Curare Alkaloids: Constituents of a Matis Dart Poison - Journal of

Article pubs.acs.org/jnp

Curare Alkaloids: Constituents of a Matis Dart Poison Gonzalo R. Malca Garcia,† Lothar Hennig,† Irina V. Shelukhina,‡ Denis S. Kudryavtsev,‡ Rainer W. Bussmann,§ Victor I. Tsetlin,*,‡ and Athanassios Giannis*,† †

Institut für Organische Chemie, Fakultät für Chemie und Mineralogie, Universität Leipzig, Johannisallee 29, 04103 Leipzig, Germany Department of Molecular Basis of Neurosignaling, Shemyakin-Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences, 117997 Moscow, Russia § William L. Brown Center, Missouri Botanical Garden, St. Louis, Missouri 63166-0299, United States ‡

S Supporting Information *

ABSTRACT: A phytochemical study of dart and arrow poison from the Matis tribe led to the identification of D-(−)-quinic acid, L-malic acid, ethyldimethylamine, magnoflorine, and five new bisbenzyltetrahydroisoquinoline alkaloids (BBIQAs), 1−5. dTubocurarine could not be identified among these products. BBIQA (3) contains a unique linkage at C-8 and C-11′. All structures were characterized by a combination of NMR and HRESIMS data. The effects of Matis poison and individual BBIQAs (1−3) on rat muscle nAChR expressed in Xenopus oocytes have been investigated using the two-electrode voltage clamp technique.

D

used around the globe.11 Introduction of curare in clinical medicine in the first half of the last century represented a major breakthrough in the areas of anesthesia and surgery.12−14 The use of curare permitted the development of the concept of the so-called “balanced anaesthesia”,15 in which the three components of general anesthesia, i.e., unconsciousness, analgesia, and muscle relaxation, are provided by separate drugs, e.g., thiopental, nitrous oxide, opioids, and curare. The Matis tribe, also known as the Jaguar People, is a hunting tribe originating from western Brazil. They are specialist hunters known to use poison from frogs (Kampo) and many toxic plants for this purpose.16,17 The tribe employs mainly blowpipes, the construction of which is rather complex. For the darts thin palm stems are tipped with poison extracted from a yet unknown vine. The vine is scraped with a stick embedded with monkey teeth, and the mixture is then boiled and concentrated for approximately 2 weeks. In the Matis tribe, some of these poisons have also been employed as ingredients in or are sources of traditional remedies.17 Arrow and ordeal poisons are still considered as conventional natural sources for future drug discovery.18 Moreover, they have been and still are an important source of pharmacological tools, as well as a valuable source of original chemical structures for the development of new drugs. Thus, a phytochemical investigation of the Matis poison was performed to identify its bioactive compounds. This led to the isolation and characterization of D-(−)-quinic acid, L-malic acid, ethyldimethylamine,

uring the course of their evolution, many plant species have survived through their ability to synthesize and accumulate toxic compounds that protect them against microorganisms, insects, herbivores, or even other plants. Some of these toxins have found a role in the subsistence activities of traditional societies, providing a source of arrow and fish poisons. In nearly all parts of the world, hunting poisons prepared from extracts of poisonous plants have been used as auxiliary weapons in the procurement of food and clothes.1,2 Indeed, people worldwide made their hunting trips more effective by applying poisonous materials to arrows, darts, spears, or javelins. Arrow poisons from South America are a rich natural source of alkaloids from several genera of the Menispermaceae [Abuta, Chondrodendron (especially C. tomentosum), Curarea, Sciadotenia, and Telitoxicum] 3,4 as well as the Longaniaceae (Strychnos).5 Frogs of the family Dendrobatidae are also known as a source of arrow poison, e.g., Phyllobates bicolor, P. aurotaenia, and P. terribilis.6,7 Chondrodendron tomentosum Ruiz. & Pav. is the main species used by South American Indians as an arrow poison in various blends of curare. Curare is, by definition, a black resin obtained from certain tropical South American trees, containing bisbenzyltetrahydroisoquinoline alkaloids,8 especially d-tubocurarine, whose action on the skeletal muscle end plate acetylcholine receptor (AchR) resulting in muscular paralysis is well known.9 Many scientists have been fascinated by curare,10 and the chemical and pharmacological studies of the active principles led to the discovery of pharmacologically active compounds still © XXXX American Chemical Society and American Society of Pharmacognosy

Received: May 21, 2015

A

DOI: 10.1021/acs.jnatprod.5b00457 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

magnoflorine, and bisbenzyltetrahydroisoquinoline alkaloids (1−5). Commercially available d-tubocurarine chloride was used to compare its spectroscopic properties with the isolated bisbenzyltetrahydroisoquinoline alkaloid derivatives. However, d-tubocurarine itself could not be identified among these products. The effects of crude Matis poison and individual bisbenzyltetrahydroisoquinoline alkaloids (BBIQAs) (1−3) on rat muscle nAChR expressed in Xenopus oocytes have been investigated using the two-electrode voltage clamp technique.

