Patterns of expression of brain-derived neurotrophic factor and tyrosine kinase B mRNAs and distribution and ultrastructural localization of their proteins in the visual pathway of the adult rat.

We have examined the cellular and subcellular distribution and the patterns of expression of brain-derived neurotrophic factor (BDNF), and of its high affinity receptor, tyrosine kinase B (TrkB), in retinorecipient regions of the brain, including the superior colliculus, the lateral geniculate nucleus and the olivary pretectal nucleus. In the retinorecipient layers of the superior colliculus, BDNF protein and mRNA were present in the cell bodies of a subpopulation of neurons, and BDNF protein was present in the neuropil as punctate or fiber-like structures. In the lateral geniculate nucleus, however, BDNF mRNA was not detected, and BDNF protein was restricted to punctate and fiber-like structures in the neuropil, especially in the most superficial part of the dorsal lateral geniculate nucleus, just below the optic tract. At the ultrastructural level, BDNF protein was localized predominantly to axon terminals containing round synaptic vesicles and pale mitochondria with irregular cristae, which made asymmetric (Gray type I) synaptic specializations (R-boutons). Enucleation of one eye was followed by loss of BDNF immunoreactivity and disappearance of BDNF-positive R-boutons in the contralateral visual centers, confirming the retinal origin of at least most of these terminals. TrkB was present in postsynaptic densities apposed to immunoreactive R-boutons in the superior colliculus and lateral geniculate nucleus, and was also associated with axonal and dendritic microtubules. These findings suggest that BDNF is synthesized by a subpopulation of retinal ganglion cells and axonally transported to visual centers where this neurotrophin is assumed to play important roles in visual system maintenance and/or in modulating the excitatory retinal input to neurons in these centers.

Effects of lipopolysaccharide-induced inflammation on expression of growth-associated genes by corticospinal neurons.

BACKGROUND: Inflammation around cell bodies of primary sensory neurons and retinal ganglion cells enhances expression of neuronal growth-associated genes and stimulates axonal regeneration. We have asked if inflammation would have similar effects on corticospinal neurons, which normally show little response to spinal cord injury. Lipopolysaccharide (LPS) was applied onto the pial surface of the motor cortex of adult rats with or without concomitant injury of the corticospinal tract at C4. Inflammation around corticospinal tract cell bodies in the motor cortex was assessed by immunohistochemistry for OX42 (a microglia and macrophage marker). Expression of growth-associated genes c-jun, ATF3, SCG10 and GAP-43 was investigated by immunohistochemistry or in situ hybridisation. RESULTS: Application of LPS induced a gradient of inflammation through the full depth of the motor cortex and promoted c-Jun and SCG10 expression for up to 2 weeks, and GAP-43 upregulation for 3 days by many corticospinal neurons, but had very limited effects on neuronal ATF3 expression. However, many glial cells in the subcortical white matter upregulated ATF3. LPS did not promote sprouting of anterogradely labelled corticospinal axons, which did not grow into or beyond a cervical lesion site. CONCLUSION: Inflammation produced by topical application of LPS promoted increased expression of some growth-associated genes in the cell bodies of corticospinal neurons, but was insufficient to promote regeneration of the corticospinal tract.

Upregulation of activating transcription factor 3 (ATF3) by intrinsic CNS neurons regenerating axons into peripheral nerve grafts.

The expression of the transcription factor ATF3 in the brain was examined by immunohistochemistry during axonal regeneration induced by the implantation of pieces of peripheral nerve into the thalamus of adult rats. After 3 days, ATF3 immunoreactivity was present in many cells within approximately 500 mum of the graft. In addition, ATF3-positive cell nuclei were found in the thalamic reticular nucleus (TRN) and medial geniculate nuclear complex (MGN), from which most regenerating axons originate. CNS cells with ATF3-positive nuclei were predominantly neurons and did not show signs of apoptosis. The number of ATF3-positive cells had declined by 7 days and further by 1 month after grafting when most ATF3-positive cells were found in the TRN and MGN. 14 days or more after grafting, some ATF3-positive nuclei were distorted and may have been apoptotic. In some experiments of 1 month duration, neurons which had regenerated axons to the distal ends of grafts were retrogradely labeled with DiAsp. ATF3-positive neurons in these animals were located in regions of the TRN and MGN containing retrogradely labeled neurons and the great majority were also labeled with DiAsp. SCG10 and c-Jun were found in neurons in the same regions as retrogradely labeled and ATF3-positive cells. Thus, ATF3 is transiently upregulated by injured CNS neurons, but prolonged expression is part of the pattern of gene expression associated with axonal regeneration. The co-expression of ATF3 with c-jun suggests that interactions between these transcription factors may be important for controlling the program of gene expression necessary for regeneration.

ATF3 upregulation in glia during Wallerian degeneration: differential expression in peripheral nerves and CNS white matter.

BACKGROUND: Many changes in gene expression occur in distal stumps of injured nerves but the transcriptional control of these events is poorly understood. We have examined the expression of the transcription factors ATF3 and c-Jun by non-neuronal cells during Wallerian degeneration following injury to sciatic nerves, dorsal roots and optic nerves of rats and mice, using immunohistochemistry and in situ hybridization. RESULTS: Following sciatic nerve injury--transection or transection and reanastomosis--ATF3 was strongly upregulated by endoneurial, but not perineurial cells, of the distal stumps of the nerves by 1 day post operation (dpo) and remained strongly expressed in the endoneurium at 30 dpo when axonal regeneration was prevented. Most ATF3+ cells were immunoreactive for the Schwann cell marker, S100. When the nerve was transected and reanastomosed, allowing regeneration of axons, most ATF3 expression had been downregulated by 30 dpo. ATF3 expression was weaker in the proximal stumps of the injured nerves than in the distal stumps and present in fewer cells at all times after injury. ATF3 was upregulated by endoneurial cells in the distal stumps of injured neonatal rat sciatic nerves, but more weakly than in adult animals. ATF3 expression in transected sciatic nerves of mice was similar to that in rats. Following dorsal root injury in adult rats, ATF3 was upregulated in the part of the root between the lesion and the spinal cord (containing Schwann cells), beginning at 1 dpo, but not in the dorsal root entry zone or in the degenerating dorsal column of the spinal cord. Following optic nerve crush in adult rats, ATF3 was found in some cells at the injury site and small numbers of cells within the optic nerve displayed weak immunoreactivity. The pattern of expression of c-Jun in all types of nerve injury was similar to that of ATF3. CONCLUSION: These findings raise the possibility that ATF3/c-Jun heterodimers may play a role in regulating changes in gene expression necessary for preparing the distal segments of injured peripheral nerves for axonal regeneration. The absence of the ATF3 and c-Jun from CNS glia during Wallerian degeneration may limit their ability to support regeneration.

NG2 proteoglycan expression in the peripheral nervous system: upregulation following injury and comparison with CNS lesions.

The chondroitin sulphate proteoglycan NG2 blocks neurite outgrowth in vitro and thus may be able to inhibit axonal regeneration in the CNS. We have used immunohistochemistry to compare the expression of NG2 in the PNS, where axons regenerate, and the spinal cord, where regeneration fails. NG2 is expressed by satellite cells in dorsal root ganglia (DRG) and in the perineurium and endoneurium of intact sciatic nerves of adult rats. Endoneurial NG2-positive cells were S100-negative. Injury to dorsal roots, ventral rami or sciatic nerves had no effect on NG2 expression in DRG but sciatic nerve section or crush caused an upregulation of NG2 in the damaged nerve. Strongly NG2-positive cells in damaged nerves were S100-negative. The proximal stump of severed nerves was capped by dense NG2, which surrounded bundles of regenerating axons. The distal stump, into which axons regenerated, also contained many NG2-positive/S100-negative cells. Immunoelectron microscopy revealed that most NG2-positive cells in distal stumps had perineurial or fibroblast-like morphologies, with NG2 being concentrated at the poles of the cells in regions exhibiting microvillus-like protrusions or caveolae. Compression and partial transection injuries to the spinal cord also caused an upregulation of NG2, and NG2-positive cells and processes invaded the lesion sites. Transganglionically labelled ascending dorsal column fibres, stimulated to sprout by a conditioning sciatic nerve injury, ended in the borders of lesions among many NG2-positive processes. Thus, NG2 upregulation is a feature of the response to injury in peripheral nerves and in the spinal cord, but it does not appear to limit regeneration in the sciatic nerve.

