Cover ArticleResearch Articles, Neurobiology of Disease

Functionally Clustered mRNAs Are Distinctly Enriched at Cortical Astroglial Processes and Are Preferentially Affected by FMRP Deficiency

Yuqin Men, Haruki Higashimori, Kathryn Reynolds, Leona Tu, Rachel Jarvis and Yongjie Yang
Journal of Neuroscience 20 July 2022, 42 (29) 5803-5814; DOI: https://doi.org/10.1523/JNEUROSCI.0274-22.2022

Abstract

Mature protoplasmic astroglia in the mammalian CNS uniquely possess a large number of fine processes that have been considered primary sites to mediate astroglia to neuron synaptic signaling. However, localized mechanisms for regulating interactions between astroglial processes and synapses, especially for regulating the expression of functional surface proteins at these fine processes, are largely unknown. Previously, we showed that the loss of the RNA binding protein FMRP in astroglia disrupts astroglial mGluR5 signaling and reduces expression of the major astroglial glutamate transporter GLT1 and glutamate uptake in the cortex of Fmr1 conditional deletion mice. In the current study, by examining ribosome localization using electron microscopy and identifying mRNAs enriched at cortical astroglial processes using synaptoneurosome/translating ribosome affinity purification and RNA-Seq in WT and FMRP-deficient male mice, our results reveal interesting localization-dependent functional clusters of mRNAs at astroglial processes. We further showed that the lack of FMRP preferentially alters the subcellular localization and expression of process-localized mRNAs. Together, we defined the role of FMRP in altering mRNA localization and expression at astroglial processes at the postnatal development (P30-P40) and provided new candidate mRNAs that are potentially regulated by FMRP in cortical astroglia.

SIGNIFICANCE STATEMENT Localized mechanisms for regulating interactions between astroglial processes and synapses, especially for regulating the expression of functional surface proteins at these fine processes, are largely unknown. Previously, we showed that the loss of the RNA binding protein FMRP in astroglia disrupts expression of several astroglial surface proteins, such as mGluR5 and major astroglial glutamate transporter GLT1 in the cortex of FMRP-deficient mice. Our current study examined ribosome localization using electron microscopy and identified mRNAs enriched at cortical astroglial processes in WT and FMRP-deficient mice. These results reveal interesting localization-dependent functional clusters of mRNAs at astroglial processes and demonstrate that the lack of FMRP preferentially alters the subcellular localization and expression of process-localized mRNAs.

Introduction

Mature protoplasmic astroglia in the mammalian CNS uniquely possess a large number of fine processes, often termed perisynaptic astroglial processes (PAPs), which are considered primary sites to mediate astroglia to neuron synaptic signaling (Derouiche et al., 2002). The PAP or “microdomain” structure also makes it feasible for a single astroglia to monitor and modulate heterogeneous synaptic activity remotely from its cell body (Grosche et al., 1999). Functional receptors, transporters, and channels are frequently found on the surface of PAPs that are imperative for homeostatic and evoked synaptic signaling in CNS physiology (Barres, 2008). In parallel, changes in astroglial Ca2+ levels, a manifestation of astroglial activity, occur much more frequently at PAPs/microdomains (Nimmerjahn et al., 2009; Kanemaru et al., 2014) and tend to have faster kinetics (Tang et al., 2015), which is considered more likely to be coupled with neuronal signaling (Bazargani and Attwell, 2016). While localized protein synthesis in remote neuronal dendritic spines is a well-established mechanism contributing to synaptic plasticity (Sutton and Schuman, 2006), localized mechanisms for regulating PAP to synapse interactions, especially for regulating the expression of functional surface proteins at PAPs, is largely unknown. Recent studies have begun to demonstrate the presence of ribosomes and low but detectable localized mRNA translation at astroglial processes (Sakers et al., 2017; Mazare et al., 2020). Such mRNAs that are localized or enriched at astroglial processes have also begun to be identified (Sakers et al., 2017; Mazare et al., 2020). However, how the localization and expression of such mRNAs at astroglial processes are regulated remains essentially unknown. In addition, whether the localization and expression of such mRNAs at astroglial processes are altered in pathologic conditions, especially with altered synaptic signaling, also remains unexplored.

Fragile X syndrome (FXS), one of the most common inherited intellectual disabilities, is caused by the CGG expansion-induced loss of FMRP protein expression (Verkerk et al., 1991) and is characterized by cognitive impairment, sensory hyperactivity/seizure, and autistic features, such as repetitive behaviors (Hagerman et al., 2009). Extensive studies in cultured neurons and brains of Fmr1 KO mice have shown that FMRP primarily functions as an RNA binding protein (RBP) that associates with many (>800) mRNAs in the brain and potentially regulates their translation (Darnell et al., 2011; Richter and Zhao, 2021). As a result of the loss of FMRP expression, altered synaptic plasticity and enhanced neuronal activity have been widely observed in cortex, hippocampus, and cerebellum of Fmr1 KO mice (Contractor et al., 2015), potentially contributing to typical FXS phenotypes. In addition to the loss of FMRP in neurons, the loss of FMRP in astroglia has also been shown to contribute to FXS pathogenesis. FMRP-deficient astroglia are capable of inducing abnormal dendritic morphology of WT hippocampal neurons in cocultures (Nimchinsky et al., 2001). Loss of FMRP in astroglia also disrupts astroglial mGluR5 signaling (Men et al., 2020) and reduces expression of the major astroglial glutamate transporter GLT1 and glutamate uptake in the cortex of Fmr1 KO mice (Higashimori et al., 2016), and exhibits several FXS-relevant behavior changes (Jin et al., 2021). As an RBP that is expressed in astroglia, whether FMRP is involved in regulating the localization and expression of mRNAs at astroglial processes is unknown. In the current study, we investigated local translation machinery in cortical astroglial processes in FMRP-deficient mice and defined the involvement of FMRP in altering mRNA localization and expression, especially at astroglial processes during postnatal development.

Materials and Methods

Animals

Fmr1-/y (FVB.129P2(B6)-Fmr1tm1Cgr/J; #003024) and WT FVB mice (#001800) were obtained from The Jackson Laboratory. Bacterial artificial chromosome Aldh1l1-eGFP-L10 (also designated as Aldh1l1-translating ribosome affinity purification [TRAP]) transgenic mice were obtained from the Gene Expression Nervous System Atlas project (Rockefeller University) through The Jackson Laboratory (#030247). Eaat2-tdT mice were generated as previously described (Yang et al., 2011). Only males were used for this study because FXS is an X-linked disorder and males are more severely affected by this disorder than females. Care and treatment of animals in all procedures strictly followed the National Institutes of Health's Guide for the care and use of laboratory animals, the Guidelines for the use of animals in neuroscience research, and the Tufts University Institutional Animal Care and Use Committee.

Immuno-electron microscopy (immuno-EM)

Immuno-EM was conducted in the Harvard Medical School EM facility. Bacterial artificial chromosome Aldh1l1-TRAP mice were anesthetized with ketamine (100 mg/kg) and xylazine (10 mg/kg) in saline by intraperitoneal injection. After anesthetization, the mice were intracardially perfused with ice-cold 0.1 m sodium PB, pH 7.4, followed by a mixture of 4% formaldehyde and 0.1% glutaraldehyde in 0.1 m sodium PB, pH 7.4, for 10 min. The brain was dissected out and postfixed in 4% formaldehyde overnight at 4°C, and brain slices (100 µm) were prepared using a vibratome. The slices were then quenched, permeabilized, and blocked in blocking buffer (3% BSA, 5% normal donkey serum, and 0.1% Triton X-100) at 4°C. Anti-GFP (Abcam, #6556) antibody was then added and incubated overnight at 4°C. After wash, slices were incubated with Protein A-gold for 1 h at 25°C. The images were taken using the JEOL 1200EX transmission electron microscope.

Synaptoneurosome (SNS) preparation

SNSs were prepared from somatosensory cortical sections (2 mm) from Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice using a mouse brain matrix, as described previously (Rao and Steward, 1991). All steps were performed on ice or at 4°C. The somatosensory cortices were homogenized in homogenization buffer (20 mm HEPES, pH 7.4, 0.32 m sucrose, 2 mm MgCl2, protease and RNase-inhibitors) with the Dounce homogenizer (10 gentle strokes). Cell debris and nuclei were removed by centrifugation at 1000 × g for 5 min at 4°C. The supernatant was centrifuged at 1300 × g for 5 min and then transferred and further centrifuged for another 10 min at 14,000 × g. The pellet was resuspended carefully with 2-2.5 ml homogenization buffer and loaded on top of a preprepared cold Ficoll-sucrose density gradient (starting from the bottom of the tube with 16%, 13%, and 5%, prepared in homogenization buffer with protease and RNase inhibitors) and spun (SW 41 rotor) at 40,000 × g for 45 min. The white band at the third density (counted from the top, between 13% and 16% Ficoll-sucrose gradient) was collected as the SNS and resuspended in 10 ml preincubation buffer (25 mm HEPES, pH 7.4, 20 mm glucose, 3.5 mm KCl, 1.2 mm Na2HPO4, 2 mm MgCl2, 129 mm NaCl with protease and RNase inhibitors) and centrifuged at 2000 × g for 10 min at 4°C. SNS was resuspended into the TRAP homogenization buffer before proceeding to TRAP isolation procedures.

