Neuromorphology
Determine the cellular and synaptic attributes of HFOs in the epileptogenic human brain tissue.
Our working hypothesis, again based on our preliminary data, is that the underlying morphological structure and synaptic profiles of cortical pyramidal and non-pyramidal neurons are significantly different in the epileptogenic zone compared to the seizure spread zone and silent areas that may contain only spikes but no HFOs. We will use post-operative criteria, such as seizure freedom, as proof that the resected area was epileptogenic.
2. Cellular and synaptic attributes of HFOs in the epileptogenic human brain tissue.
Pyramidal cell morphology in seizure onset zone, seizure spread zone and silent cortical areas. We examined high-resolution structure of human neocortical pyramidal neurons in three salient cortical areas associated with seizures, namely the SOZ, the seizure spread zone (SSZ) and the silent area with no seizure propagation. All regions were located within the inferior parietal lobule of two patients with epilepsy. The type of seizure locus was determined based on iEEG recording and mapping by electrical stimulation of each subdural electrode. The null hypothesis of no significant morphological changes in the synaptic connectivity of each area and each spine type (mushroom and thin) was tested, as measured by three spine morphological features (spine length, spine volume, and spine head diameter). Detailed structural data for layer II and III and layer V pyramidal neurons of each area were obtained through morphometric analysis of Lucifer Yellow filled principal neurons in brain slices prepared from surgical resections (300 µm thick). These neurons were imaged at high magnification (63X, 1.4 N.A. 100 nm cubic voxel) using confocal laser scanning microscopy (CLSM). The neuronal morphology, dendritic arbor and dendritic spine morphology were extracted from stacks of scanned images using a custom-designed algorithm developed in one of our collaborator's laboratory (PRH) (NeuronStudio, REF). Morphometric analysis provided unbiased measurements of both local and global structure of the pyramidal neurons in the SOZ, SSZ and the silent zone. We found the head diameter of thin and the length of mushroom dendritic spines to be significantly smaller in the SOZ and SSZ compared to the silent area. The head diameter of mushroom dendritic spines was significantly higher in the SOZ compared to SSZ and the silent area. The volume of mushroom dendritic spines was significantly higher in SOZ compared to SSZ and in SSZ compared to the silent area. The length of thin dendritic spines was significantly lower in the SOZ compared to SSZ and the silent area. Corrected p-values were ≤ 0.05. The null hypothesis was therefore rejected based on significant changes in SOZ, such as strengthened synaptic input to stable mushroom spines. These findings provide morphological evidence on how and why the SOZ is more prone to seizures, which is a necessary step to understand the pathophysiology of epileptogenesis. Figure 5 shows a layer III pyramidal neurons from the inferior temporal neocortex of an epilepsy patient that is loaded intracellularly with LY and imaged at high resolution using CLSM. Rigorous morphometric analysis of similarly dye loaded neurons using NeuronStudio is shown in B. These analysis provided accurate morphometric measures of dendrites and dendritic spines (each spine color coded according to its type: mushroom, thin, stubby) and represented in 2D (C), 3D (D) and high resolution (E). Scale bar 20 µm in A, B; 4 µm in C, D; 2 µm in E.
