DEVELOPMENTAL BIOLOGY 119,290-298 (1987)

[3H]Acetylcholine Synthesis in Cultured Ciliary Ganglion Neurons: Effects of Myotube Membranes

D.BRUCE GRAY* ANDJEREMY B.TUTTLE~

*The University of Connecticut, Department of Physiology and Neurobiology, Storrs, Connecticut 06268, and tThe University of Virginia School of Medicine, Department of Physiology, Charlottesville. Virginia 22903

Received July 13, 1984; accepted in revised form July 21, 19X6

Avian ciliary ganglion neurons in cell culture were examined for the capacity to synthesize acetylcholine (ACh) from the exogenously supplied precursor, choline. Relevant kinetic parameters of the ACh synthetic system in cultured neurons were found to be virtually the same as those of the ganglionic terminals in the intact iris. Neurons were cultured in the presence of and allowed to innervate pectoral muscle; this results in an enhanced capacity for ACh synthesis. In particular, the ability to increase ACh synthesis upon demand after stimulation is affected by interaction with the target. This effect is shown to be an acceleration of the maturation of the cultured neurons. Lysed and washed membrane remnants of the muscle target were able to duplicate, in part, this effect of live target tissue on neuronal transmitter metabolism. Culture medium conditioned by muscle, and by the membrane remnants of muscle, was without significant effect. Thus, substances secreted into the medium do not play a major role in this interaction, Neurons cultured with either muscle or muscle membrane remnants formed large. elongate structures on the target membrane surface. These were not seen in the absence of the target at the times examined. This morphological difference in terminal-like structures may parallel the developmental increases in size and vesicular content of ciliary ganglion nerve terminals in the chick iris, and may relate to the increased ACh synthetic activity. The results suggest that direct contact with an appropriate target membrane has a profound, retrograde influence upon neuronal metabolic and morphological maturation. G 1987 Academic Press, Inc

INTRODUCTION

The signaling function of peripheral neurons requires a steady supply of transmitter for secretion. Thus, a metabolic system able to provide transmitter during pe- riods of high output is necessary if output is to be main- tained. Embryonic neurons lack this functional capacity. Developing neurons form synapses that do not maintain transmitter output (Pilar et at, 1982) and that have low rates of spontaneous release and low quanta1 content (Bennett and Pettigrew, 1974; Diamond and Miledi, 1962; Letinsky, 1974; Fambrough, 1974). The detection of spontaneously occurring miniature end-plate potentials, and thus the sign diagnostic for the initial stage of syn- apse formation, is not accompanied by either special- ization in structure (Weldon and Cohen, 1979; Cohen and Weldon, 1980) or function (Chow and Poo, 1985). Our knowledge of the events that intervene between the ini- tial formation of synapses and the expression of a ma- ture, stable, and fatigue-resistant junction is incomplete. The role of cellular interaction in this process has not been defined. In particular, the acquisition of a mature transmitter synthetic capacity by the neuron has not been studied in sufficient detail. The mechanisms un- derlying these events are also unclear (for review see Bunge et aL., 1978).

Chick ciliary ganglion neurons in cell culture have re- cently been found to modulate acetylcholine synthetic capacity in response to an interaction with muscle target tissue (Tuttle et al., 1983). This culture system may therefore serve as a useful model in which to examine the mechanisms of neuron-target interaction that are involved in the acquisition of adult transmitter synthetic capacity. The ciliary ganglion in vitro model may be particularly useful since much information regarding the functional and structural characteristics of ciliary ganglion development in vivo is available (see Pilar and Tuttle, 1982).

Functional synapses are formed by ciliary neurons on iris and ciliary body muscle fibers in the embryo prior to hatching (Landmesser and Pilar, 1978). However, synaptic efficacy, basal AChl synthesis, and the ability to increase ACh synthesis in response to demand remain low throughout the embryonic period. Significant in- creases in these indices of synaptic function occur only at or just after hatching, when the reflex pathway begins

i Abbreviations used: ACh, acetylcholine; ChAT, choline acetyl- transferase; MEM, Minimal essential medium (Eagle’s); HH, Ham- burger and Hamilton (stages of the development of the chick embryo); NMJ, neuromuscular junction.

OOlZ-1606/87 $3.00 Copyright C 1987 by Academic Press, Inc. All rights of reproduction in any form reserved.

290

GRAY AND TUTTLE ACh Synthesis cl& Toy/et Membranes 291

to operate (Pilar et ah, 1981). The capacity for ACh syn- thesis was examined in cultured ciliary ganglion neurons in order to evaluate the developmental role of interaction with the muscle target in a simplified system, where various components of the target could be presented to isolated developing neurons.

In the nerve terminals of mature ciliary ganglion neu- rons, greater than 90% of ACh synthesis is blocked in the absence of external Nat ion, due to the sodium re- quirement for high-affinity transport of the ACh pre- cursor choline (Suszkiw and Pilar, 1976; Vaca and Pilar, 1979). Choline transport has been characterized in this and other systems and shown to be closely related to the synthesis of ACh (e.g., Yamamura and Snyder, 1973; Barald and Berg, 1979a,b). However, this high-affinity transport system is absent in the neural soma, and somal ACh synthesis is independent of the external Na+ con- centration. Thus, in order to measure synaptically re- lated ACh synthesis, we have determined the amount of ACh synthesized from externally supplied choline both in the presence and absence of external Na’.

