DOI: 10.1093/brain/awg126
Advanced Access publication April 8, 2003
Brain (2003), 126, 1382±1391
Impaired remyelination and depletion of
oligodendrocyte progenitors does not occur
following repeated episodes of focal demyelination
in the rat central nervous system
Jacques Penderis,1,2 Simon A. Shields1 and Robin J. M. Franklin1
1Cambridge
Centre for Brain Repair and Department of
Clinical Veterinary Medicine, University of Cambridge,
Madingley Road, Cambridge, 2Animal Health Trust,
Lanwades Park, Kentford, Newmarket, UK
Summary
It has been hypothesized that the progressive failure of
remyelination in chronic multiple sclerosis is, in part,
the consequence of repeated episodes of demyelination
at the same site, eventually depleting oligodendrocyte
progenitor cells (OPCs) and exhausting the remyelinating capacity. We investigated the effect of previous
focal, ethidium bromide-induced demyelination of brain
stem white matter (with intervening recovery) on the
ef®ciency of the remyelination process during second
and third subsequent episodes of demyelination, and the
OPC response during a second episode of demyelination. Previous focal demyelinating lesions followed by
recovery did not result in any retardation of the remyelination process, nor did they alter the proportion of
Correspondence to: Dr R. J. M. Franklin, Department of
Clinical Veterinary Medicine, University of Cambridge,
Madingley Road, Cambridge, CB3 0ES, UK
E-mail: [email protected]
Schwann cell versus oligodendrocyte remyelination. The
OPC response during remyelination was quanti®ed by
in situ hybridization using a probe to platelet-derived
growth factor-a receptor (PDGFaR), an OPC-expressed
mRNA. Following recovery from focal, toxin-induced
CNS demyelination, the OPC density returned to levels
equivalent to those in normal white matter. Furthermore, there was no depletion of OPCs following
repeated episodes of focal, toxin-induced CNS demyelination at the same site. These results indicate that
repeated CNS demyelination, which has the opportunity
to repair in the intervening period, is not characterized
by impaired remyelination or depletion of OPCs.
Keywords: demyelination, multiple sclerosis, oligodendrocyte progenitor, platelet-derived growth factor-a receptor,
remyelination
Abbreviations: CCP = caudal cerebellar peduncle; EAE = experimental autoimmune encephalomyelitis; EB = ethidium
bromide; FGF = ®broblast growth factor; MBP = myelin basic protein; MOG = myelin oligodendrocyte glycoprotein; OPC
= oligodendrocyte progenitor cell; PLP = proteolipid protein; PDGF = platelet-derived growth factor; PDGFaR = plateletderived growth factor-a receptor
Introduction
Although remyelination occurs in many acute plaques of
multiple sclerosis (Prineas and Connell, 1979; Prineas et al.,
1993a; Lassmann et al., 1997), remyelination failure
becomes an increasingly prominent feature of the pathology
of chronic multiple sclerosis (Prineas et al., 2002). The
progressive failure of remyelination during the later stages of
the disease results in impaired axonal conduction (Smith and
McDonald, 1999), and may contribute to the axonal atrophy
that is an important component of the secondary progressive
phase of the disease (Kornek et al., 2000). The reasons for
Brain 126 ã Guarantors of Brain 2003; all rights reserved
remyelination failure in multiple sclerosis are not fully
understood, and understanding why a relatively robust
regenerative process should lose momentum is an important
prerequisite for developing an effective therapeutic solution
(Franklin, 2002). One widely entertained hypothesis is that
repeated episodes of demyelination at the same site eventually exhaust the capacity of the tissue to mount a remyelinating response (Ludwin, 1980), possibly due to the gradual
depletion of oligodendrocyte progenitor cells (OPCs) in and
around the lesion (Johnson and Ludwin, 1981; Carroll et al.,
Remyelination following repeat demyelination
1383
However, these limitations are substantially resolved in a
model involving stereotaxic injections of gliotoxins into the
caudal cerebellar peduncles (CCPs; equivalent to the inferior
cerebellar peduncle in humans) of adult rats (Woodruff and
Franklin, 1999). Using this model we have been able to test
three predictions arising from the repeat demyelination-OPC
depletion hypothesis of remyelination failure: (i) that following an episode of focal demyelination and remyelination,
there will be an impairment in the extent of remyelination
after subsequent episodes of demyelination affecting the
same area; (ii) that the local OPC population will be depleted
following successful remyelination of a demyelinating lesion;
and (iii) that this depletion of the local OPC population results
in an impaired OPC response to subsequent episodes of
demyelination.
Materials and methods
Induction of focal demyelination
Fig. 1 Schematic representation of the timing and location of
demyelinating lesions in order to assess the extent and nature of
the remyelination process by standard histological techniques.
