Murine Spinal Cord Explants: A Model
For Evaluating Axonal Growth and
Myelination In Vitro
Christine E. Thomson,*Anne M. Hunter, Ian R. Griffiths,
Julia M. Edgar, and Mailis C. McCulloch
Applied Neurobiology Group, Institute of Comparative Medicine, Division of Cell Sciences,
University of Glasgow Veterinary School, Glasgow, Scotland, United Kingdom
In vitro models of myelinating central nervous system
axons have mainly been of two types, organotypic or
dissociated. In organotypic cultures, the tissue frag-
ment is thick and usually requires sectioning (physically
or optically) before visual examination. In dissociated
cultures, tissue is dispersed across the culture surface,
making it difficult to measure the extent of myelinated
fiber growth. We aimed to develop a method of cultur-
ing myelinated CNS fibers in defined medium that
could be 1) studied by standard immunofluorescence
microscopy (i.e., monolayer type culture), 2) used to
measure axonal growth, and 3) used to evaluate the
effect of substrate and media components on axonal
growth and myelination. We used 120-lm slices of em-
bryonic murine spinal cord as a focal source of CNS
tissue from which myelinated axons could extend in a
virtual monolayer. Explants were cultured on both poly-
L-lysine and astrocytes. The latter were used because
they are the scaffold on which axonal growth and mye-
lination occurs during normal development. Outgrowth
from the explant and myelination of axons was poor on
poly-L-lysine but was promoted by an astrocyte bed
layer. The best myelin formation occurred in defined
media based on DMEM using N2 mix; it was not pro-
moted by Sato mix or Neurobasal medium with B27
supplement. Neuronal survival was poor in serum-con-
taining medium. This tissue culture model should facili-
tate the study of factors involved in promoting out-
growth of CNS axons and their myelination. As such it
is relevant to studies on myelination and spinal cord
C 2006 Wiley-Liss, Inc.
Key words: myelin; mouse; spinal cord repair; remyelina-
Studying the growth and myelination of central
nervous system (CNS) axons is pertinent to research into
both myelination and spinal cord repair. There are several
published methods for culturing myelinated axons from
CNS tissue, which have various attributes and limitations.
Such culture methods include organotypic (slice) cultures
from rodent brain and spinal cord (Bornstein, 1973; Bunge
and Wood, 1973; Seil, 1979; Billings-Gagliardi et al.,
1980; Stanhope et al., 1986; Munoz-Garcia and Ludwin,
1986; Seil and Herndon, 1991; Ga ¨hwiler et al., 1992,
1997; Notterpek et al., 1993; Shrager and Novakovic,
1995; Phelps et al., 1996; Ghoumari et al., 2003) and ag-
gregate cultures derived from embryonic forebrain (Mat-
thieu et al., 1978). However, most myelinated axons de-
velop within the bulk of the tissue fragment, which makes
it difficult to observe axon–glial interaction by conven-
tional light microscopy and to track individual myelinated
axons and determine their growth parameters. Con-
versely, myelinated fibers develop in spinal ganglia/oligo-
dendrocyte cocultures (Mithen et al., 1983; Wood and
Williams, 1984; Wood and Bunge, 1986) and dissociated
cultures of embryonic CNS (Yavin and Yavin, 1977;
Lubetzki et al., 1993; Svenningsen et al., 2003). These
fibers are readily observable by conventional light micros-
copy. However, spinal ganglia axons are primarily com-
ponents of the peripheral nervous system and do not have
the same molecular and morphological properties as
intrinsic CNS neurons (Chan et al., 2004). In dissociated
cultures of CNS tissue, the cells grow in a diffuse network
across the coverslip, making it difficult to measure their
growth parameters. Finally, virtually all these models use
serum-containing media rather than defined media. Se-
rum contains undefined components, which add an ex-
perimental variable to the protocol.
Our aim was to establish a method in which myelin-
ated CNS axons grew out from a focal tissue source, in a
virtual monolayer, in defined medium. Such a method
would permit 1) manipulation of the microenvironment,
in that all the outgrowing myelinated axons would be
Contract grant sponsor: The Wellcome Trust; Contract grant number:
*Correspondence to: Dr. Christine Thomson, Comparative Physiology
and Anatomy, Institute of Veterinary, Animal, and Biomedical Sciences,
Private Bag 11 222, Palmerston North, New Zealand.
Received 18 May 2006; Revised 3 August 2006; Accepted 5 August
Published online 30 October 2006 in Wiley InterScience (www.
interscience.wiley.com). DOI: 10.1002/jnr.21084
Journal of Neuroscience Research 84:1703–1715 (2006)
' 2006 Wiley-Liss, Inc.
exposed to culture conditions; 2) observation by conven-
tional light microscopy; 3) measurement of myelinated
axon growth; and 4) experimental reproducibility, in that
the culture medium is serum free during myelination. Ini-
tially, we cultured thin slices of embryonic murine spinal
cord on poly-L-lysine; however, axonal growth out from
the explant was limited, and myelination was poor exter-
nal to the explant. We compared this with normal embry-
onic development in vivo, in which axonal growth and
myelination occur in a neuropil made up largely of astro-
cytes. Thus we hypothesised that astrocytes might pro-
mote both events. We have now tested this hypothesis by
culturing the explants on a bed layer of astrocytes derived
from neurospheres. Under such conditions, myelinated
axons grow out from a focal tissue source in a virtual
monolayer, for measurable distances, in defined medium.
Thus, this model is suitable for investigating the require-
ments for both growth and myelination of CNS axons.
Consequently, it is pertinent to the study of myelination
and spinal cord repair.
MATERIALS AND METHODS
The mice used were of the hybrid C3H/101 strain,
with no known mutations. Timed mating was performed,
with the day of plugging denoted as embryonic day 0.5
(E0.5). Postnatal animals were used for generating neuro-
spheres, postnatal day 1 (P1) being the day on which the birth
of the litter was first noted.
Cell Culture Media
The concentrations of nutrients given here are the final
concentrations in that solution. Basal media (HBSS, DMEM,
and DMEM-Advanced), amphotericin B, NaHCO3, gluta-
mine, fetal calf serum, horse serum, epidermal growth factor
(EGF), Neurobasal medium, and B27 supplement were
obtained from Gibco, Invitrogen (Paisley, Scotland). Pig skin
gelatin, Triton X-100, gentamicin, penicillin-streptomycin,
glucose, HEPES, bovine serum albumin, apotransferrin, insu-
lin, putrescine, sodium selenite, progesterone, thyroxine, tri-
iodothyronine, biotin, and hydrocortisone were obtained from
Sigma (Poole, England). Unless otherwise specified, all DMEM
was 1,000 mg/liter glucose containing 0.8 lg/ml amphoteri-
cin B and 50 lg/ml gentamicin.
