deletion of camp-binding site b in the regulatory subunit of camp

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1988 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 263, No. 34, Issue of December 5, pp. 18247-18252,1988 Printed in U. S.A.

Deletion of CAMP-binding Site B in the Regulatory Subunit of CAMP- dependent Protein Kinase Alters the Photoaffinity Labeling of Site A*

(Received for publication, March 25, 1988)

Garth E. RingheimS, Lakshmi D. Saraswate, Jo& Bubisv, and Susan S. Taylor(( From the Department of Chemistry, University of California, San Diego, La Jolh, California, 92093

Photoaffinity labeling with 8-azidoadenosine 3':5'- monophosphate is a highly selective method for prob- ing the CAMP-binding sites of the regulatory subunits of CAMP-dependent protein kinase and for identifying specific residues that are in close proximity to the CAMP-binding sites. The CAMP-binding site of a mutant R'-subunit has been characterized here and contrasted to the native R'-subunit. This mutant R'-subunit was generated by oligonucleotide-directed mutagenesis and lacks the entire second CAMP-binding domain which includes both of the residues, Trp'" and TyrS7', that are photolabeled in the native R'-subunit. The mutant R'-subunit, nevertheless, is photoaffinity- labeled with high efficiency, and the residue covalently modified was identified as The position of

based on a computer graphic model of CAMP- binding site A is proposed and correlated with the presumed locations of TyrS7l and TrpZBo in the native R-subunit.

Photoaffinity labeling also can be used to detect functional CAMP-binding sites following electrophoretic transfer of the denatured protein to nitrocellulose. Labeling of the immobilized protein on nitrocellulose required a functional CAMP-binding site A that can be photoaffinity-labeled in solution based on the following criteria. 1) The type I R-subunit is photolabeled, whereas the type I1 R-subunit is not. A primary feature which distinguishes these two R-subunits is that the R'-subunit is photolabeled at both sites A and B, whereas covalent modification of the R"-subunit occurs only at site B. 2) The truncated mutant of the R'- subunit which lacks the entire second CAMP-binding domain can be photolabeled on nitrocellulose. 3) A mutant R'-subunit which can no longer be photolabeled in site B is still photolabeled on nitrocellulose. 4) A mutation which abolished cAMP binding to site A also abolished photoaffinity labeling after transfer to nitrocellulose.

* This work was supported in part by United States Public Health Service Grant GM-34921. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$Supported in part by University of California Biotechnology grant.

8 Supported in part by United States Public Health Service Train- ing Grant AM07233. Present address: Dept. of Biochemistry, Bran- deis University, Waltham, MA 02154.

II Supported in part by the "Consejo Nacional de Investigaciones

Dept. of Chemistry, MIT, Cambridge, MA 02139. Cientificas y Technol&icas," Caracas, Venezuela. Present address:

)I Member of the University of California-San Diego Molecular Genetics Center. To whom reprint requests should be addressed.

Photoaffinity labeling of the regulatory (R) subunits of CAMP-dependent protein kinase with 8-N3cAMP1 has proved to be a remarkably specific method for mapping the CAMP- binding sites (1-4). Each regulatory subunit contains two tandem CAMP-binding sites, sites A and B; and two different forms of regulatory subunit, skeletal muscle R'-subunit and heart R"-subunit, have been characterized extensively (5-7). Covalent modification of both subunits occurs with a high stoichiometry and is also highly selective in that a single residue is targeted in each CAMP-binding site. This approach has provided specific information about the amino acid side chains that are in close proximity to C-8 of the adenine ring. In addition, affinity labeling of proteolytically generated forms of the R-subunit shows altered patterns of labeling, indicating subtle differences in the CAMP-binding sites even though overall cAMP binding is retained (4).

The results of photoaffinity labeling with 8-NScAMP can be summarized as follows. The type I1 R-subunit from porcine heart is covalently modified at a single residue, Tyr3", with a stoichiometry that approaches 1 mol of 8-N3cAMP/mol of R"-monomer (1). This modification is due to 8-NacAMP bound to site B. Although site A is occupied in the native protein by 8-N3cAMP, no stable covalent cross-links are generated following photolysis. Only when the R"-subunit is modified by limited proteolysis does covalent modification of site A occur. Under these conditions, Tyrlg6 is covalently modified (4).

In the type I R-subunit, 2 residues are covalently modified following affinity labeling with B-N~cAMP. One residue,

is homologous to the single residue of covalent modification in the R"-subunit. The other site, TrpZW, lies at the beginning of site B. Tyr371 is modified by 8-NzcAMP bound to site B, whereas TrpZW is modified by 8-N3cAMP bound to site A (2, 3). Similar covalent modifications are seen for porcine and bovine R'-subunits and for R'-subunit that has been overexpressed in Escherichia coli (8).

Now that an expression vector has been constructed for the R'-subunit, it is possible to generate specific mutant forms of the protein that have altered CAMP-binding properties. Con- sequently, several additional proteins are now available which have been generated by directed mutagenesis of the cloned and overexpressed R'-subunit (9-11). One of these mutant R*-subunits was generated by introducing a stop codon at residue 260. This mutation results in the expression of a stable R'-subunit that is lacking in CAMP-binding site B. The mutant R'-subunit retains a single high affinity CAMP-binding site and still forms a complex with the C-subunit (11). However, this truncated R'-subunit lacks both of the sites that are typically covalently modified by 8-NScAMP in the native full-length R'-subunit. Suprisingly, this truncated R'-

The abbreviations used are: 8-N3cAMP, 8-azidoadenosine 3':5'- monophosphate; HPLC, high performance liquid chromatography.

18247

18248 Deletion of CAMP-binding Site B

subunit was photoaffinity-labeled very well with 8-N3[32P] CAMP. The new site of covalent modification has been identified in this truncated R'-subunit, and its location correlated with the overall model of binding site A. In addition, the photoaffinity labeling of the R'-subunit that has been immobilized on nitrocellulose was characterized more fully using the R"-subunit and a series of mutant forms of the R'-subunit. A functional CAMP-binding site A was required in order to observe efficient labeling of the R'-subunit once it had been denatured and immobilized on nitrocellulose.

EXPERIMENTAL PROCEDURES

Materials-Reagents were purchased from the following sources: 8-N3[32P]cAMP (50 Ci/mmol) and Biotrans nylon membranes, ICN; 8-N3t3H]cAMP (50 Ci/mmol), DuPont-New England Nuclear; Nu-p- tosyl-L-phenylalanine chloromethyl ketone-treated trypsin, Sigma; trifluoroacetic acid (Sequanal-grade), Pierce Chemical Co.; acetonitrile (HPLC-grade) and Luria broth (L-broth) agar, Fisher; nitrocellulose (0.45 pm), Bio-Rad; and ampicillin, Boehringer Mannheim.

Proteins-R'-subunit was purified from porcine skeletal muscle, and the R"-subunit from porcine heart as described previously (12, 13). A proteolytic fragment of the R'-subunit was generated by limited proteolysis with thermolysin (14). The type I R-subunit also was purified from E. coli 222 using a construction vector, pLST-2, which was constructed by inserting a bovine R'-subunit cDNA into pUC7 (8,15). This expressed R'-subunit represents a fusion protein with 10 residues of B-galactosidase fused to the amino terminus of the entire R'-subunit. The proteolytic fragment of the R"-subunit was generated by limited proteolysis with chymotrypsin (4). Three mutant R'- subunits also were used 1) Y371/F, where the photolabeled residue in site B, Tyr3'l, was replaced by Phe (9); 2) a truncated mutant, W26O/St, where the Trp codon at residue 260 was replaced by a stop codon (11); and 3) a mutant, R209/K, that replaced ArgW in site A with Lys (10).

