Expression of the avian antigen SMP (Schwann cell Myelin Protein, Mr 75–80 000), first characterized in the PNS with a monoclonal antibody as an early and strictly specific Schwann cell marker, was further studied in the CNS.

Comparing SMP immunoreactive areas in the different parts of the CNS with those expressing the Myelin Basic Protein (MBP), we showed a strict colocalisation of both phenotypes. In vitro, MBP+ oligodendrocytes express the surface antigen SMP as well.

SMP cellular expression was followed in situ and in culture using nervous tissues from embryos at different stages. We were thus able to detect an early expression of this marker by oligodendroblasts before the first appearance of MBP immunoreactivity.

We have also identified a subpopulation of SMP+/ MBP– and SMP+/GC– cells, which persists under our culture conditions as precursors remaining in an immature state.

The neural epithelium yields, besides neurons, cells that are closely associated with axons, which ensheath them and produce myelin, a structure that increases efficiency of nerve impulse conduction. These cells are Schwann cells in the peripheral nervous system (PNS) and oligodendrocytes in the central nervous system (CNS). Precursors of oligodendrocytes, generated in the subventricular zone of the developing CNS, migrate to the future white matter areas where they undergo terminal differentiation in contact with the axons. The segregation of astrocyte and oligodendrocyte lineages in the CNS has been studied in vivo and in vitro by following the onset of expression of specific molecular markers such as various myelin components and surface antigens (e.g. galactocerebrosides (GC)) for oligodendrocytes and the glial fibrillary acidic protein (GFAP) for astrocytes (Raff et al. 1983; Raff, 1989; Goldman et al. 1986; LeVine and Goldman, 1988a).

Like most other cell types that form the PNS, Schwann cells arise from the neural crest, a transient and pluripotent embryonic structure that emerges from the lateral ridges of the neural epithelium. Besides neurons and Schwann cells, the neural crest also gives rise to the satellite glial cells of the peripheral ganglia, the enteric glia, melanocytes, endocrine, paraendocrine cells and, at the cephalic level, to the so-called mesectoderm. Comparative studies aiming to understand the cellular and molecular events controlling glial cell differentiation in the PNS, e.g. the maturation of myelinating and non-myelinating Schwann cells along axons in the peripheral nerves, of satellite cells or enteric glia, respectively, in peripheral ganglia or in the gut, have so far been delayed by the lack of specific markers like those already known in the CNS.

In a previous study (Dulac et al. 1988), we reported that a monoclonal antibody (Mab), raised in mouse against a glycoprotein fraction of myelin purified from adult quail peripheral nerves, recognizes an antigenic determinant that we called SMP (Schwann cell Myelin Protein), which is expressed in the PNS by both myelinating and non-myelinating Schwann cells but not by intraganglionic satellite cells. The SMP antigen of the sciatic nerve appears in Western blot, under nonreducing conditions, as a doublet of Mr 75–80 000 and is detectable by immunocytochemistry on Schwann cells earlier than any other specific myelin component identified so far. Immunopurification of the SMP protein allowed us to determine several amino acid sequences derived from proteolytic digests that do not present any homology with already known proteic sequence (unpublished). SMP also exists in the CNS and, in SDS-PAGE, forms a single band at Mr80 000. We also described the existence of SMP-positive areas, corresponding to myelinated zones, in brain and spinal cord. However, the cell type expressing this antigen in the CNS was not clearly identified. The aim of the present work was to see whether or not SMP expression in the CNS is strictly specific to the oligodendrocyte cell type. We show that there is a strict colocalisation of myelin and SMP expression both in situ and in vitro at the cellular level by using anti-Myelin Basic Protein (anti-MBP) immunoreactivity as a marker for myelin. Furthermore, we demonstrate that the SMP phenotype appears in situ before the onset of MBP expression and that, in vitro, mature MBP- or GC-expressing oligodendrocytes as well as a subpopulation of immature oligodendrocytes synthesise the SMP antigen.

