Role of nerve growth factor in the development of rat sympathetic neurons in vitro. I. Survival, growth, and differentiation of catecholamine production.

To study the effect of nerve growth factor (NGF) on neuronal survival, growth, and differentiation, cultures of dissociated neonatal rat sympathetic neurons virtually free of other cell types were maintained for 3-4 wk. In the absence of NGF, the neurons did not survive for more than a day. Increased levels of NGF increased neuronal survival and growth (total protein and total lipid phosphate); saturation occurred at 0.5 microgram/ml 7S NGF. Neuronal differentiation examined by measuring catecholamine (CA) production from tyrosine also depended on the level of NGF in the culture medium. As the NGF concentration was raised, CA production per neuron, per nanogram protein, or per picomole lipid phosphate increased until saturation was achieved between 1 and 5 microgram/ml 7S NGF. Thus, NGF induces neuronal survival, growth, and differentiation of CA production in a dose-dependent fashion. Neuronal growth and differentiation were quantitatively compared in the presence of the high and low molecular weight forms of NGF; no significant functional differences were found.

Role of nerve growth factor in the development of rat sympathetic neurons in vitro. II. Developmental studies.

Adrenergic sympathetic neurons were grown for 4 wk in submaximal and saturating concentrations of nerve growth factor (NGF) in the virtual absence of non-neuronal cells. In 0.2 or 5 microgram/ml 7S NGF, the neurons gradually decreased in number during the first week, although fewer neurons died at the higher level. No significant change in cell number was observed thereafter. Total neuronal protein, a measure of cell growth, increased linearly with age in both concentrations of NGF. At each age, neurons in high NGF exhibited greater growth per cell than those in low NGF. The ability of neurons to produce catecholamine (CA) increased dramatically during the second and third weeks in both concentrations of NGF, and along a similar time-course, although neurons in submaximal NGF developed a lesser capacity for CA production. As neurons developed in culture, they became less dependent on NGF for survival and CA production, but even in older cultures, approximately 50% of the neurons died when NGF was withdrawn.

Role of nerve growth factor in the development of rat sympathetic neurons in vitro. III. Effect on acetylcholine production.

The effect of nerve growth factor (NGF) on the development of cholinergic sympathetic neurons was studied in cultures grown either on monolayers of dissociated rat heart cells or in medium conditioned by them. In the presence of rat heart cells the absolute requirement of neurons for exogenous NGF was partially spared. The ability of heart cells to support neuronal survival was due at least in part to production of a diffusable NGF-like substance into the medium. Although some neurons survived on the heart cell monolayer without added NGF, increased levels of exogenous NGF increased neuronal survival until saturation was achieved at 0.5 microgram/ml 7S NGF. The ability of neurons to produce acetylcholine (ACh) from choline was also dependent on the level of exogenous NGF. In mixed neuron-heart cell cultures, NGF increased both ACh and catecholamine (CA) production per neuron to the same extent; saturation occurred at 1 microgram/ml 7S NGF. As cholinergic neurons developed in culture, they became less dependent on NGF for survival and ACh production, but even in older cultures approximately 40% of the neurons died when NGF was withdrawn. Thus, NGF is as necessary for survival, growth, and differentiation of sympathetic neurons when the neurons express cholinergic functions as when the neurons express adrenergic functions (4, 5).

Glial and neuronal forms of the voltage-dependent sodium channel: characteristics and cell-type distribution.

Two functionally different forms of the voltage-dependent sodium channel were observed in glia and in neurons of the mammalian nervous system. Both forms had identical conductance and tetrodotoxin sensitivity and displayed steady-state inactivation, a strongly voltage-dependent rate of activation, and a faster but weakly voltage-sensitive rate of inactivation. However, the glial form had significantly slower kinetics and a more negative voltage dependence, suggesting that it was functionally specialized for glia. This form was found in most glial types studied, while the neuronal form was observed in retinal ganglion cells, cortical motor neurons, and O2A glial progenitor cells. Both forms occurred in type-2 astrocytes. The presence of the glial form correlated with the RAN-2 surface antigen.

Immunological, morphological, and electrophysiological variation among retinal ganglion cells purified by panning.

