Introduction

Summary Introduction Materials & Methods Results Discussion References

Natural peptides have been invaluable tools for pharmacological and biochemical investigations of a wide range of physiological functions (1). Calcium channels have been investigated in large part because of the existence of a great number of molecules acting on these channels. They have been classified according to their electrophysiological and pharmacological properties (2, 3) and represent an important family of ionic channels because they are directly acting on calcium homeostasis of the cell. Biological functions such as muscle contraction, hormone secretion, or neurotransmitter release are directly dependent on intracellular calcium variations. These channels are involved in a wide range of pathologies, including cardiovascular system diseases (4, 5). Different types of calcium channels can be expressed in the same cellular type (6-8). L- and T-type calcium channels are, for example, found in most excitable cells, although the physiological role of the T-type channel remains unclear. The L-type channel is mainly involved in excitation/contraction or excitation/secretion couplings and is sensitive to dihydropyridines, a class of calcium antagonists widely used in the treatment of cardiovascular pathologies. L- and T-type calcium channels are also present in smooth muscle cells (6, 7). Interestingly, several types of voltage-dependent calcium channels have been described in gastrointestinal smooth muscles (9, 10). P-type channels are mainly found in Purkinje cells and are sensitive to a spider toxin, the -agatoxin IVA (11, 12). N-type channels are present on a wide variety of neurons and are blocked by -conotoxin GVIA (13). Both P- and N-type calcium channels play an important role in the control of neurotransmitters release (12, 14). Another type of calcium channel that presents with similarities to the P-type channel is the Q-type; it has been described in the granular layer of cerebellum. It could be involved in synaptic transmission between neurons from CA3 and CA1 layers in hippocampus and is sensitive to -conotoxin MVII C (15). Genetic investigations demonstrated the existence of several genes encoding for a same 1 subunit (16). This phenomenon could be responsible for the functional differences observed for a same channel type expressed in different cell types. The actual situation seems to be complex because there are many calcium channel subtypes for which functional and biological properties remain unclear, in part because of the absence of a consequent pharmacology (3). Marine sponges are known to synthesize many organic and peptidic molecules with interesting biological properties; for example, latrunculin A interacts with actin polymerization (17). Peptides from Bahamian sponge, discobahamin A and B, or halicylindramides from Halichondria cylindrata have antifungal properties (18, 19). Protease inhibitory peptides are also found in marine sponge (20). Many other peptides with biological activity have been isolated from sponges, but as far as we know, no calcium channel effectors have been described until now. A previous report has shown that sponge extracts are able to block carbachol-induced contractions in intestinal smooth muscle, which are insensitive to dihydropyridines (21). These observations suggest that these organisms could synthesize calcium channel effectors. In this work, we describe, on the one hand, the existence of both L-type and non-L-type calcium currents in mouse duodenal smooth muscle cells. The non-L-type calcium current is insensitive to a variety of calcium channel blockers. On the other hand, we describe the purification and sequence of a dimeric protein (mapacalcine) isolated from C. vastifica sponge that is able to specifically block the duodenal non-L-type calcium current, suggesting the existence of a new class of calcium channels.

	Materials and Methods

Summary Introduction Materials & Methods Results Discussion References

C. vastifica extraction and purification. The sponges (C. vastifica) were collected from the Red Sea or obtained from Latoxan (Rosans, France). The entire sponge and its stony substrate were ground and soaked in 50% ethanol for 24 hr. The ethanol was then evaporated in a Rota Vapor concentrator (Bioblock, Illkirsh, France) and the aqueous phase was lyophilized. The powder obtained was stored at 20° until use.

Gel filtration chromatography. Typical purification was performed using 2 g of powder that was resuspended in 15 ml of 1% acetic acid and centrifuged for 1 hr at 140,000 × g. The supernatant was loaded onto a Sephadex G50 (fine) column (2.5 × 100 cm) equilibrated with 1% acetic acid. The column was then eluted at a rate of 0.7 ml/min for 12 hr. Then, 15-ml fractions were collected, and the elution was monitored by absorbance measurement at 280 nm. All subsequent purification steps were performed with HPLC.

Ion exchange chromatography. The active fractions obtained from gel filtration were loaded onto a TSK gel SP-5PW column (2.1 × 15 cm; Toyo Soda, Tokyo, Japan) equilibrated with 1% acetic acid (buffer A). The column was then washed at a rate of 8 ml/min for 10 min with buffer A and eluted by a gradient of 0-100% 1 M ammonium acetate (buffer B) during 100 min. Fractions were collected every 2 min. Elution was followed by on-line absorbance recording at 280 nm.

