The use of transgenic mouse models of amyotrophic lateral sclerosis in preclinical drug studies

Journal of Neurological Sciences 152 Suppl. 1 (1997) S67–S73

The use of transgenic mouse models of amyotrophic lateral sclerosis in preclinical drug studies Mark E. Gurney* CNS Diseases Research Unit, Pharmacia Upjohn Inc., Kalamazoo, MI 49001, USA

Abstract The discovery of mutations in the human SOD1 gene encoding Cu,Zn superoxide dismutase (Cu,Zn SOD) in patients with familial amyotrophic lateral sclerosis (ALS) has made possible the development of etiological models of the disease. Expression of mutant SOD1 genes in transgenic mice causes a progressive paralytic disease whose general features resemble ALS in humans. We have used the transgenic model to explore etiological mechanisms and to screen potential therapeutics. Our results and those of others show that familial ALS mutations cause a gain-of-function in Cu,Zn SOD that enhances the generation of damaging oxygen radicals. This may render motor neurons sensitive to the excitotoxic effects of ambient glutamate, as a putative glutamatergic inhibitor such as riluzole has therapeutic efficacy both in the transgenic model and in human ALS. This finding highlights the utility of the SOD1-G93A transgenic mouse model for preclinical drug studies in ALS.  1997 Elsevier Science B.V. Keywords: Transgenic mice; Superoxide dismutase; Riluzole; Amyotrophic lateral sclerosis

1. Introduction With the discovery, in patients with the familial form of amyotrophic lateral sclerosis (ALS), of mutations in the SOD1 gene encoding Cu,Zn superoxide dismutase (Cu,Zn SOD), came the promise that the mutant gene could be used to produce a transgenic model of the disease (Rosen et al., 1993). Of course, success in this endeavor would depend on the nature of the SOD1 mutations. Were they loss-of-function mutations, as suggested by the decrease in Cu,Zn SOD activity measured in patient tissues and red blood cells (Bowling et al., 1993; Deng et al., 1993), or were they gain-of-function mutations as suggested by the fact that the mutations were predominantly missense mutations that substitute one amino acid for another? Since no deletions of the SOD1 gene were found in familial ALS kindreds, this implied that the presence of the mutant protein was required for pathogenesis. The experiments to distinguish between the competing hypotheses of loss- versus gain-of-function were quickly completed. Expression of mutant human Cu,Zn SOD in transgenic mice at sufficient levels causes motor neuron *Tel.: 11 616 8330925; fax: 11 616 8332525; e-mail: [email protected] 0022-510X / 97 / $17.00  1997 Elsevier Science B.V. All rights reserved. PII S0022-510X( 97 )00247-5

disease, while expression of wild-type human Cu,Zn SOD at similar levels does not (Gurney et al., 1994; Ripps et al., 1995; Wong et al., 1995). Since some of the mutations explored in this way, such as the G93A and G37R mutations, have little effect on Cu,Zn SOD enzymatic activity, both wild-type and mutant transgenic lines show similar elevation in total dismutase activity (Borchelt et al., 1994; Gurney et al., 1994; Wong et al., 1995; WiedauPazos et al., 1996). Yet, only the mutant transgenic lines develop disease which implies that mutation has caused some pathogenic alteration in the protein. Conversely, knock-out experiments in mice fail to cause motor neuron disease. Disruption of the mouse SOD1 gene by homologous recombination has allowed the production of heterozygous ‘KO’ mice that have 50% of normal levels of Cu,Zn SOD as well as homozygous null mice which lack Cu,Zn SOD completely (Reaume et al., 1996). Both types of mice are viable and live well past 1 year of age. Female homozygous null mice are sterile, apparently because the uterine wall fails to support implantation, but neither type of mouse develops clinically overt motor neuron disease. Cu,Zn SOD is the cell’s first-line defense against superoxide-initiated, oxidative damage so one might expect such animals to be more sensitive to injury or oxidative stress. Indeed, that appears to be the case as

S68

M.E. Gurney / Journal of Neurological Sciences 152 Suppl. 1 (1997) S67 –S73

motor neurons in homozygous SOD1 null mice are more sensitive to axotomy. Thus, these manipulations of the mouse genome support the view that familial ALS is caused by gain-of-function mutations in Cu,Zn SOD. Since mice have such short life spans in comparison to man, we cannot rule out that a 50% decrease in Cu,Zn SOD levels does not have pathogenic consequences in man. Perhaps over the 40 or so years that elapse before clinical disease develops in persons with SOD1 mutations, there is sufficient time for a partial reduction in anti-oxidant systems to lead to disease. Alternatively, the reduction in antioxidant defences may act as cofactor for the pathogenic action of mutant Cu,Zn SOD. Although, the recessive SOD1 mutation of D90A described in Sweden apparently has no discernible effect on tissue levels of Cu,Zn SOD or Cu,Zn SOD enzymatic activity, yet still causes motor neuron disease (Andersen et al., 1995). This argues that, as in the murine transgenic model, familial ALS is caused by gain-of-function mutations in Cu,Zn SOD.

