This study was carried out in order to determine what factors control the differentiation of certain neural crest cells in the chick embryo. Emphasis was placed on the morphologically and biochemically divergent sensory and sympathetic pathways of differentiation. Embryos were precisely staged according to Hamburger & Hamilton (1951) and it was observed that sensory ganglia with somites, explanted at stages 21–24, gave rise to cells showing formaldehyde-induced fluorescence in more than 25% of explants. These cells were identical in properties to the fluorescent cells of the sympathetic system of embryos of similar age, and appeared by 12 days in vitro. These fluorescent cells did not appear when somites and sensory ganglia explants were maintained separately.

The incidence of fluorescent cells in combined explants was considerably reduced or absent when cultures were maintained for 7 days or less, or when the explants were obtained from stage 25–26 embryos. Furthermore, when neural tube was also included in the cultures, the appearance of fluorescent cells was markedly inhibited. The requirement for somitic tissue to induce fluorescent cells in combined explants can be replaced by forelimb-bud tissue.

The origin of these cells and the factors that control their differentiation in vitro are discussed with reference to the neural crest origin of the sensory ganglion, and the possible conditions pertaining in vivo in this region.

After their initial condensation above the neural tube, cells of the vertebrate neural crest undergo a period of extensive migration, becoming widely dispersed from their original position. Derivatives of the neural crest are diverse, including sensory ganglion neurons, sympathetic neurons and the chromaffin cells of the suprarenal medullary tissue. The neural crest also gives rise to the intrinsic neurons of the heart, lungs and gut, some Schwann cells and possibly some supportive cells of the ganglia of the peripheral nervous system, non-retinal pigment cells, as well as a variety of mesenchymal cell types (see Weston, 1970, for review). After localization, the development of diverse morphological and histochemical properties enable neural crest cells to be easily identified in many cases (Hamburger & Levi-Montalcini, 1949; Gunn, 1951; Strumia & Baima-Bollone, 1964; Enemar, Falck & Hakanson, 1965; Pearse & Polak, 1971; Polak, Rost & Pearse, 1971; Fernholm, 1971, 1972).

Despite knowledge of the extent of neural crest developmental potential, and detailed study of the differentiated state of many derivatives, the factors controlling the differentiation of any specific cell type or the choice between one developmental pathway rather than another are still largely unknown. Heterochronic grafting of chick neural tube and crest suggest that, at any one level trunk neural crest cells are labile with respect to their later differentiation, at least along the morphologically and biochemically divergent sensory and sympathetic pathways (Weston & Butler, 1966). There is a preponderance of pigment cell differentiation in vitro from amphibian (Model & Dalton, 1968), and chick (Dorris, 1938) neural crest. This can also be interpreted from the view that at any particular axial level the environment in which neural crest cells find themselves during and after migration, controls their subsequent differentiation (see Weston, 1970). However, it has been shown that some regionalization exists down the embryonic axis, e.g. cartilage can be formed only from head neural crest in amphibians (Chibon, 1967), and possibly also in the chick (Johnston, 1966). Similarly chick suprarenal medullary cells are apparently provided only by trunk neural crest between somites 12 and 29 (Chevallier, 1972).

Initial direct experimental work on factors controlling neural crest cell differentiation consisted of grafting onto the chorio-allantoic membrane (CAM) (Cohen, 1972), or growing in tissue culture (Norr, 1973) various portions of the to 2-day-old chick embryo.

These indicated that cells, apparently identical to sympathetic neurons, still appeared even when the normal region of localization of the primary sympathetic chain was not included in the graft. Furthermore a requirement for somitic mesenchyme in this developmental pathway which could not be replaced by either heart or limb-bud mesenchyme in these conditions was also demonstrated. Ventral neural tube also appeared to favour this line of differentiation. However, neural tube and crest in an organ culture system (Bjerre, 1973) produced sympathetic-like cells even when explanted alone, suggesting that an inductive interaction with somitic tissue may not be absolutely necessary for sympathetic cell differentiation.

