Antinociceptive and anxiolytic-like effects of a neo-clerodane diterpene from Salvia semiatrata aerial parts

Abstract Context Salvia semiatrata Zucc. (Lamiaceae) is a species used as a tranquilizer and to relieve pain in folk medicine in Santiago Huauclilla, Oaxaca, Mexico. Objective To evaluate the antinociceptive and anxiolytic-like effects of S. semiatrata extracts and identify a bioactive metabolite. Materials and methods The extracts were obtained by maceration of S. semiatrata aerial parts using solvents in increasing polarity (hexane, ethyl acetate and methanol). A neo-clerodane diterpene was extracted from the ethyl acetate fraction using open column chromatography. Identification of this metabolite was performed by crystallography, 1H NMR, 13C NMR, ATR-IR, ECD, MS and elemental analysis. The antinociceptive activity was explored using the writhing and formalin tests. Whereas, the anxiolytic-like responses were analysed in the open-field, hole-board and plus-maze tests. All the treatments were administered using oral gavage in male CD1 mice and explored 30 min after administration of the individual extracts (300 mg/kg) or the compound 1 (0.1, 1 or 10 mg/kg). Results All the extracts produced significant reduction in the nociceptive and anxiety-like behaviour compared to mice treated with the vehicle (0.5% tween 80 in saline solution). The spectroscopic analysis corroborated the presence of the neo-clerodane diterpene 7-keto-neoclerodan-3,13-dien-18,19:15,16-diolide (1), as partial responsible of the antinociceptive and anxiolytic-like effects, which produced a dose-dependent response in the writhing test with an ED50=4.15 mg/kg. Discussion and conclusions: These results reinforce the medicinal properties of S. semiatrata in folk medicine, where participation of a neo-clerodane diterpene was evidenced in the inhibitory central nervous system activity of this species.


Introduction
Salvia is the most diverse genus of the Lamiaceae family. In Mexico, it is represented by 306 species, of which 236 (77%) are endemic and most of them belong to the subgenus Calosphace (Mart ınez- Gordillo et al. 2017). From the past to the present time, sages have been used for their wide spectrum of medicinal properties in folk Mexican medicine. It is known that species, such as Salvia coccinea Juss. ex Murray, Salvia cinnabarina M. Martens & Galeotti, Salvia fulgens Cav., Salvia microphylla Kunth, Salvia purpurea Cav., Salvia lavanduloides Kunth, Salvia hispanica L., Salvia rubiginosa Benth., Salvia urica Epling and Salvia elegans Vahl are used to treat gastrointestinal alterations; for example, dysentery, diarrhoea, stomachache, inflammation and cramping (Dom ınguez-V azquez and Castro-Ram ırez 2002; Jenks and Seung-Chul 2013). While, S. coccinea, Salvia divinorum Epling & J ativa, S. elegans, S. fulgens, S. lavanduloides, Salvia leucantha Cav., S. microphylla and Salvia polystachya Ortega have been reported for their tranquilizing effects and their use to treat insomnia (Jenks and Seung-Chul 2013;Casselman et al. 2014).
Preclinical studies of extracts and compounds (ursolic acid, b-sitosterol, cirsiliol and apigenin) from Salvia species have demonstrated several pharmacological effects at level of the central nervous system (CNS), such as anxiolytic, sedative, antidepressant and antinociceptive, among others (Imanshahidi and Hosseinzadeh 2006). Anti-inflammatory response has been included in the antinociceptive effects explored in Salvia plebeia R. Br. (Jung et al. 2009), Salvia officinalis L. (Rodrigues et al. 2012) and Salvia circinata Cav. (Moreno-P erez et al. 2019). The antinociceptive effects of S. officinalis (Ghorbani and Esmaelizadeh 2017) and S. circinata (Moreno-P erez et al. 2019) have been associated with the presence of diterpenes, triterpenes and flavonoids. In addition, the hydroalcoholic extract of Salvia reuterana Boiss. and the essential oil of Salvia miltiorrhiza Bunge produced sedative and anxiolytic-like effects (Rabbani et al. 2005;Liu et al. 2015), whereas antidepressant and anxiolytic-like activities were informed by S. elegans (Herrera-Ruiz et al. 2006).
Salvia semiatrata Zucc. is a perennial shrub, 1-2 m tall, small leaves and purple inflorescences. This species is known in Santiago Huaclilla, Oaxaca, with the common name of 'mirto morado' (purple myrtle). It is distributed in the dry areas of the state of Oaxaca and is used by residents, as a healing, anti-inflammatory, for earache, stomach-ache and for nervous disorders (Nambo 2015). From the medium polar extract of the S. semiatrata aerial parts, two neoclerodane diterpenoids have been isolated and characterized, one of them was semiatrin and the other was 7a-hydroxy-neo-clerodan-3, 13-dien-18,19:15,16-diolide (Esquivel et al. 1986;Soriano-Garc ıa et al. 1987). From the root extract, the diterpenoids tilifolidione, horminone and 20-norinuroyleanol were also identified (Esquivel et al. 2005). Regarding biological activities, antifeedant effects of tilifolidione and horminone have been reported (Simmonds et al. 1996), as well as cytotoxic for tilifolidione (Esquivel et al. 2005) and antifungal and antibacterial for horminone (Ulubelen et al. 2001;Bufalo et al. 2016).
However, the pharmacological effects on the CNS of S. semiatrata or its metabolites are not found in the literature. The purpose of this research was to evaluate the antinociceptive and anxiolytic-like activities of several organic crude extracts of the S. semiatrata aerial parts and a bioactive metabolite (1) using several experimental models.

