* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Topography of Modular Subunits in the Mushroom Bodies of the
Haemodynamic response wikipedia , lookup
Molecular neuroscience wikipedia , lookup
Metastability in the brain wikipedia , lookup
Apical dendrite wikipedia , lookup
Clinical neurochemistry wikipedia , lookup
Synaptic gating wikipedia , lookup
Subventricular zone wikipedia , lookup
Holonomic brain theory wikipedia , lookup
Stimulus (physiology) wikipedia , lookup
Nervous system network models wikipedia , lookup
Circumventricular organs wikipedia , lookup
Neuroanatomy of memory wikipedia , lookup
Development of the nervous system wikipedia , lookup
Neuropsychopharmacology wikipedia , lookup
Synaptogenesis wikipedia , lookup
Feature detection (nervous system) wikipedia , lookup
Optogenetics wikipedia , lookup
Axon guidance wikipedia , lookup
THE JOURNAL OF COMPARATIVE NEUROLOGY 399:153–161 (1998) Topography of Modular Subunits in the Mushroom Bodies of the Cockroach MAKOTO MIZUNAMI,1* MASAYUKI IWASAKI,2 RYUICHI OKADA,1 AND MICHIKO NISHIKAWA2 1Laboratory of Neuro-Cybernetics, Research Institute for Electronic Science, Hokkaido University, Sapporo 060–0812, Japan 2Faculty of Science, Fukuoka University, Fukuoka 814–0180, Japan ABSTRACT The mushroom body (MB), a conspicuous neuropil structure in the insect brain, is implicated in associative memory and in some aspects of motor control. Intrinsic neurons of the MB (Kenyon cells) extend dendrites into the calyx, and their axons run through the pedunculus and then bifurcate to form the a and the b lobes. At the pedunculus and the lobes, Kenyon cells make synaptic connections with dendrites of extrinsic (output) neurons. Previously, we reported that the a lobe of the cockroach MB consists of repetitive modular subunits (Mizunami et al. [1997] Neurosci. Lett. 229:153–156). Each subunit is composed of a dark layer and a light layer, and the layers are refereed to as slabs. Each slab is composed of axons of a specific subset of Kenyon cells. In the present study, we examined serial sections of reduced silver preparations and found that each dark and light slab continues throughout the length of the pedunculus and the a and b lobes. We also found that Golgi-impregnated Kenyon cells often exhibit a characteristic grouping, forming a thin sheet interlaced by dozens or hundreds of axons. The sheet is much thinner than the slab, and each sheet remains within a particular slab throughout the length of the pedunculus and the lobes. Thus, the sheet is a component forming the slab. In the pedunculus and the b lobe, a class of Golgi-impregnated extrinsic neurons exhibit segmented dendritelike arbors that interact with every other slab, i.e., either with only dark or light slabs. Because each neuron of this class interacts with each particular set of dark or light slabs, we conclude that the slabs are units for transmitting output signals from the MB. J. Comp. Neurol. 399:153–161, 1998. r 1998 Wiley-Liss, Inc. Indexing terms: insect; brain; neuroanatomy; associative memory; higher center The insect brain contains a pair of prominent neuropils referred to as mushroom bodies (MBs). In the honey bee, the establishment of olfactory memory is related to normal functioning of MBs during the consolidation period following olfactory associative learning (Erber et al., 1980). In the fly, mutants with chemical or structural deficits in MBs exhibit poor learning performance (Heisenberg et al., 1985; Nighorn et al., 1991; de Belle and Heisenberg, 1994; Davis, 1996). In the cockroach, surgical ablation of MBs resulted in poor performance in navigation based on visual place memory, and observations of the activity of its extrinsic neurons of the lobes while the cockroach is walking freely suggest that the MB is involved in the control, and possibly the planning, of locomotory actions (Mizunami et al., 1993). The MB neuropil has four distinct parts, i.e., calyces, pedunculus, and the a and b lobes (Fig. 1A). In the calyces, axon terminals of input neurons, the majority of which derive from the deutocerebrum (Weiss, 1974; Mobbs, 1982) and some from the protocerebrum (Nishikawa et al., 1998; r 1998 WILEY-LISS, INC. Nishino and Mizunami, 1998; Yamazaki et al., 1998), make synapses onto dendrites of intrinsic neurons, called Kenyon cells (Steiger, 1967; Schürmann, 1974; Strausfeld, 1976). Axons of Kenyon cells