2. How do SOD1 mutations cause motor neuron disease? Our current view is that SOD1 mutations allow the mutant enzyme to catalyze alternative reactions that generate damaging free radicals. The copper catalytic center of Cu,Zn SOD is buried deep within an active site channel ˚ at the surface of the protein to which narrows from 24A ˚ in the depth of the channel. The channel also is lined 10A by three positively charged lysine or arginine residues which provide electrostatic guidance to the normal substrate, superoxide (O ?2 ), and exclude nonpolar and positively charged substrates (Getzoff et al., 1992). Since the more than 50 missense mutations found in familial ALS primarily hit the polypeptide backbone of Cu,Zn SOD (Deng et al., 1993; Siddique and Deng, 1996), they may relax spatial constraints on the width of the active site channel and unlock the natural chemistry of the copper catalytic center. This normally is prevented from catalyzing Fenton type reactions by being kept in the Cu 21 oxidation state. However, SOD1 mutations alter the redox behavior of the protein and facilitate reduction of the Cu 21 to Cu 1 by ascorbate and perhaps other anionic reductants (Lyons et al., 1996). In addition, familial ALS mutations alter the zinc binding site and thereby decrease the affinity of zinc for the apoenzyme (Lyons et al., 1996). Loss of zinc may also alter the conformation of the active site channel. One consequence of such conformational changes is that familial ALS mutations enhance Fenton catalysis of hydroxyl radical formation from hydrogen peroxide (Wiedau-Pazos et al., 1996; Yim et al., 1996). This creates a bound pro-oxidant, Cu 21 –OH ? , which normally oxidizes a histidine residue in the active site leading to inactivation of the enzyme (Hodgson and Fridovich, 1975). However,

anionic scavengers such as formate, glutamate, and chlorate can protect the enzyme from inactivation by reacting with the bound pro-oxidant in competition with the active site histidine (Hodgson and Fridovich, 1975; Yim et al., 1990, 1993). This has the potential to generate a cascade of secondary free radical products which could oxidize proteins, lipids, or nucleic acids. In addition to enhancing free radical generation, SOD1 mutations in familial ALS may also facilitate peroxynitrite-mediated catalysis of protein nitration by Cu,Zn SOD (Beckman et al., 1993).

3. Selection of animal models for preclinical drug studies In addition to the transgenic model of familial ALS described above, several other models of motor neuron disease have been described. These include the murine mutants wobbler (Duchen and Strich, 1968), motor neuron disease, MND (Messer et al., 1987) and peripheral motor neuronopathy, PMN (Schmalbruch et al., 1991), as well as transgenic mice that express human neurofilament subunits at elevated levels (Cote et al., 1993; Xu et al., 1993). Wobbler and PMN mice have been used for the preclinical evaluation of various neurotrophic factors (Sendtner et al., 1992; Mitsumoto et al., 1994; Sagot et al., 1996). Studies of these various models have focused on the extent to which they reproduce the pathological features of human ALS. Is there motor neuron loss? Are there alterations of the neuronal cytoskeleton leading to accumulation of phosphorylated neurofilaments in the motor neuron cell body or spheroids in the proximal axon? Are there Lewy-like bodies in ventral horn neurons? Is there fragmentation of the Golgi apparatus as detected with a monoclonal antibody to a Golgi component, etc.? Such studies are largely descriptive in nature and reveal stronger or weaker parallels with the human disease. Extensive study of SOD1 transgenic lines expressing different levels of mutant human Cu,Zn SOD indicates that the tempo and timing of disease varies systematically with protein expression as does the character of the pathology observed in the spinal cord. In SOD1-G93A or SOD1G37R transgenic mice, the high expressors develop a primarily vacuolar pathology which differs qualitatively from the pathology seen in ALS spinal cord (Dal Canto and Gurney, 1994; Wong et al., 1995). Disease has a rapid tempo with death occurring by 4–5 months of age (Chiu et al., 1995). Both endoplasmic reticulum and the mitochondria are affected. Changes in the mitochondria feature an unusual splitting of the inner and outer mitochondrial membranes (Dal Canto and Gurney, 1994). A third membranous organelle, the Golgi apparatus, also is affected. The murine disease causes fragmentation of the Golgi much like that previously reported in ALS (Mourelatos et al., 1996). Fragmentation of the Golgi may precede the vacuolar changes, and both of these precede motor neuron

