neurons of vertebrates. Pflugers Arch 1987; 409: 52–9. Hawthorn M, Ferrante J, Luchowski E, Rutledge A, Wei XY, Triggle DJ. The actions of peppermint oil and menthol on calcium channel dependent processes in intestinal, neuronal and cardiac preparations. Aliment Pharmacol Ther 1988; 2: 101–18. Hall AC, Turcotte CM, Betts BA, Yeung WY, Agyeman AS, Burk LA. Modulation of human GABAA and glycine receptor currents by menthol and related monoterpenoids. Eur J Pharmacol 2004; 506: 9–16. Hans M, Wilhelm M, Swandulla D. Menthol suppresses nicotinic acetylcholine receptor functioning in sensory neurons via allosteric modulation. Chem Senses 2012; 37: 463–9. van Wyk B-E, Wink M. Medicinal Plants of the World. Pretoria: Briza, 2004.

10

20 Speisky H, Squella JA, Nunez-Vergara LJ. Activity of boldine on rat ileum. Planta Med 1991; 57: 519–22. 21 Speisky H, Cassels BK. Boldo and boldine: an emerging case of natural drug development. Pharmacol Res 1994; 29: 1–12. 22 Asencio M, Delaquerriere B, Cassels BK, Speisky H, Comoy E, Protais P. Biochemical and behavioral effects of boldine and glaucine on dopamine systems. Pharmacol Biochem Behav 1999; 62: 7–13. 23 Chung LY, Lo MW, Mustafa MR, Goh SH, Imiyabir Z. 5-Hydroxytryptamine2A receptor binding activity of compounds from Litsea sessilis. Phytother Res 2009; 23: 330–4. 24 Iturriaga-Vasquez P, Perez EG, Slater EY, Bermudez I, Cassels BK. Aporphine metho salts as neuronal nicotinic acetylcholine receptor blockers. Bioorg Med Chem 2007; 15: 3368–72. 25 Wink M. Interference of alkaloids with neuroreceptors and ion channels. In: Atta-ur-Rahman , ed. Studies in Natural Products Chemistry; Bioactive Natural Products. Amsterdam: Elsevier, 2000; 21B: 3–129. 26 Walstab J, Kruger D, Stark T, Hofmann T, Demir IE, Ceyhan GO, Feistel B, Schemann M, et al. Ginger and its pungent constituents noncompetitively inhibit activation of human recombinant and native 5-HT3 receptors of enteric neurons. Neurogastroenterol Motil 2013; 25: 439–47, e302. 27 Combrink S, Kostanian A, Walstab J, Barann M, Bruss M, Gothert M, Bonisch H. Characterization of the naturally occurring Arg344His variant of the human 5-HT 3A receptor. Pharmacol Rep 2009; 61: 785–97. 28 Walstab J, Combrink S, Bruss M, Gothert M, Niesler B, Bonisch H. Aequorin luminescence-based assay for 5-hydroxytryptamine (serotonin) type 3 receptor characterization. Anal Biochem 2007; 368: 185–92. 29 Cheng Y, Prusoff WH. Relationship between the inhibition constant (K1) and the concentration of inhibitor which causes 50 per cent inhibition

© 2014 John Wiley & Sons Ltd

Boldine and menthol inhibit human 5-HT3 receptors

30

31

32

33

34

(I50) of an enzymatic reaction. Biochem Pharmacol 1973; 22: 3099–108. Lew MJ, Angus JA. Analysis of competitive agonist-antagonist interactions by nonlinear regression. Trends Pharmacol Sci 1995; 16: 328–37. Kapeller J, Moller D, Lasitschka F, Autschbach F, Hovius R, Rappold G, Bruss M, Gershon ND et al. Serotonin receptor diversity in the human colon: expression of serotonin type 3 receptor subunits 5-HT3C, 5-HT3D, and 5-HT3E. J Comp Neurol 2011; 519: 420–32. Chomczynski P, Sacchi N. Singlestep method of RNA isolation by acid guanidinium thiocyanate-phenolchloroform extraction. Anal Biochem 1987; 162: 156–9. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A, Spieleman F. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 2002; 3: RESEARCH0034. Brady CA, Stanford IM, Ali I, Lin L, Williams JM, Dubin AE, Hope AG, Barnes NM. Pharmacological comparison of human homomeric 5-HT3A receptors versus heteromeric 5-HT3A/3B receptors. Neuropharmacology 2001; 41: 282–4.

35 Thompson AJ, Price KL, Lummis SC. Cysteine modification reveals which subunits form the ligand binding site in human heteromeric 5-HT3AB receptors. J Physiol 2011; 589: 4243– 57. 36 Thompson AJ, Lummis SC. Discriminating between 5-HT(3)A and 5-HT (3)AB receptors. Br J Pharmacol 2013; 169: 736–47. 37 Wink M. Evolutionary advantage and molecular modes of action of multicomponent mixtures used in phytomedicine. Curr Drug Metab 2008; 9: 996–1009. 38 Pan R, Tian Y, Gao R, Li H, Zhao X, Barrett JE, Hu H. Central mechanisms of menthol-induced analgesia. J Pharmacol Exp Ther 2012; 343: 661– 72. 39 Heimes K, Hauk F, Verspohl EJ. Mode of action of peppermint oil and ( )-menthol with respect to 5-HT3 receptor subtypes: binding studies, cation uptake by receptor channels and contraction of isolated rat ileum. Phytother Res 2011; 25: 702–8. 40 Ashoor A, Nordman J, Veltri D, Susan Yang KH, Shuba Y, Al Kury L, Sadek B, Howarth FC et al. Menthol inhibits 5-HT3 receptor-mediated currents. J Pharmacol Exp Ther 2013; 347: 398– 409.

41 Zetler G. Neuroleptic-like, anticonvulsant and antinociceptive effects of aporphine alkaloids: bulbocapnine, corytuberine, boldine and glaucine. Arch Int Pharmacodyn Ther 1988; 296: 255–81. 42 Haeseler G, Maue D, Grosskreutz J, Bufler J, Nentwig B, Piepenbrock S, Dengler R, Leuwer M. Voltagedependent block of neuronal and skeletal muscle sodium channels by thymol and menthol. Eur J Anaesthesiol 2002; 19: 571–9. 43 Sidell N, Verity MA, Nord EP. Menthol blocks dihydropyridine-insensitive Ca2+ channels and induces neurite outgrowth in human neuroblastoma cells. J Cell Physiol 1990; 142: 410–9. 44 Gelal A, Jacob P 3rd, Yu L, Benowitz NL. Disposition kinetics and effects of menthol. Clin Pharmacol Ther 1999; 66: 128–35. 45 Michel K, Zeller F, Langer R, Nekarda H, Kruger D, Dover TJ, Brady CA, Barnes NM et al. Serotonin excites neurons in the human submucous plexus via 5-HT3 receptors. Gastroenterology 2005; 128: 1317–26.

SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article: Figure S1. Concentration-dependent 5-HT-induced Ca2+ influx through 5-HT3A- and 5-HT3AB receptors. Figure S2. nCounter analysis of marker genes in total RNA from microdissected tissue of the human colon. Data S1. Methods: selection and maintenance of stably transfected cells; Results: serotonin-induced Ca2+ influx through 5-HT3A and 5-HT3AB. Table S1. Gene names, accession numbers and nucleotide sequences of nCounter probes used for detection of mRNA expression.

© 2014 John Wiley & Sons Ltd

11