Olfaction and olfactory learning in Drosophila: recent progress Andre´ Fiala

The olfactory system of Drosophila resembles that of experience. For example, an odor repetitively paired with vertebrates in its overall anatomical organization, but is a food reward becomes attractive. Conversely, an odor considerably reduced in terms of cell number, making it an ideal that often occurs concurrently with a punishment model system to investigate odor processing in a brain becomes a predictor for a negative situation and will be [Vosshall LB, Stocker RF: Molecular architecture of smell avoided. Drosophila melanogaster can easily perform such and taste in Drosophila. Annu Rev Neurosci 2007, 30:505- learning tasks and represents an excellent organism to 533]. Recent studies have greatly increased our knowledge investigate the neuronal mechanisms underlying such about odor representation at different levels of integration, from olfactory learning processes for two reasons. First, con- olfactory receptors to ‘higher brain centers’. In addition, siderable progress has already been made during recent Drosophila represents a favourite model system to study the years in analyzing how odors are represented in the fly’s neuronal basis of olfactory learning and memory, and brain [1]. Second, the powerful genetic techniques by considerable progress during the last years has been made in which structure and function of identified neurons can be localizing the structures mediating olfactory learning and observed and manipulated makes Drosophila an ideal memory [Davis RL: Olfactory memory formation in neurobiological model system to characterize a neuronal Drosophila: from molecular to systems neuroscience. Annu network that mediates olfactory learning and memory [2– Rev Neurosci 2005, 28:275-302; Gerber B, Tanimoto H, 4]. The scope of this review is to summarize recent Heisenberg M: An engram found? Evaluating the evidence advances and to point out gaps and caveats in our current from fruit flies. Curr Opin Neurobiol 2004, 14:737-744; Keene understanding of olfactory coding and olfactory learning AC, Waddell S: Drosophila olfactory memory: single genes in Drosophila. to complex neural circuits. Nat Rev Neurosci 2007, 8:341- 354]. This review summarizes recent progress in analyzing Olfactory representations in the Drosophila olfactory processing and olfactory learning in Drosophila. brain Fruitflies perceive odors through olfactory sensory neurons Addresses (OSNs) that reside in sensillae of diverse morphological Department of Genetics and Neurobiology, Theodor-Boveri-Institut, Julius-Maximilians-Universita¨ tWu¨ rzburg, Biozentrum, Am Hubland, types on the third antennal segments and the maxillary 97074 Wu¨ rzburg, Germany palps. Each of these sensory neurons expresses usually one, but sometimes two or three out of 61 specific olfactory Corresponding author: Fiala, Andre´ receptor proteins (ORs). In addition, the non-specific re- (afi[email protected]) ceptor Or83b is expressed in almost all olfactory sensory neurons and mediates targeting and functionality of the Current Opinion in Neurobiology 2007, 17:720–726 heterodimers it forms with the odor-selective receptor [5– 7]. The principle of connecting the OSNs to the brain is This review comes from a themed issue on simple: OSNs expressing the same ORs converge onto one Neurobiology of behaviour Edited by Edvard Moser and Barry Dickson or very few out of 40–50 spherical synaptic modules, the glomeruli of the antennal lobe [8,9]. Available online 1st February 2008

0959-4388/$ – see front matter The response profiles of these OSNs are reflected in the # 2007 Elsevier Ltd. All rights reserved. range of volatile compounds with which their ORs interact. Some of the ORs are highly specialized with respect to DOI 10.1016/j.conb.2007.11.009 their response profile. For example, the two receptors Gr21a and Gr63a which actually belong to the family of gustatory receptors are co-expressed in the same antennal Introduction OSNs. These neurons target one specific glomerulus, the Animals such as fruitflies navigate in a complex chemo- V-glomerulus [10] and mediate very specifically the fly’s sensory environment. Some odors can act as signals for olfactory detection of CO2 [10–12]. Artificial activation of food or danger or as pheromones released by conspecifics just these neurons using the light-sensitive cation channel eliciting innate behavioral responses. However, many ‘channelrhodopsin-2’ is sufficient to induce the fly’s very odor stimuli are not informative per se to optimally guide pronounced avoidance response normally elicited by CO2 the animal’s behavior. The brain has to make sense of the [13]. complexity of odor signals by interpreting their relevance. Associative learning represents one process by which new In contrast to such an odor selectivity among fruitfly OR or altered relevance is assigned to a stimulus through subtypes, other ORs can be more broadly activated by a

Current Opinion in Neurobiology 2007, 17:720–726 www.sciencedirect.com Odor learning in Drosophila Fiala 721

