Extracts from peppermint leaves, lemon balm leaves and in particular angelica roots mimic the pro-secretory action of the herbal preparation STW 5 in the human intestine

Phytomedicine 22 (2015) 1063–1070

Contents lists available at ScienceDirect

Phytomedicine journal homepage: www.elsevier.com/locate/phymed

Extracts from peppermint leaves, lemon balm leaves and in particular angelica roots mimic the pro-secretory action of the herbal preparation STW 5 in the human intestine Shady Allam a, Dagmar Krueger a, Ihsan Ekin Demir b, Gueralp Ceyhan b, Florian Zeller c, Michael Schemann a,∗ a

Human Biology, Technische Universität München, Freising, Germany Surgery, Klinikum Rechts der Isar Technische Universität München, Munich, Germany c Surgery Department, Clinic Freising, Freising, Germany b

a r t i c l e

i n f o

Article history: Received 5 April 2015 Revised 14 August 2015 Accepted 17 August 2015

Keywords: Intestinal secretion STW 5 Chloride channels Angelica Lemon balm Peppermint

a b s t r a c t Aim: The herbal preparation STW 5 contains fresh plant extracts from bitter candytuft whole plant, extracts from greater celandine herb, angelica root, lemon balm leaves, peppermint leaves, caraway fruit, liquorice root, chamomile flower and milk thistle fruit. We recently reported that STW 5 increased intestinal chloride secretion and proposed that this action may be involved in its clinical efficacy in the treatment of irritable bowel syndrome. The aim of this study was to identify the extracts responsible for the secretory action in order to provide the basis to develop novel target oriented herbal combinations. Methods: We used the Ussing chamber voltage clamp technique to study the effects of individual extracts of STW 5 on short circuit current (Isc, reflecting electrogenic ion transport across epithelial cells) in mucosal/submucosal preparations of human small or large intestinal specimens and the human epithelial cell line T84. Results: STW 5 at concentrations of 512 μg/ml and 5120 μg/ml evoked an increase in Isc. The increase at the lower concentration was due to pro-secretory effects of angelica which were nerve mediated. The increase at the higher concentration was additionally mimicked by peppermint and lemon balm. The remaining extracts did not influence ISC in the large intestine. The results were similar in T84 cells except that angelica had no effect while chamomile induced secretion. These pro-secretory effects were reduced by adenylate cyclase inhibitor MDL-12330A, cystic fibrosis transmembrane conductance regulator (CFTR) inhibitor CFTRinh -172 and calcium activated chloride channels blocker 4-acetamido-4-isothiocyanatostilbene-2,2-disulphonic acid (SITS). Liquorice decreased ISC only in small intestine which was reversed by the epithelial sodium channel blocker amiloride. Conclusions: Results suggested that the pro-secretory action of STW 5 is mainly due to angelica with lesser contribution of peppermint and lemon balm. Their effects involve activation of cAMP- and Ca++ -activated Cl− channels. We suggest that peppermint, lemon balm and in particular angelica may be the basis to develop novel herbal preparations to specifically treat secretory disorder based on impaired epithelial secretion, such as constipation. © 2015 Elsevier GmbH. All rights reserved.

Introduction

Abbreviations: APL, angelica + peppermint + liquorice; APLC, angelica + peppermint + liquorice + chamomile; cAMP, cyclic adenosine monophosphate; Ca-Cl, calcium dependent chloride channel; CFTR, cystic fibrosis transmembrane conductance regulator; DMSO, dimethyl sulphoxide; ISC , short circuit current; SITS, 4-acetamido-4isothiocyanatostilbene-2,2-disulphonic acid; TTX, terodotoxin. ∗ Corresponding author. Tel.: +49 8161 715483; fax: +49 8161 715785. E-mail address: [email protected] (M. Schemann). http://dx.doi.org/10.1016/j.phymed.2015.08.008 0944-7113/© 2015 Elsevier GmbH. All rights reserved.

Irritable bowel syndrome is a functional bowel disorder characterized by stool irregularities. Some newly developed drugs, such as lubiprostone and linaclotide specifically target intestinal secretion in order to improve in particular constipation (Andresen et al. 2007; Johanson et al. 2008; Thomas and Allmond 2013). Herbal medicine is increasingly used to treat in particular functional gastrointestinal diseases (Brierley and Kelber 2011; Rahimi and Abdollahi 2012). One of these medications is the herbal preparation STW 5, which is successfully used for over 50 years to treat

