Skip to main content
Download PDF

CPEB2-activated axonal translation of VGLUT2 mRNA promotes glutamatergic transmission and presynaptic plasticity

Journal of Biomedical Science volume 31, Article number: 69 (2024) Cite this article

Abstract

Background

Local translation at synapses is important for rapidly remodeling the synaptic proteome to sustain long-term plasticity and memory. While the regulatory mechanisms underlying memory-associated local translation have been widely elucidated in the postsynaptic/dendritic region, there is no direct evidence for which RNA-binding protein (RBP) in axons controls target-specific mRNA translation to promote long-term potentiation (LTP) and memory. We previously reported that translation controlled by cytoplasmic polyadenylation element binding protein 2 (CPEB2) is important for postsynaptic plasticity and memory. Here, we investigated whether CPEB2 regulates axonal translation to support presynaptic plasticity.

Methods

Behavioral and electrophysiological assessments were conducted in mice with pan neuron/glia- or glutamatergic neuron-specific knockout of CPEB2. Hippocampal Schaffer collateral (SC)-CA1 and temporoammonic (TA)-CA1 pathways were electro-recorded to monitor synaptic transmission and LTP evoked by 4 trains of high-frequency stimulation. RNA immunoprecipitation, coupled with bioinformatics analysis, were used to unveil CPEB2-binding axonal RNA candidates associated with learning, which were further validated by Western blotting and luciferase reporter assays. Adeno-associated viruses expressing Cre recombinase were stereotaxically delivered to the pre- or post-synaptic region of the TA circuit to ablate Cpeb2 for further electrophysiological investigation. Biochemically isolated synaptosomes and axotomized neurons cultured on a microfluidic platform were applied to measure axonal protein synthesis and FM4-64FX-loaded synaptic vesicles.

Results

Electrophysiological analysis of hippocampal CA1 neurons detected abnormal excitability and vesicle release probability in CPEB2-depleted SC and TA afferents, so we cross-compared the CPEB2-immunoprecipitated transcriptome with a learning-induced axonal translatome in the adult cortex to identify axonal targets possibly regulated by CPEB2. We validated that Slc17a6, encoding vesicular glutamate transporter 2 (VGLUT2), is translationally upregulated by CPEB2. Conditional knockout of CPEB2 in VGLUT2-expressing glutamatergic neurons impaired consolidation of hippocampus-dependent memory in mice. Presynaptic-specific ablation of Cpeb2 in VGLUT2-dominated TA afferents was sufficient to attenuate protein synthesis-dependent LTP. Moreover, blocking activity-induced axonal Slc17a6 translation by CPEB2 deficiency or cycloheximide diminished the releasable pool of VGLUT2-containing synaptic vesicles.

Conclusions

We identified 272 CPEB2-binding transcripts with altered axonal translation post-learning and established a causal link between CPEB2-driven axonal synthesis of VGLUT2 and presynaptic translation-dependent LTP. These findings extend our understanding of memory-related translational control mechanisms in the presynaptic compartment.

Background

Encoding transient learned information into long-term memory requires de novo protein synthesis to maintain long-lasting modifications of activated synapses [1]. To achieve input-specific reorganization of the proteome only at participating synapses for strengthening memory engrams, local translation is a more effective means than long-distance convoy of proteins from the soma, given that axonal terminals can be, to the extreme, a meter away from soma [2,3,4,5]. Nevertheless, the study of local translation mechanisms has focused on postsynaptic dendrites because in electron micrographs of the adult mammalian brain, no polyribosome was confidently observed at presynaptic axonal terminals as compared with postsynaptic dendritic shafts [6]. In contrast, regulated translation has been found in invertebrate axons [7, 8], growth cones of developing central nervous system axons [9,10,11,12], and regenerating dorsal root ganglion axons [13,14,15,16]. Several axonal transcriptomes and proteomes from cultured neurons, developing retinal ganglia, and developing callosal projection of the cortex reinforced the existence of a unique repertoire of mRNAs in axons [9, 10, 17, 18].

Although axonal translation was historically thought to exist only in developing and regenerating mammalian axons, recent biochemical and imaging studies have detected ribosomes in mature axons, some of which exist in monosomes [19,20,21]. Likewise, despite the lack of axonal ultrastructure like endoplasmic reticulum and Golgi apparatus, mRNAs encoding transmembrane and secretory proteins were found in axons [17, 18, 20, 22] and membrane-targeting and secretion of axonal proteins could be blocked locally by brefeldin A [23, 24], showing the presence of functional endoplasmic reticulum and Golgi routes in axons. Moreover, cannabinoid retrograde signaling increased presynaptic protein synthesis to support specific forms of long-term depression in hippocampal and striatal slices prepared from juvenile rodents [21, 25]. Importantly, a study using the translating ribosome affinity purification (TRAP) technique first profiled a learning-associated axonal translatome in cortico-amygdala projections (i.e., TE3 cortical region to lateral amygdala) of cued fear-conditioned adult rats [20]. Although these studies confirm that axonal protein synthesis in mature neurons is associated with synaptic plasticity and modulated by the animal’s learning experience, there is no direct evidence for which RBP in axons controls target-specific mRNA translation to regulate LTP and memory. To bridge this gap, we investigated the presynaptic role of CPEB2.

CPEB2 belongs to the CPEB family of 4 proteins in vertebrates, all of which bind to the CPE (UUUUA1-2U) and control translation of target-specific mRNAs in neurons [26,27,28,29,30,31,32,33,34,35]. Although most identified CPEB-regulated mRNAs function at postsynaptic sites, CPEB1-mediated translation in Aplysia axons and growth cones of developing hippocampal neurons is essential for the persistence of long-term facilitation [7, 8] and the growth and branching of processes [36], respectively. The loss of CPEB1, CPEB3 or CPEB4 in the presynaptic CA3 neurons in the SC-CA1 pathway did not affect paired-pulse facilitation (PPF) [26, 37, 38], a short form of presynaptic plasticity as a transient increase in the probability of vesicle release [39]. Because most CPEB2-knockout (KO) mice die within 3 days after birth due to respiratory failure [40, 41], the SC-CA1 circuit was only investigated in CPEB2-conditional KO (cKOCamk2) mice whose Cpeb2 was ablated in postsynaptic CA1 but not presynaptic CA3 neurons [35]. CPEB2-promoted postsynaptic translation of GRIP-associated protein 1 (GRASP1) mRNA is essential to control the surface expression of alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid-type glutamate receptors (AMPARs) to support long-lasting LTP and spatial memory [35]. However, the role of CPEB2 in presynaptic translation and plasticity has not been investigated.

In this study, we identified that CPEB2 deficiency in presynaptic neurons reduced fiber volley amplitude, enhanced PPF, and impaired protein synthesis-dependent LTP in both SC-CA1 and TA-CA1 pathways. Cross-comparison of the CPEB2-immunoprecipitated transcriptome with the learning-associated axonal translatome [20] revealed Slc17a6, whose expression was translationally upregulated by CPEB2. Slc17a6 encodes VGLUT2, a presynaptic protein that takes up glutamate into synaptic vesicles [42,43,44]. Moreover, high-frequency stimulation induced CPEB2-mediated translation in transected VGLUT2-positive (VGLUT2+) axons, which was sufficient to support protein synthesis-dependent LTP in the TA-CA1 pathway and replenish the releasable pool of VGLUT2+ vesicles in cultured neurons. Together with our previous findings [35], CPEB2 regulates not only post- but also pre-synaptic translation to coordinately strengthen glutamatergic circuits, thereby leading to long-term modifications in behavior.

Methods

Animals and genotyping

All procedures were approved by Institutional Animal Care and Use Committee (protocol number: 12–12-491). Mice were housed in a 12-h light–dark (lights on from 8:00 to 20:00) room with ad libitum access to food and water. Nestin-Cre (JAX#003771) and Vglut2-Cre (JAX#028863) mice were from the Jackson Laboratory [45, 46]. CPEB2-cWT and -cKO mice from the mating of Cpeb2f/f, Nestin−Cre/+ (cKONes) or Cpeb2f/f, Vglut2−Cre/+ (cKOVglut2) male and Cpeb2f/f, +/+ (cWT) female mice were used for the study. CPEB2-WT and -KO embryos or adult mice were collected from Cpeb2+/- heterozygous mating. The genotypes were determined by PCR of tail biopsies as described [41].

Intracranial injection of adeno-associated virus (AAV)

Recombinant AAV viruses, ssAAV9-Cre-ires-GFP (5.4 × 1010 vg/μl, GFP expression is mediated by internal ribosome entry site), dsAAV8-GFP (1.4 × 1010 vg/μl) and ssAAV9-GFP (5.4 × 1010 vg/μl), were purchased from the institutional AAV core. Stereotaxic injection was applied to male mice as previously mentioned with modifications [35]. In brief, the entorhinal cortex (EC) region (AP—2.46 mm, ML ± 4.29 mm, DV—4.25 mm relative to bregma) and CA1 region (AP—2.00 mm, ML ± 1.25 mm, DV—1.20 mm relative to bregma) were located by using the Robot Stereotaxic system (Neurostar). A 33-gauge needle syringe was used to deliver 0.3 μl virus to each region at the rate of 12 μl/h. After surgery, the mice were monitored until they were fully ambulatory and then recovered for at least 4 weeks in housing cages before field recording.

Behavioral assays and electrophysiology

All behavioral assays and field recording in the SC-CA1 circuit were previously described [26, 35] (Additional File 1: Supplemental method). To avoid the influence of estrous cycle in behavioral performance, only male littermates between 2 and 4 months old were used for behavioral tasks during a 13:30–17:30 light phase with the observer blinded to their genotypes. Littermates of both sexes were used for SC-CA1 recording, but only male mice were used for TA-CA1 recording because sex differences were observed in the latter circuit [47]. The electrophysiological responses were recorded at stratum radiatum (SC pathway) or stratum lacunosum-moleculare (TA pathway) of CA1 neurons after SC axons and TA axons were severed from the soma of CA3 and entorhinal cortical neurons, respectively [47, 48]. The input–output responses were measured with increasing stimulus intensity from 10 to 100 μA at a 10-μA interval. PPF was measured with indicated interpulse interval ranging from 10 to 250 ms. The baseline was defined by the first 20-min recording under 0.017 Hz and 0.1-ms pulse duration, then LTP was evoked by 1 or 4 trains of 1-s 100-Hz high-frequency stimulation (1X HFS or 4X HFS) with 5-min inter-train intervals. For field-recording on hippocampal slices prepared from AAV-injected brains, GFP signal in the stratum lacunosum moleculare or the CA1 cell layer was used to confirm the delivery of AAV at the EC or CA1 region, respectively. Only GFP-positive slices were used for electro-recording.