Drugs and reagents
Ketorolac was purchased from SupraDOL V R , Liomont (Mexico City, Mexico); Clonazepam was purchased from Gabaclotec V R , Tecnofarma (Mexico City, Mexico); Diclofenac, acetic acid and 37% formalin were purchased from Merck (Mexico City, Mexico). Tween 80 and ceric sulphate were purchased from Sigma (St. Louis, MO). Hexane, ethyl acetate and methanol were purchased from Tecsiquim, S.A. de C.V. (Mexico City, Mexico). Drugs and methanol extract were dissolved in saline solution (0.9% NaCl, Sigma, St. Louis, MO). The hexane and ethyl acetate extracts, as well as the purified compound 1 were resuspended in Tween 80 at 0.5% in saline solution (vehicle). All treatments were freshly prepared and administered by oral gavage (p.o.). Nociceptive agents, 1% acetic acid and 1% formalin were diluted in saline solution and injected by intraperitoneal (i.p., 10 mL/kg) or intraplantar (20 mL/paw) route, respectively.

Vegetal material
S. semiatrata aerial parts were collected in Santiago Huauclilla, Oaxaca, in July 2017. This region is located at the parallels 17 25 0 and 17 34 0 latitude north and meridians 96 56 0 and 97 08 0 longitude west, and at altitude between 1200 and 2700 m. A voucher specimen (no. 16362) was identified by Dra. Martha J. Mart ınez Gordillo and deposited in the IMSS Herbarium of CDMX, Mexico.

Preparation of the extracts
Three organic extracts (hexane, ethyl acetate and methanol) of S. semiatrata aerial parts were obtained by a maceration process using 1466 g of the dry and ground plant material. For this, powder was kept in a container immersed in each solvent for 48 h at room temperature (22 C). Each filtrate was concentrated under vacuum to eliminate solvents by evaporation in a rotary evaporator RII (B€ uchi Labortechnik AG, Flawil, Switzerland). A final yield of the extracts was obtained as follows: hexane (1.11%), ethyl acetate (5.52%) and methanol (2.13%) (Figure 1).
Extraction and purification of a neo-clerodane diterpene (1) The extraction of the active metabolite was carried out by fractionation of the ethyl acetate extract using conventional opencolumn chromatography packed with silica gel 60 GF 254 (Macherey-Nagel, D€ uren, Germany) ( Figure 1). The elution system consisted of 100% hexane, followed by hexane-ethyl acetate mixtures of increasing polarity until eluting with ethyl acetate (100%). Finally, it was eluted with 100% methanol. A total of 214 fractions of 100 mL each were collected, which were concentrated in a rotary evaporator for later analysis by thin layer chromatography. The plates were revealed with ceric sulphate to gather those with similar profile obtaining 42 subfractions.

