The mushroom body (MB) in the insect brain is composed of a large number of densely packed neurons called Kenyon cells (KCs) (Drosophila, 2500; honeybee, 170,000). In most insect species, the MB consists of two cap-like dorsal (developmentally: frontal) structures, the calyces, which contain the dendrites of KCs and two to four lobes formed by collaterals of branching KC axons. Although the MB receives input and provides output throughout its whole structure, the neuropil part of the calyx receives predominantly input from sensory projection neurons (PNs) of second or higher order, and the lobes send output neurons to many other parts of the brain, including recurrent neurons to the MB calyx (Fig. 44.1B). Widely branching supposedly modulatory neurons innervate the MB at all levels (calyx, peduncle, lobes), including the somata of KCs in the calyx (dopamine) (Fig. 44.1A).
Two major classes of KCs, KI and KII, can be distinguished with different dendritic morphologies and axonal projections with respect to their location in the peduncle and lobes. KII cells exhibit claw-like specializations in narrow dendritic trees, whereas wide-field branching KI cells span larger areas in the calyx neuropil. The different subtypes of both cell types are found in all calycal neuropils. KII cell somata are located outside the calycal wall, whereas KI cell somata are clustered in the central (small somata) and the adjacent zones (large-diameter somata) within the calyces. KII axons project through the peduncle into the ventral α-lobe and anterior β-lobe, whereas KI cell axons occupy horizontally stratified areas of the dorsal α-lobe and posterior β-lobe, representing the calycal zone’s lip, collar, and basal ring (Rybak and Menzel, 1993; Strausfeld, 2002).
Projection neurons originating in the sensory neuropils innervate the calyx in an orderly fashion. Olfactory PNs branch predominantly in the lip, visualPNs in the collar, mechanosensory PNs in the basal ring, and gustatory PNs in a small region between lip and collar. The intrinsic structure imposed by the PNs onto the KCs is kept throughout the MB with respect to the course of the KI cells leading to an orderly subdivision of the lobes with the KCs of the lip projecting to the ventral zone of the dorsal α-lobe, those of the collar in the layer above, and those of the basal ring in the most dorsal part. A corresponding layering is found in the KC collaterals forming the β-lobe (Fig. 44.1A). In summary, due to the subdivision of KCs into two classes, the ventral part of the α-lobe is composed of KII-type KCs homogenously distributed throughout all calycal zones, and the dorsal α-lobe is layered by the orderly projection of the KI cells from distinct calycal subcompartments (Fig. 44.1A).
The branching pattern of extrinsic neurons of the α-lobe (ext in Fig. 44.1A) and the density of synaptic contacts between KCs and extrinsic neurons lead to a stratification (Fig. 44.1A and B). One class of extrinsic neurons (A3 neurons) is at least partly immunoreactive (ir) to a GABA antibody and is considered to provide inhibitory recurrent information from the α-lobe, β-lobe, and peduncle to the calyx besides local inhibitory connections between KCs and extrinsic neurons in both α- and β-lobe (Grünewald, 1999; Ganeshina and Menzel, 2001) (Figs. 44.1B and 44.2B). A3 neurons forming dendritic strata within the α-lobe also project parallel to the KC type KI into the peduncle and the β-lobe. Outside the MB they run to the calyces and terminate in all subcompartments close to KC type KI and KII. Other GABA-ir cells (ext in Fig. 44.1B) invade the ventral α-lobe and peduncle and run parallel to axons from KII cells (Fig. 44.1B)
Functional Organization of the Calyx
Projection neurons diverge and converge on KCs in the calyx, leading to a matrix of connectivity (Fig. 44.2A). On average, one KC receives input from 10–15 olfactory PNs, and each of these PNs provides input to 50–100 KCs. Each PN presynaptic bouton comprises a microcircuit composed ofmultiple outputs to KCs, input and output from and to GABA ir profiles representing the recurrent A3 neurons, output from these profiles onto KC spines, and tentative modulatory neurons with en passant synapses containing dense core vesicles (Fig. 44.2B). Putative inhibitory input from A3 neurons also reaches PNs on axodendrons (Fig. 44.2C). Olfactory PN boutons respond to odor stimulation with excitation and/or inhibition in an odor identity–specific way. Olfactory, gustatory, and tactile stimuli excite KCs (type KII) transiently and lead to Off rebound excitation, indicating delayed release from inhibition (Fig. 44.2D). Odors are coded in KCs (type KII) in a sparse way both in the temporal and the population domain, indicating that inhibition via recurrent A3 neurons is a prominent feature of the circuit (Szyszka et al., 2005).
Repeated stimulation leads to stimulus-specific decrease of KC responses. Pairing an odor as conditioned stimulus with sucrose reward in an associative learning situation induces prolonged KC responses. After conditioning, KC responses to the rewarded odor recover from repetition-induced decrease, while the responses to a nonrewarded odor decrease further. The spatiotemporal pattern of activated KCs changes both for the learned and the specifically not learned odor (Szyszka et al., 2008). These results document that KC responses are subject to nonassociative plasticity during odor repetition and undergo associative plasticity after appetitive odor learning.
