Diverse interactions between the individual mutations in a double mutant at the active site of staphylococcal nuclease.

In principle, the quantitative effect of a second mutation on a mutant enzyme may be antagonistic, absent, partially additive, additive, or synergistic with respect to the first mutation. Depending on the kinetic or thermodynamic parameter measured, the D21E and R87G mutations of staphylococcal nuclease exhibit four of these five categories of interaction in the double mutant. While Vmax of the R87G single mutant of staphylococcal nuclease is 10(4.8)-fold lower than that of the wild-type enzyme and the Vmax of the D21E single mutant is 10(3.0)-fold below that of wild type, the double mutant D21E + R87G was found to lose a factor of only 10(4.1) in Vmax relative to wild type, rather than the product of the two single mutations (10(7.8)). These results suggest antagonistic structural effects of the individual R87G and D21E mutations. An alternative explanation for the nonadditivity of effects, namely, the separate functioning of these residues in a stepwise mechanism involving the prior attack of water on phosphorus followed by protonation of the leaving group by Arg-87, is unlikely since no enzyme-bound phosphorane intermediate (less than 1% of [enzyme]) was found under steady-state conditions on the R87G mutant by 31P NMR at 242.9 MHz. Like the effects on Vmax, quantitatively similar antagonistic effects of the two mutations were detected on the binding of divalent cations in binary enzyme-Ca2+ and enzyme-Mn2+ complexes and in the ternary enzyme-Ca2(+)-5'-pdTdA complex, suggesting that the effects on Vmax result from antagonistic structural changes at the Ca2+ binding site. Simple additive weakening effects of the two mutations were found on the binding of the substrate 5'-pdTdA, in both the absence and the presence of the divalent cations, Mn2+ and Ca2+. However, synergistic effects of the two mutations were found on the binding of the substrate analogue 3',5'-pdTp, profoundly weakening its binding to the double mutant in both the absence and the presence of divalent cations. Such synergistic effects of the two mutations may result from negative cooperativity or strain in the binding of 3',5'-pdTp to the wild-type enzyme. It is concluded that the quantitative interactions of two active-site mutations of an enzyme can vary greatly depending on which parameter of the enzyme is measured. When the two mutations interact in the same way on several parameters, a common underlying mechanism is suggested.     

Stereochemistry of the concerted enolization catalyzed by delta 5-3-ketosteroid isomerase

The reaction catalyzed by delta 5-3-ketosteroid isomerase has been shown to occur via the concerted enolization of the delta 5-3-ketosteroid substrate to form a dienolic intermediate, brought about by Tyr-14, which hydrogen bonds to and protonates the 3-keto group, and Asp-38, which removes and axial (beta) proton from C-4 of the substrate, in the same rate-limiting step [Xue, L., Talalay, P., & Mildvan, A.S. (1990) Biochemistry 29, 7491-7500; Kuliopulos, A., Mildvan, A.S., Shortle, D., & Talalay, P. (1989) Biochemistry 26, 3927-3937]. Since the axial C-4 proton is removed by Asp-38 from above the substrate, a determination of the complete stereochemistry of this rapid, concerted enolization requires information on the direction of approach of Tyr-14 to the enzyme-bound steroid. The double mutant enzyme, Y55F + Y88F, which retains Tyr-14 as the sole Tyr residue, was prepared and showed only a 4.5-fold decrease in kcat (12,000 s-1) and a 3.6-fold decrease in KM (94 microM) for delta 5-androstene-3, 17,dione, in comparison with the wild-type enzyme. Deuteration of the aromatic rings of the 10 Phe residues further facilitated the assignment of the aromatic proton resonances of Tyr-14 in the 600-MHz TOCSY spectrum at 6.66 +/- 0.01 ppm (3,5H) and at 6.82 +/- 0.01 ppm (2,6H). Variation of the pH from 4.9 to 10.9 did not alter these shifts, indicating that the pKa of Tyr-14 exceeds 10.9. Resonances assigned to the three His residues titrated with pKa values very similar to those found with the wild-type enzyme. The binding of 19-nortestosterone, a product analogue and substrate of the reverse isomerase reaction, induced downfield shifts of -0.12 and -0.06 ppm of the 3,5-and 2,6-proton resonances of Tyr-14, respectively, possibly due to deshielding by the 3-keto group of the steroid, but also induced +0.29 to -0.41 ppm changes in the chemical shifts of 8 of the 10 Phe residues and smaller changes in 10 of the 12 ring-shifted methyl resonances, indicating a steroid-induced conformation change in the enzyme. NOESY spectra in H2O revealed strong negative Overhauser effects from the 3,5-proton resonance of Tyr-14 to the overlapping 2 alpha-, 2 beta-, or 6 beta-proton resonances of the bound steroid but no NOE's to the 4- or 6 alpha-protons of the steroid.(ABSTRACT TRUNCATED AT 400 WORDS)     

