Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Nov 3.
Published in final edited form as: Chembiochem. 2016 Sep 22;17(21):2000–2006. doi: 10.1002/cbic.201600345

The Thr-His Connection on the Distal Heme of Catalase-related Hemoproteins: a Hallmark of Reaction with Fatty Acid Hydroperoxides

Zahra Mashhadi [a], Marcia E Newcomer [b], Alan R Brash [a],*
PMCID: PMC5267355  NIHMSID: NIHMS843147  PMID: 27653176

Abstract

This review focuses on a group of heme peroxidases that retain the catalase fold in structure, yet show little or no reaction with hydrogen peroxide. Instead of a role in oxidative defense, generally these enzymes are involved in secondary metabolite biosynthesis. The prototypical enzyme is the catalase-related allene oxide synthase (cAOS), an enzyme that converts a specific fatty acid hydroperoxide to the corresponding allene oxide (epoxide). Other catalase-related enzymes form allylic epoxides, aldehydes or a bicyclobutane fatty acid. In all catalases, including these catalase relatives, a His residue on the distal face of the heme is absolutely required for activity. Its immediate neighbor in sequence as well as in three-dimensional space is conserved as Val in true catalases and changed to Thr in the fatty acid hydroperoxide-metabolizing enzymes. As explained herein, the Thr-His connection on the distal face of the heme is critical in switching the substrate specificity from H2O2 to the transformation of fatty acid hydroperoxide.

Keywords: allene oxide synthase, biosynthesis, catalase, fatty acid peroxides, heme proteins

Biosynthesis in the catalase fold

Catalase-related hemoproteins have a threonine on the distal heme in place of the usual valine, eliminating reaction with H2O2 and supporting a role in fatty acid hydroperoxide metabolism and secondary metabolite biosynthesis.

graphic file with name nihms843147u1.jpg

Introduction

Mechanistic studies on catalase are among the earliest and most detailed of any hemoprotein [1–6]. The early structural analyses proved that the selectivity of catalase for reactions with hydrogen peroxide or other small substrates is dependent on the highly restricted access route to the active site heme [7]. The restricted channel available for H2O2 was confirmed in many subsequent structures (e.g [8–10], and is especially well illustrated in a side view of the ~25 Å channel leading from the surface of human catalase down to the distal face of the heme [11].

With this extensive history backing up catalase structure-function, it was a complete surprise in the mid-1990s when an enzyme from coral catalyzing the conversion of 8R-hydroperoxy-eicosatetraenoic acid (8R-HPETE) to a fatty acid allene oxide was identified as a structural relative of catalase [12]. The hydroperoxide substrate and allene oxide product are derivatives of arachidonic acid, the 20-carbon fatty acid with four double bonds. In fact the catalase-related allene oxide synthase (cAOS) exists naturally as the N-terminal domain of a fusion protein with the C-terminus being the 8R-lipoxygenase (8R-LOX) that oxygenates arachidonic acid to 8R-HPETE. Both domains of the fusion protein express well as separate entities and the X-ray crystal structures were solved by 2005 [13, 14]. The cAOS domain reacts specifically with 8R-HPETE with a turnover number around 1400 s−1, yet remarkably shows no reaction with hydrogen peroxide [12, 15]. Indeed, the characteristic fizz (the evolution of O2) when weak solutions of hydrogen peroxide are exposed to catalase is completely absent with cAOS. As discussed later, a key structural feature of cAOS that prevents reaction with H2O2 is a conserved amino acid substitution on the distal face of the heme. True catalases have a Val residue next to the catalytically essential distal heme His, and this is changed to Thr in cAOS and its relatives (Fig. 1).

Figure 1.

Figure 1

Open in a new tab

Views of the heme in (A) human catalase, and (B) cAOS from the coral Plexaura homomalla.

