The integrated stress response in brain diseases: A double-edged sword for proteostasis and synapses

The integrated stress response (ISR) is a highly conserved biochemical pathway that regulates protein synthesis. The ISR is activated in response to diverse stressors to restore cellular homeostasis. As such, the ISR is implicated in a wide range of diseases, including brain disorders. However, in the brain, the ISR also has potent influence on processes beyond proteo-stasis, namely synaptic plasticity, learning and memory. Thus, in the setting of brain diseases, ISR activity may have dual effects on proteostasis and synaptic function. In this review, we consider the ISR ’ s contribution to brain disorders through the lens of its potential effects on synaptic plasticity. From these examples, we illustrate that at times ISR activity may be a “ double-edged sword ” . We also highlight its potential as a therapeutic target to improve circuit function in brain diseases independent of its role in disease pathogenesis.


Introduction
The integrated stress response (ISR) is widely known for its role in restoring protein quality control, or proteostasis, when cells experience cell stress due to a wide variety of insults [1].The nature of cell insults triggering the ISR is revealed by conditions that activate the four eukaryotic translation initiation factor 2 alpha (eIF2a) kinases, protein kinase RNA-like ER kinase (PERK), general control nonderepressible 2 (GCN2), heme-regulated eIF2a kinase (HRI), and protein kinase R (PKR), that respond to ER stress and misfolded proteins, amino acid deprivation, iron deficiency and double stranded RNA viruses, respectively [2] (Figure 1).When eIF2a is phosphorylated, global protein synthesis rates are markedly reduced, accompanied by a switch in which open reading frames are translated such that a select group of mRNAs, which previously did not translate, now do [3].These ISR effector proteins include the transcription factor, ATF4/CREB2.ATF4 regulates the transcription of many stress response genes that function to restore cellular homeostasis.Cells modulate ISR activation through phosphatase activity of protein phosphatase 1 (PP1) complexed with either the growth arrest and DNA damage-inducible protein (GADD34) or the constitutive repressor of eIF2a phosphorylation (CReP) regulatory subunits [4].GADD34, specifically, is upregulated upon ISR activation as a negative feedback loop to limit the duration of ISR activation.Prolonged or excessive activation of the ISR leads to CHOP dependent apoptosis [1].
Cell stress and protein quality control disruption are major themes in the pathogenesis of many brain disorders.As such, the ISR has been tested as a candidate therapeutic target for a wide range of neurological disorders, with the most obvious application to neurodegenerative diseases.However, evolving studies have uncovered roles for translational regulation by the ISR pathway that appear to extend well beyond cell stress response.Such non-canonical roles include control of synaptic plasticity [5e7], circadian rhythms [8e10], axonal pathfinding [11] and neuromodulation [12,13].These brain-specific functions of this conserved biochemical pathway provide a lens to reconsider the significance of ISR contributions in brain diseases e solely as protector of protein quality control/cell stress, regulator of plasticity, or both.
an inciting mechanism, such as Alzheimer's disease (AD) and Parkinson's disease (PD), and others that may involve the ISR more through its roles in synaptic plasticity than through proteotoxic stressors, such as substance use disorders, dystonia, and autism spectrum disorders (ASD).There are a number of recent reviews that compliment this discussion; these include reviews focused on the ISR [1], genetic disorders involving the ISR [14], the ISR in brain disease [7,15,16], and the ISR in Alzheimer's disease [17].Understanding the multiple functional consequences and cellular sites of ISR activity in the brain is crucial to harness its therapeutic potential.

