|
|
|
|||||||
|
|||||||||
(Received for publication, September 21, 1994; and in revised form, January 3, 1995) From the
The nature of the membrane attachment site(s) for protein 4.1 in
the human erythrocyte membrane has yet to be fully elucidated. In this
paper we show that the major attachment site is glycophorin (GP) C/D,
and that purified protein 4.1 can bind to two distinct sites on
glycophorin C/D. One of these interactions is direct, involving
residues 82-98 on glycophorin C (61-77 on glycophorin D),
while the other interaction is mediated by p55. We have localized the
binding site for p55 on glycophorin C to residues 112-128
(glycophorin D 91-107). We also provide evidence that band 3 is
an additional, minor, protein 4.1 binding site. The binding sites for
band 3, glycophorin C/D, and p55 are all located within the 30-kDa
domain of protein 4.1. We estimate that the relative utilization of the
three sites in normal membranes comprises 40% to p55, 40% to GPC/D, and
20% to band 3. The same region of protein 4.1 binds GPC/D and band 3,
while the p55 binding site is distinct. The interactions involving
protein 4.1 with p55 and p55 with GPC/D are of high affinity
(nM), while those involving GPC/D and band 3 are 100-fold
lower (µM). These results suggest that the most
significant interactions between protein 4.1 and the membrane are those
involving p55.
The structural integrity and deformability of the human
erythrocyte, essential for its survival, are maintained by the
skeleton, a network of proteins underlying the lipid bilayer membrane.
The skeleton consists predominantly of spectrin, actin, and protein
4.1, as well as the minor components adducin, protein 4.2, protein 4.9,
tropomyosin, and myosin (reviewed in (1) and (2) ).
The membrane is connected to the skeleton by at least two protein
interactions. Ankyrin links band 3 to While there has been much circumstantial evidence for a protein
4.1-GPC/D interaction for several
years(7, 8, 9) , direct evidence has only
recently become available. This evidence suggests that GPC/D is the
major membrane attachment site for protein
4.1(13, 14) . It has also been suggested that a third
component, p55, is required for such an
interaction(15, 16) , and that there are two protein
4.1 binding sites on GPC(13, 17) . In this paper we
show that protein 4.1 can bind to GPC through two sites, one is a
direct interaction between the two proteins, the other is an indirect
interaction mediated by p55. We also provide evidence that band 3 is a
minor, low affinity, protein 4.1 binding site. Fresh normal erythrocytes were available from the National
Blood Service (Bristol, UK). Fresh Leach phenotype erythrocytes (which
lack GPC/D) were obtained from donor PL(18) . Rabbit
anti-protein 4.1 antibodies against the N- and C-terminal regions of
the protein were prepared using synthetic peptides
Cys Antibodies to band 3
(BRIC 169), (
Figure 1:
SDS-polyacrylamide gel
electrophoresis analysis of membranes and purified protein 4.1 and p55
and immunoblotting with anti-protein 4.1 and anti-p55. Samples were run
on a 4-16% polyacrylamide gradient Laemmli gel. Samples in lanes a-c were stained with Coomassie Blue. Samples in lanes d-f were probed with anti-protein 4.1 and samples
in lanes g-j with anti-p55 (as described under
``Materials and Methods''). Lanes a, e, and g, normal membranes; lanes b, f, and j, purified protein 4.1; lanes c, d, and i, purified p55; lane h, Leach phenotype
membranes.
Figure 2:
A, binding of synthetic peptides to
protein 4.1. GPC peptides were incubated with protein 4.1 (150
µg/ml) overnight at 4 °C. Binding was measured as described
under ``Materials and Methods.'' Results are plotted with error bars corresponding to mean ± S.D. (n = 4) X, GPC-1;
Figure 3:
p55
binding protein 4.1. p55 was incubated with protein 4.1 (150 µg/ml)
overnight at 4 °C. Binding was measured as described under
``Materials and Methods.'' Results are plotted with error
bars corresponding to mean ± S.D. (n = 5).
Binding capacity = 48 ± 3 µg of p55/mg of protein
4.1; K
Figure 4:
A, p55 binding to alkali-stripped normal
and Leach phenotype membranes and to trypsin/chymotrypsin (5
µg/ml)-treated alkali-stripped normal membranes. Binding was
measured as described under ``Materials and Methods.''
Results are plotted with error bars corresponding to mean
± S.D. (n = 4).
The p55 binding site on
alkali-stripped normal membranes is lost if the membranes are
pretreated with trypsin or chymotrypsin (Fig. 4A),
which cleave the C terminus of GPC(13) . When unstripped normal
membranes are trypsin-treated, only 88% ± 4% (n = 6) GPC is cleaved while 100% (n = 6) GPC
is cleaved when stripped normal membranes are digested with trypsin.
