Borna disease virus infection alters synaptic input of neurons in rat dentate gyrus

Granule cells are major targets of entorhinal afferents terminating in a laminar fashion in the outer molecular layer of the dentate gyrus. Since Borna disease virus (BDV) infection of newborn rats causes a progressive loss of granule cells in the dentate gyrus, entorhinal fibres become disjoined from their main targets. We have investigated the extent to which entorhinal axons react to this loss of granule cells. Unexpectedly, anterograde DiI tracing has shown a prominent layered termination of the entorhinal projection, despite an almost complete loss of granule cells at 9 weeks after infection. Combined light- and electron-microscopic analysis of dendrites at the outer molecular layer of the dentate gyrus at 6 and 9 weeks post-infection has revealed a transient increase in the synaptic density of calbindin-positive granule cells and parvalbuminergic neurons after 6 weeks. In contrast, synaptic density reaches values similar to those of uninfected controls 9 weeks post-infection. These findings indicate that, after BDV infection, synaptic reorganization processes occur at peripheral dendrites of the remaining granule cells and parvalbuminergic neurons, including the unexpected persistence of entorhinal axons in the absence of their main targets.     

Glutamate spillover mediates excitatory transmission in the rat olfactory bulb.

In the CNS, glutamate typically mediates excitatory transmission via local actions at synaptic contacts. In the olfactory bulb, mitral cell dendrites release glutamate at synapses formed only onto the dendrites of inhibitory granule cells. Here, I show excitatory transmission mediated solely by transmitter spillover between mitral cells in olfactory bulb slices. Dendritic glutamate release from individual mitral cells causes self-excitation via local activation of N-methyl-D-aspartate (NMDA) receptors. Paired recordings reveal that glutamate release from one cell generates NMDA receptor-mediated responses in neighboring mitral cells that are enhanced by blockade of glutamate uptake. Furthermore, spillover generates spontaneous NMDA receptor-mediated population responses. This simultaneous activation of neighboring mitral cells by a diffuse action of glutamate provides a mechanism for synchronizing olfactory principal cells.     

Miniature neurotransmission stabilizes synaptic function via tonic suppression of local dendritic protein synthesis

Activity deprivation in neurons induces a slow compensatory scaling up of synaptic strength, reflecting a homeostatic mechanism for stabilizing neuronal activity. Prior studies have focused on the loss of action potential (AP) driven neurotransmission in synaptic homeostasis. Here, we show that the miniature synaptic transmission that persists during AP blockade profoundly shapes the time course and mechanism of homeostatic scaling. A brief blockade of NMDA receptor (NMDAR) mediated miniature synaptic events ("minis") rapidly scales up synaptic strength, over an order of magnitude faster than with AP blockade alone. The rapid scaling induced by NMDAR mini blockade is mediated by increased synaptic expression of surface GluR1 and the transient incorporation of Ca2+-permeable AMPA receptors at synapses; both of these changes are implemented locally within dendrites and require dendritic protein synthesis. These results indicate that NMDAR signaling during miniature synaptic transmission serves to stabilize synaptic function through active suppression of dendritic protein synthesis.     

Lamina-specific synaptic connections of hippocampal neurons in vitro

By using slice cultures as a model, we demonstrate here that different target selectivities exist among the various afferent fibers to the hippocampus. As in intact animals, septohippocampal cholinergic fibers, provided by a slice culture of septum, innervate a co-cultured slice of hippocampus diffusely, that is, without forming distinct layers of termination. As in vivo, the septal cholinergic fibers establish synapses with a variety of target cells. Conversely, fibers from an entorhinal slice co-cultured to a hippocampal slice display their normal laminar specificity. They preferentially terminate in the outer molecular layer of the fascia dentata, thereby selectively contacting peripheral dendrites of the granule cells. This preferential termination on peripheral dendritic segments is remarkable, since these fibers do not have to compete with commissural fibers, hypothalamic fibers, and septal afferents for dendritic space under these culture conditions. Moreover, in triplet cultures in which first two hippocampal slices were co-cultured and then, with a delay of 5 days, an entorhinal slice was added, the fibers from the entorhinal slice and those from the hippocampal culture terminated in their appropriate layers in the hippocampal target culture. However, in this approach the normal sequence of ingrowth of these two afferents was reversed. In normal ontogenetic development, entorhinal afferents arrive in the hippocampus before the commissural fibers. The results show that there are different degrees of target selectivity of hippocampal afferents and that the characteristic lamination of certain afferent fibers in the hippocampus is not determined by their sequential ingrowth during development. © 1995 John Wiley & Sons, Inc.     

