Fear keeps humans alive. It is one of our most basic instincts. A tiger roars, a gunshot rings out, or a truck rattles around a corner too close to where we are standing, and our bodies leap into high alert, jaw clenched, hair standing on end. Often, we remain jangled long after the threat has passed, as though the tiger were still creeping around in the back of our minds.
For decades, scientists have considered the amygdala, an almond-shaped cluster of brain cells near the brain stem, to be the seat of our fear responses. But the amygdala presents a puzzle: The nerve cells in this region of the brain only fire briefly when an animal, human or otherwise, is first startled. Then they quit. How does the brain sustain the fear response—and prolong the fight, flight, or freeze behavior that accompanies it?
Read more: “When We Were Lunch”
Recently, a team of neuroscientists from the United States and China discovered a tiny brain region in male mice that sits right next to the amygdala and seems to hold the answer. They named it the amygdalostriatal transition zone, or ASt. Neurons in this region fire with far more power and duration in response to a fear signal than neurons in the amygdala do, and the ASt communicates directly with the striatum, a brain region that plays a central role in planning, initiating and controlling voluntary movement. They published their results in Neuron.
Mice and humans have so far been found to have similar brain structures, processing, and behavioral responses when it comes to fear. So the findings may have implications for the treatment of fear and panic disorders, the researchers say. I spoke with study coauthor Kay Tye, a professor in the Systems Neurobiology Laboratory at the Salk Institute for Biological Studies, about why scientists missed this region until now, what it tells us about the fight or flight response, and how it could shake up the study of fear.
You found this region of the brain that keeps fear alive after the initial high-alert signal has come and gone. Why do you think neuroscientists had missed this until now?
Honestly, I feel embarrassed that I missed it because I’ve been studying the two brain regions that lie on either side of it, the amygdala and the striatum. We’ve always ignored this little region in between because it doesn’t have a clear boundary when you look at it under a light microscope. It doesn’t obviously look like a separate section, whereas the amygdala has dense cell clusters and fiber bundles that make it look like a different section. But when you look at the molecular identity of the cells in the ASt, you can tell that it is its own unique brain region.
This little ASt could help to explain why there was always a kind of mismatch in terms of the importance of the amygdala to fear. It’s right next to the amygdala, so if you did a lesion of the amygdala, or if you infused drugs into the amygdala, if you did any manipulation of the amygdala, it’s very likely that you also hit this region. If you chop the amygdala out, all these functions are severely impaired. But if you record signaling there, it’s only giving you this transient little blip of onset responding. It doesn’t have the continued sustained response.
So the amygdala is responsible for the onset of fear, but the ASt keeps it going?
The analogy I like to use is that it’s like if there were a fire alarm. You’d get this initial startle response. That’s the amygdala going off. Then you evacuate and stand outside waiting until the alarm stops. You’re still in defense mode, because the threat is still present, but you’re not startled anymore. That’s the ASt. Until now, we never actually knew what was holding us there in fear response. Being in a fear state or a defensive state is very energetically demanding. You’re very alert, aware of all your sensory stimuli. It’s not efficient to default to a hypervigilant state. It makes more sense that you would have something that has to actively be engaged to keep you there.
We found these neurons in the ASt zone that have screaming responses to fear cues. By comparison, the amygdala is relatively subdued. Thousands and thousands of papers have been written about the amygdala in fear, and you get one standard deviation above the mean in terms of activity. If you look in the ASt, you get closer to 20 standard deviations. That’s something you don’t see in most parts of the brain.
Do you have hypotheses about why this region is separate from the amygdala?
You need something that’s flexible for encoding, which is the amygdala, and you need something that’s in rapid, default mode, which is the ASt. The amygdala is more at the processing level. Sometimes you need to think about whether you really should be scared of this thing or if you share a different association. The ASt is not super flexible in terms of connectivity. It’s closer to our striatal circuits, which is the final stop for integrating information prior to motor output. It’s very evolutionarily adaptive if something’s threatening you to prioritize that thing. If I’m about to get hit by a truck, I can worry about talking to my friend later. I can worry about checking this message or eating this snack or drinking this drink later. Literally everything else can be done later. We prioritize bad things. And so it should make sense that there’s some system that’s just like, “Wait, okay, it’s fine. It’s fine.” You know, you need some bottom-up signal to make you alert anytime there’s a risk. And it’s maybe better to err on the side of vigilance than on the side of dismissal.
You found that this region primarily responds to cues, not environments. What does that tell you about how fear works and is learned?
