THE JOURNAL OF Bvxoc~cn~ CHEMISTRY Vol. 254, No 13. Isue of July 10, pp. 6144-6150, 1979 Prmted in USA.

Reactive Sulfhydryl Groups of the Band 3 Polypeptide from Human Erythrocyte Membranes LOCATION IN THE PRIMARY STRUCTURE*

(Received for publication, July 13, 1978, and in revised form, December 27, 1978)

Anjana Rao$ and Reinhart A. F. Reithmeierg

From the Biological Laboratories, Harvard University, Cambridge, Massachusetts 02138

Human erythrocyte membranes contain a major transmembrane protein, known as Band 3, that is in- volved in anion transport. This protein contains a total of five reactive sulfhydryl groups, which can be as- signed to either of two classes on the basis of their susceptibility to release from the membrane by trypsin. Two of the groups are located in the region COOH- terminal to the extracellular chymotrypsin-sensitive site of the protein and remain with a membrane-bound 55,000-dalton fragment generated by trypsin treat- ment. The three sulfhydryl groups NHz-terminal to the extracellular chymotrypsin site are released from the cytoplasmic surface of the membrane by trypsin. All three groups are present in a 20,000-dalton tryptic frag- ment of Band 3. Two of these groups are located very close to the sites of trypsin cleavage that generate the 20,000-dalton fragment. The third reactive group is probably located about 15,000 daltons from the most NHz-terminal sulfhydryl group.

Two other well defined fragments of the protein do not contain reactive sulfhydryl groups. They are a 23,000-dalton fragment derived from the NHz-terminal end that is also released by trypsin from the cytoplas- mic surface of the membrane and a 19,000-dalton mem- brane-bound region of the protein that is produced by treatment with chymotrypsin in ghosts. The 20,000-dal- ton tryptic fragment may, therefore, constitute a sulfhydryl-containing domain of the Band 3 protein.

The Band 3 protein of human erythrocyte membranes is a 95,000-dalton transmembrane protein that is involved in anion transport (l-4). Previous reports from this and other labora- tories have been aimed at elucidating the structure of this protein both as a preliminary to the study of anion transport and to provide a model for the structure of other membrane transport proteins that are not as easily accessible to biochem- ical analysis. The approaches used have included cleavage by enzymatic or chemical means after labeling of regions internal or external to the membrane with impermeant reagents (5-8).

It was shown in earlier work (9) that Band 3 contains a total of five sulfhydryl groups reactive under appropriate conditions with N-ethylmaleimide. They can be classified into two sets

* This research was supported by Grant HL-08893 from the United States Public Health Service and Grant BMS 75.09919 from the National Science Foundation to Guido Guidotti, Harvard University. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “adoertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Supported by Training Grant GM 00782 from the National Insti- tutes of Health.

5 Medical Research Council of Canada Postdoctoral Fellow.

of reactive groups, separated by an extracellular site for cleav- age with chymotrypsin, that are distinguished by the condi- tions required for their reaction with N-ethylmaleimide. All five groups are accessible to a membrane-impermeant maleim- ide reagent (10) only from the cytoplasmic surface of the membrane. Three of the groups are present in a 60,000-dalton fragment generated from Band 3 by cleavage with chymotryp- sin at the external site; the other two are present in the complementary 35,000-dalton fragment. The sulfhydryl groups in the 35,000-dalton fragment are unreactive in ghosts; they can be labeled with [“H]N-ethylmaleimide in cells or after reduction of membranes with /?-mercaptoethanol. It is possible to take advantage of this difference in reactivity to label selectively one population or another of the sulfhydryl groups of the protein.

The experiments described here were designed to investi- gate the distribution of the N-ethylmaleimide-reactive sites of Band 3 in the fragments produced from it by treatment of membranes with proteolytic enzymes. Evidence will be pre- sented here to show that all three sulfhydryl groups NH2- terminal to the external chymotrypsin site are cytoplasmic, are released from the membrane by trypsin, and are located in a 20,000-dalton fragment that may represent a sulfhydryl- containing domain of Band 3. The NHt-terminal 23,000 dal- tons of the protein and the membrane-bound 19,000-dalton fragment produced by chymotrypsin treatment of ghosts con- tain no reactive sulfhydryl groups. The two reactive -SH groups located COOH-terminal to the external site for chy- motrypsin are contained in a 55,000-dalton tryptic fragment of Band 3 that remains tightly membrane bound.

This paper deals mainly with the three NHi-terminal sulfhydryl groups. These groups are located in the primary structure of the protein in relation to the sites for cleavage with trypsin. A map of the Band 3 protein showing the relative locations of sites of proteolytic cleavage and of CNBr frag- mentation is shown in Fig. 1 for reference. It incorporates the conclusions of the present work but in its basic form has been modified from Reference 6. The fragments have been assigned an orientation with respect to the NH2 terminus of the protein on the basis of previous work from this laboratory (7).

