Search

1. PAP-1 (5-(4-phenoxybutoxy)psoralen), a potent small-molecule blocker of the voltage-gated potassium Kv1.3 channel, is currently in preclinical development for psoriasis. This study was undertaken to identify the major phase I metabolites of PAP-1 in Sprague-Dawley (SD) rats. 2. Five phase I metabolites, that is 5-(oxybutyric-acid)psoralen (M1), 5-[4-(4-hydroxybutoxy)]psoralen (M2), 5-[4-(4-hydroxyphenoxy)butoxy]psoralen (M3), 5-[4-(3-hydroxyphenoxy)butoxy]psoralen (M4), and 8-hydroxyl-5-(4-phenoxybutoxy)psoralen (M5), were isolated from the bile of rats and identified by mass spectrometry and NMR spectroscopy. The last four metabolites are new compounds. 3. Incubation of PAP-1 with SD rat liver microsomes rendered the same five major metabolites in a nicotinamide adenine dinucleotide phosphate (NADPH)-dependent manner suggesting that cytochrome P450 (CYP) enzymes are involved in PAP-1 metabolism. Inhibitors of rat CYP1A1/2 (alpha-naphthoflavone) and CYP3A (ketoconazole) but not CYP2D6 (quinidine), CYP2E (diethyldithiocarbamate), or CYP2C9 (sulphaphenazole) blocked the metabolism of PAP-1 in rat microsomes. 4. Of the five metabolites M3, M4, and M5 were found to inhibit Kv1.3 currents with nanomolar IC50s, while M1 and M2 were inactive. Our results identified the Kv1.3-inactive M1 as the major phase I metabolite, and suggest that hydroxylation and O-dealkylation are the major pathways of PAP-1 metabolism. 5. We further conducted a 6-month repeat-dose toxicity study with PAP-1 at 50 mg/kg in both male and female Lewis rats and did not observe any toxic effects.

This article reviews the structural and functional diversity of neuronal calcium channels and the therapeutic potential of antagonizing such channels. Through spatial and temporal control of intracellular calcium concentration, voltage-sensitive calcium channels regulate a host of neuronal processes, including neurotransmitter secretion, electrical activity, cytoskeletal function, cell metabolism and proliferation, and gene expression. Several genes elaborate a number of calcium channel isoforms or subtypes--each tailored to specific roles in neuronal function and possessing distinct biophysical properties, distribution, modulation, and pharmacological sensitivity. This diversity has raised the possibility that subtype-specific antagonists could provide novel treatments for some neuropathologies. In fact, neuroprotective and analgesic actions of N-type channel blockers in animals appear to confirm this supposition. These properties prompted human clinical studies evaluating these agents for prevention of neuronal degeneration following ischemic brain trauma and for relief of pain. Future medical applications for these blockers and antagonists of other channels subtypes are discussed.

Neuronal voltage-sensitive calcium channels (VSCCs) are a diverse family of proteins that regulate entry of Ca2+ into neurons. Selective antagonists of VSCCs have proven to be powerful pharmacological tools for identifying and characterizing these channels. A new VSCC antagonist, SNX-230 (also known as ω-conopeptide MVIIC), binds with high affinity to receptors in rat brain and blocks one or more high-threshold VSCCs that are neither L- nor N-type. We have defined the neuroanatomical distribution of the high-affinity non-L, non-N VSCC receptors for SNX-230 using [125I]SNX-230 bound to rat brain sections and compared it with that of [125I]SNX-111, a reversible blocker of N-type VSCCs. Highest densities of binding for both ligands were seen in areas rich in synaptic connections, such as the oriens, radiatum and molecular layers of the hippocampus. In general, the density of [125I]SNX-230-binding was higher in cerebellum compared with that in forebrain. In contrast, this general distribution of density was reversed for [125I]SNX-111. In the glomeruli of the olfactory bulb, binding of [125I]SNX-230 was undetectable compared with the high density of [125I]SNX-111-binding. Differential localization of the two ligands was also seen in cervical spinal cord. The clearly different localization of [125I]SNX-230 compared with that of [125I]SNX-111 in the olfactory bulb and spinal cord suggested that the binding sites for [125I]SNX-230 in other brain regions, while co-localized macroscopically, are also distinct from those for [125I]SNX-111. This was confirmed when addition of saturating concentrations of SNX-111 did not affect the distribution pattern of [125I]SNX-230-binding. These observations demonstrate that the high-affinity [125I]SNX-230 receptors in the brain and those of the N-type VSCCs are distinct molecular entities. Thus, SNX-230 (ω-MVIIC) offers a useful ligand for investigating the distribution and functions of non-L-, non-N-type VSCCs in the brain.

The level of intracellular Ca regulates many cellular processes, including neurotransmitter and hormone secretion, the activity of ion channels and enzymes, cytoskeletal function, cell proliferation, and gene expression. Voltage-sensitive Ca channels are among the most heterogeneous of ion channels. In neurons, Ca channels differ in cellular location, biophysical and pharmacological properties, and modulation. A single neuron generally contains multiple types of Ca channels, and such channels are central to the integration and expression of activity in the nervous system. It is clearly important to understand the functional significance of Ca channel diversity; a major research effort that is under way has made clear that different calcium channel types are all part of a family of multisubunit ion channels.