variety of volatiles. For example, the response range of vioral responses cannot be deduced from the knowledge OSNs expressing one particular receptor (Or22a), ana- of combinatorial OR activations and inhibitions. Rather, lysed in a detailed calcium imaging study by Pelz et al. processing of olfactory information might alter the [14], covers 39 out of 104 tested odorants. However, the particular contribution of defined OSNs to the ultimate calcium response evoked by ethyl hexanoate and methyl behavior-releasing effect of an odor. hexanoate, components of several fruit odors, clearly exceeded those evoked by the other volatiles tested. A In fact, the antennal lobe comprises a complex network of systematic analysis of 24 odorant receptors has been several types of neurons. Around 150, mostly uniglomer- reported in an impressive and very comprehensive study ular, olfactory projection neurons (PNs) in each hemi- by Hallem and Carlson [15], expanding results from a sphere receive inputs from 1300 OSNs and convey odor previous publication [16]. Here, the authors misexpressed information from the antennal lobe to the lateral horn and the different ORs in a neuron that lacks its endogenous the mushroom body, but provide also local output within receptor and recorded electrophysiologically neuronal the antennal lobe. Multiglomerular local interneurons activity in the corresponding sensillum evoked by 110 which are diverse with respect to their transmitter project different volatiles across various concentrations. This throughout large parts of the antennal lobe. Inhibitory study revealed that a sharp division into generalist and actions mediated by GABAergic local neurons shape the specialist receptors cannot be actually made. Rather, PN responses [18]. More recently, excitatory, cholinergic tuning breadths of ORs are variable, with some receptors interneurons have also been found that broaden the being activated by one or very few compounds, others response spectra of individual PNs [19,20]. The net effect responding to a wide variety of compounds, and a con- of these processing circuits has been subject to a con- tinuum of response spectra in between. In addition, troversial discussion. Whereas optical imaging studies inhibition of a receptor by a particular odorant occurs have suggested that the antennal lobe’s input provided relatively often, demonstrating that odor representations via OSNs is essentially mirrored by PN output [21,22], at this initial level of processing are characterized by both electrophysiological studies have indicated a more com- excitation and inhibition of particular OSNs. Taken plex transformation of odor information [23]. A recent together, these studies offer the fascinating prospect that publication by Bandhavat et al. [24] clearly demon- a nearly complete knowledge of possible OSN responses strates that response spectra of PNs receiving their main providing odor information to the fly’s brain is within input from ‘non-specialist’ OSNs are considerably reach. Moreover, as OSNs expressing the same ORs broader compared to their input counterparts. Interest- target the same identified glomeruli in the antennal lobe, ingly, the odorant evoking the strongest activity in a one can create a map of odotopic representations in this particular OSN is not necessarily the most efficient odor primary olfactory neuropil. Two laboratories have started to activate its corresponding PN [24]. doing just that [8,9]. By tracing the projections of sensory neurons to their antennal lobe targets a detailed receptor- PNs transmit olfactory information to the lateral horn, a to-glomerulus map has been proposed. This anatomical brain region whose exact mode of function is not well connectivity scheme could be combined with the already understood. Most of the PNs also target en passent the established response profiles of ORs to provide the first calyx, the main olfactory input region of the mushroom framework for a functional atlas of the antennal lobe [8,9]. body. The mushroom body of each hemisphere consist of 2500 intrinsic neurons, the Kenyon cells, which can be These advances made over the last years have led to a divided into various classes due to their birth order, gene more and more comprehensive description of the periph- expression and axonal projections: early a/b-, late a/b-, eral mechanisms underlying the ‘odor space’ of a fly. a0/b0- and g-neurons. Several anatomical studies have However, little information is available as to what extent revealed that individual PNs originating from identified individual ORs contribute to the odor-evoked behavior of antennal lobe glomeruli send their terminal arborisations a fly. In this context a publication by Keller and Vosshall stereotypically to distinct regions of the lateral horn [25– [17] has added important new results. Changes in both 28] and the mushroom body calyx [27–29], in agreement locomotor activity and distance to an odor source were with distinct odor-evoked activity patterns in the calyx monitored to determine the flies’ behavioral responses to [30]. However, whether the mushroom body maintains a 73 different odors. Interestingly, the deletion of a single segregation of distinct olfactory channels remains unclear. class of ORs (Or22a or Or43b), had only very subtle Tanaka et al. [27] describe the mushroom body as a effects on the odor-induced behavior and, more impor- mainly integrative center, with all types of Kenyon cells tantly, did not alter the response to the odors that activate receiving input from all input regions, although with the respective OR most strongly. These results indicate slight differences between particular Kenyon cell types first a strong redundancy within the system, reflecting the in their arborisation across the calyx. By contrast, Lin et al. fact that some odorants activate multiple ORs, and single [29] report a more strict separation of Kenyon cell popu- ORs are often activated by several odorants. Secondly, lations receiving input from distinct groups of projection they point towards the possibility that particular beha- neurons, which argues for a categorization or segregation www.sciencedirect.com Current Opinion in Neurobiology 2007, 17:720–726 722 Neurobiology of behaviour