Dendritic integration in hippocampal dentate granule cells

Hippocampal granule cells are important relay stations that transfer information from the entorhinal cortex into the hippocampus proper. This process is critically determined by the integrative properties of granule cell dendrites. However, their small diameter has so far hampered efforts to examine their properties directly. Using a combination of dual somatodendritic patch-clamp recordings and multiphoton glutamate uncaging, we now show that the integrative properties of granule cell dendrites differ substantially from other principal neurons. Due to a very strong dendritic voltage attenuation, the impact of individual synapses on granule cell output is low. At the same time, integration is linearized by voltage-dependent boosting mechanisms, only weakly affected by input synchrony, and independent of input location. These experiments establish that dentate granule cell dendritic properties are optimized for linear integration and strong attenuation of synaptic input from the entorhinal cortex, which may contribute to the sparse activity of granule cells in vivo.     

mRNA distribution within dendrites: Relationship to afferent innervation

The majority of neuronal mRNAs are confined to cell bodies, but a few mRNAs are present at high levels in dendrites. Here we report an initial analysis of the relationship between afferent innervation and the distribution of mRNA within dendritic fields. In situ hybridization techniques were used to compare the subcellular distribution of dendritic mRNAs in principal neurons of the hippocampal formation in vivo. The mRNA encoding the α subunit of calcium/calmodulin dependent protein kinase II (CAMII kinase) was present at high levels throughout the layers that contain the dendrites of hippocampal pyramidal cells and dentate granule cells. In contrast, the mRNA encoding the high molecular weight microtubule-associated protein MAP2 had a more limited distribution. In the dentate gyrus, labeling for MAP2 was present in a discrete band in the lamina containing proximal dendrites and decreased to low levels in laminae containing distal dendrites. This laminar pattern resembles the distinct terminations of the commissural/associational projection (high MAP2 labeling) and the entorhinal projection (lower MAP2 labeling) upon dendrites of granule cells. To determine if the differential distribution of dendritic mRNAs was regulated by either the presence or activity of afferents, we evaluated mRNA distribution in the dentate molecular layer following (1) removal of the entorhinal input by lesions of the entorhinal cortex or (2) prolonged delivery of potentiating stimulation to entorhinal afferents. Denervation led to modest decreases in the levels of mRNAs for both CAMII and MAP2 but did not lead to detectable alterations in mRNA distribution. Also, prolonged stimulation did not lead to detectable alterations in MAP2 or CAMII mRNA distribution, although such stimulation clearly elevated the expression of mRNA for glial fibrillary acidic protein (GFAP). © 1995 John Wiley & Sons, Inc.     

Cranial irradiation alters dendritic spine density and morphology in the hippocampus

Therapeutic irradiation of the brain is a common treatment modality for brain tumors, but can lead to impairment of cognitive function. Dendritic spines are sites of excitatory synaptic transmission and changes in spine structure and number are thought to represent a morphological correlate of altered brain functions associated with hippocampal dependent learning and memory. To gain some insight into the temporal and sub region specific cellular changes in the hippocampus following brain irradiation, we investigated the effects of 10 Gy cranial irradiation on dendritic spines in young adult mice. One week or 1 month post irradiation, changes in spine density and morphology in dentate gyrus (DG) granule and CA1 pyramidal neurons were quantified using Golgi staining. Our results showed that in the DG, there were significant reductions in spine density at both 1 week (11.9%) and 1 month (26.9%) after irradiation. In contrast, in the basal dendrites of CA1 pyramidal neurons, irradiation resulted in a significant reduction (18.7%) in spine density only at 1 week post irradiation. Analysis of spine morphology showed that irradiation led to significant decreases in the proportion of mushroom spines at both time points in the DG as well as CA1 basal dendrites. The proportions of stubby spines were significantly increased in both the areas at 1 month post irradiation. Irradiation did not alter spine density in the CA1 apical dendrites, but there were significant changes in the proportion of thin and mushroom spines at both time points post irradiation. Although the mechanisms involved are not clear, these findings are the first to show that brain irradiation of young adult animals leads to alterations in dendritic spine density and morphology in the hippocampus in a time dependent and region specific manner.     

