Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Sep 26.
Published in final edited form as: Mol Psychiatry. 2025 Nov 21;31(4):2069–2080. doi: 10.1038/s41380-025-03366-6

Enhanced CB1 receptor function in GABAergic neurons mediates hyperexcitability and impaired sensory-driven synchrony of cortical circuits in Fragile X Syndrome model mice

D Gonzalez 1,*, CR Jonak 2,*, M Bernabucci 1, G Molinaro 1, K Collins 1, SA Assad 2, JR Gibson 1, DK Binder 2,3,#, KM Huber 1,#
PMCID: PMC13613361  NIHMSID: NIHMS2211207  PMID: 41272343

Abstract

Electroencephalographic (EEG) recordings in individuals with Fragile X Syndrome (FXS) and the mouse model of FXS (Fmr1 KO) display cortical hyperexcitability at rest, as well as deficits in sensory-driven cortical network synchrony. A form of circuit hyperexcitability is observed in ex vivo cortical slices of Fmr1 KO mice as prolonged persistent activity, or Up, states. It is unknown if the circuit mechanisms that cause prolonged Up states contribute to FXS-relevant EEG phenotypes. Here we examined the role of endocannabinoids (eCB) in prolonged Up states in slices and resting and sensory-driven EEG phenotypes in awake Fmr1 KO mice. Bidirectional changes in eCB function are reported in the Fmr1 KO that depend on synapse type (excitatory or inhibitory). We demonstrate that pharmacological or genetic reduction of Cannabinoid Receptor 1 (CB1R) in GABAergic neurons rescues prolonged cortical Up states and deficits in sensory-driven cortical synchrony in Fmr1 KO mice. In support of these findings, recordings from Fmr1 KO cortical Layer (L) 2/3 pyramidal neurons revealed enhanced CB1R-mediated suppression of inhibitory synaptic currents. In contrast, genetic reduction of Cnr1 in glutamatergic neurons did not affect Up state duration, but deletion of Fmr1 in the same neurons was sufficient to cause long Up states. These findings support a model where loss of Fmr1 in glutamatergic neurons leads to enhanced CB1R-mediated suppression of GABAergic synaptic transmission, prolonged cortical circuit activation and reduced sensory-driven circuit synchronization. Results suggest that antagonism of CB1Rs may be a therapeutic strategy to correct sensory processing deficits in FXS.

Introduction

Fragile X syndrome (FXS) is the most common form of inherited intellectual disability and a leading monogenetic cause of autism 1. Individuals with FXS and the FXS mouse model - the Fmr1 knockout (KO) display altered sensory processing and sensory hypersensitivity in both the auditory and somatosensory modalities 2-4. Electroencephalographic (EEG) recordings in FXS individuals as well as Fmr1 KO mice provide evidence for enhanced cortical activity at rest, observed as enhanced resting state gamma power5-9. Cortical circuits also have impaired temporal processing in response to sensory inputs. This can be observed in FXS individuals and Fmr1 KO mice as reduced synchronization to time-varying auditory stimuli in the gamma frequency range (40-80Hz), termed intertrial phase coherence or ITPC 6-9. Importantly, the severity of EEG alterations in FXS individuals correlates with sensory sensitivities, and deficits in social communication and executive function7, 10-12, suggesting the EEG changes are informative clinically. The conserved EEG phenotypes across species suggest a common dysfunction of sensory circuits across FXS individuals and the Fmr1 KO. Thus, revealing circuit mechanisms of the EEG phenotypes in mice may provide knowledge of circuit dysfunction in humans with FXS.

Findings from acute slices and in vivo recordings in Fmr1 KO mice provide evidence of hyperexcitable cortical circuits that may underlie the EEG phenotypes 13-18. Multi-unit recordings in vivo in the primary somatosensory or auditory cortex of Fmr1 KO mice report higher baseline firing rates 19, 20. Network hyperexcitability is observed in somatosensory cortex slices or anesthetized Fmr1 KO mice as an increased duration of persistent activity states, or “Up states”. Up states are a rhythmic oscillation of cortical network activity driven by local excitatory and inhibitory cortical circuits and thus reflect the balanced functioning of excitation and inhibition 13, 14. The duration of both spontaneous and thalamically-evoked Up states is prolonged in Fmr1 KO suggesting that circuit mechanisms that control cortical circuit timing are deficient in the Fmr1 KO and/or circuits are hyperexcitable 13, 14. If or how the circuit mechanisms of prolonged Up states contribute to the EEG phenotypes is unclear. The Up state phenotype is robust in a reduced preparation such as the cortical slice and is amenable to probe the cellular and synaptic mechanisms of network dysfunction in Fmr1 KO.

Pathological endocannabinoid (eCB) signaling through cannabinoid 1 receptors (CB1Rs) has been demonstrated in the Fmr1 KO and contributes to behavioral phenotypes observed in Fmr1 KO mice including audiogenic seizures 21-25. If or how altered CB1R function contributes to altered sensory cortical networks in the Fmr1 KO is unknown. A major mechanism by which CB1Rs affect circuit function is through suppression of presynaptic glutamate and GABA release 26-30. Activation of postsynaptic Gq-coupled, Group 1 metabotropic glutamate receptors (mGluR1 and mGluR5), 31, 32 stimulates synthesis of the eCB, 2-arachidonoylglycerol (2-AG) via diacylglycerol lipase-α (DAGL-α) 33. 2-AG traverses the synaptic cleft, activates CB1Rs on presynaptic terminals of glutamatergic and GABAergic neurons which then suppress neurotransmitter release 26, 27, 34, 35. Thus, eCBs have the capability to regulate the balance of excitatory and inhibitory synaptic transmission. In the Fmr1 KO, CB1R-mediated regulation of synaptic transmission is altered, but the direction of change seems to depend on synapse type. In the Fmr1 KO, mGluR1/5 and CB1R-dependent suppression of inhibitory synaptic transmission is enhanced, whereas suppression of excitatory synaptic transmission is reduced 21 22-24. The different CB1R effects on inhibitory vs excitatory synaptic function were performed in different studies and brain regions (hippocampus, striatum, cortex), but effects are consistent for synapse type across regions. Furthermore, both increasing and decreasing eCB/CB1R function have been suggested to be therapeutic strategies for FXS 21,25, 36, 37. The imbalance of CB1R-dependent suppression of inhibitory and excitatory synaptic transmission in the Fmr1 KO may be expected to lead to hyperexcitability of circuits and results support this view 22, 25. However, it is unclear if altered CB1R function in excitatory or inhibitory neurons or both contribute to hyperexcitability related phenotypes associated with FXS.

Towards this goal, we discovered that pharmacological or genetic reduction of mGluR5 activity corrected Up state duration in the Fmr1 KO 14. Because eCB synthesis is a major downstream effector of mGluR5, we investigated the role of eCB synthesis and CB1Rs in prolonged Up states. We hypothesize that the circuit mechanisms of prolonged cortical Up states contribute to the EEG phenotypes of enhanced resting state gamma power and/or deficient ITPC. We test this hypothesis by determining if similar pharmacological and genetic manipulations of CB1R function correct both Up state and EEG phenotypes in Fmr1 KO mice. Our results indicate that enhanced CB1R-mediated suppression of inhibitory synaptic transmission in Fmr1 KO cortex contributes to prolonged Up states as well as deficits in cortical circuit synchronization to auditory stimuli in vivo. These results provide insight into the cellular and synaptic mechanisms of cortical network hyperexcitability and altered temporal processing. Results suggest that antagonism of CB1Rs may aid sensory processing deficits in FXS.

Methods and Materials

Mice:

We used the following mouse lines: Fmr1 KO 38, floxed Fmr1 39, floxed Cnr1 40, provided by Dr. Joel Elmquist (UT Southwestern Medical Center), VGlut1-Cre (Slc17a7-IRES-Cre; Jackson Labs, stock# 023527) 41, and VGlut2-Cre (Slc17a6-IRES-Cre; Jackson Labs, stock# 016963)42. All mice were maintained on a C57Bl/6J background by backcrossing with WT mice (Jackson Labs #: 000664). All procedures were approved by the UT Southwestern or UC Riverside Institutional Animal Care and Use Committee. Experiments were conducted in accordance with the NIH Guide for the Care and Use of Laboratory Animals.

Neocortical Slice preparation, Up state recordings and pharmacological treatments:

Up state experiments were performed in acute somatosensory neocortical slices prepared from male (P18-77) littermates of each genotype as described 14 (See Supplemental methods). Rimonabant (SR141716A, 5 μM, Tocris, Cat 0923. 0.1% DMSO) and DO34 (10 μM, Glixx Laboratories Inc., 0.1% DMSO) were included in recovery ACSF immediately after slicing and remained in the ACSF during recording (>2 hours).

