Lester / Davidson Lab 1999 Annual Report

The inward rectifying potassium channel Kir4.1 is thought to play a major role in establishing and maintaining potassium homeostasis in the brain by passive uptake of K⁺into the glia cells. In situ hybridization experiments in rats have shown that this channel is predominantly expressed in glial cells of the cerebellum, to a lesser extend in the cerebral cortex and hippocampus. One specific function of this channel may be to "siphon" K⁺ ions released by neurons toward the distal portion of the glial cell, and then to the extracellular space. Additionally, a possible role of this particular channel in glial proliferation and neurogenesis has been discussed. We have constructed a knock-out mouse lacking the Kir4.1 channel by gene targeting inactivation. A lambda genomic library derived from 129/SveTac mice was screened using a Kir 4.1 cDNA probe and several individual genomic clones were isolated. Restriction mapping and sequencing analysis indicated that the Kir4.1 polypeptide is encoded by a single exon. A genomic fragment of 5.8 kb was used to construct a replacement type targeting vector, followed by homologous recombination in ES cells and blastocyst injection. Heterozygous animals for the targeted gene were mated with each other to produce mice homozygous for the targeted gene Kir 4.1. There are no marked phenotypic differences between the heterozygous and wild type animals. Mice homozygous for the inactivated Kir4.1 display a strong phenotype, presenting with low growth rates, ataxia, general underdevelopment, and death by 3 weeks postnatal. In order to understand the physiological and pathophysiological role of this channel at the cellular level, we are using patch-clamp recording in the whole-cell configuration in acutely dissociated cells of the cerebellum and the hippocampus. The results will be compared to those for neurons and glia in culture. Furthermore a detailed histology of the CNS and other organs expressing the Kir4.1 channel will be performed, as well as behavioral studies. This knock-out mouse allows us to investigate the function of this particular channel for the development of the CNS and its role in the mature CNS, and might provide a model for neurodegeneration of the cerebellum.

*Department of Physiology, University of Minnesota, Minneapolis

ROMK1 (Kir1.1) silences central neurons in vitro and in vivo

Hendrickje Nadeau, Gabriela Greif, David Crotty, and David J. Anderson

Lentiviral vectors were constructed to express the weakly rectifying kidney K⁺ channel ROMK1, either fused to green fluorescent protein or as a bicistronic message (RomK1-IRES-GFP). The channel was stably expressed in rat hippocampal neurons. The level of functional expression, quantified as the inward current at –130 mV in 5.4 mM external K⁺ blockable by 500 mM Ba⁺², followed approximately a Gaussian distribution. Infected cells showed an average membrane hyperpolarization of 5 mV relative to uninfected cells and a halving of whole-cell input resistance; this led to an increased injected current threshold for spikes, without change in spike shape, size, or timing. The relation between injected current and spike frequency shifted linearly with the amount of ROMK1 expressed. Infected cells were maintained in culture for up to four weeks without evidence of viral toxicity or change in the levels of functional expression. The same viral vectors were also introduced into animals, leading to similar long-term stable expression ("localized transgenics"). Delivery of the channel to neural progenitor cells, in either embryonic or P1 mice , led to hypoexcitable phenotypes . Adult cells were also infected and investigated electrophysiologically.

Selective Silencing of Neurons in Mouse Brain: Strategies using Chloride Channels

Eric Slimko, David J. Anderson

Selectively reducing the excitability of glutamatergic neurons will (1) allow for the creation of animal models of human neurological disorders and (2) provide insight into possible treatments for conditions such as epilepsy, stroke, and excitotoxin poisoning. Because there are no pharmacological agents that target subsets of glutamatergic cells, we are focusing on genetic approaches. Our strategy is to express invertebrate ivermectin-sensitive chloride channels (GluCl a and b) under tissue-specific promoters and to activate them with the drug to produce the silencing effect through a hyperpolarizing chloride current and a decrease in cell resistance. Ivermectin is a well-known antihelminthic/insecticide used to treat a variety of parasitic diseases in both humans and animals. Although high concentrations of the drug in brain can be lethal, we believe we can activate the channels with low enough concentrations to avoid the toxic effect. Our goal is to construct in vitro models that mimic as much as possible conditions that can be obtained in transgenic mice and in human gene therapy. We have expressed the a subunit in HEK293 cells and recorded ivermectin activated chloride currents. Also, we have delivered the gene using microinjection into cultured rat sympathetic cervical ganglion neurons. We are performing patch-clamp experiments on the injected cultures, examining the effects of the inserted gene on the cell’s excitability. This method has shown intriguing evidence of reducing the excitability target cells. We are now working with both the a and b subunit of the chloride channel, which we believe will provide more robust silencing currents than the a subunit alone. Further goals involve the creation of retroviral delivery systems, modification of the chloride channels to increase their efficacy, slice work with slices expressing the genes, the creation of transgenic mice, and the infection of monkey brain with the GluCl containing virus.

Selective neuronal degeneration induced by transgene expression of mutant ion channels

Johannes Schwarz, Hong Dang, Jingshan Chen*, Eric J. Nestler*

The etiology of various neurodegenerative disorders is unknown. At present, most animal models involve acute chemical or mechanical lesions. Positive results in these models are rarely reproduced in clinical trails. Therefore there is a need for animal models with chronic neurodegeneration in selected brain regions. The weaver mutation of GIRK2 (wvGIRK2) induces depolarization of affected neurons which represents a common pathway in degenerating neurons.

We use the tetracycline inducible system to drive transgene expression of the weaver mutation in striatal, cerebellar neurons, or both. Two lines of mice were generated with expression of the tetracycline transactivator (tTA) restricted to these brain regions (Chen et al. 1998). In addition, we generated transgenic mouse lines carrying the mutant wvGIRK2 gene under control of a tetracycline inducible promotor. Mating of the tTA and wvGIRK2 mice shall target wvGIRK2 expression to selected neurons in striatum and/or cerebellum.

