| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From the Departments of Neurology*
and
Immunology,
and the Molecular Neuroscience
Program,
Mayo Clinic and Foundation,
Rochester, Minnesota
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Theiler’s murine encephalomyelitis virus (TMEV) causes a chronic progressive demyelinating disease in susceptible strains of mice that is pathologically indistinguishable from MS.4-9 Intracerebral injection of TMEV into mice of the prototypic susceptible background (SJL/J) results in a biphasic disease characterized by an acute neuronal polioencephalitis followed by chronic white matter demyelination and neurological deficits.10 Twenty-one days after postinfection, viral persistence is observed in oligodendrocytes, not neurons, and therefore, should not contribute directly to progressive neuronal disruption.11,12 During the chronic phase of disease, multifocal demyelinating lesions are primarily localized to the white matter of the brain stem and spinal cord, and the demyelinating process is at least in part immune mediated.13 Small, focal demyelinating lesions are detectable by 15 days postinfection,4 and the demyelinating phase of the disease plateaus by 100 days postinfection.14 After this time point, susceptible SJL/J mice continue to decline in neurological function, and this decline is associated with significant increases in spinal cord atrophy and a reduction in myelinated axon frequencies.
Disease course after TMEV infection of susceptible SJL/J mice is heterogeneous. Mice studied during the chronic phase of disease vary in lesion load, axonal loss, and neurological function. The goal of this study was to determine whether pathological variables assessed from a single demyelinated lesion serve as an indicator of neurological dysfunction and the severity of disease course. This hypothesis was based on recent studies in MS indicating that mechanisms of immunopathogenesis within individual patients may serve as an indicator of disease severity.15,16 We tested this hypothesis by conducting quantitative ultrastructural analyses of individual demyelinated spinal cord lesions from SJL/J mice infected for 192 days. We previously demonstrated significant total spinal cord demyelination (a measure of lesion load), spinal cord atrophy, myelinated axon loss in the normal-appearing white matter, and neurological dysfunction in the chronically infected SJL/J mice used in this study.14 This provided the unique opportunity to conduct an independent, blinded study to determine whether correlative relationships existed between variables assessed within a single demyelinated lesion and global assessments of spinal cord pathology and neurological function.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
At 8 weeks of age, SJL/J mice (n = 7) (Jackson Laboratories, Bar Harbor, ME) used in these studies were anesthetized and injected intracranially with 2 x 106 pfu of the Daniel’s strain of TMEV in a 10 µl volume. Six age-matched SJL/J mice were sham-infected intracranially with 10 µl of phosphate-buffered saline and used as controls. Care and handling of mice conformed to the guidelines of both the National Institutes of Health and the Mayo Clinic Animal Care and Use Committee.
Tissue Processing
Mice were anesthetized intraperitoneally with 10 mg of
pentobarbital at 192 days postinfection and perfused via intracardiac
puncture with 50 ml of Trump’s fixative (phosphate-buffered 4%
formaldehyde with 1% glutaraldehyde, pH 7.2). The spinal cords were
sectioned transversely into 1-mm blocks, postfixed with osmium
tet-roxide, and embedded in Araldite (Polysciences; Warrington,
PA). Cross-sections (1-µm thick) were cut from every other 1-mm block
to obtain a complete representation of all cord regions. These 1-µm
cross-sections were stained with 4% paraphenylenediamine to label
myelin. An average of 15 sections per mouse was analyzed by light
microscopy to determine percentages of total spinal cord demyelination
and C7 combined lateral and anterior column areas. For quantitative
ultrastructural analyses using electron microscopy, a single
mid-thoracic spinal cord lesion was selected by light microscopy from
the anterolateral columns (Figure 1)
using the 1-µm paraphenylenediamine-stained cross-sections for all
infected and sham-infected SJL/J mice. The lesion was selected from the
anterolateral columns of the thoracic spinal cord because this is the
most common location to observe lesions.17
Thin sections
were cut at 0.1 µm and counterstained with uranyl acetate and lead
citrate.