The ability of p55 to block the trypsin cleavage site on GPC was
demonstrated by incubating p55 with alkali-stripped normal membranes
prior to trypsin digestion. Under these conditions, only 43% ±
6% (n = 3) GPC was cleaved. In an attempt to
localize the p55 binding site on GPC, we investigated the ability of
GPC peptides to bind to purified p55. The results (Fig. 4B) clearly show a concentration-dependent
binding of GPC-1 to p55. The binding was saturable at 16 µg of
GPC-1/mg of p55. This compares with a theoretical value of 22 µg/mg
p55, assuming one GPC-1 binding site per molecule of p55. Some binding
of the other peptides was observed, but this was consistently much
lower than that of GPC-1. Further, GPC-1 can totally inhibit p55
binding to alkali-stripped normal membranes, while GPC-2 and GPC-3 have
no effect (data not shown). Taken together, these results indicate
that p55 binds to the extreme C-terminal region of GPC.
Figure 5:
Binding of chymotrypsin-digested protein
4.1 to p55 (A) and alkali-stripped (B) normal
membranes. Protein 4.1 was partially digested with chymotrypsin, and
the fragments were purified as described under ``Materials and
Methods.'' Results are plotted with error bars corresponding to mean ± S.D. (n = 4).
In order to determine whether the binding sites
were in the same or different regions of the 30-kDa domain, protein 4.1
was saturated with GPC-3, and its ability to bind p55 and
alkali-stripped Leach phenotype membranes (band 3 sites) was studied.
While there was no inhibition of binding to p55 (data not shown),
binding to alkali-stripped Leach phenotype membranes was completely
inhibited (Fig. 6). These results indicate that there are two
distinct membrane binding sites on protein 4.1, one binds GPC-3 and
band 3, while the other binds p55.
Figure 6:
Inhibition of protein 4.1 binding to
alkali-stripped Leach membranes. Protein 4.1 (150 µg/ml) was
incubated with GPC peptides for 5 h at 4 °C, before incubation with
Leach phenotype stripped membranes. Binding was measured as described
under ``Materials and Methods.'' Peptide concentrations are
depicted on a logarithmic scale. Results are plotted with error
bars corresponding to mean ± S.D. (n = 4).
X, GPC-1;
Protein 4.1 binding to trypsin-treated
membranes (GPC-3 sites) represents 55% of the total binding; protein
4.1 binding in the absence of p55 (band 3 + GPC-3 sites), 72% (Fig. 7; Table 1). A direct measure of the protein 4.1
binding to p55 was obtained by determining the amount of
GPC-3-saturated protein 4.1 bound to p55-saturated alkali-stripped
membranes. The expected value of 27% was obtained (Fig. 7; Table 1).
Figure 7:
Binding of protein 4.1 to alkali-stripped
normal membranes. X, purified protein 4.1 was incubated overnight at 4
°C with alkali-stripped normal membranes, binding capacity 192
± 3 µg/mg membrane protein; K
The location of protein 4.1 binding sites in the red cell
membrane has been a matter of some controversy with
GPC/D(7, 8, 9) , band
3(4, 5, 6) , glycophorin A(10) , and
the lipid bilayer itself (11, 12) all being
implicated. Under the conditions of the experiments reported in this
paper, we find no significant binding to either glycophorin A or the
lipid bilayer (all GPC independent binding is trypsin-sensitive).
However, we provide evidence that protein 4.1 binds to alkali-stripped
normal erythrocyte membranes through at least three distinct sites. Two
sites are located on GPC/D, one is a direct interaction involving
residues 82-98 of GPC, and the other an indirect interaction,
mediated by p55. The third binding site is most likely located on the
N-terminal cytoplasmic domain of the anion transport protein band 3
(syn AE-1). Under the conditions of our assay, the proportion of
protein 4.1 occupying each of these sites is approximately 55% directly
to GPC, 28% through p55, and 17% through band 3 (Table 1).
