The ISR downstream target ATF4 represses long-term memory in a cell type–specific manner
Contributed by Nahum Sonenberg; received April 14, 2024; accepted June 25, 2024; reviewed by Clive Bramham and Satoshi Kida
Significance
The activation of the integrated stress response (ISR) underlies memory deficits in various cognitive disorders, but how ISR regulates long-term memory (LTM) remains largely unknown. Here, we show that deleting activating transcription factor 4 (ATF4), a downstream target of ISR, in excitatory neurons but not in inhibitory and cholinergic neurons or astrocytes bolsters LTM-associated behaviors. Therefore, in excitatory neurons, ATF4 plays a major role in regulating ISR-mediated mnemonic processes.
Abstract
The integrated stress response (ISR), a pivotal protein homeostasis network, plays a critical role in the formation of long-term memory (LTM). The precise mechanism by which the ISR controls LTM is not well understood. Here, we report insights into how the ISR modulates the mnemonic process by using targeted deletion of the activating transcription factor 4 (ATF4), a key downstream effector of the ISR, in various neuronal and non-neuronal cell types. We found that the removal of ATF4 from forebrain excitatory neurons (but not from inhibitory neurons, cholinergic neurons, or astrocytes) enhances LTM formation. Furthermore, the deletion of ATF4 in excitatory neurons lowers the threshold for the induction of long-term potentiation, a cellular model for LTM. Transcriptomic and proteomic analyses revealed that ATF4 deletion in excitatory neurons leads to upregulation of components of oxidative phosphorylation pathways, which are critical for ATP production. Thus, we conclude that ATF4 functions as a memory repressor selectively within excitatory neurons.
Sign up for PNAS alerts.
Get alerts for new articles, or get an alert when an article is cited.
The integrated stress response (ISR) is a phylogenetically conserved signaling network that maintains cellular proteostasis by tuning protein synthesis (1). A large body of evidence documents the ISR as a central molecular switch that regulates long-term memory (LTM) formation (1, 2). Genetic or pharmacological suppression of the ISR enhances the formation of LTM (3–6). In contrast, ISR activation, which leads to global reduction in protein synthesis but increase in the translation of activating transcription factor 4 (ATF4) (7) (Fig. 1A), impairs LTM (3, 8–10). However, it remains to be determined whether ATF4 serves as the primary downstream effector of ISR during the formation of LTM.
Fig. 1.
ATF4 belongs to the basic leucine zipper domain (bZip) family of transcription factors and functions in a context-dependent manner either as a transcriptional repressor or activator (11). ISR activation impaired protein synthesis–dependent long-term potentiation (LTP) in hippocampal slices from control (wild type; WT) mice but not in ATF4-deficient slices (3), suggesting that ATF4 is a major factor driving this process. Moreover, expressing a dominant negative inhibitor of ATF4 [and CCAAT/enhancer-binding proteins (C/EBPs)] in murine forebrain neurons promoted long-term synaptic plasticity and memory formation (12). In contrast to these findings, shRNA-mediated Atf4 knockdown in the mouse hippocampus impaired synaptic plasticity and LTM (13). It is noteworthy that many of the methods employed to investigate the involvement of ATF4 in the formation of LTM have important limitations as they can potentially a) target other proteins (e.g., C/EBPs) (12) and b) exhibit shRNA-mediated off-target effects (14–16), all of which may confound the interpretation of the results. Moreover, germline deletion of Atf4 in mice results in defects in ocular, skeletal, and hematopoietic development (17), rendering them unsuitable for studying LTM. In addition, compensatory/adaptive mechanisms could occur when a gene is deleted in the germline (18). To circumvent these shortcomings, we deleted Atf4 in different cell populations (excitatory neurons, inhibitory interneurons, cholinergic neurons, and astrocytes) to dissect the role of ATF4 in different brain cell types and investigated how it controls LTM. Unexpectedly, our results demonstrate that ATF4 represses the formation of LTM in a cell type–specific manner.
Results
Cell Type–Specific Expression of Atf4 mRNA in the Mouse Brain.
Atf4 mRNA is ubiquitously expressed in the mouse brain (Allen Mouse Brain Atlas; SI Appendix, Fig. S1A). To examine Atf4 levels in distinct cell types in the brain, we performed combined RNA in situ hybridization (ISH; RNAscope) and immunohistochemistry (IHC) on mouse hippocampal sections. This approach was chosen as commercial ATF4 antibodies do not provide reliable IHC signals in mouse brain sections. The combined ISH/IHC revealed that Atf4 mRNA was coexpressed with markers of various neuronal subtypes, such as calcium/calmodulin-dependent protein kinase type II subunit alpha (CaMKIIα) for excitatory neurons, glutamic acid decarboxylase 67 (GAD67) for inhibitory neurons, and choline acetyltransferase (ChAT) for cholinergic neurons (SI Appendix, Fig. S1B). However, only a small fraction of Atf4 mRNAs colocalized with glial fibrillary acidic protein (GFAP) (SI Appendix, Fig. S1B), indicating low basal ATF4 expression in the astrocytes, in accordance with previous reports (19, 20).
