Rochefort, Nathalie

Nolan, Matthew

Francioni, Valerio

2020-03-16T10:48:39Z

2020-03-16T10:48:39Z

2020-07-07

Cortical layer 5 excitatory neurons are characterized by long apical dendrites receiving inputs from multiple long-range cortical and subcortical connections. In vitro and in vivo recordings have shown that the dendrites of layer 5 pyramidal neurons support both distance-dependent filtering and local dendritic non linearities, including NMDA, calcium and sodiummediated spikes. Additionally, the coincident occurrence of back-propagating action potentials and tuft depolarization was shown to generate widespread calcium transients in the apical tuft dendrites which leads to bursts of action potentials in the soma. In the primary visual cortex (V1), layer 5 pyramidal neurons display selective responses to physical features of visual stimuli, such as the orientation and direction of movement. In addition, layer 5 neurons activity is gain modulated by locomotion both in darkness and during visual stimulation. This gain modulation was shown to be mediated through a recurrent VIP-SST cortical circuit which was suggested to produce a net disinhibition of the apical tuft dendrites of pyramidal neurons. So far however, the dendritic activity underlying gain modulation of layer 5 pyramidal neurons during locomotion remains unexplored. Additionally, the extent to which dendritic activity is compartmentalised from the activity in other sibling branches and from the activity in the soma is a matter of debate. In vitro studies suggest that apical tuft branches should be highly compartmentalised, however in vivo studies have returned controversial results about the extent of somato-dendritic coupling. To address these questions, I sparsely labelled layer 5 neurons of the primary visual cortex with a genetically-encoded calcium indicator (GCaMP6). I used multiplane, two-photon calcium imaging to monitor the activity in different apical tuft branches and different compartments of the neuron (soma, trunk and tuft) semi-simultaneously. I acquired data in head-fixed mice freely running on a cylindrical treadmill both in darkness and during the presentation of drifting gratings. Finally, I performed offline morphological reconstructions of the neurons imaged, in order to extract anatomical information about the neurons and dendrites I imaged. These results showed that the apical tuft dendrites increase their activity in response to visual stimulation and locomotion. However, I found that the activity of different sibling branches belonging to one neuron had highly correlated activity. Branch-specific events were rare, small, and independent of visual stimulation and locomotion. This high correlation persisted not only between different apical tuft branches, but also between different compartments of the neurons showing that dendritic calcium activity is systematically coincident with global events spreading throughout the entire neuron. Neither locomotion nor visual stimulation altered this high coupling between somatic and dendritic activity. However, the results showed that activity levels between soma and the apical tuft were asymmetric. While almost all dendritic events were detected in the soma, up to 40% of somatic events could not be detected in the apical tuft dendrites, suggesting that somatic signals attenuated from the soma to the apical tuft. Throughout all compartments, smaller events were more likely to decay below the detection threshold, suggesting that signals attenuated in a distance and amplitude-dependent manner from the soma to the apical tuft. These results provide important insights about the mechanisms of dendritic integration of individual layer 5 neurons in the visual cortex. They suggest that the entire neuron behaves as a single computational unit rather than many independent ones and suggest that activity in the compartments is largely driven by somatic action potentials regardless of the animal’s behavioural state. Nonetheless, the extent to which these findings apply to other neuronal types, other cortical areas and different behavioural and perceptual states will have to be determined by future experiments.

https://hdl.handle.net/1842/36879

http://dx.doi.org/10.7488/era/180

en

The University of Edinburgh

Francioni, V., Padamsey, Z. and Rochefort, N. L. (2019) ‘High and asymmetric somatodendritic coupling of v1 layer 5 neurons independent of visual stimulation and locomotion’, eLife. doi: 10.7554/eLife.49145.

Pakan, J. M., Francioni, V. and Rochefort, N. L. (2018) ‘Action and learning shape the activity of neuronal circuits in the visual cortex’, Current Opinion in Neurobiology, 52, pp. 88–97. doi: 10.1016/j.conb.2018.04.020.

dendrites

GCaMP6

layer 5

visual cortex

dendritic integration

Somato-dendritic coupling in layer 5 pyramidal neurons of the mouse primary visual cortex

Thesis or Dissertation

Doctoral

PhD Doctor of Philosophy