Electroencephalogram (EEG):

A multi-electrode (36) array of leads was implanted on the skull surface in P80-90 mice as 9 (See Supplemental methods). Resting EEG, sensory responses to Chirp, 40Hz and 80Hz pulses were obtained and analyzed as described 9. For rimonabant EEG recordings, approximately 2 hour recordings were performed immediately before and after a 7-day treatment of rimonabant (after is on day 8). Once per day, a dose of 1mg/kg IP was applied as in previous studies 36, 43, 44. For the chirp auditory stimulus in these experiments, a 1 second sound intensity ramp preceding the chirp stimulus was not included. The ASSR was the same as for reduced gene dosage experiments except the click train was maintained for 3 seconds.

RNAscope in situ hybridization (ISH) combined with immunofluorescence:

RNAscope was performed in the UT Southwestern Metabolic Core. Two littermate pairs of WT and Fmr1 KO mice were anesthetized with a ketamine (80 mg/kg) –xylazine (10 mg/kg) cocktail; i.p.) and transcardially perfused with phosphate buffer saline (PBS) followed by 10% formalin. Brains were post-fixed for 24 hours at 4°C and then in 30% sucrose for another 24 hours at 4°C. Brains sections (25 μm) were cut with a cryostat and collected in PBS before being treated with hydrogen peroxide for 10 minutes. Sections were mounted onto SuperFrost slides and desiccated overnight at room temperature. On the following day, pretreatment and ISH were performed following the recommendations from the manufacturer (Advanced Cell Diagnostics, USA) and the reagents included in kit cat#323110. The probe for Cnr1 (cat# 420721-C2) was applied in a solution of probe diluent for 2 hours at 40°C (HybEZ oven). Slides were incubated with amplification reagents and Opal dyes 570 (1/1,500; Akoya Biosciences). Following the RNAScope procedure, neurons that are GFP-positive were labeled by incubating the brain slides with an antiserum against GFP (Aves Lab; cat# GFP-1020). Slides were next incubated one hour (1/1,000) with a biotinylated anti-chicken secondary (Jackson ImmunoResearch #703-065-155) and one hour (1/1,000) with a streptavidin AlexaFluor488 (Invitrogen, cat# S32354). Slides were rinsed and EcoMount (BioCare Medical, USA) was applied before adding a coverslip.

Statistical Analysis:

Unless stated otherwise, data are presented as the mean ± SEM. Significant differences were determined using 2 sided ttests, one-way or two-way ANOVA, and repeated-measures ANOVA when appropriate. Post hoc multiple comparison tests were performed where indicated. Sample sizes were chosen based on results from our previous studies required to observe genotypic effects 8, 14. Variance was similar between experimental groups as determined by Spearman’s test for heteroscedasticity in GraphPad.

Results

Antagonism of either CB1R or DAGLα is sufficient to rescue cortical prolonged Up states in Fmr1 KO mice.

We examined spontaneously occurring activity states, or Up states, in acutely prepared brain slices containing somatosensory cortex. Up state duration is longer in Fmr1 KO mice, and we interpret this to reflect circuit hyperexcitability in the cortex 13. To determine if longer Up states in Fmr1 KO mice are mediated by altered eCB signaling, we pharmacologically blocked eCB signaling. If longer Up states involved enhanced eCB signaling – like that occurring at GABAergic synapses – then blocking CB1Rs would be expected to reduce Up state duration to normal WT levels. We refer to this as a “rescue.” On the other hand, if the long Up states were a consequence of diminished eCB signaling – like that reported for glutamatergic synapses – then blocking eCB signaling may either prolong Up states or have no effect in the Fmr1 KO.

As a first test, we blocked CB1Rs by preincubating slices (2 hrs) with the CBR1 an inverse agonist, rimonabant (5 μM, 0.1% DMSO), or vehicle (0.1% DMSO). In vehicle treated slices from young mice (3 weeks of age or postnatal day; P18-24), Up states were longer in Fmr1 KO slices compared to WT, as expected (Fig. 1A,B,C). Rimonabant treatment reduced Up state duration in the Fmr1 KO as compared to vehicle-treated controls (Fig. 1B,C). Similar results were observed in slices from adult mice (9-11 weeks of age; Fig. 1D,E). These results are consistent with enhanced, eCB-mediated suppression of GABAergic transmission contributing to long Up states in the Fmr1 KO. Interestingly, rimonabant had no effect on Up state duration or frequency in WT slices, but reduced Up state amplitude across both genotypes (2 way ANOVA, Main effect of rimonabant; F (1, 109) = 4.071; p< 0.05; Fig. Supplementary (S) 1A,B) similar to previous observations 45.

Figure 1. Antagonism of CB1Rs and DAGLα rescues prolonged Up states in Fmr1 KO cortical slices.

Figure 1.

Open in a new tab

A) Up states were recorded extracellularly in layer 4 of somatosensory (S1) cortex. B) Example Up states from S1 cortical slices from 3 week old mice collected in vehicle or the CBR1 antagonist, Rimonabant (5 μM; Rim) Scale bar= 80μV/0.5 s. C) Group data of Up state durations (WT vehicle (veh); N=20/4 (slices/mice); WT Rim; N=20/4; Fmr1 KO veh; N= 34/7; Fmr1 KO Rim; 42/9). D) Example Up states in slices from adult (9-11 week old) mice in vehicle or the CBR1 antagonist, Rimonabant (5 μM) Scale bar= 100μV/0.5 s. E) Group data of Up state durations from adult mice (WT veh; N=19/4; WT Rim; N=16/4; Fmr1 KO veh; N= 18/5; Fmr1 KO Rim; 18/5). F) Example Up states in vehicle or the diacylglycerol lipase α inhibitor, DO34 (10 μM). G) Group data of Up state durations (WT veh; N=21/5; WT DO34; N=22/5; Fmr1 KO veh; N= 26/7; Fmr1 KO DO34; 29/7). Scale bar= 50 μV/1 s. 2-way ANOVA with Sidak's multiple comparisons test. H,I) Enhanced CB1R-mediated suppression of inhibitory synaptic transmission in Fmr1 KO cortex. H) Examples mIPSCs from a L2/3 pyramidal neuron before (baseline) and after Rimonabant (5 μM) application. Scale bars = 5 pA and 50 ms. I) Average mIPSC frequency (normalized to pre-rimonabant baseline) in WT and Fmr1 KO neurons. (WT: N 20/15; cells/mice; Fmr1 KO: N= 14/7). Repeated measures 2 way ANOVA; Rimonabant X genotype interaction (F (3, 96) = 3.366; p<0.05); Sidak’s posthoc tests; *, p<0.05 **, p<0.01 ****, p<0.001.

mGluR5 stimulates synthesis of the eCB 2-Arachidonoylglycerol (2-AG) by generating its precursor diacylglycerol. Diacylglycerol is then converted to 2-AG by diacylglycerol lipase-α (DAGLα) 46. To test if 2-AG synthesis was required for long Up states, we pretreated slices from 3 week old mice with a selective DAGLα inhibitor, DO34 (10 μM, 0.2% DMSO), or vehicle (0.2% DMSO) 47-49. As observed for rimonabant, DO34 treatment reduced Up state duration in the Fmr1 KO (Fig. 1F,G) and had no effect on duration in WT slices. DO34 also did not affect Up state frequency or amplitude in either genotype (Fig. S1C,D). These results are consistent with enhanced, ongoing eCB-mediated suppression at GABAergic synapses in the Fmr1 KO that leads to longer Up states.

Changes in GABAergic synaptic currents with application of a CB1R antagonist are consistent with greater eCB-mediated suppression in the Fmr1 KO.

The rescue of Up state duration with rimonabant suggests that CB1Rs are driving circuit hyperexcitability. We hypothesized that the CB1R-mediated suppression of inhibitory synaptic transmission may be enhanced in Fmr1 KO cortex similar to what has been observed in striatum and hippocampal CA1 22, 23. We tested the hypothesis by examining the effects of rimonabant on spontaneous miniature (m) IPSCs. While CB1Rs typically suppress evoked synaptic transmission 50, effects can also be observed on mIPSCs in the presence of Ca2+ 51-53. We measured mIPSC amplitude and frequency before and after wash-in of rimonabant (5 μM, Fig. 1H,I, 0.1% DMSO). Baseline mIPSC frequency and amplitude was not different between WT and Fmr1 KO mice (Fig. S1E,F). If chronic, CB1R-mediated suppression of GABAergic transmission exists in the Fmr1 KO, a greater increase in either mIPSC amplitude or frequency would be expected in Fmr1 KO slices with rimonabant. We observed an increase in mIPSC frequency in both WT and Fmr1 KO slices, but the effects of rimonabant on mIPSC frequency were greater in the Fmr1 KO (Fig. 1I; 2way ANOVA; Main effect of genotype; F (1, 32) = 4.598; p<0.05; genotype X time interaction; F (3, 96) = 3.366; p< 0.05). This result is consistent with enhanced CB1R- mediated suppression of inhibitory synaptic transmission in the Fmr1 KO cortex and suggests that this mechanism contributes to network hyperexcitability.

