23 Sep Hebrew University Study Discovers How Cocaine Hijacks the Brain’s Action-Selection Circuit
MedicalResearch.com Interview with:
Dr. Ben Jerry Gonzales
Dr. Ben Jerry Gonzales PhD
Mr. Itay Shalom
Prof. Ami Citri.jpg
Mr. Itay Shalom
Prof. Ami Citri PhD
Citri Laboratory, Hebrew University of Jerusalem
When mice were repeatedly exposed to high doses of cocaine in a new study from the Hebrew University of Jerusalem, their diverse behavioral repertoire progressively narrowed until a single action dominated: licking the floor and walls of the arena. By the fifth day of exposure, the animals spent approximately 62% of their time performing this one repetitive behavior. The research, published in Current Biology, identifies the striatal circuit mechanism driving this behavioral capture and demonstrates that manipulating the circuit can interrupt cocaine-induced rigidity within a second. According to the NIDA (National Institute on Drug Abuse), cocaine exerts its behavioral effects primarily through dopaminergic circuits in the basal ganglia, and understanding how those circuits translate drug exposure into compulsive, repetitive behavior is a central question in addiction neuroscience.
MedicalResearch.com: What is the background for this study? What types of behaviors may be hijacked by cocaine?
Gonzales, Shalom, and Prof. Citri: Animals constantly select actions appropriate to their current needs and environment and switch between them as circumstances change. When this ability breaks down, behavior can become rigid and repetitive. Behavioral rigidity and repetitive actions are features of several psychiatric and neurological conditions, including compulsive disorders, Tourette syndrome, Parkinson’s disease, and drug-induced movement disorders, but the circuit mechanisms that transform flexible action selection into persistent repetition remain poorly understood.
A central structure for action selection is the striatum, part of the basal ganglia. Its lateral portion is organized roughly as a map of the body, with the ventrolateral striatum (VLS) particularly associated with movements of the mouth and tongue. Classic studies showed that applying psychostimulants directly to this region can produce repetitive oral behaviors. We wanted to know how the same circuit participates in normal behavior and what happens when cocaine drives behavior toward rigidity.
In our study, the behavioral domain captured by cocaine was orofacial. Before cocaine exposure, mice moved among many actions — exploration, rearing, grooming, and body licking — with relatively little licking of the floor and walls. With repeated high-dose cocaine exposure, this repertoire progressively narrowed toward persistent surface licking. By the fifth day, mice spent approximately 62% of their time licking the floor or walls. Importantly, cocaine does not appear to invent a new behavior. The same VLS circuit participates in natural orofacial actions. Cocaine instead causes an existing action channel to dominate at the expense of other behavioral options.
[HEADSHOT: Dr. Ben Jerry Gonzales — Credit: Hyewon Kim]
MedicalResearch.com: Can you describe the program you used to monitor these changes?
Gonzales, Shalom, and Prof. Citri: We developed a deep-learning platform called STEREO to describe behavior in terms that are immediately recognizable to a human observer — exploring, rearing, grooming, body licking, or licking of the surfaces of the arena. This was important because conventional behavioral experiments often measure only one or two predetermined variables, such as distance traveled. Many automated approaches instead track body points and subsequently infer behavioral states from patterns of movement. STEREO directly classifies semantically meaningful actions from video while also taking their temporal context into account.
The scale was also important: the cocaine dataset alone contained approximately 13.7 million video frames from 256 sessions. STEREO gave us a continuous, frame-by-frame description of what the animals were doing, with performance comparable to the agreement between independent human observers. We could then align those behavioral transitions with neural activity and precisely timed circuit manipulations. That changed the question we could ask. Rather than simply asking whether cocaine increased licking or locomotion, we could watch the entire behavioral repertoire reorganize over time and see one class of actions progressively displace the others.
MedicalResearch.com: What happened when the cocaine was withdrawn?
Gonzales, Shalom, and Prof. Citri: We did not conduct a long-term withdrawal or abstinence study, so we want to be careful about the term “withdrawal.” We did, however, ask whether the behavioral reorganization outlasted the acute effects of cocaine. Twenty-four hours after the final cocaine exposure, the animals received saline instead and were returned to the same arena. Surface licking remained elevated relative to their original pre-cocaine baseline. This tells us that the behavioral change was not simply an acute effect present only while cocaine was acting. What we cannot yet distinguish is why it persisted — repeated cocaine could have produced lasting changes in the relevant neural circuits, or the animals could have learned an association between the environment and the effects of cocaine. Testing animals in an environment that had never been paired with cocaine would help distinguish these possibilities.
MedicalResearch.com: Do you think the same effects are found with other drugs?
