Progress in Neurobiology 95 (2011) 427–447

The neurobiology of insect olfaction: Sensory processing in a comparative context

Joshua P. Martin *, Aaron Beyerlein, Andrew M. Dacks 1, Carolina E. Reisenman,Jeffrey A. Riffell 2, Hong Lei, John G. Hildebrand

Department of Neuroscience, College of Science, University of Arizona, 1040 East Fourth Street, Tucson, AZ 85721-0077, USA

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428

2. Olfactory world of an insect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429

2.1. Chemical composition of natural olfactory stimuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429

2.2. Spatio-temporal structure of olfactory stimuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429

3. Olfaction-based behavior of insects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430

4. Evolution and speciation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431

4.1. Evolution of olfactory receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431

4.2. Olfactory specialization and speciation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431

4.2.1. Peripheral adaptations for olfactory specialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432

4.2.2. Central adaptations for olfactory specialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432

5. Central olfactory pathways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432

5.1. Antennal lobe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432

5.1.1. Inhomogeneous interactions between glomeruli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433

5.2. Higher-order olfactory centers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

A R T I C L E I N F O

Article history:

Received 5 June 2011

Received in revised form 10 September 2011

Accepted 19 September 2011

Available online 24 September 2011

Keywords:

Olfaction

Neuroethology

Comparative anatomy

Antennal lobe

Protocerebrum

Sensory coding

A B S T R A C T

The simplicity and accessibility of the olfactory systems of insects underlie a body of research essential to

understanding not only olfactory function but also general principles of sensory processing. As insect

olfactory neurobiology takes advantage of a variety of species separated by millions of years of evolution,

the field naturally has yielded some conflicting results. Far from impeding progress, the varieties of

insect olfactory systems reflect the various natural histories, adaptations to specific environments, and

the roles olfaction plays in the life of the species studied.

We review current findings in insect olfactory neurobiology, with special attention to differences

among species. We begin by describing the olfactory environments and olfactory-based behaviors of

insects, as these form the context in which neurobiological findings are interpreted. Next, we review

recent work describing changes in olfactory systems as adaptations to new environments or behaviors

promoting speciation. We proceed to discuss variations on the basic anatomy of the antennal (olfactory)

lobe of the brain and higher-order olfactory centers. Finally, we describe features of olfactory

information processing including gain control, transformation between input and output by operations

such as broadening and sharpening of tuning curves, the role of spiking synchrony in the antennal lobe,

and the encoding of temporal features of encounters with an odor plume. In each section, we draw

connections between particular features of the olfactory neurobiology of a species and the animal’s life

history. We propose that this perspective is beneficial for insect olfactory neurobiology in particular and

sensory neurobiology in general.

� 2011 Elsevier Ltd. All rights reserved.

Abbreviations: OR, olfactory receptor protein; ORC, olfactory receptor cell; Orco, olfactory coreceptor; AL, antennal lobe; LN, local interneuron; PN, projection neuron; OBP,

odorant-binding protein; MGC, macroglomerular complex; MB, mushroom body; LH, lateral horn of the protocerebrum; APT, antenno-protocerebral tract; MBC, mushroom

Contents lists available at SciVerse ScienceDirect

Progress in Neurobiology

jo u rn al ho m epag e: ww w.els evier . c om / lo cat e/pn eu ro b io

body calyx; LH, lateral horn; KC, kenyon cell; LAL, lateral accessory lobe; LFP, local field potential; AC, antennal commissure; ILPC, inferior lateral protocerebrum.

* Corresponding author. Tel.: +1 520 621 6643; fax: +1 520 621 8282.

E-mail address: jpmartin@email.arizona.edu (J.P. Martin).1 Present address: Department of Neuroscience, Mount Sinai School of Medicine, New York, NY 10029, USA.2 Present address: Department of Biology, University of Washington, Seattle, WA 98195, USA.

0301-0082/$ – see front matter � 2011 Elsevier Ltd. All rights reserved.

doi:10.1016/j.pneurobio.2011.09.007

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447428

5.2.1. Antenno-protocerebral tracts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

5.2.2. Mushroom-body calyx. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

5.2.3. Lateral horn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436

5.2.4. Protocerebrum and beyond. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437

6. Comparative coding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437

6.1. Gain control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437

6.2. Tuning/coding transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439

6.2.1. Sharpening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439

6.2.2. Broadening. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439

6.2.3. Sharpening versus broadening in two species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439

6.3. Synchrony . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440

6.4. Encoding temporal features of olfactory stimuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441

7. Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442

1. Introduction

Olfaction plays a central role in insect behaviors of interest tohumans. Feeding and reproduction of insect pests, pollinators, andvectors of disease are strongly regulated by the volatile organiccompounds (hereinafter ‘‘volatiles’’) an insect encounters. The tollof malaria and other diseases mediated by insect vectors, famineattributable to insect crop pests, and the alarming collapse of amajor pollinator species (honey bees) impels investigations intothe neurobiology underlying the sense of smell in insects (vanNaters and Carlson, 2006).

As a topic of study in neurobiology per se, olfaction shares thebenefits of all neurobiological research in insects and otherinvertebrates: reduced numerical complexity of the nervoussystem, the corresponding advantages of identifiable nerve cells(Comer and Robertson, 2001) and neuropil structures, and a strongconnection between neurobiology and behavior. The sharedprinciples of organization and function of olfactory systems ininvertebrates and vertebrates (Ache and Young, 2005; Hildebrandand Shepherd, 1997; Kaupp, 2010) make discoveries in insects ofgeneral interest to researchers studying other animals.

In this review, we argue that the diversity of insect speciesprovides another benefit for the study of olfactory neurobiology.Insects are believed to be the most speciose and diverse class ofanimals on earth: approximately one million species arecurrently described (and many more are expected to be found)in 30 orders, separated by more than 400 million years ofevolution (Grimaldi and Engel, 2005 – see Fig. 1), adapted to moreenvironments and niches, and correspondingly expressing morediverse behaviors, than any other closely related group ofanimals. Most insect behavior relies to some degree onchemosensory information. Research on the olfactory neurobiol-ogy of insects has focused on flies, cockroaches, honey bees,moths, and locusts, as well as other representative species from atleast seven orders spanning holometabola and hemimetabola(respectively, insect taxa that undergo complete or incompletemetamorphosis – see Fig. 1).

Converging evidence from multiple species maps out therelatively simple, elegant structure of a ‘canonical’ insect olfactorysystem. A single type of olfactory receptor protein (OR) withcharacteristic affinity for volatiles is expressed in a subset of theolfactory receptor cells (ORCs) housed in a head appendage (e.g.antenna or palp). The axons of all ORCs expressing the samereceptor converge in a single glomerulus within the antennal lobe(AL), processing and propagating olfactory information throughthe first level of processing in the CNS. Glomeruli interact througha population of local interneurons (LNs), many or most of whichare inhibitory, that shape the output of the AL conveyed byprojection neurons (PNs) to down-stream olfactory foci elsewhere

in the brain. At subsequent stages of neural processing,glomerular output is integrated, and olfactory information isformatted for memory and association with other modalities andultimately drives or modulates the activity of circuits that controlbehavior.

For nearly every one of these canonical principles of structureand function of insect olfactory systems, an exception can be foundamong insect species. Some ORCs express multiple receptors(Benton et al., 2009; Couto et al., 2005; Fishilevich and Vosshall,2005; Vosshall et al., 2000), and in a few species ORC axonsterminate within multiple, overlapping sets of glomeruli (Antonand Hansson, 1996). Glomeruli in the honey-bee AL interactthrough a network of histaminergic local neurons not found ininsects outside of the Hymenoptera (Dacks et al., 2010). Glomerulimay excite rather than inhibit their neighbors, partially decouplingthe output of the AL from the input patterns of ORCs (Olsen et al.,2007). Output from the AL to higher-order centers follows non-homologous pathways in different species, with an underlyingorganization that almost certainly has consequences for olfactoryprocessing (Galizia and Rossler, 2010).

Far from impeding investigations into the neurobiology ofolfaction, the diversity of insect models offers a suite of ‘‘naturalexperiments’’ wherein olfactory systems have evolved to performefficiently the tasks most crucial to the animal’s survival. Naturalbehavior reveals how an animal uses olfactory information: whatis emphasized, what is filtered out, and what is integrated orextracted from the background. Insects present perhaps thegreatest range of behaviors mediated by volatile chemicals.Included in this single class are species that use olfaction verylittle or perhaps not at all and have correspondingly reduced brainstructures (c.f. Strausfeld et al., 1998), species highly specialized ona plant or animal food source or host (c.f. Dekker et al., 2006), andspecies with remarkable abilities to associate new food resourceswith corresponding olfactory cues (c.f. Dukas, 2008; Giurfa, 2007).Correlating unique or exaggerated olfactory behavior with uniqueor exaggerated features of olfactory systems can be a powerful tool.Insects offer exceptionally fertile grounds for such a comparativeapproach.

In this review, we consider the remarkable progress of researchon the anatomy, physiology, and information processing in theolfactory systems of insects in a comparative context. We beginwith a discussion of the quality and temporal structure of olfactorystimuli important to various insects (Section 2). We survey theolfaction-based behaviors exhibited by insects in their naturalenvironments, especially those that demonstrate the challengesfaced by an olfactory system (Section 3). Next we examine recentevidence of how the OR repertoire is shaped by both the salience ofvolatiles in an insect’s ecological niche and the roles of thosecompounds in the insect’s evolutionary history (Section 4).

Fig. 1. The phylogeny of the neopteran insects (after Grimaldi and Engel, 2005), highlighting orders most commonly studied in olfactory neurobiology. The four orders most

commonly used in studies of central olfactory neurobiology (Orthoptera, Hymenoptera, Diptera, and Lepidoptera) are highlighted in large font, and the lineage connecting

them is bolded in the phylogenetic tree. Three additional orders, less commonly studied but nonetheless important, are highlighted in bold (Blattaria, Hemiptera, and

Coleoptera). Numbers along the timeline indicate deep time events for reference (Grimaldi and Engel, 2005; Williams et al., 2010). Model animals from each of the four most

commonly used orders are depicted in the inset photos: (A) S. americana (‘‘locust’’) (B) A. mellifera (‘‘honey-bee’’) (C) D. melanogaster (‘‘fruit fly’’) and (D) M. sexta

(‘‘hawkmoth’’). Photos: Charles Hedgcock.

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447 429

Information is passed and processed from ORCs, through theprimary olfactory center (the AL), to higher-order foci in theprotocerebrum. We highlight idiosyncrasies in structure andfunction of these areas and the insights about olfactory processingthey provide (Section 5). In the final section (Section 6), we suggestthat insects employ neural codes for olfactory stimuli andprocessing mechanisms that reflect the demands presented bothby their environment and by the structure and function of theneural systems that produce those codes.

2. Olfactory world of an insect

Natural, behaviorally significant olfactory stimuli typically aremixtures of volatiles whose concentrations co-vary dynamically intime and space. The volatiles released by conspecifics, foodsources, and oviposition hosts are chemically diverse, andmixtures often contain compounds unique to their source, as wellas volatiles common to the chemical signature of multiple sources(c.f. Bruce et al., 2005; Raguso, 2008). The olfactory world is also anarena of constant movement and flux. Once emitted by a source,volatiles are dispersed, mixed, and diluted by the ambient motionof air to form a shifting and filamentous plume. In this section, wereview new insights into the physico-chemical properties ofselected volatiles that have known roles in insect behavior.Identification and description of behaviorally relevant volatiles asthey are received by insects establish additional parameters forinvestigation of olfactory neurobiology.

2.1. Chemical composition of natural olfactory stimuli

Contemporary analytical technology has revealed considerablechemical diversity in the olfactory world of insects. Analytical toolsare particularly powerful when directly combined with neuro-physiological (e.g. Ghaninia et al., 2008) or behavioral (e.g.Allmann and Baldwin, 2010; Turlings et al., 2004) assays toidentify effective volatiles in a complex mixture released by asource. These studies also reveal commonality among olfactorystimuli. For example, olfactory systems of phytophagous insectsacross many orders respond to many of the same, ubiquitous plantvolatiles despite biases toward different hosts, suggesting thatcomplex, higher-order properties of olfactory stimuli are necessaryfor volatile source identification (Bruce et al., 2005; Raguso, 2008).

2.2. Spatio-temporal structure of olfactory stimuli

Immediately following their emission from a source, volatilesbecome subject to the physical forces of moving air that imposestructure on them. At spatial scales greater than 1 cm the olfactoryenvironment is dynamic. Both fluid-dynamic analysis andanalytic technologies have shown that a plume of volatiles isnot a uniform concentration gradient but rather a filamentous anddiscontinuous structure containing regions of volatile-laden airinterspersed with regions of volatile-free air (Murlis and Jones,1981; Riffell et al., 2008a). Insect olfactory systems apparentlyhave evolved to process the resulting intermittent olfactory

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447430

stimuli, and similarities in behavior and neural processing arenoteworthy in phylogenetically distant species (Budick andDickinson, 2006; Carde and Willis, 2008).

The ability of many insects to track a discontinuous olfactorystimulus and locate its unseen source is based on a common searchstrategy comprising alternating crosswind ‘casts’ to locatevolatiles, and upwind ‘surges’ following contact with volatiles(Murlis et al., 1992). In searching for the source of behaviorallysignificant volatiles, moths and flies use this strategy overlandscape scales (Reynolds and Frye, 2007; Reynolds et al.,2007), and their flight paths resemble a mathematically optimalsearch strategy (Reynolds, 2005) for following a plume of volatiles.Insects rely on the information contained in plume structure tovarying degrees when searching for a source: in an artificiallyhomogeneous plume that lacks spatio-temporal structure, fliesnavigate upwind to the source (Budick and Dickinson, 2006), butmany moths are unsuccessful (Willis and Baker, 1984). In addition,the temporal patterns in which insects encounter volatiles areaffected by the size and velocity of the animal (Koehl, 2001, 2006),active interaction with the plume by wing beating (Loudon andKoehl, 2000; Sane, 2006; Sane and Jacobson, 2006), and antennalflicking (Hillier and Vickers, 2004; Nishiyama et al., 2007; see alsoDethier, 1987). Olfactory stimuli that are important for insects thushave chemical and temporal characteristics particular to eachspecies. Investigations in olfactory neurobiology should considerthese features in designing and interpreting experiments.

3. Olfaction-based behavior of insects

A goal of efforts to understand the olfactory neurobiology ofinsects is to analyze neural mechanisms that underlie olfaction-modulated behavior. In addition, a clear description of theproblems an olfactory system must solve in the natural worldprovides the parameters for investigations into neural processingof olfactory information. In general, olfaction-based behaviorcritical for survival and reproduction first involves recognition(either innate or learned) of a significant olfactory stimulus.Subsequent processing takes into account the context of thesensation, both the external environment and the internal state ofthe animal.

Much has been gained by studying the innate responses ofinsects to olfactory cues. Heritable, stereotyped, robust behavioraloutput, and the consistent neural architecture that underlies it,are hallmarks of insect neurobiology and of olfactory neurobiolo-gy in particular. The pheromonal signals that guide many insectsto conspecific mating partners have been studied most thorough-ly, and recent work has continued to expand on the complexityand diversity of these signals. The chemical identity of insectpheromones can be as simple as the monomolecular compound11-cis-vaccenyl acetate, an apparently multifunctional phero-mone that mediates aggregation in Drosophila melanogaster

(Bartelt et al., 1985; Wertheim et al., 2002), indicates to malesthe mating state of a female (Ejima et al., 2007; Ha and Smith,2006) and increases female receptiveness to male courtshipattempts (Kurtovic et al., 2007), or as complex as the mixtures ofsex-pheromone components released by some adult femalemoths (Baker, 2008; Vickers et al., 1998) and beetles (Leal,1996; Yuko and Walter, 2008). The Hymenoptera employ perhapsthe greatest variety of social olfactory signals studied thus far (LeConte and Hefetz, 2008; Slessor et al., 2005). In honey bees, forexample, recent work has elaborated how pheromones organizethe defense of the hive (Hunt, 2007), recruit foragers (Thom et al.,2007), and reinforce the primacy of the queen (Strauss et al., 2008;Vergoz et al., 2007). The diversity of volatile semiochemicals andcuticular hydrocarbons used by ants (Hymenoptera: Formicidae),and their roles in behaviors ranging from recognition and

navigation to alarm and courtship remain an active area of study(c.f. Endler et al., 2004; recently reviewed in Holldobler andWilson, 2009). Recent advances in the field have extended to theneuroanatomy (Zube et al., 2008) and neurophysiology (Yamagataand Mizunami, 2010) of the olfactory system in ants. Evidence ofcues for intraspecific communication in other insects, such asstimuli described as stress-, aggregation-, and sex-relatedpheromones in D. melanogaster (Bartelt et al., 1985; Ejimaet al., 2007; Suh et al., 2007, 2004), has been accumulating.Two common threads emerge: these signals often are complexmixtures of volatiles, with multiple physical and behavioraleffects (Ferveur, 2005; Siwicki et al., 2005; Slessor et al., 1990),and processing of the stimuli may be segregated anatomically(Jefferis et al., 2007; Seki et al., 2005) or by the expression ofcommon genes, as in the case of D. melanogaster stress (Suh et al.,2004) and sex pheromones (Manoli et al., 2005).

Finally, insects are innately attracted to the complex odors ofcertain host animals (e.g. Zwiebel and Takken, 2004) or plants (e.g.Raguso and Willis, 2002), suggesting that foraging, like socialbehavior, relies on heritable responses to olfactory stimuli. Insectsalso can learn associations between volatile stimuli and rewards inthe natural environment. These abilities often are related to thedemands of the animal’s environment, e.g. the diversity of foodsources. Honey bees can learn the scent of flowers that areprofitable to visit (i.e. are yielding nectar and/or pollen) at anytime, either through direct experience (Chittka and Raine, 2006) orby exposure to the volatiles associated with nectar or pollentransferred to prospective foragers by returning bees during thewaggle-dance (Farina et al., 2007; Gil and De Marco, 2005). Locustsare voracious, generalist feeders, with a robust ability to associateolfactory stimuli with the quantity and nutritional quality of foodsources (Behmer et al., 2005). The sphinx moth Manduca sexta

exhibits an innate attraction to, and preference for, certain hostflowers but can learn to feed from others when the preferredflowers are scarce (Riffell et al., 2008b). Finally, in two relatedspecies of parasitic wasps, the ability to associate volatiles with thepresence of host caterpillars correlates with the behavioralplasticity demanded by the distribution of their respective hostspecies (Bleeker et al., 2006; Smid et al., 2007). Although theparticular form and capacity for olfactory learning vary among theinsects, the benefits for growth (Dukas and Bernays, 2000) andmating success (Dukas, 2005) suggest that learning is widespread.

Whether an olfaction-based behavior is innate or learned, itsexpression may be affected by the context in which the olfactorystimulus is experienced. A single volatile often has little effect oninsect behavior when detected in isolation but gains behavioralsignificance in the context of other volatiles. In response toherbivory by the sawfly Diprion pini, the Scots pine tree emitselevated quantities of the terpenoid (E)-b-farnesene (Mumm andHilker, 2005). While neither (E)-b-farnesene nor the other head-space volatiles of the Scots pine tree are attractive to parasitoids ofD. pini, mixtures of (E)-b-farnesene with other head-spacevolatiles are highly attractive to the parasitic wasp Chrysonotomyia

ruforum (Mumm and Hilker, 2005). Volatiles also may elicitdifferent behaviors in different contexts. For example, 11-cis-vaccenyl acetate acts as an aggregation pheromone for D.

melanogaster in the context of plant volatiles (Bartelt et al.,1985) and inhibits mating in the context of other males or matedfemales (Jallon et al., 1981).

Olfactory stimuli can elicit different, even antagonisticresponses based on the internal physiological state of an individualinsect. Many factors, such as age, time in the circadian cycle, andfeeding and mating status, may change the salience or associationof stimuli that are important for the survival of an individual insect.Thus the nervous system must temper its sensitivity, resolution, orprecision to suit the needs of the individual based on the current

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447 431

physiological or behavioral context (Anton et al., 2007; Bodin et al.,2008). Attraction to olfactory stimuli can be diminished aftermating (e.g. male moth Agrotis ipsilon; Barrozo et al., 2010;Gadenne et al., 2001) or feeding (e.g. female mosquito Anopheles

gambiae; Klowden, 1996). The neural mechanisms underlyingthese similar behavioral changes appear, respectively, to be central(i.e. decreased response of AL neurons; Gadenne et al., 2001) andperipheral (i.e. decreased antennal sensitivity; Takken et al., 2001)owing to down-regulation of olfactory-receptor proteins (Fox et al.,2001). Innate attraction to food odors in D. melanogaster ismediated chiefly by a single glomerulus (Semmelhack and Wang,2009). Recent work has demonstrated a mechanism in whichhunger, signaled by low insulin, upregulates a peptide receptor onthe ORCs innervating this glomerulus, facilitating vesicle releaseand ultimately influencing the animal to search for food (Root et al.,2011). Finally, locusts exhibit state-dependent valuation of foodresources, such that the locust Schistocerca gregaria prefersvolatiles associated with food last encountered when the animalwas nutrient stressed (Pompilio et al., 2006). Thus, different neuralmechanisms can produce similar behavioral effects, and theinternal state of an insect likely alters olfactory function atmultiple levels, in different ways for different species.