Estimation of the dissociation constants for these interactions suggest
that binding through p55 is of higher affinity (K The involvement of p55 in the
GPC-protein 4.1 interaction has recently been studied by Marfatia et al.(16) . These authors concluded that a
recombinant fusion protein containing p55 was able to bind with high
affinity to the 30-kDa domain of protein 4.1 and to GPC. They also
concluded that the 30-kDa domain of protein 4.1 was able to bind to GPC
with equally high affinity. While our data support the view of Marfatia et al.(16) that protein 4.1 can bind to both p55 and
GPC, we suggest that the GPC-protein 4.1 interaction is of a lower
affinity to the p55-protein 4.1 interaction. This conclusion is broadly
consistent with the recent finding of Gascard and Cohen (17) that GPC contains both high and low affinity binding sites
for protein 4.1. However, the proportion of high affinity sites on GPC
reported by Gascard and Cohen (17) (10% on average) is much
lower than our findings. This is likely to have arisen because Gascard
and Cohen (17) did not consider the possible role of p55, and,
thus, the high affinity sites detected are probably due to p55
contamination in their protein 4.1 preparations and/or IOVs used for
reassociation assays. Gascard and Cohen (17) also describe low
affinity binding sites on Leach inside out vesicles, again in agreement
with our findings. The results presented here were obtained using an in vitro assay. Hence, it does not automatically follow that
the results of these binding assays can be directly related to the
situation in native membranes. In particular, the amounts of protein
4.1 and p55 which are bound at saturation (after correction for loss of
peripheral protein by alkali extraction) are 103 µg/mg for protein
4.1 and 25 µg/mg for p55, almost 2-fold higher than the amounts of
protein 4.1 and p55 (59 µg/mg and 16 µg/mg), respectively, in
normal membranes. Nevertheless, there is compelling evidence that GPC
is a major site for protein 4.1 binding in native membranes. Up to 75%
of protein 4.1 can be extracted from membranes of Leach phenotype under
conditions of low ionic strength in comparison with approximately 25%
from normal membranes(13, 14) . This result argues
that up to 50% of protein 4.1 binding in normal membranes involves
direct binding to GPC or binding via p55 to GPC. Consideration of the
relative affinity of these interactions and the likely stoichiometry
argues the latter since p55 is absent from both Leach phenotype
membranes and protein 4.1-deficient membranes(15) , suggesting
that all the p55 in normal membranes is simultaneously bound to GPC and
protein 4.1. Estimates of the number of molecules of p55 in normal
membranes are of the order 80,000/cell(34) , while protein 4.1
is 200,000/cell. Thus, a maximum of 40% of protein 4.1 molecules are
likely to be involved in this high affinity interaction (assuming 1:1
stoichiometry). The marked increase in the protein 4.1 extracted by low
ionic strength from Leach phenotype membranes may, for the most part,
reflect the absence of the high affinity protein 4.1 binding site on
GPC mediated by p55. Peptide GPC-3 binds directly to Leach phenotype
membranes in amounts sufficient to bind more than 85% of protein 4.1
present. In contrast, GPC-3 does not bind at all to normal
membranes(14) . If the distribution is similar in normal
membranes, then these results suggest that only a small proportion of
protein 4.1 (approximately 15%) is bound to GPC-independent sites in
these cells (presumably via band 3). The work presented here shows that
the GPC-3 site on protein 4.1 is distinct from the p55 binding site
(see above). Therefore, these results suggest that in normal red cells
the utilization of the three protein 4.1 binding sites is 45% GPC, 40%
p55, and 15% band 3. These conclusions are in agreement with those of
Pinder et al.(13) who concluded that band 3 and GPA
were not major binding sites for protein 4.1 and that GPC possesses two
protein 4.1 binding sites. The nature of protein 4.1 binding sites
in the normal erythrocyte membrane has been the subject of considerable
controversy over recent years. The realization that an additional
protein (p55) also participates in this interaction (15) provides a possible explanation for many of the
discrepancies since most studies have involved protein 4.1
reassociation with protein 4.1-depleted membranes without monitoring
the level of p55 contamination in the isolated protein 4.1 or the
protein 4.1-depleted membranes. Therefore, in most studies, the
majority of sites examined have been low affinity sites and, of these,
about 50% are trypsin-sensitive GPC-independent sites on band 3 with
the possibility that a small number are associated with GPA and the
lipid bilayer (this paper, 4, 35, 36). The remaining low affinity
trypsin-resistant sites are likely to be located on GPC at the GPC-3
site (the GPC-3 and GPC-2 regions of GPC survive extensive protease
digestion (papain, trypsin, and chymotrypsin, 50 µg/ml, 24 h at 37
°C) as judged by their failure to affect binding of specific
antibodies to GPC-2 and GPC-3, data not shown). It has been
suggested that the binding site on protein 4.1 for band 3 and possibly
also for GPC involves the negatively charged motif LEEDY (residues
37-41) on protein 4.1 and the oppositely charged motifs IRRRY
(residues 386-390) and LRRRY(343-347) on band 3 and YHRKG
(residues 85-89) on GPC(5) . The results presented here
would be consistent with this hypothesis since the LEEDY sequence is
located on the 30-kDa N-terminal domain of protein 4.1 and the GPC-3
peptide which contains the YHRKG motif completely inhibits protein 4.1
interaction with alkali-stripped normal and Leach membranes. Since
Leach membranes lack GPC, it is necessary to postulate that the
inhibition is due to blocking of the site on protein 4.1 which
interacts with band 3. The binding sites involved in protein 4.1-p55
interaction are unknown, but it is interesting to note that p55
contains an SH3 domain (29) and that the 30-kDa domain of
protein 4.1 contains a proline-rich motif (PPDP residues 81-84)
which might serve as a binding site for such an SH3
domain(32) . The results presented here suggest two models
of GPC-protein 4.1 interaction in native membranes (Fig. 8). In
the first model (Fig. 8A), two molecules of protein 4.1
can, in some cases, bind to a single GPC molecule. One molecule of
protein 4.1 binds through the GPC-3 site and the other through p55. In
the second model (Fig. 8B), some GPC molecules have a
single protein 4.1 molecule bound through the GPC-3 site while others
have a single protein 4.1 molecule bound through the GPC-3 site and
through p55. The first model (Fig. 8A) is compatible
with the reassociation assays involving alkali-stripped membranes where
the amount of protein 4.1 bound to alkali-stripped normal membranes at
saturation is 1.75-fold greater than that found in normal membranes.