1064

S. Allam et al. / Phytomedicine 22 (2015) 1063–1070

patients with functional dyspepsia (Madisch et al. 2001; Melzer et al. 2004; Madisch et al. 2004a). STW 5 and is a fixed herbal combination of a fresh plant extract from bitter candytuft (Iberis amara) and extracts from dried greater celandine herb, angelica root, lemon balm leaves, peppermint leaves, caraway fruit, liquorice root, chamomile flower and milk thistle fruit (Wegener and Wagner 2006). There is one clinical study suggesting that STW 5 was also beneficial in patients with irritable bowel syndrome (IBS) improving visceral pain as well as the stool irregularities (Madisch et al. 2004b). We reported recently that STW 5 increased ion secretion in the human small and large intestine and proposed that this pro-secretory activity may be involved in its clinical efficacy (Krueger et al. 2009). The increased secretion was mediated by opening of cAMP and calcium dependent chloride channels because the adenylate cyclase inhibitor MDL-12,330A as well as 4-acetamido-4-isothiocyanatostilbene-2,2disulphonic acid (SITS) antagonized the actions of STW 5. In the previous study we did not address the question which of the individual components was responsible for the enhanced ion secretion. Results of such a study may further demonstrate the potential for selective disease-targeted combinations and particularly stimulate development of specific herbal preparations that target epithelial chloride secretion. We therefore aimed to identify the plant extract or extracts that were responsible for the secretory action of STW 5 and studied its effect on human small and large intestinal preparations as well as in the T84 epithelial cell line. Material and methods All experimental procedures have been previously published in detail (Krueger et al. 2009, Krueger et al. 2013). Tissue samples Tissue preparations were obtained from 119 patients with a mean age of 64 ± 15 years (ranging from 14 to 93 years) who underwent routine surgical operations at the hospitals in Freising and Klinikum Rechts der Isar of the Technische Universität München. Diagnoses that led to surgeries were (number of patients in parenthesis): stomach carcinoma (15), pancreatic carcinoma (10), large intestine carcinoma (55), diverticulitis (7), ovarian carcinoma (3), polyps (3), allergic eosinophilic gastroenteritis (AEG) (3), large intestinal stenosis (2), elonageted sigma (2), ileostoma reversal (3), unspecified reasons (4), gall bladder carcinoma (2), small intestine carcinoma (2), chronic ileus (1), esophageal shift (1), severe motility disorder after gastrectomy (1), fistula (1), small intestine volvulus (1), small intestine perforation (1), large intestinal stoma (1) and intestinal obstruction (1). Experiments were performed in 543 mucosal/submucosal preparations 264 and 279 of which were derived from small and large intestine, respectively. All procedures were approved by the ethics committee of the Technische Universität München (1748/07 and 2595/09) with the informed patient’s consent. The preparations were from macroscopically unaffected areas as determined by the pathologists. Immediately after resection, samples were transferred to the laboratory under aseptic conditions in cold oxygenated sterile Krebs buffer. The Krebs solution contained (in mM) 117 NaCl, 4.7 KCl, 1.2 MgCl2 , 1.2 NaH2 PO4 , 25 NaHCO3 , 2.5 CaCl2 and 11 glucose (all from Sigma-Aldrich, Steinheim, Germany). Transferred specimen were washed three times with icecold, carbogen-aerated Krebs buffer and then dissected to obtain mucosal/submucosal preparations containing the submucous plexus. Human epithelial cell line T84 T84 cells (ECACC, Salisbury, UK) were seeded on Millipore filters (Bedford, MA, USA) with 0.45 μm pore size, and incubated at

37 °C and 95% O2 and 5% CO2 (Carbogen) in Dulbecco’s modified Eagle medium (DMEM) / Ham’s Nutrient Mixture F-12, supplemented with 10% heat-inactivated fetal calf serum, 100 IU ml-1 penicillin, 100 μg ml−1 streptomycin and 2.75 μg ml−1 amphotericin B (all from Sigma-Aldrich, Schnelldorf, Germany) to attain a monolayer. The medium was replaced daily. After 12–14 days, filters were mounted in Ussing chambers for ion transport studies. Experiments were performed on 760 T84 cell filter discs. Ussing chamber experiments To test the effects of STW 5 and its individual extracts on ion transport in intact human mucosal/submucosal preparations and T84 cells, we used the Ussing voltage clamp technique (Easy Mount chambers, Physiologic Instruments, San Diego, CA, USA). The tissue specimens were mounted into plexiglass Ussing chambers with exposure area of 0.5 cm2 . Mucosal and serosal sides were bathed separately in 5 ml Carbogen-bubbled Krebs solution maintained at 37 °C. The set-up allowed simultaneous measurements of up to eight mucosal/submucosal preparations dissected from one specimen. We recorded short-circuit current (Isc ) as a measure for the transepithelial electrogenic transport (expressed in μA/cm2 ; the values were corrected for bath resistance). As previously described for the effects of STW 5 positive Isc indicated a net anion current from the serosa to the lumen (Krueger et al. 2009). Tissues were electrically stimulated by silver electrodes placed on either side of the tissues and connected to a constant voltage stimulator (Grass SD-9; Astro-Med Inc., West Warwick, RI, USA). Such an electrical field stimulation (EFS) with short pulses selectively induced nerve mediated transepithelial ion fluxes. The EFS was achieved by delivering a train of pulses with supramaximal stimulus parameters: pulse amplitude, 20 V; pulse frequency 10 Hz; pulse duration 1 ms; train duration 10 s. Last but not least we measured tissue resistances. Before starting the actual measurements, human tissues and T84 cells were allowed to equilibrate for at least 45 min or 20 min, respectively. STW 5, its individual extracts, combination of extracts and all other drugs were applied basolateraly to the serosal bathing solution except CFTRinh -172, SITS and amiloride which were added luminally to the mucosal bathing solution. Neither pH nor osmolality of the Krebs solution changed after addition of any of the drugs. Drugs and herbal extracts STW 5 (batch number E90722; Iberogast® ) and its individual extracts were provided by Steigerwald Arzneimittelwerk GmbH, Darmstadt, Germany (Table 1). They belong according to the guidelines of the European Medicines Agency to “other herbal substances”. The extraction processes as well as the quality controls were as previously described in detail (Kroll and Cordes 2006). Briefly, the quality of the compounds was controlled according to Good Manufacturing Practice and Good Agricultural Practice of Medicinal and Aromatic Plants (cited and outlined in Kroll and Cordes 2006). The quality of each extract is tested according to individual specification, among which is the identity of the used drug (thin layer chromatography fingerprint) and the content of marker substances (Table 1) within a defined range of ±5% measured by high performance liquid chromatography or gasliquid chromatography. The lyophilized extracts were dissolved in Krebs buffer and added basolaterally. We applied the individual extracts of angelica root, caraway fruit, chamomile flower, greater celandine herbs, bitter candytuft, lemon balm leaves, liquorice root, milk thistle fruit and peppermint leaves at final concentrations of 89.5, 28.4, 114.3, 62.9, 27.3, 57.9, 80.1, 14.4 and 37.2 μg/ml respectively, which corresponded to their concentrations in 512 μg/ml STW 5; this concentration evoked a reliable pro-secretory action (Krueger et al. 2009). The tenfold higher

