Inflammatory demyelination alters subcortical visual circuits
© The Author(s). 2017
Received: 26 May 2017
Accepted: 7 August 2017
Published: 18 August 2017
Skip to main content
We’re sorry, something doesn't seem to be working properly.
Please try refreshing the page. If that doesn't work, please contact us so we can address the problem.
© The Author(s). 2017
Received: 26 May 2017
Accepted: 7 August 2017
Published: 18 August 2017
We’re sorry, something doesn't seem to be working properly.
Please try refreshing the page. If that doesn't work, please contact us so we can address the problem.
Multiple sclerosis (MS) is an inflammatory demyelinating disease classically associated with axonal damage and loss; more recently, however, synaptic changes have been recognized as additional contributing factors. An anatomical area commonly affected in MS is the visual pathway; yet, changes other than those associated with inflammatory demyelination of the optic nerve, i.e., optic neuritis, have not been described in detail.
Adult mice were subjected to a diet containing cuprizone to mimic certain aspects of inflammatory demyelination as seen in MS. Demyelination and inflammation were assessed by real-time polymerase chain reaction and immunohistochemistry. Synaptic changes associated with inflammatory demyelination in the dorsal lateral geniculate nucleus (dLGN) were determined by immunohistochemistry, Western blot analysis, and electrophysiological field potential recordings.
In the cuprizone model, demyelination was observed in retinorecipient regions of the subcortical visual system, in particular the dLGN, where it was found accompanied by microglia activation and astrogliosis. In contrast, anterior parts of the pathway, i.e., the optic nerve and tract, appeared largely unaffected. Under the inflammatory demyelinating conditions, as seen in the dLGN of cuprizone-treated mice, there was an overall decrease in excitatory synaptic inputs from retinal ganglion cells. At the same time, the number of synaptic complexes arising from gamma-aminobutyric acid (GABA)-generating inhibitory neurons was found increased, as were the synapses that contain the N-methyl-d-aspartate receptor (NMDAR) subunit GluN2B and converge onto inhibitory neurons. These synaptic changes were functionally found associated with a shift toward an overall increase in network inhibition.
Using the cuprizone model of inflammatory demyelination, our data reveal a novel form of synaptic (mal)adaption in the CNS that is characterized by a shift of the excitation/inhibition balance toward inhibitory network activity associated with an increase in GABAergic inhibitory synapses and a possible increase in excitatory input onto inhibitory interneurons. In addition, our data recognize the cuprizone model as a suitable tool in which to assess the effects of inflammatory demyelination on subcortical retinorecipient regions of the visual system, such as the dLGN, in the absence of overt optic neuritis.
Multiple sclerosis (MS) has long been considered a chronic inflammatory and demyelinating disease of the CNS that is ultimately associated with axonal degeneration and neuronal loss . More recently, however, synaptic changes, possibly reflecting forms of maladaptive plasticity, have been implicated in contributing to the pathophysiology of the disease, even at early stages of disease progression when such changes may still be reversible [2, 3]. In this context, studies using experimental autoimmune encephalomyelitis (EAE) as a model suggest that synaptic changes in MS may include a shift toward increased excitation, which may be caused, at least in part, by T cell-initiated adaptive immune responses [2–7].
In the healthy CNS, appropriate synaptic functions and neural circuitries are regulated by resident micro- and astroglia cells, which contribute to synapse formation, elimination and plasticity, as well as neurotransmission . Activation of CNS-resident glia cells under neuroinflammatory conditions, on the other hand, is thought to cause adaptive synaptic alterations ultimately leading to the malfunction of neural circuits [2, 6, 8, 9]. Synchronization of impulse traffic within neural circuits is likely also regulated by the extent of myelination . Consistent with this point of view, demyelination has been implicated in causing delayed spike arrival timing as well as impaired activity coincidence and synaptic efficacy . In MS, it is quite probable that maladaptive synaptic changes result from a combination of both neuroinflammation and demyelination, whereby the contribution of each of the two factors may depend on the affected brain area and/or the stage of the disease.
