close

Вход

Забыли?

вход по аккаунту

?

Degeneration of spinocerebellar neurons in amyotrophic lateral sclerosis.

код для вставкиСкачать
ORIGINAL ARTICLES
Degeneration of Spinocerebellar Neurons in
Amyotrophc Lateral Sclerosis
Celia Williams, BSc,* Mark A. Kozlowski, BS," David R. Hinton, MD,*tS and Carol A. Miller, MDVS
~~
~~~
~ _ _ _ _ _ _ _ ~
~~
The selective involvement of spinocerebellar neurons in sporadic amyotrophic lateral sclerosis was investigated using
two monoclonal antibodies that have neuronal subset specificity in human spinal cord. In normal control subjects,
monoclonal antibody 6A2 showed specificity for neurons of the dorsal nucleus of Clarke, the cells of origin of the
dorsal spinocerebellar tract. Immunoreactive neurons were also observed in locations corresponding to the central
cervical nucleus and spinal border region, containing neurons of the cervicospinocerebellar and ventral spinocerebellar tracts, respectively. The latter two neuronal subsets are indistinguishable from surrounding neurons when conventional histological stains are used. Antigen 6A2 was distributed on surfaces of neuronal somas and proximal neurites
and extended into the extracellular space. A second antibody, monoclonal antibody 44.1, labeled the cytoplasm of
neuronal somas and neurites, including all monoclonal antibody 6A2-reactive cells and alpha motoneurons. In spinal
cords of all 5 patients with amyotrophic lateral sclerosis, monoclonal antibody 6A2 reactivity in the majority of
spinocerebellar neurons was absent or localized to the somal cytoplasm, which still stained with monoclonal antibody
44.1. In more severely involved tissues, there was loss of some spinocerebellar neurons and a corresponding loss of
monoclonal antibody 44.1 reactivity. These findings confirm involvement of the spinal cord components of the spine
cerebellar system at all levels in sporadic amyotrophic lateral sclerosis and suggest that some surface molecules are
modified during the degenerative process.
Williams C, Kozlowski MA, Hinton DR, Miller CA. Degeneration of spinocerebellar neurons in
amyotrophic lateral sclerosis. Ann Neurol 1990;27:215-225
In amyotrophic lateral sclerosis (ALS), a disease of unknown cause, the upper and lower motoneurons selectively degenerate. Possible etiological agents include
toxic and trophic substances, viruses, or factors mediated through autoimmune and genetic mechanisms
[ l , 2). None of these is unique to the sporadic form of
ALS. Although most interest has focused on the alpha
motoneuron, Averback and Crocker suggested that
degeneration in patients with ALS also includes
neurons of the dorsal nucleus of Clarke. They reported loss of over one-third of the cells in this nucleus
in all 12 cases of sporadic ALS reviewed [3f. In their
study, only neurons of the dorsal nucleus of Clarke
were examined and standard. histological techniques
that do not distinguish neuronal subsets were used.
One limiting factor in the phenotypic characterization of spinal cord neurons has been the paucity of
molecular markers that identify neuronal subsets in the
central nervous system (CNS). Recently, Hinton and
colleagues described a panel of monoclonal antibodies
(MAbs) reactive with distinct neuronal subpopulations
throughout the CNS [4f. One of these MAbs has defined a subset of pyramidal neurons that undergoes
degeneration in patients with Alzheimer's disease [Sf.
In the gray matter of the spinal cord, another MAb
(44.1) reacts strongly with many neurons, including
the alpha motoneurons and their axons in the ventral
roots. We used this MAb in the spinal cords of patients with ALS to identify neurons heterotopically
located in white matter tracts subserving motor function 161. This observation suggests an abnormal developmental substrate in this disease.
Components of the human spinocerebellar system
have been identified by another MAb, 6A2. Neurons
of the dorsal nucleus of Clarke in the spinal cord, and
the globose and emboliform nuclei and Purkinje cells
in the cerebellum [4f are labeled by this MAb.
In the spinal cords of other mammals, the spinocerebellar system includes neurons in the central cervical
nucleus, the dorsal nucleus of Clarke, and the spinal
border region, which are the neurons of origin of the
cervical, dorsal, and ventral spinocerebellar tracts, respectively. Spinal border cells were initially described
in the monkey by Cooper and Sherrington [73. More
recently, the distribution of spinocerebellar neurons
throughout the entire length of the spinal cord has
From the *Department of Pathology, School of Medicine, University of Southern California, tLos Angeles County HospitalUniversity of Southern California Medical Center, h s Angeles, and
the ?-Department of Biology, California Institute of Technology,
Pasadena, CA.
Received Mar 8, 1989, and in revised form Jul 31. Accepted for
publication Jul 31, 1989.
Address correspondence fo Dr Miuer, University of Southern
California, School of Medicine, Depmment of pathology, 201 1
Zonal Avenue, McKibben Annex 345, Los Angeles, CA 90033.
been mapped in the rat, cat, and squirrel monkey by
retrograde transport of horseradish peroxidase [S, 91.
