Differential Labeling and Identification of the Cysteine-containing Tryptic Peptides of Catalytic Subunit from Porcine Heart CAMP-dependent Protein Kinase*

The cysteine-containing tryptic peptides from the catalytic subunit of CAMP-dependent protein kinase were isolated and characterized with the aid of the sulfhydryl specific reagent iodoacetic acid. When the catalytic subunit was reacted with [‘4C)iodoacetic acid, cleaved with trypsin, and chromatographed on Sephadex G-50, three distinct components of radioactivity (C-1, C-2, and C-3) were identified. These components were further purified using a combination of high performance liquid chromatography and paper chromatogra- phy. Subsequent analysis revealed that only two of these components ((3-2 and (2-3) contained carboxy- methylcysteine. The sequences of these two peptides are as follows: (C-2) -Cys-Gly-Lys-Glu-Phe-Ser-Glu- Phe; (C-3)”I’hr-Trp-Thr-Leu-Cys-Gly-Asn-Pro-Gln-Tyr-X-Ala-Pro-Glu-ne-ne-Leu-X-Lys. The third com- ponent (C-1) was shown definitely not to contain carboxymethylcysteine. The alkylation of the catalytic subunit with [14C]io-doacetic acid was also studied with respect to inhibition of the enzymatic activity. Incubation of the enzyme with a 1500-fold molar excess of [‘4C]iodoacetic acid at 37 “C resulted in essentially complete loss of phospho- transferase activity within 2 min. the various peptide maps were determined by either counting direct aliquots (up to 0.4 ml) in 5 ml of Bray's solution (180 g of naphthalene, 12 g of PPO, 0.6 g of POPOP, 300 ml of absolute methanol, brought to 3 liters with dioxane) or by drying down aliquots (up to 1.6 ml), redissolving, and counting these in 5 ml of a solution of 15 g of PPO and 0.9 g of POPOP in 3 liters of toluene. All samples were counted in a Beckman LS-233 liquid scintillation system.

Differential Labeling and Identification of the Cysteine-containing Tryptic Peptides of Catalytic Subunit from Porcine Heart CAMPdependent Protein Kinase* (Received for publication, November 14, 1980, and in revised form, December 15, 1980) Norman C. Nelson  The cysteine-containing tryptic peptides from the catalytic subunit of CAMP-dependent protein kinase were isolated and characterized with the aid of the sulfhydryl specific reagent iodoacetic acid. When the catalytic subunit was reacted with ['4C)iodoacetic acid, cleaved with trypsin, and chromatographed on Sephadex G-50, three distinct components of radioactivity (C-1, C-2, and C-3) were identified. These components were further purified using a combination of high performance liquid chromatography and paper chromatography. Subsequent analysis revealed that only two of these components ((3-2 and (2-3) contained carboxymethylcysteine. The sequences of these two peptides are as follows: (C-2) -Cys-Gly-Lys-Glu-Phe-Ser-Glu-
The alkylation of the catalytic subunit with [14C]iodoacetic acid was also studied with respect to inhibition of the enzymatic activity. Incubation of the enzyme with a 1500-fold molar excess of ['4C]iodoacetic acid at 37 "C resulted in essentially complete loss of phosphotransferase activity within 2 min. These conditions resulted in the complete alkylation of C-1, C-2, and C-3. Incubation of the enzyme with a 500-fold molar excess of ["C]iodoacetic acid at 37 "C resulted in essentially no loss of phosphotransferase activity, even after 90 min of incubation. Under these conditions, only C-1 was alkylated, with no significant alkylation of C-2 or C-3 occurring. This eliminates the possibility of C-1 being involved in the phosphotransferase activity of the enzyme. The substrate MgATP protected the enzyme against loss of activity due to modification with iodoacetic acid and blocked the alkylation of not only C-2 and C-3, but also C-1. These data suggest that both cysteine residues contained in the catalytic subunit are located in or near the active site of the molecule.