M.E. Gurney / Journal of Neurological Sciences 152 Suppl. 1 (1997) S67 –S73

loss (Chiu et al., 1995; Mourelatos et al., 1996). In addition to the vacuolar changes, gliosis, cytoplasmic ubiquitination, axonal swellings, somatic phosphorylation of neurofilaments, occasional spheroids, and Lewy-like bodies are observed (Dal Canto and Gurney, 1995; Tu et al., 1996, 1997). There also is Wallerian degeneration of the ascending tracts in the sensory columns in a pattern reminiscent of familial ALS (Gurney et al., 1994; Chiu et al., 1995). That pattern of dorsal column involvement is not seen in sporadic ALS (Hirano et al., 1967). Muscle denervation, compensatory axonal sprouting, decreases in compound muscle action potentials, increases in spontaneous fibrillation potentials, and decreases in the number of motor units all keep pace with the loss of motor neurons (Chiu et al., 1995; Azzouz et al., 1997). In addition, there is loss of midbrain dopaminergic neurons as also is found in familial ALS (Kostic et al., 1997). This observation is consistent with the known sensitivity of dopaminergic neurons to oxidative attack. In contrast to the high expressors, transgenic SOD1-G93A mice that express the mutant protein at lower levels have a slower tempo of disease with protracted survival as long as 1 year. Pathology in the spinal cord of these mice more closely resembles human ALS with an absence of vacuolar changes. The mice show loss of motor neurons, accompanying gliosis, and the presence of many ubiquitinated Lewy-like bodies and axonal swellings (Dal Canto and Gurney, 1995, 1997). Which type of mutant SOD1 transgenic mouse more closely resembles human ALS? Certainly on the basis of pathology the lines of mice that express lower levels of mutant human Cu,Zn SOD and which have a more protracted course of disease are the best mimics of human ALS. Which type of mouse is more useful for preclinical drug studies? Probably the high expressors as the shorter course of disease makes more feasible the relatively rapid evaluation of potential therapeutics. Also, demonstration of therapeutic benefit in the high expressors probably provides the most stringent criterion for predicting success in the clinic. Given the expense and time required to organize human clinical trials, only the best candidate drugs should be brought forward for evaluation in patients. The underlying assumption, of course, is that disease in both types of mice has a common etiology even though details of the pathology differ in character. Studies of spontaneous mutations in mice that cause motor neuron disease suffer from a lack of knowledge of the disease gene. Since familial ALS is genetically heterogeneous (Siddique et al., 1991), and since SOD1 mutations only account for 20% of familial cases, it may well be the case that mutations in the wobbler, MND or PMN genes also occur in man and also cause motor neuron disease. However, we do not yet have those data. Wobbler has been used extensively for the preclinical evaluation of neurotrophic factors in motor neuron disease. The experiments have had great success. Ciliary neurotrophic factor (CNTF), brain-derived neurotrophic factor (BDNF), and

S69

their combination transiently preserve motor function in the wobbler mouse model (Mitsumoto et al., 1994). As reported at meetings, however, similar dosing regimens for CNTF, BDNF, or CNTF with BDNF, fail to extend survival or transiently preserve motor function in mutant SOD1-G93A transgenic mice. Nor have CNTF or BDNF had success in human clinical trials in ALS. Why the neurotrophic factors should be successful in one murine model but not another and why they failed in the clinic is unclear. Certainly delivery is a major issue. Although the axon of the motor neuron lies in the periphery, its cell body lies in the spinal cord behind the blood–brain barrier. Neurotrophic factors delivered in the periphery have access to the motor nerve terminal and can be retrogradely transported to the cell body, but is that transport system sufficient to meet the requirements for delivery of a therapeutic? Also, pathology in wobbler mice is fairly restricted, while it extends into the brain stem in mutant SOD1-G93A transgenic mice, and has a central component in human ALS. Damage occurring in those compartments may have been inaccessible to neurotrophic factor therapy. Transgenic mice expressing elevated levels of human neurofilament (NF) subunits have been used in imaginative ways to explore pathogenic mechanisms. Expression of the human NF-L subunit at high levels causes rapidly progressive motor neuron disease and death by 3–4 weeks of age (Xu et al., 1993). Expression of human NF-H also causes motor neuron loss, but only over an extended period of greater than 1 year (Cote et al., 1993). The NF-H transgenic mice develop massive accumulation of neurofilaments in the motor neuron cell body which apparently are sufficient to exclude the transport within the axon of organelles such as mitochondria (Collard et al., 1995). A failure in energy production and / or axonal transport could explain the loss of motor neurons in these mice. To date, NF transgenic mice have not been used for preclinical drug studies.

