ARCHIVES

OF BIOCHEMISTRY

AND

BIOPHYSICS

Vol. 296, No. 2, August 1, pp. 685-690,1992

Epitope Mapping of Monoclonal Antibodies to the Escherichia co/i F, ATPase CYSubunit in Relation to Activity Effects and Location in the Enzyme Complex Based on Cryoelectron Microscopy R. Aggeler,*

R. A. Capaldi,*,’

S. Dunn,?

and E. P. Gogol**2

*Institute of Molecular Biology, University of Oregon, Eugene, Oregon 97403; and TDepartment Health Sciences Center, University of Western Ontario, London, Ontario, Canada N6A 5Cl

of Biochemistry

Received March 27, 1992, and in revised form April 20,1992

The F1 ATPases of the bacterial plasma membrane, mitochondrial inner membrane, and chloroplast thylakoid membrane are similar and organized from five different

subunits, (Y, p, y, 6, and t, in the stoichiometry 3:3:1:1:1 (l-3). There are a total of six nucleotide binding sites per F1, three catalytic sites located mostly on the @subunits, and three noncatalytic sites located mostly on the a subunits (reviewed in (1, 2)). The (Ysubunit has been studied extensively by chemical modification and by site-directed mutagenesis (4-8). Senior has summarized this work in a model in which the (Ysubunit is divided into several domains along the linear sequence, including an N-terminal membrane binding domain, followed by a nucleotide binding domain (noncatalytic sites) and then an a-0 signal transmission region (9). One approach to studying structure-function relationships in multisubunit enzymes, such as the ATP synthase, is the use of monoclonal antibodies (mAbs).3 To date, mAbs have been used to examine the subunit homologies of F1 from different organisms (e.g., 10,ll) and to locate functionally important regions of the protein by their effect on activity (e.g., 11, 12). Intact mAbs and their Fab’ fragments have also been useful in mapping the topology of Escherichia coli F1 ATPase (ECF,) in negative staining (13) and in cryoelectron microscopy studies (14,15). In top view, ECFl shows six major masses arranged hexagonally around a central cavity, with a seventh mass located asymmetrically in this cavity (14). Fab fragments generated from a mAb to the (Y subunit labeled every other peripheral mass, indicating that the three (Y and three /3 subunits alternate in the ECFl structure (14). Fab fragments made from mAbs to the y, 6, and c subunits, respectively, showed that these

’ To whom correspondence should be addressed. * Present address: Program in Molecular and Cell Biology, University of Texas at Dallas, F031, P.O. Box 830688, Richardson, TX 75083.

3 Abbreviations used: mAbs, monoclonal antibodies; ECFl , Escherichia coli F, ATPase; SDS, sodium dodecyl sulfate.

The interaction of Escherichia coli F1 ATPase (ECF1) with several different monoclonal antibodies (mAbs) specific for the a subunit has been examined. The epitopes for each of the mAbs have been localized by using molecular biological approaches to generate fragments of the a subunit. The binding of several of the mAbs has also been examined by cryoelectron microscopy of ECFl Fab complexes. One of the mAbs, au, bound in the region Asn 109-Val 153 without affecting ATPase activity. Most of the mAbs bound in the C-terminal third of the a subunit. MAb aI bound between residues Gln 443 and Trp 513. This mAb activated ATPase activity and was visualized in cryoelectron microscopy, superimposed on the a subunit, indicating that the epitope was on the top or bottom of ECFl in the hexagonal projection. Other mAbs to the C-terminus, including ao which also activated the enzyme, reacted between Gly 371 and Trp 513 but failed to bind to small overlapping fragments within this sequence. The epitopes for these mAbs are probably formed by the folded polypeptide which occurs only in Western analysis when long stretches of the a subunit are present., suggesting that the C-terminus of a is a self-folding domain. In cryoelectron microscopy, Fab fragments for an were seen extending from the sides of the ECFl complex in hexagonal projection. 0 1992 Academic

Press,

Inc.

0003-9861/92 $5.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form

685 Inc. reserved.

686

AGGELER ET AL.

smaller subunits are located close to fi subunits in the hexagonal projections (14, 15). Several mAbs to the (Y subunit of ECFl are now available. Here we describe experiments in which functional effects, and the binding location based on cryoelectron microscopy, are related to segments of the a subunit by epitope mapping. Our findings are discussed in terms of a domain structure of the (Y subunit.

SDS-polyacrylamide gel electrophoresis was performed using 15% polyacrylamide gels and the buffer system of Laemmli (25). Proteins were electrophoretically transferred to Immobilon membranes using carbonate blotting buffer as described previously (26). Immunochemical detection of proteins was performed as described previously (11) except that the second antibody was conjugated to alkaline phosphatase, and the bands were developed by immersion of the membrane in a solution containing 1 mg/ml of Fast Blue RR and 1 mg/ml of a-naphthyl phosphate in 50 mM Tris-HCl, pH 8.0, 100 mM NaCl, 5 mM MgC&.