ACh synthesis following a preincubation period in high (55 mM) [K’],, measures the response of the syn- thetic system to depletion of the existing stores of transmitter. In vuivo, this responsiveness does not reach significant proportions until hatching, and the acquisi- tion of this property may depend upon developmentally relevant target interaction (Pilar et al., 1982).

Our previous experiments have shown that coculture of ganglionic neurons with striated muscle for increased basal ACh synthesis and the responsiveness of ACh synthesis to prior depletion relative to neurons cultured alone (Tuttle et al., 1983; Gray and Tuttle, 1983). The chemosensitivity of the neuronal soma to applied ACh is also maintained in coculture with muscle, but lost in its absence (Crean et al., 1982). The exact nature of these trophic interactions has yet to be demonstrated. Two possibilities for the mechanism of interaction include the influence of some soluble protein factors produced and released by the target tissue, and membrane-bound factors involved in a cell-to-cell contact-mediated inter- action. Evidence for the former derives from experi- ments suggesting embryo eye extract or fractions thereof stimulate ciliary neuron ChAT activity as measured in homogenates (Nishi and Berg, 1981), and other soluble factors from muscle and culture medium conditioned by muscle that promote ACh synthesis (Tuttle et ah, 1983; Nishi and Berg, 1981). Evidence that direct cell contact may also play a role comes from experiments in which myotubes (polynuclear fused muscle cells) have been raised in culture, lysed by hypoosmotic shock, and washed, leaving only semiattached muscle membrane on the collagen surface of the culture. Ciliary neurons plated upon this substrate also show enhanced retention

of chemosensitivity to ACh (Tuttle, 1983) suggesting a profound trophic effect of the target membrane. The question now arises whether such contact with appro- priate target membrane results in significant trophic stimulation of neuronal cholinergic metabolism. Before meaningful study of the acquisition of mature trans- mitter synthetic capacity can proceed, however, the va- lidity and comparability of the biochemical properties of the ACh synthetic system of the cultured neurons must be established in relation to similar results from ciliary nerve terminals in viva Some of the results con- tained in this report have been previously presented in abstract form (Gray and Tuttle, 1983, 1984).

METHOIX?

Tissue culture. Ciliary ganglion neurons were derived from g-day-old chick embryo ciliary ganglia via gentle trypsinization and mechanical dissociation, as described previously (Tuttle et al., 1983). Cells were resuspended in growth medium (consisting of 80 ml MEM, 10 ml chick embryo extract, 10 ml horse serum (GIBCO), 2 ml MEM amino acid mix, and 1 ml MEM vitamin mix), and plated in collagen-coated, 16-mm microwells (Linbro), and then incubated at 37°C and 5%) CO,.

Pectoral muscles were dissected from 11-day chick embryos, rinsed with sterile Tyrode’s solution, trans- ferred to growth medium, and dissociated by trituration and plated (after Tuttle et a,Z., 1983) into collagen-coated microwells. Muscle cultures were treated with 10 PM cytosine arabinoside after 2 days in culture, and the an- timetabolite removed after 48 hr. After 4-5 days in cul- ture, myotubes had formed and ciliary ganglion neurons were added. The number of neurons in the cultures was estimated by counting phase-bright cell somata under phase-contrast microscopy. Neuron density was rou- tinely 7000-12,000 neurons per 2.0-cm2 microwell (3500- 6000 neurons per cm’).

In some cases medium “conditioned” from cultures of chick ciliary ganglion neurons, myotubes, or lysed myo- tube membranes was substituted for normal growth medium. Cultures used for medium conditioning were similar in nature, composition, cell density, and age to the cultures described above. Medium was routinely transferred to the test cultures after a 2-day exposure (“conditioning”) to the host culture.

Preparation of the lysed myotube substrate. Eleven- day chick embryo pectoral muscle was dissociated, plated, treated with arabinoside, and incubated for 5-7 days in standard growth medium as described above. On the same day as ciliary ganglion dissection, myotube cultures were washed three times in sterile distilled wa- ter and incubated for 3 hr at 4°C to osmotically lyse and wash the myotubes. Cultures were then washed three times in growth medium at 37”C, and ciliary ganglion

292 DEVELOPMENTAL BIOLOGY VOLUME 119, 1987

neurons were added. Lysed fibroblast substrates were prepared in exactly the same way, except that fibroblast cultures were obtained from dissociated cranial skin tissue of 11-day chick embryos. These primary cultures were subcultured at 5-day intervals to eliminate myotube contamination (Tuttle, 1983).