EXPERIMENTAL PROCEDURES

Materials

Aquasol, Liquifluor, Protosol, and N[ethyZ-2-“Hlethylmaleimide were obtained from New England Nuclear. N-Ethylmaleimide and phenylmethanesulfonyl fluoride were obtained from Sigma; acryl- amide, ,&mercaptoethanol, and sodium dodecyl sulfate from Bio-Rad; cyanogen bromide from Aldrich Chemical Co., Inc.; trifluoroacetic acid from Pierce Chemical Co.; Sepharose and Sephadex resins from Pharmacia; and L-1-tosylamido-2.phenylethyl chloromethyl ketone- trypsin and chymotrypsin from Worthington. 2.Nitro-5.thiocyano-

6144

by guest on February 2, 2019http://w

ww

.jbc.org/D

ownloaded from

Reactive -SH Groups of Band 3 6145

SH.SH

SH St- SH

N, ------7c

i

0 n

z 6 7 7

T I

T T*C, CO

60.000 35.000

23.000 I 20,000 _ 55.000

19,000

FIG. 1. A map of the Band 3 protein, showing the relative positions of the sites for proteolytic cleavage and the N-ethylmaleimide-reac- tive sulfhydryl groups. The sites in the 20,000-dalton tryptic fragment that are targets for CNBr cleavage have also been located. Details are discussed in the text. T,, cytoplasmic trypsin-sensitive sites; C, and C,,, cytoplasmic and extracellular chymotrypsin-sensitive sites.

benzoic acid was a gift from L. K. Drickamer, Harvard University. All other chemicals were reagent grade.

Methods

All procedures were performed at 0-4°C unless otherwise indicated. Preparation of Cells and Ghosts-Packed human red cells from

the Massachusetts Red Cross Blood Center were washed three times in isotonic saline and lysed in 10 to 20 volumes of 5 mM sodium phosphate, pH 8.0. Ghosts were washed 3 to 4 times in this buffer (4).

Extraction of Ghosts-Extraction of membranes with NaOH or EDTA/high salt was performed as described (11).

Proteolytic Cleavages-The conditions for proteolytic cleavage were modified from Reference 6.

Chymotrypsin Treatment of Ghosts-Ghosts at a concentration of 1 to 2 mg of membrane protein/ml were treated with 100 pg/ml of chymotrypsin for 1 h at room temperature. Phenylmethanesulfonyl fluoride was added to a final concentration of 1 to 2 mM, the mem- branes were incubated for 20 min on ice, and then extracted with 10 volumes of cold 0.1 N NaOH.

Trypsin Treatment of Ghosts-Membranes depleted of Bands 1,2, 5, and 6 by extraction with EDTA/high salt (11) were treated at a protein concentration of 1 to 2 mg/ml with 1 or 5 pg/ml of trypsin for 1 h at 0°C. Phenylmethanesulfonyl fluoride was added to a final concentration of 1 to 2 mM, the membranes were kept for 20 min on ice, and diluted with 10 volumes of 5 mM sodium phosphate, pH 8.0, or with water before centrifugation at 27,000 x g for 30 min. The supernatant was made 0.1% in SDS,’ boiled, lyophilized, and the residue resuspended to the original volume of membranes. The pellet was re-extracted with 0.1 N NaOH (11).

Purification of the 20,000-Dalton Tryptic Fragment-The 20,000- dalton tryptic fragment was purified by preparative slab gel electro- phoresis on 12.5% Laemmli gels. Gels were stained and destained and the region containing the fragment sliced out. The protein was eluted into 20 ml of 0.25 M NH,HCO:s, 10% P-mercaptoethanol, 0.2% SDS, at 37°C overnight. The eluate was lyophilized, dissolved in 1 ml of 0.25 M NH4HCO:s, and precipitated with 10 volumes of ice-cold acetone at -20°C overnight.

Cleavage with CNBr-Cleavage with CNBr of Band 3 and its fragments was performed essentially as described (8). It was necessary to subject the precipitated protein, purified from slab gels, to a vigorous reduction procedure before reproducible CNBr patterns were achieved. The precipitate was dissolved in 0.25 M NH,HCO:+, 1% SDS, centrifuged to remove insoluble material, made 30% in ,&mer- captoethanol, and reduced at 45°C for 24 h under nitrogen. Most of the P-mercaptoethanol was removed by lyophilization, the residue was dissolved in 70% trifluoroacetic acid, and the CNBr treatment was performed as usual (8). SDS was added to a concentration of 0.1% before removal of the CNBr by lyophilization. Recoveries of radio- activity after CNBr treatment were essentially loo%, indicating that the protein-bound N-ethylmaleimide is stable to the conditions of cleavage.

The material from CNBr treatment of the [“H]N-ethylmaleimide- labeled 20,000-dalton peptide was analyzed by gel filtration on a column of Sephadex G-100 (0.7 x 30 cm), run in 0.1% SDS, 5 mru

’ The abbreviations used are: SDS, sodium dodecyl sulfate; NTCB, 2-nitro-5-thiocyanobenzoic acid.

sodium phosphate, pH 8.0. Fractions (0.3 ml each) were collected at a flow rate of 0.09 ml/min, and O.l-ml aliquots from each fraction were counted in 5 ml of Aquasol.