The cAOS enzyme has a similar catalytic activity to certain cytochrome P450s that are specialized for reaction with fatty acid hydroperoxides [16–19]: both cAOS and the P450-AOS transform a fatty acid hydroperoxide substrate to a fatty acid allene oxide (Scheme 1). In the case of the P450-AOS, the substrate is a hydroperoxide of the C18 polyunsaturated fatty acids of plants and fungi [16, 20]. Allene oxides were named in the 1960s from the chemical reaction, oxidation of the allene moiety giving the eponymous allene oxide (Scheme 1A) [21].

Scheme 1.

Scheme 1

Open in a new tab

Allene oxide generation. A) Chemical epoxidation of an allene produces an allene oxide. B) In biochemistry, allene oxides are generated by enzymatic dehydration of a specific fatty acid hydroperoxide. This may be followed by spontaneous or enzyme-controlled cyclization by allene oxide cyclase (AOC) to produce a cyclopentenone. C) Key steps in the jasmonate pathway in plants involve 13S-lipoxygenase (13S-LOX), P450-type AOS, and then cyclization by AOC. D) The equivalent steps in the clavulone pathway in coral involve 8R-LOX, cAOS, and a (yet to be characterized) enzymatic cyclization to give, preclavulone A.

More unstable than most epoxides, allene oxides have a versatile chemistry, the product outcome dependent on the chemical structure, the presence of nearby reactive molecules, and on transformation by other enzymes such as allene oxide cyclase. With a second double bond in conjugation with the epoxyene moiety as in the fatty acid allene oxides, spontaneous or enzyme-controlled cyclization (allene oxide cyclase, AOC) gives a cyclopentenone (Scheme 2B), the next step in biosynthesis of jasmonic acid in plants (Scheme 2C) [22] and the prostanoid-related clavulones in corals (Scheme 2D) [17, 23].

Scheme 2.

Scheme 2

Open in a new tab

Monooxygenase activity of cAOS activated by iodosylbenzene (PhIO)

Sequence and overall structural similarities in catalase and cAOS

The X-ray crystal structure reveals the cAOS structure is a remarkable match for the core catalase fold [13]. cAOS lacks the loops and wrapping domains of catalase monomers that facilitate melding of the four monomers into the catalase homotetramer. Lacking these, the cAOS monomer is smaller in size (43 kD versus 60 kD of a human catalase monomer) and crystallizes as a homodimer. The percent amino acid sequence identity of the coral cAOS to mammalian catalase is low – about 15 – 20%. Nonetheless, the conservation of key features of the catalase amino acid sequence and the three dimensional structure establish the relationship (Fig. 2). The matches include the critical distal heme His67, the ancillary catalytic distal heme Asn137, and the proximal heme ligand, Tyr353, in the sequence context of R(x)3Y(x)6R. The distal heme His (H75 in human catalase, H67 in cAOS) is essential for any catalytic activity in true catalases and in cAOS [7, 24]. The importance of the well-conserved asparagine residue in the distal face of the heme was tested by mutagenesis in catalase (Asn 148) and cAOS (Asn 137) [24, 25]. Although the distal heme Asn mutation to Ala reduces the reaction rate of cAOS and catalase to 4% and 24% of that of the wild-type enzymes, respectively, it is not required for activity [24].

Figure 2.

Figure 2

Open in a new tab

The overall structure of a subunit of human catalase (A) and monomeric cAOS (B) with the similar β-barrel (purple), α-helixes (cyan), and the α-helix of the proximal heme ligand (amber). C) The sequence alignment of cAOS with representatives of the three clades of monofunctional catalases [48]: human catalase (Hcat), E. coli hydroperoxidase II (HPII), and Pseudomonas syringae cat-F. Conserved residues in the heme environment are marked with a red box, and distal heme His, distal Asn, and proximal heme Tyr are starred. The secondary structural elements of cAOS and human catalase are shown above their respective sequences.