Primary neuron culture and luciferase reporter assay

Cortical and hippocampal tissues of E18.5 embryos from Cpeb2+/- heterozygous mating or cortical tissues of E18.5 rat embryos were collected for neuron culture as reported [35]. The percent of neurons in DIV14-DIV22 cultures was more than 85% based on counting the number of MAP2-positive neurons and GFAP-positive glia. Neurons were seeded at 3 × 106 cells/60-mm dish for immunoblotting and 106 cells/35-mm dish for RNA isolation by using TRIzol. Neurons were plated at 105 or 5 × 105 cells/18-mm coverslip in a 12-well plate for fluorescence in situ hybridization or FM4-64FX loading assay, respectively, and 7 × 105 cells/well in a 12-well plate for 4X HFS-induced protein expression. The Slc17a6 3'-UTR was amplified from mouse brain cDNA with the primers 5'-GGACTAGTCCGATGCTAGTTGCTGGATTCATTTG and 5'-CGGGATCCGTTTTGATTTTACAAAAGGGTAAATAGAA, and cloned into the pGL3-promoter vector (Promega). Human embryonic kidney 293 T (HEK293T) or Neuro-2a (N2a) cells were transfected with 0.5 μg plasmid mixture for dual luciferase assay (Promega).

Immunohistochemistry, imaging acquisition, and quantification

The detailed procedures were previously described [26, 35]. Briefly, brains isolated from mice after cardiac perfusion of PBS and 4% formaldehyde/PBS were post-fixed in 4% formaldehyde overnight, then immersed in 30% (w/v) sucrose/PBS at 4 °C overnight. Brains were embedded in Tissue-Tek OCT compound and sectioned coronally at 30 μm or 10 μm by using a Leica cryostat. Selected sections were processed for immunohistochemistry (30-μm thickness) with indicated primary antibodies (Additional File 1: Table S1) or FISH (10-μm thickness) as described in the next paragraph. Images were acquired by AxioImager Z1 (Zeiss) or a laser-scanning LSM780 (Zeiss) confocal microscope and quantified by using ImageJ (NIH).

Fluorescence in situ hybridization (FISH)

Stellaris FISH Probes, mouse Slc17a6 with CAL Fluor Red 610 (VSMF-30156–5, LGC Biosearch Technologies), were hybridized to brain slices or cultured neurons, following the manufacturer’s instructions and the published study with modifications [49]. In brief, all solutions were prepared in diethyl pyrocarbonate (DEPC)-based PBS unless otherwise specified. DIV17 neurons on coverslips were fixed with 2% formaldehyde for 15 min, then washed with PBS for 5 min × 3 times. Fixed samples were permeabilized with 0.3% Triton X-100 for 10 min, washed with PBS for 3 times and then rinsed 3 times with the hybridization buffer (30 mM sodium citrate pH 7, 0.3 M NaCl, 10% dextran sulfate, 2 mM ribonucleoside vanadyl complex, 10% formamide, 200 μg/ml tRNA, 80 μg/ml bovine serum albumin [BSA] in DEPC-H2O). The FISH Probes were diluted to a final 0.25 μM in the blocking buffer (1% w/v Blocking Reagent [11096176001, Roche], 10 mM maleic acid pH 7.5, 15 mM NaCl, 0.2 mM Tris–HCl pH 8, 20 μM EDTA in hybridization buffer) along with the designated primary antibodies (Additional File 1: Table S1). After 37 °C hybridization in a humidified dark chamber for at least 16 h, samples were washed with 0.3% Triton X-100 in PBS for 5 min × 3 times, followed by incubation with secondary antibodies for 1 h. After one rinse with PBS, samples were post-fixed with 2% formaldehyde for 15 min and washed with PBS for 5 min × 3 times before mounting with ProLong Gold Antifade Mountant (ThermoFisher). Similar procedures were performed to detect Slc17a6 RNA in brain sections, with the inclusion of 1 μg/ml proteinase K treatment at 37 °C for 10 min subsequent to permeabilization in 0.3% Triton X-100 to enhance RNA detection.

Activity-induced FM4-64FX loading assay

DIV17-23 neurons on coverslips or in microfluidic chambers were activated ± 4X HFS ± 200 μg/ml cycloheximide, then harvested at designated times for immunoblotting or incubated for 2 h, followed by the loading of FM4-64FX as described [50]. All solutions were prepared in Tyrode’s buffer (25 mM HEPES pH 7.4, 119 mM NaCl, 2.5 mM KCl, 2 mM CaCl2, 2 mM MgCl2 and 30 mM D-glucose) unless otherwise indicated. Briefly, coverslips were rinsed twice with Tyrode’s buffer and once with the loading buffer (10 μM FM4-64FX [Thermo Fisher], 50 μM APV and 10 μM CNQX) and then stimulated for 90 s of 10-Hz in the same buffer. After 20-min incubation, neurons were washed twice with Tyrode’s buffer, then incubated for 10 min in 1 mM ADVASEP-7 (binds and removes dye from external membranes, Merck), 50 μM APV and 10 μM CNQX [51]. After one wash of Tyrode’s buffer, neurons were fixed with 4% formaldehyde/PBS for 10 min, permeabilized with 0.2% Triton X-100/PBS for 10 min, and blocked with 5% BSA/PBS for 1 h, followed by overnight incubation of primary antibodies (Additional File 1: Table S1) in 1% BSA/PBS at 4 °C, then secondary antibodies for 2 h at room temperature before mounting with ProLong Gold Antifade Mountant.

Synaptosome preparation and stimulation

All steps were performed at 4 °C unless otherwise specified, and buffers were oxygenated with 95% O2 and 5% CO2 for approximately 2 h before use. Cortical and hippocampal tissues of adult mice were fractionated to obtain crude synaptosomes [26]. Tissues were homogenized (600 μl buffer/100 mg tissue) in the buffer containing 8.33 mM D-glucose, 250 mM sucrose, 20 mM HEPES pH 7.4, 25 mM Tris–HCl pH 7.4, 10 mM KCl, 1.5 mM MgCl2, 1 mM EDTA, 1 mM DTT, 1X protease inhibitor (Sigma-Aldrich), 1X PhosSTOP (Roche), and 60 U/ml RNase Inhibitor (TOOLS), by using a Dounce homogenizer with the loose pestle for 50 strokes twice at an interval of 5 min. The homogenates were centrifuged at 300 xg for 2 min to remove debris, 700 xg for 5 min to remove nuclei, and then 9200 xg for 20 min to collect pellets as crude synaptosomes, which were resuspended in Krebs buffer (212.7 mM D-glucose, 118.5 mM NaCl, 24.9 mM NaHCO3, 1.18 mM Na2SO4, 1.18 mM KH2PO4, 2.5 mM CaCl2, 15 μl/ml 1N HCl, 1X protease inhibitor, 1X PhosSTOP, and 60 U/ml RNase Inhibitor) and aliquoted into 200 μl per tube. Synaptosomal samples were pre-warmed at 37 °C for 5 min, then stimulated without (Ctrl) or with 50 μM NMDA and 10 μM glutamate for 30 s, followed by the addition of 120 μM APV ± 200 μg/ml cycloheximide, then incubated at 37 °C for designated times according to the published protocol [52]. All samples were centrifuged at 9200 xg for 20 min to collect synaptosomal pellets, which were lysed and sonicated in the buffer containing 50 mM Tris–HCl pH 6.8, 0.2% Triton X-100, 0.04% sodium deoxycholate, 1.5% SDS and 5% glycerol, then used for immunoblotting.

Immunoblotting and quantification

Adult mouse cortex or DIV21 cultured neurons was homogenized in the lysis buffer (50 mM HEPES pH 7.4, 150 mM NaCl, 1 mM MgCl2, 0.5% Triton X-100, 1 mM DTT, 1 mM EDTA, 1 mM EGTA, 2X protease inhibitor [Sigma-Aldrich] and 1X PhosSTOP [Roche]) for 30 min on ice before centrifugation at 16,000 xg for 5 min at 4 °C. The protein concentration of cortical, neuronal and synaptosomal lysates was quantified by using the Bradford protein assay (Bio-Rad) or Pierce BCA assay kit (Thermo Fisher). Equal amounts of samples were mixed with Laemmli sample buffer and heated at 55 °C for 20 min, separated on 10% SDS-PAGE, then transferred to nitrocellulose membranes (Millipore). Primary antibodies were incubated overnight at 4 °C and then horseradish peroxidase-conjugated secondary antibodies (Additional File 1: Table S1). Immunostained signals detected with chemiluminescence substrates (Millipore) were acquired by using ImageQuant LAS 4000 Mini (GE Healthcare) and then quantified by using ImageJ (NIH).

RNA immunoprecipitation (RIP) and quantitative RT-PCR (RT-qPCR)

The previous protocols were followed with modifications [30, 35]. In brief, cortical and hippocampal lysates from adult male mice were divided and incubated with control or CPEB2 IgG-bound protein G beads for 3 h. One tenth of the beads were kept for immunoblotting, and the rest were isolated for RNA for microarray analysis. Similarly, to validate selected CPEB2-binding targets, hippocampal lysates from adult male mice were used. One third of the beads were used for immunoblotting, and the rest were isolated for RNA for RT-qPCR. Total RNA from adult cortex and DIV21 neurons were isolated by TRIzol (Invitrogen). RT-qPCR was performed by using ImPromII Reverse Transcriptase (Promega) and Universal Probe Library and Lightcycler 480 system (Roche). Primers and probes for qPCR are: Slc17a6, 5'-TGTTGGTGCAATGACAAAGAA, 5'-CCTGAGGCAAATAGTGCATAAA, and UPL probe #22; Slc17a7, 5'-GCAGGAGGAGTTTCGGAAG, 5'-CCTGCCGCTTCTCCAGTA, and UPL probe #75; Syt1, 5'-GCGATCTCCAGAGTGCTGAG, 5'-GACAACAGTCAGCTTGCCG, and UPL probe #7; Gapdh, 5'-GCCAAAAGGGTCATCATCTC, 5'-CACACCCATCACAAACATGG, and UPL probe #29.

RIP-microarray and gene ontology (GO) analysis

The procedure with use of RIP samples for microarray analysis (Agilent SurePrint G3 Mouse GE 8 × 60 K, Welgene, Taiwan) was described before [30]. Briefly, 2,685 transcripts with raw signals > 1000, > 1.5-fold enrichment in CPEB2 IgG-precipitated substances, and at least one CPE in 3'-UTR were identified from 2 independent RIP-microarray experiments (Additional File 2). The learning-associated translatome (GSE124592) was reanalyzed to identify 2,438 genes, showing ± 1.5-fold change in expression only in the axonal compartment after fear learning (Additional File 3). CPEB2-bound axon-specific transcripts were identified by cross-comparison of 2,685 CPEB2-RIP targets with 2,438 axon-translated mRNAs (Additional File 4) and then GO-analyzed by Metascape [53] with the key word ‘glutamatergic transmission’ included to filter the results (Additional File 4).

