Neuron

Review

Evolution of Insect Olfaction

Bill S. Hansson1,* and Marcus C. Stensmyr1,*
1Department of Evolutionary Neuroethology, Max Planck Institute for Chemical Ecology, Hans Knoell Strasse 8, 07749 Jena, Germany

*Correspondence: [email protected] (B.S.H.), [email protected] (M.C.S.)

DOI 10.1016/j.neuron.2011.11.003

Neuroethology utilizes a wide range of multidisciplinary approaches to decipher neural correlates of natural
behaviors associated with an animal’s ecological niche. By placing emphasis on comparative analyses of
adaptive and evolutionary trends across species, a neuroethological perspective is uniquely suited to uncov-
ering general organizational and biological principles that shape the function and anatomy of the nervous
system. In this review, we focus on the application of neuroethological principles in the study of insect olfac-
tion and discuss how ecological environment and other selective pressures influence the development of
insect olfactory neurobiology, not only informing our understanding of olfactory evolution but also providing
broader insights into sensory processing.

Introduction tance across closely related taxa, can also be an efficient way
In order to locate and evaluate food, shelter, mates, and to identify the functional significance of specific features, be
breeding substrates as well as to avoid predators and other they genes or neurons, correlated with the trait under study.
dangers—or simply just to move around—animals rely on Knowing the ecology of the study animal can provide clues as
a wide range of sensory systems. These senses supply the to the natural context in which a given set of neurons comes
animal’s nervous system with information subsequently used into importance, and to relevant external stimuli, in turn providing
to generate a simplified internal representation of the complex access to specialized circuits underlying specific behaviors. The
external world, which in turn allows the animal to decide upon ecology can moreover assist in creating improved behavioral
and execute the appropriate behavioral response given the situ- assays, better reflecting the behavioral complexity of animals
ation. Identification and functional dissection of neural circuits operating in a natural setting, yielding improved behavioral
underlying specific behaviors is currently a hot topic in neurosci- readout possibilities.
ence, an interest in part fuelled by recent methodological Neuroethological approaches have provided significant
advances allowing for in vivo manipulation of activity from insights into mechanisms underlying a wide variety of neural
precisely defined neuronal circuits, or even from single neurons. processes. A classic example is the auditory map of the barn
Technical advances aside, prerequisites for these types of owl (Tyto alba) (Knudsen and Konishi, 1978). The nocturnal
endeavors are (1) that the behavioral repertoire of the animal barn owls are masters at localizing prey through auditory infor-
under scrutiny is understood and (2) an understanding of which mation and are capable of hunting in complete darkness (Payne,
external stimuli or situations cause the behavior of interest to 1971). By recording from the midbrain, while presenting sounds
be elicited. In principle, one needs accordingly to have at least akin to those an owl would encounter in its natural habitat, from
a rudimentary grasp of the ecology of the species under study. various locations in space, Knudsen and Konishi managed to
For example, if one wishes to study innate fear in the house localize an area in the inferior colliculus, housing a set of neurons,
mouse (Mus musculus), it is obviously important to know what so called space-specific neurons, which would only fire once
a mouse would be scared of, i.e., to know which potential auditory stimuli were delivered from a specific spatial position.
dangers and predators a mouse would face in the arid regions The cells in this region were found to be organized in a precise
of the northern Indian subcontinent, the evolutionary cradle of topographic array, with cell clusters arranged to represent the
the species (Boursot et al., 1996). vertical and horizontal location of the sound. Although the barn
The study of neuronal circuits in a behavioral context, specif- owl is a highly specialized animal, showing some neuronal
ically in a comparative, ecological, and/or evolutionary frame- features with respect to auditory processing not present in other
work, is usually termed neuroethology. Typically, neuroetholog- brain regions or species, the owl’s auditory system nevertheless
ical studies are concerned with natural behaviors and are often relies on neural strategies for, e.g., coincidence detection and
performed in less established ‘‘model’’ systems. Although a enhancement of reliability, which are probably essential for the
species like the duck-billed platypus (Ornithorhynchus anatinus) operation of many other circuits within as well as outside the
might be impractical as a model overall, or offer no direct owl auditory system. The barn owl and its auditory localization
general advantage over established systems, species like this pathway have also provided fundamental insights into neuronal
may offer unique insights with respect to specific questions, in computation and in particular how these computations are
this instance mammalian electroreception (Scheich et al., affected by experience.
1986). Moreover, expanding neuroscientific studies beyond es- The neuroethological approach is, however, not without its
tablished laboratory models is naturally also of importance to drawbacks. The disadvantage of working with natural behaviors
verify the generality of processes and functions. Comparative is that these are indeed natural behaviors, and as such in some
approaches, as in exploring a given trait with differing impor- cases only exhibited by free-ranging animals, i.e., wild animals

698 Neuron 72, December 8, 2011 ª2011 Elsevier Inc.

Neuron

Review

B C E G

D F

Figure 1. Diversity of Insects, Antennae, and Olfactory Sensilla

(A) Insects come in a plethora of shapes and forms. Elaborate antennal structures are found across many different insect taxa and highlight the general
importance of olfaction in this group of animals. Depicted species: (1) longhorn beetle, Diastocera wallichi (Coleoptera; Cerambycidae); (2) tortoise beetle,
Cassidini tribe, undetermined (Coleoptera: Chrysomelidae); (3) European cockchafer, Melolontha melolontha (Coleoptera: Scarabaeidae); (4) Nevada buck moth,
Hemileuca nevadensis (Lepidoptera: Saturniidae); (5) snout beetle, Euphotos bennetti (Curcolionidae), (6) Ground Beetle, undetermined (Coleoptera: Paussidae);
(7) velvet ant, undetermined (Hymenoptera; Mutillidae); (8) click beetle, undetermined (Coleoptera: Elateridae); and (9) leaf-footed bug, undetermined (Hemiptera:
Coreidae). Scale bar represents 1 cm. All photos by M.C.S.
(B) A male swede midge, Contarinia nasturtii (Diptera: Cecidomyiidae) with impressively sized antenna. Insert shows the peculiar sensilla circumfila (arrow), which
are composed of multiple fused and bifurcated sensilla that form elongated loops. The s. circumfila are unique to gall midges and their functional significance
remains unclear (Boddum et al., 2010). Image courtesy of Y. Hillbur, SLU Alnarp.
(C–F) Examples of principal types of olfactory sensilla, here from the antennae of the mosquitoes Anopheles gambiae s.s. and An. Quadriannulatus. (C) blunt
sensilla trichodea (arrow), (D) large s. coeloconia, (E) s. basiconica, and (F) small s. coeloconica (arrow). Adapted from Pitts and Zwiebel (2006), with permission
from the authors.
(G) Longitudinal section through a s. trichodea from D. melanogaster. Color code as per Figure 2A. Adapted from Shanbhag et al. (1999), with permission from
Elsevier.

roaming their habitat. Simply observing animals in nature is often netic framework. Insects are in short ideal for neuroethological
a complex task; to carefully monitor behaviors and subject studies and have consequently also received considerable
these to experimental manipulation is often a herculean task. attention in this respect. In particular, insects have proven
In addition, natural behaviors are typically complex composites a particularly successful model in studying the sense of smell.
of distinct subroutines. Even a fairly simple creature like the Here we aim to review work addressing insect olfaction from
honeybee worker Apis mellifera shows a considerable behavioral a neuroethological perspective, highlighting particularly salient
repertoire, with at least 59 distinct and recognizable behaviors findings that inform our broader understanding of olfactory
on the menu (Chittka and Niven, 2009). Differentiating among evolution and neurobiology specifically and sensory processing
the behaviors and determining which stimuli elicit which behavior more generally. Specifically, we will cover how insects decode
is in many cases challenging. Even if distinct behaviors can be their chemical environment, how the peripheral olfactory system
discerned, monitored, and subjected to manipulation, finding adapts and evolves, and in turn how this reflects the adaptive
the neural correlates might often be hard. Neuroscience tools forces acting on the system over evolutionary time.
readily available in established systems, such as the fly or the
mouse, are in many instances not directly transferable to other Environment and Function Drive Variability in Peripheral
species, at least not without considerable efforts. Insects, Olfactory Morphology
however, in spite of their minute size, display a wide span of The sense of smell is of pivotal importance to most insects
behaviors of which most are stereotype and executed in an (Dethier, 1947). The importance of olfaction is evident from the
obligate manner pending the presentation of the correct stim- elaborate antennal structures, the functional equivalents of the
ulus, even in a laboratory setting. Insects in addition comprise human nose, found in many insects. Apart from antennae,
a remarkably diverse group of organisms. Within a given family, insects also detect odors with their maxillary palps and/or labial
one can often find a wide variety of lifestyles and habitats palps. The antennae (and palps) come in a multitude of shapes
(Grimaldi and Engel, 2005), thus providing excellent entry points (Figure 1A) but nevertheless conform to the same basic princi-
for comparative studies within a narrow and defined phyloge- ples (Schneider, 1964). The distal segment of the antennae is

