The amygdala is a deep nucleus located in the temporal lobe of the forebrain that governs negative emotions such as fear and anxiety. Neuroscientists classify various neurons according to the transmitters they release, and these different types of neurons often have different functions. The amygdala is home to a variety of neurons that secrete different transmitters. Among these many neurons, we focus on a class of neurons that release corticotropin-releasing factor (CRF). Not only this transmitter is particularly distributed in the amygdala, but it is closely related to the fear and anxiety that the amygdala involved in.

When people experience stress responses such as fright or threat, the hypothalamus will release CRF to promote the secretion of adrenaline, making the organism more alert and enhancing its metabolism to cope with these external stimuli. Since CRF itself is highly correlated with various stresses, and negative emotions such as fear and anxiety processed by the amygdala are a series of stresses for the body, it is necessary to study the functions of these CRF-releasing neurons in the amygdala in the various behaviors in which the amygdala is involved.

The neurons in the brain responsible for the same behavior all form connections and transmit information from one by one and function together. However, even the same batch of neurons receiving information from different superiors may play different roles in various behaviors. Although previous studies have reported the functions of these CRF neurons in some behaviors, in order to systematically and comprehensively analyze the functions of these neurons, the first thing to do is to parse out the connections of them.

There is a close relationship between the location of neurons and the information they transmit. Neurons with the same function always tend to be distributed together. For example, most neurons in the prefrontal cortex are involved in higher cognitive functions, while neurons in the cerebellum are mostly involved in motor regulation. However, in general, neurons transmit only two types of information, one that promotes the activity of downstream neurons receiving that information, the so-called excitatory information, and the corresponding other that inhibits the activity of downstream neurons, which is the inhibitory information. Therefore, figuring out the distribution of these input neurons and the type of information they transmit are the two primary concerns of neuroscientists. In addition to the analysis of the types of information transmitted, our study also focused on two input patterns. Overall top-down inputs tend to converge excitatory information, while bottom-up inputs tend to diffuse inhibitory information. This unique information input patterns enables CeA-CRF neurons to integrate more high-level input information from the top-down regions, while the basic physiological information delivered from bottom-up regions can achieve sufficient response intensity in time.

Other than that, seeing is believing. Neurons are composed of a cell body as well as long axons that connect to downstream neurons. We hope to restore the whole picture of a long axon starting from the cell body of the input neuron and eventually terminating at the CRF neuron. However, we encountered difficulties such as weak fluorescence of axonal terminals, insufficient imaging accuracy, and long imaging time. We eventually solved these problems with fast high-resolution light-sheet imaging, and could quickly trace the complete neuron in three-dimensional space within the existing framework. However, this was only the first step in our vision to restore as much information as possible about the neuronal pathways, so the next challenge was to demonstrate this projection process in situ in a whole brain atlas. Then the existing Allen Mouse Brain Atlas recognized by the academic community was used to precisely restore the entire spatial information of neurons by anchoring three three-dimensional coordinate points of it in this brain atlas, and this result revealed that these axons do not connect to their target neuron in the most economical and efficient way with the shortest straight line between these two points, but choose to pass through some nuclei in the brain. We speculate that the reason behind this may be partly a developmental change, but more likely because these long axons also exchange information with these passing nuclei for the purpose of some kind of information modification. We then identified the likely form of information stream from the passing nuclei to the long axons by pre- and postsynaptic marker staining, but with the current technical means we are still unable to specifically manipulate the flow of information from these inputs to see how it might affect behavioral functions.

Our study identified all the neurons that transmit information to CRF neurons in amygdala throughout the brain. Starting from the connections of these neurons, by analyzing the locations of these input neurons and the transmitters they released, we could infer the information they might transmit. In addition, their input patterns and connectivity are revealed in our study. All this information provides anatomical and structural references for subsequent functional studies of these CRF neurons.