TRAP isolation of RNA

Ribosome-bound mRNAs were isolated from somatosensory cortical sections (1-2 mm) or SNS of Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice (P40) as previously described (Doyle et al., 2008). Briefly, cortex homogenates and SNSs were mixed with the eGFP antibody (mouse clone HtzGFP-19C8, Memorial Sloan Kettering Cancer Center)-coupled beads (Dynabeads M270, Invitrogen) for immunoprecipitation. Following immunoprecipitation, RNA was isolated using TRIzol reagent (Thermo Fisher Scientific), according to the manufacturer's instructions. Total RNA was also directly isolated from the somatosensory cortex tissue (10 mg) using TRIzol reagent. The quality and quantity of all isolated RNA were analyzed with the Agilent 2100 Bioanalyzer system.

Preparation of RNA sequencing libraries and RNA-Seq analysis

Preparation of RNA sequencing libraries and sequencing was conducted by the Tufts University Core Facility Genomics facility. Briefly, the TruSeq Stranded Total RNA with RiboZero Gold (Illumina) was used in library preparation. The 100 bp paired-end sequencing was conducted on an Illumina HiSEq 2500 Sequencer. RNA-seq libraries were assigned randomly to sequencing lanes to avoid lane bias. The FastQC program was used to examine the quality of generated reads. For mouse samples, we usually obtained 20-30 million reads for each library with a Phred quality score >32. The total number of mapped reads for each sample was at least 60 million, sufficient for reliable calculations of counts per million mapped reads (CPM) values and subsequent differential gene expression analysis. For RNA-seq analysis, reads were aligned to the reference mouse genome (GRCm38/mm10) using STAR (version 2.6.0) software. RSEM (version 1.3) was used to quantify transcripts from mapped reads, and the CPM values were generated and normalized in EdgeR (version 3.5). The genes with an averaged CPM >1 were considered detectable transcripts. EdgeR was also used to identify significant changes in mRNA expression among different samples using the biological replicates with a default q value (FDR-adjusted p value, Padj) cutoff of 0.05 (q < 0.05).

Brain slice preparation and puromycin incorporation assay

Brain slices were prepared from Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice (P40 age). Following the cervical decapitation, the cortex was then quickly removed and 300 µm cortical slices were cut using a vibratome (Leica VT1000, Leica Microsystems) in iced aCSF in mm as follows: 2.5 KCl, 125 NaCl, 25, NaHCO3, 1.25 NaH2PO4, 10 D+ glucose, 2 CaCl2, 1 MgCl2, with osmolarity at 300-305 mOsm, equilibrated with 95% oxygen (O2)−5% CO2. The brain slices were recovered at room temperature for 1 h. After recovering, slices were incubated with 3 µm puromycin diluted in aCSF for 10 min and then immediately fixed with 4% PFA overnight. The slices were then cryoprotected with 30% sucrose for 4 h and sectioned (20 µm) with a cryostat. Immunostaining was performed using anti-puromycin antibody (clone 3RH11, Kerafast #EQ0001) to detect puromycin.

Immunohistochemistry and confocal imaging

Mice were deeply anesthetized with ketamine (100 mg/kg) + xylazine (10 mg/kg) in saline by intraperitoneal injection and perfused intracardially with 4% PFA in 1 × PBS. The brains were dissected and kept in 4% PFA overnight at 4°C, then cryoprotected by immersion in 30% sucrose for 24-48 h. Brains were embedded and frozen in OCT-Compound Tissue-Tek (Sakura). Coronal sections (10-20 µm) were prepared with a cryostat (Leica HM525). No immunostaining was performed to amplify eGFP and tdT fluorescence signals. For detecting puromycin signals, anti-puromycin antibody was applied overnight at 4°C in blocking buffer following incubation with the blocking buffer (1% BSA, 5% goat serum, and 0.1% Triton X-100 in 1 × PBS). After washing slides 3 times in PBS, corresponding secondary antibody (1:1000) was added for 120 min at room temperature (RT). For FMRP immunostaining, slides with sections were dried at 32.5°C for 30 min, briefly washed in PBS, then incubated in sodium citrate antigen retrieval buffer (0.01 m, pH 6.0) for 10 min at RT followed by an additional 5 min at 85°C. Sections were then washed 3 times with PBS and blocked (1% BSA, 5% NGS, 0.2% Triton X-100 in PBS) for 1 h at RT. Anti-FMRP antibody (mouse 2F5 clone, Developmental Studies Hybridoma Bank, 1:1 in blocking buffer) was applied overnight at 4°C. The next day, sections were washed 3 times with PBS, incubated with AlexaFluor-488 secondary antibody (goat anti-mouse, 1:1000 in blocking buffer, Invitrogen) for 2 h at RT, rinsed in PBS and dH2O, and coverslipped using ProLong Gold Antifade Mountant with DAPI (Thermo Fisher Scientific). All images were obtained under 40× or 63× oil immersion objective lens with a confocal laser scanning microscope (15-20 µm Z stack with 1 µm step, A1R Nikon or Leica SP8 Falcon).

Image analysis and quantification

For the quantification of eGFP+ immunogold particles within astroglial processes, EM images were analyzed in ImageJ. Only those synapses that had a >100 nm synapse interface (determined by postsynaptic density line) were >500 nm away from the edge of the image, and contained clear synaptic vesicles at the presynaptic terminal were included in quantification. A 600-nm-radius circle was drawn from the center of the synapse (middle of the PSD line), and the number of eGFP+ immunogold signals in each circle was counted. For the analysis of astroglial domains, 3D reconstruction of tdT+ astroglia from confocal Z-stack images (40× or 63×) was performed using the surface tool in Imaris (version 9.7, Bitplane). This function uses an automatic smoothing of the image with the Gaussian filter. The tdT fluorescence negative area in each of the confocal stack images was used as the internal control to determine the background fluorescence. The sensitivity threshold (absolute intensity) was manually adjusted so that the generated astroglial domain in the 3D image best matched with that in the original confocal image. The cell somas were then detected based on size (≥6 µm in diameter) and used as seeding points to build the 3D domains. The quality (intensity) threshold was also manually adjusted to ensure the optimal detection of cell somas in a given image. Cells that were only partially included in confocal and 3D images were excluded from analysis, as well as cells that were difficult to separate in confocal images were also excluded. Overlapped domains were also carefully examined to ensure accuracy. The volume size of individual astroglia can be directly measured from generated 3D domains in Imaris. For the quantification of ISH signals and eGFP-L10 fluorescence at astroglial processes and in soma, Glt1 or Epha4 in situ signals or eGFP fluorescence was measured inside the full astroglial 3D domains as well as inside the soma-only 3D domains, and the Glt1 or Epha4 in situ signal or eGFP fluorescence intensity in the astroglial processes was calculated by subtracting soma-only intensity from the full astroglial domain intensity for each individual astroglia.

ISH

Animals were deeply anesthetized with ketamine (100 mg/kg) and xylazine (10 mg/kg) in saline by intraperitoneal injection and perfused intracardially with 4% PFA in 1× PBS. The brains from Fmr1+/yEaat2-tdT+ and Fmr1-/yEaat2-tdT+ mice were postfixed in 4% PFA overnight at 4°C, and then cryoprotected by immersion in 30% sucrose for 48 h or until the tissue sank to the bottom. The brain tissues were frozen in the OCT embedding media on dry ice. Frozen brain tissues were then sectioned at 10 µm thickness using a cryostat (Leica HM525). After sectioning, ISH was performed following the instruction of the RNAscope fluorescent multiplex assay (Advanced Cell Diagnostics). Briefly, the slides with brain sections were washed in 1× PBS for 10 min, boiled in 1× target retrieval solution for 5 min, then immediately transferred into distilled water and washed by moving the rack up and down 5 times. The slides were then washed in fresh 100% ethanol by moving the rack up and down 5 times. After air drying, Protease IV was added on each section and then washed with fresh distilled water. Next, the sections were hybridized with RNAscope probes (1×) overnight at 40°C in a humid hybridization oven, then washed in wash buffer 2 × 2 min at RT. Finally, the brain slices were hybridized with the amplification (Amp): −1 (30 min; 40°C), −2 (15 min; 40°C), −3 (30 min; 40°C), and −4 (15 min; 40°C) probes. After each individual amplification step, the slides were washed with washing buffer and were mounted after the final amplification.

Immunoblot

The total protein amount was determined by the Bradford protein assay (Bio-Rad). Samples (10 µg) were homogenized in lysis buffer (20 mm Tris-HCl, 140 mm NaCl, 1 mm EDTA, SDS 0.1%, Triton 1%, and glycerol 10%) with a protease inhibitor cocktail and loaded into a 4%-15% gradient SDS-PAGE gel. Separated proteins were transferred onto a PVDF membrane (Bio-Rad) for 1.5-2 h. The membrane was blocked with 3% milk in TBS with 0.1% Tween 20 (TBST) and then incubated with the appropriate primary antibody overnight at 4°C. Anti-GLT1 (1:5000, rabbit) was a generous gift from Jeffrey Rothstein (John Hopkins University). Anti-β-actin (1:1000, Sigma), anti-Iba1 (rabbit, Wako #019-19741), anti-PSD95 (Cell Signaling Technology #2507), anti-Histone H2A (rabbit, Cell Signaling Technology #2718), anti-GFAP (1:1000; Dako), and anti-eGFP (NeuroMab clone N86/8) were obtained commercially. Following incubation with the primary antibody, the membrane was exposed to the HRP-conjugated goat anti-rabbit secondary antibody (1:5000), diluted in TBST. Bands were visualized on CL-XPosureTM film (Thermo Fisher Scientific) by ECL Plus chemiluminescent substrate (Thermo Fisher Scientific). Different exposure times were used for detecting different proteins. Glutamate transporter immunoblots often show a monomer (62 kDa), a dimer (120 kDa), and sometimes multimers (250 kDa), as previously described (Rothstein et al., 1994).

qRT-PCR

RNA isolated with the TRAP procedure from cortex tissue, SNS preparation, or total cortex was converted to cDNA using a high-capacity cDNA synthesis kit (Thermo Fisher Scientific). The relative expression levels of select genes were measured by qPCR using SYBR Green (Invitrogen) reagent, and specific primers for analyzed genes were chosen from the PrimerBank (https://pga.mgh.harvard.edu/primerbank/) as follows: Glt1, forward: 5′-ACAATATGCCCAAGCAGGTAGA-3′, reverse: 5′-GACACCAAACACAGTCAGTGA-3′; Kir4.1, forward: 5′-GTCGGTCGCTAAGGTCTATTACA-3′, reverse: 5′-GGCCGTCTTTCGTGAGGAC-3′; Arc, forward: 5′-AAGTGCCGAGCTGAGATGC-3′, reverse: 5′-CGACCTGTGCAACCCTTTC-3′; Oligo2, forward: 5′-TCCCCAGAACCCGATGATCTT-3′, reverse: 5′-CGTGGACGAGGACACAGTC-3′; Fabp7, forward: 5′-GGACACAATGCACATTCAAGAAC-3′, reverse: 5′-CCGAACCACAGACTTACAGTTT-3′; β-actin, forward: GGCTGTATTCCCCTCCATCG, reverse: CCAGTTGGTAACAATGCCATGT; β-actin was used as endogenous control for normalization.

