Animal Models for Motor Neuron Disease

Special Topic Overview

Sherril L. Green and Ravi J. Tolwani* Abstract: Motor neuron disease is a general term applied to a broad class of neurodegenerative diseases that are characterized by fatally progressive muscular weakness, atrophy, and paralysis attributable to loss of motor neurons. At present, there is no cure for most motor neuron diseases, including amyotrophic lateral sclerosis (ALS), the most common human motor neuron disease—the cause of which remains largely unknown. Animal models of motor neuron disease (MND) have significantly contributed to the remarkable recent progress in understanding the cause, genetic factors, and pathologic mechanisms proposed for this class of human neurodegenerative disorders. Largely driven by ALS research, animal models of MND have proven their use- fulness in elucidating potential causes and specific pathogenic mechanisms, and have helped to advance promising new treatments from “benchside to bedside.” This review summarizes important features of selected established animal models of MND: genetically engineered mice and inherited or spontaneously occurring MND in the murine, canine, and equine species.

The term “motor neuron disease” (MND) is used to designate Table 1. General classification of selected common human motor a variety of neurologic disorders (Table 1), the principal fea- neuron diseases (MND)* tures of which are attributable to dysfunction of upper and Primary MND lower motor neurons. Patients with amyotrophic lateral sclero- Idiopathic Sporadic amyotrophic lateral sclerosis sis (ALS), the most common motor neuron disease, manifest Western Pacific amyotrophic lateral variable combinations of both upper and lower motor neuron sclerosis-parkinsonism-dementia-complex signs, including spasticity, hyperreflexia, and extensive plantar Amyotrophic lateral sclerosis with frontal lobe dementia Monomelic MND signs (upper motor neuron signs); and progressive muscular Heritable weakness, fasciculations, and atrophy (lower motor neuron Autosomal dominant Familial amyotrophic lateral sclerosis signs), leading to fatal paralysis (1). Cognitive function is not Familial amyotrophic lateral sclerosis with dementia affected, and death usually occurs from complications attribut- Autosomal recessive able to difficulties in swallowing and respiration. Electromyo- Spinal muscular atrophy Neuroaxonal dystrophy graphy reveals fibrillations, and muscle biopsy reveals Pseudobulbar palsy denervation atrophy. However, there are no specific antemor- X-linked tem diagnostic tests. The histopathologic findings include den- Bulbospinal neuronopathy (Kennedy’s syndrome) ervation atrophy of skeletal muscle, and loss of motor neurons Secondary MND in the anterior horn of the spinal cord, motor nuclei, and the Infections motor cortex (2). At the light microscopic level, the motor neu- HIV, HTLV-1 Syphilis rons almost always contain axonal swellings or spheroids filled Poliomyelitis virus with misaligned neurofilament (NF) protein and accumula- Metabolic disease Hyperparathyroidism tions of NFs in the cell body (Figure 1) (2). Hexosaminidase deficiency Amyotrophic lateral sclerosis was first described in 1869 Autoimmune disease by the French physician Jean-Martin Charcot (3). The Greek Lymphoma Paraproteinemia derivation of the word amyotrophic means “muscles without Antigangliosides antibodies nourishment” (a means without, myo means muscle, and Neurotoxins trophic means nourishment). Lateral refers to the region of Aluminum, lead, mercury Lathyrus sativus toxin: BOAA the spinal cord where the axons of diseased and Cycad toxin: BMAA dysfunctioning motor neurons have degenerated and are re- HIV = human immunodeficiency virus; HTLV-1 = human T-cell placed by sclerosis, or scars. Also known as Lou Gehrig’s Dis- lymphotropic virus 1; BOAA = ␤-N-oxalyl-amino-L-alanine; BMAA = ␤- methyl-amino-alanine. *Adapted with permission from J. Lowe, G. Lennox, ease, ALS affects 4 to 6 individuals per 100,000 people and P. N. Leigh. 1997. Disorders of movement and system degenerations, p. 321. In D. I. Graham and P. L. Lantos (ed.), Greenfield’s Neuropathol- ogy, Sixth Edition. Oxford University Press, Inc., New York. Department of Comparative Medicine, Stanford University School of Medi- cine, Stanford, California *Address correspondence to: Dr. Ravi J. Tolwani, 287 Campus Drive, RAF worldwide (4). The disease usually affects individuals dur- 1, Quad 7, Building 330, Stanford, CA 94305-5410. ing the fifth decade of life, is progressive, and is uniformly fatal