Figure 1. Structure of magnoflorine.

number of benzyltetrahydroisoquinoline alkaloids originate via phenol oxidative coupling of an N-methylcoclaurine and a reticuline unit. In all instances, the original bridging occurs by an ether bridge, a biaryl bond, or a methyleneoxy bridge depending upon the biosynthetic route.26 The difference between seco (1) and cyclic forms (2−5) is evidenced by their 1H NMR spectra. In the seco form the protons of the 1,4-disubstituted aromatic moiety shows the expected AA′XX′ spin system, whereas in the cyclic form the four aromatic protons give separate signals caused by hindered rotation of the aryl residue. Furthermore, in the seco form H-1 and H-1′ have similar chemical shifts, but in the cyclized structures they have vastly different chemical shifts. Additionally, in the cyclic form (2 and reference d-tubocurarine) H-8 is strongly shielded and resonates near δH 5.0. BBIQA 1 has the same molecular formula as the known BBIQA lindoldhamine.27−29

■

RESULTS AND DISCUSSION Spot tests were used for the qualitative determination of secondary metabolites present in the Matis poison.19,20 We identified steroids and triterpenoids (dark green) by the Liebermann−Burchard test, flavonoids (red) by the Shinoda test, and alkaloids by using Dragendorff’s, Mayer’s, and Wagner’s reagents and the picric acid test. No saponins were detected. However, BBIQAs can be confused with saponins in the foam test, but no saponins were detected by 1H NMR spectroscopy. The dried poison (5.29 g) was purified to yield three fractions. Fraction A consisted of small pieces of bark, stem, and leaves. Fraction B yielded 12.3 mg of D-(−)-quinic acid and 3.4 mg of L-malic acid. The structures of D-(−)-quinic acid and Lmalic acid were elucidated by means of spectroscopic data comparison with literature values.21 Fraction C yielded ethyldimethylamine, magnoflorine, and compounds 1−5. Ethyldimethylamine was identified by means of data comparison with literature results.

Figure 2. Structure of lindoldhamine.

Lindoldhamine was isolated for the first time from the leaves of Lindera oldhamii Hemsl. (Lauraceae)27−29 and thereafter also from Polyalthia nitidissima (Dunal) Benth. (Annonaceae),30 Triclisia sacleuxii (Pierre) Diels (Menispermaceae),31 Albertisia papuana Becc (Menispermaceae),32 Guatteria gaumeri (Mexican Annonaceae),33 Xylopia parvif lora (A. Rich.) Benth. (Annonaceae),34 and Abuta pahni (Mart.) Krukoff & Barneby (Menispermaceae).35 The structure of lindoldhamine was established as a tail-totail bisbenzyltetrahydroisoquinoline by spectroscopic data and chemical degradations.27−29 However, there are no X-ray diffraction data and only limited information about 1H NMR data in pyridine-d5, MS, and physical properties. Our isolated compound 1 has the same molecular formula as lindoldhamine determined by HRESIMS {m/z 569.26475 [M + H]+ (calcd for C34H37N2O6+, 569.26461)}, but has different solubility and specific rotation {[α]33D +35 (c 0.1, EtOH),28,29 [α]22D +15.22 (c 0.3, MeOH)}. The 1H NMR data of lindoldhamine from different reports are summarized together with our own results for compound 1 in methanol-d4 in Table 1. Compound 1 is insoluble in CDCl3. Thus, it was not possible to compare our 1H NMR spectra directly with literature data of lindoldhamine.27−29,31,35,36 13 C(APT), edited HSQC, and HMBC spectra of BBIQA 1 displayed 34 carbon resonances, assignable to seven aromatic

Magnoflorine was obtained as a pale yellow resin. 13C NMR and HRESIMS data indicated a molecular formula of C20H24NO4+ based upon the protonated molecular ion at m/ z 342.1706 (calculated for C20H24NO4+: 342.1722). The 1H and 13C NMR data are in accordance with literature data for magnoflorine.22 From 1H, 13C, and edited HSQC spectra characteristic absorptions for three methylene groups (δH 3.22/ 2.80, δC 23.33; 3.61/3.46, 61.12 and 3.08/2.62, 30.40), two methoxy groups (δH 3.80, δC 54.68; 3.82, 54.91), two N-methyl groups (δH 3.30, δC 52.53 and 2.90, 42.11), and three aromatic protons (δH 6.61 s, δC 108.25; 6.62 d, 116.05; 6.74 d, 109.24; J = 8.0 Hz) were evident. The HMBC data showed the expected correlations corresponding to the condensed ring system. Magnoflorine has been described as a “weak” neuromuscular blocking agent and a not-so-weak ganglionic blocker.23−25 Compounds 1−5 were identified as BBIQA derivatives. The BBIQAs represent a large and important class of natural products. Most are formed via phenol oxidative coupling of two N-methylcoclaurine (or coclaurine) units, while a much smaller B