Neurosphere culture medium (NM). This balanced
salt medium was supplemented with glucose and hormones
(Zhang et al., 1998) and consisted of a 1:1 mixture of DMEM
(with 4,500 mg/liter glucose)/F12 nutrient media, 0.105%
NaHCO3, 2 mM glutamine, 5,000 IU/ml penicillin, 5 lg/ml
streptomycin, 0.6% glucose, 5.0 mM HEPES, 0.001% bovine
serum albumin, 100 lg/ml apotransferrin, 25 lg/ml insulin,
60 lM putrescine, 30 nM sodium selenite, 20 nM progester-
one, and 20 ng/ml EGF.
Astrocyte culture medium (AM). This medium
consisted of DMEM, 2 mM glutamine and 10% heat-inacti-
vated fetal calf serum.
Plating medium (PM). To encourage attachment of
explants, a serum-rich medium was used. This consisted of
50% DMEM, 25% horse serum, 25% HBSS (with calcium
and magnesium), and 2 mM glutamine. The horse serum was
used to encourage neuronal growth (Banker and Goslin,
1992); it has also been found to encourage survival of mouse
oligodendroglia (Shaw et al., 1996).
Differentiation medium (DfM). This medium was
used to encourage oligdodendroglial differentiation and con-
sisted of DMEM (4,500 mg/liter glucose), 0.5% hormone mix
(see below), 10 ng/ll biotin, and 50 nM hydrocortisone.
DMEM-Advanced (Gibco) was also used in lieu of standard
Hormone mix. The hormone mix was apotransferrin
at 1 mg/ml, 20 mM putrescine, 4 lM progesterone, 6 lM se-
lenium. This formulation is based on the N2 mix (Bottenstein
and Sato, 1979).
Sato-H medium. This
DMEM (4,500 mg/liter glucose) to which was added Sato
mix, giving a final concentration of 2 mM glutamine, 100
lg/ml apotransferrin, 0.0286% bovine albumin-Path-O-Cyte,
0.2 lM progesterone, 100 lM putrescine, 0.45 lM thyroxine,
00.224 lM sodium selenite, and 0.5 lM trioiodothyronine.
The medium was supplemented with 50 nM hydrocortisone.
In all explant cultures, 10 lg/ml insulin was added to the cul-
ture medium for the first 12 days in vitro and then removed
from the medium.
Poly-L-lysine (PLL; MW 70,000–150,000; 13.3 lg/ml;
Sigma) was used to coat the coverslips. Sterile 13-mm round
glass coverslips were immersed in PLL for 30–60 min in a
laminar flow hood, after which the PLL was aspirated and
replaced with sterile ddH2O. The coverslips were then trans-
ferred to 35-mm sterile plastic Petri dishes (three per dish) or
24-well tissue culture plates and left to air dry in the laminar
All cultures were maintained in a humidified incubator
at 378C and 5% CO2.
Neurospheres and astrocyte monolayers. Neuro-
spheres were generated from neonatal mouse striatum by using
methods modified from Reynolds and Weiss (1996) and
Zhang et al. (1998). Approximately 2.5 weeks before the
explant cultures were to be set up, the striatum was dissected
from neonatal mice less than 36 hr old into neurosphere me-
dium (NM; 1 pair striata/ml medium). The tissue was dissoci-
ated by trituration with a Pasteur pipette and added to 4 ml
of NM, in a T25 flask. EGF was added to a final concentra-
tion of 20 ng/ml. At 2 days in vitro, 2.5 ml NM + EGF was
added to the T25 flask; at 4 days of culture, the suspension
was centrifuged, two-thirds of the medium was discarded, and
the cells were resuspended by triturating gently through a Pas-
teur pipette in the remaining medium and inoculated into a
T85 flask containing 10 ml NM + EGF. The cells were main-
tained by passaging and triturating with a Pasteur pipette in
NM + EGF (changing two-thirds of the medium three times
per week) for a total of 8–12 days. At this time, the spheres
from each T85 flask were harvested by centrifugation, resus-
pended in 1.5 ml AM, and triturated using a Pasteur pipette
to produce small spheres/cell clusters. The suspension volume
1704Thomson et al.
Journal of Neuroscience Research DOI 10.1002/jnr
was increased to 2.5ml and the cells were seeded as 100 ll ali-
quots into 500 ll AM on PLL-coated coverslips in 24-well
plates. This resulted in approximately 20–30 small spheres
being added to each well. The cultures were fed twice weekly
with AM, and in that medium they differentiated into astro-
cytes within a few days. Confluent astrocyte monolayers de-
veloped within 7–10 days of plating.
pregnant mice at day 13.5–14 of gestation were killed by an-
esthetic overdose (halothane) and CO2inhalation. The uterus
was removed, and, under a dissecting microscope, the fetuses
were extruded and decapitated just caudal to the cerebellum,
leaving the cervical flexure attached to the cord. The fetus
was placed on its ventral surface in a dry sterile Petri dish.
The intact spinal cord, extending from the caudal medulla
oblongata to the midlumbar region, was removed by blunt
dissection with fine forceps (No. 5; Taab, Aldermaston, Eng-
land). The cords were immediately transferred to a fresh Petri
dish containing DMEM buffered with 5 mM HEPES. The
cervical flexure was gently pressed against the floor of the Pe-
tri dish, so the dorsal aspect of the cord was uppermost. By
holding the cervical flexure with forceps, the meninges over
the dorsal midline could be split longitudinally by using a light
stroking motion with fine forceps. Although this procedure
tended to split the roof plate, resulting a U-shaped cord, it did
not appear to affect the culture viability. After the meninges
were split, the cervical flexure was still held gently with for-
ceps, and the meninges were stripped off in a continuous
sheet, leaving an intact spinal cord. A 3.0% solution of agar
(type VII, low gelling temperature; Sigma) was made up in
DMEM in a sterile bijoux with screw-top lid, melted in a
658C water bath, and then transferred to 378C water bath 5–
10 min before required. (In our laboratory, it took approxi-
mately 8 min for the agar to cool to 378C in the water bath.)
About 2 ml of the liquid agar was poured into a sterile alu-
minium foil mould and checked for temperature by touching
the outside of the boat to exposed skin. Up to eight cords
were embedded parallel to each other in the liquid agar, and
the mould was set on ice to gel. The foil mould was removed
and the agar block trimmed with a sterile scalpel blade and
mounted on the tray of a Vibratome 1000 Plus (General Sci-
entific, Redhill, United Kingdom) using Super Glue (cyanoa-
crylate) so that the cords were perpendicular to the plane of
sectioning and yielded transverse sections. The bath surround-
ing the tray was packed with ice. With the block kept moist
at all times with DMEM, the cords were sliced in transverse
section at 110–120 lm thickness and the agar-embedded slices
transferred into a Petri dish containing DMEM/5 mM
HEPES/2% fetal calf serum. The dish was gently rocked by
hand to encourage the slices to detach from the agar.