Three neurotoxic peptides from the venom of Conus striatus have been purified, biochemically characterized, and chemically synthesized. One of these, an acetylcholine receptor blocker designated alpha-conotoxin SII, has the sequence GCCCNPACGPNYGCGTSCS. In contrast to all other alpha-conotoxins, SII has three disulfide bonds (instead of two), has no net positive charge, and has a free C-terminus. The other two paralytic peptides are Ca channel-targeted omega-conotoxins, SVIA and SVIB. omega-SVIA is the smallest natural omega-conotoxin so far characterized and has the sequence CRSSGSPCGVTSICCGRCYRGKCT-NH2. Although omega-conotoxin SVIA is a potent paralytic toxic in lower vertebrate species, it was much less effective in mammals. The third toxin, omega-conotoxin SVIB, has the sequence CKLKGQSCRKTSYDCCSGSCGRSGKC-NH2. This peptide has a different pharmacological specificity from other omega-conotoxins previously purified from Conus venoms; only omega-conotoxin SVIB has proven to be lethal to mice upon ic injection. Binding competition experiments with rat brain synaptosomal membranes indicate that the high-affinity binding site for omega-conotoxin SVIB is distinct from the high-affinity omega-conotoxin GVIA or MVIIA site.

Voltage-sensitive Ca2+ channels that control neurotransmitter release are blocked by omega-conotoxin (omega-CgTx) GVIA from the marine snail Conus geographus, the most widely used inhibitor of neurotransmitter release. However, many mammalian synapses are omega-CgTx-GVIA insensitive. We describe a new Conus peptide, omega-CgTx-MVIIC, that is an effective inhibitor of omega-CgTx-GVIA-resistant synaptic transmission. Ca2+ channel targets that are inhibited by omega-CgTx-MVIIC but not by omega-CgTx-GVIA include those mediating depolarization-induced 45Ca2+ uptake in rat synaptosome preparations, "P" currents in cerebellar Purkinje cells, and a subset of omega-CgTx-GVIA-resistant currents in CA1 hippocampal pyramidal cells. The characterization of omega-CgTx-MVIIC by a combination of molecular genetics and chemical synthesis defines a general approach for obtaining ligands with novel receptor subtype specificity from Conus.

Our drug discovery efforts for N-type calcium channel blockers in the 4-piperidinylaniline series led to the discovery of an orally active analgesic agent 26.1-[4-Dimethylamino-benzyl)-piperidin-4-yl]-[4-(3,3-dimethyl-but yl)-phenyl]-(3-methyl-but-2-enyl)amine (26) showed high affinity to functionally block N-type calcium channels (IC50=0.7 microM in the IMR32 assay) and exhibited high efficacy in the anti-writhing analgesia test with mice (ED50=12 mg/kg by po and 4 mg/kg by iv). In this report, the rationale for the design, synthesis, biological evaluation, and pharmacokinetics of this series of blockers is described.

Ziconotide is a selective peptide antagonist of the N-type calcium channel currently in clinical trials for analgesia. Ziconotide reached a maximal brain concentration of between 0.003 and 0.006% of the injected material per gram of tissue at 3–20 min after i.v. injection, and this decayed to below 0.001%/g after 2 h. The structurally distinct conopeptide SNX-185 (synthetic TVIA) was considerably more persistent in brain after i.v. administration, with 0.0035% of the injected material present at 2–4 h after i.v. injection, and 0.0015% present at 24 h. Similar results (i.e. greater persistence of SNX-185) were obtained when the peptides were perfused through in vivo dialysis probes implanted into the hippocampus. Image analysis and serial sectioning showed that diffusion of Ziconotide in the extracellular fluid around the dialysis probe was minimal, with the peptide located within 1 mm of the probe after 2 h. In vitro diffusion through cultured bovine brain microvessel endothelial cells (BBMEC) verified that a close structural analog of Ziconotide (SNX-194) passed through this blood–brain barrier (BBB) model as expected for peptides of similar physical properties (permeability coefficient of 6.5 × 10−4 cm/g). Passage from blood to brain was also verified by in situ perfusion through the carotid artery. A statistically greater amount of radioactivity was found to cross the BBB after perfusion of radioiodinated Ziconotide compared to [14C]inulin. Capillary depletion experiments and HPLC analysis defined the brain location and stability.

Selective N-Type Voltage Sensitive Calcium Channel (VSCC) antagonists have shown utility in several models of pain and ischemia. We report the structure-activity relationship at the proximal phenyl group in a series of non-peptidyl VSCC blockers.

Selective N-type voltage sensitive calcium channel (VSCC) blockers have shown efficacy in several animal models of stroke and pain. In the process of searching for small molecule N-type calcium channel blockers, we have identified a series of N-methyl-N-aralkyl-peptidylamines with potent functional activity at N-type VSCCs. The most active compound discovered in this series is PD 173212 (11, IC50 = 36 nM in the IMR-32 assays). SAR and pharmacological evaluation of this series are described.

