Molecular mechanisms of cyclic nucleotide-gated channels

Cyclic nucleotide-gated (CNG) channels are highly specialized to carry out their unique role in cell signalling. Significant progress has been made in the last several years determining the molecular mechanisms for these specializations. The activation of the channels begins with the binding of cyclic nucleotide to a domain in the carboxyl terminal region. This binding, in turn, produces an induced fit of the protein that involves a movement of the C-helix portion of the binding domain. The induced fit of the binding domain is coupled to an allosteric conformational change that opens the channel pore. The pore is formed primarily from the sequence between the S5 and S6 segments. A single glutamic acid in the pore represents the binding site for multiple monovalent cations, the blocking site for external divalent cations, and the site for the effect of protons on permeation.     

Ca2+ permeation in cyclic nucleotide-gated channels

Cyclic nucleotide-gated (CNG) channels conduct Na+, K+ and Ca2+ currents under the control of cGMP and cAMP. Activation of CNG channels leads to depolarization of the membrane voltage and to a concomitant increase of the cytosolic Ca2+ concentration. Several polypeptides were identified that constitute principal and modulatory subunits of CNG channels in both neurons and non-excitable cells, co-assembling to form a variety of heteromeric proteins with distinct biophysical properties. Since the contribution of each channel type to Ca2+ signaling depends on its specific Ca2+ conductance, it is necessary to analyze Ca2+ permeation for each individual channel type. We have analyzed Ca2+ permeation in all principal subunits of vertebrates and for a principal subunit from Drosophila melanogaster. We measured the fractional Ca2+ current over the physiological range of Ca2+ concentrations and found that Ca2+ permeation is determined by subunit composition and modulated by membrane voltage and extracellular pH. Ca2+ permeation is controlled by the Ca2+-binding affinity of the intrapore cation-binding site, which varies profoundly between members of the CNG channel family, and gives rise to a surprising diversity in the ability to generate Ca2+ signals.     

Stoichiometry and arrangement of heteromeric olfactory cyclic nucleotide-gated ion channels

Native cylic nucleotide-gated (CNG) channels are composed of alpha and beta subunits. Olfactory CNG channels were expressed from rat cDNA clones in Xenopus oocytes and studied in inside-out patches. Using tandem dimers composed of linked subunits, we investigated the stoichiometry and arrangement of the alpha and beta subunits. Dimers contained three subunit types: alphawt, betawt, and alpham. The alpham subunit lacks an amino-terminal domain that greatly influences gating, decreasing the apparent affinity of the channel for ligand by 9-fold, making it a reporter for inclusion in the tetramer. Homomeric channels from injection of alphawtalphawt dimers and from alphawt monomers were indistinguishable. Channels from injection of alphawtalpham dimers had apparent affinities 3-fold lower than alphawt homomultimers, suggesting a channel with two alphawt and two alpham subunits. Channels from coinjection of alphawtalphawt and betabeta dimers were indistinguishable from those composed of alpha and beta monomers and shared all of the characteristics of the alpha+beta phenotype of heteromeric channels. Coinjection of alphawtalpham and beta beta dimers yielded channels also of the alpha+beta phenotype but with an apparent affinity 3-fold lower, indicating the presence of alpham in the tetramer and that alpha+beta channels have adjacent alpha-subunits. To distinguish between an alpha-alpha-alpha-beta and an alpha-alpha-beta-beta arrangement, we compared apparent affinities for channels from coinjection of alphawtalphawt and betaalphawt or alphawtalphawt and betaalpham dimers. These channels were indistinguishable. To further argue against an alpha-alpha-alpha-beta arrangement, we quantitatively compared dose-response data for channels from coinjection of alphawtalpham and beta beta dimers to those from alpha and beta monomers. Taken together, our results are most consistent with an alpha-alpha-beta-beta arrangement for the heteromeric olfactory CNG channel.     

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.     