The degree to which the lability of neural crest differentiation is retained after localization has been examined by Cowell & Weston (1970). In an analysis of cells of the 4-day-old chick embryo sensory ganglion in vitro, it was observed that many cells differentiated as melanocytes, although normally these cells are fated to develop as sensory neurons or their supporting cells. This ability declined with donor age, no melanocytes being observed in explants from donors older than six days’ incubation. Older ganglion cells also spread far less in culture than the younger ganglion cells. A reduced amount of pigmentation was also observed when the younger ganglia were explanted on agar, which restricts cellular outgrowth. This suggests that cell dispersal may be a necessary prerequisite for the melanocyte trait to appear. In addition, some intrinsic restriction on the ability to form this unusual cell type intervenes. This investigation is concerned with the ability of the chick sensory ganglion to produce sympathetic-like cells in vitro, and the effect of other embryonic structures on this unusual developmental pathway.

Preparation

White Leghorn-Black Australorp cross chick embryos of –5 days’ incubation (stages 21–26, Hamburger & Hamilton, 1951) were transversely sectioned into three somite-wide blocks between somite 11 and somite 22. Each piece was then frontally sectioned immediately ventral to the neural tube. The dorsal and ventral pieces so obtained were then divided along the mid-line. Further dissection of the ventral piece (VP) was restricted to trimming away lateral somitic mesenchyme, heart, gut, limb-buds and notochord where these occurred. The dorsal piece, consisting of half neural tube (NT), three sensory ganglia (SG) and dorsal somites with ectoderm (S), could then be divided into these respective components as desired for various recombination experiments. Occasionally a reduced amount of neural tube (nt) of only one somite length was used. Forelimb-bud tissue (FLB) used in some experiments was obtained from the middle section of the limb-bud, and was approximately the same size as the somite pieces.

The dissections were performed with cataract knives under a binocular dissecting microscope, with the tissues completely immersed in a dissecting medium consisting of Eagle’s Basal Medium with 10% horse serum (Commonwealth Serum Laboratories, Melbourne, hereafter referred to as C.S.L.).

Tissue culture

Tissue fragments were explanted singly or recombined in Rose (1954) chambers on glass coverslips (Lomb Co. Sydney) under 1 cm wide strips of dialysis cellophane (Visking Co. Chicago, size 27/32, average pore radius 2–4 nm), as described by Rose, Pomerat, Shindler & Trunnell (1958). Cultures were maintained for between 4 and 19 days at 37 °C in a medium of 10% horse serum and 4% 9-day-old chick embryo extract in Eagle’s Basal Medium (C.S.L.). Some cultures were provided with medium 199 (Salk, Younger & Ward, 1954) with 20% foetal calf serum (C.S.L.); 5 mg/ml glucose; 0-05 units/ml insulin; 100 units/ml penicillin; and 1 unit/ml Nerve Growth Factor (Burroughs Wellcome, U.K.). The culture medium was changed every three days.

Explants were routinely examined with a Wild M 40 inverted phase contrast microscope. Living cultures were photographed with a Zeiss Ikon camera mounted on a Zeiss Standard RA microscope.

Histochemistry

Formaldehyde-induced fluorescence (FIF) of catecholamines was demonstrated after removing the strip of dialysis cellophane and washing the cultures briefly in balanced salt solution. Cultures were dried on the coverslips in a vacuum dessicator over phosphorus pentoxide for at least 1 h, and then incubated over paraformaldehyde powder (Merck, Darmstadt) at 80 °C for 1 h. The coverslips were mounted on microscope slides with liquid paraffin and examined with a Leitz Ortholux microscope with HBO 200 mercury lamp and 3 mm BG 38, 3 mm BG 12 and K 530 filters, and a light field condenser. A Leitz Orthomat automatic camera was used for photomicrography.

Non-specific autofluorescence was detected by incubating some cultures without paraformaldehyde. In addition, all cultures with, fluorescent cells were washed gently with running water for 1–2 min then re-examined. Structural autofluorescence is not diminished by this treatment whereas the intensity of catecholamine fluorescence declines markedly.

After FIF treatment, some cultures were stained with toluidine blue (Humason, 1962) to reveal cartilage, or a von Kossa method (Mallory, 1961), slightly modified, to detect calcium deposits.

