PURIFICATION AND N-TERMINAL SEQUENCING OF TWO PRESYNAPTIC NEUROTOXIC PLA

Two presynaptic phospholipases A2 (PLA2), neuwieditoxin-I (NeuTX-I) and neuwieditoxin-II (NeuTX-II), were isolated from the venom of Bothrops neuwiedi pauloensis (BNP). The venom was fractionated using molecular exclusion HPLC (Protein-Pak 300SW column), followed by reverse phase HPLC (μBondapak C18 column). Tricine-SDS-PAGE in the presence or absence of dithiothreitol showed that NeuTX-I and NeuTX-II had a molecular mass of approximately 14 kDa and 28kDa, respectively. At 10μg/ml, both toxins produced complete neuromuscular blockade in indirectly stimulated chick biventer cervicis isolated preparation without inhibiting the response to acetylcholine, but NeuTX-II reduced the response to KCl by 67.0±8.0% (n=3; p<0.05). NeuTX-I and NeuTX-II are probably responsible for the presynaptic neurotoxicity of BNP venom in vitro. In fact, using loose patch clamp technique for mouse phrenic nerve-diaphragm preparation, NeuTX-I produced a calcium-dependent blockade of acetylcholine release and caused appearance of giant miniature end-plate potentials (mepps), indicating a pure presynaptic action. The N-terminal sequence of NeuTX-I was DLVQFGQMILKVAGRSLPKSYGAYGCYCGWGGRGK (71% homology with bothropstoxin-II and 54% homology with caudoxin) and that of NeuTX-II was SLFEFAKMILEETKRLPFPYYGAYGCYCGWGGQGQPKDAT (92% homology with Basp-III and 62% homology with crotoxin PLA2). The fact that NeuTX-I has Q-4 (Gln-4) and both toxins have F-5 (Phe-5) and Y-28 (Tyr-28) strongly suggests that NeuTX-I and NeuTX-II are Asp49 PLA2.


INTRODUCTION
Out of the four genera of venomous snakes in Brazil (Bothrops, Crotalus, Lachesis and Micrurus), only Crotalus and Micrurus cause failure of the neuromuscular junction, producing peripheral muscular weakness.However, the main complications in lethal envenomations by Bothrops species are acute renal failure, shock and sepsis (6,41).Respiratory failure is also observed sometimes but the mechanism is not well understood.Since Bothrops venoms produce no signs of neurotoxicity after snakebite, respiratory failure has been related to pulmonary edema (41).
The neuromuscular actions of various Bothrops venoms have been extensively investigated in recent years.Cogo et al. (9) reported the neurotoxicity of Bothrops insularis venom on mice and chicks and on mouse phrenic nerve-diaphragm (MPND) and chick biventer cervicis (CBC) isolated preparations.After a lethal intramuscular (i.m.) dose of venom, chicks showed flaccidity followed by head-drop, dyspnea, apnea, convulsions and death.The venom produced irreversible blockade of MPND and CBC, and increased the frequency of mepps in mouse diaphragm.
Subsequently, a neurotoxic presynaptic PLA 2 fraction, which blocked indirectlyevoked twitches in CBC but did not affect the responses to acetylcholine (ACh) and KCl, was isolated from this venom (10).
Zamuner et al. (53) reported that B. neuwiedi venom caused head-drop, loss of balance and respiratory failure in chicks, after an i.m. dose of 0.55 mg/kg.The venom also inhibited the twitch-tension responses in CBC preparations, but did not decrease the contractures to exogenous ACh or KCl, thus suggesting a presynaptic action.
Borja-Oliveira et al. (5) subsequently investigated the neurotoxicity of several lots of B. neuwiedi venoms on CBC and observed that not all B. neuwiedi venoms exhibited neurotoxicity.Furthermore, analysis of the neurotoxic action of a sample of B. neuwiedi pauloensis (BNP) venom on CBC showed that the presynaptic action of the venom was temperature-dependent (4).More recently, we have observed that the presynaptic neurotoxicity of BNP on MPND is Ca 2+ -dependent and accompanied by a pronounced increase in the frequency of mepps and the presence of giant mepps (12).PLA 2 (E.C. 3.1.14)are Ca 2+ -dependent enzymes that catalyze the hydrolysis of phospholipids at position sn-2 (51).Snake venom PLA 2 occurs in groups I (Elapidae and Hydrophidae) and II (Viperidae).Group II contains the catalytically-active Asp49 PLA 2 , as well as the Lys49 isoforms.The latter type of proteins are considered PLA 2 homologues since they are probably devoid of (or may have very low) catalytic activity.In addition to their enzymatic activity, venom PLA 2 may have other actions, such as neurotoxicity and myotoxicity (25).
In the present work, two presynaptic neurotoxic PLA 2 , named neuwieditoxin-I (NeuTX-I) and neuwieditoxin-II (NeuTX-II), were isolated from BNP venom and partially sequenced.This is the first time to our knowledge that a pure presynaptically active neurotoxin such as NeuTX-I has been isolated from botropic venom and partially sequenced.

