ARTICLES

LINGO-1 antagonist promotes spinal cord remyelination

© 2007 Nature Publishing Group http://www.nature.com/naturemedicine

and axonal integrity in MOG-induced experimental

autoimmune encephalomyelitis
Sha Mi1,7, Bing Hu2,7,8, Kyungmin Hahm1, Yi Luo1, Edward Sai Kam Hui6, Qiuju Yuan2, Wai Man Wong2,
Li Wang 2, Huanxing Su2, Tak-Ho Chu2, Jiasong Guo2, Wenming Zhang2, Kwok-Fai So2–4, Blake Pepinsky1,
Zhaohui Shao1, Christilyn Graff1, Ellen Garber1, Vincent Jung1, Ed Xuekui Wu6 & Wutian Wu2,3,5,7

Demyelinating diseases, such as multiple sclerosis, are characterized by the loss of the myelin sheath around neurons, owing to
inflammation and gliosis in the central nervous system (CNS). Current treatments therefore target anti-inflammatory mechanisms
to impede or slow disease progression. The identification of a means to enhance axon myelination would present new therapeutic
approaches to inhibit and possibly reverse disease progression. Previously, LRR and Ig domain–containing, Nogo receptor–
interacting protein (LINGO-1) has been identified as an in vitro and in vivo negative regulator of oligodendrocyte differentiation and
myelination. Here we show that loss of LINGO-1 function by Lingo1 gene knockout or by treatment with an antibody antagonist of
LINGO-1 function leads to functional recovery from experimental autoimmune encephalomyelitis. This is reflected biologically by
improved axonal integrity, as confirmed by magnetic resonance diffusion tensor imaging, and by newly formed myelin sheaths, as
determined by electron microscopy. Antagonism of LINGO-1 or its pathway is therefore a promising approach for the treatment of
demyelinating diseases of the CNS.

In multiple sclerosis, the myelin and oligodendrocytes of brain and multiple sclerosis. The model has previously been used to demon-
spinal cord white matter are the targets of T cell–mediated immune strate that fostering remyelination can be effective in moderating
attacks1, resulting in demyelination and the consequent progressive disease progression4. Approaches that enhance myelination include
disabilities and paralysis. Immunomodulatory and immuno- the promotion of oligodendrocyte precursor cell (OPC) differ-
suppressive agents can slow, but not reverse, disease progression2–4, entiation8–11 and the transplantation of OPCs, Schwann cells,
pointing to the need to develop new treatment paradigms that involve olfactory ensheathing cells and neural stem cells into primary demye-
myelin repair. linated lesions12. Here, we demonstrate that blocking LINGO-1
Recently, we demonstrated that LINGO-1 is a key negative regulator function, either through Lingo1 knockout or through treatment
of oligodendrocyte differentiation and myelination5,6. In vitro, the with anti-LINGO-1, promotes functional recovery in the EAE
overexpression of LINGO-1 inhibits oligodendrocyte differentiation model. Additionally, this functional recovery is correlated with
and myelination, whereas attenuation of its function with a dominant- improved axonal integrity and axonal myelination, as visualized by
negative allele of Lingo1, Lingo1 RNA-mediated interference, soluble magnetic resonance diffusion tensor imaging (DTI) and electron
LINGO-1, or LINGO-1 antagonist antibody (anti-LINGO-1) enhances microscopy, respectively.
oligodendrocyte differentiation and myelination6. LINGO-1 has been Lingo1-knockout mice show earlier onset of myelination of
shown to have a role in oligodendrocyte differentiation and myelina- CNS axons than the wild type, with no apparent developmental or
tion in vivo through the analysis of Lingo1-knockout and transgenic behavioral abnormalities6, suggesting that a LINGO-1 antagonist could
mice6,7. These studies suggest that inhibition of LINGO-1 function be developed into an effective and specific therapeutic approach for
could comprise a therapeutic approach for the treatment of demye- treating demyelinating diseases. To test this hypothesis, we first
linating disease. determined whether Lingo1-knockout mice show greater resistance to
Myelin oligodendrocyte glycoprotein (MOG)-induced murine the development of MOG-induced EAE than wild-type mice. An EAE
experimental autoimmune encephalomyelitis (EAE) is a widely score, which quantifies disease progression by measuring motor
accepted model for studying the clinical and pathological features of dysfunction, was used as a surrogate indicator of demyelination.