Statistical analysis

Sample sizes were determined by G*power analysis (version 3.1) with effect size determined empirically and 80% power and are sufficient for statistical analysis. The statistical analysis approach used for each experiment is described in each method section and in the figure legends. All values were plotted as mean ± SEM unless otherwise noted. All volcano plots, percent stacked bar graphs, scatter plots, and heat maps were generated in Prism 9 from the analyzed RNA-Seq data.

Results

Differential perisynaptic astroglial localization of eGFP-tagged ribosome subunit L10a in WT and FMRP-deficient cortex

Previously, Aldh1l1-TRAP transgenic mice in which an eGFP-tagged ribosome subunit L10a is selectively expressed in astroglia have been generated (Doyle et al., 2008), which allow selective tagging of ribosomes and isolation of ribosome-bound translating mRNA in astroglia. To visualize the subcellular localization of labeled ribosome subunits, especially at astroglial processes, in WT and FMRP-deficient brains, we generated Fmr1+/yAldh1l1-TRAP+Eaat2-tdT+ and Fmr1-/yAldh1l1-TRAP+Eaat2-tdT+ mice, in which the full astroglial domain morphology, including fine processes, is clearly illustrated based on the Eaat2 promoter driven tdT expression, as previously demonstrated (Morel et al., 2014). We particularly analyzed cortical astroglia from somatosensory layers IV-V, as the tdT labeling of astroglia in these layers is more separated for domain analysis and elongated cortical UP state in these layers has been previously observed in Fmr1 KO mice (Hays et al., 2011). The overall cortical astroglial domain size and distribution, based on the tdT fluorescence, are highly similar between WT and FMRP-deficient mice (Fig. 1A) at postnatal day 30 (P30). While the eGFP-tagged L10a is observed in both soma (Fig. 1Ba,Bb, yellow arrows) and processes (both primary and secondary) (Fig. 1Ba,Bb, white arrows) of astroglia, the eGFP-L10a fluorescence intensity is substantially higher at processes than in soma of astroglia in FMRP-deficient mice but more similar in WT mice (Fig. 1C). Additional significant increase of the eGFP-L10a fluorescence at astroglial processes of FMRP-deficient mice compared with WT mice was also observed (Fig. 1C). We also measured tdT fluorescence from astroglial domains in Figure 1A and found modest (18%) reduction (p = 0.19, Fig. 1D) between WT and FMRP-deficient cortical astroglia. These results suggest a ribosome redistribution between soma and processes induced by the FMRP deficiency in cortical astroglia.

Figure 1.
Figure 1.

PAP localization of eGFP-tagged ribosome subunit L10a in WT and FMRP-deficient cortical astroglia. A, Cumulative probability curve of astroglial domain size from Fmr1+/yEaat2-tdT+ and Fmr1-/yEaat2-tdT+ mice at P30. Inset, Bar graph represents the average domain size for each group. N = 74-88 cells from 3 mice per group. Representative confocal images (B) and quantification (C) of eGFP-tagged ribosome subunit L10a localization in astroglial soma (S) and distal processes (P), as indicated by the tdT+ reporter from Fmr1+/yAldh1l1-TRAP+Eaat2-tdT+ and Fmr1-/yAldh1l1-TRAP+Eaat2-tdT+ mice at P30. White arrows indicate eGFP-tagged L10a at astroglial processes. Yellow arrows indicate eGFP-tagged L10a in astroglial soma. The total intensity of eGFP was quantified. Scale bar, 10 µm. N = 17-26 cells from >3 images per mouse in each group. D, Quantification of tdT fluorescence intensity in cortical astroglia of Fmr1+/yAldh1l1-TRAP+Eaat2-tdT+ and Fmr1-/yAldh1l1-TRAP+Eaat2-tdT+ mice at P30. N > 80 cells from 5 mice/group. Representative immuno-EM images (E) and quantification (F) of eGFP+ immunogold labeling at PAPs in somatosensory cortex of Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice. Scale bar, 100 nm. A, Astroglial processes; T, axon terminal; D, dendritic spine. Quantification of eGFP+ immunogold particles was conducted within a 600 nm radius (white dashed circle) from the center of the synapse in cortex. N = 18-32 images/3 or 4 mice per group. G, Quantification of postsynaptic dendritic size from somatosensory cortex of Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice. White dashed line indicates the individual quantified dendritic spine area. N = 11 images/3 or 4 mice per group. H, Representative confocal images of incorporated puromycin in both astroglial soma and processes. White arrows indicate puromycin signal in astroglial processes. Yellow arrows indicate puromycin signal in astroglial soma. Gray arrows indicate puromycin signal in neuronal soma. Scale bar, 10 µm. A, D, F, G, p values determined by unpaired Student's t test. C, p values determined by one-way ANOVA followed by post hoc Tukey's test.

To specifically examine the localization of the ribosome subunit L10a at PAPs, we next performed immunogold labeling of eGFP (tagged with L10a) on the cortex of Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice at P30 and observed abundant eGFP+ immunogold particles (Fig. 1Ea,Eb, white arrows) in PAPs. The synapse is clearly identified by the presence of abundant presynaptic vesicles and postsynaptic density area (Fig. 1Ea,Eb). A modest and not statistically significant increase in the percentage of synapses contacted by astroglial process in cortex (Fmr1+/yAldh1l1-TRAP+: 37.5%±12.9%; Fmr1-/yAldh1l1-TRAP+: 50.2%±8.1%, N = 42-55 synapses/3 mice per group) was observed. However, the average number of eGFP+ immunogold particles within PAPs, quantified within a 600 nm (adequate distance to include astroglial PAPs) circle from the center of individual synapses (Fig. 1Ea), was significantly less (Fig. 1F) in cortex of Fmr1-/yAldh1l1-TRAP+ mice compared with Fmr1+/yAldh1l1-TRAP+ mice, suggesting a reduced number of ribosomes localized specifically at PAPs in FMRP-deficient cortex. This is not necessarily in disagreement with the observed increase of the overall eGFP-L10 fluorescence in FMRP-deficient mice (Fig. 1C), as they vary in quantification method (fluorescent intensity vs immunogold particle number) as well as location of quantified astroglial processes (perisynaptic vs all). Dendritic spines in FXS have been commonly described as “immature,” based on their elongated and thin morphology (Comery et al., 1997). However, these studies often used Golgi staining or were based on fluorescent reporters (Portera-Cailliau, 2012), which are not optimal for examining spine size. The cross-section of dendritic spines in FMRP deficient conditions remains little examined. We thus quantified postsynaptic dendritic spine size (Fig. 1Eb, white dashed line) from eGFP immunoEM images. Interestingly, we found a 30% increase (p = 0.052, N = 26-31 synapses/11 images per group, Fig. 1G) of dendritic spine area in FMRP-deficient (0.12 ± 0.01 µm2) mice compared with WT (0.08 ± 0.01 µm2) mice, possibly influenced by enhanced cortical neuronal excitability as previously reported (Contractor et al., 2015).

To further determine whether translation occurs at astroglial processes, we next performed a puromycin incorporation assay on cortical slices prepared from Fmr1+/yEaat2-tdT+ and Fmr1-/yEaat2-tdT+ mice (P30). Similar to aminoacylated tRNA, puromycin can be efficiently incorporated into the elongating peptide and stop the translation. As a result, detection of puromycylated peptides by its specific antibody has been correlated to newly synthesized proteins and translation activity (Aviner, 2020), which facilitates spatial detection of new protein synthesis, such as at astroglial processes. Together with the illustration of tdT+ astroglial processes, consistently low but visible puromycin signals were observed at tdT+ astroglial processes (Fig. 1H, white arrows) as well as astroglial soma (Fig. 1H, yellow arrows) in both Fmr1+/yEaat2-tdT+ and Fmr1-/yEaat2-tdT+ cortical slices, suggesting that FMRP deficiency likely has no or low impact on overall baseline protein synthesis at astroglial processes and soma. In contrast, strong positive puromycin immunoreactivity was seen in nearby neuronal soma (Fig. 1Ha, gray arrows) in both WT and FMRP-deficient slices. Subsequent quantification confirmed a low level (∼1%-2%) of puromycin immunoreactivity at astroglial processes (regardless of the Fmr1 genotype), compared with neuronal soma. These results are in line with previous reports that only a low level of baseline protein synthesis occurs at astroglial processes (Mazare et al., 2020). Overall, these confocal and EM images provide evidence that ribosome subunits can be commonly localized at astroglial processes, including perisynaptic processes, and that baseline protein synthesis occurs at such processes in both WT and FMRP-deficient conditions.

