Refine
Has Fulltext
- yes (39)
Is part of the Bibliography
- yes (39)
Year of publication
Document Type
- Doctoral Thesis (39) (remove)
Language
- English (39) (remove)
Keywords
- Drosophila (39) (remove)
Institute
Sonstige beteiligte Institutionen
Past experience contributes to behavioural organization mainly via learning: Animals learn otherwise ordinary cues as predictors for biologically significant events. This thesis studies such predictive, associative learning, using the fruit fly Drosophila melanogaster. I ask two main questions, which complement each other: One deals with the processing of those cues that are to be learned as predictors for an important event; the other one deals with the processing of the important event itself, which is to be predicted. Do fruit flies learn about combinations of olfactory and visual cues? I probe larval as well as adult fruit flies for the learning about combinations of olfactory and visual cues, using a so called ‘biconditional discrimination’ task: During training, one odour is paired with reinforcement only in light, but not in darkness; the other odour in turn is reinforced only in darkness, but not in light. Thus, neither the odours nor the visual conditions alone predict reinforcement, only combinations of both do. I find no evidence that either larval or adult fruit flies were to solve such task, speaking against a cross-talk between olfactory and visual modalities. Previous studies however suggest such cross-talk. To reconcile these results, I suggest classifying different kinds of interaction between sensory modalities, according to their site along the sensory-motor continuum: I consider an interaction ‘truly’ cross-modal, if it is between the specific features of the stimuli. I consider an interaction ’amodal’ if it instead engages the behavioural tendencies or ‘values’ elicited by each stimulus. Such reasoning brings me to conclude that different behavioural tasks require different kinds of interaction between sensory modalities; whether a given kind of interaction will be found depends on the neuronal infrastructure, which is a function of the species and the developmental stage. Predictive learning of pain-relief in fruit flies Fruit flies build two opposing kinds of memory, based on an experience with electric shock: Those odours that precede shock during training are learned as predictors for punishment and are subsequently avoided; those odours that follow shock during training on the other hand are learned as signals for relief and are subsequently approached. I focus on such relief learning. I start with a detailed parametric analysis of relief learning, testing for reproducibility as well as effects of gender, repetition of training, odour identity, odour concentration and shock intensity. I also characterize how relief memories, once formed, decay. In addition, concerning the psychological mechanisms of relief learning, first, I show that relief learning establishes genuinely associative conditioned approach behaviour and second, I report that it is most likely not mediated by context associations. These results enable the following neurobiological analysis of relief learning; further, they will form in the future the basis for a mathematical model; finally, they will guide the researchers aiming at uncovering relief learning in other experimental systems. Next, I embark upon neurogenetic analysis of relief learning. First, I report that fruit flies mutant for the so called white gene build overall more ‘negative’ memories about an experience with electric shock. That is, in the white mutants, learning about the painful onset of shock is enhanced, whereas learning about the relieving offset of shock is diminished. As they are coherently affected, these two kinds of learning should be in a balance. The molecular mechanism of the effect of white on this balance remains unresolved. Finally, as a first step towards a neuronal circuit analysis of relief learning, I compare it to reward learning and punishment learning. I find that relief learning is distinct from both in terms of the requirement for biogenic amine signaling: Reward and punishment are respectively signalled by octopamine and dopamine, for relief learning, either of these seem dispensible. Further, I find no evidence for roles for two other biogenic amines, tyramine and serotonin in relief learning. Based on these findings I give directions for further research.
The learned helplessness phenomenon is a specific animal behavior induced by prior exposure to uncontrollable aversive stimuli. It was first found by Seligman and Maier (1967) in dogs and then has been reported in many other species, e.g. in rats (Vollmayr and Henn, 2001), in goldfishes (Padilla, 1970), in cockroaches (Brown, 1988) and also in fruit flies (Brown, 1996; Bertolucci, 2008). However, the learned helplessness effect in fruit flies (Drosophila melanogaster) has not been studied in detail. Thus, in this doctoral study, we investigated systematically learned helplessness behavior of Drosophila for the first time.
