The role of thalamic inhibitory interneurons in sensory processing

The thalamus has traditionally been described as a simple relay station for sensory information directed to the cortex. However, this simplistic view, which implies that complex information processing is restricted to the cortex, has been questioned by a growing number of researchers (Basso et al., 2005). Recent work supports a revised model that views the thalamus as critical for information processing (Groh et al., 2014; Rompani et al., 2017; Sherman, 2016) and as a coordinator of cortico-thalamo-cortical loops during on-going cortical computations (Yu et al., 2006). For instance, experiments that interfered with thalamic function specifically in the time window between cortical acquisition of sensory information and the generation of a behavioural output revealed a clear deterioration of behavioural performance (Bolkan et al., 2017; Guo et al., 2017; Schmitt et al., 2017). While it is now being appreciated that the thalamus plays a major role in information processing, the historic lag in research on thalamic cellular organisation has hampered efforts to fully understand the mechanisms and extent of this contribution. In particular, while it is clear that diverse classes of GABA interneurons (INs) (Fishell and Rudy, 2011) are of fundamental importance for information processing in cortical microcircuitry (Harris and Shepherd, 2015) and that their dysfunction leads to brain pathologies (Marin, 2012), far less is known about the complexity of thalamic inhibition, let alone explaining how this could contribute to information processing within the thalamo-cortical axis or dysfunctions in sensory perception, attention and psychoses.


The research in this project will use an established Sox14-Cre recombinase knockin mouse as a major experimental resource. All experiments will rely upon the progeny of a cross between heterozygous Sox14-Cre males and females. Because this is a knockin line, the homozygous Sox14-Cre mice do not express Sox14 and therefore do not develop thalamic interneurons, as well as other populations of Sox14+ inhibitory neurons throughout the brain. These mice will be compared, blind to genotype, with their wild-type littermates in several experiments targeting visual function as a first pass assessment of the importance of Sox14+ neurons. The remaining ~50% of the progeny will be Cre-expressing heterozygotes that have functional Sox14+ neurons throughout the brain that express Cre recombinase (as established by the Delogu lab). These mice will be used for experiments targeting just the inhibitory neurons in the primary thalamic nucleus of the visual system - the dorsal lateral geniculate nucleus (dLGN) - using viral vectors to express chemogenetic or optogenetic actuators. In all cases, the experimental design will assess Cre-heterozygotes in which Cre-dependent AAV viral vectors have been used to selectively express the actuator restricted to dLGN inhibitory neurons as a more targeted approach to dLGN inhibitory neurons. For an ideal control we will always use an empty AAV vector of the same serotype and expressing the same promoter to conditionally express just a matched fluorophore in the same cells. Again, the experiments will be conducted blind to treatment, so the controls will receive the same protocol of drug delivery or light stimulation, which is essential to control for potential off-target effects of drug or light (of particular importance in the visual system). Group sizes will be ~15 mice for each group in all these experiments, in line with power calculations made for supporting grants for the Delogu lab and the Cooke lab.