close

Вход

Забыли?

вход по аккаунту

?

Enhanced oxygen radical production in a transgenic mouse model of familial amyotrophic lateral sclerosis.

код для вставкиСкачать
En.hanced Oxygen Radical I?'roduction in a
Transgenic Moiise Mod.el of Famili.a1
h y o t r o p h i c Lateral Sclerosis
Rugao Liu, PhD,*§ John S. Althaus, MS,t Brenda R. Ellerbrock, BA,? David A. Becker, PhDJ
and Mark E. Gurney, PhD*t$
Mutations of the SODl gene encoding copper/zinc superoxide dismutase (CuZnSOD) cause an inherited form of amyotrophic lateral sclerosis. When expressed in transgenic mice, the same SODl mutations cause progressive loss of spinal
motor neurons with consequent paralysis and death. In vitro biochemical studies indicate that SODl mutations enhance
free radical generation by the mutant enzyme. We investigated those findings in vivo by using a novel, brain-permeable
spin trap, azulenyl nitrone. Reaction of azulenyl nitrone with a free radical forms a nitroxide adduct that then fiagments
to yield the corresponding azulenyl aldehyde. Transgenic mice expressing mutant SOD1-G93A show enhanced free
radical content in spinal cord but not brain. This correlates with tissue-specific differences in the level of transgene
expression. In spinal cord, the increase in free radical content is in direct proportion to the age-dependent increase in
mutant human CuZnSOD expression. This increase precedes motor neuron degeneration. The higher level of human
CuZnSOD expression seen in spinal cord compared with brain, and consequent difference in free radical generation,
provides a basis for understanding the selective vulnerability of the spinal cord in this disease model.
Liu
R, Althaus JS, Ellerbrock BR, Becker DA, Gurney ME. Enhanced oxygen radical production in a transgenic
mouse model of familial amyotrophic lateral sclerosis. A n n Neurol 1998;44:763-770
Amyotrophic lateral sclerosis (ALS) selectively affects
motor neurons in cortex, brainstem, and spinal cord. It
occurs in both sporadic and familial forms, which are
clinically and pathologically similar.' The major advance in understanding the disease has come from genetic studies that identified a subset of familial ALS
(FALS) patients bearing autosomal dominant mutations in the S O D l gene encoding copperlzinc superoxide dismutase ( CUZ~SOD).'.~
CuZnSOD is a cytosolic enzyme whose major function is to dismutate
superoxide to hydrogen peroxide. In cooperation with
other antioxidant enzymes, for example, catalase and
glutathione peroxidase, deleterious reactive oxygen species are scavenged to harmless water and molecular oxygen. Of the 50 or more S O D l mutations described in
FALS, most are missense mutations, which substitute
one amino acid for a n ~ t h e rMany
.~
of these mutations,
including the SODl-G93A mutation that is the subject of this study, have relatively little effect on dismutation activity toward superoxide, although such in
vitro measurements do not assess directly the kinetics
of superoxide dismutation in motor neurons.425
Expression of mutant human CuZnSOD in trans-
genic mice causes a motor neuron disease that resembles FALS.'.' SOD 1-G93A transgenic mice expressing
the mutant protein develop progressive paralysis in association with motor neuron loss that progresses to
death. The timing of disease, duration of survival, and
aspects of the pathology in spinal cord vary with the
level of expression of the mutant enzyme.'-' In contrast, SOD 1 transgenic mice expressing wild-type human CuZnSOD do not develop motor neuron disease.' As both wild-type and mutant SODl-G93A
transgenic mice show similar elevation of CuZnSOD
enzymatic activity, disease is thought to be caused by
gain-of-function mutations in the SOD 1 gene.'
More recently, S O Dl mutations have been shown
to enhance in vitro generation of hydroxyl radical from
hydrogen peroxide. 1 0 3 1 1 In addition to the dismutation
of superoxide to hydrogen peroxide, wild-type CuZnSOD can use hydrogen peroxide as a substrate to initiate a Fenton-like reaction with production in vitro of
hydroxyl
The SODl-G93A and A4V mutations enhance this minor catalytic activity" by decreasing the & for hydrogen peroxide" and by altering the redox behavior of the enzyme.I5
From the *Department of Biological Sciences, Western Michigan
University, and tPhatmacia and Upjohn, Central Nervous System
Diseases Research, Kalamazoo, MI; and $Department of Chemistry,
Florida International University, Miami, FL.
Received Sep 29, 1997, and in revised form Mar 11, 1998. Accepted for publication Apr 16, 1998.
5Present address: Hughes Institute, Roseville, MN 551 13.
Address correspondence to Dr Gurney, Pharmacia and Upjohn, Inc,
Central Nervous System Diseases Research, 301 Henrietta Street,
Kalamazoo, MI 49007.
Copyright
0 1998 by the American Neurological Association 763
We investigated the characteristics of free radical production in the FALS transgenic model by usin a newly
developed spin trap, azulenyl nitrone (AZN)."17 Trapping of free radicals by AZN results, as with other nitrones, in addition of the radicals to the nitrone double
bond to yield nitroxide adducts. These nitroxides, depending on the type of free radical trapped, can then
fragment via several pathways to yield the corresponding azulenyl aldehyde (AZA) (Fig 1). T h e novelty of
AZN resides in the ultraviolet-visible (UV-VIS) spectral changes of the compound as it reacts with a n d
consequently tags free radicals. This provides a unique
approach to monitor oxygen radical production. l 6 The
strongly absorbing chromophores within azulenyl derivatives such as AZN a n d AZA,which possess a comm o n visible absorbance maxima at 390 nm, also allow
the ready detection of AZN a n d related metabolites by
high-performance liquid chromatography (HPLC).l7
T h e success of this technique also hinges on the highly
lipophilic character of A Z N , which allows it t o cross
the blood-brain barrier. l 7 This enabled us to sample
oxygen radical content in the brain and spinal cord of
SOD1 transgenic mice by measuring the formation of
AZA after intravenous injection of AZN.
Subjects and Methods
Transgenic Mice
TgN(SODl-G93A) lGur mice expressing the G93A mutant
of human CuZnSOD (the G1H strain described by Chiu
and associates") and TgN(S0D 1)2Gur mice expressing
wild-type human CuZnSOD (the N29 strain in the study by
Chiu and associates") were housed under virus antibodyfree conditions. Both strains are available through the Induced Mutant Resource, The Jackson Laboratory (Bar Harbor, ME). Selective breeding maintained each transgene in
the hemizygous state on an F, hybrid C57BL6 X SJL genetic background.'8 The colony was free of known mouse
pathogens.
Assay of Free Radical Production In Vitro
For in vitro assays, mice were killed at 90 days of age by
transcardial perfusion with cold phosphate-buffered saline,
using ethically approved procedures after anesthetization
with Metofane (Pittman-Moore). The spinal cord was harvested, cut into small pieces in 50 mM potassium phosphate
buffer (pH 7.5), and then disrupted by sonication twice for
15 seconds on ice. Debris were pelleted by centrifugation in
a 1.5-mi microfuge tube at: 8,000 rpm for 30 seconds and
the supernatant was transferred to a new tube. Protein concentration was determined by the bicinchoninic acid (BCA)
method (Pierce Chemical Co). To assay free radical production in vitro, spinal cord extracts (5 mg of proteinlml final)
were incubated with 30 m M hydrogen peroxide in the presence of 5 mM AZN in 5 m M potassium phosphate buffer
(pH 7.5) containing 0.2 mM deferoxamine (Sigma Chemical) and 1 m M EDTA at 37°C for 2 hours. AZN was dissolved initially at 100 mM in Intralipid carrier (Pharmacia)
containing 0.5% ethanol and then diluted into the assay
buffer. CuZnSOD was inactivated in some experiments by
20-minute preincubation wit:h 1 mM diethyldithiocarbamate
(DDC; Sigma) at room temperature before the addition of
hydrogen peroxide and AZN. Reaction products were extracted with an equal volume of 100 p.1 ethyl acetate and the
AZAlAZN ratio was analyzed by HPLC. A control with no
protein added was included in each experiment to subtract
background oxidation of AZN.
Assay of CuZnSOD
Analysis of CuZnSOD by western blot was as described previously.'8 The S O D zymographic assay used the protocol of
Beauchamp and Fridovich." The assay of CuZnSOD activity, based on inhibition of nitroblue tetrazolium reduction in
a coupled xanthinelxanthine oxidase system, and the SOD
enzyme immunoassay have been described previously.6
HPLC Assay o f AZAIRZN Ratio
For HPLC analysis of the AZAIAZN ratio, 50 ml of the
sample resuspended in mobile phase was injected onto the
HPLC system. The mobile phase was methanollwater 90:lO
with 1 g of ammonium acetate added per liter. Mobile phase
flow of 0.5 ml/min carried the analyte onto a C18 Symmetry
column (Waters Instrument Company). The eluant was
spectrophotometrically monitored by using a W detector
(Spectroflow 757) at 390 nm. The wavelength of 390 nm
was selected based on comparison of the common absorption
maxima for both the AZN and the AZA product. Peak area
1.5
m
b
.r-
X
v
g
1.0
0
cn
"(A).