Neuron 72, December 8, 2011 ª2011 Elsevier Inc. 699

 Neuron

Review

A B

Figure 2. The Anatomy of the Insect and Mammalian Peripheral Olfactory Systems
(A) In insects, the primary olfactory detecting organs are the antennae, here illustrated by a male luna moth Actias luna (Lepidoptera; Saturniidae) (A1) (Photo:
B. Bumgartner). Although not visible on this image, insects, also detect odors with their palps (shown in Figure 1B). The antennae are in several insect orders
highly branched structures (A2) (Photo: M.M. Cordts); densely covered in olfactory sensilla (A3). The sensilla, here illustrated by a longitudinal section through a s.
basiconica from D. melanogaster, houses the olfactory sensory neurons (OSN) whose dendritic part express the chemosensory receptors that binds the odor
molecules. Volatile chemicals are in insects primarily detected by two different receptor types (A4); the odorant receptors (ORs) and the ionotropic receptors (IRs).
The gustatory receptors (GRs), as the name implies, mediate taste; however, members of this gene family also detect CO2. Image A3 is adapted from Shanbhag
et al., 1999, with permission from Elsevier.
(B) Mammals, here illustrated by Harvey the house mouse (B1), detect odors via multiple organs located in the nasal cavity (B2). The main olfactory epithelium and
the septal organ detect general odorants, whereas the vomeronasal organ and the Gruenenberg ganglion primarily detect compounds mediating sexual and
social behaviors. The main olfactory epithelium is a highly convoluted structure (B3), with the outer cell layer composed of closely packed OSNs (B4). Images B3
and B4 comes from a cross between two gene-targeted mice expressing fluorescent reporters from the M71 and the P2 odorant receptor loci respectively.
Neurons expressing the M71 receptor are labeled by RFP (red arrow) and P2 expressing neurons by GFP (green arrow). Mammals rely on a variety of odor
receptors, of which the vomeronasal receptor (V1R and V2R), the trace amine associated receptor (TAAR), and the formyl peptide receptor (FPR) molecules are
involved in social and sexual communication and assessment. Images B1, B3, and B4 courtesy of professor P. Mombaerts, MPI Biophysics, Frankfurt, Germany.

covered, to various extents with olfactory sensilla, which show (OSNs) (Zacharuk, 1980) (Figure 2A). Although fulfilling the
a wide variety of shapes and structures (Schneider and Stein- same role, the organization of the peripheral olfactory system
brecht, 1968) (Figures 1B–1F). Irrespective of form, the olfactory of insects is quite different from that of mammals (Figure 2B).
sensilla all share the same function, namely, to encapsulate and The insect antennae have presumably evolved from structures
protect the sensitive dendrites of the olfactory sensory neurons that predominantly mediated mechanosensory input. In primitive

700 Neuron 72, December 8, 2011 ª2011 Elsevier Inc.

Neuron

Review

terrestrial arthropods, the antennae have great flexibility of families, of which only two will be discussed here. The odor
movement due to the presence of intrinsic musculature, but molecules initially bind to so-called odorant binding proteins
owing to the small number of sensilla, quite a poor capacity for (OBPs, Vogt and Riddiford, 1981). OBPs are secreted in large
chemoreception. The sensillum-rich flagellar antennae found in quantities by support cells surrounding the OSNs and show
most insects are, however, void of intrinsic muscles, and are in specific binding properties (Swarup et al., 2011). Although their
most lineages specialized structures for detecting odor mole- exact function remains to be elucidated (but see Laughlin
cules (Schneider, 1964). Exemptions are naturally found, such et al., 2008, for their role in pheromone communication), these
as in the aquatic water scavenger beetles (Coleoptera: Hydro- proteins are probably involved in transporting the odor ligands
phillidae), whose antennae actually lack an olfactory function to the receptor sites, situated in the dendritic membrane of the
altogether and instead serve as ‘‘snorkels,’’ which are used to OSNs. The OBPs form a large insect-exclusive gene family
refill internal air reservoirs (Schaller, 1926). Whether antennal with conserved structure, but which otherwise shows a high
architecture is shaped by the evolutionary necessity to detect degree of sequence diversity. The OBP family is possibly as
certain odor molecules is uncertain. Most likely, the variability old as the insects themselves, having evolved in response to
in antennal shapes (as seen in Figure 1A) reflects constraints demands imposed by the conquest of land (Vieira and Rozas,
imposed by the physical, rather than the chemical environment 2011, but see Forêt and Maleszka, 2006). So-called odorant
of the insects. For example, the delicate plumose antennae of binding proteins are also found in vertebrates; these, however,
the volant Nevada buck moth in Figure 1A has very likely evolved belong to the lipocalin family and show no structural similarity
to capture volatile molecules with high efficiency in air, but would to the insect OBPs (Bianchet et al., 1996). The OBP family in
be ill suited to fulfilling the same function for a ground- or soil- the vinegar fly comprises 51 members (Hekmat-Scafe et al.,
dwelling insect. 2002), and similar numbers have been found in other insects
As to why insects are equipped with a second nose, i.e., the so far investigated. Although subfamilies can be discerned within
maxillary and/or the labial palps, remains unclear. In several the OBP family, examination of these genes across broader
insect species, including the hawk moth Manduca sexta taxonomic range reveals that the OBPs largely cluster according
(Lepidoptera: Sphingidae) and the African malaria mosquito to phylogeny, with groupings representing independent, lineage
Anopheles gambiae (Diptera: Culicoidae), these organs serve specific radiations of specific OBPs. Clear orthologs, present
a distinct function as they house OSNs detecting CO2, which across different insect orders, are hence essentially lacking
in both species is a crucial sensory cue for locating resources (Vieira and Rozas, 2011). Analyses of the OBP repertoires from
(Thom et al., 2004; Lu et al., 2007). However, in the vinegar fly the 12 complete Drosophila spp. genomes have shown that in
Drosophila melanogaster (Diptera: Drosophilidae), CO2 detec- ecologically specialized species, OBPs evolve more rapidly
tion is accomplished via OSNs on the antennae, and the palp’s than in their generalist relatives (Vieira et al., 2007), suggesting
OSNs show overlapping response spectra with those of the that OBPs play an important role in ecological diversification.
antennae (de Bruyne et al., 1999). In the vinegar fly, the palps In our view, elucidating the precise role of this interesting gene
have instead been suggested to play a role in taste enhancement family should be a prioritized task for the field.
(Shiraiwa, 2008). How general such a function would be across Like the OBPs, the gene family encoding odorant receptors
insects remains to be investigated. Whether sensillum architec- (ORs) in insects is also an insect exclusive radiation. The ORs
ture confers a functional advantage in detecting a specific class form a large and highly divergent gene family (Clyne et al.,
of chemicals also remains unclear. Trichoid sensilla (Figure 1C) in 1999; Vosshall et al., 1999), which shows no homology to the
many insects, including vinegar flies and most (if not all) moths, OR families of vertebrates and nematodes. The insect ORs and
house OSNs tuned to pheromones (van der Goes van Naters the related gustatory receptors (GRs, Clyne et al., 2000; Scott
and Carlson, 2007; Kaissling et al., 1989). However, whether et al., 2001) together form an arthropod-specific chemoreceptor
the trichoid structure itself is advantageous for the detection of superfamily, in which the ORs constitute a single highly
this type of chemicals is uncertain. Likewise, OSNs housed in expanded branch (Robertson et al., 2003). Members of this
coeloconic sensilla (Figure 1G) respond mostly to water-soluble superfamily essentially share no homology to any known gene
amines and acids (see below), but what role, if any, the actual family, and encodes for seven transmembrane-domain recep-
coeloconic architecture play is unknown. tors with an inverted transmembrane topology as compared to
The peripheral olfactory system of insects shows a remarkable the G protein-coupled olfactory receptors of vertebrates (Benton
morphological diversity at all levels. The role of this diversity et al., 2006). In contrast to the ORs of vertebrates, the insect ORs
remains unclear but probably reflects selection pressures for form heteromeric complexes typically composed of a single
high sensitivity, phylogenetic and/or developmental constraints, ligand-binding OR (Störtkuhl and Kettler, 2001; Dobritsa et al.,
and imposed by the physical environment, rather than adapta- 2003, but see Goldman et al., 2005) and the OR coreceptor
tions to detect specific volatile chemicals. Orco (Vosshall et al., 2000; Larsson et al., 2004). Orco acts as
a chaperone (Larsson et al., 2004) and also takes part in signal
The Molecular Machinery of Insect Olfaction Is Uniquely transduction (Sato et al., 2008; Wicher et al., 2008). The rise of
Composed the OR family is assumed to date back to the early Devonian
The odor molecules pass through pores or slits in the sensillum and the first insects as an adaptation to terrestrial life (Robertson
cuticle and enter the sensillum lymph (Steinbrecht, 1997). From et al., 2003). However, one could also envision that the OR
here on, the typically hydrophobic chemicals that constitute radiation occurred at a later stage (perhaps first with the rise of
odor ligands on land interact with members from multiple gene Neoptera); being driven by the diversification of vascular plants

Neuron 72, December 8, 2011 ª2011 Elsevier Inc. 701

 Neuron

Review

ecological specialization. Firmly categorizing these species in

terms of ecology and demography is however difficult. For
example, although D. erecta is specialized upon fruit from
Pandanus spp. screwpines, this resource is not continuously
available in the habitat. Accordingly, this species must also
utilize other resources. Moreover, D. erecta has a restricted
and patchy distribution and may thus in fact have a small effec-
tive population size (Lachaise et al., 1988). Consequently, exam-
ining OR repertoires of additional drosophilid taxa is undoubt-
edly necessary before any firm conclusions can be drawn.
In short, the molecular basis of insect olfaction shows a
number of unique features and is characterized by two large
gene families, the OBPs and the ORs, which are presumably
exclusive to this group of animals. When these two gene families
first appear in the insect lineage and whether the initial conquest
of land or the diversification of land plants drove their evolution
remains to be determined. All insect genomes to date stem
from derived orders. Deep sequencing of species from basal
insect orders, as well as from allied hexapod orders is thus
needed in order to understand the evolutionary history of these
gene families.

