Molecular mechanism of cAMP modulation of HCN pacemaker channels.

Hyperpolarization-activated cation channels of the HCN gene family contribute to spontaneous rhythmic activity in both heart and brain. All four family members contain both a core transmembrane segment domain, homologous to the S1-S6 regions of voltage-gated K+ channels, and a carboxy-terminal 120 amino-acid cyclic nucleotide-binding domain (CNBD) motif. Homologous CNBDs are responsible for the direct activation of cyclic nucleotide-gated channels and for modulation of the HERG voltage-gated K+ channel--important for visual and olfactory signalling and for cardiac repolarization, respectively. The direct binding of cyclic AMP to the cytoplasmic site on HCN channels permits the channels to open more rapidly and completely after repolarization of the action potential, thereby accelerating rhythmogenesis. However, the mechanism by which cAMP binding modulates HCN channel gating and the basis for functional differences between HCN isoforms remain unknown. Here we demonstrate by constructing truncation mutants that the CNBD inhibits activation of the core transmembrane domain. cAMP binding relieves this inhibition. Differences in activation gating and extent of cAMP modulation between the HCN1 and HCN2 isoforms result largely from differences in the efficacy of CNBD inhibition.

Molecular and functional heterogeneity of hyperpolarization-activated pacemaker channels in the mouse CNS.

The hyperpolarization-activated cation current (termed I(h), I(q), or I(f)) was recently shown to be encoded by a new family of genes, named HCN for hyperpolarization-activated cyclic nucleotide-sensitive cation nonselective. When expressed in heterologous cells, each HCN isoform generates channels with distinct activation kinetics, mirroring the range of biophysical properties of native I(h) currents recorded in different classes of neurons. To determine whether the functional diversity of I(h) currents is attributable to different patterns of HCN gene expression, we determined the mRNA distribution across different regions of the mouse CNS of the three mouse HCN genes that are prominently expressed there (mHCN1, 2 and 4). We observe distinct patterns of distribution for each of the three genes. Whereas mHCN2 shows a widespread expression throughout the CNS, the expression of mHCN1 and mHCN4 is more limited, and generally complementary. mHCN1 is primarily expressed within neurons of the neocortex, hippocampus, and cerebellar cortex, but also in selected nuclei of the brainstem. mHCN4 is most highly expressed within neurons of the medial habenula, thalamus, and olfactory bulb, but also in distinct neuronal populations of the basal ganglia. Based on a comparison of mRNA expression with an electrophysiological characterization of native I(h) currents in hippocampal and thalamic neurons, our data support the idea that the functional heterogeneity of I(h) channels is attributable, in part, to differential isoform expression. Moreover, in some neurons, specific functional roles can be proposed for I(h) channels with defined subunit composition.

The HCN gene family: molecular basis of the hyperpolarization-activated pacemaker channels.

The molecular basis of the hyperpolarization-activated cation channels that underlie the anomalous rectifying current variously termed Ih, Iq, or I(f) is discussed. On the basis of the expression patterns and biophysical properties of the newly cloned HCN ion channels, an initial attempt at defining the identity and subunit composition of channels underlying native Ih is undertaken. By comparing the sequences of HCN channels to other members of the K channel superfamily, we discuss how channel opening may be coupled to membrane hyperpolarization and to direct binding of cyclic nucleotide. Finally, we consider some of the questions in cardiovascular physiology and neurobiology that can be addressed as a result of the demonstration that Ih is encoded by the HCN gene family.

Constraining ligand-binding site stoichiometry suggests that a cyclic nucleotide-gated channel is composed of two functional dimers.

Cyclic nucleotide-gated ion channels are composed of four pore-forming subunits. Binding of cyclic nucleotide to a site in the intracellular carboxyl terminus of each subunit leads to channel activation. Since there are four subunits, four binding events are possible. In this study, we investigate the effects of individual binding events on activation by studying channels containing one, two, three, or four functional binding sites. The binding of a single ligand significantly increases opening, although four ligands are required for full activation. The data are inconsistent with models in which the four subunits activate in a single concerted step (Monod-Wyman-Changeux model) or in four independent steps (Hodgkin-Huxley model). Instead, the four subunits may associate and activate as two independent dimers.