Activity-dependent regulation of h channel distribution in hippocampal CA1 pyramidal neurons

The hyperpolarization-activated cation current, I(h), plays an important role in regulating intrinsic neuronal excitability in the brain. In hippocampal pyramidal neurons, I(h) is mediated by h channels comprised primarily of the hyperpolarization-activated cyclic nucleotide-gated (HCN) channel subunits, HCN1 and HCN2. Pyramidal neuron h channels within hippocampal area CA1 are remarkably enriched in distal apical dendrites, and this unique distribution pattern is critical for regulating dendritic excitability. We utilized biochemical and immunohistochemical approaches in organotypic slice cultures to explore factors that control h channel localization in dendrites. We found that distal dendritic enrichment of HCN1 is first detectable at postnatal day 13, reaching maximal enrichment by the 3rd postnatal week. Interestingly we found that an intact entorhinal cortex, which projects to distal dendrites of CA1 but not area CA3, is critical for the establishment and maintenance of distal dendritic enrichment of HCN1. Moreover blockade of excitatory neurotransmission using tetrodotoxin, 6-cyano-7-nitroquinoxaline-2,3-dione, or 2-aminophosphonovalerate redistributed HCN1 evenly throughout the dendrite without significant changes in protein expression levels. Inhibition of calcium/calmodulin-dependent protein kinase II activity, but not p38 MAPK, also redistributed HCN1 in CA1 pyramidal neurons. We conclude that activation of ionotropic glutamate receptors by excitatory temporoammonic pathway projections from the entorhinal cortex establishes and maintains the distribution pattern of HCN1 in CA1 pyramidal neuron dendrites by activating calcium/calmodulin-dependent protein kinase II-mediated downstream signals.     

Homeostatic changes of short-term plasticity of gabaergic synaptic transmission in rat hippocampal cell cultures

It is well documented that prolonged alteration of activity in neuronal networks initiates a number of homeostatic mechanisms including compensatory changes of excitatory and inhibitory synaptic strength. We studied whether this also evokes compensatory changes of short-term synaptic transmission. Using patch-clamp technique in hippocampal cell cultures we examined the effects: of prolonged decrease of neuronal firing evoked by sodium channel blocker: tetrodotoxin (TTX) and ionotropic glutamate receptor antagonist - kynurenate; prolonged enhancement ofneuronal firing evoked by antagonist GABAA receptors - bicuculline on short-term depression of GABAergic synaptic transmission evoked by train of stimuli (5 Hz). We found that both TTX and kynurenate treatments enhance depression of GABAergic transmission, while bicuculline treatment does not. We conclude that alteration of depression of GABAergic transmission evoked by the prolonged decrease of neuronal activity may contribute to homeostatic plasticity in hippocampal neuronal networks.     

Viral interference with neuronal integrity: what can we learn from the Borna disease virus?

The neurotropic Borna disease virus (BDV) is unusual in that it can persistently infect neurons of the central nervous system (CNS) without causing general cell death, reflecting its favourable adaptation to the brain. The activity-dependent enhancement of neuronal network activity is however disturbed after BDV infection, possibly by its effect on the protein kinase C signalling pathway. The best model for studying BDV, which has a non-cytolytic replication strategy in primary neurons, is the rat. Infection of adult rats results in a fatal immune-mediated disease, whereas BDV establishes persistent infection of the brain in newborn rats resulting in progressive neuronal cell loss in defined regions of the CNS. Our recently developed system of BDV-infected hippocampal slice cultures has clearly shown that the onset of granule cell loss begins after the formation of the mossy fibre projection. Quantitative analysis has revealed a significant increase in synaptic density on identified remaining granule cell dendrites at 6?weeks after infection, followed by a decline. Granule cells are the major target of entorhinal afferents. However, despite an almost complete loss of dentate granule cells during BDV infection, entorhinal axons persist in their correct layer, both in vivo and in slice cultures, possibly exploiting rewiring capabilities and thereby allowing new synapse formation with available targets. These morphological observations, together with electrophysiological and biochemical data, indicate that BDV is a suitable model virus for studying virus-induced morphological and functional changes of neurons and connectivity patterns.     

Environment matters: synaptic properties of neurons born in the epileptic adult brain develop to reduce excitability

Neural progenitors in the adult dentate gyrus continuously produce new functional granule cells. Here we used whole-cell patch-clamp recordings to explore whether a pathological environment influences synaptic properties of new granule cells labeled with a GFP-retroviral vector. Rats were exposed to a physiological stimulus, i.e., running, or a brain insult, i.e., status epilepticus, which gave rise to neuronal death, inflammation, and chronic seizures. Granule cells formed after these stimuli exhibited similar intrinsic membrane properties. However, the new neurons born into the pathological environment differed with respect to synaptic drive and short-term plasticity of both excitatory and inhibitory afferents. The new granule cells formed in the epileptic brain exhibited functional connectivity consistent with reduced excitability. We demonstrate a high degree of plasticity in synaptic inputs to adult-born new neurons, which could act to mitigate pathological brain function.     