2. Cellular and synaptic attributes of HFOs in the epileptogenic human brain tissue.
Pyramidal cell morphology in seizure onset zone, seizure spread zone and silent cortical areas. We examined high-resolution structure of human neocortical pyramidal neurons in three salient cortical areas associated with seizures, namely the SOZ, the seizure spread zone (SSZ) and the silent area with no seizure propagation. All regions were located within the inferior parietal lobule of two patients with epilepsy. The type of seizure locus was determined based on iEEG recording and mapping by electrical stimulation of each subdural electrode. The null hypothesis of no significant morphological changes in the synaptic connectivity of each area and each spine type (mushroom and thin) was tested, as measured by three spine morphological features (spine length, spine volume, and spine head diameter). Detailed structural data for layer II and III and layer V pyramidal neurons of each area were obtained through morphometric analysis of Lucifer Yellow filled principal neurons in brain slices prepared from surgical resections (300 µm thick). These neurons were imaged at high magnification (63X, 1.4 N.A. 100 nm cubic voxel) using confocal laser scanning microscopy (CLSM). The neuronal morphology, dendritic arbor and dendritic spine morphology were extracted from stacks of scanned images using a custom-designed algorithm developed in one of our collaborator's laboratory (PRH) (NeuronStudio, REF). Morphometric analysis provided unbiased measurements of both local and global structure of the pyramidal neurons in the SOZ, SSZ and the silent zone. We found the head diameter of thin and the length of mushroom dendritic spines to be significantly smaller in the SOZ and SSZ compared to the silent area. The head diameter of mushroom dendritic spines was significantly higher in the SOZ compared to SSZ and the silent area. The volume of mushroom dendritic spines was significantly higher in SOZ compared to SSZ and in SSZ compared to the silent area. The length of thin dendritic spines was significantly lower in the SOZ compared to SSZ and the silent area. Corrected p-values were ≤ 0.05. The null hypothesis was therefore rejected based on significant changes in SOZ, such as strengthened synaptic input to stable mushroom spines. These findings provide morphological evidence on how and why the SOZ is more prone to seizures, which is a necessary step to understand the pathophysiology of epileptogenesis. Figure 5 shows a layer III pyramidal neurons from the inferior temporal neocortex of an epilepsy patient that is loaded intracellularly with LY and imaged at high resolution using CLSM. Rigorous morphometric analysis of similarly dye loaded neurons using NeuronStudio is shown in B. These analysis provided accurate morphometric measures of dendrites and dendritic spines (each spine color coded according to its type: mushroom, thin, stubby) and represented in 2D (C), 3D (D) and high resolution (E). Scale bar 20 µm in A, B; 4 µm in C, D; 2 µm in E.
Pyramidal cell morphology in human cortical dysplasia. Malformations of cortical development underlie a significant portion of adult and pediatric forms of intractable epilepsy. Severe forms of focal cortical dysplasia are characterized by dysmorphic/cytomegalic neurons and balloon cells in the context of pervasive laminar disorganization. To date, direct measures of the morphology of human dysplastic neurons have not been possible on a large scale. We examined high-resolution structure of dysplastic pyramidal neurons in the SOZ of the inferior parietal lobule of two adult patients with epilepsy. Pyramidal neurons of presumably normal cortical as determined by iEEG, cortical mapping, and staining for non-phosphorylated neurofilament staining (showing normal cortical lamination) were used as control (Fig. 6). The null hypothesis of no significant morphological changes in the synaptic connectivity of the SOZ and each spine type (mushroom and thin) were tested, as measured by spine length, spine volume, and spine head diameter. Detailed structural data for layer II and III and layer V dysplastic pyramidal neurons of the SOZ were obtained through morphometric analysis of Lucifer Yellow filled neurons in brain slices prepared from surgical resections (300 µm thick). Figure 6 shows 2D projection high-resolution CLSM of the human parietal neocortex stained immunohistochemically for non-phosphorylated neurofilament proteins (NPNFP, Green label) and for NeuroTrace (red label). (A) Broad overview showing the full thickness cortical mantle (layers I-VI). (B) Boxed region from A digitally magnified. The high-resolution imaging technique allows viewing and recording of all cortical layers for analysis of cortical dysplasias. Scale bar 125 µm in A; 30 µm in B.
Morphometric analysis provided unbiased measurements of both local and global structure of the pyramidal neurons. Basal thin spines were found to be significantly longer in the control than in the dysplastic cells (p ≤ 0.001, corrected), with differences in mean lengths differing by as much as 0.2 micrometers at 95% confidence. The mean volume of basal mushroom spines, however, differed by as much as 0.04 cubic micrometers (95% confidence bound), nearly 30% of the mean volume of dysplastic basal mushroom spines (p ≤ 0.01, corrected), with the control being larger. Thin spine head diameter also varied by a difference in means of 0.1 micrometers and 0.12 micrometers (apical, basal) between the dysplastic cells and control cell (p ≤ 0.001, corrected), with control being larger. The null hypothesis was therefore rejected based on significant changes in the dysplastic pyramidal neurons of the SOZ, such as significant loss of synaptic input to stable mushroom spines. These findings provide morphological evidence on how and why the dysplastic tissue is more prone to seizures in intractable epilepsy. Figure 7 shows 2D projection high-resolution CLSM of the dysplastic human parietal neocortex stained immunohistochemically for non-phosphorylated neurofilament proteins (NPNFP, Green label) and for NeuroTrace (red label). Dystrophic neurons were observed in all cortical layers. Morphological distortions include corkscrewing of the dendrites (arrowheads; A, C), and neuronal clustering both within a layer (B, horizontally) and across layers (C, vertically). These clustered dystrophic neurons typically demonstrate both distorted apical and basal dendrites, and morphologically are considerably larger in size (D). Scale bar 100 µm in A, B, C; 20 µm in D.
