CHEMISTRY & BIODIVERSITY – Vol. 8 (2011) 155 Tropane Alkaloids from Erythroxylum caatingae Plowman by Steno L. de Oliveira a ), Josean F. Tavares a ), Marianna V. S. Castello Branco a ), Hellane F. S. Lucena a ), José M. Barbosa-Filho a ), Maria de F. Agra a ), Silene C. do Nascimento b ), Jaciana dos S. Aguiar b ), Teresinha G. da Silva b ), Carlos A. de Simone c ), João X. de Araújo-Júnior c ) d ), and Marcelo S. da Silva* a ) a ) Departamento de Ciências Farmacêuticas, Laboratório de Tecnologia Farmacêutica, Universidade Federal da Paraba, C. P. 5009, 58051-970, João Pessoa, Paraba, Brazil (phone: þ 55 83 3216 7427; fax: þ 55 83 3216 7365; marcelosobral@ltf.ufpb.br) b ) Departamento de Antibióticos, Universidade Federal de Pernambuco, 50670-901, Recife, PE, Brazil c ) Instituto de Qumica e Biotecnologia, Universidade Federal de Alagoas, 57072-970, Maceió, AL, Brazil d ) ESENFAR, Universidade Federal de Alagoas, 57072-970, Maceió, AL, Brazil Three tropane alkaloids, 1 – 3, were isolated from Erythroxylum caatingae, i.e., 6b-benzoyloxy-3a[(4-hydroxy-3,5-dimethoxybenzoyl)oxy]tropane (1), a new tropane alkaloid, along with the known alkaloids 3a,6b-dibenzoyloxytropane (2) and 6b-benzoyloxy-3a-[(3,4,5-trimethoxybenzoyl)oxy]tropane (catuabine B; 3). Their structures were determined by 2D- (1H and 13C) NMR. By LC/ESI-MS/MS analysis of the fractions of alkaloids 1 – 3, it was possible to detect five more alkaloids, 4 – 8, two of these, 4 and 8, possibly being new natural products. X-Ray crystallography of the chloride derivate of 1, i.e., 6bbenzoyloxy-3a-(4-hydroxy-3,5-dimethoxybenzoyloxy)tropane hydrochloride (1a) confirmed the structure of 1. Cytotoxicity was tested against the cell lines HEp-2, NCI-H292, and KB for the MeOH extract and alkaloid 3, and antitumor activity was tested against Sarcoma 180 only for the MeOH extract. Introduction. – The family Erythroxylaceae comprises four genera and ca. 240 species with pantropical distribution, with its main centers of diversity and endemism in Venezuela, Brazil, and Madagascar. The Erythroxylaceae consist of the genera Aneulophus, Erythroxylum, Nectaropetalum, and Pinacopodium, where Erythroxylum is the largest and most representative genus, with ca. 230 species and wide distribution in the tropical regions of the world, and South America is the center of diversity and endemism. Erythroxylum caatingae is a species with distribution restricted to the Northeast Brazil, being only recorded for the states of Bahia, Ceara, Paraba, Pernambuco, and Rio Grande do Norte [1]. Tropane alkaloids occur frequently in the families Convolvulaceae, Erythroxylaceae, Proteaceae, Rhizophoraceae, and Solanaceae and have occasionally been reported in plants of the families Brassicaceae, Euphorbiaceae, and Oleaceae [2]. Recently, 17 tropane alkaloids were reported from E. vacciniifolium, collected in Paraba (Brazil) [3] [4]. Studies with the CHCl3 extract of E. pervillei and E. rotundifolium revealed cytotoxicity against the multidrug-resistant (MDR) cell line KB-V1 in the presence of vinblastine [2 – 5]. Recently, we described a new trachylobane diterpene and its cytotoxicity against V79 cells and rat hepatocytes [6]. As part of a continuing investigation of new bioactive molecules from plants of Paraba (Brazil), we describe here the isolation and structural identification of three tropane alkaloids,  2011 Verlag Helvetica Chimica Acta AG, Zrich 156 CHEMISTRY & BIODIVERSITY – Vol. 8 (2011) among which 6b-benzoyloxy-3a-(4-hydroxy-3,5-dimethoxybenzoyloxy)tropane (1) is a new tropane alkaloid, whereas 3a,6b-dibenzoyloxytropane (2) and 6b-benzoyloxy-3a(3,4,5-trimethoxybenzoyloxy)tropane (catuabine B; 3) are known compounds. The hydrochloride of 1, i.e., 1a, was obtained by acidification with HCl. Furthermore, it was possible to detect, in the fractions of alkaloids 1 – 3, five additional alkaloids, 4 – 8, partially characterized by LC/ESI-MS/MS. Cytotoxicity of the MeOH extract and alkaloid 3 against HEp-2, NCI-H292, and KB cells, and in vivo antitumor activity of the MeOH extract of E. caatingae in the presence of methotrexate were determined. Results and Discussion. – Compound 1 was isolated as an amorphous solid, with a melting point of 220 – 2238. HR-MS exhibited a molecular ion at m/z 442.1848 ([M þ H] þ ; calc. 442.1866), compatible with the molecular formula C24H27NO7. The IR spectrum showed absorptions at 3425 (OH), and 1724 and 1277 cm  1 characteristic of an ester group. The 13C-NMR spectrum of 1 showed signals at d(C) 60.1 (C(1)), 65.7 (C(5)), and 40.1 (MeN), characteristics of a tropane skeleton. In addition, the signals at d(C) 67.3 and 79.7 are compatible with the tropane skeleton substituted at C(3) and C(6) [7]. The HSQC spectrum showed direct correlations between d(C) 60.1 (C(1)) and d(H) 3.45 (HC(1)), and 65.7 (C(5)), and 3.43 (HC(5)). The presence of two COO groups was supported by the signals at d(C) 165.5 and 166.0. The 1H-NMR CHEMISTRY & BIODIVERSITY – Vol. 8 (2011) 157 spectra revealed signals of two O-bearing CH groups at d(H) 5.34 (HbC(3)) and 5.93 (HaC(6)), two O-bearing CH groups at d(H) 3.45 (HC(1)) and 3.43 (HC(5)), three O-bearing CH2 groups at d(H) 2.29 and 2.37 (HaxC(2) and HaxC(4), resp.), 1.82 and 2.03 (HeqC(2) and HeqC(4), resp.), and 2.33 and 2.82 (HaC(7) and HbC(7), resp.), and a MeN group at d(H) 2.46. The complete assignment of all the Cand H-atoms is compiled in Table 1. The relative configuration of 1 was established by the signal at d(H) 5.34 (br. t, HC(3)) with a coupling constant of J ¼ 5.0 Hz indicating the a-orientation of the substituent at C(3) [3]. On the other hand, the relative configuration of the substituent at C(6) was established by the analysis of multiplicity and of the coupling constants of HC(6). HC(6) (d(H) 5.93) showed two couplings (dd, J ¼ 3.0, 7.5) with the two H-atoms with signals at d(H) 2.82 and 2.33 (HaC(7) and HbC(7), resp.) and did not show any coupling with the signal of HC(5) (d(H) 3.43). This observation is in line with the b-orientation of the substituent, and a dihedral angle of 908 between H-atoms HC(5) and HaC(6) [3]. These data were corroborated by Xray crystallography of 1a (Figs. 1 and 2). The chemical shifts d(H) 8.01 (dd, J ¼ 1.0, 7.5, HC(2’’,6’’)), 7.43 (t, J ¼ 7.5, HC(3’’,5’’)), and 7.55 (t, J ¼ 7.5, HC(4’’)) are characteristic of a benzoyloxy (BzO) group. The chemical shifts at d(H) 7.32 (s, HC(2’,6’)) and Table 1. NMR Data for Compound 1 a ) Position 1 2ax 2eq 3b 4ax 4eq 5 6a 7a 7b 1’ 2’ 3’ 4’ 5’ 6’ 7’ 1’’ 2’’ 3’’ 4’’ 5’’ 6’’ 7’’ MeN MeO a d(C ) 60.12 34.59 67.28 33.25 65.74 79.69 36.73 121.15 106.55 146.97 139.60 146.97 106.55 165.46 130.28 129.46 128.34 132.92 128.34 129.46 166.00 40.07 56.50 d( H) HMBC 3.45 (m) 2.29 (m) 1.82 (br. d, J ¼ 15.0) 5.34 (br. t, J ¼ 5.0) 2.37 (m) 2.03 (br. d, J ¼ 15.0) 3.43 (m) 5.93 (dd, J ¼ 3.0, 7.5) 2.82 (dd, J ¼ 7.5, 14.0) 2.33 (m) Me HC(7) 7.32 (s) 7.32 (s) 8.01 (dd, J ¼ 1.0, 7.5) 7.43 (t, J ¼ 7.5) 7.55 (t, J ¼ 7.5) 7.43 (t, J ¼ 7.5) 8.01 (dd, J ¼ 1.0, 7.5) 1 H,1H-COSY HC(1) HC(2)/HC(4) Me/HC(3)/HC(7) HC(7) HC(4) HC(2’)/HC(6’) HC(6’) MeO/HC(2’) MeO/HC(6’) HC(2’) HC(3) HC(3’’)/HC(5’’) HC(4’’)/HC(6’’) HC(5’’) HC(2’’)/HC(6’’) HC(3’’) HC(2’’)/HC(4’’) HC(2’’)/HC(6’’) 2.46 (s) 4.00 (s) ) CDCl3 ; at 500 MHz for 1H-NMR and 125 MHz for 13C-NMR. HC(3’) HC(2’) HC(6’) HC(5’) 158 CHEMISTRY & BIODIVERSITY – Vol. 8 (2011) Fig. 1. An ORTEP3 projection of the molecule 1a, showing the atom numbering and displacement ellipsoids at the 50% probability level Fig. 2. H-Bonding interactions in the crystal packing CHEMISTRY & BIODIVERSITY – Vol. 8 (2011) 159 4.00 (2 MeO) are characteristic of a 3’,5’-dimethoxybenzoyloxy (Hdmb) group. The HMBC spectrum showed correlations between C(3’) and C(5’) (d(C) 147.0) and MeO groups (d(H) 4.00). Therefore, compound 1 was identified as 6b-benzoyloxy-3a-(4hydroxy-3,5-dimethoxybenzoyloxy)tropane, a new tropane alkaloid. Compound 1a was obtained as white crystals, with a melting point of 196 – 1988. Analysis of the 1H- and 13C-NMR data showed chemical shifts compatible with a 3,6disubstituted tropane skeleton. Unlike alkaloid 1, derivative 1a showed chemical shifts for HC(1) at d(H) 4.11 and HC(5) at d(H) 4.12. These chemical shifts are compatible with those of cocaine hydrochloride [8] (Table 2). The structure of 1a was established by X-ray crystallography (Figs. 1 and 2). Table 2. NMR Data for Compounds 1a, 2, and 3 a ) Position 1a d(C ) 1 2ax 2eq 3b 4ax 4eq 5 6a 7a 7b 2 d( H ) 67.18 4.11 (br. s) 33.02 2.40 (br. d, J ¼ 15.5) 64.48 5.46 (br. t, J ¼ 4.5) 34.14 2.20 (br. d, J ¼ 16.0) 63.10 4.12 (br. s) 35.40 6.09 (dd, J ¼ 3.0, 8.0) 74.85 3.19 (dd, J ¼ 8.0, 15.0) 2.51 (m) d(C ) 119.75 106.58 7.35 (s) 147.12 140.25 164.98 1’’ 2’’, 6’’ 3’’, 5’’ 4’’ 7’’ MeN m-MeO p-MeO 128.51 129.44 128.67 133.89 165.10 40.34 a 3.10 (s) d(C ) Tmb 125.38 106.63 7.39 (s) 153.13 153.13 165.32 Bz 130.37 129.48 128.32 132.89 166.37 40.11 8.03 (dd, J ¼ 1.5, 8.0) 7.43 (t, J ¼ 8.0) 7.53 (m) 2.60 (s) 56.41 4.00 (s) d( H ) 60.03 3.43 (m) 34.66 2.26 (m) 1.83 (br. d, J ¼ 15.0) 67.65 5.34 (br. t, J ¼ 5.0) 33.31 2.33 (m) 2.03 (br. d, J ¼ 15.0) 65.74 3.42 (m) 79.84 5.92 (dd, J ¼ 3.0, 7.0) 36.69 2.80 (dd, J ¼ 7.0, 15.0) 2.30 (m) Bz 130.41 129.50 8.11 (dd, J ¼ 1.5, 8.0) 128.58 7.49 (t, J ¼ 8.0) 133.00 7.58 (m) 165.70 Bz 7.98 (br. d, J ¼ 7.5) 7.49 (t, J ¼ 7.5) 7.63 (t, J ¼ 7.0) d( H ) 59.92 3.42 (m) 34.65 2.26 (m) 1.82 (br. d, J ¼ 15.0) 67.57 5.34 (br. t, J ¼ 5.0) 33.16 2.29 (m) 2.06 (br. d, J ¼ 15.0) 65.85 3.39 (br. s) 80.10 5.87 (dd, J ¼ 3.0, 7.5) 36.16 2.78 (dd, J ¼ 7.5, 14.5) 2.33 (m) Hdmb 1’ 2’, 6’ 3’, 5’ 4’ 7’ 3 Bz 130.31 129.48 128.35 132.94 166.05 40.16 56.31 60.89 8.02 (dd, J ¼ 1.5, 8.5) 7.42 (t, J ¼ 8.0) 7.55 (t, J ¼ 7.5) 2.60 (s) 3.98 (s) 3.92 (s) ) CDCl3 ; at 500 MHz for 1H-NMR and 125 MHz for 13C-NMR. Compound 2 was isolated as white crystals, with a melting point of 119 – 1218. The MS showed a molecular-ion peak of m/z 366.1 ([M þ H] þ ), compatible with the molecular formula C22H23NO4 . The 1H- and 13C-NMR spectra exhibited chemical-shift values similar to those for 1 (Table 1), the chemical shifts of which are in line with a tropane skeleton disubstituted at C(3) and C(6). The principal difference between the 160 CHEMISTRY & BIODIVERSITY – Vol. 8 (2011) NMR data of compounds 1 and 2 is the type of substituent at C(3). The signals at d(H) 8.11 (dd, J ¼ 1.5, 8.0, HC(2’,6’)), 7.49 (t, J ¼ 8.0, HC(3’,5’)), and 7.58 (t, J ¼ 7.5, HC(4’)) are compatible with a BzO group at C(3) (Table 2). The HMBC spectrum showed long-range correlations between signals at 165.7 with 5.34 (HbC(3)), and 166.4 with 