Two different monoclonal antibodies to the Thy-1 antigen, T11D7 and 2G12, were used to purify and characterize retinal ganglion cells from postnatal rat retina. Although Thy-1 has been reported to be a specific marker for ganglion cells in retina, retinal cell suspensions contained several other types of Thy-1-positive cells as well. Nevertheless, a simple two-step "panning" procedure allowed isolation of ganglion cells to nearly 100% purity. We found that postnatal ganglion cells differed in antigenic, morphological, and intrinsic electrophysiological characteristics, and that these properties were correlated with one another. Minor variations of this panning protocol should allow rapid, high yield purification to homogeneity of many other neuronal and glial cell types.

Ion channel expression by white matter glia: the type-1 astrocyte.

We describe the electrophysiological properties of acutely isolated type-1 astrocytes using a new "tissue print" dissociation procedure. Because the enzymes used did not destroy or modify the ion channels, and the cells retained many processes, the properties may reflect those in vivo. The types of ion channels in type-1 astrocytes changed rapidly during the first 10 postnatal days, when they attained their adult phenotype. This change was dependent on the presence of neurons. In culture, most of these channel types were not expressed, but a phenotype more typical of that in vivo could be induced by co-culture with neurons. The electrophysiological properties of astrocytes make some existing hypotheses of astrocyte function less likely.

Ion channel expression by white matter glia: the O-2A glial progenitor cell.

We describe electrophysiological properties of the O-2A glial progenitor cell in a new serum-free culture system. O-2A progenitors have many properties characteristic of neurons: they have glutamate-activated ion channels, express the neuronal form of the sodium channel, fire single regenerative potentials, and synthesize the neurotransmitter GABA by an alternative synthetic pathway. Nearly identical properties were observed in acutely isolated O-2A progenitors, indicating that this phenotype is not an artifact of culture. The O-2A did not express a simple subset of channel types found in its descendant cells, the type-2 astrocyte and oligodendrocyte, studied in the same culture system. During development, these electrophysiological properties may contribute to O-2A function in vivo.

The murine cone photoreceptor: a single cone type expresses both S and M opsins with retinal spatial patterning.

Mice express S and M opsins that form visual pigments for the detection of light and visual signaling in cones. Here, we show that S opsin transcription is higher than that of M opsin, which supports ultraviolet (UV) sensitivity greater than midwavelength sensitivity. Surprisingly, most cones coexpress both S and M opsins in a common cone cell type throughout the retina. All cones express M opsin, but the levels are graded from dorsal to ventral. The levels of S opsin are relatively constant. However, in the far dorsal retina, S opsin is repressed stochastically, such that some cones express M opsin only. These observations indicate that two different mechanisms control M and S opsin expression. We suggest that a common cone type is patterned across the retinal surface to produce phenotypic cone subtypes.

Developmental study of Müller cells in the rat retina using a new monoclonal antibody, RT10F7.

We produced the monoclonal antibody RT10F7, characterized its antigenic specificity and expression in the adult and developing retina, in cultured retinal cells and in other parts of the central nervous system. In metabolically-labelled retinal cultures RT10F7 immunoprecipitated a protein of approximately 36,000 mol. wt. In the adult, RT10F7 stained endfeet of Müller cells in the ganglion cell layer, four horizontal bands in the inner plexiform layer, and radial fibres in the outer plexiform layer which terminated at the outer limiting membrane. In the inner nuclear layer, most somata were underlined by Müller processes that wrapped around them, but some cell bodies were immunoreactive for RT10F7 in the cytoplasm. During development, postnatal day 21 was the first age at which the adult pattern of immunoreactivity was present, although a fourth band in the inner plexiform layer was less clear than for the adult. By 14 and eight days after birth, the pattern of RT10F7 immunoreactivity approximated that of the adult; however, only three bands and one band were present, respectively, in the inner plexiform layer. At earlier ages, postnatal days 4, 1 and embryonic ages 19 and 15, the monoclonal antibody stained Müller cell endfeet and radial fibres, from the inner plexiform layer through the neuroblastic layer to the outer limiting membrane. At these ages, the immunoreactivity was more prominent at the level of Müller cell endfeet. The monoclonal antibody stained glia in preparations of dissociated retinal cells maintained in culture but not astrocytes or oligodendrocytes from optic nerve cultures. In brain sections, tanycytes exhibited RT10F7 immunoreactivity. The monoclonal antibody RT10F7 recognized a specific cell type in the retina, the Müller cell. In the adult and developing retina, RT10F7 recognized an antigen that is present primarily in Müller cell processes. This feature allowed us to follow the maturation of the Müller cell and correlate it with developmental events in the retina. RT10F7 is a specific marker for Müller cells in vivo and in vitro and may be useful for studies of function of Müller cells after ablation or after injuries that are known to activate Müller cells.

