Probing cAMP-dependent protein kinase holoenzyme complexes I alpha and II beta by FT-IR and chemical protein footprinting

Although individual structures of cAMP-dependent protein kinase (PKA) catalytic (C) and regulatory (R) subunits have been determined at the atomic level, our understanding of the effects of cAMP activation on protein dynamics and intersubunit communication of PKA holoenzymes is very limited. To delineate the mechanism of PKA activation and structural differences between type I and II PKA holoenzymes, the conformation and structural dynamics of PKA holoenzymes Ialpha and IIbeta were probed by amide hydrogen-deuterium exchange coupled with Fourier transform infrared spectroscopy (FT-IR) and chemical protein footprinting. Binding of cAMP to PKA holoenzymes Ialpha and IIbeta leads to a downshift in the wavenumber for both the alpha-helix and beta-strand bands, suggesting that R and C subunits become overall more dynamic in the holoenzyme complexes. This is consistent with the H-D exchange results showing a small change in the overall rate of exchange in response to the binding of cAMP to both PKA holoenzymes Ialpha and IIbeta. Despite the overall similarity, significant differences in the change of FT-IR spectra in response to the binding of cAMP were observed between PKA holoenzymes Ialpha and IIbeta. Activation of PKA holoenzyme Ialpha led to more conformational changes in beta-strand structures, while cAMP induced more apparent changes in the alpha-helical structures in PKA holoenzyme IIbeta. Chemical protein footprinting experiments revealed an extended docking surface for the R subunits on the C subunit. Although the overall subunit interfaces appeared to be similar for PKA holoenzymes Ialpha and IIbeta, a region around the active site cleft of the C subunit was more protected in PKA holoenzyme Ialpha than in PKA holoenzyme IIbeta. These results suggest that the C subunit assumes a more open conformation in PKA holoenzyme IIbeta. In addition, the chemical cleavage patterns around the active site cleft of the C subunit were distinctly different in PKA holoenzymes Ialpha and IIbeta even in the presence of cAMP. These observations provide direct evidence that the R subunits may be partially associated with the C subunit with the pseudosubstrate sequence docked in the active site cleft in the presence of cAMP.     

Expression and characterization of mutant forms of the type I regulatory subunit of cAMP-dependent protein kinase. The effect of defective cAMP binding on holoenzyme activation

The mouse wild type and four mutant regulatory type I (RI) subunits were expressed in Escherichia coli and subjected to kinetic analyses. The defective RI subunits had point mutations in either cAMP-binding site A (G200/E), site B (G324/D, R332/H), or in both binding sites. In addition, a truncated form of RI which lacked the entire cAMP-binding site B was generated. All of the mutant RI subunits which bound [3H]cAMP demonstrated more rapid rates of cAMP dissociation compared to the wild type RI subunit. Dissociation profiles showed only a single dissociation component, suggesting that a single nonmutated binding site was functional. The mutant RI subunits associated with purified native catalytic subunit to form chromatographically separable holoenzyme complexes in which catalytic activity was suppressed. Each of these holoenzymes could be activated but showed varying degrees of cAMP responsiveness with apparent Ka values ranging from 40 nM to greater than 5 microM. The extent to which the mutated cAMP-binding sites were defective was also shown by the resistance of the respective holoenzymes to activation by cAMP analogs selective for the mutated binding sites. Kinetic results support the conclusions that 1) Gly-200 of cAMP-binding site A and Gly-324 or Arg-332 of site B are essential to normal conformation and function, 2) activation of type I cAMP-dependent protein kinase requires that only one of the cAMP-binding sites be functional, 3) mutational inactivation of site B (slow exchange) has a much more drastic effect than that of site A on increasing the Ka of the holoenzyme for cAMP, as well as in altering the rate of cAMP dissociation from the remaining site of the free RI subunit. The strong dependence of one cAMP-binding site on the integrity of the other site suggests a tight association between the two sites.     