Gonzales, Shalom, and Prof. Citri: We tested cocaine, so we cannot say that the same mechanism applies to other drugs. There is, however, good reason to ask the question for other psychostimulants. Classical studies showed that amphetamine delivered directly into the ventrolateral striatum, or dopaminergic stimulation within this region, can also produce repetitive oral behaviors in rodents. Our results suggest a possible circuit framework for understanding those observations: sufficiently strong dopaminergic stimulation may disproportionately recruit the striatal territory controlling orofacial actions. Importantly, dose matters — in our study, a moderate cocaine dose predominantly recruited the dorsomedial striatum, whereas the high dose additionally recruited the ventrolateral striatum. It will also be interesting to ask whether related circuit principles operate in neurological or medication-induced movement disorders characterized by repetitive movements, including levodopa-induced and tardive dyskinesias. Their causes and pharmacology are very different from cocaine exposure, so this remains a question rather than a conclusion from our study.
MedicalResearch.com: What should readers take away from your report?
Gonzales, Shalom, and Prof. Citri: The central message is that rigid behavior can emerge through the capture of a circuit that normally participates in natural action selection. In our mice, repeated cocaine progressively restricted a diverse behavioral repertoire until orofacial actions dominated. We also found that the two major striatal pathways exert opposing causal influences over these actions. Both direct- and indirect-pathway neurons were recruited around licking, including during natural licking. But when we manipulated them selectively, increasing direct-pathway activity promoted orofacial actions, whereas increasing indirect-pathway activity rapidly interrupted cocaine-induced repetitive licking and allowed alternative behaviors to re-emerge. Conversely, suppressing the direct pathway reduced cocaine-induced licking, while suppressing the indirect pathway intensified it.
This distinction is important. We are not simply observing a neural signal that correlates with stereotyped behavior — manipulating the circuit can push behavior in either direction. The study was performed in mice receiving experimenter-administered cocaine. It is therefore a model of cocaine-induced behavioral rigidity, not a model demonstrating human addiction or compulsivity. Whether analogous forms of “channel capture” contribute to repetitive behavior in people is an important question for future research.
MedicalResearch.com: What recommendations do you have for future research as a result of this study?
Gonzales, Shalom, and Prof. Citri: The major next question is how one action channel comes to dominate behavioral control. Upstream, we want to understand how dopamine changes the cortical information reaching the striatum: does dopamine simply change overall striatal activity, or can it change which cortical inputs are able to influence action selection? Downstream, we are interested in following different somatotopic action channels through the basal ganglia — do channels representing different behavioral domains remain functionally separate, do they compete with one another, and where do their signals eventually converge? The body-map organization of the striatum also gives us a testable broader hypothesis: if the principle we describe is general, repetitive actions involving other body domains may preferentially recruit the corresponding striatal channels. Approaches such as STEREO should make it possible to test these ideas without restricting analysis to a single predetermined behavior.
MedicalResearch.com: Is there anything else you would like to add?
Gonzales, Shalom, and Prof. Citri: One aspect of the study that is particularly striking is how rapidly circuit manipulation can alter behavior. Activating the indirect pathway can interrupt ongoing cocaine-induced surface licking within a second, and the effect reverses almost immediately when stimulation ends. The supplementary videos accompanying the paper show this, along with direct-pathway stimulation promoting orofacial actions and indirect-pathway stimulation interrupting natural body licking. The STEREO classifier is available to academic researchers upon reasonable request under a material transfer agreement through Yissum, the technology-transfer company of the Hebrew University of Jerusalem. Processed data and analysis code are publicly available through Figshare and GitHub.
Disclosures: Ami Citri, Itay Shalom, and Ben Gonzales declare competing interests related to the invention of the STEREO behavioral-analysis platform, which is the subject of U.S. Provisional Application No. 63/813,843. The Citri laboratory provides a behavioral-analysis service using STEREO to the contract research organization Pharmaseed; Ami Citri serves as a consultant to Pharmaseed, and Itay Shalom receives a supplementary stipend associated with this service. The other authors declare no competing interests.
Citation:
Gonzales BJ, Shalom I, Lipton DM, Turm H, Noble J, Festuccia M, Groysman M, Citri A. Opponent ventrolateral striatal circuits regulate behavioral flexibility and rigidity.
Current Biology. 2026. doi:10.1016/j.cub.2026.08.068
For a broader overview of how dopaminergic circuits in the basal ganglia shape addiction, compulsive behavior, and the neuroscience of action selection, see this MedicalResearch.com overview of dopamine and the basal ganglia — what the research shows about addiction and compulsive behavior.
Disclaimer: The information on MedicalResearch.com is provided for educational purposes only, and is in no way intended to diagnose, cure, or treat any medical or other condition. Some links are sponsored. MedicalResearch.com and Eminent Domains Inc. do not warrant or endorse products or claims made by third party links. Always seek the advice of your physician or other qualified health provider and ask your doctor any questions you may have regarding a medical condition. In addition to all other limitations and disclaimers in this agreement, service provider and its third party providers disclaim any liability or loss in connection with the content provided on this website.
Last Updated on September 23, 2026 by Marie Benz MD FAAD