ACh synthesis assay. Microwell cultures were washed three times with modified Tyrode’s solution (consisting of 130 mMNaC1,20 mM HC03,4 mM KCl, 2 mM CaC12, 1 mM MgC12, 10 mM glucose, 1 mM glutamine, and 10 mM Hepes, pH 7.4), and then preincubated for 10 min with either the above Tyrode’s or a similar solution with 55 mM K+, and 78 mM NaCl. The cultures were then washed with normal Tyrode’s and incubated with 1 yM [3H]choline (8.4 Ci/mmole) (New England Nuclear) in either normal Tyrode’s or zero Na+ Tyrode’s (NaCl re- placed by LiCI) at 37°C for 30 min. In experiments wherein the substrate concentration was altered, the specific activity of choline was varied as choline concen- trations ranged from 0.1 PM to 1.0 mM. In experiments concerning the time course of synthesis, the incubation periods in [3H]choline ranged from 2 min to 45 min, with a total choline concentration of 1 PM.

In all experiments, cells were harvested at the end of the incubation period by washing twice (in either normal or zero Naf Tyrode’s buffer) and were immediately in- cubated in 0.5 ml extracting solution of formic acid-ace- tone (15/85 v/v) containing ‘*C-ACh for the estimation of recovery. Cells were scraped from the bottom of the microwells and the suspension was homogenized on ice. 3H-ACh was separated from the [3H]choline precursor by high-voltage paper electrophoresis, eluted, and quantified by liquid scintillation (Vaca and Pilar, 1979).

These procedures produced overlap between choline and ACh of less than l%, and recovery of sample as measured by recovery of the internal standard (14C-ACh) routinely fell between 50 and 60% .The following choline metabolites were tested for possible contamination in the above assay: choline, phosphocholine, betaine, and butyrlcholine. No significant migratory overlap of these metabolites could be demonstrated.

ChATassay. ChAT activity was assayed by measuring the synthesis of 3H-ACh from [3H]acetyl coenzyme A, using an adaptation of the method of Fonnum as pre- viously described (Tuttle et ab, 1983).

Dye injections. The fluorescent dye Lucifer Yellow CH was injected iontophoretically into randomly selected ciliary ganglion neurons under each of the culture con- ditions. For these experiments, cells were plated onto plastic coverslips coated with dried rat tail collagen and cultured in 35-mm Falcon tissue culture dishes in 2 ml standard medium. High resistance microelectrodes (80- 150 MR with 3 M KCl, >300 MO with the dye solution) containing 4% Lucifer Yellow in a 0.2 M LiCl solution

with 2% paraformaldehyde were used. (The Lucifer Yel- low was a gift of Dr. Stewart, NIH). Employing pulsed currents at 1 Hz, successful filling of an impaled neuron required 30-120 sec. Following a 2-hr diffusion period, the cultures were fixed in 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.2, for an additional 2 hr. After mounting in glycerol, cultures were viewed with a flu- orescence microscope, and photographed.

RESULTS

Neurons under all of the culture and coculture con- ditions synthesized significant amounts of ACh both in the presence and in the absence of extracellular sodium. Figure 1A shows the time course of ACh synthesis for neurons in culture alone (no coculture) for 7 days. Values were obtained under conditions expected to produce ex- tremes in the amount of synthesis: preincubation with 55 mM Kf and assay in normal Tyrode’s (high), versus preincubation and assay in zero Na+ Tyrode’s (low). In both cases, a 30-min incubation period in [3H]choline was chosen because accumulation of labeled ACh was approximately linear for this time. Also, the total amount of 3H-ACh synthesized gave a satisfactory signal to noise ratio for samples of relatively few cells. For all of the culture conditions used, whether with or in the absence of muscle, and at 4-12 days in culture, similar curves of 3H-ACh accumulation were observed (data not shown). It should be noted that this assay may not in the strictest sense be measuring rates of 3H-ACh syn- thesis under defined substrate conditions, but rather re- flects the capacity of the neurons to accumulate 3H-ACh. This accumulation is the net sum of [3H]choline uptake and conversion to 3H-ACh, minus any loss due to deg- radation, efflux, or other cellular processes (Mains and Patterson, 1973; Patterson et al., 1975).

Cultures of myotubes alone or of fibroblasts without neural cells did not synthesize detectable amounts of 3H-ACh under these assay conditions. These nonneural cells did, however, take up [3H]choline, predominately via a sodium-independent pathway (data not shown).

Figures 1B and C show the effect of varying choline concentrations on neuronal 3H-ACh synthesis assayed in the presence and absence of Nat, both in culture alone and in coculture with myotubes. An Eadie-Hofstee plot of a similar experiment wherein the substrate concen- tration was varied for a neuron-myotube coculture (Fig. 1D) clearly indicates there are two separable components of ACh synthesis. The apparent Km of the sodium-de- pendent component is approximately 2 pM. Estimates of the apparent Km, derived similarly, of this high af- finity process from neurons cultured alone and cultured upon myotube membranes (see below) were always be- tween 1 and 4 pM. Choline concentrations for half-max-

GRAY AND TUTTLE

lncubotion Time (min)

6 c t

I ,

A’ / /’ -oL

/ I/O ’ /I

‘0, ’

A

293 ACh Synthesis and Target Membranes

cl l,e;

,‘;/A B /’ 8 I A I

I I 1 0.2 0.4 0.6 0.0 IO

0.1 I IO 100 1000 V

Choline Concentrollon (PM) /c (fmole/hr/neuron/pm choline)