A common clinical manifestation in MS is visual dysfunction, which is often caused by inflammatory demyelination of the optic nerve, a condition known as optic neuritis . However, not all patients with visual impairments exhibit optic neuritis, suggesting that neuroinflammation and demyelination of visual centers located within the brain may critically contribute to visual dysfunction in MS . Likewise, in a common animal model of inflammatory demyelination, the cuprizone model, white matter areas and the cortex have been found affected while the optic nerve and tract are largely spared [14–16]. Thus, we used the cuprizone model to assess pathological changes, including potential synaptic alterations that are associated with inflammatory demyelination in the visual thalamus, an essential relay nucleus necessary for processing light-derived signals from the retina and transmitting those signals to the primary visual cortex . Of note, inflammatory demyelination in deep gray matter including the thalamus has been reported to occur frequently in MS [18, 19].
Six-week-old C57BL/6 male mice (The Jackson Laboratory) were fed with chow containing 0.2% cuprizone (Sigma-Aldrich) or control chow for 3 or 5 weeks. All animal studies were approved by the Institutional Animal Care and Use Committee at the Virginia Commonwealth University.
Anti-MBP (Aves Labs), anti-GFAP (DAKO), anti-F4/80 (clone Cl:A3-1; AbD Serotec), anti-PSD-95 (Abcam), anti-vGlut2 (Abcam), anti-GAD67 (Abcam), and Alexa488- and/or Alexa594-conjugated secondary antibodies (Molecular Probes) were used for immunohistochemistry. Anti-GluN2B (Synaptic Systems) antibodies were used for Western blot analysis and immunohistochemistry. Anti-GAPDH (EMD Millipore) and horseradish peroxidase (HRP)-labeled secondary antibodies (Vector Laboratories) were used for Western blot analysis.
Mice were deeply anesthetized using Avertin (2,2,2-Tribromoethanol, Sigma-Aldrich), and the brains were dissected into ice-cold PBS. Three hundred-micrometer coronal brain sections were prepared using a McIlwain tissue chopper (Ted Pella, Inc.), and the tissues, corpus callosum (CC) and dorsal lateral geniculate nucleus (dLGN), were dissected manually using small scalpel blades. The optic nerves were collected after removal of the brain. All tissue samples were flash frozen in liquid nitrogen or directly used for RNA isolation. RNA isolation and real-time RT-qPCR analysis was performed as described previously  using the following gene-specific primer pairs: Mbp: forward (5′-CTTGGCCACAGCAAGTACCATGGACC-3′) and reverse (5′-TTGTACATGTGGCACAGCCCGGGAC-3′); Plp1: forward (5′-CCACACTAGTTTCCCTGCTCACCT-3′) and reverse (5′-GGTGCCTCGGCCCATGAGTT-3′); Il1b: forward (5′-TGAAGAAGAGCCCATCCTCTGTGA-3′) and reverse (5′-GGTCCGACAGCACGAGGCTT 3′); Tnf: forward (5′-GCCCACGTCGTAGCAAACCACC-3′) and reverse (5′-CCCATCGGCTGGCACCACTA-3′); Nos2: forward (5′-TCCAGAATCCCTGGACAAGCTGC-3′) and reverse (5′-TGCAAGTGAAATCCGATGTGGCCT-3′); Ptgs2: forward: (5′-TTGCTGGCCGGGTTGCTGG-3′) and reverse: (5′-CAGGGAGAAGCGTTTGCGGT-3′); Grin2a: forward: (5’CGCATCCATGGCTTGGTGTTT-3′) and reverse (5′-TGTCGGATCCTTGTCAGCCAT-3′); Grin2b: forward (5′-ATGATGCCTTGCTCTCCCTG-3′) and reverse (5′-ATGCCATAGCCCGTAGAAGC-3′); Pgk1 (reference gene): forward: (5′-ATGCAAAGACTGGCCAAGCTAC-3′) and reverse: (50-AGCCACAGCCTCAGCATATTTC-30). Relative expression levels were determined using the∆∆CT method .
Mice were transcardially perfused with saline (0.9% NaCl) followed by 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). Brains were removed, post-fixed for 24 h, and coronal vibratome sections were collected at 40 μm (Leica VS1000S). The sections were permeabilized and blocked in phosphate-buffered saline (PBS) containing 3% bovine serum albumin (Sigma-Aldrich), 5% normal goat serum (NGS; Invitrogen), and 0.3% Triton X-100 and then incubated with primary (overnight at 4 °C) and secondary antibodies (2 h at RT). Primary and secondary antibodies were used at the following dilutions: chicken anti-MBP (1:500); rabbit anti-GFAP (1:500); rat anti-F4/80 (1:100); rabbit anti-PSD-95 (1:100); mouse anti-vGlut2 (1:400); rabbit anti-GluN2B (1:50); mouse anti-GAD67(1:500); Alexa Fluor 594 (goat anti-mouse IgG, 1:1000; goat anti-rabbit IgG, 1:1000); Alexa Fluor 488 (goat anti-chicken IgY, 1:500; rabbit anti-goat IgG, 1:300; goat anti-rabbit IgG, 1:400; goat anti-mouse IgG, 1:400). Sections were mounted using aqueous mounting medium (DAKO), and images were collected using a Leica TCS SPE confocal laser scanning microscope.