project anteroventrally to the pedunculus and then bifurcate to form two lobes: the dorsally projecting a lobe and the medially projecting b lobe (Fig. 1A). At the pedunculus and lobes, Kenyon cells make synapses onto dendrites of extrinsic (output) neurons (Schürmann, 1970; Frontali and Mancini, 1970). Axons of extrinsic neurons project into different areas of the brain (Mobbs, 1984; Rybak and Menzel, 1993), including lateral Grant sponsor: Naitoh Foundation; Grant sponsor: Narishige Zoological Science Award; Grant sponsor: Ministry of Education, Science, Culture and Sports of Japan. *Correspondence to: Dr. Makoto Mizunami, Laboratory of NeuroCybernetics, Research Institute for Electronic Science, Hokkaido University, Sapporo 060-0812, Japan. E-mail: [email protected] Received 29 December 1997; Revised 6 May 1998; Accepted 10 May 1998 154 M. MIZUNAMI ET AL. Fig. 1. A: Camera lucida reconstruction of a Golgi-impregnated Kenyon cell in serial frontal sections of a cockroach brain. The neuron gives off arborizations in the calyx (C), and its axon runs anteroventrally through the pedunculus (Ped) and then bifurcates; one branch runs dorsally through the a lobe and the other runs medially through the b lobe. B: Horizontal section of the a lobe in a reduced silverstained brain. Fifteen dark (D) and 15 light (L) slabs alternate in this individual. Dark and light slabs are numbered 1–15 from posterior to anterior (A). M, medial. Scale bars 5 200 µm in A, 20 µm in B. proto- and deutocerebri from which descending neurons originate to supply thoracic motor centers (Li and Strausfeld, 1997). Each MB of the cockroach Periplaneta americana contains 200,000 Kenyon cells (Neder, 1959), the largest number reported thus far for insect MBs: the reported number of Kenyon cells in each MB is 170,000 in the honey bee Apis (Witthöft, 1967), 50,000 in the cricket Acheta (Schürmann, 1973), 50,000 in the locust Schistocerca (Leitch and Laurent, 1996), 21,000 in the fly Musca (Strausfeld, 1976), and 2,500 in the fly Drosophila (Balling et al., 1987). Our recent study using reduced silver stains showed that 15 modular subunits, each consisting of a pair of dark and light slabs, are present in the a lobe of the cockroach (Mizunami et al., 1997). By comparing reduced silver and Golgi preparations, we concluded that each modular subunit is formed from the axons of a specific subset of Kenyon cells. It was unclear, however, if the modular organization is restricted to the a lobe or is also present in other output neuropils. In the present study, we report that these modular subunits are maintained throughout the length of the output neuropils, i.e., the pedunculus and the a and b lobes. In addition, smaller groupings of Kenyon cells, which we refer to as sheets, are also maintained throughout the length of the output neuropils. MATERIALS AND METHODS Studies were done on adult male and female cockroaches, Periplaneta americana, from a colony raised in Hokkaido University. The cockroaches were maintained at 25–27°C and at 12-hour light–dark cycles. For Golgi impregnations (Strausfeld, 1980), each cockroach was anesthetized by cooling with ice, and the head was removed and mounted, dorsal side up, on a small dish. A small piece of cuticle on the front of the head was removed to expose the brain. The head was immersed in 2.5% potassium dichromate for 1 hour, and the brain was dissected out. The isolated brain was then transferred to a bath containing five parts 2.5% potassium dichromate and one part 25% glutaraldehyde (Polysciences, Tokyo, Japan) for 4–7 days in the dark at 4°C. The tissues were subsequently treated by one of two methods: (1) the brains were immersed in 0.75% silver nitrate for 3–4 days at 4°C; or (2) the brains were washed in 2.5% potassium dichromate for 1 hour, followed by 4–7 days of incubation in 100 parts of 2.5% potassium dichromate and one part of 1% osmium tetroxide for 6–10 days at 4°C, and then the brains were immersed in 0.75% silver nitrate for 4–7 days at 4°C. The brains were dehydrated through a graded series of ethanol, immersed in propylene oxide, and embedded in soft Araldite (Nisshin EM, Tokyo, Japan). Serial sections were cut at 50–110 µm in the three major planes, frontal, MODULAR SUBUNITS IN THE COCKROACH MUSHROOM BODIES 155 horizontal and sagittal, with respect to the axis of the head. For reduced silver impregnation, the heads