Cleavage with NTCB-The conditions for cleavage with NTCB (7, 12, 13) were modified to permit direct analysis of the products on the Laemmli SDS gel electrophoresis system used. The reaction was performed at a final concentration of 1 to 2 mg of protein/ml with 0.5 mM NTCB in 10 mM Tris-acetate, pH 8.0, 1 mM EDTA for 30 min at room temperature. The pH was raised to about 9.0 with one-half volume of 1 M NH,HCO:,, and the tubes were incubated for cleavage for 6 to 20 h at 37°C. Modification in the presence of 1% SDS was performed in the same way. At the end of the incubation P-mercap- toethanol was added to a concentration of I%, and SDS was added if necessary to a concentration of 2%. The samples were boiled, lyoph- ilized, and the residues dissolved in Laemmli sample buffer.

Cleavage with NTCB of the purified 20,000-dalton fragment was performed in 1% SDS after vigorous reduction of the acetone precip- itate as described for CNBr fragmentation.

SDS Gel Electrophoresis-SDS polyacrylamide slab gel electro- phoresis was performed according to Laemmli (14). SDS-urea poly- acrylamide slab gel electrophoresis (13% acrylamide, 2O:l acrylamide: NJ’-methylene bisacrylamide) was performed according to Swank and Munkries (15). The molecular weight standards used, listed with their molecular weights, were: phosphorylase (Worthington), 94,ooO, bovine serum albumin (Sigma), 68,000; catalase (Worthington), 58,ooO, aldolase (Worthington), 40,000; carbonic anhydrase (Mann), 29,000; myoglobin (Mann), 17,200; and lysozyme (Worthington), 14,300. P-Galactosidase (Mann), 130,000, was also used for the exper- iment of Fig. 5. Insulin and the CNBr fragments of myoglobin were used as standards in the low molecular weight region.

After electrophoresis the gels were stained with Coomassie blue and destained. Gels were cut out for scanning in a Zeiss or Gilford spectrophotometer. Gel slices were swelled overnight in 5 to 10 ml of a mixture of 40 ml of Liquifluor and 50 ml of Protosol in 1 liter of toluene, and analyzed for bound radioactivity in a Beckman LS-233 liquid scintillation counter.

Quantitation of the Protein in Slab Gels-Three methods were used for the quantitation of the 20,000-dalton fragment in slab gels that had been stained with Coomassie blue and destained. 1) The gel was scanned at 550 nm using a Gilford spectrophotometer. The integrated absorbance of the 20,000.dalton peak was estimated as a fraction of the total absorbance on the gel. Knowledge of the total protein loaded on the gel (I 6) made it possible to estimate the weight of protein in the 20,000-dalton band. 2) The 20,000-dalton band was scanned and its absorbance compared with that of a series of bands containing increasing amounts of bovine serum albumin that were run on the same gel. The absorbance of the standard was linear up to at least 5 pg. The protein in the 20,000-dalton band was obtained by interpolation along this line. This method does not require a knowl- edge of the total protein loaded on the gel. 3) The 20,000-dalton band and the series of bands containing bovine serum albumin were sliced out and eluted overnight into 0.5 ml of 0.1 N NaOH, 0.2% SDS. The absorbance was read at 600 nm and was linear up to 5 yg of the standard.

Conversion into moles was made by using a molecular weight of 20,000 for this fragment. After methods 1 and 2 the gels were directly analyzed for radioactivity; for method 3 duplicate slots were sliced.

Labeling with [3HJN-ethylmaleimide-Labeling of membranes with [“H]N-ethylmaleimide has been described (9). For specific la- beling of the previously inaccessible -SH groups-in the COOH- terminal region of Band 3, membranes at 1 to 2 mg of protein/ml were reacted with 2 mM N-ethylmaleimide in 5 mM sodium phosphate, pH 7.0, for 1 h at 37’C, reduced with 5% /&mercaptoethanol under nitrogen for 1 h at 37”C, washed to remove the /3-mercaptoethanol, and reacted with 1 mM r’H]N-ethylmaleimide at the indicated specific activity for 1 h at 37’C. The membranes were then washed and extracted as required.

RESULTS AND DISCUSSION

Proteolytic Cleavage of Band 3 in Membranes

Fig. 2 shows the results of proteolytic digestion of Band 3 in

cells and in membranes. The conditions used were modified from those of Steck and coworkers (6); the fragmentation patterns produced agree with theirs except for slight differ- ences in apparent molecular weights due to the different

by guest on February 2, 2019http://w

ww

.jbc.org/D

ownloaded from

6146 Reactive -SH Groups of Band 3

electrophoresis systems used. It should be noted that Lanes 1 of both parts of Fig. 2 contain unextracted membranes, and, therefore, include spectrin, Bands 4, and other extrinsic pro- teins. Tracks 2 and 3 contain membranes extracted after proteolysis with 0.1 N NaOH; such treatment removes extrin- sic proteins from the membrane and leaves Band 3 or its membrane-associated fragment as the major peptides staining with Coomassie blue (6, 11).