5.87 (HC(6)), defining the chemical shifts of the CO groups C(7’) and C(7’’), respectively. Thus, compound 2 was identified as 3a,6b-dibenzoyloxytropane. Compound 3 was isolated as white crystals, with a melting point of 154 – 1578. The MS showed a molecular-ion peak of m/z 456.1 ([M þ H] þ ), compatible with the molecular formula C25H29NO7. The 1H- and 13C-NMR spectra of this alkaloid also displayed chemical-shift values similar to those for alkaloids 1 and 2, but the signals at 7.39 (s, 2 H), 3.98 (s, 6 H), and 3.92 (s, 3 H) indicated that the substituent at C(3) is a 3,4,5-trimethoxybenzoyloxy group (Table 2). The HMBC spectrum showed correlations between the signals of C(7’) (d(C) 165.3) and HC(3) (d(H) 5.34), and between the signals of C(7’’) (d(C) 166.1) and HC(6) (d(H) 5.92). The relative configuration of 3 was assigned as in compound 1. Thus, the structure of compound 3 is 6b-benzoyloxy3a-(3,4,5-trimethoxybenzoyloxy)tropane (catuabine B). The 1H- and 13C-NMR spectra of the fractions of alkaloids 1 – 3 showed the presence of other constituents in smaller proportions. In an attempt to identify these constituents, these fractions were analyzed by LC/ESI-MS/MS (Table 3). Supported by the fragmentation model of alkaloid 1 (Scheme), it was possible to partially identify alkaloids 4 – 8. The selective MS2 of the three fractions analyzed showed ions at m/z 140.1 and 122.2, suggesting the presence of disubstituted tropane alkaloids [9]. The chromatogram of total ions of fraction of 1, showed three peaks, with the molecular-ion peaks at m/z 420.1, 442.1, and 366.1, where the largest component ion, with the peak at m/z 442.1, was identified as alkaloid 1. The MS2 experiment of the ion with the peak at m/z 420.1 resulted in the formation of ions with peaks at m/z 320.2 and 222.1, among others. These fragments [420.1 ! 320.2] and [420.1 ! 222.1] are compatible with the loss of the Me2CCHCO and Hdmb substituents from the tropane ring, respectively, suggesting the molecular formula C22H29NO7 for compound 4. Similarly, for fraction of 2, the chromatogram of total ions showed three peaks with molecular-ion peaks at m/z 278.1, 262.1, and 366.1, where the largest component ion, with the peak at m/z 366.1, was identified as alkaloid 2. The MS2 experiment of the ion with the peak at m/z 278.1 resulted in the formation of ions with peaks at m/z 156.1 and 138.1, among others. These fragments [278.1 ! 156.1 ! 138.1] are compatible with the loss of the BzO and OH substituents, respectively, suggesting the molecular formula C15H19NO4 for compound 5. The MS2 experiment of ion with the peak at m/z 262.1 resulted in the formation of ions with peaks at m/z 140.1 and 122.2, among others. These fragments [262.1 ! 140.1 ! 122.1] are compatible with the loss of the BzO and OH substituents, suggesting the molecular formula C15H19NO3 for compound 6. The chromatogram of total ions of fraction of 3, showed four peaks with molecular-ion peaks at m/z 262.1, 352.1, 426.1, and 456.1, where the largest component ion, with the peak at m/z 456.1, was identified as alkaloid 3. The MS2 experiment of the ion with the peak at m/z 352.1 resulted in the formation of ions with peaks at m/z 195.0 and 140.2, among others. The fragment [352.1 ! 140.2] is compatible with the loss of the substituent Tmb, thus suggesting the molecular formula C10H11NO4 for compound 7. The MS2 experiment of the ion with the peakt at m/z 426.1, resulted in the formation of ions with peaks at m/z 304.1 and 122.2, CHEMISTRY & BIODIVERSITY – Vol. 8 (2011) 161 among others. These fragments [426.1 ! 