introduction
Scientists have been working to understand how primate brains support natural behaviors, especially social interaction and communication behaviors. However, there is great complexity and spontaneity in these natural behaviors, which makes it difficult for traditional task-oriented experimental and analytical methods to unravel neural activity patterns in natural behavior research. In this study, the authors used natural behavior-oriented experimental and analytical methods to record the neural activity of free-moving marmosets in conversation, revealing how the prefrontal cortex and premotor cortex are involved in natural communication and vocal behavior.
近日，美国加州大学圣地亚哥分校Cory T. Miller和Jingwen Li团队在Neuron上发表了文章Representing the dynamics of natural marmoset vocal behaviors in frontal cortex。
By recording the neural activity of marmosets as they make natural vocalizations, the authors discovered a complex pattern of activity in a group of neurons in the prefrontal cortex. Unlike previous studies conducted on fixed tasks, the researchers used generalized linear models (GLMs) to quantify the sequential changes of these neurons in natural vocalizations to better understand the functional distribution of the prefrontal and anterior motor cortex and their support for complex social behaviors. The study found that different clusters of neurons responded significantly to different aspects of the marmosets' natural communication behavior—hearing, vocalization, and social status—and that these clusters were not confined to specific cortical regions, but were widely distributed throughout the prefrontal and anterior motor cortex.
The authors used a generalized linear model (GLM) to analyze the neural activity of marmosets in their natural vocal behavior. GLM can predict the firing rate of neurons in a continuous natural behavior by introducing variables such as behavioral events, internal states, and neuronal firing history as factors. The analysis shows that GLM can not only capture the neural responses of listening and vocal behavior, but also reveal neuronal activity related to social states that cannot be found by traditional analysis methods (PSTH), so as to more comprehensively characterize the brain-behavior interaction. Through GLM analysis, the researchers identified multiple clusters of neurons associated with natural vocal behavior, including clusters of neurons associated with austemy, responsive occurrence, spontaneous vocalization, and internal behavioral states such as conversation or spontaneous states.
狨猴自然发声实验与GLM分析模型（Credit: Neuron）
A surprising finding is that these functional clusters are not confined to a specific area of the brain, but are widely distributed throughout the prefrontal lobe and anterior motor cortex. The authors further analyzed the predictive power of these neurons on whether marmosets will make vocalizations after hearing a sound. The results showed that nearly one-third of the neurons were able to predict the conversational vocalization behavior of marmosets, and some of them had a prediction accuracy (AUC) of more than 0.8. Notably, neurons with high predictive performance are distributed in different functional clusters, showing the important role of the entire prefrontal cortex and the premotor cortex in natural social behavior.
In conclusion, this study reveals a wide range of functional clusters in the frontal cortex in the natural vocal behavior of marmosets, and points out that these clusters are distributed throughout the cortex and not just in specific areas. This discovery breaks the traditional framework of task-oriented neuroscience research and provides a new perspective on how primates interact socially in their natural environment.
bibliography
https://doi.org/10.1016/j.neuron.2024.08.020
Editor-in-charge|Explore Jun
Typography |
Article Source | BioArt”
End