Manganese(II) and substrate interaction with unadenylylated glutamine synthetase (Escherichia coli W). I. Temperature and frequency dependent nuclear magnetic resonance studies
A comprehensive study of solvent interaction with unadenylylated glutamine synthetase (E1.7) was conducted using the enzyme isolated from E. coli W. The longitudinal, (1/T1p)b, and transverse, (1/T2p)b, proton relaxation rates were measured with various enzyme samples as a function of frequency (6-48 MHz) and temperature (1-40.degree. C). With Mn(II) bound at the tight metal ion site approximately 2 H2O molecules are rapidly exchanging with bulk solvent. This number is reduced to approximately 1 in the presence of glutamine. All data were successfully analyzed according to the Solomon-Bloembergen-Morgan (SBM) scheme for dipolar relaxation of water protons interacting with enzyme-bound Mn(II). The correlation time for this process varies from 1-3 .times. 10-9 for the complexes described above. Significant contributions to the correlation time arise from both 1/.tau.m, the exchange rate for H2O molecules bound at the metal site, and from l/.tau.s, the electron spin relaxation rate for Mn(II) with the latter rate showing a frequency dependence at the magnetic field strengths used in this study. A study of Mn(II) binding to E1.7 at 25.degree. C revealed 2 classes of metal ion sites, a tight set of 1/subunit with KD = 5.0 .times. 10-7 M and a weak set of 1/subunit with KD = 4.5 .times. 10-5 M. In the presence of glutamine the affinity of the 1st site for Mn(II) was unchanged but the KD value for the weak site changed to 3 .times. 10-6 M. In E1.7 samples with Mn(II) bound at both the tight and weak metal ion sites the data are interpretable with 2 rapidly exchanging H2O molecules interacting with each bound Mn(II) ion. With saturating amounts of glutamine or of ADP or of glutamine plus ADP plus arsenate, the proton relaxation rates progressively decreased suggesting that the substrates or inhibitors used were interacting with the bound Mn(II) ions resulting in diminished solvent accessibility to these bound ions. These results are interpretable in terms of ligand substitution into the coordination sphere of the bound Mn(II) ions. This is probably the case for Mn(II) at the weak metal ion site since Hunt et al. showed that Mn(II) can bind as the Mn(II)-ADP complex to the 2nd metal ion site. Results of proton relaxation rate data on E1.7 with Mn(II) bound at both the tight and weak metal ion sites led to the conclusion that these metal ion sites are > 6 .ANG. apart. In comparison with proton relaxation rate data on fully adenylylated glutamine synthetase (E11.8) as studied by Villafranca and Wedler, the 1st tight metal ion site in E11.8 has 3 rapidly exchanging H2O molecules. Mn(II) has a weaker binding constant to E11.8 (KD .apprx. 5 .times. 10-6 M) at the pH value used in both studies and a suggestion is made that that an additional protein ligand is binding to Mn(II) in glutamine synthetase when the subunits are not adenylylated.