The dissociated catalytic subunit of CAMP-dependent protein kinase (EC 2.7.1.37) is a monomeric protein having a molecular weight of approximately 40,000 (1-4). It is this dissociated form of the catalytic subunit that is fully active as an ATP-protein phosphotransferase. The holoenzyme, on the other hand, is an inactive tetramer containing two regulatory and two catalytic subunits. Although a t least two major and distinct forms of the holoenzyme exist, the dissociated catalytic subunits from both have been shown to be highly homologous in marked contrast to the regulatory subunits ( 5 ) .
Characterization of the phosphorylation sites on native protein substrates as well as kinetic studies with synthetic peptides (6)(7)(8)(9) have demonstrated certain features that are common to recognition sites for this enzyme. In particular, one and frequently two basic amino acid residues followed by a hydrophobic residue usually precede the site of phosphorylation, which can be either a serine or a threonine hydroxyl group. In native protein substrates, no more than two intervening sequences have been found between the basic residues and the phosphorylated residue.
Various affinity-labeling studies have identified potential residues that may play a role in the catalytic functioning of this enzyme. p-Fluorosulfonylbenzoyl 5"adenosine irreversibly inactivated the catalytic subunit, and a single modified lysine residue was identified (10). Treatment of the catalytic subunit with N*-tosyl-L-lysine chloromethyl ketone was also shown to specifically and irreversibly inactivate the catalytic subunit although no specifically modified amino acid residue was identified (11). Kochetkov et al., using a brain histone kinase, identified a lysine residue that was modified with the dialdehyde analog of ATP (12). Witt and Roskoski utilized acetic anhydride to inhibit the catalytic subunit and suggested the involvement of a tyrosine residue (13).
In addition to these affinity-labeling studies, alkylation with a variety of reagents has indicated that at least one, and possibly two, cysteine residues are situated near the active site of the enzyme (1-3, 11, 14, 15). The present study was designed to identify the cysteine-containing tryptic peptides from porcine heart catalytic subunit, to distinguish the cysteine residues on the basis of reactivity, and to determine which of these cysteine residues was in close proximity to the ATP-binding site on the catalytic subunit.

EXPERIMENTAL PROCEDURES
Assays-Protein kinase was assayed as described previously 14) according to a modification of the precedure of Wastilla et al. (16), using histone type I1 (Sigma) as substrate. One unit of activity represents 1 nmoi of y-"P incorporated into protein per min.
Purification of the Catalytic Subunit of CAMP-dependent Protein Kinase-The catalytic subunit used for these studies was prepared from fresh porcine heart tissue (predominantly type I1 protein kinase) as described previously (5), with minor modifications. The Whatman DE11 column was eluted wi6h a 15-liter gradient from 40 mM phosphate buffer (pH 6.5), to 40 m M phosphate buffer (pH 6.5) +350 m~ NaCl. The pooled kinase peak was immediately diluted to a conductivity of 4 mmho at which point 1.5 liters of DEAE-cellulose (Whatman DE52) was added and the whole was mechanically stirred for 2 h. The supernatant was poured off and the resin was poured into a column and washed extensively with 40 mM phosphate buffer (pH 6.5), followed by extensive washing with 15 mM phosphate buffer (pH 6.1). The catalytic subunit was then eluted with 3 liters of 2.2 X 3743 M CAMP followed by 8 liters of 15 mM phosphate buffer (pH 6.1), and the eluent was passed through an in-tandem cross-linked CM-Sepharose column and eluted as described by Zoller et al. (IO). The typical yield of catalytic subunit from porcine heart is 8 to 12 mg of pure subunit/kg of wet tissue.
High Performance Liquid Chromatography-HPLC' was carried out using an Altex 3200 system with a Waters Clxp-Bondapak column rodt) (pH 2.2) and (b) CHaCN (Fischer, HPLC grade) as described by (0.39 X 30 cm). The buffers employed were: (a) 0.1% H:IP04 (Mallinck-Fullmer et al. (17). The majority of the tryptic peptides were eluted with a 30-min linear gradient from 0 to 50% CH:%CN at a flow rate of 2 ml/min. Peptides were monitored at 219 nm with the use of a Hitachi spectrophotometer equipped with a flow-through cell.