ISR-regulated protein synthesis in synaptic plasticity
Experience-dependent synaptic strength remodeling, known as synaptic plasticity, is crucial for an organism's ability to adaptively interact with its environment through learning and memory processes.Long-lasting synaptic plasticity involves two main forms: long-term potentiation (LTP) and long-term depression (LTD), which respectively strengthen and weaken synaptic strength.De-novo protein synthesis is required for most long-lasting forms of synaptic plasticity [18].Local protein synthesis either pre-[19,20] or post-synaptically at the synapses undergoing plasticity is one major site for this requirement [21e23].Protein synthesis regulatory pathways, such as the ISR and the mechanistic target of rapamycin complex I (mTORC1), play significant roles in plasticity mechanisms [7,24].To date, the effects of ISR manipulations on learning and synaptic plasticity have been observed across a broad range of species (sea slug [25], mouse, chicken [26], rat [27]) and brain regions (hippocampus, amygdala, ventral tegmental area, striatum, and cortex) and are more extensively discussed in a number of recent reviews [6,7,16,17].
Early evidence showcasing that the ISR pathway influenced synaptic plasticity through ATF4/CREB2 was in the sea snail, Aplysia [25].Inhibition of ATF4 reduced the amount of activity required for induction of longterm facilitation without altering its magnitude.Subsequent pharmacological and genetic manipulations targeting the ISR pathway continue to reinforce and expand its role in modulating synaptic plasticity, learning, and memory [5,26,28e34].ISR alterations most often seem to recalibrate the activity threshold required to induce long-term synaptic plasticity, with reductions in ISR activation favoring LTP and impairing LTD, as most comprehensively described in Ref. [5].Notably, several learning and memory tasks demonstrate that diminishing ISR activity also enhances performance by reducing the amount of experience/training needed to form memories [5,31,32,34].Conversely, studies in the hippocampus and striatum demonstrate that ISR activation is necessary for forms of LTD [35,36].These findings indicate that while ISR inhibition has shown memory enhancing effects in a range of tasks, behaviors that rely on LTD mechanisms would be predicted to be enhanced by ISR-boosting manipulations.Lastly, while we present a general model for ISR effects on LTP and LTD (Figure 1); there are studies that suggest this rule may not be universal [13,37,38].Currently, the specific Basic elements of the integrated stress response pathway and a schematic highlighting additional roles for the ISR in the brain for synaptic plasticity.a.The four main ISR-activating kinases are HRI, GCN2, PERK, and PKR, all of which phosphorylate eIF2a at Serine 51.There are two regulatory subunits, GADD34 and CReP, that dephosphorylate eIF2a when complexed with Protein phosphatase 1.When eIF2a is phosphorylated, global protein translation is halted and there is a selective increase in the translation of specific proteins, one of which is ATF4, that mediate recovery of cellular homeostasis.Prolonged or excessive ISR activation can lead to CHOP-dependent apoptosis.b.In the brain, the ISR has been identified as a powerful modulator of synaptic plasticity and experience-dependent learning.In addition to canonical ISR functions of cell stress response, ISR activation and inhibition influence thresholds for induction of long-term synaptic plasticity.In many, but not all, examples studied to date, inhibiting the ISR favors induction of long-term potentiation (LTP), whereas activation the ISR favors long-term depression (LTD).
relationships between ISR activity and the underlying cells, synapses, circuits, and behaviors involved are not fully established.
Recent studies have begun to fill these gaps by defining cell and circuit-specific effects of ISR activity [12,13,28,39,40] (Figure 2).For example, in the hippocampus, ISR actions in excitatory neurons and somatostatin interneurons, but not parvalbumin positive interneurons, facilitated LTP of excitatory hippocampal neurons and enhanced memory [39].Importantly, this and several other recent studies further define non-cell autonomous sites for ISR effects on synaptic plasticity, learning and memory.These studies illustrate roles for cholinergic [12] and dopaminergic [13] neuromodulatory neurons, and even astrocytes [40].

ISR contributions to synaptic plasticity perturbations in disease
In the examples that follow, we aim to highlight 3 themes in considering the contribution of the ISR to synaptic function in disease.In many diseases that are associated with aging, mitochondrial dysfunction and protein misfolding such as Alzheimer's disease(AD), Parkinson's disease(PD), amyotrophic lateral sclerosis, prion disease, multiple sclerosis, traumatic brain injury (TBI) and others, the cellular response to diseaserelated pathologies activates the ISR and changes proteostasis [16].This cell pathology-driven change to ISR activity can have subsequent consequences for synaptic function.There also appear to be a second group of diseases in which ISR synaptic effects may be the primary mechanism, without obvious cell stress processes driving the disease.And finally, even if the ISR is not responsible for synaptic dysfunction associated with a given disease, the ISR may nonetheless offer a corrective course for therapeutic intervention through its influence on synaptic plasticity.

