Project Summary
Benzodiazepines (BZs) are effective and safe when used appropriately, but their utility is limited by unwanted
side effects like misuse and reduced safety when combined with other drugs. Individuals with substance-use
disorders (SUDs) misuse BZs as much as 20x greater than the general population, and the rising number of
overdose deaths attributed to BZs are largely driven by opioid co-administration. Given the high rates of polydrug
use among individuals who also misuse BZs, preclinical evaluations of BZs should consider the polydrug
scenarios in which they are being misused. The goal of this research is to identify pharmacological determinants
of BZ misuse, with the goal of identifying BZ-type ligands with reduced abuse potential in polydrug situations.
We reported previously that nonselective, partial-efficacy BZ ligands or those that lack intrinsic efficacy at a1-
subunit containing GABAA (a1GABAA) receptors have reduced abuse potential relative to traditional BZs.
However, our data suggest that the degree to which these BZ-type ligands exhibit abuse potential depends on
the subject’s drug history. This new application will use choice models to evaluate the overall hypothesis that
drug experience is a key determinant of the role of GABAA receptor subtypes in the abuse potential of BZ-type
ligands. A key finding from our choice research is that efficacy at a1GABAA receptors may be necessary for self-
administration of BZs in cocaine-experienced subjects, but not required for enhancement of cocaine choice. It is
unknown whether this pattern of effects is observed with other drugs of abuse, in particular opioids. Moreover,
the pharmacological mechanism underlying these effects is unknown, with possibilities including (1) a differential
role for a1GABAA receptors in reinforcement-enhancing vs. reinforcing effects of BZs alone, or (2) differences in
overall intrinsic efficacy, irrespective of subtype selectivity. We will address these potential mechanisms in two
Aims organized by unique approaches. In Aim 1, we will use drug vs. drug choice to evaluate the extent to which
BZ-type ligands varying in efficacy and selectivity will enhance or attenuate drug choice when delivered as a
combination with cocaine, heroin, or alprazolam in separate groups of subjects. In Aim 2, we will use drug vs.
nondrug choice to evaluate the hypothesis that nonselective, partial modulators or a1-sparing BZs will have
reduced reinforcing effects when delivered alone in subjects with a history of cocaine, heroin, alprazolam, or
food choice. A significant addition to our groups of monkeys with different reinforcement histories will be the
food-experienced animals, providing a quantitative model of drug-naïve individuals’ initial exposure to drug
taking. BZ use among individuals with a SUD, in or out of treatment, is increasingly common and is associated
with increased risk of BZ misuse and overdose. Testing the hypotheses proposed will provide critical information
for understanding how past and current drug use affects the abuse liability of BZ-type drugs.