Crystallographic investigations
Crystallographic data for 1 were collected on a Bruker APEX II CCD Diffractometer (Billerica, MA) at 150 K using Cu-Ka radiation (k ¼ 1.54178 Å) from an Incoatec ImS source and Helios optic monochromator. Suitable crystals were coated with hydrocarbon oil (Parabar), picked up with a nylon loop, and mounted in the cold nitrogen stream of the diffractometer. The structure was solved using intrinsic phasing (SHELXT) (Sheldrick 2015) and refined by full-matrix least-squares on F2 using the shelXle GUI (H€ ubschle et al. 2011). The hydrogen atoms of the C-H bonds were placed in idealized positions. Crystallographic data were deposited at the Cambridge Crystallographic Data Center (CCDC) with the number 1953351.

Computational details of ECD
3D model for compound 1 was built and geometry optimized using Spartan'10 with a PM3 semiempirical method. Conformational analysis was performed via Monte Carlo search protocol under the same force field. The conformers were checked for redundancy. All conformers within 10 kcal/mol were minimized using DFT at the B3LYP/DGDZVP level of theory employing Gaussian 09 (Gaussian, Inc., Wallingford, CT). The conformers were optimized, and the IR, vibrational analyses and thermochemical properties were obtained. The TD-SCF with default solvent model was used to perform the ECD calculations of the major conformers in MeOH solution, using a B3LYP-6-31 þ G(d) force field. The calculated rotatory strength (R) in dipole velocity form (Rvel), and the excitation energy (nm) were simulated into an ECD curve using SpecDis software (Berlin, Germany), which uses the Harada-Nakanishi approximation (Equation (1)) (Bruhn et al. 2013). The spectra were generated with a r/c value of 0.28 eV and a UV correction of -5.0 nm. All calculations were performed on the HP Cluster Platform 3000SL 'Miztli', a parallel supercomputer with a Linux operating system, containing 25,312 cores and a total of 15,000 GB of RAM. The Harada-Nakanishi equation:

Pharmacological evaluation
Animals Male CD1 mice (25-30 g of body weight) provided by the Bioterium of the Faculty of Sciences-UNAM were used in groups of six animals. Mice were placed in acrylic boxes with water and food ad libitum and kept at a controlled temperature of 22 ± 2 C and with a 12 h light/dark cycle. The protocol was accepted on 4 June 2018 by the Research Ethics Committee for Animal Use of the Instituto Nacional de Psiquiatr ıa Ram on de la Fuente Muñiz (COMBIOETICA-09.CEI-010-20170316). The experiments were carried out following national (NOM-062-ZOO-1999) and international ethical guidelines for the use of laboratory animals.

Antinociceptive-like effects
Mice were divided into six groups to receive the following treatments by oral administration (p.o.): vehicle (0.5% Tween 80 in saline solution), the reference drug diclofenac or ketorolac (10 mg/kg, p.o.), the crude extract (hexane, ethyl acetate or methanol at 300 mg/kg, p.o.) or the neo-clerodane diterpene (0.1, 1 or 10 mg/kg, p.o.). Thirty minutes after the treatment, mice were independently subjected to the nociceptive agent to induce pain-like behaviour.

Writhing test
This test consisted of an i.p. administration of 1% acetic acid (0.1 mL/10 g body weight) to cause irritation in the peritoneal cavity, which in turn produced stereotyped behaviour in mice, where stretching of the hind limbs and abdominal contractions were characteristic. The number of stretches (writhes) was counted every 5 min for a total period of 30 min after injection of the nociceptive agent (Collier et al. 1968). The decrease in writhing behaviour was interpreted as an antinociceptive effect.