Alzheimer's disease
The contributions of the ISR to AD have been a central focus in the field due to AD's misfolded protein pathology, oxidative stress, and other cell stress events [16,17,41e46].In both mouse models of AD and post- mortem human tissue, levels of p-eIF2a, p-PERK, and ATF4 are elevated, consistent with ISR activation [45e48].Although relationships between ISR activity and AD pathology and pathogenesis have been established, many studies targeting the ISR have not altered amyloid or tau pathologies in mouse models [43,47,49e52].Nonetheless, some ISR manipulations in AD models ameliorate learning and memory impairments and aberrant LTP [47,53e60].This is supported in the familial AD 5xFAD mouse model, where inhibiting PERK or PKR ameliorated impaired spatial memory and corrected LTP plasticity induction [47,53,57,61].However, it should also be mentioned that other studies failed to show that reducing the ISR, with phospho-mutant forms of eIF2a, by GCN2 knockout, or pharmacologically was therapeutically beneficial in models of AD [49,62,63].
Where, or in which cell type, the ISR may be exerting its influence on plasticity in AD progression remains unexplored.Notably, the cholinergic system is a focal point in AD research due to its role in cognition [64].Recent work found that striatal cholinergic interneurons [12] tonically activate the ISR in the normal brain.Cellselective ISR inhibition in those neurons causes enhanced performance in Morris water maze and level press task, raising the possibility that ISR actions in the cholinergic system could contribute to some of the behavioral effects of ISR inhibition.It also points to another potential site for ISR intervention, where modulating the ISR might combat cholinergic hyperactivity driven by Ab [65e67].

Traumatic brain injury
TBI results from excessive acceleration and deceleration forces to the brain and is pathologically accompanied by diffuse axonal injury, which can lead to death in severe cases or a range of long-lasting cognitive impairments.Rodent models demonstrate neuronal injury and increases in ISR components such as p-eIF2a, p-PERK and ATF4 from at least 1 to 30 days post injury [68e70].
TBI rodent models show less hippocampal LTP induction [71].Acute ex-vivo treatment with ISRIB applied to hippocampal brain slices was sufficient to increase LTP magnitude, using a low induction protocol [68].Although there is a focus on white matter injury and loss in TBI, the acuity of this effect on LTP argues for a different mechanism.Accordingly, prior studies suggest that LTP induction deficits in TBI are more likely related to synaptic dysfunction than lack of axonal input.In hippocampal immunohistochemistry, the neuronal pyramidal layer shows the highest increase in p-eIF2a [70].Thus, while rationale that contributions from microglia, oligodendrocytes, and astrocytes exists [72], ISR actions in neurons may play a substantial role in LTP induction deficits.Studies of the ISR in TBI highlight that both white matter and synaptic effects may coexist [73] and targeting the ISR may be one avenue to therapeutically address both of these pathologies.
In TBI, two distinct ISR-modifying approaches yield benefit: Guanabenz, a GADD34 phosphatase inhibitor, when administered around the initial injury to boost p-eIF2a levels, improved motor performance in treated mice as soon as one day after treatment and lasting for several days thereafter [70].Conversely, post-injury treatment with ISRIB enhanced learning rates and memory retrieval when administered 2e4 weeks after injury but just prior to and during the learning paradigm [68].These findings suggest that the beneficial directionality of ISR modulation may vary with disease phases, yet bidirectional tests in the same phase remain unexplored.
The beneficial effects of opposing ISR manipulations during different phases of disease suggest they might act on different underlying mechanisms.If ISR effects solely on plasticity are considered, pre-and peri-insult ISR boosting might lower potential for "excitotoxicity" by inhibiting excessive LTP induction.Conversely, weeks following TBI, LTP-dependent behavioral learning could be enhanced by inhibiting the ISR during learning.