NMDA-like receptors are localized throughout the bee brain, including the MB. Silencing the expression of the NR1 subunit of the NMDA receptor (NMDAR) in the MB by RNAi impairs selectively the acquisition phase and the formation of middle-term memory leaving long-term memory intact. It is concluded that NMDARs are not coincidence detectors in the MB but are rather involved in the formation of particular memories.
A putative octopaminergic neuron, the VUMmx1, represents the reinforcing function during olfactory conditioning of the bee. It receives input from sucrose receptors in the suboesophageal ganglion. The axon arborizes bilaterally in both antennal lobes, the lateral horns, and the lip and basal ring regions of the calyces (Hammer, 1993).
Functional Organization of the α-Lobe
One of the α-lobe extrinsic neurons, the PE1, is an identified single neuron projecting to the lateral protocerebral lobe of the bee brain (Fig. 44.2E). It responds to a large range of stimuli (olfactory, gustatory, tactile, and visual) with prolonged excitation. Pairing electrical stimulation of KCs with depolarization of PE1 leads to associative long-term potentiation (LTP) (Menzel and Manz, 2005) (Fig. 44.2F). PE1 also changes its response properties during olfactory reward learning. In such a paradigm PE1 reduces its response specifically to the learned odor. PE1 receives putative inhibitory input fromA3 neurons (Okada et al., 2007). Since A3 neurons also change their odor responses during associative odor learning, it is possible that the reduced PE1 responses to the learned odor result from an increase of inhibition induced by the learned odor via A3 neurons. However, it is also possible that associative plasticity is an intrinsic property of PE1 itself. In such a case, behavioral learning would have to induce long-term depression (LTD) rather than LTP. The transition between LTD and LTP may be controlled by local Ca2+ activities in spines because it is known to occur in mammalian cortical neurons. The main limitation in delineating the processes of long-term synaptic plasticity in MB extrinsic neurons and its relation to behavioral learning lies in the fact that the transmitter(s) of KCs and the respective receptors in the extrinsic neurons are not known.
Global Properties of the Mushroom Body
The high divergence of sensory input to KCs, their sparse coding properties, and the necessity of coincident activity via PNs makes it likely that KCs code sensory modalities in a highly specific way. The MB extrinsic neurons, however, integrate across sensory modalities and change their properties during learning in multiple ways. It is, therefore, concluded that the MB recodes the highly dimensional sensory space at it input sites (calyx) into a low-dimensional space of meaning and value with a special emphasis on acquired forms of recoding at its output sites (lobes). This view is supported by the finding that the MB in insects is a necessary structure for olfactory learning and memory consolidation. In honeybees, reversible blocking neural activity in the MB by local cooling leads to retrograde amnesia within a few minutes after a single learning trial. Furthermore, several cellular pathways in MB neurons are known to be related to learning and memory formation, and manipulation of the structural integrity of the MB during ontogeny leads to compromised olfactory learning in adult bees.
Ganeshina OT, Menzel R (2001) GABA-immunoreactive neurons in the mushroom bodies of the honeybee: An electron microscopic study. J Comp Neurol 437:335–349.Find this resource:
Grünewald B (1999) Morphology of feedback neurons in the mushroom body of the honeybee, Apis mellifera. J Comp Neurol 404:114–126.Find this resource:
Grünewald B, Wersing A, Wüstenberg DG (2004) Learning channels. Cellular physiology of odor processing neurons within the honeybee brain. Acta Biol Hungarica 55:53–63.Find this resource:
Hammer M (1993) An identified neuron mediates the unconditioned stimulus in associative olfactory learning in honeybees. Nature 366:59–63.Find this resource:
Menzel R, Manz G (2005) Neural plasticity of mushroom body-extrinsic neurons in the honeybee brain. J Exp Bio 208(22):4317–4332.Find this resource:
Rybak J, Menzel R (1993) Anatomy of the mushroom bodies in the honey bee brain: the neuronal connections of the alpha-lobe. J Comp Neurol 334:444–465.Find this resource:
Strausfeld NJ (2002) Organization of the honey bee mushroom body: representation of the calyx within the vertical and gamma lobes. J Comp Neurol 450:4–33.Find this resource:
Szyszka P, Ditzen M, Galkin A, Galizia CG, Menzel R (2005) Sparsening and temporal sharpening of olfactory representations in the honeybee mushroom bodies. J Neurophysiol 94:3303–3313.Find this resource:
Szyszka P, Galkin A, Menzel R (2008) Associative and non-associative plasticity in Kenyon cells of the honeybee mushroom body. Front Syst Neurosci 2:1–10.Find this resource: