Altered cortical Cytoarchitecture in the Fmr1 knockout mouse
Fragile X syndrome (FXS) is a neurodevelopmental disorder caused by silencing of the FMR1 gene and subsequent loss of its protein product, fragile X retardation protein (FMRP). One of the most robust neuropathological findings in post-mortem human FXS and Fmr1 KO mice is the abnormal increase in dendritic spine densities, with the majority of spines showing an elongated immature morphology. However, the exact mechanisms of how FMRP can regulate dendritic spine development are still unclear. Abnormal dendritic spines can result from disturbances of multiple factors during neurodevelopment, such as alterations in neuron numbers, position and glial cells. In this study, we undertook a comprehensive histological analysis of the cerebral cortex in Fmr1 KO mice. They displayed significantly fewer neuron and PV-interneuron numbers, along with altered cortical lamination patterns. In terms of glial cells, Fmr1 KO mice exhibited an increase in Olig2-oligodendrocytes, which corresponded to the abnormally higher myelin expression in the corpus callosum. Iba1-microglia were significantly reduced but GFAP-astrocyte numbers and intensity were elevated. Using primary astrocytes derived from KO mice, we further demonstrated the presence of astrogliosis characterized by an increase in GFAP expression and astrocyte hypertrophy. Our findings provide important information on the cortical architecture of Fmr1 KO mice, and insights towards possible mechanisms associated with FXS.
KeywordsFragile X syndrome Fmr1 KO mice Cortical architecture Astrocytes
Autism Spectrum Disorder
Central Nervous System
Fragile X Mental Retardation 1
Fragile X Mental Retardation Protein
Fragile X Syndrome
Glial Fibrillary Acidic Protein
Myelin Basic Protein
Fragile X syndrome (FXS) is the most common inherited form of intellectual disability and one of the leading genetic causes of autism spectrum disorder (ASD), affecting approximately 1 in 4000 males and 1 in 6000 females [1, 2, 3]. It is characterized by a wide spectrum of clinical symptoms, including mild to severe intellectual disability, susceptibility to seizures, hyperactivity, hypersensitivity to sensory stimuli, autistic behaviors such as social anxiety and attention deficits, macroorchidism and abnormal facial features (Reviewed in ). FXS is a neurodevelopmental disorder caused by a CGG repeat expansion in the X-linked fragile X mental retardation 1 (FMR1) gene, which results in the transcriptional silencing of FMR1, and subsequent reduction of its protein product, fragile X retardation protein (FMRP) [4, 5].
FMR1 and FMRP are ubiquitously expressed in the mammalian CNS, beginning in early embryogenesis and persisting throughout development into adulthood . At the cellular level, they are detected in different cell populations from proliferating cells of the developing brain, and later within pyramidal neurons, to GABAergic interneurons and glial cells of microglia, oligodendrocytes and astrocytes [6, 7, 8, 9, 10, 11]. Functionally, FMRP regulates mRNA expression by binding and then suppressing the translation of its target mRNAs . Genome-wide microarrays and high-throughput sequencing studies have identified more than 800 mRNA targets of FMRP, many of which are linked to neurodevelopmental processes including neurite growth, spine development, synaptic function and neuronal signaling [12, 13, 14]. The broad expression of FMRP in multiple cell types and brain regions, together with the vast number of interacting mRNA targets, suggest that it is immensely important for brain development and maturation.
Animal models are crucial in understanding the biological functions of genetic mutations. The Fmr1 knockout (KO) mouse, containing the loss of a functional FMRP protein, has been well-established in displaying behavioral abnormalities reminiscent of human FXS traits [15, 16]. More importantly, many studies have since been using this model as a tool to investigate the pathophysiological mechanisms underlying FXS. One of the most robust neuropathological findings in post-mortem human FXS and Fmr1 KO mice is the abnormal increase in dendritic spine densities, with the majority of spines showing an elongated immature morphology [17, 18, 19]. However, evidence of other histological defects associated with FXS are still lacking, and inconsistent results have been described on neurotransmission properties, such as NMDA and AMPA receptor expression/function .
Dendritic spine development is a complicated process beginning with synaptogenesis in early childhood, synapse elimination/pruning in adolescence and spine maintenance in adulthood . Proper development and maturation requires strict spatial and temporal regulation, involving multiple factors at each stage. Therefore, alterations in spine number and/or morphology could emerge from secondary effects or compensatory responses of any dysfunctional events occurring during that period. For example, neuron and interneuron numbers, cortical lamination pattern, and axonal connections and myelination have all been shown to affect spinogenesis [21, 22, 23]. Besides, recent evidence has directed microglia and astrocytes as key players in modulating synaptic formation and pruning via distinct pathways. For example, microglia and astrocytes mediate synapse elimination by direct engulfment of synapses through the complement system and MEGF10/MERTK respectively [24, 25]. Moreover, alterations in cytokine production and astrocyte-specific glutamate transporters can have pronounced effects on dendritic spines [26, 27, 28].
