Changes in responsiveness to extracellular ATP in chick skeletal muscle during development and upon denervation

Skeletal muscles in developing chick embryos were tested for responsiveness to adenosine 5'-triphosphate (ATP), a substance known to depolarize chick skeletal muscle in culture. The sensitivity to extracellular ATP was tested at various stages of development in five different muscles; pectoralis superficia, anterior latissimus dorsi, posterior latissimus dorsi, sartorious, and gastrocnemius. At the earliest time that muscles were tested (Embryonic Day 6, stage 30 of Hamburger and Hamilton, 1951) application of ATP(50-100 microM) elicited vigorous contractions in all five muscles, but within a few days (Embryonic Day 17, stage 43) none of the muscles contracted in response to ATP. Sensitivity declined at approximately the same time in all five of these muscles. Intracellular recordings made from muscle fibers near the time of hatching (Embryonic Days 18-21 or Postnatal Days 1-2) indicated that the loss of the ability to contract in response to ATP was due to the total loss of responsiveness to ATP. Surgical denervation of the anterior latissimus dorsi and posterior latissimus dorsi was performed in a series of chicks 1-2 days after hatching, and the ability of these muscles to contract in response to ATP was tested 3-10 days after the surgery. Contractions in response to ATP were present in many of the muscles. Thus denervation of muscles in newly hatched chicks led to the reappearance of sensitivity to ATP. The disappearance of ATP responsiveness shortly after muscles become innervated and the reappearance of ATP responsiveness following denervation suggest that the expression of ATP responsiveness is regulated by motor neurons.

Dihydropyridines and verapamil inhibit voltage-dependent K+ current in isolated outer hair cells of the guinea pig

Dihydropyridines and verapamil are widely used as blockers of voltage-dependent Ca++ channels. In this work we show that these compounds can have a direct blocking action on a class of voltage-activated potassium channels. Voltage-dependent whole-cell currents were recorded from isolated guinea-pig outer hair cells (OHCs) under conditions such that the free Ca++ concentration in both the internal and external solutions was minimized. A substantial Ca(++)-independent K+ current was revealed by this procedure. Both conventional K+ and Ca++ channel ligands inhibited this current. The order of potency (in terms of the half inhibitory concentrations (IC50) of channel inhibitors) was: nimodipine (6 microM) > Bay K 8644 (8 microM) > verapamil (11 microM) > 4-aminopyridine (22 microM) > nifedipine (32 microM) > quinine (49 microM) > TEA (10236 microM). Except for verapamil, these channel ligands reduced the size of the K+ currents without much alteration of the time course of the currents. In contrast, verapamil caused a more than 10-fold increase in the apparent inactivation rate of the K+ currents without significantly altering the activation of the currents. The observation that relatively low concentrations of calcium channel ligands can directly inhibit potassium currents in isolated OHCs indicates that caution should be taken when these pharmacological agents are used as tools for studying cochlear hair cell physiology.

Contact with myelin evokes a release of calcium from internal stores in neonatal rat oligodendrocytes in vitro

The response of neonatal rat oligodendrocytes to contact with myelin extracts prepared from the central and peripheral nervous system was examined. Contact with either CNS myelin or PNS myelin resulted in collapse of the fine structure of the leading edge of oligodendrocytes in vitro. The collapse of the fine structure of oligodendrocyte processes was preceded by a substantial (approximately fivefold) increase in intracellular free calcium concentration. The calcium concentration increase was due, at least in part, to a release of calcium from internal stores, since it persisted when extracellular calcium was removed by chelation with EGTA. The increase in calcium concentration and the coincident morphological change suggest that oligodendrocytes might be able to recognize and react to specific molecules on the surface of other oligodendrocytes.

Locus coeruleus neuron growth cones and spinal cord regeneration

One of the major impediments to successful recovery of function after a spinal cord injury is thought to be the reaction of the neuronal growth cone to inhibitory influences in the local environment in or around the site of the injury. The growth cones of locus coeruleus neurons studied in vitro collapsed upon contact with an extract of CNS myelin but did not collapse on contact with an extract of PNS myelin. Coincident with the collapse of the growth cone, was an increase in internal free calcium concentration that was predominantly the result of an influx of calcium through channels in the plasma membrane. Omega-conotoxin, which specifically blocks N-type voltage-gated calcium channels, blocked both the myelin-induced calcium influx and the coincident collapse of the growth cone.