The Insect Olfactory System Selectively Detects

Figure 3. Phylogenetic Relationships within the Insect Odorant Signature Features of the Habitat
Receptor Family Insects have to detect specific volatile information in a very
The insect ORs forms a highly divergent gene family. With the exception of the complicated chemical environment. How is this feat accom-
coreceptor Orco, close orthologs and conserved subfamily structures are
absent across insect orders. The shown tree was computed (using FastTree
plished? In the vinegar fly and the African malaria mosquito,
based on a MAFFT-nsi alignment) from 1069 OR genes from nine species more or less the complete OR repertoires have been deor-
belonging to six different orders. phaned, i.e., their key odorant stimuli have been identified. In
both species, the ORs display a varying degree of specificity,
with certain receptors showing a high degree of selectivity, while
and the increasing abundance of volatile chemicals in the envi- others respond to a broad spectrum of compounds (Carey et al.,
ronment. The latter scenario is in our view more likely. 2010, Hallem and Carlson, 2006). Response profiles of OSNs,
Insect ORs form a large and highly divergent gene family, with obtained through single sensillum recordings (SSRs) from
no close orthologies (apart from Orco) or apparent subfamily numerous other insects also suggest a spectrum of OR binding
structure conserved across insect orders Figure 3). Thus-far- affinities. Perhaps the most well-known specialist OSNs are
identified OR repertoires range in size from ten in Phthiraptera those detecting pheromones, where OSNs capable of sepa-
(i.e., lice, Kirkness et al., 2010) to
200–400 in Hymenoptera rating two enantiomers with a specificity spanning over more
(i.e., bees, ants, and wasps; Robertson et al., 2010; Wurm than four decadic concentration steps have been found (Wojta-
et al., 2011). As with the OBPs, the OR family is characterized sek et al., 1998). Highly specialized OSNs tuned to host volatiles
by species-specific expansions of single genes or gene have been identified from a number of insect species (e.g.,
subfamilies. Recently duplicated OR loci gain novel functions Mustaparta et al., 1979; Todd and Baker, 1993; Tanaka et al.,
through positive selection, presumably driven by needs arising 2009). OSN response spectra are however dependent on
from host shifts or host specializations (see below, Gardiner concentration, number of, and the ecological relevance of the
et al., 2008). These processes may also render previous adapta- screened volatiles to the organism under scrutiny. Most, if not
tions in the chemosensory repertoire void, resulting in the loss of all OSNs can be triggered to respond to almost any compound
OR genes that no longer serve a functional purpose. Analysis of if presented with high enough doses. Thus screening of volatiles
the OR repertoires of the five closest relatives of D. melanogaster at inappropriately high concentrations would give misleading
also revealed that in the food source specialist species exam- results, as would a screen with a too small stimulus battery or
ined, the OR family evolve faster as well as show a higher rate one comprised of chemicals of no relevance to the animal as
of pseudogene conversion as compared to generalist siblings key ligands might be missing.
(McBride et al., 2007). However, by extending the analysis to A solution to this problem is to use gas chromatography (GC)
all 12 sequenced Drosophila species, Gardiner and colleagues for stimulus delivery, which enables rapid screening of large
(2008) found that the proportion of pseudogenized genes did numbers of compounds selected from the habitat and ecology
not differ between the specialist and generalist taxa, whereas of the species (Figure 4). GC-linked SSR (Wadhams, 1982)
the endemic species showed significantly more losses than the experiments indeed also suggest a very high degree of speci-
mainland species. In their view, small effective population size ficity of ORs across many insect species (e.g., Wibe et al.,
and genetic drift may rather account for OR gene loss than 1997; Kristoffersen et al., 2008; Ghaninia et al., 2008). In these

702 Neuron 72, December 8, 2011 ª2011 Elsevier Inc.

Neuron

Review

Figure 4. The Native Breeding Substrates of D. melanogaster

The vinegar fly, although today cosmopolitan and a human commensal has its ancestral home in east equatorial Africa. Odor ligands that the fly’s nose has
evolved to detect should be found among the listed hosts, of which all are native to equatorial Africa and reportedly utilized by semi-wild populations of D.
melanogaster as breeding (and feeding) sites (Buruga and Olembo, 1971; Lachaise, 1974, 1979; Lachaise et al., 1982; Lachaise and Tsacas, 1983; Rio et al.,
1983). Of note is the surprisingly high number of flowers used. Very few of the listed plants have had their volatile profile examined.

experiments, hundreds (or more) volatiles were screened; unsuitable and avoided objects, such as unripe fruit, foliage,
however, only a minute fraction produced responses, with and mammals. The former group of compounds elicits positive
each OSN typically responding to few compounds of structural chemotaxis when screened individually, whereas the latter are
proximity. The detected compounds also make sense in light either ignored or repellent (Larsson et al., 2003; Bengtsson
of the examined animal’s ecology. OSNs in the vinegar fly for et al., 2009) (Figure 5C). Selective OSNs detecting odorants
example, which feeds on fermentative yeasts (typically from inhibiting host attraction have also been found in many other
fruit), accordingly detect volatiles associated with microbial insects, such as the spruce bark beetle (Ips typographus) (Ander-
activity and alcoholic fermentation, as well as compounds, which sson et al., 2009).
even though more generally occurring in nature, nevertheless are Assuming the fraction of insect species examined so far is
typical for fruit (Stensmyr et al., 2003a). The two African scarabs representative, the ORs appears to be largely divided into those
Pachnoda marginata and P. interrupta (Figure 5A), which both that detect chemicals specifically associated with key aspects of
can be found on a wide variety of flowers and rotting fruits, hence the host (or of unsuitable hosts) and to those that detect
also display OSNs narrowly tuned to compounds typical of these compounds of more general nature. The ORs tuned to specific
resources (Figures 5B and 5C). The scarabs are also equipped host odors also appear to be the most selective. In the African
with selective OSNs indicative of aspects representative of malaria mosquito, the most narrowly tuned ORs detect volatiles

Neuron 72, December 8, 2011 ª2011 Elsevier Inc. 703

 Neuron

Review

A C olfactory system appears to show a similar level of specificity.

Investigated ORs from rodents respond selectively to a small
number of structurally related compounds (e.g., Araneda et al.,
2000; but see Grosmaitre et al., 2009). Furthermore, GC-linked
electrophysiology experiments performed in the house mouse
Mus musculus, suggest narrow receptor tuning (Lin et al.,
2006) that correlates with ecologically significant odorants, sug-
B gesting a similar evolutionary strategy shapes odorant selectivity
in the mammalian olfactory system. A difference to the insect
system may however be that mammalian odorant receptors to
a larger extent appear to be tuned to sample select chemical
features, rather than select compounds. In mammals accord-
ingly, the identity of a specific chemical is likely to depend
more on combinatorial activation of a number of ORs (Buck,
2005) than is the case in insects.
To further understand the general principles underlying insect
olfactory coding, we suggest to expand the number of species
investigated, particularly from poorly sampled insect orders,
and to take evolutionary and ecological facts into careful consid-
eration. When performing these experiments it is also highly
important to use odors in relevant concentrations.

Evolutionary Trends in the Peripheral Olfactory System

of Insects
How does the insect olfactory system then respond to selective
pressure? A classic case of adaptation is the peripheral phero-
mone-detecting system often found in male insects locating
mates using female-produced sex pheromones as cues. To
Figure 5. Scarab Noses in Tune with Their Environment
(A) The fruit chafer Pachnoda marginata and the sorghum chafer P. interrupta detect the low concentrations of pheromones released by the
are native to tropical equatorial Africa and the semiarid Sahel, respectively. females (often around 0.1–10 ng per hour; e.g., Lacey and
Both species feed on a wide variety of fruit and flowers. Sanders, 1992), male insects have often added surface area to
(B) Response spectra of identified functional olfactory sensory neuron (OSN)
classes from the two species obtained through single sensillum recordings.
the antennae. This means that the antennae have become highly
Most of the OSN classes are present in both species with retained function. pectinate or multilayered leaf shaped (as seen in Figure 1A).
The heatmap illustrates the high degree of selectivity of the individual OSN Such an evolution can be compared to the vertebrate system,
groups. Numbers refer to (C). Plot based on data from Bengtsson et al. (2011). where a similar process has occurred in the nasal cavity. Animals
(C) Behavioral effect and the ecological correlates of the physiologically active
compounds. A majority of the single compounds eliciting positive chemotaxis relying heavily on high olfactory sensitivity, such as rabbits, have
(green), in either field or laboratory experiments are also indicative of objects a highly convoluted structure (Allison and Warwick, 1949), while
that the scarabs are known to utilize, whereas the repellent compounds animals less reliant on odor information, such as humans, display
(magenta) are on the other hand characteristically found in objects, which the
beetles are known to avoid. Behavioral data from Larsson et al. (2003) and
much less complicated nasal cavities (Negus, 1957). All of these
Bengtsson et al. (2009). processes, in all types of animals, serve the basic purpose
of making room for more sensor elements (OSNs), thereby
enhancing capacity to detect salient environmental cues. By
of acute biological relevance for the species, such as AgOr2 that having more detecting neurons, the chance of capturing the
is narrowly tuned to indole, a major component of human sweat information residing in an odor molecule increases. In addition
(Carey et al., 2010). Furthermore, we can assume that ORs to a numerical increase in detectors, male insects have also often
assigned to detect key features of what constitutes a host (or increased the sensitivity of each detector, thereby gaining
an unsuitable host), provide input to neuronal circuits mediating a multiplicative effect. The mechanisms behind sensitivity
innate behaviors, whereas receptors tuned to more general augmentation are still unclear, but could reside both in the
odors likely drive behavior in a context-dependent manner. number of chemoreceptors expressed on the dendritic surface
That certain nonpheromone receptors indeed mediate critical and/or in the transduction mechanisms translating the signal
information has also been illustrated in the vinegar fly, where from chemical to electrical. Analogous sexual dimorphism is
activation of a single OR gene is sufficient to elicit attraction also found in the other sensory modalities, such as the visual
toward vinegar (Semmelhack and Wang, 2009). system. For example, male insects typically have both larger
Insects appear to detect odors via a specific detection system, and morphologically more complex eyes than females (e.g.,
which is not configured to broadly sample chemical space, Beersma et al., 1977), reflecting the importance of visual cues
but constitutes a discriminating machinery tuned to select for locating mates and for securing matings (Thornhill and
compounds of relevance to the animal, where each chemore- Alcock, 1983). In certain firefly species), striking sexual dimor-
ceptor has a direct ecological correlate. The mammalian phism with respects to eye size is found, with one sex having

