Plumeran Alkaloids and Glycosides from the Seeds of Aspidosperma pyrifolium Mart

Departamento de Química Orgânica e Inorgânica, Universidade Federal do Ceará, CP 12200, 60021-940 Fortaleza-CE, Brazil Instituto de Química, Universidade Federal do Rio Grande do Norte, CP 1524, 59072-970 Natal-RN, Brazil Laboratório de Neurofarmacologia, Universidade Federal do Ceará, CP 3157, 60430-270 Fortaleza-CE, Brazil Laboratório de Neurologia Experimental, Faculdade de Ciências da Saúde (FACS), Universidade do Estado do Rio Grande do Norte (UERN), 59607-360 Mossoró-RN, Brazil Laboratório de Ciências Químicas, Universidade Estadual do Norte Fluminense, Campos dos Goytacazes, 28013-602 Rio de Janeiro-RJ, Brazil


Introduction
The Aspidosperma genus is reported as a prolific source of indole alkaloids, substances of great interest as a function of their structural diversity.Several biological activities associated with this class of compounds have been reported, [1][2][3][4][5] for example, the hypotensive and analgesic activities of A. quebracho-blanco, 4 and the antimicrobial and cytotoxic activities of A. marcgranianum. 4In addition, anti-inflammatory, anticancer, antimalarial, antiulcer, antileishmanial and healing activities are attributed to alkaloids from other Aspidosperma species. 1,2,4,5spidosperma pyrifolium Mart., popularly known as "pereiro preto" (Port.lit.= dark pereiro), is a shrub, sometimes a small tree, widely distributed in the northeastern Brazil flora.It is largely used in carpentry due to the excellent quality of its wood. 6Previous works report on the hypotensive effect 4,6,7 shown by the alkaloids present in the root bark and leaves of A. pyrifolium, as well the antiplasmodial activity 8 revealed by some alkaloids with aspidospermane skeleton, for instance aspidospermine, isolated from the stem bark.No antinociceptive or antiinflammatory activities have been reported so far.Despite the report of about 27 dihydroindole alkaloids with plumeran skeleton, isolated from leaves, roots and trunk bark of A. pyrifolium, 4,6,7 there is no report in the literature for the phytochemical study of its seeds.

General experimental procedures
The 1D and 2D NMR spectra were acquired on a Bruker Avance DRX-500 and/or DPX-300 spectrometers, using CD 3 OD, C 5 D 5 N, (CD 3 ) 2 SO or CDCl 3 as solvents.All standard pulse sequences were provided by the TopSpin software from Bruker.The Fourier transform infrared (FT-IR) spectra were obtained on a Perkin Elmer Spectrum 1000 spectrometer, using an universal attenuated total reflectance accessory (UATR).The high-resolution electrospray ionization mass spectra (HRESIMS) were acquired using a Shimadzu LCMS-IT-TOF (225-07100-34) spectrometer.Specific rotations were measured on a Jasco polarimeter model P-2000.Column chromatography was performed using silica gel 60 (EMD, 70-230 mesh) and Sephadex LH-20 from Pharmacia Fine Chemicals.Thin layer chromatography (TLC) was performed on precoated silica gel aluminum plates (TLC Silica gel 60 F 254 , from Merck).The compound spots were visualized by UV light detection and/or sprayed with a solution of vanillin/perchloric acid/EtOH followed by heating, as well as with the Dragendorff reagent.The analyses by high-performance liquid chromatography (HPLC) were performed on a Waters chromatograph, using semi-preparative (250 × 10 mm, 5 mm) and analytical (250 × 4.6 mm, 5 mm) Phenomenex RP-18 columns, with flow rate of 4.72 and 1.0 mL min -1 , respectively.

Plant material
The seeds of A. pyrifolium were collected in Cabrobó city neighborhoods (Pernambuco, Brazil), in June 2010.The material was identified by comparison with a specimen collected in December 2004 from "Fazenda não me deixes", in Quixadá (Ceará, Brazil).A voucher specimen has been deposited at the Herbarium Prisco Bezerra at Universidade Federal do Ceará, under the registration number 35524.

