Identification and isolation of soluble histones from bovine milk and serum

Abstract

An immunoassay for soluble histones as trace components of biological fluids was developed on the basis of the dual capacity of histones to bind solid-phase DNA and monoclonal anti-histone antibody. Application of this histone-capture assay to bovine milk resulted in a positive signal, and DNA-cellulose chromatography was used to isolate histone-like material in microgram quantities. Western-blot analysis using a panel of anti-histone antibodies demonstrated the presence of histones H2A, H2B and H4 in apparently intact form. DNAase digestion experiments indicated that at least a portion of milk histone was complexed to DNA. Bovine serum was analysed in the same manner on serial DNA-cellulose columns, and H4 and partially degraded H2A were detected by Western-blot analysis. The finding of soluble histones in bovine milk and serum may account for unexpected results when these biological fluids are used as blocking reagents in Western blots and other immunoassays and may have ramifications in the origin and significance of anti-histone antibodies in human disease.

Full text

Biochem.
J.
(1987)
244,
675-682
(Printed
in
Great
Britain)
Identification
and
isolation
of
soluble
histones
from
bovine
milk
and
serum
Shinobu
WAGA,
Eng
M.
TAN
and
Robert
L.
RUBIN*
Scripps
Clinic
and
Research
Foundation,
10666
N.
Torrey
Pines
Road,
La
Jolla,
CA
92037,
U.S.A.
An
immunoassay
for
soluble
histones
as
trace.
components
of
biological
fluids
was
developed
on
the
basis
of
the
dual
capacity
of
histones
to
bind
solid-phase
DNA
and
monoclonal
anti-histone
antibody.
Application
of
this
histone-capture
assay
to
bovine
milk
resulted
in
a
positive
signal,
and
DNA-cellulose
chromatography
was
used
to
isolate
histone-like
material
in
microgram
quantities.
Western-blot
analysis
using
a
panel
of
anti-histone
antibodies
demonstrated
the
presence
of
histones
H2A,
H2B
and
H4
in
apparently
intact
form.
DNAase
digestion
experiments
indicated
that
at
least
a
portion
of
milk
histone
was
complexed
to
DNA.
Bovine
serum
was
analysed
in
the
same
manner
on
serial
DNA-cellulose
columns,
and
H4
and
partially
degraded
H2A
were
detected
by
Western-blot
analysis.
The
finding
of
soluble
histones
in
bovine
milk
and
serum
may
account
for
unexpected
results
when
these
biological
fluids
are
used
as
blocking
reagents
in
Western
blots
and
other
immunoassays
and
may
have
ramifications
in
the
origin
and
significance
of
anti-histone
antibodies
in
human
disease.
INTRODUCTION
Macromolecules
normally
found
within
cells
have
also
been
reported
in
other
compartments
in
vivo.
DNA
was
detected
in
human
serum
(Tan
et
al.,
1966)
at
concentrations
ranging
from
0.004
to
0.4
,ug/ml
(McCoubrey-Hoyer
et
al.,
1984)
and
in
the
,ug/ml
range
in
some
patients
with
systemic
lupus
erythematosus
(Raptis
&
Menard,
1980).
Naked
DNA
has
a
half-life
of
less
than
2
min
in
mice
(Emlen
&
Mannik,
1980),
but
the
stability
of
endogenous
DNA
in
the
circulation
may
be
longer.
Protection
against
nucleolytic
degradation
of
serum
DNA
may
be
afforded
by
non-covalently
bound
proteins,
and
histones
would
be
natural
candidates.
However,
there
is
little
evidence
for
the
presence
of
histones
in
vivo
outside
the
cell
nucleus.
Soluble
DNA
is
secreted
by
lymphocytes
activated
with
phytohaemagglutinin
(Rogers
et
al.,
1972),
and
a
portion
of
lymphocyte
membrane
DNA
is
associated
with
histone-like
proteins
(Rogers
&
Kawahara,
1981).
Cell-membrane-associated
nucleohistone
or
DNA
has
been
confirmed
by
numerous
studies
(Rekvig
&
Han-
nestad,
1980;
Horneland
et
al.,
1983;
Jacob
et
al.,
1984;
Holers
&
Kotzin,
1985;
Bennett
et
al.,
1985),
and
nucleohistone-like
material
was
detected
in
tissue-culture
media
of
mouse
spleen
cells
(Atkinson
et
al.,
1985)
and
neural
retinal
cells
(Shubert
&
LaCorbiere,
1985).
It
appears
likely,
therefore,
that
nucleohistones
and
DNA
may
not
be
uncommon
in
extracellular
fluids,
raising
the
possibility
that
soluble
histones
may
exist
in
the
circulation.
In
addition
to
being
of
fundamental
interest,
histones
in
certain
biological
fluids
may
have
practical
ramifi-
cations.
Bovine
serum
or
non-fat
dry
milk
have
come
into
increasing
use
as
protein
carriers
or
blocking
agents
in
solid-phase
immunoassays
for
autoantibodies
(Johnson
et
al.,
1984).
Discrepancies
in
the
literature
on
anti-histone-antibody
specificities
led
to
the
suspicion
that
the
blocking
medium
may
affect
the
signal
in
these
assays.
Previous
studies
demonstrating
DNA
in
serum,
and
the
possibility
that
this
DNA
may
be
complexed
with
histones,
suggested
that
soluble
histones
may
account
for
these
artefacts.
For
these
reasons
a
study
was
undertaken
to
search
for
histone-like
material
in
bovine
serum
and
milk.
MATERIALS
AND
METHODS
Culture
assay
for
soluble
histones
Calf
thymus
DNA
(Calbiochem-Behring,
La
Jolla,
CA,
U.S.A.)
was
diluted
in
PBS
(0.01
M-sodium
phosphate
buffer/0.
14
M-NaCl,
pH
7.2)
to
a
concen-
tration
of
2.5
,tg/ml,
followed
by
boiling
for
15
min
and
immediate
cooling
in
an
ice
bath.
