The Journal of Neuroscience, June 13, 2012 • 32(24):8317– 8330 • 8317
Neurobiology of Disease
Targeted Ablation of Oligodendrocytes Induces Axonal
Pathology Independent of Overt Demyelination
Laura-Jane Oluich,1,2 Jo Anne S. Stratton,1,2 Yao Lulu Xing,1,2 Sze Woei Ng,1 Holly S. Cate,1,2 Pankaj Sah,3
François Windels,3 Trevor J. Kilpatrick,1,2 and Tobias D. Merson1,2
Florey Neuroscience Institutes, University of Melbourne, Parkville, Victoria 3010, Australia, 2Centre for Neuroscience Research, University of Melbourne,
Parkville, Victoria 3010, Australia, and 3Queensland Brain Institute, The University of Queensland, Brisbane, Queensland, 4072, Australia
1
The critical role of oligodendrocytes in producing and maintaining myelin that supports rapid axonal conduction in CNS neurons is well
established. More recently, additional roles for oligodendrocytes have been posited, including provision of trophic factors and metabolic
support for neurons. To investigate the functional consequences of oligodendrocyte loss, we have generated a transgenic mouse model of
conditional oligodendrocyte ablation. In this model, oligodendrocytes are rendered selectively sensitive to exogenously administered
diphtheria toxin (DT) by targeted expression of the diphtheria toxin receptor in oligodendrocytes. Administration of DT resulted in
severe clinical dysfunction with an ascending spastic paralysis ultimately resulting in fatal respiratory impairment within 22 d of DT
challenge. Pathologically, at this time point, mice exhibited a loss of ⬃26% of oligodendrocyte cell bodies throughout the CNS. Oligodendrocyte cell-body loss was associated with moderate microglial activation, but no widespread myelin degradation. These changes were
accompanied with acute axonal injury as characterized by structural and biochemical alterations at nodes of Ranvier and reduced
somatosensory-evoked potentials. In summary, we have shown that a death signal initiated within oligodendrocytes results in subcellular
changes and loss of key symbiotic interactions between the oligodendrocyte and the axons it ensheaths. This produces profound functional consequences that occur before the removal of the myelin membrane, i.e., in the absence of demyelination. These findings have
clear implications for the understanding of the pathogenesis of diseases of the CNS such as multiple sclerosis in which the oligodendrocyte is potentially targeted.
Introduction
The oligodendrocyte is a specialized glial cell within the CNS that
is responsible for producing the myelin that ensheaths axons and
forms the nodes of Ranvier that permit saltatory nerve conduction. Dysfunction and death of oligodendrocytes has been described in various demyelinating pathologies including the
leukodystrophies and multiple sclerosis (MS) (Lucchinetti et
al., 1996; Dowling et al., 1999; Feigenbaum et al., 2000).
Autoimmune-induced inflammatory demyelination has traditionally been the assumed pathogenic mechanism in MS
(Weiner, 2004; Frohman et al., 2006). However, recent detailed
examinations of pathological specimens have challenged this
view. On the one hand, significant axonal degeneration occurs
early in disease (Trapp et al., 1998; 1999; Kuhlmann et al., 2002),
Received March 2, 2012; revised March 22, 2012; accepted March 28, 2012.
Author contributions: H.S.C., P.S., T.J.K., and T.D.M. designed research; L.-J.O., J.A.S.S., Y.L.X., S.W.N., F.W., and
T.D.M. performed research; L.-J.O. and T.D.M. analyzed data; T.J.K. and T.D.M. wrote the paper.
This work was supported by the Australian Postgraduate Award and Rotary Health Scholarship (L.J.O), the
National Health and Medical Research Council of Australia (NHMRC), Multiple Sclerosis Research Australia (MSRA)
Betty Cuthbert Fellowship (T.D.M.), NHMRC and MSRA Project Grants (T.D.M.), and an NHMRC Program Grant (T.J.K.
and P.S.). Anna Friedhuber and Dennis Kemper provided technical assistance with electron microscopy. Charityworks for MS and the Victorian Operational Infrastructure Support Program provided additional support.
The authors declare no competing financial interests.
Correspondence should be addressed to either Tobias D. Merson or Trevor J. Kilpatrick, Multiple Sclerosis
Division, Kenneth Myer Building, University of Melbourne, Parkville, Victoria 3010, Australia. E-mail:
tmerson@unimelb.edu.au or tkilpat@unimelb.edu.au.
DOI:10.1523/JNEUROSCI.1053-12.2012
Copyright © 2012 the authors 0270-6474/12/328317-14$15.00/0
even in the absence of demyelination (Nikić et al., 2011), raising
the possibility of axonal attack by activated microglia and/or
macrophages. On the other hand, oligodendrocyte death in acute
MS can occur without significant immune cell infiltration and may
precede the development of inflammatory demyelinated plaques
(Barnett and Prineas, 2004; Marik et al., 2007; Henderson et al.,
2009). Moreover, periplaque tissue in evolving MS lesions exhibits
oxidative damage that is often localized within apoptotic oligodendrocytes (Haider et al., 2011). Thus, oligodendrocytopathy could
represent the primary event in newly forming MS lesions.
Experimental models of primary oligodendrocytopathy could
assist in unraveling the complex pathological sequelae in MS. To
this end, specific oligodendrocyte-targeting models have recently
been developed using diphtheria toxin (DT) as the driver of
oligodendrocyte death (Buch et al., 2005; Traka et al., 2010;
Pohl et al., 2011). These binary genetic approaches involve
oligodendrocyte-restricted Cre or CreER-regulated expression
of the simian DT receptor (DTR) or the A subunit of DT (DT-A).
In the former scenario, expression of DTR in oligodendrocytes
renders naturally DT-resistant mouse cells susceptible to intraperitoneally injected DT. Once internalized, DT catalyzes ADPribosylation of eukaryotic elongation factor 2, which results in
inhibition of protein translation and ultimately, apoptosis
(Collier, 2001). Induction of oligodendrocyte apoptosis in these
mice causes demyelination, myelin splitting, and/or vacuolization resulting in motor pathologies of variable severity (Buch et
al., 2005; Traka et al., 2010; Pohl et al., 2011).
8318 • J. Neurosci., June 13, 2012 • 32(24):8317– 8330
Oluich et al.• Axonal Pathology Independent of Demyelination following Oligodendrocyte Apoptosis
We have generated a novel transgenic mouse denoted MBPDTR (Line 100A) for targeted ablation of oligodendrocytes that
does not rely upon Cre/loxP transgenics. In these mice, expression of human DTR is regulated by the proximal 1.96 kb promoter of mouse myelin basic protein (MBP) gene which directs
specific expression to mature oligodendrocytes (Gow et al., 1992;
Shults et al., 2005; Emery et al., 2006). Administration of DT to
MBP-DTR mice leads to loss of oligodendrocyte cell bodies and
the development of severe clinical pathology characterized by
spastic paralysis of the hindlimbs, tremor, ataxia, and kyphosis
that correlates with impairment of electrophysiological properties of neurones and ultrastructural alterations at the nodes of
Ranvier. Importantly, these changes occurred in the absence of
inflammatory demyelination and demonstrate that axonal pathology occurring in the absence of demyelination can still be
secondary to primary pathology of the oligodendrocyte.
R26R-YFP⫹/⫺:MBP-DTR⫹/⫺ and Pdgfra-CreERT2 ⫹/⫺:R26R-YFP ⫹/⫺:
wild-type littermates. Pdgfra-CreERT2 mice (Rivers et al., 2008) and R26RYFP mice (Srinivas et al., 2001) were kindly provided by Professor William
D. Richardson (Wolfson Institute for Biomedical Research, University College of London) and Professor Frank Costantini (Columbia University Medical Center, New York), respectively. Experimental procedures were
conducted in accordance with National Health and Medical Research Council guidelines and were approved by the animal ethics committees of the
Florey Neuroscience Institutes and University of Queensland.
Intraperitoneal injection and gavaging. For the initial experimental cohorts, mice aged 8 –10 weeks were given a single intraperitoneal injection
of 200 ng DT (2 g/ml in saline; Sigma) or an equal volume (100 l) of
saline as control. For subsequent cohorts, DT was administered as a
single intraperitoneal injection at a fixed dose of 10 g/kg (prepared at 1
g/ml in saline); vehicle controls received saline at a dose of 10 l/g.
Three experimental groups were typically assessed: MBP-DTR mice administered DT, MBP-DTR mice administered saline, and wild-type littermates administered DT. To induce recombination in Pdgfra-CreERT2:
R26R-YFP:MBP-DTR and Pdgfra-CreERT2:R26R-YFP:wild-type mice,
tamoxifen (prepared at 40 mg/ml in 50°C corn oil; Sigma) was administered by oral gavage (300 mg/kg/d) for 4 consecutive days, commencing
7 d before DT challenge. Following DT or saline challenge, mice were
weighed and monitored daily for signs of neurological impairment. The
clinical course was scored as follows: 0 ⫽ healthy, 1⫽ pelvis lowered, 2 ⫽
pelvis on ground, 3 ⫽ hindpaw slippage, 4 ⫽ kyphosis, 5 ⫽ tail spasticity,
6 ⫽ upper torso lowered, and 7 ⫽ labored breathing.
Immunohistochemistry. Mice were deeply anesthetized with 100 mg/kg
sodium pentobarbitone then transcardially perfused with PBS followed
by 4% paraformaldehyde (PFA)/PBS. Extracted brains, spinal cords, and
sciatic nerves were postfixed in 4% PFA/PBS for 2 h on ice, rinsed briefly
in PBS, and then equilibrated in 20% sucrose/PBS for 24 h before embedding in Tissue-Tek O.C.T. compound (Sakura FineTek) and freezing
on an isopentane bath over dry ice. Tissue was stored at ⫺80°C until
sectioned. Ten-micron-thick cryosections of the brain (coronal), spinal
cord (coronal and longitudinal), and sciatic nerve (longitudinal) were
collected onto Superfrost Plus slides (Menzel Gläser) and air-dried for
1 h before storing at ⫺80°C until stained.
Cryosections were blocked for 1 h with PBS containing 0.3% Triton
X-100 and 10% normal serum (goat or donkey depending upon host of
secondary antibodies). In some instances serum was replaced with 1%
BSA and 0.1% gelatin. Primary antibodies diluted in 10% normal serum/
PBS were applied for 16 h at 4°C or at room temperature. Slides were
rinsed three times in PBS (5 min each) before application of secondary
antibodies for 30 min at room temperature. Slides were rinsed again,
counterstained with Hoechst 33342 (1 g/ml; Invitrogen), and coverslipped with Mowiol mounting medium. For DTR immunohistochemistry, an Alexa Fluor 488-conjugated goat anti-FITC antibody (1:200;
Invitrogen) was used as a tertiary antibody after the secondary antibody
rinse to amplify the DTR signal.
