Neuroprotective effects of creatine in a transgenic animal model of amyotrophic lateral sclerosis
Peter Klivenyi1, Robert J. Ferrante2, 3, Russell T. Matthews1, Mikhail B. Bogdanov1, Autumn M. Klein2, 3, Ole A. Andreassen1, Gerald Mueller1, Marieke Wermer1, Rima Kaddurah-Daouk4
& M. Flint Beal1, 5
1 Neurochemistry Laboratory, Neurology Service,Massachusetts General Hospital and Harvard Medical School, 32 Fruit Street, Boston, Massachusetts 02118, USA
2 Departments of Neurology, Pathology and Psychiatry, Boston University School of Medicine, 715 Albany Street, Boston, Massachusetts
02118, USA
3 Department of Veterans Affairs, 200 Spring Road,
Bedford, Massachusetts 01730, USA
4 The Avicena Group, Inc., One Broadway, Suite 600,
Cambridge, Massachusetts 02142, USA
5 Department of Neurology and Neuroscience, Cornell University Medical College, 525 East 68th Street, New York, New York
10021, USA
Mitochondria are particularly vulnerable to oxidative stress, and mitochondrial swelling and vacuolization are among the earliest pathologic features found in two strains of transgenic amyotrophic lateral sclerosis (ALS) mice with SOD1 mutations1,
2. Mice with the G93A human SOD1 mutation have altered electron transport enzymes, and expression of the mutant enzyme in vitro results in a loss of mitochondrial membrane potential and elevated cytosolic calcium concentration3. Mitochondrial dysfunction may lead to ATP depletion, which may contribute to cell death. If this is true, then buffering intracellular energy levels could exert neuroprotective effects. Creatine kinase and its substrates creatine and phosphocreatine constitute an intricate cellular energy buffering and transport system connecting sites of energy production (mitochondria) with sites of energy consumption4, and creatine administration stabilizes the mitochondrial creatine kinase and inhibits opening of the mitochondrial transition pore5. We found that oral administration of creatine produced a dose-dependent improvement in motor performance and extended survival in G93A transgenic mice, and it protected mice from loss of both motor neurons and substantia nigra neurons at 120 days of age. Creatine administration protected G93A transgenic mice from increases in biochemical indices of oxidative damage. Therefore, creatine administration may be a new therapeutic strategy for ALS.
Oral administration of 1% creatine or 2% creatine in the diet resulted in dose-dependent, significant improvements in the survival of G93A mice compared with the survival of mice fed unsupplemented diets (Fig. 1a and b). The mean survival in mice on unsupplemented diets increased from 143.7 2.3 days to 157.2 2.8 days with 1% creatine (P < 0.05) and to 169.3 4.7 days with 2% creatine (P < 0.001). Survival was extended by 13 days with 1% creatine and by 26 days with 2% creatine, which is better than the improvement with riluzole, which extends survival by 13 days in this model6. Carboxyfullerenes improve survival by 9 days7, whereas penicillamine improves survival by 10 days8. Therefore, of the pharmacologic interventions assessed in this model so far, creatine administration results in the best improvement in survival. We also determined the effects of 1% and 2% creatine on rotorod performance (Fig. 1b and c). Mice fed creatine supplementation had significantly better performance from 116 to 136 days of age than mice fed unsupplemented diets. There were no significant differences in motor performance with 2% creatine between 70 and 110 days of age (data not shown).
Figure 1. a and b, Effects of 1% or 2% creatine supplementation on survival in G93A transgenic mice.
a, Cumulative probability for survival. b, Mean survival. and , control (unsupplemented diet); , 1% creatine diet; and , 2% creatine diet. Survival was significantly increased in mice fed creatine. *, P < 0.05; #, P < 0.001. c and d, Effects of 1% (c) and 2% (d) creatine supplementation on rotorod performance. There was improved performance with creatine supplementation () at most time points between 116 and 136 days of age. RPM, revolutions per minute of rod. *, P < 0.05, compared with G93A transgenic mice fed normal diets ().
To determine whether creatine supplementation exerts neuroprotective effects, we counted neurons in G93A mice fed either unsupplemented diets or creatine-supplemented diets and their littermate controls, killed at 120 days of age. There was significant neuronal loss (49.3% reduction) in the ventral horns of G93A transgenic mice compared with G93A nontransgenic littermate controls (Table and Fig. 2). Large ventral horn neurons were the most profoundly affected (95% loss); medium and small neurons were less affected (34.5% and 47.2% loss, respectively). In contrast, G93A transgenic mice fed 1% creatine showed complete protection and did not differ significantly from control mice (Table and Fig. 2).