4. Experience with preclinical drug evaluation in SOD1-G93A transgenic mice Initial studies of therapeutics in the SOD1-G93A transgenic mice were organized through a collaborative working group with sponsorship by the Muscular Dystrophy Association (MDA). In addition, the transgenic mice were distributed to the majority of companies with drug discovery efforts in ALS, as well as to the Jackson Laboratory for distribution through the Induced Mutant Resource (IMR) to the research community by anonymous request. Characteristics of the SOD1 transgenic mice available through the IMR are given in Table 1 and Fig. 1. TgN(SOD1G93A)1Gur mice contain the mutant human SOD1 gene and develop motor neuron disease. As a control for biochemical studies, TgN(SOD1)2Gur mice are also avail-

S70

M.E. Gurney / Journal of Neurological Sciences 152 Suppl. 1 (1997) S67 –S73

Table 1 Characteristics of SOD1 transgenic lines deposited with the Induced Mutant Resource (The Jackson Laboratory) IMR designation

Laboratory designation

Transgene copy number a (per diploid genome)

Spinal cord Cu,Zn SOD a,b (U / mg)

TgN(SOD1)2Gur TgN(SOD1-G93A)1Gur B6SJL F 1 hybrid

N1029 c or N29 d G1H d

861.5 2561.5

7.762.5 8.962.7 1.360.3

a

Mean6SD. Xanthine / xanthine oxidase assay of Cu,Zn SOD activity measured at 90 days of age (Gurney et al., 1994). c Gurney et al. (1994). d Chiu et al. (1995). b

able. These contain a wild-type human SOD1 gene that does not cause disease. Both types of mice have similar elevation of superoxide dismutase activity in spinal cord since the G93A mutation has little effect on Cu,Zn SOD dismutase activity towards superoxide (Table 1). The first therapeutic success in preclinical drug studies with the mutant SOD1 transgenic model was obtained by the MDA collaborative working group (Gurney et al., 1996). An antioxidant diet containing elevated amounts of vitamin E and selenium was found to delay the onset of clinical disease and to maintain wheel running, while survival was not extended. In contrast, two putative glutamatergic inhibitors, riluzole and gabapentin, had no effect on disease onset, but did extend survival with riluzole having greater efficacy than gabapentin. Riluzole had about the same efficacy in the transgenic mice as seen in human ALS trials (Bensimon et al., 1994). This led us to develop the concept of ‘initiators’ and ‘propagators’ of disease in familial ALS. In this view, disease is initiated by oxidative attack on proteins or membrane lipids due to the generation of oxygen radicals by mutant Cu,Zn SOD. This damage accumulates over time until it reaches a threshold for disease. At that point, motor neurons become sensitive to the excitotoxic effects of ambient levels of glutamate.

Fig. 1. Kaplan-Meir analysis of cumulative survival for TgN(SOD1] mice. G93A)1Gur and TgN(SOD1-G93A)1Gur del

The cycle of oxygen radical production and glutamate excitotoxicity may become reinforcing (Pelligrini-Giampietro et al., 1990). Although glutamate excitotoxicity has been identified as a propagator of disease, the pathway of motor neuron death is not yet known. The endpoints used in the study described above were onset of disease as assessed by noting shaking in one or more limbs when the mouse is suspended in the air, behavioral output as assessed by spontaneous running in a wheel, and end-stage disease. The shaking in the limbs is associated with disinhibition of spinal reflexes and Wallerian degeneration in the descending spinal tracts (Chiu et al., 1995). Mice given access to wheels run amazing distances of 7–10 km per night. We find that running declines monotonically with increasing paralysis due to motor neuron loss, so the behavioral measure provides a quantifiable endpoint for assessing strength and mobility. End-stage disease is assessed by determining if the mouse can roll over when placed on its side. When mice fail that test, they should be euthanized by ethically approved procedures as they have reached a degree of impairment that prevents them from foraging for food or water. The riluzole data were included in support of the new drug application filed by Rhone-Poulenc-Rorer with the United States Food and Drug Administration. With such widespread distribution of this particular model, we can begin to assess its utility in the preclinical evaluation of ALS therapeutics. The clinical phenotype of the mutant SOD1 transgenics depends critically on the amount of mutant protein expressed in the individual mouse, Since high levels of expression generally are achieved with high copy number (if expression is dependent upon transgene copy number and not site of integration), the transgene locus may be unstable over time. When transgenics are made by microinjection of DNA into a fertilized mouse egg, generally only one or two integration events occur with a variable number of transgene copies integrating at each locus. If integration occurred on an autosome, then a Mendelian pattern of inheritance should be observed for the transgene locus. However, intralocus recombination events during meiosis may either expand or contract the number of transgene copies contained within the locus. Such recombination events occur with low frequency, but can have confusing or damaging consequences. For

M.E. Gurney / Journal of Neurological Sciences 152 Suppl. 1 (1997) S67 –S73

example, our first try at establishing the TgN(SOD1G93A)1Gur line at the IMR resulted in a line with reduced copy number which has since been designated TgN(SOD1] . With the fall in copy number from about 25 G93A)1Gur del SOD1 transgene copies per haploid genome to about eight, survival lengthened from 132611 days in the parental strain to 251628 in the deleted strain (Fig. 1). At Pharmacia Upjohn, in 1479 meiosis with transmission of the SOD1-G93A transgene by the heterozygote, we have found two additional intralocus rearrangements with a decrease in copy number. One of those also had occurred in a male selected for breeding, yet since we maintain our breeding records in a relational database, it was relatively simple to identify and remove his progeny from our colony. Our experience with intralocus rearrangement in high copy number strains has prompted us to implement two quality control criteria. First, we use a quantitative EIA to genotype all progeny for human Cu,Zn SOD expression. This allows us to identify gross intralocus rearrangements that substantially alter protein expression. Second, we monitor the survival of all animals selected for breeding. We discard the progeny of any mouse whose survival falls outside the 95% confidence interval that we have previously established for that particular transgenic line. A second concern in widespread use of the SOD1-G93A transgenic model is the robustness of the clinical phenotype in different laboratories. To a certain extent this depends upon the clinical endpoint chosen for assessment. For that reason, we have preferred to select robust endpoints such as survival, or objective, quantitative measures of clinical impairment such as running in a wheel. For example, the duration of survival of the TgN(SOD1G93A)1Gur line has been fairly stable across different laboratories, housing conditions, and animal facilities (Table 2). In practice, we have found treatment groups of 12 animals sufficient to detect a change in survival of about one SD of the mean in a two-tailed t-test with a significance level of 0.05. Behavioral assessments are expected to vary more widely across laboratories, but evaluation of SOD1-G93A for onset of clinical disease using as an endpoint shaking or tremor in one or more Table 2 Survival statistics for TgN(SOD1-G93A)1Gur mice in different laboratories, animal facilities, and institutions Mean6SD

Count

Location

Ref.

136615 132612 137611 13568 138610 138616

23 85 9 11 15 12

NWU PNU REGN MGH ULP COL

Gurney et al., 1996

Azzouz et al., 1997 Kostic et al., 1997

NWU, Northwestern University; PNU, Pharmacia Upjohn; REGN, Vivien Wong, Regeneron Pharmaceuticals; MGH, Flint Beal, Massachusetts General Hospital; ULP, Universite Louis Pasteur, France; COL, Columbia University.

S71

limbs yielded a similar result in two published studies by different laboratories (Gurney et al., 1996; Kostic et al., 1997). Running in a wheel is a quantitative measure of behavioral output that we find to be both robust and sensitive, but it requires investment in a computerized data collection system (Gurney et al., 1996).

5. Future avenues for therapeutic intervention in ALS Since multiple drugs can be tested cheaply and relatively rapidly in the transgenic model, what other types of therapeutics should be evaluated? Certainly combination therapy with antioxidants and glutamatergic inhibitors should be tried. Vitamin E is a potent antioxidant, but a poor drug due to its limited transport into the brain. Other types of antioxidants may have greater therapeutic benefit and should be screened. These include coenzyme Q10, procysteine, N-acetylcysteine, lipoic acid, and various types of nitrones. Copper chelators such as penicillamine which might reduce Cu,Zn SOD activity may also have therapeutic benefit. Inhibitors of nitric oxide production and scavengers of peroxynitrite also should be evaluated in the model as a test of the hypothesis proposed by Beckman et al. (1993). Riluzole has a variety of pharmacological actions in addition to inhibition of glutamate release (Wokke, 1996), such as inhibition of GABA reuptake, so it may be worthwhile to investigate GABA reuptake inhibitors and GABA receptor antagonists. In addition, inhibitors of the different types of ionotropic glutamate receptors should be evaluated for therapeutic benefit. Although neurotrophic factors have proved ineffective in ALS when administered by injection, delivery into the cerebrospinal fluid by pumps or cellular implants may prove more effective and could be evaluated in the transgenic model. Since we know so little about the actual pathway of motor neuron death in this model, it also will be interesting to test transgenic expression of bcl2 or crmA for protection in SOD1-G93A transgenic mice. Hopefully, one or more of these avenues of research will result in effective therapeutics for this devastating disease.

