THE JOURNAL OF BXOLOGICAL CHEMISTRY Vol. 257, No. 8, Issue of April 2 5 , pp. 4564-4569,1!382 Printed in U.S.A.

Erythrocyte Membrane Skeletal Protein Bands 4.1 a and b Are Sequence-related Phosphoproteins *

(Received for publication, November 4, 1981)

Steven R. Goodman@, John Yulll, Carol F. Whitfield+, Elsie N. Gulp.$, and Edward J. PosnakS From the $Milton S. Hershey Medical Center, The Pennsylvania State University, Department of Physiology, Hershey, Pennsylvania 17033 and the 1[ Scripps Clinic and Research Foundation, Department of Immunopathology, La Jolla, California 92037

Bands 4.1 a and b are proteins of 80,000 and 78,000 molecular weight, which are both present at -100,000 copies per erythrocyte ghost. Both proteins are com- ponents of the erythrocyte membrane skeleton. Bands 4.1 a and b are labeled when intact erythrocytes are incubated with [32P]orthophosphorie acid, and, there- fore, are phosphoproteins. One-dimensional partial proteolytic mapping analysis of 32P-labeled bands 4.1 a and 4.1 b and two-dimensional peptide mapping anal- ysis of ‘261-labeled bands 4.1 a and 4.1 b clearly dem- onstrated that the two proteins are sequence-related phosphoproteins. Band 4.1 purified by standard tech- niques (Tyler, J. M., Hargreaves, W. R., and Branton, D. (1979) Proc. Nutl. Acad Sci. U. S. A. 76, 5192-5196) contains bands 4.1 a and 4.1 b. Bands 4.1 a and 4.1 b bind to spectrin heterodimers in solution. We conclude that the erythrocyte skeletal proteins bands 4.1 a and 4.1 b are sequence-related phosphoproteins, both ca- pable of binding spectrin.

Spectrin is the major component of the erythrocyte mem- brane skeleton, which also contains actin, band 4.1, syndein (bands 2.1 -+ 2.6’), band 3, band 4.9, and band 7 (1-3). The spectrin membrane skeleton is attached to the cytoplasmic surface of the human erythrocyte membrane (4), and appears to play an important role in the maintenance of the mem- branes discoid shape (for a review, see Ref 5) and in restricting the lateral mobility of its macromolecules (for a review, see Ref. 2 ) . Human erythrocyte spectrin heterodimer is composed of two large nonidentical polypeptides chains (bands 1 and 2, or a! and /3) (240,~00 and 220,000 daltons), and appears to be an elongated flexible rod of 1000 A contour length (6). The spectrin ( ~ ~ $ 2 tetramer, which is formed by head-to-head association of two heterodimers with little or no overlap (6), appears to be the physiologically relevant unit of spectrin on the erythrocyte membrane (7-9). The spectrin tetramer, and hence the membrane skeleton, is bound to the membrane by a high affinity association with the syndeins (bands 2.1 + 2.6) (lo-12j. The syndeins are anchored to membrane by attach-

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Health. 0 Recipient of Grant HL-26059 from the National Institutes of

11 Recipient of Grant HL-21845 from the National Institutes of Health, and J. Yu is a recipient of an Established Investigator Award from the American Heart Association.

The major polypeptides of the erythrocyte membrane are desig- nated by Steck’s nomenclature (15). Band 2.1 has also been named ankyrin (11).

ment to -10% of the transmembrane band 3 molecules (13). Since spectrin tetramers are formed by head-to-head associ- ation of heterodimers, other linking proteins are required to form the membrane skeletal network. Actin protofilaments and band 4.1 bind to the terminal ends of the tetramer, cross- linking adjacent spectrin molecules, possibly by forming a spectrin-band 4.1-actin ternary complex (for a review, see Ref. 14).