1998; Keirstead et al., 1998; Franklin, 1999). We call this
hypothesis the repeat demyelination-OPC depletion hypothesis. Some supportive evidence has been obtained from the
examination of multiple sclerosis lesions (Prineas et al.,
1993b), and from experimental models such as repeated or
sustained exposure to the dietary demyelinating agent
cuprizone, which lead to impaired remyelination (Ludwin,
1980; Johnson and Ludwin, 1981). A similar impairment of
remyelination occurs in a chronic relapsing experimental
autoimmune encephalomyelitis (EAE) model induced by
repeated adoptive transfer of encephalitogenic T-cells and
anti-MOG (myelin oligodendrocyte glycoprotein) antibodies
(Linington et al., 1992). However, in most of these experimental models, cuprizone being an exception, it is often
dif®cult to establish whether the same area of CNS white
matter has undergone successive bouts of demyelination,
remyelination and further demyelination. This makes it
dif®cult to address the repeat demyelination-OPC depletion
hypothesis of remyelination failure directly.
Experimental models of demyelination/remyelination that
involve direct injection of demyelinating agents such as the
gliotoxins ethidium bromide (EB) and lysolecithin should, in
principle, be amenable to repeat demyelination by reinjection at the same site. For the most part, these models
have involved injection into spinal cord white matter, where
the relatively small size of the tracts makes the tissue
vulnerable to axonal injury following repeat injections.
Female Sprague±Dawley rats (young adults, 8±10 weeks of
age, ~ 200 g) were used in all experiments. Experiments were
performed in compliance with UK Home Of®ce regulations
and institutional guidelines. Anaesthesia was induced using
4% iso¯uorane in oxygen. Immediately following loss of
consciousness the rats were removed from the 4%
iso¯uorane±oxygen mixture and general anaesthesia was
maintained by a continuous intravenous infusion of propofol
(Rapinovet; Mallinckrodt Veterinary Ltd, Uxbridge, UK) via
the lateral tail vein. Diazepam (C.P. Pharmaceuticals Ltd,
Wrexham, UK) was administered at 1.7 mg/kg i.p. to
facilitate anaesthesia, and buprenorphine hydrochloride
(Vetergesic; Animalcare Ltd, Hull, UK) was administered
at 0.03 mg/kg i.m. to provide analgesia. Three millilitres of
warmed sterile saline was injected (i.p.) to maintain normal
hydration during the surgical procedure and recovery.
Demyelination was induced unilaterally or bilaterally by
stereotaxic injection of 4 ml of 0.01% ethidium bromide (EB)
into the CCP as described previously in detail (Woodruff and
Franklin, 1999). In the experiments involving remyelination
analysis following a second injection of EB, we con®rmed
that stereotaxic coordinates were correct for both the ®rst and
second injection by injecting Evans Blue dye into the
peduncles of an age-matched control rat and identifying the
location of the dye following sectioning of the brain on a
sledge microtome.
Histological analysis of remyelination
Groups of four to six animals were used for histological
comparison of the character and extent of remyelination
following focal demyelination in the presence of no, one or
two previous demyelinating lesions at the same site, with
complete recovery between, as detailed in Fig. 1. Animals
were perfused with 4% glutaraldehyde in phosphate buffer
and the brains were post-®xed overnight. Tissue blocks
1384
J. Penderis et al.
Fig. 2 Schematic representation of the timing and location of
demyelinating lesions for quanti®cation of the PDGFaR-positive
oligodendrocyte progenitor cell response.
encompassing the CCP were cut as detailed previously
(Woodruff and Franklin, 1999). While maintaining their
correct orientation and sequence, blocks were processed
through osmium tetroxide, dehydrated and embedded in
TAAB (TAAB Laboratories, Aldermaston, UK) resin.
Sections (1 mm) were cut and stained with alkaline toluidine
blue. In these sections, remyelinated axons can be readily
distinguished from normally myelinated axons outside the
lesion by the decreased thickness of the myelin sheath.
Within the lesion, remyelinated axons can be distinguished
from demyelinated axons because the former possess myelin
sheaths recognizable as a dark staining rim around the axon.
The myelin of Schwann cell remyelination has a slightly
darker staining hue than that of central myelin, and the
proximity of the Schwann cell nucleus frequently gives the
myelinating Schwann cell unit in transverse section a
characteristic `signet ring' appearance, a combination of
features that enable Schwann cell remyelination to be
distinguished from oligodendrocyte remyelination. Using
these morphological criteria, the estimated proportions of
axons that had been remyelinated by oligodendrocytes, by
Schwann cells, or had remained demyelinated were recorded
on two separate occasions, blinded to status and survival time.
The mean scores and SEM were calculated for each group of
animals, and statistical analysis was performed using the
Mann±Whitney test.
PDGFaR in situ hybridization
The platelet-derived growth factor-a receptor (PDGFaR)
probe was transcribed from a 1637 bp EcoRI cDNA fragment,
encoding most of the extracellular domain of mouse
PDGFaR cloned into pBluescript KS+ (a gift from Dr N. P.
Pringle and Professor W. D. Richardson, University College
London, UK). Groups of four animals were used for the
comparison of PDGFaR-positive (+) cells in the CCP. As
detailed in Fig. 2, comparisons were performed between
PDGFaR+ cells in normal white matter and following
complete remyelination (10 weeks post lesioning), as well
as 5, 10 and 21 days following demyelination in the presence
or absence of a previous demyelinating lesion at the same site.