Paper Mapping-For two-dimensional mapping, electrophoresis was carried out at pH 6.5 for 40 min at 3 KV in a high voltage apparatus. Descending chromatography was performed in butanol: acetic acid:water:pyridine (153:12:10) for 16 h. One-dimensional electrophoresis was carried out at either pH 6.5 or pH 1.9 for the times indicated in the text. Peptides were visualized with cadmium:ninhydrin (18) or sequentially with phenanthrenequinone for arginine, Pauly stain for histidine, and then ninhydrin (19). Tryptophan-containing peptides were identified by UV fluorescence. ['4C]Carboxymethylcysteine-containing peptides were identified by autoradiography. Peptides were eluted from paper with 1 to 2 ml of 30% acetic acid. Selected electrophoresis strips were scanned for radioactivity using a Packard model 7200 radiochromatogram scanner.
Sephadex G-50"Peptides were eluted using either 1% acetic acid or 50 mM NH,OH from a column (2.2 cm X 185 em) of Sephadex G-50 fine resin at a flow rate of 30 ml/h, collecting 3-ml fractions.
Alkylation-Complete alkylation of sulfhydryl groups was achieved according to the basic method of Crestfield et al. (20). The protein solution was dialyzed against 100 mM Tris, 0.5 mM EDTA, pH 8.5. The solution was then brought to 6 M in guanidine HCl by the addition of a number of grams of guanidine HC1 equal to the volume of the dialyzed protein solution. The protein was then incubated with a 2-fold excess of dithiothreitol over the total sulfhydryl groups for 2 h at room temperature with stirring under nitrogen and in the dark. A 10-fold excess of [14C]iodoacetic acid over the sulfhydryl groups was added and incubated for an additional 2 h at room temperature with stirring under nitrogen and in the dark. The reaction was terminated by the addition of a 10-fold excess of P-mercaptoethanol and the solution was then dialyzed exhaustively against 50 mM NH,HCO.,, pH 8.5.
Performic Acid Oxidation-Performic acid oxidation was carried out according to Hirs (21). Lyophilized protein was redissolved in 0.3 ml of 88% formic acidmethanol (4:l). A 5% solution of 30% H202 in formic acid was incubated at room temperature for 2 h. 0.1 ml was added to the protein solution and the mixture was incubated for 2 h at -7 "C, then divided into 3 separate aliquots which were dried and hydrolyzed for 24.48, and 72 h. Amino Acid Analyses-Analyses were performed on a Beckman model 118C automatic amino acid analyzer using a single column system. Samples were hydrolyzed in vacuo at 110 "C in 6 N HCI for the times detailed in the text.
Edman procedure (22). Dansyl amino acids were identified by two-Sequencing-Manual sequencing was carried out using the dansyldimensional thin layer chromatography on polyamide sheets (8 X 8 cm; Pierce) according to the method of Hartley (23).
Solid phase sequencing was carried out on a Sequemat automatic sequenator as described by Laursen et al. (24). The peptides were coupled to P-aminopropyl glass (75" pore diameter) that had been activated with p-phenylene diisothiocyanate as described previously (25, 26). The anilinothiazolinone amino acids were converted to phenylthiohydantoin derivatives by incubation with 0.1 N HCI for 10 min at 80 "C followed by extraction with ethyl acetate. The PTHamino acids were subsequently identified by HPLC using the gradient system of Bhown et al. (27).
Cyanogen Bromide Cleavage-Reduced and carboxymethylated protein was dialyzed against water and lyophilized to dryness. The residue was dissolved in 70% formic acid and incubated with a 50-fold molar excess of CNBr (MCB) for 24 h according to Steers et al. (28). The CNBr mixture was diluted 2-fold with water before application to the Sephadex-G50 column.
Radioactivity-Radioactive peaks for the various peptide maps were determined by either counting direct aliquots (up to 0.4 ml) in 5 ml of Bray's solution (180 g of naphthalene, 12 g of PPO, 0.6 g of POPOP, 300 ml of absolute methanol, brought to 3 liters with dioxane) or by drying down aliquots (up to 1.6 ml), redissolving, and counting these in 5 ml of a solution of 15 g of PPO and 0.9 g of POPOP in 3 liters of toluene. All samples were counted in a Beckman LS-233 liquid scintillation system.