Modulation of rod photoreceptor cyclic nucleotide-gated channels by tyrosine phosphorylation

Cyclic nucleotide-gated (CNG) channels in vertebrate photoreceptors are crucial for transducing light-induced changes in cGMP concentration into electrical signals. In this study, we show that both native and exogenously expressed CNG channels from rods are modulated by tyrosine phosphorylation. The cGMP sensitivity of CNG channels, composed of rod alpha-subunits expressed in Xenopus oocytes, gradually increases after excision of inside-out patches from the oocyte membrane. This increase in sensitivity is inhibited by a protein tyrosine phosphatase (PTP) inhibitor and is unaffected by three different Ser/Thr phosphatase inhibitors. Moreover, it is suppressed or reversed by application of ATP but not by a nonhydrolyzable ATP analog. Application of protein tyrosine kinase (PTK) inhibitors causes an increase in cGMP sensitivity, but only in the presence of ATP. Taken together, these results suggest that CNG channels expressed in oocytes are associated with active PTK(s) and PTP(s) that regulate their cGMP sensitivity by changing phosphorylation state. The cGMP sensitivity of native CNG channels from salamander rod outer segments also increases and decreases after incubation with inhibitors of PTP(s) and PTK(s), respectively. These results suggest that rod CNG channels are modulated by tyrosine phosphorylation, which may function as a novel mechanism for regulating the sensitivity of rods to light.     

Subunit interactions in the activation of cyclic nucleotide-gated ion channels

Cyclic nucleotide-gated (CNG) ion channels of retinal photoreceptors and olfactory neurons are multimeric proteins of unknown stoichiometry. To investigate the subunit interactions that occur during CNG channel activation, we have used tandem cDNA constructs of the rod CNG channel to generate heteromultimeric channels composed of wild-type and mutant subunits. We introduced point mutations that affect channel activation: 1) D604M, which alters the relative ability of agonists to promote the allosteric conformational change(s) associated with channel opening, and 2) T560A, which primarily affects the initial binding affinity for cGMP, and to a lesser extent, the allosteric transition. At saturating concentrations of agonist, heteromultimeric channels were intermediate between wild-type and mutant homomultimers in agonist efficacy and apparent affinity for cGMP, cIMP, and cAMP, consistent with a model for the allosteric transition involving a concerted conformational change in all of the channel subunits. Results were also consistent with a model involving independent transitions in two or three, but not one or four, of the channel subunits. The behavior of the heterodimers implies that the channel stoichiometry is some multiple of 2 and is consistent with a tetrameric quaternary structure for the functional channel complex. Steady-state dose-response relations for homomultimeric and heteromultimeric channels were well fit by a Monod, Wyman, and Changeux model with a concerted allosteric opening transition stabilized by binding of agonist.     

Tetracaine reports a conformational change in the pore of cyclic nucleotide-gated channels

Local anesthetics are a diverse group of clinically useful compounds that act as pore blockers of both voltage- and cyclic nucleotide-gated (CNG) ion channels. We used the local anesthetic tetracaine to probe the nature of the conformational change that occurs in the pore of CNG channels during the opening allosteric transition. When applied to the intracellular side of wild-type rod CNG channels expressed in Xenopus oocytes from the alpha subunit, the local anesthetic tetracaine exhibits state-dependent block, binding with much higher affinity to closed states than to open states. Here we show that neutralization of a glutamic acid in the conserved P region (E363G) eliminated this state dependence of tetracaine block. Tetracaine blocked E363G channels with the same effectiveness at high concentrations of cGMP, when the channel spent more time open, and at low concentrations of cGMP, when the channel spent more time closed. In addition, Ni2+, which promotes the opening allosteric transition, decreased the effectiveness of tetracaine block of wild-type but not E363G channels. Similar results were obtained in a chimeric CNG channel that exhibits a more favorable opening allosteric transition. These results suggest that E363 is accessible to internal tetracaine in the closed but not the open configuration of the pore and that the conformational change that accompanies channel opening includes a change in the conformation or accessibility of E363.     

Movement of gating machinery during the activation of rod cyclic nucleotide-gated channels

