UC San Diego

The fundamental function of the brain is to produce useful movements. Movement has become progressively more directly controlled by descending systems through evolutionary history, culminating in the mammalian motor cortex which contains projections to the spinal cord. The activity of neurons within the motor cortex is highly heterogeneous, with different neurons being correlated with different aspects of movement. Consequently, activity within the motor cortex is most comprehensively described at the level of large neuronal populations. The motor cortex is also capable of substantial plasticity after injury and during learning, suggesting that the role of activity in guiding movements is flexible. A key component of understanding the function of the motor cortex therefore is to monitor changes in activity of motor cortex neuronal populations across learning.

The experiments described within this dissertation follow this approach primarily through the use of in vivo two-photon calcium imaging. Mice were trained in a forelimb motor learning task over the course of two weeks, and the activity of a consistent population of hundreds of neurons was recorded during each training session through calcium indicator fluorescence. Movements of the forelimb during each training session were continuously monitored by displacement of a lever, allowing for a comparison between activity patterns and corresponding movements.

Three separate neuronal populations were analyzed. These include excitatory pyramidal neurons in layer 2/3, inhibitory interneurons within layer 2/3, and corticospinal neurons within layer 5b. Layer 2/3 was found to be active primarily around movement, while corticospinal neurons were active either during movement or quiescence. Inhibitory interneurons were found to be relatively stable across recording, being primarily correlated with local excitatory neuron activity. Excitatory neurons in layer 2/3 and corticospinal neurons, on the other hand, were modulated by learning. This included an initial increase in the number of movement-related cells, followed by a decreased but stable population. The temporal activity of individual cells also changed across time but became more consistent late in learning. These dynamics resulted in a differing relationship between activity and movement across learning, suggesting that the influence of motor cortex activity on movement is shaped by learning.