Figure 2. Neuronal loss in the ventral horns of the lumbar spinal cord.
G93A transgenic mouse (a), G93A transgenic mouse fed 1% creatine (b) and a nontransgenic littermate control (c), at 120 days of age. There is profound neuronal loss in the ventral horn of the G93A transgenic mouse. In contrast, there are no substantial cellular differences between the G93A transgenic mouse fed 1% creatine and the nontransgenic littermate control. Scale bar represents 100 m.
We determined whether the creatine effects might be due to an effect on muscle. There were no substantial differences in the weight of the hind limb extensor muscle at 84 days of age between control and G93A mice fed either 2% creatine or unsupplemented diets. For control mice fed a normal diet, the percentage of body weight for this muscle was 0.76 0.04%; for control mice fed creatine, 0.66 0.06%; for G93A mice fed a normal diet, 0.70 0.06%; and for G93A mice fed creatine, 0.61 0.03%. By histomorphometry of muscle at the same time point, there were no substantial differences in muscle fiber diameter for either control or G93A mice fed 2% creatine or unsupplemented diets. There were no substantial differences in body weight between the G93A mice fed 2% creatine (24.7 2.3 g) and those fed unsupplemented diets (27.5 2.6 g) at 84 days of age. There were no substantial differences in body weight between mice fed 2% creatine and those fed unsupplemented diets at time points between 70 and110 days of age (data not shown).
We also confirmed the observation of a significant loss of tyrosine hydroxylase (TH)-positive neurons in the substantia nigra at end-stage in G93A transgenic mice9. At 120 days of age, there was a substantial loss of neurons immunopositive for Nissl (23.8%), TH (27.5%) and dopamine transporter (DAT) (31.8%) in the substantia nigra pars compacta of G93A transgenic mice compared with littermate controls (Table). The G93A mice fed 1% creatine showed no significant neuronal loss compared with the littermate controls (Table). These results also indicate that energy dysfunction is involved in the loss of the substantia nigra neurons, and that creatine administration might be a useful therapeutic strategy for Parkinson disease.
Creatine supplementation (2%) in the G93A mice resulted in increased creatine levels compared with those in both wild-type and G93A mice fed unsupplemented diets (P < 0.01 for both comparisons). The creatine levels in control mice, G93A mice fed normal diets and G93A mice fed 2% creatine were 51.2 3.5 mmol/g protein, 48.5 4.7 mmol/g protein and 66.8 13.8 mmol/g protein, respectively.
We determined whether creatine supplementation could affect oxidative injury in G93A mice. The levels of 3-nitrotyrosine are significantly increased in the spinal cords of G93A mice10. At 120 days of age, mice fed 1% creatine showed no increase in spinal cord 3-nitrotyrosine levels compared with mice fed unsupplemented diets (Fig. 3). We also assessed free-radical generation in vivo using microdialysis. Administration of the mitochondrial toxin 3-nitropropionic acid results in a significant increase in the conversion of salicylate to 2,3-DHBA in the striatum, which is blocked in mice overexpressing Cu, Zn SOD (11). Here we found that systemic administration of 3-nitropropionic acid (3-NP) resulted in a significant increase in the conversion of 4-HBA to 3,4-DHBA in G93A transgenic mice fed unsupplemented diets (Fig. 4). In mice fed 1% creatine-supplemented diets, there was no significant increase in 3,4DHBA/4HBA after 3-NP administration.
Figure 3. Effects of 1% creatine supplementation starting at 70 days of age on spinal cord
3-nitrotyrosine/tyrosine concentrations at 120 days of age.
, wild-type mice; , G93A transgenic mice fed a normal diet; , G93A transgenic mice fed a 1%-creatine diet. *, P < 0.05, compared withlittermate controls; #, P < 0.05, compared with creatine supplementation.
Figure 4. Effect of 1% creatine supplementation on the conversion of 4HBA to 3,4DHBA in G93A transgenic mice after systemic administration of 3-NP.
Administration of 3-NP resulted in a significant increase in the ratio of 3,4DHBA to 4HBA in mice fed normal diets, which was significantly attenuated in mice receiving 1% creatine supplementation. , basal levels; , after 3-NP. *, P < 0.05, compared with basal levels; #, P < 0.05, compared with mice treated with 3-NP and fed a control diet.