1Biogen Idec Inc., 14 Cambridge Center, Cambridge, Massachusetts 02142, USA. 2Department of Anatomy, 3State Key Laboratory of Brain and Cognitive Sciences,
4Research Centre of Heart, Brain, Hormone and Healthy Aging, 5Research Center of Reproduction, Development and Growth, Li Ka Shing Faculty of Medicine, and
6Department of Electrical and Electronic Engineering, Faculty of Engineering, The University of Hong Kong, Pokfulam, Hong Kong SAR, China. 7These authors

contributed equally to this manuscript. 8Present address: School of Life Science, The University of Science and Technology of China, Hefei, Anhui 230027, China.
Correspondence should be addressed to W.W. ([email protected]) or S.M. ([email protected]).
Received 6 April; accepted 31 August; published online 30 September 2007; doi:10.1038/nm1664

1228 VOLUME 13 [ NUMBER 10 [ OCTOBER 2007 NATURE MEDICINE

 ARTICLES

Both wild-type (WT) and Lingo1-knockout mice developed more often in the demyelinated areas of WT than in those of knockout
EAE symptoms; however, EAE scores were significantly lower in EAE mice (arrows in Fig. 1e). In WT mice, damaged myelin sheaths
Lingo1-knockout mice throughout all stages of disease progression often had loose and separated layers or degraded sheath structures
(P o 0.02; Fig. 1a). (arrows in Fig. 1f, left panel) or were completely absent (Fig. 1g, left
EAE is a complex model for demyelination, as it involves both panel). In contrast, Lingo1-knockout EAE animals showed an abun-
immune and neurological components12. Although the evidence to dance of newly formed myelin sheaths that were notably thinner
date indicates a CNS-specific role for LINGO-1, the possibility (arrowheads in Fig. 1f,g). Quantitative analysis of myelinated axons
remains that loss of LINGO-1 moderates EAE by altering the genera- indicated that there were more myelinated fibers in Lingo1-knockout
tion and the infiltration into the CNS of encephalogenic T cells EAE mice than in the WT EAE controls (P ¼ 0.0002; Fig. 1h). The
involved in EAE pathology. These possibilities can be excluded by combined data show that LINGO-1 deficiency probably results in an
© 2007 Nature Publishing Group http://www.nature.com/naturemedicine

testing for functional differences between T cells from WT and altered CNS compartment in which remyelination is encouraged. The
knockout mice, followed by examination of the ability of encephalo- deficiency did not influence the ability of encephalogenic T cells and
genic T cells from WT and knockout mice to confer EAE through inflammatory immune effector cells to cause EAE.
adoptive transfer. We isolated T cells from both WT and Lingo1- The mitigation of EAE progression in Lingo1-knockout mice
knockout mice after MOG immunization and measured their pro- suggests that exogenous means, such as the use of antibody antago-
liferation and cytokine release after MOG stimulation. As expected, nists to block endogenous LINGO-1 function, could slow EAE
LINGO-1 deficiency did not alter Lingo1-knockout T-cell responses to progression. Anti-LINGO-1 is an IgG1-isotype antibody that binds
MOG in the proliferation assay as compared to the WT control specifically to LINGO-1 (Supplementary Fig. 1 online) to promote
(Fig. 1b). Also, there was no difference in the amounts of cytokines oligodendrocyte differentiation and myelination in vitro (S.M.,
(interleukin (IL)-2, IL-4, IL-5, IL-6, IL-10, IL-
13, IL-17, tumor necrosis factor (TNF)-a, WT Lingo1 KO
interferon (IFN)-g and granulocyte mono- a e
KO n = 11
cyte colony stimulating factor (GM-CSF)) 3 WT n = 9
released by T cells obtained from MOG
EAE clinical score

stimulation of WT or Lingo1-knockout mice 2

(data not shown). Consistent with normal
1
T-cell function, MOG-immunized encepha-
logenic T cells isolated from Lingo1-knockout 0
mice that were adoptively transferred into 0 5 10 15 20 25 30
Time (d) after immunization
WT mice were able to induce EAE simi-
larly to those from WT mice (Fig. 1c). In
contrast, encephalogenic T cells isolated from b WT n = 5

