Dopamine’s role in fear extinction is more precise than we thought
A mouse study published in PNAS and reported by MIT’s Picower Institute shows how one dopamine pathway helps the brain learn that a once-dangerous place is now safe. The finding is causal, not just correlational, but it should not be mistaken for a ready-made treatment for PTSD or anxiety.
A place can change meaning.
A room where something painful happened can later be harmless. A street once linked with danger can become ordinary again. A sound, a smell, a context, a pattern of light—any of these can carry fear long after the threat is gone.
The brain’s problem is not only how to learn danger. It is how to stop treating the past as if it were still happening.
That process is called fear extinction. The name can be misleading. Extinction does not mean the original fear memory has been erased. It means the brain has learned something new: this context, here and now, no longer predicts harm.
A 2025 study in mice, “Dopamine induces fear extinction by activating the reward-responding amygdala neurons,” gives that process a sharper biological shape. The key signal is dopamine, but not dopamine in the vague pop-science sense of pleasure, motivation, reward, or addiction. Here, dopamine behaves more like a teaching signal. It helps the brain register that an expected threat did not arrive.
More precisely, the study identifies a dopamine pathway from the ventral tegmental area (VTA) to the posterior basolateral amygdala (pBLA) as a causal driver of fear extinction. (news.mit.edu)
That precision is the point.
This is not a story about “more dopamine equals less fear.” It is a story about one dopamine circuit acting on one amygdala population at the moment when the brain has to decide whether a place still means danger—or has become safe.
As co-author Michele Pignatelli di Spinazzola put it in MIT’s report, “Dopamine is essential to initiate fear extinction.” The important word is initiate. The signal seems to matter early, when the animal returns to a place where it expects something bad to happen, and nothing bad happens. (news.mit.edu)
That absence is not empty. It is information.
The amygdala does not store only fear
The basolateral amygdala is often treated as a fear center. That is too simple.
Earlier work from Susumu Tonegawa’s lab identified different neuronal populations in the basolateral amygdala that can assign different emotional meanings to experience. The 2020 Picower Institute article “With these neurons, extinguishing fear is its own reward” described how Ppp1r1b-expressing neurons in the posterior basolateral amygdala are involved in fear extinction and reward-related valence, while Rspo2-expressing neurons are linked to negative valence and fear-related memory. (Institutul Picower)
In simplified terms, the fear-linked population says: this place is dangerous.
The extinction-linked population says something different: this place is safe now.
This changes how we should think about extinction. Fear is not simply drained out of the brain like water from a tank. A new safety memory is built beside the old fear memory, and the two compete for control over behavior.
When the fear memory dominates, the animal freezes.
When the safety memory gains control, freezing drops.
The original fear association may still be there. It is just no longer the only interpretation available.
That matters for trauma, anxiety, and relapse. Fear can return because extinction is not deletion. It is a new layer of learning.
As former lab member Xiangyu Zhang explained in the MIT report: “Our study uncovers a precise mechanism by which dopamine helps the brain unlearn fear. We found that dopamine activates specific amygdala neurons tied to reward, which in turn drive fear extinction. We now see that unlearning fear isn’t just about suppressing it — it’s a positive learning process powered by the brain’s reward machinery. This opens up new avenues for understanding and potentially treating fear-related disorders, like PTSD.” (news.mit.edu)
The dopamine input is not random
The 2025 study asked what kind of signal helps these amygdala populations encode fear or safety. The answer was not simply “dopamine,” but dopamine arriving through different VTA-to-amygdala routes. (news.mit.edu)
According to MIT’s summary of the paper, the fear-linked Rspo2 neurons received dopaminergic input mainly from anterior and lateral VTA regions, while the extinction-linked Ppp1r1b neurons received denser dopaminergic projections from central and posterior VTA regions. The researchers also found that both cell populations expressed D1 dopamine receptors, with higher receptor expression in the Ppp1r1b neurons. (news.mit.edu)
In plain terms, the safety-coding neurons were especially well positioned to respond to this dopamine signal.