Band 4.1 is a phosphoryiated peripheral membrane protein which is present in 1.8-2.3 X lo5 copies per erythrocyte ghost (15, 16), and has a molecular weight of -78,000 based on sodium dodecyl sulfate-polyacrylamide gel electrophoresis us- ing a continuous buffer system (15, 17). Band 4.1, which is pure by the criteria of SDS-PAGE2 utilizing this continuous buffer system, has been demonstrated to bind with high affinity to purified spectrin (18,131. The binding site has been localized by rotary shadowing and electron microscopy to the tail ends of the spectrin tetramer (18, 19), which is also the site of spectrin’s association with F-actin (20). Spectrin tetra- mers can bind to and cross-link F-actin in solution in the absence of band 4.1 (20-221, but a growing body of evidence suggests that the addition of band 4.1 strengthens and restores ea2* sensitivity to this interaction (23-25). Band 4.1 can be resolved into two components (referred to as bands 4.1 a and b) by SDS-gel electrophoresis using discontinuous buffer sys- tems (26). The question of whether bands 4.1 a and b are structurally and functionally related, or whether they are heterogenous proteins of similar moIecular weight which co- incidentally co-migrate on SDS-PAGE in continuous buffer systems had not been answered. Furthermore, the question of whether the “pure band 4.1” utilized in previous studies contained band 4.1 a or b or both had not been answered. These unanswered questions seriously complicated the inter- pretation of previous in vitro studies on the spectrin-band 4.1 interaction and the role which band 4.1 may play in the spectrin-actin interaction, and, therefore, obfuscated our un- derstanding of the molecular interactions of band 4.1 in the erythrocyte membrane skeleton. The importance of the spec- trin-band 4.1 interaction to normal membrane stability has been suggested by our recent finding that several patients with abnormally shaped, osmotically fragile red cells due to hereditary spherocytosis demonstrate a defective spectrin- band 4.1 interaction (27). Therefore, to clarify previous work on the spectrin-band 4.1 interaction by other investigators, and our own findings concerning a defect in this interaction in hereditary spherocytosis, it became necessary to elucidate the relationship between bands 4.1 a and b. In this study we

The abbreviations used are: SDS-PAGE, sodium dodecyl sulfate- polyacrylamide gel electrophoresis; EDTA, ethylenediaminetetraac- etate; DFP, diisopropyl fluorophosphate; TEMED, N,N,N‘,N’-tetra- methylenediamine.

4564

by guest on January 12, 2020http://w

ww

.jbc.org/D

ownloaded from

Erythrocyte Membran

demonstrate that bands 4.1 a and b are sequence-related phosphoproteins, which are both 1) associated with the eryth- rocyte membrane skeleton, 2) present in preparations of band 4.1 isolated by conventional techniques, and 3) found associ- ated with spectrin after incubation of purified band 4.1 and spectrin in vitro.

EXPERIMENTAL PROCEDURES

Materials ["P]Orthophosphoric acid (20 mCi/ml in 0.02 N HCI) and mon-

oiodinated '*'I-labeled Bolton-Hunter reagent (2000 Ci/mmol) were from New England Nuclear. '*'I (-17 mCi/mmol) was from Amer- sham Corp. Adenosine, DFP, Coomassie brilliant blue (R), diphenyl carbamyl chloride-treated trypsin, a chymotrypsin, and thermolysin were from Sigma. Acrylamide, N,N'-methylenebisacryylamide, am- monium persulfate, TEMED, and SDS were from Bio-Rad.

Methods Preparation of Erythrocyte Ghosts-Human erythrocyte ghosts

were prepared from freshly drawn blood anticoagulated in acid cit- rate/dextrose essentially by the procedure of Dodge et al. (28). except that the lysing buffer consisted of 5 mM NaPOr, 1 m~ EDTA, 0.4 mM DFP, pH 7.6.

Preparation of Triton Membrane Skeletons-Triton membrane skeletons were prepared by the method of Yu et al. (1). Briefly, erythrocytes were lysed and washed in 5 mM NaPOr, 1 mM EDTA, 0.4 mM DFP, pH 8.0. Ghosts were extracted for 20 min at 4 "C in 5-7 volumes of 1% Triton X-100,56 mM Na borate, 1 mM EDTA, 0.4 mM DFP, pH 8.0. Membrane skeletons were pelleted at 20,000 rpm, 40 min, 2 "C (Beckman-Spinco 52-21 centrifuge, J A 20 rotor). The membrane skeletons were washed once in lysing buffer and were solubilized in 1% SDS, 10 mM Tris, 1 mM EDTA, 32 mM dithiothreitol, pH 8.0 (100 "C, 10 rnin) for SDS-PAGE.