Chymotrypsin

Chymotrypsin treatment of ghosts followed by NaOH ex- traction produces a very simple pattern containing a 19,000- dalton fragment representing a membrane-bound portion of Band 3 and a series of poorly staining bands in the 35,000- dalton region that may also be derived from Band 3 (Fig. 2A, Lane 2).

these cells with trypsin produces a membrane-bound 19,000- dalton fragment identical in molecular weight and in lack of associated label with the 19,000-dalton fragment produced by chymotrypsin from the same membranes or from ghosts (Fig. 2B, Lanes 2 and 3; Fig. 2A, Lane 2). The internal trypsin and chymotrypsin sites must, t,herefore, be close t,ogether (a dif- ference in molecular weight of -500 is normally detectable) and do not include a reactive cysteine residue between them (Fig. 1). The results of Fig. 2B also confirm a conclusion drawn previously (6) that the 20,000-, 23,000-, and 19,000-dalton fragments are derived from the 60,000-dalton chymotryptic fragment (Fig. 1).

Fig. 3A shows that the 19,000-dalton fragment produced by chymotrypsin treatment of [3H]N-ethylmaleimide-reacted membranes contains no associated radioactivity. There is some label comigrating with the 35,000-dalton bands; the radioactivity in this region increases if the ghosts are reduced before reaction with [“H]N-ethylmaleimide. Quantitation of the radioactivity in this region of the gel was not attempted because the yields of the fragments appear variable, and there is considerable uncertainty, on account of their diffuseness and poor staining, as to the fraction of total membrane protein that they constitute.

Trypsin Digestion of the Chymotryptic 60,000-Dalton Frag- ment

Chymotrypsin treatment of cells labeled with [‘H]N-ethyl- maleimide cleaves Band 3 to a 60,000-dalton fragment (Fig. 2B, Lane 1). Subsequent treatment of the membranes from

Supernatant from Trypsin Digestion of Band 3

Trypsin treatment of membranes extracted with EDTA/ high salt produces two major fragments of M, 23,000 and 20,000 in the supernatant, as well as a 55,000-dalton fragment that remains associated with the membrane (Fig. 2.4, Lanes 3 and 4). The 41,000-dalton fragment described by Steck (6) was only occasionally seen in these experiments and appears to be cleaved under the conditions used to the 23,000-dalton and 20,000-dalton fragments.

Fig. 3C contains the pattern of radioactivity appearing in the supernatant after trypsin digestion of membranes reacted with [‘H]N-ethylmaleimide. The NH?-terminal 23,000-dalton fragment contains only background levels of radioactivity while the 20,000-dalton fragment is highly labeled and ac- counts for all the radioactivity originally associated with Band 3. There is no increase in the labeling of the 20,000-dalton fragment when reduced ghosts are reacted with [“H]N-ethyl- maleimide, confirming the previous conclusion (9) that the sites that become accessible to N-ethylmaleimide upon reduc- tion are not located in this region of the protein. The 41,000-

FIG. 2. A, SDS polyacrylamide gel electrophoresis (12.5% Laem- mli) of proteolytic fragments generated from Band 3 in membranes. 1, untreated membranes, 12.5 pg. 2, NaOH-extracted pellet from chymotrypsin treatment of membranes, 7.5 pg. 3, NaOH-extracted pellet from trypsin treatment (5 pg/ml) of EDTA-extracted mem- branes, 10.5 pg. 4, supernatant from trypsin treatment (5 pg/ml) of EDTA-extracted membranes, 4.5 pg. 5, molecular weight standards, 0.5 pg each. B, SDS polyacrylamide gel elertrophoresis (12.5% La- emmli) of proteolytic fragments generated from the 60,000-dalton

12345

60Kb

35KD

19K b j 423K : 420K

chymotryptic fragment of Band 3. Membranes containing the 60,000 dalton fragment (“60K membranes”) were prepared from cells treated with 1 mg/ml of chymotrypsin for 90 min at 37°C. 1, “60K mem- branes,” 16 pg. 2, NaOH-extracted pellet from chymotrypsin treat- ment of “60K membranes,” 16 pg. 3, pellet from trypsin treatment (5 pg/ml) of EDTA-extracted “60K membranes,” 17 pg. 4, supernatant from trypsin treatment (5 pg/ml) of EDTA-extracted “60K mem- branes,” 7 pg. 5, molecular weight standards, 0.5 pg each. The stand- ards used are listed under “Experimental Procedures.”

by guest on February 2, 2019http://w

ww

.jbc.org/D

ownloaded from

Reactive -SH Groups of Band 3 6147

j2OK LJiJL r-

L...J

1 0

1 0

0

1 0

40

.2

.1

8

a!

.2

‘.l

-4

.3

.2

.1

2 4 6 s

CM FROM TOP OF GEL

FIG. 3. Radioactivity associated with the proteolytic fragments of Band 3 derived from membranes reacted with [“H]N-ethylmaleimide (15,750 cpm/nmol in A, B, and C; 2,130 cpm/nmol in D). Slab gels like those in Fig. 2 were scanned and sliced after staining. A, NaOH- extracted pellet from chymotrypsin treatment of membranes, 9 pg. B, supernatant from trypsin treatment (1 pg/ml) of EDTA-extracted membranes, 5 pg. C, supernatant from trypsin treatment (5 pg/ml) of EDTA-extracted membranes, 5 pg. D, NaOH-extracted pellet from trypsin digestion (5 pg/ml) of EDTA-extracted membranes specifi- cally labeled with [3H]N-ethylmaleimide (see under “Experimental Procedures”). Protein, 37 pg.

dalton fragment produced with low levels of trypsin is also labeled with [3H]N-ethylmaleimide (Fig. 3B), thus confirming that it contains the 20,000-dalton fragment (6).

