Introduction

Top Summary Introduction Materials & Methods Results Discussion References

High affinity block of calcium channels by a number of polypeptide toxins has aided in the pharmacological dissection of endogenous whole-cell currents present in different neuronal preparations. Toxins also have played an important role in the molecular isolation and characterization of different calcium channel subtypes in expression systems (for review see Olivera et al., 1994). Neuronal calcium channels are multiprotein complexes composed of at least three subunits (1, 2, and ), of which the distinct pore-forming 1 subunit (modulated by 2 and  subunits) determines the major biophysical and pharmacological properties of the channel (Stea et al., 1995b). Most neurons coexpress multiple types of different calcium channel 1 subunits (1A-1E, 1G; Snutch et al., 1990; Soong et al., 1993; Perez-Reyes et al., 1998). Biophysical and pharmacological criteria have been used to isolate individual components of native whole-cell calcium currents (T-, L-, N-, and P/Q-type see Bean, 1989; Zhang et al., 1993) and to identify their cloned counterparts; P/Q-type (1A), N-type (1B), L-type (1C and 1D), and T-type (1G) (Stea et al., 1995b; Perez-Reyes et al., 1998). Irreversible inhibition by -CgTX GVIA, a peptide fraction isolated from the venom of the fish-hunting cone snail Conus geographus (Olivera et al., 1984), has been widely adopted as a defining characteristic of the 1B (N-type) calcium channel (Fox et al., 1987; Dubel et al., 1992; Boland et al., 1994). -Aga IVA, a peptide toxin isolated from the venom of the spider Agelenopsis aperta (Mintz et al., 1992a, 1992b), and another peptide fraction from the cone shell Conus magus, -CTx MVIIC (Hillyard et al., 1992), have been used to identify and investigate the role of 1A (P/Q-type) currents in a number of different neuronal preparations and expression systems (Mori et al., 1991; Turner et al., 1992; Takahashi and Momiyama, 1993; Stea et al., 1994; Randall and Tsien, 1995).

Although peptide toxins have proved to be useful in distinguishing different calcium channel subtypes, the specificity of a number of these compounds is not absolute. For example, -CTx MVIIC inhibits both high threshold Q- and N-type currents (Hillyard et al., 1992; Schwartz et al., 1993). In addition, two other peptide fractions isolated from the venom of A. aperta have been shown to block a variety of different mammalian calcium channels; -Aga IA inhibits T-, L-, and N-type currents in rat dorsal root ganglion neurons (Scott et al., 1990), and -Aga IIIA produces a high affinity inhibition of N-, L-, and P-type channels in neurons as well as in cardiac L-type channels (Mintz et al., 1991; Mintz, 1994).

In the current study, we focused on identifying and characterizing the subtype specificity of a novel peptide toxin fraction, designated DW13.3, isolated from the venom of the spider Filistata hibernalis (DW). We described the extraction and complete amino acid sequence of DW13.3 (molecular weight, 8668; 74 amino acids, 12 cysteines) and investigated DW13.3 block of four different neuronal calcium channels expressed in Xenopus laevis oocytes (1A, 1B, 1C, and 1E). The inhibitory action of DW13.3 in oocytes was also compared and contrasted to that observed in a human embryonic kidney cell line stably expressing 1B together with 2 and 1b (1B HEK cells) and on a number of cultured mammalian neuronal preparations. DW13.3 produced a high affinity block of all the calcium channel currents investigated in this study (with the exception of low threshold T-type currents recorded in GH3 cells). The kinetics of toxin block and the relative efficacy of the toxin varied among preparations. In addition, DW13.3 produced a partial block of the 1A current expressed in oocytes and P-type current recorded in cultured rat Purkinje neurons.