The following primary antibodies were used: rabbit anti-panneurofascin (1:2000; a generous gift from Professor Peter Brophy, University of Edinburgh), mouse anti-APC/CC-1 (1:100; Calbiochem),
mouse anti-ankyrin G (1:200; Santa Cruz Biotechnology), mouse antioligodendrocytes (1:5000, clone NS-1/RIP; Millipore), rabbit anti-amyloid precursor protein (1:300; Zymed), rat anti-PDGFR␣ (1:500,
clone APA5; BD PharMingen), rabbit anti-NG2 (1:200; Millipore), rabbit anti-cleaved caspase-3 (1:200; Cell Signaling Technology), rabbit
anti-CD3 (1:200; Dako), rabbit anti-Krox20 (1:200; Covance), rat antiCD11b (1:200; BD PharMingen), mouse anti-glial fibrillary acidic protein (GFAP) (1:200, clone GA5; Millipore), rabbit anti-GFAP (1:500;
Dako), goat anti-HB-EGF (1:200; R&D Systems), rabbit anti-Iba1 (1:
1000; Wako Industries), mouse anti-myelin-associated glycoprotein
(MAG) (1:1000; Millipore), rabbit anti-MBP (1:200; Millipore Bioscience Research Reagents), mouse anti-NeuN (1:100; Millipore Bioscience
Research Reagents), mouse anti-S100 (1:1000, clone SH-B1; Sigma),
rabbit anti-Olig2 (1:200; Millipore), mouse anti-Kv1.2 (1:500; NeuroMab), rabbit anti-Nav1.6 (1:100; Alomone Labs), and mouse antiSMI312 (1:1000; Covance). Secondary antibodies raised in goat or
Materials and Methods
Generation of MBP-DTR transgenic mice. The MBP-DTR construct was
generated by excising a LacZ transgene from the pMG2 vector (kindly
provided by Associate Professor Alexander Gow, Wayne State University, Detroit, Michigan; Gow et al., 1992) and replacing it with the coding
sequence (CDS) for human heparin-binding epidermal growth factor
(HB-EGF), a high-affinity DTR. The HB-EGF/DTR CDS was amplified
by high-fidelity PCR from the pRC/CMV/hHB-EGF plasmid (kindly
provided by Professor Eisuke Mekada, Osaka University, Japan). The
resultant MBP-DTR construct was validated by DNA sequencing to
comprise the 1.94 kb proximal promoter of the mouse myelin basic
protein promoter (MBP, corresponding to bases ⫺1907 to ⫹36 relative
to ATG), the 627 bp human HB-EGF/DTR CDS, and 1.43 kb 3⬘ splicing
and the polyadenylation sequence derived from the human -globin
gene (exon 2, intron 2, exon 3). MBP-DTR plasmid DNA isolated from
maxiprep culture (Qiagen) was further purified by CsCl gradient centrifugation. The DNA was digested with ScaI and SphI producing a 1.7 kb
fragment comprising the plasmid backbone and a 4.7 kb fragment comprising the proximal MBP promoter, human HB-EGF (DTR) CDS, and
human -globin polyadenylation sequence. The DNA fragments were
separated on a 1.5% agarose (Lonza SeaPlaque; Fisher Scientific)/TBE gel
and the 4.7 kb ScaI/SphI MBP-DTR fragment was extracted from the
corresponding gel slice by -Agarase I digestion (New England BioLabs).
The extracted DNA was dialyzed (“v” series membrane, 0.05 m pore
size; Millipore) with microinjection buffer and diluted to 2 g/ml before
microinjecting into the male pronuclei of fertilized C57BL/6 oocytes. Tail
biopsies from the offspring of oocyte recipients were assessed for incorporation of the transgene by Southern blot and PCR analysis. For Southern blot analysis, genomic DNA was digested with BstXI followed by
agarose gel electrophoresis. Gel blots prepared using nitrocellulose membranes (Hybond N⫹; GE Healthcare) were hybridized with a 32P-labeled
DNA probe generated by PCR amplification of a 629 bp product from
the MBP-DTR construct that encompassed part of the murine MBP
promoter and human DTR sequence (sense primer: 5⬘-CAGGCCCACATT
CATATC-3⬘; antisense primer: 5⬘-TCTTCCCTAGCCCCTTGCCTTTCT3⬘). PCR genotyping of tail biopsy DNA was performed using primers
targeting a 512 bp amplicon encompassing part of the vector between the
promoter and HB-EGF CDS and part of the human HB-EGF CDS with 80%
homology to murine HB-EGF (sense primer: 5⬘-GCTCGAATAATTCTAG
GGTC-3⬘; antisense primer: 5⬘-GGTCATAGGTATATAAGCGA-3⬘). Of
the 18 founders that exhibited germline transmission of the MBP-DTR
transgene, we describe the characterization of one line denoted MBP-DTR
Line 100A which is hereafter referred to as MBP-DTR mice.
Animal cohorts. All experimental mice were maintained on a pure
C57BL/6 background on a 12 h light-dark cycle and provided with rodent
chow (Barastoc Products) and water ad libitum. Experimental cohorts
were generated by crossing MBP-DTR heterozygotes with wild-type
C57BL/6 to produce heterozygous MBP-DTR and wild-type littermates or, alternatively, by crossing MBP-DTR heterozygotes with
Pdgfra-CreERT2⫹/⫹:R26R-YFP⫹/⫹ mice to generate Pdgfra-CreERT2⫹/⫺:
Oluich et al.• Axonal Pathology Independent of Demyelination following Oligodendrocyte Apoptosis
J. Neurosci., June 13, 2012 • 32(24):8317– 8330 • 8319
donkey and conjugated to FITC, TRITC, or NL557 were purchased from
Jackson ImmunoResearch or R&D Systems and used at 1:200 dilution.
Slides were examined using an AxioPlan2 fluorescent microscope
(Zeiss), and images were captured using an Axiocam HRc camera (Zeiss)
using Axiovision 7.2 software. For comparative analysis of the intensity
of myelin protein labeling, photomicrographs were converted to grayscale and analyzed using ImageJ software (National Institutes of Health;
NIH). Intensity was defined as the average gray value obtained from all
pixels within the region of interest above background. Confocal analysis
of nodal protein expression was performed using a FluoView FV1000
confocal microscope (Olympus) under a 40⫻ oil objective using FluoView confocal software (Olympus).
TUNEL. For combined immunohistochemistry and TUNEL, 10-mthick coronal cryosections of fresh frozen brain tissue were collected onto
Superfrost Plus slides, air-dried for 5 min, postfixed in 4% PFA/PBS for
20 min at room temperature, and rinsed three times in PBS. Immunohistochemistry was conducted as previously described using TRITCconjugated secondary antibodies then processed for TUNEL using the
Fluorescein In Situ Cell Death Detection Kit (Roche) according to the
manufacturer’s protocol. Slides were counterstained with Hoechst 33342
and coverslipped as described previously. To quantify the extent of apoptosis, the number of TUNEL ⫹ Hoechst ⫹ cells in the cerebral cortex and
corpus callosum were counted (bregma level ⫹0.5 mm) in MBP-DTR
and wild-type mice (n ⫽ 4 per genotype) at both 5 and 15 d following DT
challenge and expressed as TUNEL ⫹ cells per millimeter squared. For
statistical analysis, the data were analyzed by one-way ANOVA followed
by Bonferroni’s post hoc tests.
Electron microscopy. Following perfusion/fixation, the lumbar spinal
cord and sciatic nerves were postfixed in Karnovsky’s buffer (2.5% glutaraldehyde, 4% PFA, 0.1 M sodium cacodylate, pH 7.3) for 16 h at 4°C,
rinsed three times in PBS, and stored in 0.1 M sodium cacodylate, pH 7.3,
at 4°C for 24 h. The tissue was postfixed in a solution of 2% w/v osmium
tetraoxide and 1.5% w/v potassium ferricyanide and embedded in
Spurr’s resin. The dorsal funiculus of the lumbar spinal cord and sciatic
nerves were sectioned at 0.5 m thickness on an ultramicrotome (Ultracut S; Reichert) in transverse orientation for analysis of g-ratios (dorsal funiculus, sciatic nerve) or longitudinally (dorsal funiculus) for
assessment of nodes of Ranvier.
Semithin (0.5 m) transverse sections were stained with methylene
blue and imaged at 10,000⫻ magnification. Fibers within a 121.5 m 2
region of interest (ROI) within each dorsal funiculus (two ROI assessed
per animal, n ⫽ 2 animals per group) or a 3.07 mm 2 ROI within each
sciatic nerve (four ROI assessed per animal, n ⫽ 2 animals per group)
were assessed using ImageJ software (NIH). For each axon, the axonal
area excluding the myelin sheath and the entire fiber area including the
myelin sheath were independently outlined and the area calculated using
the “measurement” function within ImageJ. In addition, as a measure of
myelin integrity, the axonal and fiber diameters were measured and used
to calculate g-ratios (ratio of axonal diameter to fiber diameter) (Sherman and Brophy, 2005). Results were analyzed using multivariate
ANOVA followed by Tukey’s post hoc analysis using SPSS 15.0.
For nodal analysis, the ROI in the dorsal funiculus was confirmed by
methylene blue staining before collection of ultrathin (90 nm) sections.
Ultrathin sections were postfixed in lead nitrate and urinyl acetate,
mounted on copper grids, and examined in an Elmskop 102 transmission
electron microscope (Siemens). Photomicrographs were captured at
25,000⫻ magnification and negatives were scanned to generate highresolution computer images for analysis. Nodes of Ranvier were examined qualitatively for indications of ultrastructural abnormalities. The
percentage of nodes within the dorsal funiculi of DT-challenged MBPDTR (n ⫽ 2) and wild-type (n ⫽ 2) mice exhibiting paranodal loop
eversion was determined on the basis of assessments by two expert myelin biologists who were blinded to the identity of each group.
Somatosensory-evoked potentials. Six DT-challenged MBP-DTR mice
at clinical end point and five matching DT-challenged wild-type littermates were anesthetized with an intraperitoneal injection of ketamine
(0.1 mg/kg) and xylazine (0.01 mg/kg). Depth of anesthesia was checked
by the loss of response to rear paw pinch and additional injection of the
anesthetic solution was provided as required. The body temperature of
animals was maintained at 36.5°C using a controlled heating pad. Animals were mounted in a stereotaxic frame (model 962; Kopf Instruments) and the scalp incised longitudinally and retracted to expose the
skull. A hole was drilled above the somatosensory cortex (anteroposterior, ⫺0.5; lateral, 2.5; and dorsal, 1 mm) to allow insertion of a borosilicate glass recording electrode filled with 3 M saline. A second hole was
drilled on the opposite side above the contralateral frontal lobe to allow
for the insertion of the reference electrode between the skull and the
meninges. Stimulating electrodes made of stainless steel (25 g) were inserted at the sciatic notch and connected to a constant current generator
(A385; World Precision Instrument). Local field potentials were recorded at 5 kHz and bandpass filtered (30 Hz–1.8 kHz). Each animal
received 50 sciatic notch stimulations (5 mA; 0.2 ms; interspike interval,
15 s); the onset of the field potential immediately following the stimulation artifact was used to measure the nerve conduction velocity (values
are mean ⫾ SEM).
Measurement of compound action potentials in isolated sciatic nerve.
Eight DT-challenged MBP-DTR mice at clinical end point and three
matching DT-challenged wild-type littermates were overdosed with sodium pentobarbitone (100 mg/kg). Sciatic nerves were isolated and
placed in saline at room temperature. Nerves were coated in liquid paraffin oil, and the proximal and distal ends were placed onto gold-plated
fixed electrodes (proximal end, stimulating electrode; distal end, recording electrode 10 mm from proximal electrode; medial position, grounding electrode) in a humidified 37°C chamber. Square-wave stimulating
pulses of 100 s duration were delivered at 30 s intervals using a
constant-voltage stimulus isolator (PowerLab 4/20; ADInstruments) at
progressive intensities from 0.2 mV up to 10 V. Monophasic compound
action potentials (CAPs) generated at suprathreshold stimulus intensity
(6 V) were amplified using a differential bioamplifier (Bio Amp; Maclab)
and digitized using Scope software (v4.1; ADInstruments). Post hoc analysis was performed using LabChart Reader for Mac (ADInstruments).