704 Neuron 72, December 8, 2011 ª2011 Elsevier Inc.

Neuron

Review

considerably enlarged eyes, and a visual sensitivity peak closely opportunity to feed on blood (Hill et al., 2010). Loss of innate
tuned to the conspecific bioluminescence flash signal (Lau repulsion has also been implied as a driving force for the
et al., 2007). D. sechellia-noni specialization. In this case however, loss of
Reproductive pheromone cues are not the only stimuli that can repulsion stems from altered expression of two OBPs confined
shape olfactory structure and function. For example, the fly to gustatory sensilla on the legs, which have rendered
species Drosophila sechellia, which is endemic to the Seychelles D. sechellia taste blind to the toxic acids of its host (Matsuo
archipelago and a close relative to the vinegar fly, has adapted to et al., 2007).
subsist on the native Morinda fruit that is generally toxic to most Adaptations are hence observed in parts of the peripheral
other drosophilids. In D. sechellia one specific type of olfactory olfactory system that directly interfaces with key features of
sensillum has been lost and instead replaced by a dramatic the species-specific host preference. However, shifts in ecology
increase in another type of sensillum. The increased type houses do not necessarily have to result in wide rearrangements of the
OSNs tuned to the odor of the single host, whereas the lost olfactory system. For example, across all nine members of the
express ORs with putative ligands not found in the fruit melanogaster species group, OSNs from large basiconic sensilla
(Stensmyr et al., 2003b; Dekker et al., 2006). Host driven sensory have largely conserved function, in spite of these species stem-
augmentations are also seen in Culex mosquitoes. Here, the ming from quite a wide geographic range and occupying
sensillum type that houses OSNs tuned to nonanal, a volatile different habitats (Stensmyr et al., 2003b). The presence of
characteristic of birds, are more numerous in ornitophilic Culex OSNs with highly conserved function has also been observed
taxa than in mammalophilic. The OSNs in these sensilla more- across owlet moths with disparate ecology (Stranden et al.,
over display a remarkable selectivity and sensitivity toward 2003). These core OSNs presumably detect compounds signi-
nonanal, on a par with or even surpassing that of pheromone fying key aspects of what makes up for a suitable host, regard-
OSNs found in moths. The amplified and sensitive nonanal less of the specific niche, or alternatively, detect common
detection system presumably provides the mosquitoes with compounds that are of general interest. Similarly, across droso-
improved long-range host detection (Syed and Leal, 2009). A philid flies the OR repertoire tuned to odorants connected with
high proportion of host-odor-tuned OSNs is also found in the yeast and microbes (the staple food item of virtually all droso-
grass-dwelling Japanese scarab beetle Phylloperta diversa, philds, irrespective of detailed preference), which indeed also
where the majority of the nonpheromonal olfactory sensilla con- seems to be functionally conserved (Stensmyr et al., 2003b; de
tain OSNs tuned to so-called green leaf volatiles. These OSNs Bruyne et al., 2010; Stökl et al., 2010).
likewise display an extreme degree of specificity and sensitivity, In summary, the insect olfactory system reflects the needs
and as with the mosquito, probably provide the scarab with imposed by the taxon-specific ecology. Host shifts and special-
improved long-range host detection (Hansson et al., 1999). ization leads to corresponding alterations in the odor detection
How these sensillum adaptations have been generated is machinery. The adaptations noted include increase as well
unknown, but hints of a possible molecular mechanism involving as decrease of select detector units. Although the olfactory
microRNAs come from work done by Cayirlioglu et al. (2008). systems from quite a number of insects have been examined
microRNAs are small noncoding RNA units, which posttranscrip- to date, properly controlled for, comparative functional studies
tionally suppress gene expression by binding to the nontrans- are actually rare. Additional examination of carefully chosen
lated 30 end of mRNAs of specific target genes. In the vinegar taxa of appropriate phylogenetic distance and with well-defined
fly, loss of the microRNA, miR-279, which regulates expression and contrasting ecology is accordingly needed before more solid
of the transcription factor Nerfin-1, causes ectopic formation of conclusions can be drawn.
CO2 sensing OSNs in the maxillary palps. It is accordingly
possible that other microRNAs, regulating other transcription Matching the Periphery—Evolution of the Antennal Lobe
factors are also underlying topographical reconfigurations of The adaptations at the antennal level are also reflected in the
sensilla and OSNs of other types. Interestingly, the loss of primary olfactory center of the insect brain, the antennal lobe
miR-279 creates a phenotype intermediate between that of the (AL). The AL, homologous to the olfactory bulb of vertebrates,
vinegar fly and the African malaria mosquito. If the ectopic is composed of typically spheroid structures, called glomeruli.
expression of CO2 receptors on the maxillary palps also confers All OSNs expressing the same receptor converge onto one out
a switch in behavior from repellent, as in the vinegar fly (Suh of these usually between 50 and 200 glomeruli (Vosshall et al.,
et al., 2004), to attractive, as in the mosquito (Gillies, 1980), 2000). The glomerulus also houses the branches of local inter-
remains unclear. neurons and the dendrites of projection neurons that transmit
Host shifts and specialization do not however only entail the processed information to higher brain areas (Tolbert and
increase of specific input channels but may also lead to, or Hildebrand, 1981; Distler and Boeckh, 1996). In 1924, Bretsch-
even be the result of, loss of detector channels. In the fruit- neider was the first to report the presence of a strong sexual
piercing moth Calyptrata thalictri (Lepidoptera: Noctuidae), dimorphism in the AL; male oak eggar moths, Lasiocampa
a subset of the males has been found to draw blood meals quercus (Lepidoptera: Lasiocampidae) displayed several
from mammalian hosts. This shift in behavior has been linked enlarged glomeruli at the entrance of the antennal nerve into
to a reduction of a specific group of OSNs tuned to repellent the AL (Bretschneider, 1924). Sixty years later, Koontz and
inducing vertebrate volatiles. Blood feeding could thus stem Schneider (1987) showed that these enlarged glomeruli, termed
from a loss of innate repulsive behavior to vertebrate odors, the macroglomerular complex (MGC; Boeckh and Boeckh,
leading to increased chance of zoophilic interactions and the 1979; Hildebrand et al., 1980) (Figure 6A), very likely served

Neuron 72, December 8, 2011 ª2011 Elsevier Inc. 705

 Neuron

Review

A B C D

Figure 6. Evolutionary Trends in the Insect Antennal Lobe

(A) The MGC of the hawk moth Manduca sexta (red), showing the two main glomeruli; the cumulus at the entrance of the antennal nerve resting on the toroid,
a doughnut-shaped glomerulus. Ordinary glomeruli are seen beneath the toroid in different shades of blue. Image courtesy of L. Kuebler, MPI Chemical Ecology.
(B) Although no pronounced sexual dimorphism is present in the AL of D. melanogaster, a range of Hawaiian drosophilids has evolved distinct male-specifc
macroglomeruli, presumably processing pheromone information. Adapted from Kondoh et al. (2003), with permission from The Royal Society.
(C) Macroglomeruli have not only evolved in the context of sex pheromones, but can also be found in circuits processing host odors, as illustrated here by
Drosophila sechellia, where the macroglomerulus process information regarding key volatiles of its sole host, the noni fruit. Adapted from Dekker et al. (2006), with
permission from Elsevier.
(D) In the leaf cutting ant, Atta vollenweideri, three enlarged glomeruli specific to males process trail-pheromones. Adapted from Kuebler et al. (2009), with
permission from Wiley-Liss, Inc.
(E) The evolutionary progression of antennal lobe morphology within the Orthoptera, from the basal suborder Ensifera to the more derived suborder Caelifera. In
lower orthopterans, such as in the great green bush-cricket (Tettigonia viridissima), the AL looks like in most other insects; around 50 glomeruli with uniglomerular
OSN projections and with uniglomerular projection neuron dendritic innervations. In the southern fieldcricket (Gryllus bimaculatus) each glomerulus has become
divided into a number of microglomeruli, and OSNs and PNs target several of these within the same glomerulus. In the slender ground-hopper (Tetrix subulata) the
number of glomeruli has increased substantially and OSNs target single or multiple of these, while PNs still are mainly uniglomerular. In the desert locust
(Schistocerca gregaria) the microglomerulization has reached an extreme proportion; thousands of minute microglomeruli are packed tightly within the lobe.
OSNs mostly target multiple microglomeruli, as do PNs. Schematic drawings are based on Ignell et al. (2001). Photos: M.C.S.