Animals were perfused with 4% paraformaldehyde in PBS
(phosphate-buffered saline) and tissue was prepared for in
situ hybridization performed as described previously
(Fruttiger et al., 1999), except that both the RNA polymerases
and the RNA transcription reactions were run as recommended by Boehringer Mannheim Biochemica (Mannheim,
Germany). After in situ hybridization, RNA hybrids were
visualized in situ by a standard technique as described
previously (Fruttiger et al., 1999). The density of the
PDGFaR+ cells within the lesion was assessed by image
analysis (MCID model M4; Imaging Research Inc., Toronto,
Canada). EB-injected or normal regions of the CCP, identi®ed by solochrome cyanine staining, were captured under the
43 objective using a red ®lter to accentuate the blue-stained
PDGFaR nuclei. The automatic target detection feature of the
MCID system was used to select positive nuclei, and the
threshold level was set according to the lesion background
level such that all positive nuclei were selected. The valid
criteria for counting a single cell were set as an area of
>20 mm2. The average area of PDGFaR+ cells was found to
be ~60 mm2. Therefore, to allow estimation of PDGFaR+
cells that appeared in `clumps', a target with an area of
>80 mm2 was counted as two or more cells. However, to
exclude areas of above-threshold non-speci®c staining such
as section tears and folds, targets with an area of >200 mm2
were excluded. The MCID image analysis system then
calculated the cell density of positive cells. Representative
sections were analysed from each lesion side of each animal
to derive a mean score (a mean of 10 sections per lesion side
were analysed per mean score). The individual mean scores
were then used to calculate the mean score and SEM for each
group of animals, and statistical analysis was performed using
the Mann±Whitney test.
Results
Previous focal demyelination±remyelination
does not impair subsequent remyelination
following a repeated episode of demyelination at
the same site
The repeat demyelination-OPC depletion hypothesis predicts
that the extent and rate of the remyelinating response would
become progressively impaired at that site following repeated
cycles of superimposed focal demyelination and remyelination. In order to test this, we used a model that involves
stereotaxic injection of EB into the CCP, a relatively large
cross-sectional tract of large-diameter myelinated proprioceptive ®bres en-route from the spinal cord to the cerebellar
Remyelination following repeat demyelination
Fig. 3 Toxin-induced CNS demyelination model. The solochrome
cyanine-stained section demonstrates focal demyelination of the
CCP induced by stereotaxic injection of EB into the CCP, with the
adjacent normally myelinated spinal tract of the trigeminal nerve
(sp5) indicated for comparison. Following successful repair, the
region of remyelination in the CCP appears paler than the
normally myelinated sp5 (insert: toluidine blue-stained section).
Scale bars: 500 mm.
cortex (Fig. 3). We have demonstrated previously that this
model results in rapid demyelination (Woodruff and Franklin,
1999; Sim et al., 2002b) associated with loss of mRNAs of
myelin basic protein (MBP), proteolipid protein (PLP) and
Gtx (markers of myelinating oligodendrocytes), PDGFaR (a
marker of OPCs) and Olig-1 (a marker of all oligodendrocyte
lineage cells) (Sim et al., 2000, 2002b). To con®rm that
remyelination had occurred prior to induction of further serial
lesions at the same site, the extent of remyelination was
assessed at 9 weeks following EB injection. As indicated in
Fig. 4, the process of remyelination was complete by this
time, predominantly through oligodendrocyte remyelination,
but also with some Schwann cell peripheral-type myelin and
with occasional axons that had not yet remyelinated (Shields
et al., 1999; Sim et al., 2002a).
To assess the rate and extent of remyelination in a
demyelinating lesion created in an area of remyelination,
EB was injected a second time (using modi®ed stereotaxic
coordinates to ensure correct positioning) in the CCP of rats
that had been injected with EB 10 weeks previously. The EB
was also injected into the normal white matter of the
contralateral CCP. The extent of remyelination was assessed
at 4 and 12 weeks following the second EB injection and
compared with the remyelination in the contralateral CCP
(Fig. 5). The data demonstrated that there was no difference
1385
in the remyelination at 4 weeks following the second serial
(double) EB injection as compared with a single EB injection.
At 12 weeks, although the extent of remyelination had
increased, there was similarly no difference between the
double and single lesions. At neither survival times was there
a signi®cant difference in the proportion of oligodendrocyte
versus Schwann cell remyelinated axon between the double
and single lesions.
Since a single episode of repeat demyelination may not be
regarded as a rigorous test of the hypothesis, we next assessed
the extent and character of remyelination 10 weeks after a
third serial EB injection at the same site, as compared with 10
weeks after a single contralateral EB injection (Fig. 6). In this
experiment all the lesions were mainly in the dorsal part of
the trigeminal tract, which lies immediately beneath the CCP
(see Fig. 3). The distribution of the lesion on the side that had
received three injections was the same as that on the side
receiving a single injection, indicating that these three
injections had all been into the same white matter area. We
found that there was no difference in remyelination following
three episodes of demyelination at the same site where
complete recovery is allowed in the intervening period, as
compared with a single episode of demyelination.
