Search

Previous experiments with purified mouse and bovine nerve growth factor (NGF) have shown that the biological activities of these two NGFs are identical, whereas the immunological cross‐reactivity of antibodies produced against the two NGF molecules is very limited. This observation, together with the fact that antibodies to mouse NGF do not affect the development of sympathetic and sensory neurons in chick embryos, suggests that the domain of the NGF molecules responsible for the biological action has been highly conserved during evolution, whereas other domains determining the immunological properties were under less rigorous evolutionary constraint. The nucleotide sequences of bovine and chick NGF were determined from a cDNA clone prepared from mRNA of bovine seminal vesicles and from cloned chick genomic DNA, and the amino acid sequences deduced therefrom were compared with the available sequences of mouse and human NGF. All six cysteine residues were conserved in agreement with the previous finding that the biological activity of NGF is conformation‐dependent requiring intact disulfide bridges. Amino acid changes are mainly confined to hydrophilic regions expected to be potential antigenic determinants, thus providing an explanation for the poor immunological cross‐reactivities between the different NGFs. One single hydrophilic region is conserved in all NGFs and this region could be involved in the biological activity. The carboxy termini of bovine and chick NGF differ from that of mouse NGF, the changes in the amino acid sequences suggest that chick and bovine NGF are probably not processed by the gamma‐subunit and that no 7S complex can be formed as in the mouse submandibular gland.

We have developed a sensitive assay for the quantification of nerve growth factor mRNA (mRNANGF) in various tissues of the mouse using in vitro transcribed RNANGF. Probes of both polarities were used to determine the specificity of the hybridization signals obtained. Comparison of NGF levels with its mRNA revealed that both were correlated with the density of sympathetic innervation. Thus, vas deferens contained high levels of both NGF and mRNANGF, whereas skeletal muscle levels were barely detectable, indicating that in peripheral tissues NGF levels are primarily regulated by the quantity of mRNANGF and not by the rate of processing of NGF precursor to NGF. However, although superior cervical ganglia contained the highest levels of NGF, its mRNA was barely detectable. Thus, the high levels of NGF in sympathetic ganglia result from retrograde axonal transport rather than local synthesis. The quantity of NGF found in the submandibular glands of female animals was three orders of magnitude higher than expected from their mRNA levels. This observation is discussed in the context of the difference between the mechanism of storage and exocytosis of exocrine glands versus the constitutive release from other tissues.

The mRNAs of nerve growth factor (NGF) and brain derived neurotrophic factor (BDNF) exhibit a similar, though not identical, regional and cellular distribution in the rodent brain. In situ hybridization experiments have shown that BDNF, like NGF, is predominantly expressed by neurons. The neuronal localization of the mRNAs of these two neurotrophic molecules raised the question as to whether neuronal activity might be involved in the regulation of their synthesis. After we had demonstrated that depolarization with high potassium (50 mM) resulted in an increase in the levels of both BDNF and NGF mRNAs in cultures of hippocampal neurons, we investigated the effect of a large number of transmitter substances. Kainic acid, a glutamate receptor agonist, was by far the most effective in increasing BDNF and NGF mRNA levels in the neurons, but neither N‐methyl‐D‐aspartic acid (NMDA) nor inhibitors of the NMDA glutamate receptors had any effect. However, the kainic acid mediated increase was blocked by antagonists of non‐NMDA receptors. Kainic acid also elevated levels of BDNF and NGF mRNAs in rat hippocampus and cortex in vivo. These results suggest that the synthesis of these two neurotrophic factors in the brain is regulated by neuronal activity via non‐NMDA glutamate receptors.

We report the purification from pig brain of a factor supporting the survival of, and fibre outgrowth from, cultured embryonic chick sensory neurons. The purified factor migrates as one single band, mol. wt. 12 300, on gel electrophoresis in the presence of sodium dodecylsulphate (SDS) and is a basic molecule (pI greater than or equal to 10.1). Approximately 1 microgram factor was isolated from 1.5 kg brain. The final degree of purification was estimated to be 1.4 X 10(6)‐fold, and the specific activity 0.4 ng/ml/unit, which is similar to that of nerve growth factor (NGF) using the same assay system. This factor is the first neurotrophic factor to be purified since NGF, from which it is clearly distinguished because it has different antigenic and functional properties.

Choline acetyltransferase (ChAT) has been purified from pig brain to greater than 95% homogeneity (purification factor: 646 000, specific activity of the purified enzyme: 128 mumol acetylcholine formed/min/mg). Gel electrophoresis of the purified enzyme in the presence of sodium dodecylsulphate and beta‐mercaptoethanol revealed a single protein band at 68 000 daltons. Immunoprecipitation and double immunodiffusion tests showed that antisera raised against this protein specifically recognize ChAT. A monoclonal antibody prepared against the enzyme specifically binds a protein from crude pig brain supernatants which has a mol. wt. of 68 000 and a specific activity of 153 mumol/min/mg. This antibody shows no species cross‐reactivity. The specificity of the immunohistochemical localization of ChAT has been established by comparing the labeling of pig retina using the antiserum with that obtained using the monoclonal antibody. Both probes specifically identify the same retinal structures: labeled cell bodies are found in the inner nuclear layer and the ganglion cell layer, while a double band is stained in the inner plexiform layer. In rat spinal cord, the antiserum labels the motoneurons and the preganglionic sympathetic neurons, located in the intermedio‐lateral nucleus, the intercalated region, and the central autonomic area.

