Research

The accumbens receives excitatory input from the ventral hippocampus, basolateral amygdala, midline thalamus and prefrontal cortex. Comparing these pathways side by side, Jon's postdoctoral work found that hippocampal input is concentrated in the medial shell, and that activating any one of them is enough to reinforce behaviour. In the lab, fiber photometry from pathway-defined axons showed that the rostral and caudal shell are innervated by distinct, intermingled populations of forebrain neurons, and that each input carries mixed, opposing signals during reward seeking.

Consumption coincides with a coordinated, network-wide reduction in excitatory drive to the rostral accumbens shell, and reducing input from individual pathways is enough to promote feeding. Measuring palatability from the microstructure of licking, the lab found that low-frequency stimulation of ventral hippocampal input to the shell enhances how palatable a food is, independently of opioid signalling, and can condition a flavour preference. The same lick-microstructure approach shows that alcohol and alcohol-paired flavours gain palatability with repeated drinking.

Bidirectional optogenetic manipulation of dopamine neurons during a trial-based choice task showed that cue-evoked dopamine helps maintain cue–value associations but does not encode expected value. More recent work finds that dopamine in the medial accumbens shell reinforces actions while simultaneously limiting how long they last, through D1 receptor signalling.

Efficient reward seeking means knowing when not to respond. Direct and indirect pathway neurons of the accumbens are both more active when behaviour is suppressed, and inhibiting either one increases unproductive responding. Amygdala and thalamic inputs, which innervate different accumbens subregions, are both required as well, pointing to complementary rather than opponent processes.

Axon-terminal silencing with archaerhodopsin is widely used to test pathway-specific contributions to behaviour. The lab showed that it also increases spontaneous transmitter release and recruits local inhibition, producing the same behavioural effects regardless of the pathway targeted, and that cell-body inhibition is a more specific alternative. A review lays out all-optical approaches that record and manipulate circuits at the same time.