Antinociceptive compounds and LC-DAD-ESIMSn profile from Dictyoloma vandellianum leaves

Limonoids, quinolone alkaloids and chromones have been reported as constituents of Dictyoloma vandellianum Adr. Juss. (Rutaceae). Although those compounds are known for their biological activities, only the anti-inflammatory activity of chromones isolated from the underground parts has been evaluated. There are no studies of the pharmacological properties of the aerial parts of D. vandellianum. The present study was carried out to determine the phytochemical profile and antinociceptive activity of the methanol extract, fractions and isolated compounds of leaves of D. vandellianum. The phytochemical profile was performed by HLPC-DAD-ESIMSn and pure substances obtained were characterized by MS and NMR spectroscopy. The antinociceptive activity was assessed using the formalin assay in mice, and the motor function in the rotarod test. ME and all the fractions obtained from ME produced antinociceptive effects. Among them, the ethyl ether fraction was the most active. Data from HPLC-DAD-ESIMSn showed that the ethyl ether fraction presented 42 compounds. The major compounds isolated from this fraction—gallic acid, methyl gallate and 1,2,6-tri-O-galloyl-β-d-glucopyranose–were tested and produced antinociceptive effects. Gallic acid, methyl gallate and 1,2,6-tri-O-galloyl-β-d-glucopyranose at antinociceptive doses did not affect the motor performance in mice in the rotarod test. This work is the first report of the occurrence of gallotanins in D. vandellianum. In addition, the pharmacological study showed that D. vandellianum leaves present antinociceptive activity, probably induced by gallic acid, methyl gallate and 1,2,6-tri-O-galloyl-β-d-glucopyranose.


Introduction
Pain is an unpleasant sensation that affects 20% of adults worldwide and it is associated with a wide range of diseases and tissue damage [1]. Currently, pharmacological management of pain a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 All solvents (analytical/HPLC grade) were purchased from Quimex (São Paulo, Brazil) and Tedia (Rio de Janeiro, Brazil) and used without further purification. Indomethacin and dimethyl sulfoxide were obtained from Sigma Chemical Company (St. Louis, MO, USA). Indomethacin was dissolved in Tris HCl 0.1 M pH 8.0 plus saline. The chloroform fractions were dissolved in 5% DMSO plus saline, and the remaining fractions and compounds were dissolved directly in saline.

Plant material
The

Liquid chromatography-mass spectrometry instrumentation and conditions
A Shimadzu 1 (Kyoto, Japan) High Performance Liquid Chromatography System, coupled with an Amazon X or micrOTOF II (Bruker Daltonics, Billerica, MA, USA) with an electrospray ion (ESI) source, was used to perform the ESI-MS n and HRESIMS analysis, respectively. The LC System consisted of a LC-20AD solvent pump unit (flow rate of 600 μL.min −1 ); a DGU-20A 5 online degasser; a CBM-20A system controller and a SPD-M20A (190-800 nm) diode array detector. The LC separation was performed on a Kromasil C-18 5 μm 100Å, 250 × 4.6 mm (Kromasil, Bohus, Sweden) analytical column. Injections (20 μL) were performed using an autosampler (SIL-10AF). The mobile phase consisted of 0.1% formic acid in water (solvent A) and methanol (solvent B). Exploratory linear gradient (5 × 100% B) was performed to elution in 90 min. The analysis parameters are as follows: capillary 4.5 kV, ESI in negative mode, final plate offset 500 V, 40 psi nebulizer, dry gas (N 2 ) with flow rate of 8 mL/ min and a temperature of 300˚C. CID fragmentation, in Amazon X, was achieved in auto MS/ MS mode using enhanced resolution mode for MS and MS/MS mode. The spectra (m/z 50-1000) were recorded every 2 s. controlled rooms (22-23˚C), under a 12:12 h light-dark cycle, with access to water and food ad libitum. Environmental enrichment was obtained with mouse igloos. Animal care and handling procedures were in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The present protocol was approved by the Institutional Animal Care and Use Committee, Ethics Committee for Animal Experimentation of FIOCRUZ (CEUA/FIOCRUZ. Permit Number: L-IGM-015/ 2013). Every effort was made to minimize the number of animals used and any discomfort. Accordingly, the animals were only used once and were sacrificed immediately after experimentation with isoflurane overdose. Behavioral tests were performed without knowing to which experimental group each mouse belonged.