Article

A molecular formula of C34H35N2O6+ was assigned to BBIQA 3 on the basis of the HRESIMS ion at m/z 567.24980 [M + H]+ (calculated at 567.24896), which is consistent with a BBIQA containing one aryl ether bridge and two N−H groups. The existence of the cyclic form is evident from the NMR data, because H-10, H-11, H-13, and H-14 and the corresponding carbons of the p-substituted aromatic ring (C-10, C-11, C-13, C-14) appear as separated signals. Most importantly, the strongly shielded proton H-8 is absent. This was confirmed by APT and edited HSQC spectra. Only nine aromatic methine protons were observed. Correlation peaks of the two aromatic proton singlets at δH 6.84 and 6.86 and the corresponding carbon atoms in the HMBC spectrum are indicative of the presence of these protons in the tetrahydroisoquinoline units and the pentasubstitution of their aromatic rings. HMBC correlations of H-1 (δH 4.87) to C-8 (δC 123.89) and H-10′ (δH 6.00) to C-8 (δC 123.89), respectively, are indicative of the unusual C−C connectivity of the two tetrahydroisoquinoline moieties. The ECD spectrum shows two high-amplitude negative Cotton effects at 203 nm (Δε ≈ −3.56 mdeg) and 239 nm (Δε ≈ −2.84 mdeg) and high-amplitude positive Cotton effects at 216 nm (Δε ≈ +2.89 mdeg), 254 nm (Δε ≈ +2.61 mdeg), and 287 nm (Δε ≈ +2.16 mdeg) (Figure S55, Supporting Information). There are no ECD data of closely related compounds available in the literature. BBIQA 4 represents also a new bisbenzyltetrahydroisoquinoline alkaloid, as shown by its UV and IR absorptions and NMR data. The molecular formula of C36H37N2O7+ was assigned to compound 4 on the basis of the HRESIMS ion at m/z 609.25941 [M + H]+ (calculated 609.25953), consistent with BBIQAs containing two aryl ether bridges, one N−H, and one N-acetyl group (δC 172.64 CO, δH 2.18 s, δC 22.57 CH3). The presence of the cyclic form is evident from the NMR data, because the NMR signals of the phenyl moieties, similar to those of compounds 2−5 and d-tubocurarine, are separated due to hindered rotation. BBIQA 5, with a molecular formula of C37H39N2O7+, based on the HRESIMS ion at m/z 623.27507 [M + H]+ (calculated at 623.27518), is consistent with a BBIQA containing two aryl ether bridges, one N-CH3 (δH 2.91 br, δC 41.50 −CH3), and one N-acetyl group (δC 172.64 CO, δH 2.18 s, δC 22.57 CH3). 1 H and 13C NMR data are similar to those of BBIQA 4, with differences involving the presence of the N-methyl group. Some years ago a similar N-acetylated BBIQA was isolated from the root bark of Tiliacora racemosa Colebr (Menispermaceae). This plant is regarded as an antidote for snake bite.39,40 Interestingly, a derivative related to compounds 4 and 5 was reported as a blocker of L-type calcium channels.41 To evaluate the muscle relaxant activity of the Matis dart poison and its BBIQA compounds 1−3, we conducted electrophysiology experiments. At 1 mg/mL Matis poison reduced the acetylcholine-evoked current by 95 ± 2%, indicating that the poison contains components with muscle relaxant activity. To establish whether the muscle-relaxant properties of the Matis poison are due to the identified compounds, we performed additional electrophysiology experiments with BBIQA compounds 1−3. These compounds share structures similar to the well-known muscle relaxant dtubocurarine. BBIQA compounds 1−3 reduced the acetylcholine-evoked current at 20 μM by 45 ± 8%, 88 ± 5%, and 86 ± 5%, respectively.

Table 1. 1H NMR Literature Data of Lindoldhamine in Comparison to Compound 1 lindoldhamine in CDCl331,35,36 δH (J in Hz) H-1

4.05 dd

H-1′

4.15 dd

H-5 H-5′ H-8 H-8′ H-10′ H-13′

6.51 6.58 6.69 6.69 6.66 6.89

H-14′

6.91 dd

H-10

7.17 d

H-14

7.17 d

H-11 H-13 CH3O-6 CH3O-6′

6.87 6.87 3.86 3.85

s s s s d d

d d s s

4.05 dd (8.4, 5.4) 4.10 dd (9.0, 4.2) 6.55 s 6.62 s 6.66 s 6.67 s 6.88 d (2.4) 6.87 d (2.4) 7.19 d (1.8) 7.19 d (1.8) 6.92 m 6.92 m 3.81 s 3.82 s

lindoldhamine in pyridine-d528,29 δH (J in Hz)

compound 1 in methanol-d4 δH (J in Hz)

4.05

4.65

4.15

4.60

6.51 6.58 6.69 6.69 6.66 6.89

6.75−7.60

6.79 6.74 6.61 6.50 6.80 6.98

6.91

7.02

7.17

7.28

7.17

7.28

6.87 6.87 3.86 3.85

6.94 6.94 3.81 3.85

3.75 3.75

tertiary carbons, six aromatic quaternary carbons, 11 aromatic methines, two aliphatic methines, six aliphatic methylenes, and two OCH3 groups connected to aromatic rings. The NMR data featured a 1,2,4-trisubstituted and a 1,4-disubstituted benzene moiety [H-10′ (δH 6.80), H-13′ (δH 6.98), H-14′ (δH 7.02) and H-13 (δH 6.94), H-14 (δH 7.28), J = 8.1 Hz AA′XX′ spin system] together with four aromatic proton singlets at δH 6.79, 6.74, 6.61, and 6.50 (Table 2). The HMBC spectrum shows only the expected C−H correlations via two and three bonds within the two tetrahydroisoquinoline units, but gives no additional information about the connectivity of the two moieties (tail-to-tail, head-to-head, or head-to-tail; no correlations via four bonds were observed; Figure 3b). This was also supported by the NOESY spectrum [cross correlations between H-1 (δH 4.65), H-8 (δH 6.61), H-15 (δH 3.08/3.44 br), and H-1′ (δH 4.60), H-8′ (δH 6.50) and H-15′ (δH 3.10/3.27 br), respectively; Figure 3a]. Collectively these data indicate that BBIQA 1 is most likely an isomer of lindoldhamine with head-to-tail bonding corresponding to dtubocurarine. It is known that, at least in Berberis, the biosynthetic formation of tail-to-tail dimers is not selective.37,38 Therefore, it cannot be ruled out that 1 may be an isomer of lindoldhamine. The HRESIMS data of BBIQA 2 give the molecular ion [M + H]+ at m/z 595.28010, corresponding to a molecular formula of C36H39N2O6+ (calculated at 595.28026). In contrast to 1, compound 2 exists in the cyclic form, indicated by the appearance of separated signals for the protons and the corresponding carbons of the 1,4-disubstituted aromatic ring, caused by steric hindrance. Additionally, H-8 is strongly shielded as in d-tubocurarine (δH 5.31; d-tubocurarine: δH 5.05). A major difference with regard to d-tubocurarine is the absence of one N-methyl group and the fact that both methoxy groups at C-12′ (δC 150.90) and C-6′ (δC 149.87) are part of one BBIQA unit as determined by HMBC. C



