The effects of specific antibody fragments on the ‘irreversible' neurotoxicity induced by Brown snake (Pseudonaja) venom

Abstract

1. Brown snake (Pseudonaja) venom has been reported to produce ‘irreversible' post synaptic neurotoxicity (Harris & Maltin, 1981; Barnett et al., 1980).

2. A murine phrenic nerve/diaphragm preparation was used to study the neurotoxic effects of this venom and pre- and post-synaptic components were distinguished by varying the temperature and frequency of nerve stimulation. There were no myotoxic effects and the neurotoxicity proved irreversible by washing alone.

3. The effects of a new Fab based ovine antivenom have been investigated and proved able to produce a complete, rapid (<1 h) reversal of the neurotoxicity induced by Brown snake venom. A reversal was also possible when the antivenom addition was delayed for a further 60 min.

4. We believe that this is the first time such a reversal has been shown.

Introduction

Australia is home to some of the most poisonous snakes in the world (White & Pounder, 1984). Brown snakes (Pseudonaja sp.) account for most bites and fatalities (Sutherland, 1992; Jelinek & Breheny, 1990) and the clinical features of such envenoming include disseminated intravascular coagulation and, on occasion, generalized neurotoxicity (Judd & White, 1994). The venom contains both pre- and post-synaptic neurotoxins (Barnett et al., 1979) and the pre-synaptic component, textilotoxin, is the most potent and complex toxin ever isolated from a snake venom. This phospholipase A₂ neurotoxin has a molecular weight of approximately 72 kD and comprises five non-covalently linked subunits (Tyler et al., 1987a; Su et al., 1983; Simpson et al., 1993). Its exact mechanism of action remains unknown, but is dependant on both temperature and the frequency of nerve stimulation. In addition, textilotoxin is said to be unaffected by antitoxin treatment once the initial binding phase is completed (Lloyd et al., 1991; Simpson et al., 1993; Hamilton et al., 1980; Cull-Candy et al., 1976).

Two post-synaptic based neurotoxins have also been isolated. One (pseudonajatoxin-a) is reported to bind ‘irreversibly' to nicotinic acetylcholine receptors (Barnett et al., 1980) and is unusually large with 117 amino acid residues, seven disulphide bonds and a molecular weight of 12 kD (Barnett et al., 1979; 1980). The second (pseudonajatoxin-b) blocks acetylcholine receptors on mammalian skeletal muscle only weakly and reversibly (Tyler et al., 1987b).

It is not known which of the three neurotoxins are clinically important, but Harris & Maltin (1981) reported that the neurotoxic effects of whole venom are of a predominantly post-synaptic type lasting, in vivo, for 2–3 days. They also reported that the effects of the venom could not be reversed in vitro, by a conventional antivenom. This has important implications for the treatment of Brown snake envenomation and, therefore, experiments were performed to determine if a new ovine Fab based antivenom could reverse its neurotoxic effects.

Methods

Antivenom

The antivenom was prepared by immunizing sheep with a maximum of 1 mg/28 days of equal amounts of venom from four Pseudonaja species; P. textilis, P. nuchalis, P. affinis and P. inframacula, in order to maximize the cross reactivity of the final product (Venom Supplies Limited, P.O. Box 547, Tanuda 5352, South Australia). Ten ml of blood kg⁻¹ body weight was collected from each sheep 2 weeks after each immunization, once adequate antibody titres had been achieved (30 weeks post primary immunization), and the serum was separated and its IgG fraction partially purified by precipitation with sodium sulphate (final concentration 18%) at 25°C for 15 min with mixing. The precipitate, after pelleting by centrifugation, was washed twice with 18% sodium sulphate, reconstituted in saline (0.9% sodium chloride) and the IgG partially digested with papain to yield Fab fragments (Rawat et al., 1994). Commercially available CSL Brown Snake Antivenom (CSL F(ab′)₂) was purchased from CSL Ltd., Melbourne, Australia, and dialyzed against saline to remove the cresol preservative.

In vitro neurotoxicity

Left phrenic nerve-hemidiaphragm preparations were isolated from 20–35 g male, out-bred white mice (Kitchen, 1984; Bulbring, 1946). Each preparation was bathed in 60 ml Krebs buffer (mM 118 NaCl; 25 NaHCO₃; 1.0 NaH₂PO₄.H₂O; 4.8 KCl; 1.9 CaCl₂.2H₂O; 1.2 MgSO₄.7H₂O; 11.1 D glucose) maintained at 32°C, and supplied with 95% O₂+5% CO₂. A silicone anti-foaming agent (Sigma) diluted 1 in 60,000 was added to prevent excess foaming in the tissue bath. Indirect stimulation (via the nerve) was supplied using a supramaximal voltage (∼3 V, 0.2 Hz, 0.2 ms) and muscle contractions were recorded using an isometric transducer linked to a Lectromed Ltd amplifier (HF cut 150 Hz) and chart recorder (MX216–100), with tissue holder-electrodes from Harvard Apparatus Ltd. This was continued for at least 30 min until a consistent response was produced after which venom was added (t₀), and washed out after 30 min (t₃₀) and again after a further 30 min (t₆₀). The decrease in contractions was calculated as a percentage based on the extent of the contractions just before venom addition. The myotoxic effects of the venom were assessed by applying a short burst of direct (muscle) stimulation (0.2 Hz, 1 ms, ∼50 V) before venom addition and at 30 min intervals thereafter.