(1) Explants of ventral piece (VP)

Cultures of VP from stages 21–26 showed rapid cellular outgrowth within a few days, and myotubes and cartilage nodules were prominent. FIF cells were present in large numbers in all cases after a period of between 4 and 12 days in vitro. Long and complex fluorescent axon networks were frequently observed, especially in the older cultures (Fig. 1). The average nuclear diameter of fluorescent cells, as measured from fluorescence photomicrographs (Fig. 2), was 6·3 μm (range 5·0–8·8 μm; 53 cells; 12 days in vitro). The average nuclear diameter of neurons and neuroblasts from phase contrast photomicrographs was 6·0 (range 5·0–7·3 μmi; 35 cells; 12 days in vitro). Not all neurons in culture were fluorescent (Fig. 3). These cells were similar in appearance to their fluorescent neighbours, although they tended to be slightly smaller when examined with phase contrast optics. It is possible that these non-fluorescent cells were not adrenergic neurons, or they may have been metabolically resting (Yamauchi, Lever & Kemp, 1973; Benitez, Murray & Cote, 1973), or simply have been immature. The fluorescent cells could not be divided into groups on the basis of their fluorescence intensity or size: all appeared small and brightly fluorescent (cf. Chamley, Mark, Campbell & Burnstock, 1972a).

Fig. 1

VP explant of stage 21 donor, 12 days in culture after FIF treatment. Note fluorescent cells and long processes. Scale = 100 μm.

Fig. 1

VP explant of stage 21 donor, 12 days in culture after FIF treatment. Note fluorescent cells and long processes. Scale = 100 μm.

Fig. 2

VP explant of stage 23 donor, 12 days in culture after FIF treatment. A small group of fluorescent cells. Scale = 10 μm.

Fig. 2

VP explant of stage 23 donor, 12 days in culture after FIF treatment. A small group of fluorescent cells. Scale = 10 μm.

Fig. 3

Same area as Fig. 2, phase contrast optics. Fluorescence is more frequently displayed by the larger cells of the group (cf Fig. 2). Scale = 10 μm.

Fig. 3

Same area as Fig. 2, phase contrast optics. Fluorescence is more frequently displayed by the larger cells of the group (cf Fig. 2). Scale = 10 μm.

(2) Explants of somite and sensory ganglia

Sensory ganglia (SG) explanted alone gave considerable numbers of neurons that normally occurred in smaller groups than those typically observed in VP cultures, and in addition the cells had much larger nuclei of 8·9,μm average diameter (range 7·0–11·3 μm; 125 cells; 12 days in vitro; phase contrast optics). The cells also appeared to have a smaller nucleo-cytoplasmic ratio. These groups of neurons were interconnected by a very extensive axonal network, which was usually accompanied by rather flattened supporting cells (Fig. 4), although definitive Schwann cells closely applied to axons and glial cells were also observed. A sheet of fibroblast-like cells separated neurons and axonal networks from the glass of the culture chamber. No clearly identifiable melanocytes were observed beneath the strip of dialysis cellophane (cf. Cowell & Weston, 1970), although occasionally distinct black or brown pigment cells were observed at the edges of the cellophane. Cultures containing even one cell with FIF after 12 days in culture were very rare (see Table 1).

Table 1

Occurrence of FIF cells in cultures of chick embryo tissue

Occurrence of FIF cells in cultures of chick embryo tissue
Occurrence of FIF cells in cultures of chick embryo tissue
Fig. 4

SG explant of stage 21 donor, 5 days in culture, phase contrast optics. Non-fluorescent sensory neurons and supporting cells are shown. Note the size difference compared to the fluorescent cells of Fig. 3. Scale = 10 μm.

Fig. 4

SG explant of stage 21 donor, 5 days in culture, phase contrast optics. Non-fluorescent sensory neurons and supporting cells are shown. Note the size difference compared to the fluorescent cells of Fig. 3. Scale = 10 μm.

Somite explants (S) showed large numbers of myotubes by 6 days in vitro, but spontaneous contractions were rarely observed. Cartilage nodules were infrequently observed in explants from donors of stage 23 or younger, even after 19 days in vitro. Older donors produced cartilage far more frequently and often in large amounts. In addition to these differentiation end states, almost all cultures initially showed condensed flattened bars or strands of contiguous cells, which by 7 days in vitro usually displayed a definite cell free central core with a surrounding layer of flattened cells. After 12 days in culture this central core was shown to have accumulated considerable amounts of calcium salts, and probably therefore represents a premature differentiation of sclerotomal cells along an osteogenic pathway. These explants were surrounded by flattened fibroblastic cells, and occasionally a few large, presumably sensory neurons could be seen, but FIF cells were rare even after 12 days in vitro (see Table 1).