of olfactory information in the mushroom body. It will be impact these correlative changes in odor representation important to clarify this point in greater detail to further might have for the experience-dependent change in understand the mushroom body’s mode of action. How- behavior. In fact, a memory in a broad sense might ever, it appears important to note that no anatomical study potentially contain multiple aspects and components, alone can reveal odor representations or functional maps. e.g. improved discrimination between CS+ and CS, Unfortunately, physiological data on the function of habituation to CS, generalization of odor-evoked mushroom body neurons in Drosophila are limited. In responses, increase in arousal caused by the US, among this respect an optical imaging study reported by Wang others. In addition it is difficult to exclude that experi- et al. [31] is of great importance. The authors have ence-dependent changes in neuronal activity observed in monitored calcium activity in Kenyon cells evoked by a given neuronal population simply reflect learning- odor stimuli of various concentrations and find that induced modifications occurring somewhere else in the chemically dissimilar odors activate stereotypically very brain. Therefore, it appears useful to differentiate be- small, non-overlapping subsets of Kenyon cells. Despite tween all possible changes in neuronal activity correlating the fact that the calcium sensor used might underestimate with the training procedure and those neuronal changes the Kenyon cells’ activity, the overall result is in good that are really necessary and sufficient to drive the con- accordance with electrophysiological data achieved from ditioned response, that is the memory trace [3]. The locusts [32,33]. There, only very few Kenyon cells necessity of intact mushroom bodies for the formation respond with very few action potentials to a given odor, of an aversive short-term memory has been demonstrated a phenomenon often referred to as ‘sparse code’, in by various classical experimental platforms: structural contrast to the combinatorial ‘ensemble’ code found in mushroom body mutants [38], flies with chemically more peripheral olfactory neurons. Under the assumption ablated mushroom bodies [39] and transgenic flies over- of a sparse code the mushroom body appears to be a well- expressing a dominant negative G-protein in the Kenyon suited substrate for associative learning because a reinfor- cells [40] all show impairments in olfactory learning. cing reward or punishment signal might interact with very specific odor representations provided by Kenyon cells. A more elaborate method has also substantiated the Alternatively, the mushroom body might be activated mushroom body’s role in aversive olfactory memory more extensively by a given odor, a hypothesis that is formation. If indeed changes in Kenyon cell synapses supported by broadly distributed odor-evoked calcium determine the fly’s behavioral change in response to the signals in a/b-lobes shown by Yu et al. [34]. odor, one would expect that blocking the mushroom body’s output during the test situation should impair Localization of olfactory memory traces memory retrieval. In fact, this has been shown by using The most commonly used associative learning paradigm the temperature-sensitive dynamin transgene shibirets, in Drosophila relies on a differential Pavlovian condition- which allows one to block synaptic transmission by ing procedure [35] in which one odor (conditioned raising the temperature [41–43]. If synaptic output from stimulus+ or CS+) is temporally paired with electric Kenyon cells during testing is prevented, memory reten- shocks (unconditioned stimulus or US). A second odor tion cannot be observed. However, if synaptic output is is presented equally often, but without any punishment blocked only during training, memory acquisition is (conditioned stimulus or CS). In a subsequent choice unaffected, supporting the idea that any memory trace test flies avoid the odor associated with the punishment. can be localized at or upstream of the mushroom body’s The only factor that distinguishes CS+ from CS is the output synapses. A recent publication by Krashes et al. coincident occurrence of the punishment. Consequently, [44] has refined this point of view. By restricting the the change in behavior can be attributed to the coinci- expression of shibirets to subsets of Kenyon cells they dence of CS+ and US in contrast to CS alone. show that output from the Kenyon cells whose axons