(CI
a
0
,
c
J
J
em
2
Fig 1. (A) Capture of a perox$, hydroxyl, or
alkoxyl radical by azulenyl nitrone (AZN)
forms a nitroxide adduct that subsequently
fiapnents to a stable azulenyl aldehyde
(B) Analysis o f the ]iwi/X?N ratio
by high-pe$ormance liquid chromatography
afier in vitro incubation of AZN with a
protein extract of SODl-G93A spinal cord
in the presence of hydrogen peroxide. The
0.5
--
SODI-G93A
determination was performed by using the Waters Maximum
software package. AZN and AZA peak identification was
based on coelution with spiked standards. Peak areas for
AZN and AZA were used to calculate the AZA/AZN ratio.
In Vivo Assay o f Oxygen Radical Content in Tissues
Mice were injected with AZN in a volume of 0.5 ml through
a tail vein. For injection, AZN was suspended in Intralipid
containing 0.5% ethanol at a concentration of 3.5 mglml
(1.75 mg/mouse). Four hours after injection, mice were
anesthetized with Metofane, blood was harvested by heart
puncture, and then the mice were killed by perfusion with
cold phosphate-buffered saline following ethically approved
procedures. Once clear of blood, the brain and spinal cord
were harvested, and then all samples were frozen in liquid
nitrogen and subsequently stored at -80°C. For analysis,
brain and spinal cord homogenates were prepared as above.
AZN and related metabolites were extracted from the aqueous sample into 0.6 ml of ethyl acetate. The organic extracts
then were filtered in 0.2-wm nylon microfilter tubes (Rainin)
and vacuum dried. For HPLC analysis, the dried samples
were resuspended in 120 ~1 of the mobile phase.
Mice were injected intraperitoneally with salicylate in
phosphate-buffered saline at a dose of 300 mg/kg. One hour
later, the mice were processed as for AZN to obtain the spinal cord. This was homogenized in mobile phase containing
deferoxamine, mannitol, and phenylalanine and then analyzed by HPLC to quantitate the content in the sample of
salicylate and its metabolites, 2,3-dihydroxybenzoic acid
(2,3-DHBA) and 2,5-DHBA.'"
Statistical Analysis
Data are reported as mean C SEM values. Statistical
comparisons were made by two-tailed Student's t test, with
the Bonferroni correction for multiple comparisons where
appropriate.
Results
Previous studies of highly purified, recombinant human C u Z n S O D containing the G93A mutation have
demonstrated enhanced catalysis of free radical formation from hydrogen peroxide in comparison with the
nonmutant enzyme.'03' Accordingly, we sought to
confirm that spinal cord extracts prepared from S O D l G93A mice that contain mutant C u Z n S O D catalyzed
the formation of free radicals from hydrogen peroxide
at a greater rate than extracts prepared from S O D l
transgenic mice that contain nonmutant human
C u Z n S O D , and that these could be captured by the
AZN spin trap. T h e G93A mutant form of human
C u Z n S O D has normal levels of dismutase activity directed toward superoxide,' so for this experiment, we
chose two lines of wild-type or mutant S O D l transgenic mice that expressed human C u Z n S O D at equivalent levels. T h e amount of human C u Z n S O D expressed in S O D l trans enic mouse lines varies with
transgene copy number!'
C u Z n S O D dismutase activity directed toward superoxide was measured by inhi-
bition of nitroblue tetrazolium reduction in a coupled
xanthine/xanthine oxidase assay system.6 This demonstrated that the two transgenic lines selected,
TgN(SODl-G93A) l G u r and TgN(SOD1)2Gur, had
comparable elevation of total C u Z n S O D activity in
spinal cord at 90 days of age, 7.5 -t 0.6 and 8.9 2 1.2
U / m g of protein, respectively, compared with 1.0 2
0.08 U / m g of protein in nontransgenic mice. This was
further confirmed by both western blot and zymographic assay, which detected similar amounts of
C u Z n S O D protein and activity in T g N ( S 0 D 1-G93A)
and TgN(SOD1) transgenic spinal cord (Fig 2).
Catalysis of free radical production from hydrogen
peroxide was measured in vitro by using extracts prepared from S O D l and SODl-G93A mouse spinal
cord at 90 days of age. AZN was used to trap oxygen
radicals formed in vitro, and H P L C analysis was used
to measure the conversion of AZN to AZA due to fragmentation of the nitroxide adduct (see Fig 1). Two
metal chelators, E D T A and deferoxamine, were included in the incubation solution to minimize Fenton
catalysis of hydroxyl radical formation by free iron or
copper in the tissue extract. T h e experiment showed
that catalysis of free radical production by SOD1G93A mouse spinal cord extracts increased linearly
with hydrogen peroxide concentration at least up to 30
Fig 2. (A) Detection of human copperlzinc superoxide dismutase (CuZnSOD) protein in nontransgenic (non Td,
SODl, and SODI-G93A spinal cord by western blot, using a
polyclonal rabbit antibody against mouse CuZnSOD. The
upper and lower bands correspond to human and mouse
CuZnSOD, respectively. (B) Zymographic assay of SOD activity in SODl and SODI-G93A spinal cord compared with
nonTg mice.
human
mouse
\
-
nonTg
nonTg
SODl
SODl SODl-G93A
SOD1-G93A
Liu et al: Oxygen Radicals in FALS Transgenic Mice
765
mM, and that with 30 mM hydrogen peroxide, catalysis of free radical production increased linearly with
protein concentration.
