Jaurnal of'Neurochhemlslry
Raven Press, Ltd., New York
0 I993 International Society for Neurochemistry
Rapid Communication
Superoxide Dismutase Activity, Oxidative Damage, and
Mitochondria1 Energy Metabolism in Familial
and Sporadic Amyotrophic Lateral Sclerosis
Allen C. Bowling, Jorg B. Schulz, *Robert H. Brown, Jr., and M. Flint Beal
Neurochemistry Laboratory and *Day Neuromuscular Research Center, Neurology Service, Massachusetts
General Hospitul and Hurvard Medical School, Boston, Massachusetts, U.S.A.
Abstract: The cause of neuronal death in amyotrophiclateralscle-
rosis (ALS) is unknown. Recently, it was found that some patients
with autosomal-dominant familial ALS (FALS) have point muta-
tions in the gene that encodes Cu/Zn superoxide dismutase
(SOD1). In this study of postmortem brain tissue, we examined
SOD activity and quantified protein carbonyl groups, a marker of
oxidative damage, in samples of frontal cortex (Brodmannarea 6)
from 10 control patients, three FALS patients with known SODl
mutations (FALS-1), one autosomal-dominant FALS patient with
no identifiable SODl mutations (FALS-0), and 11 sporadic ALS
(SALS) patients. Also, we determined the activities of compo-
nents of the electron transport chain (complexes1, 11-111, and IV) in
these samples. The cytosolic SOD activity, which is primarily
SODl activity, was reduced by 38.8% (p < 0.05) in the FALS-1
patients and not significantly altered in the SALS patients or the
FALS-0 patient relative to the control patients. The mitochondrial
SOD activity, which is primarily SOD2 activity, was not signifi-
cantly altered in the FALS-I, FALS-0, or SALS patients. The pro-
tein carbonyl content was elevated by 84.8% (p < 0.01) in the
SALS patients relative to the control patients. Finally, the com-
plex I activity was increased by 55.3% @ < 0.001) in the FALS-1
patients relative to the control patients. These results from corti-
cal tissue demonstrate that SODl activity is reduced and com-
plex I activity is increased in FALS-1 patients and that oxidative
damage to proteins is increased in SALS patients. Key Words:
Amyotrophic lateral sclerosis-Neurodegeneration-Free radi-
cals-Oxidative damage-Superoxide dismutase-Oxidative
phosphorylation.
J. Neurochem. 61,2322-2325(1993).
The cause of neuronal death in amyotrophic lateral sclero-
sis (ALS) is not known. Recently, it was discovered that
some patients with autosomal-dominant familial ALS
(FALS) have point mutations in the gene that encodes Cu/
Zn superoxide dismutase (SODI) (Rosen et al., 1993).
SOD 1 catalyzes the formation of hydrogen peroxide
through the dismutation of superoxide free radicals; by this
reaction, SODl is thought to play an important role in regu-
lating oxidative damage to cells (Halliwell, 1992). Two cen-
tral questions unanswered by the initial report of Rosen et
al. ( 1 993) were how the SOD I point mutations affect the
activity of SODl and whether there is evidence of increased
oxidative damage in either FALS or sporadic ALS (SALS).
2322
There are two types of inherited ALS. The autosomal-
dominant form is more common and constitutes 10- 15%of
all ALS cases (Horton et al., 1976; Mulder et al., 1986),
whereas the autosomal-recessive form is very rare and
usually only present in consanguineous families (Ben Ha-
mida et al., 1990). Most autosomal-dominant FALS cases
are indistinguishable from SALS on the basis ofclinical and
pathological criteria (Horton et al., 1976; Mulder et al.,
1986). This similarity suggests that autosomal-dominant
FALS and SALS may share similar pathogenetic mecha-
nisms, such as increased oxidative damage to cellular com-
ponents.
To determine how mutations in SODl affect its activity
and to quantify oxidative damage in FALS and SALS, we
investigated postmortem brain tissue from FALS patients
with known SOD1 mutations (FALS-I), one autosomal-do-
minant FALS patient with no identifiable SODl mutations
(FALS-0), SALS patients, and controls. In this study, we
determined SOD activity in cytosolic and mitochondrial
fractions as well as oxidative damage to cytosolic proteins.
In addition, as mitochondria1 energy metabolism may be
particularly susceptible to oxidative damage (Richter et al.,
1988; Linnane et al., 1989), we determined the activities of
the mitochondria1 electron transport chain complexes.
MATERIALS AND METHODS
Postmortem brain tissue, which was stored at -8O"C, was
obtained from 10 control patients (mean _t SD age, 69.9
-t 13.2 years; postmortem interval, 11. I +. 5.4h; sex ratio, six
males: four females), 1 I SALS patients (mean k SD age,
64.7 -t 12.7 years; postmortem interval, 16.0 I+_ 4.0 h; five
males: six females), three FALS patients with mutations in
Resubmitted manuscript received August 27, 1993; accepted
September 6, 1993.
Address correspondenceand reprint requests to Dr. M. F. Beal at
Neurology Service, Warren 408, Massachusetts General Hospital,
Boston, MA 02 I 14, U S A .
Abbreviations used: ALS, amyotrophic lateral sclerosis; FALS,
familial ALS; SALS, sporadic ALS; SOD, superoxide dismutase.
 SOD ACTIVITY AND OXIDATIVE DAMAGE IN ALS
exon I ofthe SOD1 gene (mean +- SD age, 41.3 t 2.1 years;
postmortem interval, 9.0 2 8.9 h two males: one female),
and one autosomal-dominant male FALS patient with no
identifiable SOD 1 mutations (age, 38 years; postmortem
interval, 12 h). For each FALS patient, all five exons of
SOD 1 were screened for mutations using single-strand con-
formational polymorphisms (Rosen et al., 1993). The diag-
nosis of SALS or FALS was made on the basis of clinical
features, electromyographic data, and postmortem histo-
logic examination of the brain and spinal cord. Approxi-
mately I g of Brodmann area 6 was homogenized; crude
mitochondrial fractions were obtained by applying the Pz
pellet to a discontinuous Ficoll gradient (Lai and Clark,
1979), and cytosolic fractions were obtained by centrifuging
the S, fraction at 100,000 g for 60 min.
SOD activity was determined by inhibition of the sponta-
neous oxidation of epinephrine to adrenochrome in 50 mA4
sodium carbonate and 0.1 mM EDTA, pH 10.2 (Misra and
Fndovich, 1972). The rate of oxidation was generally
0.02 1-0.023/min at 480 nm, and studies with exogenously
added bovine erythrocyte SOD demonstrated that inhibi-
tion was additive in the 2 0 4 0 % inhibition range for cyto-
solic and mitochondrial fractions. Samples were used rou-
tinely at concentrations that would produce 40-60% inhibi-
tion, and the percent inhibition was converted to units of
enzyme activity. Because the assays were conducted at pH
10.2, which is not an optimal pH for SOD2 activity, the
time of exposure to the buffer was limited to 30-40 s before
initiation of the reaction. Nevertheless, it is possible that
SOD2 was partially inactivated in these studies. Protein car-
bonyl derivatives were quantified using a 2,4-dinitrophenyl-
hydrazine assay with minor modifications (Levine et al.,
1990). Previously described assays were used for activity
measurements of citrate synthase and complexes I, 11-111,
and IV (Bowling et al., 1993). In the mitochondrial frac-
tions, the activities of SOD and the electron transport chain
complexes were corrected for the amount of citrate synthase
activity to account for variable degrees of mitochondrial
enrichment (Bowling et al., 1993).
In the control group, there was no significant difference
between males and females for cytosolic or particulate SOD
activity, protein carbonyl derivatives, citrate synthase activ-
ity, or electron transport chain complex activities. As a re-
sult, the data were analyzed without separating by sex. In
addition, the protein carbonyl derivatives and enzyme activ-
ities did not exhibit a significant correlation with age or
postmortem interval in the control group. The lack of corre-
lation with age for some of these measurements may be due
to the absence of younger control patients. The youngest
control patient was 51 years old, whereas previous aging
studies of protein carbonyl content (Smith et al., 1991) and
oxidative phosphorylation enzyme activities (Byrne et al.,
1991; Bowling et al., 1993) indicate that the inclusion of
younger subjects may be necessary to demonstrate age de-
pendence.
Protein concentration was determined by the method of
Bradford (1 976). All reagents were obtained from Sigma (St.
Louis, MO, U.S.A.) except the Bio-Rad (Richmond, CA,
U.S.A.)protein assay reagent and ubiquinone- 1, which was
the kind gift of Eisai Chemical Co. (Tokyo, Japan). Linear
regression analysis was used for correlations of measure-
ments with age and postmortem interval. Differences be-
tween groups were analyzed by ANOVA followed by
Fisher’s PLSD post hoc test.
