Autism Spectrum Disorder: The Engine Room

In this episode, we go deeper than behavior and deeper than brain structure. We're heading into the engine room, the cellular and chemical systems that researchers believe are central to understanding why the brain in autism develops and functions the way it does.

We look at three systems. The brain's primary chemical messenger and what happens when its calibration is off. The cellular structures responsible for energy production and why the brain feels their dysfunction most acutely. And the signaling pathway that controls how the brain builds and maintains its own connections, and what goes wrong when it runs in the wrong direction.

We also connect all three back to what we covered in Episodes 1 and 2, because none of these systems operate in isolation. They interact. They feed into each other. And together they start to explain something the Lord et al. Lancet review pointed at but couldn't yet answer, why we can see patterns in autism across large groups, but still can't make reliable predictions for any individual person.

Based on Shuid et al., Update on Atypicalities of Central Nervous System in Autism Spectrum Disorder. Brain Sciences, 2020. Cross-referenced with Lord et al., Autism Spectrum Disorder. The Lancet, 2018.

Transcript

SPEAKER_00 0:00

Think about the last time you were in an airport late at night. Not during the rush hour, not when every gate is full, and you can barely hear yourself think. I'm talking late at night, when the terminal has emptied out and the lights have dimmed and the shops have their shutters down. It's quiet in a way airports almost never are. And in the quiet, you start noticing things you couldn't hear before. The low hum of the ventilation system running somewhere above the ceiling, the crackle of the PA before it speaks, like it's clearing its throat, the faint vibration of something mechanical beneath the floor. The airport is still working. All of its systems are still running. You just couldn't hear them before because everything else was so loud. That's where we're going today. Not to the check-in desk, not through security, not to the departure board. We're going through a door that says staff only, down a corridor that passengers never see, and into the part of the airport that makes everything else possible. We're headed to the engine room. Over the past two episodes, we've been building a picture of autism from the outside. We started what autism looks like, the behavior, the diagnosis, the enormous range of how it shows up in different people. Then we went behind the walls and looked at the brain structures themselves, the regions that are shaped differently, the connectivity patterns that run differently, the inflammation that sometimes activates and doesn't stop. Today, we go one level deeper into the chemistry, into the cellular machinery, into three systems that researchers believe are central to understanding why the brain in autism develops and functions the way it does. If this is your first episode, welcome. You don't need to have listened to the others. Everything you need, I'll give you as we go. But if you've been here since the beginning, I think today is the episode where a lot of things are going to start connecting. So let's get started. I'm your host, Nafisa Golwala. I'm not a clinician, I'm someone who believes that good research deserves a wider audience, and that most people, given clear and honest information, are capable of doing something meaningful with it. Today's primary paper is Shoot It All, update on atypicalities of central nervous system in autism spectrum disorder, published in Brain Sciences in 2020. We'll also be connecting it back to the Lancet paper from episode one. Here's what you need to know before we go in, because what we're looking at today only makes sense if you understand what we already know about autism. Autism spectrum disorder is a neurodevelopmental condition defined by differences in social communication and in patterns of behavior and sensory experiences. It is biological in origin. It begins early in brain development and it looks completely different from one person to the next. Some people are diagnosed at 2, others at 45, and many are never diagnosed at all. That enormous range isn't just about how autism shows up on the outside. It also reflects what's happening at the biological level. And that is exactly what today's episode is about. In episode one, we looked at the Lancet review by Lorde and colleagues. One of its key findings was this: genetics and neuroscience have identified intriguing patterns of risk in autism, but without much practical benefit yet. We know a lot is going on inside the brain. We can see patterns, but we cannot yet translate those patterns into reliable predictions for individual people. In episode two, we looked at the brain structure and its immune response. We looked at regions like the amygdala and the hippocampus and what happens when the brain's own immune cells, the microglia, activate in ways that don't resolve. We looked at neuroinflammation, what it is, what the research shows, and why it matters for how the brain develops. Today's paper, Shoot It All, starts to explain why. Because what's happening inside the brain with autism isn't one thing. It's multiple systems all running differently, all interacting with each other, and we're going to look at three of them. The PA system, the power generators, and the maintenance protocol. Imagine you're standing in the middle of the terminal and the PA system turns on. Most of the time you barely register