That anatomy fits the behavioral result.
The brain is not simply bathing the amygdala in dopamine and hoping for the best. The system is organized. Different VTA projections reach different amygdala populations, and those populations push behavior in different directions.
This is why the finding is interesting. The same molecule can help extinguish fear in one circuit and help preserve or revive fear in another.
Dopamine is not the message by itself.
The circuit is the message.
The decisive moment comes when danger fails to appear
The experiment used a three-day fear-conditioning and extinction design. On the first day, mice received mild foot shocks in a specific chamber, linking that context with danger. On the second day, they returned to the same chamber for an extinction session, but no shocks were delivered. On the third day, the researchers tested whether the animals had retained the extinction learning. (news.mit.edu)
During the initial shock learning, dopamine-related activity was stronger in the fear-linked Rspo2 neurons. That makes sense. The animal was learning that the context predicted threat.
But during the early part of the extinction session, the pattern shifted. The mouse returned to the chamber expecting danger, but the expected shock did not arrive. In that window, dopamine activity rose more strongly in the Ppp1r1b neurons—the neurons linked to extinction and reward-related safety learning. (news.mit.edu)
That is the heart of the study.
The brain was not just recording calm. It was detecting a mismatch: I expected something bad, and it did not happen.
That mismatch is powerful. It is the moment when an old prediction becomes available for revision.
The mice that showed the strongest dopamine signal onto Ppp1r1b neurons were also the mice that extinguished fear most effectively. (news.mit.edu)
At that point, the result could still have been only a correlation. Dopamine might have been present during extinction without actually causing it.
So the researchers tested causality.
Turning the circuit down impaired extinction. Turning it up accelerated it.
The causal experiments are what make the study more than an observation.
Using optogenetic tools, the researchers reduced or increased dopaminergic input from the VTA to the posterior basolateral amygdala. Reducing that input impaired fear extinction. Increasing it accelerated extinction. (news.mit.edu)
The contrast was equally important. When the researchers activated dopaminergic projections into the anterior basolateral amygdala, extinction was impaired, and fear could be reinstated even without new shocks. (news.mit.edu)
This is the part that prevents a cheap interpretation.
The lesson is not: dopamine reduces fear.
The lesson is: dopamine can reduce or reinforce fear depending on the circuit it acts through.
The receptor experiments made the same point at the cellular level. In Ppp1r1b neurons, increasing D1 receptor expression promoted extinction and reduced fear recall, while reducing D1 receptor expression impaired extinction. In Rspo2 neurons, reducing D1 receptors lowered freezing behavior. (news.mit.edu)
As the authors wrote in the study, quoted by MIT: “We showed that fear extinction requires VTA dopaminergic activity in the pBLA Ppp1r1b neurons by using optogenetic inhibition of VTA terminals and cell-type-specific knockdown of D1 receptors in these neurons.” (news.mit.edu)
Put together, the evidence supports a narrow but strong conclusion: dopamine from the VTA promotes fear extinction by acting on Ppp1r1b neurons in the posterior basolateral amygdala, through D1 receptor-dependent mechanisms.
That is why the study matters. It does not merely show that dopamine is active while fear fades. It shows that changing this circuit changes the animal’s behavior.
This is not a trauma delete button
It is tempting to turn a finding like this into a therapeutic promise. That would be premature.
Fear extinction is not erasure. The original fear memory can remain. What changes is the brain’s ability to form and use a competing safety memory.
That is why fear can return after stress, in a new context, or when a reminder appears. The fear memory was not deleted. It was inhibited under certain conditions by newer learning.
This is frustrating, but it is also realistic.
The brain does not work like a hard drive. It does not simply remove a file called “trauma.” It keeps old predictions and builds new ones. Recovery often depends on which prediction controls behavior in the present.
The study helps explain one way the brain may shift that balance.