FIG. 1. sH-ACh synthesis by cultured ciliary ganglion neurons: Time course and dependence on substrate concentration. (A) Neurons cultured alone, 7 days in vitro. Triangles refer to values obtained in normal saline following preincubation (10 min) in 55 mM[K’]; circles refer to values with no preincubation but obtained in zero external [Na’]. The concentration of choline was maintained at 1 pM during periods of incubation. (B) Neurons cultured alone, 7 days in vitro. Triangles refer to values obtained in zero external [Na’], circles to Na+-dependent values (data in zero Na+ subtracted from data in normal saline). In this and in (C), the specific activity of [8H]choline was varied to accommodate the widr range of choline concentration. (C) Neurons cocultured with myotubes, ‘7 days, symbols are the same as in (B). (D) Neurons in coculture with myotubes, 7 days. An Eadie-Hofstee plot of the concentration dependence of *H-ACh synthesis. The symbols are the same as in (B) and (C). Note the clear separation of the sodium-dependent and independent values for ‘H-ACh synthesis,

imal synthesis under zero Na+ conditions ranged from 15-50 PM, but Eadie-Hofstee plots of these values were not always linear, thus precluding acceptable estimation of this value for the lower affinity process. The values for the higher affinity, sodium-dependent process were invariably linear in the analysis. Thus, Na+-dependent 3H-ACh synthesis appears to be highly correlated, on a kinetic basis, with Na+-dependent, high-affinity choline uptake (Suszkiw and Pilar, 1976).

Figure 2 shows 3H-ACh synthesis under basal condi-

tions and following stimulation by preincubation in 55 mM K+, both in the presence and absence of Na+, and for three different times in culture. It is important to note that although Na+ independent synthesis is very low (40% of the total) in ciliary ganglion nerve ter- minals in vivo, Na+-independent synthesis accounts for close to 50% of the basal 3H-ACh synthesis in all cases in culture. However, in culture, the assay cannot distin- guish between synthesis in the neuronal soma and that limited to the processes.

294 DEVELOPMENTAL BIOLOGY VOLUME 119,1987

4 da 7 da 12 da

FIG. 2. 3H-ACh synthesis in ciliary ganglion neurons cultured alone for 4, ‘7, and 12 days, under four assay conditions. Hatched bar: basal,

no preincubation, normal saline. Solid bar: basal, no preincubation, zero [Na’] saline. Clear bar: Preincubation in 55 mM [K+] (10 min) for depolarization stimulus, synthesis in normal saline. Cross-hatched bar: preincubation in 55 mM [K+] (10 min), synthesis in zero [Na+] saline. The values represent the mean + SE of at least 9 experiments. *Significantly different mean, Student’s t test, P, 0.05. It is only after 12 days in culture that neurons cultured alone respond to the depolarizing preincubation with an increase in ‘H-ACh synthesis. The stimulation of synthesis is primarily in the [Na+]-dependent portion.

At 4 and 7 days in vitro, no stimulation of 3H-ACh synthesis can be detected after preincubation with 55 mM K+. By 12 days in vitro, however, accumulation of 3H-ACh is markedly higher in the 30-min period after the preincubation in K+-rich buffer. This effect is espe- cially apparent for the Na+-dependent portion (total 3H- ACh accumulation minus 3H-ACh accumulation in zero Na+). The sodium-dependent fraction of the total 3H- ACh accumulation accounts for over 80% of the increased synthesis after high-K+ incubation.

Figure 3 shows a comparison of the Na+ dependent portion of the 3H-ACh accumulation values during basal and stimulated periods for cultures of neurons alone, neurons in coculture with myotubes, and in coculture with myotube membranes. These experiments were all performed on cultures at 7 days in vitro. Basal 3H-ACh accumulation is two times higher for neurons in cocul- ture with myotubes than for neurons cultured alone, and K+-stimulated accumulation is enhanced by a factor of more than four. In fact, the responsiveness to depolar- izing preincubation of the 3H-ACh synthetic and accu- mulation processes is already apparent at only 4 days of neuronal culture on myotubes. Na+-dependent basal accumulation of 3H-ACh was 0.48 f 0.06 (mean + SEM) fmole/cell/hr while stimulated values were 0.89 +- 0.16 fmole/cell/hr for such cultures at 4 days. Coculture of ciliary ganglion neurons with live myotubes clearly en- hances Na+-dependent 3H-ACh accumulation, an effect shown previously (Tuttle et ah, 1983). The enhanced syn-

thetic responsiveness is shown here to be observable soon after the neurons are plated. However, culture on the membranes of myotubes also enhances the neuronal ability to produce ACh, but to a level intermediate be- tween neurons alone in culture and neurons in coculture with the live myotubes (Fig. 3). The beneficial effect of myotube membranes on the capacity for 3H-ACh accu-

A B C

FIG. 3. Sodium-dependent ‘H-ACh synthesis by neurons cultured alone (A), neurons in coculture with myotubes (B), and cultured on the membrane remnants of myotubes (C) for ‘7 days in vitro. Clear bars refer to values without preincubation, the dotted bars to exper- iments with preincubation in 55 mM[K+] (10 min) prior to measurement of synthesis. Solid bars represent total ChAT activity in homogenates of cell cultures and refer to the ordinate scale on the right. Values of synthesis are mean f SE of at least 9 experiments, n = 4 for the ChAT measurements.