Gel Electrophoresis-Samples were prepared for gel electrophoresis by boiling for 1 min in the presence of 1% sodium dodecyl sulfate, 0.01% phenol red, 10% sucrose, and 5% 2-mercaptoethanol. Slab gels were prepared with 12.5% acrylamide as described by Laemmli (16). After electrophoresis, the gels were either stained with 10% glacial acetic acid, 25% isopropyl alcohol, and 0.25% Coomassie Blue R-250 or transferred electrophoretically to nitrocellulose using a Hoefer Scientific Instruments electrotransfer apparatus. The electrophoretic transfer was continued overnight a t 4 "C at 50 V in 20 mM Tris-C1 (pH 8.3), 154 mM glycine, and 20% methanol. The nitrocellulose was then incubated for 1 h at room temperature in 0.05% Tween 20, 150 mM sodium chloride, and 10 mM Tris-C1 (pH 7.4) to block nonspecific binding sites.

Photoaffinity Labeling-Photoaffinity labeling in solution was car- ried out with a hand-held UV light source using conditions described previously (1). In the case of the nitrocellulose sheets, the sheets were blocked with Tween 20 as described above and then incubated once with 10 mM Tris-C1 (pH 7.4) containing 8-N3[32P]cAMP in the dark in a sealed bag. After 10 min, the nitrocellulose was washed with ice- cold buffer alone and then exposed to UV light for 5 min. The nitrocellulose was then dried and exposed to x-ray film.

Identification of Photolabeled Residue-Two mg of the truncated R'-subunit, W26O/St (1 mg/ml), was incubated on ice with 8-N3[3H] CAMP for 60 min and then photolyzed for 5 min as described previously (1, 3). The sample was dialyzed exhaustively (four buffer changes, 2 liters each) against 10 mM potassium phosphate (pH 8.0) and then digested for 4 h a t 37 "C with Nu-p-tosyl-L-phenylalanine chloromethyl ketone-treated trypsin (1:50, w/w). A second equivalent aliquot of trypsin was added after 2 h. The resulting tryptic peptides were resolved by HPLC.

The tryptic peptides were resolved on a Vydac Cls reverse-phase column using a flow rate of 1 ml/min and a 10 mM sodium phosphate (pH 6.8) gradient of 0-30% acetonitrile in 120 min, followed by 30- 50% acetonitrile in 40 min. The radioactive fractions were pooled and reinjected into a second identical reverse-phase column using a 0.1% trifluoroacetic acid gradient of 12-15% acetonitrile in 60 min.

The labeled peptide was sequenced using an Applied Biosystems Gas-Phase Sequencer with an on-line HPLC system for detection of phenylthiohydantoin-derivatives.

RESULTS

Both the native R'-subunit (8) and a truncated R'-subunit (11) are expressed in large amounts in E. coli 222 following transformation with pLST-2 and pLST-2(W260/St), respec- tively. When cell extracts containing equivalent amounts of the native R'-subunit and the truncated R'-subunit (W260/ St) were denatured with sodium dodecyl sulfate, electrophoresed on polyacylamide gels, electrophoretically transferred to nitrocellulose, and photoaffinity-labeled with ~-Ns[~ 'P] CAMP, the results shown in Fig. 1 were observed. The mutant R'-subunit was photolabeled with an efficiency that paralleled the native R'-subunit even though both of the sites of covalent modification in the intact R'-subunit, Tyr371 and TrpZ6', are missing in this mutant.

In order to characterize more precisely this novel photoaffinity labeling, 2 mg of the mutant protein was photolyzed in solution with 8-N3[3H]cAMP. The stoichiometry of covalent incorporation was 0.9 mol of labeled nucleotide to 1 mol of truncated R'-subunit (data not shown). After removing the excess reagent by dialysis, the remaining protein was digested with Nu-p-tosyl-L-phenylalanine chloromethyl ketone-treated trypsin. The resulting tryptic peptides were resolved by HPLC, as indicated in Fig. 2. A single major radioactive peptide was observed, representing 34% of the total radioactivity loaded onto the column, and this entire region was pooled and rechromatographed using the same column but a different gradient. From this second column, a doublet containing two radioactive peptides was isolated (Fig. 3), representing 68% of the total radioactivity loaded onto the column. The sequence of the peptide in the first radioactive fraction indicated by the arrow is shown in Fig. 4 in addition to the actual chromatographs of phenylthiohydantoin-derivatives for steps 1-3. The sequence is noteworthy in that no residue is identified at position 2, which is clearly resolved as a tyrosine in the peptide that is isolated from the unmodified R'-subunit. This residue at step 2 can thus be identified as the site of covalent modification. The radioactivity remained bound to the filter; and hence, it was not possible to correlate directly the radioactivity with step 2 in the sequence. The site of modification was thus identified as Tyr244. Each fraction containing radioactivity in Fig. 3 was sequenced and found to correspond to the same site of modification. The reason for the appearance of a doublet based on the absorbance is not apparent.