Reverse-phase purification step. Selected fractions from ion exchange chromatography underwent a purification step on a reverse-phase C 8 column (Lichrosorb 100 RP 8, Merck, Nogent sur Marne, France; 10 × 250 mm) equilibrated with water containing 0.1% TFA (buffer A) at 3 ml/min. The column was washed 10 min with buffer A; then, elution was performed by a gradient of 0-80% acetonitrile containing 0.1% TFA over 40 min. One-minute fractions were collected, and the chromatogram was recorded at 280 nm.

Desalting operation. Before being tested by electrophysiological techniques, fractions obtained from gel filtration or ion exchange steps were desalted using SepPak cartridges (Waters, Milford, MA). The cartridge was previously washed by 5 ml of acetonitrile and then rinsed with 10 ml of 0.1% TFA in water. Samples to be desalted were mixed with TFA (final concentration, 0.1%) and loaded onto the cartridge. After washing with 10 ml of 0.1% TFA in water, the retained material was eluted by 1 ml of water/acetonitrile (50:50) and 1 ml of acetonitrile, both containing 0.1% TFA. The elutates were evaporated using a Christ concentrator (Bioblock), and the desalted material obtained was used for electrophysiology.

Sequence analysis. The purity of the toxin obtained after reverse-phase HPLC purification was assessed by the use of capillary electrophoresis (270A-HT Capillary Electrophoresis System; Applied Biosystems, Norwalk, CT) in 20 mM sodium citrate buffer, pH 2.5, at 30 kV and 30° for 10 min with a 50-cm, 50-µm-diameter capillary. UV detection was performed at 200 nm.

Amino acid analysis. Amino acid analyses were performed with a Beckman Instruments (Palo Alto, CA) 6300 amino acid analyzer after hydrolysis in 6 M HCl (Pierce Chemical, Rockford, IL) in the presence of 0.25% phenol at 110° in sealed/evacuated tubes for 24, 48, and 72 hr.

Reduction and carboxamidoethylation. The toxin (0.4 mg, ~20 nmol) was dissolved in 100 µl of 0.1 M ammonium bicarbonate buffer, pH 8.0, containing 6 M guanidium chloride and 0.1 M dithiothreitol and maintained at 70° for 30 min. The mixture was then brought to room temperature, and alkylation of the reduced toxin was performed for 45 min by the addition of 50 µl of 6 M acrylamide in water. The carboxamidoethylated toxin was desalted by reverse-phase HPLC on a C18 Vydac column (200 × 2.1 mm, 5-µm particle size, 300-Å pore diameter) at a flow rate of 100 µl/min. After washing out the salts with 0.05% TFA, the protein was eluted with 80% acetonitrile in 0.05% TFA. The elution was monitored at 210 nm. The carboxamidoethylated toxin (0.1 mg, ~10 nmol) dissolved in 100 µl of 0.1 M ammonium bicarbonate, pH 8.0, was then hydrolyzed for 2 hr at 37° with TosPheCH₂Cl-treated trypsin using an enzyme-to-substrate ratio of 1:100 (w/w).

Peptide separation. The tryptic peptides were separated by reverse-phase HPLC on a C18 Vydac column (200 × 2.1 mm, 5-µm particle size, 300-Å pore diameter) using a gradient of acetonitrile of 8-40% in 0.05% TFA for 90 min at a flow rate of 100 µl/min. Elution of peptides was monitored by UV absorption at 210 nm.

Ion spray mass spectrometry. The molecular mass of the native toxin and that of its carboxamidoethylated derivative were determined by ion spray mass spectrometry. Samples (10-20 pmol/µl) were dissolved in 20% acetonitrile in water/formic 0.1% acid. Ion spray mass spectra were recorded on a simple quadrupole mass spectrometer (API I; Perkin-Elmer Cetus, Norwalk, CT) equipped with an ion-spray (nebulizer-assisted electrospray) source (Sciex, Toronto, Ontario, Canada). The solutions were continuously infused with a medical infusion pump (model 11; Harward Apparatus, South Natick, MA) at a flow rate of 5 µl/min.