Selective N-type voltage sensitive calcium channel (VSCC) blockers have shown utility in several models of stroke and pain. We are especially interested in small molecule N-type calcium channel blockers for therapeutic use. Herein, we report a series of N,N-dialkyl-dipeptidylamines with potent functional activity at N-type VSCCs and in vivo efficacy. The synthesis, SAR, and pharmacological evaluation of this series are discussed.

We describe the first potent and selective blocker of the class E Ca2+channel. SNX-482, a novel 41 amino acid peptide present in the venom of the African tarantula, Hysterocrates gigas, was identified through its ability to inhibit human class E Ca2+ channels stably expressed in a mammalian cell line. An IC50 of 15-30 nM was obtained for block of the class E Ca2+ channel, using either patch clamp electrophysiology or K+-evoked Ca2+ flux. At low nanomolar concentrations, SNX-482 also blocked a native resistant or R-type Ca2+ current in rat neurohypophyseal nerve terminals, but concentrations of 200-500 nM had no effect on R-type Ca2+ currents in several types of rat central neurons. The peptide has the sequence GVDKAGCRYMFGGCSVNDDCCPRLGCHSLFSYCAWDLTFSD-OH and is homologous to the spider peptides grammatoxin S1A and hanatoxin, both peptides with very different ion channel blocking selectivities. No effect of SNX-482 was observed on the following ion channel activities: Na+ or K+ currents in several cultured cell types (up to 500 nM); K+ current through cloned potassium channels Kv1.1 and Kv1. 4 expressed in Xenopus oocytes (up to 140 nM); Ca2+ flux through L- and T-type Ca2+ channels in an anterior pituitary cell line (GH3, up to 500 nM); and Ba2+ current through class A Ca2+ channels expressed in Xenopus oocytes (up to 280 nM). A weak effect was noted on Ca2+ current through cloned and stably expressed class B Ca2+ channels (IC50500 nM). The unique selectivity of SNX-482 suggests its usefulness in studying the diversity, function, and pharmacology of class E and/or R-type Ca2+ channels.

The interaction between the proteins syntaxin 1A and SNAP-25 is a key step in synaptic vesicle docking and fusion. To define the SNAP-25 binding domain on syntaxin, we have prepared peptides that span the syntaxin H3 domain (residues 191-266), the region previously shown to be important for binding to SNAP-25, and then determined the affinities of these peptides for binding to SNAP-25. A minimal binding domain was identified within a region of 32 amino acids (residues 189-220). Its affinity for SNAP-25 is substantially enhanced by C-terminal extension (residues 221-266). Circular dichroism revealed the presence of substantial alpha-helicity in the H3 domain and in the 32-mer minimal binding domain, but not in H3 peptides that do not bind to SNAP-25. At temperatures that denature the alpha-helix of the minimal binding domain peptide, SNAP-25 binding is lost. Selected mutations in evolutionarily conserved residues of the amphiphilic alpha-helix within the minimal binding domain (e.g., residues 205 and 209) greatly reduce the affinity for SNAP-25 but have no major effect on secondary structure, suggesting that these residues may interact directly with SNAP-25. The H3 domain peptide and the minimal binding domain peptide inhibit norepinephrine release from PC12 cells. These results suggest that specific amino acid residues in the H3 domain, positioned by the underlying alpha-helical structure, are important for its binding to SNAP-25 and support the notion that this interaction is important for presynaptic vesicular exocytosis.

Biochemical properties of the alpha 1 subunits of class A brain calcium channels (alpha 1A) were examined in adult rat brain membrane fractions using a site-directed anti-peptide antibody (anti-CNA3) specific for alpha 1A. Anti-CNA3 specifically immunoprecipitated high affinity receptor sites for omega-conotoxin MVIIC (Kd approximately 100 pM), but not receptor sites for the dihydropyridine isradipine or for omega-conotoxin GVIA. In immunoblotting and immunoprecipitation experiments, anti-CNA3 recognized at least two distinct immunoreactive alpha 1A polypeptides, a major form with an apparent molecular mass of 190 kDa and a minor, full-length form with an apparent molecular mass of 220 kDa. The 220- and 190-kDa alpha 1A polypeptides were also specifically recognized by both anti-BI-Nt and anti-BI-1-Ct antibodies, which are directed against the NH2- and COOH-terminal ends of alpha 1A predicted from cDNA sequence, respectively. These data indicate that the predicted NH2 and COOH termini are present in both size forms and therefore that these isoforms of alpha 1A are created by alternative RNA splicing rather than post-translational proteolytic processing of the NH2 or COOH termini. The 220-kDa form was phosphorylated preferentially by cAMP-dependent protein kinase, whereas protein kinase C and cGMP-dependent protein kinase preferentially phosphorylated the 190-kDa form. Our results identify at least two distinct alpha 1A subunits with different molecular mass, demonstrate that they may result from alternative mRNA splicing, and suggest that they may be differentially regulated by protein phosphorylation.

