Content uploaded by Sian M Thomas
Author content
All content in this area was uploaded by Sian M Thomas on Sep 16, 2015
Content may be subject to copyright.
Biochem.
J.
(1986)
240,
489-494
(Printed
in
Great
Britain)
Free-radical-mediated
fragmentation
of
monoamine
oxidase
in
the
mitochondrial
membrane
Roles
for
lipid
radicals
Roger
T.
DEAN,*
Sian
M.
THOMAS*
and
Anthony
GARNERt
*Cell
Biology
Research
Group
and
tDepartment
of
Biochemistry,
Brunel
University,
Uxbridge,
Middlesex
UB8
3PH,
U.K.
A
flux
of
hydroxyl
radicals
generated
by
y-irradiation
can
fragment
monoamine
oxidase
in
the
membrane
of
submitochondrial
particles.
This
fragmentation
can
be
inhibited
by
mannitol
and
in
addition
is
more
extensive
in
monoamine
oxidase
preparations
that
have
been
depleted
of
lipid.
This
latter
observation
is
consistent
with
the
higher
yields
of
fragmentation
induced
by
hydroxyl
radicals
in
soluble
proteins
in
the
absence
of
added
lipids.
In
the
absence
of
oxygen,
y-irradiation
of
submitochondrial
particles
leads
to
cross-linking
reactions.
A
flux
of
hydroperoxyl
radicals
also
causes
fragmentation,
whereas
one
of
superoxide
is
virtually
inactive
in
this
respect.
The
irradiation
of
submitochondrial
particles
leads
in
addition
to
the
accumulation
of
products
of
lipid
peroxidation.
When
these
irradiated
preparations
are
exposed
to
ferrous
or
cupric
salts
a
further
fragmentation
of
monoamine
oxidase
ensues,
especially
at
acid
pH.
These
transition-metal-catalysed
reactions
do
not
occur
with
irradiated
preparations
depleted
of
lipid,
and
the
post-irradiation
protein
modifications
are
concomitant
with
further
lipid
peroxidation.
The
data
indicate
roles
for
lipid
radicals
in
both
fragmentation
and
cross-linking
reactions
of
proteins
in
biological
membranes.
These
reactions
may
have
an
important
bearing
on
control
of
protein
activity
and
of
protein
turnover
in
membranes.
INTRODUCTION
It
is
known
that
oxygen-centred
free
radicals
can
readily
fragment
proteins
in
solution
and
even
in
complex
biological
structures
such
as
cartilage
[1-6].
Amino
acid
modification
and
cross-linking
reactions
(the
latter
particularly
in
the
absence
of
oxygen
[3,5])
also
occur.
The
products
of
these
reactions
are
usually
more
susceptible
to
enzymic
hydrolysis
than
the
native
molecule
[6].
Consequently,
free
radical
damage
to
proteins
occurring
within
cells
may
positively
influence
the
basal
rate
of
protein
breakdown
therein.
We
have
recently
shown
[7]
for
proteins
synthesized
in
isolated
mitochondria
that
this
prediction
is
upheld:
degradation
is
fastest
when
radical
fluxes
are
greatest,
in
the
case
of
mitochondria
in
State
4.
The
majority
of
the
endogenously
synthesized
mito-
chondrial
polypeptides
are
membrane-bound
[8]
so
the
influence
of
radicals
on
protein
breakdown
may
apply
not
only
to
proteins
in
aqueous
environments
but
perhaps
also
within
biological
membranes.
In
the
present
study,
we
have
therefore
followed
the
reaction
between
free
radicals
and
an
intrinsic
outer
membrane
protein
of
the
mitochondria,
monoamine
oxidase.
We
have
con-
sidered
whether
oxygen-centred
radical
attack
may
directly
fragment
the
monoamine
oxidase
in
situ.
In
biological
membranes
lipid
peroxidation
is
frequently
a
consequence
of
radical
attack
[9]
and
so
we
have
considered
the
possibility
that
radicals
produced
in
lipid
molecules
may
also
interact
with
monoamine
oxidase
to
cause
degradation.
Our
results
provide
evidence
both
for
direct
oxygen-centred
radical
fragmentation
of
mono-
amine
oxidase
and
for
a
role
of
some
lipid-derived
radicals
in
fragmentation
of
membrane-bound
proteins.
Such
reactions
are
likely
to
influence
the
overall
rate
of
Vol.
240
protein
turnover
in
membranes
(consistent
with
our
earlier
evidence
[7]),
and
may
also
relate
to
regulation
of
protein
activity.
MATERIALS
AND
METHODS
Materials
Chelex-treated
water
and
glow-cleaned
glass
were
used
for
all
experiments.
[3H]Pargyline
was
from
New
England
Nuclear,
and
non-radioactive
pargyline
from
Sigma.
