Basal and Spasmolytic Effects of a Hydroethanolic Leaf Extract of Melissa officinalis L. on Intestinal Motility: An Ex Vivo Study

Abstract Melissa officinalis L. (lemon balm) has been used for decades with symptomatic benefits in patients with digestive disorders. However, very little is known on the effects of M. officinalis on the gastrointestinal (GI) tract. In this study, the basal and spasmolytic properties of a hydroethanolic leaf extract (HLE) of M. officinalis were assessed ex vivo on different segments of the GI tract of mice after phytochemical characterization of the extract. M. officinalis HLE had site- and dose-dependent effects on the contractile activity of the GI tract, the motility response being impacted in the jejunum and ileum but not in the antrum and colon. The observed effects could be caused by the phenolic compounds (mainly rosmarinic acid) detected in the extract.


INTRODUCTION
M elissa officinalis L., also known as lemon balm, is a medicinal plant that has long been used in traditional medicine for digestive comfort and its impact upon mood and cognitive performance. [1][2][3] As for gastrointestinal (GI) function, the European Medicines Agency acknowledged M. officinalis folium as a traditional herbal medicinal product for the symptomatic treatment of mild GI complaints including bloating and flatulence, 4 and the European Scientific Cooperative On Phytotherapy, for the symptomatic treatment of digestive disorders such as minor spasms. 5 Although M. officinalis has been used for decades with symptomatic benefits in patients with digestive disorders, to date, little is known on the effects of M. officinalis on the GI tract. The pathophysiological concepts concerning functional disorders of the GI tract include disturbances in motility and acid production as underlying mechanisms.
Results of the few studies that have been performed to characterize the effects of M. officinalis, mainly on motility, are scarce and often contradictory. Most of the studies available were performed with M. officinalis folium extract that is a constituent of a marketed fixed combination of nine herbal extracts. [6][7][8][9][10] The observations made in these studies were dependent on the gut part tested and were in part species dependent. For instance, Forster et al. did not show spasmolytic activity of M. officinalis in the guinea pig ileum while Heinle et al. observed a reduction in histamineinduced contractile response. 6,7 Furthermore, under basal conditions, Schemann et al. reported increased motility in proximal and distal stomach while Sibaev et al. observed lemon balm leaf had no effects in basic electrophysiological properties of smooth muscle cells in the ileum. 8,9 Altogether, no comprehensive studies on the effects of M. officinalis hydroethanolic leaf extracts (HLEs) upon basal contractile activity and spasmolytic activity in various gut segments have been performed. Therefore, the basal and spasmolytic properties of HLE of M. officinalis were assessed ex vivo on different segments of the GI tract of mice after phytochemical characterization of the extract.

MATERIALS AND METHODS
M. officinalis leaves were collected in Aubiat (France) in June 2015 and identified by Gilles Thébault from the herbarium of Museum d'Histoire Naturelle Henri-Lecoq (Clermont-Ferrand, France) where a voucher specimen was deposited (CLF106452). This herbarium is referenced with the international association for plant taxonomy whose head office is in the New York botanical garden.
Preparation of M. officinalis HLE M. officinalis HLE was produced by PiLeJe Industrie (France) according to the patented process WO2001056 584A1. The procedure was standardized and the extract quality was controlled by thin layer chromatography, and by high-performance liquid chromatography (HPLC) for determination of rosmarinic acid content as described in the European Pharmacopeia monograph (ref 01/2011:1447). M. officinalis fresh frozen leaves were extracted with 20% to 70% (v/v) ethanolic leaching. The extract was concentrated under reduced pressure to evaporate ethanol and then ethanol concentration was adjusted to 30%. M. officinalis HLE in ethanol was used in ex vivo experiments. The specific batch used in the study (no. C-14J297) contained 8.6% of dry material that held 5.3% of rosmarinic acid. The yield of extraction as a percentage weight of the dry matter of starting fresh plant material was 24%. After addition of glycerol, the M. officinalis HLE is a rosmarinic acid standardized extract of M. officinalis (EPS Mélisse; PiLeJe Laboratoire, France).

