molecular genetic approaches to the targeted suppression of neuronal activity

Review R1041

Molecular genetic approaches to the targeted suppression ofneuronal activityBenjamin White, Thomas Osterwalder and Haig Keshishian

Understanding how the diverse cells of the nervoussystem generate sensations, memories and behaviors isa profound challenge. This is because the activity ofmost neurons cannot easily be monitored or individuallymanipulated in vivo. As a result, it has been difficult todetermine how different neurons contribute to nervoussystem function, even in simple organisms likeDrosophila. Recent advances promise to change thissituation by supplying molecular genetic tools formodulating neuronal activity that can be deployed in aspatially and temporally restricted fashion. In somecases, targeted groups of neurons can be ‘switched off’and back ‘on’ at will in living, behaving animals.

Address: Molecular, Cellular, and Developmental Biology Department,Yale University, New Haven, Connecticut 06520, USA.

Correspondence: Benjamin WhiteE-mail: [email protected]

Current Biology 2001, 11:R1041–R1053

0960-9822/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved.

IntroductionThe nervous system has an enormously complex task. Itmust monitor and evaluate changes both in an organism’sinternal state and in the surrounding world and use thisinformation, together with previously stored information, togenerate appropriate behaviors. One general approach tounderstanding how a nervous system accomplishes thisinvolves selectively perturbing its molecular and cellularcomponents and determining how these perturbationsaffect its function. A particularly critical implementation ofthis last approach involves inactivating specific neurons todetermine their roles in development, information pro-cessing and behavior. Recently, progress in this directionhas been aided by the development of generally applica-ble methods for suppressing the activity of arbitrary groupsof neurons, in some cases in a reversible fashion.

Two developments have led to the introduction of suchtechniques within the last several years: one is the identifi-cation and characterization of several genes useful forinhibiting neural activity; the other is the exploitation ofDNA regulatory sequences governing cell-type specificgene expression. Genes whose products suppress neuralactivity can be introduced into genetically tractable organ-isms and selectively overexpressed in targeted cell types.In the fruit fly Drosophila, genetic suppressors of activitycan be targeted to specific groups of neurons with remark-able selectivity [1–3]. The introduction of drug-inducibletranscription factors that can be deployed in a cell-typespecific manner is also beginning to permit temporal controlof suppression [4].

In this review, we discuss molecular genetic approaches tocontrolling neuronal activity and describe in detail severalquestions they are being used to address. We focus primar-ily on the genetic tools that can be employed to inhibitactivity, describing only briefly the tools required for cell-type specific and temporally regulated expression, whichhave been extensively reviewed elsewhere [5,6]. While wemention developments in both invertebrate and vertebratemodel organisms, we concentrate on recent work using thefruitfly, Drosophila melanogaster. It is in Drosophila that theideal of general, cell type-specific, graded, and reversiblemanipulation of neural activity is most rapidly beingapproached and where the application of emerging tech-niques is lending new insight to old problems. Investiga-tions into the homeostatic regulation of synaptic functionand the cellular basis of memory are both benefiting fromthe new approaches. We describe these developmentsin detail.

R1042 Current Biology Vol 11 No 24

Methods for suppressing electrical activityusing K+ channelsSuppression of neuronal activity involves the inhibition ofeither electrical or synaptic activity (Figure 1). Tools forsuppressing electrical activity target processes underlyingthe maintenence of membrane excitability and can affect,in principle, virtually all aspects of neuronal function. Incontrast, tools for suppressing synaptic activity specificallytarget processes involved in neurotransmission and affectonly neuronal output or input. Because of their potentiallybroader scope of application, considerable effort has beenmade to develop tools that universally block membraneexcitability. Although neurotransmitter-gated Cl- channelshave found limited use in this regard [7], most efforts havefocused on K+-conducting channels, which typically do notrequire neurotransmitter for activation (Figure 2).

Methods using naturally occurring K+ channelsDiverse K+ channel types occur in nature. Their selectiv-ity for K+, and the negative equilibrium potential for thision (EK) in neurons, underlie their natural roles in limitingor modulating membrane depolarization (see Box 1). Infact, neurons tailor their patterns of electrical activity byexpressing different subsets of K+ channels, and deployingthem to appropriate subcellular sites. The mechanisms by

which neurons regulate channel expression, membranetargeting and levels of surface expression are highly regu-lated [8]. However, as early as 1992, Kandel and his col-leagues [9,10] showed that cultured Aplysia neuronstransfected with native K+ channel genes would function-ally express the channels at levels high enough to alter,and even suppress, electrical activity.

The work from Kandel’s laboratory used voltage-gatedchannels, which open in response to depolarization. Sub-sequent K+ channel overexpression studies in mammalianneurons have focused primarily on inward rectifier K+

channels (see Box 2). These channels tend to be open atrest, and to close in response to depolarization. Mostinward rectifier channels are subject to regulation by otherfactors [11]. Using inward rectifiers regulated by G-pro-teins, or GIRKs, Lester’s laboratory [12] first showed thatoverexpressing virally transduced GIRK genes in hip-pocampal neurons reduced excitability after G-proteinactivation by neurotransmitter. Marban’s group [13], build-ing on methods for modulating electrical activity in cardiaccells by K+ channel overexpression, similarly demonstratedthat human Kir2.1, another inward rectifier, efficiently sup-pressed excitability of superior cervical ganglion neuronsafter induction of channel gene expression by ecdysone

Figure 1

The functional anatomy of neurons andstrategies for the suppression of activity.Neurons receive and integrate inputs, thengenerate and transmit outputs. Theseprocesses are typically carried out in separatecompartments of the neuron, but all relyfundamentally on electrical changes in theneuronal membrane. While electrical activityforms the basis of all neuronal activity, inputand output occurs at synapses, whereelectrical signals are typically converted tochemical signals for transmission. Neuronalactivity can be suppressed by targeting themachinery underlying either electricalconduction or chemical transmission. Thefigure shows the anatomy of a typicalDrosophila motor neuron, which receivessynaptic inputs on its dendritic arbor andforms synaptic outputs on a muscle cell.Integration is carried out by the dendrites andsignals propagate down the axon to theneuromuscular synapse. Rectangles indicatethat membrane conductances, due primarily tochannels selective for Na+ and K+ ions asillustrated in the inset (a), underlie electricalactivity in all compartments as well as in non-neuronal cells such as muscles. Circlesindicate synaptic regions and are enlarged inthe inset (b), which shows the processesunderlying release of neurotransmitter fromvesicles.

Synaptic inputs

Neuron

Muscle

Synapticoutput

Current Biology

Na+

Na+

Ca++

Na+

K+K+

K+K+

(a) (b)

[14,15]. In addition, fusion of the Kir2.1 channel to GreenFluorescent Protein (GFP) allowed visualization of geneexpression and channel localization without impairing theability of the channel to potently inhibit excitability.

Targeted expression of the GFP-Kir2.1 gene using theGal4–UAS system has since been used to suppressexcitability in vivo in Drosophila. Baines et al. [16] showedthat Kir2.1 expression in embryonic motor neurons couldinhibit synaptic transmission at developing neuromuscularsynapses. More recently, GFP-Kir2.1 has been used toattenuate both muscle [17] and photoreceptor [18] excitabil-ity. However, not all inward rectifiers suppress excitabilitywith equal efficacy, and some have deleterious effectswhen overexpressed in neurons. For example, the inwardrectifier channel Kir1.1 has been shown to induce apopto-sis when overexpressed in cultured mammalian neurons,evidently by promoting K+ efflux [19].