Deletion of ATF4 in Forebrain Excitatory Neurons Leads to Enhanced LTM.
To study the role of ATF4 in LTM formation in forebrain excitatory neurons, which are critically required for this process (21), we crossed Atf4fl/fl mice with Camk2a-Cre mice that express Cre postnatally in excitatory forebrain neurons (22) to generate cell type-specific knockout (Atf4fl/fl:Camk2a-Cre; hereafter denoted Atf4Ex cKO) and control (Atf4+/+:Camk2a-Cre) mice, respectively (Fig. 1 B and C). The selective deletion of Atf4 in excitatory neurons was confirmed by combined ISH/IHC (Fig. 1 D and E).
We first examined spatial LTM using the Morris water maze (MWM) test, in which mice use visual cues to find a hidden platform in a circular pool (23). We studied LTM as it requires protein synthesis which is controlled by ISR. Previous studies demonstrated that ISR inhibition facilitates spatial LTM formation under a weak training paradigm (one training session per day for 5 to 6 d; Fig. 1F) (3, 4). Atf4Ex cKO mice reached the hidden platform significantly faster than control mice (Fig. 1G; 45.44 ± 12.19% reduction in escape latency in Atf4Ex cKO vs. control on day 5 of training), indicating that ATF4 represses spatial LTM acquisition. Accordingly, unlike control mice, Atf4Ex cKO mice spent more time swimming in the target quadrant during the probe test (Fig. 1H; Atf4Ex cKO and control mice spent 32.76 ± 1.98% and 26.46 ± 1.17% of the total time in the target quadrant, respectively). Additional analysis revealed that Atf4Ex cKO mice, but not control mice, spend significantly more time in the target quadrant than chance, further demonstrating that weak training induces LTM in these animals. Notably, the enhanced LTM of Atf4Ex cKO mice cannot be attributed to enhanced locomotion since both genotypes swam a similar distance in the circular pool (Fig. 1I). Moreover, both control and Atf4Ex cKO mice performed comparably in a version of the MWM where the platform was visible to the mice (Fig. 1J). Thus, the deletion of Atf4 in excitatory forebrain neurons promotes spatial LTM.
We also examined spatial LTM using a more robust training paradigm (referred to as the standard MWM test hereafter), consisting of three training trials per day for 5 d (SI Appendix, Fig. S2A). Both control and Atf4Ex cKO mice performed similarly during the LTM acquisition phase (SI Appendix, Fig. S2B), and no difference was observed in the time spent in the target quadrant during the probe test (SI Appendix, Fig. S2C). Both control and Atf4Ex cKO mice spent significantly more time in the target quadrant than the baseline, indicating the formation of robust LTM in both these groups in the standard MWM protocol. Taken together, these data support the notion that the threshold for forming LTM is lowered in Atf4Ex cKO mice.
We next investigated hippocampus-dependent contextual fear memory. In this task, mice receive a foot shock (the unconditioned stimulus; US) in a chosen context (conditioned stimulus; CS). Twenty-four hours after training, mice are exposed to the CS, and their fear response is measured (4). We first used a weak training protocol, which consisted of a mild foot shock (0.35 mA for 1 s) (Fig. 1K). Twenty-four hours after training, Atf4 Ex cKO froze more (~1.9-fold increase) than control mice (Fig. 1L), demonstrating enhanced contextual LTM. The freezing behavior of Atf4Ex cKO mice was also increased (1.5-fold) when measured 24 h after a stronger training protocol (two shocks of 0.7 mA for 2 s) (SI Appendix, Fig. S2 D and E). No anxiety-like behavior or locomotor alterations were observed in Atf4Ex cKO mice (SI Appendix, Fig. S2 F–K). Thus, long-term fear memory is facilitated in mice lacking ATF4 in forebrain excitatory neurons.
Given that Atf4Ex cKO mice exhibited enhanced LTM, we next studied hippocampal LTP, a cellular model underlying LTM formation (24). A single train of high-frequency stimulation (1 × HFS) of the Schaffer collateral pathway elicited a long-lasting LTP only in slices from Atf4Ex cKO mice (Fig. 1 M and N) demonstrating that deletion of ATF4 in forebrain excitatory neurons lowered the threshold for LTP. Therefore, consistent with the genetic inhibition of ISR (5), deletion of Atf4 in the forebrain excitatory neurons bolsters synaptic plasticity and LTM formation.
Deletion of ATF4 in Inhibitory Neurons, Cholinergic Neurons, or Astrocytes Fails to Enhance LTM.
