- Email: [email protected]
Motor neurons are rich in non-phosphorylated neurofilaments: cross-species comparison and alterations in ALS
Brain Research 861 Ž2000. 45–58 www.elsevier.comrlocaterbres
Research report
Motor neurons are rich in non-phosphorylated neurofilaments: cross-species comparison and alterations in ALS Yee Man Tsang a , Freddie Chiong a , Daniel Kuznetsov a , Edward Kasarskis b , Changiz Geula a
a, )
Laboratory for NeurodegeneratiÕe and Aging Research, Department of Medicine, HarÕard Medical School and Section of Gerontology, Beth Israel Deaconess Medical Center, 21-27 Burlington AÕenue, P.O. Box 15709, Boston, MA 02215, USA b Department of Neurology, UniÕersity of Kentucky Medical School, Lexington, KY, USA Accepted 21 December 1999
Abstract The localization and distribution of non-phosphorylated neurofilaments ŽNP-NF. in the upper and lower motor neurons was investigated in the rat, the common marmoset, the rhesus monkey and man using the SMI-32 antibody. Within the spinal cord of all species studied, the most intense NP-NF immunoreactivity was observed within the ventral horn a-motor neurons. Concurrent staining for the cholinergic marker choline acetyltransferase ŽChAT. demonstrated that virtually all of the ChAT-positive a-motor neurons contain NP-NF immunoreactivity. Although NP-NF staining was also observed in other neurons within the ventral and intermediate horns, these neurons were loosely scattered and contained a considerably lower staining intensity. The only other prominent NP-NF staining in the spinal cord occurred within the neurons of the dorsal nucleus of Clark and the intermediolateral cell column. Phosphorylated neurofilament ŽP-NF. immunoreactivity was found primarily in neuronal processes. Occasionally, a solitary motor neuron contained weak P-NF immunoreactivity. Within the brainstem, neurons in all cranial nerve motor nuclei contained intense NP-NF immunoreactivity. The distribution and apparent density of NP-NF immunoreactive neurons in these nuclei was virtually identical to that observed for neurons immunoreactive for ChAT. NP-NF immunoreactive neurons of relatively lower intensity were found in many other regions of the brainstem. All of the giant Betz cells of layer ŽL. V in the motor cortex contained dark NP-NF immunoreactivity. Within the spinal cord of amyotrophic lateral sclerosis ŽALS. patients, both Nissl and NP-NF staining demonstrated the dramatic loss of a-motor neurons characteristic of this disorder. Some of the remaining motor neurons contained intense P-NF immunoreactivity. These observations suggest that NP-NF immunoreactivity is a good marker for motor neurons in health and disease and may be a useful tool for studies of motor neuron degeneration ŽMND.. q 2000 Elsevier Science B.V. All rights reserved. Keywords: Brainstem; Cholinergic; Choline acetyltransferase; Motor cortex; Phosphorylated neurofilament; Spinal cord
1. Introduction Skeletal muscles are under the direct influence of lower motor neurons within the brainstem Žcranial nerve motor nuclei. and the ventral horns of the spinal cord Ž a-motor neurons.. The lower motor neurons receive substantial input from the upper motor neurons in layer ŽL. V of motor cortex ŽBrodmann’s area 4. w4,26x. Damage to neurons in various parts of this system or the axons which emanate from them produces distinct motor deficits w1x. In addition, some components of this system are vulnerable to neurodegenerative processes. For example, a-motor neurons, some cranial nerve motor nuclei, and to a smaller ) Corresponding author. [email protected]
Fax:
q1-617-632-1828;
e-m ail:
extent the upper motor neurons are vulnerable to degeneration in motor neuron diseases such as amyotrophic lateral sclerosis ŽALS. w1,9x. Because of their vulnerability to damage and degeneration at multiple levels of the neuraxis, and their crucial role in the execution of behavioral responses, the upper and lower motor neurons have been subject to intense investigation. One problem encountered in these investigations has been the correct identification of motor neurons w6,8x. In some locations, such as the ventral horns of the spinal cord, motor neurons are not always confined to closed nuclear boundaries w4x and must be distinguished from neurons in adjacent regions. In the past, several markers have been utilized for the identification of motor neurons. The large size of these neurons can be used in Nissl preparations as one such
0006-8993r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 6 - 8 9 9 3 Ž 0 0 . 0 1 9 5 4 - 5
46
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
marker w4,23x. However, all neurons are enriched in Nissl substance and adjacent, large non-motor neurons may be mistakenly identified as motor neurons. Because of their cholinergic phenotype, the cholinergic markers choline acetyltransferase ŽChAT. and acetylcholinesterase ŽAChE. have been used to identify somatic motor neurons within the brainstem and spinal cord w6,8,23x. The synthetic cholinergic enzyme ChAT is specifically localized in cholinergic neurons w12x. However, this enzyme is exquisitely sensitive to fixatives and therefore difficult to visualize, particularly in formalin-fixed pathologic specimens ŽGeula, unpublished observations. w12,18x. The hydrolytic enzyme AChE is also present in all cholinergic neurons, but it is not a specific cholinergic marker and is found in a substantial population of non-cholinergic neurons within the ventral horns of spinal cord and in the brainstem w8,12,13x. Furthermore, the upper motor neurons are not cholinergic and are therefore devoid of ChAT immunoreactivity but contain a high concentration of AChE, particularly in the human w13,20x. Differentiating brainstem and ventral horn motor neurons in the chick have been shown to express the homeobox gene Islet-1 and to contain Islet-1 protein w10,27,35x. However, the presence of Islet-1 in adult motor neurons, particularly in the rat and the human, has not been established. A reliable marker of somatic motor neurons should have several characteristics. First, it should allow the identification of virtually all motor neurons. Second, it should be relatively resistant to variations in tissue parameters such as duration and mode of fixation, post-mortem interval, etc. Third, it should identify motor neurons in a range of species. Here we report that immunohistochemical staining for non-phosphorylated neurofilaments ŽNPNF., identified by the specific antibody SMI-32 w30,31x, can serve as a relatively reliable marker of motor neurons, satisfying all of the criteria listed above. The enrichment of a-motor neurons in NP-NF has important implications for ALS in which these neurons have been shown to contain inclusions of the phosphorylated form of neurofilaments ŽP-NF. w19,22,24x.
2. Materials and methods 2.1. Tissue preparation Four Sprague–Dawley rats Ž3 months old., four common marmosets Ž Callithrix jacchus, 2–8 years old., two
47
rhesus monkeys Ž Macaca mulata, 5 years old. and postmortem tissue from six neurologically normal human cases and three patients with the clinical and pathological symptoms of ALS were used in this study. Spinal cord, brainstem and motor cortex was available in all animals. Of the normal human cases, spinal cord was available in two and cortex and brainstem in four cases. From the ALS cases, only the lower cervicalrupper thoracic cord was available. Rats and marmosets were deeply anesthetized with an overdose of sodium pentobarbital Ž100 mgrkg, i.p.; after 12 mgrkg ketamine, i.m. in the marmoset. and perfused transcardially with 50 ml of saline, followed by 300 ml of 4% paraformaldehyde in 0.1 M sodium phosphate buffer and 100 ml of 10% sucrose in phosphate buffer. The rhesus monkeys Žanesthetized in the same manner as the marmoset. were perfused with 750 ml of saline, followed by 1.5–2 l of 4% paraformaldehyde and 1 l of 10% sucrose. The spinal cord and brain of each animal were removed and taken through sucrose gradients Žup to 30%. for cryoprotection. Spinal cord, brainstem and hemispheric blocks from the human cases were fixed by immersion in 4% paraformaldehyde for 30–40 h Žbrainstem and cortex. or in 10% formalin for 6 months to 3 years Žspinal cords.. All fixed human tissue was taken through sucrose gradients Ž10–40%. for cryoprotection. Whole spinal cords were separated into seven blocks Župper cervical, cervical enlargement, upper thoracic, midthoracic, lower thoracic, lumbar and sacral. and sectioned at 40 mm on a freezing microtome. The brainstem and whole hemispheres or hemispheric blocks were also sectioned at 40 mm. Adjacent series of 1 in 10–15 sections were saved in 0.1 M phosphate buffer containing 0.02% sodium azide. One series of each cut block or hemisphere was stained for Nissl substance using Cresyl violet and employed for delineation of cytoarchitectonic boundaries and quantitative determination of the number of Nisslstained putative motor neurons. 2.2. Immunohistochemistry Immunohistochemistry was performed according to the avidin–biotin–peroxidase complex ŽABC. method employing the Vectastain Elite ABC kit ŽVector Laboratories, Burlingame, CA.. Free floating sections were rinsed three times in 0.1 M phosphate-buffered saline ŽPBS., at pH 7.4. This rinse was repeated after every incubation step. Sections were treated with 0.4% Triton X-100 in PBS for 30 min at room temperature and then soaked for 1 h in the
Fig. 1. NP-NF staining of ventral horn motor neurons at the cervical enlargement ŽA, C, E and G. and thoracic ŽB, D, F, and H. levels of the human ŽA and B., rhesus monkey ŽC and D., marmoset ŽE and F. and the rat ŽG and H. spinal cord. All motor neurons display intense NP-NF immunoreactivity. Within cervical levels of the cord, both the lateral and medial Žarrows in A, C and E. motor groups are stained. NP-NF staining in other neurons within the spinal cord is of considerably lower intensity. Note that the intensity of staining within motor neurons is lowest in the rat ŽG and H. and greatest in the human ŽA and B.. Arrows in ŽB. and ŽD. point to the neurons of the dorsal nucleus of Clark which also display prominent NP-NF staining. The magnification in ŽA. – ŽE. is 35 = , in ŽF. is 85 = , and in ŽG. and ŽH. is 170 = .