form the a/b-lobes is in fact necessary only for memory Optical imaging experiments have revealed that in fact retrieval, whereas output from a0/b0-Kenyon cells is the glomerular activity patterns in the antennal lobe required during acquisition and consolidation for the evoked by the odor used as CS+ transiently changes formation of a stable memory, but not for its read-out. for up to 7 min after training. In response to the CS+ These results imply differential roles of the various an additional glomerulus is recruited as a result of train- Kenyon cell subtypes and a sequential use of diverse ing, which can be attributed to an increase in transmitter lobe systems during learning and memory retrieval. To release in PNs [36]. In addition, optical imaging exper- explain these results, the authors formulate a working iments focussing on a particular optical section within the model in which a pair of mushroom body innervating a-lobe show a change in activity as a result of a training neurons (DPM neurons), whose constitutive activity is procedure that induces long-term memory [34]. This is essential for aversive and appetitive memory stabiliz- in agreement with a proposed role of a-lobe neurons in ation [45,46], provide a feedback loop between different long-term memory formation [37]. As fascinating as these lobe systems as a consolidation signal (see [4] for a results indeed are, it remains to be investigated what detailed discussion of this idea).

Current Opinion in Neurobiology 2007, 17:720–726 www.sciencedirect.com Odor learning in Drosophila Fiala 723

To clarify the sufficiency of intact Kenyon cells for forming odor with a sugar reward. In a test situation the flies an olfactory memory, a genetic trick has been used. The approach the odor associated with the reward. As already classical learning mutant rutabaga, which is deficient in the shown by Schwaerzel et al. [51] both aversive and appeti- activity of a type I adenylate cyclase, acquires a signifi- tive memories are rutabaga-dependent, and both types of cantly reduced associative olfactory memory. By expres- memories can be rescued by expressing rutabaga in mutant sing the wild type rutabaga gene in Kenyon cells in the Kenyon cells. However, the appetitive but not the aversive mutant background intact memory formation could be memory can also be rescued by expressing rutabaga in PNs. restored [47–49], which demonstrates that Kenyon cells This suggests that multiple redundant memory traces can are sufficient for a rutabaga-dependent formation of olfac- be formed. Referring to the so far known physiological tory memory. In a recent report Thum et al. [50]have differences between the response properties of PNs and expanded this concept. In addition to punishing the Kenyon cells one might speculate that the memory trace in animals using electric shocks, the authors have exploited PNs is less specific for a particular odorant, but further work an appetitive learning paradigm in which they paired an is required to validate this idea.

Figure 1

(Left side) Three-dimensional reconstruction of the Drosophila brain [56]. Olfactory projection neurons (red) convey odor information from the antennal lobe (light blue) to the mushroom body (dark blue) and the lateral horn (brown). Besides these olfactory neuropils structures depicted are the medulla (pink), lobula (orange), and central complex (green). (Right side) Schematic working model of the olfactory pathway in the Drosophila brain as a substrate for olfactory learning. Odors are perceived by 1300 olfactory sensory neurons located on the antenna which project to the spherical glomeruli of the antennal lobe. Those sensory neurons that are activated by an odor are depicted in red. The antennal lobe network including local inhibitory and excitatory neurons transform olfactory information so that projection neurons respond to a broader range of odorants compared to sensory neurons (illustrated by the pink projection neuron). Many of the 250 projection neurons form synapses with a large proportion of the 2500 mushroom body neurons. However, due to the physiological properties of these connections only those intrinsic mushroom body neurons that receive input from multiple activated projection neurons are activated by a given odor. During olfactory learning modulatory neurons releasing octopamine or dopamine are driven by reinforcing rewarding or punitive stimuli and change those synapses within the olfactory pathway that are simultaneously activated by an odor, ultimately altering the activity of mushroom body output neurons. www.sciencedirect.com Current Opinion in Neurobiology 2007, 17:720–726 724 Neurobiology of behaviour