S. Allam et al. / Phytomedicine 22 (2015) 1063–1070

1065

Table 1 Individual extracts of the herbal preparation STW 5 used in this study. Name (batch number)

Botanical namea

Plant part

Extraction solvent (V/V)

Drug-extract ratio

% in 100 ml Iberogast

Characteristic compounds (marker substances)

Bitter candytuft (92064) Peppermint (82115) Chamomile (82174) Liquorice (82130) Angelica (92061) Caraway (82184) Milk thistle (92042)

Iberis amara L.

50 % Ethanol

1 : 1.5–2.5

15

Mentha piperita L. Matricaria chamomilla L. Glycyrrhiza glabra L. Angelica archangelica L. Carum carvi L. Silybum marianum (L.) Gaertn. Melissa officinalis L.

Fresh whole plant Dried herbs Dried flowers Dried roots Dried roots Dried fruits Dried fruits

30 % Ethanol 30 % Ethanol 30 % Ethanol 30 % Ethanol 30 % Ethanol 30 % Ethanol

1 : 2.5–3.5 1 : 2–4 1 : 2.5–3.5 1 : 2.5–3.5 1 : 2.5–3.5 1 : 2.5–3.5

5 20 10 10 10 10

Kaempferol-3,4-di-O-β -glucopyranoside-7O-α -rhamnopyranoside Menthol Bisabolol oxide A Glycyrrhizic acid Osthole Carvone N-malonyltryptophan

Dried leaves

30 % Ethanol

1 : 2.5–3.5

10

Rosmarinic acid

Chelidonium majus L.

Dried herbs

30 % Ethanol

1 : 2.5–3.5

10

Chelidonic acid

Lemon balm (82142) Greater celandine (92043) a

Plant names have been checked on: http://www.theplantlist.org./

concentration corresponded to 5120 μg/ml STW 5, which is still subtherapeutic, representing one tenth of the recommended therapeutic dose of 51.3 mg STW 5 provided by 20 drops (equivalent to 1 ml Iberogast® ). Stock solutions of amiloride, 4-acetamido-4isothiocyanatostilbene 2 (SITS), forskolin, MDL-12,330A and CFTRinh 172 (all from Sigma-Aldrich, Schnelldorf, Germany) were prepared in dimethyl sulphoxide (DMSO). Tetrodotoxin (TTX; Tocris Cookson, Bristol, UK) was dissolved in distilled water. At the final concentrations used none of the solvents had any effect on tissue functions. Data analysis and statistics Changes in Isc after basolateral applications of STW 5, its individual extracts, and their combinations were tested with one sample t-test for normally distributed data and with one sample signed rank for non-normally distributed data. For experiments with blockers which were added 20 min prior to test drugs we used one way repeated measurement analysis of variance (one way RM-ANOVA) followed by correction for multiple comparisons versus control by Bonferroni for normally distributed data. Kruskal–Wallis One Way ANOVA on Ranks was used for non-normally distributed data. For the remaining experiments we used paired t-test for normally and Mann–Whitney Rank Sum test for non-normally distributed data. Normally or non-normally distributed data are presented as means ± SEM or medians with the 25 and 75 quartiles in square brackets, respectively. Results Effects of individual extracts of STW 5 on epithelial secretion As previously reported by us (Krueger et al. 2009), we confirmed in the present study that STW 5 at a concentration of 512 μg/ml has a reliable pro-secretory action in human intestinal preparations and T84 cells. In the first set of experiments the nine individual extracts were individually added at concentrations corresponding to their concentrations in 512 μg/ml STW 5. At this concentration only angelica evoked a significant pro-secretory action in human tissues (Figs. 1A-I, 2 and Table 2). The increase in ISC was comparable between angelica and STW 5 [512 μg/ml] (13.2[4.9/21.4] μA/cm2 , n = 14 vs 25.8[6.6/38.6] μA/cm2 , n = 12; P = 0.06). Angelica and STW 5 had similar effects in both small and large intestinal preparations (Table 2). To ensure that the production processes did not affect the efficacy of the components we remixed the nine lyophilized extracts each at a concentration that corresponded to 512 μg/ml STW 5. This