Voltage-gated calcium channels in small-cell lung carcinomas may initiate autoimmunity in the paraneoplastic neuromuscular disorder Lambert-Eaton syndrome. The calcium-channel subtype that is responsible is not known.We compared the effects of antagonists of L-type, N-type, and P/Q-type neuronal calcium channels on the depolarization-dependent influx of calcium-45 in cultured carcinoma cells. Serum samples from patients with various disorders were tested for reactivity with P/Q-type channels solubilized from carcinoma and cerebellar membranes and N-type channels from cerebral cortex.P/Q-type calcium-channel antagonists were the most potent inhibitors of depolarization-induced 45Ca influx in cultured small-cell carcinoma cell lines. Anti-P/Q-type calcium-channel antibodies were found in serum from all 32 patients with Lambert-Eaton syndrome and a diagnosis of cancer and in 91 percent of the 33 patients with Lambert-Eaton syndrome without cancer. Anti-N-type calcium-channel antibodies were found in 49 percent of the 65 patients with the Lambert-Eaton Syndrome. Lower titers of anti-P/Q-type and anti-N-type calcium-channel antibodies were found in 54 percent of 70 patients with a paraneoplastic encephalomyeloneuropathic complication of lung, ovarian, or breast carcinoma, 24 percent of 90 patients with cancer but no evident neurologic complications, 23 percent of 78 patients with sporadic amyotrophic lateral sclerosis, and less than 3 percent of 69 patients with myasthenia gravis, epilepsy, or scleroderma.The high frequency of P/Q-type calcium-channel antibodies found in patients with Lambert-Eaton syndrome implies that antibodies of this specificity have a role in the presynaptic pathophysiology of this disorder.

Voltage-sensitive Ca2+ channels are essential to transmitter release at the chemical synapse. To demonstrate the localization of voltage-sensitive Ca2+ channels in relation to the site of transmitter release, mouse neuromuscular junctions were double labelled with alpha-bungarotoxin and a novel voltage-sensitive Ca2+ channel probe, SNX-260, a synthetic analog of omega-conopeptide MVIIC. Similar to omega-conopeptide MVIIC, biotinylated SNX-260 blocked nerve-stimulated transmitter release at the mouse neuromuscular junction. Fluorescently-tagged biotinylated SNX-260 labelled the nerve terminal which appeared thinner than and was outlined by acetylcholine receptor clusters as seen in en face view. This SNX-260 labelling was inhibited by preincubation with unconjugated SNX-260. Side-views of the neuromuscular junction indicated that the SNX-260 labelling was on the synaptic side facing the acetylcholine receptor rather than on the nonsynaptic side of the nerve terminal. This presynaptic binding was confirmed by the absence of SNX-260 labelling in denervated muscles following a nerve cut or disjunction after collagenase treatment. Confocal microscopy revealed spots of SNX-260 labelling that may correlate with active zones. The SNX-260 labelling pattern was not affected by preincubation with unconjugated SNX-111 (omega-conopeptide MVIIA), an N-type voltage-sensitive Ca2+ channel blocker. These findings suggest that SNX-260 is a novel probe for localizing non-N type voltage-sensitive Ca2+ channels and that these voltage-sensitive Ca2+ channels are localized near the transmitter release sites at the mammalian motor nerve terminal membrane. The results are consistent with the suggestion that non-N, probably P/Q type voltage-sensitive Ca2+ channels mediate evoked transmitter release at the mammalian neuromuscular junction.

High-threshold voltage-sensitive calcium channels of the N-type, L-type, and P-type have been distinguished in the mammalian CNS predominantly on the basis of their sensitivity to selective antagonists. Matching them with genes identified by molecular cloning is an ongoing undertaking. Whereas L-type channels are characterized by their sensitivity to dihydropyridines and P-type channels by sensitivity to the funnel-web spider toxin AgaIVA, the N-type channel has been shown to be recognized by the omega-conopeptides GVIA and MVIIA. Recently, two new members of the family of omega-conopeptides--MVIIC from the marine snail Conus magus and SVIB from Conus striatus--have been described. Binding and electrophysiological data suggest that these two peptides, in addition to interacting with N-type calcium channels, interact with a widely distributed receptor in neuronal membranes that is distinct from N-type channels. In this report we demonstrate through biochemical and pharmacological differentiation at individual receptor polypeptide resolution, by affinity cross-linking, SDS-PAGE, and autoradiography, that SNX-230 (synthetic MVIIC) binds with high affinity to a calcium channel alpha 1 subunit distinct from the high-affinity alpha 1 target of SNX-111 (synthetic MVIIA). SNX-183 (synthetic SVIB) interacts with both alpha 1 subunits with lower affinity. Whereas the alpha 1 subunit recognized with high affinity by MVIIA corresponds to the N-type channel, the other represents a novel calcium channel distinct from N-, L-, and perhaps P-type channels.