Reinforcement signalling memory traces found for appetitive and aversive learning Considerable progress has been made to determine aver- [50] correlate with the presumed sites of coincidence sive and appetitive memory traces, that is to identify between odor representation and reinforcement. One can those synapses whose plasticity is necessary and sufficient propose a model in which the coincident activity of for the expression of olfactory memory. It appears likely activated Kenyon cells and the reinforcing signal that a memory trace localizes to the sites of interaction mediated by dopaminergic or octopaminergic neurons, between CS and US, where the coincidence between a respectively, lead to changes in the efficacy of Kenyon representation of the to-be-learned odor and a repres- cell output synapses, which ultimately causes a change in entation of the reinforcing aspect of the punishment or behavior. In addition, octopaminergic neurons might also the reward takes place. Which neurons might mediate the modify PN synapses in the antennal lobe and/or the reinforcing properties of the reward or the punishment? mushroom body calyx, which in itself is sufficient to Schwaerzel et al. [51] have shown that release of dopa- cause the conditioned behavior (see Figure 1). mine is necessary for aversive memory formation, but dispensable for appetitive learning. Conversely, a mutant Conclusions lacking octopamine is unaffected in aversive learning, but The deciphering of a brain circuitry mediating associative impaired in reward learning [51]. This prompts one to olfactory learning can be approached from two sides. speculate that these two types of modulatory transmitters First, neuronal odor representations at various levels of might mediate antagonistic reinforcement during the processing can be described and neurons responding to associative olfactory learning task (but see [52] for contra- relevant rewarding or punitive events can be character- dictory results). In optical imaging experiments it could ized. A description of how odors are represented at the be demonstrated that dopaminergic neurons that project level of OSNs and how odor information is transformed across parts of the mushroom body lobes are indeed through the antennal lobe network has far advanced. It responsive to punitive electric shocks, but only weakly will be of importance to clarify how different odors are to odor stimuli [53]. It will be interesting to test whether represented at the level of the mushroom body. Second, dopaminergic neurons are activated only by punitive one can aim to define those sites of plasticity that are stimuli or whether rewarding stimuli lead to dopamine necessary and/or sufficient for mediating the experience- release as well. Interestingly, the weak activation by odors dependent change in behavior. A growing body of evi- changes in the course of a training procedure. After dence suggests that aversive olfactory memory can be repetitive pairing of one odor with an electric shock localized to the mushroom body, with different Kenyon the response of dopaminergic neurons to the odor is cell subpopulations playing different roles in acquisition prolonged. This prolongation occurs in the test situation and stabilization of olfactory memory. However, an at a time at which the punishment has occurred in the additional memory trace found for appetitive, but not training, thus indicating a predictive property of dopa- aversive memory formation in PNs strongly suggests that minergic neurons. It is tempting to speculate that during multiple memory traces can co-exist within the brain. It learning the ‘punished’ odor has acquired reinforcing will be fascinating to see in the future how these two sites properties and can be used as a punitive signal in a of memory formation relate to each other in terms of second-order conditioning experiment. However, as men- ‘memory content’. tioned above, correlative monitoring of neuronal activities alone never reveals causal connections between a change Acknowledgements in activity and a change in behavior. The sufficiency of I am grateful to Erich Buchner, Martin Heisenberg, Hiromu Tanimoto and Peter Bengtson for helpful and stimulating discussions on the manuscript. I octopaminergic and dopaminergic neurons to induce regret that – because of space constraints – not all papers contributing to the appetitive or aversive olfactory memories, respectively, advancement in the field could be cited. has been demonstrated in Drosophila larvae [54]. Here, a sugar reward can be substituted by light-induced acti- References and recommended reading vation of octopaminergic and/or tyraminergic neurons. Papers of particular interest, published within the period of review, have been highlighted as: Conversely, the reinforcing effect of a punitive salt stimulus can be substituted by activating dopaminergic of special interest of outstanding interest neurons. This directly proves a causal link between neuronal activity and memory formation. 1. Vosshall LB, Stocker RF: Molecular architecture of smell and taste in Drosophila. Annu Rev Neurosci 2007, 30:505-533. Another interesting finding by Riemensperger et al. [53] 2. Davis RL: Olfactory memory formation in Drosophila: from was that dopaminergic neurons extensively innervate the molecular to systems neuroscience. Annu Rev Neurosci 2005, 28:275-302. mushroom body, but dopaminergic arborisation in the antennal lobe is barely detectable. By contrast, octopa- 3. Gerber B, Tanimoto H, Heisenberg M: An engram found? Evaluating the evidence from fruit flies. Curr Opin Neurobiol minergic neurons branch extensively throughout the 2004, 14:737-744. antennal lobe, in addition to the innervation of parts of 4. Keene AC, Waddell S: Drosophila olfactory memory: single genes the mushroom body lobes [55]. In this respect, the to complex neural circuits. Nat Rev Neurosci 2007, 8:341-354.