Predicted structures of cAMP binding domains of type I and II regulatory subunits of cAMP-dependent protein kinase

The mammalian cAMP-dependent protein kinases have regulatory (R) subunits that show substantial homology in amino acid sequence with the catabolite gene activator protein (CAP), a cAMP-dependent gene regulatory protein from Escherichia coli. Each R subunit has two in-tandem cAMP binding domains, and the structure of each of these domains has been modeled by analogy with the crystal structure of CAP. Both the type I and II regulatory subunits have been considered, so that four cAMP binding domains have been modeled. The binding of cAMP in general is analogous in all the structures and has been correlated with previous results based on photolabeling and binding of cAMP analogues. The model predicts that the first cAMP binding domain correlates with the previously defined fast dissociation site, which preferentially binds N6-substituted analogues of cAMP. The second domain corresponds to the slow dissociation site, which has a preference for C8-substituted analogues. The model also is consistent with cAMP binding in the syn conformation in both sites. Finally, this model has targeted specific regions that are likely to be involved in interdomain contacts. This includes contacts between the two cAMP binding domains as well as contacts with the amino-terminal region of the R subunit and with the catalytic subunit.     

Activation of type I cyclic AMP-dependent protein kinases with defective cyclic AMP-binding sites

Two S49 mouse lymphoma cell variants hemizygous for expression of mutant regulatory (R) subunits of type I cyclic AMP-dependent protein kinase were used to investigate functional consequences of lesions in the putative cAMP-binding sites of R subunit. Kinase activation properties of wild-type and mutant enzymes were compared using cAMP and six site-selective analogs of cAMP. Kinases from both mutant sublines were relatively resistant to cyclic nucleotide-dependent activation, but they were fully activable by at least some effectors. Relative resistances of the mutant kinases varied from about 5-fold for analogs selective for their nonmutated sites to as much as 700-fold for analogs selective for their mutated sites; resistance to cAMP was intermediate. Apparent affinities of wild-type and mutant R subunits for [3H]cAMP were not appreciably different, but competition experiments with site-selective analogs of cAMP suggested that binding of cAMP to mutant R subunits was primarily to their nonmutated sites. Analyses of cooperativity in cyclic nucleotide-dependent activation of mutant kinases, synergism between site I- and site II-selective analogs in activating the mutant enzymes, and dissociation of bound cAMP from mutant R subunits provided additional evidence that the mutations in these strains selectively inactivated single classes of cAMP-binding sites: phenomena attributable in wild-type enzyme to intrachain interactions between sites I and II were always absent or severely diminished in experiments with the mutant enzymes. These results confirm that R subunit sequences implicated in cAMP binding by homology with other cyclic nucleotide-binding proteins actually correspond to functional cAMP-binding sites. Furthermore, occupation of either cAMP-binding site I or II is apparently sufficient for activation of cAMP-dependent protein kinase. The presence of four functional cAMP-binding sites in wild-type kinase enhances the cooperativity and sensitivity of cAMP-mediated activation.     

Properties of cyclic AMP-independent catabolite gene activator proteins of Escherichia coli

The protein products of two crp alleles encoding mutationally altered catabolite gene activator proteins CAP and CAPc, which are functionally active in vivo in the absence of cAMP, were purified by an immunoaffinity purification procedure. These proteins bind cAMP with the same affinity as does the wild-type catabolite gene activator protein. From their susceptibility to the proteolytic enzyme subtilisin, we conclude that the two mutationally altered proteins adopt structural features adequate for biological activity and similar to the conformation that cAMP elicits or stabilizes in wild-type catabolite gene activator protein. We note, however, that their conformation is not unique and can be modulated by cAMP. The two altered proteins, CAP and CAPc, bind to the lactose promoter, giving rise to specific DNA-protein complexes in the absence of cAMP and promote initiation of specific lac messenger RNA synthesis.     

An increase of cAMP-dependent protein kinase during development in Polysphondylium pallidum

Polysphondylium pallidum is a cellular slime mold in which, unlike in Dictyostelium discoideum, cAMP is not the chemotactic agent. The occurrence of a cAMP-dependent protein kinase in D. discoideum was demonstrated earlier and we suggested that it may mediate the intracellular effects of cAMP on the development of the organism, particularly since an increase in the amount of the enzyme during development was noted. In D. discoideum cAMP plays a dual role insofar as it serves both as chemotactic agent and as second messenger; it was of interest therefore, to determine whether a cAMP-dependent protein kinase occurred in P. pallidum. We found a cAMP-dependent protein kinase in P. pallidum using Kemptide as substrate. The regulatory subunit of the enzyme has an apparent molecular weight of 41,000 and seems to be similar in its properties with that isolated earlier from D. discoideum. The cAMP-dependent protein kinase catalytic subunits from the two species are also similar. Furthermore, there is a developmentally regulated, parallel, two- to threefold increase in the two subunits of the cAMP-dependent protein kinase in P. pallidum. The increase occurs before aggregates are formed. These findings are compatible with a role of the intracellular cAMP and of the cAMP-dependent protein kinase in the development of P. pallidum.     