Three groups of flies were tested in heatbox. Control group was in the chambers experiencing constant, mild temperature. Second group, master flies were punished in their chambers by being heated if they stopped walking for 0.9s. The heat pulses ended as soon as they resumed walking again. A third group, the yoked fly, was in their chambers at the same time. However, their behavior didn’t affect anything: yoked flies were heated whenever master flies did, with same timing and durations. After certain amount of heating events, yoked flies associated their own behavior with the uncontrollability of the environment. They suppressed their innate responses such as reducing their walking time and walking speed; making longer escape latencies and less turning around behavior under heat pulses. Even after the conditioning phase, yoked flies showed lower activity level than master and control flies. Interestingly, we have also observed sex dimorphisms in flies. Male flies expressed learned helplessness not like female flies. Differences between master and yoked flies were smaller in male than in female flies. Another interesting finding was that prolonged or even repetition of training phases didn’t enhance learned helplessness effect in flies.
Furthermore, we investigated serotonergic and dopaminergic nervous systems in learned helplessness. Using genetic and pharmacological manipulations, we altered the levels of serotonin and dopamine in flies’ central nervous system. Female flies with reduced serotonin concentration didn’t show helpless behavior, while the learned helplessness effect in male flies seems not to be affected by a reduction of serotonin. Flies with lower dopamine level do not display the learned helplessness effect in the test phase, suggesting that with low dopamine the motivational change in learned helplessness in Drosophila may decline faster than with a normal dopamine level.
In this work, a behavioural analysis of different mutants of the fruit fly Drosophila melanogaster has been carried out. Primarily, the gap climbing behaviour (Pick & Strauss, 2005) has been assayed as it lends itself for the investigation of decision making processes and the neuronal basis of adaptive behaviour. Furthermore it shows how basic motor actions can be combined into a complex motor behaviour. Thanks to the neurogenetic methods, Drosophila melanogaster has become an ideal study object for neurobiological questions. Two different modules of climbing control have been examined in detail. For the decision making, the mutant climbing sisyphus was analysed. While wild-type flies adapt the initiation of climbing behaviour to the width of the gap and the probability for a successful transition. climbing sisyphus flies initiate climbing behaviour even at clearly insurmountable gap widths. The climbing success itself is not improved in comparison to the wild-type siblings. The mutant climbing sisyphus is a rare example of a hyperactive mutant besides many mutants that show a reduced activity. Basic capabilities in vision have been tested in an optomotor and a distance-estimation paradigm. Since they are not affected, a defect in decision making is most probably the cause of this behavioural aberration. A second module of climbing control is keeping up orientation towards the opposite side of the gap during the execution of climbing behaviour. Mutants with a structural defect in the protocerebral bridge show abnormal climbing behaviour. During the climbing attempt, the longitudinal body axis does not necessarily point into the direction of the opposite side. Instead, many climbing events are initiated at the side edge of the walking block into the void and have no chance to ever succeed. The analysed mutants are not blind. In one of the mutants, tay bridge1 (tay1) a partial rescue attempt used to map the function in the brain succeeded such that the state of the bridge was restored. That way, a visual targeting mechanism has been activated, allowing the flies to target the opposite side. When the visibility of the opposing side was reduced, the rescued flies went back to a tay1 level of directional scatter. The results are in accord with the idea that the bridge is a central constituent of the visual targeting mechanism. The tay1 mutant was also analysed in other behavioural paradigms. A reduction in walking speed and walking activity in this mutant could be rescued by the expression of UAS-tay under the control of the 007Y-GAL4 driver line, which concomitantly restores the structure of the protocerebral bridge. The separation of bridge functions from functions of other parts of the brain of tay1 was accomplished by rescuing the reduced optomotor compensation in tay1 by the mb247-GAL4>UAS-tay driver. While still having a tay1-like protocerebral bridge, mb247-GAL4 rescue flies are able to compensate at wild-type levels. An intact compensation is not depended on the tay expression in the mushroom bodies, as mushroom body ablated flies with a tay1 background and expression of UAS-tay under the control of mb247-GAL4 show wild-type behaviour as well. The most likely substrate for the function are currently unidentified neurons in the fan-shaped body, that can be stained with 007Y-GAL4 and mb247-GAL4 as well.
