Behavior is a manifestation of temporally and spatially defined neuronal activities. To understand how behavior is controlled by the nervous system, it is important to identify the neuronal substrates responsible for these activities, and to elucidate how they are integrated into a functional circuit. I introduce a novel and general method to conditionally perturb anatomically defined neurons in intact Drosophila. In this method, a temperature-sensitive allele of shibire (shi(ts1)) is overexpressed in neuronal subsets using the GAL4/UAS system. Because the shi gene product is essential for synaptic vesicle recycling, and shi(ts1) is semidominant, a simple temperature shift should lead to fast and reversible effects on synaptic transmission of shi(ts1) expressing neurons. When shi(ts1) expression was directed to cholinergic neurons, adult flies showed a dramatic response to the restrictive temperature, becoming motionless within 2 min at 30 degrees C. This temperature-induced paralysis was reversible. After being shifted back to the permissive temperature, they readily regained their activity and started to walk in 1 min. When shi(ts1) was expressed in photoreceptor cells, adults and larvae exhibited temperature-dependent blindness. These observations show that the GAL4/UAS system can be used to express shi(ts1) in a specific subset of neurons to cause temperature-dependent changes in behavior. Because this method allows perturbation of the neuronal activities rapidly and reversibly in a spatially and temporally restricted manner, it will be useful to study the functional significance of particular neuronal subsets in the behavior of intact animals.

A paired presentation of an odor and electric shock induces aversive odor memory in Drosophila melanogaster. Electric shock reinforcement is mediated by dopaminergic neurons, and it converges with the odor signal in the mushroom body (MB). Dopamine is synthesized in approximately 280 neurons that form distinct cell clusters and is involved in a variety of brain functions. Recently, one of the dopaminergic clusters (PPL1) that includes MB-projecting neurons was shown to signal reinforcement for aversive odor memory. As each dopaminergic cluster contains multiple types of neurons with different projections and physiological characteristics, functional understanding of the circuit for aversive memory requires cellular identification. Here, we show that MB-M3, a specific type of dopaminergic neurons in the PAM cluster, is preferentially required for the formation of labile memory. Strikingly, flies formed significant aversive odor memory without electric shock when MB-M3 was selectively stimulated together with odor presentation. In addition, we identified another type of dopaminergic neurons in the PPL1 cluster, MB-MP1, which can induce aversive odor memory. As MB-M3 and MB-MP1 target the distinct subdomains of the MB, these reinforcement circuits might induce different forms of aversive memory in spatially segregated synapses in the MB.

Surgical, pharmacological and genetic lesion studies have revealed distinct anatomical sites involved with different forms of learning. Studies of patients with localized brain damage and work in rodent model systems, for example, have shown that the hippocampal formation participates in acquisition of declarative tasks but is not the site of their long-term storage. Such lesions are usually irreversible, however, which has limited their use for dissecting the temporal processes of acquisition, storage and retrieval of memories. Studies in bees and flies have similarly revealed a distinct anatomical region of the insect brain, the mushroom body, that is involved specifically in olfactory associative learning. We have used a temperature-sensitive dynamin transgene, which disrupts synaptic transmission reversibly and on the time-scale of minutes, to investigate the temporal requirements for ongoing neural activity during memory formation. Here we show that synaptic transmission from mushroom body neurons is required during memory retrieval but not during acquisition or storage. We propose that the hebbian processes underlying olfactory associative learning reside in mushroom body dendrites or upstream of the mushroom body and that the resulting alterations in synaptic strength modulate mushroom body output during memory retrieval.

Mutations in the amnesiac gene in Drosophila affect both memory retention and ethanol sensitivity. The predicted amnesiac gene product, AMN, is an apparent preproneuropeptide, and previous studies suggest that it stimulates cAMP synthesis. Here we show that, unlike other learning-related Drosophila proteins, AMN is not preferentially expressed in mushroom bodies. Instead, it is strongly expressed in two large neurons that project over all the lobes of the mushroom bodies, a finding that suggests a modulatory role for AMN in memory formation. Genetically engineered blockade of vesicle recycling in these cells abbreviates memory as in the amnesiac mutant. Moreover, restoration of amn gene expression to these cells reestablishes normal olfactory memory in an amn deletion background. These results indicate that AMN neuropeptide release onto the mushroom bodies is critical for normal olfactory memory.

Biogenic amines, such as serotonin and dopamine, can be important in reinforcing associative learning. This function is evident as changes in memory performance with manipulation of either of these signals. In the insects, evidence begins to argue for a common role of dopamine in negatively reinforced memory. In contrast, the role of the serotonergic system in reinforcing insect associative learning is either unclear or controversial. We investigated the role of both of these signals in operant place learning in Drosophila. By genetically altering serotonin and dopamine levels, manipulating the neurons that make serotonin and dopamine, and pharmacological treatments we provide clear evidence that serotonin, but not dopamine, is necessary for place memory. Thus, serotonin can be critical for memory formation in an insect, and dopamine is not a universal negatively reinforcing signal.biogenic amines ͉ dopamine ͉ learning ͉ white-ABC transporter ͉ reinforcement T he neural systems containing biogenic amines, such as dopamine and serotonin, may mediate reinforcement information to influence memory performance. In the monkey for example, activity in the dopaminergic system is modulated based on expected reward (1), and the phasic output of these neurons may regulate memory performance (1, 2). In some invertebrates the biogenic amines have also been shown to be critical for conditioning (3)(4)(5). Within the insects, however, dopamine is the only biogenic amine clearly implicated in negatively reinforced associative memory (6-8). Indeed, and interestingly, dopaminergic system activation can be a sufficient reinforcing signal for olfactory conditioning in Drosophila larvae (9). Thus, support grows for a general function of the dopaminergic system in negatively reinforced memory. Whether serotonin has a role in insect learning is less clear (10), and in Drosophila it is controversial (11-13). Here, we investigated the influence of serotonin and dopamine on reinforcement of place learning in Drosophila.The ''heat box'' can be used to rapidly condition place memories in Drosophila (14,15). In this paradigm, single flies are allowed to wander in a chamber that is lined top and bottom with Peltier heating elements ( Fig. 1) (16, 17). A series of light sensors on one side of the chamber tracks the behavior of a fly, and when the animal moves to a predetermined half, the whole chamber heats to a nonpreferred (aversive) temperature. With experience, normal flies avoid the chamber-half associated with rising temperatures (15,16,18). A test performed after conditioning, when the danger of rising temperature is removed, is used to measure place memory. Importantly, one can dissociate acquisition from reinforcement processing defects by the performance of mutant flies after short and long training sessions (19). Flies that are mutant for a type-1 adenylyl cyclase (i.e., rutabaga) show poor memory performance after short periods of conditioning but normal memory after longer training, emphasizing the memory acquisition function for...