The means by which SOD1 mutations cause damage to motor neurons may involve a complex interplay between oxidative damage, mitochondrial dysfunction and excitotoxicity. There are abnormalities in mitochondria in both sporadic ALS and familial ALS (refs. 12,13). If mitochondrial dysfunction contributes to the pathogenesis of motor neuron loss in ALS, then creatine administration may both help to buffer intracellular energy stores and inhibit mitochondrial transition pore opening, which is linked to both excitotoxic and apoptotic cell death14. Creatine stabilizes mitochondrial creatine kinase in an octomeric form, which inhibits the opening of the mitochondrial transition pore by calcium15. Creatine administration can also stimulate mitochondrial respiration and phosphocreatine synthesis5. Phosphocreatine diffuses to the cytoplasm, where it serves as both a temporal and spatial energy buffer, maintaining ATP levels used by the Na+/K+ ATPase and the Ca2+ ATPase4. Its importance in brain function is supported by in vivo31P-NMR transfer measurements showing correlations between creatine kinase flux and brain activity measured by EEG, as well as with brain 2-deoxyglucose uptake16,
17. Phosphocreatine also serves as a direct energy source for glutamate uptake into synaptic vesicles18, and impaired glutamate uptake has been implicated in the pathogenesis of sporadic ALS, as well as in transgenic animal models of ALS (19).
The therapeutic benefit of creatine on survival and motor function in G93A SOD1 transgenic mice provides further evidence that mitochondrial dysfunction and oxidative damage may play a part in the pathogenesis of ALS. Creatine is well tolerated in man, and is reported to be beneficial in several neurologic diseases20,
21. Our results indicate that creatine supplementation might be a new therapeutic strategy for the treatment of ALS.
Methods Mice. Transgenic mice with the G93A human SOD1 mutation (G1H/+) line were obtained from Jackson Laboratories2 (Bar Harbor, Maine). G93A mice were bred with female littermates. The offspring were genotyped by PCR assay of DNA obtained from tail tissue. Transgenic mice were housed in micro-isolator cages in a modified barrier facility, and were seronegative for common mouse and bacterial pathogens. Mice were fed either a diet supplemented with 1% creatine or 2% creatine starting at 70 days of age. Six mice fed unsupplemented diets were compared with seven mice fed 1% creatine and seven mice fed 2% creatine.
Behavioral testing (rotorod). Mice were given two days to become acquainted with the rotorod apparatus (Columbus Instruments, Columbus, Ohio). Then testing began with the mice trying to stay on a rod that was rotating at 1 rpm. The speed was then increased by 1 rpm every 10 seconds until the mouse fell off. Each mouse was given three trials. The speed of rod rotation at which the mouse fell off was used as the measure of competency on this task. Mice were tested every other day until they could no longer perform the task.
Survival. G93A transgenic mice initially show a high-frequency resting tremor, which progresses to gait abnormalities, paralysis of the hindlimbs, then to paralysis of the forelimbs, and finally to complete paralysis. Mice were killed when they could no longer roll over within 10 seconds of being pushed on their side. This time point was used as the time of death.
Histologic evaluation. G93A transgenic mice, ten fed 1% creatine starting at 70 days of age and nine fed a normal diet, and seven G93A nontransgenic littermate control mice were killed at 120 days of age for histologic evaluation. These mice were deeply anesthetized and then transcardially perfused with 4% buffered paraformaldehyde. The brains and spinal cords were removed, post-fixed with the perfusant for 2 h, and cryoprotected in a graded series of 10% and 20% glycerol/2% DMSO solution. The brain and spinal cord tissue specimens were subsequently cut on a cryostat into sections 50 m in thickness and stained for Nissl and immunohistochemical markers as described10. Cut tissue sections of the midbrain were immunostained for tyrosine hydroxylase (TH antisera, 1:1,000 dilution; Eugene Tech International, Ridgefield, New Jersey) and dopamine transporter (DAT antisera, 1:500 dilution; Chemicon International, Temecula, California). DAT labels neurons of the substantia nigra pars compacta21. Midbrain sections (two to three microscopic sections) from each mouse, through both the left and right substantia nigra pars compacta from the Bregma levels 3.08 mm to 3.16 mm and intraaural levels 0.72 mm to 0.64 mm, were analyzed by microscopic videocapture. Nissl-, TH- and DAT-positive neurons were counted using Neurolucida (Microbrightfield, Colchester, Vermont) image analysis software. These counts were within a homogenous structure, making the tenets of stereology valid. Nissel-stained neurons were also counted (using the Neurolucida system) at a magnification of 250 in both ventral horn areas from six tissue sections of the lumbar spinal cord of each mouse, with size discrimination into diameter classes of >25 m (L, large), between 25 and15 m (M, medium) and <15 microns (S, small). All cells were counted from within the ventral horn below a lateral line across the spinal cord from the central canal. Correction for tissue section thickness was made in all specimens.
