Dynamic nonlinearities enable direction opponency in Drosophila elementary motion detectors

Abstract

Direction-selective neurons respond to visual motion in a preferred direction. They are directionopponent if they are also inhibited by motion in the opposite direction. In flies and vertebrates, direction opponency has been observed in second-order direction-selective neurons, which achieve this opponency by subtracting signals from first-order direction-selective cells with opposite directional tunings. Here, we report direction opponency in Drosophila that emerges in first-order direction-selective neurons, the elementary motion detectors T4 and T5. This opponency persists when synaptic output from these cells is blocked, suggesting that it arises from feedforward, not feedback, computations. These observations exclude a broad class of linear-nonlinear models that have been proposed to describe direction-selective computations. However, they are consistent with models that include dynamic nonlinearities. Simulations of opponent models suggest that direction opponency in first-order motion detectors improves motion discriminability by suppressing noise generated by the local structure of natural scenes. Users may view, print, copy, and download text and data-mine the content in such documents, for the purposes of academic research, subject always to the full Conditions of use:

A minimal synaptic model for direction selective neurons in Drosophila

“…We sought to maximally constrain our choices of model parameters by published physiological measurements and to show that the model is relatively robust to the specific values of remaining unconstrained parameters. The set of input parameters used here is equal to that used in a model developed to explain direction-opponency in T4 cells (Badwan et al, 2019). To approximately match the spatial acceptance functions of photoreceptors in the fly eye (Stavenga, 2003), we set the full width at half maximum of the input spatial filters to 5.7 degrees.…”

“…Broadening the input spatial receptive fields or decreasing the input spacing would increase the correlation between the input conductances in this model. We would expect that to affect the spatiotemporal correlations to which the model is sensitive (Badwan et al, 2019;Fitzgerald & Clark, 2015). Previous functional studies of the inputs to T4 cells have suggested that their spatial receptive fields may be somewhat wider than that of one photoreceptor (Arenz et al, 2017;Behnia et al, 2014;Gruntman et al, 2018), but we use this simplified spatial processing because our goal is to construct a minimal model.…”

“…Previous functional studies of the inputs to T4 cells have suggested that their spatial receptive fields may be somewhat wider than that of one photoreceptor (Arenz et al, 2017;Behnia et al, 2014;Gruntman et al, 2018), but we use this simplified spatial processing because our goal is to construct a minimal model. To produce peak responses to PD sinusoidal gratings at approximately 1 Hz (Badwan et al, 2019;Maisak et al, 2013;Strother et al, 2017), we fixed the time constant of the temporal filters to 150 ms. The true temporal filters are unlikely to have same time constants or be perfect derivatives of one another (Arenz et al, 2017;Behnia et al, 2014).…”

“…The procedure used to select the values of these parameters is described in detail in Appendix B. As shown previously (Badwan et al, 2019), there exists a broad region of parameter space for which this model displays strong directional responses to sinusoid gratings with a temporal frequency of 1 Hz and a spatial wavelength of 45 degrees. We note that our choice of filter normalization, which differs from that in the previous use of this model (Badwan et al, 2019), affects the parameter values chosen, as it scales g 1 , g 2 , and g 3 relative to g leak .…”

“…This wealth of data about T4 and T5 cells and their inputs has led to many different models that seek to describe the response properties of T4 and T5 cells (Arenz et al, 2017;Badwan, Creamer, Zavatone-Veth, & Clark, 2019;Behnia et al, 2014;Clark et al, 2011;Creamer, Mano, & Clark, 2018; Figure 1. An anatomically constrained synaptic model for T4 cells.…”