Vector Trace cells in the Subiculum of the hippocampal formation

众所周知，海马体支持空间记忆、导航、规划和想象。了解海马体如何支持这些功能是全球范围内激烈、引人注目的研究的主要目标。在这里，我们专注于一种特殊的空间记忆：记住空间定义线索的位置有多远，在什么方向。这被称为矢量记忆，我们认为它对海马体功能至关重要。理论和实证研究表明，广义上讲，空间映射依赖于两种感官输入：一种来自自我运动线索，一种来自外部环境线索。关于自我运动，我们可以通过估计我们已经移动了多远，朝哪个方向移动，例如通过运动和感觉系统告诉我们迈步的力度和次数，以及迈步扩展的程度等线索，来了解我们所处的位置。当然，我们周围的空间并不是“空的”，而是充满了各种外部线索，如边界、物体和目标（如本垒）。为了有效地导航，动物还必须对这些定义空间的外部线索定位的距离和方向进行编码。像路径规划和想象这样的认知过程需要记住这些向量表示的能力，但证据表明存在能够提供这种向量记忆的神经元。重要的是，我们最近发现了矢量记忆的证据：一种新的神经元类型，我们称之为矢量跟踪细胞（VTC），它位于海马体形成的一个区域，叫做下丘。VTCs在物体被移走后数小时内仍能记住物体的距离和方向。我们怀疑至少有一些vtc可以记住这几天，我们将对此进行测试。VTCs似乎为计算动物与多个线索（或不同线索之间）之间的空间关系提供了记忆，从而摆脱了必须直接感知这些线索的限制。与无法记住这些空间关系的系统相比，这可以实现更强大的空间规划和富有想象力的认知。事实上，我们推测可能有一些认知功能不是明显的空间性的，而是依赖于多维度的向量编码，而VTCs可以支持，如果输入的是口头信息。显然，如果我们想要准确理解海马体是如何支持空间记忆、导航、规划和想象的，那么我们就需要了解海马体是如何产生基于矢量的空间记忆的。如果不感谢耻骨下的特殊贡献，这是不可能实现的，我们在那里发现了矢量跟踪细胞，这是海马形成的关键输出区域。至少在人类中，它可能是最关键的海马体区域，在一个被称为“默认模式网络”的网络中与大脑的其他部分相互作用。自传式记忆、创造性思维、白日梦、想象力以及面向未来的推测和计划都被认为高度依赖于这种“默认模式网络”。我们怀疑基于向量的认知和记忆支持这些更广泛的基于记忆和创造性的功能。这项资助旨在通过提供以下见解来逐步改变对海马体矢量编码的理解：海马体下带如何在功能上组织成不同的部分，其中一个部分支持矢量记忆；单个矢量跟踪细胞如何产生矢量编码，形成和保持矢量记忆的机制，以及这种记忆持续多久；最后，外部线索的矢量编码如何影响依赖于自我运动的空间映射。

It is well established that the hippocampus supports spatial memory, navigation, planning, and imagination. Understanding how the hippocampus supports these functions is a major goal of intense, high-profile research globally. Here, we focus on a particular kind of spatial memory: remembering how far away and in what direction space-defining cues are located. This is called vector memory, and we think it is crucial to hippocampal function.Theoretical and empirical work has indicated that spatial mapping relies, broadly speaking, on two kinds of sensory inputs: one derived from self-motion cues, one from external environmental cues. Regarding self-motion, we can get some idea of where we are by estimating how far we have moved and in what direction, for instance from clues such as the motor and sensory systems informing us of the vigour and number of strides taken, and the degree of stride extension. Of course, the spaces around us are not 'empty', but filled with various external cues such as boundaries, objects and goals (like home bases). To navigate effectively, an animal must also encode how far away and in what direction such space-defining external cues are located. Cognitive processes such as path-planning and imagination entail the ability to remember such vector representations, but evidence that there are neurons that can provide such vector memory has been lacking.Importantly, then, we have recently found exactly such evidence of vector memory: a novel neuron type, which we call the Vector Trace cell (VTC), located in a region of the hippocampal formation called the subiculum. VTCs remember the distance and direction to an object for hours after that object is removed. We suspect that at least some VTCs can remember this for days, which we will test. It seems likely that VTCs provide memories for computing the spatial relationships between an animal and multiple cues (or between different cues), freed from the constraints of having to directly perceive those cues. This enables more powerful spatial planning and imaginative cognition than would be the case for a system that could not remember these spatial relationships. Indeed we speculate there might be some cognitive functions that are not obviously spatial, but rely on vector coding in more than one dimension, that VTCs could support, if the input to them was for example verbal information.Clearly, if we want to understand exactly how hippocampus supports spatial memory, navigation, planning, and imagination, which is intensively researched globally, then we will need to understand how the hippocampus produces vector-based spatial memory. This cannot be achieved without appreciating the particular contribution of the subiculum, where we have discovered Vector Trace cells, and which is a key output region of the hippocampal formation. In humans at least, it may be the most crucial hippocampal region for interacting with the rest of the brain in a network called the 'default mode network'. Autobiographical memories, creative thought, daydreaming, imagination and future-oriented speculation and plans are all thought to be highly dependent upon this 'default mode network'. We suspect that vector-based cognition and memories support these wider memory-based and creative functions. This grant aims to deliver a step-change in the understanding of hippocampal vector coding by delivering insights into: how the subiculum is organised functionally into different divisions, with one of these divisions supporting vector memory; the mechanisms of how individual vector trace cells generate vector coding and form and maintain vector memories, and how long this memory lasts; and finally how vector coding of external cues influences spatial mapping that is reliant, for its moment-to-moment activity, on self-motion.