Neuronal mechanisms of integrated flight control and goal-directed behaviour in butterfly

稳定姿势的反射如何与自愿行为相结合？在运动过程中，包括人类在内的所有动物在各种环境条件下都保持默认的身体方向。当在不平坦的地形上行走，在湍流中飞行或在水流中游泳时，强大的内环控制系统不断测量外部扰动并产生负反馈信号以稳定运动。但是，如果动物想改变其轨迹，例如避免碰撞或转向有吸引力的目标，会发生什么呢？达到这些效果的动作会立即触发反射，稳定原来的身体方向。为了解决这个问题，von Holst和Mittelstaedt（1950）提出，动物会产生一个“传出复制”，预测对自我产生的运动的感觉反应。传出副本中和了由随意运动引起的“传入”感觉信号，从而防止稳定反射发挥作用。使用传出副本的优点是，而不是被完全封锁，内环继续稳定对不必要的扰动在volitional behaviors.Although传出副本已被建议，以帮助在脊椎动物和无脊椎动物的感觉处理，实验证据表明，它们的使用在运动控制的神经回路中获得最近在果蝇。视觉interneurons感知广域视流（LPTCs）被发现调制他们的活动时，苍蝇作出了一个自愿的转身，无论是自发的，或在一个威胁的对象。这些调制具有适当的符号和时间，以作为传出副本，消除在转弯期间经历的传入信号。然而，这些神经元是如何被传出拷贝所靶向的，传出拷贝在大脑中是如何计算的，以及它们是否反映了适用于其他行为环境的一般机制，这些都是未知的。最终，大脑中用于协调行为动作的任何信息都必须传递到体内的运动系统，这在昆虫中是由下行神经元完成的。在果蝇中，少数下行神经元与LPTC形成双向突触，并对宽视野视流（WFDNs）的特定模式敏感。WFDNs位于胸部的视觉系统和运动系统之间，被策略性地放置用于通过参与产生意志行为的中枢脑区域进行传出复制调制。也有证据表明，在下行神经元的水平上，内环和外环通路之间存在串扰。因此，WFDNs可能代表了一个通路的传出复制传输到LPTCs在更外围阶段的感觉运动pathway.Recently，我们已经发现了大量的WFDNs在蝴蝶，这可能起到稳定的特定组件的自我运动经历这些生物的不稳定的飞行特性。蝴蝶是敏捷的飞行者，配备了一个视觉系统，可以精确地调整颜色视觉，在某些物种中，可以使用各种天窗线索来解决导航外环任务。在这个项目中，我们的目标是了解内部和外部循环的行为是如何集成在蝴蝶WFDNs。为此，我们将研究蝴蝶WFDN神经生理学和自由飞行运动学之间的关系，并直接探测它们在系留飞行中的内环和外环行为中的功能。总的来说，该项目将提供概念框架，以促进我们对神经元小回路如何解决昆虫反射陷阱的理解，引导我们走向更普遍的综合感觉运动控制设计原则。

How are reflexes that stabilise posture integrated with voluntary behaviours? During locomotion, all animals, including humans, maintain a default body orientation under various environmental conditions. When walking on uneven terrain, flying in turbulent air, or swimming through water currents, powerful inner-loop control systems constantly measure external perturbations and generate negative feedback signals to stabilise locomotion. But what happens if an animal wants to change its trajectory, for instance to avoid collisions or to turn towards attractive targets? Movements to those effects would immediately trigger reflexes stabilizing the original body orientation. As a result, the animal would be trapped by its own reflexes.As a solution to this problem, von Holst and Mittelstaedt (1950) proposed that animals generate an 'efference copy' that predicts the sensory response to self-generated movements. The efference copy neutralizes 'reafferent' sensory signals caused by the voluntary movement, thus preventing stabilization reflexes from kicking in. The advantage of using efference copies is that rather than being blocked altogether, the inner-loop continues to stabilise against unwanted perturbations during volitional behaviours.Although efference copies have been suggested to aid sensory processing in vertebrates and invertebrates, experimental evidence demonstrating their use in neural circuits for locomotor control was obtained only recently in fruitflies. Visual interneurons sensing wide-field optic flow (LPTCs) were found to modulate their activity whenever the fly made a volitional turn, either spontaneously, or in response to a threatening object. These modulations were of the appropriate sign and timing to function as efference copies, cancelling reafferent signals experienced during a turn. However, it remains unknown how these neurons are targeted by efference copies, how efference copies are calculated within the brain, and whether they reflect a general mechanism applicable for other behavioural contexts.Ultimately, any information in the brain that is used to coordinate behavioural action must be relayed to motor systems in the body, which in insects is done by descending neurons. In flies, a small number of descending neurons form bi-directional synapses with LPTCs and are sensitive to specific patterns of wide-field optic flow (WFDNs). Positioned between the visual system and motor systems in the thorax, WFDNs are strategically placed for efference copy modulation by central brain regions involved in generating volitional behaviour. Evidence is also emerging of crosstalk between inner- and outer-loop pathways at the level of descending neurons. Thus, WFDNs may represent a pathway for efference copy transmission to LPTCs at a more peripheral stage of the sensorimotor pathway.Recently, we have discovered a multitude of WFDNs in the butterfly, which may function to stabilise specific components of self-motion experienced during the erratic flight characteristic of these creatures. Butterflies are agile fliers equipped with a visual system exquisitely tuned to colour vision and, in some species, to solve navigational outer-loop tasks using various skylight cues. In this project, we aim to understand how inner- and outer-loop behaviours are integrated in butterfly WFDNs. To this end we will characterise the relationship between the butterfly WFDN neurophysiology and free-flight kinematics, and directly probe their function during inner- and outer-loop behaviours in tethered flight. Overall, this project will provide the conceptual framework to advance our understanding of how small circuits of neurons solve the reflex trap in insects, leading us towards more generalisable design principles for integrated sensorimotor control.