304.1 ! 122.2] are compatible with the loss of the BzO and Dmb substituents, suggesting the molecular formula C24H27NO6 for compound 8. Thus, of the alkaloids only partially characterized, 4 and 8 are possibly new natural products. Scheme. Proposed Fragmentation of Alkaloid 1 Table 3. LC/MS/MS Data for Each Ion Analyzed and Its Fragmentations Alkaloid Peak No. Mþ MS2 Fragments 1 1 2 3 420 442 366 320, 222, 181, 140, 122 320, 244, 181, 140 244, 122 2 1 2 3 278 262 366 156, 138, 94 140, 122, 91 244, 122 3 1 2 3 4 262 352 426 456 140, 105 195, 140, 122 244, 165, 140, 122 244, 195, 167 162 CHEMISTRY & BIODIVERSITY – Vol. 8 (2011) The cytotoxic activities of the MeOH extract of Erythroxylum caatingae stems and alkaloid 3 were evaluated against human cancer cell lines (HEp-2, NCI-H292, and KB). Alkaloid 3 exhibited significant cytotoxicity only against NCI-H292 (IC50 ¼ 50 mg/ ml), and the MeOH extract of E. caatingae stems was inactive against all cell lines tested (IC50 > 50 mg/ml). The effects of MeOH extract of E. caatingae stems on mice transplanted with Sarcoma 180 are shown in Table 4. There was a significant reduction in tumor weight in animals treated with different doses. These reductions gave tumor inhibition percentages of 59.4 to 66.4%. The MeOH extract tested showed IC50 values greater than 10 mg/ml for all tumor cell lines tested, suggesting that the in vivo anticancer properties were not related to direct antiproliferative effects. Table 4. In vivo Antitumor Activity against Sarcoma 180 Using MeOH Extract of Erythroxylum caatingae Stems a ) Compound Dose [mg/kg] Weight of tumor [g] Inhibition [%] MeOH extract of stems 100 200 400 2.5 – 0.698  0.05 0.577  0.10 0.614  0.11 0.659  0.08 1.718  0.20 59.4 66.4 64.3 61.7 – Methotrexate Control a ) Data are presented as means  SEM for seven animals. *: P < 0.05 compared with control group. The authors thank the Conselho Nacional de Desenvolvimento Cientfico e Tecnológico (CNPq), CoordenaÅão de AperfeiÅoamento Pessoal de Nvel Superior (CAPES) for financial support, Vicente Carlos de O. Costa (UFPB) for the NMR spectra, and Norberto P. Lopes (Faculdade de Ciências Farmacêuticas da USP de Ribeirão Preto) for recording the mass spectra. The Escola de Enfermagem e Farmácia/Instituto de Qumica e Biotecnologia (UFAL) helped carry out the X-ray crystallography. Mr. P. Wehrung of the Plateforme de Chimie Biologique Intégrative de Strasbourg (PCBiS) IFR 85 and Mrs. M. Schmitt, Laboratoire de Pharmacochimie, Faculté de Pharmacie, 74 route du Rhin, F-67400 Illkirch, helped to perform the LC/ESI-MS/MS. Experimental Part General. M.p.: Microquimica digital melting-point apparatus, model MQAPF-302, with the Pt block in a REICHERT Kofler-type light microscope, model R3279, with a temp. that varies from 0 to 3508; values uncorrected. IR Spectra: BOMEM SERIE 100 MB spectrometer, in the range of 4000 – 400 cm  1, with KBr pellets (0.5 mg of the sample/100 mg of KBr). NMR Spectra: VARIAN-NMR SYSTEM spectrometer, operating at 500 MHz for 1H and at 125 MHz for 13C; recorded in CDCl3 , with TMS as internal standard. LC/ESI-MS/MS: HPLC Agilent 1200RRLC and Bruker HCT Ultra mass spectrometer. X-Ray crystallography: Enraf-Nonius Kappa-CCD diffractometer. Plant Material. The stems of Erythroxylum caatingae were collected in Picui, Paraba, Brazil. The botanical material was identified by M. F. A. A dried specimen was deposited with the Herbario Prof. L. P. Xavier (JPB), Universidade Federal da