Isolation and validation of mRNAs localized at cortical astroglial processes with the SNS/TRAP procedure from Aldh1l1-TRAP mice

Although prior studies have begun to identify localized mRNAs at astroglial processes (Sakers et al., 2017; Mazare et al., 2020), how the localization and trafficking of these mRNAs to astroglial processes are regulated remains unknown. As an RBP, FMRP has been shown to traffic to neuronal dendrites on activity stimulation (Dictenberg et al., 2008), which has been suggested to be a required step for mGluR activation-induced mRNA dendritic localization (Bassell and Warren, 2008). Whether and how FMRP regulates subcellular localization and expression of mRNAs in astroglia remains essentially unknown. Previously, we and others have found FMRP expression in developing astroglia (Pacey and Doering, 2007; Higashimori et al., 2016). To specifically examine FMRP immunoreactivity at astroglial processes, we performed FMRP immunostaining on cortical sections of Eaat2-tdT mice at P15, as FMRP expression is highest in glia during development (Pacey and Doering, 2007). As expected, we found that FMRP is primarily expressed in neuronal soma (Fig. 2Aa, yellow arrows). FMRP-immunopositive puncta at different branches of astroglial processes were also clearly observed (Fig. 2Ab, white arrows). We next decided to identify and compare mRNAs enriched at cortical astroglial processes in WT and FMRP-deficient conditions. By combining a modified Ficoll-sucrose gradient SNS preparation (Rao and Steward, 1991) and TRAP procedures in Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice, we adapted a SNS/TRAP approach to selectively isolate mRNA transcripts localized at astroglial processes (Sakers et al., 2017). Although the SNS procedure typically isolates synaptic structures, perisynaptic astroglial microdomains can also be isolated because of their contact with or close proximity to synapses, as shown in Figure 1E. The SNS fraction was clearly visible following gradient ultracentrifugation (Fig. 2B). The successful and specific preparation of SNS was further confirmed by the detection and enrichment of the typical neuronal synaptic protein PSD95 and the perisynaptic astroglial glutamate transporter GLT1 in the SNS fraction compared with the P1 pellet fraction and cortical lysates (Fig. 2C). In addition, we found that GFAP, which preferentially labels the major but not distal and fine astroglial processes (Morel et al., 2014), is not detected in the SNS fraction, whereasa β-actin, which penetrates to the fine astroglial processes, is detected in the SNS fraction (Fig. 2C). The selective detection of astroglial GLT1 and β-actin, but not GFAP, further confirms the inclusion of perisynaptic fine (but not major) astroglial processes in the SNS fraction. Importantly, eGFP was detected and also enriched in the SNS fraction compared with the P1 pellet and total lysates (Fig. 2C), validating the presence of eGFP-L10a at PAPs shown in Figure 1. Additionally, the nuclear protein histone H2A and microglial protein Iba1 are detected only in cortical lysates and the P1 pellet fraction but not in the SNS fraction, suggesting minimal contamination with cellular soma and other glial cells in the SNS fraction. Overall, these immunoblot results suggest that the SNS fraction specifically contains eGFP+ PAPs (GLT1+, β-actin+, and GFAP–) and neuronal synapses (PSD95+).

Figure 2.
Figure 2.

Isolation and validation of cortical astroglial process-enriched mRNAs using SNS/TRAP from Aldh1l1-TRAP mice. A, Representative confocal images of FMRP immunostaining at astroglial processes of Eaat2-tdT mouse somatosensory cortex (P15). Aa, Overlay of FMRP immunoreactivity and tdT fluorescence (40× confocal image). Scale bar, 20 µm. Ab, Magnified view of astroglial processes (white box) in Aa, taken at 63× with 2 times zoom-in. Scale bar, 10 µm. Aa, Yellow arrows indicate FMRP immunostaining in cortical neurons. Ab, White arrows indicate colocalized FMRP immunostaining signals with tdT fluorescence. B, Representative image of pelleted SNS fractions from Ficoll-sucrose density gradient ultracentrifugation of Fmr1+/yAldh1l1-TRAP+ (P40) somatosensory cortical homogenates. C, Representative immunoblot of synaptic and astroglial process proteins for validation of SNS fractions. P1 pellet fraction: pellet from 1000 × g centrifugation following tissue homogenization. D, Diagram of synaptic compartments and potential mRNA localization in astroglial processes (SNS/TRAP) or nonastroglial processes (SNS/Sup). SNS/Sup refers to flowthrough from the TRAP procedure, likely containing neuronal and potentially other glial mRNAs. qRT-PCR quantification of representative astroglial (E) and nonastroglial (F) enriched mRNAs isolated from SNS/TRAP, SNS/Sup, and somatosensory cortical lysate (total input). β-actin was used as endogenous control, and the comparison was based on total input (E) or SNS/TRAP (F). N = 3-6 biologically independent samples per condition. p values determined using one-way ANOVA and post hoc Tukey test.

By performing the TRAP procedure from the SNS fraction, we isolated sufficient mRNA (typically 12-15 ng mRNA/sample) that contained highly integral 18s and 28s ribosomal RNA (RIN > 9, 28s/18s > 2). Previously, the Fabp7 mRNA has been found to be enriched at PAPs (Gerstner et al., 2012). In addition, GLT1 and potassium channel Kir4.1 proteins are known to be highly expressed at astroglial processes near synapses. We therefore compared the relative abundance of Fabp7, Glt1, and Kir4.1 mRNA from SNS/TRAP (mRNAs at astroglial processes), SNS/Sup (flow-through from the SNS/TRAP, presumably mRNAs in neurons and other CNS cells) (as illustrated in Fig. 2D), and total mRNA input (mRNAs isolated from total cortical homogenates) samples from the somatosensory cortex of Aldh1l1-TRAP mice by qRT-PCR. Our qRT-PCR results showed that the threshold of cycle (Ct) value for Fabp7, Glt1, and Kir4.1 mRNA from SNS/TRAP samples was in the 20-26 range, indicating an abundant amount of these mRNAs. The relative level of Fabp7, Glt1, and Kir4.1 mRNA was indeed 56-, 18-, and 10-fold higher, respectively, than the level in SNS/Sup and total mRNA input samples (Fig. 2E), suggesting a significant enrichment of these known astroglial mRNAs in SNS/TRAP samples. In contrast, mRNA levels of typical neuronal (Arc) and oligodendroglial (Olig2) genes were barely detectable in SNS/TRAP samples, although clearly detected in SNS/Sup (Arc) or total mRNA input samples (Arc and Olig2) (Fig. 2F), confirming very limited, if any, contamination of neuronal synaptic mRNAs in SNS/TRAP samples.

Identification of mRNAs enriched at astroglial processes of WT and FMRP-deficient somatosensory cortex

Encouraged by the sufficient quantity of isolated mRNA and the astroglial mRNA specificity from the SNS/TRAP procedure on Aldh1l1-TRAP mice, we prepared sequencing libraries from total isolated cortical RNA (T, n = 2), astroglia ribosome-bound mRNAs (TRAP, n = 3), ribosome-bound mRNAs from astroglial processes (SNS/TRAP, n = 3), and flow-through mRNAs from the SNS/TRAP procedure (SNS/Sup, n = 3) from somatosensory cortex of both Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice. A total of 50-70 million reads for each sample group was obtained, which was sufficient for subsequent mapping and bioinformatic analysis. We first performed multidimensional scaling analysis to compare the similarity of all RNA-Seq datasets and found consistent clustering of RNA-Seq datasets from samples prepared by the same procedure (Fig. 3A), suggesting a unique mRNA profile for each sample group and confirming the isolation of specific mRNA transcripts using these distinct procedures. Additionally, representative astroglia-specific mRNAs, but not other CNS cell-type specific mRNAs, were selectively enriched only in astroglial TRAP and SNS/TRAP samples (highlighted by the red line, Fig. 3B), further validating astroglial specificity of the TRAP and SNS/TRAP procedure. Moreover, the total number of different mRNAs detected (CPM > 1) from both TRAP and SNS/TRAP RNA-Seq datasets was highly comparable (Fig. 3C), indicating a very similar composition complexity of transcribed mRNAs, but not an overrepresentation of a small subset of mRNAs, in these samples.

Figure 3.
Figure 3.

Identification of mRNAs enriched at astroglial processes of Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ somatosensory cortex. A, Multidimensional scaling plot of RNA-Seq datasets from different sample groups. T, Total RNA isolated from somatosensory cortex; TRAP, RNA isolated by the TRAP procedure from somatosensory cortex; SNS/TRAP, RNA isolated by the TRAP procedure from SNS samples; SNS/Sup, flowthrough from the TRAP procedure from the SNS samples. N = 3 biologically independent samples per condition, except total RNA samples (n = 2). B, Heatmap of representative CNS cell type mRNAs for astroglia (A), neurons (N), oligodendrocytes (O), microglia (M), and endothelial (E) cells in RNA-Seq datasets from different sample groups. All RNAs were isolated from Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice at P40. T, Total input. C, Number of detected individual mRNAs (CPM > 1) in RNA-Seq datasets from different sample groups. Volcano plot of the mRNAs significantly enriched in SNS/TRAP (astroglial processes, highlighted in green) or in TRAP (highlighted in red) identified by RNA-Seq in cortical astroglia of Fmr1+/yAldh1l1-TRAP+ (D, Extended Data Fig. 3-1) and Fmr1-/yAldh1l1-TRAP+ cortex (E, Extended Data Fig. 3-2). P-E, Process-enriched; NP-E, non–process-enriched. F, Volcano plot of the mRNAs significantly enriched in SNS/TRAP or in TRAP datasets from Fmr1+/yAldh1l1-TRAP+ cortex with known FMRP-associated mRNAs highlighted in blue. See also Extended Data Figure 3-3.

