Crystal structure of protein isoaspartyl methyltransferase: a catalyst for protein repair.

BACKGROUND: Formation of isoaspartyl residues is one of several processes that damage proteins as they age. Protein L-isoaspartate (D-aspartate) O-methyltransferase (PIMT) is a conserved and nearly ubiquitous enzyme that catalyzes the repair of proteins damaged by isoaspartyl formation. RESULTS: We have determined the first structure of a PIMT from crystals of the T. maritima enzyme complexed to S-adenosyl-L-homocysteine (AdoHcy) and refined it to 1.8 A resolution. Although PIMT forms one structural unit, the protein can be divided functionally into three subdomains. The central subdomain closely resembles other S-adenosyl-L-methionine-dependent methyltransferases but bears a striking alteration of topological connectivity, which is not shared by any other member of this family. Rather than arranged as a mixed beta sheet with topology 6 upward arrow7 downward arrow5 upward arrow4 upward arrow1 upward arrow2 upward arrow3 upward arrow, the central sheet of PIMT is reorganized to 7 upward arrow6 downward arrow5 upward arrow4 upward arrow1 upward arrow2 upward arrow3 upward arrow. AdoHcy is largely buried between the N-terminal and central subdomains by a conserved and largely hydrophobic loop on one rim of the binding cleft, and a conserved Ser/Thr-rich beta strand on the other. The Ser/Thr-rich strand may provide hydrogen bonds for specific interactions with isoaspartyl substrates. The side chain of Ile-206, a conserved residue, crosses the cleft, restricting access to the donor methyl group to a deep well, the putative isoaspartyl methyl acceptor site. CONCLUSIONS: The structure of PIMT reveals a unique modification of the methyltransferase fold along with a site for specific recognition of isoaspartyl substrates. The sequence conservation among PIMTs suggests that the current structure should prove a reliable model for understanding the repair of isoaspartyl damage in all organisms.

Reaction of the isosteric methylenephosphonate analog of alpha-D-glucose 1-phosphate with phosphoglucomutase. Induced-fit specificity revisited.

The phospho form of phosphoglucomutase reacts with the isosteric methylenephosphonate analog of alpha-D-glucose 1-phosphate to produce the corresponding analog of alpha-D-glucose 1,6-bisphosphate plus the dephosphoenzyme. In a coupled reaction, kcat/Km = 1.7 x 10(3) M-1 s-1, which is about 2 x 10(-5) times that for the corresponding reaction with alpha-D-glucose 1-phosphate. The decrease in kcat/Km is divided more or less evenly between less efficient PO3- transfer and decreased binding, although smaller phosphates and phosphonates bind approximately equally. There is a much smaller difference in the binding of glucose 1-methylenephosphonate 6-phosphate and glucose 1,6-bisphosphate to the dephosphoenzyme: the binding ratio is < 1:35 when the glucose ring is oriented similarly. Preferred binding patterns for a number of substrates/inhibitors, studied by 31P NMR and UV-difference spectroscopy, suggest that in the ground state the phosphonate group is tolerated to a much greater extent at the catalytic subsite than at the phosphate-binding subsite, where binding specificity appears to be directed toward a tetrahedral-PO3(2-) group attached to a bridging atom that can act as a hydrogen-bond acceptor. Binding specificity at the catalytic subsite apparently is directed toward a different array, possibly (-O...PO3...O-)2-. Some of these results are considered in terms of a modified version of the "induced fit" concept of enzymic specificity, which is reexamined in view of implied thermodynamic restrictions. The internal rearrangement whereby the positions of the anionic groups of the phosphate/phosphonate are exchanged is compared with the analogous rearrangements involving glucose 1,6-bisphosphate and 1,4-butanediol bisphosphate. The supplementary material describes a three-step synthesis of 1-deoxy-alpha-D-glucose 1-methylenephosphonate together with a procedure for phosphorylating the phosphonate to produce an analog of alpha-D-glucose 1,6-bisphosphate and also describes a facile procedure for the qualitative conversion of organic phosphonates to inorganic phosphate.

Removal of salt from a salt-induced protein crystal without cross-linking. Preliminary examination of "desalted" crystals of phosphoglucomutase by X-ray crystallography at low temperature.