M R

NATIVE R-)

MUTANT R+

FIG. 1. Photoaffinity labeling of immobilized native and truncated R-subunits; E. coli 222 was transformed either with pLST-2 ( l a n e R ) , which expressed the native R-subunit, or with pLST-2(W260/St) ( l a n e M), which codes for a truncated form of the R-subunit. Cell extracts from both cultures were solubilized and electrophoresed on polyacylamide gels in the presence of sodium dodecyl sulfate. The proteins were then transferred electrophoretically to nitrocellulose and photoaffinity-labeled with 8-N3[32P]cAMP. Each lane contained equivalent amounts of the R-subunit (-5 pg) based on staining with Coomassie Blue R-250 or immunodetection with anti-R'-subunit serum antibodies.

Deletion of CAMP-binding Site B 18249

20 40 60 IO 100 120 140

FRACTION NUMBER

FIG. 2. HPLC elution of tryptic peptides from truncated R- subunit following photoaffinity labeling with 8-Ns['H]cAMP. The pooled fractions containing radioactive peptide are indicated by the arrow. The radioactive peaks that eluted earlier correspond to unincorporated reagent and its breakdown products. No peptide was associated with these fractions.

w U Z

CL

$2 2 02

5 10 15 k 20 25

TIME ( m m )

FIG. 3. Purification of labeled peptide by HPLC. The pooled fractions indicated in Fig. 2 were rechromatographed using a different gradient system. The radioactive fractions indicated were each sequenced separately.

Since this site of modification, differed from what was found in the native R'-subunit, we decided to characterize further the requirements for photolabeling. It was particularly important to establish whether photoaffinity labeling of the immobilized protein on nitrocellulose was specific for either or both CAMP-binding sites. As was indicated in Fig. 1, the type I R-subunit and the mutant R'-subunit can be detected readily and highly selectively in a total cell extract. Protein purification of the R'-subunits was not required. The extract shown here is a total extract from E. coli cells that were expressing high levels of R'-subunit; however, the immobilized R'-subunit also can be detected in mammalian cell extracts with comparable selectivity (data not shown). When the covalent incorporation of 8-N3[32P]cAMP was contrasted to the binding of 8-N3[32P]cAMP, as seen in Fig. 5, the efficiency of covalent modification was found to be very high. Typically, greater than 50% of the 8-N3[32P]cAMP that was bound to

E P !n N" g 9 :0-015J 0.010 -0111

0.005 TYr

10 Tlme (mln)

FIG. 4. Identification of site of covalent modification with S-Ns['H]cAMP. The peptide shown in Fig. 3 was sequenced in its entirety (upper), and the first three steps of that sequence are indicated here. Step 2, which typically shows Tyr in the unmodified protein, gave no identifiable phenylthiohydantoin-derivative.

- R-Subunit

CAMP - FIG. 5. Comparison of binding of 8-Ns['2P]cAMP with co-

valent modification. The R-subunit (4 pg/lane) was electrophoresed, transferred electrophoretically to nitrocellulose, and blocked with Tween 20. The nitrocellulose lanes were then cut into strips and incubated for 30 min at 4 "C with 8-N3[32P]cAMP as indicated under

then irradiated (hu). Unlabeled CAMP M) was then added to "Experimental Procedures." Two of the samples as indicated (+) were

the two lanes shown on the right, and these two samples were incubated for 10 min at 30 "C. The filters were then all washed with cold buffer and exposed to film.