How does EcoRI cleave its recognition site on DNA?

The two protein subunits of the EcoRI restriction enzyme interact symmetrically with the recognition site on DNA, so that each subunit is in position to cleave one strand of the DNA. But each subunit seems to require a protein conformation change before it can cleave DNA. Depending upon whether one or both subunits change conformation during the life-time of the enzyme-DNA complex, a single reaction of the EcoRI enzyme cleaves either one or both strands of the DNA. Reaction profiles with other restriction enzymes differ from EcoRI, though the underlying mechanisms may be the same.     

Kinetic and structural studies of the allosteric conformational changes induced by binding of cAMP to the cAMP receptor protein from Escherichia coli

The cAMP receptor protein, allosterically activated by cAMP, regulates the expression of more than 100 genes in Escherichia coli. CRP is a homodimer of two-domain subunits. It has been suggested that binding of cAMP to CRP leads to a long-distance signal transduction from the N-terminal cAMP binding domain to the C-terminal domain of the protein responsible for interaction with specific sequences of DNA. In this study, the stopped-flow and time-resolved fluorescence lifetime measurements were used to observe the kinetics of the distance changes between the N-terminal and C-terminal domain of CRP induced by binding of cAMP to high-affinity binding sites. In these measurements, we used the constructed CRP heterodimer, which possesses a single Trp85 residue localized at the N-terminal domain of one CRP subunit, and fluorescently labeled by 1,5-I-AEDANS Cys178 localized at the C-terminal domain of the same subunit or at the opposite one. The F?rster resonance energy transfer method has been used to study the distance changes, induced by binding of cAMP, between Trp85 (fluorescence donor) and Cys178-AEDANS (fluorescence acceptor) in the CRP structure. The obtained results show that the allosteric transitions of CRP at micromolar cAMP concentrations follow the sequential binding model, in which binding of cAMP to high-affinity sites causes a 4 A movement of the C-terminal domain toward N-terminal domains of the protein, with kinetics faster than 2 ms, and CRP adopts the "closed" conformation. This fast process is followed by the slower reorientation of both CRP subunits.     

Polyamines and basic proteins stimulate activation by cAMP and catalytic activity of Mucor rouxii cAMP-dependent protein kinase

Partial activation of Mucor rouxii cAMP-dependent protein kinase by cAMP was obtained when kemptide was used as substrate, but complete activation was attained with cAMP plus protamine or histone. Full activation could not be achieved by increasing kemptide or cAMP concentration. Complete activation by cAMP could be obtained by addition of 10 microM polylysine, 10 microM lysine-rich histone or 0.5 mM spermine plus spermidine. The degree of stimulation could be up to 5-fold, depending on the amount of enzyme in the assay. The same concentrations of polycations increased 1.5-2.3-fold the Vmax of kemptide phosphorylation by the free catalytic subunits of both Mucor and bovine heart protein kinases; 10 microM polyarginine inhibited completely the activity of both enzymes.     

Studies on the subunits of Escherichia coli coenzyme A transferase. Reconstitution of an active enzyme.

The alpha and beta subunits of the acetyl-CoA:acetoacetate-CoA transferase were purified by isoelectric focusing of the enzyme in the presence of 6 M urea. The purified beta subunit, in which the active center of the enzyme is located, exhibits low catalytic activity (2% of the specific activity of the native enzyme) which is stimulated 5-6-fold in the presence of an equimolar concentration of alpha subunit. The presence of the substrate,acetoacetyl-CoA, is required to recover the catalytic activity of the beta subunit and mixtures containing purified alpha and beta subunits. When the enzyme is dissociation in the presence of 6 M urea and the subunits are not fractioned, removal of the urea by dialysis results in the recovery of 88-98% of enzymic activity and the native alpha2beta2 subunit structure. However, analysis of this renatured enzyme by immunochemical techniques shows that the enzyme does not refold to a completely native conformation. This renatured enzyme exhibits an immunological reactivity more closely resembling the isolated alpha subunit. The results indicate that the alpha subunit serves as a structural subunit, or possible a maturation subunit, imposing a conformation on the beta subunit that is catalytically more competent.     