48
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
carrier medium consisting of 10% normal goat serum and 0.1% Triton X-100 in PBS. Tissue was incubated in the primary antibody at appropriate dilutions for 24 h at 48C. Sections were then incubated in biotinylated goat secondary IgG Ž1r500. for 15 h, followed by the ABC complex Ž1r100. for 2 h. The resultant peroxidase labeling was visualized by incubating the sections in 0.005% diaminobenzidine ŽDAB. and 0.01% H 2 O 2 in 50 mM Tris– HCl ŽpH 7.6.. Following termination of the reaction by
rinsing in the Tris–HCl buffer, sections were mounted on slides, air-dried, dehydrated in graded alcohols, cleared in xylene and coverslipped under permount. Control sections were processed using non-specific IgG in place of the primary antibody or by omitting the primary antibody. A specific polyclonal antibody Žgift of Dr. Louis Hersh, University of Kentucky Medical School, 1r300–1r500. was used to visualize ChAT containing neurons. NP-NFcontaining neurons were visualized using the mono-
Fig. 2. NP-NF ŽA, C and E. and ChAT ŽB, D and F. immunoreactivities in adjacent sections of cervical enlargement ŽA and B. and lumbar enlargement ŽC and D. of the rhesus and cervical enlargement of the marmoset spinal cord ŽE and F.. In these and all other levels of the spinal cord, large ventral horn neurons stained for NP-NF showed an identical pattern of distribution and apparent density to the a-motor neurons immunoreactive for the cholinergic enzyme ChAT. Note that other NP-NF-positive neurons within the spinal cord display considerably less staining intensity Žarrows in A, C and E. when compared with motor neurons. Identical results were obtained in the rat. The magnification in ŽA. – ŽD. is 35 = and in ŽE. and ŽF. is 85 = .
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
clonal SMI-32 antibody ŽSternberger Immunochemicals, 1r5000–1r10 000.. The SMI-31 antibody was used to stain P-NF-containing structures ŽSternberger Immunochemicals, 1r5000–1r10 000.. As appropriate, one series of sections from each subject was stained for ChAT, NP-NF and P-NF. For concurrent visualization of ChAT and NP-NF within the same section, the double-immunohistochemical method of Levey et al. w17x was used. For this purpose, tissue sections were first processed for ChAT immunoreactivity as described above. After the development of the DAB brown reaction product, the tissue sections were processed for immunohistochemistry using the SMI-32 antibody, except the peroxidase labeling was visualized using benzidine dihydrochloride, which results in a granular blue reaction product. 2.3. QuantitatiÕe analysis Adjacent tissue sections stained for Nissl substance, ChAT and NP-NF, were subjected to careful microscopic
49
examination for qualitative analysis of the extent of overlap between these markers in motor neurons. Within the spinal cord sections, quantitative analysis of the degree of overlap between the above markers was carried out in two ways. First, the number of positive motor neuron profiles for each marker was counted in single-stained adjacent sections. For this purpose, ChAT- or Nissl Žin formalinfixed tissue.-stained sections were used to define the boundaries in the ventral horns within which motor neurons were located. An ocular grid placed in the eyepiece of a microscope was systematically moved through this defined area and the number of positive profiles within each grid counted. This analysis was carried out for all of the 1 in 10–15 series of sections at each level of the spinal cord and the percentage of ChAT- andror Nissl-stained a-motor neurons, which contained NP-NF immunoreactivity determined. Second, the percentage of ChAT-positive amotor neurons containing NP-NF immunoreactivity was determined in double-stained sections in the rhesus spinal cord using the same method of counting.
Fig. 3. Concurrent immunostaining for NP-NF and ChAT in the rhesus spinal cord demonstrated complete overlap between the two markers. In this photomicrograph, the presence of the smooth brown DAB product of ChAT immunoreactivity Žparticularly noticeable in dendrites. can be seen co-localized with the granular blue BDHC product of NP-NF immunoreactivity in motor neurons. NP-NF immunoreactivity was found in virtually all ChAT-positive a-motor neurons. Magnification is 300 = .
50
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
3. Results Immunohistochemistry, using the NP-NF antibody SMI-32, stained a substantial population of neurons throughout the brain. The majority of these were large projection neurons. NP-NF immunoreactivity was granular in appearance, was distributed throughout the cytoplasm and extended into dendrites. The intensity of staining in NP-NF-positive neurons displayed a systematic increase as the phylogenetic scale was ascended. Thus, NP-NF-positive neurons in the rat brain displayed the lightest staining relative to the other species examined. The lower primate, the common marmoset and the rhesus displayed progressively more intense NP-NF neuronal staining, and the human brain contained the most intense staining ŽFig. 1.. 3.1. Spinal cord Many neurons within the spinal cord of all species studied displayed NP-NF immunoreactivity. In the perfusion-fixed tissue Žrat, marmoset and rhesus., a prominent collection of neuronal processes Žprimarily dendrites. was also NP-NF-positive throughout the cord ŽFig. 1C–H.. Within the prolonged formalin-fixed human cords, NP-NF immunoreactivity in neuronal processes was less conspicuous ŽFig. 1A and B.. The most intense staining in all species studied was observed in collections of large neurons of the ventral horns with the characteristic morphology of a-motor neurons ŽFigs. 1–3.. The ventral root fibers travelling within the ventral white matter were also intensely NP-NF immunoreactive. Two other collections of neurons displayed moderate to high staining intensity. These were the intermediolateral cell column and the neurons of the Clark’s column within the thoracic and upper lumbar regions ŽFigs. 1 and 4.. Individual neurons with light to moderate staining intensity were scattered within the upper portion of the ventral horns, the intermediate gray and the dorsal gray matter of the cord ŽFigs. 1 and 2.. Consistent with their cholinergic phenotype, the ventral horn a-motor neurons and their axons within the ventral white matter and the ventral roots contained prominent ChAT immunoreactivity ŽFig. 2B, D and F.. ChAT immunoreactivity was also observed within the neurons of the intermediolateral cell column and scattered neurons within the more medial aspects of the intermediate horns, close to the central canal. Qualitative comparison of adjacent sections stained for ChAT and NP-NF revealed nearly identical distribution
51
and apparent density within the ventral horns. Both immunoreactivities were found within the large a-motor neurons ŽFig. 2A–F.. Quantitative determination of ChAT and NP-NF immunoreactivity in the ventral horns of the rat, marmoset and rhesus revealed that on average 82– 100% of the ChAT-positive ventral horn neurons also contained NP-NF immunoreactivity. A similar result was obtained comparing the numbers of a-motor neurons in Nissl preparations to NP-NF-positive neurons in the human formalin-fixed specimens Žin which ChAT staining was absent due to prolonged fixation.. Double immunohistochemistry for ChAT and NP-NF within the rhesus spinal cord resulted in concurrent localization of the above markers in virtually all a-motor neurons ŽFig. 3.. In double-stained sections, 92–95% of ChAT-positive ventral horn neurons also contained NP-NF immunoreactivity. In addition, 99% of the large ventral horn NP-NF-positive neurons contained ChAT immunoreactivity. 3.2. Brainstem A large population of NP-NF immunoreactive neurons was observed throughout the brainstem of all species examined. As with the spinal cord, immunoreactivity was also observed in the proximal neuronal processes. NP-NFpositive neurons were observed in most brainstem structures, including the superior colliculus, the reticular formation, the vestibular nuclei and the red nucleus. The most prominent NP-NF immunoreactivity, however, was observed within neurons of all of the cranial nerve motor nuclei ŽFig. 5A, C and E.. NP-NF-positive neurons were observed in the oculomotor Žincluding its Edinger–Westphal subdivision. ŽFig. 5E., trochlear, abducens ŽFig. 5B., trigeminal motor, facial, vagal motor and hypoglossal nuclei ŽFig. 5A.. Comparison with matching or adjacent Nissl-stained sections indicated that virtually all of the neurons within these cranial nerve nuclei were also NPNF-positive. A number of brainstem nuclei contained ChAT-positive neurons. Prominent staining was observed within the pedunculopontine and laterodorsal tegmental nuclei. In addition, all of the cranial nerve motor nuclei contained ChAT-positive neurons. Qualitative comparison of adjacent sections stained for ChAT and NP-NF revealed that virtually all of the ChAT-positive motor neurons within the brainstem also contained NP-NF immunoreactivity ŽFig. 5A–F..
Fig. 4. The dorsal nucleus of Clark ŽA–D. and the intermediolateral cell column ŽE and F. were the only other neuronal groups which displayed prominent NP-NF immunoreactivity within the spinal cord. ŽA. and ŽB. are examples of NP-NF immunoreactivity in the dorsal nucleus of Clark in the human spinal cord. ŽC. shows NP-NF staining within the same nucleus in the rhesus and ŽD. in the marmoset. In the lower species studied, including the rhesus shown here ŽE., the neurons of the intermediolateral cell column contained an intermediate intensity of NP-NF immunoreactivity. In contrast, the same neuronal group in the human brain ŽF. displayed intense NP-NF immunoreactivity. The magnification in ŽA. and ŽB. is 85 = and in ŽC. – ŽF. is 170 = .
52
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
53
3.3. Motor cortex
4. Discussion
Neurons containing NP-NF immunoreactivity were observed throughout the cortical mantle in the rat, marmoset, rhesus and man. The most prominent staining was observed within the cell bodies and proximal processes of pyramidal neurons in LIII and LV of neocortex and throughout limbic and paralimbic cortical zones. Consistent with this pattern of staining, neurons within LIII and LV of motor cortex ŽBrodmann’s area 4. were NP-NFpositive. In particular, intense NP-NF immunoreactivity was observed within the giant LV motor neurons ŽBetz cells in the primate. ŽFig. 5G and H. from which the corticospinal tract originates. Comparison with adjacent Nissl-stained sections revealed that nearly all of these giant neurons of motor cortex contained NP-NF immunoreactivity in the rat, marmoset, rhesus and human.