The interaction of two synthetic omega-conopeptides SNX-111 (MVIIA) and SNX-230 (MVIIC) both derived from the marine snail Conus magus, with non-L-type neuronal voltage-sensitive calcium channels (VSCC) in rat brain synaptosomal preparations has been investigated with the aid of well-characterized 125I derivatives of the two peptides. To assess the effects of iodination on the binding characteristics of SNX-111 and SNX-230, the corresponding peptides containing monoiodotyrosine in place of tyrosine, namely, SNX-259 ([127I]SNX-111) and SNX-260 ([127I]SNX-230), respectively, were prepared by solid-phase synthesis. Saturation analysis showed that [125I]SNX-111 and [125I]SNX-230 bound to two distinct classes of high-affinity sites with apparent Kd's of 9 and 11 pM and Bmax's of 0.54 and 2.2 pmol/mg protein, respectively. Kinetic analysis revealed that both peptides exhibited high association rates as well as rapid dissociation rates in contrast to the 125I derivative of the synthetic omega-conopeptide from Conus geographus, GVIA (SNX-124), which binds irreversibly to N-type channels in rat brain synaptosomes. Competition binding experiments with [125I]SNX-111 and [125I]SNX-124 established that both of them bind to the same site, namely, N-type VSCC. The site detected by the binding of [125I]SNX-230 is distinct from N-type VSCC since SNX-111 has very low affinity (K(i) = 135 nM) in competition studies. Recent findings that a novel high-voltage-activated calcium channel in rat cerebellar granule neurons is resistant to blockers of L-, N-, and P-type VSCC but is highly sensitive to SNX-230 suggest that the [125I]SNX-230 binding site may represent this novel type of calcium channel or another, as yet undescribed, VSCC.

A cDNA clone encoding a new omega-conotoxin was identified from Conus magus. The predicted peptide was chemically synthesized using a novel strategy that efficiently yielded the biologically active disulfide-bonded isomer. This peptide, omega-conotoxin MVIID, targets other voltage-gated calcium channels besides the N-subtype and exhibits greater discrimination against the N-channel subtype than any other omega-conotoxin variant to date. Consequently, omega-conotoxin MVIID may be a particularly useful ligand for calcium channel subtypes that are not of the L- or N-subclasses. Of the eight major sequence variants of omega-conotoxins that have been elucidated, four come from Conus magus venom. We suggest that sequence variants from the same venom may be designed to optimally interact with different molecular variants of calcium channels; such omega-conotoxin sets from a single venom may therefore be useful for helping to identify novel calcium channel subtypes.

We report here the solution structure of omega-conotoxin GVIA, a peptide antagonist of the N-type neuronal voltage-sensitive calcium channel. The structure was determined using two-dimensional NMR in combination with distance geometry and restrained molecular dynamics. The full relaxation matrix analysis program MARDIGRAS was used to generate maximum and minimum distance restraints from the crosspeak intensities in NOESY spectra. The 187 restraints obtained were used in conjunction with 23 angle restraints from vicinal coupling constants as input for the structure calculations. The backbones of the best 21 structures match with an average pairwise RMSD of 0.58 A. The structures contain a short segment of triple-stranded beta-sheet involving residues 6-8, 18-21, and 24-27, making this the smallest published peptide structure to contain a triple-stranded beta-sheet. Conotoxins have been shown to be effective neuroprotective agents in animal models of brain ischemia. Our results should aid in the design of novel nonpeptide compounds with potential therapeutic utility.

Previous work has demonstrated that the neurotoxin leptinotarsin elicits release of neurotransmitter from mammalian nerve terminals, and it has been suggested that the toxin may act either as a direct agonist of voltage-sensitive calcium channels in these terminals (Crosland et al., 1984) or as a calcium ionophore (Madeddu et al., 1985a,b). Preliminary studies (Yeager et al., 1987) demonstrated that leptinotarsin also evokes transmitter release from isolated elasmobranch electric organ nerve terminals. We now report further investigations of the effects of leptinotarsin in this system. The action of the toxin is saturable, releasing about the same small fraction of total transmitter as that released by depolarization. An upper limit for the concentration for half maximal release is estimated to be 4 nM. Leptinotarsin-evoked transmitter release exhibits behavior very similar to depolarization-evoked release with respect to dependence on Ca2+, Ba2+, and Sr2+ and blockade by Co2+, Cd2+, and trifluoperazine. Leptinotarsin also promotes the uptake of calcium into synaptosomes to a degree similar to that caused by depolarization by K+. The binding of leptinotarsin to nerve terminals is probably Ca2+ dependent and receptor mediated. Taken together with the behavior of leptinotarsin-evoked release in other preparations, these results are consistent with the hypothesis that this toxin acts by opening a presynaptic calcium channel. However, the possibility that leptinotarsin is a calcium ionophore cannot be excluded.

The transmembrane distribution of the major aminophospholipids in the bovine retinal rod outer segment disk membrane, phosphatidylethanolamine and phosphatidylserine, was determined using a novel pair of permeable and impermeable covalent modification reagents. The values for the percentages of phosphatidylethanolamine and phosphatidylserine in the outer monolayer were calculated from a simple expression which takes into account the leakage of impermeable reagent into the disk lumen as monitored by the extent of labeling of lysine entrapped in the lumen. We infer from our results that at least 73 to 87% of the disk phosphatidylethanolamine and 77 to 88% of the disk phosphatidylserine are in the outer disk membrane monolayer. The fatty acid composition of the inner aminophospholipids is slightly more saturated than the outer aminophospholipids. Calculations using the lateral surface areas occupied by the disk membrane lipids suggest that 65 to 100% of the disk phosphatidylcholine is on the inner membrane surface. Since the disk phosphatidylcholine is also somewhat more saturated than the phosphatidylethanolamine and phosphatidylserine of the outer monolayer, the total inner membrane monolayer fatty acid composition is more saturated than that of the outer monolayer fatty acid composition.

