What happens next.
Once that feedback control fails, the system does not simply stay activated. It starts a pathological chain that can end in neurodegeneration decades before symptoms appear.
First in class · Patent pending ALPHA 2A PAM The brain has a built-in brake for stress. No approved drug restores it.
The brain's stress system has a built-in brake.
There is a part of your brain that controls how alert you are, how strongly you react to stress, and whether you can calm down after. It is called the locus coeruleus, a small cluster of neurons deep in the brainstem. The chemical it releases is called norepinephrine, and it touches virtually every region of the brain.
On the surface of each of these neurons sits a receptor whose only job is to sense how much norepinephrine the neuron just released and tell it to slow down. That receptor is called the α2A autoreceptor. When the neuron fires, some of that norepinephrine loops back to the receptor and pulls the cell toward quiet. More firing means a stronger brake signal.
That loop is what keeps stress proportional. It lets the system react, then return toward baseline.
Chronic stress breaks that brake.
Under sustained stress, the autoreceptors are pulled inside the cell and the signalling partners the receptor needs to send its brake signal (Gi/o proteins and GIRK channels, for those following the pharmacology) deplete alongside them.
Without that brake, the norepinephrine release is no longer spiking in response to threats and then calming down. It is constant. Always on. What researchers call "tonic." The system that was supposed to react and recover is now just running, all the time, at a level it was never designed to sustain.
The person living inside this system can't wind down. Sleep doesn't come easily. Focus fragments. The heart rate sits higher than it should. Doctors call it anxiety, or hypertension, or ADHD. It's none of those. It's all of those, because they're all downstream of the same failure. That pattern, a set of symptoms driven by a single mechanical cause, is what researchers call a biotype.
And three self-reinforcing feedback loops keep feeding back into the locus coeruleus, locking the system in overdrive.
This is not ordinary anxiety. It is a failure of the brain's own regulatory architecture.
Autoregulation fails
Chronic stress pulls the α2A autoreceptors inside the cell and depletes the signalling partners they need to work. In chronic stress models, 88% of locus coeruleus neurons begin firing in uncontrolled bursts, a pattern virtually absent in healthy controls.
Tonic NE escalates
Without the brake, norepinephrine release becomes constant, no longer spiking and calming but running at full volume all the time. The system designed to respond and recover is now permanently on.
Stress hormone loops amplify
The body's stress hormone systems, CRF signalling and the HPA axis, stop resolving the response and start feeding back into the locus coeruleus. Instead of calming the system down, they lock it in overdrive.
MAO-A expands in LC neurons
Stress hormones switch on a protein called KLF11 that increases production of an enzyme called MAO-A, while the system that normally clears excess MAO-A is suppressed. The enzyme that creates the toxic byproduct becomes physically more abundant inside the very neurons being pushed hardest.
Excess NE becomes DOPEGAL
Excess norepinephrine is broken down into a toxic byproduct called DOPEGAL, produced inside the very neurons that are being pushed beyond their design limits.
DOPEGAL modifies tau
DOPEGAL covalently bonds to tau at K353, triggers aggregation, and activates an enzyme that cleaves tau into toxic fragments. Blocking this single bond prevented locus coeruleus neuron death in rescue experiments.
LC neurons begin to die
What started as a broken feedback loop, something that could have been fixed, becomes permanent physical damage. The locus coeruleus begins losing neurons, and they don't come back.
Early Alzheimer's pathology appears
Tau pretangles appear in locus coeruleus neurons in the second and third decades of life, 30 to 50 years before clinical Alzheimer's onset. By the time symptoms appear, roughly 30% of these neurons are already gone.
Why hasn't this been fixed?
Direct agonists
Clonidine, guanfacine, and dexmedetomidine activate α2 receptors directly. They suppress locus coeruleus firing whether or not the system actually needs to quiet down. The consequence is predictable: receptors desensitise, tolerance develops, and withdrawal unmasks the same dysregulated circuit because the underlying failure was never repaired.
Downstream drugs
Some drugs called SNRIs actually raise norepinephrine levels by preventing the neuron from recycling it back. They increase the very chemical that is already in excess. Beta-blockers damp peripheral output. Benzodiazepines suppress the system more broadly. None of them repair the autoinhibitory failure at the locus coeruleus.
If you've taken several of these and none of them fully resolved what you're experiencing, this may be why. They were each managing a different symptom of the same unaddressed cause.
No approved compound restores the brain's own α2A autoregulation.
What would actually fix it would need to work differently. Not force the receptor from outside, but amplify the brain's own signal so the brake works again. It would need to engage only when the system is actually overdriving and go quiet when it is not. And it would need to fade as the system normalises, so the dose is effectively set by the disease state itself.
AVX-1 works with the existing feedback system, not in place of it.
Mechanism
That is what AVX-1 is designed to do. It is what pharmacologists call a positive allosteric modulator of the α2A autoreceptor. It doesn't activate the receptor. It amplifies the brain's own norepinephrine signal at the feedback site.
Signal dependence
When the neuron fires and norepinephrine hits the autoreceptor, AVX-1 makes that signal stronger. A larger inhibitory response from the same natural signal. In the absence of norepinephrine, the drug is pharmacologically silent. It literally cannot work unless the brain is providing the input.
Phasic preservation
It preserves phasic signalling. The brake takes about 150 milliseconds to kick in. The fast bursts your brain uses for attention and reacting to real threats finish before the brake can catch them. That means AVX-1 only slows down the constant background noise, not the signals you actually need.
Self-limiting
As tonic firing comes down, there's less norepinephrine to potentiate. The drug's own effect fades as the system normalises. The dose is effectively titrated by the disease state itself.