Plating the explants. Three coverslips (PLL coated or
astrocyte monolayers) were placed in each 35-mm Petri dish,
and 1 ml of DMEM was added to each dish; this medium was
removed just prior to adding the explants. The spinal cord sli-
ces (without the agar) were transferred onto the coverslips
(PLL-coated or with the astrocytic monolayer) by using a pip-
ette fitted with a 200-ll tip, from which the distal 5 mm had
been cut to enlarge the aperture. The excess medium was
removed, and 200 ll of PM was gently added to the edges of
each Petri dish, taking care not to dislodge the slices. Five
minutes later, a second 200-ll aliquot was added. A third
200-ll aliquot was carefully added 20 min after plating of the
explant. The slices were left to attach for 2–3 hr in the incu-
bator, after which 600 ll of DfM with insulin was carefully
added. The slices were viewed under the dissecting micro-
scope to monitor for detachment while medium was being
added at each stage. In some experiments, alternative media
were used for explant culture (Table I).
Thereafter, the cultures were fed three times per week
with DfM by changing half the medium, resulting in a total
volume/Petri dish of approximately 1 ml. The DfM was sup-
plemented with insulin for the first 12 days of culture.
Primary oligodendrocyte culture. Primary oligoden-
drocytes were prepared from P6 spinal cord by enzymatic and
mechanical dissociation as described previously (Thomson
et al., 1999).
The following antibodies (with their dilutions) were
used: A2B5 (1:10; Schnitzer and Schachner, 1982), O4, O1
TABLE I. Effect of Media Composition on Growth of Myelinated Fibers
Differentiation medium with insulin for first 12 days of culture Excellent outgrowth of axons; some outgrowth of oligodendroglia that
myelinate axons; few oligodendroglia growing isolated from axonal contact
Good axonal outgrowth, but increased numbers of oligodendroglia growing
isolated from axonal contact; These cells tended to obscure the myelinated
Moderate axonal outgrowth and myelination, with few isolated
oligodendroglia (Fig. 6)
Early myelination within the body of the explant (beginning at 8–10 days in
vitro), but reduced oligodendrocyte outgrowth and no myelination
external to the explant; axons began to degenerate from 14 days in vitro
Good axonal outgrowth, but the myelin sheaths were thinner and fewer;
there were more oligodendroglia growing isolated from axonal contact
even when insulin was withdrawn at 12 days in vitro (Figs. 5A,B, 6)
Good axonal outgrowth, and development of many immature
oligodendroglia external to the explant but little myelin formation
Differentiation medium with insulin for full duration of culture
Differentiation medium made up in advanced DMEM
Serum rich (> 5.0%) for the duration of the culture
Sato medium with hydrocortisone.
Neurobasal medium with B27 and N2 hormone mix supplements
In Vitro Axonal Growth and Myelination 1705
Journal of Neuroscience Research DOI 10.1002/jnr
(1:5; Sommer and Schachner, 1981), glial fibrillary acidic pro-
tein (GFAP; 1:750; DakoPatts, Glostrup, Denmark), myelin basic
protein (MBP; 1:500; Prof, N.P. Groome, Oxford, United King-
dom), AA3 against PLP/DM20 (1:10; Yamamura et al., 1991;
from Dr. S. Pfeiffer, University of Connecticut), monoclonal anti-
neurofilament 68 (1:800; Sigma), and SMI 31 against phosphoryl-
ated neurofilament (1:1,500; Affiniti, Exeter, United Kingdom).
Secondary antibodies were obtained from Southern Biotech (Cam-
bridge Bioscience, Cambridge, United Kingdom) and included
goat anti-mouse IgM Texas-red (1:50), goat anti-rat IgG fluores-
cein isothiocyanate (FITC; 1:50), goat anti-mouse IgG1 Texas-red
(1:80). Goat anti-mouse IgG Cascade blue (1:80) was obtained
from Molecular Probes (Leiden, The Netherlands). All procedures
were performed at room temperature unless otherwise stated.
Cell surface antigens. Cultures were rinsed briefly in
DMEM; primary antibody was applied for 30 min; cultures
were washed briefly in DMEM; secondary antibody was
added for 30 min; and cultures were washed in DMEM and
fixed in4% paraformaldehyde/phosphate-buffered
(PBS), pH 7.5, for 15 min. Primary and secondary antibodies
were diluted in DMEM.
Cytoskeletal and membranous antigens. The fol-
lowing method was used either in isolation from or subse-
quent to labelling of cell surface antigens. Cultures were fixed
in 4% paraformaldehyde/PBS for 15 min; washed three times
in PBS, 10 min/wash; permeabilized with 0.5% Triton X in
PBS for 30 min; blocked using a blocking buffer (0.2% pig
skin gelatin, 0.1% Triton X in PBS) for 30 min; and primary
antibody, diluted in blocking buffer, was added and left over-
night at 48C. On the following day, the coverslips were
washed three times in PBS 20 min/wash, and the secondary
antibody diluted in blocking buffer was added for 1–2 hr.
The coverslips were washed in PBS for 5 min and briefly in
distilled water, mounted in AF1 glycerol/PBS (Citifluor, Lon-
don, United Kingdom), and sealed with clear nail varnish.
Plastic sections and ultrastructure. For electron mi-
croscopy (EM), cultures were grown for 4–5 weeks and then
fixed in 2% paraformaldehyde/5% glutaraldehyde in 0.08 M
sodium cacodylate buffer (pH 7.2). A 3% solution of agar
(type VII, low gelling temperature; Sigma) was made up in
PBS, and a drop was placed in a plastic petri dish and allowed
to gel. Coverslips with fixed myelinating cultures were placed
in PBS in a Petri dish, and, under a dissecting microscope, the
cells of a single culture were gently flushed and scraped off to
produce a single sheet of tissue that was swirled into a clump
and placed on the agar drop. A second drop of agar was
placed on top of the first drop, sandwiching the clumped cul-
ture material. The agar sandwich was trimmed to a 2 3 3
mm block, processed routinely for EM, and embedded into
Araldite blocks for sectioning. The clumping of the culture
material into the agar block made it easier to find the myelin-
ated fibers under EM.
Light Microscopy and Photography
Evaluation of the sections was performed on an Olym-
pus IX70 research fluorescence microscope with standard epi-
fluorescence optics. Confocal images were captured on a Leica
DMIRE2 microscope. Thin sections (70 nm) were evaluated
on a JEOL JEM-100CX II electron microscope. Images for
analysis were digitally captured.
