Activated Glia May Induce Developmentally Regulated Inflammation and Synaptic Remodeling in a TSC Mouse Model of Epilepsy

cover page image

By: Karbi L. Choudhury


Brain development requires an initial excess formation of synapses, which is followed by synaptic elimination or pruning before adult synaptic connections are established. Both processes are necessary for proper maturation and function of neuronal networks. There is growing evidence that glial cells play an important role in both synaptic formation and pruning, and several age-specific signaling pathways have been identified. Glial-dependent synaptic modeling occurs mainly during the first 3 postnatal weeks in mice, but may extend during adolescence (3-8 weeks). In addition activated microglia and reactive astrocytes are a major source of pro-inflammatory cytokines in the brain, which can enhance excitatory synaptic transmission, while suppressing the inhibitory synaptic responses.

Here, I evaluated whether there are specific changes in glial cell numbers, morphology, and expression profiles during key steps of synaptogenesis in a Tuberous Sclerosis Complex (TSC) mouse model of epilepsy. The activation status of mTOR signaling pathway was assessed by Western blotting for phospho-S6, while glial cell densities were assessed by immunohistochemistry for glial cell markers GFAP, CD68, ALDH1L1, and Iba-1. I found that the mTOR pathway was activated in the TSC mutants at all ages studied, in both neurons and astrocytes. This was accompanied by increased densities of GFAP-expressing reactive astrocytes, but no significant changes were observed in the densities of CD68-expressing activated microglia. However, there was a marked increase in CD68-expressing cells and a striking change in microglial morphology, suggestive of a phagocytic phenotype. The counts of neither ALDH1L1-nor Iba-1-expressing cells were significantly altered, demonstrating no change in total numbers of astrocytes and microglia. A better understanding of altered glial responses in TSC may help develop new therapies for this disorder.


Tuberous sclerosis complex (TSC) is a rare genetic disorder in which mutations of the TSC1 and TSC2 genes lead to formation of benign tumors throughout the body. The upregulation of the mammalian target of rapamycin (mTOR) pathway in response to TSC1/TSC2 inactivation has a primary role in the pathology of TSC. The neurologic symptoms of the disease include seizures, intellectual disability and autism. Epilepsy is the most devastating symptom for TSC patients, as 90% of them will develop seizures within their first year of life, which are often therapy-resistant. Effective seizure control in TSC is critical due to the strong association between early seizure onset, refractory epilepsy, and poor developmental outcome.6 To date, agents inhibiting mTOR signaling pathway have emerged as the most promising therapeutic agents for TSC; however, their potential significant side effects may limit their clinical use. Thus, novel approaches to treat TSC related epilepsy are in high demand. mTOR is well known for regulating many functions related to immune defense and inflammation. Indeed, analysis of human and TSC mouse model brain tissue has shown glial cells activation and upregulation of proteins involved in complement and cytokine signaling. A highly simplified model of the mTOR pathway and its relation to inflammation is presented in Figure 1.

Although it is known that there is a significant increase in inflammation in TSC brain tissue, the relationship between glial cell activation, synaptic development, seizure susceptibility and epilepsy has not been studied. We hypothesize that in TSC, activated microglia and reactive astrocytes may partially retain an immature molecular profile, which will allow them to express a variety of factors required for synaptic remodeling. We anticipate that this study will reveal new mechanisms of vulnerability for epilepsy development in TSC, which could serve as targets for novel treatments administered within an optimal time window.


Animals. Tsc1cc Nestin-rtTA+ tet-OP-cre+ mice were produced by timed inactivation of Tsc1 in neuronal progenitor cells at embryonic day (E)13. This was achieved by treating pregnant dams with doxycycline to induce recombination at the Tsc1 conditional (c) allele in the pups. Levels of recombination were determined by quantitative copy number analysis of the conditional Tsc1c and null Tsc1 alleles through a multiplex ligation-dependent probe assay (MLPA). Only mice with over 12% recombination were utilized.

