An Enantiocomplementary Dirigent Protein for the Enantioselective Laccase‐Catalyzed Oxidative Coupling of Phenols

(+)- or (−)-pinoresinol: that is the question. Which of the two enantiomeric lignans is formed during laccase-catalyzed phenol coupling of (E)-coniferyl alcohol (1) depends on the dirigent protein. In the presence of the first enantiocomplementary dirigent protein AtDIR6, (−)-2 is formed (78 % ee). Preferential formation of (+)-2 is observed in the presence of the dirigent protein FiDIR1, whereas only racemic 2 is formed in the absence of dirigent proteins. The oxidative coupling of phenols marks a key step in the biosynthesis of lignans, flavonolignans, and alkaloids and plays a central role in plant secondary metabolism.1 The dimerization of propenylphenol derivatives to form lignans generally imparts high regio-, diastereo-, and enantioselectivity in vivo,2 whereas in vitro enantioselectivity is negligible.3, 4 For example, after enzymatic oxidation of (E)-coniferyl alcohol (1) in vitro coupling results in the formation of only racemic pinoresinol ((±)-2). Regio- and stereoselective phenol coupling is observed not only in plant secondary metabolism but also in bacteria, lichen, and fungi.5 Examples include the regioselective formation of the isomers vioxanthin in Penicillium citreoviride and pigmentosin A in Hypotrachyna immaculate,6 and the formation of the enantiomeric perylene quinones, hypocrellin and hypocrellin A, in Hypocrella bambusae and Shiraia bambusicola, respectively.7 The rationale for the dramatically different routes taken during oxidative phenol coupling in vivo and in vitro remains an open question. In 1997, Lewis et al. showed that in the presence of a dirigient protein (DP) from Forsythia intermedia (FiDIR1), the oxidative coupling of 1 (Scheme 1) results in enantiomerically pure (+)-pinoresinol ((+)-2) as well as (±)-dehydrodiconiferyl alcohol ((±)-3) and erythro/threo-(±)-guaiacylglycerol 8-O-4′ coniferyl ether ((±)-4).8, 9 Additional DPs were subsequently found in Thuja plicata. Like FiDIR1, the T. plicata DPs lack catalytic activity but mediate the enantioselective formation of (+)-2 in the course of enzymatic oxidation of 1.2, 10 There are also indications for a protein in Linum usitatissimum that allows the preferential formation of (−)-2.11 The atropselective coupling of hemigossypol to (+)-gossypol in Gossypium hirsutum further shows that enantioselective phenol coupling in plants is not restricted to the formation of 2.12 Oxidative coupling of 1 with FiDIR1.8 Continuing studies of DPs and their effects on phenol coupling selectivity are required to achieve a better understanding of lignan biosynthesis and for the development of phenol coupling as a generally applicable tool in enantioselective synthesis. The latter aspect is of particular importance, since there is a high demand, but no general solution, for enantioselective phenol coupling in organic synthesis, even though some promising approaches exist.13 We present herein the cloning, expression, purification, and functional characterization of a DP from Arabidopsis thaliana, which mediates the laccase-catalyzed enantioselective oxidative phenol coupling of 1 to (−)-2. We further show that the enantioselectivity of the newly characterized Arabidopsis DP is opposed to that of the known Forsythia DP. In analogy to enantiocomplementary enzymes14 we would like to propose the term enantiocomplementary dirigent protein (EDP) to describe such proteins. Starting point in our search for EDPs was the recent report on a pinoresinol reductase from Arabidopsis thaliana specifically converting (−)-2 into (−)-lariciresinol.15 This finding suggested the existence of an EDP responsible for the formation of (−)-2. Data base searches identified two Arabidopsis genes, At1g64160 and At4g23690, coding for proteins highly similar to known DPs (Figure 1). Adopting the nomenclature proposed by Ralph et al., we will refer to the two proteins as AtDIR5 (At1g64160) and AtDIR6 (At4g23690), respectively.16 With an amino acid sequence conservation of 52 % with FiDIR1, the two proteins are good candidates for the enantioselective coupling of 1 to (−)-2 in Arabidopsis (Scheme 2). Sequence comparison between DPs and EDPs from F. intermedia, T. plicata, and A. thaliana. Residues conserved in all sequences are indicated in black. Sequence conservation between Thuja and Forsythia is highlighted in green and conservation within Arabidopsis sequences in yellow. Predicted N-terminal signal peptides (www.cbs.dtu.dk) are shown in italics. Oxidative coupling of 1 with AtDIR6. The open reading frame of AtDIR6 was cloned into an expression vector, transformed into cultivated plant (Solanum peruvianum) cells, and a suspension cell culture was established from the cell line exhibiting the highest AtDIR6 expression level. The