Sequence analysis. Sequencing of protein in native form or after reduction and carboxamidoethylation of cysteines and of tryptic peptides was carried out on a gas-phase sequencer (model 470A, Applied Biosystems) using the 03RPTH program. Phenylthiohydantoin derivatives of amino acids were identified with an on-line phenylthiohydantoin amino acid analyzer (model 120A, Applied Biosystems).

Cell preparation. Swiss mice (20-25 g) and Wistar rats (150-160 g) were stunned and then killed by cervical dislocation. The longitudinal layer of the mouse duodenal smooth muscle (or the rat portal vein) was cut into several pieces and incubated for 10 min in a low-Ca²⁺ (40 µM) physiological solution. Then, 0.8 mg/ml collagenase, 0.2 mg/ml pronase E, and 1 mg/ml bovine serum albumin were added, and the mixture was maintained at 37° for 20 min. The solution was then renewed, and the pieces of mouse duodenum (or rat portal vein) were incubated under the same conditions for an additional 20-min period. Tissues were then placed in enzyme-free solution and dissociated using a fire-polished Pasteur pipette to release cells. Cells were maintained in short term primary culture in Medium 199 containing 10% fetal calf serum (duodenum) or 5% fetal calf serum (portal vein), 1 mM glutamine, 1 mM pyruvate, 20 units/ml penicillin, and 20 µg/ml streptomycin. Cells were maintained in culture at 37° in an incubator under controlled atmosphere (5% CO₂) and used for electrophysiology within 36 hr.

Patch-clamp measurements. Voltage-clamp and membrane current recordings were made with a standard patch-clamp technique using a List EPC-7 patch-clamp amplifier (List Electronics, Darmstadt, Germany). The whole-cell recording mode was performed with patch pipettes of 1-3 M resistance. Membrane potential and current records were stored and analyzed using an IBM-PC computer (pClamp system).

Fluorescence measurements. Whole-cell membrane currents and intracellular calcium concentration were measured simultaneously as previously reported (22). Briefly, 50 µM Indo-1 was added to the pipette solution and entered into the cells after establishment of the whole-cell recording mode. Indo-1 repartition was usually homogeneous over the cytoplasm, and no compartmentalization of the dye was observed. Cells were mounted in a perfusion chamber and placed on the stage of an inverted microscope (Nikon Diaphot). The cell studied was illuminated at 360 nm. Emitted light from a window slightly larger than the cell was counted simultaneously at 405 and 480 nm by two photomultipliers (P₁, Nikon). Intracellular calcium concentration was estimated on the basis of the 405 nm/480 nm ratio using a calibration for Indo-1 determined within cells. Some experiments were carried out in the presence of 1 µM oxodipine (a light-stable dihydropyridine derivative) to inhibit L-type voltage-dependent calcium channels. All measurements were made at 25 ± 1°.

Solutions. The normal physiological solution contained 30 mM NaCl, 5.6 mM KCl, 1 mM MgCl₂, 2 mM CaCl₂, 11 mM glucose, and 10 mM HEPES/NaOH, pH 7.4. The basic pipette solution contained 130 mM CsCl and 10 mM HEPES/CsOH, pH 7.3. For calcium current recordings, CsCl was used instead of KCl in the pipette and external solutions to block outward potassium currents. In addition, 10 mM EGTA, 5 mM Na₂ATP, and 1 mM MgCl₂ were added to the basic pipette solution. For potassium channel current recordings, KCl was used in the external and pipette solutions. Caffeine, C. vastifica toxin, and all pharmacological agents were applied to the recorded cell by pressure ejection from a glass pipette for the period indicated on the records. Before each experiment, an application of physiological solution alone was tested, and cells with movement artifacts were excluded. In addition, C. vastifica toxin and all pharmacological agents were tested by their addition to the perfusion solution.

Intestinal membrane preparation. Mice were killed by cervical dislocation. Small intestines were removed and rinsed with ice-cold buffer containing 140 mM NaCl, 20 mM Tris·HCl, pH 7.4, and 0.1 mM phenylmethylsulfonyl fluoride (washing buffer). The intestines were then opened and scraped to remove the epithelial cell layer. Intestinal smooth muscles were diluted 10-fold (w/v) with the washing buffer at 4° and homogenized 30 sec in a Warring Blender and then twice at 30 sec with a Polytron homogenizer. Homogenate was centrifuged at 4000 × g for 7 min, and the supernatant was centrifuged for 45 min at 140,000 × g. All centrifugation steps were performed at 4°. The pellets were resuspended in the washing buffer and frozen at 80° until use. Proteins were measured using BioRad (Hercules, CA) protein assay reagent and lysozyme as standard.