Quantification of immunohistochemical data was performed using ImageJ  by determining immunostained densities and the number of immuno-positive cells. Two fields per slice and two slices per animal were analyzed from four control and five cuprizone-treated mice. Synaptic puncta analysis was performed as described previously  using 20 serial optical sections captured at 0.71 μm intervals (Leica TCS SPE).
Mice were transcardially perfused sequentially with PBS and 4% paraformaldehyde/2% glutaradehyde in 0.1 M cacodylate buffer. Brains were immediately removed and vibratomed (300-μm coronal sections), and dLGN tissues were dissected from 300-μm coronal vibratome sections. Tissues were then stained, embedded, sectioned, and imaged by Renovo Neural Inc. (Cleveland, OH). Images were acquired at a resolution of 5 nm/pixel and image sets included > 200 serial sections (with each section representing 75 nm in the z axis). Serial block face scanning electron microscopy (SBFSEM) data sets were 40 μm × 40 μm × 12–20 μm. Retinal terminals and axons were identified and traced in TrakEM2 as described previously .
Mice were deeply anesthetized using Avertin (2,2,2-Tribromoethanol, Sigma-Aldrich), and brains were dissected into ice-cold PBS. Three hundred-micrometer coronal brain sections were then prepared using a McIlwain tissue chopper (Ted Pella, Inc.), and dLGN tissue samples were obtained by manual dissection using small scalpel blades. The tissue samples were flash frozen in liquid nitrogen or directly homogenized in lysis buffer (150 mM NaCl, 10 mM KCl, 20 mM HEPES, pH 7.0, 1 mM MgCl2, 20% glycerol, and 1% Triton X-100, including the complete protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific)). Twelve micrograms of protein per sample were separated by gel electrophoresis. After transfer and incubation with primary antibodies, HRP-conjugated secondary antibodies in combination with the ECL Prime Western blotting detection reagent (GE Healthcare Life Sciences) were used. Chemiluminescent signals were detected by exposure to photographic film (Kodak BioMax MR) and quantified by densitometry (Scion Image; Scion Corporation).
Mice were transcardially perfused with ice-cold high-sucrose artificial cerebrospinal fluid (aCSF; in mM—2.5 KCl, 10 MgSO2, 0.5 CaCl2, 1.25 NaH2PO4, 26 NaHCO3, 11 glucose, and 234 sucrose; pH 7.4 when saturated with 95% O2–5% CO2). Parasagittal brain slices containing the optic tract reaching the dLGN were obtained as described by Chen and Regehr, 2000 . Briefly, the brain hemispheres were separated by a 10° angled cut relative to the medial line. The intact hemisphere was attached to a 25° angled platform to prepare 300 μm thick vibratome slices. Using this procedure, one to two slices containing the OT reaching the dLGN could be collected from each brain. The slices were allowed to recover in oxygenated aCSF (in mM: 126 NaCl, 3.0 KCl, 2.0 MgCl2, 2.0 CaCl2, 1.25 NaH2PO4, 10 glucose, and 26 NaHCO3) at 34 °C for 1 h and at room temperature thereafter until placed in the recording chamber. Field potential recordings were made in slices placed in an interface chamber (Harvard Apparatus) with an Axoclamp2B amplifier and Cygnus ER1 for an overall gain of 1000X and digitized with a Digidata 1322A and pClamp software (all except Cygnus from Molecular Devices). During the recordings, the slices were maintained at 34 °C and continuously perfused with oxygenated aCSF. The tungsten stimulation and recording electrodes were positioned on the optic tract and dLGN, respectively, separated by a minimum distance of 100 μm. In order to assess the GluN2B component of the retinogeniculate field potential, baseline recordings were made in normal aCSF every 30 s for 5 min. Thereafter, Ro25-6981 (1 mM, Tocris Bioscience), a selective activity-dependent blocker of NMDA receptors containing the GluN2B subunit, was added to the perfusing aCSF, which likely required 5 min to reach the chamber. After 10 min, the recordings were made in the presence of the drug every 30 s for 5 min. In each slice, the peak amplitude of the synaptic-dependent field negativity post-drug was normalized using pre-drug values, whereby the second of two evoked negative field potential peaks was identified as the synaptic-dependent field potential by comparing recordings in normal aCSF to those in aCSF containing the competitive NMDA receptor antagonist APV (50 μM, Tocris Bioscience) and the competitive non-NMDA glutamate receptor antagonist DNQX (20 μM, Tocris Bioscience).
