Toxicon 46 (2005) 786–796 www.elsevier.com/locate/toxicon

Isolation and chemical characterization of PwTx-II: A novel alkaloid toxin from the venom of the spider Parawixia bistriata (Araneidae, Araneae) Lilian M.M. Cesara, Maria A. Mendesa, Claudio F. Tormenab, Maurı´cio R. Marquesa, Bibiana M. de Souzaa, Daniel Menezes Saidemberga, Jackson C. Bittencourtc, Mario S. Palmaa,* a

Institute of Biosciences, Sa˜o Paulo State University (UNESP), Avenue 24 A, 1515-Bela vista, 13506-900 Rio Claro, SP, Brazil b Department of Chemistry, Faculdade de Filosofia Cieˆncias e Letras da Universidade de Sa˜o Paulo, FFCLRP-USP, Av. Bandeirantes, 3900, 14040-901 Ribeira˜o Preto, SP, Brazil c Laboratory of Chemical Neuroanatomy, Department of Anatomy, Institute of Biomedical Sciences, University of Sa˜o Paulo (USP), Sa˜o Paulo 05508-900, Brazil Received 9 June 2005; revised 10 August 2005; accepted 11 August 2005 Available online 23 September 2005

Abstract Brazil has many species of spiders belonging to Araneidae family however, very little is known about the composition, chemical structure and mechanisms of action of the main venom components of these spiders. The main objective of this work was to isolate and to perform the chemical characterization of a novel b-carboline toxin from the venom of the spider Parawixia bistriata, a typical species of the Brazilian ‘cerrado’. The toxin was purified by RP-HPLC and structurally elucidated by using a combination of different spectroscopic techniques (UV, ESI-MS/MS and 1H NMR), which permitted the assignment of the molecular structure of a novel spider venom toxin, identified as 1-4-guanidinobutoxy-6-hydroxy-1,2,3,4-tetrahydro-bcarboline, and referred to here as PwTx-II. This compound is toxic to insects (LD50Z12G3 hg/mg honeybee), neurotoxic, convulsive and lethal to rats (LD50Z9.75 mg/kg of male Wistar rat). q 2005 Elsevier Ltd. All rights reserved. Keywords: Spider venom; Spider toxin; Natural compound; Alkaloid; Structure determination; Neurotoxicity; Natural product; Spectroscopy

1. Introduction Studies of Arthropod defensive chemistries continue to bring to light novel structures and unanticipated biosynthetic capabilities. In the last decade, insecticidal toxins from Arthropod venoms have been the subject of considerable emphasis in the literature. Spider venoms generally are constituted of complex mixtures of * Corresponding author. Tel.: C55 19 3526 4163; fax: C55 19 3534 8523. E-mail address: [email protected] (M.S. Palma).

0041-0101/$ - see front matter q 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.toxicon.2005.08.005

biologically active toxins, which may be grouped into three major classes of compounds according to their chemical nature: high molecular mass proteins (MwO 10 kDa), peptides (Mw 3–10 kDa) and low molecular mass compounds (Mw!1 kDa) (Kawai and Nakajima, 1993). The latter class can be subdivided into other subclasses, according to the chemical nature of these toxins. (Marques et al., 2004) The acylpolyamine toxins constitute a large family of neuroblockers of ionotropic glutamate receptors, occurring in the venoms of solitary wasps and spiders, acting at the level of the neuromuscular junction of Arthropods in