Preparation of '"P-labeled Erythrocyte Ghosts-Erythrocyte membrane phosphoproteins were ''*P-labeled by metabolically label- ing intact erythrocytes (6 ml of packed cells) with ["Plorthophos- phoric acid (10 mCi) by the method of Bennett and Branton (29). '"P- labeled erythrocyte ghosts were prepared in the Same manner as unlabeled ghosts.

Preparation of [32PJSpectrin Heterodimer-"P-labeled erythro- cyte ghosts were washed in 0.3 mM NaPOn, 0.2 mM EDTA, pH 7.6. The ghosts were resuspended in 0.3 mM NaP04, 0.2 mM EDTA, 50 p DFP, pH 7.6 (8 ml final volume), and were incubated at 37 "C for 25 min. The crude spectrin extract was separated from spectrin- depleted inverted vesicles by centrifugation at 55,000 rpm, 25 min, 2 "C (SW 60 rotor, Beckman-Spinco L2-65B centrifuge). ["P]Spectrin heterodimer was purified by rate zonal sedimentation as described by Bennett and Branton (29). The "2P-labeled crude spectrin extract was layered (2 m1/10 ml of gradient) onto a 5-2076 linear sucrose gradient in 10 mM NaPOr, 20 mM NaCI, 130 mM KCI, 1 mM MgCL, 1 mM EDTA, 1 mM dithiothreitol, 0.5 mM NaNn, pH 7.5. The gradients were centrifuged at 286,000 X g for 18 h (SW 41 rotor, 41,000 rpm, 2 "C), and were dripped into 16 fractions. Electrophoretically pure 9 S ['*PI spectrin heterodimer was obtained.

Preparation of "'Z-labeled Band4.l-Isolation of '2sII-labeled band 4.1 was by the standard technique of Tyler et al. (la), with the following important modification. 1) All buffers were brought to pH 7.6 after addition of DFP; if this is not done DFP lowers the buffer pH causing decreased spectrin extraction and low yields of band 4.1. 2) Labeling of band 4.1 with 0.5 mCi of '*'I-labeled Bolton-Hunter reagent is performed after purification by DEAE-cellulose chroma- tography.

Binding Analysis of the Spectrin-4.1 Interaction by Rate Zonal Sedimentation-Pure [32P]spectrin and "'I-labeled band 4.1 were dialyzed against 5 mM NaPOn, 20 mM KCI, 1 mM EDTA, pH 7.6 (12 h, 4 "C). The binding of "'I-labeled band 4.1 to ['*P]spectrin was assayed by rate zonal sedimentation through sucrose gradients as described by Tyler et al. (18). Protein mixtures were incubated in a 4 0 0 4 volume containing 5 mM NaPO,, 20 mM KCI, 1 mM EDTA, pH 7.6, for 120 min at 2 "C, prior to layering onto 520% linear sucrose gradients (11.8 ml) containing the same buffer. Following centrihga- tion at 286,000 X g for 15 h (41,000 rpm, SW 41 rotor, 2 "C),O.83-ml fractions were collected. Gradient fractions were assayed for ItiI and '*P activity, making standard corrections for '*sI/'*P crosstalk. In addition 50- to 100-pl aliquots from peak fractions were prepared for SDS-PAGE.