Omega-conotoxin prevents myelin-evoked growth cone collapse in neonatal rat locus coeruleus neurons in vitro

The response of neonatal rat locus coeruleus neurons to contact with myelin extracts prepared from the CNS and PNS was examined. The growth cones of these neurons collapsed following contact with central myelin, but continued to elongate on contact with peripheral myelin. Central myelin elicited an increase in the intracellular free calcium concentration in these growth cones, while peripheral myelin did not. This increase appeared to require transmembrane calcium flux, since it was blocked by extracellular EGTA, and also by extracellular cobalt. These neurons express N- and L-type calcium channels. Exposure to 5 microM omega-conotoxin GVIA, a specific blocker of N-type channels, prevented both the myelin-evoked increase in growth cone calcium concentration and the collapse of growth cones.

Voltage-dependent block by neomycin of the ATP-induced whole cell current of guinea-pig outer hair cells

1. The effects of externally applied ATP and neomycin on whole cell currents of isolated guinea pig cochlear outer hair cells (OHCs) were studied using the whole cell voltage-clamp technique. In OCHs held at -70 mV, ATP activated a large inward current. In the presence of neomycin, the ATP-induced whole cell current activated along a relatively unaltered time course, but the current then decreased to a reduced steady level. The neomycin inhibition of the ATP-induced current was dose dependent. The half-inhibitory concentration (IC50) of neomycin measured at steady state was estimated to be 90 microM. 2. Neomycin inhibition of the ATP response could not be reversed by increasing the concentration of ATP, indicating that the effect was noncompetitive. The inhibition was voltage dependent and was greatly reduced when OHCs were held at positive potentials. 3. Cells treated with 100 microM ATP gave maximal current responses. Addition of neomycin substantially increased membrane current noise of the 100 microM ATP responses. When neomycin concentration was varied from 10 to 500 microM, the current noise level peaked between 50 and 100 microM. The noise increase was observed at negative holding potentials but not at positive potentials. 4. The neomycin-induced whole cell current noise was used to estimate the size of the underlying elementary current. The ATP-induced single channel current of OHCs at -70 mV was estimated to be approximately 0.3 pA. The number of ATP-activated channels in a single OHC was estimated to be in the range of a few thousand. 5. The characteristics of the neomycin inhibition of ATP-induced currents were consistent with an open channel blocking mechanism. Analysis of the voltage dependence of the steady state neomycin inhibition suggested a neomycin binding site at an electrical distance of 0.3 from the extracellular side.

Single potassium channel currents activated by extracellular ATP in developing chick skeletal muscle: a role for second messengers

1. In developing chick skeletal muscle, extracellular ATP activates an early excitatory current and a delayed potassium current. Previous work had shown that the potassium current elicited by ATP is sensitive to temperature and activates with a delay of nearly 1 s, suggesting that a second messenger is involved. The existence of a second messenger was confirmed by the observation that single potassium channels were activated in cell-attached patches, when ATP was applied outside of the patch pipette. 2. Two classes of ATP-activated potassium-channel currents were observed in cell-attached patches: one had a slope conductance of 23 pS, whereas the other had a slope conductance of 51 pS. 3. Pharmacological manipulations suggested that activation of the whole-cell potassium current by ATP did not require cyclic adenosine monophosphate (cAMP), cyclic guanosine monophosphate (cGMP), inositol 1,4,5-triphosphate (IP3), nitric oxide, or a rise in internal free calcium. Additional pharmacological experiments suggested that activation of the whole-cell potassium current might not require activation of a G protein and probably did not involve intracellular protein phosphorylation. 4. The ability of arachidonic acid and its metabolites to activate potassium channels in chick skeletal muscle was also tested. Arachidonic acid, several prostaglandins and several leukotrienes activated whole-cell potassium currents. However, results with several inhibitors suggested that arachidonic acid and its metabolites are not necessary for activation of the whole-cell potassium current by ATP. 5. In excised outside-out membrane patches, ATP activated a single class of potassium channels. The slope conductance of these channels indicated that they are likely to be identical to the smaller of the two classes of second messenger activated potassium channels observed in cell-attached patches. 6. The observation that the larger class of potassium channels observed in cell-attached patches was absent in excised patches suggests that activation of these channels by ATP requires a cytosolic factor that is easily dialyzed away. In contrast, the observation that the smaller class of potassium channels could still be activated by ATP in excised patches suggests that the two classes of potassium channels are activated by different mechanisms. These results also indicate that all the molecules involved in coupling ATP receptor activation to opening of the smaller class of potassium channels remain closely associated with an excised patch. One possible explanation is that there might be an intramembranous second messenger.