These cells receive type I afferents, as do motoneurons. To date, there is no evidence for direct synaptic
contact between the alpha motoneuron and cells of the
spinocerebellar system. With the paucity of specific
neuronal subset markers and the unavailability of physiological data, the distribution of these neurons in locations other than the dorsal nucleus of Clarke has not,
to our knowledge, been previously confirmed in humans.
In this study, using light and electron microscopy,
we used MAbs 6A2 and 44.1 immunocytochemically
to identify, in normal human spinal cord, neurons of
the spinocerebellar system, including the central cervical nucleus, the dorsal nucleus of Clarke, and the spinal
border cells. A comparison with tissues from patients
with ALS revealed involvement of the spinocerebellar
system and associated changes in alpha motoneurons.
Methods
M A b Production
The specificities of MAbs 6A2 and 44.1 have been previously described in some detail [47. MAbs 6A2 and 44.1
were prepared according to the method of Kohler and Milstein {lo}. MAb 6A2 was generated by using homogenates
of Drosophila melanagaster heads as primary immunogen { 1I],
while MAb 44.1 was prepared by an in vitro sensitization
technique { 127 using neuron-enriched fractions of frozen,
unfixed ventral gray matter from the lumbar region of human
spinal cord as the immunogen {4]. Supernatant fluids from
hybridoma cell lines were screened immunohistologicallyon
10-pm cryostat sections of unfixed human spinal cord, and
cell lines of interest were cloned by limiting dilution. MAbs
6A2 and 44.1 were both of the IgM class.
Tissue Preparation
Postmortem tissues were obtained from 5 patients with ALS
(age range, 44-70 years), 5 neurologically normal control
subjects (age range, 27-71 years), and 5 patients with Alzheimer’s Disease (AD) (age range, 69-92 years). Postmortem intervals averaged 5 hours for patients with ALS, 4
hours for patients with AD, and 8 hours for normal control
subjects.
The neuropathological diagnosis for each patient was
based on two observers’ independent assessments of 6pm-thick sections from formdin-fured, paraffin-embedded
tissue. Sections were stained with hematoxylin-eosin, the
Lux01 fast blue/periodic acid-Schiff (LFB/PAS) reaction, and
the Bodian stain. Cryostat sections (10 pm thick) of fresh
frozen tissues were also stained with Sudan black for evaluation of demyelination.
For immunostaining, cross sections of unfuted dorsal root
ganglia and spinal cord at cervical, thoracic, and lumbar levels
were examined. One-half-centimeter blocks of tissue were
frozen in liquid nitrogen-chilled isopentane for 20 to 30
seconds. The tissue was mounted on a bed of frozen embedding medium (OCT compound, Tissue Tek, Miles Inc, Elk216 Annals of Neurology Vol 27 No 3 March 1990
hart, IN) and 10-pm-thick sections were cut on a cryotome
and thaw mounted onto untreated microscope slides.
Immunocytochtmistry
Immunostaining on tissue sections was done by the avidinbiotin complex immunoperoxidase method. Cryostat sections (10 pm thick) were fixed in acetone and sequentially
incubated at room temperature with full-strength supernatant, biotinylated goat anti-mouse IgM diluted 1:200 in phosphate-buffered saline (PBS), avidin-biotin complex (Vector
Labs, Burlingame, CA), and then 3-amino-9-ethylcarbaole
(AEC) (0.04%) with hydrogen peroxide (0.015%). Sections
were washed with three changes of PBS (pH 7.4) between
each of the antibody incubations, and after exposure to AEC.
Some of the sections were counterstained with hematoxylin
prior to mounting.
For double staining, one antigen was identified by the avidin-biotin method and the second antigen by fluorescein
immunofluorescence. The avidin-biotin stain was as just described, except that the section went directly from the final
wash in PBS to incubation with the second primary antibody
for 30 minutes, followed by two rinses in PBS, incubation
with fluorescein-isothiocyanate-conjugatedgoat anti-mouse
IgM (Organon Teknika-Cappel, Westchester, PA) diluted
1:50 in PBS for 30 minutes, and two final rinses in PBS. The
sections were mounted in 90% glycerol in PBS containingpphenylenediamine [131 and viewed under direct bright-field
illumination, epifluorescence, or a combination of both. Because of the possibility that the avidin-biotin reaction product might block antigenic sites reactive with the second primary MAb, serial sections were stained using each primary
MAb alone or were double stained using a reversed order of
the primary antibodies.
A major limiting factor was the paucity of neurologically
normal control subjects available for the study that came to
autopsy within a brief postmortem interval, and where tissues were appropriately frozen to preserve antigen 6A2
immunoreactivity. Antigen preservation in control tissue
samples was checked by the reproducibility of MAb 6A2
staining in the dorsal nucleus of Clarke. Of the two MAbs
used, MAb 6A2 reactivity was the more sensitive to prolonged agonal hypoxia and postmortem autolysis in tissues
prepared after 12 hours postmortem (C. Williams, unpublished observation, 1988).Reactivity of both MAbs was lost
after prolonged formalin fixation, or with paraffin embedding.