Tissue Collection and Perfusions. Animals were anesthetized with Nembutol (3 uL per gram). Once the animal reached a surgical plane of anesthesia, it was perfused with ice cold PBS. Brain tissue collected for Western blot was dissected and frozen directly after perfusion with PBS. The animal was further perfused with 4% paraformaldehyde if the brain was being collected for fixation. Fixed brains were post fixed overnight in 4% paraformaldehyde and cryoprotected in 30% sucrose for 2 days.

Western Blot. Tissue was homogenized with a glass dounce homogenizer. Concentrations of protein in homogenized tissue samples were determined using a Bradford assay with standards from 0-2 mg/ml and the Microplate Manager Software. Samples were then equalized with 1X Lysis buffer and 4X Sample Buffer to .8µg/ 1uL in a total volume of 35uL. 30uL of each sample was loaded into a precast 4-20% Tris-Glycine precast gel and run at 115 V for 90 minutes. The gel was then transferred to a membrane overnight. The membrane was probed with markers of the mTOR signaling pathway. The primary antibodies were applied overnight at 4°C. Actin or S6 were used as loading controls. Blots were incubated for 1 hour at room temperature in peroxidase- conjugated mouse or rabbit secondary antibody at a 1:2000 dilution.

Tissue Sectioning and Immunocytochemistry. Fixed tissue was mounted in Tissue-Tek, sectioned at 60 μm using the cryostat, and processed as free-floating sections. Antigen retrieval was performed on the free-floating sections using Citrate Buffer and washed in PBS. Sections were double labeled in primary (See Table 1) and corresponding mouse or rabbit secondary antibody (1:1000) using 5% goat serum in PBS and 0.1% Triton. Upon staining, the sections were mounted onto slides. Slides were then cover slipped using Fluoromount with Dapi, a blue nuclear stain, and viewed using an epifluorescence microscope. Images were captured at 10x and 20x magnification (n=5/brain). Single and double labeled cells were counted for every image using ImageJ software.

Statistical Analyses. For Western blot analysis, TSC-specific measures were normalized to the appropriate age-matched control samples run on the same blot. To determine the statistical significance, group differences were tested using t-tests within individual age groups. For quantitative immunohistochemistry analysis, an average cell count per mm2 was calculated for each brain, and cell densities where compared using two-way ANOVA adjusted by multiple comparisons.


Differential expression of mTOR activity marker phospho-S6 in TSC mutants and controls during early postnatal brain development
The first step in our research was to establish where exactly pS6 is expressed in a cell-specific manner in the cortex of both TSC mutants and wild-type controls (Figure 3).

Utilizing 3 time points in the mouse developmental period (P7, P15, and P30), we assessed levels of phosphorylated ribosomal protein S6 (pS6), a protein downstream of TSC1/TSC2 and mTOR (Figure 3; mutants are represented by the red bar graphs and controls by the blue bar graphs). PS6 has been widely implicated in cell proliferation and growth, and the phosphorylated form over the total protein (S6) levels constitutes an established readout of mTOR activity. At the P7 time point, we see an increase in the levels of pS6 normalized to S6 in the mutants (403±163%, n=5) over the controls (100±41%, n=5), although this increase is not significant (p=0.109). At P15, however, the overall increase in pS6 to S6 ratio is significantly higher in the TSC mutants (324±15%, n=8; p <0.0001) over controls (100±9%, n=8). Finally, P30 mutant mice also exhibit a significantly elevated pS6 to S6 ratio (350±10%, n=6; p<0.0001 over controls (100±12%, n=6).

Next, we sought to establish where exactly pS6 is expressed in a cell-specific manner in the cortex of both TSC mutants and wild-type controls (Figure 2).