480 Special Topic Overview

Figure 1. (A) Perikaryl accumulation of neurofilaments in motor neuron disease. Silver impregnation; magnification x400. (B) Spheroids (arrowheads) in the anterior horn of an A ALS patient. One of the spheroids is compressing an adja- cent motor neuron (arrow). Silver impregnation; magnification x300. (With permission from: M. Iwata and A. Hirano, 1979. Current problems in the pathology of amyotrophic lateral sclerosis, p. 288–289. In H. M. Zimmerman [ed.], Progress in Neuropathology, Vol. 4. Raven Press, New York.) (5, 6). Most people die within 5 years after the onset of symptoms (5). Although ALS is familial in approximately 10% of patients, the cause of sporadic ALS (SALS), which accounts for the remaining 90% of all cases, is unknown (7, 8). The remarkable clinical similarity between familial ALS (FALS) and SALS, however, suggests that they share common pathways of motor neuron death. In 1993, Rosen et al. (9) discovered that 20% of patients with FALS carried a mutation in the Cu/Zn superoxide dismutase 1 (Cu/Zn SOD1) gene, which is located on chromosome 21. This gene is a metalloen- zyme and is one of the principal oxygen-derived free radical scavengers and protectants against oxidant injury to the nervous system. This discovery strengthened the prominent hypothesis that dysfunction and subse- quent death of the motor neuron is related to anoma- B lous enzyme activity and cellular damage attributable to oxidant stress. However, this hypothesis recently has been challenged by Bruijn et al. (10), who propose that mutant SOD1 toxicity does not arise from superoxide- mediated oxidant stress but from some other as yet uni- dentified biochemical mechanism. This does not completely discount the role of oxidant stress in MND because many disease-related and normal biological processes, including normal aging, produce reactive oxygen species capable of damaging DNA proteins, car- bohydrates, and lipids (11). Nevertheless, why “at-risk” populations of motor neurons are selectively vulnerable to injury and the specific mechanisms that subse- quently lead to cytoskeletal dysfunction and cell death is not fully understood. Proposed pathways involving in MND are schematized in Figure 2. The catalogue of animal models for human MND is extensive (4, 12–19) and is generally divided into two categories: those that develop spontaneously, which are naturally acquired, inherited, or sporadic, and those that are Genetically Engineered Animal Models experimentally induced (Table 2). Here we summarize the of Motor Neuron Disease contributions and the limitations of selected experimentally To date, over 50 SOD1 mutations have been associated induced and spontaneously occurring animal models for with FALS (20), and the number will likely continue to grow. MND: murine transgenic motor neuron disease, and sponta- Genetically engineered mice with the most common muta- neously occurring murine, canine, and equine motor neuron tions—A4V, G93A, G37R or G86R (21–23)—have been the disease. Ideally, animal models should reflect the clinical most extensively studied and have contributed substantially to and neuropathologic features of the human disease as advancing our understanding of human MND. All these muta- closely as possible. Selected features of the aforementioned tions have induced MND in mice that is clinically and patho- animal models, including clinical characteristics and general logically similar to ALS. However, the gene copy number and pathologic features, are summarized and compared with level of gene expression significantly affects the pathologic those of ALS (Table 3). Extensive descriptions of the neuro- changes and age at onset in the various transgenic lines. Mice pathologic findings, particularly the variations in the animal expressing the G93A SOD1 mutation have been proposed to models at the light and electron microscopic levels, can be best reproduce ALS-like features (24, 25). found elsewhere (4, 11–18 and 4, 12–27, respectively) and The initial expectation was that these mutations cause are beyond the scope of this review. disease by lowering Cu/Zn SOD1 activity, as was observed

481 Vol 49, No 5 Laboratory Animal Science October 1999

Figure 2. Schematic model of proposed mechanisms leading to motor neuron disease (ER = endoplasmic reticulum, SOD = superoxide dismutase, Bcl-2 = a family of genes that regulates apoptosis). in the erythrocytes and tissues of ALS patients(26, 27). (10) may explain why the efficacy of antioxidant therapy in However, in several of the transgenic mouse lines (the G93A Cu/Zn SOD1 transgenic mice or in FALS patients is moder- and G37R mice in particular), Cu/Zn SOD1 activity was nor- ate at best. Treatment of the transgenic mice with two mal. In addition, Cu/Zn SOD1 knockout mice or mice glutaminergic inhibitors, riluzole and gabapentin, had no ef- overexpressing the wild-type protein do not develop MND (21). fect on disease onset but did extend survival (28). The use of More recently, Bruijn et al. (10) reported that mice which riluzole in human ALS clinical trials was similarly effica- completely lack endogenous Cu/Zn SOD1 and those that cious and is presently approved for use in the treatment of have six times the normal concentrations of Cu/Zn SOD1 ALS (29). develop identical ALS-like pathologic changes in the pres- In addition to the SOD1 transgenic mice, transgenic mice ence of mutant SOD1 (10). Thus, combination of the results that express mutant NF or overexpress wild-type NF pro- of these experiments suggests that motor neurons degener- teins have documented that primary NF alterations can ate because mutant Cu/Zn SOD1 may itself be toxic or because MND that resembles ALS (30, 31). However, the direct cause it gains a toxic property or aberrant function. role of NFs in the etiopathogenesis is unclear, because NF Genetically engineered mice allow testing of specific hy- gene mutations associated with FALS appear to be rare (32, potheses relevant to human disease. Thus Cu/Zn SOD1 33). To the authors’ knowledge, these transgenic lines have transgenic mice have been used for preclinical drug screen- not been studied in preclinical trials to date. ing to test new treatment strategies for ALS (24). For ex- Targeted disruption of the genes that encode for brain-de- ample, disease onset was delayed (but survival was not rived neurotrophic factor (BDNF), ciliary neurotrophic fac- extended) in SOD1 transgenic mice fed an antioxidant diet tor (CNTF), and the tyrosine kinase B (trkB) receptor have containing high amounts of vitamin E and selenium (28). enabled investigators to examine how these elements influ- However, the data suggesting that the toxicity of mutant ence development and maintenance of motor neurons (34–38). Cu/Zn SOD1 may not directly mediate oxidant stress injury Although mutations in genes encoding for these proteins have