secreted AtDIR6 protein was purified to apparent homogeneity by fractionated ammonium sulfate precipitation and conventional chromatographic techniques. It was obtained in a yield of 0.2 mg L−1 as five isoforms differing in the degree of glycosylation (Figure 2). The five isoforms with masses of 20.4, 20.9, 21.4, 21.9, and 22.4 kDa were deglycosylated using CF3SO3H resulting in a single protein species of 18.6 kDa. The N-terminus of this protein was identified by mass spectrometry (MS) confirming the predicted processing site of the signal peptide (Figure 1). The mass calculated for mature AtDIR6 is consistent with 18.6 kDa determined for the deglycosylated protein. Purification of AtDIR6. a) Stripe domain structure polyacrylamide gel electrophoresis (SDS-PAGE) analysis of the crude extract (A), purified AtDIR6 (B), and a size marker (L). b) MALDI-TOF-MS analysis of the purified protein prior to (blue) and after (red) deglycosylation using CF3SO3H. To investigate a potential “dirigent” (directing) activity of AtDIR6 we first analyzed the product spectrum generated during the oxidative coupling of 1 catalyzed by laccase from Trametes versicolor alone, with O2 as the oxidizing agent. In addition to (±)-pinoresinol ((±)-2) we isolated and identified (±)-3 and (±)-4. Using chiral HPLC (Chiralpak IB) the pinoresinol formed in absence of DPs was shown to be a racemic mixture of (+)- and (−)-2 (Figure 3 a). Demonstration of dirigent activity. The enantiomeric composition of 2 was analyzed by HPLC (A at 280 nm) after the reaction of 1 with a) laccase, b) laccase and AtDIR6, c) laccase and FiDIR1. d) (+)-2 standard. The spectrum of products was the same when laccase-catalyzed oxidation of 1 with O2 as oxidizing agent was carried out in presence of AtDIR6, however, the yield of 2 increased with a concomitant decrease in 3 and 4. More importantly, in the presence of AtDIR6, the oxidative coupling of 1 resulted in the preferential formation of (−)-2 (Figure 3 b). The enantiomeric excess of (−)-2 was (26.8±3.1) % ee when AtDIR6 was present at a concentration of 1.9 μM. The ee value increased to (49.2±2.1) % when the concentration of 1 was reduced and that of AtDIR6 kept constant (Figure 4 b). With increasing concentrations of AtDIR6, the enantiomeric excess could be improved to (78±3.6) % ee at 11.7 μM AtDIR6, and the maximum was not reached (Figure 4 a). Enantiomeric excess of (−)-2 as a result of varying concentrations of a) AtDIR6 and b) 1. In (a) the concentration of 1 was kept constant at 1.7 mM and in (b) the concentration of AtDIR6 was kept constant at 1.9 μM. The data clearly show that AtDIR6 is a genuine DP with a novel dirigent activity. Directing the formation of 1 to (−)-2, AtDIR6 is enantiocomplementary to known DPs. To further corroborate this finding, we also cloned and purified FiDIR1, the dirigent protein characterized by Lewis and co-workers.8 AtDIR6 and FiDIR1 were then compared with respect to their dirigent activities. During laccase-catalyzed coupling of 1 to 2 we observed opposite enantioselectivities for FiDIR1 and AtDIR6, that is, the preferential formation of (+)-2 by FiDIR1 and (−)-2 by AtDIR6 (Figure 3 b,c). With AtDIR6 and FiDIR1, two DPs with opposite enantioselectivity are now available in recombinant form enabling the enantioselective synthesis of both (+)-2 and (−)-2. During the formation of (−)-2 by AtDIR6 and (+)-2 by FiDIR1 high (E)DP concentrations are required to achieve high ee-values. It is assumed that high (E)DP concentrations are necessary because unselective coupling of free radicals competes with the directed coupling of (E)DP-bound radicals.2, 9 The mechanism of action of DPs and EDPs and, particularly, the molecular basis of their opposite enantioselectivity will be the subject of future investigations. However, first indications can be obtained from a sequence comparison between DPs and EDPs (Figure 1). A number of amino acid residues stand out that are conserved between FiDIR1 and the T. plicata DP, but differ from the Arabidopsis sequences. Since Forsythia and Arabidopsis are phylogenetically much closer, the sequence differences cannot be explained by phylogenetic distance but may rather reflect functional differences of the proteins. The respective amino acids may thus be relevant for the observed enantiocomplementarity of DPs and EDPs. The analysis and comparison of DP and EDP structures will ultimately resolve this question. On basis of structural data it may become possible to devise minimal DPs suited for the development of a universal enantioselective phenol coupling process. Detailed facts of importance to specialist readers are published as ”Supporting Information”. Such documents are peer-reviewed, but not copy-edited or typeset. They are made available as submitted by the authors. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.