Somite and sensory ganglia explants cultured as one piece (S:SG) showed a combination of the differentiation end states already described, although the neurons were confined largely to the outer fibroblastic layer. One significant deviation from the above observations, however, was the appearance of cells with FIF in over 25% of cultures from stage 21–24 donors after 12 days in vitro (S: SG versus S, P < 0·1% per cent; S: SG versus SG, P = 1%, using x2 test with Yates’ correction; see Table 1). Furthermore, virtually no fluorescent cells were observed in explants from stages 21-24 in the first 7 days in vitro (12 days S:SG versus 7 day S:SG, P = 1%), or when the explants were taken from embryonic stages 25–26 (stages 21–24 S:SG versus stages 25–26 S: SG, P < 5%). In addition a slightly increased frequency of FIF cellcontaining cultures was observed on prolonged cultivation of explants (19 days), or by the utilization of enriched medium, or by dividing and then recombining somite and sensory ganglia explants prior to cultivation (S/SG). These increases however were not significant in the numbers performed.

These fluorescent cells when present, occurred in the same axonal network as the large non-fluorescent ganglion cells, and were in small numbers (average 8·3 per culture, range 1–33), in one or two loose groups confined to a small proportion of the total neural area. These fluorescent cells were of small size (average nuclear diameter 6·7 μm, range 5·5–8·5 μm, 27 cells measured after FIF treatment), and were often fusiform in outline (Fig. 5) although cells with larger numbers of short fluorescent processes were also common (Fig. 6). Cells with fluorescent processes exceeding 50 μm in length were very rare, but it could not be established whether this was due to the absence of such long processes or merely to the lack of catecholamine. No increase in the frequency of long fluorescent cell processes was observed even in cultures of 19 days duration.

Fig. 5

S:SG explant of stage 23 donor, 12 days in culture, after FIF treatment. Single fusiform fluorescent cell. Scale = 10 μm.

Fig. 5

S:SG explant of stage 23 donor, 12 days in culture, after FIF treatment. Single fusiform fluorescent cell. Scale = 10 μm.

Fig. 6

S:SG expiant of stage 23 donor, 19 days in culture, after FIF treatment. Group of cells with short fluorescent processes. Scale = 10 μm.

Fig. 6

S:SG expiant of stage 23 donor, 19 days in culture, after FIF treatment. Group of cells with short fluorescent processes. Scale = 10 μm.

(3) Explants of forelimb-bud (FLB) and forelimb-bud combined with sensory ganglia (FLB/SG)

Fragments of FLB from embryos of stages 21–24 showed marked similarities in cell types represented, to explants of somite, although cartilage was present in great amounts even in explants grown from stage 21 embryos. Exposure to formaldehyde gas after 12 days in vitro revealed no fluorescent cells. Similar fragments when combined with sensory ganglion (FLB/SG) produced fluorescent cells identical in appearance and distribution to those observed in somite recombined with sensory ganglia (S/SG) cultures (Fig. 7). Recombinants from donors of stages 25–26 again did not produce fluorescent cells (Table 1).

Fig. 7

FLB/SG explant of stage 23 donor, 12 days in culture, after FIF treatment. Single cell with long fluorescent process. Scale = 10 μm.

Fig. 7

FLB/SG explant of stage 23 donor, 12 days in culture, after FIF treatment. Single cell with long fluorescent process. Scale = 10 μm.

(4) Explants involving neural tube

Explants of stages 21–24 S:SG with attached neural tube (S:SG:NT) showed mesenchymal derivatives similar to S: SG cultures, although contractions of myotubes were common after 6 days in vitro. As anticipated, the axonal networks were much more extensive as compared to S:SG cultures, with a large percentage of axons apparently without supporting cells (cf. SG and S:SG cultures). Large neurons similar in appearance to the non-fluorescent ganglion cells of SG explants, occurred in small groups near to the neural tube, apparently in similar or slightly fewer numbers as compared to explants without NT. It is possible that some neurons were overlooked due to their proximity to the neural tube. After 12 days in vitro no fluorescent cells could be found in these cultures.

Explants of S: SG with NT dissected free and recombined at explantation in random orientation to the other tissue fragment (S: SG/NT), also showed this lack of fluorescent cells after 12 days in vitro. Extension of the culture period to 19 days gave the same result, but reduction of the amount of NT tissue to about one third (S:SG:nt) showed fluorescent cells in a few cultures. This was far below the frequency of fluorescent cells observed in explants without neural tube (S:SG versus S:SG:nt, P < 1%), and suggests that the neural tube has an inhibitory effect on the differentiation of FIF cells, at least in these conditions.