Statistics

Analysis was performed using SigmaPlot software. The statistical methods and analyzed results are listed (Additional File 5).

Results

CPEB2 regulated both pre- and post-synaptic transmission in the SC pathway

To explore an RBP-mediated presynaptic/axonal translation for plasticity and memory, we first examined whether CPEB2 affects presynaptic electrophysiological responses in hippocampal slices. We used a Nestin-Cre mouse line to generate CPEB2-cKONes mice with Cpeb2 ablated in most neurons and glia including those in CA1 and CA3 regions (Fig. 1A). In our previous study, CPEB2 was only depleted in hippocampal CA1 pyramidal neurons but not CA3 in CPEB2-cKOCamk2 mice. Hence, they only exhibited impaired postsynaptic responses and long-lasting LTP due to lack of postsynaptic CPEB2 in the SC pathway [35]. Here, we measured the input–output relationship in the SC pathway which measures presynaptic (fiber volley [FV] amplitude) and postsynaptic (the slope of field excitatory postsynaptic potential [fEPSP]) responses of CA3 and CA1 neurons, respectively. Both fEPSP slopes (F(1,160) = 32.264, P < 0.001) and FV amplitudes (F(1,160) = 9.928, P = 0.002) were reduced in CPEB2-cKONes mouse hippocampal slices (Fig. 1B). Moreover, PPF, which monitored presynaptic neurotransmitter release and was inversely correlated with the release probability [54], was enhanced in CPEB2-cKONes mouse hippocampal slices (F(1,154) = 27.033, P < 0.001, Fig. 1C). Translation-dependent LTP elicited by 4X HFS [55] lasted for 3 h in the CPEB2-cWT but not CPEB2-cKONes SC pathway (F(1,40) = 15.149, P < 0.001, Fig. 1D). As compared with what we previously observed in CPEB2-cKOCamk2 mice [35], the additional deletion of CPEB2 in CA3 neurons of CPEB2-cKONes mice caused abnormal presynaptic responses, including FV amplitude and PPF (Fig. 1B, C), and failed to maintain long-lasting LTP for more than 1 h (Fig. 1D), rather than merely decreasing LTP magnitude in CPEB2-cKOCamk2 mice [35], so CPEB2 also played a presynaptic role in modulating glutamatergic transmission.

Fig. 1
figure 1

Aberrant synaptic transmission and LTP in CPEB2-cKONes SC-CA1 pathway. A Mice carrying the Cpeb2 allele with two loxP sites flanking exons 3–5 (fCPEB2) were crossed with Nestin-Cre mice to generate CPEB2-cWT and -cKONes mice. CPEB2 immunohistochemistry of coronal brain slices was from adult CPEB2-cWT and -cKONes mice. B-D Electrophysiological responses in the SC-CA1 pathway of adult (3- to 5-month-old) mice, including the input–output curve of the fEPSP slope and fiber volley amplitude in response to increasing stimulus intensity (B), paired-pulse facilitation (C), and LTP evoked by 4X HFS (D). Sample traces were the average fEPSP of 5 min before (gray) and the very last 5 min (black) after stimulation. Numbers in parentheses (n/N) represent the number of recorded slices (n) and mice (N). Sample traces with vertical scales of 0.5 mV and horizontal scales of 10 ms (i-o curve and LTP) and 50 ms (PPF). Data are mean ± SEM. *P < 0.05 and **P < 0.01, two-way ANOVA with Fisher’s LSD post-hoc comparison

Full size image

CPEB2-cKONes mice exhibited defective hippocampus-dependent learning and memory

To assess whether CPEB2-cKONes mice exhibited cognitive impairment, we conducted behavioral assays previously used in CPEB2-cKOCamk2 mice, which showed defective consolidation of spatial memory in the Morris water maze test and fear memory after contextual fear conditioning [35], but excluded the fear conditioning test because hemizygous Nestin-Cre mice showed defective contextual- and cued-conditioned fear [45]. The exploratory activity in the open field (Additional File 1: Fig. S1A) and the anxiety level in the elevated plus maze (Additional File 1: Fig. S1B) were comparable between CPEB2-cKONes mice and their cWT littermates. After 4-day acquisition training in the Morris water maze, both groups learned to find the hidden platform in the target quadrant (F(3,52) = 14.609, P < 0.001), but CPEB2-cKONes mice performed worse than their littermates (F(1,52) = 7.173, P = 0.010, Additional File 1: Fig. S1C). In the day 5 probe test, we analyzed the percentage of time mice spent in the target quadrant where the platform was previously placed. CPEB2-cKONes mice showed reduced memory consolidation (F(3,52) = 4.182, P = 0.010, P = 0.020 in target quadrant, Additional File 1: Fig. S1D), but their swimming ability and visual acuity were normal (Additional File 1: Fig. S1E).

Systemic identification of CPEB2-bound transcripts involved in learning-associated presynaptic translation

Since SC axons were transected from the soma of CA3 neurons during electro-recording and additional depletion of presynaptic CPEB2 severely impaired protein synthesis-dependent LTP (Fig. 1D), CPEB2 must regulate axonal translation to support enduring neurotransmitter release for LTP. To identify the candidate mRNAs regulated by CPEB2, we used RIP with adult mouse cortical and hippocampal tissues. CPEB2- and control IgG-precipitated materials were used for immunoblotting of CPEB2 (Fig. 2A) and RNA isolation for microarray analysis. From 2 independent RIP experiments, 3,309 transcripts with raw signal > 1,000 were enriched more than 1.5-fold in CPEB2-immunoprecipitated samples, and 2,685 contained at least 1 CPE in their 3'-untreanslated region (3'-UTR) (Fig. 2B and Additional File 2). To investigate whether CPEB2 could regulate the translation of specific axonal mRNAs to affect presynaptic plasticity and memory, we cross-compared CPEB2-bound transcriptome with the only learning-associated axonal translatome (GSE124592). In this study, Ostroff et al. used the TRAP method [56] to isolate ribosome-associated transcripts (i.e., translatome) in axons and somas of auditory cortical TE3 neurons before and after cued fear conditioning [20]. We analyzed 14,520 TRAP-isolated transcripts from this dataset and found 7,454 transcripts with more than ± 1.5-fold change after fear conditioning (Fig. 2B and Additional File 3). Moreover, a total of 2,438 transcripts in the trained translatome showed changed expression only in the axonal compartment (Fig. 2C and Additional File 3). Cross-comparison of CPEB2-bound transcriptome with the 2,438 transcripts identified 272 mRNAs being CPEB2-bound and axon-translated targets (Fig. 2D and Additional File 4). Because of aberrant glutamatergic transmission in the CPEB2-cKONes SC circuit (Fig. 1), we included the ‘glutamatergic transmission’ keyword in Metascape gene annotation analysis [53] and identified 7 candidates (Additional File 4). We excluded 4 of them, with expression mainly in astrocytes (i.e., Glul) [57] or whose function was mainly at the postsynapse (i.e., Homer1, Gria3 and Grin3a) [58,59,60,61]. The remaining 3 candidates, Ext1, Slc17a6 and Syt1, encode proteins localized in the presynaptic compartment and contain 6, 3 and 6 CPEs, respectively, in their 3'-UTRs (Fig. 2D). Ext1 encodes exostosin glycosyltransferase 1, which functions in both pre- and post-synaptic domains [62, 63], so we focused on Slc17a6 and Syt1, which encode VGLUT2 for transporting glutamate to synaptic vesicles [64, 65] and SYT1 for calcium-dependent exocytosis of glutamate- or GABA-containing vesicles [66,67,68].

Fig. 2
figure 2

Identification and validation of learning-associated presynaptic Slc17a6 is translationally upregulated by CPEB2. A Adult mouse cortices and hippocampi were used for RNA immunoprecipitation with CPEB2 or control (Ctrl) IgG. The precipitated substances were processed for western blot analysis or RNA isolation for microarray assay. B The flowchart denoted the criteria used to compare the CPEB2-immunoprecipitated transcriptome (GSE208738) and the learning-related translatome (GSE124592). C Among 7454 transcripts, 2438 showed more than 1.5-fold change in expression only in the axonal compartment: 1477 were upregulated (red dot) and 961 downregulated (blue dot). D Among the 2438 axon-specific translated mRNAs, 272 were CPEB2-binding targets with at least one CPE in their 3'-UTRs (purple and green dots). Three were known to function at glutamatergic presynapses and contain multiple CPEs in their 3'-UTRs of varying lengths. E Adult cortices and cultured neurons at 21 days in vitro (DIV) from global CPEB2-wild type (CPEB2-WT) and -KO mice were used for immunoblotting. The protein levels were normalized to that of LRP130. F RT-qPCR analysis of Slc17a6 mRNA level (normalized to Gapdh level) in adult cortex and DIV 21 neurons. G Dual luciferase reporter assay. The firefly luciferase reporter appended with mouse Slc17a6 3'-UTR and Renilla luciferase plasmids were co-transfected with the plasmid expressing EGFP, myc-tagged full-length (CP2) or C-terminus (CP2C) of CPEB2 into HEK293T or N2a cells (duplicate transfections in 3 or 5 independent cultures, respectively). Data are mean ± SEM. *P < 0.05, **P < 0.01 and ***P < 0.01, one-way ANOVA with Fisher’s LSD post-hoc test or Student’s t test

Full size image

Expression of VGLUT2 was translationally activated by CPEB2

VGLUT2 and SYT1 are exclusively concentrated in synaptic vesicles, so they may rely on local translation to maintain their presynaptic demands. Using CPEB2-WT and -KO adult cortices and cultured neurons, we identified decreased protein level of VGLUT2 but not SYT1 (Fig. 2E) and the change was at the protein rather than mRNA level (Fig. 2F). VGLUT1, encoded by Slc17a7, is more widely expressed than VGLUT2 in the hippocampus and cortex (Additional File 1: Fig. S2A) and also takes up glutamate into synaptic vesicles [64, 65]. Although the 3'-UTR of VGLUT1 mRNA contained no CPE and was not enriched in the CPEB2-immunoprecipitated transcriptome, we also examined and found that CPEB2 deficiency did not affect VGLUT1 level (Fig. 2E). Moreover, using the firefly luciferase reporter appended to the 3'-UTR of mouse Slc17a6, we found that the full-length but not C-terminal RNA-binding domain of CPEB2 enhanced the synthesis of firefly luciferase as compared with the GFP control (Fig. 2G). Therefore, CPEB2 bound to and promoted the translation of VGLUT2 mRNA.