a purpose in receiving and processing information regarding the glomeruli have been added over evolutionary time. Such
female sex pheromone. In 1992 Hansson et al. showed that a process is suggested by the finding that OSNs carrying genet-
OSNs tuned to different pheromone components target specific ically similar ORs project to adjacent glomeruli in the antennal
glomeruli of the MGC (Hansson et al., 1992). This was indeed the lobe of the vinegar fly (Couto et al., 2005), and a similar arrange-
first clear evidence of the functional role of glomeruli as projec- ment could be envisaged in the moth MGC. Specific factors
tion areas of OSNs putatively expressing the same receptor. determining glomerulus formation has been identified both
The MGC serves as an example of how strong selection pres- morphologically (e.g., Oland and Tolbert, 1996) and molecularly
sure, here to increase the sensitivity toward sex pheromones, (e.g., Rodrigues and Hummel, 2008). These do, however, still
can create pronounced size differences among olfactory not provide a conclusive picture of how the glomerular array
glomeruli. Since the early 1990s a large number of moth species might change over evolutionary time. Interestingly, in the hawk
have been studied, and it has been shown that very often input moth and the American cockroach Periplaneta americana
regarding the main component of a sex pheromone mixture is (Blattaria: Blattidae) a subdivision of the major glomerulus (the
processed by an enlarged glomerulus, the cumulus (e.g., Hans- cumulus) has been observed (Christensen et al., 1995; Hösl,
son et al., 1991). This MGC part can then be surrounded by 1990). In both species differential innervation patterns seem to
a number of smaller satellite MGC-glomeruli receiving informa- be connected to topographical representation of the antennal
tion regarding the presence of other pheromone components, length axis.
or of behavioral antagonists preventing interspecific attraction Sexual dimorphism in the AL is not only restricted to the
(e.g., Kárpáti et al., 2008). Pheromone sensing evolved has prob- Lepidoptera. Also in drosophilid flies, sexual dimorphism with
ably evolved from simple, single-component processing, and respect to specific glomeruli has been observed (Figure 6B).
thereby one-glomerulus MGCs, to the more complex state that An investigation across 37 species of drosophilids from the
we see in most species today. This means that the MGC may Hawaiian Islands found two glomeruli enlarged in males across
be created through a budding process, where new, satellite several of the investigated species (Kondoh et al., 2003). The

706 Neuron 72, December 8, 2011 ª2011 Elsevier Inc.

Neuron

Review

homologous glomeruli in D. melanogaster (the DA1 and DL3) in life style. We propose intensified comparative studies of key
have also been shown to receive pheromonal input (van der groups, as, e.g., the orthopterans, in combination with the
Goes van Naters and Carlson, 2007). A phylogenetic comparison molecular developmental studies presently being performed in
further revealed that the noted sexual dimorphism has evolved the vinegar fly. Such a combination will allow us to reach
independently in two of the lineages. Male-specific macroglo- a considerably deeper understanding of evolutionary processes
merulus/macroglomeruli have also been found in several other molding antennal lobe architecture.
insect groups, such as, e.g., cockroaches, wasps (Hymenop-
tera: Vespidae), and bees (Hymenoptera: Apidae) (Jawlovski, Exploring Nature for Relevant Odor Ligands
1948), but is probably a much more widespread phenomenon, To understand the relevance and significance of a given neural
having evolved wherever a need for long-distance detection of circuit, one needs to know the sensory stimuli that activate it.
female produced volatile pheromones is present. In the case of the olfactory circuitry, this initially means finding
Other environmental selection pressures beyond phero- a relevant odor ligand. For the pathways mediating sexual
mones, including food and oviposition site-associated odors, behaviors, the ligand is typically a pheromone, and the isolation
can also shape glomerular organization and structure. For and identification of which is nowadays mostly a technical
example, the two glomeruli (DM2 and VM5d) in the fly matter. Identifying odor ligands activating circuits underlying
D. sechellia, targeted by OSNs tuned to its singular food source, other important behaviors is however in many cases a more
the noni-fruit, are 200% larger in both sexes relative to daunting task even if detailed knowledge of the animal’s ecology
D. melanogaster (Dekker et al., 2006) (Figure 6C). Interestingly, is at hand. Help can however be drawn from a slightly unex-
the expansion of the noni-fruit specific detection system in pected direction, namely deceptive plants. A wide range of plant
D. sechellia not only provides higher sensitivity to the fruit odors, is known to trick insects into pollination without providing
but it also makes the fly tolerant to much higher odor concentra- a reward. To accomplish this feat, these plants all rely on being
tions that would inhibit attraction in all other fruit flies. The mech- able to trigger and to exploit neural circuits underlying obligate
anisms underlying this dual function are still unclear. and innate attraction in the targeted insects. In short, the plants
Similar alterations in glomerular size for odors other than sex copy signals that the intended victims of the deception cannot
pheromones has also been observed in workers and males of afford to ignore. Although visual and tactile cues are in many
leaf-cutting ants (Hymenoptera: Formicidae), which have one instances important, most often the key to success resides
and three greatly enlarged glomerulus respectively (Kleineidam with the plants being able to mimic odors of importance to the
et al., 2005), which presumably process trail-pheromone insects (Urru et al., 2011). Accordingly, deceptive plants can
components (Kuebler et al., 2010) (Figure 6D). Female M. sexta provide unique insights into what constitutes a critical resource
also show two enlarged glomeruli, which are specific to a set to the targeted insect and what sensory cues mediate the attrac-
of host plant volatiles and accordingly assumed to be involved tion to this resource.
in behaviors specific to the females, probably in locating and The dead horse arum (Helicodiceros muscivorus) and the
selecting suitable oviposition sites (King et al., 2000). Solomon’s lily (Arum palaestinum) serves as excellent examples
An interesting example of AL evolution is found within the of how deceptive plants can be used to identify important odor
order Orthoptera, which includes, e.g., grasshoppers, crickets, ligands. The former produces a ghastly smell, reminiscent
and wetas. When comparing the grasshopper and locust to other of rotting flesh and also attracts carrion blowflies (Diptera:
orthopteran insects it is clear that a strong evolutionary trend Calliphoridae), the latter has in contrast a pleasant smell, similar
from a ‘‘normal’’ glomerular system with unbranched OSN axons to fruity wine and instead attracts drosophilid flies. The apparent
in primitive orthopterans to a microglomerular system with carrion mimicry is remarkably simply accomplished, via the
branched input neurons in grasshoppers and locusts is present production of just three compounds, namely dimethyl mono-,
in the AL structure (Ignell et al., 2001) (Figure 6E). The functional di-, and trisulfide (Stensmyr et al., 2002). The mimicry of alcoholic
significance of a system evolving from a glomerular architecture fermentation is likewise accomplished via only a handful of odor-
with unbranched OSNs and with most PNs targeting single ants, including e.g., acetoin acetate and 2,3-butanediol acetate
glomeruli, into a system with thousands of microglomeruli (Stökl et al., 2010). The deception nevertheless works since the
innervated by highly branched OSNs and PNs is still unclear. copied odors are diagnostic for the targeted insects favored
By allowing a much more diverse interaction between OSNs oviposition sites (i.e., decomposing animals and rotting fruit
and PNs such a system could potentially increase the coding respectively), whereas they are very rarely present in other
capacity. The functional characteristics among orthopteran substrates. These plants hence nicely demonstrate the principle
olfactory systems, however, still remain to be elucidated, and that insects rely on a select set of chemicals to localize essential
this is an area where we see progress adding significantly to resources.
our understanding of the evolution of the insect sense of smell. Systems built on sensory deceit are thus excellent sources of
In general, the insect antennal lobe offers an excellent information regarding key stimuli for the dupe. The mimicking
substrate to study evolutionary processes in olfaction. Even flowers produce odors to which olfactory receptors in insects
though insects have radiated into so many different species very likely have evolved high affinity. Having access to such
and life forms, the antennal lobe of neopteran insects has main- ligands is of course of utmost importance when dissecting the
tained its basic architecture with incremental steps of change neural function of the olfactory system, from periphery to brain,
introduced over evolutionary time. This fact makes it possible and further deepens our understanding of insect behavior. Inves-
to follow these changes and often to connect them to changes tigations of such systems should be carefully selected among

