Vol.
174,
No.
January
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
31, 1991
Pages
THE ISOLATED
CYTOPLASHIC
AT PHYSIOLOGICAL
James
DOMAIN OF BAND 3 BINDS CALCIUM
SALT CONCENTRATION AND NEUTRAL pH
M. Salhany
Department
975-982
and
Karen
A. Cordes
Affairs Medical Center and Departments of Internal Medicine and Biochemistry University of Nebraska Medical Center, Omaha, Nebraska, 68105 Received
November
26,
of Veterans
1990
Calcium is known to be a potent but partial intracellular inhibitor of band 3 anion exchange. Here we test the hypothesis that the cytoplasmic domain of band 3 ( CDB3 ) contains a calcium binding site. Calcium binding to CDB3 was monitored by measuring the formation of the Arsenazo III-calcium complex at various constant CDB3 concentrations. These experiments were performed at physiological salt and neutral pH. The calcium-CDB3 dissociation constant was estimated to be < 24 PM. We also found that the Arsenazo III-calcium complex Finds to CDB3, while the free dye does not bind. We conclude that CDB3 contains a site which is capable of binding free calcium under physiological conditions. A specific role for this site in inhibition of band 3 anion exchange is suggested, but that role remains to be established. 0 1991Academic PlxS5, Inc.
Human
erythrocyte
domains,
each
monomer
( 1 ).
exchange (
CDB3
kinases porter
comprising The
ankyrin,
enzymes ( 2 ) and ( 2-5
essential
for
). the
Abbreviations:
tetraacetate; (hydroxymethyl)methane.
the
is
the
Although anion
is
the
of
and
heavily
mass for
domain certain binds
activity
of
the of
of
the anion
band
the
red
3
cell
tryrosine-specific
phosphorylated
cytoplasmic
exchange
distinct
the
responsible
CDB3 also
most
functionally
cytoplasmic
hemoglobin
( 1 ).
two
one-half
domain
while
) binds
3 has
about
integral
activity,
glycolytic
band
domain ( 6,
EGTA:ethyleneglycol-bis-(B-aminoethylether)N,N~bistris: N,N-bis(2-hydroxyethyl)iminotris-
site is 7 ),
on the
clearly there
not is,
Vol.
174,
No.
2, 1991
Elucidation
of
identification modulate
but
mechanism
of
ligands anion
( 1 ).
(
of
< 20 p
has
recently
( 10 ) showed A"
of
that
alters In
the
with
erythrocyte
binding
to
( 14 ) of
Amino the
certain
X
ion
band
loop
most
acids
in
calcium
of the binding
z
glu
leu
gln
asp
asp
2
3
4
5
6
7
this
paper,
we test
the
at physiological
hypothesis
MATERIALS
calcium
CDB3 ( 12 ).
acid
location
sequence
of
the
acidic
N-
for
a calcium
a comparable
charge
structural
motif
"EF-hand" ( 1 ):
glu 8
with
of
to
amino
14 have
of
microcalorimetric
-X
tyr
salt
reactivity reaction
identified
-Y
glu
the
addition,
addition
proteins
and effect
In
assigned the
proteases
1.
is
that
and
2 through
intracellular
its
there
probable
region
Y
tightly
domain
),
at
9
alters
we searched
3 ( 13
CDB3 as the
site.
homology
In
( 1 ),
a potent
inhibitory
calcium
showing
cytoplasmic is
of
(
( 11 ) specifically
review
of
confirmed
ghosts,
and
CDB3
calcium
cytosolic
Finally,
unsealed
a recent
terminus
activation
integral
an endotherm
human
where
trans
to
is
that
exchange
been
the
porter.
potential
This
1-fluoro-2,4-dinitrobenzene. evidence
this
( 1 ).
requires
bind
anion )
domains
of the
showed
was minimal.
calcium
activity
has
COMMUNICATIONS
communication
specifically
exchange which
RESEARCH
between
such
Low ( 8 ) first
transglutaminase
"lysine
for
BIOPHYSICAL
which
inhibitor
concentrations
Passow
communication
effector
partial
of
for
the
the
One calcium
AND
evidence
nevertheless,
also
BIOCHEMICAL
-2
asp
met
met
10
11
12
9
that
concentration
CDB3 can bind and neutral
glu
glu
13
14
calcium pH.