Taken together, the results reported above show that all three reactive --SH groups of the 60,000-dalton chymotryptic fragment of the protein are likely to be contained within a

pmol 20K

FIG. 4. Quantitation of radioactivity in the 20,000-dalton tryptic fragment. Slab gels (12.5% Laemmli) like those of Fig. 3C were analyzed for radioactivity in the 20,00@dalton fragment and for protein by the three methods described under “Experimental Proce- dures.” A, by estimation of the absorbance of the 20K band as a fraction of the total absorbance in the scan. X, comparison of the absorbance of the 20K band to the absorbance of bands containing bovine serum albumin. 0, elution of dye bound to the 20K band and comparison of its absorbance with the absorbance of dye eluted from bovine serum albumin. p, represents the mean value obtained for the stoichiometry of labeling (2.5 f 0.5 mol of N-ethylmaleimide reacted/m01 of 20,000-dalton peptide).

central 20,000-dalton region, that may represent a sulfhydryl- containing domain of Band 3. The amino-terminal 23,000 daltons and the membrane-associated 19,000 daltons of the protein do not contain reactive sulfhydryl groups.

Trypsin Digestion of Band 3: Membrane-bound Peptides

Trypsin treatment of EDTA-extracted membranes leaves a major 55,000-dalton fragment associated with the membrane (Fig. 2A, Lane 3). This represents the portion of Band 3 that remains after the release of the 23,000- and 20,000-dalton fragments into the supernatant. It contains the two -SH groups that are inaccessible in ghosts but become reactive after reduction; they can be labeled specifically by the prior reaction of all accessible -SH groups with unlabeled N- ethylmaleimide, followed by reduction and reaction with ra- dioactive N-ethylmaleimide.

Fig. 30 shows the radioactivity remaining in the pellet after trypsin treatment of these specifically labeled membranes and extraction with NaOH. The 55,000-dalton peptide is seen to contain the major peak of radioactivity. The 20,000-dalton fragment released into the supernatant in these experiments contained less than 3% of the radioactivity of the control (membranes reacted with [“H]N-ethylmaleimide without prior reduction), although the protein released was equivalent; this confirms the specificity of labeling of the -SH groups in the 55,000-dalton region.

The labeled low molecular weight fragments seen in Fig. 30 are further degradation products of the 55,000-dalton frag- ment and have been characterized in more detail.’

The Number of Reactive -SH Groups in the 20,000-Dalton Domain

Quantitation of the Radioactivity in the 20,000-Dalton Fragment-A quantitative estimate of the radioactivity in the 20,000-dalton fragment released by trypsin seemed desirable in order to establish whether this fragment contains all three reactive -SH groups of this region of the protein.

’ Rao, A. (1978) Ph.D. thesis, Harvard University.

by guest on February 2, 2019http://w

ww

.jbc.org/D

ownloaded from

6148

4 20K

4

Reactive -SH Groups of Band 3

B

The methods used for quantitation of radioactivity and of protein are described in detail under “Experimental Proce- dures.” The results of the analysis are shown in Fig. 4. The stoichiometry of labeling was 2.5 f 0.5 mol of N-ethylmaleim- ide incorporated per mol of the 20,000-dalton peptide. It proved difficult by these methods to distinguish between the values of 2 and 3 for the stoichiometry of labeling of the 20,000-dalton fragment. This may reflect either a genuine heterogeneity of cleavage or the errors inherent in the meth- ods used for quantitation.

Cleavage with CNBr of the 20,000-Dalton Fragment-A better approach to estimating the number of labeled sites in a protein is to cleave with chemical or proteolytic agents and to find the number of labeled fragments in a limit cleavage. This number provides a minimum value for the number of reactive sites.

It is known from previous work (9) that one -SH group is located in each of the three major peptides derived by CNBr cleavage of the intact protein. It follows that the 20,000-dalton tryptic fragment, as the only one labeled, should contain regions of each CNBr peptide if no -SH groups are lost in small fragments of the protein during digestion with trypsin.

The purified 20,000-dalton peptide was treated with CNBr, and the products were analyzed on SDS-urea gels (Fig. 5). The 20,000-dalton fragment appears to be quantitatively cleaved to a fragment of M, = 14,500 which migrates with or very near the second CNBr fragment (CB-B, M, = 14,500) generated from the intact protein (Fig. 5B, Lanes 2 and 3). No peptides smaller than this are detected in stained gels. When CNBr cleavage was performed on the 20,000-dalton fragment derived from [“H]N-ethylmaleimide-reacted ghosts, about 50% of the original radioactivity was recovered from the gel, of which 60% was in the n/r, = 14,500 CNBr fragment (data not shown).