Future analyses will show whether ectopic expression of the weaver GIRK2 can induce selective neurodegeneration and, thus, serve as a tool to develop animal models for disorders of unknown etiology. In addition, conditioned gene expression may enable us to study the effect of such mutants on developing versus adult neurons.

* Department of Psychiatry and Neurobiology, Yale University, New Haven

Assessing the in vivo function of hyperpolarization activated cation channels using designer mutants

Johannes Schwarz, Baljit S. Khakh

Rhythms in the heart and brain depend on specialized cells that act as pacemakers. Defects in the cardiac pacemaker lead to cardiac arrhythmias and similar defects in the brain may underlie sleep disorders and possibly epilepsy. Hyperpolarization-activated cyclic nucleotide-gated (HCN) channels belong to the superfamily of voltage-gated K+ selective channels with 6 hydrophobic domains per subunit. However, they possess an unusual P segment: HCN channels contain the sequence CIGYG (residues 347-351) instead of the K+ channel signature sequence of TIGYG.

We have replaced the GYG triplet with AYG, GAG or AAA and expressed these mutant channels in HEK 293 cells: no hyperpolarization activated currents were detected for any of the mutants alone. When coexpressed with wild type HCN channels these mutants suppressed by 90 % the total hyperpolarization activated current. In the simplest interpretation the mutant channels act as dominant negatives in the heteromer formed between the mutants and wild type HCN channels. Surprisingly, the point mutant C347A showed faster kinetics of activation and deactivation: the mutant channel displays a gain-of-function phenotype. Possibly cysteine-cysteine interactions in wild type channels contribute to the classically slow kinetics of the HCN channels.

In future work we will test the utility of using dominant negative and gain-of-function mutants of HCN channels to modify endogenous hyperpolarization activated currents , e. g. pain sensing neurons, thalamic neurons. Our ultimate goal is to understand the role of neuronal HCN channels in vivo, and the transgenic expression of mutant HCN channels in mice may expedite this.

The role of CAG-repeats in hyperpolarization activated cyclic nucleotide gated (HCN) channels

Johannes Schwarz, Baljit S. Khakh

HCN channels are a recently identified class of mixed cation channels with unique electrophysiological properties. In the mouse mHCN1 is predominantly and abundantly expressed in the brain. This gene has a stretch of 37 CAG repeats in the C-terminal region. Repeats longer than 35 CAG triplets cause degenerative disorders in humans (e. g., Huntington’s disease, spinocerebellar ataxias). However, these CAG repeats are in the N-terminal region of the respective gene.

We designed two mutants of mHCN1 with an eliminated or reduced (13 repeats) CAG repeat. When expressed in HEK 293 cells, both mutants showed a reduction of hyperpolarization induced cation currents. Compared to the wild type channel the 13 repeat mutant exhibited 60 % and the 0 repeat mutant 10 % of wild type currents.

These data suggest that the CAG repeat in the C-terminal region of mHCN1 is necessary for channel function.

Future work will investigate the mechanism by which the elimination of the C-terminal CAG repeat inhibits proper channel function.

Knock-in Mice Carrying A Targeted Mutation in the 5HT3A Subunit.

Hong Dang, Purnima Despande, Zoufeng Chen, Nicholas Guy¹, David Julius¹

The 5HT3 receptor is a member of the LGIC superfamily whose subunits all contain 4 transmembrane domains. To further understand the role of this receptor in behaving animals, we used the targeted exon replacement technique to generate mouse lines containing a hyperactive 5HT3A mutation. The mutation V13’S in the M2 channel lining domain caused a ~ 70-fold decrease in serotonin EC₅₀ in the homomeric 5HT3A receptor expressed in Xenopus oocytes. The mutant also showed no desensitization in contrast to the wild type receptor. Furthermore, when the mutant subunit was coexpressed with the recently discovered 5HT3B subunit (Davies et al, 1999, Nature, 397: 359), there was even more marked hyperactivity. A genomic construct containing the mutation and a Neo-resistant selection cassette flanked by 2 LoxP sites in the nearby upstream intron was used for targeted exon replacement in mouse ES cells. Two clones of ES cells containing the correct construct were used to generate mouse lines. The Neo cassette in the intron was deleted from the ES cell genome by transient transfection with a plasmid containing the CRE recombinase gene. Two mouse lines, neoVS retaining the Neo cassette and VS with Neo deleted, were established. Preliminary results show that both hetero- and homo-zygotic mice of the neoVS construct appear grossly normal, but display reduced expression of the 5HT3 mRNA and responses. Both the hetero- and homo-zygotic mice from the VS lineage are also viable. These mice are being characterized in detail.

¹Dept. Cellular and Molecular Pharmacology, UCSF, San Francisco, CA 94143.

Generation of "knock-in" mice with mutations in the M2 region of the a4 neuronal nicotinic acetylcholine receptor gene

Cesar. Labarca¹, Purnima Deshpande¹, Paulo Kofuji², Baljit.S. Khakh¹, Mark. Nowak³, Zoufeng. Chen, James. Boulter⁴.

We are examining the physiological role of nicotinic acetylcholine receptors (nAChR) in the nervous system by introducing mutations into the corresponding mouse genes that significantly alter the function of the receptors. The Leu9’Ser mutation in the M2 transmembrane region of the nAChR increases the apparent sensitivity to ACh about 10-fold for every mutated subunit in the mouse muscle receptor. Naturally occurring mutations in the human muscle nAChR M2 domain produce myasthenic syndromes caused by excessive activation. We have generated mice with a Leu9’Ser mutation in the M2 region of the a4 subunit of the AChR. a4b2 receptors carrying this mutation have ACh sensitivity about 100-fold higher than wild type receptors when examined in the Xenopus oocyte system. The M2 transmembrane region was mapped to exon 5 of mouse a4 genomic DNA. A genomic fragment of 9.5 kb containing exon 5 was used to construct a targeting vector, with a neomycin cassette flanked by loxP sites for positive selection, and a diphtheria toxin cassette for negative selection of random insertion. The Leu9’Ser mutation was introduced by site directed mutagenesis. ES cell clones were screened for homologous recombination. Selected positive clones were transfected with a plasmid carrying Cre recombinase to excise the neomycin cassette and injected into blastocysts to generate chimeric mice carrying the mutation. Heterozygous mice generated from these chimeras have no obvious phenotypic differences with wild type animals; they are being mated with each other to generate mice homozygous for the mutation. Homozygous mice die at birth. Heterozygous and homozygous mice will be analyzed for changes in physiology, nicotine responses, and behavior.