Antinociceptive activity-Formalin test
The mice were placed in an open Plexiglas observation chamber for 30 min to acclimate to their surroundings. They were then removed and gently restrained while 20 μL of 2.5% formalin (1:100 dilution of stock formalin solution, 37% formaldehyde in 0.9% saline) was injected subcutaneously into the dorsal surface of the hind paw using a 30 gauge needle. Following injection, the mice were returned to the observation chamber for a 30 min observation period. The nociceptive score was determined by counting the time the animal spent licking the injected paw during the early phase (0-10 min) and the late phase (10-30 min) [28]. The effects of ME (1.95-125 mg/kg) and its ether, ethyl acetate and chloroform fractions (100 mg/ kg) were evaluated on the formalin test. Next, gallic acid, methyl gallate and 1,2,6-tri-O-galloyl-β-D-glucopyranose (0.19-200 mg/kg) obtained from the ethyl ether fraction were also evaluated. All treatments were administered once, by intraperitoneal route, 40 min before the injection of formalin. Indomethacin (10 mg/kg) and morphine (5 mg/kg) were used as the reference drugs. Vehicle group was treated with saline or 5% DMSO plus saline, as appropriate.

Motor function assay-Rotarod test
To evaluate possible non-specific muscle-relaxant or sedative effects of the treatments, immediately before the formalin assay, the mice were submitted to the rotarod test, in a modified form as previously described [29]. The rotarod apparatus (Insight, Ribeirão Preto, SP, Brazil) consisted of a bar with a diameter of 3 cm, subdivided into five compartments. The bar rotated at a constant speed of eight revolutions per min. Mice were trained 24 h before the experiment to remain on the bar for 120 s. Those not remaining on the bar for two consecutive periods of 120 s were not included in the study. Forty minutes after the intraperitoneal injection of diazepam (10 mg/kg, reference drug), ME (200 mg/kg), gallic acid (200 mg/kg), methyl gallate (200 mg/kg), 1,2,6-tri-O-galloyl-β-D-glucopyranose (200 mg/kg) or vehicle, the animals were placed on the rotating rod and the latency to fall was measured for up to 120 s. The results are expressed as the average time(s) the animals remained on the rotarod in each group.

Statistical analysis
Data are presented as means ± standard error of the means (SEM) of measurements made on six animals in each group. Comparisons between three or more treatments were made using one-way ANOVA with Tukey's post hoc test. All data were analyzed using Prism 5 Computer Software (GraphPad, San Diego, CA, USA). Statistical differences were considered to be significant at p < 0.05. The ED 50 (dose of an agonist that produces 50% of the maximal possible effect of that agonist) values were expressed with a confidence interval (CI). Individual doseresponse curves were fitted with the Hill logistic equation. ED 50 values were obtained as the dose at which a half maximal reduction in nociceptive score occurred, calculated with 95% confidence limits.

Phytochemical study
The HPLC-DAD-ESIMS n analysis detected the presence of 42 compounds in the ether ethyl fraction (see Fig 1 and Table 1). The three known compounds that were isolated from this fraction had their structures confirmed by 1 H and 13 C NMR and mass spectrometry. The isolated compounds were identified as gallic acid, methyl gallate and 1,2,6-tri-O-galloyl-β-Dglucopyranose.