The levels of nerve growth factor (NGF) and its mRNA in the rat central nervous system were determined by two‐site enzyme immunoassay and quantitative Northern blots, respectively. Relatively high NGF levels (0.4‐1.4 ng NGF/g wet weight) were found both in the regions innervated by the magnocellular cholinergic neurons of the basal forebrain (hippocampus, olfactory bulb, neocortex) and in the regions containing the cell bodies of these neurons (septum, nucleus of the diagonal band of Broca, nucleus basalis of Meynert). Comparatively low, but significant NGF levels (0.07‐0.21 ng NGF/g wet weight) were found in various other brain regions. mRNANGF was found in the hippocampus and cortex but not in the septum. This suggests that magnocellular cholinergic neurons of the basal forebrain are supplied with NGF via retrograde axonal transport from their fields of innervation. These results, taken together with those of previous studies showing that these neurons are responsive to NGF, support the concept that NGF acts as trophic factor for magnocellular cholinergic neurons.

The survival of cultured chick sympathetic neurons and the outgrowth of neurites were stimulated by the basement membrane protein laminin coated onto polyornithine culture substrates. The survival‐potentiating activity was dependent on the presence of nerve growth factor. Both effects of laminin could be completely inhibited by affinity‐purified antibodies against laminin fragment 3, the product of a limited proteolysis that corresponds to the heparin‐binding globular domain at the end of the long arm of the laminin molecule. Antibodies against other laminin fragments were inactive, including those against previously determined cell‐binding domains. A large laminin fragment, E8, was produced by brief elastase digestion and shown to consist of fragment 3 and an adjacent rod‐like structure. Although lacking the cell binding domains, fragment E8 potentiated both neuronal survival and neurite outgrowth, and these effects could be blocked by antibodies against fragment 3. Weak survival and neurite potentiating activity was also detected in another fragment corresponding to the short arms of laminin, but as these effects were not inhibited by any of the antibodies tested they probably arose de novo during proteolysis. The heparin‐binding domain of laminin is therefore responsible for its effects on neurons.

Nerve growth factor (NGF) and elevated K+ concentrations (35 mM) support the survival of the same population of chick embryonic sympathetic neurons. We have used methyltransferase inhibitors, which block protein methylation in intact cells, to investigate the mechanism(s) by which NGF and high K+ exert their effects. Methyltransferase inhibitors selectively blocked NGF‐but not high K+‐mediated survival of neurons. The ability of neurons, plated on laminin, to respond rapidly to NGF with neurite outgrowth was used to demonstrate that the blockade of the effects of NGF by methyltransferase inhibitors was reversible. At the molecular level, we studied the rapid decrease in phosphorylation of p70, a 70‐kd phosphoprotein of sympathetic neurons regulated by both NGF and high K+. Methyltransferase inhibitors blocked the decrease in p70 phosphorylation induced by NGF but not that by high K+. We conclude that the early molecular events of NGF‐mediated neuronal survival differ from those of high K+‐mediated neuronal survival in that they involve protein methylation, whereas at a later step, possibly at the level of protein phosphorylation, the two pathways leading to survival of sympathetic neurons converge.

A very sensitive and specific method for in situ hybridization has been developed. This method detects low copy numbers of mRNA(NGF) transcripts in both tissue sections and cultured cells using 35S‐labelled cRNA and oligonucleotide probes. In order to reduce the high nonspecific background occurring with 35S‐labelled probes, prehybridization in the presence of non‐labelled thio alpha UTP at pH 5.5 proved to be essential, together with a series of additional changes in the standard procedures for in situ hybridization. With this improved method it was possible to demonstrate that in tissues densely innervated by sensory (whisker pad) or both sympathetic and sensory (iris) fibers, NGF is synthesized not only by Schwann cells ensheathing these fibers, but also‐‐and even to a much larger extent‐‐by the target cells of the sensory and sympathetic neurons, i.e. epithelial cells, smooth muscle cells and fibroblasts. Moreover, in the sciatic nerve of newborn rats (where the mRNA(NGF) levels are 15 X higher than in adults) it was demonstrated that all Schwann cells have the capacity to express mRNA(NGF), not just those ensheathing the axons of NGF‐responsive neurons.

We have analyzed the appearance of neurons and glial cells in chick dorsal root ganglia during development. Neurons were identified by the presence of polysialogangliosides recognized by tetanus toxin (GD1b, GT1) or by the monoclonal antibody Q211 directed against polysialogangliosides containing four, five and six sialic acid residues. Glial cells were identified by the presence of 04 antigen. A population of undifferentiated cells, i.e., cells which express neither neuronal nor glial cell surface antigens, present in dorsal root ganglia until embryonic day 7, was separated from the neuronal and glial population. This cell population contains neuronal progenitor cells which differentiate to neurons within 1 day in culture. This differentiation process is characterized by the appearance of neuronal morphology, of neuron‐specific gangliosides and by the appearance of voltage‐dependent sodium and calcium channels.

We report the purification from pig brain of a factor supporting the survival of, and fibre outgrowth from, cultured embryonic chick sensory neurons. The purified factor migrates as one single band, mol. wt. 12 300, on gel electrophoresis in the presence of sodium dodecylsulphate (SDS) and is a basic molecule (pI greater than or equal to 10.1). Approximately 1 microgram factor was isolated from 1.5 kg brain. The final degree of purification was estimated to be 1.4 X 10(6)‐fold, and the specific activity 0.4 ng/ml/unit, which is similar to that of nerve growth factor (NGF) using the same assay system. This factor is the first neurotrophic factor to be purified since NGF, from which it is clearly distinguished because it has different antigenic and functional properties.