Parkinson's disease
Accumulation of alpha-synuclein into Lewy bodies is a pathological hallmark of PD.These Lewy bodies trigger the UPR and ISR responses in the brain [74].In small human pathology studies, findings suggestive of an activated ISR have been reported.Elevated levels of p-PERK and p-eif2a have been found in dopaminergic neurons of people with PD, especially around areas with high a-synuclein aggregates [75] and increased PKR phosphorylation has been described in hippocampi of PD and HD subjects [76].
In PD model systems, both increased and decreased ISR manipulations have been reported to be beneficial in slowing disease progression.These findings suggest the ISR's potential role as a modifier of PD pathogenesis, yet the optimal directionality and timing of treatments are not fully established [77e84].Similar to the varying effects of ISR modulation in TBI discussed earlier, PD model studies suggest distinct consequences of the ISR at different disease stages.Early on, prior to toxin exposure (6-hydroxydopamine, 6-OHDA), priming the ISR may help the cell respond to a pathogenesisinducing stressor and generate less apoptosis [77,81].Recently, this idea has been put forth to explain the epidemiological finding that PD is inversely correlated with nicotine usage, as chronic nicotine exposure blunted subsequent acute ISR activation by cell stressors [85,86].
In PD models, dendritic atrophy and postsynaptic spine loss have been observed in indirect, but not direct, spiny projection neurons [87,88].Among other possibilities, this is a synaptic phenotype that could arise from excessive synaptic weakening via an LTD-like mechanism.Since ISR activation is required for corticostriatal LTD on indirect pathway projection neurons [36], increased ISR activity in PD might act in indirect pathway neurons to cause a bias away from LTP and towards LTD.Thus, independently of whether the ISR is effectively ameliorating proteostatic challenges in PD, chronic ISR activation can be hypothesized to have collateral effects on synaptic function that contribute to the symptomatology of PD.

