Open Access
How to translate text using browser tools
1 December 2004 Further exploration into the adaptive design of the arthropod “microbrain”: I. Sensory and memory-processing systems
Makoto Mizunami, Fumio Yokohari, Masakazu Takahata
Author Affiliations +

Arthropods have small but sophisticated brains that have enabled them to adapt their behavior to a diverse range of environments. In this review, we first discuss some of general characteristics of the arthropod “microbrain” in comparison with the mammalian “megalobrain”. Then we discuss about recent progress in the study of sensory and memory-processing systems of the arthropod “microbrain”. Results of recent studies have shown that (1) insects have excellent capability for elemental and context-dependent forms of olfactory learning, (2) mushroom bodies, higher olfactory and associative centers of arthropods, have much more elaborated internal structures than previously thought, (3) many genes involved in the formation of basic brain structures are common among arthropods and vertebrates, suggesting that common ancestors of arthropods and vertebrates already had organized head ganglia, and (4) the basic organization of sensori-motor pathways of the insect brain has features common to that of the mammalian brain. These findings provide a starting point for the study of brain mechanisms of elaborated behaviors of arthropods, many of which remain unexplored.


Arthropods are the most diverse and abundant animal group, representing about 80% of all known animal species (Brusca and Brusca, 1990), and they play a vital role in the Earth's ecosystem. These animals have successfully adapted themselves to virtually every habitat and have developed amazingly diverse forms of behavior. The enormous evolutionary success of arthropods in terms of species richness and diversity depends, at least in part, on the sophistication of their brains. We have suggested that many characteristic features of the arthropod “microbrain” compared with the mammalian “megalobrain” can be ascribed not only to the smaller number of its constituting neurons but also to the optimization to life with a smaller body (Mizunami et al., 1999).

Recently, a research project aimed at revealing adaptive designs of the arthropod “microbrain” was performed in Japan. The project focused on two subjects, sensory and memory-processing systems in the head ganglia, i.e., the “microbrain” proper, and the distributed processing in the whole central nervous system, including thoracic and abdominal ganglia. Here we review some of the recent progress in the understanding of sensory and memory-processing systems in the arthropod “microbrain”, focusing on the accomplishments of the “microbrain” research project. First, we briefly discuss functional characteristics of the arthropod “microbrain” in comparison with the mammalian “megalobrain”, and then we proceed to results of experimental studies.

Characteristic features of the arthropod “microbrain” in comparison with the mammalian “megalobrain”

Proposed functional characteristics of the arthropod “microbrain” in comparison with those of the mammalian “megalobrain” are shown in Table 1. These proposed functional characteristics are based mainly on the assumption that the arthropod “microbrain” is optimized to support life style and behavior of relatively small and short-lived animals (relatively r-selected animals; Pianka, 1974), while the mammalian “megalobrain” is optimized for the life of large and long-lived animals (relatively K-selected animals), as discussed by Mizunami et al. (1999). For example, it is thought that (1) a singly identifiable neuron or a small number of neurons often serve as a functional unit for information processing in the arthropod “microbrain”, while a large number of neurons that comprise a column or layer serve as a functional unit in the mammalian “megalobrain”, (2) the hierarchies in sensory and motor systems are much shallower in the “microbrain” than in the “megalobrain”, (3) a major objective of visual processing in arthropods is extraction of a relatively small number of biologically significant features (Wehner 1987), while that in mammals is formation of map-like representations of the external sensory features, (4) motion vision is important for small and fast-moving animals, while discrimination of detailed features of targets by utilizing large number of neurons is achieved in large brains, (5) learning capability is less important for short-lived animals, including many arthropods, than for long-lived animals, including many mammals (Mayr, 1974), (6) information storage in small brains is often achieved by modifying particular synapses in particular sensorimotor pathways, while mammalian brains are equipped with neural circuits for general memory storage, (7) priority in motor control for small and light-weight animals is the achievement of rapid movement, while that for large and heavy animals is the achievement of precise movement and (8) adaptive changes in motor programs in small brains may be achieved by switching among a small number of relatively stereotyped motor programs, while more flexible changes in motor programs are achieved in large brains. Although the validity of each item in Table 1 needs to be confirmed in future experimental studies, the proposed functional characteristics summarized in Table 1 should provide a starting point to elucidate adaptive design of the arthropod “microbrain”.

Table 1

Some possible correlates of the microbrain and the megalobrain


On account of distinct features of the arthropod “micro-brain” from the mammalian “megalobrain” as shown in Table 1, it is becoming evident that many features of behavior and the brain are common among arthropods and mammals. For example, recent studies have shown that many short-lived arthropods exhibit excellent learning capability, especially in the context of foraging behavior, as is discussed in the next section. Mayr (1974) argued that social or communicative behavior (e.g., mating) of animals tends to be relatively inflexible so as to insure unambiguous communication, while resource-directed behavior (e.g., foraging) is more likely to be flexible because the environment is variable. This argument explains why even relatively short-lived animals exhibit well-developed learning capability in the context of resource-directed behavior.

Olfactory discrimination and learning

Most previous studies on olfactory learning and memory in arthropods were carried out on only two species, i.e., the honey bee (Menzel, 1999; Giurfa, 2003) and the fruitfly Drosophila (Dubnau and Tully, 2001; Heisenberg, 2003; Isabel et al., 2004). Mizunami and colleagues examined whether cockroaches and crickets, known as excellent materials for electrophysiological studies of brain neurons (Staudacher and Schildberger, 1998; Mizunami et al., 1998b; Okada et al., 1999; Nishino et al., 2003), are suitable for the study of olfactory learning and memory.