Quantification of myelinated fiber outgrowth from
explants. Digital images of immunostained myelinated fibers
that had grown out from the explant (two images/explant)
were captured. Lines were superimposed on the images at 100
and 200 lm from the edge of the explant. The numbers of
myelinating axons crossing these two lines were counted and
summed. On average 80 images were captured for each cul-
ture condition in each experiment, and the experiments were
repeated three times. Data were analyzed by ANOVA and
Dunnett’s multiple-comparisons test, comparing different cul-
ture conditions with the standard of cultures grown in DfM
G ratios. Thin sections and a diffraction grating for
calibration were photographed at 38,000. Image Pro Plus 4.0
(Media Cybernetics, Silver Springs, MD) software was used to
calculate the G ratio (axonal diameter to total myelinated fiber
diameter) from 75 myelinated fibers.
Axonal Growth and Myelination of
Axons Are Minimal on PLL
By 48 hr in vitro, approximately half the explants
plated on PLL-coated coverslips detached. Around those
explants that remained attached, the outgrowth formed a
short fringe extending up to 300 lm (Fig. 1A). After 3
Fig. 1. Explants cultured on poly-L-lysine. Outgrowth of axons was
limited, and very little myelination occurred, but moderate numbers
of myelinated axons did form within the explants. A: Phase-contrast
image depicting that outgrowth from an explant onto poly-L-lysine
after 48 hr is restricted to a few hundred micrometers. B: In some
areas where cellular outgrowth was reasonable, axons and oligoden-
droglia grew superimposed on top of each other, yet the two cell
types failed to interact, and myelin sheaths did not form. Arrows
identify axons, open arrowheads indicate the expanded plasma mem-
brane of a mature oligodendrocyte stained with O1 antibody, and a
solid arrowhead indicates the perinuclear region, which is brightly
stained for PLP/DM20. C,D: Same images identified by immuno-
staining and phase optics. Axons growing within the body of an
explant (Exp) were myelinated (open arrowheads); however, after
leaving the explant, they were unmyelinated despite being sur-
rounded by a cluster of glial cells (top left, D). Solid arrowhead indi-
cates the same axon in each image. Myelin was stained for PLP/
DM20. E–H: Explants cultured on poly-L-lysine for 28 days cos-
tained for myelin (E,G) and axons (F,H). Few axons grew out onto
the coverslip, but moderate numbers of myelinated axons (arrow-
heads) developed within the body of the explant, in which astrocytes
were also present (data not shown). Solid arrowheads indicate the
edge of the explants. E,G: Myelin sheaths stained for PLP-DM20 (p/
d); F,H: axons stained for phosphorylated neurofilaments (nf). Scale
bars ¼ 50 lm in A,B; 50 lm in C (applies to C,D); 50 lm in E
(applies to E,F); 50 lm in G (applies to G,H).
1706Thomson et al.
Journal of Neuroscience Research DOI 10.1002/jnr
weeks in vitro, outgrowth of axons was irregular and of
low density, with large areas surrounding the explant
devoid of axons. Outgrowth of astrocytes and oligoden-
droglia was also poor. Occasional myelin sheaths devel-
oped, but oligodendroglia were also found superimposed
on axons but not interacting with them (Fig. 1B–D).
Moderate numbers of myelin sheaths developed within
the explants (Fig. 1E–H).
Astrocytes Support Axonal and Oligodendroglial
Outgrowth and Subsequent Myelination
When grown on an astrocytic bed layer, axons
extended up to 1,000 lm from the explant (Fig. 2A)
within 48 hr of plating. Phase-bright glia grew out fol-
lowing the leading front of the axonal growth cones.
Within the first 2 weeks, the pattern of differentiating oli-
godendrocytes emerging from the explant recapitulated in
vivo oligodendroglial differentiation. At the leading edge
of the glial cell outgrowth were the bi-or multipolar
A2B5+cells, followed by O4+multipolar cells, wherreas
O1+/PLP-DM20+cells were found closest to the explant
(data not shown). This sequential pattern of oligodendro-
glial differentiation is consistent with that noted in the lit-
erature (Schachner et al., 1981). At 2.5 weeks in vitro,
myelin sheaths started to develop on outgrowing axons.
By 3 weeks in culture, axonal density was greatest within
800 lm of the explant; however, small bands of axons
extended up 2,500 lm from the explant, whereas PLP-
DM20+oligodendroglia migrated up to 1,200 lm. Nerve
cell bodies were found mainly within the explant. Good
outgrowth of myelinated fibers developed from the major-
ity of explants (>70%) in each experiment. Maximal de-
velopment occurred at 23–28 days in vitro; after 5 weeks,
the cultures started to deteriorate. Myelin stained intensely
with antibodies against PLP-DM20 (Fig. 2B, 3A,B) and
MBP (Fig. 2D) and was seen surrounding axons stained for
phosphorylated neurofilament (Fig. 2C,E, 3A,B). The
bulk of the myelin was formed on axons located within a
600-lm radius from the explants, but individual myelinat-
ing oligodendroglia were also observed myelinating some
of the longer axons up to 1,200 lm from the explant (Fig.
2F). Under higher magnification, the oligodendroglial cell
body, its cell processes, and associated myelin sheaths could
be seen (Fig. 2G). Well-formed, compacted myelin lamel-
lae were identified by electron microscopy (Fig. 4A–C).
The G ratio of myelinated fibers in the explants was
0.81 6 0.07 (mean 6 SEM) at 33 days in culture, which
is approximately 2.5 weeks after myelination had ensued.
This was compared with data collected by other research-
ers in our laboratory. The G ratio of large-diameter fibers
of the ventral funiculus of P20 wild-type (C57) mice was
0.711 6 0.01 (Prof. I.R. Griffiths, unpublished data),
which is approximately 20 days after myelination begins
in the ventral funiculus. The G ratio of small-diameter
fibers in the optic nerve from wild-type mice at P20 was
0.78 6 0.06, which is approximately 13 days after myeli-
nation begins in the optic nerve (Dr. J. Edgar, personal
Dense Networks of Myelinated Fibers, With Nodes
of Ranvier, Develop Between Adjacent Explants
Four to six slices were cultured on each coverslip.
Occasionally, between some explants, dense, bridging
networks of myelinated axons developed (Fig. 4D). These
interconnected explants were mostly located within 800
lm of each other. Along these axons, contiguous myelin
sheaths developed. These sheaths were separated by small
gaps in the myelin, which had the appearance of nodes of
Ranvier (Fig. 4E).