L.M.M. Cesar et al. / Toxicon 46 (2005) 786–796

general, causing paralysis/death (Nakanishi et al., 1990; Skinner et al., 1990; Quistad et al., 1993; Palma et al., 1997; Palma and Nakajima, 2005). Nucleoside toxins were identified in the venom of the spider Hololena curta; these toxins are constituted of mono- and disulphate derivatives of guanosine or xantosine, bearing one or two D-fucose units, which have the uncommon property of effectively blocking kainate receptors, in addition to weakly blocking L-type calcium channels (McCormick et al., 1999). The nucleoside inosine was also demonstrated to be a component of Parawixia bistriata venom, which causes paralysis in termites (Rodrigues et al., 2004). The organometallic diazenaryl compounds are potent insecticidal toxins also identified in the venom of P. bistriata, presenting lethal effect even when topically applied on spiders’ preys (Marques et al., 2005). The tetrahydrobetacarboline ring, 1,2,3,4-tetrahydro-7Hpyrido/3,4-b/indole (THBC) constitutes the basic element of several natural indolyl alkaloids, such as those occurring in some ascidians and plants such as Peganum harmala, or Rauwolfia serpentine (Abramovitch and Spencer, 1964; Szanta´y et al., 1986; Carmona et al., 2000). Tetrahydro-b-carboline compounds are endogenous in some animals, and found at trace levels in mammalian brain (Barker et al., 1981; Johnson et al., 1985). These supposed mammalian THBC alkaloids probably arise endogenously from the condensation of central nervous system indolamines (or their precursor amino acid, L-tryptophan) with an aldehyde or a-keto acid via Pictet–Spengler reaction (Carmona et al., 2000). These endogenously formed alkaloids act on various aspects of neurotransmission modulation, and are neurotoxic since they constitute a family of high affinity ligands of the benzodiazepine (BDZ) receptors which is a sub-type of GABA receptor) (Robertson, 1980). P. bistriata (Araneidae, Araneae) is a very common species of social spider in Brazil (Levi, 1992) and its venom seems to constitute a rich source of neurotoxins. It was reported recently that intracerebroventricular (i.c.v.) application of the crude venom caused limbic seizures in rats (Rodrigues et al., 2001); in addition to this, an isolated fraction of this venom (not structurally characterized) inhibited the uptaking of GABA in rat cortical sinaptossomes (Rodrigues et al., 2002). P. bistriata uses tetrahydrob-carboline compounds as part of its chemical weaponry to kill/paralyse the Arthropod preys (Marques et al., 2005). These molecules are natural analogues of trypargine [1-(3 0 guanidinopropyl)-1,2,3,4-tetrahydro-b-carboline]; an alkaloid compound previously isolated from the skin of the African frog Kassina senegalensis (Akizawa et al., 1982). Recently, an indoleyl alkaloid compound was reported in the venom of this spider and characterized as a potent insecticide toxin (Cesar et al., 2005). A novel alkaloid toxin has been now identified in the crude venom of the P. bistriata; it was isolated from acetonitrile (CH3CN) extracts of the crude venom by chromatographic techniques,

787

and its structure has been elucidated through the combined use of UV-spectrophotometry, 1H NMR spectroscopy and ESI mass spectrometry. The structural features of 1,4guanidinobutoxi-6-hydroxy-1,2,3,4-tetrahydro-b-carboline and the study of its insecticidal action may be used, as a starting point for the development of new molecules for pest control.

2. Material and methods 2.1. Spider collection, venom extraction and purification P. bistriata specimens were collected in Rio Claro, SP, southeast Brazil. Spiders were sacrificed by freezing at K20 8C. Venom glands were removed with surgical microscissors and the venom was extracted with 1:1 acetonitrile (CH3CN)/ultra pure water. The extract was centrifuged through AMICON 3 spin filters (Millipore) at 8000!g during 15 min at 4 8C and the low molecular mass (LMM) fraction (!3 kDa) was collected, lyophilized and stored at K10 8C. The LMM fraction was then solubilized in 5% (v/v) CH3CN in bi-distilled water (containing 0.1% TFA) and fractionated in a HPLC system (SHIMADZU, mod. LC10Advp) equipped with a diode array detector (SHIMADZU, mod. SPD-10Avp), using a reversed-phase semi-preparative column CapCell Pack-C18 (10!250 mm, 5 mm). Elution was carried out by linear gradient from 5 to 60% (v/v) CH3CN in bi-distilled water (containing 0.1% TFA) during 60 min at 30 8C. The UV absorbance was monitored at 220, 254 and 280 nm, at a flow rate of 2.5 mL minK1. 2.2. Mass spectrometry analysis Mass spectra were acquired on a triple quadrupole (Quatro II) mass spectrometer instrument (Micromass, UK), equipped with a standard electrospray probe, adjusted to ca. 5 mL minK1. During all experiments the source temperature was maintained at 80 8C and the needle voltage at 3.6 kV, applying a drying gas flow (nitrogen) of 200 L hK1 and a nebulizer gas flow (nitrogen) of 20 L hK1. The mass spectrometer was calibrated with intact horse heart myoglobin and its typical cone-voltage induced fragments. The cone sample to skimmer lens voltage, controlling the ion transfer to the mass analyzer, was maintained at 30 V. About 50 pmol of each sample was injected into electrospray transport solvent. The ESI spectra were obtained in the continuous acquisition mode, scanning from m/z 100 to 2000 with a scan time of 5 s. 2.3. MS/MS spectrometry experiments Typical conditions were: a capillary voltage of 3 kV, a cone voltage of 30 V, collision gas pressure of 3.5!10K3 mbar and a desolvation gas temperature of 80 8C. In these