Olfactory learning capabilities of the cockroach Periplaneta americana and the cricket Gryllus bimaculatus were demonstrated in studies using operant (Matsumoto and Mizunami, 2000; Sakura and Mizunami, 2001) and classical (Matsumoto and Mizunami, 2002a; Watanabe et al., 2003) conditioning paradigms. In both species, the preference for the odor associated with reward significantly increased after only one training session in both classical and operant conditioning paradigms. Memory formed by three training sessions lasted at least four weeks after training. In cockroaches, the classical conditioning procedure was effective even when they were harnessed in plastic tubes. Thus, this procedure may be useful for future electro-physiological studies of activity changes in brain neurons associated with olfactory learning.

By using an operant conditioning paradigm, Sakura et al. (2002) studied the capability of the cockroach to discriminate (1) three normal aliphatic alcohols with different numbers of carbon (1-pentanol, 1-hexanol and 1-octanol), (2) three C6 aliphatic alcohols (1-hexanol, 2-hexanol and trans-2-hexen-1-ol), (3) binary mixtures of two of these three alcohols and their components and (4) 1-hexanol solution of three different concentrations. In all of these tests, cockroaches exhibited higher preferences for the odors associated with reward. It was concluded that cockroaches can discriminate odors of structurally similar alcohols nearly as well as humans, even though their olfactory system consists of a much smaller number of neurons than that in humans.

Matsumoto and Mizunami (2002a) studied temporal determinants of olfactory long-term memory retention in the cricket using a differential conditioning procedure in which one odor was associated with reward and another odor was associated with punishment. It was shown that (1) four conditioning trials are sufficient to lead to a saturated level of acquisition, (2) conditioning is successful when the conditioned stimulus is presented immediately or at 5 sec before the onset of presentation of the unconditioned stimulus and (3) the optimal interval between trials are 2–5 min. The formation of long-term olfactory memory (>5 hr) in crickets required de novo protein synthesis, while the formation of the short-term memory (<4 hr) did not (Matsumoto et al., 2003).

Matsumoto and Mizunami (2002b) trained third- or fourth-instar nymph crickets to associate one odor with water and another odor with saline solution. At six and ten weeks after training, adult crickets exhibited significantly greater preferences for the odor associated with water over that associated with saline solution. It was therefore concluded that crickets are capable of retaining olfactory memory for practically their entire lifetime.

Matsumoto and Mizunami (2004) found that cricket had a capability for context-dependent olfactory learning. One group of crickets was trained to associate one of a pair of odors (conditioned stimulus, CS1) with water reward (appetitive unconditioned stimulus, US+) and another odor (CS2) with saline solution (aversive US, US-) under illumination and to associate CS1 with US- and CS2 with US+ in the dark (Fig. 1). Another group of crickets received training of the opposite stimulus arrangement. One day after the training for three days, the former group significantly preferred CS1 over CS2 under illumination but preferred CS2 over CS1 in the dark, and the latter group exhibited the opposite odor preference.

Fig. 1

Simplified diagrams of the experimental setup for context-dependent olfactory conditioning in crickets. Crickets were subjected to training or testing in the light or dark conditions. For training (A and B), each cricket was presented with peppermint or vanilla odor before the presentation of water (reward) or saline (non-reward) in a beaker. For the odor preference test (C and D), each cricket was allowed to freely choose between peppermint and vanilla sources in a box. For training or testing under illumination (A and C), a white fluorescent lamp on the ceiling was turned on. Group A crickets received training to associate peppermint odor with water and vanilla odor with saline solution (P+ and V–conditioning trials) under illumination and to associate vanilla odor with water and peppermint odor with saline solution (P–and V+ conditioning trials) in the dark. Group B crickets received training of the opposite stimulus arrangement. Modified from Matsumoto and Mizunami (2004).


In honey bees, Ichikawa and Sasaki (2003) demonstrated the importance of social stimuli for the development of learning capability. Newly emerged and very young worker bees could not associate a given odor (CS) with a sucrose reward (US). When workers are reared individually in a confined condition and deprived of colony odor and other social stimuli, they do not develop the ability even after 9 days after emergence. The important period for acquiring the learning ability was found to be from day 2 to day 6 after emergence.

Information processing in the antennal lobe

In insects, olfactory information is relayed from olfactory receptor neurons in the antennae to the antennal lobe (AL) and then to two areas of the protocerebral neuropils, i.e., the lateral protocerebrum (the lateral horn) and the mushroom body (MB) (Hildebrand and Shepherd, 1997). In the AL, the receptor neurons make synaptic connections with postsynaptic neurons in specialized compartments called glomeruli. In many species, the AL of the male comprises a number of ordinary glomeruli and a sexually dimorphic glomerulus called the macroglomerular complex (MGC), the latter receiving axon terminals of receptor neurons of sex-pheromones.

Kanzaki and colleagues revealed some aspects of functional organization of the antennal lobe of the silkmoth. The MGC of the male silkmoth consists of three subdivisions: the cumulus, toroid and horseshoe. Intracellular recording and staining revealed that responses of MGC projection neurons to pheromonal stimulation correlate with their dendritic arborizations in the subdivisions of the MGC and that each subdivision-specific projection neuron transmits information to different regions in the calyces of the MB and the inferior lateral protocerebrum (Kanzaki et al., 2003). Major pheromone component information is transferred to the medial part of the inferior lateral protocerebrum. Thus, this part may be an important processing site for major pheromonal information.