UltrastructuralImmunocytochemistry
Thoracic segments of the spinal cord were dissected from
recently formalin-fixed spinal cord of normal control patients. The cord was cut at 2-mm intervals and postfixed in
4% paraformaldehyde and 0.1% glutaraldehyde in PBS for 1
hour. After washing in PBS, 50-pm-thick Vibratome (Technical Products, International, St Louis, MO) sections were
cut and incubated at 4°C overnight with either MAb 6A2
or 44.1 diluted 1:l with saponin (0.1%)-PBS (S-PBS). Sections were washed with PBS and incubated for 1 hour with
1:200 dilution of biotinylated anti-mouse IgM antibody
(Vector). Following a further washing with S-PBS, the sections were reacted with streptavidin-peroxidase conjugate
(Zymed, San Francisco, CA) for 30 minutes. After a final
wash with S-PBS, the sections were washed with PBS alone
and the reaction product visualized with diaminobenzidine
(0.05%) and hydrogen peroxidase (0.03%). The sections
were then washed with PBS and reacted with 2% osmium
tetroxide for 30 minutes. After two washes in PBS, the sections were postfixed for 15 minutes in PBS containing 2%
glutaraldehyde. To augment the signal, many of the sections
were further processed by the silver-gold intensification
method reported by Liposits and colleagues [14]. The sections were dehydrated and then flat-embedded in Epon
(Eponate 12, Ted. Pella Inc, Redding, CA). Thin plastic sections were cut after an initial semithin section and viewed on
a Zeiss EM-I0 (Carl Zeiss, New York, NY).
Results
Spinocerebellar System Is Defined by MAb 6A2
In the spinal cords of all control patients, MAb 6A2
stained the apparent surfaces of neuronal somas and
promixal dendrites. At the cervical, thoracic, and lumbar levels (Fig 1; A l , B1, Cl), reactivity was restricted
to neurons comprising the spinocerebellar system, including neurons of the central cervical nucleus (see Fig
1; A2), the dorsal nucleus of Clarke (see Fig 1; B2),
and the spinal border cells (see Fig 1; C2).
Neurons of the central cervical nucleus and the
spinal border cells were sparsely scattered in a rostralcaudal organization, rather than in a discrete nucleus,
with only one or two cells visible bilaterally in a single
section. In the lumbar region of the spinal cord, the
spinal border cells were in close proximity to the alpha
motoneurons (Fig 2A). In contrast, neurons of the
dorsal nucleus of Clarke composed a defined nuclear
group that could be observed in a single cross section
(see Fig 2B).
Antigen 6A2 Expression in Spinal Cord of Patients
with ALS
CNS tissues from 4 of 5 of the patients with ALS
showed striking changes in the spinocerebellar system.
In most neurons of the central cervical nucleus, dorsal
nucleus of Clark, and the spinal border region, cell
surface staining with MAb 6A2 (see Fig 1; A3, B3,
C3) was either partially lost or totally absent. In 2 cases
(Cases 1 and 5 ) (Table), neuronal loss was severe and
unilateral in the dorsal nucleus of Clarke and in one of
them (Case 5), unilateral loss was also noted in the
central cervical nucleus. A few spinocerebellar neurons in all three cord regions contained foci of immunopositive cytoplasmic material, interspersed among
lipofuscin granules (see Fig 1; 3B)-a feature never
found in control tissue samples.
Antigen 44.1-A Second Neuronal Marker
Neurons of the dorsal nucleus of Clarke reactive with
MAb 6A2 also stained with MAb 44.1 (see Fig 2D).
Double staining on single sections (levels T4-Tl2)
confirmed colocahzation of their corresponding antigens to over 80% of the same neurons. In control
tissues, labeling of the neuronal surfaces with MAb
6A2 and the somal cytoplasm with MAb 44.1 was
apparent in the dorsal nucleus of Clarke (Fig 3; A l ,
A2). In some sites in tissues from patients with ALS,
MAb 44.1 neuronal staining often remained intact, but
either antigen 6A2 was localized to the cytoplasm or
immunoreactivity was absent (see Fig 3; B1, B2). In no
instance either in control tissues or in ALS tissues was
MAb 6A2 reactivity present alone without MAb 44.1
staining. In severely involved sites devoid of MAb
6A2 reactivity, and with decreased numbers of MAb
44.1 reactive neurons, hematoxylin-eosin-stained sections confirmed an overall reduction in neurons (see
Fig 3; A3, B3).
In control spinal cords, MAb 44.1 was strongly reactive with the somal cytoplasm and neurites of large,
multipolar neurons in the ventral horn, characteristic
of the alpha motoneurons (see Fig 2C). Diffuse cytoplasmic reactivity was confirmed ultrastructurally (data
not shown). These neurons could be easily distinguished from the surrounding neuropil; they were not
stained with MAb 6A2. As shown in Figure 4, there
was predictable loss of MAb 44.1-reactive cells in anterior horns of ALS tissues, particularly in lumbar and
cervical regions. All ALS tissues showed moderate to
severe loss of alpha motoneurons.