Standard histological techniques have demonstrated a clear separation between the cell aggregates of the sensory ganglion and the primary sympathetic chain in the chick embryo, at least from about stage 19 onwards. Autoradiography has revealed the presence of a few neural crest or neural tube derived cells lying along the ventral nerve roots. These cells are probably presumptive Schwann cells, rather than neuronal precursors (Weston, 1963).

Cells containing catecholamine are confined to the primary sympathetic chain and its ventral extensions into the adrenal medulla, between day 3 and 4 of incubation (Enemar et al. 1965). No cells with demonstrable FIF occur in the sensory ganglion even after loading with DOPA, a catecholamine precursor (Polak et al. 1971). Migration of cells from the primary sympathetic chain to form the secondary sympathetic complex closer to the sensory ganglion has commenced after 4 days’ incubation (about stage 25), when fluorescent cells can be seen extending dorsally from the primary position (Enemar et al. 1965). It is possible however that non-fluorescent cells have commenced this migration at an earlier stage though these cannot be revealed by loading with a-methyl noradrenaline or DOPA (Allen & Newgreen, unpublished observations). The formation of the secondary chain and increased recruitment of FIF cells into it, continues from day 5 to day 7 of incubation (stage 26 onwards; Enemar et al. 1965). It is therefore unlikely that sympathetic precursor cells are included in the dorsal explants at this time of development (stages 21–26). The scarcity of FIF cells in explants of somite (S) or sensory ganglia (SG) tends to confirm this, since sympathetic ganglion cells survive well for many weeks in vitro even when denied pre- and post-ganglionic connexions (Chamley et al. 1972a; Benitez et al. 1973). Indeed the definitive sympathetic system of embryos of the same age showed no significant decrease in fluorescent cell numbers between 4 and 19 days in culture.

The fluorescent cells observed on culturing somite (S) combined with sensory ganglia (SG) are therefore unlikely to be derived from cells normally fated as sympathetic neurons, although they bear a great resemblance to the latter. Their exact status could not be precisely established since both these cells and the normal sympathetic cells in culture resembled SIF cells in their small size and bright fluorescence (cf. Chamley et al. 1972b). The absence of long fluorescent processes was also common to many true sympathetic cells.

The most likely source of those FIF cells is therefore the sensory ganglion. This view is supported by the FLB/SG results, since the mid-FLB explants used here could be expected to contain virtually no neural crest derived cells except for the first of the melanoblasts or their precursors in the epidermis, before stage 24 (Fox, 1949). This view parallels the observation that the sensory ganglion at a similar stage in development contains a population of cells capable of differentiating as melanocytes under certain conditions, rather than as sensory neurons or their supporting cells as would be the case normally (Cowell & Weston, 1970). Presumably these cells are numbered amongst the medio-dorsal cells of Hamburger & Levi-Montalcini (1949). The length of time necessary for the appearance of FIF cells, greater than 7 days in vitro, is consistent with the view that these cells are expressing an entirely new cell phenotype as a result of the conditions in culture. Again, like the production of melanocytes from the sensory ganglion (Cowell & Weston, 1970), the appearance of FIF cells is dependent upon the age of the donor, although this ability declines with age even more rapidly than that of melanin synthesis.

In these explants of sensory ganglion, comparable in age to Cowell & Weston’s (1970) study, but in different culture conditions, overtly differentiated pigment cells were lacking. This cannot be due to genetic factors (see Dorris (1938) on pigment cell differentiation in vitro from this breed), nor by a prevention of dispersal of their cellular precursors as occurs in explants on agar (Cowell & Weston, 1970) or on the CAM (Dorris, 1941). On the contrary, dispersal seems to be enhanced under the cellophane strip. It is therefore of interest to note the occasional presence of pigment in some cells that had escaped from under the cellophane. It is uncertain whether the absence of pigmented melanocytes indicates that more cells could be available for other lines of differentiation, but evidently none or few of these express themselves as catecholamine synthesizing cells in SG explants. In plasma clot cultures the decrease in pigmentation observed when somitic tissue is explanted with sensory ganglion could indicate that fewer cells embark upon this line of differentiation, and in fact mature ganglion cell numbers seem to be increased (Peterson & Murray, 1955). A reciprocal relationship between pigmentation and neuronal development does seem to exist in sensory ganglia in vitro, the degree of cell dispersal being influential in the development of one phenotype rather than the other (Weston, 1971).