|
All ultrastructural analyses were performed in a blinded and
nonbiased manner on x3,000 photographs captured using an electron
microscope (JEOL 1200; JEOL, Peabody, MA). For all mice infected with
TMEV, nonoverlapping photographs were taken of an entire demyelinated
lesion from the anterolateral columns (mean lesion area, 18,355
µm2) (see example in Figure 1B
). Comparable
areas (mean area, 25,606 µm2) were photographed
from the normal white matter in the anterolateral columns of
sham-infected mice to serve as a control (example in Figure 1A
). An
average of 36 photographs was analyzed per mouse. The boundaries of
lesions were identified on the electron microscope by the presence of
the surrounding normal-appearing white matter.
Calculation of Axonal Areas
Cross-sectional axons were traced on the electron
microscope photographs by placing a transparency (3
mol/L; Austin, TX) over the photograph and then outlining the
axolemma of each axon in the field. Longitudinal axons and astrocytic
processes were excluded from the analysis. An average of 3,756 axons
was traced for each mouse. The total number of axons traced for
sham-infected and infected mice was 32,377 and 16,450, respectively.
Axons traced in the demyelinated lesions of infected mice were divided
into three separate categories: demyelinated, remyelinated, and
normally myelinated. A demyelinated axon was defined as one that
completely lacked a surrounding myelin sheath (examples in Figure 1D
).
Oligodendrocyte-mediated remyelination was identified by the appearance
of abnormally thin myelinated axons relative to axonal diameter when
compared to normally myelinated axons (examples in Figure 1E
).18
To further ensure that remyelinated axons were
categorized separately from normally myelinated axons, we calculated
the ratio of myelin thickness to axonal diameter in remyelinated axons
using a x10 magnifier that contained a micrometer. We reported
previously that remyelinated axons have a myelin thickness to axonal
diameter ratio of < 0.14 (mean, 0.08).19
Axons
remyelinated by infiltrating Schwann cells were identified by thicker
myelin sheaths, a surrounding basement membrane, and a one-to-one
relationship between the Schwann cell body and the axon (Figure 1F)
.
After tracing all of the axons for each of the photographs,
transparencies were digitized using a Hewlett Packard scanner.
Axonal areas for normally myelinated, remyelinated, and demyelinated
fibers were then quantified using a program written for the KS400 image
analysis software (Kontron Elektronik Gmbh; Munich, Germany). The
program calculated the number and area of axons from each transparency.
Axonal data for each mouse were represented as axonal frequency
distributions. Axonal frequency distributions were calculated by
dividing the number of axons with areas ranging from 0
µm2
to 85 µm2
(using
0.5-µm2
intervals as bins) by the total number
of axons sampled. Relative frequencies were calculated for all fibers
sampled and for individual fiber types (normally myelinated,
remyelinated, and demyelinated). For example, axonal frequency
distributions for myelinated axons were calculated by dividing the
number of normally myelinated axons in each
0.5-µm2
bin by all of the normally myelinated
axons sampled. Frequency histograms were further divided into three
different size categories to facilitate comparisons: 0 to 4
µm2
(small fibers); 4
µm2
to 10 µm2
(medium
fibers); and >10 µm2
(large fibers).
Quantification of Mitochondria
Intraaxonal mitochondria were counted in all normally myelinated, remyelinated, and demyelinated axons sampled. The data were represented as the number of mitochondria per axonal area (in µm2).
Quantification of Degenerating Axons
Degenerating axons were identified as axons containing an
axoplasmic mass filled with numerous floccular dense bodies, swelling
mitochondria, and granular disintegration of neurofilaments and
vesicles (see examples in Figure 3, A and B
).20
Myelin
ovoids devoid of underlying axolemma or axoplasm were also included as
degenerating axons. Ovoids within macrophages were not included as
degenerating axons. Data were represented as the total number of
degenerating axons counted.
|
The percentage of total spinal cord demyelination (or total lesion load) is a measure of all lesions in the spinal cord white matter and was assessed as described.14 Briefly, a Zeiss interactive digital analysis system and camera lucida attached to a Zeiss photomicroscope (Carl Zeiss Inc., Thornwood, NY) were used to first determine the total white matter area from 10 to 12 paraphenylenediamine-stained spinal cord sections per mouse by tracing all of the white matter. The area of total spinal cord demyelination was determined by tracing each of the demyelinated lesions. Data were represented as percent total spinal cord demyelination per mouse by dividing the total area of demyelination by the total area of white matter sampled and then multiplying by 100.
