UC Berkeley

Lipoxygenases are a remarkable class of C–H activating enzymes, cleaving that bond with high specificity and further controlling the reaction with molecular oxygen to form a single fatty acid hydroperoxide product under certain conditions. Glycine max} (soybean) lipoxygenase-1 (SLO), a model system for this two-step reaction, is characterized by a large, nearly temperature-independent kinetic isotope effect (KIE) on the hydrogen-abstraction step. Theoretical modeling of the KIE and its temperature dependence in wild-type SLO and several mutants has provided significant insight into the mechanistic details of this enzyme. We have applied two state-of-the-art biophysical techniques to SLO. In doing so, we have tested the predictions made by the nonadiabatic full tunneling model developed by Hammes-Schiffer and coworkers about substrate–cofactor distances, the dynamics of those distances, and the mechanism by which distal residues affect those distances.

A novel application of electron–nuclear double resonance spectroscopy, a pulsed EPR technique, is used to determine not only the electron–nuclear distances but also the variance of those distances in the enzyme–substrate complex, both for wild-type SLO and a single-site mutant (Chapter 2). The experimental substrate–cofactor distance (“donor–acceptor distance”) for wild-type SLO is the same as for the mutant, which is counter to model predictions. We have also measured distances for a second substrate carbon and a global minimum distance for substrate hydrogens, which, together, allow us to orient the substrate within the active site. These constraints collectively imply a tightly compressed active site with significant van der Waals overlap and moderately distorted bond geometries, validating qualitative concepts of active-site compaction in this system. The modification of SLO and development of sample conditions for these experiments are also reported.

The enormous KIE in a double mutant of SLO, over 500, is partially explained by a crystal structure of that mutant presented herein (Chapter 3). This structure shows a much-expanded active site with no changes in the main chain and minimal changes in side-chain rotamers. The crystal structures of the Ile553 series of SLO mutants are re-examined (Chapter 4). This series shows a progressively greater temperature dependence of their KIEs as the residue size is decreased (Ile→Va→Ala→Gly); we show that this change is correlated with an expanding network of residues that occupy multiple conformers, a distinctly different result than for the double mutant.

Finally, large and temperature-independent KIEs in the wild type and a single-site mutant of a cyanobacterial lipoxygenase are reported (Chapter 5). This finding expands the generality of the large, temperature-dependent KIEs in plant, human, and fungal lipoxygenases to include a new biological kingdom.

Collectively, these results validate a particular model of lipoxygenase catalysis, illuminate the role of the remainder of the protein in causing that catalysis, and firmly establish ground-state nuclear tunneling as the essential mechanism of C–H activation by lipoxygenase.