Formalin test
This test consisted of an intraplantar injection of 20 mL of 1% formalin in the right posterior paw of the mice to produce licking behaviour as painful reaction. Then, the time spent in licking the limb administered with the nociceptive agent was counted as a pain-like behaviour for 1 min every 5 min for a 30 min period. Two stages of licking behaviour, considered as painful response, were recorded in this test. The first was the early or neurogenic phase (0-10 min) and the second was the late or inflammatory phase (10-30 min) (Tjolsen et al. 1992). The decrease in licking behaviour in either phase was interpreted as an antinociceptive effect.

Evaluation of the anxiolytic-like effects
Mice were divided into groups of six animals to receive one of the following treatments: vehicle (0.5% tween 80 in saline solution), the reference drug clonazepam (0.5 mg/kg, p.o.) and the organic extracts (hexane, ethyl acetate or methanol at 300 mg/kg, p.o.) or the neo-clerodane diterpene (0.1, 1 or 10 mg/kg, p.o.). Thirty minutes after the treatment, animals were consecutively tested in the anxiety models, such as open-field, hole-board, and finally in the plus-maze test.

Open field test
This test consisted of placing mice in an individual manner in the acrylic box divided into 12 squares (6 Â 6 cm) to record the number of squares explored by each animal in a 2 min period (Prut and Belzung 2003). A significant decrease exploration was indicative of the sedative-like effects.

Hole-board test
This test consisted of an acrylic box with wooden floor with evenly distributed holes. Mice were placed on the board and the number of times the animals put their head into the holes in a period of three min was counted. The decreased exploration is indicative of the anxiolytic-like effects (Lister 1990).

Statistical analysis
Data from pharmacological evaluations are presented as the mean ± standard error of the mean (S.E.M.) of each treatment. Statistical analysis was carried out using one-way analysis of variance (ANOVA) followed by Dunnett's post hoc test to compare treatments against the vehicle group. Data from the temporal course curve of writhing test were analysed by two-way ANOVA followed by Dunnett's post hoc test. The analysis was done using GraphPad version 6 program (GraphPad Software, La Jolla, CA). A value of p < 0.05 was considered significant.