It seemed possible that radioactivity was being lost from the gels during staining and destaining. Gels containing the 20,000-dalton fragment before and after CNBr treatment were

FIG. 5. SDS polyacrylamide gel elec- trophoresis of the fragments generated from the purified 20,000-dalton tryptic peptide by chemical cleavage in SDS. A, Cleavage with NTCB (13.5% Laemmh gel). 1, molecular weight standards, 0.5 pg each. 2, purified 20K peptide, 1 pg. 3, purified 20K peptide treated with NTCB, 2 pg. B, cleavage with CNBr (16% Laemmli gel). 1, purified 20K peptide, 1 pg. 2, purified 20K peptide treated with CNBr, 1 pg. 3, purified Band 3 treated with CNBr, 5 pg. 4, molecular weight standards, 0.5 ag each. The standards used are listed under “Experimental Pro- cedures.” CB-C is not visible as a distinct band in Lane 3 of part B since it migrates anomalously on Laemmli gels.

sliced immediately after electrophoresis without staining; the patterns obtained are shown in Fig. 6. Essentially all the radioactivity placed on the gel is recovered in the single peak of the untreated 20,000-dalton fragment. After CNBr treat- ment about 70% of the original counts is recovered from the gel, distributed equally between two peaks of M, 14,500 and 5,000. The low recoveries in earlier experiments can probably be explained by losses of the 5,000-dalton fragment from the gels during staining and destaining.

About 30% of the radioactivity was still unaccounted for after CNBr treatment. In order to resolve this question, the material from a CNBr treatment of labeled 20,000-dalton fragment was analyzed by gel filtration on a Sephadex G-100 column run in 0.1% SDS (Fig. 7). Ninety-two per cent of the radioactivity applied was recovered from the column, of which

A

20K

B 20K CNBr

I 2 4 6 8 2 4 6 8

cm from top of gel

1

J

FIG. 6. Cleavage with CNBr of the 20,000-dalton tryptic peptide purified from membranes reacted with [‘H]iV-ethylmaleimide (2,130 cpm/nmol). SDS-urea polyacrylamide gels were sliced immediately after electrophoresis without staining. A, 20K peptide, 8 pg. Recovery of radioactivity from gel, 98%. B, 20K peptide treated with CNBr, 15.5 pg. Recovery of radioactivity from gel, 71%

by guest on February 2, 2019http://w

ww

.jbc.org/D

ownloaded from

Reactive -SH Groups of Band 3 6149

IO 20 30 40 50 fraction number

FIG. 7. Gel filtration chromatography on Sephadex G-100 (see under “Experimental Procedures”) of material from a CNBr treat- ment of the 20,000-dalton fragment purified from membranes reacted with r”HlN-ethvlmaleimide (2.130 cnm/nmol). A machine back- ground of40 cpm was subtracted from’all fractions. Total recovery of radioactivity from column, 92%. The arrows mark the positions of the void and included volumes of the column.

65% migrated in a broad peak with an RF of 0.18. SDS gel electrophoresis of material pooled from this peak gave a pattern of radioactivity identical with Fig. 6B. The remainder (28%) of the counts applied to the column chromatograph at a position close to the included volume of the column (RF 0.91). None of the counts in this peak could be located on an SDS gel.

It is possible from these data to conclude that at least two and probably three sulfhydryl-containing CNBr peptides are derived from the 20,000-dalton tryptic fragment. One of these has a molecular weight of 14,500 and must correspond to the CNBr fragment CB-B (Mr = 14,500) derived also from the intact protein by CNBr treatment (8). The fragment of M, 5,000 probably represents that portion of CB-A that overlaps into the 20,000-dalton fragment; its molecular weight is slightly low compared to previous estimates (8). The remain- ing third of the radioactivity of the 20,000-dalton fragment after CNBr treatment elutes near the column volume of a Sephadex G-100 column run in SDS and must contain oligo- peptides or smaller material. The peptide character of this material has not been unambiguously established but is likely since the peak contained ninhydrin-positive material, since S- succinylcysteine was detected upon amino acid analysis after prolonged hydrolysis, and since protein-bound radioactivity in N-ethylmaleimide is stable to the conditions of CNBr treatment (see under “Experimental Procedures”).

Location of the Reactive -SH Groups in Relation to the Sites for Cleavage with Trypsin

The Major -SH Group Reactive with NTCB-The major sulfhydryl group of Band 3 that is reactive with NTCB is located 24,000 daltons from the NHP-terminal end and pro- duces a 24,000-dalton fragment upon cleavage of the NTCB- reacted protein in SDS (6, 7, 17). Cleavage of Band 3 with the same reagent in membranes extracted with NaOH, EDTA, or lithium diiodosalicylate produces fragments of M, 24,000 and 23,000 (data not shown). The 23,000-dalton fragment comi- grates with the 23,000-dalton product of trypsin cleavage. It is produced in lower yields if the incubation for cleavage is carried out at higher temperatures (80°C) or for shorter times, suggesting that it is generated by proteolytic cleavage of the 24 OOO-dalton NTCB fragment by a membrane-associated pro- tease.