The mRNAs of nerve growth factor (NGF) and brain derived neurotrophic factor (BDNF) exhibit a similar, though not identical, regional and cellular distribution in the rodent brain. In situ hybridization experiments have shown that BDNF, like NGF, is predominantly expressed by neurons. The neuronal localization of the mRNAs of these two neurotrophic molecules raised the question as to whether neuronal activity might be involved in the regulation of their synthesis. After we had demonstrated that depolarization with high potassium (50 mM) resulted in an increase in the levels of both BDNF and NGF mRNAs in cultures of hippocampal neurons, we investigated the effect of a large number of transmitter substances. Kainic acid, a glutamate receptor agonist, was by far the most effective in increasing BDNF and NGF mRNA levels in the neurons, but neither N‐methyl‐D‐aspartic acid (NMDA) nor inhibitors of the NMDA glutamate receptors had any effect. However, the kainic acid mediated increase was blocked by antagonists of non‐NMDA receptors. Kainic acid also elevated levels of BDNF and NGF mRNAs in rat hippocampus and cortex in vivo. These results suggest that the synthesis of these two neurotrophic factors in the brain is regulated by neuronal activity via non‐NMDA glutamate receptors.

Nerve growth factor (NGF) and elevated K+ concentrations (35 mM) support the survival of the same population of chick embryonic sympathetic neurons. We have used methyltransferase inhibitors, which block protein methylation in intact cells, to investigate the mechanism(s) by which NGF and high K+ exert their effects. Methyltransferase inhibitors selectively blocked NGF‐but not high K+‐mediated survival of neurons. The ability of neurons, plated on laminin, to respond rapidly to NGF with neurite outgrowth was used to demonstrate that the blockade of the effects of NGF by methyltransferase inhibitors was reversible. At the molecular level, we studied the rapid decrease in phosphorylation of p70, a 70‐kd phosphoprotein of sympathetic neurons regulated by both NGF and high K+. Methyltransferase inhibitors blocked the decrease in p70 phosphorylation induced by NGF but not that by high K+. We conclude that the early molecular events of NGF‐mediated neuronal survival differ from those of high K+‐mediated neuronal survival in that they involve protein methylation, whereas at a later step, possibly at the level of protein phosphorylation, the two pathways leading to survival of sympathetic neurons converge.

Although ciliary neurotrophic factor (CNTF) is a potent survival factor for many types of neurons and glial cellsin vitro,there is currently no evidence that it participates in normal development. Here we show that CNTF greatly enhances the rate of oligodendrocyte generation. Proliferation of oligodendrocyte precursor cells purified from rodent optic nerves and cultured in platelet-derived growth factor-containing medium is significantly increased by CNTF. Similarly, the number of proliferating oligodendrocyte precursor cells in developing optic nerves of transgenic mice lacking CNTF is decreased by up to threefold and the number of oligodendrocytes is transiently decreased; proliferation is restored to normal by the delivery of exogenous CNTF into the developing optic nerve. Both oligodendrocyte number and myelination ultimately attain wild-type values in CNTF-deficient adult mice, indicating that CNTF is not necessary for either oligodendrocyte differentiation or myelination, although it normally accelerates oligodendrocyte development by enhancing the proliferation of oligodendrocyte precursor cells.

The fate of tetanus toxin (mol wt 150,000) subsequent to its retrograde axonal transport in peripheral sympathetic neurons of the rat was studied by both electron microscope autoradiography and cytochemistry using toxin-horseradish peroxidase (HRP) coupling products, and compared to that of nerve growth factor (NGF), cholera toxin, and the lectins wheat germ agglutinin (WGA), phytohaemagglutinin (PHA), and ricin. All these macromolecules are taken up by adrenergic nerve terminals and transported retrogradely in a selective, highly efficient manner. This selective uptake and transport is a consequence of the binding of these macromolecules to specific receptive sites on the nerve terminal membrane. All these ligands are transported in the axons within smooth vesicles, cisternae, and tubules. In the cell bodies these membrane compartments fuse and most of the transported macromolecules are finally incorporated into lysosomes. The cell nuclei, the parallel golgi cisternae, and the extracellular space always remain unlabeled. In case the tetanus toxin, however, a substantial fraction of the labeled material appears in presynaptic cholinergic nerve terminals which innervate the labeled ganglion cells. In these terminals tetanus toxin-HRP is localized in 500-1,000 A diam vesicles. In contrast, such a retrograde transsynaptic transfer is not at all or only very rarely detectable after retrograde transport of cholera toxin, NGF, WGA, PHA, or ricin. An atoxic fragment of the tetanus toxin, which contains the ganglioside-binding site, behaves like intact toxin. With all these macromolecules, the extracellular space and the glial cells in the ganglion remain unlabeled. We conclude that the selectivity of this transsynaptic transfer of tetanus toxin is due to a selective release of the toxin from the postsynaptic dendrites. This release is immediately followed by an uptake into the presynaptic terminals.

gamma-Aminobutyric acid (GABA) is the major inhibitory neurotransmitter in the adult mammalian central nervous system. However, GABA depolarizes immature rat hippocampal neurons and increases intracellular Ca2+ ([Ca2+]i). Here we show, that GABA and the GABAA receptor agonist muscimol induce c-Fos immunoreactivity and increase BDNF mRNA expression in embryonic hippocampal neurons cultured for 5 days. In contrast, after 3 weeks in culture, GABA and muscimol failed to induce c-fos and BDNF expression. Fura-2 fluorescence microscopy revealed that muscimol produces a dihydropyridine-sensitive transient increase in [Ca2+]i, comparable to the effect of the non-NMDA receptor agonist kainic acid in neurons cultured for 5 days, but not in 3-week-old cultures. The increase in c-Fos immunoreactivity and BDNF mRNA levels by GABA were dependent upon the activation of voltage-gated Ca2+ channels, as shown using the L-type specific Ca2+ channel blocker nifedipine. The differential regulation of c-fos and BDNF expression by GABA and muscimol in developing and mature hippocampal neurons is due to a switch in the ability of GABAA receptors to activate voltage-gated Ca2+ channels. These observations support the hypothesis that GABA might have neurotrophic effects on embryonic or perinatal hippocampal neurons, which are mediated by BDNF.