GRAY AND TUTTLE ACh Synthesis and Target Membranes 295

mulation is less evident at the earlier time (4 days) in culture, but even then, the target membrane substrate appears to elevate neuronal accumulation of labeled transmitter.

Activity levels of the enzyme responsible for ACh synthesis (acetyl-CoA:choline D-acetyltransferase (ChAT, EC 2.3.1.6)) in neurons cocultured with myotube membranes for 7 days showed a 72% increase over ac- tivity in neurons cultured alone (neurons alone, 0.22 pmole ACh/cell/hr; neurons on membranes, 0.38 pmole ACh/cell/hr). These values of ChAT activity also show an intermediate effect, since coculture of the neurons with live myotubes results in greater than doubled ChAT activity, to 0.57 pmole ACh synthesized/cell/hr.

The ability of myotube membrane to enhance the neu- ronal system for transmitter synthesis and accumulation would appear to implicate a direct contact interaction with some component of the muscle membrane as an operative agency, since soluble or releasable products of the myotubes are presumably subtracted during osmotic shock lysis and washing. However, incomplete washing, or the release of degradation products of the myotube membranes into the culture medium may remain as a possible contributory factor for the effect on neuronal maturation. Thus, this possibility was examined by cul- ture of neurons plated alone in medium that had been exposed to both live myotubes and to the lysed mem- branes of myotubes (medium thus “conditioned” by the muscle tissue or membranes and their products would presumably contain any molecules active in solution). As shown in Fig. 4, ‘Lconditioned” culture medium from either the live myotubes or their membranes failed to have any trophic, beneficial effect upon the neuronal ac- cumulation of 3H-ACh. It might be noted, however, that the “conditioned” medium was not concentrated in any way, and thus a product at very low concentration or very labile would have escaped detection.

The effect of culture on myotube membranes appears to be specific for muscle, as coculture of ciliary ganglion neurons with lysed fibroblast cell membranes produces no increase in neuronal accumulation of 3H-ACh follow- ing stimulation in 55 mM [K+],: 0.78 -t 0.28 fmole 3H- ACh/cell/hr (basal) and 0.66 & 0.15 fmole/cell/hr (after 55 mM K+ preincubation) (n = 5). Thus, even though fibroblasts are also mesenchymal, and share many of the same antigenic membrane epitopes with myoblasts and myotubes (D. Fambrough, personal communication), coculture with their membranes or with live fibroblasts (Tuttle et al., 1983) does not elevate transmitter accu- mulation. Therefore, the effect of myotube membranes does not appear to be due an increase in general adhesion to the cellular substrate, or to some other nonspecific aspect of the lysed cell substrate.

Ciliary ganglion neurons in culture have been shown to form synaptic structures upon one another, and to

NEURONAL CM MYOTUEE CM MEMBRANE CM

FIG. 4. Effect of conditioned media upon 3H-ACh synthesis by ciliary ganglion neurons. Assay conditions and symbols are the same as in Fig. 2. See Methods for details of conditioned medium preparation. No significant increases in 3H-ACh synthesis were caused by the media as compared to synthesis by neurons in routine growth medium, and depolarization did not significantly elevate synthesis of cells grown in any of these test media. The values are expressed as means +- SE. rl = 4.

rapidly form functional neuromuscular synaptic con- tacts when in coculture with myotubes (Betz, 1976; Nishi and Berg, 1977; Crean et al., 1982). In order to better visualize the morphological consequences (if any) of the enhanced maturation of transmitter synthesis and ac- cumulation seen in the cultures of neurons with myo- tubes and their membranes, Lucifer Yellow CH dye was injected into individual neurons, and allowed to diffuse throughout the entire cell. After fixation and mounting, the entire extent and ramification of individual neurons could be examined under fluorescence microscopy, even in complex cultures. Figure 5 is a montage of some of these cultured neurons under the culture conditions of experimental interest: neurons plated alone (A, B), with myotubes (C), and on myotube membranes (D). The neurons in culture without muscle tissue formed a rather typical meshwork of processes with varicosities. These varicosities were found along the processes of neurons under all culture conditions. They range from 0.2 to 1 pm in size, and were associated with the surfaces of other cells and the collagen substrate. In the absence of muscle target or myotube membrane, the neuritic processes formed complex branching patterns (Fig. 5A). Presumed interneuronal contacts were common (Fig. 5B) and of the size expected from the interneuronal synaptic struc- tures examined by transmission electron microscopy in a previous study (Crean et al., 1982).