Outgrowth Is Affected by Astrocyte Density, Slice
Thickness, and Spinal Cord Orientation at Cutting
Myelinated axons grow out from the majority of
explants; however, certain technical factors were associ-
ated with maximal outgrowth. These factors included
ensuring that the astrocyte bed layer was confluent, rather
than patchy, at the time of plating the explants. The opti-
mal thickness for spinal cord slices was 110–125 lm; thin-
ner than this and they were too fragile to handle, whereas
increased cell death occurred in thicker slices presumably
because of limitations in nutrient diffusion. Large numbers
of myelinated fibers formed within the explants. In slices
that were 110 lm thick, these myelinated fibers were
individually visible, whereas there was obvious superim-
position of fibers in 120-lm-thick slices. Optimal out-
growth was obtained when spinal cords sections were
transverse rather than tangential or longitudinal. Explants
were obtained from cervical, thoracic, and cranial lumbar
Fig. 2. Explants cultured on a bed layer of astrocytes. There was sub-
stantial outgrowth and myelination of axons. A: Phase-contrast image
depicting outgrowth from an explant after 48 hr. Compared with
that on poly-L-Lysine, axonal outgrowth was extensive (arrowheads),
with axons reaching up to 1,000 lm. Arrows indicate phase-bright
glia emerging from the explant. B–E: Costained images of myelin-
ated axons (arrowheads) that have grown out from the explant body
(Exp). B,D: Cultures stained for myelin proteins PLP/DM20 (p/d) at
24 days in vitro and MBP after 21 days in vitro, respectively. C,E:
Same cultures as in B,D stained for phosphorylated neurofilament
(nf). F: The bulk of the myelinated fibers were located within a 600-
lm radius of the explant; however, beyond that zone, individual
myelinating units (axons myelinated by a single oligodendrocyte)
were observed. Myelin sheaths stained with PLP/DM20 (p/d) after
23 days in vitro; arrows indicate oligodendroglial cell bodies.
G: Image depicting a mature myelinating unit stained for PLP-
DM20. Arrow indicates the oligodendroglial cell body with the neg-
atively stained nuclear area, and arrowhead indicates the lightly
stained oligodendroglial process connecting to the intensely stained
myelin sheath. Scale bars ¼ 100 lm in A; 50 lm in B (applies to B,C);
50 lm in D (applies to D,E); 50 lm in F; 20 lm in G.
1708Thomson et al.
Journal of Neuroscience Research DOI 10.1002/jnr
In Vitro Axonal Growth and Myelination1709
cord. The effect of anatomical region on outgrowth was
not specifically evaluated.
Embryonic Age Is Important for
Optimizing Cell Cultures
The optimal embryonic age for culture in the mouse
was E13.5–14, with the day of plugging being day 0.5 of
gestation. There was less outgrowth from E15.5 tissue,
and removing the spinal cord was more difficult, insofar
as the overlying tissues were better developed and did not
dissect as easily as at E13.5–14.5.
Neurospheres Give Rise to Astrocyte-Rich
The astrocyte bed layer cultures were immuno-
stained to ascertain whether the neurospheres could have
contributed other cell types to the myelinating cultures.
Neurosphere-derived astrocyte cultures, without explants,
were stained at 7, 14, and 21 days in vitro. Small numbers
of MBP+oligodendrocytes differentiated from the neuro-
spheres in the first 2 weeks, but these numbers declined
progressively during the culture period. At 3 weeks, cul-
tures were stained for GFAP, neurofilament (NF68), O4.
and O1. Almost 100% of cells were GFAP positive, and
only an occasional O4+cell was noted; cultures were neg-
ative for more differentiated oligodendrocytes and for
neurons. Although it is possible that cellular differentia-
tion in the astrocyte bed layer is altered by the presence of
the explant, our data suggest that the bed layer is unlikely
to contribute oligodendroglia or neurons to the myelin-
ated fibers in the explant-astrocyte cocultures.
Neurospheres were used as a source of astrocytes,
because they reliably differentiated into a population of
astrocytes that could be derived from a source with \cell-
line"-like qualities; i.e., cultures that could be passaged
and expanded for up to 2 weeks. However, after approxi-
mately five to seven passages, the neurospheres would
tend to attach to the flask and start differentiating. Use of
neurospheres limited the number of animals required, in
that one neonatal mouse yielded sufficient cells to plate
24 coverslips, each of which could support four to eight
explants. Neurospheres also resulted in a highly enriched
population of astrocytes.
Axon–Glial Interaction Is Affected by
Composition of Growth Medium
Although the bed layer of astrocytes was critical for
supporting outgrowth of axons, media composition deter-
mined subsequent myelination of those axons (Figs. 5, 6).
Serum-rich media were used at the initial plating of the
explants to encourage explant attachment and astrocytic
bed layer survival. However, we found that prolonged
use of high serum concentrations (>5%) resulted in neu-
ronal and oligodendroglial loss (data not shown).
Several different media compositions were tried
(Table I), but DfM made up in standard DMEM yielded
the best cultures with the thickest myelin and the fewest
nonmyelinating oligodendroglia (Fig. 6). The removal of
Fig. 3. Confocal images (1 lm slice) of spinal cord explants. The
myelin is brightly stained with anti-PLP/DM20, and the axons are
more lightly stained with antibody against phosphorylated neurofila-
ment. The axons extend to the edge of the image. The edge of the
explant is delineated with a dashed line. Scale bars ¼ 100 lm.
1710 Thomson et al.
Journal of Neuroscience Research DOI 10.1002/jnr
insulin after 12–13 days in vitro was critical for minimiz-
ing the numbers of nonmyelinating oligodendroglia.
Other defined media based on Sato medium (Figs. 5A,B,
6) or Neurobasal medium with B27 and the N2 hormone
mix supplement (Fig. 5C,D) supported good axonal and
oligodendroglial outgrowth, but very few myelin sheaths
were formed. The number of myelinated fibers was also
reduced by approximately 50% when explants were cul-
tured in DfM made up in advanced-DMEM (Fig. 6).
Hydrocortisone has been shown to support oligo-
dendroglial growth and differentiation (Warringa et al.,
1987). In a preliminary study, the effect of hydrocortisone
Fig. 4. Mature myelinated axons developed in and around explants
cultured on neurosphere-derived astrocytes. A–C: Electron micro-
graphs obtained of myelinated fibers in transverse (A,C) and longitu-
dinal (B) section, cultured for 33 days on a bed layer of astrocytes.
Myelin sheaths (solid arrowheads) were well compacted, with regular
periodicity. Arrow in C indicates inner tongue of oligodendrocyte
cytoplasm. g, Edge of EM grid; a, axons. D,E: Dense mats of myelin
grew between adjacent explants (arrows indicate direction of
explants, which were located just out of the photograph in D), and
contiguous myelin sheaths developed along the axons. Arrowheads in
E point to probable nodes of Ranvier (arrowheads in E). Myelin was
stained for PLP/DM20 (p/d). Scale bars ¼ 0.5 lm in A; 1 lm in B;
100 nm in C; 50 lm in D; 20 lm in E.
In Vitro Axonal Growth and Myelination1711
Journal of Neuroscience Research DOI 10.1002/jnr
on dissociated oligodendroglial cultures generated from
early postnatal mouse spinal cord was evaluated. Addition
of hydrocortisone increased the proportion of MBP+cells
(70–80% c/w 60%) at 10 days in vitro (data not shown).