In the visual and olfactory systems, cyclic nucleotide-gated (CNG) ion channels convert stimulus-induced changes in the internal concentrations of cGMP and cAMP into changes in membrane potential. Although it is known that significant activation of these channels requires the binding of three or more molecules of ligand, the detailed molecular mechanism remains obscure. We have probed the structural changes that occur during channel activation by using sulfhydryl-reactive methanethiosulfonate (MTS) reagents and N-ethylmaleimide (NEM). When expressed in Xenopus oocytes, the alpha-subunit of the bovine retinal channel forms homomultimeric channels that are activated by cGMP with a K1/2 of approximately 100 microM. Cyclic AMP, on the other hand, is a very poor activator; a saturating concentration elicits only 1% of the maximum current produced by cGMP. Treatment of excised patches with MTS-ethyltrimethylamine (MTSET) or NEM dramatically potentiated the channel's response to both cyclic nucleotides. After MTSET treatment, the dose-response relation for cGMP was shifted by over two orders of magnitude to lower concentrations. The effect on channel activation by cAMP was even more striking. After modification, the channels were fully activated by cAMP with a K1/2 of approximately 60 microM. This potentiation was abolished by conversion of Cys481 to a nonreactive alanine residue. Potentiation occurred more rapidly in the presence of saturating cGMP, indicating that this region of the channel is more accessible when the channel is open. Cys481 is located in a linker region between the transmembrane and cGMP-binding domains of the channel. These results suggest that this region of the channel undergoes significant movement during the activation process and is critical for coupling ligand binding to pore opening. Potentiation, however, is not mediated by the recently reported interaction between the amino- and carboxy-terminal regions of the alpha-subunit. Deletion of the entire amino-terminal domain had little effect on potentiation by MTSET.     

Perfusion system components release agents that distort functional properties of rod cyclic nucleotide-gated ion channels

In switching from studying native cyclic nucleotide-gated (CNG) ion channels in rod cells to studying the corresponding cloned channels expressed in Xenopus oocytes, we changed our perfusion system to a more efficient one. This change involved replacing culture flasks and a small plexiglass/glass chamber with plastic syringes, metal needles, and plastic petri dishes. We now report that these new perfusion system components release agents that distort or obscure measured functional properties of rod CNG channels. The magnitude and time course of appearance of the artifacts vary widely among individual components (e.g. from syringe to syringe). The effects most resemble voltage-dependent block of the channels, giving a decrease in current at positive potentials, and producing distortions of the kinetics and voltage dependence of channel activation.     

Three amino acids in the C-linker are major determinants of gating in cyclic nucleotide-gated channels

The activation of cyclic nucleotide-gated (CNG) channels is a complex process comprising the initial ligand binding and a consecutive allosteric transition from a closed to an open configuration. The cone and olfactory CNG channels differ considerably in cyclic nucleotide affinity and efficacy. In each channel, the cyclic nucleotide-binding site is connected to the last transmembrane segment of the channel by a linker peptide (C-linker) of approximately 90 amino acids. Here we report that replacement of three amino acids in the cone C-linker by the corresponding amino acids of the olfactory channel (I439V, D481A and D494S) profoundly enhanced the cAMP efficacy and increased the affinities for cAMP and cGMP. Unlike the wild-type cone channel, the mutated channel exhibited similar single-channel kinetics for both cGMP and cAMP, explaining the increase in cAMP efficacy. We thus conclude that the identified amino acids are major determinants of channel gating.     

Synthesis and use of a biotinylated 3-azidophenothiazine to photolabel both amino- and carboxyl-terminal sites in calmodulin

The biotinylated probe, 3-azido-10-(4-(4-biotinyl-1-piperazinyl)butyl)phenothiazine, was used to examine the phenothiazine binding domains in calmodulin (CaM) by photolabeling. This phenothiazine, synthesized from 3-azido-10-(4-(1-piperazinyl)butyl)phenothiazine and d-biotinyl tosylate, inhibited the CaM-mediated activation of phosphodiesterase (PDE) with an I50 of 12.5 (+/- 2.8) microM. Photolabeling of CaM produced covalent adducts in excellent yield (32%) in a light- and Ca2+-dependent manner. Studies performed over a range of drug concentrations suggested a 2:1 stoichiometry for the binding of the phenothiazine probe to CaM. Limited trypsin digestion and purification of the resulting fragments by either SDS-PAGE or HPLC provided two principal phenothiazine-containing peptides. Amino acid composition and sequence analyses performed on these two peptides established that both the N- and C-terminal domains in CaM, particularly the regions amino terminal to Ca2+-binding loops 1 and 3, were modified by the photoactivated phenothiazine derivative. These data, particularly for the C-terminal domain, are in excellent agreement with the X-ray structure analysis of a 1:1 CaM-trifluoperazine complex.     

