A thorny question: how does activity maintain dendritic spines?

The immediate large-scale dendritic plasticity of cortical pyramidal neurons subjected to acute epidural compression

Head trauma and acute disorders often instantly compress the cerebral cortex and lead to functional abnormalities. Here we used rat epidural bead implantation model and investigated the immediate changes following acute compression. The dendritic arbors of affected cortical pyramidal neurons were filled with intracellular dye and reconstructed 3-dimensionally for analysis. Compression was found to shorten the apical, but not basal, dendrites of underlying layer III and V cortical pyramidal neurons and reduced dendritic spines on the entire dendritic arbor immediately. Dendrogram analysis showed that in addition to distal, proximal apical dendrites also quickly reconfigured. We then focused on apical dendritic trunks and explored how proximal dendrites were rapidly altered. Compression instantly twisted the microtubules and deformed the membrane contour of dendritic trunks likely a result of the elastic nature of dendrites as immediate decompression restored it and stabilization of microtubules failed to block it. Subsequent adaptive remodeling restored plasmalemma and microtubules to normal appearance in 3 days likely via active mechanisms as taxol blocked the restoration of microtubules and in addition partly affected plasmalemmal reorganization which presumably engaged recycling of excess membrane. In short, the structural dynamics and the associated mechanisms that we revealed demonstrate how compression quickly altered the morphology of cortical output neurons and hence cortical functions consequently.

The cytoarchitecture and soma-dendritic arbors of the pyramidal neurons of aged rat sensorimotor cortex: an intracellular dye injection study

We studied the cytoarchitecture and dendritic arbors of the output neurons of the sensorimotor cortex of aged rats and found that although individual cortical layer became thinner, the overall cytoarchitecture and neuron densities remained comparable to those of young adults. To find out whether aging affects cortical outputs we studied the soma-dendritic arbors of layers III and V pyramidal neurons, main output neurons of the cerebral cortex, using brain slice intracellular dye injection technique. With a fluorescence microscope, selected neurons were filled with fluorescence dye under visual guidance. Injected slices were resectioned into thinner sections for converting the injected dye into non-fading material immunohistochemically. The long apical dendritic trunk and branches could be routinely revealed. This allowed us to reconstruct and study the dendritic arbors of these neurons in isolation in 300-microm-thick dimension. Analysis shows that their cell bodies did not shrink, but the densities of spines on dendrites and the total dendritic length significantly reduced. Among spines, those with long thin stalks thought to be involved in memory acquisition appeared to be reduced. These could underlie the compromise of sensorimotor functions following aging.

Differential expression of NMDA and AMPA/KA receptor subunits in the inferior olive of postnatal rats

We have employed immunohistochemistry to determine the expression patterns of receptor subunits of N-methyl-d-aspartate (NMDA-NR1 and NR2A/B) and alpha-amino-3-hydroxy-5-methyl-isoxazole-4-propionic acid/kainic acid (AMPA/KA-GluR1, GluR2, GluR2/3, GluR4, and GluR5/6/7) in the inferior olive of postnatal rats up to adulthood. Immunoreactivity for distinct receptor subunits was predominantly localized in the soma and dendrites of neurons. Semi-quantification showed that the overall immunoreactivity in the inferior olive of adults was intense for GluR1, moderate for NR1 and NR2A/B, and low for GluR2, GluR2/3, GluR4, and GluR5/6/7. At P7, GluR1 was restricted to the dorsomedial cell column, subnucleus beta, principal nucleus and ventrolateral protrusion while the other subunits were found in all subnuclei of the inferior olive. The immunoreactivities for all glutamate receptor subunits ranged from low to moderate. As the rats matured, the immunoreactivity of GluR4 decreased after the second postnatal week, while those of the other subunits showed a general trend of increase, reaching adult level during the third postnatal week. Double immunofluorescence revealed that all NR1-containing neurons exhibited NR2A/B immunoreactivity, indicating that native NMDA receptors comprise of hetero-oligomeric combinations of NR1 and NR2A/B. Furthermore, co-localization of NMDA and AMPA/KA receptor subunits was demonstrated in individual neurons of the inferior olive. All NR1-containing neurons exhibited GluR1 immunoreactivity, and all NR2A/B-containing neurons showed GluR5/6/7 immunoreactivity. Our data suggest that NMDA and AMPA/KA receptors are involved in glutamate-mediated neurotransmission, contributing to synaptic plasticity and reorganization of circuitry in the inferior olive during postnatal development.