Phytochemical analysis of M. officinalis HLE
Chemical composition of M. officinalis HLE was further analyzed using ultra-HPLC (UHPLC). Chromatographic analyses by UHPLC were performed on an Ultimate 3000 RSLC UHPLC system (Thermo Fisher Scientific, Inc., MA, USA) coupled to a quaternary rapid separation pump (Ultimate autosampler) and a rapid separation diode array detector. Compounds were separated on a Uptisphere Strategy C18 column (250 · 4.6 mm, 5 lm; Interchim), which was controlled at 30°C. The mobile phase was a mixture of 0.1% (v/v) formic acid in water (phase A) and 0.1% (v/v) formic acid in acetonitrile (phase B). The gradient of phase A was 100% (0-39 min), 0% (40-49 min), 100% (50-60 min), and then was held at 100%. The flow rate was 0.8 mL/min, and the injection volume was 5 lL. The UHPLC system was connected to an Orbitrap (Thermo Fisher Scientific, Inc.) mass spectrometer, operated in negative electrospray ionization mode. Source operating conditions were as follows: 3 kV spray voltage; 320°C heated capillary temperature; 400°C auxiliary gas temperature; sheath, sweep, and auxiliary gas (nitrogen) flow rate of 50, 10 and 2 arbitrary units, respectively; and collision cell voltage between 10 and 50 eV. Full scan data were obtained at a resolution of 70,000, whereas tandem mass spectrometry (MS 2 ) data were obtained at a resolution of 17,500. Data were processed using Xcalibur software (Thermo Fisher Scientific, Inc.).
Compounds detected were characterized according to their retention times, mass spectral data, and comparison with authentic standards when available or with data found in the literature. 3

Effects of M. officinalis HLE on GI motility
Male C57BL/6J Rj mice aged 8 weeks (n = 40; Janvier Laboratory, Le Genest-Saint-Isle, France) were maintained on a 12 h light-12 h dark cycle (22°C) and had free access to food and water. After 1-week adaptation period, mice were killed by cervical dislocation. Whole stomach and intestine of each mouse were quickly excised and placed in an icecold Hank's buffer saline solution. Circular muscle strips (10 mm in length and 0.2 mm in width) were cut from the distal antrum region and longitudinal strips from jejunum, ileum, and proximal colon. Muscle preparations were suspended vertically in an organ bath filled with Krebs solution (NaCl, 117 mM; KCl, 4.7 mM; MgCl 2 , 1.2 mM; NaH 2 PO 4 , 1.2 mM; NaHCO 3 , 25 mM; CaCl 2 , 2.5 mM; and glucose, 11 mM) warmed at 37°C and gassed with 95% O 2 +5% CO 2 . Muscle strips were stretched with a tension of 4-6 mN. Isometric contractions were continuously recorded by using isometric force transducers (No. TRI202PAD; Panlab, Cornellã, Spain) coupled to a computer equipped with the Pow-erLab 8/30 System and the Labchart data analysis software (AD Instruments, Spechbach, Germany).
Basal effects were assessed after a 30-min equilibration period. Muscle strips (n = 10 per condition) were treated with increasing concentrations of M. officinalis HLE (1, 5, 10, 25, and 50 mg/mL corresponding, respectively, to 0.077, 0.39, 0.77, 1.9, and 3.87 mg/mL of dry material and 0.004, 0.02, 0.04, 0.1, and 0.2 mg/mL of rosmarinic acid) or ethanol (at corresponding concentrations, i.e., 0.3, 1.5, 3, 7.5, and 15 mg/mL). Muscle strips were then washed in Krebs solution for 15 min and treated with nifedipine 10 lM (myorelaxant positive control) after a 15-min equilibration period. The protocol for the assessment of spasmolytic effects was identical except steady-state muscle strips were precontracted with the muscarinergic agonist bethanechol (100 lM) before treatment with M. officinalis HLE or ethanol then, after washing with Krebs solution, treated with bethanechol and nifedipine for control.
Frequency of spontaneous phasic contractions was evaluated for 2 min and contractile activity was assessed by measuring the area under the curve (AUC) for 2 min. These two parameters were assessed before and after addition of M. officinalis HLE or ethanol. Values are expressed as meanstandard error of the mean. Graphs and statistical analysis were performed using GraphPad Prism 5.00 (GraphPad Software, Inc., La Jolla, CA, USA). Friedman's test followed by Dunn's test was used to analyze the dose-response effect of M. officinalis HLE or ethanol. Treatment effects were compared using two-way analysis of variance (ANOVA) for repeated measures followed by Bonferroni post hoc test. Nifedipine effects were analyzed with Wilcoxon test. Results were considered statistically significant when P < .05. All values were normalized to tissue weight and expressed in 654 AUBERT ET AL.
percentage compared with initial value before treatment with M. officinalis HLE or ethanol.