In comparison with S O D l or nontransgenic mice,
spinal cord extracts from SODl-G93A mice catalyzed
excess free radical production from hydrogen peroxide
(Fig 3). Thus, mutant CuZnSOD in the SODl-G93A
spinal cord extracts appears to catalyze excess free radical
production as previously shown for the highly purified,
recombinant enzyme.",'
To assure ourselves that the
catalysis of excess free radical production was mediated
by mutant CuZnSOD, we asked if the reaction could be
inhibited by DDC, a specific inhibitor of CuZnSOD
that acts by chelating the copper in the active site." Preincubation with 1 mM DDC before the addition of
AZN and hydrogen peroxide completely abolished the
CuZnSOD dismutation activity toward superoxide
present in the S O D l and SODl-G93A spinal cord extracts. This reduced catalysis of free radical production
in SOD 1-G93A spinal cord extracts to the baseline levels seen in S O D l or nontransgenic mouse spinal cord
extracts (see Fig 3 ) . Thus, these experiments show that
mouse spinal cord extracts containing the G93A mutant
form of human CuZnSOD catalyze excess production
of free radicals from hydrogen peroxide and that this
can be inhibited by DDC, an active site inhibitor of
CuZnSOD superoxide dismutation activity.
We next sought to determine if the tissue content of
free radicals in the spinal cord of SODl-G93A mice
was elevated in vivo. We used two spin traps to probe
free radical levels in vivo, AZN, which theoretically
should be able to capture alkoxyl, peroxyl and hydroxyl
radicals, and salicylate, which is a highly specific probe
for hydroxyl radicals. AZN was administered intravenously while salicylate was administered by intraperitoFig 3. Detection of free radical production by spinal cord extracts prepared from 90-day-old mice on incubation with hy< 0.001; triplicate assays). The endrogen peroxide ("9
hanced production ofpee radicals catalyzed by SODl-G93A
spinal cord extracts is blocked by preincubation with diethyldithiocarbamate (DDC), a specific inhibitor of CuZnSOD
(p < 0.0001; triplicate assays).
neal injection. We initially sampled nontransgenic,
SOD1, and SODl-G93A mice at 90 days of age. Significant elevation of free radical production occurred
only in SODl-G93A transgenic mice, and this was detected only with the AZN spin trap (Fig 4). Conversion of AZN to AZA was as high as 4.5% in the 90day-old SODl-G93A mouse spinal cord, whereas the
conversion of salicylate to 2,5-DHBA and 2,3-DHBA
was less than 0.01% in the same mice. Only baseline
conversion levels of AZN to AZA or salicylate to 2,5DHBA or 2,3-DHBA were seen in nontransgenic and
SOD 1 transgenic mice. This indicates that excess production of free radical species is associated with the expression of mutant G93A CuZnSOD. This was not a
consequence of simply increasing tissue levels of CuZnSOD dismutation activity directed toward superoxide
as is the case for S O D l transgenic mice. At 90 days of
age, 50% of SODl-G93A mice have developed symptomatic disease, and loss of motor neurons is evident in
the spinal cord."
We next probed the free radical content of spinal
cord and brain in SODl-G93A mice in comparison
with nontransgenic mice at 30, 60, and 90 days of age
(Fig 5). The experiment was designed to obtain some
insight into the time dependence of free radical production in relation to the tissue content of CuZnSOD
and to progression of spinal cord pathology. A slight
(about 35%) increase in free radical production was
observed at 30 days of age in the spinal cords of
SODl-G93A mice. Conversion of AZN to AZA continued to increase monotonically with age, indicating
an increase in the rate of free radical production in the
Fig 4. Spinal cord conversion of azulenyl nitrone ('AZN) or
salicyhte to azulenyl aldehyde (AZA) or 2,5-dihydroxybenzoic
acid (2,5-DHBA), respectively, afzer intravenous injection of
nontransgenic (non Td, SODl, and SODl -G93A transgenic
mice (*+p < 0.001; 6 mice per group). Only data for conversion o f salicyhte to 2S-DHBA are shown; values for conversion of salicyhte to 2,3-DHBA were smaller and did not differ between the different types of mice.