Structural determinants of barium permeation and rectification in non-NMDA glutamate receptor channels

A single site in recombinant glutamate receptor channels of the GluR1-GluR4 family has been previously identified as a key regulator of ion permeation. The natural amino acid at this position (arginine in GluR2 but glutamine in GluR1, GluR3, and GluR4) determines both the ability to pass outward current and the divalent cation permeability of kainate-activated receptor channels. By mutagenesis of GluR6, we demonstrated that the same site also controls the ability to pass outward current in another non-NMDA receptor family. Additional mutations at and near this site in GluR3 indicated that the position of the arginine is critical to function, that the ability to pass outward current is not necessarily linked to low barium permeability, and that the size as well as the charge of the side chain at this position influences barium permeation. These results provide evidence that this site forms part of the selectivity filter of glutamate receptor channels.

Identification of a site in glutamate receptor subunits that controls calcium permeability

The neurotransmitter glutamate mediates excitatory synaptic transmission throughout the brain. A family of genes encoding subunits of the non-N-methyl-D-aspartate (non-NMDA) type of glutamate receptor has been cloned. Some combinations of these subunits assemble into receptors with a substantial permeability to calcium, whereas others do not. To investigate the structural features that control ion permeation through these ligand-gated channels, mutant receptor subunits with single-amino acid changes were constructed. Mutation of a certain amino acid that results in a net charge change (from glutamine to arginine or vice versa) alters both the current-voltage relation and the calcium permeability of non-NMDA receptors. A site has thus been identified that regulates the permeation properties of these glutamate receptors.

A receptor that is highly specific for extracellular ATP in developing chick skeletal muscle in vitro

1. Extracellular adenosine 5'-triphosphate (ATP) activated an early excitatory conductance followed by a late potassium conductance in developing chick skeletal muscle. A series of ATP analogues were tested for their ability to activate these two conductances. All compounds tested were either agonists for both responses or for neither. Furthermore, the potency of agonists was similar for the two responses. 2. The order of potency for agonists was ATP approximately adenosine 5'-O-(3-thiotriphosphate) (ATP-gamma-S) approximately 2-methylthio-ATP (2-CH3S-ATP) greater than 2'-deoxy-ATP approximately 3'-deoxy-ATP greater than adenosine 5'-tetraphosphate (ATP-OPO3) approximately adenosine 5'-diphosphate (ADP). Many other ATP analogues were not agonists. 3. Activation of the excitatory response did not require divalent cations. Furthermore, the concentration-response relation of the excitatory response was similar when ATP was applied as the free anion of ATP (ATP4-) or complexed with a divalent cation (M.ATP2-). 4. Three antagonists of the ATP response were characterized. 8-Br-ATP was a weak antagonist, while 2',3'-dialdehyde-ATP and DIDS (4,4'-diisocyanatostilbene-2,2'-disulphonic acid) were potent irreversible inhibitors. The two conductances were equally affected by these antagonists. 5. These results suggest that both ATP responses are activated through the same receptor type, or two very similar receptors.

Physiological properties of newly formed synapses between sympathetic preganglionic neurons and sympathetic ganglion neurons

We have examined the physiological properties of transmission at newly formed synapses between sympathetic preganglionic neurons and sympathetic ganglion neurons in vitro. Chick neurons were labeled with fluorescent carbocyanine dyes before they were placed into culture (Honig and Hume, 1986), and were studied by making intracellular recordings during the first 2 weeks of coculture. Evoked monosynaptic excitatory postsynaptic potentials (EPSPs) were not observed until 48 h of coculture. Beyond this time, the frequency with which connected pairs could be found did not vary greatly with time. With repetitive stimulation, the evoked monosynaptic EPSPs fluctuated in amplitude from trial to trial and showed depression at frequencies as low as 1 Hz. To gain further information about the quantitative properties of transmission at newly formed synapses, we analyzed the pattern of fluctuations of delayed release EPSPs. In mature systems, delayed release EPSPs are known to represent responses to single quanta, or to the synchronous release of a small number of quanta. For more than half of the connections we studied, the histograms of delayed release EPSPs were extremely broad. This result suggested that either quantal responses are drawn from a continuous distribution that has a large coefficient of variation or that there are several distinct size classes of quantal responses. The pattern of fluctuations of monosynaptic EPSPs was consistent with both of these possibilities, and was inconsistent with the possibility that monosynaptic EPSPs are composed of quantal subunits with very little intrinsic variation. Although variation in the size of responses to single quanta might arise in a number of ways, one attractive explanation for our results is that the density and type of acetylcholine receptors varies among the different synaptic sites on the surface of developing sympathetic ganglion neurons.

