Proc. Nati. Acad. Sci. USA Vol. 83, pp. 493-497, January 1986 Neurobiology Central action of dendrotoxin: Selective reduction of a transient K conductance in hippocampus and binding to localized acceptors (K channels/facilitatory neurotoxin/neuronal binding protein) JAMES V. HALLIWELL*, IEKHSAN B. OTHMANt, ANNEGRET PELCHEN-MATTHEWSt, AND J. OLIVER DOLLYtt *Medical Research Council Neuropharmacology Research Group, School of Pharmacy, London WC1 lAX, and tDepartment of Biochemistry, Imperial College, London SW7 2AZ, United Kingdom Communicated by Sir Andrew Huxley, September 3, 1985 ABSTRACT Dendrotoxin, a small single-chain protein MATERIALS AND METHODS from the venom of Dendroaspis angusticeps, is highly toxic Purification and Radioiodination of DTX. The neurotoxin following intracerebroventricular injection into rats. Voltage- was purified to homogeneity from the venom of D. angusti- clamp analysis of CA1 neurons in hippocampal slices, treated ceps (Sigma), characterized as detailed (8) and 'l25-iodinated with tetrodotoxin, revealed that nanomolar concentrations of using a modification of the chloramine-T method that avoids dendrotoxin reduce selectively a transient, voltage-dependent loss of neurotoxicity (2). K conductance. Epileptiform activity known to be induced by Electrophysiological Recording in Hippocampal Slices. dendrotoxin can be attributed to such an action. Membrane Methods for preparation and maintenance ofbrain slices from currents not affected directly by the toxin include (i) Ca- guinea pig or rat have been described (9, 10). DTX and activated K conductance; (it) noninactivating voltage-depen- 4-aminopyridine (4AP) were dissolved in Krebs medium and dent K conductance; (Wi) inactivating and noninactivating Ca administered for required periods by superfusion at 280C. conductances; (iv) persistent inward (anomalous) rectifier Cells in the CA1 region of the slice were impaled with glass current. Persistence of the effects of the toxin when Cd was micropipettes filled with 3 M KCl, 4 M KOAc or 3 M CsCl, included to suppress spontaneous transmitter release indicates giving dc resistance values of 50-80, <120, and 40 MW, a direct action on the neuronal membrane. Using biologically respectively. Intracellular recordings were made by using a active, 12'I-labeled dendrotoxin, protein acceptor sites of high single electrode voltage-clamp amplifier (10). Measurements affinity were detected on cerebrocortical synaptosomal mem- of current and voltage were routinely taken at the end of 0.5- branes and sections of rat brain. In hippocampus, toxin or 1-sec clamp steps to obtain steady-state current-voltage binding was shown autoradiographically to reside in synapse- relations (CVRs). For the construction ofnear-instantaneous rich and white matter regions, with lower levels in cell body CVRs, measurements were made at the earliest opportunity layers. This acceptor is implicated in the action oftoxin because after the initiation of a voltage jump when the voltage had its affinities for dendrotoxin congeners are proportional to their settled to within 1 mV ofthe steady-state command potential. central neurotoxicities and potencies in reducing the transient, Assay of 125I-Labeled DTX Binding to Synaptosomal Mem- voltage-dependent K conductance. branes and Brain Sections. Freeze-thawed synaptosomes, purified from rat cortex (8), were resuspended in Krebs- phosphate buffer, pH 7.4 ("0.7 mg of protein/ml); where Neurotransmitter release is facilitated at peripheral (1) and stated, Ca2+ was omitted and 1.3 mM EGTA included. central (2, 3) synapses by a homologous group ofsingle-chain Binding of 251I-labeled DTX (1251-DTX) was assayed in neurotoxins (4), which are devoid of known enzymatic duplicate samples by rapid centrifugation of the membranes activity (1). Two of these mamba snake proteins, toxin I and through oil (2, 8); nonspecific binding was determined like- dendrotoxin (DTX), from Dendroaspis polylepis and wise by adding 100-fold excess of DTX. For competition angusticeps, respectively, are highly neurotoxic when inject- experiments, a standard centrifugation assay (11) was used ed intracerebroventricularly into rats (5). Electrophysiolog- where 2.5 nM 1251-DTX was incubated as above with synap- ical recordings in hippocampal slices from guinea pig or rat tosomes in the absence and presence ofpure preparations (2, have demonstrated that DTX is a potent convulsant (2, 3) that 11) of various proteins. 