mixture is referred to as sSTW 5. In human intestinal preparations, the pro-secretory actions of STW 5 (25.8[6.6/38.6] μA/cm2 , n = 12) was comparable to that evoked by sSTW 5 (17.7[11.1/36] μA/cm2 , n = 16; P = 1.0) (Fig. 2). This was also true in T84 cells where STW 5 and sSTW 5 showed comparable increases in ISC (5.2[3.3/8.1] vs 4.2[2.4/5.4] μA/cm2 , n = 6, P = 0.4) (Fig. 3). In the human epithelial cell line T84, only chamomile had a meaningful pro-secretory action when the extracts were used at concentrations corresponding to 512 μg/ml STW 5 (Fig. 1A-I). The amplitude of the Isc was comparable to 512 μg/ml STW 5 (3.8[2.7/5.7] vs 5.2[3.3/8.1] μA/cm2 , n = 6; P = 0.5) (Figs. 2 and 3). At concentrations present in 512 μg/ml STW 5 all other extracts evoked marginal or no secretory responses. It was striking to find that angelica potently enhanced ISC in freshly dissected intestinal preparations but not in T84 cells whereas chamomile had effects only in T84 cells. We previously reported that STW 5 activated enteric neurons (Krueger et al. 2009). Consequently, we asked the question of whether the lack of effect of angelica in T84 cells may be due to the lack of nerves in these cell line preparations. Indeed, blockade of nerve activity by TTX (1 μM) dramatically reduced the response to angelica in tissue preparations by 79% (P = 0.03). Additionally, chamomile may inhibit the nerve-driven secretion in human intestine while directly activating epithelial cells. This however was not the case because the effect of chamomile did not change after TTX (P = 0.7) (Fig. 4). In a second set of experiments we tested the effects of 512 μg/ml STW 5 and its individual extracts on secretory responses in human intestinal preparations. At concentrations corresponding to 5120 μg/ml, peppermint and lemon balm in addition to angelica significantly increased Isc in small and large intestinal preparations. Liquorice significantly decreased Isc in small intestine but still had no effect in large intestinal preparations (Fig. 2 and Table 2). The other extracts remained ineffective even at these higher concentrations. In T84 cells, peppermint, angelica, and lemon balm evoked in addition to chamomile an increase in Isc (Fig. 3). The other extracts still remained ineffective. Combined effects of pro-secretory extracts on intestinal secretion Basolateral co-application of angelica, peppermint and lemon balm [APL] had strong pro-secretory action in human intestinal preparations. At concentrations corresponding to 512 μg/ml STW 5, APL evoked increase in ISC was 16.3[9.0/21.1] μA/cm2 (n = 12) and lower than the increase observed at concentrations corresponding to 5120 μg/ml STW 5 (23.5[18.3/32.6] μA/cm2 , n = 16; P = 0.06). We found no evidence for synergistic effects of APL because the response was within the range one would expect by adding the median values

1066

S. Allam et al. / Phytomedicine 22 (2015) 1063–1070

Human tissue

T84 cells

(A)

Human tissue

(B)

Angelica (A) as in 512 µg/ml STW 5

STW 5 512 µg/ml

(C)

(E)

T84 cells

(D) Angelica (A) as in 5120 µg/ml STW 5

Lemon balm (L) as in 5120 µg/ml STW 5

(G)

Peppermint (P) as in 5120 µg/ml STW 5

(F)

Chamomile (C) as in 5120 µg/ml STW 5 Chamomi le (C) as in 5120 µg/ml STW 5

(H) A+P+L A+P+L+C Combination APL(C) as in 512 µg/ml STW 5

A+P+L A+P+L+C Combination APL(C) as in 5120 µg/ml STW 5

ΔIsc 10 µA/cm2

(I)

10 min

Liquorice as in 5120 µg/ml STW 5 Fig. 1. Representative traces illustrating the effects of STW 5 (A), its components (B-F and I) and their combination (G and H) on secretion in human intestine (left panels) and T84 cells (right panels). The traces were taken from different experiments. The scale bar in the lower right corner applies to all traces. The arrows indicate the start of basolateral drug application. A = angelica, P = peppermint, L = lemon balm and C = chamomile.

Fig. 2. Effects of STW 5 and its individual components on secretion (ISC ) in human intestinal preparations. The pro-secretory effects of at 512 μg/ml STW 5 is only mimicked by angelica (A) when applied at concentration corresponding to 512 μg/ml STW 5. At 10x higher concentrations, additionally, peppermint (P) and lemon balm (L) increased ion secretion while liquorice decreased it. The decrease with liquorice was due to its inhibitory action in small intestinal preparations only. sSTW 5 induced a comparable effect to STW 5. Combined application of extracts A + P + L (corresponding to 512 or 5120 μg/ml STW5) evoked responses which were comparable to STW 5. Numbers in parentheses indicate the number of tissues (equal to number of patients) studied. First and second values represent number of tissues studied with or corresponding to concentrations of 512 and 5120 μg/ml STW 5, respectively. ∗ mark significant changes compared to baseline Isc. Data presented as medians with 75/25 quartiles.

S. Allam et al. / Phytomedicine 22 (2015) 1063–1070

1067

Table 2 Prosecretory actions of STW 5, its components and their combinations in human small and large intestinal preparations.

STW 5 (512 μg/ml) sSTW 5 (512 μg/ml) Angelica (A) as in 512 μg/ml Angelica (A) as in 5120 μg/ml Peppermint (P) as in 5120 μg/ml Lemon balm (L) as in 5120 μg/ml Liquorice as in 5120 μg/ml APLa as in 512 μg/ml APLa as in 5120 μg/ml

Small intestine Isc [μA/cm2 ]

Large intestine Isc [μA/cm2 ]

P value

20.3 ± 8.0 30 [9.5/44.8] 15.4 ± 2.8 19.4 ± 5.1 15.3 ± 5.8 10.1 [6.8/38.0] -9.7 ± 1.7 17.4 ± 3.3 26.6 [22.0/31.8]

30.1 ± 7.6 16 [13.0/20.2] 14.1 ± 5.4 13.9 ± 2.8 8.2 ± 2.7 6.0 [5.0/10.0] -2.4 ± 1.3 14.7 ± 2.6 23.2 [18.2/24.4]

