Localization of the high-affinity ATP site in adenosine-3':5'-monophosphate-dependent protein kinase type I. Photoaffinity labelling studies with 8-azidoadenosine 5'-triphosphate

Viroid-induced phosphorylation of a host protein related to a dsRNA-dependent protein kinase

Viroids are very small, unencapsidated RNAs that replicate and induce severe disease in plants without encoding for any proteins. The mechanisms by which the viroid RNA regulates these events and interacts with host factors are unknown. An Mr 68,000 host-encoded protein has been identified that is differentially phosphorylated in extracts from viroid-infected and mock-inoculated tissues. This phosphoprotein is immunologically related to a double-stranded (ds) RNA-dependent protein kinase from virus-infected, interferon-treated human cells. Further, nucleotide photoaffinity labeling indicates that the protein has an ATP binding site. This protein is similar to dsRNA-dependent protein kinases implicated in mammalian systems in the regulation of protein synthesis and virus replication.

8-Azido-ATP (alpha 32P) binding to rod outer segment proteins

ATP has important roles in the vertebrate rod outer segment (ROS) physiological response to light. One of them is the quench of light-activated cGMP-phosphodiesterase activity. How ATP quenches PDE is not established; however, leading hypotheses favor the intervention of a 48-kDa ATP-binding protein and/or an ATP-utilizing rhodopsin kinase in this reaction. We have investigated the binding of [alpha 32P]8-azido-ATP to rat ROS proteins in the presence and absence of various divalent cations and competitive nucleotides. An event we have detected which might further clarify the role of ATP in PDE inactivation is a zinc-induced binding of azido-ATP to rhodopsin. Manganese is also effective in inducing this binding, while magnesium and calcium are not. The azido-ATP binding is eliminated by the addition of ATP, but not GTP, UTP, cGMP, or cAMP. A nucleotide-binding site on the rim protein is also suggested from these studies.

Identification of a MAP 2-like ATP-binding protein associated with axoplasmic vesicles that translocate on isolated microtubules

Axoplasmic vesicles were purified and observed to translocate on isolated microtubules in an ATP-dependent, trypsin-sensitive manner, implying that ATP-binding polypeptides essential for force generation were present on the vesicle surface. To identify these proteins [alpha 32P]8-azidoadenosine 5'-triphosphate ([alpha 32P]8-N3ATP), a photoaffinity analogue of ATP, was used. The results presented here identify and characterize a vesicle-associated polypeptide having a relative molecular mass of 292 kD that bound [alpha 32P]8-N3ATP. The incorporation of label is ultraviolet light-dependent and ATP-sensitive. Moreover, the 292-kD polypeptide could be isolated in association with vesicles or microtubules, depending on the conditions used, and the data indicate that the 292-kD polypeptide is similar to mammalian brain microtubule-associated protein 2 (MAP 2) for the following reasons: The 292-kD polypeptide isolated from either squid axoplasm or optic lobe cross-reacts with antiserum to porcine brain MAP 2. Furthermore, it purifies with taxol-stabilized microtubules and is released with salt. Based on these characteristics, the 292-kD polypeptide is distinct from the known force-generating molecules myosin and flagellar dynein, as well as the 110-130-kD kinesin-like polypeptides that have recently been described (Brady, S. T., 1985, Nature (Lond.), 317:73-75; Vale, R. D., T. S. Reese, and M. P. Sheetz, 1985b, Cell, 42:39-50; Scholey, J. M., M. E. Porter, P. M. Grissom, and J. R. McIntosh, 1985, Nature (Lond.), 318:483-486). Because the 292-kD polypeptide binds ATP and is associated with vesicles that translocate on purified MAP-free microtubules in an ATP-dependent fashion, it is therefore believed to be involved in vesicle-microtubule interactions that promote organelle motility.