Fibroblasts that proliferate near denervated synaptic sites in skeletal muscle synthesize the adhesive molecules tenascin(J1), N-CAM, fibronectin, and a heparan sulfate proteoglycan

Four adhesive molecules, tenascin(J1), N-CAM, fibronectin, and a heparan sulfate proteoglycan, accumulate in interstitial spaces near synaptic sites after denervation of rat skeletal muscle (Sanes, J. R., M. Schachner, and J. Covault. 1986. J. Cell Biol. 102:420-431). We have now asked which cells synthesize these molecules, and how this synthesis is regulated. Electron microscopy revealed that mononucleated cells selectively accumulate in perisynaptic interstitial spaces beginning 2 d after denervation. These cells were identified as fibroblasts by ultrastructural and immunohistochemical criteria; [3H]thymidine autoradiography revealed that their accumulation results from local proliferation. Electron microscopic immunohistochemistry demonstrated that N-CAM is associated with the surface of the fibroblasts, while tenascin(J1) is associated with collagen fibers that abut fibroblasts. Using immunofluorescence and immunoprecipitation methods, we found that fibroblasts isolated from perisynaptic regions of denervated muscle synthesize N-CAM, tenascin(J1), fibronectin, and a heparan sulfate proteoglycan in vitro. Thus, fibroblasts that selectively proliferate in interstitial spaces near synaptic sites are likely to be the cellular source of the interstitial deposits of adhesive molecules in denervated muscle. To elucidate factors that might regulate the accumulation of these molecules in vivo, we analyzed the expression of tenascin(J1) and fibronectin by cultured fibroblasts. Fibroblasts from synapse-free regions of denervated muscle, as well as skin, lung, and 3T3 fibroblasts accumulate high levels of tenascin(J1) and fibronectin in culture, showing that perisynaptic fibroblasts are not unique in this regard. However, when they are first placed in culture, fibroblasts from denervated muscle bear more tenascin(J1) than fibroblasts from innervated muscle, indicating that expression of this molecule by fibroblasts is regulated by the muscle's state of innervation; this difference is no longer apparent after a few days in culture. In 3T3 cells, accumulation of tenascin(J1) is high in proliferating cultures, depressed in confluent cultures, and reactivated in cells stimulated to proliferate by replating at low density or by wounding a confluent monolayer. Thus, synthesis of tenascin(J1) is regulated in parallel with mitotic activity. In contrast, levels of fibronectin, which increase less dramatically after denervation in vivo, are similar in fibroblasts from innervated and denervated muscle and in proliferating and quiescent 3T3 cells.(ABSTRACT TRUNCATED AT 400 WORDS)     

Glutamate producing aspartate aminotransferase in glutamatergic perforant path terminals of the rat hippocampus

Summary The enzyme aspartate aminotransferase was demonstrated cytochemically in the rat hippocampus 4, 7, and 14 days after unilateral entorhinal cortex lesion. At the light microscopic level the enzyme showed a significant activity decrease in the ipsilateral entorhinal terminal field which was similar at all postlesion times investigated. Non-denervated areas, i.e. the inner one-third of the dentate gyrus molecular layer and the radiatum layer of CA2/3, showed an increase of aminotransferase activities. At the electron microscopic level in the entorhinal terminal field of the control (unoperated) side aspartate aminotransferase was localized preferentially in a great number of boutons, containing the cytoplasmic and mitochondrial isoenzymes. Following entorhinal lesion a significant loss of these positively reacting boutons was seen. Most of the degenerating boutons contained reaction product but a small number was negative for aspartate aminotransferase. From 4 to 14 postlesion days the positively reacting boutons of the non-denervated supragranular zone expanded outward into the denervated area according to the known terminal proliferation of the commissural and associational systems. The remaining denervated entorhinal terminal field was reinnervated predominantly by negatively reacting boutons (probably terminal proliferations of septal afferents) and by a small number of positively reacting boutons (probably terminal proliferations of the crossed temporodentate pathway). The presence of cytoplasmic aspartate aminotransferase in the terminals of a well-known glutamatergic system is discussed in relation to the possible importance of this enzyme for the production of releasable glutamate.     