Acknowledgements The FALS Therapeutics Working Group was supported by a grant from the Muscular Dystrophy Association.

References Andersen, P.M., Nilsson, P., Ala-Hurula, V., Keranen, M.-L., Tarvainnen, I., Haltia, T., Nilsson, L., Binzer, M., Forsgren, L., Marklund, S.L., 1995. Amyotrophic lateral sclerosis associated with homozygosity for

S72

M.E. Gurney / Journal of Neurological Sciences 152 Suppl. 1 (1997) S67 –S73

an Asp90Ala mutation in CuZn-superoxide dismutase. Nature Genet. 10, 61–66. Azzouz, M., Leclerc, N., Gurney, M., Warter, J.-M., Poindron, P., Borg, J., 1997. Progressive motor neuron impairment in an animal model of familial amyotrophic lateral sclerosis. Muscle Nerve 20, 45–51. Beckman, J.S., Carson, M., Smith, C.D., Koppenol, W.H., 1993. ALS, SOD and peroxynitrite. Nature 364, 584. Bensimon, G., Lacomblez, L., Meininger, V., 1994. A controlled trial of riluzole in amyotrophic lateral sclerosis (in collaboration the ALS / Riluzole Study Group). New Engl. J. Med. 330, 585–591. Borchelt, D.R., Lee, M.K., Slunt, H.S., Guarnieri, M., Xu, Z.-S., Wong, P.C., Brown, R.H., Price, D.L., Sisodia, S.S., Cleveland, D.W., 1994. Superoxide dismutase 1 with mutations linked to familial amyotrophic lateral sclerosis possesses significant activity. Proc. Natl. Acad. Sci. USA 91, 8292–8296. Bowling, A.C., Schulz, J.B., Brown, Jr. R.H., Beal, M.F., 1993. Superoxide dismutase activity, oxidative damage, and mitochondrial energy metabolism in familial and sporadic amyotrophic lateral sclerosis. J. Neurochem. 61, 2322–2325. Chiu, A.Y., Zhai, P., Dal Canto, M.C., Peters, T.W., Kwon, Y.W., Prattis, S.W., Gurney, M.E., 1995. Age dependent penetrance of disease in a transgenic mouse model of familial amyotrophic lateral sclerosis. Mol. Cell. Neurosci. 6, 349–362. Collard, J.F., Cote, F., Julien, J.P., 1995. Defective axonal transport in a transgenic mouse model of amyotrophic lateral sclerosis. Nature 375, 61–64. Cote, F., Collard, J.F., Julien, J.P., 1993. Progressive neuronopathy in transgenic mice expressing the human neurofilament heavy gene: a mouse model of amyotrophic lateral sclerosis. Cell 73, 35–46. Dal Canto, M.C., Gurney, M.E., 1994. The development of CNS pathology in a murine transgenic model of human ALS. Am. J. Pathol. 145, 1271–1279. Dal Canto, M.C., Gurney, M.E., 1995. Neuropathological changes in two lines of mice carrying a transgene for mutant human Cu,Zn SOD, and in mice overexpressing wild type human SOD: a model of familial amyotrophic lateral sclerosis (FALS). Brain Res. 676, 25–40. Dal Canto, M.C., Gurney, M.E., 1997. A low expressor line of transgneic mice carrying a mutant human Cu,Zn superoxide dismutase (SOD1) gene develops pathological changes which most closely resemble those in human ALS. Acta Neuropathol. 93, 537–550. Deng, H.-X., Hentati, A., Tainer, J.A., Iqbal, Z., Cayabyab, A., Hung, W.-Y., Getzoff, E.D., Hu, P., Herzfeldt, B., Roos, R.P., Warner, C., Deng, G., Soriano, E., Smyth, C., Parge, H.E., Ahmed, A., Roses, A.D., Sidddique, T., 1993. Amyotrophic lateral sclerosis and structural defects in Cu,Zn superoxide dismutase. Science 261, 1047–1051. Duchen, L.W., Strich, S.J., 1968. An hereditary motor neuron disease with progressive denervation of muscle in the mouse: The mutant ‘wobbler’. J. Neurol. Neurosurg. Psychiatry 31, 535–547. Getzoff, E.D., Cabelli, D.E., Fisher, C.L., Parge, H.E., Viezzoli, M.S., Banci, L., Hallewell, R.A., 1992. Faster superoxide dismutase mutants designed by enhancing electrostatic guidance. Nature 358, 347–351. Gurney, M.E., Pu, H., Chiu, A.Y., Dal-Canto, M.C., Polchow, C.Y., Alexander, D.D., Caliendo, J., Hentati, A., Kwon, Y.W., Deng, H.-X., Chen, W., Zhai, P., Sufit, R.L., Siddique, T., 1994. Motor neuron degeneration in mice that express a human Cu,Zn superoxide dismutase mutation. Science 264, 1772–1775. Gurney, M.E., Cutting, F.B., Zhai, P., Doble, A., Taylor, C.P., Andrus, P.K., Hall, E.D., 1996. Benefit of Vitamin E, Riluzole, and Gabapentin in a transgenic model of familial amyotrophic lateral sclerosis. Ann. Neurol. 39, 147–157. Hirano, A., Kurland, L.T., Sayre, G.P., 1967. Familial amyotrophic lateral sclerosis. A subgroup characterized by posterior and spinocerebellar tract involvement and hyaline inclusions in the anterior horn cells. Arch. Neurol. 16, 232–243. Hodgson, E.K., Fridovich, I., 1975. The interaction of bovine erythrocyte superoxide dismutase with hydrogen peroxide: Inactivation of the enzyme. Biochemistry 14, 5294–5299.