WT mice were markedly reduced in their 20

KO n = 5 f
Stimulation index

ability to induce EAE in Lingo1-knockout 15

mice, in comparison to WT recipients
10
(P o 0.006; Fig. 1d).
5
We used electron microscopy to show that
lower EAE scores do indeed reflect remyeli- 0
0 5 10 25 50
nation. Characteristic ‘dying-back’ oligoden- MOG concentration (µg/ml)
drogliopathy presenting as an ‘inside-out’
type of myelin damage13,14 was observed c
WT to WT n = 9
4 KO to WT n = 11
g
Figure 1 Lower EAE clinical scores and
EAE clinical score

increased remyelination in MOG-induced EAE 3

Lingo1-knockout mice. (a) Mean clinical 2
scores of MOG-induced EAE in WT and
LINGO-1–deficient mice. *P o 0.02. (b) T-cell 1
proliferation response in the presence of MOG 0
as determined by 3H-thymidine incorporation. 0 10 20 30 40
(c) Adoptive transfer of EAE into WT mice by Time (d) after transfer
MOG-immunized encephalogenic T cells isolated
from Lingo1-knockout or WT mice. (d) Adoptive d
transfer of EAE in Lingo1-knockout mice or WT 4 h 900
mice by encephalogenic T cells isolated from WT WT to WT n = 10
EAE clinical score

myelinated axons

mice. *P o 0.006. (e–g) EM images showing 3 WT to KO n = 10

Number of

600 *
remyelination in the lumbar spinal cords of 2
Lingo1-knockout mice (right) and demyelination
300
in WT control mice (left). (h) Quantification of 1
toluidine blue–stained sections, showing
0 0
significantly more myelinated axons in the 0 10 20 30 40
T

KO
T

W
W

spinal cords of Lingo1-knockout EAE mice Time (d) after transfer

o1
E
al

EA
m

ng
or

(n ¼ 7) than in WT EAE mice (n ¼ 6).

Li
N

E
EA

*P ¼ 0.0002. KO, knockout.

NATURE MEDICINE VOLUME 13 [ NUMBER 10 [ OCTOBER 2007 1229

 ARTICLES

unpublished data). Here, we tested whether LINGO-1 antagonist The decreased EAE severity after anti-LINGO-1 antagonist treat-
antibody can promote functional recovery and myelination in the ment indicates improved axonal function and integrity, which can be
rat EAE model. We first determined whether the local delivery of anti- visualized by DTI. We collected DTI images of the lumbar segments of
LINGO-1 can be prophylactic for the onset or the progression of EAE. post-fixed spinal cords. Low fractional anisotropy values were
Anti-LINGO-1 was introduced 3 d after MOG immunization by obtained from the dorsal area of the spinal cord in IgG control–
intrathecal delivery by osmotic pump, and the experiment was treated EAE rats (Fig. 2c, top center), which correlated with the
stopped after 36 d for tissue examination by histology and electron demyelination area seen in the toluidine blue–stained section (Fig. 2c,
microscopy. Similarly to the results in Lingo1-knockout mice, anti- bottom center). In contrast, substantially higher fractional anisotropy
LINGO-1 did not alter EAE onset, but did significantly mitigate values were obtained from anti-LINGO-1–treated (Fig. 2c, top right)
disease severity across all stages of disease progression, in comparison or normal control (Fig. 2c, upper left) rats, which correlated with the
© 2007 Nature Publishing Group http://www.nature.com/naturemedicine