CPEB2 regulated synaptic plasticity in the TA-CA1 pathway

Hippocampal CA1 neurons receive glutamatergic afferents from CA3 neurons in the stratum radiatum and from layer III entorhinal cortex neurons in the stratum lacunosum moleculare [69]. Because most presynaptic terminals express VGLUT2 in the stratum lacunosum moleculare (Fig. 3A and Additional File 1: Fig. S2A), we then determined whether CPEB2-controlled protein synthesis, such as for VGLUT2, was important for presynaptic glutamate release and plasticity by recording the TA-CA1 circuit [47, 48] (Fig. 3B). Similar to what we observed in the SC circuit (Fig. 1), the input–output curve of CPEB2-cKONes slices was diminished in both fEPSP slopes (F(1,260) = 58.060, P < 0.001) and FV amplitudes (F(1,260) = 17.149, P < 0.001, Fig. 3C) and the CPEB2-cKONes TA-CA1 pathway exhibited enhanced PPF (F(1,161) = 19.319, P < 0.001, Fig. 3D) and defective 4X HFS-evoked LTP (F(1,51) = 8.771, P = 0.005, Fig. 3E). We then used a Vglut2-Cre mouse line to generate CPEB2-cKOVglut2 mice and investigated TA-CA1 plasticity under CPEB2 deficiency in the subset of glutamatergic neurons. Due to transient expression of VGLUT2 in the developing hippocampus [64, 65, 70], CPEB2 was depleted in most pre- (i.e., entorhinal cortex) and post-synaptic (i.e., hippocampus) glutamatergic neurons in the TA-CA1 pathway (Additional File 1: Fig. S2B), including all CaMK2α-expressing CA1 neurons (Additional File 1: Fig. S2C). Similarly, we observed diminished fEPSP slopes (F(1,200) = 35.106, P < 0.001) and FV amplitudes (F(1,200) = 25.076, P < 0.001, Fig. 3F), enhanced PPF (F(1,133) = 12.826, P < 0.001, Fig. 3G) and defective 4X HFS-evoked LTP (F(1,42) = 11.022, P = 0.002, Fig. 3H) in CPEB2-cKOVglut2 TA-CA1 circuit. However, protein synthesis-independent LTP elicited by 1X HFS was unaffected in CPEB2-cKONes and -cKOVglut2 slices (Additional File 1: Fig. S3A and S3B), reinforcing the role of CPEB2 in translation-dependent long-term synaptic plasticity [35].

Fig. 3
figure 3

CPEB2 deficiency affects glutamatergic transmission and translation-dependent LTP in the TA-CA1 circuit. A Differential distribution of VGLUT1+ and VGLUT2+ afferents on hippocampal CA1 neurons. SO, stratum oriens; SR, stratum radiatum; SLM, stratum lacunosum moleculare; SM, stratum moleculare. Scales, 0.2 mm. B The illustration denoted SC-CA1 and TA-CA1 pathways. C-H CPEB2-cKONes and -cKOVglut2 male mice (3 to 5 months old) were used for field recording. The input–output responses (C and F), paired-pulse facilitation (D and G) and 4X HFS-evoked LTP (E and H) in the TA-CA1 pathway of CPEB2-cKONes and -cKOVglut2 hippocampal slices, respectively. Numbers in parentheses (n/N) represent the number of recorded slices (n) and mice (N). Sample traces were presented in the same manner as described in Fig. 1. Data are mean ± SEM. *P < 0.05, **P < 0.01 and ***P < 0.001, two-way ANOVA with Fisher’s LSD post hoc test

Full size image

CPEB2-cKOVglut2 mice showed defective memory consolidation

The loss of CPEB2 in glutamatergic neurons sufficiently impaired glutamatergic transmission and long-term plasticity (Fig. 3F-H), so we examined the behavior of CPEB2-cKOVglut2 mice and found no abnormalities in their exploratory activity in the open field (Fig. 4A), anxiety level in the elevated plus maze (Fig. 4B), and spatial learning in the Morris water maze (Fig. 4C). However, CPEB2-cKOVglut2 mice had reduced memory consolidation (F(3,52) = 4.127, P = 0.011, P = 0.006 in target quadrant, Fig. 4D) despite normal swimming ability and visual acuity (Fig. 4E).

Fig. 4
figure 4

Impaired spatial memory consolidation in CPEB2-cKOVglut2 mice. A-E Adult male mice at 2 to 4 months old were used for behavior assays. A Open-field. Representative moving traces in the open arena during the first 10 min and the quantified entry times, duration, moving distance and velocity. B Elevated plus maze (EPM). Representative 5-min moving traces in the EPM and the quantified entry times, duration and moving distance in the open versus closed arms. C Morris water maze (MWM). Mice were trained with 4 trials per day for 4 consecutive days, and the escape latency to the platform was averaged from 4 trials. D The MWM probe test on day 5 recorded the percentage of time (total 60 s) spent in each quadrant. E In the visual MWM task, the velocity and latency reflected the swimming ability and visual acuity, respectively. Data are mean ± SEM. **P < 0.01, Student’s t test and two-way ANOVA with Fisher’s LSD post-hoc test. The number of mice for behavioral assays was denoted in parentheses

Full size image

CPEB2 regulated both pre- and post-synaptic plasticity in the TA-CA1 pathway

Because CPEB2-activated GRASP1 mRNA translation is important for recycling of AMPARs to postsynaptic sites of CA1 neurons [35], impaired translation-dependent LTP in the SC (Fig. 1D) and TA circuits (Fig. 3E, H) of CPEB2-cKONes and -cKOVglut2 hippocampi could be attributed to both pre- and post-synaptic dysregulation. To determine whether the loss of CPEB2 at the presynaptic TA afferents (CPEB2-cKOEC) or the postsynaptic CA1 neurons (CPEB2-cKOCA1) sufficiently affected translation-dependent LTP, we injected adeno-associated virus expressing GFP or Cre recombinase-ires-GFP (Cre) at the lateral entorhinal cortex (@EC) or the CA1 region (@CA1) of CPEB2-cWT mice, respectively. The injection sites were revealed by the expression of GFP (Fig. 5A) and recordings were conducted using only slices of GFP signal in TA afferents onto the stratum lacunosum moleculare (@EC) or in CA1 neurons (@CA1). CPEB2 deficiency in TA afferents (Cre@EC) reduced FV amplitudes (F(1,180) = 6.095, P = 0.014, Fig. 5B), enhanced PPF (F(1,126) = 12.933, P < 0.001, Fig. 5C) and decreased 4X HFS-evoked LTP (F(1,54) = 12.656, P < 0.001, Fig. 5D) but did not affect 1X HFS-induced LTP (Additional File 1: Fig. S3C). On the other hand, depleting CPEB2 in postsynaptic CA1 (Cre@CA1) reduced fEPSP slopes (F(1,140) = 15.193, P < 0.001, Fig. 5E) and 4X HFS-evoked LTP (F(1,42) = 6.289, P = 0.016, Fig. 5G). To verify that the effect of AAV-Cre on reducing 4X HFS-elicited LTP in CPEB2-cWT mice was not caused by Cre-induced toxicity or by using different AAV serotypes to express GFP or Cre-ires-GFP, we performed experiments in wild-type mice with AAV-injected @EC. No differences were found in 4X HFS-evoked LTP regardless of whether AAV9-GFP (1.62 × 1010 vg), AAV9-Cre-ires-GFP (1.62 × 1010 vg) or AAV8-GFP (0.42 × 109 vg) was used (Additional File 1: Fig. S4). Additionally, cWT mice but not WT mice infected with AAV9-Cre-ires-GFP @EC, exhibited reduced 4X HFS-induced LTP in the TA-CA1 pathway (Additional File 1: Fig. S4D). These results supported the critical role of CPEB2-controlled translation in both pre- and post-synaptic compartments to regulate glutamatergic plasticity.

Fig. 5
figure 5

Pre- or post-synaptic deletion of CPEB2 impairs glutamatergic transmission and LTP in the TA-CA1 circuit. A The virus injection sites in the entorhinal cortex (EC, left) and CA1 (right) were marked in green dots. B-G CPEB2-cWT mice were injected with AAV8-GFP (4.2 × 109 vg) or AAV9-Cre-ires-GFP (1.62 × 1010 vg) to generate cWT control or CPEB2-cKOEC and -cKOCA1 mice, respectively. Four weeks after intracranial delivery of AAV, 3–5-month-old male mice were used for field recording. The input–output responses (B and E), paired-pulse facilitation (C and F) and 4X HFS-evoked LTP (D and G) in the TA-CA1 pathway of CPEB2-cKOEC and -cKOCA1 hippocampal slices, respectively. Numbers in parentheses (n/N) represent the number of recorded slices (n) and mice (N). Sample traces were presented in the same manner as described in Fig. 1. Data are mean ± SEM. *P < 0.05, **P < 0.01 and ***P < 0.001, two-way ANOVA with Fisher’s LSD post hoc test

Full size image

Activity-triggered synaptosomal VGLUT2 synthesis depended on CPEB2

Aberrant presynaptic responses in severed CPEB2-lacking SC and TA afferents were similar to those reported in VGLUT2-cKO mice [70]. Because CPEB2 controls Slc17a6 translation (Fig. 2E-G), we tested whether such a regulation could be axon-localized to affect the function of VGLUT2-containing vesicles. Therefore, we detected Slc17a6 mRNA by FISH, followed by immunostaining of CPEB2 and an axonal marker, tau, in CPEB2-WT neurons (representative images in Fig. 6A). Because only a portion of pyramidal neurons express Slc17a6 mRNA, we randomly selected 12 image fields with at least one soma expressing Slc17a6 puncta for quantification. The number of Slc17a6-positive (Slc17a6+) puncta in somas and axons was counted and expressed as a relative percentage in each image field. Approximately 65% to 75% Slc17a6+ puncta contained a CPEB2 signal in both somatic and axonal compartments (Fig. 6B). Moreover, the number of Slc17a6+ puncta in the stratum lacunosum moleculare layer of CA1 neurons was comparable between CPEB2-cWT and -cKONes hippocampi (Fig. 6C), indicating that CPEB2 deficiency did not affect axonal distribution of Slc17a6 mRNA. Axonal colocalization of CPEB2 and Slc17a6 mRNA supports the possibility of CPEB2-dependent local Slc17a6 translation, so we isolated synaptosomes from cortical and hippocampal tissues and stimulated them with N-methyl-D-aspartate (NMDA) and glutamate to induce chemical LTP [52], which is accompanied by elevated phosphorylation of calcium/calmodulin-dependent protein kinase 2 subunit alpha (CaMK2α) at Thr 286 (Fig. 6D). CPEB2-WT but not CPEB2-KO synaptosomes showed NMDA/glutamate-induced expression of VGLUT2, and this increase could be blocked by cycloheximide, an inhibitor of eukaryotic translation (Fig. 6D).

