Three New Triterpenes from the Resinous Bark of Protium kleinii and their Antinociceptive Activity

Três triterpenos pentacíclicos inéditos foram isolados da resina de Protium kleinii. Suas estruturas, 3-oxo-11β,16β-diidróxi-urs-12-eno (1), 3-oxo-11β-hidróxi-urs-12-eno (2) e 3-oxo-11β-hidróxiolean-12-eno (3), foram elucidadas por espectroscopia de RMN. A administração sistêmica destes compostos (10-100 mg kg, i.p.) foi capaz de inibir tanto a primeira quanto a segunda fase da nocicepção induzida pela formalina em camundongos. Em adição, foram isolados os triterpenos conhecidos 3β,16β-dihidróxi-urs-12-eno (breina) (4) e uma mistura de α (5)e β-amirina (6).


Introduction
Plants of the genus Protium (Burseraceae) are known for the production of a resin exudate used for skin diseases, healings of ulcers, scirrhous and as an analgesic. 1Essential oils from the leaves and resin of various Protium spp.have been evaluated for their anti-inflammatory activity. 2,3riterpenes, mainly of the oleane and ursane series, occur as binary mixtures in the resin. 4,5Protium kleinii Cuatrecas is endemic in southern Brazil, where it is known as "almecega", "almiscar", "pau-de-breu" and "guapoi", inter alia. 6n an earlier paper 7 we reported the antinociceptive activity of the ether soluble portion and of 3β,16βdihydroxy-urs-12-ene (brein) obtained from its resin.

Results and Discussion
Conventional chromatographic separation of the ethyl ether soluble portion of the resinous bark led to the isolation of a mixture of α (5)-and β-amyrin (6) and of 3β,16β-dihydroxy-urs-12-en (brein) (4) as the main components.Compound 1 and an unseparable mixture of 2 and 3 were also obtained by extended purification and repeated crystallizations.
The molecular formula C 30 H 48 O 3 was assigned to compound 1 for the molecular peak at m/z 456 in the EIMS, for the presence of thirty signals in the 13 C NMR spectrum and for elemental analysis.The 1 H NMR spectrum showed signals attributable to eight methyls, two of which in secondary carbon (δ 0.95, d, J 7.0 Hz; δ 0.89, d, J 6.3 Hz), to an olefinic proton (δ 5.44, d, J 2.7 Hz) and to two hydroxymethines (δ 4.25, dd, J 11.1 and 5.2 Hz; δ 4.59, dd, J 9.8 and 2.7 Hz).In the 13 C NMR spectrum, inter alia, signals for a saturated carbonyl (δ 217.9), two oximethines (δ 81.7 and 66.7) and a trisubstituted double bond (δ 125.4 and 144.2) were evident.The unequivocal assignements of each NMR signal are reported in Table 1 as a result of 1D-and 2D-NMR experiments.Cumulatively, these data suggested an urs-12-ene triterpenoid with a carbonyl and two hydroxy substituents.Comparison of ring A signal with those of methyl-3-oxo-ursolate 8 permitted the location of the carbonyl on position 3.
Compounds 2 and 3 were obtained as a 2:3 mixture after repeated attempts of chromatographic purification and crystallization.The mixture showed a molecular peak in the mass spectrum at m/z 440 and analyzed for C 30 H 48 O 2 , but exhibited two sets of signals both in the 1 H and 13 C NMR spectra.In the proton spectrum were evident two doublets at δ 5.38 and 5.45 (J 2.8 Hz) (ratio 2:3) and two double doublet at δ 4.56 and 4.59 (J 9.4 and 2.8 Hz), which highly resemble the signals H-12 and H-11, respectively, of compound 1.The carbon spectrum disclosed fifty-six signals with a pair of signals at δ 124.8/121.2, at δ 146.1/ 152.8, at δ 218.0/217.8 and at δ 81.8/82.0.Therefore the mixture was considered to be constituted by two triterpenes of the ursane/oleane series with a carbonyl and an hydroxyl.Careful examination of the multiplicity of the signals in the DEPT experiment permitted the carbon assignments of 2 and of 3 as reported in Table 2; for instance, C-18 and C-19 are CH in the ursane 2, while they are CH and CH 2 , respectively, in the oleane 3.According to the method developed by Maia et al. 4 and Olea and Roque, 9 the assignments were corroborated by comparison with the data of ursane/oleane pairs in the literature, 10 by the relative intensity of the signals and by the Hetcor spectrum.In addition, in the Inept experiment the irradiation at δ 4.56- During the search in the literature a compound which claimed to be 3α,11α-dihydroxy-olean-12-ene, it was found. 13Nevertheless, the signal for C-11 is reported at δ 81.8 so that its structure should be revised to 3α,11βdihydroxy-olean-12-ene.
Recently was reported the isolation of 3β,16β,11αtrihydroxyurs-12-ene from Trattinnickia burserifolia. 14he signals of 13 C NMR were δ 68.2 for C-11α and δ 66.7 for C-16β, being these values confirmed by NOESY experiments.Such important data confirm our previous conclusions.