That Kir2.1 overexpression does not appear to lead to neu-ronal demise suggests that differences in the conductionproperties of channels of the same type can have strong

effects on a channel’s ability to suppress excitability. Evenclosely related channels can produce what appear to be verydifferent effects, depending on the context of expression.For example, Sutherland et al. [20] report that overexpres-sion of Aplysia Kv1.1 in the hippocampus of transgenic miceresults in complex changes in endogenous channel expres-sion and paradoxical hyperexcitability at the systems level.In contrast, overexpression of the equivalent mammalianchannel in sensory neurons of the nematode Caenorhabditiselegans yields developmental results consistent with thesimple suppression of excitability [21].

Methods using mutant K+ channelsWhile most efforts to use K+ channels to inhibit excitabil-ity have involved overexpressing native channels, twogroups have used mutant, voltage-gated channels for sup-pression [3,22]. Voltage-gated K+ channels often open tooslowly and at potentials too positive to be useful in oppos-ing depolarizing currents. However, point mutations inregions involved in voltage sensing can generate channelsthat activate at more negative potentials [23]. Two suchmutations in distinct K+ channel genes have been found tounderlie deficits in egg-laying and motor coordination inC. elegans, evidently by decreasing muscular and neuronalexcitability [24,25]. Capitalizing on this finding, Zhao et al.

Review Molecular genetic suppression of neuronal activity R1043

Regulation of electrical potentialsNeurons generally maintain a negative voltage across theirmembranes. This electrical resting potential is largely establishedby ion channels that selectively allow K+ ions to permeate anotherwise non-conducting cell membrane. As the concentration ofK+ inside neurons is usually 15–30 times greater than levelsoutside, there is a strong tendency for K+ to diffuse out of the cell.However, the outward diffusion of K+ has an electricalconsequence: positive ions flowing out of the cell leave behindunbalanced negative charges. The accumulation of unbalancedcharges — negative on the inside, positive on the outside — exertsa powerful electrical force on K+ ions traversing the membrane.This force retards the outward, and promotes the inward, passageof K+ ions. The net outward flow slows to a halt at the K+

equilibrium potential, given by the Nernst equation, where theconcentration-dependent efflux of K+ ions is exactly balanced bythe electrically driven influx. Most cells have their restingmembrane potentials close to the K+ equilibrium potential.

In excitable cells, such as neurons, excitation involves atransient change in the membrane voltage to a more positivevalue, often through the entry of cations such as Na+ or Ca2+. Theconcentration of Na+ outside the cell is usually 10 times greaterthan on the inside. As a result, when channels selective for Na+

open, the influx of Na+ neutralizes the net negative charge insidethe cell and depolarizes the membane, causing its voltage tobecome more positive. During an action potential the peak voltagemay approach the (positive) Na+ equilibrium potential. Actionpotentials are transient, and are terminated both by inactivation ofthe voltage-dependent Na+ channels that sustain depolarization,and by membrane repolarization mediated by the delayedactivation of voltage-dependent K+ channels.

Box 1Figure 2

Overexpression strategies for suppressing electrical activity. The mostgeneral approach to attenuating electrical activity involves enhancing theconductances normally used by neurons to oppose excitation. Thismeans augmenting either Cl-, or more generally K+, conductances byincreasing the number of channels in the membrane that conduct theseions. (a) In a typical neuron, voltage-sensitive Na+ channels (red) allowpositive charges to flow into the cell and depolarize the membrane.Depolarization is opposed or modulated by K+ channels, which allow theflow of positive charges back out of the cell. Some of these channels aregated by voltage (blue with +-sign), others serve as ‘leak channels’ or areregulated by other factors (blue). The inset shows the characteristicdepolarization and repolarization of an electrically active membraneduring an action potential. (b) The effects of overexpressing a voltage-sensitive K+ channel in the membrane. Membrane depolarization now notonly initiates Na+ influx, but also potently enhances K+ efflux whichcounteracts the depolarization and blocks the generation andpropagation of the electrical signal, or action potential (inset).

Current Biology

Na+

K+ K+

K+ channel overexpression

K+ K+ K+ K+ K+

Na+

(a)

(b)

[22] targeted expression of one of the mutant channels tocholinergic neurons in C. elegans to study the effects ofactivity suppression on axonal sprouting.

Similarly, White et al. [3] genetically modified the rapidlyactivating Shaker K+ channel of Drosophila, introducing pre-viously characterized mutations [26,27] that cause it to openat more negative voltages and to remain open upon sus-tained depolarization. These investigators showed that thischannel, which they call the ‘Electrical Knock-Out’ or EKOchannel, substantially attenuates cellular excitability incentral and peripheral neurons, as well as in muscles, ofDrosophila when targeted to these cell types. In subsequentwork, Osterwalder et al. [4] have shown that induction ofEKO channel expression in Drosophila muscle using thedrug-inducible GeneSwitch Gal4 transcription factor can beused to regulate excitability in a targeted and temporallyrestricted fashion in vivo. This work builds on previousefforts, which showed the temporal regulation of K+ channelexpression using either heat shock [28–30] or tetracycline-sensitive [31] promoters. However, unlike earlier methods,the GeneSwitch technique achieves both cell-type specificand temporally controlled expression of channel constructs.

General considerations of K+ channel based techniquesWhile both the Kir2.1 and EKO channels have shownbroad utility in Drosophila, and other K+ channel typeshave been shown to suppress excitability in isolated appli-cations, the cell biological mechanisms that permit pertur-bations of excitability are incompletely understood. As thestudy by Nadeau et al. [19] indicates, suppression ofexcitability is not always tolerated, an issue that may be of

particular importance in mammalian neurons. And it is asyet unclear which parameters are relevant to successfulsuppression. Channel conductance properties — when thechannel is open and how much K+ it conducts — are likelyto be relevant, but correct channel biogenesis and surfaceexpression are also important.

Both GFP-Kir2.1 and EKO contain membrane-targetingsignals — PDZ binding domains — that direct them tospecific subcellular sites [3,17]. It is not yet clear, however,that all neurons appropriately recognize these targetingmotifs, or that the sites to which the channels are directeduniversally permit attenuation of electrical activity. Indeed,White et al. [3] found that even high levels of EKO channelexpression in photoreceptors were unable to attenuate thephotoresponse by more than 50%. An important challengeis to determine whether different targeting motifs can beused to direct these channels to other subcellular domains.If so, it may soon be possible to efficiently suppressexcitability in specific membrane compartments such asaxons, dendrites and somata, and selectively perturb spe-cific neuronal functions.

The efficacy of a channel in suppressing the excitability of agiven cell type may also be affected by the cell’s ability tomodulate channel function. Kir2.1, for example, requires thephospholipid phosphotidylinositol 4,5 bisphosphate (PIP2)for its activity [32], and limiting quantities of this lipid canreduce channel activity. However, this property can some-times be useful in its own right. For example, Hardie et al.[18] have cleverly taken advantage of the phospholipiddependence of GFP–Kir2.1, using the channel as a PIP2sensor in wildtype and mutant Drosophila photoreceptors.

Methods for suppressing synaptic activityWhile both pre- and postsynaptic manipulations can be usedto suppress synaptic activity, the most generally applicabletools target the neurotransmitter release machinery, which isessentially the same at all chemical synapses (Figure 3a).Compared to the relative simplicity of the ionic processesgoverning membrane excitability, the processes that controlneurotransmission are complex. Dozens of proteins regulatethe storage, docking, priming, fusion, and recovery of synap-tic vesicles [33,34]. Fortunately, nature has offered clues asto which of these proteins are essential and, in some cases,has provided molecular tools for disrupting them.