Fig. 6
figure 6

CPEB2 is present in most Slc17a6 mRNA puncta and promotes VGLUT2 synthesis in NMDA/glutamate-stimulated synaptosomes. A DIV17 neurons were probed with Slc17a6 mRNA by FISH, followed by immunostaining of CPEB2 and the axonal marker tau and Hoechst labeling. Representative images showed the colocalization of Slc17a6 mRNA and CPEB2 in somas (defined by bright-field images) and axons (tau+ areas) with magnified images outlined in blue and yellow squares, respectively. Scales, 10 μm and 1 μm in magnified images. B The percentage of Slc17a6+ puncta containing CPEB2 signal in neurons. Twelve image fields from 4 independent cultures were quantified. C Representative images and quantified results showed the number of Slc17a6+ puncta in the SLM layer of CPEB2-cWT and cKONes hippocampi. Scale, 10 μm. Numbers in parentheses (n/N) represent the number of slices (n) and mice (N). D Synaptosomes from adult cortical and hippocampal tissues (8 independent preparations) were stimulated without (Ctrl) or with 50 μM NMDA and 10 μM glutamate for 30 s, followed by the addition of 120 μM APV ± 200 μg/ml cycloheximide (CHX) for the indicated times and then harvested for immunoblotting. Data are mean ± SEM. *P < 0.05 and ***P < 0.001, Mann–Whitney Rank Sum Test in (B), Student’s t test in (C), and two-way ANOVA with Fisher’s LSD post-hoc test in (D)

Full size image

Reduced VGLUT2 in adult CPEB2-cKONes and -cKOVglut2 hippocampi

Since VGLUT2 was mainly detected in presynaptic axons of EC neurons projecting to hippocampal stratum lacunosum moleculare (Fig. 3A and Additional File 1: Fig. S2A), we used hippocampal tissues and confirmed that CPEB2 also bound to Slc17a6 and Syt1 transcripts but not Slc17a7 mRNA (Additional File 1: Fig. S5A). Moreover, the protein level of VGLUT2 but not SYT1 and VGLUT1 was also diminished in CPEB2-cKONes and -cKOVglut2 adult hippocampi (Additional File 1: Fig. S5B). Therefore, CPEB2 binds to and enhances the expression of VGLUT2 in the TA afferents in the hippocampus.

CPEB2 promoted axonal VGLUT2 synthesis to replenish VGLUT2+ vesicles

Defective protein synthesis-dependent LTP in the CPEB2-cKOEC TA-CA1 pathway (Fig. 5D) strongly indicated that CPEB2-regulated presynaptic translation was important to sustain long-lasting LTP, in part, by affecting activity-induced expression of VGLUT2 in the transected TA axons (Fig. 6D). To support this notion further, we analyzed activity-elicited uptake of FM4-64FX dye into axonal VGLUT2+ vesicles in intact and axotomized neurons. We first confirmed that 4X HFS also increased VGLUT2 expression in CPEB2-WT but not CPEB2-KO cultured neurons (Additional File 1: Fig. S6), then analyzed the endocytosed FM4-64FX signal in VGLUT2+ vesicles after 900 pulse (10 Hz × 90 s)-evoked glutamate release. Therefore, only synaptic vesicles that undergo active release and recycling could be marked by the loading of FM4-64FX from the extracellular space. The loading of FM4-64FX into VGLUT2+ vesicles was comparable between CPEB2-WT and -KO neurons under the basal condition, but 4X HFS increased the FM4-64FX signal in VGLUT2+ vesicles only in CPEB2-WT neurons (Fig. 7A). To ensure that CPEB2-activated local Slc17a6 translation is important to support the function of VGLUT2+ vesicles, neurons cultured in a microfluidic chamber [71], which only permits axons growing into the other side, were severed to separate axons from somas before 4X HFS (Additional File 1: Fig. S7A). Activity-enhanced FM4-64FX-loading signal in VGLUT2+ but not VGLUT2-negative (VGLUT2−) vesicles was sensitive to cycloheximide and defective in CPEB2-KO axons (Fig. 7B and Additional File 1: Fig. S7B). Moreover, the newly synthesized VGLUT2 molecules were mostly used to increase the pool of VGLUT2+ vesicles rather than increase VGLUT2 intensity or vesicular size of existing vesicles (Fig. 7C). Because VGLUT2 is a membrane protein, this finding agreed with the role of CPEB2 to promote translation of VGLUT2 mRNA presumably on ribosomes attached to endoplasmic reticulum and consequently produce new VGLUT2+ vesicles. Hence, activity- and CPEB2-dependent axonal translation was important to support the synaptic function of VGLUT2+ vesicles, thereby contributing to the persistence of long-term presynaptic plasticity (Fig. 8).

Fig. 7
figure 7

Activity- and CPEB2-dependent axonal VGLUT2 synthesis increases functional VGLUT2+ vesicles. A DIV17-23 neurons were stimulated without (basal) or with 4X HFS and incubated for 2 h before activity (10 Hz for 90 s)-stimulated FM4-64FX loading. The FM4-64FX puncta signals in VGLUT2+ vesicles were quantified in 10 image fields from 3 independent cultures and expressed as relative ratios. The region of interest at 4X magnification is shown in the insert. Scales, 10 μm and 1 μm in magnified images. B-C The experimental procedure was illustrated in the upper panel. CPEB2-WT and -KO neurons were seeded in microfluidic chambers and cultured for 18 to 19 DIV. After axotomy, the same stimulation and loading protocol were applied without (control) or with 50 μg/ml cycloheximide (CHX). The axonal region was outlined with an axonal marker, βIII-tubulin (marked by white dashed lines). The signal intensity of FM4-64FX and VGLUT2 was color-coded. The FM4-64FX puncta signals in VGLUT2+ and VGLUT2− vesicles were quantified in 20 image fields from 5 independent cultures and expressed as a relative ratio. Data are mean ± SEM. *P < 0.05, **P < 0.01 and ***P < 0.001, two-way ANOVA with Fisher’s LSD post-hoc test. Scale, 5 μm

Full size image
Fig. 8
figure 8

CPEB2-controlled axonal translation governs glutamatergic transmission and long-term presynaptic plasticity. CPEB2- and activity-dependent axonal translation of VGLUT2 mRNA replenishes synaptic vesicles and maintains presynaptic translation-dependent long-term potentiation in the TA-CA1 circuit

Full size image

Discussion

In this study, we revealed that CPEB2-controlled translation governs presynaptic plasticity and demonstrated that activity- and CPEB2-dependent axonal Slc17a6 translation is critical to replenish VGLUT2 locally and support presynaptic glutamate transmission and translation-dependent LTP.

Unlike developing and regenerating axons, the axonal compartment of mature mammalian neurons is traditionally considered not to possess protein synthesis capacity [6]. However, because of its extensive architecture, activity-driven local synthesis of axonal proteins is intuitively appealing to support presynaptic plasticity in a timely and precise manner. By contrast to polysomes in dendritic spines, monosomes appear to dominate in the axonal compartment, and several omics studies with various approaches have demonstrated that the axonal translatome changes dynamically in response to activity input [10, 17,18,19,20,21]. The maintenance of several forms of memory-associated LTP requires de novo protein synthesis, including 4X HFS-evoked LTP in TA-CA1 and SC-CA1 pathways used in our study, but whether the newly produced proteins are contributed from presynaptic, postsynaptic or both compartments to support long-lasting synaptic strengthening is not entirely clear. Taking the example from our analysis, many learning-affected and axon-specifically translated mRNAs encode proteins in mitochondria to function in both pre- and post-synapses (Fig. 2C and Additional File 3), so identifying candidate mRNAs and determining which compartmental translation supports LTP is difficult. Moreover, most RBPs are widely expressed in the brain; region-specific ablation by crossing with different Cre transgenic mice or AAV delivery is needed to dissect this issue. For instance, the widely used Camk2-Cre mice [72] help in clarifying the postsynaptic contribution in the SC and TA circuits. In this study, we initially used AAV8-GFP as a control for AAV9-GFP-ires-Cre due to the strong neuronal tropism of both AAV serotypes after intracranial delivery [73]. However, several studies have indicated that AAV8 has higher glial tropism than AAV9 [73,74,75,76], so we conducted additional experiments (Additional File 1: Fig. S3C and Fig. S4D) to confirm that the defective LTP observed in Cre@EC mice (Fig. 5D) was not due to Cre toxicity or the differential tropisms of AAV8 and AAV9. Although no significant differences were observed between the use of AAV8-GFP and AAV9-GFP in our study, it is important to mention that using the same AAV serotype and titer is a standard practice in the field. Because presynaptic CPEB2 in VGLUT2-dominated TA circuit supported translation-dependent LTP (Fig. 5D), and 4X HFS- and CPEB2-driven axonal VGLUT2 synthesis increased the number of functional VGLUT2+ vesicles (Fig. 7B, C), so we expect that CPEB2-activated Slc17a6 translation in transected TA axons is important to support long-lasting glutamate release to maintain LTP. Together with our previous finding in CPEB2-cKOCamk2 mice [35] and CPEB2-cKOCA1 (Fig. 5G), CPEB2 plays an important role in coordinating both pre- and post-synaptic translation for glutamatergic plasticity.

VGLUT2-KO mice die immediately after birth because of failure to generate respiratory rhythms [77, 78]. Similarly, most CPEB2-KO mice die postnatally because of elevated airway constriction induced by hyperparasympathetic signaling and impaired pulmonary development [40, 41]. We found no postnatal mortality in CPEB2-cKOVglut2 mice, so the VGLUT2-glutamatergic circuit is unlikely defective in the respiratory system. By crossing with Emx1-Cre mice to ablate Slc17a6 in excitatory neurons in the cortex and hippocampus, VGLUT2-cKOEmx1 mice showed impaired learning and memory in the Morris water maze and aberrant basal transmission, increased PPF and decreased LTP in the SC-CA1 circuit [70]. Despite many similarities in the SC-CA1 connection between VGLUT2-cKOEmx1 and CPEB2-KONes mice, these electrophysiological defects are not merely caused by impaired VGLUT2 mRNA translation in CA3 neurons. Glutamatergic transmission in most excitatory synapses depends on VGLUT1 and VGLUT2, which display complementary expression in the adult brain (Fig. 3A and Additional File 1: Fig. S2A). However, in VGLUT2-cKOEmx1 mice, impaired SC circuit responses could be attributed to abnormal dendritic development and decreased VGLUT1 expression [70], which are not affected in CPEB2-deficient neurons [35] (Fig. 2E). Because SC afferents synapsing onto the stratum radiatum of CA1 neurons comprise mostly a population of excitatory terminals that express VGLUT1 (Fig. 3A), increased PPF and decreased LTP in VGLUT2-cKOEmx1 mice resulted from decreased VGLUT1 level; whereas in CPEB2-cKONes mice, they are likely caused by impaired translation of CPEB2-binding transcripts other than Slc17a6 in SC afferents as well as reduced postsynaptic AMPAR level [35]. Although CPEB2 is expressed more predominantly in neurons than glia [79], astroglial transcripts, Glul, Slc1a2 and Slc1a3, encoding glutamine synthetase and glutamate transporters, were also identified in CPEB2-immunoprecipitated transcriptome. Since astrocytes play a key role in glutamatergic transmission by controlling the glutamate-glutamine cycle and glutamate uptake in the synaptic cleft, further studies are needed to understand whether the loss of CPEB2 in astrocytes contributes to aberrant SC-CA1 transmission in CPEB2-cKONes hippocampus [80]. To our knowledge, the TA-CA1 circuit has not been examined under the VGLUT2-deficient condition. VGLUT1 and VGLUT2 have the same function to control vesicular transport and synaptic release of glutamate [44, 64], so depletion of VGLUT2 in the TA-CA1 circuit is expected to cause similar electrophysiological defects. However, in this study, we identified that the synthesis of VGLUT2 but not VGLUT1 depends on CPEB2 (Fig. 2E). Moreover, VGLUT2 has a shorter half-life than VGLUT1 in cultured cortical neurons (Additional File 1: Fig. S8), implying that local translation may be more important for VGLUT2 than VGLUT1 to control glutamatergic plasticity.

