Laccase (benzenediol: oxygen oxidoreductase, EC188.8.131.52), belongs to a family of multicopper oxidases found in bacteria, fungi, green plants, and insects1-4. The active center of laccase is composed of four inorganic copper ions that are assigned respectively to T1 and T2/T3 sites (Fig. 1). The T1 site consists of one redox-active Cu2+ ion which is reduced to Cu1+ after accepting an electron from the substrate donor, and then is turned back to thermal-stable Cu2+ after delivering the electron to downstream acceptor; The T2/T3 sites contain 3 Cu clusters that reduce O2 to water after accepting four electrons from T1 in steps1-3, 5. The redox changes of T1 and T2/T3 Cu ions coupled with electron transfer within laccase has been elucidated1-3, 5, 6.
Because of the high redox potential, laccase-catalyzed transformations have recently become an active research field in chemistry7-10 and biology11-14. For example, an ever-increasing focus on laccase mediated reaction has been shown in oxidation, polymerization and adduction reactions1, 2, 15-17. These processes are always involved with the cleavage and formation of chemical bonds and the electron transfer. Unfortunately, the substrate-dependent mechanism of substrate transformation in T1 site is seldom studied. This is thus unclear when compared to the well-established electron transfer pathway from T1 to O2 via T2/T3 sites1-3, 5. As for the potential application and protein designment, the mechanism must be elucidated experimentally. Moreover, the product/intermediates released from Cu-T1 are more important than the O2-to-water transformation at T2/T3 sites. This question was addressed herein with catechol and analogs as laccase substrate.
Among laccase′s substrates, phenols are of importance because of their uses as the building blocks in industry, agriculture, medicine, synthesis and polymer chemistry18. Catechol (o-dihydroxybenzene), a typical derivative of phenol, plays a key role in physiological activities18-21. A number of substrate analogs, m-and p-dihydroxybenzene (resorcinoland hydroquinone), 3-or 4-methyl catechol, 3-or 4-nitroso catechol, and 2, 3-dihydroxynaphthalene, were also used to study the mechanism clearly.
Laccase (EC184.108.40.206, > 10 U∙mg-1) is purchased from Sigma-Aldrich. Catechol (99%) and analogs in Scheme 1 are purchased from J & K Scientific. Trapping reactant, 5, 5-dimethyL-1-pyrroline-N-oxide (DMPO) (> 99.0%), is from Dojindo International. The reaction solution with double stilled water contains laccase and catechol or its analogues with the final pH~6.3. After 5-or 15-min reaction, 200 μL of mixture was sampled for the following experiments. Analogs (Scheme 1), as well as m-and p-dihydroxybenzene (resorcinoland hydroquinone), 3-or 4-methyl catechol (98%), 3-or 4-nitroso catechol (98%), and 2, 3-dihydroxynaphthalene (98%), were used as controls. Ethyl acetate was high performance liquid chromatography (HPLC) grade and purchased from Tianjin Guangfu fine chemical research institute.
After 5-or 15-min procedure reactions at room temperature, 200 μL of solution contains laccase (0.5 mg∙mL-1) and catechol or its analogues (0.5 mg∙mL-1) was sampled, and then stored quickly in liquid nitrogen until electron paramagnetic resonance (EPR) measurement. The freeze-trapping EPR spectra were recorded with a Bruker X-band EMX EPR spectrometer at 20 K.
For spin-trapping, 200 μL of reaction mixture has laccase (0.25 mg∙mL-1), catechol or phenol derivative (0.25 mg∙mL-1), DMPO (100 mmol∙L-1). The reaction was preceded at room temperature about 3-4 min, then trapping reagent DMPO was added swiftly to trap the radicals for EPR measurements (Scheme 2) at room temperature.
HPLC-MS (mass spectrometry) was performed with the Waters Xevo G2 QTof. The catalytic reaction containing 1 mg∙mL-1 laccase and the corresponding substrates were preceded about 5 min at room temperature. The sampled mixture was filtered first through a 0.45-μm filter membrane before injection. The HPLCMS settings were: mobile phase, acetonitrile/water (20%/80%); flow velocity, 0.3 mL∙min-1; inject volume, 2 μL. The other settings are: scanning mode, anion mode; ion source, electrospray ionization (ESI). The actual mass is to be added by one proton unit (1.0078).
The 1H NMR experiments were performed with the (400-MHz) Bruker AVANCE Ⅲ 400 NMR spectrometer. The samples were prepared as follows. Aliquots of catalysis (100 mL in 500-mL Erlenmeyer flasks) were prepared using 50 μg∙mL-1 catechol and 0.2 mg∙mL-1 laccase. After 30 min procedure reactions in an orbital shaker maintained at 200 r∙min-1 and 30 ℃, products were extracted 3 times with the same volume of precooled ethyl acetate (4 ℃). Then the products were purified by silica gel column chromatography and C-18 reversed-phase chromatography. Extracts were evaporated to dryness using an Eppendorf Vacufuge (Eppendorf Scientific, Westbury, NY) at 30 ℃. The sample was then subjected to 1H NMR analysis.
