Medizinische Universität Graz Austria/Österreich - Forschungsportal - Medical University of Graz
Gewählte Publikation:
Wernitznig, S.
Studying a collision sensing pathway in the locust visual system so that it can be modelled and adapted for a device that helps blind or visually impaired persons
Doktoratsstudium der Medizinischen Wissenschaft; Humanmedizin; [ Dissertation ] Graz Medical University; 2016. pp. 97
[OPEN ACCESS]
FullText
- Autor*innen der Med Uni Graz:
- Wernitznig Stefan
- Betreuer*innen:
- Leitinger Gerd
- Altmetrics:
- Abstract:
- A particular neuronal circuit in the locust’s optic lobe helps the animal avoid collisions. Although this circuit has been investigated for almost three decades, it still holds some secrets. The work of this thesis focuses on neurons which are part of this circuit and are located in the three neuropils of the optic lobe, the lamina, the medulla, and the lobula complex. Main objective was to describe their morphology and synaptology, in order to lead to an expansion of the known connectivity within the neuronal circuit. Furthermore, electrophysiological experiments were conducted with the aim of characterising inhibitory connections within the circuit. Inhibition had been assumed as a vital part of the circuit’s function when a model of the neuronal connectivity within the circuit had been made. To study the neurons’ morphology and synaptology a novel electron microscopy technique was used for data acquisition. This so-called serial block-face scanning electron microscopy (SBEM) enables the production of several thousand serial sections from which the branches and synapses of the neurons are 3D-reconstructed. A special protocol to increase the contrast in the samples was adapted and modified. In order to carry out the electrophysiological experiments a stimulation device was designed for single ommatidia stimulations. By that it was possible to present different stimulation patterns onto a small field of the animal’s compound eye. The responses to the stimuli were recorded extracellularly from a neuron in the thoracic nerve cord, the descending contralateral movement detector (DCMD), which is postsynaptic to the lobula giant movement detector (LGMD). The LGMD receives its inputs from afferent neurons, sums them up and when it becomes excited, the LGMD spikes are transmitted 1-to-1 onto the DCMD. Two species of grasshoppers were used for the investigations. In adult Schistocerca gregaria the first neuropil, or lamina was examined and the two so-called lamina monopolar neurons (L1 and L2) reconstructed. These are the first interneurons in the circuit and they receive their input from the photoreceptor cells (R cells) from the retina. The particular finding was that approximately 40% of the input synapses onto the L neurons were tetrads which were previously thought to be only present in flies. In fourth instar Locusta migratoria, neurons, which connect the two neuropils, the medulla and lobula complex, were investigated. Main focus was on the trans-medullary afferent (TmA) neurons. These interneurons are thought to be a crucial part in the collision sensing pathway as they give direct input the LGMD in the lobula complex. The reconstruction of two TmA neurons showed that there are possible other neurons within the medulla associated with the neuronal circuit. The TmA neurons were not only postsynaptic in the medulla but also presynaptic, giving evidence that there is not only a downstream pathway between the eye and the lobula complex, but also feedback towards the eye. This evidence indicates an additional computation step within the medulla, which was not considered in the model. The electrophysiological experiments, conducted in adult L. migratoria showed the effect of single ommatidia stimulation. By presenting OFF stimuli to just 30 ommatidia an inhibition of the neighbouring ommatidia was observed. The effect was strongest when the stimuli were approximately 100 ms apart. Both, the results of morphology and physiology gave evidence that the proposed model of the collision sensing pathway in the locust is based on correct assumptions. There is a clear inhibitory effect observable within the system. This effect is dependent on the speed of the spreading stimulus. Concerning the information computation, the model might have to be enhanced as there is strong evidence of additional involved neurons in the medulla.