Identification of a gene encoding a hyperpolarization-activated pacemaker channel of brain.

The generation of pacemaker activity in heart and brain is mediated by hyperpolarization-activated cation channels that are directly regulated by cyclic nucleotides. We previously cloned a novel member of the voltage-gated K channel family from mouse brain (mBCNG-1) that contained a carboxy-terminal cyclic nucleotide-binding domain (Santoro et al., 1997) and hence proposed it to be a candidate gene for pacemaker channels. Heterologous expression of mBCNG-1 demonstrates that it does indeed code for a channel with properties indistinguishable from pacemaker channels in brain and similar to those in heart. Three additional mouse genes and two human genes closely related to mBCNG-1 display unique patterns of mRNA expression in different tissues, including brain and heart, demonstrating that these channels constitute a widely expressed gene family.

A state-independent interaction between ligand and a conserved arginine residue in cyclic nucleotide-gated channels reveals a functional polarity of the cyclic nucleotide binding site.

Activation of cyclic nucleotide-gated channels is thought to involve two distinct steps: a recognition event in which a ligand binds to the channel and a conformational change that both opens the channel and increases the affinity of the channel for an agonist. Sequence similarity with the cyclic nucleotide-binding sites of cAMP- and cGMP-dependent protein kinases and the bacterial catabolite activating protein (CAP) suggests that the channel ligand binding site consists of a beta-roll and three alpha-helices. Recent evidence has demonstrated that the third (or C) alpha-helix moves relative to the agonist upon channel activation, forming additional favorable contacts with the purine ring. Here we ask if channel activation also involves structural changes in the beta-roll by investigating the contribution of a conserved arginine residue that, in CAP and the kinases, forms an important ionic interaction with the cyclized phosphate of the bound ligand. Mutations that conserve, neutralize, or reverse the charge on this arginine decreased the apparent affinity for ligand over four orders of magnitude but had little effect on the ability of bound ligand to open the channel. These data indicate that the cyclized phosphate of the nucleotide approaches to within 2-4 A of the arginine, forming a favorable ionic bond that is largely unaltered upon activation. Thus, the binding site appears to be polarized into two distinct structural and functional domains: the beta-roll stabilizes the ligand in a state-independent manner, whereas the C-helix selectively stabilizes the ligand in the open state of the channel. It is likely that these distinct contributions of the nucleotide/C-helix and nucleotide/beta-roll interactions may also be a general feature of the mechanism of activation of other cyclic nucleotide-binding proteins.

Allosteric activation and tuning of ligand efficacy in cyclic-nucleotide-gated channels.

Despite recent advances in the identification of ligand-binding and voltage-sensing regions of ion channels, the domains that couple such regions to channel opening have not been identified. Moreover, it is uncertain whether ligand binding or depolarization are obligatory steps that must precede channel opening (according to linear reaction schemes) or whether they act to stabilize the channel in an open state that can exist independently of ligand binding or depolarization (according to cyclic allosteric models). By comparing ligand-independent and ligand-dependent channel openings, we now show that retinal and olfactory cyclic-nucleotide-gated channels are activated by a cyclic allosteric mechanism. We further show that an amino-terminal domain, distinct from the pore and ligand-binding motifs, participates in the allosteric gating transition, accounting for differences in the free energy of gating of the two channels. The allosteric transition provides an important mechanism for tuning the physiological response of ligand-binding proteins, such as cyclic-nucleotide-gated channels, to different biological signals.

Evidence for the induction of repetitive action potentials in synaptosomes by K+-channel inhibitors: an analysis of plasma membrane ion fluxes.