Sections

Early chick and quail embryos and tissues dissected from late embryos and adults were fixed in 4% paraformaldehyde in Ca2+-Mg2+-free PBS, embedded in OTC (Miles), quickly frozen and sectioned at 7 to 20 μm with a cryostat. Heads from young embryos (E5 to Ell) or isolated brains from older embryos and adult animals were cut in sagittal sections. The spinal cord and surrounding tissues were cut transversally.

Sections were collected on gelatin-coated microscope slides, air dried and immediately processed for immunofluorescence staining or stored at –20°C until used.

Cultures

Quail embryo brain and spinal cord were dissected at various ages, ranging from E5 to E16, chopped into pieces and mechanically dissociated into single-cell suspensions by passing through a Pasteur pipette and then repeatedly through an elongated pipette. The undissociated tissue pieces were allowed to sediment for 5 min and the resulting cell suspension was diluted to approximately 104 cells ml−1 in culture medium consisting of Dulbecco’s modified Eagle’s medium (DMEM, GIBCO) supplemented with 10 % newborn calf serum (NCS, GIBCO) and 2% Ell chick embryo extract (CEE); 100μl of the suspension were plated on 35 mm tissue culture dishes (Nunc). Cultures were maintained at 37°C in a humidified atmosphere of 5 % CO2 in air. The medium was completed to 1 ml the next day. For some experiments, dishes precoated with laminin (BRL) (20μml−1 for 2h at 37°C) were used. This greatly increased the attachment of cells from older embryos.

Alternatively, tissue dissociation was achieved by enzymatic digestion with 0·1% trypsin (GIBCO), 0·1% EDTA (SERVA) for 10 min at 37 °C followed by 5 to 20 min at room temperature. The reaction was stopped with FCS-supplemented medium, tissues were dissociated by gentle pipetting and cells recovered by centrifugation (5 min, 900g).

In some experiments, the brain was dissected into its major morphological parts, and each was dissociated and cultured separately.

Antibodies

Anti-SMP Mab was used as undiluted hybridoma supernatant or as 1:200 to 1:500 diluted ascitic fluid. Anti-tetanus toxin antiserum (rabbit) was prepared in our laboratory as previously described (Ziller et al. 1983); anti-Myelin Basic Protein (anti-MBP) antiserum (rabbit) was a gift from Dr K. Mikoshiba (University of Osaka, Osaka, Japan). Anti-GC antiserum (rabbit) was a gift from Dr G. Labourdette (INSERM Strasbourg, France). Fluorescein isothiocyanate-conjugated goat anti-mouse immunoglobulin (GAM–F1TC, Nordic) and tetra-methylrhodamine isothiocyanate-conjugated immunoglobulin (GAR–TRITC, Nordic) were used at 1:50 dilution.

Immunostaining

For immunostaining of adult brain and spinal cord sections, the slides were immersed in 95% ethanol or 0·2% Triton X-100 for 30 min and rinsed in PBS before applying anti-SMP or anti-MBP. This treatment was not necessary in the case of embryonic sections, which were directly incubated in a moist chamber for 1 h at room temperature or overnight at 4°C with anti-SMP monoclonal antibody, followed by GAM-FITC for Ih. When double staining was required, sections were subsequently incubated with anti-MBP and GAR-TRITC (Ih each) and mounted in Mowiol (Hoechst Chemical).

Cultures were fixed for I h, and subsequently incubated with anti-SMP followed by GAM-FTTC for I h each. All incubations were performed at room temperature. For anti-SMP and anti-MBP double staining, cultures were incubated for 20min in 0·1% Triton X-100 in PBS before anti-MBP antiserum followed by GAR-TRITC were applied for 1 h each. When double staining was performed with anti-SMP and anti-tetanus toxin antiserum, cultures were treated for 30 min with the toxin diluted in DMEM at 37 °C before fixation. Anti-toxin antiserum and GAR-TRITC were then applied as usual, before or after anti-SMP staining. Anti-GC immunostaining was performed on living cells before fixation and anti-SMP immunolabelling.