0A W A Z N
2,5 DHBA lSAL
nonTg
SOD1
766 Annals of Neurology Vol 44
SODI-G93A
No 5
nonTg
November 1998
SOD1
SODI-G93A
L
k*
n
L
A
sc
Age(days)
30
sc
60
SC
90
Brain
Brain
60
30
Brain
90
Fig 5. Oxygen radical content of SODl-G93A and nontransgenic mouse spinal cord and brain at different ages as assessed
by conversion of azulenyl nitrone ( M N ) to azulenyl aldehyde
( X U ) after intravenous injection (,*p < 0.001; 6 mice per
group).
spinal cord of SODl-G93A mice. This reached a maximum value of 4.5% conversion at 90 days of age. This
increase was not seen in nontransgenic mice. Brain
production of free radicals rose slightly from 30 to 90
days of age but did not differ between SODl-G93A
and nontransgenic mice. The percentage conlersion of
AZN to AZA only reached a maximum of 0.6 to 0.7%
in the brain at 90 days of age, compared with the spinal cord where the percent conversion reached 4.5%, a
difference of nearly six- to eightfold. We also failed to
detect excess production of free radicals in blood despite the high content of human CuZnSOD in the red
blood cells of SODl-G93A mice. The ratio of AZA to
AZN in blood averaged around 0.4% in both types of
mice at all three ages.
Measurements of mutant human CuZnSOD in
SOD 1-G93A mice by enzyme immunoassay indicate
that the increase in free radical content in spinal cord
parallels an age-dependent increase in human CuZnSOD expression. Between 30 and 90 days of age, human CuZnSOD content in spinal cord increases about
twofold (Table). The increase was seen in two experiments with different groups of SODl-G93A mice. In
contrast to spinal cord, the level of human CuZnSOD
expression in the brain of SOD 1-G93A transgenic
mice is about threefold less and does not change significantly with age. The increase in human CuZnSOD
expression is seen in both SODl-G93A and SODl
mice, although it is more pronounced in the SOD1G93A mice (see Table). The enzyme immunoassay for
human CuZnSOD does not cross-react with mouse
CuZnSOD, so we were unable to determine if expression of endogenous mouse CuZnSOD also increases
with age. As shown in Figure 6 , the free radical content
of SODl-G93A transgenic mouse spinal cord and
brain is directly proportional to the amount of mutant
human CuZnSOD expressed in the tissue. These findings link the generation of free radicals in situ to the
tissue content of mutant human CuZnSOD in SODIG93A transgenic mice.
Discussion
Previous in vitro biochemical studies of the G93A and
other CuZnSOD mutations linked to FALS indicate
that the mutations enhance catalysis of a Fenton-like
reaction with consequent production of damaging free
radicals. 'Ox The narrow, positively charged, active-site
channel normally limits the access of reactants to the
copper redox center, and thus locks away much of the
natural chemistry of copper. As the more than 50 missense mutations of CuZnSOD linked to FALS alter
residues located primarily in the polypeptide backbone
of the enzyme, they probably relax the conformation of
the protein and allow the copper catalytic center
Table. Human CuZnSOD in Brain and Spinal Cord of SODI and SODl-G93A Transgenic Mice
Transgenic Line
Age (days)
Spinal Cord
(ng/pg of protein)
Brain
(ng/pg of protein)
Percent Increase
_ _ _ ~
~
Experiment 1
Experiment 2
Experiment 1
SODl-G93A
SOD1-G93A
SOD 1-G93A
SOD1-G93A
SODl-G93A
SODl-G93A
SODl
SODl
SOD 1
30
60
90
30
60
90
30
60
90
9.1 t- 0.2
14.8 t- 0.7;
29.3 t- 1.7
14.5 t- 1.0
20.6 t- 0.9"
28.3 -C 1.6b
17.7 2 1.0
17.9 t- 2.8
27.8 t- 5.8
160
320
142
195
101
156
6.7 t- 0.6
7.8 t- 0.8
8.8 2 1.9
8.8 t- 0.8
9.9 t- 0.3
9.7 C 0.3
9.1 & 0.3
5.4 t- 0.4
1.3 t- 0.1
Data are mean 2 SEM values; n = 6 for each group.
"p < 0.005; ' p < 0.001.
CuZnSOD
=
coppedzinc superoxide dismutase.
Liu et al: Oxygen Radicals in FALS Transgenic Mice
767
+
5
4l
I
I
I
I
15
20
25
30
0
0
5
10
ng Human Cu.Zn SODIpg protein
Fig 6 Oxygen radical content in SODl-G93A brain and
spinal cord at different ages is directly proportional to the tissue content of human copperhinc superoxide dismutase (Cu,Zn
SOD). Filled symbols = spinal cord open symbols =
brain; A = 30, = 60, a n d 0 = 90 days ofage.
greater access to solvent. In support of this hypothesis,
FALS-linked mutations alter the zinc site and the redox
behavior of CuZnSOD.15 Our experiments extend in
vitro studies of the highly purified enzyme by showing
that spinal cord extracts containing G93A mutant
CuZnSOD catalyze Fenton-type production of hydroxyl radical (or other secondary free radicals) from
hydrogen peroxide in vitro and that enhanced rates of
free radical production can be measured in situ in the
spinal cords of SODl-G93A mice. This supports the
hypothesis that SODI-linked FALS is due to free radical-mediated attack and consequent oxidative damage
to vulnerable neuronal populations.