	Materials and Methods

Top Summary Introduction Materials & Methods Results Discussion References

Preparation, purification, and sequence analysis of DW13.3. Venom was obtained by electrical stimulation of the cephalothorax of live CO₂-anesthetized spiders (Bascur et al., 1980), and pooled venom stored at 80°. Mass spectral information was obtained using ES-MS on a Finnigan TSQ-700 mass spectrometer fitted with an Analytica of Branford (Branford, CT) ES ionization source. Samples were dissolved in 0.1% trifluoroacetic acid to a concentration of ~10 µM. The flow rate of this solution into the mass spectrometer was 1 µl/min with a sheath of liquid of 2-methoxyethanol. The instrument was scanned over the range of m/z 600-2200 or 400-1800 in the profile mode, and the resulting data were deconvoluted using Finnigan Biotech software. The resolution of the instrument was set such that average mass data (as opposed to monoisotopic data) were collected. For higher molecular mass samples (>4000 Da), accuracy was within 0.01%, whereas at lower masses, the accuracy was ~0.4 Da. The averaging of multiple molecular ions of different charge states accounts for the improved accuracy at higher mass. Peptide sequencing was performed on a Hewlett-Packard G1005A sequencing system consisting of a G1000A sequencer, an on-line 1090 HPLC unit, and an associated computer system and software. Amino acid analysis was performed with an Applied Biosystems Model 420H hydrolyzer/derivatizer and a model 130 analyzer. Data were acquired and processed using the PE Nelson data system.

Transient expression of calcium channel cDNAs in X. laevis oocytes. Stage V and VI X. laevis oocytes were prepared and nuclear injections with cDNAs were performed as described previously (Stea et al., 1995a). Direct intranuclear injections were undertaken with 10 nl of a mixture of rat brain cDNAs encoding calcium channel 1, 2, and  subunits in a 1:1:1 molar mix. The cDNA constructs have been described previously (Bourinet et al., 1996, and references thererin). Two-microelectrode voltage-clamp experiments were performed after 3-6-day incubations with a GeneClamp 500 amplifier (Axon Instruments, Burlingame, CA). Macroscopic currents were recorded as described (Stea et al., 1995a) in a solution containing 4 mM BaCl ₂, 38 mM KCl, 36 mM TEA-Cl, 5 mM 4-aminopyridine, 0.4 mM niflumic acid, 0.02 mM 5-nitro-2-(3-phenylpropylamine)benzoic acid, 5 mM HEPES, and 500 µg/ml cytochrome c (Sigma Chemical, St. Louis, MO), pH 7.6. The endogenous oocyte calcium-activated Cl current was completely suppressed by the injection of 10-30 nl of a solution containing 100 mM BAPTA-free acid (10 mM HEPES, pH titrated to 7.2 with CsOH) to chelate intracellular calcium (Charnet et al., 1994). Samples of DW13.3 were resuspended as stock solutions (115 µM) in double-distilled water and stored at 20° between experiments. The toxin was dissolved in the recording saline before application and perfused into the bath. An effective exchange of the chamber solution was achieved within 1-2 sec as judged by superperfusion of a solution containing 100 mM Cd⁺ and monitoring the development of block. Unless otherwise stated, currents were elicited by a 0.067-Hz train of 400-msec pulses from a holding potential of 100 mV. Acquisition and data analysis were performed using pCLAMP (ver. 6.03) software (Axon Instruments). Data were filtered at 500 Hz, and in most cases, leak subtraction was carried out on-line using a P/4 protocol.

HEK and GH3 cells. Whole-cell patch-clamp recordings were carried out on a HEK 293 cell line stably expressing 1B, 2, and 1b (Zamponi et al., 1997). Cells were split and plated at ~10% confluency on glass coverslips in DMEM supplemented with 10% fetal bovine serum and 0.4 mg/ml neomycin. GH3 cells were purchased from American Tissue Culture Collection (Rockville, MD), plated at ~30% confluency on glass coverslips, and incubated at 37° in Ham's F-10 medium supplemented with 15% horse serum and 2.5% fetal bovine serum for up to 1 week. Whole-cell patch-clamp recordings were performed using an Axopatch 200-A amplifier (Axon Instruments) linked to a personal computer equipped with pCLAMP v 6.0. Patch pipettes (Sutter borosilicate glass) showed typical resistances of 1.7-3.5 M. The internal pipette solution contained 105 mM CsCl, 25 mM TEA-Cl, 1 mM CaCl₂, 11 mM EGTA, and 10 mM HEPES, pH 7.2. The external recording solution contained 5 mM BaCl₂, 1 mM MgCl₂, 10 mM HEPES, 40 mM TEA-Cl, 10 mM glucose, and 87.5 mM CsCl, pH 7.2. DW13.3 was perfused directly into the vicinity of the cells by means of a solenoid-driven microperfusion system. Currents typically were elicited from a holding potential of 100 mV to various test potentials using Clampex (Axon Instruments). Data were filtered at 1 kHz and recorded directly onto the hard drive of a computer.

