TL;DR
This paper presents a computational model of brainstem circuits that control locomotion speed and gait in mice. By integrating existing knowledge of brainstem-spinal cord interactions, the model proposes two distinct pathways for speed and gait control, offering a mechanistic explanation for how the brainstem orchestrates complex movements.
What Problem Does This Paper Address?
Locomotion, the ability to move, is fundamental for survival. While we know that the spinal cord generates the basic rhythmic patterns for limb movement, the brainstem plays a crucial role in initiating, modulating, and adapting locomotion to different speeds and gaits (e.g., walking vs. running). However, the precise neural circuits within the brainstem and how they interact with the spinal cord to achieve this control remain poorly understood. Specifically, how different brainstem nuclei coordinate to control both the speed of movement and the specific pattern of limb coordination (gait) is a key question. This paper aims to bridge this gap by building a computational model that integrates known brainstem structures and their connections to the spinal cord.
Methods
The authors extended their previous computational model of spinal locomotor circuits. The new model incorporates key brainstem structures: the cuneiform nucleus (CnF) and pedunculopontine nucleus (PPN) within the mesencephalic locomotor region (MLR), and the lateral paragigantocellular nucleus (LPGi) in the pontomedullary reticular formation (RF). These brainstem nuclei are modeled as populations of excitatory (glutamatergic) and inhibitory (GABAergic/glycinergic) neurons.
The model proposes two distinct glutamatergic pathways originating from the LPGi:
- LPGi-Glu-1: Relays excitation from CnF and PPN to spinal rhythm generators (RGs), controlling locomotor frequency and speed.
- LPGi-Glu-2: Relays excitation from CnF to spinal interneurons that inhibit commissural interneurons (CINs) and long propriospinal neurons (LPNs), controlling gait.
The model uses an "activity-based" approach for neuron populations, where the average membrane potential is described by differential equations. Synaptic connection weights were adapted from previous work and tuned to reproduce experimental data. Simulations were performed by applying controlled excitatory drives to specific brainstem populations to mimic experimental manipulations (e.g., optogenetic stimulation).
Datasets
Not discussed in the paper. The study relies on existing experimental data from the literature for model validation.
Key Insights and Ideas
- Dual Pathway Control: The model strongly suggests that brainstem control of locomotion is mediated by two parallel pathways: one for speed (via RGs) and one for gait (via CINs and LPNs). This provides a clear mechanistic hypothesis for how these two aspects of locomotion are independently yet coordinately controlled.
- Differential Roles of MLR Nuclei: The model highlights distinct roles for the CnF and PPN. CnF activation can drive a wide range of speeds and gaits, while PPN activation is primarily associated with slower, alternating gaits. This aligns with experimental findings suggesting the CnF is more involved in fast, escape-like locomotion.
- LPGi as a Key Relay: The LPGi in the RF emerges as a critical relay station, integrating signals from the MLR and distributing them to the spinal cord via distinct pathways for speed and gait.
- Inhibitory Neuron Modulation: Inhibitory neurons in the brainstem (CnF, PPN, LPGi) are shown to play a crucial role in modulating and even stopping locomotion, with different nuclei exhibiting distinct effects. For instance, CnF and LPGi inhibitory neurons can halt locomotion, while PPN inhibitory neurons primarily reduce speed.
Key Results
- CnF Stimulation: Unilateral stimulation of glutamatergic CnF neurons in the model successfully reproduced a wide range of locomotor frequencies and gait transitions from walk to trot, gallop, and bound, consistent with experimental observations.
- PPN Stimulation: Unilateral stimulation of glutamatergic PPN neurons resulted in lower locomotor frequencies and only alternating gaits (walk and trot), without inducing synchronous gaits.
- PPN Inactivation: Inactivating the PPN shifted the transition from alternating to synchronous gaits to lower locomotor frequencies, suggesting the PPN plays a role in gating the transition to faster gaits.
- LPGi Stimulation: Unilateral stimulation of LPGi glutamatergic neurons mimicked the effects of CnF stimulation, initiating locomotion across a range of frequencies and gaits.
- Inhibitory Neuron Effects: Simulations showed that activating inhibitory neurons in the CnF, PPN, or LPGi progressively reduced locomotor frequency and induced orderly gait transitions towards slower gaits, with CnF and LPGi inhibition capable of halting locomotion.
Limitations and Future Work
- Model Simplification: The model uses "activity-based" neuron populations rather than detailed spiking models, which simplifies analysis but may miss nuances of neural dynamics.
- Limited Brainstem Structures: The model focuses on MLR and LPGi, acknowledging that other brainstem circuits might also contribute to locomotion.
- No Biomechanics: The model does not incorporate biomechanical constraints or sensory feedback, which are important for real-world locomotion.
- Future Work: The authors suggest developing simplified network models for systems-level mathematical analysis, incorporating more brainstem circuits, and integrating biomechanical models and sensory feedback. They also highlight the need for experimental validation of the proposed pathways and mechanisms.
Summary updated