Mandelate racemase from Pseudomonas putida. Magnetic resonance and kinetic studies of the mechanism of catalysis.

The interactions of mandelate racemase with divalent metal ion, substrate, and competitive inhibitors were investigated. The enzyme was found by electron paramagnetic resonance (EPR) to bind 0.9 Mn2+ ion per subunit with a dissociation constant of 8 muM, in agreement with its kinetically determined activator constant. Also, six additional Mn2+ ions were found to bind to the enzyme, much more weakly, with a dissociation constant of 1.5 mM. Binding to the enzyme at the tight site enhances the effect of Mn2+ on the longitudinal relaxation rate (1/T1p) of water protons by a factor of 11.9 at 24.3 MHz. From the frequency dependence of 1/T1p, it was determined that there are similar to 3 water ligands on enzyme-bound Mn2+ which exchange at a rate larger than or equal to 10-7 sec-1. The correlation time for enzyme-bound Mn2+-water interaction is frequency-dependent, indicating it to be dominated by the electron spin relaxation time of Mn2+. Formation of the ternary enzyme-Mn2+-mandelate complex decreases the number of fast exchanging water ligands by similar to 1, but does not affect tau-c, suggesting the displacement or occlusion of a water ligand. The competitive inhibitors D,L-alpha-phenylglycerate and salicylate produce little or no change in the enzyme-Mn2+-H2O interaction, but ternary complexes are detected indirectly by changes in the dissociation constant of the enzyme-Mn2+ complex and by mutual competition experiments. In all cases the dissociation constants of substrates and competitive inhibitors from ternary complexes determined by magnetic resonance titrations agree with K-M and K-i values determined kinetically and therefore reflect kinetically active complexes. From the paramagnetic effects of Mn2+ on 1/T1 and 1/T2 of the 13C-enriched carbons of 1-[13C]-D,L-mandelate and 2-[13C]-D,L-mandelate, Mn2+ to carboxylate carbon and Mn2+ to carbinol carbon distances of 2.93 plus or minus 0.04 and 2.71 plus or minus 0.04 A, respectively, were calculated, indicating bidentate chelation in the binary Mn2+-mandelate complex. In the active ternary complex of enzyme, Mn2+, and D,L-mandelate, these distances increase to 5.5 plus or minus 0.2 and 7.2 plus or minus 0.2 A, respectively, indicating the presence of at least 98.9% of a second sphere complex in which Mn2+, and C1 and C2 carbon atoms are in a linear array. The water relaxation data suggest that a water ligand is immobilized between the enzyme-bound Mn2+ and the carboxylate of the bound substrate. This intervening water ligand may polarize or protonate the carboxyl group. From 1/T2p the rate of dissociation of the substrate from this ternary complex (larger than or equal to 5.6 times 10-4 sec-1) is at least 52 times greater than the maximal turnover number of the enzyme (1070 sec-1), indicating that the complex detected by nuclear magnetic resonance (NMR) is kinetically competent to participate in catalysis. Relationships among the microscopic rate constants are considered.     

