Experimental Correlation of Substrate Position with Reaction
Outcome in the Aliphatic Halogenase, SyrB2
Ryan J. Martinie,† Jovan Livada,† Wei-chen Chang,† Michael T. Green,† Carsten Krebs,†,‡
J. Martin Bollinger, Jr.,†,‡ and Alexey Silakov*,†
Departments of †Chemistry and of ‡Biochemistry and Molecular Biology, The Pennsylvania State University, University Park,
Pennsylvania 16802, United States *S Supporting Information
ABSTRACT: The iron(II)- and 2-(oxo)glutarate-dependent (Fe/2OG) oxygenases catalyze an array of challenging transformations, but how individual members of the enzyme family direct different outcomes is poorly understood. The Fe/ 2OG halogenase, SyrB2, chlorinates C4 of its native substrate,
L-threonine appended to the carrier protein, SyrB1, but hydroxylates C5 of L-norvaline and, to a lesser extent, C4 of Laminobutyric acid when SyrB1 presents these non-native amino acids. To test the hypothesis that positioning of the targeted carbon dictates the outcome, we defined the positions of these three substrates by measuring hyperfine couplings between substrate deuterium atoms and the stable, EPR-active iron−nitrosyl adduct, a surrogate for reaction intermediates. The
Fe−2H distances and N−Fe−2H angles, which vary from 4.2 Å and 85° for threonine to 3.4 Å and 65° for norvaline, rationalize the trends in reactivity. This experimental correlation of position to outcome should aid in judging from structural data on other
Fe/2OG enzymes whether they suppress hydroxylation or form hydroxylated intermediates on the pathways to other outcomes. ■ INTRODUCTION
Mononuclear1−4 and dinuclear1,5,6 nonheme iron enzymes, cytochromes P450,7 and radical SAM enzymes8,9 activate and functionalize inert C−H bonds with a remarkable degree of specificity and selectivity. Members of the iron- and 2(oxo)glutarate-dependent (Fe/2OG) oxygenase family catalyze a variety of transformations at unactivated carbon centers, including hydroxylation, desaturation, cyclization, stereoinversion, and halogenation; these reactions play crucial roles in microbial metabolism and biosynthesis,2 as well as oxygen and body mass homeostasis,10−12 DNA repair,13−15 epigenetic inheritance, and control of transcription in humans.16−18 The mechanistic strategy employed by this family was first elucidated for the hydroxylases.4 These enzymes activate oxygen at their common Fe(II) cofactor, which is coordinated by a (His)2(Glu/Asp)1 “facial triad” ligand set, 19,20 to form a high-spin (S = 2) Fe(IV)−oxo (ferryl) intermediate.21−24 This ferryl unit abstracts a hydrogen atom (H•) from the substrate,25 yielding an Fe(III)−hydroxo/substrate−radical intermediate; this radical then couples with the hydroxo ligand (formally
HO•), producing the hydroxylated product and an Fe(II) complex.4,26,27 Current understanding of the mechanisms employed by other Fe/2OG oxygenases to direct this potent reactivity to only one of several alternative reaction outcomes (e.g., halogenation, stereoinversion, etc.) is incomplete. A robust understanding of this control is a major unmet challenge and will be required for these systems to be exploited for potential biotechnological applications, including production of new drug compounds.
The Fe/2OG aliphatic halogenases provide an ideal system to study enzymatic discrimination between accessible reactivities. SyrB2 from Pseudomonas syringae B301D is the founding member of the Fe/2OG aliphatic halogenases.28−31
It catalyzes chlorination of the C4 position of L-threonine appended via a thioester linkage to the phosphopantetheine arm of the companion aminoacyl carrier protein, SyrB1 (hereafter, all L-aminoacyl-S-SyrB1 substrates are abbreviated by designating only the appended amino acid in boldface type, e.g., Thr; Figure 1). In SyrB2, the sequence position that normally provides the carboxylate of the canonical facial triad of protein ligands is occupied by an alanine (Ala118), and the cosubstrate, chloride (Cl−), occupies the vacated site in the iron coordination sphere.32 Fe/2OG halogenases mechanistically parallel the hydroxylases in that both employ ferryl intermediates as the H•-abstracting species.33−35 However, following this step, the reactivities diverge: in the halogenases it is the Cl• ligand, rather than the HO•, that couples with the substrate radical. Thus, halogenases generate the substrate radical in a manner similar to the hydroxylases but are faced with the more difficult challenge of directing it to chlorination rather than hydroxylation. Efficient H• transfer to the ferryl would seemingly imply proximity to the hydroxo ligand in the subsequent step, and HO•/substrate radical coupling occurs readily in the hydroxylases. Yet, the native substrate of SyrB2,
Thr, is almost exclusively chlorinated (Figure 1),36 although
Received: March 31, 2015
Published: May 12, 2015
Article pubs.acs.org/JACS © 2015 American Chemical Society 6912 DOI: 10.1021/jacs.5b03370
J. Am. Chem. Soc. 2015, 137, 6912−6919 non-native substrates undergo hydroxylation as well as chlorination. Therefore, SyrB2 represents an intriguing case in which two different reaction outcomes catalyzed by this enzyme family (hydroxylation and halogenation) are observed, making it an ideal system for investigating how the enzymes discriminate between reactivities.
Our previous work showed that the cis-chloroferryl complex in SyrB2 reacts more rapidly with SyrB1 presenting Laminobutyric acid (Aba) or L-norvaline (Nva) than with Thr, suggesting a programmed inefficiency in H• abstraction from the native substrate. Formed with Nva, the complex decays 130-fold more rapidly (9.5 s−1 at 5 °C) than with Thr (0.07 s−1); with Aba, the rate is intermediate (0.9 s−1) (Figure 1).36
Selectivity for chlorination is also strongly modulated: Thr is almost exclusively chlorinated, Aba is chlorinated and hydroxylated at C4 to similar extents, and Nva is predominately hydroxylated at the C5 position.36 It was reasoned that the decrease in ferryl decay rates could most simply arise from an increase in the distance between the H• donor and acceptor.