After
overnight
incubation
in
Immulon
II
microtitre
plates
(Dynatech
Laboratories,
Alexandria,
VA,
U.S.A.),
the
DNA
solution
was
decanted,
and
0.25
ml
of
gelatin
(Baker)
at
1
mg/ml
was
incubated
in
the
wells
for
at
least
18
h.
Solutions
to
be
tested
for
histone
activity
were
incubated
in
DNA-coated
wells
for
2
h
at
room
temperature
with
agitation.
An
IgG2a
mouse
monoclonal
anti-histone
antibody
reactive
with
histones
H2A,
H2B
and
H2A-H2B
dimers
(antibody
'b';
Table
1)
was
then
used
to
detect
DNA-captured
histones.
After incubation
for
1.5
h,
bound
anti-histone
antibody
was
detected
with
peroxidase-conjugated
goat
anti-mouse
IgG
(Tago,
Burlingame,
CA,
U.S.A.)
according
to
e.l.i.s.a.
methodo-
logy
previously
described
(Rubin
et
al.,
1983).
Total
histone
(Calbiochem-Behring)
and
polylysine
(Miles
Laboratories,
Elkhart,
IN,
U.S.A.)
were
employed
as
controls
in
this
assay.
DNA-cellulose
affinity
chromatography
Non-fat
dry
milk
(Von's
Grocery
Co.,
Los
Angeles,
CA,
U.S.A.)
was
dissolved
in
PBS
at
150
mg/ml,
and
Abbreviations
used:
e.l.i.s.a.,
enzyme-linked
immunosorbent
assay;
FBS,
*
fo
whom
correspondence
and
reprint
requests
should
be
sent.
fetal-bovine
serum;
HI,
H2
etc.,
histone
HI,
H2
etc.
Vol.
244
675

S.
Waga,
E.
M.
Tan
and
R.
L.
Rubin
Table
1.
Individual-histone-binding
specificities
of
monoclonal
anti-histone
antibodies
Individual
histones
were
purified
from
calf
thymus
as
previously
described
(Rubin,
1986b),
and
their
reactivities
with
murine
monoclonal
anti-histone
antibodies
were
determined
by
e.l.i.s.a.
Antibody
binding
to
histone/polypeptide
(A)
Monoclonal
antibody
HI
H2A
H2B
H3
H4
H2A-H2B
Polylysine
Polyarginine
a
b
c
0.03
0.01
0.06
3.74
0.07
0.12
0.03
0.04
5.26
0.14
0.03
0.07
4.15
0.06
1.25
2.57
1.82
4.55
0.05
0.01
0.06
0.05
0.03
0.02
whey
was
prepared
from
this
solution
by
ultracentri-
fugation
in
the
SW
41
rotor
(r
11
cm)
of
a
Beckman
centrifuge
at
39000
rev./min
for
1
h
at
4
IC
(Lonnerdal
et
al.,
1985).
DNA-cellulose
was
prepared
by
the
method
of
Biogioni
et
al.
(1978).
Briefly,
cellulose
(Cellex
N-1;
Bio-Rad
Laboratories,
Richmond,
CA,
U.S.A.)
was
activated
with
cyanuric
chloride
(Sigma
Chemical
Co.,
St.
Louis,
MO,
U.S.A.),
which
had
been
purified
by
extraction
with
chloroform.
Dichlorotriazinyl-cellulose
(activated
cellulose)
was
suspended
at
a
concentration
of
1
g/4
ml
of
calf
thymus
DNA
solution
(0.5
mg/ml)
with
stirring
at
4
'C.
Spectrophotometric
analysis
showed
that
more
than
80%
of
the
DNA
was
bound
to
cellulose.
A
5
ml
portion
of
settled
DNA-cellulose
containing
2.7
mg
of
DNA
was
suspended
in
either
800
ml
of
whey
or
1000
ml
of
FBS
(Irvine
Scientific,
Santa
Ana,
CA,
U.S.A.),
and
incubated
overnight
at
4
'C
with
gentle
stirring.
The
DNA-cellulose
was
collected
by
centri-
fugation
at
3000
rev./min
(1100
g)
for
20
min
at
4
'C
and
transferred
to
a
9
mm-diameter
glass
column.
After
washing
with
500
ml
of
PBS
at
a
rate
of
0.5
ml/min,
the
column
matrix
was
treated
with
0.1
M-HCl,
and
1
ml
fractions
were
collected
at
a
flow
rate
of
0.14
ml/min.
Each
fraction
was
evaluated
by
its
A230,
was
neutralized
with
0.01
M-Tris/NaOH,
and
tested
in
the
capture
assay
for
soluble
histone.
Histone-positive
fractions
were
pooled,
dialysed
against
0.05
M-acetic
acid
and
freeze-
dried.
Protein
determination
on
the
water-soluble
material
was
by
the
method
of
Lowry
et
al.
(1951),
with
bovine
serum
albumin
as
standard.
As
a
control,
DNA-cellulose
to
which
no
biological
fluid
was
applied
was
subjected
to
the
same
washing
and
elution
procedure;
no
protein
or
histone
activity
was
detected.
SDS/polyacrylamide-gel
electrophoresis
SDS/polyacrylamide-gel
electrophoresis
on
a
15%0
(w/v)-acrylamide
slab
gel
with
a
6%
-acrylamide
stacking
gel
was
performed
by
the
method
of
Laemmli
(1970).
Disulphide
bonds
in
the
samples
were
reduced
by
boiling
for
5
min
in
0.0625
M-Tris
buffer,
pH
6.8,
containing
2.3%
(w/v)
SDS,
5%
(v/v)
2-mercaptoethanol
and
10%
(v/v)
glycerol.