to an isotype antibody control or a no-treatment control, as indicated relatively normal myelination seen in the histological sections (Fig. 2c,
by EAE scores (P o 0.05; Fig. 2a). bottom right or left, respectively). Fractional anisotropy values from
Bolstered by the prophylactic data, we undertook a more clinically projection DTI images of the anti-LINGO-1–treated group (Fig. 2d,
relevant assessment of the therapeutic potential of LINGO-1 antagon- right) were significantly higher than those of the IgG control–treated
ism in rats already exhibiting EAE symptoms. MOG-induced rats with group (Fig. 2d, left, and Fig. 2e; P ¼ 0.0034), indicating improved
EAE scores of 1.0 were treated with either anti-LINGO-1 or an isotype axonal health and remyelination. Similarly, spinal cords obtained
control antibody delivered systemically by intraperitoneal injection. from mice that had systemic intraperitoneal delivery of anti-
After a 2-week treatment, the anti-LINGO-1 group showed signifi- LINGO-1 had higher fractional anisotropy values than the isotype
cantly lower EAE scores than the control group (P o 0.05; Fig. 2b). antibody control group (P ¼ 0.0083; Fig. 2f). The health of remye-
Most notably, disease progression appeared stabilized, with a percep- linated axons was also confirmed by immunostaining for amyloid-b
tible downtrend (Fig. 2b). precursor protein (APP), a marker for axonal damage shown to be
increased in multiple sclerosis lesions15. Fewer
APP-positive elements were observed in the
a 3 b 3
white matter of the anti-LINGO-1–treated
No treatment n = 8 IgG control n = 8
2.5 IgG control n = 13 2.5 Anti-Lingo-1 n = 9 EAE spinal cords than in that of the IgG
EAE clinical score

EAE clinical score

2
Anti-Lingo-1 n = 15 2 control–treated cords (Supplementary
1.5 1.5
Fig. 2 online). This LINGO-1 functional
antagonist therefore has potential therapeutic
1 1
efficacy whether it is delivered locally or
0.5 0.5
systemically and is efficacious before or after
0 0 the manifestation of EAE symptoms.
0
2
4
6
8
10
12
14
16
18
20
21
22
23
24
25
26
27
28
29
30
31
32
33
0
1
2
3
4
5
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36

We next determined whether the apparent

Time (d) after immunization Time (d) after immunization
functional improvement after anti-LINGO-1
c Normal IgG control Anti-LINGO-1 treatment (indicated by lowered EAE scores
and increased DTI image densities) correlates
DTI images

Figure 2 Treatment with an antibody antagonist

to LINGO-1 function leads to functional recovery
and increased integrity of axons in MOG-induced
1 mm 1 mm 1 mm
EAE rats. (a) Local delivery of anti-LINGO-1 or
IgG control antibody, or no treatment, before the
onset of clinical EAE symptoms. P o 0.05.
Toluidine blue

(b) Two-week systemic treatment with

anti-LINGO-1 or IgG control after the onset of
clinical EAE. P o 0.05. Arrows indicate dosing
regimen. (c) DTI images (top) from normal (left),
1 mm 1 mm 1 mm IgG control–treated (center) and anti-LINGO-
1–treated (right) rats matched with toluidine
IgG control Anti-LINGO-1 blue–stained semi-thin spinal cord transverse
d e f sections (bottom, from left to right, respectively).
1.0 1.0 (d) Projection DTI images of lower thoracic and
lumbar spinal cords, showing a weak DTI signal
Fractional anisotropy

Fractional anisotropy

0.9 0.9 in the dorsal column of the lumbar region from

IgG control–treated rats (arrows, left) and an
0.8 * 0.8 increased DTI signal in anti-LINGO-1–treated rats
DTI images

* (arrows, right). (e) Fractional anisotropy values

0.7 0.7 from the dorsal spinal cord areas of rats treated
by local delivery of antibody before the clinical
0.6 0.6 onset of EAE symptoms (n ¼ 7 in each group).
-1

-1
l

*P ¼ 0.0034. (f) Fractional anisotropy values

ro

ro
O

O
nt

nt
G

G
co

co
N

from the dorsal spinal cord areas of spinal cords

LI

LI
G

G
ti-

ti-
Ig

Ig

from rats treated by systemic delivery of antibody

An

An

after the clinical onset of EAE symptoms (n ¼ 7

1.5 mm 1.5 mm
in each group). *P ¼ 0.0083.

1230 VOLUME 13 [ NUMBER 10 [ OCTOBER 2007 NATURE MEDICINE

 ARTICLES

a Normal IgG control Anti-LINGO-1 Figure 3 Histochemical detection of

remyelination after LINGO-1 antibody treatment.
(a) Paraffin spinal cord sections from normal
(left), IgG control–treated (center) and
anti-LINGO-1–treated (right) animals were
stained with LFB. Blue areas indicate
myelination, whereas pale areas indicate
demyelination (arrows indicate the dorsal column
1 mm 1 mm 1 mm of spinal cord). (b,c) Toluidine blue staining of
1-mm semi-thin sections to visualize myelination
b in normal (left), IgG control–treated EAE (center)
© 2007 Nature Publishing Group http://www.nature.com/naturemedicine

or anti-LINGO-1–treated EAE (right) animals.