The ability of the antivenom to neutralize the neurotoxic effects of the venom was assessed by mixing a fixed venom concentration (5 mg l⁻¹) with antivenom and incubating at 37°C for 30 min before addition to the hemidiaphragm preparation (t₀). An identical dose cycle was used to that shown above. The ability of the antivenom to reverse neurotoxicity was assessed by exposing the preparation to venom (5 mg l⁻¹) for 30 min before washing and then replacing the bathing solution with Krebs buffer containing antivenom for the remainder of the experiment. Late reversal of neurotoxicity was assessed by exposing the preparation to venom as above, washing after 30 min (t₃₀) and stimulating for a further 60 min before replacing the bathing solution with Krebs buffer containing antivenom (t₉₀) for the remainder of the experiment. Finally reversal of neurotoxicity by antivenom was assessed under more favourable conditions for pre-synaptic (textilotoxin) neurotoxicity, namely by stimulation at a higher temperature (37°C) and frequency (1.0 Hz).

Control responses under identical conditions but without venom or antivenom were also performed.

In vivo toxicity

In vivo toxicity and neutralization was determined by intravenous LD₅₀ and ED₅₀ assays (Theakston & Reid, 1983; Laing et al., 1992) using 18–20 g male mice.

Results

Venom effects on the phrenic nerve/diaphragm

A dose of 5 mg l⁻¹ venom produced a rapid neurotoxic effect (Figure 1) with 95% blockade of the contractions within 30 min which was unaffected by two washes with Krebs buffer. No direct (myotoxic) action was found at this, or higher venom concentrations (25 mg l⁻¹, data not shown).

Figure 1.

Effects of Pseudonaja venom (5 mg l⁻¹) on the mouse phrenic nerve/diaphragm at 32°C, with a stimulation frequency of 0.2 Hz (±s.e.mean, n=5), with controls (n=4).

Neutralization of neurotoxicity by premixing venom and antivenom

Antivenom at a concentration of 200 mg l⁻¹ Fab produced complete neutralization of the venom (5 mg l⁻¹) neurotoxicity (Figure 2). Halving the antivenom concentration to 100 mg l⁻¹ Fab resulted in only a 40–60% blockade of the response which was reversed by washing. Complete neutralization using 200 mg l⁻¹ Fab antivenom was also possible at a higher temperature and frequency (37°C 1.0 Hz⁻¹; data not shown). A concentration of 400 mg l⁻¹ of CSL F(ab′)₂ antivenom resulted in a 40–50% blockade of the response which was reversed by washing.

Figure 2.

Neutralization of the neurotoxicity of Pseudonaja venom at 32°C with a stimulating frequency of 0.2 Hz via the nerve (±s.e.mean). Five mg l⁻¹ venom alone (n=5) and premixed with Fab antivenom (n=4) or CSL F(ab′)₂ (n=3). Control without venom or antivenom (n=4).

Reversal of neurotoxicity

Antivenom, at a concentration of 200 mg l⁻¹ of Fab, produced complete reversal of neurotoxicity within 50 min of its addition (Figure 3) and an equimolar concentration of specific IgG (600 mg l⁻¹) produced a similar effect (data not shown). A concentration of 400 mg l⁻¹of CSL F(ab′)₂ antivenom produced a significantly slower reversal. The effect of the Fab antivenom was slightly slower and less marked (94% control) at 37°C and, when combined with a high stimulating frequency (1.0 Hz, 37°C), only reversed 44% of the blockade (Figure 4).

Figure 3.

Antivenom reversal of Pseudonaja venom neurotoxicity at 32°C with a stimulating frequency of 0.2 Hz via the nerve (±s.e.mean). Venom (5 mg l⁻¹) induced neurotoxicity with antivenom added after 30 min, 200 mg l⁻¹ Fab (n=4) or 400 mg l⁻¹ CSL F(ab′)₂ (n=4) and the delayed addition of 200 mg l⁻¹ Fab after 90 min (n=4). Control without venom or antivenom (n=4).

Figure 4.

Effects of temperature and stimulation frequency on the reversal of neurotoxicity. Antivenom addition 30 min (open bars) and 90 min after venom (hatched bars), ±s.e.mean (n=4).