Tetanus toxin light chain blocks synaptic vesicle releaseParticularly interesting are the clostridial toxins producedby the pathogenic bacteria responsible for tetanus and bot-ulism, which act by disrupting synaptic transmission (forreview see [35]). The catalytically relevant portions of thegenes for tetanus and botulinum toxins encode metallo-proteases that cleave critical proteins in the transmitterrelease machinery (Figure 3b). The proteolytic region, the


K+ channelsNeurons use a wide array of K+ channels to tailor their electricalproperties. As described in Box 1, such channels are essential insetting the resting properties of neurons and in shaping their activeresponses. The two principal types of K+ channel discussed in thisreview are voltage-gated K+ channels and inward rectifiers. Voltage-gated K+ channels are typically closed at rest and open only inresponse to membrane depolarization. When not over-expressednor mutated to behave otherwise they have little effect on the restingproperties of a neuron, and instead influence characteristics of theactive response such as the height, duration or frequency of actionpotentials. In contrast, inward rectifiers normally help to maintain andstabilize the resting properties of a neuron, and can stronglyinfluence its responsiveness to depolarization. Channels of this typetend to be regulated by other cellular factors, and when not over-expressed their effects on excitability are limited by their tendency tostop conducting at more depolarized potentials, due to theirblockage by intracellular Mg2+ ions. In addition to these two classes,many other types of K+ channel are used by neurons, but becauseof their direct regulation by cellular factors such as intracellularCa2+, phosphorylation or ATP, they have been less extensivelyinvestigated for use as suppressors of excitability.

Box 2

light chain, of tetanus toxin (TeTxLC) cleaves synapto-brevin, while the botulinum toxins cleave either synapto-brevin, SNAP-25 or syntaxin.

Mochida et al. [36] first recognized the potential of truncatedtoxin genes for inhibiting synaptic release, showing thatsynaptic transmission was suppressed in Aplysia neurons het-erologously expressing light chain mRNAs. Sweeney et al.[37] extended this approach, elegantly showing that neu-ronally targeted expression of the TeTxLC gene potentlyinhibits synaptic activity in vivo in Drosophila. TeTxLC hassince proved to be a powerful tool in that organism, both inprobing the functional roles of specific neurons and in eluci-dating the role of synaptic activity in developmentalprocesses [1,16,38–45]. The recent introduction of a methodfor developmentally regulating TeTxLC activity furtheraugments the power of this technique [46].

Temperature-sensitive dynamin mutant depletes synapticvesiclesWhile naturally occurring toxins have supplied one tool formanipulating synaptic activity, other tools have emergedfrom mutagenesis studies designed to identify genes

critical for neurotransmission. The proteins encoded bythese genes are obvious targets for inhibition of synapticactivity, and in some cases simple overexpression of theseproteins can reduce neurotransmission [47]. The targetedexpression of mutant genes represents another approach toblocking the function of synaptic proteins, and Drosophilaagain provides an elegant example of this approach, devel-oped by Kitamoto [48], using a conditional mutant of thekey protein dynamin (Figure 3c).

Dynamin is a mechanoenzyme essential for vesicular endo-cytosis, and a Drosophila dynamin mutant named Shibire(Shits1) was isolated by Suzuki and colleagues [49–51] some30 years ago, as a temperature-sensitive paralytic. Shits1 fliesare viable and motile at room temperature, but paralyzewithin seconds at the restrictive temperature of 29ºC, dueto cessation of vesicular endocytosis and depletion of thesynaptic vesicle pool [52,53]. Importantly, the Shits1 muta-tion is genetically semidominant, implying that the mutantdynamin protein functions in a dominant-negative fashion.

Because of dynamin’s general role in membrane retrieval,the Shits1 mutation has pleiotropic effects, disrupting a


Figure 3

Strategies for suppressing synaptictransmission. (a) Neurotransmission atchemical synapses involves the presynapticrelease of neurotransmitter in response todepolarization-mediated Ca2+ entry. Thebinding of neurotransmitter by postsynapticreceptors, which also act as ion channels,initiates a new electrical impulse.Neurotransmitter release requires (1) uptakeof neurotransmitter (red circles) into synapticvesicles; (2) docking and (3) priming ofsynaptic vesicles; (4) Ca2+-mediatedexocytosis of neurotransmitter; (5) clathrin-mediated endocytosis; and (6) recycling ofsynaptic vesicles. Many of the proteins thatmediate these processes are found at allchemical synapses, including the three criticalSNARE proteins — synaptobrevin, syntaxin,and SNAP-25 — which mediate membranedocking, and dynamin, which mediates vesiclerecycling. (b) The SNARE proteins are targetsof tetanus and botulinum neurotoxins, andpresynaptic expression of the gene for tetanustoxin light chain (TeTxLC) inhibits synapticvesicle docking by proteolytically cleavingsynaptobrevin, thus blocking neurotransmitterrelease. (c) The Drosophila mutant alleleShibire, Shits1, encodes a temperaturesensitive, dominant negative form of thedynamin protein. Presynaptic expression ofShits1 leads to depletion of the synapticvesicle pool at restrictive temperatures, andrapidly abolishes chemical synaptictransmission. Current Biology

Ca2+

Na+Na+

K+K+

SynaptobrevinSyntaxin, SNAP-25

1

23 4

5

6

Dynamin

TeTxLC Shits1

(a)

(b) (c)

Ca2+Ca2+

variety of cellular functions [54–57]. However, the rapidreversibility of the Shits1 phenotype has made it a valuabletool for the episodic silencing of synaptic transmissionthroughout the nervous system. It was Kitamoto [48],however, who recognized that the semidominant nature ofthe Shits1 mutation meant that it could be used morespecifically. Targeting expression of the Shits1 gene to spe-cific neurons using the Gal4–UAS system, Kitamoto showedthat one could effectively turn neurotransmission on andoff in cholinergic neurons by literally moving animalsbetween incubators. Exquisite temporal control of synap-tic activity can thus be achieved in a cell-specific fashionwithout induction of gene expression.

General considerations of techniques that block synaptictransmissionTeTxLC and Shits1 represent elegant and potent tools forsuppressing synaptic activity. The former tool, which is notintrinsically subject to temporal regulation, has clear advan-tages in developmental studies, where prolonged inhibitionof synaptic function is required. Under these conditions, thepleiotropic effects of prolonged block of dynamin funtioncan lead to defects in Shits1 expressing tissues [48]. The tar-geted Shits1 technique, however, with its capacity for acuteinhibition of neurotransmission is perfect for investigationsinto the neuronal basis of behavior.

The great challenge to all approaches for suppressingsynaptic transmission is that many of the molecules essen-tial for neurosecretion, or their relatives, play non-synapticroles in vesicle trafficking. It is perhaps worth noting inthis context that the TeTxLC and Shits1 techniques side-step this challenge in different ways. TeTxLC, which inmany animals cleaves multiple synaptobrevin isoforms,some of which are involved in constitutive vesicle secre-tion [58], cleaves in Drosophila only the synaptic isoform[37]. In contrast, the targeted Shits1 technique obviates thepleiotropic effects of the Shibire mutation by restrictinginhibition to short times, where the effects on neurosecre-tion are profound, but those on other cellular functions arenot. It is unfortunate that neither approach can be readilyadapted for use in mammals.