P = 0.4 P = 0.3 P = 0.8 P = 0.4 P = 0.3 P = 0.1 P = 0.005∗ P = 0.52 P = 0.5

N=6 N=9 N=8 N=7 N=6 N=6 N=6 N=5 N=5

N=6 N=7 N=6 N=6 N=6 N=7 N = 10 N=7 N = 11

a

APL is a mixture of angelica, peppermint and lemon balm. N is the number of tissues (equal to number of patients). Marks significant difference between small and large intestine. Values are presented as mean ± SEM or medians with their 25/75 quartiles. ∗

Fig. 3. Effects of STW 5 and its individual components on secretion (ISC ) in T84 cells. At concentrations corresponding to 512 μg/ml STW 5, chamomile (C) evoked a secretory response comparable to STW 5 and sSTW 5. At concentrations corresponding to 5120 μg/ml STW 5 also angelica (A), peppermint (P) and lemon balm (L) evoked increases in Isc. Co-application of A + P + L + C at concentrations corresponding to 512 and 5120 μg/ml STW 5 evoked an increase in Isc which is comparable to the calculated sum of their individual responses. ∗ Statistically significant from zero (P ≤ 0.001), ¥ significantly different from 5120 μg/ml STW 5 (P ≤ 0.001), ‡ significantly different from APLC as in 512 μg/ml STW 5 (P ≤ 0.001). At least n = 6 filters for each data point. Data presented as medians with 75/25 quartiles.

application was not significantly different from the calculated sum of the responses recorded after single extract application (47.5 ± 3.4 μA/cm2 vs. 41.6 μA/cm2 ; P = 0.1) (Fig. 3). Because of limited availability of human tissue, we used the T84 cell model to confirm that even the combined application of the five extracts with no effect on ISC evoked no secretory response (0.3 ± 0.1 μA/cm2 ; n = 5, P = 0.1). None of the extracts or their combinations had any influence on the resistance of human intestinal preparations or T84 cell line (data not shown). Only the high concentration of STW 5 (5120 μg/ml) significantly reduced the nerve-evoked secretion in response to EFS whereas all other extracts and their combinations had no effect on EFS (data not shown). Pharmacology of pro-secretory components of STW 5 Fig. 4. Angelica-evoked increase in Isc in human intestinal preparations is partly nerve mediated whilst chamomile is still inactive after nerve blockade. Pre-incubation of TTX (1 μM) significantly reduced angelica-evoked ion secretion. Chamomile did not evoke any secretory response in TTX. Numbers in parentheses indicate the number of tissues (equal to number of patients) studied. ∗ marks significant difference (P < 0.05). Data illustrated as medians and 75/25 quartiles.

for the individual extracts. The pro-secretory response of APL was comparable between small and large intestinal preparations (Table 2, Fig. 2). In T84 cells, basolateral co-application of the four pro-secretory extracts angelica, peppermint, lemon balm, and chamomile (APLC) also induced significant increases in Isc . This response to combined

In our previous study we concluded that STW 5-induced increases in Isc in human tissues and T84 cells were due to increased anion secretion via activation of two epithelial chloride channels (Krueger et al. 2009). These were the cAMP-dependent cystic fibrosis transmembrane conductance regulator (CFTR) and calcium-activated chloride channels (ClCa) (Krueger et al. 2009). To further characterize the prosecretory mode of action of basolateral applications of individual extracts or their combinations, blockers of these ion channels were applied in human intestine and T84 cells. Pre-treatment of human intestinal preparations with the adenylate cyclase inhibitor MDL-12,330A significantly reduced the increase in Isc evoked by STW 5, sSTW 5, the prosecretory individual extracts and their combinations (Fig. 5). The results were similar in

1068

S. Allam et al. / Phytomedicine 22 (2015) 1063–1070

Fig. 5. Pro-secretory effects of STW 5 and its individual extracts in human intestinal preparations are significantly decreased by the adenylate cyclase blocker MDL 12,330A, the CFTR blocker CFTRinh -172 or the CaCl channel inhibitor SITS. The liquorice induced decrease in Isc in the small intestine is significantly reduced by the epithelial sodium channel blocker amiloride. Please note that the STW 5 data were taken from our previous study (Krueger et al. 2009). APL is a combined application of angelica (A), peppermint (P) and lemon balm (L). Numbers in parentheses indicate the number of tissues (equal to number of patients) studied. Extracts and blockers were applied in separate tissues from the same patient. ∗ marks significant differences from controls. Data illustrated as means ± SEM.

Fig. 6. Pro-secretory effects of STW 5 and its individual extracts in T84 cells are significantly decreased by the adenylate cyclase blocker MDL 12,330A, the CFTR blocker CFTRinh -172, the CaCl channel inhibitor SITS (all panel A) or the combination of CFTRinh -172 and SITS (panel B). Please note that the STW 5 data were taken from our previous study (Krueger et al. 2009). APLC is a combined application of angelica (A), peppermint (P), lemon balm (L) and chamomile (C). At least n = 6 filters for each data point. Extracts and blockers were applied in separate filters from the same batch. ∗ marks significant differences from controls. Data illustrated as means ± SEM.

T84 cells and supported the involvement of cAMP-dependent CFTR chloride channels in chloride secretion (Fig. 6). The involvement of CFTR chloride channels was further confirmed by apical application of CFTRinh -172 (20 μM) in human intestine and T84 cells. Similar to MDL-12,330A, the pro-secretory responses were significantly reduced in both human preparations and T84 cells (Figs. 5 and 6A, B). In both, human intestine and in T84 cells, none of the two inhibitors of CFTR chloride channels blocked the response. This observation suggested the involvement of additional anion channels. We

therefore tested the CaCl antagonist SITS (1 mM) which had an inhibitory effect on STW 5 evoked secretion (Krueger et al. 2009). The results showed that SITS significantly reduced the Isc increased in human tissue as well as T84 cells (Figs. 5 and 6A, B). Combined application of CFTRinh -172 and SITS in T84 cells significantly reduced but did not abolish the pro-secretory responses (Fig. 6B). The liquorice-induced decrease in Isc in small intestinal preparations was reduced in the presence of the epithelial sodium channel blocker amiloride (10 μM) by 72% (P = 0.01) (Fig. 5).

