NEURONAL CONTROL OF INTERACTIONS AMONG BEHAVIOR

我们建议在以下层面考察行为交互作用 单个的，已识别的神经细胞在神经系统中的药物 利奇。我们将首先通过以下方式展示各种行为如何交互 确定引出四种不同的 行为--游泳、步行、缩短和局部弯曲 对机械感觉刺激的反应。市场中的可变性 响应将为我们提供一种行为集的度量 动物。然后，我们将通过以下方式展示这四种行为如何交互 同时或顺序地传递成对的刺激， 它们各自都会非常可靠地引出其中一种行为。 如果这对行为是相互排斥的，这些研究 将测量动物的行为选择；从不同的 配对，我们将确定动物的行为等级。这个 行为可能会被证明是兼容的，因此元素是 不干涉的或协调的。看起来游泳和游泳 爬行是协调的，即局部弯曲不干扰 爬行或游泳，而所有其他行为都是 被组织成一个等级。我们还将确定在多大程度上 这些行为相互作用被神经调节剂改变， 尤其是5-羟色胺，通过学习。 在平行研究中，我们将描述细胞和网络的特征 负责启动和生成每个 四种行为。对于其中两个项目，此过程即将完成 行为，游泳和局部弯曲，对 另外两个。我们将从两到三个神经元进行穿透和记录 同时处于半完整的、表现为水蚤的或孤立的 神经索。我们将测试神经元的特征 通过比较记录在 动物到数字化模型网络，在正常和 令人不安的条件。微扰将是：1)偏振 一次一到三个神经元，或者2)消除相互作用 通过光消融将神经元完全分开。到时候我们会的 确定四种行为中神经元之间的相互作用 电路，为了找到兼容性的神经元基础， 协调，或行为的相互排他性。基于 这些信息，我们将能够确定地点和性质 神经元的变化导致这些基因的改变 神经调节剂和学习的相互作用。这样的神经元 这种机制很可能在所有动物身上都能找到，包括人类。 当行为相互作用时。

We propose to examine behavioral interactions at the level of single, identified neurons in the nervous system of the medicinal leech. We will first show how various behaviors interact by determining the probability of eliciting each of four different behaviors--swimming, walking, shortening and local bending--in response to mechanosensory stimuli. The variability in the responses will provide us with a measure of behavioral set of the animal. We will then show how the four behaviors interact by delivering pairs of stimuli, either simultaneously or in sequence, which individually would elicit one of the behaviors very reliably. If the pairs of behaviors are mutually exclusive, these studies will measure the animal's behavioral choice; from different pairings, we will determine the animal's behavioral hierarchy. The behaviors may prove to be compatible, so that elements are noninterfering or coordinated. It appears that swimming and crawling are coordinated, that local bending does not interfere with crawling or swimming, and that all other behaviors are organized into a hierarchy. We will also determine to what extent these behavioral interactions are modified by neuromodulators, particularly serotonin, and by learning. In parallel studies, we will characterize the cellular and network properties responsible for initiating and generating each of the four behaviors. This process is nearly complete for two of the behaviors, swimming and local bending, and is well begun for the other two. We will impale and record from two or three neurons simultaneously in semi-intact, behaving leeches or in isolated nerve cords. We will test the neuronal characterizations for the completeness by comparing the activity pattern recorded in the animal to a digitized model network, under both normal and perturbed conditions. The perturbations will be: 1) polarizing one to three neurons at a time, or 2) eliminating the interactions between neurons altogether by photoablation. We will then determine the interactions between neurons in the four behavioral circuits, to find the neuronal basis for the compatibility, coordination, or mutual exclusivity of the behaviors. Based on this information, we will be able to determine the site and nature of neuronal changes responsible for the modification of these interactions by neuromodulators and by learning. Such neuronal mechanisms are likely to be found in all animals, including humans, when behaviors interact.