Le Proteins 4.1 a and b 4565

Gel Electrophoresis-SDS-PAGE in a continuous buffer system was performed by the method of Fairbanks et al. (17). as modified by Steck and Yu (30). SDS-PAGE was also performed using the discon- tinuous system of Laemmli (31). 1.5-mm thick slab gels contained a 13-cm separation gel consisting of 6 or 7% acrylamide (Ac/bis = 37.5),

and ammonium persulfate, and a 3-cm stacking gel consisting of 3% acrylamide, 0.08% bisacrylamide, 0.125 M Tris-HCI, pH 6.8,0.1% SDS, 1 mM EDTA, 0.05% TEMED, and ammonium persulfate. All gels were stained and destained as described by Fairbanks et al. (17). Scanning of gels and autoradiographs was performed with an LKB Zeineh soft laser scanning densitometer which has a beam width of 3-10 p n , and, therefore, can easily discriminate bands 4.1 a and b. Peak areas were determined by using the instrument's integrator, and by cutting and weighing peaks.

Peptide Mapping-The '"P-labeled spectrin p chain was excised from SDS-polyacrylamide slab gels of purified spectrin. '*P-labeled bands 4.1 a and 4.1 b were excised from discontinuous (31) SDS- polyacrylamide slab gels of ''*P-labeled ghosts. One-dimensional pep- tide maps of ["P]spectrin, band 4.1 a, and 4.1 b were obtained by limited proteolysis in SDS-polyacrylamide gels by the technique of Cleveland et al. (32).

For two-dimensional peptide mapping analysis, the polypeptides were separated in 1.5-mm thick 5-15% acrylamide gradient slab gels containing SDS (31). Individual polypeptides were cut from the gel and radioiodinated directly within the gel slice by the method of Elder et al. (33). After extensive washing, the proteins in the gel slices were cleaved with 0.1 ml of a IO-mm dithiothreitol solution containing

0.375 M Tris-HC1, pH 8.8, 0.1% SDS, 1 mM EDTA, 0.05% TEMED,

/l -2 - 2.1

-3 -4.1 a -4.1 b - 4.2

- 5

A B FIG. 1. Electrophoretic analysis of erythrocyte membrane

and Triton membrane skeleton protein on discontinuous poly- acrylamide gels. Samples prepared as described under "Methods" were analyzed on SDS-polyacrylamide (6%) gels utilizing the discon- tinuous buffer system (31), with Coomassie blue staining. A, eryth- rocyte membrane protein (60 pg); B, Triton membrane skeletal pro- tein (60 pg). The major polypeptides of the erythrocyte membrane are designated by Steck's nomenclature (15).

by guest on January 12, 2020http://w

ww

.jbc.org/D

ownloaded from

4566 Erythrocyte Membrane Proteins 4.1 a and b

3 4.la 4.1 b

of band 4.1 a/band 4.1 b = 1.2/1.0 in the membrane skeleton (5 gels scanned).

Bands 4.1 a and 4.1 b Are Both Phosphoproteins-Intact erythrocytes were metabolically labeled with ["'P]orthophos- phoric acid, "'P-labeled ghosts were prepared, and membrane proteins were separated by SDS-PAGE in the discontinuous buffer system (31). The autoradiogram presented in Fig. 2 demonstrates that bands 4.1 a and 4.1 b are phosphoproteins, which incorporate '"P to approximately the same extent.

5 -

6 - " 4 A B

FIG. 2. Bands 4.1 a and 4.1 b are phosphoproteins. 'lZP-labeled erythrocyte membranes prepared as described under "Methods" were analyzed on SDS-polyacrylamide ( 7 8 ) gels, utilizing the discontinuous buffer system (31). Autoradiograms were exposed for 3 days without any intensifying screen. A, er.ythrocyte membrane protein (64 pg), Coomassie blue staining; B, matching '"P autoradiogram. Arrows designate bands 4.1 a and 4.1 b.

5 pg of trypsin or chymotrypsin. Peptide analysis was performed by loading 1 X lo6 cpm onto each cellulose-coated thin layer chromatog- raphy plate.

Autoradiography-Autoradiograms were exposed for 1-14 days a t -70 "C, using Kodak X-Omat XAR-5 film with a Dupont Cronex Lighting Plus intensifying screen.