Isolation of sympathetic neurons, Purkinje cells, and cerebellar granule cells. Sympathetic neurons were isolated from the superior cervical ganglia of 28-35-day-old male Sprague Dawley rats as described in by Shapiro and Hille (1993) with the exception that after 20 min in papain, the tissue was placed in 3 ml of Hanks' solution containing 3 mg/ml type I collagenase (Worthington Biochemicals, Freehold, NJ) and 10 mg/ml dispase type II (Boehringer-Mannheim, Indianapolis, IN). Purkinje cells from the cerebellar vermis were isolated according to the method of Regan (1991) with the following modifications. Tissue was incubated in 5.5 units/ml papain (Worthington) for 30 min at 32° and then transferred to room-temperature MEM containing 10% FCS and 10 mM glucose. The tissue was cut into ~2-mm pieces and stored in the MEM solution under an atmosphere of 95% O₂/5% CO₂ at room temperature until needed. Cells were isolated for recording by gentle trituration in MEM containing 1 mg/ml DNase. Cerebellar granule cells were prepared by isolating the cerebellum from 8-day-old pups and mincing the tissue into ~1-mm pieces in room-temperature Tyrode's solution. The tissue was washed in calcium/magnesium-free Tyrode's solution and incubated in 0.1% trypsin at 37° for 15 min. After incubation, the tissue was washed in Tyrode's solution with 10% FCS and triturated in the presence of DNase, spun, and plated onto poly-d-lysine-coated glass coverslips. The cells were maintained in an atmosphere of 5% CO₂ at 37° in MEM with Earle's salts supplemented with 10% FCS, 2 mM L-glutamine, 25 mM KCl, and 100 units/ml penicillin-streptomycin. Cultures were treated with 10 µM cytosine-arabinofuranoside 24 hr after plating to inhibit non-neuronal cells. Recordings were made 5-6 days after isolation.

Electrophysiology of native calcium currents. Currents were recorded at room temperature in the sympathetic, Purkinje, and cerebellar granule cells using whole-cell patch-clamp with an Axopatch 1D amplifier and pCLAMP software (Axon Instruments). The electrodes had tip diameters of 1-2 µm with a resistance of 1.5-2.5 M. Corrections were made for junction potentials, and leak currents were subtracted using a standard P/N protocol provided by the software. Seals were formed in a standard extracellular buffer containing 150 mM NaCl, 4 mM KCl, 10 mM HEPES, 10 mM D-glucose, 2 mM MgCl₂, and 2 mM CaCl₂, pH 7.4, and adjusted to 335 mOsM using sucrose. Barium currents were recorded using a buffer containing 154 mM TEA-Cl, 10 mM HEPES, 10 mM D-glucose, and 2 mM MgCl₂, pH 7.4, adjusted to 335 mOsm with sucrose and containing either 5 mM BaCl₂ for sympathetic and Purkinje cells or 10 mM BaCl₂ for granule cells. The intracellular buffer contained 108 mM CsCl, 9 mM HEPES, 9 mM EGTA, 4 mM MgCl₂·6H₂O, 14 mM creatine-PO₄, 4 mM MgATP, and 0.3 mM Tris-GTP, pH 7.4, and adjusted to 320 mOsM with sucrose.

Data analysis. Current recordings were analyzed using pCLAMP software (Axon Instruments), and curve fitting was carried out with GraphPAD Prism Software (San Diego, CA). Preparation of figures was carried out in Freelance graphics (Windows). Unless otherwise stated, error bars represent the standard error.

	Results

Top Summary Introduction Materials & Methods Results Discussion References

Isolation and sequence analysis of DW13.3. An ion exchange chromatogram of crude F. hibernalis (DW) venom is shown in Fig. 1. A protocol using both ion exchange and reverse-phase HPLC separated DW13.3 from the other Filistata (DW) peptides. Accordingly, crude F. hibernalis (DW) venom was applied to an ion exchange HPLC column (see Fig. 1, legend). The desired fraction containing DW13.3 was collected from 34 to 35.5 min, and pooled fractions were desalted without concentration. Subsequently, the material from ion exchange fractionation of 3360 µl of crude venom was applied to a Vydac C-18 HPLC column (300 Å, 5 µm, 22 × 250 mm; Nest Group, Southborough, MA) using a linear gradient program (A = 0.1% CF₃CO₂H; B = CH₃CN) with detection at 220 nM and a flow rate of 15 ml/min. The gradient was 23-30% B over 60 min, and the desired fraction containing DW13.3 was collected from 29 to 31 min. Pooled like fractions from individual runs were concentrated by lyophilization. Observed mass (ES-MS) = 8668 (calc. 8668.29).