GABAergic inhibitory neurons play a critical role in LTM formation by regulating the activity of excitatory neurons (25). Astrocytes are believed to play a key role in LTM by modulating synaptic functions (26, 27). Genetic inhibition of the ISR in either GABAergic neurons or astrocytes facilitates LTM formation (5, 28). If ATF4 is the major effector of the ISR in inhibitory neurons and astrocytes, it is conceivable that deletion of the Atf4 gene in these cell types should lead to an enhanced LTM phenotype similar to that of excitatory neurons. To this end, we deleted Atf4 in inhibitory neurons by crossing Atf4fl/fl with mice expressing Cre recombinase under the glutamic-acid decarboxylase 2 (Gad2) promoter to generate GABAergic-specific Atf4 knockout (Atf4fl/fl:Gad2-Cre; hereafter denoted as Atf4In cKO) and control (Atf4+/+:Gad2-Cre) mice (Fig. 2A). To delete Atf4 in astrocytes, Atf4fl/fl and Gfap-Cre mice were cross-bred to obtain astrocyte-specific Atf4 knockout (Atf4fl/fl:Gfap-Cre; hereafter denoted as Atf4Astro cKO) and control (Atf4+/+:Gfap-Cre) mice (Fig. 3A). Combined ISH/IHC confirmed that Atf4 was selectively deleted from GABAergic inhibitory neurons in Atf4In cKO (Fig. 2 B and C) and GFAP-expressing astrocytes in Atf4Astro cKO mice (Fig. 3 B and C).
Fig. 2.
Fig. 3.
Strikingly, deletion of the Atf4 gene from either inhibitory neurons or astrocytes failed to facilitate LTM upon weak MWM (Fig. 2 D and E for inhibitory neurons and Fig. 3 D and E for astrocytes) and CFC paradigms (Figs. 2G and 3G for inhibitory neurons and astrocytes, respectively). Moreover, both genotypes behaved similarly in the standard version of the MWM (SI Appendix, Fig. S3 A and B for inhibitory neurons and SI Appendix, Fig. S4 A and B for astrocytes) or CFC (SI Appendix, Fig. S3C for inhibitory neurons and SI Appendix, Fig. S4C for astrocytes). No visual or anxiety-related behavioral abnormalities were observed upon ATF4 deletion in inhibitory neurons or astrocytes (for inhibitory neurons: Fig. 2F and SI Appendix, Fig. S3 D–H; for astrocytes: Fig. 3F and SI Appendix, Fig. S4 D–H). Thus, deletion of ATF4 from inhibitory neurons or astrocytes fails to facilitate LTM.
Finally, regulation of the ISR in cholinergic neurons is also important for LTM formation (29). To study the effects of selective inhibition of the ISR in cholinergic neurons on contextual LTM formation, we crossed Eif2s1A/A;ftg mice, which allow for cell type–specific mutation of the phosphorylation site at serine 51 in the alpha subunit of eukaryotic translation initiation factor 2 (eIF2) to alanine (S51A) and inhibition of ISR signaling (30, 31), with mice expressing Cre recombinase under the choline acetyltransferase (ChAT) promoter (Eif2s1Chol cKI) (SI Appendix, Fig. S5 A and B). ISR inhibition in cholinergic neurons facilitated LTM formation, as determined by increased freezing 24 h after a weak fear conditioning training (SI Appendix, Fig. S5E). To investigate whether memory enhancement as a consequence of ISR inhibition was mediated via ATF4, we generated mice in which ATF4 is deleted in cholinergic neurons, by crossing Atf4fl/fl mice and ChAT-Cre mice (Atf4Chol cKO) (SI Appendix, Fig. S5 C and D). In contrast to the Eif2s1Chol cKI mice, the Atf4Chol cKO mice performed similarly to control mice in the CFC test (SI Appendix, Fig. S5F). Thus, the enhanced LTM caused by inhibition of the ISR in cholinergic neurons is ATF4-independent. Taken together, our results show that deletion of Atf4 solely in excitatory neurons facilitates LTM formation, which supports the notion that ATF4 represses LTM in a cell type–specific manner.
ATF4 Deletion in Excitatory Neurons Engenders Enhanced Expression of Proteins Involved in Oxidative Phosphorylation.
To gain insight into the molecular mechanism underlying the enhanced LTM in mice lacking ATF4 in excitatory neurons, we conducted RNA sequencing (RNA-Seq) on the hippocampus from control and Atf4Ex cKO mice (SI Appendix, Methods and Materials). A total of 110 genes were differentially expressed (DEGs) in the Atf4Ex cKO compared to the control group, of which 94 were up-regulated and 16 down-regulated (SI Appendix, Fig. S6A and Dataset S1). Importantly, 19% of the DEGs (21 out of the 110) contain documented ATF4 binding sites near the transcription start sites (TSSs) (distance up to ±1,000 bp from TSSs) as determined by the list of targets obtained from publicly available human and mouse ATF4 ChIP-Atlas datasets (32) (SI Appendix, Fig. S6B). ATF4 represses the cAMP response element-binding protein (CREB) (33), a transcription factor involved in LTM formation (34). Indeed, 36 out of the 110 DEGs contain known CREB binding sites (SI Appendix, Fig. S6C), suggesting that ATF4 deletion derepresses CREB-mediated gene expression in excitatory neurons. Accordingly, phosphorylation of CREB was also increased by 2.8-fold in the hippocampal lysates of Atf4Ex cKO mice compared to controls (SI Appendix, Fig. S6D).