Nevertheless, the small numbers of fluorescent cells in cultures involving three sensory ganglia as well as somite, compared with the large numbers of pigment cells obtained by Cowell & Weston (1970), from similar aged donors, indicate that only a small percentage of cells potentially available for transformation actually expresses the sympathetic line. This is unlikely to be due to immaturity of the catecholamine synthesizing or storage mechanisms (see Ignarro & Shideman, 1968), since 19 day cultures showed little increase in fluorescent cell numbers over 12 day cultures. It is known however that not all adrenergic cells in sympathetic ganglia exhibit FIF at any one time, both in vivo (Yamauchi et al. 1973) and in vitro (Benitez et al. 1973). Rather than the possibility that the sensory ganglion is inherently less capable of producing fluorescent cells than melanocytes, it is more likely that this culture system does not easily allow this line of differentiation, in much the same way as melanocyte differentiation is inhibited at some level in the same cultures. It is clear, however, from observations of VP cultures, that the method is capable of cell maintenance if not the expression of final differentiated end states. The experiments of Cohen (1972) and Norr (1973) indicate that early somitic tissue and ventral neural tube induce normal sympathetic differentiation, although Bjerre (1973) has reported the appearance of fluorescent cells in cultures of hind brain neural crest alone. In the present cultures, slightly older somitic mesenchyme could also elicit this type of differentiation, but the disorder produced by explantation and the distribution of fluorescent cells in small isolated groups suggest that the mere presence of somitic tissue is insufficient in itself. Perhaps more precise microenvironmental factors must be present, which are met by only a relatively few cells potentially capable of responding to them and subsequently expressing FIF differentiation.

Limb-bud or heart mesenchyme could not duplicate the effects of early somitic mesenchyme in promoting sympathetic differentiation from neural crest on the CAM (Cohen, 1972), although crest cell migration was extensive. In the present work, forelimb-bud tissue proved just as effective as older somitic mesenchyme in the development of sympathetic cells from the sensory ganglion. It is uncertain whether this represents some response factor inherent to the sensory ganglion. However, the similarities of somite and forelimb-bud explants under these culture conditions have already been noted.

Although somite and neural tube are thought to be important for sympathetic cell differentiation (Cohen, 1972) the sensory ganglion, lying adjacent to the neural tube and embedded in the somite, shows a complete absence of these characteristics. No fluorescent cells can be detected from the time when these first appear in the primary sympathetic chain (Enemar et al. 1965), even after loading with the catecholamine precursor DOPA (Polak et al. 1971) or with a-methyl noradrenaline (Allen & Newgreen, unpublished observations). The ability to take up catecholamines is lacking at later stages (Burdman, 1968; England & Goldstein, 1969).

The different characteristics of sensory and sympathetically fated cells appear to be enforced only after these cells have left the neural crest (Weston & Butler, 1966). Although derived from the same axial levels (Yntema & Hammond, 1954, 1955), the cells ultimately destined as sympathetic cells migrate before those fated to form the sensory ganglion (Weston, 1963). Sympathetic cell precursors pass through the somite close to the neural tube before localizing near the dorsal aorta, where the differentiated state in the form of catecholamine accumulation is normally first manifested (Enemar et al. 1965). Neural crest cells destined to form the sensory ganglion do not migrate so extensively, and aggregate next to the neural tube with which they are continuously associated. In tissue culture, the presence of the neural tube drastically inhibited the appearance of FIF cells from explants of sensory ganglion and somites (Table 1). It is possible that the neural tube has a similar inhibitory effect in vivo. It should be noted that the migration of sympathetic cells to the position of the secondary sympathetic chain, closer to the neural tube, occurs only after most of the cells have to a large degree differentiated, and are already capable of displaying fluorescence.

The action of the neural tube however may not be this specific. In these cultures, neural tube did not suppress FIF cells by significantly promoting sensory neuron differentiation from the limited pool of cells available. Peterson & Murray (1955) found that central nervous tissue actually increased the number of degenerating neurons in explants of embryonic sensory ganglia. Thus the suppression of FIF cells by the neural tube may merely reflect an inhibition of neuronal differentiation on a population of cells which, differentiating from a relatively labile state entirely in vitro, may be especially sensitive to its action. The condensation of neural crest cells may be promoted adjacent to the neural tube (Weston, 1970), but subsequent differentiation, although restricted to neuronal lines (Weston, 1971), may be inhibited or retarded. Indeed, sensory neuronal differentiation appears first in the ganglion cells furthest from the neural tube (Hamburger & Levi-Montalcini, 1949), and possibly lends further support to the concept of the selective inhibitory action of the neural tube on the differentiation of various neuronal elements.