In the explant cultures, addition of 50 nM hydrocortisone
appeared to promote myelin formation, generating more,
and thicker, sheaths.
In these studies, we sought to develop a method for
culturing thin layers of myelinated CNS fibers from a
focal source of tissue, in defined medium. Such a model
could then be used to determine the factors required to
support both outgrowth and myelination of axons. Data
from such studies are pertinent to research into myelina-
tion and spinal cord repair. Spinal cord injury and demye-
linating diseases, such as multiple sclerosis, are two major
CNS disorders. In both of these diseases, therapeutic
options are still limited. Having an in vitro model that
permits easy visualization and measurement of growing,
myelinated axons should help mechanistic and therapeutic
studies in these areas.
Fig. 5. Myelination of axons growing out from the explant on an
astrocyte bed layer was determined by the composition of the media.
Despite abundant oligodendroglia and vigorous axonal outgrowth af-
ter 3 weeks in vitro, there was minimal ensheathment and myelin
formation in a medium based on Sato mix (A,B) or Neurobasal me-
dium with B27 supplement and N2 hormone mix (C,D). A,C: Mye-
lin was stained for PLP/DM20 (p/d). B,D: Axons were stained for
phosphorylated neurofilament (nf). Exp, explant; arrows indicate lim-
ited association between the axons and oligodendroglia. Scale bars ¼
Fig. 6. Graph depicting the effect of different culture conditions on
outgrowth of myelinated fibers from the spinal cord explants.
Explants were grown on a bed layer of astrocytes derived from neu-
rospheres (NSA; columns 1–3) or on poly-L-lysine (PL)-coated cov-
erslips (column 4). The different media used included differentiation
medium (DfM), DfM made up in advanced-DMEM (DfM-Ad), and
Sato-hydrocortisone (Sato-H) medium. The number of myelinated
fibers crossing grid lines superimposed on digital images of the out-
growth was counted. Compared with the standard (DfM on NSA),
there were significantly fewer myelinated fibers growing out from
the explants in DfM-Ad and Sato-H medium (P < 0.05) and even
fewer fibers in cultures grown on poly-L-lysine in DfM (P < 0.01).
1712 Thomson et al.
Journal of Neuroscience Research DOI 10.1002/jnr
In previous studies (as cited in the introductory
paragraphs), myelinated fibers have been generated in
organotypic/thick slice cultures, reaggregation cultures,
and dissociated forebrain cultures. With the first two cul-
ture methods, the tissue may be relatively thick, and, for
morphological evaluation, electron microscopy has been
required. Dissociated forebrain cultures are amenable to
examination by immunofluorescence and light micros-
copy, in that cultures are only a few cells thick. However,
measuring the growth of axons in dissociated cultures is
difficult, because they form a diffuse network across the
culture surface. In this paper, we describe a model in
which thin slices of spinal cord were cultured on astro-
cytes derived from neurospheres. Neurosphere-derived
astrocytes were selected as the substrate based on the
results of initial experiments in which a variety of other
substrates were given trials. Outgrowth of myelinated
axons was minimal on substrates such as laminin-2,
extracellular matrix, and astroglial-derived matrix. In pilot
studies, we also tried astrocytes cultured from postnatal
day 1 mouse cortical astrocytes and obtained modest out-
growth. On neurosphere-derived astrocytes, a virtual
monolayer of myelinated axons grew out from the
explant, and this monolayer was amenable to both light
microscopy and measuring axonal length. Growing such
fibers in thin layers extending from a tissue fragment per-
mits study into the effect of bound substrate, or soluble
factors, on axonal growth and myelination. Although my-
elinated axons do develop within the explant, it is not
easy to manipulate the microenvironment within the
explant and test the effect of various factors either soluble
or substrate bound. It is also hard to measure axonal
length. Finally, serum is not required for the bulk of the
culture period, removing an unknown variable from the
method. Based on the results of this study, astrocytes
derived from neurospheres seem to promote the growth
of axons in developing murine CNS cultures, whereas
specific media is required to promote myelination of these
Growth and Myelination of Axons Is Better on
Astrocytes Than on PLL
We began these studies by culturing thin slices of
embryonic spinal cord on PLL-coated coverslips, but the
outgrowth of axons and oligodendroglia was limited. My-
elinated fibers developed within the explants, but they
were growing within the microenvironment of a tissue
fragment, with its attendant limitations, as discussed previ-
In vivo, astrocytes account for up to 50% of the vol-
ume of the CNS (Hof et al., 1999) and are important in
the embryological growth of axons. They are a known
source of oligodendroglial survival and differentiation fac-
tors (Raff et al., 1988; Barres et al., 1992; Barres and Raff,
1994) and may also encourage myelination (Meyer-
Franke and Barres, 1996; Meyer-Franke et al., 1999). We
hypothesized that astrocytes could promote axonal out-
growth and myelination by oligodendrocytes in vitro. We
found that astrocytes derived from neurospheres can sup-
port the outgrowth of myelinated axons from embryonic
spinal cord slices.
In this model, axons can extend for up to 2,500 lm
from the explant, but the bulk of the myelination occurs
within 600–1,200 lm. The reason for this limitation is
unknown, but may involve either the outgrowth of other
cells (e.g., explant derived astrocytes) or the diffusion of
soluble substances from the explant into this zone.
The myelin produced in these cultures was well
formed at the ultrastructural level and achieved a reasona-
ble thickness as evidenced by the G ratio. Given more
time, it might have continued to thicken, but cultures
begin to degenerate at about 5 weeks. The reason for the
degeneration is not known, but presumably it is due to
some imbalance of factors or deficiency in the culture
When grown on neurosphere-derived astrocytes,
myelinated fibers were most dense between adjacent ex-
plants, where fibers could form dense bridging networks.
This development of dense mats of myelinated fibers may
be related to synapse formation and the development of
electrical activity, which has been shown to promote
myelination (Gyllensten and Malmfors, 1963; Demerens
et al., 1996). In our explant cultures, the role of electrical
activity could be evaluated by using tetrodotoxin as
described by Demerens et al. (1996).
Media Composition Affects Myelin Formation
The aim of the method was to generate cultures
with myelinated axons but minimal numbers of oligoden-
droglia growing isolated from the axons. Additionally, we
wished to develop a serum-free (defined) culture me-
dium, because serum composition can vary, and we found
that prolonged use of serum decreased axonal and oligo-
dendroglial survival. Serum has been found in other stud-
ies markedly to decrease lipid synthesis and oligodendro-
glial differentiation in primary glial cultures (Sykes and
Lopes-Cardozo, 1988). However, wanting to use defined
medium in our myelinating cultures generated a conflict
in media requirement, insofar as astrocytes traditionally
prefer serum-containing media. Thus, cultures were ini-
tially established in serum-rich medium, and a feeding
strategy was devised such that the concentration of serum
declined with each feeding so that, by 10–12 days in
vitro, it was less than 1%. Outgrowth of myelinated axons
from the explants was best supported with differentiation
medium made up in standard DMEM.