Notably, pathway analyses revealed an upregulation of genes encoding proteins that function in oxidative phosphorylation (OXPHOS) (SI Appendix, Fig. S6E). qPCR analysis showed an upregulation of several genes of the OXPHOS pathway (Cox5b, Ndufa13, and Ndufb6) in the hippocampus from Atf4Ex cKO compared to control mice (SI Appendix, Fig. S6F), consistent with previous results (35).
Given the poor correlation between mRNA levels and their encoded proteins in complex biological samples (36, 37), we examined the proteome of the synaptic compartments from control and Atf4Ex cKO mice. Sixty-nine synaptic proteins were differentially expressed in the brain of Atf4Ex cKO mice, of which 33 were up-regulated and 36 down-regulated (Fig. 4 A and B and Dataset S2). Homer3 (but not Homer1 or Homer2), a postsynaptic density-associated protein involved in the formation of fear memory (38) was elevated in synaptosome proteomics Atf4Ex cKO mice, and validated by western blotting (Fig. 4 C and D).
Fig. 4.
Consistent with the RNA-seq analyses, pathway analysis of the synaptic proteome showed an upregulation of proteins involved in OXPHOS in Atf4Ex cKO mice (Fig. 4 E and F). To investigate the link between ATF4 and OXPHOS, primary neurons from WT and Atf4Ex cKO embryos were stained with the cell-permeable tetramethylrhodamine ethyl ester (TMRE) dye to measure the mitochondrial membrane potential (ΔΨm), a marker of mitochondrial activity and ATP synthesis (39). Atf4Ex cKO neurons displayed a twofold increase in TMRE signal compared to the WT neurons (Fig. 4G) at 21 d in vitro (DIV21), a time point when CaMKIIα protein is readily expressed in culture (40) and ATF4 is ~80% depleted in the lysates obtained from primary neurons (Fig. 4H). Together, we demonstrate that Atf4 deletion in excitatory neurons results in enhanced expression of OXPHOS pathway proteins, engendering more ATP production to promote neuronal functions underlying LTM formation.
Discussion
The ISR is a pivotal protein network that controls LTM formation across diverse phyla. Convergent and orthogonal studies demonstrated that genetic inhibition of the ISR pathway by either deleting the ISR kinase general control nonderepressible 2 (GCN2) (41) or protein kinase R (PKR) (6, 42), or by generating an eIF2α heterozygous knock-in (Ser51Ala) (3) or a cell type–specific eIF2α homozygous knock-in (5, 28), boost LTM formation in mice. In addition, inhibiting the ISR kinases: PKR or protein kinase RNA-like endoplasmic reticulum kinase (PERK) pharmacologically or activating eIF2B, which is converted into an inhibitor of the guanosine diphosphate (GDP) to guanosine triphosphate (GTP) exchange upon ISR activation, with an ISR inhibitor (ISRIB), augments LTM in rodents (4, 6, 42, 43). Conversely, genetic or pharmacological activation of ISR impairs LTM formation (3, 8–10). The ISR bidirectionally controls LTM formation in birds (10). The clinical relevance of the ISR is highlighted by a) the identification of mutations in ISR cardinal genes (encoding eIF2γ and constitutive repressor of eIF2α phosphorylation, CReP) that activate the ISR and are associated with intellectual disabilities in humans (44–47) and b) inhibition of the ISR reverses the LTM decline in several cognitive disorders, including Down syndrome (48) and Alzheimer’s disease (49).
A salient feature of the ISR is the enhancement of LTM upon its inhibition in diverse cell types, e.g., excitatory and inhibitory neurons (5) as well as cholinergic neurons (this work) and astrocytes (28). Using a combination of molecular, genetic, behavioral, electrophysiological, and biochemical approaches, we conclusively show that the ISR downstream target ATF4 is an LTM repressor only in excitatory neurons, which comprise the largest percentage (~70 to 80%) of the neuronal cells in the neocortex (50) (key findings are summarized in SI Appendix, Fig. S7). The findings that ATF4 is not a general ISR effector that regulates LTM formation across cell types were unexpected. Thus, ISR downstream targets other than ATF4 may repress LTM in other cell types. For instance, oligophrenin-1 (OPHN1), a protein whose synthesis is up-regulated upon mGluR-mediated ISR activation (31), may function as an ISR effector in dopaminergic neurons to control reward-related learning and memory (51, 52). Future studies should aim to elucidate how ISR inhibition in cholinergic neurons, inhibitory neurons, and astrocytes promotes LTM. To this end, one could either use a gene candidate approach (using known ISR targets) or an unbiased genome-wide approach using cell type–specific ribosome profiling. In the latter case, once the ISR downstream targets are identified in different cell types, functional studies are needed to determine their roles during the formation of LTM.
