Structural relationships in the adenylate kinase family

The sequences of five distantly related adenylate kinases have been aligned. The local conservation of amino acids is discussed in the light of the known three-dimensional structure of one of the enzymes, the cytosolic isoenzyme 1 (AK1) from porcine muscle. The similarity profile outlines clearly the active site in the cleft of the spatial structure of AK1. The alignment reveals further that the enzyme family can be subdivided into small and large variants according to the presence or absence of a particular segment of about 30 residues in the middle of the chain. The extra segments of the large variants are strongly conserved.     

Refined structure of porcine cytosolic adenylate kinase at 2.1 A resolution

The crystal structure of porcine cytosolic adenylate kinase has been established at 2.1 A resolution using a restrained least-squares refinement method. Based on 11,251 independent reflections of better than 10 A resolution, a final R-factor of 19.3% was obtained with a model obeying standard geometry within 0.026 A in bond lengths and 3.3 degrees in bond angles. In comparison with the previous structure at 3 A resolution, there is a significant improvement. The high resolution structure has been used to rationalize the strictly conserved residues in the adenylate kinase family. Among these is the glycine-rich loop, which forms a giant anion hole accommodating a sulfate ion which mimics a phosphoryl group of a substrate. Such a structure seems to occur in a large group of mononucleotide binding proteins. Moreover, a conserved cis-proline has been detected in the active center. A structural comparison with the complex between adenylate kinase from yeast and a substrate-analog at medium resolution indicates that this kinase performs appreciable mechanical movements during a catalytic cycle. The reported structure presumably represents an open form of the enzyme, similar to that in solution in the absence of substrates. However, since there are large intermolecular contacts in the crystal, some deviation from the solution structure has to be expected.     

Inhibition of adenylate kinase by P1-(lin-benzo-5'-adenosyl)-P4-(5'-adenosyl) tetraphosphate and P1-(lin-benzo-5'-adenosyl)-P5-(5'-adenosyl pentaphosphate.

P1-(lin-Benzo-5'-adenosyl)-P5-(5'-adenosyl) penraphosphate and P1-(lin-benzo-5'-adenosyl)-P4-(5'-adenosyl) tetraphosphate have been synthesized from lin-benzoadenosine 5'-monophosphoromorpholidate and adenosine 5'-tetraphosphate and adenosine 5'-triphosphate. These mixed dinucleoside polyphosphates are potent inhibitors of porcine muscle adenylate kinase, with association constants of 2 x 10(5) M-1 for the pentaphosphate and 2 x 10(6) M-1 for the tetraphosphate, respectively, as determined by kinetics and fluorescence experiments. The increase in fluorescence intensities and fluorescence lifetimes of both inhibitors upon binding to adenylate kinase results from a breaking of the intramolecular stacking interaction observed when these ligands are free in solution and implicates their binding to the enzyme in an "open" or "extended" form. These results and the dimensional requirements of these inhibitors are discussed in relation to our current knowledge of the active site of adenylate kinase and to the known inhibitors of adenylate kinase, P1,P5-bis(5'-adenosyl) pentaphosphate and P1,P4-bis-(5'-adenosyl) tetraphosphate.     

The amino acid sequence of adenylate kinase from Paracoccus denitrificans and its relationship to mitochondrial and microbial adenylate kinases

The complete amino acid sequence of adenylate kinase (MgATP + AMP in equilibrium MgADP + ADP) from Paracoccus denitrificans has been determined. 1. The S-[14C]carboxymethylated protein was cleaved with clostripain, cyanogen bromide and endoproteinase Lys-C; 18, 9 and 6 fragments, respectively, were analyzed. Some of these peptides were further degraded by trypsin, Staphylococcus aureus V8 protease and carboxypeptidases A and B. The fragments were separated by HPLC and sequenced with a gas-phase sequencer. 2. Sequencing the whole unblocked protein yielded the N-terminal region. The C-terminal residues were obtained by carboxypeptidase-Y digestion in agreement with the sequence of tryptic and cyanogen bromide peptides. 3. The final sequence shows 217 amino acids with Mr = 23,609 and contains one free cysteine and a disulfide bond. 4. The comparison of the P. denitrificans sequence with other known adenylate kinases shows highest similarity with the structurally known Escherichia coli enzyme (47%). The only and catalytically relevant His in the paracoccal enzyme is close to the site of binding of adenosine(5')pentaphospho(5')adenosine to E. coli adenylate kinase. The disulfide bridge is located in the 30-residue segment, which is indicative of the large variants and is absent in cytosolic adenylate kinase. The similarity to the mitochondrial intermembrane-space and matrix adenylate kinase isoenzymes is only 40% and 30%, respectively, while 39% of redidues are identical to those of yeast cytosolic adenylate kinase. Therefore, adenylate kinases do not support the hypothesis of a close relationship between Paracoccus and mitochondria.     