Thr66 in cAOS

A closer look at the heme pocket of true catalases and comparison with those in the catalase-related hemoproteins that metabolize fatty acid hydroperoxides shows a key conserved difference in the residue adjacent to the essential distal His. This residue is typically Val in true catalases (V74 in human catalase) but changed to Thr in catalase-related hemoproteins (T66 in cAOS) (Fig. 1 and Fig. 3). T66 of cAOS is in H-bonding contact with H67 which, compared to human catalase, is seen in flipped conformation in the crystal structure [13]. Mutation of T66 of cAOS to Val has a dramatic impact on the reactivity with hydrogen peroxide. Indeed, quite obviously, whereas wild-type cAOS is completely inert to a 3% solution of H2O2, the T66V mutant elicits a brief fizz (indicative of catalatic activity with evolution of O2). The activity is 2–3 orders of magnitude less than with true catalases and is associated with almost simultaneous degradation of the heme and complete loss of the Soret band in the UV-Vis spectrum [26]. The clear inference is that the wild-type T66 of cAOS both prevents reaction with H2O2 and protects the enzyme from degradation in its presence.

Figure 3.

Figure 3

Open in a new tab

Primary structure of catalase-related proteins illustrating their comparative amino acid sequence lengths, the sequence context of the distal heme Histidine (RxV/TH), and the position of the proximal heme ligand (Tyr in the context R(x3)Y(x6)R.

The T66V mutation reduces cAOS metabolism of 8R-HPETE by 85% although allene oxide is still formed as product [26]. The reciprocal mutation in human catalase (Val74 to Thr, Ser, Met, Pro, or Ala) does not eliminate the catalatic nor peroxidatic activity (as might be expected in strict analogy to cAOS) but reduces the activities by 3-fold [27].

Spectral analysis of the heme environment in cAOS and catalase

There are distinct differences in the heme environments of cAOS and catalase, the implications of which are not completely understood. Comparison of the MCD spectra of cAOS and catalase suggested a flipped heme configuration in cAOS [28] and this was confirmed in the crystal structure [13]. In contrast with true catalases that possess 5-coordinate/high spin heme, the resonance Raman spectrum of cAOS revealed a 6-coordinate/high spin heme as a water molecule is liganded to the sixth position of the heme iron [26]. EPR and UV-Vis spectral evidence shows that wild-type cAOS has ~500–1000 fold reduced affinity for cyanide, azide or peracetic acid [28]. It is perhaps of significance that the T66 mutation to Val in cAOS results in the appearance (to a minor extent) of catalase-like features: a small proportion of 5-coordinate/high spin heme species, an 8-fold higher binding affinity for cyanide along with the modest catalatic activity, interpreted as a partly changed orientation of the His67 [26].

cAOS, the T66V mutation, and reaction with H2O2

The question of why cAOS does not exhibit catalase activity, despite the remarkable conservation of the catalase heme environment, has been addressed experimentally and computationally. In both enzymes the distal His (H67 in cAOS, H75 in human catalase) sits just above one quadrant of the heme, encircled by three amino acids: in cAOS T66, A104 and N137 and in catalase V74, S114, N148 [11, 13]. In the crystal structure of cAOS H67 is within H-bonding distance to both T66 and N137 (2.9 and 2.8 Å, respectively). In contrast, only S114 is within H-bonding distance to the distal His in catalase. To satisfy the two patterns of H-bonding, the imidazole ring must be “flipped” in one enzyme relative to the orientation in the other (Fig. 1 and Fig. 4). Tosha et al. [26] proposed that tethering of H67 by two H-bond partners in cAOS might explain the lack of reaction with hydrogen peroxide, and indeed found that the T66V mutant displayed catalase activity.

Figure 4.