Results
MBP-DTR mice express DTR exclusively in oligodendrocytes
in the CNS
To assess the consequences of oligodendrocyte ablation in the
adult CNS, we generated transgenic mice that express DTR under
the regulatory control of the proximal promoter for the MBP
gene (see Materials and Methods; Fig. 1A). We used a 1.94 kb
MBP promoter that had previously been shown to direct transgene expression specifically to the mature oligodendrocyte lineage (Gow et al., 1992; Wrabetz et al., 1998; Emery et al., 2006).
Eighteen germline-transmitting MBP-DTR transgenic founders
were identified by Southern blot (Fig. 1 B) and PCR analysis (Fig.
1C; see Materials and Methods). We performed a series of immunohistochemical analyses to assess the expression pattern of DTR
in the adult CNS of MBP-DTR and wild-type littermates. Immunolabeling was conducted using an antibody that specifically detects human HB-EGF/DTR, but not its murine homolog. Of the
MBP-DTR lines identified, five lines expressed DTR in 100% of
the oligodendrocytes. Here we characterize the phenotype of one
of these five transgenic lines denoted MBP-DTR Line 100A. Assessment of coronal sections of the adult forebrain in MBP-DTR
Line 100A mice, referred to hereafter as MBP-DTR, revealed that
DTR was highly expressed in white matter tracts (Fig. 1 D),
whereas only low-level, nonspecific labeling was observed in
wild-type littermates (Fig. 1 E). We confirmed that DTR was
expressed in mature oligodendrocytes in MBP-DTR mice by
multilabel immunohistochemistry. Specifically, all CC-1 ⫹ oligodendrocytes identified in sections from MBP-DTR mice were
found to coexpress DTR (Fig. 1 F), whereas no DTR was detected
in any CC-1 ⫹ oligodendrocytes in wild-type littermates (Fig.
1G). DTR expression in transgenic mice was found not only
within the cell bodies of CC-1 ⫹ oligodendrocytes but also within
the processes of oligodendrocytes identified by colocalization of
8320 • J. Neurosci., June 13, 2012 • 32(24):8317– 8330
Oluich et al.• Axonal Pathology Independent of Demyelination following Oligodendrocyte Apoptosis
DTR and MBP (Fig. 1 H–J ) as well as
of DTR and RIP (Fig. 1 K–M ). In contrast,
DTR was not expressed by any other neural cell population examined including
NeuN ⫹ neurons (Fig. 1N), PDGFR␣ ⫹ oligodendrocyte progenitors cells (Fig. 1O),
Iba1 ⫹ microglia (Fig. 1P), or GFAP ⫹ astrocytes (data not shown). Together, these
experiments demonstrate the successful
generation of MBP-DTR transgenic mice
that express DTR exclusively within the mature oligodendrocyte population of the
CNS, in a pattern that closely reflects the expression of myelin proteins.
Severe clinical and motor deficits in
MBP-DTR mice following DT challenge
To ablate oligodendrocytes, we administered a single 200 ng dose of DT to MBPDTR transgenic mice (n ⫽ 9) and wild-type
littermates (n ⫽ 8), or an equivalent volume of saline to MBP-DTR mice (n ⫽ 6).
The two control groups, wild-type mice
administered DT and MBP-DTR mice administered saline, did not shown any clinical changes over the subsequent 3 weeks.
In contrast, MBP-DTR mice injected with
DT exhibited progressive motor deficits
that typically began at 9 d after DT administration and became profound in the next
5– 6 d. The maximum survival time following the onset of clinical impairment
was 8 d. DT-challenged transgenic animals exhibited hindlimb clasping when
held by the tail, an ataxic gait, hindlimb
paralysis, tail spasticity, and thoracic kyphosis (Fig. 2A). The phenotype rapidly
progressed to include forelimb weakness,
respiratory impairment (labored breathing), reduced mobility, lethargy, and
weight loss, at which time disabled animals and their matching controls were
killed, a time point we have termed the
clinical end point. A plot of survival time
to clinical end point following DT challenge demonstrates that all MBP-DTR
4
Figure 1. Generation of MBP-DTR transgenic mice expressing DTR within the mature oligodendrocyte population of the adult
CNS. A, Schematic representation of the ScaI/SphI-linearized MBP-DTR construct used to generate MBP-DTR founders. The construct comprises the 1.94 kb proximal promoter of the mouse myelin basic protein promoter, the 627 bp human HB-EGF/DTR CDS,
and the 1.43 kb 3⬘ splicing and the polyadenylation sequence derived from the human -globin gene. Arrowheads depict the
sense and antisense primers used for PCR genotyping. The solid black line indicates the size and position of the 32P-labeled probe
used for Southern blot analysis of BstXI-digested mouse tail DNA. B, Southern blot analysis of BstXI-digested genomic DNA isolated
from mouse tail biopsies used to identify MBP-DTR (tg) founders; note the absence of signal at 1.7 kb in the wild-type (wt) sample.
Hybridization was performed using a 629 bp 32P-labeled DNA probe that identified a 1.7 kb BstXI fragment generated from the
MBP-DTR sequence. C, PCR genotyping using primers that amplify a 512 bp product from the MBP-DTR transgene enabled
discrimination of MBP-DTR (tg) from wild-type (wt) mice. D, E, Immunohistochemical assessment of DTR expression in the
forebrain using an anti-human HB-EGF antibody revealed DTR expression within white matter tracts of MBP-DTR (D) but not
wild-type (E) mice. F, G, Immunohistochemical assessment of
DTR (green) expression among CC-1 (red) immunoreactive oligodendrocytes in the corpus callosum of MBP-DTR (F) versus
wild-type (G) mice. Nuclei were stained with Hoechst 33258
(blue). H–J, DTR expression (H, green) in the corpus callosum
colocalized with MBP (I, red), as revealed in the merged image
(J). K–M, DTR expression (K, green) was localized to both the
cell body and processes of RIP-labeled oligodendrocytes (L,
red), as shown in the merged image (M). Nuclei were counterstained with Hoechst. N–P, DTR expression (green) was not
observed in NeuN-labeled (red) neuronal cell bodies (N), in
PDGFR␣-labeled (red) oligodendrocyte progenitor cells (O) or
in Iba1-labeled (red) microglia (P, confocal stack). CC, corpus
callosum, Ctx, neocortex; Str, striatum; z, z-axis projection.
Scale bars: F, G, K–P, 50 m; (in J) H–J, 200 m.
Oluich et al.• Axonal Pathology Independent of Demyelination following Oligodendrocyte Apoptosis
J. Neurosci., June 13, 2012 • 32(24):8317– 8330 • 8321
Figure 2. DT-challenged MBP-DTR mice exhibit progressive motor deficits and oligodendrocyte loss. A, MBP-DTR mice challenged with DT developed progressive clinical symptoms from day 9
postchallenge onward (see Materials and Methods for description of clinical scores). Data points represent the average score for all animals (n ⫽ 9 MBP-DTR mice, DT-injected; n ⫽ 6 MBP-DTR mice,
saline-injected; n ⫽ 8 WT mice, DT-injected) on each day. The apparent decrease in average disease severity at day 17 reflects the need to kill the most severely affected animals (n ⫽ 2) on day 16
(B), leaving animals with more moderate disease severity to influence average clinical score. No clinical changes were observed in wild-type controls injected with DT or in MBP-DTR controls injected
with saline. B, Plot of survival time to clinical end point after DT challenge in MBP-DTR mice. The clinical end point is defined as the time point at which animals were killed due to the severity of the
clinical phenotype. C, Latency of sciatic notch-evoked local field potentials measured in the somatosensory cortex was increased by 58.5 ⫾ 16% in symptomatic DT-challenged MBP-DTR mice
compared with DT-challenged wild-type mice ( p ⬍ 0.05, Student’s t test). D, Amplitude of sciatic notch-evoked local field potentials measured in the somatosensory cortex was reduced by 11.5 ⫾
2.39% in symptomatic DT-challenged MBP-DTR mice compared with DT-challenged wild-type mice ( p ⬍ 0.0016, Student’s t test). E, Density of TUNEL ⫹ cells in MBP-DTR versus wild-type controls
assessed 5 and 15 d after DT challenge (n ⫽ 4 per genotype per time point). Two-way ANOVA revealed an overall effect of genotype ( p ⬍ 0.0001), time ( p ⫽ 0.005), and an interaction between
these 2 factors ( p ⫽ 0.0003). Bonferroni’s post hoc: MBP-DTR ⫹ DT (day 15) versus wild-type ⫹ DT (day 15), p ⬍ 0.0001. F–N, Combined labeling for TUNEL and cell type-specific antigens revealed
that a subset of TUNEL ⫹ cells (F, I, L, green) in DT-challenged MBP-DTR mice expressed the oligodendroglial lineage marker Olig2 ⫹ (G, red) or the microglial/macrophage (Figure legend continues.)
8322 • J. Neurosci., June 13, 2012 • 32(24):8317– 8330
Oluich et al.• Axonal Pathology Independent of Demyelination following Oligodendrocyte Apoptosis
mice were killed between 12 and 22 d after DT injection (Fig. 2 B),
with a median survival time of 16 d.
drocytes were undergoing apoptosis following DT challenge. In
addition, some TUNEL ⫹ cells expressed CD11b (Fig. 2 I–K ) or
Iba1 (data not shown), consistent with a microglial/macrophage
identity. Notably, we never observed colocalization of TUNEL
with NeuN ⫹ neuronal nuclei (Fig. 2 L–N ), and TUNEL was
not identified in GFAP ⫹ astrocytes or NG2⫹ oligodendrocyte
progenitor cells (data not shown). Since DTR expression in
MBP-DTR mice is restricted to the mature oligodendrocyte
population, the presence of TUNEL ⫹ microglia likely reflects
an indirect response to the induction of DT-mediated oligodendrocyte apoptosis.
Disabled mice exhibit electrophysiological abnormalities
To directly assess whether symptomatic MBP-DTR mice challenged with DT exhibited electrophysiological abnormalities,
we performed in vivo recordings of somatosensory-evoked potentials (SSEPs) by stimulating at the sciatic notch and recording the response in the somatosensory cortex of both MBPDTR (n ⫽ 6) and wild-type animals (n ⫽ 5) challenged with
DT. In control animals the onset of the local field potential was
detected 8.11 ⫾ 0.29 ms after sciatic notch stimulation,
whereas in symptomatic MBP-DTR animals SSEPs were recorded 13.42 ⫾ 0.30 ms after stimulation, representing a
39.6% increase in latency (Fig. 2C). Statistical analysis demonstrated that transmission of the SSEP was significantly faster
in control versus symptomatic MBP-DTR animals (unpaired
Student’s t test; p ⫽ 0.037). We also observed a statistically
significant reduction in the peak amplitude of SSEPs in symptomatic DT-challenged MBP-DTR animals (31.98 ⫾ 0.86 mV)
compared with control animals (36.13 ⫾ 0.97 mV, p ⫽ 0.0016,
unpaired Student’s t test) (Fig. 2 D).
Induction of oligodendrocyte apoptosis in MBP-DTR mice
following DT challenge
Administration of DT would be expected to result in widespread
oligodendrocyte loss throughout the CNS of MBP-DTR transgenic mice. To confirm this, apoptosis within the brains of MBPDTR mice was assessed 5 and 15 d after administration of DT,
corresponding to time points before and during overt clinical
disease, respectively (n ⫽ 4 per time point). Wild-type animals
administered DT and assessed 5 and 15 d postinjection served as
controls (n ⫽ 4 per time point). At day 15, the number of
TUNEL ⫹ cells in the neocortex of MBP-DTR mice was increased
17.5 relative to wild-type controls (Fig. 2 E). Two-way ANOVA of
TUNEL ⫹ cell density identified a significant effect of both genotype ( p ⬍ 0.0001) and time ( p ⫽ 0.0005) and also a significant
interaction ( p ⫽ 0.0003) between these two variables. Bonferroni’s post hoc analysis confirmed that the increase in apoptotic cell
density in MBP-DTR mice compared with wild-type mice at 15 d
post-DT was highly significant ( p ⬍ 0.0001).