Prion disease
Another subset of neurodegenerative disorders related to protein misfolding is prion diseases.Prion diseases are caused by the misfolding of prion protein (PrP) into a pathogenic form, PrP Sc [89,90].There is a long incubation time for the disease, however, after disease onset, there is a rapid deterioration of cognition and eventually death [91].PrP formation and aggregation would be predicted to potently activate the ISR/UPR through misfolded proteins.
In mouse models of prion disease, animals begin to show structural and electrophysiological signs of synaptic loss/ weakening and behavioral abnormalities 7e9 weeks after inoculation, whereas neuronal degeneration appears later, followed by death around week 13 [92].This period of early synaptic dysfunction has been identified as a key window of opportunity to correct and prevent further synaptic loss and neuronal degeneration.Depletion or inhibition of PrP in this window can reverse prion pathology [93e95].In-vitro and in-vivo treatment with the ISR boosting drug, Sephin1, reduces the amount of PrP Sc and subsequently increased viability [96].Synaptic impairment has also been found to correlate with the severity of degeneration, and interventions that prevent or correct synaptic dysfunction are able to slow disease progression and prevent neurodegeneration in prion disease [91,97].For example, a study aimed at treating synaptic dysfunction alone, using M1 muscarinic acetylcholine receptor positive allosteric modulators, corrected memory deficits, improved motor function and extended lifespan, further suggesting that maintaining synaptic function is a key site of therapeutic intervention [98].
In mouse models, synaptic loss in prion disease is accompanied by signs of increased ISR activation, raising a question of cause and effect for the ISR on synaptopathology in this protein misfolding disorder.Multiple studies treating PrP-inoculated mice with ISR inhibiting molecules, such as, ISRIB [99], PERK inhibitors [100], and Trazodone [60,101] (which has been shown to act through the ISR to restore global protein synthesis) have shown beneficial effects on a wide range of readouts that include synaptic function, behavior, neurodegeneration and lifespan.Intriguingly, synaptic dysfunction in prion disease may develop through non-cell autonomous effects.In prion disease mouse model, astrocytes were found to show features of ISR/UPR activation, impaired synaptogenesis properties in an in vitro assay and a grossly altered profile of secreted proteins [102].Homeostatic astrocytes secrete factors that promote synaptogenesis and synaptic maintenance [103,104], while the secretome from astrocytes undergoing elevated ISR/UPR activation had a significant reduction of known synaptogenic molecules [102].Astrocyte-selective ISR inhibition showed beneficial effects on behaviors associated with synaptic dysfunction and limited neurodegeneration.Interestingly, in this study [102], cell-specifically reducing the ISR in astrocytes had similar protective effects as globally genetically or pharmacologically reducing the ISR [60,92,99e101].In summary, prion disorders which involve protein misfolding as an inciting event, also show ISR activation and synaptic weakening.The summarized prion studies highlight that these ISR effects on synapses are likely to play a central role in downstream consequences of behavior and neurodegeneration, and when targeted early enough, are sufficient to delay pathogenesis.These studies provide another example where ISR activation due to misfolded proteins may exert clinically relevant sequelae through ISR "bystander" effects on synapses.
Neurodevelopmental disorders: autism, Down syndrome, and Fragile X syndrome Mechanisms for autism spectrum disorders (ASD) often involve the regulation of protein synthesis and synaptic function [105,106].Many disease models show synaptic plasticity disruptions, which are the hypothesized mechanisms for clinical symptomatology [7,107].Recent studies in Down's syndrome (DS) and Fragile X syndrome (FXS) specifically implicate the ISR as a contributor to disease phenotypes.
DS arises from the triplication of chromosome 21.Mouse models show synaptic plasticity deficits of reduced LTP and increased LTD in hippocampal slices [108] and impaired behavior [109].Human and mouse tissue show elevated levels of phosphorylated eIF2a, increased levels of PKR, and reduced protein synthesis, suggesting elevated ISR activity in the disease [110].In a DS mouse model, both pharmacological and genetic reductions of the ISR corrected these elevated ISR phenotypes and also enhanced LTP and rescued memory impairment [110].
Fragile X messenger ribonucleoprotein 1 (FMR1) mutations cause FXS and are associated with a net increased protein synthesis [111,112].A hallmark of this disorder is excessive induction of a form of LTD initiated by metabotropic glutamate receptors (mGluR) [113].Recent studies identify contributions of dysregulated ISR activity to FXS phenotypes [114].This cellspecific study revealed lowered p-eIF2a levels driven by upregulated signaling of the mechanistic/mammalian target of rapamycin complex 1 (mTORC1) pathway in excitatory, but not inhibitory, neurons [114].Manipulations enhancing ISR activity (CReP shRNA) in excitatory neurons rescued the FXS LTD phenotype as well as behavioral deficits associated with ASD behaviors [114].
In the FXS model, excessive LTD signaling is thought to arise from a lack of FMRP's negative regulation of LTD transcripts [113].The enhanced protein synthesis seen in FMR1-KO mice suggest a decreased ISR state and that would predict LTD impairment [35].However, this model shows enhanced hippocampal LTD.Notably, increased mGluR LTD has also been reported with PERK deletion [51] another manipulation that would reduce the ISR.Thus, a blunted ISR may contribute to this specific FXS phenotype.These observations also illustrate that a fuller understanding of the ISR in specific cells and specific plasticity paradigms are likely to reveal a more nuanced appreciation of its role in plasticity beyond ISR activation favoring LTD and ISR inhibition favoring LTP.
In addition to genetic forms of neurodevelopmental disorders, it is well established that environmental factors such as maternal diet [115], environmental toxins [116], and maternal infections [117] increase the risk of neurodevelopmental abnormalities in offspring in both mouse models and human epidemiological studies [118e120].Interestingly, male offspring are more susceptible to developmental abnormalities as a result of maternal insults [121].Male offspring in a viral mimetic model of maternal immune activation (MIA) display impaired sociability and increased marble burying.In addition, there are sexually dimorphic effects on protein synthesis in this model; males, but not females, show reduced translation and ISR activation (p-eIF2a levels) in the fetal brain [122].Evidence suggests that ISR activation may play a role behaviorally, as these behavioral abnormalities did not develop when MIA was performed in eIF2a phosphomutant mice [122].The sexual dimorphism in ISR biology in this model and others and in autism-related disorders clinically suggest that there are likely to be additional insights to be gained from further understanding sex-specific differences in stress responses and translational regulation.