The introduction of a syringe needle and the injection of EB
inevitably cause some degree of axonal injury. It might be
argued that second and third injections into the same area will
result in enough axonal damage to cause a signi®cant decrease
in the lesion size. For a lesion that remyelinates from its edge
to its centre (Sim et al., 2000, 2002b), this would decrease the
time required for complete remyelination in the multiple
injection lesions, thereby potentially masking a decrease in
remyelination ef®ciency. However, when we compared the
area of the trigeminal tract containing both the remyelinated
lesion and the surrounding intact white matter in single- (n =
4) and triple- (n = 4) injected animals with the trigeminal tract
in unlesioned animals (n = 8) we could ®nd no evidence of a
signi®cant decrease in lesion size [control area = 0.86 mm2 6
0.01 (SEM); single injection = 0.85 mm2 6 0.02; triple
injection = 0.89 mm2 6 0.01]. Thus, while there is likely to be
cumulative axonal damage it is unlikely to be suf®ciently
large to alter the interpretation of the results signi®cantly.
PDGFaR+ oligodendrocyte progenitor cell
numbers return to levels equivalent to those in
normal white matter following recovery from
EB-induced focal demyelination
The OPC depletion hypothesis predicts that reduced numbers
of OPCs occur within and around an area of remyelination
when compared with normal white matter. This was
addressed by quantifying the numbers of PDGFaR mRNAexpressing OPCs at 10 weeks following an EB demyelinating
lesion in a group of four animals, as compared with levels in
the unlesioned CCP. At 10 weeks following demyelinating
injury we could detect no difference between the numbers of
PDGFaR mRNA-expressing OPCs within the area of
1386
J. Penderis et al.
Fig. 4 (A) Graph demonstrating successful repair of EB-induced demyelination in the CCP, 9 weeks
post-lesion induction (6 SEM). (B) The normal CCP is characterized by predominantly large-diameter
myelinated axons arranged in parallel. (C) Oligodendrocyte-remyelinated axons can be readily
distinguished from normally myelinated axons outside the lesion by the thinness of the myelin sheath.
(D) The myelin of Schwann cell-remyelination has a slightly darker staining hue than that of central
myelin, and the proximity of the Schwann cell nucleus frequently gives the myelinating Schwann cell
unit in transverse section a characteristic `signet ring' appearance (arrows). (E) Demyelinated axons are
identi®ed by the absence of a surrounding myelin sheath (2 weeks post-lesion induction, for illustration
purposes); large numbers of lipid-®lled macrophages are present in this early lesion. All sections are 1mm toluidine blue resin sections. Scale bars: 10 mm.
remyelination and in the equivalent area of the contralateral
intact CCP (Fig. 7). Both the area of remyelination and the
intact CCP were identi®ed on adjacent solochrome cyaninestained sections. These ®ndings indicate that following
successful remyelination, PDGFaR mRNA-expressing
OPCs return to levels equivalent to those in normal white
matter, with no demonstrable sustained depletion of OPCs.
Moreover, since there was a similar distribution of OPCs
throughout the CCP containing an area of remyelination, we
could ®nd no evidence that remyelinated areas of white
matter contain different densities of OPCs to normally
myelinated white matter.
The PDGFaR+ oligodendrocyte progenitor cell
response following focal demyelination is not
impaired by a previous episode of remyelination
at the same site
The OPC depletion hypothesis also predicts that the OPC
response will be impaired during the repair response to
subsequent episodes of superimposed demyelination. The EB
model of demyelination is characterized by depletion of
PDGFaR and Olig-1 mRNA-expressing OPCs within the
lesion, and recruitment of new OPCs occurs from the
surrounding intact white matter (Sim et al., 2002b). We
could therefore use this model to establish whether a previous
episode of successful remyelination altered the number of
PDGFaR mRNA-expressing OPCs generated following a
subsequent episode of EB-induced demyelination at the same
site. We determined the PDGFaR mRNA+ cell density
within the area of demyelination in groups of four animals 5,
10 and 21 days after EB injection in the presence or absence
of a previous demyelinating lesion at the same site 10 weeks
previously. The area of demyelination for determining
PDGFaR mRNA+ cell density was delineated on adjacent
solochrome cyanine-stained sections. We found that there
was no depletion of the PDGFaR mRNA+ OPC response in
the presence of a previous demyelinating lesion (Fig. 8) and
that the absolute densities were comparable to levels that we
have previously been demonstrated in rats of a similar age
and strain (Sim et al., 2002b). This observation suggests that
the OPC response is the same regardless of the previous
history of demyelination±remyelination at the same site.
Discussion
The variable nature of the reparative response following
primary demyelination of the CNS, where some lesions
Remyelination following repeat demyelination
1387
Fig. 5 Comparison of the character and extent of remyelination (6 SEM) 4 and 12 weeks following EB
injection into a remyelinated area (®lled bars) and normal white matter (shaded bars). No retardation of
either the rate or extent of remyelination is evident at 4-weeks and 12-weeks post lesion induction, with
remyelination primarily oligodendrocyte-derived.
Fig. 6 No evidence of impaired remyelination is present following
three episodes of demyelination at the same site (black bars)
where complete recovery is allowed in the intervening period, as
compared with a single episode of demyelination (grey bars) (6
SEM).
should undergo extensive and even complete remyelination
while others remain largely demyelinated, constitutes one of
the most frustrating aspects of multiple sclerosis. If each new
lesion were to remyelinate fully then in all likelihood the
disease would carry a more favourable prognosis.
Understanding why remyelination is incomplete or fails
entirely is critically important in devising therapeutic
approaches through which it might be improved.
