Universitätsbibliothek

Cataglyphis ants are famous for their navigational abilities. They live in hostile habitats where they forage as solitary scavengers covering distances of more than hundred thousand times their body lengths. To return to their nest with a prey item – mainly other dead insects that did not survive the heat – Cataglyphis ants constantly keep track of their directions and distances travelled. The navigational strategy is called path integration, and it enables an ant to return to the nest in a straight line using its home vector. Cataglyphis ants mainly rely on celestial compass cues, like the position of the sun or the UV polarization pattern, to determine directions, and they use an idiothetic step counter and optic flow to measure distances. In addition, they acquire information about visual, olfactory and tactile landmarks, and the wind direction to increase their chances of returning to the nest safe and sound. Cataglyphis’ navigational performance becomes even more impressive if one considers their life style. Most time of their lives, the ants stay underground and perform tasks within the colony. When they start their foraging careers outside the nest, they have to calibrate their compass systems and acquire all information necessary for navigation during subsequent foraging. This navigational toolkit is not instantaneously available, but has to be filled with experience. For that reason, Cataglyphis ants perform a striking behavior for up to three days before actually foraging. These so-called learning walks are crucial for the success as foragers later on. In the present thesis, both the ontogeny and the fine-structure of learning walks has been investigated. Here I show with displacement experiments that Cataglyphis ants need enough space and enough time to perform learning walks. Spatially restricted novices, i. e. naïve ants, could not find back to the nest when tested as foragers later on. Furthermore, ants have to perform several learning walks over 1-3 days to gain landmark information for successful homing as foragers. An increasing number of feeder visits also increases the importance of landmark information, whereas in the beginning ants fully rely on their path-integration vector. Learning walks are well-structured. High-speed video analysis revealed that Cataglyphis ants include species-specific rotational elements in their learning walks. Greek Cataglyphis ants (C. noda and C. aenescens) inhabiting a cluttered pine forest perform voltes, small walked circles, and pirouettes, tight turns about the body axis with frequent stopping phases. During the longest stopping phases, the ants gaze back to their nest entrance. The Tunisian Cataglyphis fortis ants inhabiting featureless saltpans only perform voltes without directed gazes. The function of voltes has not yet been revealed. In contrast, the fine structure of pirouettes suggests that the ants take snapshots of the panorama towards their homing direction to memorize the nest’s surroundings. The most likely hypothesis was that Cataglyphis ants align the gaze directions using their path integrator, which gets directional input from celestial cues during foraging. To test this hypothesis, a manipulation experiment was performed changing the celestial cues above the nest entrance (no sun, no natural polarization pattern, no UV light). The accurately directed gazes to the nest entrance offer an easily quantifiable readout suitable to ask the ants where they expect their nest entrance. Unexpectedly, all novices performing learning walks under artificial sky conditions looked back to the nest entrance. This was especially surprising, because neuronal changes in the mushroom bodies and the central complex receiving visual input could only be induced with the natural sky when comparing test animals with interior workers. The behavioral findings indicated that Cataglyphis ants use another directional reference system to align their gaze directions during the longest stopping phases of learning walk pirouettes. One possibility was the earth’s magnetic field. Indeed, already disarraying the geomagnetic field at the nest entrance with an electromagnetic flat coil indicated that the ants use magnetic information to align their looks back to the nest entrance. To investigate this finding further, ants were confronted with a controlled magnetic field using a Helmholtz coil. Elimination of the horizontal field component led to undirected gaze directions like the disarray did. Rotating the magnetic field about 90°, 180° or -90° shifted the ants’ gaze directions in a predictable manner. Therefore, the earth’s magnetic field is a necessary and sufficient reference system for aligning nest-centered gazes during learning-walk pirouettes. Whether it is additionally used for other navigational purposes, e. g. for calibrating the solar ephemeris, remains to be tested. Maybe the voltes performed by all Cataglyphis ant species investigated so far can help to answer this question..

