ISRIB

ISRIB stands for integrated stress response inhibitor. It emerged from a screen for compounds that prevented stress-induced ATF4 production, and its first report showed enhanced long-term memory in mice and rats. The experiments used trans-ISRIB, the more active isomer of a symmetric molecule built around a cyclohexane ring with two glycolamide arms [8].

Healthy and aged mice. In the original weak-training water maze, mice given 0.25 mg/kg after training reached the platform faster and remembered its location better than controls. A separate 2.5 mg/kg experiment improved contextual fear memory at 24 hours but not at 1 hour, pointing to memory consolidation rather than an immediate performance boost [8]. In approximately 19-month-old mice, three 2.5 mg/kg injections improved spatial learning and later performance on a different working-memory task, assessed 18–21 days after the last injection. Young mice did not benefit under the same training conditions. The study also found changes in neuronal excitability and dendritic spines; it did not establish permanent reversal of aging [3].

Traumatic brain injury. A 2017 study began treatment 27 days after focal injury or 14 days after concussive injury. Three or four injections improved the respective spatial and working-memory tasks, and focal-injury mice retained the learned location a week after training without further dosing [9]. A later concussive-injury study used four injections on days 14–17 and found improved object recognition on day 18 and again on day 53. It also normalized abnormal cortical spine turnover. These results concern mouse injury models, not a demonstrated treatment for human concussion [4].

Alzheimer's models give conflicting answers. The differences matter more than a count of positive papers.

Prion disease. In infected mice treated after synapse loss had begun, ISRIB partly restored brain protein synthesis and preserved neurons without reducing prion accumulation. The mean endpoint was delayed from 84 to 96 days after infection. However, treated animals reached the humane endpoint through 20% weight loss, despite remaining protected from the usual neurological signs; this was not an uncomplicated survival benefit [13].

Human evidence. A 2025 ALS review found no published ISRIB trials. It described an online spreadsheet of 42 self-users, with subjective reports and no standardized functional, respiratory or laboratory measurements. Those reports cannot establish efficacy [7]. A paper describing an assay for trans-ISRIB in human plasma is analytical-method validation, not evidence of a human treatment trial [14].

Cells use the integrated stress response to alter protein production under stress. Several stress-sensing kinases converge on phosphorylation of eIF2α. Phosphorylated eIF2 inhibits eIF2B, the exchange factor that recycles eIF2 for translation initiation. This suppresses much ordinary protein synthesis while favoring translation of selected stress-response proteins, including ATF4 [2][8].

ISRIB binds eIF2B directly. Structural work placed it in a pocket at the interface between eIF2B subcomplexes, helping stabilize the active ten-subunit complex. Subsequent biochemical work showed allosteric opposition between ISRIB and phosphorylated eIF2: they bind separate sites but favor different eIF2B conformations. ISRIB can therefore make eIF2B less sensitive to the inhibitory signal, rather than simply preventing that signal from being generated [1][2]. It did not directly block PERK activation or eIF2α phosphorylation in the discovery experiments [8].

The effect has limits. Cell experiments found that sufficiently high eIF2α phosphorylation overcame ISRIB, even when the concentration was raised. Its activity therefore depends on the intensity of stress and the state of the translation machinery, rather than switching the entire response off at every stress level [15]. This helps explain why an effect in one disease model cannot be assumed in another; some negative Alzheimer's studies found little evidence of the elevated baseline stress signaling the treatment was intended to counter [5][10].

Trans-ISRIB inhibited the original cellular ATF4 reporter at 5 nM, compared with 600 nM for the cis-isomer. The founding investigators used the trans form for subsequent experiments [8].

Poor solubility complicates exposure. In a mouse study, increasing the injected dose from 2.5 to 5 mg/kg did not increase measured plasma or brain concentrations. The authors identified solubility as a possible explanation and a barrier to development. Rodent injection results do not demonstrate that an oral product reaches equivalent concentrations [13].

Human safety remains uncharacterized by controlled ISRIB trials in the 2025 review. The ALS self-use reports mentioned mild constipation and headache, but cannot provide reliable adverse-event rates or establish causality [7].

The animal findings depend on the model and regimen. Daily 0.25 mg/kg preserved pancreatic tissue and glucose control in prion-infected mice, unlike the direct PERK inhibitor used as a comparator. Nevertheless, the ISRIB-treated animals lost enough weight to require euthanasia; the cause was unresolved [13]. In a different study, 5 mg/kg daily increased mortality in APPSwe mice and their nontransgenic littermates within eight days. The same dose was much better tolerated in another mouse background, so neither result defines a universal safety threshold [10].

The stress response can also protect cells. In the discovery experiments, ISRIB increased cell death during persistent experimentally induced endoplasmic-reticulum stress. That is a measured laboratory liability, not evidence that routine human use has a known toxicity rate [8]. The increased hippocampal phosphorylated tau in treated PS19 mice is another reason to retain the negative findings alongside positive memory results [10].