Binding of [³H]-(+)-isradipine. Intestinal membranes (0.5 mg of protein/ml) were incubated at 25° in 0.1% bovine serum albumin and 20 mM HEPES/NaOH, pH 7.4, in the presence of 100 pM labeled (+)-isradipine and increasing concentrations of unlabeled isradipine, oxodipine, and mapacalcine. After a 1-hr incubation, 0.8-ml aliquots were filtered through Whatman GF/C filters and washed three times with 0.1 M ice-cold Tris/Cl buffer, pH 7.4. Radioactivity retained on the filters was estimated with a Packard scintillation counter (Meriden, CT).

Chemicals and drugs. All reagents and solvents were of the highest purity available. Collagenase (E.C. 3.4.24.3) and trypsin (E.C. 3.4.21.4) treated with TosPheCH₂Cl were from Worthington (Freehold, NJ). Carboxypeptidase P sequencing grade (E.C. 3.4.16.1) was purchased from Boehringer-Mannheim Biochemica (Mannheim, Germany). Pronase (E.C. 3.4.24.31), bovine serum albumin, tetrodotoxin, and amiloride were from Sigma Chemical (St. Louis, MO). -Conotoxin GVIA, -agatoxin IVA, and -conotoxin MVII C were from Latoxan (Rosans, France). Fenoverine was from Delalande (Paris, France). Diltiazem was from Synthelabo (Paris, France). Desmethoxyverapamil was from Knoll (Ludwigshafen, Germany). Isradipine was from Sandoz (Rueil-Malmaison, France). Medium 199 was from Flow Laboratories (Puteaux, France). Fetal bovine serum was from Flobio (Courbevoie, France). Streptomycin, penicillin, glutamine, and pyruvate were from GIBCO (Paisley, UK). Oxodipine was a gift from Dr. A. Galiano (IQB, Madrid, Spain). Caffeine and HEPES were from Merck. Indo-1 was from Calbiochem (Meudon, France).

Data analysis. The results are expressed as mean ± standard error. Significance was tested with the Student's t test. Values of p < 0.05 were considered statistically significant. Inhibition and concentration-response curves were analyzed by a nonlinear least-squares fitting program according to models involving one or two binding sites.

	Results

Summary Introduction Materials & Methods Results Discussion References

Calcium currents in duodenal myocytes. With CsCl in the pipette and bathing solutions to inhibit potassium currents, depolarizing pulses were applied from two holding potentials (80 or 60 mV) to 0 mV for 150 msec at 0.05 Hz. With 2 mM CaCl₂, the inward current obtained from a holding potential of 80 mV was larger than that obtained from a holding potential of 60 mV (Fig. 1A). Decay of the calcium currents elicited from 60 and 80 mV was well fitted with a two-exponential function. For inward currents evoked by depolarizations to 0 mV, the time constants of inactivation were 18 ± 5 and 96 ± 29 msec for the current elicited from 60 mV (12 experiments) and 12 ± 2 and 81 ± 21 msec for the current elicited from 80 mV (18 experiments). The effects of several compounds that inhibit calcium channels were tested on the maximal calcium currents elicited from the two holding potentials. As shown in Table 1, the maximal calcium current elicited from 60 mV was suppressed by isradipine, oxodipine, fenoverine, diltiazem, and desmethoxyverapamil. In contrast, when the cells were held at 80 mV, high concentrations of these compounds inhibited the calcium current by only 40-50%, suggesting that a component of inward current resistant to calcium channel antagonists was unmasked by negative holding potentials.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Calcium currents in mouse duodenal myocytes. A, Currents elicited from holding potentials of (a) 60 mV and (b) 80 mV to 0 mV with 2 mM CaCl2 in the external solution. B, Time course of calcium currents elicited from a holding potential of 60 to 0 mV at stimulation frequencies of () 0.05 Hz and () 0.1 Hz. C, Effect of 10 µM isradipine on the calcium current elicited by depolarizing steps from 80 to 0 mV at a stimulation frequency of 0.1 Hz. D, Effect of 0.1 mM Cd2+ on the calcium current elicited by depolarization steps from 80 to 0 mV at a stimulation frequency of 0.1 Hz. Calcium current was expressed as a fraction of control current (I/Ic) obtained 3 min after the membrane was ruptured under the pipette.