A number of hypotheses have been proposed to explain
remyelination failure (reviewed in Franklin, 2002), one of
which has as its basis the predisposition in multiple sclerosis
for demyelinating episodes to repeatedly affect the same area
of white matter (Prineas et al., 1993b). The repair of focal
demyelination involves the local recruitment of OPCs, which
engage demyelinated axons and differentiate into remyelinating oligodendrocytes. It has been proposed that should the
same area of white matter be required to mount a repair
response repeatedly following cycles of demyelination and
remyelination then the number of OPCs available for
recruitment declines, resulting in an impairment of remyelination. Indeed, one study reported a persistent decrease in OPC
number around a demyelinating lesion following a single
episode of demyelination (Keirstead et al., 1998), although
this observation is at variance with other studies indicating
the converse (Cenci di Bello et al., 1999; Levine and
Reynolds, 1999; Sim et al., 2002b). We have called this
1388
J. Penderis et al.
Fig. 7 (A) The CCP at 70 days post-lesion (70 dpl) was identi®ed histologically by solochrome cyanine
(SC) staining on adjacent sections, and this was then used to de®ne the region for oligodendrocyte
progenitor determination following in situ hybridization for PDGFaR mRNA. (B) No depletion of
oligodendrocyte progenitors is evident when comparing the density of PDGFaR mRNA-positive cells in
the CCP 10 weeks following a demyelinating lesion (grey bar) with normal CCP (white bar) (6SEM).
Scale bars: 500 mm.
Fig. 8 (A) Density of PDGFaR mRNA-positive oligodendrocyte progenitors 10 days post (dp) a single
EB lesion, with the extent of the demyelinating lesion de®ned by solochrome cyanine (SC) staining of
the adjacent serial section, and (B) 21 days after an EB lesion superimposed on a remyelinated region.
(C) Quanti®cation of the PDGFaR mRNA-positive cell density 5, 10 and 21 days following an EB
lesion demonstrated no signi®cant alteration in the presence (black bars) or absence (grey bars) of a
demyelinating lesion at the same site 10 weeks previously (6 SEM). Scale bars: 500 mm.
Remyelination following repeat demyelination
hypothesis the repeat demyelination-OPC depletion hypothesis, and have tested it using a model in which the same area
of white matter can be repeatedly targeted for demyelination
by the stereotaxic injection of a toxin that causes acute
demyelination. We have found that the same region of white
matter can undergo complete remyelination after each of
three sequential episodes of demyelination, implying that an
area of remyelination will remyelinate as ef®ciently as an
area of normally myelinated tissue should they become
demyelinated. While it is possible that one or more of the
injections failed to induce a lesion, we regard this as unlikely
given that failure to induce a lesion by stereotaxic injection of
EB into brain stem white matter occurs infrequently.
Moreover, we have also shown that when remyelination is
complete the numbers of OPCs have returned to levels
equivalent to those that existed before demyelination and
remyelination occurred, and that a repeat wave of remyelination does not cause a progressive depletion of OPCs.
These results are consistent with recent studies indicating
ef®cient repopulation by OPCs of areas of normal white
matter from which they have been depleted by X-irradiation
(Hinks et al., 2001; Chari and Blakemore, 2002), and provide
further evidence of the robust capacity of OPCs to repopulate
areas from which they are depleted even though the rate at
which this occurs may decline with age (Sim et al., 2002b).
The toxin model we have used, however, differs from
multiple sclerosis in many ways and is not intended to
provide an experimental `facsimile' of multiple sclerosis,
which is in any case a disease of considerable diversity. It
does, however, provide a powerful model for studying the
biology of the repair process in a manner that is not
confounded by the persistence of a demyelinating stimulus.
One can argue that results obtained with these models should
be extrapolated to immune-mediated in¯ammatory disease
such as multiple sclerosis with some caution, and should be
supported by data from a model more similar to multiple
sclerosis in its aetiology and pathogenesis. On the other
hand, one can also argue that it provides insights into the
fundamental principles of the myelin repair process, broadly
applicable across the spectrum of demyelinating disease, that
do not emerge with similar clarity in more complex models.
If the process of remyelination per se does not inhibit
subsequent remyelination, why then do some models of repeat
demyelination, such as the EAE model based on repeated cotransfer of MBP-speci®c T cells in combination with
antibodies speci®c to MOG (Linington et al., 1992), lead to
persistent demyelination? A key feature of the experimental
design in the present study was the period of time between
successive episodes of demyelination. These times were
chosen to allow ample opportunity for the regenerative
response to occur within an environment free of a demyelinating stimulus. The results indicate that in such circumstances,
ef®cient remyelination occurs regardless of whether the
region has undergone remyelination previously. A more
severe challenge to the remyelinating process would occur if
the region were exposed to demyelinating stimuli while
1389
remyelination was still in progress, where its repeated
interruption may indeed then lead to remyelination failure.