The results of the present study demonstrate the presence of a relatively high density of NP-NF immunoreactivity within virtually all motor neurons of the spinal cord, brainstem and motor cortex. Although neurons in some adjacent structures also contained NP-NF immunoreactivity, NP-NF staining in motor neurons was considerably more intense. NP-NF immunoreactivity in motor neurons was observed in all of the species examined. In addition, perikaryal NP-NF staining was obtained in motor neurons regardless of the mode and strength of fixation. These results indicate that the presence of intense NP-NF immunoreactivity can be used as a reliable marker of all motor neurons. Thus, immunohistochemical staining for NP-NF using the SMI-32 or other specific antibodies may prove a useful tool in experimental studies of motor neurons, particularly in cases where identification of lower motor neurons using ChAT is difficult due to variable fixation or absence of species-appropriate antibodies. A few recent reports had indicated staining for NP-NF in spinal cord motor neurons using the SMI-32 antibody in situ w6,14,23x. The extent of NP-NF staining in motor neurons, however, was unclear. The findings of the present experiment confirm and extend the results of the above studies, by showing that NP-NF immunoreactivity is present in virtually all of the lower and upper motor neurons. A few in vitro investigations had used the SMI-32 antibody to identify motor neurons in dissociated spinal cord neuronal cultures w5,6x. Our findings of NP-NF staining of moderate intensity within a subpopulation of non-motor spinal cord neurons suggest that this in vitro approach will most likely lead to the identification of non-motor as well as motor neurons.
3.4. P-NF protein Throughout the brain and spinal cord, P-NF immunoreactivity was found in axon-like processes ŽFig. 6A., including those within the white matter and major fiber tracts. P-NF immunoreactivity was rarely observed in neurons, and when present, was of relatively low intensity. Consistent with this pattern of staining, P-NF immunoreactivity was only rarely observed within the motor neurons of the spinal cord ŽFig. 6A., brainstem and motor cortex. 3.5. Amyotrophic lateral sclerosis Within the spinal cord specimens from the three ALS patients, a dramatic loss of a-motor neurons was apparent in Nissl-stained sections ŽFig. 7A and B.. Adjacent sections stained for NP-NF contained the same pattern and density of ventral horn motor neurons as the Nissl-stained sections ŽFig. 7C and D.. Other NP-NF-positive neuronal populations, such as those of the Clark’s column, appeared to be relatively well preserved in ALS cord when compared with normal specimens ŽFig. 7E–H.. However, unlike the normal spinal cords, some of the remaining ventral horn motor neurons in the ALS cord also contained immunoreactivity for P-NF ŽFig. 6B and C..
4.1. Functional significance NF are a major and specific structural element of neuronal cytoskeleton. They are composed of three subunits of approximately 68 Žlow; NFL., 145 Žmedium; NFM. and 200 kDa Žhigh; NFH. molecular weight. Each of these subunits is the product of a distinct gene. The NF is made of a core filament, composed of all three subunits, and projections, which are made up of NFM and NFH and project out perpendicularly from the core filament. Apposed projections from different core filaments connect to
Fig. 5. NP-NF ŽA, C and E. and ChAT ŽB, D and F. immunoreactivity in adjacent sections of brainstem at the level of the dorsal Žmotor. nucleus of vagus Župper nuclei. and hypoglossal nucleus ŽA and B. and the abducens nucleus ŽC and D. of the rhesus and the oculomotor nuclear complex of man ŽE and F.. Note that the distribution and apparent density of the two markers in spinal cord motor neurons is virtually identical, indicating the presence of NP-NF in almost all of these motor neurons. Arrow in ŽE. points to the autonomic division of the oculomotor nucleus ŽEdinger–Westphal.. Identical results were obtained in other cranial nerve motor nuclei of all species studied. Intense NP-NF immunoreactivity was present in the giant Betz cells Župper motor neurons. in all species examined, including the human ŽG. and rhesus ŽH. shown here. Note that smaller pyramidal neurons in the same cortical areas Žarrows in H. display considerably less intense immunoreactivity. The magnification in ŽA. – ŽD. is 35 = , in ŽE. and ŽF. is 20 = and in G and H is 85 = .
54
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
create cross-bridges. The structure thus created serves as cytoskeletal support for neurons w14,15,25x. NF, particularly the NFM and NFH subunits, are heavily phosphorylated ŽP-NF. w25,30x. P-NF are found preferentially within axons while NP-NF are observed primarily in neuronal perikarya and dendrites w14,31x. The SMI-32 and SMI-31 antibodies used in the present study recognize the NFM and NFH subunits within NP-NF and P-NF, respectively w31x. Previous work has shown that although most neurons stain positively for NP-NF when the SMI-32 antibody is used, large projection neurons, such as cortical pyramidal neurons are particularly enriched in NP-NF w2,21x. This observation indicates that large neurons, with their fast-conducting, large caliber axons require a high content of NF for structural support and integrity. In fact, accumulation of NF within axons has been shown to be associated with enlargement of axonal caliber during development w15,38x. Although qualitative in nature, our observations regarding the relative intensity of NP-NF immunoreactivity in different populations of neurons allows certain generalizations. Consistent with the above observations, we found motor neurons, which are among the largest neurons in the nervous system, to be highly enriched in NP-NF. Our observations indicate that the association between the gradient of staining intensity and the size of neurons can be detected in comparisons within and between various species. Thus, in each species studied, we found the motor neurons to display among the highest intensities of staining while adjacent and nearby neurons, which were smaller in size, displayed considerably less intense staining. In addition, we observed qualitatively increased NP-NF staining intensity in motor neurons as the phylogenetic scale was ascended. Motor neurons in the rat spinal cord, which are the smallest among the species studied, displayed the lowest staining intensity. In contrast, the motor neurons of the human spinal cord, which were the largest motor neurons observed, appeared to contain the highest staining intensity. These observations are consistent with the findings of Campbell and Morrison w3x, which indicated a strong positive correlation between intensity of NP-NF immunoreactivity and neuronal size. In agreement with the
primary localization of P-NF in axons, in normal brains we observed staining for P-NF mainly in axons. P-NF staining was virtually absent from neurons. Depletion of NF has been shown to result in the disruption and fragmentation of microtubules within neurons w32,33x. Since one of the primary functions of microtubules is axoplasmic transport of materials, the aforementioned finding suggests that NF are essential for such transport. Thus, in addition to providing structural support, the high content of NP-NF within motor neurons is likely to contribute to transport of nutrients and essential chemicals to the very long axons of these neurons. 4.2. Implications for motor neuron disease Our observations in ALS spinal cord must be regarded as preliminary since they are qualitative in nature and are based on three cases only. However, our findings do indicate that staining for NP-NF may provide a reliable means for assessing the extent of motor neuron loss in motor neuron diseases such as ALS. Within the spinal cord of ALS patients, we observed a marked loss of motor neurons when compared to normal controls. The results obtained with NP-NF immunostaining were similar to that obtained using the Nissl stain. Neurons in other spinal cord regions which also contain a relatively high intensity of NP-NF staining, such as those within the dorsal nucleus of Clark and the intermediolateral cell column were relatively well preserved in ALS tissue stained with either method. Recent evidence indicates that the high content of NF may be a source of liability for spinal cord motor neurons in ALS. Consistent with the results of this study, a number of investigations have found accumulations of P-NF in the surviving spinal cord motor neurons in ALS w19,22,24x. Such accumulations are thought to cause reduced content and disorganization of NF in motor axons leading to abnormalities in axonal cytoskeleton. Similar P-NF accumulations have been reported in spinal cord motor neurons of transgenic animals expressing mutations in superoxide dismutase 1 ŽSOD1. gene found in familial ALS w23,40x. Deletion of NFL in SOD1 transgenic mice has been shown to significantly slow the progression and extent of motor
Fig. 6. P-NF immunoreactivity within the ventral horn of thoracic sections of the spinal cord in a normal subject ŽA. and cases of ALS ŽB and C.. In the normal subject ŽA., virtually all P-NF immunoreactivity is confined to neuronal processes Žmostly axons.. Only rare and isolated neurons resembling motor neurons were found to contain a low intensity of staining Žarrow in A.. Identical results were obtained in all species studied. In ALS ventral horn, P-NF axons appeared to be disrupted ŽB and C. and a few of the remaining large neurons Žputative a-motor neurons. displayed intense P-NF immunoreactivity Žlarge arrow in B and C.. In some instances, such neurons seemed to contain P-NF-positive inclusions, as parts of the neuron were free of staining Žempty arrow in B.. In addition, the ventral horns of ALS cases contained P-NF-positive bodies Žsmall arrows in B and C. which are most likely axonal enlargements andror remaining debris from degenerated motor neurons. The magnification in ŽA. – ŽC. is 200 = . Fig. 7. Žp.12. The ventral horn of ALS patients ŽB and D; thoracic. displayed a marked loss of motor neurons when compared with normal subjects ŽA and C. in both Nissl-stained sections ŽA and B. and NP-NF-immunostained material ŽC and D.. Note that the remaining motor neurons are considerably smaller than normal. Other spinal cord neurons which contain a relatively high intensity of NP-NF immunoreactivity in the normal Žsee Fig. 3., such as the dorsal nucleus of Clark ŽE and G. and the intermediolateral cell column ŽF and H. were relatively well preserved in ALS in tissue processed for the visualization of Nissl substance ŽE and F. or NP-NF ŽG and H.. ŽG. and ŽH. are at a lower level of the cord as compared with ŽE. and ŽF.. The magnification in ŽA. – ŽH. is 170 = .
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
55
56
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
Fig. 7
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
neuron disorder w37x. Motor neuron degeneration ŽMND. mice which naturally develop motor neuron disease have been found to show a substantial acceleration of the disease process when expressing a NF-disrupting transgene w28x. Transgenic animals expressing a mutant NFL subunit w16x or overexpressing the wild type human NFH w7x or the wild type mouse NFL w39x have been shown to display relatively selective motor neuron death. Finally, mutations Žpolymorphisms. have been reported in the repeat motif of the NFH gene in a number of ALS cases w11,34x. Taken together, the evidence summarized above has been suggested to indicate that the enrichment of NF in motor neurons may be a natural risk factor for the vulnerability of these neurons in ALS. However, it should be noted that some studies have failed to detect polymorphisms in the NFH gene in large samples of ALS patients w29,36x. Furthermore, some groups of motor neurons, such as those in the oculomotor nucleus, which display intense NP-NF immunoreactivity, are relatively well preserved in ALS w1x.
w12x
w13x
w14x
w15x w16x
w17x
w18x
Acknowledgements We thank Nicholas Nagykery for expert technical assistance. This work is supported in part by the Milton Fund of the Harvard Medical School.