Figure 5C shows that larger and more extensive structures are formed by neuronal processes in contact with a myotube. These larger swellings and thickened processes were never seen in the cultures of neurons without muscle tissue or membranes at less than 8 days

296 DEVELOPMENTAL BIOLOGY VOLUME 119. 1987

FIG. 5. Prints of color slide photomicrographs of Lucifer Yellow-injected cultured neurons taken under transmission fluorescence microscopy. (A) Ciliary ganglion neuron in culture without muscle showing the typical branched network of processes decorated with numerous small swellings. These varicosities are not artifacts of the dye injection, as they are also present under scanning electron and transmission electron microscopy of unfilled neurons (Tuttle, 1984). Bar = 10 pm. (B) Higher magnification of an interneuronal contact between two neurons in a culture of neurons alone. The spherical gray form on the right is the cell soma of an unfilled neuron. Note the diameter of the process and the size of the bouton-type contact. Bar = 5 pm. (C) Complex of processes extending from a single filled neuronal soma and coursing over a live myotube. Bar = 5 pm. (D) Complex of processes similar to those in (C), but formed by a single neuron on the membrane remnant of a myotube. Bar = 5 Wm.

in vitro. These structures were also seen only approxi- mated to myotubes or their membranes, and not on other neurons in the cocultures or on fibroblasts. The neuronal processes in the myotube cocultures often followed the long axis of a contacted myotube for a considerable dis- tance, without the extensive larger angle branching seen in the neuronal cultures without myotubes (Fig. 5A). Finally, Fig. 5D is an example of a filled neuronal struc- ture upon the lysed myotube membrane substrate. The large swellings and thick neuritic processes seen on the live myotubes were also formed on the myotube mem- branes. While these enlarged processes were prominent features of neuron-myotube membrane cocultures, they occurred less frequently than in the myotube cocultures.

DISCUSSION

These results demonstrate that ciliary ganglion neu- rons plated upon a substrate of osmotically lysed myo- tubes have an ACh synthetic capability intermediate between that of neurons growing in culture with no muscle target and neurons in coculture with myotubes. It is interesting to speculate that trophic regulation of neuronal transmitter metabolism may be mediated via membrane associated cell contact as well as by soluble factors. In cultures of embryonic nigrostriatal tissue (Prochiantz et CL, 1979; Prochiantz et aZ., 1981) tissue from the target areas and membrane preparations of

GRAY AND TUTTLE ACh Synthesis and Target Membranes 297

that tissue both increased dopamine uptake, thus sug- gesting a beneficial influence of membranes from the target area on neuronal maturation.

The present results of culture in conditioned medium appear to rule out a significant role for soluble factors as the trophic mediators for the partial effect on en- hanced ACh synthesis. This parallels recent data on the trophic regulation of chemosensitivity from stage 36 (11 days in ouo) ciliary ganglion neurons in vitro (Tuttle, 1983). However, it is important to note that soluble fac- tors may be present in much higher concentrations at cell surfaces than throughout the culture medium. Also, soluble factors in conditioned media may have influences that are masked or altered by the components of horse serum and embryo extract present in the culture me- dium. Soluble factors have been reported to elevate ac- tivity of ChAT in cultured neurons (e.g., Nishi and Berg, 1981). The partial effectiveness of the myotube mem- brane substrate in duplicating the influence of interac- tion with live myotubes could reflect either the absence of exchanged soluble factors, or some other qualitative or quantitative deficiency in the membrane substrate.

The exact component of the cell membrane responsible for producing this trophic effect is unknown. Likely can- didates for this component range from N-CAMS (Edel- man, 1983) to ACh receptors (cf. Bekoff and Betz, 1976), or almost any other molecule or molecular complex on the muscle surface. Experiments involving tests of myo- tube cultures after specific enzymatic digestion both be- fore and after lysis should demonstrate the extracellular exposure of the active moiety and its molecular class. Although basal lamina is not seen in electron microscopy of these embryonic myotubes (Crean et ab, 1982), extra- cellular matrix molecules may be involved. Extracellular matrix material has been implicated in fostering the growth and differentiation of mammary and epithelial cells in vitro (Wicha et al, 1982). Also, p-formaldehyde- fixed nonneural cells from several sources are reported to increase ACh synthesis by normally adrenergic sym- pathetic neurons in culture (Hawrot, 1980). Lysed sub- strates were not tested for cholinergic induction. In the present case, the active moiety would have to be specific for muscle or myotube, as fibroblasts are ineffective.

The formation or stabilization of a neuromuscular junction may also play a central role. Neuromuscular junctions are formed rapidly and readily by ciliary gan- glion neurons in vitro (Betz, 1976; Nishi and Berg, 1977; Crean et al, 1982). The Lucifer Yellow dye injection al- lowed morphological contacts to be visualized that re- semble in size those seen in neuron-muscle coculture. Contacts of this large extent were not seen in the cultures lacking myotubes or myotube membranes. A previous study of the embryonic iris demonstrated a temporal correlation between nerve terminal size and the ap- pearance of ACh synthesis that increases in response to

depolarization (Pilar et ah, 1981). Fewer of the large ax- onal swellings and terminal-like structures were appar- ent in the myotube remnant cultures than in the cocul- tures. Thus, if a similar relationship between ACh syn- thesis and terminal size obtends in culture, this may explain why the myotube membrane substrate is less effective at increasing ACh synthesis than coculture with live myotubes. The fate of these terminal-like contacts on the myotube membranes was not followed longer than 7 days of culture, for the membrane substrate gradually loses its adhesion to the collagen and is progressively lost in the older cultures.