Efficient delivery of Cre-recombinase to neurons in vivo and stable transduction of neurons using adeno-associated and lentiviral vectors.

BACKGROUND: Inactivating genes in vivo is an important technique for establishing their function in the adult nervous system. Unfortunately, conventional knockout mice may suffer from several limitations including embryonic or perinatal lethality and the compensatory regulation of other genes. One approach to producing conditional activation or inactivation of genes involves the use of Cre recombinase to remove loxP-flanked segments of DNA. We have studied the effects of delivering Cre to the hippocampus and neocortex of adult mice by injecting replication-deficient adeno-associated virus (AAV) and lentiviral (LV) vectors into discrete regions of the forebrain. RESULTS: Recombinant AAV-Cre, AAV-GFP (green fluorescent protein) and LV-Cre-EGFP (enhanced GFP) were made with the transgene controlled by the cytomegalovirus promoter. Infecting 293T cells in vitro with AAV-Cre and LV-Cre-EGFP resulted in transduction of most cells as shown by GFP fluorescence and Cre immunoreactivity. Injections of submicrolitre quantities of LV-Cre-EGFP and mixtures of AAV-Cre with AAV-GFP into the neocortex and hippocampus of adult Rosa26 reporter mice resulted in strong Cre and GFP expression in the dentate gyrus and moderate to strong labelling in specific regions of the hippocampus and in the neocortex, mainly in neurons. The pattern of expression of Cre and GFP obtained with AAV and LV vectors was very similar. X-gal staining showed that Cre-mediated recombination had occurred in neurons in the same regions of the brain, starting at 3 days post-injection. No obvious toxic effects of Cre expression were detected even after four weeks post-injection. CONCLUSION: AAV and LV vectors are capable of delivering Cre to neurons in discrete regions of the adult mouse brain and producing recombination.

Corticospinal neurons up-regulate a range of growth-associated genes following intracortical, but not spinal, axotomy.

The failure of some CNS neurons to up-regulate growth-associated genes following axotomy may contribute to their failure to regenerate axons. We have studied gene expression in rat corticospinal neurons following either proximal (intracortical) or distal (spinal) axotomy. Corticospinal neurons were retrogradely labelled with cholera toxin subunit B prior to intracortical lesions or concomitantly with spinal lesions. Alternate sections of forebrain were immunoreacted for cholera toxin subunit B or processed for mRNA in situ hybridization for ATF3, c-jun, GAP-43, CAP-23, SCG10, L1, CHL1 or krox-24, each of which has been associated with axotomy or axon regeneration in other neurons. Seven days after intracortical axotomy, ATF3, c-jun, GAP-43, SCG10, L1 and CHL1, but not CAP-23 or krox-24, were up-regulated by layer V pyramidal neurons, including identified corticospinal neurons. The maximum distance between the lesion and the neuronal cell bodies that up-regulated genes varied between 300 and 500 microm. However, distal axotomy failed to elicit changes in gene expression in corticospinal neurons. No change in expression of any molecule was seen in the neocortex 1 or 7 days after corticospinal axotomy in the cervical spinal cord. The expression of GAP-43, CAP-23, L1, CHL1 and SCG10 was confirmed to be unaltered after this type of injury in identified retrogradely labelled corticospinal neurons. Thus, while corticospinal neuronal cell bodies fail to respond to spinal axotomy, these cells behave like regeneration-competent neurons, up-regulating a wide range of growth-associated molecules if axotomized within the cerebral cortex.

FKBP12 mRNA expression is upregulated by intrinsic CNS neurons regenerating axons into peripheral nerve grafts in the brain.