Reagents and venom
Acetylcholine chloride, solvents (HPLC grade) and other reagents were obtained from Sigma and Aldrich Chemical Company (St Louis, MO, USA).Bothrops neuwiedi pauloensis venom, collected in the state of São Paulo, was provided by Butantan Institute (São Paulo, São Paulo State, Brazil).The venom was obtained from adult specimens, and was desiccated and stored at 2-4°C.

Purification of Neuwieditoxins I and II
Molecular exclusion HPLC: Bothrops neuwiedi pauloensis venom (25mg) was loaded onto a Protein-Pak 300SW (Waters, Milford, MA, USA) column (0.78cm X 30cm) and eluted with 0.25M ammonium bicarbonate, pH 7.9, at a flow rate of 0.3ml/min.The elution profile was monitored at 280nm using a Waters 490 programmable multiwavelength detector (Waters, Milford, MA, USA); fractions were collected using a Foxy 200 automatic fraction collector (Isco, Inc., Lincoln, NE, USA).The chromatographic runs were recorded on a 746 data module dual channel recorder (Waters, Milford, MA, USA).
Reverse-phase HPLC: Lyophilized fractions from molecular exclusion HPLC were screened for neuromuscular activities and those of interest were then fractionated using a μBondapak C18 (Waters, MA, USA) column (0.78cm X 30cm).The column was equilibrated with buffer A (0.1% trifluoroacetic acid -TFA, in water) and the proteins were eluted with a linear gradient of buffer B (0.035% TFA in 66% acetonitrile) at a flow rate of 1.0ml/min.The elution profile was monitored at 280nm.
The fractions from reverse-phase HPLC were screened for neuromuscular activities.
Those of interest were then pooled, lyophilized and stored at -20°C.
Phospholipase A 2 activity PLA 2 activity was measured using the assay previously described by Cho and Kezdy (7) and Holzer and Mackessy (20), modified for 96-well plates (3).The enzyme activity, expressed as the reaction initial velocity (Vo), was calculated based on the increase in absorbance after 20 min.Absorbances at 425nm were measured using a SpectraMax 340 multiwell plate reader (Molecular Devices, Sunnyvale, CA, USA).

Electrophoresis
Tricine-PAGE in a discontinuous gel and buffer system (45) was used to estimate the molecular mass of the proteins, in the presence or absence of dithiothreitol (DTT).
Marker proteins were included in the runs and the gels were stained with coomassie blue.

N-terminal sequencing
Direct sequencing of the N-terminal was done with reduced and carboxymethylated protein, using a Procise protein sequencing system (Applied Biosystems, CA, USA).
Phenylthiohydantoin amino acids were identified by comparing their retention times with those of 20 phenylthiohydatoin amino acid standards.
Contractures to exogenously applied submaximal concentrations of acetylcholine (110μM) and KCl (13.4mΜ) were obtained in the absence of nerve stimulation prior to the addition of toxins and at the end of the experiment in order to test for the presence of neurotoxic and myotoxic activities (18).