Monoclonal antibodies with defined specificities for Torpedo nicotinic acetylcholine receptor cross-react with Drosophila neural tissue.

A panel of monoclonal antibodies with known specificity for the well-characterized nicotinic acetylcholine receptor from the electroplax of Torpedo californica, many of which cross-react with the mammalian muscle acetylcholine receptor, were examined for cross-reactivity in the fly, Drosophila melanogaster. Monoclonal antibodies with specificities for different epitopes on the transmembrane receptor complex from Torpedo cross-react with different regional subsets of neural tissue in Drosophila. Axonal tracts, neuropil, mechano-sensory bristle elements and photoreceptors, each are detected by separate monoclonal antibody classes corresponding to different epitope domains. A preliminary characterization of an antigenic determinant in Drosophila heads recognized by one of the cross-reacting monoclonal antibodies is presented. Monoclonal antibodies such as these may be useful in identifying molecules of homologous structure or function, possibly including a neuronal acetylcholine receptor.

The influence of non-neuronal cells on catecholamine and acetylcholine synthesis and accumulation in cultures of dissociated sympathetic neurons.

The effects of several non-neuronal cell types on neurotransmitter synthesis in cultures of dissociated sympathetic neurons from the new-born rat were studied. Acetylcholine synthesis from radioactive choline was increased 100- to 1000-fold in the presence of non-neuronal cells from sympathetic ganglia. This increase was roughly dependent on the number of ganglionic non-neuronal cells present. The effect did not appear to be due to an increased plating efficiency of neurons, since the non-neuronal cells were capable of increasing acetylcholine synthesis after only 48-hr contact with neurons that had been previously grown without non-neuronal cells for 2 weeks. C6 rat glioma cells were also able to stimulate acetylcholine synthesis, but 3T3 mouse fibroblast cells had little or no effect. None of the non-neuronal cell types synthesized detectable acetylcholine in the absence of the neurons. The ganglionic non-neuronal cells had no significant effect on catecholamine synthesis (which occurs in the absence of non-neuronal cells).

Immunological function of the blood-cerebrospinal fluid barrier.

Because the brain lacks a true lymphatic system, it is unclear how peripheral lymphocytes recognize foreign antigens present in the central nervous system. This report demonstrates that the choroid plexus, which constitutes the blood-cerebrospinal fluid barrier, is able to present foreign antigen to, and stimulate the proliferation of, peripheral helper T lymphocytes through an Ia-dependent, major histocompatibility complex-restricted mechanism. Furthermore, in vivo, choroid plexus epithelial cells have access to, and are capable of taking up, virus-sized particles injected elsewhere into the cerebrospinal fluid. Thus these data suggest that the blood-cerebrospinal fluid barrier may play a role in immunological communication between the central nervous system and periphery, a function relevant to the initiation of immunological responses to central nervous system infections and autoimmune processes and for the surveillance of tumor cells in the cerebrospinal fluid.

Ion channel expression by white matter glia: I. Type 2 astrocytes and oligodendrocytes.