Article

Table 2. NMR Spectroscopic Data of Compounds 1−3 Recorded in Methanol-d4 at 400 MHz for 1H and 100 MHz for 13C (δ in ppm) compound 1 position 1 3α 3β 4α 4β 4* 5 6 7 8 8* 9 10 11 12 13 14 2-NCH3 6-OCH3 1′ 3α′ 3β′ 4α′ 4β′ 4*′ 5′ 6′ 7′ 8′ 8*′ 9′ 10′ 11′ 12′ 13′ 14′ 2 × 2′-NCH3 6′-OCH3 12′-OCH3 15α/15β 15α′/15β′ a

δH (J in Hz)

compound 2

δC

mult

57.80 40.95

CH CH2

25.91

CH2

123.77 112.66 149.25 146.71 114.24 124.89 130.69 132.00 119.04 158.63 119.04 132.00

C CH C C CH C C CH CH C CH CH

56.40 57.43 40.59

CH3 CH CH2

25.78

CH2

123.78 112.66 149.20 146.51 114.37 124.63 128.26 123.15 145.24 149.60 118.68 127.18

C CH C C CH C C CH C C CH CH

3.82 s

56.41

CH3

3.08/3.44 br 3.10/3.27 br

40.40 40.23

CH2 CH2

a

4.65 3.32 3.55 3.05 3.09

t (7.1) br br br br

6.79 s

6.61 s

7.28 d (8.3) 6.94 d (8.3) 6.94 d (8.3) 7.28 d (8.3) 3.85 4.60 3.32 3.48 3.03 3.08

s t (7.2) br br br br

6.74 s

6.50 s

6.80 br

6.98 d (8.2) 7.02 dd (8.2, 1.5)

δH (J in Hz) 4.64 3.46 3.56 3.03 3.03

δC

a

d (11.3) br br br br

6.71 s

5.30 s

7.23 d (8.5) 6.81 d (8.5) 6.20 d (8.5) 6.56 d (8.5)

5.37 3.66 4.01 3.26 3.26

br br br br br

6.87 s

6.46 br

7.01 d (8.4) 7.50 dd (8.4, 1.5) 3.29 s 3.31 s 3.90 s 3.62 s 2.88/3.31 br 3.00/3.72 br

compound 3 mult

55.13 38.18

CH CH2

24.49

CH2

124.48 115.82 146.00 145.32 112.93 122.03 129.11 128.65 117.30 155.91 112.49 130.84

C CH C C CH C C CH CH C CH CH

65.12 54.59

CH CH2

23.27

CH2

118.87 107.84 149.87 138.21 138.07 120.42 129.26 123.70 142.81 150.90 112.70 123.32 50.34/51.53 55.79 56.29 39.56 37.76

C CH C C C C C CH C C CH CH CH3 CH3 CH3 CH2 CH2

δH (J in Hz)a

δC

mult

56.02 35.83

CH CH2

23.52

CH2

121.49 111.12 148.32 142.64 123.89 124.99 129.85 129.11 111.67 156.41 118.01 129.83

C CH C C C C C CH CH C CH CH

55.40 50.61 37.17

CH3 CH CH2

23.71

CH2

121.84 108.80 149.27 139.32 137.43 122.04 127.86 136.03 120.91 151.16 115.38 125.84

C CH C C C C C CH C C CH CH

3.90 s

55.68

CH3

2.84/3.25 br 2.47/3.36 br

39.37 39.59

CH2 CH2

4.87 3.49 3.68 3.22 3.22

br br m br br

6.86 s

6.55 d (8.3) 6.18 d (8.3) 6.59 d (8.5) 6.87 d (8.5) 3.86 5.16 3.45 3.65 3.11 3.11

s t (3.8) br br br br

6.84 s

6.00 d (1.9)

6.79 d (8.4) 7.17 dd (8.4, 1.9)

Signal peak broadening by hindered rotation: not all the coupling constants could be accurately determined. with a Jasco J-715 spectropolarimeter, and UV−vis spectra were recorded with a JASCO V-630. IR spectra were recorded with a JASCO FT/IR-4100 instrument. NMR spectra were measured using a Varian Mercury Plus (400 MHz) and a Bruker Avance III HD (400 MHz) spectrometer, respectively. HRESIMS spectra were generated with a Bruker Daltonics instrument in the ESI positive mode. Silica gel 60 (0.040−0.063 mm, 230−400 mesh, ASTM, Fluka) and C18 reversed-phase silica gel (0.040−0.063 mm, 230−400 mesh, SigmaAldrich) were used for column chromatography. TLC was performed on analytical aluminum sheets (H2O/HOAc/CH3OH and H2O/HCl/ CH3OH) and, for preparative work, a 20 × 20 cm RP-18 F254S glass plate (Merck KGaA, 64271 Darmstadt, Germany). The chemicals were purchased from ABCR, Acros, Aldrich, Alfa Aesar, Fluorochem, TCL Europe, and VWR at the highest commercially available purity.