Three residues predicted by molecular modeling to interact with the purine moiety alter ligand binding and channel gating in cyclic nucleotide-gated channels

Cytoplasmic cAMP and cGMP are soluble cellular messengers that directly activate cyclic nucleotide-gated (CNG) channels. These channels mediate sensory transduction in photoreceptors and olfactory neurons. The closely related CNG channels in these cell types have different nucleotide activation profiles, and we have investigated the molecular basis of their nucleotide selectivity properties. Previously, we predicted that the purine moiety of the nucleotide interacts with residues F533, K596, and D604 (bovine rod alpha CNG channel subunit sequences) of the nucleotide binding domain. In this study, we replaced these three residues with the corresponding residues of the bovine olfactory CNG channel. Mutations at each position altered the nucleotide activation of the rod CNG channels. In a mutant where K596 was replaced with arginine, cAMP-activated currents were enhanced 8-12-fold, suggesting that residue 596 influences channel gating. Thermodynamic cycle analysis of the data showed that (1) the residues are energetically coupled and (2) energetic coupling exists between the potentiating effects of Ni2+ and the replacement of F533 with tyrosine. These data suggest that changes in one of the residues alter the purine contacts with the other residues and that F533 communicates with the C-linker region of the channel involved in Ni2+ potentiation.     

Cloning and characterization of an olfactory cyclic nucleotide-gated channel expressed in mouse heart

Regulation of ionic currents in the heart is partly achieved by signaling cascades which alter intracellular levels of cyclic nucleotides. Changes in cyclic nucleotide levels can regulate channels either directly, like the direct binding of cAMP to the i(f) channel in pacemaker tissues, or indirectly through phosphorylation of channels by cAMP-dependent, or cGMP-dependent protein kinases. These types of regulation generally alter the voltage sensitivities of channels. A class of voltage-insensitive channels, first discovered in retinal rods and olfactory neurons, were recently identified in the heart. These channels are opened by the direct binding of cyclic nucleotides, providing a means of regulating ionic currents outside the influence of membrane voltage. Since different isoforms have different affinities for cAMP and cGMP, it is important to determine which isoforms are expressed in heart in order to predict their roles in heart function. We have cloned the olfactory channel from mouse heart, and find that although the message is very rare, Western blot analysis indicates the olfactory channel protein is stable in heart sarcolemma. Our data also suggest the olfactory channel protein forms homomeric channels in the heart since other isoforms or splice variants were not detected either by PCR amplification or by RNase protection. In addition, we have isolated and sequenced the mouse olfactory cyclic nucleotide-gated channel gene, and show the genomic organization is remarkably similar to that found in the human retinal channel gene. Part of this work was presented in abstract form.     

A site of interaction between pleckstrin's PH domains and G beta gamma

Pleckstrin is a 40 kDa substrate for protein kinase C found in platelets and neutrophils. Based upon its sequence, pleckstrin contains two of the recently-described PH domains that are thought to be binding motifs for phosphatidyl 4,5-bisphosphate (PIP2) and/or G protein beta gamma heterodimers (G beta gamma). In the present studies we have examined the interaction between pleckstrin and G beta gamma by incubating pleckstrin fusion proteins with lysates from human platelets. In this analysis, both the N-terminal and C-terminal PH domains from pleckstrin bound G beta gamma in vitro, as did peptides containing as little as the first 30 residues of the C-terminal pleckstrin PH domain. Introduction of a point mutation into this region, analogous to the mutation in the Btk PH domain that causes X-linked immunodeficiency disease (XID) in mice, dramatically disrupted this interaction. We propose that pleckstrin may interact with G beta gamma, and that one potential site for this interaction involves the first 30 residues of pleckstrin's C-terminal PH domain.     

Mapping of interaction domains between human repair proteins ERCC1 and XPF

ERCC1-XPF is a heterodimeric protein complexinvolved in nucleotide excision repair and recombinational processes. Like its homologous complex in Saccharomyces cerevisiae , Rad10-Rad1, it acts as a structure-specific DNA endonuclease, cleaving at duplex-single-stranded DNA junctions. In repair, ERCC1-XPF and Rad10-Rad1 make an incision on the the 5'-side of the lesion. No humans with a defect in the ERCC1 subunit of this protein complex have been identified and ERCC1-deficient mice suffer from severe developmental problems and signs of premature aging on top of a repair-deficient phenotype. Xeroderma pigmentosum group F patients carry mutations in the XPF subunit and generally show the clinical symptoms of mild DNA repair deficiency. All XP-F patients examined demonstrate reduced levels of XPF and ERCC1 protein, suggesting that proper complex formation is required for stability of the two proteins. To better understand the molecular and clinical consequences of mutations in the ERCC1-XPF complex, we decided to map the interaction domains between the two subunits. The XPF-binding domain comprises C-terminal residues 224-297 of ERCC1. Intriguingly, this domain resides outside the region of homology with its yeast Rad10 counterpart. The ERCC1-binding domain in XPF maps to C-terminal residues 814-905. ERCC1-XPF complex formation is established by a direct interaction between these two binding domains. A mutation from an XP-F patient that alters the ERCC1-binding domain in XPF indeed affects complex formation with ERCC1.     