In vivo, Na+-dependent, high-affinity choline uptake appears to be limited primarily to neuronal terminals (Suszkiw and Pilar, 1976). Thus, a trophic effect on the terminals of the cultured neurons might be taken as a prerequisite for the results described. However, the lo- calization of active choline uptake and the correlated ACh synthesis solely to terminal sites has not been es- tablished for this culture system. Autoradiographic la- beling of high-affinity choline uptake in cultures of cil- iary ganglion neurons was reported to produce grain formation over both the cell soma and the extended pro- cesses (Barald and Berg, 1979b). Because considerable diffusion of the label occurred during fixation and pro- cessing, these results failed to localize uptake to specific cellular domains. Thus, in culture, high-affinity choline uptake may be significant in any or all parts of the neu- ron. The close association between choline uptake and synthesis, in this and other systems (e.g., McIntosh and Collier, 1976), may suggest that the dynamics of choline uptake are an important limiting factor in the regulation of ACh synthesis.

The kinetics of ACh synthesis in our cultures closely parallel those of ciliary neuron terminals in the iris muscle. The time course of labeled ACh synthesis and the substrate dependency of the pathway show the same results in vitro and in vivo (Vaca and Pilar, 1979), with a high affinity constant for choline of 2 ym. Neuronal ChAT levels both in vivo and in culture are thought to be adequate to allow ACh synthesis at rates several or- ders of magnitude higher than observed (Vaca and Pilar, 1979; Tuttle et al., 1983), thus the ACh synthesis mea- sured in the cultures may be more related to an intra- cellular equilibrium between choline and ACh than to levels of the synthesizing enzyme, ChAT. How this may relate to the apparent temporal correlation of synthesis with terminal size, and presumably the size of the ve- sicular compartment, is unclear.

Cultures of neurons without myotubes or target membranes increased ACh synthetic responsiveness to transmitter depletion by 2 weeks of culture. Chronolog- ically, this occurs almost exactly at the same time as the acquisition of a demand-responsive synthetic mech- anism in vivo (this is arrived at by adding the days of

298 DEVELOPMENTAL BIOLOGY VOLUME 119, 1987

incubation in ouo (9) to the days in culture (12-14)). It is thus unclear whether the enhancement of neuronal ACh synthesis by interaction with myotubes or their membranes is an acceleration of the same maturational process as occurs in vivo, or a phenomenon unique to the culture environment. In either case, experiments de- signed to reveal the operative moieties involved in, or the mechanism of the interaction should prove worth- while and revealing.

We thank Dr. G. Pilar for his encouragement and support. Supported by the U.S. Army Research Office; J.B.T. is the recipient of a Research Career Development Award (PHS-NINCDS).

REFERENCES

BARALD, K. F., and BERG, D. K. (1979a). Ciliary ganglion neurons in cell culture: High affinity choline uptake and autoradiographic la- beling. Dev. BioL 72,15-23.

BARALD, K. F., and BERG, D. K. (1979b). Autoradiographic labeling of spinal cord neurons with high affinity choline uptake in cell culture. Dev. BioL 72,1-14.

BEKOFF, A., and BETZ, W. J. (1976). Acetylcholine hot spots: Devel- opment on myotubes cultured from aneural limb bud. Science 193, 915-917.

BENNET, M. R., and PE~IGREW, A. G. (1974). The formation of synapses on striated muscle during development. J. PhysioL (London) 241, 515-545.

BETZ, W. J. (1976). The formation of synapses between chick embryo skeletal muscle and ciliary ganglion grown in vitro. J. PhysioL (Lor- don) 254,63-73.

BUNGE, R., JOHNSON, M., and Ross, C. D. (1978). Nature and nurture in the development of the autonomic neuron. Science 199.1409-1416.

CHOW, I., and Poo, M. M. (1985). Release of acetylcholine from em- bryonic neurons upon contact with muscle cell. J. Neurosci. 5,1076- 1082.

COHEN, M. W., and WELDON, P. R. (1980). Localization of acetylcholine receptors and synaptic ultrastructure at nerve-muscle contacts in culture-dependence on nerve type. J. Cell BioL 86,388-401.

CREAN, G., PILAR, G., TUITLE, J. B., and VACA, K. (1982). Enhanced chemosensitivity of chick parasympathetic neurons in coculture with myotubes. J. Physiol. (London) 331,87-104.

DIAMOND, J., and MILEDI, R. (1962). A study of foetal and new-born rat muscle fibers. J. PhysioL (Lundon) 162,393-408.

EDELMAN, G. (1983). Cell adhesion molecules. Science 219, 450-457. FAMBROUGH, D. (1974). Development of cholinergic innervation of

skeletal, cardiac, and smooth muscle. In “Biology of Cholinergic Function” (A. M. Goldberg and I. Hanin, eds.), Raven Press, London.

FONNUM, F. (1975). A rapid radiochemical method for the determination of choline acetyltransferase. J. Neurochem. 24,407-409.

GRAY, D. B., and TUTTLE, J. B. (1983). Target influences upon trans- mitter synthesis in cultured chick ciliary ganglion neurons. 5’0~. Neurosd Abstr. 9, 846.

