Mouse models of SYNGAP1-related intellectual disability.

SYNGAP1 is a Ras-GTPase-activating protein highly enriched at excitatory synapses in the brain. De novo loss-of-function mutations in SYNGAP1 are a major cause of genetically defined neurodevelopmental disorders (NDDs). These mutations are highly penetrant and cause SYNGAP1-related intellectual disability (SRID), an NDD characterized by cognitive impairment, social deficits, early-onset seizures, and sleep disturbances. Studies in rodent neurons have shown that Syngap1 regulates developing excitatory synapse structure and function, and heterozygous Syngap1 knockout mice have deficits in synaptic plasticity, learning, and memory and have seizures. However, how specific SYNGAP1 mutations found in humans lead to disease has not been investigated in vivo. To explore this, we utilized the CRISPR-Cas9 system to generate knock-in mouse models with two distinct known causal variants of SRID: one with a frameshift mutation leading to a premature stop codon, SYNGAP1; L813RfsX22, and a second with a single-nucleotide mutation in an intron that creates a cryptic splice acceptor site leading to premature stop codon, SYNGAP1; c.3583-9G>A. While reduction in Syngap1 mRNA varies from 30 to 50% depending on the specific mutation, both models show ~50% reduction in Syngap1 protein, have deficits in synaptic plasticity, and recapitulate key features of SRID including hyperactivity and impaired working memory. These data suggest that half the amount of SYNGAP1 protein is key to the pathogenesis of SRID. These results provide a resource to study SRID and establish a framework for the development of therapeutic strategies for this disorder.

All-or-none disconnection of pyramidal inputs onto parvalbumin-positive interneurons gates ocular dominance plasticity.

Disinhibition is an obligatory initial step in the remodeling of cortical circuits by sensory experience. Our investigation on disinhibitory mechanisms in the classical model of ocular dominance plasticity uncovered an unexpected form of experience-dependent circuit plasticity. In the layer 2/3 of mouse visual cortex, monocular deprivation triggers a complete, "all-or-none," elimination of connections from pyramidal cells onto nearby parvalbumin-positive interneurons (Pyr→PV). This binary form of circuit plasticity is unique, as it is transient, local, and discrete. It lasts only 1 d, and it does not manifest as widespread changes in synaptic strength; rather, only about half of local connections are lost, and the remaining ones are not affected in strength. Mechanistically, the deprivation-induced loss of Pyr→PV is contingent on a reduction of the protein neuropentraxin2. Functionally, the loss of Pyr→PV is absolutely necessary for ocular dominance plasticity, a canonical model of deprivation-induced model of cortical remodeling. We surmise, therefore, that this all-or-none loss of local Pyr→PV circuitry gates experience-dependent cortical plasticity.

An ultrasensitive biosensor for high-resolution kinase activity imaging in awake mice.

Protein kinases control nearly every facet of cellular function. These key signaling nodes integrate diverse pathway inputs to regulate complex physiological processes, and aberrant kinase signaling is linked to numerous pathologies. While fluorescent protein-based biosensors have revolutionized the study of kinase signaling by allowing direct, spatiotemporally precise kinase activity measurements in living cells, powerful new molecular tools capable of robustly tracking kinase activity dynamics across diverse experimental contexts are needed to fully dissect the role of kinase signaling in physiology and disease. Here, we report the development of an ultrasensitive, second-generation excitation-ratiometric protein kinase A (PKA) activity reporter (ExRai-AKAR2), obtained via high-throughput linker library screening, that enables sensitive and rapid monitoring of live-cell PKA activity across multiple fluorescence detection modalities, including plate reading, cell sorting and one- or two-photon imaging. Notably, in vivo visual cortex imaging in awake mice reveals highly dynamic neuronal PKA activity rapidly recruited by forced locomotion.

The C9orf72 repeat expansion disrupts nucleocytoplasmic transport.