Results
Fractionation, purification and structural elucidation of a neo-clerodane diterpene From the pharmacological evaluation of the extracts, it was obtained that the medium polar extract showed the highest yield, being 5-and 2-fold more abundant than the hexane and methanol, respectively (Figure 1). The chromatographic separation of the ethyl acetate extract lead to the isolation of the of the most abundant metabolite, identified as 7-keto-neo-clerodan-3,13-dien-18,19: 15,16-diolide (1), a neo-clerodane diterpene. This compound was purified for the first time from S. semiatrata as a crystalline form (Figure 2). The compound 1 was isolated and crystallized as white rectangles grown by slow solvent evaporation from a chloroform solution, m.p. (205-206 C). The compound 1 was considered pure according to 1 H NMR and 13 C NMR spectra ( Supplementary Figures 1-3). Spectroscopic data agreed to those previously reported for this compound isolated from Baccharis species (Asteraceae), such as Baccharis trinervis Pers. (Kuroyanagi et al. 1993), Baccharis trimera (Less.) DC. (Herz et al. 1977) and Baccharis gaudichaudiana DC. (Akaike et al. 2003). The ATR-IR spectrum (Supplementary Figure 4) showed two strong bands for a,b-unsaturated c-lactone and one band for carbonyl group (1767, 1739 and 1701 cm À1 ). Here in, we reported for the first time full structural details of the trans-neoclerodane diterpene 1 by X-ray single-crystal analysis (Supplementary Table 1). Figure 2 shows a perspective view of the molecular structure of 1 where the compound crystallized in a monoclinic system with the P2 1 space group. The cyclohexene and cyclohexanone rings are trans-fused and adopt twist boat and chair conformations, respectively. The c lactone has an envelope conformation and the 13(16H) furanone ring is practically planar. The absolute stereochemistry at C-5, C-8, C-9 and C-10 was determined as S, S, R and R, respectively, based on the comparison of the calculated and experimental circular dichroism (ECD) spectrum (Figure 3), as well as the Flack parameter was 0.06. The molecular models for the most stable conformers (Supplementary Figure 5) of compound 1 were used to calculate the ECD spectrum. The experimental ECD spectrum for this compound showed two negative Cotton effects at 244 and 295 nm, which were in good agreement with those displayed for enantiomer 5S8S9R10R. The methyl substituent at C-17 is axial and the methyl group C-20 occupies the more sterically favoured equatorial position. The crystal structure is stabilized by weak CHÁ Á ÁO contacts forming infinite chains as it is shown in Figure 4. The C(6)Á Á ÁO(3) and H(6)Á Á ÁO(3) distances are 3.67(4) and 2.69(6) Å and the angle C(6)-HÁ Á ÁO(3) is 168.6 (5) .
Antinociceptive effects of S. semiatrata crude extracts and the bioactive compound 1 Much clinical pain occurs in deep muscle tissues and/or visceral organs, because of that, the acute and tonic pain models to specifically study muscle or visceral nociception are the writhing and formalin tests. To select the most active extract during the preliminary screening of the antinociceptive effect of the S. semiatrata, it was evaluated each in a single dose of 300 mg/kg, p.o. using both, writhing and formalin tests. In the writhing test, it was observed that all the extracts at 300 mg/kg, p.o. and the ketorolac (reference drug at 10 mg/kg, p.o.) produced significant reduction (F 4,25 ¼ 18.90, p < 0.001) in the nociceptive behaviour compared to mice treated with the vehicle (Table 1). The antinociceptive response was confirmed in the formalin test, where S. semiatrata extracts produced a significant reduction in the nociceptive response in both neurogenic (F 4,25 ¼ 11.57, p < 0.001) and inflammatory (F 4,25 ¼ 13.27, p < 0.001) phases of the test (Table 1).
Regarding compound 1, it was purified for the first time from the S. semiatrata aerial parts as a crystalline form. According to the temporal course curves, the nociceptive response after administration of 1 was significantly reduced using doses of 0.1, 1 or 10 mg/kg, p.o. (F 4,25 ¼ 46.35, p < 0.001), it was observed from the first 5 or 10 min after administration of the algogenic agent (time: F 6,30 ¼ 129.3, p < 0.0001; treatment: F 4,20 ¼ 50.18, p < 0.0001; interaction: F 24,120 ¼ 19.23, p < 0.0001) (Figure 5(A)). The nociceptive behavioural response observed in the vehicle group was almost totally inhibited at the dosage of 10 mg/kg, p.o. of 1. The effect was maintained throughout the entire period of the experiment resembling the antinociceptive response of the reference drug tested at the same dosage ( Figure 5(A)). Antinociceptive response of 1 in the writhing test was observed in a dose-dependent manner producing an ED 50 ¼ 4.15 mg/kg ( Figure 5(B)).
Antinociceptive effects of 1 were also observed in the nociceptive response induced in the 1% formalin test. It was reduced in a significant manner at all doses tested resembling the effect of diclofenac, but this effect was not observed in a dose-dependent manner in both neurogenic (F 4,25 ¼ 11.57, p < 0.001) ( Figure  6(A)) and inflammatory (F 4,25 ¼ 12.40, p < 0.001) (Figure 6(B)) phases like in the writhing test.
Anxiolytic-like effects of S. semiatrata crude extracts and bioactive compound 1 Anxiolytic-like effects of S. semiatrata were preliminary explored by administrating crude extracts (hexane, ethyl acetate or methanol) at a dosage of 300 mg/kg, p.o. The hexane and ethyl acetate extracts produced an equivalent significative anxiolytic-like response to that obtained with the reference drug clonazepam at 0.5 mg/kg, p.o. in the plus-maze test as compared to the vehicle group (F 4,25 ¼ 41.66, p < 0.0001) ( Table 1). Tranquilizing response produced with all the extracts was not associated with the sedative-like effects, and it was not as intense like that observed with the reference drug in the open-field (F 4,25 ¼ 3.14, p < 0.03) and hole-board (F 4,25 ¼ 4.19, p < 0.001) tests (Table 1).
In the case of compound 1, it produced a significant increase in the open-sided arms examination resembling the reference drug clonazepam response in comparison to the vehicle group (F 4,25 ¼ 19.64, p < 0.0001) (Figure 7(A)). Unlike the extracts, compound 1 significantly modified ambulatory activity of mice in the open-field test (F 4,25 ¼ 62.91, p < 0.0001) (Figure 7(B)), but not in the dipping-like behaviour of the hole-board, where the highest dosage of 1 almost reached similar response to that observed with clonazepam (F 4,25 ¼ 4.72, p ¼ 0.006) (Figure 7(C)). These results together suggest an increase in the tranquilizing effect when compound was administered.