it. A gate change, a boarding announcement, your brain filters it into the background alongside the noise of the crowd and the hum of the building. It's there, it's doing its job, and you move on. But now, imagine the volume starts climbing. The announcements keep coming faster than you can process them, or in some parts of the terminal, the system is so quiet nobody can hear it. While in others it's so loud people are covering their ears. The system isn't doing what it's supposed to do. Not because it's catastrophically broken, but because the calibration is off. That's broadly what a lot of studies are pointing towards when it comes to glutamate in autism. So, what is glutamate? Glutamate is the brain's primary excitatory neurotransmitter. Think of a neurotransmitter as a chemical signal that carries messages between brain cells. And glutamate is the one that says activate, fire, pay attention. It's essential for brain development, for learning, for memory. Without enough of it, signals don't get through. With too much, the system gets overwhelmed. And across studies in autism, something about glutamate regulation seems off. Although not always in the same way. Researchers can usually measure glutamate levels in a living brain without surgery using a neuroimaging technique called proton magnetic resonance spectroscopy. It works by detecting the chemical signatures of different compounds in specific brain regions. What they measure is usually referred to as GLX, a combined measure of glutamate and a related compound called glutamine, which together gives us a picture of how the glutamate system is functioning. This is where things start to get a bit strange. Some studies find elevated glutamate levels in people with autism, particularly in regions of the brain involved in sensory processing and social cognition. The auditory cortex, for example, the part of the brain that processes sound, showed higher glutamate levels in adults with autism compared to controls. When you think about it, how many people with autism describe the experience of sound as overwhelming, as physically painful, as something that can't be filtered or turned down? Elevated glutamate in the auditory cortex may offer one possible biological clue for why sensory processing can differ in autism, though the paper does not establish a direct explanation for individual sensory experiences. But other studies find the opposite, reduced glutamate in different brain regions. One study found significantly lower levels in the basal ganglia, which is a region involved in movement, reward, and social behavior. And that reduction correlated specifically with impairment in social communication. So it's not just too loud or too quiet, it's both, depending on where you're standing. In different parts of the terminal, it's doing both at the same time, which is exactly why it's so hard to map out, and part of why we can't yet look at a scan and make a reliable prediction about any individual person. There's also evidence that glutamate levels are elevated in the blood of people with autism, not just in the brain. And what's particularly striking is that the same elevation shows up in their parents and siblings, which suggests this isn't purely something that happens in autism. It may reflect an underlying genetic tendency that runs in families, even in people who don't have a diagnosis themselves. This actually ties back to something from episode one. Lorde and colleagues noted that genetics in autism involves hundreds of overlapping variants, not one single gene. What the glutamate research adds to that picture is the possibility that some genetic risk may intersect with how glutamate signaling is regulated. But this review does not identify a single direct pathway that explains that line. The PA system's calibration problem may be partly written into the wiring from the beginning. One thing that matters here, and it's easy to miss, when glutamate levels are consistently too high for too long, it stops being a signaling problem and becomes a damage problem. Excess glutamate can be neurotoxic, meaning it contributes to the death of nerve cells. The PA system, if it runs at a maximum volume indefinitely, starts damaging the speakers, which is one reason why glutamate dysfunction remains an important area for future research. Now, one important caveat that the paper makes, and this is a theme throughout the research on autism neurobiology. The findings vary a lot, depending on the age of the participants, their IQ level, their gender, whether they're taking psychoactive medication, and which brain region is being measured. The same study design in two different labs can produce opposite results. That's not a reason to dismiss the research. It's a reason to hold it carefully. The PA system is dysregulated. We just haven't fully mapped out the wiring yet. Okay, so if you think about it, every piece of equipment in an airport runs on power. The check-in screens, the security scanners, the PA systems we just talked about, the gate displays that tell you where you're going, all of it draws from the same electrical grid. Now, imagine those generators are running at 70% capacity. Not broken, not off, just not producing what the system actually needs. Most of the airport keeps functioning, the lights are on, the screens are working, but the areas that demand the most power start to feel it first. Things start to run a little slower, systems that need high energy to perform complex tasks start operating below their best. That's one way researchers think mitochondrial dysfunction might show up