For statistical analysis, the GraphPad Prism Software (GraphPad Software) was used. In the case data were compared with a normalized control value (1 or 100%), one-sample t tests were used [26, 27]. In the case of comparing two groups of data composed of variable samples, unpaired two-tailed Student’s t tests  were used. Normal distribution of the data was confirmed using Shapiro-Wilk and Kolmogorov-Smirnov normality tests. Sample size (n) refers to the number of animals and is indicated in each of the figure legends. Differences were considered significant for p ≤ 0.05.
Taken together, the above data demonstrate that cuprizone-induced demyelination in the dLGN is associated with an inflammatory and reactive astrocyte response that is similar to the responses previously described for the corpus callosum.
The above data reveal robust changes in both excitatory and inhibitory inputs into the dLGN, whereby these changes appear to temporally follow the initial inflammatory demyelinating events induced by cuprizone.
Together, the above findings sustain the idea that inflammatory demyelination impacts the excitatory/inhibitory balance and can lead to an increase in inhibitory contributions to network activity in the dLGN.
Our findings demonstrate that in the cuprizone model demyelination occurs in retinorecipient regions of the subcortical visual system, in particular the dLGN, where it is associated with inflammatory micro- and astrogliosis. In contrast, the anterior parts of the pathway, i.e., the optic nerve and tract, appear largely unaffected in this model. These findings are consistent with previous observations [14–16, 44]. On a functional level, inflammatory demyelination within the dLGN was found accompanied by synaptic alterations leading to a shift toward retinogeniculate network inhibition. These observations are distinct from those reported for models of EAE in which inflammatory demyelination of the optic nerve [45, 46] and unbalanced synaptic hyperexcitation  have been reported to contribute to the overall pathology. Thus, our findings unravel a novel type of synaptic alterations accompanying inflammatory demyelination within the CNS, i.e., an excitation/inhibition balance that is shifted toward inhibitory network activity and associated with an increase in GABAergic inhibitory synapses as well as a potentially enhanced excitatory input onto inhibitory interneurons. In addition, our findings establish cuprizone-mediated inflammatory demyelination as a model for investigating pathological visual pathway changes under non-optic neuritis conditions.
Our observation that inflammatory demyelination leads to damage and/or loss of excitatory retinal inputs into the dLGN (see Fig. 4) may, at least in part, be similar to the findings made in some EAE models. In these models, synapses have been considered early targets due to their high sensitivity to even subtle inflammation-induced changes [3, 7]. Such synaptic alterations are thought to occur largely independent of axonal demyelination and neuronal cell death. Instead, they are thought to be mediated by inflammatory cytokines, in particular TNF-α, that are released from infiltrating T cells as well as activated microglia and astrocytes. In light of the increased levels of Tnf mRNA (see Fig. 2), a similar mechanism may drive the excitatory synapse damage/loss in the dLGN of cuprizone-treated mice. However, it is noteworthy that the reduction in the number of vGluT2-expressing synapses was observed at a stage at which pro-inflammation and Tnf mRNA levels began to subside. In this context, it is interesting to note that complement receptor 3-dependent synaptic engulfment by non-pathological microglia has been highlighted as a key mechanism for the elimination of ineffective synapses during development . In addition, despite a lack of visible myelin breakdown and axonal degeneration in optic nerves from cuprizone-treated mice, some structural alterations and a temporal dispersion of action potentials along optic nerve axons have been reported . These changes could potentially result in ineffective retinal inputs into the dLGN. It is, therefore, conceivable that microglia-mediated and complement-dependent elimination of ineffective synapses plays a role in the synaptic changes seen in the dLGN of cuprizone-treated mice. If so, such elimination would appear to uniquely affect retinal terminals in the dLGN of cuprizone-treated mice, despite the notion that microglia are capable of eliminating inhibitory synapses and have been found to do so in other injury models [48–50]. Independent of the exact mechanism, our data demonstrate that inflammatory demyelination can cause significant changes affecting the excitatory retinal inputs into the dLGN even in the absence of overt optic neuritis.
