The ethylene-forming enzyme (EFE) from the plant pathovar, Psuedomonas syringae pv. phaseolicola PK2 (PDB ID: 5V31), a member of the mononuclear Fe(II)- and 2-oxogluatrate (2OG)-dependent oxygenase superfamily, catalyzes two reactions (Figure 1) (1). The first, and minor, of these reactions is the C5 hydroxylation of L-arginine (L-Arg), driven by coupling to the oxidative decarboxylation of 2OG to form succinate and carbon dioxide (CO₂) (1). Hydroxylation reactions such as this one are characteristic of Fe(II)- 2OG-dependent oxygenases (2).  The second, and major, reaction catalyzed by EFE is the decomposition of 2OG into ethylene and three CO₂ molecules (1). This reaction is unique to EFE among the Fe(II)- and 2OG-dependent oxygenases ­(1).

Ethylene is biologically most prevalent as a plant steroid that plays a role in growth and maturation (3). The significance of ethylene production in bacteria such as Psuedomonas syringae is still unclear, though it most likely contributes to the pathogenicity of these bacteria in plants (4). In addition to its biological role as a plant steroid or disease agent, ethylene is widely used in a variety of industrial reactions such as the polymerization of plastics, most notably low-density polyethylene (LDPE) and high-density polyethylene (HDPE) (1,5). The primary means of obtaining ethylene is through thermal cracking of petroleum hydrocarbons with steam, which has a negative environmental impact due to the release of greenhouse gases (5). In plants, the synthesis of ethylene also produces toxic cyanide, which limits the use of the plant ethylene-production pathway as a biosynthetic agent (3). Use of biomass expressing EFE, which only utilizes and produces non-toxic substrates, has been proposed as a potential, alternative “green” method of ethylene production (1).

EFE is a 117.15 kDa, 352 residue monomer with an isoelectric point of 5.49 (6). The secondary structure is 31% helical and 19% β-sheet (1). Like all Fe(II)- 2OG-dependent oxygenases, the core structural motif of EFE is a double-stranded β-helix (DSBH) (7). The core of the DSBH is composed of a β-sandwich comprised of two anti-parallel β-sheets (7). The major sheet is comprised of 6 β-strands and the minor sheet is comprised of 3 β-strands (1). The β-sandwich is stabilized by 10 α-helices (1). The cavity between the major and minor sheets of the DSBH serves as the binding site for the ligands involved in the reactions: Fe(II), 2OG, and L-Arg (1). EFE has been crystallized in complex with nickel in place of iron (PDB ID: 5V2V), in complex with manganese (in place of iron) and L-Arg (PDB ID: 5V31), and in complex with manganese (in place of iron), L-Arg, and 2OG (PDB ID: 5V2Y) (1). The metal binding site of EFE is a 2-histidine-1-carboxylate facial triad typical of Fe(II)- and 2OG-dependent oxygenases, in which the manganese ion is hexacoordinated to His-189, Asp-191, and His-268 from the β-8 sheet  (1,5). Binding of the metal ion induces residues 80-93 to close over the cavity like a “lid,” shielding the active site from the solvent (1). In the absence of L-Arg, 2OG coordinates with the metal center in a monodentate fashion with its C1 carboxyl group (1). Upon the binding of L-Arg to the active site, the coordinated metal center is able to bind the substrate 2OG bidentate (1). This bidentate coordination, through the C1 carboxyl group and C2 carbonyl of 2OG, is the typical binding mode found in 2OG-dependent oxygenases (9). The open configuration of the active site established by the bidentate binding of 2OG to manganese allows the coordination of a molecule of dioxygen (O₂) to the metal ion as well (1).