were isolated and mounted in a small dish, and the brains were exposed by removing a piece of cuticle. The heads were immersed in cockroach saline (Yamasaki and Narahashi, 1959) containing 3% paraformaldehyde for 1 hour, and then the brains were dissected out. The isolated brains were fixed in a solution containing 4% paraformaldehyde, 5% glacial acetic acid, and 85% ethanol for 2 days, dehydrated, and embedded in paraffin. The original (Bodian, 1936) and a variation (Otsuka, 1962) of the Bodian reduced silver impregnation method were used for 12-µm sections. For staining with toluidine blue/azure II, the brains were fixed in ethanol for 4 days and embedded in Araldite. Serial sections were cut at 12 µm and briefly stained with a mixture of 1% toluidine blue and 1% azure II. The specimens were rinsed, dehydrated with a graded series of ethanol, cleared in xylol, and coverslipped under Canada balsam. The preparations were observed under Nomarski interference contrast (Olympus, Tokyo, Japan), real-time three-dimensional (Edge Scientific Instruments, Santa Monica, CA), and conventional microscopes. A high magnification observation at 1,0003 was made by using an oil-immersion objective lens (1003). Trajectories of stained cells were traced by using a camera lucida. RESULTS Topography of modular subunits in the MB Observations of reduced silver preparations of the cockroach brain have shown that there are repetitive modular subunits, each consisting of a pair of dark and light slabs in the a lobe of the MB (Fig. 1B; Mizunami et al., 1997). More or less similar slablike striations have been described in some anatomical accounts of the cockroach MB (Hanström, 1928; Howse, 1975; Weiss, 1981; Li and Strausfeld, 1997), but their detailed morphological features have not been reported. Comparisons of reduced silver and Golgi preparations have shown that different slabs are formed by axons of different subsets of Kenyon cells (Mizunami et al., 1997). In the cross section of the a lobe, the slabs at the posterior region are arched, with the largest curvature at the posterior end, and those at the anterior region are straight (Fig. 1B). The number of modular subunits differs among individuals; the typical number seen in about 60% of individuals is 15 and the minimum and maximum numbers observed among 24 cockroaches are 14 and 19, respectively. The number of modular subunits in the right and left MBs, however, is the same for all individuals examined thus far. No prominent difference was found in the number of subunits between males and females. The MBs of juveniles exhibited fewer modular subunits: juveniles with a body length of 8–10 mm exhibited 9–12 subunits and those with a body length of 12–18 mm exhibited 13–14 subunits. This observation indicates an increase in the number of modular subunits during ontogenetic processes. To describe the modular structures quantitatively, the averaged width at the center (Fig. 2A) and the area (Fig. 2B) of each of the dark and light slab was measured in cross sections of the a lobe. We set a specific assumption to assess borders between neighboring slabs: the borders were set so that the total area of 15 dark slabs was equal to that of 15 light slabs. Measurements were made at each of three locations of the right and left a lobes in serial Fig. 2. The width (A) and the area (B) in the cross section of slabs, measured at the a lobe. Dark (D) and light (L) slabs are numbered 1–15 from posterior to anterior. Border lines between neighboring dark and light slabs are defined by assuming that the total area of 15 dark slabs is equal to that of 15 light slabs. The widths at the center and the areas were measured at three locations of the left and right a lobes in serial horizontal sections of a reduced silver-stained brain, and averages and standard deviations of six measurements are shown. horizontal sections of a reduced silver-stained brain. Dark (D) and light (L) slabs are numbered 1–15 from posterior to anterior. The width of slabs is smallest in the median part (3–4 µm) and larger in the anterior or posterior parts. The widths of the slabs at the posterior end (1L) and anterior end (15D) are approximately 15 µm and 20 µm, respectively. The area in the cross section is the smallest at the posterior end and largest at the anterior end: the area of each of the last four slabs at the anterior end (14L–15D) is