2323
RESULTS
SOD assays were performed on the cytosolic and mito-
chondnal fractions. As in previous studies in brain tissue
(Saggu et al., 1989), 5.0 W K C N inhibited SOD activity by
70-80% in the cytosolic fraction (“cytosolic SOD’) and by
10-20% in the mitochondrial fraction (“particulate SOD’).
For the mitochondrial fractions, assays were performed rou-
tinely in the presence of 5.0 mA4KCN. Therefore, the cyto-
solic SOD measurement represented primarily SOD1 activ-
ity, whereas the particulate SOD measurement represented
primarily SOD2 activity.
Cytosolic SOD activity was reduced by 38.8% (p < 0.05)
in the FALS-I patients relative to the controls (Table I). In
contrast, relative to the controls, cytosolic SOD activity of
the FALS-0 and SALS patients and citrate synthase-
corrected particulate SOD activity of the SALS, FALS-I,
and FALS-0 patients were not significantly altered. Protein
carbonyl content was elevated by 84.8% (r, < 0.01) in the
SALS patients (Table 1). Citrate synthase-corrected com-
plex I activity was increased by 55.3% (p < 0.001) in the
FALS-1 patients relative to the controls (Table 2). Complex
I activity was not significantly altered in the SALS patients.
There were no significant differences for citrate synthase
activities or for citrate synthase-corrected activities of com-
plexes 11-111 and IV in the controls, SALS patients, and
FALS- 1 patients.
DISCUSSION
Some cases of autosomal-dominant FALS have been as-
sociated with mutations of the SODl gene (Rosen et al.,
1993). This gene, which encodes a 16,000-Da polypeptide,
consists of five exons (Levanon et al., 1985). Initially, I 1
different missense mutations were identified in exons 2 and
4 in 13 different families (Rosen et al., 1993). Subsequently,
several additional FALS-associated SODI mutations have
been identified, including an exon 1 mutation (Deng et al.,
1993; D. R. Rosen et al., unpublished observations). Three
of the FALS patients (FALS-1) in the present study have the
exon 1 mutation. The results of this study demonstrate de-
creased cytosolic SOD activity in frontal cortex (Brodmann
area 6) from these patients and therefore indicate that the
exon 1 mutation results in decreased SOD1 activity. The
finding of decreased activity is consistent with a recent study
that reports decreased SOD activity in erythrocyte lysates of
FALS patients with six different SODl mutations (Deng et
al., 1993) and our recent demonstration of reduced SOD
activity in erythrocyte lysates, lymphoblastoid cell lines,
and brain tissue (Brodmann area 1 I ) of FALS patients with
exon 1 mutations (A. C. Bowling et al., unpublished obser-
vations). As FALS patients generally exhibit similar clinical
and pathological features (Horton et al., 1976; Mulder et al.,
1986), it is possible that all FALS-associated SOD 1 muta-
tions result in reduced activity.
The identification of SOD 1 mutations and decreased
SOD 1 activity in some FALS patients suggests that free radi-
cals may play a central role in this disorder. SODl is a me-
talloenzyme that produces hydrogen peroxide by the dis-
mutation of superoxide, a free radical (Halliwell, 1992).
Through this reaction, SODl reduces the levels of superox-
ide. Thus, it is theoretically possible that decreased SOD1
activity may increase free radical levels and thereby increase
oxidative damage. In this study, the level of protein car-
bonyl groups, a marker of oxidative damage to proteins, was
J. Neurochem., Vol. 61, No. 6 . 1993
2324 A . C. BOWLING ET AL.

TABLE 1. SOD activities and protein carbonyl content in controls,

SA LS palients, and FALS patients
~~~~~

Age (years) Cytosolic SOD Particulate SOD Protein carbonyl

~ ~~~~

Controls 69.9 13.2 438.0 f 149.1 53.6 f 6.2 1.51 t 0.80

SALS 64.1 f 12.7 394.5 f 68.5 49.7 f 8.3 2 . 7 9 k 1.13"
FALS-1 41.3 f 2.1 267.9 _+ 56.7' 53.0 f 9.0 1.82 F 0.33
FALS-0 38 564. I 53.4 1.98
- ~~~~

Data are mean f SD values. Cytosolic SOD activity is expressed as U/mg of protein, particulate
SOD activity is expressed as SOD activity (U/mg of protein)/citratesynthase activity (pmol/min/
mg of protein), and protein carbonyl content is expressed as nmol/mg of protein. Samples were
assayed in triplicate.