Again, pS6 elevation indicates a general increase in activation of the mTOR pathway. Using immunohistochemical staining methods and epifluorecsence imaging, we qualitatively visualized expression of pS6 (green) in the TSC and control brain at both P7 (Panels A-F) and P15 (Panels G-I). In Panels A-F, sections were double labeled with NeuN (red), a neuronal marker. Panels G-I show double labeling with GFAP (red), an astrocyte-specific marker. In Panels A-C, we see normal expression of pS6 in control cortical neurons, In contrast, the levels of pS6 are elevated in a dramatic fashion in the TSC cortical neurons, as indicated by Panels D-F, although the total number of neurons appears unchanged (A and D). Panels G-I show that pS6 induction is not limited to neurons, but is also quite robust in astrocytes in the TSC mutants. Magnification values for each image are indicated in the upper left corner.

Age-specific astroglial phenotypes in TSC mutants and controls
As activation of mTOR in TSC astrocytes may impact their normal development, in this set of experiments I aimed to assess potential changes in astrocytes numbers, along with alterations in cell morphology and expression profile indicative of cellular activation. Therefore, I performed qualitative and quantitative immunohistochemical analysis at different ages, spanning from ages P7 to P50. Reactive astrocytes (in response to mTOR upregulation) were identified by GFAP staining, while total astrocytes were identified by expression of ALDH1L1. At all time points analyzed, the densities of GFAP+ reactive astrocytes are significantly higher in the cortex of TSC mutants (p<0.001), compared to controls (P7 control vs mutant: 166 cells/mm2, n=3 vs 817 cells/mm2, n=4; P15 control vs mutant: 287 cells/mm2 , n=4 vs 857 cells/mm2, n=4; P30-50 control vs mutant: 268 cells/mm2, n=4 vs 1183 cells/mm2, n=3).

In contrast, the total number of astrocytes was not significantly different in the mutants relative to age-matched controls (p>0.05), but there was an obvious decreasing trend from P7 to P30 in both genotypes (P7 control vs mutant: 1080 cells/mm2, n=2 vs 897 cells/mm2, n=2; P15 control vs mutant: 459 cells/mm2, n=2 vs 623 cells/mm2, n=2; P30-50 control vs mutant: 461 cells/mm2, n=1 vs 485 cells/mm2, n=1). TSC mutants do not seem to express any significant changes in total number of astrocytes from their respective age-matched controls.

Finally, regardless of age, we observed a significant change in astrocyte morphology. In control mice, most astrocytes displayed a stellated shape with relatively small cell bodies and thin processes, denoting a resting state. In contrast, in the TSC mice, astrocytes showed enlarged cell bodies and shorter, thicker processes.

Characterization of microglial phenotypes in TSC mutants and controls during brain development
Although Tsc1 levels are preserved in microglia in our mouse model, we also studied microglial reactions in the TSC and control mice at the same ages, as this is a very sensitive indicator of brain inflammation. Activated microglia can be visualized by probing for CD68 in the immunohistochemistry, while total microglia are measured by visualization of Iba1.

In Figure 4,

the quantification of activated microglia with CD68 (Panel A) indicates no significant changes in cell densities in TSC mutants over controls (p>0.05). Also, they trended higher in mutants at P15 (P7 control vs mutant: 324 cells/mm2, n=2 vs 361 cells/mm2, n=5; P15 control vs mutant: 325 cells/mm2, n=2 vs 541 cells/mm2, n=2; P30-50 control vs mutant: 397 cells/mm2, n=2 vs 381 cells/mm2, n=2). In the corresponding images, however, this difference is much more stark and convincing. The mutants (right) at all time points express much more CD68, a lysosomal protein associated with phagocytosis, than their age-matched controls (Panels B, C, & D), suggesting a robust phagocytic microglial phenotype in the mutants. This is accompanied by striking morphological changes (hypertrophied, amoeboid cell bodies), which may allow for the engulfment and elimination of targeted brain structures. However, whether these microglial cells indeed have phagocytic properties requires further investigation.

For total microglial densities (Figure 4, Panel E), there seem to be no differences between mutants and controls, however, there does seem to be more sparse Iba-1+ cells at P30-50 in the mutants compared to controls (P7 control vs mutant: 594 cells/mm2, n=2 vs 535 cells/mm2, n=5; P15 control vs mutant: 539 cells/mm2, n=2 vs 632 cells/mm2, n=2; P30-50 control vs mutant:

499 cells/mm2, n=5 vs 308 cells/mm2, n=7). The corresponding immunostaining also demonstrates this trend (Panels F, G, & H).