Pharmacological analysis
Animals.Male Swiss mice (25-35 g), housed at 22 ± 2°C temperature under a 12 h light/12 h dark cycle and with access to food and water ad libitum, were used.Experiments were performed during the light phase of the cycle.The animals were allowed to adapt to the laboratory for at least 2 h before testing and were used only once.Experiments reported in this study were carried out in accordance with current ethical guidelines for the investigation of experimental pain in conscious animals. 16ormalin test.The procedure used was essentially similar to that described previously. 17,1820 mL of 2.5% formalin solution (0.92% of formaldehyde), made up in PBS (NaCl 137 m7 mmol L -1 , KCl 2.7 mmol L -1 .and phosphate buffer, 10 m7 mmol L -1 ), were injected intraplantarly under the surface of the right hindpaw.Animals were treated with triterpenes (1) (10-30 mg kg -1 , i.p.) and (mixture 2, 3) (10-60 mg kg -1 , i.p.) or vehicle (10 mL kg -1 , i.p.), 0.5 h before formalin injection.After intraplantar injection of formalin, the animals were immediately placed in a glass cylinder 20 cm in diameter and the time spent licking the injected paw was measured with a chronometer, during 30 mins.The initial nociceptive scores normally peacked after 5 min (first phase, representing the neurogenic pain), and after 15-30 min after formalin injection (second phase, representing the inflammatory pain), this time being considered as indicative of nociception.
Statistical analysis.The results are presented as means ± S.E.M., except the ID 50 values (it. the dose of (1) reducing the pain responses by 50% relative to the control value) which are reported as geometric means accompanied by their respective 95% confidence limits.Data were analyzed by analysis of variance (ANOVA) or t-test and complemented by Dunnett's or Newman Keul's post-hoc test.P values less than 0.05 (P < 0.05) were considered as indicative of significance.The ID 50 values were determined by linear regression from individual experiments using "GraphPad Software", and are reported as geometric means accompanied by their respective 95% confidence limits.
Drugs.The following substances were used: formalin (Merck, AG, Darmstadt, Germany), was dissolved in 0.9% m/v of NaCl solution just before injection while capsaicin and ether fraction which were dissolved in absolute ethanol

Figure 2 .
Figure2.Effects of (mixture 2, 3) given intraperitoneally (10-100 mg kg -1 ) against the early (panel A) and the late phase (panel B) of formalin-induced licking in mice.The total time (mean ± S.E.M.) spent licking the hindpaw was measured in the first phase (0-5 min) and the second phase (15-30 min) after subplantar injection of formalin into the hindpaw.Each point represents the mean ± S.E.M. for 6-10 animals.The asterisks denote the significance levels when compared with control groups.Significantly different from controls, **P < 0.01.

Figure 1 .
Figure 1.Effects of(1) given intraperitoneally (10-60 mg kg -1 ) against the early (panel A) and the late phase (panel B) of formalininduced licking in mice.The total time (mean ± S.E.M.) spent licking the hindpaw was measured in the first phase (0-5 min) and the second phase (15-30 min) after subplantar injection of formalin into the hindpaw.Each point represents the mean ± S.E.M. for 6-10 animals.The asterisks denote the significance levels when compared with control groups.Significantly different from controls, *P< 0.05 and **P < 0.01.