in the brain in autism, and evidence for it is substantial. So, we all know that the mitochondria is the powerhouse of the cell. But what are they actually doing? Inside almost every cell, you've got these structures called mitochondria. Their job is to produce energy, specifically a molecule called ATP, which is essentially the fuel that powers everything a cell does. They produce it through a process that requires oxygen and a series of specialized proteins working in sequence. It's efficient, it's elegant, and it's absolutely essential for the brain to function. And the brain is especially vulnerable here for a pretty simple reason. The brain accounts for roughly 20% of the body's total energy demand, even though it's only about 2% of its mass. Neurons, aka brain cells, are extraordinarily hungry. And the parts of neurons that communicate with each other, the terminals where signals pass from one cell to the next, are especially dependent on a constant, reliable energy supply. When that supply is compromised, communication suffers. In autism, multiple studies have found signs that mitochondrial energy production is impaired. One of the more commonly reported signs is elevated lactate, along with pyruvate and alanine abnormalities in the blood, cerebrospinal fluid, or brain samples for some groups with autism. Lactate is a byproduct that accumulates when cells can't process energy the way they're supposed to, like exhaust from an engine that isn't burning fuel cleanly. Finding high lactate is a well-established indicator that something in the energy production system isn't working properly. Researchers have also looked directly at brain tissue in postmortem studies and found reduced activity of the protein complexes that drive energy production, specifically in the frontal and temporal lobes and the cerebellum. These are regions involved in decision making, language, sensory processing, and coordination, the same regions that show up repeatedly across autism research as functioning differently. Now, this is where it gets more complicated and honestly more interesting. Mitochondria are unusual among the structures inside cells because they carry their own DNA, separate from the DNA in the cell's nucleus. This is a remnant of their evolutionary origin as independent organisms that were absorbed into cells billions of years ago. And in autism, studies have found abnormalities in the mitochondrial DNA itself, inside copy numbers, deletions, and damage. One study found these empty DNA deletions co-occurring with other genetic alterations already associated with autism, which supports an association between mitochondrial alteration and ASD, though it does not prove causation. This is the part I want you to keep in mind, because it links directly back to the last section. The paper proposes that excess glutamate, what we described as the PA system running too loud, may itself be contributing to mitochondrial stress. Glutamate is a known mitochondrial disruptor. So the signaling problem and the energy problem aren't separate stories. One may be making the other worse. The generators are running below capacity. The highest demand areas of the building are feeling it most. And the source of that reduced capacity may partly be the other systems we're talking about today, feeding back into each other in a loop that is so far very difficult to untangle. And that's where the third system comes in. Every airport has a facilities team, a system that decides what gets built, what gets maintained, and what gets cleared away when it's no longer needed. When a new section of the terminal is added, this system coordinates structure. When old infrastructure becomes redundant, it schedules removal. And critically, it decides what gets pruned. Because airports like brains regularly need to remove things that were useful temporarily, but would then cause congestion if they stayed permanently. Now imagine that system malfunctions. In some parts of the airport, it goes into overdrive, building constantly, never cleaning. Temporary structures become permanent. Old scaffolding stays up indefinitely. The building gets denser and more complex in ways that weren't ever planned. And in other parts of the airport, the opposite happens. Maintenance stops, structures that should be repaired are left to degrade, and gaps appear where there should be connections. This is the MTOR signaling pathway, and it is one of the pathways repeatedly implicated in autism research, though the direction of dysregulation is not uniform across studies. Before we go further, we need to unpack what MTOR actually is, because it's one of those terms that sounds more complicated than the concept it describes. MTOR stands for mechanistic target of rapamycin, which is essentially a molecular switch inside cells. It sits at the center of a signaling network that controls how cells grow, how they survive, how they use energy, and how they organize their internal structures. In the brain specifically, MTOR regulates how neurons shape themselves, how connections between neurons form and strengthen, and a process called autophagy, which is the cellular equivalent of taking out the trash, clearing away old, damaged, or unnecessary structures so the cell can function cleanly. And this is the part that really matters. MTOR dysregulation in autism doesn't go in one direction, it goes in both. And both directions seem to cause problems, just in very different ways. One study looked at brain tissue from temporal lobe of children and adolescents with autism, which is essentially the region involved in