Immunocytochemistry on sections of CNS tissues

(1) Post-hatching and adult quail

When applied on sections of brain and spinal cord, anti-SMP Mab stained the white matter and the fibres that cross the gray matter. Examination of the same sections under polarized light to show up the birefringence of myelinated areas revealed that SMP-positive regions coincided with those where birefringence was observed.

Double-staining experiments carried out with anti-SMP Mab and anti-MBP antiserum confirmed the colocalisation of SMP and myelin. Single fibres and bundles of fibres clearly showed a bright fluorescence with both reagents (Fig. 1). No SMP staining was ever found in neurons and MBP-negative areas in general (Fig. 2).

Fig. 1.

Double immunofluorescent labelling of a section of adult quail cerebellum with anti-SMP Mab (A) and anti-MBP antiserum (B). Fibres of the cerebellar white matter are strongly immunoreactive with both reagents. Bar=100μm.

Fig. 1.

Double immunofluorescent labelling of a section of adult quail cerebellum with anti-SMP Mab (A) and anti-MBP antiserum (B). Fibres of the cerebellar white matter are strongly immunoreactive with both reagents. Bar=100μm.

Fig. 2.

Anti-SMP immunoreactivity in a parasagittal section of adult quail brain through a cerebellar deep nucleus. Neurons, which can be easily recognized by their morphology (arrows), are strictly SMP-negative, whereas fibres adjacent to the nucleus and those crossing it are brightly immunoreactive (arrowheads).Bar=150μm.

Fig. 2.

Anti-SMP immunoreactivity in a parasagittal section of adult quail brain through a cerebellar deep nucleus. Neurons, which can be easily recognized by their morphology (arrows), are strictly SMP-negative, whereas fibres adjacent to the nucleus and those crossing it are brightly immunoreactive (arrowheads).Bar=150μm.

(2) Quail and chick embryos

Sparse positive fibres and cell bodies became discernible in spinal cord sections from E9 in the quail and from E12 in the chick. They were confined to a narrow lateral strip of the ventral white matter along either side of the anterior median fissure. By E10 in the quail and E13 in the chick, the strip was thicker near the anterior median fissure and extended laterally as a thin rim (Fig. 3A). At this stage, the first MBP immunoreactivity was detected (Fig. 3B). By Ell in the quail and E15 in the chick, SMP-positive fibres were localized in the external area of the white matter around the whole spinal cord including the dorsomedial region. The width of the stained areas increased over the following days as did the intensity of immunoreactivity. Similarly, a few positive fibres were detectable in the brain stem and cerebellum from E12 in the chick; their number and staining intensity increased rapidly over the next few days.

Fig. 3.

Transverse section of spinal cord from E10 quail embryo. Detail of the ventral region. (A) Anti-SMP immunoreactivity; (B) anti-MBP immunoreactivity. SMP-positive cells and fibres (A) form a strip along either side of the anterior median fissure (arrow) that extends laterally (arrowheads), whereas the rare MBP-positive cells (B) detectable at this stage are confined to both sides of the fissure (arrow). Bar=50μm.

Fig. 3.

Transverse section of spinal cord from E10 quail embryo. Detail of the ventral region. (A) Anti-SMP immunoreactivity; (B) anti-MBP immunoreactivity. SMP-positive cells and fibres (A) form a strip along either side of the anterior median fissure (arrow) that extends laterally (arrowheads), whereas the rare MBP-positive cells (B) detectable at this stage are confined to both sides of the fissure (arrow). Bar=50μm.

In the cerebral cortex of the chick embryo, no SMP-positive fibres were seen at E15. However, some were present at E18.

Brain and spinal cord cultures

Brain (cerebrum, medulla oblongata, cerebellum, optic nerve, olfactive and optic lobes) and spinal cord cultures from quail embryos of various ages, ranging from E5 to E16 were examined after one to 10 days in vitro and SMP-positive cells were counted. Immunoreactivity was found on the cell surface and could be evidenced both on living and fixed cells.

(1) Onset of SMP expression

The appearance of SMP antigen in vitro was found to be related to the total age of the cells (i.e. the age of the embryo from which they had been taken, plus the length of time in culture) and also, for a given final age, to the age of the embryo from which they had been taken (Table 1).