Because the in vivo acute toxicity experiments indicated breathing suppression and the electrophysiology experiments showed that the Matis poison components are blockers of muscle nAChR, it seems reasonable that the toxicity of the Matis poison is associated with d-tubocurarine-like muscle relaxants. These results show for the first time that an isomer of lindoldhamine (1) inhibits muscle nAChR.

■

EXPERIMENTAL SECTION

General Experimental Procedures. Melting points were measured with a Büchi Melting Point B-540. Optical rotation data were obtained with a Schmidt + Haensch Polartronic MHz-8 at the sodium-D line (589 nm) using a 50 mm path-length cell with the solvent and concentration indicated. The ECD spectrum was recorded D



Article

Figure 3. (a) NOE correlations of compounds 1−5. (b) Key HMBC correlations of compounds 1−5. material was authenticated and deposited at the William L. Brown Center, Missouri Botanical Garden, in November 2011. Powdered material was maintained at room temperature (22−25 °C) and protected from light until required for extraction and analysis. Extraction and Isolation. Dried poison (5.29 g) was treated for 1 h with (100 mL) H2O using ultrasound. The water suspension was dissolved with 2% aqueous HOAc (100 mL) for 5 h to produce an acidic solution. On the surface of the solution small pieces of bark, stem, and leaves were present. Those solids were separated by filtration through Whatmann No. 1 filter paper to yield fraction A (201.4 mg). The acidic solution was extracted several times with Et2O (100 mL portions) in order to remove nonalkaloidal compounds, wax, fat, and impurities until the acidic phase was slightly brown colored. The Et2O fraction (fraction B) was dried using Na2SO4 (about 5 to 10 g of Na2SO4 for each 100 mL), filtered, and concentrated under reduced pressure at 40 °C. The residue was afforded in a nonalkaloidal fraction (343.5 mg; negative test with Dragendorff’s reagent), containing D-(−)-quinic acid, L-malic acid, and other unidentified organic material. Afterward the acidic solution was basified to pH 7 with gradual addition of a 25% ammonia solution under stirring. After the neutralization, the warm solution was allowed to stand for 2 h. The basic solution was extracted several times with 100 mL of CH2Cl2, CHCl3, and EtOAc. The organic layers were combined, dried with Na2SO4, filtered, and concentrated to dryness under vacuum using a rotary evaporator (40 °C), to yield 3.89 g of an alkaloid fraction (fraction C; brown oil; positive test with Dragendorff’s reagent).

Scheme 1. Two-Electrode Voltage-Clamp Measurement of the Effects of Matis Poison and Individual BBIQA Compounds 1−3 on Xenopus Oocytes Expressing Rat Muscle nAChRa

a

Control acetylcholine-evoked responses are shown on the right of each graph; the left part of each graph represents response inhibited by Matis poison or isolated BBIQA compounds 1−3. Matis poison applied at 1 mg/mL for 5 min reduced acetylcholine-evoked current through muscle nAChR by 95 ± 2%. BBIQA compounds 1−3 at 20 μM reduced acetylcholine-evoked current by 45 ± 8%, 88 ± 5%, and 86 ± 5%, respectively. Black bars represent 20 μM acetylcholine application. Poison Collection. The Matis poison was obtained in Leticia, at the Colombian−Brazilian border, by Dr. Rainer W. Bussmann. The E



Article

Table 3. NMR Spectroscopic Data of Compounds 4 and 5 and d-Tubocurarine Recorded in Methanol-d4 at 400 MHz for 1H and 100 MHz for 13C (δ in ppm) compound 4 position 1 3α 3β 4α 4β 4* 5 6 7 8 8* 9 10 11 12 13 14 2-NCH3 6-OCH3 1′ 3α′ 3β′ 4α′ 4β′ 4*′ 5′ 6′ 7′ 8′ 8*′ 9′ 10′ 11′ 12′ 13′ 14′ 2 × 2′-NCH3 2′-NCOCH3 6′-OCH3/12′-OCH3 15α/15β 15α′/15β′ a

δH (J in Hz) 5.08 3.47 3.76 3.18 3.31

a

d (11.1) br br br br

6.83 s

7.17 dd (8.7, 2.5) 6.64 dd (8.7, 2.5) 5.79 dd (8.5, 2.2) 6.28 dd (8.5, 2.2) 3.86 5.94 3.32 3.99 2.95 3.32

s d (9.7) br br br br

6.84 s

7.37 dd (8.5, 2.2) 6.56 dd (8.5, 2.2) 5.81 dd (8.2, 2.5) 6.21 dd (8.2, 2.2) 2.18 s 3.86 s 2.98/3.28 br 2.68/3.04 br

d-tubocurarine

compound 5

δC

mult

53.61 37.43

CH CH2

24.19

CH2

δH (J in Hz)

a

4.90 d (11.4) a

δC

mult

62.33 42.85

CH CH2

23.70

CH2

124.14 109.59 151.01 138.61 139.22 122.84 127.70 129.88 118.98 156.31 115.36 130.28 41.50 56.78 52.84 45.88

