Ketamine repairs reward circuitry to reverse stress-induced anhedonia

A single low dose of the anesthetic ketamine restored the ability to enjoy sweet treats and social contact in mice that had been made apathetic by long-term stress, according to new research published in Neuron. The same injection also repaired weakened connections onto a specific group of reward-related brain cells. When those repaired connections were blocked, the animals’ recovery disappeared—suggesting the synaptic fix is a key part of ketamine’s sustained antidepressant effects.

Originally developed as an anesthetic in the 1960s, ketamine has more recently drawn attention for its fast-acting antidepressant properties. Over the past two decades, clinicians have found that sub-anesthetic doses can lift mood within hours in many people with major depression who do not respond to conventional medications. That rapid action stands in contrast to commonly prescribed drugs like selective serotonin reuptake inhibitors, which often take weeks to alleviate symptoms.

One symptom of depression that appears to respond especially well to ketamine is anhedonia—the loss of pleasure in normally rewarding experiences. Yet it remains unclear which brain circuits are responsible for the drug’s lasting effects on mood and motivation. The new study set out to identify those changes down to the level of individual synapses.

“Depression is a leading cause of morbidity and mortality worldwide, and it is projected to become the second leading cause of disability by 2030,” said study author Marco Pignatelli, an assistant professor of psychiatry at Washington University School of Medicine in St. Louis.

“Despite the urgent need to address depression as a public health priority, current pharmacotherapies require prolonged administration—weeks, if not months—for clinical improvement, and are often associated with high non-response rates. In contrast, a single sub-anesthetic dose of ketamine induces a rapid antidepressant effect in about 70% of treatment-resistant patients.”

Free daily newsletter

“Importantly, this improvement often occurs within the context of anhedonia,” he added. “Conventional antidepressants do poorly in relieving anhedonia, which is one of the two core symptoms used to diagnose major depression. That makes ketamine a promising and unique pharmacological option. However, without understanding the circuits and synapses that support this effect, our ability to design safer, more targeted medications is limited.”

To model anhedonia in mice, the researchers implanted slow-release corticosterone pellets under the animals’ skin to mimic chronic stress. Over three weeks, the hormone reduced the animals’ interest in sweetened water, decreased time spent with other mice, and lowered their willingness to work for rewards in a progressive ratio task where nose-pokes earned sugar pellets.

Twenty-four hours before each test, half the stressed mice received a single injection of ketamine at 10 milligrams per kilogram—a dose well below what’s used for anesthesia. The rest received saline, as did a group of unstressed control mice. Ketamine restored reward-seeking behavior across all tasks, while saline had no effect. The drug did not increase overall activity levels, ruling out general stimulation as an explanation.

To uncover what had changed in the brain, the researchers prepared thin slices from another group of similarly treated mice and recorded electrical activity in the nucleus accumbens, a key hub for processing reward. They focused on medium spiny neurons that express dopamine receptor type 1, a subtype linked to motivation and approach behavior.

Chronic stress reduced the strength and frequency of excitatory inputs to these neurons, but ketamine reversed this effect within a day—possibly as soon as one hour after injection. These changes were not observed in neighboring neurons that express dopamine receptor type 2, highlighting the cell-type specificity of the effect.

To test whether these restored synapses were necessary for ketamine’s behavioral benefits, the researchers used a molecular technique to block them. They introduced a HaloTag protein into dopamine receptor type 1 neurons, allowing them to tether a glutamate receptor blocker to just those cells. Delivering the blocker into the nucleus accumbens 24 hours after ketamine erased the previously observed improvements in sugar preference, sociability, and motivation. Blocking the same receptors on dopamine receptor type 2 neurons had no effect, pinpointing the importance of the type 1 cells.

The team then asked whether strengthening the same synapses—without ketamine—would be enough to reverse anhedonia. They inserted a light-sensitive version of Rac1, a protein that clusters glutamate receptors, into the dopamine receptor type 1 neurons. Mice exposed to blue light showed restored motivation and sociability, while exposure to red light had no effect. The findings suggest that targeted synaptic enhancement alone can substitute for ketamine in relieving anhedonia.

To identify where the restored signals were coming from, the researchers injected a light-activated protein into several brain areas that send input to the nucleus accumbens. They found that chronic stress weakened glutamatergic connections from the medial prefrontal cortex and ventral hippocampus—regions involved in decision-making and memory. Ketamine restored both pathways but did not affect inputs from the amygdala, thalamus, or ventral tegmental area.

Further analysis revealed distinct mechanisms for each pathway. The prefrontal input gained stronger unitary synaptic responses, while the hippocampal input showed both stronger responses and increased release events. These changes indicate different modes of adaptation at the two inputs.

To confirm that both inputs were required for ketamine’s effects, the team used a two-virus strategy to express a designer inhibitory receptor specifically in either the prefrontal or hippocampal projection to the nucleus accumbens. Administering a drug that silenced the targeted neurons before ketamine blocked the behavioral rescue.

Interestingly, the type of failure depended on which pathway was silenced. Turning off the ventral hippocampal input delayed the animals’ first approach to a social partner and the first attempt to obtain a reward. In contrast, silencing the prefrontal input did not affect initiation, but reduced overall engagement. These results suggest that hippocampal input helps trigger reward-seeking, while prefrontal input helps sustain it.

“We’ve uncovered a brain mechanism through which ketamine restores reward-related behavior after chronic stress in mice, and we believe these mechanisms are likely conserved across species,” Pignatelli said. “That makes this discovery relevant for clinical applications.”

But, as with all research, there are limitations. “It is always important to highlight that our mechanistic studies are taking place by using murine animal models,” Pignatelli noted. “In general, the findings of a study in mice may not directly translate to humans or other species, but they offer valuable insights into biological processes within that model.”

Future studies could use brain imaging to examine whether similar changes occur in patients who recover from anhedonia after ketamine treatment. Identifying specific synapses that support antidepressant effects may eventually allow researchers to design drugs that improve motivation without the dissociative side effects associated with ketamine.

“I hope that results from this body of work will have a sustained impact on the field by propelling the rational design of more targeted treatments, thereby facilitating more effective and safer therapies aiming at alleviating anhedonia,” Pignatelli said.

The study, “Ketamine rescues anhedonia by cell-type- and input-specific adaptations in the nucleus accumbens,” was authored by Federica Lucantonio, Jacob Roeglin, Shuwen Li, Jaden Lu, Aleesha Shi, Katherine Czerpaniak, Francesca R. Fiocchi, Leonardo Bontempi, Brenda C. Shields ,Carlos A. Zarate, Jr., Michael R. Tadross, and Marco Pignatelli.