²Neuroscience, U.of Minn. Med. Sch., Minneapolis MN 55455; ³Psychiatry, Med. U. of S. C., Charleston, SC 29425; ⁴Psychiatry & Biobehavioral Sci., UCLA, L. A., CA 90024

Backbone Mutations in Transmembrane Domains of a Ligand-Gated Ion Channel: Implications for the Mechanism of Gating

Pamela M. England, Yinong Zhang, Dennis A. Dougherty¹

An approach to identify backbone conformational changes underlying nicotinic acetylcholine receptor (nAChR) gating was developed. Specific backbone peptide bonds were replaced with an ester, which disrupts backbone hydrogen bonds at the site of mutation. At a conserved proline residue (aPro221) in the first transmembrane (M1) domain, the amide-to-ester mutation provides receptors with near-normal sensitivity, although the natural amino acids tested other than Pro produce receptors that gate with a much larger EC₅₀ than normal. Therefore a backbone hydrogen bond at this site may interfere with normal gating. In the aM2 domain, the amide-to-ester mutation yielded functional receptors at 15 positions, three of which provided receptors with >10-fold lower EC₅₀than wild type. These results support a model for gating that includes significant changes of backbone conformation within the M2 domain.

¹Division of Chemistry and Chemical Engineering

Probing the role of a conserved M1 proline residue in 5HT3 ligand-gated ion channel gating

Hong Dang, Pamela M. England, Dennis A. Dougherty

A conserved proline residue is found in the first transmembrane (M1) domain of every subunit in the ligand-gated ion channel superfamily. The position of this proline between the N-terminal extracellular agonist binding and the M2 channel lining domains in the primary sequence suggests its possible involvement in the gating of the receptor. Replacing this proline with alanine, glycine, or leucine in the 5HT3A homomeric receptors expressed in Xenopus oocytes resulted in the absence of 5-HT induced whole cell currents. Combining these proline mutations with an additional valine to serine mutation in the M2 channel domain produced constitutively open channels responding marginally to 5HT3 ligands. To determine what properties of the conserved proline are critical for the function of the channel, an a-hydroxy acid and two imino acids were incorporated at the proline position using the nonsense suppression method. Trans-3-methyl-proline, pipecolic acid, and leucic acid were able to replace the conserved proline to produce active channels with EC₅₀s similar to that for the wild type receptor. These trends are preserved in the heteromeric receptors consisting of 5HT3A and 5HT3B subunits in oocytes. The prominent common feature among these residues and proline is the lack of hydrogen bond donor activity, potentially resulting in a flexible secondary structure in the M1 region. Thus lack of hydrogen bond donor activity may be a key element in channel gating and may explain the high degree of conservation of this M1 proline.

Mapping Disulfide Connectivity Using Backbone Ester Hydrolysis

Pamela M. England, Dennis A. Dougherty¹

The site-specific incorporation of a-hydroxy acids into proteins using nonsense suppression can provide a powerful probe of protein structure and function. The resulting backbone ester may be selectively hydrolyzed in the presence of the peptide backbone, providing an "orthogonal" chemistry that can be useful both as an analytical tool and as a structural probe. Here we detail a substantial substituent effect on this hydrolysis reaction. Consistent with mechanistic expectations, the steric bulk of the amino acid immediately N-terminal of the hydroxy acid has a large effect on the hydrolysis rate. Based on these results, we also describe a simple protocol for identifying disulfide loops in soluble and membrane proteins, exemplified by the a subunit of the muscle nicotinic acetylcholine receptor (nAChR). If a backbone ester is incorporated outside a disulfide loop, hydrolysis alone gives two fragments; but if the ester is incorporated within a disulfide loop, both hydrolysis and reduction are required for cleavage. This test could be useful in characterizing the disulfide topology of complex, membrane proteins.

¹Division of Chemistry and Chemical Engineering

The Tethered Agonist Approach to Mapping Ion Channel Proteins:

Toward a Structural Model for the Agonist Binding Site of the Nicotinic Receptor

Lintong Li*, Wenge Zhong*, Dennis A. Dougherty*

Using the in vivo nonsense suppression method for unnatural amino acid incorporation, a series of tethered quaternary ammonium derivatives of tyrosine (Tyr-OnQ) has been incorporated at three residues in the nicotinic acetylcholine receptor. At three sites a constitutively active receptor results, but the pattern of activation as a function of chain length is different. At position a 149, there is a clear preference for a three carbon tether, while at position a 93 tethers of n = 2 - 5 carbons are comparably effective. At position g 55/d 57, all tethers except n = 2 are effective. A model for the receptor binding site can be developed by analogy with the acetylcholinesterase crystal structure.

*Division of Chemistry and Chemical Engineering

The role of the cation-p interaction in nicotine binding to the nAChR

Gabriel S. Brandt^†, Niki M. Zacharias^†, Caroline Gibbs², Dennis A. Dougherty^†

Previous work in this laboratory has established that acetylcholine binds Trp149 in the alpha subunit of the mouse muscle nicotinic acetylcholine receptor (nAChR). This binding is mediated largely by a cation-p interaction between the tryptophan residue and the quaternary ammonium center of ACh. Interestingly, other data suggested that nicotine binding differs from that of ACh, because there is no evidence of a cation-p interaction between nicotine and Trp149. A significant difference between the two agonists is that nicotine contains a tertiary ammonium center rather than the quaternary center of ACh. Furthermore, the pKa of this nitrogen is close to physiological pH. Here, we address whether the degree of amine substitution plays an important role in nicotine binding to Trp149.