482 Special Topic Overview

Table 2. Classification of selected animal models for MND portant in the survival and maintenance of motor neurons (38). Experimentally induced Thus these animal models have contributed to under- Genetically engineered mice standing the dependence of the motor neuron on neurotrophins Cu/Zn SOD gene mutations Neurofilament gene mutations and overexpression and related factors. Treatment of neurodegenerative diseases CNTF, BDNF, and trkB gene knockouts with trophic or growth-related factors for the purpose of res- Neurotoxins cuing damaged or diseased neurons is an expanding field of IDPN Heavy metals—aluminum, lead, mercury research. To date and to our knowledge, results of ongoing Cycad toxin: BMAA clinical trials testing neurotrophic factors in patients with Lathyrus sativus toxin: BOAA Doxorubicin, vincristine, vinblastine ALS have not been published. As suggested by the afore- Nutritional deficiencies mentioned animal models, it will be crucial to continue to Calcium and magnesium investigate which cell types are capable of responding to Copper Ascorbic acid particular factors and at what point they are likely to do so Viral (e.g., in early development, during the postnatal period, or Poliomyelitis virus during adulthood) (39). Murine retrovirus Lactate dehydrogenase-elevating virus Although genetically engineered animal models offer Autoimmune enormous new insights into specific pathologic pathways, Experimental allergic gray matter disease (EAGMD) Experimental allergic MND (EAMND) they do have limitations. There is significant genetic strain Axotomy induced variation in transgenic mice commonly used as experimental animal models. This can make phenotypic expression Spontaneously occurring Heritable and comparisons confusing. In addition, owing to species dif- Murine ferences in life span and physiology, transgenic mice might Motor neuron degeneration (mnd) not truly replicate all aspects of the human disease. Despite Wasted mutant Progressive motor neuronopathy (pmn) these disadvantages, genetically engineered mice offer enor- Wobbler mous potential to test specific treatments. Their small size Canine and ready availability make them especially useful for test- Hereditary canine spinal muscular atrophy Acquired ing new therapeutic modalities, particularly novel com- Equine MND pounds and delivery systems with enhanced ability to cross CNTF = ciliary neurotrophic factor; BDNF = brain-derived neurotrophic the blood brain barrier. factor; trkB = tyrosine kinase B receptor; IDPN = ␤'␤'-iminodipropionitrile; BMAA = ␤-methyl-amino-alanine; BOAA = ␤-N-oxalyl-amino-L-alanine Spontaneously Occurring Mouse not been reported in association with ALS, experiments using knockout mice have yielded relevant information regard- Models of Motor Neuron Disease ing the survival and maintenance of motor neurons in vivo. Four spontaneously occurring, inherited mouse models of For example, homozygous BDNF mutant (-/-) mice usually MND have received considerable attention: the motor neuron degenerative (mnd) mouse (40, 41); the wobbler mouse die within 48 h of birth and have reduced numbers of cranial (42); the wasted mutant (43); and the progressive motor and sensory neurons, but no motor neuron loss (34). Some neuronopathy (pmn) mouse (44). All are well-established may survive up to 25 days of age, but have severe ataxia and models, characterized over the last decade or more, and all reduced and irregular breathing attributed to neuron loss in are autosomal recessive mutants. Recent advances in immuno- the vestibular ganglia, and petrose and nodose ganglia, re- cytochemical and molecular technologies have further indi- spectively. In contrast, trkB-deficient homozygous mutants cated the extent and the nature of motor neuron have significant loss of motor neurons in the facial nucleus involvement. Some are more similar to ALS than was origi- and lumbar part of the spinal cord (35, 36). These animals nally observed, and others differed greatly with further feed poorly, and almost all die by 2 days of age. Thus, the characterization. For example, the mnd mice have adult late trkB receptor appears to be necessary for normal postnatal onset motor neuron degeneration, and intraneuronal accu- motor neuron development, and motor neuron survival does mulation of NFs and lipofuscin-like material. However, not appear to be dependent on BDNF alone. It has been hy- these mice do not have skeletal muscle denervation and mo- pothesized that signaling through trkB by growth or trophic tor neuron loss characteristic of ALS. The lipofuscin-like mate- factors other than BDNF is required to maintain the adult rial contains a subunit of mitochondrial adenosine motor neuron (37). In contrast, CNTF is highly expressed in Schwann cells of triphosphate synthase characteristic of neuronal ceroid peripheral myelin and in certain astrocytes. Ciliary neu- lipofuscinosis (45, 46). Thus the mnd mouse is a more appro- rotrophic factor does not appear to be necessary for normal priate animal model of Batten’s disease (47), the most com- motor neuron development, but it may have a role in their mon human neuronal lipofuscinosis. The mnd trait maps to postnatal maintenance. The CNTF (-/-) animals develop the centromeric region of mouse chromosome 8, and the mouse lysosomal protein 1 gene is a candidate locus (48). progressive muscular atrophy and motor neuron degenera- In contrast to the mnd mouse, the wobbler mouse, first re- tion by 8 weeks of age (38). By 28 weeks of age, there is nu- ported in 1956 (49), has anterior horn cell loss and muscular merical reduction of motor neurons in the facial nucleus, denervation atrophy. Wobbler mice develop generalized along with loss of muscular strength. These data suggest tremors and forelimb weakness. The term wobbler is used to that CNTF is at least one of the several possible factors im-