788

L.M.M. Cesar et al. / Toxicon 46 (2005) 786–796

experiments Q1 was used to select the parent ion and was not scanned. The ion of interest was individually selected in Q1 and structurally characterized by collision-induced dissociation (CID). It was subjected to about 25 eV collision energy and 5!10K3 mbar collision gas pressure (argon) in Q2. The CID fragments were analyzed by scanning Q3. 2.4. Nuclear magnetic resonance experiments The 1H NMR spectrum was recorded at 25 8C on a Varian INOVA 500 spectrometer, operating at 499.88 MHz for 1H. Spectrum was of ca. 500 mg cmK3 solutions in D2O, which was used as deuterium lock and reference for spectrum. The signal for remaining H2O was partially suppressed applying a presaturation sequence (Braun et al., 1996). 2.5. Rat toxicity Adult male Wistars rats of 60 days (250–300 g) were housed two per cage in our institutional animal care facility and allowed to adapt at least 7 days prior to the onset of experiments. Animals were maintained on a 12 h light-dark cycle, in a temperature-controlled environment (21 8C and 55% of humidity), with free access to water and food. All experiments were carried out in accordance with the guidelines of the National Academy of Sciences ‘Guide for the Care and Use of Laboratory Animals’ (N.A.S., 1996) and the University of Sa˜o Paulo State Committee for Ethics and Animal Care in Experimental Research. The guide was implanted in the lateral ventricle (-0,4 (AP); 1,4 (ML); 3,4 (DV)) under anaesthesia by ketamine, xylazine and acepromazine (25:5:1 mg KgK1 body weight), seven days before the i.c.v injection. Before the experiment, animals were manipulated, at least twice a day for 10 min, as a way to avoid the stress caused by the manipulation induced by the i.c.v injection. The injection was introduced through the guide 2 h before the perfusion, in order to minimize any effect during the experiment. A volume of 10 mL of THBC was injected at a concentration from 2 to 13 mg KgK1 rat (NZ5, per concentration), dissolved in saline solution (0.9%); in control animals, only saline solution was injected (0.9%). Estimation of median lethal dose (LD50) was performed by using the minimal number of animals following a protocol in accordance with the Test Guidelines 423 from OECD (Organization for Economic Co-operation and Development) Guidelines for testing of chemicals (OECD, 2001). 2.6. Insecticide activity Different doses of the natural compound (from 2 to 100 ng mgK1 of insect) (NZ10, per concentration) were injected in a final volume of 1 mL into the pronotum of honeybees (Africanized Apis mellifera), using a Hamilton

microsyringe (10 mL). The insects were kept in a Petri dish up to 4 h in the presence of candy (food) and water supply. During this period the toxicity effects and/or the lethal action of the new toxins were observed. Control experiments were performed by injecting the physiological solution into the insect pronotum. Toxicity levels were calculated according to the Probit method (Chou and Chou, 1987) and expressed as 50% lethal doses (LD50). Results are expressed as meansGS.D. of five experiments. Differences between the two extracts (reconstituted and non-reconstituted) were evaluated with Students’ t-test. Values of P! 0.05 were considered significant. 2.7. Trypargine syntheses Trypargine was synthesized from 2-benzyl-3-(methoxycarbonyl)-1,2,3,4-tetrahydro-9H-pyrido[3,4-b]indole-1-propionic acid, prepared by asymmetric Pictet–Spengler reaction of Nb-benzyl-(D)-tryptophan methyl ester. The synthetic hydrochloride form of the toxin was identified by its physico-chemical properties (mp 211–213 8C; [a]DZC 378 (CH3OH); MS m/z: 272 [MCH]C) by comparison with the characterization data from previous work (Shimizu et al., 1982a,b).

3. Results and discussion The LMM fraction from the crude venom of the spider P. bistriata was initially fractionated by RP-HPLC (C-18). The chromatographic separation resulted in 14 distinct fractions (Fig. 1(a)), which were submitted to the insecticidal bioassay. Fraction 7 was among the most abundant ones and the results show it as a very active component of this venom, responsible for causing the death of honeybees and convulsions in rats. Fraction 7 was submitted to ESI-MS analysis, which revealed a series of different compounds; among these compounds, those of molecular masses 288 and 316 in the [MCH]C form, were the major components (Fig. 1(c)). Afterwards, this fraction was submitted to re-chromatography under isocratic conditions at 26% (v/v) CH3CN [containing 0.1% (v/v) TFA] with a semipreparative column (CapCell Pack-C18 10!250 mm, 5 mm) at a flow rate of 1.0 mL minK1 and the elution was monitored at 215 nm. The chromatographic profile presented two fractions 7.1 and 7.2 (Fig. 1b), which were analyzed by ESI-MS and had molecular masses of 315 and 287 Da (data not shown), respectively. These fractions were submitted to UV spectrophotometry, mass spectrometry (ESI-MS/MS) and 1H NMR spectroscopy. The compound of molecular mass 287 Da (fraction 7.2), was previously characterized as N-[2-(6hydroxy-2,3,4,9-tetrahydro-1H-b-carbolin-1-yl) ethyl] guanidine, which also may be referred to hydroxytrypargine (PwTx-I) (Cesar et al., 2005).

L.M.M. Cesar et al. / Toxicon 46 (2005) 786–796

The UV spectrum of the compound present in fraction 7.1 had UVmax (log 3): 224 nm (4.52), 275 nm (3.71) and 289 nm (3.42), which when compared, to a library of UV spectra, suggested the presence of an indole moiety.

789

Synthetic trypargine toxin was used as a reference for the spectroscopic analyses to help in the structural elucidation of the natural toxin from fraction 7.1.

