Variability and robustness of functional neuronal properties in visual motion-detection circuitry

苍蝇可以对不同的视觉刺激做出强健的反应，但行为健壮性背后的神经元底物和计算在很大程度上是未知的。在果蝇中，已经绘制了核心运动检测电路的开路径和离路径的神经元。其中，Tm9神经元是视桥定位排列在视外通路的柱状中间神经元，并指导视运动行为。最近的研究表明，与其他TM神经元不同，Tm9可以接受各种各样的突触前输入，不同的Tm9神经元的身份可能不同。我们的初步工作还揭示了一种非柱状、广野细胞类型的Tm9突触前的形态和功能变异，Tm9是远端延髓神经元DM4。此外，光学功能记录揭示了单个Tm9功能特性的相应变异性。这些发现提出了两个需要在P5中检验的假设：(1)可变突触输入对于确定Tm9神经元的异质性功能特性是必不可少的，(2)Tm9神经元反过来是电路功能所必需的，并导致对挑战性刺激条件的行为健壮性。因此，P5的目标是了解突触前电路中的不精确在多大程度上促进了对发育噪声的健壮性以及对可变视觉输入的功能健壮性。在该项目的两个平行部分中，我们将首先量化和操纵TM9的可变突触前柱状电路结构，并将分析DM4广域神经元作为TM9的主要异质输入的贡献。要做到这一点，我们将结合解剖学研究和苍蝇眼的生理测量。我们将对可变输入进行基因操作，并测量Tm9生理学和对一系列视觉刺激的行为反应。当结束时，我们将测试不精确的突触形成和不同的Tm9功能特性对运动检测对挑战性刺激条件的稳健性所起的作用。

Flies can respond robustly to variable visual stimuli, but the neuronal substrate and computation underlying behavioral robustness are largely unknown. Neurons of the ON and OFF pathways of core-motion detection circuitry have been mapped in Drosophila. Of these, Tm9 neurons are retinotopically arranged columnar interneurons of the OFF pathway, and guide optomotor behavior. Recent work has shown that Tm9, unlike other Tm neurons, can receive a wide variety of presynaptic inputs, whose identity may differ for individual Tm9 neurons. Our preliminary work has also revealed morphological and functional variability in one non-columnar, wide-field cell type presynaptic to Tm9, the distal medulla neuron Dm4. Furthermore, optical functional recordings revealed a corresponding variability of individual Tm9 functional properties. These findings give rise to two hypotheses to be tested in P5: (1) variable synaptic inputs are essential to determine heterogeneous functional properties of Tm9 neurons, which (2) are in turn required for circuit function and lead to behavioral robustness to challenging stimulus conditions. Hence, the goal of P5 is to understand to what extent imprecisions in the presynaptic circuitry facilitates robustness to developmental noise as well as functional robustness to variable visual inputs. In two parallel parts of the project, we will first quantify and manipulate the variable presynaptic columnar circuit architecture of Tm9, and we will analyze the contribution of the Dm4 wide-field neuron as a major heterogeneous input to Tm9. To do so, we will combine anatomical studies with physiological measurements across the fly eye. We will genetically manipulate variable inputs and measure Tm9 physiology as well as behavioral responses to a range of visual stimuli. When concluded, we will have tested the role of imprecise synapse formation and heterogeneous Tm9 functional properties for the robustness of motion detection to challenging stimulus conditions.