Carboxypeptidase Treatment-Treatment with carboxypeptidase was carried out according to Ambler (29). Incubations were carried out at 37 "C for 1 h unless otherwise stated.

RESULTS
Amino Acid Composition-The amino acid composition of protein kinase I1 catalytic subunit isolated from porcine heart was determined on the basis of hydrolyses carried out for 24, 48, and 72 h. In order to quantitate the cysteine content, the protein was fist oxidized with performic acid, thereby converting the cysteines to cysteic acid. The results shown in Table I indicated that each monomeric catalytic subunit contained two to three cysteine residues.
Peptide Mapping-In order to specifically identify the cysteine-containing tryptic peptides, the catalytic subunit was reduced, alkylated with ['4C]iodoacetic acid, and digested with trypsin. The resulting mixture of tryptic peptides was resolved using several different methods. Initial mapping of the tryptic peptides using two-dimensional paper chromatography gave the peptide map indicated in Fig. 1. Autoradiography of this map showed three well resolved radioactive spots designated as (2-1, C-2, and C-3. Differential staining for histidine and arginine indicated that the radioactive spots did not contain either of these two residues. Peptide C-3 was distinguished by its fluorescence and positive Ehrlich stain indicating the presence of tryptophan.
When this mixture of tryptic peptides was subjected to high performance liquid chromatography on a Waters C,"-pBondapak reverse phase column and eluted as described under "Experimental Procedures," the profile shown in Fig. 2 was Based on 40,000 g/mol of molecular weight for catalytic subunit.
Determined from other analyses.  Radioactivity-containing spots are marked C-l, C-2, and C-3, as described in the text. procedures.
NH,-terminal dansylation, amino acid composition, and paper electrophoresis showed that C-l could not be purified using the HPLC method described above. A variety of other methods were therefore used in order to further characterize C-l. Although acid hydrolysis demonstrated that the radioactive pool from HPLC was not homogeneous, glycine was identified as the major component. Paper electrophoresis at pH 1.9 with subsequent radioactivity scanning revealed a r--I----- obtained. When the fractions were measured for radioactivity, three peaks were identified. Electrophoresis of an aliquot of each of these three peaks at pH 6.5 under conditions identical with those used for peptide mapping above identified the three radioactive peaks as C-l, C-2, and C-3 corresponding to the peptide map in Fig. 1.
When the complete mixture of tryptic peptides was chromatographed on Sephadex G-50, the profne shown in Fig. 3 was observed. Again, three well resolved peaks of radioactivity were identified.
Electrophoresis of these radioactive peaks at pH 6.5 identified the fist peak, corresponding to the largest peptide fraction, as C-3, the middle peak as C-2, and the third peak, corresponding to the smallest Sephadex fraction, as Cl. The radioactive peaks were also subjected to HPLC as shown in Fig. 4. The first and second peaks run coincidentally with the peaks previously identified with HPLC (Fig. 2) as C-3 and C-2, respectively.
The third peak of radioactivity from the Sephadex column eluted from HPLC at the same position C-l had been eluted previously, and is not shown in Fig. 4. These data confirm the identification of the radioactive peaks as C-3, C-2, and C-l, respectively, in order of their elution from Sephadex G-50.  Fig. 3 from HPLC. Peak regions were pooled and lyophihzed to near dryness and then eluted from HPLC. Radioactivity was counted as in Fig. 2. Radioactivity was coincident with the HPLC peaks previously identified as C-I, C-2, and C-3 (smallest radioactive fraction eluted at 3% CH.CN, coincident with position of C-I in Fig. 2; data not shown).
radioactive peak that migrated slightly faster than glycine (RF glycine = 0.55, RF radioactive peak = 0.62). When the sample was acid-hydrolyzed and then electrophoresed under the Same conditions, the radioactive peak was coincident with glycine. Chromatography of the sample without hydrolysis revealed a radioactive peak with RF = 0.16 compared to glycine with an RF = 0.17. As stated earlier, the sample ran as a neutral spot on paper electrophoresis at pH 6.5, both with and without hydrolysis. In all cases, there was no radioactivity detected a t the carboxymethylcysteine position. As a control, aliquots of C-2 and C-3 were acid-hydrolyzed and electrophoresed, in which case radioactivity ran coincident with carboxymethylcysteine.
