TL;DR
This study uses fMRI in macaque monkeys to investigate how different sub-regions of the premotor cortex (F5) specialize for hand versus mouth movements. By comparing task-based and resting-state fMRI, they found evidence for distinct functional networks supporting hand and mouth actions, suggesting a specialized organization within F5 that aligns with anatomical and previous electrophysiological findings.
What Problem Does This Paper Address?
The ventral premotor cortex (area F5) in macaques is known to be involved in both hand and mouth movements. However, the precise functional organization and specialization of its sub-regions (F5p, F5c, F5a) for these different motor actions have been debated. While anatomical and single-neuron studies suggest a gradient of specialization, non-invasive fMRI studies using data-driven clustering haven't always captured this fine-grained functional distinction. This paper aims to clarify the functional specialization of F5 subfields by combining task-based fMRI (during actual hand and mouth movements) with seed-based resting-state fMRI (examining intrinsic functional connectivity) to see if these methods converge on similar conclusions about F5's role in hand versus mouth motor control.
Methods
The researchers employed two primary fMRI approaches:
- Task-based fMRI: Two monkeys were trained to perform specific motor tasks while in the scanner.
- Hand Movement Localizer: Monkeys performed a reach-and-grasp task in the dark, contrasted with a reach-only task.
- Mouth Movement (Taste) Localizer: Monkeys fixated on a screen and received different flavored liquids, with the contrast being sweet liquid versus artificial saliva. This task was chosen because it reliably elicits licking and jaw movements.
- Resting-state fMRI: All eight monkeys underwent resting-state scans.
- Seed-based Functional Connectivity Analysis: This involved defining specific "seed" regions (small spheres) within F5 subfields (F5p, dorsal F5c, ventral F5c, F5a) and other hand/mouth-related brain areas. The temporal correlation of the fMRI signal between these seeds and the rest of the brain was then calculated to map functional networks. They also performed seed-to-seed correlation analyses to examine how different F5 seeds related to each other and to broader hand/mouth motor networks.
- Multidimensional Scaling (MDS): This technique was used to visualize the clustering of functional relationships between the F5 seeds and other hand/mouth motor seeds in a 2D space, helping to reveal patterns of connectivity.
Data preprocessing included motion correction, co-registration to a template, and smoothing. Statistical analysis involved general linear models (GLM) for task fMRI and Fisher's r-to-z transformation for resting-state correlations, with group-level analyses performed.
Datasets
The study utilized fMRI data acquired from eight rhesus monkeys (Macaca mulatta). Two of these monkeys also participated in the task-based fMRI experiments. No new datasets were introduced; the study relied on data collected from these subjects.
Key Insights and Ideas
- Convergence of Task and Resting-State fMRI: A key finding is the remarkable agreement between task-based fMRI and resting-state functional connectivity in revealing the functional specialization of F5 subfields. This suggests that intrinsic brain networks observed during rest can reflect the functional roles of these areas during active behavior.
- Dorsal-Ventral Specialization within F5c: The study provides strong evidence for a functional gradient within F5c, with ventral F5c showing a clear preference for mouth-related functional connectivity and task responses, while dorsal F5c is more involved in hand-related actions.
- F5p and F5a Lean Towards Hand Actions: F5p consistently showed stronger functional correlations with hand-related networks, and task fMRI indicated its involvement in hand grasping. F5a also showed a tendency towards hand-related connectivity and task responses.
- Area 44's Distinct Profile: The area 44 seed, located in the fundus of the arcuate sulcus, showed a different functional connectivity profile compared to F5a, suggesting a distinct functional role, particularly with a stronger link to mouth-related networks in its anterior portion (44/GrFO).
- Seed-Based Resting-State fMRI as a Powerful Tool: The authors highlight that seed-based resting-state fMRI can reveal fine-grained functional organization that might be missed by purely data-driven clustering approaches.
Key Results
- Task fMRI:
- Hand grasping movements elicited significant responses in F5p, dorsal F5c, and F5a.
- Ingestive mouth movements (licking/swallowing) showed particularly pronounced responses in ventral F5c.
- Resting-State fMRI:
- F5p exhibited predominantly functional correlations with the hand-motor network.
- Ventral F5c showed strong functional correlations with the mouth-motor network.
- Dorsal F5c showed robust functional correlations with both hand- and mouth-motor networks.
- F5a showed functional correlations with hand-related areas, but also with some mouth-related areas.
- Area 44 showed a distinct connectivity profile, with the 44/GrFO seed showing strong correlations with mouth-related regions.
- Comparison of Seed-to-Brain Correlations: Ventral F5c had significantly stronger correlations with mouth areas than hand areas, while F5p showed the opposite. Dorsal F5c showed stronger correlations with mouth areas than hand areas on average.
Limitations and Future Work
- Mouth Movement Task: The mouth movement task involved both motor actions (licking, jaw movements) and taste perception. The authors acknowledge that the contrast between sweet and artificial saliva likely captured both motor and taste-related responses, which might influence the interpretation of purely motor specialization. Future work could involve tasks focusing solely on mouth movements without taste.
- Communicative vs. Non-Communicative Movements: The study focused on non-communicative mouth movements. The authors suggest that future studies could investigate communicative gestures (e.g., lip smacking) to further explore the functional repertoire of F5.
- Data-Driven vs. Seed-Based Approaches: While the study advocates for seed-based resting-state fMRI, they acknowledge that future data-driven approaches with more refined clustering might also reveal finer functional distinctions.
- Anatomical vs. Functional Connectivity: The study primarily relies on functional connectivity. While functional connectivity is constrained by anatomical connections, direct comparisons with tract-tracing data in the same animals would provide a more complete picture.
The authors suggest that future research could explore the role of these F5 subfields in more complex actions, including communicative gestures and the integration of sensory information with motor planning.
Summary updated