- + +

the nitrocellulose-immobilized protein was covalently incor- porated.

Since the type I and I1 R-subunits are known to vary in the specific sites that are photoaffinity-labeled in solution, photolabeling of the two subunits was compared following electrophoretic transfer to nitrocellulose. The specificity of covalent incorporation is seen in Fig. 6, where the modification of the immobilized R1- and R"-subunits are compared. Even though both of these proteins are modified readily in solution by 8- N3cAMP, only the R'-subunit is efficiently modified following electrophoretic transfer to nitrocellulose.

In order to establish whether the covalent modification of the immobilized protein was site-specific, several variant

18250 Deletion of CAMP-binding Site B - a b c d e a b c d e ,m ',l'."s'q. ~ -

f

RII- j

PROTEIN 8NfcAMP

FIG. 6. Photoaffinity labeling of immobilized type I and I1 R-subunits and proteolytic fragments of R'- and R"-subunits. Samples were elctrophoresed, electrophoretically transferred to nitrocellulose, blocked with Tween 20, and photoaffinity-labeled with 8- N3[32P]~AMP as described under "Experimental Procedures." Lane a, carboxyl-terminal chymotryptic fragment of the R"-subunit; lane b, partial chymotryptic digest of the R"-subunit; lane c, RII-subunit with some endogenous proteolysis; lane d, carboxyl-terminal ther- molytic fragment of the porcine R'-subunit; lane e, porcine R'-subunit with some endogenous proteolysis. The mobilities of the R'-subunit and of the phosphorylated (upper) and dephosphorylated (lower) R"- subunits are indicated by the arrowheads. Staining with Coomassie Blue R-250 is shown on the left; radioactivity is shown on the right.

forms of the R'- and R"-subunits were investigated. In the case of the R"-subunit, affinity labeling of the native protein and of a proteolytically generated fragment of the R"-subunit have been compared previously and have been shown to differ in the modification that is associated with site A (4). As seen in Fig. 6, when these two proteins are immobilized, electrophoretically transferred to nitrocellulose, and photoaffinity- labeled with 8-N3cAMP, neither the proteolytically generated fragment of the R"-subunit nor the native R"-subunit is covalently modified to any significant extent. In contrast, the proteolytic fragment of the R'-subunit, like the native R'- subunit, is photolabeled readily.

In addition to the native R'-subunit, several mutant forms of the R'-subunit have been constructed which show altered CAMP-binding properties. One such mutant, R209/K, contains a point mutation that has abolished CAMP binding to site A, but which retains a functional CAMP-binding site B (9). Another mutant R'-subunit was formed which replaced Tyr371 with Phe (15). This mutation abolished photolabeling of site B. All of these mutant R'-subunits were electrotransferred to nitrocellulose following polyacrylamide gel electrophoresis, blocked with Tween 20, and then photoaffinity- labeled with 8-N3[32P]cAMP. As shown in Fig. 7, the truncated protein and the R'-subunit which did not photolabel in CAMP-binding site B were photolabeled to an extent that was equivalent to the native R'-subunit. On the other hand, mutant R209/K, which lacks a functional site A, was not photolabeled by 8-N3[32P]cAMP.

DISCUSSION

Photoaffinity labeling has proven to be a very specific probe for mapping regions of the CAMP-binding sites in the R- subunits of CAMP-dependent protein kinase. Not only has it enabled us to identify portions of the CAMP-binding sites in the native proteins, but it also appears to be very sensitive to subtle perturbations in the CAMP-binding sites whether in- duced by limited proteolysis or by site-directed mutagenesis of the cloned gene for the R'-subunit. The mutant form of the R'-subunit that is described here has been generated by intro-

. .