References w1x J.H. Adams, L.W. Duchen ŽEds.., Greenfield’s Neuropathology, Oxford Univ. Press, New York, 1992, 1557 pp. w2x M.J. Campbell, P.R. Hof, J.H. Morrison, A subpopulation of primate corticocortical neurons is distinguished by somatodendritic distribution of neurofilament protein, Brain Res. 539 Ž1991. 133–136. w3x M.J. Campbell, J.H. Morrison, Monoclonal antibody to neurofilament protein ŽSMI-32. labels a subpopulation of pyramidal neurons in the human and monkey neocortex, J. Comp. Neurol. 282 Ž1989. 191–205. w4x M.B. Carpenter, J. Sutin, Human Neuroanatomy, Williams and Wilkins, Baltimore, 1983, 872 pp. w5x S.G. Carriedo, H.Z. Yin, R. Lamberta, J.H. Weiss, In vitro kainate injury to large, SMI-32Žq. spinal neurons is Ca2q dependent, NeuroReport 6 Ž1995. 945–948. w6x S.G. Carriedo, H.Z. Yin, J.H. Weiss, Motor neurons are selectively vulnerable to AMPArkainate receptor-mediated injury in vitro, J. Neurosci. 16 Ž1996. 4069–4079. w7x F. Cote, J.F. Collard, J.P. Julien, Progressive neuropathy in transgenic mice expressing the human neurofilament heavy gene: a mouse model of amyotrophic lateral sclerosis, Cell 73 Ž1993. 35–46. w8x J.R. Delfs, D.M. Saroff, Y. Nishida, J. Friend, C. Geula, Effects of NMDA and its antagonists on ventral horn cholinergic neurons in organotypic roller tube spinal cord cultures, J. Neural Transm. 104 Ž1997. 31–51. w9x A. Eisen, C. Krieger, Pathogenic mechanisms in sporadic amyotrophic lateral sclerosis, Can. J. Neurol. Sci. 20 Ž1993. 286–296. w10x J. Ericson, S. Thor, T. Edlund, T.M. Jessell, T. Yamada, Early stages of motor neuron differentiation revealed by expression of homeobox gene Islet-1, Science 256 Ž1992. 1555–1560. w11x D.A. Figlewicz, A. Krizus, M.G. Martinoli, V. Meininger, M. Dib,
w19x
w20x
w21x
w22x
w23x
w24x
w25x w26x w27x
w28x
w29x
w30x w31x
57
G.A. Rouleau, J.P. Julien, Variants of the heavy neurofilament subunit are associated with the development of amyotrophic lateral sclerosis, Hum. Mol. Genet. 3 Ž1994. 1757–1761. C. Geula, M.-M. Mesulam, Cholinergic systems and related neuropathological predilection patterns in Alzheimer disease, in: R.D. Terry, R. Katzman, K.L. Bick ŽEds.., Alzheimer Disease, Raven Press, New York, 1994, pp. 263–291. C. Geula, M.-M. Mesulam, C.C. Kuo, H. Tokuno, Postnatal development of cortical acetylcholinesterase-rich neurons in the rat brain: permanent and transient patterns, Exp. Neurol. 134 Ž1995. 157–178. T. Gotow, J. Tanaka, Phosphorylation of neurofilament H subunit as related to arrangement of neurofilaments, J. Neurosci. Res. 37 Ž1994. 691–713. M.K. Lee, D.W. Cleveland, Neuronal intermediate filaments, Annu. Rev. Neurosci. 19 Ž1996. 187–217. M.K. Lee, J.R. Marszalek, D.W. Cleveland, A mutant neurofilament subunit causes massive, selective motor neuron death: implications for the pathogenesis of human motor neuron disease, Neuron 13 Ž1994. 975–988. A.I. Levey, J.P. Bolam, D.B. Rye, A.E. Hallanger, R.M. Demuth, M.-M. Mesulam, B.H. Wainer, A light and electron microscopic procedure for sequential double antigen localization using diaminobenzidine and benzidine dihydrochloride, J. Histochem. Cytochem. 34 Ž1986. 1449–1457. A.I. Levey, B.H. Wainer, D.B. Rye, E.J. Mufson, M.-M. Mesulam, Choline acetyltransferase-immunoreactive neurons intrinsic to rodent cortex and distinction from acetylcholinesterase-positive neurons, Neuroscience 13 Ž1984. 341–353. V. Manetto, N.H. Sternberger, G. Perry, L.A. Sternberger, P. Gambetti, Phosphorylation of neurofilaments is altered in amyotrophic lateral sclerosis, J. Neuropathol. Exp. Neurol. 47 Ž1988. 642–653. M.-M. Mesulam, C. Geula, Acetylcholinesterase-rich neurons of the human cerebral cortex: cytoarchitectonic and ontogenetic patterns of distribution, J. Comp. Neurol. 306 Ž1991. 193–220. M.-M. Mesulam, C. Geula, Differential distribution of a neurofilament protein epitope in acetylcholinesterase-rich neurons of human cerebral neocortex, Brain Res. 544 Ž1991. 169–173. H. Mizusawa, S. Matsumoto, S.H. Yen, A. Hirano, R.R. RojasCorona, H. Donnenfeld, Focal accumulation of phosphorylated neurofilaments within anterior horn cell in familial amyotrophic lateral sclerosis, Acta Neuropathol. 79 Ž1989. 37–43. B.M. Morrison, J.W. Gordon, M.E. Ripps, J.H. Morrison, Quantitative immunocytochemical analysis of the spinal cord in G86R superoxide dismutase transgenic mice: neurochemical correlates of selective vulnerability, J. Comp. Neurol. 373 Ž1996. 619–631. D.G. Munoz, C. Greene, D.P. Perl, D.J. Selkoe, Accumulation of phosphorylated neurofilaments in anterior horn motoneurons of amyotrophic lateral sclerosis patients, J. Neuropathol. Exp. Neurol. 47 Ž1988. 9–18. H.C. Pant, Veeranna, Neurofilament phosphorylation, Biochem. Cell Biol. 73 Ž1995. 575–592. G. Paxinos, The Rat Nervous System, Vol. 2, Academic Press, Australia, 1985, 362 pp. S.L. Pfaff, M. Mendelsohn, C.L. Stewart, T. Edlund, T.M. Jessell, Requirement for LIM homeobox gene Isl1 in motor neuron generation reveals a motor neuron-dependent step in interneuron differentiation, Cell 84 Ž1996. 309–320. J. Plummer, A. Peterson, A. Messer, Accelerated and widespread neuronal loss occurs in motor neuron degeneration Žmnd. mice expressing a neurofilament-disrupting transgene, Mol. Cell. Neurosci. 6 Ž1995. 532–543. K. Rooke, D.A. Figlewicz, F.Y. Han, G.A. Rouleau, Analysis of the KSP repeat of the neurofilament heavy subunit in familiar amyotrophic lateral sclerosis, Neurology 46 Ž1996. 789–790. L.A. Sternberger, Immunocytochemistry, Wiley, New York, 1986, 524 pp. L.A. Sternberger, N.H. Sternberger, Monoclonal antibodies distin-
58
w32x
w33x
w34x
w35x
w36x
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58 guish phosphorylated and nonphosphorylated forms of neurofilaments in situ, Proc. Natl. Acad. Sci. U.S.A. 80 Ž1983. 6126–6130. X T. Tashiro, R. Imai, Y. Komiya, Early effects of beta,beta -iminodipropionitrile on tubulin solubility and neurofilament phosphorylation in the axon, J. Neurochem. 63 Ž1994. 291–300. T. Tashiro, Y. Komiya, Impairment of cytoskeletal protein transport X due to aging or beta,beta -iminodipropionitrile intoxication in the rat sciatic nerve, Gerontology 40 Ž1994. 36–45, Suppl. J. Tomkins, P. Usher, J.Y. Slade, P.G. Ince, A. Curtis, K. Bushby, P.J. Shaw, Novel insertion in the KSP region of the neurofilament heavy gene in amyotrophic lateral sclerosis ŽALS., NeuroReport 9 Ž1998. 3967–3970. T. Tsuchida, M. Ensini, S.B. Morton, M. Baldassare, T. Edlund, T.M. Jessell, S.L. Pfaff, Topographic organization of embryonic motor neurons defined by expression of LIM homeobox genes, Cell 79 Ž1994. 957–970. J.D. Vechio, L.I. Bruijn, Z. Xu, R.H. Brown, D.W. Cleveland, Sequence variants in human neurofilament proteins: absence of linkage to familial amyotrophic lateral sclerosis, Ann. Neurol. 40 Ž1996. 603–610.
w37x T.L. Williamson, L.I. Bruijn, Q. Zhu, K.L. Anderson, S.D. Anderson, J.P. Julien, D.W. Cleveland, Absence of neurofilaments reduces the selective vulnerability of motor neurons and slows disease caused by a familial amyotrophic lateral sclerosis-linked superoxide dismutase 1 mutant, Proc. Natl. Acad. Sci. U.S.A. 95 Ž1998. 9631– 9636. w38x T.L. Williamson, J.R. Marszalek, J.D. Vechio, L.I. Bruijn, M.K. Lee, Z. Xu, R.H. Brown Jr., D.W. Cleveland, Neurofilaments, radial growth of axons, and mechanisms of motor neuron disease, Cold Spring Harbor Symp. Quant. Biol. 61 Ž1996. 709–723. w39x Z. Xu, L.C. Cork, J.W. Griffin, D.W. Cleveland, Increased expression of neurofilament subunit NF-L produces morphological alterations that resemble the pathology of human motor neuron disease, Cell 73 Ž1993. 23–33. w40x B. Zhang, P. Tu, F. Abtahian, J.Q. Trojanowski, V.M. Lee, Neurofilaments and orthograde transport are reduced in ventral root axons of transgenic mice that express human SOD1 with a G93A mutation, J. Cell Biol. 139 Ž1997. 1307–1315.