The olfactory sense is of utmost importance for honeybees, Apis mellifera. Honeybees use olfaction for communication within the hive, for the identification of nest mates and non-nest mates, the localization of food sources, and in case of drones (males), for the detection of the queen and mating. Honeybees, therefore, can serve as excellent model systems for an integrative analysis of an elaborated olfactory system. To efficiently filter odorants out of the air with their antennae, honeybees possess a multitude of sensilla that contain the olfactory sensory neurons (OSN). Three types of olfactory sensilla are known from honeybee worker antennae: Sensilla trichoidea, Sensilla basiconica and Sensilla placodea. In the sensilla, odorant receptors that are located in the dendritic arborizations of the OSNs transduce the odorant information into electrical information. Approximately 60.000 OSN axons project in two parallel bundles along the antenna into the brain. Before they enter the primary olfactory brain center, the antennal lobe (AL), they diverge into four distinct tracts (T1-T4). OSNs relay onto ~3.000-4.000 local interneurons (LN) and ~900 projection neurons (PN), the output neurons of the AL. The axons of the OSNs together with neurites from LNs and PNs form spheroidal neuropil units, the so-called glomeruli. OSN axons from the four AL input tracts (T1-T4) project into four glomerular clusters. LNs interconnect the AL glomeruli, whereas PNs relay the information to the next brain centers, the mushroom body (MB) - associated with sensory integration, learning and memory - and the lateral horn (LH). In honeybees, PNs project to the MBs and the LH via two separate tracts, the medial and the lateral antennal-lobe tract (m/lALT) which run in parallel in opposing directions. The mALT runs first to the MB and then to the LH, the lALT runs first to the LH and then to the MB. This dual olfactory pathway represents a feature unique to Hymenoptera. Interestingly, both tracts were shown to process information about similar sets of odorants by extracting different features. Individual mALT PNs are more odor specific than lALT PNs. On the other hand, lALT PNs have higher spontaneous and higher odor response action potential (AP) frequencies than mALT PNs. In the MBs, PNs form synapses with ~184.000 Kenyon cells (KC), which are the MB intrinsic neurons. KCs, in contrast to PNs, show almost no spontaneous activity and employ a spatially and temporally sparse code for odor coding. In manuscript I of my thesis, I investigated whether the differences in specificity of odor responses between m- and lALT are due to differences in the synaptic input. Therefore, I investigated the axonal projection patterns of OSNs housed in S. basiconica in honeybee workers and compared them with S. trichoidea and S. placodea using selective anterograde labeling with fluorescent tracers and confocal- microscopy analyses of axonal projections in AL glomeruli. Axons of S. basiconica-associated OSNs preferentially projected into the T3 input-tract cluster in the AL, whereas the two other types of sensilla did not show a preference for a specific glomerular cluster. T3- associated glomeruli had previously been shown to be innervated by mALT PNs. Interestingly, S. basiconica as well as a number of T3 glomeruli lack in drones. Therefore I set out to determine whether this was associated with the reduction of glomeruli innervated by mALT PNs. Retrograde tracing of mALT PNs in drones and counting of innervated glomeruli showed that the number of mALT-associated glomeruli was strongly reduced in drones compared to workers. The preferential projections of S. basiconica-associated OSNs into T3 glomeruli in female workers together with the reduction of mALT-associated glomeruli in drones support the presence of a female-specific olfactory subsystem that is partly innervated by OSNs from S. basiconica and is associated with mALT projection neurons. As mALT PNs were shown to be more odor specific, I suppose that already the OSNs in this subsystem are more odor specific than lALT associated OSNs. I conclude that this female-specific subsystem allows the worker honeybees to respond adequately to the enormous variety of odorants they experience during their lifetime. In manuscript II, I investigated the ion channel composition of mALT and lALT PNs and KCs in situ. This approach represents the first study dealing with the honeybee PN and KC ion channel composition under standard conditions in an intact brain preparation. With these recordings I set out to investigate the potential impact of intrinsic neuronal properties on the differences between m- and lALT PNs and on the sparse odor coding properties of KCs. In PNs, I identified a set of Na+ currents and diverse K+ currents depending on voltage and Na+ or Ca2+ that support relatively high spontaneous and odor response AP frequencies. This set of currents did not significantly differ between mALT and lALT PNs, but targets for potential modulation of currents leading to differences in AP frequencies were found between both types of PNs. In contrast to PNs, KCs have very prominent K+ currents, which are likely to contribute to the sparse response fashion observed in KCs. Furthermore, Ca2+ dependent K+ currents were found, which may be of importance for coincidence detection, learning and memory formation. Finally, I conclude that the differences in odor specificity between m- and lALT PNs are due to their synaptic input from different sets of OSNs and potential processing by LNs. The differences in spontaneous activity between the two tracts may be caused by different neuronal modulation or, in addition, also by interaction with LNs. The temporally sparse representation of odors in KCs is very likely based on the intrinsic KC properties, whereas general excitability and spatial sparseness are likely to be regulated through GABAergic feedback neurons.