Effects on GI tract motility
Antrum. Both in basal and precontracted conditions, effects of M. officinalis HLE on antrum were not significantly different from those of ethanol (vehicle; Table 2).
In the precontracted antrum, there was no significant change in frequency of phasic contractions with M. officinalis HLE whatever the concentration (Friedman's test: P = .12 [n = 10]). Although ethanol induced a significant decrease in that frequency at corresponding concentration of 50 mg/mL (Friedman's test: P = .0002 [n = 10]; Dunn's test: P < .05 vs. control [n = 10]), the difference between two treatments was not significant (two-way ANOVA, P = .83; M. officinalis HLE n = 10; ethanol n = 10). Similar observations were made for the contractile activity (AUC): no effect of M. officinalis, significant effect of ethanol at corresponding concentrations of 25 and 50 mg/mL, and no difference between the two treatments.
Nifedipine significantly inhibited phasic contractions and AUCs in all conditions (for antrum and all GI parts).

DISCUSSION
Our results show that the HLE of M. officinalis tested has site-and dose-dependent effects on the contractile activity of the GI tract that do not depend on its vehicle (ethanol). In particular, M. officinalis HLE affected motility response in jejunum and ileum, whereas it had no effect on antrum and colon.
As for the small intestine, M. officinalis HLE inhibited AUC and contractile frequency at low concentrations for which ethanol had no effect. In addition, at higher concentration, the amplitude of effects was also significantly higher than that observed with ethanol (when it had an effect).
Whether these changes observed ex vivo impact intestinal transit remains to be assessed in vivo. However, this decrease in contractile activity observed both under basal condition and after precontraction could lead to slower intestinal transit. Conversely, spasmolytic activity could contribute to reduce intestinal sensitivity. Consistent with our findings, inhibitory effects of M. officinalis on ileum have been reported in several studies. 6,7,11 In brief, an ethanol extract of M. officinalis folium inhibited histamine-induced contractions of guinea pig ileum, whereas an aqueous extract was inactive. 11 Interestingly, a hydroethanolic extract (30% ethanol dose unknown) also decreased histamine-induced contractions and amplitude of spontaneous contractions in the guinea pig ileum in comparison with solvent control. 7 However, this extract had no effect on amplitude and frequency of slow waves in circular smooth muscle of mouse small intestine, suggesting other mechanisms of action underlying functional effects of the extract (such as altering intracellular calcium homeostasis). 10 In addition, no antispasmodic activity was observed for a hydroethanolic extract (30% ethanol; 2.5 and 10 mL/L) in the guinea pig ileum. 6 Finally, our study also extended spasmolytic activity of M. officinalis HLE to the jejunum that seems to be more sensitive to the extract both under basal and precontracted conditions. Of interest, M. officinalis HLE had no effects on antrum and colon contractile activity but seemed to antagonize the effects of ethanol. As several studies showed that gastric emptying was slowed by ethanol (e.g., Jian et al. 12 ), our results suggest that M. officinalis could prevent ethanolinduced slowing of gastric emptying. In contrast to our findings, a M. officinalis extract used as an ethanol-free lyophilisate induced a small but significant contractile response in the antrum. 8 In our study, M. officinalis HLE had no significant effects on proximal colon motility; these results are similar to those of Sibaev et al. who reported that an ethanolic extract of lemon balm had no effect on amplitude and frequency of slow waves in the colon. 