Neurostimulation, including invasive methods like deep brain stimulation (DBS), is an increasingly important approach to treating mental illness. It offers the possibility of directly targeting the circuit dysfunctions that produce mental disorders. The clinical use of brain stimulation, and DBS in particular, has been limited by a lack of mechanistic understanding. We do not know why/how stimulating a specific region or pathway leads to symptom improvement. This limits our ability to correctly “dose” stimulation, or to verify biological target engagement. Further, mechanisms are difficult to dissect in humans, because (A) we cannot easily capture stimulation’s effects on brain circuits and (B) human patients cannot tolerate the many repeated experiments needed to map stimulation parameter space. Detailed network/mechanism mapping is possible in animals, but common animal models of psychiatric illness are not strong mirrors of human disease. This project attempts to overcome those barriers by modeling psychiatric DBS’ mechanisms through the lens of cognitive function – behaviors that can be more rigorously measured in both humans and animals. Specifically, we recently showed that DBS of the ventral internal capsule/ventral striatum (VCVS), acts in part by improving patients’ executive function. In two human studies, we showed that VCVS DBS augments cognitive control – the ability to withhold a habitual/default response in favor of a more goal-aligned option. Further, this augmentation translates across species. We applied DBS-like stimulation to a rat homologue of VCVS, during a cognitive control task similar to our human paradigm. We saw improvements that very closely tracked our human results. We now propose to use that reverse-translational model to identify which cellular/circuit elements and neural activity variables mediate this behavioral improvement. A dominant theory argues that DBS in psychiatry works through white matter, e.g. by retrograde modulation of PFC through cortico-thalamic axons in VCVS. Our data suggest, however, that changes in local striatal activity are also important. We will replace electrical DBS with population-restricted optogenetic stimulation, mapping the contributions of individual cortico-thalamic circuits (Aim 1) and striatal sub-regions (Aim 2). During those manipulations, we will record spikes and LFP, at multiple sites within the cortico-striatal cognitive control circuitry. We will identify which variables are most strongly changed by DBS-like stimulation and correlate with behavior change (Aim 3). Success would identify putative mechanisms of a promising neurostimulation therapy, with near-term clinical implications. Technologies already exist to steer the DBS electric field to target specific axons/nuclei, and to titrate stimulation based on physiologic measures. The PI is a DBS psychiatrist, and could use those approaches in his own clinical program to target the mechanisms we identify.