Spinal Cord Atrophy
We measured the C7 combined lateral and anterior column area as described.14 There is atrophy in both cervical and thoracic spinal cord of 192-day-infected SJL/J mice. C7 was selected because this spinal cord level was reproducibly obtained for all mice used in the study and allowed for assessment of both ascending and descending fiber tracts. Briefly, an Olympus Provis AX70 microscope fitted with a SPOT color digital camera and a x1.25 objective was used to digitize the spinal cord cross-section corresponding to C7 for each mouse. A program written for the KS400 image analysis software was then used to calculate the anterior and lateral white matter area (which included the anterior, lateral, and anterolateral columns) from each section after manually outlining the region. Data were represented as the C7 anterior and lateral column area (in mm2) for each mouse.
Rotarod Analysis
The Rotamex rotarod (Columbus Instruments, Columbus, OH) measures motor coordination/balance and was used as described.14 Briefly, sham-infected and infected SJL/J mice were trained at 191 days after injection using a constant speed rotarod assay (speed, 5 rpm; time, 3 minutes; trials, 3) to familiarize them with the rotarod. This was followed by analysis of all mice at 192 days using an accelerated (7 rpm/minutes) rotarod assay (start speed, 5 rpm; end speed, 40 rpm; time, 5 minutes; trials, 3). The amount of time on the rotarod (seconds) was recorded for each of the three trials per mouse and then averaged.
Statistical Analyses
Sham-infected and infected relative axonal frequencies were compared using an unpaired Student’s t-test (P < 0.05). Mitochondria per axonal area for sham-infected and infected mice were statistically compared using a one-way analysis of variance. Pairwise comparisons were made using the Student-Newman-Keuls Method (P < 0.05). All correlation coefficients were calculated using a Pearson product moment correlation (P < 0.05).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We used electron microscopy to conduct a detailed analysis of
individual lesions from SJL/J mice infected with TMEV for 192 days.
Comparable areas from the anterolateral columns were selected from
sham-infected, age-matched control mice (Figure 1, A and B
, black
boxes). Individual spinal cord lesions sampled from TMEV-infected mice
ranged in size from 7,296 to 35,000 µm2
(Table 1)
. These lesions contained varying
degrees of pathology that included macrophage infiltration, astrocytic
scars, collagen deposition, denuded axons, axonal degeneration, and
remyelinated axons (Table 1)
. Using electron micrographs, we accurately
classified axons into three different categories based on myelin sheath
thickness: normally myelinated, remyelinated, and demyelinated.
Normally myelinated axons have thick myelin sheaths relative to the
axon diameter, and were observed in the white matter of sham-infected
mice (Figure 1C)
and in the lesions of 192-day-infected mice (Figure 1D
, arrows). An average of 45% of the axons sampled from the lesions
of infected mice were normally myelinated (Table 1
; range, -11 to
73%). Demyelinated axons represented 25% of the axons sampled in
infected mice (Table 1
; range, -7 to 54%) and were identified as
axons with axolemma but no myelin sheaths (Figure 1D
, asterisks).
|
or infiltrating Schwann cells (Figure 1F)
.
Oligodendrocyte-mediated remyelination was identified by a thin myelin
sheath relative to axon diameter (Figure 1E
, asterisks).
Oligodendrocyte-mediated remyelination was responsible for >99% of
the total remyelination observed in the lesions of infected SJL/J mice.