MAb 44.1 reacted selectively with the large ganghon cells of the dorsal root ganglia of control and
ALS tissues. Random sampling of dorsal root gangha
did not reveal any obvious neuronal loss or change in
antigen 44.1 distribution in the latter group. None of
the ganglion cell somas were reactive with MAb
6A2 either in control or in ALS tissues (data not
shown).
Quantitative Assessment of Spinocerebellar Neuronal Loss
in ALS
In the control patients, staining with MAb 6A2 of
three thoracic regions (Tl-T3, T4-T8, T10-Tl2) revealed, bilaterally, from 10 to 40 positively staining
neurons per 10O-Fm length of cord in the dorsal nucleus of Clarke (Fig 5). The numbers of neurons were
progressively greater caudally with the most neurons
in the T10-Tl2 levels, in both control and ALS tissues.
To determine if there was actual loss of antigen
6A2-bearing neurons in the dorsal nucleus of Clarke,
or simply loss of immunoreactivity, the second neuron-specific marker, MAb 44.1, was used. At each
level, adjacent pairs of 10-pm-thick sections were
made at 100-pm intervals over a distance of 1 111111,and
stained with MAbs 6A2 and 44.1. Differences between counts of MAb 6A2 and 44.1 immunoreactive
neurons and in immunoreactivity between control and
Williams et
al:
Spinocerebellar Neurons in ALS 217
B
A
218 Annals of Neurology
Vol 27
No 3 March 1990
C
4F
~ n o c l ~ ~ ~ (MAb)
b o GA2-positive
d y
cells in s p i i l
cord. The topographical distributions of MAb 6A2-immunoreuctiue neurons (@) are represented in A1 (cervical),B l (thoracic),
and C l (lumbar segments). Each diagram is a composite of multiple segments: Cl-C4, TI -T12, and L3-Lj. Photomicrographs of each cord region are from the enclosed areas. ( A l ) The
central cervical nucleus. ( B l ) Dorsal nucleus of Clarke. ( C l )
Spinal border region. (A2,
B2, C2) In controls at all levels,
neurons show immunoreactivity around the peripheries of somas
and proximal neurites. In A2,the unstained, circular profile is
a corpus amylacea. (A3,B3, C3) Neurons in the tissues from
patients with amyotrophic lateral sclerosis show focal loss of surface reactiuity (arrows), and in some neurons (B3), the antigen
lies within the somal cytoplasm. A spinal border cell neuron
(C3) shows focal loss of antigen on the somal surface (arrows)
with reactivity remaining along one neurite (arrowheads). (Immunoperoxidase.) Bar = 40 pm.
Fig 2 . Monoclonal antibody (MAb) reactivity in the dorsal nucleus of Clarke and ventral horn of controlspinal cords. MAb
6A2 labels the somalsurfaces of neurons at low magnification.
(a) The spinal border region in the ventral horn of the lumbar
region of the spinal cord. (b) The dorsal nucleus of Clarke in
thoracic spinal cord. Note that in panel a the immunoreactive
spinal bordw cell (arrow) is adjacent to unlabeled motoneurons
(smallarrows). (c, d) Cytoplasm of neurons in the ventral horn
of the lumbar region of the spinal cord and the dorsal nucleus of
CLrke is immunoreactiue with MA6 44.1. (Immunoperoxidae.)
Bar = 100 b.m.
Williams et al: Spinocerebellar Neurons in ALS
219
Monoclonal Antibody 6A2 Positive Cells in the Central Cervical Nucleus and Spinal Border Regiona
Spinal Border Region L3-L5
Central Cervical Nucleus Cl-C4
Clinical
Duration
Case No. Age (yr) Diagnosis (yr)
Length (pm) of
Length (pm) of
Spinal Cord
MAb 6A2 Positive Spinal Cord
MAb 6A2 Positive
Section Evaluated Neurondl00 pm Section Evaluated Neurondl00 pm
~~
1
2
3
4
5
6
7
8
9
10
11
12
13
14
I5
70
60
44
64
44
31
69
27
71
61
69
92
79
83
78
ALS
ALS
ALS
ALS
ALS
Control
Control
Control
Control
Control
AD
AD
AD
AD
AD
1
2
2
3
12
-
5
7
8
15
16
1
0
2
NA
5
NA
21
NA
NA
NA
25
13
400
400
400
NA
800
NA
440
NA
NA
NA
610
400
400
400
NA
16
10
NA
880
100
NA
480
400
300
NA
460
330
400
720
350
NA
NA
160
1
Ob
NA
1
3
15
NA
13
9
28
6
9
NA
NA
7b
"At least 30 or more serial 10-pm-thick sections of each cervical or lumbar section of the spinal cord were stained with MAb 6A2. Note the
marked loss of MAb 6A2 reactivity in both spinal cord regions in tissues of patients with ALS (Cases 1-4), which were all of short clinical
duration. Relatively high counts obtained for Case 5 result from averaging cell counts from both sides of the cord, although loss was actually
severe unilaterally.
bLess than 20 serial sections were counted.