Vanishing white matter disease
Vanishing white matter disease (VWMD) is a rare, autosomal recessive leukodystrophy that arises from mutations in eIF2B.VWMD pathology is characterized by sudden myelin degeneration often triggered by an external stressor.Symptoms of VWMD include cognitive and motor impairments.Since eIF2B is part of the eIF2 ternary complex in which eIF2a acts to regulate ISR state, disease mechanisms for this disorder likely include the ISR.Characterizations of how the human mutation changes the ISR have shown excessive ISR response and heightened basal ISR activation [123,124].In a mouse model for VWMD, hypomyelination has been reported early in development despite abundant oligodendrocytes [125].As discussed later, white matter changes are seen in other ISR disorders like dystonia.A role for ISR dysregulation in VWMD pathogenesis is supported by improvement of brain myelin in response to the drugs ISRIB and 2BAct [123].The exact cellular sites involved and whether there is concomitant synaptic pathology is unknown.As yet, no cell autonomous studies have been performed to assign the deleterious myelination effects or restorative effects directly to oligodendrocytes.At a minimum, this disorder highlights another non-canonical ISR role in the brain related to myelination.However, even in this white matter disease, a synaptic contribution can also be speculated based on the Hebbian principle that functionally connected circuits promote neuronal coupling.During development and throughout the lifespan, oligodendrocytes and oligodendrocyte precursor cells exhibit activity-dependent myelination [126].In this case, excessive ISR activity could inhibit LTP induction and drive excessive synaptic weakening with downstream consequences of reducing myelination.While synapses and plasticity have not been studied, to our knowledge, in the VWMD mouse model, there is the potential for multiple modes of ISR contributions to this disorder's pathogenesis.

Dystonia
Dystonia is a movement disorder that is characterized by involuntary, uncoordinated twisting movements and abnormally sustained postures, often involving simultaneous co-contraction of agonist and antagonist muscle groups [127].Broad disruptions of motor network circuitry are observed, including the basal ganglia.When dystonia is the sole clinical presentation, typically there is no overt cell pathology or neurodegeneration.Rather, disruption of synaptic plasticity processes is thought to contribute [128e130].Rodent models of a monogenic form of dystonia, DYT-TOR1A, show increased magnitude of LTP and impaired LTD at corticostriatal synapses and disrupted dopamine signaling in striatal cholinergic interneurons [131e134].Recent studies coalesce on the ISR as a common mechanism for multiple dystonias supported by human genetics [36,135e137], and additional functional phenotypes observed in mouse models of genetic forms not obviously related to ISR genes [36,138,139].Though, it is not yet clear whether all dystonias share a common directionality of ISR impairment.For example, DYT-PRKRA mutations are thought to excessively activate the ISR [135], while ISR dysregulation suggestive of weakened activity has been reported in DYT-TOR1A patient fibroblasts [36], DYT-THAP1 [139] and ERstressed DYT1 mouse models [138].In DYT-TOR1A cell and animal models, responses to ISR tool compounds support the directionality of ISR boosting being beneficial and ISR inhibiting manipulations being sufficient to reproduce disease phenotypes [12,36,140,141].One possible unifying idea is that dystonia mechanisms cause chronic ISR activation that, through feedback mechanisms, weaken the magnitude of acute phasic ISR activation.For example, DYT-TOR1A human fibroblasts show elevated basal levels of CReP, a regulatory protein that supports PPP1 phosphatase activity on eIF2a [36].
Pharmacological inhibition of eIF2a phosphatases rescues striatal LTD deficits in mouse models of DYT-TOR1A [36] and DYT-THAP6 [139] and ISRIB blocks corticostriatal LTD in wildtype conditions [36].Additionally, recent studies directly implicate the ISR in another physiological hallmark of dystonia models involving striatal cholinergic interneurons.These cells were recently found to be somewhat unusual in their high steady state activation of the ISR in the normal brain [12].In that study, lowering the ISR in cholinergic interneurons (CReP overexpression) specifically recapitulated the abnormalities in dopamine signaling in cholinergic interneurons that characterize dystonia models [132,142].These observations introduce the idea of certain cell types that may be more susceptible to ISR-modifying perturbations because of high basal demand.In addition, since dopamine signaling in cholinergic neurons facilitates the form of corticostriatal LTD that is impaired in dystonia models [143], these findings again highlight the possibility that ISR actions on synaptic plasticity may be exerted through non-cell autonomous mechanisms.