9 The underlying mechanisms of the organ-specific effects of M. officinalis reported in our study remains currently unknown. It could however reflect differences in properties of smooth muscle cells and/or interstitial cells of Cajal between different gut segments. These differences are probably not because of region-specific differences in the neurochemical phenotype of the enteric nervous system as previous studies have shown that SWT 5 (that contains among others M. officinalis) effects were not modified by synaptic transmission blockers. 13 Main compounds found in M. officinalis HLE were rosmarinic acid, and other phenolic compounds, as previously reported in literature. 3,14,15 Our phytochemical analysis also revealed hydroxyl cinnamic acids such as caffeic acid and derivatives, and of several compounds containing carboxylic acid function including quinic, citric, and salicylic acids. Luteolin 3 0 -O-b-d-glucuronide was also found; this substance was previously reported as the major flavone present in M. officinalis. 14,15 Which individual component or group of components of M. officinalis influences GI activity and is responsible for which specific effect will have to be answered in future investigations. However, effects of M. officinalis are mainly attributed to phenolic compounds and its essential oil. 3 Studies performed with the essential oil of M. officinalis showed a myorelaxant effect on different parts of the small intestine in various species. [16][17][18][19] However, as shown in our phytochemical analysis our extract does not contain essential oil (no terpenes were detected). Therefore, effects observed cannot be attributed to essential oil, but are likely because of rosmarinic acid and/or other phenolic compounds. An aqueous extract of M. officinalis was reported to have a vasorelaxant effect on the rat isolated thoracic aorta. 20 Rosmarinic acid was the most abundant compound in this extract, and the authors confirmed that the compound exerted a vasorelaxant effect. Rosmarinic acid was also found to have spasmolytic effects on rat uterus. 21 However, rosmarinic acid (from thyme) had no antispasmodic activity in a preconstricted rat smooth muscle trachea model; the effect observed was attributed to luteolin. 22 Therefore, luteolin 3 0 -O-b-d-glucuronide present in our extract could be, at least partly, at the origin of observed effects.
Overall, the effects of M. officinalis HLE tested on gut motility are site and dose dependent. Whether the effects observed ex vivo have an impact on intestinal transit will have to be assessed in vivo. These effects could be because of an action of phenolic compounds but further experiments M. officinalis HLE Ethanol FIG. 4. Effects of M. officinalis HLE on the ileum. Friedman's test followed by Dunn's test were used to analyze the dose-response effect of M. officinalis HLE or ethanol and values that are significantly different from control (i.e., previous treatment) are indicated respectively by # and F. Treatment effects were compared by using twoway ANOVA for repeated measures followed by Bonferroni post hoc test and values that are significantly different are indicated by *.
will have to confirm this hypothesis. Benefits of M. officinalis could be related to site-dependent spasmolytic actions, besides its previously described anti-inflammatory, antinociceptive, and prosecretory action. 3,23,24 ACKNOWLEDGMENTS Part of the animal experiments was carried out within the small animal exploration facility Therassay (Nantes), which is supported by the GIS-IBiSA (Groupement d'Intérêt Scientifique-Infrastructure en Biologie Santé et Agronomie) program. The research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sector.
AUTHOR DISCLOSURE STATEMENT I.G. and C.B. are, respectively, research project manager and medical writer at PiLeJe Laboratoire. Valérie Bardot (PiLeJe Industrie), I.R. and P.C. were involved in the chromatographic analyses. P.A. and M.N. performed the animal experiments for PiLeJe Laboratoire.