Schwann cell remyelination accounted for <1% of the total
remyelination quantified (Figure 1F
, asterisks). An average of 30% of
the axons sampled from the lesions of infected mice were remyelinated
(Table 1
; range, -9 to 56%). Reductions in Large Axonal Frequencies Are Observed in Demyelinated Lesions
We determined the severity of axonal loss from individual
demyelinated lesions. All axons (normally myelinated, remyelinated, and
demyelinated) were measured from the anterolateral columns of
sham-infected and 192-day-infected mice and plotted as
axonal frequency distributions (Figure 2, A–C)
. We represented the data as axonal frequency distributions rather
than axonal densities (axons per unit area) because axonal frequencies
are not influenced by factors in the lesions that increase
extracellular space, such as edema, inflammatory cells, and gliosis.
When axonal frequencies were compared between sham-infected and
infected SJL/J mice, no reductions were observed in small axons (Figure 2A)
. A 9% decrease was observed in the medium axons (Figure 2B)
but
did not reach statistical significance. The most significant decrease
(63%) was observed in the large axons (Figure 2C)
of 192-day-infected
mice, which is consistent with the 62% decrease in large myelinated
fibers estimated previously from the normal-appearing white matter of
these same mice.14
This supports the hypothesis that
axonal loss quantified in the normal-appearing white matter results
from axons that traverse inflammatory lesions.
|
The Number of Degenerating Axons in a Lesion Correlates with Spinal Cord Atrophy and Neurological Deficits
Normally myelinated, remyelinated, and demyelinated
degenerating axons were frequently observed in the lesions of
192-day-infected SJL/J mice (Table 1
and Figure 3, A and B
). Axons were identified as
degenerating if their axoplasm contained numerous floccular dense
bodies, swelling mitochondria, and granular disintegration of
neurofilaments and vesicles (Figure 3, A and B
, single arrows). Some
degenerating axons also had a dark amorphous axoplasm (Figure 3B
,
double arrows). Myelin ovoids devoid of axoplasm were occasionally
engulfed by macrophages (Figure 3B
, asterisks). The number of
degenerating axons counted in the lesions of infected mice ranged from
94 to 469 (Table 1)
.
To determine whether degenerating axons correlated with other
pathological or functional variables, we assessed the relationships
between the number of degenerating axons and total spinal cord
demyelination, atrophy, and rotarod performance. No correlation was
observed between the percentage of total spinal cord demyelination and
the number of degenerating axons sampled from a single demyelinated
lesion (plot not shown). The C7 combined lateral and anterior column
area has been shown to correlate almost perfectly with the frequency of
medium and large myelinated axons (
4 µm2)
measured in the normal-appearing spinal cord white matter, suggesting
that this measure of spinal cord atrophy is a strong indicator of
axonal loss.14
In the present study, a near perfect
negative correlation was found between the C7 combined lateral and
anterior column area and the number of degenerating axons within a
single lesion (Figure 3C)
(r = -0.90,
P = 0.014). These data suggest that axonal degeneration
results in reduced spinal cord areas and are consistent with the
hypothesis that the C7 combined lateral and anterior column area is a
measure of spinal cord atrophy and axonal loss.
The rotarod is a sensitive measure of motor coordination and balance
and has been shown to correlate with measures of spinal atrophy and
axonal loss in 192-day-infected SJL/J mice.14
Interestingly, the number of degenerating axons within a single
demyelinated lesion correlated almost perfectly with rotarod
performance (Figure 3D)
(r = -0.91,
P = 0.005), demonstrating that quantification of the
number of degenerating axons within one lesion can predict the overall
neurological function in a 192-day-infected mouse. The number of
degenerating axons quantified in a lesion likely reflects both axonal
loss and lesion size, as a positive linear correlation approaching
statistical significance was obtained between lesion size (Table 1)
and
the number of degenerating axons (plot not shown)
(r = 0.73, P = 0.063).