Visual information is essential for Drosophila to navigate its environment. The visual system of the fly has been studied for many decades and has yielded many insights about vision in general. However, visual information can be ambiguous and the system processing it needs to be able to cope with that. In this study, the visual orientation behavior of Drosophila is challenged by panoramic incoherent motion stimuli to which the fly can respond in three different, equally adaptive ways. The study is conducted in a well-established setup, the so-called flight simulator (Heisenberg and Wolf, 1993), where the fly can control its visual surroundings in stationary flight with its yaw torque, which is simultaneously recorded. The fly can either use one of two incoherently moving panorama patterns or the integrated motion of both as its reference for straight flight. It is observed that flies use all three of these behavioral alternatives for orientation. Previous models of fly motion vision do not predict a bimodal tuning to incoherent wide-field motion stimuli (Joesch et al., 2008, Borst et al., 1995), however, a recent study on blowflies could suggests that they show component selectivity to the individual moving gratings in a compound plaid stimulus (Saleem et al., 2012). Here, it can be shown that the same bimodal tuning manifests in Drosophila, although the stimuli used are different and most of the experiments are conducted in closed loop. It is found that the extent to which the Drosophila expresses this component selectivity in its orientation behavior, i.e. how often it stabilizes a single panorama pattern instead of the integrated motion of both, depends on two properties of the panorama stimuli, pattern contrast and horizontal pattern element distance. Single pattern stabilization decreases with increasing contrast and increasing pattern element distance. In the latter case, it increases again when there are very few horizontal pattern elements, although that appears to be the result of a lack of rivalry between the patterns due to the low number of pattern elements. Both increased pattern contrast and pattern element distance increase the salience of the single pattern elements. A single element in a compound visual stimulus, like a dot within a dot pattern, can be interpreted as a standalone figure or a part of a bigger unit. Previous studies on Drosophila vision have concentrated on how the fly discriminates a figure from the background (Heisenberg and Wolf, 1984, Bahl et al., 2013, Aptekar et al., 2012), but have hardly touched the question of what qualifies a figure or a background (i.e. a panorama) stimulus as such. In the present study, it is observed that, when exposed to incoherent panoramic motion stimuli, the flies prefer to orient themselves towards the average of the two motions when the panorama stimuli possess strong figure features and towards the single patterns when they do not and single pattern elements are therefore less salient.
The above-mentioned plaid stimuli are a well-known multistable percept in human psychophysics. Multistability is a property of higher visual systems and considered an indicator of endogenous activity in vision. As Drosophila expresses behavioral multistability in the IPMP, it is evaluated in this respect. The results show several parallels to human multistable perception. For one, the frequency and duration with which a behavior occurs, can be influenced, but the occurrence of the behaviors is non-deterministic and not coupled to the stimulus. It can also be shown that the switches between behaviors do not stem from a rivalry of the two visual hemispheres of the fly, although monocularity does also influence the likelihood with which the behaviors occur. Secondly, like in human perceptual rivalry, individual flies exhibit strong idiosyncrasies regarding the overall durations they spend with the different behaviors and the frequencies with which they switch between them. Finally, the distribution of the durations between the behavioral switches can be fit to the same function as the distribution of percept durations in human multistable perception, the gamma function, although it has a different shape and therefore also differing parameters. The Drosophila mutant radish, which has been shown to have attention-like deficits (van Swinderen and Brembs, 2010, Koenig et al., 2016a), does also express an altered behavior in the IPMP compared to wildtype flies. As these behavioral alterations resemble effects on multistable perception found in humans suffering from ADHD (Amador-Campos et al., 2015) and perceptual multistability is generally considered to be closely related to attention (Leopold and Logothetis, 1999), attentional processes are also very likely to play a role in the flies’ behavior in the IPMP.
In conclusion, the visual system of Drosophila is capable disentangle incoherent motion stimuli even if they overlap and cover the entire visual field, i.e. it shows component selectivity of wide-field motion. Whether it uses a single wide-field motion component or the average of two as its reference for straight flight depends on pattern contrast and horizontal pattern element density, which indicates an involvement of a figure-background rivalry. This rivalry and the one between the two wide-field motion components elicit a multistability in the orientation behavior of the fly the temporal dynamics of which partially resemble the temporal dynamics of human multistable perception and which also suggests the involvement of attentional processes.