Insects offer numerous case studies, each revealing connectionsamong olfactory stimuli, neural mechanisms, and behavior. Acomprehensive description of natural, olfaction-based behaviorsand of the volatiles that control or modulate them can guideexperimental analysis of the neural mechanisms that underlieolfactory guided behaviors.

4. Evolution and speciation

Insects have evolved richly diverse olfaction-based behaviors asadaptations to diverse ecological challenges and opportunities andexhibit correspondingly diverse olfactory structures and mecha-nisms. A first step toward understanding the differences in theolfactory neurobiology of different insect taxa is to question howthese differences may have arisen through evolution. Recent workin closely related species has described changes in olfactoryprocessing that accompany speciation, and a picture of olfactorysystem evolution in insects is emerging.

4.1. Evolution of olfactory receptors

Insect ORs appear to be functionally and genetically distinctfrom those discovered in other taxa (Touhara and Vosshall, 2009).The topology of insect ORs is inverted within the cell membranewith respect to the N-terminal orientation of the classical G-protein-coupled receptor structure exhibited by mammalian ORs.Binding of volatile molecules by insect ORs is accomplished by aheterodimeric complex of one ‘typical’ insect OR and one ‘common’receptor expressed in nearly all ORCs, designated Or83b in D.

melanogaster (Benton et al., 2006). Or83b and its orthologues,hereafter ‘‘Orco’’ (after Vosshall and Hansson, 2011), are necessaryfor both ORC function (Jones et al., 2005; Larsson et al., 2004) andtrafficking of ORs to and insertion in the membrane of ORC cilia.Recent work has suggested that this heterodimeric complex itselfserves as an ion channel that might be both ligand- and cyclic-nucleotide-activated (Nakagawa and Vosshall, 2009; Sato et al.,2008; Wicher et al., 2008). Both the heterodimeric pairing of aninsect OR and an Orco protein, and the sequence of the Orco gene,are highly conserved among insects studied to date (Jones et al.,2005; Krieger et al., 2003; Melo et al., 2004; Nakagawa et al., 2005;Pitts et al., 2004; Robertson et al., 2010). Experiments in whichOrco genes from other insects rescue receptor-neuron responses inmutant D. melanogaster lacking copies of Or83b demonstrate thatthe function of Orco proteins is conserved. (Jones et al., 2005).

From this common point, the OR repertoire of insects diverges.Intraspecies OR sequences are very diverse. The amino-acidsequences of ORs in D. melanogaster are only 15% similar onaverage (Robertson et al., 2003), illustrating the diversity of ORsneeded to detect and discriminate a wide variety of volatiles. Theresponse of an ORC expressing one of these ORs, processed throughsynaptic circuitry in the olfactory pathways of the CNS, forms thebasis for encoding all relevant volatile stimuli in the animal’senvironment. How OR tuning in animals with widely differentolfactory environments has evolved is still poorly understood, butmuch progress has been made in describing how the OR repertoireof species within a group diverges.

An insect’s ORs are products of evolutionary pressures andprocesses that enhance, remove, or alter receptor-protein expres-sion in ORCs. The growing number of genomes available for speciesin the genus Drosophila provides the best example of this process.Odorant-binding proteins (OBPs; Vieira et al., 2007) and ORs(Nozawa and Nei, 2007) are subject to ‘‘birth and death’’ evolution,wherein new genes are produced through duplication andmodification, and some genes ‘‘die off’’ by deletion or mutationto pseudogenes. Certain species of fruit flies in the genusDrosophila, however, exhibit remarkable conservation of ortholo-gous OR groups, sharing 22–79% sequence similarity (Guo and Kim,2007). Many of the differences between orthologous ORs areattributable to synonymous substitution, suggesting that thefunction of these ORs is conserved between groups. In fact, themolecular receptive ranges (olfactory ‘‘tuning’’) of ORCs in thelarge basiconic sensilla of several Drosophila species appear to benearly identical (Stensmyr et al., 2003). In comparison, while themosquito species A. gambiae and Aedes aegypti share 21orthologous ORs, only one of these is shared with D. melanogaster

(Bohbot et al., 2007). Based on these findings, Guo and Kim (2007)proposed that the ORs conserved between species constitute a‘‘basis set’’ (Pouget and Sejnowski, 1997), in analogy with the threecone receptors for color vision. A basis set is a set of coding unitsthat, together, can be combined to produce a unique code for anyolfactory stimulus ‘object’ an animal may encounter. Although wemake no formal claim that ORs fit all of the requirements of a basisset, we note that coding schemes involving true basis functionshave been proposed for olfaction (Hopfield, 1995).

4.2. Olfactory specialization and speciation

The marked dependence of many insect species on one or a fewplant species for feeding and/or oviposition is a common basis forreproductive isolation and thus is likely to accompany speciation(Berlocher and Feder, 2002). In many of these insect–plant hostrelationships, volatiles play a key role in location of hosts. In the D.

melanogaster species subgroup, an extreme example of a host-plant specialist is Drosophila sechelia, a species endemic to theSeychelles Islands that lays eggs exclusively on the local Morinda

plant, whose fruit is toxic to the larvae of other Drosophila species(Tsacas and Baechli, 1981). Adult D. sechelia exhibit multipleadaptations of their olfactory system to Morinda volatiles(discussed further in Sections 4.2.1 and 4.2.2). The geographicisolation of D. sechelia from most other Drosophila species issufficient, albeit not necessary, for host specialization of theolfactory system. For instance, the apple maggot Rhagoletis

pomonella and its sibling species complex constitute a well-documented example of sympatric speciation (i.e. divergencewithout geographic isolation) that is driven by species-specificbehavioral preferences and adaptations of the olfactory system(discussed further in Sections 4.2.1 and 4.2.2) for the fruit volatilesof host plants that flower at different times (Linn et al., 2003).

Should closely related insect species share both a geographicrange and common food sources, reproductive isolation might be

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447432

maintained by multi-component sex-attractant pheromone mix-tures that evoke mate-seeking behavior, a well-studied phenome-non among the Lepidoptera (moths and butterflies), althoughconsiderably fewer volatile pheromones have been identifiedamong butterfly species than in moths. Reproductive isolation ismaintained among several sympatric moth species of the genusHeliothis by a system in which females of multiple species release acommon 16-carbon aldehyde (Z-11-hexadecenal) as the maincomponent of their sex pheromone, but each species-specific sex-pheromone mixture contains different secondary components.These minor components have been shown both to evoke mate-seeking behavior (Vetter and Baker, 1983, 1984) and to antagonizesuch behavior in sister species (Vickers and Baker, 1997). Studies ofthe reliance of insect species on a limited set of plants and onspecies-specific pheromones is of particular interest to neurobiol-ogists, as specialized behavior often is reflected in specialization ofthe peripheral, and even, central chemosensory system (de Bruyneand Baker, 2008).

4.2.1. Peripheral adaptations for olfactory specialization

The peripheral olfactory systems of closely related species ofinsects are likely to exhibit multiple structural and functionalhomologies. However, recent work using insect groups that areeither undergoing or have relatively recently undergone shifts inhost-plant specialization demonstrates that a variety of adapta-tions of the olfactory system can subserve exclusive hostrelationships. Male orchid bees (Euglossini) express novel matingbehavior in which they harvest blends of orchid volatiles andsubsequently release them in order to attract mates (reviewed inCameron, 2004). Distinct ‘collections’ of volatiles attractive todifferent females may act as isolating mechanisms during incipientspeciation, and two male morphotypes have been shown toharvest volatile mixtures that differ by the inclusion of a majorcomponent. Divergent behaviors among males are reflected inolfactory function, as GC-EAD experiments have shown distinctresponses to the differing component between male morphotypes(Eltz et al., 2008).

In the apple maggot fly R. pomonella, ORCs from three races thatexhibit distinct behavioral preferences for three different hostplants are differentiated, not by functional type or relativeabundance in the antenna (Olsson et al., 2006a) but by theirresponse thresholds and temporal firing patterns to volatilesemitted by their preferred host plant (Olsson et al., 2006b). Incontrast, specialization of D. sechelia on Morinda is reflected notonly in a shift of the abundance of a single type of sensillum (ab3),apparently at the expense of other types of sensilla found inheterospecifics (Stensmyr et al., 2003), but also in the alteredaffinity of an ab3 ORC for a host volatile. This drastic shift in ORCtuning corresponds to a change of only nine amino acids in the ab3ORC receptor sequence (Dekker et al., 2006). Although frequent,adaptations of the peripheral olfactory system that emphasizerepresentations of host- and mate-related volatiles at the expenseof other volatiles may take on diverse forms: shifts in receptorabundance, shifts in receptor affinity, or both acting in parallel.

4.2.2. Central adaptations for olfactory specialization

Common organizational principles of insect olfactory systemspermit prediction of differences in the anatomy and neuralcircuitry of the AL based on peripheral olfactory adaptations inclosely related species. Owing to convergence of axons of ORCs of asingle receptor phenotype in a particular AL glomerulus, anincrease in the abundance of one type of ORC is likely to result in anincrease in the volume of its target glomerulus. The sameconvergence is also likely to transform a change in the molecularreceptive range of an ORC to a change in the tuning properties ofPNs arborizing in the corresponding glomerulus. In D. sechelia, the

DM2 glomerulus, in which ab3 ORC axons terminate, is corre-spondingly enlarged relative to the homologous glomerulus in D.

melanogaster (Dekker et al., 2006). Although ab3 ORCs can detecttheir preferred volatile, methyl hexanoate, at amounts as low asfive fg, DM2 PNs may benefit from convergence of ORC axons toexhibit an even greater sensitivity by improvement of the signal tonoise ratio in response to the compound (Dekker et al., 2006).

Male moths of two closely related species, Heliothis virescens

and Heliothis subflexa, have a subset of antennal ORCs that aremorphologically identical but sensitive to chemically differentsecondary pheromone components (Baker et al., 2004). Neverthe-less, the male-specific macroglomerular complexes (MGCs) in theALs of both species share identical volumes and a common spatialorganization (Vickers and Christensen, 2003). Within the MGCs ofH. virescens and H. subflexa, the PNs originating in an orthologouslarge glomerulus (the cumulus) are tuned to a major pheromonecomponent common to both species (Vickers and Christensen,2003). The PNs of an adjacent glomerulus (the DM glomerulus) ineach species, however, are functionally different, dedicated toprocessing information about a secondary component unique tothe conspecific pheromone blend. DM tuning may have a complexgenetic basis, as hybrid males of these two species exhibitresponses to both secondary components while maintainingparental cumulus tuning (Vickers, 2006a,b). Divergent mate-seeking behavior in H. virescens and H. subflexa relies oncombinatorial activation of a glomerular array, providing a modelin which the AL exploits small differences in the quality andproportion of the olfactory stimulus in order to execute species-specific behavior (Lei and Vickers, 2008).

5. Central olfactory pathways

Recent progress in describing the architecture of the majorcenters for olfaction in the insect brain – the AL, mushroom body(MB), and lateral horn (LH) of the protocerebrum – in severalspecies has suggested that olfactory centers at all levels are nothomogeneous processing units. Instead, in different species theyare divided into anatomical or functional subsystems that interactto varying degrees. Here we review findings on the structure andfunction of olfactory brain organization in insect taxa investigatedto date.

5.1. Antennal lobe

ORCs are anatomically and functionally isolated from oneanother at the periphery (but see Getz and Akers, 1994, andAndersson et al., 2010, for possible exceptions), and axons of ORCsexpressing the different ORs terminate in different glomeruli in theAL, the first brain region in which olfactory information channelstypically interact. In D. melanogaster, it is well established thatglomeruli receive input from receptors expressing one OR, and allORCs expressing a particular OR terminate in one glomerulus(Couto et al., 2005; Fishilevich and Vosshall, 2005; Gao et al., 2000;Vosshall et al., 2000). Where more than one receptor is expressedin ORCs that terminate in a glomerulus, the combination is stillunique to that glomerulus (Couto et al., 2005; Fishilevich andVosshall, 2005; Goldman et al., 2005; Vosshall et al., 2000).Experiments with transplantation of antennae between male andfemale moths (Rossler et al., 1999) or related species of moths(Vickers et al., 2005), demonstrate that the identity of the ORCdetermines the functional identity of the glomerulus in theseanimals as well. Evidence for ORC tuning governing glomerularfunction in other insects is currently sparse (Ghaninia et al., 2007;Kelber et al., 2006; Robertson and Wanner, 2006), but it isreasonable to assume that the principle is conserved amonginsects, with the exception of some hemipteran and orthopteran

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447 433

species (Ignell et al., 2001; Kristoffersen et al., 2008a). Glomerulithus represent olfactory information channels with uniqueresponse profiles (Hallem and Carlson, 2006) that interact via alocal network in the AL.

Certain insect taxa lack this anatomical segregation of inputchannels. At one extreme, the reduced ALs of the Psyllidae (plantlice) and members of the Aphidoidea (both Hemiptera: Sternor-rhyncha) apparently do not have glomeruli, despite their relianceon olfaction to locate plant hosts (Kristoffersen et al., 2008a). TheALs of locusts (Orthoptera: Acrididae), exhibit microglomerularstructure but not segregation of ORC input (Ignell et al., 2001)(Fig. 2A). Antennal ORCs, likely expressing different ORs, project to1–3 AL ‘‘microglomeruli’’ (Hansson et al., 1996), and the dendritesof PNs sample from 10 to 25 anatomically distant microglomeruli(Anton and Hansson, 1996). Dendrites of PNs arborize in roughlyconcentric rings from the outside to the inside of the AL, but thefunctional significance of that pattern is unclear (Anton andHansson, 1996; Jortner et al., 2007). Several species of theAcrididae exhibit a novel aglomerular region within the AL, ofunknown function in olfactory processing (Ignell et al., 2001).Locusts and some other hemimetabolous insects also have ananatomically separate chemosensory center, the glomerular lobe,which receives primary afferents from ORCs in the maxillary palp(Frambach and Schurmann, 2004; Schachtner et al., 2005). Amongholometabolous insects, the glomerular lobe is fused to the AL, andglomeruli receiving inputs from non-antennal sources (e.g. thepalps) are distributed among ‘antennal’ glomeruli.

In a majority of insects that have been studied, AL glomerulireceive sensory input from a single type of ORC and relay outputmainly through uniglomerular PNs (Hildebrand and Shepherd,1997). In this manner, input to glomeruli from the peripheryrepresents an ensemble of independent sensory channels. How-ever, higher-order organization of glomerular innervation andclustering of behaviorally related glomeruli complicate the‘independence’ of glomeruli by regrouping them in the AL intoanatomical, and perhaps functional, subdivisions. In honey beesand ants (Kirschner et al., 2006; Zube et al., 2008, respectively), theantennal nerve divides into distinct tracts upon entering the AL,each tract feeding a cluster of neighboring glomeruli (Fig. 2C).Similarly, the ALs of several lepidopteran species exhibit MGCs—regions in which sex-pheromonal information is processed inenlarged, neighboring glomeruli in males (see Christensen andHildebrand, 2002 for a review) (Fig. 2B)—as well as sexuallydimorphic regions in the female AL, possibly related to volatilesimportant for oviposition (Reisenman et al., 2009; Rospars andHildebrand, 2000).

The AL of D. melanogaster exhibits no obvious anatomicaldivisions, but groups of glomeruli are distinguished by develop-mental origins (Jefferis et al., 2001), expression of common genes,or the type of sensilla housing the ORCs that innervate theglomerulus (Benton et al., 2007; Couto et al., 2005). Among thesegroups is a trio of glomeruli that in male flies are sites ofpheromone-information processing (Datta et al., 2008). Both ORCsand PNs in these glomeruli express fruitless, a gene involved incourtship behavior (Manoli et al., 2005). Divisions in an AL can alsorepresent areas of thermo- or hygro-sensitive input (Nishino et al.,2003; Zeiner and Tichy, 2000; Ruchty et al., 2010). Thus, within theAL, channels devoted to different modalities or to volatilesassociated with different behaviors may be segregated, althoughsome divisions are likely parallel systems in which olfactorystimuli activate many glomeruli across divisions (reviewed inGalizia and Rossler, 2010).

The neural circuitry of the AL in different species providesvaried mechanisms for interaction among glomeruli. LNs typicallyare multiglomerular, with neurites that extend to many orall glomeruli (often termed ‘homo’ or ‘global’ LNs) or that

interconnect only a smaller number of specific glomeruli(‘oligoglomerular’ or ‘hetero’ LNs) (Abel et al., 2001; Galizia andKimmerle, 2004; Matsumoto and Hildebrand, 1981; Reisenmanet al., 2008; Seki and Kanzaki, 2008; Seki et al., 2010; Wilson andLaurent, 2005). In moths (M. sexta [Hoskins et al., 1986] andBombyx mori [Seki and Kanzaki, 2008]), flies (D. melanogaster [Sekiet al., 2010]), and cockroaches (Periplaneta americana [Distler,1989]), a majority of LNs are global and GABAergic. Hymenopter-ans have a comparatively novel population of LNs; ‘hetero’ LNs aremore common (Fonta et al., 1993), may receive ORC input in onlyone glomerulus through dense innervation of its apical zone, andproject sparsely to several other glomeruli (Dacks et al., 2010;Galizia and Kimmerle, 2004). LNs produce Na+-based actionpotentials in response to olfactory stimulation in most insects inwhich they have been studied (c.f. Christensen et al., 1993; Galiziaand Kimmerle, 2004; Wilson and Laurent, 2005). By contrast, LNsin locust ALs (MacLeod and Laurent, 1996), as well as a subset ofLNs in cockroach ALs (Husch et al., 2009), respond to antennalstimulation with subthreshold membrane oscillations and voltage-activated calcium currents. Notably, a majority of such nonspikingcells in the cockroach (P. americana) AL are not GABAergic (Huschet al., 2009).

LNs also exhibit additional, diverse neurotransmitter pheno-types. All but the most basal hymenopterans (Dacks et al., 2010),and some cockroaches (Leucophaea maderae) (Losel and Homberg,1999; Nassel, 1999), have a number of histaminergic LNs.Histamine inhibits responses to olfactory stimuli in the glomeruliof honey bees, possibly by presynaptic inhibition of ORCs (Sachseet al., 2006). A class of excitatory, cholinergic LNs has beenidentified in D. melanogaster (Huang et al., 2010; Shang et al.,2007). Excitatory LNs have yet to be demonstrated decisively inother insect species, but many of the functions of excitatoryinterconnections can be performed by disinhibition based on serialinhibitory synapses formed by LNs (Christensen et al., 1993; Avronand Rospars, 1995). Small groups of LNs are further differentiatedby the expression of one or more putative neuropeptides, likely co-released with GABA (Nassel and Homberg, 2006). Finally, ALs areinnervated by various centrifugal modulatory neurons, affectingthe function of the AL according to circadian, appetitive, orassociative information from other brain areas (Dacks et al., 2005,2006; Ignell, 2001; Nassel, 1999; Sinakevitch et al., 2005).