We have examined the expression of the immunophilin FKBP12 in adult rat intrinsic CNS neurons stimulated to regenerate axons by the implantation of segments of autologous tibial nerve into the thalamus or cerebellum. After survival times of 3 days to 6 weeks, the brains were fresh-frozen. In some animals the regenerating neurons were retrogradely labelled with cholera toxin subunit B 1 day before they were killed. Sections through the thalamus or cerebellum were used for in situ hybridization with digoxygenin-labelled riboprobes for FKBP12 or immunohistochemistry to detect cholera toxin subunit B-labelled neurons. FKBP12 was constitutively expressed by many neurons, and was very strongly expressed in the hippocampus and by Purkinje cells. Regenerating neurons were found in the thalamic reticular nucleus and deep cerebellar nuclei of animals that received living grafts. Neurons in these nuclei upregulated FKBP12 mRNA; such neurons were most numerous at 3 days post grafting but were most strongly labelled at 2 weeks post grafting. Regenerating neurons identified by retrograde labelling were found to have upregulated FKBP12 mRNA. No upregulation was seen in neurons in animals that received freeze-killed grafts, which do not support axonal regeneration. We conclude that FKBP12 is a regeneration-associated gene in intrinsic CNS neurons.

Expression of regeneration-related molecules in injured and regenerating striatal and nigral neurons.

Peripheral nerve grafts in the neostriatum promote axonal regeneration from restricted classes of CNS neuron, principally cells in the substantia nigra pars compacta (SNpc) and striatal cholinergic interneurons. We have examined the molecular responses of CNS neurons induced to regenerate axons by tibial nerve grafting to the neostriatum of adult rats. Brain sections were probed for mRNAs for the transcription factor c-jun, and the cell recognition molecule CHL1, or immunoreacted for TrkA or p75, 1 day to 29 weeks after grafting (dpo; wpo). In unoperated rats, scattered neurons throughout the neostriatum showed weak signals for CHL1 mRNA and slightly stronger signals for c-jun mRNA. Cells of similar appearance strongly expressed TrkA but possessed little p75. By 1 dpo, many neostriatal neurons of various sizes and GFAP + glial cells near the host/graft interface had upregulated CHL1 mRNA, c-jun mRNA and p75. Most of the larger (20-25 microm diameter) CHL1 mRNA+ cells were also TrkA+, indicating that they were NGF-sensitive cholinergic interneurons. From two weeks postgrafting, high levels of CHL1 and c-jun mRNAs and p75 in the neostriatum were confined to a few presumptive cholinergic interneurons; p75+ cells were also TrkA+ and were larger than TrkA+ neurons on the contralateral side. Retrograde labelling showed that most p75+ and some TrkA+ neurons regenerated axons through the graft. Neurons in the SNpc showed a moderate to strong signal for CHL1 mRNA, weaker signal for c-jun mRNA, and no p75 or TrkA. Some SNpc cells upregulated c-jun mRNA after graft implantation, although they did not upregulate CHL1 mRNA, p75 or TrkA. Since neostriatal neurons which regenerate axons into grafts express receptors for NGF, and grafts mimic the effects of NGF treatment on these cells, sensitivity to graft-derived NGF may be a determinant of their high regenerative capacity. The finding that c-jun and CHL1 are consistently expressed by CNS neurons induced to regenerate their axons strongly supports the idea that these molecules are directly involved in axonal regeneration.

Transcriptional upregulation of SCG10 and CAP-23 is correlated with regeneration of the axons of peripheral and central neurons in vivo.

We have compared SCG10 and CAP-23 expression with that of GAP-43 during axonal regeneration in the peripheral and central nervous systems (PNS, CNS) of adult rats. SCG10, CAP-23, and GAP-43 mRNAs were strongly upregulated by motor and dorsal root ganglion (DRG) neurons following sciatic nerve crush, but not after dorsal rhizotomy. When the sciatic nerve was cut and ligated to prevent reinnervation of targets, expression of all three mRNAs was prolonged. Neurons in the thalamic reticular nucleus and deep cerebellar nuclei transiently upregulated these mRNAs after axotomy, and showed prolonged upregulation of all three molecules when regenerating axons into peripheral nerve grafts inserted into the thalamus of cerebellum. Neurons in the dorsal thalamus and cerebellar cortex showed poor regenerative capacity and most did not upregulate any of these mRNAs. Thus, in both PNS and CNS neurons, the transcription of SCG10, CAP-23, and GAP-43 mRNAs is coregulated following axotomy and during regeneration. Signals from living peripheral nerve appear to maintain expression of all three mRNAs in regenerating neurons, and in PNS neurons downregulation correlates with target reinnervation. Thus, SCG10 and CAP-23, as well as GAP-43, are likely to be important neuronal determinants of regenerative ability.