Figure 3-1

Enriched mRNAs at cortical astroglial process in Fmr1+/yAldh1l1-TRAP+ mice. Download Figure 3-1, XLSX file.

Figure 3-2

Enriched mRNAs at cortical astroglial processes in Fmr1-/yAldh1l1-TRAP+ mice. Download Figure 3-2, XLSX file.

Figure 3-3

FMRP-associated mRNAs that are enriched at astroglial processes or non-processes. Download Figure 3-3, XLSX file.

Prior studies have demonstrated selective and enriched isolation of astroglial mRNAs from total brain region tissues using the TRAP procedure (Doyle et al., 2008; Morel et al., 2017). As eGFP-L10a is distributed both in soma and processes in astroglia (Fig. 1), TRAP based isolation includes soma and process located mRNAs in astroglia from Aldh1l1-TRAP cortex. In contrast, the SNS/TRAP procedure preferentially isolates ribosome-bound mRNAs localized at astroglial (fine) processes as shown in Figure 2. Thus, a direct comparison between SNS/TRAP and TRAP RNA-Seq datasets allows identification of mRNAs that are particularly enriched at astroglial processes (Sakers et al., 2017). This comparison is also reasoned based on the similar mRNA composition complexity (Fig. 3C) in both TRAP and SNS/TRAP RNA-Seq datasets. Indeed, a large number of mRNAs were found to be enriched (determined by Padj < 0.05) at astroglial processes (designated as P-E) in both WT (4179) and FMRP-deficient (4524) mice (Fig. 3D,E; Extended Data Figs. 3-1 and 3-2). There are also many mRNAs that are enriched in TRAP than in SNS/TRAP RNA-Seq datasets, which are either enriched in the soma or similarly present at both processes and in soma. We designated these mRNAs as non–process-enriched (NP-E). The identification of several thousand mRNAs enriched at astroglial processes is exciting but also surprising given the typically small diameter of astroglial (fine) processes that are conventionally considered spatially limited for harboring a large number of mRNA transcripts. Since all (except mitochondrial) mRNAs are initially transcribed from the soma-located nucleus, the quite large number of mRNAs enriched at processes also implies that intracellular mRNA trafficking to processes is a common cellular process in astroglia in vivo. As an RBP, FMRP has been shown to associate with many mRNAs from the developing mouse brain (Darnell et al., 2011). To determine whether FMRP-associated mRNAs are also localized at astroglial processes that can be near synapses, we compared FMRP-associated mRNAs with our TRAP and SNS/TRAP RNA-Seq datasets. Indeed, we found that 445 FMRP-associated mRNAs are differentially expressed either in TRAP or SNS/TRAP datasets (Fig. 3F; Extended Data Fig. 3-3), among which, 315 such mRNAs are preferentially localized at astroglial processes (Fig. 3F; Extended Data Fig. 3-3), indicating a potentially more important role of FMRP in regulating process-enriched mRNAs.

To confirm that mRNAs identified from the SNS/TRAP versus TRAP comparison are indeed astroglial process-enriched, we next performed ISH of two identified mRNAs, Glt1 and Epha4, with their specific RNAscope probes in Fmr1+/yEaat2-tdT+ and Fmr1-/yEaat2-tdT+ cortical sections. Probe-specific in situ signals were clearly observed in cortical sections with no visible background (Fig. 4Ab′-Ae′ compared with Fig. 4Aa′). Although both Glt1 and Epha4 mRNAs are not selectively expressed in astroglia, visualization of ISH signals of these two mRNAs in astroglial soma and processes was facilitated by the tdT reporter in cortical astroglia (Fig. 4Aa-Ae). We further generated 3D domains of individual astroglia using Imaris software (Fig. 4Ab″-Ae″) and filtered out Glt1 or Epha4 mRNA ISH signals outside of individual astroglial domains, so that ISH signals either in astroglial soma (S) or at processes (P) (by subtracting signals in soma from signals in total astroglia) can be selectively quantified (Fig. 4Ab‴-Ae‴). As expected, the overall Glt1 in situ signals in astroglia are much more evident than that of overall Epha4 mRNA. Representative Epha4 mRNA in situ signals colocalized (white arrows, Fig. 4Ad′,Ad′,Adm,Ad‴) or not colocalized (not in the same focal plane, blue arrows, Fig. 4Ae′,Aem,Ae‴) were shown. Quantification of both Glt1 and Epha4 ISH signals confirmed that both mRNAs have substantially higher expression levels at processes than in soma in both Fmr1+/yEaat2-tdT+ and Fmr1-/yEaat2-tdT+ mice (Fig. 4B,C). Interestingly, both Glt1 and Epha4 mRNAs have a reduced mRNA level in FMRP-deficient astroglial processes than in WT astroglial processes (Fig. 4B,C). However, the overall changes of Glt1 and Epha4 mRNA levels in WT and FMRP-deficient conditions are distinct. The total Glt1 mRNA levels in cortical astroglia are significantly decreased in the absence of FMRP (Fig. 4D), consistent to our previous observation that FMRP deficiency reduces functional expression of cortical GLT1 in Fmr1-/y mice (Higashimori et al., 2013). On the other hand, total astroglial Epha4 mRNA levels remain unchanged by the lack of FMRP (Fig. 4E).

Figure 4.
Figure 4.

ISH of representative mRNAs preferentially expressed at astroglial processes of Fmr1+/yEaat2-tdT+ and Fmr1-/yEaat2-tdT+mice. A, Representative confocal and Imaris 3D images of Glt1 and Epha4 mRNA ISH signals in somatosensory cortical astroglia of Fmr1+/yEaat2-tdT+ and Fmr1-/yEaat2-tdT+mice (P40). Scale bar, 10 µm. Ad′, Adm, and Ad‴, White arrows indicate Epha4 mRNA in situ signals colocalized with astroglia. Ae′, Aem, Ae‴, Blue arrows indicate Epha4 mRNA in situ signals not colocalized with astroglia. Relative Glt1 (B) and Epha4 (C) mRNA levels in soma and processes of somatosensory cortical astroglia of Fmr1+/yEaat2-tdT+ and Fmr1-/yEaat2-tdT+mice. The astroglial 3D domain was generated using Imaris. The total (sum) intensity of Glt1 or Epha4 mRNA in situ signals in soma or within individual astroglial domain was quantified in Imaris. The total (sum) intensity at astroglial processes was calculated by subtracting the soma sum intensity from the overall sum intensity within the astroglial domain. The sum intensity of control (Fmr1+/yEaat2-tdT+) astroglial soma was used as a calibrator for comparison. Relative Glt1 (D) and Epha4 (E) mRNA levels (mean intensity) in somatosensory cortical astroglia of Fmr1+/yEaat2-tdT+ and Fmr1-/yEaat2-tdT+mice. The mean intensity was calculated by dividing the sum intensity of the all (soma and processes) in situ signals in individual astroglia by the total volume of that astroglia. For Glt1 in situ, N = 7 or 8 images (total 72-94 astroglia)/4 mice per group. For Epha4 in situ, N = 6 images (total 40-68 astroglia)/3 mice per group. B, C, p values were determined by one-way ANOVA and post hoc Tukey test. D, E, p values were determined by unpaired Student's t test.

Distinct functional categories of mRNAs are selectively clustered at cortical astroglial processes and are preferentially affected by FMRP deficiency

Previously, a subset of mRNAs has been identified to be selectively expressed in astroglia by comparing RNA-Seq datasets from different CNS cell types (Zhang et al., 2014). We decided to examine whether the compartmental localization of these mRNAs (Extended Data Fig. 5-1) is closely associated with their putative functions. We calculated the process enrichment index (%) based on the CPM values in TRAP or SNS/TRAP RNA-Seq datasets and generated the percent stacked bar graph (Fig. 5A). Interestingly, mRNAs (e.g., Pla2g3, Grm3) encoding surface functional proteins tend to preferentially localize at processes (index < 50%) but mRNAs (e.g., Ptx3, Otx1) encoding transcriptional factors are highly soma enriched (index > 75%). To further analyze the association between the subcellular localization and function of mRNAs expressed in astroglia, mRNAs enriched in either TRAP or SNS/TRAP RNA-Seq datasets from WT and FMRP-deficient mice were analyzed in Ingenuity Pathway Analysis software to identify their functional categories. Of all functionally annotated mRNAs, top functional categories of mRNAs that are not enriched at astroglial processes (both WT and FMRP-deficient conditions) are typically involved in somatic functions (Fig. 5B; Extended Data Fig. 5-2), such as transcription (transcriptional regulators), cellular metabolism (enzymes), and intracellular signaling (kinases). However, these functional categories, in particular, transcriptional regulators, become far less dominant in mRNAs enriched at astroglial processes. Instead, categories of surface functional proteins, including GPCRs, ion channels, transmembrane receptors, and transporters, become highly represented (Fig. 5B; Extended Data Fig. 5-3).

Figure 5.
Figure 5.