Research report
Motor neurons are rich in non-phosphorylated neurofilaments: cross-species comparison and alterations in ALS Yee Man Tsang a , Freddie Chiong a , Daniel Kuznetsov a , Edward Kasarskis b , Changiz Geula a
a, )
Laboratory for NeurodegeneratiÕe and Aging Research, Department of Medicine, HarÕard Medical School and Section of Gerontology, Beth Israel Deaconess Medical Center, 21-27 Burlington AÕenue, P.O. Box 15709, Boston, MA 02215, USA b Department of Neurology, UniÕersity of Kentucky Medical School, Lexington, KY, USA Accepted 21 December 1999
Abstract The localization and distribution of non-phosphorylated neurofilaments ŽNP-NF. in the upper and lower motor neurons was investigated in the rat, the common marmoset, the rhesus monkey and man using the SMI-32 antibody. Within the spinal cord of all species studied, the most intense NP-NF immunoreactivity was observed within the ventral horn a-motor neurons. Concurrent staining for the cholinergic marker choline acetyltransferase ŽChAT. demonstrated that virtually all of the ChAT-positive a-motor neurons contain NP-NF immunoreactivity. Although NP-NF staining was also observed in other neurons within the ventral and intermediate horns, these neurons were loosely scattered and contained a considerably lower staining intensity. The only other prominent NP-NF staining in the spinal cord occurred within the neurons of the dorsal nucleus of Clark and the intermediolateral cell column. Phosphorylated neurofilament ŽP-NF. immunoreactivity was found primarily in neuronal processes. Occasionally, a solitary motor neuron contained weak P-NF immunoreactivity. Within the brainstem, neurons in all cranial nerve motor nuclei contained intense NP-NF immunoreactivity. The distribution and apparent density of NP-NF immunoreactive neurons in these nuclei was virtually identical to that observed for neurons immunoreactive for ChAT. NP-NF immunoreactive neurons of relatively lower intensity were found in many other regions of the brainstem. All of the giant Betz cells of layer ŽL. V in the motor cortex contained dark NP-NF immunoreactivity. Within the spinal cord of amyotrophic lateral sclerosis ŽALS. patients, both Nissl and NP-NF staining demonstrated the dramatic loss of a-motor neurons characteristic of this disorder. Some of the remaining motor neurons contained intense P-NF immunoreactivity. These observations suggest that NP-NF immunoreactivity is a good marker for motor neurons in health and disease and may be a useful tool for studies of motor neuron degeneration ŽMND.. q 2000 Elsevier Science B.V. All rights reserved. Keywords: Brainstem; Cholinergic; Choline acetyltransferase; Motor cortex; Phosphorylated neurofilament; Spinal cord
1. Introduction Skeletal muscles are under the direct influence of lower motor neurons within the brainstem Žcranial nerve motor nuclei. and the ventral horns of the spinal cord Ž a-motor neurons.. The lower motor neurons receive substantial input from the upper motor neurons in layer ŽL. V of motor cortex ŽBrodmann’s area 4. w4,26x. Damage to neurons in various parts of this system or the axons which emanate from them produces distinct motor deficits w1x. In addition, some components of this system are vulnerable to neurodegenerative processes. For example, a-motor neurons, some cranial nerve motor nuclei, and to a smaller ) Corresponding author. [email protected]
Fax:
q1-617-632-1828;
e-m ail:
extent the upper motor neurons are vulnerable to degeneration in motor neuron diseases such as amyotrophic lateral sclerosis ŽALS. w1,9x. Because of their vulnerability to damage and degeneration at multiple levels of the neuraxis, and their crucial role in the execution of behavioral responses, the upper and lower motor neurons have been subject to intense investigation. One problem encountered in these investigations has been the correct identification of motor neurons w6,8x. In some locations, such as the ventral horns of the spinal cord, motor neurons are not always confined to closed nuclear boundaries w4x and must be distinguished from neurons in adjacent regions. In the past, several markers have been utilized for the identification of motor neurons. The large size of these neurons can be used in Nissl preparations as one such
0006-8993r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 6 - 8 9 9 3 Ž 0 0 . 0 1 9 5 4 - 5
46
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
marker w4,23x. However, all neurons are enriched in Nissl substance and adjacent, large non-motor neurons may be mistakenly identified as motor neurons. Because of their cholinergic phenotype, the cholinergic markers choline acetyltransferase ŽChAT. and acetylcholinesterase ŽAChE. have been used to identify somatic motor neurons within the brainstem and spinal cord w6,8,23x. The synthetic cholinergic enzyme ChAT is specifically localized in cholinergic neurons w12x. However, this enzyme is exquisitely sensitive to fixatives and therefore difficult to visualize, particularly in formalin-fixed pathologic specimens ŽGeula, unpublished observations. w12,18x. The hydrolytic enzyme AChE is also present in all cholinergic neurons, but it is not a specific cholinergic marker and is found in a substantial population of non-cholinergic neurons within the ventral horns of spinal cord and in the brainstem w8,12,13x. Furthermore, the upper motor neurons are not cholinergic and are therefore devoid of ChAT immunoreactivity but contain a high concentration of AChE, particularly in the human w13,20x. Differentiating brainstem and ventral horn motor neurons in the chick have been shown to express the homeobox gene Islet-1 and to contain Islet-1 protein w10,27,35x. However, the presence of Islet-1 in adult motor neurons, particularly in the rat and the human, has not been established. A reliable marker of somatic motor neurons should have several characteristics. First, it should allow the identification of virtually all motor neurons. Second, it should be relatively resistant to variations in tissue parameters such as duration and mode of fixation, post-mortem interval, etc. Third, it should identify motor neurons in a range of species. Here we report that immunohistochemical staining for non-phosphorylated neurofilaments ŽNPNF., identified by the specific antibody SMI-32 w30,31x, can serve as a relatively reliable marker of motor neurons, satisfying all of the criteria listed above. The enrichment of a-motor neurons in NP-NF has important implications for ALS in which these neurons have been shown to contain inclusions of the phosphorylated form of neurofilaments ŽP-NF. w19,22,24x.
2. Materials and methods 2.1. Tissue preparation Four Sprague–Dawley rats Ž3 months old., four common marmosets Ž Callithrix jacchus, 2–8 years old., two
47
rhesus monkeys Ž Macaca mulata, 5 years old. and postmortem tissue from six neurologically normal human cases and three patients with the clinical and pathological symptoms of ALS were used in this study. Spinal cord, brainstem and motor cortex was available in all animals. Of the normal human cases, spinal cord was available in two and cortex and brainstem in four cases. From the ALS cases, only the lower cervicalrupper thoracic cord was available. Rats and marmosets were deeply anesthetized with an overdose of sodium pentobarbital Ž100 mgrkg, i.p.; after 12 mgrkg ketamine, i.m. in the marmoset. and perfused transcardially with 50 ml of saline, followed by 300 ml of 4% paraformaldehyde in 0.1 M sodium phosphate buffer and 100 ml of 10% sucrose in phosphate buffer. The rhesus monkeys Žanesthetized in the same manner as the marmoset. were perfused with 750 ml of saline, followed by 1.5–2 l of 4% paraformaldehyde and 1 l of 10% sucrose. The spinal cord and brain of each animal were removed and taken through sucrose gradients Žup to 30%. for cryoprotection. Spinal cord, brainstem and hemispheric blocks from the human cases were fixed by immersion in 4% paraformaldehyde for 30–40 h Žbrainstem and cortex. or in 10% formalin for 6 months to 3 years Žspinal cords.. All fixed human tissue was taken through sucrose gradients Ž10–40%. for cryoprotection. Whole spinal cords were separated into seven blocks Župper cervical, cervical enlargement, upper thoracic, midthoracic, lower thoracic, lumbar and sacral. and sectioned at 40 mm on a freezing microtome. The brainstem and whole hemispheres or hemispheric blocks were also sectioned at 40 mm. Adjacent series of 1 in 10–15 sections were saved in 0.1 M phosphate buffer containing 0.02% sodium azide. One series of each cut block or hemisphere was stained for Nissl substance using Cresyl violet and employed for delineation of cytoarchitectonic boundaries and quantitative determination of the number of Nisslstained putative motor neurons. 2.2. Immunohistochemistry Immunohistochemistry was performed according to the avidin–biotin–peroxidase complex ŽABC. method employing the Vectastain Elite ABC kit ŽVector Laboratories, Burlingame, CA.. Free floating sections were rinsed three times in 0.1 M phosphate-buffered saline ŽPBS., at pH 7.4. This rinse was repeated after every incubation step. Sections were treated with 0.4% Triton X-100 in PBS for 30 min at room temperature and then soaked for 1 h in the
Fig. 1. NP-NF staining of ventral horn motor neurons at the cervical enlargement ŽA, C, E and G. and thoracic ŽB, D, F, and H. levels of the human ŽA and B., rhesus monkey ŽC and D., marmoset ŽE and F. and the rat ŽG and H. spinal cord. All motor neurons display intense NP-NF immunoreactivity. Within cervical levels of the cord, both the lateral and medial Žarrows in A, C and E. motor groups are stained. NP-NF staining in other neurons within the spinal cord is of considerably lower intensity. Note that the intensity of staining within motor neurons is lowest in the rat ŽG and H. and greatest in the human ŽA and B.. Arrows in ŽB. and ŽD. point to the neurons of the dorsal nucleus of Clark which also display prominent NP-NF staining. The magnification in ŽA. – ŽE. is 35 = , in ŽF. is 85 = , and in ŽG. and ŽH. is 170 = .