Electrophysiological procedure
Preparation: Left hemidiaphragm of mouse was prepared as previously described by Re et al. (37,38).Briefly, Charles River male mice, 30-40 days old, were killed and the left hemidiaphragm was dissected together with a short length of phrenic nerve.
The preparation was bathed in Krebs' solution of the following composition (mM): , pH 7.4, gassed with 95% O 2 -5% CO 2 and kept at room temperature (18-22°C).Concentrations of MgCl 2 (5 to 15mM) and CaCl 2 (0.9 to 2mM) were adjusted in order to abolish the twitch of the muscle fiber.
The muscle was pinned on Sylgard resin, and placed on the stage of a Leitz inverted microscope.End-plates were visible by transillumination of the preparation using an optic fiber system.The preparation was equilibrated in saline for 30 minutes before starting the experiments.
End-plate signals: Spontaneous and evoked end-plate currents were recorded using a focal extracellular pipette pressed against the edge of an end-plate.Drummond 100µl measuring pipettes of soft glass (1.4mm) were pulled with a Kopf 700C Puller, which had a final tip diameter ranging from 3 to 15µm and was fire polished with a Narishighe MF83 Microforge.After filling with physiological solution, the electrode was connected to a List LM EPC7 current to voltage converter.Pipette resistances were of 100-300KΩ, and seal resistances, measured after pressing the pipette against the sarcolemma, ranged between 300-600KΩ.The loose patch clamp method (47) enables a good control of the series resistance all over the experiment (40).Furthermore, the voltage of the muscle fiber is well monitored using the same technique and virtually clamped at the resting value.Indirect stimulation of the The parameters were obtained using the protocol previously described by Re et al. (38).The technique led to automatic evaluation of the mean τ value of the mepc's decay, the mean peak value of the miniature events (mepc), the spontaneous release frequency (f), and the mean epc's peak value (epc).The resting membrane potential (RMP) was measured using conventional intracellular microelectrodes (3M KCl) connected to a P16 differential amplifier (GRASS).

Statistical analysis
Each experiment was repeated at least three times.Results were expressed as mean ± standard error of mean (S.E.M.).Student's t-test was used for statistical analysis of the data, and values of p<0.05 indicated significance.

RESULTS
Bothrops neuwiedi pauloensis (BNP) venom was initially separated into 8 fractions by molecular exclusion (Figure 1A).Since the venom contains PLA 2 activity (12.7nmol/min/mg), this enzyme was assayed along the venom elution profile.Only The N-terminal sequences were determined up to the 36 th for NeuTX-I, and up to the 41 st for NeuTX-II amino acid residues (Figures 3 and 4).
To better clarify the possible molecular event involved in the NeuTX-I (the only toxin that induced complete neuromuscular blockade without inhibiting ACh and KClinduced contracture) action, we studied the effects of the toxin (5, 10 and 50μg/ml) on mouse neuromuscular junction using the loose patch clamp technique (LPC).The results obtained were summarized in Table 1.The toxin induced an increase in the amplitude and frequency of the spontaneous release.Giant mepc's were frequently observed at NeuTX-I concentration of 10μg/ml (Figure 5).After this apparent facilitation, a complete block of the evoked release was always observed (Figures 6   a, b, c).Usually, the recovery of the evoked signals was incomplete after the final wash out.However, the block was overcome by adding calcium to the physiological solution (Figures 6 d, e).
During treatments with NeuTX-I, resting membrane potential values were not different from those obtained during control (toxin-free medium) experiments.* Indicates non-significant differences.Values in bold indicate significant mean with p<0.05 .
Presynaptic inhibition is conveniently studied using dissected BCP preparation.A pure presynaptically active neurotoxin would abolish nerve-evoked twitches, without affecting responses to cholinoceptor agonists or the responses to direct muscle stimulation, but would not affect the responses to elevated K + concentration (18).As a pure presynaptically neurotoxin, NeuTX-I induced complete neuromuscular blockade without inhibiting the responses to ACh and KCl.Consequently, its effects were studied using loose patch clamp technique on a mammalian neuromuscular junction.The electrophysiological data were indicative of a pure presynaptic effect.
After an initial facilitation leading to depletion of ACh release, a complete calciumdependent blockade was observed.The presynaptic action is confirmed by the constancy of the mepc's decay time, i.e. the rate constant of the receptor-ion channel complex conformational change, reflecting the closing of the ACh-sensitive channels (24,29) at the postsynaptic sites.
In general, snake venoms contain a mixture of toxins active at prejunctional and postjunctional sites, as well as myotoxic components.Therefore, probably more than one component is capable of contributing to the neuromuscular action produced by the venom.Myotoxic components reduce the response of biventer cervicis muscle to exposure to elevated K + concentration and/or initiate contractures in the muscle (18).
These effects were produced by BNP venom (4) and NeuTX-II.
In previous reports, the neuromuscular blocking effect of BNP venom in BCP and MPND was temperature and Ca 2+ -dependent and this fact also indicated that enzymatic activity is necessary for neuromuscular action (4,12).The neuromuscular blocking effect of the fraction corresponding to peak 3, from the molecular exclusion step, which showed PLA 2 activity and revealed NeuTX-I and NeuTX-II after reversephase HPLC, was temperature-dependent, indicating that its neurotoxicity may require enzymatic activity, as suggested for the venom (4).This is the first report of purification and N-terminal sequence of presynaptic PLA  Although the neurotoxic action of Bothrops venoms is not a clinical problem, an understanding of the effects of other species may provide insights into a variety of physiological processes.Snake presynaptic neurotoxins with PLA 2 activity affect the release of ACh and block the neurotransmission in an unknown way.Once their mechanism of action is firmly established, presynaptic PLA 2 neurotoxins could be employed as tools to investigate specific aspects of neuroexocytosis.In addition, the isolation, identification, and pharmacological and biochemical characterization of PLA 2 enzymes help to understand their structure-function relationships, which can be useful in developing prototypes of novel tools and pharmaceutical drugs (33).