Mechanism of activation of phenylalanine and synthesis of P1, P4-bis(5'-adenosyl) tetraphosphate by yeast phenylalanyl-tRNA synthetase

The activation of L-phenylalanine by yeast phenylalanyl-tRNA synthetase using adenosine 5'-[(S)-alpha-17O,alpha,alpha-18O2]triphosphate is shown to proceed with inversion of configuration at P alpha of ATP. This observation taken together with the lack of positional isotope exchange when adenosine 5'-[beta,beta-18O2]triphosphate is incubated with the enzyme in the absence of phenylalanine and in the presence of the competitive inhibitor phenylalaninol indicates that activation of phenylalanine occurs by a direct "in-line" adenylyl-transfer reaction. In the presence of Zn2+, yeast phenylalanyl-tRNA synthetase also catalyzes the phenylalanine-dependent hydrolysis of ATP to AMP and the synthesis of P1,P4-bis(5'-adenosyl) tetraphosphate (Ap4A). With adenosine 5'-[(S)-alpha-17O,alpha,alpha-18O2]triphosphate, the formation of AMP and Ap4A is shown to occur with inversion and retention of configuration, respectively. It is concluded that phenylalanyl adenylate is an intermediate in both processes, Zn2+ promoting AMP formation by hydrolytic cleavage of the C-O bond and Ap4A formation by displacement at phosphorus of phenylalanine by ATP.     

The switch between two conformations of adenylate kinase

Crystalline adenylate kinase from porcine muscle cytosol can assume two interconvertible structures. Here, we report the refined structure of crystal form B at 3.3 A resolution and compare it with crystal form A. Crystal forms A and B can be interconverted by protonation and deprotonation of His36, which is located deep in the active center cleft. The changes concern the molecular packing as well as the polypeptide chain conformation. On conversion from crystal form A to B, the N-terminal alpha-helix unwinds, the active center cleft opens to some extent and the nucleotide-binding glycine-rich loop 15-22 at the active center is detached from the bulk protein. This loop has counterparts in various important mononucleotide-binding proteins and is known to bind a phosphoryl group in adenylate kinase and in the oncogenic ras proteins. It is most likely involved in the phosphoryl transfer and the concomitant conformational changes. it is suggested that the two observed conformations are relevant for enzyme action in solution: they represent two of a series of three known snapshots depicting the enzyme during the substrate binding process.     

Three-dimensional structure of the complex of guanylate kinase from yeast with its substrate GMP

The enzyme guanylate kinase was isolated from baker's yeast and crystallized as a complex with its substrate GMP. The crystal structure was solved by multiple isomorphous replacement, solvent-flattening, restrained least-squares refinement, and simulated annealing. The current R-factor is 28.9% at a resolution of 2.0 A. The model is given as a backbone tracing, the GMP binding site is shown in atomic detail. In its major domain (residues 1 to 32 and 82 to 186), the chain fold is closely similar to the adenylate kinases, while the minor domain (residues 33 to 81) differs grossly from the 3-helix fold of the adenylate kinases. Structural homology and mechanistical similarity allow us to assign the AMP site of the adenylate kinases on the basis of the GMP site.     

Complexes of yeast adenylate kinase and nucleotides investigated by 1H NMR.