Hill et al. (2002) found a serotonin-immunoreactive neuron in the brain of the male silkmoth Bombyx mori that branched in many important brain neuropils such as the MB, central body, lateral accessory lobe and AL (Fig. 2). The neuron responded to mechanosensory stimuli to the antennae. In order to examine the possible modulatory effects of serotonin in the AL, Hill et al. (2003) utilized optical imaging with a voltage-sensitive dye combined with bath application of serotonin. Serotonin caused increases in the optical responses in both the MGC and the ordinary glomeruli evoked by electrical stimulation of the antennal nerve. Serotonin exerted a significantly greater enhancing effect in the toroid glomerulus of the MGC than in the cumulus or horseshoe, and the effects of serotonin were also non-homogeneously distributed in the ordinary glomeruli.

Fig. 2

Confocal image of the brain of the silkmoth Bombyx mori (frontal view). (A) This neuron has its soma in the posterior portion of the lateral cell cluster of one AL. The primary neurite projects through the ipsilateral AL, where it has a few fine branches in the posterior coarse neuropil. Branches are present in the ipsilateral lateral accessory lobe (LAL) and in both ipsi- and contralateral parts of the protocerebrum (PC), including the calyces (Ca) of both MBs, and also in the central body (CB). More extensive branching is seen in the contralateral superior PC. In the contralateral AL, branches are present in every ordinary glomeruli (Gs) as well as in each compartment of the macroglomerular complex (MGC). Scale bar: 100 μm. D: dorsal, V: ventral. (B) Schematic diagram of the brain regions innervated by this neuron. Oe: oesophagus, OL: optic lobe, SOG: suboesophageal ganglion, SPC: superior protocerebrum. Modified from Hill et al. (2002).


Ai and Kanzaki (2004) studied the functional synaptic organization of the MGC of the male silkmoth by optical recording with a voltage-sensitive dye. Sensory neurons in the antennae send their axons down either the medial nerve or lateral nerve, depending on whether they are located on the medial or lateral flagella. Pheromone-sensitive fibers in the medial nerve are biased towards the medial MGC, and those in the lateral nerve are biased towards the lateral MGC in the AL. When the medial nerve was stimulated, the amplitudes of the excitatory postsynaptic activities in the central toroid, the medial toroid and the medial cumulus were higher than those in the other areas. In contrast, when the lateral nerve was stimulated, excitatory postsynaptic activities were evoked in areas in both the lateral toroid and the lateral cumulus. It was concluded that information from the two main branches of the antenna is segregated to different sub-regions of the MGC glomeruli.

Odor-induced oscillatory signals have been observed in olfactory neural networks of vertebrates and invertebrates and implicated in olfactory processing (Stopfer et al., 1997). Okada and Kanzaki (2001) examined the localization of oscillatory signals in the AL using an optical recording technique with a voltage-sensitive dye in the bumblebee, and they found that the oscillatory signals were usually localized to regions corresponding to a single or several glomerular structures.

In the fruitfly Drosophila melanogaster, mate choice during courtship depends on detection of sex pheromones, which are initially processed in the AL, but no sexually-dimorphic glomeruli have been found in the AL. Kondo et al. (2003) compared the brain anatomies of 37 species of Drosophilidae from the islands of the Hawaiian archipelago, uncovering a sexual dimorphism within the AL in which two out of the 51 identifiable glomeruli were markedly enlarged in males. A phylogeny indicated that the sexual dimorphism of the homologous glomeruli arose independently in two species groups of Hawaiian endemic Drosophilidae. The corresponding glomeruli in D. melanogaster were also found to be sexually dimorphic. The formation of glomeruli of male size is prevented by the ectopic expression of female-type transformer (tra) cDNA in males. A defined set of glomeruli in Drosophila may enlarge in response to sex-determination genetic signals, mutations of which may result in species differences in sexual dimorphism of the brain.

Much progress has been attained to identify olfactory receptor genes in Drosophila (Vosshall et al., 2000) and to characterize odor coding in the primary olfactory center (the antennal lobe) in Drosophila (Keller and Vosshall, 2003; Wang et al., 2003), honeybees (Sachse and Galizia, 2003), moths (Daly et al., 2004) and locusts (Stopfer et al., 2003). The results provide solid basis for future studies of odor processing at a more central level.

Structure, development and gene expression in the mushroom body

MBs are higher olfactory and multisensory centers of the arthropod brain (Strausfeld et al., 1998) and implicated in odor discrimination (Laurent, 2002) and in olfactory and other forms of associative learning (Erber et al., 1980; Davis, 1996; Mizunami et al., 1998c; Giurfa, 2003; Heisenberg, 2003).

Yasuyama et al. (2002) investigated the synaptic circuits of the calyx (a major input neuropil) of the MB in the fluitfly using immuno-electron microscopy. The calyx neuropil contained (1) extrinsic neurons with immunoreactivity to choline acetyltransferase that correspond to projection neurons from the AL, (2) extrinsic neurons with GABA-like immunoreactivity that arise from the protocerebrum, and (3) intrinsic neurons of the MB called Kenyon cells. It was shown that Kenyon cells may receive major excitatory input from cholinergic boutons and considerable postsynaptic inhibition from GABAergic terminals as well as, more rarely, presynaptic inhibitory signaling.