Table 1.

Time of appearance of the SMP antigen in vitro (in days)

Time of appearance of the SMP antigen in vitro (in days)
Time of appearance of the SMP antigen in vitro (in days)
(a) Spinal cord cultures

cells whose total age was 10 days did not express SMP if they had been taken from an E5 embryo but did so if they had been removed from an E7 embryo. In the latter cultures, a few scarce SMP-positive cells with few processes appeared in the cultures when the total age of the cells had reached 10 days. Two days later, their number had increased and at 17 days many cells displayed strong SMP immunoreactivity. Numerous positive cells were detected in certain older cultures, but they were too dense to be counted accurately.

(b) Brain cultures

the first positive cells were not seen before 12 days of final age (regardless of the age at which the tissue had been removed). Their number did not increase considerably until 4 days later and, as in the case of spinal cord cultures, the number of SMP-positive cells at 16 days was much higher when they were derived from E13 embryos than when they were taken 3 days earlier.

(2) Morphology of the SMP-positive cells

The SMP-positive cells were of variable morphology, ranging from small round to large and irregular cell bodies with more or less branched processes. They were found singly or in groups of 2–3 cells. Staining was less intense on isolated cells than on the clusters. Younger cells had few short processes while older cells had many thin branched processes. In a given culture, cells with a variety of shapes were found (Fig. 4).

Fig. 4.

Anti-SMP immunoreactivity in vitro. (A) Four-day-old culture of spinal cord from E10 quail embryo. (B) Three-day-old culture of optic nerve from E13 quail embryo. SMP-positive cells display a great variety of cell shapes from small round cell bodies with few thin processes (A) to a more complex morphology with flat membranous extensions and numerous processes (B). Bar=25μm.

Fig. 4.

Anti-SMP immunoreactivity in vitro. (A) Four-day-old culture of spinal cord from E10 quail embryo. (B) Three-day-old culture of optic nerve from E13 quail embryo. SMP-positive cells display a great variety of cell shapes from small round cell bodies with few thin processes (A) to a more complex morphology with flat membranous extensions and numerous processes (B). Bar=25μm.

(3) Identification of the SMP positive cells

Some cultures from each series were tested, simultaneously with anti-SMP Mab, for the presence of tetanus-toxin-binding sites, and anti-MBP or anti-GC immunoreactivity in order to identify neurons and mature oligodendrocytes, respectively. Virtually all the tetanus-toxin-positive cells had a neuron-like morphology; they were abundant and distributed in clusters on underlying glial cells. No double-positive cells were found. SMP-positive branched cells were often associated with the clusters of tetanus-toxin-positive neurons.

Double staining was performed with anti-SMP Mab and anti-MBP antiserum in 3- and 5-day-old cultures from E13 and E16 quails, respectively; all MBP-positive cells were also SMP-positive.

In lumbar and cervical spinal cord cultures, the SMP+/MBP+ cells with several thin processes were found on areas of densely aggregated negative cells. Some isolated SMP+/MBP–cells were also detected, with the same morphology as MBP+ cells, but no SMP–/MBP+ cells were found (Fig. 5).

Fig. 5.

Double staining of a spinal cord cell culture with anti-SMP Mab (A) and anti-MBP antiserum (B). Spinal cord was removed from an E13 quail embryo and the culture fixed after 3 days. All MBP+ oligodendrocytes are SMP+ (arrows), whereas some SMP+ cells displaying the same morphology are MBP– (arrowheads). According to the distribution of SMP immunoreactivity in situ, we consider these SMP+ MBP– cells to be still immature non-myelinating oligodendrocytes. Bar=25 μm.

Fig. 5.

Double staining of a spinal cord cell culture with anti-SMP Mab (A) and anti-MBP antiserum (B). Spinal cord was removed from an E13 quail embryo and the culture fixed after 3 days. All MBP+ oligodendrocytes are SMP+ (arrows), whereas some SMP+ cells displaying the same morphology are MBP– (arrowheads). According to the distribution of SMP immunoreactivity in situ, we consider these SMP+ MBP– cells to be still immature non-myelinating oligodendrocytes. Bar=25 μm.