Samples
applied
to
the
gel
(180
mm
x
150
mm
x
1.5
mm)
were
electrophoresed
at
10
mA
for
14
h
and
were
stained
with
0.05
0
Coomassie
Blue
R250
in
methanol/acetic
acid/water
(9:2:9,
by
vol.)
for
3
h.
Western-blot
analysis
After
SDS/polyacrylamide-gel
electrophoresis
the
samples
were
electrophoretically
transferred
to
nitro-
cellulose
by
the
method
of
Towbin
et
al.
(1979).
Transfer
took
place
at
260
mA
for
3
h
at
5
'C.
The
nitrocellulose
was
then
cut
into
strips,
which
were
subsequently
soaked
in
0.25%
(w/v)
gelatin
(Baker,
Phillipsburg,
NJ,
U.S.A.)
in
PBS
for
2h
at
room
temperature
to
block
non-specific
binding
sites.
A
set
of
monoclonal
antibodies
or
human
serum
with
anti-histone
activity
(described
below),
diluted
with
0.25%
gelatin
in
PBS/0.05
%
-Tween-20,
pH
7.4,
were
allowed
to
react
with
the
nitrocellulose
strips
for
1
h
at
room
temperature.
The
strips
were
then
washed
by
agitation
in
two
changes
of
PBS/Tween
for
1.5
h.
Bound
antibodies
were
detected
by
reaction
with
either
goat
anti-(mouse
K-chain)
(Southern
Biotech-
nology
Associates,
Birmingham,
AL,
U.S.A.)
for
the
monoclonal
antibodies,
or
goat
anti-human
IgG
(Tago,
Burlingame,
CA,
U.S.A.)
for
the
human
antibodies.
The
detecting
antibodies
were
radiolabelled
with
1251
(McConahey
&
Dixon,
1980)
and
diluted
in
0.25%
gelatin
in
PBS/Tween
to
2
x
105
c.p.m./ml.
After
extensive
washing
with
PBS/Tween
for
2.0
h,
the
strips
were
dried
and
transferred
to
a
cassette
fitted
with
enhancing
screen
and
Kodak
X-Omat
film
for
autoradio-
graphic
exposure
for
approx.
24
h
at
-70
'C.
Mouse
monoclonal
antibodies
and
human
sera
used
for
Western-blot
analysis
The
mouse
monoclonal
anti-histone
and
anti-DNA
antibodies
were
derived
from
autoimmune
mice
by
hybridoma
technology.
Anti-histone
antibody
'b'
and
the
anti-DNA
antibody
were
obtained
from
an
NZB/NZW
mouse
as
previously
described
(Kotzin
et
al.,
1984).
Monoclonal
anti-histone
antibodies
'a'
and
'c'
were
derived
from
MRL/lpr/lpr
mice.
As
shown
in
Table
1
and
Fig.
1,
antibody
'a',
an
IgG3,
showed
predominant
reactivity
with
histones
H2A
and
H4.
Antibody
'c'
was
of
the
IgG2b
isotype
and
was
essentially
monospecific
for
H2B.
The
specificity
of
the
monoclonal
anti-histone
antibodies
evaluated
by
e.l.i.s.a.
using
purified
individual
histones
(Table
1)
largely
agreed
with
their
reactivity
in
Western-blot
format
using
total
histones
separated
by
SDS/polyacrylamide-gel
electro-
phoresis
(Fig.
1).
Antibody
'b',
which
showed
pre-
dominant
reactivity
by
e.l.i.s.a.
with
the
H2A-H2B
complex,
by
Western
blot
detects
HI,
H2A,
H2B
and
a
smear
in
the
H2A/H2B
region
of
the
gel.
The
smear
is
due
to
cross-contamination
of
H2A
and
H2B
during
electrophoretic
transfer,
resulting
in
H2A-H2B-complex
formation
on
the
nitrocellulose
(Portanova
et
al.,
1986).
Human
sera
with
anti-histone
antibodies
were
obtained
from
patients
with
procainamide-induced
lupus
(Rubin
et
al.,
1985).
1987
676

Histones
in
milk
and
serum
a
b
c
6.0
5.0
4.0
HiL
E
H3
H2B-
H2A
/
H4-
0
..:
....
Fig.
1.
Western
blot
of
monoclonal
anti-histone
antibodies
Monoclonal
antibodies
'a',
'b'
and
'c'
were
allowed
to
react
with
total
calf
thymus
histones
after
separation
by
SDS/polyacrylamide-gel
electrophoresis
and
transfer
to
nitrocellulose.
DNAase
I
treatment
Samples
were
subjected
to
DNAase
treatment
in
a
solvent
consisting
of
0.25
mM-CaCl2/1
mM-MgCl2/
0.5
mM-phenylmethanesulphonyl
fluoride
(Calbiochem-
Behring)
in
PBS.
DNAase
I
(Sigma)
was
added
at
0.2
unit/ml
and
incubated
for
4
h
at
37
°C,
followed
by
addition
of
EDTA
to
4
mm
to
stop
the
reaction.
RESULTS
Detection
of
histone
activity
in
milk
It
was
anticipated
that,
if
histones
were
a
component
of
milk,
they
would
be
present
in
low
concentrations
and
detection
would
require
an
assay
of
high
sensitivity
and
specificity.
This
was
achieved
by
concentrating
the
putative
histone
components
by
their
natural
tendency
to
bind
to
DNA
in
a
histone
'capture'
assay.
The
assay
format
consisted
of
DNA
immobilized
to
polystyrene,
followed
by
incubation
with
the
test
material
and
detection
of
bound
histone
with
a
monoclonal
anti-
histone
antibody.
As
shown
in
Fig.
2,
a
total-histone
concentration
of
approx.
0.3
,ug/ml
was
detectable,
with
the
signal
increasing
linearly
up
to
1.3
,tg/ml.
Above
this
con-
centration,
histone
detection
decreased.
Polylysine,
a
basic
DNA-binding
polypeptide
which
does
not
react
with
the
monoclonal
anti-histone
antibody
(Table
1),
did
not
show
significant
activity.