5.1.1. Inhomogeneous interactions between glomeruli

There is increasing evidence that the network of glomerularinteractions, while stereotyped across individuals, is not ahomogenous all-to-all network. Inhomogeneity of interglomerularconnections may provide a basis for species-specific glomerularinteractions that facilitate processing of sensory information aboutnatural volatile stimuli. The strength of local excitatory inputs toPNs (Olsen et al., 2007) and presynaptic inhibition of ORC inputs(Olsen and Wilson, 2008) are not correlated in different glomeruliresponding to the same olfactory stimuli, as would be expected ifboth glomeruli were receiving only inputs carrying informationabout total afferent input. LNs in D. melanogaster branch withheterogeneous density patterns in multiple glomeruli and exhibit adegree of olfactory selectivity (Wilson and Laurent, 2005). Imagingof a population of LNs in D. melanogaster reveals patterns ofsynaptic activity that are specific to particular olfactory stimuli (Nget al., 2002). Finally, the expression of GABAB-like receptors in ORCterminals is heterogeneous throughout the D. melanogaster AL, andis, for example, particularly dense in pheromone-responsiveglomeruli and nearly absent in a CO2-responsive glomerulus (Rootet al., 2008). These studies suggest that inhibitory and excitatoryinputs carry additional information that is glomerulus- andstimulus-specific. Consistent with this idea, a model of thehoney-bee AL most accurately predicts the experimentally

Fig. 2. Schematic representations of the olfactory systems of four commonly studied animals. In each, the AL, (black circle) MBC, and LH (black ovals) are depicted, along with

the l-, ml-, and m-APT (dark grey, light grey, and medium grey lines, respectively), which comprise the output of the AL (after Galizia and Rossler, 2010). Particular,

characteristic features of the organization of the olfactory system of each order are illustrated in color. Features are chosen to illustrate differences between model insects, and

do not represent an exhaustive representation of the organizational principles of each olfactory system. (A) Orthoptera (S. americana, locust). The AL of the locust is

characterized by ORCs (in blue, redrawn from Hansson et al., 1996) which project to 1–3 microglomeruli (grey circles), and exclusively multiglomerular PNs (red, redrawn

from Anton and Hansson, 1996) arborizing in 10–25 microglomeruli. This is in contrast to the uniglomerular projections of ORCs and majority uniglomerular PNs in the other

orders illustrated. PNs project via the mAPT to the MBC and LH, and diffusely though the ml-APT to as-yet-unknown targets in the lateral protocerebrum. The MBC and LH of

locusts exhibit no obvious anatomical or functional subdivision. (B) Lepidoptera (B. mori, M. sexta, S. littoralis, others). The AL of male moths is typically subdivided into a

‘‘main’’ AL (green-shaded circles) and a macroglomerular complex consisting of pheromone-responsive glomeruli (blue-shaded ovals). Uniglomerular PNs in the main AL

project through the m-APT and arborize throughout the MBC and LH, while uPNs from the MGC arborize chiefly in a more conscribed region of the MBC and in the inferior

lateral protocerebrum (ILPC), neighboring the LH. (C) Hymenoptera (A. mellifera, C. floridanus, others). ORCs enter the ALs of bees and ants in several tracts (four tracts

illustrated, as found in A. mellifera). Each tract terminates in a distinct set of neighboring glomeruli (cyan, blue, orange and yellow circles, numbers not representative). The AL

is divided into two hemi-lobes (magenta and green shading) by associated output tract. uPNs from glomeruli in one hemi-lobe send axons through the m-APT (magenta line),

arborizing (magenta shading), in the outer portion of the lip (l) and the middle portion of the basal ring (br) of the MBC (one calyx depicted) before terminating in the LH. uPNs

from the other hemi-lobe send axons through the l-APT (green line), arborizing (green shading) in the LH before terminating in the inner portion of the lip and the basal ring of

the MBC. In the LH, each tract occupies a segregated region with some overlap between them. (D) Diptera (D. melanogaster). ORCs in flies send axons to single glomeruli in the

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447434

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447 435

observed output patterns of PNs, not when inhibitory connectionsare weighted homogeneously or when they vary with spatialproximity but instead when they scale according to the similarityof ORC input to the glomerulus (Linster et al., 2005). In the moth M.

sexta, sexually dimorphic glomeruli have strong, inhibitoryconnections to the main AL but apparently do not receivereciprocal inhibition from isomorphic glomeruli (Reisenmanet al., 2008), and in D. melanogaster, wide-field LNs are less likelyto innervate a glomerulus associated with social communication(Chou et al., 2010; Wilson and Laurent, 2005). Finally, the strengthof connections between glomeruli may be affected by learning-related plasticity observed in the AL (Daly et al., 2004; Thum et al.,2007; Yu et al., 2004), perhaps reinforcing relations betweenoutput channels associated with reward or punishment. LNsappear to provide a network of inhomogeneous inhibitoryconnections among glomeruli, reflecting underlying principles ofcomputation not yet fully understood.

The AL network has an inhomogeneous, species-specificarchitecture allowing for interactions among units, ranging fromindividual glomeruli to functionally or anatomically relatedclusters of glomeruli. Information processing is likely to besimilarly complex and species-specific, with common functionsmediated by different mechanisms in different species, as well asfunctions adapted for the particular olfactory world of a species.

5.2. Higher-order olfactory centers

From the AL, information about olfactory stimuli is relayed toseveral higher-order centers for subsequent processing. Recentwork has produced maps of the pathways through whichinformation flows in the olfactory system and allows for apreliminary analysis of the input/output characteristics of eachlevel of processing. Across species, these maps show varyingdegrees of segregation of information channels. The organizationof the AL proper and its output tracts, as well as the projections ofaxons of AL PNs to other brain areas and the integration ofinformation from the AL by third-order neurons, all are organizedaccording to characteristics specific to groups of insects. In thissection, we review those organizing principles. Chief among themis the emerging evidence that, as in the AL, higher-order regionsexhibit diverse patterns of segregation of olfactory channels.

5.2.1. Antenno-protocerebral tracts

AL PNs project to higher-order centers through several axonaltracts. In the following sections, we employ nomenclaturesuggested by Galizia and Rossler (2010), who identify threeantenno-protocerebral tracts (APTs) with a common position,although not necessarily common function or homology, acrossmany insects (Fig. 2). As the traditional nomenclature of ‘‘antenno-cerebral tracts’’ differs among species, this newer nomenclaturefacilitates comparison across taxa. The largest tract is the medialAPT (m-APT), which is present in the Archaeognatha andZygentoma, the most basal insects examined (Strausfeld, 2009;Strausfeld et al., 2009). In flies, moths, ants, honey bees, andcockroaches, most uniglomerular PNs send axons through the m-APT, have collateral projections to the mushroom-body calyces(MBCs), and terminate in the lateral horn (LH) of the protocer-ebrum (Homberg et al., 1988; Malun et al., 1993; Marin et al., 2002;Rø et al., 2007; Stocker et al., 1990). The multiglomerular PNs of theAL of locusts also follow this tract (Ignell et al., 2001). The lateral

ipsilateral AL, and through a unique antennal commissure (AC, dark blue lines) to the c

related by neuroblast origin of PNs, the type of sensilla that houses the ORCs that innerva

and purple shading in AL). The majority of uPNs project through the m-APT, and arborize

cyan, yellow, green and purple shading in MBC). The m-APT continues to the LH, where uP

the MBC (grey ovals in the LH). A segregated region (indicated by an asterisk) receives

APT (l-APT) is composed of the less-common multiglomerular PNaxons in moths and cockroaches and a mixture of uni- andmultiglomerular PNs in flies. In ants and bees uniglomerular PNsfrom one hemi-lobe of the AL project through the l-APT,constituting a distinct, parallel olfactory system thus far describedonly in the Hymenoptera (Galizia and Rossler, 2010; Kirschneret al., 2006; Zube et al., 2008) (Fig. 2C). This system is also presentin some basal, non-social hymenoptera and is thought to be anadaptation to complex olfactory environments and behaviors(Rossler and Zube, 2011). The l-APT in cockroaches and fliesterminates in the LH, but in moths, bees, and ants it also projects tothe MBCs. Finally, additional, minor tracts compose a medial-lateral APT (ml-APT) that carries axons of a variety of uni- ormultiglomerular PNs and projects in species-specific patterns tothe MBCs, LH, or other regions in the protocerebrum (reviewed inGalizia and Rossler, 2010).

Although insects share a common Bauplan of APTs, each taxonexhibits variations on the theme. Tracts sharing a projectionpattern but composed of uni- or multiglomerular PNs in differentspecies, for example, can facilitate the testing of hypotheses aboutthe functions of different types of PNs. Similarly, features unique toa taxon, e.g. the parallel uniglomerular PN system in bees and ants,may well be correlated with behavior particular to those species.

5.2.2. Mushroom-body calyx

Upon reaching their synaptic targets, PN axons terminate withvarying degrees of segregation. We consider first the calyces of themushroom bodies (MBCs), where PNs synapse with Kenyon cells(KCs). Recent work has provided new insights into the structure ofMBCs, and especially the compartmentalization of PN inputs and theKCs with which they synapse, demonstrating that the MBC is not ahomogeneous structure. Instead, in many insects it containsmultiple subsystems, with various architectures of segregationand integration in different species. Here, we consider recent workdescribing olfactory networks in the MB. We propose that a detailedunderstanding of these networks is required for investigations ofolfactory information processing in this region, as well as processingin the AL that shapes the input to this associative center.

In locusts (Schistocerca americana, Jortner et al., 2007; Laurentand Naraghi, 1994) and several moth species (Homberg et al.,1988; Rø et al., 2007), axon collaterals of AL PNs spread throughoutthe MBC, but a detailed analysis of their branching patterns is notavailable. In contrast, certain pheromone-responsive PNs of themoth B. mori terminate in circumscribed regions of the MBCs(Kanzaki et al., 2003; Seki et al., 2005) (Fig. 2B). For D. melanogaster,the only insect species studied systematically to date, axons of PNsfrom the same glomerulus terminate in remarkably similarlocations in the MBCs, in apparently fewer and more circumscribedregions than in moths or locusts (Jefferis et al., 2001, 2007; Linet al., 2007; Marin et al., 2002; Wong et al., 2002).

MBCs of ants and bees are subdivided into three regions: lip,collar, and basal ring (Ehmer and Gronenberg, 2002; Gronenberg,1999, 2001; Gronenberg and Holldobler, 1999; Mobbs, 1982).Axons of AL PNs projecting in both the l-APT and m-APT terminatein the lip and basal ring (Abel et al., 2001; Ehmer and Gronenberg,2002; Gronenberg, 1999, 2001; Gronenberg and Holldobler, 1999;Kirschner et al., 2006; Zube and Rossler, 2008) (Fig. 2C). The basalring has three concentric layers, receiving (respectively from insideto outside) input from the l-APT, m-APT, and visual neurons fromthe medulla (Gronenberg, 1999; Kirschner et al., 2006; Zube et al.,

orresponding glomerulus in the contralateral AL. Glomeruli in the AL are variously

te the glomerulus, and the MBC regions that the PNs target (red, cyan, yellow, green

in regions of the MBC defined by the KC type that inhabit them (corresponding red,

Ns terminate in conscribed regions, in groupings related but not identical to those in

innervation from pheromone-responsive PNs.

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447436

2008). Little overlap is evident between the layers. In the lip region,l-APT PNs innervate the inner core, while m-APT PNS innervate theoutermost layer and produce diffuse arborization throughout theMBCs (Kirschner et al., 2006). Several species of hemimetabolousinsects also exhibit functional division of the MBCs. In crickets(Gryllus bimaculatus), output neurons from the glomerular lobe(anatomically distinct from the AL and receiving chemosensoryinput from the mouth parts) terminate in the posterior calyx, whilePNs from the AL terminate in the anterior (Frambach andSchurmann, 2004). The MBCs of cockroaches (P. americana) aredivided into compartments receiving input from two distinctpopulations of uniglomerular PNs, male-specific PNs, and someneurons from the optic lobes (Strausfeld and Li, 1999).

Anatomical separation of input channels may not indicateseparation of information if the third-order neurons sample acrosschannels. Characteristics of integration depend on the overlap ofPN axonal terminals and KC dendritic fields. At one end of thespectrum, locust KCs have wide dendritic fields covering a largeportion of the MBC (Jortner et al., 2007; Perez-Orive et al., 2002).Each KC is estimated to receive input from ca. 50% of PNs, whichJortner et al. (2007) identified as approaching the theoretical idealfor a general associative network (Shannon and Weaver, 1964).This suggests a lack of segregation in the locust KC network. In themoth species Spodoptera littoralis and B. mori, KCs with widedendritic fields send axons to the a/b and a0/b0 lobes of the MBs,while KCs with narrower fields project to the g lobe (Fukushimaand Kanzaki, 2009; Sjoholm et al., 2006). The calyx is furthersubdivided by classes of KCs distinguished by immunoreactivity,also with distinct projection patterns (Fukushima and Kanzaki,2009; Sjoholm et al., 2005). Given the projection patterns of PNs,KCs of different classes may integrate information from varyingnumbers of PNs, but it appears that they may sample from anycombination of PNs. This simplification may yet be challenged by amore detailed study of the MBC in these insects.

KCs are arranged in the MBC of worker honey bees in a patternsimilar to that in moths. Class-II KCs integrate across the lip, basalring, and collar of the MBC and thus can receive input from alluniglomerular PN tracts (m- and l-APT) as well as from the visualsystem (Strausfeld, 2002), while dendrites of class-I KCs arerestricted to regions exclusively receiving projections of axonsfrom the l-APT, the m-APT, or both tracts. All classes project to thevertical lobe (analogous to a/b and a0/b0 lobes of other insects),but the upper two-thirds of the lobe is divided into stratainnervated by axons originating in either the lip, collar, or basalring of the MBC, and the lower third, proposed to be analogous tothe g lobe in D. melanogaster, is innervated by the axons of class-IIKCs (Strausfeld, 2002). The vertical lobe thus contains strata thatreceive projections of KCs that integrate either within or betweenthe two main input tracts. Further subdivision according to ALdivisions is not apparent, and KCs may sample across divisions thatshare a tract. Each class-II KC receives input from an estimated 10PNs, consistent with a greater degree of segregation between inputchannels in this animal (Szyszka et al., 2005).

The MBC of D. melanogaster exhibits a further degree oforganization (Fig. 2D). As mentioned above, PNs of the sameglomerulus have terminals in corresponding, circumscribedregions of the MBC. The MBC can be divided into four verticalsections, according to neuroblast origin (Ito et al., 1997; Technauand Heisenberg, 1982), each of which is laterally divided intoregions containing KCs exclusively from one of five groupsdistinguished by target lobe and developmental origin: g lobe,a0/b0 lobe, and pioneer, early, and late a/b lobe KCs (Lee et al.,1999; Zhu et al., 2003). A detailed analysis of a subset of PNs from13 glomeruli revealed that PNs target KCs across the verticaldivisions, but typically target KCs of one group (Lin et al., 2007) andcontribute to distinct, isolated clusters of PN–KC synapses called

microglomeruli (Leiss et al., 2009). Extension of these results to thePN groups identified by Jefferis et al. (2007) suggests that the KCsintegrate input only within PN groups and that different groups ofPNs are associated with each lobe of the MB. Each KC is estimatedto receive input from 10 PNs (ca. 5%) (Turner et al., 2008). If eachclass of KCs integrates information from only a small group of PNs,this arrangement might produce a system of parallel associativenetworks. Within each subset, connectivity may more closelyresemble the locust MBC, with ca. 50% of input channels makingconnections with each KC.

Detailed findings about fine divisions of the MBC, directlycomparable to those in D. melanogaster, are not yet available forother insects. Available evidence, however, describes a spectrum ofinput/output segregration within the MBC. At one pole, D.

melanogaster PNs and KCs are isolated in multiple, parallelsubsystems with little or no overlap, as opposed to locust MBCsin which interaction between any PN and any KC is anatomicallypossible. As a first approximation, this spectrum suggestsfunctional consequences of neural architecture that can informcomparison between species.

5.2.3. Lateral horn

In contrast to the MBC, the lateral horn of the protocerebrum inmany insects appears to be a diffuse, aglomerular neuropil (e.g. Sunet al., 1997; Yasuyama et al., 2003). In D. melanogaster, PNsterminate in the LH in a stereotyped pattern resembling that in theMBC (Jefferis et al., 2007; Marin et al., 2002; Tanaka et al., 2004;Wong et al., 2002) (Fig. 2D). As in the MBC, PNs with a commonneuroblast origin project to similar regions of the LH (Marin et al.,2002). Consistent with this, PNs that terminate in the same regionof the LH tend to originate in neighboring glomeruli, although thereciprocal arrangement is not observed (Marin et al., 2002). Inaddition, PNs cluster in similar groups in the MBC and LH (Jefferiset al., 2007). Finally, PNs that receive input from ORCs in certaintypes of sensilla project to similar areas of the LH (Jefferis et al.,2007). Thus at least in D. melanogaster, PNs terminate instereotyped but overlapping patterns.

Segregation of inputs in the LH is largely correlated with the inputtract through which axons project, or by a distinction between food-derived and social or pheromonal stimuli. The m-APT and l-APT PNsin honey bees and ants terminate in distinct regions of the LH andshare a diffuse region of overlap between them (Kirschner et al.,2006; Zube et al., 2008) (Fig. 2C). The LH is similarly divided by inputtracts in species of moths (Homberg et al., 1988; Rø et al., 2007)(Fig. 2B) and D. melanogaster (Wong et al., 2002) (Fig. 2D). PNssupplying the ml-APTs in bees and ants produce a unique ‘‘lateralnetwork’’ located between the g lobe of the MB, the LH, and otherareas of the protocerebrum (Kirschner et al., 2006; Zube et al., 2008).Some PNs in this group terminate in the LH, exclusively in the areaalso innervated by the l-APT. The function of this network is not yetunderstood, but it provides a substrate for interaction between theAL, MB, and LH not present in other insect species.

Input to the LH from pheromone-responsive PNs in severalspecies is segregated to varying degrees. In D. melanogaster, axonsfrom two glomeruli project to a restricted region of the LH thatcontains few arborizations of other PNs (Jefferis et al., 2007)(Fig. 2D). These glomeruli express a male-specific form of thefruitless gene required for male sexual behavior (Demir andDickson, 2005; Manoli et al., 2005; Stockinger et al., 2005). Thisregion is enlarged in males and contains an additional PNarborization produced when the male isoform of fruitless isexpressed both in PNs and other cells in the brain (Datta et al.,2008). The pheromone-responsive PNs in moths project to a regionthat is more isolated from PNs originating in the main AL (Homberget al., 1988; Kanzaki et al., 2003; Seki et al., 2005) (Fig. 2B). Insilkworm moths (B. mori), this region is further divided into two

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447 437

partially overlapping segments receiving input from each of twoMGC glomeruli (Kanzaki et al., 2003; Seki et al., 2005). Finally, inthe LH of the cockroach, PNs responding to cold, dry, or moist air,respectively, terminate in a distinct region of the LH, with areasoverlapping those innervated by uni- and multiglomerular PNsfrom other glomeruli (Nishino et al., 2003).

It has been hypothesized that the LH supports integration ofmore ‘‘stereotyped’’ olfactory information than the MB (Jefferiset al., 2007). Comparing the recent, detailed findings about theanatomy of the LH and MBC (Jefferis et al., 2007; Lin et al., 2007),we suggest that this distinction lacks adequate support. KCs in theMBC of D. melanogaster integrate information from a restrictedsubset of AL glomeruli and exhibit pronounced input/outputsegregation. Similarly to what is observed in the MBC, groups of LHneurons in D. melanogaster integrate input from PNs that arborizein limited numbers of glomeruli and terminate in spatiallyrestricted zones (Jefferis et al., 2007; Tanaka et al., 2004). Thus,available information suggests that neither structure supports anentirely general integration scheme in which information origi-nating in any one glomerulus may be integrated with thatoriginating in any other glomerulus by third-order neurons.Recent investigation of a small, genetically identifiable group ofKC neurons, however, reveals that members of the group do notrespond similarly to a battery of volatile stimuli (Murthy et al.,2008). Thus, while groups of KCs may sample from the samerestricted subset of PN inputs, each may receive inputs from aparticular combination of PNs within that subset, leading tounique, non-stereotyped responses. The fine structure of the MBC,consisting of many microglomeruli that segregate small groups ofinput, output, and local circuit synapses (Leiss et al., 2009; Steiger,1967; Ganeshina and Menzel, 2001), supports this hypothesis.These structures are a common feature of the MBC in neopteraninsects (Groh and Rossler, 2011), and they represent an order ororganization not seen in the LH.

Physiological and behavioral experiments suggest an addition-al, distinct role for the LH, different from that of the MB andpossibly species-specific. In locusts (S. americana) the LH is thesource of a strong, feed-forward inhibition to KCs (Perez-Oriveet al., 2002). A cluster of GABAergic LH interneurons (LHIs)reportedly connects the LH to the MBC and likely underlies thisfeed-forward inhibition. LHIs are driven by the same PN inputs thatdrive the KCs but with a greater latency. LHIs then inhibit KCs,closing the temporal window during which they can integrateinput. GABAergic neurons that may be analogous to the locust LHIsalso have been described in flies (D. melanogaster; Yasuyama et al.,2003), moths (M. sexta; Homberg et al., 1987), and cockroaches (P.

americana; Nishino and Mizunami, 1998). No evidence of feed-forward inhibition in D. melanogaster KCs has been found, however(Turner et al., 2008), and GABAergic LHIs in P. americana areinhibited by multimodal stimulation and thus are likely tosubserve a different function (Nishino and Mizunami, 1998).While honey-bee KCs exhibit periodic inhibition similar to that inlocust KCs, it is likely driven by KCs themselves, via feedbackneurons of the MB peduncle (Grunewald, 1999; Szyszka et al.,2005). It remains to be seen whether the LHI function described inlocusts is conserved in other species.