MATERIALS

RESULTS

AND METHODS

ECFi was isolated from E. coli strain AN1460 by a modification of the method of Wise et al. (16) to be described elsewhere. Antibody purification and characterization was described in Dunn et al. (11) and Gogol et al. (14). Electron microscopy and image analysis was conducted as described by Gogol et al. (14). ATPase activity measurements were made as described in Aggeler et al. (12). Strains and plusmid construction. Portions of the (Y subunit were expressed as fusion proteins using vectors pUC8 (17), pTZ18U or pTZ19U (18), or pUR291(19). Recombinant DNA procedures were carried out by standard methods (20). The sequence of the uric operon has been determined (21). pSD7 was constructed by inserting the 2.5kb EcoRI fragment of pAP55 (22) into the EcoRI site of pUC8. The insert in pSD7 is inverted relative to expression from the lot promoter and bears most of uncA, all of uncG, and the beginning of uncD. pSD48 was constructed by inserting the 1.0.kb RsaI fragment of pSD7 into the H&II site of pUC8 and selecting a transformant bearing a plasmid with the insert in the proper orientation for expression. pSD52 was constructed by inserting the 1.4-kb EcoRI-Sal1 fragment of pSD7 into pUC8 that had been cut with EcoRI and SalI. pSD53 was constructed by inserting the 0.41-kb Sal1 fragment of pSD7 into the Sal1 site of pUC8 and selecting a transformant bearing the insert in the proper orientation for expression. pSD58 was constructed by inserting the 0.70kb SmuI-Es&J1 fragment of pSD7 into pTZ18U which had been cut with HincII and selecting a transformant bearing a plasmid with the insert in the proper orientation for expression. pSD62 was constructed by partially digesting pSD52 with HincII, religating, and selecting a clone which retained the uric sequence encoding amino acid residues N4-V153 of the o subunit. pSD63 was constructed by digesting pSD52 to completion with HincII and religating. The resultant plasmid contains 3026 bp, including those encoding amino acid residues N4-V108 of the 01subunit. pSD64 was constructed by inserting the 0.28-kb HinPl fragment of pSD58 into the AccI site of pTZ18U and selecting a transformant bearing a plasmid with the insert in the proper orientation for expression. pSD65 was constructed by inserting the 0.30-kb HpuII fragment of pSD58 into the AccI site of pUC8 and selecting a transformant bearing the insert in the proper orientation for expression. pSD66 was constructed by transferring the uric sequence from pSD53 into pUR291, using the BumHI and Hind111 sites. pSD73 was constructed by inserting the 135bp HirzcII fragment of pSD52 into pUC8 that had been cut with HincII and selecting a transformant bearing the insert in the proper orientation for expression. pSD75 was constructed by transferring the uric sequence from pSD73 into pUR291, using the BumHI and Hind111 sites. All constructs were verified by restriction endonuclease mapping and, except for pSD53 and pSD73, by expression of fusion proteins that were recognized by polyclonal rabbit antiicu antibodies. Plasmids derived from pUC8, pTZ18U, and pTZ19U were expressed in E. coli JM103 (23). Plasmids derived from pUR291 were expressed in E. coli F’ 11 recA (24). Western blots. Cultures of strains bearing plasmids encoding o( subunit fusion proteins were grown in L-broth to early logarithmic phase, induced with 1 mM isopropyl-@-D-thiogalactoside, and grown an additional 4 h. Cells were harvested by centrifugation, resuspended in polyacrylamide gel sample buffer containing 2% SDS, and extracted for 5 min in a boiling water bath. Extracts were stored at -20°C and used for Western blots without further treatment.

The three mAbs studied in detail are aI and aII described in Gogol et al. (14) and a-4 described in the study of Dunn et al. (11) (here called ao). Both aI and an activate ECFl (11, 12). In the case of aI, this effect had been suggested to result from an altered binding of the E subunit, which acts as an inhibitor of ATPase activity in isolated ECFl (12). The third mAb studied, cyII, had no effect on activity, although immunotiter studies show that this antibody binds to the native ECFl complex (12, 14). Epitope Mapping The binding sites of the different mAbs with the (Ysubunit were localized by epitope mapping using a molecular biological approach. Thus a series of plasmids were constructed, most bearing sequences encoding portions of the (Y subunit fused in frame either to the ,8-galactosidase apeptide encoded by vectors pUC8 and pTZ18U or to the C-terminus of the entire P-galactosidase protein encoded by pUR291. Table I lists the sizes and the residues of a included in those fusion proteins which were most useful for the epitope mapping. In two cases, pSD53 and pSD73, the fragments of (Y fused to short polypeptides were not detectable by Western blotting, presumably due to instability in the cells (data not shown). The sequences encoding these fragments were therefore fused into the 3’ end of the entire Zuc.Zgene in plasmids pSD66 and pSD75,