Discussion
The genus Salvia has been a source of medicinal plants with activity on the CNS (Imanshahidi and Hosseinzadeh 2006), regarding to the antinociceptive activity of different species of Salvia. It has been explored in several organic extracts. However, the medium polar extracts have demonstrated higher activity Figure 3. Experimental and calculated ECD spectra for 1-5S8S9R10R (red line) and 1-5R8R9S10S (blue line). A negative shift of 5 nm was required to match the spectra. among them. For example, acetone extract showed greater reduction of the nociceptive response than methanol extract of Salvia aegyptiaca L. evaluated at doses from 250 to 2000 mg/kg, p.o. in the writhing and formalin tests in mice (Al-Yousuf et al. 2002). Hexane and chloroform extracts of S. officinalis leaves produced dose-dependent anti-inflammatory response by inhibiting croton oil-induced oedema in the mouse ear, where chloroform extract was the most active in comparison to a lower response obtained with methanol extract (Baricevic et al. 2001). In a similar manner, a chloroformic extract of Salvia wiedemannii Boiss. at 500 mg/kg, i.p. produced significant antinociceptive effect in the writhing and tail-flick tests in mice while, polar extracts like aqueous, ethanol or butanol were not effective (Ustun and Sezik 2011). Nevertheless, some Salvia species have been tested prepared only as a polar extract, for example, the ethanol of S. plebeia evaluated at doses of 50-200 mg/kg, p.o., which showed activity in abdominal pain and inflammation induced by carrageenan in mice (Jung et al. 2009). The antinociceptive effect in different types of nociception has been also reported for S. divinorum, partly due to the presence of salvinorins found in the medium polar extracts (Tlacomulco-Flores et al. 2020). All these results together suggest that medium polar constituents of Salvia species are mainly responsible of their antinociceptive effects and potential source of analgesic drugs.
Although there are several studies reporting the antinociceptive effects of different extracts and species of Salvia, there is not enough research on the main responsible active metabolites. In the case of neo-clerodane terpenes, there is also scarce information about its potential to relieve pain. An exception is S. divinorum, which is well known because of the presence of salvinorin A, a neo-clerodane terpene widely studied for its depressant activity on the CNS (Casselman et al. 2014). The dose-dependent antinociceptive effects of salvinorin A have been explored using the tail-flick and the hot plate tests (Tlacomulco-Flores et al. 2020), formalin (Guida et al. 2012) and writhing tests (McCurdy et al. 2006) in mice. In the present study, it should be noted that 1 tested at 10 mg/kg, p.o. produced a dose-dependent antinociceptive effect that remained until the end of the experiment (30 min) in the writhing test in mice ( Figure 5(A)). Our data agree to that reported for tilifolidione (1-100 mg/kg, p.o.), a clerodane diterpene isolated from Salvia tiliifolia Vahl, which produced a dose-dependent antinociceptive effect in the writhing and formalin tests in mice (Gonz alez-Ch avez et al. 2018). Whereas, amarisolide (1, 5 and 10 mg/kg, i.p.), a neo-clerodane isolated from the aerial parts of S. circinata, produced significant but non-dose-dependent antinociceptive effects in the writhing test in mice (Moreno-P erez et al. 2019). These data together  reinforce the participation of neo-clerodane diterpenes as responsible for the depressant CNS activity of sage and their potential as alternative for pain treatment. Anxiolytic and sedative-like activities in Salvia species and their bioactive metabolites have not been enough explored in preclinical or clinical studies. Nevertheless, anxiolytic-like effects for the essential oil of the S. miltiorrhiza aerial parts (100 and 200 mg/kg, p.o.) using the plus-maze test in rats (Liu et al. 2015) have been reported. While, this effect has also been observed in the hydroalcoholic