Correlation between putative inhibitory molecules at the dorsal root entry zone and failure of dorsal root axonal regeneration.

The molecular mechanisms involved in preventing regenerating dorsal root axons from entering the spinal cord at the dorsal root entry zone (DREZ) are obscure. We used immunohistochemistry, in situ hybridization, and electron microscopy to study axonal regeneration after dorsal rhizotomy in adult rats and its relationship to cellular changes and the distribution of putative growth inhibitory molecules in this region. Astrocyte processes, ending as bulb-shaped expansions, grew up to 700 microm into the basal lamina tubes of injured roots, where regenerating axons were also present. Some of these axons approached or reached the DREZ but grew no further; others turned back toward the ganglion, suggesting the presence of repulsive cues in or near the DREZ. Tenascin-C mRNA and protein and CSPG stub immunoreactivity were strongly upregulated in the roots after rhizotomy, but were only weakly expressed in the DREZ. Tenascin-R immunoreactivity was confined to CNS tissue, and unaffected by rhizotomy. Large, rounded GFAP-negative, NG2-immunoreactive cells, a few of which were OX42 positive, were found in the DREZ following rhizotomy. Astrocyte processes projecting into the roots were tenascin-R and NG2 negative. Hence, only NG2-expressing cells and tenascin-R were appropriately situated to inhibit regeneration through the DREZ.

Glutamatergic components of the retrosplenial granular cortex in the rat.

The ultrastructural characteristics, distribution and synaptic relationships of identified, glutamate-enriched thalamocortical axon terminals and cell bodies in the retrosplenial granular cortex of adult rats is described and compared with GABA-containing terminals and cell bodies, using postembedding immunogold immunohistochemistry and transmission electron microscopy in animals with injections of cholera toxin- horseradish peroxidase (CT-HRP) into the anterior thalamic nuclei. Anterogradely labelled terminals, identified by semi-crystalline deposits of HRP reaction product, were approximately 1 microm in diameter, contained round, clear synaptic vesicles, and established asymmetric (Gray type I) synaptic contacts with dendritic spines and small dendrites, some containing HRP reaction product, identifying them as dendrites of corticothalamic projection neurons. The highest densities of immunogold particles following glutamate immunostaining were found over such axon terminals and over similar axon terminals devoid of HRP reaction product. In serial sections immunoreacted for GABA, these axon terminals were unlabelled, whereas other axon terminals, establishing symmetric (Gray type II) synapses were heavily labelled. Cell bodies of putative pyramidal neurons, containing retrograde HRP label, were numerous in layers V-VI; some were also present in layers I-III. Most were overlain by high densities of gold particles in glutamate but not in GABA immunoreacted sections. These findings provide evidence that the terminals of projection neurons make synaptic contact with dendrites and dendritic spines in the ipsilateral retrosplenial granular cortex and that their targets include the dendrites of presumptive glutamatergic corticothalamic projection neurons.

Axonal regeneration from CNS neurons in the cerebellum and brainstem of adult rats: correlation with the patterns of expression and distribution of messenger RNAs for L1, CHL1, c-jun and growth-associated protein-43.