AND METHODS
CDB3 was isolated in pure form as described in a previous CDB3 was found to contain After isolation, report ( 15 ). Our methods for stripping the significant amounts of calcium. protein of calcium form part of the results described below. Experiments were conducted in Buffer A [ 150 mM KCl, 20 mM bistris, 1 PM EGTA, pH 7.0 ] at 25OC. EGTA was included in order to complex traces of calcium ( < 1 PM ) determined with Arsenazo
976
Vol.
174,
No.
BIOCHEMICAL
2, 1991
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
In Buffer A containing III ( 16 ) in the 20 mM bis-tris buffer. 1 c"M EGTA, no Arsenazo III-calcium complex was detected. Calcium binding to CDB3 was monitored by measuring calcium titration of Arsenazo III ( 16 ) at various constant, concentrations. The apparent dissociation increasing CDB3 complex was computed by fitting the constant for the dye-calcium data to a hyperbolic function of the form:
(1)
AA
=(AAmax*f
Ca+’
using the Enzfitter spectra were collected spectrophotometer.
)I/(
Kapp.
curve fitting program using a Hitachi
(
+
17
Model
(Ca+‘)) Optical recording
).
100-60
RESULTS AND DISCUSSION Stripping
CDB3
CDS3 of
by
dialyzing
extensive
dialyzed
the
dialysis
concentration Buffer
calcium.
A
( only
We attempted
protein
against
against see
legend
contains
CDB3 contained
Buffer to
Fig.
about about
ARSENAL0
to
remove
1 mM EGTA, A
to
1 for
details
PM
calcium
using
from
followed
lower
1 pM EGTA-calcium 70
calcium
the
by EGTA
).
While
( Methods the
spectra
III
Figure
1. Absorbance spectrum of Arsenazo III in the absence and wresence of the cvtowlasmic domain of band 3 I CDB3 ). The concentration of ghe' dye was 20 pM for both' spectra. One spectrum is for the calcium-free dye in Buffer A. The second spectrum is for the dye in Buffer A plus dialyzed CDB3 ( 69 pM in monomer ). Before addition of the dye, the protein was dialyzed overnight in 1 mM EGTA, 5 mM bistris, pIi 7. The sample was volume reduced to 1 ml of 180 pM in CDB3 monomer and it was dialyzed against 250 mls of Buffer A. The buffer was changed 3 additional times ( 250 mls each ) to dilute EGTA. Control experiments with an empty dialysis bag showed a dye spectrum which was indistinguishable from that of the buffer. L
977
),
Vol.
174,
No.
BIOCHEMICAL
2, 1991
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
0.6 ARSENAZO
Ill
I
0
200
400 [Ca+2].vM
I
600
800
Figure 2. Absorbance change of Arsenazo III at 600 nm as a function of increasing calcium concentration. Arsenazo III ( 20 uM ) was titrated with calcium in the absence and presence of 60 pM CDB3 ( monomer ), all in Buffer A ( see text ). The lines drawn through the data are fits to a hyperbolic function ( see the Materials and Methods Section ). The values which gave the best fit were: ( No CDB3 ), ( Plus
in
KI = 104 + 7 PM;
CDB3 ),
Figure
Kapp.
based
1,
equation
equivalent
to
indicating
that to
calcium then
the
it
was exposed
we added
CDB3,
removed
the
excess
during
buffer
Calcium
alone binding
column-stripped
approach physiological
to
( Fig. to
determine conditions.
of is
monomer
from
to
the
the
about
present,
the
problem
1 mM EGTA directly
EGTA on a G-25
1
liter
of
to
column such
of the
stripping protein
and
equilibrated
with
column-stripped from that of
CDB3 the dye
1 ).
Arsenazo
CDB3.
validity
dialysis.
Buffer A. Addition of Arsenazo III to gave a spectrum which was indistinguishable in
CDB3
calcium
approach
the
+ 0.002
concentration of
concentration
a successful
from
This
1.
18
assuming
CDB3 accumulates
which
In
(
+ 0.006
A A max = 0.26
= 146 + 6 PM;
on an estimate
Rose-Drag0
buffer
Amax = 0.38
A
III
in
We used if
the
this 978
and
a competitive
calcium With
absence
binds method,
presence
of
spectroscopic to the
CDB3 dye
is
under fully
Vol.