Cerebellum cultures consisted of densely packed, hemispherical aggregates of tetanus-toxin-positive neurons on glial cells. Small SMP+/MBP+ cells were found on top of neuronal aggregates. As in the spinal cord cultures, we detected some SMP+/MBP– cells in the same areas, but no SMP–/MBP+ cells were found (not shown).

Cultures from other brain regions (optic lobes, cerebrum, olfactive lobe, brainstem) confirmed that all MBP+ cells were also SMP+, although again the reverse was not true; in fact, here, SMP+/MBP– cells were more numerous than MBP+ cells (Fig. 6). In general, SMP+ cells were more abundant in cultures of cerebellum or spinal cord cells than in cultures of other parts of the CNS. As expected from the observations made in situ, they were particularly rare in cultures of cerebral hemispheres.

Fig. 6.

Anti-SMP (A) and anti-GC (B) immunoreactivities in a 4-day-old culture of brain removed from an E10 quail embryo. All GC+ cells express SMP but some rare SMP+ cells are GC– (arrow). As for SMP+/MBP-cells (Fig. 5) we consider the SMP+/GC-subpopulation as immature precursors of oligodendrocytes. Bar=20μm.

Fig. 6.

Anti-SMP (A) and anti-GC (B) immunoreactivities in a 4-day-old culture of brain removed from an E10 quail embryo. All GC+ cells express SMP but some rare SMP+ cells are GC– (arrow). As for SMP+/MBP-cells (Fig. 5) we consider the SMP+/GC-subpopulation as immature precursors of oligodendrocytes. Bar=20μm.

Double staining experiments performed with anti-SMP Mab and anti-GC antiserum in 2-, 5- and 7-day-old cultures from E10 quail embryo showed both the expression of SMP by all GC+ oligodendrocytes. Once again some SMP+ cells were GC–. This SMP+/GC– subpopulation of cells is smaller than the SMP+/ MBP-one.

We demonstrated in a previous article (Dulac et al.1988) that, in the PNS, the avian glial antigen SMP is expressed by Schwann cells from embryonic day 6, i.e. before the onset of myelination in the quail embryo. Hence, this antigen could be considered as one of the earliest appearing marker enabling myelinating and non-myelinating Schwann cells to be distinguished from other peripheral glial cells, i.e. satellite cells of sensory and autonomic ganglia and all enteric glial cells. As a very early marker of the Schwann cell lineage, SMP can thus be used to study the differentiation of this cell type in clonal cultures of neural crest cells (Dupin et al.1989) .

SMP immunoreactivity was also detected in the CNS. The present report is focused on the further characterisation of cells expressing this antigen in the brain and in the spinal cord.

We show that there is a strict colocalisation of myelin and SMP expression on brain and spinal cord sections from adult and post-hatching quail: only myelinated areas, detected by their birefringence under polarized fight or their MBP synthesis, displayed anti-SMP immunoreactivity. Moreover, every MBP+ fibre that we were able to identify was also SMP+.

We confirm in vitro the expression of SMP by mature MBP+ oligodendrocytes. We prepared tissue cultures from late embryonic (E16) quail brain and spinal cord and fixed them after 5 days. Following the observations that the two antigens are expressed in culture (Mirsky et al. 1980; Dulac et al. 1988) and that, at the correspond-ing age in vivo (P5), SMP and MBP are both detectable, we performed double immunofluorescence labelling with the anti-SMP Mab and the anti-MBP antiserum. As expected, every oligodendrocyte identified by anti-MBP staining was also found to express SMP. Some SMP+/MBP– cells were found in every culture examined, irrespective of the age of the embryo (E13 or E16) and the region of the CNS from which the nervous tissue had been removed. We obtained similar results with double-labelling experiments carried out on younger cells (2-, 5- and 7-day-old cultures from E10 quail), to detect GC, which is an earlier oligodendrocyte marker than MBP. The morphology displayed by the SMP+ cells matches the descriptions of oligoden-drocytes found in the literature: small round to large and irregular cel) bodies with more- or-less branched processes. We were not able to attribute a particular cell morphology to the immature non-myelinating oligodendrocytes MBP–/SMP+ or GC–/SMP+, although some authors have tended to correlate the cell shape of oligodendrocytes with a defined stage of differentiation (Kuhlmann-Krieg et al. 1988).