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Preparative ion exchange chromatogram of DW13.3 isolation from crude venom. Crude F. hibernalis (DW) venom (~80 µl) was applied to a Poly LC polysulfoethyl aspartamide HPLC column (300 Å, 5 µm, 9.4 × 200 mm) using a triphasic linear gradient program (B = CH3CN, C = 5 mM H3PO4/H2O at pH 4.5, D = C + 1 M NaCl) with detection at 220 nm and a flow rate of 3.5 ml/min. The gradient was 20% B/80% C/0% D to 20% B/0% C/80% D over 45 min.

View larger version (27K): [in this window] [in a new window]   Fig. 2.   ES analysis of peptide fragments generated in the Glu-C digest of pyridylethylated DW13.3. Pyridethylated DW13.3 (10 µg) in H2O (10 µl) was combined with 100 µl of buffer (25 mM NH4OAc, pH 4.0) and 10 µl of endopeptidase Glu-C solution (0.1 µg/ml) and maintained at 37° for 24 hr. The fragments were separated by reversed-phase HPLC (Vydac C-18, 300 Å, 5 µm, 4.6 × 250 mm) using a biphasic linear gradient program (A = 0.1% CF3CO2H; B = CH3CN) with detection at 220 nm and a flow rate of 1 ml/min. The gradient was 0-30% B over 35 min and then 30-60% B over 25 min. Pyridylethylated derivative of DW13.3, suitable for amino-terminal sequencing, was prepared in the following manner: DW13.3 peptide (300 µg) was dissolved in 20 µl of buffer (1:3 ratio of 1 M Tris, pH 8.4, 4 mM EDTA-dibasic and 8 M guanidine hydrochloride), treated with 7.28 µl of a 10% v/v solution of 2-mercaptothanol in buffer, and kept in the dark at room temperature for 3 hr. The reaction then was treated with 11.19 µl of a 10% v/v solution of 4-vinylpyridine in buffer and kept at room temperature in the dark for an additional 18 hr. The reaction was diluted to 600 µl with 1% trifluoroacetic acid-H2O and applied to an HPLC column (Vydac C-18, 4.6 × 250 mm), which was operated using a biphasic linear gradient program (A = 0.1% CF3CO2H, B = CH3CN) with detection at 220 nm and a flow rate of 1.0 ml/min. The gradient was 0-30% B for 35 min and 30-60% B for 25 min. The desired fraction was collected from 34 to 35 min and concentrated by lyophilization. Approximate yield (based on amino acid analysis) was 194 µg. Observed mass was 9941 (calc. 9941.29).

View this table: [in this window] [in a new window]   TABLE 1 Sequences and cysteine alignment of DW 13.3 and other calcium channel toxins Primary amino acid sequences of DW13.3 (Ahlijanian et al., 1995); CT-III, a neurotoxic insecticidal polypeptide isolated from the venom of the funnel web spider Hololena curta (Stapleton et al., 1990); -Aga IVa, a selective blocker of P-type Ca2+ channels isolated from the venom of the funnel web spider Agelenopsis aperta (Mintz et al., 1992a, 1992b); -CgTX GVIA, a selective blocker of N-type Ca2+ channels isolated from the venom of the fish-hunting cone snail Conus geographus (Olivera et al., 1984); -CTx MVIIC, a blocker of N- and P/Q-type Ca2+ channels whose structure was provided by a venom duct cDNA library (Hillyard et al., 1992); and -Aga IIIA, a nonselective blocker of N-, L-, and P-type Ca2+ channels isolated from the venom of Agelenopsis aperta (Mintz et al., 1991).