The development of ethanol tolerance is due to changes in synaptic plasticity. Since the mechanisms mediating synaptic plasticity are probably defective in the mutant hangAE10, it was a goal of the present study to find out how HANG contributes to synaptic plasticity. In particular, it was important to clarify in which neuronal process HANG plays a role. Antibody stainings against HANG revealed that the protein is localized in all neuronal nuclei of larval and adult brains; the staining is absent in hangAE10, thus confirming that this P-element insertion stock is a protein null for HANG. Detailed analysis of the subnuclear distribution of HANG showed that HANG immunoreactivity is enriched at distinct spots in the nucleus in a speckled pattern; these speckles are found at the inside of the nuclear membrane and do not colocalize with chromatin nor with the nucleolus; thus, HANG is probably involved in the stabilization, processing or export of RNAs. As synaptic plasticity can be studied in single neurons at the larval neuromuscular junction, the morphology of the synaptic terminals of hangAE10 mutants was analyzed at muscle 6/7, segment A4. These studies revealed that hangAE10 mutants display a 40 % increase in bouton number and axonal branch length; in addition, some boutons have an abnormal hourglass-like shape, suggesting that they are arrested in a semi-separated state following the initiation of bouton division. The increase in bouton number of hang mutants is mainly due to an increase in numbers of type Ib boutons. The analysis of the distribution of several synaptic markers in hang mutants did not show abnormalities. The presynaptic expression of HANG in hang mutants rescues the increase in bouton number and axonal branch length, thus proving that the phenotypes seen in the P-element insertion hangAE10 are attributable to the lack of HANG rather than to effects of the P-element marker rosy or to a secondary hit on the same chromsome during mutagensis. This finding is further supported by the fact that postsynaptic expression of HANG does not rescue the abnormal NMJ morphology of hangAE10. Alterations in cAMP levels regulate the number of boutons; since hang mutants display an increase in bouton number, the questions was whether this morphological abnormality was due to defects in cAMP signalling. To test this hypothesis, hangAE10 NMJs were compared to those of the hypomorphic allele dnc1 that has a defective cAMP cascade. Some aspects of the NMJ phenotype (e.g. the increase in bouton number and the unaltered ratio of active zones per bouton area) are similar in hangAE10 and dnc1, other differ. Expression of a UAS-dnc transgene in hangAE10 mutants does not modify the phenotype. In summary, the results of this study indicate that nuclear protein HANG might be involved in isoform-specific splicing of genes required for synaptic plasticity at the NMJ.
During my PhD I studied two principal biological aspects employing Drosophila melanogaster. Therefore, this study is divided into Part I and II.
Part I: Bruchpilot and Complexin interact to regulate synaptic vesicle tethering to the
active zone cytomatrix
At the presynaptic active zone (AZ) synaptic vesicles (SVs) are often physically linked to an electron-dense cytomatrix – a process referred to as “SV tethering”. This process serves to concentrate SVs in close proximity to their release sites before contacting the SNARE complex for subsequent fusion (Hallermann and Silver, 2013). In Drosophila, the AZ protein Bruchpilot (BRP) is part of the proteinous cytomatrix at which SVs accumulate (Kittel et al., 2006b; Wagh et al., 2006; Fouquet et al., 2009). Intriguingly, truncation of only 1% of the C-terminal region of BRP results in a severe defect in SV tethering to this AZ scaffold (hence named brpnude; Hallermann et al., 2010b).
Consistent with these findings, cell-specific overexpression of a C-terminal BRP fragment, named mBRPC-tip (corresponds to 1% absent in brpnude; m = mobile) phenocopied the brpnude mutant in behavioral and functional experiments. These data indicate that mBRPC-tip suffices to saturate putative SV binding sites, which induced a functional tethering deficit at motoneuronal AZs. However, the molecular identity of the BRP complement to tether SVs to the presynaptic AZ scaffold remains unknown. Moreover, within larval motoneurons membrane-attached C-terminal portions of BRP were sufficient to tether SVs to sites outside of the AZ. Based on this finding a genetic screen was designed to identify BRP interactors in vivo. This screen identified Complexin (CPX), which is known to inhibit spontaneous SV fusion and to enhance stimulus evoked SV release (Huntwork and Littleton, 2007; Cho et al., 2010; Martin et al., 2011). However, so far CPX has not been associated with a function upstream of priming/docking and release of SVs. This work provides morphological and functional evidence, which suggests that CPX promotes recruitment of SVs to the AZ and thereby curtails synaptic short-term depression. Together, the presented findings indicate a functional interaction between BRP and CPX at Drosophila AZs.