The active form of nicotine is probably the protonated form. By alkylating the pyrrolidinium nitrogen, we generated a derivative of nicotine containing a permanently charged quaternary ammonium center. We then used this N-methylnicotinium derivative in our usual protocol for examining cation-p interactions. Briefly, this protocol involves introducing uncoded amino acids of varying degrees of p electron density into the site of interest. In the case of Trp149, the wild-type tryptophan was replaced with 5-CN-Trp and 5-Br-Trp. The cyano- derivative is much less electronegative over the indole ring than the bromo- derivative, even though the molecules are sterically very similar. A large difference in dose-response relation between the two mutants implies a significant electrostatic role in agonist binding. In the case of nicotine, for which there is very little effect suggestive of a cation-p interaction, the ratio between these two mutants is less than 3. For ACh, the same ratio is around 50.

Preliminary results suggest that the EC₅₀ of N-methylnicotinium for 5-CN-Trp is around 238 µM, while that of the 5-Br-Trp derivative is 7 µM. With a ratio of 34, N-methylnicotinium appears to be experiencing a cation-p interaction with Trp149. Studies are currently being carried out to understand the implications of this interaction for nAChR structure and function. The combination of unnatural amino acid mutagenesis and synthesis of agonist analogs is uniquely suited to elucidate the precise nature of the non-covalent intermolecular interactions that are responsible for almost all biological recognition events.

Reference:

1. Zhong, W.G.; Gallivan, J.P.; Zhang, Y.N.; Li, L.T.; Lester, H.A.; Dougherty, D.A. From ab initio quantum mechanics to molecular neurobiology: A cation-p binding site in the nicotinic receptor. PNAS, 95, 12088 (1998).

†Division of Chemistry and Chemical Engineering

²Caltech Undergraduate

Cross-linking studies on the nAChR expressed in Xenopus oocytes

Gabriel S. Brandt^†, John A. Wendel, Dennis A. Dougherty^†

Chemical cross-linking is often proposed as a method for investigating protein-protein or intersubunit interactions. However, cross-linking typically suffers from limited specificity. Unnatural amino acid mutagenesis allows for the introduction of uniquely reactive amino acid side chains at a specified position in a protein and could therefore be used for site-directed cross-linking. The in vivo unnatural amino acid mutagenesis technique developed previously¹ could be used to investigate protein-protein interactions in a biologically relevant context. In order to make this technique available for the study of ion channel assembly in Xenopus oocytes, we have generated nicotinic acetylcholine receptor (nAChR) subunits with a C-terminal His7 tag and the HA reporter epitope in the M3-M4 loop.

Following an established protocol for protein isolation and subsequent cross-linking, we have utilized Ni-NTA beads to isolate nAChR expressed in oocytes. Following elution of the protein from the beads, the intact receptor pentamers are eluted from the beads and treated with a variety of cross-linking reagents in the presence of 0.5% b-dodecylmaltoside, a mild detergent. The samples are subject to SDS-PAGE and then detected with Mab HA.11 in a Western blot. These blots clearly show the presence of higher-MW bands in the expected range of oligomers between alpha and other nAChR subunits. However, under certain conditions, we see these bands appearing in the absence of cross-linker. It remains unclear whether we are detecting covalent products of crosslinking in our assay. Studies employing stronger denaturing conditions are under way.

†Division of Chemistry and Chemical Engineering

Disulfide Crosslinking in the a and g subunits of the Nicotinic Acetylcholine Receptor

Niki M. Zacharias^*, Gabriel S. Brandt^*, Dennis A. Dougherty^*

To gain chemical-scale information on subunit-subunit contacts, nicotinic acetylcholine receptors with single cysteine mutations in each of the a or g subunits are expressed in Xenopus oocytes and treated with copper sulfate and 1,10- phenanthroline to catalyze the formation of a disulfide bond. Only residues that are close in space will be responsive to such treatment. The initial target is the M2 region of the subunits, which is believed to be the transmembrane spanning region that lines the pore of the channel. To date, nine mutations have been studied using copper phenanthroline treatments (a T244C, a L245C, a S246C, a S248C, a V249C, a L250C, a L251C, g T253C, g L260C). Electrophysiological experiments have revealed that current is inhibited in a T244C mutants by 42 + 3%, a L251C mutants by 27 + 9%, a T244C/g T253C mutants by 58 + 1%, and a L251C/g L260C mutants by 26 + 6% compared to wildtype after a four minute application of copper in the presence of acetylcholine. The inhibition of current in mutants indicates that formation of a disulfide bond could be occurring. Changes in molecular weight (dimer formation) are probed by the use of Western blots to verify the disulfide bond formation.

*Division of Chemistry and Chemical Engineering

Properties of channels formed by coexpression of nicotinic and P2X subunits

Baljit S. Khakh, Jason Sydes, and Cesar Labarca

In the nervous system P2X and nicotinic channels underlie purinergic and cholinergic transmission, respectively. Studies from neurons that express these channels indicate that they may not operate independently. We have used a null cell type to express and study the function of nicotinic and P2X channels. We find that acetylcholine (ACh) and ATP evoke robust a3b4 nicotinic and P2X2 channel mediated currents when both channel types are expressed separately, or in the same cell. If the a3b4 nicotinic and P2X2 channels acted independently then the simultaneous addition of ACH and ATP would evoke a current that is the linear sum of the two-individual currents. However, we find that the current evoked by co-application of ATP and ACh is always smaller than the expected sum of the two currents alone, suggesting an interaction between these two structurally distinct cation channels. The mechanistic basis of this is under further investigation.