Because of our concern that the lack of serum in
DfM might negatively influence survival of the astrocytic
bed layer, we also tried making up DfM in Advanced-
DMEM. This is a defined, serum-free medium that has
been formulated for cells that prefer serum (www.
invitrogen.com). However, we found that this medium
decreased the number of myelinated fibers by more
Insulin was found to promote oligodendroglial sur-
vival in the first 12 days of the culture period but after
In Vitro Axonal Growth and Myelination1713
Journal of Neuroscience Research DOI 10.1002/jnr
that time had a negative effect, in that it promoted the
survival of excess isolated oligodendroglia (i.e., not en-
sheathing/myelinating axons). These additional cells
stained positively with myelin antibodies, and their pres-
ence obscured myelinating oligodendroglia. Insulin is a
survival factor for oligodendroglia (Barres et al., 1992),
and, with its absence in our cultures, the surviving oligo-
dendroglia were mainly those ensheathing/myelinating
the axons. This mimics the in vivo situation in the devel-
oping optic nerve (Barres et al., 1993). We tried a number
of other medium formulations, including Sato mix (with
added hydrocortisone), which we have traditionally used
for culturing oligodendrocytes (Fanarraga et al., 1993;
Thomson et al., 1999), and Neurobasal medium with
N27 supplement, which is used to promote neuronal sur-
vival (Brewer et al., 1993; www.invitrogen.com). Both
media promoted axonal and oligodendroglial outgrowth
but did not promote myelination. The addition of hydro-
cortisone appeared to be associated with increased myelin
sheath formation. The positive effect on myelination may
be due to the general trophic effect that hydrocortisone
has on cultured oligodendroglia (Warringa et al., 1987).
The Technique Minimizes Animal Use
This technique yields a large number of explants
from one pregnant mouse. The average litter size for the
C3H/101 mouse strain in our facility is seven (average of
100 litters). Approximately 35–45 slices cut at 120 lm
can be obtained from one spinal cord. Multiple cords can
be mounted in the agar block for slicing (usually four to
eight cords). After plating, four to six explants can be
placed on each 13-mm coverslip. Thus this method is ef-
ficient, generating sufficient numbers of samples to test
multiple parameters in a single experiment.
The astrocyte bed layer is derived from neuro-
spheres, which have \cell-line"-like characteristics. They
expand with time in culture such that, from one postnatal
day 1 mouse, enough neurospheres can be generated
within 2 weeks to plate 24 3 13 mm coverslips, while
also maintaining some neurospheres for passage and fur-
ther expansion. When switched to DMEM/10%, neuro-
spheres differentiate into astrocytes that reach confluence
by 7–10 days.
Conclusions and Future Studies
We have generated a model that should be useful
for studying growth and myelination of spinal cord axons
by standard light microscopy and immunofluorescence
techniques. The myelinated axons grow in thin, virtual
monolayers on top of the astrocytes and as such do not
require confocal microscopy or electron microscopy for
visual evaluation. Large numbers of samples can be gener-
ated from one or two postnatal day 1 mice and one em-
bryonic litter. The myelin is well formed and compacted
as depicted by electron microscopy.
These studies indicate that astrocytes derived from
neurospheres promote outgrowth of spinal cord axons,
whereas myelin sheath formation is very dependent on
medium composition. Future studies could include more
detailed investigation into the effect of other substrates,
such as laminin, extracellular matrix, and astrocytes cul-
tured from neonatal mouse forebrain, although these sub-
strates do not appear to be as supportive as neurosphere-
derived astrocytes. We have just concluded studies on the
effect of astrocyte-conditioned medium on myelination
(manuscript in preparation).
Thus, this model should aid in research into axonal
growth and methods of promoting spinal cord repair.
Additionally, it should promote studies on mechanisms of
myelination and dysmyelination as seen in myelin mutants
such as jimpy (Plpjp) and shiverer (Mbp gene mutant) mice.
Identifying factors required for myelination might also
contribute to promoting remyelination in diseases such as
Banker G, Goslin K. 1992. Primary dissociated cell cultures of nervous
tissue. In: Banker G, Goslin K, editors. Culturing nerve cells. London:
The MIT Press. p 41–73.
Barres BA, Raff MC. 1994. Control of oligodendrocyte number in the
developing rat optic nerve. Neuron 12:935–942.
Barres BA, Hart IK, Coles HSR, Burne JF, Voyvodic JT, Richardson
WD, Raff MC. 1992. Cell death and control of cell survival in the oli-
godendrocyte lineage. Cell 70:31–46.
Barres BA, Jacobson MD, Schmid R, Sendtner M, Raff MC. 1993. Does
oligodendrocyte survival depend on axons? Curr Biol 3:489–497.
Billings-Gagliardi S, Adcock LH, Schwing GB, Wolf MK. 1980. Hypo-
myelinated mutant mice. II. Myelination in vitro. Brain Res 200:135–
Bornstein MB. 1973. Organotypic mammalian central and peripheral
nerve tissue. In: Kruse PF, Patterson MK, editors. Tissue culture.
Methods and applications. New York: Academic Press. p 86–92.
Bottenstein JE, Sato GH. 1979. Growth of a rat neuroblastoma cell line
in serum-free supplemented medium. Proc Natl Acad Sci U S A
Brewer GJ, Torricelli JR, Evege EK, Price PJ. 1993. Optimized survival
of hippocampal neurons in B27-supplemented Neurobasal, a new se-
rum-free medium combination. J Neurosci Res 35:567–576.
Bunge RP, Wood P. 1973. Studies on the transplantation of spinal cord
tissue in the rat. 1. The development of a culture system for hemisec-
tions of embryonic spinal cord. Brain Res 57:261–276.
Chan JR, Watkins TA, Cosgaya JM, Zhang C-Z, Chen L, Reichardt LF,
Shooter EM, Barres BA. 2004. NGF controls axonal receptivity to myeli-
nation by Schwann cells or oligodendrocytes. Neuron 43:183–191.
Demerens C, Stankoff B, Logak M, Anglade P, Allinquant B, Couraud
F, Zalc B, Lubetzki C. 1996. Induction of myelination in the central
nervous system by electrical activity. Proc Natl Acad Sci U S A
Fanarraga ML, Sommer I, Griffiths IR, Montague P, Groome NP, Nave
K-A, Schneider A, Brophy PJ, Kennedy PGE. 1993. Oligodendrocyte
development and differentiation in the rumpshaker mutation. Glia
Ga ¨hwiler BH, Thompson SM, Audinat E, Robertson RT. 1992. Orga-
notypic slice cultures of neural tissue. In: Banker G, Goslin K, editors.
Culturing nerve cells. London: The MIT Press. p 379–411.