DW13.3 potently blocks all four types of transiently expressed calcium channel. Bath application of DW13.3 (230 nM) produced a significant block of calcium channel currents transiently expressed in X. laevis oocytes. Four different types of calcium channel were investigated (1A, 1B, 1C, and 1E), each coexpressed with 2 and 1b subunits. DW13.3 produced a potent and reversible inhibition of all the calcium channels tested, although the degree of block and the rate of reversal of inhibition varied significantly among the subtypes (Fig. 3). Although the development of toxin block of the peak I_Ba (carried by 4 mM Ba²⁺ in all cases) was complete within 2 min of toxin application, the rate of recovery varied among the channel types. Three of the four calcium channels (1B, 1C, and 1E) showed rapid and nearly complete reversal of toxin inhibition, recovering to ~70-80% of the control current within 4 min. In contrast, the recovery of 1A was markedly slower (p < 0.01), and in oocytes with particularly stable currents, it was still not complete after 20-25 min of perfusion with control solution (data not shown). The average time constants of recovery were 1506 ± 342, 77 ± 14, 66 ± 8, and 45 ± 5 sec for 1A, 1B, 1C, and 1E, respectively (7, 11, 7, and 6 determinations; Fig. 3, inset).

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Inhibition of transiently expressed calcium channels by DW13.3. Four different cloned calcium channels (1A, ; 1B, ; 1C, ; 1E, ) were transiently coexpressed with 2 and 1b subunits in X. laevis oocytes. Bath perfusion of a solution containing DW13.3 (230 nM) produced a rapid (within 2 min) and reversible inhibition of the peak Ba2+ current (4 mM Ba2+, Vh = 100 mV, Vc = 0, +10 mV, 400-msec step, activated every 15 sec), with 1A exhibiting a much slower rate of recovery from toxin block. Inset, comparison of the mean time constants (± standard error) for recovery (off) for the four different calcium channel clones. Numbers in parentheses, number of determinants for each sample. **, p < 0.01.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Dose dependence of DW13.3 inhibition of expressed calcium channels. Current traces, DW13.3 inhibition of transiently expressed peak control Ba2+ currents (1A, 1B, 1C, and 1E) by a saturating concentration of DW13.3 (1.15 µM; , control; , DW13.3; , recovery). Opposite, dose-response data are summarized for all four calcium channel types. Data points represent mean ± standard error values (2-10 determinations) of the percent current inhibited after the perfusion of DW13.3 (2-16 min). Solid line fits were made to the following equation and assumed 1:1 binding of the toxin with the channel: Y = max/[1 + (IC50/[toxin])], where max is maximum inhibition. Barium currents (4 mM) were activated every 15 sec with a 400-msec step depolarization from 100 mV to 0 or 10 mV. DW13.3 inhibition of 1A was the most potent (IC50 = 4.3 nM) yet remained incomplete.

Inhibition of 1B in HEK cells. In contrast to that for X. laevis oocytes, DW13.3 was found to be markedly more potent when tested on 1B calcium channel currents expressed in mammalian HEK cells (1B HEK). Whole-cell currents were inhibited by concentrations of toxin as low as 300 pM, and analysis of the dose-response data yielded an IC₅₀ value of just 2.3 nM (Fig. 5, A and B). The rate of onset of block was rapid in 1B HEK cells, with maximal inhibition occurring within 1-2 min of application. However, the kinetics of toxin block were dissimilar compared with those observed in oocytes. The actions of DW13.3 were readily reversible in oocytes (Fig. 3), but there was little or no discernible recovery of 1B currents recorded in HEK cells within the 2-4-min recovery period (0-10%, four determinations; Fig. 5C). The relatively slow reversal of DW13.3 inhibition in 1B HEK cells yields an effective potency of this toxin 1 order of magnitude greater than that observed for the oocyte expression system (Fig. 5B).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   DW13.3 inhibition of 1B channels expressed in HEK cells. A, Currents were recorded from a single cell exposed to progressive applications of increasing toxin concentrations (300 pM, 1 nM, 3 nM). Currents were activated every 15 sec by 100-msec step depolarizations to +10 mV from a holding potential of 100 mV. B, Dose-response relation for 1B currents recorded in HEK cells (solid line; IC50 = 2.3 nM). Also included for comparison are data points and fits obtained for 1B currents expressed in X. laevis oocytes (intermittent broken line; IC50 = 14.4 nM) and the nifedipine-resistant (mainly N-type) currents recorded in rat sympathetic neurons (short broken line; IC50 = 22 nM). All data point represent mean values (± standard error, 2-9 determinations) of the percent inhibition of the control current and were fitted with the following equation assuming 1:1 binding of the toxin with the channel: Y = max/[1 + (IC50/[toxin])]. C, Time course showing irreversible saturating block of the 1B current recorded in HEK cells (Vh = 100 mV, Vc = 0 mV, 100-msec step depolarizations activated every 15 sec).