Such a situation is likely to occur in the chronic relapsing form
of MBP-induced EAE augmented by MOG antibodies, where
demyelinating stimuli may persist while remyelination takes
place and are supplemented at ~3-week intervals (Linington
et al., 1992). More deleterious still would be a situation where
there are attempts at remyelination in the face of a constant
exposure to a demyelinating stimulus. As soon as new
remyelinating oligodendrocytes are generated they would
succumb to the demyelinating stimulus and thus lead to a
relentless demand for the provision of new OPCs. Such a
situation has been modelled in mice fed the demyelinating
agent cuprizone over a prolonged period (6±7 months), a
protocol that leads to sustained demyelination with very few
oligodendrocytes and hence very little remyelination
(Ludwin, 1980). However, whether the poor remyelination
in chronic cuprizone is due to a reduction in OPCs has not
been demonstrated and, despite the apparent ultrastructural
normality of both the axons and astrocytes in the chronic
cuprizone model (Ludwin, 1980), it may still be the case that
the expression of remyelinating signals from both these cell
types is altered. For example, the expression pro®le of
secreted growth factors involved in remyelination by astrocytes, such as FGF or PDGF, may change without there being
any histologically obvious change in their morphology
(Redwine and Armstrong, 1998; Hinks and Franklin, 1999;
Hinks and Franklin, 2000; Armstrong et al., 2002).
A confounding feature of models in which demyelinating
agents are repeatedly or chronically administered is the effect
on remyelination attributable to ageing. As animals age there
is a profound decrease in the rate at which remyelination
occurs (Gilson and Blakemore, 1993; Shields et al., 1999;
Sim et al., 2000). This phenomenon is associated with an
impairment of both OPC recruitment and differentiation, and
not with an age-associated decline in the number of OPCs in
normal white matter available for recruitment (Sim et al.,
2002b). Experiments conducted over several months in
rodents are of suf®cient duration for there to be a signi®cant
slowing of remyelination rate as the animals age, and this
alone may contribute to an apparent failure of remyelination,
independent of the pattern or manner of demyelination.
The results of the present study challenge the concept that
remyelination fails because of OPC depletion. Further
evidence challenging this notion has come from studies on
multiple sclerosis tissue, which indicate that areas of chronic
demyelination can exist in which OPCs and immature
oligodendrocytes can be detected (Scolding et al., 1998;
Wolswijk, 1998; Chang et al., 2000, 2002). In these situations
it would appear that remyelination may fail not as a result of a
shortage of OPCs, but rather due to their inability to
differentiate fully into myelin-forming cells. Whether this is
due to the absence of differentiation inducing signals, the
presence of differentiation inhibitors such as the notch-jagged
signalling pathway (Wang et al., 1998; John et al., 2002) and
PSA-NCAM (polysialylated neural cell adhesion molecule)
1390
J. Penderis et al.
(Charles et al., 2002), or due to changes in the properties of
demyelinated axons that render them incapable of being
remyelinated (Chang et al., 2002) is not fully resolved.
A case can thus be made that the onus of blame for the
eventual failure of remyelination lies more with the environmental signals to which the OPCs respond during remyelination than with the OPCs themselves. The environment that
leads to remyelination is likely to have many diverse
components, the timing and extent of their expression being
critical for the success of the repair process. If these signals
are inappropriately expressedÐthe dysregulation hypothesis
of remyelination failure (Franklin, 2002)Ðthen remyelination may fail. Recent years have seen some signi®cant
advances in the identity of growth factors critical for
remyelination (reviewed in Franklin et al., 2001), although
whether there is a therapeutic role for individual growth
factors is not clear, contrasting results being obtained with
different models and different routes of delivery (Yao et al.,
1995; Cannella et al., 1998; Cannella et al., 2000; O'Leary
et al., 2002). A signi®cant development has been the
identi®cation of the pro-remyelinating role of the in¯ammatory response to demyelination (Ludwin, 1980; Graca and
Blakemore, 1986; Warrington et al., 2001; Franklin et al.,
2002). Depletion of macrophages or of key in¯ammatory
cytokines such as interleukin-1b and tumour necrosis factor-a
results in an impairment of remyelination following toxininduced demyelination (Arnett et al., 2001; Kotter et al.,
2001; Mason et al., 2001), while an impaired macrophage
response in old animals compared with young following
lysolecithin-induced demyelination is associated with less
ef®cient remyelination (Hinks and Franklin, 2000). An
argument has been made that in large demyelinated plaques
in multiple sclerosis, which are many fold greater in volume
than those that can be obtained in rodent models, then the
abatement of the pro-remyelinating environment is more
likely to occur before remyelination is complete, especially in
those lesions in which OPCs are de®cient (Blakemore et al.,
2002). Identifying the key environmental signals for remyelination and those whose absence accounts for remyelination
failure is thus important for developing therapeutic strategies
by which remyelination can be enhanced.
Blakemore WF, Chari DM, Gilson JM, Crang AJ. Modelling large
areas of demyelination in the rat reveals the potential and possible
limitations of transplanted glial cells for remyelination in the CNS.
Glia 2002; 38: 155±68.
Acknowledgements
Franklin RJM, Zhao C, Sim FJ. Ageing and CNS remyelination.
[Review]. Neuroreport 2002; 13: 923±8.
The authors wish to thank Anil Kalupahana for technical
assistance and Dr Chao Zhao for his help. This work was
funded by the Wellcome Trust and Cambridge Neuroscience
Inc., MA, USA.
References
Armstrong RC, Le TQ, Frost EE, Borke RC, Vana AC. Absence of
®broblast growth factor 2 promotes oligodendroglial repopulation
of demyelinated white matter. J Neurosci 2002; 22: 8574±85.