Upstream effect
It works upstream of everything. Reducing tonic norepinephrine at the source is predicted to attenuate the stress-hormone amplification, the MAO-A expansion, the DOPEGAL production, and the tau modification that initiates neurodegeneration.
The same drug that treats the functional symptoms, the anxiety, hyperarousal, and cognitive dysfunction, is predicted to simultaneously reduce the neurotoxic load that drives Alzheimer's pathology decades later. Treatment and prevention become the same intervention in the same patient.
No α2A PAM exists anywhere in development.
AVX‑1 would be the first.
Who this reaches.
The conditions below look different in a clinic. They share the same upstream failure.
The primary biotype
The person whose resting heart rate has always been too high. Who has been told they have anxiety, and hypertension, and maybe ADHD. Who takes multiple medications and still doesn't feel right. Who has never gotten a satisfying answer for why their body runs like this. That person may be living with a single upstream failure that no current treatment addresses.
PTSD
Someone whose stress system locked into overdrive after trauma and never came back down. Who still startles too easily, sleeps too lightly, and can't stop scanning for threats years after the danger passed. Current treatments manage the symptoms. None of them restore the circuit that trauma broke.
Resistant hypertension
Someone whose blood pressure stays elevated no matter what their doctor prescribes. Who has tried multiple medications and the numbers still won't come down, because the signal driving it isn't coming from the heart or the kidneys. It's coming from the brain.
Already on medication
Someone already on an SNRI or guanfacine or another norepinephrine-active drug that helps partially but never fully resolves the problem. Their current medication may be raising the very chemical that is already in excess. Restoring the brake could protect the circuit that current therapy is stressing.
Anxiety
Someone whose anxiety has never fully responded to SSRIs, benzodiazepines, or therapy alone. Whose baseline is tension, vigilance, and a body that won't quiet down. If the driver is the locus coeruleus rather than serotonin, restoring the brake at the source could reach what downstream drugs cannot.
AVX-1
AVX-1 is a mechanism-first program designed to restore α2A autoregulation without flattening the phasic norepinephrine signalling required for attention, salience, and acute response.
The work is staged from receptor-state modelling to translational validation. Each stage has explicit criteria for what advances and what stops. The program is built to produce a compound whose selectivity, delivery, and circuit logic survive direct challenge, or to terminate early enough that nothing is wasted on a false lead.
- Target
- α2A adrenergic autoreceptor
- Modality
- Gi/o-GIRK-biased, β-arrestin-sparing positive allosteric modulator
- Delivery
- Intranasal with CNS-biased exposure
- IP
- Patent pending. USPTO provisional filed March 2026.
The search begins with receptor conformations that distinguish positive allosteric modulation from direct agonism. Ensemble docking, molecular dynamics, and state mapping are used to find chemotypes that strengthen endogenous α2A signalling only in the receptor states that matter.
A series does not advance because it binds. It advances only if the modelling supports a plausible allosteric pose, tractable chemistry, and a bias profile worth testing at the bench. If the state dependence collapses, the series ends there.
Hits are screened for endogenous selectivity, Gi/o-GIRK amplification, and β-arrestin sparing under conditions that mimic the circuit Aeviant is trying to restore. The requirement is stronger autoreceptor feedback when norepinephrine is present, with silence when it is not.
This stage is built to eliminate easy but unacceptable wins: direct agonism, tonic suppression, or apparent activity that disappears once endogenous ligand dependence is enforced. If the molecule cannot stay conditional, it does not progress.
The translational package has to show that the mechanism survives outside the assay plate: CNS-relevant exposure, locus coeruleus-linked readouts, and a real separation between tonic normalisation and phasic flattening.
That matters because the same upstream mechanism should support both symptomatic relief and neuroprotective logic. If exposure is peripheral, the readouts are ambiguous, or the tonic-phasic distinction cannot be demonstrated, the program does not earn the next study.
Each stage has prewritten stop conditions tied to selectivity, delivery, and circuit-level interpretation. If phasic signalling is materially flattened, CNS delivery fails, or the mechanism cannot be defended cleanly, the program stops.
The point is not to preserve momentum. If the evidence says stop, the program stops.
Design constraints.
AVX-1 is designed to amplify autoreceptor signalling only when endogenous norepinephrine is already engaging the receptor. That keeps the mechanism inside the circuit's native logic rather than replacing it with a permanent pharmacological brake.
The cost of that constraint is real. Patients whose locus coeruleus tonic firing is already low should be pharmacologically invisible to the compound. In that population it would be inert by design, not by failure.
Mechanism
No norepinephrine, no effect.
The intervention point is the locus coeruleus, before the system cascades into CRF/HPA amplification, MAO-A expansion, DOPEGAL formation, and tau modification.
If the mechanism is real, functional treatment and neuroprotection are not separate programs. Reduction in hyperarousal now and interruption of the downstream toxic cascade that follows are different endpoints of the same intervention.
Program consequence
Functional treatment and neuroprotection share the same mechanism.
Every claim in the program is paired with an experiment and an explicit failure mode. If AVX-1 erases phasic signalling, misses CNS-relevant exposure, or loses its mechanistic selectivity, it should not advance.
Kill criteria are written before the experiment runs and enforced regardless of what the program needs to be true.
Program rule
If the data says stop, the program stops.
ÆVIANT BIOSCIENCES
Founded to test whether restoring α2A autoinhibition can break a pathological cascade that current pharmacotherapy does not address.
Founded by Immanuel Martins. Based in British Columbia, Canada.
Scientific and partnering inquiries welcome.
Preclinical stage. No products are available for sale or clinical use.