Concerning the neurons that we identified in our cultures by their tetanus-toxin-binding sites in doublestaining experiments, we were able to exclude that they express the SMP antigen.

We have studied in some detail the appearance of SMP immunoreactivity in different parts of the CNS during ontogeny in both quail and chick. We were thus able to compare our results with published descriptions of the myelination process in the chick embryo (Bensted et al. 1957; Hartman et al. 1979; Macklin and Weill, 1985; Cochran et al. 1983).

SMP immunoreactivity appeared almost simultaneously in spinal cord, brain stem and cerebellum. It was first detected in quail and chick CNS, at 9 days and 12 days of incubation, respectively, which are the equivalent developmental stages according to Zacchei (1961) where Bensted et al. (1957) detected at the same time a localized concentration of glial cells in the white matter close to the anterior median fissure. At E13 Sudan Black staining revealed the first imperfectly formed myelin sheaths. Their description of the increasingly thicker strip of myelinated spinal cord from E13 to E17 matches the pattern of distribution of SMP that we found on sections: in the different regions of the CNS, the faint SMP reactivity, when it first appeared, was limited to some cell bodies and some fibres or bundles. The number of stained fibres, the thickness of the bundles and the staining intensity increased with the age of the embryo, reaching the adult pattern around hatching.

Macklin and Weill (1985) confirmed this time course by studying appearance of myelin-specific markers. MBP appeared at E13 in the chick in the spinal cord and in the brain stem, at E16 in the cerebellum, at E17 in the optic nerve, at hatching in the cortex. 2’-3’-cyclic nucleotide 3’ phosphohydrolase (CNPase) activity appeared one day later. We can thus assert that (i) SMP is expressed by immature oligodendrocytes in the different regions of the CNS several days before the onset of myelination; (ii) the chronology of the appearance of SMP, particularly at the different levels of the spinal cord and in the cerebellum, does not exactly coincide with that of myelination in the chick, which occurs at different stages in different regions of the CNS, following a general caudal-to-rostral order (Macklin and Weill, 1985; Hartman et al. 1979).

Abney et al. (1981), studying the differentiation of cell culture of embryonic rat brain, established that there was a correlation between the first expression of GC by oligodendrocytes and GFAP by astrocytes in vitro and the timing of glial differentiation in vivo. We show here that SMP is first detectable in 3-day-old cultures of spinal cord from E7 quail embryos, i.e. when the total age of the cells is 10 days; in cultures of brain, a few positive cells can be detected at a total age of 12 days (Table 1). Thus, particularly as regards the spinal cord, the onset of SMP expression by oligodendrocytes appears nearly on schedule, compared to the ontogeny in vivo. On the other hand, the delay of SMP expression by 10-day-old cells when they were removed at E5 instead of E7, the small number of SMP+ cells of 16 days of total age when they were removed at E10 rather than E13 and the persistence in every culture of immature oligodendrocytes SMP+/MBP– or GC-, suggests that the culture medium lacks certain physiological signals that would ensure a complete oligodendrocyte differentiation.