Genetic reduction of Cnr1 in GABAergic neurons, but not glutamatergic neurons, rescues prolonged Up states in Fmr1 KO mice.

To determine if enhanced function of CB1Rs in inhibitory neurons contributes to prolonged Up states in Fmr1 KO, we conditionally deleted Cnr1 in GABAergic neurons using the VGAT-Cre mouse line. To confirm this genetic targeting strategy, we bred Cnr1 floxed mice (Cnr1fl/fl) with both VGATCre and Rosa26-EYFP Cre reporter (Rosa26EYFP) mice and performed RNAscope for Cnr1 together with immunohistochemistry for EYFP to mark Cre expressing GABAergic neurons. On the wildtype background, (VGATCre:Rosa26EYFP:Cnr1+/+) Cnr1 expression is high in cortical GABAergic neurons as shown 54. With VGAT-Cre mediated deletion of Cnr1 (VGATCre:Rosa26EYFP:Cnr1fl/fl) these high Cnr1 expressing cells were absent and there was no eYFP-Cnr1 co-expressing neurons (Fig. 2A). Because CB1Rs are implicated in development of cortical circuits 55, 56, we used a heterozygous deletion strategy to reduce, but not completely eliminate, CB1Rs in GABAergic neurons in Fmr1 KO mice (VGATCre:Cnr1fl/+:Fmr1 KO). Heterozygous Cnr1 expression results in half the CB1R protein25.

Figure 2: Genetic reduction of Cnr1 in GABAergic neurons, but not glutamatergic neurons, rescues prolonged Up states in the Fmr1 KO.

Figure 2:

Open in a new tab

A) Cnr1 gene deletion in GABAergic neurons was accomplished by crossing the VGAT-Cre and floxed- Cnr1 mouse lines. Images of homozygous Cnr1 deletion in VGAT-Cre (+) somatosensory cortical neurons using IHC for the Cre reporter EYFP (Green) and RNAscope for Cnr1 RNA (Red). Filled arrowheads: Cre/EYFP(+), or GABAergic, neurons. Open arrowheads= Cre/EYFP (−), or putative excitatory, neurons. Scale bar= 50μm. B,C) Heterozygous deletion of Cnr1 in GABAergic neurons (VGATCre Cnr1fl/+) rescues Up state duration in Fmr1 KO slices. B) Example traces of the four genotypic groups C) Group average (±SEM) of Up state duration (Cre(−) WT; N=30/6; VGATCre:Cnr1fl/+: WT; N=17/5; Cre (−) Fmr1 KO; N= 32/6; VGATCre:Cnr1fl/+: Fmr1 KO; 20/5). D,E) Heterozygous deletion of Cnr1 in glutamatergic neurons (VGlut2Cre Cnr1fl/+) did not affect Up state duration in WT or Fmr1 KO (Cre(−) WT; N=39/9; VGlut2Cre:Cnr1fl/+: WT; N=28/6; Cre (−) Fmr1 KO; N= 39/9; VGlut2Cre:Cnr1fl/+: Fmr1 KO; 21/5). F,G) Deletion of Fmr1 in glutamatergic neurons is sufficient to cause long Up states. Floxed (fl) Fmr1 mice were crossed to VGlut1-Cre mice to delete Fmr1 in cortical glutamatergic neurons (VGlut1Cre Fmr1fl/y). F) Example traces from each genotype. G) Scatterplot and bar-graph showing mean duration is increased with glutamatergic neuron-specific Fmr1 deletion (Cre(−) Fmr1 fl/y: N=19/5; VGlut1Cre:Fmr1 fl/y N= 25/6). Scale bars for all panels = 50 μV and 1 s. *p<0.05; **p<0.01; ***p<0.005. 2-way ANOVA with posthoc comparisons.

If CB1R-mediated, presynaptic suppression at GABAergic synaptic transmission is enhanced in Fmr1 KO mice, we would expect reduced Cnr1 dosage in GABAergic neurons to increase inhibition and consequently, reduce and normalize Up state duration in Fmr1 KO slices. In Cre(−) controls, with wildtype levels of Cnr1 in GABAergic neurons, we reproduced the long Up state phenotype in Fmr1 KO mice (Fig. 2B,C). Heterozygous deletion of Cnr1 selectively in GABAergic neurons (VGATCre:Cnr1fl/+: Fmr1 KO) reduced Up state duration as compared to Cre (−) Fmr1 KO and durations were similar to Cre(−) WT. (Fig. 2B,C). These results are consistent with enhanced CB1R-mediated signaling in GABAergic neurons causing longer Up states. Interestingly, reduction of Cnr1 in GABAergic neurons on a Fmr1 WT background (VGATCre:Cnr1fl/+: WT) had no effect on Up state duration suggesting that CB1R-mediated suppression of GABAergic transmission does not normally contribute to Up state duration in this context. Up state amplitude was reduced in VGATCre:Cnr1fl/+ mice on both the WT and Fmr1 KO backgrounds (2way ANOVA, Main effect of Cnr1; F (1, 69) = 9.150; p< 0.01; Fig. S2A,B), similar to what we observed with rimonabant (Fig. S1A).

CB1R-mediated, presynaptic suppression of glutamatergic synaptic transmission is decreased in the frontal cortex of Fmr1 KO mice which may contribute to cortical circuit hyperexcitability21. To examine the contribution of reduced CB1R- function in excitatory neurons to Up states in Fmr1 KO neurons, we used the same strategy as above except we used VGlut2-Cre mice to delete Cnr1 in cortical glutamatergic neurons 42. Deletion of Cnr1 in glutamatergic neurons was confirmed using RNAscope (Fig. S2C). To examine effects on Up states, we again used a heterozygous deletion strategy (VGlut2Cre:Cnr1fl/+) to reduce Cnr1. If reduced Cnr1 function in excitatory neurons contributes to long Up states in the Fmr1 KO, then we may expect Up states in VGlut2Cre:Cnr1fl/+: Fmr1 −/y mice to be longer than in Cre(−) Fmr1 KO slices. In contrast to this prediction, a reduction of Cnr1 in glutamatergic neurons had no effect on Up state duration on either a WT or Fmr1 KO background (Fig. 2D,E). VGlut2Cre:Cnr1fl/+ mice also had normal Up state amplitude and frequency on both the WT or Fmr1 KO backgrounds (Fig. S2D,E). These results suggest that reduced CB1R function in glutamatergic neurons does not contribute to long Up states in Fmr1 KO mice and supports a role for enhanced CB1R function specifically in inhibitory neurons.

Deletion of Fmr1 in glutamatergic neurons is sufficient to cause long Up states.

Although our results implicate CB1R function in inhibitory neurons, we hypothesize that loss of FMRP in glutamatergic neurons leads to increased synthesis or release of endocannabinoids near GABAergic presynaptic terminals to suppress release. In support of this idea, our previous study found that simultaneous deletion of Fmr1 in glutamatergic neurons and glia, but not deletion restricted to GABAergic neurons, recapitulated the long Up state duration phenotype 14, 57. Subsequent work implicated Fmr1 deletion in astrocytes in circuit hyperexcitability, including long Up states 58. To re-address this issue, we created mice with conditional deletion of Fmr1 in forebrain glutamatergic neurons (VGlut1Cre:Fmr1 fl/y) using VGlut1-Cre 41 and floxed Fmr1 (Fmr1 fl/y) mice and measured Up states in comparison to Cre(−) male littermates. We found that Fmr1 deletion in glutamatergic neurons was sufficient to induce longer Up states and mimic the Fmr1 KO phenotype (Fig. 2F,G). Up state frequency and amplitude in VGlut1Cre:Fmr1 fl/y slices were not different from Cre(−) littermates (Fig. S2F,G). Therefore, our results support the hypothesis that postsynaptic loss of Fmr1 in glutamatergic neurons leads to enhanced CB1R-mediated presynaptic suppression of inhibitory synapses.

Reduced Cnr1 dosage and CB1R antagonism rescue auditory-driven synchronous cortical activity in Fmr1 KO mice

Our results in slices support a role for enhanced CB1R function in GABAergic neurons underlying altered cortical network activity in the Fmr1 KO. We next utilized EEG recordings to determine if enhanced CB1R function in GABAergic neurons contributed to altered resting and sensory-driven cortical activity in vivo in the Fmr1 KO. Importantly, these EEG phenotypes are observed in individuals with FXS and results may inform mechanism and therapeutic targets for human EEG alterations 5, 7, 10. EEG measurements using a 30-channel array of surface electrodes contacting the skull were performed as previously described 8. The array was divided up into 3 regions in each hemisphere (Frontal, Medial and Temporal) (Fig. 3A). We measured the ability of auditory stimuli to consistently drive synchronized cortical activity among all stimulus trials of sound presentation using intertrial phase coherence (ITPC) – a measure that focuses on the timing of frequency components independent of raw power.