483 Vol 49, No 5 Laboratory Animal Science October 1999

Table 3. Features of selected animal models of MND, compared with those of amyotrophic lateral sclerosis SOD mutant Wobbler Wasted Feature mouse mnd mouse mouse mutant pmn mouse HCSMA EMND ALS AD for (FALS) Inheritance AR AR AR AR AR AD Not inherited Not inherited (all other ALS) Mutation/ chromosome SOD1 gene (G93R) Ch8 Ch 11 Eef1a2 Ch 13 Ch Linkage Not inherited Ch 21/ gene group 1 SOD1 gene Onset of weakness Focal/ general Hind limb Forelimb Hind limb Hind limb General General Focal/general Age at onset of weakness Adult Adult Weanling Neonate Weanling At weaninga Adult Adult Adultb Survival time 4–6 months Not affected 6 months 12–14 days 1 month 8–16 weeksa Variable 2–5 yrs

2–5 yrsb Skeletal muscular atrophy + - + + + + + + UMN degeneration +/- +/- +/- - - - +/- + aHomozygotes; bheterozygotes. MND = motor neuron disease; pmn = progressive motor neuronopathy; HCSMA = hereditary canine spinal muscular atrophy; EMND = equine motor neuron disease; AR = autosomal recessive; AD = autosomal dominant; FALS = familial amyotrophic lateral sclerosis; SOD1 = superoxide dismutase 1; Ch = chromosome; Eef1a2 = gene that encodes for translation elongation factor EF-1␣; UMN = upper motor neuron; + = present; - = absent; +/- = varies quantitatively and qualitatively from ALS. describe the resulting gait abnormality. The predominant central-peripheral axonopathy in which proximal axons and forelimb involvement differs from that of other animal mod- cell bodies in the anterior horn are spared (44). It is not a els of ALS. Alterations in amino acid concentrations in the primary disease of the cell soma, which is different from the spinal cord also have been reported and are similar to those lesion of ALS. The pmn mouse appears to be a model of dy- observed in ALS patients (50). Receptor-binding assays have ing-back axonal disease. The mouse mutation pmn maps to indicated that distribution of glutamate receptors is altered chromosome 13 (57). in the wobbler mouse spinal cord (51), similar to the altered Although none of the aforementioned spontaneously oc- distribution of glutamate receptors in ALS patients. Wobbler curring mouse models is an exact replica of ALS, certain fea- mice may yield clues to the genetic basis of MND in a mouse tures of each may provide insights and help characterize model that shares important biochemical features with ALS. pathways involved in the etiopathogenesis of MND. The role The mapping of the wobbler gene to chromosome 11 (52), of oxidant stress, if any, in these murine models remains to and its close linkage to the ras-1 gene and the glutamine be clarified. The lack of identification of the specific genetic synthetase pseudogene (53), has allowed detection of af- mutations in the wobbler, mnd, and pmn mouse models is fected progeny before clinical signs of disease develop. something of a disadvantage. It makes it difficult to define Identification of affected animals before the onset of clinical specific pathophysiologic mechanisms beyond descriptive signs of disease is a distinct advantage for any animal model. characterizations, to detect affected progenies before the on- Unlike the mnd or wobbler mice, the wasted mutant is set of the disease, and to test specific treatments (e.g., those characterized by aggressive disease (e.g., onset at an early that have a reasonable chance of showing efficacy in human age and rapid progression to death). Wasted mutant mice clinical trials) (58). However, the known mutation in the are characterized by wasting and immunologic and neuro- wasted mutant, and those that will undoubtedly be discov- logic abnormalities starting 21 days after birth, with death ered in the other mouse model genes, may also occur in hu- ensuing by 28 days. Selective vacuolar degeneration of the mans and cause motor neuron disease. Therefore, identifying anterior horn cells and motor nuclei of the brainstem are the responsible genes and gene products in the spontaneously prominent features of this model and are characteristic of occurring murine models will contribute important advances ALS (54). It has been proposed that the wasted mutant has in MND research. With the recent advances toward comple- decreased ability for DNA repair (55). Chambers et al. (56) tion of a mouse genomic map, such discoveries are eminent. recently reported that the lethal mutation of the mouse Similar to the transgenic mice, spontaneously occurring wasted mutant is a deletion that abolishes expression of a mouse models offer the advantage of their small size and tissue-specific isoform of translation elongation factor 1, en- availability. They have also been used to evaluate potential coded by the Eefla2 gene. Future studies using this model treatments for MND. For example, pmn mice treated with may elucidate the mechanism(s) in this mutant that make CNTF have improved motor function, prolonged survival, the motor neuron especially vulnerable to injury. and decreased motor neuron loss (59). Recently, results of Muscular atrophy and paralysis are present in the pmn gene therapy using adenoviral delivery of neurotrophin 3 mouse. However, there is little loss of motor neurons in the and CNTF indicated that these factors are capable of pro- spinal cord. Pathologically this model is characterized by a moting distal axon degeneration and improving neuromus-