Desferrioxamine
was
a
kind
gift
from
Ciba-
Geigy,
and
Trolox
was
from
Aldrich.
Other
chemicals
were
the
purest
commercial
grade.
Preparation
of
labelled
monoamine
oxidase
within
mitochondrial
membranes
Rat
liver
mitochondria
were
isolated
as
described
previously
[7];
on
each
occasion,
two
rats
were
used.
Monoamine
oxidase
was
then
labelled
in
situ
using
the
active-site-directed,
covalent-binding
inhibitor
pargyline.
This
provides
a
convenient
tracer
for
the
fate
of
molecules
of
the
enzyme
[7,10].
SDS/polyacrylamide-gel
electrophoresis
in
reducing
conditions
showed
that
polypeptides
of
approx.
60
kDa
were
labelled,
providing
that
pargyline
was
used
at
low
concentrations
(1
guM).
This
corresponds
well
to
the
reported
polypeptides.
Mitochondria
were
sonicated,
to
give
submitochondrial
particles,
before
use
in
experiments
and
were
always
washed
(by
sedimentation
in
an
Eppendorf
Microfuge
for
5
min)
immediately
before
use.
Lipid-depleted
mitochondria
were
made
from
portions
of
preparations
of
pargyline-labelled
intact
mitochondria,
by
extraction
with
aqueous
methyl
ethyl
ketone
by
a
modification
of
the
procedure
of
Erkstedt
&
Oreland
[11].
After
489
R.
T.
Dean,
S.
M.
Thomas
and
A.
Gamer
extraction,
the
residue
was
resuspended
in
0.1
M-
potassium
phosphate
buffer
containing
1
mM-EDTA.
The
membranes
were
sedimented
at
27000
g
for
10
min,
and
then
washed
three
times
with
distilled
water,
using
the
same
sedimentation
conditions.
More
than
80%
of
the
lipid
phosphorus
was
removed
by
this
procedure.
All
the
preparations
were
stored
at
-20
°C
for
up
to
4
weeks,
in
small
portions
so
that
repeated
freeze-thawing
was
not
needed.
Radical-generating
systems
Oxygen-centred
radicals
were
generated
as
in
our
previous
work
using
steady
state
radiolysis
at
50
Gy/min
with
a
60Co
source
[4].
This
permitted
the
selective
generation
of
hydroxyl,
superoxide
and
hydroperoxyl
radicals,
by
the
addition
of
N20/02
(4:
1),
10
mM-formate
(pH
7.2)
gassed
with
air
and
10
mM-formate
(pH
4)
gassed
with
air,
respectively,
and
the
investigation
of
their
effects
on
the
labelled
enzyme
in
situ.
Experiments
were
normally
done
in
10
mM-potassium
phosphate
buffer,
pH
7.2;
the
only
protein
present
was
that
of
the
mitochondria
(1.0-2.5
mg/ml).
Oxygen
was
not
supplied
for
some
irradiations.
Measurement
of
degradation
of
monoamine
oxidase
Initially,
we
measured
trichloroacetic
acid-soluble
fragments
produced
from
the
radioactively
labelled
enzyme.
Carrier
bovine
serum
albumin
(0.5
%,
w/v)
was
added
and
mitochondria
were
then
precipitated
with
a
final
concentration
of
trichloroacetic
acid
of
5%
(w/v).
The
precipitate
was
sedimented
on
a
Microfuge
and
the
supernatant
sampled
for
scintillation
counting.
The
pellet
was
redissolved
in
formic
acid
and
counted
separately.
Degradation
is
calculated
as
acid-soluble
radioactivity
as
a
percentage
of
the
total
in
the
system.
This
method
only
detects
small
fragments
of
monoamine
oxidase.
Larger
fragments
were
detected
by
SDS/poly-
acrylamide
gel
electrophoresis
(9
%
polyacrylamide)
under
reducing
conditions
[12].
Gels
were
silver
stained
or
sectioned
for
radioactivity
counting
after
dissolution
in
100
vol.
H202
(1
ml/gel
slice)
at
60
°C
overnight.
For
technical
reasons
(see
results)
it
was
later
necessary
to
avoid
the
acid
pH
of
the
trichloroacetic
acid
method;
instead,
ZnSO4
was
added
(2.5
g/100
ml
final
concentration),
followed
by
100
,u
of
saturated
barium
hydroxide/ml.
The
resulting
suspension
was
centrifuged
as
above.
The
total
radioactivity
in
each
incubation
was
measured
by
counting
an
unprecipitated
sample.
Two-phase
degradation
experiments
(phase
1:
oxygen-
centred
radical
attack;
phase
2:
transition
metal
attack
on
the
products)
Mitochondrial
membranes
containing
labelled
mono-
amine
oxidase
were
exposed
to
1000
Gy
of
irradiatio6
in
the
system
generating
hydroxyl
radicals
in
the
presence
of
oxygen.