Animals
Male Swiss mice (weighing 25-30 g), 5-8 per group, were used for the tests.The animals were housed in standard environmental conditions (22 ± 1 °C, humidity 60 ± 5%, 12 h light, 12 h dark cycle) with free access to water and food.All experiments were performed in accordance with the NIH Guide for Care and Use of Laboratory Animals.

Experimental protocol
The animals were divided into six groups and received distilled water (control), APSE-Aq (1, 10, 50, 100 and/or 200 mg kg -1 ) and indomethacin (10 mg kg -1 ) or morphine (5 mg kg -1 ) intraperitoneally (ip).After 30 min of the last treatment, the animals were subjected to the following tests: abdominal writhing induced by acetic acid, the formalin test and paw edema induced by carrageenan.

Abdominal writhing induced by acetic acid
The animals received 0.6% acetic acid intraperitoneally (10 μL per g of weight).After 10 min of acetic acid administration, the number of writhings over a period of 20 min was recorded for each animal.A writhing was identified as an extension of the hind legs accompanied by constriction of the abdomen. 23

Formalin test
Mice were injected with formalin (20 μL 1% formalin) intraperitoneally under the ventral surface of the right hind paw.The amount of time spent licking the injected paw was timed with a chronometer and was considered as indicative of nociception.The initial nociceptive response peaked 5 min after formalin injection (early phase) and 20-25 min after formalin injection (late phase), representing the tonic and inflammatory pain responses, respectively. 24

Paw edema induced by carrageenan
The animals received an intraplantar injection of 1% carrageenan (100 μL) to induce the edema in the right hind paw.The volume of the paws was measured before and 1, 2, 3, 4 and 24 h after the administration of carrageenan. 25he volume of the edema, in milliliters, was measured using a pletismometer (Ugo Basile, Italy), where the right hind paw was submerged up to the tibio-tarsal joint, in the measuring chamber of the device.The volume of fluid displaced was recorded and considered the volume of the paw.The results were expressed as the difference between the volume of the paw at the specified time intervals and the volume before the carrageenan injection (t = 0).

Statistical analyses
The results are presented as the mean ± the standard error of the mean (SEM).The statistical differences between the groups were analyzed by one-way analysis of variance (ANOVA) followed by the Student-Newman-Keuls multiple comparisons test.To analyze the data from the paw edema induced by carrageenan tests, two-way ANOVA followed by a Bonferroni post hoc test was used.GraphPad software (GraphPad Software, San Diego, CA, USA) was used in these analyses.Values of p < 0.05 were considered significant.was coupling to H-3 besides the 3H-14 and 2H-15, thus suggesting that N-4 should be protonated as a consequence of the HPLC isolation methodology using TFA.In this case, the simple rotaevaporation of the eluent from the HPLC, to remove the solvent and the volatile TFA, had not been enough.Thus, compound 1 was treated with an alkali solution and then, after workup, submitted to another NMR analysis (Table 1).As expected, a general shielding of all protons was observed, particularly higher for H- 323.2123).The FT-IR spectrum exhibited a tertiary amide I band at 1646 cm -1 , as well as bands at 1177 and 1124 cm -1 attributed to C-N bonds.Absorptions at 1481 and 1379 cm -1 were assigned to symmetrical bending of methylene and methyl groups, respectively.Analysis of the CPD, attached proton test (APT) and HSQC 13 C NMR spectra (Table 3) allowed identification of 21 carbons, in agreement with the suggested molecular formula, which could be correlated with six non-hydrogenated (3 sp 2 and 3 sp 3 ), five monohydrogenated (1 sp 3 and 4 sp 2 ), nine methylene and one methyl carbons.The signal at d 171.7 (C-22), typical of carbonyl, in addition to a signal of a methyl at d 25.6 (C-23), suggested the presence of an N-acetyl group, in agreement with the band at 1646 cm -1 in the IR spectrum.The 1 H NMR spectrum (Table 3) exhibited three aromatic proton signals at d 7.59 (d, J 7.6 Hz, H-9), 7.28 (t, J 7.6 Hz, H-11) and 7.15 (t, J 7.6 Hz, H-10).Their multiplicities indicated the presence of a 1,2-disubstituted benzene ring, despite the non-detection of the fourth proton, even on the HSQC spectrum or by changing the solvent (see Supplementary Information).A singlet at d 3.99 (H-21), a broad quartet at d 3.80 (J 10.0 Hz, H-5α), a triple doublet at    3) showed considerable deshielding for the protons of the nitrogenbearing carbons of the protonated form.Similarly to that already observed for compound 1, the C-8 carbon also exhibited a higher shielding in the N-protonated molecule.A significant shielding for C-2 and C-15, of approximately 3 ppm, was also observed.