Ligand-induced conformational changes in the Escherichia coli F1 adenosine triphosphatase probed by trypsin digestion

Digestion of the F1-ATPase of Escherichia coli with trypsin stimulated ATP hydrolytic activity and removed the delta and epsilon subunits of the enzyme. A species represented by the formula alpha 1(3) beta 1(3) gamma 1, where alpha 1, beta 1 and gamma 1 are forms of the native alpha, beta and gamma subunits which have been attacked by trypsin, was formed by trypsin digestion in the presence of ATP. In the presence of ATP and MgCl2, conversion of gamma to gamma 1 was retarded and the enzyme retained the epsilon subunit. These results imply that binding of ATP to the beta subunits alters the conformation of ECF1 to increase the accessibility of the gamma subunit to trypsin. The likely trypsin cleavage sites in the alpha, beta and gamma subunits are discussed. ECF1 from the alpha subunit-defective mutant uncA401, or after treatment with N,N'-dicyclohexylcarbodiimide or 4-chloro-7-nitrobenzofurazan, was present in a conformation in which the gamma subunit was readily accessible to trypsin and could not be protected by the presence of ATP and MgCl2. In a similar manner to native E. coli F1-ATPase, the hydrolytic activity of the trypsin-digested enzyme was stimulated by the detergent lauryldimethylamine N-oxide. Since the digested enzyme lacked the epsilon subunit, a putative inhibitor of hydrolytic activity, a mechanism for the stimulation which involves loss or movement of this subunit is untenable.     

Purification and characterization of an inactive form of cAMP-dependent protein kinase containing bound cAMP

By a new procedure, the holoenzyme of bovine heart type II cAMP-dependent protein kinase was purified to homogeneity as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). A high performance liquid chromatography-DEAE purification step resolved two distinct peaks of protein kinase activity, which were designated Peak 1 and Peak 2 based on their order of elution. The two peaks exhibited similar Stokes radii and sedimentation coefficients. They had similar ratios of regulatory to catalytic subunits both by densitometric scanning of SDS-PAGE bands and by the ratios of equilibrium [3H]cAMP binding to maximal kinase activity. These results suggested that the holoenzyme of each peak contained two regulatory subunits and two catalytic subunits, although a subpopulation of holoenzyme lacking one catalytic subunit also appeared to be present in Peak 2. Assays of cAMP indicated that the Peak 1 holoenzyme was cAMP-free, but half of the Peak 2 holoenzyme cAMP binding sites contained cAMP. Determination of [3H]cAMP dissociation rates showed that the cAMP was equally distributed in binding Site 1 and Site 2 of Peak 2. Although SDS-PAGE analysis ruled out conversions by proteolysis or autophosphorylation-dephosphorylation, Peak 1 could be partially converted to Peak 2 by the addition of subsaturating amounts of cAMP. Interconvertibility of the two holoenzyme peaks strongly suggested that the difference between the two peaks was caused by the presence of cAMP in Peak 2. Peak 2 holoenzyme, as compared to Peak 1, had enhanced binding in nonequilibrium [3H]cIMP and [3H]cAMP binding assays, as was expected due to the presence of cAMP and to the known positive cooperativity in binding of cyclic nucleotides to the kinase. The positive cooperativity in kinase activation, as indicated by the Hill coefficient, was greater for Peak 2 than Peak 1, but the cAMP concentration required for half-maximal activation (Ka) of each of the two peaks was very similar. In conclusion, Peak 2 is an inactive ternary complex of cAMP, regulatory subunit, and catalytic subunit, and Peak 1 is a cAMP-free holoenzyme. The cAMP-bound form may represent a major cellular form of the enzyme which is primed for activation.     

Probing conformations of the beta subunit of F0F1-ATP synthase in catalysis

A subcomplex of F0F1-ATP synthase (F0F1), alpha3beta3gamma, was shown to undergo the conformation(s) during ATP hydrolysis in which two of the three beta subunits have the "Closed" conformation simultaneously (CC conformation) [S.P. Tsunoda, E. Muneyuki, T. Amano, M. Yoshida, H. Noji, Cross-linking of two beta subunits in the closed conformation in F1-ATPase, J. Biol. Chem. 274 (1999) 5701-5706]. This was examined by the inter-subunit disulfide cross-linking between two mutant beta(I386C)s that was formed readily only when the enzyme was in the CC conformation. Here, we adopted the same method for the holoenzyme F0F1 from Bacillus PS3 and found that the CC conformation was generated during ATP hydrolysis but barely during ATP synthesis. The experiments using F0F1 with the epsilon subunit lacking C-terminal helices further suggest that this difference is related to dynamic nature of the epsilon subunit and that ATP synthesis is accelerated when it takes the pathway involving the CC conformation.     