Functionally clustered mRNAs in cortical astroglial processes are altered by FMRP deficiency. A, Percent stacked bar graph of astroglia enriched mRNAs (Extended Data Fig. 5-1) based on their TRAP and SNS/TRAP RNA-Seq CPM values from Fmr1+/yAldh1l1-TRAP+ mice. Percentage was calculated by dividing total (TRAP + SNS/TRAP) CPM by the TRAP CPM values. B, Pie charts represent functional categories of mRNAs enriched at astroglial processes in Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice identified by the Ingenuity Pathway Analysis (IPA) software. TRAP- or SNS/TRAP-enriched genes were determined by fold change (FC > 2) based on CPM and uploaded to IPA for functional category analysis. Percentage was calculated by dividing the total number of annotated genes by the number of genes in each functional category. NP-E, Non–process-enriched; P-E, process-enriched. See also Extended Data Figures 5-2 and 5-3. Volcano plot of the differentially expressed mRNAs at astroglial processes (C, SNS/TRAP, Extended Data Fig. 5-4) or astroglia (D, TRAP, Extended Data Fig. 5-5) between Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice; 1: Pcdha4; 2: Ube2o; 3: Epha4 (C); a: Junb; b: Fos; c: Npas4 (D). E, Scatter plot of all detected mRNAs with their preferential subcellular localization in Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice. mRNAs that were differentially expressed at astroglial process between Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice are highlighted in different colors, depending on their subcellular location. Blue represents mRNAs enriched in WT but not in FMRP-deficient astroglial processes. Red represents mRNAs not enriched in either WT or FMRP-deficient astroglial processes. Purple represents mRNAs enriched in WT or FMRP-deficient astroglial processes. Green represents mRNAs enriched in FMRP-deficient but not in WT astroglial processes. See also Extended Data Figure 5-6. N = 3 biologically independent samples per group.

Figure 5-1

Relative percentage of expression levels of astroglial enriched genes at astroglial process or non-process of Fmr1+/yAldh1l1-TRAP+ mice. Download Figure 5-1, XLSX file.

Figure 5-2

Percentage of different functional categories in IPA analysis_Fmr1+/yAldh1l1-TRAP+ mice. Download Figure 5-2, XLSX file.

Figure 5-3

Percentage of different functional categories in IPA analysis_Fmr1-/yAldh1l1-TRAP+ mice. Download Figure 5-3, XLSX file.

Figure 5-4

Differentially expressed mRNAs from SNS/TRAP between Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice. Download Figure 5-4, XLSX file.

Figure 5-5

Differentially expressed mRNAs in cortical astroglia (from TRAP) between Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice. Download Figure 5-5, XLSX file.

Figure 5-6

Preferential process (P-E or NP-E) localization of 87 genes from Figure 5C. NP-E: non–process-enriched; P-E: process-enriched. Download Figure 5-6, XLSX file.

Although the overall distribution pattern of different functional clusters from astroglial process-enriched mRNAs in both WT and FMRP-deficient mice is similar, whether the identity of such mRNAs in WT and FMRP-deficient mice is highly conserved or distinct is unknown. We then directly compared mRNAs detected at astroglial processes by using SNS/TRAP RNA-Seq datasets from Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice and found 86 mRNAs that are differentially expressed, as summarized in Extended Data Figure 5-4 and shown in Figure 5C. Specifically, 3 mRNAs (Pcdha4, Epha4, and Ube2o), which have also been previously found to be associated with FMRP (Darnell et al., 2011), are all present at significantly higher levels at astroglial processes in the FMRP-deficient condition. Although Epha4 mRNA in situ signals are reduced at FMRP-deficient astroglial processes (Fig. 4C), it is important to point out that SNS/TRAP RNA-Seq primarily assesses mRNA levels in perisynaptic, but not all, astroglial processes. Indeed, our quantification from eGFP immunoEM images found astroglial processes ensheath 38%-50% synapses in WT or FMRP-deficient cortex. In contrast, quantification of Epha4 in situ signals from astroglial processes included essentially all astroglial processes (perisynaptic and nonperisynaptic). As a result, the discrepancy in Epha4 mRNA levels at WT and FMRP-deficient astroglial processes reflects a different pool of astroglial Epha4 mRNA analyzed in these approaches.

On the other hand, a mere 8 potential protein coding mRNAs (Fig. 5D; Extended Data Fig. 5-5) were found to be differentially expressed from total ribosome bound mRNAs (TRAP RNA-Seq datasets) between Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ mice. Almost all these 8 mRNAs (except Fmr1) are involved in somatic functions, such as transcriptional regulation (Junb, Fos, Npas4, Pagr1a, and Samd11), Ca2+ binding (S100a10), and kinase activity (Lats2), and were not previously found to be associated with FMRP. We further generated a scatter plot based on relative fold changes between TRAP and SNS/TRAP RNA-Seq datasets from both Fmr1+/yAldh1l1-TRAP+ and Fmr1-/yAldh1l1-TRAP+ cortex to highlight the localization preference of these 86 mRNAs identified in Figure 5C. Indeed, the majority (69) of these 86 mRNAs are either enriched at astroglial processes (46, highlighted in purple) or undergo a localization switch (highlighted in blue and green, 23) in the absence of FMRP (Fig. 5E; Extended Data Fig. 5-6). The identity of these mRNAs is indicated in Extended Data Figure 5-6. Overall, these data suggest that substantially more mRNAs localized at astroglial processes are preferentially affected than mRNAs localized in the soma by the FMRP deficiency. In particular, the lack of FMRP induces not only expression level changes of mRNAs enriched at astroglial processes (highlighted in purple) but also changes in localization (highlighted in blue and green).

Discussion

In the current study, we investigated how the lack of FMRP alters the localization of ribosomes at PAPs and the subcellular localization/expression of mRNAs in cortical astroglia. Our results showed interesting localization-dependent functional clusters of mRNAs at astroglial processes. In addition, we found that the lack of FMRP preferentially alters the subcellular localization and expression of process-localized mRNAs, thus providing new candidate mRNAs that are potentially regulated by FMRP in cortical astroglia.

Localized astroglial signaling to synapses has been increasingly recognized to play functionally more significant roles in influencing neuronal activity (Bazargani and Attwell, 2016). However, the molecular understanding of such localized signaling regulation has been quite limited. This is partially because of the conventional view that such localized space, often the fine processes or microdomains, have small diameters (∼50 nm) and are typically organelle free (Bushong et al., 2002). However, a recent study using 3D-STED microscopy in living organotypic brain slices as well as in vivo found a reticular meshwork of nodes and shafts with median diameters of 330 and 202 nm, respectively (range from 200 to 800 nm and 200 to 300 nm, respectively) within the astroglial spongiform domain (Arizono et al., 2020), supporting the spatial feasibility that many mRNAs could be harbored at processes. Other studies have also showed the localization of mitochondria at astroglial processes (Jackson and Robinson, 2018), potentially serving as a local Ca2+ source (Agarwal et al., 2017). As a large number of mRNAs are isolated from the SNS/TRAP fraction of Aldh1l1-TRAP mice here and in prior studies (Sakers et al., 2017; Mazare et al., 2020) and only mature ribosomes are typically associated with mRNAs, observed eGFP-L10a immunogold particles are conceivably assembled mature ribosomes (typically 20-30 nm diameter), not just the L10a subunit, that associate with these mRNAs. Although the baseline translation activity at astroglial processes is low (based on puromycin incorporation), this translation can be very demand-dependent in response to specific stimuli for synthesizing a subset of proteins. This notion is supported not only by the large number of mRNAs located at processes, but also by the identity of these mRNA that preferentially encode functional surface proteins, such as transporters, receptors, and channels, which are more involved in synaptic signaling. The presence of mitochondria at astroglial processes also provides a robust energy source for the energy-demanding translation process. Functionally, as a single astroglia often receives heterogeneous synaptic inputs through different fine processes, this potential synthesis-on-demand mechanism may facilitate an input-specific local interaction between astroglia and different synapses.

How FMRP regulates mRNA translation in astroglia has just begun to be understood. Although a large number of mRNAs were previously found to be associated with FMRP (Darnell et al., 2011), it remained unknown whether some of these mRNAs were expressed in astroglia. By comparing FMRP-associated mRNAs with astroglial TRAP and SNS/TRAP RNA-Seq datasets, here we found 445 such mRNAs to be expressed in astroglia, with the majority (71%) of them enriched at astroglial processes, suggesting an important involvement of such gene functions at astroglial processes in FXS. Consistently, our previous studies have found that FMRP deficiency (especially from astroglia) downregulates functional expression of astroglial mGluR5 (Men et al., 2020) and GLT1 (Higashimori et al., 2013), both of which are highly enriched on the surface of astroglial processes. In addition, a recent study found increased expression of other surface functional proteins, such as purinergic P2Y2 and P2Y6 receptors, in astroglia of FMRP-deficient mice (Reynolds et al., 2021). Although the overall tdT+ astroglial domain size is not affected by the lack of FMRP (Fig. 1A), this fluorescence-based domain analysis may not adequately resolve the dynamic changes of astroglial fine processes in Fmr1-/y mice. Indeed, the percentage of synapses contacted by astroglial process in cortex is increased in FMRP-deficient mice, as well as redistribution of ribosomes between soma and processes in astroglia (Fig. 1). On the other hand, our RNA-Seq analysis showed that FMRP deficiency preferentially altered expression of mRNAs localized at astroglial processes by either changing their subcellular localization or expression levels, further supporting the importance of functional surface proteins at astroglial processes, encoded from process-localized mRNAs, in contributing to FXS pathogenesis. In particular, FMRP is associated with three such mRNAs, Pcdha4, Ube2o, and Epha4, with Epha4 and Ube2o having similar expression levels in astroglia and in neurons. Given the very low baseline protein translation activity at astroglial processes, it is not surprising that FMRP deficiency appears not to change the overall level of protein translation. It is also possible that FMRP only specifically regulates translation of a subset of mRNAs at processes, which is unable to be detected by the puromycin incorporation assay.