We next assessed the phenotype of TUNEL ⫹ cells and identified two distinct cell types by combined TUNEL and immunohistochemical analysis. Some TUNEL ⫹ nuclei coexpressed the
transcription factor Olig2 (Fig. 2 F–H ) suggesting that oligoden-
4
(Figure legend continued.) marker CD11b (J, red) but not the neuronal marker NeuN (M, red).
Merged images with Hoechst 33342 nuclear counterstain (blue) reveal a TUNEL ⫹ Olig2 ⫹ cell
(H) in the cerebral cortex, a TUNEL ⫹ CD11b ⫹ cell in the corpus callosum (K), and a TUNEL ⫹
NeuN ⫺ cell in the cortex (N) of an MBP-DTR mouse assessed 15 d post-DT. O, P, Representative
photomicrographs of immunolabeling against the oligodendrocyte marker CC-1 (green) in the
dorsal column (white outline) of wild-type ⫹ DT (O) and MBP-DTR ⫹ DT (P) mice. The dorsal
funiculus is outlined in yellow. Nuclei were counterstained with Hoechst 33342 (blue). Q, Density of YFP ⫹ cells that are immunoreactive for the oligodendrocyte precursor cell marker
(PDGFR␣) or mature oligodendrocyte marker (CC-1) in the white and gray matter of the lumbar
spinal cord of DT-challenged Pdgfra-CreERT2:R26R-YFP:MBP-DTR (n ⫽ 4) or DT-challenged
Pdgfra-CreERT2:R26R-YFP:wild-type (n ⫽ 3) mice. R, S, Density of CC-1 ⫹ oligodendrocyte somata at day 5 (R) and at clinical end point (S) in DT-challenged MBP-DTR mice compared with
controls. Nd, not determined. Error bars indicate mean ⫾ SE of the mean. Cb, cerebellum; CC,
corpus callosum; Ctx, cerebral cortex; Str, striatum; SpW, spinal cord white matter; SpG, spinal
cord gray matter; Mb, midbrain; Hb, hindbrain; Bs, brainstem; Df, dorsal funiculus of spinal cord
white matter. Scale bars: (in N) F–N, 20 m; (in P) O, P, 100 m.
Numbers of CC-1-positive oligodendrocytic cell bodies are
reduced in DT-challenged MBP-DTR mice
To assess the extent of oligodendrocyte ablation following DT
challenge in MBP-DTR mice, we quantified the density of CC-1 ⫹
oligodendrocytes in the forebrain, cerebellum, and spinal cord at
5 d post-DT injection, and in the forebrain, cerebellum, spinal
cord, midbrain, hindbrain, and brainstem at the clinical end
point. MBP-DTR mice injected with saline and wild-type mice
injected with DT were assessed at the same time points. Representative photomicrographs of CC-1 staining in the dorsal funiculus at the clinical end point are shown (Fig. 2O,P). At the day 5
time point, the number of oligodendrocytes per millimeter
squared was not different in any region of the CNS of DTchallenged MBP-DTR mice when compared with controls (Fig.
2 R; multivariate ANOVA, p ⫽ 0.915). On the other hand, there
was a significant reduction in the total number of CC-1 ⫹ oligodendrocytes in the CNS of the DT-challenged MBP-DTR mice at
the clinical end point, with an average reduction in cell density of
26 ⫾ 1.6% (multivariate ANOVA, p ⫽ 0.030). A region-specific
analysis indicated that the numbers of CC-1-positive somata
were significantly reduced in the neocortex (Tukey-HSD, p ⫽
0.006), striatum ( p ⬍ 0.001), cerebellar white matter ( p ⫽ 0.001),
spinal cord white matter ( p ⫽ 0.001), spinal cord gray matter
( p ⬍ 0.001), and hindbrain ( p ⫽ 0.05), but were not significantly
reduced in the corpus callosum ( p ⫽ 0.102), midbrain ( p ⫽
0.396), and brainstem ( p ⫽ 0.373; Fig. 2S). Interestingly, in the
dorsal funiculus of the spinal cord white matter which contains
the corticospinal tract there was a 39.5 ⫾ 5.2% loss of CC-1positive oligodendrocytes in comparison with DT-challenged
wild-type mice ( p ⫽ 0.05; Fig. 2S).
To assess whether newly differentiated oligodendrocyte
precursor cells (OPCs) contribute to the pool of mature oligodendrocytes at clinical end point, we used Pdgfra-CreERT2:R26R-YFP:
MBP-DTR transgenic mice and Pdgfra-CreERT2:R26R-YFP:wildtype littermates to map the fate of OPCs, which all express the
PDGFR␣ receptor, following DT-challenge. Under these conditions we have shown that 52.8% of all PDGFR␣ ⫹ cells express
YFP within 7 d from the first tamoxifen gavage. We found no
difference in the density of YFP ⫹ cells that were immunoreactive
for the oligodendrocyte precursor cell marker, PDGFR␣ ( p ⫽
0.910; Fig. 2Q), or mature oligodendrocyte marker, CC-1 ( p ⫽
0.124; Fig. 2Q), in the lumbar spinal cord of Pdgfra-CreERT2:
R26R-YFP:MBP-DTR (n ⫽ 4) or Pdgfra-CreERT2:R26R-YFP:
wild-type (n ⫽ 3) mice. These findings indicate that OPCs present
at the time of oligodendrocyte ablation do not appreciably differentiate into mature oligodendrocytes in compensation for the
mature oligodendrocytes that are lost following DT challenge in
the MBP-DTR mice.
Oluich et al.• Axonal Pathology Independent of Demyelination following Oligodendrocyte Apoptosis
J. Neurosci., June 13, 2012 • 32(24):8317– 8330 • 8323
Figure 3. Increased microglial density and GFAP immunoreactivity following oligodendrocyte apoptosis. A–C, Immunohistochemistry against the microglial/macrophage marker Iba1 in the
ventrolateral aspect of the spinal cord of MBP-DTR ⫹ saline (A), wild-type ⫹ DT (B), and MBP-DTR ⫹ DT (C) mice. MBP-DTR ⫹ DT mice were collected at clinical end point and matching controls
were examined at the same time point following saline or DT injection. Nuclei were identified with Hoechst (blue). Note the increased intensity of Iba1 staining in the section from a DT-challenged
MBP-DTR mouse (C). D, Density of Iba1 ⫹ cells in gray and white matter regions of the brain and spinal cord were increased in DT-challenged MBP-DTR mice at clinical end point. E–G, Multicellular
aggregates of Iba1 ⫹ cells (E, red) among CC-1 ⫹ oligodendrocytes (F, green) assessed 5 d after DT challenge in MBP-DTR mice. Note the close cell– cell contact between Iba1 ⫹ microglia and CC-1 ⫹
oligodendrocytes (G, arrowheads). H–J, Assessment of GFAP immunoreactivity (green) in the ventrolateral aspect of the spinal cord of MBP-DTR ⫹ saline (H), wild-type ⫹ DT (I), and MBP-DTR ⫹
DT (J) mice reveals evidence of reactive astrogliosis in MBP-DTR ⫹ DT mice (J) at clinical end point. Scale bars: (in C, J) A–C, H–J, 100 m; (in G) E–G, 25 m. Cb, cerebellum; CC, corpus callosum;
Ctx, cerebral cortex; GM, gray matter; Str, striatum; SpG, spinal cord gray matter; SpW, spinal cord gray matter; WM, white matter.
Reactive changes in microglia/macrophages and astrocytes in
disabled MBP-DTR mice
We next assessed whether oligodendrocyte loss induced an innate
immune response by examining Iba1 immunoreactivity at the
clinical end point in DT-challenged MBP-DTR mice and
matched controls (Fig. 3A–C). Morphological changes in the
Iba1-positive microglial cell population were seen in MBP-DTR;
DT-challenged mice that were not present in control mice. These
changes were characterized by shortened thickened processes and
increased Iba1 immunoreactivity; however, we did not observe
any microglia with the amoeboid morphology typically associated with phagocytic activity (Stence et al., 2001). We also found
that the average density of Iba ⫹ cells throughout the CNS of
DT-challenged MBP-DTR mice relative to DT-challenged wildtype mice was increased by 39 ⫾ 7% (Fig. 3D; multivariate
ANOVA, p ⫽ 0.015). Examination of specific regions of interest
revealed that the density of Iba1 ⫹ cells was significantly increased
in the neocortex (150% of control; p ⫽ 0.001), striatum (137%;
p ⬍ 0.001), corpus callosum (151%; p ⬍ 0.001), cerebellar white
matter (140%; p ⫽ 0.037), and spinal cord white (124%; p ⫽
0.035) and gray matter (134%; p ⫽ 0.009) (Fig. 3D).
We performed similar morphological and quantitative analyses of Iba1 ⫹ cells at the day 5 time point and observed an overall
35 ⫾ 10.6% increase in Iba1 ⫹ cells specifically in DT-challenged
MBP-DTR mice relative to controls; however, this increase was
not statistically significant (multivariate ANOVA, p ⫽ 0.737). We
also assessed the relationship between Iba1 ⫹ microglia and oli-
godendrocytes at this early time point by colabeling for Iba1 and
CC-1 (Fig. 3E–G). Microglia were frequently in close association
with CC-1 ⫹ somata. Occasional local aggregations of microglia
in contact with CC-1 ⫹ oligodendrocyte cell bodies were identified in oligodendrocyte-dense areas such as the corpus callosum
and cerebellar white matter (Fig. 3G, arrowheads).
We next assessed for evidence of reactive astrogliosis following
oligodendrocyte ablation by qualitative analysis of GFAP immunoreactivity in sections of the brain and spinal cord in DT-challenged
MBP-DTR mice at clinical end point and in matched controls.
Increased cellular GFAP immunoreactivity was evident in numerous regions, including the neocortex, striatum, and spinal
cord (Fig. 3H–J ); although in some regions, most notably the
corpus callosum, no such response was identified.
Absence of CD3 ⴙ T-cell infiltration into the CNS following
oligodendrocyte ablation
We next sought to assess whether the oligodendrocyte-initiated
injury in the MBP-DTR model was associated with the recruitment of peripheral T-cells into the CNS. We examined the brains
and spinal cords of DT-challenged MBP-DTR (n ⫽ 10), salineinjected MBP-DTR (n ⫽ 6), and DT-challenged wild-type controls (n ⫽ 9) by immunohistochemical analysis for the T-cell
antigen CD3. Animals were assessed both at 5 d following injection and at the clinical end point. CD3 ⫹ cells were rarely observed in any tissue samples examined. At both time points only
one CD3 ⫹ T-cell was identified for every two or three tissue
8324 • J. Neurosci., June 13, 2012 • 32(24):8317– 8330
Oluich et al.• Axonal Pathology Independent of Demyelination following Oligodendrocyte Apoptosis
Figure 4. Maintenance of normal levels of CNS myelin and neuronal preservation at clinical end point in symptomatic mice. A–C, Immunohistochemistry against the myelin marker RIP (CNPase)
on coronal forebrain sections of MBP-DTR ⫹ saline (A), wild-type ⫹ DT (B), and MBP-DTR ⫹ DT (C) mice revealed no distinguishable change in RIP immunoreactivity in the primary motor cortex
(M1) among the 3 experimental groups. D, Quantitative analysis of photomicrographs of RIP-labeled CNS tissue revealed no significant change in mean fluorescent pixel density in gray and white
matter regions of the brain and spinal cord. E–G, Immunohistochemistry against the myelin marker MAG in spinal cord sections from MBP-DTR ⫹ saline (E), wild-type ⫹ DT (F), and MBP-DTR ⫹
DT (G) mice revealed no obvious change in immunoreactivity among the 3 experimental groups. H, Quantitative analysis of photomicrographs of MAG-labeled CNS tissue revealed no significant
change in mean fluorescent pixel density in gray and white matter regions of the brain and spinal cord. I, J, Electron micrographs of ultrathin transverse sections of the dorsal funiculus of the lumbar
spinal cord of wild-type ⫹DT (I) and MBP-DTR ⫹ DT (J) mice at clinical end point revealed normal-appearing myelinated axons. K, Comparison of g-ratios for axons in the dorsal funiculus of the
spinal cord revealed a 2.1% reduction in mean g-ratio for axons of DT-challenged MBP-DTR mice versus DT-challenged wild-type controls. L, Scatterplot of g-ratio against axonal diameter of neurons
in the dorsal funiculus. Linear regression analysis revealed that the slope of the line of best fit was significantly lower for DT-challenged MBP-DTR mice compared with wild-type mice ( p ⫽ 0.0057).