Figure 4

Open in a new tab

A slice of surface renderings of catalase and cAOS to reveal the solvent accessible heme. Left: The narrow channel in catalase can help direct H2O2 towards a productive orientation with the heme (shown in green with position of the heme iron in red). The two Phe residues (khaki, top left) are virtually invariant in catalases. They rank just below the heme Tyr ligand in terms of conservation, whereas in cAOS homologues there is much more variability at these positions. His75 of catalase shows hydrogen bonding to Ser114. Right: in cAOS the channel is larger and much more open, and the heme is accessible from various angles. Notice also the “flipped” His67 in cAOS, which is hydrogen bonded to Asn137 and Thr66. The black regions represent solvent inaccessible volumes.

In contrast, De Luna et al [29] took a purely computational approach, with Molecular Dynamics (MD) simulations, to explain the lack of catalase activity in cAOS. As no structure of cAOS with fatty acid or substrate mimetic is currently available, they began by modeling the 8R-HPETE substrate into the active site, and then “pruned” it of the fatty acid moiety to leave the hydroperoxide as positioned in the putative enzyme-substrate complex. Their results suggested that the presence of a T66 ultimately affects the position of the Fe3+ in the heme such that the formation of Compound I, an obligate intermediate in the catalase reaction, is disfavored. Their modeling supports a role for T66 in proper positioning of the fatty acid peroxide, and in their view it is T66 H-bonding to H2O2 that results in the lack of catalase activity, rather than a T66 interaction with H67. It is important to note that in their MD model H67 is no longer within H-bonding distance to T66 ( >6 Å) [29]. It is not clear if this repositioning was induced by the modeled 8R-HPETE:cAOS complex, the progenitor of the H2O2:cAOS model, or only in the H2O2:cAOS simulations.

While a role for the Thr hydroxyl in obviating catalase activity has been investigated by two groups, neither approach addresses the role played by the distinctly shaped active site cavities in conferring enzyme activities. Catalase has a narrow channel that can effectively steer its small substrate directly to the reactive heme center (Fig. 4). The distinct narrowing of the catalase channel is due to the presence of Phe at amino acids 153 and 161 (1DGF, human erythrocyte catalase, Putnam et al. [11]) These amino acids are highly conserved in catalases, falling in the top 5% when ranked according to conservation by the evolutionary trace algorithm [30]. In contrast, cAOS must accommodate a bulky fatty acid hydroperoxide and accordingly has a relatively large, aqueous cavity. In the MD simulations of De Luna et al. [29], as mentioned above, the cAOS-peroxide complex was modeled directly from a computational model of cAOS with its fatty acid substrate. However, in the absence of the bulky fatty acid moiety, it is not clear if there are sufficient interactions to tether a much smaller substrate in a productive conformation, as the large, open cavity cannot impose steric constraints on H2O2.

Formation of Compound I in cAOS and activity as a monooxygenase

When true catalases react with hydrogen peroxide the ferric heme (FeIII) is oxidized to Compound I (FeV=O) and the H2O2 reduced to water; reaction with a second molecule of H2O2 then reduces the enzyme back to the ferric form with evolution of O2, all this occurring at diffusion-limited rates of reaction [31]. Although Compound I of hemoproteins is considered to be a potent oxidizing species capable of monooxygenation of alkanes and epoxidation of double bonds [32, 33], the restricted access of potential substrates to the catalase active site precludes these possibilities.

The natural reaction of cAOS in catalyzing transformation of fatty acid hydroperoxide occurs through the intermediacy of Compound II (FeiV-OH), reaction continuing with a fast electron transfer back to the ferric enzyme and release of the products, allene oxide and water [18, 34, 35]. The possibility that cAOS could be oxidized to Compound I and thus might exhibit monooxygenase activity that is precluded for true catalases was tested by using iodosylbenzene as an artificial oxygen donor, a well established oxidant used for mechanistic studies on cytochrome P450 monooxygenases [36, 37]. Indeed, when the iodosylbenzene experiment was conducted with cAOS the enzyme displayed monooxygenase activity and, moreover, with regio- and stereo-specificity that reflect the interaction with its usual substrate, 8R-HPETE [38]. The iodosylbenzene-activated cAOS oxygenated arachidonic acid to 8R-HETE and 8R,9S-epoxy-eicosatrienoic acid (8R,9S-EET), and with 8R-HETE as substrate (the hydroxy fatty acid being completely inactive for AOS-type activity), the activated cAOS selectively epoxidized the 9,10-double bond forming the 8R-hydroxy-9R,10R-trans-epoxide (Scheme 2).