Kostic, V., Gurney, M.E., Deng, H.-X., Siddique, T., Epstein, C.J., Przedborski, S., 1997. Midbrain dopaminergic neuronal degeneration in a transgenic model of familial amyotrophic lateral sclerosis. Ann. Neurol. 41, 497–504. Lyons, T.J., Liu, H., Goto, J.J., Nersissian, A., Roe, J.A., Graden, J.A., Cafe, C., Ellerby, L.M., Bredesen, D.E., Gralla, E.B., Valentine, J.S., 1996. Mutations in coper-zinc superoxide dismutase that cause amyotrophic lateral sclerosis alter the zinc binding site and the redoc behavior of the protein. Proc. Natl. Acad. Sci. USA 93, 12240–12244. Messer, A., Strominger, N.L., Mazurkiewicz, J.E., 1987. Histopathology of the late-onset motor neuron degeneration (Mnd) mutant in the mouse. J. Neurogenet. 4, 201–213. Mitsumoto, H., Ikeda, K., Klinkosz, B., Cedarbaum, J.M., Wong, V., Lindsay, R.M., 1994. Arrest of motor neuron disease in wobbler mice cotreated with CNTF and BDNF. Science 265, 1107–1110. Mourelatos, Z., Gonatas, N.K., Stieber, A., Gurney, M.E., Dal Canto, M.C., 1996. The Golgi apparatus of spinal cord motor neurons in transgenic mice expressing mutant Cu,Zn superoxide dismutase becomes fragmented in early, preclinical stages of the disease. Proc. Natl. Acad. Sci. USA 93, 5472–5477. Pelligrini-Giampietro, D.E., Cherici, G., Alesiani, M., Carla, V., Moroni, F., 1990. Excitatory amino acid release and free radical formation may cooperate in the genesis of ischemia-induced neuronal damage. J. Neurosci. 10, 1035–1041. Reaume, A.G., Elliott, J.L., Hoffman, E.K., Kowall, N.W., Ferrante, R.J., Siwek, D.F., Wilcox, H.M., Flood, D.G., Beal, M.F., Brown, Jr. R.H., Scott, R.W., Snider, W.D., 1996. Motor neurons in Cu / Zn superoxide dismutase-deficient mice develop normally but exhibit enhanced cell death after axonal injury. Nat. Genet. 13, 43–47. Ripps, M.E., Huntley, G.W., Hof, P.R., Morrison, J.H., Gordon, J.W., 1995. Transgenic mice expressing an altered murine superoxide dismutase gene provide an animal model of amyotrophic lateral sclerosis. Proc. Natl. Acad. Sci. U.S.A. 92, 689–693. Rosen, D.R., Siddique, T., Patterson, D., Figlewicz, D.A., Sapp, P., Hentati, A., Donaldson, D., Goto, J., O’Regan, J.P., Deng, H.-X., Rahamani, Z., Krizus, A., McKenna-Yasek, D., Cayabyab, A., Gaston, S.M., Berger, R., Tanzi, R.E., Halperin, J.J., Herzfeldt, B., Van den Bergh, R., Hung, W.-Y., Bird, T., Deng, G., Mulder, D.W., Smyth, C., Laing, N.G., Soriano, E., Pericak-Vance, M.A., Haines, J., Rouleau, G.A., Gusella, J.S., Horvitz, H.R., Brown, R.H., 1993. Mutations in Cu,Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature 362, 59–62. Sagot, Y., Tan, S.A., Hammang, J.P., Aebischer, P., Kato, A.C., 1996. GDNF slows loss of motoneurons but not axonal degeneration or premature death of pmn / pmn mice. J. Neurosci. 16, 2335–2341. Schmalbruch, H., Jensen, H.J., Bjaerg, M., Kamieniecka, Z., Kurland, L., 1991. A new mouse mutant with progressive motor neuronopathy. J. Neuropathol. Exp. Neurol. 50, 192–204. Sendtner, M., Schmalbruch, H., Stockli, K.A., Carroll, P., Kreutzberg, G.W., Thoenen, H., 1992. Ciliary neurotrophic factor prevents degeneration of motor neurons in mouse-mutant progressive motor neuronopathy. Nature 358, 502–504. Siddique, T., Deng, H.-X., 1996. Genetics of amyotrophic lateral sclerosis. Hum. Mol. Genet. 5, 1465–1470. Siddique, T., Figlewicz, D.A., Pericak-Vance, M.A., Haines, J.L., Rouleau, G., Jeffers, A.J., Sapp, P., Hung, W.Y., Bebout, J., McKennaYasek, D., Deng, G., Horvitz, H.R., Gusella, J.F., Brown, R.H., Roses, A.D., 1991. Linkage of a gene causing familial amyotrophic lateral sclerosis to chromosome 21 and evidence of genetic heterogeneity (written with other collaborators). New Engl. J. Med. 324, 1381–1384. Tu, P.H., Raju, P., Robinson, K.A., Gurney, M.E., Trojanowski, J.Q., Lee, V.M.Y., 1996. Transgenic mice carrying a human mutant superoxide dismutase transgene develop neuronal cytoskeletal pathology resembling human amyotrophic lateral sclerosis. Proc. Natl. Acad. Sci. USA 93, 3155–3160. Tu, P.H., Gurney, M.E., Julien, J.P., Lee, V.M.Y., Trojanowski, J.Q., 1997. Oxidative stress, mutant SOD1, and neurofilament pathology in