When
milk
solution
was
applied
to
this
assay,
histone
reactivity
was
initially
detectable
at
a
milk-solids
concentration
of
1
mg/ml,
followed
by
an
approximately
linear
increase
and
plateau
at
6
mg/ml.
This
binding
curve
resembled
the
histone
dose-response
curve,
although
displaying
approx.
4-fold
less
signal
at
maximum
response.
In
contrast,
up
to
10%
(w/v)
milk
applied
to
wells
devoid
of
DNA
did
not
produce
histone
activity
(results
not
0//
0.4
1.6
6.3
25
100
Milk
(mg/ml)
0.08
0.08
0.3
1.3
5.0
20
[Histone
or
polylysinel
(gg/mr)
Fig.
2.
Histone
capture
assay
Increasing
concentrations
of
total
histones
(0),
polylysine
(0)
or
milk
(A)
were
added
to
microtitre
wells
coated
with
denatured
DNA.
Histone
binding
was
detected
by
an
indirect
e.l.i.s.a.
using
a
mouse
monoclonal
anti-histone
antibody.
shown),
indicating
that
the
histone-like
material
in
milk
bound
to
the
DNA
component
of
the
plate.
Isolation
of
histone
activity
from
milk
The
histone-capture-assay
results
suggested
that
DNA-cellulose
affinity
chromatography
may
be
a
promising
methodology
for
isolating
putative
histones
from
milk.
However,
flow
problems
of
concentrated
milk
solution
(presumably
due
to
casein
micelles)
hampered
this
method,
but
the
whey
component
prepared
by
ultracentrifugation
surmounted
this
problem.
As
shown
in
Fig.
3,
the
whey
obtained
from
ultracentrifugation
(U)
PBS
0.1
M-HCI
I
I
15.0k
*-
10.0-
5
5.0F
S
U
F
1
"300302
304306
308310
312
314
Fraction
no.
12.0
.-
I)
-
Fig.
3.
Isolation
of
histone
activity
from
milk
by
DNA-cellulose
affinity
chromatography
The
starting
milk
solution
(S)
was
ultracentrifuged
(U),
and
800
ml
of
whey
was
applied
to
the
DNA-cellulose
column.
After
the
initial
flow-through
(F),
the
column
was
extensively
washed
with
PBS,
followed
by
elution
of
protein
with
0.1
M-HCI.
Histone
activity
(L1)
and
protein
concentration
(0)
were
measured
in
fractions
after
appropriate
dilution.
Vol.
244
677

S.
Waga,
E.
M.
Tan
and
R.
L.
Rubin
gave
the
same
histone
signal
as
the
starting
milk
solution
(S),
suggesting
that
putative
histone
was
not
associated
with
insoluble
aggregates.
Histone
activity
of
the
flow-through
material
(F)
from
the
DNA-cellulose
column
was
decreased,
indicating
adsorption
of
histone
activity
on
to
the
column.
After
extensive
washing
with
PBS,
the
column
was
subjected
to
elution
with
0.1
M-HCI.
A
single
large
peak
w-as
detected
at
the
apparent
void
volume.
A
good
correspondence
between
A2M
(protein
concentration)
and
A410
(histone
activity
in
the
capture
immunoassay)
was
observed.
Histone-positive
fractions
were
pooled,
dialysed
and
freeze-dried
before
subsequent
analysis.
This
isolation
procedure
was
performed
three
separate
times.
Starting-
with
15
g
of
milk
protein,
the
maximum
yield
of
DNA-binding
proteins
was
4.1
mg,
which
contained
375
jug
of
histone
activity
by
the
histone-
capture
assay.
Characterization
of
DNA-binding
proteins
from
milk
Spectrophotometric
analysis
of
material
from
the
DNA-cellulose
column
(Fig.
4)
showed
an
absorption
maximum
at
273
nm
and
an
A280/A230
ratio
of
0.20.
No
peak
or
shoulder
was
observed
at
260
nm,
indicating
the
absence
of
significant
amount
of
DNA
in
the
preparation.
This
spectrum
can
be
compared
with
that
for
total
histones,
which
have
an
absorption
maximum
at
274
nm
and
an
A280/A230
ratio
of
0.13.
It
appears
likely,
therefore,
that
the
isolated
material
has
a
relatively
low
aromatic-amino-acid
content,
approximately
inter-
mediate
between
that
of
histones
and
a
protein
such
as
albumin.
By
using
SDS/
15%
-polyacrylamide-gel
electro-
phoresis the
isolated
DNA-binding
proteins
produced
three
dominant
bands
at
positions
corresponding
to
3.0
F
2.0
F
A
a
b
c
Molecular
mass
(kDa)
92.5-
_
66.2-
-
45.
.
....
J
Hi
31.
0-
21.5-
,--
H3
-H2B
*
-H2A
14.4-
-HA
Fig.
5.
SDS/polyacrylamide-gel
electrophoresis
of
proteins
isolated
from
milk
(lane
b)
Molecular
masses
of
marker
proteins
(lane
a)
and
total
calf
thymus
histones
(lane
c)
are
shown
for
comparison.
The
18
kDa
and
16
kDa
bands
in
the
milk-derived
material
had
positions
similar
to,
but
not
identical
with,
those
of
histones
H3,
H2B
and/or
H2A.
78
kDa,
18
kDa
and
16
kDa,
as
well
as
numerous
faint
bands
throughout
the
gel
(Fig.
5).
By
comparison
with
total
histones
run
on
a
parallel
gel,
the
18
kDa
and
16
kDa
bands
were
similar
to,
although
not
identical
with,
histones
H3,
H2B
and/or
H2A,
but
no
definite
bands
corresponding
to
HI
or
H4
were
observed.
However,
when
Western-blot
analysis
(Fig.
6)
was
used,
binding
of
monoclonal
anti-histone
antibodies
specific
for
H2A,
H4
and
H2B
was
demonstrated.