Effect of Alkylation on Catalytic Activity-Iodoacetic acid has been shown to be an effective irreversible inhibitor of the catalytic subunit ( 3 ) . As seen in Fig. 5, incubation of the catalytic subunit with a 500-fold molar excess of iodoacetic acid had no effect on enzymatic activity whereas incubation with 1500-fold excess led to almost immediate inactivation. At intermediate concentrations (1250-fold), a more gradual inhibition was observed, which was essentially complete after 30 min. In the presence of 0.1 mM MgATP, the catalytic subunit was completely protected from irreversible inactivation (Fig.  6). Peters et al., using beef heart catalytic subunit, identified 3 groups that appeared to be sulfhydryl in nature (based on reaction with the sulfhydryl specific reagent DTNB), and that one of these three sulfhydryl groups in that enzyme was very reactive and could be titrated with DTNB with no concomitant loss of activity (2). A similar observation was made by Shaltiel (11). In this study we established that incubation of the porcine catalytic subunit with a 500-fold excess of iodoacetic acid for 30 min resulted in no loss of activity (Fig. 5). In order to establish whether the porcine catalytic subunit also had a reactive sulfhydryl group that could be alkylated with no concomitant loss of activity, the effect of a 500-fold excess of iodoacetic acid on the alkylation of C-1, C-2, and C-3 was determined. As indicated in Fig. 7, if the native catalytic  The NH2-terminal residue as well as internal lysine residues remain coupled to glass.
presence of ATP, no incorporation of radioactivity was observed ( Fig. 9 The reactivities of the cysteine residues contained in C-2 and C-3 towards alkylation with iodoacetic acid were also compared. The differences in reactivity were measured by comparing the per cent alkylation of the two cysteines as a function of time. Alkylation at pH 8.5 yielded essentially no difference in reactivity between C-2 and C-3 (data not shown). At pH 7.5, however, a distinct difference in the reactivities of the two sulfhydryl groups towards iodoacetic acid was observed (Fig. 10). Under the particular conditions of the reaction, the sulfhydryl group contained in C-3 reacted with iodoacetic acid at a higher rate than the sulfhydryl contained in C-2. Differences in per cent alkylation between C-2 and C-3 were observed over the entire time course with the largest difference-68% labeling of C-3 compared to 25% labeling of C-2-occurring at 120 min. Procedures"). An aliquot was eluted using HPLC and 200 ~1 of the indicated tubes were measured for radioactivity as in Fig. 2. acid, an amount insufficient to inactivate the enzyme, resulted in incorporation of radioactivity into only C-l with no significant alkylation of C-2 and C-3 (Fig. 8). These results identified C-l as being the most reactive group on porcine catalytic subunit towards iodoacetic acid alkylation and also eliminated C-l as being directly or peripherally involved in the catalytic activity of the enzyme.
To determine the location of the cysteine residues contained in C-2 and C-3 with respect to the ATP-binding site in catalytic subunit, ['4C]iodoacetic acid alkylation was carried out in the presence of MgATP. As indicated in Fig. 6, MgATP was shown to be effective in protecting the catalytic subunit from inhibition with iodoacetic acid. Therefore, a parallel experiment was run under conditions (cited in Fig. 5) where C-l, C-2, and C-3 were readily alkylated in the native enzyme, this time in the presence of MgATP. Again, no loss of activity was observed as compared to the control without iodoacetic acid. Furthermore, whereas, in the absence of ATP, all activity was lost and three peaks of radioactivity were observed, in the Cyanogen Bromide Cleavage of the Catalytic Subunit- The catalytic subunit of procine CAMP-dependent protein kinase contains 6 methionine residues. When the protein was reduced, carboxymethylated, cleaved with cyanogen bromide, and subjected to gel filtration, the results seen in Fig. 11 were obtained. Seven cyanogen bromide peptides were obtained: I, II, IV, III, A and B, and V, A and B. Fractions III and V from the Sephadex G-50 column were each resolved into two components, A and B (Fig. 12), following HPLC carried out according to Fullmer and Wasserman (17). When the protein was alkaylated with ['?]iodoacetic acid prior to cyanogen bromide cleavage, the radioactivity was associated either with peaks I and II or eluted at the included volume of the column. CNBr I was a large fragment (>lOO residues) having an NH2terminal phenylalanine. Digestion of CNBr I with trypsin followed by HPLC revealed that C-3 was located in this

fragment.