A model procedure for removing salt from relatively fragile salt-induced protein crystals is proposed. The procedure is based on physical principles and is validated by using millimeter-size crystals of rabbit muscle phosphoglucomutase grown from a 2.1 M solution of ammonium sulfate. Three types of operations are included in the procedure: initial transfer to salt solutions of reduced concentration; transfer to the organic-rich phase of an equilibrium biphasic mixture obtained with aqueous solutions of polyoxyethylene and the salt; and addition of various replacement cosolutes in aqueous solutions of polyoxyethylene to reduce osmotic stress on the crystal as the remaining salt is removed. A critical feature of the overall procedure is maintenance of near equilibrium throughout by using a large number of steps involving small changes in solute concentration. The conditions used in the actual transfer were adjusted to eliminate the fracturing of crystals by visually distinguishing between two opposing types of fracture patterns: those produced by osmotic crushing as opposed to osmotic expansion. Basic requirements for a successful procedure with other protein crystals are a high permeability toward small solutes and a relatively slow dissolution rate at salt concentrations for which biphasic mixtures can be obtained. Desalted crystals of phosphoglucomutase have no visible fractures, are stable in the final solution for at least a week, and exhibit no noticeable change in the resolution of their X-ray diffraction pattern. In fact, desalted crystals can be rapidly cooled to 160 K, whereas untreated crystals are almost completely disordered by the same cooling procedure. The component of the desalting mixture whose presence is crucial to the success of the cooling process is polyoxyethylene, which apparently impedes the formation of ice within the protein crystal. Diffraction data obtained with an area-detector diffractometer did not differ significantly, either in terms of quality or resolution range, between crystals in 2.3 M ammonium sulfate at room temperature and crystals at 160 K in which ammonium sulfate had been replaced by glycine. The successful use of the following replacement solutes, instead of glycine, also is documented: sucrose, glycerol, and a low molecular weight poly(ethylene glycol) (PEG-400).

Formation of substrate and transition-state analogue complexes in crystals of phosphoglucomutase after removing the crystallization salt.

Crystals of phosphoglucomutase, grown in 2.1 M ammonium sulfate, "desalted", and suspended in a 30% polyoxyethylene-8000/1 M glycine solution as described in the accompanying paper [Ray, W. J., Jr., Puvathingal, J. M., Bolin, J. T., Minor, W., Liu, Y., & Muchmore, S. W. (1991) Biochemistry 30 (preceding paper in this issue)], were treated with glucose phosphates to form an equilibrium mixture of the catalytically active substrate/product complexes. However, this treatment extensively fractured the crystals, even when very dilute solutions of glucose phosphates were used. But formation of the desired complexes was achieved, without fracturing, by introducing the glucose phosphates at high salt concentration, where they do not bind significantly to the enzyme, and maintaining their presence during subsequent sulfate-removal steps, in order to obtain essentially uniform binding throughout the crystal at all times. Although this procedure produced unfractured crystals of the catalytically active complexes, an adjustment in water activity was required to prevent the crystals from slowly liquefying in the presence of the added glucose phosphates. After this adjustment, the quality of diffraction-grade crystals subjected to this treatment was not significantly altered. An even larger adjustment in water activity was required to stabilize crystals that had been largely converted into a mixture of vanadate-based transition-state analogue complexes [cf. Ray, W. J., Jr., & Puvathingal, J. M. (1990) Biochemistry 29, 2790-2801] by means of an analogous procedure. The rationale for, and the implications of, this adjustment of water activity are discussed. The phenomenon of lattice-based binding cooperativity also is discussed together with a possible role for such cooperativity in the fracturing of protein crystals during formation of ligand complexes and possible ways to circumvent such fracturing based on the annealing of crystals at fractional saturation. An assay for quantifying the extent of formation of the vanadate-based transition-state analogue complexes in crystals of phosphoglucomutase is described. A solution to problems associated with producing and maintaining a steady-state in treated crystals is discussed within the context of maximizing the fraction of the crystalline enzyme present as a complex with one such inhibitor, glucose alpha-1-phosphate-6-vanadate. One of these problems, achieving a substantial reduction in sulfate concentration, could not be successfully addressed by employing the desalting procedure used to produce the substrate/product complexes, because of reduced diffusional rates in the final solution.(ABSTRACT TRUNCATED AT 400 WORDS)

Characterization of a vanadate-based transition-state-analogue complex of phosphoglucomutase by kinetic and equilibrium binding studies. Mechanistic implications.