Neuron 72, December 8, 2011 ª2011 Elsevier Inc. 707

 Neuron

Review

plants duping interesting target species. Vinegar flies is a natural The combination of modern molecular methods, applied in
candidate, but, relating to our suggestions above, finding flowers model insects as the vinegar fly and the malaria mosquito, in
that target primitive insects as pollinators would be highly valu- combination with well-designed comparative studies with a
able, as would identifying plants/flowers that could be used as well-founded phylogenetic background is what we suggest will
deceptive traps for insects of public health (e.g., mosquitoes) allow us to optimize our search for deeper understanding of
and agricultural economic concern (e.g., beetles). olfaction and its evolution. All of these investigations should
always be approached with the natural behavior and habitat of
Conclusion the organism in mind.
The insect olfactory system and its ability to evolve over relatively
short time spans is probably an important part of the explanation ACKNOWLEDGMENTS
why insects are such successful organisms. This success is
manifested in the fact that insects have occupied almost every The writing of this review was funded by the Max Planck Society.
imaginable ecological niche. The fact that insects adapt to all
REFERENCES
these different conditions at the same time provides us with
a plethora of fascinating examples of adaptations, both in the Allison, A.C., and Warwick, R.T.T. (1949). Quantitative observations on the
peripheral sensory organs and the brain, and it allows us to olfactory system of the rabbit. Brain 72, 186–197.
observe evolution in action. The development of sensitive
Andersson, M.N., Larsson, M.C., and Schlyter, F. (2009). Specificity and
peripheral detection systems seems to be important in shaping redundancy in the olfactory system of the bark beetle Ips typographus:
also the primary central centers. Glomeruli are added to accom- single-cell responses to ecologically relevant odors. J. Insect Physiol. 55,
556–567.
modate OSNs expressing newly evolved receptor proteins, and
glomeruli expand or contract as the number of OSNs expressing Araneda, R.C., Kini, A.D., and Firestein, S. (2000). The molecular receptive
a certain receptor change in absolute numbers. Enigmatic range of an odorant receptor. Nat. Neurosci. 3, 1248–1255.
architectures, such as the Orthopteran antennal lobe and its Beersma, D.G.M., Stavenga, D.G., and Kuiper, J.W. (1977). Retinal lattice,
innervation do, however, still puzzle those of us studying insect visual field, and binocularity in flies: dependence on species and sex. J.
Comp. Physiol. 119, 207–220.
olfaction and its evolution.
These differences in structure show us how relatively fast Bengtsson, J.M., Wolde-Hawariat, Y., Khbaish, H., Negash, M., Jembere, B.,
sensory systems can adapt to altered external conditions or Seyoum, E., Hansson, B.S., Larsson, M.C., and Hillbur, Y. (2009). Field attrac-
tants for Pachnoda interrupta selected by means of GC-EAD and single
new lifestyles. Still, however, we lack insights into how the neural sensillum screening. J. Chem. Ecol. 35, 1063–1076.
circuitry, both at the micro and the macro scale, adapts to these
Bengtsson, J.B., Khbaish, H., Wolde-Hawariat, Y., Reinecke, A., Negash, M.,
changes. Future comparative studies must therefore make use Seyoum, E., Hansson, B.S., Hillbur, Y., and Larsson, M.C. (2011). Conserved,
of high-resolution techniques, combining detailed investigations highly specialized olfactory receptor neurons for food compounds in 2 conge-
of connectivity in primary olfactory centers with functional neric scarab beetles, Pachnoda interrupta and Pachnoda marginata. Chem.
Senses 36, 499–513.
studies of the elements identified. Only then can we obtain
conclusive information regarding the connection between neural Benton, R., Sachse, S., Michnick, S.W., and Vosshall, L.B. (2006). Atypical
membrane topology and heteromeric function of Drosophila odorant receptors
function and behavior, and of the evolution of olfactory function. in vivo. PLoS Biol. 4, e20.
These kinds of data are presently being produced in the model
insect, D. melanogaster, but we still lack any kind of detailed Bianchet, M.A., Bains, G., Pelosi, P., Pevsner, J., Snyder, S.H., Monaco, H.L.,
and Amzel, L.M. (1996). The three-dimensional structure of bovine odorant
information from other insects. A future goal must therefore be binding protein and its mechanism of odor recognition. Nat. Struct. Biol. 3,
to identify species that will provide data from both an adaptive 934–939.
and a phylogenetic standpoint, and use these to build a database Boddum, T., Skals, N., Hill, S.R., Hansson, B.S., and Hillbur, Y. (2010). Gall
where neuroethologically and evolutionarily relevant data can be midge olfaction: pheromone sensitive olfactory neurons in Contarinia nasturtii
gathered and compared. and Mayetiola destructor. J. Insect Physiol. 56, 1306–1314.

When a system evolves toward high efficiency, it will also be Boeckh, J., and Boeckh, V. (1979). Threshold and odor specificity of phero-
highly suited to trigger innate attraction and/or repulsion. The mone-sensitive neurons in the deutocerebrum of Antheraea pernyi and
A. polyphemus (Saturniidae). J. Comp. Physiol. A Neuroethol. Sens. Neural
system can be ‘‘trusted’’ to deliver reliable information regarding Behav. Physiol. 132, 235–242.
a resource. Such specificity also opens up for exploitation.
Boursot, P., Din, W., Anand, R., Darviche, D., Dod, B., Von Deimling, F., Talwar,
Flowers dupe insects into doing their bidding by imitating irre- G.P., and Bonhomme, F. (1996). Origin and radiation of the house mouse:
sistible odors. These deceptive systems offer us unique opportu- Mitochondrial DNA phylogeny. J. Evol. Biol. 9, 391–415.
nities to explore how olfactory sensitivies are tuned through
Bretschneider, F. (1924). Über die Gehirne des eichenspinners und des
evolution, whereby certain odorants come to represent key Seidenspinners (Lasiocampa quercus L. und Bombyx mori L.). Jena.
behaviorally salient cues. Our aim with the present review is to Z. Naturw. 60, 563–570.
generally raise awareness as to the interesting and unique Buck, L.B. (2005). Unraveling the sense of smell (Nobel lecture). Angew. Chem.
cross-disciplinary neurobiological insights that can be gained Int. Ed. Engl. 44, 6128–6140.
from neurethological paradigms, particularly as they relate to
Buruga, J.H., and Olembo, R. (1971). Plant food preferences of some
olfaction. As is obvious from our discussion, much still remains sympatric Drosophilidae of Tropical Africa. Biotropica 3, 151–158.
to be discovered regarding how olfaction works and evolves,
Carey, A.F., Wang, G., Su, C.-Y., Zwiebel, L.J., and Carlson, J.R. (2010).
and with three million species of insects probably still not Odorant reception in the malaria mosquito Anopheles gambiae. Nature 464,
described, numerous interesting cases await to be examined. 66–71.

708 Neuron 72, December 8, 2011 ª2011 Elsevier Inc.

Neuron

Review
Cayirlioglu, P., Kadow, I.G., Zhan, X., Okamura, K., Suh, G.S., Gunning, D., Lai, Hansson, B.S., Ljungberg, H., Hallberg, E., and Löfstedt, C. (1992). Functional
E.C., and Zipursky, S.L. (2008). Hybrid neurons in a microRNA mutant are specialization of olfactory glomeruli in a moth. Science 256, 1313–1315.
putative evolutionary intermediates in insect CO2 sensory systems. Science
319, 1256–1260. Hansson, B.S., Larsson, M.C., and Leal, W.S. (1999). Green leaf volatile-
detecting olfactory receptor neurones display very high sensitivity and speci-
Chittka, L., and Niven, J. (2009). Are bigger brains better? Curr. Biol. 19, R995– ficity in a scarab beetle. Physiol. Entomol. 24, 121–126.
R1008.
Hekmat-Scafe, D.S., Scafe, C.R., McKinney, A.J., and Tanouye, M.A. (2002).
Christensen, T.A., Harrow, I.D., Cuzzocrea, C., Randolph, P.W., and Hilde- Genome-wide analysis of the odorant-binding protein gene family in
brand, J.G. (1995). Distinct projections of two populations of olfactory receptor Drosophila melanogaster. Genome Res. 12, 1357–1369.
axons in the antennal lobe of the sphinx moth Manduca sexta. Chem. Senses
20, 313–323. Hildebrand, J.G., Matsumoto, S.G., Camazine, S.M., Tolbert, L.P., Blank, S.,
Ferguson, H., and Ecker, V. (1980). Organisation and physiology of antennal
Clyne, P.J., Warr, C.G., Freeman, M.R., Lessing, D., Kim, J., and Carlson, J.R. centres in the brain of the moth Manduca sexta. In Insect Neurobiology and
(1999). A novel family of divergent seven-transmembrane proteins: candidate Pesticide Action (Neurotox 79) (London: Society of Chemical Industry),
odorant receptors in Drosophila. Neuron 22, 327–338. pp. 375–382.