CPEB1 can facilitate dendritic mRNA transport and activate NMDAR-dependent postsynaptic mRNA translation [28, 81], but the role of CPEBs in axonal mRNA transport and translation remains largely unexplored. Distinct from dendritic RNA granules, axonal mRNA is often co-transported with membrane-bound organelles, such as late endosomes and mitochondria, for the long-range delivery of mRNA molecules to distal axons [4]. Moreover, such a transport complex is an important regulon to link local translation and axonal function. For example, recent studies showed that Pink1 mRNA, which encodes PTEN-induced putative kinase 1 (PINK1) for mitophagy of damaged mitochondria, is co-transported on mitochondria to axons, and its local translation is important to supply PINK1 for axonal mitophagy [82, 83]. Although axonal distribution of Slc17a6 transcript was not affected in CPEB2-deficient TA afferents (Fig. 6C), further investigation is needed to determine whether CPEB2 may dispatch other identified mRNA candidates (Fig. 2C and Additional File 4) or co-transport with other membrane-bound organelles to axons. Interestingly, several mRNAs encoding transcriptional regulators, such as NF-kB and CREB-regulated transcription coactivator 1 (CRTC1), were identified in the learning-induced axonal translatome (Additional File 3). Since de novo transcription is also required to support long-lasting LTP and memory, activity-dependent nuclear transport of postsynaptically localized transcriptional regulators is a means of linking synaptic stimulation to transcriptional changes [84,85,86,87]. Nevertheless, whether such an activity-coupled nuclear transport mechanism exists in the presynaptic compartment of mature mammalian neurons and whether activity-regulated local translation contributes to the level of presynaptically resident transcriptional regulators to affect learning and memory remain to be explored.

Our reanalysis of learning-associated translatome (GSE124592) revealed 2,438 axon-specifically translated mRNAs, whereas Ostroff et al. found 454 transcripts. The duplicate TRAP experiments were conducted by using 2 different GFP antibodies to immunoprecipitate transcripts from eYFP-L10a (TRAP)- or YFP (control)-expressing cortical and amygdala tissues [20]. Because different antibodies can lead to different immunoprecipitation backgrounds, we chose to analyze the data separately rather than together as in the previous study [20]. Slc17a6 was not identified previously because Slc17a6 signals in the two TRAP experiments varied widely to cause a non-significant P value [20]. Nevertheless, activity-induced axonal translation of Slc17a6 was experimentally demonstrated in this study, so the learning-associated axonal translatome appears to be more expanded and covers genes directly involved in presynaptic transmission. Besides CPEB2, FMR1 was recently shown to affect protein synthesis in hippocampal mossy fibers (MFs) and FMR1 deficiency impaired protein synthesis-dependent LTP in the MF-CA3 pathway [88], but which mRNAs are translationally regulated by FMR1 in MF afferents has not been identified. We therefore speculate that many other RBPs known to play a role in synaptic plasticity and memory may also be involved in presynaptic translation to build up the synaptic proteome for memory consolidation.

Conclusions

This study identified a local translation mechanism contributing to the regulation of presynaptic plasticity. CPEB2-controlled translation governs presynaptic glutamatergic transmission and LTP in transected SC and TA afferents. CPEB2 binds to 272 mRNAs whose axonal translation changes after learning. Moreover, activity- and CPEB2-dependent axonal translation of VGLUT2 mRNA replenishes synaptic vesicles and sustains presynaptic translation-dependent LTP. These findings provide evidence linking local translational control in the regulation of glutamatergic transmission and presynaptic plasticity.

Availability of data and materials

Microarray data can be viewed by the GEO accession GSE208738: please access https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE208738 . Analyzed gene expression profiles from GSE208738, learning-associated translatome (GSE124592) and in combination are listed in Additional files 2, 3 and 4. All materials are available from the corresponding author upon reasonable request.

Abbreviations

3′-UTR:

3′-Untreanslated region

4X HFS:

4 Trains of high-frequency stimulation

AAV:

Adeno-associated virus

AMPAR:

Alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid-type glutamate receptor

APV:

(2R)-amino-5-phosphonovaleric acid

BCA:

Bicinchoninic acid

BSA:

Bovine serum albumin

CA1:

Cornu ammonis 1

CA3:

Cornu ammonis 3

CaMK2α:

Calcium/calmodulin-dependent protein kinase 2 subunit alpha

CHX:

Cycloheximide

cKO:

Conditional knockout

CNQX:

Cyanquixaline

CPEB2:

Cytoplasmic polyadenylation element binding protein 2

CRTC1:

CREB regulated transcription coactivator 1

cWT:

Conditional wild type

DEPC:

Diethyl pyrocarbonate

DIV:

Days in vitro

DTT:

Dithiothreitol

EDTA:

Ethylenediamine tetraacetic acid

EGTA:

Ethylene glycol-bis(β-aminoethyl ether)-N,N,N′,N′-tetraacetic acid

EPM:

Elevated plus maze

FISH:

Fluorescence in situ hybridization

FMR1:

Fragile X messenger ribonucleoprotein 1

fEPSP:

Field excitatory postsynaptic potential

FV:

Fiber volley

GFP:

Green fluorescent protein

GO:

Gene ontology

GRASP1:

GRIP-associated protein 1

HEK293T:

Human embryonic kidney 293 T 

IRES:

Internal ribosome entry site

KO:

Knockout

LTP:

Long-term potentiation

MF:

Mossy fiber

MWM:

Morris water maze

N2a:

Neuro-2a

NMDA:

N-methyl-D-aspartate

PBS:

Phosphate buffered saline

PINK1:

PTEN-induced putative kinase 1

PPF:

Paired-pulse facilitation

RBP:

RNA-binding protein

RIP:

RNA immunoprecipitation

SC:

Schaffer collateral

SLM:

Stratum lacunosum moleculare

SM:

Stratum moleculare

SO:

Stratum oriens

SR:

Stratum radiatum

SYT1:

Synaptotagmin 1

TA:

Temporoammonic

TRAP:

Translating ribosome affinity purification

vg:

Viral genome

VGLUT:

Vesicular glutamate transporter

WT:

Wild type

References

  1. Flexner JB, Flexner LB, Stellar E. Memory in mice as affected by intracerebral puromycin. Science. 1963;141(3575):57–9.

    Article  CAS  PubMed  Google Scholar 

  2. Biever A, Donlin-Asp PG, Schuman EM. Local translation in neuronal processes. Curr Opin Neurobiol. 2019;57:141–8.

    Article  CAS  PubMed  Google Scholar 

  3. Richter JD, Klann E. Making synaptic plasticity and memory last: mechanisms of translational regulation. Genes Dev. 2009;23(1):1–11.

    Article  CAS  PubMed  Google Scholar 

  4. Dalla Costa I, Buchanan CN, Zdradzinski MD, Sahoo PK, Smith TP, Thames E, et al. The functional organization of axonal mRNA transport and translation. Nat Rev Neurosci. 2021;22(2):77–91.

    Article  CAS  PubMed  Google Scholar 

  5. Holt CE, Schuman EM. The central dogma decentralized: new perspectives on RNA function and local translation in neurons. Neuron. 2013;80(3):648–57.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Steward O, Levy WB. Preferential localization of polyribosomes under the base of dendritic spines in granule cells of the dentate gyrus. J Neurosci. 1982;2(3):284–91.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Miniaci MC, Kim JH, Puthanveettil SV, Si K, Zhu H, Kandel ER, et al. Sustained CPEB-dependent local protein synthesis is required to stabilize synaptic growth for persistence of long-term facilitation in Aplysia. Neuron. 2008;59(6):1024–36.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Si K, Giustetto M, Etkin A, Hsu R, Janisiewicz AM, Miniaci MC, et al. A neuronal isoform of CPEB regulates local protein synthesis and stabilizes synapse-specific long-term facilitation in aplysia. Cell. 2003;115(7):893–904.

    Article  CAS  PubMed  Google Scholar 

  9. Cagnetta R, Frese CK, Shigeoka T, Krijgsveld J, Holt CE. Rapid cue-specific remodeling of the Nascent Axonal Proteome. Neuron. 2018;99(1):29-46 e4.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Cagnetta R, Wong HH, Frese CK, Mallucci GR, Krijgsveld J, Holt CE. Noncanonical modulation of the eIF2 pathway controls an increase in local translation during neural wiring. Mol Cell. 2019;73(3):474-89 e5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Shigeoka T, Jung H, Jung J, Turner-Bridger B, Ohk J, Lin JQ, et al. Dynamic axonal translation in developing and mature visual circuits. Cell. 2016;166(1):181–92.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Shigeoka T, Koppers M, Wong HH, Lin JQ, Cagnetta R, Dwivedy A, et al. On-site ribosome remodeling by locally synthesized ribosomal proteins in axons. Cell Rep. 2019;29(11):3605-19 e10.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Hanz S, Perlson E, Willis D, Zheng JQ, Massarwa R, Huerta JJ, et al. Axoplasmic importins enable retrograde injury signaling in lesioned nerve. Neuron. 2003;40(6):1095–104.

    Article  CAS  PubMed  Google Scholar 

  14. Jung H, Yoon BC, Holt CE. Axonal mRNA localization and local protein synthesis in nervous system assembly, maintenance and repair. Nat Rev Neurosci. 2012;13(5):308–24.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Patel P, Buchanan CN, Zdradzinski MD, Sahoo PK, Kar AN, Lee SJ, et al. Intra-axonal translation of Khsrp mRNA slows axon regeneration by destabilizing localized mRNAs. Nucleic Acids Res. 2022;50(10):5772–92.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Twiss JL, Fainzilber M. Ribosomes in axons–scrounging from the neighbors? Trends Cell Biol. 2009;19(5):236–43.

    Article  CAS  PubMed  Google Scholar 

  17. Hafner AS, Donlin-Asp PG, Leitch B, Herzog E, Schuman EM. Local protein synthesis is a ubiquitous feature of neuronal pre- and postsynaptic compartments. Science. 2019;364(6441):eaau3644.