Methods of targeting suppressionThe genetic tools developed to suppress neuronal activityin vivo would have little advantage over pharmacologicalagents, such as the Na+ channel blocker tetrodotoxin(TTX), were it not possible to express them in a cell-typespecific manner. While it is beyond the scope of thisreview to describe in detail the techniques for cell-typespecific expression, one is outlined in Figure 4a. This isthe elegant and powerful Gal4–UAS system of Drosophilaintroduced by Brand and Perrimon [59], which allows arbi-trary genes of interest to be expressed in defined groups ofneurons, and permits gene expression to be incrementally

increased by increasing the UAS–transgene dosage. Therecent introduction of an inducible Gal4 transcriptionfactor into this system (Figure 4b) permits temporal aswell as cell-type specific regulation of the patterns of geneexpression [4,60]. Being able to control expression tempo-rally is particularly important because of the capacity ofthe nervous system to compensate for perturbations inelectrical and synaptic activity.

Applications of techniques for suppressingactivityThe tools described above for attenuating neuronal activityare being used to study problems ranging from nervoussystem development to the neural basis of specific behav-iors (Table 1). In addition, they are permitting analysis ofthe mechanisms underlying the maintenance of synapticefficacy and cellular excitability. To illustrate the impactthat the new techniques are having, we describe below twoareas in which they are making fundamental contributions:the study of activity-dependent mechanisms in neuromus-cular development, and the study of memory processing.

Activity-dependent mechanisms underlying developmentaland functional plasticityThe functional development of the nervous system dependscritically on electrical activity, with the establishment andmaintenance of correct connections often depending uponspontaneously generated or sensory-driven activity [61,62].While tools for suppressing neural activity have been usedin Drosophila to examine its role in synaptogenesis in boththe giant fiber system [41] and the visual system [40], mostof the work has focused on the larval neuromuscular junc-tion [3,17,37,42].

Activity-dependent plasticity is a well-known feature ofthe developing Drosophila neuromuscular junction (reviewedin [63]). Both synaptic connectivity [64,65] and synapticmorphology [66–68] at the larval neuromuscular junctionare strongly influenced by the levels of activity in eithermotor neurons, muscles, or both. In addition, postsynapticactivity is known to regulate homeostatically the physiol-ogy of the synapse [69], with clear evidence of communi-cation between the muscle and the motor terminal tocontrol levels of transmitter release [70–72]. A spate ofrecent papers, in which TeTxLC [16,37,42], GFP–Kir2.1[16,17] or EKO [3] have been selectively expressed oneither side of the neuromuscular synapse, have con-tributed new insights into the mechanisms underlyingneuromuscular junction plasticity in anatomy and function(Figure 5).

Plasticity in the developmental pattern of neuromuscularjunction connectivity has been investigated by White et al.[3] who have expressed the EKO channel independentlyin motor neurons and in muscles. While wild-type larvae


display a highly stereotyped pattern of neuromuscularsynapses [64,73], suppression of excitability in embryonicmotor neurons with EKO led to the appearance of ectopicsynaptic connections onto muscle fibers. This result is con-sistent with earlier observations of Jarecki and Keshishian[64], who silenced neuronal activity using either TTX orNa+ channel mutations.

Examining the motor ending vitally, White et al. [3] foundthat the electrically silenced motor neurons sproutedlonger and more stable processes as they contacted theirmuscle targets. Surprisingly, connectivity was essentiallynormal when muscle excitability was specifically sup-pressed using the EKO channel. Similarly, there are noerrors in connectivity when muscle excitation is blockedby suppression of synaptic release using TeTxLC [37] orby pharmacological blockade of postsynaptic glutamatereceptors [64]. These results suggest that the mis-wiringfollowing suppression of excitability results from a cellautonomous effect in the motor neuron.

Cell autonomous ectopic sprouting of sensory neuron axonshas been reported in the nematode by Peckol et al. [21],who used mammalian voltage-gated K+ channels to suppressneuronal excitability. These authors show that mutantswith defective voltage-gated Ca2+ channels display aberra-tions in axonal sprouting, implicating inhibition of presy-naptic Ca2+ entry in the mechanism of ectopic synapseformation. Interestingly, while sprouting of neuromuscu-lar contacts is observed upon suppression of neural activ-ity in C. elegans, the mechanisms appear to be differentfrom those in Drosophila, as suppression of muscle excitabil-ity and block of neuromuscular transmission also lead tosprouting [22].

The homeostatic mechanisms involved in maintainingsynaptic efficacy at the neuromuscular junction have alsorecently been investigated using the new tools for the sup-pression of excitability. While previous work showed thatmuscles monitor presynaptic input and adjust it to keepthe amplitude of synaptic potentials within a physiologi-cally appropriate range of values, it has not been clear howthe muscle ‘senses’ input strength. To test the possibilitythat membrane depolarization, perhaps via Ca2+ entry, actsas the primary sensor, Paradis et al. [17] directed overex-pression of GFP–Kir2.1 specifically to larval muscles toinhibit postsynaptic excitability without completely sup-pressing it.


Figure 4

Techniques for cell-type specific and conditional expression oftransgenes in Drosophila. (a) In Drosophila, the most widely usedsystem for tissue-specific transgene expression is the bipartiteGal4–UAS system. Transgenic flies expressing the yeast transcriptionfactor Gal4 (blue) in a cell-type specific fashion (e.g. the larval nervoussystem), are mated with transgenic flies carrying a gene of interest(gene X, green) placed behind the upstream activating sequence (UAS,purple) of Gal4. The progeny then express gene X in the same patternin which Gal4 is expressed in the parental line. Many thousands of so-called Gal4 driver lines, with unique expression patterns in the nervoussystem, now exist. (b) In a variation of this technique, the Gal4transcription factor, which is normally constitutively active, is replacedby the conditional GeneSwitch transcription factor to generateGeneSwitch drivers. The chimeric GeneSwitch transcription factor hasthe same UAS target specificity as the Gal4 protein, but requires theligand RU486 for its transcriptional activity. In the absence of RU486(upper panel), GeneSwitch is transcriptionally incompetent (blue) andgene X is not expressed (grey nervous system). Systemic application ofRU486 (red dots, lower panel), however, activates GeneSwitch (red)and gene X is expressed with a tissue specificity conferred by theexpression pattern of GeneSwitch (green nervous system).

EnhancerPRGeneSwitch

GeneSwitch

UAS Gene X

RU486

Enhancer

Current Biology

PR

+RU486

UAS Gene X

UAS Gene X

EnhancerPRGAL4

(a)

(b)

Despite a 50-fold smaller input resistance in the GFP–Kir2.1expressing muscle fibers, nerve stimulation depolarizedthese fibers to exactly the same potentials as wild-typefibers. This required the excitatory postsynaptic potentialsin GFP–Kir2.1 expressing animals to increase by 30%.Postsynaptic currents were also correspondingly larger, butwithout apparent change in the number or sensitivity ofglutamate receptors, confirming an upregulation of motorneuron neurotransmitter release. These results confirmthe homeostatic coupling of presynaptic release to muscleresponse and provide strong evidence that membranedepolarization acts as the primary response sensor of inputstrength in the muscle, though a role for glutamate recep-tors cannot be ruled out (see for example [71]).

While several mechanisms might enhance presynapticrelease when muscle response is suppressed, one

mechanism might involve upregulating the excitabilityof the motor neuron. Baines et al. [16] have recentlyshown that complete block of neuromuscular transmis-sion using TeTxLC leads to upregulation of both Na+

and K+ conductances with a net increase in excitability inembryonic motor neurons. Even more surprising is theobservation that this manipulation also results in thesimultaneous silencing of the presynaptic inputs to themotor neurons [42]. This suggests that motor neuronsincapable of exciting muscles become functionally ‘iso-lated’ and fail to receive synaptic inputs through theircentral dendrites. One possible interpretation of theseresults is that retrograde signals from the muscle exercisecontrol over both the physiology and connectivity of theinnervating motor neuron, regulating its competence toreceive innervation, as well as its levels of excitabilityand presynaptic release.