RESULTS

Bands 4.1 a and b Are Components of the Erythrocyte Membrane Skeleton-Human erythrocyte membranes were prepared in the presence of EDTA and DFP to inhibit pro- teolysis. The membrane proteins were separated by SDS- PAGE in a discontinuous buffer system (31). As demonstrated in Fig. 1 (gel A) band 4.1 can be resolved into two distinct components, which have been referred to as bands 4.1 a and b (26). Based on migration within these discontinuous gels, we have calculated molecular weight values of 80,000 for band 4.1 a and 78,000 for band 4.1 b. In our preparations of normal human erythrocyte ghosts the mole ratio of band 4.1 a/band 4.1 b = 1.0/1.0 (determined by scanning densitometry, 10 gels scanned). Since band 4.1 is present in 1.8-2.3 X l o 5 copies per erythrocyte ghost (16, 17), our ghost preparation contains -100,000 copies of band 4.1 a and 100,000 copies of band 4.1 b.

Also demonstrated in Fig. 1 (gel B) is the fact that bands 4.1 a and b are both components of Triton membrane skele- tons prepared by the method of Yu et al. (1). The mole ratio

A B C D E F G H I FIG. 3. One-dimensional partial proteolytic mapping of "P-

labeled bands 4.1 a, 4.1 b, and the spectrin /3 chain. Partial proteolytic mapping was performed by the method of Cleveland et af. (32). 5 pg of protease were added per lane. ')?P autoradiograms were exposed for 14 days with an intensifying screen. A, trypsin cleavage of band 4.1 a; B, trypsin cleavage of band 4.1 b; C, chymotrypsin cleavage of band 4.1 a; D, chymotrypsin cleavage of band 4.1 b; E , thermolysin cleavage of band 4.1 a; F, thermolysin cleavage of band 4.1 b; G, trypsin cleavage of the spectrin p chain; H , chymotrypsin cleavage of the spectrin p chain; I , thermolysin cleavage of the spectrin /3 chain.

A - w e

C

- FIG. 4. 'zSI-labeled tryptic and chymotryptic peptides of

bands 4.1 a and 4.1 b. Two-dimensional peptide mapping of '?$I- labeled bands 4.1 a and 4.1 b was performed by the method of Elder et al. (33). A, tryptic analysis of band 4.1 a; B, tryptic analysis of band 4.1 b; C, chymotryptic analysis of band 4.1 a; D, chymotryptic analysis of band 4.1 b.

by guest on January 12, 2020http://w

ww

.jbc.org/D

ownloaded from

Erythrocyte Membrane Proteins 4.1 a and b 4567

a b c d e 4 f "

b C

# c *

" I_

d e f - -

FIG. 5. Electrophoretic analysis of band 4.1 purification. Stages in the purification of band 4.1 performed by the method of Tyler et al. (18) were analyzed on A, SDS-polyacrylamide (5%) cylindrical gels with a continuous buffer system (17); or B, SDS-polyacrylamide (7%) slab gel with a discontinuous buffer system (31). a, human erythrocyte ghosts; 6, band 6-depleted ghosts; c, spectrin-depleted inverted vesicles; d, 1 M KC1 stripped vesicles; e, 1 M KC1 extract; column-purified 4.1. Arrows designate bands 4.1 a and 4.1 b.

2.1 4:". A

C

:m' - 2.1

GRADIENT FRACTION

FIG. 6. Spectrin-4.1 binding analysis. A, '"I-labeled band 4.1 (W) or heat-denatured band 4.1 (M) (25 pg/ml, 35,000 cpm/pg) was incubated with ["*P]spectrin heterodimers (150 pg/ml, 1000 cpm/pg) in a 400-pl volume containing 5 mM NaP04, 20 mM KCI, 1 mM EDTA, pH 7.6 (2 h, 4 "C). Bound and unbound '"I-labeled band 4.1 were separated by rate zonal sedimentation on a 5-20% linear sucrose gradient (see under "Methods"). The arrow denotes tbe velocity sedimentation position of purified free spectrin. Aliquots (100 X) of bound (fx 7) and unbound (fx 13, 14) '2sI-labeled band 4.1