Previous experiments with purified mouse and bovine nerve growth factor (NGF) have shown that the biological activities of these two NGFs are identical, whereas the immunological cross‐reactivity of antibodies produced against the two NGF molecules is very limited. This observation, together with the fact that antibodies to mouse NGF do not affect the development of sympathetic and sensory neurons in chick embryos, suggests that the domain of the NGF molecules responsible for the biological action has been highly conserved during evolution, whereas other domains determining the immunological properties were under less rigorous evolutionary constraint. The nucleotide sequences of bovine and chick NGF were determined from a cDNA clone prepared from mRNA of bovine seminal vesicles and from cloned chick genomic DNA, and the amino acid sequences deduced therefrom were compared with the available sequences of mouse and human NGF. All six cysteine residues were conserved in agreement with the previous finding that the biological activity of NGF is conformation‐dependent requiring intact disulfide bridges. Amino acid changes are mainly confined to hydrophilic regions expected to be potential antigenic determinants, thus providing an explanation for the poor immunological cross‐reactivities between the different NGFs. One single hydrophilic region is conserved in all NGFs and this region could be involved in the biological activity. The carboxy termini of bovine and chick NGF differ from that of mouse NGF, the changes in the amino acid sequences suggest that chick and bovine NGF are probably not processed by the gamma‐subunit and that no 7S complex can be formed as in the mouse submandibular gland.

The survival of cultured chick sympathetic neurons and the outgrowth of neurites were stimulated by the basement membrane protein laminin coated onto polyornithine culture substrates. The survival‐potentiating activity was dependent on the presence of nerve growth factor. Both effects of laminin could be completely inhibited by affinity‐purified antibodies against laminin fragment 3, the product of a limited proteolysis that corresponds to the heparin‐binding globular domain at the end of the long arm of the laminin molecule. Antibodies against other laminin fragments were inactive, including those against previously determined cell‐binding domains. A large laminin fragment, E8, was produced by brief elastase digestion and shown to consist of fragment 3 and an adjacent rod‐like structure. Although lacking the cell binding domains, fragment E8 potentiated both neuronal survival and neurite outgrowth, and these effects could be blocked by antibodies against fragment 3. Weak survival and neurite potentiating activity was also detected in another fragment corresponding to the short arms of laminin, but as these effects were not inhibited by any of the antibodies tested they probably arose de novo during proteolysis. The heparin‐binding domain of laminin is therefore responsible for its effects on neurons.

The characteristics of the constitutive and activity- mediated secretion of NGF from native hippocampal slices are the same as those from hippocampal cultures transfected with an NGF-overexpressing plasmid (Blöchl and Thoenen, 1995). In these cultures, the distribution of intracellular NGF immunoreactivity—including the colocalization with endoplasmic reticulum (ER) and Golgi markers—in soma, dendrites, and axons, as visualized by confocal microscopy, is compatible with a localization of NGF in an ER-like compartment. Since the positively charged NGF molecule is bound, at the sites of its release, to the negatively charged neuronal surface, at low salt buffer concentrations, it was possible to attribute the different release mechanisms to specific neuronal surface sites. Constitutive secretion of NGF is confined to the neuronal soma and the very proximal parts of dendrites. In contrast, the activity-dependent secretion, initiated by high potassium or glutamate also occurs all along the neuronal processes, in particular dendrites. This release is independent of extracellular calcium, but depends on calcium released from intracellular calcium stores and is mediated by sodium influx via voltage-gated sodium channels and non-NMDA glutamate receptors. The confocal intensity analysis of the NGF surface staining permitted quantitative assessment of the different release mechanisms to different neuronal domains.

A series of specific macromolecules (tetanus toxin, cholera toxin, nerve growth factor [NGF], and several lectins) have been shown to be transported retrogradely with high selectivity from terminals to cell bodies in various types of neurons. Under identical experimental conditions (low protein concentrations injected), most other macromolecules, e.g. horseradish peroxidase (HRP), albumin, ferritin, are not transported in detectable amounts. In the present EM study, we demonstrate selective binding of tetanus toxin to the surface membrane of nerve terminals, followed by uptake and subsequent retorgrade axonal transport. Tetanus toxin or albumin was adsorbed to colloidal gold particles (diam 200 A). The complex was shown to be stable and well suited as an EM tracer. 1-4 h after injection into the anterior eye chamber of adult rats, tetanus toxin-gold particles were found to be selectively associated with membranes of nerve terminals and preterminal axons. Inside terminals and axons, the tracer was localized mainly in smooth endoplasmic reticulum (SER)-like membrane compartments. In contrast, association of albumin-gold complexes with nervous structures was never observed, in spite of extensive uptake into fibroblasts. Electron microscope and biochemical experiments showed selective retrograde transport of tetanus toxin-gold complexes to the superior cervical ganglion. Specific binding to membrane components at nerve terminals and subsequent internalization and retrograde transport may represent an important pathway for macromolecules carrying information from target organs to the perikarya of their innervating neurons.

The lymphokine interleukin 1 (IL-1) rapidly increases the levels of mRNA encoding nerve growth factor (NGF) in primary cultures of fibroblasts isolated from adult rat sciatic nerves. The effect of IL-1 on NGF-mRNA was not inhibited by cycloheximide, which itself increased the NGF-mRNA. From the blocking ability of actinomycin D, it was deduced that the IL-1 action depends on continuous RNA synthesis and that the half-life of NGF-mRNA is about 90 min. Nuclear run-on studies showed that, although IL-1 enhances the low basic NGF-mRNA transcription, the main effect of IL-1 results from stabilization of the NGF-mRNA. In addition, indirect evidence is presented indicating that the IL-1 effect of NGF-mRNA levels involves activation of phospholipase A2.