The pedunculus and lobes of the cockroach MB consist of 15 repetitive modular subunits, each consisting of a light slab and a dark slab (Mizunami et al. 1997, 1998). Iwasaki et al. (1999) compared the ultrastructures of the light and the dark slabs in the pedunculus of the cockroach. Quantitative analyses of transversely sectioned Kenyon cell axons showed that average diameters of Kenyon cell axons in the light slabs were 5–18% larger than those in adjacent dark slabs, the difference being statistically significant. The results suggest that diameter of Kenyon cell axons is one of major factors contributing to the varied density of slabs. Varied conduction velocity due to different axon diameter may be utilized for signal processing.

In mammals, the Pax6 gene is expressed in various regions of the forebrain and is important for olfactory memory processing and emotional learning (Stoykova and Gruss, 1994). Kurusu et al. (2000) showed that Pax6 homologs eyeless (ey), twin of eyeless and dachshund (dac) are expressed in the developing MBs of the fruitfly Drosophila. Mutations of ey completely disrupted the MB neuropils, and a null mutation of dac resulted in marked disruption and aberrant axonal projections. Genetic analyses demonstrated that, whereas ey and dac synergistically control the structural development of the MBs, the two genes are regulated independently in the course of MB development. These findings suggest conserved roles of Pax6 homologs in the genetic programs of the olfactory learning centers.

Kurusu et al. (2002) also studied the embryonic development and post-embryonic development of the fruitfly MBs. In the mid to late embryonic stages, the pioneer MB tracts extend along Fasciclin II (FAS II)-expressing cells to form the primordia for the peduncle and the medial lobe. As development proceeds, the axonal projections of the larval MBs are organized in layers surrounding a characteristic core, which harbors bundles of actin filaments (Fig. 3). Mosaic analyses revealed sequential generation of the MB layers, in which newly produced Kenyon cells project into the core to shift to more distal layers as they undergo further differentiation. Whereas the initial extension of the embryonic MB tracts is intact, loss-of-function mutations of fas II cause abnormal formation of the larval lobes. Mosaic studies demonstrated that FAS II is intrinsically required for the formation of the coherent organization of the internal MB fascicles.

Fig. 3

(A) The MB primordia in the embryonic brain. Frontal view of the embryonic brain at late stage 16. (B,C) Layer organization of the second and third instar larval MBs. Dorsal images of Kenyon cell clusters and cross sections of the peduncle and lobes. Corresponding subdivisions are shown in the same lightness. Relative expression levels of various MB markers are summarized in the table. The second instar MBs can be concentrically subdivided in two layers surrounding the core. With increase in the numbers of Kenyon cells and their projections, the third instar MBs can be subdivided into three layers surrounding the core. Note that the distoproximal concentric subdivisions of each of the four Kenyon cell clusters topologically correspond to the unified concentric subdivisions in the lobes and the peduncle. The core consists of a bundle of newly formed axon fibers that contain densely packed actin filaments. AF, actin filaments; ACT, antennocerebral tract; DCT, dorsal commissural tract; DNC, dnc-lacZ; KCs, Kenyon cells; LPT, lateral protocerebral tract; MPT, medial protocerebral tract; P1, P4l, P4m, and D/T, fiber tract founder clusters; Ped, peduncle; VCT, ventral commissural tract. Modified from Kurusu et al. (2002).


A detailed comparison of the development of the MB of cockroaches (Farris and Strausfeld, 2001), Drosophila (Strausfeld, et al., 2003), and honeybees (Farris et al., 2004) revealed unexpected similarities among species, including the emergence of class II (clawed) Kenyon cells in an early stage of development.

Several genes that are expressed in the MB of the honey bee have been characterized by Kubo and colleagues. One such gene is Mblk-1, which is expressed preferentially in large Kenyon cells of the MB, encodes a putative transcription factor with two DNA-binding motifs, and might function downstream of the Ras/MAPK pathway (Takeuchi et al., 2001; Park et al., 2002, 2003).

Fujiyuki et al. (2004) identified a candidate gene involved in the aggressive behavior of worker honey bees. They used the differential display method to search for RNAs exclusively detected in the brains of aggressive workers that had attacked a hornet, and they identified a novel RNA, termed Kakugo RNA. It is a plus-strand RNA of a novel picorna-like virus, and the brains of aggressive workers were found to be infected by this novel virus. Kakugo RNA was detected in aggressive workers but not in nurse bees or foragers. In aggressive workers, Kakugo RNA was detected in the brain but not in the thorax or abdomen, suggesting a close relation between viral infection in the brain and aggressive worker behavior.

In order to explore how olfactory information is processed in the brain and finally converted into behavior, it is useful to establish an isolated whole brain preparation that includes both olfactory organs and motor output. This was accomplished using the terrestrial mollusk Limax by Inoue et al. (2004). They identified an in vitro index of odor-evoked behavior in the Limax and determined that shortening of the mantle muscles is one of the withdrawal responses selectively induced by aversive odors and that the shortening is mediated by a pair of parietal nerves. They also identified a motoneuron that innervates the mantle muscles. When they applied various odors to the nose in these isolated molluscan brains, only aversive odors induced discharges in this motoneuron. These results indicate that the discharges of the motoneuron can serve as an in vitro index of odor-induced aversive behavior.