Neuronal P2X transmitter-gated cation channels change their ion selectivity in seconds.

Baljit S. Khakh & Cesar Labarca

Fast synaptic transmission depends on the selective ionic permeability of transmitter-gated ion channels. We show changes in the ion selectivity of neuronal P2X transmitter-gated cation channels on the second time scale, and dependent on the history of exposure to ATP. The heterologously expressed neuronal P2X2, P2X2/P2X3, and P2X4 channels, as well as native P2X channels in neurons, display varying combinations of phenomena including biphasic responses, permeability to NMDG⁺, and permeability to a fluorescent dye. Furthermore, in P2X4 receptors this process can be both up- and down regulated by altering an amino acid residue thought to line the ion permeation pathway, localizing a region that governs the activity-dependent change.

Dynamic selectivity filters in ion channels.

Baljit S. Khakh

Membrane ion channels contain integral pores that precisely select their permeant ions. The selectivity of ion channel pores has generally been viewed as fixed. However, recent studies on disparate classes of ion channels challenge the generality of this idea, and show that some ion channels can change their ion selectivity such that normally impermeant ions do in fact permeate under some circumstances. A survey of published papers shows that ion selectivity changes occur for P2X, proton-gated channels and mutant NMDA receptors from the superfamily of transmitter-gated ion channels, Kv and cardiac sodium channels from the superfamily of voltage-gated ion channels, and cyclic nucleotide-gated channels from the family of channels that are gated by intracellular messengers. An assessment of these channels reveals that ion channels can change their selectivity, such that normally impermeant ions can also flow, or that normally permeant ions can flow with greater ease. These data imply a dynamic selectivity filter in these channels. In no case is the mechanism fully understood, but the phenomenon represents both a new functional aspect of ion channels and a suggestion about hitherto unrealized mechanisms by which channels may process information encoding in the nervous system.

Allosteric control of gating and kinetics at P2X₄ receptor-channels

Baljit S. Khakh, William R. Proctor¹ & Thomas V. Dunwiddie¹

The central nervous system abundantly expresses P2X receptor-channels for ATP; of these the most widespread in the brain is the P2X₄ channel. We show that ivermectin (IVM) is a specific positive allosteric effector of heterologously expressed P2X₄ and possibly of heteromeric P2X₄/P2X₆ channels, but not of P2X₂, P2X₃, P2X₂/P2X₃ or P2X₇ channels. In the submicromolar range (EC₅₀ ~ 250 nM) the action of IVM was rapid and reversible, resulting in increased amplitude and slowed deactivation of P2X₄ channel currents evoked by ATP. IVM also markedly increased the potency of ATP, and that of the normally low-potency agonist a ,b -methylene-ATP, in a use- and voltage-independent manner without changing the ion selectivity of P2X₄ channels. Therefore IVM evokes a potent pharmacological gain-of-function phenotype that is specific for P2X₄ channels. We also tested whether IVM could modulate endogenously expressed P2X channels in the adult trigeminal mesencephalic nucleus and hippocampal CA1 neurons. Surprisingly, IVM produced no significant effect on the fast ATP-evoked inward currents in either type of neuron, despite the fact that IVM modulated P2X₄ channels heterologously expressed in embryonic hippocampal neurons. These results suggest that homomeric P2X₄ channels are not the primary subtype of P2X receptor in the adult trigeminal mesencephalic nucleus and in hippocampal CA1 neurons.

¹Univ. Colo. Hlth. Sci. Ctr. & Denver V. A. Med. Ctr.

Assessment of P2X₂ channel function using genetically engineered optical probes

Baljit S. Khakh

One of the seven cloned P2X subunits, the P2X₂ subunit, is found abundantly in nervous system, suggesting an important role in purinergic transmission. A heteromer formed between P2X₂ and P2X₃ channels may also be the molecular target for pain caused by ATP released during inflammation/cell death. In order to gain a better understanding of the distribution (on the distance scale of ~1 mm in live cells) and function of P2X₂ channels, we are employing genetically engineered optical probes. Our goals are to track P2X₂ channels in the living cells and make a high resolution P2X₂ channel map of the nervous system. In order to achieve this we are pursuing two avenues of research. 1) We have made and tested a P2X₂-FLAG-EGFP protein, this channel functions like wild type P2X₂ channels, localizes to the plasma membrane and faithfully reports the presence of P2X₂ channels in a non-invasive manner by measuring EGFP fluorescence. 2) We are generating P2X₂-ires-EGFP gene targeting constructs, which we anticipate will allow the generation of a map of the neurons that express P2X₂. We will be making knock-in mice with both of these constructs. Our goals are to understand, in living tissue, 1) which neurons express P2X₂ channels, 2) at what stages of development P2X₂ channels are expressed, 3) P2X₂ channel density in synapses, 4) whether patterns of P2X₂ expression change during physiological manipulation (eg. synaptic plasticity) or pathology (eg. inflammation). Comparison of these data with behavioral data from P2X₂ knock-out mice will provide a unique opportunity to study P2X₂ channels in living neurons with hitherto unrealized precision.

The effects of microglial P2X₇ channel activation in a model of Huntington's disease

Sigrid C. Schwarz, Johannes Schwarz & Baljit S. Khakh

P2X₇ channels are activated by extracellular ATP and are found almost exclusively in immunocompetent cells, such as macrophages and brain microglia. Activation of P2X₇ channels causes cell death in microglia via pathways that are poorly understood but may involve cell swelling. Activation of microglia up-regulates P2X₇ but the pathophysiological role of this induction is not known. To explore the role of these channels in the brain's injury response we are using the mouse striatum as a model system. We use BZATP to activate P2X₇ channels, and KN04 and KN64 to block the channels by local infusion into the striatum. We employ a model for Huntington’s disease that involves systemic injections of 3-nitropropionic acid (NPA). The questions under study include (i) what are the consequences for microglia and neurons of P2X₇ activation in the striatum, (ii) does this activation affect typical behaviors associated with unilateral insults to the striatum such as circling, (iii) which genes are regulated as result of P2X₇ activation, (iv) do NPA lesions of striatum affect P2X₇ expression, and (v) do P2X₇ blockers rescue neurons from NPA lesion?