Part II: The Adhesion-GPCR Latrophilin/CIRL shapes mechanosensation
The calcium independent receptor of α-latrotoxin (CIRL), also named Latrophilin, represents a prototypic Adhesion class G-protein coupled-receptor (aGPCR). Initially, Latrophilin was identified based on its capacity to bind the α-component of latrotoxin (α-LTX; Davletov et al., 1996; Krasnoperov et al., 1996), which triggers massive exocytotic activity from neurons of the peripheral nervous system (Scheer et al., 1984; Umbach et al., 1998; Orlova et al., 2000). As a result Latrophilin is considered to play a role in synaptic transmission. Later on, Latrophilins have been associated with other biological processes including tissue polarity (Langenhan et al., 2009), fertility (Prömel et al., 2012) and synaptogenesis (Silva et al., 2011). However, thus far its subcellular localization and the identity of endogenous ligands, two aspects crucial for the comprehension of Latrophilin’s in vivo function, remain enigmatic.
Drosophila contains only one latrophilin homolog, named dCirl, whose function has not been investigated thus far.
This study demonstrates abundant dCirl expression throughout the nervous system of Drosophila larvae. dCirlKO animals are viable and display no defects in development and neuronal differentiation. However, dCirl appears to influence the dimension of the postsynaptic sub-synaptic reticulum (SSR), which was accompanied by an increase in the postsynaptic Discs-large abundance (DLG). In contrast, morphological and functional properties of presynaptic motoneurons were not compromised by the removal of dCirl. Instead, dCirl is required for the perception of mechanical challenges (acoustic-, tactile- and proprioceptive stimuli) through specialized mechanosensory devices, chordotonal organs (Eberl, 1999). The data indicate that dCirl modulates the sensitivity of chordotonal neurons towards mechanical stimulation and thereby adjusts their input-output relation. Genetic interaction analyses suggest that adaption of the molecular mechanotransduction machinery by dCirl may underlie this process. Together, these results uncover an unexpected function of Latrophilin/dCIRL in mechanosensation and imply general modulatory roles of aGPCR in mechanoception.
Neuroanatomical data in fly brain research are mostly available as spatial gene expression patterns of genetically distinct fly strains. The Drosophila standard brain, which was developed in the past to provide a reference coordinate system, can be used to integrate these data. Working with the standard brain requires advanced image processing methods, including visualisation, segmentation and registration. The previously published VIB Protocol addressed the problem of image registration. Unfortunately, its usage was severely limited by the necessity of manually labelling a predefined set of neuropils in the brain images at hand. In this work I present novel tools to facilitate the work with the Drosophila standard brain. These tools are integrated in a well-known open-source image processing framework which can potentially serve as a common platform for image analysis in the neuroanatomical research community: ImageJ. In particular, a hardware-accelerated 3D visualisation framework was developed for ImageJ which extends its limited 3D visualisation capabilities. It is used for the development of a novel semi-automatic segmentation method, which implements automatic surface growing based on user-provided seed points. Template surfaces, incorporated with a modified variant of an active surface model, complement the segmentation. An automatic nonrigid warping algorithm is applied, based on point correspondences established through the extracted surfaces. Finally, I show how the individual steps can be fully automated, and demonstrate its application for the successful registration of fly brain images. The new tools are freely available as ImageJ plugins. I compare the results obtained by the introduced methods with the output of the VIB Protocol and conclude that our methods reduce the required effort five to ten fold. Furthermore, reproducibility and accuracy are enhanced using the proposed tools.