Table 1

Applications of tools for suppressing neuronal activity.

Tool* Preparation†

1. In vitro applicationsSuppression of excitability in cultured aKv1.1 Aplysia neurons [9]cells using K+ channels aKv5.1 Aplysia neurons [10,30]

GFP-Kv1.4 Skeletal myoballs [81]dKv1.1 Cardiac myocytes [13]

rKir3.x (GIRKs) Hippocampal neurons [12]GFP-Kir2.1 SCG neurons [14,15], hair cells [80]

HERG Cardiac myocytes [85]EKO Aplysia neurons [3]

rKir1.1 Hippocampal neurons [19]Suppression of excitability in cultured hGABACR Hippocampal neurons [7]cells using Cl– channels ceGluR-C Mammalian neurons [79]Synaptic suppression in cultured cells TeTxLC Aplysia neurons [36]

2. In vivo applicationsNeuronal differentiation xKv1.1, xKv1.2 Xenopus embryos [86,87]Synaptogenesis EKO Embryonic neurons [3], larval muscles [3,4]

GFP-Kir2.1 Embryonic motor neurons [16], larval muscles [17]ceKv3 (egl36) C. elegans motor neurons [22]rKv1.1, rKv1.2 C. elegans sensory neurons [21]

TeTxLC Embryonic neurons [37,42], photoreceptors [40]Synaptic function EKO Larval muscles, photoreceptors [3]

GFP-Kir2.1 Larval muscles [17]Shits1 Photoreceptors [48]

Behavior Eclosion EKO Subset of neurons [3]Locomotion and reflex TeTxLC Various CNS and sensory neurons [39,43,88]

Shits1 Cholinergic neurons [48]Olfaction TeTxLC Larval and adult neurons [37,38]

Circadian rhythms TeTxLC Ventral lateral neurons [1,45]Sensitization TeTxLC Ddc-expressing neurons [44]

Memory Shits1 Peptidergic neurons [75], MB neurons [2,76]General physiology dKv1.1 Muscles [28], neurons [29]

aKv1.1 Mouse CNS [20]rSK3 Mouse myotubes [31]

GFP-Kir2.1 Photoreceptors [18]TeTxLC Giant fiber neurons [41], photoreceptors [89]dunc-18 Neurons [47]

*Lower case letters preceding the names of the genetic tools indicate species: a, Aplysia; d, Drosophila; r, rat; h, human; and ce, C. elegans.†Unless otherwise indicated the preparation is Drosophila.

Examining the cellular basis of behavior: learning andmemoryAnother natural application of genetic tools that inhibitneural activity is to examine the cellular basis of behavior.This application parallels the investigation of the geneticbasis of behavior, pioneered by Benzer and his colleagues,in which randomly mutagenized flies are subjected tobehavioral screens to isolate genes underlying specificbehaviors (see [74]). In a similar way, Gal4 enhancer traplines [59] driving the expression of activity suppressor con-structs in random cell types, can be used with the samebehavioral screens to isolate neurons underlying thesebehaviors. This approach, which we refer to as ‘neurotrap-ping,’ can in principle be used to determine the completeneuronal circuitry underlying a given behavior. Althoughsystematic neurotrapping approaches are still in theirinfancy, preliminary results along these lines have beenreported for a variety of behaviors, including olfactory

escape behaviors [37], locomotor function [39,43], andwing expansion [3].

Suppressors of activity such as TeTxLC, GFP-Kir2.1,EKO, and Shits1, can also be used in more directedapproaches to the study of specific behaviors. For example,one can manipulate the activity of neurons known to beinvolved in a given behavior by directing expression of sup-pressors of activity to these neurons using Gal4 lines withdefined expression patterns. The Shits1 technique with itscapacity for rapid and reversible control of suppression is aparticularly useful tool for this purpose, as illustrated by itsapplication to the problem of memory processing in thefruit fly [2,75,76].

The mushroom bodies (Figure 6a) of adult flies are pairedbrain structures which receive multimodal afferent input,and are essential for learning and memory in Drosophila


Figure 5

Investigation of activity-dependent processesin neuromuscular development. TeTxLC, GFP-Kir2.1 and EKO have been used to investigatesynaptogenesis and synaptic function at theneuromuscular junction in Drosophila (seetext). The effects of expressing these agents inmuscles, motor neurons, or all neurons isshown. The top panel shows the synapticconnectivity of the relevant cell types, with acholinergic interneuron synapsing onto aglutamatergic motor neuron, which in turnforms a neuromuscular junction on a bodywallmuscle. Each of the lower panels illustratesthe effects of suppressing either electrical orsynaptic activity by manipulating expression ofthe gene indicated on the left. The cell typesexpressing the gene are shown in blue.Electrical silencing is indicated by a crossedout lightning bolt; synaptic silencing isindicated by blue crosses. The morphologicaland physiological changes observed in eachcase are shown on the right.

Current Biology

Pan-neuronalelectrical silencingby over-expressionof EKO in neurons

Synaptic silencingat NMJ byoverexpression ofTeTxLC inmotor neurons

Hypoinner-vation ofmotor neuronby interneurons

Ectopicinnervationof muscles

Unmanipulated ManipulatedElectrically silenced

Synaptically silenced

Inter-neuron

Motorneuron

MuscleManipulation Synapticchange

Postsynapticelectrical silencingat NMJ byoverexpressionof KIR in muscles

EnhancedNT releasefrommotor neurons

None None

[77,78]. Associative learning is usually shown by pairing anaversive foot shock stimulus to an otherwise benignodorant such as octanol or benzaldehyde. Trained flies fleefrom the odorant, and retain this associative memory fordays. Although it is well established that the mushroom

bodies are essential for associative learning, it has remaineduncertain whether the structures are involved in the estab-lishment, encoding or retrieval of memories.

The Tully and Davis laboratories [2,76] have addressedthis problem by expressing the Shits1 protein in the mush-room bodies of adult flies. This makes it possible todisrupt chemical synaptic connections within these bodiesduring specific phases of the learning process (Figure 6b,c).For example, one can train flies with synaptically silencedmushroom bodies, allow the animals to recover, and thentest for retained memories. Since this manipulation doesnot affect the mushroom bodies’ afferent inputs, it is pos-sible to test whether memory establishment depends onlocal circuitry within the bodies, perhaps through reverber-ating neural feedback circuitry.

Both studies obtained the same, clear-cut result. Despitethe synaptic-silencing within the mushroom bodies, theflies managed to learn. Upon recovery, the associationbetween the odorant and the aversive stimulus was fullyestablished, and the animals’ performance was as good asthat of control flies. In contrast, the ability to recall amemory remained blocked as long as the intrinsic mush-room body circuitry was synaptically silenced. Thus,chemical synaptic circuitry within the mushroom bodies isnot essential for the establishment of an associative memory,but is required for its expression.