extract (100 mg/kg, i.p.) of the S. reuterana aerial parts (Rabbani et al. 2005) and the flowers and leaves of S. elegans tested at 50 and 100 mg/kg, p.o. in mice, using both the plus-maze and open field tests, where ambulatory activity was not modified (Herrera Ruiz et al. 2006). An interesting case was that observed in S. divinorum leaves by an electroencephalography study in rats, which produced depressant activity and potentiation of the sodium pentobarbital-induced sleep. S. divinorum was prepared as hexane, ethyl acetate and methanol extracts and tested at doses of 10, 30 and 100 mg/kg, i.p. Its central effects were emphasized due to the presence of diterpenes such as salvinorins isolated in the medium polar extract (Gonz alez-Trujano et al. 2016).
In general, anxiolytic drugs are intended to reduce anxiety or tension at doses which do not cause sedation or sleep. It is because oversedation, tremor, ataxia and confusion are the adverse effects much more common in patients consuming some anxiolytics drugs like benzodiazepines. Medicinal plants might produce its anxiolytic like activity without sedative effects at certain doses. In our study, to explore ambulatory and exploratory activities in mice at several doses was very important in order to identify if anxiolytic-like response is associated to sedative effects and at what range of doses it will be produced. Literature has reported compounds from different chemical nature, such as non-polar and polar, as responsible for the depressant-like activity of these medicinal plants, but diterpenoids have been the most diverse metabolites characterized in Salvia genus. According to their structure, these are classified into abietanes, labdanes, clerodanes, pimaranes and icetexanes. Biological activities, such as antimicrobial, antibacterial and cytotoxic, have been reported from these diterpenes (Wu et al. 2012). However, there are few studies which explore their effects on the CNS, specifically for neo-clerodanes diterpenes (Moreno-P erez et al. 2019). The most studied case is salvinorin A, isolated from the leaves of S. divinorum, which has been reported for its therapeutic effects on Alzheimer's or bipolar disorders (Casselman et al. 2014), as well as sedative (Fantegrossi et al. 2005), anxiolytic (Braida et al. 2009) and antinociceptive (McCurdy et al. 2006;Guida et al. 2012) activities. Salvinorin A is known as an agonist of the kappa opioid receptors, but it also possesses a partial affinity for CB1 cannabinoid receptors (Braida et al. 2009). It shows similar chemical structure than compound 1 and it is known to produce anxiolytic-like effects at doses of 0.1, 1.0, 40, 80 or 160 lg/kg, s.c. in the plus-maze test, without effects on the ambulatory activity of mice tested at 0.01, 0.1 and 1 lg/kg, s.c., probably because of the use of a lower dosage. We have not explored mechanism of action in the present study. However, to know the doses in which an anxiolytic drug produces its effects without sedation is very important since both sedative and anxiolytic drugs produce CNS inhibitory effects. Activation of the kappa opioid receptor relieves pain, but it is also associated to dysphoria (feeling of discomfort or aversion) and sedation  and hole-board test (C, Ã p < 0.006) after ANOVA followed by Dunnett's post hoc test. ns: non-significant differences. (McDonald and Lambert 2005). Whereas, it has been reported that a mild activation of CB1 receptors in the prefrontal cortex and ventral hippocampus attenuates anxiety (Rubino et al. 2008). In the future, it will be interesting to explore about the mechanisms of action of compound 1 to produce its CNS depressant effects.
In conclusion, the present study gives evidence of the antinociceptive and anxiolytic-like effects of S. semiatrata, in part due to the presence of the diterpene neo-clerodane 7-keto-neoclerodan-3,13-dien-18,19: 15,16-diolide obtained for the first time as a crystal, reinforcing the medicinal properties of this species as a tranquilizer and for pain relief.