Studying Anxiety Behavior in Mice Using Cell Type-Specific Diazepam

Abstract

Anxiety disorders affect millions worldwide, leading to profound personal and societal impacts. Pharmacotherapy is a common treatment, but existing medications often fail to work for all patients and have undesirable side effects. To advance the field, an improved understanding of the neurobiological basis of anxiety is necessary. This study focuses on the unique ability of benzodiazepines to reduce anxiety in manner that tightly corresponds to the concentration of drug in the brain. This rapid mode of action suggests that benzodiazepines can provide direct insights into the neurobiological basis of anxiety. Multiple studies in humans and rodents indicate that the behavioral effects of benzodiazepines correlate strongly with their impact on neural activity in the basolateral amygdala (BLA). However, these correlative studies leave many unanswered questions. First, it remains unclear whether changes in BLA activity are caused by benzodiazepines binding directly to cells in the BLA, or via a convergence of altered neural activity from other brain regions. Second, it remains unclear whether altered BLA activity causes or merely reports behavioral changes. Finally, because systemic diazepam has a complex set of behavioral effects, it remains unclear whether the BLA contributes most to the desired anxiolytic effects versus other off-target behavioral effects such as increased impulsivity. Addressing these questions requires a causal test of benzodiazepine effects localized to the BLA. Traditionally, this has been explored via micro-infusion of drugs into the BLA and other brain regions in animal models. Despite numerous studies, interpretation has been equivocal due to drug diffusion beyond the target area, a particular concern given the ubiquitous expression of the benzodiazepine-sensitive GABAA receptors. Here, we use an innovative technology, DART (drug acutely restricted by tethering), to precisely target diazepam (a benzodiazepine) to the BLA. We express an AAV viral vector in the BLA which enables cells to capture and concentrate diazepam.1DART.2 to levels ~1000 times higher than anywhere else. This approach achieves a potent and stable drug effect that is not subject to diffusion. Slice electrophysiology demonstrates that diazepam.1DART.2 has similar effects on GABAA receptors as regular diazepam, but with selectivity for BLA neurons targeted by DART. In the elevated plus maze (EPM), diazepam.1DART.2 applied specifically to the BLA significantly increased open-arm exploration, consistent with reduced anxiety. This establishes a role for the BLA in mediating the anxiolytic effects of diazepam. However, in comparison to the BLA-specific effects of diazepam.1DART.2, the behavioral effects of systemic diazepam were stronger and appeared qualitatively multifaceted. To quantify these observations, we developed a Markovian transition analysis of EPM behavior, which first accounts for overall locomotor effects of a drug treatment and then analyzes the transition rates between distinct zones of the EPM. This analysis provides finer-scale information, allowing behavioral approach and avoidance to be independently quantified. Remarkably, we found that diazepam.1DART.2 in the BLA recapitulates the full effect of diazepam with regard to open-arm avoidance. By contrast, diazepam.1DART.2 in the BLA had only a partial effect on open-arm approach. Taken together, our findings establish that the BLA explains the majority of diazepam’s ability to reduce open-arm avoidance, commensurate with the desired anxiolytic effects of diazepam. Systemic diazepam has additional behavioral effects that likely arise from other brain regions, and which may reflect undesired effects on behavioral impulsivity. Better anxiolytic medications may be realized by drug development aimed at engaging specific brain circuits shown to mediate clinically desirable effects. Such as the anti-avoidance effect of benzodiazepine, which appears to be fully mediated by the BLA.