Fig. 1. (a): RP-HPLC profile of the venom from the spider Parawixia bistriata with a linear gradient from 5 to 60% (v/v) MeCN [0.1% (v/v/) TFA] on a semi-preparative column CapCell Pack-C18 (10!250 mm, 5 mm). The flow rate was 2.5 mL minK1 and the elution was monitored at 215 nm. (b): RP-HPLC profile of fraction 7 under isocratic conditions, at 26% (v/v) MeCN [0.1% (v/v) TFA] on a semi-preparative column (CapCell Pack-C18 10!250 mm, 5 mm) with a flow rate of 1.0 mL minK1 and the elution was monitored at 215 nm. (c): ESI-MS spectra of fraction 7 from de venom of P. bistriata spider; the asterisks indicate the main molecular ions.

790

L.M.M. Cesar et al. / Toxicon 46 (2005) 786–796

The 1H NMR spectra for trypargine (Fig. 2a) and for the natural toxin (Fig. 2b) were compared to each other, chemical shifts (ppm) and coupling constants (Hz) for both compounds and also for hydroxytrypargine are

shown in Table 1. All the spectra shown two distinct regions: aromatic (6–8 ppm) and aliphatic (2–5 ppm). The aliphatic regions are quite similar for three compounds as can be observed by analyzing spectra from Fig. 2(a)

Fig. 2. (a): 1H NMR spectrum for the synthetic trypargine at 500.13 MHz in D2O. Expanded regions: (a) 7.7–7.1 ppm; (b) 3.8–1.7 ppm and (c) 4.73–4.69 ppm. (b): 1H NMR spectrum (500.13 MHz in D2O) of the natural toxin (fraction 7.1), purified from the venom of the spider P. bistriata at a concentration of ca. 500 mg cmK3 in D2O. The spectrum was acquired at 499.88 MHz.

L.M.M. Cesar et al. / Toxicon 46 (2005) 786–796

791

Table 1 1 H NMR chemical shifts (ppm) and coupling constants (Hz) for the compounds PwTx-II, trypargine (1) and hydroxytrypargine (2) 5

HO

4 4b

7 8

8a N H

9a

3

O

7

1 H N

3' 2'

4 4b

4a

3

6

NH2 1'

PwTx-II

5

R

4a

6

8a 8

6'

NH2

N H

NH2

9a

1 1'

4'

2'

NH

NH

R= H Trypargina (1) R= OH Hydroxytrypargine (2)

Position

1

1 3

4.65 (s) 3.69 (dt, JZ12.2, 4.7) 3.37 (ddd, JZ12.2, 5.8, 4.7) 2.97 (m) – – 6.97 (d, JZ2.7) – 6.81 (dd, JZ10.5, 2.7) 7.34 (d, JZ10.5) – – – 2.17 (t, JZ6.3) 1.76 (m) 3.24 (t, JZ6.6)

4 4a 4b 5 6 7 8 8a 9a 10 20 30 40

H (Pw-Tx-II)

1

3'

5'

NH2

1

H (1)

H (2)

4.70 (t, JZ5.0) 3.74 (dt, JZ12.6, 4.9) 3.42 (ddd, JZ12.6, 6.0, 4.9) 3.08 (m) – – 7.51 (d, JZ8.1) 7.29 (t, JZ8.1) 7.21 (t, JZ8.1) 7.62 (d, JZ8.1) – – 2.03, 2.25 (2m) 1.79 (m) 3.27 (t JZ6.75) –

and (b). The greater difference was observed in the aromatic region, where there are four different hydrogens for trypargine at 7.21, 7.29, 7.51 and 7.62 ppm (Table 1), while the natural toxin and hydroxytrypargine presented only three different hydrogens at 6.81, 6.97 and 7.34 ppm for the compound from the fraction 7.1 and 6.98 6.82 and 7.32 (ppm) for hydroxytrypargine. The 1H NMR spectrum for trypargine (Fig. 2(a)) is consistent with a 1,2disubstituted aromatic moiety; meanwhile, for natural toxin and hydroxytrypargine the pattern shows two doublets and a singlet (Fig. 2(b)) consistent with a 1,2,4-trisubstituted aromatic moiety. The interpretation of the tandem mass spectrum of fraction 7.1 (Fig. 3) suggests the presence of a hydroxyl substituent in the chromophore moiety, that was corroborated by the comparison between the 1H NMR spectra of natural toxin and the synthetic trypargin. Some signals of 1 H NMR spectrum of natural toxin (Fig. 2(b)) are not exactly coincident with those observed for trypargine (Fig. 2(a)) mainly due to the different solvents used for each sample [acetic acid-d4 (CD3COOD) for the natural toxin and D2O for the analysis of trypargine]. Thus, taking together, the results of UV-spectrophotometry and 1H NMR spectroscopy analysis suggest the presence of a 6-hydroxy-1,2,3,4-tetrahydro-b-carboline as the chromophore group of the natural alkaloid toxin present in the fraction 7.1. According to the known structures of the most tetrahydro-b-carboline compounds, other signals due