Hyaluronan-based extracellular matrix under conditions of homeostatic plasticity

Neuronal networks are balanced by mechanisms of homeostatic plasticity, which adjusts synaptic strength via molecular and morphological changes in the pre- and post-synapse. Here, we wondered whether the hyaluronic acid-based extracellular matrix (ECM) of the brain is involved in mechanisms of homeostatic plasticity. We hypothesized that the ECM, being rich in chondroitin sulfate proteoglycans such as brevican, which are suggested to stabilize synapses by their inhibitory effect on structural plasticity, must be remodelled to allow for structural and molecular changes during conditions of homeostatic plasticity. We found a high abundance of cleaved brevican fragments throughout the hippocampus and cortex and in neuronal cultures, with the strongest labelling in perineuronal nets on parvalbumin-positive interneurons. Using an antibody specific for a brevican fragment cleaved by the matrix metalloprotease ADAMTS4, we identified the enzyme as the main brevican-processing protease. Interestingly, we found ADAMTS4 largely associated with synapses. After inducing homeostatic plasticity in neuronal cell cultures by prolonged network inactivation, we found increased brevican processing at inhibitory as well as excitatory synapses, which is in line with the ADAMTS4 subcellular localization. Thus, the ECM is remodelled in conditions of homeostatic plasticity, which may liberate synapses to allow for a higher degree of structural plasticity.     

Glutamate producing aspartate aminotransferase in glutamatergic perforant path terminals of the rat hippocampus. Cytochemical and lesion studies

The enzyme aspartate aminotransferase was demonstrated cytochemically in the rat hippocampus 4, 7, and 14 days after unilateral entorhinal cortex lesion. At the light microscopic level the enzyme showed a significant activity decrease in the ipsilateral entorhinal terminal field which was similar at all postlesion times investigated. Non-denervated areas, i.e. the inner one-third of the dentate gyrus molecular layer and the radiatum layer of CA2/3, showed an increase of aminotransferase activities. At the electron microscopic level in the entorhinal terminal field of the control (unoperated) side aspartate aminotransferase was localized preferentially in a great number of boutons, containing the cytoplasmic and mitochondrial isoenzymes. Following entorhinal lesion a significant loss of these positively reacting boutons was seen. Most of the degenerating boutons contained reaction product but a small number was negative for aspartate aminotransferase. From 4 to 14 postlesion days the positively reacting boutons of the non-denervated supragranular zone expanded outward into the denervated area according to the known terminal proliferation of the commissural and associational systems. The remaining denervated entorhinal terminal field was reinnervated predominantly by negatively reacting boutons (probably terminal proliferations of septal afferents) and by a small number of positively reacting boutons (probably terminal proliferations of the crossed temporo-dentate pathway). The presence of cytoplasmic aspartate aminotransferase in the terminals of a well-known glutamatergic system is discussed in relation to the possible importance of this enzyme for the production of releasable glutamate.     

Local presynaptic activity gates homeostatic changes in presynaptic function driven by dendritic BDNF synthesis

Homeostatic synaptic plasticity is important for maintaining stability of neuronal function, but heterogeneous expression mechanisms suggest that distinct facets of neuronal activity may shape the manner in which compensatory synaptic changes are implemented. Here, we demonstrate that local presynaptic activity gates a retrograde form of homeostatic plasticity induced by blockade of AMPA receptors (AMPARs) in cultured hippocampal neurons. We show that AMPAR blockade produces rapid (<3 hr) protein synthesis-dependent increases in both presynaptic and postsynaptic function and that the induction of presynaptic, but not postsynaptic, changes requires coincident local activity in presynaptic terminals. This "state-dependent" modulation of presynaptic function requires postsynaptic release of brain-derived neurotrophic factor (BDNF) as a retrograde messenger, which is locally synthesized in dendrites in response to AMPAR blockade. Taken together, our results reveal a local crosstalk between active presynaptic terminals and postsynaptic signaling that dictates the manner by which homeostatic plasticity is implemented at synapses.     