5.2.4. Protocerebrum and beyond

Little is known about olfactory information processing beyondthe MB and LH. The extrinsic neurons of the MB and LHNs in D.

melanogaster project to overlapping regions in diverse parts of thebrain (Tanaka et al., 2008). Olfactory information spreadsthroughout the brain and participates in the organization ofnumerous behaviors. The participation of MBs in associativememory has been well documented (reviewed in Heisenberg,2003; Margulies et al., 2005). The varying degrees of segregation

within and integration across the parallel divisions of olfactoryinput suggest that the underlying rules of association are complexand species-specific. The LH is similarly organized to integratewithin or between input channels and may perform multipleprocessing and feedback functions controlling behavior.

Further downstream, a pair of symmetrical brain regions hasbeen closely linked to an important olfactory behavior. Circuits inthe lateral accessory lobe (LAL) and the ventral protocerebrum(VPC) of moths (primarily investigated in M. sexta and B. mori)generate a ‘‘flip-flopping’’ pattern of activity in descending neuronsthat have arborizations in the LAL (Kanzaki et al., 1991, 1994;Olberg, 1983). These neurons switch between high- and low-activity phases on each successive stimulus pulse of sex phero-mone. This pattern of activity corresponds to the ‘‘zig-zagging’’turns that moths, and many other insects, make when searching forthe source of an odor plume (reviewed in Carde and Willis, 2008)and furthermore is correlated with the activity of motor neuronsresponsible for head movements during a turn (Kanzaki andMishima, 1996). Recent work has described the circuitry of theLAL-VPC system as a pair of mutually inhibitory units thatalternate, generating long-lasting activity in one hemisphere andquieting the other on each encounter with a pulse of olfactorystimulus (Iwano et al., 2010). Because diverse insects exhibitsimilar behavior when searching for the unseen source of an odorplume, this mechanism may be conserved across species. Evidencethat some insects can locate odor sources in the absence of a plume(e.g. D. melanogaster, Budick and Dickinson, 2006), however,suggests that additional mechanisms specific to other odor-searchstrategies may be revealed by comparative work.

6. Comparative coding

By combining findings about how an animal interacts withvolatile chemical stimuli in its environment with knowledge of thefunctional organization of the olfactory system, we can begin tounderstand how different animals encode olfactory information.Here, we consider current progress on olfactory informationprocessing, focusing on these processes in the AL: gain control andbroadening or sharpening of tuning curves between input andoutput, synchrony between AL PNs, and encoding of the temporalfeatures of the stimulus. Animals may use different codingschemes, or variations of similar schemes, that are proximallydetermined by the particular structure and mechanisms present inthe species and ultimately determined by adaptation to aparticular niche and evolution of solutions to specific challenges.

6.1. Gain control

The range of output of an ORC is a function of both the identityand intensity of an olfactory stimulus. ORCs may respond weaklyto a wide range of volatiles or to a low concentration of a preferredcompound, yet respond robustly to higher concentrations ofcertain volatiles (c.f. Bhandawat et al., 2007; Bichao et al., 2005; deBruyne et al., 2001; Ito et al., 2009; Kristoffersen et al., 2008b;Ochieng and Hansson, 1999; Pelz et al., 2006; Schlief and Wilson,2007; Wilson et al., 2004). Between these extremes of activation,the circuitry of the AL needs to maintain sensitivity, preventsaturation, encode stimulus intensity, and maintain a consistentrepresentation of a stimulus across a range of concentrations.Underlying these functions is the control of gain, or the conversionof signal strength between input and output to make efficient useof the dynamic range of PN output (Fig. 3A).

The efficient use of a PN’s dynamic range begins at the ORC-to-PN synapse. In D. melanogaster, a spike in an ORC produces a strong,reliable response in the postsynaptic PN (Bhandawat et al., 2007).This synapse has a high probability of neurotransmitter release and

Fig. 3. Principles of processing in the antennal lobe. (A) Gain. The circuitry of the

antennal lobe and synaptic properties of AL neurons determine the relationship

between ORC input to a glomerulus and PN output. Gain denotes two distinct, yet

related concepts: (i) the nonlinear relationship between the firing rate of individual

ORCs and the firing rate of the PNs on which it synapses (from Bhandawat et al., 2007),

which underlies, in part, (ii) the relationship between odor concentration and PN

output. The slope of these functions is the gain. Neuromodulation can alter this

relationship (dark grey and light grey lines in ii). (B) Sharpening and broadening of the

population output. (i) Lateral inhibition can produce an activity-dependent reduction

in the firing rate of PNs across the responding population. Glomeruli receiving weak

input (orange line ORC) may be completely silenced (grey dashed line PN). (ii) The

effect is to ‘‘sharpen’’ the representation of the stimulus from the input (light grey

curve) to the output (dark grey curve) layers. Population activity is represented as a

Gaussian curve with a central peak, in analogy to other sensory systems, although

patterns of activity in insect ALs may take a different form. (iii) Lateral excitation

spreads activity throughout the AL, increasing the response of some PNs, and in some

cases producing output in PNs (orange line) that receive no input from their cognate

ORCs (dashed grey line). (iv) The population response in the PN output layer (dark grey

line) is ‘‘broadened’’ with respect to the ORC input layer (light grey line).

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447438

therefore depresses quickly during high-frequency spiking(Kazama and Wilson, 2008). As a result, the transformationbetween ORC input and PN output is highly nonlinear, amplifyingweak and suppressing strong inputs (Bhandawat et al., 2007)(Fig. 3Aii). In response to a given stimulus, input to the AL istypically a combination of weak activation of many ORCs andstrong activation of only a few (Bhandawat et al., 2007). Thenonlinear transformation amplifies small differences between theweaker inputs, making PN responses equally likely across allpossible firing rates, a phenomenon known as histogramequalization that also is observed in the visual system of insects(Laughlin, 1981). Similar conclusions can be drawn for the ORC/PNtransformation in moths (Ito et al., 2009; Jarriault et al., 2010) andhoney bees (Krofczik et al., 2009).

Inhibitory and excitatory synaptic connections in the AL,mediated by LNs that may integrate inputs in many or allglomeruli, can adaptively regulate the dynamic range of PNresponses relative to the total input to the AL. In D. melanogaster,both presynaptic (Olsen et al., 2010; Olsen and Wilson, 2008; Rootet al., 2008) and postsynaptic (Silbering and Galizia, 2007)inhibition in glomeruli scales with the total primary-afferentinput to the AL. Similar global inhibition has been observed inhoney bees (Deisig et al., 2006, 2010; Sachse and Galizia, 2002,2003), and the presence of wide-field LNs in the ALs of many insecttaxa suggests that this function may be universal. Because globalinhibition increases with total ORC input, the representation of anolfactory stimulus is maintained by the relative activation of PNsacross the AL over a wide range of concentrations (Sachse andGalizia, 2003). However, the differential expression of GABAB-likereceptors that mediate presynaptic inhibition provides forindependent gain control in each glomerulus (Root et al., 2008).Indeed, there is evidence that the weighting of global, inter-glomerular inhibition varies between glomeruli (Olsen et al.,2010).

A class of excitatory LNs in D. melanogaster also encodes totalafferent input across glomeruli and may increase PN output inweakly activated glomeruli to a level that can be detected bydownstream neurons (Huang et al., 2010; Olsen et al., 2007; Shanget al., 2007). These competing, global excitatory and inhibitoryinputs may be balanced in a PN, such that they increase togetherwith the concentration of an olfactory stimulus (Root et al., 2007).The simultaneous adjustment of balanced, background excitatoryand inhibitory inputs can serve as a form of gain modulation as hasbeen hypothesized for mammalian cortical neurons (Chance et al.,2002), such that gain is increased with low background anddecreased with high background. Weak inputs to a PN thus wouldbe amplified relative to the total input to the AL, enhancing weakportions of the representation of a stimulus at low concentrationbut suppressing weak, likely noisy inputs at high total stimulusconcentration. Such dynamic modulation of gain in the AL has yetto be demonstrated.

Finally, serotonin (5-hydroxytryptamine) has been shown toincrease the gain and enhance the responses of PNs in the ALs ofmoths (Dacks et al., 2008) (Fig. 3Aiii). In adult M. sexta

(Kloppenburg et al., 1999) and B. mori (Gatellier et al., 2004),levels of serotonin in the AL and protocerebrum peak during theinsect’s active phase in the circadian cycle. Similar work on D.

melanogaster found glomerulus- and odor-specific enhancement ofPN responses but no effect of serotonin on gain, suggesting subtledifferences between species (Dacks et al., 2009). Serotonin-enhanced presynaptic inhibition is known to be involved in gaincontrol (see above), however, and might play an additional role inmodulating gain in fruit flies. Serotonin-immunoreactive neuronshave been described in the ALs of many insect species (Dacks et al.,2010; Ignell, 2001; Kent et al., 1987; Salecker and Distler, 1990;Siju et al., 2008; Wegerhoff, 1999). Depending on the species, these

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447 439

apparently serotonergic cells innervate a variable number ofglomeruli and have species-specific connections with other areasin the brain. Gain in the ALs of diverse insects thus might bemodulated through the influence of species-specific brain areas,and under conditions specific to each species.

6.2. Tuning/coding transformation

Olfactory stimuli elicit characteristic patterns of ORC inputacross glomeruli (c.f. Deisig et al., 2006; Hallem and Carlson, 2006;Ng et al., 2002; Sachse et al., 1999; Silbering and Galizia, 2007;Silbering et al., 2008; Wang et al., 2003). The fidelity with whichORC input is transformed to PN output, however, is debatable. Doesthe dense interconnection of glomeruli by AL interneurons performa transformation, either sharpening or broadening, between theORC input and the PN output of the AL? In this section, we reviewthe evidence that inhibitory and excitatory circuits in the AL shapeolfactory codes. In addition, we suggest that apparently conflictingresults stem from either volatile- or species-specific interactionsand likely could be resolved by a comparative approach consider-ing the natural olfactory ecology of each species.

This discussion necessarily assumes that information aboutodors is encoded in the pattern of activity across the AL. It is usefulto note at this point that some innate behaviors, e.g. mate-seekingin moths (B. mori; Kanzaki and Shibuya, 1983) and flies (D.

melanogaster; Kurtovic et al., 2007) and food-seeking in flies (D.

melanogaster; Semmelhack and Wang, 2009), depend on singleglomeruli that apparently provide all information necessary todrive those behaviors.

6.2.1. Sharpening

A primary effect of inhibition in a neural circuit is to preventunder-stimulated units from participating in the output. In thisway, a stimulus is represented by the most strongly activatedneurons, and overlap between codes for different stimuli isreduced by a ‘‘sharpening’’ of the population code (Urban, 2002)(Fig. 3Bi and ii). Ca2+-imaging studies in ALs of both honey bees(Apis mellifera) and fruit flies (D. melanogaster) have shown thatvolatile-evoked activation patterns of ORCs largely match theactivation patterns observed in PNs (Ng et al., 2002; Sachse andGalizia, 2002, 2003; Silbering and Galizia, 2007; Wang et al., 2003).In both species, global inhibition reduces the responses of allactivated glomeruli (Sachse and Galizia, 2002, 2003; Silbering andGalizia, 2007). In A. mellifera, however, PN output is suppressed byinhibition despite weak or intermediate activation of ORCsprojecting to the same glomerulus (Sachse and Galizia, 2002)(Fig. 3Bi). Similar suppression of ORC-to-PN transmission by globalinhibition is not seen in D. melanogaster (Silbering and Galizia,2007; Silbering et al., 2008). Global inhibition reduces PNresponses in both A. mellifera and D. melanogaster but sharpensthe population code only in the former.

A second form of inhibition exerts glomerulus- and stimulus-specific effects on ORC-to-PN output in D. melanogaster (Wilsonand Laurent, 2005). In A. mellifera, blockade of GABAA-likereceptors fails to eliminate suppression of ORC-to-PN transmission(Sachse and Galizia, 2002), and the remaining odor- andglomerulus-specific suppression of responses may be mediatedby histaminergic LNs (Sachse et al., 2006). Inhibitory interactionbetween glomeruli in ALs of D. melanogaster is most evident inpatterns of mixture-suppression of PN output: activation of ORCinput by one mixture component inhibits PN output in glomeruliactivated by the complementary component of a binary mixture(Silbering and Galizia, 2007). Similar mixture effects are observedin A. mellifera (Deisig et al., 2010; Joerges et al., 1997; Silbering andGalizia, 2007). In both species the effect is volatile- andglomerulus-specific, consists primarily of suppression of PN

output, and decreases the similarity between responses fordifferent mixtures. Complex, heterogeneous, inhibitory networksappear to shape PN output similarly in both D. melanogaster and A.

mellifera but may utilize species-specific cell types.

6.2.2. Broadening

The response profiles of several PNs in D. melanogaster arebroader than those of their cognate ORCs (Schlief and Wilson,2007; Wilson et al., 2004, but see Root et al., 2007; Wang et al.,2003). This remarkable result suggests that network mechanismsbroaden the molecular receptive range of PNs with respect to thatof the ORCs that provide sensory input to them (Fig. 3Biii and iv).The molecular receptive range of a neuron can be broadened in twodistinct ways: (1) the PN may respond more strongly to olfactorystimuli that weakly activate the cognate ORCs, or (2) the PN mayrespond to volatiles that do not activate cognate ORCs. These twoeffects likely are mediated by two network mechanisms. First,ORC-PN synapses amplify small, probabilistic inputs from multi-ple, weakly responsive ORCs to a greater degree than the large,sustained input of multiple firing ORCs (Bhandawat et al., 2007;Kazama and Wilson, 2008). The nonlinear transformation betweenORC and PN responses to volatiles thus broadens the molecularreceptive range of a PN and underlies the first type of broadening.

In D. melanogaster, excitatory LNs (Chou et al., 2010; Shanget al., 2007) provide indirect, excitatory input to PNs from otheractivated glomeruli (Fig. 3B). Such excitatory LNs receive directinput from ORCs (Huang et al., 2010) and are broadly (Yaksi andWilson, 2010) or more narrowly (Huang et al., 2010) responsive toolfactory stimuli. Excitation is spread to PNs across the AL throughelectrical synapses (Huang et al., 2010; Yaksi and Wilson, 2010).The broadening of PN responses from their cognate ORC inputoccurs for specific stimulus-glomerulus pairs, while in response toother volatiles there remains a direct, linear relationship betweenORC and PN activity within the glomerulus (Silbering et al., 2008).Such excitatory lateral interactions have not been observed inother species, but LNs that are not GABA-immunoreactive havebeen observed in the AL of a moth (B. mori; Seki and Kanzaki, 2008)and a cockroach (P. americana; Husch et al., 2009) and may belongto a similar class of eLNs. Finally, disinhibition, involving serialinhibitory synaptic connections between glomeruli, might effec-tively broaden the molecular receptive range of a PN (c.f.Christensen et al., 1993; Avron and Rospars, 1995).

6.2.3. Sharpening versus broadening in two species

An apparent contradiction exists in the different transforma-tions imposed upon ORC inputs in the ALs of honey bees (A.

mellifera) and fruit flies (D. melanogaster). The dominant mode ofinteraction in A. mellifera appears to be inhibition that suppressesweakly activated glomeruli and sharpens the representation of anolfactory stimulus, while in D. melanogaster evidence suggests thatlateral excitation spreads activation and broadens the representa-tion. These two modes of interaction differ, one seeminglydecreasing the similarity between representations of differentolfactory stimuli and the other spreading activation across the ALand reducing the chemosensory specificity of the output channels.A possible resolution of the difference can come from consideringthe anatomy and natural context of olfaction in these animals.

As discussed above, the olfactory system of D. melanogaster mayrepresent extreme segregation, as PNs from subsets of glomerulihave synaptic connections with subsets of KCs in the MBC,producing segregated, parallel systems that interact via localcircuitry. If these channels did not interact, any subsystem wouldreceive information from only the ORCs that directly feed into it.The molecular receptive range of ORCs in turn is limited by ligand–receptor interactions, further reducing the information available toan already impoverished system. The widespread excitatory

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447440

interactions reported in the D. melanogaster AL can shareinformation between subsystems in the AL, expanding thereceptive range of each subsystem while allowing the KC networksto perform parallel, separate functions that are still poorlyunderstood. For instance, activation of the VA71 glomerulus,whose PNs project to g-lobe KCs, excites both another glomerulus(DL1) connected to the g-lobe and glomerulus VA1d, whose PNsproject to KCs associated with the a/b lobe (Lin et al., 2007; Olsenet al., 2007). By contrast, the AL of A. mellifera contains 150–180glomeruli, nearly four times the number found in an AL in D.

melanogaster (Galizia et al., 1999). Each of the main AL outputtracts in A. mellifera contains PNs axons from more glomeruli thanthe total number in the D. melanogaster AL, and while some KCclasses receive information from only one tract, others apparentlycan integrate across all glomeruli (Kirschner et al., 2006). With allavailable information feeding into the system, separation ofoverlapping representations becomes more critical and perhapsaccounts for the predominance of inhibitory interactions betweenglomerular channels in ALs of A. mellifera.

A. mellifera and other hymenoptera use an exceptionally widerange of olfactory cues, because of the number and diversity offlowers they visit throughout their range, and an additional,complex array of social odors. These insects also can performremarkable feats of olfactory discrimination, including learning toidentify rewarding from unrewarding volatile stimuli fromdifferent cultivars of the same flower based on the relative ratiosof the volatiles common to each (Wright et al., 2005). Although fliesof the genus Drosophila include generalists (e.g., D. melanogaster)and specialists in wide variety, including a species adapted to liveexclusively on the mouthparts of crabs (reviewed in Stensmyr andHansson, 2007), individual species likely do not employ finediscrimination over a wide range of olfactory stimuli like thatobserved in A. mellifera. Efficient coding in a network is a functionof the number of encoding units, the molecular receptive range ofeach, and most importantly for our analysis, the number andsimilarity of stimuli that must be encoded (c.f. Olshausen and Field,2000; Simoncelli and Olshausen, 2001). The olfactory system of A.

mellifera encodes a larger number of olfactory stimuli with a largernumber of receptors and glomeruli, while D. melanogaster employsa less numerically complex AL, and perhaps even smallersubsystems within the AL, to encode a smaller range ofbehaviorally important stimuli. The AL network of each animallikely adjusts the receptive range of each channel to achieveoptimal coding.

Theoretical studies have demonstrated that optimal tuning ofneurons in a population is a function of stimulus-dependentbehaviors (Salinas, 2006), stimulus dimensionality (Zhang andSejnowski, 1999), and the degree of noise in the sensory signal(Pouget et al., 1999). The influence of these factors on olfactorycoding can be approximated by analysis of the interaction of theanimal with volatiles in its environment, placing coding in theproper, natural context. The neurobiology of olfaction in insectsdemonstrates how evolution, from a common starting point ofinput, output, and interneurons, can adjust the properties of thenetwork to suit the particular demands of a species’ life history andenvironment.

6.3. Synchrony

Synchronous production of action potentials by the outputneurons of a neural circuit is increasingly recognized as a criticalfeature of coding in many sensory systems (Singer, 1999).Synchronous spikes in separate output channels can ‘‘bind’’ theelements of the representation of a stimulus, allowing higher-order centers to process the inputs as a single percept (Engel et al.,1997; Engel and Singer, 2001). Among insect olfactory systems,

those of locusts, moths, and flies have become favorable models forstudies of spiking synchrony. Synchronous activity is organizeddifferently in these systems, but in all three synchrony isassociated with inhibitory local circuitry (Ito et al., 2009; Leiet al., 2002; MacLeod and Laurent, 1996; Tanaka et al., 2009).

In contrast to LNs in most other insects that have been studied,LNs in the ALs of several locust species produce only graded,oscillating potentials in response to olfactory stimuli (Anton andHansson, 1996; Laurent and Naraghi, 1994). Strong, oscillatoryinhibition from these neurons constrains PNs to firing mostly inphase with the 20-Hz oscillations of the local field potential (LFP)(Laurent and Davidowitz, 1994; Wehr and Laurent, 1996) andproduces slower temporal patterns of excitation and inhibitionduring prolonged stimulation (Bazhenov et al., 2001; Laurent andDavidowitz, 1994; Wehr and Laurent, 1999). The representation ofolfactory stimuli in the locust AL thus is distributed across anevolving ensemble of PNs, a coding space that employs both theidentity of responding neurons and the temporal evolution of theresponse eventually to produce a unique output for each of severalsimilar, correlated inputs (Laurent, 1999, 2002; Mazor and Laurent,2005; Stopfer et al., 2003).