3 4.1 - 4.2 - J

5 - Q 7 8

9

a b C a b c were removed for SDS-polyacrylamide gel electrophoresis and auto- radiography. SDS-polyacrylamide gel electrophoresis of bound and unbound 'ZsI-labeled band 4.1 was performed using: B, continuous buffer system (17), or C, discontinuous buffer system (31) (stacking gel 7% acrylamide). Autoradiograms were exposed for (B) 4 days, or (0 1 day with an intensifying screen. a, Coomassie blue-stained erythrocyte membrane protein; 6, autoradiogram of bound band 4.1; c, autoradiogram of unbound band 4.1.

by guest on January 12, 2020http://w

ww

.jbc.org/D

ownloaded from

4568 Erythrocyte Membrane Proteins 4.1 a and b

Peptide-mapping Analysis of Bands 4.1 a and 4.1 b-We have demonstrated that bands 4.1 a and 4.1 b are sequence- related proteins by two methods: 1) one-dimensional partial proteolytic mapping of 32P-labeled bands 4.1 a and 4.1 b, and 2) two-dimensional peptide mapping of ‘251-labeled bands 4.1 a and 4.1 b. One-dimensional partial proteolytic mapping of 32P-labeled bands 4.1 a and 4.1 b performed by the method of Cleveland et al. (32) is shown in Fig. 3. The autoradiogram presented clearly demonstrates that bands 4.1 a and 4.1 b yield nearly identical 32P-labeled peptide maps when partial proteolytic cleavage is carried out with trypsin, chymotrypsin, or thermolysin (Fig. 3, columns A-F). While the 32P-labeled maps of bands 4.1 a and b look nearly identical, they are clearly distinct from the ”P-labeled tryptic, chymotryptic, and thermolysin maps of the spectrin p chain (Fig. 3, columns G-Z); therefore, this technique can distinguish unrelated pro- teins. The one-dimensional peptide mapping demonstrates that bands 4.1 a and b are sequence-related phosphoproteins, a point which is further substantiated by our two-dimensional peptide mapping.

The two-dimensional peptide mapping of proteins radioio- dinated within the gel slice was performed by the method of Elder et al. (33), as modified by Yu and Goodman (10). We have previously demonstrated (10) that the erythrocyte mem- brane proteins bands 1, 2, 3, 4.1, 4.2, 5, and 6 all demonstrate distinct and characteristic 1251-labeled tryptic and chymotryp- tic two-dimensional maps. The syndeins (bands 2.1 -+ 2.6), however, exhibited a close similarity in the ‘251-labeled tryptic and chymotryptic two-dimensional maps, leading to the con- clusion that they were a family of sequence-related proteins (10). In Fig. 4, we demonstrate that both tryptic and chymo- tryptic two-dimensional peptide maps of 1251-labeled peptides of bands 4.1 a and 4.1 b exhibit close similarity. Therefore, bands 4.1 a and b are sequence-related phosphoproteins, rem- iniscent of the syndeins.

Are Bands 4.1 a and 4.1 b Both Present in “Purified” Band 4.12-Band 4.1 purified by the method of Tyler et al. (18) appears homogenous by the criterion of SDS-PAGE using a continuous buffer system (17). In agreement with Tyler et al. (18), band 4.1, which is extracted at high ionic strength from spectrin-depleted inverted vesicles and isolated by DEAE-cellulose chromatography, appears as an homogenous single polypeptide band on these gels (Fig. 5A). However, whether the “pure” band 4.1 contains band 4.1 a or b or both was not known. When various stages of the purification were electrophoresed using a discontinuous buffer system (31), we found that bands 4.1 a and b were both present in the purified band 4.1 preparation (Fig. 5B). The molar ratio of band 4.1 a/ 4.1 b in the purified preparation was 1.2/1.0 (5 gels scanned).