Inhibition of inward rectifier potassium channel Kir2.1 via tyrosine phosphorylation and endocytosis: an unnatural amino acid probe

YanheTong, Eric. Slimko, Gabriel Brandt, Dennis Dougherty, Andreas Karschin

The amplitude of potassium currents can be modulated by phosphorylation. Wischmeyer et al. (1998) identified tyrosine 242 of the inward rectifier potassium channel Kir2.1 as a putative phosphorylation site for receptor tyrosine kinases. The present experiments show that Kir2.1 can also be modulated by v-Src and PYK2 kinase. We applied the nonsense codon suppression method to incorporate the unnatural amino acid o-nitrobenzyl tyrosine (Tyr(ONB)) at position 242 of Kir2.1 expressed in Xenopus oocytes (Miller et al. Neuron, 1998). v-Src was coexpressed, and phenylarsine oxide was present to inhibit tyrosine phosphatases. The Kir2.1-Y242Tyr(ONB) currents have functional characteristics similar to the wild-type currents: the o-nitrobenzyl group presumably prevents both (1) phosphorylation and (2) tyrosine-based interaction with adaptor protein at the position 242. Tyrosine-242 was decaged by 3 s irradiation. The K⁺ currents then decreased by ~50% over ~30 min. Interestingly, over the same period, the membrane capacitance decreased by ~20%, signifying endocytosis. The capacitance change was enhanced and eliminated by coexpression of wild-type (WT) dynamin and dominant negative (DN) dynamin, respectively; the current inhibition was unchanged by WT dynamin and decreased only 2-fold by DN dynamin. We conclude that flash decaging, and subsequent tyrosine phosphorylation, at a single position inhibits the channel, then leads to massive endocytosis with further decreases in current, and we measure the kinetics of these events.

Expression of GIRK (Kir3.1/Kir3.4) channels in mouse fibroblastic cells with and without b1 integrin

Tatiana Ivanina, Clemens Neusch, Yong.-Xin Li, Yanhe Tong, Deane F. Mosher¹, Cesar Labarca.

Recently it was shown that GIRK1 and GIRK4 channels bind directly to integrin b 1 and that the integrin-binding RGD sequence is essential for functional GIRK expression (McPhee et al, 1998). To extend this observation, we compared the expression and subcellular distribution of heteromultimeric GIRK1/GIRK4 channels in two mouse fibroblastic cell lines: one expresses b 1 integrin, and the other (GD10) lacks expression of b 1-integrin due to disruption of the b 1 integrin subunit gene (Sakai et al, 1998). We tagged the carboxyl terminus of the GIRK1 and GIRK4 subunits with Green Fluorescent Protein (GFP). Cells with and without b 1 integrin were transfected with cDNAs for GIRK-GFP fusion proteins. Fluorescent microscopy at 2 days post-transfection showed plasma membrane localization of GIRK1/GIRK4 channels in both cell lines. For whole-cell patch experiments, cells were cotransfected with cDNAs for wild type GIRK1 and GIRK4, m₂muscarinic acetylcholine (ACh) receptors, and GFP. Currents induced by 5 m M ACh (25 mM external K⁺) were 766.3 + 214.5 pA in cells expressing b 1 integrin and 588 + 99.0 pA in cells without b 1 integrin (mean + SEM, n = 10 or 11). There was also no significant difference in basal GIRK currents. These data show that b 1 integrin is not necessary for functional GIRK channels in fibroblastic cells and leave open the possibility of a necessary interaction with other integrins expressed in these cells.

¹Med. & Biomol. Chem.,Univ. of Wisconsin

Co-expression of Gb5 Enhances the Function of the GGL Domain–Containing RGS7

Abraham Kovoor, Ching-Kang Chen, Theodore G. Wenzel¹, Melvin I. Simon

Regulators of G protein signaling (RGS proteins) stimulate the GTPase activity of G protein Ga subunits. Several RGS proteins, including RGS7, contain a Gg subunit-like (GGL) domain, which mediates a specific interaction between these RGS proteins and Gb5. In previous experiments, GIRK deactivation kinetics were accelerated by coexpression of several RGS proteins, indicating that the endogenous rate of GTP hydrolysis rather than intrinsic GIRK channel kinetics determine the time-course of channel deactivation following agonist removal. We used GIRK channels (Kir31/Kir3.4) and m2-muscarinic acetylcholine receptors (m₂-mAchR) heterologously expressed in Xenopus oocytes to evaluate functional effects of the complex between RGS7 and Gb5. We found that the coexpression of RGS7 / GIRK1 / GIRK4 / m₂-mAchR also accelerated recovery compared with absence of RGS7, but considerably less than the acceleration produced by RGS4. When we coexpressed Gb5 / RGS7 / GIRK1 / GIRK4 / m₂-mAchR , however, the acceleration of GIRK channel kinetics was significantly increased and approached that produced by RGS4 / GIRK1 / GIRK4 / m₂-mAchR. This enhancement of RGS7 function by Gb5 was not achieved by co-expression of Gb1, and Gb5 did not modulate the function of RGS4 at any level of RGS4 expression. The mechanism by which Gb5 promotes RGS7 function is under investigation. Support: GM-29836, GM-34236.

¹Baylor College of Medicine, Houston TX 77030

A Classification of Ion Channel Kinetic Models and Identification of Extractable Parameters using Group Theory Methods

G. Shapovalov

The kinetic data obtained in single channel measurements are usually analyzed based on Markovian stochastic behavior and discrete state assumptions. Typical single ion channel kinetic models involve states of equal conductance combined into conductance classes (Let be a total number of states and n_a be a number of states of the a-th class). This fact allows one to construct the transformations (consisting of irreducible blocks n_axn_a) relying on different generators of Markovian flow (or transition matrices), thus making some models indistinguishable from other in a kinetic sense. Moreover this implies that not all the kinetic parameters (i.e. not all the entries of the transition matrices) can be extracted from the available data.