Unlike
the
position
of
the
bands
in
the
Coomassie
Blue-stained
gels
(Fig.
5),
the
monoclonal
antibodies
marked
positions
on
the
blot
which
corresponded
precisely
to
the
electro-
phoretic
position
of
authentic
histones.
No
other
bands
were
detected,
including
bands
corresponding
to
the
positions
of
HI
and
H3.
These
results
indicate
that
at
least
some
histones
were
present
in
bovine
milk
in
apparently
undegraded
form.
In
addition,
other
unidenti-
fied
non-histone
DNA-binding
proteins
were
present.
1.01-
0
230
250
270 290
310
33
Wavelength
(nm)
Fig.
4.
Absorption
spectrwn
of
DNA-binding
pi
milk:
comparison
with
those
of
authentic
albumin
Evidence
for
the
co-presence
of
DNA
in
milk
Albumin
Histone
in
milk
was
readily
detected
by
exploiting
its
capacity
to
bind
to
DNA
in
solid-phase
assay
(Fig.
2).
The
reciprocal
assay,
detection
of
DNA
by
virtue
of
its
capacity
to
interact
with
solid-phase
histone,
was
also
Total
histone
demonstrable
by
using
a
monoclonal
anti-DNA
antibody
(results
not
shown).
When
whey
from
milk
was
applied
Purified
to
solid-phase
histone
and
probed
with
anti-DNA
material
antibody,
a
strong
signal
was
obtained
(Table
2).
The
from
milk
presence
of
DNA
in
whey
was
confirmed
by
elimination
of
anti-DNA
antibody
binding
by
pre-digestion
with
30
DNAase.
Interestingly,
when
a
replicate
sample
was
probed
with
a
monoclonal
anti-histone
antibody,
roteins
from
enhanced
antibody
binding
was
also
detected,
and
this
histones
and
was
decreased
3-fold
by
DNAase
pretreatment.
These
results
suggest
that
DNA
is
present
in
whey
and
that
at
1987
678

Histones
in
milk
and
serum
a
b
c
d
e
f
Hi1
H3
H2B
H4-
"
Fig.
6.
Western
blot
of
isolated
proteins
from
milk
DNA-binding
proteins
were
separated
by
SDS/poly-
acrylamide-gel
electrophoresis,
electrophoretically
trans-
ferred
to
nitrocellulose
and
allowed
to
react
with
the
following
antibodies:
a,
mouse
monoclonal
anti-
H2A/anti-H4;
b,
mouse
monoclonal
anti-(H2A/H2B);
c,
mouse
monoclonal
anti-H2B;
d,
normal
mouse
serum;
e,
human
polyclonal
anti-(H2A/H2B);
f,
normal
human
serum.
The
position
of
the
bands
on
these
blots
corresponded
to
the
positions
of
authentic
histones
(not
shown
in
this
immunoblot).
least
some
of
this
DNA
exists
as
a
complex
with
histones.
Total
histone
did
not
bind
anti-DNA
antibody
or
bind
anti-histone
antibody
in
a
DNAase-sensitive
manner.
DNA-binding
material
from
milk
behaved
in
a
similar
way,
indicating
absence
of
DNA
from
the
purified
material.
Isolation
of
histone-like
materials
from
FBS
Preliminary
studies
using
Western-blot
analysis
sug-
gested
that
FBS
may
also
contain
histone-like
material
(Waga
et
al.,
1986),
although
FBS
displayed
only
marginal
activity
in
the
histone-capture
assay.
Initial
attempts
to
isolate
histone
from
FBS
by
DNA-cellulose
chromatography
were
largely
unsuccessful,
although
a
large
number
of
non-histone
bands
were
detected
by
SDS/polyacrylamide-gel
electrophoresis.
However,
when
the
material
which
did
not
bind
to
DNA-cellulose
was
re-applied
to
a
second
DNA-cellulose
column,
sub-
stantially
greater
yields
of
histone
activity
were
obtained.
Typical
results
of
this
sequential
DNA-cellulose
chromatography
are
shown
in
Fig.
7.
The
flow-through
material
from
column
1
(Fl)
when
applied
to
a
second
column
displayed
high
histone
content
in
the
HCI
eluate.
In
addition,
residual
histone
activity
in
the
column-2
flow-through
material
(F2)
was
detected
and
could
be
recovered
on
a
third,
and
subsequently
a
fourth,
DNA-cellulose
column.
The
total
amounts
of
protein
and
histone
activity
recovered
from
the
four
columns
were
1.4
mg
and
38
,ug
respectively.
When
subjected
to
SDS/polyacrylamide-gel
electro-
phoresis,
13-14
major
bands
were
detectable,
with
another
7-8
minor
bands
ranging
in
apparent
molecular
mass
from
under
10
kDa
to
over
100
kDa
(Fig.
8,
lane
b).
Multiple
bands
in
the
region
of
the
gel
corresponding
to
H3,
H2B
and
H2A
(lane
a)
were
discernible
in
the
material
isolated
from
both
FBS
(lane
b)
and
milk
(lane
c).
However,
most
of
the
proteins
comprising
the
FBS
isolate
were
substantially
different
from
those
in
the
milk
isolate.
Despite
the
lack
of
a
clear
histone
profile
in
the
FBS
isolate
by
Coomassie
Blue
staining,
Western-blot
analysis
showed
definite
reaction
of
discrete
bands
with
monoclonal
anti-histone
antibodies.
The
monoclonal
antibody
reactive
with
H4
and
H2A
stained
bands
weakly
at
the
positions
corresponding
to
these
histones
(Fig.
9,
lane
la).
However,
the
monoclonal
antibody
to
H2B
was
non-reactive
(lane
lc).
Monoclonal
antibody
'b'
showed
weak
reaction
with
H2A
as
well
as
strong
staining
on
a
band
migrating
slightly
slower
than
H4
(lane
lb).