NA = not available; AD = Alzheimer's disease; ALS
=
amyotrophic lateral sclerosis; MAb = monoclonal antibody.
ALS groups were compared by analysis of variance. In
ALS tissues, there were consistently reduced numbers
of immunoreactive neurons at all thoracic levels. At
levels Tl-T3, T4-T8, and T10-Tl2, a significant decrease in the numbers of MAb 44.1-reactive neurons
( p < 0.001) represented 32%, 64%, and 46% of control values, respectively (see Fig 5). The numbers of
MAb 6A2-staining cells at these levels represented a
significant decrease ( p < 0.01) of 6096, 7996, and
58% of control values.
In the lumbar and cervical regions, the normally
scattered distribution of spinocerebellar neurons and
the paucity of neurons per section, even in control
tissues, allowed only general estimates of neuronal
loss. For semiquantitative studies, MAb 6A2-stained,
10-pm-thick sections of cervical and lumbar spinal
cord neurons were evaluated bilaterally. Immunostained neurons, up to 100 pm in diameter, could easily be seen on serial sections and were counted only
once. Spinal cord segments from 5 patients with ALS,
5 patients with AD, and 5 control subjects were evaluated, using as many as 80 serial sections from each
region of the cord. The number of MAb 6A2-reactive
neurons per 100 pm of tissue was calculated.
In the central cervical nucleus of all control subjects,
including those with Alzheimer's disease, there were
10 to 25 immunoreactive cells per 100 pm, and in the
220 Annals of Neurology Vol 27
N o 3 March 1990
spinal border region, a range of 6 to 28 neurons. As
shown in the Table, the average number of neurons
stained with MAb 6A2 in the central cervical nucleus
of patients with ALS represented only 9% of control
values, and in the lumbar region an average of 6% of
immunoreactive cells remained. Although the number
of immunoreactive neurons was small even in control
subjects, between 300 and 900 pm of tissue was evaluated in all but two of the twenty spinal cord segments
examined.
SpinocerebeIIar Tract Degeneration
Since there was an apparent loss of many spinocerebellar neurons, the spinocerebellar white matter tracts in
the spinal cord were also reviewed in cervical and thoracic regions from 4 patients with ALS (Cases 1, 3, 4,
and 5) but were not available from the fifth (Case 2). In
the 4 (see Table), there was mild, diffuse pallor of the
dorsal and ventral spinocerebellar tracts in LFB/PASstained sections and mild, diffusely reduced neurite
population in Bodian-stained preparations. Pallor was
confined to one side of the spinal cords from 2 patients
(Cases 1 and 5). Two (Cases 3 and 4) had positive
reaction with the Sudan black stain. In the lumbar region, the ventral spinocerebellar fibers did not form a
discrete tract and demyelination and axonal loss could
not be easily discerned.
A
Fzg 3. Thorucic mgion dthe qiznulcord dorsulnuchus o f
Ckwke: do#& stuining with monoc~nulanti6o&es(MA&
44.2 and 6Az.In the spinulcord ofu control (A I), su#aces o f
neumm are &-oratedLs/ MA6 6Az(immunopemdzse),whih
in the sume Jection (Az)the qtoj%&smzi stuznedwith MA6
44.2. (Immunofluorescence.)Bar = 40 pm. In the spinal cord
of a patient with amyotrophic lateral rclerosis (Bl), there is residual MAb 642 immunoreactiuity in the cytoplasm of one
neuron (arrows)while two neurons nearby (arrowheads)show
B
negligible immunoreactiuity.(B2)MAb 44.1 stains the cytoplasm of all three neurons. (Immunojluorescence.)Bar = 40 pm.
In u hematoxylin-eosin-stained section from control tissues (A3),
the dorsal nucleus of Clarke appears at low magn&cation as a
cluster o f neurons. In tissue from a patient with umyotrophic
lateral sclerosis (83) at the same segmental level, there aref w e r
neurons, and many remaining cells are shrunken. Bar = 20
CLm.
Williams et al: Spinocerebellar Neurons in ALS
221
1
I
Control
T1-3
T4-8
T10-12
ALs
T1-3
T4-8
T10-12
Fig 5 . Cell counts in the dorsal nucleus of Clarke: control versus
amyotrophic lateral sclerosis (ALS). Twenty or more 10-pmthick sections were cut, each 100 p m apart,from thoracic regions
of spinal cordsfrom 4 control subjects and 4 patients with ALS.
Neurons are immunoreactive with either monoclonal antibody
(MAb) 6A2 or MA6 44.1 in three regions: Tl-T3, T4-T8,
and TIO-T12. Results are expressed as the average numbers of
neuronsll mm, and each column represents countsfrom 4 patients. By analysis of variance, the number of MA6 44.1-immunoreactive neurons is significantly larger at all levels in both controlsubjects andpatients with ALS (p < 0.005) except in
control h e l s T4-T8 where a wide range of counts belween patients was observed. In the ALS tissues, there were significantly
reduced numbers of neurons immunoreactive with each MA6 at
all levels compared with controls (p < 0.01), but most mrked in
the mid and lower thoracic levels. Note that the caudul-most thoracic levels of the spinal cord contain the greatest total number of
cells in the nucleus. (0= MAb 6A2; El = MAb 44.1)
necessary to determine whether the processes were
synaptic terminals.