C CH C C C C C CH CH C CH CH CH3 CH3 CH CH2

29.13

CH2

128.60 108.77 149.31 138.27 139.22 124.14 132.14 130.34 118.57 155.27 114.97 129.29 172.64 22.57 56.78 33.04 44.75

C CH C C C C C CH CH C CH CH CO CH3 CH3 CH2 CH2

a a a

124.15 109.55 150.78 138.54 139.47 119.97 128.32 129.88 118.92 156.28 115.31 130.07

C CH C C C C C CH CH C CH CH

56.77 52.81 45.88

CH3 CH CH2

29.13

CH2

128.60 108.75 149.31 138.26 139.25 124.15 132.06 130.37 118.55 155.43 114.90 129.33 172.64 22.57 56.77 39.53 44.72

C CH C C C C C CH CH C CH CH CO CH3 CH3 CH2 CH2

6.85 s

7.19 dd (8.5, 2.3) 6.63 dd (8.5, 2.3) 5.82 6.30 2.91 3.87 5.93 3.32 3.99 2.95 3.31

d (8.0) dd (8.4, 2.1) br s dd (10.9, 1.5) br br br br

6.84 s

7.37 dd (8.6, 2.2) 6.56 dd (8.6, 2.8) 5.80 d (8.0) 6.21 dd (8.4, 2.2) 2.18 s 3.86 s 2.88/3.44 br 2.69 m/3.04 br

δH (J in Hz) 4.14 3.62 4.27 3.21 3.21

a

br br br br br

7.00 s

5.05 s

6.28 d (8.3) 7.08 d (8.3) br 6.67 7.21 2.88 3.96 5.28 3.65 4.29 3.27 3.27

dd (8.4, 2.2) dd (8.4, 2.2) br s d (9.5) br br br br

6.86 s

6.80 br

6.79 d (8.1) 7.02 dd (8.1, 2.2) 3.01 s; 3.17 s 3.93 s 3.41/4.00 br 3.07 br/3.38 m

δC

mult

65.22 br 53.66

CH CH2

21.13 br

CH2

123.63 112.67 br 149.62 146.40 br 115.21 br 118.90 128.47 133.56 116.40 156.33 111.92 130.27, 40.07 br 55.67 68.25 53.76

C CH C C CH C C CH CH C CH CH CH3 CH3 CH CH2

23.05 137.13 107.86 149.90 138.74 137.3 br 119.86 129.40 123.96 br 142.25 br 148.81 117.57 127.05 br 53.66; 50.41 55.65 45.16 br 39.56

CH2 C CH C C C C C CH C C CH CH CH3 CH3 CH2 CH2

Signal peak broadening by hindered rotation: not all the coupling constants could be accurately determined. (254 and 366 nm), and confirmation with Dragendorff’s reagent] to afford BBIQA 1 (15.8 mg; 0.79%), BBIQA 2 (26.2 mg; 1.31%), and BBIQA 3 (16.5 mg; 0.82%). After RPC of fractions 4−9, a dark green mixture was obtained. Using RPC (eluent 1.0 M HCl) to remove the dark green color afforded the corresponding fraction 10. Fraction 10 (110.0 mg) was eluted [RPC with H2O/HCl/CH3OH (50:0.1:50 to 0.1:1:99.9); 2.5 M HCl and purified by preparative RPTLC; H2O/HCl/CH3OH (60:0.1:40)] to afford BBIQA 4 (10.1 mg; 0.51%) and BBIQA 5 (15.0 mg; 0.75%). BBIQA Derivative 1: slightly yellow, amorphous solid; [α]22D +15.22 (c 0.3, MeOH); UV (MeOH) λmax (log ε) 284 (0.03), 224 (0.18), 202.5 (0.54) nm; IR (KBr) νmax 3443, 1622, 1508, 1383, 1273, 1217, 1120, 1024 cm−1; 1H and 13C NMR data, see Table 2; HRESIMS m/z 569.26475 [M + H]+ (calcd for C34H37N2O6+, 569.26461). [Note: [α]33D +35 (c 0.1, EtOH); m/z 568 (M+) colorless, data corresponding to lindoldhamine from the literature.28,29]

Silica gel 60 (40 g; Fluka) was used to separate fraction B by column chromatography with CH2Cl2/CH3OH (10:1), yielding D(−)-quinic acid (12.3 mg) and L-malic acid (3.4 mg). Subsequently, fraction C (alkaloid fraction; 2.0 g) was separated by RPC [C-18 reversed-phase silica gel (Sigma-Aldrich); eluted from an RPC column of 2.5 cm diameter and 25 cm length] with H2O/HOAc/ CH3OH (99:1:1), resulting in 58 fractions (alkaloid test using Dragendorff’s reagent). Fractions with similar Rf (TLC) values were combined into nine fractions. After concentration of these fractions, fraction 1 (67.0 mg) was purified by RPC (H2O/HOAc/CH3OH, 98:1:2) to afford ethyldimethylamine (11.9 mg). Fraction 3 (208.0 mg) was separated with H2O/HOAc/CH3OH (95:1:5) and further purification using RPC and preparative RP-TLC (H2O/HOAc/ CH3OH, 80:1:20) to give magnoflorine (12.1 mg). Fractions 4−9 (608.0 mg) were separated [by RPC with H2O/HCl/CH3OH (55:0.1:45 to 0.5:0.1:99.5); 2.5 M HCl, and by preparative RP-TLC; H2O/HCl/CH3OH (1:0.1:1)], alkaloid spot detection by UV light F