484 Special Topic Overview

cular function in the pmn mouse (60). Glial cell-derived neu- pathologic changes (e.g., axonal swellings) (70). This is an rotrophic factor also was documented to support survival of important finding because it suggests an early therapeutic motor neuron cell bodies in the pmn mouse (61). In wobbler window. mice, BDNF and CNTF also have been documented to arrest One disadvantage of the canine animal model is the cost degeneration of motor neurons (62). On the basis of results of maintaining a large breeding colony with long generation of these animal studies and others, clinical trials have been time. At present, there is no presymptomatic test for the dis- initiated to evaluate the efficacy of these agents in ALS pa- ease before clinical signs become apparent, and it may not tients (63). become phenotypically apparent in the heterozygotes (the breeding stock) until 2 years of age or later. However, large Inherited Canine MND canine pedigrees with dozens of offspring can allow localiz- Inherited motor neuron disease in dogs, hereditary canine ing of the related gene(s) and selective inbreeding to elimi- spinal muscular atrophy (HCSMA), was first reported in nate the heterogenous genetic background that complicates Brittany Spaniels in 1979 (64, 65). The present incidence in the phenotypic assessment of the mouse models (71). the pet population is unknown, but a colony of Brittany Spaniels with HCSMA has been maintained as an animal Equine MND model of MND. Affected dogs manifest clinical and neuro- Spontaneously occurring equine MND was first described pathologic features that are strikingly similar to those seen in 1990 (72). Since then, the epidemiologic, clinical, and neu- in ALS patients, with the exception that the motor cortex, ropathologic data gathered from over 100 cases suggest com- which is affected in humans, is spared in dogs. Over the last monalties with the sporadic form of ALS (73, 74). Neuronal 19 years, the HCSMA kindred of more than 250 character- loss is less extensive than that associated with SALS, and ized members has been studied. Pedigree analysis indicates there is no degeneration of the pyramidal pathway, which is that HCSMA is inherited as an autosomal dominant disease rudimentary in the horse. However, as similar to that associ- (66), similar to the mode of inheritance of FALS. The ated with ALS, oxidant injury to the motor neuron appears HCSMA trait has been recently mapped to the canine chro- to have a role. Pathologic evidence of oxidative injury in mosomal linkage group 1 (67, 68). The HCSMA pedigree is horses with MND includes the predominance of highly oxi- currently a focus of canine genetic mapping efforts, and in dative type-1 fibers in skeletal muscles with neurogenic at- all likelihood, the specific gene(s) responsible for HCSMA rophy, and lipopigment accumulations in the endothelium of will be identified in the near future. Syntenic studies indi- spinal cord capillaries (75). Vitamin E concentration is low cate that the markers linked to spinal muscular atrophy in some horses, and is believed to be attributable to nutri- (SMA; the human MND that is clinically similar to HCSMA) tional depletion and increased consumption in the face of are not linked to the HCSMA trait (67, 68). It is not known oxidative stress. Mutations in the Cu/Zn SOD1 gene have whether mutations in the canine Cu/Zn SOD1 gene are not been linked to equine MND (76). Epidemiologic and ret- linked to HCSMA. Thus, the role of oxidant injury in rospective clinicopathologic studies indicate that the disease HCSMA is uncertain at this time, but studies are underway is acquired, and environmental factors, specifically dietary to further characterize this aspect of canine MND. deficiency of vitamin E, along with other husbandry prac- The rate of progression of clinical HCSMA is dependent tices, are involved (73, 75, 77). on gene dose (e.g., homozygous pups develop an accelerated An important clinical feature in horses is the apparent form of the disease by 6 to 8 weeks of age that rapidly leads tendency of the disease to “burn itself out” in some animals to paralysis and necessitates euthanasia by 12 to 16 weeks (74). These animals typically have sudden onset and rapid of age [65, 69]). Heterozygotes may inherit the chronic form advancement of the clinical signs; the disease then appears of the disease, with an onset at 6 months to 2 years of age, to “stabilize” without further progression. These animals and may survive up to 7 years or longer (65, 69). This is the may survive for years, although they never completely re- only spontaneously occurring dominantly inherited animal cover. Some human ALS patients also experience this clini- MND in which onset and severity of the disease varies with cal scenario. Determining what triggers the disease, what the gene dose. sustains the progression, and what halts progression of the Skeletal muscle degeneration develops in the accelerated disease in some patients are important experimental ques- and chronic forms of HCSMA, but spinal cord motor neuron tions. Horses are not a particularly useful animal model be- loss is most apparent late in the disease in the heterozygotes cause of their size. However, equine MND is, at present, the (64). Axonal transport studies indicate that transport of only spontaneously occurring animal MND that could pro- cytoskeletal proteins in dogs with HCSMA is impaired (69). vide clues to the pathogenesis of sporadic ALS on the basis This may contribute to the characteristic massive neuro- of the epidemiologic and clinical data for the naturally ac- filamentous accumulations and swellings in the proximal quired disease. axons typically seen in the homozygous dogs (69). Axonal transport studies of HCSMA indicate that transport of Summary cytoskeletal proteins is impaired and may contribute to the Aside from nonhuman primates, a species in which inher- neurofilamentous accumulations and swellings in the proxi- ited or naturally acquired MND has not been reported, none mal axons (69). Electrophysiologic studies indicate that the of the aforementioned animal species have a well-developed motor units in affected homozygotes have evidence of failure corticospinal tract, a system that is uniquely and specifically of transmission at the level of the synapse, in the absence of involved in ALS. Nevertheless, the potential for animal mod-