language, social cognition, and how we process other people. What they found was a higher density of dendric spines than expected. Dendric spines are the small protrusions on neurons where connections form. Normally, the brain prunes them during development, removes the ones that aren't being used, so the important ones can strengthen. This pruning process depends partly on MTOR, and in this tissue, MTOR was overactive. Pruning wasn't happening properly, and the result was a denser, more tangled network than a typically developing brain would have at the same age. The maintenance team is running at full speed, but only doing half its job building, building, building, but never cleaning. Another study found the opposite in a different brain region. One involved in face recognition and visual social processing. There, MTOR signaling was reduced. The proteins responsible for organizing synaptic connections were lower than normal. The maintenance team had effectively stopped showing up, and the result was underdeveloped connections in exactly the area that processes other people's faces. So, you've got the same pathway, but behaving completely differently depending on where you look, in different parts of the same brain, overbuilt in some areas, undermaintained in others, and both of those states are thought to affect neuronal networks in ways that may contribute to behavioral and cognitive differences seen in autism. This is where it connects back from episode one. Lorde and colleagues noted that we can see patterns in autism across large groups, but we cannot look at an individual person's brain and predict what their life will look like. The MTOR research helps explain why, because the dysregulation isn't uniform. It runs in different directions in different people, in different brain regions, at different developmental stages. There is no single signature to look for. Why this matters beyond just theory is that MTOR is actually something we can target. Rapamycin, the compound MTOR, is named after an existing medication that modulates this pathway. Studies are currently underway in specific subtypes of autism where MTOR dysregulation is part of a known genetic syndrome. The possibility of targeted interventions, not cures, not one size fits all treatment, but specific modulations of specific pathways for specific people, is one of the most promising directions the field is moving in, which is exactly what good science should be building towards. Okay, so let's tie all the information together. We've been in the engine room, let's come back out into the terminal. Three systems, each showing a different way the brain might be functioning differently in autism at a cellular and chemical level. The PA system, glutamate, running at the wrong volume in different parts of the building simultaneously, too loud in some regions, contributing to sensory overwhelm, too quiet in others, contributing to difficulties with social communication and reward, and the signal dysregulation may itself be causing damage over time. The power generators, mitochondria, running below capacity, the brain which has the highest energy demand of any organ in the body, feeling the shortfall most acutely, with the genetic material inside the generators themselves showing signs of damage and deletion. Finally, the maintenance protocol, MTOR, running out of calibration in both directions, overbuilding in some regions, under maintaining in others, creating a brain that is simultaneously denser than it should be in some places and less connected than it should be in others. And these systems may not be independent. They may be feeding into each other. The signaling issues possibly affecting energy, and maybe the other way around too. The generator problem affecting every other system that depends on energy. The maintenance problem shaping the very architecture that everything else runs on. And all three connecting back to the neuroinflammation we covered in episode two. The immune response that activates in the brain and doesn't fully resolve. This is what the Lorden colleagues paper from episode one was pointing towards when it said, genetics and neuroscience have identified intriguing patterns of risk, but without much practical benefit yet. The patterns are real, the systems are identifiable, but because they interact in so many different ways in different people at different developmental stages, translating that knowledge into something that reliably helps individuals is still the hard part. The paper we look at today ends with a call for individualized medicine, not a one size fits all approach to autism, but targeted interventions for specific biological mechanisms in specific people. That's a long way from where we are, but it's a direction. And it's one that the research is increasingly. Increasingly building toward. The airport wasn't built with one blueprint, and understanding how it actually works, all of its systems, all of their interactions, requires that we keep going deeper, which is exactly what we're gonna keep doing. If this episode helps something click, share it with one person who would find it useful. That is how the show reaches the people who need it the most. Next episode, we're stepping back from the molecular level and looking at the lived experiences, sensory processing, masking, and what the research tells us about what it actually costs to navigate a world that wasn't designed with you in mind. I've been looking forward to this one. So I guess I'll see you then. And remember, things can be redesigned. They just have to be built brick by brick.