A computer model of the complex between G2'p5'G and barnase, the recombinant ribonuclease of Bacillus amyloliquefaciens, was constructed, based on the known structure of the complex RNAase T1.G2'p5'G. This model suggests that the conserved residue Glu-60 plays an important role in the specificity of barnase for guanosine. A barnase mutant was therefore made in which Glu-60 was replaced by Gln. This mutation increases the Km for the dinucleotides GpC and GpA, by a factor of 10, but does not change the kcat. For ApA, the kcat/Km decreases by a similar factor, but the individual parameters could not be determined. The mutation, however, has no influence on the kcat and the Km of barnase action towards RNA and poly(A). This demonstrates that the interactions between the substrate and the residue at position 60 must be different in the case of ApA and poly(A). For RNA, this conclusion is also likely, but not absolutely certain, because barnase/RNA might be a Briggs-Haldane type enzyme/substrate pair. Therefore, if the effect of the mutation were limited to an increase of the dissociation rate constant of the substrate (k-1), this would not be evident in Km or kcat/Km. In view of the clear cut situation with poly(A), the pH profile for and the effect of salt concentration on the kinetic parameters of the mutant barnase were studied for this substrate. The influence of salt on the Km can be interpreted via the linked function concept and shows a cooperative dissociation of 7-10 counterions upon poly(A) binding. The binding of the substrate is strongly reduced at high pH, and the pKa involved decreases strongly at high salt concentrations. Poly(A) and RNA show a pH dependency of their absorbance spectrum, indicating a pH-dependent change of base stacking, which may influence the catalytic parameters.

The acrylamide quenching of the tryptophan fluorescence of apo and holo glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was studied. In the case of apo-GAPDH, the steady state fluorescence quenching cannot be described by the classical Stern-Volmer equation: strong cooperative quenching is observed. In the presence of Pi and/or cofactor NAD+, an inaccessible fraction appears. Cooperative quenching is partially suppressed in the presence of Pi and fully absent in the presence of NAD+. The measurements of the fluorescence lifetimes of the holo-enzyme by phasefluorometry allow the resolution of two lifetimes. The long-lived component is quenched by acrylamide, the short-lived component is not. Quenching induces a red shift of the steady state emission peak. The quenching parameters from the lifetime measurements allow the quantitative description of the steady state fluorescence quenching data. In agreement with the observations of Orstan and Gafni (Photochemistry and Photobiology, (1990) 31, 725-731), we find that acrylamide causes a slow, irreversible loss of activity and a reduction of titratable thiol groups when it acts on the apo-enzyme. This inactivation is strongly reduced in the presence of NAD+. We show that this inactivation is also slowed down by the presence of Pi, and that it is accompanied by a loss of the NAD+ binding site. Blocking the thiol groups with 5,5'-dithio-bis-(2-nitrobenzoic acid) does not lead to a protection against the irreversible inactivation by acrylamide, showing that reactions other than thiol modifications are involved in the irreversible effect. A fraction of the inactivation can be reversed by treatment with mercapto-ethanol.

No abstract