Antinociceptive activity
The antinociceptive properties of test compounds were investigated in the formalin test in mice. Initially, the effect of systemic injection of methanol extract of D. vandellianum leaves (ME) on the formalin-induced nociception was evaluated (Fig2). Administration of formalin in control mice induced a biphasic flinching response, with early phase ranging from 0 to 10 min and late phase from 10 to 30 min after the injection. Intraperitoneal injection of ME (7.8-125 mg/kg), 40 min prior to the formalin injection, produced a significant antinociceptive effect in both early and late phases of the test (Fig 2A and 2B, respectively). ME at 1.95 mg/kg did not induce antinociception in the formalin test. In the late phase of the test, the MEinduced antinociceptive effect showed a dose-dependent profile, with ED 50 value of 8.03 mg/ kg (CI 4.62 to 13.94).
Antinociceptive effects of the three fractions (chloroform, ethyl ether and ethyl acetate) obtained from ME were evaluated next (Fig 3). A decrease in nociceptive score was induced by the systemic injection of ethyl ether (FEE; 100 mg/kg) and ethyl acetate (FEA; 100 mg/kg) fractions in both early and late phases of formalin test. Inhibitory effect of the ethyl ether fraction was of 76% and 95% on early and late phases of formalin test, respectively. Ethyl acetate fraction induced an inhibitory effect of 63% and 75% on the early and late phases, respectively. On the other hand, the pretreatment with chloroform fraction (FCHCl 3 , 100 mg/kg) produced antinociceptive effect only in late phase of the test, displaying 66% of nociception inhibition.
Considering that FEE induced a relevant antinociceptive effect in the formalin assay, its phytochemical profile was next evaluated to identify constituents within this fraction with  antinociceptive properties. A phase-reverse chromatography of the fraction was performed, allowing isolation of gallic acid (GA), methyl gallate (MG) and 1,2,6-tri-O-galloyl-β-D-glucopyranose (TGG). GA at 12.5 (p < 0.01), 50 (p < 0.001) and 200 (p < 0.001) mg/kg, administered by intraperitoneal route 40 min before formalin injection, reduced the nociceptive behavior of mice in late, but not early phase of the test (Fig 4). GA at 3.12 and 0.78 mg/kg did not induce effect on formalin test. Pretreatment with indomethacin (10 mg/kg, ip), a standard nonsteroidal anti-inflammatory drug, produced a similar inhibition profile of late phase (p < 0.001). As expected, the pretreatment with morphine (5 mg/kg, ip), a gold standard opioid, inhibited both the early (p < 0.01) and late (p < 0.001) phase of the formalin test. Intraperitoneal administration of MG (3.1-200 mg/kg), 40 min before formalin injection, inhibited the late phase of formalin test (p < 0.001; p < 0.01; p < 0.05). MG at 0.78 and 0.19 mg/kg showed no antinociceptive effect. MG, at all doses tested, did not induce a statistically significant reduction in the early phase of the formalin test. Systemic pretreatment with TGG at doses between 0.78 and 200 mg/kg inhibited the late, but not the early phase of the formalin test (p < 0.001; p < 0.01). TGG at 0.19 mg/kg had no effect in this assay. In the late phase of the formalin test, the TGG-induced antinociception was a dose-dependent effect, as indicated by the statistically significant difference between active doses (p < 0.01), displaying an ED 50 value of 0.80 mg/kg (CI 0.22 to 2.57).
In the rotarod test, the intraperitoneal administration of ME, gallic acid, methyl gallate and 1,2,6-tri-O-galloyl-β-D-glucopyranose (200 mg/kg), did not reduce the run time of the mice, indicating that these treatments did not induce motor performance alterations ( Fig 5). As expected, the central nervous system depressant diazepam (10 mg/kg) reduced the time of mice on the rotarod after 40 min of intraperitoneal treatment with this standard drug.

Fig 2. Effects of the methanol extract of D. vandellianum leaves (ME) on formalin test in mice.
Mice were treated with vehicle (saline, control group) or ME (125-1.95 mg/kg) by intraperitoneal route 40 min before the intraplantar injection of formalin (injected at time zero). Mice were observed from 0 to 10 min (early phase; A) and from 10 to 30 min (late phase; B), and a nociceptive score was determined for each period by counting the time in seconds that the animal spent licking the injected limb during the observation time. Data are expressed as mean times ± S.E.M.; n = 6 mice per group. Statistical significance relative to the control group: � (p < 0.05); �� (p < 0.01); ��� (p < 0.001). Statistical significance relative to the 7.81 mg/kg group: # (p < 0.05). Statistical significance relative to the 125 and 31.25 mg/kg groups: $ (p < 0.05). ANOVA followed by Tukey's test.