Spectrin-Band 4.1 Interaction-As all previous in vitro binding studies on the spectrin-band 4.1 interaction have been performed using this preparation of band 4.1 (18), it became necessary to ascertain whether bands 4.1 a and b are both capable of binding to spectrin. As previously demonstrated by Tyler et al. (18), when purified [32P]spectrin heterodimer is incubated with pure ‘”I-labeled band 4.1, a spectrin-band 4.1 complex is formed which can be separated from unbound 1251- labeled band 4.1 by rate zonal sedimentation on 5-20% linear sucrose gradients (Fig. 6A). The specificity of this interaction is demonstrated by the fact that binding is eliminated if heat- denatured (60 “C, 15 min) ‘“I-labeled band 4.1 is substituted in the incubation (Fig. 6A). Bound and free lZ5I-labeled band 4.1 were electrophoresed on SDS-polyacrylamide slab gels in a continuous buffer system, and the gels were autoradi- ographed (Fig. 6B). *251-labeled band 4.1 and [32P]spectrin /3 chain can be visualized on the autoradiogram of the bound material which sediments into the gradient, while only lZ5I-

labeled band 4.1 is visualized on the autoradiogram of the unbound material which remains on top of the gradient. The same fractions were electrophoresed on an SDS-polyacryl- amide slab gel in a discontinuous buffer system (31) (Fig. 613. AS demonstrated in Fig. 6C both bands 4.1 a and 4.1 b are bound to spectrin. As the unbound bands 4.1 a and 4.1 b are present in the same ratio (1.2/1.0) as the bound bands 4.1 a and 4.1 b, we conclude that there does not appear to be preferential binding of one or the other sequence-related peptide to the spectrin molecule.

DISCUSSION

We have demonstrated that erythrocyte membrane skeletal proteins bands 4.1 a and 4.1 b are sequence-related phospho- proteins. Both proteins are present in band 4.1 preparations isolated by the method of Tyler et al. (18), and both become associated to spectrin on in vitro incubation of purified spec- trin heterodimer with purified band 4.1. Our results are essen- tial to the interpretation of previous spectrin-band 4.1 binding studies, and suggest that it is appropriate to use band 4.1 isolated by this method (18) in these binding studies, in the absence of homogenous bands 4.1 a and 4.1 b. The purification of bands 4.1 a and 4.1 b as distinct proteins will represent a formidable task because of the similarity in size, charge, and function of these proteins. Recent in vitro spectrin-band 4.1 binding studies have shown that 2 mol of band 4.1 bind to 1 mol of spectrin heterodimer in solution (19). Our own finding that bands 4.1 a and 4.1 b become attached to spectrin during in vitro incubation can be interpreted in either of two ways: 1) both bands 4.1 a and 4.1 b are capable of binding directly to spectrin, or 2) band 4.1 a binds directly to spectrin and band 4.1 b indirectly through attachment to 4.1 a (or vice versa).

Bands 4.1 a and 4.1 b are sequence-related phosphoproteins, and their different migration on discontinuous SDS-polyacryl- amide gels could be due either to: 1) the proteolytic cleavage of an -2000-dalton terminal end of band 4.1 a, yielding band 4.1 b; 2) a difference in phosphate content yielding a charge difference; or 3) some other post-translational modification leading to a charge difference. Further structural analysis and total phosphate analysis on bands 4.1 a and b must be per- formed before this question can be unambiguously answered. In any case, the fact that the spectrin-binding proteins bands 4.1 a and 4.1 b are sequence-related, just as the syndeins have been found to be a family of sequence-related proteins (10, 34), raises the interesting question of the significance of a family of sequence-related spectrin-binding protein to the physiology of the erythrocyte.

Acknowledgments-We thank Jeffrey Ettinger, Kathy Shiffer, and Linda Casoria for their help on various aspects of this report. We also thank Zainab Khan for photographic assistance, and Jeanette Schwartz for preparing this manuscript.

REFERENCES 1. Yu, J., Fishman, D. A., and Steck, T. L. (1973) J. Supramol.