EPR study of 17O nuclear hyperfine interaction in cobalt-oxyhemoglobin: conformation of bound oxygen.

The electron spin resonance spectra of cobalt-oxyhemoglobin, with oxygen enriched to 95 atom % with ¹⁷O, are broadened due to ¹⁷O nuclear hyperfine interaction with the unpaired electron spin and show at X-band and 7°K the presence of several well-resolved ¹⁷O satellites at low and high field extremes, all of which are absent in ¹⁶O cobaltoxyhemoglobin. The ¹⁷O EPR spectra are interpreted in terms of two nonequivalent hyperfine interaction constants of 65± 5 G and 93 ± 5 G for the two oxygen atoms. The observed broadening and hyperfine splittings in the EPR spectra indicate complete transfer of the unpaired spin to oxygen orbitals. The observed difference in the hyperfine coupling constants of the two oxygen atoms establishes an asymmetric linkage of the oxygen molecule to the metal, consistent with the Pauling model of oxyhemoglobin.     

Nuclear magnetic resonance studies of substrate interaction with cobalt substituted alcohol dehydrogenase from liver.

The role of zinc in liver alcohol dehydrogenase has been studied by replacement of 1.3 and 3.5 of the four Zn(II) ions with Co(II) and measuring the effects of the paramagnetic Co(II) on the relaxation rates of the protons of water, ethanol, and isobutyramide. Water relaxation studies at 8, 24, 100, and 220 MHz indicate two classes of bound Co(II). The similar to 2 readily replaced Co(II) ions retain one fast exchanging water proton in their inner coordination spheres, while the similar to 2 slowly exchanging Co(II) ions coordinate no detectable water protons, indicating that the former replaced Zn(II) at the "catalytic sites" and the latter replaced Zn(II) at the "structural sites" detected crystallographically. Ethanol, acetaldehyde, and isobutyramide bind with appropriate affinities to the Co(II) substituted alcohol dehydrogenases decreasing the number of fast exchanging protons at the catalytic Co(II) site by greater than or equal to 54 percent. Coenzyme binding causes smaller changes in the water relaxation rate which may be due to local conformation changes. The paramagnetic effects of Co(II) at the catalytic site on the relaxation rates of the methyl protons of isobutyramide at 100 and 220 MHz indicate that this analog binds at a site 9.1 A from the catalytic Co(II). This distance decreases to 6.9 A when NADH is bound, and a Co(II) to methyne proton distance of 6.6 A is determined indicating a conformation change leading to the formation of a second sphere enzyme-Co(II)-isobutyramide complex in which a hydroxyl or water ligand intervenes between the metal and the substrate analog. Similar behavior is observed in the enzyme-ethanol complexes. The paramagnetic effects of Co(II), at the catalytic site, on the relaxation rates of the protons of ethanol at 100 and 220 MHz, indicate that this substrate bind at a site 12-14 A distant from the catalytic Co(II) but that this distancedecreases to 6.3 A in the abortive enzyme-NADH-ethanol complex. The role of the catalytic Co(II) thus appears to be the activation of a hydroxyl or water ligand which polarizes the aldehyde carbonyl group by hydrogen bonding. The role of the structural Co(II), which is more distant from isobutyramide (9-11 A), may be that of a template for protein conformation changes. By combining the present distances with those from previous magnetic resonance studies on the liver enzyme, the arrangement of coenzyme, metal, and substrate at the active site in solution can be constructed. This arrangement is consistent with that of ADP-ribose and zinc in the crystalline complex of liver alcohol dehydrogenase as determined by X-ray crystallography (Branden et al., (1973), Proc. Natl. Acad. Sci. U.S.A.70, 2439).     