Changes of Dendritic Spine Density and Morphology in the Superficial Layers of the Medial Entorhinal Cortex Induced by Extremely Low-Frequency Magnetic Field Exposure

In the present study, we investigated the effects of chronic exposure (14 and 28 days) to a 0.5 mT 50 Hz extremely low-frequency magnetic field (ELM) on the dendritic spine density and shape in the superficial layers of the medial entorhinal cortex (MEC). We performed Golgi staining to reveal the dendritic spines of the principal neurons in rats. The results showed that ELM exposure induced a decrease in the spine density in the dendrites of stellate neurons and the basal dendrites of pyramidal neurons at both 14 days and 28 days, which was largely due to the loss of the thin and branched spines. The alteration in the density of mushroom and stubby spines post ELM exposure was cell-type specific. For the stellate neurons, ELM exposure slightly increased the density of stubby spines at 28 days, while it did not affect the density of mushroom spines at the same time. In the basal dendrites of pyramidal neurons, we observed a significant decrease in the mushroom spine density only at the later time point post ELM exposure, while the stubby spine density was reduced at 14 days and partially restored at 28 days post ELM exposure. ELM exposure-induced reduction in the spine density in the apical dendrites of pyramidal neurons was only observed at 28 days, reflecting the distinct vulnerability of spines in the apical and basal dendrites. Considering the changes in spine number and shape are involved in synaptic plasticity and the MEC is a part of neural network that is closely related to learning and memory, these findings may be helpful for explaining the ELM exposure-induced impairment in cognitive functions.     

Immunocytochemistry of glycine in small neurons of the granule cell areas of the guinea pig dorsal cochlear nucleus: a post-embedding ultrastructural study

The axon terminals of the acoustic nerve contact different part of the cochlear nucleus including granule cell areas. Little is known of the cell composition and neural circuits of granule cell areas present in the fusiform and upper polymorphic layers of the dorsal cochlear nucleus in the guinea pig. The present ultrastructural immunocytochemical study exploits the technique of post-embedding immunogold and silver intensification to reveal the characteristics of small neurons in granule cell areas. Few neurons (Golgi-stellate cells) use glycine as inhibitory neurotransmitter which is present in symmetric synaptic boutons with pleomorphic and flat vesicles. In contrast, most neurons (granule and unipolar brush cells) are not glycine-positive, and presumably not excitatory. Most of the large axons (mossy fibres) in granule areas are probably excitatory (glycine-negative and storing round synaptic vesicles) and contact unipolar brush cells forming large synapses or granule cell dendrites by small synapses. A few large glycinergic boutons (inhibitory) also contact unipolar brush cells. The excitatory circuit of mossy fibre-unipolar brush and granule cells may be inhibited by the glycinergic terminals from the few glycinergic cells (Golgi-stellate neurons) present within the granule cell areas. The latter are not contacted by large mossy-like glycine terminals.     

Development of cell type-specific connectivity patterns of converging excitatory axons in the retina

To integrate information from different presynaptic cell types, dendrites receive distinct patterns of synapses from converging axons. How different afferents in?vivo establish specific connectivity patterns with the same dendrite is poorly understood. Here, we examine the synaptic development of three glutamatergic bipolar cell types converging onto?a common postsynaptic retinal ganglion cell. We find that after axons and dendrites target appropriate synaptic layers, patterns of connections among these neurons?diverge through selective changes in the conversion of axo-dendritic appositions to synapses. This process is differentially regulated by neurotransmission, which is required for the shift from single to multisynaptic appositions of one bipolar cell type but not for maintenance and elimination, respectively, of connections from the other two types. Thus, synaptic specificity among converging excitatory inputs in the?retina emerges via differential synaptic maturation of axo-dendritic appositions and is shaped by neurotransmission in a cell type-dependent manner.     

Terminal Schwann cells guide the reinnervation of muscle after nerve injury

Schwann cells and axons labeled by transgene-encoded, fluorescent proteins can be repeatedly imaged in living mice to observe the reinnervation of neuromuscular junctions. Axons typically return to denervated junctions by growing along Schwann cells contained in the old nerve sheaths or “Schwann cell tubes”. These axons then commonly “escape” the synaptic sites by growing along the Schwann cell processes extended during the period of denervation. These “escaped fibers” grow to innervate adjacent synaptic sites along Schwann cells bridging these sites. Within the synaptic site, Schwann cells, originally positioned above the synaptic site continue to cover the acetylcholine receptors (AChRs) immediately following denervation, but gradually vacate portions of this site. When regenerating axons return, they first deploy along the Schwann cells and ignore sites of AChRs vacated by Schwann cells. In many cases these vacated sites are never reinnervated and are ultimately lost. Following partial denervation, Schwann cells grow in an apparently tropic fashion from denervated to nearby innervated synaptic sites and serve as the substrates for nerve sprouting. These experiments show that Schwann cells provide pathways that stimulate axon growth and insure the rapid reinnervation of denervated or partially denervated muscles.