Morphological Constraints on Cerebellar Granule Cell Combinatorial Diversity.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Cerebellar Physiology

Background:

• The cerebellar granule cell layer (GCL) is theorized to contribute to motor control and learning via combinatorial expansion.
• Granule cells (GrCs) are believed to form combinatorial codes from mossy fiber inputs for pattern separation and learning.

Purpose of the Study:

• To investigate how GCL architecture influences GrC combinatorial diversity using a spatially realistic model.
• To understand the impact of anatomical features like dendrite length and mossy fiber clustering on GrC input combinations.

Main Methods:

• Construction of a spatially realistic computational model of the cerebellar GCL.
• Analysis of GrC combinatorial diversity and redundancy under varying input conditions.
• Incorporation of novel anatomical measurements from mice to inform modeling of mossy fibers.

Main Results:

• GrC combinatorial diversity saturates with increasing mossy fiber input diversity due to short dendrites and local input sampling.
• Mossy fiber clustering and local sampling lead to redundant GrC input combinations.
• GCL architecture supports both combinatorial and temporal expansion of information, potentially aiding learned timing.
• Sparse and filopodia-bearing mossy fibers enhance GrC diversification and redundancy.

Conclusions:

• Anatomically realistic input patterns constrain combinatorial diversity in the GCL, producing redundant combinations.
• This redundancy may support temporal diversification of information, crucial for learned timing.
• The GCL's architecture provides a neural substrate for both combinatorial and temporal information processing in the cerebellum.