In EAE and possibly relapsing-remitting MS (see below), suppression of GABAergic tone, resulting in a stimulation of excitation, is thought to significantly contribute to synaptic loss and dysfunction. These pathophysiological changes are likely mediated by the inflammatory environment in the CNS and particularly by IL-1β release from activated microglia and infiltrating T cells. In addition, synaptic changes associated with increased levels of IL1-β have been described to include significant reductions in GluN2B localizations [51–53]. In contrast, we observed here an increase in both GluN2B and the number of vGAT-positive inhibitory interneurons, likely reflecting an increase in GABAergic tone in the dLGN of cuprizone-treated mice. Distinctive from EAE, neither T cell infiltration nor increased levels of IL-1β are considered major players in driving the pathophysiological processes in the CNS of cuprizone-treated mice [16, 30]. Thus, variations in the inflammatory demyelinating milieu may play a critical role in defining the GABAergic tone of synaptic adaptations under pathological conditions. Alternatively and/or additionally, regional heterogeneity in microglia immune alertness and responsiveness to immune stimulation has been recently reported . In this context, it is worth mentioning that synaptic changes in EAE have so far only been described for CNS regions other than the dLGN. In addition, inflammatory demyelination in the EAE thalamus has been described as subtle and to occur in the absence of perivascular immune cell infiltrates typically seen in the EAE spinal cord and cerebellum . Thus, regional variations in immune responses may contribute to the apparent differences in synaptic (mal)adaptions that have so far been reported for the EAE and cuprizone models.
While retinal terminals provide the primary excitatory drive onto thalamocortical relay cells in the dLGN, they comprise only a small fraction (~ 10%) of the synapses in this region [55, 56]. However, an equal (if not larger) proportion of synapses in the dLGN is GABAergic and arises from local inhibitory interneurons as well as GABAergic neurons in the nearby thalamic reticular nucleus (TRN) [57, 58]. Our data demonstrate that the dLGN of cuprizone-treated mice exhibits on the one hand reductions in the levels of VGluT2 and the numbers of excitatory retinal synapses, while at the same time the numbers of inhibitory terminals are elevated. We interpret these results to suggest a decrease in connectivity between the retina and thalamocortical relay cells that are associated with an increase in the number of inhibitory contacts onto these principle neurons. This concept is supported by our electrophysiological data (see Fig. 7). At present, it remains unclear whether the increase in vGAT-positive inhibitory nerve terminals arises from local interneurons, projection neurons in the TRN, or both. During development, impairment or loss of retinal inputs has been described to lead to exuberant and aberrant innervation by non-retinal inputs in the mouse dLGN [59, 60]. It is, therefore, conceivable that loss of excitatory drive in the visual thalamus of cuprizone-treated mice may lead to (mal)adaptive inhibitory synaptogenesis of local interneurons and/or terminal arbors arising from the TRN that is driven by an upregulation of synaptogenic signals. In the search for potential candidate molecules, BDNF has been identified as a potent inducer of inhibitory synapses that is upregulated in developing retinorecipient nuclei of enucleated rats [61–63]. In addition, BDNF has been found upregulated upon cuprizone treatment, at least in the corpus callosum where it is released by activated astrocytes [64, 65].