In this work I consider the action of a general transformation on the space composed of vectors representing probability distributions of occupancy of different states. It appears that a transformation of a general form acts on this space as a contraction forming a semigroup that effectively connects different generators only in the case of equilibrium kinetics. It is however possible to identify a nontrivial [sub-]group describing inherent symmetry of the system which appears to be , where S_n denotes the permutation group of order n. Further analysis with application of this symmetry allows one to identify conjugate classes of transition matrices and to build 1-parametric generators with respect to the symmetry produced by equivalent conductances of different states. This also allows one to identify parameters that have no influence on the observable channel kinetics and thus to calculate the number of accessible parameters of the transition matrix. Take for example the (simplest) case of three states with two of the same conductance (typically referred as C-C-O model). Out of total of 6 parameters, one cannot generally be accessed and another one cannot be found from the measurements of equilibrium channel activity, thus leaving only 4 parameters--a well known result for such a model.

Single-Molecule Measurements to Correlate Ion Channel Kinetic and Structural Properties

George Shapovalov

Mechanosensitive (MS) channels comprise an important class of ion channels. It is believed that the action of MS channels represents a first step in an animal’s sensation of touch and even in sensing such basic forces as membrane stretch by osmotic pressure. At present several MS channels have been cloned and studied, for example E. coli MscL (L stands here for channel of large conductance) by S. Sukharev (U. of Maryland).

The structure of another bacterial mechanosensitive channel, Tb-MscL, has been determined by the Rees group at Caltech. This accomplishment allows one to plan experiments on the gating of an ion channel whose structure is known. The mechanosensitive nature of the channel allows for the idea of attempting direct mechanical gating. I intend to use single-channel patch clamping and micromechanical simulation with laser tweezers to test hypotheses about the motions involved in gating of Tb and possibly E. coli MscL channels. At leas two different hypotheses have been proposed, one in the original Chang et. al. paper and another by H. Robert Guy and his associates. In particular, Guy’s model predicts that channel opening would shift to lower osmotic pressure if the tethers are attached to subunits at the extracellular loop between TM1 and TM2, then pulled apart along a line parallel to the plane of the membrane.

The choice of optical tweezers as the tool for pulling on the channel in the patch can be explained by the following: the tweezers can be easily manipulated in 3 dimensions and also their compliance range (sub-pN to ~100 pN) is appropriate for protein conformational changes. I plan to attach the bead to the protein via a DNA or actin tether. The rationale for application of such a technique is possibility of thermal damage to the channel when it is placed it in the laser beam as would be necessary with beads directly attached to the channel. The elastic properties of DNA are now being worked out, so that known forces would be applied to the channel for a given stretching of DNA.

Mutation based cross-linking studies to determine the multiplicity of MscL.

Joshua A. Maurer*, Donald E. Elmore*, Dennis A. Dougherty*

The subunit number observed in the recent crystal structure [1] of the large conductance mechanosensitive channel (MscL) from M. tuberculosis (Tb) is in stark contrast to high conductance and previous pore size estimates of the channel. The crystal structure shows that the channel is a homopentamer, while previous biochemical studies and crystal structures suggest that the channel exists as a homohexamer.

Standard cross-linking studies involving photoreactive or lysine based cross-linking reagents have given variable results with regard to channel subunit number. In previous studies, the outcome of studies with non-specific cross-linking reagents was extremely condition dependent. An inter-subunit hydrogen bond has been identified from the Tb-MscL crystal structure between residues 45 and 51 in adjacent peptide chains. Exploitation of this hydrogen bond should allow for site-specific cross-linking of MscL. To this end, mutations of these residues to either a Cys-Cys or an Arg-Glu pair have been prepared to allow cross-linking by copper phenanthroline or peptide forming reagents (EDC/DCC). Cross-linking will be detected by SDS-PAGE gel electrophoreses and MALDI-TOF mass spectral analysis.

[1] G. Chang, R. H. Spencer, A. T. Lee, M. T. Barclay, D. C. Rees, Science282, 2220 (1998).

*Department of Chemistry, Division of Chemistry and Chemical Engineering

Functional Reconstitution Studies with MscL.

Joshua A. Maurer*, George G. Shapovalov, Dennis A. Dougherty*

The recent crystal structure of the large conductance mechanosensitive ion-channel (MscL) form M. tuberculosis (TB) provides a unique opportunity to study ion-channel structure function relations [1]. In order to elucidate structure function relationships it is necessary to have a variety of experimental techniques to assay channel activity.

Patch clamp electrophysiology on reconstituted MscL and E. coli speroplasts provides insight into channel kinetics and has been used extensively to study mutations in E. coli MscL. In addition to examining mutations in Tb MscL, this technique should provides a method for sieving experiments to determine pore size and a method to screen for potential open state channel blockers.

The development of spectroscopic methods to assay for channel opening in reconstituted vesicles should provide a quick screening method for channel mutations and an additional method to examine potential open state channel blockers. Spectroscopic methods have the additional advantage that the gating tension required to open MscL can easily be determined in these experiments.

The structure-function relations of MscL can also be probed using unnatural amino acids. The channel has been expressed in vitro, providing one potential method for the incorporation of unnatural amino acid. Additionally the possibility of expressing MscL in oocytes is being explored.

[1] G. Chang, R. H. Spencer, A. T. Lee, M. T. Barclay, D. C. Rees, Science282, 2220 (1998).