174,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
, CALCIUM 160 -
BINDING
TO
ARSENAZO
Ill
3. Plot of measured K values as a function of CDB3 concentration. Experiments a%ke those in Figure 2, were performed at several protein values determined by curve hyperbolically with increasing CDB3 drawn through the data is based on a weighted fit to a simple increasing hyperbolic function whose saturating K ( Equation 3 of the text ) was determined to be 158 + %?'pM~!k~~ insert shows the hyperbolic decrease in P A with increasing CDB3. The curve drawn through this data is %%ed on a weighted fit to a decreasing hyperbolic function with K = 36 + 3 PM. Figure
titrated
with
If
protein
the
calcium
dissociation
constant
were value
result
of
the
in
a
that
that
CDB3 must
dye.
We have
already
(
1
calcium perturbation
also
Fig. upon
the
complex
( Kapp.
in
is
of
significantly the
mixing
Furthermore, dye
and
the
complex
) should
of
free
calcium shown
the protein
lack only,
with
CDB3.
Two
decrease
in
protein.
This
binds
to
increased,
qualitatively ).
apparent
absorbance
a substantial
presence
either
dye
presence
dye-calcium
Kapp . bind
calcium,
and
first
the
the
of
change
absence
The
concentrations.
CDB3 concentration.
the
in
CDB3
dye-calcium
the
Gax
value
free
increasing
observed.
suggests
Second,
the
shows
calcium
effects the
2,
constant
binds
of
with
Figure increasing
various
simply
linearly
increase
at
ion
indicating or
that of argues
CDB3.
the
CDB3 a
free binds
spectral strongly
Vol.
174,
No.
2, 1991
BIOCHEMICAL
ARSENAZO
o! 400
,
,
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
III
,
(
t
5
, 600
500
I
I
r
I
700
NANOMETERS
Figure 4. Absorbance spectrum of Arsenazo III ( 18 PM presence of column-purified CDB3 under various conditions. spectra were collected in Buffer A. In the absence of addition of dye to CDB3 did not cause a change in spectrum. At saturating calcium ( 4.5 mM ), increasing protein concentration caused a change in the spectrum of calcium complex . See text for discussion.
free
that
dialysis EGTA
showed Thus,
In
evidence calcium
Figure
3,
the
calcium
complex
of
CDB3-(dye-calcium)
Binding
for
the
binding
binding
to
CDB3 is
binding
Finally, buffer of
the All
calcium, the dye the the dye-
equilibrium containing
free
1 mM
dye
sufficient
( Fig. the
that the of the model
increasing
the in
Kam - increases the increase
of
both
complex
systematic
increasing spectrum
to CDB3 in
by CDB3 ( see below
concentration confirm
CDB3.
However,
with
observed
to
we show that
concentration.
the
bind
in
to
to
the
explain
in Kapp . .
consistent
behavior
not
of dye binding
no
increase
CDB3
does
studies
protein. the
dye
)
decrease
in
3,
).
insert
dye-calcium
protein
is
spectra
dye-
the
formation by
increasing shown
binds
to
in
the CDB3
Figure
CDB3,
4,
since
alters
the
complex. Figure
CDB3 leads of 980
the
indicated
analysis.
formation
and
strongly with
with
hyperbolic,
for
qualitatively
dye-calcium
of
calcium
b Amax The
is
Evidence
complex
and quantitative
Kapp . for
).
concentration
saturated
concentration
free
uniformly
the
to
3 indicates apparent
Arsenazo
that
saturation III-calcium
Vol.
174,
No.
2, 1991
BIOCHEMICAL
Such
complex. assume
that
affinity
behavior
free
than
does
for
dissociated
state,
thus
Kapp.
would
in
competes
with
representing
these
of
then
to
dye.
Such
the
dye-calcium
for
competitive
when the
I
+
L = calcium; dye-calcium
complex;
the
association
complexes
complex
with
CDB3.
( 2 1
Kapp.
=
At saturating
value
Figure for
K
to
Concluding
to
isolated
toward
the
Kapp..
Saturation
dye-calcium
reactions
is
as follows:
complex
protein.
A model
3 to
am.