Figure 3. Genetic reduction of Cnr1 in GABAergic neurons or rimonabant treatment rescue the decreased synchronization of chirp-driven cortical activity in the Fmr1 KO.

Figure 3.

Open in a new tab

A) Diagram of EEG electrode array laid over scalp of mice. B) Chirp stimuli were 2 seconds of broadband noise sinusoidally modulated starting at 1 Hz and ramping up to 100 Hz. C) Cnr1 genetic reduction experiment. Top: ITPC average plots of induced activity in the right medial region (red box in A) in the 3 genotypes examined (WT, Fmr1 KO, and Fmr1 KO with heterozygous deletion of Cnr1 in GABAergic neurons; VGATCre/Cnr1fl/+/Fmr1 KO). Bottom: Average difference plots (ΔITPC) based on plots above indicate decreased ITPC at higher frequencies in Fmr1 KO mice (left) and the rescue of this phenotype with a genetic reduction of Cnr1 in GABAergic neurons. N=17,17,17 mice. D) Rimonabant treatment experiment. Left: ITPC average plots of induced activity for pre- and post-rimonabant treatment obtained from the left medial region (blue box in A) in Fmr1 KO mice (N=14). Right: Difference plot indicating increases in ITPC with rimonabant treatment. Solid lines border statistically different areas of the plot. Scales in C apply to D.

We first examined ITPC of activity induced by a chirp sound (Fig. 3B). We performed the same Cnr1 gene dosage reduction in GABAergic neurons and compared ITPC among littermates of these 3 genotypes: Cre (−) “WT”, Cre(−) Fmr1 KO, and Fmr1 KO with heterozygous deletion of Cnr1 in GABAergic neurons (VGATCre:Cnr1fl/+:Fmr1 KO). We focus first on the result averaged from the right medial region of the electrode array (Fig. 3A, red box). The average ITPC plots have a notable diagonal of higher ITPC that corresponds to modulation at the frequency currently occurring in the chirp (Fig. 3C, top). From the ITPC plots, average difference plots with plot areas of statistical difference marked by bold black line were calculated for each of the three possible genotypic comparisons (Fig. 3C, bottom). As reported previously for both FXS individuals and Fmr1 KO mice, the “Fmr1 KO – WT” difference plot revealed a decrease in ITPC in the higher frequency components corresponding to the frequency modulation occurring later in the chirp (Fig. 3C, bottom left) 6, 7 8. This phenotype was observed in all 6 electrode array regions (Fig. S3). This decrease indicated an impaired ability for higher frequency sound modulation to consistently drive cortical synchrony from one chirp trial to the next.

The two other difference plots (Fig. 3C, bottom) indicate that the decrease in Cnr1 dosage rescued the phenotype. First, the “VGATCre:Cnr1fl/+:Fmr1 KO– WT” difference plot (Fig. 3C, bottom middle) did not have any regions of difference indicating that reduced Cnr1 dosage rescued the decreased ITPC in the Fmr1 KO mouse. Second, consistent with the changes just described, the “VGATCre:Cnr1fl/+:Fmr1 KO– Fmr1 KO” difference plot (Fig. 3C, bottom right) indicates increased ITPC in Fmr1 KO mice with decreased Cnr1 dosage. Therefore, decreased Cnr1 dosage in GABAergic neurons increased the ability of sounds to consistently drive cortical activity in Fmr1 KO mice. Interestingly, this rescue occurred in both medial regions of the electrode array but did not occur in frontal or temporal regions (Fig. S3).

Acute pharmacological blockade of CB1R rescues Up state duration in Fmr1 KO slices (Fig. 1). We next determined if rimonabant treatment corrects the ITPC deficit in the Fmr1 KO. To determine the potential of chronic CB1R antagonism as a therapeutic strategy, we tested a 7-day treatment with the CBR1 antagonist. We recorded pre- and post-treatment from Fmr1 KO mice. Rimonabant increased ITPC in the Fmr1 KO, which is consistent with a rescue to wild-type levels (Fig. 3D). Unlike gene dosage, this rescue occurred in all regions of the electrode array (Fig. S4).

We obtained similar results examining ITPC of an auditory steady state response (ASSR). The ASSR was induced with either a 40 Hz or 80 Hz click train. The clearest data were collected with the 80 Hz ASSR, but similar results were also observed with 40 Hz ASSR (Figs. 4,5, S5-S8). There was a robust decrease in ITPC in the 80 Hz band in Cre (−) Fmr1 KO compared to Cre (−) WT as seen in the difference plot (Fig. 4C, bottom left). ITPC was not different when comparing VGATCre:Cnr1fl/+:Fmr1 KO to WT (Fig. 4C, bottom middle) and ITPC in the VGATCre:Cnr1fl/+:Fmr1 KO mice was greater than that observed in Fmr1 KO littermates (Fig. 4C, bottom right). The 80 Hz ASSR decrease in the Fmr1 KO mice was seen and rescued by heterozygous deletion of Cnr1 in GABAergic neurons in all brain regions measured with the electrode array (Fig. S5). Similar results were observed with ITPC of a 40 Hz ASSR (Figs. 5C; S7). Like that observed with genetic reduction of CB1Rs in GABAergic neurons, a 7-day rimonabant treatment of Fmr1 KO mice increased ITPC for 40 and 80 Hz ASSR (Figs. 4D, 5D, S6, S8) suggesting that rimonabant treatment could also rescue normal auditory-driven synchrony.

Figure 4. Both genetic reduction of Cnr1 in GABAergic neurons and rimonabant rescue decreased synchronization of pulse-driven cortical activity at 80 Hz in the Fmr1 KO.

Figure 4.

Open in a new tab

A) Diagram of EEG electrode array. Data shown are from the left temporal region electrodes (red box). B) An 80 Hz pulse train was used to induce an ASSR (auditory steady-state response). C) Top: ITPC average plots of induced activity in the 3 genotypes examined. Bottom: Average difference plots indicate decreased ITPC during the ASSR in Fmr1 KO mice (left) and the rescue of this phenotype in Fmr1 KO mice with genetic reduction of Cnr1 in GABAergic neurons (VGATCre/Cnr1fl/+/Fmr1 KO). N=17,17,17 mice. D) Left: ITPC average plots of induced activity for pre- and post-rimonabant treatment in Fmr1 KO mice (N=14). Right: Difference plot indicating an increase in ITPC with rimonabant treatment. Solid lines border statistically different areas of the plot. Scales in C apply to D.

Figure 5. Both genetic reduction of Cnr1 in GABAergic neurons and rimonabant rescue decreased synchronization of pulse-driven cortical activity at 40 Hz in the Fmr1 KO.

Figure 5.

Open in a new tab

A) Diagram of EEG electrode array. The data shown are from the left frontal region electrodes (red box). B) A 40 Hz pulse train was used to induce an ASSR. C) Genetic reduction experiment. Top: ITPC average plots of induced activity in the 3 genotypes examined. Bottom: Average difference plots indicate decreased ITPC during the ASSR in Fmr1 KO mice (left) and the rescue of this phenotype in Fmr1 KO with genetic reduction of Cnr1 in GABAergic neurons (VGATCre/Cnr1fl/+/Fmr1 KO). N=17,17,17 mice. D) Rimonabant treatment experiment. Left: ITPC average plots of induced activity for pre- and post-rimonabant treatment obtained from the left medial region in Fmr1 KO mice (N=14). Right: Difference plot indicating increases in ITPC with rimonabant treatment. Solid lines border statistically different areas of the plot. Scales in C apply to D.

In summary, these data indicate that enhanced CB1R signaling at GABAergic synapses is involved in systems level auditory response changes in the Fmr1 KO mouse. Also, a pharmacological block of CBR1 later in life can aid synchronization deficits.

CBR1 antagonism may rescue resting activity in vivo, but reduced Cnr1 dosage in GABAergic neurons does not

Under resting conditions when no experimental and salient stimuli are being presented, power in the standard EEG frequency bands – and gamma in particular - is higher in both FXS individuals and in the Fmr1 KO mouse 5, 6, 59. We determined if reduced Cnr1 dosage in GABAergic neurons could rescue this resting power phenotype. For most electrode array regions, increased power was observed in the Fmr1 KO across all frequency bands when compared to WT controls, but most consistently in the lower gamma band (Fig. 6A, S9). But there was no clear example of rescue in any of the standard frequency bands. Almost all the average power values measured in VGATCre:Cnr1fl/+:Fmr1 KO mice were different from those in WT mice, but not Fmr1 KO mice, indicating that heterozygous deletion of Cnr1 in GABAergic neurons did not rescue the resting state power phenotypes in the Fmr1 KO.