As described above (Section 5.1), each ORC in S. americana

terminates in 10–25 small glomeruli, and the dendrites of a PNsample from 10 to 25 of these channels. Although it is not yetknown how many different ORs are expressed in the ORCs of thisspecies of locust, this arrangement theoretically could produce aunique ORC input profile for each of the approximately 830 PNs inthe AL (Laurent et al., 1996a,b; Leitch and Laurent, 1996). The lackof glomeruli representing unique channels with redundant outputneurons thus allows for a huge expansion of the available codingspace between input to and output from the AL. The tuning of anindividual ORC is necessarily limited by the interaction between itsOR and the volatile ligands it binds. A PN that samples from severalORCs thus can have a greatly broadened molecular receptive range.Furthermore, a unique combination of responsive PNs, synchro-nized across many cycles of a common oscillatory signal, can beproduced for stimuli with small differences in the ORC response(Laurent, 2002).

Although LFP oscillations have been recorded in the ALs of mothsand flies, most spikes in AL PNs in these species are not fullyentrained to LFP oscillations (Christensen et al., 2003; Heinbockelet al., 1998; Ito et al., 2009; Tanaka et al., 2009; Turner et al., 2008;Wilson and Laurent, 2005). Importantly, while locust PNs producespikes at or below the 20-Hz frequency of the LFP oscillation, PNs inM. sexta (c.f. Heinbockel et al., 1998; Ito et al., 2009; Reisenman et al.,2005; Vickers et al., 1998), D. melanogaster (Bhandawat et al., 2007),and A. mellifera (Muller et al., 2002) can respond to olfactory stimuliat much higher frequencies, up to ca. 200 s�1. When responding to anatural plume of volatiles, M. sexta PNs typically respond with shortbursts of spikes and an instantaneous spiking frequency of ca.100 Hz (Vickers et al., 2001). The firing rate of single PNs in severalinsects other than locusts roughly scales with stimulus concentra-tion (see Section 6.1) (Bhandawat et al., 2007; Bichao et al., 2005;Reisenman et al., 2004, 2005; Schlief and Wilson, 2007; Silberinget al., 2008; Wilson et al., 2004), while individual S. americana PNsrespond at roughly the same firing rate across concentrations (Assisiet al., 2007; Stopfer et al., 2003). Instead, ensembles of S. americana

PNs respond to different concentrations of the same olfactorystimulus along trajectories that, while similar, evolve into distinctrepresentations over time (Stopfer et al., 2003). Thus, while LFPoscillations exist in the ALs of several species, the strict phase-lockedcode observed in S. americana is not evident in other insects andfurthermore is not easily compatible with frequency coding ofstimulus concentration.

Firing synchrony among PNs in moths (M. sexta: Christensenet al., 2003, 2000; Ito et al., 2009; Lei et al., 2002) and flies (D.

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447 441

melanogaster—the only other species studied in detail: Tanakaet al., 2009; Turner et al., 2008; Wilson and Laurent, 2005) is notfully entrained to the LFP. In response to a brief olfactory stimulus,mimicking a pulse of volatiles encountered in a natural plume(Vickers, 2006c), PNs in moth ALs produce high-frequency spikesearly in the response that are synchronized, and then synchronyfalls off quickly thereafter (Christensen et al., 2003; Lei et al., 2002).This has been examined most closely in the sex-pheromonespecific MGC in ALs of male M. sexta. In this subsystem, each of thetwo key components of the female’s sex-pheromone mixtureactivates PNs in one of the two largest MGC glomeruli and inhibitsPNs in the other (Christensen and Hildebrand, 1997; Lei et al.,2002). Stimulation with the natural mixture, thus activating boththe inhibitory and excitatory inputs to MGC PNs, increases firingsynchrony (Lei et al., 2002). Similar synchrony has been observedin the main AL in response to natural mixtures of floral volatiles,depends on the identity and intensity of volatiles, and occurs in aunique pattern for individual odors (Christensen et al., 2000; Leiet al., 2004; Riffell et al., 2009). Synchrony may subserve yetanother function in the ALs of moths. Synchronous firing betweenneurons is enhanced for pheromone (Lei et al., 2002) or plant-odor(Riffell et al., 2009) stimuli that are innately attractive. This mayrepresent a common encoding dimension for behaviorally signifi-cant odors from diverse sources such as food or conspecific mates(Martin and Hildebrand, 2010).

Multiple stimuli of longer duration (>1 s) at long (20 s)intervals evoke a two-stage response in PNs in the plant-odorresponsive glomeruli of M. sexta (Ito et al., 2009). After a briefperiod of high-frequency firing, the PN exhibits prolonged,temporally complex, lower-frequency firing. The LFP in themushroom body similarly shifts from high- to low-frequency.Under similar stimulus conditions, PNs in the ALs of D.

melanogaster also transition from high- to low-frequency spiking(Tanaka et al., 2009). In both species, individual PNs often fire inphase with these oscillations but are much less entrained than arePNs of S. americana and typically fire at higher frequency than thatof the LFP. It seems, therefore, that control of firing phase byinhibition from LNs is much weaker in ALs of M. sexta and D.

Melanogaster than in those of S. americana and at most only weaklyentrains PN firing to a global oscillation. We further discuss theimplications of the stimulus regimen necessary to generate phase-locking in these animals, similar to that in locusts, in Section 6.4.

Finally, although administration of the GABAA-receptor antag-onist picrotoxin (PTX) disrupts LFP oscillations in the AL of A.

mellifera and increases the likelihood that A. mellifera generalizesbetween closely related odors (Stopfer et al., 1997), more recentwork has shown that PTX also allows additional, weaklystimulated glomeruli to respond to an olfactory stimulus (Sachseand Galizia, 2002). Based on this side-effect, treatment with PTXhas the potential to disrupt a neural code based either onoscillatory synchronization of PNs or on the identity of responsivePNs, making behavioral pharmacological experiments difficult tointerpret. Similar work on moths (M. sexta) demonstrates that PTXdisrupts discrimination (behaviorally distinct from increasinggeneralization) between both similar and dissimilar odors, afinding that is also consistent with physiological data (Mwilariaet al., 2008). Across species, behavioral evidence for the impor-tance of LFP oscillations in olfactory coding remains elusive.

6.4. Encoding temporal features of olfactory stimuli

A flying insect experiences olfactory stimuli as brief, discontin-uous pulses, as its antennae encounter filaments of volatiles in aplume. Encoding these spatio-temporal features of a stimulus isnecessary for the flying insect to follow a plume to its source (Cardeand Willis, 2008). In this section, we consider how the olfactory

system encodes such brief, stochastic stimuli. We also suggest howdifferences in the spatio-temporal structure of stimuli importantto an insect may shape processing of olfactory information.

The ALs of moths exhibit several adaptations that improve theresolution of pulses in a sex-pheromone plume, subserving thevital task of locating the source, a calling, conspecific female moth.PNs in A. ipsilon ALs respond with extremely short lag to ORC input(Jarriault et al., 2010). In M. sexta, the ability of an MGC PN to followthe frequency and duration of pulses of sex pheromone iscorrelated with the strength of inhibitory input to the neuron(Christensen and Hildebrand, 1988; Heinbockel et al., 1999, 2004).Temporal fidelity is further improved when a moth is stimulatedwith the correct ratio of the two key pheromone components,eliciting a specific balance of excitation and inhibition (Christensenand Hildebrand, 1997; Heinbockel et al., 2004). This combinationof excitatory and inhibitory inputs causes some MGC PNs torespond with a shorter latency and higher rate of spiking,compared to stimulation with the excitatory component alone.Temporal fidelity is also enhanced by longer-lasting inhibitionfollowing a burst of spikes, during which spontaneous activity, butnot responses to subsequent pulses of pheromone, is reduced(Christensen et al., 1998; Lei and Hansson, 1999). Disrupting thisinhibition impairs a moth’s ability to follow a pheromone plume toits source (Lei et al., 2009).

ORC-to-PN synapses in D. melanogaster also act as high-passfilters, combining a high probability of neurotransmitter releasewith rapid synaptic depression to long-lasting stimuli to empha-size the onset of an olfactory stimulus (Bhandawat et al., 2007;Kazama and Wilson, 2008). Intracellular recordings in D. melano-

gaster (Kazama and Wilson, 2008) and the moth A. ipsilon (Jarriaultet al., 2010), as well as imaging of ORC and PN activity andintracellular recording of PNs in A. mellifera (Galizia and Kimmerle,2004; Sachse and Galizia, 2003), reveal that PNs respond to therising phase of the ORC response, and PN responses decline beforethe peak of ORC activity. It is likely that the inhibitory network,along with the properties of synapses in the AL, improves theability to encode rapid spatio-temporal fluctuations of a plume ofvolatiles.

The responses of KCs in the MB also reflect the emphasis on theinitial, brief contact with volatiles, and thus the brief availability ofinformation about olfactory stimuli. The responses of KCs in M.

sexta (Ito et al., 2008) and A. mellifera (Szyszka et al., 2005) occurwithin a few milliseconds of the onset of a stimulus pulse, followedby an unresponsive period. The output neurons of the MB in A.

mellifera respond with patterns differentiating rewarded fromunrewarded odors, delayed only tens of milliseconds from the ALinput to the MB (Strube-Bloss et al., 2011). A coding scheme inwhich olfactory stimuli are recognized on the basis of the identityand relative activation of PNs associated with individual glomeruli,as in classic population coding (Sanger, 2003), consolidates muchof the information in the initial response. This could allow rapid,robust classification of olfactory stimuli in a plume, where brief,unpredictable pulses of volatiles are encountered. Evidence forrapid behavioral responses to olfactory stimuli in flies (Budick andDickinson, 2006) and honey bees (Fernandez et al., 2009; Wrightet al., 2009) suggests that the information contained in shortencounters with stimuli is all that is necessary for some behaviors.

The responses of KCs in S. americana, however, are not limited tothe onset of the stimulus. An individual KC requires a large numberof coincident PN inputs in order to fire, and this criterion may notbe reached in most or even all cycles of PN responses to olfactorystimuli (Jortner et al., 2007; Perez-Orive et al., 2004, 2002).Oscillatory PN input also drives feed-forward inhibition to KCs in S.

americana, preventing them from firing in all but a brief part of thecycle (Perez-Orive et al., 2002). The net result of these processes isstimulus-driven KC activity consisting of one or a few spikes, in

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447442

stimulus-specific temporal patterns (Stopfer et al., 2003), duringeven a sustained olfactory stimulus. The identity of a stimulus thusis encoded first in a distributed population of PNs and then in asparse population of KCs, in which the membership of bothpopulations evolves over several cycles of the oscillatory signal. Animportant outcome of this processing scheme is the theoreticalability of the olfactory system of S. americana to produce a uniquecode for each of many similar stimuli, which could in turn beassociated with a variety of other information about a source ofvolatiles (Laurent, 2002).

The locust S. americana therefore instantiates a unique code in aunique olfactory network, distinct from those of other insects suchas sphinx moths, fruit flies, and honey bees. While olfactoryinformation is concentrated early in a stimulus pulse for thosespecies, representations of olfactory stimuli become more distinctover time in both the AL and MBs of S. americana (Perez-Orive et al.,2002; Stopfer et al., 2003). We suggest that this difference may beadaptive in the natural contexts in which acridids such as locustsuse olfactory information.

The olfactory systems of moths, bees, and flies appear to beadapted to encode olfactory stimuli upon brief, stochasticencounters with discontinuous pulses in plumes of volatiles.Locusts, however, impose an internal timing signal to produce areliable temporal code throughout prolonged contact with astimulus (Laurent, 2002). The addition of the temporal dimensionin the olfactory coding scheme used by locusts can greatly expandthe coding space allotted for classifying stimuli. As opposed to acode based on the identity and relative activation of respondingglomeruli (Galan et al., 2004; Sachse and Galizia, 2003), integrationover many cycles creates an exponentially large number of ‘virtual’PN populations, allowing for a different active ensemble at eachcycle (Laurent et al., 2001). This increased coding space may beadaptive for a voracious, generalist insect such as a locust. Locustshave a wide range and a diverse diet of vegetation (Simpson andRaubenheimer, 2000). Vegetative tissues of plants emit complex,species-specific mixtures of volatiles, both when intact and whendamaged as the insects feed (Bruce et al., 2005). Locusts canassociate volatiles with the nutritional content of their source(Behmer et al., 2005) and can use associations to balance their diet(Dukas and Bernays, 2000; Raubenheimer and Tucker, 1997). Thereis little evidence for olfaction-guided aerial navigation in theAcrididae (Helms et al., 2003), and several authors have suggestedthat locust foraging is primarily visually driven (Chapman, 1988).Thus, for locusts, exposure to significant olfactory stimuli occursprimarily in the context of prolonged feeding bouts in close contactwith the source of volatiles, allowing for full use of the coding spaceavailable in the temporal structure of olfactory responses.Although other insects encounter volatiles in similar circum-stances, aerial foragers must rely more frequently on identificationof brief, stochastic pulses of volatiles and therefore require a codingsystem adapted to the spatio-temporal structure of a plume.

Finally, observations of PN spikes phase-locked to LFP oscilla-tions in D. melanogaster and M. sexta are instructive (Ito et al., 2009;Tanaka et al., 2009). Oscillations emerged only after multiple(>10), 2-s pulses of an olfactory stimulus delivered at a lowerairspeed than used in other investigations. Oscillatory synchronywas also observed after an initial, phasic spiking response toolfactory stimuli, suggesting that the two coding schemes mayoperate in parallel in a single species. Perhaps encoding based onoscillations requires longer contact with a stimulus, such as mayoccur when an animal is feeding from a flower. A finalreconciliation of these two schemes would require a mechanismwherein encoding of stimuli associated with reward underconditions of long, repeated exposure (facilitating oscillations)are transformed to allow identification under odor-plume condi-tions that do not allow for oscillations to develop. Honey bees

appear to be capable of this transformation. Bees trained to longpulses of an odor perform well in a discrimination test, even whentested with short pulses of the odor (Fernandez et al., 2009).

In summary, we suggest that a trade-off is likely between acoding scheme optimized for discriminating fewer olfactorystimuli while flying and one that can produce unique, decorrelatedcodes for volatiles from many, highly similar food sources uponprolonged exposure to the stimuli.

7. Concluding remarks

Progress to date in insect olfactory neurobiology and neu-roethology points to future direction for the field. We suggest thatthe principles and models of insect olfaction will be expanded andelaborated as the range of variation among insect species isconsidered. A comparative approach, grounded in natural behav-ior, can resolve conflicting results and guide inquiry in fruitfuldirections. Future progress will be made through not only newdiscoveries but also direct comparison between species whenpossible. In evaluating comparative work, however, one mustalways be wary that findings in one species may not be applicableto another. Finally, we have attempted to illustrate the broad,interdisciplinary nature of work in this field. Insect olfaction isespecially fertile ground for interplay among studies in ecology,evolution, and behavior that set the parameters for the problemsinsect olfactory systems must solve.

Acknowledgments

We thank Tom Christensen and Nick Strausfeld for helpfulcomments in preparing this manuscript, and Charles Hedgcock forassisting in creating the figures. JPM was supported by NIHfellowship F31DC009722 during the preparation of this review.Our current research in this area is supported by NIH grant R01-DC-02751 (to JGH).

References

Abel, R., Rybak, J., Menzel, R., 2001. Structure and response patterns of olfactoryinterneurons in the honeybee, Apis mellifera. J. Comp. Neurol. 437, 363–383.

Ache, B.W., Young, J.M., 2005. Olfaction: diverse species, conserved principles.Neuron 48, 417–430.

Allmann, S., Baldwin, I.T., 2010. Insects betray themselves in nature to predators byrapid isomerization of green leaf volatiles. Science 329, 1075–1078.

Andersson, M.N., Larsson, M.C., Blazenec, M., Jakus, R., Zhang, Q.-H., Schlyter, F.,2010. Peripheral modulation of pheromone response by inhibitory host com-pound in a beetle. J. Exp. Biol. 213, 3332–3339.

Anton, S., Hansson, B.S., 1996. Antennal lobe interneurons in the desert locustSchistocerca gregaria (Forskal): processing of aggregation pheromones in adultmales and females. J. Comp. Neurol. 370, 85–96.

Anton, S., Dufour, M.C., Gadenne, C., 2007. Plasticity of olfactory-guided behaviourand its neurobiological basis: lessons from moths and locusts. Entomol. Exp.Appl. 123, 1–11.

Assisi, C., Stopfer, M., Laurent, G., Bazhenov, M., 2007. Adaptive regulation ofsparseness by feedforward inhibition. Nat. Neurosci. 10, 1176–1184.

Avron, E., Rospars, J.P., 1995. Modeling insect olfactory neuron signaling by anetwork utilizing disinhibition. Bio Systems 36, 101–108.

Baker, T.C., Ochieng’, S.A., Cosse, A.A., Lee, S.G., Todd, J.L., Quero, C., Vickers, N.J.,2004. A comparison of responses from olfactory receptor neurons of Heliothissubflexa and Heliothis virescens to components of their sex pheromone. J. Comp.Physiol. A190, 155–165.

Baker, T.C., 2008. Balanced olfactory antagonism as a concept for understandingevolutionary shifts in moth sex pheromone blends. J. Chem. Ecol. 34, 971–981.

Barrozo, R.B., Gadenne, C., Anton, S., 2010. Switching attraction to inhibition:mating-induced reversed role of sex pheromone in an insect. J. Exp. Biol.213, 2933–2939.

Bartelt, R.J., Schaner, A.M., Jackson, L.L., 1985. Cis-vaccenyl acetate as an aggregationpheromone in Drosophila melanogaster. J. Chem. Ecol. 11, 1747–1756.

Bazhenov, M., Stopfer, M., Rabinovich, M., Huerta, R., Abarbanel, H.D.I., Sejnowski,T.J., Laurent, G., 2001. Model of transient oscillatory synchronization in thelocust antennal lobe. Neuron 30, 553–567.

Behmer, S.T., Belt, C.E., Shapiro, M.S., 2005. Variable rewards and discriminationability in an insect herbivore: what and how does a hungry locust learn? J. Exp.Biol. 208, 3463–3473.

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447 443

Benton, R., Sachse, S., Michnick, S.W., Vosshall, L.B., 2006. Atypical membranetopology and heteromeric function of Drosophila odorant receptors in vivo.PLoS Biol. 4, 240–257.

Benton, R., Vannice, K.S., Vosshall, L.B., 2007. An essential role for a cd36-relatedreceptor in pheromone detection in Drosophila. Nature 450, 289–293.

Benton, R., Vannice, K.S., Gomez-Diaz, C., Vosshall, L.B., 2009. Variant ionotropicglutamate receptors as chemosensory receptors in Drosophila. Cell 136, 149–162.

Berlocher, S.H., Feder, J.L., 2002. Sympatric speciation in phytophagous insects:moving beyond controversy? Annu. Rev. Entomol. 47, 773–815.

Bhandawat, V., Olsen, S.R., Gouwens, N.W., Schlief, M.L., Wilson, R.I., 2007. Sensoryprocessing in the Drosophila antennal lobe increases reliability and separabilityof ensemble odor representations. Nat. Neurosci. 10, 1474–1482.

Bichao, H., Borg-Karlson, A.K., Wibe, A., Araujo, J., Mustaparta, H., 2005. Molecularreceptive ranges of olfactory receptor neurones responding selectively toterpenoids, aliphatic green leaf volatiles and aromatic compounds, in thestrawberry blossom weevil Anthonomus rubi. Chemoecology 15, 211–226.

Bleeker, M.A.K., Smid, H.M., Steidle, J.L.M., Kruidhof, H.M., Van Loon, J.J.A., Vet,L.E.M., 2006. Differences in memory dynamics between two closely relatedparasitoid wasp species. Anim. Behav. 71, 1343–1350.

Bodin, A., Barrozo, R.B., Couton, L., Lazzari, C.R., 2008. Temporal modulation andadaptive control of the behavioural response to odours in Rhodnius prolixus. J.Insect Physiol. 54, 1343–1348.

Bohbot, J., Pitts, R.J., Kwon, H.W., Rutzler, M., Robertson, H.M., Zwiebel, L.J., 2007.Molecular characterization of the Aedes aegypti odorant receptor gene family.Insect Mol. Biol. 16, 525–537.