The ratio of protein 4.1-GPC in normal membranes is approximately
1:1(33) . Since 55% of the protein 4.1 binding is directly to
GPC through the GPC-3 site under these conditions (Table 1), most
GPC molecules will have one protein 4.1 molecule bound at this site and
about half the GPC molecules will have a second protein 4.1 molecule
bound through p55. In native membranes, this model is less attractive.
All of the p55 is likely to be bound to GPC (80,000 molecules/cell)
accounting for approximately 40% of the protein 4.1 and GPC molecules,
and the remaining protein 4.1 could be accommodated on other GPC
molecules and band 3 rather than on the same GPC molecules. The
observation that binding sites for GPC-3 are not available on normal
membranes (14) would be consistent with this latter hypothesis.
Figure 8:
Models of protein 4.1-GPC/D-p55
interactions. A, two molecules of protein 4.1 can bind to a
single GPC molecule. One molecule of protein 4.1 binds through the
GPC-3 site, the other through p55. B, a single molecule of
protein 4.1 binds simultaneously to GPC-3 and
p55.
The observations reported in this paper do not address the
functional significance of protein 4.1-GPC interactions. Complete
deficiency of protein 4.1 (and therefore, p55) results in elliptocytic
cells and in several cases in hemolytic anemia(8) . In the case
of the Leach phenotype, absence of GPC/D (and therefore, p55) results
in a proportion of elliptocytic red cells(18) . Since Leach
phenotype red cells also have a reduced content of protein 4.1 (80% of
normal), it has been suggested that the slight elliptocytosis is a
consequence of the protein 4.1 deficiency rather than the absence of
GPC-protein 4.1-p55 interactions(15) . Support for this
hypothesis has come from experiments in which the abnormal
deformability properties (in the ektacytometer) of Leach phenotype red
cells were corrected by addition of the spectrin binding domain of
protein 4.1(34) . The functional role of interactions involving
GPC/D, p55, and protein 4.1 in the red cell remains an enigma since,
unlike protein 4.1 deficiency, there is no evidence of any pathological
consequences of GPC/p55 absence in Leach phenotype. The absence of
GPC/D in Leach phenotype results from a structural gene mutation (19) and so the concomitant absence of p55 is likely to be a
consequence of the GPC deficiency rather than an abnormality in the p55
gene itself. It seems possible that p55 and analogues of protein 4.1 in
other cells and tissues may interact with integral membrane proteins
other than GPC so that the relatively benign consequences of GPC
deficiency in red cells do not result in pathological consequences in
other cells and tissues. It is also possible that alternative
regulatory mechanisms operate in Leach phenotype red cells which
assuage the deleterious consequences of GPC/p55 deficiency. At the
present time, little is known about the regulatory pathways
(phosphorylation, palmitoylation) that may involve p55, protein 4.1,
and GPC in normal red cells.
Volume 270,
Number 10,
Issue of March 10, 1995 pp. 5360-5366
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
-spectrin, an interaction
which has been well studied(3) . The interaction of protein
4.1, the other protein linker, with the membrane has not been so well
studied, and the nature of the membrane attachment site(s) has yet to
be fully elucidated. Possible membrane attachment sites include band
3(4, 5, 6) , glycophorin C/D
(GPC/D)(
)(7, 8, 9) , glycophorin
A(10) , and the lipid bilayer(11, 12) .
-Lys
in the 30-kDa N-terminal domain and
His
-Glu
in 22/24-kDa domain. Rabbit
anti-p55 was prepared using a synthetic peptide corresponding to amino
acid residues 438-453
(Gly-Val-Asp-Glu-Thr-Leu-Lys-Lys-Leu-Gln-Glu-Ala-Phe-Asp-Gln-Ala-Cys-).