Results and Discussion
In order to explain the unexpected behavior of H-12 on the 1 H NMR spectrum of compound 5, as noticed earlier, a series of variable-temperature 1 H NMR experiments were performed.As can be noticed from Figure 4, the expected splitting pattern (two doublets and two triplets) of the four contiguous hydrogens of the aromatic system starts to rise up around 50 ºC (spectrum (a)), to be completely observed at 80 ºC (spectrum (f)).The integration of each absorption of the aromatic region at d 7.06-7.88now reads one proton.The heteronuclear multiple quantum correlation (HMQC) spectrum (see Supplementary Information Figure S39) run at 70 ºC does show the correlation of the doublet at d 7.55 with the carbon at d 116.0, and the COSY spectrum (Figure S38) now shows the complete coupling of the aromatic system.Thus, with these experiments it was proved that somehow an aromatic proton could not break through during a routine 1 H NMR experiment.
Compound 2 was identified as (+)-aspidospermine, a plumeran alkaloid, [α] D 20 +89.0 (c 0.19, CHCl 3 ) {lit. 6α] D +92}.Comparison of the 1 H and 13 C NMR data (Table 2) with those published by Zèches-Hanrot et al. 27 for the plumeran alkaloid aspidospermine showed that the chemical shifts of carbons C-8 (d 128.0) and C-13 (d 141.0) should be reversed to C-8 (d 143.7) and C-13 (d 129.8), since C-13 is ortho to the methoxy group, and so should be more shielded than C-8, that is meta.Two other observed chemical shifts need to be emphasized.9][30] The value of d 160.0, annotated for Zèches-Hanrot et al., 27 does not seem compatible, and no explanation for this difference is given.A similar strong shielding effect is observed for C-2, that in our case appears at d 69.7 and in the literature 27 at d 64.0.We did not find any reasonable theoretical argument to explain this behavior, which may suggest that the NMR data, and their assignments, should be revised.
Compound 3 was characterized as the plumeran alkaloid (+)-demethoxyaspidospermine, otherwise named N-acetylaspidospermidine, 9 [α] D 20 +25.4 (c 0.11, MeOH) {lit. 10 [α] D +10; c 0.009, CHCl 3 }. 1 H and 13 C NMR data comparison (Table 2) with those from the literature 9 revealed the chemical shift misassignments of both protons Another mistake in the literature 9 was the assignment of C-8, a substituted benzene carbon to which a value of d 109.3 is annotated versus d 134.9, experimentally observed, which is consistent with several other examples of plumeran alkaloids from the literature. 28,30-33Thus, the revision of the NMR data assignments is suggested.
Compound 4 was characterized as (-)-aspidofractinine, another plumeran alkaloid, [α] D 20 -8.35 (c 0.23, MeOH) {lit. 12[α] D -14; c 0.28, CHCl 3 }.Comparison of the 1 H and 13 C NMR data (Table 2) with those reported in the literature 11 showed that the slightly deshielded methylene at d 35.9, to which the protons at d 1.52/1.29   .All these dipolar couplings permitted to determine the β-configuration for protons H-3, H-5, the N-CH 3 , and for the carboxyl methyl group, as well as the α-configuration for proton H-15.The E geometry of the double bond was confirmed by the dipolar interaction between the protons H-15 (d 3.39) and 3H-18 (d 1.70) (Figure 5).There are two reports about the 1 H and 13 C NMR data of N-methylakuammidine/macusine A that are found in the literature. 15,16Interestingly, the chemical shifts differ about 2 ppm, and no structure assignments have been done.Comparison of the current data (Table 2) with those from Hu, Zhu and Hesse 15 showed a better compatibility.