Benitez
,
H. H.
,
Murray
,
M. R.
&
Cote
,
L. J.
(
1973
).
Responses of sympathetic chain ganglia isolated in organotypic culture to agents affecting adrenergic neuron fluorescence histochemistry
.
Expl Neurol
.
39
,
424
448
.
Bjerre
,
B.
(
1973
).
The production of catecholamine-containing cells in vitro by young chick embryos studied by the histochemical fluorescence method
.
J. Anat
.
115
,
119
131
.
Burdman
,
J. A.
(
1968
).
Uptake of 3H catecholamine by chick embryo sympathetic ganglia in tissue culture
.
J. Neurochem
.
15
,
1321
1323
.
Chamley
,
J. H.
,
Mark
,
G. E.
,
Campbell
,
G. R.
&
Burnstock
,
G.
(
1972a
).
Sympathetic ganglia in culture. I. Neurons
.
Z. Zellforsch. mikrosk. Anat
.
135
,
287
314
.
Chamley
,
J. H.
,
Mark
,
G. E.
&
Burnstock
,
G.
(
1972b
).
Sympathetic ganglia in culture. IL Accessory cells
.
Z. Zellforsch. mikrosk. Anat
.
135
,
315
327
.
Chevallier
,
A.
(
1972
).
Localisation et durée des potentialités médullo-surrénaliennes des crêtes neurales chez l’embryon de poulet
.
J. Embryol. exp. Morph
.
27
,
603
614
.
Chibon
,
P.
(
1967
).
Marquage nucléaire la thymidine tritiée des derives de la crête neurale chez l’Amphibien Urodèle Pleurodeles wait Hi Michah
.
J. Embryol. exp. Morph
.
18
,
343358
.
Cohen
,
A. M.
(
1972
).
Factors directing the expression of sympathetic nerve traits in cells of neural crest origin
.
J. exp. Zool
.
179
,
167
182
.
Cowell
,
L. C.
&
Weston
,
J. A.
(
1970
).
An analysis of melanogenesis in cultured chick embryo spinal ganglia
.
Devi Biol
.
22
,
670
697
.
Dorris
,
F.
(
1938
).
The production of pigment in vitro by chick neural crest
.
Wilhelm Roux Arch. EntwMech. Org
.
138
,
323
334
.
Dorris
,
F.
(
1941
).
The behaviour of chick neural crest in grafts to the chorio-allantoic membrane
.
J. exp. Zool
.
86
,
205
223
.
Enemar
,
A.
,
Falck
,
B.
&
Hakanson
,
R.
(
1965
).
Observations on the appearance of noradrenaline in the sympathetic nervous system of the chick embryo
.
Devi Biol
.
11
,
268
283
.
England
,
J. M.
&
Goldstein
,
M. N.
(
1969
).
The uptake and localisation of catecholamines in chick embryo sympathetic neurons in tissue culture
.
J. Cell Sci
.
4
,
677
691
.
Fernholm
,
M.
(
1971
).
On the development of the sympathetic chain and the adrenal medulla in the mouse
.
Z. Anat. EntwGesh
.
133
,
305
317
.
Fernholm
,
M.
(
1972
).
On the appearance of monoamines in the sympathetic systems and the chromaffin tissue in the mouse embryo
.
Z. Anat. EntwGesh
.
135
,
350
361
.
Fox
,
M. H.
(
1949
).
Analysis of some phases of melanoblast migration in Barred Plymouth Rock embryos
.
Physiol. Zool
.
22
,
1
22
.
Gunn
,
M.
(
1951
).
A study of the enteric plexuses in some amphibians
.
Q. JI Microsc. Sci
.
92
,
55
77
.
Hamburger
,
V.
&
Hamilton
,
H. L.
(
1951
).
A series of normal stages in the development of the chick embryo
.
J. Morph
.
88
,
49
92
.
Hamburger
,
V.
&
Levi-Montalcini
,
R.
(
1949
).
Proliferation, differentiation and degeneration in the spinal ganglia of the chick embryo in the absence of sensory input
.
J. exp. Zool
.
111
,
457
502
.
Humason
,
G. L.
(
1962
).
Animal Tissue Techniques
.
San Francisco
:
Freeman
.
Ignarro
,
L. J.
&