Despite extensive research focusing on the functional role of FMRP in FXS, how it regulates spine development remains unclear. In this study, we undertook a comprehensive histological analysis of the cerebral cortex in the Fmr1 KO mice. They displayed reductions in cortical neuron and PV-interneuron numbers, along with altered cortical lamination patterns. In terms of glial cells, KO mutants exhibited distinct changes in Olig2-oligodendrocyte, Iba-1 microglia and GFAP-astrocyte numbers. Using primary astrocytes from Fmr1 KO mice, we further demonstrated the presence of astrogliosis characterized by the significant increase in GFAP expression and astrocyte hypertrophy. Our findings provide important information on the cortical architecture of Fmr1 KO mice, and insights towards possible mechanisms that could be responsible for the increase in spine densities associated with FXS.
Fewer neuron and parvalbumin (PV)-interneuron numbers in the Fmr1 KO mouse cortex
Fmr1 KO mice show altered cortical lamination
Distinct changes in glial cell numbers of Fmr1 KO mice
FMRP-depleted astrocytes exhibit a more pronounced reactive state with higher GFAP expression and hypertrophy
There is strong evidence indicating the diverse roles of FMRP in regulating many essential neurodevelopmental processes . The most robust histopathological finding in FXS patients and Fmr1 KO mice is the abnormal increase in immature dendritic spine numbers [17, 18, 19]. However, the mechanisms of how FMRP can modulate spine development are still unclear. Disturbances in many neurodevelopmental factors can have profound effects on dendritic spines. In this study, we examined the cortical cytoarchitecture in the Fmr1 KO mouse cortex that could be associated with spine abnormalities. Alterations in neuronal number and positioning can affect proper spine formation. There is evidence suggesting that neuronal populations inherently respond to lower neuronal densities by increasing the number of synapses . We observed a small but significant decrease in neuronal density only within the cingulate cortex of KO mutants but not the lateral regions of motor and somatosensory cortex. Consistently, Hall et al. reported that patients with FXS had reduced gray matter density in the cingulate cortex using voxel-based morphometry . In the Fmr1 KO mouse, a study found no change in gross cortical neuronal morphology or overall cortical thickness of the primary somatosensory cortex . Earlier studies have documented the involvement of FMRP in embryonic neurogenesis, where embryonic Fmr1 KO mice displayed accelerated neurogenesis, increased neuronal differentiation and enhanced accumulation of progenitors in the subventricular zone [40, 41, 42]. These newly differentiated FMRP-deficient neurons exhibit abnormal functional properties such as altered calcium responses and bursts activity [40, 43]. As a result, non-functional neurons that could not be integrated into the local circuitry are eliminated, which may explain the reduction in neuron number.
FMRP is also prominently expressed in GABAergic neuron populations, implying that it may have crucial roles in interneuron development and function [8, 11]. Indeed, using the Fmr1 KO mouse model, many studies have illustrated defects in the inhibitory GABAergic circuitry [11, 44]. We found significantly fewer PV-interneurons in all regions across the neocortex, which was coherent with the results reported by Selby et al. . Furthermore, according to Selby et al., this reduction mainly occurred in somatosensory cortical layers of II/III/IV, but not in deeper layers V and VI where PV numbers were increased . These findings suggest that there could be differential histological deficits within cortical laminar organization. We noticed an aberrant cortical lamination pattern in the Fmr1 KO mutants compared to WT. Neurons destined for deeper layers were mis-positioned to superficial layers while those for superficial remained in deeper regions. Newborn neurons are established in an ‘inside-out’ pattern with the earliest born neurons forming the deep cortical layers, followed by later born neurons that progressively migrate into more superficial layers [32, 46, 47]. Our data corroborated with previous findings of more cortical neurons mis-positioned in deeper layers of Fmr1 KO embryonic brains, and further indicate that this effect persists into adulthood . It is possible that FMRP deficiency may contribute to an accelerated neurogenesis in which more premature neurons are misdirected to deeper localization. Another explanation for neuronal mis-positioning in the cortex could be due to deficits in the migration process. Although some research have documented the role of FMRP in regulating neuronal migration [49, 50, 51], La Fata et al. reported that FMRP depletion did not affect radial migration of excitatory neurons . A better understanding towards how FMRP modulates these processes associated with FXS is necessary. Nonetheless, this study provides a clearer histopathological perspective of neuron and interneuron numbers, and cortical laminar positioning in Fmr1 KO mice, as well as reflecting the possible neurodevelopmental processes affected by FMRP depletion.