The inhibitor complex produced by the binding of alpha-D-glucose 1-phosphate 6-vanadate to the dephospho form of muscle phosphoglucomutase exhibits an unusually small dissociation constant: about 15 fM for the Mg2+ enzyme at pH 7.4, when calculated in terms of the tetraanion. Such tight binding suggests that the enzyme/vanadate/glucose phosphate complex mimics a state that at least approaches the transition state for (PO3-) transfer in the normal enzymic reaction. This hypothesis also is supported by the observation that replacement of Mg2+, the normal metal ion activator, by Li+, a poor activator, substantially reduces the binding constant for the glucose phosphate/vanadate mixed diester. Other indicators that support this hypothesis are described. One is the derived equilibrium constant for replacement of a PO4(2-) group in bound glucose bisphosphate by VO4(2-): 3 x 10(6) when the replaced group is the phosphate at the (PO3-) transfer site of the Mg2+ enzyme--in contrast to about 10 for the same replacement (of PO4(2-) by VO4(2-)) in an aqueous solution of a phosphate ester. Another is the greatly decreased rate at which Mg2+ dissociates from the glucose phosphate/vanadate complex of the enzyme, relative to the rate at which it dissociates from the corresponding bisphosphate complex (rate ratio less than or equal to 3 x 10(-4)), presumably because Mg2+ binds more tightly to the glucose phosphate/vanadate complex than to the corresponding bisphosphate complex. This apparent increase in Mg2+ binding occurs in spite of what appears to be a reduced charge density at the bound vanadate grouping, relative to the bound phosphate grouping, and in spite of the somewhat weaker binding of Mg2+ by dianionic vanadate than by the phosphate dianion. Although a direct assessment of the binding constant for Mg2+ was not possible, the equilibrium constant for Mg2+/Li+ exchange could be evaluated for the complexes of dephospho enzyme with glucose bisphosphate or glucose 1-phosphate 6-vanadate. The results suggest that the glucose phosphate/vanadate complex of the Mg2+ enzyme mimics a state about halfway between the ground state and the transition state for (PO3-) transfer. This estimate also is in accord with the binding of glucose phosphate/vanadate relative to that expected for transition-state binding of glucose bisphosphate. A possible scenario for the (PO3-) transfer catalyzed by the Mg2+ form of phosphoglucomutase is discussed, on the basis of these observations, together with possible reasons why the bound vanadate group appears to mimic an intermediate state for (PO3-) transfer rather than the ground state for phosphate binding.

Comparison of rate constants for (PO3-) transfer by the Mg(II), Cd(II), and Li(I) forms of phosphoglucomutase.

Net rate constants that define the steady-state rate through a sequence of steps and the corresponding effective energy barriers for two (PO3-)-transfer steps in the phosphoglucomutase reaction were compared as a function of metal ion, M, where M = Mg2+ and Cd2+. These steps involve the reaction of either the 1-phosphate or the 6-phosphate of glucose 1,6-bisphosphate (Glc-P2) bound to the dephosphoenzyme (ED) to produce the phosphoenzyme (EP) and the free monophosphates, glucose 1-phosphate (Glc-1-P) or glucose 6-phosphate (Glc-6-P): EP.M + Glc-1-P----ED.M.Glc-P2----EP.M.Glc-6-P6. Before this comparison was made, net rate constants for the Cd2+ enzyme, obtained at high enzyme concentration via 31P NMR saturation-transfer studies [Post, C. B., Ray, W. J., Jr., & Gorenstein, D. G. (1989) Biochemistry (preceding paper in this issue)], were appropriately scaled by using the observed constants to calculate both the expected isotope-transfer rate at equilibrium and the steady-state rate under initial velocity conditions and comparing the calculated values with those measured in dilute solution. For the Mg2+ enzyme, narrow limits on possible values of the corresponding net rate constants were imposed on the basis of initial velocity rate constants for the forward and reverse directions plus values for the equilibrium distribution of central complexes, since direct measurement is not feasible. The effective energy barriers for both the Mg2+ and Cd2+ enzymes, calculated from the respective net rate constants, together with previously values for the equilibrium distribution of complexes in both enzymic systems [Ray, W. J., Jr., & Long, J. W. (1976) Biochemistry 15, 4018-4025], show that the 100-fold decrease in the kappa cat for the Cd2+ relative to the Mg2+ enzyme is caused by two factors: the increased stability of the intermediate bisphosphate complex and the decreased ability to cope with the phosphate ester involving the 1-hydroxyl group of the glucose ring. In fact, it is unlikely that the efficiency of (PO3-) transfer to the 6-hydroxyl group of bound Glc-1-P (thermodynamically favorable direction) is reduced by more than an order of magnitude in the Cd2+ enzyme. By contrast, the efficiency of the Li+ enzyme in the same (PO3-)-transfer step is less than 4 x 10(-8) that of the Mg2+ enzyme.(ABSTRACT TRUNCATED AT 400 WORDS)

The concentrations of free Mg2+ and free Zn2+ in equine blood plasma.

