A novel channel of information transmission in the brain: correlations between the activities of medullary neurons encode behaviorally relevant stimuli in weakly electric fish

我们在日常生活的几乎每个方面都依赖于我们的感官，然而我们对大脑如何处理感官刺激以产生可靠行为的理解非常有限。由于单个神经元的活动通常表现出实质性的可变性，因此认为通过结合神经群体的活动（即群体编码）来产生稳健的行为反应。为了研究群体编码，科学家们同时记录了许多神经元的活动，揭示了神经元的活动通常是相关的，既包括它们对刺激的平均反应（即信号相关性），也包括它们的可变性（即噪声相关性）。虽然人们普遍认为相关性会影响编码，但它们是有益的还是有害的，仍然是科学界激烈争论的焦点。理解相关活动是如何影响编码的，因为相关性不是固定的，而是根据感觉和行为环境或大脑内部状态而变化的，这一事实进一步复杂化了。然而，这种可塑性允许一个有趣的假设，即相关性本身可以传递有关感官输入的信息，并构成大脑中独立于单个神经元属性（例如放电率）的额外信息传递通道。然而，直到今天，这一理论仍未经实验验证。在我的研究期间，我将在弱电鱼Apteronotus leptorhynchus的电感觉系统中测试这一假设。Apteronotus leptorhynchus是一种南美裸子动物，它通过主动产生并不断包围其身体的电场来导航和交流。已知髓质内的电感觉神经元表现出很强的相关性。基于我的初步数据，我假设噪声相关性逐渐编码行为相关的刺激属性，从而在大脑中提供了一个额外的信息传递渠道。经过彻底的表征，我将阐明神经回路的性质，使这种新的编码机制。先前的研究表明，许多科学突破都是通过研究具有相对简单且描述良好的神经系统和具有良好特征的自然刺激的模型动物而实现的，这些都适用于瘦吻翼龙。此外，这种神经系统的功能原理和分子机制与包括人类在内的高等脊椎动物相似。因此，我希望我的发现将广泛适用于其他模型系统和大脑区域，并将对理解相关活动如何影响种群编码产生变革。

We rely on our senses in almost every aspect of our everyday live, yet our understanding of how the brain processes sensory stimuli in order to generate reliable behavior is very limited. As the activity of the single neuron typically displays substantial variability, the generation of robust behavioral responses is thought to be achieved by combining the activities of neural populations (i.e. population coding). To investigate population coding, scientists have recorded from many neurons simultaneously which revealed that the activities of neurons are often correlated, both in terms of their average response to a stimulus (i.e. signal correlations) as well in terms of their variabilities (i.e. noise correlations). While it is generally agreed that correlations impact coding, whether they are beneficial or detrimental is still focus of intense debates within the scientific community. Understanding how correlated activity affects coding is furthermore complicated by the fact that correlations are not fixed but change depending on sensory and behavioral context or the internal state of the brain. This plasticity however allows for the interesting hypothesis that correlations themselves could convey information about sensory input and constitute an additional channel of information transmission in the brain that works independent of single neuron attributes (e.g. firing rate). However, to this day this theory remains experimentally untested. During my research fellowship, I will test this hypothesis in the electrosensory system of the weakly electric fish Apteronotus leptorhynchus, a South-American Gymnotid species that navigates and communicates by means of an electric field that is actively generated and constantly surrounds its body. Electrosensory neurons within the medulla are known to display strong correlations. Based on my preliminary data, I hypothesize that noise correlations gradually encode behaviorally relevant stimulus attributes and thus providing an additional channel of information transmission in the brain. After a thorough characterization, I will then elucidate the nature of the neural circuits that enable this novel coding mechanism.Previous work has shown that many scientific breakthroughs were achieved by working on model animals with a relatively simple and well described nervous system and well-characterized natural stimuli, which all apply to Apteronotus leptorhynchus. Furthermore, the functional principles and molecular machinery of this nervous system are similar to those found in higher vertebrates including humans. Consequently, I expect that my findings will have a broad applicability to other model systems and brain areas and will be transformative towards understanding of how correlated activity impacts population coding.