|Home | About | Journals | Submit | Contact Us | Français|
Tuberous Sclerosis Complex (TSC) and severe cortical dysplasia (CD), or CD Type II according to Palmini’s classification, share histopathologic similarities, specifically the presence of cytomegalic neurons and balloon cells. In this study we examined the morphological and electrophysiological properties of cells in cortical tissue samples from pediatric cases with TSC and CD Type II that underwent surgery for pharmacoresistant epilepsy. Normal-appearing pyramidal neurons from TSC and CD Type II cases had similar passive membrane properties. However, the frequency of excitatory postsynaptic currents (EPSCs) was higher in neurons from TSC compared to severe CD cases, particularly the frequency of medium- and large-amplitude synaptic events. In addition, EPSCs rise and decay times were slower in normal cells from TSC compared to severe CD cases. Balloon cells were found more frequently in TSC cases, whereas cytomegalic pyramidal neurons occurred more often in CD Type II cases. Both cell types were similar morphologically and electrophysiologically in TSC and severe CD. These results suggest that even though the histopathology in TSC and severe CD is similar, there are subtle differences in spontaneous synaptic activity and topographic distribution of abnormal cells. These differences may contribute to variable mechanisms of epileptogenesis in patients with TSC compared with CD Type II.
Tuberous sclerosis complex (TSC) is an autosomal dominant disease associated with mutations in TSC1 or TSC2 genes, which encode for the proteins hamartin and tuberin respectively. When defective, these proteins have been implicated in the genesis of TSC through the mTOR pathway that affects cellular migration, proliferation, and differentiation in multiple organs (Leung and Robson, 2007; Curatolo et al., 2008). In the brain TSC involves cortical and subcortical tubers, as well as subependymal giant cell astrocytomas. Epilepsy is a major neurological complication of TSC, affecting as many as 85% of patients at some point in their lives (Sparagana and Roach, 2000). Cortical tubers and subependymal nodules are thought to be a major determinant of seizure severity (Jansen et al., 2008). Seizures can begin as early as the first day of life, are often associated with infantile spasms, and may be refractory to medical treatment. Cortical dysplasia (CD) of Taylor is another frequent cause of seizures (Taylor et al., 1971) and in pediatric epilepsy surgery cases it is the most prevalent pathologic substrate (Lerner et al., 2009). According to the Palmini classification system (Palmini et al., 2004), two main types of CD can be differentiated based on histopathologic findings: Type I, where cortical dyslamination occurs, and Type II where, in addition to dyslamination, dysplastic cells can be found. Dysplastic cells include cytomegalic neurons (CD Type IIA) and balloon cells (CD Type IIB). The histopathologic cortical abnormalities observed in CD Type IIA/B resemble those found in TSC. Indeed, it has been suggested that severe CD represents a forme fruste or phenotypic variant of TSC (Vinters et al., 1993).
The electrophysiological properties of normal and abnormal cells from surgical CD tissue are beginning to be elucidated (Cepeda et al., 2006; Andre et al., 2007). However, little is known about the electrophysiological properties of cells in cortical tissue from TSC patients. In this study we compare and contrast the biophysical membrane properties and synaptic activity of normal-appearing pyramidal neurons and abnormal, dysplastic cells obtained from TSC and CD Type II patients with intractable epilepsy. This comparison is based on 13 TSC cases (ages 1.1–10.1 yr) and 14 CD Type II A/B cases (ages 1.4–11.9 yr). The clinical protocols used to evaluate patients have been described elsewhere (Mathern et al., 1999; Cepeda et al., 2005). The standardized pre-surgery evaluation included detailed history and neurological evaluations, ictal and interictal scalp EEG recordings, and neurodevelopment assessments. Neuroimaging studies included high resolution magnetic resonance imaging (MRI) and Fluoro-2-deoxyglucose positron emission tomography (FDG-PET).
Neocortical sample sites were excised for in vitro electrophysiological and histological evaluation based on abnormal neuroimaging and electrocorticography assessments. CD tissue samples were classified as most (MA) and least abnormal (LA) according to published criteria (Cepeda et al., 2003). Sample sites were removed microsurgically and directly placed in low Ca2+ artificial cerebrospinal fluid (ACSF). Slices (350 μm) were cut and placed in regular ACSF for at least 1h and then placed in a custom-designed chamber. Individual cells (in layers II–VI) were visualized using infrared videomicroscopy and differential interference contrast (DIC) optics. Patch electrodes were filled with Cs-methanesulfonate- or K-gluconate- based solutions for voltage and current clamp recordings respectively. Electrodes also contained 0.2% biocytin (Sigma, St. Louis, MO) in the internal solution to label visualized cells.