Possible common origin of arthropod microbrain and vertebrate megalobrain

The molecular mechanisms of head development are a central question in vertebrate and invertebrate developmental biology. The anteriorly expressed homeobox gene otd in Drosophila and its homolog Otx in the mouse are required for early development of the most-anterior part of the body, suggesting that a fundamental genetic program of cephalic development might be conserved between vertebrates and invertebrates (Hirth et al., 1995). This hypothesis was examined by introducing human Otx genes into flies (Nagao et al., 1998). By inducing expression of human Otx homologs with a heat shock promoter, they found that both Otx1 and Otx2 functionally complement the cephalic defects of a fly otd mutant through specific activation and inactivation of downstream genes. The results are consistent with the view that a common molecular ground plan of cephalization was invented before the diversification of the protostome (that includes arthropods) and the deuterostome (that includes vertebrates) in the course of metazoan evolution.

Another way to address the question of whether the central nervous system (CNS) was derived from a common ancestor of deuterostomes and protostomes is to identify genes in the CNS of the putative common ancestor and to trace the evolutionary divergence of genes expressed in the CNS. Mineta et al. (2003) took a comparative approach using different species, focusing on one of the lower bilateral animals, the planarian, which is known to possess a CNS. They determined the nucleotide sequence of ESTs from the head portion of planarians, and, as a result of homology searches, they found that 116 clones had significant similarity to known genes related to the nervous system. They compared these 116 planarian EST clones with all open reading frames of the complete genome sequences of the human, fruitfly, and nematode and showed that >95% of these 116 nervous system-related genes, including genes involved in brain or neural morphogenesis, were commonly shared among these organisms, thus providing evidence at the molecular level for the existence of a common ancestral CNS (Fig. 4). This implies that the origin of nervous system-related genes greatly predated the emergence of the nervous system.

Fig. 4

A possible scenario of the process of evolution of the CNS. The phyla not analyzed are indicated in parenthesis. Possible events during evolution of the NS-related genes are noted. Modified from Mineta et al. (2003).


Basic organization of the arthropod microbrain

Elucidation of the basic organization of sensorimotor pathways in the arthropod brain was hampered by the lack of precise knowledge about the brain areas from which descending neurons originate to supply thoracic locomotory centers. Okada et al. (2003) examined the distribution of dendrites of descending neurons in the cockroach brain by retrogradely filling their axons from the cervical connective. The dendrites of descending neurons were found to be distributed in most brain areas, including lateral and medial protocerebra, which are major termination areas of output neurons of the MB, but not in the optic lobe, the AL, the MB, the central complex, or the postero-ventral part of the lateral horn. The last area is the termination area of major types of olfactory projection neurons from the AL, i.e., uni- and macroglomerular projection neurons, and, thus, these neurons have no direct connections with descending neurons. Based on these findings, Okada et al. have proposed that there are numerous parallel processing streams from cephalic sensory areas to thoracic locomotory centers, many of which are via premotor brain areas from which descending neurons originate. In addition, output from the MB, central complex and postero-ventral part of the lateral horn converge on some of the premotor areas presumably to modulate the activity of some sensori-motor pathways (Fig. 5). They proposed that many parallel processing streams function in various forms of reflexive and relatively stereotyped behaviors, while indirect pathways govern some forms of experience-dependent modification of behavior. Notably, this organization has fundamental similarities to basic organizations of the mammalian brain: Parallel and hierarchical features of sensorimotor pathways of the brains of vertebrates, including mammals, have been well documented (Romer and Persons, 1977; Brooks, 1986) and it has been suggested that the brains of mammals comprise amalgamating neural systems for reflexive behavior (brainstem), those for instinctive or emotional behavior (limbic system) and those for higher functions (neocortex) (MacLean, 1990). The similarities in the basic organization of the arthropod “microbrain” and mammalian “megalobrain” may reflect their common origin, if the suggestion from molecular and genetic studies that the common ancestors of vertebrates and invertebrates had more or less organized head ganglia is correct.

Fig. 5

A simplified diagram of neural pathways in the cockroach brain. Some neural pathways that have not yet been found in cockroaches but have been well established in other species of insects are also shown, by dashed lines, in order to better represent current knowledge on neural organizations of the insect brain. For simplification, descending pathways from the brain to the subesophageal ganglion (sog) and ascending pathways from the thoracic ganglia (tg) to the brain are not shown. Sensorimotor pathways in the sog and the tg, including proprioceptive sensory pathways, are also not shown. The l pr represents all of the l pr area except for the postero-ventral part of the lateral horn (p-v l ho). Brain neuropils are characterized as sensory (s), association (a), premotor (p) and motor (m) areas. The numbers indicate references described in Okada et al., 2003. They are mostly confined to study using intracellular recording and staining techniques; dendritic and termination areas of the neurons are estimated from their morphology and from their neural activities. Modified from Okada et al. (2003).


Other findings

Simple nervous systems of arthropods or other invertebrates are useful materials for the study of cellular and molecular mechanisms of basic brain functions. In studies using the fruitfly Drosophila, a sugar receptor gene was identified (Ueno et al., 2001), receptor neurons of bitter tastes were characterized (Meunier et al., 2003) and molecular mechanisms of synaptic growth during development were examined (Awasaki et al., 2000; Ritzenthaler et al., 2000; Sone et al., 2000). In addition, in studies using the nematode Caenorhabditis elegans, the involvement of Ras/MAPK signaling in olfaction was demonstrated (Hirotsu et al., 2000), and a secretory protein that regulates sensory integration and learning was identified (Ishihara et al., 2002). Other notable findings from our research project include the first behavioral demonstration of color vision by spiders (Nakamura and Yamashita, 2000) and identification of a cerebral neuron that contributes to synaptic modulation during taste aversion learning in Aplysia (Narusuye and Nagahama, 2002).