Paraba, under the identification label AGRA 5666. Extraction and Isolation. The crude MeOH extract (500 g) was dissolved in H2O and defatted with hexane. The defatted aq. extract was acidified with 3% HCl by mechanical mixing and filtered through Celite, yielding a residue that was discarded and an acidic soln. This acidic soln. was extracted with CHCl3 (3  500 ml) resulting in an acidic CHCl3 phase and an aq. phase that was neutralized to pH 7.0 with NH4Cl. The aq. phase at pH 7.0 was then extracted with CHCl3 to yield an aq. phase and basic CHCl3 CHEMISTRY & BIODIVERSITY – Vol. 8 (2011) 163 phase FA1 (4.0 g), which was submitted to column chromatography (CC) utilizing SiO2 as the stationary phase, and CHCl3 and MeOH as the eluents alone or in binary mixtures with increasing polarity. The result were 55 fractions of 100 ml. The 55 fractions were monitored by TLC, eluted with various solvent systems (CHCl3 and/or CHCl3/MeOH in order of increasing polarity) in chambers pre-saturated with NH3 vapor, revealed with Dragendorffs reagent, and placed in 21 groups based on their Rf values. Frs. 25, 45, and 46 were submitted to repeated recrystallization with acetone and Et2O, to yield compounds 1, 2, and 3. Compound 1 was acidified with HCl to yield the alkaloid hydrochloride, referred to as 1a. 6b-Benzoyloxy-3a-(4-hydroxy-3,5-dimethoxybenzoyloxy)tropane ( ¼ (3S,6S)-8-Methyl-6-[(phenylcarbonyl)oxy]-8-azabicyclo[3.2.1]oct-3-yl 4-Hydroxy-3,5-dimethoxybenzoate; 1). White crystals. IR (KBr): 3425 (OH), 1724 (C¼O), 1277. 1H- and 13C-NMR: see Table 1. ESI-MS: 442 ([M þ H] þ ), 320 ([M  BzOH] þ ), 244 ([M  HdmbOH] þ ), 181 [M  BzOH  C8H13ON] þ ), 140 ([M  HdmbOH  C7H5O] þ ). 3a,6b-Dibenzoyloxytropane ( ¼ (3S,6S)-8-Methyl-8-azabicyclo[3.2.1]octane-3,6-diyl Dibenzoate; 2). White crystals. IR (KBr): 1724 (C¼O), 1600 – 1585 (C¼C). 1H- and 13C-NMR: see Table 2. ESI-MS: 366 ([M þ H] þ ), 244 ([M  BzOH] þ ), 122 ([M  BzOH  C8H14N] þ ). Catuabine B ( ¼ 6b-Benzoyloxy-3a-(3,4,5-trimethoxybenzoyloxy)tropane ¼ (3S,6S)-8-Methyl-6[(phenylcarbonyl)oxy]-8-azabicyclo[3.2.1]oct-3-yl 3,4,5-Trimethoxybenzoate; 3). White crystals. IR (KBr): 1709 (C¼O), 1600 – 1585 (C¼C), 1281. 1H- and 13C-NMR: see Table 2. ESI-MS: 456 ([M þ H] þ ), 334 ([M  BzOH] þ ), 244 ([M  BzOH  3 MeO] þ ), 195 ([M  BzOH  C8H14N] þ ). X-Ray Crystallographic Analyses 1). X-Ray diffraction data collections were performed on an EnrafNonius Kappa-CCD diffractometer (95-mm CCD camera on k-goniostat) using graphite monochromated MoKa radiation (0.71073 ), at r.t. Data collections were carried out using the COLLECT software [10] up to 508 in 2q. Final unit-cell parameters were based on 9048 reflections. Integration and scaling of the reflections, and correction for Lorentz and polarization effects were performed with the HKL DENZO-SCALEPACK system of programs [11]. The structure of the compound was solved by direct methods with SHELXS-97 [12]. The models were refined by full-matrix least squares on F 2 using SHELXL-97 [13]. The program ORTEP-3 [14] was used for graphic representation and the program WINGX [15] to prepare materials for publication. All H-atoms were located by geometric considerations placed (d(CH) ¼ 0.93 – 0.98 ) and refined as riding with Uiso(H) ¼ 1.5 Ueq(Cmethyl) or 1.2 Ueq(other). An ORTEP-3 diagram of the molecule is shown in Fig. 1, and Table 5 contains the main crystallographic parameters. The compound crystallized with one Cl-atom that forms NH1N · · · Cl1i and O3’H3 · · · Cl1ii Hbonding interactions, where: [i ¼ x þ 1/2, y þ 