Phosphorus nuclear magnetic resonance studies of phosphoglucomutase and its metal ion complexes.

The ³¹P nuclear magnetic resonance of the covalently bound phosphate group at the active site of phosphoglucomutase has been examined by means of Fourier transform nuclear magnetic resonance spectroscopy. At a pD of 7.9, the chemical shift of the ³¹P nucleus is 3.8 ± 0.1 ppm downfield from 85% H₃PO₄; this shift is close to that of phosphoserine (dianionic form). Proton decoupling experiments suggest that the phosphorus of the enzymic phosphate group is coupled to protons with chemical shifts similar to those of phosphoserine. In D₂O, with proton decoupling, the ratio of the longitudinal and transverse diamagnetic relaxation times in solutions of 1.6 mm phosphoenzyme yields an approximate correlation time of 10^?7s for the ³¹P nucleus of the enzyme. This is within the range of values expected for tumbling of the entire protein molecule and suggests that the covalently attached phosphate group is immobilized or “frozen” at the active site of the enzyme by means of noncovalent interactions with adjacent groups. Consistent with this, the pK_a of the enzymic phosphate is significantly lower than that of phosphoserine. Binding of the diamagnetic activator, Mg²⁺, causes little or no change in the chemical shift of the resonance of the enzymic phosphorus from pD = 5.3 to 7.6, a downfield shift (?0.5 ± 0.1 ppm) at pD = 8.6, but an upfield shift (0.8 ±0.1 ppm) for that of phosphoserine, suggesting that bound Mg²⁺ is not coordinated to the enzymic phosphate. Independent evidence against direct coordination is provided by the paramagnetic effects of Ni²⁺ bound at the active site on the relaxation rates of the enzymic phosphorus. By assessing the paramagnetic effect of bound Ni²⁺ on both the longitudinal and transverse relaxation rates of the observed resonance, and by using correlation times determined for water proton relaxation induced by the Ni²⁺ complex, a range of Ni²⁺ to phosphorus distances of 4 to 6 Å is calculated. These distances suggest a second sphere interaction between the enzyme-bound metal and the enzymic phosphate group. Bound Ni²⁺ also markedly decreases the integrated intensity of the ³¹P resonance. Although the reason for this intensity decrease is incompletely explained, the present data establish the close proximity of the bound metal ion and the active site phosphoserine on phosphoglucomutase.