cAOS has evolved to interact with the hydroperoxide 8R-HPETE and has no means to utilize molecular oxygen for oxygenase activity. Nonetheless, the results of the oxygen donor experiments with iodosylbenzene show the latent potential for monooxygenation within the catalase fold and foreshadow the remarkable catalytic versatility described below.

Range of activities with fatty acid hydroperoxides

cAOS and its close relatives are present in selected cyanobacteria and lower animals, chiefly marine invertebrates. The sequences from coral contain introns and therefore are eukaryotic and not attributable to symbiotic bacteria. The genes are not present in higher animals and higher plants. Higher plants and fungi synthesize allene oxides using cytochrome P450s with the equivalent chemical activity and utilizing species-specific fatty acid hydroperoxides as substrate.

After discovery of the first LOX-cAOS hemoprotein from the coral Plexaura homomalla [12], other members of this group with novel activities continue to be identified and characterized, Scheme 3.

Scheme 3.

Scheme 3

Open in a new tab

Reactions of catalase-related hemoproteins from coral and cyanobacteria with Thr-His in the distal face of the heme. Abbreviations: 8R-HPETE, 8R-hydroperoxy-eicosatetraenoic acid; 9R- or 12R-HPOTE, 9R- or 12R-hydroperoxy-octadecatrienoic acids; 10S-HPODE, 10S-hydroperoxy-octadecadienoic acid; LTA, leukotriene A.

As illustrated in Scheme 3, several catalase-related hemoproteins from coral have allene oxide synthase activity with 8R-HPETE [12, 39, 40]. The same catalytic activities, formation of 8R-HPETE and its conversion the 8,9-epoxy allene oxide, are detected in starfish oocytes [41], although this has yet to be connected to a gene sequence. While sequencing of the starfish genome is yet to be reported, TBLASTN searches reveal other invertebrate genomes harboring cAOS-LOX sequences including the polychaete worm Capitella imbricata, the marine sponge Acropora digitifera, the tailed mussel Lingula anatina, the leech Helobdella robusta, and there are several cAOS-LOX sequences in the lancelet Branchiostoma floridae (considered as a primitive evolutionary precursor of the vertebrates). The cAOS-LOX sequence in the unicellular planktonic alga Ostreococcus lucimarinus is the only one identified in a plant lineage. Regarding enzymatic activities identified by experiment, the cyanobacterium Acaryochloris marina has a cAOS-LOX fusion protein that forms 12R-hydroperoxy-linolenic acid and transforms this to the corresponding 12,13-epoxy allene oxide (illustrated in Scheme 3) [42].

The coral C. imbricata contains both a cAOS-LOX fusion protein, and a closely related gene that exhibits a catalase-related hydroperoxide lyase (cHPL) that cleaves 8R-HPETE to an aldehyde acid and alkyl aldehyde. This type of fatty acid aldehyde synthesis parallels the HPL activity of CYP74 cytochrome P450s in plants [18]. There is also a stand-alone cHPL enzyme in the cyanobacterium Nostoc punctiforme PCC 73102 that cleaves its substrate, 10S-HPODE (10S-hydroperoxy-linoleic acid, formed by the adjacent heme dioxygenase gene), to an aldehyde acid and alcohols [43]. This combination of heme dioxygenase and cHPL mimics a long-recognized activity in mushrooms [44, 45], the genes for which are yet to be identified; indeed, the main C8 cleavage product of the Nostoc enzymes, 1-octen-3-ol, is known by the common name “mushroom alcohol” [46]. The most spectacular enzymatic activities with fatty acid hydroperoxide are exhibited by the fusion protein of Anabaena PCC 7120: the lipoxygenase converts α-linolenic acid to its 9R-hydroperoxide and the catalase-related domain transforms this to a bicyclobutane fatty acid together with an allylic epoxide related to leukotriene A (Scheme 3) [47]. A bicyclobutane is highly strained in structure and previously unknown in the world of biology.