The effects of four K+-channel inhibitors on synaptosomal free Ca2+ concentrations and 86Rb+ fluxes are analysed. 4-Aminopyridine, alpha-dendrotoxin, charybdotoxin, and tetraethylammonium all increase the free Ca2+ concentration, although their potencies differ widely. In each case, the elevation in free Ca2+ concentration is reversed by the subsequent addition of tetrodotoxin. The transient 86Rb+ efflux from preequilibrated synaptosomes induced with high concentrations of veratridine is partially inhibited by 4-aminopyridine and alpha-dendrotoxin. In contrast, when 4-aminopyridine or alpha-dendrotoxin is added to polarized synaptosomes, and enhanced 86Rb+ flux is seen, both for uptake and for efflux with no change in the total 86Rb+/K+ content of the synaptosomes and with only a slight time-averaged plasma membrane depolarization (6.4 and 3.3 mV, respectively). The enhancements of flux by 4-aminopyridine or alpha-dendrotoxin are sensitive to ouabain and/or to tetrodotoxin. Furthermore, these flux changes show the same concentration dependencies as the blocked component of veratridine-stimulated 86Rb+ efflux, the elevation of free Ca2+ concentration, and the facilitation of glutamate exocytosis that are elicited by 4-aminopyridine or alpha-dendrotoxin. It is concluded that these findings support the proposal of spontaneous, repetitive firing of synaptosomes evoked by K+-channel inhibitors and that the enhanced 86Rb+ flux is a consequence of the activity of 4-aminopyridine- and alpha-dendrotoxin-insensitive K+ channels during these action potentials.

Subunit stoichiometry of cyclic nucleotide-gated channels and effects of subunit order on channel function.

Cyclic nucleotide-gated (CNG) ion channels are multimeric structures containing at least two subunits. However, the total number of subunits per functional channel is unknown. To determine the subunit stoichiometry of CNG ion channels, we have coexpressed the 30 pS conductance bovine retinal channel (RET) with an 85 pS conductance chimeric retinal channel containing the catfish olfactory channel P region (RO133). When RO133 and RET monomers are coexpressed, channels with four distinct intermediate conductances are observed. Dimer constructs reveal that two of these conductance levels arise from channels with the same subunit composition (2 RO133:2 RET) but distinct subunit order (like subunits adjacent to each other versus like subunits across from each other). Thus, the data demonstrate that cyclic nucleotide-gated ion channels are tetrameric like the related voltage-gated potassium ion channels; the order of subunits affects the conductance of the channel; and the channel has 4-fold symmetry in which four asymmetric subunits assemble head to tail around a central axis.

Antagonists of cyclic nucleotide-gated channels and molecular mapping of their site of action.

Activation of photoreceptor and olfactory cyclic nucleotide-gated (CNG) channels involves distinct ligand-binding and channel-gating reactions. To dissociate binding from gating, we identified the first competitive antagonists of CNG channels: specific phosphorothioate derivatives of cAMP and cGMP. We also identified membrane-permeant forms of these molecules that are antagonists and that will be useful for elucidating physiological roles for CNG channels in intact cells. The photoreceptor and olfactory CNG channels determine which of the phosphorothioate derivatives are agonists and which are antagonists based on different structural features of the ligand. The photoreceptor channel uses the nature of the purine ring (adenine vs guanine), whereas the olfactory channel uses the isomeric position of the thiophosphate S atom (Rp vs Sp). Interestingly, the same ligand, Rp-cGMPS, has opposite effects on the two channels, activating the photoreceptor channel and antagonizing the olfactory channel. Because Rp-cGMPS binds to both channels but activates only one, the channels must differ in a protein region that couples binding to gating. Chimeric photoreceptor and olfactory CNG channels reveal that the cytoplasmic C-terminal domain determines whether bound ligand activates the channel successfully. Hence, the C terminus contains not only the cyclic nucleotide-binding site, but also a region that couples ligand binding to channel gating.

Molecular mechanism of cyclic-nucleotide-gated channel activation.