The enzyme phosphoglucomutase can be used as a metal ion indicator to measure the concentrations of free Mg2+ and free Zn2+ in physiological fluids. In horse plasma, the concentration of free Mg2+ is close to 0.5 mM, whereas that of free Zn2+ is about 2 X 10(-10) M, although numerous physiological roles for Zn2+ have been postulated that would require free Zn2+ concentration orders of magnitude higher than this. A titration of plasma with Zn2+ shows that the fractional increase in free Zn2+ is essentially the same as the fractional increase in total exchangeable Zn2+, and the results are consistent with a model in which essentially all of the Zn2+ in plasma is bound to albumin. Regardless of the model, the buffering capacity of plasma for free Zn2+ is intrinsically low; however, its capacity relative to the total (exchangeable) Zn2+ present is maximal. The implications of this type of buffering for homeostasis of plasma Zn2+ are considered. Treatment of plasma with a strong reducing agent such as dithiothreitol (0.1 mM) substantially increases the apparent binding of Zn2+ and thus reduces the free Zn2+ concentration. However, the concentration of free Zn2+ appears to be insensitive to decreases in the physiological concentrations of reduced glutathione and cysteine. The concentrations of free Zn2+ and free Mg2+ in plasma are similar to those that have been reported for muscle tissue (rabbit). Their ratio is about 4 X 10(-7). The physiological implications of these concentrations are considered. In some cases, if the Zn2+ and Mg2+ complexes of an uncharacterized vertebrate protein exhibit significantly different properties, their relative importance under physiological conditions can be approximated by evaluating those of the mixed complexes present in a solution that contains the physiological concentration of free Mg2+, plus Zn2+ buffered with histidine, at the appropriate pH and ionic strength. Other metal ion/chelon systems that come close to reproducing the concentrations of free Mg2+ and free Zn2+ in horse plasma also are considered.

The effect of polyethylene glycol on the growth and dissolution rates of a crystalline protein at high salt concentration. Phosphoglucomutase.

Although low concentrations of polyethylene glycol (1-5%, w/v) are essential for growing crystals of phosphoglucomutase from ammonium sulfate solutions (at close to 50% of saturation), the observed rate constant for short-term crystal growth on a defined, microcrystalline surface is essentially independent of polyethylene glycol concentration under these conditions. But this cosolute produces a substantial increase in the observed rate constant for the dissolution process and thus a corresponding increase in the solubility of the crystalline phase. These observations can be rationalized in terms of a decrease in the thermodynamic activity of the soluble form of phosphoglucomutase at high salt due to favorable interactions with polyethylene glycol (PEG) at the protein surface, coupled with a difference in accessibility of protein surfaces in the crystalline and solution states. Surfaces with a differential exposure in these two phases likely include both groups that interact favorably with polyethylene glycol relative to water (nonpolar groups) as well as those that interact unfavorably (ionic groups), but favorable PEG-protein interactions produced on dissolution must outweigh unfavorable ones. A PEG-induced increase in protein solubility at high salt concentration is likely to be general; PEG also may affect the growth of other protein crystals at high salt concentrations as it affects phosphoglucomutase.

A simple procedure for removing contaminating aldehydes and peroxides from aqueous solutions of polyethylene glycols and of nonionic detergents that are based on the polyoxyethylene linkage.

Commercial samples of polyethylene glycol as well as detergents incorporating the polyoxyethylene linkage may contain substantial concentrations of peroxides and aldehydes, as well as aldehyde precursors. Facile procedures for essentially eliminating either the first or all three of these contaminants from concentrated aqueous stock solutions are described. Under some conditions, the purified solutions can be stored at room temperature under aerobic conditions for time periods of weeks without significant decomposition. Possible antioxidants that might be used under other conditions to stabilize such solutions during long-term aerobic studies are considered.

The complete amino acid sequence of rabbit muscle phosphoglucomutase.

The complete amino acid sequence of rabbit muscle phosphoglucomutase has been determined by isolating the 11 peptide fragments produced by the cyanogen bromide cleavage reaction and subjecting these to automated sequencing procedures. Products produced by treatment of some of these fragments with hydroxylamine, iodosobenzoic acid, mild acid, cyanogen bromide in formic and heptafluorobutyric acids, Staphylococcus aureus V8 protease, and trypsin (with or without blocking at lysine residues) were used to complete the sequence for each of the cyanogen bromide fragments. The cyanogen bromide fragments were ordered by isolating the four tryptic peptides produced by a limited tryptic digest of the native enzyme in the presence of its substrates and its bivalent metal ion activator, Mg2+, degrading these by means of trypsin, after blocking digestion at lysine residues, and isolating and identifying all fragments thus produced that contained 10 or more residues. The 561-residue sequence thus obtained is one of the longest that has been determined by chemical means. There is excellent agreement between this sequence and published compositions after appropriate normalization. The absorbance of the enzyme is about 7.0 at 278 nm for a 1% solution; this value is 9% lower than that previously used.