The indicated p values are for comparison of the indicated group with the controls: "p < 0.0 I , ' p
< 0.05.

not significantly increased in the FALS-1 patients. It ispossi-
ble that FALS-I patients exhibit only a slight increase in
oxidative damage to protein in this brain region. Other areas
with more marked pathologic changes, such as Brodmann
area 4 or spinal cord, may exhibit more oxidative damage. It
is also possible that a large increase in oxidative protein
damage was not detected in this study because the FALS- 1
patients were not age-matched with the controls.
In the SALS patients in this study, levels of the protein
carbonyl derivatives were significantly increased. This find-
ing suggests that increased oxidative damage may play an
important role in SALS. Brain cytosolic SOD activity was
not significantly altered in SALS. This finding is consistent
with a previous study demonstrating no significant change
in SOD activity in erythrocyte lysates and muscle biopsy
specimens from SALS patients (Di6szeghy et al., 1989). An-
other study demonstrated that total SOD activity in CSF,
which contains primarily SOD1 activity (Strand and Mark-
lund, 1992), was decreased in SALS patients (Bracco et al.,
1991). This apparent discrepancy may be due to the fact
that SOD activity was assayed differently in these studies or
that SODl levels in CSF and brain cytosol respond differ-
ently in SALS.
The mechanism by which free radicals produce cell death
is not known. All types of macromolecules are subject to
oxidative damage, including proteins, DNA, and lipids
(Halliwell, 1992). Cell death is likely to occur through the
impairment of the function of macromolecules or organ-
elles that are especially sensitive to oxidative damage.
Among proteins, those that contain [4Fe-4S] clusters, such
as dehydratases, are very sensitive to inactivation by super-
oxide (Gardner and Fndovich, 199la,b). Among organ-
elles, the mitochondria appear to be especially vulnerable to
free radical damage because most of the intracellular free
radicals are generated by components of the electron trans-
port chain and mitochondria1 DNA is particularly suscepti-
ble to oxidative damage (Richter et al., 1988; Linnane et al.,
1989). In this investigation, the increase in complex I activ-
ity that we observed in the FALS- 1 patients may be compen-
satory and may indicate that the mitochondria are targeted
by the pathologic process. Previous studies have identified
ultrastructural abnormalities in the mitochondria of FALS
and SALS patients (Hirano et al., 1984a,b; Nakano et al.,
1987), but the significance of these findings is not known
(Kusaka and Hirano, 1985).
It is not clear how a mutation in SODl, an enzyme that is
constitutively expressed in all cells, could selectively dam-
age motor neurons. A generalized decrease in enzyme activ-
ity is suggested by the finding that SOD activity is reduced
in erythrocytes, lymphoblastoid cells, and Brodmann areas
6 and 11 (Deng et al., 1993; present study; A. C. Bowling et
al., unpublished observations). Decreased enzyme activity
in Brodmann areas 6 and I I , brain regions that exhibit little
or no pathology in FALS, suggests that an additional feature
of motor neurons must make these cells particularly vulner-
able to this defect. A macromolecule may be involved that is
very sensitive to oxidative damage and is also essential for
motor neuron survival. In addition, excitatory amino acid
toxicity, which has been implicated in the pathogenesis of
ALS (Rothstein et al., 1992), could be enhanced in motor
neurons by free radical-induced impairment of mitochon-
drial energy metabolism (Beal et al., 1993) or by increased
nitric oxide toxicity due to elevated superoxide levels (Beck-
man et al., 1990; Dawson et al., 1993).
The finding of increased oxidative damage in SALS pa-
tients indicates that oxidative stress may be a final common

TABLE 2. Activities of oxidative phosphorylation enzymes and citrate synthase

Complex I Complex 11-111 Complex IV Citrate svnthase
Controls 43.4 f 9.8 234.5 k 93.1 *
567.8 76.0 374.6 t 104.7
SALS 43.0 f 7.4 24 1.7 f 30.4 613.8 t 88.4 390.5 k 79.8
FALS- 1 67.4 f 6.2' 310.2 f 51.7 *
690.3 263.3 369.6 -+ 95.6
FALS-0 53.7 267.0 692.0 371.1

Data are mean f SD values. For citrate synthase, activity is in nmol/min/mg of protein. For all
other enzymes, values are citrate synthase-corrected and represent enzyme activity (nmol/min/mg
ofprotein)/citratesynthase activity (pmol/min/mg of protein).
