Feb 8, 1995 - Inhibits the Death of Rat Sympathetic ... of thinking about the ..... for the following reasons. First,. Western blots clearly demonstrated the ex-. A.
0026.895X/95/061095.06$3.O0/0
Copyright All rights
© by The American Society of reproduction in any form
MOLECULAR
for Pharmacology
ACCELERATED
and
Experimental
Therapeutics
reserved. (1995).
47:1095-1100
PHARMACOLOGY,
COMMUNICATION
Expression of Human Copper/Zinc-Superoxide Inhibits the Death of Rat Sympathetic Neurons Withdrawal of Nerve Growth Factor JOAQUIN RICHARD
JORDAN,1 J. MILLER
Departments
GHANASHYAM
of Pharmacological
and
Chicago,
Chicago,
Illinois
Received
February
8, 1 995; Accepted
D. GHADGE,1
Physiological
JOCHEN
Sciences
(J.J.,
H. M.
J.H.M.P.,
PREHN,
P.T.T.,
PETER
R.J.M.)
and
Dismutase Caused by
T. TOTH,
Neurology
RAYMOND
(G.D.G.,
P. ROOS,
R.P.R.),
The
and
University
of
60637 March
1 0, 1995
SUMMARY Rat superior
cervical ganglion neurons require the presence of factor (NGF) to develop and survive in culture. If NGF is removed from the culture medium, then the neurons die of programmed cell death. We investigated the potential role of
sion of the Ca2-binding protein calbindin D28k or the enzyme -galactosidase did not. We also observed that treatment of the cells with the cytokine transforming growth factor-31 , which has been shown to protect neurons against oxidative injury,
Ca2
delayed
nerve growth and
reactive
oxygen
species
in this
process.
We
found
that overexpression of human wild-type copper/zinc-superoxide dismutase in cultured superior cervical ganglion neurons, using an adenovirus-based vector, substantially protected the cells
from
the
effects
Elucidation tion
of NGF
of the
is of critical
development
withdrawal,
molecular
importance
and
basis
disease.
Recently,
have
general under
for understanding conditions. For
citotoxicity useful The
directed
involved
in the normal
now
well
way
might
important
and
that
activated
inappropriate
under neural
dismutase;
PCD,
programmed
BHK, baby hamster
zineethanesulfonic
kidney;
cell death;
produced
widely the
types
by
or die of ex-
to be extremely
NGF
withdrawal.
These
data
types
common
features
roles
for
a program is
of
species
a process
been
system is same path-
pathological
effects
con-
degeneration,
that
has
NGF,
TGF, transforming
nerve
growth
growth
intracellular
in
the
amyotrophic
factor;
that
include
role
of the enzyme
i.e.,
in culture
(8, 9) and
by the
cervical
PBS, phosphate-buffered
the
removal and
ganglion;
the
discovery
induces
apoptotic
linked
,
to
present model
of NGF investigated
SOD-i
reducing
in humans,
directly
(12),
recently protein
or
SOD
and
oxygen
antiapoptotic
Moreover,
are
Ca24
of reactive
formation
enzyme (10).
some
possible
has
the
radicals
also
including
The
the
oxygen
system,
superior
to
different as being
are
neurodegeneration idea
neurons
neurons SCG,
6, 7).
inhibiting
same
These
lateral sclerosis (11). In the we have used a well established
nervous
in
discussed there
mediators,
the
down-regulation
experiments
factor;
by
of reactive
tions
important
necrosis
usually
considered.
and by
act
clearly
process,
(1,
apoptosis
of cultured
thetic
be
species
death
the
to
several
may
is and
are
of degenerative
highlighted
Bcl-2
necrosis
and
oxygen in
It
of PCD
degeneration.