S. Allam et al. / Phytomedicine 22 (2015) 1063–1070

Discussion We provide the first systematic study on the influence of the nine herbal extracts present in STW 5 on epithelial secretion in human small and large intestinal preparations as well as in T84 cells. Our study showed that the pro-secretory action of STW 5 was mainly due to the angelica root extract. Among all extracts, only angelica evoked an increased secretion when used at concentrations present in 512 μg/ml STW 5, a dose that was about twice the EC50 value for its prosecretory action (Krueger et al. 2009). Only at higher concentrations, which however are still subtherapeutic, peppermint and lemon balm also had pro-secretory actions. Subtherapeutic in this context means that even the highest concentration of STW 5 or its components used in our in vitro study was 1/10 of what is the recommended single dose of 20 drops of Iberogast® which contained 51.3 mg STW 5. Admittedly, it is difficult to discuss drug concentrations in vitro versus in vivo as it remains unknown which concentrations of STW 5 reach the intestine and in particular which concentrations exist after adsorption in the gut wall. The systemic availability defined as sufficient plasma levels of the individual components may not be relevant as it is rather their local concentrations in the gut wall after adsorption which determines their action profile. We recently discussed the importance of this difference with regard to the mode of action of the muscarinic antagonist butylscopolamine (buscupan® ) (Krueger et al. 2013). Although its plasma level is very low it has potent antisecretory and spasmolytic actions very likely because it reaches much higher concentration locally in the gut wall. The pro-secretory action of angelica, peppermint and lemon balm extracts was comparable in both small and large intestine. Their effects involved activation of cAMP- and calcium-activated chloride channels based on the antagonistic actions of MDL- 12,330A, CFTRinh 172 and SITS. At least for angelica these effects were mediated by nerves. This finding agreed with our previous study in which we showed that the pro-secretory effect of 512 μg/ml STW 5 was reduced by tetrodotoxin (Krueger et al. 2009). Moreover, we demonstrated in the same study that STW 5 directly activated enteric neurons. Results of the present study suggested that mainly angelica is responsible for the nerve mediated pro-secretory action of STW 5. Its direct effect on epithelial cells may be relatively low as suggested by the rather small increase of secretion in T84 cells. Only in small intestine, liquorice at higher concentrations had an anti secretory effect. Without doubt it would be interesting and very challenging to identify the phytochemical compound(s) responsible for the actions of the extracts. The nine herbal extracts contain more than 300 known phytochemical compounds (Wegener and Wagner 2006). It is impracticable to study all of them in human preparations within an acceptable time period. What seems feasible and immediately relevant is to initiate further studies with those extracts that have pro-or anti-secretory actions. The generation of hydroethanolic extracts was described previously in detail (Kroll and Cordes 2006). By showing comparable actions of STW 5 and sSTW 5, we confirmed that the activity of STW 5 and its components are not affected by the production processes. Using aqueous solutions of lyophilized hydroethanolic extracts had the advantage to prevent unspecific in vitro effects of ethanol on secretion and epithelial barrier (Sommansson et al. 2014). Only peppermint oil has been previously studied for its action on epithelial transport in rat jejunum. This study did not find effects of serosal peppermint oil at a concentration of 1 mg/ml on (Beesley et al. 1996). Peppermint oil is in no way comparable in its chemical composition to the hydroethanolic peppermint extract tested here, which may be an explanation for this discrepant finding, next to species differences. The inhibitory action of peppermint oil on acetylcholine induced secretion may be also due to the high concentration used by Beesley and colleagues (Beesley et al. 1996).

