Role of Substrate Positioning in the Catalytic Reaction of 4-Hydroxyphenylpyruvate Dioxygenase—A QM/MM Study

Abstract

Ring hydroxylation and coupled rearrangement reactions catalyzed by 4-hydroxyphenylpyruvate dioxygenase were studied with the QM/MM method ONIOM(B3LYP:AMBER). For electrophilic attack of the ferryl species on the aromatic ring, five channels were considered: attacks on the three ring atoms closest to the oxo ligand (C1, C2, C6) and insertion of oxygen across two bonds formed by them (C1-C2, C1-C6). For the subsequent migration of the carboxymethyl substituent, two possible directions were tested (C1→C2, C1→C6) and two different mechanisms were sought (step-wise radical, single-step heterolytic). In addition, formation of an epoxide (side)product and benzylic hydroxylation, as catalyzed by the closely related hydroxymandelate synthase, were investigated. From the computed reaction free energy profiles it follows that the most likely mechanism of 4-hydroxyphenylpyruvate dioxygenase involves electrophilic attack on the C1 carbon of the ring and subsequent singlestep heterolytic migration of the substituent. Computed values of the kinetic isotope effect for this step are inverse, consistent with available experimental data. Electronic structure arguments for the preferred mechanism of attack on the ring are also presented.

Protein effects in non-heme iron enzyme catalysis: insights from multiscale models

“…This was studied using structures obtained from MD simulations. [5] The QM/MM calculations highlight the contributions of the second-sphere residues to reactivity, with Thr163 stabilizing the formation of the HPPD product, while interestingly the equivalent residue in HMS stabilizes its native reaction. Although the neighboring Asn216 has no direct affect on the reaction barriers, it plays an important role in arranging the local structure of the active site.…”

“…For this reaction, a radical path was proposed from QM modeling, while the QM/MM model favored a heterolytic path. [5,24] In this case, the important difference is that the QM/MM model includes second-sphere ligands that stabilize iron-substrate electron transfer by forming stronger hydrogen bonds with the polarized metal-ligands. These ligands were not included in the original QM model, designed prior to the presence of any X-ray structure.…”

“…[4] Some structurally similar enzymes even catalyze different reactions using the same substrate. [5] A key question is therefore the importance of the protein matrix in controlling reactivity and selectivity.…”

“…For non-heme iron enzymes, a related observation is that in several cases apparently contradictory mechanistic conclusions have been drawn from cluster and QM/MM models, sometimes from the same groups using the same QM methods, [5,16,17] and sometimes in studies by different authors. [18][19][20][21] The origins of these discrepancies are considerably easier to analyze in the latter cases because they minimize the differences in how the reaction is modeled.…”

“…[17,23] In 4-Hydroxyphenylpyruvate dioxygenase (HPPD), a radical path for the formation of the hydroxyphenylacetate (HPA) product was proposed from an early cluster model, while a subsequent QM/MM study favored a heterolytic path. [5,24] These differences could be interpreted as that there are large effects of the surrounding protein, and that these effects are critical for a correct description of any enzymatic reaction mechanism. This mini-review will use these examples, as well as others, to discuss what multiscale models reveal about the role of the protein in non-heme iron enzymes.…”