PCD
different
(5).
roles
of neuronal
reactive this
discussed
relative
Although
This work was supported by a postdoctoral fellowship from NATO to J.J., by a grant from the German Research Foundation to J.H.M.P., by United States Public Health Service Grants DA02575, DA02121, and MH40165 to R.J.M., and by United States Public Health Service Grant N521442 and the ALS Association to R.P.R. 1 JJ and G.D.G. contributed equally to this study. ABBREVIATIONS:
been
consider
possible
neurons live the concepts
proven
now
areas
some
the problem. die by initiating
suicide
be inadvertently to
example,
degeneraof neuronal
development of the nervous (4). The possibility that the
established leading
cell
us with why
(3) have
about can
cell
several
to provide
ofPCD
of thinking that neurons
genetically
ditions,
started
(1, 2) and ways idea
of nerve
death
overexpres-
to an understanding
ofinvestigation principles different
although
cell
suggest a role for reactive oxygen species in triggering programmed cell death of rat sympathetic neurons upon growth factor withdrawal.
mutafamilial
series ofPCD
of in
from
sympa-
the
possible
copper/zinc-superoxide
saline; HEPES, 4-(2-hydroxyethyl)-i
-pipera-
acid.
1095
I 096 roles
of
some
degeneration. based system in
et aL
Jordan
of
sympathetic
oxygen
these
To do for the
factors
this, we efficient
neurons.
species
may
in
Our
play
the
resulting
have used expression results
a key
role
nerve
a novel of genes indicate
in
this
that type
cell
adenovirusof interest reactive
of
neuronal
death.
Materials
and
Methods
em
Cell culture. SCGs were isolated from 1-5-day-old Holzman rats (13). Rat pups were killed by a cut made into the thorax and through the heart, under anesthesia. This procedure reduced the bleeding from the carotid artery during dissection of the ganglia. After dissection, ganglia were incubated for 40 mm, at 37#{176}, in Ca2and Mg-free
Hanks’
solution
containing
0.2%
trypsin
(Worthington,
Freehold, NJ) and were then dissociated into single cells by gentle trituration. Dissociated cells were washed twice with L-15 culture medium containing 0.1% (w/v) bovine serum albumin (CalbiochemNovabiochem, La Jolla, CA) and were plated on poly-L-lysineand laminin-treated
glass
coverslips.
Glass
coverslips
were
prepared
by
incubation for a minimum of 15 mm with 0.5 mg/ml poly-L-lysine (Sigma Chemical Co., St. Louis, MO) dissolved in borate buffer. The coverslips
were
then
rinsed
three
times
with
water
and
covered
with
0.02 mg/ml laminin in L-15 medium, for a minimum of 1 hr, before the cells were replated. Cells were maintained in L-15 culture medium supplemented with 1 p.g/ml NGF (United States Biochemicals, Cleveland, OH), 5% rat serum (GIBCO, Grand Island, NY), glucose, bicarbonate, penicillin-streptomycin, and vitamins, as described (14). Construction of recombinant, replication-defective adenovirus containing wild-type copper/zinc-SOD (SOD-i). A SOD-i cDNA clone was obtained from the American Type Culture Collection (15). A blunt-ended PstI-NheI fragment was inserted into the EcoRV site of the transfer vector pAdKN downstream of the elongation factor i-a promotor and upstream of the cellular heavy chain enhancer (4F2) and the bovine growth hormone polyadenylation site (pA+), to generate pAdKN.SODWT plasmid. The pAd.KN plasmid also contains 0-i and 9-16 map units ofDNA sequence of adenovirus type 5. To construct a replication-defective adenovirus (AdEF.SODWT) containing SOD-i cDNA, NheI-digested pAdKN.SODWT and XbaIand Clal-digested adenovirus type 5 sub36O were co-transfected into human embryonic kidney 293 cells using a calcium phosphate-DNA precipitation method. The recombinant AdEF.SODWT was plaque purified three times to isolate a homogeneous population ofthe virus. The virus used in the studies was further purified by CsC1 isopycmc ultracentrifugation
and
line (10 mM HEPES, 10% glycerol. Purified titered
by
plaque
was
dialyzed
against
HEPES-buffered
sa-
mM NaCl, 2 mM MgCl2, pH 7.5) containing virus was stored at -70#{176} in small aliquots and
140 assay.