The effect of epidural compression on cerebral cortex: a rat model

We developed a rat model of epidural plastic bead implantation to study the effect of physical compression on the cerebral cortex. Epidural implantation of a bead of appropriate size compressed the underlying sensorimotor cortex without apparent ischemia, since the capillary density of the cortex was increased. Although the thickness of all layers of the compressed cortex was significantly decreased, no apparent changes in the number of NADPH-diaphorase reactive neurons, reactive astrocytes, or microglial cells were observed, nor were apoptotic neurons observed. In fact, the densities of the neurons in most cortical layers apparently increased. To determine how epidural compression affects neuronal morphology, the dendritic arbors of layer III and V pyramidal neurons were evaluated using a fixed tissue intracellular dye injection technique. Neurons in both layers remained pyramidal in shape and their somatic sizes remained unaltered for at least a month after compression. On the other hand, their total dendritic length was significantly reduced beginning at 3 days post implantation. These analyses showed that apical dendrites were affected sooner than basal ones. The reduction of dendritic length was associated with a drop in the number of dendritic branches rather than dendritic trunks, suggesting the trimming of the peripheral part of the dendritic arbor. Detailed analysis showed that dendritic spines on all dendrites were reduced as early as 3 days following implantation. These results suggest that cortical neurons remodel their structures substantially within 3 days after being subjected to epidural compression.

Co-localization of NMDA receptors and AMPA receptors in neurons of the vestibular nuclei of rats

We are interested in studying the co-localization of NMDA glutamate receptor subunits (NR1, NR2A/B) and AMPA glutamate receptor subunits (GluR1, GluR2, GluR2/3 and GluR4) in individual neurons of the rat vestibular nuclei. Immunoreactivity for NR1, NR2A/B, GluR1, GluR2, GluR2/3 and GluR4 was found in the somata and dendrites of neurons in the four major subdivisions (superior, medial, lateral, and spinal vestibular nuclei) and in two minor groups (groups x and y) of the vestibular nuclei. Double immunofluorescence showed that all the NR1-containing neurons exhibited NR2A/B immunoreactivity, indicating that native NMDA receptors are composed of NR1 and NR2A/B in a hetero-oligomeric configuration. Co-expression of NMDA receptor subunits and AMPA receptor subunits was demonstrated by double labeling of NR1/GluR1, NR1/GluR2/3, NR1/GluR4 and NR2A/B/GluR2 in individual vestibular nuclear neurons. All NR1-containing neurons expressed GluR2/3 immunoreactivity, and all NR2A/B-containing neurons expressed GluR2 immunoreactivity. However, only about 52% of NR1-immunoreactive neurons exhibited GluR1 immunoreactivity and 46% of NR1-containing neurons showed GluR4 immunoreactivity. The present data reveal that NMDA receptors are co-localized with variants of AMPA receptors in a large proportion of vestibular nuclear neurons. These results suggest that cross-modulation between NMDA receptors and AMPA receptors may occur in individual neurons of the vestibular nuclei during glutamate-mediated excitatory neurotransmission and may in turn contribute to synaptic plasticity within the vestibular nuclei.

Development of salt-responsive neurons in the nucleus of the solitary tract

The rodent gustatory system has become a popular and useful model for the study of brain development because of this system's protracted period of postnatal maturation and its sensitivity to subtle changes in the animal's sensory environment. The goal of this investigation was to improve our understanding of dendritic remodeling exhibited by second-order gustatory neurons by presenting a comprehensive and definitive description of the development of the dendritic architecture of taste-sensitive neurons in the rostral nucleus of the solitary tract. Extracellular and intracellular recording and intracellular labeling techniques were used to examine the structure and function of individual gustatory neurons in three groups of rats: (1) Postnatal day 13-21 (PND13-21), (2) Postnatal day 22-28 (PND22-28), and (3) Adult (postnatal day 60-90). We found that neurons that responded to all three of the salts in our taste array ("Salt Sensitive") exhibited a striking increase in the number of dendritic branch points, maximum branch order, swelling density, and spine density between the PND13-21 and PND22-28 periods. These increases were followed by a period of dendritic remodeling during which the values for all measures except spine density decreased significantly. The neurons that did not respond to all three salts exhibited no change in the number of dendritic branches, branch order, or spine density during development, but they did undergo a decrease in swelling density. We also found that there was a significant decrease in the total dendritic length and cell volume of Salt Sensitive neurons between the PND22-28 and Adult periods, whereas the cells that did not respond to all three salts exhibited an increase in dendritic length and cell volume between postnatal day 28 and adulthood. Finally, we found that the dendrites of the Adult Salt Sensitive neurons were more restricted in the rostrocaudal axis than either the PND13-21 or PND22-28 Salt Sensitive cells. In contrast, there were no significant changes in the rostrocaudal extent of the dendritic arbors of cells that did not respond to all three salts. When viewed in the context of the extant literature and our own preliminary studies that used modified salt diets, we propose that these results provide strong support for the hypothesis that there is a relationship between postnatal dendritic development (particularly remodeling) and the animal's sensitivity to salts.