We have previously shown that the calcium channel antagonist omega-conotoxin M-VII-A blocks neurotransmitter release from isolated nerve terminals (synaptosomes) from the electric organ of the electric ray (Yeager et al., J. Neurosci., 7 (1987) 2390-2396). We now demonstrate that a related but more readily available peptide, omega-conotoxin G-VI-A (CgTx), also blocks the release of transmitter from these terminals and, in addition, inhibits depolarization-dependent uptake of Ca2+ into these terminals. The half-maximal inhibitory concentration (IC50 for block of depolarization-evoked release and for depolarization-dependent uptake of Ca2+ are approximately 3 and 2 microM, respectively. These results suggest the inhibitory effects of CgTx are due to the inhibition of Ca2+ entry into synaptosomes through voltage-sensitive calcium channels. Assays of radioiodinated CgTx binding to electric organ synaptosomal membranes and synaptosomes appear to show a single binding site with an apparent dissociation constant (Kd) of 3-5 microM and toxin receptor densities of 290 and 52 pmol/mg protein, respectively. These CgTx receptor densities are equivalent to 6% of the total synaptosomal membrane protein and 1% of the total synaptosomal protein (assuming a molecular weight of 200 kDa for the toxin receptor). If the observed CgTx receptor densities reflect the actual densities of voltage-sensitive calcium channels in electric organ synaptosomal membranes and synaptosomes, these preparations would be the richest source of these channels yet described.

A large fraction of the phospholipid protons of bovine retinal rod outer segment (ROS) disc membrane vesicles yield well-resolved nuclear magnetic resonance lines near physiological temperature. The spin-lattice (T1) relaxation rates of the resolved sharp resonance of ROS disc membranes appear biphasic above 10 degrees C. The rate of the more rapidly relaxing component of each resonance matches closely the relaxation rate of the corresponding resonances of liposomes of purified ROS phospholipids. The slowly relaxing component of each disc membrane resonance is most likely due to phospholipids whose motion is affected by rhodopsin. The primary difference in the relaxation behavior of phospholipids in the ROS membrane vesicles and ROS liposomes appears to be in T1, rather than T2, since the corresponding sharp resonances of both preparations have similar linewidths. These observations suggest that the interaction of rhodopsin with the more fluid membrane phospholipids predominantly affects relatively high frequency segmental motions, which determine T1, while having minimal effects on the lower frequency segmental motions, which influence T2. This conclusion can be rationalized by assuming that a substantial fraction of the interacting phospholipids are relatively fluid with respect to less frequent, larger amplitude segmental motions, but that the more frequent segmental motions (such as beta-coupled trans-gauche isomerizations) are significantly restricted by interaction with protein.

The solution structure of omega-conotoxin MVIIA (SNX-111), a peptide toxin from the fish hunting cone snail Conus magus and a high-affinity blocker of N-type calcium channels, was determined by 2D NMR spectroscopy. The backbones of the best 44 structures match with an average pairwise RMSD of 0.59 angstroms. The structures contain a short segment of triple-stranded beta-sheet involving residues 6-8, 20-21, and 24-25. The structure of this toxin is very similar to that of omega-conotoxin GVIA with which is has only 40% sequence homology, but very similar calcium channel binding affinity and selectivity.

Well resolved proton nuclear magnetic resonance ('H NMR) spectra of bovine retinal rod outer segment (ROS) disk membranes have been obtained at 100 and 360 MHz. The resolved IH resonances of the ROS membranes are due to phospholipids, with little contribution from rhodopsin. The spectra of both the ROS membranes and bilayer vesicles pre- pared from purified ROS phospholipids (liposomes) appear to represent a superposition of relatively sharp resonance components and a broad, underlying background. The distri- bution between sharp and broad spectral components is sen- sitive to sonication and temperature. The percentage of choline methyl protons which are resolved in the ROS membrane spectra as sharp resonance components increases from ap- proximately 35 to 100% and the average of the lipid hydro- carbon chain protons from approximately 20 to 40% over the Nuclear magnetic resonanie (NMR)~ has been used to elucidate details of motion (Horwitz et al., 1972; 1973b; Levine et al., 1972), phase behavior (Chapman, 1975), and asymmetry (Michaelson et al., 1974) in pure and mixed phospholipid bi- layer membranes. In addition, well resolved NMR spectra have been obtained of several natural membranes (see, e.g., Lee et al., 1974a; Davis et al., 1976). A relatively simple membrane system which appears to be particularly amenable to magnetic resonance studies is the rod outer segment (ROS) disk mem- brane from the vertebrate retina (for reviews see Hagins, 1972; Daemen, 1973; Hubbell, 1975; Ebrey and Honig, 1975; Dratz, 1977). A single protein, rhodopsin, comprises at least 85% of the total protein content. The ROS phospholipids are partic- ularly rich in polyunsaturated fatty acids, with docosahexenoic acid (C22:6w3) comprising at least 3 1-37% of the total fatty acid composition of bovine ROS disk membranes* (Nielson et al., 1970; Anderson and Sperling, 1971). The visual pigment rhodopsin rotates and translates freely in the ROS disk