White matter is a compact structure consisting primarily of neuronal axons and glial cells. As in other parts of the nervous system, the function of glial cells in white matter is poorly understood. We have explored the electrophysiological properties of two types of glial cells found predominantly in white matter: type 2 astrocytes and oligodendrocytes. Whole-cells and single-channel patch-clamp techniques were used to study these cell types in postnatal rat optic nerve cultures prepared according to the procedures of Raff et al. (Nature, 303:390-396, 1983b). Type 2 astrocytes in culture exhibit a "neuronal" channel phenotype, expressing at least six distinct ion channel types. With whole-cell recording we observed three inward currents: a voltage-sensitive sodium current qualitatively similar to that found in neurons and both transient and sustained calcium currents. In addition, type 2 astrocytes had two components of outward current: a delayed potassium current which activated at 0 mV and an inactivating calcium-dependent potassium current which activated at -30 mV. Type 2 astrocytes in culture could be induced to fire single regenerative potentials in response to injections of depolarizing current. Single-channel recording demonstrated the presence of an outwardly rectifying chloride channel in both type 2 astrocytes and oligodendrocytes, but this channel could only be observed in excised patches. Oligodendrocytes expressed only one other current: an inwardly rectifying potassium current that is mediated by 30- and 120-pS channels. Because these channels preferentially conduct potassium from outside to inside the cell, and because they are open at the resting potential of the cell, they would be appropriate for removing potassium from the extracellular space; thus it is proposed that oligodendrocytes, besides myelinating axons, play an important role in potassium regulation in white matter. The conductances present in oligodendrocytes suggest a "modulated Boyle and Conway mechanism" of potassium accumulation.

Preliminary studies on the use of monoclonal antibodies as probes for sympathetic development.

The precise structural organization and proper functioning of the adult nervous system depend on the ability of neurones to make highly ordered synaptic connexions. To define molecules involved in the development of these connexions and to study their functional roles, we use primary cultures of dissociated rat sympathetic neurones grown in the virtual absence of non-neuronal cells. These neurones can develop adrenergic or cholinergic properties, depending on the environment in which they are grown. This ability to manipulate neuronal phenotype is being used in an attempt to identify cell surface macromolecules that are important in the development or function of adrenergic and cholinergic properties. We have produced monoclonal antibodies against the surface membranes of these neurones and are in the process of characterizing them. Results are presented on the binding specificity of one of these antibodies and on the effect of two other antibodies on neurotransmitter synthesis, uptake, and release.

Calcium current in cortical astrocytes: induction by cAMP and neurotransmitters and permissive effect of serum factors.

Voltage-dependent L-type calcium channels were induced in highly purified, cultured, cortical astrocytes by exposure to substances known to increase their intracellular cAMP: 8-bromo-cAMP, forskolin, isoproterenol, and vasoactive intestinal peptide. In untreated control cultures, L-type calcium currents were entirely absent. The induction of this calcium current was specific to cortical astrocytes and a closely related astrocyte type in white matter and did not occur in meningeal cells or oligodendrocytes. The ability of forskolin to induce L-type calcium current in astrocytes depended on previous culture of these cells in a permissive lot of serum for at least 48 hr. In addition, certain lots of sera caused expression of a voltage-dependent T-type calcium channel in untreated control cultures. Several possible mechanisms of calcium channel induction by cAMP are consistent with our data: modification of a silent channel already present in the membrane, indirect effect of cytoskeletal alteration, or insertion of channels from submembrane stores. If the environment in vivo is permissive, regulation of glial ion channels may be under neuronal control.

Immunohistochemical localization of a neuronal nicotinic acetylcholine receptor in mammalian brain.

A monoclonal antibody generated against purified acetylcholine receptor from Torpedo electric organ was used to immunohistochemically localize a neuronal nicotinic acetylcholine receptor. Regions of the rat brain stained with this antibody paralleled those areas of the brain exhibiting [3H]nicotine binding sites and corresponded to areas in which mRNAs encoding for alpha subunits of the neuronal nicotinic acetylcholine receptor are present. Thus, the anteroventral thalamus, cortex, hippocampus, medial habenula, interpeduncular nucleus, and substantia nigra/ventral tegmental area exhibited significant immunoreactivity. Neurons of the medial habenula and substantia nigra were densely stained, and processes were prominently delineated. Furthermore, in the projection areas of the medial habenula (interpeduncular nucleus and median raphe) axons were strongly immunoreactive and were distributed to distinct subdivisions of the target sites. The present data suggest that there are several discrete neuronal systems in which nicotinic acetylcholine receptors have functional importance. These immunohistochemical studies delineate at the single-cell level the localization within the mammalian central nervous system of certain nicotinic acetylcholine receptors.