Why PTSD and anxiety researchers will care
Fear learning is not a malfunction. It is protective. The problem begins when fear persists after the danger has passed, or spreads to situations that are not actually threatening.
That is central to PTSD and many anxiety disorders.
A circuit that helps the brain learn “this is safe now” is therefore highly relevant. The VTA → pBLA pathway, and especially dopamine’s action on Ppp1r1b neurons, may become an important target for future research. The 2020 Picower report had already framed Ppp1r1b neurons as a possible target for fear-related disorders, and the 2025 study adds a dopamine mechanism upstream of that extinction-related population. (Institutul Picower)
But a target is not a treatment.
This distinction matters because dopamine is too broad a system to manipulate casually. It is involved in many aspects of learning, motivation, movement, salience, and psychiatric state. Raising dopamine globally would not reproduce what this study found. It might do nothing useful. It might make things worse.
The study itself shows why. Activating one dopamine pathway helped extinction. Activating another impaired extinction and could bring fear back. (news.mit.edu)
So the real implication is not “boost dopamine.”
It is much more demanding: find a way to support the right circuit, at the right time, while safety learning is actually happening.
That is a harder idea to sell.
It is also the more honest one.
As Pignatelli di Spinazzola noted in MIT’s report: “Fear learning and fear extinction provide a strong framework to study generalized anxiety and PTSD. Our study investigates the underlying mechanisms suggesting multiple targets for a translational approach, such as pBLA and use of dopaminergic modulation.” (news.mit.edu)
What the study does—and does not—show
The finding is strong because the researchers did more than observe a signal. They traced the circuit, measured activity, manipulated the pathway, and changed behavior. (news.mit.edu)
Still, the scope is limited.
This was a mouse study. Optogenetic and genetic tools allow a level of precision that is not directly available in human treatment. Fear extinction in people also involves broader networks: prefrontal cortex, hippocampus, amygdala, stress systems, memory context, attention, sleep, and individual history.
The VTA → pBLA pathway is not the whole story.
It is one important piece.
And that is enough.
The responsible conclusion is not that we now have a treatment for PTSD. We do not. The responsible conclusion is that researchers have identified a specific dopamine circuit that causally promotes fear extinction in mice.
That is already a meaningful advance.
A better way to think about dopamine
The study also gives us a better language for dopamine.
Dopamine is not just pleasure. It is not just reward. It is not just craving, drive, or motivation. In this case, dopamine helps mark the moment when the world turns out to be safer than expected.
The animal returns to the feared place.
The shock does not come.
The absence becomes meaningful.
That is the quiet sophistication of the finding. Safety is not passive. The brain has to learn it. It has to detect that the old threat prediction failed, then build a new interpretation strong enough to compete with fear.
Fear is not erased.
Safety is learned.
And in mice, at least, one dopamine pathway appears to help make that possible.
Study details and sources used
This article was prepared using the following sources:
- “Dopamine signals when a fear can be forgotten” — MIT News / The Picower Institute for Learning and Memory. This source was used for the 2025 study summary, author quotations, experimental design, VTA-to-amygdala circuit findings, optogenetic results, receptor-manipulation findings, funding information, and image credit. (news.mit.edu)
- “Dopamine induces fear extinction by activating the reward-responding amygdala neurons” — Proceedings of the National Academy of Sciences, DOI: 10.1073/pnas.2501331122. This source was used as the primary research citation for the published study, including title, journal, DOI, publication details, and core scientific claim. (pnas.org)
- “With these neurons, extinguishing fear is its own reward” — The Picower Institute. This source was used for background on the 2020 work identifying Ppp1r1b neurons in the posterior basolateral amygdala as involved in fear extinction and reward-related valence, and Rspo2 neurons as part of the competing fear-related population. (Institutul Picower)
- “Susumu Tonegawa” — The Picower Institute profile page. This source was used to identify Tonegawa’s institutional role and research context at MIT, including his work on learning, memory, circuit mapping, neuromodulation, engrams, amygdala, and valence. (Institutul Picower)