Analysis of Medium and Large Axons within a Single Lesion Correlates with Total Lesion Load, Spinal Cord Atrophy, and Neurological Dysfunction
We hypothesized that the frequency of medium and large axons (4
µm2) in a lesion would serve as the strongest
indicator of other global pathological and functional variables,
because axonal fibers in this size category were preferentially
affected in the normal-appearing white matter of 192-day-infected
mice.14
Lesion axonal frequency distributions were
calculated for all fibers and normally myelinated, remyelinated, and
demyelinated fibers, separately (see Materials and Methods). The
percentage of total spinal cord demyelination serves as a measure of
total lesion load. A strong negative correlation was obtained between
lesion load and the frequency of all medium and large (4
µm2) axons (Figure 4A)
(r = -0.84,
P = 0.020). To determine which fibers were responsible
for this correlation, we plotted the percentage of total spinal cord
demyelination versus the frequency of medium and large
normally myelinated, remyelinated, and demyelinated fibers, separately.
Each was expressed as a function of the total number of normally
myelinated, remyelinated, and demyelinated fibers sampled,
respectively. For example, the frequency of medium and large normally
myelinated fibers was calculated by dividing the number of normally
myelinated fibers >4 µm2
by the total number
of normally myelinated fibers sampled. Thus, the frequency of medium
and large normally myelinated fibers only reflects normally myelinated
fibers, but not demyelinated or remyelinated fibers. The strongest
negative correlation was obtained between percentage of total spinal
cord demyelination and the frequency of medium and large (4
µm2) normally myelinated axons (Figure 4B)
(r = -0.91, P = 0.005).
Correlation coefficients did not reach statistical significance for
frequencies of medium and large remyelinated (r
= -0.67, P = 0.103) or demyelinated
(r = -0.37, P = 0.415) fibers
(plots not shown). These data suggest that the frequency of medium and
large normally myelinated fibers measured within a single lesion is a
powerful indicator of total lesion load in a 192-day-infected mouse,
and that involving more of the spinal cord white matter in the
demyelinating process results in greater reductions of medium and large
normally myelinated fibers.
|
4 µm2) axons assessed
within a single lesion (Figure 4C)
(r = 0.93,
P = 0.007). To determine which fibers were responsible
for this correlation, we plotted C7 combined lateral and anterior
column areas versus the frequency of medium and large
normally myelinated, remyelinated, and demyelinated fibers, separately.
A near perfect positive correlation was obtained between C7 combined
lateral and anterior column areas and the frequency of medium and large
remyelinated fibers (Figure 4D)
(r = 0.95,
P = 0.003). Correlation coefficients did not reach
statistical significance for the frequency of medium and large normally
myelinated (r = 0.58, P = 0.226)
or demyelinated (r = 0.72, P =
0.106) fibers (plots not shown). No significant linear relationships
were obtained if the small fibers (
4 µm2)
were used in any correlation. These data indicate that frequency of
medium and large remyelinated fibers measured from a single thoracic
demyelinated lesion reflects the degree of atrophy measured at a
different level of the spinal cord.
We hypothesized that the frequency of medium and large remyelinated
fibers within a lesion serves as a surrogate marker for the
preservation of medium and large axons in the entire cord. In other
words, higher frequencies of medium and large axon remyelinated fibers
within a lesion may signify less axonal loss in this size category
during the course of disease, and thus, less spinal cord atrophy.
Because spinal cord atrophy correlates with rotarod performance (a
measure of motor coordination), we tested this hypothesis by plotting
rotarod performance versus the frequency of medium and large
axons in the lesion. Interestingly, a positive correlation approaching
statistical significance was obtained between rotarod performance and
the frequency of all medium and large fibers in the lesion (Figure 4E)
(r = 0.75, P = 0.055). The
strongest correlation was found between rotarod performance and the
frequency of medium and large remyelinated fibers (Figure 4F)
(r = 0.79, P = 0.033).