The proto-oncogene bcl-2 and its family members, bcl-x and bax are recognized as major regulators of cell death and survival. Although Bcl-2 and Bcl-x are expressed in brain, little is known how they are regulated in neurons. Here we have studied the expression of bcl-2, bcl-xL and bax mRNA in rat cerebellar granule neurons cultured under conditions which influence neuron survival. Insulin-like growth factor-1 and brain-derived neurotrophic factor supported the survival of these neurons, but affected neither the expression of bcl-2, bcl-xL nor bax mRNA. In contrast, bcl-2, and bcl-xL mRNAs were up-regulated in cerebellar granule neurons plated at high density exhibiting an increased neuronal survival. Western blots showed that cell density also increased Bcl-2 protein level. However, conditioned medium from dense cultures did not affect the level of bcl-2 mRNA nor survival of the neurons.

We have detected scattered brain-derived neurotrophic factor mRNA-producing neurons in the medial septal nucleus, which contains cholinergic neurons that are responsive to brain-derived neurotrophic factor and nerve growth factor. In the brainstem, many adrenergic neurons showed a positive signal for brain-derived neurotrophic factor messenger RNA. Several hypothalamic nuclei contain brain-derived neurotrophic factor messenger RNA-positive neurons, among them paraventricular, median preoptic, vetromedial and dorsomedial nuclei.

We have investigated the effects of substrate-bound laminin on levels of enzymes of the catecholamine biosynthetic pathway in primary cultures of calf adrenal chromaffin cells. Laminin increases the levels of the enzymes tyrosine hydroxylase, dopamine-beta-hydroxylase, and phenylethanolamine-N-methyl-transferase. This effect is selective, in that levels of other enzymes (lactate dehydrogenase, aromatic amino acid decarboxylase, and acetylcholinesterase) are not increased. The effect of laminin can be blocked by antibodies directed against a fragment of the heparin-binding domain of the molecule, whereas antibodies directed against other fragments do not block the increase in tyrosine hydroxylase. Thus the laminin domain involved in enzyme regulation in chromaffin cells is apparently the same as that previously implicated in laminin's interactions with neurons to potentiate survival and stimulate neurite outgrowth (Edgar, D., R. Timpl, and H. Thoenen, 1984, EMBO (Eur. Mol. Biol. Organ.) J., 3:1463-1468). The increase in chromaffin cell tyrosine hydroxylase levels is preceded by an activation of the enzyme in which the Vmax (but not the Km) is altered. The effects of laminin appear to be developmentally regulated, since neither activation nor increased levels of tyrosine hydroxylase occur in adult adrenal chromaffin cells exposed to laminin.

Laminin derived from the Engelbreth-Holm-Swarm (EHS) tumor and a lamininlike molecule synthesized by RN22 Schwannoma cells both stimulate rapid neurite outgrowth, consistent with a common neurite-promoting site. However, antilaminin antisera can only inhibit the activity of the EHS laminin. The blocking antibodies in such sera are directed against the terminal heparin-binding domain of the laminin long arm (Edgar, D., R. Timpl, and H. Thoenen. 1984. EMBO [Eur. Mol. Biol. Organ.] J. 3: 1463-1468). These epitopes are demonstrated by immunoblotting to be part of the A chain and to be absent in RN22 laminin, showing (through metabolic labeling) that the cells synthesized little if any 440-kD A chain. This indicates that the antibody inhibition was probably due to steric hindrance, a common neurite-promoting site, apparently not being antigenic in native molecules. Antibodies raised against a 25-kD proteolytic fragment derived from the long arm of laminin were then used as probes to identify other potential neurite-promoting structures. Although these antibodies do not cross-react with native laminin, they recognized the B chains of denatured EHS and RN22 molecules on immunoblots. The antibodies also bound to the large proteolytic fragment, derived from the long arm of laminin that contains the neurite-promoting site, thus inhibiting its activity. Taken together, these results point to the localization of normally nonantigenic, defined, B chain sequences within or close to the neurite-promoting site of laminin.

To decide whether adenosine 3′:5′-cyclic monophosphate (cyclic AMP) plays a role as a second messenger in the trans-synaptic induction of tyrosine 3-monooxygenase (EC 1.14.16.2), it is desirable to discriminate between neuronal and extraneuronal changes in cyclic AMP concentration. Treatment of newborn rats with nerve growth factor antiserum or 6-hydroxydopamine, leading to destruction of 61-85% of the adrenergic nerve cell bodies in the superior cervical ganglion, led to a decrease in cyclic AMP of only 16-28%. This observation demonstrates that a relatively small portion of cyclic AMP is localized in the adrenergic neurons. However, administration of isoproterenol produced an increase (12-fold) in cyclic AMP only in this neuronal pool. Neither single nor repeated injections of isoproterenol led to induction of tyrosine monoxygenase. This, together with previous observations that experimental conditions leading to induction of the enzyme do not produce significant increases of cyclic AMP in the whole ganglion, is taken as an indication that cyclic AMP is not acting as a second messenger in the trans-synaptic induction of tyrosine monooxygenase in the rat superior cervical ganglion. In the rat adrenal medulla, treatment with reserpine led to both a shortlasting (60-90 min) increase in cyclic AMP and a subsequent induction of tyrosine monooxygenase. However, the increase in cyclic AMP was almost completely prevented (40 compared to 320%) by pretreatment of the rats with propranolol while the induction of tyrosine monooxygenase was not diminished. This observation also argues against an exclusive key-function of cyclic AMP in trans-synaptic induction of tyrosine monooxygenase in the adrenal medulla.

