Ketol acid reductoisomerase (KARI), a metalo-enzyme (with NADPH as cofactor), is the second common enzyme of the branched chain amino acid biosynthesis (BCAA) pathway. This enzyme catalyzes the isomerization and reduction of α-acetolactate and α-aceto-α-hydroxybutyrate and converts these to α,β-dihydroxy isovalerate and α,-β-dihydroxy-β-methylvalerate, respectively. KARI has been widely studied as a target for the development of herbicides with two potent inhibitors, N-hydroxy-N-isopropyloxamate (IpOHA) (22 pM Ki for E.coli KARI) and 2-(dimethylphosphinoyl)-2-hydroxyacetic acid (Hoe704) (70 nM Ki for spinach KARI) being tested as experimental herbicides. To better understand the structure and
enzymatic activity of KARI a number of experiments were undertaken including, metal ion replacement and site directed mutagenesis studies. In addition to these studies, a fragment based inhibitor discovery program was also initiated using X-ray crystallography and spectrophotometric assays.
Isothermal titration calorimetry was used to determine Kd values of 0.8 μM and 4 μM for Mg2+ and NADPH for binding to rice KARI, respectively. The Kd values for both NADPH and Mg2+ are unaltered in the presence or absence of the other binding partner. The preliminary data from the metal ion binding study for Mn2+ and Co2+ shows that, Co2+ binds the enzyme most tightly followed by Mn2+ and Mg2+. Based on metal ion replacement studies, Mn2+ was found to have the highest turnover for the reduction reaction followed by Co2+ and Mg2+with each having kcat values of 290.8 min-1, 64.8 min-1 and 40.6 min-1, respectively. Mn2+, Co2+ and Mg2+ had Km values of 16 μM, 43 μM and 102 μM, respectively. A pH profile study for rice KARI in the presence of Mg2+, Co2+ and Mn2+ was undertaken. The results show that for Mg2+, two pKa values (6.9 and 9.9) were observed. However, when a similar profile was determined for Co2+, these values shifted to 7.4 and 7.8 and for Mn2+ were 6.5 and 8.7. It appears that the second pKa value is shifted the most when Mg2+ is replaced by Co2+ or Mn2+. This difference could therefore contribute to the fact that Mn2+ and Co2+ cannot carryout the isomerization reaction. The most likely candidate to attribute this pKa shift would be a water molecule coordinated to the metal ions. This is because, within the active site, there are no amino acid residues that would be likely to have an ionizable group at that pKa value. A site directed mutagenesis study performed on Arg589 indicated that, though Arg589 is important for NADPH binding it does not affect the binding of Mg2+ nor does it affect the turnover number. The Kd value for NADPH decreases four- fold when Arg589 is mutated to alanine. The catalytic efficiency decreased from 1.40 min-1μM-1 for α-acetolactate to 0.63 min-1μM-1 and to 1.0 min-1μM-1 as observed for Arg589Ala and Arg589Glu mutants, respectively. This data supports the hypothesis that the C- terminal tail is critical for clamping shut the N- and C- domains together, resulting in a more catalytically efficient enzyme.
In this study, the crystal structures of rice KARI – Mg2+ in complex with five fragments from Zenobia library, 3-aminopyridine (ZT038), 3-hydroxypyridine (ZT042), 3-aminobenzonitrile (ZT389), 2-amino-4-methylphenol (ZT398) and 3,4-diaminotoluene (ZT393) have been solved to 1.5 - 2.5 Å resolution. The results show that ZT398 is bound near the active site and the other fragments bound on the surface of the protein. The binding site of ZT038 and ZT042 is near the Lys105, a part of the protein that may be important for domain motion while ZT393 binds near the entrance of the active site. In addition, through the screening of a Maybridge fragment library, two new rice KARI inhibitors, 3-[(2-thienylthio)methyl]benzoic acid (KM02425) and
2-hydroxy-2-phenylacetic acid (JFD3933) have been identified with Ki values of 212 ± 132.4 μM and 311.5 ± 122.5 μM, respectively. To better understand the mode of binding of KM02425, the crystal structure of KM02425 in complex with rice KARI – Mg2+ enzyme was determined to 2.45 Å. The results show that this molecule binds within the enzyme’s active site. Docking studies were also undertaken and suggest an alternative mode of binding near the active site for this molecule is possible.