485 Vol 49, No 5 Laboratory Animal Science October 1999 els to contribute to understanding the pathogenesis of MND 16. Pioro, E. P., and H. Mitsumoto. 1995. Animal models of is enormous. A presymptomatic diagnosis is currently not ALS. Clin. Neurosci. 3:375–385. possible in ALS patients, and clinical signs are usually 17. Mitsumoto, H., and E. P. Pioro. 1995. Animal models of amyotrophic lateral sclerosis. Adv. Neurol. 68:73–87. readily apparent before treatment regimens are imple- 18. Ludolph, A. C. 1996. Animal models for motor neuron dis- mented. Earlier treatment may substantially improve the eases: research directions. Neurology 47:S228–S232. quality of life and survival time, but an accurate diagnosis 19. Cork, L. C., J. C. Troncoso, G. G. Klavano, et al. 1988. before clinical signs are distinctly present can be difficult. Neurofilamentous abnormalities in motor neurons in spontaneously occurring animal disorders. J. Neuropathol. Exp. In addition, human autopsy material, the main source of Neurol. 47:420–431. tissue for the diagnosis and study of human MND, reflects end- 20. Tu, P. H., M. E. Gurney, J. P. Julien, et al. 1997. Oxidative stage disease processes. Thus, study of the presymptomatic pe- stress, mutant SOD1, and neurofilament pathology in riod, the early disease process, and the temporal changes is transgenic mouse models of human motor neuron disease. important and would not be possible without experimental Lab. Invest. 76:441–456. 21. Gurney, M. E., H. Pu, A. Y. Chiu, et al. 1994. Motor neuron models. In addition, why motor neurons appear to be selec- degeneration in mice that express a human Cu,Zn superox- tively vulnerable to injury is a question that remains unan- ide dismutase mutation. Science 264:1772–1775. swered and difficult to address experimentally without the 22. Ripps, M. E., G. W. Huntley, P. R. Hof, et al. 1995. use of animal models. The composite information from the Transgenic mice expressing an altered murine superoxide dismutase gene provide an animal model of amyotrophic lat- mouse, canine, and equine animal models for MND will con- eral sclerosis. Proc. Natl. Acad. Sci. USA 92:689–693. tinue to yield important insights into the pathogenesis of 23. Wong, P. C., C. A. Pardo, D. R. Borchelt, et al. 1995. An this disease and to aid in the development of new treatment adverse property of a familial ALS-linked SOD1 mutation strategies for a disease that was, until recently, considered causes motor neuron disease characterized by vacuolar de- untreatable. generation of mitochondria. Neuron 14:1105–1116. 24. Gurney, M. E. 1997. The use of transgenic mouse models of amyotrophic lateral sclerosis in preclinical drug studies. J. Neurol. Sci. 152(Suppl. 1):S67–S73. References 25. Gurney, M. E. 1997. Transgenic animal models of familial 1. Mitsumoto, H. 1994. Classification and clinical features of amyotrophic lateral sclerosis. J. Neurol. 244(Suppl. 2): amyotrophic lateral sclerosis. In Amyotrophic lateral sclero- S15–S20. sis. A comprehensive guide to management. Demos Publica- 26. Bowling, A. C., J. B. Schulz, R. H. J. Brown, et al. 1993. tions, New York. Superoxide dismutase activity, oxidative damage, and mito- 2. Hirano, A. 1996. Neuropathology of ALS: an overview. Neu- chondrial energy metabolism in familial and sporadic amyo- rology 47:S63–S66. trophic lateral sclerosis. J. Neurochem. 61:2322–2325. 3. Charcot, J. M., and A. Joffroy. 1869. Deux cas d’atrophie 27. Deng, H. X., A. Hentati, J. A. Tainer, et al. 1993. Amyo- musculaire progressive avec les lésions de la substance grise trophic lateral sclerosis and structural defects in Cu,Zn su- et faisceaux antérolatéraux de la moelle épinière. Arch. peroxide dismutase. Science 261:1047–1051. Physiol. Norm. Pathol. 