View this table: [in this window] [in a new window]   TABLE 1 Inhibition of maximal Ca2+ currents evoked by depolarizing pulses from holding potentials (h) of 60 or 80 mV Value represents the inhibition percentage of Ca2+ current [(I/Ic)  1 × 100], in which Ic is the control current in the absence of drug and I is the current in the presence of drug for 5 min. Values are mean ± standard error of n experiments.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Current-voltage relationships and steady state inactivation curves for the two types of calcium currents. A, Calcium currents elicited from a holding potential of 60 mV at a stimulation frequency of 0.05 Hz. B, Calcium currents elicited from a holding potential of 80 mV, () at a stimulation frequency of 0.1 Hz for 10 min or () in the continuous presence of 10 µM oxodipine. Currents were expressed as a fraction of the peak current. C, Steady state inactivation curve obtained with the two-pulse protocol (inset) for the L-type calcium current from a holding potential of 60 mV () and for the non-L-type calcium current from a holding potential of 90 mV, after a 10-min period of stimulation frequency at 0.1 Hz (). Data were fitted by curves of form 1/1[1 + exp(Vm  Vh)/k], in which Vh is the potential at which half of current was inactivated, Vm is the membrane potential, and k is the slope factor. Calcium current was expressed as a fraction of maximal current (I/Imax). Points, mean response of 5-15 cells with the standard error (vertical lines).

C. vastifica extract purification and effects on calcium current. A gel filtration (Sephadex G50) chromatogram is shown in Fig. 3A. Four pools were made with the fractions that were collected (P1-4). The different pools were desalted as described above and tested by the patch-clamp technique. Pool P3 (arrows) was found to inhibit the non-L-type calcium current elicited by depolarizations from 80 to 0 mV at a stimulation frequency of 0.1 Hz for 10 min and in the continuous presence of 10 µM oxodipine with an IC₅₀ value of ~27 µg/ml. No variation of leak current and holding current was detected in the presence of C. vastifica extracts. Measurements were made only when the cell reached a steady state within 5-6 min. Pool P3 was loaded onto the TSK gel SP-5PW ion exchange column. The chromatogram that was obtained (Fig. 3B) shows, as expected, a multiple peak pattern. Different pools were collected and tested for calcium channel current inhibition after a desalting step. Inhibition of the calcium current was found under the peak eluted between 21 and 23 min under conditions used in Fig. 3B. No activity was detectable in the other fractions. The active fraction was named P3 C and loaded onto a C8 reverse-phase column. After elution of the C8 column, as described Materials and Methods, the chromatogram that was obtained (Fig. 3C) demonstrated the presence of a single symmetrical peak. The IC₅₀ values obtained on the non-L-type calcium current varied from 5 ± 1 (5-12 experiments) to 3.6 ± 0.7 (5-15 experiments) µg/ml among the tested fractions. Electrophysiological experiments showed that all the calcium-blocking activity was eluted in this fraction; this peak was named P3 C3. At this stage, we submitted the material contained in P3 C3 to capillary electrophoresis analysis, and data showed that P3 C3 was homogeneous. The peptide (mapacalcine) was then submitted to sequence analysis.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Purification procedure of mapacalcine. A, The crude extract was centrifuged and loaded onto a Sephadex G50 (2.5 × 100 cm) column. Elution was performed at 0.7 ml/min with 1% acetic acid. The material shown eluting between the arrows contained the calcium channel-blocking activity. B, Ion exchange chromatography of the active material obtained from the Sephadex G50 purification step. The TSK gel SP-5PW column was eluted using a gradient from 0 to 1 M ammonium acetate over 100 min. The active material was eluted at 21-23 min (arrows). C, Reverse-phase chromatography of the active fraction obtained from the ion exchange step. Elution was performed at 3 ml/min on a reverse-phase C8 column (Lichrosorb 100 RP 8, Merck; 10 × 250 mm) by a gradient of 0-80% acetonitrile containing 0.1% TFA over 40 min. One symmetrical peak was eluted. and all the calcium channel-blocking activity was obtained under this peak. C, Material was sequenced and submitted to mass spectroscopy analysis before and after reduction and alkylation to demonstrate its dimeric structure. Data obtained demonstrated the presence of an homodimer (see Results).