We studied the synthesis and release of nerve growth factor (NGF) in cultured rat iris with a two-site enzyme immunoassay by measuring the time course of NGF levels remaining in the iris and relased into the medium up to 72 h. For up to 3 h, the NGF levels in the iris did not change significantly. After that, they increased to a maximal level of 350 +/- 30 pg NGF/iris at 19 h, which is 200 times higher than the in vivo content. Between 20 and 72 h in culture, the NGF level decreased to 130 +/- 10 pg NGF/iris, whereas general protein synthesis did not change during that time period. Maximal rate of NGF production (203 pg NGF/h/iris) was seen between 9 and 12 h in culture. In the medium, NGF levels were first detectable after 6 h. Levels then increased with a time course similar to that seen within the iris, reaching a maximal level of 1,180 +/- 180 pg after 19 h in vitro, and then did not significantly change for up to 48 h. The NGF production of the densely sympathetically innervated dilator was three times higher than that of the predominantly cholinergically innervated sphincter. The NGF production was blocked by inhibitors of messenger RNA synthesis (actinomycin D) and of polyadenylation (9-beta-D-arabinofuranosyladenine) as well as by inhibitors of translation (cycloheximide). Monensin, which interferes with the transport of proteins through the Golgi apparatus, decreased NGF levels to 8-12% of controls in the medium, suggesting that the Golgi apparatus is involved in the intracellular processing of NGF.

Specific sensory input has profound transient and long-lasting effects on the function of corresponding sensory cortical areas both during development and in adulthood. To study whether neurotrophic factors might play a role in such processes, we investigated the effects of light on the nerve growth factor and brain-derived neurotrophic factor (BDNF) mRNA levels in rat visual cortex. Keeping adult rats in the dark or preventing normal activity of retinal ganglion cells by intraocular injection of tetrodotoxin significantly decreased the levels of BDNF mRNA in the visual cortex but not in other cortical areas. Exposure to light after a period in darkness rapidly restored the mRNA to control levels. These alterations in visual input had no effect on nerve growth factor mRNA. The mRNA of trkB, the putative signal-transducing receptor unit for BDNF, was also decreased in darkness, although less than BDNF mRNA. BDNF mRNA levels increased in the visual cortex of newborn rats after eye-opening. This increase is retarded, although not completely abolished, by rearing the pups in darkness. Thus, the levels of BDNF mRNA are rapidly regulated by sensory input during development and in adulthood. BDNF may therefore play an important role in formation and in activity-dependent modulation of specific connections in the visual cortex.

Concomitant daily treatment of newborn rats for a 2-week to 1-month period with 10 mug/g of body weight of nerve growth factor and 100 mug/g of body weight of 6-hydroxydopamine produces in the cell bodies of adrenergic neurons the characteristic effects of the growth factor but in the nerve terminals the characteristic effects of 6-hydroxydopamine. The dual opposite effects result in a striking volume increase of sympathetic ganglia which far exceeds that produced by nerve growth factor alone. The selective induction of tyrosine hydroxylase [L-tyrosine, tetrahydropteridine:oxygen oxidoreductase (3-hydroxylating), EC 1.14.16.2] in these chemically axotomized adrenergic neurons is even more pronounced than that produced by nerve growth factor alone in intact neurons.

The trans-synaptic induction of tyrosine hydroxylase [tyrosine 3-monooxygenase; EC 1.14.16.2, L-tyrosine, tetrahydropteridine: oxygen oxidoreductase (3-hydroxylating)] in adrenal medulla and sympathetic ganglia by short-term (1-2 hr) cold stress (4 degrees) exhibits a circadian rhythm which seems to be causally related to the diurnal changes in adrenal glucocorticoid synthesis. In induction is maximal during the morning hours, when plasma corticoid concentrations (reflecting corticoid synthesis in the adrenal cortex) are minimal. In contrast, initiation of tyrosine hydroxylase induction in sympathetic ganglia is only possible in the afternoon. These observations suggest that tyrosine hydroxylase inducibility in the adrenal medulla is optimal during periods of low corticoid synthesis (the adrenal medulla is exposed to excessively high corticoid concentrations directly originating from the adjacent cortex), whereas in sympathetic ganglia an induction is only possible during the period of high plasma corticoid concentrations. This assumption is supported by the observation that in the first postnatal weeks, when the pituitary--adrenocortical system is not yet operating and plasma corticoid concentrations are low, initiation of tyrosine hydroxylase induction in the adrenal medulla is possible at any time of the day, whereas in sympathetic ganglia it is not possible at all. However, after administration of glycocorticoids initiation of tyrosine hydroxylase induction by short-term cold stress is also possible in newborn animals and in adults during the morning hours. The importance of glucocorticoids as modulators for the initiation of trans-synaptic tyrosine hydroxylase induction can also be deduced from the observation that in sympathetic ganglia kept in organ cultures and induction of the hydroxylase by cholinomimetics is only possible when glycocorticoids are added to the culture medium.

Addition of nerve growth factor to cultures of dissociated rat adrenal medullary cells caused fiber outgrowth from chromaffin cells. These fibers exhibited all the characteristics of neurites, particularly the formation of typical growth cones exhibiting intense catecholamine-specific fluorescence. Because this nerve growth factor-mediated neurite outgrowth could be abolished by physiological concentrations of glucocorticoids, it is concluded that the high glucocorticoid concentrations normally present in the adrenal medulla prevent the fiber outgrowth from medullary chromaffin cells in vivo. In dissociated sympathetic neurons the same concentrations of glucocorticoids markedly reduce but do not completely abolish neuronal fiber outgrowth.