FMRP expression is not only restricted to neuronal populations, but also present in developing and mature glial cells including oligodendrocytes, microglia and astrocytes [6, 7, 10]. Oligodendrocytes are primarily responsible for myelinating axons which enables faster and more efficient propagation of nerve impulses . Hence, proper myelination is essential for normal neuronal development including dendritic spines . Although different studies have indicated abnormal white matter structure in FXS patients and Fmr1 KO mice, the findings are still inconsistent . We detected a pronounced increase in Olig2-positive pre-myelinating oligodendrocytes in the neocortex but not the corpus callosum of Fmr1 KO mice. Correspondingly, myelination only within the medial part of the corpus callosum was abnormally enhanced. In support of our results, a study found that oligodendrocyte ablation could lead to an increase in synapse numbers, possibly via modification of ionotropic glutamate receptors localization and function . Both diffusion tensor imaging studies in FXS patients and magnetic resonance imaging studies in Fmr1 KO mice have shown an increase in white matter tract density [56, 57], despite others have reported reduced myelination patterns [9, 58]. A more comprehensive study from Hall et al. suggested that white matter abnormalities associated with FXS are region-specific . As the developmental origins of different brain regions are distinct, the functional role of FMRP on oligodendrocyte development and myelination may also differ. Moreover, there is evidence illustrating that FMRP can bind and regulate the translation of myelin basic protein (MBP) and CNPase (a myelin-associated enzyme) [10, 13]. However, it remains unclear how FMRP modulate the different aspects of oligodendrocyte development, for instance the differentiation of oligodendrocyte precursor cells to mature oligodendrocytes, or apoptosis of oligodendrocytes. A more detail investigation on these events can provide important clues on the relationship between FMRP and white matter function associated with FXS.
Accumulating evidence strongly indicates that both microglia and astrocytes are involved in many neurodevelopmental processes, thus are immensely crucial in maintaining normal brain functions [59, 60, 61]. With regards to FXS, microglia can directly mediate dendritic spine elimination via the complement system where excessive synapses are tagged with complement proteins (C1q and C3) and consequently engulfed by microglial cells expressing complement receptors [24, 62, 63]. There is clinical evidence suggesting excessive microglial activation in multiple brains regions of patients with ASD [64, 65], but other studies did not detect any microglial activation or altered cytokine production in the Fmr1 KO mice [66, 67]. In our study, we found a slight reduction in the number of Iba1-positive cells across the neocortex of Fmr1 KO mutants, further justifying the absence of microglial activation. It is likely that the reduction in microglia leads to less synapse elimination mediated by the complement pathway , and other mediators such as IL-33 and CX3CR1s .
The importance of FMRP in astrocytes was validated by a recent study that showed astrocyte-specific FMRP knockout in mice resulted in increased spine density and behaviors reminiscent of FXS . Astrocytes are a major source of complement C3, and therefore could participate in removing synapses by secreting C3 . Similar to microglia, a study by Chung et al. described a direct phagocytic role of astrocytes on CNS synapses through phagocytic receptors, such as MEGF10 and MERTK . In addition, astrocytes can also contribute to synaptic elimination via inositol 1,4,5-triphosphate receptor type 2 (IP3R2) and P2ry1 dependent signaling . Here, we observed that Fmr1 KO mice exhibited prominent GFAP expression, while in vitro astrocyte cultures displayed hypertrophy and increased GFAP levels, all of which are hallmark features of reactive astrogliosis. In support of our findings, separate studies have provided evidence for astrocytic activation in the cerebellum of Fmr1 KO mouse [66, 67] and in patients with autism [65, 72]. Zamanian et al. conducted an extensive analysis on the genomic profile of reactive astrocytes, where they identified numerous genes of different signaling pathways were significantly induced . FMRP-deficient reactive astrocytes are likely to exhibit similar alterations in gene expression that could contribute to the abnormal increase in dendritic spines. For instance, changes in genes of the complement pathway can affect the process of synapse tagging and removal ; and induction of cytokine production such as IL-6 is evidently linked to dendritic spine abnormalities as well as behaviors relevant to autism [26, 27]. Another possible mechanism can be attributed to the reduced expression of the astrocyte-specific glutamate transporter GLT1 in FMRP-deleted astrocytes, as selective re-expression of FMRP rescued abnormal spine morphology . The detail mechanisms of how reactive astrogliosis in Fmr1 KO can cause an abnormal increase in dendritic spines would require more research on deciphering the downstream pathways.