The most parsimonious explanation for this result is thatthe plasticity associated with training occurs at the connec-tion between the afferent inputs, which remain functionaland able to release transmitter, and their target cells withinthe mushroom body. Perhaps Hebbian-like electrical activ-ity in the excited mushroom body dendrites, immediately


Figure 6

Spatial and temporal dissection of associative learning in Drosophila.(a) Schematic representation of the Drosophila mushroom body (MB)circuitry. Olfactory information is relayed through the glomeruli in eachantennal lobe (AL) to MB neurons located in the calyx, while sensoryinformation representing the foot-shock enters by an unknown pathway.MB neurons send axons through the MB lobes to other brain regions,which coordinate motor responses and generate odor avoidancebehavior. While foot shock and olfactory cues can elicit behaviors bypathways independent of the MB (hatched arrows), the MB isnecessary for the two sensory inputs to become associated togenerate odor avoidance. (b) MB-specific expression of the Shits1

gene can be used to block synaptic activity of MB neurons (bluecrosses) at the restrictive (lower panel, blue circuits), but not at thepermissive temperature (upper panel, red circuits) during variousphases of the learning process. (c) Blocking synaptic activity of MBneurons during the training phase (upper trace) or between the trainingand testing phases (middle trace) had no effect on either memoryacquisition or memory consolidation. However, blocking synapticactivity during the testing phase (lower trace) eliminated the ability toretrieve the memory.

Odor

Foot shock

Behaviour

MB

AL

Training Testing

Temperature shift

(a)

(b)

(c)

T=30oC(restrictive)

T=20oC(permissive)

Associativememoryreadout


?


Current Biology

Pairedinput

Pairedinput

20oC

30oC

20oC

30oC

20oC

30oC

postsynaptic to the afferents, is key to associative learning.Evidently, putative reverberatory or feedback circuits withinthe mushroom bodies that depend on chemical synaptictransmission can remain silent and the animals will stilllearn. In contrast, the results indicate that the retrieval ofmemory depends on mushroom body circuits or functionalsynaptic transmission by mushroom body efferents. While itremains unclear whether the memory ‘engram’ actuallyresides within the mushroom body, it is clear that mush-room body function is essential for evoking it.

Directions for the futureThe above studies on the Drosophila mushroom bodyprovide a unique advance in our understanding of howmemories are established and retrieved. Never before hasit been possible to investigate with such spatial and tem-poral resolution how a memory is accessed. These studies,together with the others described here, show how thetechniques presented in this review promise to inform ourunderstanding of how the brain works. But what we havedescribed is hopefully only a beginning. We close by men-tioning some of the tools still required for further progress.

First, tools for manipulating cellular excitability ontimescales similar to those attainable for synaptic transmis-sion with the Shits1 technique are necessary if we are tounderstand how neurons process and encode information.This is because electrical activity, and not synaptic activ-ity, governs the processes of integration and signal encod-ing in neurons. The targeted expression of channels whoseactivity can be directly manipulated by pharmacologicalmeans is one possible approach to this goal [79].

Second, the development of tools that can be used in ver-tebrate genetic model systems such as zebrafish and miceare badly needed. As techniques for inducible, cell-typespecific expression advance in these organisms it will bedesirable to have tools available for the reliable suppres-sion of both synaptic and electrical activity. A Ca2+-acti-vated K+ channel has been shown to partially inhibitactivity in vivo in mice [31]; other K+ channels with greaterpotential for achieving suppression have been used in cul-tured mammalian muscles and neurons [14,80,81]. Furtherwork will have to determine whether these channels canbe successfully applied to the in vivo suppression of activity.

Finally, we have described here only techniques for sup-pressing neural activity. General techniques for enhancingactivity will also be of considerable value in deter-mining the function of neurons. While dominant negativeapproaches to knocking out individual K+ channels haveproved useful in particular cases [82,83], a promising generalapproach has recently been described in mice. Kearneyet al. [84] have introduced mutations into the rat brain IIANa+ channel to slow its inactivation, and have shown that

pan-neuronal expression of this channel leads to hyperex-citability and seizures in transgenic mice.

Clearly there is much yet to do, both in creating new toolsand in applying the ones we have. What is certain is thatthe tools already in hand powerfully augment our ability toprobe nervous system function, and that they point to apromising future. It is perhaps not too soon for the physio-logically minded neuroscientist to begin to dream of the daywhen genetic switches will replace the knobs of a stimula-tor, and all the manipulations that are now readily per-formed on nerve cells in isolation will be performed ongroups of neurons in living, behaving animals.

References1. Kaneko M, Park JH, Cheng Y, Hardin PE, Hall JC: Disruption of

synaptic transmission or clock-gene-product oscillations incircadian pacemaker cells of Drosophila cause abnormalbehavioral rhythms. J Neurobiol 2000, 43:207-233.

2. Dubnau J, Grady L, Kitamoto T, Tully T: Disruption ofneurotransmission in Drosophila mushroom body blocks retrievalbut not acquisition of memory. Nature 2001, 411:476-480.

3. White BH, Osterwalder TP, Yoon KS, Joiner WJ, Whim MD,Kaczmarek LK, Keshishian H: Targeted attenuation of electricalactivity in Drosophila using a genetically modified K++ channel.Neuron 2001, 31:699-711.

4. Osterwalder TP, Yoon KS, White BH, Keshishian H: A conditionaltissue-specific transgene expression system using inducibleGAL4. Proc Natl Acad Sci USA 2001, 98:12596-12601.

5. Saez E, No D, West A, Evans RM: Inducible gene expression inmammalian cells and transgenic mice. Curr Opin Biotechnol 1997,8:608-616.

6. Clackson T: Controlling mammalian gene expression with smallmolecules. Curr Opin Chem Biol 1997, 1:210-218.

7. Cheng Q, Kulli JC, Yang J: Suppression of neuronal hyperexcitabilityand associated delayed neuronal death by adenoviral expressionof GABA(C) receptors. J Neurosci 2001, 21:3419-3428.

8. Griffith LC: Potassium channels: the importance of transportsignals. Curr Biol 2001, 11:R226-228.

9. Kaang BK, Pfaffinger PJ, Grant SG, Kandel ER, Furukawa Y:Overexpression of an Aplysia shaker K++ channel gene modifiesthe electrical properties and synaptic efficacy of identified Aplysianeurons. Proc Natl Acad Sci USA 1992, 89:1133-1137.

10. Zhao B, Rassendren F, Kaang BK, Furukawa Y, Kubo T, Kandel ER: Anew class of noninactivating K++ channels from aplysia capable ofcontributing to the resting potential and firing patterns of neurons.Neuron 1994, 13:1205-1213.

11. Ruppersberg JP: Intracellular regulation of inward rectifier K++

channels. Pflugers Arch 2000, 441:1-11.12. Ehrengruber MU, Doupnik CA, Xu Y, Garvey J, Jasek MC, Lester HA,

Davidson N: Activation of heteromeric G protein-gated inwardrectifier K++ channels overexpressed by adenovirus gene transferinhibits the excitability of hippocampal neurons. Proc Natl AcadSci USA 1997, 94:7070-7075.

13. Johns DC, Nuss HB, Chiamvimonvat N, Ramza BM, Marban E,Lawrence JH: Adenovirus-mediated expression of a voltage-gatedpotassium channel in vitro (rat cardiac myocytes) and in vivo (ratliver). A novel strategy for modifying excitability. J Clin Invest 1995,96:1152-1158.

14. Johns DC, Marx R, Mains RE, O’Rourke B, Marban E: Induciblegenetic suppression of neuronal excitability. J Neurosci 1999,19:1691-1697.

15. Johns DC, Marban E, Nuss HB: Virus-mediated modification ofcellular excitability. Ann New York Acad Sci 1999, 868:418-422.

16. Baines RA, Uhler JP, Thompson A, Sweeney ST, Bate M: Alteredelectrical properties in Drosophila neurons developing withoutsynaptic transmission. J Neurosci 2001, 21:1523-1531.

17. Paradis S, Sweeney ST, Davis GW: Homeostatic control ofpresynaptic release is triggered by postsynaptic membranedepolarization. Neuron 2001, 30:737-749.