Future perspectives

The finding that the basic organization of the arthropod “microbrain” has some fundamental similarities to that of the mammalian “megalobrain” suggests that the arthropod brain is a useful material for studying basic neural mechanisms of brain functions common among many phyla of animals, not only at the molecular and cellular levels but also at the level of neural circuits. Arthropod ‘microbrain’ consists of relatively small number of neurons and is advantageous for clarifying neural mechanisms of brain functions in terms of singly identifiable neurons.

Some species of arthropods exhibit unusually sophisticated behaviors and provide opportunities to elucidate neural basis of some of the most sophisticated functions of animal brains evolved through natural selection. Some of fascinating behavioral capabilities of arthropods include: dance communication (by honey bee foragers), skylight compass orientation, navigation on the basis of visual landmark memory, context-dependent olfactory learning and sophisticated communication by means of various kinds of pheromones. The accumulation of information on basic structure and functions of the arthropod brain as well as newly emerging experimental techniques such as optical recording and RNA interference paves the way for proceeding into the study of elaborated brain functions of arthropods.


This work was supported by a Grant-in-Aid for Scientific Research on Priority Areas (#383) from the Ministry of Education, Science, Culture, Sports and Technology of Japan.



H Ai and R Kanzaki . 2004. Modular organization of the silkmoth antennal lobe macroglomerular complex revealed by voltage-sensitive dye imaging. J Exp Biol 207:633–644. Google Scholar


T Awasaki, M Saitoh, M Sone, E Suzuki, R Sakai, K Itoh, Y–I Nabeshima, and C Hama . 2000. The Drosophila TRIO plays an essential role for patterning of axons by regulating their directional extension. Neuron 26:119–132. Google Scholar


VB Brooks 1986. The neural basis of motor control. Oxford University Press. New York. Google Scholar


RC Brusca and GJ Brusca . 1990. Invertebrates. Sinauer. Sunderland. Google Scholar


KC Daly, TA Christensen, H Lei, B Smith, and JG Hildebrand . 2004. Learning modulates the ensamble representations for odors in primary olfactory networks. Proc Natl Acad Sci USA 101:10476–10481. Google Scholar


RL Davis 1996. Physiology and biochemistry of Drosophila learning mutants. Physiol Rev 76:299–317. Google Scholar


J Dubnau and T Tully . 2001. Functional anatomy: From molecule to memory. Curr Biol 11:R240–243. Google Scholar


J Erber, T Masuhr, and R Menzel . 1980. Localization of short-term memory in the brain of the bee, Apis mellifera. Physiol Entomol 5:343–358. Google Scholar


SM Farris and NJ Strausfeld . 2001. Development of laminar organization in the mushroom bodies of the cockroach: Kenyon cell proliferation, outgrowth, and maturation. J Comp Neurol 439:331–351. Google Scholar


SM Farris, AI Abrams, and NJ Strausfeld . 2004. Development and morphology of class II Kenyon cells in the mushroom bodies of the honey bee, Apis mellifera. J Comp Neurol 474:325–339. Google Scholar


T Fujiyuki, H Takeuchi, M Ono, S Ohka, T Sasaki, A Nomoto, and T Kubo . 2004. Novel insect picorna-like virus identified in the brains of aggressive worker honeybees. J Virol 78:1093–1100. Google Scholar


M Giurfa 2003. Cognitive neuroethology: dissecting non-elemental learning in a honeybee brain. Curr Opin Neurobiol 13:1–10. Google Scholar


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


JG Hildebrand and GM Shepherd . 1997. Mechanisms of olfactory discrimination: converging evidence for common principles across phyla. Ann Rev Neurosci 20:595–631. Google Scholar


ES Hill, Iwano, L Gatellier, and R Kanzaki . 2002. Morphology and physiology of the serotonine-immunoreactive putative antennal lobe feedback neuron in the male silkmoth Bombyx mori. Chem Senses 27:475–483. Google Scholar


ES Hill, K Okada, and R Kanzaki . 2003. Visualization of modulatory effects of serotonin in the silkmoth antennal lobe. J Exp Biol 206:345–352. Google Scholar


T Hirotsu, S Saeki, M Yamamoto, and Y Iino . 2000. The Ras-MAPK pathway is important for olfaction in C. elegans. Nature 404:289–293. Google Scholar


F Hirth, S Therianos, T Loop, W Gehring, H Reichert, and K FurukubuTokunaga . 1995. Developmental defects in brain segmentation caused by mutations of the homeobox gene orthodenticle and empy spiracles in Drosophila. Neuron 15:769–778. Google Scholar


N Ichikawa and M Sasaki . 2003. Importance of social stimuli for the development of learning capability in honeybees. Appl Entomol Zool 38:203–209. Google Scholar


T Inoue, Y Inokuma, S Watanabe, and Y Kirino . 2004. In vitro study of odor-evoked behavior in a terrestrial mollusk. J Neurophysiol 91:372–381. Google Scholar


G Isabel, A Pascual, and T Preat . 2004. Exclusive consolidated memory phases in Drosophila. Science 304:1024–1027. Google Scholar


T Ishihara, Y Iino, A Mohri, I Mori, K Gengyo-Ando, S Mitani, and I Katsura . 2002. HEN-1, a secretory protein with an LDL receptor motif, regulates sensory integration and learning in Caenorhabditis elegans. Cell 109:639–649. Google Scholar