2:354–367, 629–649, 744–760. 28. Gurney, M. E., F. B. Cutting, P. Zhai, et al. 1996. Benefit of 4. Price, D. L., D. W. Cleveland, and V. E. Koliatsos. 1994. Mo- vitamin E, riluzole, and gabapentin in a transgenic model of tor neuron disease and animal models. Neurobiol. Dis. 1:3–11. familial amyotrophic lateral sclerosis. Ann. Neurol. 39: 5. Tandan, R., and W. G. Bradley. 1985. Amyotrophic lateral 147–157. sclerosis: part 1. Clinical features, pathology, and ethical is- 29. Bensimon, G., L. Lacomblez, and V. Meininger. 1994. A sues in management. Ann. Neurol. 18:271–280. controlled trial of riluzole in amyotrophic lateral sclerosis. 6. Tandan, R., and W. G. Bradley. 1985. Amyotrophic lateral ALS/Riluzole Study Group. N. Engl. J. Med. 330:585–591. sclerosis: part 2. Etiopathogenesis. Ann. Neurol. 18:419–431. 30. Xu, Z., L. C. Cork, J. W. Griffin, et al. 1993. Increased ex- 7. Emery, A. E., and S. Holloway. 1982. Familial motor neu- pression of neurofilament subunit NF-L produces morphologi- ron diseases. Adv. Neurol. 36:139–147. cal alterations that resemble the pathology of human motor 8. Siddique, T., M. A. Pericak-Vance, B. R. Brooks, et al. neuron disease. Cell 73:23–33. 1989. Linkage analysis in familial amyotrophic lateral scle- 31. Cote, F., J. F. Collard, and J. P. Julien. 1993. Progressive rosis. Neurology 39:919–925. neuronopathy in transgenic mice expressing the human 9. Rosen, D. R., T. Siddique, D. Patterson, et al. 1993. Muta- neurofilament heavy gene: a mouse model of amyotrophic lat- tions in Cu/Zn superoxide dismutase gene are associated with eral sclerosis. Cell 73:35–46. familial amyotrophic lateral sclerosis. Nature 362:59–62. 32. Vechio, J. D., L. I. Bruijn, Z. Xu, et al. 1996. Sequence vari- 10. Bruijn, L. I., M. K. Houseweart, S. Kato, et al. 1998. Ag- ants in human neurofilament proteins: absence of linkage to fa- gregation and motor neuron toxicity of an ALS-linked SOD1 milial amyotrophic lateral sclerosis. Ann. Neurol. 40:603–610. mutant independent from wild-type SOD1. Science 281: 33. Meyer, M. A., and N. T. Potter. 1995. Sporadic ALS and chro- 1851–1854. mosome 22: evidence for a possible neurofilament gene de- 11. Sies, H., W. Stahl, and A. R. Sundquist. 1992. Antioxidant fect. Muscle Nerve 18:536–539. functions of vitamins. Vitamins E and C, beta-carotene, and 34. Jones, J. M., R. L. Albin, E. L. Feldman, et al. 1993. mnd2: other carotenoids. Ann. N. Y. Acad. Sci. 669:7–20. a new mouse model of inherited motor neuron disease. 12. Sillevis, S. P., and J. M. de Jong. 1989. Animal models of Genomics 16:669–677. amyotrophic lateral sclerosis and the spinal muscular atro- 35. Klein, R. 1994. Role of neurotrophins in mouse neuronal de- phies. J. Neurol. Sci. 91:231–258. velopment. FASEB J. 8:738–744. 13. Messer, A. 1994. Mutant mouse models of ALS. Neurobiol. 36. Klein, R., R. J. Smeyne, W. Wurst, et al. 1993. Targeted Aging 15:247–248. disruption of the trkB neurotrophin receptor gene results in 14. Price, D. L., V. E. Koliatsos, P. C. Wong, et al. 1996. Motor nervous system lesions and neonatal death. Cell 75:113–122. neuron disease and model systems: aetiologies, mechanisms 37. Davies, A. M. 1994. Neurobiology. Tracking neurotrophin and therapies. CIBA Found. Symp. 196:3–13. function. Nature 368:193–194. 15. Ludolph, A. C., and P. S. Spencer. 1996. Toxic models of up- 38. Masu, Y., E. Wolf, B. Holtmann, et al. 1993. Disruption of per motor neuron disease. J. Neurol. Sci. 139(Suppl.):53–59. the CNTF gene results in motor neuron degeneration. Na- ture 365:27–32.