Spinach leaf phosphoenolpyruvate carboxylase: purification, properties, and kinetic studies

Spinach leaf phosphoenolpyruvate carboxylase has been purified to homogeneity using salt fractionatjon, chromatography, and immunologie procedures to remove contaminating ribulose diphosphate carboxylase. From gel filtration and isoelectric focusing, the molecular weight (~560,000) and isoelectric point (pI = 4.9) are indistinguishable from those of ribulose diphosphate carboxylase. The subunit molecular weight of phosphoenolpyruvate carboxylase (130,000) suggests that the native enzyme is a tetramer.Kinetic studies using Mg²⁺ or Mn²⁺ as the activator indicate that the divalent cation lowers the K_m of the substrate phosphoenolpyruvate by an order of magnitude and conversely, that the presence of the substrate similarly lowers the K_m of the metal ion, suggesting an enzyme-metal-substrate bridge complex. Three analogs of phosphoenolpyruvate, lphospholactate, d-phospholactate, and phosphoglycolate are potent competitive inhibitors. The inhibitor constant (K_i) of l-phospholactate (2 μm) is 49-fold lower with Mn²⁺ as the activator than with Mg²⁺. An analysis of the competitive inhibition by portions of the l-phospholactate molecule (i.e., l-lactate, methyl phosphate, and phosphite) indicates this 49-fold lowering is due to increased interaction of the phosphoryl group and, to a lesser extent, of the carboxyl and C-O-P bridge oxygen of l-phospholactate with the enzyme metal complex. The results provide indirect evidence for phosphoryl coordination by the enzyme-bound divalent cation.     

NMR studies of conformations and interactions of substrates and ribonucleotide templates bound to the large fragment of DNA polymerase I

The large fragment of DNA polymerase I (Pol I) effectively uses oligoribouridylates and oligoriboadenylates as templates, with kinetic properties similar to those of poly(U) and poly(A), respectively, and has little or no activity in degrading them. In the presence of such oligoribonucleotide templates, nuclear Overhauser effects (NOE's) were used to determine interproton distances within and conformations of substrates bound to the large fragment of Pol I, as well as conformations and interactions of the enzyme-bound templates. In the enzyme-oligo(rU)54 +/- 11-Mg2+dATP complex, the substrate dATP has a high anti-glycosidic torsional angle (chi = 62 +/- 10 degrees) and an O1'-endo/C3'-endo sugar pucker (delta = 90 +/- 10 degrees) differing only slightly from those previously found for enzyme-bound dATP in the absence of template [Ferrin, L.J., & Mildvan, A.S. (1985) Biochemistry 24, 4680-4694]. Both conformations are similar to those of deoxynucleotidyl units of B DNA but differ greatly from those of A or Z DNA. The conformation of the enzyme-bound substrate analogue AMPCPP (chi = 50 +/- 10 degrees, delta = 90 +/- 10 degrees) is very similar to that of enzyme-bound dATP and is unaltered by the binding of the template oligo(rU)54 +/- 11 or by the subsequent binding of the primer (Ap)9A. In the enzyme-oligo(rA)50-Mg2+TTP complex, the substrate TTP has an anti-glycosidic torsional angle (chi = 40 +/- 10 degrees) and an O1'-endo sugar pucker (delta = 100 +/- 10 degrees), indistinguishable from those found in the absence of template and compatible with those of B DNA but not with those of A or Z DNA. In the absence of templates, the interproton distances on enzyme-bound dGTP cannot be fit by a single conformation but require a 40% contribution from a syn structure (chi = 222 degrees) and a 60% contribution from one or more anti structures. The presence of the template oligo(rU)43 +/- 9 simplifies the conformation of enzyme-bound dGTP to a single structure with an anti-glycosyl angle (chi = 32 +/- 10 degrees) and an O1'-endo/C3'-endo sugar pucker (delta = 90 +/- 10 degrees), compatible with those of B DNA, possibly due to the formation of a G-U wobble base pair. However, no significant misincorporation of guanine deoxynucleotides by the enzyme is detected with oligo(rU) as template.(ABSTRACT TRUNCATED AT 400 WORDS)