Unraveling mechanisms underlying memory formation has been a challenge in neuroscience for many decades. In the new paper just published in eLife, Vincenzo Romano and colleagues studied the role of strengthening of synaptic inputs to Purkinje cells in learning new whisker reflexes. This adaptation process is important for rodents to explore their environment. The authors found that strengthening the parallel fiber to Purkinje cell input facilitates adaptation of the whisker reflexes and that this memory is expressed by long-lasting increases in Purkinje cell activity. The behavioural effects of this learning process can be mimicked by artificially increasing the Purkinje cell activity. These data suggest that strengthening of Purkinje cell synapses can lead to long-term changes in adaptation of reflexes by increasing their spiking output.
Most studies have been dedicated at finding cellular forms of plasticity that underlie induction of memory formation. For example, weakening synapses in cerebellum is instrumental in Pavlovian conditioning. This form of procedural learning is a form of learning that one can use to associate particular sensory inputs so as to improve motor behavior (e.g. Pavlov’s dogs learned to recognize a sound so they could produce saliva before food arrives). Ultimately, a learned behavior is expressed in neuronal activity. To what extent strengthening synapses in the cerebellum can also play a role in procedural memory formation and how this is expressed in neuronal activity when memory is read-out, is unclear. Moreover, it is also not clear to what extent the process of induction and expression can be correlated.
To date, all textbooks still claim that cerebellar learning is exclusively mediated through the weakening of synaptic inputs of Purkinje cells rather than strengthening them.
Because of the new cell physiological and systems data presented in the current paper a new picture of cerebellar function emerges: Strengthening of cerebellar synapses and enhancing cerebellar spiking output can underlie the induction and expression of procedural motor learning just as well as weakening of cerebellar synapses and enhancing cerebellar spiking output; the plasticity modus that will dominate depends on and can be deduced from the polarity of the downstream circuitry involved.