486 Special Topic Overview

39. Aguzzi, A., S. Brandner, S. Marino, et al. 1996. Transgenic 58. 1997. Mutant mice and neuroscience: recommendations con- and knockout mice in the study of neurodegenerative diseases. cerning genetic background. Banbury Conference on Genetic J. Mol. Med. 74:111–126. Background in Mice. Neuron 19:755–759. 40. Messer, A., N. L. Strominger, and J. E. Mazurkiewicz. 1987. 59. Sendtner, M., H. Schmalbruch, K. A. Stockli, et al. 1992. Histopathology of the late-onset motor neuron degeneration Ciliary neurotrophic factor prevents degeneration of motor (Mnd) mutant in the mouse. J. Neurogenet. 4:201–213. neurons in mouse mutant progressive motor neuronopathy. 41. Messer, A., and L. Flaherty. 1986. Autosomal dominance in Nature 358:502–504. a late-onset motor neuron disease in the mouse. J. Neurogenet. 60. Haase, G., P. Kennel, B. Pettmann, et al. 1997. Gene 3:345–355. therapy of murine motor neuron disease using adenoviral 42. Mitsumoto, H., and W. G. Bradley. 1982. Murine motor vectors for neurotrophic factors. Nat. Med. 3:429–436. neuron disease (the wobbler mouse): degeneration and regen- 61. Sagot, Y., S. A. Tan, J. P. Hammang, et al. 1996. GDNF eration of the lower motor neuron. Brain 105(Pt. 4): slows loss of motoneurons but not axonal degeneration or pre- 811–834. mature death of pmn/pmn mice. J. Neurosci. 16:2335–2341. 43. Shultz, L. D., H. O. Sweet, M. T. Davisson, et al. 1982. 62. Mitsumoto, H., K. Ikeda, T. Holmlund, et al. 1994. The ‘Wasted’, a new mutant of the mouse with abnormalities char- effects of ciliary neurotrophic factor on motor dysfunction in acteristic to ataxia telangiectasia. Nature 297:402–404. wobbler mouse motor neuron disease. Ann. Neurol. 36: 44. Schmalbruch, H., H. J. Jensen, M. Bjaerg, et al. 1991. A 142–148. new mouse mutant with progressive motor neuronopathy. 63. Sendtner, M. 1997. Gene therapy for motor neuron disease. J. Neuropathol. Exp. Neurol. 50:192–204. Nat. Med. 3:380–381. 45. Pardo, C. A., B. A. Rabin, D. N. Palmer, et al. 1994. Accu- 64. Cork, L. C., J. W. Griffin, J. F. Munnell, et al. 1979. He- mulation of the adenosine triphosphate synthase subunit C reditary canine spinal muscular atrophy. J. Neuropathol. Exp. in the mnd mutant mouse. A model for neuronal ceroid Neurol. 38:209–221. lipofuscinosis. Am. J. Pathol. 144:829–835. 65. Lorenz, M. D., L. C. Cork, J. W. Griffin, et al. 1979. He- 46. Faust, J. R., J. S. Rodman, P. F. Daniel, et al. 1994. Two reditary spinal muscular atrophy in Brittany spaniels: clini- related proteolipids and dolichol-linked oligosaccharides ac- cal manifestations. J. Am. Vet. Med. Assoc. 175:833–839. cumulate in motor neuron degeneration mice (mnd/mnd), a 66. Sack, G. H. J., L. C. Cork, J. M. Morris, et al. 1984. Auto- model for neuronal ceroid lipofuscinosis. J. Biol. Chem. somal dominant inheritance of hereditary canine spinal mus- 269:10150–10155. cular atrophy. Ann. Neurol. 15:369–373. 47. Bronson, R. T., B. D. Lake, S. Cook, et al. 1993. Motor 67. Blazej, R. G., C. S. Mellersh, L. C. Cork, et al. 1998. He- neuron degeneration of mice is a model of neuronal ceroid reditary canine spinal muscular atrophy is phenotypically lipofuscinosis (Batten’s disease). Ann. Neurol. 33:381–385. similar but molecularly distinct from human spinal muscular 48. Bermingham, N. A., J. E. Martin, and E. M. Fisher. 1996. atrophy. J. Hered. 89:531–537. The mouse lysosomal membrane protein 1 gene as a candi- 68. Mellersh, C. S., A. A. Langston, G. M. Acland, et al. 1997. date for the motor neuron degeneration (mnd) locus. Genomics A linkage map of the canine genome. Genomics 46:326–336. 32:266–271. 69. Cork, L. C., D. L. Price, J. W. Griffin, et al. 1990. Heredi- 49. Falconer, D. S. 1956. Wobbler (WR). Mouse News 15:23. tary canine spinal muscular atrophy: canine motor neuron 50. Krieger, C., T. L. Perry, S. Hansen, et al. 1991. The wobbler disease. Can. J. Vet. Res. 54:77–82. mouse: amino acid contents in brain and spinal cord. Brain 70. Pinter, M. J., R. F. Waldeck, N. Wallace, et al. 1995. Motor Res. 551:142–144. unit behavior in canine motor neuron disease. J. Neurosci. 51. Tomiyama, M., K. Kannari, J. Nunomura, et al. 1994. 15:3447–3457. Quantitative autoradiographic distribution of glutamate re- 71. Ostrander, E. A., and E. Giniger. 1997. Semper fidelis: what ceptors in the cervical segment of the spinal cord of the wobbler man’s best friend can teach us about human biology and dis- mouse. Brain Res. 650:353–357. ease. Am. J. Hum. Genet. 61:475–480. 52. Kaupmann, K., D. Simon-Chazottes, J. L. Guenet, et al. 72. Cummings, J. F., A. de Lahunta, C. George, et al. 1990. 1992. Wobbler, a mutation affecting motoneuron survival and Equine motor neuron disease: a preliminary report. Cornell gonadal functions in the mouse, maps to proximal chromo- Vet. 80:357–379. some 11. Genomics 13:39–43. 73. Divers, T. J., H. O. Mohammed, J. F. Cummings, et al. 53. Wichmann, H., H. Jockusch, J. L. Guenet, et al. 1992. 1994. Equine motor neuron disease: findings in 28 horses and The mouse homolog to the ras-related yeast gene YPT1 maps proposal of a pathophysiological mechanism for the disease. on chromosome 11 close to the wobbler (WR) locus. Mamm. Equine Vet. J. 26:409–415. Genome 3:467–468. 74. Divers, T. J., H. O. Mohammed, and J. F. Cummings. 1997. 54. Lutsep, H. L., and M. Rodriguez. 1989. Ultrastructural, Equine motor neuron disease. Vet. Clin. North Am.: Equine morphometric, and immunocytochemical study of anterior Pract. 13:97–105. horn cells in mice with “wasted” mutation. J. Neuropathol. 75. Cummings, J. F., A. de Lahunta, H. O. Mohammed, et al. Exp. Neurol. 48:519–533. 1995. Endothelial lipopigment as an indicator of alpha-toco- 55. Padilla, M., C. Libertin, C. Krco, et al. 1990. Radiation pherol deficiency in two equine neurodegenerative diseases. sensitivity of T-lymphocytes from immunodeficient “wasted” Acta Neuropathol. (Berlin) 90:266–272. mice. Cell. Immunol. 130:186–194. 76. de la Rua-Domenech, R., M. Wiedmann, H. O. 56. Chambers, D. M., J. Peters, and C. M. Abbott. 1998. The Mohammed, et al. 1996. Equine motor neuron disease is not lethal mutation of the mouse wasted (wst) is a deletion that linked to Cu/Zn superoxide dismutase mutations: sequence abolishes expression of a tissue-specific isoform of transla- analysis of the equine Cu/Zn superoxide dismutase cDNA. tion elongation factor 1alpha, encoded by the Eef1a2 gene. Gene 178:83–88. Proc. Natl. Acad. Sci. USA 95:4463–4468. 77. de la Rua-Domenech, R., H. O. Mohammed, J. F. 57. Brunialti, A. L., C. Poirier, H. Schmalbruch, et al. 1995. Cummings, et al. 1997. Intrinsic, management, and nutri- The mouse mutation progressive motor neuronopathy (pmn) tional factors associated with equine motor neuron disease. maps to chromosome 13. Genomics 29:131–135. J. Am. Vet. Med. Assoc. 211:1261–1267.

487