    Article  CAS  PubMed  Google Scholar 

  18. Poulopoulos A, Murphy AJ, Ozkan A, Davis P, Hatch J, Kirchner R, et al. Subcellular transcriptomes and proteomes of developing axon projections in the cerebral cortex. Nature. 2019;565(7739):356–60.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Biever A, Glock C, Tushev G, Ciirdaeva E, Dalmay T, Langer JD, et al. Monosomes actively translate synaptic mRNAs in neuronal processes. Science. 2020;367(6477):eaay4991.

    Article  CAS  PubMed  Google Scholar 

  20. Ostroff LE, Santini E, Sears R, Deane Z, Kanadia RN, LeDoux JE, et al. Axon TRAP reveals learning-associated alterations in cortical axonal mRNAs in the lateral amgydala. Elife. 2019;8:e51607.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Younts TJ, Monday HR, Dudok B, Klein ME, Jordan BA, Katona I, et al. Presynaptic protein synthesis is required for long-term plasticity of GABA release. Neuron. 2016;92(2):479–92.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Jung J, Ohk J, Kim H, Holt CE, Park HJ, Jung H. mRNA transport, translation, and decay in adult mammalian central nervous system axons. Neuron. 2023;111(5):650-68.e4.

    Article  CAS  PubMed  Google Scholar 

  23. Gonzalez C, Canovas J, Fresno J, Couve E, Court FA, Couve A. Axons provide the secretory machinery for trafficking of voltage-gated sodium channels in peripheral nerve. Proc Natl Acad Sci U S A. 2016;113(7):1823–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Merianda TT, Lin AC, Lam JS, Vuppalanchi D, Willis DE, Karin N, et al. A functional equivalent of endoplasmic reticulum and Golgi in axons for secretion of locally synthesized proteins. Mol Cell Neurosci. 2009;40(2):128–42.

    Article  CAS  PubMed  Google Scholar 

  25. Yin HH, Davis MI, Ronesi JA, Lovinger DM. The role of protein synthesis in striatal long-term depression. J Neurosci. 2006;26(46):11811–20.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Chao HW, Tsai LY, Lu YL, Lin PY, Huang WH, Chou HJ, et al. Deletion of CPEB3 enhances hippocampus-dependent memory via increasing expressions of PSD95 and NMDA receptors. J Neurosci. 2013;33(43):17008–22.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Fioriti L, Myers C, Huang YY, Li X, Stephan JS, Trifilieff P, et al. The persistence of Hippocampal-based memory requires protein synthesis mediated by the prion-like protein CPEB3. Neuron. 2015;86(6):1433–48.

    Article  CAS  PubMed  Google Scholar 

  28. Huang YS, Jung MY, Sarkissian M, Richter JD. N-methyl-D-aspartate receptor signaling results in Aurora kinase-catalyzed CPEB phosphorylation and alpha CaMKII mRNA polyadenylation at synapses. EMBO J. 2002;21(9):2139–48.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Huang YS, Mendez R, Fernandez M, Richter JD. CPEB and translational control by cytoplasmic polyadenylation: impact on synaptic plasticity, learning, and memory. Mol Psychiatry. 2023;28(7):2728–36.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Lu WH, Chao HW, Lin PY, Lin SH, Liu TH, Chen HW, et al. CPEB3-dowregulated Nr3c1 mRNA translation confers resilience to developing posttraumatic stress disorder-like behavior in fear-conditioned mice. Neuropsychopharmacology. 2021;46(9):1669–79.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Chiu SL, Diering GH, Ye B, Takamiya K, Chen CM, Jiang Y, et al. GRASP1 regulates synaptic plasticity and learning through endosomal recycling of AMPA receptors. Neuron. 2017;93(6):1405-19 e8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Parras A, Anta H, Santos-Galindo M, Swarup V, Elorza A, Nieto-Gonzalez JL, et al. Autism-like phenotype and risk gene mRNA deadenylation by CPEB4 mis-splicing. Nature. 2018;560(7719):441–6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Tseng CS, Chao HW, Huang HS, Huang YS. Olfactory-experience- and developmental-stage-dependent control of CPEB4 regulates c-Fos mRNA translation for granule cell survival. Cell Rep. 2017;21(8):2264–76.

    Article  CAS  PubMed  Google Scholar 

  34. Wu L, Wells D, Tay J, Mendis D, Abbott MA, Barnitt A, et al. CPEB-mediated cytoplasmic polyadenylation and the regulation of experience-dependent translation of alpha-CaMKII mRNA at synapses. Neuron. 1998;21(5):1129–39.

    Article  CAS  PubMed  Google Scholar 

  35. Lu WH, Yeh NH, Huang YS. CPEB2 activates GRASP1 mRNA translation and promotes AMPA receptor surface expression, long-term potentiation, and memory. Cell Rep. 2017;21(7):1783–94.

    Article  CAS  PubMed  Google Scholar 

  36. Kundel M, Jones KJ, Shin CY, Wells DG. Cytoplasmic polyadenylation element-binding protein regulates neurotrophin-3-dependent beta-catenin mRNA translation in developing hippocampal neurons. J Neurosci. 2009;29(43):13630–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Alarcon JM, Hodgman R, Theis M, Huang YS, Kandel ER, Richter JD. Selective modulation of some forms of schaffer collateral-CA1 synaptic plasticity in mice with a disruption of the CPEB-1 gene. Learn Mem. 2004;11(3):318–27.

    Article  PubMed  PubMed Central  Google Scholar 

  38. Tsai LY, Chang YW, Lin PY, Chou HJ, Liu TJ, Lee PT, et al. CPEB4 knockout mice exhibit normal hippocampus-related synaptic plasticity and memory. PLoS ONE. 2013;8(12):e84978.

    Article  PubMed  PubMed Central  Google Scholar 

  39. Satake S, Inoue T, Imoto K. Paired-pulse facilitation of multivesicular release and intersynaptic spillover of glutamate at rat cerebellar granule cell-interneurone synapses. J Physiol. 2012;590(22):5653–75.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Lai YT, Chao HW, Lai AC, Lin SH, Chang YJ, Huang YS. CPEB2-activated PDGFRalpha mRNA translation contributes to myofibroblast proliferation and pulmonary alveologenesis. J Biomed Sci. 2020;27(1):52.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Lai YT, Su CK, Jiang ST, Chang YJ, Lai AC, Huang YS. Deficiency of CPEB2-confined Choline Acetyltransferase expression in the dorsal motor nucleus of Vagus causes hyperactivated parasympathetic signaling-associated bronchoconstriction. J Neurosci. 2016;36(50):12661–76.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Aihara Y, Mashima H, Onda H, Hisano S, Kasuya H, Hori T, et al. Molecular cloning of a novel brain-type Na(+)-dependent inorganic phosphate cotransporter. J Neurochem. 2000;74(6):2622–5.

    Article  CAS  PubMed  Google Scholar 

  43. Fremeau RT Jr, Troyer MD, Pahner I, Nygaard GO, Tran CH, Reimer RJ, et al. The expression of vesicular glutamate transporters defines two classes of excitatory synapse. Neuron. 2001;31(2):247–60.

    Article  CAS  PubMed  Google Scholar 

  44. Herzog E, Bellenchi GC, Gras C, Bernard V, Ravassard P, Bedet C, et al. The existence of a second vesicular glutamate transporter specifies subpopulations of glutamatergic neurons. J Neurosci. 2001;21(22):RC181.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Giusti SA, Vercelli CA, Vogl AM, Kolarz AW, Pino NS, Deussing JM, et al. Behavioral phenotyping of Nestin-Cre mice: implications for genetic mouse models of psychiatric disorders. J Psychiatr Res. 2014;55:87–95.

    Article  PubMed  Google Scholar 

  46. Tronche F, Kellendonk C, Kretz O, Gass P, Anlag K, Orban PC, et al. Disruption of the glucocorticoid receptor gene in the nervous system results in reduced anxiety. Nat Genet. 1999;23(1):99–103.

    Article  CAS  PubMed  Google Scholar 

  47. Qi X, Zhang K, Xu T, Yamaki VN, Wei Z, Huang M, et al. Sex differences in long-term potentiation at Temporoammonic-CA1 synapses: potential implications for memory consolidation. Plos One. 2016;11(11):e0165891.

    Article  PubMed  PubMed Central  Google Scholar 

  48. DeNardo LA, de Wit J, Otto-Hitt S, Ghosh A. NGL-2 regulates input-specific synapse development in CA1 pyramidal neurons. Neuron. 2012;76(4):762–75.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  49. Spillane M, Ketschek A, Merianda TT, Twiss JL, Gallo G. Mitochondria coordinate sites of axon branching through localized intra-axonal protein synthesis. Cell Rep. 2013;5(6):1564–75.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Chen CY, Chen YT, Wang JY, Huang YS, Tai CY. Postsynaptic Y654 dephosphorylation of beta-catenin modulates presynaptic vesicle turnover through increased n-cadherin-mediated transsynaptic signaling. Dev Neurobiol. 2017;77(1):61–74.

    Article  CAS  PubMed  Google Scholar 

  51. Kay AR, Alfonso A, Alford S, Cline HT, Holgado AM, Sakmann B, et al. Imaging synaptic activity in intact brain and slices with FM1-43 in C. elegans, lamprey, and rat. Neuron. 1999;24(4):809–17.

    Article  CAS  PubMed  Google Scholar 

  52. Kuzniewska B, Cysewski D, Wasilewski M, Sakowska P, Milek J, Kulinski TM, et al. Mitochondrial protein biogenesis in the synapse is supported by local translation. EMBO Rep. 2020;21(8):e48882.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Zhou Y, Zhou B, Pache L, Chang M, Khodabakhshi AH, Tanaseichuk O, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun. 2019;10(1):1523.

    Article  PubMed  PubMed Central  Google Scholar 

  54. Thomson AM. Facilitation, augmentation and potentiation at central synapses. Trends Neurosci. 2000;23(7):305–12.

    Article  CAS  PubMed  Google Scholar 

  55. Raymond CR. LTP forms 1, 2 and 3: different mechanisms for the “long” in long-term potentiation. Trends Neurosci. 2007;30(4):167–75.

    Article  CAS  PubMed  Google Scholar 

  56. Heiman M, Kulicke R, Fenster RJ, Greengard P, Heintz N. Cell type-specific mRNA purification by translating ribosome affinity purification (TRAP). Nat Protoc. 2014;9(6):1282–91.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  57. Martinez-Hernandez A, Bell KP, Norenberg MD. Glutamine synthetase: glial localization in brain. Science. 1977;195(4284):1356–8.

    Article  CAS  PubMed  Google Scholar 

  58. Luo P, Li X, Fei Z, Poon W. Scaffold protein Homer 1: implications for neurological diseases. Neurochem Int. 2012;61(5):731–8.

