Ras-catalyzed hydrolysis of GTP: a new perspective from model studies.

Despite the biological and medical importance of signal transduction via Ras proteins and despite considerable kinetic and structural studies of wild-type and mutant Ras proteins, the mechanism of Ras-catalyzed GTP hydrolysis remains controversial. We take a different approach to this problem: the uncatalyzed hydrolysis of GTP is analyzed, and the understanding derived is applied to the Ras-catalyzed reaction. Evaluation of previous mechanistic proposals from this chemical perspective suggests that proton abstraction from the attacking water by a general base and stabilization of charge development on the gamma-phosphoryl oxygen atoms would not be catalytic. Rather, this analysis focuses attention on the GDP leaving group, including the beta-gamma bridge oxygen of GTP, the atom that undergoes the largest change in charge in going from the ground state to the transition state. This leads to a new catalytic proposal in which a hydrogen bond from the backbone amide of Gly-13 to this bridge oxygen is strengthened in the transition state relative to the ground state, within an active site that provides a template complementary to the transition state. Strengthened transition state interactions of the active site lysine, Lys-16, with the beta-nonbridging phosphoryl oxygens and a network of interactions that positions the nucleophilic water molecule and gamma-phosphoryl group with respect to one another may also contribute to catalysis. It is speculated that a significant fraction of the GAP-activated GTPase activity of Ras arises from an additional interaction of the beta-gamma bridge oxygen with an Arg side chain that is provided in trans by GAP. The conclusions for Ras and related G proteins are expected to apply more widely to other enzymes that catalyze phosphoryl (-PO(3)2-) transfer, including kinases and phosphatases.

Spectroscopic studies of arsenic(III) binding to Escherichia coli RI methyltransferase and to two mutants, C223S and W183F.

The interactions of an arsenic (III) reagent, (CH3)2AsSCH2CONH2, with two Escherichia coli RI methyltransferase mutants, W183F and C223S, have been studied by phosphorescence, optically detected magnetic resonance, and fluorescence spectroscopy. The phosphorescence spectrum of the W183F mutant containing only one tryptophan at position 225 reveals a single 0,0-band that is red-shifted by 9.8 nm upon binding of As(III). Fluorescence titration of W183F with (CH3)2AsSCH2CONH2 produces a large tryptophan fluorescence quenching. Analysis of the quenching data points to a single high-affinity As(III) binding site that is associated with the fluorescence quenching. Triplet-state kinetic measurements performed on the perturbed tryptophan show large reductions in the lifetimes of the triplet sublevels, especially that of the T chi sublevel. As(III) binding to the enzyme at a site very close to the Trp225 residue induces an external heavy-atom effect, showing that the perturber atom is in van der Waals contact with the indole chromophore. In the case of the C223S mutant, a single tryptophan 0,0-band also is observed in the phosphorescence spectrum, but no change occurs upon addition of the As(III) reagent. Fluorescence titration of C223S with As(III) shows essentially no quenching of tryptophan fluorescence, in contrast with W183F. These results, along with previous triplet-state and biochemical studies on the wild-type enzyme [Tsao, D. H.H., & Maki, A. H. (1991) Biochemistry 30, 4565-4572], show that As(III) binds with high affinity to the Cys223 residue and that the Trp225 side chain is located close enough to that of Cys223 to produce a heavy-atom perturbation when As(III) is bound.

Cofactor and DNA interactions in EcoRI DNA methyltransferase. Fluorescence spectroscopy and phenylalanine replacement for tryptophan 183.

EcoRI DNA methyltransferase contains tryptophans at positions 183 and 225. Tryptophan 225 is adjacent to residues previously implicated in S-adenosylmethionine (AdoMet) binding and to cysteine 223, previously shown to be the site of N-ethyl maleimide-mediated inactivation of the enzyme (Reich, N. O., and Everett, E. (1990) J. Biol. Chem. 265, 8929-8934; Everett, E. A., Falick, A. M., and Reich, N. O. (1990) J. Biol. Chem. 265, 17713-17719). The fluorescence spectra of the wild-type enzyme is centered at 338 nm indicating partial tryptophan solvent accessibility. Substitution of tryptophan 183 with phenylalanine results in a 45% drop in fluorescence intensity, but no shift in lambda max. DNA binding to the wild-type methyltransferase caused an increase in the fluorescence intensity, while binding to the tryptophan 183 mutant had a quenching effect, suggesting that DNA binding induces a conformational change near both tryptophans. Binding of AdoMet and various AdoMet analogs to the wild-type methyltransferase results in no change in the fluorescence spectrum when excitation occurs at 295 nm, suggesting that no conformational change occurs, and AdoMet does not interact with either tryptophan. In contrast, quenching was observed when excitation occurred at 280 nm, suggesting that AdoMet and its analogs may be quenching tyrosine to tryptophan energy transfer. Protein-ligand complexes were titrated with acrylamide, and the data also implicate conformational changes upon DNA binding but not upon AdoMet binding, consistent with previous limited proteolysis results (Reich, N. O., Maegley, K. A., Shoemaker, D.D., and Everett, E. (1991) Biochemistry 30, 2940-2946).

Structural and functional analysis of EcoRI DNA methyltransferase by proteolysis.

Native EcoRI DNA methyltransferase (Mtase, Mr 38,050) is proteolyzed by trypsin to generate an intermediate 36-kDa fragment (p36) followed by the formation of two polypeptides of Mr 23,000 and 13,000 (p23 and p13, respectively). Protein sequence analysis of the tryptic fragments indicates that p36 results from removal of the first 14 or 16 amino acids, p23 spans residues 15-216, and p13 spans residues 217-325. The relative resistance to further degradation of p23 and p13 suggests stable domain structures. This is further supported by the generation of similar fragments with SV8 endoprotease which has entirely different peptide specificities. Our results suggest the Mtase is a two-domain protein connected by a highly flexible interdomain hinge. The putative hinge region encompasses previously identified peptides implicated in AdoMet binding [Reich, N.O., & Everett, E. (1990) J. Biol. Chem. 265, 8929-8934] and catalysis [Everett et al. (1990) J. Biol. Chem. 265, 17713-17719]. Protection studies with DNA, S-adenosylmethionine (AdoMet), S-adenosylhomocysteine (AdoHcy), and sinefungin (AdoMet analogue) show that the Mtase undergoes significant conformational changes upon ligand binding. Trypsinolysis of the AdoMet-bound form of the Mtase generates different fragments, and the AdoMet-bound form is over 800 times more stable than unbound Mtase. The sequence-specific ternary complex (Mtase-DNA-sinefungin) is 2000 times more resistant to degradation by trypsin; cleavage eventually generates 26- and 12-kDa fragments which span residues 104-325 and 1-103, respectively (p26 and p12). The first 14 or 16 amino acids of the Mtase are not essential since p36 retains activity. Activity analysis of the p26 and p12 mixture also indicates retention of activity.(ABSTRACT TRUNCATED AT 250 WORDS)