Cell interactions regulate dendritic morphology and responses to neurotransmitters in embryonic chick sympathetic preganglionic neurons in vitro

The influence of non-neuronal cells and interneurons on the morphological development of chick sympathetic preganglionic neurons (SPNs) and on the responsiveness of these neurons to the neurotransmitters GABA, glycine, and glutamate was studied. SPNs were retrogradely labeled with the fluorescent dyes dil and diO, then separated from spinal-cord non-neuronal cells and interneurons by fluorescence-activated cell sorting. SPNs were grown in culture, either alone or in coculture with non-neuronal cells alone, with interneurons alone, or with both of these cell types (control cultures). The responsiveness of SPNs to neurotransmitters was assessed by whole-cell recording, while cell morphology was assessed after intracellular staining with 6-carboxyfluorescein. Cell size and morphology were affected by non-neuronal cells. In the absence of non-neuronal cells, SPNs had smaller cell bodies and fewer major processes, whether or not interneurons were present. In contrast, responses to the 3 neurotransmitters were affected by both non-neuronal cells and interneurons, but in ways that differed slightly for each transmitter. In the absence of both non-neuronal cells and interneurons, responses to all 3 transmitters were much smaller than in control cultures, with responses to glutamate most profoundly affected. The addition of either non-neuronal cells or interneurons slightly increased the amplitude of SPN responses to glutamate, but the level of responsiveness with either cell type alone was much lower than for SPNs grown in the presence of both cell types. The addition of interneurons also slightly increased the responsiveness of SPNs to GABA, but non-neuronal cells alone had no significant effect on the responses of SPNs to GABA. Finally, the glycine responsiveness of SPNs was raised to control levels when either non-neuronal cells or interneurons were added. These experiments demonstrate that, though interneurons can have a significant inductive effect on the responses of SPNs to neurotransmitters, not all of the changes in neurotransmitter responsiveness can be related to the formation of functional synapses.

Irreversible desensitization of ATP responses in developing chick skeletal muscle

1. In developing chick skeletal muscle, extracellular adenosine 5'-triphosphate (ATP) elicits an early excitatory conductance increase followed by a late potassium conductance increase. Both of these responses desensitize profoundly. Intracellular recordings and whole-cell voltage-clamp recordings were made in order to examine the time course and mechanism of desensitization and the recovery from desensitization. 2. Most of the loss of responsiveness to ATP occurred during the first minute of exposure to ATP. For the excitatory conductance, the loss of responsiveness to ATP resulted in part from long-lasting activation of the ATP-sensitive channels and in part from entrance into an inactive (non-conducting) state. In contrast, desensitization of the potassium conductance was entirely the result of a relatively fast transition to an inactive state. 3. Recovery from desensitization took many hours for both responses and was quite sensitive to temperature. 4. Recovery from desensitization for both responses was prevented by preincubation with the glycosylation inhibitor, tunicamycin. Several lines of evidence suggest that tunicamycin treatment blocked the delivery of new ATP receptors to the cell surface. 5. The recovery of the early response to ATP following exposure to two non-competitive inhibitors of the ATP response was also examined. These two compounds are thought to covalently modify the receptor. After exposure to either of these inhibitors, responsiveness to ATP returned over a time course that was similar to the time course of recovery from desensitization. 6. These results indicate that, following activation, ATP receptors do not become available for reactivation, and that recovery from desensitization is due to the insertion of newly synthesized receptors into the plasma membrane.