(d,e) Anti-LINGO-1 treatment significantly
reduced the demyelinated areas (*P o 0.05)
and increased the amount of myelinated axons
(*P o 0.002) compared to the results of IgG
control treatment.

c
re-form myelin sheaths in demyelinated white
matter regions (Fig. 4a,b, left). In contrast,
after anti-LINGO-1 treatment, oligodendro-
cytes were in contact with surrounding axons
and were able to form myelin sheaths
(Fig. 4a,b, right). Notably, we observed
substantially more oligodentrocyte remyelina-
d 10.0
e 2,000
tion in the area closer to the site of antibody
Number of myelinated axons
Average demyelinated area
of lumbar spinal cord (%)

delivery than in more distant regions, which

7.5 1,500
further suggests that this remyelination is
* specifically induced by anti-LINGO-1 (data
5.0 1,000
not shown). The presence of thinly myeli-
nated axons in demyelinated areas provides
2.5 500
* strong evidence that anti-LINGO-1 treatment
promotes axon remyelination. Indeed, dif-
0 0
ferent remyelination phases were observed
-1

-1
l

l
al
ro

ro
O

O
m
nt

nt

in which axons were undergoing initial

G

G
or
co

co
N

N
N
LI

LI
G

G
ti-

ti-

wrapping, partial remyelination and almost

Ig

Ig
An

An

complete remyelination (Fig. 4c, top) in

anti-LINGO-1–treated animals. We observed
with real morphological changes in the myelination of the affected a morphological structure that is typical of newly formed myelination,
lumbar spinal cord. Luxol fast blue (LFB) staining was used to in which the outer (Co) and inner (Ci) ends of the spiraling
visualize myelinated regions (Fig. 3). In transverse paraffin spinal cytoplasmic processes of oligodendrocytes (Fig. 4c, bottom left) are
cord sections from normal rats (Fig. 3a, left), blue areas indicate wrapped around axons (Fig. 4c, bottom left and right) in anti-
intact myelin, whereas pale areas indicate demyelination. Poor LFB LINGO-1–treated animals. Similarly, intraperitoneal delivery of anti-
staining, or extensive myelin loss, was clearly evident in the control- LINGO-1 also resulted in a considerable increase in the amount thinly
treated animals (Fig. 3a, center). In contrast, increased LFB staining remyelinated axons (data not shown).
was seen with anti-LINGO-1 treatment (Fig. 3a, right). Demyeli- The clinical manifestations of MOG-induced EAE, which mirror
nation in the lumbar spinal cord of the anti-LINGO-1–treated multiple sclerosis, result from an autoimmune inflammatory compo-
EAE rat group was five times lower than in the control group nent and a neurological component that involves demyelination and
(P o 0.05, Fig. 3d). axon loss16. New approaches that focus on the neurological compo-
Toluidine blue–stained sections were used to visualize myelinated nents of myelination and axon survival are beginning to be
rat axons by light microscopy (Fig. 3b,c). Loss of myelin was apparent explored12,17,18. In the present study, we demonstrate that diminished
in the control-treated axons (Fig. 3b,c, center), whereas more myeli- LINGO-1 function in mice and rats (caused by Lingo1 knockout or
nated and/or remyelinated axons were found in anti-LINGO- treatment with LINGO-1 antagonist antibody) alleviates the symp-
1–treated rats (Fig. 3c, right). These newly formed myelin sheaths toms of clinical neuropathology associated with MOG-induced EAE.
were thinner than those in the normal control (Fig. 3c, left). Aside from the lower EAE scores observed in these mice and rats,
Quantitative analysis of myelinated axons revealed significantly fewer physiological improvements in axonal integrity were revealed by
myelinated axons in the isotype antibody control–treated group than DTI, and at the cellular level, enhanced and new myelination were
in the anti-LINGO-1–treated group (P o 0.002, Fig. 3e). revealed by histological staining and electron microscopy, respectively.
Electron microscopy was used to demonstrate increased axon The effect of LINGO-1 antagonism appears to be CNS specific, as
remyelination in anti-LINGO-1–treated rats as compared to control- LINGO-1 deficiency did not affect the induction phase of EAE, as
treated rats. After IgG control treatment, oligodendrocytes were shown by T-cell proliferation assays and by the release of T-cell
clearly in physical contact with naked axons but were unable to cytokines after MOG immunization. Also, EAE can be induced by