Figure 1 .
Figure 1.Purification of NeuTXs-I and II.A: Molecular exclusion HPLC of Bothrops neuwiedi pauloensis venom (25mg).Fractions were collected and the elution profile was monitored at 280nm.B: Repurification of fraction 3 by reverse phase HPLC.NeuTX-I (*) was recovered in peak c and NeuTX-II (**) was recovered in peak b.The inset shows a Tricine-SDS-PAGE gel of the toxins in the absence and presence of dithiothreitol (DTT).

Figure 2 .
Figure 2. Neuromuscular blockade of chick biventer cervicis preparations at 37°C by NeuTXs-I and II.A: Time-dependent blockade produced by each toxin (10μg/ml), on indirectly stimulated preparations, compared to the Krebs control (*p<0.05).Each point represents the mean ± S.E.M. of 3-5 experiments.B and C: Representative recordings of the responses of indirectly stimulated chick preparation to NeuTX-I and NeuTX-II (10 μg/ml), respectively.Responses to exogenous acetylcholine (▲, 110μM) and KCl ( , 13.4mM) were obtained before and after the addition of toxins.Note that, in contrast to the response to KCl, response to acetylcholine after the blockade by NeuTX-II was still identical to the control.This result is representative of 3 experiments.D: Muscle contractures to KCl (13.4mM) and acetylcholine (ACh, 110μM) after 120min incubation with each toxin (10μg/ml) (mean ± S.E.M. of 3-5 experiments), expressed as a percentage of the pre-toxin values (*p<0.05).

Figure 5 .
Figure 5. Spontaneous miniature end-plate currents recorded at mouse neuromuscular junction.The figure shows digitized raw data related to control and treatment with NeuTX-I concentration of 10μg/ml.Note the presence of giant mepc's.Calibration bars, vertical: 250pA, horizontal: 1ms.

Figure 6 .
Figure 6.Averaged signals obtained by 200 evoked end-plate currents recorded at mouse neuromuscular junction.The figure shows digitized raw data related to control (A), treatment with NeuTX-I concentration of 5μg/ml (B) and 10μg/ml (C); the effects obtained after a wash out (D) and a subsequent flow of a physiological solution with calcium concentration increased from 1 to 1.5mM (E).Calibration bars, vertical: 2nA, horizontal: 2ms.

5 )
+1.36 (±7.75; 5)* +0.64 (±15.1;5)* +4.7 (±17.5;5)* -1.84 (±9.42; 5)* Values show the effects of the indicated NeuTX-I concentrations on four parameters related to the function of mouse neuromuscular junction compared with control and final wash out.The amplitudes of the evoked (epc) and spontaneous (mepc) release, the frequency of the quantal release (f) and the mepc's decay time constant (τ) were analyzed.Data were expressed as mean percentage variations in respect to mean control values.Standard deviations and number of experiments are indicated between parentheses.

Table 1 .
Effects of NeuTX-I on mouse neuromuscular junction.
(44)cent comparative study of the pharmacological activities of B. insularis, B. neuwiedi and C. d. terrificus in chick neuromuscular preparations(44)indicated that B. neuwiedi and B. insularis venoms may contain components that act C