Overall, the preliminary data presented here seem to support a trend of increased glial activation in TSC, preceding the age-related onset of seizures (P21-P20) and suggesting a possible contributing role of these cells in seizure generation. Such favorable trends, in all, look promising and are an exciting next step for new findings in epilepsy and TSC.

Astrocytes and microglia have been implicated in the pathophysiology of many neurodevelopmental diseases such as epilepsy, autism and schizophrenia, which are all characterized by defects in synaptic function. Thus, these cell types may be targets for the development of new therapeutic agents to treat these diseases. However, the mechanism by which glial processes mediate synaptic dysfunction and overall cognitive decline in these disorders are poorly understood. Many hypotheses have been suggested regarding astrocyte-induced formation of aberrant glutamatergic connections and cooperative pruning of unwanted synapses by microglia and astrocytes.5

However the role of these processes in epilepsy and Tuberous sclerosis has not been extensively studied. This is a great avenue for research for precisely this reason. How synaptic remodeling can precisely lead to the disease phenotype of epilepsy in TSC is a future direction that my lab and other labs hope to pursue further.


1.Allen N.J., Bennett M.L., Foo L.C., Wang G.X., Chakraborty C., Smith S.J., Barres B.A. Astrocyte glypicans 4 and 6 promote formation of excitatory synapses via GluA1 AMPA receptors. Nature. 2012 May 27; 486(7403): 410
2.Bialas, A.R. & Stevens, B. TGF- signaling regulates neuronal C1q expression and developmental synaptic refinement. Nature Neuroscience. 2013 October 27; 16: 1773-82.
3.Boer, K., Crino, P.B., Gorter J.A., Nellist, M., Jansen FE, Spliet WG, van Rijen PC, Wittink FR, Breit TM, Troost D, Wadman WJ, Aronica E. Gene expression analysis of tuberous sclerosis complex cortical tubers reveals increased expression of adhesion and inflammatory factors. Brain Pathol, 2010. 20(4): p. 704-19.
4.Boer, K., Jansen F., Nellist M., Redeker S., van den Ouweland A.M., Spliet W.G., van Nieuwenhuizen O., Troost D., Crino P.B., Aronica E. Inflammatory processes in cortical tubers and subependymal giant cell tumors of tuberous sclerosis complex. Epilepsy Res, 2008. 78(1): p. 7-21.
5.Clarke, L.E. and B.A. Barres, Emerging roles of astrocytes in neural circuit development. Nature Neuroscience, 2013. 14: p. 311-21.
6.DiMario, F.J., Jr., M. Sahin, and D. Ebrahimi-Fakhari, Tuberous sclerosis complex. Pediatr Clin North Am, 2015. 62(3): p. 633-48.
7.Dube, C., Vezzani, A., Behrens, M., Bartfai, T., Baram, T.Z. Interleukin-1beta contributes to the generation of experimental febrile seizures. Ann Neurol, 2005. 57(1): p. 152-5.
8.Galic, M.A., Riazi K., Heida J.G., Mouihate, A., Fournier N.M., Spencer S.J., Kalynchuk, L.E., Teskey G.C., Pittman Q.J. Postnatal inflammation increases seizure susceptibility in adult rats. J Neurosci, 2008. 28(27): p. 6904-13.
9.Goto, J., Talos D.M., Klein P., Qin W., Chekaluk Y.I., Anderl S., Malinowska I.A., Di Nardo A., Bronson R.T., Chan J.A., Regulable neural progenitor-specific Tsc1 loss yields giant cells with organellar dysfunction in a model of tuberous sclerosis complex. Proceedings of the National Academy of Sciences of the United States of America, 2011. 108(45): p. E1070-9.
