The mediodorsal nucleus of the thalamus (MD) is reciprocally connected with the prefrontal cortex (PFC), and although the MD has been implicated in a range of PFC-dependent cognitive functions (Watanabe and Funahashi, 2012; Mitchell and Chakraborty, 2013; Parnaudeau et al., 2018), little is known about how MD neurons in the primate participate specifically in cognitive control, a capability that reflects the ability to use contextual information (such as a rule) to modify responses to environmental stimuli. To learn how the MD-PFC thalamocortical network is engaged to mediate forms of cognitive control that are selectively disrupted in schizophrenia, we trained male monkeys to perform a variant of the AX continuous performance task, which reliably measures cognitive control deficits in patients (Henderson et al., 2012) and used linear multielectrode arrays to record neural activity in the MD and PFC simultaneously. We found that the two structures made clearly different contributions to distributed processing for cognitive control: MD neurons were specialized for decision-making and response selection, whereas prefrontal neurons were specialized to preferentially encode the environmental state on which the decision was based. In addition, we observed that functional coupling between MD and PFC was strongest when the decision as to which of the two responses in the task to execute was being made. These findings delineate unique contributions of MD and PFC to distributed processing for cognitive control and characterized neural dynamics in this network associated with normative cognitive control performance.

Wound closure in the Drosophila larval epidermis mainly involves nonproliferative, endocyling epithelial cells. Consequently, it is largely mediated by cell growth and migration. We discovered that both cell growth and migration in Drosophila require the cochaperone-encoding gene cdc37. Larvae lacking cdc37 in the epidermis failed to close wounds, and the cells of the epidermis failed to change cell shape and polarize. Likewise, wound-induced cell growth was significantly reduced, and correlated with a reduction in the size of the cell nucleus. The c-Jun N-terminal kinase (JNK) pathway, which is essential for wound closure, was not typically activated in injured cdc37 knockdown larvae. In addition, JNK, Hep, Mkk4, and Tak1 protein levels were reduced, consistent with previous reports showing that Cdc37 is important for the stability of various client kinases. Protein levels of the integrin β subunit and its wound-induced protein expression were also reduced, reflecting the disruption of JNK activation, which is crucial for expression of integrin β during wound closure. These results are consistent with a role of Cdc37 in maintaining the stability of the JNK pathway kinases, thus mediating cell growth and migration during Drosophila wound healing.

Studies of the brain promise to be revolutionized by new experimental strategies that harness the combined power of optical techniques and genetics. We have mapped the circuitry of the mouse brain by using both optogenetic actuators that control neuronal activity and optogenetic sensors that detect neuronal activity. Using the light-activated cation channel, channelrhodopsin-2, to locally photostimulate neurons allows high-speed mapping of local and long-range circuitry. For example, with this approach we have mapped local circuits in the cerebral cortex, cerebellum and many other brain regions. Using the fluorescent sensor for chloride ions, Clomeleon, allows imaging of the spatial and temporal dimensions of inhibitory circuits in the brain. This approach allows imaging of both conventional "phasic" synaptic inhibition as well as unconventional "tonic" inhibition. The combined use of light to both control and monitor neural activity creates unprecedented opportunities to explore brain function, screen pharmaceutical agents, and potentially to use light to ameliorate psychiatric and neurological disorders.