In addition, it is pertinent to examine how ATF4 regulates memory formation in forebrain excitatory neurons. Consistent with the idea that ATF4 is a CREB repressor (3, 33), we have found an increase in CREB-regulated targets in Atf4Ex cKO vs. control mice (SI Appendix, Fig. S6C). Intriguingly, we unveiled that several components of the OXPHOS pathway are up-regulated in Atf4Ex cKO mice. Previous studies have shown that inhibition of mitochondrial complex I of the OXPHOS pathway impairs LTP in rodent hippocampal slices (53). Dysregulation of the OXPHOS pathway is also associated with several memory-related disorders. For example, protein components of the OXPHOS pathway are down-regulated in Ts65Dn mice (54, 55), a mouse model of Down’s syndrome in which ISR is activated and fear memory is impaired (6). Furthermore, components of OXPHOS such as Cox5b, Ndufa13, and Ndufb6, which exhibit increased expression upon ATF4 deletion in excitatory neurons (SI Appendix, Fig. S6F), are significantly down-regulated in late-onset Alzheimer’s disease (56). Thus, it is conceivable that the ISR–ATF4 axis regulates the energy levels required for excitatory synapses to support LTM formation.
Considering that a) LTM formation depends on the intricate interplay of a diverse array of brain cell types and b) ATF4 does not repress LTM via all cell types, it stands to reason that targeting ATF4 across multiple cell types would yield different outcomes compared to targeting it in a single cell type. Accordingly, a prior study employed lentivirus to express a shRNA targeting Atf4 specifically in the hippocampus, which led to a ~60% reduction in Atf4 levels and an impairment (not enhancement) in spatial LTM (13). Since non-neuronal cells comprise at least half of the total brain cells (57), this effect could be mediated via non-neuronal cells such as astrocytes, microglia, and oligodendrocytes, which all have roles in memory formation (26, 27, 58, 59). Moreover, a major technical limitation of RNA interference is the difficulty in achieving efficient and specific knockdown of the target gene in vivo (14). This is particularly challenging in the brain because of the complex neural circuits and diverse cell types involved in LTM formation. Furthermore, the temporal and spatial specificity required to manipulate gene expression either in specific brain areas or time points adds another layer of complexity. For example, even though the MWM is primarily known as a hippocampus-dependent task, impaired connections with other brain regions impact LTM formation (reviewed in ref. 60). It is plausible that targeting ATF4 in one brain region is insufficient to dissect its role in memory. The targeted molecular genetics approach, which we used, selectively deletes Atf4 in different cell types and mitigates, at least in part, many of the aforementioned limitations. A recent study demonstrated that deletion of Atf4 in excitatory neurons by injecting Cre-expressing viruses into the brain of Atf4fl/fl mice impairs LTP (61); an immediate explanation of the discrepancy is not clear. Our observation that a single train of high-frequency stimulation is sufficient to induce long-lasting LTP only in slices from Atf4Ex cKO mice is consistent with previous studies with eIF2α knock-in mice (S51A mutant) where both heterozygous (3) and excitatory neuron-specific knock-ins (5) exhibited long-lasting LTP following 1 × HFS, whereas the same stimulation in the control mice induced a short-lasting LTP (5). These results indicate that the inhibition of ISR lowers the threshold for eliciting long-lasting LTP.
Methods and Materials
Mice were housed in cages on ventilated racks at a temperature of 20 to 22 °C and a humidity level of approximately 55%, with a standard 12-h light/dark cycle. They had ad libitum access to food (standard rodent chow) and water. At postnatal day 21, mice were separated by sex, weaned, and placed in different cages (2 to 5 mice/cage). All experimental procedures were conducted in compliance with the guidelines established by the animal care committees of McGill University and Baylor College of Medicine. Male mice aged 2- to 5-mo were used for the experiments unless otherwise stated.
Detailed information on experimental procedures is available in SI Appendix, Materials and Methods.
Ethics Approval
The animal care and experimental procedures were conducted in compliance with the regulations of the animal care committees of McGill University and Baylor College of Medicine.
Data, Materials, and Software Availability
The data generated in the study data are included in the article and/or SI Appendix. Raw sequencing reads are deposited to the Sequence Read Archive (SRA) under the accession PRJNA1134721 (62).
Acknowledgments
We thank Annie Sylvestre, Annik Lafrance, Isabelle Harvey, and Eva Migon for technical assistance and animal handling. The research was funded by a grant from the Canadian Institutes of Health Research to N.S. (FND-148423). N.M. is supported by a postdoctoral fellowship from the Charlotte and Leo Karassik Foundation. J-H.C. is supported by a Conrad F. Harrington Fellowship.
Author contributions
N.M., M.C.-M., and N.S. designed research; N.M., J.-H.C., P.Y.W., S.W.D., T.A.W., Z.H., J.L., H.Z., A.Y., and J.S. performed research; C.K. and R.A.M. contributed new reagents/analytic tools; N.M., J.-H.C., P.Y.W., S.W.D., Y.I., V.S., J.-C.L., W.S.S., A.K., and R.A.M. analyzed data; J.-H.C. and S.W.D. assisted in editing the manuscript; V.S., J.-C.L., W.S.S., and A.K. conceptual support; R.A.M. supervised the electrophysiology experiment; N.S. obtained funding; and N.M., M.C.-M., and N.S. wrote the paper.