African trypanosomiasis is a disease endemic to sub-Saharan Africa. It affects humans as well as wild and domestic animals. The human form of the disease is known as sleeping sickness and the animal form as nagana, which are usually fatal if left untreated. The cause of African trypanosomiasis is the unicellular parasite Trypanosoma brucei. During its life cycle, Trypanosoma brucei shuttles between a mammalian host and the tsetse fly vector. In the mammalian host the parasite multiplies as bloodstream form (BSF) extracellularly in the bloodstream or the lymphatic system. Survival of BSF parasites relies on immune evasion by antigenic variation of surface proteins because its extracellular lifestyle leads to direct exposure to immune responses. At any given time each BSF cell expresses a single type of variant surface glycoprotein (VSG) on its surface from a large repertoire. The active VSG is transcribed from one of 15 specialized subtelomeric domains, termed bloodstream expression sites (BESs). The remaining 14 BESs are silenced. This monoallelic expression and periodic switching of the expressed VSG enables to escape the immune response and to establish a persistent infection in the mammalian host. During developmental differentiation from BSF to the insect vector-resident procyclic form (PCF), the active BES is transcriptionally silenced to stop VSG transcription. Thus, all 15 BESs are inactive in the PCF cells as surface protein expression is developmentally regulated. Previous reports have shown that the telomere complex components TbTRF, TbRAP1 and TbTIF2 are involved in VSG transcriptional regulation. However, the precise nature of their contribution remains unclear. In addition, no information is available about the role of telomeres in the initiation and regulation of developmental BES silencing. To gain insights into the regulatory mechanisms of telomeres on VSG transcription and developmental repression it is therefore essential to identify the complete composition of the trypanosome telomere complex. To this end, we used two complementary biochemical approaches and quantitative label-free interactomics to determine the composition of telomere protein complexes in T. brucei. Firstly, using a telomeric pull-down assay we found 17 potential telomere-binding proteins including the known telomere-binding proteins TbTRF and TbTIF2. Secondly, by performing a co-immunoprecipitation experiment to elucidate TbTRF interactions we co-purified five proteins. All of these five proteins were also enriched with telomeric DNA in the pull-down assay. To validate these data, I characterized one of the proteins found in both experiments (TelBP1). In BSF cells, TelBP1 co-localizes with TbTRF and interacts with already described telomere-binding proteins such as TbTRF, TbTIF2 and TbRAP1 indicating that TelBP1 is a novel component of the telomere complex in trypanosomes. Interestingly, protein interaction studies in PCF cells suggested a different telomere complex composition compared to BSF cells. In contrast to known members of the telomere complex, TelBP1 is dispensable for cell viability indicating that its function might be uncoupled from the known telomere-binding proteins. Overexpression of TelBP1 had also no effect on cell viability, but led to the discovery of two additional shorter isoforms of TelBP1. However, their source and function remained elusive. Although TelBP1 is not essential for cell viability, western blot analysis revealed a 4-fold upregulation of TelBP1 in the BSF stage compared to the PCF stage supporting the concept of a dynamic telomere complex composition. We observed that TelBP1 influences the kinetics of transcriptional BES silencing during developmental transition from BSF to PCF. Deletion of TelBP1 caused faster BES silencing compared to wild-type parasites. Taken together, TelBP1 function illustrates that developmental BES silencing is a fine-tuned process, which involves stage-specific changes in telomere complex formation.

The honeybee Apis mellifera is a social insect well known for its complex behavior and the ability to learn tasks associated with central place foraging, such as visual navigation or to learn and remember odor-reward associations. Although its brain is smaller than 1mm² with only 8.2 x 105 neurons compared to ~ 20 x 109 in humans, bees still show amazing social, cognitive and learning skills. They express an age – related division of labor with nurse bees staying inside the hive and performing tasks like caring for the brood or cleaning, and foragers who collect food and water outside the hive. This challenges foragers with new responsibilities like sophisticated navigation skills to find and remember food sources, drastic changes in the sensory environment and to communicate new information to other bees. Associated with this plasticity of the behavior, the brain and especially the mushroom bodies (MBs) - sensory integration and association centers involved in learning and memory formation – undergo massive structural and functional neuronal alterations. Related to this background my thesis on one hand focuses on neuronal plasticity and underlying molecular mechanisms in the MBs that accompany the nurse – forager transition. In the first part I investigated an endogenous and an internal factor that may contribute to the nurse - forager phenotype plasticity and the correlating changes in neuronal network in the MBs: sensory exposure (light) and juvenile hormone (JH). Young bees were precociously exposed to light and subsequently synaptic complexes (microglomeruli, MG) in the MBs or respectively hemolymph juvenile hormone (JH) levels were quantified. The results show that light input indeed triggered a significant decrease in MG density, and mass spectrometry JH detection revealed an increase in JH titer. Interestingly light stimulation in young bees (presumably nurse bees) triggered changes in MG density and JH levels comparable to natural foragers. This indicates that both sensory stimuli as well as the endocrine system may play a part in preparing bees for the behavioral transition to foraging. Considering a connection between the JH levels and synaptic remodeling I used gene knockdown to disturb JH pathways and artificially increase the JH level. Even though the knockdown was successful, the results show that MG densities remained unchanged, showing no direct effect of JH on synaptic restructuring. To find a potential mediator of structural synaptic plasticity I focused on the calcium-calmodulin-dependent protein kinase II (CaMKII) in the second part of my thesis. CaMKII is a protein known to be involved in neuronal and behavioral plasticity and also plays an important part in structural plasticity reorganizing synapses. Therefore it is an interesting candidate for molecular mechanisms underlying MG reorganization in the MBs in the honeybee. Corresponding to the high abundance of CaMKII in the learning center in vertebrates (hippocampus), CaMKII was shown to be enriched in the MBs of the honeybee. Here I first investigated the function of CaMKII in learning and memory formation as from vertebrate work CaMKII is known to be associated with the strengthening of synaptic connections inducing long term potentiation and memory formation. The experimental approach included manipulating CaMKII function using 2 different inhibitors and a specific siRNA to create a CaMKII knockdown phenotype. Afterwards bees were subjected to classical olfactory conditioning which is known to induce stable long-term memory. All bees showed normal learning curves and an intact memory acquisition, short-term and mid-term memory (1 hour retention). However, in all cases long-term memory formation was significantly disrupted (24 and 72 hour retention). These results suggests the necessity of functional CaMKII in the MBs for the induction of both early and late phases of long-term memory in honeybees. The neuronal and molecular bases underlying long-term memory and the resulting plasticity in behavior is key to understanding higher brain function and phenotype plasticity. In this context CaMKII may be an important mediator inducing structural synaptic and neuronal changes in the MB synaptic network.