While our results showed that a large number of mRNAs are localized or even enriched at astroglial processes, very little is currently known about how the intracellular transport of mRNAs to astroglial processes is regulated. Studies in other cell types have well demonstrated the importance of various 3′UTR Cis elements in regulating the subcellular localization of mRNAs (Bramham and Wells, 2007). Analysis of such 3′UTR elements from process-enriched mRNAs is only at its infancy in astroglia (Sakers et al., 2017) that requires further examination and testing in future studies. Ultimately, these studies will promote our understanding of mRNAs' interaction with various RBPs and intracellular mRNA transport in astroglia.

Footnotes

  • This work was supported by National Institutes of Health R01MH106490 and R01NS118747 to Y.Y. We thank Zehua Chen (Broad Institute and Ling Chen from Massachusetts General Hospital) for the initial bioinformatic analysis; Tufts Center for Neuroscience Research and Tufts University Core Facility Genomics for confocal imaging, image analysis, and RNA-Seq support; and Harvard Medical School EM core facility for immunogold labeling of eGFP and EM imaging.

  • The authors declare no competing financial interests.

  • Correspondence should be addressed to Yongjie Yang at yongjie.yang{at}tufts.edu

References

  1. ↵
    2. Agarwal A,
    3. Wu PH,
    4. Hughes EG,
    5. Fukaya M,
    6. Tischfield MA,
    7. Langseth AJ,
    8. Wirtz D,
    9. Bergles DE
    (2017) Transient opening of the mitochondrial permeability transition pore induces microdomain calcium transients in astrocyte processes. Neuron 93:587–605.e587. doi:10.1016/j.neuron.2016.12.034 pmid:28132831
    OpenUrlCrossRefPubMed
  2. ↵
    2. Arizono M,
    3. Inavalli V,
    4. Panatier A,
    5. Pfeiffer T,
    6. Angibaud J,
    7. Levet F,
    8. Ter Veer MJ,
    9. Stobart J,
    10. Bellocchio L,
    11. Mikoshiba K,
    12. Marsicano G,
    13. Weber B,
    14. Oliet SH,
    15. Nagerl UV
    (2020) Structural basis of astrocytic Ca(2+) signals at tripartite synapses. Nat Commun 11:1906. doi:10.1038/s41467-020-15648-4 pmid:32312988
    OpenUrlCrossRefPubMed
  3. ↵
    2. Aviner R
    (2020) The science of puromycin: from studies of ribosome function to applications in biotechnology. Comput Struct Biotechnol J 18:1074–1083. doi:10.1016/j.csbj.2020.04.014 pmid:32435426
    OpenUrlCrossRefPubMed
  4. ↵
    2. Barres BA
    (2008) The mystery and magic of glia: a perspective on their roles in health and disease. Neuron 60:430–440. doi:10.1016/j.neuron.2008.10.013 pmid:18995817
    OpenUrlCrossRefPubMed
  5. ↵
    2. Bassell GJ,
    3. Warren ST
    (2008) Fragile X syndrome: loss of local mRNA regulation alters synaptic development and function. Neuron 60:201–214. doi:10.1016/j.neuron.2008.10.004 pmid:18957214
    OpenUrlCrossRefPubMed
  6. ↵
    2. Bazargani N,
    3. Attwell D
    (2016) Astrocyte calcium signaling: the third wave. Nat Neurosci 19:182–189. doi:10.1038/nn.4201 pmid:26814587
    OpenUrlCrossRefPubMed
  7. ↵
    2. Bramham CR,
    3. Wells DG
    (2007) Dendritic mRNA: transport, translation and function. Nat Rev Neurosci 8:776–789. doi:10.1038/nrn2150 pmid:17848965
    OpenUrlCrossRefPubMed
  8. ↵
    2. Bushong EA,
    3. Martone ME,
    4. Jones YZ,
    5. Ellisman MH
    (2002) Protoplasmic astrocytes in CA1 stratum radiatum occupy separate anatomical domains. J Neurosci 22:183–192. doi:10.1523/JNEUROSCI.22-01-00183.2002
    OpenUrlAbstract/FREE Full Text
  9. ↵
    2. Comery TA,
    3. Harris JB,
    4. Willems PJ,
    5. Oostra BA,
    6. Irwin SA,
    7. Weiler IJ,
    8. Greenough WT
    (1997) Abnormal dendritic spines in fragile X knockout mice: maturation and pruning deficits. Proc Natl Acad Sci USA 94:5401–5404. doi:10.1073/pnas.94.10.5401 pmid:9144249
    OpenUrlAbstract/FREE Full Text
  10. ↵
    2. Contractor A,
    3. Klyachko VA,
    4. Portera-Cailliau C
    (2015) Altered neuronal and circuit excitability in Fragile X syndrome. Neuron 87:699–715. doi:10.1016/j.neuron.2015.06.017 pmid:26291156
    OpenUrlCrossRefPubMed
  11. ↵
    2. Darnell JC,
    3. Van Driesche SJ,
    4. Zhang C,
    5. Hung KY,
    6. Mele A,
    7. Fraser CE,
    8. Stone EF,
    9. Chen C,
    10. Fak JJ,
    11. Chi SW,
    12. Licatalosi DD,
    13. Richter JD,
    14. Darnell RB
    (2011) FMRP stalls ribosomal translocation on mRNAs linked to synaptic function and autism. Cell 146:247–261. doi:10.1016/j.cell.2011.06.013 pmid:21784246
    OpenUrlCrossRefPubMed
  12. ↵
    2. Derouiche A,
    3. Anlauf E,
    4. Aumann G,
    5. Muhlstadt B,
    6. Lavialle M
    (2002) Anatomical aspects of glia-synapse interaction: the perisynaptic glial sheath consists of a specialized astrocyte compartment. J Physiol Paris 96:177–182. doi:10.1016/S0928-4257(02)00004-9 pmid:12445894
    OpenUrlCrossRefPubMed
  13. ↵
    2. Dictenberg JB,
    3. Swanger SA,
    4. Antar LN,
    5. Singer RH,
    6. Bassell GJ
    (2008) A direct role for FMRP in activity-dependent dendritic mRNA transport links filopodial-spine morphogenesis to fragile X syndrome. Dev Cell 14:926–939. doi:10.1016/j.devcel.2008.04.003 pmid:18539120
    OpenUrlCrossRefPubMed
  14. ↵
    2. Doyle JP,
    3. Dougherty JD,
    4. Heiman M,
    5. Schmidt EF,
    6. Stevens TR,
    7. Ma G,
    8. Bupp S,
    9. Shrestha P,
    10. Shah RD,
    11. Doughty ML,
    12. Gong S,
    13. Greengard P,
    14. Heintz N
    (2008) Application of a translational profiling approach for the comparative analysis of CNS cell types. Cell 135:749–762. doi:10.1016/j.cell.2008.10.029 pmid:19013282
    OpenUrlCrossRefPubMed
  15. ↵
    2. Gerstner JR,
    3. Vanderheyden WM,
    4. LaVaute T,
    5. Westmark CJ,
    6. Rouhana L,
    7. Pack AI,
    8. Wickens M,
    9. Landry CF
    (2012) Time of day regulates subcellular trafficking, tripartite synaptic localization, and polyadenylation of the astrocytic Fabp7 mRNA. J Neurosci 32:1383–1394. doi:10.1523/JNEUROSCI.3228-11.2012 pmid:22279223
    OpenUrlAbstract/FREE Full Text
  16. ↵
    2. Grosche J,
    3. Matyash V,
    4. Moller T,
    5. Verkhratsky A,
    6. Reichenbach A,
    7. Kettenmann H
    (1999) Microdomains for neuron-glia interaction: parallel fiber signaling to Bergmann glial cells. Nat Neurosci 2:139–143. doi:10.1038/5692 pmid:10195197
    OpenUrlCrossRefPubMed
  17. ↵
    2. Hagerman RJ,
    3. Berry-Kravis E,
    4. Kaufmann WE,
    5. Ono MY,
    6. Tartaglia N,
    7. Lachiewicz A,
    8. Kronk R,
    9. Delahunty C,
    10. Hessl D,
    11. Visootsak J,
    12. Picker J,
    13. Gane L,
    14. Tranfaglia M
    (2009) Advances in the treatment of fragile X syndrome. Pediatrics 123:378–390. doi:10.1542/peds.2008-0317 pmid:19117905
    OpenUrlAbstract/FREE Full Text
  18. ↵
    2. Hays SA,
    3. Huber KM,
    4. Gibson JR
    (2011) Altered neocortical rhythmic activity states in Fmr1 KO mice are due to enhanced mGluR5 signaling and involve changes in excitatory circuitry. J Neurosci 31:14223–14234. doi:10.1523/JNEUROSCI.3157-11.2011 pmid:21976507
    OpenUrlAbstract/FREE Full Text
  19. ↵
    2. Higashimori H,
    3. Morel L,
    4. Huth J,
    5. Lindemann L,
    6. Dulla C,
    7. Taylor A,
    8. Freeman M,
    9. Yang Y
    (2013) Astroglial FMRP-dependent translational down-regulation of mGluR5 underlies glutamate transporter GLT1 dysregulation in the fragile X mouse. Hum Mol Genet 22:2041–2054. doi:10.1093/hmg/ddt055 pmid:23396537
    OpenUrlCrossRefPubMed
  20. ↵
    2. Higashimori H,
    3. Schin CS,
    4. Chiang MS,
    5. Morel L,
    6. Shoneye TA,
    7. Nelson DL,
    8. Yang Y