N H

4.68 (t, JZ5.1) 3.72 (dt, JZ12.0, 4.5) 3.40 (ddd, JZ12.0, 6.5, 4.5) 2.97 (m) – – 6.98 (d, JZ3.0) – 6.82 (dd, JZ10.9, 3.0) 7.32 (d, JZ10.9) – – 2.17 (m) 1.75 (m) 3.22 (t, JZ6.4) –

to the presence of aliphatic H-atoms would be expected. The absence of these signals may be due to the very small amount of sample (w500 mg) that the spectrum was acquired with. All signals referring to these aliphatic H-atoms would split in more than six lines due to geminal and vicinal coupling constants. Then, these signals appear at the base line, due to a low intensity of the signal for each H-atom. Thus, only the structure of the chromophore moiety could be elucidated using 1H NMR. The assignments of the aliphatic chain were performed through the interpretation of the fragmentation pattern of the ESI-MS/ MS spectrum of the synthetic alkaloid toxin (trypargine) compared to the fragmentation pattern of the natural toxins, as described below. The precursor ion of m/z 316 [MCH]C (fraction 7.1) was selected and submitted to ESI-MS/MS analysis under CID conditions. The MS/MS spectrum is shown in Fig. 3(b), where it is possible to observe the characteristic fragment ions of m/z values: 43, 58, 72, 86, 100, 128, 170, 187, 215, 229, 243, 257, 272, 300 and 316. The interpretation of the patterns of fragmentation of the synthetic trypargine and of the natural compound from fraction 7.1 is presented in Fig. 4. Despite the chemical structures of the tetrahydro-b-carbolines compounds being relatively simple, sometimes the fragmentation pattern of these compounds in the mass spectrometer is complex, due to the frequent neutral hydrogen rearrangements that occur in the alkyl chain substituent of the chromophore,

792

L.M.M. Cesar et al. / Toxicon 46 (2005) 786–796

Fig. 3. ESI-MS/MS spectra obtained with a ramp of energy collision from 10 to 40 eV. (a): synthetic trypargine and (b): compound present in the fraction 7.1 (m/z 316) as [MCH]C ion, purified from the venom of the spider P. bistriata.

and the final m/z value is one or two units of mass lower than the expected hypothetical value. To help in the interpretations of these results, the mass spectra fragmentation pattern for the compound present in fraction 7.1 was compared to the fragmentation pattern of synthetic trypargine (Fig. 3(a)) (m/z 272 as [MCH]C) with the characteristic fragment ions of m/z values: 43, 58, 70, 85, 99, 113, 158, 172, 181, 196, 213 and 255. First of all, a comparison between the ESI-MS/MS spectra of the natural toxin and synthetic trypargine reveals a high similarity, suggesting a high level of structural similarity between both alkaloid toxins. The interpretation of the fragmentation pattern of synthetic trypargine (Fig. 4(a)) shows some characteristics that must be emphasized: (i) the occurrence of sequential and successive fragmentations in the alkyl chain located between the chromophore and the guanidine group (probably influenced by the strong basicity of the guanidine group). Thus, the ion fragments of m/z 43, 58, 70, 85 and 99 observed in the MS/ MS spectrum of the synthetic trypargine (Fig. 4(a)) permit to assign the chemical structure of the alkyl substituent as a propyl-guanidino; (ii) the ion fragment of m/z 113 (Fig. 3(a)) may be used to assign the connectivity between the propylguanidine substituent and the chromophore moiety for the synthetic trypargine. By analogy, the ion-fragments of m/z 43, 58, 71, 85 and 100 were observed for natural toxin (Fig. 4(b)) and may be

used to assign the presence of an n-butoxyl-guanidine moiety as one of the substituents of the chromophore group of the natural toxin. The fragment-ion of m/z 187, presenting a charge retention on the chromophore side, suggests the presence of a hydroxylated b-carboline moiety for the natural toxin, which is corroborated by the fragment ion of m/z 170, which suggests the lost of the hydroxyl group from the chromophore (Fig. 4(b)) as already determined by 1H NMR analysis. Therefore, the natural toxin seems to correspond to the compound 1-4-guanidinobutoxy-6-hydroxy-1,2,3,4-tetrahydro-b-carboline, which will be referred to PwTx-II. Its structure is shown in Fig. 5. This compound belongs to the family of the b-carboline toxins, isolated for the first time from the skin of the frog K. senegalensis, originally named trypargine, which is a 1,2,3,4-tetrahydro-b-carboline linked to a guanidine group (Shimizu et al., 1982a,b). It is characterized as an optically active metabolite; derived from tryptophan. Two b-carboline compounds were recently reported as spider toxins, one from the venom of P. bistriata [N-[2-(6-hydroxy-2,3,4,9-tetrahydro-1H-b-carbolin-1-yl) ethyl] guanidine] named PwTx-I (Cesar et al., 2005) and another one from the droplets of the capture web from Nephila clavipes [1,2-guanidinoethyl-3-hydroxymethyl-1,2,3,4-tetrahydro-b-carboline]

L.M.M. Cesar et al. / Toxicon 46 (2005) 786–796

793

(a) 99

+

196 85 181 70

NH

43 58 NH

NH

NH2

113 NH

158 172

255

213

H 229

(b) 215 187

300

243

+

272 257

170

NH

HO

NH NH

NH

NH2

O 43 58 72

128

86 100

H

Fig. 4. Interpretation of the fragmentation pattern, of the ESI-MS/MS spectra for: (a): the synthetic toxin trypargine. (b): for the natural bcarboline toxin PwTx-II purified from the venom of P. bistriata.