Muscle evaluation. Control and G93A mice (three to five mice per group) fed either 2% creatine or unsupplemented diets were killed at 84 days of age for measurements of muscle mass and for muscle histomorphometry. The left extensor muscles (quadriceps and tibialis anterior) were dissected and weighed. The right extensor muscles were post-fixed in 4% buffered formaldehyde. Muscle samples were embedded in paraffin and cross-sections of muscle 10 mm in thickness were cut and stained with hematoxylin and eosin.
3-Nitrotyrosine measurements. G93A transgenic mice and littermate controls (eight mice per group) were fed 1% creatine or unsupplemented diets at 70 days of age and then killed at 120 days of age for spinal cord measurements of 3-nitrotyrosine as described10.
Measurements of creatine. G93A transgenic mice, littermate control mice or G93A transgenic mice (eight mice per group) fed 1% creatine starting at 70 days of age were killed at 120 days of age by the freeze-clamp procedure for measurements of creatine22.
Microdialysis studies. G93A transgenic mice fed either a control diet or a 1%-creatine diet (six mice per group) starting at 70 days of age underwent microdialysis at 120 days of age as described23, to assess free radical generation.
Statistical analysis. Statistical comparisons were by one-way analysis of variance (ANOVA) or by repeated measures ANOVA followed by Fisher's least protected significant difference test.
Received 1 November 1998; Accepted 18 January 1999
REFERENCES
Wong, P.C. et al. An adverse property of a familial ALS-linked SOD1 mutation causes motor neuron disease characterized by vacuolar degeneration of mitochondria. Neuron14, 1105-1116 (1995). | Article | PubMed | ISI | ChemPort |
Gurney, M.E. et al. Motor neuron degeneration in mice that express a human Cu,Zn superoxide dismutase mutation. Science264, 1772-1775 (1994). | PubMed | ISI | ChemPort |
Carri, M.T. et al. Expression of a Cu,Zn superoxide dismutase typical of familial amyotrophic lateral sclerosis induces mitochondrial alteration and increase of cytosolic Ca2+ concentration in transfected neuroblastoma SH-SY5Y cells. FEBS Lett.414, 365-368 (1997). | Article | PubMed | ISI | ChemPort |
Hemmer, W. & Wallimann, T. Functional aspects of creatine kinase in brain. Dev. Neurosci.15, 249-260 (1993). | PubMed | ISI | ChemPort |
O'Gorman, E., Beutner, G., Wallimann, T. & Brdiczka, D. Differential effects of creatine depletion on the regulation of enzyme activities and on creatine-stimulated mitochondrial respiration in skeletal muscle, heart, and brain. Biochim. Biophys. Acta1276, 161-170 (1996). | Article | PubMed | ChemPort |
Gurney, M.E. et al. Antioxidants and inhibitors of glutamatergic transmission have therapeutic benefit in a transgenic model of familial amyotrophic lateral sclerosis. Ann. Neurol.39, 147-157 (1996). | PubMed | ISI | ChemPort |
Dugan, L.L. et al. Carboxyfullerenes as neuroprotective agents. Proc. Natl. Acad. Sci. USA94, 9434-9439 (1997). | Article | PubMed | ChemPort |
Hottinger, A.F., Fine, E.G., Gurney, M.E., Zurn, A.D. & Aebischer, P. The copper chelator d-penicillamine delays onset of disease and extends survival in a transgenic moouse model of familial amyotrophic lateral sclerosis. Eur. J. Neurosci.9, 1548-1551 (1997). | PubMed | ISI | ChemPort |
Kostic, V. et al. Midbrain dopaminergic neuronal degeneration in a transgenic mouse model of familial amyotrophic lateral sclerosis. Ann. Neurol.41, 497-504 (1997). | PubMed | ISI | ChemPort |