This analysis indicates that the g-ratios of larger caliber axons are reduced in DT-challenged MBP-DTR mice relative to controls. M, Evidence of a left shift in the frequency distribution of axons with
a given g-ratio divided into 0.025 bins suggests moderate swelling of oligodendrocyte membranes at the myelin internode. N, Frequency distribution of axon diameters (divided into 0.05 m bins)
was similar for DT-challenged MBP-DTR mice and controls indicating no differences in the proportion of axons of a given diameter. O, Immunohistochemical analysis of the spinal cord revealed no
difference in the number of neuronal cell bodies (NeuN ⫹) per millimeter squared in DT-challenged MBP-DTR versus wild-type mice. Immunohistochemical analysis of the spinal cord revealed no
difference in the number of axons (SMI312 ⫹) per millimeter squared in DT-challenged MBP-DTR versus wild-type mice. Nd, not determined. Scale bars: (in A, G) A–C, E–G, 200 m; (in J) I, J, 200
m. AC, anterior commissure; CbW, cerebellar white matter; CC, corpus callosum; Ctx, cerebral cortex; SpW, spinal cord gray matter.
sections assessed, thus providing no evidence for T-cell recruitment into the CNS in this model.
Maintenance of myelin and neuronal preservation following
oligodendrocyte ablation
Demyelination with neuronal preservation can occur as a result
of oligodendrocyte death or dysfunction (Blakemore, 1973;
Traka et al., 2010). To investigate demyelination in the MBPDTR model, sections from DT-challenged MBP-DTR mice (n ⫽
7) and wild-type counterparts (n ⫽ 6), as well as saline-injected
MBP-DTR mice (n ⫽ 3), were analyzed immunohistochemically
for expression of the myelin proteins, 2⬘,3⬘-cyclic nucleotide 3⬘phosphodiesterase (CNPase; Fig. 4 A–C) and MAG (Fig. 4 E–G)
at clinical end point. CNPase and MAG proteins were chosen for
analysis as they are both lost early during experimental demyelination (Felts et al., 2005) and in pattern III lesions in MS (Luc-
chinetti et al., 2000; Aboul-Enein et al., 2003; Marik et al., 2007),
and are therefore considered robust immunohistochemical
markers of demyelination. Quantitative analysis of the fluorescence intensity in photomicrographs using ImageJ revealed no
changes in RIP (CNPase) or MAG labeling in any of the analyzed
regions of the DT-challenged MBP-DTR transgenic mice when
compared with controls (Fig. 4 D, H; two-way ANOVA, p ⫽ 0.728
and p ⫽ 0.602, respectively). Immunohistochemical analysis
therefore failed to provide any evidence of overt demyelination in
symptomatic mice.
To further investigate the integrity of the myelin sheath, we
examined axial sections of the dorsal funiculus of the lumbar
spinal cord using electron microscopic analysis. The dorsal funiculus was selected for ultrastructural analysis since it contains axons integrally involved in the control of motor function and
CC-1 ⫹ cell counts revealed a 39.5 ⫾ 5.2% reduction in oligoden-
Oluich et al.• Axonal Pathology Independent of Demyelination following Oligodendrocyte Apoptosis
J. Neurosci., June 13, 2012 • 32(24):8317– 8330 • 8325
drocyte cell bodies in this region (see
above). Sections were prepared from both
MBP-DTR (n ⫽ 2) and wild-type (n ⫽ 2)
mice challenged with DT and examined at
the clinical end point for the MBP-DTR
mice (Fig. 4 I, J ). Measurements of fiber
and axon areas were used to calculate
g-ratios (Fig. 4 K–M ). The mean g-ratio of
axons from DT-challenged MBP-DTR
mice was 2.1% less than that of DTchallenged wild-type controls (Fig. 4 K;
p ⫽ 0.0103, Mann–Whitney test). Plotting g-ratios against axonal diameters revealed that g-ratios of larger caliber axons
were reduced in DT-challenged MBPDTR mice relative to controls (Fig. 4 L;
slope of line of best fit: MBP-DTR ⫹ DT,
0.035 ⫾ 0.014; wild-type ⫹ DT, 0.087 ⫾
0.013; p ⫽ 0.0057). Consistent with these
data, frequency distributions of g-ratios
revealed a left shift in the population
suggestive of moderate swelling of oligodendrocyte membranes at the myelin
internode (Fig. 4 M). This conclusion was
supported by the finding that mean axon
diameter was not different in the two
groups (MBP-DTR ⫹ DT, 0.72 ⫾ 0.02;
wild-type ⫹ DT, 0.68 ⫾ 0.02; p ⫽ 0.223,
Mann–Whitney test) and the distribution
of axonal diameters remained unaltered
(Fig. 4 N). Collectively, these findings indicate that in regions of oligodendrocyte
cell body loss in symptomatic MBP-DTR
mice challenged with DT there is no overt
demyelination and that compact myelin
remains essentially intact (Fig. 4 I, J ).
To investigate neuronal loss in the
MBP-DTR model, spinal cord sections
from DT-challenged MBP-DTR mice
(n ⫽ 7) and wild-type counterparts (n ⫽
6) were analyzed immunohistochemically
for expression of neuronal markers (Fig.
4O,P). This revealed no difference in the
number of neuronal cell bodies (NeuN ⫹,
p ⫽ 0.4714) or the number of axons
(SMI31 2⫹, p ⫽ 0.7753) per millimeter
squared in DT-challenged MBP-DTR versus wild-type mice. Similarly, assessment
4
Figure 5. Pathological changes in axons and nodes of Ranvier following oligodendrocyte ablation. A–F, Immunohistochemistry against APP (A, D, red) and RIP (B, E, green) in the cerebral cortex (A–C) and corpus callosum (D–F) of a DT-challenged MBP-DTR
mouse. Note the APP ⫹ spheroid (A, arrow) proximal to an RIP-labeled myelinated fiber (B, arrowheads) shows evidence of APP
accumulation along its length. C and F show the merged images with Hoechst 33342 counterstain. G, H, Immunohistochemical
assessment of Nav1.6 (red) and Kv1.2 (green) expression at nodes of Ranvier in longitudinal sections of the dorsal funiculus of the
lumbar spinal cord revealed lengthening of the Nav1.6-expressing domain at the node in symptomatic DT-challenged MBP-DTR
mice (H) compared with the DT-challenged wild-type mice (G). Note the difference in linear measures of Nav1.6 labeling for given
nodes in H versus G. I, J, Immunohistochemical labeling of neurofascin (Nfasc, green) and ankyrin G (red) in longitudinal sections
of the dorsal funiculus of the lumbar spinal cord revealed lengthening of the ankyrin G-labeled domain at the node of DTchallenged MBP-DTR mice (J, indicated by white bars) and reduced or abnormally distributed Nfasc expression. K, Quantification of
the density of nodal Nav1.6 expression revealed no change in node density within dorsal, lateral, or ventral aspects of the lumbar
spinal cord. L, Measurement of the length of ankyrin G expression at nodes demonstrated a significant overall effect of genotype
(p ⬍ 0.0001, two-way ANOVA) and post hoc analysis confirmed genotype-specific differences in the dorsal funiculus
(p ⬍ 0.0001) and lateral column (p ⬍ 0.01, Bonferroni’s post
hoc analysis). M, Frequency plot of node length demonstrated
a right shift in the distribution of node lengths in MBP-DTR ⫹
DT mice as opposed to wild-type ⫹ DT mice. N, Percentage of
nodes defined by discrete ankyrin G expression that exhibited
bilateral neurofascin (Nfasc) labeling. Two-way ANOVA revealed a significant overall effect of genotype ( p ⫽ 0.0001)
and a significant effect of genotype in both dorsal ( p ⬍ 0.01)
and lateral ( p ⬍ 0.05) spinal cord (Bonferroni’s post hoc analysis). Dors., dorsal; Lat., lateral; Vent., ventral. Data represent
mean ⫾ SEM. Scale bars: (in C) A–E, 25 m; (in J) G–J, 5 m.
8326 • J. Neurosci., June 13, 2012 • 32(24):8317– 8330
Oluich et al.• Axonal Pathology Independent of Demyelination following Oligodendrocyte Apoptosis
of axonal density in coronal sections of the
hindbrain of DT-challenged MBP-DTR
and wild-type mice, a region that exhibited a
23.3% reduction in CC-1 ⫹ cell bodies, did
not reveal any difference in the density of
SMI312 ⫹ axons (data not shown). Collectively these data indicate that functional deficits at the clinical end point are not due to
neuronal or axonal loss.
Axonal injury is observed following
oligodendrocyte ablation
The severe clinical phenotype observed in
DT-challenged MBP-DTR mice suggests
that a significant neuropathology induced
by oligodendrocyte cell death could occur
even though gross structural integrity and
myelination were preserved. To investigate this issue further, we looked for evidence of an underlying axonal pathology Figure 6. Dissociation of paranodal loops from the axolemma of myelinated neurons in the spinal cord following oligodendroin symptomatic DT-challenged MBP- cyte ablation. A–D, Electron microscopic assessment of nodes of Ranvier in longitudinal sections of the dorsal funiculus of the
DTR mice. First, we assessed sections for lumbar spinal cord revealed the nature of the interaction of paranodal loops and the axolemma of DT-challenged wild-type (A) and
the accumulation of axonal spheroids DT-challenged MBP-DTR mice (B–D). In all nodes examined in DT-challenged wild-type mice the paranodal loops closely abut the
containing amyloid precursor protein axon (A, arrowheads). A subset of nodes in DT-challenged MBP-DTR mice also maintained this regular structure (B); however,
(APP), a robust indicator of acute damage many nodes in these animals demonstrated alterations in axonal caliber, accumulation of organelles at the node and paranode (C,
to axonal structure (Trapp et al., 1998). D, arrows), and unfurling or frank disconnection of paranodal loops from the axon (C, D, arrowheads). There was a 59.4% increase
Throughout the CNS of symptomatic in the number of nodes described as abnormal (unpaired Student’s t test). N, node; PN, paranode. Scale bar: (in D) A–D, 1 m.
