In vivo optogenetic strategies have redefined our ability to assay how neural circuits govern behavior. Although acutely implanted optical fibers have previously been used in such studies, long-term control over neuronal activity has been largely unachievable. Here we describe a method to construct implantable optical fibers to readily manipulate neural circuit elements with minimal tissue damage or change in light output over time (weeks to months). Implanted optical fibers readily interface with in vivo electrophysiological arrays or electrochemical detection electrodes. The procedure described here, from implant construction to the start of behavioral experimentation, can be completed in approximately 2-6 weeks. Successful use of implantable optical fibers will allow for long-term control of mammalian neural circuits in vivo, which is integral to the study of the neurobiology of behavior.

Forming and breaking associations between emotionally salient environmental stimuli and rewarding or aversive outcomes is an essential component of learned adaptive behavior. Importantly, when cue-reward contingencies degrade, animals must exhibit behavioral flexibility to extinguish prior learned associations. Understanding the specific neural circuit mechanisms that operate during the formation and extinction of conditioned behaviors is critical because dysregulation of these neural processes is hypothesized to underlie many of the maladaptive and pathological behaviors observed in various neuropsychiatric disorders in humans. The medial prefrontal cortex (mPFC) participates in the behavioral adaptations seen in both appetitive and aversive-cue-mediated responding, but the precise cell types and circuit mechanisms sufficient for driving these complex behavioral states remain largely unspecified. Here, we recorded and manipulated the activity of parvalbumin-positive fast spiking interneurons (PVϩ FSIs) in the prelimbic area (PrL) of the mPFC in mice. In vivo photostimulation of PVϩ FSIs resulted in a net inhibition of PrL neurons, providing a circuit blueprint for behavioral manipulations. Photostimulation of mPFC PVϩ cells did not alter anticipatory or consummatory licking behavior during reinforced training sessions. However, optical activation of these inhibitory interneurons to cues associated with reward significantly accelerated the extinction of behavior during non-reinforced test sessions. These data suggest that suppression of excitatory mPFC networks via increased activity of PVϩ FSIs may enhance reward-related behavioral flexibility.