In previous experiments, it has been demonstrated that, in rat irides in culture, a rapid increase in nerve growth factor (NGF) levels occurred (see Barth, E.-M., Korsching, S., and Thoenen, H. (1984) J. Cell Biol. 99, 839-843). We have now determined the levels of mRNANGF in rat irides as a function of time in culture as well. After an initial lag period of 2 h, mRNANGF levels were transiently increased, so that after 12 h, they had increased 35-fold with respect to zero time. In contrast, poly(A)+ RNA levels dropped to 55% of the zero time values within 5 h, recovered to 85% after 24 h, and remained constant until the end of the observation period. Total ribosomal RNA was found to remain constant, indicating that there was no nonspecific decline of overall metabolic function. Actinomycin D prevented the increase in mRNANGF without reducing the basic mRNANGF levels over a 5-h time period, indicating that the enhanced synthesis of NGF in the rat iris in culture is primarily mediated by an augmented production of mRNANGF. The increases of mRNANGF, cellular NGF, and NGF released into the medium were found to be strictly sequential. Monensin selectively abolished the increased production of mature NGF (see Barth et al.) but not of mRNANGF, suggesting that the processing of NGF precursor is prevented.

High levels of nerve growth factor (NGF) determined by competition radioimmunoassay do not agree with values obtained by bioassay. This discrepancy is illustrated here with rat and mouse serum as examples in which values up to 1000 ng/ml have been found by using competition radioimmunoassays. An explanation for the discordant results is presented: serum components bind NGF with an intermediate affinity (Kd = 10(-7) M) but with a very large capacity (up to 0.5 mg of NGF per ml of rat serum). The binding of 125I-labeled NGF to serum components competes with the binding to the solid-phase antibodies (Kd = 10(-9)M) present in limiting amounts, according to the principle of competition radioimmunoassays. Thus, less radioactivity is recovered bound to the antibodies and this gives the erroneous impression that NGF is present. To overcome this difficulty, a two-site radioimmunoassay has been developed which utilizes nonlimiting numbers of antibody binding sites. This assay provides a reliable determination of NGF levels in serum and it can be shown that in rat and mouse serum (either sex) there is less than 5 ng of NGF per ml, in agreement with the results of the bioassay.

The effect of nerve growth factor (NGF) and medium conditioned by glioma cells (GCM) on the survival of chicken sensory neurons in culture was investigated. Neurons were isolated from embryos 8 days (E8) to 16 days (E16) old and the proportion of surviving neurons was determined after 2 days in culture. In the absence of NGF or GCM, essentially no neurons survived at any age. In the presence of NGF, survival increased from 25% of the neurons at E8 to 40% between E10 and E12 and then decreased to background level (5%) at E16. In contrast, in the presence of GCM, survival increased continuously from 10% of the neurons at E8 to 75% at E16. At early developmental stages, the effect of NGF and GCM together was greater than the sum of their individual effects: at E8, about 80% of the neurons survived, double the number expected for a simple additive effect. Thus, a significant proportion of chicken neurons from dorsal root ganglia require both NGF and GCM for survival, and this may well include neurons from the ventro-lateral population, which do not respond to NGF alone. As neurons matured, the double requirement progressively decreased and, by E16, NGF no longer increased the number of neurons over that surviving in response to GCM alone. The facts that rat brain extracts mimicked the effect of GCM and that the potency of the brain extracts of rat in the postnatal period increased in parallel with the development of the glial cells suggest that glial cells produce a factor(s) both immunologically and functionally different from NGF which supports the survival of sensory neurons.

A two-site enzyme immunoassay is described which does not suffer from artifacts inherent in previous assays and has the necessary high sensitivity to determine the endogenous levels of nerve growth factor (NGF) in the sympathetic nervous system and its target organs. Monoclonal and affinity-purified polyclonal antibodies against mouse NGF (mNGF) were covalently linked to glass beads as the first site and coupled to the enzyme beta-galactosidase as the second site. Detection of the fluorescent beta-galactosidase reaction product permitted the determination of 0.01-0.02 fmol of mNGF per assay. The recovery of mNGF added to homogenates varied between 50% and 100%, depending on the tissue. Rat superior cervical and stellate ganglia were found to contain (mean +/- SEM) 25 +/- 4 and 19 +/- 3 ng of NGF per g wet weight, respectively, and the densely innervated submandibular gland, heart atrium, and iris contained 0.5 +/- 0.1, 1.0 +/- 0.1, and 1.9 +/- 0.3 ng of NGF per g wet weight, respectively. Heart ventricle and skeletal muscle, which are poorly innervated by the sympathetic nervous system, did not contain detectable levels of NGF (less than 0.3 ng/g wet weight). Serum contained less than 0.05 ng of NGF per ml. The correlation between NGF levels and density of innervation is consistent with the concept that the production of NGF in target organs determines their density of innervation by the sympathetic nervous system.

Nerve growth factor (NGF) was localized in the submandibular gland of adult male mice by a direct immunocytochemical method using highly purified antibodies against NGF coupled to horseradish peroxidase. In light microscopic sections the reaction product was entirely confined to the cells of the secretory tubules. The acinar part of the gland was free of reaction product. This finding was confirmed by electron microscopy. Within the cells NGF was localized exclusively in the apical secretory granules.

1.Isolated superior cervical ganglia of the rat were incubated for 2–30 min (37°C) in Krebs' solution or tissue culture medium (BGJb) containing 22Na and then washed for 30 min in ice-cold 22Na-free Krebs' solution (to clear the extracellular space). The radioactivity remaining in the ganglia was taken as a measure of 22Na influx into the intracellular compartment of the ganglion.2.Addition of cholinomimetics (100 µM nicotine or 100 µM carbachol) to the incubation medium led to an increase in 22Na influx. This increase reached maximal values after 10 min of incubation; it was more pronounced after incubation in Krebs' solution than in BGJb medium.3.While chlorisondamine (3 µM) or dopamine (100 µM) greatly reduced the carbachol-induced 22Na influx, tetrodotoxin (2 µM) did not have any effect.4.In ganglia obtained from animals treated with 6-hydroxydopamine in the early postnatal phase (resulting in an extensive destruction of peripheral sympathetic neurons) neither carbachol (100 µM) nor nicotine (100 µM) produced an increase in 22Na influx demonstrating that the intraneuronal compartment is responsible for this enhanced influx.5.The effects of dopamine, chlorisondamine and tetrodotoxin on the carbachol-induced 22Na uptake into superior cervical ganglia are similar to their effects on carbachol-mediated induction of tyrosine hydroxylase in superior cervical ganglia kept in tissue culture (Thoenen and Otten 1977b). It is concluded that the induction of tyrosine hydroxylase via nicotinic receptors is closely linked to the enhanced sodium influx into the adrenergic neurons mediated by the same receptors.