DT-challenged transgenic mice, but not
challenged MBP-DTR mice and controls in the dorsal, lateral, or
controls, APP ⫹ axonal spheroids of ⬃3 to 5 m in diameter were
ventral column of the lumbar spinal cord (Fig. 5K; overall effect
visible but at very low density (Fig. 5 A, D; approximately two to
of genotype, p ⫽ 0.5202; 2-way ANOVA), consistent with the
three per coronal section). When sections were colabeled with the
normal density of SMI312 ⫹ axons previously described. Meamyelin marker RIP and visualized under high magnification,
surement of the length of the ankyrin G-expressing domain at
APP ⫹ spheroids were found in regions that remained heavily
nodes revealed an increase in node length in symptomatic DTmyelinated (Fig. 5 B, E). Accumulation of APP along the length of
⫹
challenged MBP-DTR animals relative to controls (Fig. 5L; overmyelinated (RIP ) axons was also observed (Fig. 5C), indicating
all effect of genotype, p ⬍ 0.0001; 2-way ANOVA). Bonferroni’s
that the infrequent observation of axonal pathology was not repost hoc analysis revealed that relative to controls, node length in
stricted to rare demyelinated fibers.
DT-challenged MBP-DTR mice was significantly increased in the
dorsal funiculus ( p ⬍ 0.0001) and lateral column ( p ⬍ 0.01) but
Axonal node length is increased and paranodal neurofascin
not in the ventral column ( p ⬎ 0.05). To analyze node lengthenlabeling is reduced in the spinal cord of symptomatic MBPing in greater detail, the lengths of ankyrin G-labeled nodes in
DTR mice
dorsal, lateral, and ventral columns were grouped into 0.4 m
We next assessed evidence for changes in the expression profile of
bins
and plotted to generate a frequency distribution of overall
proteins normally clustered at the node of Ranvier and paranode
node lengths in the lumbar spinal cord of DT-challenged MBPthat are required for fast axonal conduction (Vabnick and
DTR animals compared with wild-type controls (Fig. 5M ). The
Shrager, 1998). Longitudinal sections of the spinal cord of DTplot reveals a right-shifted distribution of node lengths for DTchallenged MBP-DTR mice (n ⫽ 5) and DT-challenged wild-type
challenged MBP-DTR animals relative to controls. Analysis of the
littermates (n ⫽ 4) were collected at the clinical end point and
curve of best fit (data not shown) revealed that the two frequency
immunolabeled for Nav1.6 and Kv1.2 (Fig. 5G,H ), proteins nordistributions were significantly different (extra sum-of-squares F
mally localized to the node and juxtaparanode, respectively (Potest, p ⬍ 0.0001).
liak and Peles, 2003). In animals treated with DT, we observed an
We also noted changes in the profile of neurofascin expression
increase in the average length of the Nav1.6 expression domain at
in regions outside the ankyrin G-expressing domain, considered
the node, suggesting dispersal of sodium channels following DT
to represent the oligodendrocyte-specific Nfasc155 isoform at the
administration (Fig. 5G,H, linear measurements). In contrast,
paranode. Many nodes in DT-challenged MBP-DTR mice demKv1.2 labeling appeared normal in symptomatic animals.
onstrated almost complete loss of Nfasc155 labeling. This was of
Given the dispersed pattern of Nav1.6 expression at the node,
particular interest given the integral role of Nfasc155 in establishwe next examined the expression of ankyrin G, a structural proing axoglial structure at the paranode (Zonta et al., 2008). In
tein that plays a critical role in clustering voltage-gated sodium
contrast, the expression profile of neurofascin within the ankyrin
channels at the node. To identify nodes, ankyrin G-stained secG-expressing domain, likely reflecting the expression of neurontions were colabeled with a pan-neurofascin (Nfasc) antibody
specific Nfasc186, was preserved. Overall, compared with DTthat detects both oligodendrocyte-specific (Nfasc155) and
challenged wild-type controls, symptomatic DT-challenged
neuron-specific (Nfasc186) isoforms (Tait et al., 2000) (Fig.
MBP-DTR mice exhibited a 12 ⫾ 1.2% reduction in the percent5 I, J ). First, we quantified node density on the basis of Nav1.6
labeling and observed no significant differences between DTage of nodes at which Nfasc155 was present (Fig. 5N; overall effect
Oluich et al.• Axonal Pathology Independent of Demyelination following Oligodendrocyte Apoptosis
J. Neurosci., June 13, 2012 • 32(24):8317– 8330 • 8327
Figure 7. Preservation of peripheral myelination and conduction in the sciatic nerve of MBP-DTR mice. A, B, Immunohistochemical analysis of DTR expression in the sciatic nerve of unchallenged
MBP-DTR (A) and wild-type control (B) mice revealed restricted foci of DTR protein within the perinodal regions of MBP-DTR nerves (B, arrowheads). C–E, Colocalization of perinodal DTR (C, green)
with the Schwann cell microvillar marker ezrin (D, red). F–H, Microvillar DTR expression (F, green) in the sciatic nerve surrounds domains of axonal Nav1.6 expression (G, red). I, J, Electron
micrographs of transverse sections of the sciatic nerve of wild-type ⫹DT (I) and MBP-DTR ⫹ DT (J) mice at clinical end point revealed normal-appearing myelinated axons. K, Scatterplot of g-ratio
against axon diameter revealed no differences between DT-challenged MBP-DTR mice (red) and DT-challenged wild-type controls (blue). L, Electrophysiological assessment of acutely isolated sciatic
nerves revealed no significant alteration in peripheral nerve CAPs in DT-challenged MBP-DTR mice versus wild-type mice. Scale bar, (in A, E, H, I) A–J, 10 m.
of genotype: p ⫽ 0.0001, 2-way ANOVA; specific effects of genotype: dorsal, p ⬍ 0.01; lateral, p ⬍ 0.05, Bonferroni’s post hoc
analysis).
teins. As a result, there are significant changes in the distribution
of ion channels that determine saltatory conduction velocity in
these axons.
Abnormalities of myelination in disabled MBP-DTR mice are
restricted to the paranodal region
To clarify the nature of axonal pathology occurring at the node
given the evidence of abnormal protein localization of Nav1.6,
ankyrin G, and neurofascin, we assessed nodes of Ranvier at the
ultrastructural level. We prepared ultrathin sections from the
dorsal spinal cord of two DT-challenged MBP-DTR mice, two
DT-challenged wild-type littermates, and one vehicle-treated
control. In control DT-challenged wild-type mice and a vehicletreated transgenic mouse, normal nodal structure was maintained (Fig. 6 A; wild-type ⫹ DT). Paranodal myelin loops were
present abutting the node and tightly juxtaposed to the axonal
membrane, and the axonal caliber was regular. Some nodes from
DT-challenged transgenic mice maintained this regular pattern
(Fig. 6 B). However, in a significant proportion of nodes examined in DT-challenged MBP-DTR mice, substantial abnormalities
were evident with disruption of axonal caliber and accumulation
of organelles at both the node and paranode (Fig. 6C,D). There
was also evidence that the most proximal paranodal myelin loops
(i.e., those closest to the node) were “disconnected” from the
axon and everted away from the axon, disrupting the regular
structure present at normal nodes (Fig. 6C,D, arrowheads). In
some cases, paranodes were completely dissociated from the axolemma, reminiscent of the loss of oligodendrocyte-specific neurofascin labeling characterized immunohistochemically. We next
extended this assessment to a quantitative analysis of ultrastructural changes in the nodes of Ranvier in longitudinal sections of
the dorsal funiculus within the lumbar spinal cord of both vehicle
and DT-challenged MBP-DTR mice. We found that 59.4% of
nodes in the DT-challenged MBP-DTR mice exhibited paranodal
loop eversion which was never observed in DT-challenged wildtype mice (n ⫽ 32 MBP-DTR ⫹ DT, n ⫽ 14 wild-type ⫹ DT
paranodal domains). Collectively, these data indicate that administration of DT leads to significant changes in structure of the
nodes and paranodes and the distribution of organizational pro-
Peripheral nerve myelination and conduction are unaffected
in DT-challenged MBP-DTR mice
Given that the endogenous MBP gene is expressed not only in
oligodendrocytes but also in myelinating Schwann cells (Lemke,
1988) we also assessed the consequences of DT challenge in the
peripheral nervous system (PNS) of the MBP-DTR mice. Immunohistochemical staining of the sciatic nerve revealed that there
was low-level expression of DTR proximal to the node of Ranvier
(Fig. 7 A, B). Double labeling confirmed that DTR was localized
within perinodal microvilli of Schwann cells defined by expression of ezrin (Fig. 7C–E), surrounding the domain of axonal
Nav1.6 expression (Fig. 7F–H ). To assess whether Schwann cells
were ablated following DT challenge, we quantified the density of
Krox20 ⫹ cells in the sciatic nerve and found a 33.1 ⫾ 7.3% reduction in the number of Krox20 ⫹ Schwann cells at clinical end
point (585 ⫾ 54/mm 2 in wild-type ⫹ DT mice versus 391 ⫾
43/mm 2 MBP-DTR ⫹ DT mice, n ⫽ 3 per group; p ⬍ 0.048,
unpaired Student’s two-tailed t test). Ultrastructural analysis of
the sciatic nerves revealed that myelination is normal in the DTchallenged MBP-DTR mice compared with DT-challenged wildtype (Fig. 7 I, J ), consistent with our observations concerning
myelination in the CNS. Plotting g-ratio against axon diameter
revealed similar distributions for both experimental groups (Fig.
7K ) with a mean g-ratio of 0.6085 ⫾ 0.0036 in MBP-DTR ⫹ DT
mice compared with a mean g-ratio of 0.6031 ⫾ 0.003 in wildtype ⫹ DT mice ( p ⫽ 0.258, unpaired Student’s two-tailed t test).
To assess peripheral nerve function in these mice we measured
compound action potential conduction in isolated sciatic nerves
from DT-challenged MBP-DTR and wild-type mice (n ⫽ 8 and
three nerves per group, respectively). In contrast to the significant reduction in the velocity and amplitude of central conduction measured by SSEP assessment, no significant alteration in
peripheral nerve CAPs was observed in nerves isolated from the
DT-challenged MBP-DTR mice (Fig. 7L; latency of CAP response: 1.247 ⫾ 0.009 ms in MBP-DTR ⫹ DT (n ⫽ 8) compared
8328 • J. Neurosci., June 13, 2012 • 32(24):8317– 8330
Oluich et al.• Axonal Pathology Independent of Demyelination following Oligodendrocyte Apoptosis
with 1.225 ⫾ 0.014 ms in wild-type ⫹ DT mice (n ⫽ 3); p ⫽
0.231, Student’s two-tailed t test). Collectively these data suggest
that Schwann cells are targeted by DT challenge but that both
myelination and peripheral nerve electrophysiological parameters are unaltered. We therefore conclude that Schwann cell targeting in the PNS is likely to be, at most, a minor contributor to
the overall phenotype of the DT-challenged mice.
population highly sensitive to DT-mediated cytotoxicity. In contrast, the binary genetic models rely upon repetitive administration of tamoxifen (Traka et al., 2010; Pohl et al., 2011) or DT
(Buch et al., 2005; Locatelli et al., 2012) over several days that
would be expected to induce an asynchronous course of DTmediated cytotoxicity and cell death rather that the acute toxicity
observed in the MBP-DTR model. In addition to differences in
kinetics, the percentage depletion of oligodendrocytes across various regions of gray and white matter will reflect the specific Cre
drivers and floxed alleles that are used and the precise regimen of
tamoxifen or DT that is administered (Saito et al., 2001; Leone et
al., 2003).
Discussion
Here we describe the generation and characterization of a novel
model of conditional oligodendrocyte ablation using transgenic
mice that express DTR in oligodendrocytes throughout the CNS.
