Abstract
Substance use disorder (SUD) is a common and debilitating condition characterized by compulsive use of an
addictive substance, inability to control the use of this substance, and the emergence of withdrawal symptoms
in the absence of the substance. Despite great advances in our understanding of changes that occur in various
brain regions in the context of addiction/SUD, there is a limited understanding of the diversity of cell types and
gene expression profiles of cells within these human brain regions, and how exposures to addictive substances
such as opioids and cocaine influence the molecular properties and functions of these cells. Over 50% of HIV-
infected patients experience neurological disorders collectively termed HIV-Associated Neurocognitive
Disorders (HAND), which ranges from the asymptomatic neurocognitive impairment to severely disabling
dementia. The use of addictive substances by HIV-infected individuals has been linked to diminished immune
function, increased neuroinflammation and neuronal injury, and exacerbation of HAND. Here, we seek to
directly dissect the common and distinct molecular bases of SUD/HIV infection/HAND effects on distinct cell
types by systematic profiling, dissection, computational integration, and experimental validation of their
transcriptional, epigenomic, and genetic signatures across individuals, brain regions, and cell types. Aim 1: We
use genetic, epigenomic, and transcriptional profiles, generating a total of ~28 million genome-wide maps at
the single-cell (sc) level using 2,800 samples with 10,000 cells per sample; these span 7 brain regions
(prefrontal cortex, nucleus accumbens, ventral tegmental area, dorsal striatum, insula, amygdala,
hippocampus), two assays (scRNA, scATAC), four phenotypic groups (SUD+/HIV+, SUD+/HIV-, SUD-/HIV+,
SUD-/HIV-), and a total of 200 subjects (50 in each phenotypic group). Aim 2: We integrate these datasets to
predict driver genes, regulatory regions, variants, and pathways, and the cell types and brain regions where
they act. Aim 3: We validate high-priority findings at the molecular level, and functionally validate the highest
priority targets’ ability to modulate addictive behaviors in mouse models of in vivo cocaine self-administration.
The resulting datasets will help guide the search for new therapeutics, by providing detailed therapeutic
targets, and the specific conditions where they are predicted to act.