Fig 4. Lumbar region of the spinal cord: alpha motoneurons
stained with monoclonal antibody (MAb) 44.1. (a) MAb 44.1
stains the cytophsm of alpha motoneurons in control tissues. (b)
There is loss of immunoreactivity of these neurons in tissuesfrom
patients with amyotrophic lateral sclerosis. Note one cell with no
Jomal zmmunoredc?ivity (arrows). (Immunoperoxidaie.)Bar =
40 p m .
Ultrastructural Distribution of Antigen 6A2
In the dorsal nucleus of Clarke, antigen 6A2 was
localized to patches along the external surfaces of
neuronal somas and proximal dendrites (Fig 6). Immunoreactive material extended into the space surrounding the neurons and enwrapped processes that
project directly onto the somas (see Fig 6B). Sensitivity of antigen 6A2 to glutaraldehyde fixation and to
postmortem autolysis precluded optimal preservation
222 Annals of Neurology Vol 27 No 3 March 1990
Discussion
A spectrum of changes occurs in the spinocerebellar
system in spinal cords of patients with the sporadic
form of ALS, including significant loss of immunoreactivity of the two neuronal antigens studied and ultimately loss of the corresponding neurons. The remarkable specificity of MAb 6A2 in the CNS made it
possible to detect the putative cells of the ventral spinocerebellar and cervicospinocerebellar tracts which
are otherwise indistinguishable from other SubPoPulations of neurons when standard histological techniques
are used. Conclusive proof of the functional specificity
of these neurons requires degeneration studies or appropriate electrophysiological data. Our results suggest
that in sporadic ALS, these neurons, in addition to
those of the dorsal nucleus of Clarke, also identified by
MAb 6A2, undergo changes in surface-associated
molecules and are eventually lost. Loss of surface reactivity of the 6A2 antigen, increased accumulation of
lipofuscin, and cytoplasmic MAb 6A2 reactivity in
some cells suggest that internalization of antigen 6A2
in these neurons may occur during the degenerative
process. The apparent absence of direct synaptic contacts with the alpha motoneuron precludes a transsyn-
Fig 6. Thoracic region of the spinal cord: ultrastructurallocalization of antigen 6A2. Vibratomesections from the dorsal nucleus of Clarke were stained with monoclonal antibody (MAb)
6A2 ,$Y the streptavidin immunoperoxiduse technique. (a) The
somal suyface of a dorsal nucleus neuron shows focal clustering of
diaminobenzidine reaction product (arrows). The nucleus (N)
and cytoplasm, including lipofuscin granules (L), are not
stained. Bar = 10 pm. (b) A t higher magnification, a similar
cell shws reaction product, intensified ,$Y the silver-gold method,
as a granular deposit confined predominantly to the cell membrane ( M ) and extracellular matrix (ECM) suwounding the cell
and adjacent cell processes (P). Note the relative lack of stain in
the neuronal cytoplasm (C). Bar = 1 pm.
aptic basis for degeneration of the spinocerebellar
neurons.
The wide range in the numbers of immunoreactive
neurons counted in lumbar regions of control spinal
cords is not age related; however, relatively low counts
were obtained from patients with AD. One possibility
is that the spinal border cells are arranged in clusters
and thus, for a small portion of spinal cord, a large
sampling error would arise. A serial section study
throughout levels L3-L5 would address this issue.
MAb 6A2 also labels a subpopulation of nonpyramidal neurons in the hippocampus and neocortex, as
well as Betz cells in the primary motor cortex. In preliminary observations comparing ALS and control tissues, the Betz cells were found to be nonreactive in
ALS, while other cells in these regions remained immunoreactive (results not shown).
Degenerative changes in Clarke’s column and the
spinocerebellar tracts are common features of familial
motoneuron diseases 115-171, including the Azorean
disease complex, an autosomal dominant degeneration
of the motor system 1181. An apparently sporadic
form of motoneuron degeneration, having many features in common with Azorean disease complex, including involvement of the spinocerebellar neurons,
was recently reported 1191.
A unlfying mechanism for neurodegeneration involving the motor and spinocerebellar systems may be
provided by recent evidence linking an enviornmental
excitotoxin, P-N-methylamino-L-alanine(L-BMAA) to
the Guam-parkinsonism-dementia ALS complex. Although the pathological features differ from those of
sporadic ALS, the neuronal mechanisms of damage
may overlap. The toxic action of L-BMAA is mediated
by the N-methyl-D-aspartate (NMDA) receptor but
may also act, to some extent, through quisqualate/
kainate receptors, as it is attenuated but not completely obliterated by the NMDA antagonists AP7 or
MK807 11. Excitatory synapses from interneurons
impinging on the motoneuron may be of both NMDA
and quisqualatelkainate types and are mediated by Lglutamate or a glutamate-like compound 1201, as are
the fast postsynaptic potentials at the quisqualate/
kainate receptors of the group Ia afferent synapse 1211.