Because the G93A mutant form of human CuZnSOD has normal dismutation activity directed toward
superoxide,' we chose for comparison lines of transgenic mice expressing wild-type and mutant human
CuZnSOD at about the same levels. We found that
spinal cord extracts from SODl-G93A mice catalyzed
excess formation of free radicals from hydrogen peroxide in vitro, and that only SODl-G93A mouse spinal
cord had significant levels of free radical production in
vivo. SOD 1 mice expressing comparable amounts of
human CuZnSOD did not differ from nontransgenic
mice in either experiment, despite a seven- to eightfold
increase of total CuZnSOD dismutation activity toward superoxide in comparison with nontransgenic
mice.6 This suggests that the gain of function in human CuZnSOD associated with FALS-linked mutations is due at least in part to increased catalysis of free
radical formation. Free radical production in SOD1G93A mice was seen only in spinal cord, not in brain
or blood, and correlated with the considerably higher
768 Annals of Neurology Vol 44
No 5 November 1998
levels of G93A mutant CuZnSOD expression in spinal
cord. These results imply that the selective vulnerability
of spinal cord is due to the higher conrent of CuZnSOD. Within the spinal cord, motor neurons express
the highest levels of CUZ~SOD.',',~~
The increase in oxygen radical content in SOD1G93A mouse spinal cord is likely to be due to the formation of peroxyl radicals, as salicylate failed to reveal
an increase in hydroxyl radical content. Salicylate may
simply not have the necessary sensitivity to detect formation of hydroxyl radicals in SODl-G93A spinal
cord, or if hydroxyl radicals are produced by mutant
CuZnSOD, they may react with membrane lipids to
form alkyl or peroxyl radicals too quickly for salicylate
to trap them. In theory, lipid peroxyl or other secondary peroxyl radicals could be trapped by AZN but
not by salicylate. As small, anionic scavengers protect
CuZnSOD from inactivation by reacting with hydroxyl radical within the active site channe1,12-14 secondary carbon radicals such as formyl or 2-glutamyl
radicals also could be produced by mutant CuZnSOD
in vivo. For example, reaction of glutamate with hydroxyl radical within the active site channel could produce a relatively stable captodative radical (RC), which
forms the corresponding peroxyl radical (RCOO') when
it reacts with oxygen.
W e found that in situ free radical production in
SODl-G93A mouse spinal cord increased with age in
proportion to an age-dependent increase in human
CuZnSOD protein expression. Human CuZnSOD expression increased about twofold in SODl-C93A spinal cord between 30 and 90 days of age, whereas a
smaller increase was seen in SODl mice. Brain levels
of human CuZnSOD expression do not change with
age. Brujin and colleagues22 also report a twofold increase in human CuZnSOD expression in the spinal
cord of SOD1-G85R mice, although they propose that
this is in response to disease. Perhaps this explains the
greater increase in expression of human CuZnSOD
that we see in SODl-G93A compared with SOD1
mice; however, the site of transgene integration into
the mouse genome may also influence its pattern of
expression. Comparative human data are lacking, although in human spinal cord, motor neurons clearly
contain the highest levels of S O D l mRNA and prorein.21 Previous studies of CuZnSOD expression in
rodent brain vary, but most studies report only minor changes in CuZnSOD activity between 1 and 6
months of age23,24;information on SOD 1 gene expression in rodent spinal cord is lacking.
The age-dependent increase in mutant CuZnSOD
expression and rates of free radical production in
SODl-G93A spinal cord correlates with the timing of
pathological changes in the mice. Subtle vacuolar
changes in the proximal axons of spinal motor neurons
are discernible by 30 days of age.'8,25.26By 60 days of
age, most ventral horn motor neurons show vacuolar
changes in their cell body, but significant loss of motor
neurons does not occur until 90 days of age." Brain
content of mutant human CuZnSOD never reaches
similar levels, no increase in oxygen radical content is
observed, and patholo
is absent from the cerebrum
of SODl-G93A
As our measurements of
free radical content in brain were averaged over the
bulk of the tissue, we would have failed to detect local
increases in free radical content in small brain structures. For example, loss of dopamine neurons occurs in
the substantia nigra of SODI-G93A mice27 and both
vacuolar changes and neuron loss are observed in
brainstem motor n u ~ l e i . ~ These
~ , ~ * structures are so
small, however, that a local increase in free radical content would probably be lost in measurements of bulk
brain.