Figure 6. Distinct effects of genetic and pharmacological reduction of CB1R activity on resting state EEG.

Figure 6.

Open in a new tab

Spectral density power (normalized to wild-type (WT) mice (A) or to pre-treatment Fmr1 KO mice (B)) is plotted for 6 standard frequency bands and obtained from the right temporal electrode group (as indicated by red box in lower right). A) An increase in power is observed in the low and high gamma frequency bands in the Fmr1 KO mice (black) and as compared to WT littermates. Fmr1 KO mice with genetic reduction of Cnr1 in GABAergic neurons (VGAT Cre/Cnr1fl/+/Fmr1 KO) (gray) have increased power across all frequency bands, as compared to WT, and is not different from Fmr1 KO, with the exception of high gamma. N=17, 17, 17 mice. B) 7 day rimonabant treatment of Fmr1 KO mice decreases power in all frequency bands except High Gamma. *, p<0.05 **, p<0.01 ****, p<0.001

We next examined the effect of 7-day rimonabant treatment on resting power observed in the Fmr1 KO (Figs. 6B, S10). Unlike the reduced Cnr1 dosage, rimonabant clearly decreased the power of all frequency bands except high gamma, and this change was consistent with an ability to rescue the power phenotypes towards WT levels. This pattern was observed at all electrode regions. While the data suggest that altered, ongoing CBR1 signaling is involved in resting power changes in the Fmr1 KO mice, this signaling is more effectively remedied pharmacologically in adults, and not by cell type specific genetic reduction throughout life.

Discussion

Our results provide evidence for enhanced, cell-type specific eCB signaling underlying cortical network function phenotypes in Fmr1 KO mice. Specifically, our findings support the hypothesis that enhanced CB1R-mediated suppression of GABAergic synaptic transmission mediates prolonged cortical Up states and deficits in sensory-driven cortical network synchrony in the Fmr1 KO mouse. This conclusion is supported by the “rescue” (or normalization) of circuit activity and EEG phenotypes using pharmacological blockade of CB1Rs or 2-AG synthesis and Cnr1 gene dosage reduction in GABAergic neurons. We also observe that GABAergic synaptic currents in Fmr1 KO L2/3 cortical neurons are more sensitive to antagonism of CBR1 receptors consistent with enhanced CB1R-suppression of GABAergic transmission. While selective Cnr1 deletion in glutamatergic neurons has no effect on Up states, Fmr1 deletion in the same neurons is sufficient to mimic the long Up state phenotype observed in constitutive Fmr1 KO. Taken together with previous work, our results support a model where loss of FMRP in cortical glutamatergic neurons leads to altered or enhanced mGluR5-dependent synthesis of 2-AG and enhanced suppression of GABA release from CB1R-positive inhibitory presynaptic terminals. Importantly, our results implicate a specific synaptic mechanism in controlling the duration of cortical Up states as well as the ability of cortical circuits to appropriately synchronize with time-modulated stimuli in the gamma frequency range, as measured with ITPC. Because sensory-driven cortical synchrony and ITPC are deficient in individuals with FXS, our results suggest that enhanced eCB function in cortical circuits may contribute to abnormal temporal processing of sensory stimuli in FXS.

Altered modulation of CB1R-regulated inhibitory circuits as a mediator of prolonged circuit activity in FXS

Up states are mediated by local cortical excitatory and inhibitory circuits and thus reflect the balance of excitation and inhibition 60. If or how inhibitory neurons control the termination of Up states or their duration has been well studied in wildtype animals61. Interestingly, pharmacological blockade of GABAa receptors reduces Up state duration, whereas blockade of GABAB receptors (GABAbRs) increases Up state duration 62, 63. This suggests that CB1R-mediated suppression of GABABR activation may contribute to Up state duration in Fmr1 KO 64. Consistent with this idea, the GABAB agonist baclofen increases cortical synchronization (chirp ITPC) in the Fmr1 KO 65. Opto- or chemogenetic inhibition of specific inhibitory neuron types, such as parvalbumin (PV) and somatostatin (SST)-positive interneurons, enhance pyramidal neuron firing during Up states and prolong Up state duration both in vivo and in slices 66, 67. While CB1Rs are highly expressed in Cholecystokinin (CCK)+ inhibitory neurons, other neocortical inhibitory neuron types, such as SST, VIP or Calbindin(+) inhibitory interneurons express CB1Rs and are regulated by endocannabinoids68-70. Therefore, enhanced CB1R suppression of the outputs of any of these inhibitory neuron types may contribute to prolonged Up states and EEG phenotypes in Fmr1 KO mice. Pharmacological or genetic reduction of CB1R activity did not affect Up state durations in WT mice, indicating that CB1Rs do not regulate Up state duration in normal developing cortex as reported in organotypic slices45. Therefore, results suggest that CB1Rs are abnormally active during Up states in Fmr1 KO cortex under our experimental conditions. Acute antagonism of CB1Rs or DAGLα reduced Up state duration in Fmr1 KO mice, but did not completely correct durations to WT levels, in contrast to genetic reduction of Cnr1 in inhibitory neurons. Similarly, genetic inhibitory neuron deletion or chronic antagonism of CB1Rs with rimonabant was efficient in correcting EEG phenotypes. These results suggest a requirement for chronic inhibition of CB1R function for the best therapeutic benefit. CB1R antagonism reduced Up state durations in slices from both young (3 week) and adult (9-11 week) mice, and corrected EEG phenotypes in adult mice. These results suggest that enhanced CB1R function contributes to cortical circuit dysfunction in Fmr1 KO mice across the postnatal lifespan.

Although CB1R-dependent suppression of excitatory synaptic transmission is reportedly reduced in Fmr1 KO cortex 21, genetic reduction of CB1Rs in cortical glutamatergic neurons did not mimic or exacerbate the long Up states in the Fmr1 KO. This result suggests that the primary effects of CB1Rs on Up states in the Fmr1 KO are via regulation of inhibitory synaptic transmission. CB1Rs are highly concentrated on presynaptic terminals of CCK+ inhibitory neurons54, 71. Because we observe strong potentiation of inhibitory synaptic currents with rimonabant treatment of Fmr1 KO slices in the absence of action potentials (in TTX), there may be enhanced tonic CB1R-dependent suppression of inhibitory synaptic transmission in the Fmr1 KO 64, 72, 73. Because the DAGLα inhibitor, DO34, rescues Up state duration in Fmr1 KO slices, there may be enhanced 2-AG synthesis in Fmr1 KO excitatory neurons that tonically suppress GABAergic transmission. While we observed enhanced suppression of IPSCs by eCB signaling in Fmr1 KO slices, this was not observed in a previous study in hippocampus25 and could be due to different experimental conditions, or brain regions.

Molecular basis for enhanced CB1R-dependent suppression of inhibition in Fmr1 KO

mGluR5 signaling contributes to prolonged Up states14 and stimulates 2-AG synthesis through activation of phospholipase Cβ and DAGLα 46. Because mGluR5-induced and CB1R-dependent suppression of inhibitory synaptic transmission is enhanced in the striatum and hippocampus of Fmr1 KO mice22-24, we hypothesize that this is the case in the neocortex. In Fmr1 KO mice, CB1R levels are normal and CBR1 agonist-induced suppression of inhibition is normal22, 24. These results argue against upregulation of CB1R function in the Fmr1 KO but enhanced, and perhaps tonic, mGluR5- dependent 2AG synthesis selectively at inhibitory synapses. In support of this idea, basal 2-AG levels are elevated or saturated in Fmr1 KO striatum and cortex, and insensitive to the stimulatory effects of mGluR5 agonism21, 23. A molecular basis for enhanced mGluR5 and CB1R-dependent suppression of inhibitory synaptic transmission in the Fmr1 KO may be related to mislocalization of DAGLα and mGluR5 away from the postsynaptic density21, 74, which may enhance their concentration and function near CCK+ GABAergic terminals with high levels of CB1Rs 75.

Enhanced CB1R function in inhibitory neurons mediates deficits in sensory-driven cortical synchronization and coherence in awake Fmr1 KO mice

Here we demonstrate that the pharmacological and genetic manipulations of CB1Rs that correct Up state duration also correct EEG phenotypes that are conserved in FXS individuals. Cortical synchronization (or ITPC) during the chirp, 40 and 80 Hz ASSR in the Fmr1 KO was strongly and consistently increased by both rimonabant and heterozygous deletion of Cnr1 in inhibitory neurons. This result suggests that the local cortical circuit mechanisms that mediate prolonged Up states in the Fmr1 KO also contribute to an inability of cortical circuits to consistently synchronize with rapidly modulated sensory stimuli. Cannabinoids are known to regulate neural synchrony and gamma oscillations; studied in the context of the effects of cannabis use on cognition or in schizophrenia76, 77. Specifically, intravenous application of Δ-9-tetrahydrocanabinol (Δ9-THC) in humans, the primary psychoactive constituent of marijuana and a CB1R agonist, reduced evoked power during the 40 Hz ASSR as well as ITPC 78. This result is consistent with the idea that overactive CB1R signaling Fmr1 KO mice leads to reduced ASSR and reduced cortical synchronization and may function similarly in FXS individuals.