TABLE a! Subunit

Fusion

Proteins

I

Used in Epitope

Plasmid

Vector

Mol wt

pSD48 pSD52 pSD53 pSD58 pSD62 pSD63 pSD64 pSD65 pSD66 pSD73 pSD75

pUC8 pUC8 pUC8 pTZ18U pUC8 pUC8 pTZ18U pUC8 pUR291 puts pUR291

14,390 61,583 5,479 18,233 26,764 22,168 10,479 22,535 > 120,000 16,189 >120,000

No. of residues 130 568 49 168 246 201 96 201 >1,050 147 >1,050

Mapping Residues ofa T395-W513 N4-D476 v475-w513 G371-W513 N4-V153 N4-V108 Q443-W513 G391-T489 v475-w513 N109-V153 N109-V153

MONOCLONAL

ANTIBODIES

TO THE

respectively. Fusion to the larger protein appeared to enhance the stability of the polypeptide regions of interest. Extracts of cells which had been induced to produce the fusion proteins were analyzed on Western blots using either the mAbs or the polyclonal rabbit anti-a antibodies. The most important results are shown in Fig. 1. Note in all strains the presence of the full-length (Y subunit encoded by the chromosome. As seen in A, polyclonal rabbit antibodies against a also recognized polypeptides of the expected sizes in extracts of cells bearing plasmids pSD52 (lane 4), pSD58 (lane 5), pSD64 (lane 7), pSD66 (lane 8), pSD65 (lane 9), pSD63 (lane lo), pSD62 (lane ll), and pSD75 (lane 12), although some intracellular degradation is evident in some cases. A faint band is visible for pSD48 at a molecular weight of about 10,000, as opposed to the expected size of 14,390. Thus the fusion peptide produced from pSD48 is highly unstable in the cell. This result is in marked contrast to the relative stability and good production of the slightly longer fusion peptide encoded by pSD58. Identical blots were probed with the various mAbs to the CYsubunit. Neither 01~(B) nor ffn (C) mAbs recognized the 61.5-kDa fusion protein which lacks the amino-terminal 3 residues and the carboxy-terminal 37 residues of the subunit (pSD52, lane 4), but both recognized the 18.2kDa fusion protein containing the C-terminal 143 residues (pSD58, lane 5), indicating that the epitopes of both mAbs are in the C-terminal portion of the subunit. (Y~(B) also recognized the partially degraded lo-kDa peptide produced from pSD48 (lane 6) and the 10.5-kDa polypeptide containing the C-terminal 71 residues from Gln 443 to Trp 513 (pSD64, lane 7). However, the polypeptides containing residues Gly 391 to Thr 489 (pSD65, lane 9) or residues Va1475 to Trp 513, the extreme C-terminal section (pSD66, lane 8, and pSD53, not shown), were not recognized. The results are graphically summarized in Fig. 2 and allow us to conclude that the epitope recognized by (Y~is included within residues Gln 443 to Trp 513 and that it extends to, or is dependent on, residues to the N-terminal side of Val 475 and to the C-terminal side of Thr 489. Antibody (Yn did not recognize any of the C-terminal sections smaller than that encoded by pSD58 (Fig. lC, lanes 6-9). From these results we conclude that the epitope recognized by au is included within residues Gly 371 to Trp 513 (Fig. 2) and that it must extend to, or be dependent on, residues to the N-terminal side of Gln 443 and to the C-terminal side of Asp 476. The same pattern of fusion protein recognition was shown by antibodies (Y1 (lB5AC) and 01-3 (3G2H5) described in Dunn et al. (11) (data not shown). Antibody (~11(Fig. 1D) was the only mAb that recognized the large fusion peptide containing most of the cysequence produced from pSD52 (lane 4). The epitope was further restricted to residues Asn 4 to Val 153 by the recognition of the fusion peptide produced from pSD62 (lane 11).

Escherichia

coli F, ATPase 1

A

2

3

4

687

01 SUBUNIT

5

6

7

8

9

101112

atA- 94



kDa L-Z - 67 kDa

-. w-cme---

--I

s

- 43 kDa 4

- 30 kDa

/d

d

- 20.1 kDa - 14.4 kDa /

/

B SW--

Epitope mapping of monoclonal antibodies to the Escherichia coli F1 ATPase alpha subunit in relation to activity effects and location in the enzyme complex based on cryoelectron microscopy.

The interaction of Escherichia coli F1 ATPase (ECF1) with several different monoclonal antibodies (mAbs) specific for the alpha subunit has been exami...
4MB Sizes 0 Downloads 0 Views