1/2, z; ii ¼ x, y, z] and H1N · · · Cl1i ¼ 2.065(2)  ; NH1N · · · Cl1i ¼ 1788 and H3 · · · Cl1ii ¼ 2.477(1) ; O3’H3 · · · Cl1ii ¼ 1408 (Fig. 2). Cytotoxicity Assay. NCI-H292 (human lung mucoepidermoid carcinoma cell line), HEp-2 (human larynx epidermoid carcinoma cell line), and KB (human mouth epidermoid carcinoma cell line) cells were obtained from the Adolph Lutz Institute (São Paulo, Brazil) and were maintained in DMEM ( ¼ Dulbeccos Modified Eagles Medium) supplemented with 10% fetal bovine serum (FBS), 1000 IU/ml of penicillin, and 250 mg/ml of streptomycin, and 1% of 200 mm glutamine at 378 with 5% CO2 . Cytotoxicity was evaluated with the colorimetric MTT ( ¼ 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide) assay [16] [17]. Cell suspensions were diluted to 105 cells/ml and were distributed in 96-well culture plates (220 ml in each well), which were incubated for 24 h at 378 in a humidified incubator with 5% CO2 . After 24 h, 22 ml of MeOH extract of Erythroxylum caatingae stems and alkaloid 3 were added, and the plates were incubated again at 378. At the end of this period, the culture medium with excess MTT was removed, and 100 ml of DMSO were added to each well to dissolve the formazan crystals. The optical density (OD) of the wells was measured at 595 nm with an ELISA plate reader and compared to the control. 1) Crystallographic data have been deposited with the Cambridge Crystallographic Data Center as supplementary publication No. CCDC-711367. Copies of available material can be obtained, free of charge, upon request through the Director, CCDC, 12 Union Road, Cambridge CH21EZ, UK (fax: þ 44-1223-336-033 or e-mail: deposit@ccdc.cam.ac.uk or http://www.ccdc.cam.ac.uk). 164 CHEMISTRY & BIODIVERSITY – Vol. 8 (2011) Table 5. Crystal Data and Structure Refinement for Compound 1a Empirical formula Formula weight [g mol  1] Temp. [ K] Crystal dimensions [mm] Crystal system Space group Unit cell dimensions a [] b [] c [] b [8] V [3 ] Z L( MoKa ) radiation [] pcalc [ Mg m  3 ] m( MoKa ) [mm  1] F 000 q Range for data collection [8] Index range Reflections collected Independent reflections [ Rint ] Reflections with I > 2s( I) Number of parameters refined R [ F 2 > 2s( F 2 )] Goodness-of-fit on F 2 Residual electron density [e A  3 ] C24H28ClNO7 477.92 295(2) 0.19  0.17  0.13 Monoclinic C2 22.5640(13) 14.8870(5) 8.3820(5) 97.135(2) 2793.8(2) 4 0.71073 1.136 0.174 1008 2.5 – 27.5  23  h  29,  15  k  19,  8  l  10 9939 5809 [0.024] 4374 299 0.085 1.07 0.93 Animals. Female Swiss albino mice (Mus musculus), 60-d-old and weighing 25  5 g, were obtained from the animal house of the Departamento de Antibioticos, UFPE, Brazil. The animals were housed under standard environmental conditions of temp., humidity, and under a light-dark cycle of 12 h. The mice were fed animal house diet (LABINE Purina, Brazil) and given water ad libidum. The animals were treated according to the ethics principles of animal experimentation of COBEA (Colegio Brasileiro de Experimentacão Animal), Brazil. The Animal Studies Committee of the Universidade Federal de Pernambuco approved the experimental protocols (No. 23076.012173/2007-77). Antitumor Assay. Sarcoma 180 tumor cells were maintained in the peritoneal cavities of the Swiss mice in the Laboratorio de Bioensaios para Pesquisa de Farmacos from the Departamento de Antibioticos, UFPE, Brazil. Ascitic tumor cells (suspension of 5  106 cells) were injected subcutaneously in the axillary region of healthy animals previously weighed, and the mice