Questions: Cerebellar Circuits and Function

Question 1

A researcher optogenetically silences all Purkinje cells in one region of a mouse's cerebellar cortex. What would you predict about the deep cerebellar nuclear neurons in that region and the resulting motor output?

• A. Deep cerebellar nuclear neurons decrease firing, producing reduced or poorly timed motor output
• B. Deep cerebellar nuclear neurons increase firing, because tonic inhibition is removed — the animal likely shows hypermetric or uncoordinated movements
• C. Deep cerebellar nuclear neurons are unaffected, because Purkinje cells only project to the brainstem
• D. Deep cerebellar nuclear neurons increase firing and movement becomes smoother and more coordinated

Answer: (B)

Purkinje cells tonically inhibit deep cerebellar nuclei via GABA. Silencing Purkinje cells removes this tonic inhibition, causing deep cerebellar nuclear neurons to fire more. The result is not improved movement but dysmetric, uncoordinated output — the cerebellum's role is to sculpt and time movement by selectively disinhibiting (allowing activity when appropriate) rather than simply activating. This disinhibitory logic means movement occurs when Purkinje cells pause, not when they fire. Option D gets the firing direction right but misinterprets the functional consequence — removing all inhibition indiscriminately produces uncoordinated hyperactivity, not smooth coordination.

Question 2

What is the proposed function of climbing fiber input to Purkinje cells in cerebellar learning?

• A. To carry high-frequency sensory updates about limb position that Purkinje cells integrate with motor commands
• B. To excite Purkinje cells strongly during correct, well-executed movements, reinforcing those patterns
• C. To signal a movement error, triggering long-term depression at the parallel fiber synapses that were active during the error
• D. To synchronize Purkinje cell firing across large populations during complex motor sequences

Answer: (C)

The climbing fiber is thought to be the error signal in Marr-Albus-Ito models of cerebellar learning. It originates from the inferior olive and fires when the executed movement does not match the intended movement. When a climbing fiber fires simultaneously with active parallel fibers (the 'teaching signal' coinciding with the 'student'), long-term depression (LTD) is induced at those parallel fiber–Purkinje cell synapses. This weakens that combination of inputs over time, shaping the Purkinje cell's response to produce more accurate motor output on future attempts. Option B reverses the learning logic — the climbing fiber fires during errors, not successes.

Question 3

Purkinje cells are inhibitory GABAergic neurons, and cerebellar output to downstream motor structures occurs primarily through disinhibition — when Purkinje cells pause firing, deep cerebellar nuclei are released to activate downstream targets.

• T. True
• F. False

Answer: True

This is the defining architectural feature of the cerebellar circuit. Purkinje cells are the sole output of the cerebellar cortex, and they continuously inhibit deep cerebellar nuclear (DCN) neurons. Motor commands emerge when Purkinje cell firing decreases — releasing DCN neurons from inhibition so they can fire and drive downstream targets. This disinhibitory logic is counterintuitive (more inhibition from Purkinje cells means less output, not more), but it gives the cerebellum fine control over the timing and amplitude of movements by adjusting when and how strongly Purkinje cells suppress the DCN.

Question 4

Climbing fibers are the primary source of convergent input onto Purkinje cells, with each Purkinje cell receiving contacts from thousands of climbing fiber axons from the inferior olive.

• T. True
• F. False

Answer: False

This reverses the actual architecture. Parallel fibers (axons of granule cells) are the source of massive convergence onto Purkinje cells — each Purkinje cell receives input from approximately 100,000 to 200,000 parallel fibers. Climbing fibers are the opposite: each Purkinje cell receives input from exactly ONE climbing fiber, which makes large, powerful synaptic contacts that produce the distinctive complex spike. The one-to-one climbing fiber arrangement is what makes it an appropriate error signal channel — it delivers a specific, unambiguous teaching signal rather than contributing to the statistical representation sampled by parallel fibers.

Question 5

Why is the ratio of parallel fiber inputs to climbing fiber inputs on a single Purkinje cell (approximately 100,000:1) important for how the cerebellum learns from errors?

Model answer: The extreme asymmetry reflects the different computational roles of the two inputs. The 100,000+ parallel fibers (from granule cells) provide a high-dimensional, combinatorial representation of the body's current state — sensory context, movement in progress, environmental conditions. Each parallel fiber synapse is individually weak, so the Purkinje cell integrates a statistical 'vote' across a vast input space. The single climbing fiber carries a completely different signal: a specific, all-or-nothing error signal from the inferior olive when the movement goes wrong. When the climbing fiber fires simultaneously with a set of active parallel fibers, it selectively weakens those parallel fiber synapses (LTD), teaching the Purkinje cell to respond differently to that exact pattern of inputs in the future. The 100,000-to-1 ratio means the system can make fine-grained adjustments — there are 100,000 adjustable weights per Purkinje cell, all tunable by a single error signal. If the ratio were reversed, learning would be coarse-grained and rapidly saturated.

This architecture implements something close to supervised learning in neural hardware: the parallel fibers are the input features, the Purkinje cell is the classifier, and the climbing fiber is the teacher providing error-corrective feedback. The large number of parallel fiber synapses gives the system enormous representational capacity; the single climbing fiber gives precise, targeted correction. Together, they allow the cerebellum to learn highly specific motor refinements from experience.