Trypsin digestion of NTCB-treated membranes results in the quantitative conversion of the 24,000-dalton NTCB frag- ment to a 23,000-dalton form, which comigrates with the 23,000-dalton fragment produced from the intact protein by trypsin treatment (data not shown). This evidence indicates, in confirmation of previous results (6), that the NTCB-reac- tive -SH group is located about 1,000 daltons COOH-termi- nal to the trypsin site in this region of the protein (Fig. 1).

The Reactive -SH Group in CB-C-The observation that the smallest CNBr fragment (CB-C) contains a reactive -SH group while the 19,000-dalton chymotryptic fragment, derived from the same region of the protein, does not, indicates that only part of CB-C is included within the 19,000-dalton frag- ment. The remainder of CB-C must be NHs-terminal to the 19,000-dalton fragment, since this fragment is bounded at its COOH terminus by the external chymotrypsin site while CB- C, being entirely included within the 60,000-dalton chymo- tryptic fragment (8), cannot extend beyond it.

Two conclusions follow: 1) The reactive -SH group of CB- C is located some distance NHs-terminal to the internal pro- tease-sensitive site that produces the 19,000-dalton fragment. 2) The product of CNBr cleavage of the 19,000-dalton frag- ment must be smaller than CB-C; the difference in molecular weight establishes an upper limit for the distance of the -SH group from the internal protease-sensitive site.

A gel comparing the fragments derived by CNBr cleavage of the 19,000-dalton fragment and of intact Band 3 is shown in Fig. 8. CB-C migrates in this system at an Mr of about 10,000 while the major product of CNBr cleavage of the 19,000-dalton fragment has an M, of 8,500; thus the -SH group of CB-C is within a region 1,500 daltons NHz-terminal to the internal protease site (Fig. 1). Other bands seen in Fig. 8, Lane 3, are the uncleaved 19,000-dalton fragment and 4 partial cleavage fragment migrating at approximately 10,000 daltons.

There is some conflict between our results and those of Drickamer (8), who placed CB-C at the COOH terminus of

1234

B-A

B-B

I CB-C ‘19K CNBr

FIG. 8. Comparison of SDS-urea polyacrylamide gels of the frag- ments generated by treatment with CNBr from the 19,000-dalton chymotryptic fragment and from intact Band 3. 1, products of CNBr treatment of myoglobin (arrowheads); protein, 5 pg. 2, products of CNBr treatment of purified Band 3; protein, 5 pg. 3, products of CNBr treatment of purified 19,000-dalton fragment; protein, 3 pg. 4, molec- ular weight standards (listed under “Experimental Procedures”) and insulin (arrowhead), 0.5 pg each.

by guest on February 2, 2019http://w

ww

.jbc.org/D

ownloaded from

6150 Reactive -SH Groups ofSand 3

the 60,000-dalton chymotryptic peptide. This was the result of an attempt to achieve internal con-

sistency in the molecular weights of the various fragments produced by chemical cleavage of Band 3 and in the amino acid compositions of the various regions of the polypeptide defined by these fragments and their sums and differences. Minor inconsistencies in molecular weights of the fragments, measured in the different gel systems used, can cause discrep- ancies in localization of fragments that amount to several thousand daltons of polypeptide chains. The discrepancies are not major, since all the available evidence can be accommo- dated if the positions of all three CNBr fragments are shifted by 5,000 to 10,000 daltons toward the NH, terminus of Band 3 (see Fig. 2 of Ref. 8).

NTCB Treatment of the Purified 20,000-Dalton Tryptic Fragment-2-Nitro-5-thiocyanobenzoic acid is a reagent that has been used to cleave polypeptide chains at cysteine resi- dues. Cleavage appears to be specific but is not always quan- titative (12, 13).

Treatment with NTCB in SDS of the 20,000-dalton frag- ment purified by preparative slab gel electrophoresis produces a fragment of M, 15,000 which is the only stained fragment visible on gels (Fig. 5). Cleavage is not quantitative, but no partial fragments appear to be produced. If cleavage is occur- ring at cysteine residues this suggests that a reactive -SH group occurs 15,000 daltons either from one end or from an adjacent --SH group. The major -SH group of the intact protein that reacts with NTCB is present in the 20,000-dalton fragment (Fig. 1 and text). I f this is also the major group reactive when the 20,000-dalton fragment is treated with NTCB, the result obtained suggests that a reactive -SH group occurs 15,000 daltons COOH-terminal to the first. This is in agreement with earlier conclusions that used partial fragments derived from NTCB cleavage of Band 3 and its 60,000-dalton chymotryptic fragment in SDS (8) and in gua- nidine hydrochloride (17).