Conclusions and perspectives

The observations are remarkably consistent that the distal heme sequence of RxTH in the small catalase-related genes is associated with the specific metabolism of fatty acid hydroperoxides. The Thr next to the distal His has a profound effect on a lack of reactivity with H2O2 and in protecting the enzyme from hydroperoxide-mediated heme degradation. The mechanisms by which the Thr residue has these protective effects require further study. Opposite views stem from spectroscopic and computational approaches. Does the Thr residue hold the distal heme His in an unproductive conformation for H2O2 dismutation, or is its role to bind H2O2 and keep the peroxide out of contact with the heme iron? The interpretation of either is hindered by the current lack of an experimental structure identifying the mode of 8R-HPETE (or a surrogate) binding in the cAOS active site; potentially this would help corroborate the validity of the computations analyses as well as revealing a role for other residues in the cAOS active site.

Besides uncertainty in the exact role of the Thr in converting catalase activity to the metabolism of fatty acid hydroperoxide, the capacity of these catalase-related enzymes to handle different reactions is also not fully explored. Even with the limited number of catalase-related enzymes characterized, some are mimicking the reactions of cytochromes p450 (e.g. fatty acid allene oxide synthesis or hydroperoxide lyase) and others catalyzing novel reactions (e.g. bicyclobutane synthesis). It is likely that new reactions remain to be discovered. The cAOS family of heme enzymes is a vivid example of the challenge of inferring function from sequence, an important goal if we wish to effectively mine the rapidly expanding sequence databases. On the other hand, efforts to understand how small changes in a highly conserved heme active site impact catalytic activity have the potential to enable the design of catalysts to generate novel oxylipin fatty acids.

Acknowledgments

This work was supported in part by NIH grants GM-74888 and GM15431 (to A.R.B.) and by HL-107887 and NS081180 (to M.E.N.).