It was the purpose of the present experiments to establish whether the reserpine-mediated changes in adrenomedullary c-AMP result from a- or ß-adrenergic activation of adenylate cyclase by catecholamines liberated from adrenal chromaffin cells and whether a causal relationship exists between these changes and subsequent induction of tyrosine hydroxylase (TH).

The uptake and retrograde transport of noradrenaline (NA) within the axons of sympathetic neurons was investigated in an in vitro system. Dissociated neurons from the sympathetic ganglia of newborn rats were cultured for 3-6 wk in the absence of non-neuronal cells in a culture dish divided into three chambers. These allowed separate access to the axonal networks and to their cell bodies of origin. [3H]NA (0.5 X 10(-6) M), added to the axon chambers, was taken up by the desmethylimipramine- and cocaine-sensitive neuronal amine uptake mechanisms, and a substantial part was rapidly transported retrogradely along the axons to the nerve cell bodies. This transport was blocked by vinblastine or colchicine. In contrast with the storage of [3H]NA in the axonal varicosities, which was totally prevented by reserpine (a drug that selectively inactivates the uptake of NA into adrenergic storage vesicles), the retrograde transport of [3H]NA was only slightly diminished by reserpine pretreatment. Electron microscopic localization of the NA analogue 5-hydroxydopamine (5-OHDA) indicated that mainly large dense-core vesicles (700-1,200-A diam) are the transport compartment involved. Whereas the majority of small and large vesicles lost their amine dense-core and were resistant to this drug. It, therefore, seems that these vesicles maintained the amine uptake and storage mechanisms characteristic for adrenergic vesicles, but have lost the sensitivity of their amine carrier for reserpine. The retrograde transport of NA and 5-OHDA probably reflects the return of used synaptic vesicle membrane to the cell body in a form that is distinct from the membranous cisternae and prelysosomal structures involved in the retrograde axonal transport of extracellular tracers.

Brain-derived neurotrophic factor (BDNF), a member of the nerve growth factor (NGF) gene family, has been shown to influence the survival and differentiation of specific classes of neurons in vitro and in vivo. The possibility that neurotrophins are also involved in processes of neuronal plasticity has only recently begun to receive attention. To determine whether BDNF has a function in processes such as long-term potentiation (LTP), we produced a strain of mice with a deletion in the coding sequence of the BDNF gene. We then used hippocampal slices from these mice to investigate whether LTP was affected by this mutation. Homo- and heterozygous mutant mice showed significantly reduced LTP in the CA1 region of the hippocampus. The magnitude of the potentiation, as well as the percentage of cases in which LTP could be induced successfully, was clearly affected. According to the criteria tested, important pharmacological, anatomical, and morphological parameters in the hippocampus of these animals appear to be normal. These results suggest that BDNF might have a functional role in the expression of LTP in the hippocampus.

We have studied the regulation of tyrosine hydroxylase by cell-cell contact in primary cultures of bovine adrenal chromaffin cells. Preparation of dissociated chromaffin cells from bovine adrenal medullae or the harvesting of cultured cells resulted in a rapid decrease of the specific mRNA(TH) (defined as the amount of mRNA(TH) (where TH represents tyrosine hydroxylase) per microgram of total RNA). The decrease in mRNA(TH) levels appears to be stimulated by the loss of cell contact, as it occurs much more rapidly than would be expected from the turnover rate of mRNA(TH) in cultured cells. Similarly an enhanced rate of tyrosine hydroxylase enzyme degradation was observed in chromaffin cells when brought into low contact cultures. In contrast to the decrease in mRNA(TH) levels, however, the decrease in tyrosine hydroxylase enzyme levels was not so rapid and could be prevented in high contact cultures. The mRNA(TH) increased 4-fold in high contact cultures relative to dissociated cells within 1 day after plating. Similarly the rate of synthesis of tyrosine hydroxylase enzyme molecules was maximal after 1 day, although the increase in the absolute amount of tyrosine hydroxylase occurred only slowly. The increase in specific mRNA(TH) by cell contact was inhibited by alpha-amanitin, indicating that cell contact evokes an increase in tyrosine hydroxylase levels by increasing the transcription of mRNA(TH), in addition to inhibiting the degradation of tyrosine hydroxylase molecules.

In the central nervous system brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF) are predominantly located in neurons. Here we demonstrate that the balance between the activity of the glutamatergic and gamma-aminobutyric acid (GABA)ergic systems controls the physiological levels of BDNF and NGF mRNAs in hippocampal neurons in vitro and in vivo. The blockade of the glutamate receptors and/or stimulation of the GABAergic system reduces BDNF and NGF mRNAs in hippocampus and NGF protein in hippocampus and septum. The reduction of NGF in the septum reflects the diminished availability of NGF in the projection field of NGF-dependent septal cholinergic neurons. These neurons do not synthesize NGF themselves but accumulate it by retrograde axonal transport. The refined and rapid regulation of BDNF and NGF synthesis by the glutamate and GABA transmitter systems suggests that BDNF and NGF might be involved in activity-dependent synaptic plasticity.