about three times that of most posterior slabs (Fig. 2B). This observation suggests that anterior slabs contain a much larger number of axons of Kenyon cells. Figure 3 shows serial horizontal sections of the MB in a reduced silver-stained brain. Although no prominent slablike laminar structures are visible in the calyx (Fig. 3A), concentrically arranged dark and light slabs alternate at the base of each lateral and medial calyx (Fig. 3B). Tracing of individual Kenyon cell axons in Golgi preparations show that the concentric arrangement of dark and light slabs 156 Fig. 3. A-E: Serial horizontal sections of the mushroom body of a reduced silver-stained brain showing topography of modular subunits. A: In the calyx, no slablike striations are visible. B: Sets of alternating dark and light slabs are concentrically arranged at the base of medial and lateral calyces. Two sets of concentric slabs uncurl at the head of the pedunculus (P) and fuse edgewise to form flattened slabs as they M. MIZUNAMI ET AL. descend through the pedunculus (C). M and L are axons of Kenyon cells derived from medial calyx (MC) and lateral calyx (LC), respectively. Flattened slabs continue throughout the length of the a (A–D) and b (E) lobes. The anterior side is at the top; the lateral side is at the right. Scale bar 5 50 µm. MODULAR SUBUNITS IN THE COCKROACH MUSHROOM BODIES 157 Fig. 4. Two serial horizontal sections, one just dorsal to (A) and the other at (B) the junction between the pedunculus and lobes of a brain stained with toluidine-blue/azure II, showing that each dark and light slab of the pedunculus (P) continues to each dark and light slab of the a lobe. The anterior side is at the top. Scale bar 5 50 µm. reflects a concentric arrangement of axons of Kenyon cells in the calyx (Mizunami et al., 1998). Concentric slabs from each calyx uncurl at the head of the pedunculus and fuse edgewise to form single flattened slabs as they descend through the pedunculus. Arrays of flattened slabs are maintained throughout the pedunculus and a and b lobes. A similar pattern of striation is visible in plastic sections stained with toluidine blue/azure II (Fig. 4) or osmiumethyl gallate, in paraffin sections stained with hematoxylin–eosin, and by Nomarski interference contrast observation of unstained sections. We attempted to examine the cause of the different staining intensity between dark and light slabs by a high magnification observation (1,0003) of these preparations, but no convincing conclusions were attained, apparently due to the limitation of the resolution of light microscopy. Observations of serial horizontal sections showed that each dark or light slab of the pedunculus is continuous with a corresponding dark or light slab in the a lobe (Fig. 4) and the b lobe (not shown); thus, each dark or light slab keeps its dark or light appearance over the entire extent in the output neuropil. In Figure 5, the topography of modular subunits is schematically drawn on the basis of observations of serial sections. Each slab continues throughout the length of the pedunculus and a and b lobes, thereby indicating that each slab consists of a particular subset of Kenyon cell axons. Note that bifurcation of slabs reflects bifurcation of individual axons of Kenyon cells (Fig. 1A). Extrinsic neurons with segmented dendritic arborizations A class of Golgi-stained extrinsic (output) neurons of the b lobe have segmented dendritelike spiny arborizations, the interval and the width of segments corresponding to the width of slabs (Fig. 6A). The segments are maintained throughout the dendritic domain of the neuron. This Fig. 5. Schematic representation of the topography of modular subunits in the pedunculus (Ped) and the a and b lobes, based on observations of serial horizontal, sagittal, and frontal sections of reduced silver-stained brains. Bifurcation of each dark and light slab at the junction between the pedunculus and lobes reflects bifurcation of axons of Kenyon cells constituting the slabs. A, anterior; P, posterior; D, dorsal; V, ventral; M, medial; L, lateral. 