References

  • 1.Theorell H. The iron-containing enzymes. B. Catalases and peroxidases. In: Sumner JB, Myrbäck K, editors. Hydroperoxidases. Part 1. II. Academic Press; New York: 1951. pp. 397–427. [Google Scholar]
  • 2.Chance B. The iron-containing enzymes. C. In: Sumner JB, Myrbäck K, editors. The enzyme-substrate compounds and mechanism of action of the hydroperoxidases. 1. II. Academic Press Inc; New York: 1951. pp. 428–453. [Google Scholar]
  • 3.Chance B, Herbert D. Biochem J. 1950;46:402–414. doi: 10.1042/bj0460402. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Nicholls P, Fita I, Loewen PC. Adv Inorg Chem. 2001;51:51–106. [Google Scholar]
  • 5.Nicholls P. Arch Biochem Biophys. 2012;525:95–101. doi: 10.1016/j.abb.2012.01.015. [DOI] [PubMed] [Google Scholar]
  • 6.Diaz A, Loewen PC, Fita I, Carpena X. Arch Biochem Biophys. 2012;525:102–110. doi: 10.1016/j.abb.2011.12.011. [DOI] [PubMed] [Google Scholar]
  • 7.Fita I, Rossman MG. J Mol Biol. 1985;185:21–37. doi: 10.1016/0022-2836(85)90180-9. [DOI] [PubMed] [Google Scholar]
  • 8.Melik-Adamyan WR, Barynin VV, Vagin AA, Borisov VV, Vainsttein BK, Fita I, Murthy MRN, Rossman MG. J Mol Biol. 1986;188:63–72. doi: 10.1016/0022-2836(86)90480-8. [DOI] [PubMed] [Google Scholar]
  • 9.Maté MJ, Zamocky M, Nykyri LM, Herzog C, Alzari PM, Betzel C, Koller F, Fita I. J Mol Biol. 1999;286:135–149. doi: 10.1006/jmbi.1998.2453. [DOI] [PubMed] [Google Scholar]
  • 10.Loewen PC, Carpena X, Rovira C, Ivancich A, Perez-Luque R, Haas R, Odenbreit S, Nicholls P, Fita I. Biochemistry. 2004;43:3089–3103. doi: 10.1021/bi035663i. [DOI] [PubMed] [Google Scholar]
  • 11.Putnam CD, Arvai AS, Bourne Y, Tainer JA. J Mol Biol. 2000;296:295–309. doi: 10.1006/jmbi.1999.3458. [DOI] [PubMed] [Google Scholar]
  • 12.Koljak R, Boutaud O, Shieh BH, Samel N, Brash AR. Science. 1997;277:1994–1996. doi: 10.1126/science.277.5334.1994. [DOI] [PubMed] [Google Scholar]
  • 13.Oldham ML, Brash AR, Newcomer ME. Proc Natl Acad Sci USA. 2005;102:297–302. doi: 10.1073/pnas.0406352102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Oldham ML, Brash AR, Newcomer ME. J Biol Chem. 2005;39:39545–39552. doi: 10.1074/jbc.M506675200. [DOI] [PubMed] [Google Scholar]
  • 15.Boutaud O, Aronoff DM, Richardson JH, Marnett LJ, Oates JA. Proc Natl Acad Sci USA. 2002;99:7130–7135. doi: 10.1073/pnas.102588199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Hamberg M. J Am Oil Chem Soc. 1989;66:1445–1449. [Google Scholar]
  • 17.Tijet N, Brash AR. Prostaglandins Other Lipid Mediators. 2002;68–69:423–431. doi: 10.1016/s0090-6980(02)00046-1. [DOI] [PubMed] [Google Scholar]
  • 18.Brash AR. Phytochemistry. 2009;70:1522–1531. doi: 10.1016/j.phytochem.2009.08.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Oliw EH, Arago M, Chen Y, Jerneren F. J Lipid Res. 2016 doi: 10.1194/jlr.M068981. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Hoffmann I, Oliw EH. J Lipid Res. 2013;54:3471–3480. doi: 10.1194/jlr.M044347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Chan TH, Ong BS. Tetrahedron. 1980;36:2369–2289. [Google Scholar]
  • 22.Wasternack C. J Plant Growth Regul. 2015;34:761–794. [Google Scholar]
  • 23.Corey EJ, Matsuda SPT, Nagata R, Cleaver MB. Tetrahedron Lett. 1988;29:2555–2558. [Google Scholar]