48
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
carrier medium consisting of 10% normal goat serum and 0.1% Triton X-100 in PBS. Tissue was incubated in the primary antibody at appropriate dilutions for 24 h at 48C. Sections were then incubated in biotinylated goat secondary IgG Ž1r500. for 15 h, followed by the ABC complex Ž1r100. for 2 h. The resultant peroxidase labeling was visualized by incubating the sections in 0.005% diaminobenzidine ŽDAB. and 0.01% H 2 O 2 in 50 mM Tris– HCl ŽpH 7.6.. Following termination of the reaction by
rinsing in the Tris–HCl buffer, sections were mounted on slides, air-dried, dehydrated in graded alcohols, cleared in xylene and coverslipped under permount. Control sections were processed using non-specific IgG in place of the primary antibody or by omitting the primary antibody. A specific polyclonal antibody Žgift of Dr. Louis Hersh, University of Kentucky Medical School, 1r300–1r500. was used to visualize ChAT containing neurons. NP-NFcontaining neurons were visualized using the mono-
Fig. 2. NP-NF ŽA, C and E. and ChAT ŽB, D and F. immunoreactivities in adjacent sections of cervical enlargement ŽA and B. and lumbar enlargement ŽC and D. of the rhesus and cervical enlargement of the marmoset spinal cord ŽE and F.. In these and all other levels of the spinal cord, large ventral horn neurons stained for NP-NF showed an identical pattern of distribution and apparent density to the a-motor neurons immunoreactive for the cholinergic enzyme ChAT. Note that other NP-NF-positive neurons within the spinal cord display considerably less staining intensity Žarrows in A, C and E. when compared with motor neurons. Identical results were obtained in the rat. The magnification in ŽA. – ŽD. is 35 = and in ŽE. and ŽF. is 85 = .
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
clonal SMI-32 antibody ŽSternberger Immunochemicals, 1r5000–1r10 000.. The SMI-31 antibody was used to stain P-NF-containing structures ŽSternberger Immunochemicals, 1r5000–1r10 000.. As appropriate, one series of sections from each subject was stained for ChAT, NP-NF and P-NF. For concurrent visualization of ChAT and NP-NF within the same section, the double-immunohistochemical method of Levey et al. w17x was used. For this purpose, tissue sections were first processed for ChAT immunoreactivity as described above. After the development of the DAB brown reaction product, the tissue sections were processed for immunohistochemistry using the SMI-32 antibody, except the peroxidase labeling was visualized using benzidine dihydrochloride, which results in a granular blue reaction product. 2.3. QuantitatiÕe analysis Adjacent tissue sections stained for Nissl substance, ChAT and NP-NF, were subjected to careful microscopic
49
examination for qualitative analysis of the extent of overlap between these markers in motor neurons. Within the spinal cord sections, quantitative analysis of the degree of overlap between the above markers was carried out in two ways. First, the number of positive motor neuron profiles for each marker was counted in single-stained adjacent sections. For this purpose, ChAT- or Nissl Žin formalinfixed tissue.-stained sections were used to define the boundaries in the ventral horns within which motor neurons were located. An ocular grid placed in the eyepiece of a microscope was systematically moved through this defined area and the number of positive profiles within each grid counted. This analysis was carried out for all of the 1 in 10–15 series of sections at each level of the spinal cord and the percentage of ChAT- andror Nissl-stained a-motor neurons, which contained NP-NF immunoreactivity determined. Second, the percentage of ChAT-positive amotor neurons containing NP-NF immunoreactivity was determined in double-stained sections in the rhesus spinal cord using the same method of counting.
Fig. 3. Concurrent immunostaining for NP-NF and ChAT in the rhesus spinal cord demonstrated complete overlap between the two markers. In this photomicrograph, the presence of the smooth brown DAB product of ChAT immunoreactivity Žparticularly noticeable in dendrites. can be seen co-localized with the granular blue BDHC product of NP-NF immunoreactivity in motor neurons. NP-NF immunoreactivity was found in virtually all ChAT-positive a-motor neurons. Magnification is 300 = .
50
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
3. Results Immunohistochemistry, using the NP-NF antibody SMI-32, stained a substantial population of neurons throughout the brain. The majority of these were large projection neurons. NP-NF immunoreactivity was granular in appearance, was distributed throughout the cytoplasm and extended into dendrites. The intensity of staining in NP-NF-positive neurons displayed a systematic increase as the phylogenetic scale was ascended. Thus, NP-NF-positive neurons in the rat brain displayed the lightest staining relative to the other species examined. The lower primate, the common marmoset and the rhesus displayed progressively more intense NP-NF neuronal staining, and the human brain contained the most intense staining ŽFig. 1.. 3.1. Spinal cord Many neurons within the spinal cord of all species studied displayed NP-NF immunoreactivity. In the perfusion-fixed tissue Žrat, marmoset and rhesus., a prominent collection of neuronal processes Žprimarily dendrites. was also NP-NF-positive throughout the cord ŽFig. 1C–H.. Within the prolonged formalin-fixed human cords, NP-NF immunoreactivity in neuronal processes was less conspicuous ŽFig. 1A and B.. The most intense staining in all species studied was observed in collections of large neurons of the ventral horns with the characteristic morphology of a-motor neurons ŽFigs. 1–3.. The ventral root fibers travelling within the ventral white matter were also intensely NP-NF immunoreactive. Two other collections of neurons displayed moderate to high staining intensity. These were the intermediolateral cell column and the neurons of the Clark’s column within the thoracic and upper lumbar regions ŽFigs. 1 and 4.. Individual neurons with light to moderate staining intensity were scattered within the upper portion of the ventral horns, the intermediate gray and the dorsal gray matter of the cord ŽFigs. 1 and 2.. Consistent with their cholinergic phenotype, the ventral horn a-motor neurons and their axons within the ventral white matter and the ventral roots contained prominent ChAT immunoreactivity ŽFig. 2B, D and F.. ChAT immunoreactivity was also observed within the neurons of the intermediolateral cell column and scattered neurons within the more medial aspects of the intermediate horns, close to the central canal. Qualitative comparison of adjacent sections stained for ChAT and NP-NF revealed nearly identical distribution
51
and apparent density within the ventral horns. Both immunoreactivities were found within the large a-motor neurons ŽFig. 2A–F.. Quantitative determination of ChAT and NP-NF immunoreactivity in the ventral horns of the rat, marmoset and rhesus revealed that on average 82– 100% of the ChAT-positive ventral horn neurons also contained NP-NF immunoreactivity. A similar result was obtained comparing the numbers of a-motor neurons in Nissl preparations to NP-NF-positive neurons in the human formalin-fixed specimens Žin which ChAT staining was absent due to prolonged fixation.. Double immunohistochemistry for ChAT and NP-NF within the rhesus spinal cord resulted in concurrent localization of the above markers in virtually all a-motor neurons ŽFig. 3.. In double-stained sections, 92–95% of ChAT-positive ventral horn neurons also contained NP-NF immunoreactivity. In addition, 99% of the large ventral horn NP-NF-positive neurons contained ChAT immunoreactivity. 3.2. Brainstem A large population of NP-NF immunoreactive neurons was observed throughout the brainstem of all species examined. As with the spinal cord, immunoreactivity was also observed in the proximal neuronal processes. NP-NFpositive neurons were observed in most brainstem structures, including the superior colliculus, the reticular formation, the vestibular nuclei and the red nucleus. The most prominent NP-NF immunoreactivity, however, was observed within neurons of all of the cranial nerve motor nuclei ŽFig. 5A, C and E.. NP-NF-positive neurons were observed in the oculomotor Žincluding its Edinger–Westphal subdivision. ŽFig. 5E., trochlear, abducens ŽFig. 5B., trigeminal motor, facial, vagal motor and hypoglossal nuclei ŽFig. 5A.. Comparison with matching or adjacent Nissl-stained sections indicated that virtually all of the neurons within these cranial nerve nuclei were also NPNF-positive. A number of brainstem nuclei contained ChAT-positive neurons. Prominent staining was observed within the pedunculopontine and laterodorsal tegmental nuclei. In addition, all of the cranial nerve motor nuclei contained ChAT-positive neurons. Qualitative comparison of adjacent sections stained for ChAT and NP-NF revealed that virtually all of the ChAT-positive motor neurons within the brainstem also contained NP-NF immunoreactivity ŽFig. 5A–F..
Fig. 4. The dorsal nucleus of Clark ŽA–D. and the intermediolateral cell column ŽE and F. were the only other neuronal groups which displayed prominent NP-NF immunoreactivity within the spinal cord. ŽA. and ŽB. are examples of NP-NF immunoreactivity in the dorsal nucleus of Clark in the human spinal cord. ŽC. shows NP-NF staining within the same nucleus in the rhesus and ŽD. in the marmoset. In the lower species studied, including the rhesus shown here ŽE., the neurons of the intermediolateral cell column contained an intermediate intensity of NP-NF immunoreactivity. In contrast, the same neuronal group in the human brain ŽF. displayed intense NP-NF immunoreactivity. The magnification in ŽA. and ŽB. is 85 = and in ŽC. – ŽF. is 170 = .
52
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
53
3.3. Motor cortex
4. Discussion
Neurons containing NP-NF immunoreactivity were observed throughout the cortical mantle in the rat, marmoset, rhesus and man. The most prominent staining was observed within the cell bodies and proximal processes of pyramidal neurons in LIII and LV of neocortex and throughout limbic and paralimbic cortical zones. Consistent with this pattern of staining, neurons within LIII and LV of motor cortex ŽBrodmann’s area 4. were NP-NFpositive. In particular, intense NP-NF immunoreactivity was observed within the giant LV motor neurons ŽBetz cells in the primate. ŽFig. 5G and H. from which the corticospinal tract originates. Comparison with adjacent Nissl-stained sections revealed that nearly all of these giant neurons of motor cortex contained NP-NF immunoreactivity in the rat, marmoset, rhesus and human.