Studies on the activation of ligand- and voltage-gated ion channels have identified regions involved in both ligand binding and voltage sensing, but relatively little is known about how such domains are coupled to channel opening. Here we investigate the structural basis for the activation of cyclic-nucleotide-gated channels, which are directly opened by cytoplasmic cyclic nucleotides but are structurally homologous to voltage-gated channels. By constructing chimaeras between cyclic-nucleotide-gated channels cloned from bovine retinal photoreceptors and catfish olfactory neurons, we identify two distinct domains that are important for ligand binding and channel gating. A putative alpha-helix in the carboxy-terminal binding domain determines the selectivity of the channel for activation by cGMP relative to cAMP. A domain in the amino-terminal region determines the ease with which channels open and thus influences agonist efficacy. We propose that channel opening is coupled to an allosteric conformational change in the binding site which enhances agonist binding. Thus, cyclic nucleotides activate the channel by binding tightly to the open state and stabilizing it.

A simple method for recording single-channel activity from synaptic plasma membranes.

Due to the small size of most nerve terminals, the ion channels which underlie presynaptic currents are usually inaccessible to investigation by conventional electrophysiological techniques. Here we describe a simple method for obtaining single-channel recordings from synaptic plasma membranes that does not require exposure of the native membranes to exogenous lipids or fusogens. To illustrate the method, we have recorded single-channel activity from rat cerebrocortical synaptosomal membranes. Under conditions designed to isolate calcium-independent currents, we describe three channel types that are most commonly observed.

Role of H5 domain in determining pore diameter and ion permeation through cyclic nucleotide-gated channels.

Ion permeation through membrane channels is thought to be governed by a narrow region of the channel pore termed the selectivity filter, which has been proposed to discriminate among ions by both specific binding and molecular sieving, as determined by pore diameter. Recent evidence suggests that a conserved domain (known as H5, P or SS1-SS2) in voltage-gated potassium, sodium and calcium channels contributes to the lining of the pore. Here we investigate whether the H5 domain determines pore diameter and examine the role of pore diameter in controlling ion permeation. These studies rely on differences in single channel conductance, ion selectivity and apparent pore diameter between cyclic nucleotide-gated channels cloned from bovine retina and catfish olfactory neurons. Using chimaeric retinal-olfactory channels, we find that the H5 domain determines these differences in permeation properties, providing structural evidence that the cyclic nucleotide-gated channels are indeed members of the voltage-gated channel family. Moreover, these results show directly that the H5 domain helps form the selectivity filter and that molecular sieving is important in controlling ion permeation.

Homologues of a K(+) channel blocker ?-dendrotoxin: characterization of synaptosomal binding sites and their coupling to elevation of cytosolic free calcium concentration.

Three polypeptides (?, ? and ?) homologous to ?-dendrotoxin, an inhibitor of certain voltage-activated K(+) channels, were found to elevate the cytosolic free concentration of calcium ([Ca(2+)](c)) in isolated central nerve terminals. Relative to ?-dendrotoxin (EC(50) ? 2.1 nM), the ?-, ?- and ?-toxins were 790-, 214- and 5.7-fold less effective; no additivity was apparent in the toxins' effects on [Ca(2+)](c). Each toxin antagonized the high affinity binding of (125)I-labelled ?- and ?-dendrotoxin to synaptosomes but with different potencies. The mutual interaction of ?- and ?-dendrotoxin with the acceptor appeared complex, the inhibition curves being noticeably extended. For the inhibition of binding of ?- and ?-dendrotoxin, the respective K(i)'s (nM) observed for ?, ?, ? and ? toxins were 0.78, 50, 99, 8, and 2.3, 95, 61, 0.53. Apparently, interactions of ?- and ?-, but not ?- or ?-dendrotoxin with the acceptor are closely coupled to an inhibition of K(+) channel(s) as observed indirectly by elevation of [Ca(2+)](c).

Repetitive action potentials in isolated nerve terminals in the presence of 4-aminopyridine: effects on cytosolic free Ca2+ and glutamate release.