In conclusion, we found that Fmr1 KO mice displayed histological deficits of fewer neuron and PV-interneuron numbers, altered cortical lamination, more Olig2-oligodendrocytes with enhanced myelination expression, less Iba1-microglia and elevated GFAP expression and astrocyte numbers. Moreover, using primary astrocyte cultures from Fmr1 KO mouse cortex, we demonstrated the presence of astrogliosis characterized by pronounced GFAP expression and hypertrophy at the cellular level. These histopathological defects are strong indicators of perturbed neurodevelopmental processes that could either directly or indirectly contribute to the abnormal elevated spine numbers associated with FXS. Further studies on neurodevelopmental pathways such as neurogenesis, neuronal migration and glial development can address this question. More importantly, our results provide a starting point for future research in investigating at the specific pathways involved in dendritic spine formation and pruning, which could ultimately lead to novel treatments in FXS.
Fmr1 KO mice on a C57/BL6J background were purchased from The Jackson Laboratory (B6.129P2-Fmr1tm1Cgr/J, Stock No: 003025), and bred at the Centre for Addiction and Mental Health (CAMH) (Toronto, Canada). Mice of approximately 4 weeks old were used for all experiments. Control groups were conducted with WT mice. All mouse protocols were approved by the CAMH Animal Care Committee and methods were carried out in accordance with the approved guidelines.
Primary astrocyte culture preparation
Cortical tissues from postnatal day 1–3 (P1–3) mouse brains were dissected out, incubated with 0.25% trypsin for 15 min at 37 °C, and dissociated by mechanical trituration. Cells were plated on T25 cell culture flasks, and grown in DMEM with 10% fetal bovine serum (FBS), 100 U/ml penicillin and 100 μg/ml streptomycin in an incubator (37 °C, 5% CO2) until astrocyte monolayers were confluent. Half of the DMEM was replaced every 3–4 days. To remove microglia and oligodendrocytes from mixed glial cultures, flasks were shaken for 24 h at 200 rpm on an orbital shaker. Afterwards, versene solution and 2.5% trypsin were added to disaggregate astrocytes completely. Purified astrocytes were resuspended in DMEM/10% FBS and plated on either glass coverslips or 60 mm culture dishes, and grown until approximately 50% confluency for immunocytochemistry or fully confluent for Western blot experiments. The purity of astrocytes was confirmed with GFAP/DAPI labeling.
Adult mouse brains (approximately 8 weeks of age) were harvested, fixed in 4% paraformaldehyde (PFA) overnight, cryoprotected in 30% sucrose and frozen at − 80 °C before further processing. 20 μm-thickness frozen coronal sections were cut using a microtome cryostat system (Leica CM3000). Free floating sections were initially blocked in 5% fetal bovine serum, 1% Triton X-100, 0.5% Tween 20 and 1% skim milk in 0.1 M PBS for 2 h at room temperature to reduce non-specific binding. This was followed by incubation with primary antibodies overnight at 4 °C and secondary antibodies for 2 h in blocking solution at room temperature. The primary antibodies used in this study include: anti-NeuN (1:200, Millipore MAB377, Etobicoke, Canada), anti-Parvalbumin (1:200, Sigma-Aldrich P3088, Oakville, Canada), anti-CDP (M-222) (Cux1) (1:200, Santa Cruz Biotechnology sc-13,024, Dallas, TX, USA), anti-Ctip2 (1:200, Abcam ab28448, Cambridge, MS, USA), anti-Olig2 (1:200, Abcam ab109186), anti-Iba1 (1:500, Wako 019–19,741, Richmond, VA, USA) and anti-GFAP (1:200, Dako Z0334, Glostrup, Denmark). Alexa 488- or 594-conjugated secondary antibodies (1:200; Thermo Fisher Scientific, Burlington, Canada) in blocking solution were used for detection of primary antibodies. Staining of myelin and nuclei was performed with Fluoromyelin green and DAPI (Thermo Fisher Scientific) respectively.
Astrocyte cultures were fixed in 4% PFA/4% sucrose, permeabilized with 0.1 M PBS containing 0.1% Triton X-100 for 15 min, and blocked for 1 h with 1% bovine serum albumin in PBS at room temperature. Likewise, they were incubated with primary antibodies overnight at 4 °C and secondary antibodies for 1 h at room temperature. Primary antibodies used for this part include anti-GFAP (1:500; Dako Z0334) and anti-Fmrp (1:200, Abcam ab17722). Alexa 488- phalloidin (Thermo Fisher Scientific A12379) and DAPI were also applied to label F-actin and nuclei respectively.