were divided into experimental groups (n ¼ 7) [18]. Twenty-four hours after inoculation, MeOH extract of E. caatingae stems (100, 200, and 400 mg/kg) was dissolved in normal saline soln. with cremophor EL (2%) and administered intraperitoneally for 7 d in mice transplanted with Sarcoma 180 tumor. Methotrexate (2.5 mg/kg) was used as the positive control. The negative control received the vehicle only. On day 8, the mice were weighed and euthanized. The tumors were dissected and weighed. The percentage of tumor inhibition was calculated in accordance with [19] as: TWI ¼ C  T/C  100, where C is the weight of the control tumor and T is the weight of the treated tumor. Statistical Analysis. Data are reported as means  standard error of the mean (SEM) of n experimental animals. Statistical differences were compared by analysis of variance (ANOVA) followed by Tukeys multiple comparison test. Differences were considered statistically significant if P < 0.05. CHEMISTRY & BIODIVERSITY – Vol. 8 (2011) 165 REFERENCES [1] M. I. B. Loiola, M. F. Agra, G. S. Baracho, R. T. de Queiroz, Acta Bot. Bras. 2007, 21, 473. [2] D. Chávez, B. Cui, H.-B. Chai, R. Garca, M. Meja, N. R. Farnsworth, G. A. Cordell, J. M. Pezzuto, A. D. Kinghorn, J. Nat. Prod. 2002, 65, 606. [3] B. Zanolari, D. Guilet, A. Marston, E. F. Queiroz, M. Q. Paulo, K. Hostettmann, J. Nat. Prod. 2003, 66, 497. [4] B. Zanolari, D. Guilet, A. Marston, E. F. Queiroz, M. Q. Paulo, K. Hostettmann, J. Nat. Prod. 2005, 68, 1153. [5] G. L. Silva, B. Cui, D. Chávez, M. You, H.-B. Chai, P. Rasoanativo, S. M. Lynn, M. J. ONeill, J. A. Lewis, J. M. Besterman, A. Monks, N. R. Farnsworth, G. A. Cordell, J. M. Pezzuto, A. D. Kinghorn, J. Nat. Prod. 2001, 64, 1514. [6] J. F. Tavares, K. F. Queiroga, M. V. B. Silva, M. F. F. M. Diniz, J. M. B. Filho, E. V. L. da-Cunha, C. A. de Simone, J. X. de A. Júnior, P. S. Melo, M. Haun, M. S. da Silva, J. Nat. Prod. 2006, 69, 960. [7] S. L. Oliveira, M. S. da Silva, J. F. Tavares, J. G. Sena-Filho, H. F. S. Lucena, M. A. V. Romero, J. M. Barbosa-Filho, Chem. Biodiversity 2010, 7, 302. [8] J. F. Muhtadi, A. A. Al-Badr, in Analytical Profiles of Drug Substances, Ed. K. Florey, Academic Press, New York, 1986, Vol. 15, pp. 151 – 229. [9] B. Zanolari, J.-L. Wolfender, D. Guilet, A. Marston, E. F. Queiroz, M. Q. Paulo, K. Hostettmann, J. Chromatogr. A 2003, 1020, 75. [10] Enraf-Nonius COLLECT, Nonius BV, Delft, The Netherlands, 1997 – 2000. [11] Z. Otwinowski, W. Minor, in Methods in Enzymology, Eds. C. W. Carter, R. M. Sweet, Academic Press, New York, 1997, Vol. 276, pp. 307 – 326. [12] G. M. Sheldrick, SHELXS-97, Program for Crystal Structure Resolution, University of Gçttingen, Gçttingen, 1997. [13] G. M. Sheldrick, SHELXL-97, Program for Crystal Structure Refinement, University of Gçttingen, Gçttingen, 1997. [14] L. J. Farrugia, J. Appl. Crystallogr. 1997, 30, 565. [15] L. J. Farrugia, J. Appl. Crystallogr. 1997, 32, 837. [16] T. Mosmann, J. Immunol. Methods 1983, 65, 55. [17] M. C. Alley, D. A. Scudiero, A. Monks, M. L. Hursey, M. J. Czerwinski, D. L. Fine, B. J. Abbott, J. G. Mayo, R. H. Shoemaker, M. R. Boyd, Cancer Res. 1988, 48, 589. [18] R. M. Ribeiro-Costa, A. J. Alves, N. P. Santos, S. C. Nascimento, E. C. P. GonÅalves, N. H. Silva, N. K. Honda, N. S. Santos-Magalhães, J. Microencapsul. 2004, 21, 371. [19] Z. Machón, L. Kuczyński, J. Giełdanowski, Z. Wieczorek, M. Zimecki, B. Błaszczyk, M. Mordarski, J. Wieczorek, L. Fiszer-Maliszewska, Arch. Immunol. Ther. Exp. 1981, 29, 217. Received December 1, 2009