Growth cones of chick sympathetic preganglionic neurons in vitro interact with other neurons in a cell-specific manner

The ability of the growth cones of sympathetic preganglionic neurons to recognize the neurons they encounter during their outgrowth and to react to them in a cell-type-specific manner may play a role in guiding them to appropriate targets during development in vivo. In this study, we examined the in vitro growth of sympathetic preganglionic neurons as they interacted with motor neurons, dorsal root ganglion neurons, and sympathetic ganglion neurons. All of these cell types might potentially be encountered by a growing preganglionic axon. The interaction of sympathetic preganglionic growth cones with each cell type was distinct. Sympathetic preganglionic growth cones fasciculated on motor-neuron neurites, collapsed after contact with the cell bodies and neurites of dorsal root ganglion neurons, and grew across the cell bodies and neurites of sympathetic ganglion neurons. These cell-type-specific responses stand in contrast to the collapse and retraction reported to be the most common growth-cone behaviors that result from contact between central and peripheral neurons in vitro and suggest that contact-mediated recognition might be sufficient for growth to and interaction with appropriate targets.

Permeation of both cations and anions through a single class of ATP-activated ion channels in developing chick skeletal muscle

Micromolar concentrations of extracellular adenosine 5'-triphosphate (ATP) elicit a rapid excitatory response in developing chick skeletal muscle. Excitation is the result of a simultaneous increase in membrane permeability to sodium, potassium, and chloride ions. In the present study we quantify the selectivity of the ATP response, and provide evidence that a single class of ATP-activated ion channels conducts both cations and anions. Experiments were performed on myoballs using the whole-cell patch-clamp technique. We estimated permeability ratios by measuring the shift in reversal potential when one ion was substituted for another. We found that monovalent cations, divalent cations, and monovalent anions all permeate the membrane during the ATP response, and that there was only moderate selectivity between many of these ions. Calcium was the most permeant ion tested. To determine if ATP activates a single class of channels that conducts both cations and anions, or if ATP activates separate classes of cation and anion channels, we analyzed the fluctuations about the mean current induced by ATP. Ionic conditions were arranged so that the reversal potential for cations was +50 mV and the reversal potential for anions was -50 mV. Under these conditions, if ATP activates a single class of channels, ATP should not evoke an increase in noise at the reversal potential of the ATP current. However, if ATP activates separate classes of cation and anion channels, ATP should evoke a significant increase in noise at the reversal potential of the ATP current. At both +40 and -50 mV ATP elicited a clear increase in noise, but at the reversal potential of the ATP current (-5 mV), no increase in noise above background was seen. These results indicate that there is only a single class of excitatory ATP-activated channels, which do not select by charge. Based on analysis of the noise spectrum, the conductance of individual channels is estimated to be 0.2-0.4 pS.

A calcium- and voltage-dependent chloride current in developing chick skeletal muscle

1. Depolarization of embryonic chick myotubes from negative potentials elicits a rapid spike followed by a long-duration after-potential. The ionic basis of the long-duration after-potential was examined by making intracellular recordings from cultured myotubes, and by making whole-cell patch-clamp recordings from myoblasts and myoballs. 2. The peak potential of the long-duration after-potential varied with the chloride gradient, suggesting that a conductance increase to chloride is involved in generating the after-potential. However, a calcium current was also implicated, since lowering the extracellular calcium or replacing extracellular calcium with cobalt abolished the after-potential. 3. When extracellular calcium was replaced with strontium or barium, short-duration spikes similar to calcium spikes were observed, but only strontium was able to support activation of long-duration after-potentials. Intracellular injection of calcium or strontium into myotubes bathed in calcium-free extracellular solutions restored the ability of depolarization to evoke an after-potential. Intracellular injection of magnesium, barium, nickel or cobalt did not restore this ability. These experiments strongly suggested that the long-duration after-potential was due to a calcium- and voltage-activated chloride current. 4. Whole-cell voltage-clamp recordings from myoballs and myoblasts showed that a large chloride conductance could be activated by depolarization when the internal free calcium concentration was buffered at levels greater than 10(-7) M. At 2.5 x 10(-7) M-calcium, the voltage dependence of activation was steepest in the range of -30 to -20 mV and the activation kinetics varied with the membrane potential. The time to half-maximal activation ranged from 0.1 s at positive potentials to greater than 1 s at more negative potentials. The time constant for deactivation was approximately 1 s at -50 mV. No inactivation was observed. 5. The selectivity of the chloride current was measured by substituting other anions for chloride. The following permeability series was found: I- greater than NO3- greater than Br- greater than Cl- greater than acetate greater than F- greater than SO4- = glucuronate. Thus anion permeability decreased as the hydration radius increased. 6. Measurements of the resting potential of developing myoblasts and myotubes under 'physiological' conditions (37 degrees C, bicarbonate buffer) suggest that the after-potential acts to depolarize these cells 10-20 mV above their resting potential (approximately -60 mV) for several seconds. 7. We discuss the possibility that the long-duration after-potential may be involved in triggering myoblast fusion and in the generation of bursts of spontaneous contractions in developing myotubes.