Ga ¨hwiler BH, Capogna M, Debanne D, McKinney RA, Thompson SM.
1997. Organotypic slice cultures: a technique has come of age. Trends
Ghoumari AM, Ibanez C, El Etr M, Leclerc P, Eychenne B, O’Malley
BW, Baulieu EE, Schumacher M. 2003. Progesterone and its metabo-
1714 Thomson et al.
Journal of Neuroscience Research DOI 10.1002/jnr
lites increase myelin basic protein expression in organotypic slice cul-
tures of rat cerebellum. J Neurochem 86:848–859.
Gyllensten L, Malmfors T. 1963. Myelinization of the optic nerve and its
dependence on visual function—a quantitative investigation in mice.
J Embryol Exp Morphol 11:255–266.
Hof PR, Trapp BD, DeVellis J, Claudio L, Colman DR. 1999. The cel-
lular components of nervous tissue. In: Zigmond MJ, Bloom FE, Landis
SC, Roberts JL, Squire LR, editors. Fundamental neuroscience. Lon-
don: Academic Press. p 41–70.
Lubetzki C, Demerens C, Anglade P, Villarroya H, Frankfurter A, Lee
VMY, Zalc B. 1993. Even in culture, oligodendrocytes myelinate solely
axons. Proc Natl Acad Sci U S A 90:6820–6824.
Matthieu J-M, Honegger P, Trapp BD, Cohen SR, Webster H. 1978.
Myelination in rat brain aggregating cell cultures. Neuroscience 3:565–572.
Meyer-Franke A, Barres BA. 1996. Oligodendrocytes do not myelinate
axons by default. Soc Neurosci Abstr 22:49.
Meyer-Franke A, Shen SL, Barres BA. 1999. Astrocytes induce oligoden-
drocyte processes to align with and adhere to axons. Mol Cell Neurosci
Mithen FA, Wood PM, Agrawal HC, Bunge RP. 1983. Immunohisto-
chemical study of myelin sheaths formed by oligodendrocytes interact-
ing with dissociated dorsal root ganglion neurons in culture. Brain Res
Munoz-Garcia D, Ludwin SK. 1986. Gliogenesis in organotypic tissue
culture of the spinal cord of the embryonic mouse. I. Immunocyto-
chemical and ultrastructural studies. J Neurocytol 15:273–290.
Notterpek LM, Bullock PN, Malek-Hedayat S, Fisher R, Rome LH.
1993. Myelination in cerebellar slice cultures: development of a system
amenable to biochemical analysis. J Neurosci Res 36:621–634.
Phelps PE, Barber RP, Vaughn JE. 1996. Nonradial migration of inter-
neurons can be experimentally altered in spinal cord slice cultures. De-
Raff MC, Lillien LE, Richardson WD, Burne JF, Noble M. 1988. Plate-
let-derived growth factor from astrocytes drives the clock that times oli-
godendrocyte development in culture. Nature 333:562–565.
Reynolds BA, Weiss S. 1996. Clonal and population analyses demon-
strate that an EGF-responsive mammalian embryonic CNS precursor is
a stem cell. Dev Biol 175:1–13.
Schachner M, Kim SK, Zehnle R. 1981. Developmental expression in
central and peripheral nervous system of oligodendrocyte cell surface
antigens (O antigens) recognized by monoclonal antibodies. Dev Biol
Schnitzer J, Schachner M. 1982. Cell type specificity of a neural cell sur-
face antigen recognized by the monoclonal antibody A2B5. Cell Tissue
Seil FJ. 1979. Cerebellum in tissue culture. In: Schneider DM, editor.
Reviews of neuroscience. New York: Raven Press. p 105–177.
Seil FJ, Herndon RM. 1991. Myelination and glial ensheathment of
Purkinje cells in cerebellar cultures are not inhibited by antibodies to
the neural cell adhesion molecule, N-CAM. Int J Dev Neurosci 9:587–
Shaw CE, Milner R, Compston AS, ffrench-Constant C. 1996. Analysis
of integrin expression on oligodendrocytes during axo-glial interaction
by using rat-mouse xenocultures. J Neurosci 16:1163–1172.
Shrager P, Novakovic SD. 1995. Control of myelination, axonal growth,
and synapse formation in spinal cord explants by ion channels and elec-
trical activity. Brain Res Dev Brain Res 88:68–78.
Sommer I, Schachner M. 1981. Monoclonal antibodies (O1–O4) to oli-
godendrocyte cell surfaces: an immunocytological study in the central
nervous system. Dev Biol 83:311–327.
Stanhope GB, Wolf MK, Billings-Gagliardi S. 1986. Genotype-specific
myelin formation around normal axons in cytosine arabinoside-treated
organotypic cultures injected with normal or shiverer optic nerve. Brain
Svenningsen AF, Shan WS, Colman DR, Pedraza L. 2003. Rapid
method for culturing embryonic neuron-glial cell cocultures. J Neurosci
Sykes JEC, Lopes-Cardozo M. 1988. The effect of serum on lipid synthesis
and on the expression of oligodendrocyte marker-enzymes in oligoden-
drocyte-enriched glial cells in culture. Neurochem Int 12:467–474.
Thomson CE, Anderson TJ, McCulloch MC, Dickinson PJ, Vouyiou-
klis DA, Griffiths IR. 1999. The early phenotype associated with the
jimpy mutation of the proteolipid protein gene. J Neurocytol 28:207–
Warringa RAJ, Hoeben RC, Koper JW, Sykes JEC, van Golde LMG.
1987. Hydrocortisone stimulates the development of oligodendrocytes
in primary glial cultures and affects glucose metabolism and lipid syn-
thesis in these cultures. Brain Res Dev Brain Res 34:79–86.
Wood PM, Bunge RP. 1986. Myelination of cultured dorsal root gan-
glion neurons by oligodendrocytes obtained from adult rats. J Neurol
Wood PM, Williams AK. 1984. Oligodendrocyte proliferation and CNS
myelination in cultures containing dissociated embryonic neuroglia and
dorsal root ganglion neurons. Brain Res 314:225–241.
Yamamura T, Konola JT, Wekerle H, Lees MB. 1991. Monoclonal anti-
bodies against myelin proteolipid protein: identification and characteri-
zation of two major determinants. J Neurochem 57:1671–1680.
Yavin Z, Yavin E. 1977. Synaptogenesis and myelinogenesis in dissoci-
ated cerebral cells from rat embryo on polylysine coated surfaces. Exp
Brain Res 29:137–147.
Zhang SC, Lundberg C, Lipsitz D, O’Connor LT, Duncan ID. 1998.
Generation of oligodendroglial progenitors from neural stem cells.
J Neurocytol 27:475–489.
In Vitro Axonal Growth and Myelination1715
Journal of Neuroscience Research DOI 10.1002/jnr