Then
they
were
transferred
to
a
37
°C
incubator
in
10
mM-potassium
phosphate
buffer,
pH
7.2,
in
the
presence
of
various
additives
such
as
ferrous
or
cupric
ions.
For
addition
to
these
experiments
ferrous
sulphate
was
made
up
as
a
fresh
10
mm
stock
in
acetate
buffer,
pH
5.6
(100
mM).
After
incubation
for
chosen
times
at
37
°C
in
the
presence
of
the
transition
metals,
monoamine
oxidase
degradation
was
measured.
Characterization
of
products
of
radical
attack
on
labelled
monoamine
oxidase
It
was
necessary
to
establish
that
the radioactive
tracer
used
to
label
monoamine
oxidase,
pargyline,
remained
associated
with
peptides
during
radical
attack.
Using
the
trichloroacetic
acid-soluble
degradation
products,
two
methods
were
applied.
Firstly,
gel
filtration
on
Bio-Gel
P-2;
secondly,
h.p.l.c.
on
Hypersil
ODS
(5
,m)
with
reverse-phase
elution
with
a
gradient
of
acetonitrile
in
7
mm-ammonium
acetate,
pH
3.8.
The
column
effluents
were
monitored
at
257
nm
(the
absorption
maximum
of
pargyline)
and
samples of
fractions
were
also
taken
for
scintillation
counting.
Measurement
of
lipid
peroxidation
in
mitochondria
during
radical
attack
In
various
relevant
conditions
lipid
peroxidation
was
measured
as
thiobarbituric-acid-reactive
material
[14].
Data
presentation
Experiments
are
representative
of
several.
They
were
performed
routinely
with
duplicate
samples
and
means
are
shown,
except
where
otherwise
indicated.
Such
duplicates
differed
from
each
other
by
less
than
3%.
In
some
other
experiments
greater
statistical
variation
was
obtained,
and
larger
numbers
of
replicates
were
used:
for
these
means
are
shown
and
degree
of
replication
stated.
In
addition
S.D.
and/or
statements
of
statistical
signifi-
cance
are
given.
Although
the
results
shown
are
all
qualitatively
reproducible,
there
were
significant
inter-
experimental
variations,
and
the
data
are
selected
to
reveal
the
maximum
extent
of
these
variations:
these
mainly
seem
to
originate
in
differences
between
successive
mitochondrial
preparations
from
pairs
of
animals.
There
are
also
differences
due
to
the
different
mitochondrial
preparations
used
(normal,
lipid-depleted,
and
the
control
preparations
obtained
in
parallel
with
the
lipid
depletion).
RESULTS
Direct
radical
attack
on
monoamine
oxidase
As
shown
in
Table
1,
the
hydroxyl
radical
was
able
to
fragment
monoamine
oxidase
only
to
a
remarkably
limited
degree
in
the
dose
range
studied.
The
superoxide
radical
was
essentially
ineffective,
while
an
intermediate
activity
was
observed
with
the
hydroperoxyl
radical
generating
system
although
this
could
be
a
result
of
conformational
changes
induced
by
acid
pH.
In
the
absence
of
oxygen,
no
fragmentation
was
detectable.
The
attack
by
hydroxyl
and
hydroperoxyl
radicals
could
be
inhibited
by
appropriate
radical
scavengers,
such
as
mannitol
in
the
case
of
the
hydroxyl
radical.
In
Table
2,
a
comparison
of
the
fragmentation
of
monoamine
oxidase
in
normal
and
lipid-depleted
submitochondrial
preparations
is
shown.
The
hydroxyl
radical
could
cleave
the
enzyme
in
the
lipid-depleted
membrane
more
readily
than
that
in
the
replete
membrane.
The
hydroperoxyl
radical
could
cleave
the
enzyme
similarly
in
both
preparations,
though
it
was
much
less
effective
than
the
hydroxyl
radical
against
the
lipid-depleted
preparation.
The
data
indicate
that
the
1986
490
Monoamine
oxidase
fragmentation
Table
1.
Degradation
of
monoamine
oxidase
in
mitochondrial
membranes
by
oxygen-radical-generating
systems
Degradation
is
measured
as
low-Mr
fragments,
and
expressed
as
the
increase
over
the
no-irradiation
blank,
as
%
of
the
total.
Degradation
observed
in
these
blanks,
which
were
treated
in
parallel,
was
always
less
than
1
%;
generally
those
at
pH
4
were
slightly
higher
than
those
at
pH
7.2.
Similar
results
were
obtained
with
either
trichloroacetic
acid
or
alkaline
zinc
processing
methods.