The hexanucleotide repeat expansion (HRE) GGGGCC (G4C2) in C9orf72 is the most common cause of amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). Recent studies support an HRE RNA gain-of-function mechanism of neurotoxicity, and we previously identified protein interactors for the G4C2 RNA including RanGAP1. A candidate-based genetic screen in Drosophila expressing 30 G4C2 repeats identified RanGAP (Drosophila orthologue of human RanGAP1), a key regulator of nucleocytoplasmic transport, as a potent suppressor of neurodegeneration. Enhancing nuclear import or suppressing nuclear export of proteins also suppresses neurodegeneration. RanGAP physically interacts with HRE RNA and is mislocalized in HRE-expressing flies, neurons from C9orf72 ALS patient-derived induced pluripotent stem cells (iPSC-derived neurons), and in C9orf72 ALS patient brain tissue. Nuclear import is impaired as a result of HRE expression in the fly model and in C9orf72 iPSC-derived neurons, and these deficits are rescued by small molecules and antisense oligonucleotides targeting the HRE G-quadruplexes. Nucleocytoplasmic transport defects may be a fundamental pathway for ALS and FTD that is amenable to pharmacotherapeutic intervention.

AMPA receptor exchange underlies transient memory destabilization on retrieval.

A consolidated memory can be transiently destabilized by memory retrieval, after which memories are reconsolidated within a few hours; however, the molecular substrates underlying this destabilization process remain essentially unknown. Here we show that at lateral amygdala synapses, fear memory consolidation correlates with increased surface expression of calcium-impermeable AMPA receptors (CI-AMPARs), which are known to be more stable at the synapse, whereas memory retrieval induces an abrupt exchange of CI-AMPARs to calcium-permeable AMPARs (CP-AMPARs), which are known to be less stable at the synapse. We found that blockade of either CI-AMPAR endocytosis or NMDA receptor activity during memory retrieval, both of which blocked the exchange to CP-AMPARs, prevented memory destabilization, indicating that this transient exchange of AMPARs may underlie the transformation of a stable memory into an unstable memory. These newly inserted CP-AMPARs gradually exchanged back to CI-AMPARs within hours, which coincided with the course of reconsolidation. Furthermore, blocking the activity of these newly inserted CP-AMPARs after retrieval impaired reconsolidation, suggesting that they serve as synaptic "tags" that support synapse-specific reconsolidation. Taken together, our results reveal unexpected physiological roles of CI-AMPARs and CP-AMPARs in transforming a consolidated memory into an unstable memory and subsequently guiding reconsolidation.

Large-scale analysis of posttranslational modifications in the hippocampus of patients with Alzheimer's disease using pI shift and label-free quantification without enrichment.

Posttranslational modifications modulate protein function in cells. Global analysis of multiple posttranslational modifications can provide insight into physiology and disease, but presents formidable challenges. In the present study, we used a technique that does not require target enrichment to analyze alterations in the phosphorylation and ubiquitination of proteins from patients with Alzheimer's disease (AD). Guided by our previous findings, we applied three strategies to further our understanding of the dysregulation of posttranslationally modified proteins. We first identified phosphorylation sites by determining peptide pI shifts using OFFGEL. Second, using tandem mass spectrometry, we determined the ubiquitination status of the proteins using an assay for a trypsin digestion remnant of ubiquitination (Gly-Gly). Third, for large-scale discovery, we quantified the global differences in protein expression. Of the proteins expressed in AD tissue at levels of 2.0 or greater compared with controls, 60 were phosphorylated and 56 were ubiquitinated. Of the proteins expressed at levels of 0.5 or lower compared with controls, 81 were phosphorylated and 56 were ubiquitinated. Approximately 98 % of the phosphopeptides exhibited a pI shift. We identified 112 new phosphorylation sites (51.38 %), and 92 new ubiquitination sites (96.84 %). Taken together, our findings suggest that analysis of the alterations in posttranslationally modified proteins may contribute to understanding the pathogenesis of AD and other diseases.

Stimulus-dependent synaptic plasticity underlies neuronal circuitry refinement in the mouse primary visual cortex.

Perceptual learning improves our ability to interpret sensory stimuli present in our environment through experience. Despite its importance, the underlying mechanisms that enable perceptual learning in our sensory cortices are still not fully understood. In this study, we used in vivo two-photon imaging to investigate the functional and structural changes induced by visual stimulation in the mouse primary visual cortex (V1). Our results demonstrate that repeated stimulation leads to a refinement of V1 circuitry by decreasing the number of responsive neurons while potentiating their response. At the synaptic level, we observe a reduction in the number of dendritic spines and an overall increase in spine AMPA receptor levels in the same subset of neurons. In addition, visual stimulation induces synaptic potentiation in neighboring spines within individual dendrites. These findings provide insights into the mechanisms of synaptic plasticity underlying information processing in the neocortex.

SynGAP regulates synaptic plasticity and cognition independently of its catalytic activity.