Synaptic and nonsynaptic (gap junctional) organization underlying HFO generating epileptogenic areas.
Figure 8 demonstrates array tomography of PSD95, GluN1 and Synapsin distribution in relation to dendritic spines in a Lucifer Yellow-filled pyramidal neuron from the human temporal neocortex. (A) The top panels demonstrates array tomography sectioning through a Lucifer Yellow-filled (green) and plastic-embedded neuron and shows (left to right) a series of four consecutive ultra thin sections (90 nm each) and a collapsed 3-D reconstruction showing two dendritic spines (arrows). (B) Lucifer Yellow (green) co-labeled with PSD95 (red). (C) Lucifer Yellow co-labeled with GluN1 (blue). (D) Lucifer Yellow co-labeled with Synapsin (cyan). Scale bar 2 µm.
Figure 9 shows serial section immunogold transmission electron microscopy (TEM) of an asymmetric (excitatory) synapse from layer III of the human temporal neocortex. Anti-GluA2/3 antibody conjugated to 10-nm immunogold particles (arrows) shows highly concentrated immunogold reaction throughout the postsynaptic density (PSD), and is commonly seen throughout the spine head as well as the presynaptic axon terminal. The asterisk in IV represents a perforated (split PSD) synapse. Note the spine apparatus in IV. This region commonly contains excitatory amino acid receptor localization as it represents the site of local protein synthesis. Note the longitudinally sectioned dendrite to the right of the excitatory synapse. ax, axon; sp, spine; sa, spine apparatus; den, dendrite. Scale bar 0.4 µm.
Figure 10 shows correlative confocal-electron microscope imaging of a LY filled pyramidal neuron from the human inferior temporal gyrus embedded in Lowicryl resin and sectioned on an ultra microtome. Serial ultra thin sections (~90 nm-thin) were mounted on a pioloform- coated cover slip and imaged using CLSM (A). The sections were lifted from the cover slip using dilute hydrofluoric acid in water, floated onto a water surface and recaptured using formvar/carbon-coated slot grids for TEM (B). An overlay of A and B is shown in C. The arrow in B and C represents the remnants of the micropipette tract from the iontophoretic cell filling. This technique can be used as an extension for both array tomography (Figure 8) and postembedding immunogold TEM (Figure 9) for ultrastructural correlations. nuc, nucleus; den, dendrite. Scale bar 6.5 µm.
Pyramidal cell morphology in human cortical dysplasia. Malformations of cortical development underlie a significant portion of adult and pediatric forms of intractable epilepsy. Severe forms of focal cortical dysplasia are characterized by dysmorphic/cytomegalic neurons and balloon cells in the context of pervasive laminar disorganization. To date, direct measures of the morphology of human dysplastic neurons have not been possible on a large scale. We examined high-resolution structure of dysplastic pyramidal neurons in the SOZ of the inferior parietal lobule of two adult patients with epilepsy. Pyramidal neurons of presumably normal cortical as determined by iEEG, cortical mapping, and staining for non-phosphorylated neurofilament staining (showing normal cortical lamination) were used as control (Fig. 6). The null hypothesis of no significant morphological changes in the synaptic connectivity of the SOZ and each spine type (mushroom and thin) were tested, as measured by spine length, spine volume, and spine head diameter. Detailed structural data for layer II and III and layer V dysplastic pyramidal neurons of the SOZ were obtained through morphometric analysis of Lucifer Yellow filled neurons in brain slices prepared from surgical resections (300 µm thick). Figure 6 shows 2D projection high-resolution CLSM of the human parietal neocortex stained immunohistochemically for non-phosphorylated neurofilament proteins (NPNFP, Green label) and for NeuroTrace (red label). (A) Broad overview showing the full thickness cortical mantle (layers I-VI). (B) Boxed region from A digitally magnified. The high-resolution imaging technique allows viewing and recording of all cortical layers for analysis of cortical dysplasias. Scale bar 125 µm in A; 30 µm in B.
Morphometric analysis provided unbiased measurements of both local and global structure of the pyramidal neurons. Basal thin spines were found to be significantly longer in the control than in the dysplastic cells (p ≤ 0.001, corrected), with differences in mean lengths differing by as much as 0.2 micrometers at 95% confidence. The mean volume of basal mushroom spines, however, differed by as much as 0.04 cubic micrometers (95% confidence bound), nearly 30% of the mean volume of dysplastic basal mushroom spines (p ≤ 0.01, corrected), with the control being larger. Thin spine head diameter also varied by a difference in means of 0.1 micrometers and 0.12 micrometers (apical, basal) between the dysplastic cells and control cell (p ≤ 0.001, corrected), with control being larger. The null hypothesis was therefore rejected based on significant changes in the dysplastic pyramidal neurons of the SOZ, such as significant loss of synaptic input to stable mushroom spines. These findings provide morphological evidence on how and why the dysplastic tissue is more prone to seizures in intractable epilepsy. Figure 7 shows 2D projection high-resolution CLSM of the dysplastic human parietal neocortex stained immunohistochemically for non-phosphorylated neurofilament proteins (NPNFP, Green label) and for NeuroTrace (red label). Dystrophic neurons were observed in all cortical layers. Morphological distortions include corkscrewing of the dendrites (arrowheads; A, C), and neuronal clustering both within a layer (B, horizontally) and across layers (C, vertically). These clustered dystrophic neurons typically demonstrate both distorted apical and basal dendrites, and morphologically are considerably larger in size (D). Scale bar 100 µm in A, B, C; 20 µm in D.
Synaptic and nonsynaptic (gap junctional) organization underlying HFO generating epileptogenic areas.
Figure 8 demonstrates array tomography of PSD95, GluN1 and Synapsin distribution in relation to dendritic spines in a Lucifer Yellow-filled pyramidal neuron from the human temporal neocortex. (A) The top panels demonstrates array tomography sectioning through a Lucifer Yellow-filled (green) and plastic-embedded neuron and shows (left to right) a series of four consecutive ultra thin sections (90 nm each) and a collapsed 3-D reconstruction showing two dendritic spines (arrows). (B) Lucifer Yellow (green) co-labeled with PSD95 (red). (C) Lucifer Yellow co-labeled with GluN1 (blue). (D) Lucifer Yellow co-labeled with Synapsin (cyan). Scale bar 2 µm.
Figure 9 shows serial section immunogold transmission electron microscopy (TEM) of an asymmetric (excitatory) synapse from layer III of the human temporal neocortex. Anti-GluA2/3 antibody conjugated to 10-nm immunogold particles (arrows) shows highly concentrated immunogold reaction throughout the postsynaptic density (PSD), and is commonly seen throughout the spine head as well as the presynaptic axon terminal. The asterisk in IV represents a perforated (split PSD) synapse. Note the spine apparatus in IV. This region commonly contains excitatory amino acid receptor localization as it represents the site of local protein synthesis. Note the longitudinally sectioned dendrite to the right of the excitatory synapse. ax, axon; sp, spine; sa, spine apparatus; den, dendrite. Scale bar 0.4 µm.
Figure 10 shows correlative confocal-electron microscope imaging of a LY filled pyramidal neuron from the human inferior temporal gyrus embedded in Lowicryl resin and sectioned on an ultra microtome. Serial ultra thin sections (~90 nm-thin) were mounted on a pioloform- coated cover slip and imaged using CLSM (A). The sections were lifted from the cover slip using dilute hydrofluoric acid in water, floated onto a water surface and recaptured using formvar/carbon-coated slot grids for TEM (B). An overlay of A and B is shown in C. The arrow in B and C represents the remnants of the micropipette tract from the iontophoretic cell filling. This technique can be used as an extension for both array tomography (Figure 8) and postembedding immunogold TEM (Figure 9) for ultrastructural correlations. nuc, nucleus; den, dendrite. Scale bar 6.5 µm.