Expressed Native FIG. 7. Photoaffinity labeling of immobilized mutant forma

of R'-subunit. Electrophoresis and photoaffinity labeling were car- ried out as indicated for Figs. 1 and 6. Labeling of 4 pg of the native R'- and R"-subunits is included as a control on the right. The expressed R'-subunit (4 pg) is on the far left, followed by the mutant R-subunits (Y371/F, 4 pg and R209/K, 4pg) and the truncated mutant R-subunit (W260/St, 12 pg). The arrows from top to bottom indicate the mobilities of the R"-subunit, the native R'-subunit, and the truncated R-subunit.

INTERACTION DIMER "HINGE"

Site A Site B -AMP Binding Sites +

t5Sw in v,m,,,/lm Truncated Mutant Site A TRP 260-Stop

FIG. 8. Summary of functional sites associated with R'-subunit and truncated R'-subunit. The location of the mutation that leads to the truncated mutant as well as the sites of covalent modification (@) by 8-NscAMP are indicated.

DOMAIN 0 v 371 Ser

378

379 al-COO'

370

362 373

I I FIG. 9. Model of C-helix in site A (upper) and site B (lower)

of R'-subunit. Sites of covalent modification are indicated by arrows. The model of this portion of each CAMP-binding domain of the R- subunit is based on the crystallographic coordinates of the catabolite gene activator protein (17).

ducing a premature stop codon which results in the deletion of the entire CAMP-binding site B, as indicated in Fig. 8. Fortuitously, this mutant R'-subunit lacks both of the residues that are photolabeled efficiently in the native R'-subunit: TrpZBO, which is labeled by 8-N3cAMP bound to site A, and Tyr371 which is labeled by 8-N3cAMP bound to site B. This

Deletion of CAMP-binding Site B 18251

mutant nevertheless retains a single high affinity CAMP- binding site and also is photoaffinity-labeled with 8-NscAMP. The new site of covalent modification was identified as Tyr244.

Models of the CAMP-binding sites in both the R'- and R"- subunits have been constructed by building the R-subunit sequences into the crystallographic coordinates of the homologous CAMP-binding domain of the catabolite gene activator protein (17). These models are consistent with the chemical modifications that have independently identified regions of the R-subunits that are in close proximity to the CAMP- binding sites. Unlike most other adenine nucleotide-binding sites, the CAMP-binding site is a &roll structure (18). One feature of each CAMP-binding site that appears to be essential for interaction of the protein with the adenine ring of cAMP is the long C-helix that extend for nearly the entire length of the domain. In the catabolite gene activator protein, one surface of this cy-helix points inward and lines a portion of the CAMP-binding pocket. In the proposed model of CAMP- binding site B, there is a tyrosine residue located here which is conserved in all of the R-subunits and which is a major residue for covalent modification with 8-N3cAMP (Fig. 9, lower). In the R"-subunit, this is the only residue that is photolabeled. When this Tyr is changed to Phe in the R'- subunit, photolabeling of site B is abolished, as well as the positive cooperativity between the two CAMP-binding sites (9).

In CAMP-binding site A, the position on the C-helix that is analogous to Tyr371 is Phe, and this residue is not photoaffinity-labeled (Fig. 9, upper). Because residue 252 in site A is Pro in the R"-subunit, it has been predicted that this C- helix terminates early in site A (17). The segment which thus extends from SerZ5' to TrpZ6O is some type of random structure that cannot be predicted readily based on the homologies with the catabolite gene activator protein. It is proposed that the end of this segment, Trp2@', which also begins the A-helix of site B, folds over and comes into close proximity to the region near Phe247. This would account for the efficient photoaffinity labeling of TrpZ6O by 8-N3cAMP that is bound to site A. Since the truncated R'-subunit terminates at LysZ5', TrpZ6O is no longer available for covalent modification by 8-N3cAMP. Un- der these conditions, Tyr244 becomes the target for photoaffinity labeling with 8-N3cAMP. Since Tyr244 is predicted to be on the same surface of the C-helix but simply one turn earlier, this observed photolabeling is feasible and consistent with the model. Without a crystal structure, it is still not possible to assess what role, if any, this Tyr plays in either the native or in the truncated mutant R'-subunit. It can only be deduced that it is in close proximity to the adenine ring of 8-NscAMP that is bound to site A.