Phytochemical profile
The 42 compounds were tentatively assigned in ethyl ether fraction from D. vandellianum by the interpretation of their fragmentation patterns obtained from mass spectra (HRESIMS, MS 2 and MS 3 experiments). Data provided by isolated compounds and literature information was also employed for the comprehensive evaluation of samples. The retention times and mass spectrum data along with peak assignments for compounds identified using negative ionization are described in Table 1. Mice were treated with vehicle (control groups) or fractions (100 mg/kg) by intraperitoneal route 40 min before the intraplantar injection of formalin (injected at time zero). FEE: ether fraction from ME (solubilized in saline); FEA: ethyl acetate fraction from ME (solubilized in 5% DMSO); FCHCl 3 : chloroform fraction from ME (solubilized in 5% DMSO). Vehicle 1: saline (control group of FEE). Vehicle 2: 5% DMSO plus saline (control group of FEA and FCHCl 3 ). Mice were observed from 0 to 10 min (early phase; A) and from 10 to 30 min (late phase; B), and a nociceptive score was determined for each period by counting the time in seconds that the animal spent licking the injected limb during the observation time. Data are expressed as mean times ± S.E.M.; n = 6 mice per group. � Significantly different from vehicle 1 group (p < 0.05); �� significantly different from vehicle 1 group (p < 0.01); ��� significantly different from vehicle 1 group (p < 0.001); # significantly different from vehicle 2 group (p < 0.01). ANOVA followed by Tukey's test.  [28,29]. Similarly, trigalloyl-hexoside (15, 19, 23, 27 and 29) and tetragalloyl (26) were tentatively assignment [28,29]. Furthermore, galloylquinic acid derivatives were identified from observation of the precursor's ions at m/z 343, m/z 495 and m/z 647, which were attributed to galloylquinic acids (4 and 7), digalloylquinic acids (11, 14 and 18) and trigalloylquinic acids (22)
Regarding to identification of flavonoids, they were classified in two groups: O-glycosides and C-glycosides, which were identified comparing their MS/MS spectra with those available in literature. It was observed two parent ions at m/z 301 and 285, both result as product an ion at m/z 151, identified as quercetin and kaempferol, respectively [28,35] and 311 corresponding to Cglycoside fragmentation pattern, which was suggested as isovitexin [32]. Peak 30 was identified as your isomer, vitexin [32].
Compound 1,2,6-tri-O-galloyl-β-D-glucopyranose was obtained as a yellowish brown amorphous powder. The 1 H NMR spectrum reveals three singlets, each one integrating for two protons at δ H 7.10, 7.15 and 7.18, assignable to the aromatic protons of the three galloyl moieties the molecule. In the carbohydrate region, the spectrum clearly shows three downfield Pharmacological and chemical profile from Dictyoloma vandellianum hydrogens resonances: a doublet at δ H 5.85 integrating one proton with large coupling constant, indicating a β-configuration of the anomeric proton; a triplet at δ H 5.31, and two signals at δ H 4.61 (d, J = 12 Hz) and 4.47 (dd, J = 12, 4.4 Hz) which were assigned to H-2 and H-6 glucose hydrogen [38]. The signals of these hydrogens are downfield compared to those of β-Dglucopyranose, indicating the location of galloyl units at these positions [39]. The structure was confirmed by the analysis of the 13 C-NMR spectrum. Three carbonyl ester signals at δ C 166.9, 168.2 and 168.3, confirm the presence of the three galloyl moieties. Signals at δ C 94.5 (C-1), 68.3 (C-2) and 62.9 (C-6) indicate that the hydroxyl groups at these positions are galloylated [35]. Its molecular formula was defined as C 27 1.7 ppm). The structure of this gallotannin was confirmed by comparison of the above described spectra data with the literature and was identified as 1,2,6-tri-O-galloyl-β-D-glucopyranose [40].
The isolated compounds gallic acid and methyl gallate also had their structures confirmed by comparing the 1 H and 13 C NMR and mass spectrometry data with the literature data [40][41][42]. In such a way, the known compounds gallic acid, methyl gallate, 1,2,6-tri-O-galloyl-β-Dglucopyranose were identified.