Some neurons in the brain and spinal cord will regenerate axons into a living peripheral nerve graft inserted at the site of injury, others will not. We have examined the patterns of expression of four molecules thought to be involved in developmental and regenerative axonal growth, in the cerebellum and brainstem of adult rats, following the implantation into the cerebellum of peripheral nerve grafts. We also determined how the expression patterns observed correlate with the abilities of neurons in these regions to regenerate axons. Three days to 16 weeks after insertion of living tibial nerve autografts, neurons which had regenerated axons into the graft were retrogradely labelled from the distal extremity of the graft with cholera toxin conjugated to horseradish peroxidase, and sections through the cerebellum and brainstem were processed for visualization of transported tracer and/or hybridized with riboprobes to detect messenger RNAs for the cell recognition molecules L1 and CHL1 (close homologue of L1), growth-associated protein-43 and the cellular oncogene c-jun. Retrogradely labelled neurons were present in cerebellar deep nuclei close to the graft and in brainstem nuclei known to project to the cerebellum. Neurons in these same nuclei were found to have up-regulated expression of all four messenger RNAs. Individual retrogradely labelled neurons also expressed high levels of L1, CHL1, c-jun or growth-associated protein-43 messenger RNAs (and vice versa), and every messenger RNA investigated was co-localized with at least one other messenger RNA. Purkinje cells did not regenerate axons into the graft or up-regulate L1, CHL1 or growth-associated protein-43 messenger RNAs, but there was increased expression of c-jun messenger RNA in some Purkinje cells close to the graft. Freeze-killed grafts produced no retrograde labelling of neurons, and resulted in only transient and low levels of up-regulation of the tested molecules, mainly L1 and CHL1. These findings show that cerebellar deep nucleus neurons and precerebellar brainstem neurons, but not Purkinje cells, have a high propensity for axon regeneration, and that axonal regeneration by these neurons is accompanied by increased expression of L1, CHL1, c-jun and growth-associated protein-43. Furthermore, although the patterns of expression of the four molecules investigated are not identical in regenerating neuronal populations, it is probable that all four are up-regulated in all neurons whose axons regenerate into the grafts and that their up-regulation may be required for axon regeneration to occur. Finally, because c-jun up-regulation is seen in Purkinje cells close to the graft, unaccompanied by up-regulation of the other molecules investigated, c-jun up-regulation alone cannot be taken to reliably signify a regenerative response to axotomy.

The cell recognition molecule CHL1 is strongly upregulated by injured and regenerating thalamic neurons.

Close homologue of L1 (CHL1) is a cell recognition molecule known to promote axonal growth in vitro. We have investigated the expression of CHL1 mRNA by regenerating central nervous system (CNS) neurons, by using in situ hybridisation 3 days to 10 weeks following the implantation of living and freeze-killed peripheral nerve autografts into the thalamus of adult rats. At all survival times after implantation of living grafts, neurons of the thalamic reticular nucleus (TRN), close to the graft tip and up to 1 mm away from it, displayed strong signal for CHL1 mRNA, even though TRN neurons show very low levels of CHL1 mRNA expression in unoperated animals. When the cell bodies of regenerating neurons were identified by retrograde labelling from the distal portion of the grafts, 4-6 weeks after operation, most of the labelled cells were found in the TRN and could be shown to haveupregulated CHL1 mRNA. In addition, some neurons in dorsal thalamic nuclei near the graft tip transiently upregulated CHL1 mRNA during the first 3 weeks after graft implantation, and glial cells showing CHL1 mRNA expression were present at the brain/graft interface 3 days to 2 weeks after operation. Freeze-killed grafts, into which axons do not regenerate, caused a transient upregulation of CHL1 in very few TRN neurons near the graft tip and in glial cells at the brain/graft interface but did not produce prolonged CHL1 mRNA expression. CHL1 can therefore be added to the list of molecules (including GAP-43, L1, and c-jun) strongly expressed by CNS neurons that regenerate their axons into nerve grafts, but not by those neurons that fail to regenerate their axons.

Overexpression of GAP-43 in thalamic projection neurons of transgenic mice does not enable them to regenerate axons through peripheral nerve grafts.