2. Goodman, S. R., and Branton, D. (1978) J. Supramol. Struct. 8,

3. Sheetz, M. P., and Sawyer, D. (1978) J. Supramol. Struct. 8,

4. Nicholson, G. L., Marchesi, V. T., and Singer, S. J . (1971) J. Cell.

5. Lux, S. E. (1979) Semin. Hematol. 16, 21-51 6. Shotton, D., Burke, B., and Branton, D. (1979) J. Mol. Biol. 131,

7. Goodman, S. R., and Weidner, S. A. (1980) J. Biol. Chem. 255,

8. Middaugh, C. R., and Ji, T. H. (1980) Eur. J. Biochem. 110,

Struct. 1, 233-248

455-463

399-412

Biol. 51, 265-272

303-329

8082-8086

587-592

by guest on January 12, 2020http://w

ww

.jbc.org/D

ownloaded from

Erythrocyte Membrane Proteins 4.1 a and b 4569

9. Liu, S. C., and Palek, J . (1980) Nature (Lond.) 285,586-588 10. Yu, J., and Goodman, S. R. (1979) Proc. Natl. Acad. Sci. U. S. A .

11. Bennett, V., and Stenbuck, P. J . (1979) J. Biol. Chem. 254,

12. Luna, E. J., Kidd, G. H., and Branton, D. (1979) J. Biol. Chem.

13. Bennett, V., and Stenbuck, P. J. (1979) Nature (Lond.) 280,

14. Branton, D., Cohen, C. M., and Tyler, J . (1981) Cell 24, 24-32 15. Steck, T. L. (1974) J. Cell Biol. 62, 1-19 16. Jones, M. N., and Nickson, J. K. (1981) Biochim. Biophys. Acta

17. Fairbanks, G., Steck, T. L., and Wallach, D. F. H. (1971) Bio-

18. Tyler, J . M., Hargreaves, W. R., and Branton, D. (1979) Proc.

19. Tyler, J. M., Reinhardt, B. N., and Branton, D. (1980) J. Biol.

20. Cohen, C. M., Tyler, J. M., and Branton, D. (1980) Cell 21,

21. Brenner, S. L., and Korn, E. D. (1979) J. Biol. Chem. 254,

76,2340-2344

2533-2541

254,2526-2532

468-473

650, 1-20

chemistry 13,2606-2617

Natl. Acad. Sci. U. S. A. 76, 5192-5196

Chem. 255,7034-7039

875-883

8620-8627

22. Brenner, S. L., and Korn, E. D. (1980) J. Biol. Chem. 255,

23. Ungewickell, E., Bennett, P. M., Calvert, R., Ohanian V., and

24. Cohen, C. M., and Korsgren, C. (1980) Biochern. Biophys. Res.

25. Fowler, V., and Taylor, D. L. (1980) J. Cell Biol. 85, 361-376 26. Mueller, T. J., and Morrison, M. (1977) J. Biol. Chern. 252,

27. Goodman, S. R., Kesselring, J . J., Weidner, S. A., and Eyster, E.

28. Dodge, J. T., Mitchell, C., and Hanahan, D. J. (1963) Arch.

29. Bennett, V., and Branton, D. (1977) J. Biol. Chem. 252,2753-2763 30. Steck, T. L., and Yu, J. (1973) J. Supramol. Struct. 1, 220-232 31. Laernmli, U. K. (1970) Nature (Lond.) 227,680-685 32. Cleveland, D. W., Fischer, S. G., Kirschner, M. W., and Laernmli,

U. K. (1977) J. Biol. Chem. 252, 1102-1106 33. Elder, J. H., Pickett, R. A., 11, Hampton, J., and Lerner, R. A.

(1977) J. Biol. Chem. 252,6510-6515 34. Siegel, D., Goodman, S. R., and Branton, D. (1980) Biochim.

Biophys. Acta 598,517-527

1670-1676

Gratzer, W. B. (1979) Nature (Lond.) 280, 811-814

Commun. 97,1429-1435

6573-6576

M. (1981) J. Supramol. Struct. Cell. Biochem. Suppl. 5, 131

Biochem. Biophys. 100,119-130

by guest on January 12, 2020http://w

ww

.jbc.org/D

ownloaded from

S R Goodman, J Yu, C F Whitfield, E N Culp and E J Posnakphosphoproteins.

Erythrocyte membrane skeletal protein bands 4.1 a and b are sequence-related

1982, 257:4564-4569.J. Biol. Chem. 

  http://www.jbc.org/content/257/8/4564Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/257/8/4564.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on January 12, 2020http://w

ww

.jbc.org/D

ownloaded from