18. Hardie RC, Raghu P, Moore S, Juusola M, Baines RA, Sweeney ST:Calcium influx via TRP channels is required to maintain PIP2 levelsin Drosophila photoreceptors. Neuron 2001, 30:149-159.

19. Nadeau H, McKinney S, Anderson DJ, Lester HA: ROMK1 (Kir1.1)causes apoptosis and chronic silencing of hippocampal neurons.J Neurophysiol 2000, 84:1062-1075.

20. Sutherland ML, Williams SH, Abedi R, Overbeek PA, Pfaffinger PJ,Noebels JL: Overexpression of a Shaker-type potassium channelin mammalian central nervous system dysregulates nativepotassium channel gene expression. Proc Natl Acad Sci USA1999, 96:2451-2455.

21. Peckol EL, Zallen JA, Yarrow JC, Bargmann CI: Sensory activityaffects sensory axon development in C. elegans. Development1999, 126:1891-1902.

22. Zhao H, Nonet ML: A retrograde signal is involved in activity-dependent remodeling at a C. elegans neuromuscular junction.Development 2000, 127:1253-1266.

23. Papazian DM, Bezanilla F: Voltage-dependent activation of ionchannels. Adv Neurol 1999, 79:481-491.

24. Johnstone DB, Wei A, Butler A, Salkoff L, Thomas JH: Behavioraldefects in C. elegans egl-36 mutants result from potassiumchannels shifted in voltage-dependence of activation. Neuron1997, 19:151-164.

25. Weinshenker D, Wei A, Salkoff L, Thomas JH: Block of an ether-a-go-go-like K++ channel by imipramine rescues egl-2 excitationdefects in Caenorhabditis elegans. J Neurosci 1999, 19:9831-9840.

26. Papazian DM, Timpe LC, Jan YN, Jan LY: Alteration of voltage-dependence of Shaker potassium channel by mutations in the S4sequence. Nature 1991, 349:305-310.

27. Hoshi T, Zagotta WN, Aldrich RW: Biophysical and molecularmechanisms of Shaker potassium channel inactivation. Science1990, 250:533-538.

28. Zagotta WN, Germeraad S, Garber SS, Hoshi T, Aldrich RW:Properties of ShB A-type potassium channels expressed inShaker mutant Drosophila by germline transformation. Neuron1989, 3:773-782.

29. Zhao ML, Sable EO, Iverson LE, Wu CF: Functional expression ofShaker K++ channels in cultured Drosophila ‘giant’; neuronsderived from Sh cDNA transformants: distinct properties,distribution, and turnover. J Neurosci 1995, 15:1406-1418.

30. Han JH, Yim SW, Lim CS, Park CW, Kaang BK: Expression of anon-inactivating K++ channel driven by a rat heat shock promoterincreased the resting potential in Aplysia silent neurons. NeurosciRes 1999, 34:13-19.

31. Bond CT, Sprengel R, Bissonnette JM, Kaufmann WA, Pribnow D,Neelands T, Storck T, Baetscher M, Jerecic J, Maylie J et al.:Respiration and parturition affected by conditional overexpressionof the Ca2++-activated K++ channel subunit, SK3. Science 2000,289:1942-1946.

32. Rohacs T, Chen J, Prestwich GD, Logothetis DE: Distinctspecificities of inwardly rectifying K++ channels forphosphoinositides. J Biol Chem 1999, 274:36065-36072.

33. Fernandez-Chacon R, Sudhof TC: Genetics of synaptic vesiclefunction: toward the complete functional anatomy of an organelle.Annu Rev Physiol 1999, 61:753-776.

34. Cremona O, De Camilli P: Synaptic vesicle endocytosis. Curr OpinNeurobiol 1997, 7:323-330.

35. Pellizzari R, Rossetto O, Schiavo G, Montecucco C: Tetanus andbotulinum neurotoxins: mechanism of action and therapeuticuses. Philos Trans R Soc Lond B Biol Sci 1999, 354:259-268.

36. Mochida S, Poulain B, Eisel U, Binz T, Kurazono H, Niemann H,Tauc L: Exogenous mRNA encoding tetanus or botulinumneurotoxins expressed in Aplysia neurons. Proc Natl Acad SciUSA 1990, 87:7844-7848.

37. Sweeney ST, Broadie K, Keane J, Niemann H, O’Kane CJ: Targetedexpression of tetanus toxin light chain in Drosophila specificallyeliminates synaptic transmission and causes behavioral defects.Neuron 1995, 14:341-351.

38. Heimbeck G, Bugnon V, Gendre N, Haberlin C, Stocker RF: Smelland taste perception in Drosophila melanogaster larva: toxinexpression studies in chemosensory neurons. J Neurosci 1999,19:6599-6609.

39. Martin JR, Ernst R, Heisenberg M: Mushroom bodies suppresslocomotor activity in Drosophila melanogaster. Learn Mem 1998,5:179-191.

40. Hiesinger PR, Reiter C, Schau H, Fischbach KF: Neuropil patternformation and regulation of cell adhesion molecules in Drosophilaoptic lobe development depend on synaptobrevin. J Neurosci1999, 19:7548-7556.

41. Allen MJ, Shan X, Caruccio P, Froggett SJ, Moffat KG, Murphey RK:Targeted expression of truncated glued disrupts giant fibersynapse formation in Drosophila. J Neurosci 1999, 19:9374-9384.

42. Baines RA, Robinson SG, Fujioka M, Jaynes JB, Bate M: Postsynapticexpression of tetanus toxin light chain blocks synaptogenesis inDrosophila. Curr Biol 1999, 9:1267-1270.

43. Martin JR, Raabe T, Heisenberg M: Central complex substructuresare required for the maintenance of locomotor activity inDrosophila melanogaster. J Comp Physiol [A] 1999, 185:277-288.

44. Li H, Chaney S, Roberts IJ, Forte M, Hirsh J: Ectopic G-proteinexpression in dopamine and serotonin neurons blocks cocainesensitization in Drosophila melanogaster. Curr Biol 2000,10:211-214.

45. Blanchardon E, Grima B, Klarsfeld A, Chelot E, Hardin PE, Preat T,Rouyer F: Defining the role of Drosophila lateral neurons in thecontrol of circadian rhythms in motor activity and eclosion bytargeted genetic ablation and PERIOD protein overexpression. EurJ Neurosci 2001, 13:871-888.

46. Keller A, Sweeney ST, Zars T, O’Kane CJ, Heisenberg M: Targetedexpression of tetanus neurotoxin interferes with behavioralresponses to sensory input in Drosophila. J Neurobiol, in press.

47. Wu MN, Littleton JT, Bhat MA, Prokop A, Bellen HJ: ROP, theDrosophila Sec1 homolog, interacts with syntaxin and regulatesneurotransmitter release in a dosage-dependent manner. EMBO J1998, 17:127-139.

48. Kitamoto T: Conditional modification of behavior in Drosophila bytargeted expression of a temperature-sensitive shibire allele indefined neurons. J Neurobiol 2001, 47:81-92.

49. Poodry CA, Hall L, Suzuki DT: Developmental properties of Shibire:a pleiotropic mutation affecting larval and adult locomotion anddevelopment. Dev Biol 1973, 32:373-386.

50. Poodry CA, Edgar L: Reversible alteration in the neuromuscularjunctions of Drosophila melanogaster bearing a temperature-sensitive mutation, shibire. J Cell Biol 1979, 81:520-527.

51. Poodry CA: shibire, a neurogenic mutant of Drosophila. Dev Biol1990, 138:464-472.

52. Koenig JH, Ikeda K: Disappearance and reformation of synapticvesicle membrane upon transmitter release observed underreversible blockage of membrane retrieval. J Neurosci 1989,9:3844-3860.