Competing interests
M.C.-M. and S.W.D. are employees of Altos Labs, Inc. M.C.-M. is a shareholder of Altos Labs, Inc. and Mikrovia, Inc. All other authors do not declare any competing/conflicting interests.
Supporting Information
Appendix 01 (PDF)
- Download
- 1.78 MB
Dataset S01 (XLSX)
- Download
- 17.97 KB
Dataset S02 (XLSX)
- Download
- 13.71 KB
References
1
M. Costa-Mattioli, P. Walter, The integrated stress response: From mechanism to disease. Science 368, eaat5314 (2020).
2
M. Costa-Mattioli, W. S. Sossin, E. Klann, N. Sonenberg, Translational control of long-lasting synaptic plasticity and memory. Neuron 61, 10–26 (2009).
3
M. Costa-Mattioli et al., eIF2α phosphorylation bidirectionally regulates the switch from short-to long-term synaptic plasticity and memory. Cell 129, 195–206 (2007).
4
C. Sidrauski et al., Pharmacological brake-release of mRNA translation enhances cognitive memory. eLife 2, e00498 (2013).
5
V. Sharma et al., eIF2α controls memory consolidation via excitatory and somatostatin neurons. Nature 586, 412–416 (2020).
6
P. J. Zhu et al., Suppression of PKR promotes network excitability and enhanced cognition by interferon-γ-mediated disinhibition. Cell 147, 1384–1396 (2011).
7
K. M. Vattem, R. C. Wek, Reinitiation involving upstream ORFs regulates ATF4 mRNA translation in mammalian cells. Proc. Natl. Acad. Sci. U.S.A. 101, 11269–11274 (2004).
8
Z. Jiang et al., eIF2α phosphorylation-dependent translation in CA1 pyramidal cells impairs hippocampal memory consolidation without affecting general translation. J. Neurosci. 30, 2582–2594 (2010).
9
M. Jian et al., eIF2α dephosphorylation in basolateral amygdala mediates reconsolidation of drug memory. J. Neurosci. 34, 10010–10021 (2014).
10
G. Batista, J. L. Johnson, E. Dominguez, M. Costa-Mattioli, J. L. Pena, Translational control of auditory imprinting and structural plasticity by eIF2α. eLife 5, e17197 (2016).
11
K. Ameri, A. L. Harris, Activating transcription factor 4. Int. J. Biochem. Cell Biol. 40, 14–21 (2008).
12
A. Chen et al., Inducible enhancement of memory storage and synaptic plasticity in transgenic mice expressing an inhibitor of ATF4 (CREB-2) and C/EBP proteins. Neuron 39, 655–669 (2003).
13
S. Pasini, C. Corona, J. Liu, L. A. Greene, M. L. Shelanski, Specific downregulation of hippocampal ATF4 reveals a necessary role in synaptic plasticity and memory. Cell Rep. 11, 183–191 (2015).
14
K. Goel, J. E. Ploski, RISC-y business: Limitations of short hairpin RNA-mediated gene silencing in the brain and a discussion of CRISPR/Cas-based alternatives. Front. Mol. Neurosci. 15, 914430 (2022).
15
J. N. Martin et al., Lethal toxicity caused by expression of shRNA in the mouse striatum: Implications for therapeutic design. Gene Ther. 18, 666–673 (2011).
16
A. M. Zimmer, Y. K. Pan, T. Chandrapalan, R. W. M. Kwong, S. F. Perry, Loss-of-function approaches in comparative physiology: Is there a future for knockdown experiments in the era of genome editing? J. Exp. Biol. 222, jeb175737 (2019).
17
H. C. Masuoka, T. M. Townes, Targeted disruption of the activating transcription factor 4 gene results in severe fetal anemia in mice. Blood 99, 736–745 (2002).
18
U. Müller, Ten years of gene targeting: Targeted mouse mutants, from vector design to phenotype analysis. Mech. Dev. 82, 3–21 (1999).
19
R. Liu et al., NOX activation in reactive astrocytes regulates astrocytic LCN2 expression and neurodegeneration. Cell Death Dis. 13, 371 (2022).
20
Y. Wang et al., Mild endoplasmic reticulum stress protects against lipopolysaccharide-induced astrocytic activation and blood–brain barrier hyperpermeability. Front. Cell Neurosci. 12, 222 (2018).
21
M. B. Kennedy, Synaptic signaling in learning and memory. Cold Spring Harb. Perspect. Biol. 8, a016824 (2013).
22
J. Z. Tsien et al., Subregion-and cell type–restricted gene knockout in mouse brain. Cell 87, 1317–1326 (1996).
23
C. V. Vorhees, M. T. Williams, Morris water maze: Procedures for assessing spatial and related forms of learning and memory. Nat. Protoc. 1, 848–858 (2006).