Identification of a nucleoside triphosphate binding site on calf thymus RNA polymerase II

A nucleoside triphosphate binding site on calf thymus RNA polymerase II was identified by using photoaffinity analogues of adenosine 5'-triphosphate and guanosine 5'-triphosphate. Both radiolabeled 8-azidoadenosine 5'-triphosphate (8-N3ATP) and radiolabeled 8-azidoguanosine 5'-triphosphate (8-N3GTP) bound to a single polypeptide of this enzyme. This polypeptide has a molecular mass of 37 kilodaltons and an isoelectric point of 5.4. Ultraviolet (UV) irradiation was necessary for photolabeling to occur. In addition, no labeling occurred when the probe was prephotolyzed or when the enzyme was inactivated. Furthermore, photolabeling of the enzyme could be decreased by preincubation with natural substrates. To provide evidence that the radiolabeled polypeptide forms a part of the domain of the nucleoside triphosphate binding site, experiments were performed using unlabeled 8-N3ATP. Although this unlabeled analogue was not a substrate for RNA polymerase II, it photoinactivated the enzyme in the presence of UV irradiation, and it inhibited transcription elongation by the enzyme in a competitive manner in the absence of UV irradiation. As in the case with photolabeling, photoinactivation by 8-N3ATP could be decreased by natural substrates; in both cases, purine ribonucleoside triphosphates were more efficient than pyrimidine nucleoside triphosphates. Furthermore, photoinactivation was saturable at about the same concentration as the inhibition constant for 8-N3ATP. Collectively, these results provide evidence that the radiolabeled polypeptide in calf thymus RNA polymerase II is an essential component for activity and suggest that this polypeptide may be part of this enzyme's purine ribonucleoside triphosphate binding site.

The function of Mg-ATP in interactions between the regulatory and catalytic subunits of type I cAMP-dependent protein kinase from rabbit skeletal muscle

The regulatory subunit of Type I cAMP-dependent protein kinase from rabbit skeletal muscle can bind [3H]cAMP to form the R-[3H]cAMP complex, and the slow phase of the enhanced exchange of free cAMP with [3H]cAMP from the R-[3H]cAMP complexes was studied under various conditions using the equilibrium isotope exchange technique. Results indicate that Mg-ATP and the catalytic subunit are absolutely required for the enhanced exchange reaction to occur, but phosphorylation of the regulatory subunit by Mg-ATP does not play a determining role in the slow rate of the dissociation/association of the Type I protein-kinase in the presence of cAMP and the catalytic subunit. We interpret the role of Mg-ATP as being one in which it may provide the structural attributes required for formation of a stabilized transient state of the cAMP-regulatory subunit-catalytic subunit ternary complex, an obligatory intermediate involved in the dissociation/association of Type I cAMP-dependent protein kinase.

The photoaffinity probe 8-N3[alpha-32P]ATP labels the ATP-binding sites of rabbit neutrophil and skeletal muscle actin

8-Azido-[alpha-32P]ATP (8-N3-ATP) was used as a photoaffinity label for ATP binding sites in the subcellular fractions of rabbit peritoneal neutrophils. The radioactive 8-N3-ATP was specifically incorporated into one major protein of 43 kDa. The isoelectric point, molecular mass and subcellular distribution of this labeled protein closely resemble those of the actin. 8-N3-[alpha-32P]ATP was further tested as a photoaffinity label for the ATP binding site in the purified rabbit skeletal muscle G-actin. The radioactive 8-N3-ATP was specifically incorporated into the actin band in SDS-polyacrylamide gel. The results indicate that 8-N3-ATP can be used as a photoaffinity label for actin.

Adenosine cyclic 3',5'-monophosphate dependent protein kinase: interaction of the catalytic subunit and holoenzyme with lin-benzoadenine nucleotides

The interaction of lin-benzoadenosine di- and triphosphates with the catalytic subunit and type II holoenzymes of adenosine cyclic 3',5'-monophosphate (cAMP) dependent protein kinase has been investigated by steady-state kinetics and fluorescence spectroscopy. lin-Benzo-ADP is a competitive inhibitor of the catalytic subunit with respect to ATP with a Ki (8.0 microM) similar to the Ki for ADP (9.0 microM). This value agrees well with the Kd (9.0 microM) determined by fluorescence polarization titration. Type II holoenzymes from bovine brain and skeletal muscle have Kd values for lin-benzo-ADP of 3.4 microM and 3.5 microM, respectively, and each binds approximately 2 mol/mol of R2C2 tetramer. Furthermore, fluorescence polarization studies indicate that both the catalytic subunit and type II holoenzyme bind lin-benzo-ADP rigidly, so that there is little or no rotation of the lin-benzoadenine portion of the molecule within the nucleotide binding site. lin-Benzo-ATP is a substrate for the phosphotransferase activities of protein kinase with peptides, water, or type II regulatory subunit as phosphoryl acceptors. With Leu-Arg-Arg-Ala-Ser-Leu-Gly as phosphoryl acceptor, the Km for lin-benzo-ATP is 11.3 microM, and that for ATP is 11.9 microM. The Vmax with lin-benzo-ATP is 20% of the Vmax with ATP as the substrate [24.9 +/- 1.8 mumol/(min . mg) vs. 5.0 +/- 1.2 mumol/(min . mg)]. Thus lin-benzo-ATP is the best nucleotide substrate (besides ATP) for the catalytic subunit reported. 1,N6-Etheno-ATP (epsilon ATP), on the other hand, is a poor substrate for the catalytic subunit with a Km of 1.8 mM and a Vmax that is 4% of the Vmax for ATP, making it unsuitable as a fluorescence probe for cAMP-dependent protein kinase.