1069

Our data indicated that angelica alone mimicked the secretagogue activity of STW 5. This makes in particular angelica an interesting candidate to isolate the active chemicals. Osthole and furocumarins are among the main active constituents in angelica archangelica (Kroll and Cordes 2006). Osthole is a very potent activator for chloride secretion in rat colonic mucosa through direct activation of CFTR channels (Yang et al. 2011). Furocoumarins act as acetylcholine esterase inhibitor and would thereby increase the endogeneous concentration of the pro-secretory transmitter acetylcholine which might eventually cause activation of CaCl channels (Sigurdsson and Gudbjarnason 2007). The clinical relevance remains to be shown but it is interesting that a dietary supplement referred to “IBS active” that contained among probiotic Lactobacilla strains, L-tryptophan, inulin and vegetal charcoal also angelica improved constipation in IBS patients (Astegiano et al. 2006). At higher concentrations corresponding to 5120 μg/ml STW 5 peppermint also increased secretion. Menthol as one of the main active constituent in peppermint activated CFTR channels in human epithelial airways Calu-3 cells (Morise et al. 2010). A similar mechanism may also be operative in the intestinal epithelium. Enteric coated peppermint oil relieved IBS symptoms in several clinical trials (Rees et al. 1979; Cappello et al. 2007; Merat et al. 2010). The pro-secretory action of 512 μg/ml or 5120 μg/ml STW 5 was mimicked in human intestinal tissue by angelica or combined application of angelica, peppermint and lemon balm, respectively. This was plausible as angelica was the only pro-secretory component when applied at concentrations corresponding to 512 μg/ml STW 5. Increasing concentrations to levels present in 5120 μg/ml STW 5 revealed in addition pro-secretory actions of peppermint and lemon balm. Hence the combination of the three extracts mimic prosecretory actions of 5120 μg/ml STW 5. Our findings would suggest that the pro-secretory actions of angelica, peppermint or lemon balm were additive and statistically not different from the pro-secretory action of STW 5. This does not rule out that some other effects of STW 5 were based on synergistic actions of the individual components as suggested previously (Wegener and Wagner 2006). For example, the antioxidative properties of STW 5 were due to a supraadditive effect of the individual extracts (Germann et al. 2006). It also needs to be emphasized that the clinical efficacy of STW 5 is very likely related to its broad action profile, including but not limited to region-dependent spasmolytic or prokinetic actions on motility, antiinflammatory actions (see Wegener and Wagner 2006) as well as prosecretory actions (Krueger et al. 2009). The extracts contributed differently to these effects (Wagner 2006 and our present study). In principle, we observed the same results in T84 cells with some noteworthy differences. First, chamomile evoked a strong prosecretory action in T84 cells only. We have no final explanation for the different responses in T84 cells and tissue preparations. It might be related to T84 specific expression of targets. This is suggested by the finding that chamomile extract significantly decrease growth in cancer cell lines but not in normal cells (Srivastava and Gupta 2007). Second, the combined application of the pro-secretory components exceeded that evoked by STW 5 or the calculated values in T84 cells but not in tissue preparations. This may be due to synergistic activation of cAMP and Ca dependent anion secretion which appeared more pronounced in T84 cells (Barrett 1993). We have shown in the present study that the pro-secretory action of individual extracts depended on activation of cAMP and Ca dependent anion secretion similar to the mode of action of STW 5 (Krueger et al. 2009). Liquorice had anti-secretory action in the small intestinal preparations but had no effect on ISC in large intestinal preparations or in T84 cells. Based on the amiloride sensitivity of this effect we assume that liquorice would increase reabsorption of sodium. It has been shown that components in liquorice possessed aldosterone-like activity (Baron et al. 1969). If this is indeed the mode of action of liquorice extracts it is plausible that liquorice had no such effect in

1070

S. Allam et al. / Phytomedicine 22 (2015) 1063–1070

T84 cells as they lack glucocorticoid as well as mineralocorticoid receptors (Tsugita et al. 2009). However, it remains puzzling that we did not observe an anti-secretory action in colonic tissue preparations which normally expresses epithelial sodium channels. Bitter candytuft had no influence on ion secretion neither in human small or large intestine nor in T84 cells. Interestingly, a monotherapy with bitter candytuft showed no clinical benefit in IBS patients (Madisch et al. 2004b). In conclusion, our results suggest that the pro-secretory action of STW 5 is mainly due to angelica with a smaller contribution of peppermint and lemon balm. Their effects involve activation of cAMPand Ca++ -activated Cl− channels. STW 5 is approved to treat functional dyspepsia and IBS both of which are not considered secretory disorders but rather multifactorial diseases associated with enhanced visceral pain, immune imbalance, altered enteric nerve sensitivity and motility disorders (Andresen et al. 2011). Our findings suggest that peppermint, lemon balm and in particular angelica may be the basis to develop novel herbal preparations to specifically treat secretory disorder based on impaired epithelial secretion, such as constipation. Of course it would be in a next step most desirable to isolate defined molecules in these extracts which may be responsible for the pro-secretory actions. Conflict of interest This study was in part funded by a research grant from Steigerwald Arzeneimittel GmbH. The authors declare that they had full control over the data and full access to the data. Steigerwald Arzneimittel GmbH did not influence data collection, analysis or interpretation. Acknowledgments Part of this study was funded by a grant from Steigerwald Arzneimittelwerk GmbH. The company had no influence and never tried to influence data acquisition, analysis or interpretation. References Andresen, V., Camilleri, M., Busciglio, I.A., Grudell, A., Burton, D., McKinzie, S., FoxxOrenstein, A., Kurtz, C.B., Sharma, V., Johnston, J.M., Currie, M.G., Zinsmeister, A.R., 2007. Effect of 5 days linaclotide on transit and bowel function in females with constipation-predominant irritable bowel syndrome. Gastroenterology 133, 761– 768. Andresen, V., Keller, J., Pehl, C., Schemann, M., Preiss, J., Layer, P., 2011. Irritable bowel syndrome–the main recommendations. Dtsch. Arztebl. Int. 108, 751–760. Astegiano, M., Pellicano, R., Terzi, E., Simondi, D., Rizzetto, M., 2006. Treatment of irritable bowel syndrome. A case control experience. Minerva Gastroenterol. Dietol. 52, 359–363. Baron, J.H., Nabarro, J.D., Slater, J.D., Tuffley, R., 1969. Metabolic studies, aldosterone secretion rate, and plasma renin after carbenoxolone sodium. Br. Med. J. 2, 793– 795. Barrett, K.E., 1993. Positive and negative regulation of chloride secretion in T84 cells. Am. J. Physiol. 265, C859–C868. Beesley, A., Hardcastle, J., Hardcastle, P.T., Taylor, C.J., 1996. Influence of peppermint oil on absorptive and secretory processes in rat small intestine. Gut 39, 214–219.