Antisera were raised against synthetic peptides in a manner analogous
to that described previously(14) .
)glycophorin A (BRIC 163(20) ), and
glycophorin C/D (BGRL 100(21) ) were available in-house.
GPC-peptides and corresponding antisera were as described
previously(14) . A synthetic peptide corresponding to the
entire cytoplasmic domain of GPA was synthesized using methods
previously described(14) .Immunoblotting
Erythrocyte membranes were
electrophoretically separated under reducing conditions on an 8% (w/v)
acrylamide homogenous Laemmli gel. Immunoblotting was as described by
Mallinson et al.(22) , except that 5% (w/v) bovine
milk powder, resuspended in phosphate-buffered saline at pH 7.4, was
used as the blocking agent.Protein Purification
Protein 4.1 was isolated
using a modified method of Tyler et al.(23) , as
described previously(14) . To prepare protein 4.1 free of p55,
the KCl concentration used to elute protein 4.1 from the DEAE-Sephacryl
column was reduced from 100 mM to 90 mM.
Contamination of protein 4.1 with p55 was detected by ELISA of the
fractions using anti-p55. p55 was purified according to the method of
Husain-Chishti et al.(24) . Protein determination was
performed using the procedure of Lowry et al.(25) .Limited Digestion of Protein 4.1 and Isolation of
30-kDa-containing Fragments
Protein 4.1 was digested with
chymotrypsin at an enzyme to substrate ratio of 1:100, according to
Leto and Marchesi(28) . Fragments containing the 30-kDa domain
of protein 4.1 were isolated by passing the digestion mixture through a
CNBr-activated Sepharose column to which anti-30-kDa antibody had been
coupled. Unbound material (fragments devoid of the 30-kDa domain) was
collected, and bound fractions (containing the 30-kDa domain) eluted
with 0.1 M glycine HCl, pH 2.7. Following neutralization (1 M Tris, pH 9.0, 5 M NaCl), the purity of the
fractions was confirmed by SDS-polyacrylamide gel electrophoresis and
subsequent staining of the gels with Coomassie Blue. These fragments
were subsequently used in the ELISA assay described below.Alkali-stripping of Human Erythrocyte
Membranes
Leaky membranes were prepared by the method of Dodge et al.(26) . Membranes were incubated with 2 mM phenylmethylsulfonyl fluoride at 0 °C for 5 min, before
stripping of peripheral proteins by washing with ice-cold 0.1 M NaOH. They were then washed three times with 5 mM sodium
phosphate buffer, pH 8.0 at 4 °C.Proteolytic Digestion of Membranes
Unstripped and
alkali-stripped membranes were diluted to 1 mg of protein/ml and
incubated with 5 µg/ml trypsin or chymotrypsin in 5 mM sodium phosphate, pH 8.0, for 1 h on ice. Digestion was stopped by
the addition of phenylmethylsulfonyl fluoride to 1 mM (final
concentration). Alkali-stripped membranes were prepared, as above, and
all membranes were washed three times with 5 mM sodium
phosphate buffer, pH 8.0 at 4 °C.Analysis of Protease-treated Membranes by
ELISA
The wells of a round bottomed microtiter plate (Sterilin)
were coated with 75 µl of lectin (from Agaricus bisporus;
Sigma) (7 µg/ml in 20 mM sodium phosphate buffer, pH 7.0)
and incubated for 1 h at room temperature(27) . All other
procedures were carried out on ice, and all washes were in PBS, pH 7.4,
0.055% Tween 20. The lectin-coated plate was washed, 75 µl of a 5%
suspension of protease-treated membranes was added, and the plate was
centrifuged at 930 
g for 2 min. The wells were
incubated with antibodies to band 3 (BRIC 169), glycophorin A (BRIC
163), or GPC/D (anti-GPC1, anti-GPC2, anti-GPC3, BGRL 100) (75
µl/well) for 2 h. After washing, the secondary antibody (rabbit
anti-mouse or swine anti-rabbit peroxidase conjugate; DAKO) was added
and incubated for 2 h. After further washing, 75 µl of substrate
was added (1 mg of o-phenylenediamine and 0.4 µl of
HO in 1 ml of 0.05 M citrate-phosphate
buffer, pH 5.0). The color reaction was terminated by the addition of
75 µl of 1 M HCl after 3 min. The plate was read on a
Titertek Multiskan ELISA reader (Flow Laboratories) at 492 nm.Binding Assays Involving Protein 4.1, Fragments of
Protein 4.1, p55, Erythrocyte Membranes, and GPC
Peptides
Binding studies were carried out using the ELISA system
previously described(14) . Preliminary experiments showed that
equilibrium was reached after 4 h and that no peptide or protein
binding was lost during washing. Within the limits of this assay, all
plots appear linear. For inhibition of protein 4.1 or p55 binding to
alkali-stripped membranes, serial dilutions of synthetic peptides were
made in 3% (w/v) bovine serum albumin and incubated with an equal
volume of protein 4.1 or p55 (final concentration 150 µg/ml)
overnight at 4 °C before addition to the microtiter plate. For p55
binding GPC-3-saturated protein 4.1: purified protein 4.1 (500
µg/ml) on microtiter wells, prepared as described above, was
incubated with saturating levels of GPC-3 (75 µl/well) overnight at
4 °C, before the addition of p55. For protein 4.1 binding
p55-saturated alkali-stripped membranes: p55 (500 µg/ml) (75
µl/well) was incubated with microtiter plate-bound alkali-stripped
membranes overnight at 4 °C, prior to the addition of protein 4.1.