In response to systemic DT administration, MBP-DTR mice progressively developed profound clinical disability characterized by
ataxia, spastic paralysis of the hindlimbs, kyphosis, weight loss,
and ultimately respiratory paralysis within 22 d of DT injection.
At clinical end point, mice exhibited the loss of ⬃26% of oligodendrocyte cell bodies throughout the CNS but maintained normal g-ratios and normal levels of MAG and CNP proteins
indicating no overt demyelination. Schwann cells were also targeted; however, their loss was not associated with abnormalities
of peripheral nerve conduction and the animals displayed a spastic rather than a flaccid paralysis, indicative of a central rather
than a peripheral clinical deficit. Disabled mice displayed eversion of paranodal loops away from the axolemma and altered
distribution of ion channels and structural proteins at both the
nodes of Ranvier and paranodes resulting in nodal lengthening.
Although axonal density within the spinal cord was unaltered,
axonal pathology was evident from APP accumulation within
axonal swellings and an increased SSEP latency indicative of reduced conduction velocity. Collectively, these data reveal a critical role for oligodendrocytes in maintaining the structural and
functional integrity of axons that is distinct from the provision of
myelin.
Comparison with other models of conditional
oligodendrocyte ablation
Previously described binary genetic models of conditional oligodendrocyte ablation were established by crossing two transgenic
lines to direct Cre-mediated expression of DT-A or DTR in mature oligodendrocytes (Buch et al., 2005; Traka et al., 2010; Pohl
et al., 2011). Despite the different genetic strategies used to induce oligodendrocyte ablation, all transgenic models exhibit a
similar clinical phenotype but differ with respect to time of symptom onset and clinical course, with MBP-DTR mice exhibiting
the most rapid and severe phenotype. Whereas the former models are associated with overt CNS demyelination and/or myelin
vacuolation, DT-challenged MBP-DTR mice exhibited profound
clinical disability and axonal pathology before these histopathological features appear. Myelin preservation in symptomatic DTchallenged MBP-DTR mice at clinical end point (⬃16 d post-DT
administration) suggests that degenerative mechanisms responsible for degradation of compact myelin following oligodendrocyte apoptosis are not fully induced before the clinical end point.
Comparatively, demyelination in MOGi-Cre:iDTR (oDTR) mice
(Buch et al., 2005; Locatelli et al., 2012) and PLP/CreERT:
ROSA26-eGFP-DTA mice (Traka et al., 2010) was first observed
28 and 35 d postchallenge, respectively, whereas reduced MBP
immunoreactivity in PLP-CreERT2:R26-LacZ/DT-A mice was
prominent at 39 – 42 d post tamoxifen (Pohl et al., 2011). The
rapid onset of clinical pathology in MBP-DTR mice within 9 –12
d after a single DT injection likely reflects the high-level constitutive expression of DTR by oligodendrocytes that renders the
Conduction impairment is associated with disruption of
nodes and paranodes
The clinical presentation of ataxia and hindlimb paralysis was
corroborated by increased latency of SSEP recordings. Structurally, we observed the presence of axonal spheroids at low density
throughout the CNS of DT-challenged MBP-DTR mice, a marker
of early axonal pathology and transection (Trapp et al., 1998).
Ultrastructural analysis revealed disorganization of paranodal
structure within spinal cord white matter characterized by eversion of the paranodal loops from the axolemma. Since node and
fiber density were unaltered in symptomatic mice, the observation that only a subset of nodes exhibited increased nodal Nav1.6
and ankyrin G expression suggests that paranodal eversion likely
precedes redistribution of Nav1.6 and ankyrin G. Current understanding of paranodal structure and function support this conclusion. Tight axoglial junctions at the paranode form via high
molecular weight complexes between the axonal proteins caspr
and contactin and oligodendrocyte Nfasc155 (Tait et al., 2000;
Bhat et al., 2001). By preventing lateral membrane diffusion, the
paranode maintains voltage-gated Na ⫹ and K ⫹ channels at high
concentration at the node and juxtaparanode, respectively,
thereby enabling fast saltatory conduction (Sherman and
Brophy, 2005). In symptomatic MBP-DTR mice, maintenance of
Kv1.2 expression within the juxtaparanode but dispersal of nodal
Nav1.6 therefore likely reflects the selective disruption of paranodal loops located closest to the node but preservation of distal
loop attachment to the axolemma.
The nodal and paranodal changes observed in MBP-DTR
mice bear similarities to defects in a number of myelin and paranodal protein-mutant mice and with changes observed in MS
tissue. Notably, mice deficient in Nfasc155 or Caspr do not form
normal paranodes and exhibit aberrant clustering of voltagegated ion channels on axons resulting in various motor deficits
(Bhat et al., 2001; Rios et al., 2003; Pillai et al., 2009). In MS
subjects, Caspr and Nfasc155 are abnormally distributed on
myelinated axons that border chronic demyelinated lesions
(Wolswijk and Balesar, 2003; Howell et al., 2006) suggesting that
changes in paranodal integrity could occur at the leading edge of
MS lesion formation before demyelination. Paranodal eversion
has also been described in transgenic mice with additional copies
of the plp gene (Tanaka et al., 2009) and in CGT mutant mice that
have aberrant but viable oligodendrocytes (Dupree et al., 1998).
Dispersal of nodal proteins away from the node and paranodal
disruption characterized by loss of the axoglial junctions therefore appear to be a common mechanism leading to functional
deficits that can occur in isolation or coincident with demyelination. Independent of the initiating event, these pathological
changes at the node and paranode are likely to contribute to
failure of saltatory conduction and reduced conduction velocity.
Oluich et al.• Axonal Pathology Independent of Demyelination following Oligodendrocyte Apoptosis
J. Neurosci., June 13, 2012 • 32(24):8317– 8330 • 8329
Immune activity and myelination
The activation, proliferation, recruitment, and conversion of microglia to a phagocytic amoeboid phenotype is an important cellular response that contributes to clearance of myelin debris in
various models of CNS demyelination (Merson et al., 2010). In
contrast, DT-challenged MBP-DTR mice exhibited only a modest increase in microglial cell density that retained a ramified
morphology with evidence of process thickening reminiscent of
“prephagocytic” lesions in early MS cases (Barnett and Prineas,
2004), and “predemyelinating” lesions described by Marik et al.
(2007). Nevertheless, close apposition between microglia and oligodendrocyte cell bodies 5 d following DT administration suggests that microglia are responsive to early alterations in
oligodendrocyte physiology. Importantly, microglia are capable
of phagocytosing damaged cells and membranes without exhibiting
an amoeboid morphology (Davalos et al., 2005; Nimmerjahn et al.,
2005; Kim and Dustin, 2006) and rapid clearance of apoptotic cell
bodies does not stimulate a proinflammatory response in macrophages (Gregory and Devitt, 2004; Town et al., 2005). Thus microglia in MBP-DTR mice could participate in restricted “surgical”
phagocytosis of the cell bodies of apoptotic oligodendrocytes without conversion to an amoeboid phenotype required for efficient
phagocytosis of compact myelin debris.
It was also notable that there was no adaptive immune response consistent with previous observations in oligodendrocyte
ablation models (Buch et al., 2005; Traka et al., 2010; Pohl et al.,
2011; Locatelli et al., 2012). The absence of lymphocyte infiltration has also been observed in acute MS lesions (Barnett and
Prineas, 2004; Henderson et al., 2009) and therefore does not
mitigate against the potential relevance of this model to MS. The
essential issue that remains in testing the veracity of Barnett and
Prineas’ (2004) hypothesis of oligodendrocytopathy as a potential cause of MS is to determine whether there are downstream
circumstances in which the adaptive immune system does become activated, as occurs in chronic MS. The recent study by
Locatelli et al. (2012), demonstrating that oligodendrocyte cell
death failed to potentiate anti-CNS immunity, even in the context of potent induction of T-cell activation and inflammation,
suggests that autoimmunity is not readily modulated by oligodendrocyte death. Thus an important issue in addressing Barnett
and Prineas’ (2004) hypothesis will be to establish how engagement with the immune system differs between models of oligodendrocytopathy that induce T-cell recruitment into the CNS (Ip
et al., 2006; Kassmann et al., 2007) and models of oligodendrocyte cell death that do not.
fore, or in the absence of, demyelination. This phenotype closely
mirrors the early MS lesion, i.e., the prephagocytic lesions described by Barnett and Prineas (2004), as well as the predemyelinating lesions described by Marik et al. (2007) and Pattern III MS
lesions (Lucchinetti et al., 2000), providing an opportunity to
elucidate the processes that lead to demyelination and axonal
degeneration and the relationship between the events. Importantly, the data also serve to illustrate that the genesis of axonal
pathology in the absence of demyelination in recent onset disease
could be induced by an oligodendrocytopathy.
Oligodendrocytes have been proposed to play an integral role
in the maintenance of axonal function by providing trophic support to axons via the provision of energy-rich metabolites to
myelinated axons (Nave, 2010). Further evaluation of the MBPDTR model will help clarify the molecular mechanisms that govern
oligodendrocyte–axon interactions, and the types of metabolites
that might be supplied to axons. Ascertaining the nature of the
cytokine profile that predominates in the MBP-DTR model, and
contrasting it with that which prevails in demyelinating models
could delineate the molecular mechanisms regulating axonal degeneration on the one hand and myelin breakdown and
microglial-mediated phagocytosis on the other. Such information could facilitate our understanding of how these various processes are controlled and the circumstances in which it is
appropriate to modulate each of them.
Concluding remarks
We have demonstrated that oligodendrocyte cell death disrupts
the normal structural and molecular integrity of paranodes and
nodes of Ranvier, leading to impaired conduction of action potentials and a gross clinical phenotype despite maintenance of
normal myelin internodes. The direct expression of DTR in mature oligodendrocytes has many attractive features as a model for
oligodendrocyte ablation. Unlike existing binary genetic approaches to ablate oligodendrocytes, the MBP-DTR model provides unlimited freedom to apply Cre/loxP genetics to further
explore pathogenic mechanisms. Moreover, this phenotype is
elicited with a single dose of DT, enabling clear dissociation of
reactive changes that could otherwise occur contemporaneously
in the binary genetic models. The MBP-DTR model will facilitate
the analysis of the early events involved in the breakdown of the
symbiotic relationship between oligodendrocytes and axons. Our
results clearly demonstrate that axonal impairment occurs be-
References
Aboul-Enein F, Rauschka H, Kornek B, Stadelmann C, Stefferl A, Brück W,
Lucchinetti C, Schmidbauer M, Jellinger K, Lassmann H (2003) Preferential loss of myelin-associated glycoprotein reflects hypoxia-like white
matter damage in stroke and inflammatory brain diseases. J Neuropathol
Exp Neurol 62:25–33.
Barnett MH, Prineas JW (2004) Relapsing and remitting multiple sclerosis:
pathology of the newly forming lesion. Ann Neurol 55:458 – 468.
Bhat MA, Rios JC, Lu Y, Garcia-Fresco GP, Ching W, St Martin M, Li J,
Einheber S, Chesler M, Rosenbluth J, Salzer JL, Bellen HJ (2001) Axonglia interactions and the domain organization of myelinated axons requires neurexin IV/Caspr/Paranodin. Neuron 30:369 –383.
Blakemore WF (1973) Demyelination of the superior cerebellar peduncle in
the mouse induced by cuprizone. J Neurol Sci 20:63–72.
Buch T, Heppner FL, Tertilt C, Heinen TJAJ, Kremer M, Wunderlich FT, Jung
S, Waisman A (2005) A Cre-inducible diphtheria toxin receptor mediates cell lineage ablation after toxin administration. Nat Meth 2:419 – 426.