The role of one of the histidine residues present in many adenylate kinases (H36 in the porcine cytosolic enzyme) is highly disputed. We thus studied the yeast enzyme (AKye) containing this His residue. AKye is highly homologous to the Escherichia coli enzyme (AKec), a protein that is already well characterized by NMR [Vetter et al. (1990) Biochemistry 29, 7459-7467] and does not contain the His residue in question. In addition, discrepancies between solution structural and X-ray crystallographic studies on the location of the nucleotide binding sites of adenylate kinases are clarified. One- and two-dimensional nuclear magnetic resonance (NMR) spectroscopy was used to investigate AKye and its complex with the bisubstrate analogue P1,P5-bis(5'-adenosyl)pentaphosphate (AP5A). The well-resolved spectra of AKye allowed identification of nearly all detectable resonances originating from aromatic side chain protons (12 out of 15 spin systems). From these studies, all aromatic residues of AKec involved in the binding of ATP.Mg2+ have functional analogues in AKye. The AMP site seems to make no contacts to aromatic side chains, neither in the AKye.AP5A.Mg2+ nor in the AKec.AP5A.Mg2+ complexes, so that it is presently not possible to localize this binding site by NMR. The ATP site of AKye is located near residues W210 and H143 in a position similar to the ATP site of the E. coli enzyme. In combination with the recent X-ray results on the AP5A complexes AKye and AKec and the GMP complex of guanylate kinase [Stehle, T., & Schultz, G. E. (1990) J. Mol. Biol. 221, 255-269], the latter one leading to the definition of the monophosphate site, the problem of the location of the nucleotide sites can be considered to be solved in a way contradicting earlier work [for a review, see Mildvan, A. S. (1989) FASEB J. 3, 1705-1714] and denying the His residue homologous to H36 in porcine adenylate kinase a direct role in substrate binding.     

Phage display selection of P1 mutants of BPTI directed against five different serine proteinases

The P1 position of protein inhibitors and oligopeptide substrates determines, to a large extent, association energy with many serine proteinases. To test the agreement of phage display selection with the existing thermodynamic data, a small library of all 20 P1 mutants of basic pancreatic trypsin inhibitor (BPTI) was created, fused to protein III, and displayed on the surface of M13 phage. The wild type of displayed inhibitor monovalently and strongly inhibited trypsin with an association constant of Ka = 3 x 10(11) M(-1). The library was applied to select BPTI variants active against five serine proteinases of different specificity (bovine trypsin and chymotrypsin, human leukocyte and porcine pancreatic elastases, human azurocidin). The results of enrichment with four proteinases agreed well with the available thermodynamic data. In the case of azurocidin, the phage display selection allowed determination of the P1 specificity of this protein with the following frequencies for selected P1 variants: 43% Lys, 36% Leu, 7% Met, 7% Thr, 7% Gln.     

Inhibition of carbamyl phosphate synthetase by P1, P5-di(adenosine 5')-pentaphosphate: evidence for two ATP binding sites.

Studies on the effect of a series of alpha, omega-diadenosine 5'-polyphosphate (ApnA; n = 2 to 6) on carbamyl phosphate synthetase showed that only Ap5A is an effective inhibitor. Ap5A also inhibits two partial reactions catalyzed by the enzyme: bicarbonate-dependent ATPase and ATP synthesis from carbamyl phosphate and ADP. The data indicate that Ap5A binds to the enzyme sites that interact with ATP. Of a variety of ATP-utilizing enzymes (kinases, hydrolases, synthetases), only adenylate kinase (Leinhard, G. E., and Secemski, I. I. (1973) J. Biol. Chem. 248, 1121--1123) and carbamyl phosphate synthetase are inhibited by Ap5A. The present findings provide strong evidence that carbamyl phosphate synthetase has two separate binding sites for ATP in which the gamma-phosphate moeities of ATP are bound in close proximity to the bicarbonate binding site of the enzyme.     

Backbone resonance assignments for a homolog of the essential ribosome biogenesis factor Fap7 from P. horikoshii in its nucleotide-free and -bound forms

The protein “factor activating Pos9 (Skn7)”, Fap7, is an essential protein in yeast and plays an important role in the biogenesis of the small ribosomal subunit. In eukaryotes, the final processing step of the small ribosomal subunit RNA is the endonucleolytic cleavage of 20S pre-rRNA at cleavage site D yielding mature 18S rRNA. Depletion of Fap7 in yeast leads to a dramatic accumulation of 20S pre-rRNA and a concomitant decrease in 18S rRNA in the cytoplasm. In addition, these cells contain higher levels of 60S, but decreased numbers of 40S ribosomal subunits. Fap7 contains a P-loop like motif placing it in a class with NTPases and kinases and a role for it as an adenylate kinase has been suggested. Up to now both the structure of Fap7 and its detailed function during ribosome biogenesis remain elusive. Here, we present the backbone NMR assignments of a Fap7 homolog from the thermophilic archaeon Pyrococcus horikoshii in its nucleotide free form and bound to the adenylate kinase inhibitor AP₅A.     