CNBr II was identified as the COOH-terminal cyanogen bromide fragment, by the absence of homoserine in the amino acid composition.
In addition, treatment with carboxypeptidases A and B released phenylalanine. On sodium dodecyl sulfate-urea gels run according to the method of Swank et al. (30), this CNBr fragment had a molecular weight of 12,000 to 13,000. Tryptic digestion coupled with HPLC revealed that this fragment contained C-2, which was also identified as the COOH-terminal tryptic peptide by the release of phenylalanine with carboxypeptidase A and B treatment. When an aliquot of the third radioactive peak was run on HPLC, all the radioactivity eluted at the position C-l had previously been shown to elute. When an aliquot was electrophoresed at pH 6.5, the radioactivity ran with a fluorescaminepositive, neutral spot. This spot was eluted and then electrophoresed at pH 1.9, either with or without prior hydrolysis. In both cases, the radioactivity ran coincident with glycine. An aliquot of this eluted sample was also dansylated (see "Exper-  imental Procedures") and then chromatographed, again either with or without prior hydrolysis. In both cases, glycine was the major dansyl spot and autoradiography identified this glycine spot as the only radioactive component on the plate. This third peak of radioactivity was further characterized by running a hydrolyzed aliquot on the amino acid analyzer, collecting the eluent in fractions, and counting these fractions for radioactivity (the eluent was not allowed to react with ninhydrin as the ninhydrin solution quenches radioactivity; known retention times were used to identify any peaks). The only radioactive peak eluted at the glycine position.
In a different CNBr digest, when an aliquot of the third radioactive peak (corresponding to C-1) was electrophoresed at pH 1.9, radioactivity was detected only slightly off the origin (RF = 0.045) with none being detected at the glycine position. The difference between digests was that in this case the enzyme was in 70% formic acid for 48 h. as compared to the usual 24 h.

DISCUSSION
The number as well as the role of the cysteine residues in the catalytic subunit of CAMP-dependent protein kinase has been somewhat unclear. On the basis of amino acid composition alone, a value of 2 to 3 cysteines per catalytic subunit has been reported in most cases (l- 3,11,14). An exception is the bovine liver enzyme where Sugden et al. (3) report only 1; however, this value may be low in that bovine heart enzyme contains 2 cysteines (31). In the case of the catalytic subunit from porcine heart type II CAMP-dependent protein kinase, amino acid analysis of the performic acid-oxidized protein also indicated a value of 2 to 3 cysteine residues/subunit. Further evidence indicating that there were 3 cysteines/subunit was derived from mapping of the tryptic peptides of ['4C]iodoacetic acid-reacted protein, where three radioactive components were observed. These three components were clearly separated from one another using several different chromatographic techniques and, in addition, were shown to be distinct on the basis of properties characterized by differential labeling experiments. However, when these three components were isolated and characterized, only two, C-2 and C-3, were found to be carboxymethylcysteine-containing peptides, thus indicating a value of two cysteine residues/catalytic subunit. This number is in agreement with the reports of Kaiser et al. (14), Titani et al. (31), and Hart1 (15), but conflicts with the reports of Peters et al. (2) and Shaltiel et al. (11). These groups indicate the presence of a third, very reactive cysteine residue which can be titrated with sulfhydryl specific reagents. This reactive group most likely corresponds to the third radioactive component identified in this study, C-1, which was also very reactive towards a sulfhydryl specific reagent (iodoacetic acid in this case), but was shown not to contain carboxymethylcysteine.
