Reciprocal projections between the globus pallidus externa and cortex span motor and nonmotor regions

Emily Ferenczi, Wengang Wang, Abhishek Biswas, Trent Pottala, Yuanlin Dong, Andrew K. Chan, Maria Albanese, Raina S. Sohur, Tao Jia, Kevin J. Mastro, Bernardo L. SabatiniProceedings of the National Academy of Sciences 2025 · DOI

TL;DR

This study reveals a direct, bidirectional communication pathway between the globus pallidus externa (GPe) and the cortex, spanning both motor and non-motor brain areas. This circuit, previously underestimated, is crucial for integrating behavioral goals, internal states, and environmental cues to rapidly modulate behavior, offering new targets for treating basal ganglia disorders.

What Problem Does This Paper Address?

The globus pallidus externa (GPe) has traditionally been viewed as a "motor relay" within the basal ganglia, primarily involved in regulating movement. However, recent evidence suggests a more complex role, implicating it in diverse behaviors and highlighting its intricate connections beyond the basal ganglia. Specifically, the precise anatomical, molecular, and electrophysiological characteristics of GPe neurons that project to the cortex (pallidocortical neurons) were not well understood. Furthermore, the extent to which these GPe-cortex circuits are reciprocal, meaning the cortex also sends signals back to the GPe, and how this bidirectional communication is organized, remained unclear. This paper aims to fill these knowledge gaps, providing a detailed map of these connections and their functional implications.

Methods

The researchers employed a multi-pronged approach combining viral tracing, in vivo electrophysiology, and molecular techniques.

  • Retrograde Viral Tracing: They used rabies virus (RV) and cholera toxin b (CTb) to trace the projections of GPe neurons to various cortical regions (sensory, motor, insular, and frontal cortices). This allowed them to map the spatial distribution of pallidocortical neurons within the GPe and determine their projection targets.
  • In Situ mRNA Hybridization (mRNA-FISH): This technique was used to identify the molecular markers expressed by pallidocortical neurons, helping to classify them into distinct subtypes.
  • Whole-Cell Electrophysiology: Ex vivo recordings from acute brain slices were performed on retrogradely labeled pallidocortical neurons to characterize their intrinsic electrophysiological properties (e.g., firing patterns, membrane resistance).
  • Optogenetics: Channelrhodopsin (ChR2) was used to stimulate cortical axons projecting to the GPe, allowing researchers to assess the functional impact of these inputs on GPe neuron activity and to identify monosynaptic glutamatergic connections.
  • Conditional Viral Tracing: Transgenic mouse lines with Cre-recombinase expression in specific cell types (e.g., ChAT+, Npr3+, Pvalb+, FoxP2+, Rbp4+, Tlx3-PL56-Cre, Sim1-KJ18-Cre) were used in conjunction with Cre-dependent viral vectors to specifically target and manipulate defined neuronal populations and their projections.

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • GPe as a Cortical Interface: The study firmly establishes the GPe not just as an internal basal ganglia nucleus but as a key interface with the cortex, capable of both sending and receiving information.
  • Topographic Organization: Pallidocortical neurons are not randomly distributed within the GPe. Their location within the GPe reflects their specific cortical projection target, suggesting a degree of spatial organization that could support targeted communication.
  • Electrophysiological and Molecular Heterogeneity: Pallidocortical neurons are not a single, uniform population. They segregate into at least two distinct electrophysiological and molecular subtypes, indicating specialized roles or processing capabilities.
  • Reciprocal Cortical Input: The GPe receives direct, monosynaptic excitatory input from deep cortical layers (layer V/VI), particularly from motor and non-motor association areas. This input is functionally significant, capable of evoking action potentials in GPe neurons.
  • Fast, Bidirectional Circuit: The existence of a rapid, two-way communication loop between the GPe and cortex is a major finding. This circuit is ideally positioned to integrate diverse information streams for rapid behavioral modulation.