*Department of Chemistry, Division of Chemistry and Chemical Engineering

A functional, correctly targeted, photochemically calibrated GABA transporter (mGAT1)-GFP fusion protein

Chi-Sung Chiu, Marc Unger¹, Emil Kartalov¹, Stephen Quake¹

We wish to measure the density of neurotransmitter transporters such as mGAT1 by constructing and studying mice whose mGAT1 has been replaced by an mGAT1-GFP fusion. Both N- terminal and C-terminal GFP (K. G. Beam lab construct) fusions of mGAT1 (from N. Nelson, in pcDNAIII) were constructed and tested by transfection into HEK 293T cells. The GFP in the N-terminal of mGAT1 does not function well and does not sort to the cell membrane. However, the C-terminal fusion, which includes a 12-residue spacer between the mGAT1 C-terminus and the GFP, functions well and localizes to the plasma membrane. The mGAT1-GFP has EC₅₀ and V_max values indistinguishable from those of WT GAT1 indicating that this fusion protein shows no functional difference with wild type. Electrophysiological characterization of this fusion construct is in process.

To study the localization of mGAT1-GFP in hippocampal neurons, a lentivirus expression construct was made. The expressed mGAT1-GFP fusion protein is targeted to neuronal processes and accumulates at synaptic sites, for instance by colocalization with synapsin I. We are now constructing a mouse gene targeting vector containing the linker-GFP in the final coding exon.

We also expressed the same GFP construct in bacteria, purified it, coupled it to agarose beads, and determined the GFP density on the beads with single-molecule fluorescence techniques. These beads will serve as calibrations for the density measurements on brain tissue.

¹Division of Engineering and Applied Science

Submolecular motions of the GABA transporter GAT1, a fluorescent kinetics study

Ming LI, Robert A. Farley

We simultaneously measured the fluorescence and voltage-clamp currents of rGAT1, expressed in Xenopus oocyte membranes and labeled with fluorescent dye, in order to study the submolecular motions during GAT1 function. The sulfhydryl-reactive fluorescent dye tetramethylrhodomine (TMRM) was used for labeling. Previous work (Yu et al., 1998) showed that Cys74 of rGAT1 is the only residue that reacts with the SH-reactive MTSET; and the signals described here did not occur with the Cys74Ala mutant. Therefore, we propose that TMRM also labels rGAT1 at Cys74. During hyperpolarizing voltage jumps, the fluorescence increases by ~1% in a two-exponential process with time constants of ~10 ms and ~120 ms. The substrates, GABA, Na⁺ and Cl^-, affect the fluorescence signal differently. Na⁺ has the most significant effect on the signal: NMDG substitution for Na⁺ blocked the slower component. The GAT1 inhibitors NO-711 and SKF-89976A also affect the fluorescence differently: while NO-711 blocked the slower component, SKF-89976A blocked the faster component. These results provide insight about the molecular mechanism of substrate and inhibitor action at GAT1. We propose that upon Na⁺ binding, the side chains near Cys74 undergo a conformational change that is important for rGAT1 function.

Efficient transfer of a dominant negative inhibitor of CREB by recombinant Sindbis virus infection of cells

Tzu-Ping Yu, Dong. Ju, and N. Davidson.

cAMP response element-binding protein (CREB), an important transcription factor that binds to cAMP response elements (CRE sites) in promoters, has been implicated in the formation of memory in many species. CREB can be phosphorylated by either cAMP- or calcium-dependent protein kinases. A dominant negative inhibitor of CREB (A-CREB) has been shown to completely inhibit cAMP-mediated transcription of c-fos expression in PC12 cells (Ahn et al., Mol. Cell. Biol., 1998). In the present study, we intend to use gene transfer techniques to study the role of CREB in cAMP-mediated synaptic plasticity in neurons.

A cDNA encoding an A-CREB insert was derived by polymerase chain reaction from a plasmid kindly provided by Drs. J. Moll and C. Vinson. The coding segment of A-CREB and a DNA segment encoding an IRES-GFP fragment were inserted into the Sindbis virus vector, pSinRep5 of the Sindbis vector kit provided by Invitrogen. In vitro capped RNA transcripts of pSin-ACREB-IRES-GFP and helper DNA (DH-26S) were synthesized and transfected into BHK cells. Viral particles were harvested 66 hours after transfection. Different dilutions of viral solution were bath applied to BHK cells. Cells were examined 18 hours after infection. About 90% of cells were infected and revealed strong fluorescence when the virus was applied with a dilution of 1:20. With dilution of 1:200, approximately 50% of cells were infected. We are currently testing pSin-ACREB-IRES-GFP using several functional assays for CREB function.

Increased GABA sensitivity: one of multiple effects of acetylcholine on GABAergic synaptic transmission in cultured rat hippocampal neurons.

Sila Konur¹, Yong-Xin Li

We investigated the effects of acetylcholine (ACh, 100 mM) on GABAergic synaptic transmission in cultured rat hippocampal neurons using whole-cell patch clamp techniques (10 mM CNQX was added in the recording medium). (1) ACh induced a transient inward current, 50-700 pA, lasting for 1-4 seconds in 13 of 30 cells. This current was blocked by 10 nM MLA, a nicotinic receptor antagonist. (2) There were 12-50 pA inward currents 1-1.5 min in duration. Without TTX, ACh increased both frequencies and amplitudes of spontaneous, action potential evoked synaptic currents. (3) With TTX in the recording medium, the change in holding currents was followed by a decrease in frequency of miniature postsynaptic currents (minis). (4) A slower increase in mini frequency, reaching a maximum of 5-10 Hz, or ~5 times the original value, appeared ~ 5 min after ACh application. (5) The increase in spontaneous mini frequency was accompanied by an increase in the amplitude of minis. There was no significant change in the waveform of the minis. Simultaneously with phase (5), the increased mini amplitude, we detected increased currents induced by exogenous applied muscimol, a GABA_A agonist, suggesting an increase in the number and/or the sensitivity of postsynaptic GABA receptors. Phase (5) was due to muscarinic receptors.

¹SURF student

Back to Table of Contents | Back to the Lester/Davidson Lab Homepage