The access resistance ranged from 8–20 MΩ. Passive membrane properties were determined in voltage clamp mode by applying a depolarizing step voltage command (10 mV) and using the membrane test function integrated in the pClamp (version 8) software (Axon Instruments, Foster City, CA). This function reports membrane capacitance (in pF), input resistance (in MΩ), and time constant (in ms). Spontaneous synaptic activity was recorded at a holding potential of −70 mV. At this membrane potential synaptic currents are mediated mostly by α-amino-3-hydroxyl-5-methyl-4-isoxazole-propionate (AMPA) receptors as they are blocked with the AMPA receptor antagonist CNQX (10 μM). In some experiments the GABAA receptor antagonist bicuculline (10 μM) was added to isolate the spontaneous excitatory postsynaptic currents (EPSCs). The Mini Analysis Program (Justin Lee, Synaptosoft, 1999) was used to calculate the frequency of spontaneous synaptic currents and to construct amplitude-frequency histograms. Threshold for event detection (5 pA) was set above root-mean-square noise (2–3 pA at Vhold=−70 mV). After the experiment the slice was fixed in 10% formaldehyde and processed according to published protocols (Horikawa and Armstrong, 1988). Differences between groups were assessed using Student’s t or Mann-Whitney Rank Sum tests, appropriate ANOVA, and χ2-test for distributions.
Cases were selected so that they were closely age-matched. Average ages were 4.3±0.9 yr for TSC (n=13) and 4.4±0.9 yr for CD Type II (n=14, p=0.981). All TSC cases had cortical tubers as demonstrated by MRI and FDG-PET evaluations and verified at histopathology. Samples were from the tuber or perituberal areas and were classified as MA (n=18). Four were classified as LA in TSC cases. Most samples from CD Type II cases were also classified as MA (n=10). Five were classified as LA. However, abnormal cells in CD Type II were seen exclusively in the MA areas. All samples were from parietal, temporal or frontal cortices.
The vast majority of cells sampled in each group were normal-appearing pyramidal neurons (Figure 1A). Based on biocytin labeling, these cells typically had a soma with triangular shape, distinct apical and basal dendrites, and abundant spines. Some cells displayed tortuous dendrites (see Figure 1A, right panel). Cytomegalic pyramidal neurons were 2–3 times larger than normal-appearing pyramidal neurons and had abundant dendrites and spines. Balloon cells were generally round and displayed profuse, thin phyllopodia-like processes that lacked spines. An axon could not be distinguished in balloon cells. Cytomegalic pyramidal neurons were more prevalent in CD Type II cases whereas balloon cells were found more frequently in TSC cases (Table I, p=0.002, χ2-test).
Passive membrane properties of normal-appearing pyramidal neurons, including cell capacitance, input resistance, and time constant, were similar in both groups, although there was a trend for neurons in CD Type II to display larger capacitances (p=0.068, t-test) and lower input resistances (p<0.05, Mann-Whitney test) than in TSC cases (Table II). There were no differences in resting membrane potentials (−62.8±2.2 mV in TSC and −62.1±3.2 mV in CD Type II) or firing properties of normal-appearing pyramidal neurons from TSC or CD Type II cases. In addition, current-voltage relationships were similar (Figure 1B). Cytomegalic pyramidal neurons had larger membrane capacitance and time constant than normal-appearing pyramidal neurons. In contrast, input resistance was lower in cytomegalic neurons. These properties did not differ between TSC and severe CD cases (Table II). Passive membrane properties of balloon cells recorded in TSC or severe CD cases also were similar (Table II). In voltage clamp recordings balloon cells did not display inward currents and in current clamp they did not fire action potentials. In contrast, cytomegalic pyramidal neurons displayed repetitive Ca2+ oscillations, a sign of hyperexcitability.