L-Arg binds near, but not to, the metal ion in a hydrophilic pocket lined by Glu-84, Val-85, Thr-86, Arg-316, Asp-191, Tyr-192, and Cys-317 (1). Asp-84, Val-85, Thr-86, Asp-191, Tyr-192, and Arg-316 all hydrogen bond to L-Arg, while Cys-317 has Van der Walls interactions with L-Arg (1). To form a hydrogen bond with L-Arg, Asp-191 alters its orientation in space, which induces a change in the configuration of the metal ion (1). This change in manganese configuration allows for the bidentate binding of 2OG to the manganese (in contrast to the unusual monodentate binding in the absence of L-Arg) (1). Additionally, Tyr-192, the residue adjacent to Asp-191, undergoes a significant conformational change to accommodate its own hydrogen bonding to L-Arg (1). This creates a twisted peptide bond between Asp-191 and Tyr-192, in which the peptide bond is twisted an average of 33.2º from its ideal bond angle of 180º (1).

twisted peptide bond between Asp-191 Phe-283 shifts closer to the metal ion in the active site (1). The phenylalanine residue may hinder flipping of the ferryl group, which results in the aberrant hydroxylation of L-Arg, producing ethylene (1).

Anthocyanidin synthase (PDB ID: 2BRT, ANS) was identified as a protein that is very closely related to EFE with a PSI-BLAST E-score of 1e-32 and a Dali Z-score of 26.0 (12,13). Like EFE, ANS is a Fe(II)- and 2OG-dependent oxygenase (14). ANS catalyzes the oxidation of trans-dihydroquercetin to quercetin, a metabolite involved in the synthesis of the anthocyanidin class of flavonoids (14). Flavonoids are a variety of pigments which are responsible for the vibrant colors in many plants and have a variety of health benefits as inhibitors of cell proliferation, antimutagenics, antimicrobials, anti-inflammatory agents, as well as a variety of other biomedicinal properties (14). ANS, as an archetypal Fe(II)- and 2OG-dependent oxygenase shares a variety of structural features with EFE (1,14). Like EFE, ANS contains a double-stranded beta helix (DSBH) involving 8 of the structure’s 13 β-strands (14). Residues 309-333 compose the alpha-helical “lid” that shield’s the enzyme’s active site (14). ANS also features the characteristic 2-histidine-1-carboxylate facial triad, in which His-232, His-288, and Asp-234 are hexacoordinated with the metal (14). 2OG also interacts in a bidentate manner with the metal ion (14).

Instead of the tyrosine involved in the twisted peptide bond with the aspartic acid (coordinated to the metal center found in EFE), the residue adjacent to the aspartic acid coordinated to the metal center in ANS (Asp-234) is a valine (14). Because the valine side chain cannot hydrogen bond to the substrate dihydroquercetin (analogous to the L-Arg present in EFE), there is no twisted peptide bond in the active site of ANS (14). This supports the potential role of a twisted peptide bond in the unique ethylene-producing enzymatic activity of EFE. The other structural feature of EFE that was preposed as relevant in ethylene production was a phenylalanine residue in the active site that potentially caused steric hindrance during the “ferryl flip” mechanism of the hydroxylation of L-Arg (1). The structure of ANS may weaken this proposed mechanism for ethylene formation, as ANS also features a phenyalanine residue in its active site (Phe-304) (14). The molecular distance between this Phe-304 in ANS from the iron metal ion is 4.7 Å, only slightly larger than the distance between the analogous Phe-283 and magnanese in EFE of 4.67 Å (15). The presence of this Phe-304 does not hinder enzymatic activity of ANS; rather, Phe-304 stabilizes the binding of the substrate dihydroquercetin (analogous to naringenin in this structure) through pi-stacking interactions (14). This suggests that the Phe-283 in EFE may not play a role in the unique ethylene-forming activity (14).

EFE's potential as a green biosynthetic source of ethylene justifies further study of this protein. A detailed understanding of EFE's structure and how this structure differs from other Fe(II)- and 2OG-dependent oxygenases is crucial to understanding, and potentially optimizing, its mechanism of ethylene formation.