Electrical Coupling between Olfactory GlomeruliIntroductionResultsOptogenetic Stimulation of LNs Produces Mixed Excitation-Inhibition in PNsIdentifying eLNsOdor Responses of eLNs Can Account for the Properties of Lateral ExcitationeLNs Make Electrical Synapses onto PNsGenetic Elimination of Synapses from eLNs onto PNseLNs Receive Cholinergic Excitation from PNseLNs Make Mixed Synapses onto iLNsPN-PN Interactions Require Electrical SynapsesGenetic Elimination of Odor-Evoked Lateral ExcitationEliminating Lateral Excitation Can Reduce PN Odor ResponsesEliminating PN-PN Interactions Can Reduce PN Odor ResponsesPerturbing Electrical Networks Can Increase PN Odor Responses by Reducing InhibitionDiscussionTarget-Cell-Specific Properties of eLN SynapsesFunctional Implications of Electrical SynapsesProperties and Functions of Odor-Evoked Lateral ExcitationInteractions among LNsA Challenge for Understanding Neural CircuitsExperimental ProceduresFly StocksElectrophysiologyImmunohistochemistryOlfactory StimulationOptogenetic StimulationDual Whole-Cell RecordingsData AnalysisSupplemental InformationAcknowledgmentsReferencesNeuronArticleElectrical Coupling between Olfactory GlomeruliEmre Yaksi1and Rachel I. Wilson1,*1Department of Neurobiology, Harvard Medical School, 220 Longwood Avenue, Boston MA, 02115, USA*Correspondence: [email protected] 10.1016/j.neuron.2010.08.041SUMMARYIn the Drosophila antennal lobe, excitation canspread between glomerular processing channels. Inthis study, we investigated the mechanism of lateralexcitation. Dual recordings from excitatory localneurons (eLNs) and projection neurons (PNs) showedthat eLN-to-PN synapses transmit both hyperpolar-ization and depolarization, are not diminished byblocking chemical neurotransmission, and are abol-ished by a gap-junction mutation . This mutation elim-inates odor-evoked lateral excitation in PNs anddiminishes some PN odor responses. This impliesthat lateral excitation is mediated by electricalsynapses from eLNs onto PNs. In addition, eLNsform synapses onto inhibitory LNs. Eliminating thesesynapses boosts some PN odor responses andreduces the disinhibitory effect of GABA receptorantagonists on PNs. Thus, eLNs have two opposingeffects on PNs, driving both direct excitation andindirect inhibition. We propose that when stimuliare weak, lateral excitation promotes sensitivity,whereas when stimuli are strong, lateral excitationhelps recruit inhibitory gain control.INTRODUCTIONSensory neurons are generally selective for particular stimulusfeatures. Neurons at the same level of sensory processing thatare tuned to different features can be thought of as representingdifferent ‘‘processing channels.’’ One of the aims of sensoryneuroscience is to understand the mechanisms and functionsof crosstalk between such channels.The notion of a sensory channel is particularly well-defined inearly olfactory processing. This is because each glomerulus inthe olfactory bulb or antennal lobe defines both an anatomicalmodule and a discrete feedforward circuit. Each olfactoryreceptor neuron (ORN) is presynaptic to a single glomerulus, andeach second-order neuron is postsynaptic to a single glomerulus(Bargmann, 2006). Functional connections between processingchannels were until recently thought to be mainly inhibitory,with little or no spread of excitation between principal neuronsin different glomeruli (Lledo et al., 2005; Schoppa and Urban,2003; but see Laurent et al., 2001). Recently, however, severalstudies in the Drosophila antennal lobe demonstrated the exis-tence of excitatory connections between second-order neuronsin different glomeruli (Olsen et al., 2007; Root et al., 2007; Shanget al., 2007). These studies found that when the ORN inputs toa second-order neuron were silenced, that neuron still receivedindirect odor-evoked excitation from other glomeruli (which wedefine here as ‘‘lateral excitation’’). These studies proposedthat lateral excitation was mediated by local neurons (LNs)that extend dendrites into many glomeruli and form dendroden-dritic synapses with second-order neurons (termed projectionneurons, or PNs).Both the mechanism and the function of lateral excitation areuncertain. It has been suggested that its function might be toboost responses to weak stimuli (Olsen et al., 2007; Shanget al., 2007), but this could not be directly tested because therewas no known way to abolish lateral excitation. It was alsoproposed that lateral excitation is mediated by the release ofacetylcholine from LNs (Shang et al., 2007 ). However, this couldnot be directly tested because cholinergic antagonists block thetransmission of all olfactory signals to the brain; this is due to thefact that ORNs are themselves cholinergic (Kazama and Wilson,2008). Moreover, lateral excitation is recruited very rapidly afterORN signals reach the brain, with a delay almost too short fora disynaptic connection (1.5 ms; Kazama and Wilson, 2008).This rapid recruitment of lateral excitation suggests that theunderlying mechanism might be unusual.In this study, we had two broad aims. Our first aim was todetermine the synaptic mechanisms responsible for the spreadof excitation between glomeruli. Our second aim was to discoverhow eliminating these mechanisms alters the output of theantennal lobe in response to olfactory stimuli.RESULTSOptogenetic Stimulation of LNs Produces MixedExcitation-Inhibition in PNsWhen direct ORN input to a PN is silenced, olfactory stimuli canstill elicit an excitatory response (Olsen et al., 2007; Root et al.,2007; Shang et al., 2007). This is thought to reflect the action ofexcitatory LNs (eLNs). However, it has not been directly demon-strated that any LNs actually have excitatory effects on otherneurons. Several studies have noted the existence of GABA-immunonegative LNs (Chou et al., 2010; Shang et al., 2007;Wilson and Laurent, 2005), which are potential candidates forexcitatory LNs. Among the Gal4 lines that reportedly driveexpression in LNs, none is specific to GABA-negative LNs, butabout half of the LNs labeled by krasavietz-Gal4 are GABA nega-tive (58%–61%; Chou et al., 2010; Shang et al., 2007), makingit a useful starting point. Shang et al. (2007) reported thatGABA-negative krasavietz LNs are immunopositive for choline1034 Neuron 67, 1034–1047, September 23, 2010 ª2010 Elsevier Inc.acetyltransferase (Cha), although a later study reported that notall are Cha-positive (Chou et al., 2010), casting some doubt