Late reversal of neurotoxicity

Addition of antivenom (200 mg l⁻¹ Fab) under the standard conditions (0.2 Hz, 32°C) 90 min after venom again produced complete reversal of neurotoxicity (Figure 3), while an 81% reversal was achieved when the temperature was raised to 37°C (Figure 4). When the higher temperature is combined with stimulation at 1.0 Hz only a 29% reversal could be attained (Figure 4).

In vivo toxicity

In vivo the venom had an LD₅₀=47 μg kg⁻¹ (95% confidence limits from probit analysis=26–79). The ovine Fab based antivenom had an ED₅₀ value of 74 mg kg⁻¹ against 2×LD₅₀ (95% confidence limits=47–100). Commercially available equine CSL F(ab′)₂ based antivenom had an ED₅₀ value of 626 mg kg⁻¹ against 2×LD₅₀ (95% confidence limits=463–789).

Discussion

In this study Brown snake venom, in agreement with previous reports, caused no myotoxicity but effectively complete neurotoxicity that could not be reversed by washing (Sutherland et al., 1981; Harris & Maltin, 1981; Su et al., 1983).

After pre-mixing, 200 mg l⁻¹ of the ovine Fab antivenom appeared to neutralize completely all the toxic components of the venom both in vitro and in vivo. Reducing the concentration in vitro to 100 mg l⁻¹ resulted in a transitory and partial reduction of the twitch response which could be reversed to control levels by washing. A higher concentration (400 mg l⁻¹) of CSL F(ab′)₂ antivenom produced similar findings.

Harris & Maltin (1981) demonstrated, by measuring endplate potentials, that Brown snake venom neurotoxicity was predominantly of a post synaptic type and, in contrast to the present studies, could not be reversed by the delayed addition of antivenom despite preventing the development of neurotoxicity when added 10 min before the venom. No apparent explanation can be found for this difference, however, antivenom which still contained the preservative cresol was used by Harris & Maltin.

We have shown for the first time that sufficient amounts of an antivenom can rapidly (<1 h) and totally reverse the neurotoxicity produced by this venom. This reversal could also be demonstrated following the late addition of antivenom, an important factor in effective snake bite therapy. A slower reversal could also be produced by the CSL F(ab′)₂ antivenom using a higher concentration (400 mg l⁻¹).

A fast antibody induced reversal of neurotoxicity has previously been described for a post synaptic neurotoxin (toxin α) purified from spitting cobra (Naja nigricollis) venom, and is unique to antibodies raised to part of the toxin which is distinct from the toxic site (Boulain & Menz 1982; Boulain et al., 1982; 1985; Gatineau et al., 1988; Guenneugues et al., 1997; Menez et al., 1982; 1984). Due to the rapid rate of reversal of Brown snake venom neurotoxicity and the polyclonal nature of the antivenoms, a similar mechanism of reversal is suggested, with a subset of antibody clones specific for an epitope on the toxin(s) which is distinct from the toxic site, resulting in a modification of a flexible region and destablizing the toxin receptor complex (Boulain & Menz 1982; Boulain et al., 1982; 1985; Gatineau et al., 1988; Guenneugues et al., 1997; Menez et al., 1982; 1984; Zinn-Justin et al., 1993). This is despite the unusual structure and large size (12 kD) of pseudonajatoxin-a, the venom's predominant post-synaptic toxin (Barnett et al., 1979; 1980; Tyler et al., 1987b).

Our findings indicate that although the antivenom could produce complete reversal of neurotoxicity normally (0.2 Hz, 32°C), it was possible to reveal the presence of an irreversible neurotoxic component at an increased temperature (37°C) and with high frequency nerve stimulation (1.0 Hz). Presumably this must represent only a minor venom constituent. Pre mixing experiments under these conditions showed that sufficient antibody was present to fully neutralize this component. These irreversible features are highly indicative of the pre-synaptic phospholipase neurotoxin, textilotoxin (Simpson et al., 1993; Lloyd et al., 1991; Su et al., 1983).

The new Fab based antivenom was found to be 8.5 times more effective in vivo and more than twice as effective in vitro compared to the current clinical treatment (CSL F(ab′)₂).

The in vitro neurotoxic effects of this venom could also be reversed by the specific IgG. However, in vivo due to their small size, Fab fragments have a different pharmacokinetic profile and are able to quickly penetrate the interstitial space resulting in a greater volume of distribution than intact IgG (Smith et al., 1979). This, it is hoped, will allow a more rapid transfer to Fab antibody into the synapse than can be achieved with conventional IgG or F(ab′)₂ based antivenoms, and would be more likely to result in a quick reversal of neurotoxicity.

In conclusion, the venom is devoid of myotoxic effects, and the ability of an antivenom to produce a full and rapid (<1 h) reversal of Brown snake venom induced neurotoxicity, is shown here for the first time.

Abbreviations

CSL F(ab′)₂

CSL Brown Snake Antivenom