(Marques et al., 2005), named NWTx-I (compounds 1 and 4 of Fig. 5). In order to characterize this novel spider toxin, its toxicity was assayed against insects and mammals. The toxicity of PwTx-II for insects was determined using honeybees as a model insect as previously described elsewhere (Manzoli-Palma et al., 2003); the LD50 value determined for PwTx-II is 12G3 ng/mg honeybee, (P! 0.05). This value is comparable to that observed for NWTx-I (9G2 ng mgK1 honeybee) (Marques et al., 2005) and smaller than that reported for PwTx-I (29G4 hg mgK1 of honeybee) (Cesar et al., 2005). In a previous work it was also reported the paralyzing effects of the crude venom of P. bistriata in termites (Fontana et al., 2000); it was observed that the termites became progressively and irreversibly immobilized, or they even died after the injection, in a dose-dependent manner with the a value of LD50Z178 ng mgK1. Thus, apparently honeybees seems to

be more sensitive than termites to P. bistriata crude venom; this observation can be explained by the fact that honeybees are natural preys of P. bistriata, while termites are not preyed by this spider. These LD50 values also suggest that the b-carboline toxins are more lethal to insects than most of crude venoms from wandering spiders (Manzoli-Palma et al., 2003) and certainly must contribute to the insect toxicity of the venom of P. bistriata for the prey capture. The LD50 value of 9.75 mg kgK1 was determined for Wistar rats after i.c.v injection of PwTx-II. During the standardization of the doses of the toxin through i.c.v application, the group of rats which received the equivalent to 2 mg KgK1, started to show convulsions 5 min after toxin application, characterized by tonic-clonic crises, which lasted up to 90 min. When doses equivalents to the LD50 were injected, the appearance of clinical symptoms followed the following time course: after

794

L.M.M. Cesar et al. / Toxicon 46 (2005) 786–796

HO

NH

NH NH

NH2

NH

NH

NH2

NH NH

(1)

NH NH

NH2

HO

NH (3)

NH

NH

NH NH

NH

NH

HO

(2)

NH2

(4)

NH

NH NH

NH

NH2

O

1,4-guanidinebutoxy-6-hydroxy-1,2,3,4-tetrahydro-β-carboline Fig. 5. Chemical structure and IUPAC nomenclature of the toxin PwTx-II isolated from the venom of the spider P. bistriata and the structures of other b-carboline toxins from animal origin: hydroxytrypargine, also known as PwTx-I, isolated from the venom of the spider P. bistriata (1): trypargine toxin isolated from the the skin of the African frog K. senegalensis. (2): tripargimine toxin isolated from the ascidian Eudistoma sp. (3): NWTx-I toxin isolated form the web of the spider N. clavipes (4).

30 min, the animal’s fur looked bristle with partially diffuse piloerection, often more localized around the neck and on the head. During the subsequent 50–60 min, the eyelids appeared fully or partially closed with porphyrin accumulation around the eyes. Generally, the contra lateral eyelid closed first. After 60 min the animals lost sensor motor reflexes, as estimated by the righting reflexes and the wire suspension test. They became flaccid, developed severe dyspnoea and died after 90 min, from respiratory failure. At doses lower than the LD50, the same clinical signs appeared in a milder form with a longer appearance time, i.e. 60–70 min after toxins injection. The observed effects were transient and rats fully recovered after 120 min. It was recently demonstrated that the crude venom of P. bistriata may contain convulsant neurotoxins probably inducing limbic seizures in rats (Rodrigues et al., 2001); thus, it is possible that the novel tetrahydro-bcarbolin toxin (PwTx-II), or even the previously described one (PwTx-I), may be involved with the reported convusant effects. Extensive studies carried out by pharmacologists have reported the endogenous formation of b-carbolines in mammals. Tetrahydro-b-carboline compounds were reported in urine, in platelets, in brain and in other tissues (Shimizu et al., 1982a,b). The b-carboline compounds are also found in some plants, such as the liana Banisteriopsis caapi (Malpighiaceae) that contains derivatives of