Mapacalcine 3 primary structure. The amino acid composition of mapacalcine was mainly characterized by large amounts of dicarboxylic residues (21%), a significantly high content of glycine (12%) and cysteine (10%), and a low content of basic residues (5%; Table 2).

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Sequence of one chain of the dimer of mapacalcine. Methods used for the determination of the sequence are indicated as follows: single letters, residues identified by automated Edman degradation; boxed letters, peptide generated from cleavage of the reduced and carboxyethylated monomer with trypsin; striped boxes, residues identified by both direct Edman degradation and tryptic peptide sequencing.

Effects of mapacalcine on calcium currents. As shown in Fig. 5A, mapacalcine (1-5 µM) had no effect on the maximal L-type calcium current elicited by a depolarization from 60 to 0 mV, at a stimulation frequency of 0.05 Hz, in duodenal myocytes. There was no change in the voltage threshold, potential for the maximal current, or apparent reversal potential, as shown by the current-voltage relationships. The absence of effect of mapacalcine on L-type calcium channels was confirmed by study of the binding of [³H]-(+)-isradipine to intestinal membranes (Fig. 5B). Inhibition of the high affinity [³H]-(+)-isradipine binding was obtained with isradipine (K_i = 0.44 ± 0.04 nM, four experiments) and oxodipine (K_i = 12 ± 0.9 nM, four experiments). In contrast, mapacalcine (0.1 nM to 10 µM) had no effect on [³H]-(+)-isradipine binding (four experiments).

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Effects of mapacalcine on the L-type calcium current and [3H]-(+)-isradipine binding. A, Calcium currents elicited from a holding potential of 60 mV at a stimulation frequency of 0.05 Hz in the () absence or () presence of 5 µM mapacalcine for 5 min. a, Maximal calcium current. b, Current-voltage relationships. Points, mean of three to seven experiments, with the standard error shown (vertical lines). B, Competition experiments between [3H]-(+)-isradipine, isradipine (), oxodipine (), and mapacalcine (). Points, mean response of four experiments, each carried out in duplicate, with the standard error shown (vertical lines). Nonspecific binding was not changed in the presence of the drugs.

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Inhibitory effects of mapacalcine on the non-L-type calcium current in duodenal myocytes. Calcium currents were elicited from a holding potential of 80 mV at a stimulation frequency of 0.1 Hz and in the presence of 10 µM oxodipine. A, Current-voltage relationships in the () absence and () presence of 0.3 µM mapacalcine for 5 min. B, Concentration-response relationship of the inhibitory action of mapacalcine. Calcium current was expressed as a fraction of control current in the absence of mapacalcine (I/Ic) and plotted against mapacalcine concentration. Continuous line, nonlinear regression of the data points assuming a Hill coefficient of 0.8. Points, mean response of 5-15 cells with the standard error shown (vertical lines). Inset, calcium current traces from three cells illustrating inhibition induced by (b) 10 nM, (c) 0.3 µM, or (d) 3 µM mapacalcine. a, Control calcium current before toxin application.

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Effects of mapacalcine on the L-type and T-type calcium currents in portal vein myocytes. A, L-type calcium currents elicited from a holding potential of 50 mV at a stimulation frequency of 0.05 Hz. a, Maximal calcium current in the () absence or () presence of 3 µM mapacalcine. b, Current-voltage relationships obtained in the () absence or () presence of 3 µM mapacalcine for 5 min. B, T-type calcium currents elicited from a holding potential of 80 mV at a stimulation frequency of 0.1 Hz for 10 min. a, Maximal calcium current in the () absence or () presence of 3 µM mapacalcine. b, Current-voltage relationships obtained in the () absence or () presence of 3 µM mapacalcine for 5 min.