  • 24.Gao B, Boeglin WE, Brash AR. Arch Biochem Biophys. 2008;477:285–290. doi: 10.1016/j.abb.2008.07.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Loewen PC, Switala J, von Ossowski I, Hillar A, Christie A, Tattrie B, Nicholls P. Biochemistry. 1993;32:10159–10164. doi: 10.1021/bi00089a035. [DOI] [PubMed] [Google Scholar]
  • 26.Tosha T, Uchida T, Brash AR, Kitagawa T. J Biol Chem. 2006;281:12610–12617. doi: 10.1074/jbc.M600061200. [DOI] [PubMed] [Google Scholar]
  • 27.Mashhadi Z, Boeglin WE, Brash AR. Biochimie. 2014;106:180–183. doi: 10.1016/j.biochi.2014.07.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Abraham BD, Sono M, Boutaud O, Shriner A, Dawson JH, Brash AR, Gaffney BJ. Biochemistry. 2001;40:2251–2259. doi: 10.1021/bi002121h. [DOI] [PubMed] [Google Scholar]
  • 29.De Luna P, Bushnell EA, Gauld JW. J Phys Chem B. 2013;117:14635–14641. doi: 10.1021/jp408486n. [DOI] [PubMed] [Google Scholar]
  • 30.Mihalek I, Res I, Lichtarge O. J Mol Biol. 2004;336:1265–1282. doi: 10.1016/j.jmb.2003.12.078. [DOI] [PubMed] [Google Scholar]
  • 31.Switala J, Loewen PC. Arch Biochem Biophys. 2002;401:145–154. doi: 10.1016/S0003-9861(02)00049-8. [DOI] [PubMed] [Google Scholar]
  • 32.Denisov IG, Makris TM, Sligar SG, Schlichting I. Chem Rev. 2005;105:2253–2277. doi: 10.1021/cr0307143. [DOI] [PubMed] [Google Scholar]
  • 33.Guengerich FP. J Biochem Mol Toxicol. 2007;21:163–168. doi: 10.1002/jbt.20174. [DOI] [PubMed] [Google Scholar]
  • 34.Ullrich V. Arch Biochem Biophys. 2003;409:45–51. doi: 10.1016/s0003-9861(02)00410-1. [DOI] [PubMed] [Google Scholar]
  • 35.Ullrich V, Brugger R. Angew Chem Int Ed Engl. 1994;33:1911–1919. [Google Scholar]
  • 36.Lichtenberger F, Nastainczyk W, Ullrich V. Biochem Biophys Res Commun. 1976;70:939–946. doi: 10.1016/0006-291x(76)90682-3. [DOI] [PubMed] [Google Scholar]
  • 37.Ortiz de Montellano PR. Chem Rev. 2010;110:932–948. doi: 10.1021/cr9002193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Boeglin WE, Brash AR. J Biol Chem. 2012;287:24139–24147. doi: 10.1074/jbc.M112.364216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Lõhelaid H, Järving R, Valmsen K, Varvas K, Kreen M, Järving I, Samel N. Biochim Biophys Acta. 2008;1780:315–321. doi: 10.1016/j.bbagen.2007.10.010. [DOI] [PubMed] [Google Scholar]
  • 40.Lõhelaid H, Teder T, Toldsepp K, Ekins M, Samel N. PloS one. 2014:9. doi: 10.1371/journal.pone.0089215. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Brash AR, Hughes MA, Hawkins DJ, Boeglin WE, Song WC, Meijer L. J Biol Chem. 1991;266:22926–22931. [PubMed] [Google Scholar]
  • 42.Gao B, Boeglin WE, Zheng Y, Schneider C, Brash AR. J Biol Chem. 2009;284:22087–22098. doi: 10.1074/jbc.M109.013151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Brash AR, Niraula NP, Boeglin WE, Mashhadi Z. J Biol Chem. 2014;289:13101–13111. doi: 10.1074/jbc.M114.555904. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Wurzenberger M, Grosch W. Z Lebensm Unters Forsch. 1982;175:186–190. [Google Scholar]
  • 45.Wurzenberger M, Grosch W. Biochim Biophys Acta. 1984;794:25–30. [Google Scholar]
  • 46.Hung R, Lee S, Bennett JW. Appl Microbiol Biotechnol. 2015;99:3395–3405. doi: 10.1007/s00253-015-6494-4. [DOI] [PubMed] [Google Scholar]
  • 47.Schneider C, Niisuke K, Boeglin WE, Voehler M, Stec DF, Porter NA, Brash AR. Proc Natl Acad Sci USA. 2007;104:18941–18945. doi: 10.1073/pnas.0707148104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Klotz MG, Loewen PC. Mol Biol Evol. 2003;20:1098–1112. doi: 10.1093/molbev/msg129. [DOI] [PubMed] [Google Scholar]

RESOURCES