■

BBIQA Derivative 2: slightly yellow, amorphous solid; [α]22D +118.30 (c 0.6, MeOH); UV (MeOH) λmax (log ε) 284 (0.10), 224.5 (0.53), 204.5 (1.11) nm; IR (KBr) νmax 3420, 1617, 1509, 1463, 1270, 1222, 1169, 1125, 1076 cm−1; 1H NMR and 13C NMR data, see Table 2; HRESIMS m/z 595.28010 [M + H] + (calcd for C36H39N2O6+, 595.28026). BBIQA Derivative 3: slightly yellow, amorphous solid; [α]22D +11.29 (c 0.3, MeOH); UV (MeOH) λmax (log ε) 284.5 (0.05), 206 (0.32) nm; ECD (c 0.037 mM, MeOH, Δε) λmax 302 (0), 287 (+2.16), 269 (+0.93), 254 (+2.61), 246 (0), 239 (−2.84), 223 (0), 216 (+2.89), 210 (0), 203 (−3.56) nm; IR (KBr) νmax 3446, 1622, 1507, 1384, 1215, 1124 cm−1; 1H NMR and 13C NMR data, see Table 2; HRESIMS m/z 567.24980 [M + H]+ (calcd for C34H35N2O6+, 567.24896). BBIQA Derivative 4: slightly yellow, amorphous solid; [α]22D +15.4 (c 0.7, MeOH); UV (MeOH) λmax (log ε) 283 (0.10), 225 (0.55), 204.5 (1.12) nm; IR (KBr) νmax 3236, 1522, 1237, 1224, 1215, 1014 cm−1; 1H NMR and 13C NMR data, see Table 3; HRESIMS m/z 609.25941[M + H]+ (calcd for C36H37N2O7+, 609.25953). BBIQA Derivative 5: slightly yellow, amorphous solid; [α]22D +8.5 (c 0.3, MeOH); UV (MeOH) λmax (log ε) 277 (0.07), 225.5 (0.34), 203 (0.39) nm; IR (KBr) νmax 3346, 1322, 1207, 1184, 1103, 1001 cm−1; 1H NMR and 13C NMR data, see Table 3; HRESIMS m/z 623.27507 [M + H]+ (calcd for C37H39N2O7+, 623.27518). Xenopus Oocytes and Electrophysiology Measurements. Oocytes were harvested from mature Xenopus frogs, manually singleted, and injected with pRBG4 plasmids, containing clones of rat muscle nAChR subunits (gift from Veit Wietzemann, Department of Molecular Neurobiology, Max Planck Institute for Medical Research, Heidelberg, Germany). After injection the oocytes were incubated at 18 °C in ND96 solution (5 mM HEPES/NaOH at pH 7.6 and 18 °C, 96 mM NaCl, 2 mM KCl, 1.8 mM CaCl2, 2 mM MgCl2)42 for 48 h, after which the recordings were performed.43 The experimental setup was as follows: several control 5−10 s pulses of 20 μM acetylcholine (one pulse per 5 min), each acetylcholine pulse was followed by several minutes of washing the oocyte with ND96 buffer; then BBIQA compounds 1−3 (20 μM) or Matis poison (1 mg/mL) was applied to the oocyte for 5 min, followed by a test pulse of 20 μM acetylcholine with the tested BBIQA.