Mechanisms of mapacalcine inhibition. A standardized protocol was used to assess the relative contribution of initial, conditioned, and tonic blockade of non-L-type calcium current by mapacalcine. As shown in Fig. 8A, the steady state inhibitory effect of 0.3 µM mapacalcine on the calcium current elicited by depolarizations from 80 mV to 0 mV, at a stimulation frequency of 0.1 Hz and in the presence of 10 µM oxodipine, was obtained within 4-6 min. The effects of mapacalcine was partially reversed after a 8-10-min return to reference solution. In Fig. 8B, a train of depolarizing pulses at 0.1 Hz was applied for 10 min, followed by a rest period of 4 min during which mapacalcine (0.5 µM) was perfused. A second identical voltage-clamp depolarization train was then applied. Blockade was estimated by measuring the difference in peak calcium current between control and test drug conditions when steady state blockade was obtained. The initial blockade, assessed as the difference in peak calcium current between the control and the first pulse after drug exposure, was 62 ± 5% (five experiments). Conditioned blockade was the difference between the peak calcium current for the first and the last pulses after drug exposure without rest period (Fig. 8C). The calcium current inhibition was 65 ± 8% (five experiments) and not different from that observed after a 4-min rest period (p > 0.05). Increasing the stimulation frequency from 0.1 to 0.2 Hz had no significant effect on the time course and steady state inhibition of calcium current in the presence of mapacalcine (67 ± 12%, six experiments; Fig. 8C). Tonic blockade was defined as the blockade of calcium current that could not be removed during a 2-min hyperpolarization at 100 mV. Because hyperpolarization of the membrane did not restore the calcium current (six experiments), we can postulate that the mapacalcine-induced blockade was largely tonic.

View larger version (21K): [in this window] [in a new window]   Fig. 8.   Properties of mapacalcine inhibition. A, Time course of inhibition of the calcium current elicited from a holding potential of 80 to 0 mV at a stimulation frequency of 0.1 Hz for 10 min in the presence of 0.25 µM mapacalcine. B, Effect of frequency of command pulses on the calcium current before and after a rest period of 4 min in the presence of 0.5 µM mapacalcine. C, Command pulses applied at () 0.1 Hz and () 0.2 Hz before and during application of 0.5 µM mapacalcine. D, Inactivation curve of the calcium current in the () absence or () presence of 3 µM mapacalcine The test potential (V2) is 0 mV, and the duration of the conditioning pulse (V1) is 20 sec (inset). Points, mean response of 612 cells with the standard error shown (vertical lines). Calcium current was expressed as a fraction of control current (I/Ic) or maximal current (I/Imax).

Effects of mapacalcine on other ionic channels. Voltage-activated potassium currents of duodenal and portal vein myocytes were obtained during depolarizations elicited from a holding potential of 80 mV. The cells were incubated for 20 min in 5 mM BaCl₂, and the pipette solution contained 130 mM KCl. In the presence of 3 µM mapacalcine for 5 min, no significant effect was observed on the voltage-dependent potassium current at any potentials tested (Fig. 9).

View larger version (13K): [in this window] [in a new window]   Fig. 9.   Effects of mapacalcine on the voltage-dependent potassium currents in duodenal and portal vein myocytes. A, In duodenal cells, potassium currents evoked by a depolarization step to (a) 0 mV and (b) current-voltage relationships in the () absence or () presence of 3 µM mapacalcine for 5 min, B, In portal vein cells, potassium currents evoked by a depolarization step to (a) 0 mV and (b) current-voltage relationships in the () absence or () presence of 3 µM mapacalcine for 5 min.

View larger version (12K): [in this window] [in a new window]   Fig. 10.   Effects of mapacalcine on both calcium release and chloride current evoked by 10 mM caffeine in portal vein myocytes. a, Control conditions. b, In the presence of 3 µM mapacalcine for 5 min. The cells were voltage-clamped at a holding potential of 50 mV. The pipette solution contained 50 µM Indo-1.

	Discussion

Summary Introduction Materials & Methods Results Discussion References

We demonstrate the existence of a new type of calcium channel coexisting along the typical L-type calcium channel in mouse duodenal smooth muscle cells. We also report the purification and characterization of a small protein from C. vastifica sponge that specifically blocks this calcium conductance.