Antinociceptive and anti-inflammatory activities
In the abdominal writhing induced by acetic acid test, 34 groups pre-treated with different doses of APSE-Aq [50 mg kg -1 (35.83 ± 2.73), 100 mg kg -1 (31.00 ± 2.58) and 200 mg kg -1 (23.25 ± 2.56)] and with indomethacin [10 mg kg -1 (16.75 ± 1.65)] exhibited a significant decrease in the number of abdominal writhes that were induced by acetic acid when compared to the group treated only with the vehicle (54.00 ± 3.20).The APSE-Aq 200 mg kg -1 group, compared to other groups, showed a better response in the reduction of writhes.The results permitted to conclude that the administration of APSE-Aq showed antinociceptive activity in an animal model of visceral pain induced by administration of acetic acid (Figure 6).In the formalin model of nociception test, 35 7).These results suggest that the aqueous fraction APSE-Aq presents possible central and peripheral effect.A similar result was found by Pereira et al. using the ethanol extract of other Aspidosperma species. 36n the paw edema induced by carrageenan test, the administration of APSE-Aq 100 mg kg -1 significantly reduced the carrageenan-induced paw edema two (p < 0.05), three (p < 0.01) and four hours (p < 0.001) after the administration of the stimulus compared to the animals treated with the vehicle.The APSE-Aq 50 mg kg -1 and 200 mg kg -1 groups reduced the paw edema three hours (APSE-Aq 50 mg p < 0.01; APSE-Aq 200 mg p < 0.001) and four hours (APSE-Aq 50 mg p < 0.001; APSE-Aq 200 mg p < 0.001) after the administration of the stimulus compared to the animals with vehicle.The indomethacin reduced the paw edema from the second hour of administration of the stimulus (p < 0.001) (Figure 8).These results suggest that APSE-Aq presents a possible anti-inflammatory activity.

Figure 1 .
Figure 1.Structures of all compounds isolated from the ethanol extract of seeds of Aspidosperma pyrifolium.

Compound 1
was isolated from APSE/Aq, after purification by HPLC.It was obtained as a brown resin, [α] D 20 -21.5 ± 0.4 (c 0.40, MeOH), and gave a positive

Figure 2 .
Figure 2. Key long-range correlations for protonated compound 1 observed through the HMBC spectrum.

Figure 3 .
Figure 3. (a) Key dipolar interactions observed in the NOESY spectrum of the deprotonated compound 1.(b) A stereoview showing the boat conformations of rings C and F.

Figure 5 .
Figure 5. Key dipolar interactions observed in the NOESY spectrum of compound 8.

Figure 7 .
Figure 7.The effect of APSE-Aq on the formalin test.The figure shows paw licking time (in seconds) at the early and late phases.Values are expressed as mean ± SEM of the number of observations.(a) vs. control; (b) vs. APSE-Aq 1 mg kg -1 ; (c) vs. APSE-Aq 10 mg kg -1 , respectively; at p < 0.0001 (one-way ANOVA followed by the Newman-Keuls pos hoc test).Indo: indomethacin.

Table 3 .
13 and13C NMR (CD 3 OD) data assignments for the protonated and deprotonated forms of compound 5 a Data were recorded at 75 MHz; b data were recorded at 500 MHz; c data were recorded at 125 MHz.n.d.: not detected.