Arnett HA, Mason J, Marino M, Suzuki K, Matsushima GK,
Ting JPY. TNF alpha promotes proliferation of oligodendrocyte
progenitors and remyelination. Nat Neurosci 2001; 4: 1116±22.
Cannella B, Hoban CJ, Gao YL, Garcia-Arenas R, Lawson D,
Marchionni M, et al. The neuregulin, glial growth factor 2,
diminishes autoimmune demyelination and enhances remyelination
in a chronic relapsing model for multiple sclerosis. Proc Natl Acad
Sci USA 1998; 95: 10100±5.
Cannella B, Pitt D, Capello E, Raine CS. Insulin-like growth factor1 fails to enhance central nervous system myelin repair during
autoimmune demyelination. Am J Pathol 2000; 157: 933±43.
Carroll WM, Jennings AR, Ironside LJ. Identi®cation of the adult
resting progenitor cell by autoradiographic tracking of
oligodendrocyte precursors in experimental CNS demyelination.
Brain 1998; 121: 293±302.
Cenci di Bello I, Dawson MRL, Levine JM, Reynolds R. Generation
of oligodendroglial progenitors in acute in¯ammatory demyelinating
lesions of the rat brain stem is associated with demyelination rather
than in¯ammation. J Neurocytol 1999; 28: 365±81.
Chang A, Nishiyama A, Peterson J, Prineas J, Trapp BD. NG2positive oligodendrocyte progenitor cells in adult human brain and
multiple sclerosis lesions. J Neurosci 2000; 20: 6404±12.
Chang A, Tourtellotte WW, Rudick R, Trapp BD. Premyelinating
oligodendrocytes in chronic lesions of multiple sclerosis. New Engl
J Med 2002; 346: 165±73.
Chari DM, Blakemore WF. Ef®cient recolonisation of progenitordepleted areas of the CNS by adult oligodendrocyte progenitor
cells. Glia 2002; 37: 307±13.
Charles P, Reynolds R, Seilhean D, Rougon G, Aigrot MS,
Niezgoda A, et al. Re-expression of PSA-NCAM by demyelinated
axons: an inhibitor of remyelination in multiple sclerosis? Brain
2002; 125: 1972±5.
Franklin RJM. RemyelinationÐa regenerative process in the CNS.
[Review]. Neuroscientist 1999; 5: 184±91.
Franklin RJM. Why does remyelination fail in multiple sclerosis?
[Review]. Nat Rev Neurosci 2002; 3: 705±14.
Franklin RJM, Hinks GL, Woodruff RH, O'Leary MT. What roles
do growth factors play in CNS remyelination? [Review]. Prog Brain
Res 2001; 132: 185±93.
Fruttiger M, Karlsson L, Hall AC, Abramsson A, Calver AR,
Bostrom H, et al. Defective oligodendrocyte development and
severe hypomyelination in PDGF-A knockout mice. Development
1999; 126: 457±67.
Gilson J, Blakemore WF. Failure of remyelination in areas of
demyelination produced in the spinal cord of old rats. Neuropathol
Appl Neurobiol 1993; 19: 173±81.
Graca DL, Blakemore WF. Delayed remyelination in the rat spinal
cord following ethidium bromide injection. Neuropathol Appl
Neurobiol 1986; 12: 593±605.
Hinks GL, Franklin RJM. Distinctive patterns of PDGF-A, FGF-2,
Remyelination following repeat demyelination
1391
IGF-I and TGF-b1 gene expression during remyelination of
experimentally-induced spinal cord demyelination. Mol Cell
Neurosci 1999; 14: 153±68.
Prineas JW, Barnard RO, Revesz T, Kwon EE, Sharer L, Cho E-S.
Multiple sclerosis: pathology of recurrent lesions. Brain 1993b;
116: 681±93.
Hinks GL, Franklin RJM. Delayed changes in growth factor gene
expression during slow remyelination in the CNS of aged rats. Mol
Cell Neurosci 2000; 16: 542±56.
Prineas JW, McDonald WI, Franklin RJM. Demyelinating diseases.
In: Graham DI, Lantos PL, editors. Green®eld's neuropathology,
Vol. 2. 7th Edn. London: Arnold; 2002. p. 471±550.
Hinks GL, Chari DM, O'Leary MT, Zhao C, Keirstead HS,
Blakemore WF, et al. Depletion of endogenous oligodendrocyte
progenitors rather than increased availability of survival factors
is a likely explanation for enhanced survival of transplanted
oligodendrocyte progenitors in X-irradiated compared to normal
CNS. Neuropathol Appl Neurobiol 2001; 27: 59±67.
Redwine JM, Armstrong RC. In vivo proliferation of
oligodendrocyte progenitors expressing PDGF-aR during early
remyelination. J Neurobiol 1998; 37: 413±28.
John GR, Shankar SL, Sha®t-Zagardo B, Massimi A, Lee SC, Raine
CS, et al. Multiple sclerosis: Re-expression of a developmental
pathway that restricts oligodendrocyte maturation. Nat Med 2002;
8: 1115±21.