In the rat and the mouse, the determination and the complete maturation of oligodendrocyte precursors have been studied in more detail both in vivo and in vitro. Le Vine and Goldman (1988b), Curtis et al. (1988) and Reynolds and Wilkin (1988) followed in situ, in the cerebrum and in the cerebellum, the successive expression of oligodendrocyte-specific markers from immature neuroectodermal cells generated in the sub-ventricular zone and expressing the GD3 ganglioside. They identified several steps in the process of oligodendrocyte differentiation. Some GD3+ cells of the sub-ventricular zone begin to express carbonic anhydrase or transferrin, which are early oligodendrocyte markers, and they migrate from the subventricular zone to the white matter areas. They eventually lose their reactivity to anti-GD3 and begin simultaneously to express GC and CNPase when they reach their final destination. Other myelin markers are subsequently expressed: MBP 2C–3 days later and proteolipid protein (PLP) and myelin-associated glycoprotein (MAG) just prior to myelin sheath formation (Verity and Campagnoni, 1988). However, in vitro studies on the differentiation capabilities of GD3+ cells (Goldman et al. 1984, 1986; Le Vine and Goldman, 1988a and 1988b) and similar work on A2B5-positive cells (Raff et al. 1983; Raff, 1989; Abney et al. 1983; Behar et al. 1988) demonstrated that these cells are bipotential precursors both of oligodendrocytes and of a subclass of astrocytes.

During development SMP can be detected in situ before MBP and only in regions containing white matter and, in vitro, on the surface of GC+ as well as some GC– cells. Thus, if the results with rodents described above are transposed to avians, one can consider SMP as a differentiation marker that appears when oligodendroblasts have just reached the future white matter tracts and before they begin to express the first myelin components.

SMP is expressed at immunocytochemically detectable levels before myelin formation and the time-lag of expression of SMP and myelin is not constant in all CNS regions. We have also noted that, in contrast to other myelin markers of the PNS, the SMP phenotype appears in neural crest cell cultures and is continuously expressed by Schwann cells in vitro. This allows us to suggest that the initiation of SMP synthesis is dissociable from the molecular mechanisms that control the process of myelination.

The authors wish to acknowledge Dr Julian Smith for critical reading of the manuscript. They also thank Yann Rantier for preparation of the illustrations. This work has been supported by the CNRS and a grant from the Ligue Nationale contre le Cancer and ARSEP. P.C.C. was the recipient of a fellowship from the Universita degli Studi di Catania (Italy).