Bruce, T.J.A., Wadhams, L.J., Woodcock, C.M., 2005. Insect host location: a volatilesituation. Trends Plant Sci. 10, 269–274.

Budick, S.A., Dickinson, M.H., 2006. Free-flight responses of Drosophila melanogasterto attractive odors. J. Exp. Biol. 209, 3001–3017.

Cameron, S.A., 2004. Phylogeny and biology of neotropical orchid bees (Euglossini).Annu. Rev. Entomol. 49, 377–404.

Carde, R.T., Willis, M.A., 2008. Navigational strategies used by insects to find distant,wind-borne sources of odor. J. Chem. Ecol. 34, 854–866.

Chance, F.S., Abbott, L.F., Reyes, A.D., 2002. Gain modulation from backgroundsynaptic input. Neuron 35, 773–782.

Chapman, R.F., 1988. Sensory aspects of host-plant recognition by Acridoidea –questions associated with the multiplicity of receptors and variability ofresponse. J. Insect Physiol. 34, 167–174.

Chittka, L., Raine, N.E., 2006. Recognition of flowers by pollinators. Curr. Opin. PlantBiol. 9, 428–435.

Chou, Y.H., Spletter, M.L., Yaksi, E., Leong, J.C.S., Wilson, R.I., Luo, L.Q., 2010. Diversityand wiring variability of olfactory local interneurons in the Drosophila antennallobe. Nat. Neurosci. 13, 439–449.

Christensen, T.A., Hildebrand, J.G., 1988. Frequency coding by central olfactoryneurons in the sphinx moth Manduca sexta. Chem. Senses 13, 123–130.

Christensen, T.A., Waldrop, B.R., Harrow, I.D., Hildebrand, J.G., 1993. Local inter-neurons and information-processing in the olfactory glomeruli of the mothManduca sexta. J. Comp. Physiol. A 173, 385–399.

Christensen, T.A., Hildebrand, J.G., 1997. Coincident stimulation with pheromonecomponents improves temporal pattern resolution in central olfactory neurons.J. Neurophysiol. 77, 775–781.

Christensen, T.A., Waldrop, B.R., Hildebrand, J.G., 1998. Multitasking in the olfactorysystem: context-dependent responses to odors reveal dual GABA-regulatedcoding mechanisms in single olfactory projection neurons. J. Neurosci. 18,5999–6008.

Christensen, T.A., Pawlowski, V.M., Lei, H., Hildebrand, J.G., 2000. Multi-unit record-ings reveal context-dependent modulation of synchrony in odor-specific neuralensembles. Nat. Neurosci. 3, 927–931.

Christensen, T.A., Hildebrand, J.G., 2002. Pheromonal and host-odor processing inthe insect antennal lobe: how different? Curr. Opin. Neurobiol. 12, 393–399.

Christensen, T.A., Lei, H., Hildebrand, J.G., 2003. Coordination of central odorrepresentations through transient non-oscillatory synchronization of glomer-ular output neurons. Proc. Natl. Acad. Sci. U.S.A. 100, 11076–11081.

Comer, C.M., Robertson, R.M., 2001. Identified nerve cells and insect behavior. Prog.Neurobiol. 63, 409–439.

Couto, A., Alenius, M., Dickson, B.J., 2005. Molecular, anatomical, and functionalorganization of the Drosophila olfactory system. Curr. Biol. 15, 1535–1547.

Dacks, A.M., Christensen, T.A., Agricola, H.J., Wollweber, L., Hildebrand, J.G., 2005.Octopamine-immunoreactive neurons in the brain and subesophageal ganglionof the hawkmoth Manduca sexta. J. Comp. Neurol. 488, 255–268.

Dacks, A.M., Christensen, T.A., Hildebrand, J.G., 2006. Phylogeny of a serotonin-immunoreactive neuron in the primary olfactory center of the insect brain. J.Comp. Neurol. 498, 727–746.

Dacks, A.M., Christensen, T.A., Hildebrand, J.G., 2008. Modulation of olfactoryinformation processing in the antennal lobe of Manduca sexta by serotonin. J.Neurophysiol. 99, 2077–2085.

Dacks, A.M., Green, D.S., Root, C.M., Nighorn, A.J., Wang, J.W., 2009. Serotoninmodulates olfactory processing in the antennal lobe of Drosophila. J. Neuro-genet. 23, 366–377.

Dacks, A.M., Reisenman, C.E., Paulk, A.C., Nighorn, A.J., 2010. Histamine-immuno-reactive local neurons in the antennal lobes of the Hymenoptera. J. Comp.Neurol. 518, 2917–2933.

Daly, K.C., Christensen, T.A., Lei, H., Smith, B.H., Hildebrand, J.G., 2004. Learningmodulates the ensemble representations for odors in primary olfactory net-works. Proc. Natl. Acad. Sci. U.S.A. 101, 10476–10481.

Datta, S.R., Vasconcelos, M.L., Ruta, V., Luo, S., Wong, A., Demir, E., Flores, J., Balonze,K., Dickson, B.J., Axel, R., 2008. The Drosophila pheromone cVA activates asexually dimorphic neural circuit. Nature 452, 473–477.

de Bruyne, M., Foster, K., Carlson, J.R., 2001. Odor coding in the Drosophila antenna.Neuron 30, 537–552.

de Bruyne, M., Baker, T.C., 2008. Odor detection in insects: volatile codes. J. Chem.Ecol. 34, 882–897.

Deisig, N., Giurfa, M., Lachnit, H., Sandoz, J.C., 2006. Neural representation ofolfactory mixtures in the honeybee antennal lobe. Eur. J. Neurosci. 24, 1161–1174.

Deisig, N., Giurfa, M., Sandoz, J.C., 2010. Antennal lobe processing increases sepa-rability of odor mixture representations in the honeybee. J. Neurophysiol. 103,2185–2194.

Dekker, T., Ibba, I., Siju, K.P., Stensmyr, M.C., Hansson, B.S., 2006. Olfactory shiftsparallel superspecialism for toxic fruit in Drosophila melanogaster sibling, D.sechellia. Curr. Biol. 16, 101–109.

Demir, E., Dickson, B.J., 2005. Fruitless splicing specifies male courtship behavior inDrosophila. Cell 121, 785–794.

Dethier, V.G., 1987. Sniff, flick, and pulse: an appreciation of interruption. Proc. Am.Philos. Soc. 131, 159–176.

Distler, P., 1989. Histochemical-demonstration of GABA-like immunoreactivity incobalt labeled neuron individuals in the insect olfactory pathway. Histochem-istry 91, 245–249.

Dukas, R., Bernays, E.A., 2000. Learning improves growth rate in grasshoppers. Proc.Natl. Acad. Sci. U.S.A. 97, 2637–2640.

Dukas, R., 2005. Experience improves courtship in male fruit flies. Anim. Behav. 69,1203–1209.

Dukas, R., 2008. Evolutionary biology of insect learning. Annu. Rev. Entomol. 53,145–160.

Ehmer, B., Gronenberg, W., 2002. Segregation of visual input to the mushroombodies in the honeybee (Apis mellifera). J. Comp. Neurol. 451, 362–373.

Ejima, A., Smith, B.P.C., Lucas, C., Van Naters, W.V., Miller, C.J., Carlson, J.R., Levine,J.D., Griffith, L.C., 2007. Generalization of courtship learning in Drosophila ismediated by cis-vaccenyl acetate. Curr. Biol. 17, 599–605.

Eltz, T., Zimmermann, Y., Pfeiffer, C., Pech, J.R., Twele, R., Francke, W., Quezada-Euan,J.J., Lunau, K., 2008. An olfactory shift is associated with male perfume differ-entiation and species divergence in orchid bees. Curr. Biol. 18, 1844–1848.

Endler, A., Liebig, J., Schmitt, T., Parker, J.E., Jones, G.R., Schreier, P., Holldobler, B.,2004. Surface hydrocarbons of queen eggs regulate worker reproduction in asocial insect. Proc. Natl. Acad. Sci. U.S.A. 101, 2945–2950.

Engel, A.K., Roelfsema, P.R., Fries, P., Brecht, M., Singer, W., 1997. Binding andresponse selection in the temporal domain – a new paradigm for neurobiologi-cal research. Theor. Biosci. 116, 241–266.

Engel, A.K., Singer, W., 2001. Temporal binding and the neural correlates of sensoryawareness. Trends Cogn. Sci. 5, 16–25.

Farina, W.M., Gruter, C., Acosta, L., Mc Cabe, S., 2007. Honeybees learn floral odorswhile receiving nectar from foragers within the hive. Naturwissenschaften 94,55–60.

Fernandez, P.C., Locatelli, F.F., Person-Rennell, N., Deleo, G., Smith, B.H., 2009.Associative conditioning tunes transient dynamics of early olfactory processing.J. Neurosci. 29, 10191–10202.

Ferveur, J.F., 2005. Cuticular hydrocarbons: their evolution and roles in Drosophilapheromonal communication. Behav. Genet. 35, 279–295.

Fishilevich, E., Vosshall, L.B., 2005. Genetic and functional subdivision of theDrosophila antennal lobe. Curr. Biol. 15, 1548–1553.

Fonta, C., Sun, X.J., Masson, C., 1993. Morphology and spatial distribution of beeantennal lobe interneurones responsive to odours. Chem. Senses 18, 101–119.

Fox, A.N., Pitts, R.J., Robertson, H.M., Carlson, J.R., Zwiebel, L.J., 2001. Candidateodorant receptors from the malaria vector mosquito Anopheles gambiae andevidence of down-regulation in response to blood feeding. Proc. Natl. Acad. Sci.U.S.A. 98, 14693–14697.

Frambach, I., Schurmann, F.W., 2004. Separate distribution of deutocerebral pro-jection neurons in the mushroom bodies of the cricket brain. Acta Biol. Hung.55, 21–29.

Fukushima, R., Kanzaki, R., 2009. Modular subdivision of mushroom bodies bykenyon cells in the silkmoth. J. Comp. Neurol. 513, 315–330.

Gadenne, C., Dufour, M.C., Anton, S., 2001. Transient post-mating inhibition ofbehavioural and central nervous responses to sex pheromone in an insect. Proc.R. Soc. Lond. B. 268, 1631–1635.

Galan, R.F., Sachse, S., Galizia, C.G., Herz, A.V.M., 2004. Odor-driven attractordynamics in the antennal lobe allow for simple and rapid olfactory patternclassification. Neural Comput. 16, 999–1012.

Galizia, C.G., McIlwrath, S.L., Menzel, R., 1999. A digital three-dimensional atlas ofthe honeybee antennal lobe based on optical sections acquired by confocalmicroscopy. Cell Tissue Res. 295, 383–394.

Galizia, C.G., Kimmerle, B., 2004. Physiological and morphological characterizationof honeybee olfactory neurons combining electrophysiology, calcium imagingand confocal microscopy. J. Comp. Physiol. A 190, 21–38.

Galizia, C.G., Rossler, W., 2010. Parallel olfactory systems in insects: anatomy andfunction. Annu. Rev. Entomol. 55, 399–420.

Ganeshina, O., Menzel, R., 2001. GABA-immunoreactive neurons in the mushroombodies of the honeybee: an electron microscopic study. J. Comp. Neurol. 437,335–349.

Gao, Q., Yuan, B.B., Chess, A., 2000. Convergent projections of Drosophila olfactoryneurons to specific glomeruli in the antennal lobe. Nat. Neurosci. 3, 780–785.

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447444

Gatellier, L., Nagao, T., Kanzaki, R., 2004. Serotonin modifies the sensitivity of themale silkmoth to pheromone. J. Exp. Biol. 207, 2487–2496.

Getz, W.M., Akers, R.P., 1994. Honeybee olfactory sensilla behave as integratedprocessing units. Behav. Neural Biol. 61, 191–195.

Ghaninia, M., Ignell, R., Hansson, B.S., 2007. Functional classification and centralnervous projections of olfactory receptor neurons housed in antennal trichoidsensilla of female yellow fever mosquitoes, Aedes aegypti. Eur. J. Neurosci. 26,1611–1623.

Ghaninia, M., Larsson, M., Hansson, B.S., Ignell, R., 2008. Natural odor ligands forolfactory receptor neurons of the female mosquito Aedes aegypti: use of gaschromatography-linked single sensillum recordings. J. Exp. Biol. 211, 3020–3027.

Gil, M., De Marco, R.J., 2005. Olfactory learning by means of trophallaxis in Apismellifera. J. Exp. Biol. 208, 671–680.

Giurfa, M., 2007. Behavioral and neural analysis of associative learning in thehoneybee: a taste from the magic well. J. Comp. Physiol. A 193, 801–824.

Goldman, A.L., van Naters, W.V., Lessing, D., Warr, C.G., Carlson, J.R., 2005. Coex-pression of two functional odor receptors in one neuron. Neuron 45, 661–666.

Grimaldi, D.A., Engel, M.S., 2005. Evolution of the Insects. Cambridge UniversityPress, Cambridge.

Groh, C., Rossler, W., 2011. Comparison of microglomerular structures in themushroom body calyx of neopteran insects. Arthropod Struct. Dev. 40, 358–367.

Gronenberg, W., 1999. Modality-specific segregation of input to ant mushroombodies. Brain Behav. Evol. 54, 85–95.

Gronenberg, W., Holldobler, B., 1999. Morphologic representation of visual andantennal information in the ant brain. J. Comp. Neurol. 412, 229–240.

Gronenberg, W., 2001. Subdivisions of hymenopteran mushroom body calyces bytheir afferent supply. J. Comp. Neurol. 435, 474–489.

Grunewald, B., 1999. Morphology of feedback neurons in the mushroom body of thehoneybee, Apis mellifera. J. Comp. Neurol. 404, 114–126.

Guo, S., Kim, J., 2007. Molecular evolution of Drosophila odorant receptor genes. Mol.Biol. Evol. 24, 1198–1207.

Ha, T.S., Smith, D.P., 2006. A pheromone receptor mediates 11-cis-vaccenyl acetate-induced responses in Drosophila. J. Neurosci. 26, 8727–8733.

Hallem, E.A., Carlson, J.R., 2006. Coding of odors by a receptor repertoire. Cell 125,143–160.

Hansson, B.S., Ochieng0 , S.A., Grosmaitre, X., Anton, S., Njagi, P.G.N., 1996. Physio-logical responses and central nervous projections of antennal olfactory receptorneurons in the adult desert locust, Schistocerca gregaria (Orthoptera: Acrididae).J. Comp. Physiol. A 179, 157–167.

Heinbockel, T., Kloppenburg, P., Hildebrand, J.G., 1998. Pheromone-evoked poten-tials and oscillations in the antennal lobes of the sphinx moth Manduca sexta. J.Comp. Physiol. A 182, 703–714.

Heinbockel, T., Christensen, T.A., Hildebrand, J.G., 1999. Temporal tuning of odorresponses in pheromone-responsive projection neurons in the brain of thesphinx moth Manduca sexta. J. Comp. Neurol. 409, 1–12.

Heinbockel, T., Christensen, T.A., Hildebrand, J.G., 2004. Representation of binarypheromone blends by glomerulus-specific olfactory projection neurons. J.Comp. Physiol. A 190, 1023–1037.

Heisenberg, M., 2003. Mushroom body memoir: from maps to models. Nat. Rev.Neurosci. 4, 266–275.

Helms, J.B., Booth, C.M., Rivera, J., Siegler, J.A., Wuellner, S., Whitman, D.W., 2003.Lubber grasshoppers, Romalea microptera (Beauvois), orient to plant odors in awind tunnel. J. Orthopt. Res. 12, 135–140.

Hildebrand, J.G., Shepherd, G.M., 1997. Mechanisms of olfactory discrimination:converging evidence for common principles across phyla. Annu. Rev. Neurosci.20, 595–631.

Hillier, N.K., Vickers, N.J., 2004. The role of heliothine hairpencil compounds infemale Heliothis virescens (Lepidoptera: Noctuidae) behavior and mate accep-tance. Chem. Senses 29, 499–511.

Holldobler, B., Wilson, E.O., 2009. The Superorganism: the Beauty, Elegance, andStrangeness of Insect Societies. W. W. Norton, New York.

Homberg, U., Kingan, T.G., Hildebrand, J.G., 1987. Immunocytochemistry of GABA inthe brain and subesophageal ganglion of Manduca sexta. Cell Tissue Res. 248, 1–24.

Homberg, U., Montague, R.A., Hildebrand, J.G., 1988. Anatomy of antenno-cerebralpathways in the brain of the sphinx moth Manduca sexta. Cell Tissue Res. 254,255–281.

Hopfield, J.J., 1995. Pattern-recognition computation using action-potential timingfor stimulus representation. Nature 376, 33–36.

Hoskins, S.G., Homberg, U., Kingan, T.G., Christensen, T.A., Hildebrand, J.G., 1986.Immunocytochemistry of GABA in the antennal lobes of the sphinx mothManduca sexta. Cell Tissue Res. 244, 243–252.

Huang, J., Zhang, W., Qiao, W.H., Hu, A.Q., Wang, Z.R., 2010. Functional connectivityand selective odor responses of excitatory local interneurons in Drosophilaantennal lobe. Neuron 67, 1021–1033.

Hunt, G.J., 2007. Flight and fight: a comparative view of the neurophysiology andgenetics of honey bee defensive behavior. J. Insect Physiol. 53, 399–410.

Husch, A., Paehler, M., Fusca, D., Paeger, L., Kloppenburg, P., 2009. Distinctelectrophysiological properties in subtypes of nonspiking olfactory local inter-neurons correlate with their cell type-specific Ca2+ current profiles. J. Neuro-physiol. 102, 2834–2845.

Ignell, R., 2001. Monoamines and neuropeptides in antennal lobe interneurons ofthe desert locust, Schistocerca gregaria: an immunocytochemical study. CellTissue Res. 306, 143–156.

Ignell, R., Anton, S., Hansson, B.S., 2001. The antennal lobe of Orthoptera – anatomyand evolution. Brain Behav. Evol. 57, 1–17.

Ito, I., Ong, R.C.Y., Raman, B., Stopfer, M., 2008. Sparse odor representation andolfactory learning. Nat. Neurosci. 11, 1177–1184.

Ito, I., Bazhenov, M., Ong, R.C.Y., Raman, B., Stopfer, M., 2009. Frequency transitionsin odor-evoked neural oscillations. Neuron 64, 692–706.

Ito, K., Awano, W., Suzuki, K., Hiromi, Y., Yamamoto, D., 1997. The Drosophilamushroom body is a quadruple structure of clonal units each of whichcontains a virtually identical set of neurones and glial cells. Development124, 761–771.

Iwano, M., Hill, E.S., Mori, A., Mishima, T., Mishima, T., Ito, K., Kanzaki, R., 2010.Neurons associated with the flip-flop activity in the lateral accessory lobe andventral protocerebrum of the silkworm moth brain. J. Comp. Neurol. 518, 366–388.

Jallon, J.M., Antony, C., Benamar, O., 1981. An anti-aphrodisiac produced by Dro-sophila melanogaster males and transferred to females during copulation. C. R.Acad. Sci. III 292, 1147–1149.

Jarriault, D., Gadenne, C., Lucas, P., Rospars, J.P., Anton, S., 2010. Transformation ofthe sex pheromone signal in the noctuid moth Agrotis ipsilon: from peripheralinput to antennal lobe output. Chem. Senses 35, 705–715.

Jefferis, G.S.X.E., Marin, E.C., Stocker, R.F., Luo, L.Q., 2001. Target neuron prespeci-fication in the olfactory map of Drosophila. Nature 414, 204–208.

Jefferis, G.S.X.E., Potter, C.J., Chan, A.I., Marin, E.C., Rohlfing, T., Maurer, C.R., Luo, L.Q.,2007. Comprehensive maps of Drosophila higher offactory centers: spatiallysegregated fruit and pheromone representation. Cell 128, 1187–1203.

Joerges, J., Kuttner, A., Galizia, C.G., Menzel, R., 1997. Representations of odours andodour mixtures visualized in the honeybee brain. Nature 387, 285–288.

Jones, W.D., Nguyen, T.A.T., Kloss, B., Lee, K.J., Vosshall, L.B., 2005. Functionalconservation of an insect odorant receptor gene across 250 million years ofevolution. Curr. Biol. 15, R119–R121.

Jortner, R.A., Farivar, S.S., Laurent, G., 2007. A simple connectivity scheme for sparsecoding in an olfactory system. J. Neurosci. 27, 1659–1669.

Kanzaki, R., Shibuya, T., 1983. Olfactory neural pathway and sexual pheromoneresponses in the deutocerebrum of the male silkworm moth, Bombyx mori(Lepidoptera, Bombycidae). Appl. Entomol. Zool. 18, 131–133.