Clyne, P.J., Warr, C.G., and Carlson, J.R. (2000). Candidate taste receptors in Hill, S.R., Zaspel, J., Weller, S., Hansson, B.S., and Ignell, R. (2010). To be or
Drosophila. Science 287, 1830–1834. not to be. a vampire: a matter of sensillum numbers in Calyptra thalictri?
Arthropod Struct. Dev. 39, 322–333.
Couto, A., Alenius, M., and Dickson, B.J. (2005). Molecular, anatomical, and
Hösl, M. (1990). Pheromone-sensitive neurons in the deutocerebrum of
functional organization of the Drosophila olfactory system. Curr. Biol. 15,
Periplaneta americana: receptive fields on the antenna. J. Comp. Physiol.
1535–1547.
A Neuroethol. Sens. Neural Behav. Physiol. 167, 321–327.
de Bruyne, M., Clyne, P.J., and Carlson, J.R. (1999). Odor coding in a model Ignell, R., Anton, S., and Hansson, B.S. (2001). The antennal lobe of
olfactory organ: the Drosophila maxillary palp. J. Neurosci. 19, 4520–4532. orthoptera - anatomy and evolution. Brain Behav. Evol. 57, 1–17.
de Bruyne, M., Smart, R., Zammit, E., and Warr, C.G. (2010). Functional and Jawlovski, H. (1948). Studies on the insect’s brain. Ann. Univ. M. Curie
molecular evolution of olfactory neurons and receptors for aliphatic esters Sklodowska C 3, 1–30.
across the Drosophila genus. J. Comp. Physiol. A Neuroethol. Sens. Neural
Behav. Physiol. 196, 97–109. Kaissling, K.-E., Hildebrand, J.G., and Tumlinson, J.H. (1989). Pheromone
receptor cells in the male moth Manduca sexta. Arch. Insect Biochem. Physiol.
Dekker, T., Ibba, I., Siju, K.P., Stensmyr, M.C., and Hansson, B.S. (2006). 10, 273–279.
Olfactory shifts parallel superspecialism for toxic fruit in Drosophila
melanogaster sibling, D. sechellia. Curr. Biol. 16, 101–109. Kárpáti, Z., Dekker, T., and Hansson, B.S. (2008). Reversed functional
topology in the antennal lobe of the male European corn borer. J. Exp. Biol.
Dethier, V.G. (1947). Chemical Insect Attractants and Repellents (Philadelphia: 211, 2841–2848.
Blakiston), pp. 289.
King, J.R., Christensen, T.A., and Hildebrand, J.G. (2000). Response charac-
Distler, P.G., and Boeckh, J. (1996). Synaptic connection between olfactory teristics of an identified, sexually dimorphic olfactory glomerulus. J. Neurosci.
receptor cells and uniglomerular projection neurons in the antennal lobe of 20, 2391–2399.
the American cockroach, Periplaneta americana. J. Comp. Neurol. 370, 35–46.
Kirkness, E.F., Haas, B.J., Sun, W., Braig, H.R., Perotti, M.A., Clark, J.M., Lee,
Dobritsa, A.A., van der Goes van Naters, W., Warr, C.G., Steinbrecht, R.A., and S.H., Robertson, H.M., Kennedy, R.C., Elhaik, E., et al. (2010). Genome
Carlson, J.R. (2003). Integrating the molecular and cellular basis of odor coding sequences of the human body louse and its primary endosymbiont provide
in the Drosophila antenna. Neuron 37, 827–841. insights into the permanent parasitic lifestyle. Proc. Natl. Acad. Sci. USA
107, 12168–12173.
Forêt, S., and Maleszka, R. (2006). Function and evolution of a gene family
encoding odorant binding-like proteins in a social insect, the honey bee Kleineidam, C.J., Obermayer, M., Halbich, W., and Rössler, W. (2005). A
(Apis mellifera). Genome Res. 16, 1404–1413. macroglomerulus in the antennal lobe of leaf-cutting ant workers and its
possible functional significance. Chem. Senses 30, 383–392.
Gardiner, A., Barker, D., Butlin, R.K., Jordan, W.C., and Ritchie, M.G. (2008).
Drosophila chemoreceptor gene evolution: selection, specialization and Knudsen, E.I., and Konishi, M. (1978). A neural map of auditory space in the
genome size. Mol. Ecol. 17, 1648–1657. owl. Science 200, 795–797.

Ghaninia, M., Larsson, M.C., Hansson, B.S., and Ignell, R. (2008). Natural odor Kondoh, Y., Kaneshiro, K.Y., Kimura, K., and Yamamoto, D. (2003). Evolution
ligands for olfactory receptor neurons of the female mosquito Aedes aegypti: of sexual dimorphism in the olfactory brain of Hawaiian Drosophila. Proc. Biol.
use of gas chromatography-linked single sensillum recordings. J. Exp. Biol. Sci. 270, 1005–1013.
211, 3020–3027.
Koontz, M.A., and Schneider, D. (1987). Sexual dimorphism in neuronal projec-
tions from the antennae of silk moths (Bombyx mori, Antherea polyphemus)
Gillies, M.T. (1980). The role of carbon dioxide in host-finding in mosquitoes
and the gypsy moth (Lymantria dispar). Cell Tissue Res. 249, 39–50.
(Diptera: Culicidae): a review. Bull. Entomol. Res. 70, 525–532.
Kristoffersen, L., Larsson, M.C., and Anderbrant, O. (2008). Functional charac-
Goldman, A.L., Van der Goes van Naters, W., Lessing, D., Warr, C.G., and teristics of a tiny but specialized olfactory system: olfactory receptor neurons
Carlson, J.R. (2005). Coexpression of two functional odor receptors in one of carrot psyllids (Homoptera: Triozidae). Chem. Senses 33, 759–769.
neuron. Neuron 45, 661–666.
Kuebler, L.S., Kelber, C., and Kleineidam, C.J. (2010). Distinct antennal lobe
Grimaldi, D., and Engel, M.S. (2005). Evolution of the Insects (Cambridge, UK: phenotypes in the leaf-cutting ant (Atta vollenweideri). J. Comp. Neurol. 518,
Cambridge Univ Press). 352–365.
Grosmaitre, X., Fuss, S.H., Lee, A.C., Adipietro, K.A., Matsunami, H., Lacey, M.J., and Sanders, C.J. (1992). Chemical composition of sex phero-
Mombaerts, P., and Ma, M. (2009). SR1, a mouse odorant receptor with an mone of oriental fruit moth and rates of release by individual female moths.
unusually broad response profile. J. Neurosci. 29, 14545–14552. J. Chem. Ecol. 18, 1421–1435.

Hallem, E.A., and Carlson, J.R. (2006). Coding of odors by a receptor reper- Lachaise, D. (1974). Les Drosophilidae des savanes preforestieres de Lamto
toire. Cell 125, 143–160. (Cote-d’Ivoire). V. Les regimes alimentaires. Ann. Soc. Entomol. Fr. 10, 3–50.

Hansson, B.S., Christensen, T.A., and Hildebrand, J.G. (1991). Functionally Lachaise, D. (1979). Speciation, Coevolution et Adaptation des populations de
distinct subdivisions of the macroglomerular complex in the antennal lobe of Drosophilides en Afrique tropicale. These Doctorat d’Etat, Universite Pierre et
the male sphinx moth Manduca sexta. J. Comp. Neurol. 312, 264–278. Marie Curie, Paris.

Neuron 72, December 8, 2011 ª2011 Elsevier Inc. 709

 Neuron

Review
Lachaise, D., and Tsacas, L. (1983). In The Genetics and Biology of Drosophila, Scheich, H., Langner, G., Tidemann, C., Coles, R.B., and Guppy, A. (1986).
M. Ashburner, H.L. Carson, and J.N. Thompson, Jr., eds. (London: Academic Electroreception and electrolocation in platypus. Nature 319, 401–402.
Press), pp. 221–331.
Schneider, D. (1964). Insect antennae. Annu. Rev. Entomol. 9, 103–122.
Lachaise, D., Tsacas, L., and Couturier, G. (1982). The Drosophilidae associ-
ated with tropical African figs. Evolution 36, 141–151. Schneider, D., and Steinbrecht, R.A. (1968). Checklist of insect olfactory
sensilla. Symp. Zool. Soc. Lond. 23, 279–297.
Lachaise, D., Cariou, M.-L., David, J.R., Lemeunier, F., Tsacas, L., and
Ashburner, M. (1988). Historical biogeography of the Drosophila melanogaster Scott, K., Brady, R., Jr., Cravchik, A., Morozov, P., Rzhetsky, A., Zuker, C., and
species subgroup. Evol. Biol. 22, 159–225. Axel, R. (2001). A chemosensory gene family encoding candidate gustatory
and olfactory receptors in Drosophila. Cell 104, 661–673.
Larsson, M.C., Stensmyr, M.C., Bice, S.B., and Hansson, B.S. (2003). Attrac-
tiveness of fruit and flower odorants detected by olfactory receptor neurons in Semmelhack, J.L., and Wang, J.W. (2009). Select Drosophila glomeruli
the fruit chafer Pachnoda marginata. J. Chem. Ecol. 29, 1253–1268. mediate innate olfactory attraction and aversion. Nature 459, 218–223.