Degradation
(%)
Radical
system
Irradiation
dose
(Gy).
...
300
600
1200
OH(N20/02)
+
10
nM-Mannitol
OH(N20)
02'-
(air)
H02'
(air,
pH
4)
+
10
mM-Mannitol
0.06
0.23
0.68
0 0
0.06
0
0
0.01
0
0
0.02
0.16
0.30
0.51
0.15
0.30
0.48
Table
2.
Comparison
of
radical
fragmentation
of
monoamine
oxidase
in
lipid-depleted
and
normal
submitochondrial
particles
From
a
single
preparation
of
labelled
mitochondria,
both
a
lipid-depleted
and
a
normal
preparation
(treated
in
parallel
throughout)
were
obtained.
Degradation
was
expressed
as
in
Table
1.
Degradation
in
blanks
treated
in parallel
was
always
less
than
1
%,
and
those
for
lipid-depleted
preparations
were
much
lower
than
those
for
the
normal
preparations.
Each
value
represents
the
mean
of
four
separate
irradiations.
*indicates
a
significant
difference
from
the
appropriate
no-irradiation
blank
where
P
<
0.01
in
a
two-tailed
t
test.
Degradation
(%)
Radical
system
Irradiation
dose
(Gy).
.
.
300
600
1200
OH(N20/02)
Normal
Lipid-depleted
H02'
(air,
pH
4)
Normal
Lipid-depleted
action
of
the
hydroxyl
radical
on
the
enzyme
may
be
restricted
by
the
presence
of
lipid,
while
that
of
the
hydroperoxyl
radical
may
depend
on
the
presence
of
lipid,
since
this
radical
is
inactive
against
bovine
serum
albumin
in
free
solution
[6].
There
was
also
a
marked
inverse
relationship
between
the
protein
concentration
of
the
mitochondrial
suspen-
sion
during
the
irradiation
and
the
extent
of
fragment-
ation
observed.
Similarly,
mannitol
was
a
more
effective
scavenger
the
lower
the
protein
concentration
(results
not
shown).
Characterization
of
the
products
of
radical
attack
on
monoamine
oxidase
On
Bio-Gel
P-2
chromatography
of
the
trichloroacetic
acid-soluble
material
from
radical
attack
on
the
monoamine
oxidase
several
radioactive
peaks
were
observed,
mostly
eluting
earlier
than
free
pargyline.
About
25%
in
most
cases
was
eluted
at
the
position
of
free
pargyline.
Thus
it
seemed
that
a
proportion
of
the
degradation
products
might
not
result
from
proteolysis
but
from
some
other
reaction
producing
free
pargyline.
However,
the
proteolytic
enzyme
trypsin
could
also
generate
a
component
eluting
in
the
free
pargyline
position.
Thus
products
other
than
free
pargyline
contaminate
the
peak
corresponding
to
free
pargyline.
0.01
0.11
0.23*
0.18*
0.25*
0.65*
0.06
0.12
0.35
0.03
0.09
0.41
*
Thus
the
vast
majority
of
the
degradation
products
were
other
than
free
pargyline
and
were
likely
to
be
pargyline
attached
to
peptides.
This
was
further
supported
by
experiments
with
h.p.l.c.
A
variety
of
acetonitrile
gradients
were
each
able
to
separate
more
than
80%
of
the
degradation
products
from
the
major
peak
of
the
radioactive
free
pargyline.
Thus
at
a
conservative
estimate
more
than
80%
of
the
acid-soluble
degradation
products
in
all
the
studied
cases
could
be
shown
by
the
combination
of
these
two
separation
methods
to
be
other
than
free
pargyline:
the
labelling
method was
thus
sufficient
to
trace
degradation
of
monoamine
oxidase.
Two-phase
experiments
on
degradation
of
monoamine
oxidase
In
view
of
the
small
degree
of
fragmentation
achieved
during
irradiation,
we
hypothesized
that
many
of
the
radicals
generated
react
with
lipid
rather
than
with
protein.
Were
this
the
case,
the
lipids
in
the
irradiated
membrane
would
be
expected
to
contain
a
residue
of
hydroperoxides
and
other
molecules
which
could
be
degraded
during
a
subsequent
transition
metal
attack,
generating
further
radicals
[15]
which
might
be
more
able
to
fragment
monoamine
oxidase.
Thus,
when
irradiated
mitochondria
were
subsequently
Vol.
240
491
R.
T.
Dean,
S.
M.
Thomas
and
A.
Garner
C
0
'-
co
L._
cn
a0
1:
7
4
pH
Fig.
1.