Dil and diO: versatile fluorescent dyes for neuronal labelling and pathway tracing

The fluorescent carbocyanine dyes dil and diO have an extensive history of use in cell biology, but their use as neuronal tracers is relatively recent. We found in 1985 that these molecules were excellent retrograde and anterograde tracers in the developing nervous system. We went on to show that these dyes were retained in neurons placed in culture, that they initially labelled the processes as well as the cell bodies of cultured neurons, and that they were seemingly non-toxic. We suggested that the major mechanism of translocation for these molecules was lateral diffusion in the membrane, rather than fast axonal transport. This suggestion was recently confirmed in a striking manner by Godement et al., when they showed that these dyes can be used to label axonal projections in fixed tissues. Labelling with carbocyanine dyes has already allowed several exciting advances in developmental neurobiology. In this article we review the properties of carbocyanine dyes and point out some of their uses and advantages.

Multiple actions of adenosine 5'-triphosphate on chick skeletal muscle

1. Extracellularly applied adenosine 5'-triphosphate (ATP) is known to have an excitatory action on chick skeletal muscle. By making intracellular recordings from cultured chick myotubes bathed with blockers of several types of voltage-dependent channels, the direct action of ATP could be observed. 2. When muscle cells were studied near their resting potential, ATP usually produced a biphasic response. There was a rapid initial depolarization, followed by a slower repolarization. The repolarization could drive cells negative to their initial resting potential, indicating that it was not due simply to desensitization of the process that produced the depolarization. Thus there are at least two distinct responses to ATP. 3. At room temperature the early response to ATP activated within 20 ms, and the second response activated with a latency of approximately 1 s. In our standard blocking solution, the average reversal potential of the early response was -17 mV, while the late response had a reversal potential that was negative to -70 mV. In a few cells the second response appeared to be absent. 4. The amplitude and time course of the late response were substantially decreased by low temperature (12 degrees C) and increased by high temperature (37 degrees C). In contrast, temperature had much smaller effects on the early response. Both the time course and temperature dependence of the late response suggest that an intracellular second messenger system may be involved in its activation. 5. Ion-substitution experiments were performed to determine the type of conductance changes that evoke each response. These indicated that the early response was due to an increased membrane permeability to sodium, potassium and chloride, but not to large cations or anions, and that the late response was due to an increased permeability to potassium. 6. Measurement of the responses of muscle cells to acetylcholine supported the conclusion that both anions and cations are permeable during the early ATP response. Under conditions in which there was a large negative reversal potential for all cations, and a large positive reversal potential for all anions, the early ATP response reversed approximately 50 mV positive to the acetylcholine response. 8. The possibility that the early ATP response is due to a channel selective for size, but not charge, is discussed.

Fluorescent carbocyanine dyes allow living neurons of identified origin to be studied in long-term cultures

A prerequisite for many studies of neurons in culture is a means of determining their original identity. We needed such a technique to study the interactions in vitro between a class of spinal cord neurons, sympathetic preganglionic neurons, and their normal target, neurons from the sympathetic chain. Here, we describe how we use two highly fluorescent carbocyanine dyes, which differ in color but are otherwise similar, to identify neurons in culture. The long carbon chain carbocyanine dyes we use are lipid-soluble and so become incorporated into the plasma membrane. Neurons can be labeled either retrogradely or during dissociation. Some of the labeled membrane gradually becomes internalized and retains its fluorescence, allowing identification of cells for several weeks in culture. These dyes do not affect the survival, development, or basic physiological properties of neurons and do not spread detectably from labeled to unlabeled neurons. It seems likely that cells become retrogradely labeled mainly by lateral diffusion of dye in the plane of the membrane. If so, carbocyanine dyes may be most useful for retrograde labeling over relatively short distances. An additional feature of carbocyanine labeling is that neuronal processes are brightly fluorescent for the first few days in culture, presumably because dye rapidly diffuses into newly inserted membrane. We have used carbocyanine dyes to identify sympathetic preganglionic neurons in culture. Our results indicate that preganglionic neurons can survive in the absence of their target cells and that several aspects of their differentiation in the absence of target appear normal.