Wasps of the genus Polistes comprise over 200 species and are nearly cosmopolitan. They show a lack of physiological caste differentiation and are therefore considered as primitively eusocial. Furthermore, paper wasps are placed between the solitary living Eumenidae and the highly social organized Vespinae. Hence, they are often called a “key genus” for understanding the evolution of sociality. Particularly, Polistes dominula, with its small easy manageable nests and its frequent occurrence and wide distribution range is often the subject of studies. In Europe, the invasion of this species into northern regions is on the rise. Since little was known about the nesting behaviour of P. dominula in Central Europe, the basic principles about nesting were investigated in Würzburg, Germany (latitude 49°) by conducting a comprehensive field-study spanning three consecutive years. Furthermore, the thermoregulation of individual wasps in their natural habitat had not yet been investigated in detail. Therefore, their ability to respond to external hazards with elevated thorax temperatures was tested. In addition, different types of nest thermoregulation were investigated using modern methods such as infrared thermography and temperature data logger. In the present work, the investigation of basic nesting principles revealed that foundress groups (1-4 foundresses) and nests are smaller and that the nesting season is shorter in the Würzburg area than in other regions. The mean size of newly founded nests was 83 cells and the average nesting season was around 4.6 months. The queens neither preferred single (54%) nor multiple founding (46%) in this study. The major benefit of multiple founding is an increased rate of survival. During the three years of observation, only 47% of single-foundress colonies survived, whereas 100% of colonies that were built by more than two queens, survived. However, an influence of the number of foundresses on the productivity of colonies in terms of number of cells and pupae per nest has not shown up. However, the length of the nesting season as well as the nest sizes varied strongly depending on the climatic conditions of the preceding winter during the three consecutive years. In order to investigate the thermoregulatory mechanisms of individual adult P. dominula wasps, I presented artificial threats by applying smoke or carbon dioxide simulating fire and predator attacks, respectively, and monitored the thorax temperature of wasps on the nest using infrared thermography. The results clearly revealed that P. dominula workers recognized smoke and CO2 and reacted almost instantaneously and simultaneously with an increase of their thorax temperature. The maximal thorax temperature was reached about 65 s after the application of both stressors, but subsequently the wasps showed a different behaviour pattern. They responded to a longer application of smoke with moving to the exit and fled, whereas in case of CO2 the wasps started flying and circling the nest without trying to escape. No rise of the thorax temperature was detectable after an air blast was applied or in wasps resting on the nest. Additionally, the thorax temperatures of queens were investigated during dominance battles. I found that the thorax temperature of the dominant queens rose up to 5°C compared to that of subordinate queens that attacked the former. The study of active mechanisms for nest thermoregulation revealed no brood incubation or clustering behaviour of P. dominula. Furthermore, I found out that wing fanning for cooling the nest was almost undetectable (4 documented cases). However, I could convincingly record that water evaporation is most effective for nest cooling. By the direct comparison of active (with brood and adults) and non-active (without brood and adults) nests, the start of cooling by water evaporation was detected above maximum outside temperatures of 25°C or at nest temperatures above 35°C. The powerful role of water in nest cooling was manifested by an average decrease of temperature of a single cell of about 8°C and a mean duration of 7 min until the cell reached again its initial temperature. The investigation of passive thermoregulatory mechanisms revealed that the nest site choice as well as nest orientation appears to be essential for P. dominula wasps. Furthermore, I was able to show that the architecture of the nests plays an important role. Based on the presented results, it can be assumed that the vertical orientation of cells helps maintaining the warmth of nests during the night, whereas the pedicel assists in cooling the nest during the day.