Fe2+-catalysed
degradation
of
mon
at
low
pH
A
large
volume
of
submitochondrial
]
protein/ml)
was
irradiated
in
glass-dist
(under
OH/02
conditions;
1000
Gy).
then
diluted
into
10
vol.
of
water
titra
pH,
with
(@)
or
without
(0)
100
/,M-
phase
incubation
was
then
at
37
°C
ft
measurement
of
degradation
used
method.
Degradation
is
expressed
as
t
0
min
to
30
min
of
incubation
in
the
se
shows
that
this
catalysis
of
fragmentation
of
the
monoamine
oxidase
is
a
feature
of
a
low
pH
and
not
of
the
presence
of
trichloracetic
acid.
Strikingly,
fragment-
ation
of
lipid
hydroperoxides
also
occurs
best
with
transition
metals
at
low
pH
[16].
It
is
also
notable
that
Fe2+/trichloroacetic
acid
fragments
non-irradiated
monoamine
oxidase
only
slightly
(result
not
shown).
The
data
indicate
a
substantial
difference
in
the
action
of
Fe2+
at
low
and
neutral
pH,
and
also
that
the
fragmentation
observed
in
irradiated
samples
in
the
two-phase
experiments
above
occurred
mainly
during
the
acid
phase.
To
avoid
this
acid-catalysed
degradation
of
the
monoamine
oxidase
we
precipitated
intact
protein
by
the
alkaline
zinc
method.
In
control
experiments
we
showed
that
incubation
of
pre-irradiated
mitochondria
in
the
presence
of
zinc
and
barium
hydroxide
with
or
without
the
presence
of
transition
metal
resulted
in
fragmentation
indistinguishable
from
that
in
comparable
controls
not
incubated
in
the
presence
of
alkaline
zinc.
This
3
2
1
precipitation
method
was
therefore
acceptable.
Fig.
2
shows
the
kinetics
of
transition-metal-catalysed
degrada-
oamine
oxidase
in
situ
tion
of
pre-irradiated
monoamine
oxidase
using
this
alkaline
zinc
precipitation
method.
A
lipid-dependent
particles
10
mgof
transition-metal-catalysed
reaction
in
pre-irradiated
partilees
(1
mg
of
samples
was
thus
observed,
which
could
be
prevented
by
hed
warticlereat
pH7
a
1.5-fold
molar
excess
of
desferrioxamine.
Non-
ted
to
each
desired
irradiated
samples
treated
in
parallel
showed
no
-Fe2+.
The
second-
degradation
until
90-120
min
of
second-phase
incuba-
or
30
min,
and
the
tion,
after
which
it
accelerates.
Under
these
experimental
the
alkaline
zinc
conditions,
iron
may
be
cycling
between
its
two
the
increment
from
oxidation
states,
by
virtue
of
several
different
reactions.
mcond
phase.
Ca2+
(100
/tM)
can
also
cause
fragmentation.
Its
effects
are
largely
inhibited
by
150
,#M-EDTA
(result
not
exposed
to ferrous
iron
at
100
/LM
much
larger
quantities
of
trichloroacetic
acid-soluble
material
(up
to
6%
of
the
total)
were
observed,
and
then
the
amount
of
acid-soluble
products
declined,
giving
a
biphasic
reaction.
The
decline
is
presumably
due
to
cross-linking
reactions.
However,
it
was
notable
that
the
zero-time
values,
determined
as
soon
as
possible
after
addition
of
the
transition
metal,
were
much
above
those
of
controls
which
did
not
receive
transition
metal.
Generation
of
acid-soluble
radioactivity
occurred
after
a
lag
in
non-irradiated
samples
also.
The
effects
of
Fe2+
in
both
irradiated
and
non-irradiated
samples
could
be
drastic-
ally
reduced
by
a
1.5-fold
molar
excess
of
desferriox-
amine.
These
data
suggested
that
degradation
might
be
continuing
even
during
the
brief
processing
in
trichloro-
acetic
acid.
Thus
we
found,
as
predicted,
that
very
rapid
production
of
acid-soluble
material
occurs
when
irradia-
ted
mitochondria
with
labelled
monoamine
oxidase
are
incubated
in
the
presence
of
transition
metal
in
5
%
(w/v)
trichloroacetic
acid,
which
had
a
pH
of
approx.
1.5.
Up
to
15
%
of
the
label
can
be
made
soluble
in
5
%
trichloroacetic
acid
within
90
min
in
these
conditions.
This
reaction
can
be
inhibited
only
40
%
by
the
iron
chelator
desferrioxamine
at
a
1.5-fold
molar
excess
added
after
the
trichloroacetic
acid.
Greater
but
still
incomplete
inhibition
was
obtained
with
a
10-fold
molar
ratio
of
desferrioxamine
to
transition
metal.
Better
inhibition
is
obtained
when
desferrioxamine
is
present
with
iron
before
the
trichloroacetic
acid
is
added.