Kanzaki, R., Arbas, E.A., Hildebrand, J.G., 1991. Physiology and morphology ofdescending neurons in pheromone-processing olfactory pathways in the malemoth Manduca sexta. J. Comp. Physiol. A 169, 1–14.

Kanzaki, R., Ikeda, A., Shibuya, T., 1994. Morphological and physiological-propertiesof pheromone-triggered flipflopping descending interneurons of the malesilkworm moth, bombyx-mori. J. Comp. Physiol. A 175, 1–14.

Kanzaki, R., Mishima, T., 1996. Pheromone-triggered ‘flipflopping’ neural signalscorrelate with activities of neck motor neurons of a male moth, Bombyx mori.Zool. Sci. 13, 79–87.

Kanzaki, R., Soo, K., Seki, Y., Wada, S., 2003. Projections to higher olfactory centersfrom subdivisions of the antennal lobe macroglomerular complex of the malesilkmoth. Chem. Senses 28, 113–130.

Kaupp, U.B., 2010. Olfactory signalling in vertebrates and insects: differences andcommonalities. Nat. Rev. Neurosci. 11, 188–200.

Kazama, H., Wilson, R.I., 2008. Homeostatic matching and nonlinear amplificationat identified central synapses. Neuron 58, 401–413.

Kelber, C., Rossler, W., Kleineidam, C.J., 2006. Multiple olfactory receptor neuronsand their axonal projections in the antennal lobe of the honeybee Apis mellifera.J. Comp. Neurol. 496, 395–405.

Kent, K.S., Hoskins, S.G., Hildebrand, J.G., 1987. A novel serotonin-immunoreactiveneuron in the antennal lobe of the sphinx moth Manduca sexta persiststhroughout postembryonic life. J. Neurobiol. 18, 451–465.

Kirschner, S., Kleineidam, C.J., Zube, C., Rybak, J., Grunewald, B., Rossler, W., 2006.Dual olfactory pathway in the honeybee, Apis mellifera. J. Comp. Neurol. 499,933–952.

Kloppenburg, P., Ferns, D., Mercer, A.R., 1999. Serotonin enhances central olfactoryneuron responses to female sex pheromone in the male sphinx moth Manducasexta. J. Neurosci. 19, 8172–8181.

Klowden, M.J., 1996. Endogenous factors regulating mosquito host-seeking behav-iour. Ciba Found. Symp. 200, 212–225.

Koehl, M.A.R., 2001. Transitions in function at low reynolds number: hair-bearinganimal appendages. Math. Meth. Appl. Sci. 24, 1523–1532.

Koehl, M.A.R., 2006. The fluid mechanics of arthropod sniffing in turbulent odorplumes. Chem. Senses 31, 93–105.

Krieger, J., Klink, O., Mohl, C., Raming, K., Breer, H., 2003. A candidate olfactoryreceptor subtype highly conserved across different insect orders. J. Comp.Physiol. A 189, 519–526.

Kristoffersen, L., Hansson, B.S., Anderbrant, O., Larsson, M.C., 2008a. Aglomerularhemipteran antennal lobes- basic neuroanatomy of a small nose. Chem. Senses33, 771–778.

Kristoffersen, L., Larsson, M.C., Anderbrant, O., 2008b. Functional characteristics of atiny but specialized olfactory system: olfactory receptor neurons of carrotpsyllids (Homoptera: Triozidae). Chem. Senses 33, 759–769.

Krofczik, S., Menzel, R., Nawrot, M.P., 2009. Rapid odor processing in the honeybeeantennal lobe network. Front. Comput. Neurosci. 2, 9.

Kurtovic, A., Widmer, A., Dickson, B.J., 2007. A single class of olfactory neuronsmediates behavioural responses to a Drosophila sex pheromone. Nature 446,542–546.

Larsson, M.C., Domingos, A.I., Jones, W.D., Chiappe, M.E., Amrein, H., Vosshall, L.B.,2004. Or83b encodes a broadly expressed odorant receptor essential for Dro-sophila olfaction. Neuron 43, 703–714.

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447 445

Laughlin, S., 1981. A simple coding procedure enhances a neurons informationcapacity. Z. Naturforsch. C 36, 910–912.

Laurent, G., Davidowitz, H., 1994. Encoding of olfactory information with oscillatingneural assemblies. Science 265, 1872–1875.

Laurent, G., Naraghi, M., 1994. Odorant-induced oscillations in the mushroombodies of the locust. J. Neurosci. 14, 2993–3004.

Laurent, G., Wehr, M., Davidowitz, H., 1996a. Temporal representations of odors inan olfactory network. J. Neurosci. 16, 3837–3847.

Laurent, G., Wehr, M., MacLeod, K., Stopfer, M., Leitch, B., Davidowitz, H., 1996b.Dynamic encoding of odors with oscillating neuronal assemblies in the locustbrain. Biol. Bull. 191, 70–75.

Laurent, G., 1999. A systems perspective on early olfactory coding. Science 286,723–728.

Laurent, G., Stopfer, M., Friedrich, R.W., Rabinovich, M.I., Volkovskii, A., Abarbanel,H.D.I., 2001. Odor encoding as an active, dynamical process: experiments,computation, and theory. Annu. Rev. Neurosci. 24, 263–297.

Laurent, G., 2002. Olfactory network dynamics and the coding of multidimensionalsignals. Nat. Rev. Neurosci. 3, 884–895.

Le Conte, Y., Hefetz, A., 2008. Primer pheromones in social Hymenoptera. Annu. Rev.Entomol. 53, 523–542.

Leal, W.S., 1996. Chemical communication in scarab beetles: reciprocal behavioralagonist-antagonist activities of chiral pheromones. Proc. Natl. Acad. Sci. U.S.A.93, 12112–12115.

Lee, T., Lee, A., Luo, L.Q., 1999. Development of the Drosophila mushroom bodies:sequential generation of three distinct types of neurons from a neuroblast.Development 126, 4065–4076.

Lei, H., Hansson, B.S., 1999. Central processing of pulsed pheromone signals byantennal lobe neurons in the male moth Agrotis segetum. J. Neurophysiol. 81,1113–1122.

Lei, H., Christensen, T.A., Hildebrand, J.G., 2002. Local inhibition modulates odor-evoked synchronization of glomerulus-specific output neurons. Nat. Neurosci.5, 557–565.

Lei, H., Christensen, T.A., Hildebrand, J.G., 2004. Spatial and temporal organization ofensemble representations for different odor classes in the moth antennal lobe. J.Neurosci. 24, 11108–11119.

Lei, H., Vickers, N., 2008. Central processing of natural odor mixtures in insects. J.Chem. Ecol. 34, 915–927.

Lei, H., Riffell, J.A., Gage, S.L., Hildebrand, J.G., 2009. Contrast enhancement ofstimulus intermittency in a primary olfactory network and its behavioralsignificance. J. Biol. 8, 21.

Leiss, F., Groh, C., Butcher, N.J., Meinertzhagen, I.A., Tavosanis, G., 2009. Synapticorganization in the adult Drosophila mushroom body calyx. J. Comp. Neurol.517, 808–824.

Leitch, B., Laurent, G., 1996. GABAergic synapses in the antennal lobe and mush-room body of the locust olfactory system. J. Comp. Neurol. 372, 487–514.

Lin, H.H., Lai, J.S.Y., Chin, A.L., Chen, Y.C., Chiang, A.S., 2007. A map of olfactoryrepresentation in the Drosophila mushroom body. Cell 128, 1205–1217.

Linn, C., Feder, J.L., Nojima, S., Dambroski, H.R., Berlocher, S.H., Roelofs, W., 2003.Fruit odor discrimination and sympatric host race formation in Rhagoletis. Proc.Natl. Acad. Sci. U.S.A. 100, 11490–11493.

Linster, C., Sachse, S., Galizia, C.G., 2005. Computational modeling suggests thatresponse properties rather than spatial position determine connectivity be-tween olfactory glomeruli. J. Neurophysiol. 93, 3410–3417.

Losel, R., Homberg, U., 1999. Histamine-immunoreactive neurons in the brain of thecockroach Leucophaea maderae. Brain Res. 842, 408–418.

Loudon, C., Koehl, M.A.R., 2000. Sniffing by a silkworm moth: wing fanningenhances air penetration through and pheromone interception by antennae.J. Exp. Biol. 203, 2977–2990.

MacLeod, K., Laurent, G., 1996. Distinct mechanisms for synchronization and temporalpatterning of odor-encoding neural assemblies. Science 274, 976–979.

Malun, D., Waldow, U., Kraus, D., Boeckh, J., 1993. Connections between thedeutocerebrum and the protocerebrum, and neuroanatomy of several classesof deutocerebral projection neurons in the brain of male Periplaneta americana.J. Comp. Neurol. 329, 143–162.

Manoli, D.S., Foss, M., Villella, A., Taylor, B.J., Hall, J.C., Baker, B.S., 2005. Male-specificfruitless specifies the neural substrates of Drosophila courtship behaviour.Nature 436, 395–400.

Margulies, C., Tully, T., Dubnau, J., 2005. Deconstructing memory in Drosophila. Curr.Biol. 15, R700–R713.

Marin, E.C., Jefferis, G.S.X.E., Komiyama, T., Zhu, H.T., Luo, L.Q., 2002. Representationof the glomerular olfactory map in the Drosophila brain. Cell 109, 243–255.

Martin, J.P., Hildebrand, J.G., 2010. Innate recognition of pheromone and food odorsin moths: a common mechanism in the antennal lobe? Front. Behav. Neurosci.4, 159.

Matsumoto, S.G., Hildebrand, J.G., 1981. Olfactory mechanisms in the moth Man-duca sexta – response characteristics and morphology of central neurons in theantennal lobes. Proc. R. Soc. Lond. B 213, 249–277.

Mazor, O., Laurent, G., 2005. Transient dynamics versus fixed points in odor repre-sentations by locust antennal lobe projection neurons. Neuron 48, 661–673.

Melo, A.C.A., Rutzler, M., Pitts, R.J., Zwiebel, L.J., 2004. Identification of a chemo-sensory receptor from the yellow fever mosquito, Aedes aegypti, that is highlyconserved and expressed in olfactory and gustatory organs. Chem. Senses 29,403–410.

Mobbs, P.G., 1982. The brain of the honeybee Apis mellifera 1. The connections andspatial-organization of the mushroom bodies. Philos. Trans. R. Soc. B 298, 309–354.

Muller, D., Abel, R., Brandt, R., Zockler, M., Menzel, R., 2002. Differential parallelprocessing of olfactory information in the honeybee, Apis mellifera. J. Comp.Physiol. A 188, 359–370.

Mumm, R., Hilker, M., 2005. The significance of background odour for an eggparasitoid to detect plants with host eggs. Chem. Senses 30, 337–343.

Murlis, J., Jones, C.D., 1981. Fine-scale structure of odor plumes in relation to insectorientation to distant pheromone and other attractant sources. Physiol. Ento-mol. 6, 71–86.

Murlis, J., Elkinton, J.S., Carde, R.T., 1992. Odor plumes and how insects use them.Annu. Rev. Entomol. 37, 505–532.

Murthy, M., Fiete, I., Laurent, G., 2008. Testing odor response stereotypy in theDrosophila mushroom body. Neuron 59, 1009–1023.

Mwilaria, E.K., Ghatak, C., Daly, K.C., 2008. Disruption of GABAA in the insectantennal lobe generally increases odor detection and discrimination thresholds.Chem. Senses 33, 267–281.

Nakagawa, T., Sakurai, T., Nishioka, T., Touhara, K., 2005. Insect sex-pheromonesignals mediated by specific combinations of olfactory receptors. Science 307,1638–1642.

Nakagawa, T., Vosshall, L.B., 2009. Controversy and consensus: noncanonical sig-naling mechanisms in the insect olfactory system. Curr. Opin. Neurobiol. 19,284–292.

Nassel, D.R., 1999. Histamine in the brain of insects: a review. Microsc. Res. Tech. 44,121–136.

Nassel, D.R., Homberg, U., 2006. Neuropeptides in interneurons of the insect brain.Cell Tissue Res. 326, 1–24.

Ng, M., Roorda, R.D., Lima, S.Q., Zemelman, B.V., Morcillo, P., Miesenbock, G., 2002.Transmission of olfactory information between three populations of neurons inthe antennal lobe of the fly. Neuron 36, 463–474.

Nishino, H., Mizunami, M., 1998. Giant input neurons of the mushroom body:intracellular recording and staining in the cockroach. Neurosci. Lett. 246, 57–60.

Nishino, H., Yamashita, S., Yamazaki, Y., Nishikawa, M., Yokohari, F., Mizunami, M.,2003. Projection neurons originating from thermo- and hygrosensory glomeruliin the antennal lobe of the cockroach. J. Comp. Neurol. 455, 40–55.

Nishiyama, K., Okada, J., Toh, Y., 2007. Antennal and locomotor responses toattractive and aversive odors in the searching cockroach. J. Comp. Physiol. A193, 963–971.

Nozawa, M., Nei, M., 2007. Evolutionary dynamics of olfactory receptor genes inDrosophila species. Proc. Natl. Acad. Sci. U.S.A. 104, 7122–7127.

Ochieng, S.A., Hansson, B.S., 1999. Responses of olfactory receptor neurones tobehaviourally important odours in gregarious and solitarious desert locust,Schistocerca gregaria. Physiol. Entomol. 24, 28–36.

Olberg, R.M., 1983. Pheromone-triggered flip-flopping interneurons in the ventralnerve cord of the silkworm moth, Bombyx mori. J. Comp. Physiol. 152, 297–307.

Olsen, S.R., Bhandawat, V., Wilson, R.I., 2007. Excitatory interactions betweenolfactory processing channels in the Drosophila antennal lobe. Neuron 54,89–103.

Olsen, S.R., Wilson, R.I., 2008. Lateral presynaptic inhibition mediates gain control inan olfactory circuit. Nature 452, 956–960.

Olsen, S.R., Bhandawat, V., Wilson, R.I., 2010. Divisive normalization in olfactorypopulation codes. Neuron 66, 287–299.

Olshausen, B.A., Field, D.J., 2000. Vision and the coding of natural images. Am. Sci.88, 238–245.

Olsson, S., Linn, C.E., Roelofs, W.L., 2006a. The chemosensory basis for behavioraldivergence involved in sympatric host shifts. I. Characterizing olfactory recep-tor neuron classes responding to key host volatiles. J. Comp. Physiol. A 192, 279–288.

Olsson, S.B., Linn, C.E., Roelofs, W.L., 2006b. The chemosensory basis for behavioraldivergence involved in sympatric host shifts II: olfactory receptor neuronsensitivity and temporal firing pattern to individual key host volatiles. J. Comp.Physiol. A 192, 289–300.

Pelz, D., Roeske, T., Syed, Z., de Bruyne, M., Galizia, C.G., 2006. The molecularreceptive range of an olfactory receptor in vivo (Drosophila melanogasterOR22a). J. Neurobiol. 66, 1544–1563.

Perez-Orive, J., Mazor, O., Turner, G.C., Cassenaer, S., Wilson, R.I., Laurent, G., 2002.Oscillations and sparsening of odor representations in the mushroom body.Science 297, 359–365.

Perez-Orive, J., Bazhenov, M., Laurent, G., 2004. Intrinsic and circuit properties favorcoincidence detection for decoding oscillatory input. J. Neurosci. 24, 6037–6047.

Pitts, R.J., Fox, A.N., Zwiebel, L.J., 2004. A highly conserved candidate chemoreceptorexpressed in both olfactory and gustatory tissues in the malaria vector Anophe-les gambiae. Proc. Natl. Acad. Sci. U.S.A. 101, 5058–5063.

Pompilio, L., Kacelnik, A., Behmer, S.T., 2006. State-dependent learned valuationdrives choice in an invertebrate. Science 311, 1613–1615.

Pouget, A., Sejnowski, T.J., 1997. Spatial transformations in the parietal cortex usingbasis functions. J. Cogn. Neurosci. 9, 222–237.

Pouget, A., Deneve, S., Ducom, J.C., Latham, P.E., 1999. Narrow versus wide tuningcurves: what’s best for a population code? Neural Comput. 11, 85–90.

Raguso, R.A., Willis, M.A., 2002. Synergy between visual and olfactory cues in nectarfeeding by naive hawkmoths, Manduca sexta. Anim. Behav. 64, 685–695.

Raguso, R.A., 2008. Wake up and smell the roses: the ecology and evolution of floralscent. Annu. Rev. Ecol. Evol. Syst. 39, 549–569.

Raubenheimer, D., Tucker, D., 1997. Associative learning by locusts: pairing of visualcues with consumption of protein and carbohydrate. Anim. Behav. 54, 1449–1459.

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447446

Reisenman, C.E., Christensen, T.A., Francke, W., Hildebrand, J.G., 2004. Enantios-electivity of projection neurons innervating identified olfactory glomeruli. J.Neurosci. 24, 2602–2611.

Reisenman, C.E., Christensen, T.A., Hildebrand, J.G., 2005. Chemosensory selectivityof output neurons innervating an identified, sexually isomorphic olfactoryglomerulus. J. Neurosci. 25, 8017–8026.

Reisenman, C.E., Heinbockel, T., Hildebrand, J.G., 2008. Inhibitory interactionsamong olfactory glomeruli do not necessarily reflect spatial proximity. J.Neurophysiol. 100, 554–564.

Reisenman, C.E., Riffell, J.A., Hildebrand, J.G., 2009. Neuroethology of ovipositionbehavior in the moth Manduca sexta. Ann. N. Y. Acad. Sci. 1170, 462–467.

Reynolds, A.M., 2005. Scale-free movement patterns arising from olfactory-drivenforaging. Phys. Rev. E 72, 041928.

Reynolds, A.M., Frye, M.A., 2007. Free-flight odor tracking in Drosophila is consistentwith an optimal intermittent scale-free search. PLoS ONE 2, e354.

Reynolds, A.M., Reynolds, D.R., Smith, A.D., Svensson, G.P., Lofstedt, C., 2007.Appetitive flight patterns of male Agrotis segetum moths over landscape scales.J. Theor. Biol. 245, 141–149.

Riffell, J.A., Abrell, L., Hildebrand, J.G., 2008a. Physical processes and real-timechemical measurement of the insect olfactory environment. J. Chem. Ecol.34, 837–853.

Riffell, J.A., Alarcon, R., Abrell, L., Davidowitz, G., Bronstein, J.L., Hildebrand, J.G.,2008b. Behavioral consequences of innate preferences and olfactory learning inhawkmoth-flower interactions. Proc. Natl. Acad. Sci. U.S.A. 105, 3404–3409.

Riffell, J.A., Lei, H., Christensen, T.A., Hildebrand, J.G., 2009. Characterization andcoding of behaviorally significant odor mixtures. Curr. Biol. 19, 335–340.

Rø, H., Muller, D., Mustaparta, H., 2007. Anatomical organization of antennal lobeprojection neurons in the moth Heliothis virescens. J. Comp. Neurol. 500, 658–675.

Robertson, H.M., Warr, C.G., Carlson, J.R., 2003. Molecular evolution of the insectchemoreceptor gene superfamily in Drosophila melanogaster. Proc. Natl. Acad.Sci. U.S.A. 100, 14537–14542.

Robertson, H.M., Wanner, K.W., 2006. The chemoreceptor superfamily in the honeybee, Apis mellifera: expansion of the odorant, but not gustatory, receptor family.Genome Res. 16, 1395–1403.

Robertson, H.M., Gadau, J., Wanner, K.W., 2010. The insect chemoreceptor super-family of the parasitoid jewel wasp Nasonia vitripennis. Insect Mol. Biol. 19,121–136.

Root, C.M., Semmelhack, J.L., Wong, A.M., Flores, J., Wang, J.W., 2007. Propagation ofolfactory information in Drosophila. Proc. Natl. Acad. Sci. U.S.A. 104, 11826–11831.

Root, C.M., Masuyama, K., Green, D.S., Enell, L.E., Nassel, D.R., Lee, C.H., Wang, J.W.,2008. A presynaptic gain control mechanism fine-tunes olfactory behavior.Neuron 59, 311–321.

Root, C.M., Ko, K.I., Jafari, A., Wang, J.W., 2011. Presynaptic facilitation by neuropep-tides signaling mediates odor-driven food search. Cell 145, 133–144.

Rospars, J.P., Hildebrand, J.G., 2000. Sexually dimorphic and isomorphic glomeruli inthe antennal lobes of the sphinx moth Manduca sexta. Chem. Senses 25, 119–129.

Rossler, W., Randolph, P.W., Tolbert, L.P., Hildebrand, J.G., 1999. Axons of olfactoryreceptor cells of transsexually grafted antennae induce development of sexuallydimorphic glomeruli in Manduca sexta. J. Neurobiol. 38, 521–541.