24
T. V. Bliss, G. L. Collingridge, A synaptic model of memory: Long-term potentiation in the hippocampus. Nature 361, 31–39 (1993).
25
L. Topolnik, S. Tamboli, The role of inhibitory circuits in hippocampal memory processing. Nat. Rev. Neurosci. 23, 476–492 (2022).
26
A. Kol et al., Astrocytes contribute to remote memory formation by modulating hippocampal–cortical communication during learning. Nat. Neurosci. 23, 1229–1239 (2020).
27
A. Adamsky et al., Astrocytic activation generates de novo neuronal potentiation and memory enhancement. Cell 174, 59–71.e14 (2018).
28
V. Sharma et al., mRNA translation in astrocytes controls hippocampal long-term synaptic plasticity and memory. Proc. Natl. Acad. Sci. U.S.A. 120, e2308671120 (2023).
29
A. R. Helseth et al., Cholinergic neurons constitutively engage the ISR for dopamine modulation and skill learning in mice. Science 372, eabe1931 (2021).
30
S. H. Back et al., Translation attenuation through eIF2alpha phosphorylation prevents oxidative stress and maintains the differentiated state in beta cells. Cell Metab. 10, 13–26 (2009).
31
G. V. Di Prisco et al., Translational control of mGluR-dependent long-term depression and object-place learning by eIF2α. Nat. Neurosci. 17, 1073–1082 (2014).
32
Z. Zou, T. Ohta, F. Miura, S. Oki, ChIP-Atlas 2021 update: A data-mining suite for exploring epigenomic landscapes by fully integrating ChIP-seq, ATAC-seq and Bisulfite-seq data. Nucleic Acids Res. 50, W175–W182 (2022).
33
D. Bartsch et al., Aplysia CREB2 represses long-term facilitation: Relief of repression converts transient facilitation into long-term functional and structural change. Cell 83, 979–992 (1995).
34
S. Kida, A functional role for CREB as a positive regulator of memory formation and LTP. Exp. Neurobiol. 21, 136–140 (2012).
35
P. M. Quirós et al., Multi-omics analysis identifies ATF4 as a key regulator of the mitochondrial stress response in mammals. J. Cell Biol. 216, 2027–2045 (2017).
36
T. Maier, M. Güell, L. Serrano, Correlation of mRNA and protein in complex biological samples. FEBS Lett. 583, 3966–3973 (2009).
37
R. de Sousa Abreu, L. O. Penalva, E. M. Marcotte, C. Vogel, Global signatures of protein and mRNA expression levels. Mol. BioSyst. 5, 1512–1526 (2009).
38
V. Rizzo et al., Encoding of contextual fear memory requires de novo proteins in the prelimbic cortex. Biol. Psychiatry: Cogn. Neurosci. Neuroimaging 2, 158–169 (2017).
39
L. D. Zorova et al., Mitochondrial membrane potential. Anal. Biochem. 552, 50–59 (2018).
40
S. Cornelia Koeberle et al., Developmental stage-dependent regulation of spine formation by calcium-calmodulin-dependent protein kinase IIα and Rap1. Sci. Rep. 7, 13409 (2017).
41
M. Costa-Mattioli et al., Translational control of hippocampal synaptic plasticity and memory by the eIF2α kinase GCN2. Nature 436, 1166–1170 (2005).
42
E. Stern, A. Chinnakkaruppan, O. David, N. Sonenberg, K. Rosenblum, Blocking the eIF2α kinase (PKR) enhances positive and negative forms of cortex-dependent taste memory. J. Neurosci. 33, 2517–2525 (2013).
43
V. Sharma et al., Local inhibition of PERK enhances memory and reverses age-related deterioration of cognitive and neuronal properties. J. Neurosci. 38, 648–658 (2018).
44
B. Abdulkarim et al., A missense mutation in PPP1R15B causes a syndrome including diabetes, short stature, and microcephaly. Diabetes 64, 3951–3962 (2015).
45
G. Borck et al., eIF2γ mutation that disrupts eIF2 complex integrity links intellectual disability to impaired translation initiation. Mol. Cell 48, 641–646 (2012).
46
M. Skopkova et al., EIF2S3 mutations associated with severe X-linked intellectual disability syndrome MEHMO. Hum. Mutat. 38, 409–425 (2017).
47
S. Moortgat et al., Two novel EIF2S3 mutations associated with syndromic intellectual disability with severe microcephaly, growth retardation, and epilepsy. Am. J. Med. Genet., Part A 170, 2927–2933 (2016).
48
P. J. Zhu et al., Activation of the ISR mediates the behavioral and neurophysiological abnormalities in Down syndrome. Science 366, 843–849 (2019).
49
M. M. Oliveira et al., Correction of eIF2-dependent defects in brain protein synthesis, synaptic plasticity, and memory in mouse models of Alzheimer’s disease. Sci. Signaling 14, eabc5429 (2021).
50
X. Tan, S. H. Shi, Neocortical neurogenesis and neuronal migration. Wiley Interdiscip. Rev. Dev. Biol. 2, 443–459 (2013).