M.E. Gurney / Journal of Neurological Sciences 152 Suppl. 1 (1997) S67 –S73 transgenic mouse models of human motor neuron disease. Lab. Invest. 76, 441–456. Wiedau-Pazos, M., Goto, J.J., Rabizadeh, S., Gralla, E.B., Roe, J.A., Lee, M.K., Valentine, J.S., Bredesen, D.E., 1996. Altered reactivity of superoxide dismutase in familial amyotrophic lateral sclerosis. Science 271, 515–518. Wokke, J., 1996. Riluzole. Lancet 348, 795–799. Wong, P.C., Pardo, C.A., Borchelt, D.R., Lee, M.K., Copeland, N.G., Jenkins, N.A., Sisodia, S.S., Cleveland, D.W., Price, D.L., 1995. An adverse property of a familial ALS-linked SOD1 mutation causes motor neuron disease characterized by vacuolar degeneration og mitochondria. Neuron 14, 1105–1116. Yim, M.B., Chock, P.B., Stadtman, E.R., 1990. Copper,zinc superoxide dismutase catalyzes hydroxyl radical production from hydrogen peroxide. Proc. Natl. Acad. Sci. USA 87, 5006–5010.

S73

Yim, M.B., Chock, P.B., Stadtman, E.R., 1993. Enzyme function of copper, zinc superoxide dismutase as a free radical generator. J. Biol. Chem. 268, 4099–4105. Yim, M.B., Kang, J.-H., Yim, H.-S., Kwak, H.-S., Chock, P.B., Stadtman, E.R., 1996. A gain-of-function of an amyotrophic lateral sclerosisassociated Cu,Zn-superoxide dismutase mutant: An enhancement of free radical formation due to a decrease in Km for hydrogen peroxide. Proc. Natl. Acad. Sci. USA 93, 5709–5714. Xu, Z., Cork, L.C., Griffin, J.W., Cleveland, D.W., 1993. Increased expression of neurofilament subunit NF-L produces morphological alterations that resemble the pathology of human motor neuron disease. Cell 73, 23–33.