10.Kawasaki, Y., Zhang, L., Cheng, J.K., Ji, R.R. Cytokine mechanisms of central sensitization: distinct and overlapping role of interleukin-1beta, interleukin-6, and tumor necrosis factor-alpha in regulating synaptic and neuronal activity in the superficial spinal cord. J Neurosci, 2008. 28(20): p. 5189-94.
11.Kucukdereli, H., Allen, N.J., Lee, A.T., Feng, A., Ozlu, M.I., Conatser, L.M., Chakraborty, C., Workman, G., Weaver, M., Sage, E.H., Barres, B.A., Eroglu, C. Control of excitatory CNS synaptogenesis by astrocyte-secreted proteins Hevin and SPARC. PNAS. 2011 Jun 21; 108(32): 440-449.
12.Lee, R.H., Mills, E.A., Schwartz, N., Bell, M.R., Deeg, K.E., Ruthazer, E.S., Marsh-Armstrong, N., Aizenman, C.D. Neurodevelopmental effects of chronic exposure to elevated levels of pro-inflammatory cytokines in a developing visual system. Neural development, 2010. 5: p. 2.
13.Liauw J., Hoang, S., Choi, M., Eroglu, C., Choi, M., Sun, G., Percy, M., Wildman-Tobriner, B., Bliss, T., Guzman, R.G., 14.Barres, B.A. Thrombospondins 1 and 2 are necessary for synaptic plasticity functional recovery after stroke. Journal of Cerebral Blood Flow and Metabolism. 2008 Oct 1; 28(10): 1722-32.
15.Mottahedin, A., Ardalan M., Chumak T., Riebe I., Ek J., Mallard C., Effect of Neuroinflammation on Synaptic Organization and Function in the Developing Brain: Implications for Neurodevelopmental and Neurodegenerative Disorders. Front Cell Neurosci, 2017. 11: p. 190.
16.Saxton, R.A. and D.M. Sabatini, mTOR Signaling in Growth, Metabolism, and Disease. Cell, 2017. 168(6): p. 960-976.
17.Schafer, D.P., Lehrman, E.K., Kautzman, A.G., Koyama, R., Mardinly, A.R., Yamasaki, R., Ransohoff, R.M., Greenberg, M.E., Barres, B.A., Stevens, B. Microglia sculpt postnatal neural circuits in an activity and complement-dependent manner. Nature. 2012 May 24; 74(4): 691-705.
18.Stevens B., Allen N.J., Vazquez L.E., Howell G.R., Christopherson K.S., Nouri N., Micheva K.D., Mehalow A.K., Huberman A.D., Stafford B. The classical complement cascade mediates CNS synapse elimination. Cell. 2007 Dec 14; 131(6): 1164-78. 19.Vezzani, A., Aronica E., Mazarati A., Pittman Q.J.. Epilepsy and brain inflammation. Exp Neurol, 2013. 244: p. 11-21.
20.Vezzani, A., French J., Bartfai T., Baram T.Z. The role of inflammation in epilepsy. Nat Rev Neurol, 2011. 7(1): p. 31-40.
21.Zhang, B., Zou J., Han L., Rensing N., Wong M., Microglial activation during epileptogenesis in a mouse model of tuberous sclerosis complex. Epilepsia, 2016. 57(8): p. 1317-25.
22.Zhang, B., Zou J., Rensing N.R., Yang M., Wong M.. Inflammatory mechanisms contribute to the neurological manifestations of tuberous sclerosis complex. Neurobiol Dis, 2015. 80: p. 70-9.

Natural Science and Mathematics: Menu

1. Activated Glia May Induce Developmentally Regulated Inflammation and Synaptic Remodeling in a TSC Mouse Model of Epilepsy by Karbi L. Choudhury

2. Effects of early-life seizures on excitatory and inhibitory markers during auditory cortex development by Evan L. Honig

About the Author

Karbi L. Choudhury
University of Pennsylvania

Table of Contents

1. Abstract

2. Introduction

3. Methods

4. Results

5. Discussion

6. References