Although ASSR ITPC deficits were rescued across all brain regions by Cnr1 deletion in GABAergic neurons, the chirp ITPC deficits were rescued in only medial cortical regions. There are reported cortical region differences in expression of CCK+ inhibitory interneurons which may contribute differently to the chirp ITPC 79. Similarly, frontal cortical areas have higher levels of CB1Rs as compared to primary sensory cortex and genetic reduction may have less of a functional impact in frontal areas80. Alternatively, because there is a known decreased function of CB1Rs at excitatory synapses in the frontal cortex, this may make a larger contribution to the frontal cortex ITPC deficits in the Fmr1 KO21. We also observed distinct effects of pharmacological vs genetic reduction of CB1R activity on resting state EEG. While rimonabant reduced resting state gamma power in Fmr1 KO mice, genetic reduction of Cnr1 in inhibitory neurons did not. This may be due to differences in the effects of CB1R antagonism in adult vs throughout development 78, 81. Alternatively, antagonism of CB1Rs on excitatory neurons by rimonabant may mediate effects on resting state EEG.

CB1Rs and behavioral phenotypes in Fmr1 KO mice

Our results are consistent with roles for enhanced CB1R function in certain behavioral phenotypes in Fmr1 KO mice, including auditory hypersensitivity (e.g. audiogenic seizures) and cognitive function25, 36. Interestingly, other results suggest that deficient CB1R function contributes to hyperlocomotion and open-field behavior in Fmr1 KO mice25 21. These results suggest that the imbalances in CB1R regulation of excitatory and inhibitory circuits varies across brain regions and thus contributes to different behaviors. Genetic and circuit-specific tools are necessary to better reveal brain region specific roles for CB1Rs in behavioral phenotypes in Fmr1 KO mice. These results also suggest that different therapeutic strategies may be required to aid different behaviors in FXS.

ASD risk genes regulate E/I balance through CB1Rs

In addition to Fmr1, loss of function of other ASD risk genes such as Neuroligin3 or Neurexin 1B affect CB1R-regulation of synaptic transmission53, 82, 83. With Nlgn3 mutations, tonic CB1R-mediated suppression of IPSCs is deficient whereas loss of function of Nrxn1B enhances tonic CB1R suppression of excitatory synaptic transmission. These results reveal how ASD risk genes can affect the E/I balance but do so through different synaptic mechanisms. Based on the known role of eCBs in regulation of neural oscillations76, 77 and our observations that we can ameliorate deficits in sensory-driven network oscillations in Fmr1 KO mice by manipulation of CB1Rs, dysfunction of CB1Rs may contribute to altered neural oscillations and deficits in sensory processing and cognition in these mouse or individuals with loss of function in these other ASD risk genes. Furthermore, our results suggest that targeting selective CB1R antagonism may be an effective therapeutic strategy to restore cortical synchronization and aid sensory processing in individuals with FXS.

Supplementary Material

Supplemental Methods and Figures

Acknowledgements:

We would like to acknowledge Jacob E. Bowles and Christopher Williams for technical assistance with genotyping and Dr. Laurent Gautron for assistance with RNAscope. This work was supported by NIH grants U54HD082008 and U54HD104461 (KMH, JRG, DKB) and P30DK127984 (UTSW Metabolic Core).

Footnotes

Conflict of Interest: The authors have no competing financial interests in relation to the work described herein.

Ethics Approval: All experimental procedures have been approved by the Institutional Animal Care and Use Committees at UT Southwestern (Protocol #’s 2015-101252; 2017-101986) or UC Riverside (Protocol #: #20190015) and conducted in full accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Data Availability:

The authors declare that all data supporting the findings of this study are available in the main text or the supplementary materials. Datasets are available from the corresponding authors upon request. Correspondence and requests for materials should be addressed to KMH. and DKB (Kimberly.Huber@UTSouthwestern.edu and devin.binder@ucr.edu). Labview software for analysis of Up states is available at https://github.com/jaygibson244/GibsonHuber-UP-states.git. The MATLAB code used to perform statistical analysis for chirp, 40 and 80Hz ASSR is available at Github: https://github.com/CarrieRJonak/MEA-EEG_ITPC.

References:

  • 1.Bassell GJ, Warren ST. Fragile X syndrome: loss of local mRNA regulation alters synaptic development and function. Neuron 2008; 60(2): 201–214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Juczewski K, von Richthofen H, Bagni C, Celikel T, Fisone G, Krieger P. Somatosensory map expansion and altered processing of tactile inputs in a mouse model of fragile X syndrome. Neurobiol Dis 2016; 96: 201–215. [DOI] [PubMed] [Google Scholar]
  • 3.He CX, Cantu DA, Mantri SS, Zeiger WA, Goel A, Portera-Cailliau C. Tactile Defensiveness and Impaired Adaptation of Neuronal Activity in the Fmr1 Knock-Out Mouse Model of Autism. J Neurosci 2017; 37(27): 6475–6487. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.McCullagh EA, Rotschafer SE, Auerbach BD, Klug A, Kaczmarek LK, Cramer KS et al. Mechanisms underlying auditory processing deficits in Fragile X syndrome. Faseb j 2020; 34(3): 3501–3518. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Wang J, Ethridge LE, Mosconi MW, White SP, Binder DK, Pedapati EV et al. A resting EEG study of neocortical hyperexcitability and altered functional connectivity in fragile X syndrome. J Neurodev Disord 2017; 9: 11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Lovelace JW, Ethell IM, Binder DK, Razak KA. Translation-relevant EEG phenotypes in a mouse model of Fragile X Syndrome. Neurobiol Dis 2018; 115: 39–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Ethridge LE, White SP, Mosconi MW, Wang J, Pedapati EV, Erickson CA et al. Neural synchronization deficits linked to cortical hyper-excitability and auditory hypersensitivity in fragile X syndrome. Mol Autism 2017; 8: 22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Jonak CR, Lovelace JW, Ethell IM, Razak KA, Binder DK. Multielectrode array analysis of EEG biomarkers in a mouse model of Fragile X Syndrome. Neurobiol Dis 2020; 138: 104794. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Jonak CR, Assad SA, Garcia TA, Sandhu MS, Rumschlag JA, Razak KA et al. Phenotypic analysis of multielectrode array EEG biomarkers in developing and adult male Fmr1 KO mice. Neurobiol Dis 2024; 195: 106496. [DOI] [PubMed] [Google Scholar]
  • 10.Ethridge LE, De Stefano LA, Schmitt LM, Woodruff NE, Brown KL, Tran M et al. Auditory EEG Biomarkers in Fragile X Syndrome: Clinical Relevance. Front Integr Neurosci 2019; 13: 60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Pedapati EV, Schmitt LM, Ethridge LE, Miyakoshi M, Sweeney JA, Liu R et al. Neocortical localization and thalamocortical modulation of neuronal hyperexcitability contribute to Fragile X Syndrome. Commun Biol 2022; 5(1): 442. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Schmitt LM, Li J, Liu R, Horn PS, Sweeney JA, Erickson CA et al. Altered frontal connectivity as a mechanism for executive function deficits in fragile X syndrome. Mol Autism 2022; 13(1): 47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Gibson JR, Bartley AF, Hays SA, Huber KM. Imbalance of neocortical excitation and inhibition and altered UP states reflect network hyperexcitability in the mouse model of fragile X syndrome. J Neurophysiol 2008; 100(5): 2615–2626. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Hays SA, Huber KM, Gibson JR. Altered Neocortical Rhythmic Activity States in Fmr1 KO Mice Are Due to Enhanced mGluR5 Signaling and Involve Changes in Excitatory Circuitry. The Journal of Neuroscience 2011; 31(40): 14223–14234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Osterweil EK, Chuang SC, Chubykin AA, Sidorov M, Bianchi R, Wong RK et al. Lovastatin corrects excess protein synthesis and prevents epileptogenesis in a mouse model of fragile X syndrome. Neuron 2013; 77(2): 243–250. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Zhang Y, Bonnan A, Bony G, Ferezou I, Pietropaolo S, Ginger M et al. Dendritic channelopathies contribute to neocortical and sensory hyperexcitability in Fmr1(-/y) mice. Nat Neurosci 2014; 17(12): 1701–1709. [DOI] [PubMed] [Google Scholar]
  • 17.Contractor A, Klyachko VA, Portera-Cailliau C. Altered Neuronal and Circuit Excitability in Fragile X Syndrome. Neuron 2015; 87(4): 699–715. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Liu X, Kumar V, Tsai NP, Auerbach BD. Hyperexcitability and Homeostasis in Fragile X Syndrome. Front Mol Neurosci 2021; 14: 805929. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Cea-Del Rio CA, Nunez-Parra A, Freedman SM, Kushner JK, Alexander AL, Restrepo D et al. Disrupted inhibitory plasticity and homeostasis in Fragile X syndrome. Neurobiol Dis 2020; 142: 104959. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Rotschafer S, Razak K. Altered auditory processing in a mouse model of fragile X syndrome. Brain Res 2013; 1506: 12–24. [DOI] [PubMed] [Google Scholar]
  • 21.Jung KM, Sepers M, Henstridge CM, Lassalle O, Neuhofer D, Martin H et al. Uncoupling of the endocannabinoid signalling complex in a mouse model of fragile X syndrome. Nat Commun 2012; 3: 1080. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Zhang L, Alger BE. Enhanced endocannabinoid signaling elevates neuronal excitability in fragile X syndrome. J Neurosci 2010; 30(16): 5724–5729. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Maccarrone M, Rossi S, Bari M, De Chiara V, Rapino C, Musella A et al. Abnormal mGlu 5 receptor/endocannabinoid coupling in mice lacking FMRP and BC1 RNA. Neuropsychopharmacology 2010; 35(7): 1500–1509. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Tang AH, Alger BE. Homer protein-metabotropic glutamate receptor binding regulates endocannabinoid signaling and affects hyperexcitability in a mouse model of fragile X syndrome. J Neurosci 2015; 35(9): 3938–3945. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Busquets-Garcia A, Gomis-Gonzalez M, Guegan T, Agustin-Pavon C, Pastor A, Mato S et al. Targeting the endocannabinoid system in the treatment of fragile X syndrome. Nat Med 2013; 19(5): 603–607. [DOI] [PubMed] [Google Scholar]
  • 26.Wilson RI, Kunos G, Nicoll RA. Presynaptic specificity of endocannabinoid signaling in the hippocampus. Neuron 2001; 31(3): 453–462. [DOI] [PubMed] [Google Scholar]
  • 27.Lafourcade M, Elezgarai I, Mato S, Bakiri Y, Grandes P, Manzoni OJ. Molecular components and functions of the endocannabinoid system in mouse prefrontal cortex. PLoS One 2007; 2(8): e709. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Robbe D, Kopf M, Remaury A, Bockaert J, Manzoni OJ. Endogenous cannabinoids mediate long-term synaptic depression in the nucleus accumbens. Proc Natl Acad Sci U S A 2002; 99(12): 8384–8388. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Bender VA, Bender KJ, Brasier DJ, Feldman DE. Two coincidence detectors for spike timing-dependent plasticity in somatosensory cortex. J Neurosci 2006; 26(16): 4166–4177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Koukouli F, Montmerle M, Aguirre A, De Brito Van Velze M, Peixoto J, Choudhary V et al. Visual-area-specific tonic modulation of GABA release by endocannabinoids sets the activity and coordination of neocortical principal neurons. Cell Rep 2022; 40(8): 111202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Bear MF, Huber KM, Warren ST. The mGluR theory of fragile X mental retardation. Trends Neurosci 2004; 27(7): 370–377. [DOI] [PubMed] [Google Scholar]
  • 32.Ronesi JA, Collins KA, Hays SA, Tsai NP, Guo W, Birnbaum SG et al. Disrupted Homer scaffolds mediate abnormal mGluR5 function in a mouse model of fragile X syndrome. Nat Neurosci 2012; 15(3): 431–440, S431. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Jung KM, Astarita G, Zhu C, Wallace M, Mackie K, Piomelli D. A key role for diacylglycerol lipase-alpha in metabotropic glutamate receptor-dependent endocannabinoid mobilization. Mol Pharmacol 2007; 72(3): 612–621. [DOI] [PubMed] [Google Scholar]
  • 34.Castillo PE, Chiu CQ, Carroll RC. Long-term plasticity at inhibitory synapses. Curr Opin Neurobiol 2011; 21(2): 328–338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Chevaleyre V, Castillo PE. Heterosynaptic LTD of hippocampal GABAergic synapses: a novel role of endocannabinoids in regulating excitability. Neuron 2003; 38(3): 461–472. [DOI] [PubMed] [Google Scholar]
  • 36.Gomis-Gonzalez M, Busquets-Garcia A, Matute C, Maldonado R, Mato S, Ozaita A. Possible Therapeutic Doses of Cannabinoid Type 1 Receptor Antagonist Reverses Key Alterations in Fragile X Syndrome Mouse Model. Genes (Basel) 2016; 7(9). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Pirbhoy PS, Jonak CR, Syed R, Argueta DA, Perez PA, Wiley MB et al. Increased 2-arachidonoyl-sn-glycerol levels normalize cortical responses to sound and improve behaviors in Fmr1 KO mice. J Neurodev Disord 2021; 13(1): 47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Bakker DB. Fmr1 knockout mice: a model to study fragile X mental retardation. The Dutch-Belgian Fragile X Consortium. Cell 1994; 78(1): 23–33. [PubMed] [Google Scholar]
  • 39.Mientjes EJ, Nieuwenhuizen I, Kirkpatrick L, Zu T, Hoogeveen-Westerveld M, Severijnen L et al. The generation of a conditional Fmr1 knock out mouse model to study Fmrp function in vivo. Neurobiol Dis 2006; 21(3): 549–555. [DOI] [PubMed] [Google Scholar]
  • 40.Castorena CM, Caron A, Michael NJ, Ahmed NI, Arnold AG, Lee J et al. CB1Rs in VMH neurons regulate glucose homeostasis but not body weight. Am J Physiol Endocrinol Metab 2021; 321(1): E146–e155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Harris JA, Hirokawa KE, Sorensen SA, Gu H, Mills M, Ng LL et al. Anatomical characterization of Cre driver mice for neural circuit mapping and manipulation. Front Neural Circuits 2014; 8: 76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Vong L, Ye C, Yang Z, Choi B, Chua S Jr., Lowell BB. Leptin action on GABAergic neurons prevents obesity and reduces inhibitory tone to POMC neurons. Neuron 2011; 71(1): 142–154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Zador F, Lenart N, Csibrany B, Santha M, Molnar M, Tuka B et al. Low dosage of rimonabant leads to anxiolytic-like behavior via inhibiting expression levels and G-protein activity of kappa opioid receptors in a cannabinoid receptor independent manner. Neuropharmacology 2015; 89: 298–307. [DOI] [PubMed] [Google Scholar]
  • 44.Reagan-Shaw S, Nihal M, Ahmad N. Dose translation from animal to human studies revisited. FASEB J 2008; 22(3): 659–661. [DOI] [PubMed] [Google Scholar]