M Iwasaki, M Mizunami, M Nishikawa, T Itoh, and Y Tominaga . 1999. Ultrastructural analysis of modular subunits in the mushroom bodies of the cockroach. J Electron Microsc 48:55–62. Google Scholar


R Kanzaki, K Soo, Y Seki, and S Wada . 2003. Projections to higher olfactory centers from subdivisions of the antennal lobe macro-glomerular complex of the male silkmoth. Chem Senses 28:113–130. Google Scholar


A Keller and LB Vosshall . 2003. Decoding olfaction in Drosophila. Curr Opin Neurbiol 13:103–110. Google Scholar


K Kirshfeld 1976. The resolution of lens and compound eye. In “Neural principles in vision”. Ed by Zettler F and Weiler R . Springer. Berlin. pp. 354–370. Google Scholar


Y Kondoh, KY Kaneshiro, K Kimura, and D Yamamoto . 2003. Evolution of sexual dimorphism in the olfactory brain of Hawaiian Drosophila. Proc R Soc Lond B Biol Sci 270:1005–1013. Google Scholar


M Kurusu, T Nagao, U Walldorf, S Flister, WJ Gehring, and K FurukuboTokunaga . 2000. Genetic control of the development of the mushroom bodies, the associative learning centers in the Drosophila brain, by the eyeless, twin of eyeless and dachshund genes. Proc Natl Acad Sci USA 97:2140–2144. Google Scholar


M Kurusu, T Awasaki, LM Masuda-Nakagawa, H Kawauchi, K Ito, and K Furukubo-Tokunaga . 2002. Embryonic and larval development of the Drosophila mushroom bodies: concentric layer subdivisions and the role of fasciclin II. Development 129:409–419. Google Scholar


G Laurent 2002. Olfactory network dynamics and the coding of multidimensional signals. Nature Neurosci 3:884–895. Google Scholar


P MacLean 1990. Triune brain in evolution. Plenum. New York. Google Scholar


Y Matsumoto and M Mizunami . 2000. Olfactory learning in the cricket Gryllus bimaculatus. J Exp Biol 203:2581–2588. Google Scholar


Y Matsumoto and M Mizunami . 2002a. Temporal determinants of olfac-tory long-term retention in the cricket Gryllus bimaculatus. J Exp Biol 205:1429–1437. Google Scholar


Y Matsumoto and M Mizunami . 2002b. Lifetime olfactory memory in the cricket Gryllus bimaculatus. J Comp Physiol A 188:295–299. Google Scholar


Y Matsumoto, S Noji, and M Mizunami . 2003. Time course of protein synthesis-dependent phase of olfactory memory in the cricket Gryllus bimaculatus. Zool Sci 20:409–416. Google Scholar


Y Matsumoto and M Mizunami . 2004. Context-dependent olfactory learning in an insect. Learn Mem 11:288–293. Google Scholar


E Mayr 1974. Behavior programs and evolutionary strategies. Am Sci 62:650–659. Google Scholar


R Menzel 1999. Memory dynamics in the honeybee. J Comp Physiol A 185:323–340. Google Scholar


K Mineta, M Nakazawa, F Cebria, K Ikeo, K Agata, and T Gojobori . 2003. Origin and evolutionary process of the CNS elucidated by comparative genomics analysis of planarian ESTs. Proc Natl Acad Sci USA 100:7666–7671. Google Scholar


N Meunier, F Marion-Poll, JP Rospars, and T Tanimura . 2003. Peripheral coding of bitter taste in Drosophila. J Neurobiol 56:139–152. Google Scholar


M Mizunami, M Iwasaki, M Nishikawa, and R Okada . 1997. Modular structures in the mushroom body of the cockroach. Neurosci Lett 229:153–156. Google Scholar


M Mizunami, M Iwasaki, R Okada, and M Nishikawa . 1998a. Topography of modular subunits in the mushroom bodies of the cock-roach. J Comp Neurol 399:153–161. Google Scholar


M Mizunami, R Okada, Y Li, and NJ Strausfeld . 1998b. Mushroom bodies of the cockroach: The activity and identities of neurons recorded in freely moving animals. J Comp Neurol 402:501–519. Google Scholar


M Mizunami, JM Weibrecht, and NJ Strausfeld . 1998c. Mushroom bodies of the cockroach: their participation in place memory. J Comp Neurol 402:520–537. Google Scholar


M Mizunami, F Yokohari, and M Takahata . 1999. Exploration into the adaptive design of the arthropod “microbrain.”. Zool Sci 16:703–709. Google Scholar


T Nagao, S Leuzinger, D Acampora, A Simeone, R Finkelstein, H Reichert, and K Furukubo-Tokunaga . 1998. Developmental rescue of Drosophila cephalic defects by the human Otx genes. Proc Natl Acad Sci USA 95:3737–3742. Google Scholar


T Nakamura and S Yamashita . 2000. Learning and discrimination of colored papers in jumping spiders (Araneae, Salticidae). J Comp Physiol A 186:897–901. Google Scholar


K Narusuye and T Nagahama . 2002. Cerebral CBM1 neuron contributes to synaptic modulation appearing during rejection of seaweed in Aplysia kurodai. J Neurophysiol 88:2778–2795. Google Scholar


H Nishino, S Yamashita, Y Yamazaki, M Nishikawa, F Yokohari, and M Mizunami . 2003. Projection neurons originating from thermoand hygrosensory glomeruli in the antennal lobe of the cockroach. J Comp Neurol 455:40–55. Google Scholar