Fig.
1
5-
4-
ae
3-
c
0
._o
I..2
-
(-
~0
0
0.5
1
Time
(h)
Fig.
2.
Lipid-dependent
degradation
of
amine
oxidase
by
Fe2+
I
1.5
2
pre-irradiated
mono-
The
two-phase
experiment
was
conducted
as
for
Fig.
1,
except
that
in
the
second
phase
samples
were
diluted
into
10
mM-potassium
phosphate
buffer,
pH
7.2,
either
with
no
further
addition
(0)
or
in
the
presence
of
100
1sM-Fe2+
(-)
in
the
second
phase.
Results
from
samples
incubated
in
the
second
phase
with
iron
plus
a
1.5-fold
molar
excess
of
desferrioxamine
were
very
similar
to
those
of
the
controls
shown
in
the
Figure.
1986
-0
0-
0
492
Monoamine
oxidase
fragmentation
Table
3.
Lack
of
iron-dependent
degradation
in
pre-irradiated
lipid
depleted
monoamine
oxidase
The
two-phase
experiment
was
conducted
as
described
in
Fig.
1.
Each
value
above
represents
the
mean
of
four
separate
irradiations
treated
separately
in
the
second
phase,
and
S.D.
values
are
given
in
parentheses.
**indicates
a
significant
difference
between
iron-containing
and
control
samples
at
P
<
0.05,
and
*similarly
at
P <
0.01.
Degradation
(%)
Sample
Time
...
1.5
h
6
h
22.5
h
Normal
Control
+
Fe2+
Lipid-depleted
Control
+
Fe2+
1.14
(0.12)
1.22
(0.23)
1.21
(0.18)
1.03
(0.09)
1.28
(0.14)**
1.60
(0.16)*
1.20
(0.50)
1.16
(0.59)
1.33
(0.53)
1.05
(0.49)
1.35
(0.45)
1.66
(0.43)
shown).
In
parallel
experiments
on
bovine
serum
albumin
in
solution
in
the
absence
of
lipids
no
such
post-irradiation
effect
of
transition
metals
on
polypeptide
fragmentation
was
observed
either
at
neutral
pH,
or
in
the
presence
of
5%
trichloroacetic
acid
or
zinc/barium
hydroxide
(results
not
shown.)
Nor
can
transition
metals
alone
fragment
bovine
serum
albumin
[6],
though
they
can
eventually
lead
to
fragmentation
of
monoamine
oxidase
in
the
membrane.
Thus
two
lines
of
evidence
indicate
that
some
of
the
fragmentation
of
monoamine
oxidase
we
observed
depends
on
the
presence
of
lipid:
firstly
the
reaction
of
the
hydroperoxyl
radical,
and
secondly
the
lack
of
effect
of
transition
metals
on
pre-irradiated
soluble
proteins
in
comparison
with
their
considerable
effectiveness
on
pre-irradiated
protein
within
lipid
membranes.
This
interpretation
was
supported
by
further
experiments
comparing
lipid-depleted
with
normal
monoamine
oxi-
dase
preparations.
Table
3
shows
that
in
pre-irradiated
lipid-depleted
preparations,
there
is
very
little
subsequent
fragmentation
during
incubation
in
the
presence
or
absence
of
iron.
In
contrast,
iron
can
induce
a
detectable
fragmentation
of
the
enzyme
in
the
pre-irradiated
normal
mitochondrial
preparations.
The
corresponding
controls
(without
iron)
showed
no
significant
degradation.
Gel
electrophoresis
of
products
of
radical
attack
on
monoamine
oxidase
During
attack
by
OH-/O2
(up
to
1000
Gy)
only
a
slight
reduction
in
the
radioactivity
associated
with
the
60
kDa
labelled
band
on
the
gels
could
be
detected,
but
no
clear
fragments.
The
H02'
and
02'-
radical-generating
systems
had
no
detectable
effect
on
the
electrophoretic
distribution
of
radioactivity.
The
OH
radical
in
the
absence
of
oxygen
gave
significant
amounts
of
radioacti-
vity
in
larger
molecules,
presumably
due
to
cross-linking.
The
products
of
the
two-phase
experiments
contained
many
labelled
components,
some
larger
and
some
smaller
than
the
enzyme
polypeptide.
Radioactivity
was
also
detected
at
the
dye
front
in
amounts
correlated
with
those
observed
by
zinc/barium
hydroxide
or
trichloro-
acetic
acid
precipitation
(as
appropriate).
Comparison
of
fragmentation
of
monoamine
oxidase
with
lipid
peroxidation
We
have
measured
thiobarbituric
acid-reactive
pro-
ducts
of
lipid
peroxidation
in
conditions
corresponding
to
those
described
above.