The results of the present study demonstrate the presence of a relatively high density of NP-NF immunoreactivity within virtually all motor neurons of the spinal cord, brainstem and motor cortex. Although neurons in some adjacent structures also contained NP-NF immunoreactivity, NP-NF staining in motor neurons was considerably more intense. NP-NF immunoreactivity in motor neurons was observed in all of the species examined. In addition, perikaryal NP-NF staining was obtained in motor neurons regardless of the mode and strength of fixation. These results indicate that the presence of intense NP-NF immunoreactivity can be used as a reliable marker of all motor neurons. Thus, immunohistochemical staining for NP-NF using the SMI-32 or other specific antibodies may prove a useful tool in experimental studies of motor neurons, particularly in cases where identification of lower motor neurons using ChAT is difficult due to variable fixation or absence of species-appropriate antibodies. A few recent reports had indicated staining for NP-NF in spinal cord motor neurons using the SMI-32 antibody in situ w6,14,23x. The extent of NP-NF staining in motor neurons, however, was unclear. The findings of the present experiment confirm and extend the results of the above studies, by showing that NP-NF immunoreactivity is present in virtually all of the lower and upper motor neurons. A few in vitro investigations had used the SMI-32 antibody to identify motor neurons in dissociated spinal cord neuronal cultures w5,6x. Our findings of NP-NF staining of moderate intensity within a subpopulation of non-motor spinal cord neurons suggest that this in vitro approach will most likely lead to the identification of non-motor as well as motor neurons.
3.4. P-NF protein Throughout the brain and spinal cord, P-NF immunoreactivity was found in axon-like processes ŽFig. 6A., including those within the white matter and major fiber tracts. P-NF immunoreactivity was rarely observed in neurons, and when present, was of relatively low intensity. Consistent with this pattern of staining, P-NF immunoreactivity was only rarely observed within the motor neurons of the spinal cord ŽFig. 6A., brainstem and motor cortex. 3.5. Amyotrophic lateral sclerosis Within the spinal cord specimens from the three ALS patients, a dramatic loss of a-motor neurons was apparent in Nissl-stained sections ŽFig. 7A and B.. Adjacent sections stained for NP-NF contained the same pattern and density of ventral horn motor neurons as the Nissl-stained sections ŽFig. 7C and D.. Other NP-NF-positive neuronal populations, such as those of the Clark’s column, appeared to be relatively well preserved in ALS cord when compared with normal specimens ŽFig. 7E–H.. However, unlike the normal spinal cords, some of the remaining ventral horn motor neurons in the ALS cord also contained immunoreactivity for P-NF ŽFig. 6B and C..
4.1. Functional significance NF are a major and specific structural element of neuronal cytoskeleton. They are composed of three subunits of approximately 68 Žlow; NFL., 145 Žmedium; NFM. and 200 kDa Žhigh; NFH. molecular weight. Each of these subunits is the product of a distinct gene. The NF is made of a core filament, composed of all three subunits, and projections, which are made up of NFM and NFH and project out perpendicularly from the core filament. Apposed projections from different core filaments connect to
Fig. 5. NP-NF ŽA, C and E. and ChAT ŽB, D and F. immunoreactivity in adjacent sections of brainstem at the level of the dorsal Žmotor. nucleus of vagus Župper nuclei. and hypoglossal nucleus ŽA and B. and the abducens nucleus ŽC and D. of the rhesus and the oculomotor nuclear complex of man ŽE and F.. Note that the distribution and apparent density of the two markers in spinal cord motor neurons is virtually identical, indicating the presence of NP-NF in almost all of these motor neurons. Arrow in ŽE. points to the autonomic division of the oculomotor nucleus ŽEdinger–Westphal.. Identical results were obtained in other cranial nerve motor nuclei of all species studied. Intense NP-NF immunoreactivity was present in the giant Betz cells Župper motor neurons. in all species examined, including the human ŽG. and rhesus ŽH. shown here. Note that smaller pyramidal neurons in the same cortical areas Žarrows in H. display considerably less intense immunoreactivity. The magnification in ŽA. – ŽD. is 35 = , in ŽE. and ŽF. is 20 = and in G and H is 85 = .
54
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
create cross-bridges. The structure thus created serves as cytoskeletal support for neurons w14,15,25x. NF, particularly the NFM and NFH subunits, are heavily phosphorylated ŽP-NF. w25,30x. P-NF are found preferentially within axons while NP-NF are observed primarily in neuronal perikarya and dendrites w14,31x. The SMI-32 and SMI-31 antibodies used in the present study recognize the NFM and NFH subunits within NP-NF and P-NF, respectively w31x. Previous work has shown that although most neurons stain positively for NP-NF when the SMI-32 antibody is used, large projection neurons, such as cortical pyramidal neurons are particularly enriched in NP-NF w2,21x. This observation indicates that large neurons, with their fast-conducting, large caliber axons require a high content of NF for structural support and integrity. In fact, accumulation of NF within axons has been shown to be associated with enlargement of axonal caliber during development w15,38x. Although qualitative in nature, our observations regarding the relative intensity of NP-NF immunoreactivity in different populations of neurons allows certain generalizations. Consistent with the above observations, we found motor neurons, which are among the largest neurons in the nervous system, to be highly enriched in NP-NF. Our observations indicate that the association between the gradient of staining intensity and the size of neurons can be detected in comparisons within and between various species. Thus, in each species studied, we found the motor neurons to display among the highest intensities of staining while adjacent and nearby neurons, which were smaller in size, displayed considerably less intense staining. In addition, we observed qualitatively increased NP-NF staining intensity in motor neurons as the phylogenetic scale was ascended. Motor neurons in the rat spinal cord, which are the smallest among the species studied, displayed the lowest staining intensity. In contrast, the motor neurons of the human spinal cord, which were the largest motor neurons observed, appeared to contain the highest staining intensity. These observations are consistent with the findings of Campbell and Morrison w3x, which indicated a strong positive correlation between intensity of NP-NF immunoreactivity and neuronal size. In agreement with the
primary localization of P-NF in axons, in normal brains we observed staining for P-NF mainly in axons. P-NF staining was virtually absent from neurons. Depletion of NF has been shown to result in the disruption and fragmentation of microtubules within neurons w32,33x. Since one of the primary functions of microtubules is axoplasmic transport of materials, the aforementioned finding suggests that NF are essential for such transport. Thus, in addition to providing structural support, the high content of NP-NF within motor neurons is likely to contribute to transport of nutrients and essential chemicals to the very long axons of these neurons. 4.2. Implications for motor neuron disease Our observations in ALS spinal cord must be regarded as preliminary since they are qualitative in nature and are based on three cases only. However, our findings do indicate that staining for NP-NF may provide a reliable means for assessing the extent of motor neuron loss in motor neuron diseases such as ALS. Within the spinal cord of ALS patients, we observed a marked loss of motor neurons when compared to normal controls. The results obtained with NP-NF immunostaining were similar to that obtained using the Nissl stain. Neurons in other spinal cord regions which also contain a relatively high intensity of NP-NF staining, such as those within the dorsal nucleus of Clark and the intermediolateral cell column were relatively well preserved in ALS tissue stained with either method. Recent evidence indicates that the high content of NF may be a source of liability for spinal cord motor neurons in ALS. Consistent with the results of this study, a number of investigations have found accumulations of P-NF in the surviving spinal cord motor neurons in ALS w19,22,24x. Such accumulations are thought to cause reduced content and disorganization of NF in motor axons leading to abnormalities in axonal cytoskeleton. Similar P-NF accumulations have been reported in spinal cord motor neurons of transgenic animals expressing mutations in superoxide dismutase 1 ŽSOD1. gene found in familial ALS w23,40x. Deletion of NFL in SOD1 transgenic mice has been shown to significantly slow the progression and extent of motor
Fig. 6. P-NF immunoreactivity within the ventral horn of thoracic sections of the spinal cord in a normal subject ŽA. and cases of ALS ŽB and C.. In the normal subject ŽA., virtually all P-NF immunoreactivity is confined to neuronal processes Žmostly axons.. Only rare and isolated neurons resembling motor neurons were found to contain a low intensity of staining Žarrow in A.. Identical results were obtained in all species studied. In ALS ventral horn, P-NF axons appeared to be disrupted ŽB and C. and a few of the remaining large neurons Žputative a-motor neurons. displayed intense P-NF immunoreactivity Žlarge arrow in B and C.. In some instances, such neurons seemed to contain P-NF-positive inclusions, as parts of the neuron were free of staining Žempty arrow in B.. In addition, the ventral horns of ALS cases contained P-NF-positive bodies Žsmall arrows in B and C. which are most likely axonal enlargements andror remaining debris from degenerated motor neurons. The magnification in ŽA. – ŽC. is 200 = . Fig. 7. Žp.12. The ventral horn of ALS patients ŽB and D; thoracic. displayed a marked loss of motor neurons when compared with normal subjects ŽA and C. in both Nissl-stained sections ŽA and B. and NP-NF-immunostained material ŽC and D.. Note that the remaining motor neurons are considerably smaller than normal. Other spinal cord neurons which contain a relatively high intensity of NP-NF immunoreactivity in the normal Žsee Fig. 3., such as the dorsal nucleus of Clark ŽE and G. and the intermediolateral cell column ŽF and H. were relatively well preserved in ALS in tissue processed for the visualization of Nissl substance ŽE and F. or NP-NF ŽG and H.. ŽG. and ŽH. are at a lower level of the cord as compared with ŽE. and ŽF.. The magnification in ŽA. – ŽH. is 170 = .
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
55
56
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
Fig. 7
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58
neuron disorder w37x. Motor neuron degeneration ŽMND. mice which naturally develop motor neuron disease have been found to show a substantial acceleration of the disease process when expressing a NF-disrupting transgene w28x. Transgenic animals expressing a mutant NFL subunit w16x or overexpressing the wild type human NFH w7x or the wild type mouse NFL w39x have been shown to display relatively selective motor neuron death. Finally, mutations Žpolymorphisms. have been reported in the repeat motif of the NFH gene in a number of ALS cases w11,34x. Taken together, the evidence summarized above has been suggested to indicate that the enrichment of NF in motor neurons may be a natural risk factor for the vulnerability of these neurons in ALS. However, it should be noted that some studies have failed to detect polymorphisms in the NFH gene in large samples of ALS patients w29,36x. Furthermore, some groups of motor neurons, such as those in the oculomotor nucleus, which display intense NP-NF immunoreactivity, are relatively well preserved in ALS w1x.
w12x
w13x
w14x
w15x w16x
w17x
w18x
Acknowledgements We thank Nicholas Nagykery for expert technical assistance. This work is supported in part by the Milton Fund of the Harvard Medical School.