Cues are a part of your environment, but they’re acute and temporally specific—a twig snap, a weird shadow. Some cues will come with contextual associations. If you’ve been shocked in this environment before, then you might not like this environment. But fear of environments and context can also be considered anxiety—if you have a heightened state of apprehension in the absence of an immediate threat.
It’s interesting, because I tend to think of lasting fear like anxiety, maybe even PTSD, as a slightly irrational response to an environment, due to fearful cues that you come to associate with that environment.
That’s right. But that’s not what’s happening in this region. It’s two different things. If I were a veteran of the Vietnam War, everything associated with Vietnam might make me feel anxious. But you might also have contextual triggers: gunshots, loud noises that sound like banging. A different system in your brain is important if you are associating contexts versus cues with negative states. It’s a different level of urgency. This is called threat imminence.
Justified versus unjustified fear is one way to think about it. The amygdala governs the justified fear, the ASt the lingering anxiety from justified fear. But fear disorders related to context and environment are processed elsewhere.
Does a specific region in the brain respond specifically to frightening environments?
The hippocampus is important for contextual fear conditioning, while cued fear conditioning tends to be linked to the amygdala. The ASt is very aptly named because it does seem to act as a shortcut between the amygdala and motor output regions. It’s not a shortcut for the hippocampus.
Does your finding change what we know about how the flight or fight response works?
The fight or flight response is multiple things. It triggers the HPA axis—the hypothalamic-pituitary-adrenal axis—which is the body’s main stress response system. It has outputs from the amygdala and the hypothalamus. This new ASt region may also contribute to that. We didn’t study that specifically. But my guess is the amygdala and the ASt zone are both part of the fight or flight response. The amygdala striatal transition zone is particularly important for maintaining the fight or the flight response. It tells us how long we need to stay in those states. It signals that we need to keep fighting or keep flying.
You found this region in the brains of male mice. How confident are you that such a region exists and behaves in a similar way in female mice?
The amygdala is very well conserved in female mice, and I am pretty confident this region exists in female mice, as well. It’s likely that certain factors will be processed differently in female mice, however. We’ve seen a lot of evidence that males and females have different responses in the amygdala to, for example, social isolation or drinking. If you isolate males, they will drink more. If you isolate females, they drink less. The function may be different in females. But the brain region exists.
What about in human brains?
The amygdalostriadal transition zone should exist in humans. I don’t have any evidence that the function is the same or the connectivity is the same, but I would be very surprised if this did not exist in humans and if it did not have a similar function.
How does your finding change what we know about disorders like PTSD and anxiety?
Anxiety and fear disorders are hugely prevalent. Some 18 percent of the population is affected by some sort of anxiety disorder. But no treatment exists without side effects. That’s because we don’t understand the circuitry that underlies anxiety and fear. We have found some of it, but being afraid of things is the number one survival skill that most species have, so it’s very robustly and redundantly expressed in the brain. Many parallel circuits are useful for detecting threats. If you want to treat anxiety, you have to know all the circuits. You can’t just know some of them because the other ones will compensate.
The drugs we currently use have lots of side effects that are sedating and impair cognition. That’s because if you don’t know all of the circuits, you’re just shooting blind. Anxiety is really probably a couple dozen different diseases, and we’re treating them all the same, trying to use a one-size-fits-all medicine for what is likely to be many different subclasses of disorders that target different circuits.
We diagnose anxiety disorders with a questionnaire. Some of the questions will be like, “Oh, do you have intrusive thoughts? Do you have panic attacks?” And some people might have panic attacks, but not intrusive thoughts. Some people might have strong reactions to certain things, but not panic attacks. These different symptoms are indicative of different pathology, and yet if you check three, then you have anxiety, and you go into this box, and you get the same treatment. Imagine a future where we tailor the treatment to the actual disease and the symptoms.
Could the finding change how scientists study fear?
It’s a wake-up call for the field. Someone said in a paper 100 years ago, “The amygdala is important for fear,” and then thousands of people studied the amygdala for fear and didn’t look elsewhere. I would’ve said this was basically a solved field. We know so much about fear conditioning.
If there’s a whole unexplored brain region we didn’t even know about, we are so far away from knowing how everything works in the brain. This is such a fundamental finding. Something so basic as fear is probably the most studied internal state of any. How fear conditioning works is one of the most studied questions in neuroscience. There are thousands of papers on the subject, and we missed a very obvious critical region. It was just tiny and no one looked at it before. That shows us how much more there is to discover, if we are still discovering in 2026 basic brain regions that do super basic functions that are essential for survival. We’re at the very tip of the iceberg.
Enjoying Nautilus? Subscribe to our free newsletter.
Lead Image: RoseRodionova / Shutterstock