Adenoviruses
carrying
plicity of infection of i500 or were not infected. Four days after infection, neurons were washed with cold PBS and scraped. Neurons were resuspended in 15 l ofH10K10D1 buffer (iO mM HEPES, iO m’vi KC1, 1 mM dithiothreitol, pH 7.4) and lysed on ice with 1% Nonidet P-40 for 15 mm. Cytoplasmic protein (10 j.g), obtained after centrifugation at rpm for iO mm at 4#{176}, was analyzed for the presence of i9-kDa SOD-i protein by Western blotting using horseradish peroxidase-conjugated anti-SOD-i polyclonal antibody (catalogue no. PP077; Binding Site) and an enhanced chemiluminescence West-
the
genes
for
cal-
bindin D28k (AdCABP-i) and f3-galactosidase (AdBAc.LacZ) have been described (i6, i7). Infection protocoL After 3 days in vitro the cultures were infected with the respective virus at a multiplicity ofinfection of i500. For this purpose, aliquots of the high titer viruses were added to the culture medium and distributed by gentle agitation. After 2 hr, the infection medium was exchanged with regular, NGF-supplemented, culture medium. After an additional 3 days, cultures either were subjected to NGF deprivation (see NGF withdrawal) or were fixed, and expression of the respective proteins was verified by immunocytochemistry. No nonspecific toxic effects ofthe viruses were apparent under the conditions used in these studies. The percentage of neurons showing positive staining for each antigen was >95% in every experiment and was usually not significantly different from 100%. In contrast, no neurons were stained for the antigens under noninfected conditions (see Fig. 2). Western blot analysis. Sympathetic neurons were infected with either AdBAc.LacZ, AdCABP-i, or AdEF.SODWT virus at a multi-
blotting
SOD
detection
system
activity
gels.
AdEF.SODWT
or not
(Amersham).
BHK-2i infected,
cells and
were
96
hr
either
after
infected
infection
with
cytoplasmic
extracts were prepared in H10K10D1 buffer containing 0.5% Nonidet P-40. One milligram of protein was electrophoresed on a 7.5% nondenaturing polyacrylamide gel at 4#{176}, and SOD enzyme activity was determined by using the nitroblue tetrazolium method (i8). Immunohistochemistry. In each experiment successful expression
of
proteins
was
verified
by
immunohistochemistry.
For
this
purpose cultures were fixed with 4% paraformaldehyde in culture medium at 37#{176} for iS mm. After cells were washed three times with 0.i M PBS (0.1 M Na2 HPO4, 0.i M NaH2 P04, 0.2 M NaCl), pH 7.4, they
were
permeabilized
with
0.i%
Triton
X-100
(Eastman
Kodak,
Rochester, NY) in PBS for 2.5 mm. The coverslips were then incubated in blocking medium [0.i% Tween-20 (Sigma) and 4% bovine serum albumin (Sigma) in 0.1 M PBS] for i hr at room temperature. Incubations with primary antibodies were performed overnight at 4#{176}, using murine monoclonal anti-human SOD-i (1/300), anti-calbinthn D28k (1/1000), and anti-3-galactosidase (111000) antibodies (Sigma), diluted in blocking medium. Monoclonal antibodies were detected using
anti-mouse
Grove,
IgG
(Jackson
ImmunoResearch
diluted 112000 in blocking ABC kit (Vector Laboratories, Burlingame, visualized using 3,3’-diaminobenzidine substrate.
Coverslips
were
mounted
Laboratories,
solution, CA). (Sigma)
PA),
in
90%
West
and a Vectastain The peroxidase was as the chromogenic glycerol
with
0.1
phosphate buffer containing 0.01% NaN3. TGF-f11 treatment. Twenty-four hours before NGF cultures received 10 ng/ml TGF-131 (recombinant human
withdrawal,
Systems,
solution
Minneapolis,
saline
containing
MN) i mg/ml
from
a
1000
ovalbumin
and
ng/ml 4 mri
form;
stock HC1).
M
R & D (in
Controls
were
treated with an equal volume of the vehicle; TGF-J31 or vehicle was also present during the period of NGF withdrawal. NGF withdrawal. For induction of PCD, the original culture medium was exchanged for fresh culture medium with (control conditions) or without (NGF withdrawal) the NGF supplement. Neuronal degeneration was analyzed by phase-contrast microscopy after 24, 48, and 72 hr. Neurons exhibiting a rough appearance with irregularly
shaped
cell
bodies,
blebs,
and
vacuoles,
followed
by
cell
shrinkage, neurite fragmentation, and loss ofphase-brightness, were considered to be damaged. Frequently, nuclear fragmentation could be observed by phase-contrast microscopy. Healthy neurons were identified as having round to oval, smooth, phase-bright cell bodies and intact neurites. A total of iOO-200 neurons were examined in three to five randomized subfields ofthe coverslips. Similar numbers of neurons were examined under all experimental conditions. Each condition was represented by six coverslips. All experiments were performed
in quadruplicate.