The mechanisms by which an elevated KCl level and the K+-channel inhibitor 4-aminopyridine induce release of transmitter glutamate from guinea-pig cerebral cortical synaptosomes are contrasted. KCl at 30 mM caused an initial spike in the cytosolic free Ca2+ concentration ([Ca2+]c), followed by a partial recovery to a plateau 112 +/- 13 nM above the polarized control. The Ca2+-dependent release of endogenous glutamate, determined by continuous fluorimetry, was largely complete by 3 min, by which time 1.70 +/- 0.35 nmol/mg was released. [Ca2+]c elevation and glutamate release were both insensitive to tetrodotoxin. KCl-induced elevation in [Ca2+]c could be observed in both low-Na+ medium and in the presence of low concentrations of veratridine. 4-Aminopyridine at 1 mM increased [Ca2+]c by 143 +/- 18 nM to a plateau similar to that following 30 mM KCl. The initial rate of increase in [Ca2+]c following 4-aminopyridine administration was slower than that following 30 mM KCl, and a transient spike was less apparent. Consistent with this, the 4-aminopyridine-induced net uptake of 45Ca2+ is much lower than that following an elevated KCl level. 4-Aminopyridine induced the Ca2+-dependent release of glutamate, although with somewhat slower kinetics than that for KCl. The measured release was 0.81 nmol of glutamate/mg in the first 3 min of 4-aminopyridine action. In contrast to KCl, glutamate release and the increase in [Ca2+]c with 4-aminopyridine were almost entirely blocked by tetrodotoxin, a result indicating repetitive firing of Na+ channels. Basal [Ca2+]c and glutamate release from polarized synaptosomes were also significantly lowered by tetrodotoxin.(ABSTRACT TRUNCATED AT 250 WORDS)

Dendrotoxin and charybdotoxin increase the cytosolic concentration of free Ca2+ in cerebrocortical synaptosomes: an effect not shared by apamin.

Nanomolar concentrations of charybdotoxin or dendrotoxin increase the cytoplasmic free Ca2+ concentration in isolated central nerve terminals. The effects of the two toxins, normally considered to be blockers of K+ channels controlled by voltage in a Ca2+-sensitive or -insensitive manner, respectively, show only marginal additivity. Apamin, and inhibitor of low conductance Ca2+-activated K+ channels, was without effect in either the absence or presence of dendrotoxin. The effect of charybdotoxin on polarized, isolated central nerve terminals seems to be mediated largely by a block of K+ channels sensitive to dendrotoxin. Apparently, these voltage-operated K+ channels make a more significant contribution to maintaining the polarized potential of synaptosomes than do those activated by Ca2+.

Dendrotoxin, 4-aminopyridine, and beta-bungarotoxin act at common loci but by two distinct mechanisms to induce Ca2+-dependent release of glutamate from guinea-pig cerebrocortical synaptosomes.

The release of endogenous glutamate from guinea-pig cerebrocortical synaptosomes evoked by dendrotoxin, beta-bungarotoxin, and 4-aminopyridine is compared. Dendrotoxin and 4-aminopyridine cause Ca2+-dependent release, representing a partial depletion of the KCl-releasable transmitter pool. The decrease in the plasma membrane potential caused by 4-aminopyridine or dendrotoxin and the evoked release of glutamate from a transmitter pool accord with the inhibitory action of these agents on certain K+ conductances. In contrast, the massive release of glutamate evoked by beta-bungarotoxin is produced in the presence of Ca2+ but not of Sr2+, a result consistent with a generalised permeabilisation of synaptosomal plasma membranes. Although dendrotoxin inhibits the binding of beta-bungarotoxin and the resultant synaptosomal lysis, demonstration of a direct effect of beta-bungarotoxin binding per se on K+ permeability is impractical owing to its phospholipase A2 activity.

The discovery of a rapidly metabolized polymeric tetraphosphate derivative of adenosine in perfused rat heart.

The predicted presence in perfused rat hearts of a rapidly metabolized but hitherto unrecognized form of adenosine phosphate has been confirmed by specific radioactive labelling. The properties of the purified compound suggest that it is a heteropolymer of a small organic acid, phosphate and purine nucleoside in the proportions 1:4:1.