11 K12
\
IL
I = Arsenazo
with
S(L)
and S(IL)
of
free
calcium
following
III;
being and
= the
respectively, dye-calcium
was derived:
1 / ( 1 + ( S )/K12
1 1
we have: =
KI
* [
K12 / Kll
1
evaluated
a simple
by
hyperbolic
fitting
function
based
saturating
on a fit
calcium on
these
of the
and
increasing
values,
we
data
the to
( 158 PM, Fig. 3 ). KI -sat. in Figure 2 ( 104 PM ). An upper
Based
IL
the
relationship
* [ ( 1 + ( S )/Kll
be numerically
K12 can be found
d %lax at Fig. 3 ). Kll
( S ),
3 can
independently for
KI
The
Kapp .-sat.
in
complex
the
S = CDB3 monomer;
free
data
the
to
(S)
L
Equation
higher
shift
binding
K1l 11
1
with
we
S(IL)
( s 1
(3
if
would
in
S(L)
Where
CDB3
binding
an increase result
COMMUNICATIONS
understood
binds
causing
calcium
RESEARCH
be qualitatively
the
formation
free
BIOPHYSICAL
initially
to
equilibrium
behavior
can
calcium
it
AND
for CDB3
estimate
primary
determine
was
a
determined
limiting the
value change
( K = 36 PM, the
value
be 5 24 PM.
Remarks.
CDB3 at
Our
results
show that
physiological
salt 981
calcium
concentration
in
binds and
tightly neutral
of
Vol.
174,
the to
2, 1991
Whether
PH. 3,
No.
and
CDB3 contains
whether
observed
BIOCHEMICAL
binding
inhibition
AND
BIOPHYSICAL
the
only
calcium
to
of
anion
calcium this
exchange
RESEARCH
COMMUNICATIONS
binding
site
site ( 8,
is
responsible 9 ),
both
on band for remain
be established.
ACKNOWLEDGMENTS. We thank Dr. Larry Schopfer for helpful discussions and for critically reading the manuscript. We also thank John Friel for drawing the figures. This work was supported by the Department of Veterans Affairs Medical Research Funds. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
Band 3 Protein" CRC Press, Salhany, J.M. ( 1990 ) "Erythrocyte Boca Raton Florida. Habib-Mohamed, A. & Steck, T.L. ( 1986 ) J. Biol. Chem. 261: 2804-2809. Dekowski, S.A., Rybicki, A. & Drickamer, R. ( 1983 ) J. Biol. Chem. 258: 2750-2753. ( 1987 ) Biochim. Vasseur, C., Piau, J.P. & Bursaux, E. Acta 899: l-8. Biophys. Chavi, P., Low, P.S., Allen, D.P., Zioncheck, T.F., Williardson, B.M., Geahlem, R.L. & Harrison, M.L. ( 1987 ) J. Biol. Chem. 262: 4592-4596. Grinstein, S., Ship, S. & Rothstein, A. ( 1978 ) Biochim. Biophys. Acta 507: 294-304. Lindsey, A.E., Schneider, K., Simmons, D.M., Baron, R., Lee, B.S. & Kopito, R. R. ( 1990 ) PrOC. Natl. Acad. Sci. USA 87: 5278-5282. Low, P.S. ( 1978 ) Biochim. Biophys. Acta 514: 264-273. Joshi, R. & Gupta, C.M. ( 1990 ) J. Membr. Biol. 117: 233242. Passow, H. ( 1986 ) Rev. Physiol. Biochem. Pharmacol. 103: 61-203. Brandts, J.F., Lusko, K., Schwarts, A.T., Erickson, L., Carlson, R.B., Vincentelli, J. & Traverna, R.D. ( 1975 ) Colloq. Int. CNRS # 246: 169-175. Appell, K.C. & Low, P.S. ( 1982 ) Biochemistry 21: 2151-2157. Tanner, M.J.A., Martin, P.G. & High, S. ( 1989 ) Biochem. J. 256: 703-712. Bairoch, A. & Cox, J.A. ( 1990 ) FEBS Letters 269: 454-456. Salhany, J.M. & Cassoly, R. ( 1989 ) J. Biol. Chem. 264 13991404. Scarpa, A. Brinley, F.J., Tiffert, T. h Dubyak, G.R. ( 1978 ) Ann. N.Y. Acad. Sci. 307: 86-111. Leatherbarrow, R.J. ( 1987 ) Enzfitter. A nonlinear regression data analysis program for the IBM PC. Elsevier Sci. Publishing Co. Amsterdam. Rose, N.J. & Drago, R.L. ( 1959 ) J. Am. Chem. Sot. 81: 61386141.
982