Generally, the oxidase-catalyzed reactions are always involved with the short-lived free radicals during the cleavage and re-arrangement of the activated bonds22-24. To trap these transient radical species, the reactions are stopped either by the in situ swiftly freezing in cryogenic liquid nitrogen (77 K), or by roomtemperature spin trapping using DMPO. The latter allows structurally configuration of the radical species through the isotropic hyperfine constants (hfcs, indicated by a) of α-nitrogen and β-hydrogen (i.e., aN and aHβ)25, 26.
Fig. 2a displays the time-course change of EPR signal stabilized by freeze trapping. A broadened and strong signal was observed after 5 min of reaction, and the signal decreased dramatically when the reaction was elongated to 15 min. This EPR signal has a symmetrical feature at g~2.004 showing a 7-G peak-to-trough width. It is assigned to the o-benzosemiquinone radical, the direct product of catechol oxidation. The continuous decreasing in EPR signal indicates that the o-benzosemiquinone radical was depleted in the sequential reaction. In the meantime, Q-Tof mass spectrometry (i.e., HPLC-MS) was performed to detect the products. Fig. 2(b, c) shows that substrate catechol (C6H6O2) was converted dominantly to the dimer (C12H8O4 and C12H10O4) and partly to the trimer (C18H12O6) (see mechanism in Scheme 3 below). Evidently, this oligomerization removes the o-benzosemiquinone radical, thereby abolishing the corresponding EPR signal (Fig. 2a). Fig. 2e shows a new absorption (marked by an arrow) observed at~5.8 min with identical molecule weight to substrate catechol (at~2.2 min). The lag effect is ascribed to its weak polarity compared with that of catechol. Under the identical condition, oligomers and the lagging signal were not observed in the parallel experiments with two isomers (m-and p-dihydroxybenzene), as well as in the parent controls with the absence of either catechol or laccase (Fig. 2b). Therefore this lagging signal indicates isomerization of the o-benzosemiquinone radical, i.e., it suggests that the formed isomer is also a radical species.
With room-temperature spin trapping protocol, two radical species were trapped in situ by DMPO (Fig. 3a). ⅰ) For DMPO-∙OH adduct radical, the isotropic hyperfine coupling constants (hfcs) of α-nitrogen and β-proton characterized by aN=aHβ=14.8 G lead to the four-line (1 : 2 : 2 : 1) spectrum; ⅱ) For carbon-centered DMPO adduct radical, aN=15.6 G and aHβ=22.8 G were obtained from the six-line hyperfine pattern. In the parallel experiments, these two transient radicals were not observed by replacement of catechol with m-or p-dihydroxybenzene isomers (Fig. 4). In order to clarify the origin of ∙OH radical, the 17O isotope-labeling solvent water (H217O, ~80% (atomic fraction) 17O) was used. Experimentally, the four-line EPR signal of the DMPO-∙16OH radical (16O, 99.757% nature abundance) was replaced by a weaker signal, due to the further hyperfine splitting of 17O (I=5/2) of DMPO-∙17OH radical (aN=aHβ=14.8 G, and a17O=4.6 G). Moreover, removal of the DMPO-∙16OH radical signal by 17O isotope unambiguously demonstrated that the other oxygencentered radicals (e.g., peroxide, acyl, aroyl or alkoxyl radicals) were not formed. Notably, the o-benzosemiquinone radical could not be trapped by DMPO. Obviously, solvent water was a substrate upon the oxidative reaction, which has not been reported so far. Simultaneously, the six-line carbon-centered signal was prominently protruded because of the elimination of the DMPO-∙ 16OH radical after adding 17O-labeled water. The remaining question lies with whether the carbon-centered radical is aromatic or non-aromatic species originating from catechol oxidation? We rule out the former which gives aN and aHβ values larger than those of the latter25, 26. Since the other oxygen-centered radicals were not observed experimentally, these two radicals were probably not released from the intra-(at C1-C2 site) or extra-diol (at C1-C6 site) cleavages which was similar to those in catalysis by catechol dioxygenase or in chemical oxidation20, 21, 27-32. Therefore, the trapped non-aromatic carbon-centered radical was assigned to the isomer of the o-benzosemiquinone radical, which was also detected in HPLC-MS (Fig. 2e). To collect the intermediates, the catalyzed reaction was stopped after 30 min by adding ethyl acetate for extraction. The crude extraction by chromatography was subjected to 1H-NMR measurement in d6-DMSO. Fig. 3b shows that a new group of 1H signal in the extraction was appeared at δ~6.0 at the expensed of those (δ~6.7) of substrate catechol, and this novel 1H signal was assigned to the oxole derivant (i.e., the furan derivative)33-36.