Collier RJ (2001) Understanding the mode of action of diphtheria toxin: a
perspective on progress during the 20th century. Toxicon 39:1793–1803.
Davalos D, Grutzendler J, Yang G, Kim JV, Zuo Y, Jung S, Littman DR, Dustin
ML, Gan WB (2005) ATP mediates rapid microglial response to local
brain injury in vivo. Nat Neurosci 8:752–758.
Dowling P, Ming X, Raval S, Husar W, Casaccia-Bonnefil P, Chao M, Cook S,
Blumberg B (1999) Up-regulated p75NTR neurotrophin receptor on
glial cells in MS plaques. Neurology 53:1676 –1682.
Dupree JL, Coetzee T, Blight A, Suzuki K, Popko B (1998) Myelin galactolipids are essential for proper node of Ranvier formation in the CNS.
J Neurosci 18:1642–1649.
Emery B, Cate HS, Marriott M, Merson T, Binder MD, Snell C, Soo PY,
Murray S, Croker B, Zhang JG, Alexander WS, Cooper H, Butzkueven H,
Kilpatrick TJ (2006) Suppressor of cytokine signaling 3 limits protection
of leukemia inhibitory factor receptor signaling against central demyelination. Proc Natl Acad Sci U S A 103:7859 –7864.
Feigenbaum V, Gélot A, Casanova P, Daumas-Duport C, Aubourg P, DuboisDalcq M (2000) Apoptosis in the central nervous system of cerebral adrenoleukodystrophy patients. Neurobiol Dis 7:600 – 612.
Felts PA, Woolston A-M, Fernando HB, Asquith S, Gregson NA, Mizzi OJ,
Smith KJ (2005) Inflammation and primary demyelination induced by
the intraspinal injection of lipopolysaccharide. Brain 128:1649 –1666.
Frohman EM, Racke MK, Raine CS (2006) Multiple sclerosis—the plaque
and its pathogenesis. N Engl J Med 354:942–955.
Gow A, Friedrich VL Jr, Lazzarini RA (1992) Myelin basic protein gene
8330 • J. Neurosci., June 13, 2012 • 32(24):8317– 8330
Oluich et al.• Axonal Pathology Independent of Demyelination following Oligodendrocyte Apoptosis
contains separate enhancers for oligodendrocyte and Schwann cell expression. J Cell Biol 119:605– 616.
Gregory CD, Devitt A (2004) The macrophage and the apoptotic cell: an
innate immune interaction viewed simplistically? Immunology 113:1–14.
Haider L, Fischer MT, Frischer JM, Bauer J, Höftberger R, Botond G, Esterbauer H, Binder CJ, Witztum JL, Lassmann H (2011) Oxidative damage
in multiple sclerosis lesions. Brain 134:1914 –1924.
Henderson AP, Barnett MH, Parratt JD, Prineas JW (2009) Multiple sclerosis: distribution of inflammatory cells in newly forming lesions. Ann Neurol 66:739 –753.
Howell OW, Palser A, Polito A, Melrose S, Zonta B, Scheiermann C, Vora AJ,
Brophy PJ, Reynolds R (2006) Disruption of neurofascin localization
reveals early changes preceding demyelination and remyelination in multiple sclerosis. Brain 129:3173–3185.
Ip CW, Kroner A, Bendszus M, Leder C, Kobsar I, Fischer S, Wiendl H, Nave
KA, Martini R (2006) Immune cells contribute to myelin degeneration
and axonopathic changes in mice overexpressing proteolipid protein in
oligodendrocytes. J Neurosci 26:8206 – 8216.
Kassmann CM, Lappe-Siefke C, Baes M, Brügger B, Mildner A, Werner HB,
Natt O, Michaelis T, Prinz M, Frahm J, Nave KA (2007) Axonal loss and
neuroinflammation caused by peroxisome-deficient oligodendrocytes.
Nat Genet 39:969 –976.
Kim JV, Dustin ML (2006) Innate response to focal necrotic injury inside
the blood-brain barrier. J Immunol 177:5269 –5277.
Kuhlmann T, Lingfeld G, Bitsch A, Schuchardt J, Brück W (2002) Acute
axonal damage in multiple sclerosis is most extensive in early disease
stages and decreases over time. Brain 125:2202–2212.
Lemke G (1988) Unwrapping the genes of myelin. Neuron 1:535–543.
Leone DP, Genoud S, Atanasoski S, Grausenburger R, Berger P, Metzger D,
Macklin WB, Chambon P, Suter U (2003) Tamoxifen-inducible gliaspecific Cre mice for somatic mutagenesis in oligodendrocytes and
Schwann cells. Mol Cell Neurosci 22:430 – 440.
Locatelli G, Wörtge S, Buch T, Ingold B, Frommer F, Sobottka B, Krüger M,
Karram K, Bühlmann C, Bechmann I, Heppner FL, Waisman A, Becher B
(2012) Primary oligodendrocyte death does not elicit anti-CNS immunity. Nat Neurosci. Advance online publication. Retrieved February 26,
2012. doi:10.1038/nn.3062.
Lucchinetti CF, Brück W, Rodriguez M, Lassmann H (1996) Distinct patterns of multiple sclerosis pathology indicates heterogeneity on pathogenesis. Brain Pathol 6:259 –274.
Lucchinetti C, Brück W, Parisi J, Scheithauer B, Rodriguez M, Lassmann H
(2000) Heterogeneity of multiple sclerosis lesions: implications for the
pathogenesis of demyelination. Ann Neurol 47:707–717.
Marik C, Felts PA, Bauer J, Lassmann H, Smith KJ (2007) Lesion genesis in
a subset of patients with multiple sclerosis: a role for innate immunity?
Brain 130:2800 –2815.
Merson TD, Binder MD, Kilpatrick TJ (2010) Role of cytokines as mediators and regulators of microglial activity in inflammatory demyelination
of the CNS. Neuromolecular Med 12:99 –132.
Nave KA (2010) Myelination and support of axonal integrity by glia. Nature
468:244 –252.
Nikić I, Merkler D, Sorbara C, Brinkoetter M, Kreutzfeldt M, Bareyre FM,
Brück W, Bishop D, Misgeld T, Kerschensteiner M (2011) A reversible
form of axon damage in experimental autoimmune encephalomyelitis
and multiple sclerosis. Nat Med 17:495– 499.
Nimmerjahn A, Kirchhoff F, Helmchen F (2005) Resting microglial cells
are highly dynamic surveillants of brain parenchyma in vivo. Science
308:1314 –1318.
Pillai AM, Thaxton C, Pribisko AL, Cheng JG, Dupree JL, Bhat MA (2009)
Spatiotemporal ablation of myelinating glia-specific neurofascin
(Nfasc NF155) in mice reveals gradual loss of paranodal axoglial junctions and concomitant disorganization of axonal domains. J Neurosci
Res 87:1773–1793.
Pohl HB, Porcheri C, Mueggler T, Bachmann LC, Martino G, Riethmacher D,
Franklin RJ, Rudin M, Suter U (2011) Genetically induced adult oligo-
dendrocyte cell death is associated with poor myelin clearance, reduced
remyelination, and axonal damage. J Neurosci 31:1069 –1080.
Poliak S, Peles E (2003) The local differentiation of myelinated axons at
nodes of Ranvier. Nat Rev Neurosci 4:968 –980.
Rios JC, Rubin M, St Martin M, Downey RT, Einheber S, Rosenbluth J,
Levinson SR, Bhat M, Salzer JL (2003) Paranodal interactions regulate
expression of sodium channel subtypes and provide a diffusion barrier for
the node of Ranvier. J Neurosci 23:7001–7011.
Rivers LE, Young KM, Rizzi M, Jamen F, Psachoulia K, Wade A, Kessaris N,
Richardson WD (2008) PDGFRA/NG2 glia generate myelinating oligodendrocytes and piriform projection neurons in adult mice. Nat Neurosci
11:1392–1401.
Saito M, Iwawaki T, Taya C, Yonekawa H, Noda M, Inui Y, Mekada E, Kimata
Y, Tsuru A, Kohno K (2001) Diphtheria toxin receptor-mediated conditional and targeted cell ablation in transgenic mice. Nat Biotechnol
19:746 –750.
Sherman DL, Brophy PJ (2005) Mechanisms of axon ensheathment and
myelin growth. Nat Rev Neurosci 6:683– 690.
Shults CW, Rockenstein E, Crews L, Adame A, Mante M, Larrea G,
Hashimoto M, Song D, Iwatsubo T, Tsuboi K, Masliah E (2005) Neurological and neurodegenerative alterations in a transgenic mouse model
expressing human alpha-synuclein under oligodendrocyte promoter: implications for multiple system atrophy. J Neurosci 25:10689 –10699.
Srinivas S, Watanabe T, Lin CS, William CM, Tanabe Y, Jessell TM, Costantini F (2001) Cre reporter strains produced by targeted insertion of
EYFP and ECFP into the ROSA26 locus. BMC Dev Biol 1:4.
Stence N, Waite M, Dailey ME (2001) Dynamics of microglial activation: a
confocal time-lapse analysis in hippocampal slices. Glia 33:256 –266.
Tait S, Gunn-Moore F, Collinson JM, Huang J, Lubetzki C, Pedraza L, Sherman DL, Colman DR, Brophy PJ (2000) An oligodendrocyte cell adhesion molecule at the site of assembly of the paranodal axo-glial junction.
J Cell Biol 150:657– 666.
Tanaka H, Ma J, Tanaka KF, Takao K, Komada M, Tanda K, Suzuki A,
Ishibashi T, Baba H, Isa T, Shigemoto R, Ono K, Miyakawa T, Ikenaka K
(2009) Mice with altered myelin proteolipid protein gene expression display cognitive deficits accompanied by abnormal neuron-glia interactions and decreased conduction velocities. J Neurosci 29:8363– 8371.
Town T, Nikolic V, Tan J (2005) The microglial “activation” continuum:
from innate to adaptive responses. J Neuroinflammation 2:24.
Traka M, Arasi K, Avila RL, Podojil JR, Christakos A, Miller SD, Soliven B,
Popko B (2010) A genetic mouse model of adult-onset, pervasive
central nervous system demyelination with robust remyelination.
Brain 133:3017–3029.
Trapp BD, Peterson J, Ransohoff RM, Rudick R, Mörk S, Bö L (1998)
Axonal transection in the lesions of multiple sclerosis. N Engl J Med
338:278 –285.
Trapp BD, Ransohoff R, Rudick R (1999) Axonal pathology in multiple
sclerosis: relationship to neurologic disability. Curr Opin Neurol
12:295–302.
Vabnick I, Shrager P (1998) Ion channel redistribution and function during
development of the myelinated axon. J Neurobiol 37:80 –96.
Weiner HL (2004) Multiple sclerosis is an inflammatory T-cell-mediated
autoimmune disease. Arch Neurol 61:1613–1615.
Wolswijk G, Balesar R (2003) Changes in the expression and localization of
the paranodal protein Caspr on axons in chronic multiple sclerosis. Brain
126:1638 –1649.
Wrabetz L, Taveggia C, Feltri ML, Quattrini A, Awatramani R, Scherer SS,
Messing A, Kamholz J (1998) A minimal human MBP promoter-lacZ
transgene is appropriately regulated in developing brain and after optic
enucleation, but not in shiverer mutant mice. J Neurobiol 34:10 –26.
Zonta B, Tait S, Melrose S, Anderson H, Harroch S, Higginson J, Sherman DL,
Brophy PJ (2008) Glial and neuronal isoforms of Neurofascin have distinct roles in the assembly of nodes of Ranvier in the central nervous
system. J Cell Biol 181:1169 –1177.