Amino Acid Analyses of the Purified Fragments-Amino acid analyses of the 19,000-dalton chymotryptic fragment, purified from stripped chymotrypsin-treated membranes (6) on a P-100 column in SDS, and of the 23,000- and 41,000- dalton fragments purified from 12.5% Laemmli-slab gels, were performed in order to confirm the assignment of cysteine residues described above. The 19,000-dalton fragment con- tained 0.5 mole %, the 23,000-dalton fragment 0.6 mole %, and the 41,000-dalton fragment 0.9 mole % of cysteic acid after performic acid oxidation (18). This corresponds to 0.9, 1.1, and 3.4 cysteine residues in the respective fragments. These num- bers are in reasonable agreement with other data on the amino acid composition of these fragments (19). It is possible that the 19,000-dalton and 23,000-dalton fragments, which are not labeled with N-ethylmaleimide (Fig. 3), contain unreactive cysteine residues, although the cysteic acid values may be an overestimate since no cysteine was detected in unoxidized samples of these fragments, and since phosphoserine and other acidic amino acid derivatives elute at the same position as cysteic acid (18). Neither of these fragments reacts with 5,5’-dithiobis(2-nitrobenzoic acid) or with [3H]N-ethylmaleim- ide in SDS,” and neither contains reactive -SH groups that are involved in cross-linking the protein to a dimer after

.I R. A. F. Reithmeier, unpublished results.

treatment with Cu*‘-o-phenanthroline (6, 20). These results suggest that if either the 19,000-dalton or the 23,000-dalton fragment contains a cysteine residue it is very unreactive.

CONCLUSIONS

The experiments reported in this paper have been used to identify a cytoplasmic sulfhydryl-containing domain of Band 3 that contains all three -SH groups normally reactive with N-ethylmaleimide in ghosts. This region is bounded by tryp- sin-sensitive sites and contains reactive --SH groups close to each end (Fig. 1). The location in Fig. 1 of the second sulfhy- dry1 group from the NH, terminus is to some extent arbitrary; it has been placed 15,000 daltons from the most NH&erminal -SH group in accordance with the results of NTCB cleavage reported here and in Refs. 7 and 8. Evidence is presented to show that the 19,000-dalton chymotryptic fragment contains no reactive -SH groups. The CNBr fragment derived from the 19,000-dalton chymotryptic peptide is the smallest iden- tified fragment of the protein that spans the membrane, since it contains sites that are located on both the cytoplasmic and extracellular surfaces of the membrane. These are the internal protease-sensitive site at the NH2 terminus of the fragment and the region labeled by a variety of impermeant reagents from the outside of the cell (5, 7, 8). Two other -SH groups of Band 3 that are reactive with N-ethylmaleimide have been located in a 55,000-dalton membrane-bound product of trypsin digestion (Fig. 1).

Acknowledgments-We are grateful to Guido Guidotti and the members of his laboratory for their advice and interest during the course of this work. Siu Tong and Michael Ho are thanked for their help with the amino acid analyses.

REFERENCES

1. Cabantchik, Z. I., and Rothstein, A. (1974) J. Me&r. Biol. 15, 207-226

2. Ho, M. K., and Guidotti, G. (1975) J. Biol. Chem. 250,675-683 3. Bretscher, M. S. (1971) J. Mol. Biol. 59, 351-357 4. Fairbanks. G.. Steck. T. L.. and Wallach. D. F. H. (1971) Bio-

chemistry lb, 260612617 5. Jenkins, R. E., and Tanner, M. J. A. (1975) Biochem. J. 14’7,393-

399 6. Steck, T. L., Ramos, B., and Strapazon, E. (1976) Biochemistry

15, 1154-1161 7. Drickamer, L. K. (1976) J. BioZ. Chem. 251, 5115-5123 8. Drickamer, L. K. (1977) J. Biol. Chem. 252,6909-6917 9. Rao, A. (1979) J. Biol. Chem. 254,3503-3511

10. Abbott, R. E., and Schachter, D. (1976) J. Biol. Chem. 251,7176- 7183

11. Steck, T. L., and Yu, J. (1973) J. Supramol. Struct. 1, 220-232 12. Degani, Y., and Patchornik, A. (1974) Biochemistry 13,3-11 13. Jacobson, G. R., Schaffer, M. H., Stark, G. R., and Vanaman, T.

C. (1973) J. BioZ. Chem. 248,6583-6591 14. Laemmli, U. K. (1970) Nature 227,680-685 15. Swank, R. T., and Munlcries, K. D. (1971) Anal. B&hem. 39,

462-477 16. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.

(1951) J. BioZ. Chem. 193,265-275 17. Fukuda, M., Eshdat, Y., Tarone, G., and Marchesi, V. T. (1978) J.

BioZ. Chem. 253,2419-2428 18. Hirs, C. W. (1967) Methods Enzymol. 11, 59-62 19. Steck, T. L., Koziaxz, J. J., Singh, M. K., Reddy, G., and KliNer,

H. (1978) Biochemistrv 17. 1216-1222 20. Reithmeier, R. A. F., &d Rae, A. (1979) J. BioZ. Chem. 254,

6151-6155

by guest on February 2, 2019http://w

ww

.jbc.org/D

ownloaded from

A Rao and R A Reithmeiermembranes. Location in the primary structure.

Reactive sulfhydryl groups of the band 3 polypeptide from human erythroycte

1979, 254:6144-6150.J. Biol. Chem. 

  http://www.jbc.org/content/254/13/6144Access 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/254/13/6144.full.html#ref-list-1

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

by guest on February 2, 2019http://w

ww

.jbc.org/D

ownloaded from