Protein 4.1-GPC/D Interaction
In a previous
report we provided evidence for a direct interaction between protein
4.1 and the region of GPC corresponding to amino acid residues
82-98(14) . However, the recent report showing that Leach
phenotype membranes and membranes from individuals with protein 4.1
deficiency lack an additional protein, p55(15) , raised the
possibility that p55 might be a contaminant of protein 4.1 preparations
from normal erythrocytes and thereby influence protein 4.1 binding
assays of the type described in our earlier study. Examination of
protein 4.1 preparations from normal erythrocytes with anti-p55, either
by ELISA or by immunoblotting, showed a small amount of p55
(approximately 5%, data not shown) and so the experiments were repeated
with protein 4.1 preparations from normal cells which were demonstrably
free of p55 (see Fig. 1and ``Materials and
Methods''). Peptide GPC-3 bound directly to these purified protein
4.1 preparations (Fig. 2A) and completely inhibited
their binding to alkali-stripped normal membranes (Fig. 2B), confirming that, under the conditions of
this assay, protein 4.1 binds directly to GPC-3.
, GPC-2; 
, GPC-3. B,
inhibition of protein 4.1 binding to alkali-stripped normal membranes.
Protein 4.1 (150 µg/ml) was incubated with GPC peptide for 5 h at 4
°C before incubation with stripped normal membranes. Binding was
measured as described under ``Materials and Methods.''
Peptide concentrations are depicted on a logarithmic scale. Results are
plotted with error bars corresponding to mean ± S.D. (n = 4). X, GPC-1; , GPC-2; ,
GPC-3.
Role of p55 in Protein 4.1-GPC/D
Interaction
Evidence that p55 is absent from Leach phenotype
membranes and membranes from individuals with protein 4.1 deficiency
suggests that p55 binds to both GPC and protein 4.1 in normal
membranes. Indeed, purified p55 (Fig. 1) binds to protein 4.1 in
a concentration-dependent, saturable (48 µg of p55/mg of protein
4.1) manner, with high affinity (K
= 1.27
± 0.17 nM) (Fig. 3) and also binds saturably (63
µg of p55/mg of membrane protein) and with high affinity (K = 4.54 ± 0.13 nM) to
alkali-stripped normal membranes (Fig. 4A). As
expected, p55 does not bind to alkali-stripped Leach phenotype
membranes (Fig. 4A).
= 1.27 ± 0.17
nM.
, normal; X, Leach;
, trypsin normal; 
, chymotrypsin normal. Normal binding
capacity = 63 ± 5 µg of p55/mg of membrane protein; K = 4.54 ± 0.13
nM. B, binding of synthetic peptides to p55. GPC
peptides were incubated with p55 (150 µg/ml) overnight at 4 °C.
Binding was measured as described under ``Materials and
Methods.'' Results are plotted with error bars corresponding to mean ± S.D. (n = 5). X,
GPC-1; , GPC-2; , GPC-3.