158 M. MIZUNAMI ET AL. Fig. 6. Examples of Golgi-stained extrinsic neurons whose dendritelike arbors exhibit prominent segmentation, observed in cross sections of the b lobe (A) and the pedunculus (B). A is a sagittal section and B is a horizontal section. Axons of Kenyon cells (not shown) run perpendicular to the plane of the sections. A: Dendritic arbors of the extrinsic neuron exhibit at least nine segments, the width and intervals of which match those of slabs. B: Dendritic arbors of the neuron exhibit approximately 14 segments (arrowheads), each of which covers less than half of the width of each slab. The dorsal side is at the top in A, and the anterior side is at the top in B. Scale bars 5 30 µm. suggests that extrinsic neurons of this class receive synaptic inputs from Kenyon cells in every other slab, i.e., in only dark slabs or light slabs. Four of 17 Golgi-impregnated extrinsic neurons in the b lobe have segmented dendritic arborizations. We reported that about 35% of output neurons in the a lobe exhibit similar segmentation (Mizunami et al., 1997). In the pedunculus, three of 11 Golgiimpregnated extrinsic neurons possess segmented dendritic arbors. In all three neurons, the dendritic arborizations cover only less than half of the thickness of each slab, thus interacting with the axons of only less than half of Kenyon cells that constitute each slab (Fig. 6B). Different extrinsic neurons with segmented dendrites appear to interact with different sets of dark or light slabs: dendrites of some neurons cover up to 13–14 slabs, whereas those of other neurons cover only a few particular slabs. Thus, we suggest that the slabs represent units to transmit output signals from the MB. In Figure 8, axons of Kenyon cells consisting of a sheet are traced to the cell body region and the synaptic neuropil of the calyx in serial horizontal sections. Their cell bodies form at least three separate clusters in the anterolateral (arrow in Fig. 8A), anterior (arrowhead), and anteromedial (not shown) parts of the cell body region dorsal to the lateral calyx, and their dendritic arborizations occupy anterolateral, anterior, and anteromedial parts of the synaptic neuropil of the lateral calyx, respectively (Fig. 8A,B). Axons of each group form a bundle as they converge to the center of the calyx, and the three bundles run side by side at the base of the calyx (Fig. 8B, arrows). Another axonal bundle derived from the posteromedial part of the lateral calyx then joins these bundles (not shown); the cell bodies could not be located. These four axon bundles are flattened and fuse together to from a thin sheet at the head of the pedunculus (Fig. 8C). The sheetlike structure is maintained throughout the length of the a (Fig. 8C–E) and b (not shown) lobes. The lateral extent of the sheet covers half of the width of the pedunculus and covers the entire width of the a and b lobes (Fig. 8C–E), indicating that axons of lateral and medial calyces are mixed together as they descend to the lobes. A comparison with serial horizontal sections of reduced silver preparations suggests that the sheet is located in slab 9D or 10L on its entire extent in the pedunculus and the a and b lobes. Thus, we consider the sheets to be elements that constitute the slabs. Topography of sheets in the MB Axons of Golgi-impregnated Kenyon cells often exhibit a characteristic grouping, a thin mesh- or sheetlike structure, which we refer to as a sheet (Fig. 7A,B). The sheet in the a lobe appears as an array of axons when observed in a horizontal section (Fig. 7A) and as a meshwork interlaced by dozens or hundreds of Kenyon cell axons when observed in a frontal section (Fig. 7B). Figure 7B shows a sheet wider than 120 µm, indicating that the sheet is extremely flat and thin. Each axon of an individual Kenyon cell remains within a particular sheet for its entire extent in the pedunculus and the a and b lobes. Observations of sheets in cross sections of the a lobe show that the sheets are flat only in the anterior and medial parts (Fig. 7B) and that they are curved in the posterior part, the largest curvature being at the posterior end. Thus, the form of the sheets precisely agrees with that of slabs at that location, except that the sheets are much thinner than slabs. The number of Kenyon cell axons counted in each sheet differed from 40 to 280 (n 5 30), the large variance possibly due to the capricious nature of the Golgi stain. DISCUSSION We previously reported that repetitive modular subunits, each consisting of a pair of a dark and a light slab, are present in the a lobe of the MB of the cockroach (Mizunami et al., 1997). Each slab represents the axons of a particular subset of Kenyon cells. We now report that each slab continues throughout the length of the output neuropils. Some extrinsic neurons have a characteristic segmented dendrites: these segments have been shown to represent interactions with specific slabs. In addition, slabs are now shown to contain smaller elements, referred MODULAR SUBUNITS IN THE COCKROACH MUSHROOM BODIES 159 Fig. 7. A: Axons of Golgi-impregnated Kenyon cells observed in a horizontal section of the a lobe, near the junction between the a and b lobes. Axons of Kenyon cells are arranged in rows. B: Golgi-stained Kenyon cell axons at the a lobe, observed in a frontal section. A thin meshwork, or sheet, formed by a number of Kenyon axons can be seen. The anterior side is at the top in A, and the dorsal side is at the top in B. Scale bars 5 20 µm. to as sheets, that arise from smaller subsets of Kenyon cells. The cause of the different staining intensities between dark and light slabs could not be determined in the present study; this is the subject of an electron microscopic study (Iwasaki et al., unpublished observations). Topography of modular subunits in the MB At the base of the calyx, alternating dark and light slabs are arranged concentrically like growth rings of a tree, and this concentric arrangement is converted into a linear array of flattened slabs in the pedunculus, as was described by Weiss (1981). The concentric arrangement of slabs at the base of the calyx reflects concentric arrangements of Kenyon cell axons in the calyx, as is discussed in an accompanying paper (Mizunami et al., 1998). Each slab differs in shape, thickness, and area in the cross section. The area in the cross section of a slab at the anterior part is three times larger than that of a slab at the posterior end, suggesting that the anterior slabs consist of a larger number of Kenyon cells. In an accompanying paper (Mizunami et al., 1998), we describe how Kenyon cells can be classified into four types, each of which forms a respective set of modular subunits. We speculate that each of the modular subunits and slabs may have specific morphological features that reflect different functional roles. A class of extrinsic (output) neurons of the pedunculus and b lobe possesses segmented dendritelike arbors that interact with every other slab, i.e., with only dark or light slabs. Similar neurons have been observed in the a lobe (Mizunami et al., 1997), indicating that this class of extrinsic neurons is distributed throughout output neuropils. Different extrinsic neurons cover different sets of dark or light slabs, suggesting that the slabs are units for transmitting output signals from the MB. Extrinsic neurons with segmented dendritic arbors of the pedunculus cover less than half of the width of each slab. This coverage appears to correspond to substructures of the slabs (discussed below). Modular organizations of the MB may not be restricted to cockroaches. In the honey bee, Mobbs (1982) reported that there are seven layers in the pedunculus and lobes and that each layer represents a different morphological type of Kenyon cell originating from a different concentric zone of the calycal neuropil. Some of these layers exhibit Phe-Met-Arg-Phe-NH2 (FMRF amide)-like immunoreactivity (Schürmann and Erber, 1990) or immunoreactivity to type II cAMP-dependent protein kinase (Müller, 1997). Dendrites of extrinsic neurons of the a lobe of the honey bee also exhibit a prominent layered arrangement (Mobbs, 1982; Rybak and Menzel, 1993). In the fruit fly Drosophila, subdivision of the MB by several different types of Kenyon cell was found when patterns of gene expression were examined (Yang et al., 1995) and ablation of a specific type of Kenyon cell resulted in a defect in sex-specific courtship behavior, whereas that of other types did not (O’Dell et al., 1995). Weiss (1981) observed a number of layers in the pedunculus and lobes in the MB of several species of Orthoptera by reduced silver stainings. These modular structures may be a common design of insect MBs. Topography of sheets in the MB In Golgi preparations, the axons of Kenyon cells are often seen to take the form of a very thin sheet in the pedunculus and the a and b lobes. This arrangement suggests a nonrandom nature of our Golgi stains: the probability that axons constituting the same sheet are concurrently impregnated appears to be high in our Golgi procedure. Bundles of axons arising from several groups of Kenyon cells, which may occupy separate areas of the cell body region and may send dendritic processes to distinct regions of the calycal neuropil, flatten and fuse together at Fig. 8. A–F: Serial horizontal sections showing groups of Golgiimpregnated Kenyon cells whose axons form a sheet. Their cell bodies are located in anterolateral (arrowheads), anterior (arrowheads), and anteromedial (not shown) parts of the cell body region dorsal to the lateral calyx (LC; A), and their dendrites cover anterolateral, anterior, and anteromedial portions of the synaptic neuropil of the lateral calyx (A,B). The axons of these cell groups form a bundle (arrows in A,B) and are arranged side by side at the base of the calyx (B). As the axonal bundles descend to the head of the pedunculus (P), they are flattened and fuse together to form a thin sheet (arrows in C–E). The sheet continues throughout the length of the a (C–E) and b lobes (not shown). The anterior side is at the top; the medial side is at the right. Scale bar 5 50 µm. MODULAR SUBUNITS IN THE COCKROACH MUSHROOM BODIES the head of the pedunculus. The sheet is much thinner than the slab, and each sheet appears to remain in a particular slab for its entire extent in the output neuropils, suggesting that the sheet is a component forming the slab. It remains unclear as to whether all Kenyon cells are organized into sheets or some Kenyon cells form sheets but others run solitary without forming a sheet. If the former is the case, the sheet can be regarded as a smaller modular substructure of the MB. A variation in the periodic pattern of modules has been noted (Li and Strausfeld, 1997) and will be explained by a different organization of substructures (Iwasaki et al., unpublished observations). The functional significance of the sheetlike structure remains to be clarified. One possibility is that Kenyon cells constituting each sheet make synaptic connections with each other. Frontali and Mancini (1970) noted in their electron microscopic study that Kenyon cells make synapses with each other in the a lobe of the cockroach. In the locust, Kenyon cells exhibit synchronous and oscillatory activity when olfactory stimulation is applied to the antenna. Reciprocal synaptic interactions among Kenyon cells may play a role in producing this synchronized electrical activity (Laurent and Naraghi, 1994). It would be interesting to investigate whether each subset of Kenyon cells constituting each sheet exhibits similar synchronous activity in the cockroach. ACKNOWLEDGMENTS We thank Dr. N.J. Strausfeld for helpful discussion, especially for pointing out the possibility that slabs represent radial arrangements in the calyces. The advice helped us to interpret our results correctly. M. Ohara provided helpful comments. LITERATURE CITED Balling, A., G.M. Technau, and M. Heisenberg (1987) Are the structural changes in adult Drosophila mushroom bodies memory traces? Studies on biochemical learning mutants. J. Neurogenet. 4:65–73. Bodian, D. (1936) A new method for staining nerve fibres and nerve endings in mounted paraffin sections. Anat. Rec. 65:89–97. Davis, R.L. (1996) Physiology and biochemistry of Drosophila learning mutants. Physiol. Rev. 76:299–317. de Belle, J.S., and M. Heisenberg (1994) Associative odor learning in Drosophila abolished by chemical ablation of mushroom bodies. Science 263:692–695. Erber, J., T. Masuhr, and R. Menzel (1980) Localization of short-term memory in the brain of the bee, Apis mellifera. Physiol. Entomol. 5:343–358. Frontali, N. and G. Mancini (1970) Studies of the neuronal organization of cockroach corpora pedunculata. J. Insect Physiol. 16:2293–2301. Hanström, B. (1928) Vergleichende Anatomie des Nervensystems der wirbellosen Tiere unter Berücksichtigung seiner Funktion. Berlin: Springer. Heisenberg, M., A. Borst, S. Wagner, and D. Byers (1985) Drosophila mushroom body mutants are deficient in olfactory learning. J. Neurogen. 2:1–30. Howse, P.E. (1975) Brain structure and behavior in insects. Annu. Rev. Entomol. 20:359–379. Laurent, G. and M. Naraghi (1994) Odorant-induced oscillations in the mushroom bodies of the locust. J. Neurosci. 14:2993–3004. Leitch, B. and G. Laurent (1996) GABAergic synapses in the antennal lobe and mushroom body of the locust olfactory system. J. Comp. Neurol. 