ISRIB and Traumatic Brain Injury: What the Preclinical Research Shows
If you've had a concussion — or several — you may know a symptom that never shows up on a scan: the thinking doesn't quite come back. The headache resolves, the imaging is clean, you get "cleared," and yet weeks or months later the fog, the word-finding pauses, the sense that recall runs a step slower are all still there. Standard care mostly offers rest and time. For many people that is enough. For a subset, it isn't — and the mechanistic reason why is one of the more interesting threads in recent neuroscience.
I want to be clear about what this article is and isn't. It's a review of what the preclinical research shows about the integrated stress response in brain injury. It is not medical advice, and nothing here describes a treatment you can buy. The studies below were done in mice and fish. There are no human trials of ISRIB — or of A15 — for traumatic brain injury. Hold that gap in mind the whole way through.
The part of the deficit that isn't lost tissue
The default model of TBI is mechanical: tissue was damaged, and what's lost is lost. That's partly true. But it doesn't explain a specific, reproducible finding — that in animal models, cognitive deficits from brain injury can be reversed by a small molecule given weeks after the injury, long after the mechanical event is over. If the deficit were purely structural loss, that shouldn't be possible.
The fact that it is possible points to something else. A portion of the chronic deficit after brain injury may not be lost tissue at all. It may be a signaling state the brain got stuck in — and, at least in animals, a state that can be switched back off.
How brain injury gets the ISR stuck on
The integrated stress response (ISR) is a cellular pathway that pauses most protein synthesis when a cell is under stress. Four sensor kinases converge on one target — the translation factor eIF2α. When eIF2α is phosphorylated, it inhibits eIF2B, and new protein production largely stops. That's protective in the short term. The problem is when the signal doesn't switch off. (For the full mechanism, see the pillar piece on what the integrated stress response is.)
Traumatic brain injury turns out to be one of the states that keeps the ISR switched on. And neurons are exactly the wrong cells to leave with suppressed protein synthesis: forming and maintaining memories is a protein-synthesis-dependent process. Keep the translational brake engaged and you get precisely the profile people describe after a head injury — encoding new information is harder, recall is slower, the machinery is running under capacity.
ISRIB (the "integrated stress response inhibitor") acts at eIF2B, stabilizing its active shape so it keeps working even when phospho-eIF2α is trying to jam it. In injury models, that releases the brake.
What the studies actually found
The landmark study, from the Rosi and Walter labs at UCSF. TBI persistently activated the ISR in mice. Treating with ISRIB reversed hippocampal-dependent memory deficits in two different injury models — focal contusion and diffuse concussive injury. Two findings stood out: ISRIB corrected the deficits even when given weeks after the injury, and the improvement was maintained after treatment stopped. Suppressed long-term potentiation in the hippocampus was fully restored.
The "weeks later, and it sticks" result is the one that reframes the problem. It's the strongest single piece of evidence that part of the chronic deficit is a maintained signaling state rather than permanent loss.
Repetitive mild head trauma — the model closest to real-world concussions — produced chronic ISR activation and cell-specific synaptic changes in the prefrontal cortex, along with altered risk-taking behavior (and the effect was sex-dependent, present in male but not female mice). A brief course of ISRIB weeks after injury relieved the ISR activation, reversed the behavioral change, and restored synaptic function — and the reversal held.
Using two-photon imaging to watch individual synapses in living mice, this study showed concussive injury scrambled dendritic-spine dynamics in the parietal cortex for up to a month. A brief course of ISRIB entirely reversed the structural changes and the associated working-memory deficits — direct structural evidence, at the level of single synapses, for what the behavioral studies had inferred.
A more recent replication in a completely different species. In a zebrafish brain-injury model, injury diminished memory and social behavior; ISRIB given after the injury markedly reduced those deficits and shifted brain gene expression away from injury-activated inflammation pathways. Cross-species replication is worth noting — it's weak evidence on its own, but it's the kind of convergence that makes a mechanism more credible.
Why does the timing matter so much?
In most injury treatments, the window is early — you intervene during or right after the event. The ISR results invert that. The benefit showed up when treatment began after the acute phase was over, and it persisted after dosing ended. That pattern only makes sense if what's being corrected is an ongoing, self-sustaining signaling state — the ISR keeping itself active — rather than a one-time injury that has already run its course. Switch the state off and the system, in these animals, largely returns to baseline.
Does any of this work in humans?
No. This is the sentence the whole article is built around. Every result above is from animals. There are no published human trials of ISRIB for traumatic brain injury, and none of these studies used A15. Animal-to-human translation in neuroscience fails far more often than it succeeds, and "reverses deficits in a mouse" has historically been a poor predictor of "helps a person." Treat this as a promising mechanism with zero human efficacy data for this indication — not as an established effect.
Can I take ISRIB A15 for a head injury?
Read the previous section again. A15 is sold as a research compound, not a treatment, and there is no human protocol for TBI to follow — dosing, timing, safety in an injured brain, none of it has been established in people. If you have a head injury or lingering post-concussion symptoms, the correct step is a clinician, not a research chemical. Nothing here is medical advice.
How is A15 related to the ISRIB in these studies?
The studies used ISRIB itself. A15 is a structural analog that hits the same eIF2B target with better oral bioavailability (more on that in the complete A15 guide). Mechanistically the two are expected to behave similarly, but "expected to" is doing real work in that sentence — the TBI data is ISRIB data, and it would be dishonest to present it as if A15 had been tested in these models. It hasn't.
The honest summary
What the research supports is narrow and genuinely interesting: in several animal models, and now across two species, chronic cognitive deficits after brain injury track with a stuck integrated stress response, and inhibiting that response — even late, even briefly — reverses the deficits in ways that persist. That is a real, replicated preclinical finding.
What it does not support is any claim about people. The distance between "reverses memory deficits in injured mice" and "does anything for a human with a concussion" is large, unbridged, and easy to underestimate when the animal data looks this clean. Both things are true at once, and the honest position holds both.
Where to buy ISRIB A15
In-house synthesized · 98%+ purity · Research compound
Individual results vary. For research purposes only.
Buy ISRIB A15 →One of the earliest independent synthesizers of ISRIB A15. Background in medicinal chemistry and small-molecule synthesis. Writing about compounds I've actually made.
Full profile →