GPC/D Independent Membrane Attachment Sites for Protein
4.1
In our previous paper(14) , we reported low levels
of protein 4.1 binding to alkali-stripped Leach phenotype membranes
(approximately 20% of normal). Since these membranes lack GPC/D and
p55, additional membrane binding sites for protein 4.1 must exist. Band
3 and GPA have been postulated as protein 4.1 binding
sites(4, 5, 6, 10) . To determine
whether these sites are utilized, Leach phenotype stripped membranes
were subjected to trypsin cleavage, which destroys the N-terminal
domain of band 3, while only having a mild effect on the cytoplasmic
domain of GPA (15% reduction) (as determined by monoclonal antibodies
to the cytoplasmic domains of band 3 (BRIC 169) and GPA (BRIC 163) data
not shown). Trypsin treatment destroyed all protein 4.1 binding to
alkali-stripped Leach phenotype membranes, suggesting that band 3
rather than GPA is a binding site for protein 4.1. This finding is
supported by the failure of a synthetic peptide corresponding to the
entire cytoplasmic domain of GPA to inhibit binding of protein 4.1 to
alkali-stripped Leach phenotype membranes and absence of binding of
this peptide to purified protein 4.1 (data not shown).Localization of Membrane Binding Sites on Protein
4.1
To identify the protein 4.1 domain(s) involved in binding to
the membrane, chymotryptic fragments of protein 4.1 were separated (as
described under ``Materials and Methods''), and their ability
to bind GPC, p55, and band 3 was examined. Fragments containing the
30-kDa domain were able to bind to p55 (Fig. 5A) as
well as to alkali-stripped normal membranes (Fig. 5B),
while those lacking the 30-kDa domain were unable to bind. Thus, the
membrane binding sites for GPC, p55, and band 3 appear to be located on
the same domain.
, 30-kDa-containing fragments; X, 22/24-kDa-containing
fragments.
, GPC-2; , GPC-3.
Relative Utilization of the Membrane Binding
Sites
We have sought to determine the proportion of protein 4.1
that binds to each of the three sites (GPC-3, p55, band 3) under the
conditions of our assay. Determination of protein 4.1 binding to
untreated and trypsin-treated alkali-stripped normal membranes which
have been incubated previously with saturating quantities of p55
provides a means of determining the relative contribution of protein
4.1 binding sites on p55 and GPC-3, respectively. Interpretation of
these experiments is possible because trypsin treatment destroys GPC
independent protein 4.1 binding sites (see above) and because p55
binds, exclusively, to a trypsin-sensitive site at the extreme C
terminus of GPC (see above).
=
0.114 ± 0.11 µM. , binding of protein 4.1 to
alkali-stripped normal membranes saturated with p55, binding capacity
258 ± 5 µg/mg membrane protein; K = 0.955 ± 0.19 nM. , binding
of protein 4.1 to trypsin-treated alkali-stripped membranes, binding
capacity 151 ± 6 µg/mg membrane protein; K = 0.125 ± 0.16 µM. ,
binding of protein 4.1 to trypsin-treated alkali-stripped normal
membranes saturated with p55, binding capacity 143 ± 8 µg/mg
membrane protein; K = 0.119 ± 0.14
µM.
for GPC-p55 of 4.54 nM and for p55-protein 4.1 of 2.5
nM) than interactions involving GPC-protein 4.1 (K = 0.125 µM) and band 3 (K = 0.11 µM). The measurement
of the K for the GPC-protein 4.1 interaction is
lower in this study than that reported previously(14) . This
discrepancy may be attributable to p55 contamination in the protein 4.1
preparations used in the earlier study. It is interesting to note that
Pasternack et al.(4) obtained similar high K values using protein 4.1 prepared by the method
of Tyler et al.(23) .
))
We acknowledge the generous cooperation of donor PL.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Gaetani, S. Mootien, S. Harper, P. G. Gallagher, and D. W. Speicher Structural and functional effects of hereditary hemolytic anemia-associated point mutations in the alpha spectrin tetramer site Blood, June 15, 2008; 111(12): 5712 - 5720. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Manno, Y. Takakuwa, and N. Mohandas Modulation of Erythrocyte Membrane Mechanical Function by Protein 4.1 Phosphorylation J. Biol. Chem., March 4, 2005; 280(9): 7581 - 7587. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. K. Parra, S. L. Gee, M. J. Koury, N. Mohandas, and J. G. Conboy Alternative 5' exons and differential splicing regulate expression of protein 4.1R isoforms with distinct N-termini Blood, May 15, 2003; 101(10): 4164 - 4171. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Shafizadeh, B. H. Paw, H. Foott, E. C. Liao, B. A. Barut, J. J. Cope, L. I. Zon, and S. Lin Characterization of zebrafish merlot/chablis as non-mammalian vertebrate models for severe congenital anemia due to protein 4.1 deficiency Development, March 11, 2003; 129(18): 4359 - 4370. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. J. Bruce, S. Ghosh, M. J. King, D. M. Layton, W. J. Mawby, G. W. Stewart, P.-A. Oldenborg, J. Delaunay, and M. J. A. Tanner Absence of CD47 in protein 4.2-deficient hereditary spherocytosis in man: an interaction between the Rh complex and the band 3 complex Blood, August 13, 2002; 100(5): 1878 - 1885. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Biederer and T. C. Sudhof CASK and Protein 4.1 Support F-actin Nucleation on Neurexins J. Biol. Chem., December 14, 2001; 276(51): 47869 - 47876. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Bennett and A. J. Baines Spectrin and Ankyrin-Based Pathways: Metazoan Inventions for Integrating Cells Into Tissues Physiol Rev, July 1, 2001; 81(3): 1353 - 1392. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Ye, D. A. Compton, M. M. Lai, L. D. Walensky, and S. H. Snyder Protein 4.1N Binding to Nuclear Mitotic Apparatus Protein in PC12 Cells Mediates the Antiproliferative Actions of Nerve Growth Factor J. Neurosci., December 15, 1999; 19(24): 10747 - 10756. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Gascard, W. Nunomura, G. Lee, L. D. Walensky, S. W. Krauss, Y. Takakuwa, J. A. Chasis, N. Mohandas, and J. G. Conboy Deciphering the Nuclear Import Pathway for the Cytoskeletal Red Cell Protein 4.1R Mol. Biol. Cell, June 1, 1999; 10(6): 1783 - 1798. [Abstract] [Full Text]
|
|
|
|
|
|
P. Gascard, G. Lee, L. Coulombel, I. Auffray, M. Lum, M. Parra, J.G. Conboy, N. Mohandas, and J.A. Chasis Characterization of Multiple Isoforms of Protein 4.1R Expressed During Erythroid Terminal Differentiation Blood, December 1, 1998; 92(11): 4404 - 4414. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. L. Borowsky and R. O. Hynes Layilin, A Novel Talin-binding Transmembrane Protein Homologous with C-type Lectins, is Localized in Membrane Ruffles J. Cell Biol., October 19, 1998; 143(2): 429 - 442. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.-J. C. Tang and T. K. Tang The 30-kD Domain of Protein 4.1 Mediates Its Binding to the Carboxyl Terminus of pICln, a Protein Involved in Cellular Volume Regulation Blood, August 15, 1998; 92(4): 1442 - 1447. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. D. Venezia, P. Maillet, L. Morle, L. Roda, J. Delaunay, and F. Baklouti A Large Deletion Within the Protein 4.1 Gene Associated With a Stable Truncated mRNA and an Unaltered Tissue-Specific Alternative Splicing Blood, June 1, 1998; 91(11): 4361 - 4367. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. F. Workman and P. S. Low Biochemical Analysis of Potential Sites for Protein 4.1-mediated Anchoring of the Spectrin-Actin Skeleton to the Erythrocyte Membrane J. Biol. Chem., March 13, 1998; 273(11): 6171 - 6176. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Nunomura, Y. Takakuwa, R. Tokimitsu, S. W. Krauss, M. Kawashima, and N. Mohandas Regulation of CD44-Protein 4.1 Interaction by Ca2+ and Calmodulin. IMPLICATIONS FOR MODULATION OF CD44-ANKYRIN INTERACTION J. Biol. Chem., November 28, 1997; 272(48): 30322 - 30328. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Marfatia, J. H. Morais-Cabral, A. C. Kim, O. Byron, and A. H. Chishti The PDZ Domain of Human Erythrocyte p55 Mediates Its Binding to the Cytoplasmic Carboxyl Terminus of Glycophorin C. ANALYSIS OF THE BINDING INTERFACE BY IN VITRO MUTAGENESIS J. Biol. Chem., September 26, 1997; 272(39): 24191 - 24197. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Saras, U. Engstrom, L. J. Gonez, and C.-H. Heldin Characterization of the Interactions between PDZ Domains of the Protein-tyrosine Phosphatase PTPL1 and the Carboxyl-terminal Tail of Fas J. Biol. Chem., August 22, 1997; 272(34): 20979 - 20981. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. W. Krauss, C. A. Larabell, S. Lockett, P. Gascard, S. Penman, N. Mohandas, and J. A. Chasis Structural Protein 4.1 in the Nucleus of Human Cells: Dynamic Rearrangements during Cell Division J. Cell Biol., April 21, 1997; 137(2): 275 - 289. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X.-L. An, Y. Takakuwa, W. Nunomura, S. Manno, and N. Mohandas Modulation of Band 3-Ankyrin Interaction by Protein 4.1. FUNCTIONAL IMPLICATIONS IN REGULATION OF ERYTHROCYTE MEMBRANE MECHANICAL PROPERTIES J. Biol. Chem., December 27, 1996; 271(52): 33187 - 33191. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Nunomura, Y. Takakuwa, M. Parra, J. Conboy, and N. Mohandas Regulation of Protein 4.1R, p55, and Glycophorin C Ternary Complex in Human Erythrocyte Membrane J. Biol. Chem., August 4, 2000; 275(32): 24540 - 24546. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |