Long-range inputome of cortical neurons containing corticotropin-releasing hormone

Peilin Zhao, Mengting Zhao, Huading Wang, Tao Jiang, Xueyan Jia, Jiaojiao Tian, Anan Li, Hui Gong, Xiangning LiScientific 2020 · DOI

TL;DR

This study maps the long-range inputs to corticotropin-releasing hormone (CRH)-expressing neurons across six key cortical areas using a novel combination of rabies virus tracing and whole-brain light-sheet microscopy. The findings reveal general patterns of cortical connectivity, with distinct regional specializations in thalamic and basal forebrain inputs, providing a foundational atlas for understanding CRH neuron function in brain circuits.

What Problem Does This Paper Address?

Understanding the intricate neural circuits of the brain is fundamental to deciphering brain function and dysfunction. Corticotropin-releasing hormone (CRH) is a crucial neuropeptide involved in stress, learning, and social behavior, and its dysregulation is linked to various neurological and psychiatric disorders. While CRH neurons are known to be GABAergic and primarily project locally, their long-range inputs and how these inputs differ across cortical regions remain largely uncharacterized. Previous studies were limited by 2D imaging, hindering a comprehensive, whole-brain, 3D understanding of these connections. This paper aims to fill this gap by providing a detailed, high-resolution, whole-brain map of the monosynaptic inputs to CRH neurons in six distinct cortical areas.

Methods

The researchers employed a sophisticated viral tracing strategy. They used Cre-dependent adeno-associated viruses (AAVs) to express the TVA receptor and rabies virus G glycoprotein (RG) specifically in CRH neurons of CRH-Cre mice. Following this, a modified rabies virus (RV) engineered with an avian virus envelope (EnvA) was injected. This RV, expressing green fluorescent protein (GFP), can only infect cells expressing the TVA receptor, thus labeling monosynaptic inputs to the CRH neurons.

To achieve whole-brain, high-resolution imaging, they utilized fluorescent micro-optical sectional tomography (fMOST). This technique allows for the acquisition of continuous 3D datasets with a resolution of 0.32 × 0.32 × 2 µm³, enabling the precise localization and reconstruction of input neurons throughout the entire brain. The acquired data was then registered to the Allen Brain Reference Atlas for quantitative analysis and comparison across different cortical regions. Immunohistochemistry was also used to identify cholinergic neurons in the basal forebrain.

Datasets

This study did not use pre-existing datasets for training or evaluation. Instead, it generated novel whole-brain inputome datasets for CRH neurons in six cortical areas (mPFC, M1, S2, S1BF, Au, V1) using the described viral tracing and imaging techniques. These datasets, comprising the spatial coordinates of thousands of input neurons, are made available through a public repository (https://atlas.brainsmatics.org).

Key Insights and Ideas

  • Whole-Brain, 3D Inputome Mapping: The combination of rabies tracing and fMOST provides an unprecedentedly detailed, 3D map of long-range inputs to specific neuronal populations across the entire brain. This is a significant advancement over previous 2D section-based analyses.
  • General Input Patterns with Regional Specificity: While CRH neurons across different cortical areas share some common input sources (ipsilateral cortex, thalamus, basal forebrain), there are distinct regional specializations, particularly in thalamic and basal forebrain projections.
  • Thalamic Input Topography: The study reveals a clear topographical organization of thalamic inputs, with anterior and middle thalamus projecting to rostral cortical areas (like mPFC) and posterior and lateral thalamus projecting to caudal areas (like V1).
  • Basal Forebrain Cholinergic Nuances: While the basal forebrain is a major input source, specific subregions show differential projections. Notably, cholinergic neurons in the diagonal band nucleus project strongly to the prefrontal and visual cortices, and the visual cortex (V1) receives a significant cholinergic input from the horizontal limb of the diagonal band (HDB).
  • Zona Incerta and Ventral Hippocampus as Key Inputs to mPFC: The prefrontal cortex (mPFC) receives dense inputs from the zona incerta (ZI) and ventral hippocampus (vHIP), suggesting a role for these regions in modulating mPFC CRH neuron activity, potentially related to feeding and stress responses.

Key Results

  • Widespread Input Distribution: Input neurons to CRH neurons were found throughout the brain, with the majority located in the ipsilateral cortex, followed by the thalamus, contralateral cortex, and basal forebrain.
  • Cortical Input Patterns: Cortical CRH neurons primarily received input from adjacent cortical regions. For example, mPFC CRH neurons received inputs from the orbitofrontal cortex (OFC) and M2, while M1 CRH neurons received inputs from sensory cortices (S2, S1BF) and M2.
  • Thalamic Input Density: Thalamic projection neurons accounted for over 20% of the total long-range input neurons. Specific thalamic nuclei showed preferential targeting of different cortical CRH neurons (e.g., anterior thalamic nuclei to rostral cortex, posterior thalamic nuclei to caudal cortex).
  • Basal Forebrain Cholinergic Input: Over half of the basal forebrain inputs to CRH neurons were cholinergic, with the exception of mPFC. The V1 showed a preference for cholinergic input from the HDB.
  • ZI and vHIP Input to mPFC: Dense inputs from the zona incerta (ZI) and ventral hippocampus (vHIP) were identified specifically for CRH neurons in the mPFC.

Limitations and Future Work

The authors acknowledge that their study focused on monosynaptic inputs and did not explore polysynaptic connections. While the viral tracing method is highly specific, there's a potential for slight overestimation of local presynaptic cells due to minor viral leakage, which they addressed by focusing on long-range inputs.

Future work could build upon this foundational atlas by:

  • Investigating the functional roles of these identified input circuits in CRH neuron activity and behavior.
  • Exploring polysynaptic connections to gain a more complete understanding of CRH neuron circuitry.
  • Examining how these input patterns are modulated by different physiological states, such as stress or learning.
  • Extending this mapping to other CRH neuron populations or other neuronal types within these cortical areas.

Summary updated

Abstract

Dissection of the neural circuits of the cerebral cortex is essential for studying mechanisms underlying brain function. Herein, combining a retrograde rabies tracing system with fluorescent micro-optical sectional tomography, we investigated long-range input neurons of corticotropin-releasing hormone containing neurons in the six main cortical areas, including the prefrontal, somatosensory, motor, auditory, and visual cortices. The whole brain distribution of input neurons showed similar patterns to input neurons distributed mainly in the adjacent cortical areas, thalamus, and basal forebrain. Reconstruction of continuous three-dimensional datasets showed the anterior and middle thalamus projected mainly to the rostral cortex whereas the posterior and lateral projected to the caudal cortex. In the basal forebrain, immunohistochemical staining showed these cortical areas received afferent information from cholinergic neurons in the substantia innominata and lateral globus pallidus, whereas cholinergic neurons in the diagonal band nucleus projected strongly to the prefrontal and visual cortex. Additionally, dense neurons in the zona incerta and ventral hippocampus were found to project to the prefrontal cortex. These results showed general patterns of cortical input circuits and unique connection patterns of each individual area, allowing for valuable comparisons among the organisation of different cortical areas and new insight into cortical functions.
BibTeX
@inproceedings{zhao2020long,
  author = {Peilin Zhao and Mengting Zhao and Huading Wang and Tao Jiang and Xueyan Jia and Jiaojiao Tian and Anan Li and Hui Gong and Xiangning Li},
  title = {{Long-range inputome of cortical neurons containing corticotropin-releasing hormone}},
  booktitle = {Scientific},
  year = {2020},
  doi = {10.1038/s41598-020-68115-x},
}
Figure