Inhibition of N-type current in rat sympathetic neurons. DW13.3 block of native calcium currents (5 mM Ba, I_Ba) was studied in rat superior cervical ganglion neurons. These cells typically exhibit a large proportion of high threshold current that is carried through -CgTX GVIA-sensitive, N-type calcium channels (Boland et al., 1994). In this study, saturating concentrations of nifedipine (10 µM) first were applied to block the L-type component of high threshold current in these cells (11 ± 1%, five determinations). Of the current that remained, 80.7 ± 1.2% (six determinations) could be blocked by saturating concentrations of -CgTX GVIA (3.2 µM). The current that was resistant to both nifedipine and -CgTX GVIA was also resistant to the P-type channel blocker -Aga IVA (200 nM).

View larger version (24K): [in this window] [in a new window]   Fig. 6.   DW13.3 block of -CgTX GVIA-sensitive, N-type current in rat sympathetic neurons. Currents were activated every 15 sec by a 35-msec step depolarization from 90 mV to +10 mV after the application of 10 µM nifedipine to block L-type current. A, Time course taken from a cell showing the reversible inhibition of nifedipine-insensitive, high threshold IBa (5 mM Ba2+) by DW13.3 (100 nM). Inset, current trace scale bar = 1 nA and 10 msec. B, Concentration dependence of DW13.3 inhibition of the nifedipine-resistant IBa. Data points reflect mean ± standard error of two to nine determinations. Solid line fit was made assuming a 1:1 binding association of the toxin with the channel and is based on 80% saturating block of the remaining current. C, Current traces showing additional block of high threshold nifedipine-resistant current by DW13.3 in the presence of -CgTX GVIA. D, Preapplication of DW13.3 interferes with additional block by -CgTX GVIA. Peak currents were recorded under control conditions (inset, current trace 1) and during the sequential addition of DW13.3 (320 nM and 1 µM; current traces 2 and 3, respectively) and -CgTX GVIA (3.2 µM; current trace 4; scale bar = 1 nA and 10 msec). Rate of onset and efficacy of additional block by -CgTX GVIA is significantly reduced in the presence of DW13.3. E, Bar graph summarizing toxin-sensitive components of high threshold, non-L-type current in rat sympathetic neurons. Saturating concentrations of DW13.3 (>320 nM) inhibit 83% of the N-type current and 73% of the remaining nifedipine-resistant current (mean ± standard error of three to six determinations).

View this table: [in this window] [in a new window]   TABLE 2 Summary of IC50, Kd, off, and maximum percent inhibition values

Partial block of 1A in oocytes and of P-type currents in Purkinje neurons. Although the IC₅₀ value for DW13.3 inhibition of the 1A calcium channel expressed in oocytes was 4.3 nM (Fig. 4), the highest saturating concentration of DW13.3 used in this study (1.15 µM) did not produce total block (68 ± 8%, three determinations). Fig. 7A shows that increasing concentrations of DW13.3 applied to a single oocyte produced a cumulative inhibition of the current that (in this case) saturated at ~55%. The dose-response data for 1A are well described assuming a Hill coefficient of 1.0, consistent with 1:1 binding of the toxin molecule with the calcium channel. Nevertheless, fit of the data did require a model based on incomplete block (even if block was not assumed to occur in a unitary manner; i.e., when the Hill coefficient was not fixed at 1.0).