For the uncoupled o-benzosemiquinone radical, the isotropic hfcs of aH4, 5=3.56 G and aH3, 6=0.9 G are known for two sets of two equivalent protons18, 24. The larger hfc of aH4, 5 indicates that a large positive spin density is localized at C4, 5 and it is negative at H4, 518, 24. Under the self-oxidized conditions, the oligomeric or adductive products were not detected. Obviously, the formed o-benzosemiquinone radical alone does not yield any adduct product. When adding laccase, a chelate complex was formed between the o-benzosemiquinone radical and the counterion Cu1+/ Cu2+ (T1)19-21, 27-32. The inductive effect caused by Cu ion activates the cation radical tendency at the C4-C5 site. This is proved by the dimer and trimmer formed efficiently from the inter-molecular adduct on the C4-C5 site37, 38. In particular, the trapped non-aromatic carbon-centered radical is probably an isomer derived from the intramolecular adduct of the o-benzosemiquinone radical. To test the hypothesis, analogues of catechol were surveyed comparatively. In Fig. 4, when isomers (m-and p-dihydroxybenzene) were oxidized by laccase, neither the radicals trapped by DMPO nor the oligomers were observed, as well as 3-or 4-nitroso catechol, and 2, 3-dihydroxynaphthalene. However, formations of DMPOtrapped radicals and dimerization were not affected in the presence of 3-or 4-methyl catechol (Fig. 1(d, f) and Fig. 4). Again, none of the other oxygen-centered radicals was observed in the parallel experiments.
To describe the laccase-catalyzed inter-and intra-molecular adducts of catechol, a mechanism is proposed in Scheme 3, with comparison to the well-established mechanism of catechol dioxygenase20, 21, 27-30. Upon catalysis, Cu2+(T1) is reduced firstly to the metastable Cu1+ (T1) and catechol is oxidized to the o-benzosemiquinone radical. Chemical turnover from Cu1+(T1) to Cu2+(T1) is accomplished after delivering an electron to the downstream. At T2/T3 sites, O2 is reduced to H2O after accepting four electrons from T1 (Fig. 1). Interaction between the incipient o-benzosemiquinone radical and the counterion Cu2+ (T1) yields a chelate intermediate as b in Scheme 3. Cu2+ in the chelate induces further spin density transfer such that the nascent cation radical character at the C4-C5 site is fulfilled actively. The intermolecular addition reaction from c to d is the major pathway which can be regarded as the mimic bio-synthesis of biological macromolecules, e.g., lignin. Inevitably, the C4-C5 bond is also attacked by the intramolecular hydroxyl group during the transition from e to g via f (depicted as the minor pathway in Scheme 3). The intermediates f and g were characterized by EPR and 1H-NMR respectively (Fig. 3). The intramolecular adduct gives to the sidechain carboncentered furan-derivative radical when the cleavage of C4-C5 bond occurs. This ring-opening is accomplished with synergy of substrate water that is ascertained when the origin of hydroxyl radical is identified by 17O-labeling water (H217O). Simultaneously, Cu2+(T1) is reduced to Cu1+(T1) once again. To further determine the inductive effect of spin density transfer, some derivatives of catechol are applied as substrates. Formations of the DMPOtrapped radicals and dimer are not affected in the presence of 3-or 4-methyl catechol substrates (methyl group causing the electron-donating inductive effects). Contrarily, these reactions are prohibited in the presence of 3-or 4-nitro catechol (-NO2 group causing the electron-withdrawing inductive effects), as well as in the presence of 2, 3-dihydroxynaphthalene. When formation of the chelate complex is prevented in the presence of m-and p-dihydroxybenzene, all these results are not observed any more.
This mechanism reminds us of another oxidation mechanism of catechol catalyzed by dioxygenase. Although the protein ligation to Cu2+(T1) is more or less similar to that around Fe2+ in catechol dioxygenase19-21, the inductive effects of spin density transfer on the substrate oxidation are distinguished from each other. For dioxygenase, catechol and O2 substrates binding and ring-opening are carried out simultaneously in adjacent position19-21. The distinct catalytic mechanisms of catechol dioxygenase and laccase provide the first-hand evidence for further understanding and utilization of Fe/Cu co-catalysis and synergy39. These comparative results show that the radical transformation can be mimicked if the electronic structure between substrate intermediate (s) and catalytic ion center can be adjusted biologically or chemically. In addition to the aromatic building block for oligomerization, the transient furan-derivative radical might have potential application in flavor chemistry, or phenols′ antioxidation and interaction of multicopper oxidase in vivo with degenerative diseases such as cancer, heart disease, inflammation and even ageing18. This also sheds light on the further comprehension of laccase-catalyzed transformation in PAHs biodegradation, green chemistry, and flavor industry.
Phenol oxidation and transformation catalyzed by laccase was studied mechanistically with mass spectrometry and EPR. The data shows that the chelate interacting between Cu2+(T1) ion and o-benzosemiquinone radical plays a pivotal role to direct the subsequent inter-or intra-molecular adducts, i.e., oligomerization or isomerization correspondingly, when the electronic structure of the chelate was altered purposely by the different modified catechol analogs. The proposed mechanism is helpful to understand the distinct catalysis of Cu and Fe that induce different electron spin transfer in the short-lived intermediates. This also shows that the radical transformation can be mimicked if the electronic structure between substrate intermediate (s) and catalytic ion center can be adjusted biologically or chemically to further exploit the utilization of laccase.