Abstract

As a fly flies around in the world the visual scene moves constantly across its eyes. Depending on its path, this elicits a particular large-field motion pattern called ‘flow field’. Since the flow-fields are characteristic for particular flight trajectories they can be used to guide behavior, in particular to control the course of the fly. In the blowfly, these visual motion cues are mediated by a set of 60 motion-sensitive neurons called lobula plate tangential cells (LPTCs). The directionally selective response of the LPTCs has been ascribed to the integration of local motion information across their extensive dendritic trees. As the lobula plate is organized retinotopically the receptive fields of the tangential cells ought to be determined by their dendritic architecture. This appears not always to be the case. Recent experiments have revealed many lateral connections among tangential cells that appear to mediate their often complex receptive fields. Here single cells were ablated in order to determine which lateral connections are functionally important. I found that the ablation of a single cell, or class of cells revealed that the lateral connections among LPTCs can be the source of their local motion input, or augment the feedfoward input from local motion elements through either dendro-dendritic and axonal-axonal connections. Other connections between LPTCs were found to have no discernable functional significance and suggest that the lobula plate circuitry is yet to be fully revealed. The specific projects are outlined below. Input Circuitry to the HS- and CH-cells A single class of the lobula plate tangential cells the CH (centrifugal horizontal) neurons, play an important role in two pathways: figure-ground discrimination and flow-field selectivity. As was recently found, the dendrites of CH-cells are electrically coupled to the dendritic tree of another class of neurons sensitive to horizontal image motion, the horizontal system cells (HS). However, whether motion information arrives independently at both of these cells or is passed from one to the other is not known. Here I examine the ipsilateral input circuitry to HS and CH neurons by selective laser ablation of individual interneurons. I find that the response of CH neurons to motion presented in front of the ipsilateral eye is entirely abolished after the ablation of HS-cells. In contrast the motion response of HS-cells persists after the ablation of CH-cells. I conclude that HS-cells receive direct motion input from local motion elements, whereas CH-cells do not; their motion response is driven by HS-cells. This connection scheme is discussed with reference as to how the dendritic networks involved in figure-ground detection and flow-field selectivity might operate. Rotational Flow-Field Selectivity The group of neurons that processes horizontal motion forms a symmetric bilateral network that is able to combine information about motion presented in front of both eyes. Here I consider a group of 16 neurons whose connections have been explicitly identified. Each of these neurons has a large dendritic tree receiving information about ipsilateral local motion events that is spatially pooled to produce a directionally selective response. In addition, some of the lobula plate neurons are also sensitive to motion cues in front of the other eye. This information is carried by the spiking neurons H1, H2 and Hu that send their axons to the other side of the brain, where the H1- and H2-cells synapse onto 2 of the 3 HS-cells, and all three contralaterally projecting cell provide input to both CH-cells. The CH-cells are known to provide inhibitory input to the H1- and H2-cells. These network interactions appear to amplify the response to rotational stimuli and reduce the response to translation. I ablate either single HS-cells or both CH-cells in order to break the path whereby information about the opposite eye reaches the H1- and H2-cell. I did not find that these ablations affected the flow-field selectivity of either H1- and H2-cells. Network modeling showed that although the described circuitry does support rotational flow-field selectivity for the HS- and CH-cells, the model H2-cell does not show the expected flow-field selectivity. This suggests that the circuitry or cellular mechanisms underlying the response properties of the H2-cell are not completely understood. Basis of the Broad Receptive Field of VS-cells As the lobula plate is organized retinotopically the receptive fields of the tangential cells ought to be determined by their dendritic architecture. This appears not always to be the case. One compelling example is the exceptionally wide receptive fields of the vertical system (VS) tangential cells. Using dual intracellular recordings Haag and Borst (2004) found VS-cells to be mutually coupled in such a way that each VS-cell is connected exclusively to its immediate neighbours. This coupling may form the basis of the broad receptive fields of VS-cells. Here I tested this hypothesis directly by photo-ablating individual VS-cells. The receptive field width of VS-cells indeed narrowed after the ablation of single VS-cells, specifically depending on whether the receptive field of the ablated cell was more frontal or more posterior to the recorded cell. In particular, the responses changed as if the neuron lost access to visual information from the ablated neuron and those VS-cells more distal than it from the recorded neuron. These experiments provide compelling evidence that the lateral connections amongst VS-cells are a crucial component in the mechanism underlying their complex receptive fields, augmenting the direct columnar input to their dendrites. Vertical-Horizontal Interactions Two heterolaterally spiking cells, the H1- and H2-cells have been shown to be sensitive to vertical motion presented in the frontal portion of their receptive fields. Receptive field measurements performed here show that the H1-, VS1- and VS2-cells all respond to vertical downward motion across an almost completely overlapping portion of the frontal visual field. Using dual intracellular recordings Haag and Borst (2003) demonstrated that the VS1-cell but not the VS2-cell supplies input to both these cells. Through current injections into different compartments of the VS1-and VS2-cells I have provided physiological evidence that the output of VS1-cell near its dendritic arbors is the likely site of its input to the H1-cell. This coupling may form the basis of the vertical sensitivity of the H1- and H2-cell. I tested this hypothesis directly by recording the sensitivity of the H1-cell to horizontal and vertical motion in the frontal visual field both before and after the ablation of single VS1-cells. After the ablation of the VS1-cell the response of the H1-cell to vertical motion disappeared but its response to horizontal motion remained robust. These experiments demonstrate that the VS1-cell provides the input to the H1-cell that makes it sensitive to vertical motion in the frontal visual field likely through connections in their dendritic trees.