The presence of several types of calcium channels in a same cellular type is a well-known situation (6-8). Even in visceral smooth muscle, the presence of both L-type and non-T/non-L-type calcium current has been suggested (9). Evidence supports that in mouse duodenal myocytes, the calcium current elicited by a depolarization from 80 mV (after complete blockade of L-type calcium channels) is a new type of calcium current, as follows. (a) This current is resistant to high stimulation frequency and to a variety of drugs, such as dihydropyridines, -conotoxin GVIA, -conotoxin MVIIC, and -agatoxin IVA, at the high concentrations used, indicating that it cannot be an L-, an N-, a P- or a Q-type calcium current. (b) The steady state inactivation curve is sigmoidal and voltage dependent. Half-maximal and complete inactivations are obtained at more negative potentials (70 and 45 mV, respectively) than those of the L-type calcium current (40 and 10 mV, respectively). (c) T-type calcium channels have been previously described in cardiac cells (25) and smooth muscle cells (7, 26). These channels are also activated when cells are maintained at a low holding potential, ~70 mV, and are resistant to high stimulation frequencies. Although T-type calcium channels were first described as dihydropyridine-resistant channels (27), recent studies report a certain sensitivity to dihydropyridines like isradipine (7, 28) and antispasmodic drugs like fenoverine (29). The non-L-type calcium current of duodenal myocytes seems completely insensitive to high concentrations of fenoverine and isradipine. (d) The new toxin, mapacalcine, selectively inhibits the non-L-type calcium current of duodenal myocytes. Mapacalcine has no effect on L-type calcium channels, potassium, or chloride channels in duodenal and portal vein myocytes. (e) Mapacalcine does not affect the calcium responses induced by caffeine, suggesting that it does not interact with the intracellular calcium stores. (f) The time constants of inactivation for the non-L-type calcium current in duodenal myocytes are slow compared with the fast inactivation decay of the T-type calcium current in other cell types (27). This may be due to the fact that the slow time constants of inactivation are very insensitive to changes in membrane potential and that only the fast time constants follow a bell-shaped curve against voltage. Another type of calcium channels, the R-type channel, has been recently described in neuronal tissues as a high voltage-activated calcium channel that is insensitive to the pharmacological drugs used in this study (15). However, this calcium current inactivates faster than the non-L-type calcium current of duodenal myocytes (30). Until now, R-type calcium channels have not been found in peripheral excitable tissues, and it is unlikely that the calcium conductance described in this work could correspond to a peripheral form of the R-type calcium channel. According to our data, mapacalcine seems to be very specific for a new type of calcium current, identified in duodenal myocytes. This current is activated at membrane potentials near 60 mV, has a slow inactivation decay, and is resistant to rundown by a high stimulation frequency and a variety of calcium channel blockers.

Our structural data show that mapacalcine is a dimeric protein composed of two identical chains of 9541 Da (corresponding to its reduced monomer with amidated carboxyl terminus). The amino acid sequence of a mapacalcine monomer does not show significant homologies with protein sequences in data banks. The polymeric structure of certain toxins seems to be associated with membrane insertion and channel-forming properties. For example, the mechanisms of action of aerolysin, a peptide secreted by Aeromonas hydrophyla, have been described. Aerolysin is secreted as a dimeric prototoxin able to cross the bacterial membranes; after proteolytic processing, it forms an heptamer that is able to insert into the cellular membrane and form a voltage-gated anion-selective channel (31, 32). The ant toxin, ectatomin, a 7928-Da peptide, is formed by two highly homologous polypeptide chains linked by a disulfide bond. It forms a bundle of four amphipatic  helices in aqueous solution (33). In its dimeric state and at concentrations of 10-50 µM, this toxin is able to form a potential-dependent nonselective cation channel in cell membranes (33). Although pore-forming peptides have been isolated from marine sponges (34), the mechanisms of action of mapacalcine cannot be attributed to a channel-forming activity, which could be expected with respect to its dimeric structure comparable to that of ectatomin. Our data clearly show that no leak current appears in the presence of mapacalcine. On the contrary, we observed a specific calcium current blockade. Mapacalcine seems, by the way, to be the first toxin sharing some structural properties of the family of channel-forming toxins but does not obviously act like them. It acts probably like classic channel-blocking toxins through direct interaction with the ionic channel or related proteins. Our experimental results are consistent with the binding of mapacalcine to resting calcium channels. Increasing the rate at which calcium channels are activated (0.2 Hz) as well as the number of inactivated calcium channels at depolarized holding potentials does not affect the inhibitory action of mapacalcine (absence of use and potential dependence). Further experiments using a labeled derivative of mapacalcine will be performed to investigate the interactions between the toxin and its receptors. Subunitary composition of the protein associated to the calcium channel described here will have to be elucidated, allowing comparison with subunitary composition of already described calcium channels. Improvement in the knowledge of the mapacalcine-sensitive calcium channel should provide informations regarding its contribution in the treatment of intestinal diseases associated to perturbations of cellular calcium homeostasis.