NATURE MEDICINE VOLUME 13 [ NUMBER 10 [ OCTOBER 2007 1231

 ARTICLES

IgG control Anti-LINGO-1 Figure 4 Electron microscopic visualization of remyelination after

a anti-LINGO-1 treatment. (a) Low-magnification electron micrographs
show borders between demyelinated and myelinated areas in control-
treated (left) and anti-LINGO-1–treated (right) rats. (b) High-magnification
electron micrographs of demyelinated areas showing naked axons (asterisks)
from IgG control–treated rats (left) and remyelinated axons (asterisks) in
anti-LINGO-1–treated rats (right). (c) Top, electron micrograph of different
phases of remyelination after anti-LINGO-1 treatment, with axons
undergoing initial wrapping (AX1), partial remyelination (AX2) and
b almost complete remyelination (AX3 and AX4). Bottom, typical
morphological structures of newly formed myelination are seen where the
© 2007 Nature Publishing Group http://www.nature.com/naturemedicine

outer (Co) and inner (Ci) ends of the spiraling cytoplasmic processes of
oligodendrocytes (bottom left) are wrapped around axons in anti-LINGO-
1–treated animals. AX, axon.

Supplementary Methods. Cell surface–binding ELISA and competitive

ELISA experiments demonstrated the specific binding to LINGO-1 (Sup-
Anti-LINGO-1
plementary Fig. 1).
c
Clinical evaluation of MOG-EAE and selection of animals. After
MOG induction of EAE, each animal was assessed by a behavioral test
based on motor functions, and an EAE score was obtained by methods
described previously3,19. Detailed steps for the evaluation are in the Supple-
mentary Methods.

T-cell function and cytokine secretion assays. We isolated cells from the
draining lymph nodes of MOG-immunized mice. Five animals were used in
each group. Cells were cultured in triplicate in the presence of MOG for 3 d. To
measure T-cell proliferation responses, we added 1 mCi 3H-thymidine to each
well for the last 18 h. Cells were harvested onto filter mats using the Tomtec
Harvester 96, and 3H-thymidine incorporation was quantified using the Wallac
1450 Microbeta Jet counter. Duplicate cultures were set up to measure the level
of cytokine secretion in the presence of 50 mg MOG. Supernatants collected
72 h after the start of the cultures were assayed for IL-2, IL-4, IL-5, IL-6, IL-10,
IL-13, IL-17, TNFa, IFNg and GM-CSF abundance using the Pierce Search
Light multiplex ELISA system.