The change of day and night is one of the challenges all organisms are exposed to, as they have to adjust their physiology and behavior in an appropriate way. Therefore so called circadian clocks have evolved, which allow the organism to predict these cyclic changes of day and night. The underlying molecular mechanism is oscillating with its endogenous period of approximately 24 hours in constant conditions, but as soon as external stimuli, so called Zeitgebers, are present, the clocks adjust their period to exactly 24h, which is called entrainment. Studies in several species, including humans, animals and plants, showed that light is the most important Zeitgeber synchronizing physiology and behavior to the changes of day and night. Nevertheless also other stimuli, like changes in temperature, humidity or social interactions, are powerful Zeitgebers for entraining the clock. This thesis will focus on the question, how light influences the locomotor behavior of the fly in general, including a particular interest on the entrainment of the circadian clock. As a model organism Drosophila melanogaster was used.
During the last years several research groups investigated the effect of light on the circadian clock and their results showed that several light input pathways to the clock contribute to wild-type behavior. Most of the studies focused on the photopigment Cryptochrome (CRY) which is expressed in about half of the 150 clock neurons in the fly. CRY is activated by light, degrades the clock protein Timeless (TIM) and hence entrains the clock to the light-dark (LD)-cycle resulting from changes of day and night. However, also flies lacking CRY are still able to entrain their clock mechanism as well as their activity-rest-rhythm to LD-cycles, clearly showing that the visual system of the fly also contributes to clock synchronization. The mechanism how light information from the visual system is transferred to the clock is so far still unknown. This is also true for so-called masking-effects which are changes in the behavior of the animal that are directly initiated by external stimuli and therefore independent of the circadian clock. These effects complement the behavior of the animals as they enable the fly to react quickly to changes in the environment even during the clock-controlled rest state.
Both of these behavioral features were analyzed in more detail in this study. On the one hand, we investigated the influence of the compound eyes on the entrainment of the clock neurons and on the other hand, we tried to separate clock-controlled behavior from masking. To do so "nature-like" light conditions were simulated allowing the investigation of masking and entrainment within one experiment. The simulation of moonlight and twilight conditions caused significant changes in the locomotor behavior. Moonlit nights increased nocturnal activity levels and shifted the morning (M) and evening (E) activity bouts into the night. The opposite was true for the investigation of twilight, as the activity bouts were shifted into the day. The simulation of twilight and moonlight within the same experiment further showed that twilight appears to dominate over moonlight, which is in accordance to the assumption that twilight in nature is one of the key signals to synchronize the clock as the light intensity during early dawn rises similarly in every season. By investigating different mutants with impaired visual system we showed that the compound eyes are essential for the observed behavioral adaptations. The inner receptor cells (R7 and R8) are important for synchronizing the endogenous clock mechanism to the changes of day and night. In terms of masking, a complex interaction of all receptor cells seems to adjust the behavioral pattern, as only flies lacking photopigments in inner and outer receptor cells lacked all masking effects. However, not only the compound eyes seem to contribute to rhythmic activity in moonlit nights. CRY-mutant flies shift their E activity bout even more into the night than wild-type flies do. By applying Drosophila genetics we were able to narrow down this effect to only four CRY expressing clock neurons per hemisphere. This implies that the compound eyes and CRY in the clock neurons have antagonistic effects on the timing of the E activity bout. CRY advances activity into the day, whereas the compound eyes delay it. Therefore, wild-type behavior combines both effects and the two light inputs might enable the fly to time its activity to the appropriate time of day.
But CRY expression is not restricted to the clock neurons as a previous study showed a rather broad distribution within the compound eyes. In order to investigate its function in the eyes we collaborated with Prof. Rodolfo Costa (University of Padova). In our first study we were able to show that CRY interacts with the phototransduction cascade and thereby influences visual behavior like phototaxis and optomotor response. Our second study showed that CRY in the eyes affects locomotor activity rhythms. It appears to contribute to light sensation without being a photopigment per se. Our results rather indicate that CRY keeps the components of the phototransduction cascade close to the cytoskeleton, as we identified a CRY-Actin interaction in vitro. It might therefore facilitate the transformation of light energy into electric signals.
In a further collaboration with Prof. Orie Shafer (University of Michigan) we were able to shed light on the significance of the extraretinal Hofbauer-Buchner eyelet for clock synchronization. Excitation of the eyelet leads to Ca2+ and cAMP increases in specific clock neurons, consequently resulting in a shift of the flies´ rhythmic activity.