Photoaffinity labeling with 8-N3cAMP has been used here to map subtle changes in a CAMP-binding site or at least changes which are not sufficient to abolish high affinity binding of CAMP. This method, photolabeling with 8- N~cAMP, also is used frequently to distinguish and identify type I and I1 protein kinases (19, 20). We describe here yet another use for photoaffinity labeling with 8-N3cAMP and that is as a screening mechanism for detecting functional CAMP-binding sites in R-subunits following electrophoretic transfer to nitrocellulose. Antibodies also can be used to identify R-subunits after electrophoretic transfer to nitrocellulose (21). However, immunological detection methods say nothing about the functional state of the protein. Affinity labeling not only identifies the protein with a comparable sensitivity but also reveals that the protein is functional in terms of cAMP binding. This latter information becomes

critical when various mutations are being introduced into the R'-subunit.

Photoaffinity labeling after electrophoretic transfer to nitrocellulose is, however, not applicable to all R-subunits. When the initial photolabeling of the R'-subunit was repeated with the type I1 R-subunit, only nominal incorporation of radioactivity was observed. This led us to speculate that a functional site A which is capable of being photolabeled is required since the native R"-subunit is only covalently modified in site B. Comparison of several variant forms of the R'- subunit is consistent with this hypothesis. For example, mutant R209/K, which does not bind cAMP to site A ( l l ) , also was not modified after the R'-subunit was electrotransferred to nitrocellulose. On the other hand, Y371/F, which has lost the ability to be photolabeled in site B, is still photoaffinity- labeled in addition to the wild-type R'-subunit after electrophoretic transfer to nitrocellulose, again emphasizing the critical need for a functional site A.

On the other hand, the results with the proteolyzed form of the R"-subunit demonstrate that the presence of a functional CAMP-binding site that can be photolabeled in solution is not sufficient to guarantee photolabeling of the immobilized protein. Although site A of the native R"-subunit is not photolabeled in solution, site A is labeled in the proteolyzed form of the R"-subunit. This site, however, apparently is not accessible to the photoaffinity label in the immobilized protein.

The two CAMP-binding sites associated with the R-subunits have been characterized extensively. Initial reports when the R-subunits were first purified indicated that each monomer had only one CAMP-binding site based on membrane filtration assays (22). It was subsequently shown using other methods that there were actually two functional CAMP- binding sites (23, 24). These two sites can be measured by equilibrium dialysis (25), by ammonium sulfate precipitation followed by filtration (26), and by filtration in the presence of high salt and histones (24). These two CAMP-binding sites differ in several significant areas. For example, they differ in their preference for binding various analogs of cAMP (27). These sites also have different dissociation rates, with site A dissociating much faster than site B (28, 29). It is only site B that is detected by the conventional membrane filtration assays and site A which was missed in the early reports (26). The best explanation for this discrepancy is that the R- subunit actually binds to the membrane and that binding to the membrane causes cAMP to dissociate even more rapidly from site A (25). Given these properties of sites A and B, it is somewhat surprising that photolabeling after transfer to nitrocellulose is associated with the fast dissociation site, site A (25). The dissociation rate does not appear to be a critical factor for photolabeling of the immobilized R'-subunit. In- stead, the requirement for a site A that can be labeled suggests that refolding which likely occurs more readily from the amino terminus of the protein may be a more important factor. The carboxyl terminus of site B simply may not refold sufficiently to place the tyrosine that photolabels in solution into the correct orientation with respect to 8-N3cAMP bound to photolabel.