Antinociceptive activity
The present study demonstrates that systemic administration of methanol extract of D. vandellianum leaves produces a consistent antinociceptive effect. Fractionation of D. vandellianum leaves showed that this effect could be attributed, at least in part, to their bioactive constituents gallic acid, methyl gallate and 1,2,6-tri-O-galloyl-β-D-glucopyranose.
The formalin test is a useful screening tool for assessment of the analgesic properties of plant extracts and compounds. In the present study, data from this assay indicated that ME induces a dose-dependent antinociceptive effect. Afterwards, this methanol extract was fractionated and the obtained fractions (ether, ethyl acetate and chloroform) were also evaluated in formalin test. All tested fractions presented antinociceptive effects in vivo, but the ethyl ether fraction was the most active. In line with this result, ethyl ether fraction was selected for the phytochemical study, evidencing three major compounds: gallic acid, methyl gallate and 1,2,6-tri-O-galloyl-β-D-glucopyranose. In order to evaluate the eventual contribution of these compounds to antinociceptive action of ME, their biological properties were next evaluated in the formalin test.
The results obtained from the present work demonstrate that GA induces antinociceptive effects in the late phase of the formalin test. From a pharmacological point of view, it has been proposed that drugs that block the nociception transmission, such as opioid analgesics, suppress both phases of formalin-induced pain response, whereas nonsteroidal anti-inflammatory drugs, such as indomethacin, seem to suppress only the late phase [43]. Based on these concepts, it is possible to propose that GA-induced antinociception is associated with antiinflammatory properties. In line with this idea, the antinociceptive effect of GA has been demonstrated in the carrageenan-induced inflammatory pain model [44]. In addition, this compound has also been reported to have relevant anti-inflammatory and immunomodulatory activities [45][46][47].
Like GA, methyl gallate presented an antinociceptive effect on the inflammatory phase of formalin test, displaying analgesic activity with a similar profile of NSAIDS. Supporting this hypothesis, it was previously demonstrated that MG has anti-inflammatory and cyclooxygenase-2 inhibitory activities [48,49]. Even though the antinociceptive properties of gallic acid ethyl ester have been previously described [50], the present data demonstrate the antinociceptive effect of MG.
Few studies have reported the pharmacological properties of TGG, and most of them describe its antibacterial and antiviral activities [40,[51][52][53]. A previous study has demonstrated antinociceptive activity of methanolic extract of Miconia minutiflora (Bonpl.) DC., which contains gallotannin class in its composition [54]. The present study displays the direct pharmacological evidence for the antinociceptive activity of gallotannin. TGG exhibited antinociceptive effect on the inflammatory phase of formalin test, and was more efficacious than indomethacin, the analgesic anti-inflammatory drug used as a gold standard in this test. Corroborating this relevant effect of TGG on the inflammatory phase of formalin, Erdèlyi and coworkers demonstrated by in vitro experiments, that this class of compound is able to reduce the inflammatory cytokines expression [55]. Importantly, gallic acid, methyl gallate and 1,2,6-tri-O-galloyl-β-Dglucopyranose at antinociceptive doses did not affect the motor performance in mice in the rotarod test. These results corroborate the antinociceptive properties pointed by formalin test.
Despite consistent antinociceptive effects induced by systemic treatment with GA, MG and TGG, the contribution of others bioactive molecules to ME-induced antinociception cannot be ruled out. The ether ethyl fraction of ME presented 39 minor compounds, and it is well accepted that minor constituents in plant extracts may contribute to pharmacological properties of these extracts through synergistic actions or independent pharmacological effects [56]. In addition, among the minor compounds identified in ME, some of them, such as kaempferol and quercetin, have well-stablished antinociceptive activity [57,58]. Antinociceptive properties of quercetin have been demonstrated in both experimental and clinical conditions. Quercetin produces dose-related antinociception in several models of chemical pain, such as acetic acid test, formalin test, nociception induced by glutamate and capsaicin [57], and oxaliplatininduced neuropathic pain [9], showing also analgesic properties in clinical conditions [59]. Similarly, kaempferol treatment is found to attenuate neuropathic pain [60] and chemical pain [61] in pre-clinical studies. Thus, it is possible that these compounds contribute to the antinociceptive effect of the ether ethyl fraction of ME, however, this hypothesis has not been investigated here.

Conclusions
The present study, using a classic model of analgesic drug screening combined with reliable methods of structural analysis, demonstrated that the methanolic extract of D. vandellianum leaves shows antinociceptive properties on experimental inflammatory pain. The isolation indicated that gallic acid, methyl gallate and 1,2,6-tri-O-galloyl-β-D-glucopyranose, major constituents of the ether fraction from ME, are the antinociceptive components of D. vandellianum. The antinociceptive properties demonstrated in formalin test were corroborated by the results of rotarod test, which did not demonstrate motor deficits and nonspecific depression in the nervous system. To the best of our knowledge this is the first report of the antinociceptive activity of 1,2,6-tri-O-galloyl-β-D-glucopyranose. The occurrence of these compounds in D. vandellianum is also described for the first time.