LINGO-1 antibody treatment. For local delivery, 3 d after MOG-induction, we

the adoptive transfer of encephalogenic T cells from MOG-immunized introduced isotype control antibody or anti-LINGO-1 intrathecally into rats
Lingo1-knockout mice to WT mice, and EAE is mitigated when with a minipump (Alzet mini–osmotic pump, Alza Corporation). Each
encephalogenic T cells from MOG-immunized WT mice are trans- minipump held 2.6 mg anti-LINGO-1 and delivered 185 mg/d for 2 weeks,
and the minipump was replaced every 2 weeks. The isotype control antibody
ferred to Lingo1-knockout mice. LINGO-1 inhibition’s nonimmune
was delivered at the same concentration and by the same method used for anti-
role in the attenuation of EAE is also supported by other studies, in
LINGO-1.
which it has been shown that LINGO-1 antagonist antibody promotes For systemic drug delivery, we intraperitoneally injected 8 mg/kg antibody
remyelination in two noninflammatory models of demyelination, the twice per week (either anti-LINGO-1 or isotype control antibody), commen-
cuprizone and lysolecithin models (S.M., unpublished data). In cing when an EAE score of 1.0 was reached. The rats were killed after 2 weeks
demyelinated CNS, diminished LINGO-1 function is therefore likely of treatment.
to enhance remyelination by promoting differentiation of OPCs into
mature myelin-forming oligodendrocytes. Together, these data have Tissue preparation for DTI, electron microscopy and histology. Animals
provided an in vivo proof-of-concept for the development of LINGO- were perfused with a fixative 30 and 36 d after MOG-EAE induction for mice
and rats, respectively, by the method described previously20. Fixed tissues were
1 antagonists as a novel therapeutic approach for the treatment of
prepared for DTI (see below), immunocytochemistry and electron microscopy
demyelinating diseases. by methods described previously16,20,21.
METHODS Quantification of demyelination area and axon preservation. Areas of
Induction of MOG-EAE model. We induced EAE in adult rats by previously
demyelination were estimated on complete transverse LFB–stained spinal cord
described methods3. Detailed procedures are provided in the Supplementary sections (lumbar 4 segments, L4). Blue areas indicate intact myelin, whereas
Methods online. All procedures were carried out according to the US National pale areas indicate demyelination (Supplementary Fig. 4a online). ImageJ
Institutes of Health Guide for the Care and Use of Laboratory Animals and
software (free software from the US National Institutes of Health) was used
were approved by the Committee on the Use of Live Animals in Teaching and
to manually trace the total cross-sectional area (yellow trace, Supplementary
Research of the University of Hong Kong and by the Biogen Idec Institutional Fig. 4b) and the demyelination area (red trace, Supplementary Fig. 4b) of each
Animal Care and Use Committee. The presence of immune cells and inflam- section. Demyelination was expressed as a percentage of the total demyelination
matory infiltrates (T cells, macrophages, microglia and others) within the CNS
area over the total spinal cord area, using random sections from the fourth
tissue of EAE rats was detected by immunocytochemistry (Supplementary
lumbar segment of each animal.
Fig. 3 online). To quantify axons, we adopted a line-sampling method similar to that
Generating LINGO-1 antibody. We generated anti-LINGO-1 in mice by described previously16. Axons intercepted by nine fixed sampling lines (red
methods described previously19. Detailed procedures are provided in the lines, Supplementary Fig. 5 online) were sampled from toluidine blue–stained

1232 VOLUME 13 [ NUMBER 10 [ OCTOBER 2007 NATURE MEDICINE

 ARTICLES

1-mm-thick sections of the L4 spinal segment of each rat. Nine tissue strips 6. Mi, S. et al. LINGO-1 negatively regulates myelination by oligodendrocytes. Nat.
extending from the gray matter to the pial surface were selected under the
100 Neurosci. 8, 745–751 (2005).
7. Lee, X. et al. NGF regulates the expression of axonal LINGO-1 to inhibit oligoden-
oil-immersion objective of a Zeiss microscope. All myelinated axons whose drocyte differentiation and myelination. J. Neurosci. 27, 220–225 (2007).
axoplasms were intercepted by a sampling line were counted. Sampled axonal 8. Wekerle, H., Kojima, K., Lannes-Vieira, J., Lassmann, H. & Linington, C. Animal
profiles from the tissue strips of each rat were pooled. Because of the resolution models. Ann. Neurol. 36 Suppl, S47–S53 (1994).
limits of light microscopy, unmyelinated and myelinated fibers of cross- 9. Kerschensteiner, M. et al. Remodeling of axonal connections contributes to recovery in
an animal model of multiple sclerosis. J. Exp. Med. 200, 1027–1038 (2004).
sectional areas smaller than 0.1 mm2 were not included in the counts. 10. Reynolds, R. et al. The response of NG-2 expressing oligodendrocyte progenitors to
demyelination in MOG-EAE and MS. J. Neurocytol. 31, 523–536 (2002).
Diffusion tensor magnetic resonance imaging. All magnetic resonance 11. Ben-Hur, T. et al. Transplanted multipotential neural precursor cells migrate into the
imaging experiments were performed on a high-field small-bore 7 Tesla inflamed white matter in response to experimental autoimmune encephalomyelitis.
rodent magnetic resonance imaging scanner (70/16 PharmaScan, Bruker Glia 41, 73–80 (2003).
© 2007 Nature Publishing Group http://www.nature.com/naturemedicine