Affinity labeling of nucleotide-binding sites on kinases and dehydrogenases by pyridoxal 5'-diphospho-5'-adenosine

A new adenine nucleotide analog, [3H]pyridoxal 5'-diphospho-5'-adenosine (PLP-AMP), has been synthesized. The effectiveness of PLP-AMP as an affinity probe has been tested using a number of nucleotide-binding enzymes. In comparison to reaction with pyridoxal 5'-phosphate, PLP-AMP binds more tightly and exhibits greater specificity of labeling for most enzymes tested. PLP-AMP is a very potent inhibitor of yeast alcohol dehydrogenase and rabbit muscle pyruvate kinase, with complete inhibition obtained upon incorporation of 1 mol of reagent/mol of catalytic subunit. The reagent is also a potent inhibitor of yeast hexokinase and phosphoglycerate kinase. When modified in the absence of substrates, these enzymes require 2 mol of reagent/mol of active site for complete inhibition. However, when modified in the presence of sugar substrates, this stoichiometry decreases to 1.1 for the hexokinase-glucose complex and 1.4 for the phosphoglycerate kinase . 3-phosphoglycerate complex. The most potent inhibition by PLP-AMP was observed with rabbit muscle adenylate kinase. Half-maximal inhibition was obtained at a concentration of approximately 1 microM. In contrast to these examples, PLP-AMP, as well as pyridoxal 5'-phosphate, fails to act as a potent or specific inhibitor of beef heart mitochondrial F1-ATP-ase. The high specificity of labeling and the ability of nucleotide substrates to decrease the rate of inactivation of the kinases and dehydrogenase are consistent with the modification of active site residues. The complete reversibility of both modification and inactivation in the absence of reduction by NaBH4 and the absorption spectra of modified enzymes prior to and following reduction indicate reaction with lysyl residues. We conclude that PLP-AMP holds considerable promise as an affinity label for exploring the structure and mechanism of nucleotide-binding enzymes.     

High-resolution structures of adenylate kinase from yeast ligated with inhibitor Ap5A, showing the pathway of phosphoryl transfer.

The structure of adenylate kinase from yeast ligated with the two-substrate-mimicking inhibitor Ap5A and Mg2+ has been refined to 1.96 A resolution. In addition, the refined structure of the same complex with a bound imidazole molecule replacing Mg2+ has been determined at 1.63 A. These structures indicate that replacing Mg2+ by imidazole disturbs the water structure and thus the complex. A comparison with the G-proteins shows that Mg2+ is exactly at the same position with respect to the phosphates. However, although the Mg2+ ligand sphere of the G-proteins is a regular octahedron containing peptide ligands, the reported adenylate kinase has no such ligands and an open octahedron leaving space for the Mg2+ to accompany the transferred phosphoryl group. A superposition of the known crystalline and therefore perturbed phosphoryl transfer geometries in the adenylate kinases demonstrates that all of them are close to the start of the forward reaction with bound ATP and AMP. Averaging all observed perturbed structures gives rise to a close approximation of the transition state, indicating in general how to establish an elusive transition state geometry. The average shows that the in-line phosphoryl transfer is associative, because there is no space for a dissociative metaphosphate intermediate. As a side result, the secondary dipole interaction in the alpha-helices of both protein structures has been quantified.     

Resistance to the CCR5 inhibitor 5P12-RANTES requires a difficult evolution from CCR5 to CXCR4 coreceptor use

Viral resistance to small molecule allosteric inhibitors of CCR5 is well documented, and involves either selection of preexisting CXCR4-using HIV-1 variants or envelope sequence evolution to use inhibitor-bound CCR5 for entry. Resistance to macromolecular CCR5 inhibitors has been more difficult to demonstrate, although selection of CXCR4-using variants might be expected. We have compared the in vitro selection of HIV-1 CC1/85 variants resistant to either the small molecule inhibitor maraviroc (MVC) or the macromolecular inhibitor 5P12-RANTES. High level resistance to MVC was conferred by the same envelope mutations as previously reported after 16-18 weeks of selection by increasing levels of MVC. The MVC-resistant mutants were fully sensitive to inhibition by 5P12-RANTES. By contrast, only transient and low level resistance to 5P12-RANTES was achieved in three sequential selection experiments, and each resulted in a subsequent collapse of virus replication. A fourth round of selection by 5P12-RANTES led, after 36 weeks, to a "resistant" variant that had switched from CCR5 to CXCR4 as a coreceptor. Envelope sequences diverged by 3.8% during selection of the 5P12-RANTES resistant, CXCR4-using variants, with unique and critical substitutions in the V3 region. A subset of viruses recovered from control cultures after 44 weeks of passage in the absence of inhibitors also evolved to use CXCR4, although with fewer and different envelope mutations. Control cultures contained both viruses that evolved to use CXCR4 by deleting four amino acids in V3, and others that maintained entry via CCR5. These results suggest that coreceptor switching may be the only route to resistance for compounds like 5P12-RANTES. This pathway requires more mutations and encounters more fitness obstacles than development of resistance to MVC, confirming the clinical observations that resistance to small molecule CCR5 inhibitors very rarely involves coreceptor switching.     