The most straightforward hypothesis that explains
these data is that in situ production of free radicals in
SOD l-G93A spinal cord simply parallels the tissue
content of mutant human CuZnSOD protein. This
precedes motor neuron loss and is consistent with the
hypothesis that motor neuron disease is initiated by oxidative injury. Furthermore, a series of therapeutic
studies with the SODl-G93A mouse model lend additional support to this hypothesis. First, copper chelation therapy with D-penicillamine delayed onset of disease and improved survival, although whether the
treatment actually reduced spinal cord CuZnSOD activity levels was not reported.28 Second, two different
studies show that treatment with antioxidants delays
onset of disease. In the first, dietary supplementation
with vitamin E and selenium delayed onset of disease
and transiently maintained motor behavior,29 whereas
in the second, treatment with carboxyfullerenes (malonic acid substituted "Bucky balls") delayed disease onset and improved survival.3o Third, "double transgenic" mice, which coexpress both SODl-G93A and
bcl-2 transgenes, show delayed onset of disease and improved s ~ r v i v a l . ~In' cell culture, increased expression
of bcl-2 decreases lipid peroxidation, suggesting that
bcl-2 may directly or indirectly regulate a cellular antioxidant
which would be consistent with
its effects in this model. Increased expression of bcl-2
also inhibits entry into an apoptotic pathway for cell
death.32933
The alternative hypothesis is that the increase in free
radical content that we measure in SODI-G93A
mouse spinal cord is a secondary response to the tissue
damage initiated by expression of the mutant protein.
In this view, disease is not initiated by free radical
mechanisms. This viewpoint also ignores the biochemical studies performed with highly purified, human
CuZnSOD, which show that the G93A mutation enhances Fenton-like catalysis of hydroxyl radical production from hydrogen peroxide."," It may be, however,
that a further increase in CuZnSOD expression is induced by tissue damage. That might create a feedforward mechanism that accelerates oxidative damage
to vulnerable cells.
These studies confirm and extend in vitro studies of
altered enzymatic function in mutant human CuZnSOD.'"'' They suggest a pathophysiological basis for
the proposed gain-of-function mechanism of disease
and indicate that AZN or other brain-penetrating antioxidants may be of therapeutic benefit in ALS and
perhaps other chronic neurodegenerative diseases in
which free radical mechanisms have been implicated.
Supported in part by grants from the National Science Foundation
to Drs Gurney and Becker.
We thank Tim Fleck, Carol Himes, and Susan Kuiper for assistance
with these experiments.
References
1. Williams DB, Windebank AJ. Motor neuron disease (amyotrophic lateral sclerosis). Mayo Clin Proc 1991;66:54-82
2. Rosen DR, Siddique T, Patterson D, et al. Murations in CuZn
superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature 1993;362:59-62
3. Deng H-X, Hentati A, Tainer JA, et al. Amyotrophic lateral
sclerosis and structural defects in CuZn superoxide dismutase.
Science 1993;362:1047-105 I
4. Borchelt DR, Lee MK, Slunt HS, et al. Superoxide dismutase 1
with mutations linked to familial amyotrophic lateral sclerosis
possesses significant activiry. Proc Natl Acad Sci USA 1994;91:
8292- 8296
5. Andersen PM, Nilsson P, &a-Hurula V, et al. Amyotrophic
larcral sclerosis associated with homozygosity for an Asp"AIa
mutation in CuZn-superoxide dismutase. Nat Gcnet 1995;lO:
61-66
6. Gurney ME, Pu H, Chiu AY, et al. Motor neuron degeneration
in mice that express a human CuZn superoxide dismutase mutation. Science 1994;264: 1772-1775
7. Wong PC, Pardo CA, Borchelt DR, et al. An adverse property
of a familial ALS linked SOD1 mutation causes motor neuron
disease characterized by vacuolar degeneration og mitochondria.
Neuron 1995;14:1105-1116
8. Dal Canto MC, Gurney ME. A low expressor line of transgenic
mice carrying a mutant human CuZn superoxide dismutase
(SOD 1) gene develops pathological changes which most closely
resemble those in human ALS. Acta Neuropathol 1997;93:537550
9. Gurney ME. Transgenic animal models of familial amyotrophic
larcral sclerosis. J Neurol 1997;244(Suppl 2):s15-S20
10. Wiedau Pazos M, Goto JJ, Rabizadeh S,et al. Altered reactivity
of superoxide dismutasc in familial amyotrophic lateral sclerosis.