125I-DTX binding to brain sections causes enhancement of cell excitability with a concomitant (see below) was performed by using tissue that was dry potentiation of transmitter release. With the aim of elucidat- mounted onto slides; after labeling with toxin and washing, ing the mechanism of action of these novel facilitatory test and control specimens were detached from the slides polypeptides, we have studied their effects on a range of prior to quantitation of 125I-DTX bound by gamma counting. membrane currents in hippocampal neurons. Our findings Light-Microscope Autoradiography. Lightly fixed whole that the pronounced central and convulsive brain from rats that had been perfused with 0.1% paraform- indicate toxicity aldehyde/phosphate-buffered saline, pH 7.4 was used to cut action of DTX can be attributed to a reduction of a transient cryostat serial sections (10 gm). After labeling with 1251-DTX, K conductance [i.e., the transient, voltage-dependent K the mounted sections were processed for autoradiography current IA (6, 7)]. Moreover, a membrane acceptor protein using LKB Ultrofilm (12). with a high affinity for DTX was identified, and its discrete localization in hippocampus established. It is concluded that Abbreviations: DTX, dendrotoxin; '25I-DTX, 125I-labeled DTX; DTX is a useful pharmacological tool for studying voltage- TTX, tetrodotoxin; CVR, current-voltage relation; 4AP, 4-amino- sensitive K channels responsible for IA, and variants of the pyridine; IA, transient K current; Ic, Ca-activated K current; IM, a latter.§ noninactivating time- and voltage-dependent K current, triggered by depolarizing the cell positive to around -60 mV; IQ, a mixed Na/K current activated by hyperpolarization. The publication costs of this article were defrayed in part by page charge *To whom reprint requests should be addressed. payment. This article must therefore be hereby marked "advertisement" §A preliminary report (2) of this work was presented at the Eighth in accordance with 18 U.S.C. §1734 solely to indicate this fact. Conference on Neurobiology, Nov. 1983, Gif-sur-Yvette, France. 493 Downloaded by guest on September 24, 2021 494 Neurobiology: Halliwell et al. Proc. Natl. Acad. Sci. USA 83 (1986) RESULTS Ca spikes by increasing flux through Ca channels responsible for the two Ca conductances described in these cells (13, 14). Effect of DTX on Evoked and Spontaneo ActivityAtiitin Resistance of Other Currents to DTX. Possible action of Hippocampal Neurons. Tetrodotoxin (TTX) wasas routinely DTX on three other identifiable currents in these cells was administered to eliminate Na action potentia lsand, hence, examined by voltage-clamp, using KCl (or in some cases simplify analysis of single neurons. Twenty-o oe cells in the KOAc) electrodes, in 17 hippocampal neurons. These includ- CA1 region were studied under current- or voltage-clamp ed a noninactivating time- and voltage-dependent K current, conditions; control resting membrane potentiadl was -68 ± 2 triggered by depolarizing the cell positive to around -60 mV mV (±SEM), input-resistance was 60 ± 4 M[el, and where (IM) (10), a Ca-activated K current (Ic), a time- and voltage- measured before the addition of TTX, action potential am- dependent K conductance activated by raised intracellular plitude was 96 ± 4 mV in the cells impaled witih K-containing Ca (16), and a mixed Na/K current (IQ) (10). IM and Ic were pipettes. The control in Fig. 1A shows the volttage responses studied by holding neurons at a potential of about -30 mV of a neuron, in the presence of 0.5 IiM TT7X, to current and making 1-sec jumps to both hyperpolarized and depolar- injection. Depolarizing current elicited broad (presumed ized levels. Hyperpolarization resulted in a time-dependent Ca-dependent) spikes; these were enhanced iin number and turn-off of IM to give an inward current relaxation and evoked with lower current strengths after 1 hr exposure to 50 associated conductance decrease (10); depolarizing steps nM DTX (Fig. LA), while little change in ther responses to initiated a further outward current partly due to Ic (16). After hyperpolarizing current or in the resting memtDrane potential more than 30 min in DTX (50-300 nM), neitherlM norIc were was observed (n = 3). Similar alterations in C2a spikes were affected materially. Holding cells close to rest and then hyperpolarizing them turns on IQ (10), which is responsible observed in six of seven other cells exposed lto higher DTX for anomalous rectification in the steady-state CVR. This, concentrations (250-300 nM); these could biepartially re- was unaffected by DTX. versed by the addition of 300 Cd (a Ca-cha nelpatillyere- too, ,uM DTX Suppresses the 'A in Hippocampal Neurons. The only suggesting involvement of Ca channels in thenedepolarizingdebolarizin consistent effect ofDTX (50-300 nM) on a membrane current response. was apparent when employing the protocol that reveals the DTX Does Not Affect Ca Channels Directly. In three IA; this activates rapidly when the cell membrane potential is neurons impaled with electrodes containing 3 1M