Cells receiving many group I afferent inputs, such as
the neurons of the dorsal nucleus of Clarke and other
spinocerebellar cells, may be susceptible to excitotoxic
damage, yet undergo a more protracted degeneration
than the motoneuron.
Degeneration in the spinocerebellar tract is less obvious than that observed in the corticospinal tract of
patients with ALS. However, the loss of 6A2 antigen
reactivity may precede neuronal demise and, thus, not
be a reliable indicator of tract viability. A slower rate
of degeneration in spinocerebellar neurons, compared
to motoneurons, as suggested already, would also exWilliams et al: Spinocerebellar Neurons in ALS
223
plain the less striking changes observed in the tracts
seen postmortem.
Our ultrastructural studies confirmed antigen 6A2
localization to the neuronal surface. We also verified
antigen localization to spinocerebellar neurons in the
murine spinal cord. At the ultrastructural level and
under more optimal conditions of preparation than in
human autopsy tissues, antigen 6A2 is externally localized to the somal surfaces (D.R. Hinton, unpublished observations, 1988). Several antigens of unknown molecular specificities have a similar surface
localization to neuronal somas and proximal neurites,
but differ in their topographical distributions in the
CNS. MAb Cat 301 stains surfaces of some medium
and large projection neurons in the mammalian spinal
cord r22, 231 and recognizes subpopulations of cells
throughout the CNS. MAb Tor 23 has a strong, but
not exclusive affinity for neurons of the motor system
in the mammalian CNS 1241. Specificity of MAb VC5
is restricted to a subset of gamma-aminobutyric acid
(GABA) neurons in visual cortex 1251.
Neuronal migration and pathfinding during development depend on surface molecules. Our recent observation of heterotopically located neurons in the spinal
cords of patients with ALS may suggest abnormalities
in these functions. It is of particular interest that the
abnormally located neurons are primarily within regions of motor function 161.
The close association of the 6A2 antigen with the
neuronal surface membrane, as well as sensitivity of
the epitope to deglycosidases and proteases in murine
CNS, suggests that it is a glycosylated protein (C. Williams, unpublished observation, 1989). Presence of the
HNK-1 epitope on the molecule has also been confirmed and infers that the 6A2 antigen could show
homology with cell adhesion molecules or the GM1
ganglioside which also bear this epitope. Glycosylated
surface determinants in motoneuron disease may be of
special importance. Polyclonal IgM antibodies that
bind to carbohydrate determinants on the GM1 and
GDlb gangliosides are present in the serum of 57% of
patients with ALS who were studied by Pestronk and
colleagues 1261.
The antigens identified by MAb 44.1 have been partially biochemically characterized in our laboratory.
MAb 44.1 recognizes a major polypeptide of 44 kDa
on Western immunoblots of homogenates of human
neural tissues 141. Determining the function of any of
the corresponding antigens depends on their future
isolation and molecular characterization.
Further confirmation of our findings in spinal cords
of additional patients with ALS is needed. Evaluation
of the brainstem and cerebellar components of the
spinocerebellar system in patients with ALS, as well as
in those with other degenerative diseases involving the
motor and spinocerebellar pathways, is also necessary
224 Annals of Neurology Vol 2 7 No 3 March 1990
to understand fully the specificity of the changes in
patients with ALS.
Our findings confirm the direct involvement of the
spinocerebellar system in sporadic ALS and offer a
possible molecular correlate with the pathogenetic
processes. Particular attention to the spinocerebellar
system may help further refine our definition of motoneuron diseases at the clinical and cellular levels.
This work was funded in part by the Muscular Dystrophy Association (to C.A.M.), the Barbara Vanderbilt Peck Foundation of the
Amyotrophic Lateral Sclerosis Society of America (to C.A.M.), the
National Institute of Mental Health (RO1-MH 39145, to C.A.M.),
the National Institute for Aging (P5O-AG 05142, to C.A.M.), the
Gordon Ross Foundation (to D.R.H.), the Medical Research Council of Canada (to D.R.H.), and the National Science Foundation
(NS-DSB-8409366, to Dr Benzer).
We gratefully acknowledge Dr Seymour Benzer for his generous gdt
of MAb 6A2 and critical review of the manuscript, and Dr Mariana
Linker-Israeli for assistance in the generation of MAh 44.1. We
thank Larry Reiter and Phillip Mora for their expert technical assistance, and Dr Linda Chan for statistical analysis. We would also like
to thank Dr Janet Blanks for the generous use of electron microscopy facilities funded by Grant EY03040 from the National Eye
Institute (a Core Center grant to the Doheny Eye Foundation).