Rossler, W., Zube, C., 2011. Dual olfactory pathway in Hymenoptera: evolutionaryinsights from comparative studies. Arthropod Struct. Dev. 40, 349–357.

Ruchty, M., Helmchen, F., Wehner, R., Kleineidam, C.J., 2010. Representation ofthermal information in the antennal lobe of leaf-cutting ants. Front. Behav.Neurosci. 4, 174.

Sachse, S., Rappert, A., Galizia, C.G., 1999. The spatial representation of chemicalstructures in the antennal lobe of honeybees: steps towards the olfactory code.Eur. J. Neurosci. 11, 3970–3982.

Sachse, S., Galizia, C.G., 2002. Role of inhibition for temporal and spatial odorrepresentation in olfactory output neurons: a calcium imaging study. J. Neu-rophysiol. 87, 1106–1117.

Sachse, S., Galizia, C.G., 2003. The coding of odour-intensity in the honeybeeantennal lobe: local computation optimizes odour representation. Eur. J. Neu-rosci. 18, 2119–2132.

Sachse, S., Peele, P., Silbering, A., Guhmann, M., Galizia, C.G., 2006. Role of histamine asa putative inhibitory transmitter in the honeybee antennal lobe. Front. Zool. 3, 22.

Salecker, I., Distler, P., 1990. Serotonin-immunoreactive neurons in the antennallobes of the american cockroach Periplaneta americana – light-microscopic andelectron-microscopic observations. Histochem. Cell Biol. 94, 463–473.

Salinas, E., 2006. How behavioral constraints may determine optimal sensoryrepresentations. PLoS Biol. 4, 2383–2392.

Sane, S.P., 2006. Induced airflow in flying insects I. A theoretical model of theinduced flow. J. Exp. Biol. 209, 32–42.

Sane, S.P., Jacobson, N.P., 2006. Induced airflow in flying insects II. Measurement ofinduced flow. J. Exp. Biol. 209, 43–56.

Sanger, T.D., 2003. Neural population codes. Curr. Opin. Neurobiol. 13, 238–249.Sato, K., Pellegrino, M., Nakagawa, T., Nakagawa, T., Vosshall, L.B., Touhara, K., 2008.

Insect olfactory receptors are heteromeric ligand-gated ion channels. Nature452, 1002–1009.

Schachtner, J., Schmidt, M., Homberg, U., 2005. Organization and evolutionarytrends of primary olfactory brain centers in Tetraconata (Crustacea plus Hex-apoda). Arthropod Struct. Dev. 34, 257–299.

Schlief, M.L., Wilson, R.I., 2007. Olfactory processing and behavior downstreamfrom highly selective receptor neurons. Nat. Neurosci. 10, 623–630.

Seki, Y., Aonuma, H., Kanzaki, R., 2005. Pheromone processing center in the pro-tocerebrum of Bombyx mori revealed by nitric oxide-induced anti-CGMP immu-nocytochemistry. J. Comp. Neurol. 481, 340–351.

Seki, Y., Kanzaki, R., 2008. Comprehensive morphological identification and GABAimmunocytochemistry of antennal lobe local interneurons in Bombyx mori. J.Comp. Neurol. 506, 93–107.

Seki, Y., Rybak, J., Wicher, D., Sachse, S., Hansson, B.S., 2010. Physiological andmorphological characterization of local interneurons in the Drosophila antennallobe. J. Neurophysiol. 104, 1007–1019.

Semmelhack, J.L., Wang, J.W., 2009. Select Drosophila glomeruli mediate innateolfactory attraction and aversion. Nature 459, 218–223.

Shang, Y.H., Claridge-Chang, A., Sjulson, L., Pypaert, M., Miesenbock, G., 2007.Excitatory local circuits and their implications for olfactory processing in thefly antennal lobe. Cell 128, 601–612.

Shannon, C.E., Weaver, W., 1964. The Mathematical Theory of Communication.University of Illinois Press, Urbana.

Siju, K.P., Hansson, B.S., Ignell, R., 2008. Immunocytochemical localization ofserotonin in the central and peripheral chemosensory system of mosquitoes.Arthropod Struct. Dev. 37, 248–259.

Silbering, A.F., Galizia, C.G., 2007. Processing of odor mixtures in the Drosophilaantennal lobe reveals both global inhibition and glomerulus-specific interac-tions. J. Neurosci. 27, 11966–11977.

Silbering, A.F., Okada, R., Ito, K., Galizia, C.G., 2008. Olfactory information processingin the Drosophila antennal lobe: anything goes? J. Neurosci. 28, 13075–13087.

Simoncelli, E.P., Olshausen, B.A., 2001. Natural image statistics and neural repre-sentation. Annu. Rev. Neurosci. 24, 1193–1216.

Simpson, S.J., Raubenheimer, D., 2000. The hungry locust. Adv. Stud. Behav. 29, 1–44.

Sinakevitch, I., Niwa, M., Strausfeld, N.J., 2005. Octopamine-like immunoreactivityin the honey bee and cockroach: comparable organization in the brain andsubesophageal ganglion. J. Comp. Neurol. 488, 233–254.

Singer, W., 1999. Neuronal synchrony: a versatile code for the definition ofrelations? Neuron 24, 49–65.

Siwicki, K.K., Riccio, P., Ladewski, L., Marcillac, F., Dartevelle, L., Cross, S.A., Ferveur,J.F., 2005. The role of cuticular pheromones in courtship conditioning ofDrosophila males. Learn. Mem. 12, 636–645.

Sjoholm, M., Sinakevitch, I., Ignell, R., Strausfeld, N.J., Hansson, B.S., 2005. Organi-zation of kenyon cells in subdivisions of the mushroom bodies of a Lepidopteraninsect. J. Comp. Neurol. 491, 290–304.

Sjoholm, M., Sinakevitch, I., Strausfeld, N.J., Ignell, R., Hansson, B.S., 2006. Functionaldivision of intrinsic neurons in the mushroom bodies of male Spodopteralittoralis revealed by antibodies against aspartate, taurine, FMRF-amide, Mas-allatotropin and DC0. Arthropod Struct. Dev. 35, 153–168.

Slessor, K.N., Kaminski, L.A., King, G.G.S., Winston, M.L., 1990. Semiochemicals of thehoneybee queen mandibular glands. J. Chem. Ecol. 16, 851–860.

Slessor, K.N., Winston, M.L., Le Conte, Y., 2005. Pheromone communication in thehoneybee (Apis mellifera l). J. Chem. Ecol. 31, 2731–2745.

Smid, H.M., Wang, G.H., Bukovinszky, T., Steidle, J.L.M., Bleeker, M.A.K., van Loon,J.J.A., Vet, L.E.M., 2007. Species-specific acquisition and consolidation of long-term memory in parasitic wasps. Proc. R. Soc. Lond. B 274, 1539–1546.

Steiger, U., 1967. Uber den Feinbau des Neuropils im Corpus pedunculatum derWaldameise. Z. Zellforsch. Mikrosk. Anat. 81, 511–536.

Stensmyr, M.C., Dekker, T., Hansson, B.S., 2003. Evolution of the olfactory code in theDrosophila melanogaster subgroup. Proc. R. Soc. Lond. B 270, 2333–2340.

Stensmyr, M.C., Hansson, B.S., 2007. Flies’ lives on a crab. Curr. Biol. 17, R743–R746.Stocker, R.F., Lienhard, M.C., Borst, A., Fischbach, K.F., 1990. Neuronal architecture of

the antennal lobe in Drosophila melanogaster. Cell Tissue Res. 262, 9–34.Stockinger, P., Kvitsiani, D., Rotkopf, S., Tirian, L., Dickson, B.J., 2005. Neural circuitry

that governs Drosophila male courtship behavior. Cell 121, 795–807.Stopfer, M., Bhagavan, S., Smith, B.H., Laurent, G., 1997. Impaired odour discrimi-

nation on desynchronization of odour-encoding neural assemblies. Nature 390,70–74.

Stopfer, M., Jayaraman, V., Laurent, G., 2003. Intensity versus identity coding in anolfactory system. Neuron 39, 991–1004.

Strausfeld, N.J., Hansen, L., Li, Y.S., Gomez, R.S., Ito, K., 1998. Evolution, discovery,and interpretations of arthropod mushroom bodies. Learn. Mem. 5, 11–37.

Strausfeld, N.J., Li, Y.S., 1999. Organization of olfactory and multimodal afferentneurons supplying the calyx and pedunculus of the cockroach mushroombodies. J. Comp. Neurol. 409, 603–625.

Strausfeld, N.J., 2002. Organization of the honey bee mushroom body: representa-tion of the calyx within the vertical and gamma lobes. J. Comp. Neurol. 450, 4–33.

Strausfeld, N.J., 2009. Brain organization and the origin of insects: an assessment.Proc. R. Soc. B 276, 1929–1937.

Strausfeld, N.J., Sinakevitch, I., Brown, S.M., Farris, S.M., 2009. Ground plan of theinsect mushroom body: functional and evolutionary implications. J. Comp.Neurol. 513, 265–291.

Strauss, K., Scharpenberg, H., Crewe, R.M., Glahn, F., Foth, H., Moritz, R.F.A., 2008.The role of the queen mandibular gland pheromone in honeybees (Apis melli-fera): honest signal or suppressive agent? Behav. Ecol. Sociobiol. 62, 1523–1531.

Strube-Bloss, M.F., Nawrot, M.P., Menzel, R., 2011. Mushroom body output neuronsencode odor-reward associations. J. Neurosci. 31, 3129–3140.

Suh, G.S.B., Wong, A.M., Hergarden, A.C., Wang, J.W., Simon, A.F., Benzer, S., Axel, R.,Anderson, D.J., 2004. A single population of olfactory sensory neurons mediatesan innate avoidance behaviour in Drosophila. Nature 431, 854–859.

J.P. Martin et al. / Progress in Neurobiology 95 (2011) 427–447 447

Suh, G.S.B., de Leon, S.B.T., Tanimoto, H., Fiala, A., Benzer, S., Anderson, D.J., 2007.Light activation of an innate olfactory avoidance response in Drosophila. Curr.Biol. 17, 905–908.

Sun, X.J., Tolbert, L.P., Hildebrand, J.G., 1997. Synaptic organization of the uniglomer-ular projection neurons of the antennal lobe of the moth Manduca sexta: a laserscanning confocal and electron microscopic study. J. Comp. Neurol. 379, 2–20.

Szyszka, P., Ditzen, M., Galkin, A., Galizia, C.G., Menzel, R., 2005. Sparsening andtemporal sharpening of olfactory representations in the honeybee mushroombodies. J. Neurophysiol. 94, 3303–3313.

Takken, W., van Loon, J.J.A., Adam, W., 2001. Inhibition of host-seeking response andolfactory responsiveness in Anopheles gambiae following blood feeding. J. InsectPhysiol. 47, 303–310.

Tanaka, N.K., Awasaki, T., Shimada, T., Ito, K., 2004. Integration of chemosensorypathways in the Drosophila second-order olfactory centers. Curr. Biol. 14, 449–457.

Tanaka, N.K., Tanimoto, H., Ito, K., 2008. Neuronal assemblies of the Drosophilamushroom body. J. Comp. Neurol. 508, 711–755.

Tanaka, N.K., Ito, K., Stopfer, M., 2009. Odor-evoked neural oscillations in Drosophilaare mediated by widely branching interneurons. J. Neurosci. 29, 8595–8603.

Technau, G., Heisenberg, M., 1982. Neural reorganization during metamorphosis ofthe corpora pedunculata in Drosophila melanogaster. Nature 295, 405–407.

Thom, C., Gilley, D.C., Hooper, J., Esch, H.E., 2007. The scent of the waggle dance. PLoSBiol. 5, 1862–1867.

Thum, A.S., Jenett, A., Ito, K., Heisenberg, M., Tanimoto, H., 2007. Multiple memorytraces for olfactory reward learning in Drosophila. J. Neurosci. 27, 11132–11138.

Touhara, K., Vosshall, L.B., 2009. Sensing odorants and pheromones with chemo-sensory receptors. Annu. Rev. Physiol. 71, 307–332.

Tsacas, L., Baechli, G., 1981. Drosophila sechellia, n, Sp., huitieme espece du sous-groupe melanogaster des iles Seychelles (Diptera, Drosophilidae). Rev. Fr.Entomol. 3, 146–150.

Turlings, T.C.J., Davison, A.C., Tamo, C., 2004. A six-arm olfactometer permittingsimultaneous observation of insect attraction and odour trapping. Physiol.Entomol. 29, 45–55.

Turner, G.C., Bazhenov, M., Laurent, G., 2008. Olfactory representations by Drosoph-ila mushroom body neurons. J. Neurophysiol. 99, 734–746.

Urban, N.N., 2002. Lateral inhibition in the olfactory bulb and in olfaction. Physiol.Behav. 77, 607–612.

van Naters, W.V., Carlson, J.R., 2006. Insects as chemosensors of humans and crops.Nature 444, 302–307.

Vergoz, V., Schreurs, H.A., Mercer, A.R., 2007. Queen pheromone blocks aversivelearning in young worker bees. Science 317, 384–386.

Vetter, R.S., Baker, T.C., 1983. Behavioral responses of male Heliothis virescens in asustained-flight tunnel to combinations of seven compounds identified fromfemale sex-pheromone glands. J. Chem. Ecol. 9, 747–759.

Vetter, R.S., Baker, T.C., 1984. Behavioral-responses of male Heliothis zea (Lepidop-tera: Noctuidae) moths in sustained-flight tunnel to combinations of fourcompounds identified from female sex-pheromone gland. J. Chem. Ecol. 10,193–202.

Vickers, N.J., Baker, T.C., 1997. Chemical communication in Heliothine moths 7.Correlation between diminished responses to point source plumes and singlefilaments similarly tainted with a behavioral antagonist. J. Comp. Physiol. A 180,523–536.

Vickers, N.J., Christensen, T.A., Hildebrand, J.G., 1998. Combinatorial odor discrimina-tion in the brain: attractive and antagonist odor blends are represented in distinctcombinations of uniquely identifable glomeruli. J. Comp. Neurol. 400, 35–56.

Vickers, N.J., Christensen, T.A., Baker, T.C., Hildebrand, J.G., 2001. Odour-plumedynamics influence the brain’s olfactory code. Nature 410, 466–470.

Vickers, N.J., Christensen, T.A., 2003. Functional divergence of spatially conservedolfactory glomeruli in two related moth species. Chem. Senses 28, 325–338.

Vickers, N.J., Poole, K., Linn, C.E., 2005. Plasticity in central olfactory processing andpheromone blend discrimination following interspecies antennal imaginal disctransplantation. J. Comp. Neurol. 491, 141–156.

Vickers, N.J., 2006a. Inheritance of olfactory preferences I. Pheromone-mediatedbehavioral responses of Heliothis subflexa � Heliothis virescens hybrid malemoths. Brain Behav. Evol. 68, 63–74.

Vickers, N.J., 2006b. Inheritance of olfactory preferences III. Processing of phero-monal signals in the antennal lobe of Heliothis subflexa � Heliothis virescenshybrid male moths. Brain Behav. Evol. 68, 90–108.

Vickers, N.J., 2006c. Winging it: moth flight behavior and responses of olfactoryneurons are shaped by pheromone plume dynamics. Chem. Senses 31, 155–166.

Vieira, F.G., Sanchez-Gracia, A., Rozas, J., 2007. Comparative genomic analysis of theodorant-binding protein family in 12 Drosophila genomes: purifying selectionand birth-and-death evolution. Genome Biol. 8, R235.

Vosshall, L.B., Wong, A.M., Axel, R., 2000. An olfactory sensory map in the fly brain.Cell 102, 147–159.

Vosshall, L.B., Hansson, B.S., 2011. A unified nomenclature system for the insectolfactory coreceptor. Chem. Senses 36, 497–498.

Wang, J.W., Wong, A.M., Flores, J., Vosshall, L.B., Axel, R., 2003. Two-photon calciumimaging reveals an odor-evoked map of activity in the fly brain. Cell 112, 271–282.

Wegerhoff, R., 1999. GABA and serotonin immunoreactivity during postembryonicbrain development in the beetle Tenebrio molitor. Microsc. Res. Tech. 45, 154–164.

Wehr, M., Laurent, G., 1996. Odour encoding by temporal sequences of firing inoscillating neural assemblies. Nature 384, 162–166.

Wehr, M., Laurent, G., 1999. Relationship between afferent and central temporalpatterns in the locust olfactory system. J. Neurosci. 19, 381–390.

Wertheim, B., Dicke, M., Vet, L.E.M., 2002. Behavioural plasticity in support of abenefit for aggregation pheromone use in Drosophila melanogaster. Entomol.Exp. Appl. 103, 61–71.

Wicher, D., Schafer, R., Bauernfeind, R., Stensmyr, M.C., Heller, R., Heinemann, S.H.,Hansson, B.S., 2008. Drosophila odorant receptors are both ligand-gated andcyclic-nucleotide-activated cation channels. Nature 452, 1007–1010.

Williams, B.A., Kay, R.F., Kirk, E.C., 2010. New perspectives on anthropoid origins.Proc. Natl. Acad. Sci. U.S.A. 107, 4797–4804.

Willis, M.A., Baker, T.C., 1984. Effects of intermittent and continuous pheromonestimulation on the flight behavior of the oriental fruit moth, Grapholita molesta.Physiol. Entomol. 9, 341–358.

Wilson, R.I., Turner, G.C., Laurent, G., 2004. Transformation of olfactory representa-tions in the Drosophila antennal lobe. Science 303, 366–370.

Wilson, R.I., Laurent, G., 2005. Role of gabaergic inhibition in shaping odor-evokedspatiotemporal patterns in the Drosophila antennal lobe. J. Neurosci. 25, 9069–9079.

Wong, A.M., Wang, J.W., Axel, R., 2002. Spatial representation of the glomerular mapin the Drosophila protocerebrum. Cell 109, 229–241.

Wright, G.A., Lutmerding, A., Dudareva, N., Smith, B.H., 2005. Intensity and the ratiosof compounds in the scent of snapdragon flowers affect scent discrimination byhoneybees (Apis mellifera). J. Comp. Physiol. A 191, 105–114.

Wright, G.A., Carlton, M., Smith, B.H., 2009. A honeybee’s ability to learn, recognize,and discriminate odors depends upon odor sampling time and concentration.Behav. Neurosci. 123, 36–43.

Yaksi, E., Wilson, R.I., 2010. Electrical coupling between olfactory glomeruli. Neuron67, 1034–1047.

Yamagata, N., Mizunami, M., 2010. Spatial representation of alarm pheromoneinformation in a secondary olfactory centre in the ant brain. Proc. R. Soc. B 277,2465–2474.

Yasuyama, K., Meinertzhagen, I.A., Schurmann, F.W., 2003. Synaptic connections ofcholinergic antennal lobe relay neurons innervating the lateral horn neuropilein the brain of Drosophila melanogaster. J. Comp. Neurol. 466, 299–315.

Yu, D.H., Ponomarev, A., Davis, R.L., 2004. Altered representation of the spatial codefor odors after olfactory classical conditioning: memory trace formation bysynaptic recruitment. Neuron 42, 437–449.

Yuko, I., Walter, S.L., 2008. Chiral discrimination of the japanese beetle sex phero-mone and a behavioral antagonist by a pheromone-degrading enzyme. Proc.Natl. Acad. Sci. U.S.A. 105, 9076–9080.

Zeiner, R., Tichy, H., 2000. Integration of temperature and olfactory information incockroach antennal lobe glomeruli. J. Comp. Physiol. A 186, 717–727.

Zhang, K.C., Sejnowski, T.J., 1999. Neuronal tuning: to sharpen or broaden? NeuralComput. 11, 75–84.

Zhu, S.J., Chiang, A.S., Lee, T., 2003. Development of the Drosophila mushroombodies: elaboration, remodeling and spatial organization of dendrites in thecalyx. Development 130, 2603–2610.

Zube, C., Kleineidam, C.J., Kirschner, S., Neef, J., Rossler, W., 2008. Organization of theolfactory pathway and odor processing in the antennal lobe of the ant Campo-notus floridanus. J. Comp. Neurol. 506, 425–441.

Zube, C., Rossler, W., 2008. Caste- and sex-specific adaptations within the olfactorypathway in the brain of the ant Camponotus floridanus. Arthropod Struct. Dev.37, 469–479.

Zwiebel, L.J., Takken, W., 2004. Olfactory regulation of mosquito-host interactions.Insect Biochem. Mol. Biol. 34, 645–652.