SynGAP is an abundant synaptic GTPase-activating protein (GAP) critical for synaptic plasticity, learning, memory, and cognition. Mutations in SYNGAP1 in humans result in intellectual disability, autistic-like behaviors, and epilepsy. Heterozygous Syngap1-knockout mice display deficits in synaptic plasticity, learning, and memory and exhibit seizures. It is unclear whether SynGAP imparts structural properties at synapses independently of its GAP activity. Here, we report that inactivating mutations within the GAP domain do not inhibit synaptic plasticity or cause behavioral deficits. Instead, SynGAP modulates synaptic strength by physically competing with the AMPA-receptor-TARP excitatory receptor complex in the formation of molecular condensates with synaptic scaffolding proteins. These results have major implications for developing therapeutic treatments for SYNGAP1-related neurodevelopmental disorders.

Long-term neural correlates of reversible fear learning in the lateral amygdala.

Fear conditioning and extinction are behavioral models that reflect the association and dissociation of environmental cues to aversive outcomes, both known to involve the lateral amygdala (LA). Accordingly, responses of LA neurons to conditioned stimuli (CS) increase after fear conditioning and decrease partially during extinction. However, the long-term effects of repeated fear conditioning and extinction on LA neuronal firing have not been explored. Here we show, using stable, high signal-to-noise ratio single-unit recordings, that the ensemble activity of all recorded LA neurons correlates tightly with conditioned fear responses of rats in a conditioning/extinction/reconditioning paradigm spanning 3 d. This CS-evoked ensemble activity increased after conditioning, decreased after extinction, and was repotentiated after reconditioning. Cell-by-cell analysis revealed that among the LA neurons that displayed potentiated responses after initial fear conditioning, some exhibited weakened CS responses after extinction (extinction-susceptible), whereas others remained potentiated (extinction-resistant). The majority of extinction-susceptible neurons exhibited strong potentiation after reconditioning, suggesting that this distinct subpopulation (reversible fear neurons) encodes updated CS-unconditioned stimulus (US) association strength. Interestingly, these reversible fear neurons displayed larger, more rapid potentiation during reconditioning compared with the initial conditioning, providing a neural correlate of savings after extinction. In contrast, the extinction-resistant fear neurons did not show further increases after reconditioning, suggesting that this subpopulation encodes persistent fear memory representing the original CS-US association. This longitudinal report on LA neuronal activity during reversible fear learning suggests the existence of distinct populations encoding various facets of fear memory and provides insight into the neuronal mechanisms of fear memory modulation.

Quantitative proteomics of auditory fear conditioning.

Auditory fear conditioning is a well-characterized rodent learning model where a neutral auditory cue is paired with an aversive outcome to induce associative fear memory. The storage of long-term auditory fear memory requires long-term potentiation (LTP) in the lateral amygdala and de novo protein synthesis. Although many studies focused on individual proteins have shown their contribution to LTP and fear conditioning, non-biased genome-wide studies have only recently been possible with microarrays, which nevertheless fall short of measuring changes at the level of proteins. Here we employed quantitative proteomics to examine the expression of hundreds of proteins in the lateral amygdala in response to auditory fear conditioning. We found that various proteins previously implicated in LTP, learning and axon/dendrite growth were regulated by fear conditioning. A substantial number of proteins that were regulated by fear conditioning have not yet been studied specifically in learning or synaptic plasticity.

Mouse models of SYNGAP1 -related intellectual disability.