    Article  CAS  PubMed  Google Scholar 

  59. Moga DE, Janssen WG, Vissavajjhala P, Czelusniak SM, Moran TM, Hof PR, et al. Glutamate receptor subunit 3 (GluR3) immunoreactivity delineates a subpopulation of parvalbumin-containing interneurons in the rat hippocampus. J Comp Neurol. 2003;462(1):15–28.

    Article  PubMed  Google Scholar 

  60. Pachernegg S, Strutz-Seebohm N, Hollmann M. GluN3 subunit-containing NMDA receptors: not just one-trick ponies. Trends Neurosci. 2012;35(4):240–9.

    Article  CAS  PubMed  Google Scholar 

  61. Sala C, Piech V, Wilson NR, Passafaro M, Liu G, Sheng M. Regulation of dendritic spine morphology and synaptic function by Shank and Homer. Neuron. 2001;31(1):115–30.

    Article  CAS  PubMed  Google Scholar 

  62. Irie F, Badie-Mahdavi H, Yamaguchi Y. Autism-like socio-communicative deficits and stereotypies in mice lacking heparan sulfate. Proc Natl Acad Sci U S A. 2012;109(13):5052–6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  63. Lin X, Wei G, Shi Z, Dryer L, Esko JD, Wells DE, et al. Disruption of gastrulation and heparan sulfate biosynthesis in EXT1-deficient mice. Dev Biol. 2000;224(2):299–311.

    Article  CAS  PubMed  Google Scholar 

  64. Fremeau RT Jr, Kam K, Qureshi T, Johnson J, Copenhagen DR, Storm-Mathisen J, et al. Vesicular glutamate transporters 1 and 2 target to functionally distinct synaptic release sites. Science. 2004;304(5678):1815–9.

    Article  CAS  PubMed  Google Scholar 

  65. Herzog E, Takamori S, Jahn R, Brose N, Wojcik SM. Synaptic and vesicular co-localization of the glutamate transporters VGLUT1 and VGLUT2 in the mouse hippocampus. J Neurochem. 2006;99(3):1011–8.

    Article  CAS  PubMed  Google Scholar 

  66. Bouazza-Arostegui B, Camacho M, Brockmann MM, Zobel S, Rosenmund C. Deconstructing Synaptotagmin-1’s distinct roles in synaptic vesicle priming and neurotransmitter release. J Neurosci. 2022;42(14):2856–71.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  67. Bouhours B, Gjoni E, Kochubey O, Schneggenburger R. Synaptotagmin2 (Syt2) drives fast release redundantly with Syt1 at the output synapses of Parvalbumin-expressing inhibitory neurons. J Neurosci. 2017;37(17):4604–17.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  68. Tang J, Maximov A, Shin OH, Dai H, Rizo J, Sudhof TC. A complexin/synaptotagmin 1 switch controls fast synaptic vesicle exocytosis. Cell. 2006;126(6):1175–87.

    Article  CAS  PubMed  Google Scholar 

  69. Halasy K, Hajszan T, Kovacs EG, Lam TT, Leranth C. Distribution and origin of vesicular glutamate transporter 2-immunoreactive fibers in the rat hippocampus. Hippocampus. 2004;14(7):908–18.

    Article  PubMed  Google Scholar 

  70. He H, Mahnke AH, Doyle S, Fan N, Wang CC, Hall BJ, et al. Neurodevelopmental role for VGLUT2 in pyramidal neuron plasticity, dendritic refinement, and in spatial learning. J Neurosci. 2012;32(45):15886–901.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  71. Taylor AM, Blurton-Jones M, Rhee SW, Cribbs DH, Cotman CW, Jeon NL. A microfluidic culture platform for CNS axonal injury, regeneration and transport. Nat Methods. 2005;2(8):599–605.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  72. Tsien JZ, Chen DF, Gerber D, Tom C, Mercer EH, Anderson DJ, et al. Subregion- and cell type-restricted gene knockout in mouse brain. Cell. 1996;87(7):1317–26.

    Article  CAS  PubMed  Google Scholar 

  73. Castle MJ, Turunen HT, Vandenberghe LH, Wolfe JH. Controlling AAV tropism in the nervous system with natural and engineered capsids. Methods Mol Biol. 2016;1382:133–49.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  74. Aschauer DF, Kreuz S, Rumpel S. Analysis of transduction efficiency, tropism and axonal transport of AAV serotypes 1, 2, 5, 6, 8 and 9 in the mouse brain. PLoS ONE. 2013;8(9):e76310.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  75. Cearley CN, Wolfe JH. Transduction characteristics of adeno-associated virus vectors expressing cap serotypes 7, 8, 9, and Rh10 in the mouse brain. Mol Ther. 2006;13(3):528–37.

    Article  CAS  PubMed  Google Scholar 

  76. Klein RL, Dayton RD, Tatom JB, Henderson KM, Henning PP. Aav8, 9, rh10, rh43 vector gene transfer in the rat brain: effects of serotype, promoter and purification method. Mol Ther. 2008;16(1):89–96.

    Article  CAS  PubMed  Google Scholar 

  77. Moechars D, Weston MC, Leo S, Callaerts-Vegh Z, Goris I, Daneels G, et al. Vesicular glutamate transporter VGLUT2 expression levels control quantal size and neuropathic pain. J Neurosci. 2006;26(46):12055–66.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  78. Wallen-Mackenzie A, Gezelius H, Thoby-Brisson M, Nygard A, Enjin A, Fujiyama F, et al. Vesicular glutamate transporter 2 is required for central respiratory rhythm generation but not for locomotor central pattern generation. J Neurosci. 2006;26(47):12294–307.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  79. Turimella SL, Bedner P, Skubal M, Vangoor VR, Kaczmarczyk L, Karl K, et al. Characterization of cytoplasmic polyadenylation element binding 2 protein expression and its RNA binding activity. Hippocampus. 2015;25(5):630–42.

    Article  CAS  PubMed  Google Scholar 

  80. Dallerac G, Zapata J, Rouach N. Versatile control of synaptic circuits by astrocytes: where, when and how? Nat Rev Neurosci. 2018;19(12):729–43.

    Article  CAS  PubMed  Google Scholar 

  81. Huang YS, Carson JH, Barbarese E, Richter JD. Facilitation of dendritic mRNA transport by CPEB. Genes Dev. 2003;17(5):638–53.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  82. Harbauer AB, Hees JT, Wanderoy S, Segura I, Gibbs W, Cheng Y, et al. Neuronal mitochondria transport Pink1 mRNA via synaptojanin 2 to support local mitophagy. Neuron. 2022;110(9):1516-31 e9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  83. Harbauer AB, Schwarz TL. Mitochondrial hitch-hiking of Pink1 mRNA supports axonal mitophagy. Autophagy. 2022;18(12):3048–9.

  84. Meffert MK, Chang JM, Wiltgen BJ, Fanselow MS, Baltimore D. NF-kappa B functions in synaptic signaling and behavior. Nat Neurosci. 2003;6(10):1072–8.

    Article  CAS  PubMed  Google Scholar 

  85. Uchida S, Shumyatsky GP. Synaptically localized transcriptional regulators in memory formation. Neuroscience. 2018;370:4–13.

    Article  CAS  PubMed  Google Scholar 

  86. Uchida S, Teubner BJW, Hevi C, Hara K, Kobayashi A, Dave RM, et al. CRTC1 nuclear translocation following learning modulates memory strength via exchange of chromatin remodeling complexes on the Fgf1 gene. Cell Rep. 2017;18(2):352–66.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  87. Ch’ng TH, Uzgil B, Lin P, Avliyakulov NK, O’Dell TJ, Martin KC. Activity-dependent transport of the transcriptional coactivator CRTC1 from synapse to nucleus. Cell. 2012;150(1):207–21.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  88. Monday HR, Kharod SC, Yoon YJ, Singer RH, Castillo PE. Presynaptic FMRP and local protein synthesis support structural and functional plasticity of glutamatergic axon terminals. Neuron. 2022;110(16):2588-606 e6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We thank the DNA Sequencing Core Facility and the AAV Core Facility, supported by the Academia Sinica Projects (AS-CFII-111-211 and AS-CFII112-204), and the institutional Electrophysiology Core for the instruments. We thank Hsiu-Chen Lin, Hsu-Wen Chao, Chin-Yi Chen, Yu-Wen Chao and Shu-Ping Lee for assistance with RNA immunoprecipitation, synaptosome isolation, FM4-64FX loading experiments, neuronal culture purity and the use of AxioImager Z1, respectively.

Funding

National Science and Technology Council (NSTC, Taiwan [108–2320-B-001–020-MY3 and 111–2311-B-001–020-MY3] and Academia Sinica [AS-GC-111-L03]. W.-H.L. was supported by a postdoctoral fellowship from NSTC (2020–2021, 110–2811-B-001–569) and Academia Sinica (2022–2023).

Author information

Authors and Affiliations

  1. Institute of Biomedical Sciences, Academia Sinica, 128 Sec. 2, Academia Rd., Taipei, 11529, Taiwan

    Wen-Hsin Lu, Tzu-Tung Chang, Yao-Ming Chang, Yi-Hsiang Liu, Chia-Hsuan Lin, Ching-Shu Suen, Ming-Jing Hwang & Yi-Shuian Huang

  2. Taiwan International Graduate Program in Interdisciplinary Neuroscience, National Yang-Ming Chao-Tung University and Academia Sinica, Taipei, 11529, Taiwan

    Chia-Hsuan Lin & Yi-Shuian Huang

  3. Neuroscience Program of Academia Sinica, Academia Sinica, Taipei, 11529, Taiwan

    Yi-Shuian Huang

Authors
  1. Wen-Hsin Lu
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Tzu-Tung Chang
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Yao-Ming Chang
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Yi-Hsiang Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Chia-Hsuan Lin
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Ching-Shu Suen
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Ming-Jing Hwang
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Yi-Shuian Huang
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

W.-H.L. designed and conducted the experiments and analyzed data. Y.-M.C., C.-S.S. and M.-J.H. performed bioinformatics analysis, T.-T.C and Y.-H.L helped with slice recording, immunoblotting, immunostaining and stereotaxic injection, and C.-H.L. conducted reporter assays. Y.-S.H. designed and supervised the study, and co-wrote the manuscript with W.-H.L. and is responsible for its content.

Corresponding author

Correspondence to Yi-Shuian Huang.

Ethics declarations

Ethics approval and consent to participate

All animal experimental procedures were approved by the Institutional Animal Care and Utilization Committee.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Lu, WH., Chang, TT., Chang, YM. et al. CPEB2-activated axonal translation of VGLUT2 mRNA promotes glutamatergic transmission and presynaptic plasticity. J Biomed Sci 31, 69 (2024). https://doi.org/10.1186/s12929-024-01061-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12929-024-01061-2

Keywords

Journal of Biomedical Science

ISSN: 1423-0127