Polyoma virus middle T antigen: relationship to cell membranes and apparent lack of ATP-binding activity

Middle T antigen of polyoma virus is associated principally with the plasma membrane. Comparison of the trypsin sensitivity of middle T in intact cells and "inside out" membrane preparations showed that middle T is oriented towards the inside of the cell. This was confirmed by labeling of middle T in permeabilized cells, but not in intact cells, using [gamma-32P]ATP. Middle T molecules active in the in vitro kinase reaction could be differentiated from the bulk (metabolically labeled) middle T based on resistance to trypsin treatment. The active fraction also behaved differently from the bulk when cell frameworks were prepared with Triton-containing buffers; whereas the bulk middle T was evenly distributed in the soluble and cell framework fractions, the kinase-active forms were largely associated with the framework. Middle T molecules labeled in vivo with 32PO4 were found largely in the framework fraction, like the molecules that show kinase activity in vitro. Experiments with ATP affinity reagents 8-azido-ATP and 2,3-dialdehyde ATP have failed to label the middle T antigen. However, 2,3-dialdehyde ATP could be used to inhibit the kinase reaction. This raises the question of whether middle T antigen possesses intrinsic kinase activity or, rather, associates with a cellular tyrosine kinase.

Polyoma virus middle T antigen: relationship to cell membranes and apparent lack of ATP-binding activity

Middle T antigen of polyoma virus is associated principally with the plasma membrane. Comparison of the trypsin sensitivity of middle T in intact cells and "inside out" membrane preparations showed that middle T is oriented towards the inside of the cell. This was confirmed by labeling of middle T in permeabilized cells, but not in intact cells, using [gamma-32P]ATP. Middle T molecules active in the in vitro kinase reaction could be differentiated from the bulk (metabolically labeled) middle T based on resistance to trypsin treatment. The active fraction also behaved differently from the bulk when cell frameworks were prepared with Triton-containing buffers; whereas the bulk middle T was evenly distributed in the soluble and cell framework fractions, the kinase-active forms were largely associated with the framework. Middle T molecules labeled in vivo with 32PO4 were found largely in the framework fraction, like the molecules that show kinase activity in vitro. Experiments with ATP affinity reagents 8-azido-ATP and 2,3-dialdehyde ATP have failed to label the middle T antigen. However, 2,3-dialdehyde ATP could be used to inhibit the kinase reaction. This raises the question of whether middle T antigen possesses intrinsic kinase activity or, rather, associates with a cellular tyrosine kinase.

A model for the chemical interactions of adenosine 3':5'-monophosphate with the R subunit of protein kinase type I. Refinement of the cyclic phosphate binding moiety of protein kinase type I

The cAMP receptor site in the regulatory subunit of adenosine 3':5'-monophosphate (cAMP)-dependent protein kinase type I was mapped using analogues of cAMP in which the ribose phosphate moiety was systematically modified. Electronical alteration of the cyclophosphate ring at the 3' and 5' positions by sulfur and nitrogen decreased the affinity of these analogues towards the kinase. Substituents at these positions are not tolerated. Testing the separated diastereomers of derivatives in which one of the exocyclic oxygens at the phosphorus has been substituted by sulfur, it was found that one diastereoisomer is preferentially recognized. Based on these results it is proposed that the hydrophylic cyclic phosphate-ribose moiety of cAMP is bound to the kinase via its 3' and 5'-oxygens, the 2'-hydroxy group and the negative charge in a fixed position. Based on our and other published results it is further proposed, that the adenine moiety is bound in a hydrophobic cleft without any hydrogen bond interactions. The chemical interactions between cAMP and the R subunit of protein kinase type I differ from those found for the binding of cAMP to the chemoreceptor of Dictyostelium discoideum [18].