An interesting complexity is provided by the elevated expression levels of specifically the GluN2B NMDAR subunit in the dLGN of cuprizone-treated mice (see Fig. 6). NMDA receptors have been described to contribute to both retinogeniculate as well as corticogeniculate transmission [66, 67]. Thus, the increase seen in GluN2B-positive terminals could arise from both retinal as well cortical projections. For example, the observed increase in GluN2B-positive puncta in local GAD67-positive interneurons may reflect a compensatory mechanism in response to a reduction in excitatory retinogeniculate drive to these cells. In the hippocampus, GluN2B-containing NMDARs have been shown to positively regulate the maturation of glutamatergic input synapses in interneurons . Thus, the responses seen in the dLGN of cuprizone-treated mice could reflect an attempt to induce the formation of new excitatory inputs from retinal ganglion cells in local inhibitory synapses. This idea is based on the assumption that a significant portion of the increased NMDARs resides on the postsynaptic membranes of synapses targeting local inhibitory synapses, and it is supported by our data indicating increased GluN2B protein levels within GAD67-positive dLGN interneurons (see Fig. 6). However, an alternative possibility (which is not mutually exclusive) is that a portion of the increased GluN2B subunits in cuprizone-treated animals reside on the presynaptic terminals of TRN projections into the dLGN. TRN neurons contain tonically active GluN2B-containing NMDARs on their GABAergic terminals . Presynaptic NMDARs in TRN neurons facilitate GABA release and inhibition . Thus, increased inhibition in the dLGN of cuprizone-treated animals may reflect elevated presynaptic NMDARs on TRN-relay cell synapses and enhanced GABA release from these terminals. As alluded to above, the synaptic changes observed under the inflammatory demyelinating conditions as they are present in the dLGN of cuprizone-treated mice may be very complex, and intricate additional studies are needed to address the mechanistic underpinnings leading to these changes. Such continuing studies would include an in vivo assessment of visual function. In this context, it is noteworthy that demyelination of the optic nerves and chiasm has been described to lead to clear alterations in both amplitude and latency of visual evoked potentials . In addition, chronic cuprizone treatment was found to be associated with an impairment of visual function by recording multifocal electroretinograms . Importantly, and consistent with our findings, no retinal ganglion cell degeneration was observed in this study.
From a translational point of view, complement C1q-C3-associated synaptic changes, such as those potentially contributing to the elimination of ineffective retinal inputs into the dLGN of cuprizone-treated mice, have been linked to progressive MS [72, 73]. In light of the distinct synaptic alterations seen in EAE (decrease in GABAergic tone and increase in excitation) versus cuprizone (increase in inhibitory network activity), it is also noteworthy that cuprizone-associated oligodendrocyte death  and an absence of significant T cell activation  mimic features of the pathophysiology described for primary progressive MS [76, 77]. Thus, it is tempting to speculate that the changes described here may be more representative for primary progressive MS, while those seen in EAE may be more reflective of alterations occurring in relapsing-remitting MS.
Using the cuprizone model, our findings unravel a novel type of synaptic alterations accompanying inflammatory demyelination within the CNS, i.e., an excitation/inhibition balance that is shifted toward inhibitory network activity associated with an increase in GABAergic inhibitory synapses and possibly an enhanced excitatory input onto inhibitory interneurons. These findings are distinct from those describing an unbalanced increase in excitatory network activity in inflammatory demyelinating models of EAE . In addition, our findings establish cuprizone-mediated inflammatory demyelination as a model for investigating pathological visual pathway changes under non-optic neuritis conditions.
Central nervous system
Dorsal lateral geniculate nucleus
Experimental autoimmune encephalomyelitis
Glutamate decarboxylase 1 (67 KDa isoform)
Glyceraldehyde 3-phosphate dehydrogenase
Glial fibrillary protein
NMDAR subunit 2B
NMDAR subunit 2A
NMDAR subunit 2B
Microglia/macrophage-specific calcium-binding protein
Interleukin 1 beta
Myelin basic protein
Inducible nitric oxide synthase
Postsynaptic density protein 95
Tumor necrosis factor alpha
Thalamic reticular nucleus
Vesicular glutamate transporter 2
This work was supported by the grants from the National Institutes of Health (NIH)/National Institute of Neurological Disorders and Stroke (NINDS; B.F.), the National Institutes of Health (NIH)/National Eye Institute (NEI; M.A.F) as well as the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and the Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ). SESA and HRM were recipients of a CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior) fellowship.
All data supporting the results presented here are included in this published article.
SESA performed the experiments, analyzed and interpreted the data, and prepared the initial draft of the manuscript. HRM was a major contributor related to the performance and analysis of the electrophysiological recordings. NAW was a major contributor related to the RT-qPCR experiments, KMJ oversaw the electrophysiological experiments and provided critical comments during the writing of the manuscript, FCAG oversaw the experiments related to the immunohistochemical synapse analysis, PCC provided critical input to the design of the experiments, MAF provided the SBFSEM data and essential guidance related to the interpretation of the data and the writing of the manuscript, BF oversaw the entire project and wrote the manuscript. All authors read and approved the final manuscript.
All animal studies were approved by the Institutional Animal Care and Use Committee at the Virginia Commonwealth University. They are compliant with the Animal Welfare Act and Regulations (Animal Welfare Assurance Number: A3281-01) and the Office of Laboratory Animal Welfare Policies and Laws (OLAW). VCU’s Institutional Animal Care and Use Committee is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC).
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.