Excitatory action of ATP on embryonic chick muscle

It has been suggested that ATP might play a role in synaptic transmission at developing vertebrate neuromuscular junctions. To increase our understanding of the events underlying synapse formation, we have used intracellular recording and patch clamp recording to examine the response of chick myoblasts and myotubes to to ATP and other nucleotides, ATP, applied at micromolar concentrations, has a potent depolarizing action on chick myoblasts and myotubes. The ATP depolarization declines during prolonged application of ATP and shows no recovery for at least 20 min after the removal of ATP. The physiological event that underlies the ATP response has a reversal potential near O mV and is due to a conductance increase. However, contrary to our expectation, in a series of nearly 200 cell-attached and outside-out patch recordings, we did not detect single-channel currents that were related to ATP. The myotube ATP receptor is pharmacologically distinct from putative ATP receptors in other systems. It is not activated by ADP, AMP, or adenosine. Furthermore, the nonhydrolyzable ATP analogs, AMP-PNP, alpha,beta-meATP, and beta,gamma-meATP (respectively, 5-adenylylimido diphosphate; alpha,beta-methylene adenosine 5'-triphosphate; and beta,gamma-methylene adenosine 5'-triphosphate), which are potent ATP agonists in other systems, have no depolarizing action on myotubes. The ATP receptor is also distinct from the nicotinic ACh receptor since responses to ATP are unaffected by the nicotinic antagonists d-tubocurarine and alpha-bungarotoxin. We therefore applied alpha-bungarotoxin to nerve-muscle co-cultures in the hope of uncovering an additional component of the postsynaptic potential, which might represent a synaptic action of ATP. Under these experimental conditions no evidence indicative of a postsynaptic action of ATP released from nerve terminals was observed.

Acetylcholine release from growth cones detected with patches of acetylcholine receptor-rich membranes

Synaptic potentials can be evoked at nerve-muscle junctions in vitro within minutes after an exploring growth cone contacts a receptive myotube. Functional transmission is also evident in vivo on the time scale of minutes after motor axons enter adjacent myotomes. The ability to release acetylcholine (ACh) may be induced in motor nerve terminals after they contact receptive target cells. Alternately, growth cones may be capable of releasing ACh before contact. To examine the development of ACh release we have used isolated patches of acetylcholine receptor(AChR)-rich membrane as sensitive detectors of ACh. We report here that the growth cones of embryonic chick ciliary ganglion neurones can release ACh, even when the cells are grown in the absence of target myotubes.

Apportionment of the terminals from single preganglionic axons to target neurones in the rabbit ciliary ganglion

We have studied the apportionment of terminals from single preganglionic axons to target neurones in the ciliary ganglion of adult rabbits. Both electrical recording and intra-axonal injection of horseradish peroxidase (HRP) showed that each preganglionic axon innervates only a small fraction of the ganglion cell population (about 10-20 of the approximately 400 ganglion cells). Examination of ganglia in whole mounts showed that neurones whose cell bodies were enveloped by HRP-labelled boutons from a single axon were often surrounded by other neurones which received no contacts from the labelled fibre. Electron microscopical examination of labelled presynaptic terminals on individual ganglion cells confirmed that the boutons of single axons were sharply confined to particular target cells. This suggests that individual target neurones (or portions of them) are the unit of innervation during the development of these synaptic connexions. Comparison of the amplitudes of synaptic responses in singly and multiply innervated ganglion cells indicated that, on average, an individual axon made a weaker synaptic connexion with a multiply innervated neurone than with neurone that received only one input. Moreover, neurones innervated by several different axons tended to have fewer synapses on their somata than neurones innervated by only one or two preganglionic axons. Individual post-synaptic profiles were often contacted exclusively by labelled terminals when examined in the electron microscope. Since many of these neurones are multiply innervated, this observation suggests some regional separation of the several inputs contacting the same cell. For several reasons, however, this inference must be regarded as tentative. Taken together, these findings provide a possible explanation of the correlation between the dendritic geometry of ganglion cells and the number of different axons that innervate them (Purves & Hume, 1981). The several axons that initially innervate ganglion cells without dendrites evidently compete during early life until only a single input remains. On ganglion cells with dendrites, however, the number of inputs that persists is proportional to dendritic complexity. The present results suggest that the diminished competition between axons innervating neurones with dendrites may result from some degree of terminal segregation on dendritic arborizations.