Current Opinion in Neurobiology 2007, 17:720–726 www.sciencedirect.com Odor learning in Drosophila Fiala 725

5. Benton R, Sachse S, Michnick SW, Vosshall LB: Atypical The publication describes very thoroughly how olfactory information is membrane topology and heteromeric function of Drosophila transformed by the antennal lobe network from olfactory sensory neurons odorant receptors in vivo. PLoS Biol 2006, 4:e20. to projection neurons with respect to a variety of parameters, such as temporal dynamics, reliability and separability of odor-evoked activity as 6. Larsson MC, Domingos AI, Jones WD, Chiappe ME, Amrein H, well as response profiles. Vosshall LB: Or83b encodes a broadly expressed odorant receptor essential for Drosophila olfaction. Neuron 2004, 25. Wong AM, Wang JW, Axel R: Spatial representation of the 43:703-714. glomerular map in the Drosophila protocerebrum. Cell 2002, 109:229-241. 7. Neuhaus EM, Gisselmann G, Zhang W, Dooley R, Stortkuhl K, Hatt H: Odorant receptor heterodimerization in the olfactory 26. Marin EC, Jefferis GS, Komiyama T, Zhu H, Luo L: Representation system of Drosophila melanogaster. Nat Neurosci 2005, 8:15- of the glomerular olfactory map in the Drosophila brain. Cell 17. 2002, 109:243-255. 8. Couto A, Alenius M, Dickson BJ: Molecular, anatomical, and 27. Tanaka NK, Awasaki T, Shimada T, Ito K: Integration of functional organization of the Drosophila olfactory system. chemosensory pathways in the Drosophila second-order Curr Biol 2005, 15:1535-1547. olfactory centers. Curr Biol 2004, 14:449-457. 9. Fishilevich E, Vosshall LB: Genetic and functional 28. Jefferis GS, Potter CJ, Chan AM, Marin EC, Rohlfing T, Maurer CR subdivision of the Drosophila antennal lobe. Curr Biol 2005, Jr, Luo L: Comprehensive maps of Drosophila higher olfactory 15:1548-1553. centers: spatially segregated fruit and pheromone representation. Cell 2007, 128:1187-1203. 10. Suh GS, Wong AM, Hergarden AC, Wang JW, Simon AF, Benzer S, Axel R, Anderson DJ: A single population of olfactory sensory 29. Lin HH, Lai JS, Chin AL, Chen YC, Chiang AS: A map of olfactory neurons mediates an innate avoidance behaviour in representation in the Drosophila mushroom body. Cell 2007, Drosophila. Nature 2004, 431:854-859. 128:1205-1217. 11. Jones WD, Cayirlioglu P, Kadow IG, Vosshall LB: Two 30. Fiala A, Spall T, Diegelmann S, Eisermann B, Sachse S, chemosensory receptors together mediate carbon dioxide Devaud JM, Buchner E, Galizia CG: Genetically expressed detection in Drosophila. Nature 2007, 445:86-90. cameleon in Drosophila melanogaster is used to visualize olfactory information in projection neurons. Curr Biol 2002, 12. Kwon JY, Dahanukar A, Weiss LA, Carlson JR: The molecular 12:1877-1884. basis of CO2 reception in Drosophila. Proc Natl Acad Sci U S A 2007, 104:3574-3578. 31. Wang Y, Guo HF, Pologruto TA, Hannan F, Hakker I, Svoboda K, Zhong Y: Stereotyped odor-evoked activity in the mushroom 13. Suh GS, Ben-Tabou de Leon S, Tanimoto H, Fiala A, Benzer S, body of Drosophila revealed by green fluorescent protein- Anderson DJ: Light activation of an innate olfactory avoidance based Ca2+ imaging. J Neurosci 2004, 24:6507-6514. response in Drosophila. Curr Biol 2007, 17:905-908. This optical imaging study provides data how intrinsic mushroom body 14. Pelz D, Roeske T, Syed Z, de Bruyne M, Galizia CG: The neurons in Drosophila respond to odor stimuli. The authors demonstrate molecular receptive range of an olfactory receptor in vivo that strong activation of Kenyon cells evoked by odor stimuli occurs in (Drosophila melanogaster Or22a). J Neurobiol 2006, 66:1544- only small fractions of Kenyon cells. 