In
agreement
with
the
literature
on
mitochondrial
peroxidation
[17,18],
little
peroxidation
was
observed
in
vitro
up
to
3
h
at
37
°C
unless
100
/uM-Fe2+
was
added,
when
thiobarbituric
acid-reactive
materials
were
rapidly
generated,
reaching
plateau
values
corresponding
to
approx.
0.7
nmol
of
malondialdehyde/mg
of
protein
within
60-120
min.
Ascorbate
(0.5
mM)
together
with
Fe2+
(5-100
,M)
did
not
enhance
peroxidation
over
that
with
iron
alone,
and
also
had
no
effect
on
proteolysis
(not
shown).
1000
Gy
of
irradiation
in
the
OH'/O2
system
gave
larger
quantities
of
malondialdehyde/mg
of
protein.
When
pre-irradiated
mitochondria
are
incubated
with
the
transition
metals
rather
little
alteration
in
the
thiobarbi-
turic
acid-reactive
material
present
is
observed.
Only
extremely
small
quantities
of
thiobarbituric
acid-reactive
materials
could
be
generated
from
the
lipid-depleted
preparations.
The
kinetics
of
lipid
peroxidation
and
protein
fragmentation
were
compared
in
many
experi-
ments.
It
was
observed
that
the
two
processes
were
concomitant.
The
lag
before
peroxidation
and
fragment-
ation
were
detected
in
unirradiated
samples
exposed
to
iron
could
be
substantially
compressed
by
repeated
thawing
and
brief
storage
at
4
°C
of
the
labelled
submitochondrial
particles,
perhaps
due
to
progressive
depletion
of
antioxidants
[171.
In
such
samples
peroxi-
dized
by
storage,
reactions
causing
cross-linking,
such
as
to
decrease
the
quantity
of
acid-soluble
pargyline-peptide
radioactivity,
could
also
be
detected.
DISCUSSION
The
evidence
above
indicates
that
free
radicals
can
fragment
and
otherwise
damage
monoamine
oxidase
when
it
is
in
the
outer
mitochondrial
membrane.
However,
it
implies
also
that
lipids
protect
monoamine
oxidase
from
fragmentation
in
comparison
with
protein
in
lipid-free
solution.
Thus
for
a
given
dose
of
hydroxyl
radical
attack,
for
instance,
a
smaller
degree
of
fragmen-
tation
of
monoamine
oxidase
is
observed
in
the
normal
mitochondrial
membrane
than
in
the
lipid-depleted
membranes,
and
a
much
smaller
degree
than
is
observed
with
bovine
serum
albumin
in
solution
[6].
One
reason
for
this
is
presumably
that
lipid
consumes,
directly
or
indirectly
in
peroxidative
reactions,
many
of
the
primary
radicals,
leaving
far
fewer
reacting
directly
with
the
protein.
The
two-phase
experiments,
however,
show
that
Vol.
240
493
494
R.
T.
Dean,
S.
M.
Thomas
and
A.
Garner
the
consequence
of
lipid
peroxidation,
the
presence
of
lipid
hydroperoxides
and
other
materials,
may
allow
another
route
for
protein
fragmentation.
Thus,
when
these
products
react
with
transition
metals
more
radical
production
is
initiated
and
some
(unknown)
components
of
this
reaction
are
relatively
effective
in
protein
fragmentation.
It
seems
likely
therefore
that
some
lipid
radicals
can
fragment
proteins
and
that
this
effect
is
occurring
during
both
phases
of
the
monoamine
oxidase
experiments
described
above,
in
conjunction
with
other
cross-linking
reactions.
Previous
literature
[2,17,19]
has
primarily
emphasized
the
possibility
of
lipid-induced
cross-linking
of
protein,
rather
than
fragmentation.
However,
a
limited
amount
of
data
indicate
fragmentation
due
to
lipid
radicals,
in
particular,
fragmentation
of
cytochrome
c
[20].
Lipid-dependent
degradation
of
gelatin
has
also
been
observed
[21];
this
may
have
involved
peptide-bond
scission,
but
this
was
not
conclusively
shown.
A
further
formal
possibility
exists
for
the
mechanism
of
fragmenta-
tion
of
monoamine
oxidase
in
the
two-phase
experi-
ments:
that
other
oxidized
proteins
generated
during
the
first
phase
give
rise
to
further
radical
fluxes
which
fragment.
This
possibility
is
perhaps
unlikely
in
view
of
the
lack
of
a
second-phase
fragmentation
in
the
lipid-depleted
enzyme,
but
requires
further
investigation
with
simpler
model
systems
containing
membrane
proteins.
The
relevance
of
lipid-dependent
polypeptide
frag-
mentation
to
intact
biological
membranes
rather
than
to
artificial
mixtures
of
proteins
and
lipids
has
not
previously
been
described.