References
1. Spenser PS, Nunn PB, Hugon J, et al. Guam amyotrophic lateral sclerosis-Parkinsonism-dementia linked to a plant excitant
neurotoxin. Science 1987;237:5 17-522
2. Rowland LP. Motor neuron diseases and amyotrophic lateral
sclerosis: research progress. Trends Neurosci 1987;10(10):393398
3. Averback P, Crocker P. Regular involvement of Clarke’s nucleus in sporadic amyotrophic lateral sclerosis. Arch Neurol
1982;39:155-156
4. Hinton DR, Henderson V W ,Blanks JC, et al. Monoclonal antibodies react with neuronal subpopulations in the human nervous system. J Comp Neurol 1988;267:398-408
5. Miller CA, Rudnicka M, Hinton DR, et al. Monoclonal antibody identification of subpopulations of cerebral cortical neurons affected in Alzheimer’s disease. Proc Natl Acad Sci USA
1987;84:8657-8661
6. Kozlowski MA, Williams C, Hinton DR, Miller CA. Heterotopic neurons in ALS spinal cord. Neurology 1989;39:644648
7. Cooper S, Sherrington CS. Gower’s tract and spinal border
cells. Brain 1940;63:123-134
8. Snyder RL,F a d RLM, Mehler WR. Comparative study of the
neurons of origin of the spinocerebellar afferents in the rat, cat,
and squirrel monkey based on the retrograde transport of horseradish peroxidase. J Comp Neurol 1978;181(4):833-852
9. Matsushita M, Hosoya Y,Ikeda M. Anatomical organization of
the spinocerebellar system of the cat, as studied by retrograde
transport of horseradish peroxidase. J Comp Neurol 1979;
184181-106
10. Kohler G, Milstein C. Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 1975;256:
495-497
11. Fujita SC, Zipursky SL, Benzer S, et al. Monoclonal antibodies
against the Drosophikz nervous system. Proc Natl Acad Sci USA
1982;79:7929-7933
12. Matthew WD, Patterson PH. The production of a monoclonal
antibody which blocks the action of a neurite outgrowth promoting factor. Cold Springs Harbor Symp Quant Biol 1983;
48(Pt 2):625-631
13. Johnson GD, de C Nogueira Araujo GM. A simple method of
reducing the fading of immunofluorescence during microscopy.
J Immunol Methods 1981;43:349-350
14. Liposits Z, Setalo G, Flerko B. Application of the silver-gold
intensified 3, 3’-diaminobenzidine chromogen to the light and
electron microscopic detection of the leuteinizing hormonereleasing hormone system of the rat brain. Neuroscience 1984;
13:513-525
15. Horton WA, Eldridge R, Brody JA. Familial motor neuron
disease. Neurology 1976;26:460-465
16. Tanaka J, Nakamura H, Tabuchi Y, Takahashi K. Familial
amyotrophic lateral sclerosis: features of multisystem degeneration. Acta Neuropathol 1984;64:22-29
17. Mulder DW, Kurland LT, Offord KP, Beard CM. Familial adult
motor neuron disease: amyotrophic lateral sclerosis. Neurology
1986;36:511-517
18. Rosenberg RN, Fowler HL. Autosomal dominant motor system
disease of the Portuguese: a review. Neurology 1984;31:11241126
19. McQuinn BA, Kemper TL. Sporadic case resembling autoso-
mal-dominant motor system degeneration (Azorean disease
complex). Arch Neurol 1987;44:341-344
20. OBrien RJ, Fischbach GD. Excitatory synaptic transmission between interneurons and motoneurons in chick spinal cord cell
cultures. J Neurosci 1986;6(11):3284-3289
2 1. Jessel TM, Yoshioka K, Jahr CE. Amino acid receptor-mediated
transmission at primary afferent synapses in rat spinal cord. J
Exp Biol 1986;124:239-258
22. McKay RDG, Hockfield SJ. Monoclonal antibodies distinguish
antigenically discrete neuronal types in the vertebrate central
nervous system. Proc Natl Acad Sci USA 1982;79:6747-6751
23. Hendry SHC, Jones EG, Hockfield S, McKay RDG. Neuronal
populations stained with monoclonal antibody Cat-301 in the
mammalian cerebral cortex and thalamus. J Neurosci 1988;
8(2):518-542
24. Stephenson DT, Kushner PD. An atlas of a rare neuronal surface antigen in the rat central nervous system. J Neurosci 1988;
8(8):3035-3056
25. Naegele JR, Arimatsu Y,Schwara P, Barnstable CJ. Selective
staining of a subset of GABAergic neurons in cat visual cortex
by monoclonal antibody VC1.l. J Neurosci 1988;8(1):79-89
26. Pestronk A, Adams N, Clawson L, et al. Serum antibodies to
GM1 ganglioside in amyotrophic lateral sclerosis. Neurology
(1988);38:1457-1461
Williams et al: Spinocerebellar Neurons in ALS
225
Документ
Категория
Без категории
Просмотров
1
Размер файла
3 807 Кб
Теги
degeneration, spinocerebellar, lateral, neurons, sclerosis, amyotrophic
1/--страниц
Пожаловаться на содержимое документа