Multimodal integration in extraocular motoneurons

持续准确的视觉感知是自我运动的重要行为要求。为了在运动过程中保持视力，视网膜图像漂移通过多感觉和内在运动感知复制信号的协同作用引起的动态反活动眼睛和/或头部调节来最小化。眼外运动神经元是眼球稳像运动输出的最后一个神经元元件，是信号整合的关键元件。目前的建议旨在通过实验/计算方法研究如何通过单个运动神经元整合多模态信号以产生功能适当的运动命令。这将揭示单个细胞水平上运动神经元决策的规则和约束。因此，在这个框架下，目前的项目旨在解读1)眼外运动神经元形态生理特性的范围和分布，作为决定神经元计算动态带宽的潜在因素；2)在被动和主动运动期间，不同的多感觉（视觉-前庭）和运动干扰复制输入对眼外运动神经元的功能相互作用和可塑性。实验方法将在目标3中得到补充和概念化表述，其中生成了一个描述多模态信号集成机制原理的计算模型。该模型的规范基础是在概率框架中传感器融合作为运动过程中最佳视线稳定机制的作用。这项研究将在非洲爪蟾蝌蚪简单而易处理的神经系统上进行，这为实验访问所有相关的感觉运动结构提供了可能性，用于电生理和光学记录以及行为分析。运动活动的表达在这些半完整的准备以及完整的眼睛和内耳允许在存在和不存在运动活动的情况下应用自然感觉刺激。这两个实验室在实验和计算研究方面的互补经验，为成功破译生物算法及其在种群和单个神经元形态和内在特性方面的神经实现提供了可能性。一种可能的运动神经元细分具有不同的属性，可以解释特定的计算能力如何影响细胞对空间和动态方面的神经元输出的决定，这是有效运动行为的基础。因此，拟议的项目将展示眼外运动神经元如何从多模态信号中产生足够的凝视稳定所需的适当活动模式，以及如何在神经元水平的计算模型中形式化潜在机制。

Continuous accurate visual perception is an important behavioral requirement during self-motion. To maintain visual acuity during locomotion, retinal image drift is minimized by dynamic counteractive eye and/or head-adjustments evoked by the concerted action of multisensory and intrinsic motor efference copy signals. As final neuronal element of motor outputs for image-stabilizing eye movements, extraocular motoneurons are the key element for signal integration. The current proposal aims to study with an experimental/computational approach how multimodal signals are integrated by individual motoneurons to generate functionally appropriate motor commands. This will reveal rules and constraints of motoneuronal decision-making at a single cell level. In this framework, the current project therefore aims to decipher 1) the range and distribution of morpho-physiological properties of extraocular motoneurons as the underlying element that determines the dynamic bandwidth for neuronal computations, 2) the functional interaction and plasticity between different multisensory (visuo-vestibular) and locomotor efference copy inputs to extraocular motoneurons during passive and active motion. The experimental approach will be complemented and conceptually formulated in aim 3) in which a computational model is generated that describes the mechanistic principles for multimodal signal integration. The normative basis for the model is the role of sensor fusion in a probabilistic framework as mechanism for optimal gaze-stabilization during locomotion. The study will be performed on the simple and tractable nervous system of Xenopus laevis tadpoles that offers the possibility to experimentally access all relevant sensory-motor structures for electrophysiological and optical recordings as well as behavioral analyses. The expression of locomotor activity in these semi-intact preparations along with intact eyes and inner ears allows the application of natural sensory stimuli in the presence and absence of locomotor activity. The complementary experience of the two involved laboratories for experimental and computational studies offers the possibility to successfully decipher the biological algorithms and their neural implementation with respect to population and single neurons morphological and intrinsic properties. A possible separation into motoneuronal subdivisions with distinct properties might be the explanation of how specific computational capabilities influence the cellular decision on spatial and dynamic aspects of neuronal output that underlies effective motor behavior. The proposed project will thus demonstrate how extraocular motoneurons generate the appropriate activity patterns required for adequate gaze stabilization from multimodal signals and how the underlying mechanisms can be formalized in a computational model at the neuronal level.