53. Kawasaki F, Hazen M, Ordway RW: Fast synaptic fatigue in shibiremutants reveals a rapid requirement for dynamin in synapticvesicle membrane trafficking. Nat Neurosci 2000, 3:859-860.

54. Chen MS, Obar RA, Schroeder CC, Austin TW, Poodry CA,Wadsworth SC, Vallee RB: Multiple forms of dynamin are encodedby shibire, a Drosophila gene involved in endocytosis. Nature1991, 351:583-586.

55. Ramaswami M, Rao S, van der Bliek A, Kelly RB, Krishnan KS:Genetic studies on dynamin function in Drosophila. J Neurogenet1993, 9:73-87.

56. Seugnet L, Simpson P, Haenlin M: Requirement for dynamin duringNotch signaling in Drosophila neurogenesis. Dev Biol 1997,192:585-598.

57. Staples RR, Ramaswami M: Functional analysis of dynaminisoforms in Drosophila melanogaster. J Neurogenet 1999,13:119-143.

58. McMahon HT, Ushkaryov YA, Edelmann L, Link E, Binz T, Niemann H,Jahn R, Sudhof TC: Cellubrevin is a ubiquitous tetanus-toxinsubstrate homologous to a putative synaptic vesicle fusionprotein. Nature 1993, 364:346-349.

59. Brand AH, Perrimon N: Targeted gene expression as a means ofaltering cell fates and generating dominant phenotypes.Development 1993, 118:401-415.

60. Roman G, Endo K, Zong L, Davis RL: P{Switch}, a novel system forspatial and temporal control of gene expression in Drosophilamelanogaster. Proc Natl Acad Sci USA 2001, 98:12602-12607.

61. Yuste R, Nelson DA, Rubin WW, Katz LC: Neuronal domains indeveloping neocortex: mechanisms of coactivation. Neuron 1995,14:7-17.

62. Aamodt SM, Constantine-Paton M: The role of neural activity insynaptic development and its implications for adult brain function.Adv Neurol 1999, 79:133-144.


63. Budnik V, Gramates LS: Neuromuscular Junctions in Drosophila. NewYork: Academic Press; 1999.

64. Jarecki J, Keshishian H: Role of neural activity duringsynaptogenesis in Drosophila. J Neurosci 1995, 15:8177-8190.

65. Chang TN, Keshishian H: Laser ablation of Drosophila embryonicmotoneurons causes ectopic innervation of target muscle fibers.J Neurosci 1996, 16:5715-5726.

66. Budnik V, Zhong Y, Wu CF: Morphological plasticity of motor axonsin Drosophila mutants with altered excitability. J Neurosci 1990,10:3754-3768.

67. Zhong Y, Budnik V, Wu CF: Synaptic plasticity in Drosophilamemory and hyperexcitable mutants: role of cAMP cascade.J Neurosci 1992, 12:644-651.

68. Zhong Y, Wu CF: Differential modulation of potassium currents bycAMP and its long-term and short-term effects: dunce andrutabaga mutants of Drosophila. J Neurogenet 1993, 9:15-27.

69. Davis GW, Goodman CS: Synapse-specific control of synapticefficacy at the terminals of a single neuron. Nature 1998,392:82-86.

70. Davis GW, DiAntonio A, Petersen SA, Goodman CS: PostsynapticPKA controls quantal size and reveals a retrograde signal thatregulates presynaptic transmitter release in Drosophila. Neuron1998, 20:305-315.

71. DiAntonio A, Petersen SA, Heckmann M, Goodman CS: Glutamatereceptor expression regulates quantal size and quantal content at theDrosophila neuromuscular junction. J Neurosci 1999, 19:3023-3032.

72. Petersen SA, Fetter RD, Noordermeer JN, Goodman CS, DiAntonio A:Genetic analysis of glutamate receptors in Drosophila reveals aretrograde signal regulating presynaptic transmitter release.Neuron 1997, 19:1237-1248.

73. Broadie K, Bate M: Activity-dependent development of theneuromuscular synapse during Drosophila embryogenesis.Neuron 1993, 11:607-619.

74. Waddell S, Quinn WG: Flies, genes, and learning. Annu RevNeurosci 2001, 24:1283-1309.

75. Waddell S, Armstrong JD, Kitamoto T, Kaiser K, Quinn WG: Theamnesiac gene product is expressed in two neurons in theDrosophila brain that are critical for memory. Cell 2000,103:805-813.

76. McGuire SE, Le PT, Davis RL: The role of Drosophila mushroombody signaling in olfactory memory. Science 2001, 293:1330-1333.

77. Heisenberg M, Borst A, Wagner S, Byers D: Drosophila mushroombody mutants are deficient in olfactory learning. J Neurogenet1985, 2:1-30.

78. de Belle JS, Heisenberg M: Associative odor learning in Drosophilaabolished by chemical ablation of mushroom bodies. Science1994, 263:692-695.

79. Slimko E, Anderson D, Lester H, Davidson N: Selective silencing ofmammalian neurons: strategies using chloride channels. SocNeurosci Abstracts 2000, 30:710-719.

80. Holt JR, Johns DC, Wang S, Chen ZY, Dunn RJ, Marban E, Corey DP:Functional expression of exogenous proteins in mammaliansensory hair cells infected with adenoviral vectors. J Neurophysiol1999, 81:1881-1888.

81. Falk T, Kilani RK, Yool AJ, Sherman SJ: Viral vector-mediatedexpression of K++ channels regulates electrical excitability inskeletal muscle. Gene Ther 2001, 8:1372-1379.

82. Johns DC, Nuss HB, Marban E: Suppression of neuronal andcardiac transient outward currents by viral gene transfer ofdominant-negative Kv4.2 constructs. J Biol Chem 1997,272:31598-31603.

83. Shibata R, Nakahira K, Shibasaki K, Wakazono Y, Imoto K, Ikenaka K:A-type K++ current mediated by the Kv4 channel regulates thegeneration of action potential in developing cerebellar granulecells. J Neurosci 2000, 20:4145-4155.

84. Kearney JA, Plummer NW, Smith MR, Kapur J, Cummins TR, WaxmanSG, Goldin AL, Meisler MH: A gain-of-function mutation in thesodium channel gene Scn2a results in seizures and behavioralabnormalities. Neuroscience 2001, 102:307-317.

85. Hoppe UC, Marban E, Johns DC: Distinct gene-specificmechanisms of arrhythmia revealed by cardiac gene transfer oftwo long QT disease genes, HERG and KCNE1. Proc Natl Acad SciUSA 2001, 98:5335-5340.

86. Jones SM, Ribera AB: Overexpression of a potassium channelgene perturbs neural differentiation. J Neurosci 1994,14:2789-2799.

87. Jones SM, Hofmann AD, Lieber JL, Ribera AB: Overexpression ofpotassium channel RNA: in vivo development rescues neuronsfrom suppression of morphological differentiation in vitro.J Neurosci 1995, 15:2867-2874.

88. Reddy S, Jin P, Trimarchi J, Caruccio P, Phillis R, Murphey RK: Mutantmolecular motors disrupt neural circuits in Drosophila. J Neurobiol1997, 33:711-723.

89. Tabuchi K, Sawamoto K, Suzuki E, Ozaki K, Sone M, Hama C,Tanifuji-Morimoto T, Yuasa Y, Yoshihara Y et al.: GAL4/UAS-WGAsystem as a powerful tool for tracing Drosophila transsynapticneural pathways. J Neurosci Res 2000, 59:94-99.


molecular genetic approaches to the targeted suppression of neuronal activity

Documents