    (2016) Selective deletion of astroglial FMRP dysregulates glutamate transporter GLT1 and contributes to fragile X syndrome phenotypes in vivo. J Neurosci 36:7079–7094. doi:10.1523/JNEUROSCI.1069-16.2016 pmid:27383586
    OpenUrlAbstract/FREE Full Text
  21. ↵
    2. Jackson JG,
    3. Robinson MB
    (2018) Regulation of mitochondrial dynamics in astrocytes: mechanisms, consequences, and unknowns. Glia 66:1213–1234. doi:10.1002/glia.23252 pmid:29098734
    OpenUrlCrossRefPubMed
  22. ↵
    2. Jin SX,
    3. Higashimori H,
    4. Schin C,
    5. Tamashiro A,
    6. Men Y,
    7. Chiang MS,
    8. Jarvis R,
    9. Cox D,
    10. Feig L,
    11. Yang Y
    (2021) Astroglial FMRP modulates synaptic signaling and behavior phenotypes in FXS mouse model. Glia 69:594–608. doi:10.1002/glia.23915 pmid:32970902
    OpenUrlCrossRefPubMed
  23. ↵
    2. Kanemaru K,
    3. Sekiya H,
    4. Xu M,
    5. Satoh K,
    6. Kitajima N,
    7. Yoshida K,
    8. Okubo Y,
    9. Sasaki T,
    10. Moritoh S,
    11. Hasuwa H,
    12. Mimura M,
    13. Horikawa K,
    14. Matsui K,
    15. Nagai T,
    16. Iino M,
    17. Tanaka KF
    (2014) In vivo visualization of subtle, transient, and local activity of astrocytes using an ultrasensitive Ca(2+) indicator. Cell Rep 8:311–318. doi:10.1016/j.celrep.2014.05.056 pmid:24981861
    OpenUrlCrossRefPubMed
  24. ↵
    2. Mazare N,
    3. Oudart M,
    4. Moulard J,
    5. Cheung G,
    6. Tortuyaux R,
    7. Mailly P,
    8. Mazaud D,
    9. Bemelmans AP,
    10. Boulay AC,
    11. Blugeon C,
    12. Jourdren L,
    13. Crom SL,
    14. Rouach N,
    15. Cohen-Salmon M
    (2020) Local translation in perisynaptic astrocytic processes is specific and changes after fear conditioning. Cell Rep 32:108076. doi:10.1016/j.celrep.2020.108076 pmid:32846133
    OpenUrlCrossRefPubMed
  25. ↵
    2. Men Y,
    3. Ye L,
    4. Risgaard RD,
    5. Promes V,
    6. Zhao X,
    7. Paukert M,
    8. Yang Y
    (2020) Astroglial FMRP deficiency cell-autonomously up-regulates miR-128 and disrupts developmental astroglial mGluR5 signaling. Proc Natl Acad Sci USA 117:25092–25103. doi:10.1073/pnas.2014080117 pmid:32958647
    OpenUrlAbstract/FREE Full Text
  26. ↵
    2. Morel L,
    3. Higashimori H,
    4. Tolman M,
    5. Yang Y
    (2014) VGluT1+ neuronal glutamatergic signaling regulates postnatal developmental maturation of cortical protoplasmic astroglia. J Neurosci 34:10950–10962. doi:10.1523/JNEUROSCI.1167-14.2014 pmid:25122895
    OpenUrlAbstract/FREE Full Text
  27. ↵
    2. Morel L,
    3. Chiang MS,
    4. Higashimori H,
    5. Shoneye T,
    6. Iyer LK,
    7. Yelick J,
    8. Tai A,
    9. Yang Y
    (2017) Molecular and functional properties of regional astrocytes in the adult brain. J Neurosci 37:8706–8717. doi:10.1523/JNEUROSCI.3956-16.2017 pmid:28821665
    OpenUrlAbstract/FREE Full Text
  28. ↵
    2. Nimchinsky EA,
    3. Oberlander AM,
    4. Svoboda K
    (2001) Abnormal development of dendritic spines in FMR1 knock-out mice. J Neurosci 21:5139–5146. doi:10.1523/JNEUROSCI.21-14-05139.2001
    OpenUrlAbstract/FREE Full Text
  29. ↵
    2. Nimmerjahn A,
    3. Mukamel EA,
    4. Schnitzer MJ
    (2009) Motor behavior activates Bergmann glial networks. Neuron 62:400–412. doi:10.1016/j.neuron.2009.03.019 pmid:19447095
    OpenUrlCrossRefPubMed
  30. ↵
    2. Pacey LK,
    3. Doering LC
    (2007) Developmental expression of FMRP in the astrocyte lineage: implications for fragile X syndrome. Glia 55:1601–1609. doi:10.1002/glia.20573 pmid:17823967
    OpenUrlCrossRefPubMed
  31. ↵
    2. Portera-Cailliau C
    (2012) Which comes first in fragile X syndrome, dendritic spine dysgenesis or defects in circuit plasticity? Neuroscientist 18:28–44. doi:10.1177/1073858410395322 pmid:21551076
    OpenUrlCrossRefPubMed
  32. ↵
    2. Rao A,
    3. Steward O
    (1991) Evidence that protein constituents of postsynaptic membrane specializations are locally synthesized: analysis of proteins synthesized within synaptosomes. J Neurosci 11:2881–2895. doi:10.1523/JNEUROSCI.11-09-02881.1991
    OpenUrlAbstract/FREE Full Text
  33. ↵
    2. Reynolds KE,
    3. Wong CR,
    4. Scott AL
    (2021) Astrocyte-mediated purinergic signaling is upregulated in a mouse model of Fragile X syndrome. Glia 69:1816–1832. doi:10.1002/glia.23997
    OpenUrlCrossRef
  34. ↵
    2. Richter JD,
    3. Zhao X
    (2021) The molecular biology of FMRP: new insights into fragile X syndrome. Nat Rev Neurosci 22:209–222. doi:10.1038/s41583-021-00432-0 pmid:33608673
    OpenUrlCrossRefPubMed
  35. ↵
    2. Rothstein JD,
    3. Martin L,
    4. Levey AI,
    5. Dykes-Hoberg M,
    6. Jin L,
    7. Wu D,
    8. Nash N,
    9. Kuncl RW
    (1994) Localization of neuronal and glial glutamate transporters. Neuron 13:713–725. doi:10.1016/0896-6273(94)90038-8 pmid:7917301
    OpenUrlCrossRefPubMed
  36. ↵
    2. Sakers K,
    3. Lake AM,
    4. Khazanchi R,
    5. Ouwenga R,
    6. Vasek MJ,
    7. Dani A,
    8. Dougherty JD
    (2017) Astrocytes locally translate transcripts in their peripheral processes. Proc Natl Acad Sci USA 114:E3830–E3838. doi:10.1073/pnas.1617782114 pmid:28439016
    OpenUrlAbstract/FREE Full Text
  37. ↵
    2. Sutton MA,
    3. Schuman EM
    (2006) Dendritic protein synthesis, synaptic plasticity, and memory. Cell 127:49–58. doi:10.1016/j.cell.2006.09.014 pmid:17018276
    OpenUrlCrossRefPubMed
  38. ↵
    2. Tang W,
    3. Szokol K,
    4. Jensen V,
    5. Enger R,
    6. Trivedi CA,
    7. Hvalby O,
    8. Helm PJ,
    9. Looger LL,
    10. Sprengel R,
    11. Nagelhus EA
    (2015) Stimulation-evoked Ca2+ signals in astrocytic processes at hippocampal CA3-CA1 synapses of adult mice are modulated by glutamate and ATP. J Neurosci 35:3016–3021. doi:10.1523/JNEUROSCI.3319-14.2015 pmid:25698739
    OpenUrlAbstract/FREE Full Text
  39. ↵
    2. Verkerk AJ,
    3. Pieretti M,
    4. Sutcliffe JS,
    5. Fu YH,
    6. Kuhl DP,
    7. Pizzuti A,
    8. Reiner O,
    9. Richards S,
    10. Victoria MF,
    11. Zhang FP
    (1991) Identification of a gene (FMR-1) containing a CGG repeat coincident with a breakpoint cluster region exhibiting length variation in fragile X syndrome. Cell 65:905–914. doi:10.1016/0092-8674(91)90397-H pmid:1710175
    OpenUrlCrossRefPubMed
  40. ↵
    2. Yang Y,
    3. Vidensky S,
    4. Jin L,
    5. Jie C,
    6. Lorenzini I,
    7. Frankl M,
    8. Rothstein JD
    (2011) Molecular comparison of GLT1+ and ALDH1L1+ astrocytes in vivo in astroglial reporter mice. Glia 59:200–207. doi:10.1002/glia.21089 pmid:21046559
    OpenUrlCrossRefPubMed
  41. ↵
    2. Zhang Y,
    3. Chen K,
    4. Sloan SA,
    5. Bennett ML,
    6. Scholze AR,
    7. O'Keeffe S,
    8. Phatnani HP,
    9. Guarnieri P,
    10. Caneda C,
    11. Ruderisch N,
    12. Deng S,
    13. Liddelow SA,
    14. Zhang C,
    15. Daneman R,
    16. Maniatis T,
    17. Barres BA,
    18. Wu JQ
    (2014) An RNA-sequencing transcriptome and splicing database of glia, neurons, and vascular cells of the cerebral cortex. J Neurosci 34:11929–11947. doi:10.1523/JNEUROSCI.1860-14.2014 pmid:25186741
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Email
Print
View Full Page PDF
Citation Tools
Respond to this article
Functionally Clustered mRNAs Are Distinctly Enriched at Cortical Astroglial Processes and Are Preferentially Affected by FMRP Deficiency
Yuqin Men, Haruki Higashimori, Kathryn Reynolds, Leona Tu, Rachel Jarvis, Yongjie Yang
Journal of Neuroscience 20 July 2022, 42 (29) 5803-5814; DOI: 10.1523/JNEUROSCI.0274-22.2022
Reddit logo Twitter logo Facebook logo Mendeley logo

Jump to section

Keywords

Responses to this article

Respond to this article

Jump to comment:

No eLetters have been published for this article.