Substance use disorders
A common feature to many drugs of abuse that lead to substance use disorders (SUDs) involves dopaminedependent remodeling of synaptic strength, especially in limbic circuitry.These mechanisms often co-opt normal synaptic plasticity mechanisms [144].Consistent with this, changes in the relative amounts of phosphorylated eIF2a have been identified at distinct temporal phases in several SUD models.In models using cocaine and opioids, levels of p-eIF2a in the basolateral amygdala decreased during drug reconsolidation and during the time period after drug re-exposure [27].Similarly, in the ventral tegmental area, eIF2a dephosphorylation was observed after cocaine administration and association training [145].Conditions that lowered p-eIF2a (Eif2a1 S/A phospho-mutant mice and adolescent, as opposed to adult, mice) also lowered the threshold of cocaine dosage needed to evoke conditioned place preference and LTP of glutamatergic synapses on dopamine neurons [145,146].These findings suggest that manipulations increasing the ISR during key periods might reduce or impede drugs of abusedependent plasticity processes.Recent studies provide examples of using the ISR therapeutically to oppose drug-induced processes of synaptic strengthening [147,148].Treatment with Sal003, an enhancer of the ISR, after drug exposure, decreases drug-induced LTP and behavioral changes, reduced drug-seeking behaviors, and prevented drug-paired stimulus-induced cravings [148].
In SUDs, components of the ISR have been shown to change, ISR manipulations have modified behavioral and plasticity outcomes, and there is some human genetic evidence that Eif2s1 genetic variants may influence tobacco use [149].The SUD field nicely illustrates that an understanding of the protein synthesisdependent phases of synaptic plasticity in behavioral adaptations can be used to target interventions and interfere with processes like incubation of craving and long-lasting memory consolidation.Additionally, the understanding of how ISR activity changes through developmental time windows highlight how basal ISR states in the brain might confer vulnerability to different neurological disorders.

Conclusion
In this review, we have considered the significance of ISR effects on synapses and synaptic plasticity in a number of diseases in which the ISR has been associated.In the examples discussed, a few themes emerge.First, in the cases of brain diseases stemming from pathology that activates cell stress responses, the ISR may be a double-edged sword, having dual effects related to its role in proteostasis and its capacity to modify synaptic plasticity processes (e.g., AD, PD, Prion disease, and TBI).These "bystander" consequences for synapses can also contribute to clinical symptomatology.In AD, it is notable that acute ISR manipulations restore LTP deficits, independent of any observable effect on cell pathology hallmarks.Second, there are a group of disorders in which the ISR's role in synaptic processes may be the primary mechanism.Autism spectrum disorders, dystonia and substance use diseases were discussed in this context.These diseases typically lack any overt cell pathology or neurodegeneration and include examples where monogenic forms are directly associated with genes regulating protein synthesis.These diseases establish relationships between ISR activity and synaptic plasticity in a range of circuits implicated in the particular clinical manifestations of each condition.Third, by highlighting the broad and potent influence the ISR can have on synaptic plasticity, learning and memory, its potential as a therapeutic avenue was highlighted as a means to modify synaptic function.In these cases, the ISR can be another "knob to turn" in the effort to restore synaptic and circuit function in a range of conditions, irrespective of its involvement in the underlying disease pathogenesis.As to how the ISR modifies synapses, we note that while we presented a relatively simple model for ISR effects that are supported by findings in a range of circuits (Figure 1b), several other results already demonstrate that this is unlikely to be universally true.Current efforts using cell-specific approaches are beginning to establish the peculiarities of ISR actions in distinct cells, synapses, and circuits (Figure 2).This roadmap will be highly instructive to anticipate and guide ISR applications for therapeutic goals.

Declaration of competing interest
The author (N.C.) has patents and patents pending related to methods for dystonia diagnosis and treatment involving the ISR.The authors have no financial disclosures.