'p < 0.001, relative to the controls.

J . Nrurochem.. i'ol. 61, No. 6, 1993

 SOD ACTIVITY AND OXIDATIVE DAMAGE IN ALS
pathway for cell death in autosomal-dominani FALS and
SALS. As SOD1 activity did not appear to b: altered in
SALS, the molecular basis for increased oxidative damage
in SALS and FALS may be very different. SOD 1 function is
involved in FALS, but various genetic and environmental
factors may be causative in SALS. In spite of d fferent pn-
mary molecular defects, the finding that increased oxidative
stress may be common to FALS and SALS sbggests that
antioxidant therapy may be effective for both disorders.
Acknowledgment: M.F.B. is supported by grants from the
U.S. Public Health Service (AGO5 134, NS 16367, and
NS10828). A.C.B. is a Fellow of the Massachiisetts Alz-
heimer's Disease Research Center (NIA AGO5 134) and the
Molecular Biology of Neurodegeneration Program at Har-
vard Medical School (NIAST32-AGO-0222). R.H.B. is
supported by the Muscular Dystrophy Associition, the
Amyotrophic Lateral Sclerosis Association, the Pierre L. de
Bourgknecht ALS Research Foundation, the Myrtle May
MacLellan ALS Research Foundation, the C. B. Day In-
vestment Co., and grant 1PO 1NS3 1248-01 from the Na-
tional Institutes of Health. We thank John M. Carney for
helpful discussions and Sharon Melanson for secr :tarial as-
sistance.
REFERENCES
Beal M. F., Hyman B. T., and Koroshetz W. (1993) Do defects in
mitochondrial energy metabolism underlie the pathology of
neurodegenerative diseases? Trends Neurosci. 16, 1:!5- 131.
Beckman J. J., Beckman T. W., Chen J., Marshall P. A., and Free-
man B. A. (1990) Apparent hydroxyl radical production by
peroxynitrite: implications for endothelial injury from nitric
oxide and superoxide. Proc. Natl. Acad. Sci. USA $17, 1620-
1624.
Ben Hamida M., Hentati F., and Ben Hamida C. (1990) k ereditary
motor system diseases (chronic juvenile amyotropt ic lateral
sclerosis). Brain 113, 347-363.
Bowling A. C., Mutisya E., Walker L. C., Price D. L., Cork L. C.,
and Beal M. F. (1993) Age-dependent impairment of mito-
chondrial function in primate brain. J. Neurochem. 60, 1964-
1967.
Bracco F., Scarpa M., Rigo A,, and Battistin L. (1991) Determina-
tion of superoxide dismutase activity by the polarugraphic
method of catalytic currents in the cerebrospinal fluid of aging
brain and neurologic degenerative diseases. Proc. Sqc. Exp.
Bid. Med. 196, 36-4 I .
Bradford M. M. ( 1976) A rapid and sensitive method for tht: quanti-
tation of microgram quantities of protein utilizing thc princi-
ple of protein-dye binding. Anal. Biochem. 72, 248-254.
Byrne E., Dennett X., and Trounce I. (1991) Oxidative ent'rgy fail-
ure in post-mitotoic cells: a major factor in senescexe. Rev.
Neurol. 147,532-5 35.
Dawson V. L., Dawson T. M., Bartley D. A., Uhl G. R., and Snyder
S. H. ( 1993) Mechanisms of nitric oxide-mediated neu -0toxic-
ity in primary brain cultures. J. Neurosci. 13, 265 1-206 I .
Deng H.-X., Hentati A., Tainer J. A,, et al. (1993) Amyotrophic
lateral sclerosis and structural defects in Cu, Zn supxoxide
dismutase. Science 261, 1047-1051.
Dibszeghy P., Imre S., and Mechler F. (1989) Lipid peroxidation
and superoxide dismutase activity in muscle and erythrocytes
in adult muscular dystrophies and neurogenic atrophks. Eur.
Arch. Psychiatry Neurol. Sci. 238, 175-177.
Gardner P. R. and Fridovich I. (1991~)Superoxide sensitivity of
2325
the Escherichia coli 6-phosphogluconate dehydratase. J. Biol.
Chem. 266, 1478-1483.
Gardner P. R. and Fridovich I. (1 99 18) Superoxide sensitivity of
the Escherichia coli aconitase. J. Bid. Chern. 266, 19328-
19333.