b-carbolines very similar to harmalan (Hostettmann et al., 1999). According to these authors the b-carbolines are powerful and reversible inhibitors of the monoamine oxidase (MAO), an enzyme involved in the regulation of the physiological levels of some biogenic amines in mammals, such as serotonine and tyramine. The b-carboline compounds usually present a wide spectrum of biological activities such as respiratory failure in mice (Akizawa et al., 1982), interaction with part of the serotonergic neurotransmitters system and blocking of voltage-gated sodium currents in squid axon membrane in a potential-dependent manner (Wagoner et al., 1999). The lethality of b-carbolines to insects and their neurotoxicity to mammals make this class of natural compounds suitable to be used as toxins by some animals, as exemplified in Fig. 5: the compound (1) is the hydroxytrypargine, PwTx-I, used as an insecticidal toxin by the spider P. bistriata; compound (2) is the toxin trypargine, used by the African frog K. senegalensis to prevent predation by snakes; compound (3) is the toxin trypargimine, used by the ascidians Eudistoma sp. to prevent its capture by marine predators; compound (4) is the insecticidal toxin NWTx-I, used by the orb-web spider N. clavipes to capture preys on its web. In the present manuscript we reported the isolation and chemical chracterization of a novel b-carboline toxin, PwTxII, from the venom of the spider P. bistriata. This compound

L.M.M. Cesar et al. / Toxicon 46 (2005) 786–796

is very toxic to insects, neurotoxic, convulsivant and lethal to rats; similarly to typical alkaloids these compounds are members of a novel family of low molecular mass toxins of spider venoms used by P. bistriata for prey capture. A pharmacological study of these toxins is in progress.

Acknowledgements The authors thank the Sa˜o Paulo State Research Foundation (FAPESP) for financial support of this research through the Program BIOprospecTA (proc. 04/07942-2) and for fellowships (to M.A.M., proc. 01/05060-4; C.F.T., proc. 02/12305-6; M.R.M, proc. 00/6879- and B.M.S, proc. 03/00985-5); and the Brazilian Council for Scientific and Technological Development [CNPq] for a scholarship (to L.M.M.C.) and fellowships (to M.S.P.).

References Abramovitch, R.A., Spencer, I.D., 1964. The carbolines. Adv. Heterocycl. Chem. 3, 79–88. Akizawa, T., Yamazaki, K., Yasuhara, T., Nakajima, T., Roseghini, M., Spasmer, G.F., 1982. Trypargine, a new tetrahydro-bcarboline of animal origin: isolation and chemical chracterization from the skin of the African rhacophorid frog, Kassina senegalensis. Biomed. Res. 3, 232–234. Barker, S.A., Harrison, R.E.W., Brown, J.G.B., Christian, S.T., 1981. Identification and quantification of 1,2,3,4-tetrahydro-b-carboline, 2-methyl-1,2,3,4-tetrahydro-b-carboline, and 6-methoxy-1, 2,3,4-tetrahydro-b-carboline as in vivo constituents of rat brain and adrenal gland Biochem. Pharmacol. 30, 9–15. Braun, S., Kalinowski, H.O., Berger, S., 1996. 100 and More Basic NMR Experiments: a Practical Course Course. VCH, New York p. 413. Carmona, C., Ghanem, R., Balo´n, M., Mun˜oz, M.A., Guardado, P., 2000. Chemical and photochemical oxidation of tetrahydrobetacarboline. J. Photochem. Photobiol. A Chem. 135, 171–177. Cesar, L.M.M., Tormena, C.F., Marques, M.R., Silva, G.V.J., Mendes, M.A., Rittner, R., Palma, M.S., 2005. Structure Determination of Hydroxytrypargine: A New Tetrahydro-bCarboline Toxin from the Venom of the Spider Parawixia bistriata. Helv. Chim. Acta 88, 796–801. Chou, J., Chou, T.C., 1987. Dose-effect analysis with microcomputers. Biosoft, Cambridge p.93. Fontana, A.C.K., Cairra˜o, M.A., Colusso, A.J., Santos, W.F., Coutinho-Netto, J., 2000. Activity of the Parawixia bistriata crude venom in a new bioassay in termites. Toxicon 38, 133– 138. Hostettmann, K., Hostettmann, M., Rodriguez, S., Wolfender, J., 1999. LC-hyphenated techiques in the search for new bioactive plant constituents. In: Nakanishi, K., Cooper, R., Snyder, J.K. (Eds.), The Biology-Chemistry Interface, p. 120 (Printed in the United States of America). Johnson, J.V., Yost, R.A., Beck, O., Faull, K.F., 1985. In: Collins, M.A. (Ed.), Aldehyde Adducts in Alcoholism. Alan R. Liss, New York, p. 161.