Immobilised monomers of human liver arginase.

Human liver arginase (L-arginine amidinohydrolase, EC 3.5.3.1) was immobilised by attachment to nylon with glutaraldehyde as a crosslinking agent. Incubation of the immobilised tetrameric enzyme with EDTA followed by dialysis resulted in the dissociation of the enzyme into inactive matrix-bound and solubilised subunits. Both species recovered enzymatic activity after incubation with Mn2+, and the activity of the reactivated matrix-bound subunits was nearly 25% of that shown by the enzyme initially attached to the support in the tetrameric form. When the reactivated bound subunits were incubated with soluble subunits in the presence of Mn2+, they 'picked-up' from the solution an amount of protein and enzymatic activity almost identical to that initially lost by the immobilised tetramer after the dissociating treatment with EDTA. This occurred only in the presence of Mn2+. It is suggested that the reactivation of the subunits of arginase involves the initial formation of an active monomer, which then acquires a conformation that favours a reassociation to the tetrameric state.     

Subunits of RNA polymerase in function and structure. 7. Structure of premature core enzyme.

The structure of premature core enzyme, an obligatory intermediate in both in vivo and in vitro assembly of Escherichia coli DNA-dependent RNA polymerase, was compared with that of native core enzyme. Though this assembled but inactive form of core enzyme harbors the gross conformation similar to that of native enzyme, minor and presumably local differences exist, which were identified by near-ultraviolet circular dichroism spectra, tritium-hydrogen exchange rate, protease sensitivity, intersubunit cross-linking rate by bifunctional reagents, sedimentation behavior, and elution profile from phosphocellulose. Taken together these results indicate that the core enzyme subunits are loosely associated in the premature core. The temperature-dependent maturation is required for the core subunits to be tightly associated, leading to the formation of structurally stable and functionally active RNA polymerase.     

Functional multiplicity of phosphoglucose isomerase from Lactobacillus casei

Phosphoglucose isomerase (D-glucose-6-phosphate ketolisomerase, EC 5.3.1.9), purified from Lactobacillus casei, showed multiplicity with respect to electrophoretic mobility, molecular weight, kinetic properties and responses to erythrose 4-phosphate. Among the three forms isolated, one having a dimeric conformation, was specific for glucose 6-phosphate. Erythrose 4-phosphate inhibited this preparation in a sigmoid fashion, while this compound activated the enzyme for isomerization of ribose 5-phosphate. In tetrameric conformation of the similar subunits, the enzyme was more specific for ribose 5-phosphate and the inhibition exerted by erythrose 4-phosphate was hyperbolic. The possible implications of these observations have been discussed.     

Closed conformation of the active site loop of rabbit muscle triosephosphate isomerase in the absence of substrate: evidence of conformational heterogeneity

The active site loop of triosephosphate isomerase (TIM) exhibits a hinged-lid motion, alternating between the two well defined "open" and "closed" conformations. Until now the closed conformation had only been observed in protein complexes with substrate analogues. Here, we present the first rabbit muscle apo TIM structure, refined to 1.5A resolution, in which the active site loop is either in the open or in the closed conformation in different subunits of the enzyme. In the closed conformation described here, the lid loop residues participate in stabilizing hydrogen bonds characteristic of holo TIM structures, whereas chemical interactions observed in the open loop conformation are similar to those found in the apo structures of TIM. In the closed conformation, a number of water molecules are observed at the projected ligand atom positions that are hydrogen bonded to the active site residues. Additives used during crystallization (DMSO and Tris molecules and magnesium atoms) were modeled in the electron density maps. However, no specific binding of these molecules is observed at, or close to, the active site and the lid loop. To further investigate this unusual closed conformation of the apo enzyme, two more rabbit muscle TIM structures, one in the same and another in a different crystal form, were determined. These structures present the open lid conformation only, indicating that the closed conformation cannot be explained by crystal contact effects. To rationalize why the active site loop is closed in the absence of ligand in one of the subunits, extensive comparison with previously solved TIM structures was carried out, supported by the bulk of available experimental information about enzyme kinetics and reaction mechanism of TIM. The observation of both open and closed lid conformations in TIM crystals might be related to a persistent conformational heterogeneity of this protein in solution.