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Partial inhibition of 1A expressed in X. laevis oocytes. A, Time course showing three sequential applications of increasing concentrations of DW13.3 on the same oocyte. Currents were elicited every 15 sec by 400-msec step depolarizations to +10 mV from a holding potential of 100 mV. Inset, current traces were taken from the same cell at points indicated (con, control, , 5.6 nM, , 56 nM, , 230 nM). Inhibition by DW13.3 saturates at <100% block. B, Time course of block in four different oocytes exposed to 2.3, 5.6, 56, and 230 nM DW13.3. Barium currents (4 mM) were activated every 15 sec with 400-msec step depolarizations from 100 mV to 0 or +10 mV. Currents were normalized to the control current amplitude recorded just before toxin application (). Dotted lines, fits of single exponentials decaying to a plateau (on). C, Dependence of 1/on on [DW13.3]. Values are mean ± standard error of four to seven determinations. Solid line fit was made to the following equation: 1/on = kon·[DW13.3] + koff, where kon is 0.000249 (nM1 sec1), koff is 0.00192 (sec1), and Kd is 7.7 nM. Dotted line, a fit in which the y-intercept was fixed to the mean value obtained from kinetic analysis of the toxin off-rate calculated from individual cells (off = 1506 ± 342 sec, seven determinations; koff = 0.000595 sec1, kon = 0.000256 nM1 sec1, and Kd = 2.6 nM).

4

1

1

4

1

1

View larger version (20K): [in this window] [in a new window]   Fig. 8.   DW13.3 partially inhibits the P-type current in Purkinje neurons but does not occlude complete block by -Aga IVA. A, Both -Aga IVA and -CTx MVIIC effectively inhibit all P-type current present in Purkinje neurons. B, Partial block by saturating concentrations of DW13.3 prevents additional inhibition by -CTx MVIIC but not by -Aga IVA. C, Current traces showing the progressive block of the control P-type current (1) after the application of 32 nM DW13.3 (2) followed by 5 µM -CTx MVIIC (3) and 200 nM -Aga IVA (4). D, Data showing that DW13.3 selectively occludes -CTx MVIIC but not -Aga IVA from binding to P-type calcium channels. Nifedipine (10 µM) was present throughout all these experiments. Peak inward currents (5 mM Ba2+) were elicited by 35-msec step depolarizations to 20 mV from a holding potential of 80 mV.

DW13.3 inhibition of pharmacologicaly unclassified components of native calcium channel currents. In addition to inhibiting components of pharmacologicaly defined high threshold current, DW13.3 exhibited a potent block of a number of other uncharacterized current types. A considerable proportion (~70%) of the non-L-, non-N-, non-P/Q-type current in rat sympathetic neurons was sensitive to block by saturating concentrations of DW13.3 (>320 nM; see above). A number of studies have suggested a correlation between the 1E channel and a component of rapidly inactivating current in cerebellar granule neurons that is resistant to L-, N-, and P/Q-type channel blockers (R-type current) (Zhang et al., 1993; Randall and Tsien, 1995). Initial experiments on expressed channels revealed a potent yet partial block of the 1E channel (Table 2). The actions of DW13.3 on the resistant component of current were investigated in cerebellar granule neurons. A combination of nifedipine (10 µM), -CgTX GVIA (1 µM), and -CTx MVIIC (5 µM) was used to block the L-, N-, and P/Q-type channels respectively. In this study L-, N-, and P/Q-type current represented 49 ± 3% (10 determinations) of the total whole-cell current. Fig. 9 shows the results of the coapplication of these three channel blockers followed by the effects of subsequent addition of DW13.3 (100 nM). DW13.3 rapidly (within 1-2 min) blocked 82 ± 4% (four determinations) of the remaining resistant current. In contrast with 1E, inhibition did not seem to reverse after washing for 5 min with control solution (three determinations, data not shown). Preliminary experiments carried out on whole-cell currents in GH3 cells revealed that this T-type current showed little or no sensitivity to a concentration of DW13.3 that significantly reduced the L-type current in these cells (230 nM; data not shown).

View larger version (20K): [in this window] [in a new window]   Fig. 9.   DW13.3 inhibits residual current in cerebellar granule neurons. A, Current traces from a single cerebellar granule neuron show additional inhibition of the resistant current by DW13.3 (100 nM) after inhibition of L-, N-, and P/Q-type currents. B, Average time course from three cells. DW13.3 (100 nM, ) produced a significant inhibition of the resistant current remaining in the presence of saturating concentrations of L-, N-, and P/Q-type channel blockers nifedipine (10 µM), -CgTX GVIA (1 µM), and -CTx MVIIC (5 µM) compared with control (). Peak currents (10 mM Ba2+) were activated every 15 sec by 35-msec step depolarizations to either 0 or +10 mV from a holding potential of 100 mV.