Biospin GmbH) at the University of Hong Kong. DTI was used for 12. Lubetzki, C., Williams, A. & Stankoff, B. Promoting repair in multiple sclerosis:
problems and prospects. Curr. Opin. Neurol. 18, 237–244 (2005).
semi-quantification of the demyelination area in the spinal cord of EAE 13. Ludwin, S.K. & Johnson, E. Evidence for a ‘‘dying-back’’ gliopathy in demyelinating
rats as described previously22–24. Detailed procedures are provided in the disease. Ann. Neurol. 9, 301–305 (1981).
Supplementary Methods. 14. Lassmann, H., Bartsch, U., Montag, D. & Schachner, M. Dying-back oligodendroglio-
pathy: a late sequel of myelin-associated glycoprotein deficiency. Glia 19, 104–110
Statistical analysis. GraphPad Prism (version 4.0, GraphPad) software was (1997).
15. Trapp, B.D. et al. Axonal transection in the lesions of multiple sclerosis. N. Engl. J.
used for statistical analysis. Data are shown as means ± s.e.m. Statistical
Med. 338, 278–285 (1998).
significance in comparing two means was tested with the unpaired Student’s 16. Papadopoulos, D., Pham-Dinh, D. & Reynolds, R. Axon loss is responsible for chronic
t-test (with Welch’s correction or Mann-Whitney modification). One-way neurological deficit following inflammatory demyelination in the rat. Exp. Neurol. 197,
analysis of variance was applied to test multiple means. P o 0.05 was set as 373–385 (2006).
17. Kornek, B. et al. Multiple sclerosis and chronic autoimmune encephalomyelitis: a
the cutoff for statistical significance.
comparative quantitative study of axonal injury in active, inactive, and remyelinated
lesions. Am. J. Pathol. 157, 267–276 (2000).
Note: Supplementary information is available on the Nature Medicine website. 18. Zhao, C., Fancy, S.P.J., Kotter, M.R., Li, W.W. & Franklin, R.J.M. Mechanisms of CNS
remyelination—the key to therapeutic advances. J. Neurol. Sci. 233, 87–91 (2005).
Published online at http://www.nature.com/naturemedicine 19. Eager, K.B. & Kennett, R.H. The use of conventional antisera in the production of
Reprints and permissions information is available online at http://npg.nature.com/ specific monoclonal antibodies. J. Immunol. Methods 64, 157–164 (1983).
reprintsandpermissions 20. Wu, W., Scott, D.E. & Reiter, R.J. Transplantation of mammalian pineal gland: studies
of survival, revascularization, reinnervation, and recovery of function. Exp. Neurol. 122,
88–99 (1993).
1. Trapp, B.D., Ransohoff, R.M., Fisher, E. & Rudick, R. Neurodegeneration in 21. Li, L. et al. Rescue of adult mouse motoneurons from injury-induced cell death by a
multiple sclerosis: relationship to neurological disability. Neuroscientist 5, 48–57 glial cell line-derived neurotrophic factor (GDNF). Proc. Natl. Acad. Sci. USA 92,
(1999). 9771–9775 (1995).
2. Noseworthy, J.H., Gold, R. & Hartung, H.P. Treatment of multiple sclerosis: recent trials 22. Hasan, K.M., Gupta, R.K., Santos, R.M., Wolinsky, J.S. & Narayana, P.A. Comparison
and future perspectives. Curr. Opin. Neurol. 12, 279–293 (1999). of gradient encoding schemes for diffusion-tensor MRI. J. Magn. Reson. Imaging 13,
3. Diem, R. et al. Combined therapy with methylprednisolone and erythropoietin in a 769–780 (2001).
model of multiple sclerosis. Brain 128, 375–385 (2005). 23. Basser, P.J. & Pierpaoli, C. Microstructural and physiological features of tissues
4. DuboisDalcq, M., ffrench-Constant, C. & Franklin, R.J.M. Enhancing central nervous elucidated by quantitative-diffusion-tensor MRI. J. Magn. Reson. B. 111, 209–219
system remyelination in multiple sclerosis. Neuron 48, 9–12 (2005). (1996).
5. Mi, S. et al. LINGO-1 is a component of the Nogo-66 receptor/p75 signaling complex. 24. Kim, J.H. et al. Detecting axon damage in spinal cord from a mouse model of multiple
Nat. Neurosci. 7, 221–228 (2004). sclerosis. Neurobiol. Dis. 21, 626–632 (2006).

NATURE MEDICINE VOLUME 13 [ NUMBER 10 [ OCTOBER 2007 1233