It is well established that some populations of neurons of the adult rat central nervous system (CNS) will regenerate axons into a peripheral nerve implant, but others, including most thalamocortical projection neurons, will not. The ability to regenerate axons may depend on whether neurons can express growth-related genes such as GAP-43, whose expression correlates with axon growth during development and with competence to regenerate. Thalamic projection neurons which fail to regenerate into a graft also fail to upregulate GAP-43. We have tested the hypothesis that the absence of strong GAP-43 expression by the thalamic projection neurons prevents them from regenerating their axons, using transgenic mice which overexpress GAP-43. Transgene expression was mapped by in situ hybridization with a digoxigenin-labeled RNA probe and by immunohistochemistry with a monoclonal antibody against the GAP-43 protein produced by the transgene. Many CNS neurons were found to express the mRNA and protein, including neurons of the mediodorsal and ventromedial thalamic nuclei, which rarely regenerate axons into peripheral nerve grafts. Grafts were implanted into the region of these nuclei in the brains of transgenic animals. Although these neurons strongly expressed the transgene mRNA and protein and transported the protein to their axon terminals, they did not regenerate axons into the graft, suggesting that lack of GAP-43 expression is not the only factor preventing thalamocortical neurons regenerating their axons.

Expression of CHL1 and L1 by neurons and glia following sciatic nerve and dorsal root injury.

Cell adhesion molecules (CAMs), particularly L1, are important for axonal growth on Schwann cells in vitro. We have used in situ hybridization to study the expression of mRNAs for L1 and its close homologue CHL1, by neurons regenerating their axons in vivo, and have compared CAM expression with that of GAP-43. Adult rat sciatic nerves were crushed (allowing functional regeneration), or cut and ligated to maintain axonal sprouting but prevent reconnection with targets. In other animals lumbar dorsal roots were transected to produce slow regeneration of the central axons of sensory neurons. In unoperated animals L1 and CHL1 mRNAs were expressed at moderate levels by small- to medium-sized sensory neurons and L1 mRNA was expressed at moderate levels by motor neurons. Many large sensory neurons expressed neither L1 nor CHL1 mRNAs and motor neurons expressed little or no CHL1 mRNA. Neither motor nor sensory neurons showed any obvious upregulation of L1 mRNA after axotomy. Increased CHL1 mRNA was found in motor neurons and small- to medium-sized sensory neurons 3 days to 2 weeks following sciatic nerve crush, declining toward control levels by 5 weeks when regeneration was complete. Cut and ligation injuries caused a prolonged upregulation of CHL1 mRNA (and GAP-43 mRNA), indicating that reconnection with target tissues may be required to signal the return to control levels. Large sensory neurons did not upregulate CHL1 mRNA after axotomy and thus regenerated within the sciatic nerve without producing CHL1 or L1. Dorsal root injuries caused a modest, slow upregulation of CHL1 mRNA by some sensory neurons. CHL1 mRNA was also upregulated by many presumptive Schwann cells in injured nerves and by some satellite cells around large sensory neurons after sciatic nerve injuries and was transiently upregulated by some astrocytes in the degenerating dorsal columns after dorsal rhizotomy.

Immunoelectron microscopic study of glutamate inputs from the retrosplenial granular cortex to identified thalamocortical projection neurons in the anterior thalamus of the rat.

We have carried out an ultrastructural study to determine the characteristics and distribution of glutamate-containing constituents of the anterodorsal (AD) and anteroventral (AV) thalamic nuclei in adult rats. We used a polyclonal antibody to glutamate and a postembedding immunogold detection method in animals in which the neurons of AD/AV projecting to the cortex had been retrogradely labelled and the terminals of corticothalamic afferents anterogradely labelled by injection of cholera toxin-horseradish peroxidase (HRP) into the retrosplenial granular cortex. The heaviest immunogold labelling was over axon terminals 0.42 to 2.2 microm in diameter containing round synaptic vesicles and establishing Gray type 1 (asymmetric) synaptic contact (type 1 terminals) on HRP-labelled or non-labelled dendrites. Mean gold particle densities over such terminals were 3-4 times higher than the densities over the dendrites to which they were presynaptic and 5-6 times higher than over terminals establishing Gray type 2 (symmetric) synaptic contacts (type 2 terminals). Gold particle densities over neuronal cell bodies and dendrites and over a subpopulation of myelinated axons were intermediate between the densities over type 1 and type 2 terminals. In adjacent serial sections immunoreacted for gamma aminobutyric acid, type 2 terminals were heavily immunolabelled whereas type 1 terminals and other profiles with moderate gold particle densities after glutamate immunoreaction displayed very low labelling. A subpopulation of small type 1 axon terminals (up to 1 microm diameter) contained HRP reaction product identifying them as cortical in origin; they contacted small dendritic profiles (most <1 microm diameter) many of which also contained HRP reaction product. We conclude that terminals of the corticothalamic projection from retrosplenial granular cortex to AD/AV are glutamatergic and innervate predominantly distal dendrites of thalamocortical projection neurons.