Correlation coefficients did not reach statistical significance for the
frequencies of medium and large normally myelinated
(r = 0.44, P = 0.318) or
demyelinated (r = 0.53, P =
0.198) fibers (plots not shown). These data are consistent with the
hypothesis that the frequency of medium and large remyelinated axons
within a single lesion is an indicator of axonal loss, spinal cord
atrophy, and thus, neurological function. In further support of this
hypothesis, a strong negative correlation (plot not shown)
(r = -0.84, P = 0.018) was
obtained between the frequency of medium and large remyelinated axons
and the number of degenerating axons (Table 1)
within a lesion. This
indicates that increased frequencies of medium and large remyelinated
fibers are associated with less degenerating axons.
Alternatively, it is possible that the frequency of medium and large
remyelinated axons within a lesion represents functionally repaired
axons that aid in the preservation or maintenance of neurological
function. To test this possibility, we asked if any relationships
existed between the actual percentage of remyelinated axons in a lesion
(expressed as a function of the total number of fibers sampled) (Table 1)
and rotarod performance. No correlation was obtained between the
percentage of remyelinated axons within a lesion and rotarod
performance (plot not shown). Furthermore, the percentage of
remyelinated medium and large fibers (expressed as function of the
total number of medium and large fibers sampled) did not correlate with
rotarod performance (plot not shown). In concert, these data are
consistent with the former hypothesis that the frequency of medium to
large remyelinated fibers within a lesion (expressed as a function of
the total number of remyelinated fibers sampled) is a surrogate marker
for the survival of all medium to large fibers in the combined lateral
and anterior column area. These fibers seem to be required for the
maintenance of neurological function assessed by the rotarod.
Intraaxonal Mitochondria Are Increased in Demyelinated Lesions
Increases in intra-axonal mitochondria have been observed
previously in demyelinated lesions and could reflect CNS injury or
repair.20,23-27
We quantified the number of intra-axonal
mitochondria for normally myelinated, remyelinated, and demyelinated
axons in infected mice to gauge the overall health of axons within a
demyelinated lesion (Figure 5A)
.
Mitochondria quantified in the myelinated axons of sham-infected mice
were used as controls. Intra-axonal mitochondria were significantly
increased in normally myelinated (mean, 0.66 ± 0.03),
remyelinated (mean, 0.81 ± 0.05), and demyelinated (mean,
0.97 ± 0.06) fibers of 192-day-infected mice when compared to the
myelinated (mean, 0.39 ± 0.03) fibers of sham-infected controls
(Figure 5A)
. Furthermore, within the demyelinated lesions, increases in
intra-axonal mitochondria were associated with decreases in myelin
surrounding the axons. For example, the highest intra-axonal
mitochondria were observed in the demyelinated fibers followed by lower
numbers of mitochondria in remyelinated and normally myelinated fibers,
respectively.
|
(r = 0.89, P = 0.007) and
demyelinated (plot not shown) (r = 0.78,
P = 0.037) axons. These data suggest that involving
more of the total spinal cord white matter area in the demyelinating
process results in increased intra-axonal mitochondria quantified
within a single lesion. A strong negative correlation was also obtained
between the C7 combined lateral and anterior column area and the
intra-axonal mitochondria for demyelinated axons (Figure 5C)
(r = -0.86, P = 0.029).
These data suggest that the relative increases in mitochondria
within demyelinated axons of a single lesion are an accurate reflection
of the severity of spinal cord atrophy. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The results from this study support the hypothesis that the assessment
of axonal loss within the normal-appearing white matter represents the
secondary effects of axons that projected through inflammatory lesions.
This concept has also been described recently in a study of three
patients with large, solitary brain demyelinating lesions of the type
seen in early MS.28
One month after observing the lesion
in the patients, reductions in NAA (a measure of neuronal
integrity)29
were observed in the normal-appearing white
matter of the contralateral hemisphere that were homologous to those
observed in the lesions. In the present study, we observed a reduction
in the frequency of large fibers (10 µm2) in
the lesions of the anterolateral columns of 192-day-infected mice when
compared to sham-infected controls. Moreover, the reduction in the
frequency of large fibers observed in individual demyelinated lesions
was comparable to the reduction observed previously14
in
the normal-appearing white matter (63% versus 62%,
respectively). A small reduction (9%) in the frequency of medium
fibers (4 µm2
to 10
µm2) was observed that did not reach
statistical significance.