The plasma membrane of electric organ nerve terminals has two domains that can be distinguished by monoclonal antibodies. A library of 111 mouse monoclonal antibodies raised to nerve terminals from Torpedo californica contains 4 antibodies that bind specifically to the outside of intact synaptosomes. The distribution of the binding sites of these monoclonal antibodies on the outside of intact nerve terminals was examined by immunofluorescence and immunoelectron microscopy. The binding sites of 3 (tor23, 25, and 132) are distributed uniformly over nerve trunks and fine terminal branches. The binding site of the fourth (tor70) is restricted to synaptic junctional regions. This antibody, but not the other 3, recognizes a major component of synaptic vesicles, a proteoglycan associated with the inner surface of the vesicle membrane. The difference in the pattern of binding of these monoclonal antibodies suggests that the region of the plasma membrane containing active zones is antigenically distinguishable from other nerve terminal plasma membrane. We suggest that the antigen recognized by tor70 is externalized by exocytosis of synaptic vesicles while other plasma antigens take a different route to the surface. The unexpected observation that the vesicle antigen remains on the surface after exocytosis and is prevented from diffusion from the synaptic junctional region would be consistent with an interaction between the vesicle proteoglycan and elements of the synaptic cleft.

The transmembrane disposition of the visual pigment rhodopsin was studied by the covalent labeling of protein amino groups with membrane-permeable and -impermeable imidoesters. A new, highly reactive permeable reagent, 2-(methylsulfonyl)ethyl acetimidate (SAI) was developed for this purpose. The permeabilities of both this compound and the "impermeable" reagent isethionyl acetimidate (IAI) across the rod outer segment disk membrane were directly measured. Our results indicate that rhodopsin contains three classes of amino groups. One class (35--55% of the total) reacts rapidly with the membrane-impermeable reagent and is presumably exposed on the outside surface of the membrane. A second class (35--55% of the total) is located on the internal surface of the disk since its rate of reaction is dependent on the relative permeabilities of the labeling reagents. The remaining 10% of the rhodopsin amino groups are inaccessible to either type of imidate and are largely accounted for by the single lysine residue which specifically binds the chromophore retinal. These findings, taken together with evidence from freeze--fracture electron microscopy, imply that rhodopsin is a transmembrane protein.

During transmitter release, synaptic vesicle membrane is specifically inserted into the nerve terminal plasma membrane only at specialized sites or "active zones." In an attempt to obtain a membrane fraction enriched in active zones, we have utilized the electric organ of the marine ray. From this organ, a fraction enriched in nerve terminals (synaptosomes) was prepared by conventional means. These synaptosomes were bound to microscopic beads by an antiserum to purified electric organ synaptic vesicles (anti-SV). The success of this immunoadsorption procedure was demonstrated by increased specific activities of bead-bound nerve terminal cytoplasmic markers and decreased specific activities of markers for contaminating membranes. To obtain a presynaptic plasma membrane (PSPM) fraction, we lysed the bead-bound synaptosomes by hypoosmotic shock and sonication, resulting in complete release of cytoplasmic markers. When the synaptosomal fraction was surface-labeled with iodine before immunoadsorption, 10% of this label remained bead-bound after lysis, compared with 2% of the total protein, indicating an approximately fivefold enrichment of bead-bound plasma membrane. Concomitantly, the specific activity of bead-bound anti-SV increased approximately 30-fold, indicating an enrichment of plasma membrane which contained inserted synaptic vesicle components. This PSPM preparation is not simply synaptic vesicle membrane since two-dimensional electrophoresis revealed that the polypeptides of the surface-iodinated PSPM preparation include both vesicle and numerous nonvesicle components. Secondly, antiserum to the PSPM fraction is markedly different from anti-SV and binds to external, nonvesicle, nerve terminal components.

Thin-layer chromatography was used to separate the major phospholipid headgroup classes of the rod outer segment disk membrane into subfractions which differ markedly in fatty acid composition. At least 18% of the rod outer segment phosphatidylcholine must contain two saturated fatty acids. Furthermore, two unsaturated fatty acids are found in at least 43% of the phosphatidylserine, 24% of the phosphatidylcholine, and 24% of the phosphatidylethanolamine. The unsaturated acids are predominantly polyunsaturated in all cases. A similar separation, but with less resolution, was achieved with silicic acid column chromatography. The temperature dependence of the polarization of the fluorescence of trans-parinaric acid (9,11,13,15-all-trans-octadecatetraenoic acid) showed that the thermal behavior of aqueous dispersions of the phosphatidylcholine subfractions was consistent with their fatty acid compositions.