The effects of FK506 on dorsal column axons following spinal cord injury in adult rats: neuroprotection and local regeneration.

There is considerable evidence that immunophilin ligands can promote the regeneration of axons in peripheral nerves and act as neuroprotective agents in the CNS. We have examined the effects of FK506 and GPI 1046 on the responses to partial transection of ascending spinal dorsal column axons at T9, in some cases combined with crush of one sciatic nerve. FK506 (0.5 or 2.0 mg/kg) and GPI 1046 (10 or 40 mg/kg) was administered subcutaneously immediately after surgery and five times a week thereafter. Some animals received methylprednisolone (MP) (two subcutaneous doses of 30 mg/kg) in addition to, or instead of, FK506. After survival times of 1-12 weeks, dorsal column axons were labeled transganglionically with cholera toxin B-HRP. There was massive axonal sprouting at the lesion sites in animals with sciatic nerve injury and immunophilin ligand treatment. In FK506-treated animals a few severed sensory axons regenerated for up to 10 mm rostral to the lesion. Of greater significance, 30% of 71 FK506-treated animals had spared axons in the dorsal column, extending to the nucleus gracilis, versus 8% of 50 control animals (P < 0.05), showing that FK506 reduces the likelihood of axonal destruction due to secondary injury. A combination of FK506 and MP afforded greater protection than MP alone (P < 0.05), but axonal survival was not affected by sciatic nerve crush, dose of FK506, or survival time after injury. GPI 1046 (n = 11) did not promote axonal survival. Thus FK506 protects axons from secondary injury following spinal cord trauma, and in this experimental model, its neuroprotective effect is greater than that of MP.

Immunoelectron microscopic study of gamma-aminobutyric acid inputs to identified thalamocortical projection neurons in the anterior thalamus of the rat.

We have carried out a semi-quantitative ultrastructural study to determine the characteristics and distribution of gamma-aminobutyric acid (GABA)-containing constituents of the anterodorsal (AD) and anteroventral (AV) thalamic nuclei in adult rats. We used a polyclonal antibody to GABA and a postembedding immunogold detection method in animals in which the cortical projection neurons of these nuclei had been labelled by retrograde transport of cholera toxin/horseradish peroxidase (HRP) injected into the retrosplenial granular cortex. Two types of GABA-immunopositive structures were identified, with gold particle densities 4-40 times higher than the highest densities over blood-vessel lumens and areas of empty resin: (1) an apparently homogeneous population of axon terminals with Gray type-2 (symmetric) synaptic contacts corresponding to F-axon terminals; and (2) small-medium sized myelinated axons scattered individually or in small groups within the neuropil which may be their parent axons. These axons and terminals may originate from the ipsilateral thalamic reticular nucleus; others may arise from the basal forebrain or brainstem. The GABA-immunopositive terminals comprised approximately 16% of all axon terminal profiles in AD and 12% in AV, a significant difference. However, because the immunoreactive axon terminals in AD were significantly larger than those in AV (1.09+/-0.47 microm2 vs 0.90+/-0.43 microm2) and would therefore be encountered more frequently, it is not possible to conclude that the GABAergic innervation of AD is heavier than that of AV. The GABA-positive terminals established synaptic contacts with cell bodies and dendrites of all sizes (some of which were HRP-labelled) with the following frequency distribution (AD/AV, no significant difference): somata 5%/7%; large dendrites (> or = 1.5 microm) 14%/9%; medium dendrites (1.00-1.49 microm) 35%/45% and small dendrites (< 1 microm) 46%/40%. Despite evidence from previous studies, we found no evidence in this study for the presence of GABAergic interneurons or for GABA-containing projection neurons in AD or AV.