One of the most interesting findings from this study was the fact that although TMEV-induced demyelinating disease is heterogeneous, the frequencies of medium and large fibers calculated within a single demyelinated lesion served as a powerful indicator of disease course in individual 192-day-infected mice. By analyzing individual lesions located in the anterolateral columns, we could accurately estimate total lesion load, spinal cord atrophy, and neurological function. A strong negative correlation (r = -0.91, P = 0.005) was obtained between the percentage of total spinal cord demyelination and the frequency of medium and large myelinated axons within the lesion. The results suggest that the animals with the most total spinal cord demyelination had the greatest medium and large normally myelinated fiber loss during the course of disease. A near perfect positive correlation (r = 0.95, P = 0.003) was obtained between the C7 combined lateral and anterior column area and the frequency of medium and large remyelinated axons within the lesion. Interestingly, the number of degenerating axons within a single lesion also correlated negatively with C7 combined lateral and anterior column area (r = -0.90, P = 0.014). This confirms that the C7 combined lateral and anterior column area is a measure of spinal cord atrophy, and thus, axonal loss. These data are also in agreement with magnetic resonance imaging literature that use brain and spinal cord atrophy as a surrogate marker for axonal loss in MS.30,31
In the TMEV model of progressive demyelination, we hypothesize that the frequency of medium and large remyelinated axons remaining in a lesion serves as a strong indicator of the medium and large axon loss that occurred during the course of disease. We propose the following as an explanation. During the course of TMEV-induced demyelinating disease, some axons in all size categories are demyelinated. Based on our studies, the small axons are relatively preserved, whereas the medium and large demyelinated axons will either remain demyelinated, degenerate, or begin the process of repair (ie, remyelination). Degeneration will eventually result in decreased medium and large normally myelinated fibers assessed in the normal-appearing white matter. The medium and large axons that survive the demyelinating process and begin the process of repair will become remyelinated. Thus, we hypothesize that if greater frequencies of medium and large remyelinated axons are observed in the lesion, then less medium and large axons were lost during the course of disease. In support of this hypothesis, a strong negative correlation (r = -0.84, P = 0.018) was obtained between the frequency of medium and large remyelinated axons and the number of degenerating axons within a lesion, indicating that less axonal degeneration is associated with greater preservation of medium and large axons. Furthermore, it is also important to note that the frequency of medium and large remyelinated fibers correlated with a measure of motor coordination (r = 0.79, P = 0.033), demonstrating the role of these medium and large fibers in the maintenance of neurological function.
The mechanism(s) that results in the preferential loss of medium to large fibers is not completely understood. We propose two hypotheses to explain this preferential loss. In murine/human amyotropic lateral sclerosis only the largest caliber, neurofilament-rich axons are lost.32,33 Abnormal accumulations of neurofilaments are observed in the motor neurons that give rise to these axons.34,35 Deficits in slow36 and fast37 axonal transport precede motor neuron degeneration and clinical disease in two murine models of amyotropic lateral sclerosis. Based on these studies, it has been suggested that axonal transport may be especially crucial to the long, large caliber axons and disruptions may result in axonal strangulation of these large, neurofilament-rich axons.38 Therefore, disruptions in axonal transport resulting from demyelination and alterations in neurofilaments could explain the preferential loss of medium and large fibers in the TMEV model.