K Okada and R Kanzaki . 2001. Localization of odor-induced oscillations in the bumblebee antennal lobe. Neurosci Lett 316:133–136. Google Scholar


R Okada, M Sakura, and M Mizunami . 2003. Distribution of dendrites of descending neurons and its implications for the basic organization of the cockroach brain. J Comp Neurol 458:158–174. Google Scholar


R Okada, J Ikeda, and M Mizunami . 1999. Sensory responses and movement-related activities in extrinsic neurons of the cockroach mushroom bodies. J Comp Physiol A 185:115–129. Google Scholar


J-M Park, T Kunieda, H Takeuchi, and T Kubo . 2002. DNA-binding properties of Mblk-1, a putative transcription factor from the honeybee. Biochem Biophys Res Commun 291:23–28. Google Scholar


J-M Park, T Kunieda, and T Kubo . 2003. The activity of Mblk-1, a mushroom body-selective transcription factor from the honeybee, is modulated by the ras/MAPK pathway. J Biol Chem 278:18689–18694. Google Scholar


ER Pianka 1970. On r and K selection. Am Nat 104:592–597. Google Scholar


S Ritzenthaler, E Suzuki, and A Chiba . 2000. Myopodia: Postsynaptic filopodia that interact with innervating motoneuron axons. Nature Neurosci 3:1012–1017. Google Scholar


AS Romer and TS Persons . 1977. The vertebrate body. W.B. Saunders. Philadelphia. Google Scholar


S Sachse and CG Galizia . 2003. The coding of odour-intensity in the honeybee antennal lobe: local computation optimizes odour representation. Eur J Neurosci 18:2119–2132. Google Scholar


M Sakura and M Mizunami . 2001. Olfactory learning and memory in the cockroach Periplaneta americana. Zool Sci 18:21–28. Google Scholar


M Sakura, R Okada, and M Mizunami . 2002. Olfactory discrimination of structurally similar alcohols by cockroaches. J Comp Physiol A 188:787–797. Google Scholar


M Sone, E Suzuki, M Hoshino, D Hou, HH Kuromi, M Fukuda, S Kuroda, K Kaibuchi, Y-I Nabeshima, and C Hama . 2000. Synaptic development is controlled in the periactive zones of Drosophila synapses. Development 127:4157–4168. Google Scholar


E Staudacher and K Schildberger . 1998. Gating of sensory responses of descending brain neurones during walking in crickets. J Exp Biol 201:559–572. Google Scholar


M Stopfer, S Bhagavan, BH Smith, and G Laurent . 1997. Impaired odour discrimination on desynchronization of odour-encoding neural assemblies. Nature 390:70–74. Google Scholar


M Stopfer, V Jayaraman, and G Laurent . 2003. Intensity versus identity coding in an olfactory system. Neuron 39:991–1004. Google Scholar


A Stoykova and P Gruss . 1994. Roles of Pax-genes in developing and adult brain as suggested by expression patterns. J Neurosci 14:1395–1412. Google Scholar


NJ Strausfeld, L Hansen, Y Li, RS Gomez, and K Ito . 1998. Evolution, discovery, and interpretation of arthropod mushroom bodies. Learn Mem 5:11–37. Google Scholar


NJ Strausfeld, I Sinakevitch, and I Vilinsky . 2003. The mushroom bodies of Drosophila melanogaster: an immunocytological and Golgi study of Kenyon cell organization in the calyces and lobes. Micros Res Tech 62:151–169. Google Scholar


H Takeuchi, E Kage, M Sawata, A Kamikouchi, K Ohashi, M Ohara, T Fujiyuki, T Kunieda, K Sekimizu, S Natori, and T Kubo . 2001. Identification of a novel gene, Mblk-1, that encodes a putative transcription factor expressed preferentially in the large-type Kenyon cells of the honeybee brain. Insect Mol Biol 10:487–494. Google Scholar


K Ueno, M Ohta, H Morita, Y Mikuni, S Nakajima, K Yamamoto, and K Isono . 2001. Trehalose sensitivity in Drosophila correlates with mutations in and expression of the gustatory receptor gene Gr5a. Curr Biol 11:1451–1455. Google Scholar


LB Vosshall, AM Wong, and R Axel . 2000. An olfactory sensory map in the fly brain. Cell 102:147–159. Google Scholar


JW Wang, AM Wong, J Flores, LB Vosshall, and R Alex . 2003. Two-photon calcium imaging reveals an odor-evoked map of activity in the fly brain. Cell 112:271–282. Google Scholar


H Watanabe, Y Kobayashi, M Sakura, Y Matsumoto, and M Mizunami . 2003. Classical olfactory conditioning in the cockroach Periplaneta americana. Zool Sci 20:1447–1454. Google Scholar


R Wehner 1987. ‘Matched filters’ - Neural models of the external world. J Comp Physiol 161:511–531. Google Scholar


K Yasuyama, IA Meinertzhagen, and F-W Schürmann . 2002. Synaptic organization of the mushroom body calyx in Drosophila melanogaster. J Comp Neurol 445:211–226. Google Scholar
Makoto Mizunami, Fumio Yokohari, and Masakazu Takahata "Further exploration into the adaptive design of the arthropod “microbrain”: I. Sensory and memory-processing systems," Zoological Science 21(12), 1141-1151, (1 December 2004).
Received: 9 August 2004; Published: 1 December 2004
antennal lobe
brain evolution
mushroom body
Back to Top