Johnson ES, Ludwin SK. The demonstration of recurrent
demyelination and remyelination of axons in the central nervous
system. Acta Neuropathol (Berl) 1981; 53: 93±8.
Keirstead HS, Levine JM, Blakemore WF. Response of
oligodendrocyte progenitor cell population (de®ned by NG2
labelling) to demyelination of the adult spinal cord. Glia 1998;
22: 161±70.
Kornek B, Storch MK, Weissert R, Wallstroem E, Stefferl A,
Olsson T, et al. Multiple sclerosis and chronic autoimmune
encephalomyelitis: a comparative quantitative study of axonal
injury in active, inactive, and remyelinated lesions. Am J Pathol
2000; 157: 267±76.
Kotter MR, Setzu A, Sim FJ, van Rooijen N, Franklin RJM.
Macrophage depletion impairs oligodendrocyte remyelination
following lysolecithin-induced demyelination. Glia 2001; 35: 204±12.
Scolding N, Franklin R, Stevens S, Heldin C-H, Compston A,
Newcombe J. Oligodendrocyte progenitors are present in the
normal adult human CNS and in the lesions of multiple sclerosis.
Brain 1998; 121: 2221±8.
Shields SA,
Remyelination
compared to
demyelination.
Gilson JM, Blakemore WF, Franklin RJM.
occurs as extensively but more slowly in old rats
young rats following gliotoxin-induced CNS
Glia 1999; 28: 77±83.
Sim FJ, Hinks GL, Franklin RJM. The re-expression of the
homeodomain transcription factor Gtx during remyelination of
experimentally induced demyelinating lesions in young and old rat
brain. Neuroscience 2000; 100: 131±9.
Sim FJ, Zhao C, Li W-W, Lakatos A, Franklin RJM. Expression of
the POU-domain transcription factors SCIP/Oct-6 and Brn-2 is
associated with Schwann cell but not oligodendrocyte remyelination
of the CNS. Mol Cell Neurosci 2002a; 20: 669±82.
Sim FJ, Zhao C, Penderis J, Franklin RJM. The age-related decrease
in CNS remyelination ef®ciency is attributable to an impairment of
both oligodendrocyte progenitor recruitment and differentiation. J
Neurosci 2002b; 22: 2451±9.
Lassmann H, Bruck W, Lucchinetti C, Rodriguez M. Remyelination
in multiple sclerosis. [Review]. Mult Scler 1997; 3: 133±6.
Smith KJ, McDonald WI. The pathophysiology of multiple
sclerosis: the mechanisms underlying the production of symptoms
and the natural history of the disease. [Review]. Philos Trans R Soc
Lond B Biol Sci 1999; 354: 1649±73.
Levine JM, Reynolds R. Activation and proliferation of endogenous
oligodendrocyte precursor cells during ethidium bromide-induced
demyelination. Exp Neurol 1999; 160: 333±47.
Wang S, Sdrulla AD, diSibio G, Bush G, Nofziger D, Hicks C, et al.
Notch receptor activation inhibits oligodendrocyte differentiation.
Neuron 1998; 21: 63±75.
Linington C, Engelhardt B, Kapocs G, Lassman H. Induction of
persistently demyelinated lesions in the rat following the repeated
adoptive transfer of encephalitogenic T cells and demyelinating
antibody. J Neuroimmunol 1992; 40: 219±24.
Warrington AE, Bieber AJ, Ciric B, Van Keulen V, Pease LR,
Mitsunaga Y, et al. Immunoglobulin-mediated CNS repair.
[Review]. J Allergy Clin Immunol 2001; 108: S121±5.
Ludwin SK. Chronic demyelination inhibits remyelination in the
central nervous system. Lab Invest 1980; 43: 382±7.
Mason JL, Suzuki K, Chaplin DD, Matsushima GK. Interleukin-1b
promotes repair of the CNS. J Neurosci 2001; 21: 7046±52.
O'Leary MT, Hinks GL, Charlton HM, Franklin RJM. Increasing
local levels of IGF-I mRNA expression using adenoviral vectors
does not alter oligodendrocyte remyelination in the CNS of aged
rats. Mol Cell Neurosci 2002; 19: 32±42.
Wolswijk G. Chronic stage multiple sclerosis lesions contain a
relatively quiescent population of oligodendrocyte precursor cells. J
Neurosci 1998; 18: 601±9.
Woodruff RH, Franklin RJM. Demyelination and remyelination of
the caudal cerebellar peduncle of adult rats following stereotaxic
injections of lysolecithin, ethidium bromide and complement/antigalactocerebroside: a comparative study. Glia 1999; 25: 216±28.
Prineas JW, Connell F. Remyelination in multiple sclerosis. Ann
Neurol 1979; 5: 22±31.
Yao D-L, Liu X, Hudson LD, Webster HD. Insulin-like growth
factor I treatment reduces demyelination and up-regulates gene
expression of myelin-related proteins in experimental
autoimmune encephalomyelitis. Proc Natl Acad Sci USA 1995;
92: 6190±4.
Prineas JW, Barnard RO, Kwon EE, Sharer LR, Cho E-S. Multiple
sclerosis: remyelination of nascent lesions. Ann Neurol 1993a; 33:
137±51.
Received November 13, 2002. Revised January 6, 2003.
Accepted January 10, 2003.