Abney
,
E. R.
,
Bartlett
,
P. P.
and
Raff
,
M. C.
(
1981
).
Astrocytes, ependymal cells, and oligodendrocytes develop on schedule in dissociated cell cultures of embryonic rat brain
.
Devi Biol
.
83
,
301
310
.
Abney
,
E. R.
,
Williams
,
B. P.
and
Raff
,
M. C.
(
1983
).
Tracing the development of oligodendrocytes from precursor cells using monoclonal antibodies, fluorescence-activated cell sorting, and cell culture
.
Devi Biol
.
100
,
166
171
.
Behar
,
T.
,
McMorris
,
F. A.
,
Novotny
,
E. A.
,
Barker
,
J. L.
and
Dubois-Dalcq
,
M.
(
1988
).
Growth and differentiation properties of O-2A progenitors purified from rat cerebral hemispheres
.
J. Neurosci. Res
.
21
,
168
180
.
Bensted
,
J. P. M.
,
Dobbing
,
J.
,
Morgan
,
R. S.
,
Reid
,
R. T. W.
and
Payling Wright
,
G.
(
1957
).
Neuroglial development and myelination in the spinal cord of the chick embryo
.
J. Embryol. exp. Morph
.
5
,
428
437
.
Cochran
,
F. B.
,
Ledeen
,
R. W.
and
Yu
,
R. K.
(
1983
).
Gangliosides and proteins in developing chicken brain myelin
.
Dev. Brain Res.
6
,
27
32
.
Curtis
,
R.
,
Cohen
,
J.
,
Fok-Seang
,
J.
,
Hanley
,
M. R.
,
Gregson
,
N. A.
,
Reynolds
,
R.
and
Wilkin
,
G. P.
(
1988
).
Development of macroglial cells in rat cerebellum. I. Use of antibodies to follow early in vivo development and migration of oligodendrocytes
.
J. Neurocytol
.
17
,
43
54
.
Dulac
,
C.
,
Cameron-Curry
,
P.
,
Zjller
,
C.
and
Le Douarin
,
N. M.
(
1988
).
A surface protein expresssed by avian myelinating and non myelinating Schwann cells but not by satellite or enteric glial cells
.
Neuron
1
,
211
220
.
Dubin
,
E.
,
Baroffio
,
A.
,
Dulac
,
C.
,
Cameron-Curry
,
P.
and
Le Douarin
,
N. M.
(
1989
).
Schwann cell differentiation in clonal cultures of the neural crest, as evidenced by the anti-SMP monoclonal antibody. Submitted
.
Goldman
,
J. E.
,
Geier
,
S. S.
and
Hirano
,
M.
(
1986
).
Differentiation of astrocytes and oligodendrocytes from germinal maxtrix cells in primary culture
.
J. Neurosci
.
6
,
52
60
.
Goldman
,
J. E.
,
Hirano
,
M.
,
Yu
,
R. K.
and
Seyfried
,
T. N.
(
1984
).
GD3 ganglioside is a glycolipid characteristic of immature neuroectodermal cells
.
J. Neuroimmunol
.
7
,
179
192
.
Hartman
,
B. K.
,
Agrawal
,
H. C.
,
Kalmbach
,
S.
and
Shearer
,
W. T.
(
1979
).
A comparative study of the immunohistochemical localization of basic protein to myelin and oligodendrocytes in rat and chicken brain
.
J. comp. Neurol
.
188
,
273
290
.
Kuhlmann-Krieg
,
S.
,
Sommer
,
I.
and
Schachner
,
M.
(
1988
).
Ultrastructural features of cultured oligodendrocytes expressing stage-specific cell surface antigens
.
Dev. Brain. Res
.
39
,
269
280
.
Le Vine
,
S. M.
and
Goldman
,
J. E.
(
1988a
).
Embryonic divergence of oligodendrocyte and astrocyte lineages in developing rat cerebrum
.
J. Neurosci
.
8
,
3992
4006
.
Le Vine
,
S. M.
and
Goldman
,
J. E.
(
1988b
).
Spatial and temporal patterns of oligodendrocyte differentiation in rat cerebrum and cerebellum
.
J. comp. Neurol
.
277
,
441
455
.
Macklin
,
W. B.
and
Weill
,
C. L.
(
1985
).
Appearance of myelin proteins during development in the chick central nervous system
.
Dev. Neurosci
.
7
,
170
178
.
Mirsky
,
R.
,
Winter
,
J.
,
Abney
,
E. R.
,
Truss
,
R. M.
,
Gavrilovic
,
J.
and
Raff
,
M. C.
(
1980
).
Myelin-specific proteins and glycolipids in Rat Schwann cells and oligodendrocytes in culture
.
J. Cell Biol
.
84
,
483
494
.
Raff
,
M. C.
(
1989
).
Glial cell diversification in the rat optic nerve
.
Science
243
,
1450
1456
.
Raff
,
M. C.
,
Miller
,
R. H.
and
Noble
,
M.
(
1983
).
A glial progenitor cell that develops in vitro into an astrocyte or an oligodendrocyte depending on culture medium
.
Nature, Lond
.
303
,
390
396
.
Reynolds
,
R.
and
Wilkin
,
G. P.
(
1988
).
Development of macroglial cells in rat cerebellum. II. An in situ immunohistochemical study of oligodendroglial lineage from precursor to mature myelinating cell
.
Development
102
,
409
425
.
Verity
,
A. N.
and
Campagnoni
,
A. T.
(
1988
).
Regional expresssion of myelin protein genes in the developing mouse brain: in situ hybridization studies
.
J. Neurosci. Res
.
21
,
238
248
.
Zacchei
,
A. M.
(
1961
).
Lo sviluppo embrionale della quaglia giapponese (Cotumix cotumix japonica T. es
.).
Arch. ital. Anat. Embriol
.
66
,
36
62
.
Ziller
,
C.
,
Dupin
,
E.
,
Brazeau
,
P.
,
Paulin
,
D.
and
Le Douarin
,
N. M.
(
1983
).
Early segregation of an neuronal precursor cell line in the neural crest as revealed by culture in a chemically defined medium
.
Cell
32
,
627
638
.