SYNGAP1 is a Ras-GTPase activating protein highly enriched at excitatory synapses in the brain. De novo loss-of-function mutations in SYNGAP1 are a major cause of genetically defined neurodevelopmental disorders (NDD). These mutations are highly penetrant and cause SYNGAP1 -related intellectual disability (SRID), a NDD characterized by cognitive impairment, social deficits, early-onset seizures, and sleep disturbances (1-5). Studies in rodent neurons have shown that Syngap1 regulates developing excitatory synapse structure and function (6-11), and heterozygous Syngap1 knockout mice have deficits in synaptic plasticity, learning and memory, and have seizures (9, 12-14). However, how specific SYNGAP1 mutations found in humans lead to disease has not been investigated in vivo. To explore this, we utilized the CRISPR-Cas9 system to generate knock-in mouse models with two distinct known causal variants of SRID: one with a frameshift mutation leading to a premature stop codon, SYNGAP1; L813RfsX22, and a second with a single-nucleotide mutation in an intron that creates a cryptic splice acceptor site leading to premature stop codon, SYNGAP1; c.3583-9G>A . While reduction in Syngap1 mRNA varies from 30-50% depending on the specific mutation, both models show ∼50% reduction in Syngap1 protein, have deficits in synaptic plasticity, and recapitulate key features of SRID including hyperactivity and impaired working memory. These data suggest that half the amount of SYNGAP1 protein is key to the pathogenesis of SRID. These results provide a resource to study SRID and establish a framework for the development of therapeutic strategies for this disorder. Significance Statement: SYNGAP1 is a protein enriched at excitatory synapses in the brain that is an important regulator of synapse structure and function. SYNGAP1 mutations cause SYNGAP1 -related intellectual disability (SRID), a neurodevelopmental disorder with cognitive impairment, social deficits, seizures, and sleep disturbances. To explore how SYNGAP1 mutations found in humans lead to disease, we generated the first knock-in mouse models with causal SRID variants: one with a frameshift mutation and a second with an intronic mutation that creates a cryptic splice acceptor site. Both models show decreased Syngap1 mRNA and Syngap1 protein and recapitulate key features of SRID including hyperactivity and impaired working memory. These results provide a resource to study SRID and establish a framework for the development of therapeutic strategies. Highlights: Two mouse models with SYNGAP1 -related intellectual disability (SRID) mutations found in humans were generated: one with a frameshift mutation that results in a premature stop codon and the other with an intronic mutation resulting in a cryptic splice acceptor site and premature stop codon. Both SRID mouse models show 35∼50% reduction in mRNA and ∼50% reduction in Syngap1 protein.Both SRID mouse models display deficits in synaptic plasticity and behavioral phenotypes found in people. RNA-seq confirmed cryptic splice acceptor activity in one SRID mouse model and revealed broad transcriptional changes also identified in Syngap1 +/- mice. Novel SRID mouse models generated here provide a resource and establish a framework for development of future therapeutic intervention.

Reactivation of fear memory renders consolidated amygdala synapses labile.

It is believed that memory reactivation transiently renders consolidated memory labile and that this labile or deconsolidated memory is reconsolidated in a protein synthesis-dependent manner. The synaptic correlate of memory deconsolidation upon reactivation, however, has not been fully characterized. Here, we show that 3,5-dihydroxyphenylglycine (DHPG), an agonist for group I metabotropic glutamate receptors (mGluRI), induces synaptic depotentiation only at thalamic input synapses onto the lateral amygdala (T-LA synapses) where synaptic potentiation is consolidated, but not at synapses where synaptic potentiation is not consolidated. Using this mGluRI-induced synaptic depotentiation (mGluRI-depotentiation) as a marker of consolidated synapses, we found that mGluRI-depotentiation correlated well with the state of memory deconsolidation and reconsolidation in a predictable manner. DHPG failed to induce mGluRI-depotentiation in slices prepared immediately after reactivation when the reactivated memory was deconsolidated. DHPG induced mGluRI-depotentiation 1 h after reactivation when the reactivated memory was reconsolidated, but it failed to do so when reconsolidation was blocked by a protein synthesis inhibitor. To test the memory-specificity of mGluRI-depotentiation, conditioned fear was acquired twice using two discriminative tones (2.8 and 20 kHz). Under this condition, mGluRI-depotentiation was fully impaired in slices prepared immediately after reactivation with both tones, whereas mGluRI-depotentiation was partially impaired immediately after reactivation with the 20 kHz tone. Consistently, microinjection of DHPG into the LA 1 h after reactivation reduced fear memory retention, whereas DHPG injection immediately after reactivation failed to do so. Our findings suggest that, upon memory reactivation, consolidated T-LA synapses enter a temporary labile state, displaying insensitivity to mGluRI-depotentiation.

Modulation of fear memory by retrieval and extinction: a clue for memory deconsolidation.