Shideman
,
F. E.
(
1968
).
Appearance and concentrations of catecholamines and their biosynthesis in the embryonic and developing chick
.
J. Pharmac. exp. Ther
.
159
,
38
48
.
Johnston
,
M. C.
(
1966
).
A radioautographic study of the migration and fate of cranial neural crest cells in the chick embryo
.
Anat. Rec
.
156
,
143
156
.
Mallory
,
F. B.
(
1961
).
Pathological Technique
.
New York
:
Hafner
.
Model
,
P. G.
&
Dalton
,
H. C.
(
1968
).
The uptake and localization of radioactive DOPA by amphibian melanoblasts in vitro
.
Devi Biol
.
17
,
245
271
.
Norr
,
S. C.
(
1973
).
In vitro analysis of sympathetic neuron differentiation from chick neural crest cells
.
Devi Biol
.
34
,
16
38
.
Pearse
,
A. G. E.
&
Polak
,
J. M.
(
1971
).
Cytochemical evidence of the neural crest origin of mammalian ultimobranchial cells
.
Histochemie
27
,
96
102
.
Peterson
,
E. R.
&
Murray
,
M. R.
(
1955
).
Myelin sheath formation in cultures of avian spinal ganglia
.
Am. J. Anat
.
96
,
319
355
.
Polak
,
J. M.
,
Rost
,
F. W.
&
Pearse
,
A. G. E.
(
1971
).
Fluorogenic amine tracing of neural crest derivatives forming the adrenal medulla
.
Gen. comp. Endocr
.
16
,
132
136
.
Rose
,
G. C.
(
1954
).
A separable and multipurpose tissue culture chamber
.
Tex. Rep. Biol. Med
.
12
,
1074
1083
.
Rose
,
G. C.
,
Pomerat
,
C. M.
,
Shindler
,
T. O.
&
Trunnell
,
J. B.
(
1958
).
A cellophane strip technique for culturing tissue in multipurpose culture chambers
.
J. biophys. biochem. Cytol
.
4
,
761
764
.
Salk
,
E. S.
,
Younger
,
J. S.
&
Ward
,
E. N.
(
1954
).
Use of color change of phenol red as the indicator in titrating poliomyelitis virus or its antibody in a tissue culture system. Appendix. Method of preparing medium 199
.
Am. J. Hyg
.
60
,
214
230
.
Strumia
,
E.
&
Baima-Bollone
,
P. L.
(
1964
).
Acetylcholinesterase activity in the spinal ganglia of the chick embryo during development
.
Acta anat
.
57
,
281
293
.
Weston
,
J. A.
(
1963
).
A radioautographic analysis of the migration and localization of trunk neural crest cells in the chick
.
Devi Biol
.
6
,
279
310
.
Weston
,
J. A.
(
1970
).
The migration and differentiation of neural crest cells
.
Adv. Morphogen
.
8
,
41
114
.
Weston
,
J. A.
(
1971
).
Neural crest cell migration and differentiation
.
In Cellular Aspects of Growth and Differentiation in Nervous Tissue
(ed.
D.
Pease
), pp.
1
22
.
Berkeley
:
University of California Press
.
Weston
,
J. A.
&
Butler
,
S. L.
(
1966
).
Temporal factors affecting localization of neural crest cells in the chicken embryo
.
Devi Biol
.
14
,
246
266
.
Yamauchi
,
A.
,
Lever
,
J. D.
&
Kemp
,
K. W.
(
1973
).
Catecholamine loading and depletion in the rat superior cervical ganglion (a formol fluorescence and enzyme histochemistry study with numerical assessments)
.
J. Anat
.
114
,
271
282
.
Yntema
,
C. L.
&
Hammond
,
W. S.
(
1954
).
Depletions and abnormalities in the cervical sympathetic system of the chick following extirpation of the neural crest
.
J. exp. Zool
.
100
,
237
263
.
Yntema
,
C. L.
&
Hammond
,
W. S.
(
1955
).
Experiments on the origin and development of sacral autonomic nerves of the chick embryo
.
J. exp. Zool
.
129
,
375
414
.