  • 45.Pava MJ, den Hartog CR, Blanco-Centurion C, Shiromani PJ, Woodward JJ. Endocannabinoid modulation of cortical up-states and NREM sleep. PLoS One 2014; 9(2): e88672. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Lu HC, Mackie K. Review of the Endocannabinoid System. Biol Psychiatry Cogn Neurosci Neuroimaging 2021; 6(6): 607–615. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Ogasawara D, Deng H, Viader A, Baggelaar MP, Breman A, den Dulk H et al. Rapid and profound rewiring of brain lipid signaling networks by acute diacylglycerol lipase inhibition. Proc Natl Acad Sci U S A 2016; 113(1): 26–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Schurman LD, Carper MC, Moncayo LV, Ogasawara D, Richardson K, Yu L et al. Diacylglycerol Lipase-Alpha Regulates Hippocampal-Dependent Learning and Memory Processes in Mice. J Neurosci 2019; 39(30): 5949–5965. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Liu X, Chen Y, Vickstrom CR, Li Y, Viader A, Cravatt BF et al. Coordinated regulation of endocannabinoid-mediated retrograde synaptic suppression in the cerebellum by neuronal and astrocytic monoacylglycerol lipase. Sci Rep 2016; 6: 35829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Liu X, Dimidschstein J, Fishell G, Carter AG. Hippocampal inputs engage CCK+ interneurons to mediate endocannabinoid-modulated feed-forward inhibition in the prefrontal cortex. Elife 2020; 9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Wilson RI, Nicoll RA. Endogenous cannabinoids mediate retrograde signalling at hippocampal synapses. Nature 2001; 410(6828): 588–592. [DOI] [PubMed] [Google Scholar]
  • 52.Trettel J, Levine ES. Cannabinoids depress inhibitory synaptic inputs received by layer 2/3 pyramidal neurons of the neocortex. J Neurophysiol 2002; 88(1): 534–539. [DOI] [PubMed] [Google Scholar]
  • 53.Speed HE, Masiulis I, Gibson JR, Powell CM. Increased Cortical Inhibition in Autism-Linked Neuroligin-3R451C Mice Is Due in Part to Loss of Endocannabinoid Signaling. PLoS One 2015; 10(10): e0140638. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Marsicano G, Lutz B. Expression of the cannabinoid receptor CB1 in distinct neuronal subpopulations in the adult mouse forebrain. Eur J Neurosci 1999; 11(12): 4213–4225. [DOI] [PubMed] [Google Scholar]
  • 55.Hedrich J, Angamo EA, Conrad A, Lutz B, Luhmann HJ. Cell type specific impact of cannabinoid receptor signaling in somatosensory barrel map formation in mice. J Comp Neurol 2020; 528(1): 3–13. [DOI] [PubMed] [Google Scholar]
  • 56.Durieux LJA, Gilissen SRJ, Arckens L. Endocannabinoids and cortical plasticity: CB1R as a possible regulator of the excitation/inhibition balance in health and disease. Eur J Neurosci 2022; 55(4): 971–988. [DOI] [PubMed] [Google Scholar]
  • 57.Gorski JA, Talley T, Qiu M, Puelles L, Rubenstein JL, Jones KR. Cortical excitatory neurons and glia, but not GABAergic neurons, are produced in the Emx1-expressing lineage. J Neurosci 2002; 22(15): 6309–6314. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Jin SX, Higashimori H, Schin C, Tamashiro A, Men Y, Chiang MSR et al. Astroglial FMRP modulates synaptic signaling and behavior phenotypes in FXS mouse model. Glia 2021; 69(3): 594–608. [DOI] [PubMed] [Google Scholar]
  • 59.Sinclair D, Oranje B, Razak KA, Siegel SJ, Schmid S. Sensory processing in autism spectrum disorders and Fragile X syndrome-From the clinic to animal models. Neurosci Biobehav Rev 2017; 76(Pt B): 235–253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Haider B, Duque A, Hasenstaub AR, McCormick DA. Neocortical network activity in vivo is generated through a dynamic balance of excitation and inhibition. J Neurosci 2006; 26(17): 4535–4545. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Lemieux M, Chauvette S, Timofeev I. Neocortical inhibitory activities and long-range afferents contribute to the synchronous onset of silent states of the neocortical slow oscillation. Journal of Neurophysiology 2015; 113(3): 768–779. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Sanchez-Vives MV, Mattia M, Compte A, Perez-Zabalza M, Winograd M, Descalzo VF et al. Inhibitory Modulation of Cortical Up States. Journal of Neurophysiology 2010; 104(3): 1314–1324. [DOI] [PubMed] [Google Scholar]
  • 63.Mann EO, Kohl MM, Paulsen O. Distinct Roles of GABAA and GABAB Receptors in Balancing and Terminating Persistent Cortical Activity. The Journal of Neuroscience 2009; 29(23): 7513–7518. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Rey AA, Purrio M, Viveros M-P, Lutz B. Biphasic Effects of Cannabinoids in Anxiety Responses: CB1 and GABAB Receptors in the Balance of GABAergic and Glutamatergic Neurotransmission. Neuropsychopharmacology 2012; 37(12): 2624–2634. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Jonak CR, Pedapati EV, Schmitt LM, Assad SA, Sandhu MS, DeStefano L et al. Baclofen-associated neurophysiologic target engagement across species in fragile X syndrome. J Neurodev Disord 2022; 14(1): 52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Neske GT, Connors BW. Distinct Roles of SOM and VIP Interneurons during Cortical Up States. Frontiers in Neural Circuits 2016; 10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Zucca S, D'Urso G, Pasquale V, Vecchia D, Pica G, Bovetti S et al. An inhibitory gate for state transition in cortex. Elife 2017; 6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Bodor AL, Katona I, Nyíri G, Mackie K, Ledent C, Hájos N et al. Endocannabinoid signaling in rat somatosensory cortex: laminar differences and involvement of specific interneuron types. J Neurosci 2005; 25(29): 6845–6856. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Hill EL, Gallopin T, Férézou I, Cauli B, Rossier J, Schweitzer P et al. Functional CB1 receptors are broadly expressed in neocortical GABAergic and glutamatergic neurons. J Neurophysiol 2007; 97(4): 2580–2589. [DOI] [PubMed] [Google Scholar]
  • 70.Wedzony K, Chocyk A. Cannabinoid CB1 receptors in rat medial prefrontal cortex are colocalized with calbindin-but not parvalbumin- and calretinin-positive GABA-ergic neurons. Pharmacol Rep 2009; 61(6): 1000–1007. [DOI] [PubMed] [Google Scholar]
  • 71.Katona I, Sperlagh B, Sik A, Kafalvi A, Vizi ES, Mackie K et al. Presynaptically located CB1 cannabinoid receptors regulate GABA release from axon terminals of specific hippocampal interneurons. J Neurosci 1999; 19(11): 4544–4558. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 72.Neu A, Földy C, Soltesz I. Postsynaptic origin of CB1-dependent tonic inhibition of GABA release at cholecystokinin-positive basket cell to pyramidal cell synapses in the CA1 region of the rat hippocampus. The Journal of Physiology 2007; 578(1): 233–247. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73.Roberto M, Cruz M, Bajo M, Siggins GR, Parsons LH, Schweitzer P. The endocannabinoid system tonically regulates inhibitory transmission and depresses the effect of ethanol in central amygdala. Neuropsychopharmacology 2010; 35(9): 1962–1972. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 74.Aloisi E, Le Corf K, Dupuis J, Zhang P, Ginger M, Labrousse V et al. Altered surface mGluR5 dynamics provoke synaptic NMDAR dysfunction and cognitive defects in Fmr1 knockout mice. Nat Commun 2017; 8(1): 1103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 75.Omiya Y, Uchigashima M, Konno K, Yamasaki M, Miyazaki T, Yoshida T et al. VGluT3-expressing CCK-positive basket cells construct invaginating synapses enriched with endocannabinoid signaling proteins in particular cortical and cortex-like amygdaloid regions of mouse brains. J Neurosci 2015; 35(10): 4215–4228. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 76.Skosnik PD, Cortes-Briones JA, Hajós M. It’s All in the Rhythm: The Role of Cannabinoids in Neural Oscillations and Psychosis. Biological Psychiatry 2016; 79(7): 568–577. [DOI] [PubMed] [Google Scholar]
  • 77.Syed SA, Schnakenberg Martin AM, Cortes-Briones JA, Skosnik PD. The Relationship Between Cannabinoids and Neural Oscillations: How Cannabis Disrupts Sensation, Perception, and Cognition. Clin EEG Neurosci 2023; 54(4): 359–369. [DOI] [PubMed] [Google Scholar]
  • 78.Cortes-Briones J, Skosnik PD, Mathalon D, Cahill J, Pittman B, Williams A et al. Δ9-THC Disrupts Gamma (γ)-Band Neural Oscillations in Humans. Neuropsychopharmacology 2015; 40(9): 2124–2134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 79.Whissell PD, Cajanding JD, Fogel N, Kim JC. Comparative density of CCK- and PV-GABA cells within the cortex and hippocampus. Front Neuroanat 2015; 9: 124. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 80.Eggan SM, Lewis DA. Immunocytochemical Distribution of the Cannabinoid CB1 Receptor in the Primate Neocortex: A Regional and Laminar Analysis. Cerebral Cortex 2006; 17(1): 175–191. [DOI] [PubMed] [Google Scholar]
  • 81.Skosnik PD, D'Souza DC, Steinmetz AB, Edwards CR, Vollmer JM, Hetrick WP et al. The effect of chronic cannabinoids on broadband EEG neural oscillations in humans. Neuropsychopharmacology 2012; 37(10): 2184–2193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 82.Foldy C, Malenka RC, Sudhof TC. Autism-associated neuroligin-3 mutations commonly disrupt tonic endocannabinoid signaling. Neuron 2013; 78(3): 498–509. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 83.Anderson GR, Aoto J, Tabuchi K, Földy C, Covy J, Yee AX et al. β-Neurexins Control Neural Circuits by Regulating Synaptic Endocannabinoid Signaling. Cell 2015; 162(3): 593–606. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Methods and Figures
NIHMS2211207-supplement-Supplemental_Methods_and_Figures.pdf (2.6MB, pdf)

Data Availability Statement

The authors declare that all data supporting the findings of this study are available in the main text or the supplementary materials. Datasets are available from the corresponding authors upon request. Correspondence and requests for materials should be addressed to KMH. and DKB (Kimberly.Huber@UTSouthwestern.edu and devin.binder@ucr.edu). Labview software for analysis of Up states is available at https://github.com/jaygibson244/GibsonHuber-UP-states.git. The MATLAB code used to perform statistical analysis for chirp, 40 and 80Hz ASSR is available at Github: https://github.com/CarrieRJonak/MEA-EEG_ITPC.

RESOURCES