Memories are fragile and easily forgotten at first, but after a consolidation period of hours to weeks, are inscribed in our brains as stable traces, no longer vulnerable to conventional amnesic treatments. Retrieval of a memory renders it labile, akin to the early stages of consolidation. This phenomenon has been explored as memory reactivation, in the sense that the memory is temporarily 'deconsolidated', allowing a short time window for amnesic intervention. This window closes again after reconsolidation, which restores the stability of the memory. In contrast to this 'transient deconsolidation' and the short-spanned amnesic effects of consolidation blockers, some specific treatments can disrupt even consolidated memory, leading to apparent amnesia. We propose the term 'amnesic deconsolidation' to describe such processes that lead to disruption of consolidated memory and/or consolidated memory traces. We review studies of these 'amnesic deconsolidation' treatments that enhance memory extinction, alleviate relapse, and reverse learning-induced plasticity. The transient deconsolidation that memory retrieval induces and the amnesic deconsolidation that these regimes induce both seem to dislodge a component that stabilizes consolidated memory. Characterizing this component, at both molecular and network levels, will provide a key to developing clinical treatments for memory-related disorders and to defining the consolidated memory trace.

Visualizing synaptic plasticity in vivo by large-scale imaging of endogenous AMPA receptors.

Elucidating how synaptic molecules such as AMPA receptors mediate neuronal communication and tracking their dynamic expression during behavior is crucial to understand cognition and disease, but current technological barriers preclude large-scale exploration of molecular dynamics in vivo. We have developed a suite of innovative methodologies that break through these barriers: a new knockin mouse line with fluorescently tagged endogenous AMPA receptors, two-photon imaging of hundreds of thousands of labeled synapses in behaving mice, and computer vision-based automatic synapse detection. Using these tools, we can longitudinally track how the strength of populations of synapses changes during behavior. We used this approach to generate an unprecedentedly detailed spatiotemporal map of synapses undergoing changes in strength following sensory experience. More generally, these tools can be used as an optical probe capable of measuring functional synapse strength across entire brain areas during any behavioral paradigm, describing complex system-wide changes with molecular precision.

Amygdala depotentiation and fear extinction.

Auditory fear memory is thought to be maintained by fear conditioning-induced potentiation of synaptic efficacy, which involves enhanced expression of surface AMPA receptor (AMPAR) at excitatory synapses in the lateral amygdala (LA). Depotentiation, reversal of conditioning-induced potentiation, has been proposed as a cellular mechanism for fear extinction; however, a direct link between depotentiation and extinction has not yet been tested. To address this issue, we applied both ex vivo and in vivo approaches to rats in which fear memory had been consolidated. A unique form of depotentiation reversed conditioning-induced potentiation at thalamic input synapses onto the LA (T-LA synapses) ex vivo. Extinction returned the enhanced T-LA synaptic efficacy observed in conditioned rats to baseline and occluded the depotentiation. Consistently, extinction reversed conditioning-induced enhancement of surface expression of AMPAR subunits in LA synaptosomal preparations. A GluR2-derived peptide that blocks regulated AMPAR endocytosis inhibited depotentiation, and microinjection of a cell-permeable form of the peptide into the LA attenuated extinction. Our results are consistent with the use of depotentiation to weaken potentiated synaptic inputs onto the LA during extinction and provide strong evidence that AMPAR removal at excitatory synapses in the LA underlies extinction.

GSK-3ß activation is required for ZIP-induced disruption of learned fear.

The myristoylated zeta inhibitory peptide (ZIP), which was originally developed as a protein kinase C/Mζ (PKCζ/PKMζ) inhibitor, is known to produce the loss of different forms of memories. However, ZIP induces memory loss even in the absence of PKMζ, and its mechanism of action, therefore, remains elusive. Here, through a kinome-wide screen, we found that glycogen synthase kinase 3 beta (GSK-3ß) was robustly activated by ZIP in vitro. ZIP induced depotentiation (a cellular substrate of memory erasure) of conditioning-induced potentiation at LA synapses, and the ZIP-induced depotentiation was prevented by a GSK-3ß inhibitor, 6-bromoindirubin-3-acetoxime (BIO-acetoxime). Consistently, GSK-3ß inhibition by BIO-acetoxime infusion or GSK-3ß knockdown by GSK-3ß shRNA in the LA attenuated ZIP-induced disruption of learned fear. Furthermore, conditioned fear was decreased by expression of a non-inhibitable form of GSK-3ß in the LA. Our findings suggest that GSK-3ß activation is a critical step for ZIP-induced disruption of memory.

SynGAP isoforms differentially regulate synaptic plasticity and dendritic development.