Fluorescent images of brain sections and astrocyte cultures were captured using a confocal microscope (Olympus FluoView FV1200) at either 10× or 60× magnification. For immunohistochemistry, slices chosen for analyses were anatomically-matched between comparing groups, and included samples from rostral to caudal regions. All images were converted to grey-scale and normalized to background staining. Fluorescently-labeled cells were quantified in specifically-defined regions of the cortex using the ITCN plugin for ImageJ (https://imagej.nih.gov/ij/), where fixed parameters of cell width and threshold are pre-set such that only cells reaching the minimum signal will be counted. For Cux1 and Ctip2, regions of interest (ROIs) were positioned over cortical regions with each ROI further subdivided into eight equal bins from the pia to the inner border of the cortex, to assess neuron distribution across the layers of the cortex (Fig. 2a) . The distribution was expressed as a percentage of the numbers of labeled-cells in each bin divided by the total numbers within each ROI . In addition, fluorescent signal intensities of myelin staining and GFAP immunoreactivity were measured by mean grey values (ImageJ). Similarly with immunocytochemistry on astrocytes, the cell perimeter was first outlined and mean grey values of GFAP fluorescence, as well as surface area were determined as previously described. All image-capturing and threshold parameters were kept the same for each measurement between comparing groups.
Proteins were first extracted from primary astrocyte cultures using RIPA buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 2 mM EDTA, 1 mM PMSF with 1% Igepal CA-630, 1% sodium deoxycholate, 1% Triton X-100 and protease inhibitor cocktail (1:100, Sigma P8340), followed by BCA protein assay (Thermo Fisher Scientific 23,225) to determine the concentration of each sample. 100 μg of solubilized proteins were boiled in Laemmli sample buffer for 5 min, subjected to sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis, and subsequently transferred onto nitrocellulose membranes. The blots were incubated with primary antibodies of anti-GFAP (Dako Z0334), anti-Fmrp (Abcam ab17722) and anti-GAPDH (Millipore MAB374) overnight at 4 °C, then with horseradish peroxidase-conjugated secondary antibodies for 2 h at room temperature. Protein bands were visualized with enhanced chemiluminescence reagents (Amersham Biosciences, Buckinghamshire, UK). The intensity of all resulting bands was quantified by densitometry using ImageJ, normalized to loading bands (GAPDH) and expressed as a percentage of WT controls. Triplicates were performed for each sample.
The Student’s two-tailed t-test was performed to determine statistical differences between WT and Fmr1 KO groups. Neuronal distributions of Cux1 and Ctip2 were analyzed with the one-way ANOVA (GraphPad 5.0), with Bonferroni’s correction for multiple testing. All images were blinded prior to analysis. Data were expressed as mean ± standard error of mean (SEM) and a significance level of p < 0.05 was used.
FHFL and FL designed the study. FHFL prepared all cells and brain tissues, performed all immunofluorescent experiments, image acquisition, data analysis and manuscript writing. FHFL and TKYL performed Western blots and data analysis. PS helped with breeding and maintaining Fmr1 KO mouse colony. FL supervised all projects and contributed in writing the manuscript. All authors read and approved the final manuscript.
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
- 8.Olmos-Serrano JL, Paluszkiewicz SM, Martin BS, Kaufmann WE, Corbin JG, Huntsman MM. Defective GABAergic neurotransmission and pharmacological rescue of neuronal hyperexcitability in the amygdala in a mouse model of fragile X syndrome. J Neurosci. 2010;30(29):9929–38.PubMedPubMedCentralGoogle Scholar
- 16.The Dutch-Belgian Fragile X Consortium, Bakker CE, Verheij C, Willemsen R, van der Helm R, Oerlemans F, et al. Fmr1 knockout mice: a model to study fragile X mental retardation. Cell. 1994;78(1):23–33.Google Scholar
- 24.Stephan AH, Barres BA, Stevens B. The complement system: an unexpected role in synaptic pruning during development and disease. Annu Rev Neurosci. 2012;35:369–89.Google Scholar
- 51.Castren ML. Cortical neurogenesis in fragile X syndrome. Front Biosci. 2016;8:160–8.Google Scholar
- 58.Barnea-Goraly N, Eliez S, Hedeus M, Menon V, White CD, Moseley M, et al. White matter tract alterations in fragile X syndrome: preliminary evidence from diffusion tensor imaging. American journal of medical genetics part B, neuropsychiatric genetics: the official publication of the international society of. Psychiatr Genet. 2003;118B(1):81–8.Google Scholar
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.