GRAY, D. B., and TULLE, J. B. (1984). Target influences upon terminal morphology and ACH synthesis in cultured chick ciliary ganglion neurons. Sot. Neurosci. Abstr. 10, 1057.

HAWROT, E. (1980). Cultured sympathetic neurons: Effects of cell-de- rived and synthetic substrata on survival and development. Dev. Biol 74, 136-151.

LANDMESSER, L., and PILAR, G. (1978). Interactions between neurons and their targets during in viva synaptogenesis. Fed Proc 37,2016- 2022.

LETINSKY, M. S. (1974). Physiological properties of developing tadpole nerve-muscle junctions during repetitive stimulation. Dev. Biol 40,

154-161. MAINS, R. E., and PATTERSON, P. H. (1973). Primary cultures of dis-

sociated sympathetic neurons. II. Initial studies on catecholamine metabolism. J. Cell Biol. 59,346-360.

MCINTOSH, F. C., and COLLIER, B. (1976). Neurochemistry of cholinergic terminals. In “Neuromuscular Junction” (E. Zaimis, ed.), pp. 99- 228. Springer-Verlag, Berlin.

NISHI, R., and BERG, D. (1977). Dissociated ciliary ganglion neurons in vitro: Survival and synapse formation. Proc. NatL Acad. Sci. USA 74,5171-5175.

NISHI, R., and BERG, D. (1981). Two components from eye tissue that differentially stimulate the growth and development of ciliary gan- glion neurons in cell culture. J. Neurosci. 1,505-513.

PATTERSON, P. H., REICHARDT, L. F., and CHUN, L. Y. (1975). Biochemical studies on the development of primary sympathetic neurons in cell culture. Cold Spring Harbor Symp. Quant. BioL 40,389-397.

PILAR, G., and TUTTLE, J. B. (1982). A simple neuronal system with a range of uses: The avian ciliary ganglion. In “Progress in Cholinergic Biology: Model Cholinergic Synapses” (I. Hanin and A. M. Goldberg, eds.), p, 213. Raven Press, New York.

PILAR, G., TUTTLE, J. B., and LANDMESSER, L. (1982). Neuromuscular interdependence during development. In “Disorders of the Motor Unit” (D. L. Schotland, ed.), pp. 119-136. Wiley, New York.

PILAR, G., TUTTLE, J., and VACA, K. (1981). Functional maturation of motor nerve terminals in the avian iris: Ultrastructure, transmitter metabolism and synaptic reliability. J. PhysioL (London) 321, 175- 193.

PROCHIANTZ, A., DAGUET, M. C., HERBET, A., and GLOWINSKI, J. (1981). Specific stimulation of in vitro maturation of mesencephalic dopa- minergic neurons by striatal membranes. Nature (London) 293,570- 572.

PROCHIANTZ, A., DIPORZIO, U., KATO, A., BERGER, B., and GLOWINSKI, J. (1979). In vitro maturation of mesencephalic dopaminergic neurons from mouse embryos is enhanced in the presence of their striatal target cells. Proc. NatL Acad. Sci. USA 76,5387-5391.

SUSZKIW, J. B., and PILAR, G. (1976). Selective localization of a high affinity choline uptake system and its role in ACh formation in cho- linergic nerve terminals. J. Neurochem. 26,1133-1138.

TUTTLE, J. B. (1977). Dissociated cell culture of chick ciliary ganglion: Survival and development with and without target tissues. Sot. Neurosci. Abstr. 3, 529.

TUTTLE, J. B. (1983). Interaction with membrane remnants of target myotubes maintains transmitter sensitivity of cultured neurons. Science 220, 977-979.

TUTTLE, J. B. (1980). Long-term survival and development of dissociated parasympathetic neurons in culture. Brain Res. 183,161-180.

TUTTLE, J. B. (1985). Regulation of cholinergic development by target contact. Fed. Proc. 44,2760-2766.

TUTTLE, J. B., VACA, K., and PILAR, G. (1983). Target influences on ‘H- ACh synthesis and release by ciliary ganglion neurons in vitro. Dev. BioL 97,255-263.

VACA, K., JOHNSON, D., and PILAR, G. (1982). Modulation of transmitter synthesis and release in cholinergic terminals. .Z PhysioL (Paris) 78,385-391.

VACA, K., and PILAR, G. (1979). Mechanisms controlling choline trans- port and acetylcholine synthesis in motor nerve terminals during electrical stimulation. J. Gen. PhysioL 73, 605-628.

VACA, K., TUTTLE, J. B., and PILAR, G. (1979). Trophic effects of muscle on nerve: I. Acetylcholine synthesis. Sot. Neurosci. Abstr. 5, 771.

WELDON, P. R., and COHEN, M. W. (1979). Development of synaptic ultrastructure at neuromuscular contacts in an amphibian cell cul- ture system. .Z Neurocytol. 8,239-259.

WICHA, M. S., LOWRIE, G., KOHN, E., BAGAVANDOSS, P., and MAHN, T. (1982). Proc. NatL Acad. Sci. USA 79, 4523.

YAMAMURA, H. I., and SNYDER, S. H. (1973). High affinity transport of choline into synaptosomes of rat brain. J. Neurochem. 21,1355-1374.