Halliwell B. (1992) Reactive oxygen species and the central nervous
system. J. Neurochern. 59, 1609- 1623.
Harman D. (1 990) Lipofuscin and ceroid formation: the cellular
recycling system, in Lipojiscin and Ceroid Pigments (Porta
E. A., ed), pp. 3- 15. Plenum Press, New York.
Hirano A., Nakano I., Kurland L. T., Mulder D. W., Holley P. W.,
and Saccomanno G. (1 984a) Fine structural study of neurofi-
brillary changes in a family with amyotrophic lateral sclerosis.
J. Neuropathol. Exp. Neurol. 43,47 1-480.
Hirano A., Donnenfeld H., Sasaki S., and Nakano I. (19848) Fine
structural observations of neurofilamentous changes in amyo-
trophic lateral sclerosis. J. Neuropathol. Exp. Neurol. 43,46 1-
470.
Horton W. A,, Eldridge R., and Brody J. A. (1976) Familial motor
neuron disease: evidence for at least three different types. Neu-
rology 26,460-465.
Kusaka H. and Hirano A. ( 1 985) Fine structure ofanterior horns in
patients without amyotrophic lateral sclerosis. J. Neuropathol.
Exp. Neurol. 44,430-438.
Lai J. C. K. and Clark J. B. (1979) Preparation of synaptic and
nonsynaptic mitochondria from mammalian brain. Methods
Enzymol. 55, 51-60.
Levanon D., Lieman-Hunvitz J., Dafni N., Wigderson M., Sher-
man L., Bernstein Y., Laver-Rudich M., Danciger E., Stein O.,
and Groner Y. (1985) Architecture and anatomy of the chro-
mosomal locus in human chromosome 21 encoding the Cu/
Zn superoxide dismutase. EMBO J. 4, 77-84.
Levine R. L., Garland D., Oliver C. N., Amici A,, Climent I., Lenz
A.-G., Ahn B.-W., Shaltiel S.,and Stadtman E. R. (1990) Deter-
mination of carbonyl content in oxidatively modified proteins.
Methods Enzyrnol. 186, 464-487.
Linnane A. W., Marzuki S., Ozawa T., and Tanaka M. (1989) Mi-
tochondrial DNA mutations as an important contribution to
aging and degenerative diseases. Lancet 1, 642-645.
Misra H. P. and Fridovich I. (1972) The role of superoxide anion in
the autooxidation of epinephrine and a simple assay for super-
oxide dismutase. J. Biol. Chern. 247, 3170-3175.
Mulder D. W., Kurland L. T., Offord K. P., and Beard C. M. ( 1 986)
Familial adult motor neuron disease: amyotrophic lateral scle-
rosis. Neurology 36, 5 1 1-5 17.
Nakano Y . ,Hirayama K., and Terao K. (1987) Hepatic ultrastruc-
tural changes and liver dysfunction in amyotrophic lateral scle-
rosis. Arch. Neurol. 44, 103- 106.
Richter C., Park J.-W., and Ames B. N. (1988) Normal oxidative
damage to mitochondrial and nuclear DNA is extensive. Proc.
Natl. Acad. Sci. USA 85,6465-6467.
Rosen D. R., Siddique T., Patterson D., et al. (1 993) Mutations in
Cu/Zn superoxide dismutase are associated with familial
amyotrophic lateral sclerosis. Nature 362, 59-62.
Rothstein J. D.. Martin L. J., and Kuncl R. W. (1992) Decreased
glutamate transport by the brain and spinal cord in amyotro-
phic lateral sclerosis. N. EngI. J. Med. 326, 1464-1468.
Saggu H., Cooksey J., Dexter D., Wells F. R., Lees A,, Jenner P.,
and Marsden C. D. (1989) A selective increase in particulate
superoxide dismutase activity in parkinsonian substantia ni-
gra. J. Neurochem. 53,692-697.
Smith C. D., Carney J. M., Starke-Reed P. E., Oliver C. N., Stadt-
man E. R., Floyd R. A., and Markesbeny W. R. ( I 99 I ) Excess
brain protein oxidation and enzyme dysfunction in normal
aging and in Alzheimer's disease. Proc. Natl. Acad. Sci. USA
88, 10540-10543.
Strand T. and Marklund S. L. ( 1 992) Release of superoxide dismu-
tase into cerebrospinal fluid as a marker ofbrain lesion in acute
cerebral infarction. Stroke 23, 5 15-5 18.
J. Neurochem., Vol. 61, No. 6, 1993