Evidence
that
this
band
was
a
degradation
product
of
H2A
and/or
H2B
was
suggested
by
the
observation
that
calf
thymus
chromatin,
partially
degraded
by
endogenous
proteinases
(Watson
&
Moud-
rianakis,
1982)
also
displayed
a
band
at
this
position
which
bound
the
monoclonal
antibody
(lane
2b).
No
Table
2.
DNA-histone
complexes
in
milk:
effect
of
DNAase
pre-treatment
on
histone-binding
capacity
A
portion
of
the
material
to
be
assayed
was
treated
with
DNAase
as
described
in
the
Materials
and
methods
section.
A
(200
,zl)
portion
of
this
or
the
untreated
preparation
was
incubated
in
histone-coated
wells.
Either
monoclonal
anti-histone
antibody
'b'
or
anti-DNA
antibody
was
subsequently
added
and
binding
detected
by
e.l.i.s.a.
The
absorbance
data
shown
were
obtained
after
subtracting
the
background
binding
of
the
monoclonal
anti-DNA
antibody
(0.02
A)
or
the
anti-histone
antibody
(0.18
A).
Monoclonal
antibody
binding
(A)
Monoclonal
Material
[Protein]
detecting
Before
After
DNAase
applied
(mg/ml)
antibody
digestion
digestion
Whey
from
milk
DNA-binding
material
from
milk
Total
histone
19
Anti-DNA
Anti-histone
0.027
Anti-DNA
Anti-histone
0.002
Anti-DNA
Anti-histone
Vol.
244
1.85
0.39
0.01
0.26
0.02
0.11
0.01
0.13
0.01
0.18
0.03
0.18
679

S.
Waga,
E.
M.
Tan
and
R.
L.
Rubin
I
2_
2.0\-
1.5
-1.5
V
.
1.0
0.5
0.5
SF1-
LO
8
r-
CDo
F2-
ov
8n
o'-
F3-
Co
U8
o
8_
F4
Xn
u)
r-
ax)
_s
Fraction
no.
Fig.
7.
DNA-cellulose
affinity
chromatography
of
proteins
from
FBS
A
1-litre
portion
of
FBS
was
applied
to
a
DNA-cellulose
column
(a)
and
elution
was
performed
as
described
in
the
legend
to
Fig.
3.
The
unbound
flow-through
protein
from
this
column
(Fl)
was
applied
to
a
second
column
(b).
Flow-through
from
this
column
(F2)
was
applied
to
a
third
(c),
and
sequentially
a
fourth
(d),
column.
Fractions
obtained
after
washing
each
column
with
PBS
and
eluting
with
HCI
were
analysed
for
total
protein
(0)
and
histone
activity
([]).
1
2
a
bcd
a
bcd
H3
H2B
-
H2A"-
:z--
H4-
'
j;
:s
..
Fig.
8.
SDS/polyacrylamide-gel
electrophoresis
of
FBS
DNA-
binding
proteins
Coomassie
Blue-stained
bands
in
the
region
of
the
gel
near
H3,
H2B
and
H2A
of
authentic
histones
(lane
a)
were
discernible
in
the
material
isolated
from
both
FBS
(lane
b)
and
milk
(lane
c).
Marker
proteins
with
molecular
masses
in
kDa
are
depicted.
binding
of
the
monoclonal
anti-H2B
to
FBS
proteins
was
detected
(lane
lc).
It
appears,
therefore,
that
only
histones
H2A
and
H4
were
detected
in
the
isolate
from
FBS,
and
H2A
was
partially
proteolysed.
DISCUSSION
Histones
were
isolated
from
bovine
milk
and
serum
by
DNA-cellulose
affinity
chromatography,
separated
by
Fig.
9.
Western
blot
of
DNA-binding
proteins
isolated
from
FBS
The
monoclonal
antibody
probes
used
in
lanes
a-d
are
described
in
the
legend
to
Fig.
6.
Proteins
from
milk
separated
by
SDS/polyacrylamide-gel
electrophoresis
and
detected
with
the
set
of
mouse
monoclonal
anti-histone
antibodies
(panel
1)
can
be
compared
with
histone
extracted
from
chromatin
and
processed
in
parallel
lanes
(panel
2).
SDS/polyacrylamide-gel
electrophoresis
and
identified
with
a
set
of
monoclonal
anti-histone
antibodies.
Over
200
macromolecules
have been
detected
in
human
or
bovine
milk
(Blanc,
1981;
Anderson
et
al.,
1982),
but
histones
have
not
been
reported.
Estimates
from
the
immunoassay
for
soluble
histones
indicated
that
the
recovered
histone
represented
only
0.0003
%
of
the
total
1987
a
b
c
Molecular
mass
(kDa)
-92.5
-66.2
-
45.0
-31.0
Hl
c
H3\
H2B
-
H2A'-
H4-
-21.5
-14.4
'Hl
/H3
-
H2B
-FH
2A
-
H4
680

Histones
in
milk
and
serum
681
protein
in
milk
and
10-fold
less
of
the
serum
protein
content,
explaining
its
lack
of
prior
detection
and
the
necessity
to
selectively
concentrate
histones
from
these
protein-rich
complex
biological
fluids.
DNA
was
readily
detected
in
unfractionated
milk
by
the
immunoassay
(Table
2),
consistent
with
the
findings
of
Sanguansermsri
et
al.
(1974)
and
Jarasch
et
al.
(1977).
DNAase-digestion
experiments
indicated
that
at
least
a
portion
of
milk
histone
occurred
as
a
complex
with
DNA.
It
is
possible
that
milk
histone
originated
from
chromatin
released
after
mammary-epithelial-cell
death
or
from
leucocytes
known
to
be
present
in
post-colostrum
milk
(Brooker,
1978;
Concha
et
al.,
1978;
Blanc,
1981).