51
W. Huang et al., Translational control by eIF2α phosphorylation regulates vulnerability to the synaptic and behavioral effects of cocaine. eLife 5, e12052 (2016).
52
A. N. Placzek et al., eIF2α-mediated translational control regulates the persistence of cocaine-induced LTP in midbrain dopamine neurons. eLife 5, e17517 (2016).
53
R. Kimura et al., Acute exposure to the mitochondrial complex I toxin rotenone impairs synaptic long-term potentiation in rat hippocampal slices. CNS Neurosci. Ther. 18, 641–646 (2012).
54
M. J. Alldred, S. H. Lee, G. E. Stutzmann, S. D. Ginsberg, Oxidative phosphorylation is dysregulated within the basocortical circuit in a 6-month old mouse model of Down syndrome and Alzheimer’s disease. Front. Aging Neurosci. 13, 707950 (2021).
55
M. J. Alldred et al., Profiling basal forebrain cholinergic neurons reveals a molecular basis for vulnerability within the Ts65Dn model of Down syndrome and Alzheimer’s disease. Mol. Neurobiol. 58, 5141–5162 (2021).
56
S. S. Adav, J. E. Park, S. K. Sze, Quantitative profiling brain proteomes revealed mitochondrial dysfunction in Alzheimer’s disease. Mol. Brain 12, 8 (2019).
57
D. Keller, C. Erö, H. Markram, Cell densities in the mouse brain: A systematic review. Front. Neuroanat. 12, 83 (2018).
58
V. Stratoulias et al., ARG1-expressing microglia show a distinct molecular signature and modulate postnatal development and function of the mouse brain. Nat. Neurosci. 26, 1008–1020 (2023).
59
M. Munyeshyaka, R. D. Fields, Oligodendroglia are emerging players in several forms of learning and memory. Commun. Biol. 5, 1148 (2022).
60
R. D’Hooge, P. P. De Deyn, Applications of the Morris water maze in the study of learning and memory. Brain Res. Rev. 36, 60–90 (2001).
61
A. Kumar et al., 2-Deoxyglucose drives plasticity via an adaptive ER stress-ATF4 pathway and elicits stroke recovery and Alzheimer’s resilience. Neuron 111, 2831–2846.e2810 (2023).
62
K. Katz et al., The Sequence Read Archive: A decade more of explosive growth. Nucleic Acids Research 50, D387–D390 (2022).
Information & Authors
Information
Published in
Classifications
Copyright
Copyright © 2024 the Author(s). Published by PNAS. This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND).
Data, Materials, and Software Availability
The data generated in the study data are included in the article and/or SI Appendix. Raw sequencing reads are deposited to the Sequence Read Archive (SRA) under the accession PRJNA1134721 (62).
Submission history
Received: April 14, 2024
Accepted: June 25, 2024
Published online: July 24, 2024
Published in issue: July 30, 2024
Keywords
Acknowledgments
We thank Annie Sylvestre, Annik Lafrance, Isabelle Harvey, and Eva Migon for technical assistance and animal handling. The research was funded by a grant from the Canadian Institutes of Health Research to N.S. (FND-148423). N.M. is supported by a postdoctoral fellowship from the Charlotte and Leo Karassik Foundation. J-H.C. is supported by a Conrad F. Harrington Fellowship.
Author contributions
N.M., M.C.-M., and N.S. designed research; N.M., J.-H.C., P.Y.W., S.W.D., T.A.W., Z.H., J.L., H.Z., A.Y., and J.S. performed research; C.K. and R.A.M. contributed new reagents/analytic tools; N.M., J.-H.C., P.Y.W., S.W.D., Y.I., V.S., J.-C.L., W.S.S., A.K., and R.A.M. analyzed data; J.-H.C. and S.W.D. assisted in editing the manuscript; V.S., J.-C.L., W.S.S., and A.K. conceptual support; R.A.M. supervised the electrophysiology experiment; N.S. obtained funding; and N.M., M.C.-M., and N.S. wrote the paper.
Competing interests
M.C.-M. and S.W.D. are employees of Altos Labs, Inc. M.C.-M. is a shareholder of Altos Labs, Inc. and Mikrovia, Inc. All other authors do not declare any competing/conflicting interests.
Notes
Reviewers: C.B., Universitetet i Bergen; and S.K., University of Tokyo.
Authors
Metrics & Citations
Metrics
Altmetrics
Citations
Cite this article
The ISR downstream target ATF4 represses long-term memory in a cell type–specific manner, Proc. Natl. Acad. Sci. U.S.A.
121 (31) e2407472121,
https://doi.org/10.1073/pnas.2407472121
(2024).
Copied!
Copying failed.
Export the article citation data by selecting a format from the list below and clicking Export.
Cited by
Loading...
View Options
View options
PDF format
Download this article as a PDF file
DOWNLOAD PDFLogin options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginRecommend to a librarian
Recommend PNAS to a LibrarianPurchase options
Purchase this article to access the full text.