Science 1996;271:515-518
11. Yitn MB, Kang JH, Yim HS, et al. A gain of function of an
amyotrophic lateral sclerosis associated Cu,Zn superoxide dismutase mutant: an enhancement of free radical formation due
to a decrease in K,, for hydrogen peroxide. Proc Natl Acad Sci
USA 1996;93:5709-5714
12. Hodgson EK, Fridovich I. The interaction of bovine erythrocyte superoxide dismutase with hydrogen peroxide: inactivation
of the enzyme. Biochemistry 1975;14:5294-5299
13. Yim MB, Chock PB, Stadtman ER. Copper zinc superoxide
Liu et al: Oxygen Radicals in FAIS Transgenic Mice
769
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
dismutase catalyzes hydroxyl radical production from hydrogen
peroxide. Proc Natl Acad Sci USA 1990;87:5006-5010
Yim MB, Chock PB, Stadtman ER. Enzyme function of copper
zinc superoxide dismutase as a free radical generator. J Biol
Chem 1993;268:4099-4 105
Lyons TJ, Liu H , Goto JJ, et al. Mutations in copper-zinc superoxide dismutase that cause amyotrophic lateral sclerosis alter
the zinc binding site and the redox behavior of the protein.
Proc Natl Acad Sci USA 1996;93:12240-12244
Becker DA. Highly sensitive colorimetric detection and facile
isolation of diamagnetic free radical adducts of novel chromotropic nitrone spin trapping agents readily derived from guaiazulene. J Am Chem SOC1996;118:905-906
Althaus JS, Fleck TJ, Becker DA, et al. Azulenyl nitrones: colorimetric detection of oxyradical end products and neuroprotection in the gerbil forebrain ischemia/reperfusion model. Free
Radic Biol Med 1998;24:738-744
Chiu AY, Zhai P, Dal Canto MC, et al. Age dependent penetrance of disease in a transgenic mouse model of familial
amyotrophic lateral sclerosis. Mol Cell Neurosci 1995;6:349362
Beauchamp C , Fridovich I. Superoxide dismutase: improved assays and an assay applicable to acrylamide gels. Anal Biochem
1971;44:276-287
Althaus JS, Andrus PK, Hall ED, et al. Improvements in the
salicylate trapping method for measurement of hydroxyl radical
levels in the brain. In: Ohnishi ST, Ohnishi T, ed. Central
nervous system trauma: research techniques. Boca Raton, FL:
CRC Press, 1995:437-444
Tsuda T, Munthasser S, Fraser PE, et al. Analysis of the functional effects of a mutation in SODl associated with familial
amyotrophic lateral sclerosis. Neuron 1994;13:727-736
Brujin LI, Becher MW, Lee MK, et al. ALS-linked SODl mutant G85R mediates damage to astrocytes and promotes rapidly
progressive disease with S O D 1-containing inclusions. Neuron
1997;18:327-338
Kellogg EW, Fridovich I. Superoxide dismutase in the rat and
770 Annals of Neurology
Vol 44
No 5
November 1998
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
mouse as a function of age and longevity. J Gerontol 1976;31:
405-408
De Haan JB, Newman JD, Kola I. Cu/Zn superoxide dismutase mRNA and enzyme activity, and susceptibility to lipid
peroxidation, increases with aging in murine brains. Mol Brain
Res 1992;13:179-1 87
Dal Canto MC, Gurney ME. The development of CNS pathology in a murine transgenic model of human ALS. Am J
Pathol 1994;145:1271-1279
Dal Canto MC, Gurney ME. Neuropathological changes in
two lines of mice carrying a transgene for mutant human CuZn
S O D and in mice overexpressing wild type human SOD: a
model of familial amyotrophic lateral sclerosis (FALS). Brain
Res 1995;676:25-40
Kostic V, Gurney ME, Deng H-X, et al. Midbrain dopaminergic neuronal degeneration in a transgenic model of familial
amyotrophic lateral sclerosis. Ann Neurol 1997;41:497-504
Hottinger AF, Fine EG, Gurney ME, et al. The copper chelator
d-penicillamine delays onset of disease and extends survival in a
transgenic mouse model of familial amyotrophic lateral sclerosis. Eur J Neurosci 1997;9:1548-1551
Gurney ME, Cutting FB, Zhai P, et al. Benefit of vitamin E
riluzole and gabapentin in a transgenic model of familial amyotrophic lateral sclerosis. Ann Neurol 1996;39:147-157
Dugan LL, Turetsky DM, D u C, et al. Carboxyfullerenes as
neuroprotective reagents. Proc Natl Acad Sci USA 1997;94:
9434-9439
Kostic V, Jackson-Lewis V, de Bilbao F, et al. Bcl-2: prolonging
life in a transgenic model of familial amyotrophic lateral sclerosis. Science 1997;277:559-562
Kane DJ, Sarafian TA, Anton R, et al. Bcl-2 inhibition of neural death: decreased generation of reactive oxygen species. Science 1993;262:1274-1277
Hockenbery DM, Oltvai ZN, Yin XM, et al. Bcl-2 functions in
an antioxidant pathway to prevent apoptosis. Cell 1993;75:
24 1-25 1
Документ
Категория
Без категории
Просмотров
2
Размер файла
881 Кб
Теги
production, mode, transgenic, lateral, mouse, familiar, enhance, radical, sclerosis, oxygen, amyotrophic
1/--страниц
Пожаловаться на содержимое документа