Functional interaction between InsP3 receptors and store-operated Htrp3 channels

Calcium ions are released from intracellular stores in response to agonist-stimulated production of inositol 1,4,5-trisphosphate (InsP3), a second messenger generated at the cell membrane. Depletion of Ca2+ from internal stores triggers a capacitative influx of extracellular Ca2+ across the plasma membrane. The influx of Ca2+ can be recorded as store-operated channels (SOC) in the plasma membrane or as a current known as the Ca2+-release-activated current (I(crac)). A critical question in cell signalling is how SOC and I(crac) sense and respond to Ca2+-store depletion: in one model, a messenger molecule is generated that activates Ca2+ entry in response to store depletion; in an alternative model, InsP3 receptors in the stores are coupled to SOC and I(crac). The mammalian Htrp3 protein forms a well defined store-operated channel and so provides a suitable system for studying the effect of Ca2+-store depletion on SOC and I(crac). We show here that Htrp3 channels stably expressed in HEK293 cells are in a tight functional interaction with the InsP3 receptors. Htrp3 channels present in the same plasma membrane patch can be activated by Ca2+ mobilization in intact cells and by InsP3 in excised patches. This activation of Htrp3 by InsP3 is lost on extensive washing of excised patches but is restored by addition of native or recombinant InsP3-bound InsP3 receptors. Our results provide evidence for the coupling hypothesis, in which InsP3 receptors activated by InsP3 interact with SOC and regulate I(crac).     

Molecular dissection of radixin: distinct and interdependent functions of the amino- and carboxy-terminal domains

The ERM proteins--ezrin, radixin, and moesin--occur in particular cortical cytoskeletal structures. Several lines of evidence suggest that they interact with both cytoskeletal elements and plasma membrane components. Here we described the properties of full-length and truncated radixin polypeptides expressed in transfected cells. In stable transfectants, exogenous full-length radixin behaves much like endogenous ERM proteins, localizing to the same cortical structures. However, the presence of full-length radixin or its carboxy-terminal domain in cortical structures correlates with greatly diminished staining of endogenous moesin in those structures, suggesting that radixin and moesin compete for a limiting factor required for normal associations in the cell. The results also reveal distinct roles for the amino- and carboxy-terminal domains. At low levels relative to endogenous radixin, the carboxy-terminal polypeptide is associated with most of the correct cortical targets except cleavage furrows. In contrast, the amino-terminal polypeptide is diffusely localized throughout the cell. Low level expression of full-length radixin or either of the truncated polypeptides has no detectable effect on cell physiology. However, high level expression of the carboxy-terminal domain dramatically disrupts normal cytoskeletal structures and functions. At these high levels, the amino-terminal polypeptide does localize to cortical structures, but does not affect the cells. We conclude that the behavior of radixin in cells depends upon activities contributed by separate domains of the protein, but also requires modulating interactions between those domains.     

Molecular cloning of a putative cyclic nucleotide-gated ion channel cDNA from Limulus polyphemus

Cyclic nucleotide-gated channels have been proposed to mediate the electrical response to light in the ventral photoreceptor cells of the horseshoe crab, Limulus polyphemus. However, a cyclic nucleotide-gated channel has not been identified from Limulus. We have cloned a putative full-length cyclic nucleotide-gated channel cDNA by screening cDNA libraries constructed from Limulus brain using a probe developed from Limulus ventral eye nerves. The putative full-length cDNA was derived from two overlapping partial cDNA clones. The open reading frame encodes 905 amino acids; the sequence shows 44% identity to that of the alpha subunit of the bovine rod cyclic GMP-gated channel over the region containing the transmembrane domains and the cyclic nucleotide binding domain. This Limulus channel has a novel C-terminal region of approximately 200 amino acids, containing three putative Src homology domain 3 binding motifs and a putative coiled-coil domain. The possibility that this cloned channel is the same as that detected previously in excised patches from the photoreceptive membrane of Limulus ventral photoreceptors is discussed in terms of its sequence and its expression in the ventral eye nerves.