We
suggest
that
this
inter-
action
is
probably
of
considerable
biological
significance.
We
imagine
that
lipid
alkoxy
and
peroxyl
radicals
may
be
involved:
the
reactive
moieties
require
identification.
In
mitochondria
and
chloroplasts,
whose
radical
fluxes
are
metabolically
controlled
(see
[7]
for
discussion),
such
reactions
may
modify
protein
activities
and
influence
overall
protein
turnover;
but
such
regulatory
effects
may
be
much
more
widespread
[24].
For
example,
the
accentuated
degradation
at
low
pH
may
magnify
the
role
of
this
lipid-dependent
process
in
low-pH
environments
such
as
lysosomes
[22],
and
the
sequestered
eroding
surface
of
inflammatory
sites
[23].
R.
T.
D.
is
grateful
for
grants
from
the
Agricultural
and
Food
Research
Council
(AFRC)
and
the
Arthritis
and
Rheumatism
Council
(U.K.),
and
S.
M.
T.
for
a
studentship
from
the
AFRC.
REFERENCES
1.
Garrison,
W.
M.
(1968)
Curr.
Top.
Radiat.
Res.
4,
43-94
2.
Schaich,
W.
M.
(1980)
CRC
Crit.
Rev.
Food
Sci.
Nutr.
13,
89-129,
131-159,
189-244
3.
Scheussler,
H.
&
Schilling,
K.
(1984)
Int.
J.
Radiat.
Biol.
45,
267-287
4.
Dean,
R.
T.,
Roberts,
C.
R.
&
Forni,
L.
G.
(1984)
Biosci.
Rep.
4,
1017-1026
5.
Dean,
R.
T.,
Roberts,
C.
R.
&
Jessup,
W.
(1985)
in
Intracellular
Protein
Catabolism
(Khairallah,
E.,
Bond,
J.
&
Bird,
J.
W.
C.,
eds.),
pp.
341-350,
A.
R.
Liss,
New
York
6.
Wolff,
S.
P.
&
Dean,
R.
T.
(1985)
Biochem.
J.
234,399-403
7.
Dean,
R.
T.
&
Pollak,
J.
K.
(1985)
Biochem.
Biophys.
Res.
Commun.
126,
1082-1089
8.
Tzagoloff,
A.
(1983)
Mitochondria,
Plenum,
New
York
9.
Slater,
T.
F.
(1984)
Biochem.
J.
222,
1-12
10.
Evans,
P.
J.
&
Mayer,
R.
J.
(1984)
Biochem.
J.
219,
61-72
11.
Erkstedt,
B.
&
Oreland,
L.
(1976)
Biochem.
Pharmacol.
25,
119-124
12.
Laemmli,
U.
K.
(1970)
Nature
(London)
277,
680-682
13.
Wray,
W.,
Boulikas,
T.,
Wray,
V.
P.
&
Hancock,
R.
(1981)
Anal.
Biochem.
118,
197-203
14.
Slater,
T.
F.
&
Sawyer,
B.
C.
(1971)
Biochem.
J.
123,
823-831
15.
O'Connell,
M.
J.
&
Garner,
A.
(1984)
Int.
J.
Radiat.
Biol.
44,
615-622
16.
O'Brien,
P.
J.
(1969)
Can.
J.
Biochem.
47,
485-492
17.
Tappel,
A.
L.
(19750
in
Pathobiology of
Cell
Membranes
(Trump,
B.
F.
&
Arstila,
A.
U.,
eds.),
pp.
145-169,
Academic
Press,
New
York
18.
Vladimirov,
Yu.
A.,
Olenev,
U.
I.,
Suslova,
T.
B.
&
Cheremisina,
Z.
P. (1980)
Adv.
Lipid
Res.
17,
174-249
19.
Karel,
M.,
Schaik,
K.
&
Roy,
R.
B.
(1975)
J.
Agr.
Food
Chem.,
23,
159-163
20.
Tappel,
A.
L.
(1973)
Fed.
Proc.
Fed.
Am.
Soc.
Exp.
Biol.
32,
1870-1878
21.
Zirlin,
A.
&
Karel,
M.
(1969)
J.
Food
Sci.
34,
160-164
22.
Dean,
R.
T.
&
Barrett,
A.
J.
(1976)
Essays
Biochem
12,
1-40
23.
Etherington,
D.
J.,
Pugh,
D.
&
Silver,
J.
A.
(1981)
Acta
Biol.
Med.
Germ.
40,
1625-1629
24.
Wolff,
S.
P.,
Gamer,
A.
&
Dean,
R.
T.
(1986)
Trends
Biochem.
Sci.
11,
27-31
Received
12
May
1986/25
July
1986;
accepted
15
August
1986
1986