1563. 32. Perez-Orive J, Mazor O, Turner GC, Cassenaer S, Wilson RI, 15. Hallem EA, Carlson JR: Coding of odors by a receptor Laurent G: Oscillations and sparsening of odor representations repertoire. Cell 2006, 125:143-160. in the mushroom body. Science 2002, 297:359-365. In this study the authors characterize the response profiles of 24 different 33. Jortner RA, Farivar SS, Laurent G: A simple connectivity scheme olfactory receptors. This work lays the foundation for a comprehensive for sparse coding in an olfactory system. J Neurosci 2007, functional analysis of the fruitfly’s olfactory receptor repertoire. 27:1659-1669. 16. Hallem EA, Ho MG, Carlson JR: The molecular basis of odor 34. Yu D, Akalal DB, Davis RL: Drosophila alpha/beta mushroom coding in the Drosophila antenna 117 . Cell 2004, :965-979. body neurons form a branch-specific, long-term cellular 17. Keller A, Vosshall LB: Influence of odorant receptor repertoire memory trace after spaced olfactory conditioning. Neuron on odor perception in humans and fruit flies. Proc Natl Acad Sci 2006, 52:845-855. USA2007, 104:5614-5619. This publication presents an optical imaging study on learning-induced changes of neuronal activity in mushroom body neurons. The authors 18. Wilson RI, Laurent G: Role of GABAergic inhibition in shaping find that a training procedure that can induce long-term memory leads odor-evoked spatiotemporal patterns in the Drosophila to a change in activity evoked by the trained odor. Interestingly, this antennal lobe. J Neurosci 2005, 25:9069-9079. change in activity is detectable only in a particular branch of Kenyon cells that has previously been shown to be involved in long-term 19. Olsen SR, Bhandawat V, Wilson RI: Excitatory interactions memory formation. between olfactory processing channels in the Drosophila antennal lobe. Neuron 2007, 54:89-103. 35. Tully T, Quinn WG: Classical conditioning and retention in normal and mutant Drosophila melanogaster. J Comp Physiol 20. Shang Y, Claridge-Chang A, Sjulson L, Pypaert M, Miesenbock G: [A] 1985, 157:263-277. Excitatory local circuits and their implications for olfactory processing in the fly antennal lobe. Cell 2007, 36. Yu D, Ponomarev A, Davis RL: Altered representation of the 128:601-612. spatial code for odors after olfactory classical conditioning; memory trace formation by synaptic recruitment. Neuron 2004, 21. Ng M, Roorda RD, Lima SQ, Zemelman BV, Morcillo P, 42:437-449. Miesenbock G: Transmission of olfactory information between three populations of neurons in the antennal lobe of the fly. 37. Pascual A, Preat T: Localization of long-term memory within the Neuron 2002, 36:463-474. Drosophila mushroom body. Science 2001, 294:1115-1117. 22. Wang JW, Wong AM, Flores J, Vosshall LB, Axel R: Two-photon 38. Heisenberg M, Borst A, Wagner S, Byers D: Drosophila calcium imaging reveals an odor-evoked map of activity in the mushroom body mutants are deficient in olfactory learning. fly brain. Cell 2003, 112:271-282. J Neurogenet 1985, 2:1-30. 23. Wilson RI, Turner GC, Laurent G: Transformation of olfactory 39. de Belle JS, Heisenberg M: Associative odor learning in representations in the Drosophila antennal lobe. Science 2004, Drosophila abolished by chemical ablation of mushroom 303:366-370. bodies. Science 1994, 263:692-695. 24. Bhandawat V, Olsen SR, Gouwens NW, Schlief ML, Wilson RI: 40. Connolly JB, Roberts IJ, Armstrong JD, Kaiser K, Forte M, Tully T, Sensory processing in the Drosophila antennal lobe increases O’Kane CJ: Associative learning disrupted by impaired Gs reliability and separability of ensemble odor representations. signaling in Drosophila mushroom bodies. Science 1996, Nat Neurosci 2007, 10:1474-1482. 274:2104-2107. www.sciencedirect.com Current Opinion in Neurobiology 2007, 17:720–726 726 Neurobiology of behaviour