Results It has survival dent culture factor
been widely demonstrated of rat sympathetic neurons
on the
form
In
is thought
medium, anism
presence
medium.
that of PCD
of the
appropriate
growth
most
paradigms
the
to be NGF.
neurons shows (12).
die
over
many We
that the development in cell culture are
When
NGF
a period of the
readily
hallmarks confirmed
factors
major days, of the this
in the
important
is removed
of a few
and depen-
from
the
by a mech“apoptotic”
to be
the
case.
SOD Neurons
in our
withdrawal criteria and
cultures
(Fig. suggestive
membrane
drawal,
died
1) and
over
many
of apoptosis, blebbing.
neurons
also
a 3-day
exhibited
incuding
Forty-eight stained
D28k,
on
neuronal
the
human
positively
death
observed
chosen
because
they
iological
ways
of perturbing
cell
D28k) (i7, 19) and reactive anions (copper/zinc-SOD-i).
oxygen
of the protein expression chemistry increased
neurons 3 days
species
made to infection
of each antigen was 24 hr after infection. over
the
next
24-48
after
(calbindin
was
then
appropriate Evidence
in
of NGF
was
using
the
rus)
(Figs.
i and
of
time
D28k
we
not
or
A
reasons.
First,
us
(24
that
to
and
Western
blots
control
48 hr)
de-
in
the
conditions vi-
virus of cells
perform neurons in these clearly
protected
as well
expression, died
any
overexpression in SCG activity
The
expression
of SOD-i with number
copper/zinctime. neurons
-galactosidase
allow
think
the that
from
in
to the control died by 72 hr
f3-galactosidase-expressing
points
treated small
man SOD-i protein increase in enzyme 2).
of by
2). Copper/zinc-SOD-i
not the
did
iO% died
the
ofSOD-i
In contrast neurons had
different D28k
to the effects
calbindin
assays,
not
over
ofoverexpression
had
1097
PCD
observed
SOD-i-overexpressing
calbindin
at earlier
In contrast
cultures
(Fig.
death
control cells Although
immunohistoof staining
stabilized
was
only cells
(i.e.,
ing
efficient expression we used adenovirusapproximately 100%
express the (Fig. 2) (17).
effect
withdrawal,
of cell
neurons
as superoxide
expression
intensity
a striking
NGF
presence
phys-
staining
SOD-1-overexpressing gree
These
such
in
these neurons (Figs. 1 and 2). situation, in which >90% of the
cal-
relatively
obtained by The intensity
hr and
with-
removal. of Ca24
13-Galactosidase
could be after viral
somata NGF
diminution
next 7 days. We observed
f3-galactosidase
NGF
handling
No
double-stranded
represent
used as an additional control. To achieve of these proteins in cultured neurons, based vectors (i7). Using these vectors,
NGF
procedure. of the proteins and
after
were
after
for
copper/zinc-SOD-i,
proteins
after
morphological
shrunken
hours
DNA breaks, using the TUNEL staining We examined the effects of expression bindin
period of the
and
(Fig.
1).
cells expressas
rapidly
(Fig. i). in the SCG conventional
as
did
neuron enzyme
of immunoreactive
hu-
was associated cells, for the
with an following
demonstrated
the
ex-
B S
I
I
1
I S..
23
I Turn. 1k)
C
caubundun
Time
(ii)
TIm.
(3,)
D
928K
I
I
I
I #{149}5#{149}
24
TIrne(h)
E
24 1mw
(3,)
FIg. I . Histograms representing the time course of survival of rat SCG neurons grown in the presence (U) or absence (Li) of NGF. A, Control conditions, without any treatment; B, cells overexpressing p-galactosidase; C, cells overexpressing calbindin D28k; D, cells overexpressing human SOD-i ; F, treatment with TGF-p1 24 hr before the start of NGF withdrawal (four experiments in each case; see Materials and Methods). *, p