Key Results

  • Pallidocortical neurons project to a wide range of cortical regions, including sensorimotor areas and higher-order association areas like the anterior cingulate and insular cortices.
  • These neurons are organized topographically within the GPe, with distinct clusters projecting to different cortical areas.
  • Electrophysiological recordings revealed at least two distinct clusters of pallidocortical neurons based on their firing properties, independent of their cortical projection target.
  • Molecular analysis identified at least two main subtypes of pallidocortical neurons, characterized by the expression of markers like Npr3/Slc32a1 and Nkx2-1/Chat.
  • The GPe receives direct glutamatergic input from deep cortical layers (V/VI), with inputs from the insular cortex being particularly prominent in some datasets.
  • This cortical input can functionally influence GPe neuron activity, evoking action potentials in some neuron types and altering firing patterns in others.
  • Crucially, pallidocortical neurons themselves receive direct excitatory input from the cortex, confirming the reciprocal nature of this circuit.

Limitations and Future Work

The authors acknowledge several limitations and suggest future research directions:

  • Axonal Projection Sparsity: While they identified subcortical projections (e.g., to the reticular nucleus of the thalamus and amygdala), the sparse nature of these projections makes detailed mapping challenging.
  • Marker Expression in Cortex: The expression of some markers (Npr3, Chat) in the cortex could lead to spurious axonal projections from cortical neurons rather than GPe neurons, requiring careful control experiments.
  • Discrepancy in Cortical Input Origin: A discrepancy was noted between their findings of intratelencephalic (IT) neuron input to the GPe and previous studies suggesting pyramidal tract (PT) neurons are the primary source. This could be due to differences in mouse lines, cortical regions, or sampling limitations.
  • Limited Cortical Targets Explored: The study focused on four main cortical projection targets and two main cortical input regions. Future work should explore other cortical areas, such as the orbitofrontal and retrosplenial cortices, which are likely to influence pallidal function.
  • Functional Significance: While they demonstrated functional impact of cortical input on GPe neuron activity, further in vivo studies are needed to understand how these reciprocal loops contribute to behavior and how they are modulated during specific tasks. Simultaneous recordings from downstream structures (SNr, thalamus, cortex) during stimulation will be crucial for building comprehensive models.
  • Therapeutic Targets: The study highlights the potential for identifying therapeutic targets for neuromodulation in basal ganglia disorders, but further research is needed to directly test the functional role of these circuits in disease states.

Summary updated

Abstract

The globus pallidus externa (GPe) is a heterogeneous nucleus of the basal ganglia, with intricate connections to other basal ganglia nuclei, as well as direct connections to the cortex. The anatomic, molecular, and electrophysiologic properties of cortex-projecting pallidocortical neurons are not well characterized. Here, we show that pallidocortical neurons project to diverse motor and nonmotor cortical regions, are organized topographically in the GPe, and segregate into at least two distinct electrophysiological and molecular phenotypes. In addition, we find that the GPe receives direct synaptic input from deep layers of diverse motor and nonmotor cortical regions, some of which form reciprocal connections onto pallidocortical neurons. These results demonstrate the existence of a fast, bidirectional circuit between the GPe and the cortex that is ideally positioned to integrate information about behavioral goals, internal states, and environmental cues to rapidly modulate behavior.
BibTeX
@inproceedings{ferenczi2025reciprocal,
  author = {Emily Ferenczi and Wengang Wang and Abhishek Biswas and Trent Pottala and Yuanlin Dong and Andrew K. Chan and Maria Albanese and Raina S. Sohur and Tao Jia and Kevin J. Mastro and Bernardo L. Sabatini},
  title = {{Reciprocal projections between the globus pallidus externa and cortex span motor and nonmotor regions}},
  booktitle = {Proceedings of the National Academy of Sciences},
  year = {2025},
  doi = {10.1073/pnas.2423367122},
}
Figure