Since
we
also
detected
histones
in
bovine
serum,
it
is
also
possible
that
this
material
was
transferred
directly
to
milk
in
the
mammary
gland,
as
are
other
blood-derived
macromolecules
(Bruder
et
al.,
1984).
The
inability
to
detect
HI
may
be
due
to
the
low
reactivity
of
this
histone
with
the
anti-histone
probes
(Fig.
1)
or
the
absence
of
HI
in
the
isolate
(due
to
its
well-known
sensitivity
to
proteolytic
degradation).
Proteinases
are
known
to
be
present
in
milk
(Shahani
et
al.,
1980)
and
serum
(Erdos
et
al.,
1965),
and,
in
fact,
H2A
and
H2B
from
serum
displayed
the
altered
electrophoretic
mobility
charac-
teristic
of
these
histones
in
chromatin
after
endogenous
proteolysis.
H3,
one
of
the
core
histones
of
the
nucleosome
(McGhee
&
Felsenfeld,
1980),
may
be
present,
but
undetected
because
of
a
lack
of
the
appropriate
antibody
probe.
Isolation
of
histones
from
serum
required
serial
DNA-cellulose
chromatography.
Presumably
removal
of
the
bulk of
the
serum
DNA-binding
proteins
was
required
before
the
minor
histone
components
could
favourably
compete
for
binding
to
available
sites
on
immobilized
DNA.
These
observations
are
similar
to
those
of
Hoch
et
al.
(1975)
and
Lewis
&
Antre
(1978),
who
showed
that
serum
albumin
and
IgG
interfere
with
DNA-cellulose
chromatography
as
a
result
of
charge-
dependent
binding
to
the
DNA.
Non-fat
dry
milk
(Johnson
et
al.,
1984)
and
fetal
bovine
serum
(Thomas
et
al.,
1984;
Gohill
et
al.,
1985)
have
come
under
increasing
use
as
blocking
agents
to
minimize
non-specific
binding
of
proteins
and
nucleic
acids
to
solid
supports
such
as
nitrocellulose.
Spinola
&
Cannon
(1985)
compared
the
effect
of
several
blocking
reagents,
including
milk,
on
the
identification
of
bacterial
proteins
using
Western-blot
technology
and
found
different
results
depending
on
which
blocking
agent
was
used.
Miskimins
et
al.
(1985)
recommended
milk
treatment
to
eliminate
background
problems
with
immunoblots,
but
also
inexplicably
observed
enhanced
reactivity
of
the
DNA
probe
to
histone
HI
in
this
system.
Blocking
media
effects
may
be
due
to
contaminating
nucleohistones,
which
may
mediate
the
binding
of
the
probe
to
reactive
components
on
the
nitrocellulose.
Detection
of
anti-histone
and
anti-DNA
antibodies
by
Western-blot
techniques
is
particularly
prone
to
en-
hanced
binding
artefacts
when
milk
or
fetal-bovine
serum
are
used
to
the
blocking
reagent
(S.
Waga,
E.
M.
Tan
&
R.
L.
Rubin,
unpublished
work).
Whether
circulating
histones
originate
from
cell
death
and
subsequent
lysis
or
as
a
result
of
an
active
process
of
secretion
(Rogers
&
Kawahawa,
1981;
Atkinson
et
al.,
1985)
may
not
be
relevant
to
their
immunological
signifi-
cance.
Histones
are
a
common
target
of
autoantibodies
in
patients
with
drug-induced
lupus
(Fritzler
&
Tan,
1978)
and
systemic
lupus
erythematosus
(Rubin,
1986a,b).
The
existence
of
histones
accessible
to
the
immune
system
may
have
consequences
with
respect
to
the
origin
of
the
anti-histone
antibodies
and
their
pathogenic
potential
to
participate
in
immune-complex
disease.
This
is
publication
4430BCR
from
the
Research
Institute
of
Scripps
Clinic.
S.W.
was
the
recipient
of
a
fellowship
funded
by
Mr.
Ho
Tim
of
Hong
Kong.
This
research
was
supported
in
part
by
the
W.
M.
Keck
Foundation
and
National
Institutes
of
Health
Grants
AM32063,
Al
10386
and
AM34358.
We
thank
Dr.
Argyrios
Theofilopoulos
for
production
of
some
of
the
monoclonal
antibodies
used
in
this
study.
REFERENCES
Anderson,
N.
G.,
Powers,
M.
T.
&
Tollaksen,
S.
L.
(1982)
Clin.
Chem.
28,
1045-1055
Atkinson,
M.
J.,
Bell,
D.
A.
&
Singhal,
S.
K.
(1985)
J.
Immunol.
135,
2524-2533
Bennett,
R.
M.,
Gabor,
G.
T.
&
Merritt,
M.
M.
(1985)
J.
Clin.
Invest.
76,
2182-2190
Biogioni,
S.,
Sisto,
R.,
Ferraro,
A.,
Caiafer,
P.
&
Turano,
C.
(1978)
Anal.
Biochem.
89,
616-619
Blanc,
B.
(1981)
World
Rev.
Nutr.
Diet.
36,
1-89
Brooker,
B.
E.
(1978)
J.
Dairy
Res.
45,
21-24
Bruder,
G.,
Jarasch,
E.-D.
&
Heid,
H.
W.
(1984)
J.
Clin.
Invest.
74,
783-794
Concha,
C.,
Holmberg,
0.
&
Morcin,
B.
(1978)
J.
Dairy
Res.
45,
287-290
Emlen,
W.
&
Mannik,
M.
(1982)
J.
Exp.
Med.
155,
1210-1215
Erdos,
E.
G.,
Wohler,
I.
M.,
Levine,
M.
I.
&
Westerman,
M.
P.
(1965)
Clin.
Chim.
Acta
11,
39-43
Fritzler,
M.
J.
&
Tan,
E.
M.
(1978)
J.
Clin.
Invest.
62,
560-567
Gohill,'
J.,
Cary,
P.
D.,
Couppez,
M.
&
Fritzler,
M.
J.
(1985)
J.
Immunol.
135,
3116-3121
Hoch,
S.
O.,
Longmire,
R.
L.
&
Hoch,
J.
A.
(1975)
Nature
(London)
255,
560-562
Holers,
V.
M.
&
Kotzin,
B.
L.
(1985)
J.
Clin.
Invest.
76,
991-998
Homeland,
M.,
Rekvig,
0.
P.,
Jorgensen,
L.
&
Hannestad,
K.
(1983)
Clin.
Exp.
Immunol.
54,
373-377
Jacob,
L.,
Tron,
F.,
Bach,
J.-F.
&
Louvard,
D.
(1984)
Proc.
Natl.
Acad.
Sci.
U.S.A.
81,
3843-3845
Jarasch,
E.-D.,
Bruder,
G.,
Keenan,
T.
&
Franke,
W.
W.
(1977)
J.
Cell
Biol.
73,
223-241
Johnson,
D.
A.,
Gautsch,
J.
W.,
Sportsman,
J.
R.
&
Elder,
J.
H.
(1984)
Gene
Anal.
Tech.
1,
3-8
Kotzin,
B.
L.,
Lafferty,
J.
A.,
Portanova,
J.
P.,
Rubin,
R.
L.
&
Tan,
E.
M.
(1984)
J.
Immunol.
133,
2554-2559
Laemmli,
U.
K.
(1970)
Nature
(London)
227,
680-685
Lewis,
J.
G.
&
Antre,
C.
M.
(1978)
FEBS
Lett.
92,
211-213
Lonnerdal,
B.,
Keen,
C.
L.
&
Hurley,
L.
S.
(1985)
Am.
J.
Clin.
Nutr.
41,
550-559
Lowry,
0.
H.,
Rosebrough,
N.
J.,
Farr,
A.
L.
&
Randall,
R.
J.
(1951)
J.
Biol.
Chem.
193,
265-275
McConahey,
P.
J.
&
Dixon,
F.
J.
(1980)
Methods
Enzymol.
70,
210-213
McCoubrey-Hoyer,
A.,
Oharma,
T.
B.
&
Holman,
H.
R.
(1984)
Am.
J.
Med.
77,
23-24
McGhee,
J.
D.
&
Felsenfeld,
G.
(1980)
Annu.
Rev.
Biochem.
39,
1115-1156
Miskimins,
W.
K.,
Roberts,
M.
P.,
McClelland,
A.
&
Ruddle,
F.
H.
(1985)
Proc.
Natl.
Acad.
Sci.
U.S.A.
82,
6741-6744
Portanova,
J.
P.,
Arndt,
R.
A.,
Tan,
E.
M.
&
Kotzin,
B.
L.
(1986)
Arthritis
Rheum.
29,
S27
Raptis,
L.
&
Menard,
H.
A.
(1980)
J.
Clin.
Invest.
66,
1391-1399
Rekvig,
0.
P.
&
Hannestad,
K.
(1980)
J.
Exp.
Med.
152,
1720-1733
Rogers,
J.
C.
&
Kawahara,
R.
5.
(1981)
Exp.
Cell
Res.
134,
1-13
Vol.
244

682
S.
Waga,
E.
M.
Tan
and
R.
L.
Rubin
Rogers,
J.
C.,
Boldt,
-D.,
Kornfeld,
S.,
Skinner,
A.
&
Valeri,
C.
R.
(1972)
Proc.
Natl.
Acad.
Sci.
U.S.A.
69,
1685-1689
Rubin,
R.
L.,
Joslin,
F.
G.
&
Tan,
E.
M.
(1983)
J.
Immunol.
Methods
63,
359-366
Rubin,
R.
L.,
McNally,
E.
M.,
Nusinow,
S.
R.,
Robinson,
C.
A.
&Tan,
E.
M.
(1985)
Clin.
Immunol.
Immunopathol.
36,
49-59
Rubin,
R.
L.
(1986a)
in
Systemic
Lupus
Erythematosus
(Lahita,
R.
G.,
ed.),
pp.
271-289,
John
Wiley
and
Sons,
New
York
Rubin,
R.
L.
(1986b)
in
Manual
of
Clinical
Immunology,
3rd
edn.
(Rose,
N.
R.,
Friedman,
H.
&
Fahey,
J.
L.,
eds.),
pp.
744-749,
American
Society
for
Microbiology,
Washington,
DC
Sanguansermsri,
J.,
Gyorgy,
P.
&
Zilliken,
F.
(1974)
Am.
J.
Clin.
Nutr.
27,
859-865
Shahani,
K.
M.,
Kwan,
A.
J.
&
Friend,
B.
A.
(1980)
Am.
J.
Clin.
Nutr.
33,
1861-1868
Shubert,
D.
&
LaCorbiere,
M.
(1985)
J.
Cell
Biol.
101,
1071-1077
Spinola,
S.
M.
&
Cannon,
J.
G.
(1985)
J.
Immunol.
Methods
81,
161-165
Tan,
E.
M.,
Schur,
P.
H.,
Carr,
R.
I.
&
Kunkel,
H.
G.
(1966)
J.
Clin.
Invest.
45,
1732-1740
Thomas,
J.
O.,
Wilson,
C.
M.
&
Hardin,
J.
A.
(1984)
FEBS
Lett.
169,
90-96
Towbin,
H.
T.,
Staehelin,
T.
&
Gordon,
J.
(1979)
Proc.
Natl.
Acad.
Sci.
U.S.A.
76,
4350-4354
Waga,
S.,
Tan,
E.
M.
&
Rubin,
R.
L.
(1986)
Arthritis
Rheum.
29,
S72
Watson,
D.
K.
&
Moudrianakis,
E.
N.
(1982)
Biochemistry
21,
248-256
Received
11
November
1986/10
February
1987;
accepted
27
February
1987
1987