The refolding pathway also could provide an explanation for the results that were seen with the chymotryptic fragment of the R"-subunit, which did not label when the protein was immobilized on nitrocellulose but which did photolabel at site A in solution. This immobilized protein may simply not fold correctly after denaturation when a large portion of the amino terminus is missing. The dissociation rate of cAMP from site A also is much faster in general for the R"-subunit than for the R'-subunit; and this, in contrast to the R'-subunit, may

modified when the protein is immobilized on nitrocellulose. 2418 The results described here do, however, establish that a functional CAMP-binding site A which has accessible sites in close 16. Laemmli, U. K. (1970) Nature 227,680-685

Enzymol. 159,325-336

proximity to the bound 8-N3cAMP is likely to be essential, 17. Weber, I. T., Steitz, T. A., Bubis, J., and Taylor, s. s. (1987) but not sufficient, for photolabeling to occur. 18. Steitz. T. A.. and Weber. I. T. (1985) in Bwkwical Macromolecules

15. Saraswat, L. D., Filutowicz, M., and Taylor, S. S. (1988) Methods

Biochemistry 26,343-351

1.

2. 3. 4. 5.

6.

7.

8.

9.

10.

11.

REFERENCES Kerlavage, A. R., and Taylor, S. S. (1980) J. Biol. Chem. 255,

Bubis, J., and Taylor, S. S. (1985) Biochemistry 24,2163-2170 Bubis, J., and Taylor, S. S. (1987) Biochemistry 2 6 , 3478-3486 Bubis, J., and Taylor, S. S. (1987) Biochemistry 2 6 , 5997-6004 Flockhart, D. A., and Corbin, J. D. (1982) CRC Crit. Rev. Biochem.

Takio, K., Wade, R. D., Smith, S. B., Krebs, E. G., Walsh, K. A., and Titani, K. (1984) Biochemistry 23,4207-4218

Titani, K., Sasagawa, T., Ericsson, L. H., Kumar, S., Smith, S. B., Krebs, E. G., and Walsh, K. A. (1984) Biochemistry 23 ,

Saraswat, L. D., Filutowicz, M., and Taylor, S. S. (1986) J. Bwl.

Bubis, J., Saraswat, L. D., and Taylor, S. S. (1988) Biochemistry

Bubis, J., Neitzel, J., Saraswat, L. D., and Taylor, S. S. (1988) J.

Saraswat, L. D., Ringheim, G. E., Bubis, J., and Taylor, S. S.

8483-8488

15,93-124

4193-4199

Chem. 261,11091-11096

27,1570-1576

Bwl. Chem. 263,9668-9673

(1988) J. Biol. Chem. 263 , 18241-18246

~~ ~ ~~~

andAssemblies: Nucleic Ac& and Interactive Proteins (Jurnak, F. A., and McPherson, A., eds) Vol. 11, pp. 290-321, Wiley, New York

19. Haley, B. E. (1977) Methods Enzyml. 46, 339-346 20. Butley, M. S., and Malkinson, A. M. (1984) J. Bwl. Chem. 2 5 9 ,

21. Mumbv. M. C.. Weldon. S. L., Scott. C. W., and Taylor, S. S. 6001-6007

(198ij Pharnkol. T&r. 28,.367-387 22. Hofmann. F.. Beavo, J. A., Bechtel, P. J., and Krebs, E. G. (1975)

J. Biol. Chkm. 250,7795-7801 23. Weber, W., and Hilz, H. (1979) Bwchem. Biophys. Res. Commun.

24. Corbin, J. D., Sugden, P. H., West, L., Flockhart, D. A., Lincoln, T. M., and McCarthy, D. (1978) J. BwL Chem. 253,3997-4003

25. Builder, S. E., Beavo, J. A., and Krebs, E. G. (1980) J. Biol. Chem.

26. Ogreid, D. (1988) Methods Enzymol., in press 27. Corbin, J. D., Rannels, S. R., Flockhart, D. A., Robinson-Steiner,

A. M., Tigani, M., Doskeland, S. O., Suva, R., and Miller, J. P. (1982) Eur. J. Biochem. 125,259-266

28. Ogreid, D., and Doskeland, S. 0. (1980) FEBS Lett. 121 , 340- 344

29. Rannels, S. R., and Corbin, J. D. (1980) J. Bwl. Chem. 2 5 5 , 7085-7088

90,1073-1081

255,2350-2354

deletion of camp-binding site b in the regulatory subunit of camp

Documents