Spontaneous post-synaptic currents (principally mediated by AMPA receptors) were reduced in normal-appearing pyramidal neurons from CD Type II compared to TSC cases. In addition, medium- and large-amplitude events occurred more frequently in pyramidal neurons from TSC than in neurons from severe CD cases (Figure 2). The kinetics of spontaneous synaptic events was different in TSC and CD Type II. Rise time (1.6±0.1 in TSC and 1.1±0.1 ms in CD Type II), decay time (7.4±0.4 in TSC and 5.2±0.3 ms in CD Type II), and half-amplitude durations (8.0±0.4 in TSC and 5.6±0.4 ms in CD Type II) of spontaneous synaptic currents were significantly slower in TSC compared to CD type II cases (p<0.01). Balloon cells did not display spontaneous synaptic activity and cytomegalic pyramidal neurons displayed low frequency of spontaneous activity.
Although TSC and CD Type II share several histopathologic similarities, particularly the presence of cytomegalic neurons and balloon cells, our study supports the notion that there are subtle functional differences between these etiologies. Although passive membrane properties appeared similar in cells from TSC and severe CD cases, spontaneous excitatory synaptic activity was increased in TSC cases. These differences indicate that normal-appearing pyramidal neurons from TSC cases receive more synaptic inputs than neurons from severe CD cases. In particular, medium- and large-amplitude events occurred more frequently in neurons from TSC cases. These events are usually induced by presynaptic action potentials and suggest more pyramidal neuron firing in TSC compared to severe CD networks. The kinetic properties of the spontaneous excitatory events were slower in TSC compared to severe CD cases, suggesting differences in glutamate receptor subunit composition or glutamate transporters between the two etiologies.
The electrophysiological properties of cytomegalic neurons and balloon cells were similar in both TSC and CD Type II cases. Cytomegalic neurons showed signs of hyperexcitability whereas balloon cells were unable to generate action potentials, confirming our previous studies (Mathern et al., 2000; Cepeda et al., 2003). Balloon cells, obtained mostly from TSC cases, were immunocytologically positive for neuronal and/or glial markers (Mathern et al., 2000) indicating their undifferentiated status. We and others have proposed that balloon cells may be remnants of radial glial progenitor cells (Cepeda et al., 2006; Lamparello et al., 2007) or from the minority of embryonic progenitors that co-express neuronal and glial markers (Zecevic, 2004). Thus, in terms of timing of differentiation and/or migration defects, it appears that both TSC and CD Type II would occur earlier in cortical neurogenesis, in contrast CD Type I would occur in later phases of brain development, once cellular differentiation has been completed (Mathern et al., 2007).
The observation that balloon cells were more prevalent in the cortex of TSC cases indicates different degrees of severity or migration patterns in both pathologies. Morphological studies have shown that in severe CD cases balloon cells are more common in the white matter (Vinters et al., 1993), an area rarely sampled in our studies due to technical limitations. This topographic distribution could lead to differential synaptic arrangements and mechanisms of epileptogenesis between TSC and CD Type II. More studies are necessary to determine if those mechanisms involve cellular or network differences or a combination of both.
A recent study in a patient with a missense mutation in the TSC2 gene, albeit without a cortical tuber identified by neuroimaging, reported a number of electrophysiological measures in normal-appearing pyramidal neurons that suggested alterations in synaptic excitation but no changes in inhibition (Wang et al., 2007). Unfortunately, the frequency of spontaneous EPSCs was not reported in that study. Excitatory activity, manifested by large amplitude inward currents, was evoked by bath application of bicuculline. Thus, a direct comparison with the present results is not possible. However, the results reported here are in line with our previous study demonstrating reduced glutamatergic relative to GABAergic synaptic activity in severe CD cases, which suggested delayed maturation of cortical networks (Cepeda et al., 2005). In conclusion, although based on histopathology TSC and CD Type II cases share many similarities, emerging evidence supports that there are subtle but significant differences between the two pathologies. Further, in a recent study we showed that TSC patients display reduced neuronal densities in lower gray matter whereas CD Type IIB patients have increased neuronal densities in the upper cortical and white matter regions (Chandra et al., 2007). Finally, it should be pointed out that the present study is limited to pediatric patients with TSC and we do not know if these results are applicable to adult cases after many years of seizures. Indeed, our recent data indicate that seizure freedom after surgery correlates with younger age and shorter seizure duration in children with TSC (Wu et al., 2009).
The authors would like to thank the patients and their parents for allowing use of resected tissue for experimentation. We also thank the UCLA Hospital Pediatric Neurology staff for their assistance. Donna Crandall helped with the illustrations. This study was supported by NIH grant NS 38992.
Disclosures: The contributing authors to this article declare no commercial conflicts of interest.