Publisher Summary This chapter discusses the fatty acid composition and pairing in phospholipids of rod outer segments (ROS). In common with the phospholipids of most other biological membranes, the phospholipids of the ROS disk membrane possess a large measure of molecular heterogeneity. The major phospholipids of the ROS disk membrane have now been subfractionated by a simple thin-layer chromatography (TLC) method based on fatty acid composition. An analysis of this subfractionation has yielded estimates of the percentages of each major phospholipid class that are present as the general subclasses containing two saturated fatty acids, one saturated and one unsaturated fatty acid, or two unsaturated fatty acids. The TLC method described in this chapter is a simple procedure that minimizes manipulations of the ROS lipids. This is an important consideration, as minimal handling of these highly polyunsaturated molecules reduces the chances for oxidative degradation.

Publisher Summary This chapter discusses the chemical modification of rhodopsin with imidoesters: synthesis of reagents, membrane permeability of reagents, and modification methods. Chemical modification of biological macromolecules has yielded a wealth of information with regard to their structure and function. For example, covalent labeling of membranes by membrane-permeable and impermeable reagents provides a useful probe for studying the molecular architecture of these structures. This chapter describes the use of this approach in the study of the topological arrangement of rhodopsin within the rod outer segment (ROS) disk membrane. The reagents employed were isethionyl acetimidate (IAI), an imidoester first developed to investigate the structure of the erythrocyte membrane, and two newly synthesized imidates, ethyl acetimidate (SAI) and choline acetimidate (CAI). IAI and CAI are charged molecules and hence have relatively low membrane permeabilities, whereas SAI is highly membrane permeable.

The partition of cis-parinaric acid (9,11,13,15-cis, trans, trans,cis-octadecatetraenoic acid, cis-PnA) and trans-parinaric acid (9,11,13,15-all-trans-octadecatetraenoic acid, trans-PnA) among aqueous, solid lipid, and fluid lipid phases has been measured by three spectroscopic parameters: absorption spectral shifts, fluorescence quantum yield, and fluorescence polarization. The solid lipid was dipalmitoylphosphatidylcholine (DPPC); the fluid lipid was palmitoyldocosahexaenoylphosphatidylcholine (PDPC). Mole fraction partition coefficients between lipid and water were determined by absorption spectroscopy to be for ci--PnA, 5.3 X 10(5) with a solid lipid and 9 X 10(5) with fluid lipid and, for trans-PnA, 5 X 10(6) with solid lipid and 1.7 X 10(6) with fluid lipid. Ratios of the solid to the fluid partition coefficients (Kps/f) are 0.6 +/- 0.2 for cis-PnA and 3 +/- 1 for trans-PnA. A phase diagram for codispersions of DPPC and PDPC has been constructed from the measurements of the temperature dependence of the fluorescence quantum yield and polarization of cis-PnA and trans-PnA and their methyl ester derivatives. A simple analysis based on the phase diagram and fluorescence data allows additional calculations of Kps/f's which are determined to be 0.7 +/- 0.2 for the cis probes and 4 +/- 1 for the trans probes. The relative preference of trans-PnA for solid phase lipids and its enhanced quantum yield in solid phase lipids make it sensitive to a few percent solid. The trans probes provide evidence that structural order may persist in dispersions of these phospholipids 10 degrees C or more above their transition temperature. It is concluded that measurements of PnA fluorescence polarization vs. temperature are better suited than measurements of quantum yield vs. temperature for determining phospholipid phase separation.

Publisher Summary This chapter focuses on the ω-conotoxins and their receptor targets. ω-conotoxins show remarkable tissue specificity, at least in comparison with other presently known Ca channel blockers, including the dihydropyridines. Thus, at the present time ω-conotoxins are known to target only voltage-sensitive Ca channels in neuronal tissue but not Ca channels in skeletal muscle, smooth muscle, or cardiac muscle. However, within vertebrate neuronal tissue, the chapter presents evidence that the ω-conotoxins have multiple receptor subtypes. The results of these studies suggest that, at presynaptic termini, there may well be several different voltage-sensitive Ca channel subtypes.

Calcium channels are represented by at least 9 distinct genes (calcium channel classes A-I), corresponding to at least 5 functional and pharmacological “types” (L, N, P/Q, R and T). Selective L-, N-, and T-type channel antagonists are either in clinical use or in late stage clinical trials, while antagonists of P/Q channels are known to be toxic. No selective ligand has been identified for the R-type (class E), and its function and pharmacology are consequently, poorly understood. We review recent work on the discovery and initial characterization of SNX-482, the first known selective antagonist of R-type calcium channels. SNX-482 is a 41 residue acidic peptide with three disulfide bonds that has been isolated from the venom of the African tarantula, Hysterocrates gigas. In cell-based assays, it is a potent and selective inhibitor of the class E or R-type calcium channel. SNX-482 blocks some but not all native R-type currents: it blocks an R-type current in vertebrate neurohypophysis, but it does not block an R-type current in cerebellar granule cells. The peptide blocks oxytocin but not vasopressin release, suggesting a possible utility for SNX-482 as a neuroendocrine modulator. The peptide possesses antiseizure activity in several animal models of epilepsy, suggesting that class E antagonists may have pharmacological use in seizure disorders.