References w1x J.H. Adams, L.W. Duchen ŽEds.., Greenfield’s Neuropathology, Oxford Univ. Press, New York, 1992, 1557 pp. w2x M.J. Campbell, P.R. Hof, J.H. Morrison, A subpopulation of primate corticocortical neurons is distinguished by somatodendritic distribution of neurofilament protein, Brain Res. 539 Ž1991. 133–136. w3x M.J. Campbell, J.H. Morrison, Monoclonal antibody to neurofilament protein ŽSMI-32. labels a subpopulation of pyramidal neurons in the human and monkey neocortex, J. Comp. Neurol. 282 Ž1989. 191–205. w4x M.B. Carpenter, J. Sutin, Human Neuroanatomy, Williams and Wilkins, Baltimore, 1983, 872 pp. w5x S.G. Carriedo, H.Z. Yin, R. Lamberta, J.H. Weiss, In vitro kainate injury to large, SMI-32Žq. spinal neurons is Ca2q dependent, NeuroReport 6 Ž1995. 945–948. w6x S.G. Carriedo, H.Z. Yin, J.H. Weiss, Motor neurons are selectively vulnerable to AMPArkainate receptor-mediated injury in vitro, J. Neurosci. 16 Ž1996. 4069–4079. w7x F. Cote, J.F. Collard, J.P. Julien, Progressive neuropathy in transgenic mice expressing the human neurofilament heavy gene: a mouse model of amyotrophic lateral sclerosis, Cell 73 Ž1993. 35–46. w8x J.R. Delfs, D.M. Saroff, Y. Nishida, J. Friend, C. Geula, Effects of NMDA and its antagonists on ventral horn cholinergic neurons in organotypic roller tube spinal cord cultures, J. Neural Transm. 104 Ž1997. 31–51. w9x A. Eisen, C. Krieger, Pathogenic mechanisms in sporadic amyotrophic lateral sclerosis, Can. J. Neurol. Sci. 20 Ž1993. 286–296. w10x J. Ericson, S. Thor, T. Edlund, T.M. Jessell, T. Yamada, Early stages of motor neuron differentiation revealed by expression of homeobox gene Islet-1, Science 256 Ž1992. 1555–1560. w11x D.A. Figlewicz, A. Krizus, M.G. Martinoli, V. Meininger, M. Dib,
w19x
w20x
w21x
w22x
w23x
w24x
w25x w26x w27x
w28x
w29x
w30x w31x
57
G.A. Rouleau, J.P. Julien, Variants of the heavy neurofilament subunit are associated with the development of amyotrophic lateral sclerosis, Hum. Mol. Genet. 3 Ž1994. 1757–1761. C. Geula, M.-M. Mesulam, Cholinergic systems and related neuropathological predilection patterns in Alzheimer disease, in: R.D. Terry, R. Katzman, K.L. Bick ŽEds.., Alzheimer Disease, Raven Press, New York, 1994, pp. 263–291. C. Geula, M.-M. Mesulam, C.C. Kuo, H. Tokuno, Postnatal development of cortical acetylcholinesterase-rich neurons in the rat brain: permanent and transient patterns, Exp. Neurol. 134 Ž1995. 157–178. T. Gotow, J. Tanaka, Phosphorylation of neurofilament H subunit as related to arrangement of neurofilaments, J. Neurosci. Res. 37 Ž1994. 691–713. M.K. Lee, D.W. Cleveland, Neuronal intermediate filaments, Annu. Rev. Neurosci. 19 Ž1996. 187–217. M.K. Lee, J.R. Marszalek, D.W. Cleveland, A mutant neurofilament subunit causes massive, selective motor neuron death: implications for the pathogenesis of human motor neuron disease, Neuron 13 Ž1994. 975–988. A.I. Levey, J.P. Bolam, D.B. Rye, A.E. Hallanger, R.M. Demuth, M.-M. Mesulam, B.H. Wainer, A light and electron microscopic procedure for sequential double antigen localization using diaminobenzidine and benzidine dihydrochloride, J. Histochem. Cytochem. 34 Ž1986. 1449–1457. A.I. Levey, B.H. Wainer, D.B. Rye, E.J. Mufson, M.-M. Mesulam, Choline acetyltransferase-immunoreactive neurons intrinsic to rodent cortex and distinction from acetylcholinesterase-positive neurons, Neuroscience 13 Ž1984. 341–353. V. Manetto, N.H. Sternberger, G. Perry, L.A. Sternberger, P. Gambetti, Phosphorylation of neurofilaments is altered in amyotrophic lateral sclerosis, J. Neuropathol. Exp. Neurol. 47 Ž1988. 642–653. M.-M. Mesulam, C. Geula, Acetylcholinesterase-rich neurons of the human cerebral cortex: cytoarchitectonic and ontogenetic patterns of distribution, J. Comp. Neurol. 306 Ž1991. 193–220. M.-M. Mesulam, C. Geula, Differential distribution of a neurofilament protein epitope in acetylcholinesterase-rich neurons of human cerebral neocortex, Brain Res. 544 Ž1991. 169–173. H. Mizusawa, S. Matsumoto, S.H. Yen, A. Hirano, R.R. RojasCorona, H. Donnenfeld, Focal accumulation of phosphorylated neurofilaments within anterior horn cell in familial amyotrophic lateral sclerosis, Acta Neuropathol. 79 Ž1989. 37–43. B.M. Morrison, J.W. Gordon, M.E. Ripps, J.H. Morrison, Quantitative immunocytochemical analysis of the spinal cord in G86R superoxide dismutase transgenic mice: neurochemical correlates of selective vulnerability, J. Comp. Neurol. 373 Ž1996. 619–631. D.G. Munoz, C. Greene, D.P. Perl, D.J. Selkoe, Accumulation of phosphorylated neurofilaments in anterior horn motoneurons of amyotrophic lateral sclerosis patients, J. Neuropathol. Exp. Neurol. 47 Ž1988. 9–18. H.C. Pant, Veeranna, Neurofilament phosphorylation, Biochem. Cell Biol. 73 Ž1995. 575–592. G. Paxinos, The Rat Nervous System, Vol. 2, Academic Press, Australia, 1985, 362 pp. S.L. Pfaff, M. Mendelsohn, C.L. Stewart, T. Edlund, T.M. Jessell, Requirement for LIM homeobox gene Isl1 in motor neuron generation reveals a motor neuron-dependent step in interneuron differentiation, Cell 84 Ž1996. 309–320. J. Plummer, A. Peterson, A. Messer, Accelerated and widespread neuronal loss occurs in motor neuron degeneration Žmnd. mice expressing a neurofilament-disrupting transgene, Mol. Cell. Neurosci. 6 Ž1995. 532–543. K. Rooke, D.A. Figlewicz, F.Y. Han, G.A. Rouleau, Analysis of the KSP repeat of the neurofilament heavy subunit in familiar amyotrophic lateral sclerosis, Neurology 46 Ž1996. 789–790. L.A. Sternberger, Immunocytochemistry, Wiley, New York, 1986, 524 pp. L.A. Sternberger, N.H. Sternberger, Monoclonal antibodies distin-
58
w32x
w33x
w34x
w35x
w36x
Y.M. Tsang et al.r Brain Research 861 (2000) 45–58 guish phosphorylated and nonphosphorylated forms of neurofilaments in situ, Proc. Natl. Acad. Sci. U.S.A. 80 Ž1983. 6126–6130. X T. Tashiro, R. Imai, Y. Komiya, Early effects of beta,beta -iminodipropionitrile on tubulin solubility and neurofilament phosphorylation in the axon, J. Neurochem. 63 Ž1994. 291–300. T. Tashiro, Y. Komiya, Impairment of cytoskeletal protein transport X due to aging or beta,beta -iminodipropionitrile intoxication in the rat sciatic nerve, Gerontology 40 Ž1994. 36–45, Suppl. J. Tomkins, P. Usher, J.Y. Slade, P.G. Ince, A. Curtis, K. Bushby, P.J. Shaw, Novel insertion in the KSP region of the neurofilament heavy gene in amyotrophic lateral sclerosis ŽALS., NeuroReport 9 Ž1998. 3967–3970. T. Tsuchida, M. Ensini, S.B. Morton, M. Baldassare, T. Edlund, T.M. Jessell, S.L. Pfaff, Topographic organization of embryonic motor neurons defined by expression of LIM homeobox genes, Cell 79 Ž1994. 957–970. J.D. Vechio, L.I. Bruijn, Z. Xu, R.H. Brown, D.W. Cleveland, Sequence variants in human neurofilament proteins: absence of linkage to familial amyotrophic lateral sclerosis, Ann. Neurol. 40 Ž1996. 603–610.
w37x T.L. Williamson, L.I. Bruijn, Q. Zhu, K.L. Anderson, S.D. Anderson, J.P. Julien, D.W. Cleveland, Absence of neurofilaments reduces the selective vulnerability of motor neurons and slows disease caused by a familial amyotrophic lateral sclerosis-linked superoxide dismutase 1 mutant, Proc. Natl. Acad. Sci. U.S.A. 95 Ž1998. 9631– 9636. w38x T.L. Williamson, J.R. Marszalek, J.D. Vechio, L.I. Bruijn, M.K. Lee, Z. Xu, R.H. Brown Jr., D.W. Cleveland, Neurofilaments, radial growth of axons, and mechanisms of motor neuron disease, Cold Spring Harbor Symp. Quant. Biol. 61 Ž1996. 709–723. w39x Z. Xu, L.C. Cork, J.W. Griffin, D.W. Cleveland, Increased expression of neurofilament subunit NF-L produces morphological alterations that resemble the pathology of human motor neuron disease, Cell 73 Ž1993. 23–33. w40x B. Zhang, P. Tu, F. Abtahian, J.Q. Trojanowski, V.M. Lee, Neurofilaments and orthograde transport are reduced in ventral root axons of transgenic mice that express human SOD1 with a G93A mutation, J. Cell Biol. 139 Ž1997. 1307–1315.