372:487–514. Li, Y. and N.J. Strausfeld (1997) Morphology and sensory modality of mushroom body extrinsic neurons in the brain of the cockroach, Periplaneta americana. J. Comp. Neurol. 387:631–650. 161 Mizunami, M., J.M. Weibrecht, and N.J. Strausfeld (1993) A new role for the insect mushroom bodies: Place memory and motor control. In R.D. Beer, R.E. Ritzmann, and T. McKenna (eds): Biological Neural Networks in Invertebrate Neuroethology and Robotics. Cambridge: Academic Press, pp. 199–225. Mizunami, M., M. Iwasaki, M. Nishikawa, and R. Okada (1997) Modular structures in the mushroom body of the cockroach. Neurosci. Lett. 229:153–156. Mizunami, M., M. Iwasaki, R. Okada, and M. Nishikawa (1998) Topography of four classes of Kenyon cells in the mushroom bodies of the cockroach. J. Comp. Neurol. 399:162-175. Mobbs, P.G. (1982) The brain of the honeybee Apis mellifera. I. The connections and spatial organization of the mushroom bodies. Phil. Trans. R. Soc. Lond. B 298:309–354. Mobbs, P.G. (1984) Neural networks in the mushroom bodies of the honeybee. J. Insect Physiol. 30:43–58. Müller, U. (1997) Neuronal cAMP-dependent protein kinase type II is concentrated in mushroom bodies of Drosophila melanogaster and the honeybee Apis mellifera. J. Neurobiol. 33:33–44. Neder, R. (1959) Allometrisches Wachstum von Hirnteilen bei drei verschieden grossen Schabenarten. Zool. Jahrb. Anat. 4:411–464. Nighorn, A., M.J. Healy, and R.L. Davis (1991) The cyclic AMP phosphodiesterase encoded by the Drosophila dunce gene is concentrated in the mushroom body neuropil. Neuron 6:455–467. Nishikawa, M., H. Nishino, M. Mizunami, and F. Yokohari (1998) Functionspecific distribution patterns of axon terminals of input neurons in the calyces of the mushroom body of the cockroach, Periplaneta americana. Neurosci. Lett. 245:33–36. Nishino, H. and M. Mizunami (1998) Giant input neurons of the mushroom body: intracellular recording and staining in the cockroach. Neurosci. Lett. 246:57–60. O’Dell, K.M.C., J.D. Armstrong, M.Y. Yang, and K. Kaiser (1995) Functional dissection of the Drosophila mushroom bodies by selective feminization of genetically defined subcompartments. Neuron 15:55–61. Otsuka, N. (1962) Histologisch-entowicklungsgeschichtliche Untersuchungen an Mauthnerschen Zellen von Fischen. Z. Zellforsh. Mikrosk. Anat. 58:33–50. Rybak, J. and R. Menzel (1993) Anatomy of the mushroom bodies in the honey bee brain: The neuronal connections of the alpha-lobe. J. Comp. Neurol. 334:444–465. Schürmann, F.W. (1970) Über die Struktur der Pilzkörper des Insektenhirns. I. Synapsen im Pedunculus. Z. Zellforsch. 103:365–381. Schürmann, F.W. (1973) Über die Struktur der Pilzkörper des Insektenhirns. III. Die Anatomie der Nervenfasern in den Corpora pedunculata bei Acheta domesticus L. (Orthoptera): eine Golgi-Studie. Z. Zellforsch. 145:247–285. Schürmann, F.W. (1974) Bemerkungen zur Funktion der Corpora pedunculata im Gehirn der Insekten aus morphologischer Sicht. Exp. Brain Res. 19:406–432. Schürmann, F.W. and J. Erber (1990) FMRFamide-like immunoreactivity in the brain of the honeybee (Apis mellifera). A light- and electron microscopical study. Neuroscience 38:797–807. Steiger U. (1967) Über den Feinbau des Neuropils im Corpus pedunculatum der Waldameise. Z. Zellforsch. 81:511–536. Strausfeld, N.J. (1976) Atlas of an Insect Brain. Berlin: Springer. Strausfeld, N.J. (1980) The Golgi method. In N.J. Strausfeld and T.A. Miller (eds): Neuroanatomical Techniques. Berlin: Springer. pp. 132–205. Weiss, M.J. (1974) Neuronal connections and the function of the corpora pedunculata in the brain of the American cockroach, Periplaneta americana (L.). J. Morphol. 142:21–69. Weiss, M.J. (1981) Structural patterns in the corpora pedunculata of Orthoptera: A reduced silver analysis. J. Comp. Neurol. 203:515–553. Witthöft, W. (1967) Absolute Anzahl und Verteilung der Zellen im Hirn der Honigsbiene. Z. Morphl. Tiere. 61:160–184 Yamasaki, T. and T. Narahashi (1959) The effects of potassium and sodium ions on the resting and action potentials of the cockroach giant axon. J. Insect Physiol. 3:146–158. Yamazaki, Y., M. Nishikawa, and M. Mizunami (1998) Three classes of GABA-like immunoreactive neurons in the mushroom body of the cockroach. Brain Res. 788:80–86. Yang, M.Y., J.D. Armstrong, I. Vilinsky, N.J. Strausfeld, and K. Kaiser (1995) Subdivision of the Drosophila mushroom bodies by enhancertrap expression patterns. Neuron 15:45–54.