■

REFERENCES

(1) Kopp, B.; Bauer, W. P.; Schnürch, A. B. J. Ethnopharmacol. 1992, 36, 57−62. (2) Bisset, N. G. Lloydia 1966, 29, 1−18. (3) Bisset, N. G.; Woods, M. C. Lloydia 1966, 29, 172−195. (4) Schultes, E. R. Supl. Acta Amazonica 1988, 18 (1−2), 393−406. (5) Philippe, G.; Angenot, L.; Tits, M.; Frédérich, M. Toxicon 2004, 44, 405−416. (6) Philippe, G.; Angenot, L. J. Ethnopharmacol. 2005, 100, 85−91. (7) Daly, J. W.; Myers, C. W.; Whittaker, N. Toxicon 1987, 25 (10), 1023−1095. (8) King, H. J. Chem. Soc. 1948, 1945−1949. (9) Lee, C. Br. J. Anaesth. 2001, 87 (5), 755−769. (10) Bisset, N. G. J. Ethnopharmacol. 1992, 36, 1−26. (11) Bisset, N. G. Alkaloids: Chemical and Biological Perspectives, Vol. 8; Springer: Berlin, 1992; pp 3−150. (12) Läwen, A. Bruns Beitr. Klin. Chir. 1912, 80, 168−180. (13) Griffith, H. R.; Johnson, E. Anesthesiology 1942, 3, 418−420. (14) Prescott, F.; Organe, G.; Rowbotham, S. Lancet 1946, ii, 80−84. (15) Gray, T. C.; Halton, J. Br. Med. J. 1950, ii, 293−295. (16) http://www.amazonrainforestnews.com/2011/06/followinghunting-matis-of-brazil.html (02/08/2015). (17) http://www.indian-cultures.com/cultures/matis-indians/ (02/ 08/2015). (18) Tulp, M.; Bohlin, L. Drug Discovery Today 2004, 9, 450−458. ́ (19) Dominguez, X. A. Métodos de Investigación Fitoquimica, 1st ed.; Editorial Limusa: México, 1973. (20) Harborne, B. J. Phytochemical Methods, a Guide to Modern Techniques of Plant Analysis, 2nd ed.; Chapman and Hall: New York, 1984. (21) Eliel, E. L.; Ramirez, M. B. Tetrahedron: Asymmetry 1997, 8 (21), 3551−3554. (22) Ishii, H.; Imai, M.; Johji, S.; Tan, S.; Chen, I. S.; Ishikawa, T. Chem. Pharm. Bull. 1994, 42 (1), 108−111. (23) Aiyar, V. N.; Benn, M. H.; Hanna, T.; Jacyno, J.; Roth, S. H.; Wilkens, J. L. Experientia 1979, 35, 1367−1368. (24) Chang, C. C.; Wang, L. C. C.; Shac, I.-T.; Pei, Y. C.; Chiang, M. Y.; Li, T.; Hsu, T. C. Acta Pharm. Sin. 1964, 11, 42−49. (25) Shimamoto, K.; Uchizumi, S.; Inoue, K. Nippon Yakurigaku Zasshi 1957, 53, 75−80. (26) Hagel, J. M.; Facchini, P. J. Plant Cell Physiol. 2013, 54 (5), 647−672. (27) Guha, K. P.; Mukherjee, B.; Mukherjee, R. J. Nat. Prod. 1979, 42 (1), 1−84. (28) Lu, S. T.; Chen, I. S. Heterocycles 1976, 4, 1073−1076. (29) Lu, S. T.; Chen, I. S. J. Chin. Chem. Soc. 1977, 24, 187−194. (30) Jossang, A.; Leboeuf, M.; Cabalion, P.; Cavé, A. Planta Med. 1983, 49, 20−24. (31) Murebwayire, S.; Ingkaninan, K.; Changwijit, K.; Frédérich, M.; Duez, P. J. Pharm. Pharmacol. 2009, 61, 103−107. (32) LeBoeuf, P. M.; Abouchacra, M. L.; Sevenet, T.; Cavé, A. Plant. Med. Phytother. 1982, 16, 280−291. (33) Dehaussy, H.; Tits, M.; Angenot, L. Planta Med. 1983, 49 (9), 25−27. (34) Nishiyama, Y.; Moriyasu, M.; Ichimaru, M.; Iwasa, K.; Kato, A.; Mathenge, S. G.; Chalo Mutiso, P. B.; Juma, F. D. Phytochemistry 2006, 67, 2671−2675. (35) Duté, P.; Weber, J. F.; Fournet, A.; Cavé, A.; Bruneton, J. Phytochemistry 1987, 26 (7), 2136−2137. (36) Guinaudeau, H.; Freyer, J.; Shamma, M. Nat. Prod. Rep. 1986, 3, 477−488. (37) (a) Stadler, R.; Zenk, M. H. J. Biol. Chem. 1993, 268, 823−831. (b) Stadler, R.; Loeffler, S.; Cassels, B. K.; Zenk, M. H. Phytochemistry 1988, 27, 2557−2565. (38) Dewick, P. M. Medicinal Natural Products: A Biosynthetic Approach, 3rd ed.; John Wiley & Sons: England, 2009; pp 315−325. (39) Kirtikar, K. R.; Basu, B. D. Indian Medicinal Plants; Lalit Mohan Basu: Allahabad, 1933; Vol. 1, p 83.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jnatprod.5b00457. HRESIMS and 1H, 13C, and 2D NMR spectra of bisbenzyltetrahydroisoquinoline alkaloids 1−5; ECD spectrum of 3 (PDF)

■

Article

AUTHOR INFORMATION

Corresponding Authors

*Tel/Fax (V. I. Tsetlin): (007) 495 335 57 33. E-mail: vits@ mx.ibch.ru. *Tel (A. Giannis): +49(0) 341 97 36 581. Fax: +49(0)341 97 36 599. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors are grateful to Prof. Dr. A. G. Beck-Sickinger and M.Sc. F. Ott (University of Leipzig, Institute of Biochemistry, Germany) for the measurements of the ECD spectra and Mrs. R. Oehme (University of Leipzig, Institute of Analytical Chemistry, Germany) for the measurement of the HRESIMS spectra. G



Article

(40) Guha, K. P.; Das, P. C.; Mukherjee, B.; Mukherjee, R.; Juneau, G. P.; Bhacca, N. S. Tetrahedron Lett. 1976, 17, 4241−4244. (41) Felix, J. P.; King, F.; Shevell, J. L.; Garcia, M. L.; Kaczorowski, G. J.; Bick, I. R. C.; Slaughter, R. S. Biochemistry 1992, 31, 11793−11800. (42) http://www.oecd.org/chemicalsafety/testing/ oecdguidelinesforthetestingofchemicals.htm (12.08.2015). (43) Lyukmanova, E. N.; Shulepko, M. A.; Buldakova, S. L.; Kasheverov, I. E.; Shenkarev, Z. O.; Reshetnikov, R. V.; Filkin, S. Y.; Kudryavtsev, D. S.; Ojomoko, L. O.; Kryukova, E. V.; Dolgikh, D. A.; Kirpichnikov, M. P.; Bregestovski, P. D.; Tsetlin, V. I. J. Biol. Chem. 2013, 288, 15888−15899.

H


Curare Alkaloids: Constituents of a Matis Dart Poison - Journal of

Recommend Documents