41. Zars T, Fischer M, Schulz R, Heisenberg M: Localization of a 49. Schwaerzel M, Heisenberg M, Zars T: Extinction antagonizes short-term memory in Drosophila. Science 2000, 288:672-675. olfactory memory at the subcellular level. Neuron 2002, 35:951- 960. 42. McGuire SE, Le PT, Osborn AJ, Matsumoto K, Davis RL: Spatiotemporal rescue of memory dysfunction in Drosophila. 50. Thum AS, Jenett A, Ito K, Heisenberg M, Tanimoto H: Multiple Science 2003, 302:1765-1768. memory traces for olfactory reward learning in Drosophila. J Neurosci 2007, 27:11132-11138. 43. Mao Z, Roman G, Zong L, Davis RL: Pharmacogenetic rescue in This publication provides evidence for two independent memory traces in time and space of the rutabaga memory impairment by using the Drosophila brain. Whereas aversive and appetitive olfactory short- Gene-Switch. Proc Natl Acad Sci U S A 2004, 101:198-203. term memory can be localized to the mushroom body, an additional memory trace for appetitive short-term memory has been found in 44. Krashes MJ, Keene AC, Leung B, Armstrong JD, Waddell S: projection neurons. Sequential use of mushroom body neuron subsets during Drosophila odor memory processing. Neuron 2007, 53:103- 51. Schwaerzel M, Monastirioti M, Scholz H, Friggi-Grelin F, Birman S, 115. Heisenberg M: Dopamine and octopamine differentiate This study demonstrates that different subsets of Kenyon cells are between aversive and appetitive olfactory memories in required for distinct processes of olfactory learning and memory. Drosophila. J Neurosci 2003, 23:10495-10502. Whereas one population of Kenyon cells is needed for acquisition and stabilization of olfactory memory, another subset of Kenyon cells is 52. Kim YC, Lee HG, Han KA: D1 dopamine receptor dDA1 is required for the retrieval of the memory. required in the mushroom body neurons for aversive and appetitive learning in Drosophila. J Neurosci 2007, 27:7640- 45. Waddell S, Armstrong JD, Kitamoto T, Kaiser K, Quinn WG: The 7647. amnesiac gene product is expressed in two neurons in the Drosophila brain that are critical for memory. Cell 2002, 53. Riemensperger T, Voller T, Stock P, Buchner E, Fiala A: 103:805-813. Punishment prediction by dopaminergic neurons in Drosophila. Curr Biol 2005, 15:1953-1960. 46. Keene AC, Krashes MJ, Leung B, Bernard JA, Waddell S: Drosophila dorsal paired medial neurons provide a general 54. Schroll C, Riemensperger T, Bucher D, Ehmer J, Voller T, mechanism for memory consolidation. Curr Biol 2006, Erbguth K, Gerber B, Hendel T, Nagel G, Buchner E et al.: Light- 16:1524-1530. induced activation of distinct modulatory neurons triggers appetitive or aversive learning in Drosophila larvae. Curr Biol 47. Dubnau J, Grady L, Kitamoto T, Tully T: Disruption of 2006, 16:1741-1747. neurotransmission in Drosophila mushroom body blocks retrieval but not acquisition of memory. Nature 2001, 55. Sinakevitch I, Strausfeld NJ: Comparison of octopamine-like 411:476-480. immunoreactivity in the brains of the fruit fly and blow fly. J Comp Neurol 2006, 494:460-475. 48. McGuire SE, Le PT, Davis RL: The role of Drosophila mushroom body signaling in olfactory memory. Science 2001, 56. Rein K, Zo¨ ckler M, Mader MT, Gru¨ bel C, Heisenberg M: 293:1330-1333. The Drosophila standard brain. Curr Biol 2002, 12:227-431.

Current Opinion in Neurobiology 2007, 17:720–726 www.sciencedirect.com