Figure 2 Summary
Figure 2

Finally
, two other themes were highlighted.First, the stage of the disease process may influence the outcome of the ISR intervention.Studies in TBI and PD include examples where ISR interventions of both directionalities, activation and inhibition, improve disease phenotypes.In these examples, differences in the timing of the intervention e such as prior to or periinsult versus in later stages of the disease may offer mechanistic clues.And second, synaptic plasticity effects may be mediated by ISR actions in cells other than those in which the plasticity manifests.Recent literature examples of cell-specific effects of the ISR in dopaminergic neurons, cholinergic interneurons, and astrocytes influencing synaptic plasticity in other cells, learning, and memory emphasize the role of non-cell autonomous contributions.CRediT authorship contribution statementElana R. Lockshin: Conceptualization, Investigation, Data curation, Writing e original draft, Writing e review & editing, Funding acquisition.Nicole Calakos: Conceptualization, Investigation, Supervision, Writing e review & editing, Funding acquisition.
Translational control in the brain in health and disease.Cold Spring Harb Perspect Biol 2019, 11.This study discovered that there are cell types that tonically engage the ISR to maintain normal cellular function and physiology, striatal cholinergic interneurons.Cell-specific ISR inhibition in striatal cholinergic interneurons led to changes in their response to dopamine and modified learning.It provides an example of tonic ISR activity in the healthy brain, that counters canonical views of the ISR being engaged transiently, such as by cell stress or experience.The integrated stress response: from mechanism to disease.Science 2020, 368.et al.: Cell-type-specific drug-inducible protein synthesis inhibition demonstrates that memory consolidation requires rapid neuronal translation.Nat Neurosci 2020, 23:281-292.This study developed novel chemogenetic tools and approach to selectively activate the ISR in a cell-specific manner.This work also determined the role of the ISR in excitatory neurons of the lateral habenula during fear memory consolidation.29.JiangZ, et al.: eIF2a phosphorylation-dependent translation in CA1 pyramidal cells impairs hippocampal memory consolidation without affecting general translation.J Neurosci 2010, 30:2582-2594.30.Sharma V, et al.: Local inhibition of PERK enhances memory and reverses age-related deterioration of cognitive and neuronal properties.J Neurosci 2018, 38:648-658.31.Stern E, Chinnakkaruppan A, David O, Sonenberg N, Rosenblum K: Blocking the eIF2a kinase (PKR) enhances positive and negative forms of cortex-dependent taste memory.J Neurosci 2013, 33:2517-2525.32.Zhu PJ, et al.: Suppression of PKR promotes network excitability and enhanced cognition by interferon-g-mediated disinhibition.Cell 2011, 147:1384-1396.33.Costa-Mattioli M, Sossin WS, Klann E, Sonenberg N: Translational control of long-lasting synaptic plasticity and memory.Neuron 2009, 61:10-26.This paper elucidated mechanisms for the ISR in LTD.36 * .Rittiner JE, et al.: Functional genomic analyses of mendelian and sporadic disease identify impaired eIF2a signaling as a generalizable mechanism for dystonia.Neuron 2016, 92: 1238-1251.This paper first proposed that the ISR is a convergent pathological mechanism for multiple forms of dystonia.Requirement of ISR activity for corticostriatal LTD and corrective effects of boosting ISR on LTD deficits in a monogenic form of dystonia were also shown.37. Trinh MA, et al.: Brain-specific disruption of the eIF2a kinase PERK decreases ATF4 expression and impairs behavioral flexibility.Cell Rep 2012, 1:676-688.38.Pasini S, Corona C, Liu J, Greene LAA, Shelanski MLL: Specific downregulation of hippocampal ATF4 reveals a necessary role in synaptic plasticity and memory.Cell Rep 2015, 11:183-191.39 * .Sharma V, et al.: eIF2a controls memory consolidation via excitatory and somatostatin neurons.Nature 2020, 586: 412-416.This is the first of a series of recent studies performing cell-specific ISR manipulations.Authors investigated the circuit and behavioral effets of selective, genetic ISR inhibtion of excitatory neurons, somatostatin interneurons, and parvalbumin interneurons in the hippocampus.This study provides an important example of the distinct effects of ISR roles in excitatory and inhibitory neurons.40 * .Sharma V, et al.: mRNA translation in astrocytes controls hippocampal long-term synaptic plasticity and memory.Proc Natl Acad Sci U S A 2023, 120.This study performs cell-specific ISR manipulations in astrocytes and provides another example of non-cell autonomous sites for ISR action in synpatic plasticity.41.Ohno M: PERK as a hub of multiple pathogenic pathways leading to memory deficits and neurodegeneration in Alzheimer's disease.Brain Res Bull 2018, 141:72-78.