795

Kawai, N., Nakajima, T., 1993. Neurotoxins from spider venoms. In: Harvey, A.L. (Ed.), Natural and Synthetic Neurotoxins. Acad. Press, London, pp. 319–345. Levi, H.W., 1992. Spiders of orb-weaver genus Parawixia in America (Araneae:araneidae). Bull. Mus. Comp. Zool. 153, 1–46. Manzoli-Palma, M.F., Gobbi, N., Palma, M.S., 2003. Insects as models to assay spider and scorpion venom toxicity. J. Venom. Anim. Toxins Incl. Tropical Diseases 9, 174–185. Marques, M.R., Mendes, M.A., Tormena, C.F., Souza, B.M., Ce´sar, L.M.M., Rittner, R., Palma, M.S., 2005. Structure Determination of a Tetrahydro-b-Carboline of Arthropod Origin: a Novel Alkaloid Toxin Subclass from the Web of Spider. Nephila Clavipes. Chem. Biodiv. 2, 525–534. Marques, M.R., Mendes, M.A., Tormena, C.F., Souza, B.M., Ribeiro, S.P., Rittner, R., Palma, M.S., 2004. Structure determination of an organometallic 1-(diazenaryl) ethanol: a novel toxin subclass from the web of the spider Nephila clavipes. Chem. Biodiv. 830, 838. McCormick, J., McCormick, K., Duysntee, H.I., van Egen, A.K., van der Marel, G.A., Ganem, B., van Boon, J.H., 1999. Structure and total synthesis of HF-7, a neuroactive glyconucleside dissulfate from funnel-web spider Hololena curta. J. Am. Chem. Soc. 121, 5661–5665. N.A.S. 1996. National Academy of Sciences. Guide for the Care and Use of Laboratory Animals; Institute of Laboratory Animal Resources; Commission on Life Sciences; National Research Council, National Academy Press; Washington, D.C.(available on line at: http://www.nap.edu/readingroom/ books/labrats/). Nakanishi, K., Goodnow, R., Konno, K., Niwa, M., Bukownik, R., Kallimpoulus, T.A., Usherwood, P., Eldefrawi, A.T., Eldefrawi, M.E., 1990. Philanthotoxin-433 (PhTX-433), a non-competitive glutamate receptor inhibitor. Pure Appl. Chem. 62, 1223–1230. OECD, 2001. Guidance document on acute oral toxicity testing. OECD Environment, Health and Safety Monograph Series on Testing and Assessment, 4 (available on-line at http://www. oecd.org/esh/). Palma, M.S., Fujita, T., Itagaki, Y., Naoki, H., Nakajima, T., 1997. Mass spectrometric structure determination of spider toxins arginine containing from the venoms of Brazilian Garden spider Nephilengys cruentata. Nat. Toxins. 5, 47–57. Palma, M.S., Nakajima, T., 2005. Natural combinatorial chemistry strategy in acylppolyamines toxins from Nephilinae Orb-Web spiders. Toxin. Rev. 24, 209–234. Quistad, G.B., Lam, W.W., Casida, J.E., 1993. Identification of bis(agmatine) oxalamide in venom from the primitive hunting spider, Plectreuris tristis (Simon). Toxicon 31, 920–924. Robertson, H.A., 1980. Harmaline-induced tremor: the benzodiazepine receptor as a site of action. Eur. J. Pharmacol. 67, 129–132. Rodrigues, M.C.A., Guizzo, R., dos Santos, W.F., Cairasco, N.G., 2001. A comparative neuroethological study of limbic seizures induced by Parawixia bistriata venom and kainic acid injections in rats. Brain Res. Bull. 55, 79–86. Rodrigues, M.C.A., Ribeiro, A.M., Pizzo, A.B., Fontana, A.C.K., Beleboni, R.O., Coutinho-Neto, J., Miranda, A., Santos, W.F., 2002. Anticovulsant and GABA uptake inhibition properties of P. bistriata and S.raptoria spider venom fractions. Pharm. Bull. 55, 79–86.

796

L.M.M. Cesar et al. / Toxicon 46 (2005) 786–796

Rodrigues, M.C.A., Guizzo, R., Gobbo-Neto, L., Ward, R.J., Lopes, N.P., dos Santos, W.F., 2004. The biological activity in mammals and insects of the nucleosidic fraction from the spider Parawixia bistriata. Toxicon 43, 375–383. Shimizu, M., Ishikawa, M., Komoda, Y., Matsubara, Y., Nakajima, T., 1982a. A new synthesis of (G) Trypargine. Chem. Pharm. Bull. 30, 4529–4533. Shimizu, M., Ishikawa, M., Komoda, Y., Nakajima, T., 1982b. Total synthesis of (-) Trypargine. Chem. Pharm. Bull. 30, 909–914.

Skinner, W.S., Dennis, P.A., Robert, A.L., Quistad, G.B., 1990. Chemical characterization of acylpolyamine toxins from venom of a trap-door spider and tarantulas. Toxicon 28, 541–546. Szanta´y, C., Blasko´, G., Honty, K., Do¨rnyei, G., 1986. In: Brossi, A. (Ed.), The Alkaloids, vol. 27. Academic Press, New York, p. 131. Wagoner, R.M.V., Jompa, J., Ireland, C.M., 1999. Trypargine alkaloid from previously undescribed Eudistoma sp. Ascidian. J. Nat. Prod. 62, 794–797.