Larsson, M.C., Domingos, A.I., Jones, W.D., Chiappe, M.E., Amrein, H., and Shanbhag, S.R., Müller, B., and Steinbrecht, R.A. (1999). Atlas of olfactory
Vosshall, L.B. (2004). Or83b encodes a broadly expressed odorant receptor organs of Drosophila melanogaster. 1. Types, external organization, innerva-
essential for Drosophila olfaction. Neuron 43, 703–714. tion, and distribution of olfactory sensilla. Int. J. Insect Morphol. Embryol.
28, 377–397.
Lau, T.F.S., Ohba, N., Arikawa, N., and Meyer-Rochow, V.B. (2007). Sexual
dimorphism in the compound eyes of Rhagophthalmus ohbai (Coleoptera; Shiraiwa, T. (2008). Multimodal chemosensory integration through the maxil-
Rhagophthalmidae): II. Physiology and function of the eye of the male. lary palp in Drosophila. PLoS ONE 3, e2191.
J. Asia Pac. Entomol. 10, 27–31.
Steinbrecht, R.A. (1997). Pore structures in insect olfactory sensilla: A review of
Laughlin, J.D., Ha, T.S., Jones, D.N., and Smith, D.P. (2008). Activation of data and concepts. Int. J. Insect Morphol. Embryol. 26, 229–245.
pheromone-sensitive neurons is mediated by conformational activation of
pheromone-binding protein. Cell 133, 1255–1265. Stensmyr, M.C., Urru, I., Collu, I., Celander, M., Hansson, B.S., and Angioy,
A.-M. (2002). Pollination: Rotting smell of dead-horse arum florets. Nature
Lin, Y., Shea, S.D., and Katz, L.C. (2006). Representation of natural stimuli in 420, 625–626.
the rodent main olfactory bulb. Neuron 50, 937–949.
Stensmyr, M.C., Giordano, E., Balloi, A., Angioy, A.-M., and Hansson, B.S.
Lu, T., Qiu, Y.T., Wang, G., Kwon, J.Y., Rutzler, M., Kwon, H.-W., Pitts, R.J., (2003a). Novel natural ligands for Drosophila olfactory receptor neurones.
van Loon, J.J., Takken, W., Carlson, J.R., and Zwiebel, L.J. (2007). Odor J. Exp. Biol. 206, 715–724.
coding in the maxillary palp of the malaria vector mosquito Anopheles
gambiae. Curr. Biol. 17, 1533–1544. Stensmyr, M.C., Dekker, T., and Hansson, B.S. (2003b). Evolution of the
olfactory code in the Drosophila melanogaster subgroup. Proc. Biol. Sci.
Matsuo, T., Sugaya, S., Yasukawa, J., Aigaki, T., and Fuyama, Y. (2007). 270, 2333–2340.
Odorant-binding proteins OBP57d and OBP57e affect taste perception and
host-plant preference in Drosophila sechellia. PLoS Biol. 5, e118. Stökl, J., Strutz, A., Dafni, A., Svatos, A., Doubský, J., Knaden, M., Sachse, S.,
Hansson, B.S., and Stensmyr, M.C. (2010). A deceptive pollination system
McBride, C.S., Arguello, J.R., and O’Meara, B.C. (2007). Five Drosophila targeting drosophilids through olfactory mimicry of yeast. Curr. Biol. 20,
genomes reveal nonneutral evolution and the signature of host specialization 1846–1852.
in the chemoreceptor superfamily. Genetics 177, 1395–1416.
Störtkuhl, K.F., and Kettler, R. (2001). Functional analysis of an olfactory
Mustaparta, H., Angst, M.E., and Lanier, G.N. (1979). Specialization of olfac-
receptor in Drosophila melanogaster. Proc. Natl. Acad. Sci. USA 98, 9381–
tory cells to insect- and host-produced volatiles in bark beetle Ips pini (Say).
9385.
J. Chem. Ecol. 5, 109–123.
Stranden, M., Liblikas, I., König, W.A., Almaas, T.J., Borg-Karlson, A.-K., and
Negus, V. (1957). The evolutionary history of man from the evidence of the nose
Mustaparta, H. (2003). (-)-Germacrene D receptor neurones in three species
and larynx. AMA Arch. Otolaryngol. 66, 414–429.
of heliothine moths: structure-activity relationships. J. Comp. Physiol.
Oland, L.A., and Tolbert, L.P. (1996). Multiple factors shape development of A Neuroethol. Sens. Neural Behav. Physiol. 189, 563–577.
olfactory glomeruli: insights from an insect model system. J. Neurobiol. 30,
Suh, G.S., Wong, A.M., Hergarden, A.C., Wang, J.W., Simon, A.F., Benzer, S.,
92–109.
Axel, R., and Anderson, D.J. (2004). A single population of olfactory sensory
Payne, R.S. (1971). Acoustic location of prey by barn owls (Tyto alba). J. Exp. neurons mediates an innate avoidance behaviour in Drosophila. Nature 431,
Biol. 54, 535–573. 854–859.

Pitts, R.J., and Zwiebel, L.J. (2006). Antennal sensilla of two female anopheline Swarup, S., Williams, T.I., and Anholt, R.R.H. (2011). Functional dissection of
sibling species with differing host ranges. Malar. J. 5, 26. Odorant binding protein genes in Drosophila melanogaster. Genes Brain
Behav. 10, 648–657.
Rio, B., Couturier, G., Lemeunier, F., and Lachaise, D. (1983). Evolution d’une
specialisation saisonniere chez Drosophila erecta (Dipt. Drosophilidae). Ann. Syed, Z., and Leal, W.S. (2009). Acute olfactory response of Culex mosquitoes
Soc. Entomol. Fr. 19, 235–248. to a human- and bird-derived attractant. Proc. Natl. Acad. Sci. USA 106,
18803–18808.
Robertson, H.M., Warr, C.G., and Carlson, J.R. (2003). Molecular evolution of
the insect chemoreceptor gene superfamily in Drosophila melanogaster. Proc. Tanaka, K., Uda, Y., Ono, Y., Nakagawa, T., Suwa, M., Yamaoka, R., and
Natl. Acad. Sci. USA 100 (Suppl 2 ), 14537–14542. Touhara, K. (2009). Highly selective tuning of a silkworm olfactory receptor
to a key mulberry leaf volatile. Curr. Biol. 19, 881–890.
Robertson, H.M., Gadau, J., and Wanner, K.W. (2010). The insect chemore-
ceptor superfamily of the parasitoid jewel wasp Nasonia vitripennis. Insect Thom, C., Guerenstein, P.G., Mechaber, W.L., and Hildebrand, J.G. (2004).
Mol. Biol. 19 (Suppl 1 ), 121–136. Floral CO2 reveals flower profitability to moths. J. Chem. Ecol. 30, 1285–1288.

Rodrigues, V., and Hummel, T. (2008). Development of the Drosophila olfac- Thornhill, R., and Alcock, J. (1983). The Evolution of Insect Mating Systems
tory system. Adv. Exp. Med. Biol. 628, 82–101. (Cambridge: Harvard University Press).

Sato, K., Pellegrino, M., Nakagawa, T., Nakagawa, T., Vosshall, L.B., and Todd, J.L., and Baker, T.C. (1993). Response of single antennal neurons
Touhara, K. (2008). Insect olfactory receptors are heteromeric ligand-gated of female cabbage loopers to behaviorally active attractants. Naturwissen-
ion channels. Nature 452, 1002–1006. schaften 80, 183–186.

Schaller, A. (1926). Sinnesphysiologische und psychologische Untersuchun- Tolbert, L.P., and Hildebrand, J.G. (1981). Organization and synaptic ultra-
gen an Wasserkafern und Fischen. Z. Vgl. Physiol. 4, 370–464. structure of glomeruli in the antennal lobes of the moth Manduca sexta: A study

710 Neuron 72, December 8, 2011 ª2011 Elsevier Inc.

Neuron

Review
using thin sections and freeze-fracture. Proc. R. Soc. Lond. B Biol. Sci. 213, Wadhams, L.J. (1982). Coupled gas chromatography - Single cell recording:
279–301. a new Technique for use in the analysis of insect pheromones. Z. Naturforsch.
C 37c, 947–952.
Urru, I., Stensmyr, M.C., and Hansson, B.S. (2011). Pollination by brood-site
deception. Phytochemistry 72, 1655–1666. Wibe, A., Borg-Karlson, A.K., Norin, T., and Mustaparta, H. (1997). Identifica-
tion of plant volatiles activating single receptor neurons in the pine weevil
van der Goes van Naters, W., and Carlson, J.R. (2007). Receptors and neurons
(Hylobius abietis). J. Comp. Physiol. A Neuroethol. Sens. Neural Behav.
for fly odors in Drosophila. Curr. Biol. 17, 606–612.
Physiol. 180, 585–595.
Vieira, F.G., and Rozas, J. (2011). Comparative genomics of the odorant-
binding and chemosensory protein gene families across the Arthropoda: origin Wicher, D., Schäfer, R., Bauernfeind, R., Stensmyr, M.C., Heller, R., Heine-
and evolutionary history of the chemosensory system. Genome Biol Evol 3, mann, S.H., and Hansson, B.S. (2008). Drosophila odorant receptors are
476–490. both ligand-gated and cyclic-nucleotide-activated cation channels. Nature
452, 1007–1011.
Vieira, F.G., Sánchez-Gracia, A., and Rozas, J. (2007). Comparative genomic
analysis of the odorant-binding protein family in 12 Drosophila genomes: Wojtasek, H., Hansson, B.S., and Leal, W.S. (1998). Attracted or repelled?—
purifying selection and birth-and-death evolution. Genome Biol. 8, R235. a matter of two neurons, one pheromone binding protein, and a chiral center.
Biochem. Biophys. Res. Commun. 250, 217–222.
Vogt, R.G., and Riddiford, L.M. (1981). Pheromone binding and inactivation by
moth antennae. Nature 293, 161–163.
Wurm, Y., Wang, J., Riba-Grognuz, O., Corona, M., Nygaard, S., Hunt, B.G.,
Vosshall, L.B., Amrein, H., Morozov, P.S., Rzhetsky, A., and Axel, R. (1999). Ingram, K.K., Falquet, L., Nipitwattanaphon, M., Gotzek, D., et al. (2011).
A spatial map of olfactory receptor expression in the Drosophila antenna. The genome of the fire ant Solenopsis invicta. Proc. Natl. Acad. Sci. USA
Cell 96, 725–736. 108, 5679–5684.

Vosshall, L.B., Wong, A.M., and Axel, R. (2000). An olfactory sensory map in Zacharuk, R.Y. (1980). Ultrastructure and Function of Insect Chemosensilla.
the fly brain. Cell 102, 147–159. Annu. Rev. Entomol. 25, 27–47.

Neuron 72, December 8, 2011 ª2011 Elsevier Inc. 711