In order to resolve the ambiguity as to the identity of the three groups on catalytic subunit reactive toward sulfhydryl specific reagents, differential labeling was coupled with isolation and characterization of tryptic peptides from [I4C]iodoacetic acid reacted protein. It was shown that C-1 could be alkylated with a 10-fold molar excess of ['4C]iodoacetic acid without C-2 or C-3 reacting. C-1 therefore apparently accounts for the very reactive residue shown by Shaltiel et al. (11), under the conditions of their experiment, to react with a rate constant of greater than 10,000 M" min". In addition, since a 500-fold molar excess of iodoacetic acid was shown to have no effect on the activity of the enzyme and alkylation of C-1 requires only a 10-fold molar excess of iodoacetic acid, loss of activity is not associated with the alkylation of C-1. This is in agreement with the results previously reported by Peters et al. (2) and Shaltiel et al. (11).
Iodoacetic acid was shown to be an effective irreversible inhibitor of catalytic activity, a 1500-fold molar excess essentially completely inhibiting the enzyme within just a few minutes. This inactivation with iodoacetic acid was completely protected against with the enzyme substrate MgATP. T o differentiate between C-2 and C-3 with respect to inactivation, the enzyme was reacted with a large excess of iodoacetic acid (which was shown by itself to label C-1, C-2, and C-3 in the native molecule) in the presence of MgATP. Not only were both C-2 and C-3 protected from alkylation, but surprisingly so was C-1. It is possible that all three reactive groups are close enough in proximity in the tertiary structure that the MgATP sterically hinders their reaction with iodoacetic acid, or that binding of ATP leads to conformational changes in the molecule that might significantly deter the reactivity of one or more of these residues.
C-2 and C-3 were further differentiated on the basis of reactivity in the native enzyme towards alkylation with iodoacetic acid. The sulfhydryl contained in C-3 was shown to react faster with iodoacetic acid than the sulfhydryl contained in C-2 at pH 7.5 and a conductivity around 1 mmho. Essentially no difference in reactivities was observed at pH 8.5. This difference in reactivities between the two sulfhydryl groups suggests the possibility that one sulfhydryl group could be completely alkylated without significant alkylation of the other. Such an alkylation pattern could provide a method for determining the role of each sulfhydryl group in the inactivation of the catalytic subunit upon alkylation with iodoacetic acid, i.e. if only one sulfhydryl group is responsible for the inactivation, or if, indeed, both sulfhydryl groups have to be alkylated to achieve complete inactivation of the enzyme. Experiments of such a nature are currently in progress.
The ambiguity as to the actual number of cysteine residues in the catalytic subunit of CAMP-dependent protein kinase can be resolved with the evidence presented here. There are indeed, three groups on the molecule that react with SO called "sulfhydryl specific reagents," but further characterization has shown that only two of these groups are cysteine residues. The third group, referred to here as C-1, is very reactive towards iodoacetic acid in both the native and the denatured enzyme. However, the fact that the reactive portion of this group is not the sulfhydryl of a cysteine residue has been clearly demonstrated in this study. The fact that neither exhaustive dialysis nor chromatography on Sephadex G-25 removed the radioactivity associated with C-1 from the protein indicates that the iodoacetic acid reacts with a group that is actually bound to the molecule. The exact nature of the reactive group on the molecule is not known at this point.
The two cysteines were also located in the primary sequence of the molecule. Carboxypeptidase treatment of (2-2 revealed this to be the COOH-terminal tryptic peptide, thus localizing one of the cysteines in close proximity to the COOH terminus of the molecule. A peptide homologous to C-3 from beef heart catalytic subunit has previously been localized in the central region of the linear sequence (32).
The fact that small sulfhydryl specific alkylating reagents such as KCN did not inactivate the enzyme (2) indicated it is likely that rather than cysteine being an integral part of the ATP binding site, addition of a large enough group onto cysteine simply sterically hinders the binding of ATP. This seems to be the case for creatine kinase (33). Also, Schwartz et al. (34), showed that inactivation of rabbit phosphofructokinase by reaction of cysteine residues was not due to direct active site modification.
The fact that the regulatory subunit of CAMP-dependent protein kinase protects the catalytic subunit from inactivation with DTNB has been shown (14). On the other hand, studies with protein kinase inhibitor (which acts competitively with respect to protein substrates) showed that inactivation by reaction of the sulfhydryl groups with iodoacetamide was not due to an alteration in the protein binding site (2). The method of differential labeling described here is presently being used to determine the effect of peptides as well as regulatory subunit on the reactivity of specific sulfhydryl groups.