High-resolution X-ray diffraction study of the complex between endothiapepsin and an oligopeptide inhibitor: the analysis of the inhibitor binding and description of the rigid body shift in the enzyme.

The conformation of the synthetic renin inhibitor CP-69,799, bound to the active site of the fungal aspartic proteinase endothiapepsin (EC 3.4.23.6), has been determined by X-ray diffraction at 1.8 A resolution and refined to the crystallographic R factor of 16%. CP-69,799 is an oligopeptide transition--state analogue inhibitor that contains a new dipeptide isostere at the P1-P1' position. This dipeptide isostere is a nitrogen analogue of the well-explored hydroxyethylene dipeptide isostere, wherein the tetrahedral P1' C alpha atom has been replaced by trigonal nitrogen. The inhibitor binds in the extended conformation, filling S4 to S3' pockets, with hydroxyl group of the P1 residue positioned symmetrically between the two catalytic aspartates of the enzyme. Interactions between the inhibitor and the enzyme include 12 hydrogen bonds and extensive van der Waals contacts in all the pockets, except for S3'. The crystal structure reveals a bifurcated orientation of the P2 histidine side chain and an interesting relative rotation of the P3 phenyl ring to accommodate the cyclohexyl side chain at P1. The binding of the inhibitor to the enzyme, while producing no large distortions in the enzyme active site cleft, results in small but significant change in the relative orientation of the two endothiapepsin domains. This structural change may represent the action effected by the proteinase as it distorts its substrate towards the transition state for proteolytic cleavage.     

Three-dimensional structure of the complex between the mitochondrial matrix adenylate kinase and its substrate AMP

Crystals of adenylate kinase from beef heart mitochondrial matrix (EC 2.7.4.10) complexed with its substrate AMP were analyzed by X-ray diffraction. The crystal structure was solved by multiple isomorphous replacement and solvent flattening at a resolution of 3.0 A. There are two enzyme-substrate molecules in the asymmetric unit. The resolution was extended to 1.9 A by model building and refinement using simulated annealing. The current R-factor is 28.4%. The model is given as a backbone tracing for residues 5-218. The enzyme can be subdivided into three domains, the relative arrangements of which differ slightly but significantly between the two crystallographically independent molecules. When compared with other adenylate kinase structures, the chain fold is similar but the observed domain arrangement differs grossly, suggesting that large parts of the enzyme move during catalysis. The observed binding site of AMP is described. Its location in conjunction with data from homologous proteins clarifies the nucleotide-binding sites of the adenylate kinases. Previous assignments of these sites derived from X-ray crystallographic and nuclear magnetic resonance analyses are discussed.     

The adenylate kinases from a mesophilic and three thermophilic methanogenic members of the Archaea.

Adenylate kinase has been isolated from four related methanogenic members of the Archaea. For each, the optimum temperature for enzyme activity was similar to the temperature for optimal microbial growth and was approximately 30 degrees C for Methanococcus voltae, 70 degrees C for Methanococcus thermolithotrophicus, 80 degrees C for Methanococcus igneus, and 80 to 90 degrees C for Methanococcus jannaschii. The enzymes were sensitive to the adenylate kinase inhibitor P1, P5-di(adenosine-5')pentaphosphate, a property that was exploited to purify the enzymes by CIBACRON Blue affinity chromatography. The enzymes had an estimated molecular mass (approximately 23 to 25 kDa) in the range common for adenylate kinases. Each of the enzymes had a region of amino acid sequence close to its N terminus that was similar to the canonical P-loop sequence reported for all adenylate kinases. However, the methanogen sequences lacked a lysine residue that has previously been found to be invariant in adenylate kinases, including an enzyme isolated from the archaeon Sulfolobus acidocaldarius. If verified as a nucleotide-binding domain, the methanogen sequence would represent a novel nucleotide-binding motif. There was no correlation between amino acid abundance and the optimal temperature for enzyme activity.     

X-ray structure of HPr kinase: a bacterial protein kinase with a P-loop nucleotide-binding domain

HPr kinase/phosphatase (HprK/P) is a key regulatory enzyme controlling carbon metabolism in Gram- positive bacteria. It catalyses the ATP-dependent phosphorylation of Ser46 in HPr, a protein of the phosphotransferase system, and also its dephosphorylation. HprK/P is unrelated to eukaryotic protein kinases, but contains the Walker motif A characteristic of nucleotide-binding proteins. We report here the X-ray structure of an active fragment of Lactobacillus casei HprK/P at 2.8 A resolution, solved by the multiwavelength anomalous dispersion method on a seleniated protein (PDB code 1jb1). The protein is a hexamer, with each subunit containing an ATP-binding domain similar to nucleoside/nucleotide kinases, and a putative HPr-binding domain unrelated to the substrate-binding domains of other kinases. The Walker motif A forms a typical P-loop which binds inorganic phosphate in the crystal. We modelled ATP binding by comparison with adenylate kinase, and designed a tentative model of the complex with HPr based on a docking simulation. The results confirm that HprK/P represents a new family of protein kinases, first identified in bacteria, but which may also have members in eukaryotes.     

Isotope-edited proton NMR study on the structure of a pepsin/inhibitor complex

A general approach is illustrated for providing detailed structural information on large enzyme/inhibitor complexes using NMR spectroscopy. The method involves the use of isotopically labeled ligands to simplify two-dimensional NOE spectra of large molecular complexes by isotope-editing techniques. With this approach, the backbone and side-chain conformations (at the P2 and P3 sites) of a tightly bound inhibitor of porcine pepsin have been determined. In addition, structural information on the active site of pepsin has been obtained. Due to the sequence homology between porcine pepsin and human renin, this structural information may prove useful for modeling renin/inhibitor complexes with the ultimate goal of designing more effective renin inhibitors. Moreover, this general approach can be applied to study other biological systems of interest such as other enzyme/inhibitor complexes, ligands bound to soluble receptors, and enzyme/substrate interactions.     

Prostaglandin-E2-induced activation of adenosine 3'-5' cyclic monophosphate-dependent protein kinases of a murine macrophage-like cell line (P388D1)

Changes in the activities of adenosine 3',5'-cyclic monophosphate (cAMP)-dependent protein kinases in response to prostaglandin (PG)E2-induced elevation of intracellular cAMP level were investigated with a murine macrophage-like cell line, P388D1. Photoaffinity labeling with 8-azido-[32P]cAMP showed that untreated P388D1 cells possess two types of cAMP-binding proteins of m.w. 49,000 and 52,000, respectively, in the cytosol fraction in a ration of 1:8. They must represent regulatory subunits (RI and RII, respectively) of cAMP-dependent protein kinases, because affinity chromatography on a column of omega-aminohexyl-agarose of the cytosol fraction clearly separated two fractions that exhibited the enzymatic activities and cAMP-binding activities. Photoaffinity labeling of these fractions with 8-azido-[32P]cAMP confirmed the separation of two types of isoenzymes, because each cAMP-dependent protein kinase active fraction was associated with only one type of regulatory subunit. The exposure of P388D1 cells to exogenously added PGE2 (1 microM) caused about 7.5-fold increase in the intracellular cAMP level within 30 sec. The cAMP level then sharply dropped to about 100 pmol/10(7) cells, remained at this level for about 20 min, and then gradually increased to 200 pmol/10(7) (about fivefold over the control). The enzyme assay of the cytosol demonstrated that the activation of cAMP-dependent protein kinases closely follows the kinetics of the intracellular cAMP level. The activation of the enzyme was specific for PGE2 and was not triggered by 1 microM PGF2 alpha or PGD2 which have been shown to be unable to activate adenylate cyclase of P388D1 cells. The PGE2-induced increase in the intracellular cAMP level appeared to activate preferentially the type I isoenzyme, inasmuch as the enzymatic activity of this type separated by the affinity chromatography of the cytosol of PGE2-exposed cells was lower in the presence than in the absence of cAMP, whereas the type II enzyme activity remained responsive to exogenously added cAMP.