Taken together, the experiments conducted in this thesis revealed new functions of different eye structures and CRY for fly behavior. We were furthermore able to show that masking complements the rhythmic behavior of the fly, which might help to adapt to natural conditions.
Is behaviour response or action? In this Thesis I study this question regarding a rather simple organism, the larva of the fruit fly Drosophila melanogaster. Despite its numerically simple brain and limited behavioural repertoire, it is nevertheless capable to accomplish surprisingly complex tasks. After association of an odour and a rewarding or punishing reinforcement signal, the learnt odour is able to retrieve the formed memory trace. However, the activated memory trace is not automatically turned into learned behaviour: Appetitive memory traces are behaviourally expressed only in absence of the rewarding tastant whereas aversive memory traces are behaviourally expressed in the presence of the punishing tastant. The ‘decision’ whether to behaviourally express a memory trace or not relies on a quantitive comparison between memory trace and current situation: only if the memory trace (after odour-sugar training) predicts a stronger sugar reward than currently present, animals show appetitive conditioned behaviour. Learned appetitive behaviour is best seen as active search for food – being pointless in the presence of (enough) food. Learned aversive behaviour, in turn, can be seen as escape from a punishment – being pointless in absence of punishment. Importantly, appetitive and aversive memory traces can be formed and retrieved independent from each other but also can, under appriate circumstances, summate to jointly organise conditioned behaviour. In contrast to learned behaviour, innate olfactory behaviour is not influenced by gustatory processing and vice versa. Thus, innate olfactory and gustatory behaviour is rather rigid and reflexive in nature, being executed almost regardless of other environmental cues. I suggest a behavioural circuit-model of chemosensory behaviour and the ‘decision’ process whether to behaviourally express a memory trace or not. This model reflects known components of the larval chemobehavioural circuit and provides clear hypotheses about the kinds of architecture to look for in the currently unknown parts of this circuit. The second chapter deals with gustatory perception and processing (especially of bitter substances). Quinine, the bitter tastant in tonic water and bitter lemon, is aversive for larvae, suppresses feeding behaviour and can act as aversive reinforcer in learning experiments. However, all three examined behaviours differ in their dose-effect dynamics, suggesting different molecular and cellular processing streams at some level. Innate choice behaviour, thought to be relatively reflexive and hard-wired, nevertheless can be influenced by the gustatory context. That is, attraction toward sweet tastants is decreased in presence of bitter tastants. The extent of this inhibitory effect depends on the concentration of both sweet and bitter tastant. Importantly, sweet tastants differ in their sensitivity to bitter interference, indicating a stimulus-specific mechanism. The molecular and cellular processes underlying the inhibitory effect of bitter tastants are unknown, but the behavioural results presented here provide a framework to further investigate interactions of gustatory processing streams.
In this thesis, I introduce the Virtual Brain Protocol, which facilitates applications of the Standard Brain of Drosophila melanogaster. By providing reliable and extensible tools for the handling of neuroanatomical data, this protocol simplifies and organizes the recurring tasks involved in these applications. It is demonstrated that this protocol can also be used to generate average brains, i.e. to combine recordings of several brains with the same features such that the common features are emphasized. One of the most important steps of the Virtual Insect Protocol is the aligning of newly recorded data sets with the Standard Brain. After presenting methods commonly applied in a biological or medical context to align two different recordings, it is evaluated to what extent this alignment can be automated. To that end, existing Image Processing techniques are assessed. I demonstrate that these techniques do not satisfy the requirements needed to guarantee sensible alignments between two brains. Then, I analyze what needs to be taken into account in order to formulate an algorithm which satisfies the needs of the protocol. In the last chapter, I derive such an algorithm using methods from Information Theory, which bases the technique on a solid mathematical foundation. I show how Bayesian Inference can be applied to enhance the results further. It is demonstrated that this approach yields good results on very noisy images, detecting apparent boundaries between structures. The same approach can be extended to take additional knowledge into account, e.g. the relative position of the anatomical structures and their shape. It is shown how this extension can be utilized to segment a newly recorded brain automatically.