Ferrante, R.J. et al. Increased 3-nitrotyrosine and oxidative damage in mice with a human Cu, Zn superoxide dismutase mutation. Ann. Neurol.42, 326-334 (1997). | PubMed | ISI | ChemPort |
Bogdanov, M.B., Ramos, L.E., Xu, X. & Beal, M.F. Elevated "hydroxyl radical" generation in vivo in an animal model of amyotrophic lateral sclerosis. J. Neurochem.71, 1321-1324 (1998). | PubMed | ISI | ChemPort |
Sasaki, S., Maruyama, S., Yamane, K., Sakuma, H. & Takeishi, M. Ultrastructure of swollen proximal axons of anterior horn neurons in motor neuron disease. J. Neurol. Sci.97, 233-240 (1990). | Article | PubMed | ISI | ChemPort |
Siklos, L. et al. Ultrastructural evidence for altered calcium in motor nerve terminals in amyotrophic lateral sclerosis. Ann. Neurol.39, 203-219 (1996). | PubMed | ISI | ChemPort |
White, R.J. & Reynolds, I.J. Mitochondrial depolarization in glutamate-stimulated neurons: an early signal specific to excitotoxin exposure. J. Neurosci.16, 5688-5697 (1996). | PubMed | ISI | ChemPort |
O'Gorman, E. et al. The role of creatine kinase inhibition of mitochondrial permeability transition. FEBS Lett.414, 253-257 (1997). | Article | PubMed | ChemPort |
Sauter, A. & Rudin, M. Determination of creatine kinase parameters in rat brain by NMR magnetization transfer: correlation with brain function. J. Biol. Chem.268, 13166-13171 (1993). | PubMed | ISI | ChemPort |
Corbett, R.J.T. & Laptook, A.R. Age-related changes in swine brain creatine kinase-catalyzed 31P exchange measured in vivo using 31P NMR magnetization transfer. J. Cereb. Blood Flow Metab.14, 1070-1077 (1994). | PubMed | ISI | ChemPort |
Xu, C.J. et al. Phosphocreatine-dependent glutamate uptake by synaptic vesicles. J. Biol. Chem.271, 13435-13440 (1996). | Article | PubMed | ISI | ChemPort |
Rothstein, J.D., M., V.K., Levey, A.I., Martin, L. & Kuncl, R.W. Selective loss of glial glutamate transporter GLT-1 in amyotrophic lateral sclerosis. Ann. Neurol.38, 73-84 (1995). | PubMed | ISI | ChemPort |
Sipila, I., Rapola, J., Simell, O. & Vannas, A. Supplementary creatine as a treatment for gyrate atrophy of the choroid and retina. N. Engl. J. Med.304, 867-870 (1981). | PubMed | ISI | ChemPort |
Blanchard, V., Raisman-Vozari, R., Vyas, S., Michel, P.P., Javoy-Agid, F. & Uhl, G. Differential expression of tyrosine hydroxylase and membrane dopamine transporter genes in subpopulations of dopaminergic neurons of rat mesencephalon. Molec. Brain Res.22, 29-40 (1994). | Article | PubMed | ISI | ChemPort |
Matthews, R.T. et al. Neuroprotective effects of creatine and cyclocreatine in animal models of Huntington's disease. J. Neurosci.18, 156-163 (1998). | PubMed | ISI | ChemPort |
Bogdanov, M.B., Ramos, L.E., Xu, X. & Beal, M.F. Elevated "hydroxyl radical" generation in vivo in an animal model of amyotrophic lateral sclerosis. J. Neurochem.71, 1321-1324 (1998). | PubMed | ISI | ChemPort |
Acknowledgments The secretarial assistance of S. Melanson is acknowledged. Photographic assistance was provided by S. Kuemmerle. This work was supported by NIH grant PO1 AG12292 (M.F.B. and R.J.F.), NS37102 (R.J.F.), the Veterans Administration (R.J.F.), the Muscular Dystrophy Association and the ALS Association, and NIMH grant MH11692 (A.M.K.).
2.3 days to 157.2 
and 
, control (unsupplemented diet); 
, 1% creatine diet; 
and 
).
m.
250 in both ventral horn areas from six tissue sections of the lumbar spinal cord of each mouse, with size discrimination into diameter classes of >25 