SynGAP is a synaptic Ras GTPase-activating protein (GAP) with four C-terminal splice variants: α1, α2, ß, and γ. Although studies have implicated SYNGAP1 in several cognitive disorders, it is not clear which SynGAP isoforms contribute to disease. Here, we demonstrate that SynGAP isoforms exhibit unique spatiotemporal expression patterns and play distinct roles in neuronal and synaptic development in mouse neurons. SynGAP-α1, which undergoes liquid-liquid phase separation with PSD-95, is highly enriched in synapses and is required for LTP. In contrast, SynGAP-ß, which does not bind PSD-95 PDZ domains, is less synaptically targeted and promotes dendritic arborization. A mutation in SynGAP-α1 that disrupts phase separation and synaptic targeting abolishes its ability to regulate plasticity and instead causes it to drive dendritic development like SynGAP-ß. These results demonstrate that distinct intrinsic biochemical properties of SynGAP isoforms determine their function, and individual isoforms may differentially contribute to the pathogenesis of SYNGAP1-related cognitive disorders.

Cortical Synaptic AMPA Receptor Plasticity during Motor Learning.

Modulation of synaptic strength through trafficking of AMPA receptors (AMPARs) is a fundamental mechanism underlying synaptic plasticity, learning, and memory. However, the dynamics of AMPAR trafficking in vivo and its correlation with learning have not been resolved. Here, we used in vivo two-photon microscopy to visualize surface AMPARs in mouse cortex during the acquisition of a forelimb reaching task. Daily training leads to an increase in AMPAR levels at a subset of spatially clustered dendritic spines in the motor cortex. Surprisingly, we also observed increases in spine AMPAR levels in the visual cortex. There, synaptic potentiation depends on the availability of visual input during motor training, and optogenetic inhibition of visual cortex activity impairs task performance. These results indicate that motor learning induces widespread cortical synaptic potentiation by increasing the net trafficking of AMPARs into spines, including in non-motor brain regions.

Extinction of cued fear memory involves a distinct form of depotentiation at cortical input synapses onto the lateral amygdala.

The amygdala is known to be a critical storage site of conditioned fear memory. Among the two major pathways to the lateral amygdala (LA), the cortical pathway is known to display a presynaptic long-term potentiation which is occluded with fear conditioning. Here we show that fear extinction results in a net depression of conditioning-induced potentiation at cortical input synapses onto the LA (C-LA synapses). Fear conditioning induced a significant potentiation of excitatory postsynaptic currents at C-LA synapses compared with naïve and unpaired controls, whereas extinction apparently reversed this potentiation. Paired-pulse low-frequency stimulation (pp-LFS) induced synaptic depression in the C-LA pathway of fear-conditioned rats, but not in naïve or unpaired controls, indicating that the pp-LFS-induced depression is specific to associative learning-induced changes (pp-LFS-induced depotentiation(ex vivo)). Importantly, extinction occluded pp-LFS-induced depotentiation(ex vivo), suggesting that extinction shares some mechanisms with the depotentiation. pp-LFS-induced depotentiation(ex vivo) required NMDA receptor (NMDAR) activity, consistent with a previous finding that blockade of amygdala NMDARs impaired fear extinction. In addition, pp-LFS-induced depotentiation(ex vivo) required activity of group II metabotropic glutamate receptors (mGluRs), known to be present at presynaptic terminals, but not AMPAR internalization, consistent with a presynaptic mechanism for pp-LFS-induced depotentiation(ex vivo). This result is in contrast with another form of ex vivo depotentiation in the thalamic pathway that requires both group I mGluR activity and AMPAR internalization. We thus suggest that extinction of conditioned fear involves a distinct form of depotentiation at C-LA synapses, which depends upon both NMDARs and group II mGluRs.

Neuropilin-2/PlexinA3 Receptors Associate with GluA1 and Mediate Sema3F-Dependent Homeostatic Scaling in Cortical Neurons.

Regulation of AMPA-type glutamate receptor (AMPAR) number at synapses is a major mechanism for controlling synaptic strength during homeostatic scaling in response to global changes in neural activity. We show that the secreted guidance cue semaphorin 3F (Sema3F) and its neuropilin-2 (Npn-2)/plexinA3 (PlexA3) holoreceptor mediate homeostatic plasticity in cortical neurons. Sema3F-Npn-2/PlexA3 signaling is essential for cell surface AMPAR homeostatic downscaling in response to an increase in neuronal activity, Npn-2 associates with AMPARs, and Sema3F regulates this interaction. Therefore, Sema3F-Npn-2/PlexA3 signaling controls both synapse development and synaptic plasticity.