The second hypothesis for the preferential loss of medium and large fibers in the TMEV model involves intra-axonal calcium levels. Electron probe X-ray microanalysis has demonstrated that large CNS and peripheral nervous system axons have threefold to fourfold more intra-axonal calcium than small fibers.39 In vitro models of white matter injury have demonstrated that axonal damage can result from increases in intra-axonal calcium.40-44 Because large axons have a pre-existing high concentration of Ca2+, a threshold may be reached more rapidly, resulting in the activation of effector proteins45 that have the potential to cause axonal injury and degeneration. The role of this mechanism of axonal damage/loss has been demonstrated recently for spinal cord injury46 in which large fibers are preferentially lost.47-50
Mitochondria are essential for oxidative processes and energy utilization in axons, and their function may be disrupted by increases in intra-axonal calcium.51 Increases in intra-axonal mitochondria have frequently been observed by electron microscopy in injured axons.20,23-27 Interestingly, we found increased intra-axonal mitochondria in normally myelinated, remyelinated, and demyelinated axons of 192-day-infected mice when compared to the myelinated axons of sham-infected controls. The reason for increased mitochondria in injured axons is not completely understood, but may reflect oxidative stress,52 impaired axonal transport,25 attempts at restoring conduction,27 and/or axonal regeneration.53 Strong correlation coefficients were obtained between the percentage of total spinal cord demyelination and intra-axonal mitochondria for normally myelinated and demyelinated axons. This may signify that more stress is being placed on individual axons at multiple levels of the spinal cord. Furthermore, strong correlation coefficients were also obtained between the C7 combined lateral and anterior column area, the number of degenerating axons, and intra-axonal mitochondria for demyelinated axons. These data suggest that increases in intra-axonal mitochondria are at least in part an indicator of axonal injury, although alternative explanations are also plausible.
In summary, this is the first study to demonstrate in a model of progressive spinal cord demyelination that quantifying pathological variables at the ultrastructural level within a single demyelinated lesion can serve as a strong indicator of functional severity. A number of factors may induce axonal loss during the course of a progressive demyelinating disease. Future studies will investigate potential mechanisms that lead to axonal loss after progressive CNS demyelination and will use therapeutic interventions in an attempt to reduce this axonal loss and the subsequent irreversible neurological dysfunction.
|
Footnotes |
|---|
Supported by the National Institutes of Health (Grants RO1 NS24180 and RO1 NS32129), and the generous contributions of Mr. and Mrs. Eugene Applebaum and Ms. Kathryn Peterson. S. S. is supported by the Faculty of Medicine, Siriraj Hospital, Mahidol University, Thailand. D. B. M. is supported by a predoctoral NRSA from the National Institute of Mental Health (Grant 1F31ME12120). We also appreciate Ms. Dyana Saenz’s contribution to the project.
S. S. and D. B. M. contributed equally to the manuscript.
Accepted for publication July 10, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. D. Budde, M. Xie, A. H. Cross, and S.-K. Song Axial Diffusivity Is the Primary Correlate of Axonal Injury in the Experimental Autoimmune Encephalomyelitis Spinal Cord: A Quantitative Pixelwise Analysis J. Neurosci., March 4, 2009; 29(9): 2805 - 2813. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. C. Ferrari, A. M. Depino, F. Prada, N. Muraro, S. Campbell, O. Podhajcer, V. H. Perry, D. C. Anthony, and F. J. Pitossi Reversible Demyelination, Blood-Brain Barrier Breakdown, and Pronounced Neutrophil Recruitment Induced by Chronic IL-1 Expression in the Brain Am. J. Pathol., November 1, 2004; 165(5): 1827 - 1837. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. L. Oleszak, J. R. Chang, H. Friedman, C. D. Katsetos, and C. D. Platsoucas Theiler's Virus Infection: a Model for Multiple Sclerosis Clin. Microbiol. Rev., January 1, 2004; 17(1): 174 - 207. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. A. Dandekar, G. F. Wu, L. Pewe, and S. Perlman Axonal Damage Is T Cell Mediated and Occurs Concomitantly with Demyelination in Mice Infected with a Neurotropic Coronavirus J. Virol., July 1, 2001; 75(13): 6115 - 6120. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. D. Murray, D. B. McGavern, S. Sathornsumetee, and M. Rodriguez Spontaneous remyelination following extensive demyelination is associated with improved neurological function in a viral model of multiple sclerosis Brain, July 1, 2001; 124(7): 1403 - 1416. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