Brierley, S.M., Kelber, O., 2011. Use of natural products in gastrointestinal therapies. Curr. Opin. Pharmacol. 11, 604–611. Cappello, G., Spezzaferro, M., Grossi, L., Manzoli, L., Marzio, L., 2007. Peppermint oil (Mintoil) in the treatment of irritable bowel syndrome: a prospective double blind placebo-controlled randomized trial. Dig. Liver Dis. 39, 530–536. Germann, I, Hagelauer, D, Kelber, O, Vinson, B, Laufer, S, Weiser, D, Heinle, H, 2006. Antioxidative properties of the gastrointestinal phytopharmaceutical remedy STW 5 (Iberogast). (2006). Phytomedicine 13 (Suppl 5), 45–50. Johanson, J.F., Drossman, D.A., Panas, R., Wahle, A., Ueno, R., 2008. Clinical trial: phase 2 study of lubiprostone for irritable bowel syndrome with constipation. Aliment. Pharmacol. Ther. 27, 685–696. Kroll, U., Cordes, C., 2006. Pharmaceutical prerequisites for a multi-target therapy. Phytomedicine 13 (Suppl 5), 12–19. Krueger, D., Gruber, L., Buhner, S., Zeller, F., Langer, R., Seidl, S., Michel, K., Schemann, M., 2009. The multi-herbal drug STW 5 (Iberogast(R)) has prosecretory action in the human intestine. Neurogastroenterol. Motil. 21 1203-e110. Krueger, D., Michel, K., Allam, S., Weiser, T., Demir, I.E., Ceyhan, G.O., Zeller, F., Schemann, M., 2013. Effect of hyoscine butylbromide (Buscopan®) on cholinergic pathways in the human intestine. Neurogastroenterol. Motil. 25, e530–e539. Madisch, A., Holtmann, G., Mayr, G., Vinson, B., Hotz, J., 2004. Treatment of functional dyspepsia with a herbal preparation. A double-blind, randomized, placebocontrolled, multicenter trial. Digestion 69, 45–52. Madisch, A., Holtmann, G., Plein, K., Hotz, J., 2004. Treatment of irritable bowel syndrome with herbal preparations: results of a double-blind, randomized, placebocontrolled, multi-centre trial. Aliment. Pharmacol. Ther. 19, 271–279. Madisch, A., Melderis, H., Mayr, G., Sassin, I., Hotz, J., 2001. [A plant extract and its modified preparation in functional dyspepsia. Results of a double-blind placebo controlled comparative study]. Z. Gastroenterol. 39, 511–517. Melzer, J., Rosch, W., Reichling, J., Brignoli, R., Saller, R., 2004. Meta-analysis: phytotherapy of functional dyspepsia with the herbal drug preparation STW 5 (Iberogast). Aliment. Pharmacol. Ther. 20, 1279–1287. Merat, S., Khalili, S., Mostajabi, P., Ghorbani, A., Ansari, R., Malekzadeh, R., 2010. The effect of enteric-coated, delayed-release peppermint oil on irritable bowel syndrome. Dig. Dis. Sci. 55, 1385–1390. Morise, M., Ito, Y., Matsuno, T., Hibino, Y., Mizutani, T., Ito, S., Hashimoto, N., Kondo, M., Imaizumi, K., Hasegawa, Y., 2010. Heterologous regulation of anion transporters by menthol in human airway epithelial cells. Eur. J. Pharmacol. 635, 204–211. Rahimi, R., Abdollahi, M., 2012. Herbal medicines for the management of irritable bowel syndrome: a comprehensive review. World J. Gastroenterol. 18, 589–600. Rees, W.D., Evans, B.K., Rhodes, J., 1979. Treating irritable bowel syndrome with peppermint oil. Br. Med. J. 2, 835–836. Sigurdsson, S., Gudbjarnason, S., 2007. Inhibition of acetylcholinesterase by extracts and constituents from Angelica archangelica and Geranium sylvaticum. Z. Naturforsch. C. 62, 689–693. Sommansson, A., Wan Saudi, W.S., Nylander, O., Sjoblom, M., 2014. The ethanolinduced stimulation of rat duodenal mucosal bicarbonate secretion in vivo is critically dependent on luminal cl-. PLoS. ONE. 9, e102654. Srivastava, J.K., Gupta, S., 2007. Antiproliferative and apoptotic effects of chamomile extract in various human cancer cells. J. Agric. Food Chem. 55, 9470–9478. Thomas, R.H., Allmond, K., 2013. Linaclotide (Linzess) for irritable bowel syndrome with constipation and for chronic idiopathic constipation. Pharm. Ther. 38, 154– 160. Tsugita, M., Iwasaki, Y., Nishiyama, M., Taguchi, T., Shinahara, M., Taniguchi, Y., Kambayashi, M., Nishiyama, A., Gomez-Sanchez, C.E., Terada, Y., Hashimoto, K., 2009. Glucocorticoid receptor plays an indispensable role in mineralocorticoid receptor-dependent transcription in GR-deficient BE(2)C and T84 cells in vitro. Mol. Cell Endocrinol. 302, 18–25. Wagner, H., 2006. Multitarget therapy–the future of treatment for more than just functional dyspepsia. Phytomedicine. 13 (Suppl 5), 122–129. Wegener, T., Wagner, H., 2006. The active components and the pharmacological multitarget principle of STW 5 (Iberogast((R))). Phytomedicine 13 (Suppl 5), 20–35. Yang, H., Xu, L.N., Sui, Y.J., Liu, X., He, C.Y., Fang, R.Y., Liu, J., Hao, F., Ma, T.H., 2011. Stimulation of airway and intestinal mucosal secretion by natural coumarin CFTR activators. Front. Pharmacol. 2, 52.