	Discussion

Top Summary Introduction Materials & Methods Results Discussion References

DW13.3 inhibition of L-, N-, and P/Q-type currents. This study shows that DW13.3 exhibits a potent block of calcium channel currents recorded in a variety of cultured mammalian cells and exogenous expression systems (summarized in Table 2). The dose-response data obtained for each of the calcium channel clones indicate that the toxin binds in a 1:1 manner, producing differing degrees of block at saturating concentrations. Of the four clones examined in this study, DW13.3 inhibition of the 1C L-type channel was most complete. Previously, DW13.3 had been shown to discriminate between 1,4-dihydropyridine-sensitive L-type currents recorded from rat cerebellar granule cells and a rat aortic cell line (A7R5; IC₅₀ = 0.4 nM and 16 nM, respectively; Ahlijanian et al., 1995). When compared, DW13.3 inhibition of the rat cerebellar granule cell L-type current was by far the most potent observed to date. This 20-fold difference in affinity between L-type channel isoforms may reflect regional differences in the expression of 1C and 1D L-type channels and provide a means of discriminating between L-type currents in neuronal versus smooth muscle preparations. A direct comparison of the affinity of DW13.3 for heterologously expressed 1C and 1D subunits may help to explain these observed differences in potency.

DW13.3 inhibition of non-L-, -N-, and -P/Q-type current. Initial experiments on expressed channels revealed a potent yet partial block of the 1E channel. A comparable component of endogenous 1E current has yet to be unequivocally identified because no selective pharmacological agent for this channel has been described. A number of studies have suggested a correlation between the 1E channel and a component of the residual current in cerebellar granule neurons (Zhang et al., 1993; Randall and Tsien, 1995); however, the biophysical and pharmacological properties of these currents do not match "R-type" currents (compare Soong et al., 1993, and Bourinet et al., 1996). In this study, 100 nM DW13.3 (a concentration that produces maximal block of 1E currents) was found to irreversibly inhibit a significant proportion (~80%) of the resistant current in cerebellar granule neurons. In addition, a proportion (~20%) of the resistant current also persisted in the presence of all four blockers. Because by definition this current is isolated by virtue of exclusion to block by nifedipine, -CgTX GVIA, and -CTx MVIIC, it remains to be elucidated whether block by DW13.3 represents complete inhibition of an additional subpopulation of previously uncharacterized channels or partial block of the resistant current.

Structural considerations for DW13.3. DW13.3 shares very little structural identity with other calcium channel peptide toxins (Table 1). The broad specificity of DW13.3 inhibition and its varied blocking efficacy are very similar to those of -Aga IIIA, a peptide spider toxin isolated from the venom of the funnel web spider A. aperta (Cohen et al., 1992; Ertel et al., 1994; Mintz, 1994; Ahlijanian et al., 1995). However, these two toxins lack any form of sequence homology. In fact, the amino terminus of DW13.3 is most closely aligned with the µ-agatoxin family. The first 33 residues of curtotoxin (CT-III) (Stapleton et al., 1990), which has the same cysteine topology as the -Aga IV toxins, are 38% identical with amino acids on the amino terminus of DW13.3.

Where does DW13.3 act?. Although the specificity of P- and N-type calcium channel toxins for their respective molecular targets has shed some light on which structural features of the peptide toxins are important for channel inhibition, these standards alone are not sufficient for solving this problem (Adams et al., 1993b; Omecinsky et al., 1996). Toxins such as DW13.3 are structurally distinct yet target multiple calcium channel subtypes. Extreme degeneracy in the sequences of different N-type specific -conotoxins has already been described (Olivera et al., 1990). -Conotoxin binding to N-type channels is thought to involve interaction of the conserved tertiary structure of the peptides with a "macro" ligand binding site. The specificity of binding is determined by a number of smaller "micro" interactions that are specific to the residues of each individual toxin. The macrobinding site is able to tolerate the hypervariable domains of each different toxin as they are constrained by the disulfide bonds formed between conserved residues forming a common overall three dimensional structure (Olivera et al., 1990, 1991). The "macrosite hypothesis" has also been proposed for a number of other peptide toxins exhibiting a broad specificity for calcium channels (Olivera et al., 1994). Both -Aga IIIA and -CTx MVIIC inhibit a number of different calcium channels with differing binding affinities and blocking efficacies (Hillyard et al., 1992; Mintz, 1994; Stea et al., 1994; Randall and Tsien, 1995).